[kernel] r12347 - in people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian: . config-update patches-update/features/all/xen patches-update/series
Bastian Blank
waldi at alioth.debian.org
Sat Oct 18 14:50:00 UTC 2008
Author: waldi
Date: Sat Oct 18 14:44:52 2008
New Revision: 12347
Log:
* debian/changelog: Update.
* debian/config-update/defines: Set abiname to xen-3.3-1.
* debian/patches-update/series/1-extra: Update patches.
* debian/patches-update/series/13etch4-extra: Remove.
* debian/patches-update/features/all/xen/console-hvc-overtake.patch,
debian/patches-update/features/all/xen/remove-4gb-warning.patch: Update.
* debian/patches-update/features/all/xen/xen-748f324a4b2d62d89fe40c4aa52861977e1a2cae.patch:
Add.
* debian/patches-update/features/all/xen/xen-3.1-15467.patch: Remove.
* debian/rules.real: Set correct location of vmlinuz file.
Added:
people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/xen-748f324a4b2d62d89fe40c4aa52861977e1a2cae.patch (contents, props changed)
- copied, changed from r12345, /people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/xen-3.1-15467.patch
Removed:
people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/xen-3.1-15467.patch
people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/series/13etch4-extra
Modified:
people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/changelog
people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/config-update/defines
people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/console-hvc-overtake.patch
people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/remove-4gb-warning.patch
people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/series/1-extra
people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/rules.real
Modified: people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/changelog
==============================================================================
--- people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/changelog (original)
+++ people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/changelog Sat Oct 18 14:44:52 2008
@@ -1,10 +1,11 @@
-linux-2.6.18-xen-3 (2.6.18.dfsg.1-23+xen.1) kernel-waldi-xen-extra; urgency=low
+linux-2.6.18-xen-3.3 (2.6.18.dfsg.1-23+xen.1) kernel-waldi-xen-extra; urgency=low
- * Update Xen patch to 3.1 branch, 3.1.1 release.
- * Allow overtaking of hvc for the console.
+ * Update Xen patch to revision 748f324a4b2d62d89fe40c4aa52861977e1a2cae of
+ the linux-2.6.18-xen repository, 3.3 release.
+ * Overtake hvc console by default.
* Remove 4gb segments warning completely.
- -- Bastian Blank <waldi at debian.org> Fri, 17 Oct 2008 15:12:44 +0200
+ -- Bastian Blank <waldi at debian.org> Sat, 18 Oct 2008 11:14:59 +0200
linux-2.6 (2.6.18.dfsg.1-23) stable; urgency=high
Modified: people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/config-update/defines
==============================================================================
--- people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/config-update/defines (original)
+++ people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/config-update/defines Sat Oct 18 14:44:52 2008
@@ -1,5 +1,5 @@
[abi]
-abiname: xen-3.1-2
+abiname: xen-3.3-1
[base]
arches:
Modified: people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/console-hvc-overtake.patch
==============================================================================
--- people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/console-hvc-overtake.patch (original)
+++ people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/console-hvc-overtake.patch Sat Oct 18 14:44:52 2008
@@ -6,16 +6,18 @@
* 'xencons=tty' [XC_TTY]: Console attached to '/dev/tty[0-9]+'.
* 'xencons=ttyS' [XC_SERIAL]: Console attached to '/dev/ttyS[0-9]+'.
* 'xencons=xvc' [XC_XVC]: Console attached to '/dev/xvc0'.
+- * default: XC_XVC
+ * 'xencons=hvc' [XC_HVC]: Console attached to '/dev/hvc0'.
- * default: DOM0 -> XC_SERIAL ; all others -> XC_TTY.
++ * default: XC_HVC
*
* NB. In mode XC_TTY, we create dummy consoles for tty2-63. This suppresses
* warnings from standard distro startup scripts.
*/
static enum {
- XC_OFF, XC_TTY, XC_SERIAL, XC_XVC
+-} xc_mode = XC_XVC;
+ XC_OFF, XC_TTY, XC_SERIAL, XC_XVC, XC_HVC
- } xc_mode;
++} xc_mode = XC_HVC;
static int xc_num = -1;
/* /dev/xvc0 device number allocated by lanana.org. */
Modified: people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/remove-4gb-warning.patch
==============================================================================
--- people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/remove-4gb-warning.patch (original)
+++ people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/remove-4gb-warning.patch Sat Oct 18 14:44:52 2008
@@ -1,7 +1,7 @@
--- a/arch/i386/kernel/Makefile
+++ b/arch/i386/kernel/Makefile
@@ -94,1 +94,0 @@
--obj-y += fixup.o
+-obj-$(CONFIG_XEN) += fixup.o
--- a/arch/i386/kernel/entry-xen.S
+++ b/arch/i386/kernel/entry-xen.S
@@ -1205,7 +1205,0 @@
Copied: people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/xen-748f324a4b2d62d89fe40c4aa52861977e1a2cae.patch (from r12345, /people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/xen-3.1-15467.patch)
==============================================================================
--- /people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/xen-3.1-15467.patch (original)
+++ people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/features/all/xen/xen-748f324a4b2d62d89fe40c4aa52861977e1a2cae.patch Sat Oct 18 14:44:52 2008
@@ -1,32 +1,12 @@
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/boot-xen/Makefile tmp-linux-2.6-xen.patch/arch/i386/boot-xen/Makefile
---- pristine-linux-2.6.18.2/arch/i386/boot-xen/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/boot-xen/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,21 @@
-+
-+OBJCOPYFLAGS := -g --strip-unneeded
-+
-+vmlinuz: vmlinux-stripped FORCE
-+ $(call if_changed,gzip)
-+
-+vmlinux-stripped: vmlinux FORCE
-+ $(call if_changed,objcopy)
-+
-+INSTALL_ROOT := $(patsubst %/boot,%,$(INSTALL_PATH))
-+
-+XINSTALL_NAME ?= $(KERNELRELEASE)
-+install:
-+ mkdir -p $(INSTALL_ROOT)/boot
-+ ln -f -s vmlinuz-$(XINSTALL_NAME)$(INSTALL_SUFFIX) $(INSTALL_ROOT)/boot/vmlinuz-$(VERSION).$(PATCHLEVEL).$(SUBLEVEL)$(XENGUEST)$(INSTALL_SUFFIX)
-+ rm -f $(INSTALL_ROOT)/boot/vmlinuz-$(XINSTALL_NAME)$(INSTALL_SUFFIX)
-+ install -m0644 vmlinuz $(INSTALL_ROOT)/boot/vmlinuz-$(XINSTALL_NAME)$(INSTALL_SUFFIX)
-+ install -m0644 vmlinux $(INSTALL_ROOT)/boot/vmlinux-syms-$(XINSTALL_NAME)$(INSTALL_SUFFIX)
-+ install -m0664 .config $(INSTALL_ROOT)/boot/config-$(XINSTALL_NAME)$(INSTALL_SUFFIX)
-+ install -m0664 System.map $(INSTALL_ROOT)/boot/System.map-$(XINSTALL_NAME)$(INSTALL_SUFFIX)
-+ ln -f -s vmlinuz-$(XINSTALL_NAME)$(INSTALL_SUFFIX) $(INSTALL_ROOT)/boot/vmlinuz-$(VERSION).$(PATCHLEVEL)$(XENGUEST)$(INSTALL_SUFFIX)
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/Kconfig tmp-linux-2.6-xen.patch/arch/i386/Kconfig
---- pristine-linux-2.6.18.2/arch/i386/Kconfig 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/Kconfig 2007-07-30 16:35:11.000000000 +0200
-@@ -16,6 +16,7 @@ config X86_32
+diff -r d894e36cfc30 -r 0aa021803deb .hgtags
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/.hgtags Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1 @@
++9e4b868027c659a2a3221eba365f3be782fde4a6 xen-3.2.0
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/Kconfig
+--- a/arch/i386/Kconfig Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -16,6 +16,7 @@
config GENERIC_TIME
bool
@@ -34,23 +14,25 @@
default y
config LOCKDEP_SUPPORT
-@@ -103,6 +104,15 @@ config X86_PC
+@@ -102,6 +103,17 @@
+ bool "PC-compatible"
help
Choose this option if your computer is a standard PC or compatible.
-
++
+config X86_XEN
+ bool "Xen-compatible"
++ select XEN
++ select X86_PAE
+ select X86_UP_APIC if !SMP && XEN_PRIVILEGED_GUEST
+ select X86_UP_IOAPIC if !SMP && XEN_PRIVILEGED_GUEST
+ select SWIOTLB
+ help
+ Choose this option if you plan to run this kernel on top of the
+ Xen Hypervisor.
-+
+
config X86_ELAN
bool "AMD Elan"
- help
-@@ -213,6 +223,7 @@ source "arch/i386/Kconfig.cpu"
+@@ -213,6 +225,7 @@
config HPET_TIMER
bool "HPET Timer Support"
@@ -58,7 +40,7 @@
help
This enables the use of the HPET for the kernel's internal timer.
HPET is the next generation timer replacing legacy 8254s.
-@@ -263,7 +274,7 @@ source "kernel/Kconfig.preempt"
+@@ -263,7 +276,7 @@
config X86_UP_APIC
bool "Local APIC support on uniprocessors"
@@ -67,7 +49,7 @@
help
A local APIC (Advanced Programmable Interrupt Controller) is an
integrated interrupt controller in the CPU. If you have a single-CPU
-@@ -288,12 +299,12 @@ config X86_UP_IOAPIC
+@@ -288,12 +301,12 @@
config X86_LOCAL_APIC
bool
@@ -82,7 +64,7 @@
default y
config X86_VISWS_APIC
-@@ -303,7 +314,7 @@ config X86_VISWS_APIC
+@@ -303,7 +316,7 @@
config X86_MCE
bool "Machine Check Exception"
@@ -91,7 +73,16 @@
---help---
Machine Check Exception support allows the processor to notify the
kernel if it detects a problem (e.g. overheating, component failure).
-@@ -402,6 +413,7 @@ config X86_REBOOTFIXUPS
+@@ -384,7 +397,7 @@
+
+ config X86_REBOOTFIXUPS
+ bool "Enable X86 board specific fixups for reboot"
+- depends on X86
++ depends on !X86_XEN
+ default n
+ ---help---
+ This enables chipset and/or board specific fixups to be done
+@@ -402,6 +415,7 @@
config MICROCODE
tristate "/dev/cpu/microcode - Intel IA32 CPU microcode support"
@@ -99,26 +90,18 @@
---help---
If you say Y here and also to "/dev file system support" in the
'File systems' section, you will be able to update the microcode on
-@@ -419,6 +431,7 @@ config MICROCODE
-
- config X86_MSR
- tristate "/dev/cpu/*/msr - Model-specific register support"
-+ depends on !X86_XEN
- help
- This device gives privileged processes access to the x86
- Model-Specific Registers (MSRs). It is a character device with
-@@ -434,6 +447,10 @@ config X86_CPUID
+@@ -433,6 +447,10 @@
+ be executed on a specific processor. It is a character device
with major 203 and minors 0 to 31 for /dev/cpu/0/cpuid to
/dev/cpu/31/cpuid.
-
++
+config SWIOTLB
+ bool
+ default n
-+
+
source "drivers/firmware/Kconfig"
- choice
-@@ -616,6 +633,7 @@ config HIGHPTE
+@@ -616,6 +634,7 @@
config MATH_EMULATION
bool "Math emulation"
@@ -126,16 +109,15 @@
---help---
Linux can emulate a math coprocessor (used for floating point
operations) if you don't have one. 486DX and Pentium processors have
-@@ -641,6 +659,8 @@ config MATH_EMULATION
+@@ -641,6 +660,7 @@
config MTRR
bool "MTRR (Memory Type Range Register) support"
+ depends on !XEN_UNPRIVILEGED_GUEST
-+ default y if X86_XEN
---help---
On Intel P6 family processors (Pentium Pro, Pentium II and later)
the Memory Type Range Registers (MTRRs) may be used to control
-@@ -675,7 +695,7 @@ config MTRR
+@@ -675,7 +695,7 @@
config EFI
bool "Boot from EFI support"
@@ -144,7 +126,7 @@
default n
---help---
This enables the the kernel to boot on EFI platforms using
-@@ -693,7 +713,7 @@ config EFI
+@@ -693,7 +713,7 @@
config IRQBALANCE
bool "Enable kernel irq balancing"
@@ -153,7 +135,7 @@
default y
help
The default yes will allow the kernel to do irq load balancing.
-@@ -741,7 +761,7 @@ source kernel/Kconfig.hz
+@@ -741,7 +761,7 @@
config KEXEC
bool "kexec system call (EXPERIMENTAL)"
@@ -162,7 +144,15 @@
help
kexec is a system call that implements the ability to shutdown your
current kernel, and to start another kernel. It is like a reboot
-@@ -793,6 +813,7 @@ config HOTPLUG_CPU
+@@ -760,6 +780,7 @@
+ bool "kernel crash dumps (EXPERIMENTAL)"
+ depends on EXPERIMENTAL
+ depends on HIGHMEM
++ depends on !XEN
+ help
+ Generate crash dump after being started by kexec.
+
+@@ -793,6 +814,7 @@
config COMPAT_VDSO
bool "Compat VDSO support"
@@ -170,16 +160,15 @@
default y
help
Map the VDSO to the predictable old-style address too.
-@@ -810,18 +831,20 @@ config ARCH_ENABLE_MEMORY_HOTPLUG
+@@ -810,18 +832,18 @@
depends on HIGHMEM
menu "Power management options (ACPI, APM)"
- depends on !X86_VOYAGER
+ depends on !(X86_VOYAGER || XEN_UNPRIVILEGED_GUEST)
-+if !X86_XEN
- source kernel/power/Kconfig
-+endif
+-source kernel/power/Kconfig
++source "kernel/power/Kconfig"
source "drivers/acpi/Kconfig"
@@ -194,7 +183,7 @@
---help---
APM is a BIOS specification for saving power using several different
techniques. This is mostly useful for battery powered laptops with
-@@ -1006,6 +1029,7 @@ choice
+@@ -1006,12 +1028,20 @@
config PCI_GOBIOS
bool "BIOS"
@@ -202,21 +191,20 @@
config PCI_GOMMCONFIG
bool "MMConfig"
-@@ -1013,6 +1037,13 @@ config PCI_GOMMCONFIG
+
config PCI_GODIRECT
bool "Direct"
-
++
+config PCI_GOXEN_FE
+ bool "Xen PCI Frontend"
+ depends on X86_XEN
+ help
+ The PCI device frontend driver allows the kernel to import arbitrary
+ PCI devices from a PCI backend to support PCI driver domains.
-+
+
config PCI_GOANY
bool "Any"
-
-@@ -1020,7 +1051,7 @@ endchoice
+@@ -1020,7 +1050,7 @@
config PCI_BIOS
bool
@@ -225,13 +213,14 @@
default y
config PCI_DIRECT
-@@ -1033,6 +1064,18 @@ config PCI_MMCONFIG
+@@ -1033,6 +1063,19 @@
depends on PCI && ACPI && (PCI_GOMMCONFIG || PCI_GOANY)
default y
+config XEN_PCIDEV_FRONTEND
+ bool
+ depends on PCI && X86_XEN && (PCI_GOXEN_FE || PCI_GOANY)
++ select HOTPLUG
+ default y
+
+config XEN_PCIDEV_FE_DEBUG
@@ -244,7 +233,7 @@
source "drivers/pci/pcie/Kconfig"
source "drivers/pci/Kconfig"
-@@ -1043,7 +1086,7 @@ config ISA_DMA_API
+@@ -1043,7 +1086,7 @@
config ISA
bool "ISA support"
@@ -253,7 +242,7 @@
help
Find out whether you have ISA slots on your motherboard. ISA is the
name of a bus system, i.e. the way the CPU talks to the other stuff
-@@ -1070,7 +1113,7 @@ config EISA
+@@ -1070,7 +1113,7 @@
source "drivers/eisa/Kconfig"
config MCA
@@ -262,7 +251,7 @@
default y if X86_VOYAGER
help
MicroChannel Architecture is found in some IBM PS/2 machines and
-@@ -1146,6 +1189,8 @@ source "security/Kconfig"
+@@ -1146,6 +1189,8 @@
source "crypto/Kconfig"
@@ -271,7 +260,7 @@
source "lib/Kconfig"
#
-@@ -1171,7 +1216,7 @@ config X86_SMP
+@@ -1171,7 +1216,7 @@
config X86_HT
bool
@@ -280,10 +269,13 @@
default y
config X86_BIOS_REBOOT
-@@ -1184,6 +1229,16 @@ config X86_TRAMPOLINE
+@@ -1182,6 +1227,17 @@
+ config X86_TRAMPOLINE
+ bool
depends on X86_SMP || (X86_VOYAGER && SMP)
- default y
-
++ depends on !XEN
++ default y
++
+config X86_NO_TSS
+ bool
+ depends on X86_XEN
@@ -292,23 +284,13 @@
+config X86_NO_IDT
+ bool
+ depends on X86_XEN
-+ default y
-+
- config KTIME_SCALAR
- bool
default y
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/Kconfig.cpu tmp-linux-2.6-xen.patch/arch/i386/Kconfig.cpu
---- pristine-linux-2.6.18.2/arch/i386/Kconfig.cpu 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/Kconfig.cpu 2007-07-30 16:35:11.000000000 +0200
-@@ -7,7 +7,6 @@ choice
-
- config M386
- bool "386"
-- depends on !UML
- ---help---
- This is the processor type of your CPU. This information is used for
- optimizing purposes. In order to compile a kernel that can run on
-@@ -252,7 +251,7 @@ config X86_PPRO_FENCE
+
+ config KTIME_SCALAR
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/Kconfig.cpu
+--- a/arch/i386/Kconfig.cpu Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/Kconfig.cpu Wed Sep 10 10:54:08 2008 +0100
+@@ -252,7 +252,7 @@
config X86_F00F_BUG
bool
@@ -317,26 +299,17 @@
default y
config X86_WP_WORKS_OK
-@@ -302,7 +301,7 @@ config X86_USE_PPRO_CHECKSUM
-
- config X86_USE_3DNOW
- bool
-- depends on (MCYRIXIII || MK7 || MGEODE_LX) && !UML
-+ depends on MCYRIXIII || MK7 || MGEODE_LX
- default y
-
- config X86_OOSTORE
-@@ -312,5 +311,5 @@ config X86_OOSTORE
+@@ -312,5 +312,5 @@
config X86_TSC
bool
- depends on (MWINCHIP3D || MWINCHIP2 || MCRUSOE || MEFFICEON || MCYRIXIII || MK7 || MK6 || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || M586MMX || M586TSC || MK8 || MVIAC3_2 || MGEODEGX1 || MGEODE_LX) && !X86_NUMAQ
+ depends on (MWINCHIP3D || MWINCHIP2 || MCRUSOE || MEFFICEON || MCYRIXIII || MK7 || MK6 || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || M586MMX || M586TSC || MK8 || MVIAC3_2 || MGEODEGX1 || MGEODE_LX) && !X86_NUMAQ && !X86_XEN
default y
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/Kconfig.debug tmp-linux-2.6-xen.patch/arch/i386/Kconfig.debug
---- pristine-linux-2.6.18.2/arch/i386/Kconfig.debug 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/Kconfig.debug 2007-07-30 16:35:11.000000000 +0200
-@@ -79,6 +79,7 @@ config X86_MPPARSE
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/Kconfig.debug
+--- a/arch/i386/Kconfig.debug Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/Kconfig.debug Wed Sep 10 10:54:08 2008 +0100
+@@ -79,6 +79,7 @@
config DOUBLEFAULT
default y
bool "Enable doublefault exception handler" if EMBEDDED
@@ -344,9 +317,112 @@
help
This option allows trapping of rare doublefault exceptions that
would otherwise cause a system to silently reboot. Disabling this
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/acpi/boot-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/acpi/boot-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/acpi/boot-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/acpi/boot-xen.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/Makefile
+--- a/arch/i386/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -71,6 +71,10 @@
+ mflags-$(CONFIG_X86_SUMMIT) := -Iinclude/asm-i386/mach-summit
+ mcore-$(CONFIG_X86_SUMMIT) := mach-default
+
++# Xen subarch support
++mflags-$(CONFIG_X86_XEN) := -Iinclude/asm-i386/mach-xen
++mcore-$(CONFIG_X86_XEN) := mach-xen
++
+ # generic subarchitecture
+ mflags-$(CONFIG_X86_GENERICARCH) := -Iinclude/asm-i386/mach-generic
+ mcore-$(CONFIG_X86_GENERICARCH) := mach-default
+@@ -102,9 +106,20 @@
+
+ boot := arch/i386/boot
+
+-PHONY += zImage bzImage compressed zlilo bzlilo \
++PHONY += zImage bzImage vmlinuz compressed zlilo bzlilo \
+ zdisk bzdisk fdimage fdimage144 fdimage288 isoimage install
+
++ifdef CONFIG_XEN
++CPPFLAGS := -D__XEN_INTERFACE_VERSION__=$(CONFIG_XEN_INTERFACE_VERSION) \
++ -Iinclude$(if $(KBUILD_SRC),2)/asm/mach-xen $(CPPFLAGS)
++all: vmlinuz
++
++# KBUILD_IMAGE specifies the target image being built
++KBUILD_IMAGE := $(boot)/vmlinuz
++
++vmlinuz: vmlinux
++ $(Q)$(MAKE) $(build)=$(boot) $(KBUILD_IMAGE)
++else
+ all: bzImage
+
+ # KBUILD_IMAGE specify target image being built
+@@ -124,6 +139,7 @@
+
+ fdimage fdimage144 fdimage288 isoimage: vmlinux
+ $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(KBUILD_IMAGE) $@
++endif
+
+ install:
+ $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(KBUILD_IMAGE) install
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/boot/Makefile
+--- a/arch/i386/boot/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/boot/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -26,7 +26,7 @@
+ #RAMDISK := -DRAMDISK=512
+
+ targets := vmlinux.bin bootsect bootsect.o \
+- setup setup.o zImage bzImage
++ setup setup.o zImage bzImage vmlinuz vmlinux-stripped
+ subdir- := compressed
+
+ hostprogs-y := tools/build
+@@ -133,5 +133,13 @@
+ cp System.map $(INSTALL_PATH)/
+ if [ -x /sbin/lilo ]; then /sbin/lilo; else /etc/lilo/install; fi
+
++$(obj)/vmlinuz: $(obj)/vmlinux-stripped FORCE
++ $(call if_changed,gzip)
++ @echo 'Kernel: $@ is ready' ' (#'`cat .version`')'
++
++$(obj)/vmlinux-stripped: OBJCOPYFLAGS := -g --strip-unneeded
++$(obj)/vmlinux-stripped: vmlinux FORCE
++ $(call if_changed,objcopy)
++
+ install:
+ sh $(srctree)/$(src)/install.sh $(KERNELRELEASE) $(BOOTIMAGE) System.map "$(INSTALL_PATH)"
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/Makefile
+--- a/arch/i386/kernel/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -43,6 +43,7 @@
+ EXTRA_AFLAGS := -traditional
+
+ obj-$(CONFIG_SCx200) += scx200.o
++obj-$(CONFIG_XEN) += fixup.o
+
+ # vsyscall.o contains the vsyscall DSO images as __initdata.
+ # We must build both images before we can assemble it.
+@@ -80,5 +81,8 @@
+ $(obj)/vsyscall-sysenter.o $(obj)/vsyscall-note.o FORCE
+ $(call if_changed,syscall)
+
++early_printk-y += ../../x86_64/kernel/early_printk.o
+ k8-y += ../../x86_64/kernel/k8.o
+
++disabled-obj-$(CONFIG_XEN) := i8259.o reboot.o smpboot.o
++%/head.o %/head.s: $(if $(CONFIG_XEN),EXTRA_AFLAGS,dummy) :=
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/acpi/Makefile
+--- a/arch/i386/kernel/acpi/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/acpi/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -4,5 +4,9 @@
+
+ ifneq ($(CONFIG_ACPI_PROCESSOR),)
+ obj-y += cstate.o processor.o
++ifneq ($(CONFIG_PROCESSOR_EXTERNAL_CONTROL),)
++obj-$(CONFIG_XEN) += processor_extcntl_xen.o
++endif
+ endif
+
++disabled-obj-$(CONFIG_XEN) := cstate.o wakeup.o
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/acpi/boot-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/acpi/boot-xen.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,1168 @@
+/*
+ * boot.c - Architecture-Specific Low-Level ACPI Boot Support
@@ -1516,20 +1592,379 @@
+
+ return 0;
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/acpi/Makefile tmp-linux-2.6-xen.patch/arch/i386/kernel/acpi/Makefile
---- pristine-linux-2.6.18.2/arch/i386/kernel/acpi/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/acpi/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -6,3 +6,7 @@ ifneq ($(CONFIG_ACPI_PROCESSOR),)
- obj-y += cstate.o processor.o
- endif
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/acpi/processor.c
+--- a/arch/i386/kernel/acpi/processor.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/acpi/processor.c Wed Sep 10 10:54:08 2008 +0100
+@@ -62,7 +62,18 @@
+ /* Initialize _PDC data based on the CPU vendor */
+ void arch_acpi_processor_init_pdc(struct acpi_processor *pr)
+ {
++#ifdef CONFIG_XEN
++ /*
++ * As a work-around, just use cpu0's cpuinfo for all processors.
++ * Further work is required to expose xen hypervisor interface of
++ * getting physical cpuinfo to dom0 kernel and then
++ * arch_acpi_processor_init_pdc can set _PDC parameters according
++ * to Xen's phys information.
++ */
++ unsigned int cpu = 0;
++#else
+ unsigned int cpu = pr->id;
++#endif /* CONFIG_XEN */
+ struct cpuinfo_x86 *c = cpu_data + cpu;
-+ifdef CONFIG_XEN
-+include $(srctree)/scripts/Makefile.xen
-+obj-y := $(call cherrypickxen, $(obj-y), $(src))
-+endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/apic-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/apic-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/apic-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/apic-xen.c 2007-07-30 16:35:11.000000000 +0200
+ pr->pdc = NULL;
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/acpi/processor_extcntl_xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/acpi/processor_extcntl_xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,227 @@
++/*
++ * processor_extcntl_xen.c - interface to notify Xen
++ *
++ * Copyright (C) 2008, Intel corporation
++ *
++ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or (at
++ * your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful, but
++ * WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
++ * General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License along
++ * with this program; if not, write to the Free Software Foundation, Inc.,
++ * 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.
++ *
++ */
++
++#include <linux/kernel.h>
++#include <linux/init.h>
++#include <linux/types.h>
++#include <linux/acpi.h>
++#include <linux/pm.h>
++#include <linux/cpu.h>
++
++#include <linux/cpufreq.h>
++#include <acpi/processor.h>
++#include <asm/hypercall.h>
++
++static int xen_cx_notifier(struct acpi_processor *pr, int action)
++{
++ int ret, count = 0, i;
++ xen_platform_op_t op = {
++ .cmd = XENPF_set_processor_pminfo,
++ .interface_version = XENPF_INTERFACE_VERSION,
++ .u.set_pminfo.id = pr->acpi_id,
++ .u.set_pminfo.type = XEN_PM_CX,
++ };
++ struct xen_processor_cx *data, *buf;
++ struct acpi_processor_cx *cx;
++
++ if (action == PROCESSOR_PM_CHANGE)
++ return -EINVAL;
++
++ /* Convert to Xen defined structure and hypercall */
++ buf = kzalloc(pr->power.count * sizeof(struct xen_processor_cx),
++ GFP_KERNEL);
++ if (!buf)
++ return -ENOMEM;
++
++ data = buf;
++ for (i = 1; i <= pr->power.count; i++) {
++ cx = &pr->power.states[i];
++ /* Skip invalid cstate entry */
++ if (!cx->valid)
++ continue;
++
++ data->type = cx->type;
++ data->latency = cx->latency;
++ data->power = cx->power;
++ data->reg.space_id = cx->reg.space_id;
++ data->reg.bit_width = cx->reg.bit_width;
++ data->reg.bit_offset = cx->reg.bit_offset;
++ data->reg.access_size = cx->reg.reserved;
++ data->reg.address = cx->reg.address;
++
++ /* Get dependency relationships */
++ if (cx->csd_count) {
++ printk("Wow! _CSD is found. Not support for now!\n");
++ kfree(buf);
++ return -EINVAL;
++ } else {
++ data->dpcnt = 0;
++ set_xen_guest_handle(data->dp, NULL);
++ }
++
++ data++;
++ count++;
++ }
++
++ if (!count) {
++ printk("No available Cx info for cpu %d\n", pr->acpi_id);
++ kfree(buf);
++ return -EINVAL;
++ }
++
++ op.u.set_pminfo.power.count = count;
++ op.u.set_pminfo.power.flags.bm_control = pr->flags.bm_control;
++ op.u.set_pminfo.power.flags.bm_check = pr->flags.bm_check;
++ op.u.set_pminfo.power.flags.has_cst = pr->flags.has_cst;
++ op.u.set_pminfo.power.flags.power_setup_done = pr->flags.power_setup_done;
++
++ set_xen_guest_handle(op.u.set_pminfo.power.states, buf);
++ ret = HYPERVISOR_platform_op(&op);
++ kfree(buf);
++ return ret;
++}
++
++static void convert_pct_reg(struct xen_pct_register *xpct,
++ struct acpi_pct_register *apct)
++{
++ xpct->descriptor = apct->descriptor;
++ xpct->length = apct->length;
++ xpct->space_id = apct->space_id;
++ xpct->bit_width = apct->bit_width;
++ xpct->bit_offset = apct->bit_offset;
++ xpct->reserved = apct->reserved;
++ xpct->address = apct->address;
++}
++
++static void convert_pss_states(struct xen_processor_px *xpss,
++ struct acpi_processor_px *apss, int state_count)
++{
++ int i;
++ for(i=0; i<state_count; i++) {
++ xpss->core_frequency = apss->core_frequency;
++ xpss->power = apss->power;
++ xpss->transition_latency = apss->transition_latency;
++ xpss->bus_master_latency = apss->bus_master_latency;
++ xpss->control = apss->control;
++ xpss->status = apss->status;
++ xpss++;
++ apss++;
++ }
++}
++
++static void convert_psd_pack(struct xen_psd_package *xpsd,
++ struct acpi_psd_package *apsd)
++{
++ xpsd->num_entries = apsd->num_entries;
++ xpsd->revision = apsd->revision;
++ xpsd->domain = apsd->domain;
++ xpsd->coord_type = apsd->coord_type;
++ xpsd->num_processors = apsd->num_processors;
++}
++
++static int xen_px_notifier(struct acpi_processor *pr, int action)
++{
++ int ret;
++ xen_platform_op_t op = {
++ .cmd = XENPF_set_processor_pminfo,
++ .interface_version = XENPF_INTERFACE_VERSION,
++ .u.set_pminfo.id = pr->acpi_id,
++ .u.set_pminfo.type = XEN_PM_PX,
++ };
++ struct xen_processor_performance *perf;
++ struct xen_processor_px *states = NULL;
++ struct acpi_processor_performance *px;
++ struct acpi_psd_package *pdomain;
++
++ /* leave dynamic ppc handle in the future */
++ if (action == PROCESSOR_PM_CHANGE)
++ return 0;
++
++ perf = &op.u.set_pminfo.perf;
++ px = pr->performance;
++
++ perf->flags = XEN_PX_PPC |
++ XEN_PX_PCT |
++ XEN_PX_PSS |
++ XEN_PX_PSD;
++
++ /* ppc */
++ perf->ppc = pr->performance_platform_limit;
++
++ /* pct */
++ convert_pct_reg(&perf->control_register, &px->control_register);
++ convert_pct_reg(&perf->status_register, &px->status_register);
++
++ /* pss */
++ perf->state_count = px->state_count;
++ states = kzalloc(px->state_count*sizeof(xen_processor_px_t),GFP_KERNEL);
++ if (!states)
++ return -ENOMEM;
++ convert_pss_states(states, px->states, px->state_count);
++ set_xen_guest_handle(perf->states, states);
++
++ /* psd */
++ pdomain = &px->domain_info;
++ convert_psd_pack(&perf->domain_info, pdomain);
++ if (perf->domain_info.num_processors) {
++ if (pdomain->coord_type == DOMAIN_COORD_TYPE_SW_ALL)
++ perf->shared_type = CPUFREQ_SHARED_TYPE_ALL;
++ else if (pdomain->coord_type == DOMAIN_COORD_TYPE_SW_ANY)
++ perf->shared_type = CPUFREQ_SHARED_TYPE_ANY;
++ else if (pdomain->coord_type == DOMAIN_COORD_TYPE_HW_ALL)
++ perf->shared_type = CPUFREQ_SHARED_TYPE_HW;
++ } else
++ perf->shared_type = CPUFREQ_SHARED_TYPE_NONE;
++
++ ret = HYPERVISOR_platform_op(&op);
++ kfree(states);
++ return ret;
++}
++
++static int xen_tx_notifier(struct acpi_processor *pr, int action)
++{
++ return -EINVAL;
++}
++static int xen_hotplug_notifier(struct acpi_processor *pr, int event)
++{
++ return -EINVAL;
++}
++
++static struct processor_extcntl_ops xen_extcntl_ops = {
++ .hotplug = xen_hotplug_notifier,
++};
++
++void arch_acpi_processor_init_extcntl(const struct processor_extcntl_ops **ops)
++{
++ unsigned int pmbits = (xen_start_info->flags & SIF_PM_MASK) >> 8;
++
++ if (pmbits & XEN_PROCESSOR_PM_CX)
++ xen_extcntl_ops.pm_ops[PM_TYPE_IDLE] = xen_cx_notifier;
++ if (pmbits & XEN_PROCESSOR_PM_PX)
++ xen_extcntl_ops.pm_ops[PM_TYPE_PERF] = xen_px_notifier;
++ if (pmbits & XEN_PROCESSOR_PM_TX)
++ xen_extcntl_ops.pm_ops[PM_TYPE_THR] = xen_tx_notifier;
++
++ *ops = &xen_extcntl_ops;
++}
++EXPORT_SYMBOL(arch_acpi_processor_init_extcntl);
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/acpi/sleep-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/acpi/sleep-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,113 @@
++/*
++ * sleep.c - x86-specific ACPI sleep support.
++ *
++ * Copyright (C) 2001-2003 Patrick Mochel
++ * Copyright (C) 2001-2003 Pavel Machek <pavel at suse.cz>
++ */
++
++#include <linux/acpi.h>
++#include <linux/bootmem.h>
++#include <linux/dmi.h>
++#include <linux/cpumask.h>
++
++#include <asm/smp.h>
++
++#ifndef CONFIG_ACPI_PV_SLEEP
++/* address in low memory of the wakeup routine. */
++unsigned long acpi_wakeup_address = 0;
++unsigned long acpi_video_flags;
++extern char wakeup_start, wakeup_end;
++
++extern unsigned long FASTCALL(acpi_copy_wakeup_routine(unsigned long));
++#endif
++
++/**
++ * acpi_save_state_mem - save kernel state
++ *
++ * Create an identity mapped page table and copy the wakeup routine to
++ * low memory.
++ */
++int acpi_save_state_mem(void)
++{
++#ifndef CONFIG_ACPI_PV_SLEEP
++ if (!acpi_wakeup_address)
++ return 1;
++ memcpy((void *)acpi_wakeup_address, &wakeup_start,
++ &wakeup_end - &wakeup_start);
++ acpi_copy_wakeup_routine(acpi_wakeup_address);
++#endif
++ return 0;
++}
++
++/*
++ * acpi_restore_state - undo effects of acpi_save_state_mem
++ */
++void acpi_restore_state_mem(void)
++{
++}
++
++/**
++ * acpi_reserve_bootmem - do _very_ early ACPI initialisation
++ *
++ * We allocate a page from the first 1MB of memory for the wakeup
++ * routine for when we come back from a sleep state. The
++ * runtime allocator allows specification of <16MB pages, but not
++ * <1MB pages.
++ */
++void __init acpi_reserve_bootmem(void)
++{
++#ifndef CONFIG_ACPI_PV_SLEEP
++ if ((&wakeup_end - &wakeup_start) > PAGE_SIZE) {
++ printk(KERN_ERR
++ "ACPI: Wakeup code way too big, S3 disabled.\n");
++ return;
++ }
++
++ acpi_wakeup_address = (unsigned long)alloc_bootmem_low(PAGE_SIZE);
++ if (!acpi_wakeup_address)
++ printk(KERN_ERR "ACPI: Cannot allocate lowmem, S3 disabled.\n");
++#endif
++}
++
++#ifndef CONFIG_ACPI_PV_SLEEP
++static int __init acpi_sleep_setup(char *str)
++{
++ while ((str != NULL) && (*str != '\0')) {
++ if (strncmp(str, "s3_bios", 7) == 0)
++ acpi_video_flags = 1;
++ if (strncmp(str, "s3_mode", 7) == 0)
++ acpi_video_flags |= 2;
++ str = strchr(str, ',');
++ if (str != NULL)
++ str += strspn(str, ", \t");
++ }
++ return 1;
++}
++
++__setup("acpi_sleep=", acpi_sleep_setup);
++
++static __init int reset_videomode_after_s3(struct dmi_system_id *d)
++{
++ acpi_video_flags |= 2;
++ return 0;
++}
++
++static __initdata struct dmi_system_id acpisleep_dmi_table[] = {
++ { /* Reset video mode after returning from ACPI S3 sleep */
++ .callback = reset_videomode_after_s3,
++ .ident = "Toshiba Satellite 4030cdt",
++ .matches = {
++ DMI_MATCH(DMI_PRODUCT_NAME, "S4030CDT/4.3"),
++ },
++ },
++ {}
++};
++
++static int __init acpisleep_dmi_init(void)
++{
++ dmi_check_system(acpisleep_dmi_table);
++ return 0;
++}
++
++core_initcall(acpisleep_dmi_init);
++#endif /* CONFIG_ACPI_PV_SLEEP */
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/apic-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/apic-xen.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,155 @@
+/*
+ * Local APIC handling, local APIC timers
@@ -1686,10 +2121,10 @@
+
+ return 0;
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/asm-offsets.c tmp-linux-2.6-xen.patch/arch/i386/kernel/asm-offsets.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/asm-offsets.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/asm-offsets.c 2007-07-30 16:35:11.000000000 +0200
-@@ -66,9 +66,14 @@ void foo(void)
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/asm-offsets.c
+--- a/arch/i386/kernel/asm-offsets.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/asm-offsets.c Wed Sep 10 10:54:08 2008 +0100
+@@ -66,9 +66,14 @@
OFFSET(pbe_orig_address, pbe, orig_address);
OFFSET(pbe_next, pbe, next);
@@ -1705,9 +2140,17 @@
DEFINE(PAGE_SIZE_asm, PAGE_SIZE);
DEFINE(VDSO_PRELINK, VDSO_PRELINK);
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/cpu/common-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/cpu/common-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/cpu/common-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/cpu/common-xen.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/cpu/Makefile
+--- a/arch/i386/kernel/cpu/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/cpu/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -17,3 +17,4 @@
+
+ obj-$(CONFIG_MTRR) += mtrr/
+ obj-$(CONFIG_CPU_FREQ) += cpufreq/
++
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/cpu/common-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/cpu/common-xen.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,743 @@
+#include <linux/init.h>
+#include <linux/string.h>
@@ -2301,7 +2744,7 @@
+#endif
+}
+
-+void __cpuinit cpu_gdt_init(struct Xgt_desc_struct *gdt_descr)
++static void __cpuinit cpu_gdt_init(const struct Xgt_desc_struct *gdt_descr)
+{
+ unsigned long frames[16];
+ unsigned long va;
@@ -2314,7 +2757,7 @@
+ make_lowmem_page_readonly(
+ (void *)va, XENFEAT_writable_descriptor_tables);
+ }
-+ if (HYPERVISOR_set_gdt(frames, gdt_descr->size / 8))
++ if (HYPERVISOR_set_gdt(frames, (gdt_descr->size + 1) / 8))
+ BUG();
+}
+
@@ -2452,22 +2895,530 @@
+ per_cpu(cpu_tlbstate, cpu).active_mm = &init_mm;
+}
+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/cpu/Makefile tmp-linux-2.6-xen.patch/arch/i386/kernel/cpu/Makefile
---- pristine-linux-2.6.18.2/arch/i386/kernel/cpu/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/cpu/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -17,3 +17,8 @@ obj-$(CONFIG_X86_MCE) += mcheck/
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/cpu/cpufreq/powernow-k8.c
+--- a/arch/i386/kernel/cpu/cpufreq/powernow-k8.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/cpu/cpufreq/powernow-k8.c Wed Sep 10 10:54:08 2008 +0100
+@@ -46,7 +46,7 @@
+
+ #define PFX "powernow-k8: "
+ #define BFX PFX "BIOS error: "
+-#define VERSION "version 2.00.00"
++#define VERSION "version 2.20.00"
+ #include "powernow-k8.h"
+
+ /* serialize freq changes */
+@@ -66,36 +66,15 @@
+ return 800 + (fid * 100);
+ }
- obj-$(CONFIG_MTRR) += mtrr/
- obj-$(CONFIG_CPU_FREQ) += cpufreq/
+-
+ /* Return a frequency in KHz, given an input fid */
+ static u32 find_khz_freq_from_fid(u32 fid)
+ {
+ return 1000 * find_freq_from_fid(fid);
+ }
+
+-/* Return a frequency in MHz, given an input fid and did */
+-static u32 find_freq_from_fiddid(u32 fid, u32 did)
++static u32 find_khz_freq_from_pstate(struct cpufreq_frequency_table *data, u32 pstate)
+ {
+- return 100 * (fid + 0x10) >> did;
+-}
+-
+-static u32 find_khz_freq_from_fiddid(u32 fid, u32 did)
+-{
+- return 1000 * find_freq_from_fiddid(fid, did);
+-}
+-
+-static u32 find_fid_from_pstate(u32 pstate)
+-{
+- u32 hi, lo;
+- rdmsr(MSR_PSTATE_DEF_BASE + pstate, lo, hi);
+- return lo & HW_PSTATE_FID_MASK;
+-}
+-
+-static u32 find_did_from_pstate(u32 pstate)
+-{
+- u32 hi, lo;
+- rdmsr(MSR_PSTATE_DEF_BASE + pstate, lo, hi);
+- return (lo & HW_PSTATE_DID_MASK) >> HW_PSTATE_DID_SHIFT;
++ return data[pstate].frequency;
+ }
+
+ /* Return the vco fid for an input fid
+@@ -139,9 +118,7 @@
+ if (cpu_family == CPU_HW_PSTATE) {
+ rdmsr(MSR_PSTATE_STATUS, lo, hi);
+ i = lo & HW_PSTATE_MASK;
+- rdmsr(MSR_PSTATE_DEF_BASE + i, lo, hi);
+- data->currfid = lo & HW_PSTATE_FID_MASK;
+- data->currdid = (lo & HW_PSTATE_DID_MASK) >> HW_PSTATE_DID_SHIFT;
++ data->currpstate = i;
+ return 0;
+ }
+ do {
+@@ -292,7 +269,7 @@
+ static int transition_pstate(struct powernow_k8_data *data, u32 pstate)
+ {
+ wrmsr(MSR_PSTATE_CTRL, pstate, 0);
+- data->currfid = find_fid_from_pstate(pstate);
++ data->currpstate = pstate;
+ return 0;
+ }
+
+@@ -738,6 +715,7 @@
+
+ data->numps = psb->numps;
+ dprintk("numpstates: 0x%x\n", data->numps);
++ data->starting_core_affinity = cpumask_of_cpu(0);
+ return fill_powernow_table(data, (struct pst_s *)(psb+1), maxvid);
+ }
+ /*
+@@ -758,15 +736,43 @@
+ #ifdef CONFIG_X86_POWERNOW_K8_ACPI
+ static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data, unsigned int index)
+ {
+- if (!data->acpi_data.state_count || (cpu_family == CPU_HW_PSTATE))
++ if (!data->acpi_data->state_count || (cpu_family == CPU_HW_PSTATE))
+ return;
+
+- data->irt = (data->acpi_data.states[index].control >> IRT_SHIFT) & IRT_MASK;
+- data->rvo = (data->acpi_data.states[index].control >> RVO_SHIFT) & RVO_MASK;
+- data->exttype = (data->acpi_data.states[index].control >> EXT_TYPE_SHIFT) & EXT_TYPE_MASK;
+- data->plllock = (data->acpi_data.states[index].control >> PLL_L_SHIFT) & PLL_L_MASK;
+- data->vidmvs = 1 << ((data->acpi_data.states[index].control >> MVS_SHIFT) & MVS_MASK);
+- data->vstable = (data->acpi_data.states[index].control >> VST_SHIFT) & VST_MASK;
++ data->irt = (data->acpi_data->states[index].control >> IRT_SHIFT) & IRT_MASK;
++ data->rvo = (data->acpi_data->states[index].control >> RVO_SHIFT) & RVO_MASK;
++ data->exttype = (data->acpi_data->states[index].control >> EXT_TYPE_SHIFT) & EXT_TYPE_MASK;
++ data->plllock = (data->acpi_data->states[index].control >> PLL_L_SHIFT) & PLL_L_MASK;
++ data->vidmvs = 1 << ((data->acpi_data->states[index].control >> MVS_SHIFT) & MVS_MASK);
++ data->vstable = (data->acpi_data->states[index].control >> VST_SHIFT) & VST_MASK;
++}
+
-+ifdef CONFIG_XEN
-+include $(srctree)/scripts/Makefile.xen
-+obj-y := $(call cherrypickxen, $(obj-y), $(src))
-+endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/cpu/mtrr/main-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/cpu/mtrr/main-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/cpu/mtrr/main-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/cpu/mtrr/main-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,197 @@
++static struct acpi_processor_performance *acpi_perf_data[NR_CPUS];
++static int preregister_valid = 0;
++
++static int powernow_k8_cpu_preinit_acpi()
++{
++ int i;
++ struct acpi_processor_performance *data;
++ for_each_possible_cpu(i) {
++ data = kzalloc(sizeof(struct acpi_processor_performance),
++ GFP_KERNEL);
++ if (!data) {
++ int j;
++ for_each_possible_cpu(j) {
++ kfree(acpi_perf_data[j]);
++ acpi_perf_data[j] = NULL;
++ }
++ return -ENODEV;
++ }
++ acpi_perf_data[i] = data;
++ }
++
++ if (acpi_processor_preregister_performance(acpi_perf_data))
++ return -ENODEV;
++ else
++ preregister_valid = 1;
++ return 0;
+ }
+
+ static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data)
+@@ -774,28 +780,29 @@
+ struct cpufreq_frequency_table *powernow_table;
+ int ret_val;
+
+- if (acpi_processor_register_performance(&data->acpi_data, data->cpu)) {
++ data->acpi_data = acpi_perf_data[data->cpu];
++ if (acpi_processor_register_performance(data->acpi_data, data->cpu)) {
+ dprintk("register performance failed: bad ACPI data\n");
+ return -EIO;
+ }
+
+ /* verify the data contained in the ACPI structures */
+- if (data->acpi_data.state_count <= 1) {
++ if (data->acpi_data->state_count <= 1) {
+ dprintk("No ACPI P-States\n");
+ goto err_out;
+ }
+
+- if ((data->acpi_data.control_register.space_id != ACPI_ADR_SPACE_FIXED_HARDWARE) ||
+- (data->acpi_data.status_register.space_id != ACPI_ADR_SPACE_FIXED_HARDWARE)) {
++ if ((data->acpi_data->control_register.space_id != ACPI_ADR_SPACE_FIXED_HARDWARE) ||
++ (data->acpi_data->status_register.space_id != ACPI_ADR_SPACE_FIXED_HARDWARE)) {
+ dprintk("Invalid control/status registers (%x - %x)\n",
+- data->acpi_data.control_register.space_id,
+- data->acpi_data.status_register.space_id);
++ data->acpi_data->control_register.space_id,
++ data->acpi_data->status_register.space_id);
+ goto err_out;
+ }
+
+ /* fill in data->powernow_table */
+ powernow_table = kmalloc((sizeof(struct cpufreq_frequency_table)
+- * (data->acpi_data.state_count + 1)), GFP_KERNEL);
++ * (data->acpi_data->state_count + 1)), GFP_KERNEL);
+ if (!powernow_table) {
+ dprintk("powernow_table memory alloc failure\n");
+ goto err_out;
+@@ -808,17 +815,32 @@
+ if (ret_val)
+ goto err_out_mem;
+
+- powernow_table[data->acpi_data.state_count].frequency = CPUFREQ_TABLE_END;
+- powernow_table[data->acpi_data.state_count].index = 0;
++ powernow_table[data->acpi_data->state_count].frequency = CPUFREQ_TABLE_END;
++ powernow_table[data->acpi_data->state_count].index = 0;
+ data->powernow_table = powernow_table;
+
+ /* fill in data */
+- data->numps = data->acpi_data.state_count;
++ data->numps = data->acpi_data->state_count;
+ print_basics(data);
+ powernow_k8_acpi_pst_values(data, 0);
+
+ /* notify BIOS that we exist */
+ acpi_processor_notify_smm(THIS_MODULE);
++
++ /* determine affinity, from ACPI if available */
++ if (preregister_valid) {
++ if ((data->acpi_data->shared_type == CPUFREQ_SHARED_TYPE_ALL) ||
++ (data->acpi_data->shared_type == CPUFREQ_SHARED_TYPE_ANY))
++ data->starting_core_affinity = data->acpi_data->shared_cpu_map;
++ else
++ data->starting_core_affinity = cpumask_of_cpu(data->cpu);
++ } else {
++ /* best guess from family if not */
++ if (cpu_family == CPU_HW_PSTATE)
++ data->starting_core_affinity = cpumask_of_cpu(data->cpu);
++ else
++ data->starting_core_affinity = cpu_core_map[data->cpu];
++ }
+
+ return 0;
+
+@@ -826,10 +848,10 @@
+ kfree(powernow_table);
+
+ err_out:
+- acpi_processor_unregister_performance(&data->acpi_data, data->cpu);
++ acpi_processor_unregister_performance(data->acpi_data, data->cpu);
+
+- /* data->acpi_data.state_count informs us at ->exit() whether ACPI was used */
+- data->acpi_data.state_count = 0;
++ /* data->acpi_data->state_count informs us at ->exit() whether ACPI was used */
++ data->acpi_data->state_count = 0;
+
+ return -ENODEV;
+ }
+@@ -837,41 +859,23 @@
+ static int fill_powernow_table_pstate(struct powernow_k8_data *data, struct cpufreq_frequency_table *powernow_table)
+ {
+ int i;
++ u32 hi = 0, lo = 0;
++ rdmsr(MSR_PSTATE_CUR_LIMIT, hi, lo);
++ data->max_hw_pstate = (hi & HW_PSTATE_MAX_MASK) >> HW_PSTATE_MAX_SHIFT;
+
+- for (i = 0; i < data->acpi_data.state_count; i++) {
++ for (i = 0; i < data->acpi_data->state_count; i++) {
+ u32 index;
+- u32 hi = 0, lo = 0;
+- u32 fid;
+- u32 did;
+
+- index = data->acpi_data.states[i].control & HW_PSTATE_MASK;
+- if (index > MAX_HW_PSTATE) {
++ index = data->acpi_data->states[i].control & HW_PSTATE_MASK;
++ if (index > data->max_hw_pstate) {
+ printk(KERN_ERR PFX "invalid pstate %d - bad value %d.\n", i, index);
+ printk(KERN_ERR PFX "Please report to BIOS manufacturer\n");
+- }
+- rdmsr(MSR_PSTATE_DEF_BASE + index, lo, hi);
+- if (!(hi & HW_PSTATE_VALID_MASK)) {
+- dprintk("invalid pstate %d, ignoring\n", index);
+- powernow_table[i].frequency = CPUFREQ_ENTRY_INVALID;
+ continue;
+ }
+
+- fid = lo & HW_PSTATE_FID_MASK;
+- did = (lo & HW_PSTATE_DID_MASK) >> HW_PSTATE_DID_SHIFT;
++ powernow_table[i].index = index;
++ powernow_table[i].frequency = data->acpi_data->states[i].core_frequency * 1000;
+
+- dprintk(" %d : fid 0x%x, did 0x%x\n", index, fid, did);
+-
+- powernow_table[i].index = index | (fid << HW_FID_INDEX_SHIFT) | (did << HW_DID_INDEX_SHIFT);
+-
+- powernow_table[i].frequency = find_khz_freq_from_fiddid(fid, did);
+-
+- if (powernow_table[i].frequency != (data->acpi_data.states[i].core_frequency * 1000)) {
+- printk(KERN_INFO PFX "invalid freq entries %u kHz vs. %u kHz\n",
+- powernow_table[i].frequency,
+- (unsigned int) (data->acpi_data.states[i].core_frequency * 1000));
+- powernow_table[i].frequency = CPUFREQ_ENTRY_INVALID;
+- continue;
+- }
+ }
+ return 0;
+ }
+@@ -880,16 +884,16 @@
+ {
+ int i;
+ int cntlofreq = 0;
+- for (i = 0; i < data->acpi_data.state_count; i++) {
++ for (i = 0; i < data->acpi_data->state_count; i++) {
+ u32 fid;
+ u32 vid;
+
+ if (data->exttype) {
+- fid = data->acpi_data.states[i].status & EXT_FID_MASK;
+- vid = (data->acpi_data.states[i].status >> VID_SHIFT) & EXT_VID_MASK;
++ fid = data->acpi_data->states[i].status & EXT_FID_MASK;
++ vid = (data->acpi_data->states[i].status >> VID_SHIFT) & EXT_VID_MASK;
+ } else {
+- fid = data->acpi_data.states[i].control & FID_MASK;
+- vid = (data->acpi_data.states[i].control >> VID_SHIFT) & VID_MASK;
++ fid = data->acpi_data->states[i].control & FID_MASK;
++ vid = (data->acpi_data->states[i].control >> VID_SHIFT) & VID_MASK;
+ }
+
+ dprintk(" %d : fid 0x%x, vid 0x%x\n", i, fid, vid);
+@@ -930,10 +934,10 @@
+ cntlofreq = i;
+ }
+
+- if (powernow_table[i].frequency != (data->acpi_data.states[i].core_frequency * 1000)) {
++ if (powernow_table[i].frequency != (data->acpi_data->states[i].core_frequency * 1000)) {
+ printk(KERN_INFO PFX "invalid freq entries %u kHz vs. %u kHz\n",
+ powernow_table[i].frequency,
+- (unsigned int) (data->acpi_data.states[i].core_frequency * 1000));
++ (unsigned int) (data->acpi_data->states[i].core_frequency * 1000));
+ powernow_table[i].frequency = CPUFREQ_ENTRY_INVALID;
+ continue;
+ }
+@@ -943,14 +947,15 @@
+
+ static void powernow_k8_cpu_exit_acpi(struct powernow_k8_data *data)
+ {
+- if (data->acpi_data.state_count)
+- acpi_processor_unregister_performance(&data->acpi_data, data->cpu);
++ if (data->acpi_data->state_count)
++ acpi_processor_unregister_performance(data->acpi_data, data->cpu);
+ }
+
+ #else
+ static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data) { return -ENODEV; }
+ static void powernow_k8_cpu_exit_acpi(struct powernow_k8_data *data) { return; }
+ static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data, unsigned int index) { return; }
++static int powernow_k8_cpu_preinit_acpi() { return -ENODEV; }
+ #endif /* CONFIG_X86_POWERNOW_K8_ACPI */
+
+ /* Take a frequency, and issue the fid/vid transition command */
+@@ -1012,22 +1017,18 @@
+ /* Take a frequency, and issue the hardware pstate transition command */
+ static int transition_frequency_pstate(struct powernow_k8_data *data, unsigned int index)
+ {
+- u32 fid = 0;
+- u32 did = 0;
+ u32 pstate = 0;
+ int res, i;
+ struct cpufreq_freqs freqs;
+
+ dprintk("cpu %d transition to index %u\n", smp_processor_id(), index);
+
+- /* get fid did for hardware pstate transition */
++ /* get MSR index for hardware pstate transition */
+ pstate = index & HW_PSTATE_MASK;
+- if (pstate > MAX_HW_PSTATE)
++ if (pstate > data->max_hw_pstate)
+ return 0;
+- fid = (index & HW_FID_INDEX_MASK) >> HW_FID_INDEX_SHIFT;
+- did = (index & HW_DID_INDEX_MASK) >> HW_DID_INDEX_SHIFT;
+- freqs.old = find_khz_freq_from_fiddid(data->currfid, data->currdid);
+- freqs.new = find_khz_freq_from_fiddid(fid, did);
++ freqs.old = find_khz_freq_from_pstate(data->powernow_table, data->currpstate);
++ freqs.new = find_khz_freq_from_pstate(data->powernow_table, pstate);
+
+ for_each_cpu_mask(i, *(data->available_cores)) {
+ freqs.cpu = i;
+@@ -1035,9 +1036,7 @@
+ }
+
+ res = transition_pstate(data, pstate);
+- data->currfid = find_fid_from_pstate(pstate);
+- data->currdid = find_did_from_pstate(pstate);
+- freqs.new = find_khz_freq_from_fiddid(data->currfid, data->currdid);
++ freqs.new = find_khz_freq_from_pstate(data->powernow_table, pstate);
+
+ for_each_cpu_mask(i, *(data->available_cores)) {
+ freqs.cpu = i;
+@@ -1082,10 +1081,7 @@
+ if (query_current_values_with_pending_wait(data))
+ goto err_out;
+
+- if (cpu_family == CPU_HW_PSTATE)
+- dprintk("targ: curr fid 0x%x, did 0x%x\n",
+- data->currfid, data->currvid);
+- else {
++ if (cpu_family != CPU_HW_PSTATE) {
+ dprintk("targ: curr fid 0x%x, vid 0x%x\n",
+ data->currfid, data->currvid);
+
+@@ -1116,7 +1112,7 @@
+ mutex_unlock(&fidvid_mutex);
+
+ if (cpu_family == CPU_HW_PSTATE)
+- pol->cur = find_khz_freq_from_fiddid(data->currfid, data->currdid);
++ pol->cur = find_khz_freq_from_pstate(data->powernow_table, newstate);
+ else
+ pol->cur = find_khz_freq_from_fid(data->currfid);
+ ret = 0;
+@@ -1164,7 +1160,7 @@
+ * an UP version, and is deprecated by AMD.
+ */
+ if (num_online_cpus() != 1) {
+- printk(KERN_ERR PFX "MP systems not supported by PSB BIOS structure\n");
++ printk(KERN_ERR PFX "Your BIOS does not provide _PSS objects. PowerNow! does not work on SMP systems without _PSS objects. Complain to your BIOS vendor.\n");
+ kfree(data);
+ return -ENODEV;
+ }
+@@ -1204,10 +1200,7 @@
+ set_cpus_allowed(current, oldmask);
+
+ pol->governor = CPUFREQ_DEFAULT_GOVERNOR;
+- if (cpu_family == CPU_HW_PSTATE)
+- pol->cpus = cpumask_of_cpu(pol->cpu);
+- else
+- pol->cpus = cpu_core_map[pol->cpu];
++ pol->cpus = data->starting_core_affinity;
+ data->available_cores = &(pol->cpus);
+
+ /* Take a crude guess here.
+@@ -1216,7 +1209,7 @@
+ + (3 * (1 << data->irt) * 10)) * 1000;
+
+ if (cpu_family == CPU_HW_PSTATE)
+- pol->cur = find_khz_freq_from_fiddid(data->currfid, data->currdid);
++ pol->cur = find_khz_freq_from_pstate(data->powernow_table, data->currpstate);
+ else
+ pol->cur = find_khz_freq_from_fid(data->currfid);
+ dprintk("policy current frequency %d kHz\n", pol->cur);
+@@ -1233,8 +1226,7 @@
+ cpufreq_frequency_table_get_attr(data->powernow_table, pol->cpu);
+
+ if (cpu_family == CPU_HW_PSTATE)
+- dprintk("cpu_init done, current fid 0x%x, did 0x%x\n",
+- data->currfid, data->currdid);
++ dprintk("cpu_init done, current pstate 0x%x\n", data->currpstate);
+ else
+ dprintk("cpu_init done, current fid 0x%x, vid 0x%x\n",
+ data->currfid, data->currvid);
+@@ -1289,7 +1281,10 @@
+ if (query_current_values_with_pending_wait(data))
+ goto out;
+
+- khz = find_khz_freq_from_fid(data->currfid);
++ if (cpu_family == CPU_HW_PSTATE)
++ khz = find_khz_freq_from_pstate(data->powernow_table, data->currpstate);
++ else
++ khz = find_khz_freq_from_fid(data->currfid);
+
+ out:
+ set_cpus_allowed(current, oldmask);
+@@ -1323,6 +1318,7 @@
+ }
+
+ if (supported_cpus == num_online_cpus()) {
++ powernow_k8_cpu_preinit_acpi();
+ printk(KERN_INFO PFX "Found %d %s "
+ "processors (" VERSION ")\n", supported_cpus,
+ boot_cpu_data.x86_model_id);
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/cpu/cpufreq/powernow-k8.h
+--- a/arch/i386/kernel/cpu/cpufreq/powernow-k8.h Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/cpu/cpufreq/powernow-k8.h Wed Sep 10 10:54:08 2008 +0100
+@@ -1,5 +1,5 @@
+ /*
+- * (c) 2003-2006 Advanced Micro Devices, Inc.
++#* (c) 2003-2006 Advanced Micro Devices, Inc.
+ * Your use of this code is subject to the terms and conditions of the
+ * GNU general public license version 2. See "COPYING" or
+ * http://www.gnu.org/licenses/gpl.html
+@@ -10,6 +10,7 @@
+
+ u32 numps; /* number of p-states */
+ u32 batps; /* number of p-states supported on battery */
++ u32 max_hw_pstate; /* maximum legal hardware pstate */
+
+ /* these values are constant when the PSB is used to determine
+ * vid/fid pairings, but are modified during the ->target() call
+@@ -21,8 +22,8 @@
+ u32 plllock; /* pll lock time, units 1 us */
+ u32 exttype; /* extended interface = 1 */
+
+- /* keep track of the current fid / vid or did */
+- u32 currvid, currfid, currdid;
++ /* keep track of the current fid / vid or pstate */
++ u32 currvid, currfid, currpstate;
+
+ /* the powernow_table includes all frequency and vid/fid pairings:
+ * fid are the lower 8 bits of the index, vid are the upper 8 bits.
+@@ -32,12 +33,13 @@
+ #ifdef CONFIG_X86_POWERNOW_K8_ACPI
+ /* the acpi table needs to be kept. it's only available if ACPI was
+ * used to determine valid frequency/vid/fid states */
+- struct acpi_processor_performance acpi_data;
++ struct acpi_processor_performance *acpi_data;
+ #endif
+ /* we need to keep track of associated cores, but let cpufreq
+ * handle hotplug events - so just point at cpufreq pol->cpus
+ * structure */
+ cpumask_t *available_cores;
++ cpumask_t starting_core_affinity;
+ };
+
+
+@@ -87,23 +89,14 @@
+
+ /* Hardware Pstate _PSS and MSR definitions */
+ #define USE_HW_PSTATE 0x00000080
+-#define HW_PSTATE_FID_MASK 0x0000003f
+-#define HW_PSTATE_DID_MASK 0x000001c0
+-#define HW_PSTATE_DID_SHIFT 6
+-#define HW_PSTATE_MASK 0x00000007
+-#define HW_PSTATE_VALID_MASK 0x80000000
+-#define HW_FID_INDEX_SHIFT 8
+-#define HW_FID_INDEX_MASK 0x0000ff00
+-#define HW_DID_INDEX_SHIFT 16
+-#define HW_DID_INDEX_MASK 0x00ff0000
+-#define HW_WATTS_MASK 0xff
+-#define HW_PWR_DVR_MASK 0x300
+-#define HW_PWR_DVR_SHIFT 8
+-#define HW_PWR_MAX_MULT 3
+-#define MAX_HW_PSTATE 8 /* hw pstate supports up to 8 */
++#define HW_PSTATE_MASK 0x00000007
++#define HW_PSTATE_VALID_MASK 0x80000000
++#define HW_PSTATE_MAX_MASK 0x000000f0
++#define HW_PSTATE_MAX_SHIFT 4
+ #define MSR_PSTATE_DEF_BASE 0xc0010064 /* base of Pstate MSRs */
+ #define MSR_PSTATE_STATUS 0xc0010063 /* Pstate Status MSR */
+ #define MSR_PSTATE_CTRL 0xc0010062 /* Pstate control MSR */
++#define MSR_PSTATE_CUR_LIMIT 0xc0010061 /* pstate current limit MSR */
+
+ /* define the two driver architectures */
+ #define CPU_OPTERON 0
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/cpu/mtrr/Makefile
+--- a/arch/i386/kernel/cpu/mtrr/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/cpu/mtrr/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -3,3 +3,4 @@
+ obj-y += cyrix.o
+ obj-y += centaur.o
+
++obj-$(CONFIG_XEN) := main.o if.o
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/cpu/mtrr/main-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/cpu/mtrr/main-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,198 @@
+#include <linux/init.h>
+#include <linux/proc_fs.h>
+#include <linux/ctype.h>
@@ -2488,7 +3439,8 @@
+
+ op.cmd = XENPF_read_memtype;
+ op.u.read_memtype.reg = reg;
-+ (void)HYPERVISOR_platform_op(&op);
++ if (unlikely(HYPERVISOR_platform_op(&op)))
++ memset(&op.u.read_memtype, 0, sizeof(op.u.read_memtype));
+
+ *size = op.u.read_memtype.nr_mfns;
+ *base = op.u.read_memtype.mfn;
@@ -2665,24 +3617,10 @@
+}
+
+subsys_initcall(mtrr_init);
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/cpu/mtrr/Makefile tmp-linux-2.6-xen.patch/arch/i386/kernel/cpu/mtrr/Makefile
---- pristine-linux-2.6.18.2/arch/i386/kernel/cpu/mtrr/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/cpu/mtrr/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -3,3 +3,10 @@ obj-y += amd.o
- obj-y += cyrix.o
- obj-y += centaur.o
-
-+ifdef CONFIG_XEN
-+include $(srctree)/scripts/Makefile.xen
-+n-obj-xen := generic.o state.o amd.o cyrix.o centaur.o
-+
-+obj-y := $(call filterxen, $(obj-y), $(n-obj-xen))
-+obj-y := $(call cherrypickxen, $(obj-y))
-+endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/crash.c tmp-linux-2.6-xen.patch/arch/i386/kernel/crash.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/crash.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/crash.c 2007-07-30 16:35:11.000000000 +0200
-@@ -90,6 +90,7 @@ static void crash_save_self(struct pt_re
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/crash.c
+--- a/arch/i386/kernel/crash.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/crash.c Wed Sep 10 10:54:08 2008 +0100
+@@ -90,6 +90,7 @@
crash_save_this_cpu(regs, cpu);
}
@@ -2690,7 +3628,7 @@
#if defined(CONFIG_SMP) && defined(CONFIG_X86_LOCAL_APIC)
static atomic_t waiting_for_crash_ipi;
-@@ -154,6 +155,7 @@ static void nmi_shootdown_cpus(void)
+@@ -154,6 +155,7 @@
/* There are no cpus to shootdown */
}
#endif
@@ -2698,7 +3636,7 @@
void machine_crash_shutdown(struct pt_regs *regs)
{
-@@ -170,10 +172,12 @@ void machine_crash_shutdown(struct pt_re
+@@ -170,10 +172,12 @@
/* Make a note of crashing cpu. Will be used in NMI callback.*/
crashing_cpu = smp_processor_id();
@@ -2711,46 +3649,16 @@
+#endif /* CONFIG_XEN */
crash_save_self(regs);
}
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/early_printk-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/early_printk-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/early_printk-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/early_printk-xen.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/early_printk-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/early_printk-xen.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,2 @@
+
+#include "../../x86_64/kernel/early_printk-xen.c"
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/entry.S tmp-linux-2.6-xen.patch/arch/i386/kernel/entry.S
---- pristine-linux-2.6.18.2/arch/i386/kernel/entry.S 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/entry.S 2007-10-14 01:51:15.000000000 +0200
-@@ -269,7 +269,7 @@ ENTRY(sysenter_entry)
- CFI_STARTPROC simple
- CFI_DEF_CFA esp, 0
- CFI_REGISTER esp, ebp
-- movl TSS_sysenter_esp0(%esp),%esp
-+ movl SYSENTER_stack_esp0(%esp),%esp
- sysenter_past_esp:
- /*
- * No need to follow this irqs on/off section: the syscall
-@@ -689,7 +689,7 @@ device_not_available_emulate:
- * that sets up the real kernel stack. Check here, since we can't
- * allow the wrong stack to be used.
- *
-- * "TSS_sysenter_esp0+12" is because the NMI/debug handler will have
-+ * "SYSENTER_stack_esp0+12" is because the NMI/debug handler will have
- * already pushed 3 words if it hits on the sysenter instruction:
- * eflags, cs and eip.
- *
-@@ -701,7 +701,7 @@ device_not_available_emulate:
- cmpw $__KERNEL_CS,4(%esp); \
- jne ok; \
- label: \
-- movl TSS_sysenter_esp0+offset(%esp),%esp; \
-+ movl SYSENTER_stack_esp0+offset(%esp),%esp; \
- pushfl; \
- pushl $__KERNEL_CS; \
- pushl $sysenter_past_esp
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/entry-xen.S tmp-linux-2.6-xen.patch/arch/i386/kernel/entry-xen.S
---- pristine-linux-2.6.18.2/arch/i386/kernel/entry-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/entry-xen.S 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,1216 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/entry-xen.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/entry-xen.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1238 @@
+/*
+ * linux/arch/i386/entry.S
+ *
@@ -3134,6 +4042,29 @@
+#endif /* !CONFIG_XEN */
+ CFI_ENDPROC
+
++ # pv sysenter call handler stub
++ENTRY(sysenter_entry_pv)
++ RING0_INT_FRAME
++ movl $__USER_DS,16(%esp)
++ movl %ebp,12(%esp)
++ movl $__USER_CS,4(%esp)
++ addl $4,%esp
++ /* +5*4 is SS:ESP,EFLAGS,CS:EIP. +8 is esp0 setting. */
++ pushl (TI_sysenter_return-THREAD_SIZE+8+4*4)(%esp)
++/*
++ * Load the potential sixth argument from user stack.
++ * Careful about security.
++ */
++ cmpl $__PAGE_OFFSET-3,%ebp
++ jae syscall_fault
++1: movl (%ebp),%ebp
++.section __ex_table,"a"
++ .align 4
++ .long 1b,syscall_fault
++.previous
++ /* fall through */
++ CFI_ENDPROC
++ENDPROC(sysenter_entry_pv)
+
+ # system call handler stub
+ENTRY(system_call)
@@ -3963,13 +4894,42 @@
+ CFI_ENDPROC
+
+.section .rodata,"a"
-+.align 4
+#include "syscall_table.S"
+
+syscall_table_size=(.-sys_call_table)
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/fixup.c tmp-linux-2.6-xen.patch/arch/i386/kernel/fixup.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/fixup.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/fixup.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/entry.S
+--- a/arch/i386/kernel/entry.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/entry.S Wed Sep 10 10:54:08 2008 +0100
+@@ -269,7 +269,7 @@
+ CFI_STARTPROC simple
+ CFI_DEF_CFA esp, 0
+ CFI_REGISTER esp, ebp
+- movl TSS_sysenter_esp0(%esp),%esp
++ movl SYSENTER_stack_esp0(%esp),%esp
+ sysenter_past_esp:
+ /*
+ * No need to follow this irqs on/off section: the syscall
+@@ -687,7 +687,7 @@
+ * that sets up the real kernel stack. Check here, since we can't
+ * allow the wrong stack to be used.
+ *
+- * "TSS_sysenter_esp0+12" is because the NMI/debug handler will have
++ * "SYSENTER_stack_esp0+12" is because the NMI/debug handler will have
+ * already pushed 3 words if it hits on the sysenter instruction:
+ * eflags, cs and eip.
+ *
+@@ -699,7 +699,7 @@
+ cmpw $__KERNEL_CS,4(%esp); \
+ jne ok; \
+ label: \
+- movl TSS_sysenter_esp0+offset(%esp),%esp; \
++ movl SYSENTER_stack_esp0+offset(%esp),%esp; \
+ pushfl; \
+ pushl $__KERNEL_CS; \
+ pushl $sysenter_past_esp
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/fixup.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/fixup.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,88 @@
+/******************************************************************************
+ * fixup.c
@@ -4019,8 +4979,8 @@
+ if (current->tgid == 1)
+ return;
+
-+ HYPERVISOR_vm_assist(
-+ VMASST_CMD_disable, VMASST_TYPE_4gb_segments_notify);
++ VOID(HYPERVISOR_vm_assist(VMASST_CMD_disable,
++ VMASST_TYPE_4gb_segments_notify));
+
+ if (test_and_set_bit(0, &printed))
+ return;
@@ -4054,14 +5014,14 @@
+
+static int __init fixup_init(void)
+{
-+ HYPERVISOR_vm_assist(
-+ VMASST_CMD_enable, VMASST_TYPE_4gb_segments_notify);
++ WARN_ON(HYPERVISOR_vm_assist(VMASST_CMD_enable,
++ VMASST_TYPE_4gb_segments_notify));
+ return 0;
+}
+__initcall(fixup_init);
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/head-xen.S tmp-linux-2.6-xen.patch/arch/i386/kernel/head-xen.S
---- pristine-linux-2.6.18.2/arch/i386/kernel/head-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/head-xen.S 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/head-xen.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/head-xen.S Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,207 @@
+
+
@@ -4270,9 +5230,9 @@
+#endif
+ ELFNOTE(Xen, XEN_ELFNOTE_LOADER, .asciz, "generic")
+ ELFNOTE(Xen, XEN_ELFNOTE_SUSPEND_CANCEL, .long, 1)
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/init_task-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/init_task-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/init_task-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/init_task-xen.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/init_task-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/init_task-xen.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,51 @@
+#include <linux/mm.h>
+#include <linux/module.h>
@@ -4325,10 +5285,10 @@
+DEFINE_PER_CPU(struct tss_struct, init_tss) ____cacheline_internodealigned_in_smp = INIT_TSS;
+#endif
+
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/io_apic-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/io_apic-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/io_apic-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/io_apic-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,2777 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/io_apic-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/io_apic-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,2771 @@
+/*
+ * Intel IO-APIC support for multi-Pentium hosts.
+ *
@@ -4378,6 +5338,7 @@
+
+#include <xen/interface/xen.h>
+#include <xen/interface/physdev.h>
++#include <xen/evtchn.h>
+
+/* Fake i8259 */
+#define make_8259A_irq(_irq) (io_apic_irqs &= ~(1UL<<(_irq)))
@@ -4406,7 +5367,7 @@
+ apic_op.apic_physbase = mp_ioapics[apic].mpc_apicaddr;
+ apic_op.reg = reg;
+ apic_op.value = value;
-+ HYPERVISOR_physdev_op(PHYSDEVOP_apic_write, &apic_op);
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_apic_write, &apic_op));
+}
+
+#define io_apic_read(a,r) xen_io_apic_read(a,r)
@@ -5591,7 +6552,7 @@
+ set_intr_gate(vector, interrupt[idx]);
+}
+#else
-+#define ioapic_register_intr(_irq,_vector,_trigger) ((void)0)
++#define ioapic_register_intr(irq, vector, trigger) evtchn_register_pirq(irq)
+#endif
+
+static void __init setup_IO_APIC_irqs(void)
@@ -5869,8 +6830,6 @@
+ return;
+}
+
-+#if 0
-+
+static void print_APIC_bitfield (int base)
+{
+ unsigned int v;
@@ -6011,11 +6970,6 @@
+ v = inb(0x4d1) << 8 | inb(0x4d0);
+ printk(KERN_DEBUG "... PIC ELCR: %04x\n", v);
+}
-+
-+#endif /* 0 */
-+
-+#else
-+void __init print_IO_APIC(void) { }
+#endif /* !CONFIG_XEN */
+
+static void __init enable_IO_APIC(void)
@@ -6844,7 +7798,7 @@
+ struct xen_platform_op op = { .cmd = XENPF_platform_quirk };
+ op.u.platform_quirk.quirk_id = sis_apic_bug ?
+ QUIRK_IOAPIC_BAD_REGSEL : QUIRK_IOAPIC_GOOD_REGSEL;
-+ HYPERVISOR_platform_op(&op);
++ VOID(HYPERVISOR_platform_op(&op));
+ }
+ return 0;
+}
@@ -7106,10 +8060,10 @@
+}
+
+#endif /* CONFIG_ACPI */
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/ioport-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/ioport-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/ioport-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/ioport-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,122 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/ioport-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/ioport-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,123 @@
+/*
+ * linux/arch/i386/kernel/ioport.c
+ *
@@ -7195,7 +8149,8 @@
+
+ set_xen_guest_handle(set_iobitmap.bitmap, (char *)bitmap);
+ set_iobitmap.nr_ports = IO_BITMAP_BITS;
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap, &set_iobitmap);
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap,
++ &set_iobitmap));
+ }
+
+ set_bitmap(t->io_bitmap_ptr, from, num, !turn_on);
@@ -7232,9 +8187,9 @@
+ set_iopl_mask(t->iopl);
+ return 0;
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/irq-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/irq-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/irq-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/irq-xen.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/irq-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/irq-xen.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,324 @@
+/*
+ * linux/arch/i386/kernel/irq.c
@@ -7560,9 +8515,9 @@
+}
+#endif
+
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/ldt-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/ldt-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/ldt-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/ldt-xen.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/ldt-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/ldt-xen.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,270 @@
+/*
+ * linux/kernel/ldt.c
@@ -7834,64 +8789,102 @@
+ }
+ return ret;
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/machine_kexec.c tmp-linux-2.6-xen.patch/arch/i386/kernel/machine_kexec.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/machine_kexec.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/machine_kexec.c 2007-07-30 16:35:11.000000000 +0200
-@@ -19,123 +19,52 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/machine_kexec.c
+--- a/arch/i386/kernel/machine_kexec.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/machine_kexec.c Wed Sep 10 10:54:08 2008 +0100
+@@ -19,123 +19,67 @@
#include <asm/desc.h>
#include <asm/system.h>
--#define PAGE_ALIGNED __attribute__ ((__aligned__(PAGE_SIZE)))
--
++#ifdef CONFIG_XEN
++#include <xen/interface/kexec.h>
++#endif
++
+ #define PAGE_ALIGNED __attribute__ ((__aligned__(PAGE_SIZE)))
++static u32 kexec_pgd[1024] PAGE_ALIGNED;
++#ifdef CONFIG_X86_PAE
++static u32 kexec_pmd0[1024] PAGE_ALIGNED;
++static u32 kexec_pmd1[1024] PAGE_ALIGNED;
++#endif
++static u32 kexec_pte0[1024] PAGE_ALIGNED;
++static u32 kexec_pte1[1024] PAGE_ALIGNED;
+
-#define L0_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY)
-#define L1_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY)
-#define L2_ATTR (_PAGE_PRESENT)
--
++#ifdef CONFIG_XEN
+
-#define LEVEL0_SIZE (1UL << 12UL)
--
++#define __ma(x) (pfn_to_mfn(__pa((x)) >> PAGE_SHIFT) << PAGE_SHIFT)
+
-#ifndef CONFIG_X86_PAE
-#define LEVEL1_SIZE (1UL << 22UL)
-static u32 pgtable_level1[1024] PAGE_ALIGNED;
--
++#if PAGES_NR > KEXEC_XEN_NO_PAGES
++#error PAGES_NR is greater than KEXEC_XEN_NO_PAGES - Xen support will break
++#endif
+
-static void identity_map_page(unsigned long address)
--{
++#if PA_CONTROL_PAGE != 0
++#error PA_CONTROL_PAGE is non zero - Xen support will break
++#endif
++
++void machine_kexec_setup_load_arg(xen_kexec_image_t *xki, struct kimage *image)
+ {
- unsigned long level1_index, level2_index;
- u32 *pgtable_level2;
--
++ void *control_page;
+
- /* Find the current page table */
- pgtable_level2 = __va(read_cr3());
--
++ memset(xki->page_list, 0, sizeof(xki->page_list));
+
- /* Find the indexes of the physical address to identity map */
- level1_index = (address % LEVEL1_SIZE)/LEVEL0_SIZE;
- level2_index = address / LEVEL1_SIZE;
--
++ control_page = page_address(image->control_code_page);
++ memcpy(control_page, relocate_kernel, PAGE_SIZE);
+
- /* Identity map the page table entry */
- pgtable_level1[level1_index] = address | L0_ATTR;
- pgtable_level2[level2_index] = __pa(pgtable_level1) | L1_ATTR;
--
++ xki->page_list[PA_CONTROL_PAGE] = __ma(control_page);
++ xki->page_list[PA_PGD] = __ma(kexec_pgd);
++#ifdef CONFIG_X86_PAE
++ xki->page_list[PA_PMD_0] = __ma(kexec_pmd0);
++ xki->page_list[PA_PMD_1] = __ma(kexec_pmd1);
++#endif
++ xki->page_list[PA_PTE_0] = __ma(kexec_pte0);
++ xki->page_list[PA_PTE_1] = __ma(kexec_pte1);
+
- /* Flush the tlb so the new mapping takes effect.
- * Global tlb entries are not flushed but that is not an issue.
- */
- load_cr3(pgtable_level2);
--}
--
+ }
+
-#else
-#define LEVEL1_SIZE (1UL << 21UL)
-#define LEVEL2_SIZE (1UL << 30UL)
-static u64 pgtable_level1[512] PAGE_ALIGNED;
-static u64 pgtable_level2[512] PAGE_ALIGNED;
--
++int __init machine_kexec_setup_resources(struct resource *hypervisor,
++ struct resource *phys_cpus,
++ int nr_phys_cpus)
++{
++ int k;
+
-static void identity_map_page(unsigned long address)
-{
- unsigned long level1_index, level2_index, level3_index;
- u64 *pgtable_level3;
--
++ /* The per-cpu crash note resources belong to the hypervisor resource */
++ for (k = 0; k < nr_phys_cpus; k++)
++ request_resource(hypervisor, phys_cpus + k);
+
- /* Find the current page table */
- pgtable_level3 = __va(read_cr3());
-+#ifdef CONFIG_XEN
-+#include <xen/interface/kexec.h>
-+#endif
-
+-
- /* Find the indexes of the physical address to identity map */
- level1_index = (address % LEVEL1_SIZE)/LEVEL0_SIZE;
- level2_index = (address % LEVEL2_SIZE)/LEVEL1_SIZE;
@@ -7908,64 +8901,36 @@
- */
- load_cr3(pgtable_level3);
-}
-+#define PAGE_ALIGNED __attribute__ ((__aligned__(PAGE_SIZE)))
-+static u32 kexec_pgd[1024] PAGE_ALIGNED;
-+#ifdef CONFIG_X86_PAE
-+static u32 kexec_pmd0[1024] PAGE_ALIGNED;
-+static u32 kexec_pmd1[1024] PAGE_ALIGNED;
- #endif
-+static u32 kexec_pte0[1024] PAGE_ALIGNED;
-+static u32 kexec_pte1[1024] PAGE_ALIGNED;
-
+-#endif
+-
-static void set_idt(void *newidt, __u16 limit)
-{
- struct Xgt_desc_struct curidt;
-+#ifdef CONFIG_XEN
-
+-
- /* ia32 supports unaliged loads & stores */
- curidt.size = limit;
- curidt.address = (unsigned long)newidt;
-+#define __ma(x) (pfn_to_mfn(__pa((x)) >> PAGE_SHIFT) << PAGE_SHIFT)
-
+-
- load_idt(&curidt);
-};
-+#if PAGES_NR > KEXEC_XEN_NO_PAGES
-+#error PAGES_NR is greater than KEXEC_XEN_NO_PAGES - Xen support will break
-+#endif
-
-+#if PA_CONTROL_PAGE != 0
-+#error PA_CONTROL_PAGE is non zero - Xen support will break
-+#endif
-
+-
+-
-static void set_gdt(void *newgdt, __u16 limit)
-+void machine_kexec_setup_load_arg(xen_kexec_image_t *xki, struct kimage *image)
- {
+-{
- struct Xgt_desc_struct curgdt;
-+ void *control_page;
-
+-
- /* ia32 supports unaligned loads & stores */
- curgdt.size = limit;
- curgdt.address = (unsigned long)newgdt;
-+ memset(xki->page_list, 0, sizeof(xki->page_list));
-
+-
- load_gdt(&curgdt);
-};
-+ control_page = page_address(image->control_code_page);
-+ memcpy(control_page, relocate_kernel, PAGE_SIZE);
-
+-
-static void load_segments(void)
-{
-#define __STR(X) #X
-#define STR(X) __STR(X)
-+ xki->page_list[PA_CONTROL_PAGE] = __ma(control_page);
-+ xki->page_list[PA_PGD] = __ma(kexec_pgd);
-+#ifdef CONFIG_X86_PAE
-+ xki->page_list[PA_PMD_0] = __ma(kexec_pmd0);
-+ xki->page_list[PA_PMD_1] = __ma(kexec_pmd1);
-+#endif
-+ xki->page_list[PA_PTE_0] = __ma(kexec_pte0);
-+ xki->page_list[PA_PTE_1] = __ma(kexec_pte1);
-
+-
- __asm__ __volatile__ (
- "\tljmp $"STR(__KERNEL_CS)",$1f\n"
- "\t1:\n"
@@ -7978,6 +8943,7 @@
- ::: "eax", "memory");
-#undef STR
-#undef __STR
++ return 0;
}
-typedef asmlinkage NORET_TYPE void (*relocate_new_kernel_t)(
@@ -7985,7 +8951,8 @@
- unsigned long reboot_code_buffer,
- unsigned long start_address,
- unsigned int has_pae) ATTRIB_NORET;
--
++void machine_kexec_register_resources(struct resource *res) { ; }
+
-extern const unsigned char relocate_new_kernel[];
-extern void relocate_new_kernel_end(void);
-extern const unsigned int relocate_new_kernel_size;
@@ -7993,7 +8960,7 @@
/*
* A architecture hook called to validate the
-@@ -163,49 +92,38 @@ void machine_kexec_cleanup(struct kimage
+@@ -163,49 +107,38 @@
{
}
@@ -8018,10 +8985,26 @@
- reboot_code_buffer = page_to_pfn(image->control_code_page)
- << PAGE_SHIFT;
- page_list = image->head;
--
++ control_page = page_address(image->control_code_page);
++ memcpy(control_page, relocate_kernel, PAGE_SIZE);
+
- /* Set up an identity mapping for the reboot_code_buffer */
- identity_map_page(reboot_code_buffer);
--
++ page_list[PA_CONTROL_PAGE] = __pa(control_page);
++ page_list[VA_CONTROL_PAGE] = (unsigned long)relocate_kernel;
++ page_list[PA_PGD] = __pa(kexec_pgd);
++ page_list[VA_PGD] = (unsigned long)kexec_pgd;
++#ifdef CONFIG_X86_PAE
++ page_list[PA_PMD_0] = __pa(kexec_pmd0);
++ page_list[VA_PMD_0] = (unsigned long)kexec_pmd0;
++ page_list[PA_PMD_1] = __pa(kexec_pmd1);
++ page_list[VA_PMD_1] = (unsigned long)kexec_pmd1;
++#endif
++ page_list[PA_PTE_0] = __pa(kexec_pte0);
++ page_list[VA_PTE_0] = (unsigned long)kexec_pte0;
++ page_list[PA_PTE_1] = __pa(kexec_pte1);
++ page_list[VA_PTE_1] = (unsigned long)kexec_pte1;
+
- /* copy it out */
- memcpy((void *)reboot_code_buffer, relocate_new_kernel,
- relocate_new_kernel_size);
@@ -8045,78 +9028,13 @@
- /* now call it */
- rnk = (relocate_new_kernel_t) reboot_code_buffer;
- (*rnk)(page_list, reboot_code_buffer, image->start, cpu_has_pae);
-+ control_page = page_address(image->control_code_page);
-+ memcpy(control_page, relocate_kernel, PAGE_SIZE);
-+
-+ page_list[PA_CONTROL_PAGE] = __pa(control_page);
-+ page_list[VA_CONTROL_PAGE] = (unsigned long)relocate_kernel;
-+ page_list[PA_PGD] = __pa(kexec_pgd);
-+ page_list[VA_PGD] = (unsigned long)kexec_pgd;
-+#ifdef CONFIG_X86_PAE
-+ page_list[PA_PMD_0] = __pa(kexec_pmd0);
-+ page_list[VA_PMD_0] = (unsigned long)kexec_pmd0;
-+ page_list[PA_PMD_1] = __pa(kexec_pmd1);
-+ page_list[VA_PMD_1] = (unsigned long)kexec_pmd1;
-+#endif
-+ page_list[PA_PTE_0] = __pa(kexec_pte0);
-+ page_list[VA_PTE_0] = (unsigned long)kexec_pte0;
-+ page_list[PA_PTE_1] = __pa(kexec_pte1);
-+ page_list[VA_PTE_1] = (unsigned long)kexec_pte1;
-+
+ relocate_kernel((unsigned long)image->head, (unsigned long)page_list,
+ image->start, cpu_has_pae);
}
+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/Makefile tmp-linux-2.6-xen.patch/arch/i386/kernel/Makefile
---- pristine-linux-2.6.18.2/arch/i386/kernel/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -44,6 +44,12 @@ EXTRA_AFLAGS := -traditional
-
- obj-$(CONFIG_SCx200) += scx200.o
-
-+ifdef CONFIG_XEN
-+vsyscall_note := vsyscall-note-xen.o
-+else
-+vsyscall_note := vsyscall-note.o
-+endif
-+
- # vsyscall.o contains the vsyscall DSO images as __initdata.
- # We must build both images before we can assemble it.
- # Note: kbuild does not track this dependency due to usage of .incbin
-@@ -65,7 +71,7 @@ SYSCFLAGS_vsyscall-int80.so = $(vsyscall
-
- $(obj)/vsyscall-int80.so $(obj)/vsyscall-sysenter.so: \
- $(obj)/vsyscall-%.so: $(src)/vsyscall.lds \
-- $(obj)/vsyscall-%.o $(obj)/vsyscall-note.o FORCE
-+ $(obj)/vsyscall-%.o $(obj)/$(vsyscall_note) FORCE
- $(call if_changed,syscall)
-
- # We also create a special relocatable object that should mirror the symbol
-@@ -77,8 +83,20 @@ $(obj)/built-in.o: ld_flags += -R $(obj)
-
- SYSCFLAGS_vsyscall-syms.o = -r
- $(obj)/vsyscall-syms.o: $(src)/vsyscall.lds \
-- $(obj)/vsyscall-sysenter.o $(obj)/vsyscall-note.o FORCE
-+ $(obj)/vsyscall-sysenter.o $(obj)/$(vsyscall_note) FORCE
- $(call if_changed,syscall)
-
- k8-y += ../../x86_64/kernel/k8.o
-
-+ifdef CONFIG_XEN
-+include $(srctree)/scripts/Makefile.xen
-+
-+obj-y += fixup.o
-+microcode-$(subst m,y,$(CONFIG_MICROCODE)) := microcode-xen.o
-+n-obj-xen := i8259.o timers/ reboot.o smpboot.o trampoline.o
-+
-+obj-y := $(call filterxen, $(obj-y), $(n-obj-xen))
-+obj-y := $(call cherrypickxen, $(obj-y))
-+extra-y := $(call cherrypickxen, $(extra-y))
-+%/head-xen.o %/head-xen.s: EXTRA_AFLAGS :=
-+endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/microcode-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/microcode-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/microcode-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/microcode-xen.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/microcode-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/microcode-xen.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,144 @@
+/*
+ * Intel CPU Microcode Update Driver for Linux
@@ -8262,9 +9180,9 @@
+module_init(microcode_init)
+module_exit(microcode_exit)
+MODULE_ALIAS_MISCDEV(MICROCODE_MINOR);
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/mpparse-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/mpparse-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/mpparse-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/mpparse-xen.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/mpparse-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/mpparse-xen.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,1185 @@
+/*
+ * Intel Multiprocessor Specification 1.1 and 1.4
@@ -9451,10 +10369,10 @@
+
+#endif /* CONFIG_X86_IO_APIC */
+#endif /* CONFIG_ACPI */
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/pci-dma-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/pci-dma-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/pci-dma-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/pci-dma-xen.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,369 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/pci-dma-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/pci-dma-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,415 @@
+/*
+ * Dynamic DMA mapping support.
+ *
@@ -9472,9 +10390,11 @@
+#include <linux/version.h>
+#include <asm/io.h>
+#include <xen/balloon.h>
++#include <xen/gnttab.h>
+#include <asm/swiotlb.h>
+#include <asm/tlbflush.h>
+#include <asm-i386/mach-xen/asm/swiotlb.h>
++#include <asm-i386/mach-xen/asm/gnttab_dma.h>
+#include <asm/bug.h>
+
+#ifdef __x86_64__
@@ -9532,6 +10452,39 @@
+ } \
+} while (0)
+
++static int check_pages_physically_contiguous(unsigned long pfn,
++ unsigned int offset,
++ size_t length)
++{
++ unsigned long next_mfn;
++ int i;
++ int nr_pages;
++
++ next_mfn = pfn_to_mfn(pfn);
++ nr_pages = (offset + length + PAGE_SIZE-1) >> PAGE_SHIFT;
++
++ for (i = 1; i < nr_pages; i++) {
++ if (pfn_to_mfn(++pfn) != ++next_mfn)
++ return 0;
++ }
++ return 1;
++}
++
++int range_straddles_page_boundary(paddr_t p, size_t size)
++{
++ extern unsigned long *contiguous_bitmap;
++ unsigned long pfn = p >> PAGE_SHIFT;
++ unsigned int offset = p & ~PAGE_MASK;
++
++ if (offset + size <= PAGE_SIZE)
++ return 0;
++ if (test_bit(pfn, contiguous_bitmap))
++ return 0;
++ if (check_pages_physically_contiguous(pfn, offset, size))
++ return 0;
++ return 1;
++}
++
+int
+dma_map_sg(struct device *hwdev, struct scatterlist *sg, int nents,
+ enum dma_data_direction direction)
@@ -9546,10 +10499,10 @@
+ rc = swiotlb_map_sg(hwdev, sg, nents, direction);
+ } else {
+ for (i = 0; i < nents; i++ ) {
++ BUG_ON(!sg[i].page);
+ sg[i].dma_address =
-+ page_to_bus(sg[i].page) + sg[i].offset;
++ gnttab_dma_map_page(sg[i].page) + sg[i].offset;
+ sg[i].dma_length = sg[i].length;
-+ BUG_ON(!sg[i].page);
+ IOMMU_BUG_ON(address_needs_mapping(
+ hwdev, sg[i].dma_address));
+ IOMMU_BUG_ON(range_straddles_page_boundary(
@@ -9568,9 +10521,15 @@
+dma_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nents,
+ enum dma_data_direction direction)
+{
++ int i;
++
+ BUG_ON(direction == DMA_NONE);
+ if (swiotlb)
+ swiotlb_unmap_sg(hwdev, sg, nents, direction);
++ else {
++ for (i = 0; i < nents; i++ )
++ gnttab_dma_unmap_page(sg[i].dma_address);
++ }
+}
+EXPORT_SYMBOL(dma_unmap_sg);
+
@@ -9587,7 +10546,7 @@
+ dma_addr = swiotlb_map_page(
+ dev, page, offset, size, direction);
+ } else {
-+ dma_addr = page_to_bus(page) + offset;
++ dma_addr = gnttab_dma_map_page(page) + offset;
+ IOMMU_BUG_ON(address_needs_mapping(dev, dma_addr));
+ }
+
@@ -9602,6 +10561,8 @@
+ BUG_ON(direction == DMA_NONE);
+ if (swiotlb)
+ swiotlb_unmap_page(dev, dma_address, size, direction);
++ else
++ gnttab_dma_unmap_page(dma_address);
+}
+EXPORT_SYMBOL(dma_unmap_page);
+#endif /* CONFIG_HIGHMEM */
@@ -9786,7 +10747,8 @@
+ if (swiotlb) {
+ dma = swiotlb_map_single(dev, ptr, size, direction);
+ } else {
-+ dma = virt_to_bus(ptr);
++ dma = gnttab_dma_map_page(virt_to_page(ptr)) +
++ offset_in_page(ptr);
+ IOMMU_BUG_ON(range_straddles_page_boundary(__pa(ptr), size));
+ IOMMU_BUG_ON(address_needs_mapping(dev, dma));
+ }
@@ -9804,6 +10766,8 @@
+ BUG();
+ if (swiotlb)
+ swiotlb_unmap_single(dev, dma_addr, size, direction);
++ else
++ gnttab_dma_unmap_page(dma_addr);
+}
+EXPORT_SYMBOL(dma_unmap_single);
+
@@ -9824,10 +10788,10 @@
+ swiotlb_sync_single_for_device(dev, dma_handle, size, direction);
+}
+EXPORT_SYMBOL(dma_sync_single_for_device);
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/process-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/process-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/process-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/process-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,853 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/process-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/process-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,877 @@
+/*
+ * linux/arch/i386/kernel/process.c
+ *
@@ -9968,7 +10932,7 @@
+ local_irq_disable();
+ cpu_clear(smp_processor_id(), cpu_initialized);
+ preempt_enable_no_resched();
-+ HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL);
++ VOID(HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL));
+ cpu_bringup();
+}
+#else
@@ -10139,7 +11103,8 @@
+ struct thread_struct *t = &tsk->thread;
+ struct physdev_set_iobitmap set_iobitmap;
+ memset(&set_iobitmap, 0, sizeof(set_iobitmap));
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap, &set_iobitmap);
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap,
++ &set_iobitmap));
+ kfree(t->io_bitmap_ptr);
+ t->io_bitmap_ptr = NULL;
+ clear_thread_flag(TIF_IO_BITMAP);
@@ -10381,8 +11346,14 @@
+#ifndef CONFIG_X86_NO_TSS
+ struct tss_struct *tss = &per_cpu(init_tss, cpu);
+#endif
++#if CONFIG_XEN_COMPAT > 0x030002
+ struct physdev_set_iopl iopl_op;
+ struct physdev_set_iobitmap iobmp_op;
++#else
++ struct physdev_op _pdo[2], *pdo = _pdo;
++#define iopl_op pdo->u.set_iopl
++#define iobmp_op pdo->u.set_iobitmap
++#endif
+ multicall_entry_t _mcl[8], *mcl = _mcl;
+
+ /* XEN NOTE: FS/GS saved in switch_mm(), not here. */
@@ -10430,9 +11401,15 @@
+
+ if (unlikely(prev->iopl != next->iopl)) {
+ iopl_op.iopl = (next->iopl == 0) ? 1 : (next->iopl >> 12) & 3;
++#if CONFIG_XEN_COMPAT > 0x030002
+ mcl->op = __HYPERVISOR_physdev_op;
+ mcl->args[0] = PHYSDEVOP_set_iopl;
+ mcl->args[1] = (unsigned long)&iopl_op;
++#else
++ mcl->op = __HYPERVISOR_physdev_op_compat;
++ pdo->cmd = PHYSDEVOP_set_iopl;
++ mcl->args[0] = (unsigned long)pdo++;
++#endif
+ mcl++;
+ }
+
@@ -10440,13 +11417,24 @@
+ set_xen_guest_handle(iobmp_op.bitmap,
+ (char *)next->io_bitmap_ptr);
+ iobmp_op.nr_ports = next->io_bitmap_ptr ? IO_BITMAP_BITS : 0;
++#if CONFIG_XEN_COMPAT > 0x030002
+ mcl->op = __HYPERVISOR_physdev_op;
+ mcl->args[0] = PHYSDEVOP_set_iobitmap;
+ mcl->args[1] = (unsigned long)&iobmp_op;
++#else
++ mcl->op = __HYPERVISOR_physdev_op_compat;
++ pdo->cmd = PHYSDEVOP_set_iobitmap;
++ mcl->args[0] = (unsigned long)pdo++;
++#endif
+ mcl++;
+ }
+
-+ (void)HYPERVISOR_multicall(_mcl, mcl - _mcl);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ BUG_ON(pdo > _pdo + ARRAY_SIZE(_pdo));
++#endif
++ BUG_ON(mcl > _mcl + ARRAY_SIZE(_mcl));
++ if (unlikely(HYPERVISOR_multicall_check(_mcl, mcl - _mcl, NULL)))
++ BUG();
+
+ /*
+ * Restore %fs and %gs if needed.
@@ -10681,9 +11669,9 @@
+ sp -= get_random_int() % 8192;
+ return sp & ~0xf;
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/quirks-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/quirks-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/quirks-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/quirks-xen.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/quirks-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/quirks-xen.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,47 @@
+/*
+ * This file contains work-arounds for x86 and x86_64 platform bugs.
@@ -10721,7 +11709,7 @@
+ printk(KERN_INFO "Disabling irq balancing and affinity\n");
+ op.cmd = XENPF_platform_quirk;
+ op.u.platform_quirk.quirk_id = QUIRK_NOIRQBALANCING;
-+ (void)HYPERVISOR_platform_op(&op);
++ WARN_ON(HYPERVISOR_platform_op(&op));
+ }
+
+ /* put back the original value for config space*/
@@ -10732,16 +11720,21 @@
+DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_E7525_MCH, quirk_intel_irqbalance);
+DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_E7520_MCH, quirk_intel_irqbalance);
+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/relocate_kernel.S tmp-linux-2.6-xen.patch/arch/i386/kernel/relocate_kernel.S
---- pristine-linux-2.6.18.2/arch/i386/kernel/relocate_kernel.S 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/relocate_kernel.S 2007-10-14 01:51:15.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/relocate_kernel.S
+--- a/arch/i386/kernel/relocate_kernel.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/relocate_kernel.S Wed Sep 10 10:54:08 2008 +0100
@@ -7,16 +7,138 @@
*/
#include <linux/linkage.h>
+#include <asm/page.h>
+#include <asm/kexec.h>
-+
+
+- /*
+- * Must be relocatable PIC code callable as a C function, that once
+- * it starts can not use the previous processes stack.
+- */
+- .globl relocate_new_kernel
+/*
+ * Must be relocatable PIC code callable as a C function
+ */
@@ -10866,12 +11859,7 @@
+ orl $PAGE_ATTR, %edx
+ movl %edx, (%eax)
+#endif
-
-- /*
-- * Must be relocatable PIC code callable as a C function, that once
-- * it starts can not use the previous processes stack.
-- */
-- .globl relocate_new_kernel
++
relocate_new_kernel:
/* read the arguments and say goodbye to the stack */
movl 4(%esp), %ebx /* page_list */
@@ -10880,7 +11868,7 @@
movl 12(%esp), %edx /* start address */
movl 16(%esp), %ecx /* cpu_has_pae */
-@@ -24,11 +146,57 @@ relocate_new_kernel:
+@@ -24,11 +146,57 @@
pushl $0
popfl
@@ -10942,7 +11930,7 @@
/* Set cr0 to a known state:
* 31 0 == Paging disabled
-@@ -113,8 +281,20 @@ relocate_new_kernel:
+@@ -113,8 +281,20 @@
xorl %edi, %edi
xorl %ebp, %ebp
ret
@@ -10967,10 +11955,10 @@
+idt_48:
+ .word 0 /* limit */
+ .long 0 /* base */
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/setup-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/setup-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/setup-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/setup-xen.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,1898 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/setup-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/setup-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1924 @@
+/*
+ * linux/arch/i386/kernel/setup.c
+ *
@@ -11039,6 +12027,7 @@
+#include <xen/interface/physdev.h>
+#include <xen/interface/memory.h>
+#include <xen/features.h>
++#include <xen/firmware.h>
+#include <xen/xencons.h>
+#include <setup_arch.h>
+#include <bios_ebda.h>
@@ -11128,6 +12117,9 @@
+};
+struct edid_info edid_info;
+EXPORT_SYMBOL_GPL(edid_info);
++#ifndef CONFIG_XEN
++#define copy_edid() (edid_info = EDID_INFO)
++#endif
+struct ist_info ist_info;
+#if defined(CONFIG_X86_SPEEDSTEP_SMI) || \
+ defined(CONFIG_X86_SPEEDSTEP_SMI_MODULE)
@@ -11252,7 +12244,12 @@
+}, {
+ .name = "keyboard",
+ .start = 0x0060,
-+ .end = 0x006f,
++ .end = 0x0060,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO
++}, {
++ .name = "keyboard",
++ .start = 0x0064,
++ .end = 0x0064,
+ .flags = IORESOURCE_BUSY | IORESOURCE_IO
+}, {
+ .name = "dma page reg",
@@ -11705,6 +12702,21 @@
+#endif
+ add_memory_region(start, size, type);
+ } while (biosmap++,--nr_map);
++
++#ifdef CONFIG_XEN
++ if (is_initial_xendomain()) {
++ struct xen_memory_map memmap;
++
++ memmap.nr_entries = E820MAX;
++ set_xen_guest_handle(memmap.buffer, machine_e820.map);
++
++ if (HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap))
++ BUG();
++ machine_e820.nr_map = memmap.nr_entries;
++ } else
++ machine_e820 = e820;
++#endif
++
+ return 0;
+}
+
@@ -11713,6 +12725,7 @@
+#ifdef CONFIG_EDD_MODULE
+EXPORT_SYMBOL(edd);
+#endif
++#ifndef CONFIG_XEN
+/**
+ * copy_edd() - Copy the BIOS EDD information
+ * from boot_params into a safe place.
@@ -11725,6 +12738,7 @@
+ edd.mbr_signature_nr = EDD_MBR_SIG_NR;
+ edd.edd_info_nr = EDD_NR;
+}
++#endif
+#else
+static inline void copy_edd(void)
+{
@@ -11995,6 +13009,35 @@
+ return 0;
+}
+
++/*
++ * This function checks if any part of the range <start,end> is mapped
++ * with type.
++ */
++int
++e820_any_mapped(u64 start, u64 end, unsigned type)
++{
++ int i;
++
++#ifndef CONFIG_XEN
++ for (i = 0; i < e820.nr_map; i++) {
++ const struct e820entry *ei = &e820.map[i];
++#else
++ if (!is_initial_xendomain())
++ return 0;
++ for (i = 0; i < machine_e820.nr_map; ++i) {
++ const struct e820entry *ei = &machine_e820.map[i];
++#endif
++
++ if (type && ei->type != type)
++ continue;
++ if (ei->addr >= end || ei->addr + ei->size <= start)
++ continue;
++ return 1;
++ }
++ return 0;
++}
++EXPORT_SYMBOL_GPL(e820_any_mapped);
++
+ /*
+ * This function checks if the entire range <start,end> is mapped with type.
+ *
@@ -12243,14 +13286,7 @@
+ unsigned long zones_size[MAX_NR_ZONES] = {0, 0, 0};
+ unsigned int max_dma, low;
+
-+ /*
-+ * XEN: Our notion of "DMA memory" is fake when running over Xen.
-+ * We simply put all RAM in the DMA zone so that those drivers which
-+ * needlessly specify GFP_DMA do not get starved of RAM unnecessarily.
-+ * Those drivers that *do* require lowmem are screwed anyway when
-+ * running over Xen!
-+ */
-+ max_dma = max_low_pfn;
++ max_dma = virt_to_phys((char *)MAX_DMA_ADDRESS) >> PAGE_SHIFT;
+ low = max_low_pfn;
+
+ if (low < max_dma)
@@ -12347,10 +13383,6 @@
+ crashk_res.end - crashk_res.start + 1);
+#endif
+#endif
-+
-+ if (!xen_feature(XENFEAT_auto_translated_physmap))
-+ phys_to_machine_mapping =
-+ (unsigned long *)xen_start_info->mfn_list;
+}
+
+/*
@@ -12518,18 +13550,8 @@
+static void __init register_memory(void)
+{
+#ifdef CONFIG_XEN
-+ if (is_initial_xendomain()) {
-+ struct xen_memory_map memmap;
-+
-+ memmap.nr_entries = E820MAX;
-+ set_xen_guest_handle(memmap.buffer, machine_e820.map);
-+
-+ if (HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap))
-+ BUG();
-+
-+ machine_e820.nr_map = memmap.nr_entries;
++ if (is_initial_xendomain())
+ e820_setup_gap(machine_e820.map, machine_e820.nr_map);
-+ }
+ else
+#endif
+ e820_setup_gap(e820.map, e820.nr_map);
@@ -12566,11 +13588,13 @@
+ /* Register a call for panic conditions. */
+ atomic_notifier_chain_register(&panic_notifier_list, &xen_panic_block);
+
-+ HYPERVISOR_vm_assist(VMASST_CMD_enable, VMASST_TYPE_4gb_segments);
-+ HYPERVISOR_vm_assist(VMASST_CMD_enable,
-+ VMASST_TYPE_writable_pagetables);
++ WARN_ON(HYPERVISOR_vm_assist(VMASST_CMD_enable,
++ VMASST_TYPE_4gb_segments));
++ WARN_ON(HYPERVISOR_vm_assist(VMASST_CMD_enable,
++ VMASST_TYPE_writable_pagetables));
+
+ memcpy(&boot_cpu_data, &new_cpu_data, sizeof(new_cpu_data));
++ pre_setup_arch_hook();
+ early_cpu_init();
+#ifdef CONFIG_SMP
+ prefill_possible_map();
@@ -12594,7 +13618,7 @@
+ ROOT_DEV = MKDEV(UNNAMED_MAJOR,0);
+ drive_info = DRIVE_INFO;
+ screen_info = SCREEN_INFO;
-+ edid_info = EDID_INFO;
++ copy_edid();
+ apm_info.bios = APM_BIOS_INFO;
+ ist_info = IST_INFO;
+ saved_videomode = VIDEO_MODE;
@@ -12607,23 +13631,12 @@
+ bootloader_type = LOADER_TYPE;
+
+ if (is_initial_xendomain()) {
-+ /* This is drawn from a dump from vgacon:startup in
-+ * standard Linux. */
-+ screen_info.orig_video_mode = 3;
-+ screen_info.orig_video_isVGA = 1;
-+ screen_info.orig_video_lines = 25;
-+ screen_info.orig_video_cols = 80;
-+ screen_info.orig_video_ega_bx = 3;
-+ screen_info.orig_video_points = 16;
-+ screen_info.orig_y = screen_info.orig_video_lines - 1;
-+ if (xen_start_info->console.dom0.info_size >=
-+ sizeof(struct dom0_vga_console_info)) {
-+ const struct dom0_vga_console_info *info =
-+ (struct dom0_vga_console_info *)(
-+ (char *)xen_start_info +
-+ xen_start_info->console.dom0.info_off);
-+ dom0_init_screen_info(info);
-+ }
++ const struct dom0_vga_console_info *info =
++ (void *)((char *)xen_start_info +
++ xen_start_info->console.dom0.info_off);
++
++ dom0_init_screen_info(info,
++ xen_start_info->console.dom0.info_size);
+ xen_start_info->console.domU.mfn = 0;
+ xen_start_info->console.domU.evtchn = 0;
+ } else
@@ -12635,8 +13648,6 @@
+ rd_doload = ((RAMDISK_FLAGS & RAMDISK_LOAD_FLAG) != 0);
+#endif
+
-+ setup_xen_features();
-+
+ ARCH_SETUP
+ if (efi_enabled)
+ efi_init();
@@ -12764,6 +13775,11 @@
+ virt_to_mfn(pfn_to_mfn_frame_list_list);
+ }
+
++ /* Mark all ISA DMA channels in-use - using them wouldn't work. */
++ for (i = 0; i < MAX_DMA_CHANNELS; ++i)
++ if (i != 4 && request_dma(i, "xen") != 0)
++ BUG();
++
+ /*
+ * NOTE: at this point the bootmem allocator is fully available.
+ */
@@ -12778,7 +13794,7 @@
+ efi_map_memmap();
+
+ set_iopl.iopl = 1;
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl));
+
+#ifdef CONFIG_ACPI
+ if (!is_initial_xendomain()) {
@@ -12830,8 +13846,6 @@
+#endif
+ }
+ tsc_init();
-+
-+ xencons_early_setup();
+}
+
+static int
@@ -12869,10 +13883,56 @@
+ * c-basic-offset:8
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/smp-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/smp-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/smp-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/smp-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,624 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/setup.c
+--- a/arch/i386/kernel/setup.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/setup.c Wed Sep 10 10:54:08 2008 +0100
+@@ -261,7 +261,12 @@
+ }, {
+ .name = "keyboard",
+ .start = 0x0060,
+- .end = 0x006f,
++ .end = 0x0060,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO
++}, {
++ .name = "keyboard",
++ .start = 0x0064,
++ .end = 0x0064,
+ .flags = IORESOURCE_BUSY | IORESOURCE_IO
+ }, {
+ .name = "dma page reg",
+@@ -955,6 +960,28 @@
+ memory_present(0, start, end);
+ return 0;
+ }
++
++/*
++ * This function checks if any part of the range <start,end> is mapped
++ * with type.
++ */
++int
++e820_any_mapped(u64 start, u64 end, unsigned type)
++{
++ int i;
++
++ for (i = 0; i < e820.nr_map; i++) {
++ const struct e820entry *ei = &e820.map[i];
++
++ if (type && ei->type != type)
++ continue;
++ if (ei->addr >= end || ei->addr + ei->size <= start)
++ continue;
++ return 1;
++ }
++ return 0;
++}
++EXPORT_SYMBOL_GPL(e820_any_mapped);
+
+ /*
+ * This function checks if the entire range <start,end> is mapped with type.
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/smp-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/smp-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,605 @@
+/*
+ * Intel SMP support routines.
+ *
@@ -13316,21 +14376,6 @@
+ on_each_cpu(do_flush_tlb_all, NULL, 1, 1);
+}
+
-+#else
-+
-+irqreturn_t smp_invalidate_interrupt(int irq, void *dev_id,
-+ struct pt_regs *regs)
-+{ return 0; }
-+void flush_tlb_current_task(void)
-+{ xen_tlb_flush_mask(¤t->mm->cpu_vm_mask); }
-+void flush_tlb_mm(struct mm_struct * mm)
-+{ xen_tlb_flush_mask(&mm->cpu_vm_mask); }
-+void flush_tlb_page(struct vm_area_struct *vma, unsigned long va)
-+{ xen_invlpg_mask(&vma->vm_mm->cpu_vm_mask, va); }
-+EXPORT_SYMBOL(flush_tlb_page);
-+void flush_tlb_all(void)
-+{ xen_tlb_flush_all(); }
-+
+#endif /* XEN */
+
+/*
@@ -13415,11 +14460,11 @@
+
+ /* Wait for response */
+ while (atomic_read(&data.started) != cpus)
-+ barrier();
++ cpu_relax();
+
+ if (wait)
+ while (atomic_read(&data.finished) != cpus)
-+ barrier();
++ cpu_relax();
+ spin_unlock(&call_lock);
+
+ return 0;
@@ -13433,9 +14478,7 @@
+ */
+ cpu_clear(smp_processor_id(), cpu_online_map);
+ local_irq_disable();
-+#if 0
-+ disable_local_APIC();
-+#endif
++ disable_all_local_evtchn();
+ if (cpu_data[smp_processor_id()].hlt_works_ok)
+ for(;;) halt();
+ for (;;);
@@ -13450,9 +14493,7 @@
+ smp_call_function(stop_this_cpu, NULL, 1, 0);
+
+ local_irq_disable();
-+#if 0
-+ disable_local_APIC();
-+#endif
++ disable_all_local_evtchn();
+ local_irq_enable();
+}
+
@@ -13497,5856 +14538,5247 @@
+ return IRQ_HANDLED;
+}
+
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/swiotlb.c tmp-linux-2.6-xen.patch/arch/i386/kernel/swiotlb.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/swiotlb.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/swiotlb.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,734 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/sysenter.c
+--- a/arch/i386/kernel/sysenter.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/sysenter.c Wed Sep 10 10:54:08 2008 +0100
+@@ -23,6 +23,10 @@
+ #include <asm/pgtable.h>
+ #include <asm/unistd.h>
+
++#ifdef CONFIG_XEN
++#include <xen/interface/callback.h>
++#endif
++
+ /*
+ * Should the kernel map a VDSO page into processes and pass its
+ * address down to glibc upon exec()?
+@@ -44,6 +48,7 @@
+
+ void enable_sep_cpu(void)
+ {
++#ifndef CONFIG_XEN
+ int cpu = get_cpu();
+ struct tss_struct *tss = &per_cpu(init_tss, cpu);
+
+@@ -57,7 +62,36 @@
+ wrmsr(MSR_IA32_SYSENTER_CS, __KERNEL_CS, 0);
+ wrmsr(MSR_IA32_SYSENTER_ESP, tss->esp1, 0);
+ wrmsr(MSR_IA32_SYSENTER_EIP, (unsigned long) sysenter_entry, 0);
+- put_cpu();
++#else
++ extern asmlinkage void sysenter_entry_pv(void);
++ static struct callback_register sysenter = {
++ .type = CALLBACKTYPE_sysenter,
++ .address = { __KERNEL_CS, (unsigned long)sysenter_entry_pv },
++ };
++
++ if (!boot_cpu_has(X86_FEATURE_SEP))
++ return;
++
++ get_cpu();
++
++ if (xen_feature(XENFEAT_supervisor_mode_kernel))
++ sysenter.address.eip = (unsigned long)sysenter_entry;
++
++ switch (HYPERVISOR_callback_op(CALLBACKOP_register, &sysenter)) {
++ case 0:
++ break;
++#if CONFIG_XEN_COMPAT < 0x030200
++ case -ENOSYS:
++ sysenter.type = CALLBACKTYPE_sysenter_deprecated;
++ if (HYPERVISOR_callback_op(CALLBACKOP_register, &sysenter) == 0)
++ break;
++#endif
++ default:
++ clear_bit(X86_FEATURE_SEP, boot_cpu_data.x86_capability);
++ break;
++ }
++#endif
++ put_cpu();
+ }
+
+ /*
+@@ -75,11 +109,6 @@
+ #ifdef CONFIG_COMPAT_VDSO
+ __set_fixmap(FIX_VDSO, __pa(syscall_page), PAGE_READONLY);
+ printk("Compat vDSO mapped to %08lx.\n", __fix_to_virt(FIX_VDSO));
+-#else
+- /*
+- * In the non-compat case the ELF coredumping code needs the fixmap:
+- */
+- __set_fixmap(FIX_VDSO, __pa(syscall_page), PAGE_KERNEL_RO);
+ #endif
+
+ if (!boot_cpu_has(X86_FEATURE_SEP)) {
+@@ -142,6 +171,13 @@
+ vma->vm_end = addr + PAGE_SIZE;
+ /* MAYWRITE to allow gdb to COW and set breakpoints */
+ vma->vm_flags = VM_READ|VM_EXEC|VM_MAYREAD|VM_MAYEXEC|VM_MAYWRITE;
++ /*
++ * Make sure the vDSO gets into every core dump.
++ * Dumping its contents makes post-mortem fully interpretable later
++ * without matching up the same kernel and hardware config to see
++ * what PC values meant.
++ */
++ vma->vm_flags |= VM_ALWAYSDUMP;
+ vma->vm_flags |= mm->def_flags;
+ vma->vm_page_prot = protection_map[vma->vm_flags & 7];
+ vma->vm_ops = &syscall_vm_ops;
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/time-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/time-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1201 @@
+/*
-+ * Dynamic DMA mapping support.
++ * linux/arch/i386/kernel/time.c
+ *
-+ * This implementation is a fallback for platforms that do not support
-+ * I/O TLBs (aka DMA address translation hardware).
-+ * Copyright (C) 2000 Asit Mallick <Asit.K.Mallick at intel.com>
-+ * Copyright (C) 2000 Goutham Rao <goutham.rao at intel.com>
-+ * Copyright (C) 2000, 2003 Hewlett-Packard Co
-+ * David Mosberger-Tang <davidm at hpl.hp.com>
-+ * Copyright (C) 2005 Keir Fraser <keir at xensource.com>
++ * Copyright (C) 1991, 1992, 1995 Linus Torvalds
++ *
++ * This file contains the PC-specific time handling details:
++ * reading the RTC at bootup, etc..
++ * 1994-07-02 Alan Modra
++ * fixed set_rtc_mmss, fixed time.year for >= 2000, new mktime
++ * 1995-03-26 Markus Kuhn
++ * fixed 500 ms bug at call to set_rtc_mmss, fixed DS12887
++ * precision CMOS clock update
++ * 1996-05-03 Ingo Molnar
++ * fixed time warps in do_[slow|fast]_gettimeoffset()
++ * 1997-09-10 Updated NTP code according to technical memorandum Jan '96
++ * "A Kernel Model for Precision Timekeeping" by Dave Mills
++ * 1998-09-05 (Various)
++ * More robust do_fast_gettimeoffset() algorithm implemented
++ * (works with APM, Cyrix 6x86MX and Centaur C6),
++ * monotonic gettimeofday() with fast_get_timeoffset(),
++ * drift-proof precision TSC calibration on boot
++ * (C. Scott Ananian <cananian at alumni.princeton.edu>, Andrew D.
++ * Balsa <andrebalsa at altern.org>, Philip Gladstone <philip at raptor.com>;
++ * ported from 2.0.35 Jumbo-9 by Michael Krause <m.krause at tu-harburg.de>).
++ * 1998-12-16 Andrea Arcangeli
++ * Fixed Jumbo-9 code in 2.1.131: do_gettimeofday was missing 1 jiffy
++ * because was not accounting lost_ticks.
++ * 1998-12-24 Copyright (C) 1998 Andrea Arcangeli
++ * Fixed a xtime SMP race (we need the xtime_lock rw spinlock to
++ * serialize accesses to xtime/lost_ticks).
+ */
+
-+#include <linux/cache.h>
-+#include <linux/mm.h>
-+#include <linux/module.h>
-+#include <linux/pci.h>
-+#include <linux/spinlock.h>
++#include <linux/errno.h>
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/param.h>
+#include <linux/string.h>
-+#include <linux/types.h>
-+#include <linux/ctype.h>
++#include <linux/mm.h>
++#include <linux/interrupt.h>
++#include <linux/time.h>
++#include <linux/delay.h>
+#include <linux/init.h>
-+#include <linux/bootmem.h>
-+#include <linux/highmem.h>
++#include <linux/smp.h>
++#include <linux/module.h>
++#include <linux/sysdev.h>
++#include <linux/bcd.h>
++#include <linux/efi.h>
++#include <linux/mca.h>
++#include <linux/sysctl.h>
++#include <linux/percpu.h>
++#include <linux/kernel_stat.h>
++#include <linux/posix-timers.h>
++#include <linux/cpufreq.h>
++
+#include <asm/io.h>
-+#include <asm/pci.h>
-+#include <asm/dma.h>
++#include <asm/smp.h>
++#include <asm/irq.h>
++#include <asm/msr.h>
++#include <asm/delay.h>
++#include <asm/mpspec.h>
+#include <asm/uaccess.h>
-+#include <xen/interface/memory.h>
++#include <asm/processor.h>
++#include <asm/timer.h>
++#include <asm/sections.h>
+
-+int swiotlb;
-+EXPORT_SYMBOL(swiotlb);
++#include "mach_time.h"
+
-+#define OFFSET(val,align) ((unsigned long)((val) & ( (align) - 1)))
++#include <linux/timex.h>
+
-+#define SG_ENT_PHYS_ADDRESS(sg) (page_to_bus((sg)->page) + (sg)->offset)
++#include <asm/hpet.h>
+
-+/*
-+ * Maximum allowable number of contiguous slabs to map,
-+ * must be a power of 2. What is the appropriate value ?
-+ * The complexity of {map,unmap}_single is linearly dependent on this value.
-+ */
-+#define IO_TLB_SEGSIZE 128
++#include <asm/arch_hooks.h>
+
-+/*
-+ * log of the size of each IO TLB slab. The number of slabs is command line
-+ * controllable.
-+ */
-+#define IO_TLB_SHIFT 11
++#include <xen/evtchn.h>
++#include <xen/interface/vcpu.h>
+
-+int swiotlb_force;
++#if defined (__i386__)
++#include <asm/i8259.h>
++#endif
+
-+static char *iotlb_virt_start;
-+static unsigned long iotlb_nslabs;
++int pit_latch_buggy; /* extern */
+
-+/*
-+ * Used to do a quick range check in swiotlb_unmap_single and
-+ * swiotlb_sync_single_*, to see if the memory was in fact allocated by this
-+ * API.
-+ */
-+static unsigned long iotlb_pfn_start, iotlb_pfn_end;
++#if defined(__x86_64__)
++unsigned long vxtime_hz = PIT_TICK_RATE;
++struct vxtime_data __vxtime __section_vxtime; /* for vsyscalls */
++volatile unsigned long __jiffies __section_jiffies = INITIAL_JIFFIES;
++unsigned long __wall_jiffies __section_wall_jiffies = INITIAL_JIFFIES;
++struct timespec __xtime __section_xtime;
++struct timezone __sys_tz __section_sys_tz;
++#endif
+
-+/* Does the given dma address reside within the swiotlb aperture? */
-+static inline int in_swiotlb_aperture(dma_addr_t dev_addr)
-+{
-+ unsigned long pfn = mfn_to_local_pfn(dev_addr >> PAGE_SHIFT);
-+ return (pfn_valid(pfn)
-+ && (pfn >= iotlb_pfn_start)
-+ && (pfn < iotlb_pfn_end));
-+}
++unsigned int cpu_khz; /* Detected as we calibrate the TSC */
++EXPORT_SYMBOL(cpu_khz);
+
-+/*
-+ * When the IOMMU overflows we return a fallback buffer. This sets the size.
-+ */
-+static unsigned long io_tlb_overflow = 32*1024;
++extern unsigned long wall_jiffies;
+
-+void *io_tlb_overflow_buffer;
++DEFINE_SPINLOCK(rtc_lock);
++EXPORT_SYMBOL(rtc_lock);
+
-+/*
-+ * This is a free list describing the number of free entries available from
-+ * each index
-+ */
-+static unsigned int *io_tlb_list;
-+static unsigned int io_tlb_index;
++extern struct init_timer_opts timer_tsc_init;
++extern struct timer_opts timer_tsc;
++#define timer_none timer_tsc
+
-+/*
-+ * We need to save away the original address corresponding to a mapped entry
-+ * for the sync operations.
-+ */
-+static struct phys_addr {
-+ struct page *page;
-+ unsigned int offset;
-+} *io_tlb_orig_addr;
++/* These are peridically updated in shared_info, and then copied here. */
++struct shadow_time_info {
++ u64 tsc_timestamp; /* TSC at last update of time vals. */
++ u64 system_timestamp; /* Time, in nanosecs, since boot. */
++ u32 tsc_to_nsec_mul;
++ u32 tsc_to_usec_mul;
++ int tsc_shift;
++ u32 version;
++};
++static DEFINE_PER_CPU(struct shadow_time_info, shadow_time);
++static struct timespec shadow_tv;
++static u32 shadow_tv_version;
+
-+/*
-+ * Protect the above data structures in the map and unmap calls
-+ */
-+static DEFINE_SPINLOCK(io_tlb_lock);
++static struct timeval monotonic_tv;
++static spinlock_t monotonic_lock = SPIN_LOCK_UNLOCKED;
+
-+static unsigned int dma_bits;
-+static unsigned int __initdata max_dma_bits = 32;
-+static int __init
-+setup_dma_bits(char *str)
++/* Keep track of last time we did processing/updating of jiffies and xtime. */
++static u64 processed_system_time; /* System time (ns) at last processing. */
++static DEFINE_PER_CPU(u64, processed_system_time);
++
++/* How much CPU time was spent blocked and how much was 'stolen'? */
++static DEFINE_PER_CPU(u64, processed_stolen_time);
++static DEFINE_PER_CPU(u64, processed_blocked_time);
++
++/* Current runstate of each CPU (updated automatically by the hypervisor). */
++static DEFINE_PER_CPU(struct vcpu_runstate_info, runstate);
++
++/* Must be signed, as it's compared with s64 quantities which can be -ve. */
++#define NS_PER_TICK (1000000000LL/HZ)
++
++static void __clock_was_set(void *unused)
+{
-+ max_dma_bits = simple_strtoul(str, NULL, 0);
-+ return 0;
++ clock_was_set();
+}
-+__setup("dma_bits=", setup_dma_bits);
++static DECLARE_WORK(clock_was_set_work, __clock_was_set, NULL);
+
-+static int __init
-+setup_io_tlb_npages(char *str)
++static inline void __normalize_time(time_t *sec, s64 *nsec)
+{
-+ /* Unlike ia64, the size is aperture in megabytes, not 'slabs'! */
-+ if (isdigit(*str)) {
-+ iotlb_nslabs = simple_strtoul(str, &str, 0) <<
-+ (20 - IO_TLB_SHIFT);
-+ iotlb_nslabs = ALIGN(iotlb_nslabs, IO_TLB_SEGSIZE);
-+ /* Round up to power of two (xen_create_contiguous_region). */
-+ while (iotlb_nslabs & (iotlb_nslabs-1))
-+ iotlb_nslabs += iotlb_nslabs & ~(iotlb_nslabs-1);
++ while (*nsec >= NSEC_PER_SEC) {
++ (*nsec) -= NSEC_PER_SEC;
++ (*sec)++;
+ }
-+ if (*str == ',')
-+ ++str;
-+ /*
-+ * NB. 'force' enables the swiotlb, but doesn't force its use for
-+ * every DMA like it does on native Linux. 'off' forcibly disables
-+ * use of the swiotlb.
-+ */
-+ if (!strcmp(str, "force"))
-+ swiotlb_force = 1;
-+ else if (!strcmp(str, "off"))
-+ swiotlb_force = -1;
++ while (*nsec < 0) {
++ (*nsec) += NSEC_PER_SEC;
++ (*sec)--;
++ }
++}
++
++/* Does this guest OS track Xen time, or set its wall clock independently? */
++static int independent_wallclock = 0;
++static int __init __independent_wallclock(char *str)
++{
++ independent_wallclock = 1;
+ return 1;
+}
-+__setup("swiotlb=", setup_io_tlb_npages);
-+/* make io_tlb_overflow tunable too? */
++__setup("independent_wallclock", __independent_wallclock);
+
-+/*
-+ * Statically reserve bounce buffer space and initialize bounce buffer data
-+ * structures for the software IO TLB used to implement the PCI DMA API.
-+ */
-+void
-+swiotlb_init_with_default_size (size_t default_size)
++/* Permitted clock jitter, in nsecs, beyond which a warning will be printed. */
++static unsigned long permitted_clock_jitter = 10000000UL; /* 10ms */
++static int __init __permitted_clock_jitter(char *str)
+{
-+ unsigned long i, bytes;
-+ int rc;
++ permitted_clock_jitter = simple_strtoul(str, NULL, 0);
++ return 1;
++}
++__setup("permitted_clock_jitter=", __permitted_clock_jitter);
+
-+ if (!iotlb_nslabs) {
-+ iotlb_nslabs = (default_size >> IO_TLB_SHIFT);
-+ iotlb_nslabs = ALIGN(iotlb_nslabs, IO_TLB_SEGSIZE);
-+ /* Round up to power of two (xen_create_contiguous_region). */
-+ while (iotlb_nslabs & (iotlb_nslabs-1))
-+ iotlb_nslabs += iotlb_nslabs & ~(iotlb_nslabs-1);
-+ }
++#if 0
++static void delay_tsc(unsigned long loops)
++{
++ unsigned long bclock, now;
+
-+ bytes = iotlb_nslabs * (1UL << IO_TLB_SHIFT);
++ rdtscl(bclock);
++ do {
++ rep_nop();
++ rdtscl(now);
++ } while ((now - bclock) < loops);
++}
+
-+ /*
-+ * Get IO TLB memory from the low pages
-+ */
-+ iotlb_virt_start = alloc_bootmem_low_pages(bytes);
-+ if (!iotlb_virt_start)
-+ panic("Cannot allocate SWIOTLB buffer!\n");
++struct timer_opts timer_tsc = {
++ .name = "tsc",
++ .delay = delay_tsc,
++};
++#endif
+
-+ dma_bits = get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT) + PAGE_SHIFT;
-+ for (i = 0; i < iotlb_nslabs; i += IO_TLB_SEGSIZE) {
-+ do {
-+ rc = xen_create_contiguous_region(
-+ (unsigned long)iotlb_virt_start + (i << IO_TLB_SHIFT),
-+ get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT),
-+ dma_bits);
-+ } while (rc && dma_bits++ < max_dma_bits);
-+ if (rc) {
-+ if (i == 0)
-+ panic("No suitable physical memory available for SWIOTLB buffer!\n"
-+ "Use dom0_mem Xen boot parameter to reserve\n"
-+ "some DMA memory (e.g., dom0_mem=-128M).\n");
-+ iotlb_nslabs = i;
-+ i <<= IO_TLB_SHIFT;
-+ free_bootmem(__pa(iotlb_virt_start + i), bytes - i);
-+ bytes = i;
-+ for (dma_bits = 0; i > 0; i -= IO_TLB_SEGSIZE << IO_TLB_SHIFT) {
-+ unsigned int bits = fls64(virt_to_bus(iotlb_virt_start + i - 1));
++/*
++ * Scale a 64-bit delta by scaling and multiplying by a 32-bit fraction,
++ * yielding a 64-bit result.
++ */
++static inline u64 scale_delta(u64 delta, u32 mul_frac, int shift)
++{
++ u64 product;
++#ifdef __i386__
++ u32 tmp1, tmp2;
++#endif
+
-+ if (bits > dma_bits)
-+ dma_bits = bits;
-+ }
-+ break;
-+ }
-+ }
++ if (shift < 0)
++ delta >>= -shift;
++ else
++ delta <<= shift;
+
-+ /*
-+ * Allocate and initialize the free list array. This array is used
-+ * to find contiguous free memory regions of size up to IO_TLB_SEGSIZE.
-+ */
-+ io_tlb_list = alloc_bootmem(iotlb_nslabs * sizeof(int));
-+ for (i = 0; i < iotlb_nslabs; i++)
-+ io_tlb_list[i] = IO_TLB_SEGSIZE - OFFSET(i, IO_TLB_SEGSIZE);
-+ io_tlb_index = 0;
-+ io_tlb_orig_addr = alloc_bootmem(
-+ iotlb_nslabs * sizeof(*io_tlb_orig_addr));
++#ifdef __i386__
++ __asm__ (
++ "mul %5 ; "
++ "mov %4,%%eax ; "
++ "mov %%edx,%4 ; "
++ "mul %5 ; "
++ "xor %5,%5 ; "
++ "add %4,%%eax ; "
++ "adc %5,%%edx ; "
++ : "=A" (product), "=r" (tmp1), "=r" (tmp2)
++ : "a" ((u32)delta), "1" ((u32)(delta >> 32)), "2" (mul_frac) );
++#else
++ __asm__ (
++ "mul %%rdx ; shrd $32,%%rdx,%%rax"
++ : "=a" (product) : "0" (delta), "d" ((u64)mul_frac) );
++#endif
+
-+ /*
-+ * Get the overflow emergency buffer
-+ */
-+ io_tlb_overflow_buffer = alloc_bootmem_low(io_tlb_overflow);
-+ if (!io_tlb_overflow_buffer)
-+ panic("Cannot allocate SWIOTLB overflow buffer!\n");
++ return product;
++}
+
-+ do {
-+ rc = xen_create_contiguous_region(
-+ (unsigned long)io_tlb_overflow_buffer,
-+ get_order(io_tlb_overflow),
-+ dma_bits);
-+ } while (rc && dma_bits++ < max_dma_bits);
-+ if (rc)
-+ panic("No suitable physical memory available for SWIOTLB overflow buffer!\n");
++#if 0 /* defined (__i386__) */
++int read_current_timer(unsigned long *timer_val)
++{
++ rdtscl(*timer_val);
++ return 0;
++}
++#endif
+
-+ iotlb_pfn_start = __pa(iotlb_virt_start) >> PAGE_SHIFT;
-+ iotlb_pfn_end = iotlb_pfn_start + (bytes >> PAGE_SHIFT);
++void init_cpu_khz(void)
++{
++ u64 __cpu_khz = 1000000ULL << 32;
++ struct vcpu_time_info *info = &vcpu_info(0)->time;
++ do_div(__cpu_khz, info->tsc_to_system_mul);
++ if (info->tsc_shift < 0)
++ cpu_khz = __cpu_khz << -info->tsc_shift;
++ else
++ cpu_khz = __cpu_khz >> info->tsc_shift;
++}
+
-+ printk(KERN_INFO "Software IO TLB enabled: \n"
-+ " Aperture: %lu megabytes\n"
-+ " Kernel range: %p - %p\n"
-+ " Address size: %u bits\n",
-+ bytes >> 20,
-+ iotlb_virt_start, iotlb_virt_start + bytes,
-+ dma_bits);
++static u64 get_nsec_offset(struct shadow_time_info *shadow)
++{
++ u64 now, delta;
++ rdtscll(now);
++ delta = now - shadow->tsc_timestamp;
++ return scale_delta(delta, shadow->tsc_to_nsec_mul, shadow->tsc_shift);
+}
+
-+void
-+swiotlb_init(void)
++static unsigned long get_usec_offset(struct shadow_time_info *shadow)
+{
-+ long ram_end;
-+ size_t defsz = 64 * (1 << 20); /* 64MB default size */
++ u64 now, delta;
++ rdtscll(now);
++ delta = now - shadow->tsc_timestamp;
++ return scale_delta(delta, shadow->tsc_to_usec_mul, shadow->tsc_shift);
++}
+
-+ if (swiotlb_force == 1) {
-+ swiotlb = 1;
-+ } else if ((swiotlb_force != -1) &&
-+ is_running_on_xen() &&
-+ is_initial_xendomain()) {
-+ /* Domain 0 always has a swiotlb. */
-+ ram_end = HYPERVISOR_memory_op(XENMEM_maximum_ram_page, NULL);
-+ if (ram_end <= 0x7ffff)
-+ defsz = 2 * (1 << 20); /* 2MB on <2GB on systems. */
-+ swiotlb = 1;
-+ }
++static void __update_wallclock(time_t sec, long nsec)
++{
++ long wtm_nsec, xtime_nsec;
++ time_t wtm_sec, xtime_sec;
++ u64 tmp, wc_nsec;
+
-+ if (swiotlb)
-+ swiotlb_init_with_default_size(defsz);
-+ else
-+ printk(KERN_INFO "Software IO TLB disabled\n");
++ /* Adjust wall-clock time base based on wall_jiffies ticks. */
++ wc_nsec = processed_system_time;
++ wc_nsec += sec * (u64)NSEC_PER_SEC;
++ wc_nsec += nsec;
++ wc_nsec -= (jiffies - wall_jiffies) * (u64)NS_PER_TICK;
++
++ /* Split wallclock base into seconds and nanoseconds. */
++ tmp = wc_nsec;
++ xtime_nsec = do_div(tmp, 1000000000);
++ xtime_sec = (time_t)tmp;
++
++ wtm_sec = wall_to_monotonic.tv_sec + (xtime.tv_sec - xtime_sec);
++ wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - xtime_nsec);
++
++ set_normalized_timespec(&xtime, xtime_sec, xtime_nsec);
++ set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec);
++
++ ntp_clear();
+}
+
-+/*
-+ * We use __copy_to_user_inatomic to transfer to the host buffer because the
-+ * buffer may be mapped read-only (e.g, in blkback driver) but lower-level
-+ * drivers map the buffer for DMA_BIDIRECTIONAL access. This causes an
-+ * unnecessary copy from the aperture to the host buffer, and a page fault.
-+ */
-+static void
-+__sync_single(struct phys_addr buffer, char *dma_addr, size_t size, int dir)
++static void update_wallclock(void)
+{
-+ if (PageHighMem(buffer.page)) {
-+ size_t len, bytes;
-+ char *dev, *host, *kmp;
-+ len = size;
-+ while (len != 0) {
-+ unsigned long flags;
++ shared_info_t *s = HYPERVISOR_shared_info;
+
-+ if (((bytes = len) + buffer.offset) > PAGE_SIZE)
-+ bytes = PAGE_SIZE - buffer.offset;
-+ local_irq_save(flags); /* protects KM_BOUNCE_READ */
-+ kmp = kmap_atomic(buffer.page, KM_BOUNCE_READ);
-+ dev = dma_addr + size - len;
-+ host = kmp + buffer.offset;
-+ if (dir == DMA_FROM_DEVICE) {
-+ if (__copy_to_user_inatomic(host, dev, bytes))
-+ /* inaccessible */;
-+ } else
-+ memcpy(dev, host, bytes);
-+ kunmap_atomic(kmp, KM_BOUNCE_READ);
-+ local_irq_restore(flags);
-+ len -= bytes;
-+ buffer.page++;
-+ buffer.offset = 0;
-+ }
-+ } else {
-+ char *host = (char *)phys_to_virt(
-+ page_to_pseudophys(buffer.page)) + buffer.offset;
-+ if (dir == DMA_FROM_DEVICE) {
-+ if (__copy_to_user_inatomic(host, dma_addr, size))
-+ /* inaccessible */;
-+ } else if (dir == DMA_TO_DEVICE)
-+ memcpy(dma_addr, host, size);
-+ }
++ do {
++ shadow_tv_version = s->wc_version;
++ rmb();
++ shadow_tv.tv_sec = s->wc_sec;
++ shadow_tv.tv_nsec = s->wc_nsec;
++ rmb();
++ } while ((s->wc_version & 1) | (shadow_tv_version ^ s->wc_version));
++
++ if (!independent_wallclock)
++ __update_wallclock(shadow_tv.tv_sec, shadow_tv.tv_nsec);
+}
+
+/*
-+ * Allocates bounce buffer and returns its kernel virtual address.
++ * Reads a consistent set of time-base values from Xen, into a shadow data
++ * area.
+ */
-+static void *
-+map_single(struct device *hwdev, struct phys_addr buffer, size_t size, int dir)
++static void get_time_values_from_xen(unsigned int cpu)
+{
++ struct vcpu_time_info *src;
++ struct shadow_time_info *dst;
+ unsigned long flags;
-+ char *dma_addr;
-+ unsigned int nslots, stride, index, wrap;
-+ struct phys_addr slot_buf;
-+ int i;
-+
-+ /*
-+ * For mappings greater than a page, we limit the stride (and
-+ * hence alignment) to a page size.
-+ */
-+ nslots = ALIGN(size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT;
-+ if (size > PAGE_SIZE)
-+ stride = (1 << (PAGE_SHIFT - IO_TLB_SHIFT));
-+ else
-+ stride = 1;
++ u32 pre_version, post_version;
+
-+ BUG_ON(!nslots);
++ src = &vcpu_info(cpu)->time;
++ dst = &per_cpu(shadow_time, cpu);
+
-+ /*
-+ * Find suitable number of IO TLB entries size that will fit this
-+ * request and allocate a buffer from that IO TLB pool.
-+ */
-+ spin_lock_irqsave(&io_tlb_lock, flags);
-+ {
-+ wrap = index = ALIGN(io_tlb_index, stride);
++ local_irq_save(flags);
+
-+ if (index >= iotlb_nslabs)
-+ wrap = index = 0;
++ do {
++ pre_version = dst->version = src->version;
++ rmb();
++ dst->tsc_timestamp = src->tsc_timestamp;
++ dst->system_timestamp = src->system_time;
++ dst->tsc_to_nsec_mul = src->tsc_to_system_mul;
++ dst->tsc_shift = src->tsc_shift;
++ rmb();
++ post_version = src->version;
++ } while ((pre_version & 1) | (pre_version ^ post_version));
+
-+ do {
-+ /*
-+ * If we find a slot that indicates we have 'nslots'
-+ * number of contiguous buffers, we allocate the
-+ * buffers from that slot and mark the entries as '0'
-+ * indicating unavailable.
-+ */
-+ if (io_tlb_list[index] >= nslots) {
-+ int count = 0;
++ dst->tsc_to_usec_mul = dst->tsc_to_nsec_mul / 1000;
+
-+ for (i = index; i < (int)(index + nslots); i++)
-+ io_tlb_list[i] = 0;
-+ for (i = index - 1;
-+ (OFFSET(i, IO_TLB_SEGSIZE) !=
-+ IO_TLB_SEGSIZE -1) && io_tlb_list[i];
-+ i--)
-+ io_tlb_list[i] = ++count;
-+ dma_addr = iotlb_virt_start +
-+ (index << IO_TLB_SHIFT);
++ local_irq_restore(flags);
++}
+
-+ /*
-+ * Update the indices to avoid searching in
-+ * the next round.
-+ */
-+ io_tlb_index =
-+ ((index + nslots) < iotlb_nslabs
-+ ? (index + nslots) : 0);
++static inline int time_values_up_to_date(unsigned int cpu)
++{
++ struct vcpu_time_info *src;
++ struct shadow_time_info *dst;
+
-+ goto found;
-+ }
-+ index += stride;
-+ if (index >= iotlb_nslabs)
-+ index = 0;
-+ } while (index != wrap);
++ src = &vcpu_info(cpu)->time;
++ dst = &per_cpu(shadow_time, cpu);
+
-+ spin_unlock_irqrestore(&io_tlb_lock, flags);
-+ return NULL;
-+ }
-+ found:
-+ spin_unlock_irqrestore(&io_tlb_lock, flags);
++ rmb();
++ return (dst->version == src->version);
++}
+
-+ /*
-+ * Save away the mapping from the original address to the DMA address.
-+ * This is needed when we sync the memory. Then we sync the buffer if
-+ * needed.
-+ */
-+ slot_buf = buffer;
-+ for (i = 0; i < nslots; i++) {
-+ slot_buf.page += slot_buf.offset >> PAGE_SHIFT;
-+ slot_buf.offset &= PAGE_SIZE - 1;
-+ io_tlb_orig_addr[index+i] = slot_buf;
-+ slot_buf.offset += 1 << IO_TLB_SHIFT;
-+ }
-+ if ((dir == DMA_TO_DEVICE) || (dir == DMA_BIDIRECTIONAL))
-+ __sync_single(buffer, dma_addr, size, DMA_TO_DEVICE);
++/*
++ * This is a special lock that is owned by the CPU and holds the index
++ * register we are working with. It is required for NMI access to the
++ * CMOS/RTC registers. See include/asm-i386/mc146818rtc.h for details.
++ */
++volatile unsigned long cmos_lock = 0;
++EXPORT_SYMBOL(cmos_lock);
+
-+ return dma_addr;
++/* Routines for accessing the CMOS RAM/RTC. */
++unsigned char rtc_cmos_read(unsigned char addr)
++{
++ unsigned char val;
++ lock_cmos_prefix(addr);
++ outb_p(addr, RTC_PORT(0));
++ val = inb_p(RTC_PORT(1));
++ lock_cmos_suffix(addr);
++ return val;
+}
++EXPORT_SYMBOL(rtc_cmos_read);
+
-+static struct phys_addr dma_addr_to_phys_addr(char *dma_addr)
++void rtc_cmos_write(unsigned char val, unsigned char addr)
+{
-+ int index = (dma_addr - iotlb_virt_start) >> IO_TLB_SHIFT;
-+ struct phys_addr buffer = io_tlb_orig_addr[index];
-+ buffer.offset += (long)dma_addr & ((1 << IO_TLB_SHIFT) - 1);
-+ buffer.page += buffer.offset >> PAGE_SHIFT;
-+ buffer.offset &= PAGE_SIZE - 1;
-+ return buffer;
++ lock_cmos_prefix(addr);
++ outb_p(addr, RTC_PORT(0));
++ outb_p(val, RTC_PORT(1));
++ lock_cmos_suffix(addr);
+}
++EXPORT_SYMBOL(rtc_cmos_write);
+
+/*
-+ * dma_addr is the kernel virtual address of the bounce buffer to unmap.
++ * This version of gettimeofday has microsecond resolution
++ * and better than microsecond precision on fast x86 machines with TSC.
+ */
-+static void
-+unmap_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
++void do_gettimeofday(struct timeval *tv)
+{
++ unsigned long seq;
++ unsigned long usec, sec;
+ unsigned long flags;
-+ int i, count, nslots = ALIGN(size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT;
-+ int index = (dma_addr - iotlb_virt_start) >> IO_TLB_SHIFT;
-+ struct phys_addr buffer = dma_addr_to_phys_addr(dma_addr);
++ s64 nsec;
++ unsigned int cpu;
++ struct shadow_time_info *shadow;
++ u32 local_time_version;
+
-+ /*
-+ * First, sync the memory before unmapping the entry
-+ */
-+ if ((dir == DMA_FROM_DEVICE) || (dir == DMA_BIDIRECTIONAL))
-+ __sync_single(buffer, dma_addr, size, DMA_FROM_DEVICE);
++ cpu = get_cpu();
++ shadow = &per_cpu(shadow_time, cpu);
+
-+ /*
-+ * Return the buffer to the free list by setting the corresponding
-+ * entries to indicate the number of contigous entries available.
-+ * While returning the entries to the free list, we merge the entries
-+ * with slots below and above the pool being returned.
-+ */
-+ spin_lock_irqsave(&io_tlb_lock, flags);
-+ {
-+ count = ((index + nslots) < ALIGN(index + 1, IO_TLB_SEGSIZE) ?
-+ io_tlb_list[index + nslots] : 0);
-+ /*
-+ * Step 1: return the slots to the free list, merging the
-+ * slots with superceeding slots
-+ */
-+ for (i = index + nslots - 1; i >= index; i--)
-+ io_tlb_list[i] = ++count;
-+ /*
-+ * Step 2: merge the returned slots with the preceding slots,
-+ * if available (non zero)
-+ */
-+ for (i = index - 1;
-+ (OFFSET(i, IO_TLB_SEGSIZE) !=
-+ IO_TLB_SEGSIZE -1) && io_tlb_list[i];
-+ i--)
-+ io_tlb_list[i] = ++count;
-+ }
-+ spin_unlock_irqrestore(&io_tlb_lock, flags);
-+}
++ do {
++ unsigned long lost;
+
-+static void
-+sync_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
-+{
-+ struct phys_addr buffer = dma_addr_to_phys_addr(dma_addr);
-+ BUG_ON((dir != DMA_FROM_DEVICE) && (dir != DMA_TO_DEVICE));
-+ __sync_single(buffer, dma_addr, size, dir);
-+}
++ local_time_version = shadow->version;
++ seq = read_seqbegin(&xtime_lock);
+
-+static void
-+swiotlb_full(struct device *dev, size_t size, int dir, int do_panic)
-+{
-+ /*
-+ * Ran out of IOMMU space for this operation. This is very bad.
-+ * Unfortunately the drivers cannot handle this operation properly.
-+ * unless they check for pci_dma_mapping_error (most don't)
-+ * When the mapping is small enough return a static buffer to limit
-+ * the damage, or panic when the transfer is too big.
-+ */
-+ printk(KERN_ERR "PCI-DMA: Out of SW-IOMMU space for %lu bytes at "
-+ "device %s\n", (unsigned long)size, dev ? dev->bus_id : "?");
++ usec = get_usec_offset(shadow);
++ lost = jiffies - wall_jiffies;
+
-+ if (size > io_tlb_overflow && do_panic) {
-+ if (dir == PCI_DMA_FROMDEVICE || dir == PCI_DMA_BIDIRECTIONAL)
-+ panic("PCI-DMA: Memory would be corrupted\n");
-+ if (dir == PCI_DMA_TODEVICE || dir == PCI_DMA_BIDIRECTIONAL)
-+ panic("PCI-DMA: Random memory would be DMAed\n");
-+ }
-+}
++ if (unlikely(lost))
++ usec += lost * (USEC_PER_SEC / HZ);
+
-+/*
-+ * Map a single buffer of the indicated size for DMA in streaming mode. The
-+ * PCI address to use is returned.
-+ *
-+ * Once the device is given the dma address, the device owns this memory until
-+ * either swiotlb_unmap_single or swiotlb_dma_sync_single is performed.
-+ */
-+dma_addr_t
-+swiotlb_map_single(struct device *hwdev, void *ptr, size_t size, int dir)
-+{
-+ dma_addr_t dev_addr = virt_to_bus(ptr);
-+ void *map;
-+ struct phys_addr buffer;
++ sec = xtime.tv_sec;
++ usec += (xtime.tv_nsec / NSEC_PER_USEC);
+
-+ BUG_ON(dir == DMA_NONE);
++ nsec = shadow->system_timestamp - processed_system_time;
++ __normalize_time(&sec, &nsec);
++ usec += (long)nsec / NSEC_PER_USEC;
+
-+ /*
-+ * If the pointer passed in happens to be in the device's DMA window,
-+ * we can safely return the device addr and not worry about bounce
-+ * buffering it.
-+ */
-+ if (!range_straddles_page_boundary(__pa(ptr), size) &&
-+ !address_needs_mapping(hwdev, dev_addr))
-+ return dev_addr;
++ if (unlikely(!time_values_up_to_date(cpu))) {
++ /*
++ * We may have blocked for a long time,
++ * rendering our calculations invalid
++ * (e.g. the time delta may have
++ * overflowed). Detect that and recalculate
++ * with fresh values.
++ */
++ get_time_values_from_xen(cpu);
++ continue;
++ }
++ } while (read_seqretry(&xtime_lock, seq) ||
++ (local_time_version != shadow->version));
+
-+ /*
-+ * Oh well, have to allocate and map a bounce buffer.
-+ */
-+ buffer.page = virt_to_page(ptr);
-+ buffer.offset = (unsigned long)ptr & ~PAGE_MASK;
-+ map = map_single(hwdev, buffer, size, dir);
-+ if (!map) {
-+ swiotlb_full(hwdev, size, dir, 1);
-+ map = io_tlb_overflow_buffer;
++ put_cpu();
++
++ while (usec >= USEC_PER_SEC) {
++ usec -= USEC_PER_SEC;
++ sec++;
+ }
+
-+ dev_addr = virt_to_bus(map);
-+ return dev_addr;
-+}
++ spin_lock_irqsave(&monotonic_lock, flags);
++ if ((sec > monotonic_tv.tv_sec) ||
++ ((sec == monotonic_tv.tv_sec) && (usec > monotonic_tv.tv_usec)))
++ {
++ monotonic_tv.tv_sec = sec;
++ monotonic_tv.tv_usec = usec;
++ } else {
++ sec = monotonic_tv.tv_sec;
++ usec = monotonic_tv.tv_usec;
++ }
++ spin_unlock_irqrestore(&monotonic_lock, flags);
+
-+/*
-+ * Unmap a single streaming mode DMA translation. The dma_addr and size must
-+ * match what was provided for in a previous swiotlb_map_single call. All
-+ * other usages are undefined.
-+ *
-+ * After this call, reads by the cpu to the buffer are guaranteed to see
-+ * whatever the device wrote there.
-+ */
-+void
-+swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr, size_t size,
-+ int dir)
-+{
-+ BUG_ON(dir == DMA_NONE);
-+ if (in_swiotlb_aperture(dev_addr))
-+ unmap_single(hwdev, bus_to_virt(dev_addr), size, dir);
++ tv->tv_sec = sec;
++ tv->tv_usec = usec;
+}
+
-+/*
-+ * Make physical memory consistent for a single streaming mode DMA translation
-+ * after a transfer.
-+ *
-+ * If you perform a swiotlb_map_single() but wish to interrogate the buffer
-+ * using the cpu, yet do not wish to teardown the PCI dma mapping, you must
-+ * call this function before doing so. At the next point you give the PCI dma
-+ * address back to the card, you must first perform a
-+ * swiotlb_dma_sync_for_device, and then the device again owns the buffer
-+ */
-+void
-+swiotlb_sync_single_for_cpu(struct device *hwdev, dma_addr_t dev_addr,
-+ size_t size, int dir)
-+{
-+ BUG_ON(dir == DMA_NONE);
-+ if (in_swiotlb_aperture(dev_addr))
-+ sync_single(hwdev, bus_to_virt(dev_addr), size, dir);
-+}
++EXPORT_SYMBOL(do_gettimeofday);
+
-+void
-+swiotlb_sync_single_for_device(struct device *hwdev, dma_addr_t dev_addr,
-+ size_t size, int dir)
++int do_settimeofday(struct timespec *tv)
+{
-+ BUG_ON(dir == DMA_NONE);
-+ if (in_swiotlb_aperture(dev_addr))
-+ sync_single(hwdev, bus_to_virt(dev_addr), size, dir);
-+}
++ time_t sec;
++ s64 nsec;
++ unsigned int cpu;
++ struct shadow_time_info *shadow;
++ struct xen_platform_op op;
+
-+/*
-+ * Map a set of buffers described by scatterlist in streaming mode for DMA.
-+ * This is the scatter-gather version of the above swiotlb_map_single
-+ * interface. Here the scatter gather list elements are each tagged with the
-+ * appropriate dma address and length. They are obtained via
-+ * sg_dma_{address,length}(SG).
-+ *
-+ * NOTE: An implementation may be able to use a smaller number of
-+ * DMA address/length pairs than there are SG table elements.
-+ * (for example via virtual mapping capabilities)
-+ * The routine returns the number of addr/length pairs actually
-+ * used, at most nents.
-+ *
-+ * Device ownership issues as mentioned above for swiotlb_map_single are the
-+ * same here.
-+ */
-+int
-+swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg, int nelems,
-+ int dir)
-+{
-+ struct phys_addr buffer;
-+ dma_addr_t dev_addr;
-+ char *map;
-+ int i;
++ if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC)
++ return -EINVAL;
+
-+ BUG_ON(dir == DMA_NONE);
++ cpu = get_cpu();
++ shadow = &per_cpu(shadow_time, cpu);
+
-+ for (i = 0; i < nelems; i++, sg++) {
-+ dev_addr = SG_ENT_PHYS_ADDRESS(sg);
-+ if (range_straddles_page_boundary(page_to_pseudophys(sg->page)
-+ + sg->offset, sg->length)
-+ || address_needs_mapping(hwdev, dev_addr)) {
-+ buffer.page = sg->page;
-+ buffer.offset = sg->offset;
-+ map = map_single(hwdev, buffer, sg->length, dir);
-+ if (!map) {
-+ /* Don't panic here, we expect map_sg users
-+ to do proper error handling. */
-+ swiotlb_full(hwdev, sg->length, dir, 0);
-+ swiotlb_unmap_sg(hwdev, sg - i, i, dir);
-+ sg[0].dma_length = 0;
-+ return 0;
-+ }
-+ sg->dma_address = (dma_addr_t)virt_to_bus(map);
-+ } else
-+ sg->dma_address = dev_addr;
-+ sg->dma_length = sg->length;
-+ }
-+ return nelems;
-+}
++ write_seqlock_irq(&xtime_lock);
+
-+/*
-+ * Unmap a set of streaming mode DMA translations. Again, cpu read rules
-+ * concerning calls here are the same as for swiotlb_unmap_single() above.
-+ */
-+void
-+swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nelems,
-+ int dir)
-+{
-+ int i;
++ /*
++ * Ensure we don't get blocked for a long time so that our time delta
++ * overflows. If that were to happen then our shadow time values would
++ * be stale, so we can retry with fresh ones.
++ */
++ for (;;) {
++ nsec = tv->tv_nsec - get_nsec_offset(shadow);
++ if (time_values_up_to_date(cpu))
++ break;
++ get_time_values_from_xen(cpu);
++ }
++ sec = tv->tv_sec;
++ __normalize_time(&sec, &nsec);
+
-+ BUG_ON(dir == DMA_NONE);
++ if (is_initial_xendomain() && !independent_wallclock) {
++ op.cmd = XENPF_settime;
++ op.u.settime.secs = sec;
++ op.u.settime.nsecs = nsec;
++ op.u.settime.system_time = shadow->system_timestamp;
++ WARN_ON(HYPERVISOR_platform_op(&op));
++ update_wallclock();
++ } else if (independent_wallclock) {
++ nsec -= shadow->system_timestamp;
++ __normalize_time(&sec, &nsec);
++ __update_wallclock(sec, nsec);
++ }
+
-+ for (i = 0; i < nelems; i++, sg++)
-+ if (sg->dma_address != SG_ENT_PHYS_ADDRESS(sg))
-+ unmap_single(hwdev,
-+ (void *)bus_to_virt(sg->dma_address),
-+ sg->dma_length, dir);
-+}
++ /* Reset monotonic gettimeofday() timeval. */
++ spin_lock(&monotonic_lock);
++ monotonic_tv.tv_sec = 0;
++ monotonic_tv.tv_usec = 0;
++ spin_unlock(&monotonic_lock);
+
-+/*
-+ * Make physical memory consistent for a set of streaming mode DMA translations
-+ * after a transfer.
-+ *
-+ * The same as swiotlb_sync_single_* but for a scatter-gather list, same rules
-+ * and usage.
-+ */
-+void
-+swiotlb_sync_sg_for_cpu(struct device *hwdev, struct scatterlist *sg,
-+ int nelems, int dir)
-+{
-+ int i;
++ write_sequnlock_irq(&xtime_lock);
+
-+ BUG_ON(dir == DMA_NONE);
++ put_cpu();
+
-+ for (i = 0; i < nelems; i++, sg++)
-+ if (sg->dma_address != SG_ENT_PHYS_ADDRESS(sg))
-+ sync_single(hwdev,
-+ (void *)bus_to_virt(sg->dma_address),
-+ sg->dma_length, dir);
++ clock_was_set();
++ return 0;
+}
+
-+void
-+swiotlb_sync_sg_for_device(struct device *hwdev, struct scatterlist *sg,
-+ int nelems, int dir)
++EXPORT_SYMBOL(do_settimeofday);
++
++static void sync_xen_wallclock(unsigned long dummy);
++static DEFINE_TIMER(sync_xen_wallclock_timer, sync_xen_wallclock, 0, 0);
++static void sync_xen_wallclock(unsigned long dummy)
+{
-+ int i;
++ time_t sec;
++ s64 nsec;
++ struct xen_platform_op op;
+
-+ BUG_ON(dir == DMA_NONE);
++ if (!ntp_synced() || independent_wallclock || !is_initial_xendomain())
++ return;
+
-+ for (i = 0; i < nelems; i++, sg++)
-+ if (sg->dma_address != SG_ENT_PHYS_ADDRESS(sg))
-+ sync_single(hwdev,
-+ (void *)bus_to_virt(sg->dma_address),
-+ sg->dma_length, dir);
-+}
++ write_seqlock_irq(&xtime_lock);
+
-+#ifdef CONFIG_HIGHMEM
++ sec = xtime.tv_sec;
++ nsec = xtime.tv_nsec + ((jiffies - wall_jiffies) * (u64)NS_PER_TICK);
++ __normalize_time(&sec, &nsec);
+
-+dma_addr_t
-+swiotlb_map_page(struct device *hwdev, struct page *page,
-+ unsigned long offset, size_t size,
-+ enum dma_data_direction direction)
-+{
-+ struct phys_addr buffer;
-+ dma_addr_t dev_addr;
-+ char *map;
++ op.cmd = XENPF_settime;
++ op.u.settime.secs = sec;
++ op.u.settime.nsecs = nsec;
++ op.u.settime.system_time = processed_system_time;
++ WARN_ON(HYPERVISOR_platform_op(&op));
+
-+ dev_addr = page_to_bus(page) + offset;
-+ if (address_needs_mapping(hwdev, dev_addr)) {
-+ buffer.page = page;
-+ buffer.offset = offset;
-+ map = map_single(hwdev, buffer, size, direction);
-+ if (!map) {
-+ swiotlb_full(hwdev, size, direction, 1);
-+ map = io_tlb_overflow_buffer;
-+ }
-+ dev_addr = (dma_addr_t)virt_to_bus(map);
-+ }
++ update_wallclock();
+
-+ return dev_addr;
++ write_sequnlock_irq(&xtime_lock);
++
++ /* Once per minute. */
++ mod_timer(&sync_xen_wallclock_timer, jiffies + 60*HZ);
+}
+
-+void
-+swiotlb_unmap_page(struct device *hwdev, dma_addr_t dma_address,
-+ size_t size, enum dma_data_direction direction)
++static int set_rtc_mmss(unsigned long nowtime)
+{
-+ BUG_ON(direction == DMA_NONE);
-+ if (in_swiotlb_aperture(dma_address))
-+ unmap_single(hwdev, bus_to_virt(dma_address), size, direction);
-+}
++ int retval;
++ unsigned long flags;
+
-+#endif
++ if (independent_wallclock || !is_initial_xendomain())
++ return 0;
+
-+int
-+swiotlb_dma_mapping_error(dma_addr_t dma_addr)
-+{
-+ return (dma_addr == virt_to_bus(io_tlb_overflow_buffer));
++ /* gets recalled with irq locally disabled */
++ /* XXX - does irqsave resolve this? -johnstul */
++ spin_lock_irqsave(&rtc_lock, flags);
++ if (efi_enabled)
++ retval = efi_set_rtc_mmss(nowtime);
++ else
++ retval = mach_set_rtc_mmss(nowtime);
++ spin_unlock_irqrestore(&rtc_lock, flags);
++
++ return retval;
+}
+
-+/*
-+ * Return whether the given PCI device DMA address mask can be supported
-+ * properly. For example, if your device can only drive the low 24-bits
-+ * during PCI bus mastering, then you would pass 0x00ffffff as the mask to
-+ * this function.
++/* monotonic_clock(): returns # of nanoseconds passed since time_init()
++ * Note: This function is required to return accurate
++ * time even in the absence of multiple timer ticks.
+ */
-+int
-+swiotlb_dma_supported (struct device *hwdev, u64 mask)
++unsigned long long monotonic_clock(void)
+{
-+ return (mask >= ((1UL << dma_bits) - 1));
++ unsigned int cpu = get_cpu();
++ struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
++ u64 time;
++ u32 local_time_version;
++
++ do {
++ local_time_version = shadow->version;
++ barrier();
++ time = shadow->system_timestamp + get_nsec_offset(shadow);
++ if (!time_values_up_to_date(cpu))
++ get_time_values_from_xen(cpu);
++ barrier();
++ } while (local_time_version != shadow->version);
++
++ put_cpu();
++
++ return time;
+}
++EXPORT_SYMBOL(monotonic_clock);
+
-+EXPORT_SYMBOL(swiotlb_init);
-+EXPORT_SYMBOL(swiotlb_map_single);
-+EXPORT_SYMBOL(swiotlb_unmap_single);
-+EXPORT_SYMBOL(swiotlb_map_sg);
-+EXPORT_SYMBOL(swiotlb_unmap_sg);
-+EXPORT_SYMBOL(swiotlb_sync_single_for_cpu);
-+EXPORT_SYMBOL(swiotlb_sync_single_for_device);
-+EXPORT_SYMBOL(swiotlb_sync_sg_for_cpu);
-+EXPORT_SYMBOL(swiotlb_sync_sg_for_device);
-+EXPORT_SYMBOL(swiotlb_dma_mapping_error);
-+EXPORT_SYMBOL(swiotlb_dma_supported);
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/sysenter.c tmp-linux-2.6-xen.patch/arch/i386/kernel/sysenter.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/sysenter.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/sysenter.c 2007-07-30 16:35:11.000000000 +0200
-@@ -23,6 +23,10 @@
- #include <asm/pgtable.h>
- #include <asm/unistd.h>
-
-+#ifdef CONFIG_XEN
-+#include <xen/interface/callback.h>
++#ifdef __x86_64__
++unsigned long long sched_clock(void)
++{
++ return monotonic_clock();
++}
+#endif
+
- /*
- * Should the kernel map a VDSO page into processes and pass its
- * address down to glibc upon exec()?
-@@ -44,6 +48,7 @@ extern asmlinkage void sysenter_entry(vo
-
- void enable_sep_cpu(void)
- {
-+#ifndef CONFIG_X86_NO_TSS
- int cpu = get_cpu();
- struct tss_struct *tss = &per_cpu(init_tss, cpu);
-
-@@ -58,6 +63,7 @@ void enable_sep_cpu(void)
- wrmsr(MSR_IA32_SYSENTER_ESP, tss->esp1, 0);
- wrmsr(MSR_IA32_SYSENTER_EIP, (unsigned long) sysenter_entry, 0);
- put_cpu();
-+#endif
- }
-
- /*
-@@ -72,6 +78,18 @@ int __init sysenter_setup(void)
- {
- syscall_page = (void *)get_zeroed_page(GFP_ATOMIC);
-
-+#ifdef CONFIG_XEN
-+ if (boot_cpu_has(X86_FEATURE_SEP)) {
-+ static struct callback_register __initdata sysenter = {
-+ .type = CALLBACKTYPE_sysenter,
-+ .address = { __KERNEL_CS, (unsigned long)sysenter_entry },
-+ };
++#if defined(CONFIG_SMP) && defined(CONFIG_FRAME_POINTER)
++unsigned long profile_pc(struct pt_regs *regs)
++{
++ unsigned long pc = instruction_pointer(regs);
+
-+ if (HYPERVISOR_callback_op(CALLBACKOP_register, &sysenter) < 0)
-+ clear_bit(X86_FEATURE_SEP, boot_cpu_data.x86_capability);
++#ifdef __x86_64__
++ /* Assume the lock function has either no stack frame or only a single word.
++ This checks if the address on the stack looks like a kernel text address.
++ There is a small window for false hits, but in that case the tick
++ is just accounted to the spinlock function.
++ Better would be to write these functions in assembler again
++ and check exactly. */
++ if (!user_mode_vm(regs) && in_lock_functions(pc)) {
++ char *v = *(char **)regs->rsp;
++ if ((v >= _stext && v <= _etext) ||
++ (v >= _sinittext && v <= _einittext) ||
++ (v >= (char *)MODULES_VADDR && v <= (char *)MODULES_END))
++ return (unsigned long)v;
++ return ((unsigned long *)regs->rsp)[1];
+ }
++#else
++ if (!user_mode_vm(regs) && in_lock_functions(pc))
++ return *(unsigned long *)(regs->ebp + 4);
++#endif
++
++ return pc;
++}
++EXPORT_SYMBOL(profile_pc);
+#endif
+
- #ifdef CONFIG_COMPAT_VDSO
- __set_fixmap(FIX_VDSO, __pa(syscall_page), PAGE_READONLY);
- printk("Compat vDSO mapped to %08lx.\n", __fix_to_virt(FIX_VDSO));
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/time-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/time-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/time-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/time-xen.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,1159 @@
+/*
-+ * linux/arch/i386/kernel/time.c
-+ *
-+ * Copyright (C) 1991, 1992, 1995 Linus Torvalds
-+ *
-+ * This file contains the PC-specific time handling details:
-+ * reading the RTC at bootup, etc..
-+ * 1994-07-02 Alan Modra
-+ * fixed set_rtc_mmss, fixed time.year for >= 2000, new mktime
-+ * 1995-03-26 Markus Kuhn
-+ * fixed 500 ms bug at call to set_rtc_mmss, fixed DS12887
-+ * precision CMOS clock update
-+ * 1996-05-03 Ingo Molnar
-+ * fixed time warps in do_[slow|fast]_gettimeoffset()
-+ * 1997-09-10 Updated NTP code according to technical memorandum Jan '96
-+ * "A Kernel Model for Precision Timekeeping" by Dave Mills
-+ * 1998-09-05 (Various)
-+ * More robust do_fast_gettimeoffset() algorithm implemented
-+ * (works with APM, Cyrix 6x86MX and Centaur C6),
-+ * monotonic gettimeofday() with fast_get_timeoffset(),
-+ * drift-proof precision TSC calibration on boot
-+ * (C. Scott Ananian <cananian at alumni.princeton.edu>, Andrew D.
-+ * Balsa <andrebalsa at altern.org>, Philip Gladstone <philip at raptor.com>;
-+ * ported from 2.0.35 Jumbo-9 by Michael Krause <m.krause at tu-harburg.de>).
-+ * 1998-12-16 Andrea Arcangeli
-+ * Fixed Jumbo-9 code in 2.1.131: do_gettimeofday was missing 1 jiffy
-+ * because was not accounting lost_ticks.
-+ * 1998-12-24 Copyright (C) 1998 Andrea Arcangeli
-+ * Fixed a xtime SMP race (we need the xtime_lock rw spinlock to
-+ * serialize accesses to xtime/lost_ticks).
++ * This is the same as the above, except we _also_ save the current
++ * Time Stamp Counter value at the time of the timer interrupt, so that
++ * we later on can estimate the time of day more exactly.
+ */
++irqreturn_t timer_interrupt(int irq, void *dev_id, struct pt_regs *regs)
++{
++ s64 delta, delta_cpu, stolen, blocked;
++ u64 sched_time;
++ unsigned int i, cpu = smp_processor_id();
++ struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
++ struct vcpu_runstate_info *runstate = &per_cpu(runstate, cpu);
+
-+#include <linux/errno.h>
-+#include <linux/sched.h>
-+#include <linux/kernel.h>
-+#include <linux/param.h>
-+#include <linux/string.h>
-+#include <linux/mm.h>
-+#include <linux/interrupt.h>
-+#include <linux/time.h>
-+#include <linux/delay.h>
-+#include <linux/init.h>
-+#include <linux/smp.h>
-+#include <linux/module.h>
-+#include <linux/sysdev.h>
-+#include <linux/bcd.h>
-+#include <linux/efi.h>
-+#include <linux/mca.h>
-+#include <linux/sysctl.h>
-+#include <linux/percpu.h>
-+#include <linux/kernel_stat.h>
-+#include <linux/posix-timers.h>
-+
-+#include <asm/io.h>
-+#include <asm/smp.h>
-+#include <asm/irq.h>
-+#include <asm/msr.h>
-+#include <asm/delay.h>
-+#include <asm/mpspec.h>
-+#include <asm/uaccess.h>
-+#include <asm/processor.h>
-+#include <asm/timer.h>
-+#include <asm/sections.h>
++ /*
++ * Here we are in the timer irq handler. We just have irqs locally
++ * disabled but we don't know if the timer_bh is running on the other
++ * CPU. We need to avoid to SMP race with it. NOTE: we don' t need
++ * the irq version of write_lock because as just said we have irq
++ * locally disabled. -arca
++ */
++ write_seqlock(&xtime_lock);
+
-+#include "mach_time.h"
++ do {
++ get_time_values_from_xen(cpu);
+
-+#include <linux/timex.h>
-+
-+#include <asm/hpet.h>
++ /* Obtain a consistent snapshot of elapsed wallclock cycles. */
++ delta = delta_cpu =
++ shadow->system_timestamp + get_nsec_offset(shadow);
++ delta -= processed_system_time;
++ delta_cpu -= per_cpu(processed_system_time, cpu);
+
-+#include <asm/arch_hooks.h>
++ /*
++ * Obtain a consistent snapshot of stolen/blocked cycles. We
++ * can use state_entry_time to detect if we get preempted here.
++ */
++ do {
++ sched_time = runstate->state_entry_time;
++ barrier();
++ stolen = runstate->time[RUNSTATE_runnable] +
++ runstate->time[RUNSTATE_offline] -
++ per_cpu(processed_stolen_time, cpu);
++ blocked = runstate->time[RUNSTATE_blocked] -
++ per_cpu(processed_blocked_time, cpu);
++ barrier();
++ } while (sched_time != runstate->state_entry_time);
++ } while (!time_values_up_to_date(cpu));
+
-+#include <xen/evtchn.h>
-+#include <xen/interface/vcpu.h>
++ if ((unlikely(delta < -(s64)permitted_clock_jitter) ||
++ unlikely(delta_cpu < -(s64)permitted_clock_jitter))
++ && printk_ratelimit()) {
++ printk("Timer ISR/%u: Time went backwards: "
++ "delta=%lld delta_cpu=%lld shadow=%lld "
++ "off=%lld processed=%lld cpu_processed=%lld\n",
++ cpu, delta, delta_cpu, shadow->system_timestamp,
++ (s64)get_nsec_offset(shadow),
++ processed_system_time,
++ per_cpu(processed_system_time, cpu));
++ for (i = 0; i < num_online_cpus(); i++)
++ printk(" %d: %lld\n", i,
++ per_cpu(processed_system_time, i));
++ }
+
-+#if defined (__i386__)
-+#include <asm/i8259.h>
-+#endif
++ /* System-wide jiffy work. */
++ while (delta >= NS_PER_TICK) {
++ delta -= NS_PER_TICK;
++ processed_system_time += NS_PER_TICK;
++ do_timer(regs);
++ }
+
-+int pit_latch_buggy; /* extern */
++ if (shadow_tv_version != HYPERVISOR_shared_info->wc_version) {
++ update_wallclock();
++ if (keventd_up())
++ schedule_work(&clock_was_set_work);
++ }
+
-+#if defined(__x86_64__)
-+unsigned long vxtime_hz = PIT_TICK_RATE;
-+struct vxtime_data __vxtime __section_vxtime; /* for vsyscalls */
-+volatile unsigned long __jiffies __section_jiffies = INITIAL_JIFFIES;
-+unsigned long __wall_jiffies __section_wall_jiffies = INITIAL_JIFFIES;
-+struct timespec __xtime __section_xtime;
-+struct timezone __sys_tz __section_sys_tz;
-+#endif
++ write_sequnlock(&xtime_lock);
+
-+unsigned int cpu_khz; /* Detected as we calibrate the TSC */
-+EXPORT_SYMBOL(cpu_khz);
++ /*
++ * Account stolen ticks.
++ * HACK: Passing NULL to account_steal_time()
++ * ensures that the ticks are accounted as stolen.
++ */
++ if ((stolen > 0) && (delta_cpu > 0)) {
++ delta_cpu -= stolen;
++ if (unlikely(delta_cpu < 0))
++ stolen += delta_cpu; /* clamp local-time progress */
++ do_div(stolen, NS_PER_TICK);
++ per_cpu(processed_stolen_time, cpu) += stolen * NS_PER_TICK;
++ per_cpu(processed_system_time, cpu) += stolen * NS_PER_TICK;
++ account_steal_time(NULL, (cputime_t)stolen);
++ }
+
-+extern unsigned long wall_jiffies;
++ /*
++ * Account blocked ticks.
++ * HACK: Passing idle_task to account_steal_time()
++ * ensures that the ticks are accounted as idle/wait.
++ */
++ if ((blocked > 0) && (delta_cpu > 0)) {
++ delta_cpu -= blocked;
++ if (unlikely(delta_cpu < 0))
++ blocked += delta_cpu; /* clamp local-time progress */
++ do_div(blocked, NS_PER_TICK);
++ per_cpu(processed_blocked_time, cpu) += blocked * NS_PER_TICK;
++ per_cpu(processed_system_time, cpu) += blocked * NS_PER_TICK;
++ account_steal_time(idle_task(cpu), (cputime_t)blocked);
++ }
+
-+DEFINE_SPINLOCK(rtc_lock);
-+EXPORT_SYMBOL(rtc_lock);
++ /* Account user/system ticks. */
++ if (delta_cpu > 0) {
++ do_div(delta_cpu, NS_PER_TICK);
++ per_cpu(processed_system_time, cpu) += delta_cpu * NS_PER_TICK;
++ if (user_mode_vm(regs))
++ account_user_time(current, (cputime_t)delta_cpu);
++ else
++ account_system_time(current, HARDIRQ_OFFSET,
++ (cputime_t)delta_cpu);
++ }
+
-+extern struct init_timer_opts timer_tsc_init;
-+extern struct timer_opts timer_tsc;
-+#define timer_none timer_tsc
++ /* Offlined for more than a few seconds? Avoid lockup warnings. */
++ if (stolen > 5*HZ)
++ touch_softlockup_watchdog();
+
-+/* These are peridically updated in shared_info, and then copied here. */
-+struct shadow_time_info {
-+ u64 tsc_timestamp; /* TSC at last update of time vals. */
-+ u64 system_timestamp; /* Time, in nanosecs, since boot. */
-+ u32 tsc_to_nsec_mul;
-+ u32 tsc_to_usec_mul;
-+ int tsc_shift;
-+ u32 version;
-+};
-+static DEFINE_PER_CPU(struct shadow_time_info, shadow_time);
-+static struct timespec shadow_tv;
-+static u32 shadow_tv_version;
++ /* Local timer processing (see update_process_times()). */
++ run_local_timers();
++ if (rcu_pending(cpu))
++ rcu_check_callbacks(cpu, user_mode_vm(regs));
++ scheduler_tick();
++ run_posix_cpu_timers(current);
++ profile_tick(CPU_PROFILING, regs);
+
-+static struct timeval monotonic_tv;
-+static spinlock_t monotonic_lock = SPIN_LOCK_UNLOCKED;
++ return IRQ_HANDLED;
++}
+
-+/* Keep track of last time we did processing/updating of jiffies and xtime. */
-+static u64 processed_system_time; /* System time (ns) at last processing. */
-+static DEFINE_PER_CPU(u64, processed_system_time);
++static void init_missing_ticks_accounting(unsigned int cpu)
++{
++ struct vcpu_register_runstate_memory_area area;
++ struct vcpu_runstate_info *runstate = &per_cpu(runstate, cpu);
++ int rc;
+
-+/* How much CPU time was spent blocked and how much was 'stolen'? */
-+static DEFINE_PER_CPU(u64, processed_stolen_time);
-+static DEFINE_PER_CPU(u64, processed_blocked_time);
++ memset(runstate, 0, sizeof(*runstate));
+
-+/* Current runstate of each CPU (updated automatically by the hypervisor). */
-+static DEFINE_PER_CPU(struct vcpu_runstate_info, runstate);
++ area.addr.v = runstate;
++ rc = HYPERVISOR_vcpu_op(VCPUOP_register_runstate_memory_area, cpu, &area);
++ WARN_ON(rc && rc != -ENOSYS);
+
-+/* Must be signed, as it's compared with s64 quantities which can be -ve. */
-+#define NS_PER_TICK (1000000000LL/HZ)
++ per_cpu(processed_blocked_time, cpu) =
++ runstate->time[RUNSTATE_blocked];
++ per_cpu(processed_stolen_time, cpu) =
++ runstate->time[RUNSTATE_runnable] +
++ runstate->time[RUNSTATE_offline];
++}
+
-+static void __clock_was_set(void *unused)
++/* not static: needed by APM */
++unsigned long get_cmos_time(void)
+{
-+ clock_was_set();
++ unsigned long retval;
++ unsigned long flags;
++
++ spin_lock_irqsave(&rtc_lock, flags);
++
++ if (efi_enabled)
++ retval = efi_get_time();
++ else
++ retval = mach_get_cmos_time();
++
++ spin_unlock_irqrestore(&rtc_lock, flags);
++
++ return retval;
+}
-+static DECLARE_WORK(clock_was_set_work, __clock_was_set, NULL);
++EXPORT_SYMBOL(get_cmos_time);
+
-+static inline void __normalize_time(time_t *sec, s64 *nsec)
++static void sync_cmos_clock(unsigned long dummy);
++
++static DEFINE_TIMER(sync_cmos_timer, sync_cmos_clock, 0, 0);
++
++static void sync_cmos_clock(unsigned long dummy)
+{
-+ while (*nsec >= NSEC_PER_SEC) {
-+ (*nsec) -= NSEC_PER_SEC;
-+ (*sec)++;
-+ }
-+ while (*nsec < 0) {
-+ (*nsec) += NSEC_PER_SEC;
-+ (*sec)--;
++ struct timeval now, next;
++ int fail = 1;
++
++ /*
++ * If we have an externally synchronized Linux clock, then update
++ * CMOS clock accordingly every ~11 minutes. Set_rtc_mmss() has to be
++ * called as close as possible to 500 ms before the new second starts.
++ * This code is run on a timer. If the clock is set, that timer
++ * may not expire at the correct time. Thus, we adjust...
++ */
++ if (!ntp_synced())
++ /*
++ * Not synced, exit, do not restart a timer (if one is
++ * running, let it run out).
++ */
++ return;
++
++ do_gettimeofday(&now);
++ if (now.tv_usec >= USEC_AFTER - ((unsigned) TICK_SIZE) / 2 &&
++ now.tv_usec <= USEC_BEFORE + ((unsigned) TICK_SIZE) / 2)
++ fail = set_rtc_mmss(now.tv_sec);
++
++ next.tv_usec = USEC_AFTER - now.tv_usec;
++ if (next.tv_usec <= 0)
++ next.tv_usec += USEC_PER_SEC;
++
++ if (!fail)
++ next.tv_sec = 659;
++ else
++ next.tv_sec = 0;
++
++ if (next.tv_usec >= USEC_PER_SEC) {
++ next.tv_sec++;
++ next.tv_usec -= USEC_PER_SEC;
+ }
++ mod_timer(&sync_cmos_timer, jiffies + timeval_to_jiffies(&next));
+}
+
-+/* Does this guest OS track Xen time, or set its wall clock independently? */
-+static int independent_wallclock = 0;
-+static int __init __independent_wallclock(char *str)
++void notify_arch_cmos_timer(void)
+{
-+ independent_wallclock = 1;
-+ return 1;
++ mod_timer(&sync_cmos_timer, jiffies + 1);
++ mod_timer(&sync_xen_wallclock_timer, jiffies + 1);
+}
-+__setup("independent_wallclock", __independent_wallclock);
+
-+/* Permitted clock jitter, in nsecs, beyond which a warning will be printed. */
-+static unsigned long permitted_clock_jitter = 10000000UL; /* 10ms */
-+static int __init __permitted_clock_jitter(char *str)
++static int timer_resume(struct sys_device *dev)
+{
-+ permitted_clock_jitter = simple_strtoul(str, NULL, 0);
-+ return 1;
++ extern void time_resume(void);
++ time_resume();
++ return 0;
+}
-+__setup("permitted_clock_jitter=", __permitted_clock_jitter);
+
-+#if 0
-+static void delay_tsc(unsigned long loops)
-+{
-+ unsigned long bclock, now;
++static struct sysdev_class timer_sysclass = {
++ .resume = timer_resume,
++ set_kset_name("timer"),
++};
+
-+ rdtscl(bclock);
-+ do {
-+ rep_nop();
-+ rdtscl(now);
-+ } while ((now - bclock) < loops);
-+}
+
-+struct timer_opts timer_tsc = {
-+ .name = "tsc",
-+ .delay = delay_tsc,
++/* XXX this driverfs stuff should probably go elsewhere later -john */
++static struct sys_device device_timer = {
++ .id = 0,
++ .cls = &timer_sysclass,
+};
-+#endif
+
-+/*
-+ * Scale a 64-bit delta by scaling and multiplying by a 32-bit fraction,
-+ * yielding a 64-bit result.
-+ */
-+static inline u64 scale_delta(u64 delta, u32 mul_frac, int shift)
++static int time_init_device(void)
+{
-+ u64 product;
-+#ifdef __i386__
-+ u32 tmp1, tmp2;
-+#endif
++ int error = sysdev_class_register(&timer_sysclass);
++ if (!error)
++ error = sysdev_register(&device_timer);
++ return error;
++}
+
-+ if (shift < 0)
-+ delta >>= -shift;
-+ else
-+ delta <<= shift;
++device_initcall(time_init_device);
+
-+#ifdef __i386__
-+ __asm__ (
-+ "mul %5 ; "
-+ "mov %4,%%eax ; "
-+ "mov %%edx,%4 ; "
-+ "mul %5 ; "
-+ "xor %5,%5 ; "
-+ "add %4,%%eax ; "
-+ "adc %5,%%edx ; "
-+ : "=A" (product), "=r" (tmp1), "=r" (tmp2)
-+ : "a" ((u32)delta), "1" ((u32)(delta >> 32)), "2" (mul_frac) );
-+#else
-+ __asm__ (
-+ "mul %%rdx ; shrd $32,%%rdx,%%rax"
-+ : "=a" (product) : "0" (delta), "d" ((u64)mul_frac) );
-+#endif
++#ifdef CONFIG_HPET_TIMER
++extern void (*late_time_init)(void);
++/* Duplicate of time_init() below, with hpet_enable part added */
++static void __init hpet_time_init(void)
++{
++ xtime.tv_sec = get_cmos_time();
++ xtime.tv_nsec = (INITIAL_JIFFIES % HZ) * (NSEC_PER_SEC / HZ);
++ set_normalized_timespec(&wall_to_monotonic,
++ -xtime.tv_sec, -xtime.tv_nsec);
+
-+ return product;
-+}
++ if ((hpet_enable() >= 0) && hpet_use_timer) {
++ printk("Using HPET for base-timer\n");
++ }
+
-+#if 0 /* defined (__i386__) */
-+int read_current_timer(unsigned long *timer_val)
-+{
-+ rdtscl(*timer_val);
-+ return 0;
++ time_init_hook();
+}
+#endif
+
-+void init_cpu_khz(void)
-+{
-+ u64 __cpu_khz = 1000000ULL << 32;
-+ struct vcpu_time_info *info = &vcpu_info(0)->time;
-+ do_div(__cpu_khz, info->tsc_to_system_mul);
-+ if (info->tsc_shift < 0)
-+ cpu_khz = __cpu_khz << -info->tsc_shift;
-+ else
-+ cpu_khz = __cpu_khz >> info->tsc_shift;
-+}
++/* Dynamically-mapped IRQ. */
++DEFINE_PER_CPU(int, timer_irq);
+
-+static u64 get_nsec_offset(struct shadow_time_info *shadow)
++extern void (*late_time_init)(void);
++static void setup_cpu0_timer_irq(void)
+{
-+ u64 now, delta;
-+ rdtscll(now);
-+ delta = now - shadow->tsc_timestamp;
-+ return scale_delta(delta, shadow->tsc_to_nsec_mul, shadow->tsc_shift);
++ per_cpu(timer_irq, 0) =
++ bind_virq_to_irqhandler(
++ VIRQ_TIMER,
++ 0,
++ timer_interrupt,
++ SA_INTERRUPT,
++ "timer0",
++ NULL);
++ BUG_ON(per_cpu(timer_irq, 0) < 0);
+}
+
-+static unsigned long get_usec_offset(struct shadow_time_info *shadow)
-+{
-+ u64 now, delta;
-+ rdtscll(now);
-+ delta = now - shadow->tsc_timestamp;
-+ return scale_delta(delta, shadow->tsc_to_usec_mul, shadow->tsc_shift);
-+}
++static struct vcpu_set_periodic_timer xen_set_periodic_tick = {
++ .period_ns = NS_PER_TICK
++};
+
-+static void __update_wallclock(time_t sec, long nsec)
++void __init time_init(void)
+{
-+ long wtm_nsec, xtime_nsec;
-+ time_t wtm_sec, xtime_sec;
-+ u64 tmp, wc_nsec;
++#ifdef CONFIG_HPET_TIMER
++ if (is_hpet_capable()) {
++ /*
++ * HPET initialization needs to do memory-mapped io. So, let
++ * us do a late initialization after mem_init().
++ */
++ late_time_init = hpet_time_init;
++ return;
++ }
++#endif
+
-+ /* Adjust wall-clock time base based on wall_jiffies ticks. */
-+ wc_nsec = processed_system_time;
-+ wc_nsec += sec * (u64)NSEC_PER_SEC;
-+ wc_nsec += nsec;
-+ wc_nsec -= (jiffies - wall_jiffies) * (u64)NS_PER_TICK;
++ switch (HYPERVISOR_vcpu_op(VCPUOP_set_periodic_timer, 0,
++ &xen_set_periodic_tick)) {
++ case 0:
++#if CONFIG_XEN_COMPAT <= 0x030004
++ case -ENOSYS:
++#endif
++ break;
++ default:
++ BUG();
++ }
+
-+ /* Split wallclock base into seconds and nanoseconds. */
-+ tmp = wc_nsec;
-+ xtime_nsec = do_div(tmp, 1000000000);
-+ xtime_sec = (time_t)tmp;
++ get_time_values_from_xen(0);
+
-+ wtm_sec = wall_to_monotonic.tv_sec + (xtime.tv_sec - xtime_sec);
-+ wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - xtime_nsec);
++ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
++ per_cpu(processed_system_time, 0) = processed_system_time;
++ init_missing_ticks_accounting(0);
+
-+ set_normalized_timespec(&xtime, xtime_sec, xtime_nsec);
-+ set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec);
++ update_wallclock();
+
-+ ntp_clear();
++ init_cpu_khz();
++ printk(KERN_INFO "Xen reported: %u.%03u MHz processor.\n",
++ cpu_khz / 1000, cpu_khz % 1000);
++
++#if defined(__x86_64__)
++ vxtime.mode = VXTIME_TSC;
++ vxtime.quot = (1000000L << 32) / vxtime_hz;
++ vxtime.tsc_quot = (1000L << 32) / cpu_khz;
++ sync_core();
++ rdtscll(vxtime.last_tsc);
++#endif
++
++ /* Cannot request_irq() until kmem is initialised. */
++ late_time_init = setup_cpu0_timer_irq;
+}
+
-+static void update_wallclock(void)
++/* Convert jiffies to system time. */
++u64 jiffies_to_st(unsigned long j)
+{
-+ shared_info_t *s = HYPERVISOR_shared_info;
++ unsigned long seq;
++ long delta;
++ u64 st;
+
+ do {
-+ shadow_tv_version = s->wc_version;
-+ rmb();
-+ shadow_tv.tv_sec = s->wc_sec;
-+ shadow_tv.tv_nsec = s->wc_nsec;
-+ rmb();
-+ } while ((s->wc_version & 1) | (shadow_tv_version ^ s->wc_version));
++ seq = read_seqbegin(&xtime_lock);
++ delta = j - jiffies;
++ if (delta < 1) {
++ /* Triggers in some wrap-around cases, but that's okay:
++ * we just end up with a shorter timeout. */
++ st = processed_system_time + NS_PER_TICK;
++ } else if (((unsigned long)delta >> (BITS_PER_LONG-3)) != 0) {
++ /* Very long timeout means there is no pending timer.
++ * We indicate this to Xen by passing zero timeout. */
++ st = 0;
++ } else {
++ st = processed_system_time + delta * (u64)NS_PER_TICK;
++ }
++ } while (read_seqretry(&xtime_lock, seq));
+
-+ if (!independent_wallclock)
-+ __update_wallclock(shadow_tv.tv_sec, shadow_tv.tv_nsec);
++ return st;
+}
++EXPORT_SYMBOL(jiffies_to_st);
+
+/*
-+ * Reads a consistent set of time-base values from Xen, into a shadow data
-+ * area.
++ * stop_hz_timer / start_hz_timer - enter/exit 'tickless mode' on an idle cpu
++ * These functions are based on implementations from arch/s390/kernel/time.c
+ */
-+static void get_time_values_from_xen(int cpu)
++static void stop_hz_timer(void)
+{
-+ struct vcpu_time_info *src;
-+ struct shadow_time_info *dst;
++ struct vcpu_set_singleshot_timer singleshot;
++ unsigned int cpu = smp_processor_id();
++ unsigned long j;
++ int rc;
+
-+ src = &vcpu_info(cpu)->time;
-+ dst = &per_cpu(shadow_time, cpu);
++ cpu_set(cpu, nohz_cpu_mask);
+
-+ do {
-+ dst->version = src->version;
-+ rmb();
-+ dst->tsc_timestamp = src->tsc_timestamp;
-+ dst->system_timestamp = src->system_time;
-+ dst->tsc_to_nsec_mul = src->tsc_to_system_mul;
-+ dst->tsc_shift = src->tsc_shift;
-+ rmb();
-+ } while ((src->version & 1) | (dst->version ^ src->version));
++ /* See matching smp_mb in rcu_start_batch in rcupdate.c. These mbs */
++ /* ensure that if __rcu_pending (nested in rcu_needs_cpu) fetches a */
++ /* value of rcp->cur that matches rdp->quiescbatch and allows us to */
++ /* stop the hz timer then the cpumasks created for subsequent values */
++ /* of cur in rcu_start_batch are guaranteed to pick up the updated */
++ /* nohz_cpu_mask and so will not depend on this cpu. */
+
-+ dst->tsc_to_usec_mul = dst->tsc_to_nsec_mul / 1000;
++ smp_mb();
++
++ /* Leave ourselves in tick mode if rcu or softirq or timer pending. */
++ if (rcu_needs_cpu(cpu) || local_softirq_pending() ||
++ (j = next_timer_interrupt(), time_before_eq(j, jiffies))) {
++ cpu_clear(cpu, nohz_cpu_mask);
++ j = jiffies + 1;
++ }
++
++ singleshot.timeout_abs_ns = jiffies_to_st(j) + NS_PER_TICK/2;
++ singleshot.flags = 0;
++ rc = HYPERVISOR_vcpu_op(VCPUOP_set_singleshot_timer, cpu, &singleshot);
++#if CONFIG_XEN_COMPAT <= 0x030004
++ if (rc) {
++ BUG_ON(rc != -ENOSYS);
++ rc = HYPERVISOR_set_timer_op(singleshot.timeout_abs_ns);
++ }
++#endif
++ BUG_ON(rc);
+}
+
-+static inline int time_values_up_to_date(int cpu)
++static void start_hz_timer(void)
+{
-+ struct vcpu_time_info *src;
-+ struct shadow_time_info *dst;
-+
-+ src = &vcpu_info(cpu)->time;
-+ dst = &per_cpu(shadow_time, cpu);
++ cpu_clear(smp_processor_id(), nohz_cpu_mask);
++}
+
-+ rmb();
-+ return (dst->version == src->version);
++void raw_safe_halt(void)
++{
++ stop_hz_timer();
++ /* Blocking includes an implicit local_irq_enable(). */
++ HYPERVISOR_block();
++ start_hz_timer();
+}
++EXPORT_SYMBOL(raw_safe_halt);
+
-+/*
-+ * This is a special lock that is owned by the CPU and holds the index
-+ * register we are working with. It is required for NMI access to the
-+ * CMOS/RTC registers. See include/asm-i386/mc146818rtc.h for details.
-+ */
-+volatile unsigned long cmos_lock = 0;
-+EXPORT_SYMBOL(cmos_lock);
-+
-+/* Routines for accessing the CMOS RAM/RTC. */
-+unsigned char rtc_cmos_read(unsigned char addr)
-+{
-+ unsigned char val;
-+ lock_cmos_prefix(addr);
-+ outb_p(addr, RTC_PORT(0));
-+ val = inb_p(RTC_PORT(1));
-+ lock_cmos_suffix(addr);
-+ return val;
-+}
-+EXPORT_SYMBOL(rtc_cmos_read);
-+
-+void rtc_cmos_write(unsigned char val, unsigned char addr)
++void halt(void)
+{
-+ lock_cmos_prefix(addr);
-+ outb_p(addr, RTC_PORT(0));
-+ outb_p(val, RTC_PORT(1));
-+ lock_cmos_suffix(addr);
++ if (irqs_disabled())
++ VOID(HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL));
+}
-+EXPORT_SYMBOL(rtc_cmos_write);
++EXPORT_SYMBOL(halt);
+
-+/*
-+ * This version of gettimeofday has microsecond resolution
-+ * and better than microsecond precision on fast x86 machines with TSC.
-+ */
-+void do_gettimeofday(struct timeval *tv)
++/* No locking required. Interrupts are disabled on all CPUs. */
++void time_resume(void)
+{
-+ unsigned long seq;
-+ unsigned long usec, sec;
-+ unsigned long flags;
-+ s64 nsec;
+ unsigned int cpu;
-+ struct shadow_time_info *shadow;
-+ u32 local_time_version;
-+
-+ cpu = get_cpu();
-+ shadow = &per_cpu(shadow_time, cpu);
-+
-+ do {
-+ unsigned long lost;
+
-+ local_time_version = shadow->version;
-+ seq = read_seqbegin(&xtime_lock);
++ init_cpu_khz();
+
-+ usec = get_usec_offset(shadow);
-+ lost = jiffies - wall_jiffies;
++ for_each_online_cpu(cpu) {
++ switch (HYPERVISOR_vcpu_op(VCPUOP_set_periodic_timer, cpu,
++ &xen_set_periodic_tick)) {
++ case 0:
++#if CONFIG_XEN_COMPAT <= 0x030004
++ case -ENOSYS:
++#endif
++ break;
++ default:
++ BUG();
++ }
++ get_time_values_from_xen(cpu);
++ per_cpu(processed_system_time, cpu) =
++ per_cpu(shadow_time, 0).system_timestamp;
++ init_missing_ticks_accounting(cpu);
++ }
+
-+ if (unlikely(lost))
-+ usec += lost * (USEC_PER_SEC / HZ);
++ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
+
-+ sec = xtime.tv_sec;
-+ usec += (xtime.tv_nsec / NSEC_PER_USEC);
++ update_wallclock();
++}
+
-+ nsec = shadow->system_timestamp - processed_system_time;
-+ __normalize_time(&sec, &nsec);
-+ usec += (long)nsec / NSEC_PER_USEC;
++#ifdef CONFIG_SMP
++static char timer_name[NR_CPUS][15];
+
-+ if (unlikely(!time_values_up_to_date(cpu))) {
-+ /*
-+ * We may have blocked for a long time,
-+ * rendering our calculations invalid
-+ * (e.g. the time delta may have
-+ * overflowed). Detect that and recalculate
-+ * with fresh values.
-+ */
-+ get_time_values_from_xen(cpu);
-+ continue;
-+ }
-+ } while (read_seqretry(&xtime_lock, seq) ||
-+ (local_time_version != shadow->version));
++int __cpuinit local_setup_timer(unsigned int cpu)
++{
++ int seq, irq;
+
-+ put_cpu();
++ BUG_ON(cpu == 0);
+
-+ while (usec >= USEC_PER_SEC) {
-+ usec -= USEC_PER_SEC;
-+ sec++;
++ switch (HYPERVISOR_vcpu_op(VCPUOP_set_periodic_timer, cpu,
++ &xen_set_periodic_tick)) {
++ case 0:
++#if CONFIG_XEN_COMPAT <= 0x030004
++ case -ENOSYS:
++#endif
++ break;
++ default:
++ BUG();
+ }
+
-+ spin_lock_irqsave(&monotonic_lock, flags);
-+ if ((sec > monotonic_tv.tv_sec) ||
-+ ((sec == monotonic_tv.tv_sec) && (usec > monotonic_tv.tv_usec)))
-+ {
-+ monotonic_tv.tv_sec = sec;
-+ monotonic_tv.tv_usec = usec;
-+ } else {
-+ sec = monotonic_tv.tv_sec;
-+ usec = monotonic_tv.tv_usec;
-+ }
-+ spin_unlock_irqrestore(&monotonic_lock, flags);
++ do {
++ seq = read_seqbegin(&xtime_lock);
++ /* Use cpu0 timestamp: cpu's shadow is not initialised yet. */
++ per_cpu(processed_system_time, cpu) =
++ per_cpu(shadow_time, 0).system_timestamp;
++ init_missing_ticks_accounting(cpu);
++ } while (read_seqretry(&xtime_lock, seq));
+
-+ tv->tv_sec = sec;
-+ tv->tv_usec = usec;
++ sprintf(timer_name[cpu], "timer%u", cpu);
++ irq = bind_virq_to_irqhandler(VIRQ_TIMER,
++ cpu,
++ timer_interrupt,
++ SA_INTERRUPT,
++ timer_name[cpu],
++ NULL);
++ if (irq < 0)
++ return irq;
++ per_cpu(timer_irq, cpu) = irq;
++
++ return 0;
+}
+
-+EXPORT_SYMBOL(do_gettimeofday);
++void __cpuexit local_teardown_timer(unsigned int cpu)
++{
++ BUG_ON(cpu == 0);
++ unbind_from_irqhandler(per_cpu(timer_irq, cpu), NULL);
++}
++#endif
+
-+int do_settimeofday(struct timespec *tv)
++#ifdef CONFIG_CPU_FREQ
++static int time_cpufreq_notifier(struct notifier_block *nb, unsigned long val,
++ void *data)
+{
-+ time_t sec;
-+ s64 nsec;
-+ unsigned int cpu;
-+ struct shadow_time_info *shadow;
++ struct cpufreq_freqs *freq = data;
+ struct xen_platform_op op;
+
-+ if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC)
-+ return -EINVAL;
-+
-+ cpu = get_cpu();
-+ shadow = &per_cpu(shadow_time, cpu);
++ if (cpu_has(&cpu_data[freq->cpu], X86_FEATURE_CONSTANT_TSC))
++ return 0;
+
-+ write_seqlock_irq(&xtime_lock);
++ if (val == CPUFREQ_PRECHANGE)
++ return 0;
+
-+ /*
-+ * Ensure we don't get blocked for a long time so that our time delta
-+ * overflows. If that were to happen then our shadow time values would
-+ * be stale, so we can retry with fresh ones.
-+ */
-+ for (;;) {
-+ nsec = tv->tv_nsec - get_nsec_offset(shadow);
-+ if (time_values_up_to_date(cpu))
-+ break;
-+ get_time_values_from_xen(cpu);
-+ }
-+ sec = tv->tv_sec;
-+ __normalize_time(&sec, &nsec);
++ op.cmd = XENPF_change_freq;
++ op.u.change_freq.flags = 0;
++ op.u.change_freq.cpu = freq->cpu;
++ op.u.change_freq.freq = (u64)freq->new * 1000;
++ WARN_ON(HYPERVISOR_platform_op(&op));
+
-+ if (is_initial_xendomain() && !independent_wallclock) {
-+ op.cmd = XENPF_settime;
-+ op.u.settime.secs = sec;
-+ op.u.settime.nsecs = nsec;
-+ op.u.settime.system_time = shadow->system_timestamp;
-+ HYPERVISOR_platform_op(&op);
-+ update_wallclock();
-+ } else if (independent_wallclock) {
-+ nsec -= shadow->system_timestamp;
-+ __normalize_time(&sec, &nsec);
-+ __update_wallclock(sec, nsec);
-+ }
++ return 0;
++}
+
-+ /* Reset monotonic gettimeofday() timeval. */
-+ spin_lock(&monotonic_lock);
-+ monotonic_tv.tv_sec = 0;
-+ monotonic_tv.tv_usec = 0;
-+ spin_unlock(&monotonic_lock);
++static struct notifier_block time_cpufreq_notifier_block = {
++ .notifier_call = time_cpufreq_notifier
++};
+
-+ write_sequnlock_irq(&xtime_lock);
++static int __init cpufreq_time_setup(void)
++{
++ if (!cpufreq_register_notifier(&time_cpufreq_notifier_block,
++ CPUFREQ_TRANSITION_NOTIFIER)) {
++ printk(KERN_ERR "failed to set up cpufreq notifier\n");
++ return -ENODEV;
++ }
++ return 0;
++}
+
-+ put_cpu();
++core_initcall(cpufreq_time_setup);
++#endif
+
-+ clock_was_set();
++/*
++ * /proc/sys/xen: This really belongs in another file. It can stay here for
++ * now however.
++ */
++static ctl_table xen_subtable[] = {
++ {
++ .ctl_name = 1,
++ .procname = "independent_wallclock",
++ .data = &independent_wallclock,
++ .maxlen = sizeof(independent_wallclock),
++ .mode = 0644,
++ .proc_handler = proc_dointvec
++ },
++ {
++ .ctl_name = 2,
++ .procname = "permitted_clock_jitter",
++ .data = &permitted_clock_jitter,
++ .maxlen = sizeof(permitted_clock_jitter),
++ .mode = 0644,
++ .proc_handler = proc_doulongvec_minmax
++ },
++ { 0 }
++};
++static ctl_table xen_table[] = {
++ {
++ .ctl_name = 123,
++ .procname = "xen",
++ .mode = 0555,
++ .child = xen_subtable},
++ { 0 }
++};
++static int __init xen_sysctl_init(void)
++{
++ (void)register_sysctl_table(xen_table, 0);
+ return 0;
+}
++__initcall(xen_sysctl_init);
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/traps-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/traps-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1190 @@
++/*
++ * linux/arch/i386/traps.c
++ *
++ * Copyright (C) 1991, 1992 Linus Torvalds
++ *
++ * Pentium III FXSR, SSE support
++ * Gareth Hughes <gareth at valinux.com>, May 2000
++ */
+
-+EXPORT_SYMBOL(do_settimeofday);
++/*
++ * 'Traps.c' handles hardware traps and faults after we have saved some
++ * state in 'asm.s'.
++ */
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/string.h>
++#include <linux/errno.h>
++#include <linux/timer.h>
++#include <linux/mm.h>
++#include <linux/init.h>
++#include <linux/delay.h>
++#include <linux/spinlock.h>
++#include <linux/interrupt.h>
++#include <linux/highmem.h>
++#include <linux/kallsyms.h>
++#include <linux/ptrace.h>
++#include <linux/utsname.h>
++#include <linux/kprobes.h>
++#include <linux/kexec.h>
++#include <linux/unwind.h>
+
-+static void sync_xen_wallclock(unsigned long dummy);
-+static DEFINE_TIMER(sync_xen_wallclock_timer, sync_xen_wallclock, 0, 0);
-+static void sync_xen_wallclock(unsigned long dummy)
-+{
-+ time_t sec;
-+ s64 nsec;
-+ struct xen_platform_op op;
++#ifdef CONFIG_EISA
++#include <linux/ioport.h>
++#include <linux/eisa.h>
++#endif
+
-+ if (!ntp_synced() || independent_wallclock || !is_initial_xendomain())
-+ return;
++#ifdef CONFIG_MCA
++#include <linux/mca.h>
++#endif
+
-+ write_seqlock_irq(&xtime_lock);
++#include <asm/processor.h>
++#include <asm/system.h>
++#include <asm/uaccess.h>
++#include <asm/io.h>
++#include <asm/atomic.h>
++#include <asm/debugreg.h>
++#include <asm/desc.h>
++#include <asm/i387.h>
++#include <asm/nmi.h>
++#include <asm/unwind.h>
++#include <asm/smp.h>
++#include <asm/arch_hooks.h>
++#include <asm/kdebug.h>
+
-+ sec = xtime.tv_sec;
-+ nsec = xtime.tv_nsec + ((jiffies - wall_jiffies) * (u64)NS_PER_TICK);
-+ __normalize_time(&sec, &nsec);
++#include <linux/module.h>
+
-+ op.cmd = XENPF_settime;
-+ op.u.settime.secs = sec;
-+ op.u.settime.nsecs = nsec;
-+ op.u.settime.system_time = processed_system_time;
-+ HYPERVISOR_platform_op(&op);
++#include "mach_traps.h"
+
-+ update_wallclock();
++asmlinkage int system_call(void);
+
-+ write_sequnlock_irq(&xtime_lock);
++struct desc_struct default_ldt[] = { { 0, 0 }, { 0, 0 }, { 0, 0 },
++ { 0, 0 }, { 0, 0 } };
+
-+ /* Once per minute. */
-+ mod_timer(&sync_xen_wallclock_timer, jiffies + 60*HZ);
-+}
++/* Do we ignore FPU interrupts ? */
++char ignore_fpu_irq = 0;
+
-+static int set_rtc_mmss(unsigned long nowtime)
-+{
-+ int retval;
-+ unsigned long flags;
++#ifndef CONFIG_X86_NO_IDT
++/*
++ * The IDT has to be page-aligned to simplify the Pentium
++ * F0 0F bug workaround.. We have a special link segment
++ * for this.
++ */
++struct desc_struct idt_table[256] __attribute__((__section__(".data.idt"))) = { {0, 0}, };
++#endif
+
-+ if (independent_wallclock || !is_initial_xendomain())
-+ return 0;
++asmlinkage void divide_error(void);
++asmlinkage void debug(void);
++asmlinkage void nmi(void);
++asmlinkage void int3(void);
++asmlinkage void overflow(void);
++asmlinkage void bounds(void);
++asmlinkage void invalid_op(void);
++asmlinkage void device_not_available(void);
++asmlinkage void coprocessor_segment_overrun(void);
++asmlinkage void invalid_TSS(void);
++asmlinkage void segment_not_present(void);
++asmlinkage void stack_segment(void);
++asmlinkage void general_protection(void);
++asmlinkage void page_fault(void);
++asmlinkage void coprocessor_error(void);
++asmlinkage void simd_coprocessor_error(void);
++asmlinkage void alignment_check(void);
++#ifndef CONFIG_XEN
++asmlinkage void spurious_interrupt_bug(void);
++#else
++asmlinkage void fixup_4gb_segment(void);
++#endif
++asmlinkage void machine_check(void);
+
-+ /* gets recalled with irq locally disabled */
-+ /* XXX - does irqsave resolve this? -johnstul */
-+ spin_lock_irqsave(&rtc_lock, flags);
-+ if (efi_enabled)
-+ retval = efi_set_rtc_mmss(nowtime);
-+ else
-+ retval = mach_set_rtc_mmss(nowtime);
-+ spin_unlock_irqrestore(&rtc_lock, flags);
++static int kstack_depth_to_print = 24;
++#ifdef CONFIG_STACK_UNWIND
++static int call_trace = 1;
++#else
++#define call_trace (-1)
++#endif
++ATOMIC_NOTIFIER_HEAD(i386die_chain);
+
-+ return retval;
++int register_die_notifier(struct notifier_block *nb)
++{
++ vmalloc_sync_all();
++ return atomic_notifier_chain_register(&i386die_chain, nb);
+}
++EXPORT_SYMBOL(register_die_notifier); /* used modular by kdb */
+
-+/* monotonic_clock(): returns # of nanoseconds passed since time_init()
-+ * Note: This function is required to return accurate
-+ * time even in the absence of multiple timer ticks.
-+ */
-+unsigned long long monotonic_clock(void)
++int unregister_die_notifier(struct notifier_block *nb)
+{
-+ int cpu = get_cpu();
-+ struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
-+ u64 time;
-+ u32 local_time_version;
-+
-+ do {
-+ local_time_version = shadow->version;
-+ barrier();
-+ time = shadow->system_timestamp + get_nsec_offset(shadow);
-+ if (!time_values_up_to_date(cpu))
-+ get_time_values_from_xen(cpu);
-+ barrier();
-+ } while (local_time_version != shadow->version);
-+
-+ put_cpu();
++ return atomic_notifier_chain_unregister(&i386die_chain, nb);
++}
++EXPORT_SYMBOL(unregister_die_notifier); /* used modular by kdb */
+
-+ return time;
++static inline int valid_stack_ptr(struct thread_info *tinfo, void *p)
++{
++ return p > (void *)tinfo &&
++ p < (void *)tinfo + THREAD_SIZE - 3;
+}
-+EXPORT_SYMBOL(monotonic_clock);
+
-+#ifdef __x86_64__
-+unsigned long long sched_clock(void)
++/*
++ * Print one address/symbol entries per line.
++ */
++static inline void print_addr_and_symbol(unsigned long addr, char *log_lvl)
+{
-+ return monotonic_clock();
++ printk(" [<%08lx>] ", addr);
++
++ print_symbol("%s\n", addr);
+}
-+#endif
+
-+#if defined(CONFIG_SMP) && defined(CONFIG_FRAME_POINTER)
-+unsigned long profile_pc(struct pt_regs *regs)
++static inline unsigned long print_context_stack(struct thread_info *tinfo,
++ unsigned long *stack, unsigned long ebp,
++ char *log_lvl)
+{
-+ unsigned long pc = instruction_pointer(regs);
++ unsigned long addr;
+
-+#ifdef __x86_64__
-+ /* Assume the lock function has either no stack frame or only a single word.
-+ This checks if the address on the stack looks like a kernel text address.
-+ There is a small window for false hits, but in that case the tick
-+ is just accounted to the spinlock function.
-+ Better would be to write these functions in assembler again
-+ and check exactly. */
-+ if (!user_mode_vm(regs) && in_lock_functions(pc)) {
-+ char *v = *(char **)regs->rsp;
-+ if ((v >= _stext && v <= _etext) ||
-+ (v >= _sinittext && v <= _einittext) ||
-+ (v >= (char *)MODULES_VADDR && v <= (char *)MODULES_END))
-+ return (unsigned long)v;
-+ return ((unsigned long *)regs->rsp)[1];
++#ifdef CONFIG_FRAME_POINTER
++ while (valid_stack_ptr(tinfo, (void *)ebp)) {
++ addr = *(unsigned long *)(ebp + 4);
++ print_addr_and_symbol(addr, log_lvl);
++ /*
++ * break out of recursive entries (such as
++ * end_of_stack_stop_unwind_function):
++ */
++ if (ebp == *(unsigned long *)ebp)
++ break;
++ ebp = *(unsigned long *)ebp;
+ }
+#else
-+ if (!user_mode_vm(regs) && in_lock_functions(pc))
-+ return *(unsigned long *)(regs->ebp + 4);
++ while (valid_stack_ptr(tinfo, stack)) {
++ addr = *stack++;
++ if (__kernel_text_address(addr))
++ print_addr_and_symbol(addr, log_lvl);
++ }
+#endif
-+
-+ return pc;
++ return ebp;
+}
-+EXPORT_SYMBOL(profile_pc);
-+#endif
+
-+/*
-+ * This is the same as the above, except we _also_ save the current
-+ * Time Stamp Counter value at the time of the timer interrupt, so that
-+ * we later on can estimate the time of day more exactly.
-+ */
-+irqreturn_t timer_interrupt(int irq, void *dev_id, struct pt_regs *regs)
++static asmlinkage int
++show_trace_unwind(struct unwind_frame_info *info, void *log_lvl)
+{
-+ s64 delta, delta_cpu, stolen, blocked;
-+ u64 sched_time;
-+ int i, cpu = smp_processor_id();
-+ struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
-+ struct vcpu_runstate_info *runstate = &per_cpu(runstate, cpu);
++ int n = 0;
+
-+ /*
-+ * Here we are in the timer irq handler. We just have irqs locally
-+ * disabled but we don't know if the timer_bh is running on the other
-+ * CPU. We need to avoid to SMP race with it. NOTE: we don' t need
-+ * the irq version of write_lock because as just said we have irq
-+ * locally disabled. -arca
-+ */
-+ write_seqlock(&xtime_lock);
++ while (unwind(info) == 0 && UNW_PC(info)) {
++ n++;
++ print_addr_and_symbol(UNW_PC(info), log_lvl);
++ if (arch_unw_user_mode(info))
++ break;
++ }
++ return n;
++}
+
-+ do {
-+ get_time_values_from_xen(cpu);
++static void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
++ unsigned long *stack, char *log_lvl)
++{
++ unsigned long ebp;
+
-+ /* Obtain a consistent snapshot of elapsed wallclock cycles. */
-+ delta = delta_cpu =
-+ shadow->system_timestamp + get_nsec_offset(shadow);
-+ delta -= processed_system_time;
-+ delta_cpu -= per_cpu(processed_system_time, cpu);
++ if (!task)
++ task = current;
+
-+ /*
-+ * Obtain a consistent snapshot of stolen/blocked cycles. We
-+ * can use state_entry_time to detect if we get preempted here.
-+ */
-+ do {
-+ sched_time = runstate->state_entry_time;
-+ barrier();
-+ stolen = runstate->time[RUNSTATE_runnable] +
-+ runstate->time[RUNSTATE_offline] -
-+ per_cpu(processed_stolen_time, cpu);
-+ blocked = runstate->time[RUNSTATE_blocked] -
-+ per_cpu(processed_blocked_time, cpu);
-+ barrier();
-+ } while (sched_time != runstate->state_entry_time);
-+ } while (!time_values_up_to_date(cpu));
++ if (call_trace >= 0) {
++ int unw_ret = 0;
++ struct unwind_frame_info info;
+
-+ if ((unlikely(delta < -(s64)permitted_clock_jitter) ||
-+ unlikely(delta_cpu < -(s64)permitted_clock_jitter))
-+ && printk_ratelimit()) {
-+ printk("Timer ISR/%d: Time went backwards: "
-+ "delta=%lld delta_cpu=%lld shadow=%lld "
-+ "off=%lld processed=%lld cpu_processed=%lld\n",
-+ cpu, delta, delta_cpu, shadow->system_timestamp,
-+ (s64)get_nsec_offset(shadow),
-+ processed_system_time,
-+ per_cpu(processed_system_time, cpu));
-+ for (i = 0; i < num_online_cpus(); i++)
-+ printk(" %d: %lld\n", i,
-+ per_cpu(processed_system_time, i));
++ if (regs) {
++ if (unwind_init_frame_info(&info, task, regs) == 0)
++ unw_ret = show_trace_unwind(&info, log_lvl);
++ } else if (task == current)
++ unw_ret = unwind_init_running(&info, show_trace_unwind, log_lvl);
++ else {
++ if (unwind_init_blocked(&info, task) == 0)
++ unw_ret = show_trace_unwind(&info, log_lvl);
++ }
++ if (unw_ret > 0) {
++ if (call_trace == 1 && !arch_unw_user_mode(&info)) {
++ print_symbol("DWARF2 unwinder stuck at %s\n",
++ UNW_PC(&info));
++ if (UNW_SP(&info) >= PAGE_OFFSET) {
++ printk("Leftover inexact backtrace:\n");
++ stack = (void *)UNW_SP(&info);
++ } else
++ printk("Full inexact backtrace again:\n");
++ } else if (call_trace >= 1)
++ return;
++ else
++ printk("Full inexact backtrace again:\n");
++ } else
++ printk("Inexact backtrace:\n");
+ }
+
-+ /* System-wide jiffy work. */
-+ while (delta >= NS_PER_TICK) {
-+ delta -= NS_PER_TICK;
-+ processed_system_time += NS_PER_TICK;
-+ do_timer(regs);
++ if (task == current) {
++ /* Grab ebp right from our regs */
++ asm ("movl %%ebp, %0" : "=r" (ebp) : );
++ } else {
++ /* ebp is the last reg pushed by switch_to */
++ ebp = *(unsigned long *) task->thread.esp;
+ }
+
-+ if (shadow_tv_version != HYPERVISOR_shared_info->wc_version) {
-+ update_wallclock();
-+ if (keventd_up())
-+ schedule_work(&clock_was_set_work);
++ while (1) {
++ struct thread_info *context;
++ context = (struct thread_info *)
++ ((unsigned long)stack & (~(THREAD_SIZE - 1)));
++ ebp = print_context_stack(context, stack, ebp, log_lvl);
++ stack = (unsigned long*)context->previous_esp;
++ if (!stack)
++ break;
++ printk("%s =======================\n", log_lvl);
+ }
++}
+
-+ write_sequnlock(&xtime_lock);
-+
-+ /*
-+ * Account stolen ticks.
-+ * HACK: Passing NULL to account_steal_time()
-+ * ensures that the ticks are accounted as stolen.
-+ */
-+ if ((stolen > 0) && (delta_cpu > 0)) {
-+ delta_cpu -= stolen;
-+ if (unlikely(delta_cpu < 0))
-+ stolen += delta_cpu; /* clamp local-time progress */
-+ do_div(stolen, NS_PER_TICK);
-+ per_cpu(processed_stolen_time, cpu) += stolen * NS_PER_TICK;
-+ per_cpu(processed_system_time, cpu) += stolen * NS_PER_TICK;
-+ account_steal_time(NULL, (cputime_t)stolen);
-+ }
++void show_trace(struct task_struct *task, struct pt_regs *regs, unsigned long * stack)
++{
++ show_trace_log_lvl(task, regs, stack, "");
++}
+
-+ /*
-+ * Account blocked ticks.
-+ * HACK: Passing idle_task to account_steal_time()
-+ * ensures that the ticks are accounted as idle/wait.
-+ */
-+ if ((blocked > 0) && (delta_cpu > 0)) {
-+ delta_cpu -= blocked;
-+ if (unlikely(delta_cpu < 0))
-+ blocked += delta_cpu; /* clamp local-time progress */
-+ do_div(blocked, NS_PER_TICK);
-+ per_cpu(processed_blocked_time, cpu) += blocked * NS_PER_TICK;
-+ per_cpu(processed_system_time, cpu) += blocked * NS_PER_TICK;
-+ account_steal_time(idle_task(cpu), (cputime_t)blocked);
-+ }
++static void show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
++ unsigned long *esp, char *log_lvl)
++{
++ unsigned long *stack;
++ int i;
+
-+ /* Account user/system ticks. */
-+ if (delta_cpu > 0) {
-+ do_div(delta_cpu, NS_PER_TICK);
-+ per_cpu(processed_system_time, cpu) += delta_cpu * NS_PER_TICK;
-+ if (user_mode_vm(regs))
-+ account_user_time(current, (cputime_t)delta_cpu);
++ if (esp == NULL) {
++ if (task)
++ esp = (unsigned long*)task->thread.esp;
+ else
-+ account_system_time(current, HARDIRQ_OFFSET,
-+ (cputime_t)delta_cpu);
++ esp = (unsigned long *)&esp;
+ }
+
-+ /* Offlined for more than a few seconds? Avoid lockup warnings. */
-+ if (stolen > 5*HZ)
-+ touch_softlockup_watchdog();
-+
-+ /* Local timer processing (see update_process_times()). */
-+ run_local_timers();
-+ if (rcu_pending(cpu))
-+ rcu_check_callbacks(cpu, user_mode_vm(regs));
-+ scheduler_tick();
-+ run_posix_cpu_timers(current);
-+ profile_tick(CPU_PROFILING, regs);
-+
-+ return IRQ_HANDLED;
++ stack = esp;
++ for(i = 0; i < kstack_depth_to_print; i++) {
++ if (kstack_end(stack))
++ break;
++ if (i && ((i % 8) == 0))
++ printk("\n%s ", log_lvl);
++ printk("%08lx ", *stack++);
++ }
++ printk("\n%sCall Trace:\n", log_lvl);
++ show_trace_log_lvl(task, regs, esp, log_lvl);
+}
+
-+static void init_missing_ticks_accounting(int cpu)
++void show_stack(struct task_struct *task, unsigned long *esp)
+{
-+ struct vcpu_register_runstate_memory_area area;
-+ struct vcpu_runstate_info *runstate = &per_cpu(runstate, cpu);
-+ int rc;
-+
-+ memset(runstate, 0, sizeof(*runstate));
-+
-+ area.addr.v = runstate;
-+ rc = HYPERVISOR_vcpu_op(VCPUOP_register_runstate_memory_area, cpu, &area);
-+ WARN_ON(rc && rc != -ENOSYS);
-+
-+ per_cpu(processed_blocked_time, cpu) =
-+ runstate->time[RUNSTATE_blocked];
-+ per_cpu(processed_stolen_time, cpu) =
-+ runstate->time[RUNSTATE_runnable] +
-+ runstate->time[RUNSTATE_offline];
++ printk(" ");
++ show_stack_log_lvl(task, NULL, esp, "");
+}
+
-+/* not static: needed by APM */
-+unsigned long get_cmos_time(void)
++/*
++ * The architecture-independent dump_stack generator
++ */
++void dump_stack(void)
+{
-+ unsigned long retval;
-+ unsigned long flags;
-+
-+ spin_lock_irqsave(&rtc_lock, flags);
-+
-+ if (efi_enabled)
-+ retval = efi_get_time();
-+ else
-+ retval = mach_get_cmos_time();
-+
-+ spin_unlock_irqrestore(&rtc_lock, flags);
++ unsigned long stack;
+
-+ return retval;
++ show_trace(current, NULL, &stack);
+}
-+EXPORT_SYMBOL(get_cmos_time);
-+
-+static void sync_cmos_clock(unsigned long dummy);
+
-+static DEFINE_TIMER(sync_cmos_timer, sync_cmos_clock, 0, 0);
++EXPORT_SYMBOL(dump_stack);
+
-+static void sync_cmos_clock(unsigned long dummy)
++void show_registers(struct pt_regs *regs)
+{
-+ struct timeval now, next;
-+ int fail = 1;
++ int i;
++ int in_kernel = 1;
++ unsigned long esp;
++ unsigned short ss;
+
++ esp = (unsigned long) (®s->esp);
++ savesegment(ss, ss);
++ if (user_mode_vm(regs)) {
++ in_kernel = 0;
++ esp = regs->esp;
++ ss = regs->xss & 0xffff;
++ }
++ print_modules();
++ printk(KERN_EMERG "CPU: %d\nEIP: %04x:[<%08lx>] %s VLI\n"
++ "EFLAGS: %08lx (%s %.*s) \n",
++ smp_processor_id(), 0xffff & regs->xcs, regs->eip,
++ print_tainted(), regs->eflags, system_utsname.release,
++ (int)strcspn(system_utsname.version, " "),
++ system_utsname.version);
++ print_symbol(KERN_EMERG "EIP is at %s\n", regs->eip);
++ printk(KERN_EMERG "eax: %08lx ebx: %08lx ecx: %08lx edx: %08lx\n",
++ regs->eax, regs->ebx, regs->ecx, regs->edx);
++ printk(KERN_EMERG "esi: %08lx edi: %08lx ebp: %08lx esp: %08lx\n",
++ regs->esi, regs->edi, regs->ebp, esp);
++ printk(KERN_EMERG "ds: %04x es: %04x ss: %04x\n",
++ regs->xds & 0xffff, regs->xes & 0xffff, ss);
++ printk(KERN_EMERG "Process %.*s (pid: %d, ti=%p task=%p task.ti=%p)",
++ TASK_COMM_LEN, current->comm, current->pid,
++ current_thread_info(), current, current->thread_info);
+ /*
-+ * If we have an externally synchronized Linux clock, then update
-+ * CMOS clock accordingly every ~11 minutes. Set_rtc_mmss() has to be
-+ * called as close as possible to 500 ms before the new second starts.
-+ * This code is run on a timer. If the clock is set, that timer
-+ * may not expire at the correct time. Thus, we adjust...
++ * When in-kernel, we also print out the stack and code at the
++ * time of the fault..
+ */
-+ if (!ntp_synced())
-+ /*
-+ * Not synced, exit, do not restart a timer (if one is
-+ * running, let it run out).
-+ */
-+ return;
++ if (in_kernel) {
++ u8 __user *eip;
+
-+ do_gettimeofday(&now);
-+ if (now.tv_usec >= USEC_AFTER - ((unsigned) TICK_SIZE) / 2 &&
-+ now.tv_usec <= USEC_BEFORE + ((unsigned) TICK_SIZE) / 2)
-+ fail = set_rtc_mmss(now.tv_sec);
++ printk("\n" KERN_EMERG "Stack: ");
++ show_stack_log_lvl(NULL, regs, (unsigned long *)esp, KERN_EMERG);
+
-+ next.tv_usec = USEC_AFTER - now.tv_usec;
-+ if (next.tv_usec <= 0)
-+ next.tv_usec += USEC_PER_SEC;
++ printk(KERN_EMERG "Code: ");
+
-+ if (!fail)
-+ next.tv_sec = 659;
-+ else
-+ next.tv_sec = 0;
++ eip = (u8 __user *)regs->eip - 43;
++ for (i = 0; i < 64; i++, eip++) {
++ unsigned char c;
+
-+ if (next.tv_usec >= USEC_PER_SEC) {
-+ next.tv_sec++;
-+ next.tv_usec -= USEC_PER_SEC;
++ if (eip < (u8 __user *)PAGE_OFFSET || __get_user(c, eip)) {
++ printk(" Bad EIP value.");
++ break;
++ }
++ if (eip == (u8 __user *)regs->eip)
++ printk("<%02x> ", c);
++ else
++ printk("%02x ", c);
++ }
+ }
-+ mod_timer(&sync_cmos_timer, jiffies + timeval_to_jiffies(&next));
-+}
++ printk("\n");
++}
+
-+void notify_arch_cmos_timer(void)
++static void handle_BUG(struct pt_regs *regs)
+{
-+ mod_timer(&sync_cmos_timer, jiffies + 1);
-+ mod_timer(&sync_xen_wallclock_timer, jiffies + 1);
-+}
++ unsigned long eip = regs->eip;
++ unsigned short ud2;
+
-+static long clock_cmos_diff, sleep_start;
++ if (eip < PAGE_OFFSET)
++ return;
++ if (__get_user(ud2, (unsigned short __user *)eip))
++ return;
++ if (ud2 != 0x0b0f)
++ return;
+
-+static int timer_suspend(struct sys_device *dev, pm_message_t state)
-+{
-+ /*
-+ * Estimate time zone so that set_time can update the clock
-+ */
-+ clock_cmos_diff = -get_cmos_time();
-+ clock_cmos_diff += get_seconds();
-+ sleep_start = get_cmos_time();
-+ return 0;
++ printk(KERN_EMERG "------------[ cut here ]------------\n");
++
++#ifdef CONFIG_DEBUG_BUGVERBOSE
++ do {
++ unsigned short line;
++ char *file;
++ char c;
++
++ if (__get_user(line, (unsigned short __user *)(eip + 2)))
++ break;
++ if (__get_user(file, (char * __user *)(eip + 4)) ||
++ (unsigned long)file < PAGE_OFFSET || __get_user(c, file))
++ file = "<bad filename>";
++
++ printk(KERN_EMERG "kernel BUG at %s:%d!\n", file, line);
++ return;
++ } while (0);
++#endif
++ printk(KERN_EMERG "Kernel BUG at [verbose debug info unavailable]\n");
+}
+
-+static int timer_resume(struct sys_device *dev)
++/* This is gone through when something in the kernel
++ * has done something bad and is about to be terminated.
++*/
++void die(const char * str, struct pt_regs * regs, long err)
+{
++ static struct {
++ spinlock_t lock;
++ u32 lock_owner;
++ int lock_owner_depth;
++ } die = {
++ .lock = SPIN_LOCK_UNLOCKED,
++ .lock_owner = -1,
++ .lock_owner_depth = 0
++ };
++ static int die_counter;
+ unsigned long flags;
-+ unsigned long sec;
-+ unsigned long sleep_length;
+
-+#ifdef CONFIG_HPET_TIMER
-+ if (is_hpet_enabled())
-+ hpet_reenable();
-+#endif
-+ sec = get_cmos_time() + clock_cmos_diff;
-+ sleep_length = (get_cmos_time() - sleep_start) * HZ;
-+ write_seqlock_irqsave(&xtime_lock, flags);
-+ xtime.tv_sec = sec;
-+ xtime.tv_nsec = 0;
-+ jiffies_64 += sleep_length;
-+ wall_jiffies += sleep_length;
-+ write_sequnlock_irqrestore(&xtime_lock, flags);
-+ touch_softlockup_watchdog();
-+ return 0;
-+}
++ oops_enter();
+
-+static struct sysdev_class timer_sysclass = {
-+ .resume = timer_resume,
-+ .suspend = timer_suspend,
-+ set_kset_name("timer"),
-+};
++ if (die.lock_owner != raw_smp_processor_id()) {
++ console_verbose();
++ spin_lock_irqsave(&die.lock, flags);
++ die.lock_owner = smp_processor_id();
++ die.lock_owner_depth = 0;
++ bust_spinlocks(1);
++ }
++ else
++ local_save_flags(flags);
+
++ if (++die.lock_owner_depth < 3) {
++ int nl = 0;
++ unsigned long esp;
++ unsigned short ss;
+
-+/* XXX this driverfs stuff should probably go elsewhere later -john */
-+static struct sys_device device_timer = {
-+ .id = 0,
-+ .cls = &timer_sysclass,
-+};
++ handle_BUG(regs);
++ printk(KERN_EMERG "%s: %04lx [#%d]\n", str, err & 0xffff, ++die_counter);
++#ifdef CONFIG_PREEMPT
++ printk(KERN_EMERG "PREEMPT ");
++ nl = 1;
++#endif
++#ifdef CONFIG_SMP
++ if (!nl)
++ printk(KERN_EMERG);
++ printk("SMP ");
++ nl = 1;
++#endif
++#ifdef CONFIG_DEBUG_PAGEALLOC
++ if (!nl)
++ printk(KERN_EMERG);
++ printk("DEBUG_PAGEALLOC");
++ nl = 1;
++#endif
++ if (nl)
++ printk("\n");
++ if (notify_die(DIE_OOPS, str, regs, err,
++ current->thread.trap_no, SIGSEGV) !=
++ NOTIFY_STOP) {
++ show_registers(regs);
++ /* Executive summary in case the oops scrolled away */
++ esp = (unsigned long) (®s->esp);
++ savesegment(ss, ss);
++ if (user_mode(regs)) {
++ esp = regs->esp;
++ ss = regs->xss & 0xffff;
++ }
++ printk(KERN_EMERG "EIP: [<%08lx>] ", regs->eip);
++ print_symbol("%s", regs->eip);
++ printk(" SS:ESP %04x:%08lx\n", ss, esp);
++ }
++ else
++ regs = NULL;
++ } else
++ printk(KERN_EMERG "Recursive die() failure, output suppressed\n");
+
-+static int time_init_device(void)
-+{
-+ int error = sysdev_class_register(&timer_sysclass);
-+ if (!error)
-+ error = sysdev_register(&device_timer);
-+ return error;
-+}
++ bust_spinlocks(0);
++ die.lock_owner = -1;
++ spin_unlock_irqrestore(&die.lock, flags);
+
-+device_initcall(time_init_device);
++ if (!regs)
++ return;
+
-+#ifdef CONFIG_HPET_TIMER
-+extern void (*late_time_init)(void);
-+/* Duplicate of time_init() below, with hpet_enable part added */
-+static void __init hpet_time_init(void)
-+{
-+ xtime.tv_sec = get_cmos_time();
-+ xtime.tv_nsec = (INITIAL_JIFFIES % HZ) * (NSEC_PER_SEC / HZ);
-+ set_normalized_timespec(&wall_to_monotonic,
-+ -xtime.tv_sec, -xtime.tv_nsec);
++ if (kexec_should_crash(current))
++ crash_kexec(regs);
+
-+ if ((hpet_enable() >= 0) && hpet_use_timer) {
-+ printk("Using HPET for base-timer\n");
-+ }
++ if (in_interrupt())
++ panic("Fatal exception in interrupt");
+
-+ time_init_hook();
-+}
-+#endif
++ if (panic_on_oops)
++ panic("Fatal exception");
+
-+/* Dynamically-mapped IRQ. */
-+DEFINE_PER_CPU(int, timer_irq);
++ oops_exit();
++ do_exit(SIGSEGV);
++}
+
-+extern void (*late_time_init)(void);
-+static void setup_cpu0_timer_irq(void)
++static inline void die_if_kernel(const char * str, struct pt_regs * regs, long err)
+{
-+ per_cpu(timer_irq, 0) =
-+ bind_virq_to_irqhandler(
-+ VIRQ_TIMER,
-+ 0,
-+ timer_interrupt,
-+ SA_INTERRUPT,
-+ "timer0",
-+ NULL);
-+ BUG_ON(per_cpu(timer_irq, 0) < 0);
++ if (!user_mode_vm(regs))
++ die(str, regs, err);
+}
+
-+static struct vcpu_set_periodic_timer xen_set_periodic_tick = {
-+ .period_ns = NS_PER_TICK
-+};
-+
-+void __init time_init(void)
++static void __kprobes do_trap(int trapnr, int signr, char *str, int vm86,
++ struct pt_regs * regs, long error_code,
++ siginfo_t *info)
+{
-+#ifdef CONFIG_HPET_TIMER
-+ if (is_hpet_capable()) {
-+ /*
-+ * HPET initialization needs to do memory-mapped io. So, let
-+ * us do a late initialization after mem_init().
-+ */
-+ late_time_init = hpet_time_init;
-+ return;
-+ }
-+#endif
++ struct task_struct *tsk = current;
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = trapnr;
+
-+ HYPERVISOR_vcpu_op(VCPUOP_set_periodic_timer, 0,
-+ &xen_set_periodic_tick);
++ if (regs->eflags & VM_MASK) {
++ if (vm86)
++ goto vm86_trap;
++ goto trap_signal;
++ }
+
-+ get_time_values_from_xen(0);
++ if (!user_mode(regs))
++ goto kernel_trap;
+
-+ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
-+ per_cpu(processed_system_time, 0) = processed_system_time;
-+ init_missing_ticks_accounting(0);
++ trap_signal: {
++ if (info)
++ force_sig_info(signr, info, tsk);
++ else
++ force_sig(signr, tsk);
++ return;
++ }
+
-+ update_wallclock();
++ kernel_trap: {
++ if (!fixup_exception(regs))
++ die(str, regs, error_code);
++ return;
++ }
+
-+ init_cpu_khz();
-+ printk(KERN_INFO "Xen reported: %u.%03u MHz processor.\n",
-+ cpu_khz / 1000, cpu_khz % 1000);
++ vm86_trap: {
++ int ret = handle_vm86_trap((struct kernel_vm86_regs *) regs, error_code, trapnr);
++ if (ret) goto trap_signal;
++ return;
++ }
++}
+
-+#if defined(__x86_64__)
-+ vxtime.mode = VXTIME_TSC;
-+ vxtime.quot = (1000000L << 32) / vxtime_hz;
-+ vxtime.tsc_quot = (1000L << 32) / cpu_khz;
-+ sync_core();
-+ rdtscll(vxtime.last_tsc);
-+#endif
++#define DO_ERROR(trapnr, signr, str, name) \
++fastcall void do_##name(struct pt_regs * regs, long error_code) \
++{ \
++ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
++ == NOTIFY_STOP) \
++ return; \
++ do_trap(trapnr, signr, str, 0, regs, error_code, NULL); \
++}
+
-+ /* Cannot request_irq() until kmem is initialised. */
-+ late_time_init = setup_cpu0_timer_irq;
++#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \
++fastcall void do_##name(struct pt_regs * regs, long error_code) \
++{ \
++ siginfo_t info; \
++ info.si_signo = signr; \
++ info.si_errno = 0; \
++ info.si_code = sicode; \
++ info.si_addr = (void __user *)siaddr; \
++ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
++ == NOTIFY_STOP) \
++ return; \
++ do_trap(trapnr, signr, str, 0, regs, error_code, &info); \
+}
+
-+/* Convert jiffies to system time. */
-+u64 jiffies_to_st(unsigned long j)
-+{
-+ unsigned long seq;
-+ long delta;
-+ u64 st;
-+
-+ do {
-+ seq = read_seqbegin(&xtime_lock);
-+ delta = j - jiffies;
-+ if (delta < 1) {
-+ /* Triggers in some wrap-around cases, but that's okay:
-+ * we just end up with a shorter timeout. */
-+ st = processed_system_time + NS_PER_TICK;
-+ } else if (((unsigned long)delta >> (BITS_PER_LONG-3)) != 0) {
-+ /* Very long timeout means there is no pending timer.
-+ * We indicate this to Xen by passing zero timeout. */
-+ st = 0;
-+ } else {
-+ st = processed_system_time + delta * (u64)NS_PER_TICK;
-+ }
-+ } while (read_seqretry(&xtime_lock, seq));
++#define DO_VM86_ERROR(trapnr, signr, str, name) \
++fastcall void do_##name(struct pt_regs * regs, long error_code) \
++{ \
++ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
++ == NOTIFY_STOP) \
++ return; \
++ do_trap(trapnr, signr, str, 1, regs, error_code, NULL); \
++}
+
-+ return st;
++#define DO_VM86_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \
++fastcall void do_##name(struct pt_regs * regs, long error_code) \
++{ \
++ siginfo_t info; \
++ info.si_signo = signr; \
++ info.si_errno = 0; \
++ info.si_code = sicode; \
++ info.si_addr = (void __user *)siaddr; \
++ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
++ == NOTIFY_STOP) \
++ return; \
++ do_trap(trapnr, signr, str, 1, regs, error_code, &info); \
+}
-+EXPORT_SYMBOL(jiffies_to_st);
+
-+/*
-+ * stop_hz_timer / start_hz_timer - enter/exit 'tickless mode' on an idle cpu
-+ * These functions are based on implementations from arch/s390/kernel/time.c
-+ */
-+static void stop_hz_timer(void)
++DO_VM86_ERROR_INFO( 0, SIGFPE, "divide error", divide_error, FPE_INTDIV, regs->eip)
++#ifndef CONFIG_KPROBES
++DO_VM86_ERROR( 3, SIGTRAP, "int3", int3)
++#endif
++DO_VM86_ERROR( 4, SIGSEGV, "overflow", overflow)
++DO_VM86_ERROR( 5, SIGSEGV, "bounds", bounds)
++DO_ERROR_INFO( 6, SIGILL, "invalid opcode", invalid_op, ILL_ILLOPN, regs->eip)
++DO_ERROR( 9, SIGFPE, "coprocessor segment overrun", coprocessor_segment_overrun)
++DO_ERROR(10, SIGSEGV, "invalid TSS", invalid_TSS)
++DO_ERROR(11, SIGBUS, "segment not present", segment_not_present)
++DO_ERROR(12, SIGBUS, "stack segment", stack_segment)
++DO_ERROR_INFO(17, SIGBUS, "alignment check", alignment_check, BUS_ADRALN, 0)
++DO_ERROR_INFO(32, SIGSEGV, "iret exception", iret_error, ILL_BADSTK, 0)
++
++fastcall void __kprobes do_general_protection(struct pt_regs * regs,
++ long error_code)
+{
-+ struct vcpu_set_singleshot_timer singleshot;
-+ unsigned int cpu = smp_processor_id();
-+ unsigned long j;
-+ int rc;
++ current->thread.error_code = error_code;
++ current->thread.trap_no = 13;
+
-+ cpu_set(cpu, nohz_cpu_mask);
++ if (regs->eflags & VM_MASK)
++ goto gp_in_vm86;
+
-+ /* See matching smp_mb in rcu_start_batch in rcupdate.c. These mbs */
-+ /* ensure that if __rcu_pending (nested in rcu_needs_cpu) fetches a */
-+ /* value of rcp->cur that matches rdp->quiescbatch and allows us to */
-+ /* stop the hz timer then the cpumasks created for subsequent values */
-+ /* of cur in rcu_start_batch are guaranteed to pick up the updated */
-+ /* nohz_cpu_mask and so will not depend on this cpu. */
++ if (!user_mode(regs))
++ goto gp_in_kernel;
+
-+ smp_mb();
++ current->thread.error_code = error_code;
++ current->thread.trap_no = 13;
++ force_sig(SIGSEGV, current);
++ return;
+
-+ /* Leave ourselves in tick mode if rcu or softirq or timer pending. */
-+ if (rcu_needs_cpu(cpu) || local_softirq_pending() ||
-+ (j = next_timer_interrupt(), time_before_eq(j, jiffies))) {
-+ cpu_clear(cpu, nohz_cpu_mask);
-+ j = jiffies + 1;
-+ }
++gp_in_vm86:
++ local_irq_enable();
++ handle_vm86_fault((struct kernel_vm86_regs *) regs, error_code);
++ return;
+
-+ singleshot.timeout_abs_ns = jiffies_to_st(j);
-+ singleshot.flags = 0;
-+ rc = HYPERVISOR_vcpu_op(VCPUOP_set_singleshot_timer, cpu, &singleshot);
-+#if CONFIG_XEN_COMPAT <= 0x030004
-+ if (rc) {
-+ BUG_ON(rc != -ENOSYS);
-+ rc = HYPERVISOR_set_timer_op(singleshot.timeout_abs_ns);
++gp_in_kernel:
++ if (!fixup_exception(regs)) {
++ if (notify_die(DIE_GPF, "general protection fault", regs,
++ error_code, 13, SIGSEGV) == NOTIFY_STOP)
++ return;
++ die("general protection fault", regs, error_code);
+ }
-+#endif
-+ BUG_ON(rc);
+}
+
-+static void start_hz_timer(void)
++static void mem_parity_error(unsigned char reason, struct pt_regs * regs)
+{
-+ cpu_clear(smp_processor_id(), nohz_cpu_mask);
-+}
++ printk(KERN_EMERG "Uhhuh. NMI received. Dazed and confused, but trying "
++ "to continue\n");
++ printk(KERN_EMERG "You probably have a hardware problem with your RAM "
++ "chips\n");
+
-+void raw_safe_halt(void)
-+{
-+ stop_hz_timer();
-+ /* Blocking includes an implicit local_irq_enable(). */
-+ HYPERVISOR_block();
-+ start_hz_timer();
++ /* Clear and disable the memory parity error line. */
++ clear_mem_error(reason);
+}
-+EXPORT_SYMBOL(raw_safe_halt);
+
-+void halt(void)
++static void io_check_error(unsigned char reason, struct pt_regs * regs)
+{
-+ if (irqs_disabled())
-+ HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL);
++ printk(KERN_EMERG "NMI: IOCK error (debug interrupt?)\n");
++ show_registers(regs);
++
++ /* Re-enable the IOCK line, wait for a few seconds */
++ clear_io_check_error(reason);
+}
-+EXPORT_SYMBOL(halt);
+
-+/* No locking required. Interrupts are disabled on all CPUs. */
-+void time_resume(void)
++static void unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
+{
-+ unsigned int cpu;
-+
-+ init_cpu_khz();
-+
-+ for_each_online_cpu(cpu) {
-+ HYPERVISOR_vcpu_op(VCPUOP_set_periodic_timer, cpu,
-+ &xen_set_periodic_tick);
-+ get_time_values_from_xen(cpu);
-+ per_cpu(processed_system_time, cpu) =
-+ per_cpu(shadow_time, 0).system_timestamp;
-+ init_missing_ticks_accounting(cpu);
++#ifdef CONFIG_MCA
++ /* Might actually be able to figure out what the guilty party
++ * is. */
++ if( MCA_bus ) {
++ mca_handle_nmi();
++ return;
+ }
-+
-+ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
-+
-+ update_wallclock();
++#endif
++ printk("Uhhuh. NMI received for unknown reason %02x on CPU %d.\n",
++ reason, smp_processor_id());
++ printk("Dazed and confused, but trying to continue\n");
++ printk("Do you have a strange power saving mode enabled?\n");
+}
+
-+#ifdef CONFIG_SMP
-+static char timer_name[NR_CPUS][15];
++static DEFINE_SPINLOCK(nmi_print_lock);
+
-+int local_setup_timer(unsigned int cpu)
++void die_nmi (struct pt_regs *regs, const char *msg)
+{
-+ int seq, irq;
-+
-+ BUG_ON(cpu == 0);
-+
-+ HYPERVISOR_vcpu_op(VCPUOP_set_periodic_timer, cpu,
-+ &xen_set_periodic_tick);
++ if (notify_die(DIE_NMIWATCHDOG, msg, regs, 0, 2, SIGINT) ==
++ NOTIFY_STOP)
++ return;
+
-+ do {
-+ seq = read_seqbegin(&xtime_lock);
-+ /* Use cpu0 timestamp: cpu's shadow is not initialised yet. */
-+ per_cpu(processed_system_time, cpu) =
-+ per_cpu(shadow_time, 0).system_timestamp;
-+ init_missing_ticks_accounting(cpu);
-+ } while (read_seqretry(&xtime_lock, seq));
++ spin_lock(&nmi_print_lock);
++ /*
++ * We are in trouble anyway, lets at least try
++ * to get a message out.
++ */
++ bust_spinlocks(1);
++ printk(KERN_EMERG "%s", msg);
++ printk(" on CPU%d, eip %08lx, registers:\n",
++ smp_processor_id(), regs->eip);
++ show_registers(regs);
++ printk(KERN_EMERG "console shuts up ...\n");
++ console_silent();
++ spin_unlock(&nmi_print_lock);
++ bust_spinlocks(0);
+
-+ sprintf(timer_name[cpu], "timer%d", cpu);
-+ irq = bind_virq_to_irqhandler(VIRQ_TIMER,
-+ cpu,
-+ timer_interrupt,
-+ SA_INTERRUPT,
-+ timer_name[cpu],
-+ NULL);
-+ if (irq < 0)
-+ return irq;
-+ per_cpu(timer_irq, cpu) = irq;
++ /* If we are in kernel we are probably nested up pretty bad
++ * and might aswell get out now while we still can.
++ */
++ if (!user_mode_vm(regs)) {
++ current->thread.trap_no = 2;
++ crash_kexec(regs);
++ }
+
-+ return 0;
++ do_exit(SIGSEGV);
+}
+
-+void local_teardown_timer(unsigned int cpu)
++static void default_do_nmi(struct pt_regs * regs)
+{
-+ BUG_ON(cpu == 0);
-+ unbind_from_irqhandler(per_cpu(timer_irq, cpu), NULL);
-+}
++ unsigned char reason = 0;
++
++ /* Only the BSP gets external NMIs from the system. */
++ if (!smp_processor_id())
++ reason = get_nmi_reason();
++
++ if (!(reason & 0xc0)) {
++ if (notify_die(DIE_NMI_IPI, "nmi_ipi", regs, reason, 2, SIGINT)
++ == NOTIFY_STOP)
++ return;
++#ifdef CONFIG_X86_LOCAL_APIC
++ /*
++ * Ok, so this is none of the documented NMI sources,
++ * so it must be the NMI watchdog.
++ */
++ if (nmi_watchdog) {
++ nmi_watchdog_tick(regs);
++ return;
++ }
+#endif
++ unknown_nmi_error(reason, regs);
++ return;
++ }
++ if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP)
++ return;
++ if (reason & 0x80)
++ mem_parity_error(reason, regs);
++ if (reason & 0x40)
++ io_check_error(reason, regs);
++ /*
++ * Reassert NMI in case it became active meanwhile
++ * as it's edge-triggered.
++ */
++ reassert_nmi();
++}
+
-+/*
-+ * /proc/sys/xen: This really belongs in another file. It can stay here for
-+ * now however.
-+ */
-+static ctl_table xen_subtable[] = {
-+ {
-+ .ctl_name = 1,
-+ .procname = "independent_wallclock",
-+ .data = &independent_wallclock,
-+ .maxlen = sizeof(independent_wallclock),
-+ .mode = 0644,
-+ .proc_handler = proc_dointvec
-+ },
-+ {
-+ .ctl_name = 2,
-+ .procname = "permitted_clock_jitter",
-+ .data = &permitted_clock_jitter,
-+ .maxlen = sizeof(permitted_clock_jitter),
-+ .mode = 0644,
-+ .proc_handler = proc_doulongvec_minmax
-+ },
-+ { 0 }
-+};
-+static ctl_table xen_table[] = {
-+ {
-+ .ctl_name = 123,
-+ .procname = "xen",
-+ .mode = 0555,
-+ .child = xen_subtable},
-+ { 0 }
-+};
-+static int __init xen_sysctl_init(void)
++static int dummy_nmi_callback(struct pt_regs * regs, int cpu)
+{
-+ (void)register_sysctl_table(xen_table, 0);
+ return 0;
+}
-+__initcall(xen_sysctl_init);
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/traps.c tmp-linux-2.6-xen.patch/arch/i386/kernel/traps.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/traps.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/traps.c 2007-10-14 01:51:15.000000000 +0200
-@@ -642,18 +642,11 @@ static void mem_parity_error(unsigned ch
-
- static void io_check_error(unsigned char reason, struct pt_regs * regs)
- {
-- unsigned long i;
--
- printk(KERN_EMERG "NMI: IOCK error (debug interrupt?)\n");
- show_registers(regs);
-
- /* Re-enable the IOCK line, wait for a few seconds */
-- reason = (reason & 0xf) | 8;
-- outb(reason, 0x61);
-- i = 2000;
-- while (--i) udelay(1000);
-- reason &= ~8;
-- outb(reason, 0x61);
-+ clear_io_check_error(reason);
- }
-
- static void unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/traps-xen.c tmp-linux-2.6-xen.patch/arch/i386/kernel/traps-xen.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/traps-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/traps-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,1186 @@
-+/*
-+ * linux/arch/i386/traps.c
-+ *
-+ * Copyright (C) 1991, 1992 Linus Torvalds
-+ *
-+ * Pentium III FXSR, SSE support
-+ * Gareth Hughes <gareth at valinux.com>, May 2000
-+ */
-+
-+/*
-+ * 'Traps.c' handles hardware traps and faults after we have saved some
-+ * state in 'asm.s'.
-+ */
-+#include <linux/sched.h>
-+#include <linux/kernel.h>
-+#include <linux/string.h>
-+#include <linux/errno.h>
-+#include <linux/timer.h>
-+#include <linux/mm.h>
-+#include <linux/init.h>
-+#include <linux/delay.h>
-+#include <linux/spinlock.h>
-+#include <linux/interrupt.h>
-+#include <linux/highmem.h>
-+#include <linux/kallsyms.h>
-+#include <linux/ptrace.h>
-+#include <linux/utsname.h>
-+#include <linux/kprobes.h>
-+#include <linux/kexec.h>
-+#include <linux/unwind.h>
-+
-+#ifdef CONFIG_EISA
-+#include <linux/ioport.h>
-+#include <linux/eisa.h>
-+#endif
-+
-+#ifdef CONFIG_MCA
-+#include <linux/mca.h>
-+#endif
-+
-+#include <asm/processor.h>
-+#include <asm/system.h>
-+#include <asm/uaccess.h>
-+#include <asm/io.h>
-+#include <asm/atomic.h>
-+#include <asm/debugreg.h>
-+#include <asm/desc.h>
-+#include <asm/i387.h>
-+#include <asm/nmi.h>
-+#include <asm/unwind.h>
-+#include <asm/smp.h>
-+#include <asm/arch_hooks.h>
-+#include <asm/kdebug.h>
-+
-+#include <linux/module.h>
-+
-+#include "mach_traps.h"
-+
-+asmlinkage int system_call(void);
++
++static nmi_callback_t nmi_callback = dummy_nmi_callback;
++
++fastcall void do_nmi(struct pt_regs * regs, long error_code)
++{
++ int cpu;
+
-+struct desc_struct default_ldt[] = { { 0, 0 }, { 0, 0 }, { 0, 0 },
-+ { 0, 0 }, { 0, 0 } };
++ nmi_enter();
+
-+/* Do we ignore FPU interrupts ? */
-+char ignore_fpu_irq = 0;
++ cpu = smp_processor_id();
+
-+#ifndef CONFIG_X86_NO_IDT
-+/*
-+ * The IDT has to be page-aligned to simplify the Pentium
-+ * F0 0F bug workaround.. We have a special link segment
-+ * for this.
-+ */
-+struct desc_struct idt_table[256] __attribute__((__section__(".data.idt"))) = { {0, 0}, };
-+#endif
++ ++nmi_count(cpu);
+
-+asmlinkage void divide_error(void);
-+asmlinkage void debug(void);
-+asmlinkage void nmi(void);
-+asmlinkage void int3(void);
-+asmlinkage void overflow(void);
-+asmlinkage void bounds(void);
-+asmlinkage void invalid_op(void);
-+asmlinkage void device_not_available(void);
-+asmlinkage void coprocessor_segment_overrun(void);
-+asmlinkage void invalid_TSS(void);
-+asmlinkage void segment_not_present(void);
-+asmlinkage void stack_segment(void);
-+asmlinkage void general_protection(void);
-+asmlinkage void page_fault(void);
-+asmlinkage void coprocessor_error(void);
-+asmlinkage void simd_coprocessor_error(void);
-+asmlinkage void alignment_check(void);
-+#ifndef CONFIG_XEN
-+asmlinkage void spurious_interrupt_bug(void);
-+#else
-+asmlinkage void fixup_4gb_segment(void);
-+#endif
-+asmlinkage void machine_check(void);
++ if (!rcu_dereference(nmi_callback)(regs, cpu))
++ default_do_nmi(regs);
+
-+static int kstack_depth_to_print = 24;
-+#ifdef CONFIG_STACK_UNWIND
-+static int call_trace = 1;
-+#else
-+#define call_trace (-1)
-+#endif
-+ATOMIC_NOTIFIER_HEAD(i386die_chain);
++ nmi_exit();
++}
+
-+int register_die_notifier(struct notifier_block *nb)
++void set_nmi_callback(nmi_callback_t callback)
+{
+ vmalloc_sync_all();
-+ return atomic_notifier_chain_register(&i386die_chain, nb);
++ rcu_assign_pointer(nmi_callback, callback);
+}
-+EXPORT_SYMBOL(register_die_notifier); /* used modular by kdb */
++EXPORT_SYMBOL_GPL(set_nmi_callback);
+
-+int unregister_die_notifier(struct notifier_block *nb)
++void unset_nmi_callback(void)
+{
-+ return atomic_notifier_chain_unregister(&i386die_chain, nb);
++ nmi_callback = dummy_nmi_callback;
+}
-+EXPORT_SYMBOL(unregister_die_notifier); /* used modular by kdb */
++EXPORT_SYMBOL_GPL(unset_nmi_callback);
+
-+static inline int valid_stack_ptr(struct thread_info *tinfo, void *p)
++#ifdef CONFIG_KPROBES
++fastcall void __kprobes do_int3(struct pt_regs *regs, long error_code)
+{
-+ return p > (void *)tinfo &&
-+ p < (void *)tinfo + THREAD_SIZE - 3;
++ if (notify_die(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP)
++ == NOTIFY_STOP)
++ return;
++ /* This is an interrupt gate, because kprobes wants interrupts
++ disabled. Normal trap handlers don't. */
++ restore_interrupts(regs);
++ do_trap(3, SIGTRAP, "int3", 1, regs, error_code, NULL);
+}
++#endif
+
+/*
-+ * Print one address/symbol entries per line.
++ * Our handling of the processor debug registers is non-trivial.
++ * We do not clear them on entry and exit from the kernel. Therefore
++ * it is possible to get a watchpoint trap here from inside the kernel.
++ * However, the code in ./ptrace.c has ensured that the user can
++ * only set watchpoints on userspace addresses. Therefore the in-kernel
++ * watchpoint trap can only occur in code which is reading/writing
++ * from user space. Such code must not hold kernel locks (since it
++ * can equally take a page fault), therefore it is safe to call
++ * force_sig_info even though that claims and releases locks.
++ *
++ * Code in ./signal.c ensures that the debug control register
++ * is restored before we deliver any signal, and therefore that
++ * user code runs with the correct debug control register even though
++ * we clear it here.
++ *
++ * Being careful here means that we don't have to be as careful in a
++ * lot of more complicated places (task switching can be a bit lazy
++ * about restoring all the debug state, and ptrace doesn't have to
++ * find every occurrence of the TF bit that could be saved away even
++ * by user code)
+ */
-+static inline void print_addr_and_symbol(unsigned long addr, char *log_lvl)
++fastcall void __kprobes do_debug(struct pt_regs * regs, long error_code)
+{
-+ printk(" [<%08lx>] ", addr);
++ unsigned int condition;
++ struct task_struct *tsk = current;
+
-+ print_symbol("%s\n", addr);
-+}
++ get_debugreg(condition, 6);
+
-+static inline unsigned long print_context_stack(struct thread_info *tinfo,
-+ unsigned long *stack, unsigned long ebp,
-+ char *log_lvl)
-+{
-+ unsigned long addr;
++ if (notify_die(DIE_DEBUG, "debug", regs, condition, error_code,
++ SIGTRAP) == NOTIFY_STOP)
++ return;
++ /* It's safe to allow irq's after DR6 has been saved */
++ if (regs->eflags & X86_EFLAGS_IF)
++ local_irq_enable();
+
-+#ifdef CONFIG_FRAME_POINTER
-+ while (valid_stack_ptr(tinfo, (void *)ebp)) {
-+ addr = *(unsigned long *)(ebp + 4);
-+ print_addr_and_symbol(addr, log_lvl);
-+ /*
-+ * break out of recursive entries (such as
-+ * end_of_stack_stop_unwind_function):
-+ */
-+ if (ebp == *(unsigned long *)ebp)
-+ break;
-+ ebp = *(unsigned long *)ebp;
-+ }
-+#else
-+ while (valid_stack_ptr(tinfo, stack)) {
-+ addr = *stack++;
-+ if (__kernel_text_address(addr))
-+ print_addr_and_symbol(addr, log_lvl);
++ /* Mask out spurious debug traps due to lazy DR7 setting */
++ if (condition & (DR_TRAP0|DR_TRAP1|DR_TRAP2|DR_TRAP3)) {
++ if (!tsk->thread.debugreg[7])
++ goto clear_dr7;
+ }
-+#endif
-+ return ebp;
-+}
+
-+static asmlinkage int
-+show_trace_unwind(struct unwind_frame_info *info, void *log_lvl)
-+{
-+ int n = 0;
++ if (regs->eflags & VM_MASK)
++ goto debug_vm86;
+
-+ while (unwind(info) == 0 && UNW_PC(info)) {
-+ n++;
-+ print_addr_and_symbol(UNW_PC(info), log_lvl);
-+ if (arch_unw_user_mode(info))
-+ break;
++ /* Save debug status register where ptrace can see it */
++ tsk->thread.debugreg[6] = condition;
++
++ /*
++ * Single-stepping through TF: make sure we ignore any events in
++ * kernel space (but re-enable TF when returning to user mode).
++ */
++ if (condition & DR_STEP) {
++ /*
++ * We already checked v86 mode above, so we can
++ * check for kernel mode by just checking the CPL
++ * of CS.
++ */
++ if (!user_mode(regs))
++ goto clear_TF_reenable;
+ }
-+ return n;
-+}
+
-+static void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
-+ unsigned long *stack, char *log_lvl)
-+{
-+ unsigned long ebp;
++ /* Ok, finally something we can handle */
++ send_sigtrap(tsk, regs, error_code);
+
-+ if (!task)
-+ task = current;
++ /* Disable additional traps. They'll be re-enabled when
++ * the signal is delivered.
++ */
++clear_dr7:
++ set_debugreg(0, 7);
++ return;
+
-+ if (call_trace >= 0) {
-+ int unw_ret = 0;
-+ struct unwind_frame_info info;
++debug_vm86:
++ handle_vm86_trap((struct kernel_vm86_regs *) regs, error_code, 1);
++ return;
+
-+ if (regs) {
-+ if (unwind_init_frame_info(&info, task, regs) == 0)
-+ unw_ret = show_trace_unwind(&info, log_lvl);
-+ } else if (task == current)
-+ unw_ret = unwind_init_running(&info, show_trace_unwind, log_lvl);
-+ else {
-+ if (unwind_init_blocked(&info, task) == 0)
-+ unw_ret = show_trace_unwind(&info, log_lvl);
-+ }
-+ if (unw_ret > 0) {
-+ if (call_trace == 1 && !arch_unw_user_mode(&info)) {
-+ print_symbol("DWARF2 unwinder stuck at %s\n",
-+ UNW_PC(&info));
-+ if (UNW_SP(&info) >= PAGE_OFFSET) {
-+ printk("Leftover inexact backtrace:\n");
-+ stack = (void *)UNW_SP(&info);
-+ } else
-+ printk("Full inexact backtrace again:\n");
-+ } else if (call_trace >= 1)
-+ return;
-+ else
-+ printk("Full inexact backtrace again:\n");
-+ } else
-+ printk("Inexact backtrace:\n");
-+ }
++clear_TF_reenable:
++ set_tsk_thread_flag(tsk, TIF_SINGLESTEP);
++ regs->eflags &= ~TF_MASK;
++ return;
++}
+
-+ if (task == current) {
-+ /* Grab ebp right from our regs */
-+ asm ("movl %%ebp, %0" : "=r" (ebp) : );
-+ } else {
-+ /* ebp is the last reg pushed by switch_to */
-+ ebp = *(unsigned long *) task->thread.esp;
-+ }
++/*
++ * Note that we play around with the 'TS' bit in an attempt to get
++ * the correct behaviour even in the presence of the asynchronous
++ * IRQ13 behaviour
++ */
++void math_error(void __user *eip)
++{
++ struct task_struct * task;
++ siginfo_t info;
++ unsigned short cwd, swd;
+
-+ while (1) {
-+ struct thread_info *context;
-+ context = (struct thread_info *)
-+ ((unsigned long)stack & (~(THREAD_SIZE - 1)));
-+ ebp = print_context_stack(context, stack, ebp, log_lvl);
-+ stack = (unsigned long*)context->previous_esp;
-+ if (!stack)
++ /*
++ * Save the info for the exception handler and clear the error.
++ */
++ task = current;
++ save_init_fpu(task);
++ task->thread.trap_no = 16;
++ task->thread.error_code = 0;
++ info.si_signo = SIGFPE;
++ info.si_errno = 0;
++ info.si_code = __SI_FAULT;
++ info.si_addr = eip;
++ /*
++ * (~cwd & swd) will mask out exceptions that are not set to unmasked
++ * status. 0x3f is the exception bits in these regs, 0x200 is the
++ * C1 reg you need in case of a stack fault, 0x040 is the stack
++ * fault bit. We should only be taking one exception at a time,
++ * so if this combination doesn't produce any single exception,
++ * then we have a bad program that isn't syncronizing its FPU usage
++ * and it will suffer the consequences since we won't be able to
++ * fully reproduce the context of the exception
++ */
++ cwd = get_fpu_cwd(task);
++ swd = get_fpu_swd(task);
++ switch (swd & ~cwd & 0x3f) {
++ case 0x000: /* No unmasked exception */
++ return;
++ default: /* Multiple exceptions */
++ break;
++ case 0x001: /* Invalid Op */
++ /*
++ * swd & 0x240 == 0x040: Stack Underflow
++ * swd & 0x240 == 0x240: Stack Overflow
++ * User must clear the SF bit (0x40) if set
++ */
++ info.si_code = FPE_FLTINV;
++ break;
++ case 0x002: /* Denormalize */
++ case 0x010: /* Underflow */
++ info.si_code = FPE_FLTUND;
++ break;
++ case 0x004: /* Zero Divide */
++ info.si_code = FPE_FLTDIV;
++ break;
++ case 0x008: /* Overflow */
++ info.si_code = FPE_FLTOVF;
++ break;
++ case 0x020: /* Precision */
++ info.si_code = FPE_FLTRES;
+ break;
-+ printk("%s =======================\n", log_lvl);
+ }
++ force_sig_info(SIGFPE, &info, task);
+}
+
-+void show_trace(struct task_struct *task, struct pt_regs *regs, unsigned long * stack)
++fastcall void do_coprocessor_error(struct pt_regs * regs, long error_code)
+{
-+ show_trace_log_lvl(task, regs, stack, "");
++ ignore_fpu_irq = 1;
++ math_error((void __user *)regs->eip);
+}
+
-+static void show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
-+ unsigned long *esp, char *log_lvl)
++static void simd_math_error(void __user *eip)
+{
-+ unsigned long *stack;
-+ int i;
++ struct task_struct * task;
++ siginfo_t info;
++ unsigned short mxcsr;
+
-+ if (esp == NULL) {
-+ if (task)
-+ esp = (unsigned long*)task->thread.esp;
-+ else
-+ esp = (unsigned long *)&esp;
++ /*
++ * Save the info for the exception handler and clear the error.
++ */
++ task = current;
++ save_init_fpu(task);
++ task->thread.trap_no = 19;
++ task->thread.error_code = 0;
++ info.si_signo = SIGFPE;
++ info.si_errno = 0;
++ info.si_code = __SI_FAULT;
++ info.si_addr = eip;
++ /*
++ * The SIMD FPU exceptions are handled a little differently, as there
++ * is only a single status/control register. Thus, to determine which
++ * unmasked exception was caught we must mask the exception mask bits
++ * at 0x1f80, and then use these to mask the exception bits at 0x3f.
++ */
++ mxcsr = get_fpu_mxcsr(task);
++ switch (~((mxcsr & 0x1f80) >> 7) & (mxcsr & 0x3f)) {
++ case 0x000:
++ default:
++ break;
++ case 0x001: /* Invalid Op */
++ info.si_code = FPE_FLTINV;
++ break;
++ case 0x002: /* Denormalize */
++ case 0x010: /* Underflow */
++ info.si_code = FPE_FLTUND;
++ break;
++ case 0x004: /* Zero Divide */
++ info.si_code = FPE_FLTDIV;
++ break;
++ case 0x008: /* Overflow */
++ info.si_code = FPE_FLTOVF;
++ break;
++ case 0x020: /* Precision */
++ info.si_code = FPE_FLTRES;
++ break;
+ }
++ force_sig_info(SIGFPE, &info, task);
++}
+
-+ stack = esp;
-+ for(i = 0; i < kstack_depth_to_print; i++) {
-+ if (kstack_end(stack))
-+ break;
-+ if (i && ((i % 8) == 0))
-+ printk("\n%s ", log_lvl);
-+ printk("%08lx ", *stack++);
++fastcall void do_simd_coprocessor_error(struct pt_regs * regs,
++ long error_code)
++{
++ if (cpu_has_xmm) {
++ /* Handle SIMD FPU exceptions on PIII+ processors. */
++ ignore_fpu_irq = 1;
++ simd_math_error((void __user *)regs->eip);
++ } else {
++ /*
++ * Handle strange cache flush from user space exception
++ * in all other cases. This is undocumented behaviour.
++ */
++ if (regs->eflags & VM_MASK) {
++ handle_vm86_fault((struct kernel_vm86_regs *)regs,
++ error_code);
++ return;
++ }
++ current->thread.trap_no = 19;
++ current->thread.error_code = error_code;
++ die_if_kernel("cache flush denied", regs, error_code);
++ force_sig(SIGSEGV, current);
+ }
-+ printk("\n%sCall Trace:\n", log_lvl);
-+ show_trace_log_lvl(task, regs, esp, log_lvl);
+}
+
-+void show_stack(struct task_struct *task, unsigned long *esp)
++#ifndef CONFIG_XEN
++fastcall void do_spurious_interrupt_bug(struct pt_regs * regs,
++ long error_code)
+{
-+ printk(" ");
-+ show_stack_log_lvl(task, NULL, esp, "");
++#if 0
++ /* No need to warn about this any longer. */
++ printk("Ignoring P6 Local APIC Spurious Interrupt Bug...\n");
++#endif
++}
++
++fastcall void setup_x86_bogus_stack(unsigned char * stk)
++{
++ unsigned long *switch16_ptr, *switch32_ptr;
++ struct pt_regs *regs;
++ unsigned long stack_top, stack_bot;
++ unsigned short iret_frame16_off;
++ int cpu = smp_processor_id();
++ /* reserve the space on 32bit stack for the magic switch16 pointer */
++ memmove(stk, stk + 8, sizeof(struct pt_regs));
++ switch16_ptr = (unsigned long *)(stk + sizeof(struct pt_regs));
++ regs = (struct pt_regs *)stk;
++ /* now the switch32 on 16bit stack */
++ stack_bot = (unsigned long)&per_cpu(cpu_16bit_stack, cpu);
++ stack_top = stack_bot + CPU_16BIT_STACK_SIZE;
++ switch32_ptr = (unsigned long *)(stack_top - 8);
++ iret_frame16_off = CPU_16BIT_STACK_SIZE - 8 - 20;
++ /* copy iret frame on 16bit stack */
++ memcpy((void *)(stack_bot + iret_frame16_off), ®s->eip, 20);
++ /* fill in the switch pointers */
++ switch16_ptr[0] = (regs->esp & 0xffff0000) | iret_frame16_off;
++ switch16_ptr[1] = __ESPFIX_SS;
++ switch32_ptr[0] = (unsigned long)stk + sizeof(struct pt_regs) +
++ 8 - CPU_16BIT_STACK_SIZE;
++ switch32_ptr[1] = __KERNEL_DS;
++}
++
++fastcall unsigned char * fixup_x86_bogus_stack(unsigned short sp)
++{
++ unsigned long *switch32_ptr;
++ unsigned char *stack16, *stack32;
++ unsigned long stack_top, stack_bot;
++ int len;
++ int cpu = smp_processor_id();
++ stack_bot = (unsigned long)&per_cpu(cpu_16bit_stack, cpu);
++ stack_top = stack_bot + CPU_16BIT_STACK_SIZE;
++ switch32_ptr = (unsigned long *)(stack_top - 8);
++ /* copy the data from 16bit stack to 32bit stack */
++ len = CPU_16BIT_STACK_SIZE - 8 - sp;
++ stack16 = (unsigned char *)(stack_bot + sp);
++ stack32 = (unsigned char *)
++ (switch32_ptr[0] + CPU_16BIT_STACK_SIZE - 8 - len);
++ memcpy(stack32, stack16, len);
++ return stack32;
+}
++#endif
+
+/*
-+ * The architecture-independent dump_stack generator
++ * 'math_state_restore()' saves the current math information in the
++ * old math state array, and gets the new ones from the current task
++ *
++ * Careful.. There are problems with IBM-designed IRQ13 behaviour.
++ * Don't touch unless you *really* know how it works.
++ *
++ * Must be called with kernel preemption disabled (in this case,
++ * local interrupts are disabled at the call-site in entry.S).
+ */
-+void dump_stack(void)
++asmlinkage void math_state_restore(struct pt_regs regs)
+{
-+ unsigned long stack;
++ struct thread_info *thread = current_thread_info();
++ struct task_struct *tsk = thread->task;
+
-+ show_trace(current, NULL, &stack);
++ /* NB. 'clts' is done for us by Xen during virtual trap. */
++ if (!tsk_used_math(tsk))
++ init_fpu(tsk);
++ restore_fpu(tsk);
++ thread->status |= TS_USEDFPU; /* So we fnsave on switch_to() */
+}
+
-+EXPORT_SYMBOL(dump_stack);
++#ifndef CONFIG_MATH_EMULATION
+
-+void show_registers(struct pt_regs *regs)
++asmlinkage void math_emulate(long arg)
+{
-+ int i;
-+ int in_kernel = 1;
-+ unsigned long esp;
-+ unsigned short ss;
++ printk(KERN_EMERG "math-emulation not enabled and no coprocessor found.\n");
++ printk(KERN_EMERG "killing %s.\n",current->comm);
++ force_sig(SIGFPE,current);
++ schedule();
++}
+
-+ esp = (unsigned long) (®s->esp);
-+ savesegment(ss, ss);
-+ if (user_mode_vm(regs)) {
-+ in_kernel = 0;
-+ esp = regs->esp;
-+ ss = regs->xss & 0xffff;
-+ }
-+ print_modules();
-+ printk(KERN_EMERG "CPU: %d\nEIP: %04x:[<%08lx>] %s VLI\n"
-+ "EFLAGS: %08lx (%s %.*s) \n",
-+ smp_processor_id(), 0xffff & regs->xcs, regs->eip,
-+ print_tainted(), regs->eflags, system_utsname.release,
-+ (int)strcspn(system_utsname.version, " "),
-+ system_utsname.version);
-+ print_symbol(KERN_EMERG "EIP is at %s\n", regs->eip);
-+ printk(KERN_EMERG "eax: %08lx ebx: %08lx ecx: %08lx edx: %08lx\n",
-+ regs->eax, regs->ebx, regs->ecx, regs->edx);
-+ printk(KERN_EMERG "esi: %08lx edi: %08lx ebp: %08lx esp: %08lx\n",
-+ regs->esi, regs->edi, regs->ebp, esp);
-+ printk(KERN_EMERG "ds: %04x es: %04x ss: %04x\n",
-+ regs->xds & 0xffff, regs->xes & 0xffff, ss);
-+ printk(KERN_EMERG "Process %.*s (pid: %d, ti=%p task=%p task.ti=%p)",
-+ TASK_COMM_LEN, current->comm, current->pid,
-+ current_thread_info(), current, current->thread_info);
-+ /*
-+ * When in-kernel, we also print out the stack and code at the
-+ * time of the fault..
-+ */
-+ if (in_kernel) {
-+ u8 __user *eip;
++#endif /* CONFIG_MATH_EMULATION */
+
-+ printk("\n" KERN_EMERG "Stack: ");
-+ show_stack_log_lvl(NULL, regs, (unsigned long *)esp, KERN_EMERG);
++#ifdef CONFIG_X86_F00F_BUG
++void __init trap_init_f00f_bug(void)
++{
++ __set_fixmap(FIX_F00F_IDT, __pa(&idt_table), PAGE_KERNEL_RO);
+
-+ printk(KERN_EMERG "Code: ");
++ /*
++ * Update the IDT descriptor and reload the IDT so that
++ * it uses the read-only mapped virtual address.
++ */
++ idt_descr.address = fix_to_virt(FIX_F00F_IDT);
++ load_idt(&idt_descr);
++}
++#endif
+
-+ eip = (u8 __user *)regs->eip - 43;
-+ for (i = 0; i < 64; i++, eip++) {
-+ unsigned char c;
+
-+ if (eip < (u8 __user *)PAGE_OFFSET || __get_user(c, eip)) {
-+ printk(" Bad EIP value.");
-+ break;
-+ }
-+ if (eip == (u8 __user *)regs->eip)
-+ printk("<%02x> ", c);
-+ else
-+ printk("%02x ", c);
-+ }
-+ }
-+ printk("\n");
-+}
++/*
++ * NB. All these are "trap gates" (i.e. events_mask isn't set) except
++ * for those that specify <dpl>|4 in the second field.
++ */
++static trap_info_t __cpuinitdata trap_table[] = {
++ { 0, 0, __KERNEL_CS, (unsigned long)divide_error },
++ { 1, 0|4, __KERNEL_CS, (unsigned long)debug },
++ { 3, 3|4, __KERNEL_CS, (unsigned long)int3 },
++ { 4, 3, __KERNEL_CS, (unsigned long)overflow },
++ { 5, 0, __KERNEL_CS, (unsigned long)bounds },
++ { 6, 0, __KERNEL_CS, (unsigned long)invalid_op },
++ { 7, 0|4, __KERNEL_CS, (unsigned long)device_not_available },
++ { 9, 0, __KERNEL_CS, (unsigned long)coprocessor_segment_overrun },
++ { 10, 0, __KERNEL_CS, (unsigned long)invalid_TSS },
++ { 11, 0, __KERNEL_CS, (unsigned long)segment_not_present },
++ { 12, 0, __KERNEL_CS, (unsigned long)stack_segment },
++ { 13, 0, __KERNEL_CS, (unsigned long)general_protection },
++ { 14, 0|4, __KERNEL_CS, (unsigned long)page_fault },
++ { 15, 0, __KERNEL_CS, (unsigned long)fixup_4gb_segment },
++ { 16, 0, __KERNEL_CS, (unsigned long)coprocessor_error },
++ { 17, 0, __KERNEL_CS, (unsigned long)alignment_check },
++#ifdef CONFIG_X86_MCE
++ { 18, 0, __KERNEL_CS, (unsigned long)machine_check },
++#endif
++ { 19, 0, __KERNEL_CS, (unsigned long)simd_coprocessor_error },
++ { SYSCALL_VECTOR, 3, __KERNEL_CS, (unsigned long)system_call },
++ { 0, 0, 0, 0 }
++};
+
-+static void handle_BUG(struct pt_regs *regs)
++void __init trap_init(void)
+{
-+ unsigned long eip = regs->eip;
-+ unsigned short ud2;
-+
-+ if (eip < PAGE_OFFSET)
-+ return;
-+ if (__get_user(ud2, (unsigned short __user *)eip))
-+ return;
-+ if (ud2 != 0x0b0f)
-+ return;
++ int ret;
+
-+ printk(KERN_EMERG "------------[ cut here ]------------\n");
++ ret = HYPERVISOR_set_trap_table(trap_table);
++ if (ret)
++ printk("HYPERVISOR_set_trap_table failed: error %d\n", ret);
+
-+#ifdef CONFIG_DEBUG_BUGVERBOSE
-+ do {
-+ unsigned short line;
-+ char *file;
-+ char c;
++ if (cpu_has_fxsr) {
++ /*
++ * Verify that the FXSAVE/FXRSTOR data will be 16-byte aligned.
++ * Generates a compile-time "error: zero width for bit-field" if
++ * the alignment is wrong.
++ */
++ struct fxsrAlignAssert {
++ int _:!(offsetof(struct task_struct,
++ thread.i387.fxsave) & 15);
++ };
+
-+ if (__get_user(line, (unsigned short __user *)(eip + 2)))
-+ break;
-+ if (__get_user(file, (char * __user *)(eip + 4)) ||
-+ (unsigned long)file < PAGE_OFFSET || __get_user(c, file))
-+ file = "<bad filename>";
++ printk(KERN_INFO "Enabling fast FPU save and restore... ");
++ set_in_cr4(X86_CR4_OSFXSR);
++ printk("done.\n");
++ }
++ if (cpu_has_xmm) {
++ printk(KERN_INFO "Enabling unmasked SIMD FPU exception "
++ "support... ");
++ set_in_cr4(X86_CR4_OSXMMEXCPT);
++ printk("done.\n");
++ }
+
-+ printk(KERN_EMERG "kernel BUG at %s:%d!\n", file, line);
-+ return;
-+ } while (0);
-+#endif
-+ printk(KERN_EMERG "Kernel BUG at [verbose debug info unavailable]\n");
++ /*
++ * Should be a barrier for any external CPU state.
++ */
++ cpu_init();
+}
+
-+/* This is gone through when something in the kernel
-+ * has done something bad and is about to be terminated.
-+*/
-+void die(const char * str, struct pt_regs * regs, long err)
++void __cpuinit smp_trap_init(trap_info_t *trap_ctxt)
+{
-+ static struct {
-+ spinlock_t lock;
-+ u32 lock_owner;
-+ int lock_owner_depth;
-+ } die = {
-+ .lock = SPIN_LOCK_UNLOCKED,
-+ .lock_owner = -1,
-+ .lock_owner_depth = 0
-+ };
-+ static int die_counter;
-+ unsigned long flags;
-+
-+ oops_enter();
++ const trap_info_t *t = trap_table;
+
-+ if (die.lock_owner != raw_smp_processor_id()) {
-+ console_verbose();
-+ spin_lock_irqsave(&die.lock, flags);
-+ die.lock_owner = smp_processor_id();
-+ die.lock_owner_depth = 0;
-+ bust_spinlocks(1);
++ for (t = trap_table; t->address; t++) {
++ trap_ctxt[t->vector].flags = t->flags;
++ trap_ctxt[t->vector].cs = t->cs;
++ trap_ctxt[t->vector].address = t->address;
+ }
-+ else
-+ local_save_flags(flags);
++}
+
-+ if (++die.lock_owner_depth < 3) {
-+ int nl = 0;
-+ unsigned long esp;
-+ unsigned short ss;
++static int __init kstack_setup(char *s)
++{
++ kstack_depth_to_print = simple_strtoul(s, NULL, 0);
++ return 1;
++}
++__setup("kstack=", kstack_setup);
+
-+ handle_BUG(regs);
-+ printk(KERN_EMERG "%s: %04lx [#%d]\n", str, err & 0xffff, ++die_counter);
-+#ifdef CONFIG_PREEMPT
-+ printk(KERN_EMERG "PREEMPT ");
-+ nl = 1;
++#ifdef CONFIG_STACK_UNWIND
++static int __init call_trace_setup(char *s)
++{
++ if (strcmp(s, "old") == 0)
++ call_trace = -1;
++ else if (strcmp(s, "both") == 0)
++ call_trace = 0;
++ else if (strcmp(s, "newfallback") == 0)
++ call_trace = 1;
++ else if (strcmp(s, "new") == 2)
++ call_trace = 2;
++ return 1;
++}
++__setup("call_trace=", call_trace_setup);
+#endif
-+#ifdef CONFIG_SMP
-+ if (!nl)
-+ printk(KERN_EMERG);
-+ printk("SMP ");
-+ nl = 1;
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/traps.c
+--- a/arch/i386/kernel/traps.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/traps.c Wed Sep 10 10:54:08 2008 +0100
+@@ -644,18 +644,11 @@
+
+ static void io_check_error(unsigned char reason, struct pt_regs * regs)
+ {
+- unsigned long i;
+-
+ printk(KERN_EMERG "NMI: IOCK error (debug interrupt?)\n");
+ show_registers(regs);
+
+ /* Re-enable the IOCK line, wait for a few seconds */
+- reason = (reason & 0xf) | 8;
+- outb(reason, 0x61);
+- i = 2000;
+- while (--i) udelay(1000);
+- reason &= ~8;
+- outb(reason, 0x61);
++ clear_io_check_error(reason);
+ }
+
+ static void unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/vm86.c
+--- a/arch/i386/kernel/vm86.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/vm86.c Wed Sep 10 10:54:08 2008 +0100
+@@ -97,7 +97,9 @@
+ struct pt_regs * FASTCALL(save_v86_state(struct kernel_vm86_regs * regs));
+ struct pt_regs * fastcall save_v86_state(struct kernel_vm86_regs * regs)
+ {
++#ifndef CONFIG_X86_NO_TSS
+ struct tss_struct *tss;
+#endif
-+#ifdef CONFIG_DEBUG_PAGEALLOC
-+ if (!nl)
-+ printk(KERN_EMERG);
-+ printk("DEBUG_PAGEALLOC");
-+ nl = 1;
+ struct pt_regs *ret;
+ unsigned long tmp;
+
+@@ -122,12 +124,16 @@
+ do_exit(SIGSEGV);
+ }
+
++#ifndef CONFIG_X86_NO_TSS
+ tss = &per_cpu(init_tss, get_cpu());
+#endif
-+ if (nl)
-+ printk("\n");
-+ if (notify_die(DIE_OOPS, str, regs, err,
-+ current->thread.trap_no, SIGSEGV) !=
-+ NOTIFY_STOP) {
-+ show_registers(regs);
-+ /* Executive summary in case the oops scrolled away */
-+ esp = (unsigned long) (®s->esp);
-+ savesegment(ss, ss);
-+ if (user_mode(regs)) {
-+ esp = regs->esp;
-+ ss = regs->xss & 0xffff;
-+ }
-+ printk(KERN_EMERG "EIP: [<%08lx>] ", regs->eip);
-+ print_symbol("%s", regs->eip);
-+ printk(" SS:ESP %04x:%08lx\n", ss, esp);
-+ }
-+ else
-+ regs = NULL;
-+ } else
-+ printk(KERN_EMERG "Recursive die() failure, output suppressed\n");
+ current->thread.esp0 = current->thread.saved_esp0;
+ current->thread.sysenter_cs = __KERNEL_CS;
+ load_esp0(tss, ¤t->thread);
+ current->thread.saved_esp0 = 0;
++#ifndef CONFIG_X86_NO_TSS
+ put_cpu();
++#endif
+
+ loadsegment(fs, current->thread.saved_fs);
+ loadsegment(gs, current->thread.saved_gs);
+@@ -251,7 +257,9 @@
+
+ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk)
+ {
++#ifndef CONFIG_X86_NO_TSS
+ struct tss_struct *tss;
++#endif
+ long eax;
+ /*
+ * make sure the vm86() system call doesn't try to do anything silly
+@@ -296,12 +304,16 @@
+ savesegment(fs, tsk->thread.saved_fs);
+ savesegment(gs, tsk->thread.saved_gs);
+
++#ifndef CONFIG_X86_NO_TSS
+ tss = &per_cpu(init_tss, get_cpu());
++#endif
+ tsk->thread.esp0 = (unsigned long) &info->VM86_TSS_ESP0;
+ if (cpu_has_sep)
+ tsk->thread.sysenter_cs = 0;
+ load_esp0(tss, &tsk->thread);
++#ifndef CONFIG_X86_NO_TSS
+ put_cpu();
++#endif
+
+ tsk->thread.screen_bitmap = info->screen_bitmap;
+ if (info->flags & VM86_SCREEN_BITMAP)
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/vmlinux.lds.S
+--- a/arch/i386/kernel/vmlinux.lds.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/kernel/vmlinux.lds.S Wed Sep 10 10:54:08 2008 +0100
+@@ -13,6 +13,12 @@
+ OUTPUT_ARCH(i386)
+ ENTRY(phys_startup_32)
+ jiffies = jiffies_64;
+
-+ bust_spinlocks(0);
-+ die.lock_owner = -1;
-+ spin_unlock_irqrestore(&die.lock, flags);
++PHDRS {
++ text PT_LOAD FLAGS(5); /* R_E */
++ data PT_LOAD FLAGS(7); /* RWE */
++ note PT_NOTE FLAGS(4); /* R__ */
++}
+ SECTIONS
+ {
+ . = __KERNEL_START;
+@@ -26,7 +32,7 @@
+ KPROBES_TEXT
+ *(.fixup)
+ *(.gnu.warning)
+- } = 0x9090
++ } :text = 0x9090
+
+ _etext = .; /* End of text section */
+
+@@ -45,10 +51,11 @@
+ __tracedata_end = .;
+
+ /* writeable */
++ . = ALIGN(4096);
+ .data : AT(ADDR(.data) - LOAD_OFFSET) { /* Data */
+ *(.data)
+ CONSTRUCTORS
+- }
++ } :data
+
+ . = ALIGN(4096);
+ __nosave_begin = .;
+@@ -184,4 +191,6 @@
+ STABS_DEBUG
+
+ DWARF_DEBUG
+
-+ if (!regs)
-+ return;
++ NOTES
+ }
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/kernel/vsyscall-note-xen.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/kernel/vsyscall-note-xen.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,32 @@
++/*
++ * This supplies .note.* sections to go into the PT_NOTE inside the vDSO text.
++ * Here we can supply some information useful to userland.
++ * First we get the vanilla i386 note that supplies the kernel version info.
++ */
+
-+ if (kexec_should_crash(current))
-+ crash_kexec(regs);
++#include "vsyscall-note.S"
+
-+ if (in_interrupt())
-+ panic("Fatal exception in interrupt");
++/*
++ * Now we add a special note telling glibc's dynamic linker a fake hardware
++ * flavor that it will use to choose the search path for libraries in the
++ * same way it uses real hardware capabilities like "mmx".
++ * We supply "nosegneg" as the fake capability, to indicate that we
++ * do not like negative offsets in instructions using segment overrides,
++ * since we implement those inefficiently. This makes it possible to
++ * install libraries optimized to avoid those access patterns in someplace
++ * like /lib/i686/tls/nosegneg. Note that an /etc/ld.so.conf.d/file
++ * corresponding to the bits here is needed to make ldconfig work right.
++ * It should contain:
++ * hwcap 0 nosegneg
++ * to match the mapping of bit to name that we give here.
++ */
++#define NOTE_KERNELCAP_BEGIN(ncaps, mask) \
++ ASM_ELF_NOTE_BEGIN(".note.kernelcap", "a", "GNU", 2) \
++ .long ncaps, mask
++#define NOTE_KERNELCAP(bit, name) \
++ .byte bit; .asciz name
++#define NOTE_KERNELCAP_END ASM_ELF_NOTE_END
+
-+ if (panic_on_oops)
-+ panic("Fatal exception");
++NOTE_KERNELCAP_BEGIN(1, 1)
++NOTE_KERNELCAP(0, "nosegneg")
++NOTE_KERNELCAP_END
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/lib/Makefile
+--- a/arch/i386/lib/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/lib/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -7,3 +7,4 @@
+ bitops.o
+
+ lib-$(CONFIG_X86_USE_3DNOW) += mmx.o
++lib-$(CONFIG_XEN_SCRUB_PAGES) += scrub.o
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/lib/scrub.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/lib/scrub.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,21 @@
++#include <asm/cpufeature.h>
++#include <asm/page.h>
++#include <asm/processor.h>
+
-+ oops_exit();
-+ do_exit(SIGSEGV);
++void scrub_pages(void *v, unsigned int count)
++{
++ if (likely(cpu_has_xmm2)) {
++ unsigned long n = count * (PAGE_SIZE / sizeof(long) / 4);
++
++ for (; n--; v += sizeof(long) * 4)
++ asm("movnti %1,(%0)\n\t"
++ "movnti %1,%c2(%0)\n\t"
++ "movnti %1,2*%c2(%0)\n\t"
++ "movnti %1,3*%c2(%0)\n\t"
++ : : "r" (v), "r" (0L), "i" (sizeof(long))
++ : "memory");
++ asm volatile("sfence" : : : "memory");
++ } else
++ for (; count--; v += PAGE_SIZE)
++ clear_page(v);
+}
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/mach-xen/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/mach-xen/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,5 @@
++#
++# Makefile for the linux kernel.
++#
+
-+static inline void die_if_kernel(const char * str, struct pt_regs * regs, long err)
++obj-y := setup.o
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/mach-xen/setup.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/mach-xen/setup.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,158 @@
++/*
++ * Machine specific setup for generic
++ */
++
++#include <linux/mm.h>
++#include <linux/smp.h>
++#include <linux/init.h>
++#include <linux/interrupt.h>
++#include <linux/module.h>
++#include <asm/acpi.h>
++#include <asm/arch_hooks.h>
++#include <asm/e820.h>
++#include <asm/setup.h>
++#include <asm/fixmap.h>
++
++#include <xen/interface/callback.h>
++#include <xen/interface/memory.h>
++
++#ifdef CONFIG_HOTPLUG_CPU
++#define DEFAULT_SEND_IPI (1)
++#else
++#define DEFAULT_SEND_IPI (0)
++#endif
++
++int no_broadcast=DEFAULT_SEND_IPI;
++
++static __init int no_ipi_broadcast(char *str)
+{
-+ if (!user_mode_vm(regs))
-+ die(str, regs, err);
++ get_option(&str, &no_broadcast);
++ printk ("Using %s mode\n", no_broadcast ? "No IPI Broadcast" :
++ "IPI Broadcast");
++ return 1;
+}
+
-+static void __kprobes do_trap(int trapnr, int signr, char *str, int vm86,
-+ struct pt_regs * regs, long error_code,
-+ siginfo_t *info)
++__setup("no_ipi_broadcast", no_ipi_broadcast);
++
++static int __init print_ipi_mode(void)
+{
-+ struct task_struct *tsk = current;
-+ tsk->thread.error_code = error_code;
-+ tsk->thread.trap_no = trapnr;
++ printk ("Using IPI %s mode\n", no_broadcast ? "No-Shortcut" :
++ "Shortcut");
++ return 0;
++}
+
-+ if (regs->eflags & VM_MASK) {
-+ if (vm86)
-+ goto vm86_trap;
-+ goto trap_signal;
-+ }
++late_initcall(print_ipi_mode);
+
-+ if (!user_mode(regs))
-+ goto kernel_trap;
++/**
++ * machine_specific_memory_setup - Hook for machine specific memory setup.
++ *
++ * Description:
++ * This is included late in kernel/setup.c so that it can make
++ * use of all of the static functions.
++ **/
+
-+ trap_signal: {
-+ if (info)
-+ force_sig_info(signr, info, tsk);
-+ else
-+ force_sig(signr, tsk);
-+ return;
-+ }
++char * __init machine_specific_memory_setup(void)
++{
++ int rc;
++ struct xen_memory_map memmap;
++ /*
++ * This is rather large for a stack variable but this early in
++ * the boot process we know we have plenty slack space.
++ */
++ struct e820entry map[E820MAX];
+
-+ kernel_trap: {
-+ if (!fixup_exception(regs))
-+ die(str, regs, error_code);
-+ return;
-+ }
++ memmap.nr_entries = E820MAX;
++ set_xen_guest_handle(memmap.buffer, map);
+
-+ vm86_trap: {
-+ int ret = handle_vm86_trap((struct kernel_vm86_regs *) regs, error_code, trapnr);
-+ if (ret) goto trap_signal;
-+ return;
++ rc = HYPERVISOR_memory_op(XENMEM_memory_map, &memmap);
++ if ( rc == -ENOSYS ) {
++ memmap.nr_entries = 1;
++ map[0].addr = 0ULL;
++ map[0].size = PFN_PHYS((unsigned long long)xen_start_info->nr_pages);
++ /* 8MB slack (to balance backend allocations). */
++ map[0].size += 8ULL << 20;
++ map[0].type = E820_RAM;
++ rc = 0;
+ }
-+}
++ BUG_ON(rc);
+
-+#define DO_ERROR(trapnr, signr, str, name) \
-+fastcall void do_##name(struct pt_regs * regs, long error_code) \
-+{ \
-+ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
-+ == NOTIFY_STOP) \
-+ return; \
-+ do_trap(trapnr, signr, str, 0, regs, error_code, NULL); \
-+}
++ sanitize_e820_map(map, (char *)&memmap.nr_entries);
+
-+#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \
-+fastcall void do_##name(struct pt_regs * regs, long error_code) \
-+{ \
-+ siginfo_t info; \
-+ info.si_signo = signr; \
-+ info.si_errno = 0; \
-+ info.si_code = sicode; \
-+ info.si_addr = (void __user *)siaddr; \
-+ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
-+ == NOTIFY_STOP) \
-+ return; \
-+ do_trap(trapnr, signr, str, 0, regs, error_code, &info); \
-+}
++ BUG_ON(copy_e820_map(map, (char)memmap.nr_entries) < 0);
+
-+#define DO_VM86_ERROR(trapnr, signr, str, name) \
-+fastcall void do_##name(struct pt_regs * regs, long error_code) \
-+{ \
-+ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
-+ == NOTIFY_STOP) \
-+ return; \
-+ do_trap(trapnr, signr, str, 1, regs, error_code, NULL); \
++ return "Xen";
+}
+
-+#define DO_VM86_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \
-+fastcall void do_##name(struct pt_regs * regs, long error_code) \
-+{ \
-+ siginfo_t info; \
-+ info.si_signo = signr; \
-+ info.si_errno = 0; \
-+ info.si_code = sicode; \
-+ info.si_addr = (void __user *)siaddr; \
-+ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
-+ == NOTIFY_STOP) \
-+ return; \
-+ do_trap(trapnr, signr, str, 1, regs, error_code, &info); \
-+}
+
-+DO_VM86_ERROR_INFO( 0, SIGFPE, "divide error", divide_error, FPE_INTDIV, regs->eip)
-+#ifndef CONFIG_KPROBES
-+DO_VM86_ERROR( 3, SIGTRAP, "int3", int3)
-+#endif
-+DO_VM86_ERROR( 4, SIGSEGV, "overflow", overflow)
-+DO_VM86_ERROR( 5, SIGSEGV, "bounds", bounds)
-+DO_ERROR_INFO( 6, SIGILL, "invalid opcode", invalid_op, ILL_ILLOPN, regs->eip)
-+DO_ERROR( 9, SIGFPE, "coprocessor segment overrun", coprocessor_segment_overrun)
-+DO_ERROR(10, SIGSEGV, "invalid TSS", invalid_TSS)
-+DO_ERROR(11, SIGBUS, "segment not present", segment_not_present)
-+DO_ERROR(12, SIGBUS, "stack segment", stack_segment)
-+DO_ERROR_INFO(17, SIGBUS, "alignment check", alignment_check, BUS_ADRALN, 0)
-+DO_ERROR_INFO(32, SIGSEGV, "iret exception", iret_error, ILL_BADSTK, 0)
++extern void hypervisor_callback(void);
++extern void failsafe_callback(void);
++extern void nmi(void);
+
-+fastcall void __kprobes do_general_protection(struct pt_regs * regs,
-+ long error_code)
++unsigned long *machine_to_phys_mapping = (void *)MACH2PHYS_VIRT_START;
++EXPORT_SYMBOL(machine_to_phys_mapping);
++unsigned int machine_to_phys_order;
++EXPORT_SYMBOL(machine_to_phys_order);
++
++void __init pre_setup_arch_hook(void)
+{
-+ current->thread.error_code = error_code;
-+ current->thread.trap_no = 13;
++ struct xen_machphys_mapping mapping;
++ unsigned long machine_to_phys_nr_ents;
++ struct xen_platform_parameters pp;
+
-+ if (regs->eflags & VM_MASK)
-+ goto gp_in_vm86;
++ init_mm.pgd = swapper_pg_dir = (pgd_t *)xen_start_info->pt_base;
+
-+ if (!user_mode(regs))
-+ goto gp_in_kernel;
++ setup_xen_features();
+
-+ current->thread.error_code = error_code;
-+ current->thread.trap_no = 13;
-+ force_sig(SIGSEGV, current);
-+ return;
++ if (HYPERVISOR_xen_version(XENVER_platform_parameters, &pp) == 0)
++ set_fixaddr_top(pp.virt_start);
+
-+gp_in_vm86:
-+ local_irq_enable();
-+ handle_vm86_fault((struct kernel_vm86_regs *) regs, error_code);
-+ return;
++ if (HYPERVISOR_memory_op(XENMEM_machphys_mapping, &mapping) == 0) {
++ machine_to_phys_mapping = (unsigned long *)mapping.v_start;
++ machine_to_phys_nr_ents = mapping.max_mfn + 1;
++ } else
++ machine_to_phys_nr_ents = MACH2PHYS_NR_ENTRIES;
++ machine_to_phys_order = fls(machine_to_phys_nr_ents - 1);
+
-+gp_in_kernel:
-+ if (!fixup_exception(regs)) {
-+ if (notify_die(DIE_GPF, "general protection fault", regs,
-+ error_code, 13, SIGSEGV) == NOTIFY_STOP)
-+ return;
-+ die("general protection fault", regs, error_code);
-+ }
++ if (!xen_feature(XENFEAT_auto_translated_physmap))
++ phys_to_machine_mapping =
++ (unsigned long *)xen_start_info->mfn_list;
+}
+
-+static void mem_parity_error(unsigned char reason, struct pt_regs * regs)
++void __init machine_specific_arch_setup(void)
+{
-+ printk(KERN_EMERG "Uhhuh. NMI received. Dazed and confused, but trying "
-+ "to continue\n");
-+ printk(KERN_EMERG "You probably have a hardware problem with your RAM "
-+ "chips\n");
-+
-+ /* Clear and disable the memory parity error line. */
-+ clear_mem_error(reason);
-+}
++ int ret;
++ static struct callback_register __initdata event = {
++ .type = CALLBACKTYPE_event,
++ .address = { __KERNEL_CS, (unsigned long)hypervisor_callback },
++ };
++ static struct callback_register __initdata failsafe = {
++ .type = CALLBACKTYPE_failsafe,
++ .address = { __KERNEL_CS, (unsigned long)failsafe_callback },
++ };
++ static struct callback_register __initdata nmi_cb = {
++ .type = CALLBACKTYPE_nmi,
++ .address = { __KERNEL_CS, (unsigned long)nmi },
++ };
+
-+static void io_check_error(unsigned char reason, struct pt_regs * regs)
-+{
-+ printk(KERN_EMERG "NMI: IOCK error (debug interrupt?)\n");
-+ show_registers(regs);
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &event);
++ if (ret == 0)
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &failsafe);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (ret == -ENOSYS)
++ ret = HYPERVISOR_set_callbacks(
++ event.address.cs, event.address.eip,
++ failsafe.address.cs, failsafe.address.eip);
++#endif
++ BUG_ON(ret);
+
-+ /* Re-enable the IOCK line, wait for a few seconds */
-+ clear_io_check_error(reason);
-+}
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &nmi_cb);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (ret == -ENOSYS) {
++ static struct xennmi_callback __initdata cb = {
++ .handler_address = (unsigned long)nmi
++ };
+
-+static void unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
-+{
-+#ifdef CONFIG_MCA
-+ /* Might actually be able to figure out what the guilty party
-+ * is. */
-+ if( MCA_bus ) {
-+ mca_handle_nmi();
-+ return;
++ HYPERVISOR_nmi_op(XENNMI_register_callback, &cb);
+ }
+#endif
-+ printk("Uhhuh. NMI received for unknown reason %02x on CPU %d.\n",
-+ reason, smp_processor_id());
-+ printk("Dazed and confused, but trying to continue\n");
-+ printk("Do you have a strange power saving mode enabled?\n");
+}
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/mm/Makefile
+--- a/arch/i386/mm/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/mm/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -8,3 +8,4 @@
+ obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o
+ obj-$(CONFIG_HIGHMEM) += highmem.o
+ obj-$(CONFIG_BOOT_IOREMAP) += boot_ioremap.o
++obj-$(CONFIG_XEN) += hypervisor.o
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/mm/fault-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/mm/fault-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,779 @@
++/*
++ * linux/arch/i386/mm/fault.c
++ *
++ * Copyright (C) 1995 Linus Torvalds
++ */
+
-+static DEFINE_SPINLOCK(nmi_print_lock);
++#include <linux/signal.h>
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/string.h>
++#include <linux/types.h>
++#include <linux/ptrace.h>
++#include <linux/mman.h>
++#include <linux/mm.h>
++#include <linux/smp.h>
++#include <linux/smp_lock.h>
++#include <linux/interrupt.h>
++#include <linux/init.h>
++#include <linux/tty.h>
++#include <linux/vt_kern.h> /* For unblank_screen() */
++#include <linux/highmem.h>
++#include <linux/module.h>
++#include <linux/kprobes.h>
+
-+void die_nmi (struct pt_regs *regs, const char *msg)
++#include <asm/system.h>
++#include <asm/uaccess.h>
++#include <asm/desc.h>
++#include <asm/kdebug.h>
++
++extern void die(const char *,struct pt_regs *,long);
++
++#ifdef CONFIG_KPROBES
++ATOMIC_NOTIFIER_HEAD(notify_page_fault_chain);
++int register_page_fault_notifier(struct notifier_block *nb)
+{
-+ if (notify_die(DIE_NMIWATCHDOG, msg, regs, 0, 2, SIGINT) ==
-+ NOTIFY_STOP)
-+ return;
++ vmalloc_sync_all();
++ return atomic_notifier_chain_register(¬ify_page_fault_chain, nb);
++}
+
-+ spin_lock(&nmi_print_lock);
-+ /*
-+ * We are in trouble anyway, lets at least try
-+ * to get a message out.
-+ */
-+ bust_spinlocks(1);
-+ printk(KERN_EMERG "%s", msg);
-+ printk(" on CPU%d, eip %08lx, registers:\n",
-+ smp_processor_id(), regs->eip);
-+ show_registers(regs);
-+ printk(KERN_EMERG "console shuts up ...\n");
-+ console_silent();
-+ spin_unlock(&nmi_print_lock);
-+ bust_spinlocks(0);
-+
-+ /* If we are in kernel we are probably nested up pretty bad
-+ * and might aswell get out now while we still can.
-+ */
-+ if (!user_mode_vm(regs)) {
-+ current->thread.trap_no = 2;
-+ crash_kexec(regs);
-+ }
-+
-+ do_exit(SIGSEGV);
-+}
-+
-+static void default_do_nmi(struct pt_regs * regs)
++int unregister_page_fault_notifier(struct notifier_block *nb)
+{
-+ unsigned char reason = 0;
-+
-+ /* Only the BSP gets external NMIs from the system. */
-+ if (!smp_processor_id())
-+ reason = get_nmi_reason();
-+
-+ if (!(reason & 0xc0)) {
-+ if (notify_die(DIE_NMI_IPI, "nmi_ipi", regs, reason, 2, SIGINT)
-+ == NOTIFY_STOP)
-+ return;
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ /*
-+ * Ok, so this is none of the documented NMI sources,
-+ * so it must be the NMI watchdog.
-+ */
-+ if (nmi_watchdog) {
-+ nmi_watchdog_tick(regs);
-+ return;
-+ }
-+#endif
-+ unknown_nmi_error(reason, regs);
-+ return;
-+ }
-+ if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP)
-+ return;
-+ if (reason & 0x80)
-+ mem_parity_error(reason, regs);
-+ if (reason & 0x40)
-+ io_check_error(reason, regs);
-+ /*
-+ * Reassert NMI in case it became active meanwhile
-+ * as it's edge-triggered.
-+ */
-+ reassert_nmi();
++ return atomic_notifier_chain_unregister(¬ify_page_fault_chain, nb);
+}
+
-+static int dummy_nmi_callback(struct pt_regs * regs, int cpu)
++static inline int notify_page_fault(enum die_val val, const char *str,
++ struct pt_regs *regs, long err, int trap, int sig)
+{
-+ return 0;
++ struct die_args args = {
++ .regs = regs,
++ .str = str,
++ .err = err,
++ .trapnr = trap,
++ .signr = sig
++ };
++ return atomic_notifier_call_chain(¬ify_page_fault_chain, val, &args);
+}
-+
-+static nmi_callback_t nmi_callback = dummy_nmi_callback;
-+
-+fastcall void do_nmi(struct pt_regs * regs, long error_code)
++#else
++static inline int notify_page_fault(enum die_val val, const char *str,
++ struct pt_regs *regs, long err, int trap, int sig)
+{
-+ int cpu;
-+
-+ nmi_enter();
-+
-+ cpu = smp_processor_id();
-+
-+ ++nmi_count(cpu);
-+
-+ if (!rcu_dereference(nmi_callback)(regs, cpu))
-+ default_do_nmi(regs);
-+
-+ nmi_exit();
++ return NOTIFY_DONE;
+}
++#endif
+
-+void set_nmi_callback(nmi_callback_t callback)
-+{
-+ vmalloc_sync_all();
-+ rcu_assign_pointer(nmi_callback, callback);
-+}
-+EXPORT_SYMBOL_GPL(set_nmi_callback);
+
-+void unset_nmi_callback(void)
++/*
++ * Unlock any spinlocks which will prevent us from getting the
++ * message out
++ */
++void bust_spinlocks(int yes)
+{
-+ nmi_callback = dummy_nmi_callback;
-+}
-+EXPORT_SYMBOL_GPL(unset_nmi_callback);
++ int loglevel_save = console_loglevel;
+
-+#ifdef CONFIG_KPROBES
-+fastcall void __kprobes do_int3(struct pt_regs *regs, long error_code)
-+{
-+ if (notify_die(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP)
-+ == NOTIFY_STOP)
++ if (yes) {
++ oops_in_progress = 1;
+ return;
-+ /* This is an interrupt gate, because kprobes wants interrupts
-+ disabled. Normal trap handlers don't. */
-+ restore_interrupts(regs);
-+ do_trap(3, SIGTRAP, "int3", 1, regs, error_code, NULL);
-+}
++ }
++#ifdef CONFIG_VT
++ unblank_screen();
+#endif
++ oops_in_progress = 0;
++ /*
++ * OK, the message is on the console. Now we call printk()
++ * without oops_in_progress set so that printk will give klogd
++ * a poke. Hold onto your hats...
++ */
++ console_loglevel = 15; /* NMI oopser may have shut the console up */
++ printk(" ");
++ console_loglevel = loglevel_save;
++}
+
+/*
-+ * Our handling of the processor debug registers is non-trivial.
-+ * We do not clear them on entry and exit from the kernel. Therefore
-+ * it is possible to get a watchpoint trap here from inside the kernel.
-+ * However, the code in ./ptrace.c has ensured that the user can
-+ * only set watchpoints on userspace addresses. Therefore the in-kernel
-+ * watchpoint trap can only occur in code which is reading/writing
-+ * from user space. Such code must not hold kernel locks (since it
-+ * can equally take a page fault), therefore it is safe to call
-+ * force_sig_info even though that claims and releases locks.
-+ *
-+ * Code in ./signal.c ensures that the debug control register
-+ * is restored before we deliver any signal, and therefore that
-+ * user code runs with the correct debug control register even though
-+ * we clear it here.
++ * Return EIP plus the CS segment base. The segment limit is also
++ * adjusted, clamped to the kernel/user address space (whichever is
++ * appropriate), and returned in *eip_limit.
+ *
-+ * Being careful here means that we don't have to be as careful in a
-+ * lot of more complicated places (task switching can be a bit lazy
-+ * about restoring all the debug state, and ptrace doesn't have to
-+ * find every occurrence of the TF bit that could be saved away even
-+ * by user code)
++ * The segment is checked, because it might have been changed by another
++ * task between the original faulting instruction and here.
++ *
++ * If CS is no longer a valid code segment, or if EIP is beyond the
++ * limit, or if it is a kernel address when CS is not a kernel segment,
++ * then the returned value will be greater than *eip_limit.
++ *
++ * This is slow, but is very rarely executed.
+ */
-+fastcall void __kprobes do_debug(struct pt_regs * regs, long error_code)
++static inline unsigned long get_segment_eip(struct pt_regs *regs,
++ unsigned long *eip_limit)
+{
-+ unsigned int condition;
-+ struct task_struct *tsk = current;
-+
-+ get_debugreg(condition, 6);
-+
-+ if (notify_die(DIE_DEBUG, "debug", regs, condition, error_code,
-+ SIGTRAP) == NOTIFY_STOP)
-+ return;
-+ /* It's safe to allow irq's after DR6 has been saved */
-+ if (regs->eflags & X86_EFLAGS_IF)
-+ local_irq_enable();
++ unsigned long eip = regs->eip;
++ unsigned seg = regs->xcs & 0xffff;
++ u32 seg_ar, seg_limit, base, *desc;
+
-+ /* Mask out spurious debug traps due to lazy DR7 setting */
-+ if (condition & (DR_TRAP0|DR_TRAP1|DR_TRAP2|DR_TRAP3)) {
-+ if (!tsk->thread.debugreg[7])
-+ goto clear_dr7;
++ /* Unlikely, but must come before segment checks. */
++ if (unlikely(regs->eflags & VM_MASK)) {
++ base = seg << 4;
++ *eip_limit = base + 0xffff;
++ return base + (eip & 0xffff);
+ }
+
-+ if (regs->eflags & VM_MASK)
-+ goto debug_vm86;
-+
-+ /* Save debug status register where ptrace can see it */
-+ tsk->thread.debugreg[6] = condition;
++ /* The standard kernel/user address space limit. */
++ *eip_limit = (seg & 2) ? USER_DS.seg : KERNEL_DS.seg;
++
++ /* By far the most common cases. */
++ if (likely(seg == __USER_CS || seg == GET_KERNEL_CS()))
++ return eip;
+
-+ /*
-+ * Single-stepping through TF: make sure we ignore any events in
-+ * kernel space (but re-enable TF when returning to user mode).
-+ */
-+ if (condition & DR_STEP) {
-+ /*
-+ * We already checked v86 mode above, so we can
-+ * check for kernel mode by just checking the CPL
-+ * of CS.
-+ */
-+ if (!user_mode(regs))
-+ goto clear_TF_reenable;
++ /* Check the segment exists, is within the current LDT/GDT size,
++ that kernel/user (ring 0..3) has the appropriate privilege,
++ that it's a code segment, and get the limit. */
++ __asm__ ("larl %3,%0; lsll %3,%1"
++ : "=&r" (seg_ar), "=r" (seg_limit) : "0" (0), "rm" (seg));
++ if ((~seg_ar & 0x9800) || eip > seg_limit) {
++ *eip_limit = 0;
++ return 1; /* So that returned eip > *eip_limit. */
+ }
+
-+ /* Ok, finally something we can handle */
-+ send_sigtrap(tsk, regs, error_code);
++ /* Get the GDT/LDT descriptor base.
++ When you look for races in this code remember that
++ LDT and other horrors are only used in user space. */
++ if (seg & (1<<2)) {
++ /* Must lock the LDT while reading it. */
++ down(¤t->mm->context.sem);
++ desc = current->mm->context.ldt;
++ desc = (void *)desc + (seg & ~7);
++ } else {
++ /* Must disable preemption while reading the GDT. */
++ desc = (u32 *)get_cpu_gdt_table(get_cpu());
++ desc = (void *)desc + (seg & ~7);
++ }
+
-+ /* Disable additional traps. They'll be re-enabled when
-+ * the signal is delivered.
-+ */
-+clear_dr7:
-+ set_debugreg(0, 7);
-+ return;
++ /* Decode the code segment base from the descriptor */
++ base = get_desc_base((unsigned long *)desc);
+
-+debug_vm86:
-+ handle_vm86_trap((struct kernel_vm86_regs *) regs, error_code, 1);
-+ return;
++ if (seg & (1<<2)) {
++ up(¤t->mm->context.sem);
++ } else
++ put_cpu();
+
-+clear_TF_reenable:
-+ set_tsk_thread_flag(tsk, TIF_SINGLESTEP);
-+ regs->eflags &= ~TF_MASK;
-+ return;
++ /* Adjust EIP and segment limit, and clamp at the kernel limit.
++ It's legitimate for segments to wrap at 0xffffffff. */
++ seg_limit += base;
++ if (seg_limit < *eip_limit && seg_limit >= base)
++ *eip_limit = seg_limit;
++ return eip + base;
+}
+
-+/*
-+ * Note that we play around with the 'TS' bit in an attempt to get
-+ * the correct behaviour even in the presence of the asynchronous
-+ * IRQ13 behaviour
++/*
++ * Sometimes AMD Athlon/Opteron CPUs report invalid exceptions on prefetch.
++ * Check that here and ignore it.
+ */
-+void math_error(void __user *eip)
-+{
-+ struct task_struct * task;
-+ siginfo_t info;
-+ unsigned short cwd, swd;
++static int __is_prefetch(struct pt_regs *regs, unsigned long addr)
++{
++ unsigned long limit;
++ unsigned long instr = get_segment_eip (regs, &limit);
++ int scan_more = 1;
++ int prefetch = 0;
++ int i;
+
-+ /*
-+ * Save the info for the exception handler and clear the error.
-+ */
-+ task = current;
-+ save_init_fpu(task);
-+ task->thread.trap_no = 16;
-+ task->thread.error_code = 0;
-+ info.si_signo = SIGFPE;
-+ info.si_errno = 0;
-+ info.si_code = __SI_FAULT;
-+ info.si_addr = eip;
-+ /*
-+ * (~cwd & swd) will mask out exceptions that are not set to unmasked
-+ * status. 0x3f is the exception bits in these regs, 0x200 is the
-+ * C1 reg you need in case of a stack fault, 0x040 is the stack
-+ * fault bit. We should only be taking one exception at a time,
-+ * so if this combination doesn't produce any single exception,
-+ * then we have a bad program that isn't syncronizing its FPU usage
-+ * and it will suffer the consequences since we won't be able to
-+ * fully reproduce the context of the exception
-+ */
-+ cwd = get_fpu_cwd(task);
-+ swd = get_fpu_swd(task);
-+ switch (swd & ~cwd & 0x3f) {
-+ case 0x000: /* No unmasked exception */
-+ return;
-+ default: /* Multiple exceptions */
-+ break;
-+ case 0x001: /* Invalid Op */
-+ /*
-+ * swd & 0x240 == 0x040: Stack Underflow
-+ * swd & 0x240 == 0x240: Stack Overflow
-+ * User must clear the SF bit (0x40) if set
-+ */
-+ info.si_code = FPE_FLTINV;
-+ break;
-+ case 0x002: /* Denormalize */
-+ case 0x010: /* Underflow */
-+ info.si_code = FPE_FLTUND;
-+ break;
-+ case 0x004: /* Zero Divide */
-+ info.si_code = FPE_FLTDIV;
++ for (i = 0; scan_more && i < 15; i++) {
++ unsigned char opcode;
++ unsigned char instr_hi;
++ unsigned char instr_lo;
++
++ if (instr > limit)
+ break;
-+ case 0x008: /* Overflow */
-+ info.si_code = FPE_FLTOVF;
++ if (__get_user(opcode, (unsigned char __user *) instr))
++ break;
++
++ instr_hi = opcode & 0xf0;
++ instr_lo = opcode & 0x0f;
++ instr++;
++
++ switch (instr_hi) {
++ case 0x20:
++ case 0x30:
++ /* Values 0x26,0x2E,0x36,0x3E are valid x86 prefixes. */
++ scan_more = ((instr_lo & 7) == 0x6);
+ break;
-+ case 0x020: /* Precision */
-+ info.si_code = FPE_FLTRES;
++
++ case 0x60:
++ /* 0x64 thru 0x67 are valid prefixes in all modes. */
++ scan_more = (instr_lo & 0xC) == 0x4;
++ break;
++ case 0xF0:
++ /* 0xF0, 0xF2, and 0xF3 are valid prefixes */
++ scan_more = !instr_lo || (instr_lo>>1) == 1;
++ break;
++ case 0x00:
++ /* Prefetch instruction is 0x0F0D or 0x0F18 */
++ scan_more = 0;
++ if (instr > limit)
++ break;
++ if (__get_user(opcode, (unsigned char __user *) instr))
++ break;
++ prefetch = (instr_lo == 0xF) &&
++ (opcode == 0x0D || opcode == 0x18);
++ break;
++ default:
++ scan_more = 0;
+ break;
++ }
+ }
-+ force_sig_info(SIGFPE, &info, task);
++ return prefetch;
+}
+
-+fastcall void do_coprocessor_error(struct pt_regs * regs, long error_code)
++static inline int is_prefetch(struct pt_regs *regs, unsigned long addr,
++ unsigned long error_code)
+{
-+ ignore_fpu_irq = 1;
-+ math_error((void __user *)regs->eip);
-+}
++ if (unlikely(boot_cpu_data.x86_vendor == X86_VENDOR_AMD &&
++ boot_cpu_data.x86 >= 6)) {
++ /* Catch an obscure case of prefetch inside an NX page. */
++ if (nx_enabled && (error_code & 16))
++ return 0;
++ return __is_prefetch(regs, addr);
++ }
++ return 0;
++}
+
-+static void simd_math_error(void __user *eip)
++static noinline void force_sig_info_fault(int si_signo, int si_code,
++ unsigned long address, struct task_struct *tsk)
+{
-+ struct task_struct * task;
+ siginfo_t info;
-+ unsigned short mxcsr;
+
-+ /*
-+ * Save the info for the exception handler and clear the error.
-+ */
-+ task = current;
-+ save_init_fpu(task);
-+ task->thread.trap_no = 19;
-+ task->thread.error_code = 0;
-+ info.si_signo = SIGFPE;
++ info.si_signo = si_signo;
+ info.si_errno = 0;
-+ info.si_code = __SI_FAULT;
-+ info.si_addr = eip;
-+ /*
-+ * The SIMD FPU exceptions are handled a little differently, as there
-+ * is only a single status/control register. Thus, to determine which
-+ * unmasked exception was caught we must mask the exception mask bits
-+ * at 0x1f80, and then use these to mask the exception bits at 0x3f.
-+ */
-+ mxcsr = get_fpu_mxcsr(task);
-+ switch (~((mxcsr & 0x1f80) >> 7) & (mxcsr & 0x3f)) {
-+ case 0x000:
-+ default:
-+ break;
-+ case 0x001: /* Invalid Op */
-+ info.si_code = FPE_FLTINV;
-+ break;
-+ case 0x002: /* Denormalize */
-+ case 0x010: /* Underflow */
-+ info.si_code = FPE_FLTUND;
-+ break;
-+ case 0x004: /* Zero Divide */
-+ info.si_code = FPE_FLTDIV;
-+ break;
-+ case 0x008: /* Overflow */
-+ info.si_code = FPE_FLTOVF;
-+ break;
-+ case 0x020: /* Precision */
-+ info.si_code = FPE_FLTRES;
-+ break;
-+ }
-+ force_sig_info(SIGFPE, &info, task);
++ info.si_code = si_code;
++ info.si_addr = (void __user *)address;
++ force_sig_info(si_signo, &info, tsk);
+}
+
-+fastcall void do_simd_coprocessor_error(struct pt_regs * regs,
-+ long error_code)
++fastcall void do_invalid_op(struct pt_regs *, unsigned long);
++
++#ifdef CONFIG_X86_PAE
++static void dump_fault_path(unsigned long address)
+{
-+ if (cpu_has_xmm) {
-+ /* Handle SIMD FPU exceptions on PIII+ processors. */
-+ ignore_fpu_irq = 1;
-+ simd_math_error((void __user *)regs->eip);
-+ } else {
-+ /*
-+ * Handle strange cache flush from user space exception
-+ * in all other cases. This is undocumented behaviour.
-+ */
-+ if (regs->eflags & VM_MASK) {
-+ handle_vm86_fault((struct kernel_vm86_regs *)regs,
-+ error_code);
++ unsigned long *p, page;
++ unsigned long mfn;
++
++ page = read_cr3();
++ p = (unsigned long *)__va(page);
++ p += (address >> 30) * 2;
++ printk(KERN_ALERT "%08lx -> *pde = %08lx:%08lx\n", page, p[1], p[0]);
++ if (p[0] & _PAGE_PRESENT) {
++ mfn = (p[0] >> PAGE_SHIFT) | (p[1] << 20);
++ page = mfn_to_pfn(mfn) << PAGE_SHIFT;
++ p = (unsigned long *)__va(page);
++ address &= 0x3fffffff;
++ p += (address >> 21) * 2;
++ printk(KERN_ALERT "%08lx -> *pme = %08lx:%08lx\n",
++ page, p[1], p[0]);
++ mfn = (p[0] >> PAGE_SHIFT) | (p[1] << 20);
++#ifdef CONFIG_HIGHPTE
++ if (mfn_to_pfn(mfn) >= highstart_pfn)
+ return;
++#endif
++ if (p[0] & _PAGE_PRESENT) {
++ page = mfn_to_pfn(mfn) << PAGE_SHIFT;
++ p = (unsigned long *) __va(page);
++ address &= 0x001fffff;
++ p += (address >> 12) * 2;
++ printk(KERN_ALERT "%08lx -> *pte = %08lx:%08lx\n",
++ page, p[1], p[0]);
+ }
-+ current->thread.trap_no = 19;
-+ current->thread.error_code = error_code;
-+ die_if_kernel("cache flush denied", regs, error_code);
-+ force_sig(SIGSEGV, current);
+ }
+}
-+
-+#ifndef CONFIG_XEN
-+fastcall void do_spurious_interrupt_bug(struct pt_regs * regs,
-+ long error_code)
++#else
++static void dump_fault_path(unsigned long address)
+{
-+#if 0
-+ /* No need to warn about this any longer. */
-+ printk("Ignoring P6 Local APIC Spurious Interrupt Bug...\n");
++ unsigned long page;
++
++ page = read_cr3();
++ page = ((unsigned long *) __va(page))[address >> 22];
++ if (oops_may_print())
++ printk(KERN_ALERT "*pde = ma %08lx pa %08lx\n", page,
++ machine_to_phys(page));
++ /*
++ * We must not directly access the pte in the highpte
++ * case if the page table is located in highmem.
++ * And lets rather not kmap-atomic the pte, just in case
++ * it's allocated already.
++ */
++#ifdef CONFIG_HIGHPTE
++ if ((page >> PAGE_SHIFT) >= highstart_pfn)
++ return;
+#endif
++ if ((page & 1) && oops_may_print()) {
++ page &= PAGE_MASK;
++ address &= 0x003ff000;
++ page = machine_to_phys(page);
++ page = ((unsigned long *) __va(page))[address >> PAGE_SHIFT];
++ printk(KERN_ALERT "*pte = ma %08lx pa %08lx\n", page,
++ machine_to_phys(page));
++ }
+}
++#endif
+
-+fastcall void setup_x86_bogus_stack(unsigned char * stk)
-+{
-+ unsigned long *switch16_ptr, *switch32_ptr;
-+ struct pt_regs *regs;
-+ unsigned long stack_top, stack_bot;
-+ unsigned short iret_frame16_off;
-+ int cpu = smp_processor_id();
-+ /* reserve the space on 32bit stack for the magic switch16 pointer */
-+ memmove(stk, stk + 8, sizeof(struct pt_regs));
-+ switch16_ptr = (unsigned long *)(stk + sizeof(struct pt_regs));
-+ regs = (struct pt_regs *)stk;
-+ /* now the switch32 on 16bit stack */
-+ stack_bot = (unsigned long)&per_cpu(cpu_16bit_stack, cpu);
-+ stack_top = stack_bot + CPU_16BIT_STACK_SIZE;
-+ switch32_ptr = (unsigned long *)(stack_top - 8);
-+ iret_frame16_off = CPU_16BIT_STACK_SIZE - 8 - 20;
-+ /* copy iret frame on 16bit stack */
-+ memcpy((void *)(stack_bot + iret_frame16_off), ®s->eip, 20);
-+ /* fill in the switch pointers */
-+ switch16_ptr[0] = (regs->esp & 0xffff0000) | iret_frame16_off;
-+ switch16_ptr[1] = __ESPFIX_SS;
-+ switch32_ptr[0] = (unsigned long)stk + sizeof(struct pt_regs) +
-+ 8 - CPU_16BIT_STACK_SIZE;
-+ switch32_ptr[1] = __KERNEL_DS;
-+}
-+
-+fastcall unsigned char * fixup_x86_bogus_stack(unsigned short sp)
-+{
-+ unsigned long *switch32_ptr;
-+ unsigned char *stack16, *stack32;
-+ unsigned long stack_top, stack_bot;
-+ int len;
-+ int cpu = smp_processor_id();
-+ stack_bot = (unsigned long)&per_cpu(cpu_16bit_stack, cpu);
-+ stack_top = stack_bot + CPU_16BIT_STACK_SIZE;
-+ switch32_ptr = (unsigned long *)(stack_top - 8);
-+ /* copy the data from 16bit stack to 32bit stack */
-+ len = CPU_16BIT_STACK_SIZE - 8 - sp;
-+ stack16 = (unsigned char *)(stack_bot + sp);
-+ stack32 = (unsigned char *)
-+ (switch32_ptr[0] + CPU_16BIT_STACK_SIZE - 8 - len);
-+ memcpy(stack32, stack16, len);
-+ return stack32;
-+}
-+#endif
-+
-+/*
-+ * 'math_state_restore()' saves the current math information in the
-+ * old math state array, and gets the new ones from the current task
-+ *
-+ * Careful.. There are problems with IBM-designed IRQ13 behaviour.
-+ * Don't touch unless you *really* know how it works.
-+ *
-+ * Must be called with kernel preemption disabled (in this case,
-+ * local interrupts are disabled at the call-site in entry.S).
-+ */
-+asmlinkage void math_state_restore(struct pt_regs regs)
++static int spurious_fault(struct pt_regs *regs,
++ unsigned long address,
++ unsigned long error_code)
+{
-+ struct thread_info *thread = current_thread_info();
-+ struct task_struct *tsk = thread->task;
-+
-+ /* NB. 'clts' is done for us by Xen during virtual trap. */
-+ if (!tsk_used_math(tsk))
-+ init_fpu(tsk);
-+ restore_fpu(tsk);
-+ thread->status |= TS_USEDFPU; /* So we fnsave on switch_to() */
-+}
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
+
-+#ifndef CONFIG_MATH_EMULATION
++ /* Reserved-bit violation or user access to kernel space? */
++ if (error_code & 0x0c)
++ return 0;
+
-+asmlinkage void math_emulate(long arg)
-+{
-+ printk(KERN_EMERG "math-emulation not enabled and no coprocessor found.\n");
-+ printk(KERN_EMERG "killing %s.\n",current->comm);
-+ force_sig(SIGFPE,current);
-+ schedule();
-+}
++ pgd = init_mm.pgd + pgd_index(address);
++ if (!pgd_present(*pgd))
++ return 0;
+
-+#endif /* CONFIG_MATH_EMULATION */
++ pud = pud_offset(pgd, address);
++ if (!pud_present(*pud))
++ return 0;
+
-+#ifdef CONFIG_X86_F00F_BUG
-+void __init trap_init_f00f_bug(void)
-+{
-+ __set_fixmap(FIX_F00F_IDT, __pa(&idt_table), PAGE_KERNEL_RO);
++ pmd = pmd_offset(pud, address);
++ if (!pmd_present(*pmd))
++ return 0;
+
-+ /*
-+ * Update the IDT descriptor and reload the IDT so that
-+ * it uses the read-only mapped virtual address.
-+ */
-+ idt_descr.address = fix_to_virt(FIX_F00F_IDT);
-+ load_idt(&idt_descr);
-+}
++ pte = pte_offset_kernel(pmd, address);
++ if (!pte_present(*pte))
++ return 0;
++ if ((error_code & 0x02) && !pte_write(*pte))
++ return 0;
++#ifdef CONFIG_X86_PAE
++ if ((error_code & 0x10) && (__pte_val(*pte) & _PAGE_NX))
++ return 0;
+#endif
+
++ return 1;
++}
+
-+/*
-+ * NB. All these are "trap gates" (i.e. events_mask isn't set) except
-+ * for those that specify <dpl>|4 in the second field.
-+ */
-+static trap_info_t trap_table[] = {
-+ { 0, 0, __KERNEL_CS, (unsigned long)divide_error },
-+ { 1, 0|4, __KERNEL_CS, (unsigned long)debug },
-+ { 3, 3|4, __KERNEL_CS, (unsigned long)int3 },
-+ { 4, 3, __KERNEL_CS, (unsigned long)overflow },
-+ { 5, 0, __KERNEL_CS, (unsigned long)bounds },
-+ { 6, 0, __KERNEL_CS, (unsigned long)invalid_op },
-+ { 7, 0|4, __KERNEL_CS, (unsigned long)device_not_available },
-+ { 9, 0, __KERNEL_CS, (unsigned long)coprocessor_segment_overrun },
-+ { 10, 0, __KERNEL_CS, (unsigned long)invalid_TSS },
-+ { 11, 0, __KERNEL_CS, (unsigned long)segment_not_present },
-+ { 12, 0, __KERNEL_CS, (unsigned long)stack_segment },
-+ { 13, 0, __KERNEL_CS, (unsigned long)general_protection },
-+ { 14, 0|4, __KERNEL_CS, (unsigned long)page_fault },
-+ { 15, 0, __KERNEL_CS, (unsigned long)fixup_4gb_segment },
-+ { 16, 0, __KERNEL_CS, (unsigned long)coprocessor_error },
-+ { 17, 0, __KERNEL_CS, (unsigned long)alignment_check },
-+#ifdef CONFIG_X86_MCE
-+ { 18, 0, __KERNEL_CS, (unsigned long)machine_check },
-+#endif
-+ { 19, 0, __KERNEL_CS, (unsigned long)simd_coprocessor_error },
-+ { SYSCALL_VECTOR, 3, __KERNEL_CS, (unsigned long)system_call },
-+ { 0, 0, 0, 0 }
-+};
-+
-+void __init trap_init(void)
++static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address)
+{
-+ HYPERVISOR_set_trap_table(trap_table);
++ unsigned index = pgd_index(address);
++ pgd_t *pgd_k;
++ pud_t *pud, *pud_k;
++ pmd_t *pmd, *pmd_k;
+
-+ if (cpu_has_fxsr) {
-+ /*
-+ * Verify that the FXSAVE/FXRSTOR data will be 16-byte aligned.
-+ * Generates a compile-time "error: zero width for bit-field" if
-+ * the alignment is wrong.
-+ */
-+ struct fxsrAlignAssert {
-+ int _:!(offsetof(struct task_struct,
-+ thread.i387.fxsave) & 15);
-+ };
++ pgd += index;
++ pgd_k = init_mm.pgd + index;
+
-+ printk(KERN_INFO "Enabling fast FPU save and restore... ");
-+ set_in_cr4(X86_CR4_OSFXSR);
-+ printk("done.\n");
-+ }
-+ if (cpu_has_xmm) {
-+ printk(KERN_INFO "Enabling unmasked SIMD FPU exception "
-+ "support... ");
-+ set_in_cr4(X86_CR4_OSXMMEXCPT);
-+ printk("done.\n");
-+ }
++ if (!pgd_present(*pgd_k))
++ return NULL;
+
+ /*
-+ * Should be a barrier for any external CPU state.
++ * set_pgd(pgd, *pgd_k); here would be useless on PAE
++ * and redundant with the set_pmd() on non-PAE. As would
++ * set_pud.
+ */
-+ cpu_init();
-+}
-+
-+void smp_trap_init(trap_info_t *trap_ctxt)
-+{
-+ trap_info_t *t = trap_table;
-+
-+ for (t = trap_table; t->address; t++) {
-+ trap_ctxt[t->vector].flags = t->flags;
-+ trap_ctxt[t->vector].cs = t->cs;
-+ trap_ctxt[t->vector].address = t->address;
-+ }
-+}
+
-+static int __init kstack_setup(char *s)
-+{
-+ kstack_depth_to_print = simple_strtoul(s, NULL, 0);
-+ return 1;
-+}
-+__setup("kstack=", kstack_setup);
++ pud = pud_offset(pgd, address);
++ pud_k = pud_offset(pgd_k, address);
++ if (!pud_present(*pud_k))
++ return NULL;
+
-+#ifdef CONFIG_STACK_UNWIND
-+static int __init call_trace_setup(char *s)
-+{
-+ if (strcmp(s, "old") == 0)
-+ call_trace = -1;
-+ else if (strcmp(s, "both") == 0)
-+ call_trace = 0;
-+ else if (strcmp(s, "newfallback") == 0)
-+ call_trace = 1;
-+ else if (strcmp(s, "new") == 2)
-+ call_trace = 2;
-+ return 1;
-+}
-+__setup("call_trace=", call_trace_setup);
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/vm86.c tmp-linux-2.6-xen.patch/arch/i386/kernel/vm86.c
---- pristine-linux-2.6.18.2/arch/i386/kernel/vm86.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/vm86.c 2007-07-30 16:35:11.000000000 +0200
-@@ -97,7 +97,9 @@
- struct pt_regs * FASTCALL(save_v86_state(struct kernel_vm86_regs * regs));
- struct pt_regs * fastcall save_v86_state(struct kernel_vm86_regs * regs)
- {
-+#ifndef CONFIG_X86_NO_TSS
- struct tss_struct *tss;
-+#endif
- struct pt_regs *ret;
- unsigned long tmp;
-
-@@ -122,12 +124,16 @@ struct pt_regs * fastcall save_v86_state
- do_exit(SIGSEGV);
- }
-
-+#ifndef CONFIG_X86_NO_TSS
- tss = &per_cpu(init_tss, get_cpu());
-+#endif
- current->thread.esp0 = current->thread.saved_esp0;
- current->thread.sysenter_cs = __KERNEL_CS;
- load_esp0(tss, ¤t->thread);
- current->thread.saved_esp0 = 0;
-+#ifndef CONFIG_X86_NO_TSS
- put_cpu();
-+#endif
-
- loadsegment(fs, current->thread.saved_fs);
- loadsegment(gs, current->thread.saved_gs);
-@@ -251,7 +257,9 @@ out:
-
- static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk)
- {
-+#ifndef CONFIG_X86_NO_TSS
- struct tss_struct *tss;
-+#endif
- long eax;
- /*
- * make sure the vm86() system call doesn't try to do anything silly
-@@ -296,12 +304,16 @@ static void do_sys_vm86(struct kernel_vm
- savesegment(fs, tsk->thread.saved_fs);
- savesegment(gs, tsk->thread.saved_gs);
-
-+#ifndef CONFIG_X86_NO_TSS
- tss = &per_cpu(init_tss, get_cpu());
-+#endif
- tsk->thread.esp0 = (unsigned long) &info->VM86_TSS_ESP0;
- if (cpu_has_sep)
- tsk->thread.sysenter_cs = 0;
- load_esp0(tss, &tsk->thread);
-+#ifndef CONFIG_X86_NO_TSS
- put_cpu();
++ pmd = pmd_offset(pud, address);
++ pmd_k = pmd_offset(pud_k, address);
++ if (!pmd_present(*pmd_k))
++ return NULL;
++ if (!pmd_present(*pmd))
++#if CONFIG_XEN_COMPAT > 0x030002
++ set_pmd(pmd, *pmd_k);
++#else
++ /*
++ * When running on older Xen we must launder *pmd_k through
++ * pmd_val() to ensure that _PAGE_PRESENT is correctly set.
++ */
++ set_pmd(pmd, __pmd(pmd_val(*pmd_k)));
+#endif
-
- tsk->thread.screen_bitmap = info->screen_bitmap;
- if (info->flags & VM86_SCREEN_BITMAP)
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/vmlinux.lds.S tmp-linux-2.6-xen.patch/arch/i386/kernel/vmlinux.lds.S
---- pristine-linux-2.6.18.2/arch/i386/kernel/vmlinux.lds.S 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/vmlinux.lds.S 2007-07-30 16:35:11.000000000 +0200
-@@ -13,6 +13,12 @@ OUTPUT_FORMAT("elf32-i386", "elf32-i386"
- OUTPUT_ARCH(i386)
- ENTRY(phys_startup_32)
- jiffies = jiffies_64;
-+
-+PHDRS {
-+ text PT_LOAD FLAGS(5); /* R_E */
-+ data PT_LOAD FLAGS(7); /* RWE */
-+ note PT_NOTE FLAGS(4); /* R__ */
++ else
++ BUG_ON(pmd_page(*pmd) != pmd_page(*pmd_k));
++ return pmd_k;
+}
- SECTIONS
- {
- . = __KERNEL_START;
-@@ -26,7 +32,7 @@ SECTIONS
- KPROBES_TEXT
- *(.fixup)
- *(.gnu.warning)
-- } = 0x9090
-+ } :text = 0x9090
-
- _etext = .; /* End of text section */
-
-@@ -45,10 +51,11 @@ SECTIONS
- __tracedata_end = .;
-
- /* writeable */
-+ . = ALIGN(4096);
- .data : AT(ADDR(.data) - LOAD_OFFSET) { /* Data */
- *(.data)
- CONSTRUCTORS
-- }
-+ } :data
-
- . = ALIGN(4096);
- __nosave_begin = .;
-@@ -184,4 +191,6 @@ SECTIONS
- STABS_DEBUG
-
- DWARF_DEBUG
-+
-+ NOTES
- }
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/kernel/vsyscall-note-xen.S tmp-linux-2.6-xen.patch/arch/i386/kernel/vsyscall-note-xen.S
---- pristine-linux-2.6.18.2/arch/i386/kernel/vsyscall-note-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/kernel/vsyscall-note-xen.S 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,32 @@
-+/*
-+ * This supplies .note.* sections to go into the PT_NOTE inside the vDSO text.
-+ * Here we can supply some information useful to userland.
-+ * First we get the vanilla i386 note that supplies the kernel version info.
-+ */
-+
-+#include "vsyscall-note.S"
+
+/*
-+ * Now we add a special note telling glibc's dynamic linker a fake hardware
-+ * flavor that it will use to choose the search path for libraries in the
-+ * same way it uses real hardware capabilities like "mmx".
-+ * We supply "nosegneg" as the fake capability, to indicate that we
-+ * do not like negative offsets in instructions using segment overrides,
-+ * since we implement those inefficiently. This makes it possible to
-+ * install libraries optimized to avoid those access patterns in someplace
-+ * like /lib/i686/tls/nosegneg. Note that an /etc/ld.so.conf.d/file
-+ * corresponding to the bits here is needed to make ldconfig work right.
-+ * It should contain:
-+ * hwcap 0 nosegneg
-+ * to match the mapping of bit to name that we give here.
++ * Handle a fault on the vmalloc or module mapping area
++ *
++ * This assumes no large pages in there.
+ */
-+#define NOTE_KERNELCAP_BEGIN(ncaps, mask) \
-+ ASM_ELF_NOTE_BEGIN(".note.kernelcap", "a", "GNU", 2) \
-+ .long ncaps, mask
-+#define NOTE_KERNELCAP(bit, name) \
-+ .byte bit; .asciz name
-+#define NOTE_KERNELCAP_END ASM_ELF_NOTE_END
-+
-+NOTE_KERNELCAP_BEGIN(1, 1)
-+NOTE_KERNELCAP(0, "nosegneg")
-+NOTE_KERNELCAP_END
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/mach-xen/Makefile tmp-linux-2.6-xen.patch/arch/i386/mach-xen/Makefile
---- pristine-linux-2.6.18.2/arch/i386/mach-xen/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/mach-xen/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,5 @@
-+#
-+# Makefile for the linux kernel.
-+#
++static inline int vmalloc_fault(unsigned long address)
++{
++ unsigned long pgd_paddr;
++ pmd_t *pmd_k;
++ pte_t *pte_k;
++ /*
++ * Synchronize this task's top level page-table
++ * with the 'reference' page table.
++ *
++ * Do _not_ use "current" here. We might be inside
++ * an interrupt in the middle of a task switch..
++ */
++ pgd_paddr = read_cr3();
++ pmd_k = vmalloc_sync_one(__va(pgd_paddr), address);
++ if (!pmd_k)
++ return -1;
++ pte_k = pte_offset_kernel(pmd_k, address);
++ if (!pte_present(*pte_k))
++ return -1;
++ return 0;
++}
+
-+obj-y := setup.o
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/mach-xen/setup.c tmp-linux-2.6-xen.patch/arch/i386/mach-xen/setup.c
---- pristine-linux-2.6.18.2/arch/i386/mach-xen/setup.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/mach-xen/setup.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,147 @@
+/*
-+ * Machine specific setup for generic
++ * This routine handles page faults. It determines the address,
++ * and the problem, and then passes it off to one of the appropriate
++ * routines.
++ *
++ * error_code:
++ * bit 0 == 0 means no page found, 1 means protection fault
++ * bit 1 == 0 means read, 1 means write
++ * bit 2 == 0 means kernel, 1 means user-mode
++ * bit 3 == 1 means use of reserved bit detected
++ * bit 4 == 1 means fault was an instruction fetch
+ */
-+
-+#include <linux/smp.h>
-+#include <linux/init.h>
-+#include <linux/interrupt.h>
-+#include <linux/module.h>
-+#include <asm/acpi.h>
-+#include <asm/arch_hooks.h>
-+#include <asm/e820.h>
-+#include <asm/setup.h>
-+#include <asm/fixmap.h>
-+
-+#include <xen/interface/callback.h>
-+#include <xen/interface/memory.h>
-+
-+#ifdef CONFIG_HOTPLUG_CPU
-+#define DEFAULT_SEND_IPI (1)
-+#else
-+#define DEFAULT_SEND_IPI (0)
-+#endif
-+
-+int no_broadcast=DEFAULT_SEND_IPI;
-+
-+static __init int no_ipi_broadcast(char *str)
++fastcall void __kprobes do_page_fault(struct pt_regs *regs,
++ unsigned long error_code)
+{
-+ get_option(&str, &no_broadcast);
-+ printk ("Using %s mode\n", no_broadcast ? "No IPI Broadcast" :
-+ "IPI Broadcast");
-+ return 1;
-+}
++ struct task_struct *tsk;
++ struct mm_struct *mm;
++ struct vm_area_struct * vma;
++ unsigned long address;
++ int write, si_code;
+
-+__setup("no_ipi_broadcast", no_ipi_broadcast);
++ /* get the address */
++ address = read_cr2();
+
-+static int __init print_ipi_mode(void)
-+{
-+ printk ("Using IPI %s mode\n", no_broadcast ? "No-Shortcut" :
-+ "Shortcut");
-+ return 0;
-+}
++ /* Set the "privileged fault" bit to something sane. */
++ error_code &= ~4;
++ error_code |= (regs->xcs & 2) << 1;
++ if (regs->eflags & X86_EFLAGS_VM)
++ error_code |= 4;
+
-+late_initcall(print_ipi_mode);
++ tsk = current;
+
-+/**
-+ * machine_specific_memory_setup - Hook for machine specific memory setup.
-+ *
-+ * Description:
-+ * This is included late in kernel/setup.c so that it can make
-+ * use of all of the static functions.
-+ **/
++ si_code = SEGV_MAPERR;
+
-+char * __init machine_specific_memory_setup(void)
-+{
-+ int rc;
-+ struct xen_memory_map memmap;
+ /*
-+ * This is rather large for a stack variable but this early in
-+ * the boot process we know we have plenty slack space.
++ * We fault-in kernel-space virtual memory on-demand. The
++ * 'reference' page table is init_mm.pgd.
++ *
++ * NOTE! We MUST NOT take any locks for this case. We may
++ * be in an interrupt or a critical region, and should
++ * only copy the information from the master page table,
++ * nothing more.
++ *
++ * This verifies that the fault happens in kernel space
++ * (error_code & 4) == 0, and that the fault was not a
++ * protection error (error_code & 9) == 0.
+ */
-+ struct e820entry map[E820MAX];
-+
-+ memmap.nr_entries = E820MAX;
-+ set_xen_guest_handle(memmap.buffer, map);
-+
-+ rc = HYPERVISOR_memory_op(XENMEM_memory_map, &memmap);
-+ if ( rc == -ENOSYS ) {
-+ memmap.nr_entries = 1;
-+ map[0].addr = 0ULL;
-+ map[0].size = PFN_PHYS((unsigned long long)xen_start_info->nr_pages);
-+ /* 8MB slack (to balance backend allocations). */
-+ map[0].size += 8ULL << 20;
-+ map[0].type = E820_RAM;
-+ rc = 0;
++ if (unlikely(address >= TASK_SIZE)) {
++#ifdef CONFIG_XEN
++ /* Faults in hypervisor area can never be patched up. */
++ if (address >= hypervisor_virt_start)
++ goto bad_area_nosemaphore;
++#endif
++ if (!(error_code & 0x0000000d) && vmalloc_fault(address) >= 0)
++ return;
++ /* Can take a spurious fault if mapping changes R/O -> R/W. */
++ if (spurious_fault(regs, address, error_code))
++ return;
++ if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
++ SIGSEGV) == NOTIFY_STOP)
++ return;
++ /*
++ * Don't take the mm semaphore here. If we fixup a prefetch
++ * fault we could otherwise deadlock.
++ */
++ goto bad_area_nosemaphore;
+ }
-+ BUG_ON(rc);
+
-+ sanitize_e820_map(map, (char *)&memmap.nr_entries);
++ if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
++ SIGSEGV) == NOTIFY_STOP)
++ return;
+
-+ BUG_ON(copy_e820_map(map, (char)memmap.nr_entries) < 0);
++ /* It's safe to allow irq's after cr2 has been saved and the vmalloc
++ fault has been handled. */
++ if (regs->eflags & (X86_EFLAGS_IF|VM_MASK))
++ local_irq_enable();
+
-+ return "Xen";
-+}
++ mm = tsk->mm;
+
++ /*
++ * If we're in an interrupt, have no user context or are running in an
++ * atomic region then we must not take the fault..
++ */
++ if (in_atomic() || !mm)
++ goto bad_area_nosemaphore;
+
-+extern void hypervisor_callback(void);
-+extern void failsafe_callback(void);
-+extern void nmi(void);
++ /* When running in the kernel we expect faults to occur only to
++ * addresses in user space. All other faults represent errors in the
++ * kernel and should generate an OOPS. Unfortunatly, in the case of an
++ * erroneous fault occurring in a code path which already holds mmap_sem
++ * we will deadlock attempting to validate the fault against the
++ * address space. Luckily the kernel only validly references user
++ * space from well defined areas of code, which are listed in the
++ * exceptions table.
++ *
++ * As the vast majority of faults will be valid we will only perform
++ * the source reference check when there is a possibilty of a deadlock.
++ * Attempt to lock the address space, if we cannot we then validate the
++ * source. If this is invalid we can skip the address space check,
++ * thus avoiding the deadlock.
++ */
++ if (!down_read_trylock(&mm->mmap_sem)) {
++ if ((error_code & 4) == 0 &&
++ !search_exception_tables(regs->eip))
++ goto bad_area_nosemaphore;
++ down_read(&mm->mmap_sem);
++ }
+
-+unsigned long *machine_to_phys_mapping;
-+EXPORT_SYMBOL(machine_to_phys_mapping);
-+unsigned int machine_to_phys_order;
-+EXPORT_SYMBOL(machine_to_phys_order);
++ vma = find_vma(mm, address);
++ if (!vma)
++ goto bad_area;
++ if (vma->vm_start <= address)
++ goto good_area;
++ if (!(vma->vm_flags & VM_GROWSDOWN))
++ goto bad_area;
++ if (error_code & 4) {
++ /*
++ * Accessing the stack below %esp is always a bug.
++ * The large cushion allows instructions like enter
++ * and pusha to work. ("enter $65535,$31" pushes
++ * 32 pointers and then decrements %esp by 65535.)
++ */
++ if (address + 65536 + 32 * sizeof(unsigned long) < regs->esp)
++ goto bad_area;
++ }
++ if (expand_stack(vma, address))
++ goto bad_area;
++/*
++ * Ok, we have a good vm_area for this memory access, so
++ * we can handle it..
++ */
++good_area:
++ si_code = SEGV_ACCERR;
++ write = 0;
++ switch (error_code & 3) {
++ default: /* 3: write, present */
++#ifdef TEST_VERIFY_AREA
++ if (regs->cs == GET_KERNEL_CS())
++ printk("WP fault at %08lx\n", regs->eip);
++#endif
++ /* fall through */
++ case 2: /* write, not present */
++ if (!(vma->vm_flags & VM_WRITE))
++ goto bad_area;
++ write++;
++ break;
++ case 1: /* read, present */
++ goto bad_area;
++ case 0: /* read, not present */
++ if (!(vma->vm_flags & (VM_READ | VM_EXEC)))
++ goto bad_area;
++ }
+
-+void __init machine_specific_arch_setup(void)
-+{
-+ int ret;
-+ struct xen_machphys_mapping mapping;
-+ unsigned long machine_to_phys_nr_ents;
-+ struct xen_platform_parameters pp;
-+ static struct callback_register __initdata event = {
-+ .type = CALLBACKTYPE_event,
-+ .address = { __KERNEL_CS, (unsigned long)hypervisor_callback },
-+ };
-+ static struct callback_register __initdata failsafe = {
-+ .type = CALLBACKTYPE_failsafe,
-+ .address = { __KERNEL_CS, (unsigned long)failsafe_callback },
-+ };
-+ static struct callback_register __initdata nmi_cb = {
-+ .type = CALLBACKTYPE_nmi,
-+ .address = { __KERNEL_CS, (unsigned long)nmi },
-+ };
++ survive:
++ /*
++ * If for any reason at all we couldn't handle the fault,
++ * make sure we exit gracefully rather than endlessly redo
++ * the fault.
++ */
++ switch (handle_mm_fault(mm, vma, address, write)) {
++ case VM_FAULT_MINOR:
++ tsk->min_flt++;
++ break;
++ case VM_FAULT_MAJOR:
++ tsk->maj_flt++;
++ break;
++ case VM_FAULT_SIGBUS:
++ goto do_sigbus;
++ case VM_FAULT_OOM:
++ goto out_of_memory;
++ default:
++ BUG();
++ }
+
-+ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &event);
-+ if (ret == 0)
-+ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &failsafe);
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (ret == -ENOSYS)
-+ ret = HYPERVISOR_set_callbacks(
-+ event.address.cs, event.address.eip,
-+ failsafe.address.cs, failsafe.address.eip);
-+#endif
-+ BUG_ON(ret);
++ /*
++ * Did it hit the DOS screen memory VA from vm86 mode?
++ */
++ if (regs->eflags & VM_MASK) {
++ unsigned long bit = (address - 0xA0000) >> PAGE_SHIFT;
++ if (bit < 32)
++ tsk->thread.screen_bitmap |= 1 << bit;
++ }
++ up_read(&mm->mmap_sem);
++ return;
+
-+ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &nmi_cb);
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (ret == -ENOSYS) {
-+ static struct xennmi_callback __initdata cb = {
-+ .handler_address = (unsigned long)nmi
-+ };
++/*
++ * Something tried to access memory that isn't in our memory map..
++ * Fix it, but check if it's kernel or user first..
++ */
++bad_area:
++ up_read(&mm->mmap_sem);
+
-+ HYPERVISOR_nmi_op(XENNMI_register_callback, &cb);
++bad_area_nosemaphore:
++ /* User mode accesses just cause a SIGSEGV */
++ if (error_code & 4) {
++ /*
++ * Valid to do another page fault here because this one came
++ * from user space.
++ */
++ if (is_prefetch(regs, address, error_code))
++ return;
++
++ tsk->thread.cr2 = address;
++ /* Kernel addresses are always protection faults */
++ tsk->thread.error_code = error_code | (address >= TASK_SIZE);
++ tsk->thread.trap_no = 14;
++ force_sig_info_fault(SIGSEGV, si_code, address, tsk);
++ return;
+ }
-+#endif
+
-+ if (HYPERVISOR_xen_version(XENVER_platform_parameters, &pp) == 0)
-+ set_fixaddr_top(pp.virt_start);
++#ifdef CONFIG_X86_F00F_BUG
++ /*
++ * Pentium F0 0F C7 C8 bug workaround.
++ */
++ if (boot_cpu_data.f00f_bug) {
++ unsigned long nr;
++
++ nr = (address - idt_descr.address) >> 3;
+
-+ machine_to_phys_mapping = (unsigned long *)MACH2PHYS_VIRT_START;
-+ machine_to_phys_nr_ents = MACH2PHYS_NR_ENTRIES;
-+ if (HYPERVISOR_memory_op(XENMEM_machphys_mapping, &mapping) == 0) {
-+ machine_to_phys_mapping = (unsigned long *)mapping.v_start;
-+ machine_to_phys_nr_ents = mapping.max_mfn + 1;
++ if (nr == 6) {
++ do_invalid_op(regs, 0);
++ return;
++ }
+ }
-+ while ((1UL << machine_to_phys_order) < machine_to_phys_nr_ents )
-+ machine_to_phys_order++;
-+}
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/Makefile tmp-linux-2.6-xen.patch/arch/i386/Makefile
---- pristine-linux-2.6.18.2/arch/i386/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -48,6 +48,11 @@ CFLAGS += $(shell if [ $(call cc-vers
-
- CFLAGS += $(cflags-y)
-
-+cppflags-$(CONFIG_XEN) += \
-+ -D__XEN_INTERFACE_VERSION__=$(CONFIG_XEN_INTERFACE_VERSION)
++#endif
+
-+CPPFLAGS += $(cppflags-y)
++no_context:
++ /* Are we prepared to handle this kernel fault? */
++ if (fixup_exception(regs))
++ return;
+
- # Default subarch .c files
- mcore-y := mach-default
-
-@@ -71,6 +76,10 @@ mcore-$(CONFIG_X86_BIGSMP) := mach-defau
- mflags-$(CONFIG_X86_SUMMIT) := -Iinclude/asm-i386/mach-summit
- mcore-$(CONFIG_X86_SUMMIT) := mach-default
-
-+# Xen subarch support
-+mflags-$(CONFIG_X86_XEN) := -Iinclude/asm-i386/mach-xen
-+mcore-$(CONFIG_X86_XEN) := mach-xen
++ /*
++ * Valid to do another page fault here, because if this fault
++ * had been triggered by is_prefetch fixup_exception would have
++ * handled it.
++ */
++ if (is_prefetch(regs, address, error_code))
++ return;
+
- # generic subarchitecture
- mflags-$(CONFIG_X86_GENERICARCH) := -Iinclude/asm-i386/mach-generic
- mcore-$(CONFIG_X86_GENERICARCH) := mach-default
-@@ -105,6 +114,19 @@ boot := arch/i386/boot
- PHONY += zImage bzImage compressed zlilo bzlilo \
- zdisk bzdisk fdimage fdimage144 fdimage288 isoimage install
-
-+ifdef CONFIG_XEN
-+CPPFLAGS := -Iinclude$(if $(KBUILD_SRC),2)/asm/mach-xen $(CPPFLAGS)
-+head-y := arch/i386/kernel/head-xen.o arch/i386/kernel/init_task-xen.o
-+boot := arch/i386/boot-xen
-+.PHONY: vmlinuz
-+all: vmlinuz
++/*
++ * Oops. The kernel tried to access some bad page. We'll have to
++ * terminate things with extreme prejudice.
++ */
+
-+vmlinuz: vmlinux
-+ $(Q)$(MAKE) $(build)=$(boot) $@
++ bust_spinlocks(1);
++
++ if (oops_may_print()) {
++ #ifdef CONFIG_X86_PAE
++ if (error_code & 16) {
++ pte_t *pte = lookup_address(address);
++
++ if (pte && pte_present(*pte) && !pte_exec_kernel(*pte))
++ printk(KERN_CRIT "kernel tried to execute "
++ "NX-protected page - exploit attempt? "
++ "(uid: %d)\n", current->uid);
++ }
++ #endif
++ if (address < PAGE_SIZE)
++ printk(KERN_ALERT "BUG: unable to handle kernel NULL "
++ "pointer dereference");
++ else
++ printk(KERN_ALERT "BUG: unable to handle kernel paging"
++ " request");
++ printk(" at virtual address %08lx\n",address);
++ printk(KERN_ALERT " printing eip:\n");
++ printk("%08lx\n", regs->eip);
++ }
++ dump_fault_path(address);
++ tsk->thread.cr2 = address;
++ tsk->thread.trap_no = 14;
++ tsk->thread.error_code = error_code;
++ die("Oops", regs, error_code);
++ bust_spinlocks(0);
++ do_exit(SIGKILL);
+
-+install:
-+ $(Q)$(MAKE) $(build)=$(boot) XENGUEST=$(XENGUEST) $@
-+else
- all: bzImage
-
- # KBUILD_IMAGE specify target image being built
-@@ -127,6 +149,7 @@ fdimage fdimage144 fdimage288 isoimage:
-
- install:
- $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(KBUILD_IMAGE) install
-+endif
-
- archclean:
- $(Q)$(MAKE) $(clean)=arch/i386/boot
-@@ -145,3 +168,4 @@ endef
- CLEAN_FILES += arch/$(ARCH)/boot/fdimage \
- arch/$(ARCH)/boot/image.iso \
- arch/$(ARCH)/boot/mtools.conf
-+CLEAN_FILES += vmlinuz vmlinux-stripped
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/mm/fault-xen.c tmp-linux-2.6-xen.patch/arch/i386/mm/fault-xen.c
---- pristine-linux-2.6.18.2/arch/i386/mm/fault-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/mm/fault-xen.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,782 @@
+/*
-+ * linux/arch/i386/mm/fault.c
-+ *
-+ * Copyright (C) 1995 Linus Torvalds
++ * We ran out of memory, or some other thing happened to us that made
++ * us unable to handle the page fault gracefully.
+ */
++out_of_memory:
++ up_read(&mm->mmap_sem);
++ if (tsk->pid == 1) {
++ yield();
++ down_read(&mm->mmap_sem);
++ goto survive;
++ }
++ printk("VM: killing process %s\n", tsk->comm);
++ if (error_code & 4)
++ do_exit(SIGKILL);
++ goto no_context;
+
-+#include <linux/signal.h>
-+#include <linux/sched.h>
-+#include <linux/kernel.h>
-+#include <linux/errno.h>
-+#include <linux/string.h>
-+#include <linux/types.h>
-+#include <linux/ptrace.h>
-+#include <linux/mman.h>
-+#include <linux/mm.h>
-+#include <linux/smp.h>
-+#include <linux/smp_lock.h>
-+#include <linux/interrupt.h>
-+#include <linux/init.h>
-+#include <linux/tty.h>
-+#include <linux/vt_kern.h> /* For unblank_screen() */
-+#include <linux/highmem.h>
-+#include <linux/module.h>
-+#include <linux/kprobes.h>
++do_sigbus:
++ up_read(&mm->mmap_sem);
+
-+#include <asm/system.h>
-+#include <asm/uaccess.h>
-+#include <asm/desc.h>
-+#include <asm/kdebug.h>
++ /* Kernel mode? Handle exceptions or die */
++ if (!(error_code & 4))
++ goto no_context;
+
-+extern void die(const char *,struct pt_regs *,long);
++ /* User space => ok to do another page fault */
++ if (is_prefetch(regs, address, error_code))
++ return;
+
-+#ifdef CONFIG_KPROBES
-+ATOMIC_NOTIFIER_HEAD(notify_page_fault_chain);
-+int register_page_fault_notifier(struct notifier_block *nb)
-+{
-+ vmalloc_sync_all();
-+ return atomic_notifier_chain_register(¬ify_page_fault_chain, nb);
++ tsk->thread.cr2 = address;
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = 14;
++ force_sig_info_fault(SIGBUS, BUS_ADRERR, address, tsk);
+}
+
-+int unregister_page_fault_notifier(struct notifier_block *nb)
++#if !HAVE_SHARED_KERNEL_PMD
++void vmalloc_sync_all(void)
+{
-+ return atomic_notifier_chain_unregister(¬ify_page_fault_chain, nb);
++ /*
++ * Note that races in the updates of insync and start aren't
++ * problematic: insync can only get set bits added, and updates to
++ * start are only improving performance (without affecting correctness
++ * if undone).
++ * XEN: To work on PAE, we need to iterate over PMDs rather than PGDs.
++ * This change works just fine with 2-level paging too.
++ */
++#define sync_index(a) ((a) >> PMD_SHIFT)
++ static DECLARE_BITMAP(insync, PTRS_PER_PGD*PTRS_PER_PMD);
++ static unsigned long start = TASK_SIZE;
++ unsigned long address;
++
++ BUILD_BUG_ON(TASK_SIZE & ~PGDIR_MASK);
++ for (address = start;
++ address >= TASK_SIZE && address < hypervisor_virt_start;
++ address += 1UL << PMD_SHIFT) {
++ if (!test_bit(sync_index(address), insync)) {
++ unsigned long flags;
++ struct page *page;
++
++ spin_lock_irqsave(&pgd_lock, flags);
++ /* XEN: failure path assumes non-empty pgd_list. */
++ if (unlikely(!pgd_list)) {
++ spin_unlock_irqrestore(&pgd_lock, flags);
++ return;
++ }
++ for (page = pgd_list; page; page =
++ (struct page *)page->index)
++ if (!vmalloc_sync_one(page_address(page),
++ address)) {
++ BUG_ON(page != pgd_list);
++ break;
++ }
++ spin_unlock_irqrestore(&pgd_lock, flags);
++ if (!page)
++ set_bit(sync_index(address), insync);
++ }
++ if (address == start && test_bit(sync_index(address), insync))
++ start = address + (1UL << PMD_SHIFT);
++ }
+}
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/mm/highmem-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/mm/highmem-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,136 @@
++#include <linux/highmem.h>
++#include <linux/module.h>
+
-+static inline int notify_page_fault(enum die_val val, const char *str,
-+ struct pt_regs *regs, long err, int trap, int sig)
++void *kmap(struct page *page)
+{
-+ struct die_args args = {
-+ .regs = regs,
-+ .str = str,
-+ .err = err,
-+ .trapnr = trap,
-+ .signr = sig
-+ };
-+ return atomic_notifier_call_chain(¬ify_page_fault_chain, val, &args);
++ might_sleep();
++ if (!PageHighMem(page))
++ return page_address(page);
++ return kmap_high(page);
+}
-+#else
-+static inline int notify_page_fault(enum die_val val, const char *str,
-+ struct pt_regs *regs, long err, int trap, int sig)
++
++void kunmap(struct page *page)
+{
-+ return NOTIFY_DONE;
++ if (in_interrupt())
++ BUG();
++ if (!PageHighMem(page))
++ return;
++ kunmap_high(page);
+}
-+#endif
-+
+
+/*
-+ * Unlock any spinlocks which will prevent us from getting the
-+ * message out
++ * kmap_atomic/kunmap_atomic is significantly faster than kmap/kunmap because
++ * no global lock is needed and because the kmap code must perform a global TLB
++ * invalidation when the kmap pool wraps.
++ *
++ * However when holding an atomic kmap is is not legal to sleep, so atomic
++ * kmaps are appropriate for short, tight code paths only.
+ */
-+void bust_spinlocks(int yes)
++static void *__kmap_atomic(struct page *page, enum km_type type, pgprot_t prot)
+{
-+ int loglevel_save = console_loglevel;
++ enum fixed_addresses idx;
++ unsigned long vaddr;
+
-+ if (yes) {
-+ oops_in_progress = 1;
++ /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */
++ inc_preempt_count();
++ if (!PageHighMem(page))
++ return page_address(page);
++
++ idx = type + KM_TYPE_NR*smp_processor_id();
++ vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
++#ifdef CONFIG_DEBUG_HIGHMEM
++ if (!pte_none(*(kmap_pte-idx)))
++ BUG();
++#endif
++ set_pte_at_sync(&init_mm, vaddr, kmap_pte-idx, mk_pte(page, prot));
++
++ return (void*) vaddr;
++}
++
++void *kmap_atomic(struct page *page, enum km_type type)
++{
++ return __kmap_atomic(page, type, kmap_prot);
++}
++
++/* Same as kmap_atomic but with PAGE_KERNEL_RO page protection. */
++void *kmap_atomic_pte(struct page *page, enum km_type type)
++{
++ return __kmap_atomic(page, type,
++ test_bit(PG_pinned, &page->flags)
++ ? PAGE_KERNEL_RO : kmap_prot);
++}
++
++void kunmap_atomic(void *kvaddr, enum km_type type)
++{
++#if defined(CONFIG_DEBUG_HIGHMEM) || defined(CONFIG_XEN)
++ unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
++ enum fixed_addresses idx = type + KM_TYPE_NR*smp_processor_id();
++
++ if (vaddr < FIXADDR_START) { // FIXME
++ dec_preempt_count();
++ preempt_check_resched();
+ return;
+ }
-+#ifdef CONFIG_VT
-+ unblank_screen();
+#endif
-+ oops_in_progress = 0;
++
++#if defined(CONFIG_DEBUG_HIGHMEM)
++ if (vaddr != __fix_to_virt(FIX_KMAP_BEGIN+idx))
++ BUG();
++
+ /*
-+ * OK, the message is on the console. Now we call printk()
-+ * without oops_in_progress set so that printk will give klogd
-+ * a poke. Hold onto your hats...
++ * force other mappings to Oops if they'll try to access
++ * this pte without first remap it
+ */
-+ console_loglevel = 15; /* NMI oopser may have shut the console up */
-+ printk(" ");
-+ console_loglevel = loglevel_save;
++ pte_clear(&init_mm, vaddr, kmap_pte-idx);
++ __flush_tlb_one(vaddr);
++#elif defined(CONFIG_XEN)
++ /*
++ * We must ensure there are no dangling pagetable references when
++ * returning memory to Xen (decrease_reservation).
++ * XXX TODO: We could make this faster by only zapping when
++ * kmap_flush_unused is called but that is trickier and more invasive.
++ */
++ pte_clear(&init_mm, vaddr, kmap_pte-idx);
++#endif
++
++ dec_preempt_count();
++ preempt_check_resched();
+}
+
-+/*
-+ * Return EIP plus the CS segment base. The segment limit is also
-+ * adjusted, clamped to the kernel/user address space (whichever is
-+ * appropriate), and returned in *eip_limit.
-+ *
-+ * The segment is checked, because it might have been changed by another
-+ * task between the original faulting instruction and here.
-+ *
-+ * If CS is no longer a valid code segment, or if EIP is beyond the
-+ * limit, or if it is a kernel address when CS is not a kernel segment,
-+ * then the returned value will be greater than *eip_limit.
-+ *
-+ * This is slow, but is very rarely executed.
++/* This is the same as kmap_atomic() but can map memory that doesn't
++ * have a struct page associated with it.
+ */
-+static inline unsigned long get_segment_eip(struct pt_regs *regs,
-+ unsigned long *eip_limit)
++void *kmap_atomic_pfn(unsigned long pfn, enum km_type type)
+{
-+ unsigned long eip = regs->eip;
-+ unsigned seg = regs->xcs & 0xffff;
-+ u32 seg_ar, seg_limit, base, *desc;
++ enum fixed_addresses idx;
++ unsigned long vaddr;
+
-+ /* Unlikely, but must come before segment checks. */
-+ if (unlikely(regs->eflags & VM_MASK)) {
-+ base = seg << 4;
-+ *eip_limit = base + 0xffff;
-+ return base + (eip & 0xffff);
-+ }
++ inc_preempt_count();
+
-+ /* The standard kernel/user address space limit. */
-+ *eip_limit = (seg & 2) ? USER_DS.seg : KERNEL_DS.seg;
-+
-+ /* By far the most common cases. */
-+ if (likely(seg == __USER_CS || seg == GET_KERNEL_CS()))
-+ return eip;
++ idx = type + KM_TYPE_NR*smp_processor_id();
++ vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
++ set_pte(kmap_pte-idx, pfn_pte(pfn, kmap_prot));
++ __flush_tlb_one(vaddr);
+
-+ /* Check the segment exists, is within the current LDT/GDT size,
-+ that kernel/user (ring 0..3) has the appropriate privilege,
-+ that it's a code segment, and get the limit. */
-+ __asm__ ("larl %3,%0; lsll %3,%1"
-+ : "=&r" (seg_ar), "=r" (seg_limit) : "0" (0), "rm" (seg));
-+ if ((~seg_ar & 0x9800) || eip > seg_limit) {
-+ *eip_limit = 0;
-+ return 1; /* So that returned eip > *eip_limit. */
-+ }
-+
-+ /* Get the GDT/LDT descriptor base.
-+ When you look for races in this code remember that
-+ LDT and other horrors are only used in user space. */
-+ if (seg & (1<<2)) {
-+ /* Must lock the LDT while reading it. */
-+ down(¤t->mm->context.sem);
-+ desc = current->mm->context.ldt;
-+ desc = (void *)desc + (seg & ~7);
-+ } else {
-+ /* Must disable preemption while reading the GDT. */
-+ desc = (u32 *)get_cpu_gdt_table(get_cpu());
-+ desc = (void *)desc + (seg & ~7);
-+ }
++ return (void*) vaddr;
++}
+
-+ /* Decode the code segment base from the descriptor */
-+ base = get_desc_base((unsigned long *)desc);
++struct page *kmap_atomic_to_page(void *ptr)
++{
++ unsigned long idx, vaddr = (unsigned long)ptr;
++ pte_t *pte;
+
-+ if (seg & (1<<2)) {
-+ up(¤t->mm->context.sem);
-+ } else
-+ put_cpu();
++ if (vaddr < FIXADDR_START)
++ return virt_to_page(ptr);
+
-+ /* Adjust EIP and segment limit, and clamp at the kernel limit.
-+ It's legitimate for segments to wrap at 0xffffffff. */
-+ seg_limit += base;
-+ if (seg_limit < *eip_limit && seg_limit >= base)
-+ *eip_limit = seg_limit;
-+ return eip + base;
++ idx = virt_to_fix(vaddr);
++ pte = kmap_pte - (idx - FIX_KMAP_BEGIN);
++ return pte_page(*pte);
+}
+
-+/*
-+ * Sometimes AMD Athlon/Opteron CPUs report invalid exceptions on prefetch.
-+ * Check that here and ignore it.
++EXPORT_SYMBOL(kmap);
++EXPORT_SYMBOL(kunmap);
++EXPORT_SYMBOL(kmap_atomic);
++EXPORT_SYMBOL(kmap_atomic_pte);
++EXPORT_SYMBOL(kunmap_atomic);
++EXPORT_SYMBOL(kmap_atomic_to_page);
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/mm/hypervisor.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/mm/hypervisor.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,602 @@
++/******************************************************************************
++ * mm/hypervisor.c
++ *
++ * Update page tables via the hypervisor.
++ *
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
-+static int __is_prefetch(struct pt_regs *regs, unsigned long addr)
-+{
-+ unsigned long limit;
-+ unsigned long instr = get_segment_eip (regs, &limit);
-+ int scan_more = 1;
-+ int prefetch = 0;
-+ int i;
-+
-+ for (i = 0; scan_more && i < 15; i++) {
-+ unsigned char opcode;
-+ unsigned char instr_hi;
-+ unsigned char instr_lo;
-+
-+ if (instr > limit)
-+ break;
-+ if (__get_user(opcode, (unsigned char __user *) instr))
-+ break;
+
-+ instr_hi = opcode & 0xf0;
-+ instr_lo = opcode & 0x0f;
-+ instr++;
++#include <linux/sched.h>
++#include <linux/mm.h>
++#include <linux/vmalloc.h>
++#include <asm/page.h>
++#include <asm/pgtable.h>
++#include <asm/hypervisor.h>
++#include <xen/balloon.h>
++#include <xen/features.h>
++#include <xen/interface/memory.h>
++#include <linux/module.h>
++#include <linux/percpu.h>
++#include <asm/tlbflush.h>
++#include <linux/highmem.h>
+
-+ switch (instr_hi) {
-+ case 0x20:
-+ case 0x30:
-+ /* Values 0x26,0x2E,0x36,0x3E are valid x86 prefixes. */
-+ scan_more = ((instr_lo & 7) == 0x6);
-+ break;
-+
-+ case 0x60:
-+ /* 0x64 thru 0x67 are valid prefixes in all modes. */
-+ scan_more = (instr_lo & 0xC) == 0x4;
-+ break;
-+ case 0xF0:
-+ /* 0xF0, 0xF2, and 0xF3 are valid prefixes */
-+ scan_more = !instr_lo || (instr_lo>>1) == 1;
-+ break;
-+ case 0x00:
-+ /* Prefetch instruction is 0x0F0D or 0x0F18 */
-+ scan_more = 0;
-+ if (instr > limit)
-+ break;
-+ if (__get_user(opcode, (unsigned char __user *) instr))
-+ break;
-+ prefetch = (instr_lo == 0xF) &&
-+ (opcode == 0x0D || opcode == 0x18);
-+ break;
-+ default:
-+ scan_more = 0;
-+ break;
-+ }
-+ }
-+ return prefetch;
++void xen_l1_entry_update(pte_t *ptr, pte_t val)
++{
++ mmu_update_t u;
++#ifdef CONFIG_HIGHPTE
++ u.ptr = ((unsigned long)ptr >= (unsigned long)high_memory) ?
++ arbitrary_virt_to_machine(ptr) : virt_to_machine(ptr);
++#else
++ u.ptr = virt_to_machine(ptr);
++#endif
++ u.val = __pte_val(val);
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
+}
++EXPORT_SYMBOL_GPL(xen_l1_entry_update);
+
-+static inline int is_prefetch(struct pt_regs *regs, unsigned long addr,
-+ unsigned long error_code)
++void xen_l2_entry_update(pmd_t *ptr, pmd_t val)
+{
-+ if (unlikely(boot_cpu_data.x86_vendor == X86_VENDOR_AMD &&
-+ boot_cpu_data.x86 >= 6)) {
-+ /* Catch an obscure case of prefetch inside an NX page. */
-+ if (nx_enabled && (error_code & 16))
-+ return 0;
-+ return __is_prefetch(regs, addr);
-+ }
-+ return 0;
-+}
++ mmu_update_t u;
++ u.ptr = virt_to_machine(ptr);
++ u.val = __pmd_val(val);
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
++}
+
-+static noinline void force_sig_info_fault(int si_signo, int si_code,
-+ unsigned long address, struct task_struct *tsk)
++#if defined(CONFIG_X86_PAE) || defined(CONFIG_X86_64)
++void xen_l3_entry_update(pud_t *ptr, pud_t val)
+{
-+ siginfo_t info;
-+
-+ info.si_signo = si_signo;
-+ info.si_errno = 0;
-+ info.si_code = si_code;
-+ info.si_addr = (void __user *)address;
-+ force_sig_info(si_signo, &info, tsk);
++ mmu_update_t u;
++ u.ptr = virt_to_machine(ptr);
++ u.val = __pud_val(val);
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
+}
++#endif
+
-+fastcall void do_invalid_op(struct pt_regs *, unsigned long);
-+
-+#ifdef CONFIG_X86_PAE
-+static void dump_fault_path(unsigned long address)
++#ifdef CONFIG_X86_64
++void xen_l4_entry_update(pgd_t *ptr, pgd_t val)
+{
-+ unsigned long *p, page;
-+ unsigned long mfn;
-+
-+ page = read_cr3();
-+ p = (unsigned long *)__va(page);
-+ p += (address >> 30) * 2;
-+ printk(KERN_ALERT "%08lx -> *pde = %08lx:%08lx\n", page, p[1], p[0]);
-+ if (p[0] & 1) {
-+ mfn = (p[0] >> PAGE_SHIFT) | (p[1] << 20);
-+ page = mfn_to_pfn(mfn) << PAGE_SHIFT;
-+ p = (unsigned long *)__va(page);
-+ address &= 0x3fffffff;
-+ p += (address >> 21) * 2;
-+ printk(KERN_ALERT "%08lx -> *pme = %08lx:%08lx\n",
-+ page, p[1], p[0]);
-+ mfn = (p[0] >> PAGE_SHIFT) | (p[1] << 20);
-+#ifdef CONFIG_HIGHPTE
-+ if (mfn_to_pfn(mfn) >= highstart_pfn)
-+ return;
-+#endif
-+ if (p[0] & 1) {
-+ page = mfn_to_pfn(mfn) << PAGE_SHIFT;
-+ p = (unsigned long *) __va(page);
-+ address &= 0x001fffff;
-+ p += (address >> 12) * 2;
-+ printk(KERN_ALERT "%08lx -> *pte = %08lx:%08lx\n",
-+ page, p[1], p[0]);
-+ }
-+ }
++ mmu_update_t u;
++ u.ptr = virt_to_machine(ptr);
++ u.val = __pgd_val(val);
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
+}
-+#else
-+static void dump_fault_path(unsigned long address)
-+{
-+ unsigned long page;
++#endif /* CONFIG_X86_64 */
+
-+ page = read_cr3();
-+ page = ((unsigned long *) __va(page))[address >> 22];
-+ if (oops_may_print())
-+ printk(KERN_ALERT "*pde = ma %08lx pa %08lx\n", page,
-+ machine_to_phys(page));
-+ /*
-+ * We must not directly access the pte in the highpte
-+ * case if the page table is located in highmem.
-+ * And lets rather not kmap-atomic the pte, just in case
-+ * it's allocated already.
-+ */
-+#ifdef CONFIG_HIGHPTE
-+ if ((page >> PAGE_SHIFT) >= highstart_pfn)
-+ return;
-+#endif
-+ if ((page & 1) && oops_may_print()) {
-+ page &= PAGE_MASK;
-+ address &= 0x003ff000;
-+ page = machine_to_phys(page);
-+ page = ((unsigned long *) __va(page))[address >> PAGE_SHIFT];
-+ printk(KERN_ALERT "*pte = ma %08lx pa %08lx\n", page,
-+ machine_to_phys(page));
-+ }
++void xen_pt_switch(unsigned long ptr)
++{
++ struct mmuext_op op;
++ op.cmd = MMUEXT_NEW_BASEPTR;
++ op.arg1.mfn = pfn_to_mfn(ptr >> PAGE_SHIFT);
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
+}
-+#endif
+
-+static int spurious_fault(struct pt_regs *regs,
-+ unsigned long address,
-+ unsigned long error_code)
++void xen_new_user_pt(unsigned long ptr)
+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte;
-+
-+ /* Reserved-bit violation or user access to kernel space? */
-+ if (error_code & 0x0c)
-+ return 0;
-+
-+ pgd = init_mm.pgd + pgd_index(address);
-+ if (!pgd_present(*pgd))
-+ return 0;
++ struct mmuext_op op;
++ op.cmd = MMUEXT_NEW_USER_BASEPTR;
++ op.arg1.mfn = pfn_to_mfn(ptr >> PAGE_SHIFT);
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++}
+
-+ pud = pud_offset(pgd, address);
-+ if (!pud_present(*pud))
-+ return 0;
++void xen_tlb_flush(void)
++{
++ struct mmuext_op op;
++ op.cmd = MMUEXT_TLB_FLUSH_LOCAL;
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++}
++EXPORT_SYMBOL(xen_tlb_flush);
+
-+ pmd = pmd_offset(pud, address);
-+ if (!pmd_present(*pmd))
-+ return 0;
++void xen_invlpg(unsigned long ptr)
++{
++ struct mmuext_op op;
++ op.cmd = MMUEXT_INVLPG_LOCAL;
++ op.arg1.linear_addr = ptr & PAGE_MASK;
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++}
++EXPORT_SYMBOL(xen_invlpg);
+
-+ pte = pte_offset_kernel(pmd, address);
-+ if (!pte_present(*pte))
-+ return 0;
-+ if ((error_code & 0x02) && !pte_write(*pte))
-+ return 0;
-+#ifdef CONFIG_X86_PAE
-+ if ((error_code & 0x10) && (pte_val(*pte) & _PAGE_NX))
-+ return 0;
-+#endif
++#ifdef CONFIG_SMP
+
-+ return 1;
++void xen_tlb_flush_all(void)
++{
++ struct mmuext_op op;
++ op.cmd = MMUEXT_TLB_FLUSH_ALL;
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
+}
+
-+static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address)
++void xen_tlb_flush_mask(cpumask_t *mask)
+{
-+ unsigned index = pgd_index(address);
-+ pgd_t *pgd_k;
-+ pud_t *pud, *pud_k;
-+ pmd_t *pmd, *pmd_k;
-+
-+ pgd += index;
-+ pgd_k = init_mm.pgd + index;
++ struct mmuext_op op;
++ if ( cpus_empty(*mask) )
++ return;
++ op.cmd = MMUEXT_TLB_FLUSH_MULTI;
++ set_xen_guest_handle(op.arg2.vcpumask, mask->bits);
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++}
+
-+ if (!pgd_present(*pgd_k))
-+ return NULL;
++void xen_invlpg_all(unsigned long ptr)
++{
++ struct mmuext_op op;
++ op.cmd = MMUEXT_INVLPG_ALL;
++ op.arg1.linear_addr = ptr & PAGE_MASK;
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++}
+
-+ /*
-+ * set_pgd(pgd, *pgd_k); here would be useless on PAE
-+ * and redundant with the set_pmd() on non-PAE. As would
-+ * set_pud.
-+ */
++void xen_invlpg_mask(cpumask_t *mask, unsigned long ptr)
++{
++ struct mmuext_op op;
++ if ( cpus_empty(*mask) )
++ return;
++ op.cmd = MMUEXT_INVLPG_MULTI;
++ op.arg1.linear_addr = ptr & PAGE_MASK;
++ set_xen_guest_handle(op.arg2.vcpumask, mask->bits);
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++}
+
-+ pud = pud_offset(pgd, address);
-+ pud_k = pud_offset(pgd_k, address);
-+ if (!pud_present(*pud_k))
-+ return NULL;
++#endif /* CONFIG_SMP */
+
-+ pmd = pmd_offset(pud, address);
-+ pmd_k = pmd_offset(pud_k, address);
-+ if (!pmd_present(*pmd_k))
-+ return NULL;
-+ if (!pmd_present(*pmd))
-+#ifndef CONFIG_XEN
-+ set_pmd(pmd, *pmd_k);
++void xen_pgd_pin(unsigned long ptr)
++{
++ struct mmuext_op op;
++#ifdef CONFIG_X86_64
++ op.cmd = MMUEXT_PIN_L4_TABLE;
++#elif defined(CONFIG_X86_PAE)
++ op.cmd = MMUEXT_PIN_L3_TABLE;
+#else
-+ /*
-+ * When running on Xen we must launder *pmd_k through
-+ * pmd_val() to ensure that _PAGE_PRESENT is correctly set.
-+ */
-+ set_pmd(pmd, __pmd(pmd_val(*pmd_k)));
++ op.cmd = MMUEXT_PIN_L2_TABLE;
+#endif
-+ else
-+ BUG_ON(pmd_page(*pmd) != pmd_page(*pmd_k));
-+ return pmd_k;
++ op.arg1.mfn = pfn_to_mfn(ptr >> PAGE_SHIFT);
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
+}
+
-+/*
-+ * Handle a fault on the vmalloc or module mapping area
-+ *
-+ * This assumes no large pages in there.
-+ */
-+static inline int vmalloc_fault(unsigned long address)
++void xen_pgd_unpin(unsigned long ptr)
+{
-+ unsigned long pgd_paddr;
-+ pmd_t *pmd_k;
-+ pte_t *pte_k;
-+ /*
-+ * Synchronize this task's top level page-table
-+ * with the 'reference' page table.
-+ *
-+ * Do _not_ use "current" here. We might be inside
-+ * an interrupt in the middle of a task switch..
-+ */
-+ pgd_paddr = read_cr3();
-+ pmd_k = vmalloc_sync_one(__va(pgd_paddr), address);
-+ if (!pmd_k)
-+ return -1;
-+ pte_k = pte_offset_kernel(pmd_k, address);
-+ if (!pte_present(*pte_k))
-+ return -1;
-+ return 0;
++ struct mmuext_op op;
++ op.cmd = MMUEXT_UNPIN_TABLE;
++ op.arg1.mfn = pfn_to_mfn(ptr >> PAGE_SHIFT);
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
+}
+
-+/*
-+ * This routine handles page faults. It determines the address,
-+ * and the problem, and then passes it off to one of the appropriate
-+ * routines.
-+ *
-+ * error_code:
-+ * bit 0 == 0 means no page found, 1 means protection fault
-+ * bit 1 == 0 means read, 1 means write
-+ * bit 2 == 0 means kernel, 1 means user-mode
-+ * bit 3 == 1 means use of reserved bit detected
-+ * bit 4 == 1 means fault was an instruction fetch
-+ */
-+fastcall void __kprobes do_page_fault(struct pt_regs *regs,
-+ unsigned long error_code)
++void xen_set_ldt(const void *ptr, unsigned int ents)
+{
-+ struct task_struct *tsk;
-+ struct mm_struct *mm;
-+ struct vm_area_struct * vma;
-+ unsigned long address;
-+ int write, si_code;
-+
-+ /* get the address */
-+ address = read_cr2();
++ struct mmuext_op op;
++ op.cmd = MMUEXT_SET_LDT;
++ op.arg1.linear_addr = (unsigned long)ptr;
++ op.arg2.nr_ents = ents;
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++}
+
-+ /* Set the "privileged fault" bit to something sane. */
-+ error_code &= ~4;
-+ error_code |= (regs->xcs & 2) << 1;
-+ if (regs->eflags & X86_EFLAGS_VM)
-+ error_code |= 4;
++/*
++ * Bitmap is indexed by page number. If bit is set, the page is part of a
++ * xen_create_contiguous_region() area of memory.
++ */
++unsigned long *contiguous_bitmap;
+
-+ tsk = current;
++static void contiguous_bitmap_set(
++ unsigned long first_page, unsigned long nr_pages)
++{
++ unsigned long start_off, end_off, curr_idx, end_idx;
+
-+ si_code = SEGV_MAPERR;
++ curr_idx = first_page / BITS_PER_LONG;
++ start_off = first_page & (BITS_PER_LONG-1);
++ end_idx = (first_page + nr_pages) / BITS_PER_LONG;
++ end_off = (first_page + nr_pages) & (BITS_PER_LONG-1);
+
-+ /*
-+ * We fault-in kernel-space virtual memory on-demand. The
-+ * 'reference' page table is init_mm.pgd.
-+ *
-+ * NOTE! We MUST NOT take any locks for this case. We may
-+ * be in an interrupt or a critical region, and should
-+ * only copy the information from the master page table,
-+ * nothing more.
-+ *
-+ * This verifies that the fault happens in kernel space
-+ * (error_code & 4) == 0, and that the fault was not a
-+ * protection error (error_code & 9) == 0.
-+ */
-+ if (unlikely(address >= TASK_SIZE)) {
-+#ifdef CONFIG_XEN
-+ /* Faults in hypervisor area can never be patched up. */
-+ if (address >= hypervisor_virt_start)
-+ goto bad_area_nosemaphore;
-+#endif
-+ if (!(error_code & 0x0000000d) && vmalloc_fault(address) >= 0)
-+ return;
-+ /* Can take a spurious fault if mapping changes R/O -> R/W. */
-+ if (spurious_fault(regs, address, error_code))
-+ return;
-+ if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
-+ SIGSEGV) == NOTIFY_STOP)
-+ return;
-+ /*
-+ * Don't take the mm semaphore here. If we fixup a prefetch
-+ * fault we could otherwise deadlock.
-+ */
-+ goto bad_area_nosemaphore;
++ if (curr_idx == end_idx) {
++ contiguous_bitmap[curr_idx] |=
++ ((1UL<<end_off)-1) & -(1UL<<start_off);
++ } else {
++ contiguous_bitmap[curr_idx] |= -(1UL<<start_off);
++ while ( ++curr_idx < end_idx )
++ contiguous_bitmap[curr_idx] = ~0UL;
++ contiguous_bitmap[curr_idx] |= (1UL<<end_off)-1;
+ }
++}
+
-+ if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
-+ SIGSEGV) == NOTIFY_STOP)
-+ return;
++static void contiguous_bitmap_clear(
++ unsigned long first_page, unsigned long nr_pages)
++{
++ unsigned long start_off, end_off, curr_idx, end_idx;
+
-+ /* It's safe to allow irq's after cr2 has been saved and the vmalloc
-+ fault has been handled. */
-+ if (regs->eflags & (X86_EFLAGS_IF|VM_MASK))
-+ local_irq_enable();
++ curr_idx = first_page / BITS_PER_LONG;
++ start_off = first_page & (BITS_PER_LONG-1);
++ end_idx = (first_page + nr_pages) / BITS_PER_LONG;
++ end_off = (first_page + nr_pages) & (BITS_PER_LONG-1);
+
-+ mm = tsk->mm;
++ if (curr_idx == end_idx) {
++ contiguous_bitmap[curr_idx] &=
++ -(1UL<<end_off) | ((1UL<<start_off)-1);
++ } else {
++ contiguous_bitmap[curr_idx] &= (1UL<<start_off)-1;
++ while ( ++curr_idx != end_idx )
++ contiguous_bitmap[curr_idx] = 0;
++ contiguous_bitmap[curr_idx] &= -(1UL<<end_off);
++ }
++}
++
++/* Protected by balloon_lock. */
++#define MAX_CONTIG_ORDER 9 /* 2MB */
++static unsigned long discontig_frames[1<<MAX_CONTIG_ORDER];
++static unsigned long limited_frames[1<<MAX_CONTIG_ORDER];
++static multicall_entry_t cr_mcl[1<<MAX_CONTIG_ORDER];
++
++/* Ensure multi-page extents are contiguous in machine memory. */
++int xen_create_contiguous_region(
++ unsigned long vstart, unsigned int order, unsigned int address_bits)
++{
++ unsigned long *in_frames = discontig_frames, out_frame;
++ unsigned long frame, flags;
++ unsigned int i;
++ int rc, success;
++ struct xen_memory_exchange exchange = {
++ .in = {
++ .nr_extents = 1UL << order,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ },
++ .out = {
++ .nr_extents = 1,
++ .extent_order = order,
++ .address_bits = address_bits,
++ .domid = DOMID_SELF
++ }
++ };
+
+ /*
-+ * If we're in an interrupt, have no user context or are running in an
-+ * atomic region then we must not take the fault..
++ * Currently an auto-translated guest will not perform I/O, nor will
++ * it require PAE page directories below 4GB. Therefore any calls to
++ * this function are redundant and can be ignored.
+ */
-+ if (in_atomic() || !mm)
-+ goto bad_area_nosemaphore;
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return 0;
+
-+ /* When running in the kernel we expect faults to occur only to
-+ * addresses in user space. All other faults represent errors in the
-+ * kernel and should generate an OOPS. Unfortunatly, in the case of an
-+ * erroneous fault occurring in a code path which already holds mmap_sem
-+ * we will deadlock attempting to validate the fault against the
-+ * address space. Luckily the kernel only validly references user
-+ * space from well defined areas of code, which are listed in the
-+ * exceptions table.
-+ *
-+ * As the vast majority of faults will be valid we will only perform
-+ * the source reference check when there is a possibilty of a deadlock.
-+ * Attempt to lock the address space, if we cannot we then validate the
-+ * source. If this is invalid we can skip the address space check,
-+ * thus avoiding the deadlock.
-+ */
-+ if (!down_read_trylock(&mm->mmap_sem)) {
-+ if ((error_code & 4) == 0 &&
-+ !search_exception_tables(regs->eip))
-+ goto bad_area_nosemaphore;
-+ down_read(&mm->mmap_sem);
-+ }
++ if (unlikely(order > MAX_CONTIG_ORDER))
++ return -ENOMEM;
+
-+ vma = find_vma(mm, address);
-+ if (!vma)
-+ goto bad_area;
-+ if (vma->vm_start <= address)
-+ goto good_area;
-+ if (!(vma->vm_flags & VM_GROWSDOWN))
-+ goto bad_area;
-+ if (error_code & 4) {
-+ /*
-+ * Accessing the stack below %esp is always a bug.
-+ * The large cushion allows instructions like enter
-+ * and pusha to work. ("enter $65535,$31" pushes
-+ * 32 pointers and then decrements %esp by 65535.)
-+ */
-+ if (address + 65536 + 32 * sizeof(unsigned long) < regs->esp)
-+ goto bad_area;
-+ }
-+ if (expand_stack(vma, address))
-+ goto bad_area;
-+/*
-+ * Ok, we have a good vm_area for this memory access, so
-+ * we can handle it..
-+ */
-+good_area:
-+ si_code = SEGV_ACCERR;
-+ write = 0;
-+ switch (error_code & 3) {
-+ default: /* 3: write, present */
-+#ifdef TEST_VERIFY_AREA
-+ if (regs->cs == GET_KERNEL_CS())
-+ printk("WP fault at %08lx\n", regs->eip);
-+#endif
-+ /* fall through */
-+ case 2: /* write, not present */
-+ if (!(vma->vm_flags & VM_WRITE))
-+ goto bad_area;
-+ write++;
-+ break;
-+ case 1: /* read, present */
-+ goto bad_area;
-+ case 0: /* read, not present */
-+ if (!(vma->vm_flags & (VM_READ | VM_EXEC)))
-+ goto bad_area;
++ set_xen_guest_handle(exchange.in.extent_start, in_frames);
++ set_xen_guest_handle(exchange.out.extent_start, &out_frame);
++
++ scrub_pages((void *)vstart, 1 << order);
++
++ balloon_lock(flags);
++
++ /* 1. Zap current PTEs, remembering MFNs. */
++ for (i = 0; i < (1U<<order); i++) {
++ in_frames[i] = pfn_to_mfn((__pa(vstart) >> PAGE_SHIFT) + i);
++ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
++ __pte_ma(0), 0);
++ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i,
++ INVALID_P2M_ENTRY);
+ }
++ if (HYPERVISOR_multicall_check(cr_mcl, i, NULL))
++ BUG();
+
-+ survive:
-+ /*
-+ * If for any reason at all we couldn't handle the fault,
-+ * make sure we exit gracefully rather than endlessly redo
-+ * the fault.
-+ */
-+ switch (handle_mm_fault(mm, vma, address, write)) {
-+ case VM_FAULT_MINOR:
-+ tsk->min_flt++;
-+ break;
-+ case VM_FAULT_MAJOR:
-+ tsk->maj_flt++;
-+ break;
-+ case VM_FAULT_SIGBUS:
-+ goto do_sigbus;
-+ case VM_FAULT_OOM:
-+ goto out_of_memory;
-+ default:
++ /* 2. Get a new contiguous memory extent. */
++ out_frame = __pa(vstart) >> PAGE_SHIFT;
++ rc = HYPERVISOR_memory_op(XENMEM_exchange, &exchange);
++ success = (exchange.nr_exchanged == (1UL << order));
++ BUG_ON(!success && ((exchange.nr_exchanged != 0) || (rc == 0)));
++ BUG_ON(success && (rc != 0));
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ /* Compatibility when XENMEM_exchange is unsupported. */
++ if (HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &exchange.in) != (1UL << order))
+ BUG();
++ success = (HYPERVISOR_memory_op(XENMEM_populate_physmap,
++ &exchange.out) == 1);
++ if (!success) {
++ /* Couldn't get special memory: fall back to normal. */
++ for (i = 0; i < (1U<<order); i++)
++ in_frames[i] = (__pa(vstart)>>PAGE_SHIFT) + i;
++ if (HYPERVISOR_memory_op(XENMEM_populate_physmap,
++ &exchange.in) != (1UL<<order))
++ BUG();
++ }
+ }
++#endif
+
-+ /*
-+ * Did it hit the DOS screen memory VA from vm86 mode?
-+ */
-+ if (regs->eflags & VM_MASK) {
-+ unsigned long bit = (address - 0xA0000) >> PAGE_SHIFT;
-+ if (bit < 32)
-+ tsk->thread.screen_bitmap |= 1 << bit;
++ /* 3. Map the new extent in place of old pages. */
++ for (i = 0; i < (1U<<order); i++) {
++ frame = success ? (out_frame + i) : in_frames[i];
++ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
++ pfn_pte_ma(frame, PAGE_KERNEL), 0);
++ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i, frame);
+ }
-+ up_read(&mm->mmap_sem);
-+ return;
+
-+/*
-+ * Something tried to access memory that isn't in our memory map..
-+ * Fix it, but check if it's kernel or user first..
-+ */
-+bad_area:
-+ up_read(&mm->mmap_sem);
++ cr_mcl[i - 1].args[MULTI_UVMFLAGS_INDEX] = order
++ ? UVMF_TLB_FLUSH|UVMF_ALL
++ : UVMF_INVLPG|UVMF_ALL;
++ if (HYPERVISOR_multicall_check(cr_mcl, i, NULL))
++ BUG();
+
-+bad_area_nosemaphore:
-+ /* User mode accesses just cause a SIGSEGV */
-+ if (error_code & 4) {
-+ /*
-+ * Valid to do another page fault here because this one came
-+ * from user space.
-+ */
-+ if (is_prefetch(regs, address, error_code))
-+ return;
++ if (success)
++ contiguous_bitmap_set(__pa(vstart) >> PAGE_SHIFT,
++ 1UL << order);
+
-+ tsk->thread.cr2 = address;
-+ /* Kernel addresses are always protection faults */
-+ tsk->thread.error_code = error_code | (address >= TASK_SIZE);
-+ tsk->thread.trap_no = 14;
-+ force_sig_info_fault(SIGSEGV, si_code, address, tsk);
-+ return;
-+ }
++ balloon_unlock(flags);
+
-+#ifdef CONFIG_X86_F00F_BUG
-+ /*
-+ * Pentium F0 0F C7 C8 bug workaround.
-+ */
-+ if (boot_cpu_data.f00f_bug) {
-+ unsigned long nr;
-+
-+ nr = (address - idt_descr.address) >> 3;
++ return success ? 0 : -ENOMEM;
++}
++EXPORT_SYMBOL_GPL(xen_create_contiguous_region);
+
-+ if (nr == 6) {
-+ do_invalid_op(regs, 0);
-+ return;
++void xen_destroy_contiguous_region(unsigned long vstart, unsigned int order)
++{
++ unsigned long *out_frames = discontig_frames, in_frame;
++ unsigned long frame, flags;
++ unsigned int i;
++ int rc, success;
++ struct xen_memory_exchange exchange = {
++ .in = {
++ .nr_extents = 1,
++ .extent_order = order,
++ .domid = DOMID_SELF
++ },
++ .out = {
++ .nr_extents = 1UL << order,
++ .extent_order = 0,
++ .domid = DOMID_SELF
+ }
-+ }
-+#endif
++ };
+
-+no_context:
-+ /* Are we prepared to handle this kernel fault? */
-+ if (fixup_exception(regs))
++ if (xen_feature(XENFEAT_auto_translated_physmap) ||
++ !test_bit(__pa(vstart) >> PAGE_SHIFT, contiguous_bitmap))
+ return;
+
-+ /*
-+ * Valid to do another page fault here, because if this fault
-+ * had been triggered by is_prefetch fixup_exception would have
-+ * handled it.
-+ */
-+ if (is_prefetch(regs, address, error_code))
-+ return;
++ if (unlikely(order > MAX_CONTIG_ORDER))
++ return;
+
-+/*
-+ * Oops. The kernel tried to access some bad page. We'll have to
-+ * terminate things with extreme prejudice.
-+ */
++ set_xen_guest_handle(exchange.in.extent_start, &in_frame);
++ set_xen_guest_handle(exchange.out.extent_start, out_frames);
+
-+ bust_spinlocks(1);
++ scrub_pages((void *)vstart, 1 << order);
+
-+ if (oops_may_print()) {
-+ #ifdef CONFIG_X86_PAE
-+ if (error_code & 16) {
-+ pte_t *pte = lookup_address(address);
++ balloon_lock(flags);
+
-+ if (pte && pte_present(*pte) && !pte_exec_kernel(*pte))
-+ printk(KERN_CRIT "kernel tried to execute "
-+ "NX-protected page - exploit attempt? "
-+ "(uid: %d)\n", current->uid);
-+ }
-+ #endif
-+ if (address < PAGE_SIZE)
-+ printk(KERN_ALERT "BUG: unable to handle kernel NULL "
-+ "pointer dereference");
-+ else
-+ printk(KERN_ALERT "BUG: unable to handle kernel paging"
-+ " request");
-+ printk(" at virtual address %08lx\n",address);
-+ printk(KERN_ALERT " printing eip:\n");
-+ printk("%08lx\n", regs->eip);
-+ }
-+ dump_fault_path(address);
-+ tsk->thread.cr2 = address;
-+ tsk->thread.trap_no = 14;
-+ tsk->thread.error_code = error_code;
-+ die("Oops", regs, error_code);
-+ bust_spinlocks(0);
-+ do_exit(SIGKILL);
++ contiguous_bitmap_clear(__pa(vstart) >> PAGE_SHIFT, 1UL << order);
+
-+/*
-+ * We ran out of memory, or some other thing happened to us that made
-+ * us unable to handle the page fault gracefully.
-+ */
-+out_of_memory:
-+ up_read(&mm->mmap_sem);
-+ if (tsk->pid == 1) {
-+ yield();
-+ down_read(&mm->mmap_sem);
-+ goto survive;
++ /* 1. Find start MFN of contiguous extent. */
++ in_frame = pfn_to_mfn(__pa(vstart) >> PAGE_SHIFT);
++
++ /* 2. Zap current PTEs. */
++ for (i = 0; i < (1U<<order); i++) {
++ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
++ __pte_ma(0), 0);
++ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i,
++ INVALID_P2M_ENTRY);
++ out_frames[i] = (__pa(vstart) >> PAGE_SHIFT) + i;
+ }
-+ printk("VM: killing process %s\n", tsk->comm);
-+ if (error_code & 4)
-+ do_exit(SIGKILL);
-+ goto no_context;
++ if (HYPERVISOR_multicall_check(cr_mcl, i, NULL))
++ BUG();
+
-+do_sigbus:
-+ up_read(&mm->mmap_sem);
++ /* 3. Do the exchange for non-contiguous MFNs. */
++ rc = HYPERVISOR_memory_op(XENMEM_exchange, &exchange);
++ success = (exchange.nr_exchanged == 1);
++ BUG_ON(!success && ((exchange.nr_exchanged != 0) || (rc == 0)));
++ BUG_ON(success && (rc != 0));
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ /* Compatibility when XENMEM_exchange is unsupported. */
++ if (HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &exchange.in) != 1)
++ BUG();
++ if (HYPERVISOR_memory_op(XENMEM_populate_physmap,
++ &exchange.out) != (1UL << order))
++ BUG();
++ success = 1;
++ }
++#endif
+
-+ /* Kernel mode? Handle exceptions or die */
-+ if (!(error_code & 4))
-+ goto no_context;
++ /* 4. Map new pages in place of old pages. */
++ for (i = 0; i < (1U<<order); i++) {
++ frame = success ? out_frames[i] : (in_frame + i);
++ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
++ pfn_pte_ma(frame, PAGE_KERNEL), 0);
++ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i, frame);
++ }
+
-+ /* User space => ok to do another page fault */
-+ if (is_prefetch(regs, address, error_code))
-+ return;
++ cr_mcl[i - 1].args[MULTI_UVMFLAGS_INDEX] = order
++ ? UVMF_TLB_FLUSH|UVMF_ALL
++ : UVMF_INVLPG|UVMF_ALL;
++ if (HYPERVISOR_multicall_check(cr_mcl, i, NULL))
++ BUG();
+
-+ tsk->thread.cr2 = address;
-+ tsk->thread.error_code = error_code;
-+ tsk->thread.trap_no = 14;
-+ force_sig_info_fault(SIGBUS, BUS_ADRERR, address, tsk);
++ balloon_unlock(flags);
+}
++EXPORT_SYMBOL_GPL(xen_destroy_contiguous_region);
+
-+#if !HAVE_SHARED_KERNEL_PMD
-+void vmalloc_sync_all(void)
++int xen_limit_pages_to_max_mfn(
++ struct page *pages, unsigned int order, unsigned int address_bits)
+{
-+ /*
-+ * Note that races in the updates of insync and start aren't
-+ * problematic: insync can only get set bits added, and updates to
-+ * start are only improving performance (without affecting correctness
-+ * if undone).
-+ * XEN: To work on PAE, we need to iterate over PMDs rather than PGDs.
-+ * This change works just fine with 2-level paging too.
-+ */
-+#define sync_index(a) ((a) >> PMD_SHIFT)
-+ static DECLARE_BITMAP(insync, PTRS_PER_PGD*PTRS_PER_PMD);
-+ static unsigned long start = TASK_SIZE;
-+ unsigned long address;
-+
-+ BUILD_BUG_ON(TASK_SIZE & ~PGDIR_MASK);
-+ for (address = start;
-+ address >= TASK_SIZE && address < hypervisor_virt_start;
-+ address += 1UL << PMD_SHIFT) {
-+ if (!test_bit(sync_index(address), insync)) {
-+ unsigned long flags;
-+ struct page *page;
++ unsigned long flags, frame;
++ unsigned long *in_frames = discontig_frames, *out_frames = limited_frames;
++ struct page *page;
++ unsigned int i, n, nr_mcl;
++ int rc, success;
++ DECLARE_BITMAP(limit_map, 1 << MAX_CONTIG_ORDER);
+
-+ spin_lock_irqsave(&pgd_lock, flags);
-+ /*
-+ * XEN: vmalloc_sync_one() failure path logic assumes
-+ * pgd_list is non-empty.
-+ */
-+ if (unlikely(!pgd_list)) {
-+ spin_unlock_irqrestore(&pgd_lock, flags);
-+ return;
-+ }
-+ for (page = pgd_list; page; page =
-+ (struct page *)page->index)
-+ if (!vmalloc_sync_one(page_address(page),
-+ address)) {
-+ BUG_ON(page != pgd_list);
-+ break;
-+ }
-+ spin_unlock_irqrestore(&pgd_lock, flags);
-+ if (!page)
-+ set_bit(sync_index(address), insync);
++ struct xen_memory_exchange exchange = {
++ .in = {
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ },
++ .out = {
++ .extent_order = 0,
++ .address_bits = address_bits,
++ .domid = DOMID_SELF
+ }
-+ if (address == start && test_bit(sync_index(address), insync))
-+ start = address + (1UL << PMD_SHIFT);
-+ }
-+}
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/mm/highmem-xen.c tmp-linux-2.6-xen.patch/arch/i386/mm/highmem-xen.c
---- pristine-linux-2.6.18.2/arch/i386/mm/highmem-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/mm/highmem-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,136 @@
-+#include <linux/highmem.h>
-+#include <linux/module.h>
++ };
+
-+void *kmap(struct page *page)
-+{
-+ might_sleep();
-+ if (!PageHighMem(page))
-+ return page_address(page);
-+ return kmap_high(page);
-+}
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return 0;
+
-+void kunmap(struct page *page)
-+{
-+ if (in_interrupt())
-+ BUG();
-+ if (!PageHighMem(page))
-+ return;
-+ kunmap_high(page);
-+}
++ if (unlikely(order > MAX_CONTIG_ORDER))
++ return -ENOMEM;
+
-+/*
-+ * kmap_atomic/kunmap_atomic is significantly faster than kmap/kunmap because
-+ * no global lock is needed and because the kmap code must perform a global TLB
-+ * invalidation when the kmap pool wraps.
-+ *
-+ * However when holding an atomic kmap is is not legal to sleep, so atomic
-+ * kmaps are appropriate for short, tight code paths only.
-+ */
-+static void *__kmap_atomic(struct page *page, enum km_type type, pgprot_t prot)
-+{
-+ enum fixed_addresses idx;
-+ unsigned long vaddr;
++ bitmap_zero(limit_map, 1U << order);
++ set_xen_guest_handle(exchange.in.extent_start, in_frames);
++ set_xen_guest_handle(exchange.out.extent_start, out_frames);
+
-+ /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */
-+ inc_preempt_count();
-+ if (!PageHighMem(page))
-+ return page_address(page);
++ /* 0. Scrub the pages. */
++ for (i = 0, n = 0; i < 1U<<order ; i++) {
++ page = &pages[i];
++ if (!(pfn_to_mfn(page_to_pfn(page)) >> (address_bits - PAGE_SHIFT)))
++ continue;
++ __set_bit(i, limit_map);
+
-+ idx = type + KM_TYPE_NR*smp_processor_id();
-+ vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
-+#ifdef CONFIG_DEBUG_HIGHMEM
-+ if (!pte_none(*(kmap_pte-idx)))
-+ BUG();
++ if (!PageHighMem(page))
++ scrub_pages(page_address(page), 1);
++#ifdef CONFIG_XEN_SCRUB_PAGES
++ else {
++ scrub_pages(kmap(page), 1);
++ kunmap(page);
++ ++n;
++ }
+#endif
-+ set_pte_at_sync(&init_mm, vaddr, kmap_pte-idx, mk_pte(page, prot));
++ }
++ if (bitmap_empty(limit_map, 1U << order))
++ return 0;
+
-+ return (void*) vaddr;
-+}
++ if (n)
++ kmap_flush_unused();
+
-+void *kmap_atomic(struct page *page, enum km_type type)
-+{
-+ return __kmap_atomic(page, type, kmap_prot);
-+}
++ balloon_lock(flags);
+
-+/* Same as kmap_atomic but with PAGE_KERNEL_RO page protection. */
-+void *kmap_atomic_pte(struct page *page, enum km_type type)
-+{
-+ return __kmap_atomic(page, type,
-+ test_bit(PG_pinned, &page->flags)
-+ ? PAGE_KERNEL_RO : kmap_prot);
-+}
++ /* 1. Zap current PTEs (if any), remembering MFNs. */
++ for (i = 0, n = 0, nr_mcl = 0; i < (1U<<order); i++) {
++ if(!test_bit(i, limit_map))
++ continue;
++ page = &pages[i];
+
-+void kunmap_atomic(void *kvaddr, enum km_type type)
-+{
-+#if defined(CONFIG_DEBUG_HIGHMEM) || defined(CONFIG_XEN)
-+ unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
-+ enum fixed_addresses idx = type + KM_TYPE_NR*smp_processor_id();
++ out_frames[n] = page_to_pfn(page);
++ in_frames[n] = pfn_to_mfn(out_frames[n]);
+
-+ if (vaddr < FIXADDR_START) { // FIXME
-+ dec_preempt_count();
-+ preempt_check_resched();
-+ return;
++ if (!PageHighMem(page))
++ MULTI_update_va_mapping(cr_mcl + nr_mcl++,
++ (unsigned long)page_address(page),
++ __pte_ma(0), 0);
++
++ set_phys_to_machine(out_frames[n], INVALID_P2M_ENTRY);
++ ++n;
+ }
-+#endif
++ if (nr_mcl && HYPERVISOR_multicall_check(cr_mcl, nr_mcl, NULL))
++ BUG();
+
-+#if defined(CONFIG_DEBUG_HIGHMEM)
-+ if (vaddr != __fix_to_virt(FIX_KMAP_BEGIN+idx))
-+ BUG();
-+
-+ /*
-+ * force other mappings to Oops if they'll try to access
-+ * this pte without first remap it
-+ */
-+ pte_clear(&init_mm, vaddr, kmap_pte-idx);
-+ __flush_tlb_one(vaddr);
-+#elif defined(CONFIG_XEN)
-+ /*
-+ * We must ensure there are no dangling pagetable references when
-+ * returning memory to Xen (decrease_reservation).
-+ * XXX TODO: We could make this faster by only zapping when
-+ * kmap_flush_unused is called but that is trickier and more invasive.
-+ */
-+ pte_clear(&init_mm, vaddr, kmap_pte-idx);
++ /* 2. Get new memory below the required limit. */
++ exchange.in.nr_extents = n;
++ exchange.out.nr_extents = n;
++ rc = HYPERVISOR_memory_op(XENMEM_exchange, &exchange);
++ success = (exchange.nr_exchanged == n);
++ BUG_ON(!success && ((exchange.nr_exchanged != 0) || (rc == 0)));
++ BUG_ON(success && (rc != 0));
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ /* Compatibility when XENMEM_exchange is unsupported. */
++ if (HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &exchange.in) != n)
++ BUG();
++ if (HYPERVISOR_memory_op(XENMEM_populate_physmap,
++ &exchange.out) != n)
++ BUG();
++ success = 1;
++ }
+#endif
+
-+ dec_preempt_count();
-+ preempt_check_resched();
-+}
++ /* 3. Map the new pages in place of old pages. */
++ for (i = 0, n = 0, nr_mcl = 0; i < (1U<<order); i++) {
++ if(!test_bit(i, limit_map))
++ continue;
++ page = &pages[i];
+
-+/* This is the same as kmap_atomic() but can map memory that doesn't
-+ * have a struct page associated with it.
-+ */
-+void *kmap_atomic_pfn(unsigned long pfn, enum km_type type)
-+{
-+ enum fixed_addresses idx;
-+ unsigned long vaddr;
++ frame = success ? out_frames[n] : in_frames[n];
+
-+ inc_preempt_count();
++ if (!PageHighMem(page))
++ MULTI_update_va_mapping(cr_mcl + nr_mcl++,
++ (unsigned long)page_address(page),
++ pfn_pte_ma(frame, PAGE_KERNEL), 0);
+
-+ idx = type + KM_TYPE_NR*smp_processor_id();
-+ vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
-+ set_pte(kmap_pte-idx, pfn_pte(pfn, kmap_prot));
-+ __flush_tlb_one(vaddr);
++ set_phys_to_machine(page_to_pfn(page), frame);
++ ++n;
++ }
++ if (nr_mcl) {
++ cr_mcl[nr_mcl - 1].args[MULTI_UVMFLAGS_INDEX] = order
++ ? UVMF_TLB_FLUSH|UVMF_ALL
++ : UVMF_INVLPG|UVMF_ALL;
++ if (HYPERVISOR_multicall_check(cr_mcl, nr_mcl, NULL))
++ BUG();
++ }
+
-+ return (void*) vaddr;
++ balloon_unlock(flags);
++
++ return success ? 0 : -ENOMEM;
+}
++EXPORT_SYMBOL_GPL(xen_limit_pages_to_max_mfn);
+
-+struct page *kmap_atomic_to_page(void *ptr)
++#ifdef __i386__
++int write_ldt_entry(void *ldt, int entry, __u32 entry_a, __u32 entry_b)
+{
-+ unsigned long idx, vaddr = (unsigned long)ptr;
-+ pte_t *pte;
++ __u32 *lp = (__u32 *)((char *)ldt + entry * 8);
++ maddr_t mach_lp = arbitrary_virt_to_machine(lp);
++ return HYPERVISOR_update_descriptor(
++ mach_lp, (u64)entry_a | ((u64)entry_b<<32));
++}
++#endif
+
-+ if (vaddr < FIXADDR_START)
-+ return virt_to_page(ptr);
++#define MAX_BATCHED_FULL_PTES 32
+
-+ idx = virt_to_fix(vaddr);
-+ pte = kmap_pte - (idx - FIX_KMAP_BEGIN);
-+ return pte_page(*pte);
-+}
++int xen_change_pte_range(struct mm_struct *mm, pmd_t *pmd,
++ unsigned long addr, unsigned long end, pgprot_t newprot)
++{
++ int rc = 0, i = 0;
++ mmu_update_t u[MAX_BATCHED_FULL_PTES];
++ pte_t *pte;
++ spinlock_t *ptl;
+
-+EXPORT_SYMBOL(kmap);
-+EXPORT_SYMBOL(kunmap);
-+EXPORT_SYMBOL(kmap_atomic);
-+EXPORT_SYMBOL(kmap_atomic_pte);
-+EXPORT_SYMBOL(kunmap_atomic);
-+EXPORT_SYMBOL(kmap_atomic_to_page);
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/mm/hypervisor.c tmp-linux-2.6-xen.patch/arch/i386/mm/hypervisor.c
---- pristine-linux-2.6.18.2/arch/i386/mm/hypervisor.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/mm/hypervisor.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,543 @@
-+/******************************************************************************
-+ * mm/hypervisor.c
-+ *
-+ * Update page tables via the hypervisor.
-+ *
-+ * Copyright (c) 2002-2004, K A Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
++ if (!xen_feature(XENFEAT_mmu_pt_update_preserve_ad))
++ return 0;
++
++ pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
++ do {
++ if (pte_present(*pte)) {
++ u[i].ptr = (__pmd_val(*pmd) & PHYSICAL_PAGE_MASK)
++ | ((unsigned long)pte & ~PAGE_MASK)
++ | MMU_PT_UPDATE_PRESERVE_AD;
++ u[i].val = __pte_val(pte_modify(*pte, newprot));
++ if (++i == MAX_BATCHED_FULL_PTES) {
++ if ((rc = HYPERVISOR_mmu_update(
++ &u[0], i, NULL, DOMID_SELF)) != 0)
++ break;
++ i = 0;
++ }
++ }
++ } while (pte++, addr += PAGE_SIZE, addr != end);
++ if (i)
++ rc = HYPERVISOR_mmu_update( &u[0], i, NULL, DOMID_SELF);
++ pte_unmap_unlock(pte - 1, ptl);
++ BUG_ON(rc && rc != -ENOSYS);
++ return !rc;
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/mm/init-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/mm/init-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,847 @@
++/*
++ * linux/arch/i386/mm/init.c
++ *
++ * Copyright (C) 1995 Linus Torvalds
++ *
++ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
+ */
+
++#include <linux/module.h>
++#include <linux/signal.h>
+#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/string.h>
++#include <linux/types.h>
++#include <linux/ptrace.h>
++#include <linux/mman.h>
+#include <linux/mm.h>
-+#include <linux/vmalloc.h>
-+#include <asm/page.h>
++#include <linux/hugetlb.h>
++#include <linux/swap.h>
++#include <linux/smp.h>
++#include <linux/init.h>
++#include <linux/highmem.h>
++#include <linux/pagemap.h>
++#include <linux/poison.h>
++#include <linux/bootmem.h>
++#include <linux/slab.h>
++#include <linux/proc_fs.h>
++#include <linux/efi.h>
++#include <linux/memory_hotplug.h>
++#include <linux/initrd.h>
++#include <linux/cpumask.h>
++#include <linux/dma-mapping.h>
++#include <linux/scatterlist.h>
++
++#include <asm/processor.h>
++#include <asm/system.h>
++#include <asm/uaccess.h>
+#include <asm/pgtable.h>
-+#include <asm/hypervisor.h>
-+#include <xen/balloon.h>
-+#include <xen/features.h>
-+#include <xen/interface/memory.h>
-+#include <linux/module.h>
-+#include <linux/percpu.h>
++#include <asm/dma.h>
++#include <asm/fixmap.h>
++#include <asm/e820.h>
++#include <asm/apic.h>
++#include <asm/tlb.h>
+#include <asm/tlbflush.h>
-+#include <linux/highmem.h>
++#include <asm/sections.h>
++#include <asm/hypervisor.h>
++#include <asm/swiotlb.h>
+
-+void xen_l1_entry_update(pte_t *ptr, pte_t val)
-+{
-+ mmu_update_t u;
-+ u.ptr = virt_to_machine(ptr);
-+ u.val = __pte_val(val);
-+ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
-+}
++extern unsigned long *contiguous_bitmap;
+
-+void xen_l2_entry_update(pmd_t *ptr, pmd_t val)
-+{
-+ mmu_update_t u;
-+ u.ptr = virt_to_machine(ptr);
-+ u.val = __pmd_val(val);
-+ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
-+}
++unsigned int __VMALLOC_RESERVE = 128 << 20;
+
-+#if defined(CONFIG_X86_PAE) || defined(CONFIG_X86_64)
-+void xen_l3_entry_update(pud_t *ptr, pud_t val)
++DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
++unsigned long highstart_pfn, highend_pfn;
++
++static int noinline do_test_wp_bit(void);
++
++/*
++ * Creates a middle page table and puts a pointer to it in the
++ * given global directory entry. This only returns the gd entry
++ * in non-PAE compilation mode, since the middle layer is folded.
++ */
++static pmd_t * __init one_md_table_init(pgd_t *pgd)
+{
-+ mmu_update_t u;
-+ u.ptr = virt_to_machine(ptr);
-+ u.val = __pud_val(val);
-+ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
-+}
++ pud_t *pud;
++ pmd_t *pmd_table;
++
++#ifdef CONFIG_X86_PAE
++ pmd_table = (pmd_t *) alloc_bootmem_low_pages(PAGE_SIZE);
++ make_lowmem_page_readonly(pmd_table, XENFEAT_writable_page_tables);
++ set_pgd(pgd, __pgd(__pa(pmd_table) | _PAGE_PRESENT));
++ pud = pud_offset(pgd, 0);
++ if (pmd_table != pmd_offset(pud, 0))
++ BUG();
++#else
++ pud = pud_offset(pgd, 0);
++ pmd_table = pmd_offset(pud, 0);
+#endif
+
-+#ifdef CONFIG_X86_64
-+void xen_l4_entry_update(pgd_t *ptr, pgd_t val)
-+{
-+ mmu_update_t u;
-+ u.ptr = virt_to_machine(ptr);
-+ u.val = __pgd_val(val);
-+ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
++ return pmd_table;
+}
-+#endif /* CONFIG_X86_64 */
+
-+void xen_pt_switch(unsigned long ptr)
++/*
++ * Create a page table and place a pointer to it in a middle page
++ * directory entry.
++ */
++static pte_t * __init one_page_table_init(pmd_t *pmd)
+{
-+ struct mmuext_op op;
-+ op.cmd = MMUEXT_NEW_BASEPTR;
-+ op.arg1.mfn = pfn_to_mfn(ptr >> PAGE_SHIFT);
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
-+}
++ if (pmd_none(*pmd)) {
++ pte_t *page_table = (pte_t *) alloc_bootmem_low_pages(PAGE_SIZE);
++ make_lowmem_page_readonly(page_table,
++ XENFEAT_writable_page_tables);
++ set_pmd(pmd, __pmd(__pa(page_table) | _PAGE_TABLE));
++ if (page_table != pte_offset_kernel(pmd, 0))
++ BUG();
+
-+void xen_new_user_pt(unsigned long ptr)
-+{
-+ struct mmuext_op op;
-+ op.cmd = MMUEXT_NEW_USER_BASEPTR;
-+ op.arg1.mfn = pfn_to_mfn(ptr >> PAGE_SHIFT);
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++ return page_table;
++ }
++
++ return pte_offset_kernel(pmd, 0);
+}
+
-+void xen_tlb_flush(void)
-+{
-+ struct mmuext_op op;
-+ op.cmd = MMUEXT_TLB_FLUSH_LOCAL;
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
-+}
-+EXPORT_SYMBOL(xen_tlb_flush);
++/*
++ * This function initializes a certain range of kernel virtual memory
++ * with new bootmem page tables, everywhere page tables are missing in
++ * the given range.
++ */
+
-+void xen_invlpg(unsigned long ptr)
++/*
++ * NOTE: The pagetables are allocated contiguous on the physical space
++ * so we can cache the place of the first one and move around without
++ * checking the pgd every time.
++ */
++static void __init page_table_range_init (unsigned long start, unsigned long end, pgd_t *pgd_base)
+{
-+ struct mmuext_op op;
-+ op.cmd = MMUEXT_INVLPG_LOCAL;
-+ op.arg1.linear_addr = ptr & PAGE_MASK;
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
-+}
-+EXPORT_SYMBOL(xen_invlpg);
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ int pgd_idx, pmd_idx;
++ unsigned long vaddr;
+
-+#ifdef CONFIG_SMP
++ vaddr = start;
++ pgd_idx = pgd_index(vaddr);
++ pmd_idx = pmd_index(vaddr);
++ pgd = pgd_base + pgd_idx;
+
-+void xen_tlb_flush_all(void)
-+{
-+ struct mmuext_op op;
-+ op.cmd = MMUEXT_TLB_FLUSH_ALL;
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
-+}
++ for ( ; (pgd_idx < PTRS_PER_PGD) && (vaddr != end); pgd++, pgd_idx++) {
++ if (pgd_none(*pgd))
++ one_md_table_init(pgd);
++ pud = pud_offset(pgd, vaddr);
++ pmd = pmd_offset(pud, vaddr);
++ for (; (pmd_idx < PTRS_PER_PMD) && (vaddr != end); pmd++, pmd_idx++) {
++ if (vaddr < hypervisor_virt_start && pmd_none(*pmd))
++ one_page_table_init(pmd);
+
-+void xen_tlb_flush_mask(cpumask_t *mask)
-+{
-+ struct mmuext_op op;
-+ if ( cpus_empty(*mask) )
-+ return;
-+ op.cmd = MMUEXT_TLB_FLUSH_MULTI;
-+ set_xen_guest_handle(op.arg2.vcpumask, mask->bits);
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++ vaddr += PMD_SIZE;
++ }
++ pmd_idx = 0;
++ }
+}
+
-+void xen_invlpg_all(unsigned long ptr)
++static inline int is_kernel_text(unsigned long addr)
+{
-+ struct mmuext_op op;
-+ op.cmd = MMUEXT_INVLPG_ALL;
-+ op.arg1.linear_addr = ptr & PAGE_MASK;
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++ if (addr >= PAGE_OFFSET && addr <= (unsigned long)__init_end)
++ return 1;
++ return 0;
+}
+
-+void xen_invlpg_mask(cpumask_t *mask, unsigned long ptr)
++/*
++ * This maps the physical memory to kernel virtual address space, a total
++ * of max_low_pfn pages, by creating page tables starting from address
++ * PAGE_OFFSET.
++ */
++static void __init kernel_physical_mapping_init(pgd_t *pgd_base)
+{
-+ struct mmuext_op op;
-+ if ( cpus_empty(*mask) )
-+ return;
-+ op.cmd = MMUEXT_INVLPG_MULTI;
-+ op.arg1.linear_addr = ptr & PAGE_MASK;
-+ set_xen_guest_handle(op.arg2.vcpumask, mask->bits);
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
-+}
++ unsigned long pfn;
++ pgd_t *pgd;
++ pmd_t *pmd;
++ pte_t *pte;
++ int pgd_idx, pmd_idx, pte_ofs;
+
-+#endif /* CONFIG_SMP */
++ unsigned long max_ram_pfn = xen_start_info->nr_pages;
++ if (max_ram_pfn > max_low_pfn)
++ max_ram_pfn = max_low_pfn;
+
-+void xen_pgd_pin(unsigned long ptr)
-+{
-+ struct mmuext_op op;
-+#ifdef CONFIG_X86_64
-+ op.cmd = MMUEXT_PIN_L4_TABLE;
-+#elif defined(CONFIG_X86_PAE)
-+ op.cmd = MMUEXT_PIN_L3_TABLE;
++ pgd_idx = pgd_index(PAGE_OFFSET);
++ pgd = pgd_base + pgd_idx;
++ pfn = 0;
++ pmd_idx = pmd_index(PAGE_OFFSET);
++ pte_ofs = pte_index(PAGE_OFFSET);
++
++ for (; pgd_idx < PTRS_PER_PGD; pgd++, pgd_idx++) {
++#ifdef CONFIG_XEN
++ /*
++ * Native linux hasn't PAE-paging enabled yet at this
++ * point. When running as xen domain we are in PAE
++ * mode already, thus we can't simply hook a empty
++ * pmd. That would kill the mappings we are currently
++ * using ...
++ */
++ pmd = pmd_offset(pud_offset(pgd, PAGE_OFFSET), PAGE_OFFSET);
+#else
-+ op.cmd = MMUEXT_PIN_L2_TABLE;
++ pmd = one_md_table_init(pgd);
+#endif
-+ op.arg1.mfn = pfn_to_mfn(ptr >> PAGE_SHIFT);
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
-+}
++ if (pfn >= max_low_pfn)
++ continue;
++ pmd += pmd_idx;
++ for (; pmd_idx < PTRS_PER_PMD && pfn < max_low_pfn; pmd++, pmd_idx++) {
++ unsigned int address = pfn * PAGE_SIZE + PAGE_OFFSET;
++ if (address >= hypervisor_virt_start)
++ continue;
+
-+void xen_pgd_unpin(unsigned long ptr)
-+{
-+ struct mmuext_op op;
-+ op.cmd = MMUEXT_UNPIN_TABLE;
-+ op.arg1.mfn = pfn_to_mfn(ptr >> PAGE_SHIFT);
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++ /* Map with big pages if possible, otherwise create normal page tables. */
++ if (cpu_has_pse) {
++ unsigned int address2 = (pfn + PTRS_PER_PTE - 1) * PAGE_SIZE + PAGE_OFFSET + PAGE_SIZE-1;
++
++ if (is_kernel_text(address) || is_kernel_text(address2))
++ set_pmd(pmd, pfn_pmd(pfn, PAGE_KERNEL_LARGE_EXEC));
++ else
++ set_pmd(pmd, pfn_pmd(pfn, PAGE_KERNEL_LARGE));
++ pfn += PTRS_PER_PTE;
++ } else {
++ pte = one_page_table_init(pmd);
++
++ pte += pte_ofs;
++ for (; pte_ofs < PTRS_PER_PTE && pfn < max_low_pfn; pte++, pfn++, pte_ofs++) {
++ /* XEN: Only map initial RAM allocation. */
++ if ((pfn >= max_ram_pfn) || pte_present(*pte))
++ continue;
++ if (is_kernel_text(address))
++ set_pte(pte, pfn_pte(pfn, PAGE_KERNEL_EXEC));
++ else
++ set_pte(pte, pfn_pte(pfn, PAGE_KERNEL));
++ }
++ pte_ofs = 0;
++ }
++ }
++ pmd_idx = 0;
++ }
+}
+
-+void xen_set_ldt(unsigned long ptr, unsigned long len)
++#ifndef CONFIG_XEN
++
++static inline int page_kills_ppro(unsigned long pagenr)
+{
-+ struct mmuext_op op;
-+ op.cmd = MMUEXT_SET_LDT;
-+ op.arg1.linear_addr = ptr;
-+ op.arg2.nr_ents = len;
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++ if (pagenr >= 0x70000 && pagenr <= 0x7003F)
++ return 1;
++ return 0;
+}
+
-+/*
-+ * Bitmap is indexed by page number. If bit is set, the page is part of a
-+ * xen_create_contiguous_region() area of memory.
-+ */
-+unsigned long *contiguous_bitmap;
++#else
+
-+static void contiguous_bitmap_set(
-+ unsigned long first_page, unsigned long nr_pages)
-+{
-+ unsigned long start_off, end_off, curr_idx, end_idx;
++#define page_kills_ppro(p) 0
+
-+ curr_idx = first_page / BITS_PER_LONG;
-+ start_off = first_page & (BITS_PER_LONG-1);
-+ end_idx = (first_page + nr_pages) / BITS_PER_LONG;
-+ end_off = (first_page + nr_pages) & (BITS_PER_LONG-1);
++#endif
+
-+ if (curr_idx == end_idx) {
-+ contiguous_bitmap[curr_idx] |=
-+ ((1UL<<end_off)-1) & -(1UL<<start_off);
-+ } else {
-+ contiguous_bitmap[curr_idx] |= -(1UL<<start_off);
-+ while ( ++curr_idx < end_idx )
-+ contiguous_bitmap[curr_idx] = ~0UL;
-+ contiguous_bitmap[curr_idx] |= (1UL<<end_off)-1;
-+ }
-+}
++extern int is_available_memory(efi_memory_desc_t *);
+
-+static void contiguous_bitmap_clear(
-+ unsigned long first_page, unsigned long nr_pages)
++int page_is_ram(unsigned long pagenr)
+{
-+ unsigned long start_off, end_off, curr_idx, end_idx;
++ int i;
++ unsigned long addr, end;
+
-+ curr_idx = first_page / BITS_PER_LONG;
-+ start_off = first_page & (BITS_PER_LONG-1);
-+ end_idx = (first_page + nr_pages) / BITS_PER_LONG;
-+ end_off = (first_page + nr_pages) & (BITS_PER_LONG-1);
++ if (efi_enabled) {
++ efi_memory_desc_t *md;
++ void *p;
+
-+ if (curr_idx == end_idx) {
-+ contiguous_bitmap[curr_idx] &=
-+ -(1UL<<end_off) | ((1UL<<start_off)-1);
-+ } else {
-+ contiguous_bitmap[curr_idx] &= (1UL<<start_off)-1;
-+ while ( ++curr_idx != end_idx )
-+ contiguous_bitmap[curr_idx] = 0;
-+ contiguous_bitmap[curr_idx] &= -(1UL<<end_off);
++ for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) {
++ md = p;
++ if (!is_available_memory(md))
++ continue;
++ addr = (md->phys_addr+PAGE_SIZE-1) >> PAGE_SHIFT;
++ end = (md->phys_addr + (md->num_pages << EFI_PAGE_SHIFT)) >> PAGE_SHIFT;
++
++ if ((pagenr >= addr) && (pagenr < end))
++ return 1;
++ }
++ return 0;
+ }
-+}
+
-+/* Protected by balloon_lock. */
-+#define MAX_CONTIG_ORDER 9 /* 2MB */
-+static unsigned long discontig_frames[1<<MAX_CONTIG_ORDER];
-+static unsigned long limited_frames[1<<MAX_CONTIG_ORDER];
-+static multicall_entry_t cr_mcl[1<<MAX_CONTIG_ORDER];
++ for (i = 0; i < e820.nr_map; i++) {
+
-+/* Ensure multi-page extents are contiguous in machine memory. */
-+int xen_create_contiguous_region(
-+ unsigned long vstart, unsigned int order, unsigned int address_bits)
-+{
-+ unsigned long *in_frames = discontig_frames, out_frame;
-+ unsigned long frame, i, flags;
-+ long rc;
-+ int success;
-+ struct xen_memory_exchange exchange = {
-+ .in = {
-+ .nr_extents = 1UL << order,
-+ .extent_order = 0,
-+ .domid = DOMID_SELF
-+ },
-+ .out = {
-+ .nr_extents = 1,
-+ .extent_order = order,
-+ .address_bits = address_bits,
-+ .domid = DOMID_SELF
-+ }
-+ };
-+
-+ /*
-+ * Currently an auto-translated guest will not perform I/O, nor will
-+ * it require PAE page directories below 4GB. Therefore any calls to
-+ * this function are redundant and can be ignored.
-+ */
-+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ return 0;
++ if (e820.map[i].type != E820_RAM) /* not usable memory */
++ continue;
++ /*
++ * !!!FIXME!!! Some BIOSen report areas as RAM that
++ * are not. Notably the 640->1Mb area. We need a sanity
++ * check here.
++ */
++ addr = (e820.map[i].addr+PAGE_SIZE-1) >> PAGE_SHIFT;
++ end = (e820.map[i].addr+e820.map[i].size) >> PAGE_SHIFT;
++ if ((pagenr >= addr) && (pagenr < end))
++ return 1;
++ }
++ return 0;
++}
+
-+ if (unlikely(order > MAX_CONTIG_ORDER))
-+ return -ENOMEM;
++#ifdef CONFIG_HIGHMEM
++pte_t *kmap_pte;
++pgprot_t kmap_prot;
+
-+ set_xen_guest_handle(exchange.in.extent_start, in_frames);
-+ set_xen_guest_handle(exchange.out.extent_start, &out_frame);
++#define kmap_get_fixmap_pte(vaddr) \
++ pte_offset_kernel(pmd_offset(pud_offset(pgd_offset_k(vaddr), vaddr), (vaddr)), (vaddr))
+
-+ scrub_pages(vstart, 1 << order);
++static void __init kmap_init(void)
++{
++ unsigned long kmap_vstart;
+
-+ balloon_lock(flags);
++ /* cache the first kmap pte */
++ kmap_vstart = __fix_to_virt(FIX_KMAP_BEGIN);
++ kmap_pte = kmap_get_fixmap_pte(kmap_vstart);
+
-+ /* 1. Zap current PTEs, remembering MFNs. */
-+ for (i = 0; i < (1UL<<order); i++) {
-+ in_frames[i] = pfn_to_mfn((__pa(vstart) >> PAGE_SHIFT) + i);
-+ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
-+ __pte_ma(0), 0);
-+ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i,
-+ INVALID_P2M_ENTRY);
-+ }
-+ if (HYPERVISOR_multicall_check(cr_mcl, i, NULL))
-+ BUG();
++ kmap_prot = PAGE_KERNEL;
++}
+
-+ /* 2. Get a new contiguous memory extent. */
-+ out_frame = __pa(vstart) >> PAGE_SHIFT;
-+ rc = HYPERVISOR_memory_op(XENMEM_exchange, &exchange);
-+ success = (exchange.nr_exchanged == (1UL << order));
-+ BUG_ON(!success && ((exchange.nr_exchanged != 0) || (rc == 0)));
-+ BUG_ON(success && (rc != 0));
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (unlikely(rc == -ENOSYS)) {
-+ /* Compatibility when XENMEM_exchange is unsupported. */
-+ if (HYPERVISOR_memory_op(XENMEM_decrease_reservation,
-+ &exchange.in) != (1UL << order))
-+ BUG();
-+ success = (HYPERVISOR_memory_op(XENMEM_populate_physmap,
-+ &exchange.out) == 1);
-+ if (!success) {
-+ /* Couldn't get special memory: fall back to normal. */
-+ for (i = 0; i < (1UL<<order); i++)
-+ in_frames[i] = (__pa(vstart)>>PAGE_SHIFT) + i;
-+ if (HYPERVISOR_memory_op(XENMEM_populate_physmap,
-+ &exchange.in) != (1UL<<order))
-+ BUG();
-+ }
-+ }
-+#endif
++static void __init permanent_kmaps_init(pgd_t *pgd_base)
++{
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
++ unsigned long vaddr;
+
-+ /* 3. Map the new extent in place of old pages. */
-+ for (i = 0; i < (1UL<<order); i++) {
-+ frame = success ? (out_frame + i) : in_frames[i];
-+ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
-+ pfn_pte_ma(frame, PAGE_KERNEL), 0);
-+ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i, frame);
-+ }
++ vaddr = PKMAP_BASE;
++ page_table_range_init(vaddr, vaddr + PAGE_SIZE*LAST_PKMAP, pgd_base);
+
-+ cr_mcl[i - 1].args[MULTI_UVMFLAGS_INDEX] = order
-+ ? UVMF_TLB_FLUSH|UVMF_ALL
-+ : UVMF_INVLPG|UVMF_ALL;
-+ if (HYPERVISOR_multicall_check(cr_mcl, i, NULL))
-+ BUG();
++ pgd = swapper_pg_dir + pgd_index(vaddr);
++ pud = pud_offset(pgd, vaddr);
++ pmd = pmd_offset(pud, vaddr);
++ pte = pte_offset_kernel(pmd, vaddr);
++ pkmap_page_table = pte;
++}
+
-+ if (success)
-+ contiguous_bitmap_set(__pa(vstart) >> PAGE_SHIFT,
-+ 1UL << order);
++static void __meminit free_new_highpage(struct page *page, int pfn)
++{
++ init_page_count(page);
++ if (pfn < xen_start_info->nr_pages)
++ __free_page(page);
++ totalhigh_pages++;
++}
+
-+ balloon_unlock(flags);
++void __init add_one_highpage_init(struct page *page, int pfn, int bad_ppro)
++{
++ if (page_is_ram(pfn) && !(bad_ppro && page_kills_ppro(pfn))) {
++ ClearPageReserved(page);
++ free_new_highpage(page, pfn);
++ } else
++ SetPageReserved(page);
++}
+
-+ return success ? 0 : -ENOMEM;
++static int add_one_highpage_hotplug(struct page *page, unsigned long pfn)
++{
++ free_new_highpage(page, pfn);
++ totalram_pages++;
++#ifdef CONFIG_FLATMEM
++ max_mapnr = max(pfn, max_mapnr);
++#endif
++ num_physpages++;
++ return 0;
+}
-+EXPORT_SYMBOL_GPL(xen_create_contiguous_region);
+
-+void xen_destroy_contiguous_region(unsigned long vstart, unsigned int order)
++/*
++ * Not currently handling the NUMA case.
++ * Assuming single node and all memory that
++ * has been added dynamically that would be
++ * onlined here is in HIGHMEM
++ */
++void online_page(struct page *page)
+{
-+ unsigned long *out_frames = discontig_frames, in_frame;
-+ unsigned long frame, i, flags;
-+ long rc;
-+ int success;
-+ struct xen_memory_exchange exchange = {
-+ .in = {
-+ .nr_extents = 1,
-+ .extent_order = order,
-+ .domid = DOMID_SELF
-+ },
-+ .out = {
-+ .nr_extents = 1UL << order,
-+ .extent_order = 0,
-+ .domid = DOMID_SELF
-+ }
-+ };
++ ClearPageReserved(page);
++ add_one_highpage_hotplug(page, page_to_pfn(page));
++}
+
-+ if (xen_feature(XENFEAT_auto_translated_physmap) ||
-+ !test_bit(__pa(vstart) >> PAGE_SHIFT, contiguous_bitmap))
-+ return;
+
-+ if (unlikely(order > MAX_CONTIG_ORDER))
-+ return;
++#ifdef CONFIG_NUMA
++extern void set_highmem_pages_init(int);
++#else
++static void __init set_highmem_pages_init(int bad_ppro)
++{
++ int pfn;
++ for (pfn = highstart_pfn; pfn < highend_pfn; pfn++)
++ add_one_highpage_init(pfn_to_page(pfn), pfn, bad_ppro);
++ totalram_pages += totalhigh_pages;
++}
++#endif /* CONFIG_FLATMEM */
+
-+ set_xen_guest_handle(exchange.in.extent_start, &in_frame);
-+ set_xen_guest_handle(exchange.out.extent_start, out_frames);
++#else
++#define kmap_init() do { } while (0)
++#define permanent_kmaps_init(pgd_base) do { } while (0)
++#define set_highmem_pages_init(bad_ppro) do { } while (0)
++#endif /* CONFIG_HIGHMEM */
+
-+ scrub_pages(vstart, 1 << order);
++unsigned long long __PAGE_KERNEL = _PAGE_KERNEL;
++EXPORT_SYMBOL(__PAGE_KERNEL);
++unsigned long long __PAGE_KERNEL_EXEC = _PAGE_KERNEL_EXEC;
+
-+ balloon_lock(flags);
++#ifdef CONFIG_NUMA
++extern void __init remap_numa_kva(void);
++#else
++#define remap_numa_kva() do {} while (0)
++#endif
+
-+ contiguous_bitmap_clear(__pa(vstart) >> PAGE_SHIFT, 1UL << order);
++pgd_t *swapper_pg_dir;
+
-+ /* 1. Find start MFN of contiguous extent. */
-+ in_frame = pfn_to_mfn(__pa(vstart) >> PAGE_SHIFT);
++static void __init pagetable_init (void)
++{
++ unsigned long vaddr;
++ pgd_t *pgd_base = (pgd_t *)xen_start_info->pt_base;
+
-+ /* 2. Zap current PTEs. */
-+ for (i = 0; i < (1UL<<order); i++) {
-+ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
-+ __pte_ma(0), 0);
-+ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i,
-+ INVALID_P2M_ENTRY);
-+ out_frames[i] = (__pa(vstart) >> PAGE_SHIFT) + i;
++ /* Enable PSE if available */
++ if (cpu_has_pse) {
++ set_in_cr4(X86_CR4_PSE);
+ }
-+ if (HYPERVISOR_multicall_check(cr_mcl, i, NULL))
-+ BUG();
+
-+ /* 3. Do the exchange for non-contiguous MFNs. */
-+ rc = HYPERVISOR_memory_op(XENMEM_exchange, &exchange);
-+ success = (exchange.nr_exchanged == 1);
-+ BUG_ON(!success && ((exchange.nr_exchanged != 0) || (rc == 0)));
-+ BUG_ON(success && (rc != 0));
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (unlikely(rc == -ENOSYS)) {
-+ /* Compatibility when XENMEM_exchange is unsupported. */
-+ if (HYPERVISOR_memory_op(XENMEM_decrease_reservation,
-+ &exchange.in) != 1)
-+ BUG();
-+ if (HYPERVISOR_memory_op(XENMEM_populate_physmap,
-+ &exchange.out) != (1UL << order))
-+ BUG();
-+ success = 1;
++ /* Enable PGE if available */
++ if (cpu_has_pge) {
++ set_in_cr4(X86_CR4_PGE);
++ __PAGE_KERNEL |= _PAGE_GLOBAL;
++ __PAGE_KERNEL_EXEC |= _PAGE_GLOBAL;
+ }
-+#endif
+
-+ /* 4. Map new pages in place of old pages. */
-+ for (i = 0; i < (1UL<<order); i++) {
-+ frame = success ? out_frames[i] : (in_frame + i);
-+ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
-+ pfn_pte_ma(frame, PAGE_KERNEL), 0);
-+ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i, frame);
-+ }
++ kernel_physical_mapping_init(pgd_base);
++ remap_numa_kva();
+
-+ cr_mcl[i - 1].args[MULTI_UVMFLAGS_INDEX] = order
-+ ? UVMF_TLB_FLUSH|UVMF_ALL
-+ : UVMF_INVLPG|UVMF_ALL;
-+ if (HYPERVISOR_multicall_check(cr_mcl, i, NULL))
-+ BUG();
++ /*
++ * Fixed mappings, only the page table structure has to be
++ * created - mappings will be set by set_fixmap():
++ */
++ vaddr = __fix_to_virt(__end_of_fixed_addresses - 1) & PMD_MASK;
++ page_table_range_init(vaddr, hypervisor_virt_start, pgd_base);
+
-+ balloon_unlock(flags);
++ permanent_kmaps_init(pgd_base);
+}
-+EXPORT_SYMBOL_GPL(xen_destroy_contiguous_region);
+
-+int xen_limit_pages_to_max_mfn(
-+ struct page *pages, unsigned int order, unsigned int address_bits)
-+{
-+ unsigned long flags, frame;
-+ unsigned long *in_frames = discontig_frames, *out_frames = limited_frames;
-+ void *v;
-+ struct page *page;
-+ int i, nr_mcl, rc, success;
++#if defined(CONFIG_SOFTWARE_SUSPEND) || defined(CONFIG_ACPI_SLEEP)
++/*
++ * Swap suspend & friends need this for resume because things like the intel-agp
++ * driver might have split up a kernel 4MB mapping.
++ */
++char __nosavedata swsusp_pg_dir[PAGE_SIZE]
++ __attribute__ ((aligned (PAGE_SIZE)));
+
-+ struct xen_memory_exchange exchange = {
-+ .in = {
-+ .nr_extents = 1UL << order,
-+ .extent_order = 0,
-+ .domid = DOMID_SELF
-+ },
-+ .out = {
-+ .nr_extents = 1UL << order,
-+ .extent_order = 0,
-+ .address_bits = address_bits,
-+ .domid = DOMID_SELF
-+ }
-+ };
++static inline void save_pg_dir(void)
++{
++ memcpy(swsusp_pg_dir, swapper_pg_dir, PAGE_SIZE);
++}
++#else
++static inline void save_pg_dir(void)
++{
++}
++#endif
+
-+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ return 0;
++void zap_low_mappings (void)
++{
++ int i;
+
-+ if (unlikely(order > MAX_CONTIG_ORDER))
-+ return -ENOMEM;
++ save_pg_dir();
+
-+ set_xen_guest_handle(exchange.in.extent_start, in_frames);
-+ set_xen_guest_handle(exchange.out.extent_start, out_frames);
++ /*
++ * Zap initial low-memory mappings.
++ *
++ * Note that "pgd_clear()" doesn't do it for
++ * us, because pgd_clear() is a no-op on i386.
++ */
++ for (i = 0; i < USER_PTRS_PER_PGD; i++)
++#if defined(CONFIG_X86_PAE) && !defined(CONFIG_XEN)
++ set_pgd(swapper_pg_dir+i, __pgd(1 + __pa(empty_zero_page)));
++#else
++ set_pgd(swapper_pg_dir+i, __pgd(0));
++#endif
++ flush_tlb_all();
++}
+
-+ /* 0. Scrub the pages. */
-+ for ( i = 0 ; i < 1UL<<order ; i++ ) {
-+ page = &pages[i];
++static int disable_nx __initdata = 0;
++u64 __supported_pte_mask __read_mostly = ~_PAGE_NX;
++EXPORT_SYMBOL(__supported_pte_mask);
+
-+ if (!PageHighMem(page)) {
-+ v = page_address(page);
-+ scrub_pages(v, 1);
-+ } else {
-+ v = kmap(page);
-+ scrub_pages(v, 1);
-+ kunmap(page);
-+ }
++/*
++ * noexec = on|off
++ *
++ * Control non executable mappings.
++ *
++ * on Enable
++ * off Disable
++ */
++void __init noexec_setup(const char *str)
++{
++ if (!strncmp(str, "on",2) && cpu_has_nx) {
++ __supported_pte_mask |= _PAGE_NX;
++ disable_nx = 0;
++ } else if (!strncmp(str,"off",3)) {
++ disable_nx = 1;
++ __supported_pte_mask &= ~_PAGE_NX;
+ }
++}
+
-+ kmap_flush_unused();
-+
-+ balloon_lock(flags);
-+
-+ /* 1. Zap current PTEs (if any), remembering MFNs. */
-+ for (i = 0, nr_mcl = 0; i < (1UL<<order); i++) {
-+ page = &pages[i];
-+
-+ out_frames[i] = page_to_pfn(page);
-+ in_frames[i] = pfn_to_mfn(out_frames[i]);
-+
-+ if (!PageHighMem(page))
-+ MULTI_update_va_mapping(cr_mcl + nr_mcl++,
-+ (unsigned long)page_address(page),
-+ __pte_ma(0), 0);
++int nx_enabled = 0;
++#ifdef CONFIG_X86_PAE
+
-+ set_phys_to_machine(out_frames[i], INVALID_P2M_ENTRY);
-+ }
-+ if (HYPERVISOR_multicall_check(cr_mcl, nr_mcl, NULL))
-+ BUG();
++static void __init set_nx(void)
++{
++ unsigned int v[4], l, h;
+
-+ /* 2. Get new memory below the required limit. */
-+ rc = HYPERVISOR_memory_op(XENMEM_exchange, &exchange);
-+ success = (exchange.nr_exchanged == (1UL << order));
-+ BUG_ON(!success && ((exchange.nr_exchanged != 0) || (rc == 0)));
-+ BUG_ON(success && (rc != 0));
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (unlikely(rc == -ENOSYS)) {
-+ /* Compatibility when XENMEM_exchange is unsupported. */
-+ if (HYPERVISOR_memory_op(XENMEM_decrease_reservation,
-+ &exchange.in) != (1UL << order))
-+ BUG();
-+ success = (HYPERVISOR_memory_op(XENMEM_populate_physmap,
-+ &exchange.out) != (1UL <<order));
++ if (cpu_has_pae && (cpuid_eax(0x80000000) > 0x80000001)) {
++ cpuid(0x80000001, &v[0], &v[1], &v[2], &v[3]);
++ if ((v[3] & (1 << 20)) && !disable_nx) {
++ rdmsr(MSR_EFER, l, h);
++ l |= EFER_NX;
++ wrmsr(MSR_EFER, l, h);
++ nx_enabled = 1;
++ __supported_pte_mask |= _PAGE_NX;
++ }
+ }
-+#endif
-+
-+ /* 3. Map the new pages in place of old pages. */
-+ for (i = 0, nr_mcl = 0; i < (1UL<<order); i++) {
-+ page = &pages[i];
-+ unsigned long pfn = page_to_pfn(page);
++}
+
-+ frame = success ? out_frames[i] : in_frames[i];
++/*
++ * Enables/disables executability of a given kernel page and
++ * returns the previous setting.
++ */
++int __init set_kernel_exec(unsigned long vaddr, int enable)
++{
++ pte_t *pte;
++ int ret = 1;
+
-+ if (!PageHighMem(page))
-+ MULTI_update_va_mapping(cr_mcl + nr_mcl++,
-+ (unsigned long)page_address(page),
-+ pfn_pte_ma(frame, PAGE_KERNEL), 0);
++ if (!nx_enabled)
++ goto out;
+
-+ set_phys_to_machine(pfn, frame);
-+ }
-+ cr_mcl[nr_mcl - 1].args[MULTI_UVMFLAGS_INDEX] = order
-+ ? UVMF_TLB_FLUSH|UVMF_ALL
-+ : UVMF_INVLPG|UVMF_ALL;
-+ if (HYPERVISOR_multicall_check(cr_mcl, nr_mcl, NULL))
-+ BUG();
++ pte = lookup_address(vaddr);
++ BUG_ON(!pte);
+
-+ balloon_unlock(flags);
++ if (!pte_exec_kernel(*pte))
++ ret = 0;
+
-+ return success ? 0 : -ENOMEM;
++ if (enable)
++ pte->pte_high &= ~(1 << (_PAGE_BIT_NX - 32));
++ else
++ pte->pte_high |= 1 << (_PAGE_BIT_NX - 32);
++ __flush_tlb_all();
++out:
++ return ret;
+}
-+EXPORT_SYMBOL_GPL(xen_limit_pages_to_max_mfn);
+
-+#ifdef __i386__
-+int write_ldt_entry(void *ldt, int entry, __u32 entry_a, __u32 entry_b)
-+{
-+ __u32 *lp = (__u32 *)((char *)ldt + entry * 8);
-+ maddr_t mach_lp = arbitrary_virt_to_machine(lp);
-+ return HYPERVISOR_update_descriptor(
-+ mach_lp, (u64)entry_a | ((u64)entry_b<<32));
-+}
+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/mm/init-xen.c tmp-linux-2.6-xen.patch/arch/i386/mm/init-xen.c
---- pristine-linux-2.6.18.2/arch/i386/mm/init-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/mm/init-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,850 @@
++
+/*
-+ * linux/arch/i386/mm/init.c
-+ *
-+ * Copyright (C) 1995 Linus Torvalds
++ * paging_init() sets up the page tables - note that the first 8MB are
++ * already mapped by head.S.
+ *
-+ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
++ * This routines also unmaps the page at virtual kernel address 0, so
++ * that we can trap those pesky NULL-reference errors in the kernel.
+ */
++void __init paging_init(void)
++{
++ int i;
+
-+#include <linux/module.h>
-+#include <linux/signal.h>
-+#include <linux/sched.h>
-+#include <linux/kernel.h>
-+#include <linux/errno.h>
-+#include <linux/string.h>
-+#include <linux/types.h>
-+#include <linux/ptrace.h>
-+#include <linux/mman.h>
-+#include <linux/mm.h>
-+#include <linux/hugetlb.h>
-+#include <linux/swap.h>
-+#include <linux/smp.h>
-+#include <linux/init.h>
-+#include <linux/highmem.h>
-+#include <linux/pagemap.h>
-+#include <linux/poison.h>
-+#include <linux/bootmem.h>
-+#include <linux/slab.h>
-+#include <linux/proc_fs.h>
-+#include <linux/efi.h>
-+#include <linux/memory_hotplug.h>
-+#include <linux/initrd.h>
-+#include <linux/cpumask.h>
-+#include <linux/dma-mapping.h>
-+#include <linux/scatterlist.h>
++#ifdef CONFIG_X86_PAE
++ set_nx();
++ if (nx_enabled)
++ printk("NX (Execute Disable) protection: active\n");
++#endif
+
-+#include <asm/processor.h>
-+#include <asm/system.h>
-+#include <asm/uaccess.h>
-+#include <asm/pgtable.h>
-+#include <asm/dma.h>
-+#include <asm/fixmap.h>
-+#include <asm/e820.h>
-+#include <asm/apic.h>
-+#include <asm/tlb.h>
-+#include <asm/tlbflush.h>
-+#include <asm/sections.h>
-+#include <asm/hypervisor.h>
-+#include <asm/swiotlb.h>
++ pagetable_init();
+
-+extern unsigned long *contiguous_bitmap;
++#if defined(CONFIG_X86_PAE) && !defined(CONFIG_XEN)
++ /*
++ * We will bail out later - printk doesn't work right now so
++ * the user would just see a hanging kernel.
++ * when running as xen domain we are already in PAE mode at
++ * this point.
++ */
++ if (cpu_has_pae)
++ set_in_cr4(X86_CR4_PAE);
++#endif
++ __flush_tlb_all();
+
-+unsigned int __VMALLOC_RESERVE = 128 << 20;
++ kmap_init();
+
-+DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
-+unsigned long highstart_pfn, highend_pfn;
++ /* Switch to the real shared_info page, and clear the
++ * dummy page. */
++ set_fixmap(FIX_SHARED_INFO, xen_start_info->shared_info);
++ HYPERVISOR_shared_info = (shared_info_t *)fix_to_virt(FIX_SHARED_INFO);
++ memset(empty_zero_page, 0, sizeof(empty_zero_page));
+
-+static int noinline do_test_wp_bit(void);
++ /* Setup mapping of lower 1st MB */
++ for (i = 0; i < NR_FIX_ISAMAPS; i++)
++ if (is_initial_xendomain())
++ set_fixmap(FIX_ISAMAP_BEGIN - i, i * PAGE_SIZE);
++ else
++ __set_fixmap(FIX_ISAMAP_BEGIN - i,
++ virt_to_machine(empty_zero_page),
++ PAGE_KERNEL_RO);
++}
+
+/*
-+ * Creates a middle page table and puts a pointer to it in the
-+ * given global directory entry. This only returns the gd entry
-+ * in non-PAE compilation mode, since the middle layer is folded.
-+ */
-+static pmd_t * __init one_md_table_init(pgd_t *pgd)
-+{
-+ pud_t *pud;
-+ pmd_t *pmd_table;
-+
-+#ifdef CONFIG_X86_PAE
-+ pmd_table = (pmd_t *) alloc_bootmem_low_pages(PAGE_SIZE);
-+ make_lowmem_page_readonly(pmd_table, XENFEAT_writable_page_tables);
-+ set_pgd(pgd, __pgd(__pa(pmd_table) | _PAGE_PRESENT));
-+ pud = pud_offset(pgd, 0);
-+ if (pmd_table != pmd_offset(pud, 0))
-+ BUG();
-+#else
-+ pud = pud_offset(pgd, 0);
-+ pmd_table = pmd_offset(pud, 0);
-+#endif
-+
-+ return pmd_table;
-+}
-+
-+/*
-+ * Create a page table and place a pointer to it in a middle page
-+ * directory entry.
-+ */
-+static pte_t * __init one_page_table_init(pmd_t *pmd)
-+{
-+ if (pmd_none(*pmd)) {
-+ pte_t *page_table = (pte_t *) alloc_bootmem_low_pages(PAGE_SIZE);
-+ make_lowmem_page_readonly(page_table,
-+ XENFEAT_writable_page_tables);
-+ set_pmd(pmd, __pmd(__pa(page_table) | _PAGE_TABLE));
-+ if (page_table != pte_offset_kernel(pmd, 0))
-+ BUG();
-+
-+ return page_table;
-+ }
-+
-+ return pte_offset_kernel(pmd, 0);
-+}
-+
-+/*
-+ * This function initializes a certain range of kernel virtual memory
-+ * with new bootmem page tables, everywhere page tables are missing in
-+ * the given range.
++ * Test if the WP bit works in supervisor mode. It isn't supported on 386's
++ * and also on some strange 486's (NexGen etc.). All 586+'s are OK. This
++ * used to involve black magic jumps to work around some nasty CPU bugs,
++ * but fortunately the switch to using exceptions got rid of all that.
+ */
+
-+/*
-+ * NOTE: The pagetables are allocated contiguous on the physical space
-+ * so we can cache the place of the first one and move around without
-+ * checking the pgd every time.
-+ */
-+static void __init page_table_range_init (unsigned long start, unsigned long end, pgd_t *pgd_base)
++static void __init test_wp_bit(void)
+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ int pgd_idx, pmd_idx;
-+ unsigned long vaddr;
-+
-+ vaddr = start;
-+ pgd_idx = pgd_index(vaddr);
-+ pmd_idx = pmd_index(vaddr);
-+ pgd = pgd_base + pgd_idx;
++ printk("Checking if this processor honours the WP bit even in supervisor mode... ");
+
-+ for ( ; (pgd_idx < PTRS_PER_PGD) && (vaddr != end); pgd++, pgd_idx++) {
-+ if (pgd_none(*pgd))
-+ one_md_table_init(pgd);
-+ pud = pud_offset(pgd, vaddr);
-+ pmd = pmd_offset(pud, vaddr);
-+ for (; (pmd_idx < PTRS_PER_PMD) && (vaddr != end); pmd++, pmd_idx++) {
-+ if (vaddr < hypervisor_virt_start && pmd_none(*pmd))
-+ one_page_table_init(pmd);
++ /* Any page-aligned address will do, the test is non-destructive */
++ __set_fixmap(FIX_WP_TEST, __pa(&swapper_pg_dir), PAGE_READONLY);
++ boot_cpu_data.wp_works_ok = do_test_wp_bit();
++ clear_fixmap(FIX_WP_TEST);
+
-+ vaddr += PMD_SIZE;
-+ }
-+ pmd_idx = 0;
++ if (!boot_cpu_data.wp_works_ok) {
++ printk("No.\n");
++#ifdef CONFIG_X86_WP_WORKS_OK
++ panic("This kernel doesn't support CPU's with broken WP. Recompile it for a 386!");
++#endif
++ } else {
++ printk("Ok.\n");
+ }
+}
+
-+static inline int is_kernel_text(unsigned long addr)
++static void __init set_max_mapnr_init(void)
+{
-+ if (addr >= PAGE_OFFSET && addr <= (unsigned long)__init_end)
-+ return 1;
-+ return 0;
++#ifdef CONFIG_HIGHMEM
++ num_physpages = highend_pfn;
++#else
++ num_physpages = max_low_pfn;
++#endif
++#ifdef CONFIG_FLATMEM
++ max_mapnr = num_physpages;
++#endif
+}
+
-+/*
-+ * This maps the physical memory to kernel virtual address space, a total
-+ * of max_low_pfn pages, by creating page tables starting from address
-+ * PAGE_OFFSET.
-+ */
-+static void __init kernel_physical_mapping_init(pgd_t *pgd_base)
++static struct kcore_list kcore_mem, kcore_vmalloc;
++
++void __init mem_init(void)
+{
++ extern int ppro_with_ram_bug(void);
++ int codesize, reservedpages, datasize, initsize;
++ int tmp;
++ int bad_ppro;
+ unsigned long pfn;
-+ pgd_t *pgd;
-+ pmd_t *pmd;
-+ pte_t *pte;
-+ int pgd_idx, pmd_idx, pte_ofs;
-+
-+ unsigned long max_ram_pfn = xen_start_info->nr_pages;
-+ if (max_ram_pfn > max_low_pfn)
-+ max_ram_pfn = max_low_pfn;
+
-+ pgd_idx = pgd_index(PAGE_OFFSET);
-+ pgd = pgd_base + pgd_idx;
-+ pfn = 0;
-+ pmd_idx = pmd_index(PAGE_OFFSET);
-+ pte_ofs = pte_index(PAGE_OFFSET);
++ contiguous_bitmap = alloc_bootmem_low_pages(
++ (max_low_pfn + 2*BITS_PER_LONG) >> 3);
++ BUG_ON(!contiguous_bitmap);
++ memset(contiguous_bitmap, 0, (max_low_pfn + 2*BITS_PER_LONG) >> 3);
+
-+ for (; pgd_idx < PTRS_PER_PGD; pgd++, pgd_idx++) {
-+#ifdef CONFIG_XEN
-+ /*
-+ * Native linux hasn't PAE-paging enabled yet at this
-+ * point. When running as xen domain we are in PAE
-+ * mode already, thus we can't simply hook a empty
-+ * pmd. That would kill the mappings we are currently
-+ * using ...
-+ */
-+ pmd = pmd_offset(pud_offset(pgd, PAGE_OFFSET), PAGE_OFFSET);
-+#else
-+ pmd = one_md_table_init(pgd);
++#if defined(CONFIG_SWIOTLB)
++ swiotlb_init();
+#endif
-+ if (pfn >= max_low_pfn)
-+ continue;
-+ pmd += pmd_idx;
-+ for (; pmd_idx < PTRS_PER_PMD && pfn < max_low_pfn; pmd++, pmd_idx++) {
-+ unsigned int address = pfn * PAGE_SIZE + PAGE_OFFSET;
-+ if (address >= hypervisor_virt_start)
-+ continue;
-+
-+ /* Map with big pages if possible, otherwise create normal page tables. */
-+ if (cpu_has_pse) {
-+ unsigned int address2 = (pfn + PTRS_PER_PTE - 1) * PAGE_SIZE + PAGE_OFFSET + PAGE_SIZE-1;
+
-+ if (is_kernel_text(address) || is_kernel_text(address2))
-+ set_pmd(pmd, pfn_pmd(pfn, PAGE_KERNEL_LARGE_EXEC));
-+ else
-+ set_pmd(pmd, pfn_pmd(pfn, PAGE_KERNEL_LARGE));
-+ pfn += PTRS_PER_PTE;
-+ } else {
-+ pte = one_page_table_init(pmd);
++#ifdef CONFIG_FLATMEM
++ if (!mem_map)
++ BUG();
++#endif
++
++ bad_ppro = ppro_with_ram_bug();
+
-+ pte += pte_ofs;
-+ for (; pte_ofs < PTRS_PER_PTE && pfn < max_low_pfn; pte++, pfn++, pte_ofs++) {
-+ /* XEN: Only map initial RAM allocation. */
-+ if ((pfn >= max_ram_pfn) || pte_present(*pte))
-+ continue;
-+ if (is_kernel_text(address))
-+ set_pte(pte, pfn_pte(pfn, PAGE_KERNEL_EXEC));
-+ else
-+ set_pte(pte, pfn_pte(pfn, PAGE_KERNEL));
-+ }
-+ pte_ofs = 0;
-+ }
-+ }
-+ pmd_idx = 0;
++#ifdef CONFIG_HIGHMEM
++ /* check that fixmap and pkmap do not overlap */
++ if (PKMAP_BASE+LAST_PKMAP*PAGE_SIZE >= FIXADDR_START) {
++ printk(KERN_ERR "fixmap and kmap areas overlap - this will crash\n");
++ printk(KERN_ERR "pkstart: %lxh pkend: %lxh fixstart %lxh\n",
++ PKMAP_BASE, PKMAP_BASE+LAST_PKMAP*PAGE_SIZE, FIXADDR_START);
++ BUG();
+ }
-+}
-+
-+#ifndef CONFIG_XEN
-+
-+static inline int page_kills_ppro(unsigned long pagenr)
-+{
-+ if (pagenr >= 0x70000 && pagenr <= 0x7003F)
-+ return 1;
-+ return 0;
-+}
++#endif
++
++ set_max_mapnr_init();
+
++#ifdef CONFIG_HIGHMEM
++ high_memory = (void *) __va(highstart_pfn * PAGE_SIZE - 1) + 1;
+#else
-+
-+#define page_kills_ppro(p) 0
-+
++ high_memory = (void *) __va(max_low_pfn * PAGE_SIZE - 1) + 1;
+#endif
-+
-+extern int is_available_memory(efi_memory_desc_t *);
-+
-+int page_is_ram(unsigned long pagenr)
-+{
-+ int i;
-+ unsigned long addr, end;
-+
-+ if (efi_enabled) {
-+ efi_memory_desc_t *md;
-+ void *p;
-+
-+ for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) {
-+ md = p;
-+ if (!is_available_memory(md))
-+ continue;
-+ addr = (md->phys_addr+PAGE_SIZE-1) >> PAGE_SHIFT;
-+ end = (md->phys_addr + (md->num_pages << EFI_PAGE_SHIFT)) >> PAGE_SHIFT;
-+
-+ if ((pagenr >= addr) && (pagenr < end))
-+ return 1;
-+ }
-+ return 0;
++ printk("vmalloc area: %lx-%lx, maxmem %lx\n",
++ VMALLOC_START,VMALLOC_END,MAXMEM);
++ BUG_ON(VMALLOC_START > VMALLOC_END);
++
++ /* this will put all low memory onto the freelists */
++ totalram_pages += free_all_bootmem();
++ /* XEN: init and count low-mem pages outside initial allocation. */
++ for (pfn = xen_start_info->nr_pages; pfn < max_low_pfn; pfn++) {
++ ClearPageReserved(pfn_to_page(pfn));
++ init_page_count(pfn_to_page(pfn));
++ totalram_pages++;
+ }
+
-+ for (i = 0; i < e820.nr_map; i++) {
-+
-+ if (e820.map[i].type != E820_RAM) /* not usable memory */
-+ continue;
++ reservedpages = 0;
++ for (tmp = 0; tmp < max_low_pfn; tmp++)
+ /*
-+ * !!!FIXME!!! Some BIOSen report areas as RAM that
-+ * are not. Notably the 640->1Mb area. We need a sanity
-+ * check here.
++ * Only count reserved RAM pages
+ */
-+ addr = (e820.map[i].addr+PAGE_SIZE-1) >> PAGE_SHIFT;
-+ end = (e820.map[i].addr+e820.map[i].size) >> PAGE_SHIFT;
-+ if ((pagenr >= addr) && (pagenr < end))
-+ return 1;
-+ }
-+ return 0;
-+}
-+
-+#ifdef CONFIG_HIGHMEM
-+pte_t *kmap_pte;
-+pgprot_t kmap_prot;
++ if (page_is_ram(tmp) && PageReserved(pfn_to_page(tmp)))
++ reservedpages++;
+
-+#define kmap_get_fixmap_pte(vaddr) \
-+ pte_offset_kernel(pmd_offset(pud_offset(pgd_offset_k(vaddr), vaddr), (vaddr)), (vaddr))
++ set_highmem_pages_init(bad_ppro);
+
-+static void __init kmap_init(void)
-+{
-+ unsigned long kmap_vstart;
++ codesize = (unsigned long) &_etext - (unsigned long) &_text;
++ datasize = (unsigned long) &_edata - (unsigned long) &_etext;
++ initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin;
+
-+ /* cache the first kmap pte */
-+ kmap_vstart = __fix_to_virt(FIX_KMAP_BEGIN);
-+ kmap_pte = kmap_get_fixmap_pte(kmap_vstart);
++ kclist_add(&kcore_mem, __va(0), max_low_pfn << PAGE_SHIFT);
++ kclist_add(&kcore_vmalloc, (void *)VMALLOC_START,
++ VMALLOC_END-VMALLOC_START);
+
-+ kmap_prot = PAGE_KERNEL;
-+}
++ printk(KERN_INFO "Memory: %luk/%luk available (%dk kernel code, %dk reserved, %dk data, %dk init, %ldk highmem)\n",
++ (unsigned long) nr_free_pages() << (PAGE_SHIFT-10),
++ num_physpages << (PAGE_SHIFT-10),
++ codesize >> 10,
++ reservedpages << (PAGE_SHIFT-10),
++ datasize >> 10,
++ initsize >> 10,
++ (unsigned long) (totalhigh_pages << (PAGE_SHIFT-10))
++ );
+
-+static void __init permanent_kmaps_init(pgd_t *pgd_base)
-+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte;
-+ unsigned long vaddr;
++#ifdef CONFIG_X86_PAE
++ if (!cpu_has_pae)
++ panic("cannot execute a PAE-enabled kernel on a PAE-less CPU!");
++#endif
++ if (boot_cpu_data.wp_works_ok < 0)
++ test_wp_bit();
+
-+ vaddr = PKMAP_BASE;
-+ page_table_range_init(vaddr, vaddr + PAGE_SIZE*LAST_PKMAP, pgd_base);
++ /*
++ * Subtle. SMP is doing it's boot stuff late (because it has to
++ * fork idle threads) - but it also needs low mappings for the
++ * protected-mode entry to work. We zap these entries only after
++ * the WP-bit has been tested.
++ */
++#ifndef CONFIG_SMP
++ zap_low_mappings();
++#endif
+
-+ pgd = swapper_pg_dir + pgd_index(vaddr);
-+ pud = pud_offset(pgd, vaddr);
-+ pmd = pmd_offset(pud, vaddr);
-+ pte = pte_offset_kernel(pmd, vaddr);
-+ pkmap_page_table = pte;
++ set_bit(PG_pinned, &virt_to_page(init_mm.pgd)->flags);
+}
+
-+static void __meminit free_new_highpage(struct page *page, int pfn)
++/*
++ * this is for the non-NUMA, single node SMP system case.
++ * Specifically, in the case of x86, we will always add
++ * memory to the highmem for now.
++ */
++#ifdef CONFIG_MEMORY_HOTPLUG
++#ifndef CONFIG_NEED_MULTIPLE_NODES
++int arch_add_memory(int nid, u64 start, u64 size)
+{
-+ init_page_count(page);
-+ if (pfn < xen_start_info->nr_pages)
-+ __free_page(page);
-+ totalhigh_pages++;
++ struct pglist_data *pgdata = &contig_page_data;
++ struct zone *zone = pgdata->node_zones + MAX_NR_ZONES-1;
++ unsigned long start_pfn = start >> PAGE_SHIFT;
++ unsigned long nr_pages = size >> PAGE_SHIFT;
++
++ return __add_pages(zone, start_pfn, nr_pages);
+}
+
-+void __init add_one_highpage_init(struct page *page, int pfn, int bad_ppro)
++int remove_memory(u64 start, u64 size)
+{
-+ if (page_is_ram(pfn) && !(bad_ppro && page_kills_ppro(pfn))) {
-+ ClearPageReserved(page);
-+ free_new_highpage(page, pfn);
-+ } else
-+ SetPageReserved(page);
++ return -EINVAL;
+}
++#endif
++#endif
+
-+static int add_one_highpage_hotplug(struct page *page, unsigned long pfn)
++kmem_cache_t *pgd_cache;
++kmem_cache_t *pmd_cache;
++
++void __init pgtable_cache_init(void)
+{
-+ free_new_highpage(page, pfn);
-+ totalram_pages++;
-+#ifdef CONFIG_FLATMEM
-+ max_mapnr = max(pfn, max_mapnr);
++ if (PTRS_PER_PMD > 1) {
++ pmd_cache = kmem_cache_create("pmd",
++ PTRS_PER_PMD*sizeof(pmd_t),
++ PTRS_PER_PMD*sizeof(pmd_t),
++ 0,
++ pmd_ctor,
++ NULL);
++ if (!pmd_cache)
++ panic("pgtable_cache_init(): cannot create pmd cache");
++ }
++ pgd_cache = kmem_cache_create("pgd",
++#ifndef CONFIG_XEN
++ PTRS_PER_PGD*sizeof(pgd_t),
++ PTRS_PER_PGD*sizeof(pgd_t),
++#else
++ PAGE_SIZE,
++ PAGE_SIZE,
+#endif
-+ num_physpages++;
-+ return 0;
++ 0,
++ pgd_ctor,
++ PTRS_PER_PMD == 1 ? pgd_dtor : NULL);
++ if (!pgd_cache)
++ panic("pgtable_cache_init(): Cannot create pgd cache");
+}
+
+/*
-+ * Not currently handling the NUMA case.
-+ * Assuming single node and all memory that
-+ * has been added dynamically that would be
-+ * onlined here is in HIGHMEM
++ * This function cannot be __init, since exceptions don't work in that
++ * section. Put this after the callers, so that it cannot be inlined.
+ */
-+void online_page(struct page *page)
++static int noinline do_test_wp_bit(void)
+{
-+ ClearPageReserved(page);
-+ add_one_highpage_hotplug(page, page_to_pfn(page));
++ char tmp_reg;
++ int flag;
++
++ __asm__ __volatile__(
++ " movb %0,%1 \n"
++ "1: movb %1,%0 \n"
++ " xorl %2,%2 \n"
++ "2: \n"
++ ".section __ex_table,\"a\"\n"
++ " .align 4 \n"
++ " .long 1b,2b \n"
++ ".previous \n"
++ :"=m" (*(char *)fix_to_virt(FIX_WP_TEST)),
++ "=q" (tmp_reg),
++ "=r" (flag)
++ :"2" (1)
++ :"memory");
++
++ return flag;
+}
+
++#ifdef CONFIG_DEBUG_RODATA
+
-+#ifdef CONFIG_NUMA
-+extern void set_highmem_pages_init(int);
-+#else
-+static void __init set_highmem_pages_init(int bad_ppro)
++void mark_rodata_ro(void)
+{
-+ int pfn;
-+ for (pfn = highstart_pfn; pfn < highend_pfn; pfn++)
-+ add_one_highpage_init(pfn_to_page(pfn), pfn, bad_ppro);
-+ totalram_pages += totalhigh_pages;
-+}
-+#endif /* CONFIG_FLATMEM */
++ unsigned long addr = (unsigned long)__start_rodata;
+
-+#else
-+#define kmap_init() do { } while (0)
-+#define permanent_kmaps_init(pgd_base) do { } while (0)
-+#define set_highmem_pages_init(bad_ppro) do { } while (0)
-+#endif /* CONFIG_HIGHMEM */
++ for (; addr < (unsigned long)__end_rodata; addr += PAGE_SIZE)
++ change_page_attr(virt_to_page(addr), 1, PAGE_KERNEL_RO);
+
-+unsigned long long __PAGE_KERNEL = _PAGE_KERNEL;
-+EXPORT_SYMBOL(__PAGE_KERNEL);
-+unsigned long long __PAGE_KERNEL_EXEC = _PAGE_KERNEL_EXEC;
++ printk("Write protecting the kernel read-only data: %uk\n",
++ (__end_rodata - __start_rodata) >> 10);
+
-+#ifdef CONFIG_NUMA
-+extern void __init remap_numa_kva(void);
-+#else
-+#define remap_numa_kva() do {} while (0)
++ /*
++ * change_page_attr() requires a global_flush_tlb() call after it.
++ * We do this after the printk so that if something went wrong in the
++ * change, the printk gets out at least to give a better debug hint
++ * of who is the culprit.
++ */
++ global_flush_tlb();
++}
+#endif
+
-+pgd_t *swapper_pg_dir;
-+
-+static void __init pagetable_init (void)
++void free_init_pages(char *what, unsigned long begin, unsigned long end)
+{
-+ unsigned long vaddr;
-+ pgd_t *pgd_base = (pgd_t *)xen_start_info->pt_base;
-+
-+ swapper_pg_dir = pgd_base;
-+ init_mm.pgd = pgd_base;
-+
-+ /* Enable PSE if available */
-+ if (cpu_has_pse) {
-+ set_in_cr4(X86_CR4_PSE);
-+ }
++ unsigned long addr;
+
-+ /* Enable PGE if available */
-+ if (cpu_has_pge) {
-+ set_in_cr4(X86_CR4_PGE);
-+ __PAGE_KERNEL |= _PAGE_GLOBAL;
-+ __PAGE_KERNEL_EXEC |= _PAGE_GLOBAL;
++ for (addr = begin; addr < end; addr += PAGE_SIZE) {
++ ClearPageReserved(virt_to_page(addr));
++ init_page_count(virt_to_page(addr));
++ memset((void *)addr, POISON_FREE_INITMEM, PAGE_SIZE);
++ free_page(addr);
++ totalram_pages++;
+ }
-+
-+ kernel_physical_mapping_init(pgd_base);
-+ remap_numa_kva();
-+
-+ /*
-+ * Fixed mappings, only the page table structure has to be
-+ * created - mappings will be set by set_fixmap():
-+ */
-+ vaddr = __fix_to_virt(__end_of_fixed_addresses - 1) & PMD_MASK;
-+ page_table_range_init(vaddr, hypervisor_virt_start, pgd_base);
-+
-+ permanent_kmaps_init(pgd_base);
++ printk(KERN_INFO "Freeing %s: %ldk freed\n", what, (end - begin) >> 10);
+}
+
-+#if defined(CONFIG_SOFTWARE_SUSPEND) || defined(CONFIG_ACPI_SLEEP)
-+/*
-+ * Swap suspend & friends need this for resume because things like the intel-agp
-+ * driver might have split up a kernel 4MB mapping.
-+ */
-+char __nosavedata swsusp_pg_dir[PAGE_SIZE]
-+ __attribute__ ((aligned (PAGE_SIZE)));
-+
-+static inline void save_pg_dir(void)
++void free_initmem(void)
+{
-+ memcpy(swsusp_pg_dir, swapper_pg_dir, PAGE_SIZE);
++ free_init_pages("unused kernel memory",
++ (unsigned long)(&__init_begin),
++ (unsigned long)(&__init_end));
+}
-+#else
-+static inline void save_pg_dir(void)
++
++#ifdef CONFIG_BLK_DEV_INITRD
++void free_initrd_mem(unsigned long start, unsigned long end)
+{
++ free_init_pages("initrd memory", start, end);
+}
+#endif
+
-+void zap_low_mappings (void)
-+{
-+ int i;
-+
-+ save_pg_dir();
-+
-+ /*
-+ * Zap initial low-memory mappings.
-+ *
-+ * Note that "pgd_clear()" doesn't do it for
-+ * us, because pgd_clear() is a no-op on i386.
-+ */
-+ for (i = 0; i < USER_PTRS_PER_PGD; i++)
-+#if defined(CONFIG_X86_PAE) && !defined(CONFIG_XEN)
-+ set_pgd(swapper_pg_dir+i, __pgd(1 + __pa(empty_zero_page)));
-+#else
-+ set_pgd(swapper_pg_dir+i, __pgd(0));
-+#endif
-+ flush_tlb_all();
-+}
-+
-+static int disable_nx __initdata = 0;
-+u64 __supported_pte_mask __read_mostly = ~_PAGE_NX;
-+EXPORT_SYMBOL(__supported_pte_mask);
-+
-+/*
-+ * noexec = on|off
-+ *
-+ * Control non executable mappings.
-+ *
-+ * on Enable
-+ * off Disable
-+ */
-+void __init noexec_setup(const char *str)
-+{
-+ if (!strncmp(str, "on",2) && cpu_has_nx) {
-+ __supported_pte_mask |= _PAGE_NX;
-+ disable_nx = 0;
-+ } else if (!strncmp(str,"off",3)) {
-+ disable_nx = 1;
-+ __supported_pte_mask &= ~_PAGE_NX;
-+ }
-+}
-+
-+int nx_enabled = 0;
-+#ifdef CONFIG_X86_PAE
-+
-+static void __init set_nx(void)
-+{
-+ unsigned int v[4], l, h;
-+
-+ if (cpu_has_pae && (cpuid_eax(0x80000000) > 0x80000001)) {
-+ cpuid(0x80000001, &v[0], &v[1], &v[2], &v[3]);
-+ if ((v[3] & (1 << 20)) && !disable_nx) {
-+ rdmsr(MSR_EFER, l, h);
-+ l |= EFER_NX;
-+ wrmsr(MSR_EFER, l, h);
-+ nx_enabled = 1;
-+ __supported_pte_mask |= _PAGE_NX;
-+ }
-+ }
-+}
-+
-+/*
-+ * Enables/disables executability of a given kernel page and
-+ * returns the previous setting.
-+ */
-+int __init set_kernel_exec(unsigned long vaddr, int enable)
-+{
-+ pte_t *pte;
-+ int ret = 1;
-+
-+ if (!nx_enabled)
-+ goto out;
-+
-+ pte = lookup_address(vaddr);
-+ BUG_ON(!pte);
-+
-+ if (!pte_exec_kernel(*pte))
-+ ret = 0;
-+
-+ if (enable)
-+ pte->pte_high &= ~(1 << (_PAGE_BIT_NX - 32));
-+ else
-+ pte->pte_high |= 1 << (_PAGE_BIT_NX - 32);
-+ __flush_tlb_all();
-+out:
-+ return ret;
-+}
-+
-+#endif
-+
-+/*
-+ * paging_init() sets up the page tables - note that the first 8MB are
-+ * already mapped by head.S.
-+ *
-+ * This routines also unmaps the page at virtual kernel address 0, so
-+ * that we can trap those pesky NULL-reference errors in the kernel.
-+ */
-+void __init paging_init(void)
-+{
-+ int i;
-+
-+#ifdef CONFIG_X86_PAE
-+ set_nx();
-+ if (nx_enabled)
-+ printk("NX (Execute Disable) protection: active\n");
-+#endif
-+
-+ pagetable_init();
-+
-+#if defined(CONFIG_X86_PAE) && !defined(CONFIG_XEN)
-+ /*
-+ * We will bail out later - printk doesn't work right now so
-+ * the user would just see a hanging kernel.
-+ * when running as xen domain we are already in PAE mode at
-+ * this point.
-+ */
-+ if (cpu_has_pae)
-+ set_in_cr4(X86_CR4_PAE);
-+#endif
-+ __flush_tlb_all();
-+
-+ kmap_init();
-+
-+ /* Switch to the real shared_info page, and clear the
-+ * dummy page. */
-+ set_fixmap(FIX_SHARED_INFO, xen_start_info->shared_info);
-+ HYPERVISOR_shared_info = (shared_info_t *)fix_to_virt(FIX_SHARED_INFO);
-+ memset(empty_zero_page, 0, sizeof(empty_zero_page));
-+
-+ /* Setup mapping of lower 1st MB */
-+ for (i = 0; i < NR_FIX_ISAMAPS; i++)
-+ if (is_initial_xendomain())
-+ set_fixmap(FIX_ISAMAP_BEGIN - i, i * PAGE_SIZE);
-+ else
-+ __set_fixmap(FIX_ISAMAP_BEGIN - i,
-+ virt_to_machine(empty_zero_page),
-+ PAGE_KERNEL_RO);
-+}
-+
-+/*
-+ * Test if the WP bit works in supervisor mode. It isn't supported on 386's
-+ * and also on some strange 486's (NexGen etc.). All 586+'s are OK. This
-+ * used to involve black magic jumps to work around some nasty CPU bugs,
-+ * but fortunately the switch to using exceptions got rid of all that.
-+ */
-+
-+static void __init test_wp_bit(void)
-+{
-+ printk("Checking if this processor honours the WP bit even in supervisor mode... ");
-+
-+ /* Any page-aligned address will do, the test is non-destructive */
-+ __set_fixmap(FIX_WP_TEST, __pa(&swapper_pg_dir), PAGE_READONLY);
-+ boot_cpu_data.wp_works_ok = do_test_wp_bit();
-+ clear_fixmap(FIX_WP_TEST);
-+
-+ if (!boot_cpu_data.wp_works_ok) {
-+ printk("No.\n");
-+#ifdef CONFIG_X86_WP_WORKS_OK
-+ panic("This kernel doesn't support CPU's with broken WP. Recompile it for a 386!");
-+#endif
-+ } else {
-+ printk("Ok.\n");
-+ }
-+}
-+
-+static void __init set_max_mapnr_init(void)
-+{
-+#ifdef CONFIG_HIGHMEM
-+ num_physpages = highend_pfn;
-+#else
-+ num_physpages = max_low_pfn;
-+#endif
-+#ifdef CONFIG_FLATMEM
-+ max_mapnr = num_physpages;
-+#endif
-+}
-+
-+static struct kcore_list kcore_mem, kcore_vmalloc;
-+
-+void __init mem_init(void)
-+{
-+ extern int ppro_with_ram_bug(void);
-+ int codesize, reservedpages, datasize, initsize;
-+ int tmp;
-+ int bad_ppro;
-+ unsigned long pfn;
-+
-+ contiguous_bitmap = alloc_bootmem_low_pages(
-+ (max_low_pfn + 2*BITS_PER_LONG) >> 3);
-+ BUG_ON(!contiguous_bitmap);
-+ memset(contiguous_bitmap, 0, (max_low_pfn + 2*BITS_PER_LONG) >> 3);
-+
-+#if defined(CONFIG_SWIOTLB)
-+ swiotlb_init();
-+#endif
-+
-+#ifdef CONFIG_FLATMEM
-+ if (!mem_map)
-+ BUG();
-+#endif
-+
-+ bad_ppro = ppro_with_ram_bug();
-+
-+#ifdef CONFIG_HIGHMEM
-+ /* check that fixmap and pkmap do not overlap */
-+ if (PKMAP_BASE+LAST_PKMAP*PAGE_SIZE >= FIXADDR_START) {
-+ printk(KERN_ERR "fixmap and kmap areas overlap - this will crash\n");
-+ printk(KERN_ERR "pkstart: %lxh pkend: %lxh fixstart %lxh\n",
-+ PKMAP_BASE, PKMAP_BASE+LAST_PKMAP*PAGE_SIZE, FIXADDR_START);
-+ BUG();
-+ }
-+#endif
-+
-+ set_max_mapnr_init();
-+
-+#ifdef CONFIG_HIGHMEM
-+ high_memory = (void *) __va(highstart_pfn * PAGE_SIZE - 1) + 1;
-+#else
-+ high_memory = (void *) __va(max_low_pfn * PAGE_SIZE - 1) + 1;
-+#endif
-+ printk("vmalloc area: %lx-%lx, maxmem %lx\n",
-+ VMALLOC_START,VMALLOC_END,MAXMEM);
-+ BUG_ON(VMALLOC_START > VMALLOC_END);
-+
-+ /* this will put all low memory onto the freelists */
-+ totalram_pages += free_all_bootmem();
-+ /* XEN: init and count low-mem pages outside initial allocation. */
-+ for (pfn = xen_start_info->nr_pages; pfn < max_low_pfn; pfn++) {
-+ ClearPageReserved(pfn_to_page(pfn));
-+ init_page_count(pfn_to_page(pfn));
-+ totalram_pages++;
-+ }
-+
-+ reservedpages = 0;
-+ for (tmp = 0; tmp < max_low_pfn; tmp++)
-+ /*
-+ * Only count reserved RAM pages
-+ */
-+ if (page_is_ram(tmp) && PageReserved(pfn_to_page(tmp)))
-+ reservedpages++;
-+
-+ set_highmem_pages_init(bad_ppro);
-+
-+ codesize = (unsigned long) &_etext - (unsigned long) &_text;
-+ datasize = (unsigned long) &_edata - (unsigned long) &_etext;
-+ initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin;
-+
-+ kclist_add(&kcore_mem, __va(0), max_low_pfn << PAGE_SHIFT);
-+ kclist_add(&kcore_vmalloc, (void *)VMALLOC_START,
-+ VMALLOC_END-VMALLOC_START);
-+
-+ printk(KERN_INFO "Memory: %luk/%luk available (%dk kernel code, %dk reserved, %dk data, %dk init, %ldk highmem)\n",
-+ (unsigned long) nr_free_pages() << (PAGE_SHIFT-10),
-+ num_physpages << (PAGE_SHIFT-10),
-+ codesize >> 10,
-+ reservedpages << (PAGE_SHIFT-10),
-+ datasize >> 10,
-+ initsize >> 10,
-+ (unsigned long) (totalhigh_pages << (PAGE_SHIFT-10))
-+ );
-+
-+#ifdef CONFIG_X86_PAE
-+ if (!cpu_has_pae)
-+ panic("cannot execute a PAE-enabled kernel on a PAE-less CPU!");
-+#endif
-+ if (boot_cpu_data.wp_works_ok < 0)
-+ test_wp_bit();
-+
-+ /*
-+ * Subtle. SMP is doing it's boot stuff late (because it has to
-+ * fork idle threads) - but it also needs low mappings for the
-+ * protected-mode entry to work. We zap these entries only after
-+ * the WP-bit has been tested.
-+ */
-+#ifndef CONFIG_SMP
-+ zap_low_mappings();
-+#endif
-+
-+ set_bit(PG_pinned, &virt_to_page(init_mm.pgd)->flags);
-+}
-+
-+/*
-+ * this is for the non-NUMA, single node SMP system case.
-+ * Specifically, in the case of x86, we will always add
-+ * memory to the highmem for now.
-+ */
-+#ifdef CONFIG_MEMORY_HOTPLUG
-+#ifndef CONFIG_NEED_MULTIPLE_NODES
-+int arch_add_memory(int nid, u64 start, u64 size)
-+{
-+ struct pglist_data *pgdata = &contig_page_data;
-+ struct zone *zone = pgdata->node_zones + MAX_NR_ZONES-1;
-+ unsigned long start_pfn = start >> PAGE_SHIFT;
-+ unsigned long nr_pages = size >> PAGE_SHIFT;
-+
-+ return __add_pages(zone, start_pfn, nr_pages);
-+}
-+
-+int remove_memory(u64 start, u64 size)
-+{
-+ return -EINVAL;
-+}
-+#endif
-+#endif
-+
-+kmem_cache_t *pgd_cache;
-+kmem_cache_t *pmd_cache;
-+
-+void __init pgtable_cache_init(void)
-+{
-+ if (PTRS_PER_PMD > 1) {
-+ pmd_cache = kmem_cache_create("pmd",
-+ PTRS_PER_PMD*sizeof(pmd_t),
-+ PTRS_PER_PMD*sizeof(pmd_t),
-+ 0,
-+ pmd_ctor,
-+ NULL);
-+ if (!pmd_cache)
-+ panic("pgtable_cache_init(): cannot create pmd cache");
-+ }
-+ pgd_cache = kmem_cache_create("pgd",
-+#ifndef CONFIG_XEN
-+ PTRS_PER_PGD*sizeof(pgd_t),
-+ PTRS_PER_PGD*sizeof(pgd_t),
-+#else
-+ PAGE_SIZE,
-+ PAGE_SIZE,
-+#endif
-+ 0,
-+ pgd_ctor,
-+ PTRS_PER_PMD == 1 ? pgd_dtor : NULL);
-+ if (!pgd_cache)
-+ panic("pgtable_cache_init(): Cannot create pgd cache");
-+}
-+
-+/*
-+ * This function cannot be __init, since exceptions don't work in that
-+ * section. Put this after the callers, so that it cannot be inlined.
-+ */
-+static int noinline do_test_wp_bit(void)
-+{
-+ char tmp_reg;
-+ int flag;
-+
-+ __asm__ __volatile__(
-+ " movb %0,%1 \n"
-+ "1: movb %1,%0 \n"
-+ " xorl %2,%2 \n"
-+ "2: \n"
-+ ".section __ex_table,\"a\"\n"
-+ " .align 4 \n"
-+ " .long 1b,2b \n"
-+ ".previous \n"
-+ :"=m" (*(char *)fix_to_virt(FIX_WP_TEST)),
-+ "=q" (tmp_reg),
-+ "=r" (flag)
-+ :"2" (1)
-+ :"memory");
-+
-+ return flag;
-+}
-+
-+#ifdef CONFIG_DEBUG_RODATA
-+
-+void mark_rodata_ro(void)
-+{
-+ unsigned long addr = (unsigned long)__start_rodata;
-+
-+ for (; addr < (unsigned long)__end_rodata; addr += PAGE_SIZE)
-+ change_page_attr(virt_to_page(addr), 1, PAGE_KERNEL_RO);
-+
-+ printk("Write protecting the kernel read-only data: %uk\n",
-+ (__end_rodata - __start_rodata) >> 10);
-+
-+ /*
-+ * change_page_attr() requires a global_flush_tlb() call after it.
-+ * We do this after the printk so that if something went wrong in the
-+ * change, the printk gets out at least to give a better debug hint
-+ * of who is the culprit.
-+ */
-+ global_flush_tlb();
-+}
-+#endif
-+
-+void free_init_pages(char *what, unsigned long begin, unsigned long end)
-+{
-+ unsigned long addr;
-+
-+ for (addr = begin; addr < end; addr += PAGE_SIZE) {
-+ ClearPageReserved(virt_to_page(addr));
-+ init_page_count(virt_to_page(addr));
-+ memset((void *)addr, POISON_FREE_INITMEM, PAGE_SIZE);
-+ free_page(addr);
-+ totalram_pages++;
-+ }
-+ printk(KERN_INFO "Freeing %s: %ldk freed\n", what, (end - begin) >> 10);
-+}
-+
-+void free_initmem(void)
-+{
-+ free_init_pages("unused kernel memory",
-+ (unsigned long)(&__init_begin),
-+ (unsigned long)(&__init_end));
-+}
-+
-+#ifdef CONFIG_BLK_DEV_INITRD
-+void free_initrd_mem(unsigned long start, unsigned long end)
-+{
-+ free_init_pages("initrd memory", start, end);
-+}
-+#endif
-+
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/mm/ioremap-xen.c tmp-linux-2.6-xen.patch/arch/i386/mm/ioremap-xen.c
---- pristine-linux-2.6.18.2/arch/i386/mm/ioremap-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/mm/ioremap-xen.c 2007-09-30 18:06:18.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/mm/ioremap-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/mm/ioremap-xen.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,443 @@
+/*
+ * arch/i386/mm/ioremap.c
@@ -19426,7 +19858,7 @@
+ * Fill in the machine address: PTE ptr is done later by
+ * apply_to_page_range().
+ */
-+ v->val = __pte_val(pfn_pte_ma(mfn, prot));
++ v->val = __pte_val(pfn_pte_ma(mfn, prot)) | _PAGE_IO;
+
+ mfn++;
+ address += PAGE_SIZE;
@@ -19791,25 +20223,10 @@
+ --nrpages;
+ }
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/mm/Makefile tmp-linux-2.6-xen.patch/arch/i386/mm/Makefile
---- pristine-linux-2.6.18.2/arch/i386/mm/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/mm/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -8,3 +8,11 @@ obj-$(CONFIG_NUMA) += discontig.o
- obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o
- obj-$(CONFIG_HIGHMEM) += highmem.o
- obj-$(CONFIG_BOOT_IOREMAP) += boot_ioremap.o
-+
-+ifdef CONFIG_XEN
-+include $(srctree)/scripts/Makefile.xen
-+
-+obj-y += hypervisor.o
-+
-+obj-y := $(call cherrypickxen, $(obj-y))
-+endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/mm/pageattr.c tmp-linux-2.6-xen.patch/arch/i386/mm/pageattr.c
---- pristine-linux-2.6.18.2/arch/i386/mm/pageattr.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/mm/pageattr.c 2007-10-14 01:51:15.000000000 +0200
-@@ -84,7 +84,7 @@ static void set_pmd_pte(pte_t *kpte, uns
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/mm/pageattr.c
+--- a/arch/i386/mm/pageattr.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/mm/pageattr.c Wed Sep 10 10:54:08 2008 +0100
+@@ -84,7 +84,7 @@
unsigned long flags;
set_pte_atomic(kpte, pte); /* change init_mm */
@@ -19818,158 +20235,50 @@
return;
spin_lock_irqsave(&pgd_lock, flags);
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/mm/pgtable.c tmp-linux-2.6-xen.patch/arch/i386/mm/pgtable.c
---- pristine-linux-2.6.18.2/arch/i386/mm/pgtable.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/mm/pgtable.c 2007-10-14 01:51:15.000000000 +0200
-@@ -12,6 +12,7 @@
- #include <linux/slab.h>
- #include <linux/pagemap.h>
- #include <linux/spinlock.h>
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/mm/pgtable-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/mm/pgtable-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,725 @@
++/*
++ * linux/arch/i386/mm/pgtable.c
++ */
++
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/mm.h>
++#include <linux/swap.h>
++#include <linux/smp.h>
++#include <linux/highmem.h>
++#include <linux/slab.h>
++#include <linux/pagemap.h>
++#include <linux/spinlock.h>
+#include <linux/module.h>
-
- #include <asm/system.h>
- #include <asm/pgtable.h>
-@@ -137,6 +138,10 @@ void set_pmd_pfn(unsigned long vaddr, un
- __flush_tlb_one(vaddr);
- }
-
-+static int nr_fixmaps = 0;
-+unsigned long __FIXADDR_TOP = 0xfffff000;
-+EXPORT_SYMBOL(__FIXADDR_TOP);
+
- void __set_fixmap (enum fixed_addresses idx, unsigned long phys, pgprot_t flags)
- {
- unsigned long address = __fix_to_virt(idx);
-@@ -146,6 +151,13 @@ void __set_fixmap (enum fixed_addresses
- return;
- }
- set_pte_pfn(address, phys >> PAGE_SHIFT, flags);
-+ nr_fixmaps++;
-+}
++#include <asm/system.h>
++#include <asm/pgtable.h>
++#include <asm/pgalloc.h>
++#include <asm/fixmap.h>
++#include <asm/e820.h>
++#include <asm/tlb.h>
++#include <asm/tlbflush.h>
++#include <asm/io.h>
++#include <asm/mmu_context.h>
+
-+void set_fixaddr_top(unsigned long top)
-+{
-+ BUG_ON(nr_fixmaps > 0);
-+ __FIXADDR_TOP = top - PAGE_SIZE;
- }
-
- pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
-@@ -214,9 +226,10 @@ void pgd_ctor(void *pgd, kmem_cache_t *c
- spin_lock_irqsave(&pgd_lock, flags);
- }
-
-- clone_pgd_range((pgd_t *)pgd + USER_PTRS_PER_PGD,
-- swapper_pg_dir + USER_PTRS_PER_PGD,
-- KERNEL_PGD_PTRS);
-+ if (PTRS_PER_PMD == 1 || HAVE_SHARED_KERNEL_PMD)
-+ clone_pgd_range((pgd_t *)pgd + USER_PTRS_PER_PGD,
-+ swapper_pg_dir + USER_PTRS_PER_PGD,
-+ KERNEL_PGD_PTRS);
- if (PTRS_PER_PMD > 1)
- return;
-
-@@ -248,6 +261,30 @@ pgd_t *pgd_alloc(struct mm_struct *mm)
- goto out_oom;
- set_pgd(&pgd[i], __pgd(1 + __pa(pmd)));
- }
++#include <xen/features.h>
++#include <asm/hypervisor.h>
+
-+ if (!HAVE_SHARED_KERNEL_PMD) {
-+ unsigned long flags;
++static void pgd_test_and_unpin(pgd_t *pgd);
+
-+ for (i = USER_PTRS_PER_PGD; i < PTRS_PER_PGD; i++) {
-+ pmd_t *pmd = kmem_cache_alloc(pmd_cache, GFP_KERNEL);
-+ if (!pmd)
-+ goto out_oom;
-+ set_pgd(&pgd[USER_PTRS_PER_PGD], __pgd(1 + __pa(pmd)));
-+ }
-+
-+ spin_lock_irqsave(&pgd_lock, flags);
-+ for (i = USER_PTRS_PER_PGD; i < PTRS_PER_PGD; i++) {
-+ unsigned long v = (unsigned long)i << PGDIR_SHIFT;
-+ pgd_t *kpgd = pgd_offset_k(v);
-+ pud_t *kpud = pud_offset(kpgd, v);
-+ pmd_t *kpmd = pmd_offset(kpud, v);
-+ pmd_t *pmd = (void *)__va(pgd_val(pgd[i])-1);
-+ memcpy(pmd, kpmd, PAGE_SIZE);
-+ }
-+ pgd_list_add(pgd);
-+ spin_unlock_irqrestore(&pgd_lock, flags);
-+ }
-+
- return pgd;
-
- out_oom:
-@@ -262,9 +299,23 @@ void pgd_free(pgd_t *pgd)
- int i;
-
- /* in the PAE case user pgd entries are overwritten before usage */
-- if (PTRS_PER_PMD > 1)
-- for (i = 0; i < USER_PTRS_PER_PGD; ++i)
-- kmem_cache_free(pmd_cache, (void *)__va(pgd_val(pgd[i])-1));
-+ if (PTRS_PER_PMD > 1) {
-+ for (i = 0; i < USER_PTRS_PER_PGD; ++i) {
-+ pmd_t *pmd = (void *)__va(pgd_val(pgd[i])-1);
-+ kmem_cache_free(pmd_cache, pmd);
-+ }
-+ if (!HAVE_SHARED_KERNEL_PMD) {
-+ unsigned long flags;
-+ spin_lock_irqsave(&pgd_lock, flags);
-+ pgd_list_del(pgd);
-+ spin_unlock_irqrestore(&pgd_lock, flags);
-+ for (i = USER_PTRS_PER_PGD; i < PTRS_PER_PGD; i++) {
-+ pmd_t *pmd = (void *)__va(pgd_val(pgd[i])-1);
-+ memset(pmd, 0, PTRS_PER_PMD*sizeof(pmd_t));
-+ kmem_cache_free(pmd_cache, pmd);
-+ }
-+ }
-+ }
- /* in the non-PAE case, free_pgtables() clears user pgd entries */
- kmem_cache_free(pgd_cache, pgd);
- }
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/mm/pgtable-xen.c tmp-linux-2.6-xen.patch/arch/i386/mm/pgtable-xen.c
---- pristine-linux-2.6.18.2/arch/i386/mm/pgtable-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/mm/pgtable-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,727 @@
-+/*
-+ * linux/arch/i386/mm/pgtable.c
-+ */
-+
-+#include <linux/sched.h>
-+#include <linux/kernel.h>
-+#include <linux/errno.h>
-+#include <linux/mm.h>
-+#include <linux/swap.h>
-+#include <linux/smp.h>
-+#include <linux/highmem.h>
-+#include <linux/slab.h>
-+#include <linux/pagemap.h>
-+#include <linux/spinlock.h>
-+#include <linux/module.h>
-+
-+#include <asm/system.h>
-+#include <asm/pgtable.h>
-+#include <asm/pgalloc.h>
-+#include <asm/fixmap.h>
-+#include <asm/e820.h>
-+#include <asm/tlb.h>
-+#include <asm/tlbflush.h>
-+#include <asm/io.h>
-+#include <asm/mmu_context.h>
-+
-+#include <xen/features.h>
-+#include <asm/hypervisor.h>
-+
-+static void pgd_test_and_unpin(pgd_t *pgd);
-+
-+void show_mem(void)
-+{
-+ int total = 0, reserved = 0;
-+ int shared = 0, cached = 0;
-+ int highmem = 0;
-+ struct page *page;
-+ pg_data_t *pgdat;
-+ unsigned long i;
-+ unsigned long flags;
++void show_mem(void)
++{
++ int total = 0, reserved = 0;
++ int shared = 0, cached = 0;
++ int highmem = 0;
++ struct page *page;
++ pg_data_t *pgdat;
++ unsigned long i;
++ unsigned long flags;
+
+ printk(KERN_INFO "Mem-info:\n");
+ show_free_areas();
@@ -20006,87 +20315,6 @@
+}
+
+/*
-+ * Associate a virtual page frame with a given physical page frame
-+ * and protection flags for that frame.
-+ */
-+static void set_pte_pfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags)
-+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte;
-+
-+ pgd = swapper_pg_dir + pgd_index(vaddr);
-+ if (pgd_none(*pgd)) {
-+ BUG();
-+ return;
-+ }
-+ pud = pud_offset(pgd, vaddr);
-+ if (pud_none(*pud)) {
-+ BUG();
-+ return;
-+ }
-+ pmd = pmd_offset(pud, vaddr);
-+ if (pmd_none(*pmd)) {
-+ BUG();
-+ return;
-+ }
-+ pte = pte_offset_kernel(pmd, vaddr);
-+ if (pgprot_val(flags))
-+ /* <pfn,flags> stored as-is, to permit clearing entries */
-+ set_pte(pte, pfn_pte(pfn, flags));
-+ else
-+ pte_clear(&init_mm, vaddr, pte);
-+
-+ /*
-+ * It's enough to flush this one mapping.
-+ * (PGE mappings get flushed as well)
-+ */
-+ __flush_tlb_one(vaddr);
-+}
-+
-+/*
-+ * Associate a virtual page frame with a given physical page frame
-+ * and protection flags for that frame.
-+ */
-+static void set_pte_pfn_ma(unsigned long vaddr, unsigned long pfn,
-+ pgprot_t flags)
-+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte;
-+
-+ pgd = swapper_pg_dir + pgd_index(vaddr);
-+ if (pgd_none(*pgd)) {
-+ BUG();
-+ return;
-+ }
-+ pud = pud_offset(pgd, vaddr);
-+ if (pud_none(*pud)) {
-+ BUG();
-+ return;
-+ }
-+ pmd = pmd_offset(pud, vaddr);
-+ if (pmd_none(*pmd)) {
-+ BUG();
-+ return;
-+ }
-+ pte = pte_offset_kernel(pmd, vaddr);
-+ if (pgprot_val(flags))
-+ /* <pfn,flags> stored as-is, to permit clearing entries */
-+ set_pte(pte, pfn_pte_ma(pfn, flags));
-+ else
-+ pte_clear(&init_mm, vaddr, pte);
-+
-+ /*
-+ * It's enough to flush this one mapping.
-+ * (PGE mappings get flushed as well)
-+ */
-+ __flush_tlb_one(vaddr);
-+}
-+
-+/*
+ * Associate a large virtual page frame with a given physical page frame
+ * and protection flags for that frame. pfn is for the base of the page,
+ * vaddr is what the page gets mapped to - both must be properly aligned.
@@ -20136,6 +20364,7 @@
+void __set_fixmap (enum fixed_addresses idx, maddr_t phys, pgprot_t flags)
+{
+ unsigned long address = __fix_to_virt(idx);
++ pte_t pte;
+
+ if (idx >= __end_of_fixed_addresses) {
+ BUG();
@@ -20143,16 +20372,16 @@
+ }
+ switch (idx) {
+ case FIX_WP_TEST:
-+#ifdef CONFIG_X86_F00F_BUG
-+ case FIX_F00F_IDT:
-+#endif
+ case FIX_VDSO:
-+ set_pte_pfn(address, phys >> PAGE_SHIFT, flags);
++ pte = pfn_pte(phys >> PAGE_SHIFT, flags);
+ break;
+ default:
-+ set_pte_pfn_ma(address, phys >> PAGE_SHIFT, flags);
++ pte = pfn_pte_ma(phys >> PAGE_SHIFT, flags);
+ break;
+ }
++ if (HYPERVISOR_update_va_mapping(address, pte,
++ UVMF_INVLPG|UVMF_ALL))
++ BUG();
+ nr_fixmaps++;
+}
+
@@ -20506,10 +20735,71 @@
+ }
+}
+
-+static inline void pgd_walk_set_prot(struct page *page, pgprot_t flags)
++static void _pin_lock(struct mm_struct *mm, int lock) {
++ if (lock)
++ spin_lock(&mm->page_table_lock);
++#if NR_CPUS >= CONFIG_SPLIT_PTLOCK_CPUS
++ /* While mm->page_table_lock protects us against insertions and
++ * removals of higher level page table pages, it doesn't protect
++ * against updates of pte-s. Such updates, however, require the
++ * pte pages to be in consistent state (unpinned+writable or
++ * pinned+readonly). The pinning and attribute changes, however
++ * cannot be done atomically, which is why such updates must be
++ * prevented from happening concurrently.
++ * Note that no pte lock can ever elsewhere be acquired nesting
++ * with an already acquired one in the same mm, or with the mm's
++ * page_table_lock already acquired, as that would break in the
++ * non-split case (where all these are actually resolving to the
++ * one page_table_lock). Thus acquiring all of them here is not
++ * going to result in dead locks, and the order of acquires
++ * doesn't matter.
++ */
++ {
++ pgd_t *pgd = mm->pgd;
++ unsigned g;
++
++ for (g = 0; g < USER_PTRS_PER_PGD; g++, pgd++) {
++ pud_t *pud;
++ unsigned u;
++
++ if (pgd_none(*pgd))
++ continue;
++ pud = pud_offset(pgd, 0);
++ for (u = 0; u < PTRS_PER_PUD; u++, pud++) {
++ pmd_t *pmd;
++ unsigned m;
++
++ if (pud_none(*pud))
++ continue;
++ pmd = pmd_offset(pud, 0);
++ for (m = 0; m < PTRS_PER_PMD; m++, pmd++) {
++ spinlock_t *ptl;
++
++ if (pmd_none(*pmd))
++ continue;
++ ptl = pte_lockptr(0, pmd);
++ if (lock)
++ spin_lock(ptl);
++ else
++ spin_unlock(ptl);
++ }
++ }
++ }
++ }
++#endif
++ if (!lock)
++ spin_unlock(&mm->page_table_lock);
++}
++#define pin_lock(mm) _pin_lock(mm, 1)
++#define pin_unlock(mm) _pin_lock(mm, 0)
++
++#define PIN_BATCH 4
++static DEFINE_PER_CPU(multicall_entry_t[PIN_BATCH], pb_mcl);
++
++static inline unsigned int pgd_walk_set_prot(struct page *page, pgprot_t flags,
++ unsigned int cpu, unsigned seq)
+{
+ unsigned long pfn = page_to_pfn(page);
-+ int rc;
+
+ if (PageHighMem(page)) {
+ if (pgprot_val(flags) & _PAGE_RW)
@@ -20517,12 +20807,18 @@
+ else
+ set_bit(PG_pinned, &page->flags);
+ } else {
-+ rc = HYPERVISOR_update_va_mapping(
-+ (unsigned long)__va(pfn << PAGE_SHIFT),
-+ pfn_pte(pfn, flags), 0);
-+ if (rc)
-+ BUG();
++ MULTI_update_va_mapping(per_cpu(pb_mcl, cpu) + seq,
++ (unsigned long)__va(pfn << PAGE_SHIFT),
++ pfn_pte(pfn, flags), 0);
++ if (unlikely(++seq == PIN_BATCH)) {
++ if (unlikely(HYPERVISOR_multicall_check(per_cpu(pb_mcl, cpu),
++ PIN_BATCH, NULL)))
++ BUG();
++ seq = 0;
++ }
+ }
++
++ return seq;
+}
+
+static void pgd_walk(pgd_t *pgd_base, pgprot_t flags)
@@ -20530,37 +20826,48 @@
+ pgd_t *pgd = pgd_base;
+ pud_t *pud;
+ pmd_t *pmd;
-+ int g, u, m, rc;
++ int g, u, m;
++ unsigned int cpu, seq;
+
+ if (xen_feature(XENFEAT_auto_translated_physmap))
+ return;
+
-+ for (g = 0; g < USER_PTRS_PER_PGD; g++, pgd++) {
++ cpu = get_cpu();
++
++ for (g = 0, seq = 0; g < USER_PTRS_PER_PGD; g++, pgd++) {
+ if (pgd_none(*pgd))
+ continue;
+ pud = pud_offset(pgd, 0);
+ if (PTRS_PER_PUD > 1) /* not folded */
-+ pgd_walk_set_prot(virt_to_page(pud),flags);
++ seq = pgd_walk_set_prot(virt_to_page(pud),flags,cpu,seq);
+ for (u = 0; u < PTRS_PER_PUD; u++, pud++) {
+ if (pud_none(*pud))
+ continue;
+ pmd = pmd_offset(pud, 0);
+ if (PTRS_PER_PMD > 1) /* not folded */
-+ pgd_walk_set_prot(virt_to_page(pmd),flags);
++ seq = pgd_walk_set_prot(virt_to_page(pmd),flags,cpu,seq);
+ for (m = 0; m < PTRS_PER_PMD; m++, pmd++) {
+ if (pmd_none(*pmd))
+ continue;
-+ pgd_walk_set_prot(pmd_page(*pmd),flags);
++ seq = pgd_walk_set_prot(pmd_page(*pmd),flags,cpu,seq);
+ }
+ }
+ }
+
-+ rc = HYPERVISOR_update_va_mapping(
-+ (unsigned long)pgd_base,
-+ pfn_pte(virt_to_phys(pgd_base)>>PAGE_SHIFT, flags),
-+ UVMF_TLB_FLUSH);
-+ if (rc)
++ if (likely(seq != 0)) {
++ MULTI_update_va_mapping(per_cpu(pb_mcl, cpu) + seq,
++ (unsigned long)pgd_base,
++ pfn_pte(virt_to_phys(pgd_base)>>PAGE_SHIFT, flags),
++ UVMF_TLB_FLUSH);
++ if (unlikely(HYPERVISOR_multicall_check(per_cpu(pb_mcl, cpu),
++ seq + 1, NULL)))
++ BUG();
++ } else if(HYPERVISOR_update_va_mapping((unsigned long)pgd_base,
++ pfn_pte(virt_to_phys(pgd_base)>>PAGE_SHIFT, flags),
++ UVMF_TLB_FLUSH))
+ BUG();
++
++ put_cpu();
+}
+
+static void __pgd_pin(pgd_t *pgd)
@@ -20588,18 +20895,18 @@
+{
+ if (xen_feature(XENFEAT_writable_page_tables))
+ return;
-+ spin_lock(&mm->page_table_lock);
++ pin_lock(mm);
+ __pgd_pin(mm->pgd);
-+ spin_unlock(&mm->page_table_lock);
++ pin_unlock(mm);
+}
+
+void mm_unpin(struct mm_struct *mm)
+{
+ if (xen_feature(XENFEAT_writable_page_tables))
+ return;
-+ spin_lock(&mm->page_table_lock);
++ pin_lock(mm);
+ __pgd_unpin(mm->pgd);
-+ spin_unlock(&mm->page_table_lock);
++ pin_unlock(mm);
+}
+
+void mm_pin_all(void)
@@ -20657,10 +20964,118 @@
+ !mm->context.has_foreign_mappings)
+ mm_unpin(mm);
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/oprofile/Makefile tmp-linux-2.6-xen.patch/arch/i386/oprofile/Makefile
---- pristine-linux-2.6.18.2/arch/i386/oprofile/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/oprofile/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -6,7 +6,14 @@ DRIVER_OBJS = $(addprefix ../../../drive
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/mm/pgtable.c
+--- a/arch/i386/mm/pgtable.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/mm/pgtable.c Wed Sep 10 10:54:08 2008 +0100
+@@ -12,6 +12,7 @@
+ #include <linux/slab.h>
+ #include <linux/pagemap.h>
+ #include <linux/spinlock.h>
++#include <linux/module.h>
+
+ #include <asm/system.h>
+ #include <asm/pgtable.h>
+@@ -137,6 +138,10 @@
+ __flush_tlb_one(vaddr);
+ }
+
++static int nr_fixmaps = 0;
++unsigned long __FIXADDR_TOP = 0xfffff000;
++EXPORT_SYMBOL(__FIXADDR_TOP);
++
+ void __set_fixmap (enum fixed_addresses idx, unsigned long phys, pgprot_t flags)
+ {
+ unsigned long address = __fix_to_virt(idx);
+@@ -146,6 +151,13 @@
+ return;
+ }
+ set_pte_pfn(address, phys >> PAGE_SHIFT, flags);
++ nr_fixmaps++;
++}
++
++void set_fixaddr_top(unsigned long top)
++{
++ BUG_ON(nr_fixmaps > 0);
++ __FIXADDR_TOP = top - PAGE_SIZE;
+ }
+
+ pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
+@@ -214,9 +226,10 @@
+ spin_lock_irqsave(&pgd_lock, flags);
+ }
+
+- clone_pgd_range((pgd_t *)pgd + USER_PTRS_PER_PGD,
+- swapper_pg_dir + USER_PTRS_PER_PGD,
+- KERNEL_PGD_PTRS);
++ if (PTRS_PER_PMD == 1 || HAVE_SHARED_KERNEL_PMD)
++ clone_pgd_range((pgd_t *)pgd + USER_PTRS_PER_PGD,
++ swapper_pg_dir + USER_PTRS_PER_PGD,
++ KERNEL_PGD_PTRS);
+ if (PTRS_PER_PMD > 1)
+ return;
+
+@@ -248,6 +261,30 @@
+ goto out_oom;
+ set_pgd(&pgd[i], __pgd(1 + __pa(pmd)));
+ }
++
++ if (!HAVE_SHARED_KERNEL_PMD) {
++ unsigned long flags;
++
++ for (i = USER_PTRS_PER_PGD; i < PTRS_PER_PGD; i++) {
++ pmd_t *pmd = kmem_cache_alloc(pmd_cache, GFP_KERNEL);
++ if (!pmd)
++ goto out_oom;
++ set_pgd(&pgd[USER_PTRS_PER_PGD], __pgd(1 + __pa(pmd)));
++ }
++
++ spin_lock_irqsave(&pgd_lock, flags);
++ for (i = USER_PTRS_PER_PGD; i < PTRS_PER_PGD; i++) {
++ unsigned long v = (unsigned long)i << PGDIR_SHIFT;
++ pgd_t *kpgd = pgd_offset_k(v);
++ pud_t *kpud = pud_offset(kpgd, v);
++ pmd_t *kpmd = pmd_offset(kpud, v);
++ pmd_t *pmd = (void *)__va(pgd_val(pgd[i])-1);
++ memcpy(pmd, kpmd, PAGE_SIZE);
++ }
++ pgd_list_add(pgd);
++ spin_unlock_irqrestore(&pgd_lock, flags);
++ }
++
+ return pgd;
+
+ out_oom:
+@@ -262,9 +299,23 @@
+ int i;
+
+ /* in the PAE case user pgd entries are overwritten before usage */
+- if (PTRS_PER_PMD > 1)
+- for (i = 0; i < USER_PTRS_PER_PGD; ++i)
+- kmem_cache_free(pmd_cache, (void *)__va(pgd_val(pgd[i])-1));
++ if (PTRS_PER_PMD > 1) {
++ for (i = 0; i < USER_PTRS_PER_PGD; ++i) {
++ pmd_t *pmd = (void *)__va(pgd_val(pgd[i])-1);
++ kmem_cache_free(pmd_cache, pmd);
++ }
++ if (!HAVE_SHARED_KERNEL_PMD) {
++ unsigned long flags;
++ spin_lock_irqsave(&pgd_lock, flags);
++ pgd_list_del(pgd);
++ spin_unlock_irqrestore(&pgd_lock, flags);
++ for (i = USER_PTRS_PER_PGD; i < PTRS_PER_PGD; i++) {
++ pmd_t *pmd = (void *)__va(pgd_val(pgd[i])-1);
++ memset(pmd, 0, PTRS_PER_PMD*sizeof(pmd_t));
++ kmem_cache_free(pmd_cache, pmd);
++ }
++ }
++ }
+ /* in the non-PAE case, free_pgtables() clears user pgd entries */
+ kmem_cache_free(pgd_cache, pgd);
+ }
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/oprofile/Makefile
+--- a/arch/i386/oprofile/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/oprofile/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -6,7 +6,14 @@
oprofilefs.o oprofile_stats.o \
timer_int.o )
@@ -20675,9 +21090,9 @@
op_model_ppro.o op_model_p4.o
oprofile-$(CONFIG_X86_IO_APIC) += nmi_timer_int.o
+endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/oprofile/xenoprof.c tmp-linux-2.6-xen.patch/arch/i386/oprofile/xenoprof.c
---- pristine-linux-2.6.18.2/arch/i386/oprofile/xenoprof.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/oprofile/xenoprof.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/oprofile/xenoprof.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/oprofile/xenoprof.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,179 @@
+/**
+ * @file xenoprof.c
@@ -20733,8 +21148,8 @@
+ counter.kernel = (uint32_t)counter_config[i].kernel;
+ counter.user = (uint32_t)counter_config[i].user;
+ counter.unit_mask = (uint64_t)counter_config[i].unit_mask;
-+ HYPERVISOR_xenoprof_op(XENOPROF_counter,
-+ &counter);
++ WARN_ON(HYPERVISOR_xenoprof_op(XENOPROF_counter,
++ &counter));
+ }
+}
+
@@ -20858,10 +21273,24 @@
+{
+ xenoprofile_exit();
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/pci/irq-xen.c tmp-linux-2.6-xen.patch/arch/i386/pci/irq-xen.c
---- pristine-linux-2.6.18.2/arch/i386/pci/irq-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/pci/irq-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,1205 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/pci/Makefile
+--- a/arch/i386/pci/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/pci/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -3,6 +3,10 @@
+ obj-$(CONFIG_PCI_BIOS) += pcbios.o
+ obj-$(CONFIG_PCI_MMCONFIG) += mmconfig.o direct.o
+ obj-$(CONFIG_PCI_DIRECT) += direct.o
++
++# pcifront should be after pcbios.o, mmconfig.o, and direct.o as it should only
++# take over if direct access to the PCI bus is unavailable
++obj-$(CONFIG_XEN_PCIDEV_FRONTEND) += pcifront.o
+
+ pci-y := fixup.o
+ pci-$(CONFIG_ACPI) += acpi.o
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/pci/irq-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/pci/irq-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1211 @@
+/*
+ * Low-Level PCI Support for PC -- Routing of Interrupts
+ *
@@ -21123,13 +21552,13 @@
+ */
+static int pirq_via586_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
+{
-+ static const unsigned int pirqmap[4] = { 3, 2, 5, 1 };
++ static const unsigned int pirqmap[5] = { 3, 2, 5, 1, 1 };
+ return read_config_nybble(router, 0x55, pirqmap[pirq-1]);
+}
+
+static int pirq_via586_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
+{
-+ static const unsigned int pirqmap[4] = { 3, 2, 5, 1 };
++ static const unsigned int pirqmap[5] = { 3, 2, 5, 1, 1 };
+ write_config_nybble(router, 0x55, pirqmap[pirq-1], irq);
+ return 1;
+}
@@ -21411,6 +21840,12 @@
+ case PCI_DEVICE_ID_INTEL_ICH8_2:
+ case PCI_DEVICE_ID_INTEL_ICH8_3:
+ case PCI_DEVICE_ID_INTEL_ICH8_4:
++ case PCI_DEVICE_ID_INTEL_ICH9_0:
++ case PCI_DEVICE_ID_INTEL_ICH9_1:
++ case PCI_DEVICE_ID_INTEL_ICH9_2:
++ case PCI_DEVICE_ID_INTEL_ICH9_3:
++ case PCI_DEVICE_ID_INTEL_ICH9_4:
++ case PCI_DEVICE_ID_INTEL_ICH9_5:
+ r->name = "PIIX/ICH";
+ r->get = pirq_piix_get;
+ r->set = pirq_piix_set;
@@ -22067,32 +22502,9 @@
+
+ return count;
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/pci/Makefile tmp-linux-2.6-xen.patch/arch/i386/pci/Makefile
---- pristine-linux-2.6.18.2/arch/i386/pci/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/pci/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -4,6 +4,10 @@ obj-$(CONFIG_PCI_BIOS) += pcbios.o
- obj-$(CONFIG_PCI_MMCONFIG) += mmconfig.o direct.o
- obj-$(CONFIG_PCI_DIRECT) += direct.o
-
-+# pcifront should be after pcbios.o, mmconfig.o, and direct.o as it should only
-+# take over if direct access to the PCI bus is unavailable
-+obj-$(CONFIG_XEN_PCIDEV_FRONTEND) += pcifront.o
-+
- pci-y := fixup.o
- pci-$(CONFIG_ACPI) += acpi.o
- pci-y += legacy.o irq.o
-@@ -12,3 +16,8 @@ pci-$(CONFIG_X86_VISWS) := visws.o fixu
- pci-$(CONFIG_X86_NUMAQ) := numa.o irq.o
-
- obj-y += $(pci-y) common.o
-+
-+ifdef CONFIG_XEN
-+include $(srctree)/scripts/Makefile.xen
-+obj-y := $(call cherrypickxen, $(obj-y))
-+endif
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/pci/pcifront.c tmp-linux-2.6-xen.patch/arch/i386/pci/pcifront.c
---- pristine-linux-2.6.18.2/arch/i386/pci/pcifront.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/pci/pcifront.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/pci/pcifront.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/i386/pci/pcifront.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,55 @@
+/*
+ * PCI Frontend Stub - puts some "dummy" functions in to the Linux x86 PCI core
@@ -22149,68 +22561,48 @@
+}
+
+arch_initcall(pcifront_x86_stub_init);
-diff -Nurp pristine-linux-2.6.18.2/arch/i386/power/Makefile tmp-linux-2.6-xen.patch/arch/i386/power/Makefile
---- pristine-linux-2.6.18.2/arch/i386/power/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/i386/power/Makefile 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/power/Makefile
+--- a/arch/i386/power/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/power/Makefile Wed Sep 10 10:54:08 2008 +0100
@@ -1,2 +1,4 @@
-obj-$(CONFIG_PM) += cpu.o
-+obj-$(CONFIG_PM_LEGACY) += cpu.o
++obj-$(subst m,y,$(CONFIG_APM)) += cpu.o
+obj-$(CONFIG_SOFTWARE_SUSPEND) += cpu.o
+obj-$(CONFIG_ACPI_SLEEP) += cpu.o
obj-$(CONFIG_SOFTWARE_SUSPEND) += swsusp.o
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/dig/setup.c tmp-linux-2.6-xen.patch/arch/ia64/dig/setup.c
---- pristine-linux-2.6.18.2/arch/ia64/dig/setup.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/dig/setup.c 2007-07-30 16:35:11.000000000 +0200
-@@ -24,6 +24,8 @@
- #include <asm/machvec.h>
- #include <asm/system.h>
+diff -r d894e36cfc30 -r 0aa021803deb arch/i386/power/cpu.c
+--- a/arch/i386/power/cpu.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/i386/power/cpu.c Wed Sep 10 10:54:08 2008 +0100
+@@ -62,11 +62,12 @@
-+#include <xen/xencons.h>
-+
- void __init
- dig_setup (char **cmdline_p)
+ static void fix_processor_context(void)
{
-@@ -67,4 +69,19 @@ dig_setup (char **cmdline_p)
- screen_info.orig_video_mode = 3; /* XXX fake */
- screen_info.orig_video_isVGA = 1; /* XXX fake */
- screen_info.orig_video_ega_bx = 3; /* XXX fake */
-+#ifdef CONFIG_XEN
-+ if (!is_running_on_xen() || !is_initial_xendomain())
-+ return;
-+
-+ if (xen_start_info->console.dom0.info_size >=
-+ sizeof(struct dom0_vga_console_info)) {
-+ const struct dom0_vga_console_info *info =
-+ (struct dom0_vga_console_info *)(
-+ (char *)xen_start_info +
-+ xen_start_info->console.dom0.info_off);
-+ dom0_init_screen_info(info);
-+ }
-+ xen_start_info->console.domU.mfn = 0;
-+ xen_start_info->console.domU.evtchn = 0;
++#ifndef CONFIG_X86_NO_TSS
+ int cpu = smp_processor_id();
+ struct tss_struct * t = &per_cpu(init_tss, cpu);
+
+ set_tss_desc(cpu,t); /* This just modifies memory; should not be necessary. But... This is necessary, because 386 hardware has concept of busy TSS or some similar stupidity. */
+-
+#endif
- }
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/Kconfig tmp-linux-2.6-xen.patch/arch/ia64/Kconfig
---- pristine-linux-2.6.18.2/arch/ia64/Kconfig 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/Kconfig 2007-07-30 16:35:11.000000000 +0200
-@@ -58,6 +58,34 @@ config GENERIC_IOMAP
+ load_TR_desc(); /* This does ltr */
+ load_LDT(¤t->active_mm->context); /* This does lldt */
+
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/Kconfig
+--- a/arch/ia64/Kconfig Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -57,6 +57,28 @@
+ config GENERIC_IOMAP
bool
default y
-
++
+config XEN
+ bool "Xen hypervisor support"
+ default y
++ select XEN_XENCOMM
+ help
+ Enable Xen hypervisor support. Resulting kernel runs
+ both as a guest OS on Xen and natively on hardware.
+
-+config XEN_IA64_VDSO_PARAVIRT
-+ bool
-+ depends on XEN && !ITANIUM
-+ default y
-+ help
-+ vDSO paravirtualization
-+
+config XEN_IA64_EXPOSE_P2M
+ bool "Xen/IA64 exposure p2m table"
+ depends on XEN
@@ -22224,14 +22616,55 @@
+ default y
+ help
+ use dtr to map the exposed p2m table
-+
+
config SCHED_NO_NO_OMIT_FRAME_POINTER
bool
- default y
-@@ -465,6 +493,21 @@ config PCI_DOMAINS
+@@ -132,6 +154,10 @@
+
+ config IA64_HP_SIM
+ bool "Ski-simulator"
++
++config IA64_XEN
++ bool "Xen guest"
++ depends on XEN
+
+ endchoice
+
+@@ -431,6 +457,29 @@
+
+ source "drivers/sn/Kconfig"
+
++config KEXEC
++ bool "kexec system call (EXPERIMENTAL)"
++ depends on EXPERIMENTAL && !IA64_HP_SIM && (!SMP || HOTPLUG_CPU) && !XEN_UNPRIVILEGED_GUEST
++ help
++ kexec is a system call that implements the ability to shutdown your
++ current kernel, and to start another kernel. It is like a reboot
++ but it is indepedent of the system firmware. And like a reboot
++ you can start any kernel with it, not just Linux.
++
++ The name comes from the similiarity to the exec system call.
++
++ It is an ongoing process to be certain the hardware in a machine
++ is properly shutdown, so do not be surprised if this code does not
++ initially work for you. It may help to enable device hotplugging
++ support. As of this writing the exact hardware interface is
++ strongly in flux, so no good recommendation can be made.
++
++config CRASH_DUMP
++ bool "kernel crash dumps (EXPERIMENTAL)"
++ depends on EXPERIMENTAL && IA64_MCA_RECOVERY && !IA64_HP_SIM && (!SMP || HOTPLUG_CPU)
++ help
++ Generate crash dump after being started by kexec.
++
+ source "drivers/firmware/Kconfig"
+
+ source "fs/Kconfig.binfmt"
+@@ -464,6 +513,21 @@
+ config PCI_DOMAINS
bool
default PCI
-
++
+config XEN_PCIDEV_FRONTEND
+ bool "Xen PCI Frontend"
+ depends on PCI && XEN
@@ -22246,11 +22679,10 @@
+ default n
+ help
+ Enables some debug statements within the PCI Frontend.
-+
+
source "drivers/pci/pcie/Kconfig"
- source "drivers/pci/Kconfig"
-@@ -528,3 +571,13 @@ source "arch/ia64/Kconfig.debug"
+@@ -528,3 +592,16 @@
source "security/Kconfig"
source "crypto/Kconfig"
@@ -22261,186 +22693,1210 @@
+if XEN
+config XEN_SMPBOOT
+ default n
++
++config XEN_DEVMEM
++ default n
+endif
+
+source "drivers/xen/Kconfig"
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/asm-offsets.c tmp-linux-2.6-xen.patch/arch/ia64/kernel/asm-offsets.c
---- pristine-linux-2.6.18.2/arch/ia64/kernel/asm-offsets.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/asm-offsets.c 2007-07-30 16:35:11.000000000 +0200
-@@ -268,4 +268,29 @@ void foo(void)
- DEFINE(IA64_TIME_SOURCE_MMIO64, TIME_SOURCE_MMIO64);
- DEFINE(IA64_TIME_SOURCE_MMIO32, TIME_SOURCE_MMIO32);
- DEFINE(IA64_TIMESPEC_TV_NSEC_OFFSET, offsetof (struct timespec, tv_nsec));
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/Makefile
+--- a/arch/ia64/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -36,6 +36,12 @@
+ endif
+
+ CFLAGS += $(cflags-y)
+
-+#ifdef CONFIG_XEN
-+ BLANK();
++cppflags-$(CONFIG_XEN) += \
++ -D__XEN_INTERFACE_VERSION__=$(CONFIG_XEN_INTERFACE_VERSION)
+
-+#define DEFINE_MAPPED_REG_OFS(sym, field) \
-+ DEFINE(sym, (XMAPPEDREGS_OFS + offsetof(mapped_regs_t, field)))
++CPPFLAGS += $(cppflags-y)
+
-+ DEFINE_MAPPED_REG_OFS(XSI_PSR_I_ADDR_OFS, interrupt_mask_addr);
-+ DEFINE_MAPPED_REG_OFS(XSI_IPSR_OFS, ipsr);
-+ DEFINE_MAPPED_REG_OFS(XSI_IIP_OFS, iip);
-+ DEFINE_MAPPED_REG_OFS(XSI_IFS_OFS, ifs);
-+ DEFINE_MAPPED_REG_OFS(XSI_PRECOVER_IFS_OFS, precover_ifs);
-+ DEFINE_MAPPED_REG_OFS(XSI_ISR_OFS, isr);
-+ DEFINE_MAPPED_REG_OFS(XSI_IFA_OFS, ifa);
-+ DEFINE_MAPPED_REG_OFS(XSI_IIPA_OFS, iipa);
-+ DEFINE_MAPPED_REG_OFS(XSI_IIM_OFS, iim);
-+ DEFINE_MAPPED_REG_OFS(XSI_IHA_OFS, iha);
-+ DEFINE_MAPPED_REG_OFS(XSI_ITIR_OFS, itir);
-+ DEFINE_MAPPED_REG_OFS(XSI_PSR_IC_OFS, interrupt_collection_enabled);
-+ DEFINE_MAPPED_REG_OFS(XSI_BANKNUM_OFS, banknum);
-+ DEFINE_MAPPED_REG_OFS(XSI_BANK0_R16_OFS, bank0_regs[0]);
-+ DEFINE_MAPPED_REG_OFS(XSI_BANK1_R16_OFS, bank1_regs[0]);
-+ DEFINE_MAPPED_REG_OFS(XSI_B0NATS_OFS, vbnat);
-+ DEFINE_MAPPED_REG_OFS(XSI_B1NATS_OFS, vnat);
-+#endif /* CONFIG_XEN */
- }
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/entry.S tmp-linux-2.6-xen.patch/arch/ia64/kernel/entry.S
---- pristine-linux-2.6.18.2/arch/ia64/kernel/entry.S 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/entry.S 2007-07-30 16:35:11.000000000 +0200
-@@ -180,7 +180,7 @@ END(sys_clone)
- * called. The code starting at .map relies on this. The rest of the code
- * doesn't care about the interrupt masking status.
- */
--GLOBAL_ENTRY(ia64_switch_to)
-+GLOBAL_ENTRY(__ia64_switch_to)
- .prologue
- alloc r16=ar.pfs,1,0,0,0
- DO_SAVE_SWITCH_STACK
-@@ -234,7 +234,7 @@ GLOBAL_ENTRY(ia64_switch_to)
- ;;
- srlz.d
- br.cond.sptk .done
--END(ia64_switch_to)
-+END(__ia64_switch_to)
+ head-y := arch/ia64/kernel/head.o arch/ia64/kernel/init_task.o
- /*
- * Note that interrupts are enabled during save_switch_stack and load_switch_stack. This
-@@ -375,7 +375,7 @@ END(save_switch_stack)
- * - b7 holds address to return to
- * - must not touch r8-r11
- */
--ENTRY(load_switch_stack)
-+GLOBAL_ENTRY(load_switch_stack)
- .prologue
- .altrp b7
+ libs-y += arch/ia64/lib/
+@@ -45,7 +51,9 @@
+ core-$(CONFIG_IA64_GENERIC) += arch/ia64/dig/
+ core-$(CONFIG_IA64_HP_ZX1) += arch/ia64/dig/
+ core-$(CONFIG_IA64_HP_ZX1_SWIOTLB) += arch/ia64/dig/
++core-$(CONFIG_IA64_XEN) += arch/ia64/dig/
+ core-$(CONFIG_IA64_SGI_SN2) += arch/ia64/sn/
++core-$(CONFIG_XEN) += arch/ia64/xen/
-@@ -510,7 +510,7 @@ END(clone)
- * because some system calls (such as ia64_execve) directly
- * manipulate ar.pfs.
- */
--GLOBAL_ENTRY(ia64_trace_syscall)
-+GLOBAL_ENTRY(__ia64_trace_syscall)
- PT_REGS_UNWIND_INFO(0)
- /*
- * We need to preserve the scratch registers f6-f11 in case the system
-@@ -582,7 +582,7 @@ strace_error:
- (p6) mov r10=-1
- (p6) mov r8=r9
- br.cond.sptk .strace_save_retval
--END(ia64_trace_syscall)
-+END(__ia64_trace_syscall)
+ drivers-$(CONFIG_PCI) += arch/ia64/pci/
+ drivers-$(CONFIG_IA64_HP_SIM) += arch/ia64/hp/sim/
+@@ -75,8 +83,8 @@
+ boot: lib/lib.a vmlinux
+ $(Q)$(MAKE) $(build)=$(boot) $@
- /*
- * When traced and returning from sigreturn, we invoke syscall_trace but then
-@@ -601,7 +601,7 @@ GLOBAL_ENTRY(ia64_strace_leave_kernel)
- .ret4: br.cond.sptk ia64_leave_kernel
- END(ia64_strace_leave_kernel)
+-install: vmlinux.gz
+- sh $(srctree)/arch/ia64/install.sh $(KERNELRELEASE) $< System.map "$(INSTALL_PATH)"
++install:
++ -yes | sh $(srctree)/arch/ia64/install.sh $(KERNELRELEASE) vmlinux.gz System.map "$(INSTALL_PATH)"
--GLOBAL_ENTRY(ia64_ret_from_clone)
-+GLOBAL_ENTRY(__ia64_ret_from_clone)
- PT_REGS_UNWIND_INFO(0)
- { /*
- * Some versions of gas generate bad unwind info if the first instruction of a
-@@ -627,7 +627,7 @@ GLOBAL_ENTRY(ia64_ret_from_clone)
- cmp.ne p6,p0=r2,r0
- (p6) br.cond.spnt .strace_check_retval
- ;; // added stop bits to prevent r8 dependency
--END(ia64_ret_from_clone)
-+END(__ia64_ret_from_clone)
- // fall through
- GLOBAL_ENTRY(ia64_ret_from_syscall)
- PT_REGS_UNWIND_INFO(0)
-@@ -635,8 +635,11 @@ GLOBAL_ENTRY(ia64_ret_from_syscall)
- adds r2=PT(R8)+16,sp // r2 = &pt_regs.r8
- mov r10=r0 // clear error indication in r10
- (p7) br.cond.spnt handle_syscall_error // handle potential syscall failure
-+ ;;
-+ // don't fall through, ia64_leave_syscall may be #define'd
-+ br.cond.sptk.few ia64_leave_syscall
-+ ;;
- END(ia64_ret_from_syscall)
-- // fall through
- /*
- * ia64_leave_syscall(): Same as ia64_leave_kernel, except that it doesn't
- * need to switch to bank 0 and doesn't restore the scratch registers.
-@@ -681,7 +684,7 @@ END(ia64_ret_from_syscall)
- * ar.csd: cleared
- * ar.ssd: cleared
- */
--ENTRY(ia64_leave_syscall)
-+GLOBAL_ENTRY(__ia64_leave_syscall)
- PT_REGS_UNWIND_INFO(0)
- /*
- * work.need_resched etc. mustn't get changed by this CPU before it returns to
-@@ -789,7 +792,7 @@ ENTRY(ia64_leave_syscall)
- mov.m ar.ssd=r0 // M2 clear ar.ssd
- mov f11=f0 // F clear f11
- br.cond.sptk.many rbs_switch // B
--END(ia64_leave_syscall)
-+END(__ia64_leave_syscall)
+ define archhelp
+ echo '* compressed - Build compressed kernel image'
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/hp/common/sba_iommu.c
+--- a/arch/ia64/hp/common/sba_iommu.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/hp/common/sba_iommu.c Wed Sep 10 10:54:08 2008 +0100
+@@ -42,6 +42,11 @@
+ #include <asm/system.h> /* wmb() */
- #ifdef CONFIG_IA32_SUPPORT
- GLOBAL_ENTRY(ia64_ret_from_ia32_execve)
-@@ -801,10 +804,13 @@ GLOBAL_ENTRY(ia64_ret_from_ia32_execve)
- st8.spill [r2]=r8 // store return value in slot for r8 and set unat bit
- .mem.offset 8,0
- st8.spill [r3]=r0 // clear error indication in slot for r10 and set unat bit
-+ ;;
-+ // don't fall through, ia64_leave_kernel may be #define'd
-+ br.cond.sptk.few ia64_leave_kernel
-+ ;;
- END(ia64_ret_from_ia32_execve)
-- // fall through
- #endif /* CONFIG_IA32_SUPPORT */
--GLOBAL_ENTRY(ia64_leave_kernel)
-+GLOBAL_ENTRY(__ia64_leave_kernel)
- PT_REGS_UNWIND_INFO(0)
- /*
- * work.need_resched etc. mustn't get changed by this CPU before it returns to
-@@ -1135,7 +1141,7 @@ skip_rbs_switch:
- ld8 r10=[r3]
- br.cond.sptk.many .work_processed_syscall // re-check
+ #include <asm/acpi-ext.h>
++#include <asm/maddr.h> /* range_straddles_page_boundary() */
++#ifdef CONFIG_XEN
++#include <xen/gnttab.h>
++#include <asm/gnttab_dma.h>
++#endif
--END(ia64_leave_kernel)
-+END(__ia64_leave_kernel)
+ #define PFX "IOC: "
- ENTRY(handle_syscall_error)
- /*
-@@ -1175,7 +1181,7 @@ END(ia64_invoke_schedule_tail)
- * be set up by the caller. We declare 8 input registers so the system call
- * args get preserved, in case we need to restart a system call.
- */
--ENTRY(notify_resume_user)
-+GLOBAL_ENTRY(notify_resume_user)
- .prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(8)
- alloc loc1=ar.pfs,8,2,3,0 // preserve all eight input regs in case of syscall restart!
- mov r9=ar.unat
-@@ -1263,7 +1269,7 @@ ENTRY(sys_rt_sigreturn)
- adds sp=16,sp
- ;;
- ld8 r9=[sp] // load new ar.unat
-- mov.sptk b7=r8,ia64_leave_kernel
-+ mov.sptk b7=r8,__ia64_leave_kernel
- ;;
+@@ -198,6 +203,9 @@
+ void __iomem *ioc_hpa; /* I/O MMU base address */
+ char *res_map; /* resource map, bit == pdir entry */
+ u64 *pdir_base; /* physical base address */
++#ifdef CONFIG_XEN
++ u64 *xen_virt_cache;
++#endif
+ unsigned long ibase; /* pdir IOV Space base */
+ unsigned long imask; /* pdir IOV Space mask */
+
+@@ -762,14 +770,21 @@
+ * on the vba.
+ */
+
+-#if 1
+-#define sba_io_pdir_entry(pdir_ptr, vba) *pdir_ptr = ((vba & ~0xE000000000000FFFULL) \
+- | 0x8000000000000000ULL)
++#ifndef CONFIG_XEN
++#define sba_io_pdir_entry(ioc, pdir_ptr, vba) *pdir_ptr = \
++ ((virt_to_bus((void *)vba) & ~0xFFFULL) | 0x8000000000000000ULL)
+ #else
+ void SBA_INLINE
+-sba_io_pdir_entry(u64 *pdir_ptr, unsigned long vba)
++sba_io_pdir_entry(struct ioc *ioc, u64 *pdir_ptr, unsigned long vba)
+ {
+- *pdir_ptr = ((vba & ~0xE000000000000FFFULL) | 0x80000000000000FFULL);
++ *pdir_ptr = ((virt_to_bus((void *)vba) & ~0xFFFULL) |
++ 0x80000000000000FFULL);
++#ifdef CONFIG_XEN
++ if (is_running_on_xen()) {
++ int pide = ((u64)pdir_ptr - (u64)ioc->pdir_base) >> 3;
++ ioc->xen_virt_cache[pide] = vba;
++ }
++#endif
+ }
+ #endif
+
+@@ -783,6 +798,12 @@
+ mark_clean (void *addr, size_t size)
+ {
+ unsigned long pg_addr, end;
++
++#ifdef CONFIG_XEN
++ /* XXX: Bad things happen starting domUs when this is enabled. */
++ if (is_running_on_xen())
++ return;
++#endif
+
+ pg_addr = PAGE_ALIGN((unsigned long) addr);
+ end = (unsigned long) addr + size;
+@@ -850,6 +871,10 @@
+ */
+ ioc->pdir_base[off] = (0x80000000000000FFULL | prefetch_spill_page);
+ #endif
++#ifdef CONFIG_XEN
++ if (is_running_on_xen())
++ ioc->xen_virt_cache[off] = 0UL;
++#endif
+ } else {
+ u32 t = get_iovp_order(byte_cnt) + iovp_shift;
+
+@@ -864,6 +889,10 @@
+ ioc->pdir_base[off] &= ~(0x80000000000000FFULL);
+ #else
+ ioc->pdir_base[off] = (0x80000000000000FFULL | prefetch_spill_page);
++#endif
++#ifdef CONFIG_XEN
++ if (is_running_on_xen())
++ ioc->xen_virt_cache[off] = 0UL;
+ #endif
+ off++;
+ byte_cnt -= iovp_size;
+@@ -894,15 +923,29 @@
+ unsigned long flags;
+ #endif
+ #ifdef ALLOW_IOV_BYPASS
+- unsigned long pci_addr = virt_to_phys(addr);
++ unsigned long pci_addr;
++#endif
++
++#ifdef CONFIG_XEN
++ if (is_running_on_xen()) {
++ void* tmp_addr = addr;
++ size_t tmp_size = size;
++ do {
++ gnttab_dma_use_page(virt_to_page(tmp_addr));
++ tmp_addr += PAGE_SIZE;
++ tmp_size -= min(tmp_size, PAGE_SIZE);
++ } while (tmp_size);
++ }
+ #endif
+
+ #ifdef ALLOW_IOV_BYPASS
++ pci_addr = virt_to_bus(addr);
+ ASSERT(to_pci_dev(dev)->dma_mask);
+ /*
+ ** Check if the PCI device can DMA to ptr... if so, just return ptr
+ */
+- if (likely((pci_addr & ~to_pci_dev(dev)->dma_mask) == 0)) {
++ if (likely((pci_addr & ~to_pci_dev(dev)->dma_mask) == 0 &&
++ !range_straddles_page_boundary(__pa(addr), size))) {
+ /*
+ ** Device is bit capable of DMA'ing to the buffer...
+ ** just return the PCI address of ptr
+@@ -944,7 +987,7 @@
+
+ while (size > 0) {
+ ASSERT(((u8 *)pdir_start)[7] == 0); /* verify availability */
+- sba_io_pdir_entry(pdir_start, (unsigned long) addr);
++ sba_io_pdir_entry(ioc, pdir_start, (unsigned long) addr);
+
+ DBG_RUN(" pdir 0x%p %lx\n", pdir_start, *pdir_start);
+
+@@ -973,14 +1016,58 @@
+ void *addr;
+
+ if (size <= iovp_size) {
+- addr = phys_to_virt(ioc->pdir_base[off] &
+- ~0xE000000000000FFFULL);
++#ifdef CONFIG_XEN
++ if (is_running_on_xen())
++ addr = (void *)ioc->xen_virt_cache[off];
++ else
++ addr = bus_to_virt(ioc->pdir_base[off] &
++ ~0xE000000000000FFFULL);
++#else
++ addr = bus_to_virt(ioc->pdir_base[off] &
++ ~0xE000000000000FFFULL);
++#endif
+ mark_clean(addr, size);
+ } else {
+ do {
+- addr = phys_to_virt(ioc->pdir_base[off] &
+- ~0xE000000000000FFFULL);
++#ifdef CONFIG_XEN
++ if (is_running_on_xen())
++ addr = (void *)ioc->xen_virt_cache[off];
++ else
++ addr = bus_to_virt(ioc->pdir_base[off] &
++ ~0xE000000000000FFFULL);
++#else
++ addr = bus_to_virt(ioc->pdir_base[off] &
++ ~0xE000000000000FFFULL);
++#endif
+ mark_clean(addr, min(size, iovp_size));
++ off++;
++ size -= iovp_size;
++ } while (size > 0);
++ }
++}
++#endif
++
++#ifdef CONFIG_XEN
++static void
++sba_gnttab_dma_unmap_page(struct ioc *ioc, dma_addr_t iova, size_t size)
++{
++ u32 iovp = (u32) SBA_IOVP(ioc,iova);
++ int off = PDIR_INDEX(iovp);
++ struct page *page;
++
++ if (size <= iovp_size) {
++ BUG_ON(!ioc->xen_virt_cache[off]);
++ page = virt_to_page(ioc->xen_virt_cache[off]);
++ __gnttab_dma_unmap_page(page);
++ } else {
++ struct page *last_page = (struct page *)~0UL;
++ do {
++ BUG_ON(!ioc->xen_virt_cache[off]);
++ page = virt_to_page(ioc->xen_virt_cache[off]);
++ if (page != last_page) {
++ __gnttab_dma_unmap_page(page);
++ last_page = page;
++ }
+ off++;
+ size -= iovp_size;
+ } while (size > 0);
+@@ -1018,7 +1105,16 @@
+
+ #ifdef ENABLE_MARK_CLEAN
+ if (dir == DMA_FROM_DEVICE) {
+- mark_clean(phys_to_virt(iova), size);
++ mark_clean(bus_to_virt(iova), size);
++ }
++#endif
++#ifdef CONFIG_XEN
++ if (is_running_on_xen()) {
++ do {
++ gnttab_dma_unmap_page(iova);
++ iova += PAGE_SIZE;
++ size -= min(size,PAGE_SIZE);
++ } while (size);
+ }
+ #endif
+ return;
+@@ -1036,6 +1132,10 @@
+ #ifdef ENABLE_MARK_CLEAN
+ if (dir == DMA_FROM_DEVICE)
+ sba_mark_clean(ioc, iova, size);
++#endif
++#ifdef CONFIG_XEN
++ if (is_running_on_xen())
++ sba_gnttab_dma_unmap_page(ioc, iova, size);
+ #endif
+
+ #if DELAYED_RESOURCE_CNT > 0
+@@ -1102,9 +1202,14 @@
+ return NULL;
+
+ memset(addr, 0, size);
+- *dma_handle = virt_to_phys(addr);
+
+ #ifdef ALLOW_IOV_BYPASS
++#ifdef CONFIG_XEN
++ if (xen_create_contiguous_region((unsigned long)addr, get_order(size),
++ fls64(dev->coherent_dma_mask)))
++ goto iommu_map;
++#endif
++ *dma_handle = virt_to_bus(addr);
+ ASSERT(dev->coherent_dma_mask);
+ /*
+ ** Check if the PCI device can DMA to ptr... if so, just return ptr
+@@ -1115,6 +1220,9 @@
+
+ return addr;
+ }
++#ifdef CONFIG_XEN
++iommu_map:
++#endif
+ #endif
+
+ /*
+@@ -1138,6 +1246,13 @@
+ */
+ void sba_free_coherent (struct device *dev, size_t size, void *vaddr, dma_addr_t dma_handle)
+ {
++#if defined(ALLOW_IOV_BYPASS) && defined(CONFIG_XEN)
++ struct ioc *ioc = GET_IOC(dev);
++
++ if (likely((dma_handle & ioc->imask) != ioc->ibase))
++ xen_destroy_contiguous_region((unsigned long)vaddr,
++ get_order(size));
++#endif
+ sba_unmap_single(dev, dma_handle, size, 0);
+ free_pages((unsigned long) vaddr, get_order(size));
+ }
+@@ -1219,7 +1334,7 @@
+ dma_offset=0; /* only want offset on first chunk */
+ cnt = ROUNDUP(cnt, iovp_size);
+ do {
+- sba_io_pdir_entry(pdirp, vaddr);
++ sba_io_pdir_entry(ioc, pdirp, vaddr);
+ vaddr += iovp_size;
+ cnt -= iovp_size;
+ pdirp++;
+@@ -1406,7 +1521,11 @@
+ if (likely((ioc->dma_mask & ~to_pci_dev(dev)->dma_mask) == 0)) {
+ for (sg = sglist ; filled < nents ; filled++, sg++){
+ sg->dma_length = sg->length;
+- sg->dma_address = virt_to_phys(sba_sg_address(sg));
++#ifdef CONFIG_XEN
++ sg->dma_address = gnttab_dma_map_page(sg->page) + sg->offset;
++#else
++ sg->dma_address = virt_to_bus(sba_sg_address(sg));
++#endif
+ }
+ return filled;
+ }
+@@ -1429,6 +1548,15 @@
+ #endif
+
+ prefetch(ioc->res_hint);
++
++#ifdef CONFIG_XEN
++ if (is_running_on_xen()) {
++ int i;
++
++ for (i = 0; i < nents; i++)
++ gnttab_dma_use_page(sglist[i].page);
++ }
++#endif
+
+ /*
+ ** First coalesce the chunks and allocate I/O pdir space
+@@ -1562,11 +1690,25 @@
+
+ memset(ioc->pdir_base, 0, ioc->pdir_size);
+
++#ifdef CONFIG_XEN
++ /* The page table needs to be pinned in Xen memory */
++ if (xen_create_contiguous_region((unsigned long)ioc->pdir_base,
++ get_order(ioc->pdir_size), 0))
++ panic(PFX "Couldn't contiguously map I/O Page Table\n");
++
++ ioc->xen_virt_cache = (void *) __get_free_pages(
++ GFP_KERNEL, get_order(ioc->pdir_size));
++ if (!ioc->xen_virt_cache)
++ panic(PFX "Couldn't allocate Xen virtual address cache\n");
++
++ memset(ioc->xen_virt_cache, 0, ioc->pdir_size);
++#endif
++
+ DBG_INIT("%s() IOV page size %ldK pdir %p size %x\n", __FUNCTION__,
+ iovp_size >> 10, ioc->pdir_base, ioc->pdir_size);
+
+ ASSERT(ALIGN((unsigned long) ioc->pdir_base, 4*1024) == (unsigned long) ioc->pdir_base);
+- WRITE_REG(virt_to_phys(ioc->pdir_base), ioc->ioc_hpa + IOC_PDIR_BASE);
++ WRITE_REG(virt_to_bus(ioc->pdir_base), ioc->ioc_hpa + IOC_PDIR_BASE);
+
+ /*
+ ** If an AGP device is present, only use half of the IOV space
+@@ -1603,7 +1745,7 @@
+ for ( ; (u64) poison_addr < addr + iovp_size; poison_addr += poison_size)
+ memcpy(poison_addr, spill_poison, poison_size);
+
+- prefetch_spill_page = virt_to_phys(addr);
++ prefetch_spill_page = virt_to_bus(addr);
+
+ DBG_INIT("%s() prefetch spill addr: 0x%lx\n", __FUNCTION__, prefetch_spill_page);
+ }
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/Makefile
+--- a/arch/ia64/kernel/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -28,6 +28,8 @@
+ obj-$(CONFIG_CPU_FREQ) += cpufreq/
+ obj-$(CONFIG_IA64_MCA_RECOVERY) += mca_recovery.o
+ obj-$(CONFIG_KPROBES) += kprobes.o jprobes.o
++obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o crash.o
++obj-$(CONFIG_CRASH_DUMP) += crash_dump.o
+ obj-$(CONFIG_IA64_UNCACHED_ALLOCATOR) += uncached.o
+ obj-$(CONFIG_AUDIT) += audit.o
+ mca_recovery-y += mca_drv.o mca_drv_asm.o
+@@ -61,3 +63,61 @@
+ # We must build gate.so before we can assemble it.
+ # Note: kbuild does not track this dependency due to usage of .incbin
+ $(obj)/gate-data.o: $(obj)/gate.so
++
++#
++# gate page paravirtualization for xen
++#
++obj-$(CONFIG_XEN) += xengate-data.o
++
++ifeq ($(CONFIG_XEN), y)
++# The gate DSO image is built using a special linker script.
++targets += xengate.so xengate-syms.o
++endif
++
++extra-$(CONFIG_XEN) += xengate.so xengate.lds xengate.o
++
++AFLAGS_xengate.o += -D__XEN_IA64_VDSO_PARAVIRT
++$(obj)/xengate.o: $(src)/gate.S FORCE
++ $(call if_changed_dep,as_o_S)
++
++CPPFLAGS_xengate.lds := -P -C -U$(ARCH) -D__XEN_IA64_VDSO_PARAVIRT
++$(obj)/xengate.lds: $(src)/gate.lds.S
++ $(call if_changed_dep,cpp_lds_S)
++
++GATECFLAGS_xengate.so = -shared -s -Wl,-soname=linux-gate.so.1 \
++ $(call ld-option, -Wl$(comma)--hash-style=sysv)
++$(obj)/xengate.so: $(obj)/xengate.lds $(obj)/xengate.o FORCE
++ $(call if_changed,gate)
++
++ifeq ($(CONFIG_XEN), y)
++$(obj)/built-in.o: $(obj)/xengate-syms.o
++$(obj)/built-in.o: ld_flags += -R $(obj)/xengate-syms.o
++$(obj)/mca_recovery.o: $(obj)/gate-syms.o $(obj)/xengate-syms.o
++endif
++
++GATECFLAGS_xengate-syms.o = -r
++$(obj)/xengate-syms.o: $(obj)/xengate.lds $(obj)/xengate.o FORCE
++ $(call if_changed,gate)
++$(obj)/xengate-data.o: $(obj)/xengate.so
++
++#
++# .tmp_gate.o to calculate padding size for __kernel_syscall_via_epc
++#
++extra-$(CONFIG_XEN) += gate-skip.s .tmp_gate.o
++
++ifeq ($(CONFIG_XEN), y)
++AFLAGS_gate.o += -D__KERNEL_SYSCALL_VIA_EPC_PADDING
++$(obj)/gate.o: $(obj)/gate-skip.s FORCE
++endif
++
++$(obj)/.tmp_gate.o: $(src)/gate.S FORCE
++ $(call if_changed_dep,as_o_S)
++
++quiet_cmd_gate_size = GATE_SIZE $@
++ cmd_gate_size = $(NM) --extern-only --print-size $(obj)/xengate.o | \
++ $(AWK) '/__kernel_syscall_via_epc/{printf "\t.skip 0x"$$2" - "}' > $@; \
++ $(NM) --extern-only --print-size $(obj)/.tmp_gate.o | \
++ $(AWK) '/__kernel_syscall_via_epc/{printf "0x"$$2"\n"}' >> $@
++
++$(obj)/gate-skip.s: $(obj)/xengate.o $(obj)/.tmp_gate.o FORCE
++ $(call if_changed,gate_size)
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/acpi.c
+--- a/arch/ia64/kernel/acpi.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/acpi.c Wed Sep 10 10:54:08 2008 +0100
+@@ -109,6 +109,10 @@
+ return "hpzx1";
+ } else if (!strcmp(hdr->oem_id, "SGI")) {
+ return "sn2";
++#ifdef CONFIG_XEN
++ } else if (is_running_on_xen() && !strcmp(hdr->oem_id, "XEN")) {
++ return "xen";
++#endif
+ }
+
+ return "dig";
+@@ -123,6 +127,8 @@
+ return "sn2";
+ # elif defined (CONFIG_IA64_DIG)
+ return "dig";
++# elif defined (CONFIG_IA64_XEN)
++ return "xen";
+ # else
+ # error Unknown platform. Fix acpi.c.
+ # endif
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/asm-offsets.c
+--- a/arch/ia64/kernel/asm-offsets.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/asm-offsets.c Wed Sep 10 10:54:08 2008 +0100
+@@ -268,4 +268,29 @@
+ DEFINE(IA64_TIME_SOURCE_MMIO64, TIME_SOURCE_MMIO64);
+ DEFINE(IA64_TIME_SOURCE_MMIO32, TIME_SOURCE_MMIO32);
+ DEFINE(IA64_TIMESPEC_TV_NSEC_OFFSET, offsetof (struct timespec, tv_nsec));
++
++#ifdef CONFIG_XEN
++ BLANK();
++
++#define DEFINE_MAPPED_REG_OFS(sym, field) \
++ DEFINE(sym, (XMAPPEDREGS_OFS + offsetof(mapped_regs_t, field)))
++
++ DEFINE_MAPPED_REG_OFS(XSI_PSR_I_ADDR_OFS, interrupt_mask_addr);
++ DEFINE_MAPPED_REG_OFS(XSI_IPSR_OFS, ipsr);
++ DEFINE_MAPPED_REG_OFS(XSI_IIP_OFS, iip);
++ DEFINE_MAPPED_REG_OFS(XSI_IFS_OFS, ifs);
++ DEFINE_MAPPED_REG_OFS(XSI_PRECOVER_IFS_OFS, precover_ifs);
++ DEFINE_MAPPED_REG_OFS(XSI_ISR_OFS, isr);
++ DEFINE_MAPPED_REG_OFS(XSI_IFA_OFS, ifa);
++ DEFINE_MAPPED_REG_OFS(XSI_IIPA_OFS, iipa);
++ DEFINE_MAPPED_REG_OFS(XSI_IIM_OFS, iim);
++ DEFINE_MAPPED_REG_OFS(XSI_IHA_OFS, iha);
++ DEFINE_MAPPED_REG_OFS(XSI_ITIR_OFS, itir);
++ DEFINE_MAPPED_REG_OFS(XSI_PSR_IC_OFS, interrupt_collection_enabled);
++ DEFINE_MAPPED_REG_OFS(XSI_BANKNUM_OFS, banknum);
++ DEFINE_MAPPED_REG_OFS(XSI_BANK0_R16_OFS, bank0_regs[0]);
++ DEFINE_MAPPED_REG_OFS(XSI_BANK1_R16_OFS, bank1_regs[0]);
++ DEFINE_MAPPED_REG_OFS(XSI_B0NATS_OFS, vbnat);
++ DEFINE_MAPPED_REG_OFS(XSI_B1NATS_OFS, vnat);
++#endif /* CONFIG_XEN */
+ }
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/crash.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/kernel/crash.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,222 @@
++/*
++ * arch/ia64/kernel/crash.c
++ *
++ * Architecture specific (ia64) functions for kexec based crash dumps.
++ *
++ * Created by: Khalid Aziz <khalid.aziz at hp.com>
++ * Copyright (C) 2005 Hewlett-Packard Development Company, L.P.
++ * Copyright (C) 2005 Intel Corp Zou Nan hai <nanhai.zou at intel.com>
++ *
++ */
++#include <linux/smp.h>
++#include <linux/delay.h>
++#include <linux/crash_dump.h>
++#include <linux/bootmem.h>
++#include <linux/kexec.h>
++#include <linux/elfcore.h>
++#include <linux/sysctl.h>
++#include <linux/init.h>
++
++#include <asm/kdebug.h>
++#include <asm/mca.h>
++
++int kdump_status[NR_CPUS];
++atomic_t kdump_cpu_freezed;
++atomic_t kdump_in_progress;
++int kdump_on_init = 1;
++
++static inline Elf64_Word
++*append_elf_note(Elf64_Word *buf, char *name, unsigned type, void *data,
++ size_t data_len)
++{
++ struct elf_note *note = (struct elf_note *)buf;
++ note->n_namesz = strlen(name) + 1;
++ note->n_descsz = data_len;
++ note->n_type = type;
++ buf += (sizeof(*note) + 3)/4;
++ memcpy(buf, name, note->n_namesz);
++ buf += (note->n_namesz + 3)/4;
++ memcpy(buf, data, data_len);
++ buf += (data_len + 3)/4;
++ return buf;
++}
++
++static void
++final_note(void *buf)
++{
++ memset(buf, 0, sizeof(struct elf_note));
++}
++
++extern void ia64_dump_cpu_regs(void *);
++
++static DEFINE_PER_CPU(struct elf_prstatus, elf_prstatus);
++
++void
++crash_save_this_cpu()
++{
++ void *buf;
++ unsigned long cfm, sof, sol;
++
++ int cpu = smp_processor_id();
++ struct elf_prstatus *prstatus = &per_cpu(elf_prstatus, cpu);
++
++ elf_greg_t *dst = (elf_greg_t *)&(prstatus->pr_reg);
++ memset(prstatus, 0, sizeof(*prstatus));
++ prstatus->pr_pid = current->pid;
++
++ ia64_dump_cpu_regs(dst);
++ cfm = dst[43];
++ sol = (cfm >> 7) & 0x7f;
++ sof = cfm & 0x7f;
++ dst[46] = (unsigned long)ia64_rse_skip_regs((unsigned long *)dst[46],
++ sof - sol);
++
++ buf = (u64 *) per_cpu_ptr(crash_notes, cpu);
++ if (!buf)
++ return;
++ buf = append_elf_note(buf, "CORE", NT_PRSTATUS, prstatus,
++ sizeof(*prstatus));
++ final_note(buf);
++}
++
++static int
++kdump_wait_cpu_freeze(void)
++{
++ int cpu_num = num_online_cpus() - 1;
++ int timeout = 1000;
++ while(timeout-- > 0) {
++ if (atomic_read(&kdump_cpu_freezed) == cpu_num)
++ return 0;
++ udelay(1000);
++ }
++ return 1;
++}
++
++void
++machine_crash_shutdown(struct pt_regs *pt)
++{
++ /* This function is only called after the system
++ * has paniced or is otherwise in a critical state.
++ * The minimum amount of code to allow a kexec'd kernel
++ * to run successfully needs to happen here.
++ *
++ * In practice this means shooting down the other cpus in
++ * an SMP system.
++ */
++ kexec_disable_iosapic();
++#ifdef CONFIG_SMP
++ kdump_smp_send_stop();
++ if (kdump_wait_cpu_freeze() && kdump_on_init) {
++ //not all cpu response to IPI, send INIT to freeze them
++ kdump_smp_send_init();
++ }
++#endif
++}
++
++static void
++machine_kdump_on_init(void)
++{
++ local_irq_disable();
++ kexec_disable_iosapic();
++ machine_kexec(ia64_kimage);
++}
++
++void
++kdump_cpu_freeze(struct unw_frame_info *info, void *arg)
++{
++ int cpuid;
++ local_irq_disable();
++ cpuid = smp_processor_id();
++ crash_save_this_cpu();
++ current->thread.ksp = (__u64)info->sw - 16;
++ atomic_inc(&kdump_cpu_freezed);
++ kdump_status[cpuid] = 1;
++ mb();
++ if (cpuid == 0) {
++ for (;;)
++ cpu_relax();
++ } else
++ ia64_jump_to_sal(&sal_boot_rendez_state[cpuid]);
++}
++
++static int
++kdump_init_notifier(struct notifier_block *self, unsigned long val, void *data)
++{
++ struct die_args *args = data;
++
++ if (!kdump_on_init)
++ return NOTIFY_DONE;
++
++ if (val != DIE_INIT_MONARCH_ENTER &&
++ val != DIE_INIT_SLAVE_ENTER &&
++ val != DIE_MCA_RENDZVOUS_LEAVE &&
++ val != DIE_MCA_MONARCH_LEAVE)
++ return NOTIFY_DONE;
++
++ /* There really ought to be a check here to see if this
++ * is a machine check rendevous. The kexec code that
++ * was merged around 2.6.20-rc1 includes such a check.
++ * But the check relies on infastructure that is not
++ * available in 2.6.16. */
++
++ switch (val) {
++ case DIE_INIT_MONARCH_ENTER:
++ machine_kdump_on_init();
++ break;
++ case DIE_INIT_SLAVE_ENTER:
++ unw_init_running(kdump_cpu_freeze, NULL);
++ break;
++ case DIE_MCA_RENDZVOUS_LEAVE:
++ if (atomic_read(&kdump_in_progress))
++ unw_init_running(kdump_cpu_freeze, NULL);
++ break;
++ case DIE_MCA_MONARCH_LEAVE:
++ /* die_register->signr indicate if MCA is recoverable */
++ if (!args->signr)
++ machine_kdump_on_init();
++ break;
++ }
++ return NOTIFY_DONE;
++}
++
++#ifdef CONFIG_SYSCTL
++static ctl_table kdump_on_init_table[] = {
++ {
++ .ctl_name = CTL_UNNUMBERED,
++ .procname = "kdump_on_init",
++ .data = &kdump_on_init,
++ .maxlen = sizeof(int),
++ .mode = 0644,
++ .proc_handler = &proc_dointvec,
++ },
++ { .ctl_name = 0 }
++};
++
++static ctl_table sys_table[] = {
++ {
++ .ctl_name = CTL_KERN,
++ .procname = "kernel",
++ .mode = 0555,
++ .child = kdump_on_init_table,
++ },
++ { .ctl_name = 0 }
++};
++#endif
++
++static int
++machine_crash_setup(void)
++{
++ static struct notifier_block kdump_init_notifier_nb = {
++ .notifier_call = kdump_init_notifier,
++ };
++ int ret;
++ if((ret = register_die_notifier(&kdump_init_notifier_nb)) != 0)
++ return ret;
++#ifdef CONFIG_SYSCTL
++ register_sysctl_table(sys_table, 0);
++#endif
++ return 0;
++}
++
++__initcall(machine_crash_setup);
++
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/crash_dump.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/kernel/crash_dump.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,48 @@
++/*
++ * kernel/crash_dump.c - Memory preserving reboot related code.
++ *
++ * Created by: Simon Horman <horms at verge.net.au>
++ * Original code moved from kernel/crash.c
++ * Original code comment copied from the i386 version of this file
++ */
++
++#include <linux/errno.h>
++#include <linux/types.h>
++
++#include <linux/uaccess.h>
++
++/**
++ * copy_oldmem_page - copy one page from "oldmem"
++ * @pfn: page frame number to be copied
++ * @buf: target memory address for the copy; this can be in kernel address
++ * space or user address space (see @userbuf)
++ * @csize: number of bytes to copy
++ * @offset: offset in bytes into the page (based on pfn) to begin the copy
++ * @userbuf: if set, @buf is in user address space, use copy_to_user(),
++ * otherwise @buf is in kernel address space, use memcpy().
++ *
++ * Copy a page from "oldmem". For this page, there is no pte mapped
++ * in the current kernel. We stitch up a pte, similar to kmap_atomic.
++ *
++ * Calling copy_to_user() in atomic context is not desirable. Hence first
++ * copying the data to a pre-allocated kernel page and then copying to user
++ * space in non-atomic context.
++ */
++ssize_t
++copy_oldmem_page(unsigned long pfn, char *buf,
++ size_t csize, unsigned long offset, int userbuf)
++{
++ void *vaddr;
++
++ if (!csize)
++ return 0;
++ vaddr = __va(pfn<<PAGE_SHIFT);
++ if (userbuf) {
++ if (copy_to_user(buf, (vaddr + offset), csize)) {
++ return -EFAULT;
++ }
++ } else
++ memcpy(buf, (vaddr + offset), csize);
++ return csize;
++}
++
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/efi.c
+--- a/arch/ia64/kernel/efi.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/efi.c Wed Sep 10 10:54:08 2008 +0100
+@@ -21,11 +21,13 @@
+ * Skip non-WB memory and ignore empty memory ranges.
+ */
+ #include <linux/module.h>
++#include <linux/bootmem.h>
+ #include <linux/kernel.h>
+ #include <linux/init.h>
+ #include <linux/types.h>
+ #include <linux/time.h>
+ #include <linux/efi.h>
++#include <linux/kexec.h>
+
+ #include <asm/io.h>
+ #include <asm/kregs.h>
+@@ -34,6 +36,11 @@
+ #include <asm/processor.h>
+ #include <asm/mca.h>
+
++#ifdef CONFIG_PROC_IOMEM_MACHINE
++#include <xen/interface/memory.h>
++#include <asm/hypercall.h>
++#endif
++
+ #define EFI_DEBUG 0
+
+ extern efi_status_t efi_call_phys (void *, ...);
+@@ -41,7 +48,7 @@
+ struct efi efi;
+ EXPORT_SYMBOL(efi);
+ static efi_runtime_services_t *runtime;
+-static unsigned long mem_limit = ~0UL, max_addr = ~0UL;
++static unsigned long mem_limit = ~0UL, max_addr = ~0UL, min_addr = 0UL;
+
+ #define efi_call_virt(f, args...) (*(f))(args)
+
+@@ -421,6 +428,8 @@
+ mem_limit = memparse(cp + 4, &cp);
+ } else if (memcmp(cp, "max_addr=", 9) == 0) {
+ max_addr = GRANULEROUNDDOWN(memparse(cp + 9, &cp));
++ } else if (memcmp(cp, "min_addr=", 9) == 0) {
++ min_addr = GRANULEROUNDDOWN(memparse(cp + 9, &cp));
+ } else {
+ while (*cp != ' ' && *cp)
+ ++cp;
+@@ -428,6 +437,8 @@
+ ++cp;
+ }
+ }
++ if (min_addr != 0UL)
++ printk(KERN_INFO "Ignoring memory below %luMB\n", min_addr >> 20);
+ if (max_addr != ~0UL)
+ printk(KERN_INFO "Ignoring memory above %luMB\n", max_addr >> 20);
+
+@@ -894,7 +905,8 @@
+ as = max(contig_low, md->phys_addr);
+ ae = min(contig_high, efi_md_end(md));
+
+- /* keep within max_addr= command line arg */
++ /* keep within max_addr= and min_addr= command line arg */
++ as = max(as, min_addr);
+ ae = min(ae, max_addr);
+ if (ae <= as)
+ continue;
+@@ -965,6 +977,11 @@
+ if (!is_available_memory(md))
+ continue;
+
++#ifdef CONFIG_CRASH_DUMP
++ /* saved_max_pfn should ignore max_addr= command line arg */
++ if (saved_max_pfn < (efi_md_end(md) >> PAGE_SHIFT))
++ saved_max_pfn = (efi_md_end(md) >> PAGE_SHIFT);
++#endif
+ /*
+ * Round ends inward to granule boundaries
+ * Give trimmings to uncached allocator
+@@ -1004,7 +1021,8 @@
+ } else
+ ae = efi_md_end(md);
+
+- /* keep within max_addr= command line arg */
++ /* keep within max_addr= and min_addr= command line arg */
++ as = max(as, min_addr);
+ ae = min(ae, max_addr);
+ if (ae <= as)
+ continue;
+@@ -1033,20 +1051,21 @@
+ *e = (u64)++k;
+ }
+
+-void
+-efi_initialize_iomem_resources(struct resource *code_resource,
+- struct resource *data_resource)
++#define EFI_INITIALISE_PHYS 0x1
++#define EFI_INITIALISE_MACH 0x2
++#define EFI_INITIALISE_ALL (EFI_INITIALISE_PHYS|EFI_INITIALISE_MACH)
++
++static void
++efi_initialize_resources(void *efi_map_start, void *efi_map_end,
++ u64 efi_desc_size, struct resource *root_resource,
++ struct resource *code_resource,
++ struct resource *data_resource, unsigned flag)
+ {
+ struct resource *res;
+- void *efi_map_start, *efi_map_end, *p;
++ void *p;
+ efi_memory_desc_t *md;
+- u64 efi_desc_size;
+ char *name;
+ unsigned long flags;
+-
+- efi_map_start = __va(ia64_boot_param->efi_memmap);
+- efi_map_end = efi_map_start + ia64_boot_param->efi_memmap_size;
+- efi_desc_size = ia64_boot_param->efi_memdesc_size;
+
+ res = NULL;
+
+@@ -1106,7 +1125,7 @@
+ res->end = md->phys_addr + (md->num_pages << EFI_PAGE_SHIFT) - 1;
+ res->flags = flags;
+
+- if (insert_resource(&iomem_resource, res) < 0)
++ if (insert_resource(root_resource, res) < 0)
+ kfree(res);
+ else {
+ /*
+@@ -1114,8 +1133,135 @@
+ * kernel data so we try it repeatedly and
+ * let the resource manager test it.
+ */
+- insert_resource(res, code_resource);
+- insert_resource(res, data_resource);
++ if (flag & EFI_INITIALISE_PHYS) {
++ insert_resource(res, code_resource);
++ insert_resource(res, data_resource);
++ }
++#ifdef CONFIG_KEXEC
++ if (flag & EFI_INITIALISE_MACH) {
++ insert_resource(res, &efi_memmap_res);
++ insert_resource(res, &boot_param_res);
++ if (crashk_res.end > crashk_res.start)
++ insert_resource(res, &crashk_res);
++#ifdef CONFIG_XEN
++ if (is_initial_xendomain())
++ xen_machine_kexec_register_resources(
++ res);
++#endif
++ }
++#endif
+ }
+ }
+ }
++
++#ifdef CONFIG_PROC_IOMEM_MACHINE
++static void
++efi_initialize_iomem_machine_resources(void)
++{
++ unsigned long size;
++ xen_memory_map_t memmap;
++ xen_ia64_memmap_info_t *memmap_info = NULL;
++ void *efi_map_start, *efi_map_end;
++ u64 efi_desc_size;
++ int ret;
++
++ /* It would be nice if it wasn't neccessary to loop like this */
++ for (size = 1024; 1; size += 1024) {
++ memmap_info = kmalloc(size, GFP_KERNEL);
++ if (memmap_info == NULL)
++ return; /* -ENOMEM, but no way to return error */
++
++ memmap.nr_entries = size;
++ set_xen_guest_handle(memmap.buffer, memmap_info);
++ ret = HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap);
++ if (!ret)
++ break;
++
++ kfree(memmap_info);
++ }
++
++ efi_map_start = &memmap_info->memdesc;
++ efi_map_end = efi_map_start + memmap_info->efi_memmap_size;
++ efi_desc_size = memmap_info->efi_memdesc_size;
++ efi_initialize_resources(efi_map_start, efi_map_end, efi_desc_size,
++ &iomem_machine_resource, NULL, NULL,
++ EFI_INITIALISE_MACH);
++
++ kfree(memmap_info);
++}
++#endif
++
++void
++efi_initialize_iomem_resources(struct resource *code_resource,
++ struct resource *data_resource)
++{
++ void *efi_map_start, *efi_map_end;
++ u64 efi_desc_size;
++
++ efi_map_start = __va(ia64_boot_param->efi_memmap);
++ efi_map_end = efi_map_start + ia64_boot_param->efi_memmap_size;
++ efi_desc_size = ia64_boot_param->efi_memdesc_size;
++
++#ifdef CONFIG_PROC_IOMEM_MACHINE
++ if (is_initial_xendomain()) {
++ efi_initialize_resources(efi_map_start, efi_map_end,
++ efi_desc_size, &iomem_resource,
++ code_resource, data_resource,
++ EFI_INITIALISE_PHYS);
++ efi_initialize_iomem_machine_resources();
++ }
++ else
++#endif
++ efi_initialize_resources(efi_map_start, efi_map_end,
++ efi_desc_size, &iomem_resource,
++ code_resource, data_resource,
++ EFI_INITIALISE_ALL);
++}
++
++
++
++#ifdef CONFIG_KEXEC
++/* find a block of memory aligned to 64M exclude reserved regions
++ rsvd_regions are sorted
++ */
++unsigned long
++kdump_find_rsvd_region (unsigned long size,
++ struct rsvd_region *r, int n)
++{
++ int i;
++ u64 start, end;
++ u64 alignment = 1UL << _PAGE_SIZE_64M;
++ void *efi_map_start, *efi_map_end, *p;
++ efi_memory_desc_t *md;
++ u64 efi_desc_size;
++
++ efi_map_start = __va(ia64_boot_param->efi_memmap);
++ efi_map_end = efi_map_start + ia64_boot_param->efi_memmap_size;
++ efi_desc_size = ia64_boot_param->efi_memdesc_size;
++
++ for (p = efi_map_start; p < efi_map_end; p += efi_desc_size) {
++ md = p;
++ if (!efi_wb(md))
++ continue;
++ start = ALIGN(md->phys_addr, alignment);
++ end = efi_md_end(md);
++ for (i = 0; i < n; i++) {
++ if (__pa(r[i].start) >= start && __pa(r[i].end) < end) {
++ if (__pa(r[i].start) > start + size)
++ return start;
++ start = ALIGN(__pa(r[i].end), alignment);
++ if (i < n-1 && __pa(r[i+1].start) < start + size)
++ continue;
++ else
++ break;
++ }
++ }
++ if (end > start + size)
++ return start;
++ }
++
++ printk(KERN_WARNING "Cannot reserve 0x%lx byte of memory for crashdump\n",
++ size);
++ return ~0UL;
++}
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/entry.S
+--- a/arch/ia64/kernel/entry.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/entry.S Wed Sep 10 10:54:08 2008 +0100
+@@ -180,7 +180,7 @@
+ * called. The code starting at .map relies on this. The rest of the code
+ * doesn't care about the interrupt masking status.
+ */
+-GLOBAL_ENTRY(ia64_switch_to)
++GLOBAL_ENTRY(__ia64_switch_to)
+ .prologue
+ alloc r16=ar.pfs,1,0,0,0
+ DO_SAVE_SWITCH_STACK
+@@ -234,7 +234,7 @@
+ ;;
+ srlz.d
+ br.cond.sptk .done
+-END(ia64_switch_to)
++END(__ia64_switch_to)
+
+ /*
+ * Note that interrupts are enabled during save_switch_stack and load_switch_stack. This
+@@ -375,7 +375,7 @@
+ * - b7 holds address to return to
+ * - must not touch r8-r11
+ */
+-ENTRY(load_switch_stack)
++GLOBAL_ENTRY(load_switch_stack)
+ .prologue
+ .altrp b7
+
+@@ -510,7 +510,7 @@
+ * because some system calls (such as ia64_execve) directly
+ * manipulate ar.pfs.
+ */
+-GLOBAL_ENTRY(ia64_trace_syscall)
++GLOBAL_ENTRY(__ia64_trace_syscall)
+ PT_REGS_UNWIND_INFO(0)
+ /*
+ * We need to preserve the scratch registers f6-f11 in case the system
+@@ -582,7 +582,7 @@
+ (p6) mov r10=-1
+ (p6) mov r8=r9
+ br.cond.sptk .strace_save_retval
+-END(ia64_trace_syscall)
++END(__ia64_trace_syscall)
+
+ /*
+ * When traced and returning from sigreturn, we invoke syscall_trace but then
+@@ -601,7 +601,7 @@
+ .ret4: br.cond.sptk ia64_leave_kernel
+ END(ia64_strace_leave_kernel)
+
+-GLOBAL_ENTRY(ia64_ret_from_clone)
++GLOBAL_ENTRY(__ia64_ret_from_clone)
+ PT_REGS_UNWIND_INFO(0)
+ { /*
+ * Some versions of gas generate bad unwind info if the first instruction of a
+@@ -627,7 +627,7 @@
+ cmp.ne p6,p0=r2,r0
+ (p6) br.cond.spnt .strace_check_retval
+ ;; // added stop bits to prevent r8 dependency
+-END(ia64_ret_from_clone)
++END(__ia64_ret_from_clone)
+ // fall through
+ GLOBAL_ENTRY(ia64_ret_from_syscall)
+ PT_REGS_UNWIND_INFO(0)
+@@ -635,8 +635,11 @@
+ adds r2=PT(R8)+16,sp // r2 = &pt_regs.r8
+ mov r10=r0 // clear error indication in r10
+ (p7) br.cond.spnt handle_syscall_error // handle potential syscall failure
++ ;;
++ // don't fall through, ia64_leave_syscall may be #define'd
++ br.cond.sptk.few ia64_leave_syscall
++ ;;
+ END(ia64_ret_from_syscall)
+- // fall through
+ /*
+ * ia64_leave_syscall(): Same as ia64_leave_kernel, except that it doesn't
+ * need to switch to bank 0 and doesn't restore the scratch registers.
+@@ -681,7 +684,7 @@
+ * ar.csd: cleared
+ * ar.ssd: cleared
+ */
+-ENTRY(ia64_leave_syscall)
++GLOBAL_ENTRY(__ia64_leave_syscall)
+ PT_REGS_UNWIND_INFO(0)
+ /*
+ * work.need_resched etc. mustn't get changed by this CPU before it returns to
+@@ -789,7 +792,7 @@
+ mov.m ar.ssd=r0 // M2 clear ar.ssd
+ mov f11=f0 // F clear f11
+ br.cond.sptk.many rbs_switch // B
+-END(ia64_leave_syscall)
++END(__ia64_leave_syscall)
+
+ #ifdef CONFIG_IA32_SUPPORT
+ GLOBAL_ENTRY(ia64_ret_from_ia32_execve)
+@@ -801,10 +804,13 @@
+ st8.spill [r2]=r8 // store return value in slot for r8 and set unat bit
+ .mem.offset 8,0
+ st8.spill [r3]=r0 // clear error indication in slot for r10 and set unat bit
++ ;;
++ // don't fall through, ia64_leave_kernel may be #define'd
++ br.cond.sptk.few ia64_leave_kernel
++ ;;
+ END(ia64_ret_from_ia32_execve)
+- // fall through
+ #endif /* CONFIG_IA32_SUPPORT */
+-GLOBAL_ENTRY(ia64_leave_kernel)
++GLOBAL_ENTRY(__ia64_leave_kernel)
+ PT_REGS_UNWIND_INFO(0)
+ /*
+ * work.need_resched etc. mustn't get changed by this CPU before it returns to
+@@ -1135,7 +1141,7 @@
+ ld8 r10=[r3]
+ br.cond.sptk.many .work_processed_syscall // re-check
+
+-END(ia64_leave_kernel)
++END(__ia64_leave_kernel)
+
+ ENTRY(handle_syscall_error)
+ /*
+@@ -1175,7 +1181,7 @@
+ * be set up by the caller. We declare 8 input registers so the system call
+ * args get preserved, in case we need to restart a system call.
+ */
+-ENTRY(notify_resume_user)
++GLOBAL_ENTRY(notify_resume_user)
+ .prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(8)
+ alloc loc1=ar.pfs,8,2,3,0 // preserve all eight input regs in case of syscall restart!
+ mov r9=ar.unat
+@@ -1263,7 +1269,7 @@
+ adds sp=16,sp
+ ;;
+ ld8 r9=[sp] // load new ar.unat
+- mov.sptk b7=r8,ia64_leave_kernel
++ mov.sptk b7=r8,__ia64_leave_kernel
+ ;;
mov ar.unat=r9
br.many b7
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/fsys.S tmp-linux-2.6-xen.patch/arch/ia64/kernel/fsys.S
---- pristine-linux-2.6.18.2/arch/ia64/kernel/fsys.S 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/fsys.S 2007-07-30 16:35:11.000000000 +0200
-@@ -516,11 +516,34 @@ ENTRY(fsys_fallback_syscall)
+@@ -1575,7 +1581,7 @@
+ data8 sys_mq_timedreceive // 1265
+ data8 sys_mq_notify
+ data8 sys_mq_getsetattr
+- data8 sys_ni_syscall // reserved for kexec_load
++ data8 sys_kexec_load
+ data8 sys_ni_syscall // reserved for vserver
+ data8 sys_waitid // 1270
+ data8 sys_add_key
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/fsys.S
+--- a/arch/ia64/kernel/fsys.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/fsys.S Wed Sep 10 10:54:08 2008 +0100
+@@ -516,11 +516,34 @@
adds r17=-1024,r15
movl r14=sys_call_table
;;
@@ -22475,7 +23931,7 @@
mov r27=ar.rsc
mov r21=ar.fpsr
mov r26=ar.pfs
-@@ -632,7 +655,25 @@ GLOBAL_ENTRY(fsys_bubble_down)
+@@ -632,7 +655,25 @@
mov rp=r14 // I0 set the real return addr
and r3=_TIF_SYSCALL_TRACEAUDIT,r3 // A
;;
@@ -22501,206 +23957,315 @@
cmp.eq p8,p0=r3,r0 // A
(p10) br.cond.spnt.many ia64_ret_from_syscall // B return if bad call-frame or r15 is a NaT
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/gate.lds.S tmp-linux-2.6-xen.patch/arch/ia64/kernel/gate.lds.S
---- pristine-linux-2.6.18.2/arch/ia64/kernel/gate.lds.S 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/gate.lds.S 2007-07-30 16:35:11.000000000 +0200
-@@ -43,6 +43,20 @@ SECTIONS
- __start_gate_brl_fsys_bubble_down_patchlist = .;
- *(.data.patch.brl_fsys_bubble_down)
- __end_gate_brl_fsys_bubble_down_patchlist = .;
-+
-+#ifdef CONFIG_XEN_IA64_VDSO_PARAVIRT
-+ __start_gate_running_on_xen_patchlist = .;
-+ *(.data.patch.running_on_xen)
-+ __end_gate_running_on_xen_patchlist = .;
-+
-+ __start_gate_brl_xen_ssm_i_0_patchlist = .;
-+ *(.data.patch.brl_xen_ssm_i_0)
-+ __end_gate_brl_xen_ssm_i_0_patchlist = .;
-+
-+ __start_gate_brl_xen_ssm_i_1_patchlist = .;
-+ *(.data.patch.brl_xen_ssm_i_1)
-+ __end_gate_brl_xen_ssm_i_1_patchlist = .;
-+#endif
- } :readable
- .IA_64.unwind_info : { *(.IA_64.unwind_info*) }
- .IA_64.unwind : { *(.IA_64.unwind*) } :readable :unwind
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/gate.S tmp-linux-2.6-xen.patch/arch/ia64/kernel/gate.S
---- pristine-linux-2.6.18.2/arch/ia64/kernel/gate.S 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/gate.S 2007-07-30 16:35:11.000000000 +0200
-@@ -13,6 +13,9 @@
- #include <asm/sigcontext.h>
- #include <asm/system.h>
- #include <asm/unistd.h>
-+#ifdef CONFIG_XEN_IA64_VDSO_PARAVIRT
-+# include <asm/privop.h>
-+#endif
-
- /*
- * We can't easily refer to symbols inside the kernel. To avoid full runtime relocation,
-@@ -32,6 +35,40 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/gate.S
+--- a/arch/ia64/kernel/gate.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/gate.S Wed Sep 10 10:54:08 2008 +0100
+@@ -31,102 +31,6 @@
+ #define BRL_COND_FSYS_BUBBLE_DOWN(pr) \
[1:](pr)brl.cond.sptk 0; \
.xdata4 ".data.patch.brl_fsys_bubble_down", 1b-.
-
-+#ifdef CONFIG_XEN_IA64_VDSO_PARAVIRT
-+ // The page in which hyperprivop lives must be pinned by ITR.
-+ // However vDSO area isn't pinned. So issuing hyperprivop
-+ // from vDSO page causes trouble that Kevin pointed out.
-+ // After clearing vpsr.ic, the vcpu is pre-empted and the itlb
-+ // is flushed. Then vcpu get cpu again, tlb miss fault occures.
-+ // However it results in nested dtlb fault because vpsr.ic is off.
-+ // To avoid such a situation, we jump into the kernel text area
-+ // which is pinned, and then issue hyperprivop and return back
-+ // to vDSO page.
-+ // This is Dan Magenheimer's idea.
-+
-+ // Currently is_running_on_xen() is defined as running_on_xen.
-+ // If is_running_on_xen() is a real function, we must update
-+ // according to it.
-+ .section ".data.patch.running_on_xen", "a"
-+ .previous
-+#define LOAD_RUNNING_ON_XEN(reg) \
-+[1:] movl reg=0; \
-+ .xdata4 ".data.patch.running_on_xen", 1b-.
+-
+-GLOBAL_ENTRY(__kernel_syscall_via_break)
+- .prologue
+- .altrp b6
+- .body
+- /*
+- * Note: for (fast) syscall restart to work, the break instruction must be
+- * the first one in the bundle addressed by syscall_via_break.
+- */
+-{ .mib
+- break 0x100000
+- nop.i 0
+- br.ret.sptk.many b6
+-}
+-END(__kernel_syscall_via_break)
+-
+-/*
+- * On entry:
+- * r11 = saved ar.pfs
+- * r15 = system call #
+- * b0 = saved return address
+- * b6 = return address
+- * On exit:
+- * r11 = saved ar.pfs
+- * r15 = system call #
+- * b0 = saved return address
+- * all other "scratch" registers: undefined
+- * all "preserved" registers: same as on entry
+- */
+-
+-GLOBAL_ENTRY(__kernel_syscall_via_epc)
+- .prologue
+- .altrp b6
+- .body
+-{
+- /*
+- * Note: the kernel cannot assume that the first two instructions in this
+- * bundle get executed. The remaining code must be safe even if
+- * they do not get executed.
+- */
+- adds r17=-1024,r15 // A
+- mov r10=0 // A default to successful syscall execution
+- epc // B causes split-issue
+-}
+- ;;
+- rsm psr.be | psr.i // M2 (5 cyc to srlz.d)
+- LOAD_FSYSCALL_TABLE(r14) // X
+- ;;
+- mov r16=IA64_KR(CURRENT) // M2 (12 cyc)
+- shladd r18=r17,3,r14 // A
+- mov r19=NR_syscalls-1 // A
+- ;;
+- lfetch [r18] // M0|1
+- mov r29=psr // M2 (12 cyc)
+- // If r17 is a NaT, p6 will be zero
+- cmp.geu p6,p7=r19,r17 // A (sysnr > 0 && sysnr < 1024+NR_syscalls)?
+- ;;
+- mov r21=ar.fpsr // M2 (12 cyc)
+- tnat.nz p10,p9=r15 // I0
+- mov.i r26=ar.pfs // I0 (would stall anyhow due to srlz.d...)
+- ;;
+- srlz.d // M0 (forces split-issue) ensure PSR.BE==0
+-(p6) ld8 r18=[r18] // M0|1
+- nop.i 0
+- ;;
+- nop.m 0
+-(p6) tbit.z.unc p8,p0=r18,0 // I0 (dual-issues with "mov b7=r18"!)
+- nop.i 0
+- ;;
+-(p8) ssm psr.i
+-(p6) mov b7=r18 // I0
+-(p8) br.dptk.many b7 // B
+-
+- mov r27=ar.rsc // M2 (12 cyc)
+-/*
+- * brl.cond doesn't work as intended because the linker would convert this branch
+- * into a branch to a PLT. Perhaps there will be a way to avoid this with some
+- * future version of the linker. In the meantime, we just use an indirect branch
+- * instead.
+- */
+-#ifdef CONFIG_ITANIUM
+-(p6) add r14=-8,r14 // r14 <- addr of fsys_bubble_down entry
+- ;;
+-(p6) ld8 r14=[r14] // r14 <- fsys_bubble_down
+- ;;
+-(p6) mov b7=r14
+-(p6) br.sptk.many b7
+-#else
+- BRL_COND_FSYS_BUBBLE_DOWN(p6)
+-#endif
+- ssm psr.i
+- mov r10=-1
+-(p10) mov r8=EINVAL
+-(p9) mov r8=ENOSYS
+- FSYS_RETURN
+-END(__kernel_syscall_via_epc)
+
+ # define ARG0_OFF (16 + IA64_SIGFRAME_ARG0_OFFSET)
+ # define ARG1_OFF (16 + IA64_SIGFRAME_ARG1_OFFSET)
+@@ -373,3 +277,154 @@
+ // invala not necessary as that will happen when returning to user-mode
+ br.cond.sptk back_from_restore_rbs
+ END(__kernel_sigtramp)
+
-+ .section ".data.patch.brl_xen_ssm_i_0", "a"
-+ .previous
-+#define BRL_COND_XEN_SSM_I_0(pr) \
-+[1:](pr)brl.cond.sptk 0; \
-+ .xdata4 ".data.patch.brl_xen_ssm_i_0", 1b-.
++GLOBAL_ENTRY(__kernel_syscall_via_break)
++ .prologue
++ .altrp b6
++ .body
++ /*
++ * Note: for (fast) syscall restart to work, the break instruction must be
++ * the first one in the bundle addressed by syscall_via_break.
++ */
++{ .mib
++ break 0x100000
++ nop.i 0
++ br.ret.sptk.many b6
++}
++END(__kernel_syscall_via_break)
+
-+ .section ".data.patch.brl_xen_ssm_i_1", "a"
-+ .previous
-+#define BRL_COND_XEN_SSM_I_1(pr) \
-+[1:](pr)brl.cond.sptk 0; \
-+ .xdata4 ".data.patch.brl_xen_ssm_i_1", 1b-.
-+#endif
++/*
++ * On entry:
++ * r11 = saved ar.pfs
++ * r15 = system call #
++ * b0 = saved return address
++ * b6 = return address
++ * On exit:
++ * r11 = saved ar.pfs
++ * r15 = system call #
++ * b0 = saved return address
++ * all other "scratch" registers: undefined
++ * all "preserved" registers: same as on entry
++ */
+
- GLOBAL_ENTRY(__kernel_syscall_via_break)
- .prologue
- .altrp b6
-@@ -76,7 +113,42 @@ GLOBAL_ENTRY(__kernel_syscall_via_epc)
- epc // B causes split-issue
- }
- ;;
-+#ifdef CONFIG_XEN_IA64_VDSO_PARAVIRT
++GLOBAL_ENTRY(__kernel_syscall_via_epc)
++ .prologue
++ .altrp b6
++ .body
++{
++ /*
++ * Note: the kernel cannot assume that the first two instructions in this
++ * bundle get executed. The remaining code must be safe even if
++ * they do not get executed.
++ */
++ adds r17=-1024,r15 // A
++ mov r10=0 // A default to successful syscall execution
++ epc // B causes split-issue
++}
++ ;;
++#ifdef __XEN_IA64_VDSO_PARAVIRT
+ // r20 = 1
+ // r22 = &vcpu->vcpu_info->evtchn_upcall_mask
-+ // r23 = &vpsr.ic
+ // r24 = &vcpu->vcpu_info->evtchn_upcall_pending
+ // r25 = tmp
-+ // r28 = &running_on_xen
-+ // r30 = running_on_xen
+ // r31 = tmp
+ // p11 = tmp
-+ // p12 = running_on_xen
-+ // p13 = !running_on_xen
+ // p14 = tmp
-+ // p15 = tmp
-+#define isXen p12
-+#define isRaw p13
-+ LOAD_RUNNING_ON_XEN(r28)
++ mov r20=1
+ movl r22=XSI_PSR_I_ADDR
+ ;;
+ ld8 r22=[r22]
+ ;;
-+ movl r23=XSI_PSR_IC
++ st1 [r22]=r20
++ rum psr.be
+ adds r24=-1,r22
-+ mov r20=1
-+ ;;
-+ ld4 r30=[r28]
-+ ;;
-+ cmp.ne isXen,isRaw=r0,r30
++#else
++ rsm psr.be | psr.i // M2 (5 cyc to srlz.d)
++#endif
++ LOAD_FSYSCALL_TABLE(r14) // X
+ ;;
-+(isRaw) rsm psr.be | psr.i
-+(isXen) st1 [r22]=r20
-+(isXen) rum psr.be
++ mov r16=IA64_KR(CURRENT) // M2 (12 cyc)
++ shladd r18=r17,3,r14 // A
++ mov r19=NR_syscalls-1 // A
++#ifdef __XEN_IA64_VDSO_PARAVIRT
++ XEN_HYPER_GET_PSR
+ ;;
++ lfetch [r18] // M0|1
++ mov r29=r8
+#else
- rsm psr.be | psr.i // M2 (5 cyc to srlz.d)
++ ;;
++ lfetch [r18] // M0|1
++ mov r29=psr // M2 (12 cyc)
+#endif
- LOAD_FSYSCALL_TABLE(r14) // X
- ;;
- mov r16=IA64_KR(CURRENT) // M2 (12 cyc)
-@@ -84,7 +156,14 @@ GLOBAL_ENTRY(__kernel_syscall_via_epc)
- mov r19=NR_syscalls-1 // A
- ;;
- lfetch [r18] // M0|1
-+#ifdef CONFIG_XEN_IA64_VDSO_PARAVIRT
-+(isRaw) mov r29=psr
-+(isXen) XEN_HYPER_GET_PSR
++ // If r17 is a NaT, p6 will be zero
++ cmp.geu p6,p7=r19,r17 // A (sysnr > 0 && sysnr < 1024+NR_syscalls)?
++ ;;
++ mov r21=ar.fpsr // M2 (12 cyc)
++ tnat.nz p10,p9=r15 // I0
++ mov.i r26=ar.pfs // I0 (would stall anyhow due to srlz.d...)
++ ;;
++ srlz.d // M0 (forces split-issue) ensure PSR.BE==0
++(p6) ld8 r18=[r18] // M0|1
++ nop.i 0
+ ;;
-+(isXen) mov r29=r8
++ nop.m 0
++(p6) tbit.z.unc p8,p0=r18,0 // I0 (dual-issues with "mov b7=r18"!)
++#ifdef __XEN_IA64_VDSO_PARAVIRT
++
++#define XEN_SET_PSR_I(pred) \
++(pred) ld1 r31=[r22]; \
++ ;; ; \
++(pred) st1 [r22]=r0; \
++(pred) cmp.ne.unc p14,p0=r0,r31; \
++ ;; ; \
++(p14) ld1 r25=[r24]; \
++ ;; ; \
++(p14) cmp.ne.unc p11,p0=r0,r25; \
++ ;; ; \
++(p11) XEN_HYPER_SSM_I;
++
++ ;;
++ XEN_SET_PSR_I(p8)
+#else
- mov r29=psr // M2 (12 cyc)
++ nop.i 0
++ ;;
++(p8) ssm psr.i
+#endif
- // If r17 is a NaT, p6 will be zero
- cmp.geu p6,p7=r19,r17 // A (sysnr > 0 && sysnr < 1024+NR_syscalls)?
- ;;
-@@ -98,9 +177,21 @@ GLOBAL_ENTRY(__kernel_syscall_via_epc)
- ;;
- nop.m 0
- (p6) tbit.z.unc p8,p0=r18,0 // I0 (dual-issues with "mov b7=r18"!)
-+#ifdef CONFIG_XEN_IA64_VDSO_PARAVIRT
-+ ;;
-+ // p14 = running_on_xen && p8
-+ // p15 = !running_on_xen && p8
-+(p8) cmp.ne.unc p14,p15=r0,r30
++(p6) mov b7=r18 // I0
++(p8) br.dptk.many b7 // B
++
++ mov r27=ar.rsc // M2 (12 cyc)
++/*
++ * brl.cond doesn't work as intended because the linker would convert this branch
++ * into a branch to a PLT. Perhaps there will be a way to avoid this with some
++ * future version of the linker. In the meantime, we just use an indirect branch
++ * instead.
++ */
++#ifdef CONFIG_ITANIUM
++(p6) add r14=-8,r14 // r14 <- addr of fsys_bubble_down entry
+ ;;
-+(p15) ssm psr.i
-+ BRL_COND_XEN_SSM_I_0(p14)
-+ .global .vdso_ssm_i_0_ret
-+.vdso_ssm_i_0_ret:
++(p6) ld8 r14=[r14] // r14 <- fsys_bubble_down
++ ;;
++(p6) mov b7=r14
++(p6) br.sptk.many b7
+#else
- nop.i 0
- ;;
- (p8) ssm psr.i
++ BRL_COND_FSYS_BUBBLE_DOWN(p6)
+#endif
- (p6) mov b7=r18 // I0
- (p8) br.dptk.many b7 // B
-
-@@ -121,9 +212,21 @@ GLOBAL_ENTRY(__kernel_syscall_via_epc)
- #else
- BRL_COND_FSYS_BUBBLE_DOWN(p6)
- #endif
-+#ifdef CONFIG_XEN_IA64_VDSO_PARAVIRT
-+(isRaw) ssm psr.i
-+ BRL_COND_XEN_SSM_I_1(isXen)
-+ .global .vdso_ssm_i_1_ret
-+.vdso_ssm_i_1_ret:
-+#else
- ssm psr.i
-+#endif
- mov r10=-1
- (p10) mov r8=EINVAL
-+#ifdef CONFIG_XEN_IA64_VDSO_PARAVIRT
-+ dv_serialize_data // shut up gas warning.
-+ // we know xen_hyper_ssm_i_0 or xen_hyper_ssm_i_1
-+ // doesn't change p9 and p10
-+#endif
- (p9) mov r8=ENOSYS
- FSYS_RETURN
- END(__kernel_syscall_via_epc)
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/head.S tmp-linux-2.6-xen.patch/arch/ia64/kernel/head.S
---- pristine-linux-2.6.18.2/arch/ia64/kernel/head.S 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/head.S 2007-07-30 16:35:11.000000000 +0200
-@@ -367,6 +367,12 @@ start_ap:
++#ifdef __XEN_IA64_VDSO_PARAVIRT
++ XEN_SET_PSR_I(p0)
++#else
++ ssm psr.i
++#endif
++ mov r10=-1
++(p10) mov r8=EINVAL
++(p9) mov r8=ENOSYS
++ FSYS_RETURN
++#ifdef __KERNEL_SYSCALL_VIA_EPC_PADDING
++ /*
++ * All values/sizes of __kernel_xxx symbol in gate.so and xengate.so
++ * must be same to each other.
++ * Adjust symbol size in gate.so to be same to the one in xengate.so.
++ */
++.include "arch/ia64/kernel/gate-skip.s"
++#endif
++END(__kernel_syscall_via_epc)
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/gate.lds.S
+--- a/arch/ia64/kernel/gate.lds.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/gate.lds.S Wed Sep 10 10:54:08 2008 +0100
+@@ -28,6 +28,24 @@
+ . = GATE_ADDR + 0x500;
+
+ .data.patch : {
++#ifdef __XEN_IA64_VDSO_PARAVIRT
++#define __start_gate_mckinley_e9_patchlist \
++ __start_gate_mckinley_e9_patchlist_xen
++#define __end_gate_mckinley_e9_patchlist \
++ __end_gate_mckinley_e9_patchlist_xen
++#define __start_gate_vtop_patchlist \
++ __start_gate_vtop_patchlist_xen
++#define __end_gate_vtop_patchlist \
++ __end_gate_vtop_patchlist_xen
++#define __start_gate_fsyscall_patchlist \
++ __start_gate_fsyscall_patchlist_xen
++#define __end_gate_fsyscall_patchlist \
++ __end_gate_fsyscall_patchlist_xen
++#define __start_gate_brl_fsys_bubble_down_patchlist \
++ __start_gate_brl_fsys_bubble_down_patchlist_xen
++#define __end_gate_brl_fsys_bubble_down_patchlist \
++ __end_gate_brl_fsys_bubble_down_patchlist_xen
++#endif
+ __start_gate_mckinley_e9_patchlist = .;
+ *(.data.patch.mckinley_e9)
+ __end_gate_mckinley_e9_patchlist = .;
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/head.S
+--- a/arch/ia64/kernel/head.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/head.S Wed Sep 10 10:54:08 2008 +0100
+@@ -366,6 +366,12 @@
+ (isBP) movl r2=ia64_boot_param
;;
(isBP) st8 [r2]=r28 // save the address of the boot param area passed by the bootloader
-
++
+#ifdef CONFIG_XEN
+ // Note: isBP is used by the subprogram.
+ br.call.sptk.many rp=early_xen_setup
+ ;;
+#endif
-+
+
#ifdef CONFIG_SMP
(isAP) br.call.sptk.many rp=start_secondary
- .ret0:
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/iosapic.c tmp-linux-2.6-xen.patch/arch/ia64/kernel/iosapic.c
---- pristine-linux-2.6.18.2/arch/ia64/kernel/iosapic.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/iosapic.c 2007-07-30 16:35:11.000000000 +0200
-@@ -159,6 +159,75 @@ static unsigned char pcat_compat __devin
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/iosapic.c
+--- a/arch/ia64/kernel/iosapic.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/iosapic.c Wed Sep 10 10:54:08 2008 +0100
+@@ -159,6 +159,75 @@
static int iosapic_kmalloc_ok;
static LIST_HEAD(free_rte_list);
@@ -22776,17 +24341,45 @@
/*
* Find an IOSAPIC associated with a GSI
*/
-@@ -653,6 +722,9 @@ register_intr (unsigned int gsi, int vec
+@@ -287,6 +356,27 @@
+ {
+ /* do nothing... */
+ }
++
++
++#ifdef CONFIG_KEXEC
++void
++kexec_disable_iosapic(void)
++{
++ struct iosapic_intr_info *info;
++ struct iosapic_rte_info *rte;
++ u8 vec = 0;
++ for (info = iosapic_intr_info; info <
++ iosapic_intr_info + IA64_NUM_VECTORS; ++info, ++vec) {
++ list_for_each_entry(rte, &info->rtes,
++ rte_list) {
++ iosapic_write(rte->addr,
++ IOSAPIC_RTE_LOW(rte->rte_index),
++ IOSAPIC_MASK|vec);
++ iosapic_eoi(rte->addr, vec);
++ }
++ }
++}
++#endif
+
+ static void
+ mask_irq (unsigned int irq)
+@@ -652,6 +742,9 @@
+ iosapic_intr_info[vector].polarity = polarity;
iosapic_intr_info[vector].dmode = delivery;
iosapic_intr_info[vector].trigger = trigger;
-
++
+ if (is_running_on_xen())
+ return 0;
-+
+
if (trigger == IOSAPIC_EDGE)
irq_type = &irq_type_iosapic_edge;
- else
-@@ -1015,6 +1087,9 @@ iosapic_system_init (int system_pcat_com
+@@ -1015,6 +1108,9 @@
}
pcat_compat = system_pcat_compat;
@@ -22796,9 +24389,9 @@
if (pcat_compat) {
/*
* Disable the compatibility mode interrupts (8259 style),
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/irq_ia64.c tmp-linux-2.6-xen.patch/arch/ia64/kernel/irq_ia64.c
---- pristine-linux-2.6.18.2/arch/ia64/kernel/irq_ia64.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/irq_ia64.c 2007-09-30 18:06:18.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/irq_ia64.c
+--- a/arch/ia64/kernel/irq_ia64.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/irq_ia64.c Wed Sep 10 10:54:08 2008 +0100
@@ -30,6 +30,9 @@
#include <linux/smp_lock.h>
#include <linux/threads.h>
@@ -22809,7 +24402,7 @@
#include <asm/delay.h>
#include <asm/intrinsics.h>
-@@ -69,6 +72,13 @@ int
+@@ -69,6 +72,13 @@
assign_irq_vector (int irq)
{
int pos, vector;
@@ -22823,7 +24416,7 @@
again:
pos = find_first_zero_bit(ia64_vector_mask, IA64_NUM_DEVICE_VECTORS);
vector = IA64_FIRST_DEVICE_VECTOR + pos;
-@@ -87,6 +97,13 @@ free_irq_vector (int vector)
+@@ -87,6 +97,13 @@
if (vector < IA64_FIRST_DEVICE_VECTOR || vector > IA64_LAST_DEVICE_VECTOR)
return;
@@ -22837,7 +24430,7 @@
pos = vector - IA64_FIRST_DEVICE_VECTOR;
if (!test_and_clear_bit(pos, ia64_vector_mask))
printk(KERN_WARNING "%s: double free!\n", __FUNCTION__);
-@@ -240,12 +257,277 @@ static struct irqaction ipi_irqaction =
+@@ -240,11 +257,341 @@
};
#endif
@@ -23098,72 +24691,45 @@
+#endif /* CONFIG_SMP */
+}
+
-+#endif /* CONFIG_XEN */
++void
++xen_irq_init(void)
++{
++ struct callback_register event = {
++ .type = CALLBACKTYPE_event,
++ .address = (unsigned long)&xen_event_callback,
++ };
+
- void
- register_percpu_irq (ia64_vector vec, struct irqaction *action)
- {
- irq_desc_t *desc;
- unsigned int irq;
-
-+#ifdef CONFIG_XEN
-+ if (is_running_on_xen())
-+ return xen_register_percpu_irq(smp_processor_id(),
-+ vec, action, 1);
++ xen_init_IRQ();
++ BUG_ON(HYPERVISOR_callback_op(CALLBACKOP_register, &event));
++ late_time_init = xen_bind_early_percpu_irq;
++#ifdef CONFIG_SMP
++ register_percpu_irq(IA64_IPI_RESCHEDULE, &resched_irqaction);
+#endif
++}
+
- for (irq = 0; irq < NR_IRQS; ++irq)
- if (irq_to_vector(irq) == vec) {
- desc = irq_desc + irq;
-@@ -259,6 +541,21 @@ register_percpu_irq (ia64_vector vec, st
- void __init
- init_IRQ (void)
- {
-+#ifdef CONFIG_XEN
-+ /* Maybe put into platform_irq_init later */
-+ if (is_running_on_xen()) {
-+ struct callback_register event = {
-+ .type = CALLBACKTYPE_event,
-+ .address = (unsigned long)&xen_event_callback,
-+ };
-+ xen_init_IRQ();
-+ BUG_ON(HYPERVISOR_callback_op(CALLBACKOP_register, &event));
-+ late_time_init = xen_bind_early_percpu_irq;
-+#ifdef CONFIG_SMP
-+ register_percpu_irq(IA64_IPI_RESCHEDULE, &resched_irqaction);
-+#endif /* CONFIG_SMP */
-+ }
-+#endif /* CONFIG_XEN */
- register_percpu_irq(IA64_SPURIOUS_INT_VECTOR, NULL);
- #ifdef CONFIG_SMP
- register_percpu_irq(IA64_IPI_VECTOR, &ipi_irqaction);
-@@ -276,6 +573,54 @@ ia64_send_ipi (int cpu, int vector, int
- unsigned long ipi_data;
- unsigned long phys_cpu_id;
-
-+#ifdef CONFIG_XEN
-+ if (is_running_on_xen()) {
-+ int irq = -1;
++void
++xen_platform_send_ipi(int cpu, int vector, int delivery_mode, int redirect)
++{
++ int irq = -1;
++ extern void xen_send_ipi(int cpu, int vec);
+
+#ifdef CONFIG_SMP
-+ /* TODO: we need to call vcpu_up here */
-+ if (unlikely(vector == ap_wakeup_vector)) {
-+ extern void xen_send_ipi (int cpu, int vec);
-+
-+ /* XXX
-+ * This should be in __cpu_up(cpu) in ia64 smpboot.c
-+ * like x86. But don't want to modify it,
-+ * keep it untouched.
-+ */
-+ xen_smp_intr_init_early(cpu);
++ /* TODO: we need to call vcpu_up here */
++ if (unlikely(vector == ap_wakeup_vector)) {
++ /* XXX
++ * This should be in __cpu_up(cpu) in ia64 smpboot.c
++ * like x86. But don't want to modify it,
++ * keep it untouched.
++ */
++ xen_smp_intr_init_early(cpu);
+
-+ xen_send_ipi (cpu, vector);
-+ //vcpu_prepare_and_up(cpu);
-+ return;
-+ }
++ xen_send_ipi (cpu, vector);
++ //vcpu_prepare_and_up(cpu);
++ return;
++ }
+#endif
+
-+ switch(vector) {
++ switch (vector) {
+ case IA64_IPI_VECTOR:
+ irq = per_cpu(ipi_to_irq, cpu)[IPI_VECTOR];
+ break;
@@ -23176,25 +24742,331 @@
+ case IA64_CPEP_VECTOR:
+ irq = per_cpu(ipi_to_irq, cpu)[CPEP_VECTOR];
+ break;
++ case IA64_TIMER_VECTOR:
++ xen_send_ipi(cpu, vector);
++ return;
+ default:
+ printk(KERN_WARNING "Unsupported IPI type 0x%x\n",
+ vector);
+ irq = 0;
+ break;
-+ }
++ }
+
-+ BUG_ON(irq < 0);
-+ notify_remote_via_irq(irq);
-+ return;
-+ }
++ BUG_ON(irq < 0);
++ notify_remote_via_irq(irq);
++ return;
++}
+#endif /* CONFIG_XEN */
+
+ void
+ register_percpu_irq (ia64_vector vec, struct irqaction *action)
+ {
+ irq_desc_t *desc;
+ unsigned int irq;
++
++#ifdef CONFIG_XEN
++ if (is_running_on_xen())
++ return xen_register_percpu_irq(smp_processor_id(),
++ vec, action, 1);
++#endif
+
+ for (irq = 0; irq < NR_IRQS; ++irq)
+ if (irq_to_vector(irq) == vec) {
+@@ -267,6 +614,10 @@
+ pfm_init_percpu();
+ #endif
+ platform_irq_init();
++#ifdef CONFIG_XEN
++ if (is_running_on_xen() && !ia64_platform_is("xen"))
++ xen_irq_init();
++#endif
+ }
+
+ void
+@@ -275,6 +626,13 @@
+ void __iomem *ipi_addr;
+ unsigned long ipi_data;
+ unsigned long phys_cpu_id;
++
++#ifdef CONFIG_XEN
++ if (is_running_on_xen()) {
++ xen_platform_send_ipi(cpu, vector, delivery_mode, redirect);
++ return;
++ }
++#endif
+
#ifdef CONFIG_SMP
phys_cpu_id = cpu_physical_id(cpu);
- #else
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/pal.S tmp-linux-2.6-xen.patch/arch/ia64/kernel/pal.S
---- pristine-linux-2.6.18.2/arch/ia64/kernel/pal.S 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/pal.S 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/machine_kexec.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/kernel/machine_kexec.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,204 @@
++/*
++ * arch/ia64/kernel/machine_kexec.c
++ *
++ * Handle transition of Linux booting another kernel
++ * Copyright (C) 2005 Hewlett-Packard Development Comapny, L.P.
++ * Copyright (C) 2005 Khalid Aziz <khalid.aziz at hp.com>
++ * Copyright (C) 2006 Intel Corp, Zou Nan hai <nanhai.zou at intel.com>
++ *
++ * This source code is licensed under the GNU General Public License,
++ * Version 2. See the file COPYING for more details.
++ */
++
++#include <linux/mm.h>
++#include <linux/kexec.h>
++#include <linux/cpu.h>
++#include <linux/irq.h>
++#include <asm/mmu_context.h>
++#include <asm/setup.h>
++#include <asm/delay.h>
++#include <asm/meminit.h>
++#ifdef CONFIG_XEN
++#include <xen/interface/kexec.h>
++#include <asm/kexec.h>
++#endif
++
++typedef void (*relocate_new_kernel_t)(unsigned long, unsigned long,
++ struct ia64_boot_param *, unsigned long);
++
++struct kimage *ia64_kimage;
++
++struct resource efi_memmap_res = {
++ .name = "EFI Memory Map",
++ .start = 0,
++ .end = 0,
++ .flags = IORESOURCE_BUSY | IORESOURCE_MEM
++};
++
++struct resource boot_param_res = {
++ .name = "Boot parameter",
++ .start = 0,
++ .end = 0,
++ .flags = IORESOURCE_BUSY | IORESOURCE_MEM
++};
++
++
++/*
++ * Do what every setup is needed on image and the
++ * reboot code buffer to allow us to avoid allocations
++ * later.
++ */
++int machine_kexec_prepare(struct kimage *image)
++{
++ void *control_code_buffer;
++ const unsigned long *func;
++
++ func = (unsigned long *)&relocate_new_kernel;
++ /* Pre-load control code buffer to minimize work in kexec path */
++ control_code_buffer = page_address(image->control_code_page);
++ memcpy((void *)control_code_buffer, (const void *)func[0],
++ relocate_new_kernel_size);
++ flush_icache_range((unsigned long)control_code_buffer,
++ (unsigned long)control_code_buffer + relocate_new_kernel_size);
++ ia64_kimage = image;
++
++ return 0;
++}
++
++void machine_kexec_cleanup(struct kimage *image)
++{
++}
++
++#ifndef CONFIG_XEN
++void machine_shutdown(void)
++{
++ int cpu;
++
++ for_each_online_cpu(cpu) {
++ if (cpu != smp_processor_id())
++ cpu_down(cpu);
++ }
++ kexec_disable_iosapic();
++}
++
++/*
++ * Do not allocate memory (or fail in any way) in machine_kexec().
++ * We are past the point of no return, committed to rebooting now.
++ */
++extern void *efi_get_pal_addr(void);
++static void ia64_machine_kexec(struct unw_frame_info *info, void *arg)
++{
++ struct kimage *image = arg;
++ relocate_new_kernel_t rnk;
++ void *pal_addr = efi_get_pal_addr();
++ unsigned long code_addr = (unsigned long)page_address(image->control_code_page);
++ unsigned long vector;
++ int ii;
++
++ if (image->type == KEXEC_TYPE_CRASH) {
++ crash_save_this_cpu();
++ current->thread.ksp = (__u64)info->sw - 16;
++ }
++
++ /* Interrupts aren't acceptable while we reboot */
++ local_irq_disable();
++
++ /* Mask CMC and Performance Monitor interrupts */
++ ia64_setreg(_IA64_REG_CR_PMV, 1 << 16);
++ ia64_setreg(_IA64_REG_CR_CMCV, 1 << 16);
++
++ /* Mask ITV and Local Redirect Registers */
++ ia64_set_itv(1 << 16);
++ ia64_set_lrr0(1 << 16);
++ ia64_set_lrr1(1 << 16);
++
++ /* terminate possible nested in-service interrupts */
++ for (ii = 0; ii < 16; ii++)
++ ia64_eoi();
++
++ /* unmask TPR and clear any pending interrupts */
++ ia64_setreg(_IA64_REG_CR_TPR, 0);
++ ia64_srlz_d();
++ vector = ia64_get_ivr();
++ while (vector != IA64_SPURIOUS_INT_VECTOR) {
++ ia64_eoi();
++ vector = ia64_get_ivr();
++ }
++ platform_kernel_launch_event();
++ rnk = (relocate_new_kernel_t)&code_addr;
++ (*rnk)(image->head, image->start, ia64_boot_param,
++ GRANULEROUNDDOWN((unsigned long) pal_addr));
++ BUG();
++}
++
++void machine_kexec(struct kimage *image)
++{
++ unw_init_running(ia64_machine_kexec, image);
++ for(;;);
++}
++#else /* CONFIG_XEN */
++void machine_kexec_setup_load_arg(xen_kexec_image_t *xki,struct kimage *image)
++{
++ xki->reboot_code_buffer =
++ kexec_page_to_pfn(image->control_code_page) << PAGE_SHIFT;
++}
++
++static struct resource xen_hypervisor_heap_res;
++
++int __init machine_kexec_setup_resources(struct resource *hypervisor,
++ struct resource *phys_cpus,
++ int nr_phys_cpus)
++{
++ xen_kexec_range_t range;
++ int k;
++
++ /* fill in xen_hypervisor_heap_res with hypervisor heap
++ * machine address range
++ */
++
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_XENHEAP;
++
++ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ return -1;
++
++ xen_hypervisor_heap_res.name = "Hypervisor heap";
++ xen_hypervisor_heap_res.start = range.start;
++ xen_hypervisor_heap_res.end = range.start + range.size - 1;
++ xen_hypervisor_heap_res.flags = IORESOURCE_BUSY | IORESOURCE_MEM;
++
++ /* The per-cpu crash note resources belong inside the
++ * hypervisor heap resource */
++ for (k = 0; k < nr_phys_cpus; k++)
++ request_resource(&xen_hypervisor_heap_res, phys_cpus + k);
++
++ /* fill in efi_memmap_res with EFI memmap machine address range */
++
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_EFI_MEMMAP;
++
++ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ return -1;
++
++ efi_memmap_res.start = range.start;
++ efi_memmap_res.end = range.start + range.size - 1;
++
++ /* fill in boot_param_res with boot parameter machine address range */
++
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_BOOT_PARAM;
++
++ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ return -1;
++
++ boot_param_res.start = range.start;
++ boot_param_res.end = range.start + range.size - 1;
++
++ return 0;
++}
++
++void machine_kexec_register_resources(struct resource *res)
++{
++ request_resource(res, &xen_hypervisor_heap_res);
++}
++#endif /* CONFIG_XEN */
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/mca.c
+--- a/arch/ia64/kernel/mca.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/mca.c Wed Sep 10 10:54:08 2008 +0100
+@@ -79,6 +79,7 @@
+ #include <asm/system.h>
+ #include <asm/sal.h>
+ #include <asm/mca.h>
++#include <asm/kexec.h>
+
+ #include <asm/irq.h>
+ #include <asm/hw_irq.h>
+@@ -160,11 +161,33 @@
+
+ static ia64_state_log_t ia64_state_log[IA64_MAX_LOG_TYPES];
+
++#ifdef CONFIG_XEN
++DEFINE_SPINLOCK(ia64_mca_xencomm_lock);
++LIST_HEAD(ia64_mca_xencomm_list);
++
++#define IA64_MCA_XENCOMM_ALLOCATE(rec, desc) \
++ if (is_running_on_xen()) { \
++ ia64_mca_xencomm_t *entry; \
++ entry = alloc_bootmem(sizeof(ia64_mca_xencomm_t)); \
++ entry->record = rec; \
++ entry->handle = desc; \
++ list_add(&entry->list, &ia64_mca_xencomm_list); \
++ }
++#define IA64_LOG_ALLOCATE(it, size) \
++ {ia64_err_rec_t *rec; \
++ ia64_state_log[it].isl_log[IA64_LOG_CURR_INDEX(it)] = rec = \
++ (ia64_err_rec_t *)alloc_bootmem(size); \
++ IA64_MCA_XENCOMM_ALLOCATE(rec, xencomm_map(rec, size)); \
++ ia64_state_log[it].isl_log[IA64_LOG_NEXT_INDEX(it)] = rec = \
++ (ia64_err_rec_t *)alloc_bootmem(size); \
++ IA64_MCA_XENCOMM_ALLOCATE(rec, xencomm_map(rec, size));}
++#else
+ #define IA64_LOG_ALLOCATE(it, size) \
+ {ia64_state_log[it].isl_log[IA64_LOG_CURR_INDEX(it)] = \
+ (ia64_err_rec_t *)alloc_bootmem(size); \
+ ia64_state_log[it].isl_log[IA64_LOG_NEXT_INDEX(it)] = \
+ (ia64_err_rec_t *)alloc_bootmem(size);}
++#endif
+ #define IA64_LOG_LOCK_INIT(it) spin_lock_init(&ia64_state_log[it].isl_lock)
+ #define IA64_LOG_LOCK(it) spin_lock_irqsave(&ia64_state_log[it].isl_lock, s)
+ #define IA64_LOG_UNLOCK(it) spin_unlock_irqrestore(&ia64_state_log[it].isl_lock,s)
+@@ -1066,7 +1089,12 @@
+ rh->severity = sal_log_severity_corrected;
+ ia64_sal_clear_state_info(SAL_INFO_TYPE_MCA);
+ sos->os_status = IA64_MCA_CORRECTED;
+- }
++ } else {
++#ifdef CONFIG_KEXEC
++ atomic_set(&kdump_in_progress, 1);
++ monarch_cpu = -1;
++#endif
++ }
+ if (notify_die(DIE_MCA_MONARCH_LEAVE, "MCA", regs, (long)&nd, 0, recover)
+ == NOTIFY_STOP)
+ ia64_mca_spin(__FUNCTION__);
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/pal.S
+--- a/arch/ia64/kernel/pal.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/pal.S Wed Sep 10 10:54:08 2008 +0100
@@ -16,6 +16,7 @@
#include <asm/processor.h>
@@ -23203,7 +25075,7 @@
pal_entry_point:
data8 ia64_pal_default_handler
.text
-@@ -53,7 +54,7 @@ END(ia64_pal_default_handler)
+@@ -53,7 +54,7 @@
* in4 1 ==> clear psr.ic, 0 ==> don't clear psr.ic
*
*/
@@ -23212,7 +25084,7 @@
.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(5)
alloc loc1 = ar.pfs,5,5,0,0
movl loc2 = pal_entry_point
-@@ -90,7 +91,7 @@ GLOBAL_ENTRY(ia64_pal_call_static)
+@@ -90,7 +91,7 @@
;;
srlz.d // seralize restoration of psr.l
br.ret.sptk.many b0
@@ -23221,97 +25093,55 @@
/*
* Make a PAL call using the stacked registers calling convention.
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/patch.c tmp-linux-2.6-xen.patch/arch/ia64/kernel/patch.c
---- pristine-linux-2.6.18.2/arch/ia64/kernel/patch.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/patch.c 2007-07-30 16:35:11.000000000 +0200
-@@ -184,6 +184,69 @@ patch_brl_fsys_bubble_down (unsigned lon
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/patch.c
+--- a/arch/ia64/kernel/patch.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/patch.c Wed Sep 10 10:54:08 2008 +0100
+@@ -184,9 +184,37 @@
ia64_srlz_i();
}
-+#ifdef CONFIG_XEN_IA64_VDSO_PARAVIRT
-+extern char __start_gate_running_on_xen_patchlist[];
-+extern char __end_gate_running_on_xen_patchlist[];
-+
-+void
-+patch_running_on_xen(unsigned long start, unsigned long end)
-+{
-+ extern int running_on_xen;
-+ s32 *offp = (s32 *)start;
-+ u64 ip;
-+
-+ while (offp < (s32 *)end) {
-+ ip = (u64)ia64_imva((char *)offp + *offp);
-+ ia64_patch_imm64(ip, (u64)&running_on_xen);
-+ ia64_fc((void *)ip);
-+ ++offp;
-+ }
-+ ia64_sync_i();
-+ ia64_srlz_i();
-+}
-+
-+static void
-+patch_brl_symaddr(unsigned long start, unsigned long end,
-+ unsigned long symaddr)
++#ifdef CONFIG_XEN
++void __init
++ia64_patch_gate_xen (void)
+{
-+ s32 *offp = (s32 *)start;
-+ u64 ip;
-+
-+ while (offp < (s32 *)end) {
-+ ip = (u64)offp + *offp;
-+ ia64_patch_imm60((u64)ia64_imva((void *)ip),
-+ (u64)(symaddr - (ip & -16)) / 16);
-+ ia64_fc((void *)ip);
-+ ++offp;
-+ }
-+ ia64_sync_i();
-+ ia64_srlz_i();
-+}
-+
-+#define EXTERN_PATCHLIST(name) \
-+ extern char __start_gate_brl_##name##_patchlist[]; \
-+ extern char __end_gate_brl_##name##_patchlist[]; \
-+ extern char name[]
-+
-+#define PATCH_BRL_SYMADDR(name) \
-+ patch_brl_symaddr((unsigned long)__start_gate_brl_##name##_patchlist, \
-+ (unsigned long)__end_gate_brl_##name##_patchlist, \
-+ (unsigned long)name)
++ extern char __start_gate_mckinley_e9_patchlist_xen[], __end_gate_mckinley_e9_patchlist_xen[];
++ extern char __start_gate_vtop_patchlist_xen[], __end_gate_vtop_patchlist_xen[];
++ extern char __start_gate_fsyscall_patchlist_xen[], __end_gate_fsyscall_patchlist_xen[];
++ extern char __start_gate_brl_fsys_bubble_down_patchlist_xen[], __end_gate_brl_fsys_bubble_down_patchlist_xen[];
++# define START(name) ((unsigned long) __start_gate_##name##_patchlist_xen)
++# define END(name) ((unsigned long)__end_gate_##name##_patchlist_xen)
+
-+static void
-+patch_brl_in_vdso(void)
-+{
-+ EXTERN_PATCHLIST(xen_ssm_i_0);
-+ EXTERN_PATCHLIST(xen_ssm_i_1);
++ patch_fsyscall_table(START(fsyscall), END(fsyscall));
++ patch_brl_fsys_bubble_down(START(brl_fsys_bubble_down), END(brl_fsys_bubble_down));
++ ia64_patch_vtop(START(vtop), END(vtop));
++ ia64_patch_mckinley_e9(START(mckinley_e9), END(mckinley_e9));
+
-+ PATCH_BRL_SYMADDR(xen_ssm_i_0);
-+ PATCH_BRL_SYMADDR(xen_ssm_i_1);
++# undef START
++# undef END
+}
+#else
-+#define patch_running_on_xen(start, end) do { } while (0)
-+#define patch_brl_in_vdso() do { } while (0)
++#define ia64_patch_gate_xen() do { } while (0)
+#endif
+
void __init
ia64_patch_gate (void)
{
-@@ -192,6 +255,10 @@ ia64_patch_gate (void)
++ if (is_running_on_xen()) {
++ ia64_patch_gate_xen();
++ return;
++ }
++
+ # define START(name) ((unsigned long) __start_gate_##name##_patchlist)
+ # define END(name) ((unsigned long)__end_gate_##name##_patchlist)
- patch_fsyscall_table(START(fsyscall), END(fsyscall));
- patch_brl_fsys_bubble_down(START(brl_fsys_bubble_down), END(brl_fsys_bubble_down));
-+#ifdef CONFIG_XEN
-+ patch_running_on_xen(START(running_on_xen), END(running_on_xen));
-+ patch_brl_in_vdso();
-+#endif
- ia64_patch_vtop(START(vtop), END(vtop));
- ia64_patch_mckinley_e9(START(mckinley_e9), END(mckinley_e9));
- }
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/perfmon.c tmp-linux-2.6-xen.patch/arch/ia64/kernel/perfmon.c
---- pristine-linux-2.6.18.2/arch/ia64/kernel/perfmon.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/perfmon.c 2007-07-30 16:35:11.000000000 +0200
-@@ -52,6 +52,28 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/perfmon.c
+--- a/arch/ia64/kernel/perfmon.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/perfmon.c Wed Sep 10 10:54:08 2008 +0100
+@@ -52,6 +52,31 @@
#include <asm/delay.h>
#ifdef CONFIG_PERFMON
++#include <asm/hypervisor.h>
+#ifdef CONFIG_XEN
+//#include <xen/xenoprof.h>
+#include <xen/interface/xenoprof.h>
@@ -23331,13 +25161,15 @@
+#define init_xenoprof_primary(is_primary) do { } while (0)
+#define is_xenoprof_primary() (0)
+#define XEN_NOT_SUPPORTED_YET do { } while (0)
-+#define HYPERVISOR_perfmon_op(cmd, arg, count) do { } while (0)
++#define HYPERVISOR_perfmon_op(cmd, arg, count) (0)
++#define HYPERVISOR_xenoprof_op(op, arg) ({(void)arg;0;})
++struct xenoprof_init { /* dummy */ };
+#endif
+
/*
* perfmon context state
*/
-@@ -1514,6 +1536,7 @@ pfm_read(struct file *filp, char __user
+@@ -1514,6 +1539,7 @@
ssize_t ret;
unsigned long flags;
DECLARE_WAITQUEUE(wait, current);
@@ -23345,7 +25177,7 @@
if (PFM_IS_FILE(filp) == 0) {
printk(KERN_ERR "perfmon: pfm_poll: bad magic [%d]\n", current->pid);
return -EINVAL;
-@@ -2112,6 +2135,15 @@ doit:
+@@ -2112,6 +2138,15 @@
*/
if (free_possible) pfm_context_free(ctx);
@@ -23361,7 +25193,7 @@
return 0;
}
-@@ -2735,6 +2767,23 @@ pfm_context_create(pfm_context_t *ctx, v
+@@ -2735,6 +2770,23 @@
*/
pfm_reset_pmu_state(ctx);
@@ -23385,7 +25217,7 @@
return 0;
buffer_error:
-@@ -2871,6 +2920,12 @@ pfm_write_pmcs(pfm_context_t *ctx, void
+@@ -2871,6 +2923,12 @@
pfm_reg_check_t wr_func;
#define PFM_CHECK_PMC_PM(x, y, z) ((x)->ctx_fl_system ^ PMC_PM(y, z))
@@ -23398,7 +25230,7 @@
state = ctx->ctx_state;
is_loaded = state == PFM_CTX_LOADED ? 1 : 0;
is_system = ctx->ctx_fl_system;
-@@ -3111,6 +3166,12 @@ pfm_write_pmds(pfm_context_t *ctx, void
+@@ -3111,6 +3169,12 @@
int ret = -EINVAL;
pfm_reg_check_t wr_func;
@@ -23411,7 +25243,7 @@
state = ctx->ctx_state;
is_loaded = state == PFM_CTX_LOADED ? 1 : 0;
-@@ -3308,6 +3369,7 @@ pfm_read_pmds(pfm_context_t *ctx, void *
+@@ -3308,6 +3372,7 @@
int is_loaded, is_system, is_counting, expert_mode;
int ret = -EINVAL;
pfm_reg_check_t rd_func;
@@ -23419,7 +25251,7 @@
/*
* access is possible when loaded only for
-@@ -3559,6 +3621,7 @@ pfm_restart(pfm_context_t *ctx, void *ar
+@@ -3559,6 +3624,7 @@
pfm_ovfl_ctrl_t rst_ctrl;
int state, is_system;
int ret = 0;
@@ -23427,7 +25259,7 @@
state = ctx->ctx_state;
fmt = ctx->ctx_buf_fmt;
-@@ -3708,6 +3771,7 @@ static int
+@@ -3708,6 +3774,7 @@
pfm_debug(pfm_context_t *ctx, void *arg, int count, struct pt_regs *regs)
{
unsigned int m = *(unsigned int *)arg;
@@ -23435,7 +25267,7 @@
pfm_sysctl.debug = m == 0 ? 0 : 1;
-@@ -3978,6 +4042,8 @@ pfm_get_features(pfm_context_t *ctx, voi
+@@ -3978,6 +4045,8 @@
{
pfarg_features_t *req = (pfarg_features_t *)arg;
@@ -23444,20 +25276,20 @@
req->ft_version = PFM_VERSION;
return 0;
}
-@@ -3989,6 +4055,12 @@ pfm_stop(pfm_context_t *ctx, void *arg,
+@@ -3988,6 +4057,12 @@
+ struct pt_regs *tregs;
struct task_struct *task = PFM_CTX_TASK(ctx);
int state, is_system;
-
++
+ if (is_running_on_xen()) {
+ if (is_xenoprof_primary())
+ return HYPERVISOR_perfmon_op(PFM_STOP, NULL, 0);
+ return 0;
+ }
-+
+
state = ctx->ctx_state;
is_system = ctx->ctx_fl_system;
-
-@@ -4077,6 +4149,11 @@ pfm_start(pfm_context_t *ctx, void *arg,
+@@ -4077,6 +4152,11 @@
struct pt_regs *tregs;
int state, is_system;
@@ -23469,7 +25301,7 @@
state = ctx->ctx_state;
is_system = ctx->ctx_fl_system;
-@@ -4159,6 +4236,7 @@ pfm_get_pmc_reset(pfm_context_t *ctx, vo
+@@ -4159,6 +4239,7 @@
unsigned int cnum;
int i;
int ret = -EINVAL;
@@ -23477,7 +25309,7 @@
for (i = 0; i < count; i++, req++) {
-@@ -4217,6 +4295,11 @@ pfm_context_load(pfm_context_t *ctx, voi
+@@ -4217,6 +4298,11 @@
int ret = 0;
int state, is_system, set_dbregs = 0;
@@ -23489,7 +25321,7 @@
state = ctx->ctx_state;
is_system = ctx->ctx_fl_system;
/*
-@@ -4465,6 +4548,12 @@ pfm_context_unload(pfm_context_t *ctx, v
+@@ -4465,6 +4551,12 @@
int prev_state, is_system;
int ret;
@@ -23502,10 +25334,421 @@
DPRINT(("ctx_state=%d task [%d]\n", ctx->ctx_state, task ? task->pid : -1));
prev_state = ctx->ctx_state;
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/setup.c tmp-linux-2.6-xen.patch/arch/ia64/kernel/setup.c
---- pristine-linux-2.6.18.2/arch/ia64/kernel/setup.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/setup.c 2007-07-30 16:35:11.000000000 +0200
-@@ -60,6 +60,12 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/relocate_kernel.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/kernel/relocate_kernel.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,338 @@
++/*
++ * arch/ia64/kernel/relocate_kernel.S
++ *
++ * Relocate kexec'able kernel and start it
++ *
++ * Copyright (C) 2005 Hewlett-Packard Development Company, L.P.
++ * Copyright (C) 2005 Khalid Aziz <khalid.aziz at hp.com>
++ * Copyright (C) 2005 Intel Corp, Zou Nan hai <nanhai.zou at intel.com>
++ *
++ * This source code is licensed under the GNU General Public License,
++ * Version 2. See the file COPYING for more details.
++ */
++#include <asm/asmmacro.h>
++#include <asm/kregs.h>
++#include <asm/page.h>
++#include <asm/pgtable.h>
++#include <asm/mca_asm.h>
++
++ /* Must be relocatable PIC code callable as a C function
++ */
++GLOBAL_ENTRY(relocate_new_kernel)
++ .prologue
++ alloc r31=ar.pfs,4,0,0,0
++ .body
++.reloc_entry:
++{
++ rsm psr.i| psr.ic
++ mov r2=ip
++}
++ ;;
++{
++ flushrs // must be first insn in group
++ srlz.i
++}
++ ;;
++ dep r2=0,r2,61,3 //to physical address
++ ;;
++ //first switch to physical mode
++ add r3=1f-.reloc_entry, r2
++ movl r16 = IA64_PSR_AC|IA64_PSR_BN|IA64_PSR_IC
++ mov ar.rsc=0 // put RSE in enforced lazy mode
++ ;;
++ add sp=(memory_stack_end - 16 - .reloc_entry),r2
++ add r8=(register_stack - .reloc_entry),r2
++ ;;
++ mov r18=ar.rnat
++ mov ar.bspstore=r8
++ ;;
++ mov cr.ipsr=r16
++ mov cr.iip=r3
++ mov cr.ifs=r0
++ srlz.i
++ ;;
++ mov ar.rnat=r18
++ rfi
++ ;;
++1:
++ //physical mode code begin
++ mov b6=in1
++#ifdef CONFIG_XEN
++ mov r28=in2 //already a physical address
++#else
++ dep r28=0,in2,61,3 //to physical address
++
++ // purge all TC entries
++#define O(member) IA64_CPUINFO_##member##_OFFSET
++ GET_THIS_PADDR(r2, cpu_info) // load phys addr of cpu_info into r2
++ ;;
++ addl r17=O(PTCE_STRIDE),r2
++ addl r2=O(PTCE_BASE),r2
++ ;;
++ ld8 r18=[r2],(O(PTCE_COUNT)-O(PTCE_BASE));; // r18=ptce_base
++ ld4 r19=[r2],4 // r19=ptce_count[0]
++ ld4 r21=[r17],4 // r21=ptce_stride[0]
++ ;;
++ ld4 r20=[r2] // r20=ptce_count[1]
++ ld4 r22=[r17] // r22=ptce_stride[1]
++ mov r24=r0
++ ;;
++ adds r20=-1,r20
++ ;;
++#undef O
++2:
++ cmp.ltu p6,p7=r24,r19
++(p7) br.cond.dpnt.few 4f
++ mov ar.lc=r20
++3:
++ ptc.e r18
++ ;;
++ add r18=r22,r18
++ br.cloop.sptk.few 3b
++ ;;
++ add r18=r21,r18
++ add r24=1,r24
++ ;;
++ br.sptk.few 2b
++4:
++ srlz.i
++ ;;
++ //purge TR entry for kernel text and data
++ movl r16=KERNEL_START
++ mov r18=KERNEL_TR_PAGE_SHIFT<<2
++ ;;
++ ptr.i r16, r18
++ ptr.d r16, r18
++ ;;
++ srlz.i
++ ;;
++
++ // purge TR entry for percpu data
++ movl r16=PERCPU_ADDR
++ mov r18=PERCPU_PAGE_SHIFT<<2
++ ;;
++ ptr.d r16,r18
++ ;;
++ srlz.d
++ ;;
++
++ // purge TR entry for pal code
++ mov r16=in3
++ mov r18=IA64_GRANULE_SHIFT<<2
++ ;;
++ ptr.i r16,r18
++ ;;
++ srlz.i
++ ;;
++
++ // purge TR entry for stack
++ mov r16=IA64_KR(CURRENT_STACK)
++ ;;
++ shl r16=r16,IA64_GRANULE_SHIFT
++ movl r19=PAGE_OFFSET
++ ;;
++ add r16=r19,r16
++ mov r18=IA64_GRANULE_SHIFT<<2
++ ;;
++ ptr.d r16,r18
++ ;;
++ srlz.i
++ ;;
++#endif /* ! CONFIG_XEN */
++
++ //copy segments
++ movl r16=PAGE_MASK
++ mov r30=in0 // in0 is page_list
++ br.sptk.few .dest_page
++ ;;
++.loop:
++ ld8 r30=[in0], 8;;
++.dest_page:
++ tbit.z p0, p6=r30, 0;; // 0x1 dest page
++(p6) and r17=r30, r16
++(p6) br.cond.sptk.few .loop;;
++
++ tbit.z p0, p6=r30, 1;; // 0x2 indirect page
++(p6) and in0=r30, r16
++(p6) br.cond.sptk.few .loop;;
++
++ tbit.z p0, p6=r30, 2;; // 0x4 end flag
++(p6) br.cond.sptk.few .end_loop;;
++
++ tbit.z p6, p0=r30, 3;; // 0x8 source page
++(p6) br.cond.sptk.few .loop
++
++ and r18=r30, r16
++
++ // simple copy page, may optimize later
++ movl r14=PAGE_SIZE/8 - 1;;
++ mov ar.lc=r14;;
++1:
++ ld8 r14=[r18], 8;;
++ st8 [r17]=r14;;
++ fc.i r17
++ add r17=8, r17
++ br.ctop.sptk.few 1b
++ br.sptk.few .loop
++ ;;
++
++.end_loop:
++ sync.i // for fc.i
++ ;;
++ srlz.i
++ ;;
++ srlz.d
++ ;;
++ br.call.sptk.many b0=b6;;
++
++.align 32
++memory_stack:
++ .fill 8192, 1, 0
++memory_stack_end:
++register_stack:
++ .fill 8192, 1, 0
++register_stack_end:
++relocate_new_kernel_end:
++END(relocate_new_kernel)
++
++.global relocate_new_kernel_size
++relocate_new_kernel_size:
++ data8 relocate_new_kernel_end - relocate_new_kernel
++
++GLOBAL_ENTRY(ia64_dump_cpu_regs)
++ .prologue
++ alloc loc0=ar.pfs,1,2,0,0
++ .body
++ mov ar.rsc=0 // put RSE in enforced lazy mode
++ add loc1=4*8, in0 // save r4 and r5 first
++ ;;
++{
++ flushrs // flush dirty regs to backing store
++ srlz.i
++}
++ st8 [loc1]=r4, 8
++ ;;
++ st8 [loc1]=r5, 8
++ ;;
++ add loc1=32*8, in0
++ mov r4=ar.rnat
++ ;;
++ st8 [in0]=r0, 8 // r0
++ st8 [loc1]=r4, 8 // rnat
++ mov r5=pr
++ ;;
++ st8 [in0]=r1, 8 // r1
++ st8 [loc1]=r5, 8 // pr
++ mov r4=b0
++ ;;
++ st8 [in0]=r2, 8 // r2
++ st8 [loc1]=r4, 8 // b0
++ mov r5=b1;
++ ;;
++ st8 [in0]=r3, 24 // r3
++ st8 [loc1]=r5, 8 // b1
++ mov r4=b2
++ ;;
++ st8 [in0]=r6, 8 // r6
++ st8 [loc1]=r4, 8 // b2
++ mov r5=b3
++ ;;
++ st8 [in0]=r7, 8 // r7
++ st8 [loc1]=r5, 8 // b3
++ mov r4=b4
++ ;;
++ st8 [in0]=r8, 8 // r8
++ st8 [loc1]=r4, 8 // b4
++ mov r5=b5
++ ;;
++ st8 [in0]=r9, 8 // r9
++ st8 [loc1]=r5, 8 // b5
++ mov r4=b6
++ ;;
++ st8 [in0]=r10, 8 // r10
++ st8 [loc1]=r5, 8 // b6
++ mov r5=b7
++ ;;
++ st8 [in0]=r11, 8 // r11
++ st8 [loc1]=r5, 8 // b7
++ mov r4=b0
++ ;;
++ st8 [in0]=r12, 8 // r12
++ st8 [loc1]=r4, 8 // ip
++ mov r5=loc0
++ ;;
++ st8 [in0]=r13, 8 // r13
++ extr.u r5=r5, 0, 38 // ar.pfs.pfm
++ mov r4=r0 // user mask
++ ;;
++ st8 [in0]=r14, 8 // r14
++ st8 [loc1]=r5, 8 // cfm
++ ;;
++ st8 [in0]=r15, 8 // r15
++ st8 [loc1]=r4, 8 // user mask
++ mov r5=ar.rsc
++ ;;
++ st8 [in0]=r16, 8 // r16
++ st8 [loc1]=r5, 8 // ar.rsc
++ mov r4=ar.bsp
++ ;;
++ st8 [in0]=r17, 8 // r17
++ st8 [loc1]=r4, 8 // ar.bsp
++ mov r5=ar.bspstore
++ ;;
++ st8 [in0]=r18, 8 // r18
++ st8 [loc1]=r5, 8 // ar.bspstore
++ mov r4=ar.rnat
++ ;;
++ st8 [in0]=r19, 8 // r19
++ st8 [loc1]=r4, 8 // ar.rnat
++ mov r5=ar.ccv
++ ;;
++ st8 [in0]=r20, 8 // r20
++ st8 [loc1]=r5, 8 // ar.ccv
++ mov r4=ar.unat
++ ;;
++ st8 [in0]=r21, 8 // r21
++ st8 [loc1]=r4, 8 // ar.unat
++ mov r5 = ar.fpsr
++ ;;
++ st8 [in0]=r22, 8 // r22
++ st8 [loc1]=r5, 8 // ar.fpsr
++ mov r4 = ar.unat
++ ;;
++ st8 [in0]=r23, 8 // r23
++ st8 [loc1]=r4, 8 // unat
++ mov r5 = ar.fpsr
++ ;;
++ st8 [in0]=r24, 8 // r24
++ st8 [loc1]=r5, 8 // fpsr
++ mov r4 = ar.pfs
++ ;;
++ st8 [in0]=r25, 8 // r25
++ st8 [loc1]=r4, 8 // ar.pfs
++ mov r5 = ar.lc
++ ;;
++ st8 [in0]=r26, 8 // r26
++ st8 [loc1]=r5, 8 // ar.lc
++ mov r4 = ar.ec
++ ;;
++ st8 [in0]=r27, 8 // r27
++ st8 [loc1]=r4, 8 // ar.ec
++ mov r5 = ar.csd
++ ;;
++ st8 [in0]=r28, 8 // r28
++ st8 [loc1]=r5, 8 // ar.csd
++ mov r4 = ar.ssd
++ ;;
++ st8 [in0]=r29, 8 // r29
++ st8 [loc1]=r4, 8 // ar.ssd
++ ;;
++ st8 [in0]=r30, 8 // r30
++ ;;
++ st8 [in0]=r31, 8 // r31
++ mov ar.pfs=loc0
++ ;;
++ br.ret.sptk.many rp
++END(ia64_dump_cpu_regs)
++
++
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/salinfo.c
+--- a/arch/ia64/kernel/salinfo.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/salinfo.c Wed Sep 10 10:54:08 2008 +0100
+@@ -375,6 +375,25 @@
+ data->open = 0;
+ return -ENOMEM;
+ }
++#ifdef CONFIG_XEN
++ if (is_running_on_xen()) {
++ ia64_mca_xencomm_t *entry;
++ unsigned long flags;
++
++ entry = vmalloc(sizeof(ia64_mca_xencomm_t));
++ if (!entry) {
++ data->open = 0;
++ vfree(data->log_buffer);
++ return -ENOMEM;
++ }
++ entry->record = data->log_buffer;
++ entry->handle = xencomm_map(data->log_buffer,
++ ia64_sal_get_state_info_size(data->type));
++ spin_lock_irqsave(&ia64_mca_xencomm_lock, flags);
++ list_add(&entry->list, &ia64_mca_xencomm_list);
++ spin_unlock_irqrestore(&ia64_mca_xencomm_lock, flags);
++ }
++#endif
+
+ return 0;
+ }
+@@ -386,6 +405,30 @@
+ struct salinfo_data *data = entry->data;
+
+ if (data->state == STATE_NO_DATA) {
++#ifdef CONFIG_XEN
++ if (is_running_on_xen()) {
++ struct list_head *pos, *n;
++ ia64_mca_xencomm_t *found_entry = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&ia64_mca_xencomm_lock, flags);
++ list_for_each_safe(pos, n, &ia64_mca_xencomm_list) {
++ ia64_mca_xencomm_t *entry;
++
++ entry = list_entry(pos, ia64_mca_xencomm_t, list);
++ if (entry->record == data->log_buffer) {
++ list_del(&entry->list);
++ found_entry = entry;
++ break;
++ }
++ }
++ spin_unlock_irqrestore(&ia64_mca_xencomm_lock, flags);
++ if (found_entry) {
++ xencomm_free(found_entry->handle);
++ vfree(found_entry);
++ }
++ }
++#endif
+ vfree(data->log_buffer);
+ vfree(data->oemdata);
+ data->log_buffer = NULL;
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/setup.c
+--- a/arch/ia64/kernel/setup.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/setup.c Wed Sep 10 10:54:08 2008 +0100
+@@ -43,6 +43,8 @@
+ #include <linux/initrd.h>
+ #include <linux/pm.h>
+ #include <linux/cpufreq.h>
++#include <linux/kexec.h>
++#include <linux/crash_dump.h>
+
+ #include <asm/ia32.h>
+ #include <asm/machvec.h>
+@@ -60,6 +62,12 @@
#include <asm/system.h>
#include <asm/unistd.h>
#include <asm/system.h>
@@ -23518,10 +25761,12 @@
#if defined(CONFIG_SMP) && (IA64_CPU_SIZE > PAGE_SIZE)
# error "struct cpuinfo_ia64 too big!"
-@@ -70,6 +76,34 @@ unsigned long __per_cpu_offset[NR_CPUS];
+@@ -68,6 +76,34 @@
+ #ifdef CONFIG_SMP
+ unsigned long __per_cpu_offset[NR_CPUS];
EXPORT_SYMBOL(__per_cpu_offset);
- #endif
-
++#endif
++
+#ifdef CONFIG_XEN
+static void
+xen_panic_hypercall(struct unw_frame_info *info, void *arg)
@@ -23548,12 +25793,10 @@
+ local_irq_disable();
+ HYPERVISOR_shutdown(SHUTDOWN_poweroff);
+}
-+#endif
-+
- extern void ia64_setup_printk_clock(void);
+ #endif
- DEFINE_PER_CPU(struct cpuinfo_ia64, cpu_info);
-@@ -242,6 +276,14 @@ reserve_memory (void)
+ extern void ia64_setup_printk_clock(void);
+@@ -242,6 +278,14 @@
rsvd_region[n].end = (unsigned long) ia64_imva(_end);
n++;
@@ -23568,59 +25811,147 @@
#ifdef CONFIG_BLK_DEV_INITRD
if (ia64_boot_param->initrd_start) {
rsvd_region[n].start = (unsigned long)__va(ia64_boot_param->initrd_start);
-@@ -402,6 +444,19 @@ setup_arch (char **cmdline_p)
+@@ -253,6 +297,56 @@
+ efi_memmap_init(&rsvd_region[n].start, &rsvd_region[n].end);
+ n++;
+
++#ifdef CONFIG_KEXEC
++ /* crashkernel=size at offset specifies the size to reserve for a crash
++ * kernel. If offset is 0, then it is determined automatically.
++ * By reserving this memory we guarantee that linux never set's it
++ * up as a DMA target.Useful for holding code to do something
++ * appropriate after a kernel panic.
++ */
++ {
++ char *from = strstr(saved_command_line, "crashkernel=");
++ unsigned long base, size;
++#ifdef CONFIG_XEN
++ if (is_initial_xendomain() && from)
++ printk("Ignoring crashkernel command line, "
++ "parameter will be supplied by xen\n");
++ else {
++#endif
++ if (from) {
++ size = memparse(from + 12, &from);
++ if (*from == '@')
++ base = memparse(from+1, &from);
++ else
++ base = 0;
++ if (size) {
++ if (!base) {
++ sort_regions(rsvd_region, n);
++ base = kdump_find_rsvd_region(size,
++ rsvd_region, n);
++ }
++ if (base != ~0UL) {
++ rsvd_region[n].start =
++ (unsigned long)__va(base);
++ rsvd_region[n].end =
++ (unsigned long)__va(base + size);
++ n++;
++ crashk_res.start = base;
++ crashk_res.end = base + size - 1;
++ }
++ }
++ }
++ efi_memmap_res.start = ia64_boot_param->efi_memmap;
++ efi_memmap_res.end = efi_memmap_res.start +
++ ia64_boot_param->efi_memmap_size;
++ boot_param_res.start = kexec_virt_to_phys(ia64_boot_param);
++ boot_param_res.end = boot_param_res.start +
++ sizeof(*ia64_boot_param);
++#ifdef CONFIG_XEN
++ }
++#endif
++ }
++#endif
+ /* end of memory marker */
+ rsvd_region[n].start = ~0UL;
+ rsvd_region[n].end = ~0UL;
+@@ -263,6 +357,7 @@
+
+ sort_regions(rsvd_region, num_rsvd_regions);
+ }
++
+
+ /**
+ * find_initrd - get initrd parameters from the boot parameter structure
+@@ -397,10 +492,48 @@
+ }
+ early_param("nomca", setup_nomca);
+
++#ifdef CONFIG_PROC_VMCORE
++/* elfcorehdr= specifies the location of elf core header
++ * stored by the crashed kernel.
++ */
++static int __init parse_elfcorehdr(char *arg)
++{
++ if (!arg)
++ return -EINVAL;
++
++ elfcorehdr_addr = memparse(arg, &arg);
++ return 0;
++}
++early_param("elfcorehdr", parse_elfcorehdr);
++#endif /* CONFIG_PROC_VMCORE */
++
+ void __init
+ setup_arch (char **cmdline_p)
{
++#ifdef CONFIG_XEN
++ shared_info_t *s = NULL;
++ if (is_running_on_xen()) {
++ s = HYPERVISOR_shared_info;
++ xen_start_info = __va(s->arch.start_info_pfn << PAGE_SHIFT);
++ }
++#endif
++
unw_init();
-
++
+#ifdef CONFIG_XEN
+ if (is_running_on_xen()) {
+ /* Must be done before any hypercall. */
-+ xencomm_init();
++ xencomm_initialize();
+
+ setup_xen_features();
+ /* Register a call for panic conditions. */
+ atomic_notifier_chain_register(&panic_notifier_list,
+ &xen_panic_block);
+ pm_power_off = xen_pm_power_off;
++
++ xen_ia64_enable_opt_feature();
+ }
+#endif
-+
+
ia64_patch_vtop((u64) __start___vtop_patchlist, (u64) __end___vtop_patchlist);
- *cmdline_p = __va(ia64_boot_param->command_line);
-@@ -478,14 +533,79 @@ setup_arch (char **cmdline_p)
- conswitchp = &vga_con;
- # endif
- }
+@@ -464,6 +597,57 @@
+ acpi_boot_init();
+ #endif
+
+#ifdef CONFIG_XEN
+ if (is_running_on_xen()) {
-+ shared_info_t *s = HYPERVISOR_shared_info;
-+
-+ xen_start_info = __va(s->arch.start_info_pfn << PAGE_SHIFT);
-+
+ printk("Running on Xen! start_info_pfn=0x%lx nr_pages=%ld "
+ "flags=0x%x\n", s->arch.start_info_pfn,
+ xen_start_info->nr_pages, xen_start_info->flags);
+
-+ if (!is_initial_xendomain()) {
-+#if !defined(CONFIG_VT) || !defined(CONFIG_DUMMY_CONSOLE)
-+ conswitchp = NULL;
-+#endif
-+ }
-+
+ /*
+ * If a console= is NOT specified, we assume using the
-+ * xencons console is desired. By default, this is ttyS0
-+ * for dom0 and tty0 for domU.
++ * xencons console is desired. By default, this is xvc0
++ * for both dom0 and domU.
+ */
+ if (!strstr(*cmdline_p, "console=")) {
-+ char *p, *q, name[5];
++ char *p, *q, name[5] = "xvc";
+ int offset = 0;
+
-+ if (is_initial_xendomain())
-+ strncpy(name, "ttyS", 4);
-+ else
++#if defined(CONFIG_VGA_CONSOLE)
++ /*
++ * conswitchp might be set intelligently from the
++ * PCDP code. If set to VGA console, use it.
++ */
++ if (is_initial_xendomain() && conswitchp == &vga_con)
+ strncpy(name, "tty", 3);
++#endif
+
+ p = strstr(*cmdline_p, "xencons=");
+
@@ -23647,47 +25978,114 @@
+ add_preferred_console(name, offset, NULL);
+ }
+ }
-+ xencons_early_setup();
+#endif
++
+ #ifdef CONFIG_VT
+ if (!conswitchp) {
+ # if defined(CONFIG_DUMMY_CONSOLE)
+@@ -483,11 +667,28 @@
#endif
-+
/* enable IA-64 Machine Check Abort Handling unless disabled */
+#ifdef CONFIG_XEN
-+ if (is_running_on_xen() && !is_initial_xendomain())
++ if (is_running_on_xen() && !is_initial_xendomain()) {
+ nomca = 1;
++#if !defined(CONFIG_VT) || !defined(CONFIG_DUMMY_CONSOLE)
++ conswitchp = NULL;
++#endif
++ }
+#endif
if (!nomca)
ia64_mca_init();
platform_setup(cmdline_p);
++#ifdef CONFIG_XEN
++ if (is_running_on_xen() && !ia64_platform_is("xen")) {
++ extern ia64_mv_setup_t xen_setup;
++ xen_setup(cmdline_p);
++ }
++#endif
paging_init();
+#ifdef CONFIG_XEN
-+ contiguous_bitmap_init(max_pfn);
++ xen_contiguous_bitmap_init(max_pfn);
+#endif
}
/*
-@@ -870,6 +990,15 @@ cpu_init (void)
+@@ -872,6 +1073,13 @@
/* size of physical stacked register partition plus 8 bytes: */
__get_cpu_var(ia64_phys_stacked_size_p8) = num_phys_stacked*8 + 8;
platform_cpu_init();
-+
+#ifdef CONFIG_XEN
-+ /* Need to be moved into platform_cpu_init later */
-+ if (is_running_on_xen()) {
-+ extern void xen_smp_intr_init(void);
-+ xen_smp_intr_init();
++ if (is_running_on_xen() && !ia64_platform_is("xen")) {
++ extern ia64_mv_cpu_init_t xen_cpu_init;
++ xen_cpu_init();
+ }
+#endif
+
pm_idle = default_idle;
}
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/smp.c tmp-linux-2.6-xen.patch/arch/ia64/kernel/smp.c
---- pristine-linux-2.6.18.2/arch/ia64/kernel/smp.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/smp.c 2007-10-14 01:51:15.000000000 +0200
-@@ -328,10 +328,14 @@ int
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/smp.c
+--- a/arch/ia64/kernel/smp.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/smp.c Wed Sep 10 10:54:08 2008 +0100
+@@ -30,6 +30,7 @@
+ #include <linux/delay.h>
+ #include <linux/efi.h>
+ #include <linux/bitops.h>
++#include <linux/kexec.h>
+
+ #include <asm/atomic.h>
+ #include <asm/current.h>
+@@ -66,6 +67,7 @@
+
+ #define IPI_CALL_FUNC 0
+ #define IPI_CPU_STOP 1
++#define IPI_KDUMP_CPU_STOP 3
+
+ /* This needs to be cacheline aligned because it is written to by *other* CPUs. */
+ static DEFINE_PER_CPU(u64, ipi_operation) ____cacheline_aligned;
+@@ -155,7 +157,11 @@
+ case IPI_CPU_STOP:
+ stop_this_cpu();
+ break;
+-
++#ifdef CONFIG_KEXEC
++ case IPI_KDUMP_CPU_STOP:
++ unw_init_running(kdump_cpu_freeze, NULL);
++ break;
++#endif
+ default:
+ printk(KERN_CRIT "Unknown IPI on CPU %d: %lu\n", this_cpu, which);
+ break;
+@@ -213,6 +219,26 @@
+ send_IPI_single(smp_processor_id(), op);
+ }
+
++#ifdef CONFIG_KEXEC
++void
++kdump_smp_send_stop()
++{
++ send_IPI_allbutself(IPI_KDUMP_CPU_STOP);
++}
++
++void
++kdump_smp_send_init()
++{
++ unsigned int cpu, self_cpu;
++ self_cpu = smp_processor_id();
++ for_each_online_cpu(cpu) {
++ if (cpu != self_cpu) {
++ if(kdump_status[cpu] == 0)
++ platform_send_ipi(cpu, 0, IA64_IPI_DM_INIT, 0);
++ }
++ }
++}
++#endif
+ /*
+ * Called with preeemption disabled.
+ */
+@@ -328,10 +354,14 @@
smp_call_function (void (*func) (void *info), void *info, int nonatomic, int wait)
{
struct call_data_struct data;
@@ -23704,22 +26102,23 @@
/* Can deadlock when called with interrupts disabled */
WARN_ON(irqs_disabled());
-@@ -343,8 +347,6 @@ smp_call_function (void (*func) (void *i
+@@ -342,8 +372,6 @@
+ data.wait = wait;
if (wait)
atomic_set(&data.finished, 0);
-
-- spin_lock(&call_lock);
-
+- spin_lock(&call_lock);
+
call_data = &data;
mb(); /* ensure store to call_data precedes setting of IPI_CALL_FUNC */
- send_IPI_allbutself(IPI_CALL_FUNC);
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/kernel/time.c tmp-linux-2.6-xen.patch/arch/ia64/kernel/time.c
---- pristine-linux-2.6.18.2/arch/ia64/kernel/time.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/kernel/time.c 2007-09-30 18:06:18.000000000 +0200
-@@ -29,6 +29,13 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/time.c
+--- a/arch/ia64/kernel/time.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/time.c Wed Sep 10 10:54:08 2008 +0100
+@@ -29,6 +29,14 @@
#include <asm/sections.h>
#include <asm/system.h>
++#include <asm/hypervisor.h>
+#ifdef CONFIG_XEN
+#include <linux/kernel_stat.h>
+#include <linux/posix-timers.h>
@@ -23730,7 +26129,7 @@
extern unsigned long wall_jiffies;
volatile int time_keeper_id = 0; /* smp_processor_id() of time-keeper */
-@@ -40,16 +47,109 @@ EXPORT_SYMBOL(last_cli_ip);
+@@ -40,16 +48,109 @@
#endif
@@ -23840,30 +26239,30 @@
if (unlikely(cpu_is_offline(smp_processor_id()))) {
return IRQ_HANDLED;
-@@ -65,6 +165,13 @@ timer_interrupt (int irq, void *dev_id,
+@@ -64,6 +165,13 @@
+ ia64_get_itc(), new_itm);
profile_tick(CPU_PROFILING, regs);
-
++
+ if (is_running_on_xen()) {
+ delta_itm = consider_steal_time(new_itm, regs);
+ new_itm += delta_itm;
+ if (time_after(new_itm, ia64_get_itc()) && delta_itm)
+ goto skip_process_time_accounting;
+ }
-+
+
while (1) {
update_process_times(user_mode(regs));
-
-@@ -88,6 +195,8 @@ timer_interrupt (int irq, void *dev_id,
+@@ -87,6 +195,8 @@
+ if (time_after(new_itm, ia64_get_itc()))
break;
}
-
-+skip_process_time_accounting: /* XEN */
+
++skip_process_time_accounting: /* XEN */
+
do {
/*
- * If we're too close to the next clock tick for
-@@ -142,6 +251,85 @@ static int __init nojitter_setup(char *s
+@@ -142,6 +252,85 @@
__setup("nojitter", nojitter_setup);
@@ -23949,7 +26348,7 @@
void __devinit
ia64_init_itm (void)
-@@ -225,6 +413,12 @@ ia64_init_itm (void)
+@@ -225,6 +414,12 @@
register_time_interpolator(&itc_interpolator);
}
@@ -23962,88 +26361,169 @@
/* Setup the CPU local timer tick */
ia64_cpu_local_tick();
}
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/Makefile tmp-linux-2.6-xen.patch/arch/ia64/Makefile
---- pristine-linux-2.6.18.2/arch/ia64/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -45,6 +45,12 @@ ifeq ($(call cc-version),0304)
- endif
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/vmlinux.lds.S
+--- a/arch/ia64/kernel/vmlinux.lds.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/kernel/vmlinux.lds.S Wed Sep 10 10:54:08 2008 +0100
+@@ -183,6 +183,12 @@
+ __start_gate_section = .;
+ *(.data.gate)
+ __stop_gate_section = .;
++#if defined(CONFIG_XEN)
++ . = ALIGN(PAGE_SIZE);
++ __start_xen_gate_section = .;
++ *(.data.gate.xen)
++ __stop_xen_gate_section = .;
++#endif
+ }
+ . = ALIGN(PAGE_SIZE); /* make sure the gate page doesn't expose kernel data */
- CFLAGS += $(cflags-y)
-+
-+cppflags-$(CONFIG_XEN) += \
-+ -D__XEN_INTERFACE_VERSION__=$(CONFIG_XEN_INTERFACE_VERSION)
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/kernel/xengate-data.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/kernel/xengate-data.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,3 @@
++ .section .data.gate.xen, "aw"
+
-+CPPFLAGS += $(cppflags-y)
++ .incbin "arch/ia64/kernel/xengate.so"
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/mm/contig.c
+--- a/arch/ia64/mm/contig.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/mm/contig.c Wed Sep 10 10:54:08 2008 +0100
+@@ -18,6 +18,9 @@
+ #include <linux/efi.h>
+ #include <linux/mm.h>
+ #include <linux/swap.h>
++#if defined(CONFIG_XEN) && defined(CONFIG_KEXEC)
++#include <linux/kexec.h>
++#endif
+
+ #include <asm/meminit.h>
+ #include <asm/pgalloc.h>
+@@ -172,8 +175,12 @@
+ /* Free all available memory, then mark bootmem-map as being in use. */
+ efi_memmap_walk(filter_rsvd_memory, free_bootmem);
+ reserve_bootmem(bootmap_start, bootmap_size);
++#if defined(CONFIG_XEN) && defined(CONFIG_KEXEC)
++ xen_machine_kexec_setup_resources();
++#endif
+
+ find_initrd();
+
- head-y := arch/ia64/kernel/head.o arch/ia64/kernel/init_task.o
+ }
- libs-y += arch/ia64/lib/
-@@ -55,9 +61,15 @@ core-$(CONFIG_IA64_GENERIC) += arch/ia6
- core-$(CONFIG_IA64_HP_ZX1) += arch/ia64/dig/
- core-$(CONFIG_IA64_HP_ZX1_SWIOTLB) += arch/ia64/dig/
- core-$(CONFIG_IA64_SGI_SN2) += arch/ia64/sn/
-+core-$(CONFIG_XEN) += arch/ia64/xen/
+ #ifdef CONFIG_SMP
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/mm/discontig.c
+--- a/arch/ia64/mm/discontig.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/mm/discontig.c Wed Sep 10 10:54:08 2008 +0100
+@@ -21,6 +21,9 @@
+ #include <linux/acpi.h>
+ #include <linux/efi.h>
+ #include <linux/nodemask.h>
++#if defined(CONFIG_XEN) && defined(CONFIG_KEXEC)
++#include <linux/kexec.h>
++#endif
+ #include <asm/pgalloc.h>
+ #include <asm/tlb.h>
+ #include <asm/meminit.h>
+@@ -502,6 +505,9 @@
+ reserve_pernode_space();
+ memory_less_nodes();
+ initialize_pernode_data();
++#if defined(CONFIG_XEN) && defined(CONFIG_KEXEC)
++ xen_machine_kexec_setup_resources();
++#endif
- drivers-$(CONFIG_PCI) += arch/ia64/pci/
-+ifneq ($(CONFIG_XEN),y)
- drivers-$(CONFIG_IA64_HP_SIM) += arch/ia64/hp/sim/
-+endif
-+ifneq ($(CONFIG_IA64_GENERIC),y)
-+drivers-$(CONFIG_XEN) += arch/ia64/hp/sim/
-+endif
- drivers-$(CONFIG_IA64_HP_ZX1) += arch/ia64/hp/common/ arch/ia64/hp/zx1/
- drivers-$(CONFIG_IA64_HP_ZX1_SWIOTLB) += arch/ia64/hp/common/ arch/ia64/hp/zx1/
- drivers-$(CONFIG_IA64_GENERIC) += arch/ia64/hp/common/ arch/ia64/hp/zx1/ arch/ia64/hp/sim/ arch/ia64/sn/
-@@ -87,8 +99,8 @@ CLEAN_FILES += vmlinux.gz bootloader
- boot: lib/lib.a vmlinux
- $(Q)$(MAKE) $(build)=$(boot) $@
+ max_pfn = max_low_pfn;
--install: vmlinux.gz
-- sh $(srctree)/arch/ia64/install.sh $(KERNELRELEASE) $< System.map "$(INSTALL_PATH)"
-+install:
-+ -yes | sh $(srctree)/arch/ia64/install.sh $(KERNELRELEASE) vmlinux.gz System.map "$(INSTALL_PATH)"
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/mm/init.c
+--- a/arch/ia64/mm/init.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/mm/init.c Wed Sep 10 10:54:08 2008 +0100
+@@ -303,16 +303,34 @@
+ setup_gate (void)
+ {
+ struct page *page;
++ void *gate_page_addr = __start_gate_section;
++
++#ifdef CONFIG_XEN
++ unsigned long unused_gate;
++ extern char __start_xen_gate_section[];
++ if (is_running_on_xen()) {
++ gate_page_addr = __start_xen_gate_section;
++ unused_gate = (unsigned long)ia64_imva(__start_gate_section);
++ } else
++ unused_gate =
++ (unsigned long)ia64_imva(__start_xen_gate_section);
++#ifndef HAVE_BUGGY_SEGREL
++ ClearPageReserved(virt_to_page(unused_gate));
++ init_page_count(virt_to_page(unused_gate));
++ free_page(unused_gate);
++ ++totalram_pages;
++#endif
++#endif
- define archhelp
- echo '* compressed - Build compressed kernel image'
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/mm/ioremap.c tmp-linux-2.6-xen.patch/arch/ia64/mm/ioremap.c
---- pristine-linux-2.6.18.2/arch/ia64/mm/ioremap.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/mm/ioremap.c 2007-07-30 16:35:11.000000000 +0200
+ /*
+ * Map the gate page twice: once read-only to export the ELF
+ * headers etc. and once execute-only page to enable
+ * privilege-promotion via "epc":
+ */
+- page = virt_to_page(ia64_imva(__start_gate_section));
++ page = virt_to_page(ia64_imva(gate_page_addr));
+ put_kernel_page(page, GATE_ADDR, PAGE_READONLY);
+ #ifdef HAVE_BUGGY_SEGREL
+- page = virt_to_page(ia64_imva(__start_gate_section + PAGE_SIZE));
++ page = virt_to_page(ia64_imva(gate_page_addr + PAGE_SIZE));
+ put_kernel_page(page, GATE_ADDR + PAGE_SIZE, PAGE_GATE);
+ #else
+ put_kernel_page(page, GATE_ADDR + PERCPU_PAGE_SIZE, PAGE_GATE);
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/mm/ioremap.c
+--- a/arch/ia64/mm/ioremap.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/mm/ioremap.c Wed Sep 10 10:54:08 2008 +0100
@@ -16,6 +16,9 @@
static inline void __iomem *
__ioremap (unsigned long offset, unsigned long size)
{
+ offset = HYPERVISOR_ioremap(offset, size);
+ if (IS_ERR_VALUE(offset))
-+ return (void __iomem*)offset;
++ return NULL;
return (void __iomem *) (__IA64_UNCACHED_OFFSET | offset);
}
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/oprofile/init.c tmp-linux-2.6-xen.patch/arch/ia64/oprofile/init.c
---- pristine-linux-2.6.18.2/arch/ia64/oprofile/init.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/oprofile/init.c 2007-07-30 16:35:11.000000000 +0200
-@@ -11,6 +11,7 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/oprofile/Makefile
+--- a/arch/ia64/oprofile/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/oprofile/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -8,3 +8,7 @@
+
+ oprofile-y := $(DRIVER_OBJS) init.o backtrace.o
+ oprofile-$(CONFIG_PERFMON) += perfmon.o
++ifeq ($(CONFIG_XEN), y)
++oprofile-$(CONFIG_PERFMON) += xenoprof.o \
++ ../../../drivers/xen/xenoprof/xenoprofile.o
++endif
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/oprofile/init.c
+--- a/arch/ia64/oprofile/init.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/oprofile/init.c Wed Sep 10 10:54:08 2008 +0100
+@@ -11,6 +11,8 @@
#include <linux/oprofile.h>
#include <linux/init.h>
#include <linux/errno.h>
++#include <asm/hypervisor.h>
+#include "oprofile_perfmon.h"
extern int perfmon_init(struct oprofile_operations * ops);
extern void perfmon_exit(void);
-@@ -20,6 +21,13 @@ int __init oprofile_arch_init(struct opr
+@@ -19,6 +21,13 @@
+ int __init oprofile_arch_init(struct oprofile_operations * ops)
{
int ret = -ENODEV;
-
++
+ if (is_running_on_xen()) {
+ ret = xen_perfmon_init();
+ if (ret)
+ return ret;
+ return xenoprofile_init(ops);
+ }
-+
+
#ifdef CONFIG_PERFMON
/* perfmon_init() can fail, but we have no way to report it */
- ret = perfmon_init(ops);
-@@ -32,6 +40,12 @@ int __init oprofile_arch_init(struct opr
+@@ -32,6 +41,12 @@
void oprofile_arch_exit(void)
{
@@ -24056,29 +26536,20 @@
#ifdef CONFIG_PERFMON
perfmon_exit();
#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/oprofile/Makefile tmp-linux-2.6-xen.patch/arch/ia64/oprofile/Makefile
---- pristine-linux-2.6.18.2/arch/ia64/oprofile/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/oprofile/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -8,3 +8,7 @@ DRIVER_OBJS := $(addprefix ../../../driv
-
- oprofile-y := $(DRIVER_OBJS) init.o backtrace.o
- oprofile-$(CONFIG_PERFMON) += perfmon.o
-+ifeq ($(CONFIG_XEN), y)
-+oprofile-$(CONFIG_PERFMON) += xenoprof.o \
-+ ../../../drivers/xen/xenoprof/xenoprofile.o
-+endif
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/oprofile/oprofile_perfmon.h tmp-linux-2.6-xen.patch/arch/ia64/oprofile/oprofile_perfmon.h
---- pristine-linux-2.6.18.2/arch/ia64/oprofile/oprofile_perfmon.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/oprofile/oprofile_perfmon.h 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,28 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/oprofile/oprofile_perfmon.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/oprofile/oprofile_perfmon.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,30 @@
+#ifndef OPROFILE_PERFMON_H
+#define OPROFILE_PERFMON_H
+
+#ifdef CONFIG_PERFMON
++#ifdef CONFIG_XEN
+int __perfmon_init(void);
+void __perfmon_exit(void);
+int perfmon_start(void);
+void perfmon_stop(void);
++#endif
+#else
+#define __perfmon_init() (-ENOSYS)
+#define __perfmon_exit() do {} while (0)
@@ -24094,14 +26565,14 @@
+#define STATIC_IF_NO_XEN static
+#define xen_perfmon_init() (-ENOSYS)
+#define xen_perfmon_exit() do {} while (0)
-+#define xenoprofile_init() (-ENOSYS)
++#define xenoprofile_init(ops) (-ENOSYS)
+#define xenoprofile_exit() do {} while (0)
+#endif /* CONFIG_XEN */
+
+#endif /* OPROFILE_PERFMON_H */
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/oprofile/perfmon.c tmp-linux-2.6-xen.patch/arch/ia64/oprofile/perfmon.c
---- pristine-linux-2.6.18.2/arch/ia64/oprofile/perfmon.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/oprofile/perfmon.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/oprofile/perfmon.c
+--- a/arch/ia64/oprofile/perfmon.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/oprofile/perfmon.c Wed Sep 10 10:54:08 2008 +0100
@@ -13,6 +13,7 @@
#include <asm/perfmon.h>
#include <asm/ptrace.h>
@@ -24110,7 +26581,7 @@
static int allow_ints;
-@@ -33,14 +34,16 @@ perfmon_handler(struct task_struct *task
+@@ -33,14 +34,16 @@
}
@@ -24129,18 +26600,17 @@
{
allow_ints = 0;
}
-@@ -75,16 +78,35 @@ static char * get_cpu_type(void)
+@@ -75,16 +78,35 @@
static int using_perfmon;
--int perfmon_init(struct oprofile_operations * ops)
+STATIC_IF_NO_XEN
+int __perfmon_init(void)
- {
- int ret = pfm_register_buffer_fmt(&oprofile_fmt);
- if (ret)
- return -ENODEV;
-
++{
++ int ret = pfm_register_buffer_fmt(&oprofile_fmt);
++ if (ret)
++ return -ENODEV;
++
+ using_perfmon = 1;
+ return 0;
+}
@@ -24154,12 +26624,13 @@
+ pfm_unregister_buffer_fmt(oprofile_fmt.fmt_uuid);
+}
+
-+int perfmon_init(struct oprofile_operations * ops)
-+{
+ int perfmon_init(struct oprofile_operations * ops)
+ {
+- int ret = pfm_register_buffer_fmt(&oprofile_fmt);
+ int ret = __perfmon_init();
-+ if (ret)
-+ return -ENODEV;
-+
+ if (ret)
+ return -ENODEV;
+
ops->cpu_type = get_cpu_type();
ops->start = perfmon_start;
ops->stop = perfmon_stop;
@@ -24167,7 +26638,7 @@
printk(KERN_INFO "oprofile: using perfmon.\n");
return 0;
}
-@@ -92,8 +114,5 @@ int perfmon_init(struct oprofile_operati
+@@ -92,8 +114,5 @@
void perfmon_exit(void)
{
@@ -24177,9 +26648,9 @@
- pfm_unregister_buffer_fmt(oprofile_fmt.fmt_uuid);
+ __perfmon_exit();
}
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/oprofile/xenoprof.c tmp-linux-2.6-xen.patch/arch/ia64/oprofile/xenoprof.c
---- pristine-linux-2.6.18.2/arch/ia64/oprofile/xenoprof.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/oprofile/xenoprof.c 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/oprofile/xenoprof.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/oprofile/xenoprof.c Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,142 @@
+/******************************************************************************
+ * xenoprof ia64 specific part
@@ -24234,9 +26705,9 @@
+}
+
+/* XXX move them to an appropriate header file. */
-+struct resource* xen_ia64_allocate_resource(unsigned long size);
-+void xen_ia64_release_resource(struct resource* res);
-+void xen_ia64_unmap_resource(struct resource* res);
++struct resource* xen_ia64_allocate_resource(unsigned long size);
++void xen_ia64_release_resource(struct resource *res);
++void xen_ia64_unmap_resource(struct resource *res);
+
+struct resource*
+xenoprof_ia64_allocate_resource(int32_t max_samples)
@@ -24256,7 +26727,7 @@
+ return xen_ia64_allocate_resource(bufsize);
+}
+
-+void xenoprof_arch_unmap_shared_buffer(struct xenoprof_shared_buffer* sbuf)
++void xenoprof_arch_unmap_shared_buffer(struct xenoprof_shared_buffer *sbuf)
+{
+ if (sbuf->buffer) {
+ xen_ia64_unmap_resource(sbuf->arch.res);
@@ -24265,11 +26736,11 @@
+ }
+}
+
-+int xenoprof_arch_map_shared_buffer(struct xenoprof_get_buffer* get_buffer,
-+ struct xenoprof_shared_buffer* sbuf)
++int xenoprof_arch_map_shared_buffer(struct xenoprof_get_buffer *get_buffer,
++ struct xenoprof_shared_buffer *sbuf)
+{
+ int ret;
-+ struct resource* res;
++ struct resource *res;
+
+ sbuf->buffer = NULL;
+ sbuf->arch.res = NULL;
@@ -24295,11 +26766,11 @@
+ return ret;
+}
+
-+int xenoprof_arch_set_passive(struct xenoprof_passive* pdomain,
-+ struct xenoprof_shared_buffer* sbuf)
++int xenoprof_arch_set_passive(struct xenoprof_passive *pdomain,
++ struct xenoprof_shared_buffer *sbuf)
+{
+ int ret;
-+ struct resource* res;
++ struct resource *res;
+
+ sbuf->buffer = NULL;
+ sbuf->arch.res = NULL;
@@ -24323,52 +26794,227 @@
+
+ return ret;
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/pci/pci.c tmp-linux-2.6-xen.patch/arch/ia64/pci/pci.c
---- pristine-linux-2.6.18.2/arch/ia64/pci/pci.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/pci/pci.c 2007-09-30 18:06:18.000000000 +0200
-@@ -165,6 +165,11 @@ new_space (u64 phys_base, int sparse)
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/pci/pci.c
+--- a/arch/ia64/pci/pci.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/pci/pci.c Wed Sep 10 10:54:08 2008 +0100
+@@ -29,6 +29,15 @@
+ #include <asm/smp.h>
+ #include <asm/irq.h>
+ #include <asm/hw_irq.h>
++
++#ifdef CONFIG_XEN
++struct ioremap_issue_list {
++ struct list_head listp;
++ unsigned long start;
++ unsigned long end;
++};
++typedef struct ioremap_issue_list ioremap_issue_list_t;
++#endif /* CONFIG_XEN */
+
+ /*
+ * Low-level SAL-based PCI configuration access functions. Note that SAL
+@@ -164,6 +173,11 @@
+ i = num_io_spaces++;
io_space[i].mmio_base = mmio_base;
io_space[i].sparse = sparse;
-
++
+#ifdef CONFIG_XEN
+ if (is_initial_xendomain())
+ HYPERVISOR_add_io_space(phys_base, sparse, i);
+#endif
-+
+
return i;
}
+@@ -332,6 +346,169 @@
+ }
+ }
-@@ -607,6 +612,14 @@ pci_mmap_page_range (struct pci_dev *dev
- else
- vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
-
-+ if (is_initial_xendomain()) {
-+ unsigned long addr = vma->vm_pgoff << PAGE_SHIFT;
-+ size_t size = vma->vm_end - vma->vm_start;
-+ unsigned long offset = HYPERVISOR_ioremap(addr, size);
-+ if (IS_ERR_VALUE(offset))
-+ return offset;
++#ifdef CONFIG_XEN
++static void __devinit
++__cleanup_issue_list(struct list_head *top)
++{
++ ioremap_issue_list_t *ptr, *tmp_ptr;
++
++ list_for_each_entry_safe(ptr, tmp_ptr, top, listp) {
++ list_del(&(ptr->listp));
++ kfree(ptr);
+ }
++}
+
- if (remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff,
- vma->vm_end - vma->vm_start, vma->vm_page_prot))
- return -EAGAIN;
-@@ -664,6 +677,14 @@ pci_mmap_legacy_page_range(struct pci_bu
- vma->vm_pgoff += (unsigned long)addr >> PAGE_SHIFT;
- vma->vm_page_prot = prot;
-
-+ if (is_initial_xendomain()) {
-+ unsigned long addr = vma->vm_pgoff << PAGE_SHIFT;
-+ size_t size = vma->vm_end - vma->vm_start;
-+ unsigned long offset = HYPERVISOR_ioremap(addr, size);
-+ if (IS_ERR_VALUE(offset))
-+ return offset;
++static int __devinit
++__add_issue_list(unsigned long start, unsigned long end, struct list_head *top)
++{
++ ioremap_issue_list_t *ptr, *new;
++
++ if (start > end) {
++ printk(KERN_ERR "%s: Internal error (start addr > end addr)\n",
++ __FUNCTION__);
++ return 0;
++ }
++
++ /*
++ * Head of the resource structure list contains
++ * dummy val.(start=0, end=~0), so skip it
++ */
++ if ((start == 0) && (end == ~0))
++ return 0;
++
++ start &= PAGE_MASK;
++ end |= ~PAGE_MASK;
++
++ /* We can merge specified address range into existing entry */
++ list_for_each_entry(ptr, top, listp) {
++ if ((ptr->start > end + 1) || (ptr->end + 1 < start))
++ continue;
++ ptr->start = min(start, ptr->start);
++ ptr->end = max(end, ptr->end);
++ return 0;
++ }
++
++ /* We could not merge, so create new entry */
++ new = kmalloc(sizeof(ioremap_issue_list_t), GFP_KERNEL);
++ if (new == NULL) {
++ printk(KERN_ERR "%s: Could not allocate memory. "
++ "HYPERVISOR_ioremap will not be issued\n",
++ __FUNCTION__);
++ return -ENOMEM;
++ }
++
++ new->start = start;
++ new->end = end;
++
++ /* Insert the new entry to the list by ascending order */
++ if (list_empty(top)) {
++ list_add_tail(&(new->listp), top);
++ return 0;
++ }
++ list_for_each_entry(ptr, top, listp) {
++ if (new->start > ptr->start)
++ continue;
++ list_add(&(new->listp), ((struct list_head *)ptr)->prev);
++ return 0;
++ }
++ list_add_tail(&(new->listp), top);
++
++ return 0;
++}
++
++static int __devinit
++__make_issue_list(struct resource *ptr, struct list_head *top)
++{
++ int ret;
++
++ if (ptr->child) {
++ ret = __make_issue_list(ptr->child, top);
++ if (ret)
++ return ret;
++ }
++ if (ptr->sibling) {
++ ret = __make_issue_list(ptr->sibling, top);
++ if (ret)
++ return ret;
++ }
++
++ if (ptr->flags & IORESOURCE_MEM) {
++ ret = __add_issue_list(ptr->start, ptr->end, top);
++ if (ret)
++ return ret;
++ }
++
++ return 0;
++}
++
++static void __devinit
++__compress_issue_list(struct list_head *top)
++{
++ ioremap_issue_list_t *ptr, *tmp_ptr, *next;
++ int compressed;
++
++ /*
++ * Merge adjacent entries, if overlapped
++ * (entries are sorted by ascending order)
++ */
++ list_for_each_entry_safe(ptr, tmp_ptr, top, listp) {
++ if (list_is_last((struct list_head *)ptr, top))
++ continue;
++
++ next = (ioremap_issue_list_t *)
++ (((struct list_head *)ptr)->next);
++ if (next->start <= (ptr->end) + 1) {
++ next->start = min(ptr->start, next->start);
++ next->end = max(ptr->end, next->end);
++
++ list_del(&(ptr->listp));
++ kfree(ptr);
++ }
++ }
++}
++
++static int __devinit
++__issue_ioremap(struct list_head *top)
++{
++ ioremap_issue_list_t *ptr, *tmp_ptr;
++ unsigned int offset;
++
++ list_for_each_entry_safe(ptr, tmp_ptr, top, listp) {
++ offset = HYPERVISOR_ioremap(ptr->start,
++ ptr->end - ptr->start + 1);
++ if (offset == ~0) {
++ printk(KERN_ERR "%s: HYPERVISOR_ioremap() failed. "
++ "Address Range: 0x%016lx-0x%016lx\n",
++ __FUNCTION__, ptr->start, ptr->end);
++ }
++
++ list_del(&(ptr->listp));
++ kfree(ptr);
++ }
++
++ return 0;
++}
++
++static int __devinit
++do_ioremap_on_resource_list(struct resource *top)
++{
++ LIST_HEAD(ioremap_issue_list_top);
++ int ret;
++
++ ret = __make_issue_list(top, &ioremap_issue_list_top);
++ if (ret) {
++ __cleanup_issue_list(&ioremap_issue_list_top);
++ return ret;
+ }
+
- if (remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff,
- size, vma->vm_page_prot))
- return -EAGAIN;
-@@ -818,3 +839,31 @@ int pci_vector_resources(int last, int n
++ __compress_issue_list(&ioremap_issue_list_top);
++
++ (void)__issue_ioremap(&ioremap_issue_list_top);
++
++ return 0;
++}
++#endif /* CONFIG_XEN */
++
+ struct pci_bus * __devinit
+ pci_acpi_scan_root(struct acpi_device *device, int domain, int bus)
+ {
+@@ -374,6 +551,18 @@
+ pbus = pci_scan_bus_parented(NULL, bus, &pci_root_ops, controller);
+ if (pbus)
+ pcibios_setup_root_windows(pbus, controller);
++
++#ifdef CONFIG_XEN
++ if (is_initial_xendomain()) {
++ if (do_ioremap_on_resource_list(&iomem_resource) != 0) {
++ printk(KERN_ERR
++ "%s: Counld not issue HYPERVISOR_ioremap "
++ "due to lack of memory or hypercall failure\n",
++ __FUNCTION__);
++ goto out3;
++ }
++ }
++#endif /* CONFIG_XEN */
+
+ return pbus;
+
+@@ -818,3 +1007,31 @@
return count;
}
@@ -24400,23 +27046,112 @@
+}
+EXPORT_SYMBOL(xen_pcibios_setup_root_windows);
+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/hypercall.S tmp-linux-2.6-xen.patch/arch/ia64/xen/hypercall.S
---- pristine-linux-2.6.18.2/arch/ia64/xen/hypercall.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/hypercall.S 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,170 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/sn/kernel/setup.c
+--- a/arch/ia64/sn/kernel/setup.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/sn/kernel/setup.c Wed Sep 10 10:54:08 2008 +0100
+@@ -763,5 +763,13 @@
+ return 0;
+ return test_bit(id, sn_prom_features);
+ }
++
++void
++sn_kernel_launch_event(void)
++{
++ /* ignore status until we understand possible failure, if any*/
++ if (ia64_sn_kernel_launch_event())
++ printk(KERN_ERR "KEXEC is not supported in this PROM, Please update the PROM.\n");
++}
+ EXPORT_SYMBOL(sn_prom_feature_available);
+
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/sn/pci/pcibr/pcibr_provider.c
+--- a/arch/ia64/sn/pci/pcibr/pcibr_provider.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/sn/pci/pcibr/pcibr_provider.c Wed Sep 10 10:54:08 2008 +0100
+@@ -15,6 +15,7 @@
+ #include <asm/sn/pcibus_provider_defs.h>
+ #include <asm/sn/pcidev.h>
+ #include <asm/sn/sn_sal.h>
++#include <asm/sn/pic.h>
+ #include <asm/sn/sn2/sn_hwperf.h>
+ #include "xtalk/xwidgetdev.h"
+ #include "xtalk/hubdev.h"
+@@ -129,9 +130,9 @@
+ }
+
+ memcpy(soft, prom_bussoft, sizeof(struct pcibus_info));
+- soft->pbi_buscommon.bs_base =
+- (((u64) soft->pbi_buscommon.
+- bs_base << 4) >> 4) | __IA64_UNCACHED_OFFSET;
++ soft->pbi_buscommon.bs_base = (unsigned long)
++ ioremap(REGION_OFFSET(soft->pbi_buscommon.bs_base),
++ sizeof(struct pic));
+
+ spin_lock_init(&soft->pbi_lock);
+
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/sn/pci/tioca_provider.c
+--- a/arch/ia64/sn/pci/tioca_provider.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/sn/pci/tioca_provider.c Wed Sep 10 10:54:08 2008 +0100
+@@ -611,7 +611,9 @@
+ return NULL;
+
+ memcpy(tioca_common, prom_bussoft, sizeof(struct tioca_common));
+- tioca_common->ca_common.bs_base |= __IA64_UNCACHED_OFFSET;
++ tioca_common->ca_common.bs_base = (unsigned long)
++ ioremap(REGION_OFFSET(tioca_common->ca_common.bs_base),
++ sizeof(struct tioca_common));
+
+ /* init kernel-private area */
+
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/sn/pci/tioce_provider.c
+--- a/arch/ia64/sn/pci/tioce_provider.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/ia64/sn/pci/tioce_provider.c Wed Sep 10 10:54:08 2008 +0100
+@@ -1006,7 +1006,9 @@
+ return NULL;
+
+ memcpy(tioce_common, prom_bussoft, sizeof(struct tioce_common));
+- tioce_common->ce_pcibus.bs_base |= __IA64_UNCACHED_OFFSET;
++ tioce_common->ce_pcibus.bs_base = (unsigned long)
++ ioremap(REGION_OFFSET(tioce_common->ce_pcibus.bs_base),
++ sizeof(struct tioce_common));
+
+ tioce_kern = tioce_kern_init(tioce_common);
+ if (tioce_kern == NULL) {
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,9 @@
++#
++# Makefile for Xen components
++#
++
++obj-y := hypercall.o xenivt.o xenentry.o xensetup.o xenpal.o \
++ hypervisor.o util.o xencomm.o xcom_hcall.o \
++ xcom_privcmd.o xcom_asm.o xen_dma.o
++
++obj-$(CONFIG_IA64_GENERIC) += machvec.o
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/hypercall.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/hypercall.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,141 @@
+/*
+ * Support routines for Xen hypercalls
+ *
+ * Copyright (C) 2005 Dan Magenheimer <dan.magenheimer at hp.com>
+ */
+
-+#include <asm/processor.h>
+#include <asm/asmmacro.h>
++#include <asm/intrinsics.h>
++
++#ifdef __INTEL_COMPILER
++# undef ASM_SUPPORTED
++#else
++# define ASM_SUPPORTED
++#endif
+
++#ifndef ASM_SUPPORTED
+GLOBAL_ENTRY(xen_get_psr)
+ XEN_HYPER_GET_PSR
+ br.ret.sptk.many rp
-+ ;;
++ ;;
+END(xen_get_psr)
+
+GLOBAL_ENTRY(xen_get_ivr)
@@ -24519,65 +27254,29 @@
+ XEN_HYPER_SET_EFLAG
+ br.ret.sptk.many rp
+END(xen_set_eflag)
-+#endif
++#endif /* CONFIG_IA32_SUPPORT */
++#endif /* ASM_SUPPORTED */
+
+GLOBAL_ENTRY(xen_send_ipi)
-+ mov r14=r32
-+ mov r15=r33
-+ mov r2=0x400
-+ break 0x1000
-+ ;;
-+ br.ret.sptk.many rp
-+ ;;
++ mov r14=r32
++ mov r15=r33
++ mov r2=0x400
++ break 0x1000
++ ;;
++ br.ret.sptk.many rp
++ ;;
+END(xen_send_ipi)
+
-+#ifdef CONFIG_XEN_IA64_VDSO_PARAVIRT
-+// Those are vdso specialized.
-+// In fsys mode, call, ret can't be used.
-+
-+ // see xen_ssm_i() in privop.h
-+ // r22 = &vcpu->vcpu_info->evtchn_upcall_mask
-+ // r23 = &vpsr.ic
-+ // r24 = &vcpu->vcpu_info->evtchn_upcall_pending
-+ // r25 = tmp
-+ // r31 = tmp
-+ // p11 = tmp
-+ // p14 = tmp
-+#define XEN_SET_PSR_I \
-+ ld1 r31=[r22]; \
-+ ld1 r25=[r24]; \
-+ ;; \
-+ st1 [r22]=r0; \
-+ cmp.ne.unc p14,p0=r0,r31; \
-+ ;; \
-+(p14) cmp.ne.unc p11,p0=r0,r25; \
-+ ;; \
-+(p11) st1 [r22]=r20; \
-+(p11) XEN_HYPER_SSM_I;
-+
-+GLOBAL_ENTRY(xen_ssm_i_0)
-+ XEN_SET_PSR_I
-+ brl.cond.sptk .vdso_ssm_i_0_ret
-+ ;;
-+END(xen_ssm_i_0)
-+
-+GLOBAL_ENTRY(xen_ssm_i_1)
-+ XEN_SET_PSR_I
-+ brl.cond.sptk .vdso_ssm_i_1_ret
-+ ;;
-+END(xen_ssm_i_1)
-+
+GLOBAL_ENTRY(__hypercall)
+ mov r2=r37
+ break 0x1000
+ br.ret.sptk.many b0
+ ;;
+END(__hypercall)
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/hypervisor.c tmp-linux-2.6-xen.patch/arch/ia64/xen/hypervisor.c
---- pristine-linux-2.6.18.2/arch/ia64/xen/hypervisor.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/hypervisor.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,1234 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/hypervisor.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/hypervisor.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1526 @@
+/******************************************************************************
+ * include/asm-ia64/shadow.h
+ *
@@ -24600,7 +27299,6 @@
+ *
+ */
+
-+//#include <linux/kernel.h>
+#include <linux/spinlock.h>
+#include <linux/bootmem.h>
+#include <linux/module.h>
@@ -24612,15 +27310,16 @@
+#include <asm/hypervisor.h>
+#include <asm/hypercall.h>
+#include <xen/interface/memory.h>
++#include <xen/xencons.h>
+#include <xen/balloon.h>
+
-+shared_info_t *HYPERVISOR_shared_info = (shared_info_t *)XSI_BASE;
++shared_info_t *HYPERVISOR_shared_info __read_mostly =
++ (shared_info_t *)XSI_BASE;
+EXPORT_SYMBOL(HYPERVISOR_shared_info);
+
+start_info_t *xen_start_info;
+EXPORT_SYMBOL(xen_start_info);
+
-+int running_on_xen;
+EXPORT_SYMBOL(running_on_xen);
+
+#ifdef CONFIG_XEN_IA64_EXPOSE_P2M
@@ -24632,15 +27331,44 @@
+
+EXPORT_SYMBOL(__hypercall);
+
-+//XXX same as i386, x86_64 contiguous_bitmap_set(), contiguous_bitmap_clear()
-+// move those to lib/contiguous_bitmap?
-+//XXX discontigmem/sparsemem
++void __init
++xen_setup(char **cmdline_p)
++{
++ struct dom0_vga_console_info *info;
++ extern void dig_setup(char **cmdline_p);
++
++ if (ia64_platform_is("xen"))
++ dig_setup(cmdline_p);
++
++ if (!is_running_on_xen() || !is_initial_xendomain())
++ return;
++
++ info = (void *)((char *)xen_start_info +
++ xen_start_info->console.dom0.info_off);
++ dom0_init_screen_info(info, xen_start_info->console.dom0.info_size);
++
++ xen_start_info->console.domU.mfn = 0;
++ xen_start_info->console.domU.evtchn = 0;
++}
++
++void __cpuinit
++xen_cpu_init(void)
++{
++ extern void xen_smp_intr_init(void);
++ xen_smp_intr_init();
++}
++
++/*
++ *XXX same as i386, x86_64 contiguous_bitmap_set(), contiguous_bitmap_clear()
++ * move those to lib/contiguous_bitmap?
++ *XXX discontigmem/sparsemem
++ */
+
+/*
+ * Bitmap is indexed by page number. If bit is set, the page is part of a
+ * xen_create_contiguous_region() area of memory.
+ */
-+unsigned long *contiguous_bitmap;
++unsigned long *contiguous_bitmap __read_mostly;
+
+#ifdef CONFIG_VIRTUAL_MEM_MAP
+/* Following logic is stolen from create_mem_map_table() for virtual memmap */
@@ -24657,16 +27385,16 @@
+ pte_t *pte;
+
+ bitmap_start = (unsigned long)contiguous_bitmap +
-+ ((__pa(start) >> PAGE_SHIFT) >> 3);
++ ((__pa(start) >> PAGE_SHIFT) >> 3);
+ bitmap_end = (unsigned long)contiguous_bitmap +
-+ (((__pa(end) >> PAGE_SHIFT) + 2 * BITS_PER_LONG) >> 3);
++ (((__pa(end) >> PAGE_SHIFT) + 2 * BITS_PER_LONG) >> 3);
+
+ start_page = bitmap_start & PAGE_MASK;
+ end_page = PAGE_ALIGN(bitmap_end);
+ node = paddr_to_nid(__pa(start));
+
+ bitmap = alloc_bootmem_pages_node(NODE_DATA(node),
-+ end_page - start_page);
++ end_page - start_page);
+ BUG_ON(!bitmap);
+ memset(bitmap, 0, end_page - start_page);
+
@@ -24674,26 +27402,26 @@
+ pgd = pgd_offset_k(address);
+ if (pgd_none(*pgd))
+ pgd_populate(&init_mm, pgd,
-+ alloc_bootmem_pages_node(NODE_DATA(node),
-+ PAGE_SIZE));
++ alloc_bootmem_pages_node(NODE_DATA(node),
++ PAGE_SIZE));
+ pud = pud_offset(pgd, address);
+
+ if (pud_none(*pud))
+ pud_populate(&init_mm, pud,
-+ alloc_bootmem_pages_node(NODE_DATA(node),
-+ PAGE_SIZE));
++ alloc_bootmem_pages_node(NODE_DATA(node),
++ PAGE_SIZE));
+ pmd = pmd_offset(pud, address);
+
+ if (pmd_none(*pmd))
+ pmd_populate_kernel(&init_mm, pmd,
-+ alloc_bootmem_pages_node
-+ (NODE_DATA(node), PAGE_SIZE));
++ alloc_bootmem_pages_node
++ (NODE_DATA(node), PAGE_SIZE));
+ pte = pte_offset_kernel(pmd, address);
+
+ if (pte_none(*pte))
+ set_pte(pte,
-+ pfn_pte(__pa(bitmap + (address - start_page))
-+ >> PAGE_SHIFT, PAGE_KERNEL));
++ pfn_pte(__pa(bitmap + (address - start_page))
++ >> PAGE_SHIFT, PAGE_KERNEL));
+ }
+ return 0;
+}
@@ -24708,7 +27436,7 @@
+}
+
+void
-+contiguous_bitmap_init(unsigned long end_pfn)
++xen_contiguous_bitmap_init(unsigned long end_pfn)
+{
+ unsigned long size = (end_pfn + 2 * BITS_PER_LONG) >> 3;
+#ifndef CONFIG_VIRTUAL_MEM_MAP
@@ -24778,9 +27506,11 @@
+ }
+}
+
-+// __xen_create_contiguous_region(), __xen_destroy_contiguous_region()
-+// are based on i386 xen_create_contiguous_region(),
-+// xen_destroy_contiguous_region()
++/*
++ * __xen_create_contiguous_region(), __xen_destroy_contiguous_region()
++ * are based on i386 xen_create_contiguous_region(),
++ * xen_destroy_contiguous_region()
++ */
+
+/* Protected by balloon_lock. */
+#define MAX_CONTIG_ORDER 7
@@ -24826,9 +27556,8 @@
+ balloon_lock(flags);
+
+ /* Get a new contiguous memory extent. */
-+ for (i = 0; i < num_gpfn; i++) {
++ for (i = 0; i < num_gpfn; i++)
+ in_frames[i] = start_gpfn + i;
-+ }
+ out_frame = start_gpfn;
+ error = HYPERVISOR_memory_op(XENMEM_exchange, &exchange);
+ success = (exchange.nr_exchanged == num_gpfn);
@@ -24910,7 +27639,7 @@
+ .domid = DOMID_SELF
+ },
+ .nr_exchanged = 0
-+ };
++ };
+
+
+ if (!test_bit(start_gpfn, contiguous_bitmap))
@@ -24928,17 +27657,16 @@
+
+ contiguous_bitmap_clear(start_gpfn, num_gpfn);
+
-+ /* Do the exchange for non-contiguous MFNs. */
++ /* Do the exchange for non-contiguous MFNs. */
+ in_frame = start_gpfn;
-+ for (i = 0; i < num_gpfn; i++) {
++ for (i = 0; i < num_gpfn; i++)
+ out_frames[i] = start_gpfn + i;
-+ }
+ error = HYPERVISOR_memory_op(XENMEM_exchange, &exchange);
+ success = (exchange.nr_exchanged == 1);
+ BUG_ON(!success && ((exchange.nr_exchanged != 0) || (error == 0)));
+ BUG_ON(success && (error != 0));
+ if (unlikely(error == -ENOSYS)) {
-+ /* Compatibility when XENMEM_exchange is unsupported. */
++ /* Compatibility when XENMEM_exchange is unsupported. */
+ error = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
+ &exchange.in);
+ BUG_ON(error != 1);
@@ -24950,15 +27678,27 @@
+ balloon_unlock(flags);
+}
+
++int
++xen_limit_pages_to_max_mfn(struct page *pages, unsigned int order,
++ unsigned int address_bits)
++{
++ return xen_create_contiguous_region((unsigned long)page_address(pages),
++ order, address_bits);
++}
+
-+///////////////////////////////////////////////////////////////////////////
-+// grant table hack
-+// cmd: GNTTABOP_xxx
-+
++/****************************************************************************
++ * grant table hack
++ * cmd: GNTTABOP_xxx
++ */
+#include <linux/mm.h>
+#include <xen/interface/xen.h>
+#include <xen/gnttab.h>
+
++void *arch_gnttab_alloc_shared(unsigned long *frames)
++{
++ return __va(frames[0] << PAGE_SHIFT);
++}
++
+static void
+gnttab_map_grant_ref_pre(struct gnttab_map_grant_ref *uop)
+{
@@ -24968,16 +27708,19 @@
+
+ if (flags & GNTMAP_host_map) {
+ if (flags & GNTMAP_application_map) {
-+ xprintd("GNTMAP_application_map is not supported yet: flags 0x%x\n", flags);
++ xprintd("GNTMAP_application_map is not supported yet:"
++ " flags 0x%x\n", flags);
+ BUG();
+ }
+ if (flags & GNTMAP_contains_pte) {
-+ xprintd("GNTMAP_contains_pte is not supported yet flags 0x%x\n", flags);
++ xprintd("GNTMAP_contains_pte is not supported yet"
++ " flags 0x%x\n", flags);
+ BUG();
+ }
+ } else if (flags & GNTMAP_device_map) {
-+ xprintd("GNTMAP_device_map is not supported yet 0x%x\n", flags);
-+ BUG();//XXX not yet. actually this flag is not used.
++ xprintd("GNTMAP_device_map is not supported yet 0x%x\n",
++ flags);
++ BUG(); /* XXX not yet. actually this flag is not used. */
+ } else {
+ BUG();
+ }
@@ -24993,19 +27736,21 @@
+ (struct gnttab_map_grant_ref*)uop + i);
+ }
+ }
-+ return xencomm_mini_hypercall_grant_table_op(cmd, uop, count);
++ return xencomm_hypercall_grant_table_op(cmd, uop, count);
+}
+EXPORT_SYMBOL(HYPERVISOR_grant_table_op);
+
-+///////////////////////////////////////////////////////////////////////////
-+// foreign mapping
++/**************************************************************************
++ * foreign mapping
++ */
+#include <linux/efi.h>
-+#include <asm/meminit.h> // for IA64_GRANULE_SIZE, GRANULEROUND{UP,DOWN}()
++#include <asm/meminit.h> /* for IA64_GRANULE_SIZE, GRANULEROUND{UP,DOWN}() */
+
+static unsigned long privcmd_resource_min = 0;
-+// Xen/ia64 currently can handle pseudo physical address bits up to
-+// (PAGE_SHIFT * 3)
-+static unsigned long privcmd_resource_max = GRANULEROUNDDOWN((1UL << (PAGE_SHIFT * 3)) - 1);
++/* Xen/ia64 currently can handle pseudo physical address bits up to
++ * (PAGE_SHIFT * 3) */
++static unsigned long privcmd_resource_max =
++ GRANULEROUNDDOWN((1UL << (PAGE_SHIFT * 3)) - 1);
+static unsigned long privcmd_resource_align = IA64_GRANULE_SIZE;
+
+static unsigned long
@@ -25040,18 +27785,18 @@
+ efi_map_end = efi_map_start + ia64_boot_param->efi_memmap_size;
+ efi_desc_size = ia64_boot_param->efi_memdesc_size;
+
-+ // at first check the used highest address
++ /* at first check the used highest address */
+ for (p = efi_map_start; p < efi_map_end; p += efi_desc_size) {
-+ // nothing
++ /* nothing */;
+ }
+ md = p - efi_desc_size;
+ privcmd_resource_min = GRANULEROUNDUP(md_end_addr(md));
+ if (xen_ia64_privcmd_check_size(privcmd_resource_min,
-+ privcmd_resource_max)) {
++ privcmd_resource_max))
+ goto out;
-+ }
+
-+ // the used highest address is too large. try to find the largest gap.
++ /* the used highest address is too large.
++ * try to find the largest gap. */
+ tmp_min = privcmd_resource_max;
+ tmp_max = 0;
+ gap_size = 0;
@@ -25065,23 +27810,21 @@
+
+ md = p;
+ end = md_end_addr(md);
-+ if (end > privcmd_resource_max) {
++ if (end > privcmd_resource_max)
+ break;
-+ }
+ if (end < prev_end) {
-+ // work around.
-+ // Xen may pass incompletely sorted memory
-+ // descriptors like
-+ // [x, x + length]
-+ // [x, x]
-+ // this order should be reversed.
++ /* work around.
++ * Xen may pass incompletely sorted memory
++ * descriptors like
++ * [x, x + length]
++ * [x, x]
++ * this order should be reversed. */
+ continue;
+ }
+ next = p + efi_desc_size;
+ next_start = next->phys_addr;
-+ if (next_start > privcmd_resource_max) {
++ if (next_start > privcmd_resource_max)
+ next_start = privcmd_resource_max;
-+ }
+ if (end < next_start && gap_size < (next_start - end)) {
+ tmp_min = end;
+ tmp_max = next_start;
@@ -25100,19 +27843,21 @@
+ privcmd_resource_max = tmp_max;
+ if (!xen_ia64_privcmd_check_size(privcmd_resource_min,
+ privcmd_resource_max)) {
-+ // Any large enough gap isn't found.
-+ // go ahead anyway with the warning hoping that large region
-+ // won't be requested.
-+ printk(KERN_WARNING "xen privcmd: large enough region for privcmd mmap is not found.\n");
++ /* Any large enough gap isn't found.
++ * go ahead anyway with the warning hoping that large region
++ * won't be requested. */
++ printk(KERN_WARNING "xen privcmd: "
++ "large enough region for privcmd mmap is not found.\n");
+ }
+
+out:
-+ printk(KERN_INFO "xen privcmd uses pseudo physical addr range [0x%lx, 0x%lx] (%ldMB)\n",
++ printk(KERN_INFO "xen privcmd uses pseudo physical addr range "
++ "[0x%lx, 0x%lx] (%ldMB)\n",
+ privcmd_resource_min, privcmd_resource_max,
+ (privcmd_resource_max - privcmd_resource_min) >> 20);
+ BUG_ON(privcmd_resource_min >= privcmd_resource_max);
+
-+ // XXX this should be somewhere appropriate
++ /* XXX this should be somewhere appropriate */
+ (void)p2m_expose_init();
+
+ return 0;
@@ -25127,8 +27872,12 @@
+
+struct xen_ia64_privcmd_range {
+ atomic_t ref_count;
-+ unsigned long pgoff; // in PAGE_SIZE
-+ struct resource* res;
++ unsigned long pgoff; /* in PAGE_SIZE */
++ struct resource *res;
++
++ /* for foreign domain p2m mapping */
++ void *private;
++ void (*callback)(struct xen_ia64_privcmd_range *range, void *arg);
+
+ unsigned long num_entries;
+ struct xen_ia64_privcmd_entry entries[0];
@@ -25136,30 +27885,30 @@
+
+struct xen_ia64_privcmd_vma {
+ int is_privcmd_mmapped;
-+ struct xen_ia64_privcmd_range* range;
++ struct xen_ia64_privcmd_range *range;
+
+ unsigned long num_entries;
-+ struct xen_ia64_privcmd_entry* entries;
++ struct xen_ia64_privcmd_entry *entries;
+};
+
+static void
-+xen_ia64_privcmd_init_entry(struct xen_ia64_privcmd_entry* entry)
++xen_ia64_privcmd_init_entry(struct xen_ia64_privcmd_entry *entry)
+{
+ atomic_set(&entry->map_count, 0);
+ entry->gpfn = INVALID_GPFN;
+}
+
+static int
-+xen_ia64_privcmd_entry_mmap(struct vm_area_struct* vma,
++xen_ia64_privcmd_entry_mmap(struct vm_area_struct *vma,
+ unsigned long addr,
-+ struct xen_ia64_privcmd_range* privcmd_range,
++ struct xen_ia64_privcmd_range *privcmd_range,
+ int i,
+ unsigned long gmfn,
+ pgprot_t prot,
+ domid_t domid)
+{
+ int error = 0;
-+ struct xen_ia64_privcmd_entry* entry = &privcmd_range->entries[i];
++ struct xen_ia64_privcmd_entry *entry = &privcmd_range->entries[i];
+ unsigned long gpfn;
+ unsigned long flags;
+
@@ -25175,21 +27924,24 @@
+ gpfn = (privcmd_range->res->start >> PAGE_SHIFT) + i;
+
+ flags = ASSIGN_writable;
-+ if (pgprot_val(prot) == PROT_READ) {
++ if (pgprot_val(prot) == PROT_READ)
+ flags = ASSIGN_readonly;
-+ }
+ error = HYPERVISOR_add_physmap_with_gmfn(gpfn, gmfn, flags, domid);
-+ if (error != 0) {
++ if (error != 0)
+ goto out;
-+ }
+
+ prot = vma->vm_page_prot;
+ error = remap_pfn_range(vma, addr, gpfn, 1 << PAGE_SHIFT, prot);
++ /*
++ * VM_PFNMAP is set in remap_pfn_range().
++ * Reset the flag to avoid BUG_ON() in do_no_page().
++ */
++ vma->vm_flags &= ~VM_PFNMAP;
++
+ if (error != 0) {
+ error = HYPERVISOR_zap_physmap(gpfn, 0);
-+ if (error) {
-+ BUG();//XXX
-+ }
++ if (error)
++ BUG(); /* XXX */
+ } else {
+ atomic_inc(&entry->map_count);
+ entry->gpfn = gpfn;
@@ -25200,60 +27952,67 @@
+}
+
+static void
-+xen_ia64_privcmd_entry_munmap(struct xen_ia64_privcmd_range* privcmd_range,
++xen_ia64_privcmd_entry_munmap(struct xen_ia64_privcmd_range *privcmd_range,
+ int i)
+{
-+ struct xen_ia64_privcmd_entry* entry = &privcmd_range->entries[i];
++ struct xen_ia64_privcmd_entry *entry = &privcmd_range->entries[i];
+ unsigned long gpfn = entry->gpfn;
-+ //gpfn = (privcmd_range->res->start >> PAGE_SHIFT) +
-+ // (vma->vm_pgoff - privcmd_range->pgoff);
++ /* gpfn = (privcmd_range->res->start >> PAGE_SHIFT) +
++ (vma->vm_pgoff - privcmd_range->pgoff); */
+ int error;
+
+ error = HYPERVISOR_zap_physmap(gpfn, 0);
-+ if (error) {
-+ BUG();//XXX
-+ }
++ if (error)
++ BUG(); /* XXX */
+ entry->gpfn = INVALID_GPFN;
+}
+
+static void
-+xen_ia64_privcmd_entry_open(struct xen_ia64_privcmd_range* privcmd_range,
++xen_ia64_privcmd_entry_open(struct xen_ia64_privcmd_range *privcmd_range,
+ int i)
+{
-+ struct xen_ia64_privcmd_entry* entry = &privcmd_range->entries[i];
-+ if (entry->gpfn != INVALID_GPFN) {
++ struct xen_ia64_privcmd_entry *entry = &privcmd_range->entries[i];
++ if (entry->gpfn != INVALID_GPFN)
+ atomic_inc(&entry->map_count);
-+ } else {
++ else
+ BUG_ON(atomic_read(&entry->map_count) != 0);
-+ }
+}
+
+static void
-+xen_ia64_privcmd_entry_close(struct xen_ia64_privcmd_range* privcmd_range,
++xen_ia64_privcmd_entry_close(struct xen_ia64_privcmd_range *privcmd_range,
+ int i)
+{
-+ struct xen_ia64_privcmd_entry* entry = &privcmd_range->entries[i];
++ struct xen_ia64_privcmd_entry *entry = &privcmd_range->entries[i];
+ if (entry->gpfn != INVALID_GPFN &&
-+ atomic_dec_and_test(&entry->map_count)) {
++ atomic_dec_and_test(&entry->map_count))
+ xen_ia64_privcmd_entry_munmap(privcmd_range, i);
-+ }
+}
+
-+static void xen_ia64_privcmd_vma_open(struct vm_area_struct* vma);
-+static void xen_ia64_privcmd_vma_close(struct vm_area_struct* vma);
++static void xen_ia64_privcmd_vma_open(struct vm_area_struct *vma);
++static void xen_ia64_privcmd_vma_close(struct vm_area_struct *vma);
++
++static struct page *
++xen_ia64_privcmd_vma_nopage(struct vm_area_struct *vma,
++ unsigned long address,
++ int *type)
++{
++ return NOPAGE_SIGBUS;
++}
+
+struct vm_operations_struct xen_ia64_privcmd_vm_ops = {
-+ .open = &xen_ia64_privcmd_vma_open,
-+ .close = &xen_ia64_privcmd_vma_close,
++ .open = xen_ia64_privcmd_vma_open,
++ .close = xen_ia64_privcmd_vma_close,
++ .nopage = xen_ia64_privcmd_vma_nopage
+};
+
+static void
-+__xen_ia64_privcmd_vma_open(struct vm_area_struct* vma,
-+ struct xen_ia64_privcmd_vma* privcmd_vma,
-+ struct xen_ia64_privcmd_range* privcmd_range)
++__xen_ia64_privcmd_vma_open(struct vm_area_struct *vma,
++ struct xen_ia64_privcmd_vma *privcmd_vma,
++ struct xen_ia64_privcmd_range *privcmd_range)
+{
+ unsigned long entry_offset = vma->vm_pgoff - privcmd_range->pgoff;
-+ unsigned long num_entries = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
++ unsigned long num_entries =
++ (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
+ unsigned long i;
+
+ BUG_ON(entry_offset < 0);
@@ -25263,41 +28022,43 @@
+ privcmd_vma->num_entries = num_entries;
+ privcmd_vma->entries = &privcmd_range->entries[entry_offset];
+ vma->vm_private_data = privcmd_vma;
-+ for (i = 0; i < privcmd_vma->num_entries; i++) {
++ for (i = 0; i < privcmd_vma->num_entries; i++)
+ xen_ia64_privcmd_entry_open(privcmd_range, entry_offset + i);
-+ }
+
+ vma->vm_private_data = privcmd_vma;
+ vma->vm_ops = &xen_ia64_privcmd_vm_ops;
+}
+
+static void
-+xen_ia64_privcmd_vma_open(struct vm_area_struct* vma)
++xen_ia64_privcmd_vma_open(struct vm_area_struct *vma)
+{
-+ struct xen_ia64_privcmd_vma* old_privcmd_vma = (struct xen_ia64_privcmd_vma*)vma->vm_private_data;
-+ struct xen_ia64_privcmd_vma* privcmd_vma = (struct xen_ia64_privcmd_vma*)vma->vm_private_data;
-+ struct xen_ia64_privcmd_range* privcmd_range = privcmd_vma->range;
++ struct xen_ia64_privcmd_vma *old_privcmd_vma =
++ (struct xen_ia64_privcmd_vma*)vma->vm_private_data;
++ struct xen_ia64_privcmd_vma *privcmd_vma =
++ (struct xen_ia64_privcmd_vma*)vma->vm_private_data;
++ struct xen_ia64_privcmd_range *privcmd_range = privcmd_vma->range;
+
+ atomic_inc(&privcmd_range->ref_count);
-+ // vm_op->open() can't fail.
++ /* vm_op->open() can't fail. */
+ privcmd_vma = kmalloc(sizeof(*privcmd_vma), GFP_KERNEL | __GFP_NOFAIL);
-+ // copy original value if necessary
++ /* copy original value if necessary */
+ privcmd_vma->is_privcmd_mmapped = old_privcmd_vma->is_privcmd_mmapped;
+
+ __xen_ia64_privcmd_vma_open(vma, privcmd_vma, privcmd_range);
+}
+
+static void
-+xen_ia64_privcmd_vma_close(struct vm_area_struct* vma)
++xen_ia64_privcmd_vma_close(struct vm_area_struct *vma)
+{
-+ struct xen_ia64_privcmd_vma* privcmd_vma =
++ struct xen_ia64_privcmd_vma *privcmd_vma =
+ (struct xen_ia64_privcmd_vma*)vma->vm_private_data;
-+ struct xen_ia64_privcmd_range* privcmd_range = privcmd_vma->range;
++ struct xen_ia64_privcmd_range *privcmd_range = privcmd_vma->range;
+ unsigned long entry_offset = vma->vm_pgoff - privcmd_range->pgoff;
+ unsigned long i;
+
+ for (i = 0; i < privcmd_vma->num_entries; i++) {
+ xen_ia64_privcmd_entry_close(privcmd_range, entry_offset + i);
++ cond_resched();
+ }
+ vma->vm_private_data = NULL;
+ kfree(privcmd_vma);
@@ -25305,12 +28066,15 @@
+ if (atomic_dec_and_test(&privcmd_range->ref_count)) {
+#if 1
+ for (i = 0; i < privcmd_range->num_entries; i++) {
-+ struct xen_ia64_privcmd_entry* entry =
++ struct xen_ia64_privcmd_entry *entry =
+ &privcmd_range->entries[i];
+ BUG_ON(atomic_read(&entry->map_count) != 0);
+ BUG_ON(entry->gpfn != INVALID_GPFN);
+ }
+#endif
++ if (privcmd_range->callback)
++ (*privcmd_range->callback)(privcmd_range,
++ privcmd_range->private);
+ release_resource(privcmd_range->res);
+ kfree(privcmd_range->res);
+ vfree(privcmd_range);
@@ -25320,7 +28084,7 @@
+int
+privcmd_enforce_singleshot_mapping(struct vm_area_struct *vma)
+{
-+ struct xen_ia64_privcmd_vma* privcmd_vma =
++ struct xen_ia64_privcmd_vma *privcmd_vma =
+ (struct xen_ia64_privcmd_vma *)vma->vm_private_data;
+ return (xchg(&privcmd_vma->is_privcmd_mmapped, 1) == 0);
+}
@@ -25331,9 +28095,9 @@
+ int error;
+ unsigned long size = vma->vm_end - vma->vm_start;
+ unsigned long num_entries = size >> PAGE_SHIFT;
-+ struct xen_ia64_privcmd_range* privcmd_range = NULL;
-+ struct xen_ia64_privcmd_vma* privcmd_vma = NULL;
-+ struct resource* res = NULL;
++ struct xen_ia64_privcmd_range *privcmd_range = NULL;
++ struct xen_ia64_privcmd_vma *privcmd_vma = NULL;
++ struct resource *res = NULL;
+ unsigned long i;
+ BUG_ON(!is_running_on_xen());
+
@@ -25343,37 +28107,34 @@
+ privcmd_range =
+ vmalloc(sizeof(*privcmd_range) +
+ sizeof(privcmd_range->entries[0]) * num_entries);
-+ if (privcmd_range == NULL) {
++ if (privcmd_range == NULL)
+ goto out_enomem0;
-+ }
+ privcmd_vma = kmalloc(sizeof(*privcmd_vma), GFP_KERNEL);
-+ if (privcmd_vma == NULL) {
++ if (privcmd_vma == NULL)
+ goto out_enomem1;
-+ }
+ privcmd_vma->is_privcmd_mmapped = 0;
+
+ res = kzalloc(sizeof(*res), GFP_KERNEL);
-+ if (res == NULL) {
++ if (res == NULL)
+ goto out_enomem1;
-+ }
+ res->name = "Xen privcmd mmap";
+ error = allocate_resource(&iomem_resource, res, size,
+ privcmd_resource_min, privcmd_resource_max,
+ privcmd_resource_align, NULL, NULL);
-+ if (error) {
++ if (error)
+ goto out_enomem1;
-+ }
+ privcmd_range->res = res;
+
+ /* DONTCOPY is essential for Xen as copy_page_range is broken. */
-+ vma->vm_flags |= VM_RESERVED | VM_IO | VM_DONTCOPY | VM_PFNMAP;
++ vma->vm_flags |= VM_RESERVED | VM_IO | VM_DONTCOPY;
+
+ atomic_set(&privcmd_range->ref_count, 1);
+ privcmd_range->pgoff = vma->vm_pgoff;
+ privcmd_range->num_entries = num_entries;
-+ for (i = 0; i < privcmd_range->num_entries; i++) {
++ privcmd_range->private = NULL;
++ privcmd_range->callback = NULL;
++ for (i = 0; i < privcmd_range->num_entries; i++)
+ xen_ia64_privcmd_init_entry(&privcmd_range->entries[i]);
-+ }
+
+ __xen_ia64_privcmd_vma_open(vma, privcmd_vma, privcmd_range);
+ return 0;
@@ -25388,15 +28149,15 @@
+
+int
+direct_remap_pfn_range(struct vm_area_struct *vma,
-+ unsigned long address, // process virtual address
-+ unsigned long gmfn, // gmfn, gmfn + 1, ... gmfn + size/PAGE_SIZE
++ unsigned long address, /* process virtual address */
++ unsigned long gmfn, /* gmfn, gmfn + 1, ... gmfn + size/PAGE_SIZE */
+ unsigned long size,
+ pgprot_t prot,
-+ domid_t domid) // target domain
++ domid_t domid) /* target domain */
+{
-+ struct xen_ia64_privcmd_vma* privcmd_vma =
++ struct xen_ia64_privcmd_vma *privcmd_vma =
+ (struct xen_ia64_privcmd_vma*)vma->vm_private_data;
-+ struct xen_ia64_privcmd_range* privcmd_range = privcmd_vma->range;
++ struct xen_ia64_privcmd_range *privcmd_range = privcmd_vma->range;
+ unsigned long entry_offset = vma->vm_pgoff - privcmd_range->pgoff;
+
+ unsigned long i;
@@ -25405,28 +28166,27 @@
+ BUG_ON(!is_running_on_xen());
+
+#if 0
-+ if (prot != vm->vm_page_prot) {
++ if (prot != vm->vm_page_prot)
+ return -EINVAL;
-+ }
+#endif
+
+ i = (address - vma->vm_start) >> PAGE_SHIFT;
+ for (offset = 0; offset < size; offset += PAGE_SIZE) {
+ error = xen_ia64_privcmd_entry_mmap(vma, (address + offset) & PAGE_MASK, privcmd_range, entry_offset + i, gmfn, prot, domid);
-+ if (error != 0) {
++ if (error != 0)
+ break;
-+ }
+
+ i++;
+ gmfn++;
-+ }
++ }
+
+ return error;
+}
+
+
-+///////////////////////////////////////////////////////////////////////////
-+// expose p2m table
++/**************************************************************************
++ * expose p2m table
++ */
+#ifdef CONFIG_XEN_IA64_EXPOSE_P2M
+#include <linux/cpu.h>
+#include <asm/uaccess.h>
@@ -25444,12 +28204,13 @@
+};
+static unsigned long p2m_assign_start_pfn __read_mostly;
+static unsigned long p2m_assign_end_pfn __read_mostly;
-+static unsigned long p2m_expose_size; // this is referenced only when resume.
-+ // so __read_mostly doesn't make sense.
-+volatile const pte_t* p2m_pte __read_mostly;
++static unsigned long p2m_expose_size; /* this is referenced only when resume.
++ * so __read_mostly doesn't make sense.
++ */
++volatile const pte_t *p2m_pte __read_mostly;
+
-+#define GRNULE_PFN PTRS_PER_PTE
-+static unsigned long p2m_granule_pfn __read_mostly = GRNULE_PFN;
++#define GRANULE_PFN PTRS_PER_PTE
++static unsigned long p2m_granule_pfn __read_mostly = GRANULE_PFN;
+
+#define ROUNDDOWN(x, y) ((x) & ~((y) - 1))
+#define ROUNDUP(x, y) (((x) + (y) - 1) & ~((y) - 1))
@@ -25459,13 +28220,13 @@
+static int xen_ia64_p2m_expose __read_mostly = 1;
+module_param(xen_ia64_p2m_expose, int, 0);
+MODULE_PARM_DESC(xen_ia64_p2m_expose,
-+ "enable/disable xen/ia64 p2m exposure optimization\n");
++ "enable/disable xen/ia64 p2m exposure optimization\n");
+
+#ifdef CONFIG_XEN_IA64_EXPOSE_P2M_USE_DTR
+static int xen_ia64_p2m_expose_use_dtr __read_mostly = 1;
+module_param(xen_ia64_p2m_expose_use_dtr, int, 0);
+MODULE_PARM_DESC(xen_ia64_p2m_expose_use_dtr,
-+ "use/unuse dtr to map exposed p2m table\n");
++ "use/unuse dtr to map exposed p2m table\n");
+
+static const int p2m_page_shifts[] = {
+ _PAGE_SIZE_4K,
@@ -25487,21 +28248,21 @@
+};
+static struct p2m_itr_arg p2m_itr_arg __read_mostly;
+
-+// This should be in asm-ia64/kregs.h
++/* This should be in asm-ia64/kregs.h */
+#define IA64_TR_P2M_TABLE 3
+
+static void
-+p2m_itr(void* info)
++p2m_itr(void *info)
+{
-+ struct p2m_itr_arg* arg = (struct p2m_itr_arg*)info;
++ struct p2m_itr_arg *arg = (struct p2m_itr_arg*)info;
+ ia64_itr(0x2, IA64_TR_P2M_TABLE,
-+ arg->vaddr, arg->pteval, arg->log_page_size);
++ arg->vaddr, arg->pteval, arg->log_page_size);
+ ia64_srlz_d();
+}
+
+static int
+p2m_expose_dtr_call(struct notifier_block *self,
-+ unsigned long event, void* ptr)
++ unsigned long event, void *ptr)
+{
+ unsigned int cpu = (unsigned int)(long)ptr;
+ if (event != CPU_ONLINE)
@@ -25525,6 +28286,12 @@
+};
+#endif
+
++static inline unsigned long
++p2m_table_size(unsigned long num_pfn)
++{
++ return ((num_pfn + PTRS_PER_PTE - 1) / PTRS_PER_PTE) << PAGE_SHIFT;
++}
++
+static int
+p2m_expose_init(void)
+{
@@ -25534,7 +28301,6 @@
+ int error = 0;
+#ifdef CONFIG_XEN_IA64_EXPOSE_P2M_USE_DTR
+ int i;
-+ unsigned long page_size;
+ unsigned long log_page_size = 0;
+#endif
+
@@ -25563,8 +28329,9 @@
+
+#ifdef CONFIG_XEN_IA64_EXPOSE_P2M_USE_DTR
+ if (xen_ia64_p2m_expose_use_dtr) {
++ unsigned long page_size = 0;
+ unsigned long granule_pfn = 0;
-+ p2m_size = p2m_max_low_pfn - p2m_min_low_pfn;
++ p2m_size = p2m_table_size(p2m_max_low_pfn - p2m_min_low_pfn);
+ for (i = 0;
+ i < sizeof(p2m_page_shifts)/sizeof(p2m_page_shifts[0]);
+ i++) {
@@ -25574,15 +28341,16 @@
+ continue;
+
+ granule_pfn = max(page_size >> PAGE_SHIFT,
-+ p2m_granule_pfn);
++ p2m_granule_pfn);
+ p2m_convert_min_pfn = ROUNDDOWN(p2m_min_low_pfn,
-+ granule_pfn);
++ granule_pfn);
+ p2m_convert_max_pfn = ROUNDUP(p2m_max_low_pfn,
-+ granule_pfn);
++ granule_pfn);
+ num_pfn = p2m_convert_max_pfn - p2m_convert_min_pfn;
+ p2m_expose_size = num_pfn << PAGE_SHIFT;
-+ p2m_size = num_pfn / PTRS_PER_PTE;
-+ p2m_size = ROUNDUP(p2m_size, granule_pfn << PAGE_SHIFT);
++ p2m_size = p2m_table_size(num_pfn);
++ p2m_size = ROUNDUP(p2m_size,
++ granule_pfn << PAGE_SHIFT);
+ if (p2m_size == page_size)
+ break;
+ }
@@ -25597,24 +28365,25 @@
+ {
+ BUG_ON(p2m_granule_pfn & (p2m_granule_pfn - 1));
+ p2m_convert_min_pfn = ROUNDDOWN(p2m_min_low_pfn,
-+ p2m_granule_pfn);
-+ p2m_convert_max_pfn = ROUNDUP(p2m_max_low_pfn, p2m_granule_pfn);
++ p2m_granule_pfn);
++ p2m_convert_max_pfn = ROUNDUP(p2m_max_low_pfn,
++ p2m_granule_pfn);
+ num_pfn = p2m_convert_max_pfn - p2m_convert_min_pfn;
+ p2m_expose_size = num_pfn << PAGE_SHIFT;
-+ p2m_size = num_pfn / PTRS_PER_PTE;
++ p2m_size = p2m_table_size(num_pfn);
+ p2m_size = ROUNDUP(p2m_size, p2m_granule_pfn << PAGE_SHIFT);
+ align = max(privcmd_resource_align,
-+ p2m_granule_pfn << PAGE_SHIFT);
++ p2m_granule_pfn << PAGE_SHIFT);
+ }
+
-+ // use privcmd region
++ /* use privcmd region */
+ error = allocate_resource(&iomem_resource, &p2m_resource, p2m_size,
-+ privcmd_resource_min, privcmd_resource_max,
-+ align, NULL, NULL);
++ privcmd_resource_min, privcmd_resource_max,
++ align, NULL, NULL);
+ if (error) {
+ printk(KERN_ERR P2M_PREFIX
+ "can't allocate region for p2m exposure "
-+ "[0x%016lx, 0x%016lx) 0x%016lx\n",
++ "[0x%016lx, 0x%016lx] 0x%016lx\n",
+ p2m_convert_min_pfn, p2m_convert_max_pfn, p2m_size);
+ goto out;
+ }
@@ -25623,8 +28392,8 @@
+ p2m_assign_end_pfn = p2m_resource.end >> PAGE_SHIFT;
+
+ error = HYPERVISOR_expose_p2m(p2m_convert_min_pfn,
-+ p2m_assign_start_pfn,
-+ p2m_expose_size, p2m_granule_pfn);
++ p2m_assign_start_pfn,
++ p2m_expose_size, p2m_granule_pfn);
+ if (error) {
+ printk(KERN_ERR P2M_PREFIX "failed expose p2m hypercall %d\n",
+ error);
@@ -25639,9 +28408,9 @@
+#ifdef CONFIG_XEN_IA64_EXPOSE_P2M_USE_DTR
+ if (xen_ia64_p2m_expose_use_dtr) {
+ p2m_itr_arg.vaddr = (unsigned long)__va(p2m_assign_start_pfn
-+ << PAGE_SHIFT);
++ << PAGE_SHIFT);
+ p2m_itr_arg.pteval = pte_val(pfn_pte(p2m_assign_start_pfn,
-+ PAGE_KERNEL));
++ PAGE_KERNEL));
+ p2m_itr_arg.log_page_size = log_page_size;
+ smp_mb();
+ smp_call_function(&p2m_itr, &p2m_itr_arg, 1, 1);
@@ -25652,10 +28421,10 @@
+ p2m_initialized = 1;
+ printk(P2M_PREFIX "assign p2m table of [0x%016lx, 0x%016lx)\n",
+ p2m_convert_min_pfn << PAGE_SHIFT,
-+ p2m_convert_max_pfn << PAGE_SHIFT);
++ (p2m_convert_max_pfn << PAGE_SHIFT) + PAGE_SIZE);
+ printk(P2M_PREFIX "to [0x%016lx, 0x%016lx) (%ld KBytes)\n",
+ p2m_assign_start_pfn << PAGE_SHIFT,
-+ p2m_assign_end_pfn << PAGE_SHIFT,
++ (p2m_assign_end_pfn << PAGE_SHIFT) + PAGE_SIZE,
+ p2m_size / 1024);
+out:
+ unlock_cpu_hotplug();
@@ -25689,8 +28458,8 @@
+ * interrupts are masked when resume.
+ */
+ error = HYPERVISOR_expose_p2m(p2m_convert_min_pfn,
-+ p2m_assign_start_pfn,
-+ p2m_expose_size, p2m_granule_pfn);
++ p2m_assign_start_pfn,
++ p2m_expose_size, p2m_granule_pfn);
+ if (error) {
+ printk(KERN_ERR P2M_PREFIX "failed expose p2m hypercall %d\n",
+ error);
@@ -25717,11 +28486,11 @@
+ }
+}
+
-+//XXX inlinize?
++/* XXX inlinize? */
+unsigned long
+p2m_phystomach(unsigned long gpfn)
+{
-+ volatile const pte_t* pte;
++ volatile const pte_t *pte;
+ unsigned long mfn;
+ unsigned long pteval;
+
@@ -25733,8 +28502,8 @@
+
+ mfn = INVALID_MFN;
+ if (likely(__get_user(pteval, (unsigned long __user *)pte) == 0 &&
-+ pte_present(__pte(pteval)) &&
-+ pte_pfn(__pte(pteval)) != (INVALID_MFN >> PAGE_SHIFT)))
++ pte_present(__pte(pteval)) &&
++ pte_pfn(__pte(pteval)) != (INVALID_MFN >> PAGE_SHIFT)))
+ mfn = (pteval & _PFN_MASK) >> PAGE_SHIFT;
+
+ return mfn;
@@ -25747,26 +28516,230 @@
+EXPORT_SYMBOL_GPL(p2m_convert_max_pfn);
+EXPORT_SYMBOL_GPL(p2m_pte);
+EXPORT_SYMBOL_GPL(p2m_phystomach);
-+#endif
+
-+///////////////////////////////////////////////////////////////////////////
-+// for xenoprof
++/**************************************************************************
++ * foreign domain p2m mapping
++ */
++#include <asm/xen/xencomm.h>
++#include <xen/public/privcmd.h>
++
++struct foreign_p2m_private {
++ unsigned long gpfn;
++ domid_t domid;
++};
++
++static void
++xen_foreign_p2m_unexpose(struct xen_ia64_privcmd_range *privcmd_range,
++ void *arg)
++{
++ struct foreign_p2m_private *private = (struct foreign_p2m_private*)arg;
++ int ret;
++
++ privcmd_range->private = NULL;
++ privcmd_range->callback = NULL;
++
++ ret = HYPERVISOR_unexpose_foreign_p2m(private->gpfn, private->domid);
++ if (ret)
++ printk(KERN_WARNING
++ "unexpose_foreign_p2m hypercall failed.\n");
++ kfree(private);
++}
++
++int
++xen_foreign_p2m_expose(privcmd_hypercall_t *hypercall)
++{
++ /*
++ * hypercall->
++ * arg0: cmd = IA64_DOM0VP_expose_foreign_p2m
++ * arg1: va
++ * arg2: domid
++ * arg3: __user* memmap_info
++ * arg4: flags
++ */
++
++ int ret = 0;
++ struct mm_struct *mm = current->mm;
++
++ unsigned long vaddr = hypercall->arg[1];
++ domid_t domid = hypercall->arg[2];
++ struct xen_ia64_memmap_info __user *u_memmap_info =
++ (struct xen_ia64_memmap_info __user *)hypercall->arg[3];
++
++ struct xen_ia64_memmap_info memmap_info;
++ size_t memmap_size;
++ struct xen_ia64_memmap_info *k_memmap_info = NULL;
++ unsigned long max_gpfn;
++ unsigned long p2m_size;
++ struct resource *res;
++ unsigned long gpfn;
++
++ struct vm_area_struct *vma;
++ void *p;
++ unsigned long prev_src_gpfn_end;
++
++ struct xen_ia64_privcmd_vma *privcmd_vma;
++ struct xen_ia64_privcmd_range *privcmd_range;
++ struct foreign_p2m_private *private = NULL;
++
++ BUG_ON(hypercall->arg[0] != IA64_DOM0VP_expose_foreign_p2m);
++
++ private = kmalloc(sizeof(*private), GFP_KERNEL);
++ if (private == NULL)
++ goto kfree_out;
++
++ if (copy_from_user(&memmap_info, u_memmap_info, sizeof(memmap_info)))
++ return -EFAULT;
++ /* memmap_info integrity check */
++ if (memmap_info.efi_memdesc_size < sizeof(efi_memory_desc_t) ||
++ memmap_info.efi_memmap_size < memmap_info.efi_memdesc_size ||
++ (memmap_info.efi_memmap_size % memmap_info.efi_memdesc_size)
++ != 0) {
++ ret = -EINVAL;
++ goto kfree_out;
++ }
++
++ memmap_size = sizeof(*k_memmap_info) + memmap_info.efi_memmap_size;
++ k_memmap_info = kmalloc(memmap_size, GFP_KERNEL);
++ if (k_memmap_info == NULL)
++ return -ENOMEM;
++ if (copy_from_user(k_memmap_info, u_memmap_info, memmap_size)) {
++ ret = -EFAULT;
++ goto kfree_out;
++ }
++ /* k_memmap_info integrity check is done by the expose foreng p2m
++ hypercall */
++
++ max_gpfn = HYPERVISOR_memory_op(XENMEM_maximum_gpfn, &domid);
++ if (max_gpfn < 0) {
++ ret = max_gpfn;
++ goto kfree_out;
++ }
++ p2m_size = p2m_table_size(max_gpfn + 1);
++
++ down_write(&mm->mmap_sem);
++
++ vma = find_vma(mm, vaddr);
++ if (vma == NULL || vma->vm_ops != &xen_ia64_privcmd_vm_ops ||
++ vaddr != vma->vm_start ||
++ (vma->vm_flags & VM_WRITE) || (vma->vm_flags & VM_EXEC) ||
++ !privcmd_enforce_singleshot_mapping(vma))
++ goto mmap_out;
++
++ privcmd_vma = (struct xen_ia64_privcmd_vma*)vma->vm_private_data;
++ res = privcmd_vma->range->res;
++ if (p2m_size > (res->end - res->start + 1) ||
++ p2m_size > vma->vm_end - vma->vm_start) {
++ ret = -EINVAL;
++ goto mmap_out;
++ }
++
++ gpfn = res->start >> PAGE_SHIFT;
++ /*
++ * arg0: dest_gpfn
++ * arg1: domid
++ * arg2: XEN_GUEST_HANDLE(char) buffer: memmap_info
++ * arg3: flags
++ * The hypercall checks its intergirty/simplfies it and
++ * copy it back for us.
++ */
++ ret = xencomm_arch_expose_foreign_p2m(gpfn, domid,
++ xencomm_map_no_alloc(k_memmap_info, memmap_size),
++ hypercall->arg[4]);
++ if (ret)
++ goto mmap_out;
++
++ privcmd_range = (struct xen_ia64_privcmd_range*)privcmd_vma->range;
++ prev_src_gpfn_end = 0;
++ for (p = k_memmap_info->memdesc;
++ p < (void*)&k_memmap_info->memdesc[0] +
++ k_memmap_info->efi_memmap_size;
++ p += k_memmap_info->efi_memdesc_size) {
++ efi_memory_desc_t* md = p;
++ unsigned long src_gpfn = md->phys_addr >> PAGE_SHIFT;
++ unsigned long src_gpfn_end =
++ (md->phys_addr + (md->num_pages << EFI_PAGE_SHIFT)) >>
++ PAGE_SHIFT;
++ unsigned long num_src_gpfn;
++ unsigned long gpfn_offset;
++ unsigned long size;
++ unsigned int i;
++
++ if (src_gpfn <= prev_src_gpfn_end)
++ src_gpfn = prev_src_gpfn_end + 1;
++ if (src_gpfn_end <= prev_src_gpfn_end)
++ continue;
++
++ src_gpfn &= ~(PTRS_PER_PTE - 1);
++ src_gpfn_end = (src_gpfn_end + PTRS_PER_PTE - 1) &
++ ~(PTRS_PER_PTE - 1);
++ num_src_gpfn = src_gpfn_end - src_gpfn;
++ gpfn_offset = src_gpfn / PTRS_PER_PTE;
++ size = p2m_table_size(num_src_gpfn);
++
++ prev_src_gpfn_end = src_gpfn_end;
++ ret = remap_pfn_range(vma,
++ vaddr + (gpfn_offset << PAGE_SHIFT),
++ gpfn + gpfn_offset, size,
++ vma->vm_page_prot);
++ if (ret) {
++ for (i = 0; i < gpfn + gpfn_offset; i++) {
++ struct xen_ia64_privcmd_entry *entry =
++ &privcmd_range->entries[i];
++ BUG_ON(atomic_read(&entry->map_count) != 1 &&
++ atomic_read(&entry->map_count) != 0);
++ atomic_set(&entry->map_count, 0);
++ entry->gpfn = INVALID_GPFN;
++ }
++ (void)HYPERVISOR_unexpose_foreign_p2m(gpfn, domid);
++ goto mmap_out;
++ }
++
++ for (i = gpfn_offset;
++ i < gpfn_offset + (size >> PAGE_SHIFT);
++ i++) {
++ struct xen_ia64_privcmd_entry *entry =
++ &privcmd_range->entries[i];
++ BUG_ON(atomic_read(&entry->map_count) != 0);
++ BUG_ON(entry->gpfn != INVALID_GPFN);
++ atomic_inc(&entry->map_count);
++ entry->gpfn = gpfn + i;
++ }
++ }
++
++ private->gpfn = gpfn;
++ private->domid = domid;
++
++ privcmd_range->callback = &xen_foreign_p2m_unexpose;
++ privcmd_range->private = private;
++
++mmap_out:
++ up_write(&mm->mmap_sem);
++kfree_out:
++ kfree(k_memmap_info);
++ if (ret != 0)
++ kfree(private);
++ return ret;
++}
++#endif
+
++/**************************************************************************
++ * for xenoprof
++ */
+struct resource*
+xen_ia64_allocate_resource(unsigned long size)
+{
-+ struct resource* res;
++ struct resource *res;
+ int error;
+
-+ res = kmalloc(sizeof(*res), GFP_KERNEL);
++ res = kzalloc(sizeof(*res), GFP_KERNEL);
+ if (res == NULL)
+ return ERR_PTR(-ENOMEM);
+
+ res->name = "Xen";
+ res->flags = IORESOURCE_MEM;
+ error = allocate_resource(&iomem_resource, res, PAGE_ALIGN(size),
-+ privcmd_resource_min, privcmd_resource_max,
-+ IA64_GRANULE_SIZE, NULL, NULL);
++ privcmd_resource_min, privcmd_resource_max,
++ IA64_GRANULE_SIZE, NULL, NULL);
+ if (error) {
+ kfree(res);
+ return ERR_PTR(error);
@@ -25776,7 +28749,7 @@
+EXPORT_SYMBOL_GPL(xen_ia64_allocate_resource);
+
+void
-+xen_ia64_release_resource(struct resource* res)
++xen_ia64_release_resource(struct resource *res)
+{
+ release_resource(res);
+ kfree(res);
@@ -25784,7 +28757,7 @@
+EXPORT_SYMBOL_GPL(xen_ia64_release_resource);
+
+void
-+xen_ia64_unmap_resource(struct resource* res)
++xen_ia64_unmap_resource(struct resource *res)
+{
+ unsigned long gpfn = res->start >> PAGE_SHIFT;
+ unsigned long nr_pages = (res->end - res->start) >> PAGE_SHIFT;
@@ -25801,8 +28774,25 @@
+}
+EXPORT_SYMBOL_GPL(xen_ia64_unmap_resource);
+
-+///////////////////////////////////////////////////////////////////////////
-+// suspend/resume
++/**************************************************************************
++ * opt feature
++ */
++void
++xen_ia64_enable_opt_feature(void)
++{
++ /* Enable region 7 identity map optimizations in Xen */
++ struct xen_ia64_opt_feature optf;
++
++ optf.cmd = XEN_IA64_OPTF_IDENT_MAP_REG7;
++ optf.on = XEN_IA64_OPTF_ON;
++ optf.pgprot = pgprot_val(PAGE_KERNEL);
++ optf.key = 0; /* No key on linux. */
++ HYPERVISOR_opt_feature(&optf);
++}
++
++/**************************************************************************
++ * suspend/resume
++ */
+void
+xen_post_suspend(int suspend_cancelled)
+{
@@ -25810,1060 +28800,1068 @@
+ return;
+
+ p2m_expose_resume();
++ xen_ia64_enable_opt_feature();
+ /* add more if necessary */
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/Makefile tmp-linux-2.6-xen.patch/arch/ia64/xen/Makefile
---- pristine-linux-2.6.18.2/arch/ia64/xen/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,9 @@
-+#
-+# Makefile for Xen components
-+#
-+
-+obj-y := hypercall.o xenivt.o xenentry.o xensetup.o xenpal.o xenhpski.o \
-+ hypervisor.o pci-dma-xen.o util.o xencomm.o xcom_hcall.o \
-+ xcom_mini.o xcom_privcmd.o mem.o
-+
-+pci-dma-xen-y := ../../i386/kernel/pci-dma-xen.o
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/mem.c tmp-linux-2.6-xen.patch/arch/ia64/xen/mem.c
---- pristine-linux-2.6.18.2/arch/ia64/xen/mem.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/mem.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,75 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/machvec.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/machvec.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,4 @@
++#define MACHVEC_PLATFORM_NAME xen
++#define MACHVEC_PLATFORM_HEADER <asm/machvec_xen.h>
++#include <asm/machvec_init.h>
++
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/swiotlb.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/swiotlb.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,906 @@
+/*
-+ * Originally from linux/drivers/char/mem.c
++ * Dynamic DMA mapping support.
+ *
-+ * Copyright (C) 1991, 1992 Linus Torvalds
++ * This implementation is for IA-64 and EM64T platforms that do not support
++ * I/O TLBs (aka DMA address translation hardware).
++ * Copyright (C) 2000 Asit Mallick <Asit.K.Mallick at intel.com>
++ * Copyright (C) 2000 Goutham Rao <goutham.rao at intel.com>
++ * Copyright (C) 2000, 2003 Hewlett-Packard Co
++ * David Mosberger-Tang <davidm at hpl.hp.com>
+ *
-+ * Added devfs support.
-+ * Jan-11-1998, C. Scott Ananian <cananian at alumni.princeton.edu>
-+ * Shared /dev/zero mmaping support, Feb 2000, Kanoj Sarcar <kanoj at sgi.com>
-+ */
-+/*
-+ * taken from
-+ * linux/drivers/char/mem.c and linux-2.6-xen-sparse/drivers/xen/char/mem.c.
-+ * adjusted for IA64 and made transparent.
-+ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
-+ * VA Linux Systems Japan K.K.
++ * 03/05/07 davidm Switch from PCI-DMA to generic device DMA API.
++ * 00/12/13 davidm Rename to swiotlb.c and add mark_clean() to avoid
++ * unnecessary i-cache flushing.
++ * 04/07/.. ak Better overflow handling. Assorted fixes.
++ * 05/09/10 linville Add support for syncing ranges, support syncing for
++ * DMA_BIDIRECTIONAL mappings, miscellaneous cleanup.
+ */
+
++#include <linux/cache.h>
++#include <linux/dma-mapping.h>
+#include <linux/mm.h>
-+#include <linux/efi.h>
++#include <linux/module.h>
++#include <linux/spinlock.h>
++#include <linux/string.h>
++#include <linux/types.h>
++#include <linux/ctype.h>
+
-+/*
-+ * Architectures vary in how they handle caching for addresses
-+ * outside of main memory.
-+ *
-+ */
-+static inline int uncached_access(struct file *file, unsigned long addr)
-+{
-+ /*
-+ * On ia64, we ignore O_SYNC because we cannot tolerate memory attribute aliases.
-+ */
-+ return !(efi_mem_attributes(addr) & EFI_MEMORY_WB);
-+}
++#include <asm/io.h>
++#include <asm/dma.h>
++#include <asm/scatterlist.h>
+
-+int xen_mmap_mem(struct file * file, struct vm_area_struct * vma)
-+{
-+ unsigned long addr = vma->vm_pgoff << PAGE_SHIFT;
-+ size_t size = vma->vm_end - vma->vm_start;
++#include <linux/init.h>
++#include <linux/bootmem.h>
+
-+
-+#if 0
-+ /*
-+ *XXX FIXME: linux-2.6.16.29, linux-2.6.17
-+ * valid_mmap_phys_addr_range() in linux/arch/ia64/kernel/efi.c
-+ * fails checks.
-+ * linux-2.6.18.1's returns always 1.
-+ * Its comments says
-+ *
-+ * MMIO regions are often missing from the EFI memory map.
-+ * We must allow mmap of them for programs like X, so we
-+ * currently can't do any useful validation.
-+ */
-+ if (!valid_mmap_phys_addr_range(addr, &size))
-+ return -EINVAL;
-+ if (size < vma->vm_end - vma->vm_start)
-+ return -EINVAL;
++#ifdef CONFIG_XEN
++#include <xen/gnttab.h>
++#include <asm/gnttab_dma.h>
++/*
++ * What DMA mask should Xen use to remap the bounce buffer pool? Most
++ * reports seem to indicate 30 bits is sufficient, except maybe for old
++ * sound cards that we probably don't care about anyway. If we need to,
++ * we could put in some smarts to try to lower, but hopefully it's not
++ * necessary.
++ */
++#define DMA_BITS (30)
+#endif
+
-+ if (is_running_on_xen()) {
-+ unsigned long offset = HYPERVISOR_ioremap(addr, size);
-+ if (IS_ERR_VALUE(offset))
-+ return offset;
-+ }
++#define OFFSET(val,align) ((unsigned long) \
++ ( (val) & ( (align) - 1)))
+
-+ if (uncached_access(file, vma->vm_pgoff << PAGE_SHIFT))
-+ vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
++#define SG_ENT_VIRT_ADDRESS(sg) (page_address((sg)->page) + (sg)->offset)
++#define SG_ENT_PHYS_ADDRESS(SG) virt_to_bus(SG_ENT_VIRT_ADDRESS(SG))
+
-+ /* Remap-pfn-range will mark the range VM_IO and VM_RESERVED */
-+ if (remap_pfn_range(vma,
-+ vma->vm_start,
-+ vma->vm_pgoff,
-+ size,
-+ vma->vm_page_prot))
-+ return -EAGAIN;
-+ return 0;
-+}
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/util.c tmp-linux-2.6-xen.patch/arch/ia64/xen/util.c
---- pristine-linux-2.6.18.2/arch/ia64/xen/util.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/util.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,105 @@
-+/******************************************************************************
-+ * arch/ia64/xen/util.c
-+ * This file is the ia64 counterpart of drivers/xen/util.c
-+ *
-+ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
-+ * VA Linux Systems Japan K.K.
-+ *
-+ * This program is free software; you can redistribute it and/or modify
-+ * it under the terms of the GNU General Public License as published by
-+ * the Free Software Foundation; either version 2 of the License, or
-+ * (at your option) any later version.
-+ *
-+ * This program is distributed in the hope that it will be useful,
-+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ * GNU General Public License for more details.
-+ *
-+ * You should have received a copy of the GNU General Public License
-+ * along with this program; if not, write to the Free Software
-+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
-+ *
++/*
++ * Maximum allowable number of contiguous slabs to map,
++ * must be a power of 2. What is the appropriate value ?
++ * The complexity of {map,unmap}_single is linearly dependent on this value.
+ */
++#define IO_TLB_SEGSIZE 128
+
-+#include <linux/mm.h>
-+#include <linux/module.h>
-+#include <linux/slab.h>
-+#include <linux/vmalloc.h>
-+#include <asm/uaccess.h>
-+#include <xen/driver_util.h>
-+#include <xen/interface/memory.h>
-+#include <asm/hypercall.h>
++/*
++ * log of the size of each IO TLB slab. The number of slabs is command line
++ * controllable.
++ */
++#define IO_TLB_SHIFT 11
+
-+struct vm_struct *alloc_vm_area(unsigned long size)
-+{
-+ int order;
-+ unsigned long virt;
-+ unsigned long nr_pages;
-+ struct vm_struct* area;
-+
-+ order = get_order(size);
-+ virt = __get_free_pages(GFP_KERNEL, order);
-+ if (virt == 0) {
-+ goto err0;
-+ }
-+ nr_pages = 1 << order;
-+ scrub_pages(virt, nr_pages);
-+
-+ area = kmalloc(sizeof(*area), GFP_KERNEL);
-+ if (area == NULL) {
-+ goto err1;
-+ }
-+
-+ area->flags = VM_IOREMAP;//XXX
-+ area->addr = (void*)virt;
-+ area->size = size;
-+ area->pages = NULL; //XXX
-+ area->nr_pages = nr_pages;
-+ area->phys_addr = 0; /* xenbus_map_ring_valloc uses this field! */
++#define SLABS_PER_PAGE (1 << (PAGE_SHIFT - IO_TLB_SHIFT))
+
-+ return area;
++/*
++ * Minimum IO TLB size to bother booting with. Systems with mainly
++ * 64bit capable cards will only lightly use the swiotlb. If we can't
++ * allocate a contiguous 1MB, we're probably in trouble anyway.
++ */
++#define IO_TLB_MIN_SLABS ((1<<20) >> IO_TLB_SHIFT)
+
-+err1:
-+ free_pages(virt, order);
-+err0:
-+ return NULL;
-+
-+}
-+EXPORT_SYMBOL_GPL(alloc_vm_area);
++/*
++ * Enumeration for sync targets
++ */
++enum dma_sync_target {
++ SYNC_FOR_CPU = 0,
++ SYNC_FOR_DEVICE = 1,
++};
+
-+void free_vm_area(struct vm_struct *area)
-+{
-+ unsigned int order = get_order(area->size);
-+ unsigned long i;
-+ unsigned long phys_addr = __pa(area->addr);
++int swiotlb_force;
+
-+ // This area is used for foreign page mappping.
-+ // So underlying machine page may not be assigned.
-+ for (i = 0; i < (1 << order); i++) {
-+ unsigned long ret;
-+ unsigned long gpfn = (phys_addr >> PAGE_SHIFT) + i;
-+ struct xen_memory_reservation reservation = {
-+ .nr_extents = 1,
-+ .address_bits = 0,
-+ .extent_order = 0,
-+ .domid = DOMID_SELF
-+ };
-+ set_xen_guest_handle(reservation.extent_start, &gpfn);
-+ ret = HYPERVISOR_memory_op(XENMEM_populate_physmap,
-+ &reservation);
-+ BUG_ON(ret != 1);
-+ }
-+ free_pages((unsigned long)area->addr, order);
-+ kfree(area);
-+}
-+EXPORT_SYMBOL_GPL(free_vm_area);
++/*
++ * Used to do a quick range check in swiotlb_unmap_single and
++ * swiotlb_sync_single_*, to see if the memory was in fact allocated by this
++ * API.
++ */
++static char *io_tlb_start, *io_tlb_end;
+
+/*
-+ * Local variables:
-+ * c-file-style: "linux"
-+ * indent-tabs-mode: t
-+ * c-indent-level: 8
-+ * c-basic-offset: 8
-+ * tab-width: 8
-+ * End:
++ * The number of IO TLB blocks (in groups of 64) betweeen io_tlb_start and
++ * io_tlb_end. This is command line adjustable via setup_io_tlb_npages.
+ */
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/xcom_hcall.c tmp-linux-2.6-xen.patch/arch/ia64/xen/xcom_hcall.c
---- pristine-linux-2.6.18.2/arch/ia64/xen/xcom_hcall.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/xcom_hcall.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,383 @@
++static unsigned long io_tlb_nslabs;
++
+/*
-+ * This program is free software; you can redistribute it and/or modify
-+ * it under the terms of the GNU General Public License as published by
-+ * the Free Software Foundation; either version 2 of the License, or
-+ * (at your option) any later version.
-+ *
-+ * This program is distributed in the hope that it will be useful,
-+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ * GNU General Public License for more details.
-+ *
-+ * You should have received a copy of the GNU General Public License
-+ * along with this program; if not, write to the Free Software
-+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
-+ *
-+ * Tristan Gingold <tristan.gingold at bull.net>
++ * When the IOMMU overflows we return a fallback buffer. This sets the size.
+ */
-+#include <linux/types.h>
-+#include <linux/errno.h>
-+#include <linux/kernel.h>
-+#include <linux/gfp.h>
-+#include <linux/module.h>
-+#include <xen/interface/xen.h>
-+#include <xen/interface/platform.h>
-+#include <xen/interface/memory.h>
-+#include <xen/interface/xencomm.h>
-+#include <xen/interface/version.h>
-+#include <xen/interface/sched.h>
-+#include <xen/interface/event_channel.h>
-+#include <xen/interface/physdev.h>
-+#include <xen/interface/grant_table.h>
-+#include <xen/interface/callback.h>
-+#include <xen/interface/acm_ops.h>
-+#include <xen/interface/hvm/params.h>
-+#include <xen/interface/xenoprof.h>
-+#include <xen/interface/vcpu.h>
-+#include <asm/hypercall.h>
-+#include <asm/page.h>
-+#include <asm/uaccess.h>
-+#include <asm/xen/xencomm.h>
-+#include <asm/perfmon.h>
++static unsigned long io_tlb_overflow = 32*1024;
+
-+/* Xencomm notes:
-+ * This file defines hypercalls to be used by xencomm. The hypercalls simply
-+ * create inlines descriptors for pointers and then call the raw arch hypercall
-+ * xencomm_arch_hypercall_XXX
-+ *
-+ * If the arch wants to directly use these hypercalls, simply define macros
-+ * in asm/hypercall.h, eg:
-+ * #define HYPERVISOR_sched_op xencomm_hypercall_sched_op
-+ *
-+ * The arch may also define HYPERVISOR_xxx as a function and do more operations
-+ * before/after doing the hypercall.
-+ *
-+ * Note: because only inline descriptors are created these functions must only
-+ * be called with in kernel memory parameters.
++void *io_tlb_overflow_buffer;
++
++/*
++ * This is a free list describing the number of free entries available from
++ * each index
+ */
++static unsigned int *io_tlb_list;
++static unsigned int io_tlb_index;
+
-+int
-+xencomm_hypercall_console_io(int cmd, int count, char *str)
-+{
-+ return xencomm_arch_hypercall_console_io
-+ (cmd, count, xencomm_create_inline(str));
-+}
++/*
++ * We need to save away the original address corresponding to a mapped entry
++ * for the sync operations.
++ */
++static unsigned char **io_tlb_orig_addr;
+
-+int
-+xencomm_hypercall_event_channel_op(int cmd, void *op)
-+{
-+ return xencomm_arch_hypercall_event_channel_op
-+ (cmd, xencomm_create_inline(op));
-+}
++/*
++ * Protect the above data structures in the map and unmap calls
++ */
++static DEFINE_SPINLOCK(io_tlb_lock);
+
-+int
-+xencomm_hypercall_xen_version(int cmd, void *arg)
++static int __init
++setup_io_tlb_npages(char *str)
+{
-+ switch (cmd) {
-+ case XENVER_version:
-+ case XENVER_extraversion:
-+ case XENVER_compile_info:
-+ case XENVER_capabilities:
-+ case XENVER_changeset:
-+ case XENVER_platform_parameters:
-+ case XENVER_pagesize:
-+ case XENVER_get_features:
-+ break;
-+ default:
-+ printk("%s: unknown version cmd %d\n", __func__, cmd);
-+ return -ENOSYS;
++ if (isdigit(*str)) {
++ io_tlb_nslabs = simple_strtoul(str, &str, 0);
++ /* avoid tail segment of size < IO_TLB_SEGSIZE */
++ io_tlb_nslabs = ALIGN(io_tlb_nslabs, IO_TLB_SEGSIZE);
+ }
-+
-+ return xencomm_arch_hypercall_xen_version
-+ (cmd, xencomm_create_inline(arg));
++ if (*str == ',')
++ ++str;
++ if (!strcmp(str, "force"))
++ swiotlb_force = 1;
++ return 1;
+}
++__setup("swiotlb=", setup_io_tlb_npages);
++/* make io_tlb_overflow tunable too? */
+
-+int
-+xencomm_hypercall_physdev_op(int cmd, void *op)
++/*
++ * Statically reserve bounce buffer space and initialize bounce buffer data
++ * structures for the software IO TLB used to implement the DMA API.
++ */
++void
++swiotlb_init_with_default_size (size_t default_size)
+{
-+ return xencomm_arch_hypercall_physdev_op
-+ (cmd, xencomm_create_inline(op));
-+}
++ unsigned long i;
+
-+static void *
-+xencommize_grant_table_op(unsigned int cmd, void *op, unsigned int count)
-+{
-+ switch (cmd) {
-+ case GNTTABOP_map_grant_ref:
-+ case GNTTABOP_unmap_grant_ref:
-+ break;
-+ case GNTTABOP_setup_table:
-+ {
-+ struct gnttab_setup_table *setup = op;
-+ struct xencomm_handle *frame_list;
++ if (!io_tlb_nslabs) {
++ io_tlb_nslabs = (default_size >> IO_TLB_SHIFT);
++ io_tlb_nslabs = ALIGN(io_tlb_nslabs, IO_TLB_SEGSIZE);
++ }
+
-+ frame_list = xencomm_create_inline
-+ (xen_guest_handle(setup->frame_list));
++#ifdef CONFIG_XEN
++ if (is_running_on_xen())
++ io_tlb_nslabs = roundup_pow_of_two(io_tlb_nslabs);
++#endif
++ /*
++ * Get IO TLB memory from the low pages
++ */
++ io_tlb_start = alloc_bootmem_low_pages(io_tlb_nslabs * (1 << IO_TLB_SHIFT));
++ if (!io_tlb_start)
++ panic("Cannot allocate SWIOTLB buffer");
++ io_tlb_end = io_tlb_start + io_tlb_nslabs * (1 << IO_TLB_SHIFT);
+
-+ set_xen_guest_handle(setup->frame_list, (void *)frame_list);
-+ break;
-+ }
-+ case GNTTABOP_dump_table:
-+ case GNTTABOP_transfer:
-+ case GNTTABOP_copy:
-+ break;
-+ default:
-+ printk("%s: unknown grant table op %d\n", __func__, cmd);
-+ BUG();
++#ifdef CONFIG_XEN
++ for (i = 0 ; i < io_tlb_nslabs ; i += IO_TLB_SEGSIZE) {
++ if (xen_create_contiguous_region(
++ (unsigned long)io_tlb_start +
++ (i << IO_TLB_SHIFT),
++ get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT),
++ DMA_BITS))
++ panic("Failed to setup Xen contiguous region");
+ }
++#endif
+
-+ return xencomm_create_inline(op);
++ /*
++ * Allocate and initialize the free list array. This array is used
++ * to find contiguous free memory regions of size up to IO_TLB_SEGSIZE
++ * between io_tlb_start and io_tlb_end.
++ */
++ io_tlb_list = alloc_bootmem(io_tlb_nslabs * sizeof(int));
++ for (i = 0; i < io_tlb_nslabs; i++)
++ io_tlb_list[i] = IO_TLB_SEGSIZE - OFFSET(i, IO_TLB_SEGSIZE);
++ io_tlb_index = 0;
++ io_tlb_orig_addr = alloc_bootmem(io_tlb_nslabs * sizeof(char *));
++
++ /*
++ * Get the overflow emergency buffer
++ */
++ io_tlb_overflow_buffer = alloc_bootmem_low(io_tlb_overflow);
++#ifdef CONFIG_XEN
++ if (xen_create_contiguous_region((unsigned long)io_tlb_overflow_buffer,
++ get_order(io_tlb_overflow), DMA_BITS))
++ panic("Failed to setup Xen contiguous region for overflow");
++#endif
++ printk(KERN_INFO "Placing software IO TLB between 0x%lx - 0x%lx\n",
++ virt_to_phys(io_tlb_start), virt_to_phys(io_tlb_end));
+}
+
-+int
-+xencomm_hypercall_grant_table_op(unsigned int cmd, void *op, unsigned int count)
++void
++swiotlb_init (void)
+{
-+ void *desc = xencommize_grant_table_op (cmd, op, count);
-+
-+ return xencomm_arch_hypercall_grant_table_op(cmd, desc, count);
++ swiotlb_init_with_default_size(64 * (1<<20)); /* default to 64MB */
+}
+
++/*
++ * Systems with larger DMA zones (those that don't support ISA) can
++ * initialize the swiotlb later using the slab allocator if needed.
++ * This should be just like above, but with some error catching.
++ */
+int
-+xencomm_hypercall_sched_op(int cmd, void *arg)
++swiotlb_late_init_with_default_size (size_t default_size)
+{
-+ switch (cmd) {
-+ case SCHEDOP_yield:
-+ case SCHEDOP_block:
-+ case SCHEDOP_shutdown:
-+ case SCHEDOP_remote_shutdown:
-+ break;
-+ case SCHEDOP_poll:
-+ {
-+ sched_poll_t *poll = arg;
-+ struct xencomm_handle *ports;
-+
-+ ports = xencomm_create_inline(xen_guest_handle(poll->ports));
++ unsigned long i, req_nslabs = io_tlb_nslabs;
++ unsigned int order;
+
-+ set_xen_guest_handle(poll->ports, (void *)ports);
-+ break;
++ if (!io_tlb_nslabs) {
++ io_tlb_nslabs = (default_size >> IO_TLB_SHIFT);
++ io_tlb_nslabs = ALIGN(io_tlb_nslabs, IO_TLB_SEGSIZE);
+ }
-+ default:
-+ printk("%s: unknown sched op %d\n", __func__, cmd);
-+ return -ENOSYS;
-+ }
-+
-+ return xencomm_arch_hypercall_sched_op(cmd, xencomm_create_inline(arg));
-+}
-+
-+int
-+xencomm_hypercall_multicall(void *call_list, int nr_calls)
-+{
-+ int i;
-+ multicall_entry_t *mce;
+
-+ for (i = 0; i < nr_calls; i++) {
-+ mce = (multicall_entry_t *)call_list + i;
++#ifdef CONFIG_XEN
++ if (is_running_on_xen())
++ io_tlb_nslabs = roundup_pow_of_two(io_tlb_nslabs);
++#endif
++ /*
++ * Get IO TLB memory from the low pages
++ */
++ order = get_order(io_tlb_nslabs * (1 << IO_TLB_SHIFT));
++ io_tlb_nslabs = SLABS_PER_PAGE << order;
+
-+ switch (mce->op) {
-+ case __HYPERVISOR_update_va_mapping:
-+ case __HYPERVISOR_mmu_update:
-+ /* No-op on ia64. */
++ while ((SLABS_PER_PAGE << order) > IO_TLB_MIN_SLABS) {
++ io_tlb_start = (char *)__get_free_pages(GFP_DMA | __GFP_NOWARN,
++ order);
++ if (io_tlb_start)
+ break;
-+ case __HYPERVISOR_grant_table_op:
-+ mce->args[1] = (unsigned long)xencommize_grant_table_op
-+ (mce->args[0], (void *)mce->args[1],
-+ mce->args[2]);
-+ break;
-+ case __HYPERVISOR_memory_op:
-+ default:
-+ printk("%s: unhandled multicall op entry op %lu\n",
-+ __func__, mce->op);
-+ return -ENOSYS;
-+ }
++ order--;
+ }
+
-+ return xencomm_arch_hypercall_multicall
-+ (xencomm_create_inline(call_list), nr_calls);
-+}
++ if (!io_tlb_start)
++ goto cleanup1;
+
-+int
-+xencomm_hypercall_callback_op(int cmd, void *arg)
-+{
-+ switch (cmd)
-+ {
-+ case CALLBACKOP_register:
-+ case CALLBACKOP_unregister:
-+ break;
-+ default:
-+ printk("%s: unknown callback op %d\n", __func__, cmd);
-+ return -ENOSYS;
++ if (order != get_order(io_tlb_nslabs * (1 << IO_TLB_SHIFT))) {
++ printk(KERN_WARNING "Warning: only able to allocate %ld MB "
++ "for software IO TLB\n", (PAGE_SIZE << order) >> 20);
++ io_tlb_nslabs = SLABS_PER_PAGE << order;
+ }
++ io_tlb_end = io_tlb_start + io_tlb_nslabs * (1 << IO_TLB_SHIFT);
++ memset(io_tlb_start, 0, io_tlb_nslabs * (1 << IO_TLB_SHIFT));
+
-+ return xencomm_arch_hypercall_callback_op
-+ (cmd, xencomm_create_inline(arg));
-+}
-+
-+static void
-+xencommize_memory_reservation (xen_memory_reservation_t *mop)
-+{
-+ struct xencomm_handle *desc;
-+
-+ desc = xencomm_create_inline(xen_guest_handle(mop->extent_start));
-+ set_xen_guest_handle(mop->extent_start, (void *)desc);
-+}
-+
-+int
-+xencomm_hypercall_memory_op(unsigned int cmd, void *arg)
-+{
-+ XEN_GUEST_HANDLE(xen_pfn_t) extent_start_va[2];
-+ xen_memory_reservation_t *xmr = NULL, *xme_in = NULL, *xme_out = NULL;
-+ int rc;
++#ifdef CONFIG_XEN
++ for (i = 0 ; i < io_tlb_nslabs ; i += IO_TLB_SEGSIZE) {
++ if (xen_create_contiguous_region(
++ (unsigned long)io_tlb_start +
++ (i << IO_TLB_SHIFT),
++ get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT),
++ DMA_BITS))
++ panic("Failed to setup Xen contiguous region");
++ }
++#endif
++ /*
++ * Allocate and initialize the free list array. This array is used
++ * to find contiguous free memory regions of size up to IO_TLB_SEGSIZE
++ * between io_tlb_start and io_tlb_end.
++ */
++ io_tlb_list = (unsigned int *)__get_free_pages(GFP_KERNEL,
++ get_order(io_tlb_nslabs * sizeof(int)));
++ if (!io_tlb_list)
++ goto cleanup2;
+
-+ switch (cmd) {
-+ case XENMEM_increase_reservation:
-+ case XENMEM_decrease_reservation:
-+ case XENMEM_populate_physmap:
-+ xmr = (xen_memory_reservation_t *)arg;
-+ xen_guest_handle(extent_start_va[0]) =
-+ xen_guest_handle(xmr->extent_start);
-+ xencommize_memory_reservation((xen_memory_reservation_t *)arg);
-+ break;
-+
-+ case XENMEM_maximum_ram_page:
-+ break;
++ for (i = 0; i < io_tlb_nslabs; i++)
++ io_tlb_list[i] = IO_TLB_SEGSIZE - OFFSET(i, IO_TLB_SEGSIZE);
++ io_tlb_index = 0;
+
-+ case XENMEM_exchange:
-+ xme_in = &((xen_memory_exchange_t *)arg)->in;
-+ xme_out = &((xen_memory_exchange_t *)arg)->out;
-+ xen_guest_handle(extent_start_va[0]) =
-+ xen_guest_handle(xme_in->extent_start);
-+ xen_guest_handle(extent_start_va[1]) =
-+ xen_guest_handle(xme_out->extent_start);
-+ xencommize_memory_reservation
-+ (&((xen_memory_exchange_t *)arg)->in);
-+ xencommize_memory_reservation
-+ (&((xen_memory_exchange_t *)arg)->out);
-+ break;
++ io_tlb_orig_addr = (unsigned char **)__get_free_pages(GFP_KERNEL,
++ get_order(io_tlb_nslabs * sizeof(char *)));
++ if (!io_tlb_orig_addr)
++ goto cleanup3;
+
-+ default:
-+ printk("%s: unknown memory op %d\n", __func__, cmd);
-+ return -ENOSYS;
-+ }
++ memset(io_tlb_orig_addr, 0, io_tlb_nslabs * sizeof(char *));
+
-+ rc = xencomm_arch_hypercall_memory_op(cmd, xencomm_create_inline(arg));
++ /*
++ * Get the overflow emergency buffer
++ */
++ io_tlb_overflow_buffer = (void *)__get_free_pages(GFP_DMA,
++ get_order(io_tlb_overflow));
++ if (!io_tlb_overflow_buffer)
++ goto cleanup4;
+
-+ switch (cmd) {
-+ case XENMEM_increase_reservation:
-+ case XENMEM_decrease_reservation:
-+ case XENMEM_populate_physmap:
-+ xen_guest_handle(xmr->extent_start) =
-+ xen_guest_handle(extent_start_va[0]);
-+ break;
++#ifdef CONFIG_XEN
++ if (xen_create_contiguous_region((unsigned long)io_tlb_overflow_buffer,
++ get_order(io_tlb_overflow), DMA_BITS))
++ panic("Failed to setup Xen contiguous region for overflow");
++#endif
++ printk(KERN_INFO "Placing %ldMB software IO TLB between 0x%lx - "
++ "0x%lx\n", (io_tlb_nslabs * (1 << IO_TLB_SHIFT)) >> 20,
++ virt_to_phys(io_tlb_start), virt_to_phys(io_tlb_end));
+
-+ case XENMEM_exchange:
-+ xen_guest_handle(xme_in->extent_start) =
-+ xen_guest_handle(extent_start_va[0]);
-+ xen_guest_handle(xme_out->extent_start) =
-+ xen_guest_handle(extent_start_va[1]);
-+ break;
-+ }
++ return 0;
+
-+ return rc;
++cleanup4:
++ free_pages((unsigned long)io_tlb_orig_addr, get_order(io_tlb_nslabs *
++ sizeof(char *)));
++ io_tlb_orig_addr = NULL;
++cleanup3:
++ free_pages((unsigned long)io_tlb_list, get_order(io_tlb_nslabs *
++ sizeof(int)));
++ io_tlb_list = NULL;
++ io_tlb_end = NULL;
++cleanup2:
++ free_pages((unsigned long)io_tlb_start, order);
++ io_tlb_start = NULL;
++cleanup1:
++ io_tlb_nslabs = req_nslabs;
++ return -ENOMEM;
+}
+
-+unsigned long
-+xencomm_hypercall_hvm_op(int cmd, void *arg)
++static inline int
++address_needs_mapping(struct device *hwdev, dma_addr_t addr)
+{
-+ switch (cmd) {
-+ case HVMOP_set_param:
-+ case HVMOP_get_param:
-+ break;
-+ default:
-+ printk("%s: unknown hvm op %d\n", __func__, cmd);
-+ return -ENOSYS;
-+ }
-+
-+ return xencomm_arch_hypercall_hvm_op(cmd, xencomm_create_inline(arg));
++ dma_addr_t mask = 0xffffffff;
++ /* If the device has a mask, use it, otherwise default to 32 bits */
++ if (hwdev && hwdev->dma_mask)
++ mask = *hwdev->dma_mask;
++ return (addr & ~mask) != 0;
+}
+
-+int
-+xencomm_hypercall_suspend(unsigned long srec)
++/*
++ * Allocates bounce buffer and returns its kernel virtual address.
++ */
++static void *
++map_single(struct device *hwdev, char *buffer, size_t size, int dir)
+{
-+ struct sched_shutdown arg;
++ unsigned long flags;
++ char *dma_addr;
++ unsigned int nslots, stride, index, wrap;
++ char *slot_buf;
++ int i;
+
-+ arg.reason = SHUTDOWN_suspend;
++ /*
++ * For mappings greater than a page, we limit the stride (and
++ * hence alignment) to a page size.
++ */
++ nslots = ALIGN(size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT;
++ if (size > PAGE_SIZE)
++ stride = (1 << (PAGE_SHIFT - IO_TLB_SHIFT));
++ else
++ stride = 1;
+
-+ return xencomm_arch_hypercall_suspend(xencomm_create_inline(&arg));
-+}
++ BUG_ON(!nslots);
+
-+int
-+xencomm_hypercall_xenoprof_op(int op, void *arg)
-+{
-+ switch (op) {
-+ case XENOPROF_init:
-+ case XENOPROF_set_active:
-+ case XENOPROF_set_passive:
-+ case XENOPROF_counter:
-+ case XENOPROF_get_buffer:
-+ break;
++ /*
++ * Find suitable number of IO TLB entries size that will fit this
++ * request and allocate a buffer from that IO TLB pool.
++ */
++ spin_lock_irqsave(&io_tlb_lock, flags);
++ {
++ wrap = index = ALIGN(io_tlb_index, stride);
+
-+ case XENOPROF_reset_active_list:
-+ case XENOPROF_reset_passive_list:
-+ case XENOPROF_reserve_counters:
-+ case XENOPROF_setup_events:
-+ case XENOPROF_enable_virq:
-+ case XENOPROF_start:
-+ case XENOPROF_stop:
-+ case XENOPROF_disable_virq:
-+ case XENOPROF_release_counters:
-+ case XENOPROF_shutdown:
-+ return xencomm_arch_hypercall_xenoprof_op(op, arg);
-+ break;
++ if (index >= io_tlb_nslabs)
++ wrap = index = 0;
+
-+ default:
-+ printk("%s: op %d isn't supported\n", __func__, op);
-+ return -ENOSYS;
-+ }
-+ return xencomm_arch_hypercall_xenoprof_op(op,
-+ xencomm_create_inline(arg));
-+}
++ do {
++ /*
++ * If we find a slot that indicates we have 'nslots'
++ * number of contiguous buffers, we allocate the
++ * buffers from that slot and mark the entries as '0'
++ * indicating unavailable.
++ */
++ if (io_tlb_list[index] >= nslots) {
++ int count = 0;
+
-+int
-+xencomm_hypercall_perfmon_op(unsigned long cmd, void* arg, unsigned long count)
-+{
-+ switch (cmd) {
-+ case PFM_GET_FEATURES:
-+ case PFM_CREATE_CONTEXT:
-+ case PFM_WRITE_PMCS:
-+ case PFM_WRITE_PMDS:
-+ case PFM_LOAD_CONTEXT:
-+ break;
++ for (i = index; i < (int) (index + nslots); i++)
++ io_tlb_list[i] = 0;
++ for (i = index - 1; (OFFSET(i, IO_TLB_SEGSIZE) != IO_TLB_SEGSIZE -1) && io_tlb_list[i]; i--)
++ io_tlb_list[i] = ++count;
++ dma_addr = io_tlb_start + (index << IO_TLB_SHIFT);
+
-+ case PFM_DESTROY_CONTEXT:
-+ case PFM_UNLOAD_CONTEXT:
-+ case PFM_START:
-+ case PFM_STOP:
-+ return xencomm_arch_hypercall_perfmon_op(cmd, arg, count);
++ /*
++ * Update the indices to avoid searching in
++ * the next round.
++ */
++ io_tlb_index = ((index + nslots) < io_tlb_nslabs
++ ? (index + nslots) : 0);
+
-+ default:
-+ printk("%s:%d cmd %ld isn't supported\n",
-+ __func__,__LINE__, cmd);
-+ BUG();
-+ }
++ goto found;
++ }
++ index += stride;
++ if (index >= io_tlb_nslabs)
++ index = 0;
++ } while (index != wrap);
+
-+ return xencomm_arch_hypercall_perfmon_op(cmd,
-+ xencomm_create_inline(arg),
-+ count);
-+}
-+
-+long
-+xencomm_hypercall_vcpu_op(int cmd, int cpu, void *arg)
-+{
-+ switch (cmd) {
-+ case VCPUOP_register_runstate_memory_area:
-+ xencommize_memory_reservation((xen_memory_reservation_t *)arg);
-+ break;
++ spin_unlock_irqrestore(&io_tlb_lock, flags);
++ return NULL;
++ }
++ found:
++ spin_unlock_irqrestore(&io_tlb_lock, flags);
+
-+ default:
-+ printk("%s: unknown vcpu op %d\n", __func__, cmd);
-+ return -ENOSYS;
++ /*
++ * Save away the mapping from the original address to the DMA address.
++ * This is needed when we sync the memory. Then we sync the buffer if
++ * needed.
++ */
++ slot_buf = buffer;
++ for (i = 0; i < nslots; i++) {
++ io_tlb_orig_addr[index + i] = slot_buf;
++ slot_buf += 1 << IO_TLB_SHIFT;
+ }
++ if (dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL)
++ memcpy(dma_addr, buffer, size);
+
-+ return xencomm_arch_hypercall_vcpu_op(cmd, cpu,
-+ xencomm_create_inline(arg));
++ return dma_addr;
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/xcom_mini.c tmp-linux-2.6-xen.patch/arch/ia64/xen/xcom_mini.c
---- pristine-linux-2.6.18.2/arch/ia64/xen/xcom_mini.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/xcom_mini.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,456 @@
++
+/*
-+ * This program is free software; you can redistribute it and/or modify
-+ * it under the terms of the GNU General Public License as published by
-+ * the Free Software Foundation; either version 2 of the License, or
-+ * (at your option) any later version.
-+ *
-+ * This program is distributed in the hope that it will be useful,
-+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ * GNU General Public License for more details.
-+ *
-+ * You should have received a copy of the GNU General Public License
-+ * along with this program; if not, write to the Free Software
-+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
-+ *
-+ * Tristan Gingold <tristan.gingold at bull.net>
++ * dma_addr is the kernel virtual address of the bounce buffer to unmap.
+ */
-+#include <linux/types.h>
-+#include <linux/errno.h>
-+#include <linux/kernel.h>
-+#include <linux/module.h>
-+#include <xen/interface/xen.h>
-+#include <xen/interface/platform.h>
-+#include <xen/interface/memory.h>
-+#include <xen/interface/xencomm.h>
-+#include <xen/interface/version.h>
-+#include <xen/interface/event_channel.h>
-+#include <xen/interface/physdev.h>
-+#include <xen/interface/grant_table.h>
-+#include <xen/interface/hvm/params.h>
-+#include <xen/interface/xenoprof.h>
-+#ifdef CONFIG_VMX_GUEST
-+#include <asm/hypervisor.h>
-+#else
-+#include <asm/hypercall.h>
-+#endif
-+#include <asm/xen/xencomm.h>
-+#include <asm/perfmon.h>
-+
-+int
-+xencomm_mini_hypercall_event_channel_op(int cmd, void *op)
++static void
++unmap_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
+{
-+ struct xencomm_mini xc_area[2];
-+ int nbr_area = 2;
-+ struct xencomm_handle *desc;
-+ int rc;
++ unsigned long flags;
++ int i, count, nslots = ALIGN(size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT;
++ int index = (dma_addr - io_tlb_start) >> IO_TLB_SHIFT;
++ char *buffer = io_tlb_orig_addr[index];
+
-+ rc = xencomm_create_mini(xc_area, &nbr_area,
-+ op, sizeof(evtchn_op_t), &desc);
-+ if (rc)
-+ return rc;
++ /*
++ * First, sync the memory before unmapping the entry
++ */
++ if (buffer && ((dir == DMA_FROM_DEVICE) || (dir == DMA_BIDIRECTIONAL)))
++ /*
++ * bounce... copy the data back into the original buffer * and
++ * delete the bounce buffer.
++ */
++ memcpy(buffer, dma_addr, size);
+
-+ return xencomm_arch_hypercall_event_channel_op(cmd, desc);
++ /*
++ * Return the buffer to the free list by setting the corresponding
++ * entries to indicate the number of contigous entries available.
++ * While returning the entries to the free list, we merge the entries
++ * with slots below and above the pool being returned.
++ */
++ spin_lock_irqsave(&io_tlb_lock, flags);
++ {
++ count = ((index + nslots) < ALIGN(index + 1, IO_TLB_SEGSIZE) ?
++ io_tlb_list[index + nslots] : 0);
++ /*
++ * Step 1: return the slots to the free list, merging the
++ * slots with superceeding slots
++ */
++ for (i = index + nslots - 1; i >= index; i--)
++ io_tlb_list[i] = ++count;
++ /*
++ * Step 2: merge the returned slots with the preceding slots,
++ * if available (non zero)
++ */
++ for (i = index - 1; (OFFSET(i, IO_TLB_SEGSIZE) != IO_TLB_SEGSIZE -1) && io_tlb_list[i]; i--)
++ io_tlb_list[i] = ++count;
++ }
++ spin_unlock_irqrestore(&io_tlb_lock, flags);
+}
-+EXPORT_SYMBOL(xencomm_mini_hypercall_event_channel_op);
+
-+static int
-+xencommize_mini_grant_table_op(struct xencomm_mini *xc_area, int *nbr_area,
-+ unsigned int cmd, void *op, unsigned int count,
-+ struct xencomm_handle **desc)
++static void
++sync_single(struct device *hwdev, char *dma_addr, size_t size,
++ int dir, int target)
+{
-+ struct xencomm_handle *desc1;
-+ unsigned int argsize;
-+ int rc;
++ int index = (dma_addr - io_tlb_start) >> IO_TLB_SHIFT;
++ char *buffer = io_tlb_orig_addr[index];
+
-+ switch (cmd) {
-+ case GNTTABOP_map_grant_ref:
-+ argsize = sizeof(struct gnttab_map_grant_ref);
++ switch (target) {
++ case SYNC_FOR_CPU:
++ if (likely(dir == DMA_FROM_DEVICE || dir == DMA_BIDIRECTIONAL))
++ memcpy(buffer, dma_addr, size);
++ else
++ BUG_ON(dir != DMA_TO_DEVICE);
+ break;
-+ case GNTTABOP_unmap_grant_ref:
-+ argsize = sizeof(struct gnttab_unmap_grant_ref);
++ case SYNC_FOR_DEVICE:
++ if (likely(dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL))
++ memcpy(dma_addr, buffer, size);
++ else
++ BUG_ON(dir != DMA_FROM_DEVICE);
+ break;
-+ case GNTTABOP_setup_table:
-+ {
-+ struct gnttab_setup_table *setup = op;
++ default:
++ BUG();
++ }
++}
+
-+ argsize = sizeof(*setup);
++void *
++swiotlb_alloc_coherent(struct device *hwdev, size_t size,
++ dma_addr_t *dma_handle, gfp_t flags)
++{
++ unsigned long dev_addr;
++ void *ret;
++ int order = get_order(size);
+
-+ if (count != 1)
-+ return -EINVAL;
-+ rc = xencomm_create_mini
-+ (xc_area, nbr_area,
-+ xen_guest_handle(setup->frame_list),
-+ setup->nr_frames
-+ * sizeof(*xen_guest_handle(setup->frame_list)),
-+ &desc1);
-+ if (rc)
-+ return rc;
-+ set_xen_guest_handle(setup->frame_list, (void *)desc1);
-+ break;
++ /*
++ * XXX fix me: the DMA API should pass us an explicit DMA mask
++ * instead, or use ZONE_DMA32 (ia64 overloads ZONE_DMA to be a ~32
++ * bit range instead of a 16MB one).
++ */
++ flags |= GFP_DMA;
++
++ ret = (void *)__get_free_pages(flags, order);
++#ifdef CONFIG_XEN
++ if (ret && is_running_on_xen()) {
++ if (xen_create_contiguous_region((unsigned long)ret, order,
++ fls64(hwdev->coherent_dma_mask))) {
++ free_pages((unsigned long)ret, order);
++ ret = NULL;
++ } else {
++ /*
++ * Short circuit the rest, xen_create_contiguous_region
++ * should fail if it didn't give us an address within
++ * the mask requested.
++ */
++ memset(ret, 0, size);
++ *dma_handle = virt_to_bus(ret);
++ return ret;
++ }
+ }
-+ case GNTTABOP_dump_table:
-+ argsize = sizeof(struct gnttab_dump_table);
-+ break;
-+ case GNTTABOP_transfer:
-+ argsize = sizeof(struct gnttab_transfer);
-+ break;
-+ case GNTTABOP_copy:
-+ argsize = sizeof(struct gnttab_copy);
-+ break;
-+ case GNTTABOP_query_size:
-+ argsize = sizeof(struct gnttab_query_size);
-+ break;
-+ default:
-+ printk("%s: unknown mini grant table op %d\n", __func__, cmd);
-+ BUG();
++#endif
++ if (ret && address_needs_mapping(hwdev, virt_to_bus(ret))) {
++ /*
++ * The allocated memory isn't reachable by the device.
++ * Fall back on swiotlb_map_single().
++ */
++ free_pages((unsigned long) ret, order);
++ ret = NULL;
+ }
++ if (!ret) {
++ /*
++ * We are either out of memory or the device can't DMA
++ * to GFP_DMA memory; fall back on
++ * swiotlb_map_single(), which will grab memory from
++ * the lowest available address range.
++ */
++ dma_addr_t handle;
++ handle = swiotlb_map_single(NULL, NULL, size, DMA_FROM_DEVICE);
++ if (swiotlb_dma_mapping_error(handle))
++ return NULL;
+
-+ rc = xencomm_create_mini(xc_area, nbr_area, op, count * argsize, desc);
-+ if (rc)
-+ return rc;
++ ret = bus_to_virt(handle);
++ }
+
-+ return 0;
++ memset(ret, 0, size);
++ dev_addr = virt_to_bus(ret);
++
++ /* Confirm address can be DMA'd by device */
++ if (address_needs_mapping(hwdev, dev_addr)) {
++ printk("hwdev DMA mask = 0x%016Lx, dev_addr = 0x%016lx\n",
++ (unsigned long long)*hwdev->dma_mask, dev_addr);
++ panic("swiotlb_alloc_coherent: allocated memory is out of "
++ "range for device");
++ }
++ *dma_handle = dev_addr;
++ return ret;
+}
+
-+int
-+xencomm_mini_hypercall_grant_table_op(unsigned int cmd, void *op,
-+ unsigned int count)
++void
++swiotlb_free_coherent(struct device *hwdev, size_t size, void *vaddr,
++ dma_addr_t dma_handle)
+{
-+ int rc;
-+ struct xencomm_handle *desc;
-+ int nbr_area = 2;
-+ struct xencomm_mini xc_area[2];
++ if (!(vaddr >= (void *)io_tlb_start
++ && vaddr < (void *)io_tlb_end)) {
++#ifdef CONFIG_XEN
++ xen_destroy_contiguous_region((unsigned long)vaddr,
++ get_order(size));
++#endif
++ free_pages((unsigned long) vaddr, get_order(size));
++ } else
++ /* DMA_TO_DEVICE to avoid memcpy in unmap_single */
++ swiotlb_unmap_single (hwdev, dma_handle, size, DMA_TO_DEVICE);
++}
+
-+ rc = xencommize_mini_grant_table_op(xc_area, &nbr_area,
-+ cmd, op, count, &desc);
-+ if (rc)
-+ return rc;
++static void
++swiotlb_full(struct device *dev, size_t size, int dir, int do_panic)
++{
++ /*
++ * Ran out of IOMMU space for this operation. This is very bad.
++ * Unfortunately the drivers cannot handle this operation properly.
++ * unless they check for dma_mapping_error (most don't)
++ * When the mapping is small enough return a static buffer to limit
++ * the damage, or panic when the transfer is too big.
++ */
++ printk(KERN_ERR "DMA: Out of SW-IOMMU space for %lu bytes at "
++ "device %s\n", size, dev ? dev->bus_id : "?");
+
-+ return xencomm_arch_hypercall_grant_table_op(cmd, desc, count);
++ if (size > io_tlb_overflow && do_panic) {
++ if (dir == DMA_FROM_DEVICE || dir == DMA_BIDIRECTIONAL)
++ panic("DMA: Memory would be corrupted\n");
++ if (dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL)
++ panic("DMA: Random memory would be DMAed\n");
++ }
+}
-+EXPORT_SYMBOL(xencomm_mini_hypercall_grant_table_op);
+
-+int
-+xencomm_mini_hypercall_multicall(void *call_list, int nr_calls)
++/*
++ * Map a single buffer of the indicated size for DMA in streaming mode. The
++ * physical address to use is returned.
++ *
++ * Once the device is given the dma address, the device owns this memory until
++ * either swiotlb_unmap_single or swiotlb_dma_sync_single is performed.
++ */
++dma_addr_t
++swiotlb_map_single(struct device *hwdev, void *ptr, size_t size, int dir)
+{
-+ int i;
-+ multicall_entry_t *mce;
-+ int nbr_area = 2 + nr_calls * 3;
-+ struct xencomm_mini xc_area[nbr_area];
-+ struct xencomm_handle *desc;
-+ int rc;
++ unsigned long dev_addr = gnttab_dma_map_virt(ptr);
++ void *map;
+
-+ for (i = 0; i < nr_calls; i++) {
-+ mce = (multicall_entry_t *)call_list + i;
++ BUG_ON(dir == DMA_NONE);
++ /*
++ * If the pointer passed in happens to be in the device's DMA window,
++ * we can safely return the device addr and not worry about bounce
++ * buffering it.
++ */
++ if (!range_straddles_page_boundary(__pa(ptr), size) &&
++ !address_needs_mapping(hwdev, dev_addr) && !swiotlb_force)
++ return dev_addr;
+
-+ switch (mce->op) {
-+ case __HYPERVISOR_update_va_mapping:
-+ case __HYPERVISOR_mmu_update:
-+ /* No-op on ia64. */
-+ break;
-+ case __HYPERVISOR_grant_table_op:
-+ rc = xencommize_mini_grant_table_op
-+ (xc_area, &nbr_area,
-+ mce->args[0], (void *)mce->args[1],
-+ mce->args[2], &desc);
-+ if (rc)
-+ return rc;
-+ mce->args[1] = (unsigned long)desc;
-+ break;
-+ case __HYPERVISOR_memory_op:
-+ default:
-+ printk("%s: unhandled multicall op entry op %lu\n",
-+ __func__, mce->op);
-+ return -ENOSYS;
-+ }
++ __gnttab_dma_unmap_page(virt_to_page(ptr));
++ /*
++ * Oh well, have to allocate and map a bounce buffer.
++ */
++ map = map_single(hwdev, ptr, size, dir);
++ if (!map) {
++ swiotlb_full(hwdev, size, dir, 1);
++ map = io_tlb_overflow_buffer;
+ }
+
-+ rc = xencomm_create_mini(xc_area, &nbr_area, call_list,
-+ nr_calls * sizeof(multicall_entry_t), &desc);
-+ if (rc)
-+ return rc;
++ dev_addr = virt_to_bus(map);
+
-+ return xencomm_arch_hypercall_multicall(desc, nr_calls);
++ /*
++ * Ensure that the address returned is DMA'ble
++ */
++ if (address_needs_mapping(hwdev, dev_addr))
++ panic("map_single: bounce buffer is not DMA'ble");
++
++ return dev_addr;
+}
-+EXPORT_SYMBOL(xencomm_mini_hypercall_multicall);
+
-+static int
-+xencommize_mini_memory_reservation(struct xencomm_mini *area, int *nbr_area,
-+ xen_memory_reservation_t *mop)
++/*
++ * Since DMA is i-cache coherent, any (complete) pages that were written via
++ * DMA can be marked as "clean" so that lazy_mmu_prot_update() doesn't have to
++ * flush them when they get mapped into an executable vm-area.
++ */
++static void
++mark_clean(void *addr, size_t size)
+{
-+ struct xencomm_handle *desc;
-+ int rc;
-+
-+ rc = xencomm_create_mini
-+ (area, nbr_area,
-+ xen_guest_handle(mop->extent_start),
-+ mop->nr_extents
-+ * sizeof(*xen_guest_handle(mop->extent_start)),
-+ &desc);
-+ if (rc)
-+ return rc;
++ unsigned long pg_addr, end;
+
-+ set_xen_guest_handle(mop->extent_start, (void *)desc);
++#ifdef CONFIG_XEN
++ /* XXX: Bad things happen when starting domUs if this is enabled. */
++ if (is_running_on_xen())
++ return;
++#endif
+
-+ return 0;
++ pg_addr = PAGE_ALIGN((unsigned long) addr);
++ end = (unsigned long) addr + size;
++ while (pg_addr + PAGE_SIZE <= end) {
++ struct page *page = virt_to_page(pg_addr);
++ set_bit(PG_arch_1, &page->flags);
++ pg_addr += PAGE_SIZE;
++ }
+}
+
-+int
-+xencomm_mini_hypercall_memory_op(unsigned int cmd, void *arg)
++/*
++ * Unmap a single streaming mode DMA translation. The dma_addr and size must
++ * match what was provided for in a previous swiotlb_map_single call. All
++ * other usages are undefined.
++ *
++ * After this call, reads by the cpu to the buffer are guaranteed to see
++ * whatever the device wrote there.
++ */
++void
++swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr, size_t size,
++ int dir)
+{
-+ int nbr_area = 4;
-+ struct xencomm_mini xc_area[4];
-+ struct xencomm_handle *desc;
-+ int rc;
-+ unsigned int argsize;
-+
-+ switch (cmd) {
-+ case XENMEM_increase_reservation:
-+ case XENMEM_decrease_reservation:
-+ case XENMEM_populate_physmap:
-+ argsize = sizeof(xen_memory_reservation_t);
-+ rc = xencommize_mini_memory_reservation
-+ (xc_area, &nbr_area, (xen_memory_reservation_t *)arg);
-+ if (rc)
-+ return rc;
-+ break;
-+
-+ case XENMEM_maximum_ram_page:
-+ argsize = 0;
-+ break;
++ char *dma_addr = bus_to_virt(dev_addr);
+
-+ case XENMEM_exchange:
-+ argsize = sizeof(xen_memory_exchange_t);
-+ rc = xencommize_mini_memory_reservation
-+ (xc_area, &nbr_area,
-+ &((xen_memory_exchange_t *)arg)->in);
-+ if (rc)
-+ return rc;
-+ rc = xencommize_mini_memory_reservation
-+ (xc_area, &nbr_area,
-+ &((xen_memory_exchange_t *)arg)->out);
-+ if (rc)
-+ return rc;
-+ break;
++ BUG_ON(dir == DMA_NONE);
++ if (dma_addr >= io_tlb_start && dma_addr < io_tlb_end)
++ unmap_single(hwdev, dma_addr, size, dir);
++ else {
++ __gnttab_dma_unmap_page(virt_to_page(dma_addr));
++ if (dir == DMA_FROM_DEVICE)
++ mark_clean(dma_addr, size);
++ }
++}
+
-+ case XENMEM_add_to_physmap:
-+ argsize = sizeof (xen_add_to_physmap_t);
-+ break;
++/*
++ * Make physical memory consistent for a single streaming mode DMA translation
++ * after a transfer.
++ *
++ * If you perform a swiotlb_map_single() but wish to interrogate the buffer
++ * using the cpu, yet do not wish to teardown the dma mapping, you must
++ * call this function before doing so. At the next point you give the dma
++ * address back to the card, you must first perform a
++ * swiotlb_dma_sync_for_device, and then the device again owns the buffer
++ */
++static inline void
++swiotlb_sync_single(struct device *hwdev, dma_addr_t dev_addr,
++ size_t size, int dir, int target)
++{
++ char *dma_addr = bus_to_virt(dev_addr);
+
-+ default:
-+ printk("%s: unknown mini memory op %d\n", __func__, cmd);
-+ return -ENOSYS;
-+ }
++ BUG_ON(dir == DMA_NONE);
++ if (dma_addr >= io_tlb_start && dma_addr < io_tlb_end)
++ sync_single(hwdev, dma_addr, size, dir, target);
++ else if (dir == DMA_FROM_DEVICE)
++ mark_clean(dma_addr, size);
++}
+
-+ rc = xencomm_create_mini(xc_area, &nbr_area, arg, argsize, &desc);
-+ if (rc)
-+ return rc;
++void
++swiotlb_sync_single_for_cpu(struct device *hwdev, dma_addr_t dev_addr,
++ size_t size, int dir)
++{
++ swiotlb_sync_single(hwdev, dev_addr, size, dir, SYNC_FOR_CPU);
++}
+
-+ return xencomm_arch_hypercall_memory_op(cmd, desc);
++void
++swiotlb_sync_single_for_device(struct device *hwdev, dma_addr_t dev_addr,
++ size_t size, int dir)
++{
++ swiotlb_sync_single(hwdev, dev_addr, size, dir, SYNC_FOR_DEVICE);
+}
-+EXPORT_SYMBOL(xencomm_mini_hypercall_memory_op);
+
-+unsigned long
-+xencomm_mini_hypercall_hvm_op(int cmd, void *arg)
++/*
++ * Same as above, but for a sub-range of the mapping.
++ */
++static inline void
++swiotlb_sync_single_range(struct device *hwdev, dma_addr_t dev_addr,
++ unsigned long offset, size_t size,
++ int dir, int target)
+{
-+ struct xencomm_handle *desc;
-+ int nbr_area = 2;
-+ struct xencomm_mini xc_area[2];
-+ unsigned int argsize;
-+ int rc;
++ char *dma_addr = bus_to_virt(dev_addr) + offset;
+
-+ switch (cmd) {
-+ case HVMOP_get_param:
-+ case HVMOP_set_param:
-+ argsize = sizeof(xen_hvm_param_t);
-+ break;
-+ default:
-+ printk("%s: unknown HVMOP %d\n", __func__, cmd);
-+ return -EINVAL;
-+ }
++ BUG_ON(dir == DMA_NONE);
++ if (dma_addr >= io_tlb_start && dma_addr < io_tlb_end)
++ sync_single(hwdev, dma_addr, size, dir, target);
++ else if (dir == DMA_FROM_DEVICE)
++ mark_clean(dma_addr, size);
++}
+
-+ rc = xencomm_create_mini(xc_area, &nbr_area, arg, argsize, &desc);
-+ if (rc)
-+ return rc;
++void
++swiotlb_sync_single_range_for_cpu(struct device *hwdev, dma_addr_t dev_addr,
++ unsigned long offset, size_t size, int dir)
++{
++ swiotlb_sync_single_range(hwdev, dev_addr, offset, size, dir,
++ SYNC_FOR_CPU);
++}
+
-+ return xencomm_arch_hypercall_hvm_op(cmd, desc);
++void
++swiotlb_sync_single_range_for_device(struct device *hwdev, dma_addr_t dev_addr,
++ unsigned long offset, size_t size, int dir)
++{
++ swiotlb_sync_single_range(hwdev, dev_addr, offset, size, dir,
++ SYNC_FOR_DEVICE);
+}
-+EXPORT_SYMBOL(xencomm_mini_hypercall_hvm_op);
+
++/*
++ * Map a set of buffers described by scatterlist in streaming mode for DMA.
++ * This is the scatter-gather version of the above swiotlb_map_single
++ * interface. Here the scatter gather list elements are each tagged with the
++ * appropriate dma address and length. They are obtained via
++ * sg_dma_{address,length}(SG).
++ *
++ * NOTE: An implementation may be able to use a smaller number of
++ * DMA address/length pairs than there are SG table elements.
++ * (for example via virtual mapping capabilities)
++ * The routine returns the number of addr/length pairs actually
++ * used, at most nents.
++ *
++ * Device ownership issues as mentioned above for swiotlb_map_single are the
++ * same here.
++ */
+int
-+xencomm_mini_hypercall_xen_version(int cmd, void *arg)
++swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg, int nelems,
++ int dir)
+{
-+ struct xencomm_handle *desc;
-+ int nbr_area = 2;
-+ struct xencomm_mini xc_area[2];
-+ unsigned int argsize;
-+ int rc;
++ void *addr;
++ unsigned long dev_addr;
++ int i;
+
-+ switch (cmd) {
-+ case XENVER_version:
-+ /* do not actually pass an argument */
-+ return xencomm_arch_hypercall_xen_version(cmd, 0);
-+ case XENVER_extraversion:
-+ argsize = sizeof(xen_extraversion_t);
-+ break;
-+ case XENVER_compile_info:
-+ argsize = sizeof(xen_compile_info_t);
-+ break;
-+ case XENVER_capabilities:
-+ argsize = sizeof(xen_capabilities_info_t);
-+ break;
-+ case XENVER_changeset:
-+ argsize = sizeof(xen_changeset_info_t);
-+ break;
-+ case XENVER_platform_parameters:
-+ argsize = sizeof(xen_platform_parameters_t);
-+ break;
-+ case XENVER_pagesize:
-+ argsize = (arg == NULL) ? 0 : sizeof(void *);
-+ break;
-+ case XENVER_get_features:
-+ argsize = (arg == NULL) ? 0 : sizeof(xen_feature_info_t);
-+ break;
++ BUG_ON(dir == DMA_NONE);
+
-+ default:
-+ printk("%s: unknown version op %d\n", __func__, cmd);
-+ return -ENOSYS;
++ for (i = 0; i < nelems; i++, sg++) {
++ addr = SG_ENT_VIRT_ADDRESS(sg);
++ dev_addr = gnttab_dma_map_virt(addr);
++ if (swiotlb_force ||
++ range_straddles_page_boundary(page_to_pseudophys(sg->page)
++ + sg->offset, sg->length) ||
++ address_needs_mapping(hwdev, dev_addr)) {
++ void *map;
++ __gnttab_dma_unmap_page(sg->page);
++ map = map_single(hwdev, addr, sg->length, dir);
++ sg->dma_address = virt_to_bus(map);
++ if (!map) {
++ /* Don't panic here, we expect map_sg users
++ to do proper error handling. */
++ swiotlb_full(hwdev, sg->length, dir, 0);
++ swiotlb_unmap_sg(hwdev, sg - i, i, dir);
++ sg[0].dma_length = 0;
++ return 0;
++ }
++ } else
++ sg->dma_address = dev_addr;
++ sg->dma_length = sg->length;
+ }
++ return nelems;
++}
+
-+ rc = xencomm_create_mini(xc_area, &nbr_area, arg, argsize, &desc);
-+ if (rc)
-+ return rc;
++/*
++ * Unmap a set of streaming mode DMA translations. Again, cpu read rules
++ * concerning calls here are the same as for swiotlb_unmap_single() above.
++ */
++void
++swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nelems,
++ int dir)
++{
++ int i;
+
-+ return xencomm_arch_hypercall_xen_version(cmd, desc);
++ BUG_ON(dir == DMA_NONE);
++
++ for (i = 0; i < nelems; i++, sg++)
++ if (sg->dma_address != SG_ENT_PHYS_ADDRESS(sg))
++ unmap_single(hwdev, (void *) bus_to_virt(sg->dma_address), sg->dma_length, dir);
++ else {
++ __gnttab_dma_unmap_page(sg->page);
++ if (dir == DMA_FROM_DEVICE)
++ mark_clean(SG_ENT_VIRT_ADDRESS(sg),
++ sg->dma_length);
++ }
+}
-+EXPORT_SYMBOL(xencomm_mini_hypercall_xen_version);
+
-+int
-+xencomm_mini_hypercall_xenoprof_op(int op, void *arg)
++/*
++ * Make physical memory consistent for a set of streaming mode DMA translations
++ * after a transfer.
++ *
++ * The same as swiotlb_sync_single_* but for a scatter-gather list, same rules
++ * and usage.
++ */
++static inline void
++swiotlb_sync_sg(struct device *hwdev, struct scatterlist *sg,
++ int nelems, int dir, int target)
+{
-+ unsigned int argsize;
-+ struct xencomm_mini xc_area[2];
-+ int nbr_area = 2;
-+ struct xencomm_handle *desc;
-+ int rc;
++ int i;
+
-+ switch (op) {
-+ case XENOPROF_init:
-+ argsize = sizeof(xenoprof_init_t);
-+ break;
-+ case XENOPROF_set_active:
-+ argsize = sizeof(domid_t);
-+ break;
-+ case XENOPROF_set_passive:
-+ argsize = sizeof(xenoprof_passive_t);
-+ break;
-+ case XENOPROF_counter:
-+ argsize = sizeof(xenoprof_counter_t);
-+ break;
-+ case XENOPROF_get_buffer:
-+ argsize = sizeof(xenoprof_get_buffer_t);
-+ break;
++ BUG_ON(dir == DMA_NONE);
+
-+ case XENOPROF_reset_active_list:
-+ case XENOPROF_reset_passive_list:
-+ case XENOPROF_reserve_counters:
-+ case XENOPROF_setup_events:
-+ case XENOPROF_enable_virq:
-+ case XENOPROF_start:
-+ case XENOPROF_stop:
-+ case XENOPROF_disable_virq:
-+ case XENOPROF_release_counters:
-+ case XENOPROF_shutdown:
-+ return xencomm_arch_hypercall_xenoprof_op(op, arg);
++ for (i = 0; i < nelems; i++, sg++)
++ if (sg->dma_address != SG_ENT_PHYS_ADDRESS(sg))
++ sync_single(hwdev, (void *) sg->dma_address,
++ sg->dma_length, dir, target);
++}
+
-+ default:
-+ printk("%s: op %d isn't supported\n", __func__, op);
-+ return -ENOSYS;
-+ }
-+ rc = xencomm_create_mini(xc_area, &nbr_area, arg, argsize, &desc);
-+ if (rc)
-+ return rc;
-+ return xencomm_arch_hypercall_xenoprof_op(op, desc);
++void
++swiotlb_sync_sg_for_cpu(struct device *hwdev, struct scatterlist *sg,
++ int nelems, int dir)
++{
++ swiotlb_sync_sg(hwdev, sg, nelems, dir, SYNC_FOR_CPU);
++}
++
++void
++swiotlb_sync_sg_for_device(struct device *hwdev, struct scatterlist *sg,
++ int nelems, int dir)
++{
++ swiotlb_sync_sg(hwdev, sg, nelems, dir, SYNC_FOR_DEVICE);
+}
-+EXPORT_SYMBOL_GPL(xencomm_mini_hypercall_xenoprof_op);
+
+int
-+xencomm_mini_hypercall_perfmon_op(unsigned long cmd, void* arg,
-+ unsigned long count)
++swiotlb_dma_mapping_error(dma_addr_t dma_addr)
+{
-+ unsigned int argsize;
-+ struct xencomm_mini xc_area[2];
-+ int nbr_area = 2;
-+ struct xencomm_handle *desc;
-+ int rc;
++ return (dma_addr == virt_to_bus(io_tlb_overflow_buffer));
++}
+
-+ switch (cmd) {
-+ case PFM_GET_FEATURES:
-+ argsize = sizeof(pfarg_features_t);
-+ break;
-+ case PFM_CREATE_CONTEXT:
-+ argsize = sizeof(pfarg_context_t);
-+ break;
-+ case PFM_LOAD_CONTEXT:
-+ argsize = sizeof(pfarg_load_t);
-+ break;
-+ case PFM_WRITE_PMCS:
-+ case PFM_WRITE_PMDS:
-+ argsize = sizeof(pfarg_reg_t) * count;
-+ break;
++/*
++ * Return whether the given device DMA address mask can be supported
++ * properly. For example, if your device can only drive the low 24-bits
++ * during bus mastering, then you would pass 0x00ffffff as the mask to
++ * this function.
++ */
++int
++swiotlb_dma_supported (struct device *hwdev, u64 mask)
++{
++#ifdef CONFIG_XEN
++ return (virt_to_bus(io_tlb_end - 1)) <= mask;
++#else
++ return (virt_to_bus(io_tlb_end) - 1) <= mask;
++#endif
++}
+
-+ case PFM_DESTROY_CONTEXT:
-+ case PFM_UNLOAD_CONTEXT:
-+ case PFM_START:
-+ case PFM_STOP:
-+ return xencomm_arch_hypercall_perfmon_op(cmd, arg, count);
++EXPORT_SYMBOL(swiotlb_init);
++EXPORT_SYMBOL(swiotlb_map_single);
++EXPORT_SYMBOL(swiotlb_unmap_single);
++EXPORT_SYMBOL(swiotlb_map_sg);
++EXPORT_SYMBOL(swiotlb_unmap_sg);
++EXPORT_SYMBOL(swiotlb_sync_single_for_cpu);
++EXPORT_SYMBOL(swiotlb_sync_single_for_device);
++EXPORT_SYMBOL_GPL(swiotlb_sync_single_range_for_cpu);
++EXPORT_SYMBOL_GPL(swiotlb_sync_single_range_for_device);
++EXPORT_SYMBOL(swiotlb_sync_sg_for_cpu);
++EXPORT_SYMBOL(swiotlb_sync_sg_for_device);
++EXPORT_SYMBOL(swiotlb_dma_mapping_error);
++EXPORT_SYMBOL(swiotlb_alloc_coherent);
++EXPORT_SYMBOL(swiotlb_free_coherent);
++EXPORT_SYMBOL(swiotlb_dma_supported);
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/util.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/util.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,102 @@
++/******************************************************************************
++ * arch/ia64/xen/util.c
++ * This file is the ia64 counterpart of drivers/xen/util.c
++ *
++ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ */
+
-+ default:
-+ printk("%s:%d cmd %ld isn't supported\n",
-+ __func__, __LINE__, cmd);
-+ BUG();
-+ }
++#include <linux/mm.h>
++#include <linux/module.h>
++#include <linux/slab.h>
++#include <linux/vmalloc.h>
++#include <asm/uaccess.h>
++#include <xen/driver_util.h>
++#include <xen/interface/memory.h>
++#include <asm/hypercall.h>
+
-+ rc = xencomm_create_mini(xc_area, &nbr_area, arg, argsize, &desc);
-+ if (rc)
-+ return rc;
-+ return xencomm_arch_hypercall_perfmon_op(cmd, desc, count);
++struct vm_struct *alloc_vm_area(unsigned long size)
++{
++ int order;
++ unsigned long virt;
++ unsigned long nr_pages;
++ struct vm_struct *area;
++
++ order = get_order(size);
++ virt = __get_free_pages(GFP_KERNEL, order);
++ if (virt == 0)
++ goto err0;
++ nr_pages = 1 << order;
++ scrub_pages(virt, nr_pages);
++
++ area = kmalloc(sizeof(*area), GFP_KERNEL);
++ if (area == NULL)
++ goto err1;
++
++ area->flags = VM_IOREMAP; /* XXX */
++ area->addr = (void*)virt;
++ area->size = size;
++ area->pages = NULL; /* XXX */
++ area->nr_pages = nr_pages;
++ area->phys_addr = 0; /* xenbus_map_ring_valloc uses this field! */
++
++ return area;
++
++err1:
++ free_pages(virt, order);
++err0:
++ return NULL;
+}
-+EXPORT_SYMBOL_GPL(xencomm_mini_hypercall_perfmon_op);
++EXPORT_SYMBOL_GPL(alloc_vm_area);
+
-+int
-+xencomm_mini_hypercall_sched_op(int cmd, void *arg)
++void free_vm_area(struct vm_struct *area)
+{
-+ int rc, nbr_area = 2;
-+ struct xencomm_mini xc_area[2];
-+ struct xencomm_handle *desc;
-+ unsigned int argsize;
-+
-+ switch (cmd) {
-+ case SCHEDOP_yield:
-+ case SCHEDOP_block:
-+ argsize = 0;
-+ break;
-+ case SCHEDOP_shutdown:
-+ argsize = sizeof(sched_shutdown_t);
-+ break;
-+ case SCHEDOP_poll:
-+ argsize = sizeof(sched_poll_t);
-+ break;
-+ case SCHEDOP_remote_shutdown:
-+ argsize = sizeof(sched_remote_shutdown_t);
-+ break;
++ unsigned int order = get_order(area->size);
++ unsigned long i;
++ unsigned long phys_addr = __pa(area->addr);
+
-+ default:
-+ printk("%s: unknown sched op %d\n", __func__, cmd);
-+ return -ENOSYS;
++ /* This area is used for foreign page mappping.
++ * So underlying machine page may not be assigned. */
++ for (i = 0; i < (1 << order); i++) {
++ unsigned long ret;
++ unsigned long gpfn = (phys_addr >> PAGE_SHIFT) + i;
++ struct xen_memory_reservation reservation = {
++ .nr_extents = 1,
++ .address_bits = 0,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
++ set_xen_guest_handle(reservation.extent_start, &gpfn);
++ ret = HYPERVISOR_memory_op(XENMEM_populate_physmap,
++ &reservation);
++ BUG_ON(ret != 1);
+ }
++ free_pages((unsigned long)area->addr, order);
++ kfree(area);
++}
++EXPORT_SYMBOL_GPL(free_vm_area);
+
-+ rc = xencomm_create_mini(xc_area, &nbr_area, arg, argsize, &desc);
-+ if (rc)
-+ return rc;
++/*
++ * Local variables:
++ * c-file-style: "linux"
++ * indent-tabs-mode: t
++ * c-indent-level: 8
++ * c-basic-offset: 8
++ * tab-width: 8
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/xcom_asm.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/xcom_asm.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,27 @@
++/*
++ * xencomm suspend support
++ * Support routines for Xen
++ *
++ * Copyright (C) 2005 Dan Magenheimer <dan.magenheimer at hp.com>
++ */
++#include <asm/asmmacro.h>
++#include <xen/interface/xen.h>
+
-+ return xencomm_arch_hypercall_sched_op(cmd, desc);
-+}
-+EXPORT_SYMBOL_GPL(xencomm_mini_hypercall_sched_op);
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/xcom_privcmd.c tmp-linux-2.6-xen.patch/arch/ia64/xen/xcom_privcmd.c
---- pristine-linux-2.6.18.2/arch/ia64/xen/xcom_privcmd.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/xcom_privcmd.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,663 @@
++/*
++ * Stub for suspend.
++ * Just force the stacked registers to be written in memory.
++ */
++GLOBAL_ENTRY(xencomm_arch_hypercall_suspend)
++ ;;
++ alloc r20=ar.pfs,0,0,6,0
++ mov r2=__HYPERVISOR_sched_op
++ ;;
++ /* We don't want to deal with RSE. */
++ flushrs
++ mov r33=r32
++ mov r32=2 // SCHEDOP_shutdown
++ ;;
++ break 0x1000
++ ;;
++ br.ret.sptk.many b0
++END(xencomm_arch_hypercall_suspend)
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/xcom_hcall.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/xcom_hcall.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,671 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
@@ -26879,8 +29877,12 @@
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ *
-+ * Authors: Hollis Blanchard <hollisb at us.ibm.com>
+ * Tristan Gingold <tristan.gingold at bull.net>
++ *
++ * Copyright (c) 2007
++ * Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ * consolidate mini and inline version.
+ */
+#include <linux/types.h>
+#include <linux/errno.h>
@@ -26889,989 +29891,1939 @@
+#include <linux/module.h>
+#include <xen/interface/xen.h>
+#include <xen/interface/platform.h>
-+#define __XEN__
-+#include <xen/interface/domctl.h>
-+#include <xen/interface/sysctl.h>
+#include <xen/interface/memory.h>
++#include <xen/interface/xencomm.h>
+#include <xen/interface/version.h>
++#include <xen/interface/sched.h>
+#include <xen/interface/event_channel.h>
-+#include <xen/interface/acm_ops.h>
++#include <xen/interface/physdev.h>
++#include <xen/interface/grant_table.h>
++#include <xen/interface/callback.h>
++#include <xen/interface/xsm/acm_ops.h>
+#include <xen/interface/hvm/params.h>
-+#include <xen/public/privcmd.h>
-+#include <asm/hypercall.h>
++#include <xen/interface/xenoprof.h>
++#include <xen/interface/vcpu.h>
++#include <xen/interface/kexec.h>
++#include <asm/hypervisor.h>
+#include <asm/page.h>
+#include <asm/uaccess.h>
+#include <asm/xen/xencomm.h>
++#include <asm/perfmon.h>
+
-+#define ROUND_DIV(v,s) (((v) + (s) - 1) / (s))
++/* Xencomm notes:
++ * This file defines hypercalls to be used by xencomm. The hypercalls simply
++ * create inlines or mini descriptors for pointers and then call the raw arch
++ * hypercall xencomm_arch_hypercall_XXX
++ *
++ * If the arch wants to directly use these hypercalls, simply define macros
++ * in asm/hypercall.h, eg:
++ * #define HYPERVISOR_sched_op xencomm_hypercall_sched_op
++ *
++ * The arch may also define HYPERVISOR_xxx as a function and do more operations
++ * before/after doing the hypercall.
++ *
++ * Note: because only inline or mini descriptors are created these functions
++ * must only be called with in kernel memory parameters.
++ */
+
-+static int
-+xencomm_privcmd_platform_op(privcmd_hypercall_t *hypercall)
++int
++xencomm_hypercall_console_io(int cmd, int count, char *str)
+{
-+ struct xen_platform_op kern_op;
-+ struct xen_platform_op __user *user_op = (struct xen_platform_op __user *)hypercall->arg[0];
-+ struct xencomm_handle *op_desc;
-+ struct xencomm_handle *desc = NULL;
-+ int ret = 0;
++ return xencomm_arch_hypercall_console_io
++ (cmd, count, xencomm_map_no_alloc(str, count));
++}
++EXPORT_SYMBOL_GPL(xencomm_hypercall_console_io);
+
-+ if (copy_from_user(&kern_op, user_op, sizeof(struct xen_platform_op)))
-+ return -EFAULT;
++int
++xencomm_hypercall_event_channel_op(int cmd, void *op)
++{
++ struct xencomm_handle *desc;
++ desc = xencomm_map_no_alloc(op, sizeof(evtchn_op_t));
++ if (desc == NULL)
++ return -EINVAL;
+
-+ if (kern_op.interface_version != XENPF_INTERFACE_VERSION)
-+ return -EACCES;
++ return xencomm_arch_hypercall_event_channel_op(cmd, desc);
++}
++EXPORT_SYMBOL_GPL(xencomm_hypercall_event_channel_op);
+
-+ op_desc = xencomm_create_inline(&kern_op);
++int
++xencomm_hypercall_xen_version(int cmd, void *arg)
++{
++ struct xencomm_handle *desc;
++ unsigned int argsize;
++
++ switch (cmd) {
++ case XENVER_version:
++ /* do not actually pass an argument */
++ return xencomm_arch_hypercall_xen_version(cmd, 0);
++ case XENVER_extraversion:
++ argsize = sizeof(xen_extraversion_t);
++ break;
++ case XENVER_compile_info:
++ argsize = sizeof(xen_compile_info_t);
++ break;
++ case XENVER_capabilities:
++ argsize = sizeof(xen_capabilities_info_t);
++ break;
++ case XENVER_changeset:
++ argsize = sizeof(xen_changeset_info_t);
++ break;
++ case XENVER_platform_parameters:
++ argsize = sizeof(xen_platform_parameters_t);
++ break;
++ case XENVER_pagesize:
++ argsize = (arg == NULL) ? 0 : sizeof(void *);
++ break;
++ case XENVER_get_features:
++ argsize = (arg == NULL) ? 0 : sizeof(xen_feature_info_t);
++ break;
+
-+ switch (kern_op.cmd) {
+ default:
-+ printk("%s: unknown platform cmd %d\n", __func__, kern_op.cmd);
++ printk("%s: unknown version op %d\n", __func__, cmd);
+ return -ENOSYS;
+ }
+
-+ if (ret) {
-+ /* error mapping the nested pointer */
-+ return ret;
-+ }
++ desc = xencomm_map_no_alloc(arg, argsize);
++ if (desc == NULL)
++ return -EINVAL;
+
-+ ret = xencomm_arch_hypercall_platform_op(op_desc);
++ return xencomm_arch_hypercall_xen_version(cmd, desc);
++}
++EXPORT_SYMBOL_GPL(xencomm_hypercall_xen_version);
+
-+ /* FIXME: should we restore the handle? */
-+ if (copy_to_user(user_op, &kern_op, sizeof(struct xen_platform_op)))
-+ ret = -EFAULT;
++int
++xencomm_hypercall_physdev_op(int cmd, void *op)
++{
++ unsigned int argsize;
++
++ switch (cmd) {
++ case PHYSDEVOP_apic_read:
++ case PHYSDEVOP_apic_write:
++ argsize = sizeof(physdev_apic_t);
++ break;
++ case PHYSDEVOP_alloc_irq_vector:
++ case PHYSDEVOP_free_irq_vector:
++ argsize = sizeof(physdev_irq_t);
++ break;
++ case PHYSDEVOP_irq_status_query:
++ argsize = sizeof(physdev_irq_status_query_t);
++ break;
++ case PHYSDEVOP_manage_pci_add:
++ case PHYSDEVOP_manage_pci_remove:
++ argsize = sizeof(physdev_manage_pci_t);
++ break;
++ case PHYSDEVOP_map_pirq:
++ argsize = sizeof(physdev_map_pirq_t);
++ break;
++ case PHYSDEVOP_unmap_pirq:
++ argsize = sizeof(physdev_unmap_pirq_t);
++ break;
+
-+ if (desc)
-+ xencomm_free(desc);
-+ return ret;
++ default:
++ printk("%s: unknown physdev op %d\n", __func__, cmd);
++ return -ENOSYS;
++ }
++
++ return xencomm_arch_hypercall_physdev_op
++ (cmd, xencomm_map_no_alloc(op, argsize));
+}
+
-+/*
-+ * Temporarily disable the NUMA PHYSINFO code until the rest of the
-+ * changes are upstream.
-+ */
-+#undef IA64_NUMA_PHYSINFO
-+
+static int
-+xencomm_privcmd_sysctl(privcmd_hypercall_t *hypercall)
++xencommize_grant_table_op(struct xencomm_mini **xc_area,
++ unsigned int cmd, void *op, unsigned int count,
++ struct xencomm_handle **desc)
+{
-+ xen_sysctl_t kern_op;
-+ xen_sysctl_t __user *user_op;
-+ struct xencomm_handle *op_desc;
-+ struct xencomm_handle *desc = NULL;
-+ struct xencomm_handle *desc1 = NULL;
-+ int ret = 0;
-+
-+ user_op = (xen_sysctl_t __user *)hypercall->arg[0];
-+
-+ if (copy_from_user(&kern_op, user_op, sizeof(xen_sysctl_t)))
-+ return -EFAULT;
-+
-+ if (kern_op.interface_version != XEN_SYSCTL_INTERFACE_VERSION)
-+ return -EACCES;
-+
-+ op_desc = xencomm_create_inline(&kern_op);
++ struct xencomm_handle *desc1;
++ unsigned int argsize;
+
-+ switch (kern_op.cmd) {
-+ case XEN_SYSCTL_readconsole:
-+ ret = xencomm_create(
-+ xen_guest_handle(kern_op.u.readconsole.buffer),
-+ kern_op.u.readconsole.count,
-+ &desc, GFP_KERNEL);
-+ set_xen_guest_handle(kern_op.u.readconsole.buffer,
-+ (void *)desc);
++ switch (cmd) {
++ case GNTTABOP_map_grant_ref:
++ argsize = sizeof(struct gnttab_map_grant_ref);
+ break;
-+ case XEN_SYSCTL_tbuf_op:
-+#ifndef IA64_NUMA_PHYSINFO
-+ case XEN_SYSCTL_physinfo:
-+#endif
-+ case XEN_SYSCTL_sched_id:
++ case GNTTABOP_unmap_grant_ref:
++ argsize = sizeof(struct gnttab_unmap_grant_ref);
+ break;
-+ case XEN_SYSCTL_perfc_op:
++ case GNTTABOP_unmap_and_replace:
++ argsize = sizeof(struct gnttab_unmap_and_replace);
++ break;
++ case GNTTABOP_setup_table:
+ {
-+ struct xencomm_handle *tmp_desc;
-+ xen_sysctl_t tmp_op = {
-+ .cmd = XEN_SYSCTL_perfc_op,
-+ .interface_version = XEN_SYSCTL_INTERFACE_VERSION,
-+ .u.perfc_op = {
-+ .cmd = XEN_SYSCTL_PERFCOP_query,
-+ // .desc.p = NULL,
-+ // .val.p = NULL,
-+ },
-+ };
-+
-+ if (xen_guest_handle(kern_op.u.perfc_op.desc) == NULL) {
-+ if (xen_guest_handle(kern_op.u.perfc_op.val) != NULL)
-+ return -EINVAL;
-+ break;
-+ }
-+
-+ /* query the buffer size for xencomm */
-+ tmp_desc = xencomm_create_inline(&tmp_op);
-+ ret = xencomm_arch_hypercall_sysctl(tmp_desc);
-+ if (ret)
-+ return ret;
-+
-+ ret = xencomm_create(xen_guest_handle(kern_op.u.perfc_op.desc),
-+ tmp_op.u.perfc_op.nr_counters *
-+ sizeof(xen_sysctl_perfc_desc_t),
-+ &desc, GFP_KERNEL);
-+ if (ret)
-+ return ret;
-+
-+ set_xen_guest_handle(kern_op.u.perfc_op.desc, (void *)desc);
++ struct gnttab_setup_table *setup = op;
+
-+ ret = xencomm_create(xen_guest_handle(kern_op.u.perfc_op.val),
-+ tmp_op.u.perfc_op.nr_vals *
-+ sizeof(xen_sysctl_perfc_val_t),
-+ &desc1, GFP_KERNEL);
-+ if (ret)
-+ xencomm_free(desc);
++ argsize = sizeof(*setup);
+
-+ set_xen_guest_handle(kern_op.u.perfc_op.val, (void *)desc1);
++ if (count != 1)
++ return -EINVAL;
++ desc1 = __xencomm_map_no_alloc
++ (xen_guest_handle(setup->frame_list),
++ setup->nr_frames *
++ sizeof(*xen_guest_handle(setup->frame_list)),
++ *xc_area);
++ if (desc1 == NULL)
++ return -EINVAL;
++ (*xc_area)++;
++ set_xen_guest_handle(setup->frame_list, (void *)desc1);
+ break;
+ }
-+ case XEN_SYSCTL_getdomaininfolist:
-+ ret = xencomm_create(
-+ xen_guest_handle(kern_op.u.getdomaininfolist.buffer),
-+ kern_op.u.getdomaininfolist.max_domains *
-+ sizeof(xen_domctl_getdomaininfo_t),
-+ &desc, GFP_KERNEL);
-+ set_xen_guest_handle(kern_op.u.getdomaininfolist.buffer,
-+ (void *)desc);
++ case GNTTABOP_dump_table:
++ argsize = sizeof(struct gnttab_dump_table);
+ break;
-+#ifdef IA64_NUMA_PHYSINFO
-+ case XEN_SYSCTL_physinfo:
-+ ret = xencomm_create(
-+ xen_guest_handle(kern_op.u.physinfo.memory_chunks),
-+ PUBLIC_MAXCHUNKS * sizeof(node_data_t),
-+ &desc, GFP_KERNEL);
-+ if (ret)
-+ return ret;
-+ set_xen_guest_handle(kern_op.u.physinfo.memory_chunks,
-+ (void *)desc);
-+
-+ ret = xencomm_create(
-+ xen_guest_handle(kern_op.u.physinfo.cpu_to_node),
-+ PUBLIC_MAX_NUMNODES * sizeof(u64),
-+ &desc1, GFP_KERNEL);
-+ if (ret)
-+ xencomm_free(desc);
-+ set_xen_guest_handle(kern_op.u.physinfo.cpu_to_node,
-+ (void *)desc1);
++ case GNTTABOP_transfer:
++ argsize = sizeof(struct gnttab_transfer);
++ break;
++ case GNTTABOP_copy:
++ argsize = sizeof(struct gnttab_copy);
++ break;
++ case GNTTABOP_query_size:
++ argsize = sizeof(struct gnttab_query_size);
+ break;
-+#endif
+ default:
-+ printk("%s: unknown sysctl cmd %d\n", __func__, kern_op.cmd);
-+ return -ENOSYS;
-+ }
-+
-+ if (ret) {
-+ /* error mapping the nested pointer */
-+ return ret;
++ printk("%s: unknown hypercall grant table op %d\n",
++ __func__, cmd);
++ BUG();
+ }
+
-+ ret = xencomm_arch_hypercall_sysctl(op_desc);
-+
-+ /* FIXME: should we restore the handles? */
-+ if (copy_to_user(user_op, &kern_op, sizeof(xen_sysctl_t)))
-+ ret = -EFAULT;
++ *desc = __xencomm_map_no_alloc(op, count * argsize, *xc_area);
++ if (*desc == NULL)
++ return -EINVAL;
++ (*xc_area)++;
+
-+ if (desc)
-+ xencomm_free(desc);
-+ if (desc1)
-+ xencomm_free(desc1);
-+ return ret;
++ return 0;
+}
+
-+static int
-+xencomm_privcmd_domctl(privcmd_hypercall_t *hypercall)
++int
++xencomm_hypercall_grant_table_op(unsigned int cmd, void *op,
++ unsigned int count)
+{
-+ xen_domctl_t kern_op;
-+ xen_domctl_t __user *user_op;
-+ struct xencomm_handle *op_desc;
-+ struct xencomm_handle *desc = NULL;
-+ int ret = 0;
-+
-+ user_op = (xen_domctl_t __user *)hypercall->arg[0];
++ int rc;
++ struct xencomm_handle *desc;
++ XENCOMM_MINI_ALIGNED(xc_area, 2);
+
-+ if (copy_from_user(&kern_op, user_op, sizeof(xen_domctl_t)))
-+ return -EFAULT;
++ rc = xencommize_grant_table_op(&xc_area, cmd, op, count, &desc);
++ if (rc)
++ return rc;
+
-+ if (kern_op.interface_version != XEN_DOMCTL_INTERFACE_VERSION)
-+ return -EACCES;
++ return xencomm_arch_hypercall_grant_table_op(cmd, desc, count);
++}
++EXPORT_SYMBOL_GPL(xencomm_hypercall_grant_table_op);
+
-+ op_desc = xencomm_create_inline(&kern_op);
++int
++xencomm_hypercall_sched_op(int cmd, void *arg)
++{
++ struct xencomm_handle *desc;
++ unsigned int argsize;
+
-+ switch (kern_op.cmd) {
-+ case XEN_DOMCTL_createdomain:
-+ case XEN_DOMCTL_destroydomain:
-+ case XEN_DOMCTL_pausedomain:
-+ case XEN_DOMCTL_unpausedomain:
-+ case XEN_DOMCTL_getdomaininfo:
++ switch (cmd) {
++ case SCHEDOP_yield:
++ case SCHEDOP_block:
++ argsize = 0;
+ break;
-+ case XEN_DOMCTL_getmemlist:
++ case SCHEDOP_shutdown:
++ argsize = sizeof(sched_shutdown_t);
++ break;
++ case SCHEDOP_remote_shutdown:
++ argsize = sizeof(sched_remote_shutdown_t);
++ break;
++ case SCHEDOP_poll:
+ {
-+ unsigned long nr_pages = kern_op.u.getmemlist.max_pfns;
++ sched_poll_t *poll = arg;
++ struct xencomm_handle *ports;
+
-+ ret = xencomm_create(
-+ xen_guest_handle(kern_op.u.getmemlist.buffer),
-+ nr_pages * sizeof(unsigned long),
-+ &desc, GFP_KERNEL);
-+ set_xen_guest_handle(kern_op.u.getmemlist.buffer,
-+ (void *)desc);
++ argsize = sizeof(sched_poll_t);
++ ports = xencomm_map_no_alloc(xen_guest_handle(poll->ports),
++ sizeof(*xen_guest_handle(poll->ports)));
++
++ set_xen_guest_handle(poll->ports, (void *)ports);
+ break;
+ }
-+ case XEN_DOMCTL_getpageframeinfo:
-+ break;
-+ case XEN_DOMCTL_getpageframeinfo2:
-+ ret = xencomm_create(
-+ xen_guest_handle(kern_op.u.getpageframeinfo2.array),
-+ kern_op.u.getpageframeinfo2.num,
-+ &desc, GFP_KERNEL);
-+ set_xen_guest_handle(kern_op.u.getpageframeinfo2.array,
-+ (void *)desc);
-+ break;
-+ case XEN_DOMCTL_shadow_op:
-+ ret = xencomm_create(
-+ xen_guest_handle(kern_op.u.shadow_op.dirty_bitmap),
-+ ROUND_DIV(kern_op.u.shadow_op.pages, 8),
-+ &desc, GFP_KERNEL);
-+ set_xen_guest_handle(kern_op.u.shadow_op.dirty_bitmap,
-+ (void *)desc);
-+ break;
-+ case XEN_DOMCTL_max_mem:
-+ break;
-+ case XEN_DOMCTL_setvcpucontext:
-+ case XEN_DOMCTL_getvcpucontext:
-+ ret = xencomm_create(
-+ xen_guest_handle(kern_op.u.vcpucontext.ctxt),
-+ sizeof(vcpu_guest_context_t),
-+ &desc, GFP_KERNEL);
-+ set_xen_guest_handle(kern_op.u.vcpucontext.ctxt, (void *)desc);
-+ break;
-+ case XEN_DOMCTL_getvcpuinfo:
-+ break;
-+ case XEN_DOMCTL_setvcpuaffinity:
-+ case XEN_DOMCTL_getvcpuaffinity:
-+ ret = xencomm_create(
-+ xen_guest_handle(kern_op.u.vcpuaffinity.cpumap.bitmap),
-+ ROUND_DIV(kern_op.u.vcpuaffinity.cpumap.nr_cpus, 8),
-+ &desc, GFP_KERNEL);
-+ set_xen_guest_handle(kern_op.u.vcpuaffinity.cpumap.bitmap,
-+ (void *)desc);
-+ break;
-+ case XEN_DOMCTL_max_vcpus:
-+ case XEN_DOMCTL_scheduler_op:
-+ case XEN_DOMCTL_setdomainhandle:
-+ case XEN_DOMCTL_setdebugging:
-+ case XEN_DOMCTL_irq_permission:
-+ case XEN_DOMCTL_iomem_permission:
-+ case XEN_DOMCTL_ioport_permission:
-+ case XEN_DOMCTL_hypercall_init:
-+ case XEN_DOMCTL_arch_setup:
-+ case XEN_DOMCTL_settimeoffset:
-+ case XEN_DOMCTL_sendtrigger:
-+ break;
+ default:
-+ printk("%s: unknown domctl cmd %d\n", __func__, kern_op.cmd);
++ printk("%s: unknown sched op %d\n", __func__, cmd);
+ return -ENOSYS;
+ }
++
++ desc = xencomm_map_no_alloc(arg, argsize);
++ if (desc == NULL)
++ return -EINVAL;
+
-+ if (ret) {
-+ /* error mapping the nested pointer */
-+ return ret;
-+ }
-+
-+ ret = xencomm_arch_hypercall_domctl (op_desc);
++ return xencomm_arch_hypercall_sched_op(cmd, desc);
++}
++EXPORT_SYMBOL_GPL(xencomm_hypercall_sched_op);
+
-+ /* FIXME: should we restore the handle? */
-+ if (copy_to_user(user_op, &kern_op, sizeof(xen_domctl_t)))
-+ ret = -EFAULT;
-+
-+ if (desc)
-+ xencomm_free(desc);
-+ return ret;
-+}
-+
-+static int
-+xencomm_privcmd_acm_op(privcmd_hypercall_t *hypercall)
++int
++xencomm_hypercall_multicall(void *call_list, int nr_calls)
+{
-+ void __user *arg = (void __user *)hypercall->arg[0];
-+ xen_acmctl_t kern_arg;
-+ struct xencomm_handle *op_desc;
-+ struct xencomm_handle *desc = NULL;
-+ int ret;
-+
-+ if (copy_from_user(&kern_arg, arg, sizeof(kern_arg)))
-+ return -EFAULT;
-+ if (kern_arg.interface_version != ACM_INTERFACE_VERSION)
-+ return -ENOSYS;
-+
-+ switch (kern_arg.cmd) {
-+ case ACMOP_getssid: {
-+ op_desc = xencomm_create_inline(&kern_arg);
++ int rc;
++ int i;
++ multicall_entry_t *mce;
++ struct xencomm_handle *desc;
++ XENCOMM_MINI_ALIGNED(xc_area, nr_calls * 2);
+
-+ ret = xencomm_create(
-+ xen_guest_handle(kern_arg.u.getssid.ssidbuf),
-+ kern_arg.u.getssid.ssidbuf_size, &desc, GFP_KERNEL);
-+ if (ret)
-+ return ret;
++ for (i = 0; i < nr_calls; i++) {
++ mce = (multicall_entry_t *)call_list + i;
+
-+ set_xen_guest_handle(kern_arg.u.getssid.ssidbuf, (void *)desc);
++ switch (mce->op) {
++ case __HYPERVISOR_update_va_mapping:
++ case __HYPERVISOR_mmu_update:
++ /* No-op on ia64. */
++ break;
++ case __HYPERVISOR_grant_table_op:
++ rc = xencommize_grant_table_op
++ (&xc_area,
++ mce->args[0], (void *)mce->args[1],
++ mce->args[2], &desc);
++ if (rc)
++ return rc;
++ mce->args[1] = (unsigned long)desc;
++ break;
++ case __HYPERVISOR_memory_op:
++ default:
++ printk("%s: unhandled multicall op entry op %lu\n",
++ __func__, mce->op);
++ return -ENOSYS;
++ }
++ }
+
-+ ret = xencomm_arch_hypercall_acm_op(op_desc);
++ desc = xencomm_map_no_alloc(call_list,
++ nr_calls * sizeof(multicall_entry_t));
++ if (desc == NULL)
++ return -EINVAL;
+
-+ xencomm_free(desc);
++ return xencomm_arch_hypercall_multicall(desc, nr_calls);
++}
++EXPORT_SYMBOL_GPL(xencomm_hypercall_multicall);
+
-+ if (copy_to_user(arg, &kern_arg, sizeof(kern_arg)))
-+ return -EFAULT;
-+ return ret;
-+ }
++int
++xencomm_hypercall_callback_op(int cmd, void *arg)
++{
++ unsigned int argsize;
++ switch (cmd)
++ {
++ case CALLBACKOP_register:
++ argsize = sizeof(struct callback_register);
++ break;
++ case CALLBACKOP_unregister:
++ argsize = sizeof(struct callback_unregister);
++ break;
+ default:
-+ printk("%s: unknown acm_op cmd %d\n", __func__, kern_arg.cmd);
++ printk("%s: unknown callback op %d\n", __func__, cmd);
+ return -ENOSYS;
+ }
+
-+ return ret;
++ return xencomm_arch_hypercall_callback_op
++ (cmd, xencomm_map_no_alloc(arg, argsize));
+}
+
+static int
-+xencomm_privcmd_memory_op(privcmd_hypercall_t *hypercall)
++xencommize_memory_reservation(struct xencomm_mini *xc_area,
++ xen_memory_reservation_t *mop)
+{
-+ const unsigned long cmd = hypercall->arg[0];
-+ int ret = 0;
++ struct xencomm_handle *desc;
++
++ desc = __xencomm_map_no_alloc(xen_guest_handle(mop->extent_start),
++ mop->nr_extents *
++ sizeof(*xen_guest_handle(mop->extent_start)),
++ xc_area);
++ if (desc == NULL)
++ return -EINVAL;
++
++ set_xen_guest_handle(mop->extent_start, (void *)desc);
++ return 0;
++}
++
++int
++xencomm_hypercall_memory_op(unsigned int cmd, void *arg)
++{
++ XEN_GUEST_HANDLE(xen_pfn_t) extent_start_va[2];
++ xen_memory_reservation_t *xmr = NULL, *xme_in = NULL, *xme_out = NULL;
++ xen_memory_map_t *memmap = NULL;
++ XEN_GUEST_HANDLE(void) buffer;
++ int rc;
++ struct xencomm_handle *desc;
++ unsigned int argsize;
++ XENCOMM_MINI_ALIGNED(xc_area, 2);
+
+ switch (cmd) {
+ case XENMEM_increase_reservation:
+ case XENMEM_decrease_reservation:
+ case XENMEM_populate_physmap:
-+ {
-+ xen_memory_reservation_t kern_op;
-+ xen_memory_reservation_t __user *user_op;
-+ struct xencomm_handle *desc = NULL;
-+ struct xencomm_handle *desc_op;
++ xmr = (xen_memory_reservation_t *)arg;
++ set_xen_guest_handle(extent_start_va[0],
++ xen_guest_handle(xmr->extent_start));
+
-+ user_op = (xen_memory_reservation_t __user *)hypercall->arg[1];
-+ if (copy_from_user(&kern_op, user_op,
-+ sizeof(xen_memory_reservation_t)))
-+ return -EFAULT;
-+ desc_op = xencomm_create_inline(&kern_op);
++ argsize = sizeof(*xmr);
++ rc = xencommize_memory_reservation(xc_area, xmr);
++ if (rc)
++ return rc;
++ xc_area++;
++ break;
+
-+ if (xen_guest_handle(kern_op.extent_start)) {
-+ void * addr;
++ case XENMEM_maximum_gpfn:
++ argsize = 0;
++ break;
+
-+ addr = xen_guest_handle(kern_op.extent_start);
-+ ret = xencomm_create
-+ (addr,
-+ kern_op.nr_extents *
-+ sizeof(*xen_guest_handle
-+ (kern_op.extent_start)),
-+ &desc, GFP_KERNEL);
-+ if (ret)
-+ return ret;
-+ set_xen_guest_handle(kern_op.extent_start,
-+ (void *)desc);
-+ }
++ case XENMEM_maximum_ram_page:
++ argsize = 0;
++ break;
+
-+ ret = xencomm_arch_hypercall_memory_op(cmd, desc_op);
++ case XENMEM_exchange:
++ xme_in = &((xen_memory_exchange_t *)arg)->in;
++ xme_out = &((xen_memory_exchange_t *)arg)->out;
++ set_xen_guest_handle(extent_start_va[0],
++ xen_guest_handle(xme_in->extent_start));
++ set_xen_guest_handle(extent_start_va[1],
++ xen_guest_handle(xme_out->extent_start));
+
-+ if (desc)
-+ xencomm_free(desc);
++ argsize = sizeof(xen_memory_exchange_t);
++ rc = xencommize_memory_reservation(xc_area, xme_in);
++ if (rc)
++ return rc;
++ xc_area++;
++ rc = xencommize_memory_reservation(xc_area, xme_out);
++ if (rc)
++ return rc;
++ xc_area++;
++ break;
+
-+ if (ret != 0)
-+ return ret;
++ case XENMEM_add_to_physmap:
++ argsize = sizeof(xen_add_to_physmap_t);
++ break;
+
-+ if (copy_to_user(user_op, &kern_op,
-+ sizeof(xen_memory_reservation_t)))
-+ return -EFAULT;
++ case XENMEM_machine_memory_map:
++ argsize = sizeof(*memmap);
++ memmap = (xen_memory_map_t *)arg;
++ set_xen_guest_handle(buffer, xen_guest_handle(memmap->buffer));
++ desc = xencomm_map_no_alloc(xen_guest_handle(memmap->buffer),
++ memmap->nr_entries);
++ if (desc == NULL)
++ return -EINVAL;
++ set_xen_guest_handle(memmap->buffer, (void *)desc);
++ break;
+
-+ return ret;
++ default:
++ printk("%s: unknown memory op %d\n", __func__, cmd);
++ return -ENOSYS;
+ }
-+ case XENMEM_translate_gpfn_list:
-+ {
-+ xen_translate_gpfn_list_t kern_op;
-+ xen_translate_gpfn_list_t __user *user_op;
-+ struct xencomm_handle *desc_gpfn = NULL;
-+ struct xencomm_handle *desc_mfn = NULL;
-+ struct xencomm_handle *desc_op;
-+ void *addr;
-+
-+ user_op = (xen_translate_gpfn_list_t __user *)
-+ hypercall->arg[1];
-+ if (copy_from_user(&kern_op, user_op,
-+ sizeof(xen_translate_gpfn_list_t)))
-+ return -EFAULT;
-+ desc_op = xencomm_create_inline(&kern_op);
-+
-+ if (kern_op.nr_gpfns) {
-+ /* gpfn_list. */
-+ addr = xen_guest_handle(kern_op.gpfn_list);
-+
-+ ret = xencomm_create(addr, kern_op.nr_gpfns *
-+ sizeof(*xen_guest_handle
-+ (kern_op.gpfn_list)),
-+ &desc_gpfn, GFP_KERNEL);
-+ if (ret)
-+ return ret;
-+ set_xen_guest_handle(kern_op.gpfn_list,
-+ (void *)desc_gpfn);
-+
-+ /* mfn_list. */
-+ addr = xen_guest_handle(kern_op.mfn_list);
+
-+ ret = xencomm_create(addr, kern_op.nr_gpfns *
-+ sizeof(*xen_guest_handle
-+ (kern_op.mfn_list)),
-+ &desc_mfn, GFP_KERNEL);
-+ if (ret)
-+ return ret;
-+ set_xen_guest_handle(kern_op.mfn_list,
-+ (void *)desc_mfn);
-+ }
-+
-+ ret = xencomm_arch_hypercall_memory_op(cmd, desc_op);
++ desc = xencomm_map_no_alloc(arg, argsize);
++ if (desc == NULL)
++ return -EINVAL;
+
-+ if (desc_gpfn)
-+ xencomm_free(desc_gpfn);
++ rc = xencomm_arch_hypercall_memory_op(cmd, desc);
+
-+ if (desc_mfn)
-+ xencomm_free(desc_mfn);
++ switch (cmd) {
++ case XENMEM_increase_reservation:
++ case XENMEM_decrease_reservation:
++ case XENMEM_populate_physmap:
++ set_xen_guest_handle(xmr->extent_start,
++ xen_guest_handle(extent_start_va[0]));
++ break;
+
-+ if (ret != 0)
-+ return ret;
++ case XENMEM_exchange:
++ set_xen_guest_handle(xme_in->extent_start,
++ xen_guest_handle(extent_start_va[0]));
++ set_xen_guest_handle(xme_out->extent_start,
++ xen_guest_handle(extent_start_va[1]));
++ break;
+
-+ return ret;
-+ }
-+ default:
-+ printk("%s: unknown memory op %lu\n", __func__, cmd);
-+ ret = -ENOSYS;
++ case XENMEM_machine_memory_map:
++ set_xen_guest_handle(memmap->buffer, xen_guest_handle(buffer));
++ break;
+ }
-+ return ret;
++
++ return rc;
+}
++EXPORT_SYMBOL_GPL(xencomm_hypercall_memory_op);
+
-+static int
-+xencomm_privcmd_xen_version(privcmd_hypercall_t *hypercall)
++unsigned long
++xencomm_hypercall_hvm_op(int cmd, void *arg)
+{
-+ int cmd = hypercall->arg[0];
-+ void __user *arg = (void __user *)hypercall->arg[1];
+ struct xencomm_handle *desc;
-+ size_t argsize;
-+ int rc;
++ unsigned int argsize;
+
+ switch (cmd) {
-+ case XENVER_version:
-+ /* do not actually pass an argument */
-+ return xencomm_arch_hypercall_xen_version(cmd, 0);
-+ case XENVER_extraversion:
-+ argsize = sizeof(xen_extraversion_t);
-+ break;
-+ case XENVER_compile_info:
-+ argsize = sizeof(xen_compile_info_t);
-+ break;
-+ case XENVER_capabilities:
-+ argsize = sizeof(xen_capabilities_info_t);
-+ break;
-+ case XENVER_changeset:
-+ argsize = sizeof(xen_changeset_info_t);
-+ break;
-+ case XENVER_platform_parameters:
-+ argsize = sizeof(xen_platform_parameters_t);
-+ break;
-+ case XENVER_pagesize:
-+ argsize = (arg == NULL) ? 0 : sizeof(void *);
-+ break;
-+ case XENVER_get_features:
-+ argsize = (arg == NULL) ? 0 : sizeof(xen_feature_info_t);
++ case HVMOP_get_param:
++ case HVMOP_set_param:
++ argsize = sizeof(xen_hvm_param_t);
+ break;
-+
+ default:
-+ printk("%s: unknown version op %d\n", __func__, cmd);
-+ return -ENOSYS;
++ printk("%s: unknown HVMOP %d\n", __func__, cmd);
++ return -EINVAL;
+ }
+
-+ rc = xencomm_create(arg, argsize, &desc, GFP_KERNEL);
-+ if (rc)
-+ return rc;
++ desc = xencomm_map_no_alloc(arg, argsize);
++ if (desc == NULL)
++ return -EINVAL;
+
-+ rc = xencomm_arch_hypercall_xen_version(cmd, desc);
++ return xencomm_arch_hypercall_hvm_op(cmd, desc);
++}
++EXPORT_SYMBOL_GPL(xencomm_hypercall_hvm_op);
+
-+ xencomm_free(desc);
++int
++xencomm_hypercall_suspend(unsigned long srec)
++{
++ struct sched_shutdown arg;
+
-+ return rc;
++ arg.reason = SHUTDOWN_suspend;
++
++ return xencomm_arch_hypercall_suspend(
++ xencomm_map_no_alloc(&arg, sizeof(arg)));
+}
+
-+static int
-+xencomm_privcmd_event_channel_op(privcmd_hypercall_t *hypercall)
++int
++xencomm_hypercall_xenoprof_op(int op, void *arg)
+{
-+ int cmd = hypercall->arg[0];
-+ struct xencomm_handle *desc;
+ unsigned int argsize;
-+ int ret;
++ struct xencomm_handle *desc;
+
-+ switch (cmd) {
-+ case EVTCHNOP_alloc_unbound:
-+ argsize = sizeof(evtchn_alloc_unbound_t);
++ switch (op) {
++ case XENOPROF_init:
++ argsize = sizeof(xenoprof_init_t);
+ break;
-+
-+ case EVTCHNOP_status:
-+ argsize = sizeof(evtchn_status_t);
++ case XENOPROF_set_active:
++ argsize = sizeof(domid_t);
++ break;
++ case XENOPROF_set_passive:
++ argsize = sizeof(xenoprof_passive_t);
+ break;
++ case XENOPROF_counter:
++ argsize = sizeof(xenoprof_counter_t);
++ break;
++ case XENOPROF_get_buffer:
++ argsize = sizeof(xenoprof_get_buffer_t);
++ break;
++
++ case XENOPROF_reset_active_list:
++ case XENOPROF_reset_passive_list:
++ case XENOPROF_reserve_counters:
++ case XENOPROF_setup_events:
++ case XENOPROF_enable_virq:
++ case XENOPROF_start:
++ case XENOPROF_stop:
++ case XENOPROF_disable_virq:
++ case XENOPROF_release_counters:
++ case XENOPROF_shutdown:
++ return xencomm_arch_hypercall_xenoprof_op(op, arg);
+
+ default:
-+ printk("%s: unknown EVTCHNOP %d\n", __func__, cmd);
-+ return -EINVAL;
++ printk("%s: op %d isn't supported\n", __func__, op);
++ return -ENOSYS;
+ }
+
-+ ret = xencomm_create((void *)hypercall->arg[1], argsize,
-+ &desc, GFP_KERNEL);
-+ if (ret)
-+ return ret;
-+
-+ ret = xencomm_arch_hypercall_event_channel_op(cmd, desc);
++ desc = xencomm_map_no_alloc(arg, argsize);
++ if (desc == NULL)
++ return -EINVAL;
+
-+ xencomm_free(desc);
-+ return ret;
++ return xencomm_arch_hypercall_xenoprof_op(op, desc);
+}
++EXPORT_SYMBOL_GPL(xencomm_hypercall_xenoprof_op);
+
-+static int
-+xencomm_privcmd_hvm_op(privcmd_hypercall_t *hypercall)
++int
++xencomm_hypercall_perfmon_op(unsigned long cmd, void* arg,
++ unsigned long count)
+{
-+ int cmd = hypercall->arg[0];
-+ struct xencomm_handle *desc;
+ unsigned int argsize;
-+ int ret;
++ struct xencomm_handle *desc;
+
+ switch (cmd) {
-+ case HVMOP_get_param:
-+ case HVMOP_set_param:
-+ argsize = sizeof(xen_hvm_param_t);
++ case PFM_GET_FEATURES:
++ argsize = sizeof(pfarg_features_t);
+ break;
-+ case HVMOP_set_pci_intx_level:
-+ argsize = sizeof(xen_hvm_set_pci_intx_level_t);
++ case PFM_CREATE_CONTEXT:
++ argsize = sizeof(pfarg_context_t);
+ break;
-+ case HVMOP_set_isa_irq_level:
-+ argsize = sizeof(xen_hvm_set_isa_irq_level_t);
++ case PFM_LOAD_CONTEXT:
++ argsize = sizeof(pfarg_load_t);
+ break;
-+ case HVMOP_set_pci_link_route:
-+ argsize = sizeof(xen_hvm_set_pci_link_route_t);
++ case PFM_WRITE_PMCS:
++ case PFM_WRITE_PMDS:
++ argsize = sizeof(pfarg_reg_t) * count;
+ break;
+
++ case PFM_DESTROY_CONTEXT:
++ case PFM_UNLOAD_CONTEXT:
++ case PFM_START:
++ case PFM_STOP:
++ return xencomm_arch_hypercall_perfmon_op(cmd, arg, count);
++
+ default:
-+ printk("%s: unknown HVMOP %d\n", __func__, cmd);
-+ return -EINVAL;
++ printk("%s:%d cmd %ld isn't supported\n",
++ __func__, __LINE__, cmd);
++ BUG();
+ }
+
-+ ret = xencomm_create((void *)hypercall->arg[1], argsize,
-+ &desc, GFP_KERNEL);
-+ if (ret)
-+ return ret;
-+
-+ ret = xencomm_arch_hypercall_hvm_op(cmd, desc);
++ desc = xencomm_map_no_alloc(arg, argsize);
++ if (desc == NULL)
++ return -EINVAL;
+
-+ xencomm_free(desc);
-+ return ret;
++ return xencomm_arch_hypercall_perfmon_op(cmd, desc, count);
+}
++EXPORT_SYMBOL_GPL(xencomm_hypercall_perfmon_op);
+
-+static int
-+xencomm_privcmd_sched_op(privcmd_hypercall_t *hypercall)
++long
++xencomm_hypercall_vcpu_op(int cmd, int cpu, void *arg)
+{
-+ int cmd = hypercall->arg[0];
-+ struct xencomm_handle *desc;
+ unsigned int argsize;
-+ int ret;
-+
+ switch (cmd) {
-+ case SCHEDOP_remote_shutdown:
-+ argsize = sizeof(sched_remote_shutdown_t);
++ case VCPUOP_register_runstate_memory_area: {
++ vcpu_register_runstate_memory_area_t *area =
++ (vcpu_register_runstate_memory_area_t *)arg;
++ argsize = sizeof(*arg);
++ set_xen_guest_handle(area->addr.h,
++ (void *)xencomm_map_no_alloc(area->addr.v,
++ sizeof(area->addr.v)));
+ break;
++ }
++
+ default:
-+ printk("%s: unknown SCHEDOP %d\n", __func__, cmd);
-+ return -EINVAL;
++ printk("%s: unknown vcpu op %d\n", __func__, cmd);
++ return -ENOSYS;
+ }
+
-+ ret = xencomm_create((void *)hypercall->arg[1], argsize,
-+ &desc, GFP_KERNEL);
-+ if (ret)
-+ return ret;
++ return xencomm_arch_hypercall_vcpu_op(cmd, cpu,
++ xencomm_map_no_alloc(arg, argsize));
++}
+
-+ ret = xencomm_arch_hypercall_sched_op(cmd, desc);
++long
++xencomm_hypercall_opt_feature(void *arg)
++{
++ return xencomm_arch_hypercall_opt_feature(
++ xencomm_map_no_alloc(arg,
++ sizeof(struct xen_ia64_opt_feature)));
++}
+
-+ xencomm_free(desc);
-+ return ret;
++int
++xencomm_hypercall_fpswa_revision(unsigned int *revision)
++{
++ struct xencomm_handle *desc;
++
++ desc = xencomm_map_no_alloc(revision, sizeof(*revision));
++ if (desc == NULL)
++ return -EINVAL;
++
++ return xencomm_arch_hypercall_fpswa_revision(desc);
+}
++EXPORT_SYMBOL_GPL(xencomm_hypercall_fpswa_revision);
+
+int
-+privcmd_hypercall(privcmd_hypercall_t *hypercall)
++xencomm_hypercall_kexec_op(int cmd, void *arg)
+{
-+ switch (hypercall->op) {
-+ case __HYPERVISOR_platform_op:
-+ return xencomm_privcmd_platform_op(hypercall);
-+ case __HYPERVISOR_domctl:
-+ return xencomm_privcmd_domctl(hypercall);
-+ case __HYPERVISOR_sysctl:
-+ return xencomm_privcmd_sysctl(hypercall);
-+ case __HYPERVISOR_acm_op:
-+ return xencomm_privcmd_acm_op(hypercall);
-+ case __HYPERVISOR_xen_version:
-+ return xencomm_privcmd_xen_version(hypercall);
-+ case __HYPERVISOR_memory_op:
-+ return xencomm_privcmd_memory_op(hypercall);
-+ case __HYPERVISOR_event_channel_op:
-+ return xencomm_privcmd_event_channel_op(hypercall);
-+ case __HYPERVISOR_hvm_op:
-+ return xencomm_privcmd_hvm_op(hypercall);
-+ case __HYPERVISOR_sched_op:
-+ return xencomm_privcmd_sched_op(hypercall);
++ unsigned int argsize;
++ struct xencomm_handle *desc;
++
++ switch (cmd) {
++ case KEXEC_CMD_kexec_get_range:
++ argsize = sizeof(xen_kexec_range_t);
++ break;
++ case KEXEC_CMD_kexec_load:
++ case KEXEC_CMD_kexec_unload:
++ argsize = sizeof(xen_kexec_load_t);
++ break;
++ case KEXEC_CMD_kexec:
++ argsize = sizeof(xen_kexec_exec_t);
++ break;
+ default:
-+ printk("%s: unknown hcall (%ld)\n", __func__, hypercall->op);
-+ return -ENOSYS;
++ printk("%s:%d cmd %d isn't supported\n",
++ __func__, __LINE__, cmd);
++ BUG();
+ }
-+}
+
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/xencomm.c tmp-linux-2.6-xen.patch/arch/ia64/xen/xencomm.c
---- pristine-linux-2.6.18.2/arch/ia64/xen/xencomm.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/xencomm.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,263 @@
++ desc = xencomm_map_no_alloc(arg, argsize);
++ if (desc == NULL)
++ return -EINVAL;
++
++ return xencomm_arch_hypercall_kexec_op(cmd, desc);
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/xcom_privcmd.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/xcom_privcmd.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,902 @@
+/*
-+ * Copyright (C) 2006 Hollis Blanchard <hollisb at us.ibm.com>, IBM Corporation
-+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
-+ *
++ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
-+ *
++ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
-+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
++ *
++ * Authors: Hollis Blanchard <hollisb at us.ibm.com>
++ * Tristan Gingold <tristan.gingold at bull.net>
+ */
-+
++#include <linux/types.h>
++#include <linux/errno.h>
++#include <linux/kernel.h>
+#include <linux/gfp.h>
-+#include <linux/mm.h>
++#include <linux/module.h>
+#include <xen/interface/xen.h>
++#include <xen/interface/platform.h>
++#define __XEN__
++#include <xen/interface/domctl.h>
++#include <xen/interface/sysctl.h>
++#include <xen/interface/memory.h>
++#include <xen/interface/version.h>
++#include <xen/interface/event_channel.h>
++#include <xen/interface/xsm/acm_ops.h>
++#include <xen/interface/hvm/params.h>
++#include <xen/interface/arch-ia64/debug_op.h>
++#include <xen/public/privcmd.h>
++#include <asm/hypercall.h>
+#include <asm/page.h>
-+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
-+#endif
-+
++#include <asm/uaccess.h>
+#include <asm/xen/xencomm.h>
+
-+static int xencomm_debug = 0;
-+
-+static unsigned long kernel_start_pa;
-+
-+void
-+xencomm_init (void)
-+{
-+ kernel_start_pa = KERNEL_START - ia64_tpa(KERNEL_START);
-+}
++#define ROUND_DIV(v,s) (((v) + (s) - 1) / (s))
+
-+/* Translate virtual address to physical address. */
-+unsigned long
-+xencomm_vaddr_to_paddr(unsigned long vaddr)
++static int
++xencomm_privcmd_platform_op(privcmd_hypercall_t *hypercall)
+{
-+#ifndef CONFIG_VMX_GUEST
-+ struct page *page;
-+ struct vm_area_struct *vma;
-+#endif
-+
-+ if (vaddr == 0)
-+ return 0;
-+
-+#ifdef __ia64__
-+ if (REGION_NUMBER(vaddr) == 5) {
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *ptep;
-+
-+ /* On ia64, TASK_SIZE refers to current. It is not initialized
-+ during boot.
-+ Furthermore the kernel is relocatable and __pa() doesn't
-+ work on addresses. */
-+ if (vaddr >= KERNEL_START
-+ && vaddr < (KERNEL_START + KERNEL_TR_PAGE_SIZE)) {
-+ return vaddr - kernel_start_pa;
-+ }
-+
-+ /* In kernel area -- virtually mapped. */
-+ pgd = pgd_offset_k(vaddr);
-+ if (pgd_none(*pgd) || pgd_bad(*pgd))
-+ return ~0UL;
++ struct xen_platform_op kern_op;
++ struct xen_platform_op __user *user_op = (struct xen_platform_op __user *)hypercall->arg[0];
++ struct xencomm_handle *op_desc;
++ struct xencomm_handle *desc = NULL;
++ int ret = 0;
+
-+ pud = pud_offset(pgd, vaddr);
-+ if (pud_none(*pud) || pud_bad(*pud))
-+ return ~0UL;
++ if (copy_from_user(&kern_op, user_op, sizeof(struct xen_platform_op)))
++ return -EFAULT;
+
-+ pmd = pmd_offset(pud, vaddr);
-+ if (pmd_none(*pmd) || pmd_bad(*pmd))
-+ return ~0UL;
++ if (kern_op.interface_version != XENPF_INTERFACE_VERSION)
++ return -EACCES;
+
-+ ptep = pte_offset_kernel(pmd, vaddr);
-+ if (!ptep)
-+ return ~0UL;
++ op_desc = xencomm_map_no_alloc(&kern_op, sizeof(kern_op));
+
-+ return (pte_val(*ptep) & _PFN_MASK) | (vaddr & ~PAGE_MASK);
++ switch (kern_op.cmd) {
++ default:
++ printk("%s: unknown platform cmd %d\n", __func__, kern_op.cmd);
++ return -ENOSYS;
+ }
-+#endif
+
-+ if (vaddr > TASK_SIZE) {
-+ /* kernel address */
-+ return __pa(vaddr);
++ if (ret) {
++ /* error mapping the nested pointer */
++ return ret;
+ }
+
++ ret = xencomm_arch_hypercall_platform_op(op_desc);
+
-+#ifdef CONFIG_VMX_GUEST
-+ /* No privcmd within vmx guest. */
-+ return ~0UL;
-+#else
-+ /* XXX double-check (lack of) locking */
-+ vma = find_extend_vma(current->mm, vaddr);
-+ if (!vma)
-+ return ~0UL;
-+
-+ /* We assume the page is modified. */
-+ page = follow_page(vma, vaddr, FOLL_WRITE | FOLL_TOUCH);
-+ if (!page)
-+ return ~0UL;
++ /* FIXME: should we restore the handle? */
++ if (copy_to_user(user_op, &kern_op, sizeof(struct xen_platform_op)))
++ ret = -EFAULT;
+
-+ return (page_to_pfn(page) << PAGE_SHIFT) | (vaddr & ~PAGE_MASK);
-+#endif
++ xencomm_free(desc);
++ return ret;
+}
+
+static int
-+xencomm_init_desc(struct xencomm_desc *desc, void *buffer, unsigned long bytes)
++xencomm_privcmd_sysctl(privcmd_hypercall_t *hypercall)
+{
-+ unsigned long recorded = 0;
-+ int i = 0;
++ xen_sysctl_t kern_op;
++ xen_sysctl_t __user *user_op;
++ struct xencomm_handle *op_desc;
++ struct xencomm_handle *desc = NULL;
++ struct xencomm_handle *desc1 = NULL;
++ int ret = 0;
+
-+ BUG_ON((buffer == NULL) && (bytes > 0));
++ user_op = (xen_sysctl_t __user *)hypercall->arg[0];
+
-+ /* record the physical pages used */
-+ if (buffer == NULL)
-+ desc->nr_addrs = 0;
++ if (copy_from_user(&kern_op, user_op, sizeof(xen_sysctl_t)))
++ return -EFAULT;
+
-+ while ((recorded < bytes) && (i < desc->nr_addrs)) {
-+ unsigned long vaddr = (unsigned long)buffer + recorded;
-+ unsigned long paddr;
-+ int offset;
-+ int chunksz;
++ if (kern_op.interface_version != XEN_SYSCTL_INTERFACE_VERSION)
++ return -EACCES;
+
-+ offset = vaddr % PAGE_SIZE; /* handle partial pages */
-+ chunksz = min(PAGE_SIZE - offset, bytes - recorded);
++ op_desc = xencomm_map_no_alloc(&kern_op, sizeof(kern_op));
+
-+ paddr = xencomm_vaddr_to_paddr(vaddr);
-+ if (paddr == ~0UL) {
-+ printk("%s: couldn't translate vaddr %lx\n",
-+ __func__, vaddr);
-+ return -EINVAL;
++ switch (kern_op.cmd) {
++ case XEN_SYSCTL_readconsole:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.readconsole.buffer),
++ kern_op.u.readconsole.count);
++ if (xen_guest_handle(kern_op.u.readconsole.buffer) != NULL &&
++ kern_op.u.readconsole.count > 0 && desc == NULL)
++ return -ENOMEM;
++ set_xen_guest_handle(kern_op.u.readconsole.buffer,
++ (void *)desc);
++ break;
++ case XEN_SYSCTL_tbuf_op:
++ case XEN_SYSCTL_sched_id:
++ case XEN_SYSCTL_availheap:
++ break;
++ case XEN_SYSCTL_perfc_op:
++ {
++ struct xencomm_handle *tmp_desc;
++ xen_sysctl_t tmp_op = {
++ .cmd = XEN_SYSCTL_perfc_op,
++ .interface_version = XEN_SYSCTL_INTERFACE_VERSION,
++ .u.perfc_op = {
++ .cmd = XEN_SYSCTL_PERFCOP_query,
++ /* .desc.p = NULL, */
++ /* .val.p = NULL, */
++ },
++ };
++
++ if (xen_guest_handle(kern_op.u.perfc_op.desc) == NULL) {
++ if (xen_guest_handle(kern_op.u.perfc_op.val) != NULL)
++ return -EINVAL;
++ break;
+ }
+
-+ desc->address[i++] = paddr;
-+ recorded += chunksz;
++ /* query the buffer size for xencomm */
++ tmp_desc = xencomm_map_no_alloc(&tmp_op, sizeof(tmp_op));
++ ret = xencomm_arch_hypercall_sysctl(tmp_desc);
++ if (ret)
++ return ret;
++
++ desc = xencomm_map(xen_guest_handle(kern_op.u.perfc_op.desc),
++ tmp_op.u.perfc_op.nr_counters *
++ sizeof(xen_sysctl_perfc_desc_t));
++ if (xen_guest_handle(kern_op.u.perfc_op.desc) != NULL &&
++ tmp_op.u.perfc_op.nr_counters > 0 && desc == NULL)
++ return -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.perfc_op.desc, (void *)desc);
++
++ desc1 = xencomm_map(xen_guest_handle(kern_op.u.perfc_op.val),
++ tmp_op.u.perfc_op.nr_vals *
++ sizeof(xen_sysctl_perfc_val_t));
++ if (xen_guest_handle(kern_op.u.perfc_op.val) != NULL &&
++ tmp_op.u.perfc_op.nr_vals > 0 && desc1 == NULL) {
++ xencomm_free(desc);
++ return -ENOMEM;
++ }
++
++ set_xen_guest_handle(kern_op.u.perfc_op.val, (void *)desc1);
++ break;
++ }
++ case XEN_SYSCTL_getdomaininfolist:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.getdomaininfolist.buffer),
++ kern_op.u.getdomaininfolist.max_domains *
++ sizeof(xen_domctl_getdomaininfo_t));
++ if (xen_guest_handle(kern_op.u.getdomaininfolist.buffer) !=
++ NULL && kern_op.u.getdomaininfolist.max_domains > 0 &&
++ desc == NULL)
++ return -ENOMEM;
++ set_xen_guest_handle(kern_op.u.getdomaininfolist.buffer,
++ (void *)desc);
++ break;
++ case XEN_SYSCTL_debug_keys:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.debug_keys.keys),
++ kern_op.u.debug_keys.nr_keys);
++ if (xen_guest_handle(kern_op.u.debug_keys.keys) != NULL &&
++ kern_op.u.debug_keys.nr_keys > 0 && desc == NULL)
++ return -ENOMEM;
++ set_xen_guest_handle(kern_op.u.debug_keys.keys,
++ (void *)desc);
++ break;
++
++ case XEN_SYSCTL_physinfo:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.physinfo.cpu_to_node),
++ kern_op.u.physinfo.max_cpu_id * sizeof(uint32_t));
++ if (xen_guest_handle(kern_op.u.physinfo.cpu_to_node) != NULL &&
++ kern_op.u.physinfo.max_cpu_id > 0 && desc == NULL)
++ return -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.physinfo.cpu_to_node,
++ (void *)desc);
++ break;
++ default:
++ printk("%s: unknown sysctl cmd %d\n", __func__, kern_op.cmd);
++ return -ENOSYS;
+ }
+
-+ if (recorded < bytes) {
-+ printk("%s: could only translate %ld of %ld bytes\n",
-+ __func__, recorded, bytes);
-+ return -ENOSPC;
++ if (ret) {
++ /* error mapping the nested pointer */
++ return ret;
+ }
+
-+ /* mark remaining addresses invalid (just for safety) */
-+ while (i < desc->nr_addrs)
-+ desc->address[i++] = XENCOMM_INVALID;
++ ret = xencomm_arch_hypercall_sysctl(op_desc);
+
-+ desc->magic = XENCOMM_MAGIC;
++ /* FIXME: should we restore the handles? */
++ if (copy_to_user(user_op, &kern_op, sizeof(xen_sysctl_t)))
++ ret = -EFAULT;
+
-+ return 0;
++ xencomm_free(desc);
++ xencomm_free(desc1);
++ return ret;
+}
+
-+static struct xencomm_desc *
-+xencomm_alloc(gfp_t gfp_mask)
++static int
++xencomm_privcmd_domctl(privcmd_hypercall_t *hypercall)
+{
-+ struct xencomm_desc *desc;
++ xen_domctl_t kern_op;
++ xen_domctl_t __user *user_op;
++ struct xencomm_handle *op_desc;
++ struct xencomm_handle *desc = NULL;
++ int ret = 0;
+
-+ desc = (struct xencomm_desc *)__get_free_page(gfp_mask);
-+ if (desc == NULL)
-+ panic("%s: page allocation failed\n", __func__);
++ user_op = (xen_domctl_t __user *)hypercall->arg[0];
+
-+ desc->nr_addrs = (PAGE_SIZE - sizeof(struct xencomm_desc)) /
-+ sizeof(*desc->address);
++ if (copy_from_user(&kern_op, user_op, sizeof(xen_domctl_t)))
++ return -EFAULT;
+
-+ return desc;
++ if (kern_op.interface_version != XEN_DOMCTL_INTERFACE_VERSION)
++ return -EACCES;
++
++ op_desc = xencomm_map_no_alloc(&kern_op, sizeof(kern_op));
++
++ switch (kern_op.cmd) {
++ case XEN_DOMCTL_createdomain:
++ case XEN_DOMCTL_destroydomain:
++ case XEN_DOMCTL_pausedomain:
++ case XEN_DOMCTL_unpausedomain:
++ case XEN_DOMCTL_resumedomain:
++ case XEN_DOMCTL_getdomaininfo:
++ break;
++ case XEN_DOMCTL_getmemlist:
++ {
++ unsigned long nr_pages = kern_op.u.getmemlist.max_pfns;
++
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.getmemlist.buffer),
++ nr_pages * sizeof(unsigned long));
++ if (xen_guest_handle(kern_op.u.getmemlist.buffer) != NULL &&
++ nr_pages > 0 && desc == NULL)
++ return -ENOMEM;
++ set_xen_guest_handle(kern_op.u.getmemlist.buffer,
++ (void *)desc);
++ break;
++ }
++ case XEN_DOMCTL_getpageframeinfo:
++ break;
++ case XEN_DOMCTL_getpageframeinfo2:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.getpageframeinfo2.array),
++ kern_op.u.getpageframeinfo2.num);
++ if (xen_guest_handle(kern_op.u.getpageframeinfo2.array) !=
++ NULL && kern_op.u.getpageframeinfo2.num > 0 &&
++ desc == NULL)
++ return -ENOMEM;
++ set_xen_guest_handle(kern_op.u.getpageframeinfo2.array,
++ (void *)desc);
++ break;
++ case XEN_DOMCTL_shadow_op:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.shadow_op.dirty_bitmap),
++ ROUND_DIV(kern_op.u.shadow_op.pages, 8));
++ if (xen_guest_handle(kern_op.u.shadow_op.dirty_bitmap) != NULL
++ && kern_op.u.shadow_op.pages > 0 && desc == NULL)
++ return -ENOMEM;
++ set_xen_guest_handle(kern_op.u.shadow_op.dirty_bitmap,
++ (void *)desc);
++ break;
++ case XEN_DOMCTL_max_mem:
++ break;
++ case XEN_DOMCTL_setvcpucontext:
++ case XEN_DOMCTL_getvcpucontext:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.vcpucontext.ctxt),
++ sizeof(vcpu_guest_context_t));
++ if (xen_guest_handle(kern_op.u.vcpucontext.ctxt) != NULL &&
++ desc == NULL)
++ return -ENOMEM;
++ set_xen_guest_handle(kern_op.u.vcpucontext.ctxt, (void *)desc);
++ break;
++ case XEN_DOMCTL_getvcpuinfo:
++ break;
++ case XEN_DOMCTL_setvcpuaffinity:
++ case XEN_DOMCTL_getvcpuaffinity:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.vcpuaffinity.cpumap.bitmap),
++ ROUND_DIV(kern_op.u.vcpuaffinity.cpumap.nr_cpus, 8));
++ if (xen_guest_handle(kern_op.u.vcpuaffinity.cpumap.bitmap) !=
++ NULL && kern_op.u.vcpuaffinity.cpumap.nr_cpus > 0 &&
++ desc == NULL)
++ return -ENOMEM;
++ set_xen_guest_handle(kern_op.u.vcpuaffinity.cpumap.bitmap,
++ (void *)desc);
++ break;
++ case XEN_DOMCTL_gethvmcontext:
++ case XEN_DOMCTL_sethvmcontext:
++ if (kern_op.u.hvmcontext.size > 0)
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.hvmcontext.buffer),
++ kern_op.u.hvmcontext.size);
++ if (xen_guest_handle(kern_op.u.hvmcontext.buffer) != NULL &&
++ kern_op.u.hvmcontext.size > 0 && desc == NULL)
++ return -ENOMEM;
++ set_xen_guest_handle(kern_op.u.hvmcontext.buffer, (void*)desc);
++ break;
++ case XEN_DOMCTL_max_vcpus:
++ case XEN_DOMCTL_scheduler_op:
++ case XEN_DOMCTL_setdomainhandle:
++ case XEN_DOMCTL_setdebugging:
++ case XEN_DOMCTL_irq_permission:
++ case XEN_DOMCTL_iomem_permission:
++ case XEN_DOMCTL_ioport_permission:
++ case XEN_DOMCTL_hypercall_init:
++ case XEN_DOMCTL_arch_setup:
++ case XEN_DOMCTL_settimeoffset:
++ case XEN_DOMCTL_sendtrigger:
++ case XEN_DOMCTL_set_opt_feature:
++ case XEN_DOMCTL_assign_device:
++ case XEN_DOMCTL_subscribe:
++ break;
++ case XEN_DOMCTL_pin_mem_cacheattr:
++ return -ENOSYS;
++ default:
++ printk("%s: unknown domctl cmd %d\n", __func__, kern_op.cmd);
++ return -ENOSYS;
++ }
++
++ if (ret) {
++ /* error mapping the nested pointer */
++ return ret;
++ }
++
++ ret = xencomm_arch_hypercall_domctl (op_desc);
++
++ /* FIXME: should we restore the handle? */
++ if (copy_to_user(user_op, &kern_op, sizeof(xen_domctl_t)))
++ ret = -EFAULT;
++
++ xencomm_free(desc);
++ return ret;
+}
+
-+void
-+xencomm_free(struct xencomm_handle *desc)
++static int
++xencomm_privcmd_xsm_op(privcmd_hypercall_t *hypercall)
+{
-+ if (desc)
-+ free_page((unsigned long)__va(desc));
++ void __user *arg = (void __user *)hypercall->arg[0];
++ xen_acmctl_t kern_arg;
++ struct xencomm_handle *op_desc;
++ struct xencomm_handle *desc = NULL;
++ int ret;
++
++ if (copy_from_user(&kern_arg, arg, sizeof(kern_arg)))
++ return -EFAULT;
++ if (kern_arg.interface_version != ACM_INTERFACE_VERSION)
++ return -ENOSYS;
++
++ switch (kern_arg.cmd) {
++ case ACMOP_getssid: {
++ op_desc = xencomm_map_no_alloc(&kern_arg, sizeof(kern_arg));
++
++ desc = xencomm_map(
++ xen_guest_handle(kern_arg.u.getssid.ssidbuf),
++ kern_arg.u.getssid.ssidbuf_size);
++ if (xen_guest_handle(kern_arg.u.getssid.ssidbuf) != NULL &&
++ kern_arg.u.getssid.ssidbuf_size > 0 && desc == NULL)
++ return -ENOMEM;
++
++ set_xen_guest_handle(kern_arg.u.getssid.ssidbuf, (void *)desc);
++
++ ret = xencomm_arch_hypercall_xsm_op(op_desc);
++
++ xencomm_free(desc);
++
++ if (copy_to_user(arg, &kern_arg, sizeof(kern_arg)))
++ return -EFAULT;
++ return ret;
++ }
++ default:
++ printk("%s: unknown acm_op cmd %d\n", __func__, kern_arg.cmd);
++ return -ENOSYS;
++ }
++
++ return ret;
+}
+
-+int
-+xencomm_create(void *buffer, unsigned long bytes,
-+ struct xencomm_handle **ret, gfp_t gfp_mask)
++static int
++xencomm_privcmd_memory_reservation_op(privcmd_hypercall_t *hypercall)
+{
-+ struct xencomm_desc *desc;
-+ struct xencomm_handle *handle;
-+ int rc;
++ const unsigned long cmd = hypercall->arg[0];
++ int ret = 0;
++ xen_memory_reservation_t kern_op;
++ xen_memory_reservation_t __user *user_op;
++ struct xencomm_handle *desc = NULL;
++ struct xencomm_handle *desc_op;
+
-+ if (xencomm_debug)
-+ printk("%s: %p[%ld]\n", __func__, buffer, bytes);
++ user_op = (xen_memory_reservation_t __user *)hypercall->arg[1];
++ if (copy_from_user(&kern_op, user_op,
++ sizeof(xen_memory_reservation_t)))
++ return -EFAULT;
++ desc_op = xencomm_map_no_alloc(&kern_op, sizeof(kern_op));
+
-+ if (buffer == NULL || bytes == 0) {
-+ *ret = (struct xencomm_handle *)NULL;
-+ return 0;
-+ }
++ if (!xen_guest_handle(kern_op.extent_start)) {
++ ret = xencomm_arch_hypercall_memory_op(cmd, desc_op);
++ if (ret < 0)
++ return ret;
++ } else {
++ xen_ulong_t nr_done = 0;
++ xen_ulong_t nr_extents = kern_op.nr_extents;
++ void *addr = xen_guest_handle(kern_op.extent_start);
++
++ /*
++ * Work around.
++ * Xencomm has single page size limit caused
++ * by xencomm_alloc()/xencomm_free() so that
++ * we have to repeat the hypercall.
++ * This limitation can be removed.
++ */
++#define MEMORYOP_XENCOMM_LIMIT \
++ (((((PAGE_SIZE - sizeof(struct xencomm_desc)) / \
++ sizeof(uint64_t)) - 2) * PAGE_SIZE) / \
++ sizeof(*xen_guest_handle(kern_op.extent_start)))
+
-+ desc = xencomm_alloc(gfp_mask);
-+ if (!desc) {
-+ printk("%s failure\n", "xencomm_alloc");
-+ return -ENOMEM;
-+ }
-+ handle = (struct xencomm_handle *)__pa(desc);
++ /*
++ * Work around.
++ * Even if the above limitation is removed,
++ * the hypercall with large number of extents
++ * may cause the soft lockup warning.
++ * In order to avoid the warning, we limit
++ * the number of extents and repeat the hypercall.
++ * The following value is determined by evaluation.
++ * Time of one hypercall should be smaller than
++ * a vcpu time slice. The time with current
++ * MEMORYOP_MAX_EXTENTS is around 5 msec.
++ * If the following limit causes some issues,
++ * we should decrease this value.
++ *
++ * Another way would be that start with small value and
++ * increase adoptively measuring hypercall time.
++ * It might be over-kill.
++ */
++#define MEMORYOP_MAX_EXTENTS (MEMORYOP_XENCOMM_LIMIT / 512)
+
-+ rc = xencomm_init_desc(desc, buffer, bytes);
-+ if (rc) {
-+ printk("%s failure: %d\n", "xencomm_init_desc", rc);
-+ xencomm_free(handle);
-+ return rc;
++ while (nr_extents > 0) {
++ xen_ulong_t nr_tmp = nr_extents;
++ if (nr_tmp > MEMORYOP_MAX_EXTENTS)
++ nr_tmp = MEMORYOP_MAX_EXTENTS;
++
++ kern_op.nr_extents = nr_tmp;
++ desc = xencomm_map
++ (addr + nr_done * sizeof(*xen_guest_handle(kern_op.extent_start)),
++ nr_tmp * sizeof(*xen_guest_handle(kern_op.extent_start)));
++ if (addr != NULL && nr_tmp > 0 && desc == NULL)
++ return nr_done > 0 ? nr_done : -ENOMEM;
++
++ set_xen_guest_handle(kern_op.extent_start,
++ (void *)desc);
++
++ ret = xencomm_arch_hypercall_memory_op(cmd, desc_op);
++ xencomm_free(desc);
++ if (ret < 0)
++ return nr_done > 0 ? nr_done : ret;
++
++ nr_done += ret;
++ nr_extents -= ret;
++ if (ret < nr_tmp)
++ break;
++
++ /*
++ * prevent softlock up message.
++ * give cpu to soft lockup kernel thread.
++ */
++ if (nr_extents > 0)
++ schedule();
++ }
++ ret = nr_done;
++ set_xen_guest_handle(kern_op.extent_start, addr);
+ }
+
-+ *ret = handle;
-+ return 0;
-+}
++ if (copy_to_user(user_op, &kern_op, sizeof(xen_memory_reservation_t)))
++ return -EFAULT;
+
-+/* "mini" routines, for stack-based communications: */
++ return ret;
++}
+
-+static void *
-+xencomm_alloc_mini(struct xencomm_mini *area, int *nbr_area)
++static int
++xencomm_privcmd_memory_op(privcmd_hypercall_t *hypercall)
+{
-+ unsigned long base;
-+ unsigned int pageoffset;
++ const unsigned long cmd = hypercall->arg[0];
++ int ret = 0;
++
++ switch (cmd) {
++ case XENMEM_increase_reservation:
++ case XENMEM_decrease_reservation:
++ case XENMEM_populate_physmap:
++ return xencomm_privcmd_memory_reservation_op(hypercall);
++ case XENMEM_maximum_gpfn:
++ {
++ domid_t kern_domid;
++ domid_t __user *user_domid;
++ struct xencomm_handle *desc;
+
-+ while (*nbr_area >= 0) {
-+ /* Allocate an area. */
-+ (*nbr_area)--;
++ user_domid = (domid_t __user *)hypercall->arg[1];
++ if (copy_from_user(&kern_domid, user_domid, sizeof(domid_t)))
++ return -EFAULT;
++ desc = xencomm_map_no_alloc(&kern_domid, sizeof(kern_domid));
+
-+ base = (unsigned long)(area + *nbr_area);
-+ pageoffset = base % PAGE_SIZE;
++ ret = xencomm_arch_hypercall_memory_op(cmd, desc);
+
-+ /* If the area does not cross a page, use it. */
-+ if ((PAGE_SIZE - pageoffset) >= sizeof(struct xencomm_mini))
-+ return &area[*nbr_area];
++ return ret;
+ }
-+ /* No more area. */
-+ return NULL;
++ case XENMEM_translate_gpfn_list:
++ {
++ xen_translate_gpfn_list_t kern_op;
++ xen_translate_gpfn_list_t __user *user_op;
++ struct xencomm_handle *desc_gpfn = NULL;
++ struct xencomm_handle *desc_mfn = NULL;
++ struct xencomm_handle *desc_op;
++ void *addr;
++
++ user_op = (xen_translate_gpfn_list_t __user *)
++ hypercall->arg[1];
++ if (copy_from_user(&kern_op, user_op,
++ sizeof(xen_translate_gpfn_list_t)))
++ return -EFAULT;
++ desc_op = xencomm_map_no_alloc(&kern_op, sizeof(kern_op));
++
++ if (kern_op.nr_gpfns) {
++ /* gpfn_list. */
++ addr = xen_guest_handle(kern_op.gpfn_list);
++
++ desc_gpfn = xencomm_map(addr, kern_op.nr_gpfns *
++ sizeof(*xen_guest_handle
++ (kern_op.gpfn_list)));
++ if (addr != NULL && kern_op.nr_gpfns > 0 &&
++ desc_gpfn == NULL)
++ return -ENOMEM;
++ set_xen_guest_handle(kern_op.gpfn_list,
++ (void *)desc_gpfn);
++
++ /* mfn_list. */
++ addr = xen_guest_handle(kern_op.mfn_list);
++
++ desc_mfn = xencomm_map(addr, kern_op.nr_gpfns *
++ sizeof(*xen_guest_handle
++ (kern_op.mfn_list)));
++ if (addr != NULL && kern_op.nr_gpfns > 0 &&
++ desc_mfn == NULL) {
++ xencomm_free(desc_gpfn);
++ return -ENOMEM;
++ }
++
++ set_xen_guest_handle(kern_op.mfn_list,
++ (void *)desc_mfn);
++ }
++
++ ret = xencomm_arch_hypercall_memory_op(cmd, desc_op);
++
++ xencomm_free(desc_gpfn);
++ xencomm_free(desc_mfn);
++
++ if (ret != 0)
++ return ret;
++
++ return ret;
++ }
++ default:
++ printk("%s: unknown memory op %lu\n", __func__, cmd);
++ ret = -ENOSYS;
++ }
++ return ret;
+}
+
-+int
-+xencomm_create_mini(struct xencomm_mini *area, int *nbr_area,
-+ void *buffer, unsigned long bytes,
-+ struct xencomm_handle **ret)
++static int
++xencomm_privcmd_xen_version(privcmd_hypercall_t *hypercall)
+{
-+ struct xencomm_desc *desc;
++ int cmd = hypercall->arg[0];
++ void __user *arg = (void __user *)hypercall->arg[1];
++ struct xencomm_handle *desc;
++ size_t argsize;
+ int rc;
-+ unsigned long res;
+
-+ desc = xencomm_alloc_mini(area, nbr_area);
-+ if (!desc)
++ switch (cmd) {
++ case XENVER_version:
++ /* do not actually pass an argument */
++ return xencomm_arch_hypercall_xen_version(cmd, 0);
++ case XENVER_extraversion:
++ argsize = sizeof(xen_extraversion_t);
++ break;
++ case XENVER_compile_info:
++ argsize = sizeof(xen_compile_info_t);
++ break;
++ case XENVER_capabilities:
++ argsize = sizeof(xen_capabilities_info_t);
++ break;
++ case XENVER_changeset:
++ argsize = sizeof(xen_changeset_info_t);
++ break;
++ case XENVER_platform_parameters:
++ argsize = sizeof(xen_platform_parameters_t);
++ break;
++ case XENVER_pagesize:
++ argsize = (arg == NULL) ? 0 : sizeof(void *);
++ break;
++ case XENVER_get_features:
++ argsize = (arg == NULL) ? 0 : sizeof(xen_feature_info_t);
++ break;
++
++ default:
++ printk("%s: unknown version op %d\n", __func__, cmd);
++ return -ENOSYS;
++ }
++
++ desc = xencomm_map(arg, argsize);
++ if (arg != NULL && argsize > 0 && desc == NULL)
+ return -ENOMEM;
-+ desc->nr_addrs = XENCOMM_MINI_ADDRS;
+
-+ rc = xencomm_init_desc(desc, buffer, bytes);
-+ if (rc)
-+ return rc;
++ rc = xencomm_arch_hypercall_xen_version(cmd, desc);
+
-+ res = xencomm_vaddr_to_paddr((unsigned long)desc);
-+ if (res == ~0UL)
-+ return -EINVAL;
++ xencomm_free(desc);
+
-+ *ret = (struct xencomm_handle*)res;
-+ return 0;
++ return rc;
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/xenentry.S tmp-linux-2.6-xen.patch/arch/ia64/xen/xenentry.S
---- pristine-linux-2.6.18.2/arch/ia64/xen/xenentry.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/xenentry.S 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,931 @@
-+/*
-+ * ia64/xen/entry.S
-+ *
-+ * Alternate kernel routines for Xen. Heavily leveraged from
-+ * ia64/kernel/entry.S
-+ *
-+ * Copyright (C) 2005 Hewlett-Packard Co
-+ * Dan Magenheimer <dan.magenheimer at .hp.com>
-+ */
+
-+#include <asm/asmmacro.h>
-+#include <asm/cache.h>
-+#include <asm/errno.h>
-+#include <asm/kregs.h>
-+#include <asm/asm-offsets.h>
-+#include <asm/pgtable.h>
-+#include <asm/percpu.h>
-+#include <asm/processor.h>
-+#include <asm/thread_info.h>
-+#include <asm/unistd.h>
++static int
++xencomm_privcmd_event_channel_op(privcmd_hypercall_t *hypercall)
++{
++ int cmd = hypercall->arg[0];
++ struct xencomm_handle *desc;
++ unsigned int argsize;
++ int ret;
+
-+#ifdef CONFIG_XEN
-+#include "xenminstate.h"
-+#else
-+#include "minstate.h"
-+#endif
++ switch (cmd) {
++ case EVTCHNOP_alloc_unbound:
++ argsize = sizeof(evtchn_alloc_unbound_t);
++ break;
+
-+/*
-+ * prev_task <- ia64_switch_to(struct task_struct *next)
-+ * With Ingo's new scheduler, interrupts are disabled when this routine gets
-+ * called. The code starting at .map relies on this. The rest of the code
-+ * doesn't care about the interrupt masking status.
-+ */
-+#ifdef CONFIG_XEN
-+GLOBAL_ENTRY(xen_switch_to)
-+ .prologue
-+ alloc r16=ar.pfs,1,0,0,0
-+ movl r22=running_on_xen;;
-+ ld4 r22=[r22];;
-+ cmp.eq p7,p0=r22,r0
-+(p7) br.cond.sptk.many __ia64_switch_to;;
-+#else
-+GLOBAL_ENTRY(ia64_switch_to)
-+ .prologue
-+ alloc r16=ar.pfs,1,0,0,0
-+#endif
-+ DO_SAVE_SWITCH_STACK
-+ .body
++ case EVTCHNOP_status:
++ argsize = sizeof(evtchn_status_t);
++ break;
+
-+ adds r22=IA64_TASK_THREAD_KSP_OFFSET,r13
-+ movl r25=init_task
-+ mov r27=IA64_KR(CURRENT_STACK)
-+ adds r21=IA64_TASK_THREAD_KSP_OFFSET,in0
-+ dep r20=0,in0,61,3 // physical address of "next"
-+ ;;
-+ st8 [r22]=sp // save kernel stack pointer of old task
-+ shr.u r26=r20,IA64_GRANULE_SHIFT
-+ cmp.eq p7,p6=r25,in0
-+ ;;
++ default:
++ printk("%s: unknown EVTCHNOP %d\n", __func__, cmd);
++ return -EINVAL;
++ }
++
++ desc = xencomm_map((void *)hypercall->arg[1], argsize);
++ if ((void *)hypercall->arg[1] != NULL && argsize > 0 && desc == NULL)
++ return -ENOMEM;
++
++ ret = xencomm_arch_hypercall_event_channel_op(cmd, desc);
++
++ xencomm_free(desc);
++ return ret;
++}
++
++static int
++xencomm_privcmd_hvm_op_track_dirty_vram(privcmd_hypercall_t *hypercall)
++{
++#if 1
+ /*
-+ * If we've already mapped this task's page, we can skip doing it again.
++ * At this moment HVMOP_track_dirty_vram isn't implemented
++ * on xen/ia64 so that it just returns -ENOSYS.
++ * Don't issue hypercall to get -ENOSYS.
++ * When the hypercall is implemented, enable the following codes.
+ */
-+(p6) cmp.eq p7,p6=r26,r27
-+(p6) br.cond.dpnt .map
-+ ;;
-+.done:
-+ ld8 sp=[r21] // load kernel stack pointer of new task
-+#ifdef CONFIG_XEN
-+ // update "current" application register
-+ mov r8=IA64_KR_CURRENT
-+ mov r9=in0;;
-+ XEN_HYPER_SET_KR
++ return -ENOSYS;
+#else
-+ mov IA64_KR(CURRENT)=in0 // update "current" application register
++ int cmd = hypercall->arg[0];
++ struct xen_hvm_track_dirty_vram *user_op = (void*)hypercall->arg[1];
++ struct xen_hvm_track_dirty_vram kern_op;
++ struct xencomm_handle *desc;
++ struct xencomm_handle *bitmap_desc;
++ int ret;
++
++ BUG_ON(cmd != HVMOP_track_dirty_vram);
++ if (copy_from_user(&kern_op, user_op, sizeof(kern_op)))
++ return -EFAULT;
++ desc = xencomm_map_no_alloc(&kern_op, sizeof(kern_op));
++ bitmap_desc = xencomm_map(xen_guest_handle(kern_op.dirty_bitmap),
++ kern_op.nr * sizeof(uint8_t));
++ if (bitmap_desc == NULL)
++ return -ENOMEM;
++ set_xen_guest_handle(kern_op.dirty_bitmap, (void*)bitmap_desc);
++ ret = xencomm_arch_hypercall_hvm_op(cmd, desc);
++ xencomm_free(bitmap_desc);
++
++ return ret;
++#endif
++}
++
++static int
++xencomm_privcmd_hvm_op(privcmd_hypercall_t *hypercall)
++{
++ int cmd = hypercall->arg[0];
++ struct xencomm_handle *desc;
++ unsigned int argsize;
++ int ret;
++
++ switch (cmd) {
++ case HVMOP_get_param:
++ case HVMOP_set_param:
++ argsize = sizeof(xen_hvm_param_t);
++ break;
++ case HVMOP_set_pci_intx_level:
++ argsize = sizeof(xen_hvm_set_pci_intx_level_t);
++ break;
++ case HVMOP_set_isa_irq_level:
++ argsize = sizeof(xen_hvm_set_isa_irq_level_t);
++ break;
++ case HVMOP_set_pci_link_route:
++ argsize = sizeof(xen_hvm_set_pci_link_route_t);
++ break;
++ case HVMOP_set_mem_type:
++ argsize = sizeof(xen_hvm_set_mem_type_t);
++ break;
++
++ case HVMOP_track_dirty_vram:
++ return xencomm_privcmd_hvm_op_track_dirty_vram(hypercall);
++
++ default:
++ printk("%s: unknown HVMOP %d\n", __func__, cmd);
++ return -EINVAL;
++ }
++
++ desc = xencomm_map((void *)hypercall->arg[1], argsize);
++ if ((void *)hypercall->arg[1] != NULL && argsize > 0 && desc == NULL)
++ return -ENOMEM;
++
++ ret = xencomm_arch_hypercall_hvm_op(cmd, desc);
++
++ xencomm_free(desc);
++ return ret;
++}
++
++static int
++xencomm_privcmd_sched_op(privcmd_hypercall_t *hypercall)
++{
++ int cmd = hypercall->arg[0];
++ struct xencomm_handle *desc;
++ unsigned int argsize;
++ int ret;
++
++ switch (cmd) {
++ case SCHEDOP_remote_shutdown:
++ argsize = sizeof(sched_remote_shutdown_t);
++ break;
++ default:
++ printk("%s: unknown SCHEDOP %d\n", __func__, cmd);
++ return -EINVAL;
++ }
++
++ desc = xencomm_map((void *)hypercall->arg[1], argsize);
++ if ((void *)hypercall->arg[1] != NULL && argsize > 0 && desc == NULL)
++ return -ENOMEM;
++
++ ret = xencomm_arch_hypercall_sched_op(cmd, desc);
++
++ xencomm_free(desc);
++ return ret;
++}
++
++static int
++xencomm_privcmd_ia64_dom0vp_op(privcmd_hypercall_t *hypercall)
++{
++ int cmd = hypercall->arg[0];
++ int ret;
++
++ switch (cmd) {
++ case IA64_DOM0VP_fpswa_revision: {
++ unsigned int revision;
++ unsigned int __user *revision_user =
++ (unsigned int* __user)hypercall->arg[1];
++ struct xencomm_handle *desc;
++ desc = xencomm_map(&revision, sizeof(revision));
++ if (desc == NULL)
++ return -ENOMEM;
++
++ ret = xencomm_arch_hypercall_fpswa_revision(desc);
++ xencomm_free(desc);
++ if (ret)
++ break;
++ if (copy_to_user(revision_user, &revision, sizeof(revision)))
++ ret = -EFAULT;
++ break;
++ }
++#ifdef CONFIG_XEN_IA64_EXPOSE_P2M
++ case IA64_DOM0VP_expose_foreign_p2m:
++ ret = xen_foreign_p2m_expose(hypercall);
++ break;
++#endif
++ default:
++ printk("%s: unknown IA64 DOM0VP op %d\n", __func__, cmd);
++ ret = -EINVAL;
++ break;
++ }
++ return ret;
++}
++
++static int
++xencomm_privcmd_ia64_debug_op(privcmd_hypercall_t *hypercall)
++{
++ int cmd = hypercall->arg[0];
++ unsigned long domain = hypercall->arg[1];
++ struct xencomm_handle *desc;
++ int ret;
++
++ switch (cmd) {
++ case XEN_IA64_DEBUG_OP_SET_FLAGS:
++ case XEN_IA64_DEBUG_OP_GET_FLAGS:
++ break;
++ default:
++ printk("%s: unknown IA64 DEBUGOP %d\n", __func__, cmd);
++ return -EINVAL;
++ }
++
++ desc = xencomm_map((void *)hypercall->arg[2],
++ sizeof(xen_ia64_debug_op_t));
++ if (desc == NULL)
++ return -ENOMEM;
++
++ ret = xencomm_arch_hypercall_ia64_debug_op(cmd, domain, desc);
++
++ xencomm_free(desc);
++ return ret;
++}
++
++static int
++xencomm_privcmd_ia64_physdev_op(privcmd_hypercall_t *hypercall)
++{
++ int cmd = hypercall->arg[0];
++ struct xencomm_handle *desc;
++ unsigned int argsize;
++ int ret;
++
++ switch (cmd) {
++ case PHYSDEVOP_map_pirq:
++ argsize = sizeof(physdev_map_pirq_t);
++ break;
++ case PHYSDEVOP_unmap_pirq:
++ argsize = sizeof(physdev_unmap_pirq_t);
++ break;
++ default:
++ printk("%s: unknown PHYSDEVOP %d\n", __func__, cmd);
++ return -EINVAL;
++ }
++
++ desc = xencomm_map((void *)hypercall->arg[1], argsize);
++ if ((void *)hypercall->arg[1] != NULL && argsize > 0 && desc == NULL)
++ return -ENOMEM;
++
++ ret = xencomm_arch_hypercall_physdev_op(cmd, desc);
++
++ xencomm_free(desc);
++ return ret;
++}
++
++int
++privcmd_hypercall(privcmd_hypercall_t *hypercall)
++{
++ switch (hypercall->op) {
++ case __HYPERVISOR_platform_op:
++ return xencomm_privcmd_platform_op(hypercall);
++ case __HYPERVISOR_domctl:
++ return xencomm_privcmd_domctl(hypercall);
++ case __HYPERVISOR_sysctl:
++ return xencomm_privcmd_sysctl(hypercall);
++ case __HYPERVISOR_xsm_op:
++ return xencomm_privcmd_xsm_op(hypercall);
++ case __HYPERVISOR_xen_version:
++ return xencomm_privcmd_xen_version(hypercall);
++ case __HYPERVISOR_memory_op:
++ return xencomm_privcmd_memory_op(hypercall);
++ case __HYPERVISOR_event_channel_op:
++ return xencomm_privcmd_event_channel_op(hypercall);
++ case __HYPERVISOR_hvm_op:
++ return xencomm_privcmd_hvm_op(hypercall);
++ case __HYPERVISOR_sched_op:
++ return xencomm_privcmd_sched_op(hypercall);
++ case __HYPERVISOR_ia64_dom0vp_op:
++ return xencomm_privcmd_ia64_dom0vp_op(hypercall);
++ case __HYPERVISOR_ia64_debug_op:
++ return xencomm_privcmd_ia64_debug_op(hypercall);
++ case __HYPERVISOR_physdev_op:
++ return xencomm_privcmd_ia64_physdev_op(hypercall);
++ default:
++ printk("%s: unknown hcall (%ld)\n", __func__, hypercall->op);
++ return -ENOSYS;
++ }
++}
++
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/xen_dma.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/xen_dma.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,190 @@
++/*
++ * Copyright (C) 2007 Hewlett-Packard Development Company, L.P.
++ * Alex Williamson <alex.williamson at hp.com>
++ *
++ * Basic DMA mapping services for Xen guests.
++ * Based on arch/i386/kernel/pci-dma-xen.c.
++ *
++ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ */
++
++#include <linux/bitops.h>
++#include <linux/dma-mapping.h>
++#include <linux/mm.h>
++#include <asm/scatterlist.h>
++#include <xen/gnttab.h>
++#include <asm/gnttab_dma.h>
++
++#define IOMMU_BUG_ON(test) \
++do { \
++ if (unlikely(test)) { \
++ printk(KERN_ALERT "Fatal DMA error!\n"); \
++ BUG(); \
++ } \
++} while (0)
++
++static int check_pages_physically_contiguous(unsigned long pfn,
++ unsigned int offset,
++ size_t length)
++{
++ unsigned long next_bus;
++ int i;
++ int nr_pages;
++
++ next_bus = pfn_to_mfn_for_dma(pfn);
++ nr_pages = (offset + length + PAGE_SIZE-1) >> PAGE_SHIFT;
++
++ for (i = 1; i < nr_pages; i++) {
++ if (pfn_to_mfn_for_dma(++pfn) != ++next_bus)
++ return 0;
++ }
++ return 1;
++}
++
++int range_straddles_page_boundary(paddr_t p, size_t size)
++{
++ extern unsigned long *contiguous_bitmap;
++ unsigned long pfn = p >> PAGE_SHIFT;
++ unsigned int offset = p & ~PAGE_MASK;
++
++ if (!is_running_on_xen())
++ return 0;
++
++ if (offset + size <= PAGE_SIZE)
++ return 0;
++ if (test_bit(pfn, contiguous_bitmap))
++ return 0;
++ if (check_pages_physically_contiguous(pfn, offset, size))
++ return 0;
++ return 1;
++}
++
++/*
++ * This should be broken out of swiotlb and put in a common place
++ * when merged with upstream Linux.
++ */
++static inline int
++address_needs_mapping(struct device *dev, dma_addr_t addr)
++{
++ dma_addr_t mask = 0xffffffff;
++
++ /* If the device has a mask, use it, otherwise default to 32 bits */
++ if (dev && dev->dma_mask)
++ mask = *dev->dma_mask;
++ return (addr & ~mask) != 0;
++}
++
++int
++xen_map_sg(struct device *dev, struct scatterlist *sg, int nents,
++ int direction)
++{
++ int i;
++
++ for (i = 0 ; i < nents ; i++) {
++ sg[i].dma_address = gnttab_dma_map_page(sg[i].page) + sg[i].offset;
++ sg[i].dma_length = sg[i].length;
++
++ IOMMU_BUG_ON(address_needs_mapping(dev, sg[i].dma_address));
++ IOMMU_BUG_ON(range_straddles_page_boundary(
++ page_to_pseudophys(sg[i].page) + sg[i].offset,
++ sg[i].length));
++ }
++
++ return nents;
++}
++EXPORT_SYMBOL(xen_map_sg);
++
++void
++xen_unmap_sg(struct device *dev, struct scatterlist *sg, int nents,
++ int direction)
++{
++ int i;
++ for (i = 0; i < nents; i++)
++ __gnttab_dma_unmap_page(sg[i].page);
++}
++EXPORT_SYMBOL(xen_unmap_sg);
++
++int
++xen_dma_mapping_error(dma_addr_t dma_addr)
++{
++ return 0;
++}
++EXPORT_SYMBOL(xen_dma_mapping_error);
++
++int
++xen_dma_supported(struct device *dev, u64 mask)
++{
++ return 1;
++}
++EXPORT_SYMBOL(xen_dma_supported);
++
++void *
++xen_alloc_coherent(struct device *dev, size_t size,
++ dma_addr_t *dma_handle, gfp_t gfp)
++{
++ unsigned long vaddr;
++ unsigned int order = get_order(size);
++
++ vaddr = __get_free_pages(gfp, order);
++
++ if (!vaddr)
++ return NULL;
++
++ if (xen_create_contiguous_region(vaddr, order,
++ fls64(dev->coherent_dma_mask))) {
++ free_pages(vaddr, order);
++ return NULL;
++ }
++
++ memset((void *)vaddr, 0, size);
++ *dma_handle = virt_to_bus((void *)vaddr);
++
++ return (void *)vaddr;
++}
++EXPORT_SYMBOL(xen_alloc_coherent);
++
++void
++xen_free_coherent(struct device *dev, size_t size,
++ void *vaddr, dma_addr_t dma_handle)
++{
++ unsigned int order = get_order(size);
++
++ xen_destroy_contiguous_region((unsigned long)vaddr, order);
++ free_pages((unsigned long)vaddr, order);
++}
++EXPORT_SYMBOL(xen_free_coherent);
++
++dma_addr_t
++xen_map_single(struct device *dev, void *ptr, size_t size,
++ int direction)
++{
++ dma_addr_t dma_addr = gnttab_dma_map_virt(ptr);
++
++ IOMMU_BUG_ON(range_straddles_page_boundary(__pa(ptr), size));
++ IOMMU_BUG_ON(address_needs_mapping(dev, dma_addr));
++
++ return dma_addr;
++}
++EXPORT_SYMBOL(xen_map_single);
++
++void
++xen_unmap_single(struct device *dev, dma_addr_t dma_addr, size_t size,
++ int direction)
++{
++ gnttab_dma_unmap_page(dma_addr);
++}
++EXPORT_SYMBOL(xen_unmap_single);
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/xencomm.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/xencomm.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,109 @@
++/*
++ * Copyright (C) 2006 Hollis Blanchard <hollisb at us.ibm.com>, IBM Corporation
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ */
++
++#include <linux/gfp.h>
++#include <linux/mm.h>
++#include <xen/interface/xen.h>
++#include <asm/page.h>
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++#include <asm/xen/xencomm.h>
++
++static unsigned long kernel_start_pa;
++
++void
++xencomm_initialize (void)
++{
++ kernel_start_pa = KERNEL_START - ia64_tpa(KERNEL_START);
++}
++
++/* Translate virtual address to physical address. */
++unsigned long
++xencomm_vtop(unsigned long vaddr)
++{
++#ifndef CONFIG_VMX_GUEST
++ struct page *page;
++ struct vm_area_struct *vma;
++#endif
++
++ if (vaddr == 0)
++ return 0;
++
++#ifdef __ia64__
++ if (REGION_NUMBER(vaddr) == 5) {
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *ptep;
++
++ /* On ia64, TASK_SIZE refers to current. It is not initialized
++ during boot.
++ Furthermore the kernel is relocatable and __pa() doesn't
++ work on addresses. */
++ if (vaddr >= KERNEL_START
++ && vaddr < (KERNEL_START + KERNEL_TR_PAGE_SIZE)) {
++ return vaddr - kernel_start_pa;
++ }
++
++ /* In kernel area -- virtually mapped. */
++ pgd = pgd_offset_k(vaddr);
++ if (pgd_none(*pgd) || pgd_bad(*pgd))
++ return ~0UL;
++
++ pud = pud_offset(pgd, vaddr);
++ if (pud_none(*pud) || pud_bad(*pud))
++ return ~0UL;
++
++ pmd = pmd_offset(pud, vaddr);
++ if (pmd_none(*pmd) || pmd_bad(*pmd))
++ return ~0UL;
++
++ ptep = pte_offset_kernel(pmd, vaddr);
++ if (!ptep)
++ return ~0UL;
++
++ return (pte_val(*ptep) & _PFN_MASK) | (vaddr & ~PAGE_MASK);
++ }
++#endif
++
++ if (vaddr > TASK_SIZE) {
++ /* kernel address */
++ return __pa(vaddr);
++ }
++
++
++#ifdef CONFIG_VMX_GUEST
++ /* No privcmd within vmx guest. */
++ return ~0UL;
++#else
++ /* XXX double-check (lack of) locking */
++ vma = find_extend_vma(current->mm, vaddr);
++ if (!vma)
++ return ~0UL;
++
++ /* We assume the page is modified. */
++ page = follow_page(vma, vaddr, FOLL_WRITE | FOLL_TOUCH);
++ if (!page)
++ return ~0UL;
++
++ return (page_to_pfn(page) << PAGE_SHIFT) | (vaddr & ~PAGE_MASK);
++#endif
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/xenentry.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/xenentry.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,931 @@
++/*
++ * ia64/xen/entry.S
++ *
++ * Alternate kernel routines for Xen. Heavily leveraged from
++ * ia64/kernel/entry.S
++ *
++ * Copyright (C) 2005 Hewlett-Packard Co
++ * Dan Magenheimer <dan.magenheimer at .hp.com>
++ */
++
++#include <asm/asmmacro.h>
++#include <asm/cache.h>
++#include <asm/errno.h>
++#include <asm/kregs.h>
++#include <asm/asm-offsets.h>
++#include <asm/pgtable.h>
++#include <asm/percpu.h>
++#include <asm/processor.h>
++#include <asm/thread_info.h>
++#include <asm/unistd.h>
++
++#ifdef CONFIG_XEN
++#include "xenminstate.h"
++#else
++#include "minstate.h"
++#endif
++
++/*
++ * prev_task <- ia64_switch_to(struct task_struct *next)
++ * With Ingo's new scheduler, interrupts are disabled when this routine gets
++ * called. The code starting at .map relies on this. The rest of the code
++ * doesn't care about the interrupt masking status.
++ */
++#ifdef CONFIG_XEN
++GLOBAL_ENTRY(xen_switch_to)
++ .prologue
++ alloc r16=ar.pfs,1,0,0,0
++ movl r22=running_on_xen;;
++ ld4 r22=[r22];;
++ cmp.eq p7,p0=r22,r0
++(p7) br.cond.sptk.many __ia64_switch_to;;
++#else
++GLOBAL_ENTRY(ia64_switch_to)
++ .prologue
++ alloc r16=ar.pfs,1,0,0,0
++#endif
++ DO_SAVE_SWITCH_STACK
++ .body
++
++ adds r22=IA64_TASK_THREAD_KSP_OFFSET,r13
++ movl r25=init_task
++ mov r27=IA64_KR(CURRENT_STACK)
++ adds r21=IA64_TASK_THREAD_KSP_OFFSET,in0
++ dep r20=0,in0,61,3 // physical address of "next"
++ ;;
++ st8 [r22]=sp // save kernel stack pointer of old task
++ shr.u r26=r20,IA64_GRANULE_SHIFT
++ cmp.eq p7,p6=r25,in0
++ ;;
++ /*
++ * If we've already mapped this task's page, we can skip doing it again.
++ */
++(p6) cmp.eq p7,p6=r26,r27
++(p6) br.cond.dpnt .map
++ ;;
++.done:
++ ld8 sp=[r21] // load kernel stack pointer of new task
++#ifdef CONFIG_XEN
++ // update "current" application register
++ mov r8=IA64_KR_CURRENT
++ mov r9=in0;;
++ XEN_HYPER_SET_KR
++#else
++ mov IA64_KR(CURRENT)=in0 // update "current" application register
+#endif
+ mov r8=r13 // return pointer to previously running task
+ mov r13=in0 // set "current" pointer
@@ -28729,33 +32681,10 @@
+#else
+END(ia64_leave_kernel)
+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/xenhpski.c tmp-linux-2.6-xen.patch/arch/ia64/xen/xenhpski.c
---- pristine-linux-2.6.18.2/arch/ia64/xen/xenhpski.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/xenhpski.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,19 @@
-+
-+extern unsigned long xen_get_cpuid(int);
-+
-+int
-+running_on_sim(void)
-+{
-+ int i;
-+ long cpuid[6];
-+
-+ for (i = 0; i < 5; ++i)
-+ cpuid[i] = xen_get_cpuid(i);
-+ if ((cpuid[0] & 0xff) != 'H') return 0;
-+ if ((cpuid[3] & 0xff) != 0x4) return 0;
-+ if (((cpuid[3] >> 8) & 0xff) != 0x0) return 0;
-+ if (((cpuid[3] >> 16) & 0xff) != 0x0) return 0;
-+ if (((cpuid[3] >> 24) & 0x7) != 0x7) return 0;
-+ return 1;
-+}
-+
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/xenivt.S tmp-linux-2.6-xen.patch/arch/ia64/xen/xenivt.S
---- pristine-linux-2.6.18.2/arch/ia64/xen/xenivt.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/xenivt.S 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,2177 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/xenivt.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/xenivt.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,2178 @@
+/*
+ * arch/ia64/xen/ivt.S
+ *
@@ -29454,16 +33383,17 @@
+ st4 [r3]=r14 // vpsr.ic = 1
+ adds r3=8,r2 // set up second base pointer
+ ;;
++ sum PSR_DEFAULT_BITS
+#else
+ mov out0=cr.ifa
+ mov out1=cr.isr
+ adds r3=8,r2 // set up second base pointer
+ ;;
+ ssm psr.ic | PSR_DEFAULT_BITS
++#endif
+ ;;
+ srlz.i // guarantee that interruption collectin is on
+ ;;
-+#endif
+#ifdef CONFIG_XEN
+
+#define MASK_TO_PEND_OFS (-1)
@@ -29969,7 +33899,7 @@
+ alloc r14=ar.pfs,0,0,2,0 // must be first in an insn group
+#ifdef CONFIG_XEN
+ ;;
-+ br.call.sptk.many rp=xen_get_ivr
++ XEN_HYPER_GET_IVR
+ ;;
+ mov out0=r8 // pass cr.ivr as first arg
+#else
@@ -30933,9 +34863,9 @@
+
+
+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/xenminstate.h tmp-linux-2.6-xen.patch/arch/ia64/xen/xenminstate.h
---- pristine-linux-2.6.18.2/arch/ia64/xen/xenminstate.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/xenminstate.h 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/xenminstate.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/xenminstate.h Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,358 @@
+#include <asm/cache.h>
+
@@ -31295,9 +35225,9 @@
+#else
+#define SAVE_MIN DO_SAVE_MIN( , mov r30=r0, )
+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/xenpal.S tmp-linux-2.6-xen.patch/arch/ia64/xen/xenpal.S
---- pristine-linux-2.6.18.2/arch/ia64/xen/xenpal.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/xenpal.S 2007-07-30 16:35:11.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/xenpal.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/xenpal.S Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,85 @@
+/*
+ * ia64/xen/xenpal.S
@@ -31384,10 +35314,10 @@
+ srlz.d // seralize restoration of psr.l
+ br.ret.sptk.many b0
+END(xen_pal_call_static)
-diff -Nurp pristine-linux-2.6.18.2/arch/ia64/xen/xensetup.S tmp-linux-2.6-xen.patch/arch/ia64/xen/xensetup.S
---- pristine-linux-2.6.18.2/arch/ia64/xen/xensetup.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/ia64/xen/xensetup.S 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,52 @@
+diff -r d894e36cfc30 -r 0aa021803deb arch/ia64/xen/xensetup.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/ia64/xen/xensetup.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,40 @@
+/*
+ * Support routines for Xen
+ *
@@ -31397,6 +35327,13 @@
+#include <asm/processor.h>
+#include <asm/asmmacro.h>
+
++ .section .data.read_mostly
++ .align 8
++ .global running_on_xen
++running_on_xen:
++ data4 0
++ .previous
++
+#define isBP p3 // are we the Bootstrap Processor?
+
+ .text
@@ -31421,19852 +35358,51883 @@
+ br.ret.sptk.many rp
+ ;;
+END(early_xen_setup)
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/Kconfig
+--- a/arch/powerpc/Kconfig Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -438,6 +438,17 @@
+ config UDBG_RTAS_CONSOLE
+ bool
+ default n
+
-+#include <xen/interface/xen.h>
-+
-+/* Stub for suspend.
-+ Just force the stacked registers to be written in memory. */
-+GLOBAL_ENTRY(xencomm_arch_hypercall_suspend)
-+ ;;
-+ alloc r20=ar.pfs,0,0,6,0
-+ mov r2=__HYPERVISOR_sched_op
-+ ;;
-+ /* We don't want to deal with RSE. */
-+ flushrs
-+ mov r33=r32
-+ mov r32=2 // SCHEDOP_shutdown
-+ ;;
-+ break 0x1000
-+ ;;
-+ br.ret.sptk.many b0
-+END(xencomm_arch_hypercall_suspend)
-diff -Nurp pristine-linux-2.6.18.2/arch/um/kernel/physmem.c tmp-linux-2.6-xen.patch/arch/um/kernel/physmem.c
---- pristine-linux-2.6.18.2/arch/um/kernel/physmem.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/um/kernel/physmem.c 2007-07-30 16:35:11.000000000 +0200
-@@ -226,7 +226,7 @@ EXPORT_SYMBOL(physmem_forget_descriptor)
- EXPORT_SYMBOL(physmem_remove_mapping);
- EXPORT_SYMBOL(physmem_subst_mapping);
++config PPC_XEN
++ bool "Enable Xen compatible kernel"
++ depends on PPC_MULTIPLATFORM && PPC64 && PPC_MAPLE && PPC_PSERIES && SMP
++ select XEN
++ select XEN_PRIVILEGED_GUEST
++ select XEN_UNPRIVILEGED_GUEST
++ select XEN_XENCOMM
++
++ help
++ This option will compile a kernel compatible with Xen hypervisor
--void arch_free_page(struct page *page, int order)
-+int arch_free_page(struct page *page, int order)
- {
- void *virt;
- int i;
-@@ -235,6 +235,8 @@ void arch_free_page(struct page *page, i
- virt = __va(page_to_phys(page + i));
- physmem_remove_mapping(virt);
- }
-+
-+ return 0;
- }
+ config XICS
+ depends on PPC_PSERIES
+@@ -1080,6 +1091,8 @@
- int is_remapped(void *virt)
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/ia32/ia32entry-xen.S tmp-linux-2.6-xen.patch/arch/x86_64/ia32/ia32entry-xen.S
---- pristine-linux-2.6.18.2/arch/x86_64/ia32/ia32entry-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/ia32/ia32entry-xen.S 2007-10-14 01:46:51.000000000 +0200
-@@ -0,0 +1,755 @@
-+/*
-+ * Compatibility mode system call entry point for x86-64.
-+ *
-+ * Copyright 2000-2002 Andi Kleen, SuSE Labs.
-+ */
+ source "security/Kconfig"
+
++source "drivers/xen/Kconfig"
+
-+#include <asm/dwarf2.h>
-+#include <asm/calling.h>
-+#include <asm/asm-offsets.h>
-+#include <asm/current.h>
-+#include <asm/errno.h>
-+#include <asm/ia32_unistd.h>
-+#include <asm/thread_info.h>
-+#include <asm/segment.h>
-+#include <asm/vsyscall32.h>
-+#include <asm/irqflags.h>
-+#include <linux/linkage.h>
+ config KEYS_COMPAT
+ bool
+ depends on COMPAT && KEYS
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/Kconfig.debug
+--- a/arch/powerpc/Kconfig.debug Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/Kconfig.debug Wed Sep 10 10:54:08 2008 +0100
+@@ -160,6 +160,20 @@
+ Select this to enable early debugging for legacy iSeries. You need
+ to hit "Ctrl-x Ctrl-x" to see the messages on the console.
+
++config PPC_EARLY_DEBUG_XEN_DOM0
++ bool "Xen Dom0 Console"
++ depends on PPC_XEN
++ help
++ Select this to enable early debugging for Xen Dom0. Setting
++ this will result in a kernel that may not work as a DomU.
+
-+#define __XEN_X86_64 1
++config PPC_EARLY_DEBUG_XEN_DOMU
++ bool "Xen DomU Console"
++ depends on PPC_XEN && XEN_UNPRIVILEGED_GUEST
++ help
++ Select this to enable early debugging for Xen DomU. Setting
++ this will result in a kernel that may not work as a Dom0.
+
-+#define IA32_NR_syscalls ((ia32_syscall_end - ia32_sys_call_table)/8)
+ endchoice
+
+ endmenu
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/Makefile
+--- a/arch/powerpc/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -65,6 +65,7 @@
+ AFLAGS-$(CONFIG_PPC32) := -Iarch/$(ARCH)
+ CFLAGS-$(CONFIG_PPC64) := -mminimal-toc -mtraceback=none -mcall-aixdesc
+ CFLAGS-$(CONFIG_PPC32) := -Iarch/$(ARCH) -ffixed-r2 -mmultiple
++CFLAGS-$(CONFIG_PPC_XEN) += -Iinclude/asm-$(ARCH)/xen
+ CPPFLAGS += $(CPPFLAGS-y)
+ AFLAGS += $(AFLAGS-y)
+ CFLAGS += -msoft-float -pipe $(CFLAGS-y)
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/boot/Makefile
+--- a/arch/powerpc/boot/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/boot/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -36,8 +36,11 @@
+ $(addprefix $(obj)/,$(zlib) main.o): $(addprefix $(obj)/,$(zliblinuxheader)) $(addprefix $(obj)/,$(zlibheader))
+ #$(addprefix $(obj)/,main.o): $(addprefix $(obj)/,zlib.h)
+
++xen_guest-y = xen_guest.S
++
+ src-boot := crt0.S string.S prom.c stdio.c main.c div64.S
+ src-boot += $(zlib)
++src-boot += $(xen_guest-$(CONFIG_XEN))
+ src-boot := $(addprefix $(obj)/, $(src-boot))
+ obj-boot := $(addsuffix .o, $(basename $(src-boot)))
+
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/boot/xen_guest.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/boot/xen_guest.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,7 @@
++ .section __xen_guest
++ .ascii "GUEST_OS=linux"
++ .ascii ",GUEST_VER=xen-3.0"
++ .ascii ",XEN_VER=xen-3.0"
++ .ascii ",VIRT_BASE=0x0"
++ .ascii ",LOADER=generic"
++ .byte 0
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/configs/xen_maple_defconfig
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/configs/xen_maple_defconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1342 @@
++#
++# Automatically generated make config: don't edit
++# Linux kernel version: 2.6.17
++# Mon Jan 15 23:48:47 2007
++#
++CONFIG_PPC64=y
++CONFIG_64BIT=y
++CONFIG_PPC_MERGE=y
++CONFIG_MMU=y
++CONFIG_GENERIC_HARDIRQS=y
++CONFIG_IRQ_PER_CPU=y
++CONFIG_RWSEM_XCHGADD_ALGORITHM=y
++CONFIG_GENERIC_HWEIGHT=y
++CONFIG_GENERIC_CALIBRATE_DELAY=y
++CONFIG_GENERIC_FIND_NEXT_BIT=y
++CONFIG_PPC=y
++CONFIG_EARLY_PRINTK=y
++CONFIG_COMPAT=y
++CONFIG_SYSVIPC_COMPAT=y
++CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER=y
++CONFIG_ARCH_MAY_HAVE_PC_FDC=y
++CONFIG_PPC_OF=y
++CONFIG_PPC_UDBG_16550=y
++CONFIG_GENERIC_TBSYNC=y
++# CONFIG_DEFAULT_UIMAGE is not set
+
-+ .macro IA32_ARG_FIXUP noebp=0
-+ movl %edi,%r8d
-+ .if \noebp
-+ .else
-+ movl %ebp,%r9d
-+ .endif
-+ xchg %ecx,%esi
-+ movl %ebx,%edi
-+ movl %edx,%edx /* zero extension */
-+ .endm
++#
++# Processor support
++#
++CONFIG_POWER4_ONLY=y
++CONFIG_POWER4=y
++CONFIG_PPC_FPU=y
++CONFIG_ALTIVEC=y
++CONFIG_PPC_STD_MMU=y
++CONFIG_VIRT_CPU_ACCOUNTING=y
++CONFIG_SMP=y
++CONFIG_NR_CPUS=32
++CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config"
+
-+ /* clobbers %eax */
-+ .macro CLEAR_RREGS
-+ xorl %eax,%eax
-+ movq %rax,R11(%rsp)
-+ movq %rax,R10(%rsp)
-+ movq %rax,R9(%rsp)
-+ movq %rax,R8(%rsp)
-+ .endm
++#
++# Code maturity level options
++#
++CONFIG_EXPERIMENTAL=y
++CONFIG_LOCK_KERNEL=y
++CONFIG_INIT_ENV_ARG_LIMIT=32
+
-+ .macro LOAD_ARGS32 offset
-+ movl \offset(%rsp),%r11d
-+ movl \offset+8(%rsp),%r10d
-+ movl \offset+16(%rsp),%r9d
-+ movl \offset+24(%rsp),%r8d
-+ movl \offset+40(%rsp),%ecx
-+ movl \offset+48(%rsp),%edx
-+ movl \offset+56(%rsp),%esi
-+ movl \offset+64(%rsp),%edi
-+ movl \offset+72(%rsp),%eax
-+ .endm
++#
++# General setup
++#
++CONFIG_LOCALVERSION="-Xen"
++CONFIG_LOCALVERSION_AUTO=y
++CONFIG_SWAP=y
++CONFIG_SYSVIPC=y
++# CONFIG_POSIX_MQUEUE is not set
++# CONFIG_BSD_PROCESS_ACCT is not set
++CONFIG_SYSCTL=y
++# CONFIG_AUDIT is not set
++CONFIG_IKCONFIG=y
++CONFIG_IKCONFIG_PROC=y
++# CONFIG_CPUSETS is not set
++# CONFIG_RELAY is not set
++CONFIG_INITRAMFS_SOURCE=""
++# CONFIG_CC_OPTIMIZE_FOR_SIZE is not set
++# CONFIG_EMBEDDED is not set
++CONFIG_KALLSYMS=y
++# CONFIG_KALLSYMS_ALL is not set
++# CONFIG_KALLSYMS_EXTRA_PASS is not set
++CONFIG_HOTPLUG=y
++CONFIG_PRINTK=y
++CONFIG_BUG=y
++CONFIG_ELF_CORE=y
++CONFIG_BASE_FULL=y
++CONFIG_RT_MUTEXES=y
++CONFIG_FUTEX=y
++CONFIG_EPOLL=y
++CONFIG_SHMEM=y
++CONFIG_SLAB=y
++# CONFIG_TINY_SHMEM is not set
++CONFIG_BASE_SMALL=0
++# CONFIG_SLOB is not set
+
-+#if defined (__XEN_X86_64)
-+#include "../kernel/xen_entry.S"
-+
-+#define __swapgs
-+#define __cli
-+#define __sti
-+#else
-+/*
-+ * Use the native instructions
-+ */
-+#define __swapgs swapgs
-+#define __cli cli
-+#define __sti sti
-+#endif
++#
++# Loadable module support
++#
++CONFIG_MODULES=y
++CONFIG_MODULE_UNLOAD=y
++# CONFIG_MODULE_FORCE_UNLOAD is not set
++CONFIG_MODVERSIONS=y
++CONFIG_MODULE_SRCVERSION_ALL=y
++CONFIG_KMOD=y
++CONFIG_STOP_MACHINE=y
+
-+ .macro CFI_STARTPROC32 simple
-+ CFI_STARTPROC \simple
-+ CFI_UNDEFINED r8
-+ CFI_UNDEFINED r9
-+ CFI_UNDEFINED r10
-+ CFI_UNDEFINED r11
-+ CFI_UNDEFINED r12
-+ CFI_UNDEFINED r13
-+ CFI_UNDEFINED r14
-+ CFI_UNDEFINED r15
-+ .endm
++#
++# Block layer
++#
++# CONFIG_BLK_DEV_IO_TRACE is not set
+
-+/*
-+ * 32bit SYSENTER instruction entry.
-+ *
-+ * Arguments:
-+ * %eax System call number.
-+ * %ebx Arg1
-+ * %ecx Arg2
-+ * %edx Arg3
-+ * %esi Arg4
-+ * %edi Arg5
-+ * %ebp user stack
-+ * 0(%ebp) Arg6
-+ *
-+ * Interrupts off.
-+ *
-+ * This is purely a fast path. For anything complicated we use the int 0x80
-+ * path below. Set up a complete hardware stack frame to share code
-+ * with the int 0x80 path.
-+ */
-+ENTRY(ia32_sysenter_target)
-+ CFI_STARTPROC32 simple
-+ CFI_DEF_CFA rsp,0
-+ CFI_REGISTER rsp,rbp
-+ __swapgs
-+ movq %gs:pda_kernelstack, %rsp
-+ addq $(PDA_STACKOFFSET),%rsp
-+ /*
-+ * No need to follow this irqs on/off section: the syscall
-+ * disabled irqs, here we enable it straight after entry:
-+ */
-+ XEN_UNBLOCK_EVENTS(%r11)
-+ __sti
-+ movl %ebp,%ebp /* zero extension */
-+ pushq $__USER32_DS
-+ CFI_ADJUST_CFA_OFFSET 8
-+ /*CFI_REL_OFFSET ss,0*/
-+ pushq %rbp
-+ CFI_ADJUST_CFA_OFFSET 8
-+ CFI_REL_OFFSET rsp,0
-+ pushfq
-+ CFI_ADJUST_CFA_OFFSET 8
-+ /*CFI_REL_OFFSET rflags,0*/
-+ movl $VSYSCALL32_SYSEXIT, %r10d
-+ CFI_REGISTER rip,r10
-+ pushq $__USER32_CS
-+ CFI_ADJUST_CFA_OFFSET 8
-+ /*CFI_REL_OFFSET cs,0*/
-+ movl %eax, %eax
-+ pushq %r10
-+ CFI_ADJUST_CFA_OFFSET 8
-+ CFI_REL_OFFSET rip,0
-+ pushq %rax
-+ CFI_ADJUST_CFA_OFFSET 8
-+ cld
-+ SAVE_ARGS 0,0,0
-+ /* no need to do an access_ok check here because rbp has been
-+ 32bit zero extended */
-+1: movl (%rbp),%r9d
-+ .section __ex_table,"a"
-+ .quad 1b,ia32_badarg
-+ .previous
-+ GET_THREAD_INFO(%r10)
-+ orl $TS_COMPAT,threadinfo_status(%r10)
-+ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%r10)
-+ CFI_REMEMBER_STATE
-+ jnz sysenter_tracesys
-+sysenter_do_call:
-+ cmpl $(IA32_NR_syscalls-1),%eax
-+ ja ia32_badsys
-+ IA32_ARG_FIXUP 1
-+ call *ia32_sys_call_table(,%rax,8)
-+ movq %rax,RAX-ARGOFFSET(%rsp)
-+ GET_THREAD_INFO(%r10)
-+ XEN_BLOCK_EVENTS(%r11)
-+ __cli
-+ TRACE_IRQS_OFF
-+ testl $_TIF_ALLWORK_MASK,threadinfo_flags(%r10)
-+ jnz int_ret_from_sys_call
-+ andl $~TS_COMPAT,threadinfo_status(%r10)
-+ /* clear IF, that popfq doesn't enable interrupts early */
-+ andl $~0x200,EFLAGS-R11(%rsp)
-+ RESTORE_ARGS 1,24,1,1,1,1
-+ popfq
-+ CFI_ADJUST_CFA_OFFSET -8
-+ /*CFI_RESTORE rflags*/
-+ popq %rcx /* User %esp */
-+ CFI_ADJUST_CFA_OFFSET -8
-+ CFI_REGISTER rsp,rcx
-+ movl $VSYSCALL32_SYSEXIT,%edx /* User %eip */
-+ CFI_REGISTER rip,rdx
-+ TRACE_IRQS_ON
-+ __swapgs
-+ XEN_UNBLOCK_EVENTS(%r11)
-+ __sti /* sti only takes effect after the next instruction */
-+ /* sysexit */
-+ .byte 0xf, 0x35 /* TBD */
++#
++# IO Schedulers
++#
++CONFIG_IOSCHED_NOOP=y
++# CONFIG_IOSCHED_AS is not set
++CONFIG_IOSCHED_DEADLINE=y
++# CONFIG_IOSCHED_CFQ is not set
++# CONFIG_DEFAULT_AS is not set
++CONFIG_DEFAULT_DEADLINE=y
++# CONFIG_DEFAULT_CFQ is not set
++# CONFIG_DEFAULT_NOOP is not set
++CONFIG_DEFAULT_IOSCHED="deadline"
+
-+sysenter_tracesys:
-+ CFI_RESTORE_STATE
-+ SAVE_REST
-+ CLEAR_RREGS
-+ movq $-ENOSYS,RAX(%rsp) /* really needed? */
-+ movq %rsp,%rdi /* &pt_regs -> arg1 */
-+ call syscall_trace_enter
-+ LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */
-+ RESTORE_REST
-+ movl %ebp, %ebp
-+ /* no need to do an access_ok check here because rbp has been
-+ 32bit zero extended */
-+1: movl (%rbp),%r9d
-+ .section __ex_table,"a"
-+ .quad 1b,ia32_badarg
-+ .previous
-+ jmp sysenter_do_call
-+ CFI_ENDPROC
-+ENDPROC(ia32_sysenter_target)
++#
++# Platform support
++#
++CONFIG_PPC_MULTIPLATFORM=y
++# CONFIG_PPC_ISERIES is not set
++# CONFIG_EMBEDDED6xx is not set
++# CONFIG_APUS is not set
++CONFIG_PPC_PSERIES=y
++# CONFIG_PPC_PMAC is not set
++CONFIG_PPC_MAPLE=y
++# CONFIG_PPC_CELL is not set
++# CONFIG_PPC_CELL_NATIVE is not set
++# CONFIG_PPC_IBM_CELL_BLADE is not set
++# CONFIG_UDBG_RTAS_CONSOLE is not set
++CONFIG_PPC_XEN=y
++CONFIG_XICS=y
++CONFIG_U3_DART=y
++CONFIG_MPIC=y
++CONFIG_PPC_RTAS=y
++CONFIG_RTAS_ERROR_LOGGING=y
++CONFIG_RTAS_PROC=y
++CONFIG_RTAS_FLASH=y
++# CONFIG_MMIO_NVRAM is not set
++CONFIG_MPIC_BROKEN_U3=y
++CONFIG_IBMVIO=y
++# CONFIG_IBMEBUS is not set
++# CONFIG_PPC_MPC106 is not set
++CONFIG_PPC_970_NAP=y
++# CONFIG_CPU_FREQ is not set
++# CONFIG_WANT_EARLY_SERIAL is not set
+
-+/*
-+ * 32bit SYSCALL instruction entry.
-+ *
-+ * Arguments:
-+ * %eax System call number.
-+ * %ebx Arg1
-+ * %ecx return EIP
-+ * %edx Arg3
-+ * %esi Arg4
-+ * %edi Arg5
-+ * %ebp Arg2 [note: not saved in the stack frame, should not be touched]
-+ * %esp user stack
-+ * 0(%esp) Arg6
-+ *
-+ * Interrupts off.
-+ *
-+ * This is purely a fast path. For anything complicated we use the int 0x80
-+ * path below. Set up a complete hardware stack frame to share code
-+ * with the int 0x80 path.
-+ */
-+ENTRY(ia32_cstar_target)
-+ CFI_STARTPROC32 simple
-+ CFI_DEF_CFA rsp,PDA_STACKOFFSET
-+ CFI_REGISTER rip,rcx
-+ /*CFI_REGISTER rflags,r11*/
-+ __swapgs
-+ movl %esp,%r8d
-+ CFI_REGISTER rsp,r8
-+ movq %gs:pda_kernelstack,%rsp
-+ /*
-+ * No need to follow this irqs on/off section: the syscall
-+ * disabled irqs and here we enable it straight after entry:
-+ */
-+ XEN_UNBLOCK_EVENTS(%r11)
-+ __sti
-+ SAVE_ARGS 8,1,1
-+ movl %eax,%eax /* zero extension */
-+ movq %rax,ORIG_RAX-ARGOFFSET(%rsp)
-+ movq %rcx,RIP-ARGOFFSET(%rsp)
-+ CFI_REL_OFFSET rip,RIP-ARGOFFSET
-+ movq %rbp,RCX-ARGOFFSET(%rsp) /* this lies slightly to ptrace */
-+ movl %ebp,%ecx
-+ movq $__USER32_CS,CS-ARGOFFSET(%rsp)
-+ movq $__USER32_DS,SS-ARGOFFSET(%rsp)
-+ movq %r11,EFLAGS-ARGOFFSET(%rsp)
-+ /*CFI_REL_OFFSET rflags,EFLAGS-ARGOFFSET*/
-+ movq %r8,RSP-ARGOFFSET(%rsp)
-+ CFI_REL_OFFSET rsp,RSP-ARGOFFSET
-+ /* no need to do an access_ok check here because r8 has been
-+ 32bit zero extended */
-+ /* hardware stack frame is complete now */
-+1: movl (%r8),%r9d
-+ .section __ex_table,"a"
-+ .quad 1b,ia32_badarg
-+ .previous
-+ GET_THREAD_INFO(%r10)
-+ orl $TS_COMPAT,threadinfo_status(%r10)
-+ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%r10)
-+ CFI_REMEMBER_STATE
-+ jnz cstar_tracesys
-+cstar_do_call:
-+ cmpl $IA32_NR_syscalls-1,%eax
-+ ja ia32_badsys
-+ IA32_ARG_FIXUP 1
-+ call *ia32_sys_call_table(,%rax,8)
-+ movq %rax,RAX-ARGOFFSET(%rsp)
-+ GET_THREAD_INFO(%r10)
-+ XEN_BLOCK_EVENTS(%r11)
-+ __cli
-+ TRACE_IRQS_OFF
-+ testl $_TIF_ALLWORK_MASK,threadinfo_flags(%r10)
-+ jnz int_ret_from_sys_call
-+ andl $~TS_COMPAT,threadinfo_status(%r10)
-+ RESTORE_ARGS 1,-ARG_SKIP,1,1,1
-+ movl RIP-ARGOFFSET(%rsp),%ecx
-+ CFI_REGISTER rip,rcx
-+ movl EFLAGS-ARGOFFSET(%rsp),%r11d
-+ /*CFI_REGISTER rflags,r11*/
-+ TRACE_IRQS_ON
-+ movl RSP-ARGOFFSET(%rsp),%esp
-+ CFI_RESTORE rsp
-+ __swapgs
-+ sysretl /* TBD */
-+
-+cstar_tracesys:
-+ CFI_RESTORE_STATE
-+ SAVE_REST
-+ CLEAR_RREGS
-+ movq $-ENOSYS,RAX(%rsp) /* really needed? */
-+ movq %rsp,%rdi /* &pt_regs -> arg1 */
-+ call syscall_trace_enter
-+ LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */
-+ RESTORE_REST
-+ movl RSP-ARGOFFSET(%rsp), %r8d
-+ /* no need to do an access_ok check here because r8 has been
-+ 32bit zero extended */
-+1: movl (%r8),%r9d
-+ .section __ex_table,"a"
-+ .quad 1b,ia32_badarg
-+ .previous
-+ jmp cstar_do_call
-+END(ia32_cstar_target)
-+
-+ia32_badarg:
-+ movq $-EFAULT,%rax
-+ jmp ia32_sysret
-+ CFI_ENDPROC
++#
++# Kernel options
++#
++CONFIG_HZ_100=y
++# CONFIG_HZ_250 is not set
++# CONFIG_HZ_1000 is not set
++CONFIG_HZ=100
++CONFIG_PREEMPT_NONE=y
++# CONFIG_PREEMPT_VOLUNTARY is not set
++# CONFIG_PREEMPT is not set
++# CONFIG_PREEMPT_BKL is not set
++CONFIG_BINFMT_ELF=y
++# CONFIG_BINFMT_MISC is not set
++CONFIG_FORCE_MAX_ZONEORDER=13
++CONFIG_IOMMU_VMERGE=y
++# CONFIG_HOTPLUG_CPU is not set
++CONFIG_ARCH_ENABLE_MEMORY_HOTPLUG=y
++# CONFIG_KEXEC is not set
++# CONFIG_CRASH_DUMP is not set
++CONFIG_IRQ_ALL_CPUS=y
++# CONFIG_PPC_SPLPAR is not set
++CONFIG_EEH=y
++CONFIG_SCANLOG=y
++CONFIG_LPARCFG=y
++CONFIG_NUMA=y
++CONFIG_NODES_SHIFT=4
++CONFIG_ARCH_SELECT_MEMORY_MODEL=y
++CONFIG_ARCH_SPARSEMEM_ENABLE=y
++CONFIG_ARCH_SPARSEMEM_DEFAULT=y
++CONFIG_SELECT_MEMORY_MODEL=y
++# CONFIG_FLATMEM_MANUAL is not set
++# CONFIG_DISCONTIGMEM_MANUAL is not set
++CONFIG_SPARSEMEM_MANUAL=y
++CONFIG_SPARSEMEM=y
++CONFIG_NEED_MULTIPLE_NODES=y
++CONFIG_HAVE_MEMORY_PRESENT=y
++# CONFIG_SPARSEMEM_STATIC is not set
++CONFIG_SPARSEMEM_EXTREME=y
++CONFIG_MEMORY_HOTPLUG=y
++CONFIG_SPLIT_PTLOCK_CPUS=4
++CONFIG_MIGRATION=y
++CONFIG_RESOURCES_64BIT=y
++CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID=y
++CONFIG_ARCH_MEMORY_PROBE=y
++# CONFIG_PPC_64K_PAGES is not set
++# CONFIG_SCHED_SMT is not set
++CONFIG_PROC_DEVICETREE=y
++# CONFIG_CMDLINE_BOOL is not set
++# CONFIG_PM is not set
++CONFIG_SECCOMP=y
++CONFIG_ISA_DMA_API=y
+
-+/*
-+ * Emulated IA32 system calls via int 0x80.
-+ *
-+ * Arguments:
-+ * %eax System call number.
-+ * %ebx Arg1
-+ * %ecx Arg2
-+ * %edx Arg3
-+ * %esi Arg4
-+ * %edi Arg5
-+ * %ebp Arg6 [note: not saved in the stack frame, should not be touched]
-+ *
-+ * Notes:
-+ * Uses the same stack frame as the x86-64 version.
-+ * All registers except %eax must be saved (but ptrace may violate that)
-+ * Arguments are zero extended. For system calls that want sign extension and
-+ * take long arguments a wrapper is needed. Most calls can just be called
-+ * directly.
-+ * Assumes it is only called from user space and entered with interrupts off.
-+ */
++#
++# Bus options
++#
++CONFIG_GENERIC_ISA_DMA=y
++CONFIG_PPC_I8259=y
++# CONFIG_PPC_INDIRECT_PCI is not set
++CONFIG_PCI=y
++CONFIG_PCI_DOMAINS=y
++# CONFIG_PCIEPORTBUS is not set
++# CONFIG_PCI_DEBUG is not set
+
-+ENTRY(ia32_syscall)
-+ CFI_STARTPROC simple
-+ CFI_DEF_CFA rsp,SS+8-RIP
-+ /*CFI_REL_OFFSET ss,SS-RIP*/
-+ CFI_REL_OFFSET rsp,RSP-RIP
-+ /*CFI_REL_OFFSET rflags,EFLAGS-RIP*/
-+ /*CFI_REL_OFFSET cs,CS-RIP*/
-+ CFI_REL_OFFSET rip,RIP-RIP
-+ __swapgs
-+ /*
-+ * No need to follow this irqs on/off section: the syscall
-+ * disabled irqs and here we enable it straight after entry:
-+ */
-+ XEN_UNBLOCK_EVENTS(%r11)
-+ __sti
-+ movq (%rsp),%rcx
-+ movq 8(%rsp),%r11
-+ addq $0x10,%rsp /* skip rcx and r11 */
-+ movl %eax,%eax
-+ pushq %rax
-+ CFI_ADJUST_CFA_OFFSET 8
-+ cld
-+/* 1: jmp 1b */
-+ /* note the registers are not zero extended to the sf.
-+ this could be a problem. */
-+ SAVE_ARGS 0,0,1
-+ GET_THREAD_INFO(%r10)
-+ orl $TS_COMPAT,threadinfo_status(%r10)
-+ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%r10)
-+ jnz ia32_tracesys
-+ia32_do_syscall:
-+ cmpl $(IA32_NR_syscalls-1),%eax
-+ ja ia32_badsys
-+ IA32_ARG_FIXUP
-+ call *ia32_sys_call_table(,%rax,8) # xxx: rip relative
-+ia32_sysret:
-+ movq %rax,RAX-ARGOFFSET(%rsp)
-+ jmp int_ret_from_sys_call
++#
++# PCCARD (PCMCIA/CardBus) support
++#
++# CONFIG_PCCARD is not set
+
-+ia32_tracesys:
-+ SAVE_REST
-+ movq $-ENOSYS,RAX(%rsp) /* really needed? */
-+ movq %rsp,%rdi /* &pt_regs -> arg1 */
-+ call syscall_trace_enter
-+ LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */
-+ RESTORE_REST
-+ jmp ia32_do_syscall
-+END(ia32_syscall)
++#
++# PCI Hotplug Support
++#
++# CONFIG_HOTPLUG_PCI is not set
++CONFIG_KERNEL_START=0xc000000000000000
+
-+ia32_badsys:
-+ movq $0,ORIG_RAX-ARGOFFSET(%rsp)
-+ movq $-ENOSYS,RAX-ARGOFFSET(%rsp)
-+ jmp int_ret_from_sys_call
++#
++# Networking
++#
++CONFIG_NET=y
+
-+quiet_ni_syscall:
-+ movq $-ENOSYS,%rax
-+ ret
-+ CFI_ENDPROC
-+
-+ .macro PTREGSCALL label, func, arg
-+ .globl \label
-+\label:
-+ leaq \func(%rip),%rax
-+ leaq -ARGOFFSET+8(%rsp),\arg /* 8 for return address */
-+ jmp ia32_ptregs_common
-+ .endm
++#
++# Networking options
++#
++# CONFIG_NETDEBUG is not set
++CONFIG_PACKET=y
++CONFIG_PACKET_MMAP=y
++CONFIG_UNIX=y
++CONFIG_XFRM=y
++# CONFIG_XFRM_USER is not set
++# CONFIG_NET_KEY is not set
++CONFIG_INET=y
++CONFIG_IP_MULTICAST=y
++CONFIG_IP_ADVANCED_ROUTER=y
++CONFIG_ASK_IP_FIB_HASH=y
++# CONFIG_IP_FIB_TRIE is not set
++CONFIG_IP_FIB_HASH=y
++CONFIG_IP_MULTIPLE_TABLES=y
++CONFIG_IP_ROUTE_FWMARK=y
++CONFIG_IP_ROUTE_MULTIPATH=y
++# CONFIG_IP_ROUTE_MULTIPATH_CACHED is not set
++# CONFIG_IP_ROUTE_VERBOSE is not set
++CONFIG_IP_PNP=y
++CONFIG_IP_PNP_DHCP=y
++CONFIG_IP_PNP_BOOTP=y
++CONFIG_IP_PNP_RARP=y
++CONFIG_NET_IPIP=y
++# CONFIG_NET_IPGRE is not set
++CONFIG_IP_MROUTE=y
++CONFIG_IP_PIMSM_V1=y
++CONFIG_IP_PIMSM_V2=y
++# CONFIG_ARPD is not set
++CONFIG_SYN_COOKIES=y
++# CONFIG_INET_AH is not set
++# CONFIG_INET_ESP is not set
++# CONFIG_INET_IPCOMP is not set
++# CONFIG_INET_XFRM_TUNNEL is not set
++CONFIG_INET_TUNNEL=y
++CONFIG_INET_XFRM_MODE_TRANSPORT=y
++CONFIG_INET_XFRM_MODE_TUNNEL=y
++CONFIG_INET_DIAG=y
++CONFIG_INET_TCP_DIAG=y
++# CONFIG_TCP_CONG_ADVANCED is not set
++CONFIG_TCP_CONG_BIC=y
+
-+ CFI_STARTPROC32
++#
++# IP: Virtual Server Configuration
++#
++# CONFIG_IP_VS is not set
++CONFIG_IPV6=y
++CONFIG_IPV6_PRIVACY=y
++CONFIG_IPV6_ROUTER_PREF=y
++CONFIG_IPV6_ROUTE_INFO=y
++CONFIG_INET6_AH=y
++CONFIG_INET6_ESP=y
++CONFIG_INET6_IPCOMP=y
++CONFIG_INET6_XFRM_TUNNEL=y
++CONFIG_INET6_TUNNEL=y
++CONFIG_INET6_XFRM_MODE_TRANSPORT=y
++CONFIG_INET6_XFRM_MODE_TUNNEL=y
++CONFIG_IPV6_TUNNEL=y
++# CONFIG_NETWORK_SECMARK is not set
++CONFIG_NETFILTER=y
++# CONFIG_NETFILTER_DEBUG is not set
++CONFIG_BRIDGE_NETFILTER=y
+
-+ PTREGSCALL stub32_rt_sigreturn, sys32_rt_sigreturn, %rdi
-+ PTREGSCALL stub32_sigreturn, sys32_sigreturn, %rdi
-+ PTREGSCALL stub32_sigaltstack, sys32_sigaltstack, %rdx
-+ PTREGSCALL stub32_sigsuspend, sys32_sigsuspend, %rcx
-+ PTREGSCALL stub32_execve, sys32_execve, %rcx
-+ PTREGSCALL stub32_fork, sys_fork, %rdi
-+ PTREGSCALL stub32_clone, sys32_clone, %rdx
-+ PTREGSCALL stub32_vfork, sys_vfork, %rdi
-+ PTREGSCALL stub32_iopl, sys_iopl, %rsi
-+ PTREGSCALL stub32_rt_sigsuspend, sys_rt_sigsuspend, %rdx
++#
++# Core Netfilter Configuration
++#
++CONFIG_NETFILTER_NETLINK=y
++CONFIG_NETFILTER_NETLINK_QUEUE=y
++CONFIG_NETFILTER_NETLINK_LOG=y
++CONFIG_NETFILTER_XTABLES=y
++CONFIG_NETFILTER_XT_TARGET_CLASSIFY=y
++CONFIG_NETFILTER_XT_TARGET_CONNMARK=y
++CONFIG_NETFILTER_XT_TARGET_MARK=y
++CONFIG_NETFILTER_XT_TARGET_NFQUEUE=y
++CONFIG_NETFILTER_XT_TARGET_NOTRACK=y
++CONFIG_NETFILTER_XT_MATCH_COMMENT=y
++CONFIG_NETFILTER_XT_MATCH_CONNBYTES=y
++CONFIG_NETFILTER_XT_MATCH_CONNMARK=y
++CONFIG_NETFILTER_XT_MATCH_CONNTRACK=y
++CONFIG_NETFILTER_XT_MATCH_DCCP=y
++CONFIG_NETFILTER_XT_MATCH_ESP=y
++CONFIG_NETFILTER_XT_MATCH_HELPER=y
++CONFIG_NETFILTER_XT_MATCH_LENGTH=y
++CONFIG_NETFILTER_XT_MATCH_LIMIT=y
++CONFIG_NETFILTER_XT_MATCH_MAC=y
++CONFIG_NETFILTER_XT_MATCH_MARK=y
++CONFIG_NETFILTER_XT_MATCH_POLICY=y
++CONFIG_NETFILTER_XT_MATCH_MULTIPORT=y
++CONFIG_NETFILTER_XT_MATCH_PHYSDEV=y
++CONFIG_NETFILTER_XT_MATCH_PKTTYPE=y
++CONFIG_NETFILTER_XT_MATCH_QUOTA=y
++CONFIG_NETFILTER_XT_MATCH_REALM=y
++CONFIG_NETFILTER_XT_MATCH_SCTP=y
++CONFIG_NETFILTER_XT_MATCH_STATE=y
++CONFIG_NETFILTER_XT_MATCH_STATISTIC=y
++CONFIG_NETFILTER_XT_MATCH_STRING=y
++CONFIG_NETFILTER_XT_MATCH_TCPMSS=y
+
-+ENTRY(ia32_ptregs_common)
-+ popq %r11
-+ CFI_ENDPROC
-+ CFI_STARTPROC32 simple
-+ CFI_DEF_CFA rsp,SS+8-ARGOFFSET
-+ CFI_REL_OFFSET rax,RAX-ARGOFFSET
-+ CFI_REL_OFFSET rcx,RCX-ARGOFFSET
-+ CFI_REL_OFFSET rdx,RDX-ARGOFFSET
-+ CFI_REL_OFFSET rsi,RSI-ARGOFFSET
-+ CFI_REL_OFFSET rdi,RDI-ARGOFFSET
-+ CFI_REL_OFFSET rip,RIP-ARGOFFSET
-+/* CFI_REL_OFFSET cs,CS-ARGOFFSET*/
-+/* CFI_REL_OFFSET rflags,EFLAGS-ARGOFFSET*/
-+ CFI_REL_OFFSET rsp,RSP-ARGOFFSET
-+/* CFI_REL_OFFSET ss,SS-ARGOFFSET*/
-+ SAVE_REST
-+ call *%rax
-+ RESTORE_REST
-+ jmp ia32_sysret /* misbalances the return cache */
-+ CFI_ENDPROC
-+END(ia32_ptregs_common)
++#
++# IP: Netfilter Configuration
++#
++CONFIG_IP_NF_CONNTRACK=y
++CONFIG_IP_NF_CT_ACCT=y
++CONFIG_IP_NF_CONNTRACK_MARK=y
++CONFIG_IP_NF_CONNTRACK_EVENTS=y
++CONFIG_IP_NF_CONNTRACK_NETLINK=y
++CONFIG_IP_NF_CT_PROTO_SCTP=y
++CONFIG_IP_NF_FTP=y
++CONFIG_IP_NF_IRC=y
++# CONFIG_IP_NF_NETBIOS_NS is not set
++CONFIG_IP_NF_TFTP=y
++CONFIG_IP_NF_AMANDA=y
++CONFIG_IP_NF_PPTP=y
++# CONFIG_IP_NF_H323 is not set
++# CONFIG_IP_NF_SIP is not set
++# CONFIG_IP_NF_QUEUE is not set
++CONFIG_IP_NF_IPTABLES=y
++CONFIG_IP_NF_MATCH_IPRANGE=y
++CONFIG_IP_NF_MATCH_TOS=y
++CONFIG_IP_NF_MATCH_RECENT=y
++CONFIG_IP_NF_MATCH_ECN=y
++CONFIG_IP_NF_MATCH_DSCP=y
++CONFIG_IP_NF_MATCH_AH=y
++CONFIG_IP_NF_MATCH_TTL=y
++CONFIG_IP_NF_MATCH_OWNER=y
++CONFIG_IP_NF_MATCH_ADDRTYPE=y
++CONFIG_IP_NF_MATCH_HASHLIMIT=y
++CONFIG_IP_NF_FILTER=y
++CONFIG_IP_NF_TARGET_REJECT=y
++CONFIG_IP_NF_TARGET_LOG=y
++CONFIG_IP_NF_TARGET_ULOG=y
++CONFIG_IP_NF_TARGET_TCPMSS=y
++CONFIG_IP_NF_NAT=y
++CONFIG_IP_NF_NAT_NEEDED=y
++CONFIG_IP_NF_TARGET_MASQUERADE=y
++CONFIG_IP_NF_TARGET_REDIRECT=y
++CONFIG_IP_NF_TARGET_NETMAP=y
++CONFIG_IP_NF_TARGET_SAME=y
++CONFIG_IP_NF_NAT_SNMP_BASIC=y
++CONFIG_IP_NF_NAT_IRC=y
++CONFIG_IP_NF_NAT_FTP=y
++CONFIG_IP_NF_NAT_TFTP=y
++CONFIG_IP_NF_NAT_AMANDA=y
++CONFIG_IP_NF_NAT_PPTP=y
++CONFIG_IP_NF_MANGLE=y
++CONFIG_IP_NF_TARGET_TOS=y
++CONFIG_IP_NF_TARGET_ECN=y
++CONFIG_IP_NF_TARGET_DSCP=y
++CONFIG_IP_NF_TARGET_TTL=y
++CONFIG_IP_NF_TARGET_CLUSTERIP=y
++CONFIG_IP_NF_RAW=y
++CONFIG_IP_NF_ARPTABLES=y
++CONFIG_IP_NF_ARPFILTER=y
++CONFIG_IP_NF_ARP_MANGLE=y
+
-+ .section .rodata,"a"
-+ .align 8
-+ia32_sys_call_table:
-+ .quad sys_restart_syscall
-+ .quad sys_exit
-+ .quad stub32_fork
-+ .quad sys_read
-+ .quad sys_write
-+ .quad compat_sys_open /* 5 */
-+ .quad sys_close
-+ .quad sys32_waitpid
-+ .quad sys_creat
-+ .quad sys_link
-+ .quad sys_unlink /* 10 */
-+ .quad stub32_execve
-+ .quad sys_chdir
-+ .quad compat_sys_time
-+ .quad sys_mknod
-+ .quad sys_chmod /* 15 */
-+ .quad sys_lchown16
-+ .quad quiet_ni_syscall /* old break syscall holder */
-+ .quad sys_stat
-+ .quad sys32_lseek
-+ .quad sys_getpid /* 20 */
-+ .quad compat_sys_mount /* mount */
-+ .quad sys_oldumount /* old_umount */
-+ .quad sys_setuid16
-+ .quad sys_getuid16
-+ .quad compat_sys_stime /* stime */ /* 25 */
-+ .quad sys32_ptrace /* ptrace */
-+ .quad sys_alarm
-+ .quad sys_fstat /* (old)fstat */
-+ .quad sys_pause
-+ .quad compat_sys_utime /* 30 */
-+ .quad quiet_ni_syscall /* old stty syscall holder */
-+ .quad quiet_ni_syscall /* old gtty syscall holder */
-+ .quad sys_access
-+ .quad sys_nice
-+ .quad quiet_ni_syscall /* 35 */ /* old ftime syscall holder */
-+ .quad sys_sync
-+ .quad sys32_kill
-+ .quad sys_rename
-+ .quad sys_mkdir
-+ .quad sys_rmdir /* 40 */
-+ .quad sys_dup
-+ .quad sys32_pipe
-+ .quad compat_sys_times
-+ .quad quiet_ni_syscall /* old prof syscall holder */
-+ .quad sys_brk /* 45 */
-+ .quad sys_setgid16
-+ .quad sys_getgid16
-+ .quad sys_signal
-+ .quad sys_geteuid16
-+ .quad sys_getegid16 /* 50 */
-+ .quad sys_acct
-+ .quad sys_umount /* new_umount */
-+ .quad quiet_ni_syscall /* old lock syscall holder */
-+ .quad compat_sys_ioctl
-+ .quad compat_sys_fcntl64 /* 55 */
-+ .quad quiet_ni_syscall /* old mpx syscall holder */
-+ .quad sys_setpgid
-+ .quad quiet_ni_syscall /* old ulimit syscall holder */
-+ .quad sys32_olduname
-+ .quad sys_umask /* 60 */
-+ .quad sys_chroot
-+ .quad sys32_ustat
-+ .quad sys_dup2
-+ .quad sys_getppid
-+ .quad sys_getpgrp /* 65 */
-+ .quad sys_setsid
-+ .quad sys32_sigaction
-+ .quad sys_sgetmask
-+ .quad sys_ssetmask
-+ .quad sys_setreuid16 /* 70 */
-+ .quad sys_setregid16
-+ .quad stub32_sigsuspend
-+ .quad compat_sys_sigpending
-+ .quad sys_sethostname
-+ .quad compat_sys_setrlimit /* 75 */
-+ .quad compat_sys_old_getrlimit /* old_getrlimit */
-+ .quad compat_sys_getrusage
-+ .quad sys32_gettimeofday
-+ .quad sys32_settimeofday
-+ .quad sys_getgroups16 /* 80 */
-+ .quad sys_setgroups16
-+ .quad sys32_old_select
-+ .quad sys_symlink
-+ .quad sys_lstat
-+ .quad sys_readlink /* 85 */
-+#ifdef CONFIG_IA32_AOUT
-+ .quad sys_uselib
-+#else
-+ .quad quiet_ni_syscall
-+#endif
-+ .quad sys_swapon
-+ .quad sys_reboot
-+ .quad compat_sys_old_readdir
-+ .quad sys32_mmap /* 90 */
-+ .quad sys_munmap
-+ .quad sys_truncate
-+ .quad sys_ftruncate
-+ .quad sys_fchmod
-+ .quad sys_fchown16 /* 95 */
-+ .quad sys_getpriority
-+ .quad sys_setpriority
-+ .quad quiet_ni_syscall /* old profil syscall holder */
-+ .quad compat_sys_statfs
-+ .quad compat_sys_fstatfs /* 100 */
-+ .quad sys_ioperm
-+ .quad compat_sys_socketcall
-+ .quad sys_syslog
-+ .quad compat_sys_setitimer
-+ .quad compat_sys_getitimer /* 105 */
-+ .quad compat_sys_newstat
-+ .quad compat_sys_newlstat
-+ .quad compat_sys_newfstat
-+ .quad sys32_uname
-+ .quad stub32_iopl /* 110 */
-+ .quad sys_vhangup
-+ .quad quiet_ni_syscall /* old "idle" system call */
-+ .quad sys32_vm86_warning /* vm86old */
-+ .quad compat_sys_wait4
-+ .quad sys_swapoff /* 115 */
-+ .quad sys32_sysinfo
-+ .quad sys32_ipc
-+ .quad sys_fsync
-+ .quad stub32_sigreturn
-+ .quad stub32_clone /* 120 */
-+ .quad sys_setdomainname
-+ .quad sys_uname
-+ .quad sys_modify_ldt
-+ .quad compat_sys_adjtimex
-+ .quad sys32_mprotect /* 125 */
-+ .quad compat_sys_sigprocmask
-+ .quad quiet_ni_syscall /* create_module */
-+ .quad sys_init_module
-+ .quad sys_delete_module
-+ .quad quiet_ni_syscall /* 130 get_kernel_syms */
-+ .quad sys_quotactl
-+ .quad sys_getpgid
-+ .quad sys_fchdir
-+ .quad quiet_ni_syscall /* bdflush */
-+ .quad sys_sysfs /* 135 */
-+ .quad sys_personality
-+ .quad quiet_ni_syscall /* for afs_syscall */
-+ .quad sys_setfsuid16
-+ .quad sys_setfsgid16
-+ .quad sys_llseek /* 140 */
-+ .quad compat_sys_getdents
-+ .quad compat_sys_select
-+ .quad sys_flock
-+ .quad sys_msync
-+ .quad compat_sys_readv /* 145 */
-+ .quad compat_sys_writev
-+ .quad sys_getsid
-+ .quad sys_fdatasync
-+ .quad sys32_sysctl /* sysctl */
-+ .quad sys_mlock /* 150 */
-+ .quad sys_munlock
-+ .quad sys_mlockall
-+ .quad sys_munlockall
-+ .quad sys_sched_setparam
-+ .quad sys_sched_getparam /* 155 */
-+ .quad sys_sched_setscheduler
-+ .quad sys_sched_getscheduler
-+ .quad sys_sched_yield
-+ .quad sys_sched_get_priority_max
-+ .quad sys_sched_get_priority_min /* 160 */
-+ .quad sys_sched_rr_get_interval
-+ .quad compat_sys_nanosleep
-+ .quad sys_mremap
-+ .quad sys_setresuid16
-+ .quad sys_getresuid16 /* 165 */
-+ .quad sys32_vm86_warning /* vm86 */
-+ .quad quiet_ni_syscall /* query_module */
-+ .quad sys_poll
-+ .quad compat_sys_nfsservctl
-+ .quad sys_setresgid16 /* 170 */
-+ .quad sys_getresgid16
-+ .quad sys_prctl
-+ .quad stub32_rt_sigreturn
-+ .quad sys32_rt_sigaction
-+ .quad sys32_rt_sigprocmask /* 175 */
-+ .quad sys32_rt_sigpending
-+ .quad compat_sys_rt_sigtimedwait
-+ .quad sys32_rt_sigqueueinfo
-+ .quad stub32_rt_sigsuspend
-+ .quad sys32_pread /* 180 */
-+ .quad sys32_pwrite
-+ .quad sys_chown16
-+ .quad sys_getcwd
-+ .quad sys_capget
-+ .quad sys_capset
-+ .quad stub32_sigaltstack
-+ .quad sys32_sendfile
-+ .quad quiet_ni_syscall /* streams1 */
-+ .quad quiet_ni_syscall /* streams2 */
-+ .quad stub32_vfork /* 190 */
-+ .quad compat_sys_getrlimit
-+ .quad sys32_mmap2
-+ .quad sys32_truncate64
-+ .quad sys32_ftruncate64
-+ .quad sys32_stat64 /* 195 */
-+ .quad sys32_lstat64
-+ .quad sys32_fstat64
-+ .quad sys_lchown
-+ .quad sys_getuid
-+ .quad sys_getgid /* 200 */
-+ .quad sys_geteuid
-+ .quad sys_getegid
-+ .quad sys_setreuid
-+ .quad sys_setregid
-+ .quad sys_getgroups /* 205 */
-+ .quad sys_setgroups
-+ .quad sys_fchown
-+ .quad sys_setresuid
-+ .quad sys_getresuid
-+ .quad sys_setresgid /* 210 */
-+ .quad sys_getresgid
-+ .quad sys_chown
-+ .quad sys_setuid
-+ .quad sys_setgid
-+ .quad sys_setfsuid /* 215 */
-+ .quad sys_setfsgid
-+ .quad sys_pivot_root
-+ .quad sys_mincore
-+ .quad sys_madvise
-+ .quad compat_sys_getdents64 /* 220 getdents64 */
-+ .quad compat_sys_fcntl64
-+ .quad quiet_ni_syscall /* tux */
-+ .quad quiet_ni_syscall /* security */
-+ .quad sys_gettid
-+ .quad sys_readahead /* 225 */
-+ .quad sys_setxattr
-+ .quad sys_lsetxattr
-+ .quad sys_fsetxattr
-+ .quad sys_getxattr
-+ .quad sys_lgetxattr /* 230 */
-+ .quad sys_fgetxattr
-+ .quad sys_listxattr
-+ .quad sys_llistxattr
-+ .quad sys_flistxattr
-+ .quad sys_removexattr /* 235 */
-+ .quad sys_lremovexattr
-+ .quad sys_fremovexattr
-+ .quad sys_tkill
-+ .quad sys_sendfile64
-+ .quad compat_sys_futex /* 240 */
-+ .quad compat_sys_sched_setaffinity
-+ .quad compat_sys_sched_getaffinity
-+ .quad sys32_set_thread_area
-+ .quad sys32_get_thread_area
-+ .quad compat_sys_io_setup /* 245 */
-+ .quad sys_io_destroy
-+ .quad compat_sys_io_getevents
-+ .quad compat_sys_io_submit
-+ .quad sys_io_cancel
-+ .quad sys_fadvise64 /* 250 */
-+ .quad quiet_ni_syscall /* free_huge_pages */
-+ .quad sys_exit_group
-+ .quad sys32_lookup_dcookie
-+ .quad sys_epoll_create
-+ .quad sys_epoll_ctl /* 255 */
-+ .quad sys_epoll_wait
-+ .quad sys_remap_file_pages
-+ .quad sys_set_tid_address
-+ .quad compat_sys_timer_create
-+ .quad compat_sys_timer_settime /* 260 */
-+ .quad compat_sys_timer_gettime
-+ .quad sys_timer_getoverrun
-+ .quad sys_timer_delete
-+ .quad compat_sys_clock_settime
-+ .quad compat_sys_clock_gettime /* 265 */
-+ .quad compat_sys_clock_getres
-+ .quad compat_sys_clock_nanosleep
-+ .quad compat_sys_statfs64
-+ .quad compat_sys_fstatfs64
-+ .quad sys_tgkill /* 270 */
-+ .quad compat_sys_utimes
-+ .quad sys32_fadvise64_64
-+ .quad quiet_ni_syscall /* sys_vserver */
-+ .quad sys_mbind
-+ .quad compat_sys_get_mempolicy /* 275 */
-+ .quad sys_set_mempolicy
-+ .quad compat_sys_mq_open
-+ .quad sys_mq_unlink
-+ .quad compat_sys_mq_timedsend
-+ .quad compat_sys_mq_timedreceive /* 280 */
-+ .quad compat_sys_mq_notify
-+ .quad compat_sys_mq_getsetattr
-+ .quad compat_sys_kexec_load /* reserved for kexec */
-+ .quad compat_sys_waitid
-+ .quad quiet_ni_syscall /* 285: sys_altroot */
-+ .quad sys_add_key
-+ .quad sys_request_key
-+ .quad sys_keyctl
-+ .quad sys_ioprio_set
-+ .quad sys_ioprio_get /* 290 */
-+ .quad sys_inotify_init
-+ .quad sys_inotify_add_watch
-+ .quad sys_inotify_rm_watch
-+ .quad sys_migrate_pages
-+ .quad compat_sys_openat /* 295 */
-+ .quad sys_mkdirat
-+ .quad sys_mknodat
-+ .quad sys_fchownat
-+ .quad compat_sys_futimesat
-+ .quad sys32_fstatat /* 300 */
-+ .quad sys_unlinkat
-+ .quad sys_renameat
-+ .quad sys_linkat
-+ .quad sys_symlinkat
-+ .quad sys_readlinkat /* 305 */
-+ .quad sys_fchmodat
-+ .quad sys_faccessat
-+ .quad quiet_ni_syscall /* pselect6 for now */
-+ .quad quiet_ni_syscall /* ppoll for now */
-+ .quad sys_unshare /* 310 */
-+ .quad compat_sys_set_robust_list
-+ .quad compat_sys_get_robust_list
-+ .quad sys_splice
-+ .quad sys_sync_file_range
-+ .quad sys_tee
-+ .quad compat_sys_vmsplice
-+ .quad compat_sys_move_pages
-+ia32_syscall_end:
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/ia32/Makefile tmp-linux-2.6-xen.patch/arch/x86_64/ia32/Makefile
---- pristine-linux-2.6.18.2/arch/x86_64/ia32/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/ia32/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -27,9 +27,25 @@ quiet_cmd_syscall = SYSCALL $@
- -Wl,-soname=linux-gate.so.1 -o $@ \
- -Wl,-T,$(filter-out FORCE,$^)
++#
++# IPv6: Netfilter Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP6_NF_QUEUE is not set
++CONFIG_IP6_NF_IPTABLES=y
++CONFIG_IP6_NF_MATCH_RT=y
++CONFIG_IP6_NF_MATCH_OPTS=y
++CONFIG_IP6_NF_MATCH_FRAG=y
++CONFIG_IP6_NF_MATCH_HL=y
++CONFIG_IP6_NF_MATCH_OWNER=y
++CONFIG_IP6_NF_MATCH_IPV6HEADER=y
++CONFIG_IP6_NF_MATCH_AH=y
++CONFIG_IP6_NF_MATCH_EUI64=y
++CONFIG_IP6_NF_FILTER=y
++CONFIG_IP6_NF_TARGET_LOG=y
++CONFIG_IP6_NF_TARGET_REJECT=y
++CONFIG_IP6_NF_MANGLE=y
++CONFIG_IP6_NF_TARGET_HL=y
++CONFIG_IP6_NF_RAW=y
++
++#
++# Bridge: Netfilter Configuration
++#
++CONFIG_BRIDGE_NF_EBTABLES=y
++CONFIG_BRIDGE_EBT_BROUTE=y
++CONFIG_BRIDGE_EBT_T_FILTER=y
++CONFIG_BRIDGE_EBT_T_NAT=y
++CONFIG_BRIDGE_EBT_802_3=y
++CONFIG_BRIDGE_EBT_AMONG=y
++CONFIG_BRIDGE_EBT_ARP=y
++CONFIG_BRIDGE_EBT_IP=y
++CONFIG_BRIDGE_EBT_LIMIT=y
++CONFIG_BRIDGE_EBT_MARK=y
++CONFIG_BRIDGE_EBT_PKTTYPE=y
++CONFIG_BRIDGE_EBT_STP=y
++CONFIG_BRIDGE_EBT_VLAN=y
++CONFIG_BRIDGE_EBT_ARPREPLY=y
++CONFIG_BRIDGE_EBT_DNAT=y
++CONFIG_BRIDGE_EBT_MARK_T=y
++CONFIG_BRIDGE_EBT_REDIRECT=y
++CONFIG_BRIDGE_EBT_SNAT=y
++CONFIG_BRIDGE_EBT_LOG=y
++CONFIG_BRIDGE_EBT_ULOG=y
++
++#
++# DCCP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_DCCP is not set
++
++#
++# SCTP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_SCTP is not set
++
++#
++# TIPC Configuration (EXPERIMENTAL)
++#
++# CONFIG_TIPC is not set
++# CONFIG_ATM is not set
++CONFIG_BRIDGE=y
++CONFIG_VLAN_8021Q=y
++# CONFIG_DECNET is not set
++CONFIG_LLC=y
++# CONFIG_LLC2 is not set
++# CONFIG_IPX is not set
++# CONFIG_ATALK is not set
++# CONFIG_X25 is not set
++# CONFIG_LAPB is not set
++# CONFIG_NET_DIVERT is not set
++# CONFIG_ECONET is not set
++# CONFIG_WAN_ROUTER is not set
++
++#
++# QoS and/or fair queueing
++#
++# CONFIG_NET_SCHED is not set
++CONFIG_NET_CLS_ROUTE=y
++
++#
++# Network testing
++#
++# CONFIG_NET_PKTGEN is not set
++# CONFIG_HAMRADIO is not set
++# CONFIG_IRDA is not set
++# CONFIG_BT is not set
++# CONFIG_IEEE80211 is not set
++
++#
++# Device Drivers
++#
++
++#
++# Generic Driver Options
++#
++CONFIG_STANDALONE=y
++CONFIG_PREVENT_FIRMWARE_BUILD=y
++CONFIG_FW_LOADER=y
++# CONFIG_DEBUG_DRIVER is not set
++# CONFIG_SYS_HYPERVISOR is not set
++
++#
++# Connector - unified userspace <-> kernelspace linker
++#
++# CONFIG_CONNECTOR is not set
++
++#
++# Memory Technology Devices (MTD)
++#
++# CONFIG_MTD is not set
++
++#
++# Parallel port support
++#
++# CONFIG_PARPORT is not set
++
++#
++# Plug and Play support
++#
++
++#
++# Block devices
++#
++# CONFIG_BLK_DEV_FD is not set
++# CONFIG_BLK_CPQ_DA is not set
++# CONFIG_BLK_CPQ_CISS_DA is not set
++# CONFIG_BLK_DEV_DAC960 is not set
++# CONFIG_BLK_DEV_UMEM is not set
++# CONFIG_BLK_DEV_COW_COMMON is not set
++CONFIG_BLK_DEV_LOOP=y
++# CONFIG_BLK_DEV_CRYPTOLOOP is not set
++# CONFIG_BLK_DEV_NBD is not set
++# CONFIG_BLK_DEV_SX8 is not set
++CONFIG_BLK_DEV_RAM=y
++CONFIG_BLK_DEV_RAM_COUNT=16
++CONFIG_BLK_DEV_RAM_SIZE=10240
++CONFIG_BLK_DEV_INITRD=y
++# CONFIG_CDROM_PKTCDVD is not set
++# CONFIG_ATA_OVER_ETH is not set
++
++#
++# ATA/ATAPI/MFM/RLL support
++#
++CONFIG_IDE=y
++CONFIG_BLK_DEV_IDE=y
++
++#
++# Please see Documentation/ide.txt for help/info on IDE drives
++#
++# CONFIG_BLK_DEV_IDE_SATA is not set
++CONFIG_BLK_DEV_IDEDISK=y
++# CONFIG_IDEDISK_MULTI_MODE is not set
++CONFIG_BLK_DEV_IDECD=y
++# CONFIG_BLK_DEV_IDETAPE is not set
++# CONFIG_BLK_DEV_IDEFLOPPY is not set
++# CONFIG_BLK_DEV_IDESCSI is not set
++CONFIG_IDE_TASK_IOCTL=y
++
++#
++# IDE chipset support/bugfixes
++#
++CONFIG_IDE_GENERIC=y
++CONFIG_BLK_DEV_IDEPCI=y
++CONFIG_IDEPCI_SHARE_IRQ=y
++# CONFIG_BLK_DEV_OFFBOARD is not set
++CONFIG_BLK_DEV_GENERIC=y
++# CONFIG_BLK_DEV_OPTI621 is not set
++# CONFIG_BLK_DEV_SL82C105 is not set
++CONFIG_BLK_DEV_IDEDMA_PCI=y
++# CONFIG_BLK_DEV_IDEDMA_FORCED is not set
++CONFIG_IDEDMA_PCI_AUTO=y
++# CONFIG_IDEDMA_ONLYDISK is not set
++# CONFIG_BLK_DEV_AEC62XX is not set
++# CONFIG_BLK_DEV_ALI15X3 is not set
++CONFIG_BLK_DEV_AMD74XX=y
++# CONFIG_BLK_DEV_CMD64X is not set
++# CONFIG_BLK_DEV_TRIFLEX is not set
++# CONFIG_BLK_DEV_CY82C693 is not set
++# CONFIG_BLK_DEV_CS5520 is not set
++# CONFIG_BLK_DEV_CS5530 is not set
++# CONFIG_BLK_DEV_HPT34X is not set
++# CONFIG_BLK_DEV_HPT366 is not set
++# CONFIG_BLK_DEV_SC1200 is not set
++# CONFIG_BLK_DEV_PIIX is not set
++# CONFIG_BLK_DEV_IT821X is not set
++# CONFIG_BLK_DEV_NS87415 is not set
++# CONFIG_BLK_DEV_PDC202XX_OLD is not set
++# CONFIG_BLK_DEV_PDC202XX_NEW is not set
++# CONFIG_BLK_DEV_SVWKS is not set
++# CONFIG_BLK_DEV_SIIMAGE is not set
++# CONFIG_BLK_DEV_SLC90E66 is not set
++# CONFIG_BLK_DEV_TRM290 is not set
++# CONFIG_BLK_DEV_VIA82CXXX is not set
++# CONFIG_IDE_ARM is not set
++CONFIG_BLK_DEV_IDEDMA=y
++# CONFIG_IDEDMA_IVB is not set
++CONFIG_IDEDMA_AUTO=y
++# CONFIG_BLK_DEV_HD is not set
++
++#
++# SCSI device support
++#
++# CONFIG_RAID_ATTRS is not set
++CONFIG_SCSI=y
++# CONFIG_SCSI_PROC_FS is not set
++
++#
++# SCSI support type (disk, tape, CD-ROM)
++#
++CONFIG_BLK_DEV_SD=y
++# CONFIG_CHR_DEV_ST is not set
++# CONFIG_CHR_DEV_OSST is not set
++CONFIG_BLK_DEV_SR=y
++# CONFIG_BLK_DEV_SR_VENDOR is not set
++CONFIG_CHR_DEV_SG=y
++# CONFIG_CHR_DEV_SCH is not set
++
++#
++# Some SCSI devices (e.g. CD jukebox) support multiple LUNs
++#
++# CONFIG_SCSI_MULTI_LUN is not set
++# CONFIG_SCSI_CONSTANTS is not set
++# CONFIG_SCSI_LOGGING is not set
++
++#
++# SCSI Transport Attributes
++#
++CONFIG_SCSI_SPI_ATTRS=y
++CONFIG_SCSI_FC_ATTRS=y
++CONFIG_SCSI_ISCSI_ATTRS=y
++CONFIG_SCSI_SAS_ATTRS=y
++
++#
++# SCSI low-level drivers
++#
++# CONFIG_ISCSI_TCP is not set
++# CONFIG_BLK_DEV_3W_XXXX_RAID is not set
++# CONFIG_SCSI_3W_9XXX is not set
++# CONFIG_SCSI_ACARD is not set
++# CONFIG_SCSI_AACRAID is not set
++# CONFIG_SCSI_AIC7XXX is not set
++# CONFIG_SCSI_AIC7XXX_OLD is not set
++# CONFIG_SCSI_AIC79XX is not set
++# CONFIG_MEGARAID_NEWGEN is not set
++# CONFIG_MEGARAID_LEGACY is not set
++# CONFIG_MEGARAID_SAS is not set
++# CONFIG_SCSI_SATA is not set
++# CONFIG_SCSI_HPTIOP is not set
++# CONFIG_SCSI_BUSLOGIC is not set
++# CONFIG_SCSI_DMX3191D is not set
++# CONFIG_SCSI_EATA is not set
++# CONFIG_SCSI_FUTURE_DOMAIN is not set
++# CONFIG_SCSI_GDTH is not set
++# CONFIG_SCSI_IPS is not set
++# CONFIG_SCSI_IBMVSCSI is not set
++# CONFIG_SCSI_INITIO is not set
++# CONFIG_SCSI_INIA100 is not set
++# CONFIG_SCSI_SYM53C8XX_2 is not set
++CONFIG_SCSI_IPR=y
++# CONFIG_SCSI_IPR_TRACE is not set
++# CONFIG_SCSI_IPR_DUMP is not set
++# CONFIG_SCSI_QLOGIC_1280 is not set
++# CONFIG_SCSI_QLA_FC is not set
++# CONFIG_SCSI_LPFC is not set
++# CONFIG_SCSI_DC395x is not set
++# CONFIG_SCSI_DC390T is not set
++# CONFIG_SCSI_DEBUG is not set
++
++#
++# Multi-device support (RAID and LVM)
++#
++# CONFIG_MD is not set
++
++#
++# Fusion MPT device support
++#
++# CONFIG_FUSION is not set
++# CONFIG_FUSION_SPI is not set
++# CONFIG_FUSION_FC is not set
++# CONFIG_FUSION_SAS is not set
++
++#
++# IEEE 1394 (FireWire) support
++#
++# CONFIG_IEEE1394 is not set
++
++#
++# I2O device support
++#
++# CONFIG_I2O is not set
++
++#
++# Macintosh device drivers
++#
++# CONFIG_WINDFARM is not set
++
++#
++# Network device support
++#
++CONFIG_NETDEVICES=y
++# CONFIG_DUMMY is not set
++# CONFIG_BONDING is not set
++# CONFIG_EQUALIZER is not set
++# CONFIG_TUN is not set
++
++#
++# ARCnet devices
++#
++# CONFIG_ARCNET is not set
++
++#
++# PHY device support
++#
++# CONFIG_PHYLIB is not set
++
++#
++# Ethernet (10 or 100Mbit)
++#
++CONFIG_NET_ETHERNET=y
++CONFIG_MII=y
++# CONFIG_HAPPYMEAL is not set
++# CONFIG_SUNGEM is not set
++# CONFIG_CASSINI is not set
++# CONFIG_NET_VENDOR_3COM is not set
++
++#
++# Tulip family network device support
++#
++# CONFIG_NET_TULIP is not set
++# CONFIG_HP100 is not set
++CONFIG_IBMVETH=y
++CONFIG_NET_PCI=y
++# CONFIG_PCNET32 is not set
++CONFIG_AMD8111_ETH=y
++# CONFIG_AMD8111E_NAPI is not set
++# CONFIG_ADAPTEC_STARFIRE is not set
++# CONFIG_B44 is not set
++# CONFIG_FORCEDETH is not set
++# CONFIG_DGRS is not set
++# CONFIG_EEPRO100 is not set
++# CONFIG_E100 is not set
++# CONFIG_FEALNX is not set
++# CONFIG_NATSEMI is not set
++# CONFIG_NE2K_PCI is not set
++# CONFIG_8139CP is not set
++# CONFIG_8139TOO is not set
++# CONFIG_SIS900 is not set
++# CONFIG_EPIC100 is not set
++# CONFIG_SUNDANCE is not set
++# CONFIG_VIA_RHINE is not set
++
++#
++# Ethernet (1000 Mbit)
++#
++# CONFIG_ACENIC is not set
++# CONFIG_DL2K is not set
++CONFIG_E1000=y
++# CONFIG_E1000_NAPI is not set
++# CONFIG_E1000_DISABLE_PACKET_SPLIT is not set
++# CONFIG_NS83820 is not set
++# CONFIG_HAMACHI is not set
++# CONFIG_YELLOWFIN is not set
++# CONFIG_R8169 is not set
++# CONFIG_SIS190 is not set
++# CONFIG_SKGE is not set
++CONFIG_SKY2=y
++CONFIG_SK98LIN=y
++# CONFIG_VIA_VELOCITY is not set
++CONFIG_TIGON3=y
++CONFIG_BNX2=y
++# CONFIG_MV643XX_ETH is not set
++
++#
++# Ethernet (10000 Mbit)
++#
++# CONFIG_CHELSIO_T1 is not set
++# CONFIG_IXGB is not set
++# CONFIG_S2IO is not set
++# CONFIG_MYRI10GE is not set
++
++#
++# Token Ring devices
++#
++# CONFIG_TR is not set
++
++#
++# Wireless LAN (non-hamradio)
++#
++# CONFIG_NET_RADIO is not set
++
++#
++# Wan interfaces
++#
++# CONFIG_WAN is not set
++# CONFIG_FDDI is not set
++# CONFIG_HIPPI is not set
++# CONFIG_PPP is not set
++# CONFIG_SLIP is not set
++# CONFIG_NET_FC is not set
++# CONFIG_SHAPER is not set
++# CONFIG_NETCONSOLE is not set
++# CONFIG_NETPOLL is not set
++# CONFIG_NET_POLL_CONTROLLER is not set
++
++#
++# ISDN subsystem
++#
++# CONFIG_ISDN is not set
++
++#
++# Telephony Support
++#
++# CONFIG_PHONE is not set
++
++#
++# Input device support
++#
++CONFIG_INPUT=y
++
++#
++# Userland interfaces
++#
++CONFIG_INPUT_MOUSEDEV=y
++# CONFIG_INPUT_MOUSEDEV_PSAUX is not set
++CONFIG_INPUT_MOUSEDEV_SCREEN_X=1600
++CONFIG_INPUT_MOUSEDEV_SCREEN_Y=1200
++# CONFIG_INPUT_JOYDEV is not set
++# CONFIG_INPUT_TSDEV is not set
++# CONFIG_INPUT_EVDEV is not set
++# CONFIG_INPUT_EVBUG is not set
++
++#
++# Input Device Drivers
++#
++# CONFIG_INPUT_KEYBOARD is not set
++# CONFIG_INPUT_MOUSE is not set
++# CONFIG_INPUT_JOYSTICK is not set
++# CONFIG_INPUT_TOUCHSCREEN is not set
++# CONFIG_INPUT_MISC is not set
++
++#
++# Hardware I/O ports
++#
++# CONFIG_SERIO is not set
++# CONFIG_GAMEPORT is not set
++
++#
++# Character devices
++#
++CONFIG_VT=y
++CONFIG_VT_CONSOLE=y
++CONFIG_HW_CONSOLE=y
++# CONFIG_VT_HW_CONSOLE_BINDING is not set
++# CONFIG_SERIAL_NONSTANDARD is not set
++
++#
++# Serial drivers
++#
++
++#
++# Non-8250 serial port support
++#
++# CONFIG_SERIAL_ICOM is not set
++# CONFIG_SERIAL_JSM is not set
++CONFIG_UNIX98_PTYS=y
++CONFIG_LEGACY_PTYS=y
++CONFIG_LEGACY_PTY_COUNT=256
++CONFIG_HVC_DRIVER=y
++CONFIG_HVC_CONSOLE=y
++# CONFIG_HVC_RTAS is not set
++# CONFIG_HVCS is not set
++
++#
++# IPMI
++#
++# CONFIG_IPMI_HANDLER is not set
++
++#
++# Watchdog Cards
++#
++# CONFIG_WATCHDOG is not set
++CONFIG_HW_RANDOM=y
++CONFIG_GEN_RTC=y
++# CONFIG_GEN_RTC_X is not set
++# CONFIG_DTLK is not set
++# CONFIG_R3964 is not set
++# CONFIG_APPLICOM is not set
++
++#
++# Ftape, the floppy tape device driver
++#
++# CONFIG_AGP is not set
++# CONFIG_DRM is not set
++# CONFIG_RAW_DRIVER is not set
++# CONFIG_HANGCHECK_TIMER is not set
++
++#
++# TPM devices
++#
++# CONFIG_TCG_TPM is not set
++# CONFIG_TELCLOCK is not set
++
++#
++# I2C support
++#
++CONFIG_I2C=y
++CONFIG_I2C_CHARDEV=y
++
++#
++# I2C Algorithms
++#
++CONFIG_I2C_ALGOBIT=y
++# CONFIG_I2C_ALGOPCF is not set
++# CONFIG_I2C_ALGOPCA is not set
++
++#
++# I2C Hardware Bus support
++#
++# CONFIG_I2C_ALI1535 is not set
++# CONFIG_I2C_ALI1563 is not set
++# CONFIG_I2C_ALI15X3 is not set
++# CONFIG_I2C_AMD756 is not set
++CONFIG_I2C_AMD8111=y
++# CONFIG_I2C_I801 is not set
++# CONFIG_I2C_I810 is not set
++# CONFIG_I2C_PIIX4 is not set
++# CONFIG_I2C_NFORCE2 is not set
++# CONFIG_I2C_OCORES is not set
++# CONFIG_I2C_PARPORT_LIGHT is not set
++# CONFIG_I2C_PROSAVAGE is not set
++# CONFIG_I2C_SAVAGE4 is not set
++# CONFIG_I2C_SIS5595 is not set
++# CONFIG_I2C_SIS630 is not set
++# CONFIG_I2C_SIS96X is not set
++# CONFIG_I2C_STUB is not set
++# CONFIG_I2C_VIA is not set
++# CONFIG_I2C_VIAPRO is not set
++# CONFIG_I2C_VOODOO3 is not set
++# CONFIG_I2C_PCA_ISA is not set
++
++#
++# Miscellaneous I2C Chip support
++#
++# CONFIG_SENSORS_DS1337 is not set
++# CONFIG_SENSORS_DS1374 is not set
++# CONFIG_SENSORS_EEPROM is not set
++# CONFIG_SENSORS_PCF8574 is not set
++# CONFIG_SENSORS_PCA9539 is not set
++# CONFIG_SENSORS_PCF8591 is not set
++# CONFIG_SENSORS_MAX6875 is not set
++# CONFIG_I2C_DEBUG_CORE is not set
++# CONFIG_I2C_DEBUG_ALGO is not set
++# CONFIG_I2C_DEBUG_BUS is not set
++# CONFIG_I2C_DEBUG_CHIP is not set
++
++#
++# SPI support
++#
++# CONFIG_SPI is not set
++# CONFIG_SPI_MASTER is not set
++
++#
++# Dallas's 1-wire bus
++#
++
++#
++# Hardware Monitoring support
++#
++# CONFIG_HWMON is not set
++# CONFIG_HWMON_VID is not set
++
++#
++# Misc devices
++#
++
++#
++# Multimedia devices
++#
++# CONFIG_VIDEO_DEV is not set
++CONFIG_VIDEO_V4L2=y
++
++#
++# Digital Video Broadcasting Devices
++#
++# CONFIG_DVB is not set
++
++#
++# Graphics support
++#
++CONFIG_FIRMWARE_EDID=y
++# CONFIG_FB is not set
++
++#
++# Console display driver support
++#
++# CONFIG_VGA_CONSOLE is not set
++CONFIG_DUMMY_CONSOLE=y
++
++#
++# Sound
++#
++# CONFIG_SOUND is not set
++
++#
++# USB support
++#
++CONFIG_USB_ARCH_HAS_HCD=y
++CONFIG_USB_ARCH_HAS_OHCI=y
++CONFIG_USB_ARCH_HAS_EHCI=y
++# CONFIG_USB is not set
++
++#
++# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support'
++#
++
++#
++# USB Gadget Support
++#
++# CONFIG_USB_GADGET is not set
++
++#
++# MMC/SD Card support
++#
++# CONFIG_MMC is not set
++
++#
++# LED devices
++#
++# CONFIG_NEW_LEDS is not set
++
++#
++# LED drivers
++#
++
++#
++# LED Triggers
++#
++
++#
++# InfiniBand support
++#
++CONFIG_INFINIBAND=y
++CONFIG_INFINIBAND_USER_MAD=y
++CONFIG_INFINIBAND_USER_ACCESS=y
++CONFIG_INFINIBAND_ADDR_TRANS=y
++CONFIG_INFINIBAND_MTHCA=y
++CONFIG_INFINIBAND_MTHCA_DEBUG=y
++CONFIG_INFINIBAND_IPOIB=y
++CONFIG_INFINIBAND_IPOIB_DEBUG=y
++CONFIG_INFINIBAND_IPOIB_DEBUG_DATA=y
++CONFIG_INFINIBAND_SRP=y
++# CONFIG_INFINIBAND_ISER is not set
++
++#
++# EDAC - error detection and reporting (RAS) (EXPERIMENTAL)
++#
++
++#
++# Real Time Clock
++#
++# CONFIG_RTC_CLASS is not set
++
++#
++# DMA Engine support
++#
++# CONFIG_DMA_ENGINE is not set
++
++#
++# DMA Clients
++#
++
++#
++# DMA Devices
++#
++
++#
++# File systems
++#
++CONFIG_EXT2_FS=y
++# CONFIG_EXT2_FS_XATTR is not set
++CONFIG_EXT2_FS_XIP=y
++CONFIG_FS_XIP=y
++CONFIG_EXT3_FS=y
++# CONFIG_EXT3_FS_XATTR is not set
++CONFIG_JBD=y
++# CONFIG_JBD_DEBUG is not set
++CONFIG_REISERFS_FS=y
++# CONFIG_REISERFS_CHECK is not set
++CONFIG_REISERFS_PROC_INFO=y
++CONFIG_REISERFS_FS_XATTR=y
++CONFIG_REISERFS_FS_POSIX_ACL=y
++CONFIG_REISERFS_FS_SECURITY=y
++# CONFIG_JFS_FS is not set
++CONFIG_FS_POSIX_ACL=y
++# CONFIG_XFS_FS is not set
++# CONFIG_OCFS2_FS is not set
++# CONFIG_MINIX_FS is not set
++# CONFIG_ROMFS_FS is not set
++CONFIG_INOTIFY=y
++CONFIG_INOTIFY_USER=y
++# CONFIG_QUOTA is not set
++CONFIG_DNOTIFY=y
++CONFIG_AUTOFS_FS=y
++# CONFIG_AUTOFS4_FS is not set
++# CONFIG_FUSE_FS is not set
++
++#
++# CD-ROM/DVD Filesystems
++#
++CONFIG_ISO9660_FS=y
++# CONFIG_JOLIET is not set
++# CONFIG_ZISOFS is not set
++# CONFIG_UDF_FS is not set
++
++#
++# DOS/FAT/NT Filesystems
++#
++CONFIG_FAT_FS=y
++CONFIG_MSDOS_FS=y
++CONFIG_VFAT_FS=y
++CONFIG_FAT_DEFAULT_CODEPAGE=437
++CONFIG_FAT_DEFAULT_IOCHARSET="iso8859-1"
++# CONFIG_NTFS_FS is not set
++
++#
++# Pseudo filesystems
++#
++CONFIG_PROC_FS=y
++CONFIG_PROC_KCORE=y
++CONFIG_SYSFS=y
++CONFIG_TMPFS=y
++# CONFIG_HUGETLB_PAGE is not set
++CONFIG_RAMFS=y
++# CONFIG_CONFIGFS_FS is not set
++
++#
++# Miscellaneous filesystems
++#
++# CONFIG_ADFS_FS is not set
++# CONFIG_AFFS_FS is not set
++# CONFIG_HFS_FS is not set
++# CONFIG_HFSPLUS_FS is not set
++# CONFIG_BEFS_FS is not set
++# CONFIG_BFS_FS is not set
++# CONFIG_EFS_FS is not set
++CONFIG_CRAMFS=y
++# CONFIG_VXFS_FS is not set
++# CONFIG_HPFS_FS is not set
++# CONFIG_QNX4FS_FS is not set
++# CONFIG_SYSV_FS is not set
++# CONFIG_UFS_FS is not set
++
++#
++# Network File Systems
++#
++CONFIG_NFS_FS=y
++CONFIG_NFS_V3=y
++CONFIG_NFS_V3_ACL=y
++CONFIG_NFS_V4=y
++# CONFIG_NFS_DIRECTIO is not set
++# CONFIG_NFSD is not set
++CONFIG_ROOT_NFS=y
++CONFIG_LOCKD=y
++CONFIG_LOCKD_V4=y
++CONFIG_NFS_ACL_SUPPORT=y
++CONFIG_NFS_COMMON=y
++CONFIG_SUNRPC=y
++CONFIG_SUNRPC_GSS=y
++CONFIG_RPCSEC_GSS_KRB5=y
++# CONFIG_RPCSEC_GSS_SPKM3 is not set
++# CONFIG_SMB_FS is not set
++# CONFIG_CIFS is not set
++# CONFIG_CIFS_DEBUG2 is not set
++# CONFIG_NCP_FS is not set
++# CONFIG_CODA_FS is not set
++# CONFIG_AFS_FS is not set
++# CONFIG_9P_FS is not set
++
++#
++# Partition Types
++#
++CONFIG_PARTITION_ADVANCED=y
++# CONFIG_ACORN_PARTITION is not set
++# CONFIG_OSF_PARTITION is not set
++# CONFIG_AMIGA_PARTITION is not set
++# CONFIG_ATARI_PARTITION is not set
++CONFIG_MAC_PARTITION=y
++CONFIG_MSDOS_PARTITION=y
++# CONFIG_BSD_DISKLABEL is not set
++# CONFIG_MINIX_SUBPARTITION is not set
++# CONFIG_SOLARIS_X86_PARTITION is not set
++# CONFIG_UNIXWARE_DISKLABEL is not set
++# CONFIG_LDM_PARTITION is not set
++# CONFIG_SGI_PARTITION is not set
++# CONFIG_ULTRIX_PARTITION is not set
++# CONFIG_SUN_PARTITION is not set
++# CONFIG_KARMA_PARTITION is not set
++# CONFIG_EFI_PARTITION is not set
++
++#
++# Native Language Support
++#
++CONFIG_NLS=y
++CONFIG_NLS_DEFAULT="utf-8"
++# CONFIG_NLS_CODEPAGE_437 is not set
++# CONFIG_NLS_CODEPAGE_737 is not set
++# CONFIG_NLS_CODEPAGE_775 is not set
++# CONFIG_NLS_CODEPAGE_850 is not set
++# CONFIG_NLS_CODEPAGE_852 is not set
++# CONFIG_NLS_CODEPAGE_855 is not set
++# CONFIG_NLS_CODEPAGE_857 is not set
++# CONFIG_NLS_CODEPAGE_860 is not set
++# CONFIG_NLS_CODEPAGE_861 is not set
++# CONFIG_NLS_CODEPAGE_862 is not set
++# CONFIG_NLS_CODEPAGE_863 is not set
++# CONFIG_NLS_CODEPAGE_864 is not set
++# CONFIG_NLS_CODEPAGE_865 is not set
++# CONFIG_NLS_CODEPAGE_866 is not set
++# CONFIG_NLS_CODEPAGE_869 is not set
++# CONFIG_NLS_CODEPAGE_936 is not set
++# CONFIG_NLS_CODEPAGE_950 is not set
++# CONFIG_NLS_CODEPAGE_932 is not set
++# CONFIG_NLS_CODEPAGE_949 is not set
++# CONFIG_NLS_CODEPAGE_874 is not set
++# CONFIG_NLS_ISO8859_8 is not set
++# CONFIG_NLS_CODEPAGE_1250 is not set
++# CONFIG_NLS_CODEPAGE_1251 is not set
++# CONFIG_NLS_ASCII is not set
++# CONFIG_NLS_ISO8859_1 is not set
++# CONFIG_NLS_ISO8859_2 is not set
++# CONFIG_NLS_ISO8859_3 is not set
++# CONFIG_NLS_ISO8859_4 is not set
++# CONFIG_NLS_ISO8859_5 is not set
++# CONFIG_NLS_ISO8859_6 is not set
++# CONFIG_NLS_ISO8859_7 is not set
++# CONFIG_NLS_ISO8859_9 is not set
++# CONFIG_NLS_ISO8859_13 is not set
++# CONFIG_NLS_ISO8859_14 is not set
++# CONFIG_NLS_ISO8859_15 is not set
++# CONFIG_NLS_KOI8_R is not set
++# CONFIG_NLS_KOI8_U is not set
++CONFIG_NLS_UTF8=y
++
++#
++# Library routines
++#
++CONFIG_CRC_CCITT=y
++# CONFIG_CRC16 is not set
++CONFIG_CRC32=y
++# CONFIG_LIBCRC32C is not set
++CONFIG_ZLIB_INFLATE=y
++CONFIG_ZLIB_DEFLATE=y
++CONFIG_TEXTSEARCH=y
++CONFIG_TEXTSEARCH_KMP=y
++CONFIG_TEXTSEARCH_BM=y
++CONFIG_TEXTSEARCH_FSM=y
++CONFIG_PLIST=y
++
++#
++# Instrumentation Support
++#
++# CONFIG_PROFILING is not set
++# CONFIG_KPROBES is not set
++
++#
++# Kernel hacking
++#
++# CONFIG_PRINTK_TIME is not set
++CONFIG_MAGIC_SYSRQ=y
++# CONFIG_UNUSED_SYMBOLS is not set
++CONFIG_DEBUG_KERNEL=y
++CONFIG_LOG_BUF_SHIFT=17
++CONFIG_DETECT_SOFTLOCKUP=y
++# CONFIG_SCHEDSTATS is not set
++CONFIG_DEBUG_SLAB=y
++CONFIG_DEBUG_MUTEXES=y
++# CONFIG_DEBUG_RT_MUTEXES is not set
++# CONFIG_RT_MUTEX_TESTER is not set
++# CONFIG_DEBUG_SPINLOCK is not set
++CONFIG_DEBUG_SPINLOCK_SLEEP=y
++# CONFIG_DEBUG_KOBJECT is not set
++CONFIG_DEBUG_INFO=y
++CONFIG_DEBUG_FS=y
++# CONFIG_DEBUG_VM is not set
++CONFIG_FORCED_INLINING=y
++# CONFIG_RCU_TORTURE_TEST is not set
++CONFIG_DEBUG_STACKOVERFLOW=y
++CONFIG_DEBUG_STACK_USAGE=y
++CONFIG_DEBUGGER=y
++CONFIG_XMON=y
++CONFIG_XMON_DEFAULT=y
++# CONFIG_IRQSTACKS is not set
++CONFIG_BOOTX_TEXT=y
++# CONFIG_PPC_EARLY_DEBUG is not set
++
++#
++# Security options
++#
++# CONFIG_KEYS is not set
++# CONFIG_SECURITY is not set
++CONFIG_XEN=y
++CONFIG_XEN_INTERFACE_VERSION=0x00030202
++
++#
++# XEN
++#
++CONFIG_XEN_PRIVILEGED_GUEST=y
++CONFIG_XEN_UNPRIVILEGED_GUEST=y
++CONFIG_XEN_PRIVCMD=y
++CONFIG_XEN_BACKEND=y
++# CONFIG_XEN_PCIDEV_BACKEND is not set
++CONFIG_XEN_BLKDEV_BACKEND=y
++CONFIG_XEN_XENBUS_DEV=y
++CONFIG_XEN_NETDEV_BACKEND=y
++# CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER is not set
++CONFIG_XEN_NETDEV_LOOPBACK=y
++# CONFIG_XEN_TPMDEV_BACKEND is not set
++CONFIG_XEN_BLKDEV_FRONTEND=y
++CONFIG_XEN_NETDEV_FRONTEND=y
++CONFIG_XEN_SCRUB_PAGES=y
++CONFIG_XEN_DISABLE_SERIAL=y
++CONFIG_XEN_SYSFS=y
++# CONFIG_XEN_COMPAT_030002_AND_LATER is not set
++CONFIG_XEN_COMPAT_LATEST_ONLY=y
++# CONFIG_XEN_COMPAT_030002 is not set
++CONFIG_HAVE_ARCH_ALLOC_SKB=y
++CONFIG_HAVE_ARCH_DEV_ALLOC_SKB=y
++CONFIG_HAVE_IRQ_IGNORE_UNHANDLED=y
++CONFIG_NO_IDLE_HZ=y
++CONFIG_XEN_DEVMEM=y
++CONFIG_XEN_SKBUFF=y
++CONFIG_XEN_REBOOT=y
++CONFIG_XEN_XENCOMM=y
++
++#
++# Cryptographic options
++#
++CONFIG_CRYPTO=y
++CONFIG_CRYPTO_HMAC=y
++# CONFIG_CRYPTO_NULL is not set
++# CONFIG_CRYPTO_MD4 is not set
++CONFIG_CRYPTO_MD5=y
++CONFIG_CRYPTO_SHA1=y
++# CONFIG_CRYPTO_SHA256 is not set
++# CONFIG_CRYPTO_SHA512 is not set
++# CONFIG_CRYPTO_WP512 is not set
++# CONFIG_CRYPTO_TGR192 is not set
++CONFIG_CRYPTO_DES=y
++# CONFIG_CRYPTO_BLOWFISH is not set
++# CONFIG_CRYPTO_TWOFISH is not set
++# CONFIG_CRYPTO_SERPENT is not set
++# CONFIG_CRYPTO_AES is not set
++# CONFIG_CRYPTO_CAST5 is not set
++# CONFIG_CRYPTO_CAST6 is not set
++# CONFIG_CRYPTO_TEA is not set
++# CONFIG_CRYPTO_ARC4 is not set
++# CONFIG_CRYPTO_KHAZAD is not set
++# CONFIG_CRYPTO_ANUBIS is not set
++CONFIG_CRYPTO_DEFLATE=y
++# CONFIG_CRYPTO_MICHAEL_MIC is not set
++# CONFIG_CRYPTO_CRC32C is not set
++# CONFIG_CRYPTO_TEST is not set
++
++#
++# Hardware crypto devices
++#
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/kernel/cpu_setup_power4.S
+--- a/arch/powerpc/kernel/cpu_setup_power4.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/kernel/cpu_setup_power4.S Wed Sep 10 10:54:08 2008 +0100
+@@ -73,6 +73,13 @@
+ blr
+
+ _GLOBAL(__setup_cpu_ppc970)
++ /*
++ * Do nothing if not running in HV mode
++ */
++ mfmsr r0
++ rldicl. r0,r0,4,63
++ beqlr
++
+ mfspr r0,SPRN_HID0
+ li r11,5 /* clear DOZE and SLEEP */
+ rldimi r0,r11,52,8 /* set NAP and DPM */
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/kernel/prom_init.c
+--- a/arch/powerpc/kernel/prom_init.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/kernel/prom_init.c Wed Sep 10 10:54:08 2008 +0100
+@@ -188,6 +188,7 @@
+ #define PLATFORM_LPAR 0x0001
+ #define PLATFORM_POWERMAC 0x0400
+ #define PLATFORM_GENERIC 0x0500
++#define PLATFORM_GENERIC_XEN (PLATFORM_GENERIC | PLATFORM_LPAR)
+
+ static int __initdata of_platform;
+
+@@ -1529,6 +1530,14 @@
+ #ifdef CONFIG_PPC64
+ phandle rtas;
+ int x;
++#endif
++#ifdef CONFIG_PPC_XEN
++ phandle xen;
++
++ xen = call_prom("finddevice", 1, 1, ADDR("/xen"));
++ if (PHANDLE_VALID(xen)) {
++ return PLATFORM_GENERIC_XEN;
++ }
+ #endif
-+$(obj)/vsyscall-int80.so \
- $(obj)/vsyscall-sysenter.so $(obj)/vsyscall-syscall.so: \
- $(obj)/vsyscall-%.so: $(src)/vsyscall.lds $(obj)/vsyscall-%.o FORCE
- $(call if_changed,syscall)
+ /* Look for a PowerMac */
+@@ -2263,6 +2272,31 @@
+ if (RELOC(of_platform) == PLATFORM_PSERIES)
+ prom_initialize_tce_table();
+ #endif
++#ifdef CONFIG_PPC_XEN
++ if (RELOC(of_platform) & PLATFORM_LPAR) {
++ phandle xen;
++
++ prom_debug("XXX:checking for Xen OF package\n");
++
++ xen = call_prom("finddevice", 1, 1, ADDR("/xen"));
++ if (PHANDLE_VALID(xen)) {
++ u64 res[2];
++ int l;
++ ulong base;
++
++ l = prom_getprop(xen, "reserved", res, sizeof (res));
++ if (l != sizeof(res)) {
++ prom_panic("Xen reserved prop not exist\n");
++ }
++
++ base = alloc_down(res[1], PAGE_SIZE, 0);
++ if (base != res[0]) {
++ prom_panic("XSI != alloc_down()\n");
++ }
++ reserve_mem(res[0], res[1]);
++ }
++ }
++#endif
--AFLAGS_vsyscall-sysenter.o = -m32 -Wa,-32
--AFLAGS_vsyscall-syscall.o = -m32 -Wa,-32
-+AFLAGS_vsyscall-sysenter.o = -m32 -Wa,-32 -Iarch/i386/kernel
-+AFLAGS_vsyscall-syscall.o = -m32 -Wa,-32 -Iarch/i386/kernel
+ /*
+ * On non-powermacs, try to instantiate RTAS and puts all CPUs
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/kernel/setup-common.c
+--- a/arch/powerpc/kernel/setup-common.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/kernel/setup-common.c Wed Sep 10 10:54:08 2008 +0100
+@@ -387,6 +387,12 @@
+ }
+ }
+
++ if (machine_is(xen)) {
++ /* something more inteligent perhaps? */
++ for (cpu = 0; cpu < NR_CPUS; cpu++)
++ cpu_set(cpu, cpu_possible_map);
++ }
++
+ #ifdef CONFIG_PPC64
+ /*
+ * On pSeries LPAR, we need to know how many cpus
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/kernel/udbg.c
+--- a/arch/powerpc/kernel/udbg.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/kernel/udbg.c Wed Sep 10 10:54:08 2008 +0100
+@@ -45,6 +45,9 @@
+ #elif defined(CONFIG_PPC_EARLY_DEBUG_ISERIES)
+ /* For iSeries - hit Ctrl-x Ctrl-x to see the output */
+ udbg_init_iseries();
++#elif defined(CONFIG_PPC_EARLY_DEBUG_XEN_DOM0) || \
++ defined(CONFIG_PPC_EARLY_DEBUG_XEN_DOMU)
++ udbg_init_xen();
+ #endif
+ }
+
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/kernel/vdso.c
+--- a/arch/powerpc/kernel/vdso.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/kernel/vdso.c Wed Sep 10 10:54:08 2008 +0100
+@@ -282,6 +282,13 @@
+ * pages though
+ */
+ vma->vm_flags = VM_READ|VM_EXEC|VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC;
++ /*
++ * Make sure the vDSO gets into every core dump.
++ * Dumping its contents makes post-mortem fully interpretable later
++ * without matching up the same kernel and hardware config to see
++ * what PC values meant.
++ */
++ vma->vm_flags |= VM_ALWAYSDUMP;
+ vma->vm_flags |= mm->def_flags;
+ vma->vm_page_prot = protection_map[vma->vm_flags & 0x7];
+ vma->vm_ops = &vdso_vmops;
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/mm/slb_low.S
+--- a/arch/powerpc/mm/slb_low.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/mm/slb_low.S Wed Sep 10 10:54:08 2008 +0100
+@@ -51,6 +51,23 @@
+ */
+ bne cr7,1f
+
++#ifdef CONFIG_PPC_XEN
++_GLOBAL(slb_miss_kernel_load_xen_nop)
++ b 3f
++ /* Need to check if it is in the part of our XEN Foreign Map */
++ rldicl r9,r3,30,63 /* get Xen region */
++ cmpldi cr7,r9,1 /* cmp this bit set to 1 */
++ bne cr7,3f
++ /* Xen Linear mapping encoding bits, the "li" instruction below
++ * could be patched below (like the other pages of the linear map)
++ * if we ever wish to map anything other that 4K pages in
++ * this region, right now it is fine as zero.
++ */
++_GLOBAL(slb_miss_kernel_load_xen_linear)
++ li r11,0
++ b slb_finish_load
++3:
++#endif
+ /* Linear mapping encoding bits, the "li" instruction below will
+ * be patched by the kernel at boot
+ */
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/Makefile
+--- a/arch/powerpc/platforms/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/platforms/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -12,6 +12,8 @@
+ obj-$(CONFIG_PPC_86xx) += 86xx/
+ obj-$(CONFIG_PPC_PSERIES) += pseries/
+ obj-$(CONFIG_PPC_ISERIES) += iseries/
++# must occur before xen hosting platforms
++obj-$(CONFIG_PPC_XEN) += xen/
+ obj-$(CONFIG_PPC_MAPLE) += maple/
+ obj-$(CONFIG_PPC_CELL) += cell/
+ obj-$(CONFIG_EMBEDDED6xx) += embedded6xx/
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/pseries/iommu.c
+--- a/arch/powerpc/platforms/pseries/iommu.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/platforms/pseries/iommu.c Wed Sep 10 10:54:08 2008 +0100
+@@ -531,6 +531,17 @@
+ * already allocated.
+ */
+ dn = pci_device_to_OF_node(dev);
++ if (dn == NULL) {
++#ifdef CONFIG_PPC_XEN
++ /* this becomes possible for Xen Dom0 */
++ DBG("%s, dev %p (%s) has no OF devtree entree\n", __func__,
++ dev, pci_name(dev));
++ return;
++#else
++ panic("%s, dev %p (%s) has no OF devtree entree\n", __func__,
++ dev, pci_name(dev));
++#endif
++ }
+
+ for (pdn = dn; pdn && PCI_DN(pdn) && !PCI_DN(pdn)->iommu_table;
+ pdn = pdn->parent) {
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,20 @@
++obj-y += gnttab.o
++obj-y += hcall.o
++obj-y += reboot.o
++obj-y += setup.o
++obj-y += smp.o
++obj-y += time.o
++obj-y += udbg_xen.o
++obj-y += xen_guest.o
++obj-y += xencomm.o
++
++# we need the latest __XEN_INTERFACE_VERSION__ (see xen-compat.h)
++CFLAGS_hcall.o += -D__XEN_TOOLS__
++
++ifndef CONFIG_XEN_BALLOON
++obj-y += balloon.o
++endif
++
++ifndef CONFIG_XEN_UTIL
++obj-y += util.o
++endif
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/balloon.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/balloon.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,82 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
++
++#include <linux/module.h>
++#include <linux/mm.h>
++#include <asm/hypervisor.h>
++#include "setup.h"
++
++/*
++ * FIXME: Port balloon driver, if ever
++ */
++
++struct page **alloc_empty_pages_and_pagevec(int nr_pages)
++{
++ struct page *page, **pagevec;
++ int i;
++
++ pagevec = kmalloc(sizeof(*pagevec) * nr_pages, GFP_KERNEL);
++ if (pagevec == NULL)
++ return NULL;
++
++ for (i = 0; i < nr_pages; i++) {
++ page = alloc_foreign_page();
++ BUG_ON(page == NULL);
++ pagevec[i] = page;
++ /* There is no real page backing us yet so it cannot
++ * be scrubbed */
++ }
++
++ return pagevec;
++}
++
++void free_empty_pages_and_pagevec(struct page **pagevec, int nr_pages)
++{
++ int i;
++
++ if (pagevec == NULL)
++ return;
++
++ for (i = 0; i < nr_pages; i++) {
++ free_foreign_page(pagevec[i]);
++ }
++
++ kfree(pagevec);
++}
++
++void balloon_dealloc_empty_page_range(
++ struct page *page, unsigned long nr_pages)
++{
++ __free_pages(page, get_order(nr_pages * PAGE_SIZE));
++}
++
++void balloon_update_driver_allowance(long delta)
++{
++}
++
++void balloon_release_driver_page(struct page *page)
++{
++ BUG();
++}
++
++EXPORT_SYMBOL_GPL(balloon_update_driver_allowance);
++EXPORT_SYMBOL_GPL(alloc_empty_pages_and_pagevec);
++EXPORT_SYMBOL_GPL(free_empty_pages_and_pagevec);
++EXPORT_SYMBOL_GPL(balloon_release_driver_page);
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/gnttab.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/gnttab.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,468 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
++
++#include <linux/config.h>
++#include <linux/vmalloc.h>
++#include <linux/memory_hotplug.h>
++#include <xen/gnttab.h>
++#include <asm/hypervisor.h>
++#include <xen/interface/grant_table.h>
++#include <asm/pgtable.h>
++#include <asm/sections.h>
++#include <asm/io.h>
++#include <asm/machdep.h>
++#include <asm/prom.h>
++#include <asm/cacheflush.h>
++#include "setup.h"
++#include "../pseries/plpar_wrappers.h"
++
++#undef DEBUG
++
++#ifdef DEBUG
++#define DBG(fmt...) printk(KERN_EMERG fmt)
++#else
++#define DBG(fmt...)
++#endif
++
++#define NR_GRANT_FRAMES 4
++
++struct address_space xen_foreign_dummy_mapping;
++
++static ulong foreign_map_pfn;
++static ulong foreign_map_pgs;
++static unsigned long *foreign_map_bitmap;
++
++
++/* hijack _mapcount */
++static inline int gnt_mapcount(struct page *page)
++{
++ return atomic_read(&(page)->_mapcount) + 1;
++}
++
++static inline int gnt_map(struct page *page)
++{
++ /* return true is transition from -1 to 0 */
++ return atomic_inc_and_test(&page->_mapcount);
++}
++
++static inline int gnt_unmap(struct page *page)
++{
++ int val;
++
++ val = atomic_dec_return(&page->_mapcount);
++ if (val < -1) {
++ atomic_inc(&page->_mapcount);
++ printk(KERN_EMERG "%s: %d\n", __func__, val);
++ }
++
++ return (val == -1);
++}
++
++
++static long map_to_linear(ulong paddr)
++{
++ unsigned long vaddr;
++ int psize;
++ unsigned long mode;
++ int slot;
++ uint shift;
++ unsigned long tmp_mode;
++
++ psize = MMU_PAGE_4K;
++ shift = mmu_psize_defs[psize].shift;
++ mode = _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_COHERENT | PP_RWXX;
++ vaddr = (ulong)__va(paddr);
++
++ {
++ unsigned long vpn, hash, hpteg;
++ unsigned long vsid = get_kernel_vsid(vaddr);
++ unsigned long va = (vsid << 28) | (vaddr & 0x0fffffff);
++
++ vpn = va >> shift;
++ tmp_mode = mode;
++
++ /* Make non-kernel text non-executable */
++ if (!in_kernel_text(vaddr))
++ tmp_mode = mode | HPTE_R_N;
++
++ hash = hpt_hash(va, shift);
++ hpteg = ((hash & htab_hash_mask) * HPTES_PER_GROUP);
++
++ BUG_ON(!ppc_md.hpte_insert);
++ slot = ppc_md.hpte_insert(hpteg, va, paddr,
++ tmp_mode, HPTE_V_BOLTED, psize);
++ if (slot < 0)
++ printk(KERN_EMERG
++ "%s: no more bolted entries "
++ "HTAB[0x%lx]: 0x%lx\n",
++ __func__, hpteg, paddr);
++ }
++ return slot;
++}
++
++static unsigned long get_hpte_vsid(ulong slot)
++{
++ unsigned long dword0;
++ unsigned long lpar_rc;
++ unsigned long dummy_word1;
++ unsigned long flags;
++
++ /* Read 1 pte at a time */
++ /* Do not need RPN to logical page translation */
++ /* No cross CEC PFT access */
++ flags = 0;
++
++ lpar_rc = plpar_pte_read(flags, slot, &dword0, &dummy_word1);
++
++ BUG_ON(lpar_rc != H_SUCCESS);
++
++ return dword0;
++}
++
++static long find_hpte_slot(unsigned long va, int psize)
++{
++ unsigned long hash;
++ unsigned long i, j;
++ long slot;
++ unsigned long want_v, hpte_v;
++
++ hash = hpt_hash(va, mmu_psize_defs[psize].shift);
++ want_v = hpte_encode_v(va, psize);
++
++ for (j = 0; j < 2; j++) {
++ slot = (hash & htab_hash_mask) * HPTES_PER_GROUP;
++ for (i = 0; i < HPTES_PER_GROUP; i++) {
++ hpte_v = get_hpte_vsid(slot);
++
++ if (HPTE_V_COMPARE(hpte_v, want_v)
++ && (hpte_v & HPTE_V_VALID)
++ && (!!(hpte_v & HPTE_V_SECONDARY) == j)) {
++ /* HPTE matches */
++ if (j)
++ slot = -slot;
++ return slot;
++ }
++ ++slot;
++ }
++ hash = ~hash;
++ }
++
++ return -1;
++}
++
++static long find_map_slot(ulong ea)
++{
++ int psize = MMU_PAGE_4K;
++ ulong vsid;
++ ulong va;
++
++ vsid = get_kernel_vsid(ea);
++ va = (vsid << 28) | (ea & 0x0fffffff);
++
++ return find_hpte_slot(va, psize);
++}
++
++
++static void gnttab_pre_unmap_grant_ref(
++ struct gnttab_unmap_grant_ref *unmap, int count)
++{
++ long slot;
++ int i;
++ ulong ea;
++ unsigned long dummy1, dummy2;
++ ulong flags;
++
++ /* paranoia */
++ local_irq_save(flags);
++
++ for (i = 0 ; i < count; i++) {
++ struct page *page;
++
++ ea = (ulong)__va(unmap[i].host_addr);
++ page = virt_to_page(ea);
++
++ if (!gnt_unmap(page)) {
++ DBG("%s[0x%x]: skip: 0x%lx, mapcount 0x%x\n",
++ __func__, i, ea, gnt_mapcount(page));
++ continue;
++ }
++ slot = find_map_slot(ea);
++ if (slot < 0) {
++ printk(KERN_EMERG "%s: PTE not found: 0x%lx\n",
++ __func__, ea);
++ continue;
++ }
++
++ DBG("%s[0x%x]: 0x%lx: mapcount: 0x%x\n",
++ __func__, i, ea, gnt_mapcount(page));
++ plpar_pte_remove(0, slot, 0, &dummy1, &dummy2);
++ }
++ local_irq_restore(flags);
++}
++
++static void gnttab_post_map_grant_ref(
++ struct gnttab_map_grant_ref *map, int count)
++{
++ int i;
++ long slot;
++ ulong flags;
++
++ /* paranoia */
++ local_irq_save(flags);
++
++ for (i = 0 ; i < count; i++) {
++ ulong pa = map[i].host_addr;
++ struct page *page;
++
++ if (map[i].status != GNTST_okay) {
++ printk(KERN_EMERG "%s: status, skip\n", __func__);
++ continue;
++ }
++
++ BUG_ON(pa < (foreign_map_pfn << PAGE_SHIFT));
++ BUG_ON(pa >= (foreign_map_pfn << PAGE_SHIFT) +
++ (foreign_map_pgs << PAGE_SHIFT));
++
++ page = virt_to_page(__va(pa));
++
++ if (gnt_map(page)) {
++#ifdef DEBUG
++ /* we need to get smarted than this */
++ slot = find_map_slot((ulong)__va(pa));
++ if (slot >= 0) {
++ DBG("%s: redundant 0x%lx\n", __func__, pa);
++ continue;
++ }
++#endif
++ slot = map_to_linear(pa);
++ DBG("%s[0x%x]: 0x%lx, mapcount:0x%x\n",
++ __func__, i, pa, gnt_mapcount(page));
++
++ } else {
++ DBG("%s[0x%x] skip 0x%lx, mapcount:0x%x\n",
++ __func__, i, pa, gnt_mapcount(page));
++ }
++ }
++ local_irq_restore(flags);
++}
++
++int HYPERVISOR_grant_table_op(unsigned int cmd, void *op, unsigned int count)
++{
++ void *desc;
++ void *frame_list = NULL;
++ int argsize;
++ int ret = -ENOMEM;
++
++ switch (cmd) {
++ case GNTTABOP_map_grant_ref:
++ argsize = sizeof(struct gnttab_map_grant_ref);
++ break;
++ case GNTTABOP_unmap_grant_ref:
++ gnttab_pre_unmap_grant_ref(op, count);
++ argsize = sizeof(struct gnttab_unmap_grant_ref);
++ break;
++ case GNTTABOP_setup_table: {
++ struct gnttab_setup_table setup;
++
++ memcpy(&setup, op, sizeof(setup));
++ argsize = sizeof(setup);
++
++ frame_list = xencomm_map(
++ xen_guest_handle(setup.frame_list),
++ (sizeof(*xen_guest_handle(setup.frame_list))
++ * setup.nr_frames));
++
++ if (frame_list == NULL)
++ return -ENOMEM;
++
++ set_xen_guest_handle(setup.frame_list, frame_list);
++ memcpy(op, &setup, sizeof(setup));
++ }
++ break;
++ case GNTTABOP_dump_table:
++ argsize = sizeof(struct gnttab_dump_table);
++ break;
++ case GNTTABOP_transfer:
++ BUG();
++ argsize = sizeof(struct gnttab_transfer);
++ break;
++ case GNTTABOP_copy:
++ argsize = sizeof(struct gnttab_transfer);
++ break;
++ case GNTTABOP_query_size:
++ argsize = sizeof(struct gnttab_query_size);
++ break;
++ default:
++ printk(KERN_EMERG "%s: unknown grant table op %d\n",
++ __func__, cmd);
++ return -ENOSYS;
++ }
++
++ desc = xencomm_map_no_alloc(op, argsize);
++ if (desc) {
++ ret = plpar_hcall_norets(XEN_MARK(__HYPERVISOR_grant_table_op),
++ cmd, desc, count);
++ if (!ret && cmd == GNTTABOP_map_grant_ref)
++ gnttab_post_map_grant_ref(op, count);
++ xencomm_free(desc);
++ }
++ xencomm_free(frame_list);
++
++ return ret;
++}
++EXPORT_SYMBOL(HYPERVISOR_grant_table_op);
++
++static ulong find_grant_maps(void)
++{
++ struct device_node *xen;
++ u64 *gm;
++ u64 _gm[2];
++ u64 expect;
++
++ /* This value is currently hardcoded into the SLB logic that
++ * it written in assempler, See
++ * slb_miss_kernel_load_xen_linear for more information.
++ * Anything else and we can not run. */
++ expect = 34 - PAGE_SHIFT;
++
++ xen = of_find_node_by_path("/xen");
++
++ /*
++ * The foreign is 2x2 Cells.
++ * The first entry is log2 of the base page frame.
++ * The second is the number of pages
++ */
++ gm = (u64 *)get_property(xen, "foreign-map", NULL);
++ if (gm == NULL) {
++ if (!is_initial_xendomain()) {
++ printk("OF: /xen/foreign-map not present\n");
++ _gm[0] = expect;
++ _gm[1] = 2048;
++ gm = _gm;
++ } else
++ panic("OF: /xen/foreign-map must be present\n");
++ }
++
++ if (gm[0] != expect)
++ panic("foreign-map is 0x%lx, expect 0x%lx\n",
++ gm[0], expect);
++
++ foreign_map_pfn = 1UL << gm[0];
++ return gm[1];
++}
++
++static void setup_foreign_segment(void)
++{
++ extern int *slb_miss_kernel_load_xen_nop;
++ ulong iaddr = (ulong)slb_miss_kernel_load_xen_nop;
++
++ /* By default Linux will branch around this logic we replace
++ * the branch with a NOP to turn the logic on */
++ *slb_miss_kernel_load_xen_nop = 0x60000000;
++ flush_icache_range(iaddr, iaddr + 4);
++}
++
++struct page *alloc_foreign_page(void)
++{
++ ulong bit;
++ do {
++ bit = find_first_zero_bit(foreign_map_bitmap,
++ foreign_map_pgs);
++ if (bit >= foreign_map_pgs)
++ return NULL;
++ } while (test_and_set_bit(bit, foreign_map_bitmap) == 1);
++
++ return pfn_to_page(foreign_map_pfn + bit);
++}
++
++void free_foreign_page(struct page *page)
++{
++ ulong bit = page_to_pfn(page) - foreign_map_pfn;
++
++ BUG_ON(bit >= foreign_map_pgs);
++ BUG_ON(!test_bit(bit, foreign_map_bitmap));
++
++ clear_bit(bit, foreign_map_bitmap);
++}
++
++static void setup_grant_area(void)
++{
++ ulong pgs;
++ int err;
++ struct zone *zone;
++ struct pglist_data *pgdata;
++ int nid;
++
++ pgs = find_grant_maps();
++ setup_foreign_segment();
++
++ printk("%s: Xen VIO will use a foreign address space of 0x%lx pages\n",
++ __func__, pgs);
++
++ /* add pages to the zone */
++ nid = 0;
++ pgdata = NODE_DATA(nid);
++ zone = pgdata->node_zones;
++
++ err = __add_pages(zone, foreign_map_pfn, pgs);
++
++ if (err < 0) {
++ printk(KERN_EMERG "%s: add_pages(0x%lx, 0x%lx) = %d\n",
++ __func__, foreign_map_pfn, pgs, err);
++ BUG();
++ }
++
++ /* create a bitmap to manage these pages */
++ foreign_map_bitmap = kmalloc(BITS_TO_LONGS(pgs) * sizeof(long),
++ GFP_KERNEL);
++ if (foreign_map_bitmap == NULL) {
++ printk(KERN_EMERG
++ "%s: could not allocate foreign_map_bitmap to "
++ "manage 0x%lx foreign pages\n", __func__, pgs);
++ BUG();
++ }
++ /* I'm paranoid so make sure we assign the top bits so we
++ * don't give them away */
++ bitmap_fill(&foreign_map_bitmap[BITS_TO_LONGS(pgs) - 1],
++ BITS_PER_LONG);
++ /* now clear all the real bits */
++ bitmap_zero(foreign_map_bitmap, pgs);
++
++ foreign_map_pgs = pgs;
++}
++
++void *arch_gnttab_alloc_shared(unsigned long *frames)
++{
++ void *shared;
++ ulong pa = frames[0] << PAGE_SHIFT;
++ static int resume;
++
++ shared = ioremap(pa, PAGE_SIZE * NR_GRANT_FRAMES);
++ BUG_ON(shared == NULL);
++ printk("%s: grant table at %p\n", __func__, shared);
++
++ /* no need to do the rest of this if we are resuming */
++ if (!resume)
++ setup_grant_area();
++
++ resume = 1;
++
++ return shared;
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/hcall.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/hcall.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,891 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006, 2007
++ *
++ * Authors: Hollis Blanchard <hollisb at us.ibm.com>
++ */
++
++#include <linux/types.h>
++#include <linux/errno.h>
++#include <linux/kernel.h>
++#include <linux/gfp.h>
++#include <linux/module.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/domctl.h>
++#include <xen/interface/sysctl.h>
++#include <xen/interface/platform.h>
++#include <xen/interface/memory.h>
++#include <xen/interface/xencomm.h>
++#include <xen/interface/version.h>
++#include <xen/interface/sched.h>
++#include <xen/interface/event_channel.h>
++#include <xen/interface/physdev.h>
++#include <xen/interface/vcpu.h>
++#include <xen/interface/xsm/acm_ops.h>
++#include <xen/interface/kexec.h>
++#include <xen/public/privcmd.h>
++#include <asm/hypercall.h>
++#include <asm/page.h>
++#include <asm/uaccess.h>
++#include <asm/hvcall.h>
++#include "setup.h"
++
++/* Xencomm notes:
++ *
++ * For kernel memory, we assume that virtually contiguous pages are also
++ * physically contiguous. This allows us to avoid creating descriptors for
++ * kernel hypercalls, such as console and event channel operations.
++ *
++ * In general, we need a xencomm descriptor to cover the top-level data
++ * structure (e.g. the domctl op), plus another for every embedded pointer to
++ * another data structure (i.e. for every GUEST_HANDLE).
++ */
++
++int HYPERVISOR_console_io(int cmd, int count, char *str)
++{
++ struct xencomm_handle *desc;
++ int rc;
++
++ desc = xencomm_map_no_alloc(str, count);
++ if (desc == NULL)
++ return -EINVAL;
++
++ rc = plpar_hcall_norets(XEN_MARK(__HYPERVISOR_console_io),
++ cmd, count, desc);
++
++ xencomm_free(desc);
++
++ return rc;
++}
++EXPORT_SYMBOL(HYPERVISOR_console_io);
++
++int HYPERVISOR_event_channel_op(int cmd, void *op)
++{
++ int rc;
++
++ struct xencomm_handle *desc =
++ xencomm_map_no_alloc(op, sizeof(evtchn_op_t));
++ if (desc == NULL)
++ return -EINVAL;
++
++ rc = plpar_hcall_norets(XEN_MARK(__HYPERVISOR_event_channel_op),
++ cmd, desc);
++
++ xencomm_free(desc);
++
++ return rc;
++
++}
++EXPORT_SYMBOL(HYPERVISOR_event_channel_op);
++
++int HYPERVISOR_xen_version(int cmd, void *arg)
++{
++ struct xencomm_handle *desc;
++ const unsigned long hcall = __HYPERVISOR_xen_version;
++ int argsize;
++ int rc;
++
++ switch (cmd) {
++ case XENVER_version:
++ /* do not actually pass an argument */
++ return plpar_hcall_norets(XEN_MARK(hcall), cmd, 0);
++ case XENVER_extraversion:
++ argsize = sizeof(xen_extraversion_t);
++ break;
++ case XENVER_compile_info:
++ argsize = sizeof(xen_compile_info_t);
++ break;
++ case XENVER_capabilities:
++ argsize = sizeof(xen_capabilities_info_t);
++ break;
++ case XENVER_changeset:
++ argsize = sizeof(xen_changeset_info_t);
++ break;
++ case XENVER_platform_parameters:
++ argsize = sizeof(xen_platform_parameters_t);
++ break;
++ case XENVER_pagesize:
++ if (arg == NULL)
++ argsize = 0;
++ else
++ argsize = sizeof(void *);
++ break;
++ case XENVER_get_features:
++ argsize = sizeof(xen_feature_info_t);
++ break;
++ default:
++ printk(KERN_ERR "%s: unknown version cmd %d\n", __func__, cmd);
++ return -ENOSYS;
++ }
++
++ /* desc could be NULL in the case of XENVER_pagesize with NULL arg */
++ desc = xencomm_map(arg, argsize);
++
++ rc = plpar_hcall_norets(XEN_MARK(hcall), cmd, desc);
++
++ xencomm_free(desc);
++
++ return rc;
++}
++EXPORT_SYMBOL(HYPERVISOR_xen_version);
++
++
++int HYPERVISOR_physdev_op(int cmd, void *op)
++{
++ struct xencomm_handle *desc =
++ xencomm_map_no_alloc(op, sizeof(physdev_op_t));
++ int rc;
++
++ if (desc == NULL)
++ return -EINVAL;
++
++ rc = plpar_hcall_norets(XEN_MARK(__HYPERVISOR_physdev_op),
++ cmd, desc);
++
++ xencomm_free(desc);
++
++ return rc;
++}
++EXPORT_SYMBOL(HYPERVISOR_physdev_op);
++
++int HYPERVISOR_sched_op(int cmd, void *arg)
++{
++ int argsize = 0;
++ int rc = -EINVAL;
++ struct xencomm_handle *desc;
++ struct xencomm_handle *ports = NULL;
++
++ switch (cmd) {
++ case SCHEDOP_yield:
++ case SCHEDOP_block:
++ return plpar_hcall_norets(XEN_MARK(__HYPERVISOR_sched_op),
++ cmd, 0);
++ break;
++
++ case SCHEDOP_poll: {
++ struct sched_poll sched_poll;
++
++ argsize = sizeof(struct sched_poll);
++
++ memcpy(&sched_poll, arg, sizeof(sched_poll));
++
++ ports = xencomm_map(
++ xen_guest_handle(sched_poll.ports),
++ (sizeof(evtchn_port_t) * sched_poll.nr_ports));
++
++ if (ports == NULL)
++ return -ENOMEM;
++
++ set_xen_guest_handle(sched_poll.ports, (evtchn_port_t *)ports);
++ memcpy(arg, &sched_poll, sizeof(sched_poll));
++
++ }
++ break;
++ case SCHEDOP_shutdown:
++ argsize = sizeof(struct sched_shutdown);
++ break;
++ case SCHEDOP_remote_shutdown:
++ argsize = sizeof(struct sched_remote_shutdown);
++ break;
++ default:
++ printk(KERN_ERR "%s: unknown sched op %d\n", __func__, cmd);
++ return -ENOSYS;
++ }
++
++ desc = xencomm_map_no_alloc(arg, argsize);
++ if (desc) {
++ rc = plpar_hcall_norets(XEN_MARK(__HYPERVISOR_sched_op),
++ cmd, desc);
++ xencomm_free(desc);
++ }
++
++ xencomm_free(ports);
++
++ return rc;
++}
++EXPORT_SYMBOL(HYPERVISOR_sched_op);
++
++int HYPERVISOR_suspend(unsigned long srec)
++{
++ int cmd = SCHEDOP_shutdown;
++ struct sched_shutdown sched_shutdown = {
++ .reason = SHUTDOWN_suspend,
++ };
++ struct xencomm_handle *desc;
++
++ desc = xencomm_map_no_alloc(&sched_shutdown, sizeof(struct sched_shutdown));
++
++ return plpar_hcall_norets(XEN_MARK(__HYPERVISOR_sched_op),
++ cmd, desc, srec);
++}
++EXPORT_SYMBOL(HYPERVISOR_suspend);
++
++int HYPERVISOR_kexec_op(unsigned long op, void *args)
++{
++ unsigned long argsize;
++ struct xencomm_handle *desc;
++
++ switch (op) {
++ case KEXEC_CMD_kexec_get_range:
++ argsize = sizeof(struct xen_kexec_range);
++ break;
++ case KEXEC_CMD_kexec_load:
++ argsize = sizeof(struct xen_kexec_load);
++ break;
++ case KEXEC_CMD_kexec_unload:
++ argsize = sizeof(struct xen_kexec_load);
++ break;
++ case KEXEC_CMD_kexec:
++ argsize = sizeof(struct xen_kexec_exec);
++ break;
++ default:
++ return -ENOSYS;
++ }
++ desc = xencomm_map_no_alloc(args, argsize);
++
++ return plpar_hcall_norets(XEN_MARK(__HYPERVISOR_kexec_op),
++ op, desc);
++}
++EXPORT_SYMBOL(HYPERVISOR_kexec_op);
++
++int HYPERVISOR_poll(
++ evtchn_port_t *ports, unsigned int nr_ports, u64 timeout)
++{
++ struct sched_poll sched_poll = {
++ .nr_ports = nr_ports,
++ .timeout = jiffies_to_ns(timeout)
++ };
++ set_xen_guest_handle(sched_poll.ports, ports);
++
++ return HYPERVISOR_sched_op(SCHEDOP_poll, &sched_poll);
++}
++EXPORT_SYMBOL(HYPERVISOR_poll);
++
++typedef ulong (mf_t)(ulong arg0, ...);
++
++static mf_t *multicall_funcs[] = {
++ [__HYPERVISOR_grant_table_op] = (mf_t *)HYPERVISOR_grant_table_op,
++};
++
++int HYPERVISOR_multicall(void *call_list, int nr_calls)
++{
++ /* we blow out the multicall because the xencomm stuff is jsut
++ * too tricky */
++ multicall_entry_t *mcl = (multicall_entry_t *)call_list;
++ multicall_entry_t *c;
++ int i;
++ mf_t *mf;
++ int res;
++ ulong flags;
++
++ /* let make sure all the calls are supported */
++ for (i = 0; i < nr_calls; i++) {
++ mf = multicall_funcs[mcl[i].op];
++ BUG_ON(mf == NULL);
++ }
++ /* disable interrupts until we are done all calls */
++ local_irq_save(flags);
++ for (i = 0; i < nr_calls; i++) {
++ /* lookup supported multicalls */
++ c = &mcl[i];
++ mf = multicall_funcs[c->op];
++ res = mf(c->args[0], c->args[1], c->args[2],
++ c->args[3], c->args[4], c->args[5]);
++ c->result = res;
++ }
++ local_irq_restore(flags);
++ return 0;
++}
++EXPORT_SYMBOL(HYPERVISOR_multicall);
++
++
++/* privcmd operations: */
++
++static int xenppc_privcmd_domctl(privcmd_hypercall_t *hypercall)
++{
++ xen_domctl_t kern_op;
++ xen_domctl_t __user *user_op = (xen_domctl_t __user *)hypercall->arg[0];
++ struct xencomm_handle *op_desc;
++ struct xencomm_handle *desc = NULL;
++ int ret = 0;
++
++ if (copy_from_user(&kern_op, user_op, sizeof(xen_domctl_t)))
++ return -EFAULT;
++
++ if (kern_op.interface_version != XEN_DOMCTL_INTERFACE_VERSION) {
++ printk(KERN_WARNING "%s: %s %x != %x\n", __func__, current->comm,
++ kern_op.interface_version, XEN_DOMCTL_INTERFACE_VERSION);
++ return -EACCES;
++ }
++
++ op_desc = xencomm_map(&kern_op, sizeof(xen_domctl_t));
++ if (op_desc == NULL)
++ return -ENOMEM;
++
++ switch (kern_op.cmd) {
++ case XEN_DOMCTL_createdomain:
++ case XEN_DOMCTL_destroydomain:
++ case XEN_DOMCTL_pausedomain:
++ case XEN_DOMCTL_unpausedomain:
++ case XEN_DOMCTL_getdomaininfo:
++ break;
++ case XEN_DOMCTL_getmemlist:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.getmemlist.buffer),
++ kern_op.u.getmemlist.max_pfns * sizeof(unsigned long));
++
++ if (desc == NULL)
++ ret = -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.getmemlist.buffer,
++ (void *)desc);
++ break;
++ case XEN_DOMCTL_getpageframeinfo:
++ break;
++ case XEN_DOMCTL_getpageframeinfo2:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.getpageframeinfo2.array),
++ kern_op.u.getpageframeinfo2.num);
++
++ if (desc == NULL)
++ ret = -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.getpageframeinfo2.array,
++ (void *)desc);
++ break;
++ case XEN_DOMCTL_shadow_op:
++
++ if (xen_guest_handle(kern_op.u.shadow_op.dirty_bitmap))
++ {
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.shadow_op.dirty_bitmap),
++ kern_op.u.shadow_op.pages * sizeof(unsigned long));
++
++ if (desc == NULL)
++ ret = -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.shadow_op.dirty_bitmap,
++ (void *)desc);
++ }
++ break;
++ case XEN_DOMCTL_max_mem:
++ break;
++ case XEN_DOMCTL_setvcpucontext:
++ case XEN_DOMCTL_getvcpucontext:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.vcpucontext.ctxt),
++ sizeof(vcpu_guest_context_t));
++
++ if (desc == NULL)
++ ret = -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.vcpucontext.ctxt,
++ (void *)desc);
++ break;
++ case XEN_DOMCTL_getvcpuinfo:
++ break;
++ case XEN_DOMCTL_setvcpuaffinity:
++ case XEN_DOMCTL_getvcpuaffinity:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.vcpuaffinity.cpumap.bitmap),
++ (kern_op.u.vcpuaffinity.cpumap.nr_cpus + 7) / 8);
++
++ if (desc == NULL)
++ ret = -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.vcpuaffinity.cpumap.bitmap,
++ (void *)desc);
++ break;
++ case XEN_DOMCTL_max_vcpus:
++ case XEN_DOMCTL_scheduler_op:
++ case XEN_DOMCTL_setdomainhandle:
++ case XEN_DOMCTL_setdebugging:
++ case XEN_DOMCTL_irq_permission:
++ case XEN_DOMCTL_iomem_permission:
++ case XEN_DOMCTL_ioport_permission:
++ case XEN_DOMCTL_hypercall_init:
++ case XEN_DOMCTL_arch_setup:
++ case XEN_DOMCTL_settimeoffset:
++ case XEN_DOMCTL_real_mode_area:
++ break;
++ default:
++ printk(KERN_ERR "%s: unknown domctl cmd %d\n", __func__, kern_op.cmd);
++ return -ENOSYS;
++ }
++
++ if (ret)
++ goto out; /* error mapping the nested pointer */
++
++ ret = plpar_hcall_norets(XEN_MARK(hypercall->op),op_desc);
++
++ if (copy_to_user(user_op, &kern_op, sizeof(xen_domctl_t)))
++ ret = -EFAULT;
++
++out:
++ xencomm_free(desc);
++ xencomm_free(op_desc);
++ return ret;
++}
++
++static int xenppc_privcmd_sysctl(privcmd_hypercall_t *hypercall)
++{
++ xen_sysctl_t kern_op;
++ xen_sysctl_t __user *user_op = (xen_sysctl_t __user *)hypercall->arg[0];
++ struct xencomm_handle *op_desc;
++ struct xencomm_handle *desc = NULL;
++ int ret = 0;
++
++ if (copy_from_user(&kern_op, user_op, sizeof(xen_sysctl_t)))
++ return -EFAULT;
++
++ if (kern_op.interface_version != XEN_SYSCTL_INTERFACE_VERSION) {
++ printk(KERN_WARNING "%s: %s %x != %x\n", __func__, current->comm,
++ kern_op.interface_version, XEN_SYSCTL_INTERFACE_VERSION);
++ return -EACCES;
++ }
++
++ op_desc = xencomm_map(&kern_op, sizeof(xen_sysctl_t));
++
++ if (op_desc == NULL)
++ return -ENOMEM;
++
++ switch (kern_op.cmd) {
++ case XEN_SYSCTL_readconsole:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.readconsole.buffer),
++ kern_op.u.readconsole.count);
++
++ if (desc == NULL)
++ ret = -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.readconsole.buffer,
++ (void *)desc);
++ break;
++ case XEN_SYSCTL_tbuf_op:
++ case XEN_SYSCTL_physinfo:
++ case XEN_SYSCTL_sched_id:
++ break;
++ case XEN_SYSCTL_perfc_op:
++ /* XXX this requires *two* embedded xencomm mappings (desc and val),
++ * and I don't feel like it right now. */
++ printk(KERN_ERR "%s: unknown sysctl cmd %d\n", __func__, kern_op.cmd);
++ return -ENOSYS;
++ case XEN_SYSCTL_getdomaininfolist:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.getdomaininfolist.buffer),
++ kern_op.u.getdomaininfolist.max_domains *
++ sizeof(xen_domctl_getdomaininfo_t));
++
++ if (desc == NULL)
++ ret = -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.getdomaininfolist.buffer,
++ (void *)desc);
++ break;
++ default:
++ printk(KERN_ERR "%s: unknown sysctl cmd %d\n", __func__, kern_op.cmd);
++ return -ENOSYS;
++ }
++
++ if (ret)
++ goto out; /* error mapping the nested pointer */
++
++ ret = plpar_hcall_norets(XEN_MARK(hypercall->op), op_desc);
++
++ if (copy_to_user(user_op, &kern_op, sizeof(xen_sysctl_t)))
++ ret = -EFAULT;
++
++out:
++ xencomm_free(desc);
++ xencomm_free(op_desc);
++ return ret;
++}
++
++static int xenppc_privcmd_platform_op(privcmd_hypercall_t *hypercall)
++{
++ xen_platform_op_t kern_op;
++ xen_platform_op_t __user *user_op =
++ (xen_platform_op_t __user *)hypercall->arg[0];
++ struct xencomm_handle *op_desc;
++ struct xencomm_handle *desc = NULL;
++ int ret = 0;
++
++ if (copy_from_user(&kern_op, user_op, sizeof(xen_platform_op_t)))
++ return -EFAULT;
++
++ if (kern_op.interface_version != XENPF_INTERFACE_VERSION) {
++ printk(KERN_WARNING "%s: %s %x != %x\n", __func__, current->comm,
++ kern_op.interface_version, XENPF_INTERFACE_VERSION);
++ return -EACCES;
++ }
++
++ op_desc = xencomm_map(&kern_op, sizeof(xen_platform_op_t));
++
++ if (op_desc == NULL)
++ return -ENOMEM;
++
++ switch (kern_op.cmd) {
++ case XENPF_settime:
++ case XENPF_add_memtype:
++ case XENPF_del_memtype:
++ case XENPF_read_memtype:
++ case XENPF_microcode_update:
++ case XENPF_platform_quirk:
++ break;
++ default:
++ printk(KERN_ERR "%s: unknown platform_op cmd %d\n", __func__,
++ kern_op.cmd);
++ return -ENOSYS;
++ }
++
++ if (ret)
++ goto out; /* error mapping the nested pointer */
++
++ ret = plpar_hcall_norets(XEN_MARK(hypercall->op), op_desc);
++
++ if (copy_to_user(user_op, &kern_op, sizeof(xen_platform_op_t)))
++ ret = -EFAULT;
++
++out:
++ xencomm_free(desc);
++ xencomm_free(op_desc);
++ return ret;
++}
++
++int HYPERVISOR_memory_op(unsigned int cmd, void *arg)
++{
++ int ret;
++ struct xencomm_handle *op_desc;
++ xen_memory_reservation_t *mop;
++
++
++ mop = (xen_memory_reservation_t *)arg;
++
++ op_desc = xencomm_map(mop, sizeof(xen_memory_reservation_t));
++
++ if (op_desc == NULL)
++ return -ENOMEM;
++
++ switch (cmd) {
++ case XENMEM_increase_reservation:
++ case XENMEM_decrease_reservation:
++ case XENMEM_populate_physmap: {
++ struct xencomm_handle *desc = NULL;
++
++ if (xen_guest_handle(mop->extent_start)) {
++ desc = xencomm_map(
++ xen_guest_handle(mop->extent_start),
++ mop->nr_extents *
++ sizeof(*xen_guest_handle(mop->extent_start)));
++
++ if (desc == NULL) {
++ ret = -ENOMEM;
++ goto out;
++ }
++
++ set_xen_guest_handle(mop->extent_start,
++ (void *)desc);
++ }
++
++ ret = plpar_hcall_norets(XEN_MARK(__HYPERVISOR_memory_op),
++ cmd, op_desc);
++
++ xencomm_free(desc);
++ }
++ break;
++
++ case XENMEM_maximum_ram_page:
++ /* arg is NULL so we can call thru here */
++ ret = plpar_hcall_norets(XEN_MARK(__HYPERVISOR_memory_op),
++ cmd, NULL);
++ break;
++ default:
++ printk(KERN_ERR "%s: unknown memory op %d\n", __func__, cmd);
++ ret = -ENOSYS;
++ }
++
++out:
++ xencomm_free(op_desc);
++ return ret;
++}
++EXPORT_SYMBOL(HYPERVISOR_memory_op);
++
++static int xenppc_privcmd_memory_op(privcmd_hypercall_t *hypercall)
++{
++ xen_memory_reservation_t kern_op;
++ xen_memory_reservation_t __user *user_op;
++ const unsigned long cmd = hypercall->arg[0];
++ int ret = 0;
++
++ user_op = (xen_memory_reservation_t __user *)hypercall->arg[1];
++ if (copy_from_user(&kern_op, user_op,
++ sizeof(xen_memory_reservation_t)))
++ return -EFAULT;
++
++ ret = HYPERVISOR_memory_op(cmd, &kern_op);
++ if (ret >= 0) {
++ if (copy_to_user(user_op, &kern_op,
++ sizeof(xen_memory_reservation_t)))
++ return -EFAULT;
++ }
++ return ret;
++}
++
++static int xenppc_privcmd_version(privcmd_hypercall_t *hypercall)
++{
++ return HYPERVISOR_xen_version(hypercall->arg[0],
++ (void *)hypercall->arg[1]);
++}
++
++static int xenppc_privcmd_event_channel_op(privcmd_hypercall_t *hypercall)
++{
++ struct xencomm_handle *desc;
++ unsigned int argsize;
++ int ret;
++
++ switch (hypercall->arg[0]) {
++ case EVTCHNOP_alloc_unbound:
++ argsize = sizeof(evtchn_alloc_unbound_t);
++ break;
++
++ case EVTCHNOP_status:
++ argsize = sizeof(evtchn_status_t);
++ break;
++
++ default:
++ printk(KERN_ERR "%s: unknown EVTCHNOP (%ld)\n",
++ __func__, hypercall->arg[0]);
++ return -EINVAL;
++ }
++
++ desc = xencomm_map((void *)hypercall->arg[1], argsize);
++
++ if (desc == NULL)
++ return -ENOMEM;
++
++ ret = plpar_hcall_norets(XEN_MARK(hypercall->op), hypercall->arg[0],
++ desc);
++
++ xencomm_free(desc);
++ return ret;
++}
++
++static int xenppc_acmcmd_op(privcmd_hypercall_t *hypercall)
++{
++ xen_acmctl_t kern_op;
++ xen_acmctl_t __user *user_op = (xen_acmctl_t __user *)hypercall->arg[0];
++ void *op_desc;
++ void *desc = NULL, *desc2 = NULL, *desc3 = NULL, *desc4 = NULL;
++ int ret = 0;
++
++ if (copy_from_user(&kern_op, user_op, sizeof(xen_acmctl_t)))
++ return -EFAULT;
++
++ if (kern_op.interface_version != ACM_INTERFACE_VERSION) {
++ printk(KERN_WARNING "%s: %s %x != %x\n", __func__, current->comm,
++ kern_op.interface_version, ACM_INTERFACE_VERSION);
++ return -EACCES;
++ }
++
++ op_desc = xencomm_map(&kern_op, sizeof(xen_acmctl_t));
++ if (op_desc == NULL)
++ return -ENOMEM;
++
++ switch (kern_op.cmd) {
++ case ACMOP_setpolicy:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.setpolicy.pushcache),
++ kern_op.u.setpolicy.pushcache_size);
++
++ if (desc == NULL)
++ ret = -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.setpolicy.pushcache,
++ desc);
++ break;
++ case ACMOP_getpolicy:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.getpolicy.pullcache),
++ kern_op.u.getpolicy.pullcache_size);
++
++ if (desc == NULL)
++ ret = -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.getpolicy.pullcache,
++ desc);
++ break;
++ case ACMOP_dumpstats:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.dumpstats.pullcache),
++ kern_op.u.dumpstats.pullcache_size);
++
++ if (desc == NULL)
++ ret = -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.dumpstats.pullcache,
++ desc);
++ break;
++ case ACMOP_getssid:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.getssid.ssidbuf),
++ kern_op.u.getssid.ssidbuf_size);
++
++ if (desc == NULL)
++ ret = -ENOMEM;
++
++ set_xen_guest_handle(kern_op.u.getssid.ssidbuf,
++ desc);
++ break;
++ case ACMOP_getdecision:
++ break;
++ case ACMOP_chgpolicy:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.change_policy.policy_pushcache),
++ kern_op.u.change_policy.policy_pushcache_size);
++ desc2 = xencomm_map(
++ xen_guest_handle(kern_op.u.change_policy.del_array),
++ kern_op.u.change_policy.delarray_size);
++ desc3 = xencomm_map(
++ xen_guest_handle(kern_op.u.change_policy.chg_array),
++ kern_op.u.change_policy.chgarray_size);
++ desc4 = xencomm_map(
++ xen_guest_handle(kern_op.u.change_policy.err_array),
++ kern_op.u.change_policy.errarray_size);
++
++ if (desc == NULL || desc2 == NULL ||
++ desc3 == NULL || desc4 == NULL) {
++ ret = -ENOMEM;
++ goto out;
++ }
++
++ set_xen_guest_handle(kern_op.u.change_policy.policy_pushcache,
++ desc);
++ set_xen_guest_handle(kern_op.u.change_policy.del_array,
++ desc2);
++ set_xen_guest_handle(kern_op.u.change_policy.chg_array,
++ desc3);
++ set_xen_guest_handle(kern_op.u.change_policy.err_array,
++ desc4);
++ break;
++ case ACMOP_relabeldoms:
++ desc = xencomm_map(
++ xen_guest_handle(kern_op.u.relabel_doms.relabel_map),
++ kern_op.u.relabel_doms.relabel_map_size);
++ desc2 = xencomm_map(
++ xen_guest_handle(kern_op.u.relabel_doms.err_array),
++ kern_op.u.relabel_doms.errarray_size);
++
++ if (desc == NULL || desc2 == NULL) {
++ ret = -ENOMEM;
++ goto out;
++ }
++
++ set_xen_guest_handle(kern_op.u.relabel_doms.relabel_map,
++ desc);
++ set_xen_guest_handle(kern_op.u.relabel_doms.err_array,
++ desc2);
++ break;
++ default:
++ printk(KERN_ERR "%s: unknown/unsupported acmctl cmd %d\n",
++ __func__, kern_op.cmd);
++ return -ENOSYS;
++ }
++
++ if (ret)
++ goto out; /* error mapping the nested pointer */
++
++ ret = plpar_hcall_norets(XEN_MARK(hypercall->op),op_desc);
++
++ if (copy_to_user(user_op, &kern_op, sizeof(xen_acmctl_t)))
++ ret = -EFAULT;
++
++out:
++ xencomm_free(desc);
++ xencomm_free(desc2);
++ xencomm_free(desc3);
++ xencomm_free(desc4);
++ xencomm_free(op_desc);
++ return ret;
++}
++
++
++/* The PowerPC hypervisor runs in a separate address space from Linux
++ * kernel/userspace, i.e. real mode. We must therefore translate userspace
++ * pointers to something the hypervisor can make sense of. */
++int privcmd_hypercall(privcmd_hypercall_t *hypercall)
++{
++ switch (hypercall->op) {
++ case __HYPERVISOR_domctl:
++ return xenppc_privcmd_domctl(hypercall);
++ case __HYPERVISOR_sysctl:
++ return xenppc_privcmd_sysctl(hypercall);
++ case __HYPERVISOR_platform_op:
++ return xenppc_privcmd_platform_op(hypercall);
++ case __HYPERVISOR_memory_op:
++ return xenppc_privcmd_memory_op(hypercall);
++ case __HYPERVISOR_xen_version:
++ return xenppc_privcmd_version(hypercall);
++ case __HYPERVISOR_event_channel_op:
++ return xenppc_privcmd_event_channel_op(hypercall);
++ case __HYPERVISOR_acm_op:
++ return xenppc_acmcmd_op(hypercall);
++ default:
++ printk(KERN_ERR "%s: unknown hcall (%ld)\n", __func__, hypercall->op);
++ /* maybe we'll get lucky and the hcall needs no translation. */
++ return plpar_hcall_norets(XEN_MARK(hypercall->op),
++ hypercall->arg[0],
++ hypercall->arg[1],
++ hypercall->arg[2],
++ hypercall->arg[3],
++ hypercall->arg[4]);
++ }
++}
++
++int HYPERVISOR_vcpu_op(int cmd, int vcpuid, void *extra_args)
++{
++ int argsize;
++ const unsigned long hcall = __HYPERVISOR_vcpu_op;
++ struct xencomm_handle *desc;
++ int rc;
++
++ switch (cmd) {
++ case VCPUOP_initialise:
++ argsize = sizeof(vcpu_guest_context_t);
++ break;
++ case VCPUOP_up:
++ case VCPUOP_down:
++ case VCPUOP_is_up:
++ return plpar_hcall_norets(XEN_MARK(hcall), cmd, vcpuid, 0);
++
++ case VCPUOP_get_runstate_info:
++ argsize = sizeof (vcpu_runstate_info_t);
++ break;
++ default:
++ printk(KERN_ERR "%s: unknown version cmd %d\n", __func__, cmd);
++ return -ENOSYS;
++ }
++
++ desc = xencomm_map_no_alloc(extra_args, argsize);
++
++ if (desc == NULL)
++ return -EINVAL;
++
++ rc = plpar_hcall_norets(XEN_MARK(hcall), cmd, vcpuid, desc);
++
++ xencomm_free(desc);
++
++ return rc;
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/reboot.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/reboot.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,53 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
++
++#include <linux/module.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/io/console.h>
++#include <xen/xencons.h>
++#include <asm/hypervisor.h>
++#include <asm/machdep.h>
++
++static void domain_machine_restart(char * __unused)
++{
++ /* We really want to get pending console data out before we die. */
++ xencons_force_flush();
++ HYPERVISOR_shutdown(SHUTDOWN_reboot);
++}
++
++static void domain_machine_power_off(void)
++{
++ /* We really want to get pending console data out before we die. */
++ xencons_force_flush();
++ HYPERVISOR_shutdown(SHUTDOWN_poweroff);
++}
++
++void xen_reboot_init(struct machdep_calls *md)
++{
++ if (md != NULL) {
++ ppc_md.restart = md->restart;
++ ppc_md.power_off = md->power_off;
++ ppc_md.halt = md->halt;
++ } else {
++ ppc_md.restart = domain_machine_restart;
++ ppc_md.power_off = domain_machine_power_off;
++ ppc_md.halt = domain_machine_power_off;
++ }
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/setup.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/setup.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,336 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
++
++#define DEBUG
++#define CONFIG_SHARE_MPIC
++
++#include <linux/module.h>
++#include <linux/rwsem.h>
++#include <linux/delay.h>
++#include <linux/console.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/sched.h>
++#include <xen/evtchn.h>
++#include <xen/features.h>
++#include <xen/xencons.h>
++#include <asm/udbg.h>
++#include <asm/pgtable.h>
++#include <asm/prom.h>
++#include <asm/iommu.h>
++#include <asm/mmu.h>
++#include <asm/abs_addr.h>
++#include <asm/machdep.h>
++#include <asm/hypervisor.h>
++#include <asm/time.h>
++#include <asm/pmc.h>
++#include "setup.h"
++
++#ifdef DEBUG
++#define DBG(fmt...) udbg_printf(fmt)
++#else
++#define DBG(fmt...)
++#endif
++
++/* Apperently on other arches this could be used before its defined,
++ * this should not be the case in PPC */
++shared_info_t *HYPERVISOR_shared_info = (shared_info_t *)NULL;
++EXPORT_SYMBOL(HYPERVISOR_shared_info);
++
++/* Raw start-of-day parameters from the hypervisor. */
++static start_info_t xsi;
++start_info_t *xen_start_info;
++EXPORT_SYMBOL(xen_start_info);
++
++extern struct machdep_calls mach_maple_md;
++extern void maple_pci_init(void);
++
++static unsigned long foreign_mfn_flag;
++
++/* Must be called with &vma->vm_mm->mmap_sem locked for write */
++int direct_remap_pfn_range(struct vm_area_struct *vma,
++ unsigned long address,
++ unsigned long mfn,
++ unsigned long size,
++ pgprot_t prot,
++ domid_t domid)
++{
++ int rc;
++
++ /* Set the MFN flag to tell Xen that this is not a PFN. */
++ printk("%s: mapping mfn 0x%lx (size 0x%lx) -> 0x%lx\n", __func__,
++ mfn, size, mfn | foreign_mfn_flag);
++ mfn = mfn | foreign_mfn_flag;
++
++ WARN_ON(!rwsem_is_locked(&vma->vm_mm->mmap_sem));
++ rc = remap_pfn_range(vma, address, mfn, size, prot);
++
++ return rc;
++}
++
++static void __init xen_fw_feature_init(void)
++{
++ DBG(" -> %s\n", __func__);
++
++ powerpc_firmware_features = 0;
++
++ powerpc_firmware_features |= FW_FEATURE_LPAR;
++ powerpc_firmware_features |= FW_FEATURE_TCE | FW_FEATURE_DABR;
++
++ printk(KERN_INFO "firmware_features = 0x%lx\n",
++ powerpc_firmware_features);
++
++ DBG(" <- %s\n", __func__);
++}
++
++/* if these were global then I could get them from the pseries/setup.c */
++static int pseries_set_dabr(unsigned long dabr)
++{
++ return plpar_hcall_norets(H_SET_DABR, dabr);
++}
++
++static int pseries_set_xdabr(unsigned long dabr)
++{
++ /* We want to catch accesses from kernel and userspace */
++ return plpar_hcall_norets(H_SET_XDABR, dabr,
++ H_DABRX_KERNEL | H_DABRX_USER);
++}
++
++/*
++ * Early initialization.
++ */
++static void __init xenppc_init_early(void)
++{
++ struct device_node *xen;
++
++ DBG(" -> %s\n", __func__);
++
++ xen = of_find_node_by_path("/xen");
++
++ xen_start_info = &xsi;
++
++ /* fill out start_info_t from devtree */
++ if ((char *)get_property(xen, "privileged", NULL))
++ xen_start_info->flags |= SIF_PRIVILEGED;
++ if ((char *)get_property(xen, "initdomain", NULL))
++ xen_start_info->flags |= SIF_INITDOMAIN;
++ xen_start_info->shared_info = *((u64 *)get_property(xen,
++ "shared-info", NULL));
++
++ /* only look for store and console for guest domains */
++ if (xen_start_info->flags == 0) {
++ struct device_node *console = of_find_node_by_path("/xen/console");
++ struct device_node *store = of_find_node_by_path("/xen/store");
++
++ xen_start_info->store_mfn = (*((u64 *)get_property(store,
++ "reg", NULL))) >> PAGE_SHIFT;
++ xen_start_info->store_evtchn = *((u32 *)get_property(store,
++ "interrupts", NULL));
++ xen_start_info->console.domU.mfn = (*((u64 *)get_property(console,
++ "reg", NULL))) >> PAGE_SHIFT;
++ xen_start_info->console.domU.evtchn = *((u32 *)get_property(console,
++ "interrupts", NULL));
++ }
++
++ HYPERVISOR_shared_info = __va(xen_start_info->shared_info);
++
++ udbg_init_xen();
++
++ DBG("xen_start_info at %p\n", xen_start_info);
++ DBG(" magic %s\n", xen_start_info->magic);
++ DBG(" flags %x\n", xen_start_info->flags);
++ DBG(" shared_info %lx, %p\n",
++ xen_start_info->shared_info, HYPERVISOR_shared_info);
++ DBG(" store_mfn %llx\n", xen_start_info->store_mfn);
++ DBG(" store_evtchn %x\n", xen_start_info->store_evtchn);
++ DBG(" console_mfn %llx\n", xen_start_info->console.domU.mfn);
++ DBG(" console_evtchn %x\n", xen_start_info->console.domU.evtchn);
++
++ xen_setup_time(&mach_maple_md);
++
++ add_preferred_console("xvc", 0, NULL);
++
++ if (get_property(xen, "power-control", NULL))
++ xen_reboot_init(&mach_maple_md);
++ else
++ xen_reboot_init(NULL);
++
++ if (is_initial_xendomain()) {
++ u64 *mfnflag = (u64 *)get_property(xen, "mfn-flag", NULL);
++ if (mfnflag) {
++ foreign_mfn_flag = (1UL << mfnflag[0]);
++ printk("OF: using 0x%lx as foreign mfn flag\n", foreign_mfn_flag);
++ } else
++ printk("OF: /xen/mfn-base must be present it build guests\n");
++ }
++
++ /* get the domain features */
++ setup_xen_features();
++
++ DBG("Hello World I'm Maple Xen-LPAR!\n");
++
++ if (firmware_has_feature(FW_FEATURE_DABR))
++ ppc_md.set_dabr = pseries_set_dabr;
++ else if (firmware_has_feature(FW_FEATURE_XDABR))
++ ppc_md.set_dabr = pseries_set_xdabr;
++
++ iommu_init_early_pSeries();
++
++ DBG(" <- %s\n", __func__);
++}
++
++/*
++ * this interface is limiting
++ */
++static int running_on_xen;
++int is_running_on_xen(void)
++{
++ return running_on_xen;
++}
++EXPORT_SYMBOL(is_running_on_xen);
++
++static void xenppc_power_save(void)
++{
++ /* SCHEDOP_yield could immediately return. Instead, we
++ * want to idle in the Xen idle domain, so use
++ * SCHEDOP_block with a one-shot timer. */
++ /* XXX do tickless stuff here. See
++ * linux-2.6-xen-sparse/arch/xen/i386/kernel/time.c */
++ u64 now_ns = tb_to_ns(get_tb());
++ u64 offset_ns = jiffies_to_ns(1);
++ int rc;
++
++ rc = HYPERVISOR_set_timer_op(now_ns + offset_ns);
++ BUG_ON(rc != 0);
++
++ HYPERVISOR_sched_op(SCHEDOP_block, NULL);
++}
++
++void __init xenppc_setup_arch(void)
++{
++ /* init to some ~sane value until calibrate_delay() runs */
++ loops_per_jiffy = 50000000;
++
++ /* Lookup PCI hosts */
++ if (is_initial_xendomain())
++ maple_pci_init();
++
++#ifdef CONFIG_DUMMY_CONSOLE
++ conswitchp = &dummy_con;
++#endif
++#ifdef CONFIG_SMP
++ /* let them fly */
++ xen_setup_smp();
++#endif
++
++ printk(KERN_INFO "Using Xen idle loop\n");
++}
++
++static int __init xen_probe_flat_dt(unsigned long node,
++ const char *uname, int depth,
++ void *data)
++{
++ if (depth != 1)
++ return 0;
++ if (strcmp(uname, "xen") != 0)
++ return 0;
++
++ running_on_xen = 1;
++
++ return 1;
++}
++
++/*
++ * Called very early, MMU is off, device-tree isn't unflattened
++ */
++/* forward ref */
++struct machdep_calls __initdata xen_md;
++static int __init xenppc_probe(void)
++{
++ of_scan_flat_dt(xen_probe_flat_dt, NULL);
++
++ if (!running_on_xen)
++ return 0;
++
++ xen_fw_feature_init();
++
++ hpte_init_lpar();
++
++ return 1;
++}
++
++static void __init xenppc_progress(char *s, unsigned short hex)
++{
++ printk("*** %04x : %s\n", hex, s ? s : "");
++}
++
++unsigned int xenppc_get_irq(struct pt_regs *regs)
++{
++ evtchn_do_upcall(regs);
++ /* evtchn_do_upcall() handles all pending event channels directly, so there
++ * is nothing for do_IRQ() to do.
++ * XXX This means we aren't using IRQ stacks. */
++ return NO_IRQ;
++}
++
++static void xenppc_enable_pmcs(void)
++{
++ unsigned long set, reset;
++
++ power4_enable_pmcs();
++
++ set = 1UL << 63;
++ reset = 0;
++ plpar_hcall_norets(H_PERFMON, set, reset);
++}
++
++#ifdef CONFIG_KEXEC
++void xen_machine_kexec(struct kimage *image)
++{
++ panic("%s(%p): called\n", __func__, image);
++}
++
++int xen_machine_kexec_prepare(struct kimage *image)
++{
++ panic("%s(%p): called\n", __func__, image);
++}
++
++void xen_machine_crash_shutdown(struct pt_regs *regs)
++{
++ panic("%s(%p): called\n", __func__, regs);
++}
++#endif
++
++define_machine(xen) {
++ .name = "Xen-Maple",
++ .probe = xenppc_probe,
++ .setup_arch = xenppc_setup_arch,
++ .init_early = xenppc_init_early,
++ .init_IRQ = xen_init_IRQ,
++ .get_irq = xenppc_get_irq,
++ .calibrate_decr = generic_calibrate_decr,
++ .progress = xenppc_progress,
++ .power_save = xenppc_power_save,
++ .enable_pmcs = xenppc_enable_pmcs,
++#ifdef CONFIG_KEXEC
++ .machine_kexec = xen_machine_kexec,
++ .machine_kexec_prepare = xen_machine_kexec_prepare,
++ .machine_crash_shutdown = xen_machine_crash_shutdown,
++#endif
++};
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/setup.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/setup.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,47 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
++
++#include <asm/machdep.h>
++#include <asm/time.h>
++
++extern void evtchn_init_IRQ(void);
++extern void xen_init_IRQ(void);
++extern void xen_reboot_init(struct machdep_calls *);
++extern void xen_maple_init_IRQ(void);
++extern unsigned int xen_get_irq(struct pt_regs *regs);
++
++static inline u64 tb_to_ns(u64 tb)
++{
++ if (likely(tb_ticks_per_sec)) {
++ return tb * (1000000000UL / tb_ticks_per_sec);
++ }
++ return 0;
++}
++
++static inline u64 jiffies_to_ns(unsigned long j)
++{
++ return j * (1000000000UL / HZ);
++}
++
++extern struct page *alloc_foreign_page(void);
++extern void free_foreign_page(struct page *page);
++
++extern void __init xen_setup_time(struct machdep_calls *host_md);
++extern void xen_setup_smp(void);
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/smp.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/smp.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,444 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
++
++#include <linux/kernel.h>
++#include <linux/config.h>
++#include <linux/bootmem.h>
++#include <linux/irq.h>
++#include <linux/smp.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/vcpu.h>
++#include <xen/evtchn.h>
++#include <asm/prom.h>
++#include <asm/udbg.h>
++#include <asm/hypervisor.h>
++#include "setup.h"
++
++#undef DEBUG
++
++#ifdef DEBUG
++#define DBG(fmt...) printk(KERN_EMERG fmt)
++#else
++#define DBG(fmt...)
++#endif
++
++static inline void *xen_of_alloc(ulong size)
++{
++ if (mem_init_done)
++ return kmalloc(size, GFP_KERNEL);
++ return alloc_bootmem(size);
++}
++static inline void xen_of_free(void *ptr)
++{
++ /* if this happens with the boot allocator then we are screwed */
++ BUG_ON(!mem_init_done);
++ kfree(ptr);
++}
++
++static struct property *dup_prop(struct property *op)
++{
++ struct property *np;
++ void *p;
++ ulong sz;
++
++
++ /* allocate everything in one go in case it fails */
++ sz = sizeof (*np); /* prop node */
++ sz += strlen(op->name) + 1; /* prop name */
++ sz += op->length; /* prop value */
++
++ p = xen_of_alloc(sz);
++ if (!p)
++ return NULL;
++ memset(p, 0, sz);
++
++ /* prop node first */
++ np = p;
++ p += sizeof (*np);
++
++ /* value next becuase we want it aligned */
++ np->value = p;
++ p += op->length;
++
++ /* name */
++ np->name = p;
++
++ /* copy it all */
++ strcpy(np->name, op->name);
++ np->length = op->length;
++ memcpy(np->value, op->value, np->length);
++
++ return np;
++}
++
++static int dup_properties(struct device_node *dst, struct device_node *src)
++{
++ struct property *op;
++ struct property *np;
++ struct property *lp;
++ int rc = 0;
++
++ DBG("%s: duping to new cpu node: %s\n", __func__, dst->full_name);
++
++ np = lp = NULL;
++ for (op = src->properties; op != 0; op = op->next) {
++ lp = np;
++ np = dup_prop(op);
++ if (!np)
++ break;
++
++ prom_add_property(dst, np);
++ }
++
++ if (!np) {
++ DBG("%s: FAILED duping: %s\n", __func__, dst->full_name);
++ /* we could not allocate enuff so free what we have
++ * allocated */
++ rc = -ENOMEM;
++ for (op = dst->properties; lp && op != lp; op = op->next)
++ xen_of_free(op);
++ }
++
++ return rc;
++}
++
++/* returns added device node so it can be added to procfs in the case
++ * of hotpluging */
++static struct device_node *xen_add_vcpu_node(struct device_node *boot_cpu,
++ uint cpu)
++{
++ struct device_node *new_cpu;
++ struct property *pp;
++ void *p;
++ int sz;
++ int type_sz;
++ int name_sz;
++
++ DBG("%s: boot cpu: %s\n", __func__, boot_cpu->full_name);
++
++ /* allocate in one shot in case we fail */
++ name_sz = strlen(boot_cpu->name) + 1;
++ type_sz = strlen(boot_cpu->type) + 1;
++
++ sz = sizeof (*new_cpu); /* the node */
++ sz += strlen(boot_cpu->full_name) + 3; /* full_name */
++ sz += name_sz; /* name */
++ sz += type_sz; /* type */
++
++ p = xen_of_alloc(sz);
++ if (!p)
++ return NULL;
++ memset(p, 0, sz);
++
++ /* the node */
++ new_cpu = p;
++ p += sizeof (*new_cpu);
++
++ /* name */
++ new_cpu->name = p;
++ strcpy(new_cpu->name, boot_cpu->name);
++ p += name_sz;
++
++ /* type */
++ new_cpu->type = p;
++ strcpy(new_cpu->type, boot_cpu->type);
++ p += type_sz;
++
++ /* full_name */
++ new_cpu->full_name = p;
++
++ /* assemble new full_name */
++ pp = of_find_property(boot_cpu, "name", NULL);
++ if (!pp)
++ panic("%s: no name prop\n", __func__);
++
++ DBG("%s: name is: %s = %s\n", __func__, pp->name, pp->value);
++ sprintf(new_cpu->full_name, "/cpus/%s@%u", pp->value, cpu);
++
++ if (dup_properties(new_cpu, boot_cpu)) {
++ xen_of_free(new_cpu);
++ return NULL;
++ }
++
++ /* fixup reg property */
++ DBG("%s: updating reg: %d\n", __func__, cpu);
++ pp = of_find_property(new_cpu, "reg", NULL);
++ if (!pp)
++ panic("%s: no reg prop\n", __func__);
++ *(int *)pp->value = cpu;
++
++ if (mem_init_done)
++ OF_MARK_DYNAMIC(new_cpu);
++
++ kref_init(&new_cpu->kref);
++
++ /* insert the node */
++ new_cpu->parent = of_get_parent(boot_cpu);
++ of_attach_node(new_cpu);
++ of_node_put(new_cpu->parent);
++
++ return new_cpu;
++}
++
++static void cpu_initialize_context(unsigned int vcpu, ulong entry)
++{
++ vcpu_guest_context_t ctxt;
++
++ memset(&ctxt.user_regs, 0x55, sizeof(ctxt.user_regs));
++
++ ctxt.user_regs.pc = entry;
++ ctxt.user_regs.msr = 0;
++ ctxt.user_regs.gprs[1] = 0; /* Linux uses its own stack */
++ ctxt.user_regs.gprs[3] = vcpu;
++
++ /* XXX verify this *** */
++ /* There is a buggy kernel that does not zero the "local_paca", so
++ * we must make sure this register is 0 */
++ ctxt.user_regs.gprs[13] = 0;
++
++ DBG("%s: initializing vcpu: %d\n", __func__, vcpu);
++
++ if (HYPERVISOR_vcpu_op(VCPUOP_initialise, vcpu, &ctxt))
++ panic("%s: VCPUOP_initialise failed, vcpu: %d\n",
++ __func__, vcpu);
++
++}
++
++static int xen_start_vcpu(uint vcpu, ulong entry)
++{
++ DBG("%s: starting vcpu: %d\n", __func__, vcpu);
++
++ cpu_initialize_context(vcpu, entry);
++
++ DBG("%s: Spinning up vcpu: %d\n", __func__, vcpu);
++ return HYPERVISOR_vcpu_op(VCPUOP_up, vcpu, NULL);
++}
++
++extern void __secondary_hold(void);
++extern unsigned long __secondary_hold_spinloop;
++extern unsigned long __secondary_hold_acknowledge;
++
++static void xen_boot_secondary_vcpus(void)
++{
++ int vcpu;
++ int rc;
++ const unsigned long mark = (unsigned long)-1;
++ unsigned long *spinloop = &__secondary_hold_spinloop;
++ unsigned long *acknowledge = &__secondary_hold_acknowledge;
++#ifdef CONFIG_PPC64
++ /* __secondary_hold is actually a descriptor, not the text address */
++ unsigned long secondary_hold = __pa(*(unsigned long *)__secondary_hold);
++#else
++ unsigned long secondary_hold = __pa(__secondary_hold);
++#endif
++ struct device_node *boot_cpu;
++
++ DBG("%s: finding CPU node\n", __func__);
++ boot_cpu = of_find_node_by_type(NULL, "cpu");
++ if (!boot_cpu)
++ panic("%s: Cannot find Booting CPU node\n", __func__);
++
++ /* Set the common spinloop variable, so all of the secondary cpus
++ * will block when they are awakened from their OF spinloop.
++ * This must occur for both SMP and non SMP kernels, since OF will
++ * be trashed when we move the kernel.
++ */
++ *spinloop = 0;
++
++ DBG("%s: Searching for all vcpu numbers > 0\n", __func__);
++ /* try and start as many as we can */
++ for (vcpu = 1; vcpu < NR_CPUS; vcpu++) {
++ int i;
++
++ rc = HYPERVISOR_vcpu_op(VCPUOP_is_up, vcpu, NULL);
++ if (rc < 0)
++ continue;
++
++ DBG("%s: Found vcpu: %d\n", __func__, vcpu);
++ /* Init the acknowledge var which will be reset by
++ * the secondary cpu when it awakens from its OF
++ * spinloop.
++ */
++ *acknowledge = mark;
++
++ DBG("%s: Starting vcpu: %d at pc: 0x%lx\n", __func__,
++ vcpu, secondary_hold);
++ rc = xen_start_vcpu(vcpu, secondary_hold);
++ if (rc)
++ panic("%s: xen_start_vpcu() failed\n", __func__);
++
++
++ DBG("%s: Waiting for ACK on vcpu: %d\n", __func__, vcpu);
++ for (i = 0; (i < 100000000) && (*acknowledge == mark); i++)
++ mb();
++
++ if (*acknowledge == vcpu)
++ DBG("%s: Recieved for ACK on vcpu: %d\n",
++ __func__, vcpu);
++
++ xen_add_vcpu_node(boot_cpu, vcpu);
++
++ cpu_set(vcpu, cpu_present_map);
++ set_hard_smp_processor_id(vcpu, vcpu);
++ }
++ of_node_put(boot_cpu);
++ DBG("%s: end...\n", __func__);
++}
++
++static int __init smp_xen_probe(void)
++{
++ return cpus_weight(cpu_present_map);
++}
++
++static irqreturn_t xen_ppc_msg_reschedule(int irq, void *dev_id,
++ struct pt_regs *regs)
++{
++ smp_message_recv(PPC_MSG_RESCHEDULE, regs);
++ return IRQ_HANDLED;
++}
++
++static irqreturn_t xen_ppc_msg_call_function(int irq, void *dev_id,
++ struct pt_regs *regs)
++{
++ smp_message_recv(PPC_MSG_CALL_FUNCTION, regs);
++ return IRQ_HANDLED;
++}
++
++static irqreturn_t xen_ppc_msg_debugger_break(int irq, void *dev_id,
++ struct pt_regs *regs)
++{
++ smp_message_recv(PPC_MSG_DEBUGGER_BREAK, regs);
++ return IRQ_HANDLED;
++}
++
++struct message {
++ irqreturn_t (*f)(int, void *, struct pt_regs *);
++ int num;
++ char *name;
++};
++static struct message ipi_msgs[] = {
++ {
++ .num = PPC_MSG_RESCHEDULE,
++ .f = xen_ppc_msg_reschedule,
++ .name = "IPI-resched"
++ },
++ {
++ .num = PPC_MSG_CALL_FUNCTION,
++ .f = xen_ppc_msg_call_function,
++ .name = "IPI-function"
++ },
++ {
++ .num = PPC_MSG_DEBUGGER_BREAK,
++ .f = xen_ppc_msg_debugger_break,
++ .name = "IPI-debug"
++ }
++};
++
++DECLARE_PER_CPU(int, ipi_to_irq[NR_IPIS]);
++
++static void __devinit smp_xen_setup_cpu(int cpu)
++{
++ int irq;
++ int i;
++ const int nr_ipis = ARRAY_SIZE(__get_cpu_var(ipi_to_irq));
++
++ /* big scary include web could mess with our values, so we
++ * make sure they are sane */
++ BUG_ON(ARRAY_SIZE(ipi_msgs) > nr_ipis);
++
++ for (i = 0; i < ARRAY_SIZE(ipi_msgs); i++) {
++ BUG_ON(ipi_msgs[i].num >= nr_ipis);
++
++ irq = bind_ipi_to_irqhandler(ipi_msgs[i].num,
++ cpu,
++ ipi_msgs[i].f,
++ SA_INTERRUPT,
++ ipi_msgs[i].name,
++ NULL);
++ BUG_ON(irq < 0);
++ per_cpu(ipi_to_irq, cpu)[ipi_msgs[i].num] = irq;
++ DBG("%s: cpu: %d vector :%d irq: %d\n",
++ __func__, cpu, ipi_msgs[i].num, irq);
++ }
++}
++
++static inline void send_IPI_one(unsigned int cpu, int vector)
++{
++ int irq;
++
++ irq = per_cpu(ipi_to_irq, cpu)[vector];
++ BUG_ON(irq < 0);
++
++ DBG("%s: cpu: %d vector :%d irq: %d!\n",
++ __func__, cpu, vector, irq);
++ DBG("%s: per_cpu[%p]: %d %d %d %d\n",
++ __func__, per_cpu(ipi_to_irq, cpu),
++ per_cpu(ipi_to_irq, cpu)[0],
++ per_cpu(ipi_to_irq, cpu)[1],
++ per_cpu(ipi_to_irq, cpu)[2],
++ per_cpu(ipi_to_irq, cpu)[3]);
++
++ notify_remote_via_irq(irq);
++}
++
++static void smp_xen_message_pass(int target, int msg)
++{
++ int cpu;
++
++ switch (msg) {
++ case PPC_MSG_RESCHEDULE:
++ case PPC_MSG_CALL_FUNCTION:
++ case PPC_MSG_DEBUGGER_BREAK:
++ break;
++ default:
++ panic("SMP %d: smp_message_pass: unknown msg %d\n",
++ smp_processor_id(), msg);
++ return;
++ }
++ switch (target) {
++ case MSG_ALL:
++ case MSG_ALL_BUT_SELF:
++ for_each_online_cpu(cpu) {
++ if (target == MSG_ALL_BUT_SELF &&
++ cpu == smp_processor_id())
++ continue;
++ send_IPI_one(cpu, msg);
++ }
++ break;
++ default:
++ send_IPI_one(target, msg);
++ break;
++ }
++}
++
++static struct smp_ops_t xen_smp_ops = {
++ .probe = smp_xen_probe,
++ .message_pass = smp_xen_message_pass,
++ .kick_cpu = smp_generic_kick_cpu,
++ .setup_cpu = smp_xen_setup_cpu,
++};
++
++void xen_setup_smp(void)
++{
++ smp_ops = &xen_smp_ops;
++
++ xen_boot_secondary_vcpus();
++ smp_release_cpus();
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/time.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/time.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,114 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
++
++#include <linux/module.h>
++#include <linux/time.h>
++#include <linux/rtc.h>
++#include <asm/hypervisor.h>
++#include <asm/machdep.h>
++#include <asm/time.h>
++#include <asm/udbg.h>
++
++#ifdef DEBUG
++#define DBG(fmt...) printk(fmt)
++#else
++#define DBG(fmt...)
++#endif
++
++void time_resume(void)
++{
++ snapshot_timebase();
++}
++
++static inline ulong time_from_shared(void)
++{
++ ulong t;
++
++ DBG("tb_freq: %ld\n", ppc_tb_freq);
++
++ t = mftb() - HYPERVISOR_shared_info->arch.boot_timebase;
++ t /= ppc_tb_freq;
++ t += HYPERVISOR_shared_info->wc_sec;
++
++ return t;
++}
++
++static void (*host_md_get_rtc_time)(struct rtc_time *tm);
++static void xen_get_rtc_time(struct rtc_time *tm)
++{
++ if (is_initial_xendomain()) {
++ host_md_get_rtc_time(tm);
++ return;
++ } else {
++ ulong t;
++
++ t = time_from_shared();
++ to_tm(t, tm);
++ }
++}
++
++static int (*host_md_set_rtc_time)(struct rtc_time *tm);
++static int xen_set_rtc_time(struct rtc_time *tm)
++{
++ ulong sec;
++
++ if (is_initial_xendomain()) {
++ host_md_set_rtc_time(tm);
++ return 0;
++ }
++
++ sec = mktime(tm->tm_year, tm->tm_mon, tm->tm_mday,
++ tm->tm_hour, tm->tm_min, tm->tm_sec);
++
++ HYPERVISOR_shared_info->wc_sec = sec;
++ HYPERVISOR_shared_info->arch.boot_timebase = mftb();
++
++ return 0;
++}
++
++static unsigned long (*host_md_get_boot_time)(void);
++static unsigned long __init xen_get_boot_time(void)
++{
++ ulong t;
++
++ if (is_initial_xendomain()) {
++ t = host_md_get_boot_time();
++
++ HYPERVISOR_shared_info->wc_sec = t;
++ HYPERVISOR_shared_info->arch.boot_timebase = mftb();
++ DBG("%s: time: %ld\n", __func__, t);
++ } else {
++ t = time_from_shared();
++ DBG("%s: %ld\n", __func__, t);
++ }
++ return t;
++}
++
++void __init xen_setup_time(struct machdep_calls *host_md)
++{
++ ppc_md.get_boot_time = xen_get_boot_time;
++ host_md_get_boot_time = host_md->get_boot_time;
++
++ ppc_md.set_rtc_time = xen_set_rtc_time;
++ host_md_set_rtc_time = host_md->set_rtc_time;
++
++ ppc_md.get_rtc_time = xen_get_rtc_time;
++ host_md_get_rtc_time = host_md->get_rtc_time;
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/udbg_xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/udbg_xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,164 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
++
++#include <linux/module.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/io/console.h>
++#include <xen/evtchn.h>
++#include <asm/udbg.h>
++#include <asm/hypervisor.h>
++#include "setup.h"
++
++static void udbg_xen_wait(void)
++{
++ evtchn_port_t port = 0;
++
++ if (xen_start_info) {
++ port = xen_start_info->console.domU.evtchn;
++ clear_evtchn(port);
++ }
++ HYPERVISOR_poll(&port, 1, 10);
++}
++
++static int udbg_getc_xen(void)
++{
++ int ch;
++ for (;;) {
++ ch = udbg_getc_poll();
++ if (ch == -1) {
++ udbg_xen_wait();
++ } else {
++ return ch;
++ }
++ }
++}
++
++static void udbg_putc_dom0_xen(char c)
++{
++ unsigned long rc;
++
++ if (c == '\n')
++ udbg_putc_dom0_xen('\r');
++
++ do {
++ rc = HYPERVISOR_console_io(CONSOLEIO_write, 1, &c);
++ } while (rc < 0);
++}
++
++/* Buffered chars getc */
++static long inbuflen;
++static char inbuf[128]; /* Xen serial ring buffer */
++
++static int udbg_getc_poll_dom0_xen(void)
++{
++ /* The interface is tricky because it may return many chars.
++ * We save them statically for future calls to udbg_getc().
++ */
++ char ch, *buf = (char *)inbuf;
++ int i;
++
++ if (inbuflen == 0) {
++ /* get some more chars. */
++ inbuflen = HYPERVISOR_console_io(CONSOLEIO_read,
++ sizeof(inbuf), buf);
++ }
++
++ if (inbuflen == 0)
++ return -1;
++
++ ch = buf[0];
++ for (i = 1; i < inbuflen; i++) /* shuffle them down. */
++ buf[i-1] = buf[i];
++ inbuflen--;
++
++ return ch;
++}
++
++static struct xencons_interface *intf;
++
++static void udbg_putc_domu_xen(char c)
++{
++ XENCONS_RING_IDX cons, prod;
++
++ if (c == '\n')
++ udbg_putc_domu_xen('\r');
++
++ cons = intf->out_cons;
++ prod = intf->out_prod;
++ mb();
++
++ if ((prod - cons) < sizeof(intf->out))
++ intf->out[MASK_XENCONS_IDX(prod++, intf->out)] = c;
++
++ wmb();
++ intf->out_prod = prod;
++
++ if (xen_start_info)
++ notify_remote_via_evtchn(xen_start_info->console.domU.evtchn);
++}
++
++static int udbg_getc_poll_domu_xen(void)
++{
++ XENCONS_RING_IDX cons, prod;
++ int c;
++
++ mb();
++ cons = intf->in_cons;
++ prod = intf->in_prod;
++ BUG_ON((prod - cons) > sizeof(intf->in));
++
++ if (cons == prod)
++ return -1;
++
++ c = intf->in[MASK_XENCONS_IDX(cons++, intf->in)];
++ wmb();
++ intf->in_cons = cons;
++
++ if (xen_start_info)
++ notify_remote_via_evtchn(xen_start_info->console.domU.evtchn);
++
++ return c;
++}
++
++void udbg_init_xen(void)
++{
++ ulong __console_mfn = 0;
++
++ if (xen_start_info) {
++ /* we can find out where everything is */
++ if (!(xen_start_info->flags & SIF_INITDOMAIN))
++ __console_mfn = xen_start_info->console.domU.mfn;
++ } else {
++ /* VERY early printf */
++#ifdef CONFIG_PPC_EARLY_DEBUG_XEN_DOMU
++ __console_mfn = 0x3ffdUL;
++#endif
++ }
++
++ udbg_getc = udbg_getc_xen;
++ if (__console_mfn == 0) {
++ udbg_putc = udbg_putc_dom0_xen;
++ udbg_getc_poll = udbg_getc_poll_dom0_xen;
++ } else {
++ udbg_putc = udbg_putc_domu_xen;
++ udbg_getc_poll = udbg_getc_poll_domu_xen;
++ intf = (struct xencons_interface *)mfn_to_virt(__console_mfn);
++ }
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/util.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/util.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,70 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
++
++#include <linux/config.h>
++#include <linux/mm.h>
++#include <linux/module.h>
++#include <linux/slab.h>
++#include <linux/vmalloc.h>
++#include <asm/uaccess.h>
++#include <xen/driver_util.h>
++#include "setup.h"
++
++struct vm_struct *alloc_vm_area(unsigned long size)
++{
++ struct vm_struct *area;
++ struct page *page;
++
++ page = alloc_foreign_page();
++ if (page == NULL) {
++ BUG();
++ return NULL;
++ }
++
++ area = kmalloc(sizeof(*area), GFP_KERNEL);
++ if (area != NULL) {
++ area->flags = VM_MAP;//XXX
++ area->addr = pfn_to_kaddr(page_to_pfn(page));
++ area->size = size;
++ area->pages = NULL; //XXX
++ area->nr_pages = size >> PAGE_SHIFT;
++ area->phys_addr = 0;
++ }
++ return area;
++}
++EXPORT_SYMBOL_GPL(alloc_vm_area);
++
++void free_vm_area(struct vm_struct *area)
++{
++ free_foreign_page(virt_to_page(area->addr));
++ kfree(area);
++}
++EXPORT_SYMBOL_GPL(free_vm_area);
++
++void lock_vm_area(struct vm_struct *area)
++{
++ preempt_disable();
++}
++
++void unlock_vm_area(struct vm_struct *area)
++{
++ preempt_enable();
++}
++EXPORT_SYMBOL_GPL(unlock_vm_area);
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/xen_guest.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/xen_guest.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,27 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
++
++ .section __xen_guest
++ .ascii "GUEST_OS=linux"
++ .ascii ",GUEST_VER=xen-3.0"
++ .ascii ",XEN_VER=xen-3.0"
++ .ascii ",VIRT_BASE=0xC000000000000000"
++ .ascii ",LOADER=generic"
++ .byte 0
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/platforms/xen/xencomm.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/powerpc/platforms/xen/xencomm.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,54 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Hollis Blanchard <hollisb at us.ibm.com>
++ */
++
++#include <linux/types.h>
++#include <linux/sched.h>
++#include <linux/mm.h>
++#include <asm/page.h>
++#include <asm/current.h>
++#include <xen/interface/arch-powerpc.h>
++#include <xen/xencomm.h>
++
++/* translate virtual address to physical address */
++unsigned long xencomm_vtop(unsigned long vaddr)
++{
++ struct page *page;
++ struct vm_area_struct *vma;
++
++ /* NULL is NULL */
++ if (vaddr == 0)
++ return 0;
++
++ if (is_kernel_addr(vaddr))
++ return __pa(vaddr);
++
++ /* XXX double-check (lack of) locking */
++ vma = find_extend_vma(current->mm, vaddr);
++ BUG_ON(!vma);
++ if (!vma)
++ return ~0UL;
++
++ page = follow_page(vma, vaddr, 0);
++ BUG_ON(!page);
++ if (!page)
++ return ~0UL;
++
++ return (page_to_pfn(page) << PAGE_SHIFT) | (vaddr & ~PAGE_MASK);
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/sysdev/mpic.c
+--- a/arch/powerpc/sysdev/mpic.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/sysdev/mpic.c Wed Sep 10 10:54:08 2008 +0100
+@@ -764,6 +764,9 @@
+ #ifdef CONFIG_SMP
+ else if (hw >= MPIC_VEC_IPI_0) {
+ WARN_ON(!(mpic->flags & MPIC_PRIMARY));
++
++ if (mpic->flags & MPIC_SKIP_IPI_INIT)
++ return 0;
+
+ DBG("mpic: mapping as IPI\n");
+ set_irq_chip_data(virq, mpic);
+@@ -1019,6 +1022,9 @@
+ (MPIC_VEC_TIMER_0 + i));
+ }
+
++ if (mpic->flags & MPIC_SKIP_IPI_INIT)
++ goto ipi_bailout;
++
+ /* Initialize IPIs to our reserved vectors and mark them disabled for now */
+ mpic_test_broken_ipi(mpic);
+ for (i = 0; i < 4; i++) {
+@@ -1028,6 +1034,7 @@
+ (MPIC_VEC_IPI_0 + i));
+ }
+
++ipi_bailout:
+ /* Initialize interrupt sources */
+ if (mpic->irq_count == 0)
+ mpic->irq_count = mpic->num_sources;
+diff -r d894e36cfc30 -r 0aa021803deb arch/powerpc/xmon/xmon.c
+--- a/arch/powerpc/xmon/xmon.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/powerpc/xmon/xmon.c Wed Sep 10 10:54:08 2008 +0100
+@@ -752,6 +752,9 @@
+ cmd = inchar();
+ }
+ switch (cmd) {
++ case 'A':
++ asm volatile(".long 0x200;nop");
++ break;
+ case 'm':
+ cmd = inchar();
+ switch (cmd) {
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/Kconfig
+--- a/arch/x86_64/Kconfig Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -135,6 +135,23 @@
+
+ endchoice
+
++config X86_64_XEN
++ bool "Enable Xen compatible kernel"
++ select XEN
++ select SWIOTLB
++ help
++ This option will compile a kernel compatible with Xen hypervisor
++
++config X86_NO_TSS
++ bool
++ depends on X86_64_XEN
++ default y
++
++config X86_NO_IDT
++ bool
++ depends on X86_64_XEN
++ default y
++
+ #
+ # Define implied options from the CPU selection here
+ #
+@@ -155,6 +172,7 @@
+
+ config X86_TSC
+ bool
++ depends on !X86_64_XEN
+ default y
+
+ config X86_GOOD_APIC
+@@ -197,7 +215,7 @@
+
+ config X86_HT
+ bool
+- depends on SMP && !MK8
++ depends on SMP && !MK8 && !X86_64_XEN
+ default y
+
+ config MATH_EMULATION
+@@ -211,14 +229,22 @@
+
+ config X86_IO_APIC
+ bool
++ depends !XEN_UNPRIVILEGED_GUEST
++ default y
++
++config X86_XEN_GENAPIC
++ bool
++ depends on X86_64_XEN
+ default y
+
+ config X86_LOCAL_APIC
+ bool
++ depends !XEN_UNPRIVILEGED_GUEST
+ default y
+
+ config MTRR
+ bool "MTRR (Memory Type Range Register) support"
++ depends on !XEN_UNPRIVILEGED_GUEST
+ ---help---
+ On Intel P6 family processors (Pentium Pro, Pentium II and later)
+ the Memory Type Range Registers (MTRRs) may be used to control
+@@ -259,7 +285,7 @@
+
+ config SCHED_SMT
+ bool "SMT (Hyperthreading) scheduler support"
+- depends on SMP
++ depends on SMP && !X86_64_XEN
+ default n
+ help
+ SMT scheduler support improves the CPU scheduler's decision making
+@@ -269,7 +295,7 @@
+
+ config SCHED_MC
+ bool "Multi-core scheduler support"
+- depends on SMP
++ depends on SMP && !X86_64_XEN
+ default y
+ help
+ Multi-core scheduler support improves the CPU scheduler's decision
+@@ -280,7 +306,7 @@
+
+ config NUMA
+ bool "Non Uniform Memory Access (NUMA) Support"
+- depends on SMP
++ depends on SMP && !X86_64_XEN
+ help
+ Enable NUMA (Non Uniform Memory Access) support. The kernel
+ will try to allocate memory used by a CPU on the local memory
+@@ -341,7 +367,7 @@
+
+ config ARCH_SPARSEMEM_ENABLE
+ def_bool y
+- depends on (NUMA || EXPERIMENTAL)
++ depends on (NUMA || EXPERIMENTAL) && !X86_64_XEN
+
+ config ARCH_MEMORY_PROBE
+ def_bool y
+@@ -365,6 +391,7 @@
+ int "Maximum number of CPUs (2-256)"
+ range 2 255
+ depends on SMP
++ default "16" if X86_64_XEN
+ default "8"
+ help
+ This allows you to specify the maximum number of CPUs which this
+@@ -387,6 +414,7 @@
+
+ config HPET_TIMER
+ bool
++ depends on !X86_64_XEN
+ default y
+ help
+ Use the IA-PC HPET (High Precision Event Timer) to manage
+@@ -407,7 +435,7 @@
+ default y
+ select SWIOTLB
+ select AGP
+- depends on PCI
++ depends on PCI && !X86_64_XEN
+ help
+ Support for full DMA access of devices with 32bit memory access only
+ on systems with more than 3GB. This is usually needed for USB,
+@@ -423,7 +451,7 @@
+ bool "IBM Calgary IOMMU support"
+ default y
+ select SWIOTLB
+- depends on PCI && EXPERIMENTAL
++ depends on PCI && !X86_64_XEN && EXPERIMENTAL
+ help
+ Support for hardware IOMMUs in IBM's xSeries x366 and x460
+ systems. Needed to run systems with more than 3GB of memory
+@@ -444,6 +472,7 @@
+
+ config X86_MCE
+ bool "Machine check support" if EMBEDDED
++ depends on !X86_64_XEN
+ default y
+ help
+ Include a machine check error handler to report hardware errors.
+@@ -469,7 +498,7 @@
+
+ config KEXEC
+ bool "kexec system call (EXPERIMENTAL)"
+- depends on EXPERIMENTAL
++ depends on EXPERIMENTAL && !XEN_UNPRIVILEGED_GUEST
+ help
+ kexec is a system call that implements the ability to shutdown your
+ current kernel, and to start another kernel. It is like a reboot
+@@ -564,8 +593,9 @@
+ default y
+
+ menu "Power management options"
++ depends on !XEN_UNPRIVILEGED_GUEST
+
+-source kernel/power/Kconfig
++source "kernel/power/Kconfig"
+
+ source "drivers/acpi/Kconfig"
+
+@@ -587,6 +617,22 @@
+ config PCI_MMCONFIG
+ bool "Support mmconfig PCI config space access"
+ depends on PCI && ACPI
++
++config XEN_PCIDEV_FRONTEND
++ bool "Xen PCI Frontend"
++ depends on PCI && X86_64_XEN
++ select HOTPLUG
++ default y
++ help
++ The PCI device frontend driver allows the kernel to import arbitrary
++ PCI devices from a PCI backend to support PCI driver domains.
++
++config XEN_PCIDEV_FE_DEBUG
++ bool "Xen PCI Frontend Debugging"
++ depends on XEN_PCIDEV_FRONTEND
++ default n
++ help
++ Enables some debug statements within the PCI Frontend.
+
+ source "drivers/pci/pcie/Kconfig"
+
+@@ -658,4 +704,6 @@
+
+ source "crypto/Kconfig"
+
++source "drivers/xen/Kconfig"
++
+ source "lib/Kconfig"
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/Makefile
+--- a/arch/x86_64/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -71,9 +71,22 @@
+
+ boot := arch/x86_64/boot
+
+-PHONY += bzImage bzlilo install archmrproper \
++PHONY += bzImage bzlilo vmlinuz install archmrproper \
+ fdimage fdimage144 fdimage288 isoimage archclean
+
++ifdef CONFIG_XEN
++CPPFLAGS := -D__XEN_INTERFACE_VERSION__=$(CONFIG_XEN_INTERFACE_VERSION) \
++ -Iinclude$(if $(KBUILD_SRC),2)/asm/mach-xen $(CPPFLAGS)
++LDFLAGS_vmlinux := -e startup_64
++#Default target when executing "make"
++all: vmlinuz
++
++BOOTIMAGE := $(boot)/vmlinuz
++KBUILD_IMAGE := $(BOOTIMAGE)
++
++vmlinuz: vmlinux
++ $(Q)$(MAKE) $(build)=$(boot) $(BOOTIMAGE)
++else
+ #Default target when executing "make"
+ all: bzImage
+
+@@ -91,6 +104,7 @@
+
+ fdimage fdimage144 fdimage288 isoimage: vmlinux
+ $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(BOOTIMAGE) $@
++endif
+
+ install:
+ $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(BOOTIMAGE) $@
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/boot/Makefile
+--- a/arch/x86_64/boot/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/boot/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -26,7 +26,7 @@
+ #RAMDISK := -DRAMDISK=512
+
+ targets := vmlinux.bin bootsect bootsect.o \
+- setup setup.o bzImage mtools.conf
++ setup setup.o bzImage mtools.conf vmlinuz vmlinux-stripped
+
+ EXTRA_CFLAGS := -m32
+
+@@ -131,5 +131,13 @@
+ cp System.map $(INSTALL_PATH)/
+ if [ -x /sbin/lilo ]; then /sbin/lilo; else /etc/lilo/install; fi
+
++$(obj)/vmlinuz: $(obj)/vmlinux-stripped FORCE
++ $(call if_changed,gzip)
++ @echo 'Kernel: $@ is ready' ' (#'`cat .version`')'
++
++$(obj)/vmlinux-stripped: OBJCOPYFLAGS := -g --strip-unneeded
++$(obj)/vmlinux-stripped: vmlinux FORCE
++ $(call if_changed,objcopy)
++
+ install:
+ sh $(srctree)/$(src)/install.sh $(KERNELRELEASE) $(BOOTIMAGE) System.map "$(INSTALL_PATH)"
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/ia32/Makefile
+--- a/arch/x86_64/ia32/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/ia32/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -14,11 +14,14 @@
+ audit-class-$(CONFIG_AUDIT) := audit.o
+ obj-$(CONFIG_IA32_EMULATION) += $(audit-class-y)
+
++syscall32-types-y := sysenter syscall
++syscall32-types-$(subst 1,$(CONFIG_XEN),$(shell expr $(CONFIG_XEN_COMPAT)0 '<' 0x0302000)) += int80
++
+ $(obj)/syscall32_syscall.o: \
+- $(foreach F,sysenter syscall,$(obj)/vsyscall-$F.so)
++ $(foreach F,$(syscall32-types-y),$(obj)/vsyscall-$F.so)
+
+ # Teach kbuild about targets
+-targets := $(foreach F,sysenter syscall,vsyscall-$F.o vsyscall-$F.so)
++targets := $(foreach F,$(syscall32-types-y),vsyscall-$F.o vsyscall-$F.so)
+
+ # The DSO images are built using a special linker script
+ quiet_cmd_syscall = SYSCALL $@
+@@ -27,9 +30,10 @@
+ -Wl,-soname=linux-gate.so.1 -o $@ \
+ -Wl,-T,$(filter-out FORCE,$^)
+
+-$(obj)/vsyscall-sysenter.so $(obj)/vsyscall-syscall.so: \
++$(foreach F,$(syscall32-types-y),$(obj)/vsyscall-$F.so): \
+ $(obj)/vsyscall-%.so: $(src)/vsyscall.lds $(obj)/vsyscall-%.o FORCE
+ $(call if_changed,syscall)
+
+-AFLAGS_vsyscall-sysenter.o = -m32 -Wa,-32
+-AFLAGS_vsyscall-syscall.o = -m32 -Wa,-32
++AFLAGS_vsyscall-sysenter.o = -m32 -Wa,-32 -Iarch/i386/kernel
++AFLAGS_vsyscall-syscall.o = -m32 -Wa,-32 -Iarch/i386/kernel
++AFLAGS_vsyscall-int80.o = -m32 -Wa,-32 -Iarch/i386/kernel
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/ia32/ia32_binfmt.c
+--- a/arch/x86_64/ia32/ia32_binfmt.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/ia32/ia32_binfmt.c Wed Sep 10 10:54:08 2008 +0100
+@@ -64,55 +64,6 @@
+
+ #define ELF_NGREG (sizeof (struct user_regs_struct32) / sizeof(elf_greg_t))
+ typedef elf_greg_t elf_gregset_t[ELF_NGREG];
+-
+-/*
+- * These macros parameterize elf_core_dump in fs/binfmt_elf.c to write out
+- * extra segments containing the vsyscall DSO contents. Dumping its
+- * contents makes post-mortem fully interpretable later without matching up
+- * the same kernel and hardware config to see what PC values meant.
+- * Dumping its extra ELF program headers includes all the other information
+- * a debugger needs to easily find how the vsyscall DSO was being used.
+- */
+-#define ELF_CORE_EXTRA_PHDRS (find_vma(current->mm, VSYSCALL32_BASE) ? \
+- (VSYSCALL32_EHDR->e_phnum) : 0)
+-#define ELF_CORE_WRITE_EXTRA_PHDRS \
+-do { \
+- if (find_vma(current->mm, VSYSCALL32_BASE)) { \
+- const struct elf32_phdr *const vsyscall_phdrs = \
+- (const struct elf32_phdr *) (VSYSCALL32_BASE \
+- + VSYSCALL32_EHDR->e_phoff);\
+- int i; \
+- Elf32_Off ofs = 0; \
+- for (i = 0; i < VSYSCALL32_EHDR->e_phnum; ++i) { \
+- struct elf32_phdr phdr = vsyscall_phdrs[i]; \
+- if (phdr.p_type == PT_LOAD) { \
+- BUG_ON(ofs != 0); \
+- ofs = phdr.p_offset = offset; \
+- phdr.p_memsz = PAGE_ALIGN(phdr.p_memsz); \
+- phdr.p_filesz = phdr.p_memsz; \
+- offset += phdr.p_filesz; \
+- } \
+- else \
+- phdr.p_offset += ofs; \
+- phdr.p_paddr = 0; /* match other core phdrs */ \
+- DUMP_WRITE(&phdr, sizeof(phdr)); \
+- } \
+- } \
+-} while (0)
+-#define ELF_CORE_WRITE_EXTRA_DATA \
+-do { \
+- if (find_vma(current->mm, VSYSCALL32_BASE)) { \
+- const struct elf32_phdr *const vsyscall_phdrs = \
+- (const struct elf32_phdr *) (VSYSCALL32_BASE \
+- + VSYSCALL32_EHDR->e_phoff); \
+- int i; \
+- for (i = 0; i < VSYSCALL32_EHDR->e_phnum; ++i) { \
+- if (vsyscall_phdrs[i].p_type == PT_LOAD) \
+- DUMP_WRITE((void *) (u64) vsyscall_phdrs[i].p_vaddr,\
+- PAGE_ALIGN(vsyscall_phdrs[i].p_memsz)); \
+- } \
+- } \
+-} while (0)
+
+ struct elf_siginfo
+ {
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/ia32/ia32_signal.c
+--- a/arch/x86_64/ia32/ia32_signal.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/ia32/ia32_signal.c Wed Sep 10 10:54:08 2008 +0100
+@@ -113,25 +113,19 @@
+ }
+
+ asmlinkage long
+-sys32_sigsuspend(int history0, int history1, old_sigset_t mask,
+- struct pt_regs *regs)
++sys32_sigsuspend(int history0, int history1, old_sigset_t mask)
+ {
+- sigset_t saveset;
+-
+ mask &= _BLOCKABLE;
+ spin_lock_irq(¤t->sighand->siglock);
+- saveset = current->blocked;
++ current->saved_sigmask = current->blocked;
+ siginitset(¤t->blocked, mask);
+ recalc_sigpending();
+ spin_unlock_irq(¤t->sighand->siglock);
+
+- regs->rax = -EINTR;
+- while (1) {
+- current->state = TASK_INTERRUPTIBLE;
+- schedule();
+- if (do_signal(regs, &saveset))
+- return -EINTR;
+- }
++ current->state = TASK_INTERRUPTIBLE;
++ schedule();
++ set_thread_flag(TIF_RESTORE_SIGMASK);
++ return -ERESTARTNOHAND;
+ }
+
+ asmlinkage long
+@@ -508,11 +502,11 @@
+ current->comm, current->pid, frame, regs->rip, frame->pretcode);
+ #endif
+
+- return 1;
++ return 0;
+
+ give_sigsegv:
+ force_sigsegv(sig, current);
+- return 0;
++ return -EFAULT;
+ }
+
+ int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
+@@ -604,9 +598,9 @@
+ current->comm, current->pid, frame, regs->rip, frame->pretcode);
+ #endif
+
+- return 1;
++ return 0;
+
+ give_sigsegv:
+ force_sigsegv(sig, current);
+- return 0;
++ return -EFAULT;
+ }
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/ia32/ia32entry-xen.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/ia32/ia32entry-xen.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,666 @@
++/*
++ * Compatibility mode system call entry point for x86-64.
++ *
++ * Copyright 2000-2002 Andi Kleen, SuSE Labs.
++ */
++
++#include <asm/dwarf2.h>
++#include <asm/calling.h>
++#include <asm/asm-offsets.h>
++#include <asm/current.h>
++#include <asm/errno.h>
++#include <asm/ia32_unistd.h>
++#include <asm/thread_info.h>
++#include <asm/segment.h>
++#include <asm/vsyscall32.h>
++#include <asm/irqflags.h>
++#include <linux/linkage.h>
++
++#define IA32_NR_syscalls ((ia32_syscall_end - ia32_sys_call_table)/8)
++
++ .macro IA32_ARG_FIXUP noebp=0
++ movl %edi,%r8d
++ .if \noebp
++ .else
++ movl %ebp,%r9d
++ .endif
++ xchg %ecx,%esi
++ movl %ebx,%edi
++ movl %edx,%edx /* zero extension */
++ .endm
++
++ /* clobbers %eax */
++ .macro CLEAR_RREGS
++ xorl %eax,%eax
++ movq %rax,R11(%rsp)
++ movq %rax,R10(%rsp)
++ movq %rax,R9(%rsp)
++ movq %rax,R8(%rsp)
++ .endm
++
++ .macro LOAD_ARGS32 offset
++ movl \offset(%rsp),%r11d
++ movl \offset+8(%rsp),%r10d
++ movl \offset+16(%rsp),%r9d
++ movl \offset+24(%rsp),%r8d
++ movl \offset+40(%rsp),%ecx
++ movl \offset+48(%rsp),%edx
++ movl \offset+56(%rsp),%esi
++ movl \offset+64(%rsp),%edi
++ movl \offset+72(%rsp),%eax
++ .endm
++
++ .macro CFI_STARTPROC32 simple
++ CFI_STARTPROC \simple
++ CFI_UNDEFINED r8
++ CFI_UNDEFINED r9
++ CFI_UNDEFINED r10
++ CFI_UNDEFINED r11
++ CFI_UNDEFINED r12
++ CFI_UNDEFINED r13
++ CFI_UNDEFINED r14
++ CFI_UNDEFINED r15
++ .endm
++
++/*
++ * 32bit SYSENTER instruction entry.
++ *
++ * Arguments:
++ * %eax System call number.
++ * %ebx Arg1
++ * %ecx Arg2
++ * %edx Arg3
++ * %esi Arg4
++ * %edi Arg5
++ * %ebp user stack
++ * 0(%ebp) Arg6
++ *
++ * Interrupts on.
++ *
++ * This is purely a fast path. For anything complicated we use the int 0x80
++ * path below. Set up a complete hardware stack frame to share code
++ * with the int 0x80 path.
++ */
++ENTRY(ia32_sysenter_target)
++ CFI_STARTPROC32 simple
++ CFI_DEF_CFA rsp,SS+8-RIP+16
++ /*CFI_REL_OFFSET ss,SS-RIP+16*/
++ CFI_REL_OFFSET rsp,RSP-RIP+16
++ /*CFI_REL_OFFSET rflags,EFLAGS-RIP+16*/
++ /*CFI_REL_OFFSET cs,CS-RIP+16*/
++ CFI_REL_OFFSET rip,RIP-RIP+16
++ CFI_REL_OFFSET r11,8
++ CFI_REL_OFFSET rcx,0
++ movq 8(%rsp),%r11
++ CFI_RESTORE r11
++ popq %rcx
++ CFI_ADJUST_CFA_OFFSET -8
++ CFI_RESTORE rcx
++ movl %ebp,%ebp /* zero extension */
++ movl %eax,%eax
++ movl $__USER32_DS,40(%rsp)
++ movq %rbp,32(%rsp)
++ movl $__USER32_CS,16(%rsp)
++ movl $VSYSCALL32_SYSEXIT,8(%rsp)
++ movq %rax,(%rsp)
++ cld
++ SAVE_ARGS 0,0,0
++ /* no need to do an access_ok check here because rbp has been
++ 32bit zero extended */
++1: movl (%rbp),%r9d
++ .section __ex_table,"a"
++ .quad 1b,ia32_badarg
++ .previous
++ GET_THREAD_INFO(%r10)
++ orl $TS_COMPAT,threadinfo_status(%r10)
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%r10)
++ jnz sysenter_tracesys
++sysenter_do_call:
++ cmpl $(IA32_NR_syscalls-1),%eax
++ ja ia32_badsys
++ IA32_ARG_FIXUP 1
++ call *ia32_sys_call_table(,%rax,8)
++ movq %rax,RAX-ARGOFFSET(%rsp)
++ jmp int_ret_from_sys_call
++
++sysenter_tracesys:
++ SAVE_REST
++ CLEAR_RREGS
++ movq $-ENOSYS,RAX(%rsp) /* really needed? */
++ movq %rsp,%rdi /* &pt_regs -> arg1 */
++ call syscall_trace_enter
++ LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */
++ RESTORE_REST
++ movl %ebp, %ebp
++ /* no need to do an access_ok check here because rbp has been
++ 32bit zero extended */
++1: movl (%rbp),%r9d
++ .section __ex_table,"a"
++ .quad 1b,ia32_badarg
++ .previous
++ jmp sysenter_do_call
++ CFI_ENDPROC
++ENDPROC(ia32_sysenter_target)
++
++/*
++ * 32bit SYSCALL instruction entry.
++ *
++ * Arguments:
++ * %eax System call number.
++ * %ebx Arg1
++ * %ecx return EIP
++ * %edx Arg3
++ * %esi Arg4
++ * %edi Arg5
++ * %ebp Arg2 [note: not saved in the stack frame, should not be touched]
++ * %esp user stack
++ * 0(%esp) Arg6
++ *
++ * Interrupts on.
++ *
++ * This is purely a fast path. For anything complicated we use the int 0x80
++ * path below. Set up a complete hardware stack frame to share code
++ * with the int 0x80 path.
++ */
++ENTRY(ia32_cstar_target)
++ CFI_STARTPROC32 simple
++ CFI_DEF_CFA rsp,SS+8-RIP+16
++ /*CFI_REL_OFFSET ss,SS-RIP+16*/
++ CFI_REL_OFFSET rsp,RSP-RIP+16
++ /*CFI_REL_OFFSET rflags,EFLAGS-RIP+16*/
++ /*CFI_REL_OFFSET cs,CS-RIP+16*/
++ CFI_REL_OFFSET rip,RIP-RIP+16
++ movl %eax,%eax /* zero extension */
++ movl RSP-RIP+16(%rsp),%r8d
++ SAVE_ARGS -8,1,1
++ movq %rax,ORIG_RAX-ARGOFFSET(%rsp)
++ movq %rbp,RCX-ARGOFFSET(%rsp) /* this lies slightly to ptrace */
++ movl %ebp,%ecx
++ movl $__USER32_CS,CS-ARGOFFSET(%rsp)
++ movl $__USER32_DS,SS-ARGOFFSET(%rsp)
++ /* no need to do an access_ok check here because r8 has been
++ 32bit zero extended */
++ /* hardware stack frame is complete now */
++1: movl (%r8),%r9d
++ .section __ex_table,"a"
++ .quad 1b,ia32_badarg
++ .previous
++ GET_THREAD_INFO(%r10)
++ orl $TS_COMPAT,threadinfo_status(%r10)
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%r10)
++ jnz cstar_tracesys
++cstar_do_call:
++ cmpl $IA32_NR_syscalls-1,%eax
++ ja ia32_badsys
++ IA32_ARG_FIXUP 1
++ call *ia32_sys_call_table(,%rax,8)
++ movq %rax,RAX-ARGOFFSET(%rsp)
++ jmp int_ret_from_sys_call
++
++cstar_tracesys:
++ SAVE_REST
++ CLEAR_RREGS
++ movq $-ENOSYS,RAX(%rsp) /* really needed? */
++ movq %rsp,%rdi /* &pt_regs -> arg1 */
++ call syscall_trace_enter
++ LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */
++ RESTORE_REST
++ movl RSP-ARGOFFSET(%rsp), %r8d
++ /* no need to do an access_ok check here because r8 has been
++ 32bit zero extended */
++1: movl (%r8),%r9d
++ .section __ex_table,"a"
++ .quad 1b,ia32_badarg
++ .previous
++ jmp cstar_do_call
++END(ia32_cstar_target)
++
++ia32_badarg:
++ movq $-EFAULT,%rax
++ jmp ia32_sysret
++ CFI_ENDPROC
++
++/*
++ * Emulated IA32 system calls via int 0x80.
++ *
++ * Arguments:
++ * %eax System call number.
++ * %ebx Arg1
++ * %ecx Arg2
++ * %edx Arg3
++ * %esi Arg4
++ * %edi Arg5
++ * %ebp Arg6 [note: not saved in the stack frame, should not be touched]
++ *
++ * Notes:
++ * Uses the same stack frame as the x86-64 version.
++ * All registers except %eax must be saved (but ptrace may violate that)
++ * Arguments are zero extended. For system calls that want sign extension and
++ * take long arguments a wrapper is needed. Most calls can just be called
++ * directly.
++ * Assumes it is only called from user space and entered with interrupts on.
++ */
++
++ENTRY(ia32_syscall)
++ CFI_STARTPROC simple
++ CFI_DEF_CFA rsp,SS+8-RIP+16
++ /*CFI_REL_OFFSET ss,SS-RIP+16*/
++ CFI_REL_OFFSET rsp,RSP-RIP+16
++ /*CFI_REL_OFFSET rflags,EFLAGS-RIP+16*/
++ /*CFI_REL_OFFSET cs,CS-RIP+16*/
++ CFI_REL_OFFSET rip,RIP-RIP+16
++ CFI_REL_OFFSET r11,8
++ CFI_REL_OFFSET rcx,0
++ movq 8(%rsp),%r11
++ CFI_RESTORE r11
++ popq %rcx
++ CFI_ADJUST_CFA_OFFSET -8
++ CFI_RESTORE rcx
++ movl %eax,%eax
++ movq %rax,(%rsp)
++ cld
++ /* note the registers are not zero extended to the sf.
++ this could be a problem. */
++ SAVE_ARGS 0,0,1
++ GET_THREAD_INFO(%r10)
++ orl $TS_COMPAT,threadinfo_status(%r10)
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%r10)
++ jnz ia32_tracesys
++ia32_do_syscall:
++ cmpl $(IA32_NR_syscalls-1),%eax
++ ja ia32_badsys
++ IA32_ARG_FIXUP
++ call *ia32_sys_call_table(,%rax,8) # xxx: rip relative
++ia32_sysret:
++ movq %rax,RAX-ARGOFFSET(%rsp)
++ jmp int_ret_from_sys_call
++
++ia32_tracesys:
++ SAVE_REST
++ movq $-ENOSYS,RAX(%rsp) /* really needed? */
++ movq %rsp,%rdi /* &pt_regs -> arg1 */
++ call syscall_trace_enter
++ LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */
++ RESTORE_REST
++ jmp ia32_do_syscall
++END(ia32_syscall)
++
++ia32_badsys:
++ movq $0,ORIG_RAX-ARGOFFSET(%rsp)
++ movq $-ENOSYS,RAX-ARGOFFSET(%rsp)
++ jmp int_ret_from_sys_call
++
++quiet_ni_syscall:
++ movq $-ENOSYS,%rax
++ ret
++ CFI_ENDPROC
++
++ .macro PTREGSCALL label, func, arg
++ .globl \label
++\label:
++ leaq \func(%rip),%rax
++ leaq -ARGOFFSET+8(%rsp),\arg /* 8 for return address */
++ jmp ia32_ptregs_common
++ .endm
++
++ CFI_STARTPROC32
++
++ PTREGSCALL stub32_rt_sigreturn, sys32_rt_sigreturn, %rdi
++ PTREGSCALL stub32_sigreturn, sys32_sigreturn, %rdi
++ PTREGSCALL stub32_sigaltstack, sys32_sigaltstack, %rdx
++ PTREGSCALL stub32_sigsuspend, sys32_sigsuspend, %rcx
++ PTREGSCALL stub32_execve, sys32_execve, %rcx
++ PTREGSCALL stub32_fork, sys_fork, %rdi
++ PTREGSCALL stub32_clone, sys32_clone, %rdx
++ PTREGSCALL stub32_vfork, sys_vfork, %rdi
++ PTREGSCALL stub32_iopl, sys_iopl, %rsi
++ PTREGSCALL stub32_rt_sigsuspend, sys_rt_sigsuspend, %rdx
++
++ENTRY(ia32_ptregs_common)
++ popq %r11
++ CFI_ENDPROC
++ CFI_STARTPROC32 simple
++ CFI_DEF_CFA rsp,SS+8-ARGOFFSET
++ CFI_REL_OFFSET rax,RAX-ARGOFFSET
++ CFI_REL_OFFSET rcx,RCX-ARGOFFSET
++ CFI_REL_OFFSET rdx,RDX-ARGOFFSET
++ CFI_REL_OFFSET rsi,RSI-ARGOFFSET
++ CFI_REL_OFFSET rdi,RDI-ARGOFFSET
++ CFI_REL_OFFSET rip,RIP-ARGOFFSET
++/* CFI_REL_OFFSET cs,CS-ARGOFFSET*/
++/* CFI_REL_OFFSET rflags,EFLAGS-ARGOFFSET*/
++ CFI_REL_OFFSET rsp,RSP-ARGOFFSET
++/* CFI_REL_OFFSET ss,SS-ARGOFFSET*/
++ SAVE_REST
++ call *%rax
++ RESTORE_REST
++ jmp ia32_sysret /* misbalances the return cache */
++ CFI_ENDPROC
++END(ia32_ptregs_common)
++
++ .section .rodata,"a"
++ .align 8
++ia32_sys_call_table:
++ .quad sys_restart_syscall
++ .quad sys_exit
++ .quad stub32_fork
++ .quad sys_read
++ .quad sys_write
++ .quad compat_sys_open /* 5 */
++ .quad sys_close
++ .quad sys32_waitpid
++ .quad sys_creat
++ .quad sys_link
++ .quad sys_unlink /* 10 */
++ .quad stub32_execve
++ .quad sys_chdir
++ .quad compat_sys_time
++ .quad sys_mknod
++ .quad sys_chmod /* 15 */
++ .quad sys_lchown16
++ .quad quiet_ni_syscall /* old break syscall holder */
++ .quad sys_stat
++ .quad sys32_lseek
++ .quad sys_getpid /* 20 */
++ .quad compat_sys_mount /* mount */
++ .quad sys_oldumount /* old_umount */
++ .quad sys_setuid16
++ .quad sys_getuid16
++ .quad compat_sys_stime /* stime */ /* 25 */
++ .quad sys32_ptrace /* ptrace */
++ .quad sys_alarm
++ .quad sys_fstat /* (old)fstat */
++ .quad sys_pause
++ .quad compat_sys_utime /* 30 */
++ .quad quiet_ni_syscall /* old stty syscall holder */
++ .quad quiet_ni_syscall /* old gtty syscall holder */
++ .quad sys_access
++ .quad sys_nice
++ .quad quiet_ni_syscall /* 35 */ /* old ftime syscall holder */
++ .quad sys_sync
++ .quad sys32_kill
++ .quad sys_rename
++ .quad sys_mkdir
++ .quad sys_rmdir /* 40 */
++ .quad sys_dup
++ .quad sys32_pipe
++ .quad compat_sys_times
++ .quad quiet_ni_syscall /* old prof syscall holder */
++ .quad sys_brk /* 45 */
++ .quad sys_setgid16
++ .quad sys_getgid16
++ .quad sys_signal
++ .quad sys_geteuid16
++ .quad sys_getegid16 /* 50 */
++ .quad sys_acct
++ .quad sys_umount /* new_umount */
++ .quad quiet_ni_syscall /* old lock syscall holder */
++ .quad compat_sys_ioctl
++ .quad compat_sys_fcntl64 /* 55 */
++ .quad quiet_ni_syscall /* old mpx syscall holder */
++ .quad sys_setpgid
++ .quad quiet_ni_syscall /* old ulimit syscall holder */
++ .quad sys32_olduname
++ .quad sys_umask /* 60 */
++ .quad sys_chroot
++ .quad sys32_ustat
++ .quad sys_dup2
++ .quad sys_getppid
++ .quad sys_getpgrp /* 65 */
++ .quad sys_setsid
++ .quad sys32_sigaction
++ .quad sys_sgetmask
++ .quad sys_ssetmask
++ .quad sys_setreuid16 /* 70 */
++ .quad sys_setregid16
++ .quad stub32_sigsuspend
++ .quad compat_sys_sigpending
++ .quad sys_sethostname
++ .quad compat_sys_setrlimit /* 75 */
++ .quad compat_sys_old_getrlimit /* old_getrlimit */
++ .quad compat_sys_getrusage
++ .quad sys32_gettimeofday
++ .quad sys32_settimeofday
++ .quad sys_getgroups16 /* 80 */
++ .quad sys_setgroups16
++ .quad sys32_old_select
++ .quad sys_symlink
++ .quad sys_lstat
++ .quad sys_readlink /* 85 */
++#ifdef CONFIG_IA32_AOUT
++ .quad sys_uselib
++#else
++ .quad quiet_ni_syscall
++#endif
++ .quad sys_swapon
++ .quad sys_reboot
++ .quad compat_sys_old_readdir
++ .quad sys32_mmap /* 90 */
++ .quad sys_munmap
++ .quad sys_truncate
++ .quad sys_ftruncate
++ .quad sys_fchmod
++ .quad sys_fchown16 /* 95 */
++ .quad sys_getpriority
++ .quad sys_setpriority
++ .quad quiet_ni_syscall /* old profil syscall holder */
++ .quad compat_sys_statfs
++ .quad compat_sys_fstatfs /* 100 */
++ .quad sys_ioperm
++ .quad compat_sys_socketcall
++ .quad sys_syslog
++ .quad compat_sys_setitimer
++ .quad compat_sys_getitimer /* 105 */
++ .quad compat_sys_newstat
++ .quad compat_sys_newlstat
++ .quad compat_sys_newfstat
++ .quad sys32_uname
++ .quad stub32_iopl /* 110 */
++ .quad sys_vhangup
++ .quad quiet_ni_syscall /* old "idle" system call */
++ .quad sys32_vm86_warning /* vm86old */
++ .quad compat_sys_wait4
++ .quad sys_swapoff /* 115 */
++ .quad sys32_sysinfo
++ .quad sys32_ipc
++ .quad sys_fsync
++ .quad stub32_sigreturn
++ .quad stub32_clone /* 120 */
++ .quad sys_setdomainname
++ .quad sys_uname
++ .quad sys_modify_ldt
++ .quad compat_sys_adjtimex
++ .quad sys32_mprotect /* 125 */
++ .quad compat_sys_sigprocmask
++ .quad quiet_ni_syscall /* create_module */
++ .quad sys_init_module
++ .quad sys_delete_module
++ .quad quiet_ni_syscall /* 130 get_kernel_syms */
++ .quad sys_quotactl
++ .quad sys_getpgid
++ .quad sys_fchdir
++ .quad quiet_ni_syscall /* bdflush */
++ .quad sys_sysfs /* 135 */
++ .quad sys_personality
++ .quad quiet_ni_syscall /* for afs_syscall */
++ .quad sys_setfsuid16
++ .quad sys_setfsgid16
++ .quad sys_llseek /* 140 */
++ .quad compat_sys_getdents
++ .quad compat_sys_select
++ .quad sys_flock
++ .quad sys_msync
++ .quad compat_sys_readv /* 145 */
++ .quad compat_sys_writev
++ .quad sys_getsid
++ .quad sys_fdatasync
++ .quad sys32_sysctl /* sysctl */
++ .quad sys_mlock /* 150 */
++ .quad sys_munlock
++ .quad sys_mlockall
++ .quad sys_munlockall
++ .quad sys_sched_setparam
++ .quad sys_sched_getparam /* 155 */
++ .quad sys_sched_setscheduler
++ .quad sys_sched_getscheduler
++ .quad sys_sched_yield
++ .quad sys_sched_get_priority_max
++ .quad sys_sched_get_priority_min /* 160 */
++ .quad sys_sched_rr_get_interval
++ .quad compat_sys_nanosleep
++ .quad sys_mremap
++ .quad sys_setresuid16
++ .quad sys_getresuid16 /* 165 */
++ .quad sys32_vm86_warning /* vm86 */
++ .quad quiet_ni_syscall /* query_module */
++ .quad sys_poll
++ .quad compat_sys_nfsservctl
++ .quad sys_setresgid16 /* 170 */
++ .quad sys_getresgid16
++ .quad sys_prctl
++ .quad stub32_rt_sigreturn
++ .quad sys32_rt_sigaction
++ .quad sys32_rt_sigprocmask /* 175 */
++ .quad sys32_rt_sigpending
++ .quad compat_sys_rt_sigtimedwait
++ .quad sys32_rt_sigqueueinfo
++ .quad stub32_rt_sigsuspend
++ .quad sys32_pread /* 180 */
++ .quad sys32_pwrite
++ .quad sys_chown16
++ .quad sys_getcwd
++ .quad sys_capget
++ .quad sys_capset
++ .quad stub32_sigaltstack
++ .quad sys32_sendfile
++ .quad quiet_ni_syscall /* streams1 */
++ .quad quiet_ni_syscall /* streams2 */
++ .quad stub32_vfork /* 190 */
++ .quad compat_sys_getrlimit
++ .quad sys32_mmap2
++ .quad sys32_truncate64
++ .quad sys32_ftruncate64
++ .quad sys32_stat64 /* 195 */
++ .quad sys32_lstat64
++ .quad sys32_fstat64
++ .quad sys_lchown
++ .quad sys_getuid
++ .quad sys_getgid /* 200 */
++ .quad sys_geteuid
++ .quad sys_getegid
++ .quad sys_setreuid
++ .quad sys_setregid
++ .quad sys_getgroups /* 205 */
++ .quad sys_setgroups
++ .quad sys_fchown
++ .quad sys_setresuid
++ .quad sys_getresuid
++ .quad sys_setresgid /* 210 */
++ .quad sys_getresgid
++ .quad sys_chown
++ .quad sys_setuid
++ .quad sys_setgid
++ .quad sys_setfsuid /* 215 */
++ .quad sys_setfsgid
++ .quad sys_pivot_root
++ .quad sys_mincore
++ .quad sys_madvise
++ .quad compat_sys_getdents64 /* 220 getdents64 */
++ .quad compat_sys_fcntl64
++ .quad quiet_ni_syscall /* tux */
++ .quad quiet_ni_syscall /* security */
++ .quad sys_gettid
++ .quad sys_readahead /* 225 */
++ .quad sys_setxattr
++ .quad sys_lsetxattr
++ .quad sys_fsetxattr
++ .quad sys_getxattr
++ .quad sys_lgetxattr /* 230 */
++ .quad sys_fgetxattr
++ .quad sys_listxattr
++ .quad sys_llistxattr
++ .quad sys_flistxattr
++ .quad sys_removexattr /* 235 */
++ .quad sys_lremovexattr
++ .quad sys_fremovexattr
++ .quad sys_tkill
++ .quad sys_sendfile64
++ .quad compat_sys_futex /* 240 */
++ .quad compat_sys_sched_setaffinity
++ .quad compat_sys_sched_getaffinity
++ .quad sys32_set_thread_area
++ .quad sys32_get_thread_area
++ .quad compat_sys_io_setup /* 245 */
++ .quad sys_io_destroy
++ .quad compat_sys_io_getevents
++ .quad compat_sys_io_submit
++ .quad sys_io_cancel
++ .quad sys_fadvise64 /* 250 */
++ .quad quiet_ni_syscall /* free_huge_pages */
++ .quad sys_exit_group
++ .quad sys32_lookup_dcookie
++ .quad sys_epoll_create
++ .quad sys_epoll_ctl /* 255 */
++ .quad sys_epoll_wait
++ .quad sys_remap_file_pages
++ .quad sys_set_tid_address
++ .quad compat_sys_timer_create
++ .quad compat_sys_timer_settime /* 260 */
++ .quad compat_sys_timer_gettime
++ .quad sys_timer_getoverrun
++ .quad sys_timer_delete
++ .quad compat_sys_clock_settime
++ .quad compat_sys_clock_gettime /* 265 */
++ .quad compat_sys_clock_getres
++ .quad compat_sys_clock_nanosleep
++ .quad compat_sys_statfs64
++ .quad compat_sys_fstatfs64
++ .quad sys_tgkill /* 270 */
++ .quad compat_sys_utimes
++ .quad sys32_fadvise64_64
++ .quad quiet_ni_syscall /* sys_vserver */
++ .quad sys_mbind
++ .quad compat_sys_get_mempolicy /* 275 */
++ .quad sys_set_mempolicy
++ .quad compat_sys_mq_open
++ .quad sys_mq_unlink
++ .quad compat_sys_mq_timedsend
++ .quad compat_sys_mq_timedreceive /* 280 */
++ .quad compat_sys_mq_notify
++ .quad compat_sys_mq_getsetattr
++ .quad compat_sys_kexec_load /* reserved for kexec */
++ .quad compat_sys_waitid
++ .quad quiet_ni_syscall /* 285: sys_altroot */
++ .quad sys_add_key
++ .quad sys_request_key
++ .quad sys_keyctl
++ .quad sys_ioprio_set
++ .quad sys_ioprio_get /* 290 */
++ .quad sys_inotify_init
++ .quad sys_inotify_add_watch
++ .quad sys_inotify_rm_watch
++ .quad sys_migrate_pages
++ .quad compat_sys_openat /* 295 */
++ .quad sys_mkdirat
++ .quad sys_mknodat
++ .quad sys_fchownat
++ .quad compat_sys_futimesat
++ .quad sys32_fstatat /* 300 */
++ .quad sys_unlinkat
++ .quad sys_renameat
++ .quad sys_linkat
++ .quad sys_symlinkat
++ .quad sys_readlinkat /* 305 */
++ .quad sys_fchmodat
++ .quad sys_faccessat
++ .quad compat_sys_pselect6
++ .quad compat_sys_ppoll
++ .quad sys_unshare /* 310 */
++ .quad compat_sys_set_robust_list
++ .quad compat_sys_get_robust_list
++ .quad sys_splice
++ .quad sys_sync_file_range
++ .quad sys_tee
++ .quad compat_sys_vmsplice
++ .quad compat_sys_move_pages
++ia32_syscall_end:
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/ia32/ia32entry.S
+--- a/arch/x86_64/ia32/ia32entry.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/ia32/ia32entry.S Wed Sep 10 10:54:08 2008 +0100
+@@ -715,8 +715,8 @@
+ .quad sys_readlinkat /* 305 */
+ .quad sys_fchmodat
+ .quad sys_faccessat
+- .quad quiet_ni_syscall /* pselect6 for now */
+- .quad quiet_ni_syscall /* ppoll for now */
++ .quad compat_sys_pselect6
++ .quad compat_sys_ppoll
+ .quad sys_unshare /* 310 */
+ .quad compat_sys_set_robust_list
+ .quad compat_sys_get_robust_list
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/ia32/syscall32-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/ia32/syscall32-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,137 @@
++/* Copyright 2002,2003 Andi Kleen, SuSE Labs */
++
++/* vsyscall handling for 32bit processes. Map a stub page into it
++ on demand because 32bit cannot reach the kernel's fixmaps */
++
++#include <linux/mm.h>
++#include <linux/string.h>
++#include <linux/kernel.h>
++#include <linux/gfp.h>
++#include <linux/init.h>
++#include <linux/stringify.h>
++#include <linux/security.h>
++#include <asm/proto.h>
++#include <asm/tlbflush.h>
++#include <asm/ia32_unistd.h>
++#include <xen/interface/callback.h>
++
++extern unsigned char syscall32_syscall[], syscall32_syscall_end[];
++extern unsigned char syscall32_sysenter[], syscall32_sysenter_end[];
++extern int sysctl_vsyscall32;
++
++char *syscall32_page;
++static int use_sysenter = -1;
++
++#if CONFIG_XEN_COMPAT < 0x030200
++extern unsigned char syscall32_int80[], syscall32_int80_end[];
++static int use_int80 = 1;
++#endif
++
++static struct page *
++syscall32_nopage(struct vm_area_struct *vma, unsigned long adr, int *type)
++{
++ struct page *p = virt_to_page(adr - vma->vm_start + syscall32_page);
++ get_page(p);
++ return p;
++}
++
++/* Prevent VMA merging */
++static void syscall32_vma_close(struct vm_area_struct *vma)
++{
++}
++
++static struct vm_operations_struct syscall32_vm_ops = {
++ .close = syscall32_vma_close,
++ .nopage = syscall32_nopage,
++};
++
++struct linux_binprm;
++
++/* Setup a VMA at program startup for the vsyscall page */
++int syscall32_setup_pages(struct linux_binprm *bprm, int exstack)
++{
++ int npages = (VSYSCALL32_END - VSYSCALL32_BASE) >> PAGE_SHIFT;
++ struct vm_area_struct *vma;
++ struct mm_struct *mm = current->mm;
++ int ret;
++
++ vma = kmem_cache_alloc(vm_area_cachep, SLAB_KERNEL);
++ if (!vma)
++ return -ENOMEM;
++
++ memset(vma, 0, sizeof(struct vm_area_struct));
++ /* Could randomize here */
++ vma->vm_start = VSYSCALL32_BASE;
++ vma->vm_end = VSYSCALL32_END;
++ /* MAYWRITE to allow gdb to COW and set breakpoints */
++ vma->vm_flags = VM_READ|VM_EXEC|VM_MAYREAD|VM_MAYEXEC|VM_MAYWRITE;
++ vma->vm_flags |= mm->def_flags;
++ vma->vm_page_prot = protection_map[vma->vm_flags & 7];
++ vma->vm_ops = &syscall32_vm_ops;
++ vma->vm_mm = mm;
++
++ down_write(&mm->mmap_sem);
++ if ((ret = insert_vm_struct(mm, vma))) {
++ up_write(&mm->mmap_sem);
++ kmem_cache_free(vm_area_cachep, vma);
++ return ret;
++ }
++ mm->total_vm += npages;
++ up_write(&mm->mmap_sem);
++ return 0;
++}
++
++static int __init init_syscall32(void)
++{
++ syscall32_page = (void *)get_zeroed_page(GFP_KERNEL);
++ if (!syscall32_page)
++ panic("Cannot allocate syscall32 page");
++
++#if CONFIG_XEN_COMPAT < 0x030200
++ if (use_int80) {
++ memcpy(syscall32_page, syscall32_int80,
++ syscall32_int80_end - syscall32_int80);
++ } else
++#endif
++ if (use_sysenter > 0) {
++ memcpy(syscall32_page, syscall32_sysenter,
++ syscall32_sysenter_end - syscall32_sysenter);
++ } else {
++ memcpy(syscall32_page, syscall32_syscall,
++ syscall32_syscall_end - syscall32_syscall);
++ }
++ return 0;
++}
++
++/*
++ * This must be done early in case we have an initrd containing 32-bit
++ * binaries (e.g., hotplug). This could be pushed upstream to arch/x86_64.
++ */
++core_initcall(init_syscall32);
++
++/* May not be __init: called during resume */
++void syscall32_cpu_init(void)
++{
++ static const struct callback_register cstar = {
++ .type = CALLBACKTYPE_syscall32,
++ .address = (unsigned long)ia32_cstar_target
++ };
++ static const struct callback_register sysenter = {
++ .type = CALLBACKTYPE_sysenter,
++ .address = (unsigned long)ia32_sysenter_target
++ };
++
++ /* Load these always in case some future AMD CPU supports
++ SYSENTER from compat mode too. */
++ if ((HYPERVISOR_callback_op(CALLBACKOP_register, &sysenter) < 0) ||
++ (HYPERVISOR_callback_op(CALLBACKOP_register, &cstar) < 0))
++#if CONFIG_XEN_COMPAT < 0x030200
++ return;
++ use_int80 = 0;
++#else
++ BUG();
++#endif
++
++ if (use_sysenter < 0)
++ use_sysenter = (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL);
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/ia32/syscall32.c
+--- a/arch/x86_64/ia32/syscall32.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/ia32/syscall32.c Wed Sep 10 10:54:08 2008 +0100
+@@ -59,6 +59,13 @@
+ vma->vm_end = VSYSCALL32_END;
+ /* MAYWRITE to allow gdb to COW and set breakpoints */
+ vma->vm_flags = VM_READ|VM_EXEC|VM_MAYREAD|VM_MAYEXEC|VM_MAYWRITE;
++ /*
++ * Make sure the vDSO gets into every core dump.
++ * Dumping its contents makes post-mortem fully interpretable later
++ * without matching up the same kernel and hardware config to see
++ * what PC values meant.
++ */
++ vma->vm_flags |= VM_ALWAYSDUMP;
+ vma->vm_flags |= mm->def_flags;
+ vma->vm_page_prot = protection_map[vma->vm_flags & 7];
+ vma->vm_ops = &syscall32_vm_ops;
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/ia32/syscall32_syscall-xen.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/ia32/syscall32_syscall-xen.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,28 @@
++/* 32bit VDSOs mapped into user space. */
++
++ .section ".init.data","aw"
++
++#if CONFIG_XEN_COMPAT < 0x030200
++
++ .globl syscall32_int80
++ .globl syscall32_int80_end
++
++syscall32_int80:
++ .incbin "arch/x86_64/ia32/vsyscall-int80.so"
++syscall32_int80_end:
++
++#endif
++
++ .globl syscall32_syscall
++ .globl syscall32_syscall_end
++
++syscall32_syscall:
++ .incbin "arch/x86_64/ia32/vsyscall-syscall.so"
++syscall32_syscall_end:
++
++ .globl syscall32_sysenter
++ .globl syscall32_sysenter_end
++
++syscall32_sysenter:
++ .incbin "arch/x86_64/ia32/vsyscall-sysenter.so"
++syscall32_sysenter_end:
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/ia32/vsyscall-int80.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/ia32/vsyscall-int80.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,58 @@
++/*
++ * Code for the vsyscall page. This version uses the old int $0x80 method.
++ *
++ * NOTE:
++ * 1) __kernel_vsyscall _must_ be first in this page.
++ * 2) there are alignment constraints on this stub, see vsyscall-sigreturn.S
++ * for details.
++ */
++#include <asm/ia32_unistd.h>
++#include <asm/asm-offsets.h>
++
++ .code32
++ .text
++ .section .text.vsyscall,"ax"
++ .globl __kernel_vsyscall
++ .type __kernel_vsyscall, at function
++__kernel_vsyscall:
++.LSTART_vsyscall:
++ int $0x80
++ ret
++.LEND_vsyscall:
++ .size __kernel_vsyscall,.-.LSTART_vsyscall
++ .previous
++
++ .section .eh_frame,"a", at progbits
++.LSTARTFRAME:
++ .long .LENDCIE-.LSTARTCIE
++.LSTARTCIE:
++ .long 0 /* CIE ID */
++ .byte 1 /* Version number */
++ .string "zR" /* NUL-terminated augmentation string */
++ .uleb128 1 /* Code alignment factor */
++ .sleb128 -4 /* Data alignment factor */
++ .byte 8 /* Return address register column */
++ .uleb128 1 /* Augmentation value length */
++ .byte 0x1b /* DW_EH_PE_pcrel|DW_EH_PE_sdata4. */
++ .byte 0x0c /* DW_CFA_def_cfa */
++ .uleb128 4
++ .uleb128 4
++ .byte 0x88 /* DW_CFA_offset, column 0x8 */
++ .uleb128 1
++ .align 4
++.LENDCIE:
++
++ .long .LENDFDE1-.LSTARTFDE1 /* Length FDE */
++.LSTARTFDE1:
++ .long .LSTARTFDE1-.LSTARTFRAME /* CIE pointer */
++ .long .LSTART_vsyscall-. /* PC-relative start address */
++ .long .LEND_vsyscall-.LSTART_vsyscall
++ .uleb128 0 /* Augmentation length */
++ .align 4
++.LENDFDE1:
++
++/*
++ * Get the common code for the sigreturn entry points.
++ */
++#define SYSCALL_ENTER_KERNEL int $0x80
++#include "vsyscall-sigreturn.S"
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/ia32/vsyscall-sigreturn.S
+--- a/arch/x86_64/ia32/vsyscall-sigreturn.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/ia32/vsyscall-sigreturn.S Wed Sep 10 10:54:08 2008 +0100
+@@ -139,5 +139,5 @@
+ .align 4
+ .LENDFDE3:
+
+-#include "../../i386/kernel/vsyscall-note.S"
++#include <vsyscall-note.S>
+
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/Makefile
+--- a/arch/x86_64/kernel/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/kernel/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -21,11 +21,13 @@
+ obj-$(CONFIG_X86_CPUID) += cpuid.o
+ obj-$(CONFIG_SMP) += smp.o smpboot.o trampoline.o
+ obj-$(CONFIG_X86_LOCAL_APIC) += apic.o nmi.o
++obj-$(CONFIG_X86_XEN_GENAPIC) += genapic.o genapic_xen.o
+ obj-$(CONFIG_X86_IO_APIC) += io_apic.o mpparse.o \
+ genapic.o genapic_cluster.o genapic_flat.o
+ obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o crash.o
+ obj-$(CONFIG_CRASH_DUMP) += crash_dump.o
+-obj-$(CONFIG_PM) += suspend.o
++obj-$(CONFIG_SOFTWARE_SUSPEND) += suspend.o
++obj-$(CONFIG_ACPI_SLEEP) += suspend.o
+ obj-$(CONFIG_SOFTWARE_SUSPEND) += suspend_asm.o
+ obj-$(CONFIG_CPU_FREQ) += cpufreq/
+ obj-$(CONFIG_EARLY_PRINTK) += early_printk.o
+@@ -55,3 +57,8 @@
+ msr-$(subst m,y,$(CONFIG_X86_MSR)) += ../../i386/kernel/msr.o
+ alternative-y += ../../i386/kernel/alternative.o
+
++time-$(CONFIG_XEN) += ../../i386/kernel/time.o
++pci-dma-$(CONFIG_XEN) += ../../i386/kernel/pci-dma.o
++
++disabled-obj-$(CONFIG_XEN) := i8259.o reboot.o smpboot.o trampoline.o
++%/head.o %/head.s: $(if $(CONFIG_XEN),EXTRA_AFLAGS,dummy) :=
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/acpi/Makefile
+--- a/arch/x86_64/kernel/acpi/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/kernel/acpi/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -5,5 +5,10 @@
+ ifneq ($(CONFIG_ACPI_PROCESSOR),)
+ obj-y += processor.o
+ processor-y := ../../../i386/kernel/acpi/processor.o ../../../i386/kernel/acpi/cstate.o
++processor-$(CONFIG_XEN) := ../../../i386/kernel/acpi/processor.o
++ifneq ($(CONFIG_PROCESSOR_EXTERNAL_CONTROL),)
++processor-$(CONFIG_XEN) += ../../../i386/kernel/acpi/processor_extcntl_xen.o
++endif
+ endif
+
++disabled-obj-$(CONFIG_XEN) := wakeup.o
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/acpi/sleep-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/acpi/sleep-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,146 @@
++/*
++ * acpi.c - Architecture-Specific Low-Level ACPI Support
++ *
++ * Copyright (C) 2001, 2002 Paul Diefenbaugh <paul.s.diefenbaugh at intel.com>
++ * Copyright (C) 2001 Jun Nakajima <jun.nakajima at intel.com>
++ * Copyright (C) 2001 Patrick Mochel <mochel at osdl.org>
++ * Copyright (C) 2002 Andi Kleen, SuSE Labs (x86-64 port)
++ * Copyright (C) 2003 Pavel Machek, SuSE Labs
++ *
++ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ */
++
++#include <linux/kernel.h>
++#include <linux/init.h>
++#include <linux/types.h>
++#include <linux/stddef.h>
++#include <linux/slab.h>
++#include <linux/pci.h>
++#include <linux/bootmem.h>
++#include <linux/acpi.h>
++#include <linux/cpumask.h>
++
++#include <asm/mpspec.h>
++#include <asm/io.h>
++#include <asm/apic.h>
++#include <asm/apicdef.h>
++#include <asm/page.h>
++#include <asm/pgtable.h>
++#include <asm/pgalloc.h>
++#include <asm/io_apic.h>
++#include <asm/proto.h>
++#include <asm/tlbflush.h>
++
++/* --------------------------------------------------------------------------
++ Low-Level Sleep Support
++ -------------------------------------------------------------------------- */
++
++#ifdef CONFIG_ACPI_SLEEP
++
++#ifndef CONFIG_ACPI_PV_SLEEP
++/* address in low memory of the wakeup routine. */
++unsigned long acpi_wakeup_address = 0;
++unsigned long acpi_video_flags;
++extern char wakeup_start, wakeup_end;
++
++extern unsigned long FASTCALL(acpi_copy_wakeup_routine(unsigned long));
++
++static pgd_t low_ptr;
++
++static void init_low_mapping(void)
++{
++ pgd_t *slot0 = pgd_offset(current->mm, 0UL);
++ low_ptr = *slot0;
++ set_pgd(slot0, *pgd_offset(current->mm, PAGE_OFFSET));
++ WARN_ON(num_online_cpus() != 1);
++ local_flush_tlb();
++}
++#endif
++
++/**
++ * acpi_save_state_mem - save kernel state
++ *
++ * Create an identity mapped page table and copy the wakeup routine to
++ * low memory.
++ */
++int acpi_save_state_mem(void)
++{
++#ifndef CONFIG_ACPI_PV_SLEEP
++ init_low_mapping();
++
++ memcpy((void *)acpi_wakeup_address, &wakeup_start,
++ &wakeup_end - &wakeup_start);
++ acpi_copy_wakeup_routine(acpi_wakeup_address);
++#endif
++ return 0;
++}
++
++/*
++ * acpi_restore_state
++ */
++void acpi_restore_state_mem(void)
++{
++#ifndef CONFIG_ACPI_PV_SLEEP
++ set_pgd(pgd_offset(current->mm, 0UL), low_ptr);
++ local_flush_tlb();
++#endif
++}
++
++/**
++ * acpi_reserve_bootmem - do _very_ early ACPI initialisation
++ *
++ * We allocate a page in low memory for the wakeup
++ * routine for when we come back from a sleep state. The
++ * runtime allocator allows specification of <16M pages, but not
++ * <1M pages.
++ */
++void __init acpi_reserve_bootmem(void)
++{
++#ifndef CONFIG_ACPI_PV_SLEEP
++ acpi_wakeup_address = (unsigned long)alloc_bootmem_low(PAGE_SIZE);
++ if ((&wakeup_end - &wakeup_start) > PAGE_SIZE)
++ printk(KERN_CRIT
++ "ACPI: Wakeup code way too big, will crash on attempt to suspend\n");
++#endif
++}
++
++#ifndef CONFIG_ACPI_PV_SLEEP
++static int __init acpi_sleep_setup(char *str)
++{
++ while ((str != NULL) && (*str != '\0')) {
++ if (strncmp(str, "s3_bios", 7) == 0)
++ acpi_video_flags = 1;
++ if (strncmp(str, "s3_mode", 7) == 0)
++ acpi_video_flags |= 2;
++ str = strchr(str, ',');
++ if (str != NULL)
++ str += strspn(str, ", \t");
++ }
++
++ return 1;
++}
++
++__setup("acpi_sleep=", acpi_sleep_setup);
++#endif /* CONFIG_ACPI_PV_SLEEP */
++
++#endif /*CONFIG_ACPI_SLEEP */
++
++void acpi_pci_link_exit(void)
++{
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/apic-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/apic-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,197 @@
++/*
++ * Local APIC handling, local APIC timers
++ *
++ * (c) 1999, 2000 Ingo Molnar <mingo at redhat.com>
++ *
++ * Fixes
++ * Maciej W. Rozycki : Bits for genuine 82489DX APICs;
++ * thanks to Eric Gilmore
++ * and Rolf G. Tews
++ * for testing these extensively.
++ * Maciej W. Rozycki : Various updates and fixes.
++ * Mikael Pettersson : Power Management for UP-APIC.
++ * Pavel Machek and
++ * Mikael Pettersson : PM converted to driver model.
++ */
++
++#include <linux/init.h>
++
++#include <linux/mm.h>
++#include <linux/delay.h>
++#include <linux/bootmem.h>
++#include <linux/smp_lock.h>
++#include <linux/interrupt.h>
++#include <linux/mc146818rtc.h>
++#include <linux/kernel_stat.h>
++#include <linux/sysdev.h>
++#include <linux/module.h>
++
++#include <asm/atomic.h>
++#include <asm/smp.h>
++#include <asm/mtrr.h>
++#include <asm/mpspec.h>
++#include <asm/desc.h>
++#include <asm/arch_hooks.h>
++#include <asm/hpet.h>
++#include <asm/idle.h>
++
++int apic_verbosity;
++
++/*
++ * 'what should we do if we get a hw irq event on an illegal vector'.
++ * each architecture has to answer this themselves.
++ */
++void ack_bad_irq(unsigned int irq)
++{
++ printk("unexpected IRQ trap at vector %02x\n", irq);
++ /*
++ * Currently unexpected vectors happen only on SMP and APIC.
++ * We _must_ ack these because every local APIC has only N
++ * irq slots per priority level, and a 'hanging, unacked' IRQ
++ * holds up an irq slot - in excessive cases (when multiple
++ * unexpected vectors occur) that might lock up the APIC
++ * completely.
++ * But don't ack when the APIC is disabled. -AK
++ */
++ if (!disable_apic)
++ ack_APIC_irq();
++}
++
++int setup_profiling_timer(unsigned int multiplier)
++{
++ return -EINVAL;
++}
++
++void smp_local_timer_interrupt(struct pt_regs *regs)
++{
++ profile_tick(CPU_PROFILING, regs);
++#ifndef CONFIG_XEN
++#ifdef CONFIG_SMP
++ update_process_times(user_mode(regs));
++#endif
++#endif
++ /*
++ * We take the 'long' return path, and there every subsystem
++ * grabs the appropriate locks (kernel lock/ irq lock).
++ *
++ * we might want to decouple profiling from the 'long path',
++ * and do the profiling totally in assembly.
++ *
++ * Currently this isn't too much of an issue (performance wise),
++ * we can take more than 100K local irqs per second on a 100 MHz P5.
++ */
++}
++
++/*
++ * Local APIC timer interrupt. This is the most natural way for doing
++ * local interrupts, but local timer interrupts can be emulated by
++ * broadcast interrupts too. [in case the hw doesn't support APIC timers]
++ *
++ * [ if a single-CPU system runs an SMP kernel then we call the local
++ * interrupt as well. Thus we cannot inline the local irq ... ]
++ */
++void smp_apic_timer_interrupt(struct pt_regs *regs)
++{
++ /*
++ * the NMI deadlock-detector uses this.
++ */
++ add_pda(apic_timer_irqs, 1);
++
++ /*
++ * NOTE! We'd better ACK the irq immediately,
++ * because timer handling can be slow.
++ */
++ ack_APIC_irq();
++ /*
++ * update_process_times() expects us to have done irq_enter().
++ * Besides, if we don't timer interrupts ignore the global
++ * interrupt lock, which is the WrongThing (tm) to do.
++ */
++ exit_idle();
++ irq_enter();
++ smp_local_timer_interrupt(regs);
++ irq_exit();
++}
++
++/*
++ * This interrupt should _never_ happen with our APIC/SMP architecture
++ */
++asmlinkage void smp_spurious_interrupt(void)
++{
++ unsigned int v;
++ exit_idle();
++ irq_enter();
++ /*
++ * Check if this really is a spurious interrupt and ACK it
++ * if it is a vectored one. Just in case...
++ * Spurious interrupts should not be ACKed.
++ */
++ v = apic_read(APIC_ISR + ((SPURIOUS_APIC_VECTOR & ~0x1f) >> 1));
++ if (v & (1 << (SPURIOUS_APIC_VECTOR & 0x1f)))
++ ack_APIC_irq();
++
++#if 0
++ static unsigned long last_warning;
++ static unsigned long skipped;
++
++ /* see sw-dev-man vol 3, chapter 7.4.13.5 */
++ if (time_before(last_warning+30*HZ,jiffies)) {
++ printk(KERN_INFO "spurious APIC interrupt on CPU#%d, %ld skipped.\n",
++ smp_processor_id(), skipped);
++ last_warning = jiffies;
++ skipped = 0;
++ } else {
++ skipped++;
++ }
++#endif
++ irq_exit();
++}
++
++/*
++ * This interrupt should never happen with our APIC/SMP architecture
++ */
++
++asmlinkage void smp_error_interrupt(void)
++{
++ unsigned int v, v1;
++
++ exit_idle();
++ irq_enter();
++ /* First tickle the hardware, only then report what went on. -- REW */
++ v = apic_read(APIC_ESR);
++ apic_write(APIC_ESR, 0);
++ v1 = apic_read(APIC_ESR);
++ ack_APIC_irq();
++ atomic_inc(&irq_err_count);
++
++ /* Here is what the APIC error bits mean:
++ 0: Send CS error
++ 1: Receive CS error
++ 2: Send accept error
++ 3: Receive accept error
++ 4: Reserved
++ 5: Send illegal vector
++ 6: Received illegal vector
++ 7: Illegal register address
++ */
++ printk (KERN_DEBUG "APIC error on CPU%d: %02x(%02x)\n",
++ smp_processor_id(), v , v1);
++ irq_exit();
++}
++
++int disable_apic;
++
++/*
++ * This initializes the IO-APIC and APIC hardware if this is
++ * a UP kernel.
++ */
++int __init APIC_init_uniprocessor (void)
++{
++#ifdef CONFIG_X86_IO_APIC
++ if (smp_found_config)
++ if (!skip_ioapic_setup && nr_ioapics)
++ setup_IO_APIC();
++#endif
++
++ return 1;
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/asm-offsets.c
+--- a/arch/x86_64/kernel/asm-offsets.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/kernel/asm-offsets.c Wed Sep 10 10:54:08 2008 +0100
+@@ -67,8 +67,10 @@
+ DEFINE(pbe_address, offsetof(struct pbe, address));
+ DEFINE(pbe_orig_address, offsetof(struct pbe, orig_address));
+ DEFINE(pbe_next, offsetof(struct pbe, next));
++#ifndef CONFIG_X86_NO_TSS
+ BLANK();
+ DEFINE(TSS_ist, offsetof(struct tss_struct, ist));
++#endif
+ BLANK();
+ DEFINE(crypto_tfm_ctx_offset, offsetof(struct crypto_tfm, __crt_ctx));
+ return 0;
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/crash.c
+--- a/arch/x86_64/kernel/crash.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/kernel/crash.c Wed Sep 10 10:54:08 2008 +0100
+@@ -92,6 +92,7 @@
+ crash_save_this_cpu(regs, cpu);
+ }
+
++#ifndef CONFIG_XEN
+ #ifdef CONFIG_SMP
+ static atomic_t waiting_for_crash_ipi;
+
+@@ -156,6 +157,7 @@
+ /* There are no cpus to shootdown */
+ }
+ #endif
++#endif /* CONFIG_XEN */
+
+ void machine_crash_shutdown(struct pt_regs *regs)
+ {
+@@ -173,6 +175,8 @@
+
+ /* Make a note of crashing cpu. Will be used in NMI callback.*/
+ crashing_cpu = smp_processor_id();
++
++#ifndef CONFIG_XEN
+ nmi_shootdown_cpus();
+
+ if(cpu_has_apic)
+@@ -181,6 +185,6 @@
+ #if defined(CONFIG_X86_IO_APIC)
+ disable_IO_APIC();
+ #endif
+-
++#endif /* CONFIG_XEN */
+ crash_save_self(regs);
+ }
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/e820-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/e820-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,798 @@
++/*
++ * Handle the memory map.
++ * The functions here do the job until bootmem takes over.
++ *
++ * Getting sanitize_e820_map() in sync with i386 version by applying change:
++ * - Provisions for empty E820 memory regions (reported by certain BIOSes).
++ * Alex Achenbach <xela at slit.de>, December 2002.
++ * Venkatesh Pallipadi <venkatesh.pallipadi at intel.com>
++ *
++ */
++#include <linux/kernel.h>
++#include <linux/types.h>
++#include <linux/init.h>
++#include <linux/bootmem.h>
++#include <linux/ioport.h>
++#include <linux/string.h>
++#include <linux/kexec.h>
++#include <linux/module.h>
++
++#include <asm/pgtable.h>
++#include <asm/page.h>
++#include <asm/e820.h>
++#include <asm/proto.h>
++#include <asm/bootsetup.h>
++#include <asm/sections.h>
++#include <xen/interface/memory.h>
++
++/*
++ * PFN of last memory page.
++ */
++unsigned long end_pfn;
++EXPORT_SYMBOL(end_pfn);
++
++/*
++ * end_pfn only includes RAM, while end_pfn_map includes all e820 entries.
++ * The direct mapping extends to end_pfn_map, so that we can directly access
++ * apertures, ACPI and other tables without having to play with fixmaps.
++ */
++unsigned long end_pfn_map;
++
++/*
++ * Last pfn which the user wants to use.
++ */
++unsigned long end_user_pfn = MAXMEM>>PAGE_SHIFT;
++
++extern struct resource code_resource, data_resource;
++
++#ifdef CONFIG_XEN
++extern struct e820map machine_e820;
++#endif
++
++/* Check for some hardcoded bad areas that early boot is not allowed to touch */
++static inline int bad_addr(unsigned long *addrp, unsigned long size)
++{
++ unsigned long addr = *addrp, last = addr + size;
++
++#ifndef CONFIG_XEN
++ /* various gunk below that needed for SMP startup */
++ if (addr < 0x8000) {
++ *addrp = 0x8000;
++ return 1;
++ }
++
++ /* direct mapping tables of the kernel */
++ if (last >= table_start<<PAGE_SHIFT && addr < table_end<<PAGE_SHIFT) {
++ *addrp = table_end << PAGE_SHIFT;
++ return 1;
++ }
++
++ /* initrd */
++#ifdef CONFIG_BLK_DEV_INITRD
++ if (LOADER_TYPE && INITRD_START && last >= INITRD_START &&
++ addr < INITRD_START+INITRD_SIZE) {
++ *addrp = INITRD_START + INITRD_SIZE;
++ return 1;
++ }
++#endif
++ /* kernel code + 640k memory hole (later should not be needed, but
++ be paranoid for now) */
++ if (last >= 640*1024 && addr < 1024*1024) {
++ *addrp = 1024*1024;
++ return 1;
++ }
++ if (last >= __pa_symbol(&_text) && last < __pa_symbol(&_end)) {
++ *addrp = __pa_symbol(&_end);
++ return 1;
++ }
++
++ if (last >= ebda_addr && addr < ebda_addr + ebda_size) {
++ *addrp = ebda_addr + ebda_size;
++ return 1;
++ }
++
++ /* XXX ramdisk image here? */
++#else
++ if (last < (table_end<<PAGE_SHIFT)) {
++ *addrp = table_end << PAGE_SHIFT;
++ return 1;
++ }
++#endif
++ return 0;
++}
++
++/*
++ * This function checks if any part of the range <start,end> is mapped
++ * with type.
++ */
++int e820_any_mapped(unsigned long start, unsigned long end, unsigned type)
++{
++ int i;
++
++#ifndef CONFIG_XEN
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++#else
++ if (!is_initial_xendomain())
++ return 0;
++ for (i = 0; i < machine_e820.nr_map; i++) {
++ const struct e820entry *ei = &machine_e820.map[i];
++#endif
++
++ if (type && ei->type != type)
++ continue;
++ if (ei->addr >= end || ei->addr + ei->size <= start)
++ continue;
++ return 1;
++ }
++ return 0;
++}
++EXPORT_SYMBOL_GPL(e820_any_mapped);
++
++/*
++ * This function checks if the entire range <start,end> is mapped with type.
++ *
++ * Note: this function only works correct if the e820 table is sorted and
++ * not-overlapping, which is the case
++ */
++int __init e820_all_mapped(unsigned long start, unsigned long end, unsigned type)
++{
++ int i;
++
++#ifndef CONFIG_XEN
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++#else
++ if (!is_initial_xendomain())
++ return 0;
++ for (i = 0; i < machine_e820.nr_map; i++) {
++ const struct e820entry *ei = &machine_e820.map[i];
++#endif
++
++ if (type && ei->type != type)
++ continue;
++ /* is the region (part) in overlap with the current region ?*/
++ if (ei->addr >= end || ei->addr + ei->size <= start)
++ continue;
++
++ /* if the region is at the beginning of <start,end> we move
++ * start to the end of the region since it's ok until there
++ */
++ if (ei->addr <= start)
++ start = ei->addr + ei->size;
++ /* if start is now at or beyond end, we're done, full coverage */
++ if (start >= end)
++ return 1; /* we're done */
++ }
++ return 0;
++}
++
++/*
++ * Find a free area in a specific range.
++ */
++unsigned long __init find_e820_area(unsigned long start, unsigned long end, unsigned size)
++{
++ int i;
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++ unsigned long addr = ei->addr, last;
++ if (ei->type != E820_RAM)
++ continue;
++ if (addr < start)
++ addr = start;
++ if (addr > ei->addr + ei->size)
++ continue;
++ while (bad_addr(&addr, size) && addr+size <= ei->addr+ei->size)
++ ;
++ last = addr + size;
++ if (last > ei->addr + ei->size)
++ continue;
++ if (last > end)
++ continue;
++ return addr;
++ }
++ return -1UL;
++}
++
++/*
++ * Free bootmem based on the e820 table for a node.
++ */
++void __init e820_bootmem_free(pg_data_t *pgdat, unsigned long start,unsigned long end)
++{
++ int i;
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++ unsigned long last, addr;
++
++ if (ei->type != E820_RAM ||
++ ei->addr+ei->size <= start ||
++ ei->addr >= end)
++ continue;
++
++ addr = round_up(ei->addr, PAGE_SIZE);
++ if (addr < start)
++ addr = start;
++
++ last = round_down(ei->addr + ei->size, PAGE_SIZE);
++ if (last >= end)
++ last = end;
++
++ if (last > addr && last-addr >= PAGE_SIZE)
++ free_bootmem_node(pgdat, addr, last-addr);
++ }
++}
++
++/*
++ * Find the highest page frame number we have available
++ */
++unsigned long __init e820_end_of_ram(void)
++{
++ int i;
++ unsigned long end_pfn = 0;
++
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++ unsigned long start, end;
++
++ start = round_up(ei->addr, PAGE_SIZE);
++ end = round_down(ei->addr + ei->size, PAGE_SIZE);
++ if (start >= end)
++ continue;
++ if (ei->type == E820_RAM) {
++ if (end > end_pfn<<PAGE_SHIFT)
++ end_pfn = end>>PAGE_SHIFT;
++ } else {
++ if (end > end_pfn_map<<PAGE_SHIFT)
++ end_pfn_map = end>>PAGE_SHIFT;
++ }
++ }
++
++ if (end_pfn > end_pfn_map)
++ end_pfn_map = end_pfn;
++ if (end_pfn_map > MAXMEM>>PAGE_SHIFT)
++ end_pfn_map = MAXMEM>>PAGE_SHIFT;
++ if (end_pfn > end_user_pfn)
++ end_pfn = end_user_pfn;
++ if (end_pfn > end_pfn_map)
++ end_pfn = end_pfn_map;
++
++ return end_pfn;
++}
++
++/*
++ * Compute how much memory is missing in a range.
++ * Unlike the other functions in this file the arguments are in page numbers.
++ */
++unsigned long __init
++e820_hole_size(unsigned long start_pfn, unsigned long end_pfn)
++{
++ unsigned long ram = 0;
++ unsigned long start = start_pfn << PAGE_SHIFT;
++ unsigned long end = end_pfn << PAGE_SHIFT;
++ int i;
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++ unsigned long last, addr;
++
++ if (ei->type != E820_RAM ||
++ ei->addr+ei->size <= start ||
++ ei->addr >= end)
++ continue;
++
++ addr = round_up(ei->addr, PAGE_SIZE);
++ if (addr < start)
++ addr = start;
++
++ last = round_down(ei->addr + ei->size, PAGE_SIZE);
++ if (last >= end)
++ last = end;
++
++ if (last > addr)
++ ram += last - addr;
++ }
++ return ((end - start) - ram) >> PAGE_SHIFT;
++}
++
++/*
++ * Mark e820 reserved areas as busy for the resource manager.
++ */
++void __init e820_reserve_resources(struct e820entry *e820, int nr_map)
++{
++ int i;
++ for (i = 0; i < nr_map; i++) {
++ struct resource *res;
++ res = alloc_bootmem_low(sizeof(struct resource));
++ switch (e820[i].type) {
++ case E820_RAM: res->name = "System RAM"; break;
++ case E820_ACPI: res->name = "ACPI Tables"; break;
++ case E820_NVS: res->name = "ACPI Non-volatile Storage"; break;
++ default: res->name = "reserved";
++ }
++ res->start = e820[i].addr;
++ res->end = res->start + e820[i].size - 1;
++ res->flags = IORESOURCE_MEM | IORESOURCE_BUSY;
++ request_resource(&iomem_resource, res);
++ if (e820[i].type == E820_RAM) {
++ /*
++ * We don't know which RAM region contains kernel data,
++ * so we try it repeatedly and let the resource manager
++ * test it.
++ */
++#ifndef CONFIG_XEN
++ request_resource(res, &code_resource);
++ request_resource(res, &data_resource);
++#endif
++#ifdef CONFIG_KEXEC
++ if (crashk_res.start != crashk_res.end)
++ request_resource(res, &crashk_res);
++#ifdef CONFIG_XEN
++ xen_machine_kexec_register_resources(res);
++#endif
++#endif
++ }
++ }
++}
++
++/*
++ * Add a memory region to the kernel e820 map.
++ */
++void __init add_memory_region(unsigned long start, unsigned long size, int type)
++{
++ int x = e820.nr_map;
++
++ if (x == E820MAX) {
++ printk(KERN_ERR "Ooops! Too many entries in the memory map!\n");
++ return;
++ }
++
++ e820.map[x].addr = start;
++ e820.map[x].size = size;
++ e820.map[x].type = type;
++ e820.nr_map++;
++}
++
++void __init e820_print_map(char *who)
++{
++ int i;
++
++ for (i = 0; i < e820.nr_map; i++) {
++ printk(" %s: %016Lx - %016Lx ", who,
++ (unsigned long long) e820.map[i].addr,
++ (unsigned long long) (e820.map[i].addr + e820.map[i].size));
++ switch (e820.map[i].type) {
++ case E820_RAM: printk("(usable)\n");
++ break;
++ case E820_RESERVED:
++ printk("(reserved)\n");
++ break;
++ case E820_ACPI:
++ printk("(ACPI data)\n");
++ break;
++ case E820_NVS:
++ printk("(ACPI NVS)\n");
++ break;
++ default: printk("type %u\n", e820.map[i].type);
++ break;
++ }
++ }
++}
++
++/*
++ * Sanitize the BIOS e820 map.
++ *
++ * Some e820 responses include overlapping entries. The following
++ * replaces the original e820 map with a new one, removing overlaps.
++ *
++ */
++static int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map)
++{
++ struct change_member {
++ struct e820entry *pbios; /* pointer to original bios entry */
++ unsigned long long addr; /* address for this change point */
++ };
++ static struct change_member change_point_list[2*E820MAX] __initdata;
++ static struct change_member *change_point[2*E820MAX] __initdata;
++ static struct e820entry *overlap_list[E820MAX] __initdata;
++ static struct e820entry new_bios[E820MAX] __initdata;
++ struct change_member *change_tmp;
++ unsigned long current_type, last_type;
++ unsigned long long last_addr;
++ int chgidx, still_changing;
++ int overlap_entries;
++ int new_bios_entry;
++ int old_nr, new_nr, chg_nr;
++ int i;
++
++ /*
++ Visually we're performing the following (1,2,3,4 = memory types)...
++
++ Sample memory map (w/overlaps):
++ ____22__________________
++ ______________________4_
++ ____1111________________
++ _44_____________________
++ 11111111________________
++ ____________________33__
++ ___________44___________
++ __________33333_________
++ ______________22________
++ ___________________2222_
++ _________111111111______
++ _____________________11_
++ _________________4______
++
++ Sanitized equivalent (no overlap):
++ 1_______________________
++ _44_____________________
++ ___1____________________
++ ____22__________________
++ ______11________________
++ _________1______________
++ __________3_____________
++ ___________44___________
++ _____________33_________
++ _______________2________
++ ________________1_______
++ _________________4______
++ ___________________2____
++ ____________________33__
++ ______________________4_
++ */
++
++ /* if there's only one memory region, don't bother */
++ if (*pnr_map < 2)
++ return -1;
++
++ old_nr = *pnr_map;
++
++ /* bail out if we find any unreasonable addresses in bios map */
++ for (i=0; i<old_nr; i++)
++ if (biosmap[i].addr + biosmap[i].size < biosmap[i].addr)
++ return -1;
++
++ /* create pointers for initial change-point information (for sorting) */
++ for (i=0; i < 2*old_nr; i++)
++ change_point[i] = &change_point_list[i];
++
++ /* record all known change-points (starting and ending addresses),
++ omitting those that are for empty memory regions */
++ chgidx = 0;
++ for (i=0; i < old_nr; i++) {
++ if (biosmap[i].size != 0) {
++ change_point[chgidx]->addr = biosmap[i].addr;
++ change_point[chgidx++]->pbios = &biosmap[i];
++ change_point[chgidx]->addr = biosmap[i].addr + biosmap[i].size;
++ change_point[chgidx++]->pbios = &biosmap[i];
++ }
++ }
++ chg_nr = chgidx;
++
++ /* sort change-point list by memory addresses (low -> high) */
++ still_changing = 1;
++ while (still_changing) {
++ still_changing = 0;
++ for (i=1; i < chg_nr; i++) {
++ /* if <current_addr> > <last_addr>, swap */
++ /* or, if current=<start_addr> & last=<end_addr>, swap */
++ if ((change_point[i]->addr < change_point[i-1]->addr) ||
++ ((change_point[i]->addr == change_point[i-1]->addr) &&
++ (change_point[i]->addr == change_point[i]->pbios->addr) &&
++ (change_point[i-1]->addr != change_point[i-1]->pbios->addr))
++ )
++ {
++ change_tmp = change_point[i];
++ change_point[i] = change_point[i-1];
++ change_point[i-1] = change_tmp;
++ still_changing=1;
++ }
++ }
++ }
++
++ /* create a new bios memory map, removing overlaps */
++ overlap_entries=0; /* number of entries in the overlap table */
++ new_bios_entry=0; /* index for creating new bios map entries */
++ last_type = 0; /* start with undefined memory type */
++ last_addr = 0; /* start with 0 as last starting address */
++ /* loop through change-points, determining affect on the new bios map */
++ for (chgidx=0; chgidx < chg_nr; chgidx++)
++ {
++ /* keep track of all overlapping bios entries */
++ if (change_point[chgidx]->addr == change_point[chgidx]->pbios->addr)
++ {
++ /* add map entry to overlap list (> 1 entry implies an overlap) */
++ overlap_list[overlap_entries++]=change_point[chgidx]->pbios;
++ }
++ else
++ {
++ /* remove entry from list (order independent, so swap with last) */
++ for (i=0; i<overlap_entries; i++)
++ {
++ if (overlap_list[i] == change_point[chgidx]->pbios)
++ overlap_list[i] = overlap_list[overlap_entries-1];
++ }
++ overlap_entries--;
++ }
++ /* if there are overlapping entries, decide which "type" to use */
++ /* (larger value takes precedence -- 1=usable, 2,3,4,4+=unusable) */
++ current_type = 0;
++ for (i=0; i<overlap_entries; i++)
++ if (overlap_list[i]->type > current_type)
++ current_type = overlap_list[i]->type;
++ /* continue building up new bios map based on this information */
++ if (current_type != last_type) {
++ if (last_type != 0) {
++ new_bios[new_bios_entry].size =
++ change_point[chgidx]->addr - last_addr;
++ /* move forward only if the new size was non-zero */
++ if (new_bios[new_bios_entry].size != 0)
++ if (++new_bios_entry >= E820MAX)
++ break; /* no more space left for new bios entries */
++ }
++ if (current_type != 0) {
++ new_bios[new_bios_entry].addr = change_point[chgidx]->addr;
++ new_bios[new_bios_entry].type = current_type;
++ last_addr=change_point[chgidx]->addr;
++ }
++ last_type = current_type;
++ }
++ }
++ new_nr = new_bios_entry; /* retain count for new bios entries */
++
++ /* copy new bios mapping into original location */
++ memcpy(biosmap, new_bios, new_nr*sizeof(struct e820entry));
++ *pnr_map = new_nr;
++
++ return 0;
++}
++
++/*
++ * Copy the BIOS e820 map into a safe place.
++ *
++ * Sanity-check it while we're at it..
++ *
++ * If we're lucky and live on a modern system, the setup code
++ * will have given us a memory map that we can use to properly
++ * set up memory. If we aren't, we'll fake a memory map.
++ *
++ * We check to see that the memory map contains at least 2 elements
++ * before we'll use it, because the detection code in setup.S may
++ * not be perfect and most every PC known to man has two memory
++ * regions: one from 0 to 640k, and one from 1mb up. (The IBM
++ * thinkpad 560x, for example, does not cooperate with the memory
++ * detection code.)
++ */
++static int __init copy_e820_map(struct e820entry * biosmap, int nr_map)
++{
++#ifndef CONFIG_XEN
++ /* Only one memory region (or negative)? Ignore it */
++ if (nr_map < 2)
++ return -1;
++#else
++ BUG_ON(nr_map < 1);
++#endif
++
++ do {
++ unsigned long start = biosmap->addr;
++ unsigned long size = biosmap->size;
++ unsigned long end = start + size;
++ unsigned long type = biosmap->type;
++
++ /* Overflow in 64 bits? Ignore the memory map. */
++ if (start > end)
++ return -1;
++
++#ifndef CONFIG_XEN
++ /*
++ * Some BIOSes claim RAM in the 640k - 1M region.
++ * Not right. Fix it up.
++ *
++ * This should be removed on Hammer which is supposed to not
++ * have non e820 covered ISA mappings there, but I had some strange
++ * problems so it stays for now. -AK
++ */
++ if (type == E820_RAM) {
++ if (start < 0x100000ULL && end > 0xA0000ULL) {
++ if (start < 0xA0000ULL)
++ add_memory_region(start, 0xA0000ULL-start, type);
++ if (end <= 0x100000ULL)
++ continue;
++ start = 0x100000ULL;
++ size = end - start;
++ }
++ }
++#endif
++
++ add_memory_region(start, size, type);
++ } while (biosmap++,--nr_map);
++
++#ifdef CONFIG_XEN
++ if (is_initial_xendomain()) {
++ struct xen_memory_map memmap;
++
++ memmap.nr_entries = E820MAX;
++ set_xen_guest_handle(memmap.buffer, machine_e820.map);
++
++ if (HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap))
++ BUG();
++ machine_e820.nr_map = memmap.nr_entries;
++ } else
++ machine_e820 = e820;
++#endif
++
++ return 0;
++}
++
++#ifndef CONFIG_XEN
++void __init setup_memory_region(void)
++{
++ char *who = "BIOS-e820";
++
++ /*
++ * Try to copy the BIOS-supplied E820-map.
++ *
++ * Otherwise fake a memory map; one section from 0k->640k,
++ * the next section from 1mb->appropriate_mem_k
++ */
++ sanitize_e820_map(E820_MAP, &E820_MAP_NR);
++ if (copy_e820_map(E820_MAP, E820_MAP_NR) < 0) {
++ unsigned long mem_size;
++
++ /* compare results from other methods and take the greater */
++ if (ALT_MEM_K < EXT_MEM_K) {
++ mem_size = EXT_MEM_K;
++ who = "BIOS-88";
++ } else {
++ mem_size = ALT_MEM_K;
++ who = "BIOS-e801";
++ }
++
++ e820.nr_map = 0;
++ add_memory_region(0, LOWMEMSIZE(), E820_RAM);
++ add_memory_region(HIGH_MEMORY, mem_size << 10, E820_RAM);
++ }
++ printk(KERN_INFO "BIOS-provided physical RAM map:\n");
++ e820_print_map(who);
++}
++
++#else /* CONFIG_XEN */
++
++void __init setup_memory_region(void)
++{
++ int rc;
++ struct xen_memory_map memmap;
++ /*
++ * This is rather large for a stack variable but this early in
++ * the boot process we know we have plenty slack space.
++ */
++ struct e820entry map[E820MAX];
++
++ memmap.nr_entries = E820MAX;
++ set_xen_guest_handle(memmap.buffer, map);
++
++ rc = HYPERVISOR_memory_op(XENMEM_memory_map, &memmap);
++ if ( rc == -ENOSYS ) {
++ memmap.nr_entries = 1;
++ map[0].addr = 0ULL;
++ map[0].size = xen_start_info->nr_pages << PAGE_SHIFT;
++ /* 8MB slack (to balance backend allocations). */
++ map[0].size += 8 << 20;
++ map[0].type = E820_RAM;
++ rc = 0;
++ }
++ BUG_ON(rc);
++
++ sanitize_e820_map(map, (char *)&memmap.nr_entries);
++
++ BUG_ON(copy_e820_map(map, (char)memmap.nr_entries) < 0);
++
++ printk(KERN_INFO "BIOS-provided physical RAM map:\n");
++ e820_print_map("Xen");
++}
++#endif
++
++void __init parse_memopt(char *p, char **from)
++{
++ int i;
++ unsigned long current_end;
++ unsigned long end;
++
++ end_user_pfn = memparse(p, from);
++ end_user_pfn >>= PAGE_SHIFT;
++
++ end = end_user_pfn<<PAGE_SHIFT;
++ i = e820.nr_map-1;
++ current_end = e820.map[i].addr + e820.map[i].size;
++
++ if (current_end < end) {
++ /*
++ * The e820 map ends before our requested size so
++ * extend the final entry to the requested address.
++ */
++ if (e820.map[i].type == E820_RAM)
++ e820.map[i].size = end - e820.map[i].addr;
++ else
++ add_memory_region(current_end, end - current_end, E820_RAM);
++ }
++}
++
++void __init parse_memmapopt(char *p, char **from)
++{
++ unsigned long long start_at, mem_size;
++
++ mem_size = memparse(p, from);
++ p = *from;
++ if (*p == '@') {
++ start_at = memparse(p+1, from);
++ add_memory_region(start_at, mem_size, E820_RAM);
++ } else if (*p == '#') {
++ start_at = memparse(p+1, from);
++ add_memory_region(start_at, mem_size, E820_ACPI);
++ } else if (*p == '$') {
++ start_at = memparse(p+1, from);
++ add_memory_region(start_at, mem_size, E820_RESERVED);
++ } else {
++ end_user_pfn = (mem_size >> PAGE_SHIFT);
++ }
++ p = *from;
++}
++
++unsigned long pci_mem_start = 0xaeedbabe;
++EXPORT_SYMBOL(pci_mem_start);
++
++/*
++ * Search for the biggest gap in the low 32 bits of the e820
++ * memory space. We pass this space to PCI to assign MMIO resources
++ * for hotplug or unconfigured devices in.
++ * Hopefully the BIOS let enough space left.
++ */
++__init void e820_setup_gap(struct e820entry *e820, int nr_map)
++{
++ unsigned long gapstart, gapsize, round;
++ unsigned long last;
++ int i;
++ int found = 0;
++
++ last = 0x100000000ull;
++ gapstart = 0x10000000;
++ gapsize = 0x400000;
++ i = nr_map;
++ while (--i >= 0) {
++ unsigned long long start = e820[i].addr;
++ unsigned long long end = start + e820[i].size;
++
++ /*
++ * Since "last" is at most 4GB, we know we'll
++ * fit in 32 bits if this condition is true
++ */
++ if (last > end) {
++ unsigned long gap = last - end;
++
++ if (gap > gapsize) {
++ gapsize = gap;
++ gapstart = end;
++ found = 1;
++ }
++ }
++ if (start < last)
++ last = start;
++ }
++
++ if (!found) {
++ gapstart = (end_pfn << PAGE_SHIFT) + 1024*1024;
++ printk(KERN_ERR "PCI: Warning: Cannot find a gap in the 32bit address range\n"
++ KERN_ERR "PCI: Unassigned devices with 32bit resource registers may break!\n");
++ }
++
++ /*
++ * See how much we want to round up: start off with
++ * rounding to the next 1MB area.
++ */
++ round = 0x100000;
++ while ((gapsize >> 4) > round)
++ round += round;
++ /* Fun with two's complement */
++ pci_mem_start = (gapstart + round) & -round;
++
++ printk(KERN_INFO "Allocating PCI resources starting at %lx (gap: %lx:%lx)\n",
++ pci_mem_start, gapstart, gapsize);
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/e820.c
+--- a/arch/x86_64/kernel/e820.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/kernel/e820.c Wed Sep 10 10:54:08 2008 +0100
+@@ -93,7 +93,7 @@
+ * This function checks if any part of the range <start,end> is mapped
+ * with type.
+ */
+-int __meminit
++int
+ e820_any_mapped(unsigned long start, unsigned long end, unsigned type)
+ {
+ int i;
+@@ -107,6 +107,7 @@
+ }
+ return 0;
+ }
++EXPORT_SYMBOL_GPL(e820_any_mapped);
+
+ /*
+ * This function checks if the entire range <start,end> is mapped with type.
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/early_printk-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/early_printk-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,302 @@
++#include <linux/console.h>
++#include <linux/kernel.h>
++#include <linux/init.h>
++#include <linux/string.h>
++#include <linux/screen_info.h>
++#include <asm/io.h>
++#include <asm/processor.h>
++#include <asm/fcntl.h>
++
++/* Simple VGA output */
++
++#ifdef __i386__
++#include <asm/setup.h>
++#define VGABASE (__ISA_IO_base + 0xb8000)
++#else
++#include <asm/bootsetup.h>
++#define VGABASE ((void __iomem *)0xffffffff800b8000UL)
++#endif
++
++#ifndef CONFIG_XEN
++static int max_ypos = 25, max_xpos = 80;
++static int current_ypos = 25, current_xpos = 0;
++
++static void early_vga_write(struct console *con, const char *str, unsigned n)
++{
++ char c;
++ int i, k, j;
++
++ while ((c = *str++) != '\0' && n-- > 0) {
++ if (current_ypos >= max_ypos) {
++ /* scroll 1 line up */
++ for (k = 1, j = 0; k < max_ypos; k++, j++) {
++ for (i = 0; i < max_xpos; i++) {
++ writew(readw(VGABASE+2*(max_xpos*k+i)),
++ VGABASE + 2*(max_xpos*j + i));
++ }
++ }
++ for (i = 0; i < max_xpos; i++)
++ writew(0x720, VGABASE + 2*(max_xpos*j + i));
++ current_ypos = max_ypos-1;
++ }
++ if (c == '\n') {
++ current_xpos = 0;
++ current_ypos++;
++ } else if (c != '\r') {
++ writew(((0x7 << 8) | (unsigned short) c),
++ VGABASE + 2*(max_xpos*current_ypos +
++ current_xpos++));
++ if (current_xpos >= max_xpos) {
++ current_xpos = 0;
++ current_ypos++;
++ }
++ }
++ }
++}
++
++static struct console early_vga_console = {
++ .name = "earlyvga",
++ .write = early_vga_write,
++ .flags = CON_PRINTBUFFER,
++ .index = -1,
++};
++
++/* Serial functions loosely based on a similar package from Klaus P. Gerlicher */
++
++static int early_serial_base = 0x3f8; /* ttyS0 */
++
++#define XMTRDY 0x20
++
++#define DLAB 0x80
++
++#define TXR 0 /* Transmit register (WRITE) */
++#define RXR 0 /* Receive register (READ) */
++#define IER 1 /* Interrupt Enable */
++#define IIR 2 /* Interrupt ID */
++#define FCR 2 /* FIFO control */
++#define LCR 3 /* Line control */
++#define MCR 4 /* Modem control */
++#define LSR 5 /* Line Status */
++#define MSR 6 /* Modem Status */
++#define DLL 0 /* Divisor Latch Low */
++#define DLH 1 /* Divisor latch High */
++
++static int early_serial_putc(unsigned char ch)
++{
++ unsigned timeout = 0xffff;
++ while ((inb(early_serial_base + LSR) & XMTRDY) == 0 && --timeout)
++ cpu_relax();
++ outb(ch, early_serial_base + TXR);
++ return timeout ? 0 : -1;
++}
++
++static void early_serial_write(struct console *con, const char *s, unsigned n)
++{
++ while (*s && n-- > 0) {
++ early_serial_putc(*s);
++ if (*s == '\n')
++ early_serial_putc('\r');
++ s++;
++ }
++}
++
++#define DEFAULT_BAUD 9600
++
++static __init void early_serial_init(char *s)
++{
++ unsigned char c;
++ unsigned divisor;
++ unsigned baud = DEFAULT_BAUD;
++ char *e;
++
++ if (*s == ',')
++ ++s;
++
++ if (*s) {
++ unsigned port;
++ if (!strncmp(s,"0x",2)) {
++ early_serial_base = simple_strtoul(s, &e, 16);
++ } else {
++ static int bases[] = { 0x3f8, 0x2f8 };
++
++ if (!strncmp(s,"ttyS",4))
++ s += 4;
++ port = simple_strtoul(s, &e, 10);
++ if (port > 1 || s == e)
++ port = 0;
++ early_serial_base = bases[port];
++ }
++ s += strcspn(s, ",");
++ if (*s == ',')
++ s++;
++ }
++
++ outb(0x3, early_serial_base + LCR); /* 8n1 */
++ outb(0, early_serial_base + IER); /* no interrupt */
++ outb(0, early_serial_base + FCR); /* no fifo */
++ outb(0x3, early_serial_base + MCR); /* DTR + RTS */
++
++ if (*s) {
++ baud = simple_strtoul(s, &e, 0);
++ if (baud == 0 || s == e)
++ baud = DEFAULT_BAUD;
++ }
++
++ divisor = 115200 / baud;
++ c = inb(early_serial_base + LCR);
++ outb(c | DLAB, early_serial_base + LCR);
++ outb(divisor & 0xff, early_serial_base + DLL);
++ outb((divisor >> 8) & 0xff, early_serial_base + DLH);
++ outb(c & ~DLAB, early_serial_base + LCR);
++}
++
++#else /* CONFIG_XEN */
++
++static void
++early_serial_write(struct console *con, const char *s, unsigned count)
++{
++ int n;
++
++ while (count > 0) {
++ n = HYPERVISOR_console_io(CONSOLEIO_write, count, (char *)s);
++ if (n <= 0)
++ break;
++ count -= n;
++ s += n;
++ }
++}
++
++static __init void early_serial_init(char *s)
++{
++}
++
++/*
++ * No early VGA console on Xen, as we do not have convenient ISA-space
++ * mappings. Someone should fix this for domain 0. For now, use fake serial.
++ */
++#define early_vga_console early_serial_console
++
++#endif
++
++static struct console early_serial_console = {
++ .name = "earlyser",
++ .write = early_serial_write,
++ .flags = CON_PRINTBUFFER,
++ .index = -1,
++};
++
++/* Console interface to a host file on AMD's SimNow! */
++
++static int simnow_fd;
++
++enum {
++ MAGIC1 = 0xBACCD00A,
++ MAGIC2 = 0xCA110000,
++ XOPEN = 5,
++ XWRITE = 4,
++};
++
++static noinline long simnow(long cmd, long a, long b, long c)
++{
++ long ret;
++ asm volatile("cpuid" :
++ "=a" (ret) :
++ "b" (a), "c" (b), "d" (c), "0" (MAGIC1), "D" (cmd + MAGIC2));
++ return ret;
++}
++
++void __init simnow_init(char *str)
++{
++ char *fn = "klog";
++ if (*str == '=')
++ fn = ++str;
++ /* error ignored */
++ simnow_fd = simnow(XOPEN, (unsigned long)fn, O_WRONLY|O_APPEND|O_CREAT, 0644);
++}
++
++static void simnow_write(struct console *con, const char *s, unsigned n)
++{
++ simnow(XWRITE, simnow_fd, (unsigned long)s, n);
++}
++
++static struct console simnow_console = {
++ .name = "simnow",
++ .write = simnow_write,
++ .flags = CON_PRINTBUFFER,
++ .index = -1,
++};
++
++/* Direct interface for emergencies */
++struct console *early_console = &early_vga_console;
++static int early_console_initialized = 0;
++
++void early_printk(const char *fmt, ...)
++{
++ char buf[512];
++ int n;
++ va_list ap;
++
++ va_start(ap,fmt);
++ n = vscnprintf(buf,512,fmt,ap);
++ early_console->write(early_console,buf,n);
++ va_end(ap);
++}
++
++static int __initdata keep_early;
++
++int __init setup_early_printk(char *opt)
++{
++ char *space;
++ char buf[256];
++
++ if (early_console_initialized)
++ return 1;
++
++ strlcpy(buf,opt,sizeof(buf));
++ space = strchr(buf, ' ');
++ if (space)
++ *space = 0;
++
++ if (strstr(buf,"keep"))
++ keep_early = 1;
++
++ if (!strncmp(buf, "serial", 6)) {
++ early_serial_init(buf + 6);
++ early_console = &early_serial_console;
++ } else if (!strncmp(buf, "ttyS", 4)) {
++ early_serial_init(buf);
++ early_console = &early_serial_console;
++ } else if (!strncmp(buf, "vga", 3)
++#ifndef CONFIG_XEN
++ && SCREEN_INFO.orig_video_isVGA == 1) {
++ max_xpos = SCREEN_INFO.orig_video_cols;
++ max_ypos = SCREEN_INFO.orig_video_lines;
++ current_ypos = SCREEN_INFO.orig_y;
++#else
++ || !strncmp(buf, "xen", 3)) {
++#endif
++ early_console = &early_vga_console;
++ } else if (!strncmp(buf, "simnow", 6)) {
++ simnow_init(buf + 6);
++ early_console = &simnow_console;
++ keep_early = 1;
++ }
++ early_console_initialized = 1;
++ register_console(early_console);
++ return 0;
++}
++
++void __init disable_early_printk(void)
++{
++ if (!early_console_initialized || !early_console)
++ return;
++ if (!keep_early) {
++ printk("disabling early console\n");
++ unregister_console(early_console);
++ early_console_initialized = 0;
++ } else {
++ printk("keeping early console\n");
++ }
++}
++
++__setup("earlyprintk=", setup_early_printk);
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/entry-xen.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/entry-xen.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1322 @@
++/*
++ * linux/arch/x86_64/entry.S
++ *
++ * Copyright (C) 1991, 1992 Linus Torvalds
++ * Copyright (C) 2000, 2001, 2002 Andi Kleen SuSE Labs
++ * Copyright (C) 2000 Pavel Machek <pavel at suse.cz>
++ *
++ * $Id$
++ *
++ * Jun Nakajima <jun.nakajima at intel.com>
++ * Asit Mallick <asit.k.mallick at intel.com>
++ * Modified for Xen
++ */
++
++/*
++ * entry.S contains the system-call and fault low-level handling routines.
++ *
++ * NOTE: This code handles signal-recognition, which happens every time
++ * after an interrupt and after each system call.
++ *
++ * Normal syscalls and interrupts don't save a full stack frame, this is
++ * only done for syscall tracing, signals or fork/exec et.al.
++ *
++ * A note on terminology:
++ * - top of stack: Architecture defined interrupt frame from SS to RIP
++ * at the top of the kernel process stack.
++ * - partial stack frame: partially saved registers upto R11.
++ * - full stack frame: Like partial stack frame, but all register saved.
++ *
++ * TODO:
++ * - schedule it carefully for the final hardware.
++ */
++
++#define ASSEMBLY 1
++#include <linux/linkage.h>
++#include <asm/segment.h>
++#include <asm/smp.h>
++#include <asm/cache.h>
++#include <asm/errno.h>
++#include <asm/dwarf2.h>
++#include <asm/calling.h>
++#include <asm/asm-offsets.h>
++#include <asm/msr.h>
++#include <asm/unistd.h>
++#include <asm/thread_info.h>
++#include <asm/hw_irq.h>
++#include <asm/page.h>
++#include <asm/irqflags.h>
++#include <asm/errno.h>
++#include <xen/interface/arch-x86_64.h>
++#include <xen/interface/features.h>
++
++#include "xen_entry.S"
++
++ .code64
++
++#ifndef CONFIG_PREEMPT
++#define retint_kernel retint_restore_args
++#endif
++
++
++.macro TRACE_IRQS_IRETQ offset=ARGOFFSET
++#ifdef CONFIG_TRACE_IRQFLAGS
++ bt $9,EFLAGS-\offset(%rsp) /* interrupts off? */
++ jnc 1f
++ TRACE_IRQS_ON
++1:
++#endif
++.endm
++
++NMI_MASK = 0x80000000
++
++/*
++ * C code is not supposed to know about undefined top of stack. Every time
++ * a C function with an pt_regs argument is called from the SYSCALL based
++ * fast path FIXUP_TOP_OF_STACK is needed.
++ * RESTORE_TOP_OF_STACK syncs the syscall state after any possible ptregs
++ * manipulation.
++ */
++
++ /* %rsp:at FRAMEEND */
++ .macro FIXUP_TOP_OF_STACK tmp
++ movq $__USER_CS,CS(%rsp)
++ movq $-1,RCX(%rsp)
++ .endm
++
++ .macro RESTORE_TOP_OF_STACK tmp,offset=0
++ .endm
++
++ .macro FAKE_STACK_FRAME child_rip
++ /* push in order ss, rsp, eflags, cs, rip */
++ xorl %eax, %eax
++ pushq %rax /* ss */
++ CFI_ADJUST_CFA_OFFSET 8
++ /*CFI_REL_OFFSET ss,0*/
++ pushq %rax /* rsp */
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rsp,0
++ pushq $(1<<9) /* eflags - interrupts on */
++ CFI_ADJUST_CFA_OFFSET 8
++ /*CFI_REL_OFFSET rflags,0*/
++ pushq $__KERNEL_CS /* cs */
++ CFI_ADJUST_CFA_OFFSET 8
++ /*CFI_REL_OFFSET cs,0*/
++ pushq \child_rip /* rip */
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rip,0
++ pushq %rax /* orig rax */
++ CFI_ADJUST_CFA_OFFSET 8
++ .endm
++
++ .macro UNFAKE_STACK_FRAME
++ addq $8*6, %rsp
++ CFI_ADJUST_CFA_OFFSET -(6*8)
++ .endm
++
++ .macro CFI_DEFAULT_STACK start=1,adj=0
++ .if \start
++ CFI_STARTPROC simple
++ CFI_DEF_CFA rsp,SS+8-(\adj*ARGOFFSET)
++ .else
++ CFI_DEF_CFA_OFFSET SS+8-(\adj*ARGOFFSET)
++ .endif
++ .if \adj == 0
++ CFI_REL_OFFSET r15,R15
++ CFI_REL_OFFSET r14,R14
++ CFI_REL_OFFSET r13,R13
++ CFI_REL_OFFSET r12,R12
++ CFI_REL_OFFSET rbp,RBP
++ CFI_REL_OFFSET rbx,RBX
++ .endif
++ CFI_REL_OFFSET r11,R11
++ CFI_REL_OFFSET r10,R10
++ CFI_REL_OFFSET r9,R9
++ CFI_REL_OFFSET r8,R8
++ CFI_REL_OFFSET rax,RAX
++ CFI_REL_OFFSET rcx,RCX
++ CFI_REL_OFFSET rdx,RDX
++ CFI_REL_OFFSET rsi,RSI
++ CFI_REL_OFFSET rdi,RDI
++ CFI_REL_OFFSET rip,RIP
++ /*CFI_REL_OFFSET cs,CS*/
++ /*CFI_REL_OFFSET rflags,EFLAGS*/
++ CFI_REL_OFFSET rsp,RSP
++ /*CFI_REL_OFFSET ss,SS*/
++ .endm
++
++ /*
++ * Must be consistent with the definition in arch-x86/xen-x86_64.h:
++ * struct iret_context {
++ * u64 rax, r11, rcx, flags, rip, cs, rflags, rsp, ss;
++ * };
++ * with rax, r11, and rcx being taken care of in the hypercall stub.
++ */
++ .macro HYPERVISOR_IRET flag
++ testb $3,1*8(%rsp)
++ jnz 2f
++ testl $NMI_MASK,2*8(%rsp)
++ jnz 2f
++
++ cmpb $0,(xen_features+XENFEAT_supervisor_mode_kernel)(%rip)
++ jne 1f
++
++ /* Direct iret to kernel space. Correct CS and SS. */
++ orl $3,1*8(%rsp)
++ orl $3,4*8(%rsp)
++1: iretq
++
++2: /* Slow iret via hypervisor. */
++ andl $~NMI_MASK, 2*8(%rsp)
++ pushq $\flag
++ jmp hypercall_page + (__HYPERVISOR_iret * 32)
++ .endm
++
++/*
++ * A newly forked process directly context switches into this.
++ */
++/* rdi: prev */
++ENTRY(ret_from_fork)
++ CFI_DEFAULT_STACK
++ push kernel_eflags(%rip)
++ CFI_ADJUST_CFA_OFFSET 4
++ popf # reset kernel eflags
++ CFI_ADJUST_CFA_OFFSET -4
++ call schedule_tail
++ GET_THREAD_INFO(%rcx)
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT),threadinfo_flags(%rcx)
++ jnz rff_trace
++rff_action:
++ RESTORE_REST
++ testl $3,CS-ARGOFFSET(%rsp) # from kernel_thread?
++ je int_ret_from_sys_call
++ testl $_TIF_IA32,threadinfo_flags(%rcx)
++ jnz int_ret_from_sys_call
++ RESTORE_TOP_OF_STACK %rdi,ARGOFFSET
++ jmp ret_from_sys_call
++rff_trace:
++ movq %rsp,%rdi
++ call syscall_trace_leave
++ GET_THREAD_INFO(%rcx)
++ jmp rff_action
++ CFI_ENDPROC
++END(ret_from_fork)
++
++/*
++ * initial frame state for interrupts and exceptions
++ */
++ .macro _frame ref
++ CFI_STARTPROC simple
++ CFI_DEF_CFA rsp,SS+8-\ref
++ /*CFI_REL_OFFSET ss,SS-\ref*/
++ CFI_REL_OFFSET rsp,RSP-\ref
++ /*CFI_REL_OFFSET rflags,EFLAGS-\ref*/
++ /*CFI_REL_OFFSET cs,CS-\ref*/
++ CFI_REL_OFFSET rip,RIP-\ref
++ .endm
++
++/*
++ * System call entry. Upto 6 arguments in registers are supported.
++ *
++ * SYSCALL does not save anything on the stack and does not change the
++ * stack pointer.
++ */
++
++/*
++ * Register setup:
++ * rax system call number
++ * rdi arg0
++ * rcx return address for syscall/sysret, C arg3
++ * rsi arg1
++ * rdx arg2
++ * r10 arg3 (--> moved to rcx for C)
++ * r8 arg4
++ * r9 arg5
++ * r11 eflags for syscall/sysret, temporary for C
++ * r12-r15,rbp,rbx saved by C code, not touched.
++ *
++ * Interrupts are enabled on entry.
++ * Only called from user space.
++ *
++ * XXX if we had a free scratch register we could save the RSP into the stack frame
++ * and report it properly in ps. Unfortunately we haven't.
++ *
++ * When user can change the frames always force IRET. That is because
++ * it deals with uncanonical addresses better. SYSRET has trouble
++ * with them due to bugs in both AMD and Intel CPUs.
++ */
++
++ENTRY(system_call)
++ _frame (RIP-0x10)
++ SAVE_ARGS -8,0
++ movq %rax,ORIG_RAX-ARGOFFSET(%rsp)
++ GET_THREAD_INFO(%rcx)
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%rcx)
++ CFI_REMEMBER_STATE
++ jnz tracesys
++ cmpq $__NR_syscall_max,%rax
++ ja badsys
++ movq %r10,%rcx
++ call *sys_call_table(,%rax,8) # XXX: rip relative
++ movq %rax,RAX-ARGOFFSET(%rsp)
++/*
++ * Syscall return path ending with SYSRET (fast path)
++ * Has incomplete stack frame and undefined top of stack.
++ */
++ .globl ret_from_sys_call
++ret_from_sys_call:
++ movl $_TIF_ALLWORK_MASK,%edi
++ /* edi: flagmask */
++sysret_check:
++ GET_THREAD_INFO(%rcx)
++ XEN_BLOCK_EVENTS(%rsi)
++ TRACE_IRQS_OFF
++ movl threadinfo_flags(%rcx),%edx
++ andl %edi,%edx
++ CFI_REMEMBER_STATE
++ jnz sysret_careful
++ /*
++ * sysretq will re-enable interrupts:
++ */
++ TRACE_IRQS_ON
++ XEN_UNBLOCK_EVENTS(%rsi)
++ RESTORE_ARGS 0,8,0
++ HYPERVISOR_IRET VGCF_IN_SYSCALL
++
++ /* Handle reschedules */
++ /* edx: work, edi: workmask */
++sysret_careful:
++ CFI_RESTORE_STATE
++ bt $TIF_NEED_RESCHED,%edx
++ jnc sysret_signal
++ TRACE_IRQS_ON
++ XEN_UNBLOCK_EVENTS(%rsi)
++ pushq %rdi
++ CFI_ADJUST_CFA_OFFSET 8
++ call schedule
++ popq %rdi
++ CFI_ADJUST_CFA_OFFSET -8
++ jmp sysret_check
++
++ /* Handle a signal */
++sysret_signal:
++ TRACE_IRQS_ON
++/* sti */
++ XEN_UNBLOCK_EVENTS(%rsi)
++ testl $(_TIF_SIGPENDING|_TIF_NOTIFY_RESUME|_TIF_SINGLESTEP),%edx
++ jz 1f
++
++ /* Really a signal */
++ /* edx: work flags (arg3) */
++ leaq do_notify_resume(%rip),%rax
++ leaq -ARGOFFSET(%rsp),%rdi # &pt_regs -> arg1
++ xorl %esi,%esi # oldset -> arg2
++ call ptregscall_common
++1: movl $_TIF_NEED_RESCHED,%edi
++ /* Use IRET because user could have changed frame. This
++ works because ptregscall_common has called FIXUP_TOP_OF_STACK. */
++ XEN_BLOCK_EVENTS(%rsi)
++ TRACE_IRQS_OFF
++ jmp int_with_check
++
++badsys:
++ movq $-ENOSYS,RAX-ARGOFFSET(%rsp)
++ jmp ret_from_sys_call
++
++ /* Do syscall tracing */
++tracesys:
++ CFI_RESTORE_STATE
++ SAVE_REST
++ movq $-ENOSYS,RAX(%rsp)
++ FIXUP_TOP_OF_STACK %rdi
++ movq %rsp,%rdi
++ call syscall_trace_enter
++ LOAD_ARGS ARGOFFSET /* reload args from stack in case ptrace changed it */
++ RESTORE_REST
++ cmpq $__NR_syscall_max,%rax
++ ja int_ret_from_sys_call
++ movq %r10,%rcx /* fixup for C */
++ call *sys_call_table(,%rax,8)
++ movq %rax,RAX-ARGOFFSET(%rsp)
++ /* Use IRET because user could have changed frame */
++ jmp int_ret_from_sys_call
++ CFI_ENDPROC
++END(system_call)
++
++/*
++ * Syscall return path ending with IRET.
++ * Has correct top of stack, but partial stack frame.
++ */
++ENTRY(int_ret_from_sys_call)
++ CFI_STARTPROC simple
++ CFI_DEF_CFA rsp,SS+8-ARGOFFSET
++ /*CFI_REL_OFFSET ss,SS-ARGOFFSET*/
++ CFI_REL_OFFSET rsp,RSP-ARGOFFSET
++ /*CFI_REL_OFFSET rflags,EFLAGS-ARGOFFSET*/
++ /*CFI_REL_OFFSET cs,CS-ARGOFFSET*/
++ CFI_REL_OFFSET rip,RIP-ARGOFFSET
++ CFI_REL_OFFSET rdx,RDX-ARGOFFSET
++ CFI_REL_OFFSET rcx,RCX-ARGOFFSET
++ CFI_REL_OFFSET rax,RAX-ARGOFFSET
++ CFI_REL_OFFSET rdi,RDI-ARGOFFSET
++ CFI_REL_OFFSET rsi,RSI-ARGOFFSET
++ CFI_REL_OFFSET r8,R8-ARGOFFSET
++ CFI_REL_OFFSET r9,R9-ARGOFFSET
++ CFI_REL_OFFSET r10,R10-ARGOFFSET
++ CFI_REL_OFFSET r11,R11-ARGOFFSET
++ XEN_BLOCK_EVENTS(%rsi)
++ TRACE_IRQS_OFF
++ testb $3,CS-ARGOFFSET(%rsp)
++ jnz 1f
++ /* Need to set the proper %ss (not NULL) for ring 3 iretq */
++ movl $__KERNEL_DS,SS-ARGOFFSET(%rsp)
++ jmp retint_restore_args # retrun from ring3 kernel
++1:
++ movl $_TIF_ALLWORK_MASK,%edi
++ /* edi: mask to check */
++int_with_check:
++ GET_THREAD_INFO(%rcx)
++ movl threadinfo_flags(%rcx),%edx
++ andl %edi,%edx
++ jnz int_careful
++ andl $~TS_COMPAT,threadinfo_status(%rcx)
++ jmp retint_restore_args
++
++ /* Either reschedule or signal or syscall exit tracking needed. */
++ /* First do a reschedule test. */
++ /* edx: work, edi: workmask */
++int_careful:
++ bt $TIF_NEED_RESCHED,%edx
++ jnc int_very_careful
++ TRACE_IRQS_ON
++/* sti */
++ XEN_UNBLOCK_EVENTS(%rsi)
++ pushq %rdi
++ CFI_ADJUST_CFA_OFFSET 8
++ call schedule
++ popq %rdi
++ CFI_ADJUST_CFA_OFFSET -8
++ XEN_BLOCK_EVENTS(%rsi)
++ TRACE_IRQS_OFF
++ jmp int_with_check
++
++ /* handle signals and tracing -- both require a full stack frame */
++int_very_careful:
++ TRACE_IRQS_ON
++/* sti */
++ XEN_UNBLOCK_EVENTS(%rsi)
++ SAVE_REST
++ /* Check for syscall exit trace */
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP),%edx
++ jz int_signal
++ pushq %rdi
++ CFI_ADJUST_CFA_OFFSET 8
++ leaq 8(%rsp),%rdi # &ptregs -> arg1
++ call syscall_trace_leave
++ popq %rdi
++ CFI_ADJUST_CFA_OFFSET -8
++ andl $~(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP),%edi
++ XEN_BLOCK_EVENTS(%rsi)
++ TRACE_IRQS_OFF
++ jmp int_restore_rest
++
++int_signal:
++ testl $(_TIF_NOTIFY_RESUME|_TIF_SIGPENDING|_TIF_SINGLESTEP),%edx
++ jz 1f
++ movq %rsp,%rdi # &ptregs -> arg1
++ xorl %esi,%esi # oldset -> arg2
++ call do_notify_resume
++1: movl $_TIF_NEED_RESCHED,%edi
++int_restore_rest:
++ RESTORE_REST
++ XEN_BLOCK_EVENTS(%rsi)
++ TRACE_IRQS_OFF
++ jmp int_with_check
++ CFI_ENDPROC
++END(int_ret_from_sys_call)
++
++/*
++ * Certain special system calls that need to save a complete full stack frame.
++ */
++
++ .macro PTREGSCALL label,func,arg
++ .globl \label
++\label:
++ leaq \func(%rip),%rax
++ leaq -ARGOFFSET+8(%rsp),\arg /* 8 for return address */
++ jmp ptregscall_common
++END(\label)
++ .endm
++
++ CFI_STARTPROC
++
++ PTREGSCALL stub_clone, sys_clone, %r8
++ PTREGSCALL stub_fork, sys_fork, %rdi
++ PTREGSCALL stub_vfork, sys_vfork, %rdi
++ PTREGSCALL stub_rt_sigsuspend, sys_rt_sigsuspend, %rdx
++ PTREGSCALL stub_sigaltstack, sys_sigaltstack, %rdx
++ PTREGSCALL stub_iopl, sys_iopl, %rsi
++
++ENTRY(ptregscall_common)
++ popq %r11
++ CFI_ADJUST_CFA_OFFSET -8
++ CFI_REGISTER rip, r11
++ SAVE_REST
++ movq %r11, %r15
++ CFI_REGISTER rip, r15
++ FIXUP_TOP_OF_STACK %r11
++ call *%rax
++ RESTORE_TOP_OF_STACK %r11
++ movq %r15, %r11
++ CFI_REGISTER rip, r11
++ RESTORE_REST
++ pushq %r11
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rip, 0
++ ret
++ CFI_ENDPROC
++END(ptregscall_common)
++
++ENTRY(stub_execve)
++ CFI_STARTPROC
++ popq %r11
++ CFI_ADJUST_CFA_OFFSET -8
++ CFI_REGISTER rip, r11
++ SAVE_REST
++ FIXUP_TOP_OF_STACK %r11
++ call sys_execve
++ RESTORE_TOP_OF_STACK %r11
++ movq %rax,RAX(%rsp)
++ RESTORE_REST
++ jmp int_ret_from_sys_call
++ CFI_ENDPROC
++END(stub_execve)
++
++/*
++ * sigreturn is special because it needs to restore all registers on return.
++ * This cannot be done with SYSRET, so use the IRET return path instead.
++ */
++ENTRY(stub_rt_sigreturn)
++ CFI_STARTPROC
++ addq $8, %rsp
++ CFI_ADJUST_CFA_OFFSET -8
++ SAVE_REST
++ movq %rsp,%rdi
++ FIXUP_TOP_OF_STACK %r11
++ call sys_rt_sigreturn
++ movq %rax,RAX(%rsp) # fixme, this could be done at the higher layer
++ RESTORE_REST
++ jmp int_ret_from_sys_call
++ CFI_ENDPROC
++END(stub_rt_sigreturn)
++
++/* initial frame state for interrupts (and exceptions without error code) */
++#define INTR_FRAME _frame (RIP-0x10); \
++ CFI_REL_OFFSET rcx,0; \
++ CFI_REL_OFFSET r11,8
++
++/* initial frame state for exceptions with error code (and interrupts with
++ vector already pushed) */
++#define XCPT_FRAME _frame (RIP-0x18); \
++ CFI_REL_OFFSET rcx,0; \
++ CFI_REL_OFFSET r11,8
++
++/*
++ * Interrupt exit.
++ *
++ */
++
++retint_check:
++ CFI_DEFAULT_STACK adj=1
++ movl threadinfo_flags(%rcx),%edx
++ andl %edi,%edx
++ CFI_REMEMBER_STATE
++ jnz retint_careful
++retint_restore_args:
++ movl EFLAGS-REST_SKIP(%rsp), %eax
++ shr $9, %eax # EAX[0] == IRET_EFLAGS.IF
++ XEN_GET_VCPU_INFO(%rsi)
++ andb evtchn_upcall_mask(%rsi),%al
++ andb $1,%al # EAX[0] == IRET_EFLAGS.IF & event_mask
++ jnz restore_all_enable_events # != 0 => enable event delivery
++ XEN_PUT_VCPU_INFO(%rsi)
++
++ RESTORE_ARGS 0,8,0
++ HYPERVISOR_IRET 0
++
++ /* edi: workmask, edx: work */
++retint_careful:
++ CFI_RESTORE_STATE
++ bt $TIF_NEED_RESCHED,%edx
++ jnc retint_signal
++ TRACE_IRQS_ON
++ XEN_UNBLOCK_EVENTS(%rsi)
++/* sti */
++ pushq %rdi
++ CFI_ADJUST_CFA_OFFSET 8
++ call schedule
++ popq %rdi
++ CFI_ADJUST_CFA_OFFSET -8
++ GET_THREAD_INFO(%rcx)
++ XEN_BLOCK_EVENTS(%rsi)
++/* cli */
++ TRACE_IRQS_OFF
++ jmp retint_check
++
++retint_signal:
++ testl $(_TIF_SIGPENDING|_TIF_NOTIFY_RESUME|_TIF_SINGLESTEP),%edx
++ jz retint_restore_args
++ TRACE_IRQS_ON
++ XEN_UNBLOCK_EVENTS(%rsi)
++ SAVE_REST
++ movq $-1,ORIG_RAX(%rsp)
++ xorl %esi,%esi # oldset
++ movq %rsp,%rdi # &pt_regs
++ call do_notify_resume
++ RESTORE_REST
++ XEN_BLOCK_EVENTS(%rsi)
++ TRACE_IRQS_OFF
++ movl $_TIF_NEED_RESCHED,%edi
++ GET_THREAD_INFO(%rcx)
++ jmp retint_check
++
++#ifdef CONFIG_PREEMPT
++ /* Returning to kernel space. Check if we need preemption */
++ /* rcx: threadinfo. interrupts off. */
++ .p2align
++retint_kernel:
++ cmpl $0,threadinfo_preempt_count(%rcx)
++ jnz retint_restore_args
++ bt $TIF_NEED_RESCHED,threadinfo_flags(%rcx)
++ jnc retint_restore_args
++ bt $9,EFLAGS-ARGOFFSET(%rsp) /* interrupts off? */
++ jnc retint_restore_args
++ call preempt_schedule_irq
++ jmp retint_kernel /* check again */
++#endif
++
++ CFI_ENDPROC
++END(retint_check)
++
++#ifndef CONFIG_XEN
++/*
++ * APIC interrupts.
++ */
++ .macro apicinterrupt num,func
++ INTR_FRAME
++ pushq $~(\num)
++ CFI_ADJUST_CFA_OFFSET 8
++ interrupt \func
++ jmp error_entry
++ CFI_ENDPROC
++ .endm
++
++ENTRY(thermal_interrupt)
++ apicinterrupt THERMAL_APIC_VECTOR,smp_thermal_interrupt
++END(thermal_interrupt)
++
++ENTRY(threshold_interrupt)
++ apicinterrupt THRESHOLD_APIC_VECTOR,mce_threshold_interrupt
++END(threshold_interrupt)
++
++#ifdef CONFIG_SMP
++ENTRY(reschedule_interrupt)
++ apicinterrupt RESCHEDULE_VECTOR,smp_reschedule_interrupt
++END(reschedule_interrupt)
++
++ .macro INVALIDATE_ENTRY num
++ENTRY(invalidate_interrupt\num)
++ apicinterrupt INVALIDATE_TLB_VECTOR_START+\num,smp_invalidate_interrupt
++END(invalidate_interrupt\num)
++ .endm
++
++ INVALIDATE_ENTRY 0
++ INVALIDATE_ENTRY 1
++ INVALIDATE_ENTRY 2
++ INVALIDATE_ENTRY 3
++ INVALIDATE_ENTRY 4
++ INVALIDATE_ENTRY 5
++ INVALIDATE_ENTRY 6
++ INVALIDATE_ENTRY 7
++
++ENTRY(call_function_interrupt)
++ apicinterrupt CALL_FUNCTION_VECTOR,smp_call_function_interrupt
++END(call_function_interrupt)
++#endif
++
++#ifdef CONFIG_X86_LOCAL_APIC
++ENTRY(apic_timer_interrupt)
++ apicinterrupt LOCAL_TIMER_VECTOR,smp_apic_timer_interrupt
++END(apic_timer_interrupt)
++
++ENTRY(error_interrupt)
++ apicinterrupt ERROR_APIC_VECTOR,smp_error_interrupt
++END(error_interrupt)
++
++ENTRY(spurious_interrupt)
++ apicinterrupt SPURIOUS_APIC_VECTOR,smp_spurious_interrupt
++END(spurious_interrupt)
++#endif
++#endif /* !CONFIG_XEN */
++
++/*
++ * Exception entry points.
++ */
++ .macro zeroentry sym
++ INTR_FRAME
++ movq (%rsp),%rcx
++ CFI_RESTORE rcx
++ movq 8(%rsp),%r11
++ CFI_RESTORE r11
++ addq $0x10,%rsp /* skip rcx and r11 */
++ CFI_ADJUST_CFA_OFFSET -0x10
++ pushq $0 /* push error code/oldrax */
++ CFI_ADJUST_CFA_OFFSET 8
++ pushq %rax /* push real oldrax to the rdi slot */
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rax,0
++ leaq \sym(%rip),%rax
++ jmp error_entry
++ CFI_ENDPROC
++ .endm
++
++ .macro errorentry sym
++ XCPT_FRAME
++ movq (%rsp),%rcx
++ CFI_RESTORE rcx
++ movq 8(%rsp),%r11
++ CFI_RESTORE r11
++ addq $0x10,%rsp /* rsp points to the error code */
++ CFI_ADJUST_CFA_OFFSET -0x10
++ pushq %rax
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rax,0
++ leaq \sym(%rip),%rax
++ jmp error_entry
++ CFI_ENDPROC
++ .endm
++
++#if 0 /* not XEN */
++ /* error code is on the stack already */
++ /* handle NMI like exceptions that can happen everywhere */
++ .macro paranoidentry sym, ist=0, irqtrace=1
++ movq (%rsp),%rcx
++ movq 8(%rsp),%r11
++ addq $0x10,%rsp /* skip rcx and r11 */
++ SAVE_ALL
++ cld
++#if 0 /* not XEN */
++ movl $1,%ebx
++ movl $MSR_GS_BASE,%ecx
++ rdmsr
++ testl %edx,%edx
++ js 1f
++ swapgs
++ xorl %ebx,%ebx
++1:
++#endif
++ .if \ist
++ movq %gs:pda_data_offset, %rbp
++ .endif
++ movq %rsp,%rdi
++ movq ORIG_RAX(%rsp),%rsi
++ movq $-1,ORIG_RAX(%rsp)
++ .if \ist
++ subq $EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp)
++ .endif
++ call \sym
++ .if \ist
++ addq $EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp)
++ .endif
++/* cli */
++ XEN_BLOCK_EVENTS(%rsi)
++ .if \irqtrace
++ TRACE_IRQS_OFF
++ .endif
++ .endm
++
++ /*
++ * "Paranoid" exit path from exception stack.
++ * Paranoid because this is used by NMIs and cannot take
++ * any kernel state for granted.
++ * We don't do kernel preemption checks here, because only
++ * NMI should be common and it does not enable IRQs and
++ * cannot get reschedule ticks.
++ *
++ * "trace" is 0 for the NMI handler only, because irq-tracing
++ * is fundamentally NMI-unsafe. (we cannot change the soft and
++ * hard flags at once, atomically)
++ */
++ .macro paranoidexit trace=1
++ /* ebx: no swapgs flag */
++paranoid_exit\trace:
++ testl %ebx,%ebx /* swapgs needed? */
++ jnz paranoid_restore\trace
++ testl $3,CS(%rsp)
++ jnz paranoid_userspace\trace
++paranoid_swapgs\trace:
++ TRACE_IRQS_IRETQ 0
++ swapgs
++paranoid_restore\trace:
++ RESTORE_ALL 8
++ iretq
++paranoid_userspace\trace:
++ GET_THREAD_INFO(%rcx)
++ movl threadinfo_flags(%rcx),%ebx
++ andl $_TIF_WORK_MASK,%ebx
++ jz paranoid_swapgs\trace
++ movq %rsp,%rdi /* &pt_regs */
++ call sync_regs
++ movq %rax,%rsp /* switch stack for scheduling */
++ testl $_TIF_NEED_RESCHED,%ebx
++ jnz paranoid_schedule\trace
++ movl %ebx,%edx /* arg3: thread flags */
++ .if \trace
++ TRACE_IRQS_ON
++ .endif
++ sti
++ xorl %esi,%esi /* arg2: oldset */
++ movq %rsp,%rdi /* arg1: &pt_regs */
++ call do_notify_resume
++ cli
++ .if \trace
++ TRACE_IRQS_OFF
++ .endif
++ jmp paranoid_userspace\trace
++paranoid_schedule\trace:
++ .if \trace
++ TRACE_IRQS_ON
++ .endif
++ sti
++ call schedule
++ cli
++ .if \trace
++ TRACE_IRQS_OFF
++ .endif
++ jmp paranoid_userspace\trace
++ CFI_ENDPROC
++ .endm
++#endif
++
++/*
++ * Exception entry point. This expects an error code/orig_rax on the stack
++ * and the exception handler in %rax.
++ */
++ENTRY(error_entry)
++ _frame RDI
++ CFI_REL_OFFSET rax,0
++ /* rdi slot contains rax, oldrax contains error code */
++ cld
++ subq $14*8,%rsp
++ CFI_ADJUST_CFA_OFFSET (14*8)
++ movq %rsi,13*8(%rsp)
++ CFI_REL_OFFSET rsi,RSI
++ movq 14*8(%rsp),%rsi /* load rax from rdi slot */
++ CFI_REGISTER rax,rsi
++ movq %rdx,12*8(%rsp)
++ CFI_REL_OFFSET rdx,RDX
++ movq %rcx,11*8(%rsp)
++ CFI_REL_OFFSET rcx,RCX
++ movq %rsi,10*8(%rsp) /* store rax */
++ CFI_REL_OFFSET rax,RAX
++ movq %r8, 9*8(%rsp)
++ CFI_REL_OFFSET r8,R8
++ movq %r9, 8*8(%rsp)
++ CFI_REL_OFFSET r9,R9
++ movq %r10,7*8(%rsp)
++ CFI_REL_OFFSET r10,R10
++ movq %r11,6*8(%rsp)
++ CFI_REL_OFFSET r11,R11
++ movq %rbx,5*8(%rsp)
++ CFI_REL_OFFSET rbx,RBX
++ movq %rbp,4*8(%rsp)
++ CFI_REL_OFFSET rbp,RBP
++ movq %r12,3*8(%rsp)
++ CFI_REL_OFFSET r12,R12
++ movq %r13,2*8(%rsp)
++ CFI_REL_OFFSET r13,R13
++ movq %r14,1*8(%rsp)
++ CFI_REL_OFFSET r14,R14
++ movq %r15,(%rsp)
++ CFI_REL_OFFSET r15,R15
++#if 0
++ cmpl $__KERNEL_CS,CS(%rsp)
++ CFI_REMEMBER_STATE
++ je error_kernelspace
++#endif
++error_call_handler:
++ movq %rdi, RDI(%rsp)
++ CFI_REL_OFFSET rdi,RDI
++ movq %rsp,%rdi
++ movq ORIG_RAX(%rsp),%rsi # get error code
++ movq $-1,ORIG_RAX(%rsp)
++ call *%rax
++error_exit:
++ RESTORE_REST
++/* cli */
++ XEN_BLOCK_EVENTS(%rsi)
++ TRACE_IRQS_OFF
++ GET_THREAD_INFO(%rcx)
++ testb $3,CS-ARGOFFSET(%rsp)
++ jz retint_kernel
++ movl threadinfo_flags(%rcx),%edx
++ movl $_TIF_WORK_MASK,%edi
++ andl %edi,%edx
++ jnz retint_careful
++ /*
++ * The iret might restore flags:
++ */
++ TRACE_IRQS_IRETQ
++ jmp retint_restore_args
++
++#if 0
++ /*
++ * We need to re-write the logic here because we don't do iretq to
++ * to return to user mode. It's still possible that we get trap/fault
++ * in the kernel (when accessing buffers pointed to by system calls,
++ * for example).
++ *
++ */
++ CFI_RESTORE_STATE
++error_kernelspace:
++ incl %ebx
++ /* There are two places in the kernel that can potentially fault with
++ usergs. Handle them here. The exception handlers after
++ iret run with kernel gs again, so don't set the user space flag.
++ B stepping K8s sometimes report an truncated RIP for IRET
++ exceptions returning to compat mode. Check for these here too. */
++ leaq iret_label(%rip),%rbp
++ cmpq %rbp,RIP(%rsp)
++ je error_swapgs
++ movl %ebp,%ebp /* zero extend */
++ cmpq %rbp,RIP(%rsp)
++ je error_swapgs
++ cmpq $gs_change,RIP(%rsp)
++ je error_swapgs
++ jmp error_sti
++#endif
++ CFI_ENDPROC
++END(error_entry)
++
++ENTRY(hypervisor_callback)
++ zeroentry do_hypervisor_callback
++END(hypervisor_callback)
++
++/*
++ * Copied from arch/xen/i386/kernel/entry.S
++ */
++# A note on the "critical region" in our callback handler.
++# We want to avoid stacking callback handlers due to events occurring
++# during handling of the last event. To do this, we keep events disabled
++# until we've done all processing. HOWEVER, we must enable events before
++# popping the stack frame (can't be done atomically) and so it would still
++# be possible to get enough handler activations to overflow the stack.
++# Although unlikely, bugs of that kind are hard to track down, so we'd
++# like to avoid the possibility.
++# So, on entry to the handler we detect whether we interrupted an
++# existing activation in its critical region -- if so, we pop the current
++# activation and restart the handler using the previous one.
++ENTRY(do_hypervisor_callback) # do_hypervisor_callback(struct *pt_regs)
++ CFI_STARTPROC
++# Since we don't modify %rdi, evtchn_do_upall(struct *pt_regs) will
++# see the correct pointer to the pt_regs
++ movq %rdi, %rsp # we don't return, adjust the stack frame
++ CFI_ENDPROC
++ CFI_DEFAULT_STACK
++11: incl %gs:pda_irqcount
++ movq %rsp,%rbp
++ CFI_DEF_CFA_REGISTER rbp
++ cmovzq %gs:pda_irqstackptr,%rsp
++ pushq %rbp # backlink for old unwinder
++ call evtchn_do_upcall
++ popq %rsp
++ CFI_DEF_CFA_REGISTER rsp
++ decl %gs:pda_irqcount
++ jmp error_exit
++ CFI_ENDPROC
++END(do_hypervisor_callback)
++
++#ifdef CONFIG_X86_LOCAL_APIC
++KPROBE_ENTRY(nmi)
++ zeroentry do_nmi_callback
++ENTRY(do_nmi_callback)
++ CFI_STARTPROC
++ addq $8, %rsp
++ CFI_ENDPROC
++ CFI_DEFAULT_STACK
++ call do_nmi
++ orl $NMI_MASK,EFLAGS(%rsp)
++ RESTORE_REST
++ XEN_BLOCK_EVENTS(%rsi)
++ TRACE_IRQS_OFF
++ GET_THREAD_INFO(%rcx)
++ jmp retint_restore_args
++ CFI_ENDPROC
++ .previous .text
++END(nmi)
++#endif
++
++ ALIGN
++restore_all_enable_events:
++ CFI_DEFAULT_STACK adj=1
++ TRACE_IRQS_ON
++ XEN_UNBLOCK_EVENTS(%rsi) # %rsi is already set up...
++
++scrit: /**** START OF CRITICAL REGION ****/
++ XEN_TEST_PENDING(%rsi)
++ CFI_REMEMBER_STATE
++ jnz 14f # process more events if necessary...
++ XEN_PUT_VCPU_INFO(%rsi)
++ RESTORE_ARGS 0,8,0
++ HYPERVISOR_IRET 0
++
++ CFI_RESTORE_STATE
++14: XEN_LOCKED_BLOCK_EVENTS(%rsi)
++ XEN_PUT_VCPU_INFO(%rsi)
++ SAVE_REST
++ movq %rsp,%rdi # set the argument again
++ jmp 11b
++ CFI_ENDPROC
++ecrit: /**** END OF CRITICAL REGION ****/
++# At this point, unlike on x86-32, we don't do the fixup to simplify the
++# code and the stack frame is more complex on x86-64.
++# When the kernel is interrupted in the critical section, the kernel
++# will do IRET in that case, and everything will be restored at that point,
++# i.e. it just resumes from the next instruction interrupted with the same context.
++
++# Hypervisor uses this for application faults while it executes.
++# We get here for two reasons:
++# 1. Fault while reloading DS, ES, FS or GS
++# 2. Fault while executing IRET
++# Category 1 we do not need to fix up as Xen has already reloaded all segment
++# registers that could be reloaded and zeroed the others.
++# Category 2 we fix up by killing the current process. We cannot use the
++# normal Linux return path in this case because if we use the IRET hypercall
++# to pop the stack frame we end up in an infinite loop of failsafe callbacks.
++# We distinguish between categories by comparing each saved segment register
++# with its current contents: any discrepancy means we in category 1.
++ENTRY(failsafe_callback)
++ _frame (RIP-0x30)
++ CFI_REL_OFFSET rcx, 0
++ CFI_REL_OFFSET r11, 8
++ movw %ds,%cx
++ cmpw %cx,0x10(%rsp)
++ CFI_REMEMBER_STATE
++ jne 1f
++ movw %es,%cx
++ cmpw %cx,0x18(%rsp)
++ jne 1f
++ movw %fs,%cx
++ cmpw %cx,0x20(%rsp)
++ jne 1f
++ movw %gs,%cx
++ cmpw %cx,0x28(%rsp)
++ jne 1f
++ /* All segments match their saved values => Category 2 (Bad IRET). */
++ movq (%rsp),%rcx
++ CFI_RESTORE rcx
++ movq 8(%rsp),%r11
++ CFI_RESTORE r11
++ addq $0x30,%rsp
++ CFI_ADJUST_CFA_OFFSET -0x30
++ movq $11,%rdi /* SIGSEGV */
++ jmp do_exit
++ CFI_RESTORE_STATE
++1: /* Segment mismatch => Category 1 (Bad segment). Retry the IRET. */
++ movq (%rsp),%rcx
++ CFI_RESTORE rcx
++ movq 8(%rsp),%r11
++ CFI_RESTORE r11
++ addq $0x30,%rsp
++ CFI_ADJUST_CFA_OFFSET -0x30
++ pushq $0
++ CFI_ADJUST_CFA_OFFSET 8
++ SAVE_ALL
++ jmp error_exit
++ CFI_ENDPROC
++#if 0
++ .section __ex_table,"a"
++ .align 8
++ .quad gs_change,bad_gs
++ .previous
++ .section .fixup,"ax"
++ /* running with kernelgs */
++bad_gs:
++/* swapgs */ /* switch back to user gs */
++ xorl %eax,%eax
++ movl %eax,%gs
++ jmp 2b
++ .previous
++#endif
++
++/*
++ * Create a kernel thread.
++ *
++ * C extern interface:
++ * extern long kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
++ *
++ * asm input arguments:
++ * rdi: fn, rsi: arg, rdx: flags
++ */
++ENTRY(kernel_thread)
++ CFI_STARTPROC
++ FAKE_STACK_FRAME $child_rip
++ SAVE_ALL
++
++ # rdi: flags, rsi: usp, rdx: will be &pt_regs
++ movq %rdx,%rdi
++ orq kernel_thread_flags(%rip),%rdi
++ movq $-1, %rsi
++ movq %rsp, %rdx
++
++ xorl %r8d,%r8d
++ xorl %r9d,%r9d
++
++ # clone now
++ call do_fork
++ movq %rax,RAX(%rsp)
++ xorl %edi,%edi
++
++ /*
++ * It isn't worth to check for reschedule here,
++ * so internally to the x86_64 port you can rely on kernel_thread()
++ * not to reschedule the child before returning, this avoids the need
++ * of hacks for example to fork off the per-CPU idle tasks.
++ * [Hopefully no generic code relies on the reschedule -AK]
++ */
++ RESTORE_ALL
++ UNFAKE_STACK_FRAME
++ ret
++ CFI_ENDPROC
++ENDPROC(kernel_thread)
++
++child_rip:
++ pushq $0 # fake return address
++ CFI_STARTPROC
++ /*
++ * Here we are in the child and the registers are set as they were
++ * at kernel_thread() invocation in the parent.
++ */
++ movq %rdi, %rax
++ movq %rsi, %rdi
++ call *%rax
++ # exit
++ xorl %edi, %edi
++ call do_exit
++ CFI_ENDPROC
++ENDPROC(child_rip)
++
++/*
++ * execve(). This function needs to use IRET, not SYSRET, to set up all state properly.
++ *
++ * C extern interface:
++ * extern long execve(char *name, char **argv, char **envp)
++ *
++ * asm input arguments:
++ * rdi: name, rsi: argv, rdx: envp
++ *
++ * We want to fallback into:
++ * extern long sys_execve(char *name, char **argv,char **envp, struct pt_regs regs)
++ *
++ * do_sys_execve asm fallback arguments:
++ * rdi: name, rsi: argv, rdx: envp, fake frame on the stack
++ */
++ENTRY(execve)
++ CFI_STARTPROC
++ FAKE_STACK_FRAME $0
++ SAVE_ALL
++ call sys_execve
++ movq %rax, RAX(%rsp)
++ RESTORE_REST
++ testq %rax,%rax
++ jne 1f
++ jmp int_ret_from_sys_call
++1: RESTORE_ARGS
++ UNFAKE_STACK_FRAME
++ ret
++ CFI_ENDPROC
++ENDPROC(execve)
++
++KPROBE_ENTRY(page_fault)
++ errorentry do_page_fault
++END(page_fault)
++ .previous .text
++
++ENTRY(coprocessor_error)
++ zeroentry do_coprocessor_error
++END(coprocessor_error)
++
++ENTRY(simd_coprocessor_error)
++ zeroentry do_simd_coprocessor_error
++END(simd_coprocessor_error)
++
++ENTRY(device_not_available)
++ zeroentry math_state_restore
++END(device_not_available)
++
++ /* runs on exception stack */
++KPROBE_ENTRY(debug)
++/* INTR_FRAME
++ pushq $0
++ CFI_ADJUST_CFA_OFFSET 8 */
++ zeroentry do_debug
++/* paranoidexit
++ CFI_ENDPROC */
++END(debug)
++ .previous .text
++
++#if 0
++ /* runs on exception stack */
++KPROBE_ENTRY(nmi)
++ INTR_FRAME
++ pushq $-1
++ CFI_ADJUST_CFA_OFFSET 8
++ paranoidentry do_nmi, 0, 0
++#ifdef CONFIG_TRACE_IRQFLAGS
++ paranoidexit 0
++#else
++ jmp paranoid_exit1
++ CFI_ENDPROC
++#endif
++END(nmi)
++ .previous .text
++#endif
++
++KPROBE_ENTRY(int3)
++/* INTR_FRAME
++ pushq $0
++ CFI_ADJUST_CFA_OFFSET 8 */
++ zeroentry do_int3
++/* jmp paranoid_exit1
++ CFI_ENDPROC */
++END(int3)
++ .previous .text
++
++ENTRY(overflow)
++ zeroentry do_overflow
++END(overflow)
++
++ENTRY(bounds)
++ zeroentry do_bounds
++END(bounds)
++
++ENTRY(invalid_op)
++ zeroentry do_invalid_op
++END(invalid_op)
++
++ENTRY(coprocessor_segment_overrun)
++ zeroentry do_coprocessor_segment_overrun
++END(coprocessor_segment_overrun)
++
++ENTRY(reserved)
++ zeroentry do_reserved
++END(reserved)
++
++#if 0
++ /* runs on exception stack */
++ENTRY(double_fault)
++ XCPT_FRAME
++ paranoidentry do_double_fault
++ jmp paranoid_exit1
++ CFI_ENDPROC
++END(double_fault)
++#endif
++
++ENTRY(invalid_TSS)
++ errorentry do_invalid_TSS
++END(invalid_TSS)
++
++ENTRY(segment_not_present)
++ errorentry do_segment_not_present
++END(segment_not_present)
++
++ /* runs on exception stack */
++ENTRY(stack_segment)
++/* XCPT_FRAME
++ paranoidentry do_stack_segment */
++ errorentry do_stack_segment
++/* jmp paranoid_exit1
++ CFI_ENDPROC */
++END(stack_segment)
++
++KPROBE_ENTRY(general_protection)
++ errorentry do_general_protection
++END(general_protection)
++ .previous .text
++
++ENTRY(alignment_check)
++ errorentry do_alignment_check
++END(alignment_check)
++
++ENTRY(divide_error)
++ zeroentry do_divide_error
++END(divide_error)
++
++ENTRY(spurious_interrupt_bug)
++ zeroentry do_spurious_interrupt_bug
++END(spurious_interrupt_bug)
++
++#ifdef CONFIG_X86_MCE
++ /* runs on exception stack */
++ENTRY(machine_check)
++ INTR_FRAME
++ pushq $0
++ CFI_ADJUST_CFA_OFFSET 8
++ paranoidentry do_machine_check
++ jmp paranoid_exit1
++ CFI_ENDPROC
++END(machine_check)
++#endif
++
++/* Call softirq on interrupt stack. Interrupts are off. */
++ENTRY(call_softirq)
++ CFI_STARTPROC
++ push %rbp
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rbp,0
++ mov %rsp,%rbp
++ CFI_DEF_CFA_REGISTER rbp
++ incl %gs:pda_irqcount
++ cmove %gs:pda_irqstackptr,%rsp
++ push %rbp # backlink for old unwinder
++ call __do_softirq
++ leaveq
++ CFI_DEF_CFA_REGISTER rsp
++ CFI_ADJUST_CFA_OFFSET -8
++ decl %gs:pda_irqcount
++ ret
++ CFI_ENDPROC
++ENDPROC(call_softirq)
++
++#ifdef CONFIG_STACK_UNWIND
++ENTRY(arch_unwind_init_running)
++ CFI_STARTPROC
++ movq %r15, R15(%rdi)
++ movq %r14, R14(%rdi)
++ xchgq %rsi, %rdx
++ movq %r13, R13(%rdi)
++ movq %r12, R12(%rdi)
++ xorl %eax, %eax
++ movq %rbp, RBP(%rdi)
++ movq %rbx, RBX(%rdi)
++ movq (%rsp), %rcx
++ movq %rax, R11(%rdi)
++ movq %rax, R10(%rdi)
++ movq %rax, R9(%rdi)
++ movq %rax, R8(%rdi)
++ movq %rax, RAX(%rdi)
++ movq %rax, RCX(%rdi)
++ movq %rax, RDX(%rdi)
++ movq %rax, RSI(%rdi)
++ movq %rax, RDI(%rdi)
++ movq %rax, ORIG_RAX(%rdi)
++ movq %rcx, RIP(%rdi)
++ leaq 8(%rsp), %rcx
++ movq $__KERNEL_CS, CS(%rdi)
++ movq %rax, EFLAGS(%rdi)
++ movq %rcx, RSP(%rdi)
++ movq $__KERNEL_DS, SS(%rdi)
++ jmpq *%rdx
++ CFI_ENDPROC
++ENDPROC(arch_unwind_init_running)
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/genapic-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/genapic-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,143 @@
++/*
++ * Copyright 2004 James Cleverdon, IBM.
++ * Subject to the GNU Public License, v.2
++ *
++ * Generic APIC sub-arch probe layer.
++ *
++ * Hacked for x86-64 by James Cleverdon from i386 architecture code by
++ * Martin Bligh, Andi Kleen, James Bottomley, John Stultz, and
++ * James Cleverdon.
++ */
++#include <linux/threads.h>
++#include <linux/cpumask.h>
++#include <linux/string.h>
++#include <linux/kernel.h>
++#include <linux/ctype.h>
++#include <linux/init.h>
++#include <linux/module.h>
++
++#include <asm/smp.h>
++#include <asm/ipi.h>
++
++#if defined(CONFIG_ACPI)
++#include <acpi/acpi_bus.h>
++#endif
++
++/* which logical CPU number maps to which CPU (physical APIC ID) */
++u8 x86_cpu_to_apicid[NR_CPUS] __read_mostly = { [0 ... NR_CPUS-1] = BAD_APICID };
++EXPORT_SYMBOL(x86_cpu_to_apicid);
++u8 x86_cpu_to_log_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = BAD_APICID };
++
++extern struct genapic apic_cluster;
++extern struct genapic apic_flat;
++extern struct genapic apic_physflat;
++
++#ifndef CONFIG_XEN
++struct genapic *genapic = &apic_flat;
++#else
++extern struct genapic apic_xen;
++struct genapic *genapic = &apic_xen;
++#endif
++
++
++/*
++ * Check the APIC IDs in bios_cpu_apicid and choose the APIC mode.
++ */
++void __init clustered_apic_check(void)
++{
++#ifndef CONFIG_XEN
++ long i;
++ u8 clusters, max_cluster;
++ u8 id;
++ u8 cluster_cnt[NUM_APIC_CLUSTERS];
++ int max_apic = 0;
++
++#if defined(CONFIG_ACPI)
++ /*
++ * Some x86_64 machines use physical APIC mode regardless of how many
++ * procs/clusters are present (x86_64 ES7000 is an example).
++ */
++ if (acpi_fadt.revision > FADT2_REVISION_ID)
++ if (acpi_fadt.force_apic_physical_destination_mode) {
++ genapic = &apic_cluster;
++ goto print;
++ }
++#endif
++
++ memset(cluster_cnt, 0, sizeof(cluster_cnt));
++ for (i = 0; i < NR_CPUS; i++) {
++ id = bios_cpu_apicid[i];
++ if (id == BAD_APICID)
++ continue;
++ if (id > max_apic)
++ max_apic = id;
++ cluster_cnt[APIC_CLUSTERID(id)]++;
++ }
++
++ /* Don't use clustered mode on AMD platforms. */
++ if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD) {
++ genapic = &apic_physflat;
++#ifndef CONFIG_HOTPLUG_CPU
++ /* In the CPU hotplug case we cannot use broadcast mode
++ because that opens a race when a CPU is removed.
++ Stay at physflat mode in this case.
++ It is bad to do this unconditionally though. Once
++ we have ACPI platform support for CPU hotplug
++ we should detect hotplug capablity from ACPI tables and
++ only do this when really needed. -AK */
++ if (max_apic <= 8)
++ genapic = &apic_flat;
++#endif
++ goto print;
++ }
++
++ clusters = 0;
++ max_cluster = 0;
++
++ for (i = 0; i < NUM_APIC_CLUSTERS; i++) {
++ if (cluster_cnt[i] > 0) {
++ ++clusters;
++ if (cluster_cnt[i] > max_cluster)
++ max_cluster = cluster_cnt[i];
++ }
++ }
++
++ /*
++ * If we have clusters <= 1 and CPUs <= 8 in cluster 0, then flat mode,
++ * else if max_cluster <= 4 and cluster_cnt[15] == 0, clustered logical
++ * else physical mode.
++ * (We don't use lowest priority delivery + HW APIC IRQ steering, so
++ * can ignore the clustered logical case and go straight to physical.)
++ */
++ if (clusters <= 1 && max_cluster <= 8 && cluster_cnt[0] == max_cluster) {
++#ifdef CONFIG_HOTPLUG_CPU
++ /* Don't use APIC shortcuts in CPU hotplug to avoid races */
++ genapic = &apic_physflat;
++#else
++ genapic = &apic_flat;
++#endif
++ } else
++ genapic = &apic_cluster;
++
++print:
++#else
++ /* hardcode to xen apic functions */
++ genapic = &apic_xen;
++#endif
++ printk(KERN_INFO "Setting APIC routing to %s\n", genapic->name);
++}
++
++/* Same for both flat and clustered. */
++
++#ifdef CONFIG_XEN
++extern void xen_send_IPI_shortcut(unsigned int shortcut, int vector, unsigned int dest);
++#endif
++
++void send_IPI_self(int vector)
++{
++#ifndef CONFIG_XEN
++ __send_IPI_shortcut(APIC_DEST_SELF, vector, APIC_DEST_PHYSICAL);
++#else
++ xen_send_IPI_shortcut(APIC_DEST_SELF, vector, APIC_DEST_PHYSICAL);
++#endif
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/genapic_xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/genapic_xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,161 @@
++/*
++ * Copyright 2004 James Cleverdon, IBM.
++ * Subject to the GNU Public License, v.2
++ *
++ * Xen APIC subarch code. Maximum 8 CPUs, logical delivery.
++ *
++ * Hacked for x86-64 by James Cleverdon from i386 architecture code by
++ * Martin Bligh, Andi Kleen, James Bottomley, John Stultz, and
++ * James Cleverdon.
++ *
++ * Hacked to pieces for Xen by Chris Wright.
++ */
++#include <linux/threads.h>
++#include <linux/cpumask.h>
++#include <linux/string.h>
++#include <linux/kernel.h>
++#include <linux/ctype.h>
++#include <linux/init.h>
++#ifdef CONFIG_XEN_PRIVILEGED_GUEST
++#include <asm/smp.h>
++#include <asm/ipi.h>
++#else
++#include <asm/apic.h>
++#include <asm/apicdef.h>
++#include <asm/genapic.h>
++#endif
++#include <xen/evtchn.h>
++
++DECLARE_PER_CPU(int, ipi_to_irq[NR_IPIS]);
++
++static inline void __send_IPI_one(unsigned int cpu, int vector)
++{
++ int irq = per_cpu(ipi_to_irq, cpu)[vector];
++ BUG_ON(irq < 0);
++ notify_remote_via_irq(irq);
++}
++
++void xen_send_IPI_shortcut(unsigned int shortcut, int vector, unsigned int dest)
++{
++ int cpu;
++
++ switch (shortcut) {
++ case APIC_DEST_SELF:
++ __send_IPI_one(smp_processor_id(), vector);
++ break;
++ case APIC_DEST_ALLBUT:
++ for (cpu = 0; cpu < NR_CPUS; ++cpu) {
++ if (cpu == smp_processor_id())
++ continue;
++ if (cpu_isset(cpu, cpu_online_map)) {
++ __send_IPI_one(cpu, vector);
++ }
++ }
++ break;
++ case APIC_DEST_ALLINC:
++ for (cpu = 0; cpu < NR_CPUS; ++cpu) {
++ if (cpu_isset(cpu, cpu_online_map)) {
++ __send_IPI_one(cpu, vector);
++ }
++ }
++ break;
++ default:
++ printk("XXXXXX __send_IPI_shortcut %08x vector %d\n", shortcut,
++ vector);
++ break;
++ }
++}
++
++static cpumask_t xen_target_cpus(void)
++{
++ return cpu_online_map;
++}
++
++/*
++ * Set up the logical destination ID.
++ * Do nothing, not called now.
++ */
++static void xen_init_apic_ldr(void)
++{
++ Dprintk("%s\n", __FUNCTION__);
++ return;
++}
++
++static void xen_send_IPI_allbutself(int vector)
++{
++ /*
++ * if there are no other CPUs in the system then
++ * we get an APIC send error if we try to broadcast.
++ * thus we have to avoid sending IPIs in this case.
++ */
++ Dprintk("%s\n", __FUNCTION__);
++ if (num_online_cpus() > 1)
++ xen_send_IPI_shortcut(APIC_DEST_ALLBUT, vector, APIC_DEST_LOGICAL);
++}
++
++static void xen_send_IPI_all(int vector)
++{
++ Dprintk("%s\n", __FUNCTION__);
++ xen_send_IPI_shortcut(APIC_DEST_ALLINC, vector, APIC_DEST_LOGICAL);
++}
++
++static void xen_send_IPI_mask(cpumask_t cpumask, int vector)
++{
++ unsigned long mask = cpus_addr(cpumask)[0];
++ unsigned int cpu;
++ unsigned long flags;
++
++ Dprintk("%s\n", __FUNCTION__);
++ local_irq_save(flags);
++ WARN_ON(mask & ~cpus_addr(cpu_online_map)[0]);
++
++ for (cpu = 0; cpu < NR_CPUS; ++cpu) {
++ if (cpu_isset(cpu, cpumask)) {
++ __send_IPI_one(cpu, vector);
++ }
++ }
++ local_irq_restore(flags);
++}
++
++#ifdef CONFIG_XEN_PRIVILEGED_GUEST
++static int xen_apic_id_registered(void)
++{
++ /* better be set */
++ Dprintk("%s\n", __FUNCTION__);
++ return physid_isset(smp_processor_id(), phys_cpu_present_map);
++}
++#endif
++
++static unsigned int xen_cpu_mask_to_apicid(cpumask_t cpumask)
++{
++ Dprintk("%s\n", __FUNCTION__);
++ return cpus_addr(cpumask)[0] & APIC_ALL_CPUS;
++}
++
++static unsigned int phys_pkg_id(int index_msb)
++{
++ u32 ebx;
++
++ Dprintk("%s\n", __FUNCTION__);
++ ebx = cpuid_ebx(1);
++ return ((ebx >> 24) & 0xFF) >> index_msb;
++}
++
++struct genapic apic_xen = {
++ .name = "xen",
++#ifdef CONFIG_XEN_PRIVILEGED_GUEST
++ .int_delivery_mode = dest_LowestPrio,
++#endif
++ .int_dest_mode = (APIC_DEST_LOGICAL != 0),
++ .int_delivery_dest = APIC_DEST_LOGICAL | APIC_DM_LOWEST,
++ .target_cpus = xen_target_cpus,
++#ifdef CONFIG_XEN_PRIVILEGED_GUEST
++ .apic_id_registered = xen_apic_id_registered,
++#endif
++ .init_apic_ldr = xen_init_apic_ldr,
++ .send_IPI_all = xen_send_IPI_all,
++ .send_IPI_allbutself = xen_send_IPI_allbutself,
++ .send_IPI_mask = xen_send_IPI_mask,
++ .cpu_mask_to_apicid = xen_cpu_mask_to_apicid,
++ .phys_pkg_id = phys_pkg_id,
++};
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/head-xen.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/head-xen.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,214 @@
++/*
++ * linux/arch/x86_64/kernel/head.S -- start in 32bit and switch to 64bit
++ *
++ * Copyright (C) 2000 Andrea Arcangeli <andrea at suse.de> SuSE
++ * Copyright (C) 2000 Pavel Machek <pavel at suse.cz>
++ * Copyright (C) 2000 Karsten Keil <kkeil at suse.de>
++ * Copyright (C) 2001,2002 Andi Kleen <ak at suse.de>
++ *
++ * $Id: head.S,v 1.49 2002/03/19 17:39:25 ak Exp $
++ *
++ * Jun Nakajima <jun.nakajima at intel.com>
++ * Modified for Xen
++ */
++
++
++#include <linux/linkage.h>
++#include <linux/threads.h>
++#include <linux/init.h>
++#include <linux/elfnote.h>
++#include <asm/desc.h>
++#include <asm/segment.h>
++#include <asm/page.h>
++#include <asm/msr.h>
++#include <asm/cache.h>
++#include <asm/dwarf2.h>
++#include <xen/interface/elfnote.h>
++
++ .section .bootstrap.text, "ax", @progbits
++ .code64
++ .globl startup_64
++startup_64:
++ movq $(init_thread_union+THREAD_SIZE-8),%rsp
++
++ /* rsi is pointer to startup info structure.
++ pass it to C */
++ movq %rsi,%rdi
++ pushq $0 # fake return address
++ jmp x86_64_start_kernel
++
++#ifdef CONFIG_ACPI_SLEEP
++.org 0xf00
++ .globl pGDT32
++pGDT32:
++ .word gdt_end-cpu_gdt_table-1
++ .long cpu_gdt_table-__START_KERNEL_map
++#endif
++ENTRY(stext)
++ENTRY(_stext)
++
++ $page = 0
++#define NEXT_PAGE(name) \
++ $page = $page + 1; \
++ .org $page * 0x1000; \
++ phys_##name = $page * 0x1000 + __PHYSICAL_START; \
++ENTRY(name)
++
++NEXT_PAGE(init_level4_pgt)
++ /* This gets initialized in x86_64_start_kernel */
++ .fill 512,8,0
++NEXT_PAGE(init_level4_user_pgt)
++ /*
++ * We update two pgd entries to make kernel and user pgd consistent
++ * at pgd_populate(). It can be used for kernel modules. So we place
++ * this page here for those cases to avoid memory corruption.
++ * We also use this page to establish the initial mapping for the
++ * vsyscall area.
++ */
++ .fill 512,8,0
++
++NEXT_PAGE(level3_kernel_pgt)
++ .fill 512,8,0
++
++ /*
++ * This is used for vsyscall area mapping as we have a different
++ * level4 page table for user.
++ */
++NEXT_PAGE(level3_user_pgt)
++ .fill 512,8,0
++
++NEXT_PAGE(level2_kernel_pgt)
++ .fill 512,8,0
++
++NEXT_PAGE(hypercall_page)
++ CFI_STARTPROC
++ .rept 0x1000 / 0x20
++ .skip 1 /* push %rcx */
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rcx,0
++ .skip 2 /* push %r11 */
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rcx,0
++ .skip 5 /* mov $#,%eax */
++ .skip 2 /* syscall */
++ .skip 2 /* pop %r11 */
++ CFI_ADJUST_CFA_OFFSET -8
++ CFI_RESTORE r11
++ .skip 1 /* pop %rcx */
++ CFI_ADJUST_CFA_OFFSET -8
++ CFI_RESTORE rcx
++ .align 0x20,0 /* ret */
++ .endr
++ CFI_ENDPROC
++
++#undef NEXT_PAGE
++
++ .data
++/* Just dummy symbol to allow compilation. Not used in sleep path */
++#ifdef CONFIG_ACPI_SLEEP
++ .align PAGE_SIZE
++ENTRY(wakeup_level4_pgt)
++ .fill 512,8,0
++#endif
++
++ .data
++
++ .align 16
++ .globl cpu_gdt_descr
++cpu_gdt_descr:
++ .word gdt_end-cpu_gdt_table-1
++gdt:
++ .quad cpu_gdt_table
++#ifdef CONFIG_SMP
++ .rept NR_CPUS-1
++ .word 0
++ .quad 0
++ .endr
++#endif
++
++/* We need valid kernel segments for data and code in long mode too
++ * IRET will check the segment types kkeil 2000/10/28
++ * Also sysret mandates a special GDT layout
++ */
++
++ .section .data.page_aligned, "aw"
++ .align PAGE_SIZE
++
++/* The TLS descriptors are currently at a different place compared to i386.
++ Hopefully nobody expects them at a fixed place (Wine?) */
++
++ENTRY(cpu_gdt_table)
++ .quad 0x0000000000000000 /* NULL descriptor */
++ .quad 0x0 /* unused */
++ .quad 0x00af9a000000ffff /* __KERNEL_CS */
++ .quad 0x00cf92000000ffff /* __KERNEL_DS */
++ .quad 0x00cffa000000ffff /* __USER32_CS */
++ .quad 0x00cff2000000ffff /* __USER_DS, __USER32_DS */
++ .quad 0x00affa000000ffff /* __USER_CS */
++ .quad 0x00cf9a000000ffff /* __KERNEL32_CS */
++ .quad 0,0 /* TSS */
++ .quad 0,0 /* LDT */
++ .quad 0,0,0 /* three TLS descriptors */
++ .quad 0 /* unused */
++gdt_end:
++ /* asm/segment.h:GDT_ENTRIES must match this */
++ /* This should be a multiple of the cache line size */
++ /* GDTs of other CPUs are now dynamically allocated */
++
++ /* zero the remaining page */
++ .fill PAGE_SIZE / 8 - GDT_ENTRIES,8,0
++
++ .section .bss.page_aligned, "aw", @nobits
++ .align PAGE_SIZE
++ENTRY(empty_zero_page)
++ .skip PAGE_SIZE
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++/*
++ * __xen_guest information
++ */
++.macro utoh value
++ .if (\value) < 0 || (\value) >= 0x10
++ utoh (((\value)>>4)&0x0fffffffffffffff)
++ .endif
++ .if ((\value) & 0xf) < 10
++ .byte '0' + ((\value) & 0xf)
++ .else
++ .byte 'A' + ((\value) & 0xf) - 10
++ .endif
++.endm
++
++.section __xen_guest
++ .ascii "GUEST_OS=linux,GUEST_VER=2.6"
++ .ascii ",XEN_VER=xen-3.0"
++ .ascii ",VIRT_BASE=0x"
++ utoh __START_KERNEL_map
++ .ascii ",ELF_PADDR_OFFSET=0x"
++ utoh __START_KERNEL_map
++ .ascii ",VIRT_ENTRY=0x"
++ utoh (__START_KERNEL_map + __PHYSICAL_START)
++ .ascii ",HYPERCALL_PAGE=0x"
++ utoh (phys_hypercall_page >> PAGE_SHIFT)
++ .ascii ",FEATURES=writable_page_tables"
++ .ascii "|writable_descriptor_tables"
++ .ascii "|auto_translated_physmap"
++ .ascii "|supervisor_mode_kernel"
++ .ascii ",LOADER=generic"
++ .byte 0
++#endif /* CONFIG_XEN_COMPAT <= 0x030002 */
++
++ ELFNOTE(Xen, XEN_ELFNOTE_GUEST_OS, .asciz, "linux")
++ ELFNOTE(Xen, XEN_ELFNOTE_GUEST_VERSION, .asciz, "2.6")
++ ELFNOTE(Xen, XEN_ELFNOTE_XEN_VERSION, .asciz, "xen-3.0")
++ ELFNOTE(Xen, XEN_ELFNOTE_VIRT_BASE, .quad, __START_KERNEL_map)
++#if CONFIG_XEN_COMPAT <= 0x030002
++ ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET, .quad, __START_KERNEL_map)
++#else
++ ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET, .quad, 0)
++#endif
++ ELFNOTE(Xen, XEN_ELFNOTE_ENTRY, .quad, startup_64)
++ ELFNOTE(Xen, XEN_ELFNOTE_HYPERCALL_PAGE, .quad, hypercall_page)
++ ELFNOTE(Xen, XEN_ELFNOTE_L1_MFN_VALID, .quad, _PAGE_PRESENT,_PAGE_PRESENT)
++ ELFNOTE(Xen, XEN_ELFNOTE_FEATURES, .asciz, "writable_page_tables|writable_descriptor_tables|auto_translated_physmap|pae_pgdir_above_4gb|supervisor_mode_kernel")
++ ELFNOTE(Xen, XEN_ELFNOTE_LOADER, .asciz, "generic")
++ ELFNOTE(Xen, XEN_ELFNOTE_SUSPEND_CANCEL, .long, 1)
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/head64-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/head64-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,162 @@
++/*
++ * linux/arch/x86_64/kernel/head64.c -- prepare to run common code
++ *
++ * Copyright (C) 2000 Andrea Arcangeli <andrea at suse.de> SuSE
++ *
++ * Jun Nakajima <jun.nakajima at intel.com>
++ * Modified for Xen.
++ */
++
++#include <linux/init.h>
++#include <linux/linkage.h>
++#include <linux/types.h>
++#include <linux/kernel.h>
++#include <linux/string.h>
++#include <linux/percpu.h>
++#include <linux/module.h>
++
++#include <asm/processor.h>
++#include <asm/proto.h>
++#include <asm/smp.h>
++#include <asm/bootsetup.h>
++#include <asm/setup.h>
++#include <asm/desc.h>
++#include <asm/pgtable.h>
++#include <asm/sections.h>
++
++unsigned long start_pfn;
++
++/* Don't add a printk in there. printk relies on the PDA which is not initialized
++ yet. */
++#if 0
++static void __init clear_bss(void)
++{
++ memset(__bss_start, 0,
++ (unsigned long) __bss_stop - (unsigned long) __bss_start);
++}
++#endif
++
++#define NEW_CL_POINTER 0x228 /* Relative to real mode data */
++#define OLD_CL_MAGIC_ADDR 0x90020
++#define OLD_CL_MAGIC 0xA33F
++#define OLD_CL_BASE_ADDR 0x90000
++#define OLD_CL_OFFSET 0x90022
++
++extern char saved_command_line[];
++
++static void __init copy_bootdata(char *real_mode_data)
++{
++#ifndef CONFIG_XEN
++ int new_data;
++ char * command_line;
++
++ memcpy(x86_boot_params, real_mode_data, BOOT_PARAM_SIZE);
++ new_data = *(int *) (x86_boot_params + NEW_CL_POINTER);
++ if (!new_data) {
++ if (OLD_CL_MAGIC != * (u16 *) OLD_CL_MAGIC_ADDR) {
++ printk("so old bootloader that it does not support commandline?!\n");
++ return;
++ }
++ new_data = OLD_CL_BASE_ADDR + * (u16 *) OLD_CL_OFFSET;
++ printk("old bootloader convention, maybe loadlin?\n");
++ }
++ command_line = (char *) ((u64)(new_data));
++ memcpy(saved_command_line, command_line, COMMAND_LINE_SIZE);
++#else
++ int max_cmdline;
++
++ if ((max_cmdline = MAX_GUEST_CMDLINE) > COMMAND_LINE_SIZE)
++ max_cmdline = COMMAND_LINE_SIZE;
++ memcpy(saved_command_line, xen_start_info->cmd_line, max_cmdline);
++ saved_command_line[max_cmdline-1] = '\0';
++#endif
++ printk("Bootdata ok (command line is %s)\n", saved_command_line);
++}
++
++static void __init setup_boot_cpu_data(void)
++{
++ unsigned int dummy, eax;
++
++ /* get vendor info */
++ cpuid(0, (unsigned int *)&boot_cpu_data.cpuid_level,
++ (unsigned int *)&boot_cpu_data.x86_vendor_id[0],
++ (unsigned int *)&boot_cpu_data.x86_vendor_id[8],
++ (unsigned int *)&boot_cpu_data.x86_vendor_id[4]);
++
++ /* get cpu type */
++ cpuid(1, &eax, &dummy, &dummy,
++ (unsigned int *) &boot_cpu_data.x86_capability);
++ boot_cpu_data.x86 = (eax >> 8) & 0xf;
++ boot_cpu_data.x86_model = (eax >> 4) & 0xf;
++ boot_cpu_data.x86_mask = eax & 0xf;
++}
++
++#include <xen/interface/memory.h>
++unsigned long *machine_to_phys_mapping;
++EXPORT_SYMBOL(machine_to_phys_mapping);
++unsigned int machine_to_phys_order;
++EXPORT_SYMBOL(machine_to_phys_order);
++
++void __init x86_64_start_kernel(char * real_mode_data)
++{
++ struct xen_machphys_mapping mapping;
++ unsigned long machine_to_phys_nr_ents;
++ char *s;
++ int i;
++
++ setup_xen_features();
++
++ xen_start_info = (struct start_info *)real_mode_data;
++ if (!xen_feature(XENFEAT_auto_translated_physmap))
++ phys_to_machine_mapping =
++ (unsigned long *)xen_start_info->mfn_list;
++ start_pfn = (__pa(xen_start_info->pt_base) >> PAGE_SHIFT) +
++ xen_start_info->nr_pt_frames;
++
++ machine_to_phys_mapping = (unsigned long *)MACH2PHYS_VIRT_START;
++ machine_to_phys_nr_ents = MACH2PHYS_NR_ENTRIES;
++ if (HYPERVISOR_memory_op(XENMEM_machphys_mapping, &mapping) == 0) {
++ machine_to_phys_mapping = (unsigned long *)mapping.v_start;
++ machine_to_phys_nr_ents = mapping.max_mfn + 1;
++ }
++ while ((1UL << machine_to_phys_order) < machine_to_phys_nr_ents )
++ machine_to_phys_order++;
++
++#if 0
++ for (i = 0; i < 256; i++)
++ set_intr_gate(i, early_idt_handler);
++ asm volatile("lidt %0" :: "m" (idt_descr));
++#endif
++
++ /*
++ * This must be called really, really early:
++ */
++ lockdep_init();
++
++ for (i = 0; i < NR_CPUS; i++)
++ cpu_pda(i) = &boot_cpu_pda[i];
++
++ pda_init(0);
++ copy_bootdata(real_mode_data);
++#ifdef CONFIG_SMP
++ cpu_set(0, cpu_online_map);
++#endif
++ s = strstr(saved_command_line, "earlyprintk=");
++ if (s != NULL)
++ setup_early_printk(strchr(s, '=') + 1);
++#ifdef CONFIG_NUMA
++ s = strstr(saved_command_line, "numa=");
++ if (s != NULL)
++ numa_setup(s+5);
++#endif
++#ifdef CONFIG_X86_IO_APIC
++ if (strstr(saved_command_line, "disableapic"))
++ disable_apic = 1;
++#endif
++ /* You need early console to see that */
++ if (__pa_symbol(&_end) >= KERNEL_TEXT_SIZE)
++ panic("Kernel too big for kernel mapping\n");
++
++ setup_boot_cpu_data();
++ start_kernel();
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/init_task.c
+--- a/arch/x86_64/kernel/init_task.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/kernel/init_task.c Wed Sep 10 10:54:08 2008 +0100
+@@ -37,6 +37,8 @@
+ struct task_struct init_task = INIT_TASK(init_task);
+
+ EXPORT_SYMBOL(init_task);
++
++#ifndef CONFIG_X86_NO_TSS
+ /*
+ * per-CPU TSS segments. Threads are completely 'soft' on Linux,
+ * no more per-task TSS's. The TSS size is kept cacheline-aligned
+@@ -50,5 +52,6 @@
+ * debugging, no special alignment required.
+ */
+ DEFINE_PER_CPU(struct orig_ist, orig_ist);
++#endif
+
+ #define ALIGN_TO_4K __attribute__((section(".data.init_task")))
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/io_apic-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/io_apic-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,2263 @@
++/*
++ * Intel IO-APIC support for multi-Pentium hosts.
++ *
++ * Copyright (C) 1997, 1998, 1999, 2000 Ingo Molnar, Hajnalka Szabo
++ *
++ * Many thanks to Stig Venaas for trying out countless experimental
++ * patches and reporting/debugging problems patiently!
++ *
++ * (c) 1999, Multiple IO-APIC support, developed by
++ * Ken-ichi Yaku <yaku at css1.kbnes.nec.co.jp> and
++ * Hidemi Kishimoto <kisimoto at css1.kbnes.nec.co.jp>,
++ * further tested and cleaned up by Zach Brown <zab at redhat.com>
++ * and Ingo Molnar <mingo at redhat.com>
++ *
++ * Fixes
++ * Maciej W. Rozycki : Bits for genuine 82489DX APICs;
++ * thanks to Eric Gilmore
++ * and Rolf G. Tews
++ * for testing these extensively
++ * Paul Diefenbaugh : Added full ACPI support
++ */
++
++#include <linux/mm.h>
++#include <linux/interrupt.h>
++#include <linux/init.h>
++#include <linux/delay.h>
++#include <linux/sched.h>
++#include <linux/smp_lock.h>
++#include <linux/mc146818rtc.h>
++#include <linux/acpi.h>
++#include <linux/sysdev.h>
++#ifdef CONFIG_ACPI
++#include <acpi/acpi_bus.h>
++#endif
++
++#include <asm/io.h>
++#include <asm/smp.h>
++#include <asm/desc.h>
++#include <asm/proto.h>
++#include <asm/mach_apic.h>
++#include <asm/acpi.h>
++#include <asm/dma.h>
++#include <asm/nmi.h>
++
++#define __apicdebuginit __init
++
++int sis_apic_bug; /* not actually supported, dummy for compile */
++
++static int no_timer_check;
++
++int disable_timer_pin_1 __initdata;
++
++#ifndef CONFIG_XEN
++int timer_over_8254 __initdata = 0;
++
++/* Where if anywhere is the i8259 connect in external int mode */
++static struct { int pin, apic; } ioapic_i8259 = { -1, -1 };
++#endif
++
++static DEFINE_SPINLOCK(ioapic_lock);
++static DEFINE_SPINLOCK(vector_lock);
++
++/*
++ * # of IRQ routing registers
++ */
++int nr_ioapic_registers[MAX_IO_APICS];
++
++/*
++ * Rough estimation of how many shared IRQs there are, can
++ * be changed anytime.
++ */
++#define MAX_PLUS_SHARED_IRQS NR_IRQ_VECTORS
++#define PIN_MAP_SIZE (MAX_PLUS_SHARED_IRQS + NR_IRQS)
++
++/*
++ * This is performance-critical, we want to do it O(1)
++ *
++ * the indexing order of this array favors 1:1 mappings
++ * between pins and IRQs.
++ */
++
++static struct irq_pin_list {
++ short apic, pin, next;
++} irq_2_pin[PIN_MAP_SIZE];
++
++int vector_irq[NR_VECTORS] __read_mostly = { [0 ... NR_VECTORS - 1] = -1};
++#ifdef CONFIG_PCI_MSI
++#define vector_to_irq(vector) \
++ (platform_legacy_irq(vector) ? vector : vector_irq[vector])
++#else
++#define vector_to_irq(vector) (vector)
++#endif
++
++#ifdef CONFIG_XEN
++
++#include <xen/interface/xen.h>
++#include <xen/interface/physdev.h>
++#include <xen/evtchn.h>
++
++/* Fake i8259 */
++#define make_8259A_irq(_irq) (io_apic_irqs &= ~(1UL<<(_irq)))
++#define disable_8259A_irq(_irq) ((void)0)
++#define i8259A_irq_pending(_irq) (0)
++
++unsigned long io_apic_irqs;
++
++static inline unsigned int xen_io_apic_read(unsigned int apic, unsigned int reg)
++{
++ struct physdev_apic apic_op;
++ int ret;
++
++ apic_op.apic_physbase = mp_ioapics[apic].mpc_apicaddr;
++ apic_op.reg = reg;
++ ret = HYPERVISOR_physdev_op(PHYSDEVOP_apic_read, &apic_op);
++ if (ret)
++ return ret;
++ return apic_op.value;
++}
++
++static inline void xen_io_apic_write(unsigned int apic, unsigned int reg, unsigned int value)
++{
++ struct physdev_apic apic_op;
++
++ apic_op.apic_physbase = mp_ioapics[apic].mpc_apicaddr;
++ apic_op.reg = reg;
++ apic_op.value = value;
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_apic_write, &apic_op));
++}
++
++#define io_apic_read(a,r) xen_io_apic_read(a,r)
++#define io_apic_write(a,r,v) xen_io_apic_write(a,r,v)
++
++#define clear_IO_APIC() ((void)0)
++
++#else
++
++#ifdef CONFIG_SMP
++static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t mask)
++{
++ unsigned long flags;
++ unsigned int dest;
++ cpumask_t tmp;
++
++ cpus_and(tmp, mask, cpu_online_map);
++ if (cpus_empty(tmp))
++ tmp = TARGET_CPUS;
++
++ cpus_and(mask, tmp, CPU_MASK_ALL);
++
++ dest = cpu_mask_to_apicid(mask);
++
++ /*
++ * Only the high 8 bits are valid.
++ */
++ dest = SET_APIC_LOGICAL_ID(dest);
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ __DO_ACTION(1, = dest, )
++ set_irq_info(irq, mask);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
++#endif
++
++#endif /* !CONFIG_XEN */
++
++/*
++ * The common case is 1:1 IRQ<->pin mappings. Sometimes there are
++ * shared ISA-space IRQs, so we have to support them. We are super
++ * fast in the common case, and fast for shared ISA-space IRQs.
++ */
++static void add_pin_to_irq(unsigned int irq, int apic, int pin)
++{
++ static int first_free_entry = NR_IRQS;
++ struct irq_pin_list *entry = irq_2_pin + irq;
++
++ BUG_ON(irq >= NR_IRQS);
++ while (entry->next)
++ entry = irq_2_pin + entry->next;
++
++ if (entry->pin != -1) {
++ entry->next = first_free_entry;
++ entry = irq_2_pin + entry->next;
++ if (++first_free_entry >= PIN_MAP_SIZE)
++ panic("io_apic.c: ran out of irq_2_pin entries!");
++ }
++ entry->apic = apic;
++ entry->pin = pin;
++}
++
++#ifndef CONFIG_XEN
++#define __DO_ACTION(R, ACTION, FINAL) \
++ \
++{ \
++ int pin; \
++ struct irq_pin_list *entry = irq_2_pin + irq; \
++ \
++ BUG_ON(irq >= NR_IRQS); \
++ for (;;) { \
++ unsigned int reg; \
++ pin = entry->pin; \
++ if (pin == -1) \
++ break; \
++ reg = io_apic_read(entry->apic, 0x10 + R + pin*2); \
++ reg ACTION; \
++ io_apic_modify(entry->apic, reg); \
++ if (!entry->next) \
++ break; \
++ entry = irq_2_pin + entry->next; \
++ } \
++ FINAL; \
++}
++
++#define DO_ACTION(name,R,ACTION, FINAL) \
++ \
++ static void name##_IO_APIC_irq (unsigned int irq) \
++ __DO_ACTION(R, ACTION, FINAL)
++
++DO_ACTION( __mask, 0, |= 0x00010000, io_apic_sync(entry->apic) )
++ /* mask = 1 */
++DO_ACTION( __unmask, 0, &= 0xfffeffff, )
++ /* mask = 0 */
++
++static void mask_IO_APIC_irq (unsigned int irq)
++{
++ unsigned long flags;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ __mask_IO_APIC_irq(irq);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
++
++static void unmask_IO_APIC_irq (unsigned int irq)
++{
++ unsigned long flags;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ __unmask_IO_APIC_irq(irq);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
++
++static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin)
++{
++ struct IO_APIC_route_entry entry;
++ unsigned long flags;
++
++ /* Check delivery_mode to be sure we're not clearing an SMI pin */
++ spin_lock_irqsave(&ioapic_lock, flags);
++ *(((int*)&entry) + 0) = io_apic_read(apic, 0x10 + 2 * pin);
++ *(((int*)&entry) + 1) = io_apic_read(apic, 0x11 + 2 * pin);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++ if (entry.delivery_mode == dest_SMI)
++ return;
++ /*
++ * Disable it in the IO-APIC irq-routing table:
++ */
++ memset(&entry, 0, sizeof(entry));
++ entry.mask = 1;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ io_apic_write(apic, 0x10 + 2 * pin, *(((int *)&entry) + 0));
++ io_apic_write(apic, 0x11 + 2 * pin, *(((int *)&entry) + 1));
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
++
++static void clear_IO_APIC (void)
++{
++ int apic, pin;
++
++ for (apic = 0; apic < nr_ioapics; apic++)
++ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++)
++ clear_IO_APIC_pin(apic, pin);
++}
++
++#endif /* !CONFIG_XEN */
++
++static u8 gsi_2_irq[NR_IRQ_VECTORS] = { [0 ... NR_IRQ_VECTORS-1] = 0xFF };
++
++/*
++ * support for broken MP BIOSs, enables hand-redirection of PIRQ0-7 to
++ * specific CPU-side IRQs.
++ */
++
++#define MAX_PIRQS 8
++static int pirq_entries [MAX_PIRQS];
++static int pirqs_enabled;
++int skip_ioapic_setup;
++int ioapic_force;
++
++/* dummy parsing: see setup.c */
++
++static int __init disable_ioapic_setup(char *str)
++{
++ skip_ioapic_setup = 1;
++ return 1;
++}
++
++static int __init enable_ioapic_setup(char *str)
++{
++ ioapic_force = 1;
++ skip_ioapic_setup = 0;
++ return 1;
++}
++
++__setup("noapic", disable_ioapic_setup);
++__setup("apic", enable_ioapic_setup);
++
++#ifndef CONFIG_XEN
++static int __init setup_disable_8254_timer(char *s)
++{
++ timer_over_8254 = -1;
++ return 1;
++}
++static int __init setup_enable_8254_timer(char *s)
++{
++ timer_over_8254 = 2;
++ return 1;
++}
++
++__setup("disable_8254_timer", setup_disable_8254_timer);
++__setup("enable_8254_timer", setup_enable_8254_timer);
++#endif /* !CONFIG_XEN */
++
++#include <asm/pci-direct.h>
++#include <linux/pci_ids.h>
++#include <linux/pci.h>
++
++
++#ifdef CONFIG_ACPI
++
++static int nvidia_hpet_detected __initdata;
++
++static int __init nvidia_hpet_check(unsigned long phys, unsigned long size)
++{
++ nvidia_hpet_detected = 1;
++ return 0;
++}
++#endif
++
++/* Temporary Hack. Nvidia and VIA boards currently only work with IO-APIC
++ off. Check for an Nvidia or VIA PCI bridge and turn it off.
++ Use pci direct infrastructure because this runs before the PCI subsystem.
++
++ Can be overwritten with "apic"
++
++ And another hack to disable the IOMMU on VIA chipsets.
++
++ ... and others. Really should move this somewhere else.
++
++ Kludge-O-Rama. */
++void __init check_ioapic(void)
++{
++ int num,slot,func;
++ /* Poor man's PCI discovery */
++ for (num = 0; num < 32; num++) {
++ for (slot = 0; slot < 32; slot++) {
++ for (func = 0; func < 8; func++) {
++ u32 class;
++ u32 vendor;
++ u8 type;
++ class = read_pci_config(num,slot,func,
++ PCI_CLASS_REVISION);
++ if (class == 0xffffffff)
++ break;
++
++ if ((class >> 16) != PCI_CLASS_BRIDGE_PCI)
++ continue;
++
++ vendor = read_pci_config(num, slot, func,
++ PCI_VENDOR_ID);
++ vendor &= 0xffff;
++ switch (vendor) {
++ case PCI_VENDOR_ID_VIA:
++#ifdef CONFIG_IOMMU
++ if ((end_pfn > MAX_DMA32_PFN ||
++ force_iommu) &&
++ !iommu_aperture_allowed) {
++ printk(KERN_INFO
++ "Looks like a VIA chipset. Disabling IOMMU. Override with \"iommu=allowed\"\n");
++ iommu_aperture_disabled = 1;
++ }
++#endif
++ return;
++ case PCI_VENDOR_ID_NVIDIA:
++#ifdef CONFIG_ACPI
++ /*
++ * All timer overrides on Nvidia are
++ * wrong unless HPET is enabled.
++ */
++ nvidia_hpet_detected = 0;
++ acpi_table_parse(ACPI_HPET,
++ nvidia_hpet_check);
++ if (nvidia_hpet_detected == 0) {
++ acpi_skip_timer_override = 1;
++ printk(KERN_INFO "Nvidia board "
++ "detected. Ignoring ACPI "
++ "timer override.\n");
++ }
++#endif
++ /* RED-PEN skip them on mptables too? */
++ return;
++ case PCI_VENDOR_ID_ATI:
++
++ /* This should be actually default, but
++ for 2.6.16 let's do it for ATI only where
++ it's really needed. */
++#ifndef CONFIG_XEN
++ if (timer_over_8254 == 1) {
++ timer_over_8254 = 0;
++ printk(KERN_INFO
++ "ATI board detected. Disabling timer routing over 8254.\n");
++ }
++#endif
++ return;
++ }
++
++
++ /* No multi-function device? */
++ type = read_pci_config_byte(num,slot,func,
++ PCI_HEADER_TYPE);
++ if (!(type & 0x80))
++ break;
++ }
++ }
++ }
++}
++
++static int __init ioapic_pirq_setup(char *str)
++{
++ int i, max;
++ int ints[MAX_PIRQS+1];
++
++ get_options(str, ARRAY_SIZE(ints), ints);
++
++ for (i = 0; i < MAX_PIRQS; i++)
++ pirq_entries[i] = -1;
++
++ pirqs_enabled = 1;
++ apic_printk(APIC_VERBOSE, "PIRQ redirection, working around broken MP-BIOS.\n");
++ max = MAX_PIRQS;
++ if (ints[0] < MAX_PIRQS)
++ max = ints[0];
++
++ for (i = 0; i < max; i++) {
++ apic_printk(APIC_VERBOSE, "... PIRQ%d -> IRQ %d\n", i, ints[i+1]);
++ /*
++ * PIRQs are mapped upside down, usually.
++ */
++ pirq_entries[MAX_PIRQS-i-1] = ints[i+1];
++ }
++ return 1;
++}
++
++__setup("pirq=", ioapic_pirq_setup);
++
++/*
++ * Find the IRQ entry number of a certain pin.
++ */
++static int find_irq_entry(int apic, int pin, int type)
++{
++ int i;
++
++ for (i = 0; i < mp_irq_entries; i++)
++ if (mp_irqs[i].mpc_irqtype == type &&
++ (mp_irqs[i].mpc_dstapic == mp_ioapics[apic].mpc_apicid ||
++ mp_irqs[i].mpc_dstapic == MP_APIC_ALL) &&
++ mp_irqs[i].mpc_dstirq == pin)
++ return i;
++
++ return -1;
++}
++
++#ifndef CONFIG_XEN
++/*
++ * Find the pin to which IRQ[irq] (ISA) is connected
++ */
++static int __init find_isa_irq_pin(int irq, int type)
++{
++ int i;
++
++ for (i = 0; i < mp_irq_entries; i++) {
++ int lbus = mp_irqs[i].mpc_srcbus;
++
++ if ((mp_bus_id_to_type[lbus] == MP_BUS_ISA ||
++ mp_bus_id_to_type[lbus] == MP_BUS_EISA ||
++ mp_bus_id_to_type[lbus] == MP_BUS_MCA) &&
++ (mp_irqs[i].mpc_irqtype == type) &&
++ (mp_irqs[i].mpc_srcbusirq == irq))
++
++ return mp_irqs[i].mpc_dstirq;
++ }
++ return -1;
++}
++
++static int __init find_isa_irq_apic(int irq, int type)
++{
++ int i;
++
++ for (i = 0; i < mp_irq_entries; i++) {
++ int lbus = mp_irqs[i].mpc_srcbus;
++
++ if ((mp_bus_id_to_type[lbus] == MP_BUS_ISA ||
++ mp_bus_id_to_type[lbus] == MP_BUS_EISA ||
++ mp_bus_id_to_type[lbus] == MP_BUS_MCA) &&
++ (mp_irqs[i].mpc_irqtype == type) &&
++ (mp_irqs[i].mpc_srcbusirq == irq))
++ break;
++ }
++ if (i < mp_irq_entries) {
++ int apic;
++ for(apic = 0; apic < nr_ioapics; apic++) {
++ if (mp_ioapics[apic].mpc_apicid == mp_irqs[i].mpc_dstapic)
++ return apic;
++ }
++ }
++
++ return -1;
++}
++#endif
++
++/*
++ * Find a specific PCI IRQ entry.
++ * Not an __init, possibly needed by modules
++ */
++static int pin_2_irq(int idx, int apic, int pin);
++
++int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin)
++{
++ int apic, i, best_guess = -1;
++
++ apic_printk(APIC_DEBUG, "querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n",
++ bus, slot, pin);
++ if (mp_bus_id_to_pci_bus[bus] == -1) {
++ apic_printk(APIC_VERBOSE, "PCI BIOS passed nonexistent PCI bus %d!\n", bus);
++ return -1;
++ }
++ for (i = 0; i < mp_irq_entries; i++) {
++ int lbus = mp_irqs[i].mpc_srcbus;
++
++ for (apic = 0; apic < nr_ioapics; apic++)
++ if (mp_ioapics[apic].mpc_apicid == mp_irqs[i].mpc_dstapic ||
++ mp_irqs[i].mpc_dstapic == MP_APIC_ALL)
++ break;
++
++ if ((mp_bus_id_to_type[lbus] == MP_BUS_PCI) &&
++ !mp_irqs[i].mpc_irqtype &&
++ (bus == lbus) &&
++ (slot == ((mp_irqs[i].mpc_srcbusirq >> 2) & 0x1f))) {
++ int irq = pin_2_irq(i,apic,mp_irqs[i].mpc_dstirq);
++
++ if (!(apic || IO_APIC_IRQ(irq)))
++ continue;
++
++ if (pin == (mp_irqs[i].mpc_srcbusirq & 3))
++ return irq;
++ /*
++ * Use the first all-but-pin matching entry as a
++ * best-guess fuzzy result for broken mptables.
++ */
++ if (best_guess < 0)
++ best_guess = irq;
++ }
++ }
++ BUG_ON(best_guess >= NR_IRQS);
++ return best_guess;
++}
++
++/*
++ * EISA Edge/Level control register, ELCR
++ */
++static int EISA_ELCR(unsigned int irq)
++{
++ if (irq < 16) {
++ unsigned int port = 0x4d0 + (irq >> 3);
++ return (inb(port) >> (irq & 7)) & 1;
++ }
++ apic_printk(APIC_VERBOSE, "Broken MPtable reports ISA irq %d\n", irq);
++ return 0;
++}
++
++/* EISA interrupts are always polarity zero and can be edge or level
++ * trigger depending on the ELCR value. If an interrupt is listed as
++ * EISA conforming in the MP table, that means its trigger type must
++ * be read in from the ELCR */
++
++#define default_EISA_trigger(idx) (EISA_ELCR(mp_irqs[idx].mpc_srcbusirq))
++#define default_EISA_polarity(idx) (0)
++
++/* ISA interrupts are always polarity zero edge triggered,
++ * when listed as conforming in the MP table. */
++
++#define default_ISA_trigger(idx) (0)
++#define default_ISA_polarity(idx) (0)
++
++/* PCI interrupts are always polarity one level triggered,
++ * when listed as conforming in the MP table. */
++
++#define default_PCI_trigger(idx) (1)
++#define default_PCI_polarity(idx) (1)
++
++/* MCA interrupts are always polarity zero level triggered,
++ * when listed as conforming in the MP table. */
++
++#define default_MCA_trigger(idx) (1)
++#define default_MCA_polarity(idx) (0)
++
++static int __init MPBIOS_polarity(int idx)
++{
++ int bus = mp_irqs[idx].mpc_srcbus;
++ int polarity;
++
++ /*
++ * Determine IRQ line polarity (high active or low active):
++ */
++ switch (mp_irqs[idx].mpc_irqflag & 3)
++ {
++ case 0: /* conforms, ie. bus-type dependent polarity */
++ {
++ switch (mp_bus_id_to_type[bus])
++ {
++ case MP_BUS_ISA: /* ISA pin */
++ {
++ polarity = default_ISA_polarity(idx);
++ break;
++ }
++ case MP_BUS_EISA: /* EISA pin */
++ {
++ polarity = default_EISA_polarity(idx);
++ break;
++ }
++ case MP_BUS_PCI: /* PCI pin */
++ {
++ polarity = default_PCI_polarity(idx);
++ break;
++ }
++ case MP_BUS_MCA: /* MCA pin */
++ {
++ polarity = default_MCA_polarity(idx);
++ break;
++ }
++ default:
++ {
++ printk(KERN_WARNING "broken BIOS!!\n");
++ polarity = 1;
++ break;
++ }
++ }
++ break;
++ }
++ case 1: /* high active */
++ {
++ polarity = 0;
++ break;
++ }
++ case 2: /* reserved */
++ {
++ printk(KERN_WARNING "broken BIOS!!\n");
++ polarity = 1;
++ break;
++ }
++ case 3: /* low active */
++ {
++ polarity = 1;
++ break;
++ }
++ default: /* invalid */
++ {
++ printk(KERN_WARNING "broken BIOS!!\n");
++ polarity = 1;
++ break;
++ }
++ }
++ return polarity;
++}
++
++static int MPBIOS_trigger(int idx)
++{
++ int bus = mp_irqs[idx].mpc_srcbus;
++ int trigger;
++
++ /*
++ * Determine IRQ trigger mode (edge or level sensitive):
++ */
++ switch ((mp_irqs[idx].mpc_irqflag>>2) & 3)
++ {
++ case 0: /* conforms, ie. bus-type dependent */
++ {
++ switch (mp_bus_id_to_type[bus])
++ {
++ case MP_BUS_ISA: /* ISA pin */
++ {
++ trigger = default_ISA_trigger(idx);
++ break;
++ }
++ case MP_BUS_EISA: /* EISA pin */
++ {
++ trigger = default_EISA_trigger(idx);
++ break;
++ }
++ case MP_BUS_PCI: /* PCI pin */
++ {
++ trigger = default_PCI_trigger(idx);
++ break;
++ }
++ case MP_BUS_MCA: /* MCA pin */
++ {
++ trigger = default_MCA_trigger(idx);
++ break;
++ }
++ default:
++ {
++ printk(KERN_WARNING "broken BIOS!!\n");
++ trigger = 1;
++ break;
++ }
++ }
++ break;
++ }
++ case 1: /* edge */
++ {
++ trigger = 0;
++ break;
++ }
++ case 2: /* reserved */
++ {
++ printk(KERN_WARNING "broken BIOS!!\n");
++ trigger = 1;
++ break;
++ }
++ case 3: /* level */
++ {
++ trigger = 1;
++ break;
++ }
++ default: /* invalid */
++ {
++ printk(KERN_WARNING "broken BIOS!!\n");
++ trigger = 0;
++ break;
++ }
++ }
++ return trigger;
++}
++
++static inline int irq_polarity(int idx)
++{
++ return MPBIOS_polarity(idx);
++}
++
++static inline int irq_trigger(int idx)
++{
++ return MPBIOS_trigger(idx);
++}
++
++static int next_irq = 16;
++
++/*
++ * gsi_irq_sharing -- Name overload! "irq" can be either a legacy IRQ
++ * in the range 0-15, a linux IRQ in the range 0-223, or a GSI number
++ * from ACPI, which can reach 800 in large boxen.
++ *
++ * Compact the sparse GSI space into a sequential IRQ series and reuse
++ * vectors if possible.
++ */
++int gsi_irq_sharing(int gsi)
++{
++ int i, tries, vector;
++
++ BUG_ON(gsi >= NR_IRQ_VECTORS);
++
++ if (platform_legacy_irq(gsi))
++ return gsi;
++
++ if (gsi_2_irq[gsi] != 0xFF)
++ return (int)gsi_2_irq[gsi];
++
++ tries = NR_IRQS;
++ try_again:
++ vector = assign_irq_vector(gsi);
++
++ /*
++ * Sharing vectors means sharing IRQs, so scan irq_vectors for previous
++ * use of vector and if found, return that IRQ. However, we never want
++ * to share legacy IRQs, which usually have a different trigger mode
++ * than PCI.
++ */
++ for (i = 0; i < NR_IRQS; i++)
++ if (IO_APIC_VECTOR(i) == vector)
++ break;
++ if (platform_legacy_irq(i)) {
++ if (--tries >= 0) {
++ IO_APIC_VECTOR(i) = 0;
++ goto try_again;
++ }
++ panic("gsi_irq_sharing: didn't find an IRQ using vector 0x%02X for GSI %d", vector, gsi);
++ }
++ if (i < NR_IRQS) {
++ gsi_2_irq[gsi] = i;
++ printk(KERN_INFO "GSI %d sharing vector 0x%02X and IRQ %d\n",
++ gsi, vector, i);
++ return i;
++ }
++
++ i = next_irq++;
++ BUG_ON(i >= NR_IRQS);
++ gsi_2_irq[gsi] = i;
++ IO_APIC_VECTOR(i) = vector;
++ printk(KERN_INFO "GSI %d assigned vector 0x%02X and IRQ %d\n",
++ gsi, vector, i);
++ return i;
++}
++
++static int pin_2_irq(int idx, int apic, int pin)
++{
++ int irq, i;
++ int bus = mp_irqs[idx].mpc_srcbus;
++
++ /*
++ * Debugging check, we are in big trouble if this message pops up!
++ */
++ if (mp_irqs[idx].mpc_dstirq != pin)
++ printk(KERN_ERR "broken BIOS or MPTABLE parser, ayiee!!\n");
++
++ switch (mp_bus_id_to_type[bus])
++ {
++ case MP_BUS_ISA: /* ISA pin */
++ case MP_BUS_EISA:
++ case MP_BUS_MCA:
++ {
++ irq = mp_irqs[idx].mpc_srcbusirq;
++ break;
++ }
++ case MP_BUS_PCI: /* PCI pin */
++ {
++ /*
++ * PCI IRQs are mapped in order
++ */
++ i = irq = 0;
++ while (i < apic)
++ irq += nr_ioapic_registers[i++];
++ irq += pin;
++ irq = gsi_irq_sharing(irq);
++ break;
++ }
++ default:
++ {
++ printk(KERN_ERR "unknown bus type %d.\n",bus);
++ irq = 0;
++ break;
++ }
++ }
++ BUG_ON(irq >= NR_IRQS);
++
++ /*
++ * PCI IRQ command line redirection. Yes, limits are hardcoded.
++ */
++ if ((pin >= 16) && (pin <= 23)) {
++ if (pirq_entries[pin-16] != -1) {
++ if (!pirq_entries[pin-16]) {
++ apic_printk(APIC_VERBOSE, "disabling PIRQ%d\n", pin-16);
++ } else {
++ irq = pirq_entries[pin-16];
++ apic_printk(APIC_VERBOSE, "using PIRQ%d -> IRQ %d\n",
++ pin-16, irq);
++ }
++ }
++ }
++ BUG_ON(irq >= NR_IRQS);
++ return irq;
++}
++
++static inline int IO_APIC_irq_trigger(int irq)
++{
++ int apic, idx, pin;
++
++ for (apic = 0; apic < nr_ioapics; apic++) {
++ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
++ idx = find_irq_entry(apic,pin,mp_INT);
++ if ((idx != -1) && (irq == pin_2_irq(idx,apic,pin)))
++ return irq_trigger(idx);
++ }
++ }
++ /*
++ * nonexistent IRQs are edge default
++ */
++ return 0;
++}
++
++/* irq_vectors is indexed by the sum of all RTEs in all I/O APICs. */
++u8 irq_vector[NR_IRQ_VECTORS] __read_mostly;
++
++int assign_irq_vector(int irq)
++{
++ unsigned long flags;
++ int vector;
++ struct physdev_irq irq_op;
++
++ BUG_ON(irq != AUTO_ASSIGN && (unsigned)irq >= NR_IRQ_VECTORS);
++
++ spin_lock_irqsave(&vector_lock, flags);
++
++ if (irq != AUTO_ASSIGN && IO_APIC_VECTOR(irq) > 0) {
++ spin_unlock_irqrestore(&vector_lock, flags);
++ return IO_APIC_VECTOR(irq);
++ }
++
++ irq_op.irq = irq;
++ if (HYPERVISOR_physdev_op(PHYSDEVOP_alloc_irq_vector, &irq_op)) {
++ spin_unlock_irqrestore(&vector_lock, flags);
++ return -ENOSPC;
++ }
++
++ vector = irq_op.vector;
++ vector_irq[vector] = irq;
++ if (irq != AUTO_ASSIGN)
++ IO_APIC_VECTOR(irq) = vector;
++
++ spin_unlock_irqrestore(&vector_lock, flags);
++
++ return vector;
++}
++
++extern void (*interrupt[NR_IRQS])(void);
++#ifndef CONFIG_XEN
++static struct hw_interrupt_type ioapic_level_type;
++static struct hw_interrupt_type ioapic_edge_type;
++
++#define IOAPIC_AUTO -1
++#define IOAPIC_EDGE 0
++#define IOAPIC_LEVEL 1
++
++static void ioapic_register_intr(int irq, int vector, unsigned long trigger)
++{
++ unsigned idx;
++
++ idx = use_pci_vector() && !platform_legacy_irq(irq) ? vector : irq;
++
++ if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
++ trigger == IOAPIC_LEVEL)
++ irq_desc[idx].chip = &ioapic_level_type;
++ else
++ irq_desc[idx].chip = &ioapic_edge_type;
++ set_intr_gate(vector, interrupt[idx]);
++}
++#else
++#define ioapic_register_intr(irq, vector, trigger) evtchn_register_pirq(irq)
++#endif /* !CONFIG_XEN */
++
++static void __init setup_IO_APIC_irqs(void)
++{
++ struct IO_APIC_route_entry entry;
++ int apic, pin, idx, irq, first_notcon = 1, vector;
++ unsigned long flags;
++
++ apic_printk(APIC_VERBOSE, KERN_DEBUG "init IO_APIC IRQs\n");
++
++ for (apic = 0; apic < nr_ioapics; apic++) {
++ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
++
++ /*
++ * add it to the IO-APIC irq-routing table:
++ */
++ memset(&entry,0,sizeof(entry));
++
++ entry.delivery_mode = INT_DELIVERY_MODE;
++ entry.dest_mode = INT_DEST_MODE;
++ entry.mask = 0; /* enable IRQ */
++ entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS);
++
++ idx = find_irq_entry(apic,pin,mp_INT);
++ if (idx == -1) {
++ if (first_notcon) {
++ apic_printk(APIC_VERBOSE, KERN_DEBUG " IO-APIC (apicid-pin) %d-%d", mp_ioapics[apic].mpc_apicid, pin);
++ first_notcon = 0;
++ } else
++ apic_printk(APIC_VERBOSE, ", %d-%d", mp_ioapics[apic].mpc_apicid, pin);
++ continue;
++ }
++
++ entry.trigger = irq_trigger(idx);
++ entry.polarity = irq_polarity(idx);
++
++ if (irq_trigger(idx)) {
++ entry.trigger = 1;
++ entry.mask = 1;
++ entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS);
++ }
++
++ irq = pin_2_irq(idx, apic, pin);
++ add_pin_to_irq(irq, apic, pin);
++
++ if (/* !apic && */ !IO_APIC_IRQ(irq))
++ continue;
++
++ if (IO_APIC_IRQ(irq)) {
++ vector = assign_irq_vector(irq);
++ entry.vector = vector;
++
++ ioapic_register_intr(irq, vector, IOAPIC_AUTO);
++ if (!apic && (irq < 16))
++ disable_8259A_irq(irq);
++ }
++ spin_lock_irqsave(&ioapic_lock, flags);
++ io_apic_write(apic, 0x11+2*pin, *(((int *)&entry)+1));
++ io_apic_write(apic, 0x10+2*pin, *(((int *)&entry)+0));
++ set_native_irq_info(irq, TARGET_CPUS);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++ }
++ }
++
++ if (!first_notcon)
++ apic_printk(APIC_VERBOSE," not connected.\n");
++}
++
++#ifndef CONFIG_XEN
++/*
++ * Set up the 8259A-master output pin as broadcast to all
++ * CPUs.
++ */
++static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, int vector)
++{
++ struct IO_APIC_route_entry entry;
++ unsigned long flags;
++
++ memset(&entry,0,sizeof(entry));
++
++ disable_8259A_irq(0);
++
++ /* mask LVT0 */
++ apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
++
++ /*
++ * We use logical delivery to get the timer IRQ
++ * to the first CPU.
++ */
++ entry.dest_mode = INT_DEST_MODE;
++ entry.mask = 0; /* unmask IRQ now */
++ entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS);
++ entry.delivery_mode = INT_DELIVERY_MODE;
++ entry.polarity = 0;
++ entry.trigger = 0;
++ entry.vector = vector;
++
++ /*
++ * The timer IRQ doesn't have to know that behind the
++ * scene we have a 8259A-master in AEOI mode ...
++ */
++ irq_desc[0].chip = &ioapic_edge_type;
++
++ /*
++ * Add it to the IO-APIC irq-routing table:
++ */
++ spin_lock_irqsave(&ioapic_lock, flags);
++ io_apic_write(apic, 0x11+2*pin, *(((int *)&entry)+1));
++ io_apic_write(apic, 0x10+2*pin, *(((int *)&entry)+0));
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ enable_8259A_irq(0);
++}
++
++void __init UNEXPECTED_IO_APIC(void)
++{
++}
++
++void __apicdebuginit print_IO_APIC(void)
++{
++ int apic, i;
++ union IO_APIC_reg_00 reg_00;
++ union IO_APIC_reg_01 reg_01;
++ union IO_APIC_reg_02 reg_02;
++ unsigned long flags;
++
++ if (apic_verbosity == APIC_QUIET)
++ return;
++
++ printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries);
++ for (i = 0; i < nr_ioapics; i++)
++ printk(KERN_DEBUG "number of IO-APIC #%d registers: %d.\n",
++ mp_ioapics[i].mpc_apicid, nr_ioapic_registers[i]);
++
++ /*
++ * We are a bit conservative about what we expect. We have to
++ * know about every hardware change ASAP.
++ */
++ printk(KERN_INFO "testing the IO APIC.......................\n");
++
++ for (apic = 0; apic < nr_ioapics; apic++) {
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_00.raw = io_apic_read(apic, 0);
++ reg_01.raw = io_apic_read(apic, 1);
++ if (reg_01.bits.version >= 0x10)
++ reg_02.raw = io_apic_read(apic, 2);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ printk("\n");
++ printk(KERN_DEBUG "IO APIC #%d......\n", mp_ioapics[apic].mpc_apicid);
++ printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw);
++ printk(KERN_DEBUG "....... : physical APIC id: %02X\n", reg_00.bits.ID);
++ if (reg_00.bits.__reserved_1 || reg_00.bits.__reserved_2)
++ UNEXPECTED_IO_APIC();
++
++ printk(KERN_DEBUG ".... register #01: %08X\n", *(int *)®_01);
++ printk(KERN_DEBUG "....... : max redirection entries: %04X\n", reg_01.bits.entries);
++ if ( (reg_01.bits.entries != 0x0f) && /* older (Neptune) boards */
++ (reg_01.bits.entries != 0x17) && /* typical ISA+PCI boards */
++ (reg_01.bits.entries != 0x1b) && /* Compaq Proliant boards */
++ (reg_01.bits.entries != 0x1f) && /* dual Xeon boards */
++ (reg_01.bits.entries != 0x22) && /* bigger Xeon boards */
++ (reg_01.bits.entries != 0x2E) &&
++ (reg_01.bits.entries != 0x3F) &&
++ (reg_01.bits.entries != 0x03)
++ )
++ UNEXPECTED_IO_APIC();
++
++ printk(KERN_DEBUG "....... : PRQ implemented: %X\n", reg_01.bits.PRQ);
++ printk(KERN_DEBUG "....... : IO APIC version: %04X\n", reg_01.bits.version);
++ if ( (reg_01.bits.version != 0x01) && /* 82489DX IO-APICs */
++ (reg_01.bits.version != 0x02) && /* 82801BA IO-APICs (ICH2) */
++ (reg_01.bits.version != 0x10) && /* oldest IO-APICs */
++ (reg_01.bits.version != 0x11) && /* Pentium/Pro IO-APICs */
++ (reg_01.bits.version != 0x13) && /* Xeon IO-APICs */
++ (reg_01.bits.version != 0x20) /* Intel P64H (82806 AA) */
++ )
++ UNEXPECTED_IO_APIC();
++ if (reg_01.bits.__reserved_1 || reg_01.bits.__reserved_2)
++ UNEXPECTED_IO_APIC();
++
++ if (reg_01.bits.version >= 0x10) {
++ printk(KERN_DEBUG ".... register #02: %08X\n", reg_02.raw);
++ printk(KERN_DEBUG "....... : arbitration: %02X\n", reg_02.bits.arbitration);
++ if (reg_02.bits.__reserved_1 || reg_02.bits.__reserved_2)
++ UNEXPECTED_IO_APIC();
++ }
++
++ printk(KERN_DEBUG ".... IRQ redirection table:\n");
++
++ printk(KERN_DEBUG " NR Log Phy Mask Trig IRR Pol"
++ " Stat Dest Deli Vect: \n");
++
++ for (i = 0; i <= reg_01.bits.entries; i++) {
++ struct IO_APIC_route_entry entry;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ *(((int *)&entry)+0) = io_apic_read(apic, 0x10+i*2);
++ *(((int *)&entry)+1) = io_apic_read(apic, 0x11+i*2);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ printk(KERN_DEBUG " %02x %03X %02X ",
++ i,
++ entry.dest.logical.logical_dest,
++ entry.dest.physical.physical_dest
++ );
++
++ printk("%1d %1d %1d %1d %1d %1d %1d %02X\n",
++ entry.mask,
++ entry.trigger,
++ entry.irr,
++ entry.polarity,
++ entry.delivery_status,
++ entry.dest_mode,
++ entry.delivery_mode,
++ entry.vector
++ );
++ }
++ }
++ if (use_pci_vector())
++ printk(KERN_INFO "Using vector-based indexing\n");
++ printk(KERN_DEBUG "IRQ to pin mappings:\n");
++ for (i = 0; i < NR_IRQS; i++) {
++ struct irq_pin_list *entry = irq_2_pin + i;
++ if (entry->pin < 0)
++ continue;
++ if (use_pci_vector() && !platform_legacy_irq(i))
++ printk(KERN_DEBUG "IRQ%d ", IO_APIC_VECTOR(i));
++ else
++ printk(KERN_DEBUG "IRQ%d ", i);
++ for (;;) {
++ printk("-> %d:%d", entry->apic, entry->pin);
++ if (!entry->next)
++ break;
++ entry = irq_2_pin + entry->next;
++ }
++ printk("\n");
++ }
++
++ printk(KERN_INFO ".................................... done.\n");
++
++ return;
++}
++
++static __apicdebuginit void print_APIC_bitfield (int base)
++{
++ unsigned int v;
++ int i, j;
++
++ if (apic_verbosity == APIC_QUIET)
++ return;
++
++ printk(KERN_DEBUG "0123456789abcdef0123456789abcdef\n" KERN_DEBUG);
++ for (i = 0; i < 8; i++) {
++ v = apic_read(base + i*0x10);
++ for (j = 0; j < 32; j++) {
++ if (v & (1<<j))
++ printk("1");
++ else
++ printk("0");
++ }
++ printk("\n");
++ }
++}
++
++void __apicdebuginit print_local_APIC(void * dummy)
++{
++ unsigned int v, ver, maxlvt;
++
++ if (apic_verbosity == APIC_QUIET)
++ return;
++
++ printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n",
++ smp_processor_id(), hard_smp_processor_id());
++ v = apic_read(APIC_ID);
++ printk(KERN_INFO "... APIC ID: %08x (%01x)\n", v, GET_APIC_ID(v));
++ v = apic_read(APIC_LVR);
++ printk(KERN_INFO "... APIC VERSION: %08x\n", v);
++ ver = GET_APIC_VERSION(v);
++ maxlvt = get_maxlvt();
++
++ v = apic_read(APIC_TASKPRI);
++ printk(KERN_DEBUG "... APIC TASKPRI: %08x (%02x)\n", v, v & APIC_TPRI_MASK);
++
++ v = apic_read(APIC_ARBPRI);
++ printk(KERN_DEBUG "... APIC ARBPRI: %08x (%02x)\n", v,
++ v & APIC_ARBPRI_MASK);
++ v = apic_read(APIC_PROCPRI);
++ printk(KERN_DEBUG "... APIC PROCPRI: %08x\n", v);
++
++ v = apic_read(APIC_EOI);
++ printk(KERN_DEBUG "... APIC EOI: %08x\n", v);
++ v = apic_read(APIC_RRR);
++ printk(KERN_DEBUG "... APIC RRR: %08x\n", v);
++ v = apic_read(APIC_LDR);
++ printk(KERN_DEBUG "... APIC LDR: %08x\n", v);
++ v = apic_read(APIC_DFR);
++ printk(KERN_DEBUG "... APIC DFR: %08x\n", v);
++ v = apic_read(APIC_SPIV);
++ printk(KERN_DEBUG "... APIC SPIV: %08x\n", v);
++
++ printk(KERN_DEBUG "... APIC ISR field:\n");
++ print_APIC_bitfield(APIC_ISR);
++ printk(KERN_DEBUG "... APIC TMR field:\n");
++ print_APIC_bitfield(APIC_TMR);
++ printk(KERN_DEBUG "... APIC IRR field:\n");
++ print_APIC_bitfield(APIC_IRR);
++
++ v = apic_read(APIC_ESR);
++ printk(KERN_DEBUG "... APIC ESR: %08x\n", v);
++
++ v = apic_read(APIC_ICR);
++ printk(KERN_DEBUG "... APIC ICR: %08x\n", v);
++ v = apic_read(APIC_ICR2);
++ printk(KERN_DEBUG "... APIC ICR2: %08x\n", v);
++
++ v = apic_read(APIC_LVTT);
++ printk(KERN_DEBUG "... APIC LVTT: %08x\n", v);
++
++ if (maxlvt > 3) { /* PC is LVT#4. */
++ v = apic_read(APIC_LVTPC);
++ printk(KERN_DEBUG "... APIC LVTPC: %08x\n", v);
++ }
++ v = apic_read(APIC_LVT0);
++ printk(KERN_DEBUG "... APIC LVT0: %08x\n", v);
++ v = apic_read(APIC_LVT1);
++ printk(KERN_DEBUG "... APIC LVT1: %08x\n", v);
++
++ if (maxlvt > 2) { /* ERR is LVT#3. */
++ v = apic_read(APIC_LVTERR);
++ printk(KERN_DEBUG "... APIC LVTERR: %08x\n", v);
++ }
++
++ v = apic_read(APIC_TMICT);
++ printk(KERN_DEBUG "... APIC TMICT: %08x\n", v);
++ v = apic_read(APIC_TMCCT);
++ printk(KERN_DEBUG "... APIC TMCCT: %08x\n", v);
++ v = apic_read(APIC_TDCR);
++ printk(KERN_DEBUG "... APIC TDCR: %08x\n", v);
++ printk("\n");
++}
++
++void print_all_local_APICs (void)
++{
++ on_each_cpu(print_local_APIC, NULL, 1, 1);
++}
++
++void __apicdebuginit print_PIC(void)
++{
++ unsigned int v;
++ unsigned long flags;
++
++ if (apic_verbosity == APIC_QUIET)
++ return;
++
++ printk(KERN_DEBUG "\nprinting PIC contents\n");
++
++ spin_lock_irqsave(&i8259A_lock, flags);
++
++ v = inb(0xa1) << 8 | inb(0x21);
++ printk(KERN_DEBUG "... PIC IMR: %04x\n", v);
++
++ v = inb(0xa0) << 8 | inb(0x20);
++ printk(KERN_DEBUG "... PIC IRR: %04x\n", v);
++
++ outb(0x0b,0xa0);
++ outb(0x0b,0x20);
++ v = inb(0xa0) << 8 | inb(0x20);
++ outb(0x0a,0xa0);
++ outb(0x0a,0x20);
++
++ spin_unlock_irqrestore(&i8259A_lock, flags);
++
++ printk(KERN_DEBUG "... PIC ISR: %04x\n", v);
++
++ v = inb(0x4d1) << 8 | inb(0x4d0);
++ printk(KERN_DEBUG "... PIC ELCR: %04x\n", v);
++}
++#endif /* !CONFIG_XEN */
++
++static void __init enable_IO_APIC(void)
++{
++ union IO_APIC_reg_01 reg_01;
++#ifndef CONFIG_XEN
++ int i8259_apic, i8259_pin;
++#endif
++ int i, apic;
++ unsigned long flags;
++
++ for (i = 0; i < PIN_MAP_SIZE; i++) {
++ irq_2_pin[i].pin = -1;
++ irq_2_pin[i].next = 0;
++ }
++ if (!pirqs_enabled)
++ for (i = 0; i < MAX_PIRQS; i++)
++ pirq_entries[i] = -1;
++
++ /*
++ * The number of IO-APIC IRQ registers (== #pins):
++ */
++ for (apic = 0; apic < nr_ioapics; apic++) {
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_01.raw = io_apic_read(apic, 1);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++ nr_ioapic_registers[apic] = reg_01.bits.entries+1;
++ }
++#ifndef CONFIG_XEN
++ for(apic = 0; apic < nr_ioapics; apic++) {
++ int pin;
++ /* See if any of the pins is in ExtINT mode */
++ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
++ struct IO_APIC_route_entry entry;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ *(((int *)&entry) + 0) = io_apic_read(apic, 0x10 + 2 * pin);
++ *(((int *)&entry) + 1) = io_apic_read(apic, 0x11 + 2 * pin);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++
++ /* If the interrupt line is enabled and in ExtInt mode
++ * I have found the pin where the i8259 is connected.
++ */
++ if ((entry.mask == 0) && (entry.delivery_mode == dest_ExtINT)) {
++ ioapic_i8259.apic = apic;
++ ioapic_i8259.pin = pin;
++ goto found_i8259;
++ }
++ }
++ }
++ found_i8259:
++ /* Look to see what if the MP table has reported the ExtINT */
++ i8259_pin = find_isa_irq_pin(0, mp_ExtINT);
++ i8259_apic = find_isa_irq_apic(0, mp_ExtINT);
++ /* Trust the MP table if nothing is setup in the hardware */
++ if ((ioapic_i8259.pin == -1) && (i8259_pin >= 0)) {
++ printk(KERN_WARNING "ExtINT not setup in hardware but reported by MP table\n");
++ ioapic_i8259.pin = i8259_pin;
++ ioapic_i8259.apic = i8259_apic;
++ }
++ /* Complain if the MP table and the hardware disagree */
++ if (((ioapic_i8259.apic != i8259_apic) || (ioapic_i8259.pin != i8259_pin)) &&
++ (i8259_pin >= 0) && (ioapic_i8259.pin >= 0))
++ {
++ printk(KERN_WARNING "ExtINT in hardware and MP table differ\n");
++ }
++#endif
++
++ /*
++ * Do not trust the IO-APIC being empty at bootup
++ */
++ clear_IO_APIC();
++}
++
++/*
++ * Not an __init, needed by the reboot code
++ */
++void disable_IO_APIC(void)
++{
++ /*
++ * Clear the IO-APIC before rebooting:
++ */
++ clear_IO_APIC();
++
++#ifndef CONFIG_XEN
++ /*
++ * If the i8259 is routed through an IOAPIC
++ * Put that IOAPIC in virtual wire mode
++ * so legacy interrupts can be delivered.
++ */
++ if (ioapic_i8259.pin != -1) {
++ struct IO_APIC_route_entry entry;
++ unsigned long flags;
++
++ memset(&entry, 0, sizeof(entry));
++ entry.mask = 0; /* Enabled */
++ entry.trigger = 0; /* Edge */
++ entry.irr = 0;
++ entry.polarity = 0; /* High */
++ entry.delivery_status = 0;
++ entry.dest_mode = 0; /* Physical */
++ entry.delivery_mode = dest_ExtINT; /* ExtInt */
++ entry.vector = 0;
++ entry.dest.physical.physical_dest =
++ GET_APIC_ID(apic_read(APIC_ID));
++
++ /*
++ * Add it to the IO-APIC irq-routing table:
++ */
++ spin_lock_irqsave(&ioapic_lock, flags);
++ io_apic_write(ioapic_i8259.apic, 0x11+2*ioapic_i8259.pin,
++ *(((int *)&entry)+1));
++ io_apic_write(ioapic_i8259.apic, 0x10+2*ioapic_i8259.pin,
++ *(((int *)&entry)+0));
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++ }
++
++ disconnect_bsp_APIC(ioapic_i8259.pin != -1);
++#endif
++}
++
++/*
++ * function to set the IO-APIC physical IDs based on the
++ * values stored in the MPC table.
++ *
++ * by Matt Domsch <Matt_Domsch at dell.com> Tue Dec 21 12:25:05 CST 1999
++ */
++
++#ifndef CONFIG_XEN
++static void __init setup_ioapic_ids_from_mpc (void)
++{
++ union IO_APIC_reg_00 reg_00;
++ int apic;
++ int i;
++ unsigned char old_id;
++ unsigned long flags;
++
++ /*
++ * Set the IOAPIC ID to the value stored in the MPC table.
++ */
++ for (apic = 0; apic < nr_ioapics; apic++) {
++
++ /* Read the register 0 value */
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_00.raw = io_apic_read(apic, 0);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ old_id = mp_ioapics[apic].mpc_apicid;
++
++
++ printk(KERN_INFO "Using IO-APIC %d\n", mp_ioapics[apic].mpc_apicid);
++
++
++ /*
++ * We need to adjust the IRQ routing table
++ * if the ID changed.
++ */
++ if (old_id != mp_ioapics[apic].mpc_apicid)
++ for (i = 0; i < mp_irq_entries; i++)
++ if (mp_irqs[i].mpc_dstapic == old_id)
++ mp_irqs[i].mpc_dstapic
++ = mp_ioapics[apic].mpc_apicid;
++
++ /*
++ * Read the right value from the MPC table and
++ * write it into the ID register.
++ */
++ apic_printk(APIC_VERBOSE,KERN_INFO "...changing IO-APIC physical APIC ID to %d ...",
++ mp_ioapics[apic].mpc_apicid);
++
++ reg_00.bits.ID = mp_ioapics[apic].mpc_apicid;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ io_apic_write(apic, 0, reg_00.raw);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ /*
++ * Sanity check
++ */
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_00.raw = io_apic_read(apic, 0);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++ if (reg_00.bits.ID != mp_ioapics[apic].mpc_apicid)
++ printk("could not set ID!\n");
++ else
++ apic_printk(APIC_VERBOSE," ok.\n");
++ }
++}
++#else
++static void __init setup_ioapic_ids_from_mpc(void) { }
++#endif
++
++/*
++ * There is a nasty bug in some older SMP boards, their mptable lies
++ * about the timer IRQ. We do the following to work around the situation:
++ *
++ * - timer IRQ defaults to IO-APIC IRQ
++ * - if this function detects that timer IRQs are defunct, then we fall
++ * back to ISA timer IRQs
++ */
++#ifndef CONFIG_XEN
++static int __init timer_irq_works(void)
++{
++ unsigned long t1 = jiffies;
++
++ local_irq_enable();
++ /* Let ten ticks pass... */
++ mdelay((10 * 1000) / HZ);
++
++ /*
++ * Expect a few ticks at least, to be sure some possible
++ * glue logic does not lock up after one or two first
++ * ticks in a non-ExtINT mode. Also the local APIC
++ * might have cached one ExtINT interrupt. Finally, at
++ * least one tick may be lost due to delays.
++ */
++
++ /* jiffies wrap? */
++ if (jiffies - t1 > 4)
++ return 1;
++ return 0;
++}
++
++/*
++ * In the SMP+IOAPIC case it might happen that there are an unspecified
++ * number of pending IRQ events unhandled. These cases are very rare,
++ * so we 'resend' these IRQs via IPIs, to the same CPU. It's much
++ * better to do it this way as thus we do not have to be aware of
++ * 'pending' interrupts in the IRQ path, except at this point.
++ */
++/*
++ * Edge triggered needs to resend any interrupt
++ * that was delayed but this is now handled in the device
++ * independent code.
++ */
++
++/*
++ * Starting up a edge-triggered IO-APIC interrupt is
++ * nasty - we need to make sure that we get the edge.
++ * If it is already asserted for some reason, we need
++ * return 1 to indicate that is was pending.
++ *
++ * This is not complete - we should be able to fake
++ * an edge even if it isn't on the 8259A...
++ */
++
++static unsigned int startup_edge_ioapic_irq(unsigned int irq)
++{
++ int was_pending = 0;
++ unsigned long flags;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ if (irq < 16) {
++ disable_8259A_irq(irq);
++ if (i8259A_irq_pending(irq))
++ was_pending = 1;
++ }
++ __unmask_IO_APIC_irq(irq);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ return was_pending;
++}
++
++/*
++ * Once we have recorded IRQ_PENDING already, we can mask the
++ * interrupt for real. This prevents IRQ storms from unhandled
++ * devices.
++ */
++static void ack_edge_ioapic_irq(unsigned int irq)
++{
++ move_irq(irq);
++ if ((irq_desc[irq].status & (IRQ_PENDING | IRQ_DISABLED))
++ == (IRQ_PENDING | IRQ_DISABLED))
++ mask_IO_APIC_irq(irq);
++ ack_APIC_irq();
++}
++
++/*
++ * Level triggered interrupts can just be masked,
++ * and shutting down and starting up the interrupt
++ * is the same as enabling and disabling them -- except
++ * with a startup need to return a "was pending" value.
++ *
++ * Level triggered interrupts are special because we
++ * do not touch any IO-APIC register while handling
++ * them. We ack the APIC in the end-IRQ handler, not
++ * in the start-IRQ-handler. Protection against reentrance
++ * from the same interrupt is still provided, both by the
++ * generic IRQ layer and by the fact that an unacked local
++ * APIC does not accept IRQs.
++ */
++static unsigned int startup_level_ioapic_irq (unsigned int irq)
++{
++ unmask_IO_APIC_irq(irq);
++
++ return 0; /* don't check for pending */
++}
++
++static void end_level_ioapic_irq (unsigned int irq)
++{
++ move_irq(irq);
++ ack_APIC_irq();
++}
++
++#ifdef CONFIG_PCI_MSI
++static unsigned int startup_edge_ioapic_vector(unsigned int vector)
++{
++ int irq = vector_to_irq(vector);
++
++ return startup_edge_ioapic_irq(irq);
++}
++
++static void ack_edge_ioapic_vector(unsigned int vector)
++{
++ int irq = vector_to_irq(vector);
++
++ move_native_irq(vector);
++ ack_edge_ioapic_irq(irq);
++}
++
++static unsigned int startup_level_ioapic_vector (unsigned int vector)
++{
++ int irq = vector_to_irq(vector);
++
++ return startup_level_ioapic_irq (irq);
++}
++
++static void end_level_ioapic_vector (unsigned int vector)
++{
++ int irq = vector_to_irq(vector);
++
++ move_native_irq(vector);
++ end_level_ioapic_irq(irq);
++}
++
++static void mask_IO_APIC_vector (unsigned int vector)
++{
++ int irq = vector_to_irq(vector);
++
++ mask_IO_APIC_irq(irq);
++}
++
++static void unmask_IO_APIC_vector (unsigned int vector)
++{
++ int irq = vector_to_irq(vector);
++
++ unmask_IO_APIC_irq(irq);
++}
++
++#ifdef CONFIG_SMP
++static void set_ioapic_affinity_vector (unsigned int vector,
++ cpumask_t cpu_mask)
++{
++ int irq = vector_to_irq(vector);
++
++ set_native_irq_info(vector, cpu_mask);
++ set_ioapic_affinity_irq(irq, cpu_mask);
++}
++#endif // CONFIG_SMP
++#endif // CONFIG_PCI_MSI
++
++static int ioapic_retrigger(unsigned int irq)
++{
++ send_IPI_self(IO_APIC_VECTOR(irq));
++
++ return 1;
++}
++
++/*
++ * Level and edge triggered IO-APIC interrupts need different handling,
++ * so we use two separate IRQ descriptors. Edge triggered IRQs can be
++ * handled with the level-triggered descriptor, but that one has slightly
++ * more overhead. Level-triggered interrupts cannot be handled with the
++ * edge-triggered handler, without risking IRQ storms and other ugly
++ * races.
++ */
++
++static struct hw_interrupt_type ioapic_edge_type __read_mostly = {
++ .typename = "IO-APIC-edge",
++ .startup = startup_edge_ioapic,
++ .shutdown = shutdown_edge_ioapic,
++ .enable = enable_edge_ioapic,
++ .disable = disable_edge_ioapic,
++ .ack = ack_edge_ioapic,
++ .end = end_edge_ioapic,
++#ifdef CONFIG_SMP
++ .set_affinity = set_ioapic_affinity,
++#endif
++ .retrigger = ioapic_retrigger,
++};
++
++static struct hw_interrupt_type ioapic_level_type __read_mostly = {
++ .typename = "IO-APIC-level",
++ .startup = startup_level_ioapic,
++ .shutdown = shutdown_level_ioapic,
++ .enable = enable_level_ioapic,
++ .disable = disable_level_ioapic,
++ .ack = mask_and_ack_level_ioapic,
++ .end = end_level_ioapic,
++#ifdef CONFIG_SMP
++ .set_affinity = set_ioapic_affinity,
++#endif
++ .retrigger = ioapic_retrigger,
++};
++#endif /* !CONFIG_XEN */
++
++static inline void init_IO_APIC_traps(void)
++{
++ int irq;
++
++ /*
++ * NOTE! The local APIC isn't very good at handling
++ * multiple interrupts at the same interrupt level.
++ * As the interrupt level is determined by taking the
++ * vector number and shifting that right by 4, we
++ * want to spread these out a bit so that they don't
++ * all fall in the same interrupt level.
++ *
++ * Also, we've got to be careful not to trash gate
++ * 0x80, because int 0x80 is hm, kind of importantish. ;)
++ */
++ for (irq = 0; irq < NR_IRQS ; irq++) {
++ int tmp = irq;
++ if (use_pci_vector()) {
++ if (!platform_legacy_irq(tmp))
++ if ((tmp = vector_to_irq(tmp)) == -1)
++ continue;
++ }
++ if (IO_APIC_IRQ(tmp) && !IO_APIC_VECTOR(tmp)) {
++ /*
++ * Hmm.. We don't have an entry for this,
++ * so default to an old-fashioned 8259
++ * interrupt if we can..
++ */
++ if (irq < 16)
++ make_8259A_irq(irq);
++#ifndef CONFIG_XEN
++ else
++ /* Strange. Oh, well.. */
++ irq_desc[irq].chip = &no_irq_type;
++#endif
++ }
++ }
++}
++
++#ifndef CONFIG_XEN
++static void enable_lapic_irq (unsigned int irq)
++{
++ unsigned long v;
++
++ v = apic_read(APIC_LVT0);
++ apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED);
++}
++
++static void disable_lapic_irq (unsigned int irq)
++{
++ unsigned long v;
++
++ v = apic_read(APIC_LVT0);
++ apic_write(APIC_LVT0, v | APIC_LVT_MASKED);
++}
++
++static void ack_lapic_irq (unsigned int irq)
++{
++ ack_APIC_irq();
++}
++
++static void end_lapic_irq (unsigned int i) { /* nothing */ }
++
++static struct hw_interrupt_type lapic_irq_type __read_mostly = {
++ .typename = "local-APIC-edge",
++ .startup = NULL, /* startup_irq() not used for IRQ0 */
++ .shutdown = NULL, /* shutdown_irq() not used for IRQ0 */
++ .enable = enable_lapic_irq,
++ .disable = disable_lapic_irq,
++ .ack = ack_lapic_irq,
++ .end = end_lapic_irq,
++};
++
++static void setup_nmi (void)
++{
++ /*
++ * Dirty trick to enable the NMI watchdog ...
++ * We put the 8259A master into AEOI mode and
++ * unmask on all local APICs LVT0 as NMI.
++ *
++ * The idea to use the 8259A in AEOI mode ('8259A Virtual Wire')
++ * is from Maciej W. Rozycki - so we do not have to EOI from
++ * the NMI handler or the timer interrupt.
++ */
++ printk(KERN_INFO "activating NMI Watchdog ...");
++
++ enable_NMI_through_LVT0(NULL);
++
++ printk(" done.\n");
++}
++
++/*
++ * This looks a bit hackish but it's about the only one way of sending
++ * a few INTA cycles to 8259As and any associated glue logic. ICR does
++ * not support the ExtINT mode, unfortunately. We need to send these
++ * cycles as some i82489DX-based boards have glue logic that keeps the
++ * 8259A interrupt line asserted until INTA. --macro
++ */
++static inline void unlock_ExtINT_logic(void)
++{
++ int apic, pin, i;
++ struct IO_APIC_route_entry entry0, entry1;
++ unsigned char save_control, save_freq_select;
++ unsigned long flags;
++
++ pin = find_isa_irq_pin(8, mp_INT);
++ apic = find_isa_irq_apic(8, mp_INT);
++ if (pin == -1)
++ return;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ *(((int *)&entry0) + 1) = io_apic_read(apic, 0x11 + 2 * pin);
++ *(((int *)&entry0) + 0) = io_apic_read(apic, 0x10 + 2 * pin);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++ clear_IO_APIC_pin(apic, pin);
++
++ memset(&entry1, 0, sizeof(entry1));
++
++ entry1.dest_mode = 0; /* physical delivery */
++ entry1.mask = 0; /* unmask IRQ now */
++ entry1.dest.physical.physical_dest = hard_smp_processor_id();
++ entry1.delivery_mode = dest_ExtINT;
++ entry1.polarity = entry0.polarity;
++ entry1.trigger = 0;
++ entry1.vector = 0;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ io_apic_write(apic, 0x11 + 2 * pin, *(((int *)&entry1) + 1));
++ io_apic_write(apic, 0x10 + 2 * pin, *(((int *)&entry1) + 0));
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ save_control = CMOS_READ(RTC_CONTROL);
++ save_freq_select = CMOS_READ(RTC_FREQ_SELECT);
++ CMOS_WRITE((save_freq_select & ~RTC_RATE_SELECT) | 0x6,
++ RTC_FREQ_SELECT);
++ CMOS_WRITE(save_control | RTC_PIE, RTC_CONTROL);
++
++ i = 100;
++ while (i-- > 0) {
++ mdelay(10);
++ if ((CMOS_READ(RTC_INTR_FLAGS) & RTC_PF) == RTC_PF)
++ i -= 10;
++ }
++
++ CMOS_WRITE(save_control, RTC_CONTROL);
++ CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT);
++ clear_IO_APIC_pin(apic, pin);
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ io_apic_write(apic, 0x11 + 2 * pin, *(((int *)&entry0) + 1));
++ io_apic_write(apic, 0x10 + 2 * pin, *(((int *)&entry0) + 0));
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
++
++int timer_uses_ioapic_pin_0;
++
++/*
++ * This code may look a bit paranoid, but it's supposed to cooperate with
++ * a wide range of boards and BIOS bugs. Fortunately only the timer IRQ
++ * is so screwy. Thanks to Brian Perkins for testing/hacking this beast
++ * fanatically on his truly buggy board.
++ *
++ * FIXME: really need to revamp this for modern platforms only.
++ */
++static inline void check_timer(void)
++{
++ int apic1, pin1, apic2, pin2;
++ int vector;
++
++ /*
++ * get/set the timer IRQ vector:
++ */
++ disable_8259A_irq(0);
++ vector = assign_irq_vector(0);
++ set_intr_gate(vector, interrupt[0]);
++
++ /*
++ * Subtle, code in do_timer_interrupt() expects an AEOI
++ * mode for the 8259A whenever interrupts are routed
++ * through I/O APICs. Also IRQ0 has to be enabled in
++ * the 8259A which implies the virtual wire has to be
++ * disabled in the local APIC.
++ */
++ apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
++ init_8259A(1);
++ if (timer_over_8254 > 0)
++ enable_8259A_irq(0);
++
++ pin1 = find_isa_irq_pin(0, mp_INT);
++ apic1 = find_isa_irq_apic(0, mp_INT);
++ pin2 = ioapic_i8259.pin;
++ apic2 = ioapic_i8259.apic;
++
++ if (pin1 == 0)
++ timer_uses_ioapic_pin_0 = 1;
++
++ apic_printk(APIC_VERBOSE,KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n",
++ vector, apic1, pin1, apic2, pin2);
++
++ if (pin1 != -1) {
++ /*
++ * Ok, does IRQ0 through the IOAPIC work?
++ */
++ unmask_IO_APIC_irq(0);
++ if (!no_timer_check && timer_irq_works()) {
++ nmi_watchdog_default();
++ if (nmi_watchdog == NMI_IO_APIC) {
++ disable_8259A_irq(0);
++ setup_nmi();
++ enable_8259A_irq(0);
++ }
++ if (disable_timer_pin_1 > 0)
++ clear_IO_APIC_pin(0, pin1);
++ return;
++ }
++ clear_IO_APIC_pin(apic1, pin1);
++ apic_printk(APIC_QUIET,KERN_ERR "..MP-BIOS bug: 8254 timer not "
++ "connected to IO-APIC\n");
++ }
++
++ apic_printk(APIC_VERBOSE,KERN_INFO "...trying to set up timer (IRQ0) "
++ "through the 8259A ... ");
++ if (pin2 != -1) {
++ apic_printk(APIC_VERBOSE,"\n..... (found apic %d pin %d) ...",
++ apic2, pin2);
++ /*
++ * legacy devices should be connected to IO APIC #0
++ */
++ setup_ExtINT_IRQ0_pin(apic2, pin2, vector);
++ if (timer_irq_works()) {
++ apic_printk(APIC_VERBOSE," works.\n");
++ nmi_watchdog_default();
++ if (nmi_watchdog == NMI_IO_APIC) {
++ setup_nmi();
++ }
++ return;
++ }
++ /*
++ * Cleanup, just in case ...
++ */
++ clear_IO_APIC_pin(apic2, pin2);
++ }
++ apic_printk(APIC_VERBOSE," failed.\n");
++
++ if (nmi_watchdog == NMI_IO_APIC) {
++ printk(KERN_WARNING "timer doesn't work through the IO-APIC - disabling NMI Watchdog!\n");
++ nmi_watchdog = 0;
++ }
++
++ apic_printk(APIC_VERBOSE, KERN_INFO "...trying to set up timer as Virtual Wire IRQ...");
++
++ disable_8259A_irq(0);
++ irq_desc[0].chip = &lapic_irq_type;
++ apic_write(APIC_LVT0, APIC_DM_FIXED | vector); /* Fixed mode */
++ enable_8259A_irq(0);
++
++ if (timer_irq_works()) {
++ apic_printk(APIC_VERBOSE," works.\n");
++ return;
++ }
++ apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | vector);
++ apic_printk(APIC_VERBOSE," failed.\n");
++
++ apic_printk(APIC_VERBOSE, KERN_INFO "...trying to set up timer as ExtINT IRQ...");
++
++ init_8259A(0);
++ make_8259A_irq(0);
++ apic_write(APIC_LVT0, APIC_DM_EXTINT);
++
++ unlock_ExtINT_logic();
++
++ if (timer_irq_works()) {
++ apic_printk(APIC_VERBOSE," works.\n");
++ return;
++ }
++ apic_printk(APIC_VERBOSE," failed :(.\n");
++ panic("IO-APIC + timer doesn't work! Try using the 'noapic' kernel parameter\n");
++}
++#else
++#define check_timer() ((void)0)
++int timer_uses_ioapic_pin_0 = 0;
++#endif /* !CONFIG_XEN */
++
++static int __init notimercheck(char *s)
++{
++ no_timer_check = 1;
++ return 1;
++}
++__setup("no_timer_check", notimercheck);
++
++/*
++ *
++ * IRQ's that are handled by the PIC in the MPS IOAPIC case.
++ * - IRQ2 is the cascade IRQ, and cannot be a io-apic IRQ.
++ * Linux doesn't really care, as it's not actually used
++ * for any interrupt handling anyway.
++ */
++#define PIC_IRQS (1<<2)
++
++void __init setup_IO_APIC(void)
++{
++ enable_IO_APIC();
++
++ if (acpi_ioapic)
++ io_apic_irqs = ~0; /* all IRQs go through IOAPIC */
++ else
++ io_apic_irqs = ~PIC_IRQS;
++
++ apic_printk(APIC_VERBOSE, "ENABLING IO-APIC IRQs\n");
++
++ /*
++ * Set up the IO-APIC IRQ routing table.
++ */
++ if (!acpi_ioapic)
++ setup_ioapic_ids_from_mpc();
++#ifndef CONFIG_XEN
++ sync_Arb_IDs();
++#endif /* !CONFIG_XEN */
++ setup_IO_APIC_irqs();
++ init_IO_APIC_traps();
++ check_timer();
++ if (!acpi_ioapic)
++ print_IO_APIC();
++}
++
++struct sysfs_ioapic_data {
++ struct sys_device dev;
++ struct IO_APIC_route_entry entry[0];
++};
++static struct sysfs_ioapic_data * mp_ioapic_data[MAX_IO_APICS];
++
++static int ioapic_suspend(struct sys_device *dev, pm_message_t state)
++{
++ struct IO_APIC_route_entry *entry;
++ struct sysfs_ioapic_data *data;
++ unsigned long flags;
++ int i;
++
++ data = container_of(dev, struct sysfs_ioapic_data, dev);
++ entry = data->entry;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ for (i = 0; i < nr_ioapic_registers[dev->id]; i ++, entry ++ ) {
++ *(((int *)entry) + 1) = io_apic_read(dev->id, 0x11 + 2 * i);
++ *(((int *)entry) + 0) = io_apic_read(dev->id, 0x10 + 2 * i);
++ }
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ return 0;
++}
++
++static int ioapic_resume(struct sys_device *dev)
++{
++ struct IO_APIC_route_entry *entry;
++ struct sysfs_ioapic_data *data;
++ unsigned long flags;
++ union IO_APIC_reg_00 reg_00;
++ int i;
++
++ data = container_of(dev, struct sysfs_ioapic_data, dev);
++ entry = data->entry;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_00.raw = io_apic_read(dev->id, 0);
++ if (reg_00.bits.ID != mp_ioapics[dev->id].mpc_apicid) {
++ reg_00.bits.ID = mp_ioapics[dev->id].mpc_apicid;
++ io_apic_write(dev->id, 0, reg_00.raw);
++ }
++ for (i = 0; i < nr_ioapic_registers[dev->id]; i ++, entry ++ ) {
++ io_apic_write(dev->id, 0x11+2*i, *(((int *)entry)+1));
++ io_apic_write(dev->id, 0x10+2*i, *(((int *)entry)+0));
++ }
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ return 0;
++}
++
++static struct sysdev_class ioapic_sysdev_class = {
++ set_kset_name("ioapic"),
++ .suspend = ioapic_suspend,
++ .resume = ioapic_resume,
++};
++
++static int __init ioapic_init_sysfs(void)
++{
++ struct sys_device * dev;
++ int i, size, error = 0;
++
++ error = sysdev_class_register(&ioapic_sysdev_class);
++ if (error)
++ return error;
++
++ for (i = 0; i < nr_ioapics; i++ ) {
++ size = sizeof(struct sys_device) + nr_ioapic_registers[i]
++ * sizeof(struct IO_APIC_route_entry);
++ mp_ioapic_data[i] = kmalloc(size, GFP_KERNEL);
++ if (!mp_ioapic_data[i]) {
++ printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
++ continue;
++ }
++ memset(mp_ioapic_data[i], 0, size);
++ dev = &mp_ioapic_data[i]->dev;
++ dev->id = i;
++ dev->cls = &ioapic_sysdev_class;
++ error = sysdev_register(dev);
++ if (error) {
++ kfree(mp_ioapic_data[i]);
++ mp_ioapic_data[i] = NULL;
++ printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
++ continue;
++ }
++ }
++
++ return 0;
++}
++
++device_initcall(ioapic_init_sysfs);
++
++/* --------------------------------------------------------------------------
++ ACPI-based IOAPIC Configuration
++ -------------------------------------------------------------------------- */
++
++#ifdef CONFIG_ACPI
++
++#define IO_APIC_MAX_ID 0xFE
++
++int __init io_apic_get_version (int ioapic)
++{
++ union IO_APIC_reg_01 reg_01;
++ unsigned long flags;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_01.raw = io_apic_read(ioapic, 1);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ return reg_01.bits.version;
++}
++
++
++int __init io_apic_get_redir_entries (int ioapic)
++{
++ union IO_APIC_reg_01 reg_01;
++ unsigned long flags;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_01.raw = io_apic_read(ioapic, 1);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ return reg_01.bits.entries;
++}
++
++
++int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int active_high_low)
++{
++ struct IO_APIC_route_entry entry;
++ unsigned long flags;
++
++ if (!IO_APIC_IRQ(irq)) {
++ apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
++ ioapic);
++ return -EINVAL;
++ }
++
++ /*
++ * Generate a PCI IRQ routing entry and program the IOAPIC accordingly.
++ * Note that we mask (disable) IRQs now -- these get enabled when the
++ * corresponding device driver registers for this IRQ.
++ */
++
++ memset(&entry,0,sizeof(entry));
++
++ entry.delivery_mode = INT_DELIVERY_MODE;
++ entry.dest_mode = INT_DEST_MODE;
++ entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS);
++ entry.trigger = edge_level;
++ entry.polarity = active_high_low;
++ entry.mask = 1; /* Disabled (masked) */
++
++ irq = gsi_irq_sharing(irq);
++ /*
++ * IRQs < 16 are already in the irq_2_pin[] map
++ */
++ if (irq >= 16)
++ add_pin_to_irq(irq, ioapic, pin);
++
++ entry.vector = assign_irq_vector(irq);
++
++ apic_printk(APIC_VERBOSE,KERN_DEBUG "IOAPIC[%d]: Set PCI routing entry (%d-%d -> 0x%x -> "
++ "IRQ %d Mode:%i Active:%i)\n", ioapic,
++ mp_ioapics[ioapic].mpc_apicid, pin, entry.vector, irq,
++ edge_level, active_high_low);
++
++ ioapic_register_intr(irq, entry.vector, edge_level);
++
++ if (!ioapic && (irq < 16))
++ disable_8259A_irq(irq);
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ io_apic_write(ioapic, 0x11+2*pin, *(((int *)&entry)+1));
++ io_apic_write(ioapic, 0x10+2*pin, *(((int *)&entry)+0));
++ set_native_irq_info(use_pci_vector() ? entry.vector : irq, TARGET_CPUS);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ return 0;
++}
++
++#endif /* CONFIG_ACPI */
++
++
++#ifndef CONFIG_XEN
++/*
++ * This function currently is only a helper for the i386 smp boot process where
++ * we need to reprogram the ioredtbls to cater for the cpus which have come online
++ * so mask in all cases should simply be TARGET_CPUS
++ */
++#ifdef CONFIG_SMP
++void __init setup_ioapic_dest(void)
++{
++ int pin, ioapic, irq, irq_entry;
++
++ if (skip_ioapic_setup == 1)
++ return;
++
++ for (ioapic = 0; ioapic < nr_ioapics; ioapic++) {
++ for (pin = 0; pin < nr_ioapic_registers[ioapic]; pin++) {
++ irq_entry = find_irq_entry(ioapic, pin, mp_INT);
++ if (irq_entry == -1)
++ continue;
++ irq = pin_2_irq(irq_entry, ioapic, pin);
++ set_ioapic_affinity_irq(irq, TARGET_CPUS);
++ }
++
++ }
++}
++#endif
++#endif /* !CONFIG_XEN */
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/ioport-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/ioport-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,100 @@
++/*
++ * linux/arch/x86_64/kernel/ioport.c
++ *
++ * This contains the io-permission bitmap code - written by obz, with changes
++ * by Linus.
++ */
++
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/capability.h>
++#include <linux/errno.h>
++#include <linux/types.h>
++#include <linux/ioport.h>
++#include <linux/mm.h>
++#include <linux/smp.h>
++#include <linux/smp_lock.h>
++#include <linux/stddef.h>
++#include <linux/slab.h>
++#include <linux/thread_info.h>
++#include <xen/interface/physdev.h>
++
++/* Set EXTENT bits starting at BASE in BITMAP to value TURN_ON. */
++static void set_bitmap(unsigned long *bitmap, unsigned int base, unsigned int extent, int new_value)
++{
++ int i;
++
++ if (new_value)
++ for (i = base; i < base + extent; i++)
++ __set_bit(i, bitmap);
++ else
++ for (i = base; i < base + extent; i++)
++ clear_bit(i, bitmap);
++}
++
++/*
++ * this changes the io permissions bitmap in the current task.
++ */
++asmlinkage long sys_ioperm(unsigned long from, unsigned long num, int turn_on)
++{
++ struct thread_struct * t = ¤t->thread;
++ unsigned long *bitmap;
++ struct physdev_set_iobitmap set_iobitmap;
++
++ if ((from + num <= from) || (from + num > IO_BITMAP_BITS))
++ return -EINVAL;
++ if (turn_on && !capable(CAP_SYS_RAWIO))
++ return -EPERM;
++
++ /*
++ * If it's the first ioperm() call in this thread's lifetime, set the
++ * IO bitmap up. ioperm() is much less timing critical than clone(),
++ * this is why we delay this operation until now:
++ */
++ if (!t->io_bitmap_ptr) {
++ bitmap = kmalloc(IO_BITMAP_BYTES, GFP_KERNEL);
++ if (!bitmap)
++ return -ENOMEM;
++
++ memset(bitmap, 0xff, IO_BITMAP_BYTES);
++ t->io_bitmap_ptr = bitmap;
++
++ set_xen_guest_handle(set_iobitmap.bitmap, (char *)bitmap);
++ set_iobitmap.nr_ports = IO_BITMAP_BITS;
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap,
++ &set_iobitmap));
++ }
++
++ set_bitmap(t->io_bitmap_ptr, from, num, !turn_on);
++
++ return 0;
++}
++
++/*
++ * sys_iopl has to be used when you want to access the IO ports
++ * beyond the 0x3ff range: to get the full 65536 ports bitmapped
++ * you'd need 8kB of bitmaps/process, which is a bit excessive.
++ *
++ */
++
++asmlinkage long sys_iopl(unsigned int new_iopl, struct pt_regs *regs)
++{
++ unsigned int old_iopl = current->thread.iopl;
++ struct physdev_set_iopl set_iopl;
++
++ if (new_iopl > 3)
++ return -EINVAL;
++
++ /* Need "raw I/O" privileges for direct port access. */
++ if ((new_iopl > old_iopl) && !capable(CAP_SYS_RAWIO))
++ return -EPERM;
++
++ /* Change our version of the privilege levels. */
++ current->thread.iopl = new_iopl;
++
++ /* Force the change at ring 0. */
++ set_iopl.iopl = (new_iopl == 0) ? 1 : new_iopl;
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl));
++
++ return 0;
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/irq-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/irq-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,197 @@
++/*
++ * linux/arch/x86_64/kernel/irq.c
++ *
++ * Copyright (C) 1992, 1998 Linus Torvalds, Ingo Molnar
++ *
++ * This file contains the lowest level x86_64-specific interrupt
++ * entry and irq statistics code. All the remaining irq logic is
++ * done by the generic kernel/irq/ code and in the
++ * x86_64-specific irq controller code. (e.g. i8259.c and
++ * io_apic.c.)
++ */
++
++#include <linux/kernel_stat.h>
++#include <linux/interrupt.h>
++#include <linux/seq_file.h>
++#include <linux/module.h>
++#include <linux/delay.h>
++#include <asm/uaccess.h>
++#include <asm/io_apic.h>
++#include <asm/idle.h>
++
++atomic_t irq_err_count;
++#ifdef CONFIG_X86_IO_APIC
++#ifdef APIC_MISMATCH_DEBUG
++atomic_t irq_mis_count;
++#endif
++#endif
++
++#ifdef CONFIG_DEBUG_STACKOVERFLOW
++/*
++ * Probabilistic stack overflow check:
++ *
++ * Only check the stack in process context, because everything else
++ * runs on the big interrupt stacks. Checking reliably is too expensive,
++ * so we just check from interrupts.
++ */
++static inline void stack_overflow_check(struct pt_regs *regs)
++{
++ u64 curbase = (u64) current->thread_info;
++ static unsigned long warned = -60*HZ;
++
++ if (regs->rsp >= curbase && regs->rsp <= curbase + THREAD_SIZE &&
++ regs->rsp < curbase + sizeof(struct thread_info) + 128 &&
++ time_after(jiffies, warned + 60*HZ)) {
++ printk("do_IRQ: %s near stack overflow (cur:%Lx,rsp:%lx)\n",
++ current->comm, curbase, regs->rsp);
++ show_stack(NULL,NULL);
++ warned = jiffies;
++ }
++}
++#endif
++
++/*
++ * Generic, controller-independent functions:
++ */
++
++int show_interrupts(struct seq_file *p, void *v)
++{
++ int i = *(loff_t *) v, j;
++ struct irqaction * action;
++ unsigned long flags;
++
++ if (i == 0) {
++ seq_printf(p, " ");
++ for_each_online_cpu(j)
++ seq_printf(p, "CPU%-8d",j);
++ seq_putc(p, '\n');
++ }
++
++ if (i < NR_IRQS) {
++ spin_lock_irqsave(&irq_desc[i].lock, flags);
++ action = irq_desc[i].action;
++ if (!action)
++ goto skip;
++ seq_printf(p, "%3d: ",i);
++#ifndef CONFIG_SMP
++ seq_printf(p, "%10u ", kstat_irqs(i));
++#else
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
++#endif
++ seq_printf(p, " %14s", irq_desc[i].chip->typename);
++
++ seq_printf(p, " %s", action->name);
++ for (action=action->next; action; action = action->next)
++ seq_printf(p, ", %s", action->name);
++ seq_putc(p, '\n');
++skip:
++ spin_unlock_irqrestore(&irq_desc[i].lock, flags);
++ } else if (i == NR_IRQS) {
++ seq_printf(p, "NMI: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ", cpu_pda(j)->__nmi_count);
++ seq_putc(p, '\n');
++#ifdef CONFIG_X86_LOCAL_APIC
++ seq_printf(p, "LOC: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ", cpu_pda(j)->apic_timer_irqs);
++ seq_putc(p, '\n');
++#endif
++ seq_printf(p, "ERR: %10u\n", atomic_read(&irq_err_count));
++#ifdef CONFIG_X86_IO_APIC
++#ifdef APIC_MISMATCH_DEBUG
++ seq_printf(p, "MIS: %10u\n", atomic_read(&irq_mis_count));
++#endif
++#endif
++ }
++ return 0;
++}
++
++/*
++ * do_IRQ handles all normal device IRQ's (the special
++ * SMP cross-CPU interrupts have their own specific
++ * handlers).
++ */
++asmlinkage unsigned int do_IRQ(struct pt_regs *regs)
++{
++ /* high bit used in ret_from_ code */
++ unsigned irq = ~regs->orig_rax;
++
++ if (unlikely(irq >= NR_IRQS)) {
++ printk(KERN_EMERG "%s: cannot handle IRQ %d\n",
++ __FUNCTION__, irq);
++ BUG();
++ }
++
++ exit_idle();
++ irq_enter();
++#ifdef CONFIG_DEBUG_STACKOVERFLOW
++ stack_overflow_check(regs);
++#endif
++ __do_IRQ(irq, regs);
++ irq_exit();
++
++ return 1;
++}
++
++#ifdef CONFIG_HOTPLUG_CPU
++void fixup_irqs(cpumask_t map)
++{
++ unsigned int irq;
++ static int warned;
++
++ for (irq = 0; irq < NR_IRQS; irq++) {
++ cpumask_t mask;
++ if (irq == 2)
++ continue;
++
++ cpus_and(mask, irq_desc[irq].affinity, map);
++ if (any_online_cpu(mask) == NR_CPUS) {
++ /*printk("Breaking affinity for irq %i\n", irq);*/
++ mask = map;
++ }
++ if (irq_desc[irq].chip->set_affinity)
++ irq_desc[irq].chip->set_affinity(irq, mask);
++ else if (irq_desc[irq].action && !(warned++))
++ printk("Cannot set affinity for irq %i\n", irq);
++ }
++
++ /* That doesn't seem sufficient. Give it 1ms. */
++ local_irq_enable();
++ mdelay(1);
++ local_irq_disable();
++}
++#endif
++
++extern void call_softirq(void);
++
++asmlinkage void do_softirq(void)
++{
++ __u32 pending;
++ unsigned long flags;
++
++ if (in_interrupt())
++ return;
++
++ local_irq_save(flags);
++ pending = local_softirq_pending();
++ /* Switch to interrupt stack */
++ if (pending) {
++ call_softirq();
++ WARN_ON_ONCE(softirq_count());
++ }
++ local_irq_restore(flags);
++}
++EXPORT_SYMBOL(do_softirq);
++
++#ifndef CONFIG_X86_LOCAL_APIC
++/*
++ * 'what should we do if we get a hw irq event on an illegal vector'.
++ * each architecture has to answer this themselves.
++ */
++void ack_bad_irq(unsigned int irq)
++{
++ printk("unexpected IRQ trap at vector %02x\n", irq);
++}
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/ldt-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/ldt-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,282 @@
++/*
++ * linux/arch/x86_64/kernel/ldt.c
++ *
++ * Copyright (C) 1992 Krishna Balasubramanian and Linus Torvalds
++ * Copyright (C) 1999 Ingo Molnar <mingo at redhat.com>
++ * Copyright (C) 2002 Andi Kleen
++ *
++ * This handles calls from both 32bit and 64bit mode.
++ */
++
++#include <linux/errno.h>
++#include <linux/sched.h>
++#include <linux/string.h>
++#include <linux/mm.h>
++#include <linux/smp.h>
++#include <linux/smp_lock.h>
++#include <linux/vmalloc.h>
++#include <linux/slab.h>
++
++#include <asm/uaccess.h>
++#include <asm/system.h>
++#include <asm/ldt.h>
++#include <asm/desc.h>
++#include <asm/proto.h>
++#include <asm/pgalloc.h>
++
++#ifdef CONFIG_SMP /* avoids "defined but not used" warnig */
++static void flush_ldt(void *null)
++{
++ if (current->active_mm)
++ load_LDT(¤t->active_mm->context);
++}
++#endif
++
++static int alloc_ldt(mm_context_t *pc, unsigned mincount, int reload)
++{
++ void *oldldt;
++ void *newldt;
++ unsigned oldsize;
++
++ if (mincount <= (unsigned)pc->size)
++ return 0;
++ oldsize = pc->size;
++ mincount = (mincount+511)&(~511);
++ if (mincount*LDT_ENTRY_SIZE > PAGE_SIZE)
++ newldt = vmalloc(mincount*LDT_ENTRY_SIZE);
++ else
++ newldt = kmalloc(mincount*LDT_ENTRY_SIZE, GFP_KERNEL);
++
++ if (!newldt)
++ return -ENOMEM;
++
++ if (oldsize)
++ memcpy(newldt, pc->ldt, oldsize*LDT_ENTRY_SIZE);
++ oldldt = pc->ldt;
++ memset(newldt+oldsize*LDT_ENTRY_SIZE, 0, (mincount-oldsize)*LDT_ENTRY_SIZE);
++ wmb();
++ pc->ldt = newldt;
++ wmb();
++ pc->size = mincount;
++ wmb();
++ if (reload) {
++#ifdef CONFIG_SMP
++ cpumask_t mask;
++
++ preempt_disable();
++#endif
++ make_pages_readonly(
++ pc->ldt,
++ (pc->size * LDT_ENTRY_SIZE) / PAGE_SIZE,
++ XENFEAT_writable_descriptor_tables);
++ load_LDT(pc);
++#ifdef CONFIG_SMP
++ mask = cpumask_of_cpu(smp_processor_id());
++ if (!cpus_equal(current->mm->cpu_vm_mask, mask))
++ smp_call_function(flush_ldt, NULL, 1, 1);
++ preempt_enable();
++#endif
++ }
++ if (oldsize) {
++ make_pages_writable(
++ oldldt,
++ (oldsize * LDT_ENTRY_SIZE) / PAGE_SIZE,
++ XENFEAT_writable_descriptor_tables);
++ if (oldsize*LDT_ENTRY_SIZE > PAGE_SIZE)
++ vfree(oldldt);
++ else
++ kfree(oldldt);
++ }
++ return 0;
++}
++
++static inline int copy_ldt(mm_context_t *new, mm_context_t *old)
++{
++ int err = alloc_ldt(new, old->size, 0);
++ if (err < 0)
++ return err;
++ memcpy(new->ldt, old->ldt, old->size*LDT_ENTRY_SIZE);
++ make_pages_readonly(
++ new->ldt,
++ (new->size * LDT_ENTRY_SIZE) / PAGE_SIZE,
++ XENFEAT_writable_descriptor_tables);
++ return 0;
++}
++
++/*
++ * we do not have to muck with descriptors here, that is
++ * done in switch_mm() as needed.
++ */
++int init_new_context(struct task_struct *tsk, struct mm_struct *mm)
++{
++ struct mm_struct * old_mm;
++ int retval = 0;
++
++ memset(&mm->context, 0, sizeof(mm->context));
++ init_MUTEX(&mm->context.sem);
++ old_mm = current->mm;
++ if (old_mm && old_mm->context.size > 0) {
++ down(&old_mm->context.sem);
++ retval = copy_ldt(&mm->context, &old_mm->context);
++ up(&old_mm->context.sem);
++ }
++ if (retval == 0) {
++ spin_lock(&mm_unpinned_lock);
++ list_add(&mm->context.unpinned, &mm_unpinned);
++ spin_unlock(&mm_unpinned_lock);
++ }
++ return retval;
++}
++
++/*
++ *
++ * Don't touch the LDT register - we're already in the next thread.
++ */
++void destroy_context(struct mm_struct *mm)
++{
++ if (mm->context.size) {
++ if (mm == current->active_mm)
++ clear_LDT();
++ make_pages_writable(
++ mm->context.ldt,
++ (mm->context.size * LDT_ENTRY_SIZE) / PAGE_SIZE,
++ XENFEAT_writable_descriptor_tables);
++ if (mm->context.size*LDT_ENTRY_SIZE > PAGE_SIZE)
++ vfree(mm->context.ldt);
++ else
++ kfree(mm->context.ldt);
++ mm->context.size = 0;
++ }
++ if (!mm->context.pinned) {
++ spin_lock(&mm_unpinned_lock);
++ list_del(&mm->context.unpinned);
++ spin_unlock(&mm_unpinned_lock);
++ }
++}
++
++static int read_ldt(void __user * ptr, unsigned long bytecount)
++{
++ int err;
++ unsigned long size;
++ struct mm_struct * mm = current->mm;
++
++ if (!mm->context.size)
++ return 0;
++ if (bytecount > LDT_ENTRY_SIZE*LDT_ENTRIES)
++ bytecount = LDT_ENTRY_SIZE*LDT_ENTRIES;
++
++ down(&mm->context.sem);
++ size = mm->context.size*LDT_ENTRY_SIZE;
++ if (size > bytecount)
++ size = bytecount;
++
++ err = 0;
++ if (copy_to_user(ptr, mm->context.ldt, size))
++ err = -EFAULT;
++ up(&mm->context.sem);
++ if (err < 0)
++ goto error_return;
++ if (size != bytecount) {
++ /* zero-fill the rest */
++ if (clear_user(ptr+size, bytecount-size) != 0) {
++ err = -EFAULT;
++ goto error_return;
++ }
++ }
++ return bytecount;
++error_return:
++ return err;
++}
++
++static int read_default_ldt(void __user * ptr, unsigned long bytecount)
++{
++ /* Arbitrary number */
++ /* x86-64 default LDT is all zeros */
++ if (bytecount > 128)
++ bytecount = 128;
++ if (clear_user(ptr, bytecount))
++ return -EFAULT;
++ return bytecount;
++}
++
++static int write_ldt(void __user * ptr, unsigned long bytecount, int oldmode)
++{
++ struct task_struct *me = current;
++ struct mm_struct * mm = me->mm;
++ __u32 entry_1, entry_2, *lp;
++ unsigned long mach_lp;
++ int error;
++ struct user_desc ldt_info;
++
++ error = -EINVAL;
++
++ if (bytecount != sizeof(ldt_info))
++ goto out;
++ error = -EFAULT;
++ if (copy_from_user(&ldt_info, ptr, bytecount))
++ goto out;
++
++ error = -EINVAL;
++ if (ldt_info.entry_number >= LDT_ENTRIES)
++ goto out;
++ if (ldt_info.contents == 3) {
++ if (oldmode)
++ goto out;
++ if (ldt_info.seg_not_present == 0)
++ goto out;
++ }
++
++ down(&mm->context.sem);
++ if (ldt_info.entry_number >= (unsigned)mm->context.size) {
++ error = alloc_ldt(¤t->mm->context, ldt_info.entry_number+1, 1);
++ if (error < 0)
++ goto out_unlock;
++ }
++
++ lp = (__u32 *) ((ldt_info.entry_number << 3) + (char *) mm->context.ldt);
++ mach_lp = arbitrary_virt_to_machine(lp);
++
++ /* Allow LDTs to be cleared by the user. */
++ if (ldt_info.base_addr == 0 && ldt_info.limit == 0) {
++ if (oldmode || LDT_empty(&ldt_info)) {
++ entry_1 = 0;
++ entry_2 = 0;
++ goto install;
++ }
++ }
++
++ entry_1 = LDT_entry_a(&ldt_info);
++ entry_2 = LDT_entry_b(&ldt_info);
++ if (oldmode)
++ entry_2 &= ~(1 << 20);
++
++ /* Install the new entry ... */
++install:
++ error = HYPERVISOR_update_descriptor(mach_lp, (unsigned long)((entry_1 | (unsigned long) entry_2 << 32)));
++
++out_unlock:
++ up(&mm->context.sem);
++out:
++ return error;
++}
++
++asmlinkage int sys_modify_ldt(int func, void __user *ptr, unsigned long bytecount)
++{
++ int ret = -ENOSYS;
++
++ switch (func) {
++ case 0:
++ ret = read_ldt(ptr, bytecount);
++ break;
++ case 1:
++ ret = write_ldt(ptr, bytecount, 1);
++ break;
++ case 2:
++ ret = read_default_ldt(ptr, bytecount);
++ break;
++ case 0x11:
++ ret = write_ldt(ptr, bytecount, 0);
++ break;
++ }
++ return ret;
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/machine_kexec.c
+--- a/arch/x86_64/kernel/machine_kexec.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/kernel/machine_kexec.c Wed Sep 10 10:54:08 2008 +0100
+@@ -15,6 +15,128 @@
+ #include <asm/mmu_context.h>
+ #include <asm/io.h>
+
++#define PAGE_ALIGNED __attribute__ ((__aligned__(PAGE_SIZE)))
++static u64 kexec_pgd[512] PAGE_ALIGNED;
++static u64 kexec_pud0[512] PAGE_ALIGNED;
++static u64 kexec_pmd0[512] PAGE_ALIGNED;
++static u64 kexec_pte0[512] PAGE_ALIGNED;
++static u64 kexec_pud1[512] PAGE_ALIGNED;
++static u64 kexec_pmd1[512] PAGE_ALIGNED;
++static u64 kexec_pte1[512] PAGE_ALIGNED;
++
++#ifdef CONFIG_XEN
++
++/* In the case of Xen, override hypervisor functions to be able to create
++ * a regular identity mapping page table...
++ */
++
++#include <xen/interface/kexec.h>
++#include <xen/interface/memory.h>
++
++#define x__pmd(x) ((pmd_t) { (x) } )
++#define x__pud(x) ((pud_t) { (x) } )
++#define x__pgd(x) ((pgd_t) { (x) } )
++
++#define x_pmd_val(x) ((x).pmd)
++#define x_pud_val(x) ((x).pud)
++#define x_pgd_val(x) ((x).pgd)
++
++static inline void x_set_pmd(pmd_t *dst, pmd_t val)
++{
++ x_pmd_val(*dst) = x_pmd_val(val);
++}
++
++static inline void x_set_pud(pud_t *dst, pud_t val)
++{
++ x_pud_val(*dst) = phys_to_machine(x_pud_val(val));
++}
++
++static inline void x_pud_clear (pud_t *pud)
++{
++ x_pud_val(*pud) = 0;
++}
++
++static inline void x_set_pgd(pgd_t *dst, pgd_t val)
++{
++ x_pgd_val(*dst) = phys_to_machine(x_pgd_val(val));
++}
++
++static inline void x_pgd_clear (pgd_t * pgd)
++{
++ x_pgd_val(*pgd) = 0;
++}
++
++#define X__PAGE_KERNEL_LARGE_EXEC \
++ _PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_PSE
++#define X_KERNPG_TABLE _PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY
++
++#define __ma(x) (pfn_to_mfn(__pa((x)) >> PAGE_SHIFT) << PAGE_SHIFT)
++
++#if PAGES_NR > KEXEC_XEN_NO_PAGES
++#error PAGES_NR is greater than KEXEC_XEN_NO_PAGES - Xen support will break
++#endif
++
++#if PA_CONTROL_PAGE != 0
++#error PA_CONTROL_PAGE is non zero - Xen support will break
++#endif
++
++void machine_kexec_setup_load_arg(xen_kexec_image_t *xki, struct kimage *image)
++{
++ void *control_page;
++ void *table_page;
++
++ memset(xki->page_list, 0, sizeof(xki->page_list));
++
++ control_page = page_address(image->control_code_page) + PAGE_SIZE;
++ memcpy(control_page, relocate_kernel, PAGE_SIZE);
++
++ table_page = page_address(image->control_code_page);
++
++ xki->page_list[PA_CONTROL_PAGE] = __ma(control_page);
++ xki->page_list[PA_TABLE_PAGE] = __ma(table_page);
++
++ xki->page_list[PA_PGD] = __ma(kexec_pgd);
++ xki->page_list[PA_PUD_0] = __ma(kexec_pud0);
++ xki->page_list[PA_PUD_1] = __ma(kexec_pud1);
++ xki->page_list[PA_PMD_0] = __ma(kexec_pmd0);
++ xki->page_list[PA_PMD_1] = __ma(kexec_pmd1);
++ xki->page_list[PA_PTE_0] = __ma(kexec_pte0);
++ xki->page_list[PA_PTE_1] = __ma(kexec_pte1);
++}
++
++int __init machine_kexec_setup_resources(struct resource *hypervisor,
++ struct resource *phys_cpus,
++ int nr_phys_cpus)
++{
++ int k;
++
++ /* The per-cpu crash note resources belong to the hypervisor resource */
++ for (k = 0; k < nr_phys_cpus; k++)
++ request_resource(hypervisor, phys_cpus + k);
++
++ return 0;
++}
++
++void machine_kexec_register_resources(struct resource *res) { ; }
++
++#else /* CONFIG_XEN */
++
++#define x__pmd(x) __pmd(x)
++#define x__pud(x) __pud(x)
++#define x__pgd(x) __pgd(x)
++
++#define x_set_pmd(x, y) set_pmd(x, y)
++#define x_set_pud(x, y) set_pud(x, y)
++#define x_set_pgd(x, y) set_pgd(x, y)
++
++#define x_pud_clear(x) pud_clear(x)
++#define x_pgd_clear(x) pgd_clear(x)
++
++#define X__PAGE_KERNEL_LARGE_EXEC __PAGE_KERNEL_LARGE_EXEC
++#define X_KERNPG_TABLE _KERNPG_TABLE
++
++#endif /* CONFIG_XEN */
++
+ static void init_level2_page(pmd_t *level2p, unsigned long addr)
+ {
+ unsigned long end_addr;
+@@ -22,7 +144,7 @@
+ addr &= PAGE_MASK;
+ end_addr = addr + PUD_SIZE;
+ while (addr < end_addr) {
+- set_pmd(level2p++, __pmd(addr | __PAGE_KERNEL_LARGE_EXEC));
++ x_set_pmd(level2p++, x__pmd(addr | X__PAGE_KERNEL_LARGE_EXEC));
+ addr += PMD_SIZE;
+ }
+ }
+@@ -47,12 +169,12 @@
+ }
+ level2p = (pmd_t *)page_address(page);
+ init_level2_page(level2p, addr);
+- set_pud(level3p++, __pud(__pa(level2p) | _KERNPG_TABLE));
++ x_set_pud(level3p++, x__pud(__pa(level2p) | X_KERNPG_TABLE));
+ addr += PUD_SIZE;
+ }
+ /* clear the unused entries */
+ while (addr < end_addr) {
+- pud_clear(level3p++);
++ x_pud_clear(level3p++);
+ addr += PUD_SIZE;
+ }
+ out:
+@@ -83,12 +205,12 @@
+ if (result) {
+ goto out;
+ }
+- set_pgd(level4p++, __pgd(__pa(level3p) | _KERNPG_TABLE));
++ x_set_pgd(level4p++, x__pgd(__pa(level3p) | X_KERNPG_TABLE));
+ addr += PGDIR_SIZE;
+ }
+ /* clear the unused entries */
+ while (addr < end_addr) {
+- pgd_clear(level4p++);
++ x_pgd_clear(level4p++);
+ addr += PGDIR_SIZE;
+ }
+ out:
+@@ -99,76 +221,28 @@
+ static int init_pgtable(struct kimage *image, unsigned long start_pgtable)
+ {
+ pgd_t *level4p;
++ unsigned long x_end_pfn = end_pfn;
++
++#ifdef CONFIG_XEN
++ x_end_pfn = HYPERVISOR_memory_op(XENMEM_maximum_ram_page, NULL);
++#endif
++
+ level4p = (pgd_t *)__va(start_pgtable);
+- return init_level4_page(image, level4p, 0, end_pfn << PAGE_SHIFT);
++ return init_level4_page(image, level4p, 0, x_end_pfn << PAGE_SHIFT);
+ }
+-
+-static void set_idt(void *newidt, u16 limit)
+-{
+- struct desc_ptr curidt;
+-
+- /* x86-64 supports unaliged loads & stores */
+- curidt.size = limit;
+- curidt.address = (unsigned long)newidt;
+-
+- __asm__ __volatile__ (
+- "lidtq %0\n"
+- : : "m" (curidt)
+- );
+-};
+-
+-
+-static void set_gdt(void *newgdt, u16 limit)
+-{
+- struct desc_ptr curgdt;
+-
+- /* x86-64 supports unaligned loads & stores */
+- curgdt.size = limit;
+- curgdt.address = (unsigned long)newgdt;
+-
+- __asm__ __volatile__ (
+- "lgdtq %0\n"
+- : : "m" (curgdt)
+- );
+-};
+-
+-static void load_segments(void)
+-{
+- __asm__ __volatile__ (
+- "\tmovl %0,%%ds\n"
+- "\tmovl %0,%%es\n"
+- "\tmovl %0,%%ss\n"
+- "\tmovl %0,%%fs\n"
+- "\tmovl %0,%%gs\n"
+- : : "a" (__KERNEL_DS) : "memory"
+- );
+-}
+-
+-typedef NORET_TYPE void (*relocate_new_kernel_t)(unsigned long indirection_page,
+- unsigned long control_code_buffer,
+- unsigned long start_address,
+- unsigned long pgtable) ATTRIB_NORET;
+-
+-extern const unsigned char relocate_new_kernel[];
+-extern const unsigned long relocate_new_kernel_size;
+
+ int machine_kexec_prepare(struct kimage *image)
+ {
+- unsigned long start_pgtable, control_code_buffer;
++ unsigned long start_pgtable;
+ int result;
+
+ /* Calculate the offsets */
+ start_pgtable = page_to_pfn(image->control_code_page) << PAGE_SHIFT;
+- control_code_buffer = start_pgtable + PAGE_SIZE;
+
+ /* Setup the identity mapped 64bit page table */
+ result = init_pgtable(image, start_pgtable);
+ if (result)
+ return result;
+-
+- /* Place the code in the reboot code buffer */
+- memcpy(__va(control_code_buffer), relocate_new_kernel,
+- relocate_new_kernel_size);
+
+ return 0;
+ }
+@@ -178,51 +252,43 @@
+ return;
+ }
+
++#ifndef CONFIG_XEN
+ /*
+ * Do not allocate memory (or fail in any way) in machine_kexec().
+ * We are past the point of no return, committed to rebooting now.
+ */
+ NORET_TYPE void machine_kexec(struct kimage *image)
+ {
+- unsigned long page_list;
+- unsigned long control_code_buffer;
+- unsigned long start_pgtable;
+- relocate_new_kernel_t rnk;
++ unsigned long page_list[PAGES_NR];
++ void *control_page;
+
+ /* Interrupts aren't acceptable while we reboot */
+ local_irq_disable();
+
+- /* Calculate the offsets */
+- page_list = image->head;
+- start_pgtable = page_to_pfn(image->control_code_page) << PAGE_SHIFT;
+- control_code_buffer = start_pgtable + PAGE_SIZE;
++ control_page = page_address(image->control_code_page) + PAGE_SIZE;
++ memcpy(control_page, relocate_kernel, PAGE_SIZE);
+
+- /* Set the low half of the page table to my identity mapped
+- * page table for kexec. Leave the high half pointing at the
+- * kernel pages. Don't bother to flush the global pages
+- * as that will happen when I fully switch to my identity mapped
+- * page table anyway.
+- */
+- memcpy(__va(read_cr3()), __va(start_pgtable), PAGE_SIZE/2);
+- __flush_tlb();
++ page_list[PA_CONTROL_PAGE] = __pa(control_page);
++ page_list[VA_CONTROL_PAGE] = (unsigned long)relocate_kernel;
++ page_list[PA_PGD] = __pa(kexec_pgd);
++ page_list[VA_PGD] = (unsigned long)kexec_pgd;
++ page_list[PA_PUD_0] = __pa(kexec_pud0);
++ page_list[VA_PUD_0] = (unsigned long)kexec_pud0;
++ page_list[PA_PMD_0] = __pa(kexec_pmd0);
++ page_list[VA_PMD_0] = (unsigned long)kexec_pmd0;
++ page_list[PA_PTE_0] = __pa(kexec_pte0);
++ page_list[VA_PTE_0] = (unsigned long)kexec_pte0;
++ page_list[PA_PUD_1] = __pa(kexec_pud1);
++ page_list[VA_PUD_1] = (unsigned long)kexec_pud1;
++ page_list[PA_PMD_1] = __pa(kexec_pmd1);
++ page_list[VA_PMD_1] = (unsigned long)kexec_pmd1;
++ page_list[PA_PTE_1] = __pa(kexec_pte1);
++ page_list[VA_PTE_1] = (unsigned long)kexec_pte1;
+
++ page_list[PA_TABLE_PAGE] =
++ (unsigned long)__pa(page_address(image->control_code_page));
+
+- /* The segment registers are funny things, they have both a
+- * visible and an invisible part. Whenever the visible part is
+- * set to a specific selector, the invisible part is loaded
+- * with from a table in memory. At no other time is the
+- * descriptor table in memory accessed.
+- *
+- * I take advantage of this here by force loading the
+- * segments, before I zap the gdt with an invalid value.
+- */
+- load_segments();
+- /* The gdt & idt are now invalid.
+- * If you want to load them you must set up your own idt & gdt.
+- */
+- set_gdt(phys_to_virt(0),0);
+- set_idt(phys_to_virt(0),0);
+- /* now call it */
+- rnk = (relocate_new_kernel_t) control_code_buffer;
+- (*rnk)(page_list, control_code_buffer, image->start, start_pgtable);
++ relocate_kernel((unsigned long)image->head, (unsigned long)page_list,
++ image->start);
+ }
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/mpparse-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/mpparse-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1011 @@
++/*
++ * Intel Multiprocessor Specification 1.1 and 1.4
++ * compliant MP-table parsing routines.
++ *
++ * (c) 1995 Alan Cox, Building #3 <alan at redhat.com>
++ * (c) 1998, 1999, 2000 Ingo Molnar <mingo at redhat.com>
++ *
++ * Fixes
++ * Erich Boleyn : MP v1.4 and additional changes.
++ * Alan Cox : Added EBDA scanning
++ * Ingo Molnar : various cleanups and rewrites
++ * Maciej W. Rozycki: Bits for default MP configurations
++ * Paul Diefenbaugh: Added full ACPI support
++ */
++
++#include <linux/mm.h>
++#include <linux/init.h>
++#include <linux/delay.h>
++#include <linux/bootmem.h>
++#include <linux/smp_lock.h>
++#include <linux/kernel_stat.h>
++#include <linux/mc146818rtc.h>
++#include <linux/acpi.h>
++#include <linux/module.h>
++
++#include <asm/smp.h>
++#include <asm/mtrr.h>
++#include <asm/mpspec.h>
++#include <asm/pgalloc.h>
++#include <asm/io_apic.h>
++#include <asm/proto.h>
++#include <asm/acpi.h>
++
++/* Have we found an MP table */
++int smp_found_config;
++unsigned int __initdata maxcpus = NR_CPUS;
++
++int acpi_found_madt;
++
++/*
++ * Various Linux-internal data structures created from the
++ * MP-table.
++ */
++unsigned char apic_version [MAX_APICS];
++unsigned char mp_bus_id_to_type [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 };
++int mp_bus_id_to_pci_bus [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 };
++
++static int mp_current_pci_id = 0;
++/* I/O APIC entries */
++struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS];
++
++/* # of MP IRQ source entries */
++struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES];
++
++/* MP IRQ source entries */
++int mp_irq_entries;
++
++int nr_ioapics;
++int pic_mode;
++unsigned long mp_lapic_addr = 0;
++
++
++
++/* Processor that is doing the boot up */
++unsigned int boot_cpu_id = -1U;
++/* Internal processor count */
++unsigned int num_processors __initdata = 0;
++
++unsigned disabled_cpus __initdata;
++
++/* Bitmask of physically existing CPUs */
++physid_mask_t phys_cpu_present_map = PHYSID_MASK_NONE;
++
++/* ACPI MADT entry parsing functions */
++#ifdef CONFIG_ACPI
++extern struct acpi_boot_flags acpi_boot;
++#ifdef CONFIG_X86_LOCAL_APIC
++extern int acpi_parse_lapic (acpi_table_entry_header *header);
++extern int acpi_parse_lapic_addr_ovr (acpi_table_entry_header *header);
++extern int acpi_parse_lapic_nmi (acpi_table_entry_header *header);
++#endif /*CONFIG_X86_LOCAL_APIC*/
++#ifdef CONFIG_X86_IO_APIC
++extern int acpi_parse_ioapic (acpi_table_entry_header *header);
++#endif /*CONFIG_X86_IO_APIC*/
++#endif /*CONFIG_ACPI*/
++
++u8 bios_cpu_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = BAD_APICID };
++
++
++/*
++ * Intel MP BIOS table parsing routines:
++ */
++
++/*
++ * Checksum an MP configuration block.
++ */
++
++static int __init mpf_checksum(unsigned char *mp, int len)
++{
++ int sum = 0;
++
++ while (len--)
++ sum += *mp++;
++
++ return sum & 0xFF;
++}
++
++#ifndef CONFIG_XEN
++static void __cpuinit MP_processor_info (struct mpc_config_processor *m)
++{
++ int cpu;
++ unsigned char ver;
++ cpumask_t tmp_map;
++
++ if (!(m->mpc_cpuflag & CPU_ENABLED)) {
++ disabled_cpus++;
++ return;
++ }
++
++ printk(KERN_INFO "Processor #%d %d:%d APIC version %d\n",
++ m->mpc_apicid,
++ (m->mpc_cpufeature & CPU_FAMILY_MASK)>>8,
++ (m->mpc_cpufeature & CPU_MODEL_MASK)>>4,
++ m->mpc_apicver);
++
++ if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) {
++ Dprintk(" Bootup CPU\n");
++ boot_cpu_id = m->mpc_apicid;
++ }
++ if (num_processors >= NR_CPUS) {
++ printk(KERN_WARNING "WARNING: NR_CPUS limit of %i reached."
++ " Processor ignored.\n", NR_CPUS);
++ return;
++ }
++
++ num_processors++;
++ cpus_complement(tmp_map, cpu_present_map);
++ cpu = first_cpu(tmp_map);
++
++#if MAX_APICS < 255
++ if ((int)m->mpc_apicid > MAX_APICS) {
++ printk(KERN_ERR "Processor #%d INVALID. (Max ID: %d).\n",
++ m->mpc_apicid, MAX_APICS);
++ return;
++ }
++#endif
++ ver = m->mpc_apicver;
++
++ physid_set(m->mpc_apicid, phys_cpu_present_map);
++ /*
++ * Validate version
++ */
++ if (ver == 0x0) {
++ printk(KERN_ERR "BIOS bug, APIC version is 0 for CPU#%d! fixing up to 0x10. (tell your hw vendor)\n", m->mpc_apicid);
++ ver = 0x10;
++ }
++ apic_version[m->mpc_apicid] = ver;
++ if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) {
++ /*
++ * bios_cpu_apicid is required to have processors listed
++ * in same order as logical cpu numbers. Hence the first
++ * entry is BSP, and so on.
++ */
++ cpu = 0;
++ }
++ bios_cpu_apicid[cpu] = m->mpc_apicid;
++ x86_cpu_to_apicid[cpu] = m->mpc_apicid;
++
++ cpu_set(cpu, cpu_possible_map);
++ cpu_set(cpu, cpu_present_map);
++}
++#else
++static void __cpuinit MP_processor_info (struct mpc_config_processor *m)
++{
++ num_processors++;
++}
++#endif /* CONFIG_XEN */
++
++static void __init MP_bus_info (struct mpc_config_bus *m)
++{
++ char str[7];
++
++ memcpy(str, m->mpc_bustype, 6);
++ str[6] = 0;
++ Dprintk("Bus #%d is %s\n", m->mpc_busid, str);
++
++ if (strncmp(str, "ISA", 3) == 0) {
++ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_ISA;
++ } else if (strncmp(str, "EISA", 4) == 0) {
++ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_EISA;
++ } else if (strncmp(str, "PCI", 3) == 0) {
++ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_PCI;
++ mp_bus_id_to_pci_bus[m->mpc_busid] = mp_current_pci_id;
++ mp_current_pci_id++;
++ } else if (strncmp(str, "MCA", 3) == 0) {
++ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_MCA;
++ } else {
++ printk(KERN_ERR "Unknown bustype %s\n", str);
++ }
++}
++
++static void __init MP_ioapic_info (struct mpc_config_ioapic *m)
++{
++ if (!(m->mpc_flags & MPC_APIC_USABLE))
++ return;
++
++ printk("I/O APIC #%d Version %d at 0x%X.\n",
++ m->mpc_apicid, m->mpc_apicver, m->mpc_apicaddr);
++ if (nr_ioapics >= MAX_IO_APICS) {
++ printk(KERN_ERR "Max # of I/O APICs (%d) exceeded (found %d).\n",
++ MAX_IO_APICS, nr_ioapics);
++ panic("Recompile kernel with bigger MAX_IO_APICS!.\n");
++ }
++ if (!m->mpc_apicaddr) {
++ printk(KERN_ERR "WARNING: bogus zero I/O APIC address"
++ " found in MP table, skipping!\n");
++ return;
++ }
++ mp_ioapics[nr_ioapics] = *m;
++ nr_ioapics++;
++}
++
++static void __init MP_intsrc_info (struct mpc_config_intsrc *m)
++{
++ mp_irqs [mp_irq_entries] = *m;
++ Dprintk("Int: type %d, pol %d, trig %d, bus %d,"
++ " IRQ %02x, APIC ID %x, APIC INT %02x\n",
++ m->mpc_irqtype, m->mpc_irqflag & 3,
++ (m->mpc_irqflag >> 2) & 3, m->mpc_srcbus,
++ m->mpc_srcbusirq, m->mpc_dstapic, m->mpc_dstirq);
++ if (++mp_irq_entries >= MAX_IRQ_SOURCES)
++ panic("Max # of irq sources exceeded!!\n");
++}
++
++static void __init MP_lintsrc_info (struct mpc_config_lintsrc *m)
++{
++ Dprintk("Lint: type %d, pol %d, trig %d, bus %d,"
++ " IRQ %02x, APIC ID %x, APIC LINT %02x\n",
++ m->mpc_irqtype, m->mpc_irqflag & 3,
++ (m->mpc_irqflag >> 2) &3, m->mpc_srcbusid,
++ m->mpc_srcbusirq, m->mpc_destapic, m->mpc_destapiclint);
++ /*
++ * Well it seems all SMP boards in existence
++ * use ExtINT/LVT1 == LINT0 and
++ * NMI/LVT2 == LINT1 - the following check
++ * will show us if this assumptions is false.
++ * Until then we do not have to add baggage.
++ */
++ if ((m->mpc_irqtype == mp_ExtINT) &&
++ (m->mpc_destapiclint != 0))
++ BUG();
++ if ((m->mpc_irqtype == mp_NMI) &&
++ (m->mpc_destapiclint != 1))
++ BUG();
++}
++
++/*
++ * Read/parse the MPC
++ */
++
++static int __init smp_read_mpc(struct mp_config_table *mpc)
++{
++ char str[16];
++ int count=sizeof(*mpc);
++ unsigned char *mpt=((unsigned char *)mpc)+count;
++
++ if (memcmp(mpc->mpc_signature,MPC_SIGNATURE,4)) {
++ printk("SMP mptable: bad signature [%c%c%c%c]!\n",
++ mpc->mpc_signature[0],
++ mpc->mpc_signature[1],
++ mpc->mpc_signature[2],
++ mpc->mpc_signature[3]);
++ return 0;
++ }
++ if (mpf_checksum((unsigned char *)mpc,mpc->mpc_length)) {
++ printk("SMP mptable: checksum error!\n");
++ return 0;
++ }
++ if (mpc->mpc_spec!=0x01 && mpc->mpc_spec!=0x04) {
++ printk(KERN_ERR "SMP mptable: bad table version (%d)!!\n",
++ mpc->mpc_spec);
++ return 0;
++ }
++ if (!mpc->mpc_lapic) {
++ printk(KERN_ERR "SMP mptable: null local APIC address!\n");
++ return 0;
++ }
++ memcpy(str,mpc->mpc_oem,8);
++ str[8]=0;
++ printk(KERN_INFO "OEM ID: %s ",str);
++
++ memcpy(str,mpc->mpc_productid,12);
++ str[12]=0;
++ printk("Product ID: %s ",str);
++
++ printk("APIC at: 0x%X\n",mpc->mpc_lapic);
++
++ /* save the local APIC address, it might be non-default */
++ if (!acpi_lapic)
++ mp_lapic_addr = mpc->mpc_lapic;
++
++ /*
++ * Now process the configuration blocks.
++ */
++ while (count < mpc->mpc_length) {
++ switch(*mpt) {
++ case MP_PROCESSOR:
++ {
++ struct mpc_config_processor *m=
++ (struct mpc_config_processor *)mpt;
++ if (!acpi_lapic)
++ MP_processor_info(m);
++ mpt += sizeof(*m);
++ count += sizeof(*m);
++ break;
++ }
++ case MP_BUS:
++ {
++ struct mpc_config_bus *m=
++ (struct mpc_config_bus *)mpt;
++ MP_bus_info(m);
++ mpt += sizeof(*m);
++ count += sizeof(*m);
++ break;
++ }
++ case MP_IOAPIC:
++ {
++ struct mpc_config_ioapic *m=
++ (struct mpc_config_ioapic *)mpt;
++ MP_ioapic_info(m);
++ mpt+=sizeof(*m);
++ count+=sizeof(*m);
++ break;
++ }
++ case MP_INTSRC:
++ {
++ struct mpc_config_intsrc *m=
++ (struct mpc_config_intsrc *)mpt;
++
++ MP_intsrc_info(m);
++ mpt+=sizeof(*m);
++ count+=sizeof(*m);
++ break;
++ }
++ case MP_LINTSRC:
++ {
++ struct mpc_config_lintsrc *m=
++ (struct mpc_config_lintsrc *)mpt;
++ MP_lintsrc_info(m);
++ mpt+=sizeof(*m);
++ count+=sizeof(*m);
++ break;
++ }
++ }
++ }
++ clustered_apic_check();
++ if (!num_processors)
++ printk(KERN_ERR "SMP mptable: no processors registered!\n");
++ return num_processors;
++}
++
++static int __init ELCR_trigger(unsigned int irq)
++{
++ unsigned int port;
++
++ port = 0x4d0 + (irq >> 3);
++ return (inb(port) >> (irq & 7)) & 1;
++}
++
++static void __init construct_default_ioirq_mptable(int mpc_default_type)
++{
++ struct mpc_config_intsrc intsrc;
++ int i;
++ int ELCR_fallback = 0;
++
++ intsrc.mpc_type = MP_INTSRC;
++ intsrc.mpc_irqflag = 0; /* conforming */
++ intsrc.mpc_srcbus = 0;
++ intsrc.mpc_dstapic = mp_ioapics[0].mpc_apicid;
++
++ intsrc.mpc_irqtype = mp_INT;
++
++ /*
++ * If true, we have an ISA/PCI system with no IRQ entries
++ * in the MP table. To prevent the PCI interrupts from being set up
++ * incorrectly, we try to use the ELCR. The sanity check to see if
++ * there is good ELCR data is very simple - IRQ0, 1, 2 and 13 can
++ * never be level sensitive, so we simply see if the ELCR agrees.
++ * If it does, we assume it's valid.
++ */
++ if (mpc_default_type == 5) {
++ printk(KERN_INFO "ISA/PCI bus type with no IRQ information... falling back to ELCR\n");
++
++ if (ELCR_trigger(0) || ELCR_trigger(1) || ELCR_trigger(2) || ELCR_trigger(13))
++ printk(KERN_ERR "ELCR contains invalid data... not using ELCR\n");
++ else {
++ printk(KERN_INFO "Using ELCR to identify PCI interrupts\n");
++ ELCR_fallback = 1;
++ }
++ }
++
++ for (i = 0; i < 16; i++) {
++ switch (mpc_default_type) {
++ case 2:
++ if (i == 0 || i == 13)
++ continue; /* IRQ0 & IRQ13 not connected */
++ /* fall through */
++ default:
++ if (i == 2)
++ continue; /* IRQ2 is never connected */
++ }
++
++ if (ELCR_fallback) {
++ /*
++ * If the ELCR indicates a level-sensitive interrupt, we
++ * copy that information over to the MP table in the
++ * irqflag field (level sensitive, active high polarity).
++ */
++ if (ELCR_trigger(i))
++ intsrc.mpc_irqflag = 13;
++ else
++ intsrc.mpc_irqflag = 0;
++ }
++
++ intsrc.mpc_srcbusirq = i;
++ intsrc.mpc_dstirq = i ? i : 2; /* IRQ0 to INTIN2 */
++ MP_intsrc_info(&intsrc);
++ }
++
++ intsrc.mpc_irqtype = mp_ExtINT;
++ intsrc.mpc_srcbusirq = 0;
++ intsrc.mpc_dstirq = 0; /* 8259A to INTIN0 */
++ MP_intsrc_info(&intsrc);
++}
++
++static inline void __init construct_default_ISA_mptable(int mpc_default_type)
++{
++ struct mpc_config_processor processor;
++ struct mpc_config_bus bus;
++ struct mpc_config_ioapic ioapic;
++ struct mpc_config_lintsrc lintsrc;
++ int linttypes[2] = { mp_ExtINT, mp_NMI };
++ int i;
++
++ /*
++ * local APIC has default address
++ */
++ mp_lapic_addr = APIC_DEFAULT_PHYS_BASE;
++
++ /*
++ * 2 CPUs, numbered 0 & 1.
++ */
++ processor.mpc_type = MP_PROCESSOR;
++ /* Either an integrated APIC or a discrete 82489DX. */
++ processor.mpc_apicver = mpc_default_type > 4 ? 0x10 : 0x01;
++ processor.mpc_cpuflag = CPU_ENABLED;
++ processor.mpc_cpufeature = (boot_cpu_data.x86 << 8) |
++ (boot_cpu_data.x86_model << 4) |
++ boot_cpu_data.x86_mask;
++ processor.mpc_featureflag = boot_cpu_data.x86_capability[0];
++ processor.mpc_reserved[0] = 0;
++ processor.mpc_reserved[1] = 0;
++ for (i = 0; i < 2; i++) {
++ processor.mpc_apicid = i;
++ MP_processor_info(&processor);
++ }
++
++ bus.mpc_type = MP_BUS;
++ bus.mpc_busid = 0;
++ switch (mpc_default_type) {
++ default:
++ printk(KERN_ERR "???\nUnknown standard configuration %d\n",
++ mpc_default_type);
++ /* fall through */
++ case 1:
++ case 5:
++ memcpy(bus.mpc_bustype, "ISA ", 6);
++ break;
++ case 2:
++ case 6:
++ case 3:
++ memcpy(bus.mpc_bustype, "EISA ", 6);
++ break;
++ case 4:
++ case 7:
++ memcpy(bus.mpc_bustype, "MCA ", 6);
++ }
++ MP_bus_info(&bus);
++ if (mpc_default_type > 4) {
++ bus.mpc_busid = 1;
++ memcpy(bus.mpc_bustype, "PCI ", 6);
++ MP_bus_info(&bus);
++ }
++
++ ioapic.mpc_type = MP_IOAPIC;
++ ioapic.mpc_apicid = 2;
++ ioapic.mpc_apicver = mpc_default_type > 4 ? 0x10 : 0x01;
++ ioapic.mpc_flags = MPC_APIC_USABLE;
++ ioapic.mpc_apicaddr = 0xFEC00000;
++ MP_ioapic_info(&ioapic);
++
++ /*
++ * We set up most of the low 16 IO-APIC pins according to MPS rules.
++ */
++ construct_default_ioirq_mptable(mpc_default_type);
++
++ lintsrc.mpc_type = MP_LINTSRC;
++ lintsrc.mpc_irqflag = 0; /* conforming */
++ lintsrc.mpc_srcbusid = 0;
++ lintsrc.mpc_srcbusirq = 0;
++ lintsrc.mpc_destapic = MP_APIC_ALL;
++ for (i = 0; i < 2; i++) {
++ lintsrc.mpc_irqtype = linttypes[i];
++ lintsrc.mpc_destapiclint = i;
++ MP_lintsrc_info(&lintsrc);
++ }
++}
++
++static struct intel_mp_floating *mpf_found;
++
++/*
++ * Scan the memory blocks for an SMP configuration block.
++ */
++void __init get_smp_config (void)
++{
++ struct intel_mp_floating *mpf = mpf_found;
++
++ /*
++ * ACPI supports both logical (e.g. Hyper-Threading) and physical
++ * processors, where MPS only supports physical.
++ */
++ if (acpi_lapic && acpi_ioapic) {
++ printk(KERN_INFO "Using ACPI (MADT) for SMP configuration information\n");
++ return;
++ }
++ else if (acpi_lapic)
++ printk(KERN_INFO "Using ACPI for processor (LAPIC) configuration information\n");
++
++ printk("Intel MultiProcessor Specification v1.%d\n", mpf->mpf_specification);
++ if (mpf->mpf_feature2 & (1<<7)) {
++ printk(KERN_INFO " IMCR and PIC compatibility mode.\n");
++ pic_mode = 1;
++ } else {
++ printk(KERN_INFO " Virtual Wire compatibility mode.\n");
++ pic_mode = 0;
++ }
++
++ /*
++ * Now see if we need to read further.
++ */
++ if (mpf->mpf_feature1 != 0) {
++
++ printk(KERN_INFO "Default MP configuration #%d\n", mpf->mpf_feature1);
++ construct_default_ISA_mptable(mpf->mpf_feature1);
++
++ } else if (mpf->mpf_physptr) {
++
++ /*
++ * Read the physical hardware table. Anything here will
++ * override the defaults.
++ */
++ if (!smp_read_mpc(isa_bus_to_virt(mpf->mpf_physptr))) {
++ smp_found_config = 0;
++ printk(KERN_ERR "BIOS bug, MP table errors detected!...\n");
++ printk(KERN_ERR "... disabling SMP support. (tell your hw vendor)\n");
++ return;
++ }
++ /*
++ * If there are no explicit MP IRQ entries, then we are
++ * broken. We set up most of the low 16 IO-APIC pins to
++ * ISA defaults and hope it will work.
++ */
++ if (!mp_irq_entries) {
++ struct mpc_config_bus bus;
++
++ printk(KERN_ERR "BIOS bug, no explicit IRQ entries, using default mptable. (tell your hw vendor)\n");
++
++ bus.mpc_type = MP_BUS;
++ bus.mpc_busid = 0;
++ memcpy(bus.mpc_bustype, "ISA ", 6);
++ MP_bus_info(&bus);
++
++ construct_default_ioirq_mptable(0);
++ }
++
++ } else
++ BUG();
++
++ printk(KERN_INFO "Processors: %d\n", num_processors);
++ /*
++ * Only use the first configuration found.
++ */
++}
++
++static int __init smp_scan_config (unsigned long base, unsigned long length)
++{
++ extern void __bad_mpf_size(void);
++ unsigned int *bp = isa_bus_to_virt(base);
++ struct intel_mp_floating *mpf;
++
++ Dprintk("Scan SMP from %p for %ld bytes.\n", bp,length);
++ if (sizeof(*mpf) != 16)
++ __bad_mpf_size();
++
++ while (length > 0) {
++ mpf = (struct intel_mp_floating *)bp;
++ if ((*bp == SMP_MAGIC_IDENT) &&
++ (mpf->mpf_length == 1) &&
++ !mpf_checksum((unsigned char *)bp, 16) &&
++ ((mpf->mpf_specification == 1)
++ || (mpf->mpf_specification == 4)) ) {
++
++ smp_found_config = 1;
++ mpf_found = mpf;
++ return 1;
++ }
++ bp += 4;
++ length -= 16;
++ }
++ return 0;
++}
++
++void __init find_intel_smp (void)
++{
++ unsigned int address;
++
++ /*
++ * FIXME: Linux assumes you have 640K of base ram..
++ * this continues the error...
++ *
++ * 1) Scan the bottom 1K for a signature
++ * 2) Scan the top 1K of base RAM
++ * 3) Scan the 64K of bios
++ */
++ if (smp_scan_config(0x0,0x400) ||
++ smp_scan_config(639*0x400,0x400) ||
++ smp_scan_config(0xF0000,0x10000))
++ return;
++ /*
++ * If it is an SMP machine we should know now, unless the
++ * configuration is in an EISA/MCA bus machine with an
++ * extended bios data area.
++ *
++ * there is a real-mode segmented pointer pointing to the
++ * 4K EBDA area at 0x40E, calculate and scan it here.
++ *
++ * NOTE! There are Linux loaders that will corrupt the EBDA
++ * area, and as such this kind of SMP config may be less
++ * trustworthy, simply because the SMP table may have been
++ * stomped on during early boot. These loaders are buggy and
++ * should be fixed.
++ */
++
++ address = *(unsigned short *)phys_to_virt(0x40E);
++ address <<= 4;
++ if (smp_scan_config(address, 0x1000))
++ return;
++
++ /* If we have come this far, we did not find an MP table */
++ printk(KERN_INFO "No mptable found.\n");
++}
++
++/*
++ * - Intel MP Configuration Table
++ */
++void __init find_smp_config (void)
++{
++#ifdef CONFIG_X86_LOCAL_APIC
++ find_intel_smp();
++#endif
++}
++
++
++/* --------------------------------------------------------------------------
++ ACPI-based MP Configuration
++ -------------------------------------------------------------------------- */
++
++#ifdef CONFIG_ACPI
++
++void __init mp_register_lapic_address (
++ u64 address)
++{
++#ifndef CONFIG_XEN
++ mp_lapic_addr = (unsigned long) address;
++
++ set_fixmap_nocache(FIX_APIC_BASE, mp_lapic_addr);
++
++ if (boot_cpu_id == -1U)
++ boot_cpu_id = GET_APIC_ID(apic_read(APIC_ID));
++
++ Dprintk("Boot CPU = %d\n", boot_cpu_physical_apicid);
++#endif
++}
++
++
++void __cpuinit mp_register_lapic (
++ u8 id,
++ u8 enabled)
++{
++ struct mpc_config_processor processor;
++ int boot_cpu = 0;
++
++ if (id >= MAX_APICS) {
++ printk(KERN_WARNING "Processor #%d invalid (max %d)\n",
++ id, MAX_APICS);
++ return;
++ }
++
++ if (id == boot_cpu_physical_apicid)
++ boot_cpu = 1;
++
++#ifndef CONFIG_XEN
++ processor.mpc_type = MP_PROCESSOR;
++ processor.mpc_apicid = id;
++ processor.mpc_apicver = GET_APIC_VERSION(apic_read(APIC_LVR));
++ processor.mpc_cpuflag = (enabled ? CPU_ENABLED : 0);
++ processor.mpc_cpuflag |= (boot_cpu ? CPU_BOOTPROCESSOR : 0);
++ processor.mpc_cpufeature = (boot_cpu_data.x86 << 8) |
++ (boot_cpu_data.x86_model << 4) | boot_cpu_data.x86_mask;
++ processor.mpc_featureflag = boot_cpu_data.x86_capability[0];
++ processor.mpc_reserved[0] = 0;
++ processor.mpc_reserved[1] = 0;
++#endif
++
++ MP_processor_info(&processor);
++}
++
++#ifdef CONFIG_X86_IO_APIC
++
++#define MP_ISA_BUS 0
++#define MP_MAX_IOAPIC_PIN 127
++
++static struct mp_ioapic_routing {
++ int apic_id;
++ int gsi_start;
++ int gsi_end;
++ u32 pin_programmed[4];
++} mp_ioapic_routing[MAX_IO_APICS];
++
++
++static int mp_find_ioapic (
++ int gsi)
++{
++ int i = 0;
++
++ /* Find the IOAPIC that manages this GSI. */
++ for (i = 0; i < nr_ioapics; i++) {
++ if ((gsi >= mp_ioapic_routing[i].gsi_start)
++ && (gsi <= mp_ioapic_routing[i].gsi_end))
++ return i;
++ }
++
++ printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi);
++
++ return -1;
++}
++
++
++void __init mp_register_ioapic (
++ u8 id,
++ u32 address,
++ u32 gsi_base)
++{
++ int idx = 0;
++
++ if (nr_ioapics >= MAX_IO_APICS) {
++ printk(KERN_ERR "ERROR: Max # of I/O APICs (%d) exceeded "
++ "(found %d)\n", MAX_IO_APICS, nr_ioapics);
++ panic("Recompile kernel with bigger MAX_IO_APICS!\n");
++ }
++ if (!address) {
++ printk(KERN_ERR "WARNING: Bogus (zero) I/O APIC address"
++ " found in MADT table, skipping!\n");
++ return;
++ }
++
++ idx = nr_ioapics++;
++
++ mp_ioapics[idx].mpc_type = MP_IOAPIC;
++ mp_ioapics[idx].mpc_flags = MPC_APIC_USABLE;
++ mp_ioapics[idx].mpc_apicaddr = address;
++
++#ifndef CONFIG_XEN
++ set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address);
++#endif
++ mp_ioapics[idx].mpc_apicid = id;
++ mp_ioapics[idx].mpc_apicver = io_apic_get_version(idx);
++
++ /*
++ * Build basic IRQ lookup table to facilitate gsi->io_apic lookups
++ * and to prevent reprogramming of IOAPIC pins (PCI IRQs).
++ */
++ mp_ioapic_routing[idx].apic_id = mp_ioapics[idx].mpc_apicid;
++ mp_ioapic_routing[idx].gsi_start = gsi_base;
++ mp_ioapic_routing[idx].gsi_end = gsi_base +
++ io_apic_get_redir_entries(idx);
++
++ printk(KERN_INFO "IOAPIC[%d]: apic_id %d, version %d, address 0x%x, "
++ "GSI %d-%d\n", idx, mp_ioapics[idx].mpc_apicid,
++ mp_ioapics[idx].mpc_apicver, mp_ioapics[idx].mpc_apicaddr,
++ mp_ioapic_routing[idx].gsi_start,
++ mp_ioapic_routing[idx].gsi_end);
++
++ return;
++}
++
++
++void __init mp_override_legacy_irq (
++ u8 bus_irq,
++ u8 polarity,
++ u8 trigger,
++ u32 gsi)
++{
++ struct mpc_config_intsrc intsrc;
++ int ioapic = -1;
++ int pin = -1;
++
++ /*
++ * Convert 'gsi' to 'ioapic.pin'.
++ */
++ ioapic = mp_find_ioapic(gsi);
++ if (ioapic < 0)
++ return;
++ pin = gsi - mp_ioapic_routing[ioapic].gsi_start;
++
++ /*
++ * TBD: This check is for faulty timer entries, where the override
++ * erroneously sets the trigger to level, resulting in a HUGE
++ * increase of timer interrupts!
++ */
++ if ((bus_irq == 0) && (trigger == 3))
++ trigger = 1;
++
++ intsrc.mpc_type = MP_INTSRC;
++ intsrc.mpc_irqtype = mp_INT;
++ intsrc.mpc_irqflag = (trigger << 2) | polarity;
++ intsrc.mpc_srcbus = MP_ISA_BUS;
++ intsrc.mpc_srcbusirq = bus_irq; /* IRQ */
++ intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid; /* APIC ID */
++ intsrc.mpc_dstirq = pin; /* INTIN# */
++
++ Dprintk("Int: type %d, pol %d, trig %d, bus %d, irq %d, %d-%d\n",
++ intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3,
++ (intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus,
++ intsrc.mpc_srcbusirq, intsrc.mpc_dstapic, intsrc.mpc_dstirq);
++
++ mp_irqs[mp_irq_entries] = intsrc;
++ if (++mp_irq_entries == MAX_IRQ_SOURCES)
++ panic("Max # of irq sources exceeded!\n");
++
++ return;
++}
++
++
++void __init mp_config_acpi_legacy_irqs (void)
++{
++ struct mpc_config_intsrc intsrc;
++ int i = 0;
++ int ioapic = -1;
++
++ /*
++ * Fabricate the legacy ISA bus (bus #31).
++ */
++ mp_bus_id_to_type[MP_ISA_BUS] = MP_BUS_ISA;
++ Dprintk("Bus #%d is ISA\n", MP_ISA_BUS);
++
++ /*
++ * Locate the IOAPIC that manages the ISA IRQs (0-15).
++ */
++ ioapic = mp_find_ioapic(0);
++ if (ioapic < 0)
++ return;
++
++ intsrc.mpc_type = MP_INTSRC;
++ intsrc.mpc_irqflag = 0; /* Conforming */
++ intsrc.mpc_srcbus = MP_ISA_BUS;
++ intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid;
++
++ /*
++ * Use the default configuration for the IRQs 0-15. Unless
++ * overridden by (MADT) interrupt source override entries.
++ */
++ for (i = 0; i < 16; i++) {
++ int idx;
++
++ for (idx = 0; idx < mp_irq_entries; idx++) {
++ struct mpc_config_intsrc *irq = mp_irqs + idx;
++
++ /* Do we already have a mapping for this ISA IRQ? */
++ if (irq->mpc_srcbus == MP_ISA_BUS && irq->mpc_srcbusirq == i)
++ break;
++
++ /* Do we already have a mapping for this IOAPIC pin */
++ if ((irq->mpc_dstapic == intsrc.mpc_dstapic) &&
++ (irq->mpc_dstirq == i))
++ break;
++ }
++
++ if (idx != mp_irq_entries) {
++ printk(KERN_DEBUG "ACPI: IRQ%d used by override.\n", i);
++ continue; /* IRQ already used */
++ }
++
++ intsrc.mpc_irqtype = mp_INT;
++ intsrc.mpc_srcbusirq = i; /* Identity mapped */
++ intsrc.mpc_dstirq = i;
++
++ Dprintk("Int: type %d, pol %d, trig %d, bus %d, irq %d, "
++ "%d-%d\n", intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3,
++ (intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus,
++ intsrc.mpc_srcbusirq, intsrc.mpc_dstapic,
++ intsrc.mpc_dstirq);
++
++ mp_irqs[mp_irq_entries] = intsrc;
++ if (++mp_irq_entries == MAX_IRQ_SOURCES)
++ panic("Max # of irq sources exceeded!\n");
++ }
++
++ return;
++}
++
++#define MAX_GSI_NUM 4096
++
++int mp_register_gsi(u32 gsi, int triggering, int polarity)
++{
++ int ioapic = -1;
++ int ioapic_pin = 0;
++ int idx, bit = 0;
++ static int pci_irq = 16;
++ /*
++ * Mapping between Global System Interrupts, which
++ * represent all possible interrupts, to the IRQs
++ * assigned to actual devices.
++ */
++ static int gsi_to_irq[MAX_GSI_NUM];
++
++ if (acpi_irq_model != ACPI_IRQ_MODEL_IOAPIC)
++ return gsi;
++
++ /* Don't set up the ACPI SCI because it's already set up */
++ if (acpi_fadt.sci_int == gsi)
++ return gsi;
++
++ ioapic = mp_find_ioapic(gsi);
++ if (ioapic < 0) {
++ printk(KERN_WARNING "No IOAPIC for GSI %u\n", gsi);
++ return gsi;
++ }
++
++ ioapic_pin = gsi - mp_ioapic_routing[ioapic].gsi_start;
++
++ /*
++ * Avoid pin reprogramming. PRTs typically include entries
++ * with redundant pin->gsi mappings (but unique PCI devices);
++ * we only program the IOAPIC on the first.
++ */
++ bit = ioapic_pin % 32;
++ idx = (ioapic_pin < 32) ? 0 : (ioapic_pin / 32);
++ if (idx > 3) {
++ printk(KERN_ERR "Invalid reference to IOAPIC pin "
++ "%d-%d\n", mp_ioapic_routing[ioapic].apic_id,
++ ioapic_pin);
++ return gsi;
++ }
++ if ((1<<bit) & mp_ioapic_routing[ioapic].pin_programmed[idx]) {
++ Dprintk(KERN_DEBUG "Pin %d-%d already programmed\n",
++ mp_ioapic_routing[ioapic].apic_id, ioapic_pin);
++ return gsi_to_irq[gsi];
++ }
++
++ mp_ioapic_routing[ioapic].pin_programmed[idx] |= (1<<bit);
++
++ if (triggering == ACPI_LEVEL_SENSITIVE) {
++ /*
++ * For PCI devices assign IRQs in order, avoiding gaps
++ * due to unused I/O APIC pins.
++ */
++ int irq = gsi;
++ if (gsi < MAX_GSI_NUM) {
++ /*
++ * Retain the VIA chipset work-around (gsi > 15), but
++ * avoid a problem where the 8254 timer (IRQ0) is setup
++ * via an override (so it's not on pin 0 of the ioapic),
++ * and at the same time, the pin 0 interrupt is a PCI
++ * type. The gsi > 15 test could cause these two pins
++ * to be shared as IRQ0, and they are not shareable.
++ * So test for this condition, and if necessary, avoid
++ * the pin collision.
++ */
++ if (gsi > 15 || (gsi == 0 && !timer_uses_ioapic_pin_0))
++ gsi = pci_irq++;
++ /*
++ * Don't assign IRQ used by ACPI SCI
++ */
++ if (gsi == acpi_fadt.sci_int)
++ gsi = pci_irq++;
++ gsi_to_irq[irq] = gsi;
++ } else {
++ printk(KERN_ERR "GSI %u is too high\n", gsi);
++ return gsi;
++ }
++ }
++
++ io_apic_set_pci_routing(ioapic, ioapic_pin, gsi,
++ triggering == ACPI_EDGE_SENSITIVE ? 0 : 1,
++ polarity == ACPI_ACTIVE_HIGH ? 0 : 1);
++ return gsi;
++}
++
++#endif /*CONFIG_X86_IO_APIC*/
++#endif /*CONFIG_ACPI*/
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/pci-swiotlb-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/pci-swiotlb-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,55 @@
++/* Glue code to lib/swiotlb.c */
++
++#include <linux/pci.h>
++#include <linux/cache.h>
++#include <linux/module.h>
++#include <asm/dma-mapping.h>
++#include <asm/proto.h>
++#include <asm/swiotlb.h>
++#include <asm/dma.h>
++
++#if 0
++int swiotlb __read_mostly;
++EXPORT_SYMBOL(swiotlb);
++#endif
++
++struct dma_mapping_ops swiotlb_dma_ops = {
++#if 0
++ .mapping_error = swiotlb_dma_mapping_error,
++ .alloc_coherent = swiotlb_alloc_coherent,
++ .free_coherent = swiotlb_free_coherent,
++ .map_single = swiotlb_map_single,
++ .unmap_single = swiotlb_unmap_single,
++ .sync_single_for_cpu = swiotlb_sync_single_for_cpu,
++ .sync_single_for_device = swiotlb_sync_single_for_device,
++ .sync_single_range_for_cpu = swiotlb_sync_single_range_for_cpu,
++ .sync_single_range_for_device = swiotlb_sync_single_range_for_device,
++ .sync_sg_for_cpu = swiotlb_sync_sg_for_cpu,
++ .sync_sg_for_device = swiotlb_sync_sg_for_device,
++ .map_sg = swiotlb_map_sg,
++ .unmap_sg = swiotlb_unmap_sg,
++ .dma_supported = NULL,
++#endif
++};
++
++void pci_swiotlb_init(void)
++{
++#if 0
++ /* don't initialize swiotlb if iommu=off (no_iommu=1) */
++ if (!iommu_detected && !no_iommu && end_pfn > MAX_DMA32_PFN)
++ swiotlb = 1;
++ if (swiotlb_force)
++ swiotlb = 1;
++ if (swiotlb) {
++ printk(KERN_INFO "PCI-DMA: Using software bounce buffering for IO (SWIOTLB)\n");
++ swiotlb_init();
++ dma_ops = &swiotlb_dma_ops;
++ }
++#else
++ swiotlb_init();
++ if (swiotlb) {
++ printk(KERN_INFO "PCI-DMA: Using software bounce buffering for IO (SWIOTLB)\n");
++ dma_ops = &swiotlb_dma_ops;
++ }
++#endif
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/process-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/process-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,848 @@
++/*
++ * linux/arch/x86-64/kernel/process.c
++ *
++ * Copyright (C) 1995 Linus Torvalds
++ *
++ * Pentium III FXSR, SSE support
++ * Gareth Hughes <gareth at valinux.com>, May 2000
++ *
++ * X86-64 port
++ * Andi Kleen.
++ *
++ * CPU hotplug support - ashok.raj at intel.com
++ *
++ * Jun Nakajima <jun.nakajima at intel.com>
++ * Modified for Xen
++ */
++
++/*
++ * This file handles the architecture-dependent parts of process handling..
++ */
++
++#include <stdarg.h>
++
++#include <linux/cpu.h>
++#include <linux/errno.h>
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/mm.h>
++#include <linux/elfcore.h>
++#include <linux/smp.h>
++#include <linux/slab.h>
++#include <linux/user.h>
++#include <linux/module.h>
++#include <linux/a.out.h>
++#include <linux/interrupt.h>
++#include <linux/delay.h>
++#include <linux/ptrace.h>
++#include <linux/utsname.h>
++#include <linux/random.h>
++#include <linux/notifier.h>
++#include <linux/kprobes.h>
++
++#include <asm/uaccess.h>
++#include <asm/pgtable.h>
++#include <asm/system.h>
++#include <asm/io.h>
++#include <asm/processor.h>
++#include <asm/i387.h>
++#include <asm/mmu_context.h>
++#include <asm/pda.h>
++#include <asm/prctl.h>
++#include <asm/kdebug.h>
++#include <xen/interface/platform.h>
++#include <xen/interface/physdev.h>
++#include <xen/interface/vcpu.h>
++#include <asm/desc.h>
++#include <asm/proto.h>
++#include <asm/hardirq.h>
++#include <asm/ia32.h>
++#include <asm/idle.h>
++
++#include <xen/cpu_hotplug.h>
++
++asmlinkage extern void ret_from_fork(void);
++
++unsigned long kernel_thread_flags = CLONE_VM | CLONE_UNTRACED;
++
++unsigned long boot_option_idle_override = 0;
++EXPORT_SYMBOL(boot_option_idle_override);
++
++/*
++ * Powermanagement idle function, if any..
++ */
++void (*pm_idle)(void);
++EXPORT_SYMBOL(pm_idle);
++static DEFINE_PER_CPU(unsigned int, cpu_idle_state);
++
++static ATOMIC_NOTIFIER_HEAD(idle_notifier);
++
++void idle_notifier_register(struct notifier_block *n)
++{
++ atomic_notifier_chain_register(&idle_notifier, n);
++}
++EXPORT_SYMBOL_GPL(idle_notifier_register);
++
++void idle_notifier_unregister(struct notifier_block *n)
++{
++ atomic_notifier_chain_unregister(&idle_notifier, n);
++}
++EXPORT_SYMBOL(idle_notifier_unregister);
++
++enum idle_state { CPU_IDLE, CPU_NOT_IDLE };
++static DEFINE_PER_CPU(enum idle_state, idle_state) = CPU_NOT_IDLE;
++
++void enter_idle(void)
++{
++ __get_cpu_var(idle_state) = CPU_IDLE;
++ atomic_notifier_call_chain(&idle_notifier, IDLE_START, NULL);
++}
++
++static void __exit_idle(void)
++{
++ __get_cpu_var(idle_state) = CPU_NOT_IDLE;
++ atomic_notifier_call_chain(&idle_notifier, IDLE_END, NULL);
++}
++
++/* Called from interrupts to signify idle end */
++void exit_idle(void)
++{
++ if (current->pid | read_pda(irqcount))
++ return;
++ __exit_idle();
++}
++
++/*
++ * On SMP it's slightly faster (but much more power-consuming!)
++ * to poll the ->need_resched flag instead of waiting for the
++ * cross-CPU IPI to arrive. Use this option with caution.
++ */
++static void poll_idle (void)
++{
++ local_irq_enable();
++
++ asm volatile(
++ "2:"
++ "testl %0,%1;"
++ "rep; nop;"
++ "je 2b;"
++ : :
++ "i" (_TIF_NEED_RESCHED),
++ "m" (current_thread_info()->flags));
++}
++
++static void xen_idle(void)
++{
++ local_irq_disable();
++
++ if (need_resched())
++ local_irq_enable();
++ else {
++ current_thread_info()->status &= ~TS_POLLING;
++ smp_mb__after_clear_bit();
++ safe_halt();
++ current_thread_info()->status |= TS_POLLING;
++ }
++}
++
++#ifdef CONFIG_HOTPLUG_CPU
++static inline void play_dead(void)
++{
++ idle_task_exit();
++ local_irq_disable();
++ cpu_clear(smp_processor_id(), cpu_initialized);
++ preempt_enable_no_resched();
++ VOID(HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL));
++ cpu_bringup();
++}
++#else
++static inline void play_dead(void)
++{
++ BUG();
++}
++#endif /* CONFIG_HOTPLUG_CPU */
++
++/*
++ * The idle thread. There's no useful work to be
++ * done, so just try to conserve power and have a
++ * low exit latency (ie sit in a loop waiting for
++ * somebody to say that they'd like to reschedule)
++ */
++void cpu_idle (void)
++{
++ current_thread_info()->status |= TS_POLLING;
++ /* endless idle loop with no priority at all */
++ while (1) {
++ while (!need_resched()) {
++ void (*idle)(void);
++
++ if (__get_cpu_var(cpu_idle_state))
++ __get_cpu_var(cpu_idle_state) = 0;
++ rmb();
++ idle = xen_idle; /* no alternatives */
++ if (cpu_is_offline(smp_processor_id()))
++ play_dead();
++ enter_idle();
++ idle();
++ __exit_idle();
++ }
++
++ preempt_enable_no_resched();
++ schedule();
++ preempt_disable();
++ }
++}
++
++void cpu_idle_wait(void)
++{
++ unsigned int cpu, this_cpu = get_cpu();
++ cpumask_t map;
++
++ set_cpus_allowed(current, cpumask_of_cpu(this_cpu));
++ put_cpu();
++
++ cpus_clear(map);
++ for_each_online_cpu(cpu) {
++ per_cpu(cpu_idle_state, cpu) = 1;
++ cpu_set(cpu, map);
++ }
++
++ __get_cpu_var(cpu_idle_state) = 0;
++
++ wmb();
++ do {
++ ssleep(1);
++ for_each_online_cpu(cpu) {
++ if (cpu_isset(cpu, map) &&
++ !per_cpu(cpu_idle_state, cpu))
++ cpu_clear(cpu, map);
++ }
++ cpus_and(map, map, cpu_online_map);
++ } while (!cpus_empty(map));
++}
++EXPORT_SYMBOL_GPL(cpu_idle_wait);
++
++void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c)
++{
++}
++
++static int __init idle_setup (char *str)
++{
++ if (!strncmp(str, "poll", 4)) {
++ printk("using polling idle threads.\n");
++ pm_idle = poll_idle;
++ }
++
++ boot_option_idle_override = 1;
++ return 1;
++}
++
++__setup("idle=", idle_setup);
++
++/* Prints also some state that isn't saved in the pt_regs */
++void __show_regs(struct pt_regs * regs)
++{
++ unsigned long fs, gs, shadowgs;
++ unsigned int fsindex,gsindex;
++ unsigned int ds,cs,es;
++
++ printk("\n");
++ print_modules();
++ printk("Pid: %d, comm: %.20s %s %s %.*s\n",
++ current->pid, current->comm, print_tainted(),
++ system_utsname.release,
++ (int)strcspn(system_utsname.version, " "),
++ system_utsname.version);
++ printk("RIP: %04lx:[<%016lx>] ", regs->cs & 0xffff, regs->rip);
++ printk_address(regs->rip);
++ printk("RSP: %04lx:%016lx EFLAGS: %08lx\n", regs->ss, regs->rsp,
++ regs->eflags);
++ printk("RAX: %016lx RBX: %016lx RCX: %016lx\n",
++ regs->rax, regs->rbx, regs->rcx);
++ printk("RDX: %016lx RSI: %016lx RDI: %016lx\n",
++ regs->rdx, regs->rsi, regs->rdi);
++ printk("RBP: %016lx R08: %016lx R09: %016lx\n",
++ regs->rbp, regs->r8, regs->r9);
++ printk("R10: %016lx R11: %016lx R12: %016lx\n",
++ regs->r10, regs->r11, regs->r12);
++ printk("R13: %016lx R14: %016lx R15: %016lx\n",
++ regs->r13, regs->r14, regs->r15);
++
++ asm("mov %%ds,%0" : "=r" (ds));
++ asm("mov %%cs,%0" : "=r" (cs));
++ asm("mov %%es,%0" : "=r" (es));
++ asm("mov %%fs,%0" : "=r" (fsindex));
++ asm("mov %%gs,%0" : "=r" (gsindex));
++
++ rdmsrl(MSR_FS_BASE, fs);
++ rdmsrl(MSR_GS_BASE, gs);
++ rdmsrl(MSR_KERNEL_GS_BASE, shadowgs);
++
++ printk("FS: %016lx(%04x) GS:%016lx(%04x) knlGS:%016lx\n",
++ fs,fsindex,gs,gsindex,shadowgs);
++ printk("CS: %04x DS: %04x ES: %04x\n", cs, ds, es);
++
++}
++
++void show_regs(struct pt_regs *regs)
++{
++ printk("CPU %d:", smp_processor_id());
++ __show_regs(regs);
++ show_trace(NULL, regs, (void *)(regs + 1));
++}
++
++/*
++ * Free current thread data structures etc..
++ */
++void exit_thread(void)
++{
++ struct task_struct *me = current;
++ struct thread_struct *t = &me->thread;
++
++ if (me->thread.io_bitmap_ptr) {
++#ifndef CONFIG_X86_NO_TSS
++ struct tss_struct *tss = &per_cpu(init_tss, get_cpu());
++#endif
++#ifdef CONFIG_XEN
++ struct physdev_set_iobitmap iobmp_op;
++ memset(&iobmp_op, 0, sizeof(iobmp_op));
++#endif
++
++ kfree(t->io_bitmap_ptr);
++ t->io_bitmap_ptr = NULL;
++ /*
++ * Careful, clear this in the TSS too:
++ */
++#ifndef CONFIG_X86_NO_TSS
++ memset(tss->io_bitmap, 0xff, t->io_bitmap_max);
++ put_cpu();
++#endif
++#ifdef CONFIG_XEN
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap,
++ &iobmp_op));
++#endif
++ t->io_bitmap_max = 0;
++ }
++}
++
++void load_gs_index(unsigned gs)
++{
++ WARN_ON(HYPERVISOR_set_segment_base(SEGBASE_GS_USER_SEL, gs));
++}
++
++void flush_thread(void)
++{
++ struct task_struct *tsk = current;
++ struct thread_info *t = current_thread_info();
++
++ if (t->flags & _TIF_ABI_PENDING) {
++ t->flags ^= (_TIF_ABI_PENDING | _TIF_IA32);
++ if (t->flags & _TIF_IA32)
++ current_thread_info()->status |= TS_COMPAT;
++ }
++
++ tsk->thread.debugreg0 = 0;
++ tsk->thread.debugreg1 = 0;
++ tsk->thread.debugreg2 = 0;
++ tsk->thread.debugreg3 = 0;
++ tsk->thread.debugreg6 = 0;
++ tsk->thread.debugreg7 = 0;
++ memset(tsk->thread.tls_array, 0, sizeof(tsk->thread.tls_array));
++ /*
++ * Forget coprocessor state..
++ */
++ clear_fpu(tsk);
++ clear_used_math();
++}
++
++void release_thread(struct task_struct *dead_task)
++{
++ if (dead_task->mm) {
++ if (dead_task->mm->context.size) {
++ printk("WARNING: dead process %8s still has LDT? <%p/%d>\n",
++ dead_task->comm,
++ dead_task->mm->context.ldt,
++ dead_task->mm->context.size);
++ BUG();
++ }
++ }
++}
++
++static inline void set_32bit_tls(struct task_struct *t, int tls, u32 addr)
++{
++ struct user_desc ud = {
++ .base_addr = addr,
++ .limit = 0xfffff,
++ .seg_32bit = 1,
++ .limit_in_pages = 1,
++ .useable = 1,
++ };
++ struct n_desc_struct *desc = (void *)t->thread.tls_array;
++ desc += tls;
++ desc->a = LDT_entry_a(&ud);
++ desc->b = LDT_entry_b(&ud);
++}
++
++static inline u32 read_32bit_tls(struct task_struct *t, int tls)
++{
++ struct desc_struct *desc = (void *)t->thread.tls_array;
++ desc += tls;
++ return desc->base0 |
++ (((u32)desc->base1) << 16) |
++ (((u32)desc->base2) << 24);
++}
++
++/*
++ * This gets called before we allocate a new thread and copy
++ * the current task into it.
++ */
++void prepare_to_copy(struct task_struct *tsk)
++{
++ unlazy_fpu(tsk);
++}
++
++int copy_thread(int nr, unsigned long clone_flags, unsigned long rsp,
++ unsigned long unused,
++ struct task_struct * p, struct pt_regs * regs)
++{
++ int err;
++ struct pt_regs * childregs;
++ struct task_struct *me = current;
++
++ childregs = ((struct pt_regs *)
++ (THREAD_SIZE + task_stack_page(p))) - 1;
++ *childregs = *regs;
++
++ childregs->rax = 0;
++ childregs->rsp = rsp;
++ if (rsp == ~0UL)
++ childregs->rsp = (unsigned long)childregs;
++
++ p->thread.rsp = (unsigned long) childregs;
++ p->thread.rsp0 = (unsigned long) (childregs+1);
++ p->thread.userrsp = me->thread.userrsp;
++
++ set_tsk_thread_flag(p, TIF_FORK);
++
++ p->thread.fs = me->thread.fs;
++ p->thread.gs = me->thread.gs;
++
++ asm("mov %%gs,%0" : "=m" (p->thread.gsindex));
++ asm("mov %%fs,%0" : "=m" (p->thread.fsindex));
++ asm("mov %%es,%0" : "=m" (p->thread.es));
++ asm("mov %%ds,%0" : "=m" (p->thread.ds));
++
++ if (unlikely(me->thread.io_bitmap_ptr != NULL)) {
++ p->thread.io_bitmap_ptr = kmalloc(IO_BITMAP_BYTES, GFP_KERNEL);
++ if (!p->thread.io_bitmap_ptr) {
++ p->thread.io_bitmap_max = 0;
++ return -ENOMEM;
++ }
++ memcpy(p->thread.io_bitmap_ptr, me->thread.io_bitmap_ptr,
++ IO_BITMAP_BYTES);
++ }
++
++ /*
++ * Set a new TLS for the child thread?
++ */
++ if (clone_flags & CLONE_SETTLS) {
++#ifdef CONFIG_IA32_EMULATION
++ if (test_thread_flag(TIF_IA32))
++ err = ia32_child_tls(p, childregs);
++ else
++#endif
++ err = do_arch_prctl(p, ARCH_SET_FS, childregs->r8);
++ if (err)
++ goto out;
++ }
++ p->thread.iopl = current->thread.iopl;
++
++ err = 0;
++out:
++ if (err && p->thread.io_bitmap_ptr) {
++ kfree(p->thread.io_bitmap_ptr);
++ p->thread.io_bitmap_max = 0;
++ }
++ return err;
++}
++
++static inline void __save_init_fpu( struct task_struct *tsk )
++{
++ asm volatile( "rex64 ; fxsave %0 ; fnclex"
++ : "=m" (tsk->thread.i387.fxsave));
++ tsk->thread_info->status &= ~TS_USEDFPU;
++}
++
++/*
++ * switch_to(x,y) should switch tasks from x to y.
++ *
++ * This could still be optimized:
++ * - fold all the options into a flag word and test it with a single test.
++ * - could test fs/gs bitsliced
++ *
++ * Kprobes not supported here. Set the probe on schedule instead.
++ */
++__kprobes struct task_struct *
++__switch_to(struct task_struct *prev_p, struct task_struct *next_p)
++{
++ struct thread_struct *prev = &prev_p->thread,
++ *next = &next_p->thread;
++ int cpu = smp_processor_id();
++#ifndef CONFIG_X86_NO_TSS
++ struct tss_struct *tss = &per_cpu(init_tss, cpu);
++#endif
++#if CONFIG_XEN_COMPAT > 0x030002
++ struct physdev_set_iopl iopl_op;
++ struct physdev_set_iobitmap iobmp_op;
++#else
++ struct physdev_op _pdo[2], *pdo = _pdo;
++#define iopl_op pdo->u.set_iopl
++#define iobmp_op pdo->u.set_iobitmap
++#endif
++ multicall_entry_t _mcl[8], *mcl = _mcl;
++
++ /*
++ * This is basically '__unlazy_fpu', except that we queue a
++ * multicall to indicate FPU task switch, rather than
++ * synchronously trapping to Xen.
++ * The AMD workaround requires it to be after DS reload, or
++ * after DS has been cleared, which we do in __prepare_arch_switch.
++ */
++ if (prev_p->thread_info->status & TS_USEDFPU) {
++ __save_init_fpu(prev_p); /* _not_ save_init_fpu() */
++ mcl->op = __HYPERVISOR_fpu_taskswitch;
++ mcl->args[0] = 1;
++ mcl++;
++ }
++
++ /*
++ * Reload esp0, LDT and the page table pointer:
++ */
++ mcl->op = __HYPERVISOR_stack_switch;
++ mcl->args[0] = __KERNEL_DS;
++ mcl->args[1] = next->rsp0;
++ mcl++;
++
++ /*
++ * Load the per-thread Thread-Local Storage descriptor.
++ * This is load_TLS(next, cpu) with multicalls.
++ */
++#define C(i) do { \
++ if (unlikely(next->tls_array[i] != prev->tls_array[i])) { \
++ mcl->op = __HYPERVISOR_update_descriptor; \
++ mcl->args[0] = virt_to_machine( \
++ &cpu_gdt(cpu)[GDT_ENTRY_TLS_MIN + i]); \
++ mcl->args[1] = next->tls_array[i]; \
++ mcl++; \
++ } \
++} while (0)
++ C(0); C(1); C(2);
++#undef C
++
++ if (unlikely(prev->iopl != next->iopl)) {
++ iopl_op.iopl = (next->iopl == 0) ? 1 : next->iopl;
++#if CONFIG_XEN_COMPAT > 0x030002
++ mcl->op = __HYPERVISOR_physdev_op;
++ mcl->args[0] = PHYSDEVOP_set_iopl;
++ mcl->args[1] = (unsigned long)&iopl_op;
++#else
++ mcl->op = __HYPERVISOR_physdev_op_compat;
++ pdo->cmd = PHYSDEVOP_set_iopl;
++ mcl->args[0] = (unsigned long)pdo++;
++#endif
++ mcl++;
++ }
++
++ if (unlikely(prev->io_bitmap_ptr || next->io_bitmap_ptr)) {
++ set_xen_guest_handle(iobmp_op.bitmap,
++ (char *)next->io_bitmap_ptr);
++ iobmp_op.nr_ports = next->io_bitmap_ptr ? IO_BITMAP_BITS : 0;
++#if CONFIG_XEN_COMPAT > 0x030002
++ mcl->op = __HYPERVISOR_physdev_op;
++ mcl->args[0] = PHYSDEVOP_set_iobitmap;
++ mcl->args[1] = (unsigned long)&iobmp_op;
++#else
++ mcl->op = __HYPERVISOR_physdev_op_compat;
++ pdo->cmd = PHYSDEVOP_set_iobitmap;
++ mcl->args[0] = (unsigned long)pdo++;
++#endif
++ mcl++;
++ }
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ BUG_ON(pdo > _pdo + ARRAY_SIZE(_pdo));
++#endif
++ BUG_ON(mcl > _mcl + ARRAY_SIZE(_mcl));
++ if (unlikely(HYPERVISOR_multicall_check(_mcl, mcl - _mcl, NULL)))
++ BUG();
++
++ /*
++ * Switch DS and ES.
++ * This won't pick up thread selector changes, but I guess that is ok.
++ */
++ if (unlikely(next->es))
++ loadsegment(es, next->es);
++
++ if (unlikely(next->ds))
++ loadsegment(ds, next->ds);
++
++ /*
++ * Switch FS and GS.
++ */
++ if (unlikely(next->fsindex))
++ loadsegment(fs, next->fsindex);
++
++ if (next->fs)
++ WARN_ON(HYPERVISOR_set_segment_base(SEGBASE_FS, next->fs));
++
++ if (unlikely(next->gsindex))
++ load_gs_index(next->gsindex);
++
++ if (next->gs)
++ WARN_ON(HYPERVISOR_set_segment_base(SEGBASE_GS_USER, next->gs));
++
++ /*
++ * Switch the PDA context.
++ */
++ prev->userrsp = read_pda(oldrsp);
++ write_pda(oldrsp, next->userrsp);
++ write_pda(pcurrent, next_p);
++ write_pda(kernelstack,
++ task_stack_page(next_p) + THREAD_SIZE - PDA_STACKOFFSET);
++
++ /*
++ * Now maybe reload the debug registers
++ */
++ if (unlikely(next->debugreg7)) {
++ set_debugreg(next->debugreg0, 0);
++ set_debugreg(next->debugreg1, 1);
++ set_debugreg(next->debugreg2, 2);
++ set_debugreg(next->debugreg3, 3);
++ /* no 4 and 5 */
++ set_debugreg(next->debugreg6, 6);
++ set_debugreg(next->debugreg7, 7);
++ }
++
++ return prev_p;
++}
++
++/*
++ * sys_execve() executes a new program.
++ */
++asmlinkage
++long sys_execve(char __user *name, char __user * __user *argv,
++ char __user * __user *envp, struct pt_regs regs)
++{
++ long error;
++ char * filename;
++
++ filename = getname(name);
++ error = PTR_ERR(filename);
++ if (IS_ERR(filename))
++ return error;
++ error = do_execve(filename, argv, envp, ®s);
++ if (error == 0) {
++ task_lock(current);
++ current->ptrace &= ~PT_DTRACE;
++ task_unlock(current);
++ }
++ putname(filename);
++ return error;
++}
++
++void set_personality_64bit(void)
++{
++ /* inherit personality from parent */
++
++ /* Make sure to be in 64bit mode */
++ clear_thread_flag(TIF_IA32);
++
++ /* TBD: overwrites user setup. Should have two bits.
++ But 64bit processes have always behaved this way,
++ so it's not too bad. The main problem is just that
++ 32bit childs are affected again. */
++ current->personality &= ~READ_IMPLIES_EXEC;
++}
++
++asmlinkage long sys_fork(struct pt_regs *regs)
++{
++ return do_fork(SIGCHLD, regs->rsp, regs, 0, NULL, NULL);
++}
++
++asmlinkage long
++sys_clone(unsigned long clone_flags, unsigned long newsp,
++ void __user *parent_tid, void __user *child_tid, struct pt_regs *regs)
++{
++ if (!newsp)
++ newsp = regs->rsp;
++ return do_fork(clone_flags, newsp, regs, 0, parent_tid, child_tid);
++}
++
++/*
++ * This is trivial, and on the face of it looks like it
++ * could equally well be done in user mode.
++ *
++ * Not so, for quite unobvious reasons - register pressure.
++ * In user mode vfork() cannot have a stack frame, and if
++ * done by calling the "clone()" system call directly, you
++ * do not have enough call-clobbered registers to hold all
++ * the information you need.
++ */
++asmlinkage long sys_vfork(struct pt_regs *regs)
++{
++ return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, regs->rsp, regs, 0,
++ NULL, NULL);
++}
++
++unsigned long get_wchan(struct task_struct *p)
++{
++ unsigned long stack;
++ u64 fp,rip;
++ int count = 0;
++
++ if (!p || p == current || p->state==TASK_RUNNING)
++ return 0;
++ stack = (unsigned long)task_stack_page(p);
++ if (p->thread.rsp < stack || p->thread.rsp > stack+THREAD_SIZE)
++ return 0;
++ fp = *(u64 *)(p->thread.rsp);
++ do {
++ if (fp < (unsigned long)stack ||
++ fp > (unsigned long)stack+THREAD_SIZE)
++ return 0;
++ rip = *(u64 *)(fp+8);
++ if (!in_sched_functions(rip))
++ return rip;
++ fp = *(u64 *)fp;
++ } while (count++ < 16);
++ return 0;
++}
++
++long do_arch_prctl(struct task_struct *task, int code, unsigned long addr)
++{
++ int ret = 0;
++ int doit = task == current;
++ int cpu;
++
++ switch (code) {
++ case ARCH_SET_GS:
++ if (addr >= TASK_SIZE_OF(task))
++ return -EPERM;
++ cpu = get_cpu();
++ /* handle small bases via the GDT because that's faster to
++ switch. */
++ if (addr <= 0xffffffff) {
++ set_32bit_tls(task, GS_TLS, addr);
++ if (doit) {
++ load_TLS(&task->thread, cpu);
++ load_gs_index(GS_TLS_SEL);
++ }
++ task->thread.gsindex = GS_TLS_SEL;
++ task->thread.gs = 0;
++ } else {
++ task->thread.gsindex = 0;
++ task->thread.gs = addr;
++ if (doit) {
++ load_gs_index(0);
++ ret = HYPERVISOR_set_segment_base(
++ SEGBASE_GS_USER, addr);
++ }
++ }
++ put_cpu();
++ break;
++ case ARCH_SET_FS:
++ /* Not strictly needed for fs, but do it for symmetry
++ with gs */
++ if (addr >= TASK_SIZE_OF(task))
++ return -EPERM;
++ cpu = get_cpu();
++ /* handle small bases via the GDT because that's faster to
++ switch. */
++ if (addr <= 0xffffffff) {
++ set_32bit_tls(task, FS_TLS, addr);
++ if (doit) {
++ load_TLS(&task->thread, cpu);
++ asm volatile("movl %0,%%fs" :: "r"(FS_TLS_SEL));
++ }
++ task->thread.fsindex = FS_TLS_SEL;
++ task->thread.fs = 0;
++ } else {
++ task->thread.fsindex = 0;
++ task->thread.fs = addr;
++ if (doit) {
++ /* set the selector to 0 to not confuse
++ __switch_to */
++ asm volatile("movl %0,%%fs" :: "r" (0));
++ ret = HYPERVISOR_set_segment_base(SEGBASE_FS,
++ addr);
++ }
++ }
++ put_cpu();
++ break;
++ case ARCH_GET_FS: {
++ unsigned long base;
++ if (task->thread.fsindex == FS_TLS_SEL)
++ base = read_32bit_tls(task, FS_TLS);
++ else if (doit)
++ rdmsrl(MSR_FS_BASE, base);
++ else
++ base = task->thread.fs;
++ ret = put_user(base, (unsigned long __user *)addr);
++ break;
++ }
++ case ARCH_GET_GS: {
++ unsigned long base;
++ unsigned gsindex;
++ if (task->thread.gsindex == GS_TLS_SEL)
++ base = read_32bit_tls(task, GS_TLS);
++ else if (doit) {
++ asm("movl %%gs,%0" : "=r" (gsindex));
++ if (gsindex)
++ rdmsrl(MSR_KERNEL_GS_BASE, base);
++ else
++ base = task->thread.gs;
++ }
++ else
++ base = task->thread.gs;
++ ret = put_user(base, (unsigned long __user *)addr);
++ break;
++ }
++
++ default:
++ ret = -EINVAL;
++ break;
++ }
++
++ return ret;
++}
++
++long sys_arch_prctl(int code, unsigned long addr)
++{
++ return do_arch_prctl(current, code, addr);
++}
++
++/*
++ * Capture the user space registers if the task is not running (in user space)
++ */
++int dump_task_regs(struct task_struct *tsk, elf_gregset_t *regs)
++{
++ struct pt_regs *pp, ptregs;
++
++ pp = task_pt_regs(tsk);
++
++ ptregs = *pp;
++ ptregs.cs &= 0xffff;
++ ptregs.ss &= 0xffff;
++
++ elf_core_copy_regs(regs, &ptregs);
++
++ boot_option_idle_override = 1;
++ return 1;
++}
++
++unsigned long arch_align_stack(unsigned long sp)
++{
++ if (randomize_va_space)
++ sp -= get_random_int() % 8192;
++ return sp & ~0xf;
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/relocate_kernel.S
+--- a/arch/x86_64/kernel/relocate_kernel.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/kernel/relocate_kernel.S Wed Sep 10 10:54:08 2008 +0100
+@@ -7,31 +7,195 @@
+ */
+
+ #include <linux/linkage.h>
++#include <asm/page.h>
++#include <asm/kexec.h>
+
+- /*
+- * Must be relocatable PIC code callable as a C function, that once
+- * it starts can not use the previous processes stack.
++/*
++ * Must be relocatable PIC code callable as a C function
++ */
++
++#define PTR(x) (x << 3)
++#define PAGE_ALIGNED (1 << PAGE_SHIFT)
++#define PAGE_ATTR 0x63 /* _PAGE_PRESENT|_PAGE_RW|_PAGE_ACCESSED|_PAGE_DIRTY */
++
++ .text
++ .align PAGE_ALIGNED
++ .code64
++ .globl relocate_kernel
++relocate_kernel:
++ /* %rdi indirection_page
++ * %rsi page_list
++ * %rdx start address
+ */
+- .globl relocate_new_kernel
+- .code64
++
++ /* map the control page at its virtual address */
++
++ movq $0x0000ff8000000000, %r10 /* mask */
++ mov $(39 - 3), %cl /* bits to shift */
++ movq PTR(VA_CONTROL_PAGE)(%rsi), %r11 /* address to map */
++
++ movq %r11, %r9
++ andq %r10, %r9
++ shrq %cl, %r9
++
++ movq PTR(VA_PGD)(%rsi), %r8
++ addq %r8, %r9
++ movq PTR(PA_PUD_0)(%rsi), %r8
++ orq $PAGE_ATTR, %r8
++ movq %r8, (%r9)
++
++ shrq $9, %r10
++ sub $9, %cl
++
++ movq %r11, %r9
++ andq %r10, %r9
++ shrq %cl, %r9
++
++ movq PTR(VA_PUD_0)(%rsi), %r8
++ addq %r8, %r9
++ movq PTR(PA_PMD_0)(%rsi), %r8
++ orq $PAGE_ATTR, %r8
++ movq %r8, (%r9)
++
++ shrq $9, %r10
++ sub $9, %cl
++
++ movq %r11, %r9
++ andq %r10, %r9
++ shrq %cl, %r9
++
++ movq PTR(VA_PMD_0)(%rsi), %r8
++ addq %r8, %r9
++ movq PTR(PA_PTE_0)(%rsi), %r8
++ orq $PAGE_ATTR, %r8
++ movq %r8, (%r9)
++
++ shrq $9, %r10
++ sub $9, %cl
++
++ movq %r11, %r9
++ andq %r10, %r9
++ shrq %cl, %r9
++
++ movq PTR(VA_PTE_0)(%rsi), %r8
++ addq %r8, %r9
++ movq PTR(PA_CONTROL_PAGE)(%rsi), %r8
++ orq $PAGE_ATTR, %r8
++ movq %r8, (%r9)
++
++ /* identity map the control page at its physical address */
++
++ movq $0x0000ff8000000000, %r10 /* mask */
++ mov $(39 - 3), %cl /* bits to shift */
++ movq PTR(PA_CONTROL_PAGE)(%rsi), %r11 /* address to map */
++
++ movq %r11, %r9
++ andq %r10, %r9
++ shrq %cl, %r9
++
++ movq PTR(VA_PGD)(%rsi), %r8
++ addq %r8, %r9
++ movq PTR(PA_PUD_1)(%rsi), %r8
++ orq $PAGE_ATTR, %r8
++ movq %r8, (%r9)
++
++ shrq $9, %r10
++ sub $9, %cl
++
++ movq %r11, %r9
++ andq %r10, %r9
++ shrq %cl, %r9
++
++ movq PTR(VA_PUD_1)(%rsi), %r8
++ addq %r8, %r9
++ movq PTR(PA_PMD_1)(%rsi), %r8
++ orq $PAGE_ATTR, %r8
++ movq %r8, (%r9)
++
++ shrq $9, %r10
++ sub $9, %cl
++
++ movq %r11, %r9
++ andq %r10, %r9
++ shrq %cl, %r9
++
++ movq PTR(VA_PMD_1)(%rsi), %r8
++ addq %r8, %r9
++ movq PTR(PA_PTE_1)(%rsi), %r8
++ orq $PAGE_ATTR, %r8
++ movq %r8, (%r9)
++
++ shrq $9, %r10
++ sub $9, %cl
++
++ movq %r11, %r9
++ andq %r10, %r9
++ shrq %cl, %r9
++
++ movq PTR(VA_PTE_1)(%rsi), %r8
++ addq %r8, %r9
++ movq PTR(PA_CONTROL_PAGE)(%rsi), %r8
++ orq $PAGE_ATTR, %r8
++ movq %r8, (%r9)
++
+ relocate_new_kernel:
+- /* %rdi page_list
+- * %rsi reboot_code_buffer
++ /* %rdi indirection_page
++ * %rsi page_list
+ * %rdx start address
+- * %rcx page_table
+- * %r8 arg5
+- * %r9 arg6
+ */
+
+ /* zero out flags, and disable interrupts */
+ pushq $0
+ popfq
+
+- /* set a new stack at the bottom of our page... */
+- lea 4096(%rsi), %rsp
++ /* get physical address of control page now */
++ /* this is impossible after page table switch */
++ movq PTR(PA_CONTROL_PAGE)(%rsi), %r8
+
+- /* store the parameters back on the stack */
+- pushq %rdx /* store the start address */
++ /* get physical address of page table now too */
++ movq PTR(PA_TABLE_PAGE)(%rsi), %rcx
++
++ /* switch to new set of page tables */
++ movq PTR(PA_PGD)(%rsi), %r9
++ movq %r9, %cr3
++
++ /* setup idt */
++ movq %r8, %rax
++ addq $(idt_80 - relocate_kernel), %rax
++ lidtq (%rax)
++
++ /* setup gdt */
++ movq %r8, %rax
++ addq $(gdt - relocate_kernel), %rax
++ movq %r8, %r9
++ addq $((gdt_80 - relocate_kernel) + 2), %r9
++ movq %rax, (%r9)
++
++ movq %r8, %rax
++ addq $(gdt_80 - relocate_kernel), %rax
++ lgdtq (%rax)
++
++ /* setup data segment registers */
++ xorl %eax, %eax
++ movl %eax, %ds
++ movl %eax, %es
++ movl %eax, %fs
++ movl %eax, %gs
++ movl %eax, %ss
++
++ /* setup a new stack at the end of the physical control page */
++ lea 4096(%r8), %rsp
++
++ /* load new code segment and jump to identity mapped page */
++ movq %r8, %rax
++ addq $(identity_mapped - relocate_kernel), %rax
++ pushq $(gdt_cs - gdt)
++ pushq %rax
++ lretq
++
++identity_mapped:
++ /* store the start address on the stack */
++ pushq %rdx
+
+ /* Set cr0 to a known state:
+ * 31 1 == Paging enabled
+@@ -134,10 +298,19 @@
+ xorq %r13, %r13
+ xorq %r14, %r14
+ xorq %r15, %r15
++ ret
+
+- ret
+-relocate_new_kernel_end:
++ .align 16
++gdt:
++ .quad 0x0000000000000000 /* NULL descriptor */
++gdt_cs:
++ .quad 0x00af9a000000ffff
++gdt_end:
+
+- .globl relocate_new_kernel_size
+-relocate_new_kernel_size:
+- .quad relocate_new_kernel_end - relocate_new_kernel
++gdt_80:
++ .word gdt_end - gdt - 1 /* limit */
++ .quad 0 /* base - filled in by code above */
++
++idt_80:
++ .word 0 /* limit */
++ .quad 0 /* base */
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/setup-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/setup-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1654 @@
++/*
++ * linux/arch/x86-64/kernel/setup.c
++ *
++ * Copyright (C) 1995 Linus Torvalds
++ *
++ * Nov 2001 Dave Jones <davej at suse.de>
++ * Forked from i386 setup code.
++ */
++
++/*
++ * This file handles the architecture-dependent parts of initialization
++ */
++
++#include <linux/errno.h>
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/mm.h>
++#include <linux/stddef.h>
++#include <linux/unistd.h>
++#include <linux/ptrace.h>
++#include <linux/slab.h>
++#include <linux/user.h>
++#include <linux/a.out.h>
++#include <linux/screen_info.h>
++#include <linux/ioport.h>
++#include <linux/delay.h>
++#include <linux/init.h>
++#include <linux/initrd.h>
++#include <linux/highmem.h>
++#include <linux/bootmem.h>
++#include <linux/module.h>
++#include <asm/processor.h>
++#include <linux/console.h>
++#include <linux/seq_file.h>
++#include <linux/crash_dump.h>
++#include <linux/root_dev.h>
++#include <linux/pci.h>
++#include <linux/acpi.h>
++#include <linux/kallsyms.h>
++#include <linux/edd.h>
++#include <linux/mmzone.h>
++#include <linux/kexec.h>
++#include <linux/cpufreq.h>
++#include <linux/dmi.h>
++#include <linux/dma-mapping.h>
++#include <linux/ctype.h>
++
++#include <asm/mtrr.h>
++#include <asm/uaccess.h>
++#include <asm/system.h>
++#include <asm/io.h>
++#include <asm/smp.h>
++#include <asm/msr.h>
++#include <asm/desc.h>
++#include <video/edid.h>
++#include <asm/e820.h>
++#include <asm/dma.h>
++#include <asm/mpspec.h>
++#include <asm/mmu_context.h>
++#include <asm/bootsetup.h>
++#include <asm/proto.h>
++#include <asm/setup.h>
++#include <asm/mach_apic.h>
++#include <asm/numa.h>
++#include <asm/sections.h>
++#include <asm/dmi.h>
++#ifdef CONFIG_XEN
++#include <linux/percpu.h>
++#include <xen/interface/physdev.h>
++#include "setup_arch_pre.h"
++#include <asm/hypervisor.h>
++#include <xen/interface/nmi.h>
++#include <xen/features.h>
++#include <xen/firmware.h>
++#include <xen/xencons.h>
++#define PFN_UP(x) (((x) + PAGE_SIZE-1) >> PAGE_SHIFT)
++#define PFN_PHYS(x) ((x) << PAGE_SHIFT)
++#include <asm/mach-xen/setup_arch_post.h>
++#include <xen/interface/memory.h>
++
++#ifdef CONFIG_XEN
++#include <xen/interface/kexec.h>
++#endif
++
++extern unsigned long start_pfn;
++extern struct edid_info edid_info;
++
++shared_info_t *HYPERVISOR_shared_info = (shared_info_t *)empty_zero_page;
++EXPORT_SYMBOL(HYPERVISOR_shared_info);
++
++extern char hypercall_page[PAGE_SIZE];
++EXPORT_SYMBOL(hypercall_page);
++
++static int xen_panic_event(struct notifier_block *, unsigned long, void *);
++static struct notifier_block xen_panic_block = {
++ xen_panic_event, NULL, 0 /* try to go last */
++};
++
++unsigned long *phys_to_machine_mapping;
++unsigned long *pfn_to_mfn_frame_list_list, *pfn_to_mfn_frame_list[512];
++
++EXPORT_SYMBOL(phys_to_machine_mapping);
++
++DEFINE_PER_CPU(multicall_entry_t, multicall_list[8]);
++DEFINE_PER_CPU(int, nr_multicall_ents);
++
++/* Raw start-of-day parameters from the hypervisor. */
++start_info_t *xen_start_info;
++EXPORT_SYMBOL(xen_start_info);
++#endif
++
++/*
++ * Machine setup..
++ */
++
++struct cpuinfo_x86 boot_cpu_data __read_mostly;
++EXPORT_SYMBOL(boot_cpu_data);
++
++unsigned long mmu_cr4_features;
++
++int acpi_disabled;
++EXPORT_SYMBOL(acpi_disabled);
++#ifdef CONFIG_ACPI
++extern int __initdata acpi_ht;
++extern acpi_interrupt_flags acpi_sci_flags;
++int __initdata acpi_force = 0;
++#endif
++
++int acpi_numa __initdata;
++
++/* Boot loader ID as an integer, for the benefit of proc_dointvec */
++int bootloader_type;
++
++unsigned long saved_video_mode;
++
++/*
++ * Early DMI memory
++ */
++int dmi_alloc_index;
++char dmi_alloc_data[DMI_MAX_DATA];
++
++/*
++ * Setup options
++ */
++struct screen_info screen_info;
++EXPORT_SYMBOL(screen_info);
++struct sys_desc_table_struct {
++ unsigned short length;
++ unsigned char table[0];
++};
++
++struct edid_info edid_info;
++EXPORT_SYMBOL_GPL(edid_info);
++struct e820map e820;
++#ifdef CONFIG_XEN
++struct e820map machine_e820;
++#endif
++
++extern int root_mountflags;
++
++char command_line[COMMAND_LINE_SIZE];
++
++struct resource standard_io_resources[] = {
++ { .name = "dma1", .start = 0x00, .end = 0x1f,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "pic1", .start = 0x20, .end = 0x21,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "timer0", .start = 0x40, .end = 0x43,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "timer1", .start = 0x50, .end = 0x53,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "keyboard", .start = 0x60, .end = 0x60,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "keyboard", .start = 0x64, .end = 0x64,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "dma page reg", .start = 0x80, .end = 0x8f,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "pic2", .start = 0xa0, .end = 0xa1,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "dma2", .start = 0xc0, .end = 0xdf,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "fpu", .start = 0xf0, .end = 0xff,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO }
++};
++
++#define STANDARD_IO_RESOURCES \
++ (sizeof standard_io_resources / sizeof standard_io_resources[0])
++
++#define IORESOURCE_RAM (IORESOURCE_BUSY | IORESOURCE_MEM)
++
++struct resource data_resource = {
++ .name = "Kernel data",
++ .start = 0,
++ .end = 0,
++ .flags = IORESOURCE_RAM,
++};
++struct resource code_resource = {
++ .name = "Kernel code",
++ .start = 0,
++ .end = 0,
++ .flags = IORESOURCE_RAM,
++};
++
++#define IORESOURCE_ROM (IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM)
++
++static struct resource system_rom_resource = {
++ .name = "System ROM",
++ .start = 0xf0000,
++ .end = 0xfffff,
++ .flags = IORESOURCE_ROM,
++};
++
++static struct resource extension_rom_resource = {
++ .name = "Extension ROM",
++ .start = 0xe0000,
++ .end = 0xeffff,
++ .flags = IORESOURCE_ROM,
++};
++
++static struct resource adapter_rom_resources[] = {
++ { .name = "Adapter ROM", .start = 0xc8000, .end = 0,
++ .flags = IORESOURCE_ROM },
++ { .name = "Adapter ROM", .start = 0, .end = 0,
++ .flags = IORESOURCE_ROM },
++ { .name = "Adapter ROM", .start = 0, .end = 0,
++ .flags = IORESOURCE_ROM },
++ { .name = "Adapter ROM", .start = 0, .end = 0,
++ .flags = IORESOURCE_ROM },
++ { .name = "Adapter ROM", .start = 0, .end = 0,
++ .flags = IORESOURCE_ROM },
++ { .name = "Adapter ROM", .start = 0, .end = 0,
++ .flags = IORESOURCE_ROM }
++};
++
++#define ADAPTER_ROM_RESOURCES \
++ (sizeof adapter_rom_resources / sizeof adapter_rom_resources[0])
++
++static struct resource video_rom_resource = {
++ .name = "Video ROM",
++ .start = 0xc0000,
++ .end = 0xc7fff,
++ .flags = IORESOURCE_ROM,
++};
++
++static struct resource video_ram_resource = {
++ .name = "Video RAM area",
++ .start = 0xa0000,
++ .end = 0xbffff,
++ .flags = IORESOURCE_RAM,
++};
++
++#define romsignature(x) (*(unsigned short *)(x) == 0xaa55)
++
++static int __init romchecksum(unsigned char *rom, unsigned long length)
++{
++ unsigned char *p, sum = 0;
++
++ for (p = rom; p < rom + length; p++)
++ sum += *p;
++ return sum == 0;
++}
++
++static void __init probe_roms(void)
++{
++ unsigned long start, length, upper;
++ unsigned char *rom;
++ int i;
++
++#ifdef CONFIG_XEN
++ /* Nothing to do if not running in dom0. */
++ if (!is_initial_xendomain())
++ return;
++#endif
++
++ /* video rom */
++ upper = adapter_rom_resources[0].start;
++ for (start = video_rom_resource.start; start < upper; start += 2048) {
++ rom = isa_bus_to_virt(start);
++ if (!romsignature(rom))
++ continue;
++
++ video_rom_resource.start = start;
++
++ /* 0 < length <= 0x7f * 512, historically */
++ length = rom[2] * 512;
++
++ /* if checksum okay, trust length byte */
++ if (length && romchecksum(rom, length))
++ video_rom_resource.end = start + length - 1;
++
++ request_resource(&iomem_resource, &video_rom_resource);
++ break;
++ }
++
++ start = (video_rom_resource.end + 1 + 2047) & ~2047UL;
++ if (start < upper)
++ start = upper;
++
++ /* system rom */
++ request_resource(&iomem_resource, &system_rom_resource);
++ upper = system_rom_resource.start;
++
++ /* check for extension rom (ignore length byte!) */
++ rom = isa_bus_to_virt(extension_rom_resource.start);
++ if (romsignature(rom)) {
++ length = extension_rom_resource.end - extension_rom_resource.start + 1;
++ if (romchecksum(rom, length)) {
++ request_resource(&iomem_resource, &extension_rom_resource);
++ upper = extension_rom_resource.start;
++ }
++ }
++
++ /* check for adapter roms on 2k boundaries */
++ for (i = 0; i < ADAPTER_ROM_RESOURCES && start < upper; start += 2048) {
++ rom = isa_bus_to_virt(start);
++ if (!romsignature(rom))
++ continue;
++
++ /* 0 < length <= 0x7f * 512, historically */
++ length = rom[2] * 512;
++
++ /* but accept any length that fits if checksum okay */
++ if (!length || start + length > upper || !romchecksum(rom, length))
++ continue;
++
++ adapter_rom_resources[i].start = start;
++ adapter_rom_resources[i].end = start + length - 1;
++ request_resource(&iomem_resource, &adapter_rom_resources[i]);
++
++ start = adapter_rom_resources[i++].end & ~2047UL;
++ }
++}
++
++/* Check for full argument with no trailing characters */
++static int fullarg(char *p, char *arg)
++{
++ int l = strlen(arg);
++ return !memcmp(p, arg, l) && (p[l] == 0 || isspace(p[l]));
++}
++
++static __init void parse_cmdline_early (char ** cmdline_p)
++{
++ char c = ' ', *to = command_line, *from = COMMAND_LINE;
++ int len = 0;
++ int userdef = 0;
++
++ for (;;) {
++ if (c != ' ')
++ goto next_char;
++
++#ifdef CONFIG_SMP
++ /*
++ * If the BIOS enumerates physical processors before logical,
++ * maxcpus=N at enumeration-time can be used to disable HT.
++ */
++ else if (!memcmp(from, "maxcpus=", 8)) {
++ extern unsigned int maxcpus;
++
++ maxcpus = simple_strtoul(from + 8, NULL, 0);
++ }
++#endif
++#ifdef CONFIG_ACPI
++ /* "acpi=off" disables both ACPI table parsing and interpreter init */
++ if (fullarg(from,"acpi=off"))
++ disable_acpi();
++
++ if (fullarg(from, "acpi=force")) {
++ /* add later when we do DMI horrors: */
++ acpi_force = 1;
++ acpi_disabled = 0;
++ }
++
++ /* acpi=ht just means: do ACPI MADT parsing
++ at bootup, but don't enable the full ACPI interpreter */
++ if (fullarg(from, "acpi=ht")) {
++ if (!acpi_force)
++ disable_acpi();
++ acpi_ht = 1;
++ }
++ else if (fullarg(from, "pci=noacpi"))
++ acpi_disable_pci();
++ else if (fullarg(from, "acpi=noirq"))
++ acpi_noirq_set();
++
++ else if (fullarg(from, "acpi_sci=edge"))
++ acpi_sci_flags.trigger = 1;
++ else if (fullarg(from, "acpi_sci=level"))
++ acpi_sci_flags.trigger = 3;
++ else if (fullarg(from, "acpi_sci=high"))
++ acpi_sci_flags.polarity = 1;
++ else if (fullarg(from, "acpi_sci=low"))
++ acpi_sci_flags.polarity = 3;
++
++ /* acpi=strict disables out-of-spec workarounds */
++ else if (fullarg(from, "acpi=strict")) {
++ acpi_strict = 1;
++ }
++#ifdef CONFIG_X86_IO_APIC
++ else if (fullarg(from, "acpi_skip_timer_override"))
++ acpi_skip_timer_override = 1;
++#endif
++#endif
++
++#ifndef CONFIG_XEN
++ if (fullarg(from, "nolapic") || fullarg(from, "disableapic")) {
++ clear_bit(X86_FEATURE_APIC, boot_cpu_data.x86_capability);
++ disable_apic = 1;
++ }
++
++ if (fullarg(from, "noapic"))
++ skip_ioapic_setup = 1;
++
++ if (fullarg(from,"apic")) {
++ skip_ioapic_setup = 0;
++ ioapic_force = 1;
++ }
++#endif
++
++ if (!memcmp(from, "mem=", 4))
++ parse_memopt(from+4, &from);
++
++ if (!memcmp(from, "memmap=", 7)) {
++ /* exactmap option is for used defined memory */
++ if (!memcmp(from+7, "exactmap", 8)) {
++#ifdef CONFIG_CRASH_DUMP
++ /* If we are doing a crash dump, we
++ * still need to know the real mem
++ * size before original memory map is
++ * reset.
++ */
++ saved_max_pfn = e820_end_of_ram();
++#endif
++ from += 8+7;
++ end_pfn_map = 0;
++ e820.nr_map = 0;
++ userdef = 1;
++ }
++ else {
++ parse_memmapopt(from+7, &from);
++ userdef = 1;
++ }
++ }
++
++#ifdef CONFIG_NUMA
++ if (!memcmp(from, "numa=", 5))
++ numa_setup(from+5);
++#endif
++
++ if (!memcmp(from,"iommu=",6)) {
++ iommu_setup(from+6);
++ }
++
++ if (fullarg(from,"oops=panic"))
++ panic_on_oops = 1;
++
++ if (!memcmp(from, "noexec=", 7))
++ nonx_setup(from + 7);
++
++#ifdef CONFIG_KEXEC
++ /* crashkernel=size at addr specifies the location to reserve for
++ * a crash kernel. By reserving this memory we guarantee
++ * that linux never set's it up as a DMA target.
++ * Useful for holding code to do something appropriate
++ * after a kernel panic.
++ */
++ else if (!memcmp(from, "crashkernel=", 12)) {
++#ifndef CONFIG_XEN
++ unsigned long size, base;
++ size = memparse(from+12, &from);
++ if (*from == '@') {
++ base = memparse(from+1, &from);
++ /* FIXME: Do I want a sanity check
++ * to validate the memory range?
++ */
++ crashk_res.start = base;
++ crashk_res.end = base + size - 1;
++ }
++#else
++ printk("Ignoring crashkernel command line, "
++ "parameter will be supplied by xen\n");
++#endif
++ }
++#endif
++
++#ifdef CONFIG_PROC_VMCORE
++ /* elfcorehdr= specifies the location of elf core header
++ * stored by the crashed kernel. This option will be passed
++ * by kexec loader to the capture kernel.
++ */
++ else if(!memcmp(from, "elfcorehdr=", 11))
++ elfcorehdr_addr = memparse(from+11, &from);
++#endif
++
++#if defined(CONFIG_HOTPLUG_CPU) && !defined(CONFIG_XEN)
++ else if (!memcmp(from, "additional_cpus=", 16))
++ setup_additional_cpus(from+16);
++#endif
++
++ next_char:
++ c = *(from++);
++ if (!c)
++ break;
++ if (COMMAND_LINE_SIZE <= ++len)
++ break;
++ *(to++) = c;
++ }
++ if (userdef) {
++ printk(KERN_INFO "user-defined physical RAM map:\n");
++ e820_print_map("user");
++ }
++ *to = '\0';
++ *cmdline_p = command_line;
++}
++
++#ifndef CONFIG_NUMA
++static void __init
++contig_initmem_init(unsigned long start_pfn, unsigned long end_pfn)
++{
++ unsigned long bootmap_size, bootmap;
++
++ bootmap_size = bootmem_bootmap_pages(end_pfn)<<PAGE_SHIFT;
++ bootmap = find_e820_area(0, end_pfn<<PAGE_SHIFT, bootmap_size);
++ if (bootmap == -1L)
++ panic("Cannot find bootmem map of size %ld\n",bootmap_size);
++ bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, end_pfn);
++#ifdef CONFIG_XEN
++ e820_bootmem_free(NODE_DATA(0), 0, xen_start_info->nr_pages<<PAGE_SHIFT);
++#else
++ e820_bootmem_free(NODE_DATA(0), 0, end_pfn << PAGE_SHIFT);
++#endif
++ reserve_bootmem(bootmap, bootmap_size);
++}
++#endif
++
++#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE)
++struct edd edd;
++#ifdef CONFIG_EDD_MODULE
++EXPORT_SYMBOL(edd);
++#endif
++#ifndef CONFIG_XEN
++/**
++ * copy_edd() - Copy the BIOS EDD information
++ * from boot_params into a safe place.
++ *
++ */
++static inline void copy_edd(void)
++{
++ memcpy(edd.mbr_signature, EDD_MBR_SIGNATURE, sizeof(edd.mbr_signature));
++ memcpy(edd.edd_info, EDD_BUF, sizeof(edd.edd_info));
++ edd.mbr_signature_nr = EDD_MBR_SIG_NR;
++ edd.edd_info_nr = EDD_NR;
++}
++#endif
++#else
++static inline void copy_edd(void)
++{
++}
++#endif
++
++#ifndef CONFIG_XEN
++#define EBDA_ADDR_POINTER 0x40E
++
++unsigned __initdata ebda_addr;
++unsigned __initdata ebda_size;
++
++static void discover_ebda(void)
++{
++ /*
++ * there is a real-mode segmented pointer pointing to the
++ * 4K EBDA area at 0x40E
++ */
++ ebda_addr = *(unsigned short *)EBDA_ADDR_POINTER;
++ ebda_addr <<= 4;
++
++ ebda_size = *(unsigned short *)(unsigned long)ebda_addr;
++
++ /* Round EBDA up to pages */
++ if (ebda_size == 0)
++ ebda_size = 1;
++ ebda_size <<= 10;
++ ebda_size = round_up(ebda_size + (ebda_addr & ~PAGE_MASK), PAGE_SIZE);
++ if (ebda_size > 64*1024)
++ ebda_size = 64*1024;
++}
++#else
++#define discover_ebda() ((void)0)
++#endif
++
++void __init setup_arch(char **cmdline_p)
++{
++#ifdef CONFIG_XEN
++ /* Register a call for panic conditions. */
++ atomic_notifier_chain_register(&panic_notifier_list, &xen_panic_block);
++
++ ROOT_DEV = MKDEV(RAMDISK_MAJOR,0);
++ screen_info = SCREEN_INFO;
++
++ if (is_initial_xendomain()) {
++ const struct dom0_vga_console_info *info =
++ (void *)((char *)xen_start_info +
++ xen_start_info->console.dom0.info_off);
++
++ dom0_init_screen_info(info,
++ xen_start_info->console.dom0.info_size);
++ xen_start_info->console.domU.mfn = 0;
++ xen_start_info->console.domU.evtchn = 0;
++ } else
++ screen_info.orig_video_isVGA = 0;
++
++ copy_edid();
++
++ WARN_ON(HYPERVISOR_vm_assist(VMASST_CMD_enable,
++ VMASST_TYPE_writable_pagetables));
++
++ ARCH_SETUP
++#else
++ ROOT_DEV = old_decode_dev(ORIG_ROOT_DEV);
++ screen_info = SCREEN_INFO;
++ edid_info = EDID_INFO;
++#endif /* !CONFIG_XEN */
++ saved_video_mode = SAVED_VIDEO_MODE;
++ bootloader_type = LOADER_TYPE;
++
++#ifdef CONFIG_BLK_DEV_RAM
++ rd_image_start = RAMDISK_FLAGS & RAMDISK_IMAGE_START_MASK;
++ rd_prompt = ((RAMDISK_FLAGS & RAMDISK_PROMPT_FLAG) != 0);
++ rd_doload = ((RAMDISK_FLAGS & RAMDISK_LOAD_FLAG) != 0);
++#endif
++ setup_memory_region();
++ copy_edd();
++
++ if (!MOUNT_ROOT_RDONLY)
++ root_mountflags &= ~MS_RDONLY;
++ init_mm.start_code = (unsigned long) &_text;
++ init_mm.end_code = (unsigned long) &_etext;
++ init_mm.end_data = (unsigned long) &_edata;
++ init_mm.brk = (unsigned long) &_end;
++
++ code_resource.start = virt_to_phys(&_text);
++ code_resource.end = virt_to_phys(&_etext)-1;
++ data_resource.start = virt_to_phys(&_etext);
++ data_resource.end = virt_to_phys(&_edata)-1;
++
++ parse_cmdline_early(cmdline_p);
++
++ early_identify_cpu(&boot_cpu_data);
++
++ /*
++ * partially used pages are not usable - thus
++ * we are rounding upwards:
++ */
++ end_pfn = e820_end_of_ram();
++ num_physpages = end_pfn; /* for pfn_valid */
++
++ check_efer();
++
++ discover_ebda();
++
++ init_memory_mapping(0, (end_pfn_map << PAGE_SHIFT));
++
++ if (is_initial_xendomain())
++ dmi_scan_machine();
++
++#ifdef CONFIG_ACPI_NUMA
++ /*
++ * Parse SRAT to discover nodes.
++ */
++ acpi_numa_init();
++#endif
++
++#ifdef CONFIG_NUMA
++ numa_initmem_init(0, end_pfn);
++#else
++ contig_initmem_init(0, end_pfn);
++#endif
++
++#ifdef CONFIG_XEN
++ /*
++ * Reserve kernel, physmap, start info, initial page tables, and
++ * direct mapping.
++ */
++ reserve_bootmem_generic(__pa_symbol(&_text),
++ (table_end << PAGE_SHIFT) - __pa_symbol(&_text));
++#else
++ /* Reserve direct mapping */
++ reserve_bootmem_generic(table_start << PAGE_SHIFT,
++ (table_end - table_start) << PAGE_SHIFT);
++
++ /* reserve kernel */
++ reserve_bootmem_generic(__pa_symbol(&_text),
++ __pa_symbol(&_end) - __pa_symbol(&_text));
++
++ /*
++ * reserve physical page 0 - it's a special BIOS page on many boxes,
++ * enabling clean reboots, SMP operation, laptop functions.
++ */
++ reserve_bootmem_generic(0, PAGE_SIZE);
++
++ /* reserve ebda region */
++ if (ebda_addr)
++ reserve_bootmem_generic(ebda_addr, ebda_size);
++
++#ifdef CONFIG_SMP
++ /*
++ * But first pinch a few for the stack/trampoline stuff
++ * FIXME: Don't need the extra page at 4K, but need to fix
++ * trampoline before removing it. (see the GDT stuff)
++ */
++ reserve_bootmem_generic(PAGE_SIZE, PAGE_SIZE);
++
++ /* Reserve SMP trampoline */
++ reserve_bootmem_generic(SMP_TRAMPOLINE_BASE, PAGE_SIZE);
++#endif
++#endif
++
++#ifdef CONFIG_ACPI_SLEEP
++ /*
++ * Reserve low memory region for sleep support.
++ */
++ acpi_reserve_bootmem();
++#endif
++#ifdef CONFIG_XEN
++#ifdef CONFIG_BLK_DEV_INITRD
++ if (xen_start_info->mod_start) {
++ if (INITRD_START + INITRD_SIZE <= (end_pfn << PAGE_SHIFT)) {
++ /*reserve_bootmem_generic(INITRD_START, INITRD_SIZE);*/
++ initrd_start = INITRD_START + PAGE_OFFSET;
++ initrd_end = initrd_start+INITRD_SIZE;
++ initrd_below_start_ok = 1;
++ } else {
++ printk(KERN_ERR "initrd extends beyond end of memory "
++ "(0x%08lx > 0x%08lx)\ndisabling initrd\n",
++ (unsigned long)(INITRD_START + INITRD_SIZE),
++ (unsigned long)(end_pfn << PAGE_SHIFT));
++ initrd_start = 0;
++ }
++ }
++#endif
++#else /* CONFIG_XEN */
++#ifdef CONFIG_BLK_DEV_INITRD
++ if (LOADER_TYPE && INITRD_START) {
++ if (INITRD_START + INITRD_SIZE <= (end_pfn << PAGE_SHIFT)) {
++ reserve_bootmem_generic(INITRD_START, INITRD_SIZE);
++ initrd_start =
++ INITRD_START ? INITRD_START + PAGE_OFFSET : 0;
++ initrd_end = initrd_start+INITRD_SIZE;
++ }
++ else {
++ printk(KERN_ERR "initrd extends beyond end of memory "
++ "(0x%08lx > 0x%08lx)\ndisabling initrd\n",
++ (unsigned long)(INITRD_START + INITRD_SIZE),
++ (unsigned long)(end_pfn << PAGE_SHIFT));
++ initrd_start = 0;
++ }
++ }
++#endif
++#endif /* !CONFIG_XEN */
++#ifdef CONFIG_KEXEC
++#ifdef CONFIG_XEN
++ xen_machine_kexec_setup_resources();
++#else
++ if (crashk_res.start != crashk_res.end) {
++ reserve_bootmem_generic(crashk_res.start,
++ crashk_res.end - crashk_res.start + 1);
++ }
++#endif
++#endif
++
++ paging_init();
++#ifdef CONFIG_X86_LOCAL_APIC
++ /*
++ * Find and reserve possible boot-time SMP configuration:
++ */
++ find_smp_config();
++#endif
++#ifdef CONFIG_XEN
++ {
++ int i, j, k, fpp;
++ unsigned long p2m_pages;
++
++ p2m_pages = end_pfn;
++ if (xen_start_info->nr_pages > end_pfn) {
++ /*
++ * the end_pfn was shrunk (probably by mem= or highmem=
++ * kernel parameter); shrink reservation with the HV
++ */
++ struct xen_memory_reservation reservation = {
++ .address_bits = 0,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
++ unsigned int difference;
++ int ret;
++
++ difference = xen_start_info->nr_pages - end_pfn;
++
++ set_xen_guest_handle(reservation.extent_start,
++ ((unsigned long *)xen_start_info->mfn_list) + end_pfn);
++ reservation.nr_extents = difference;
++ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &reservation);
++ BUG_ON (ret != difference);
++ }
++ else if (end_pfn > xen_start_info->nr_pages)
++ p2m_pages = xen_start_info->nr_pages;
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* Make sure we have a large enough P->M table. */
++ phys_to_machine_mapping = alloc_bootmem_pages(
++ end_pfn * sizeof(unsigned long));
++ memset(phys_to_machine_mapping, ~0,
++ end_pfn * sizeof(unsigned long));
++ memcpy(phys_to_machine_mapping,
++ (unsigned long *)xen_start_info->mfn_list,
++ p2m_pages * sizeof(unsigned long));
++ free_bootmem(
++ __pa(xen_start_info->mfn_list),
++ PFN_PHYS(PFN_UP(xen_start_info->nr_pages *
++ sizeof(unsigned long))));
++
++ /*
++ * Initialise the list of the frames that specify the
++ * list of frames that make up the p2m table. Used by
++ * save/restore.
++ */
++ pfn_to_mfn_frame_list_list = alloc_bootmem_pages(PAGE_SIZE);
++
++ fpp = PAGE_SIZE/sizeof(unsigned long);
++ for (i=0, j=0, k=-1; i< end_pfn; i+=fpp, j++) {
++ if ((j % fpp) == 0) {
++ k++;
++ BUG_ON(k>=fpp);
++ pfn_to_mfn_frame_list[k] =
++ alloc_bootmem_pages(PAGE_SIZE);
++ pfn_to_mfn_frame_list_list[k] =
++ virt_to_mfn(pfn_to_mfn_frame_list[k]);
++ j=0;
++ }
++ pfn_to_mfn_frame_list[k][j] =
++ virt_to_mfn(&phys_to_machine_mapping[i]);
++ }
++ HYPERVISOR_shared_info->arch.max_pfn = end_pfn;
++ HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list =
++ virt_to_mfn(pfn_to_mfn_frame_list_list);
++ }
++
++ /* Mark all ISA DMA channels in-use - using them wouldn't work. */
++ for (i = 0; i < MAX_DMA_CHANNELS; ++i)
++ if (i != 4 && request_dma(i, "xen") != 0)
++ BUG();
++ }
++
++ if (!is_initial_xendomain()) {
++ acpi_disabled = 1;
++#ifdef CONFIG_ACPI
++ acpi_ht = 0;
++#endif
++ }
++#endif
++
++#ifndef CONFIG_XEN
++ check_ioapic();
++#endif
++
++ zap_low_mappings(0);
++
++ /*
++ * set this early, so we dont allocate cpu0
++ * if MADT list doesnt list BSP first
++ * mpparse.c/MP_processor_info() allocates logical cpu numbers.
++ */
++ cpu_set(0, cpu_present_map);
++#ifdef CONFIG_ACPI
++ /*
++ * Initialize the ACPI boot-time table parser (gets the RSDP and SDT).
++ * Call this early for SRAT node setup.
++ */
++ acpi_boot_table_init();
++
++ /*
++ * Read APIC and some other early information from ACPI tables.
++ */
++ acpi_boot_init();
++#endif
++
++ init_cpu_to_node();
++
++#ifdef CONFIG_X86_LOCAL_APIC
++ /*
++ * get boot-time SMP configuration:
++ */
++ if (smp_found_config)
++ get_smp_config();
++#ifndef CONFIG_XEN
++ init_apic_mappings();
++#endif
++#endif
++#if defined(CONFIG_XEN) && defined(CONFIG_SMP) && !defined(CONFIG_HOTPLUG_CPU)
++ prefill_possible_map();
++#endif
++
++ /*
++ * Request address space for all standard RAM and ROM resources
++ * and also for regions reported as reserved by the e820.
++ */
++ probe_roms();
++#ifdef CONFIG_XEN
++ if (is_initial_xendomain())
++ e820_reserve_resources(machine_e820.map, machine_e820.nr_map);
++#else
++ e820_reserve_resources(e820.map, e820.nr_map);
++#endif
++
++ request_resource(&iomem_resource, &video_ram_resource);
++
++ {
++ unsigned i;
++ /* request I/O space for devices used on all i[345]86 PCs */
++ for (i = 0; i < STANDARD_IO_RESOURCES; i++)
++ request_resource(&ioport_resource, &standard_io_resources[i]);
++ }
++
++#ifdef CONFIG_XEN
++ if (is_initial_xendomain())
++ e820_setup_gap(machine_e820.map, machine_e820.nr_map);
++#else
++ e820_setup_gap(e820.map, e820.nr_map);
++#endif
++
++#ifdef CONFIG_XEN
++ {
++ struct physdev_set_iopl set_iopl;
++
++ set_iopl.iopl = 1;
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl));
++
++ if (is_initial_xendomain()) {
++#ifdef CONFIG_VT
++#if defined(CONFIG_VGA_CONSOLE)
++ conswitchp = &vga_con;
++#elif defined(CONFIG_DUMMY_CONSOLE)
++ conswitchp = &dummy_con;
++#endif
++#endif
++ } else {
++#if defined(CONFIG_VT) && defined(CONFIG_DUMMY_CONSOLE)
++ conswitchp = &dummy_con;
++#endif
++ }
++ }
++#else /* CONFIG_XEN */
++
++#ifdef CONFIG_VT
++#if defined(CONFIG_VGA_CONSOLE)
++ conswitchp = &vga_con;
++#elif defined(CONFIG_DUMMY_CONSOLE)
++ conswitchp = &dummy_con;
++#endif
++#endif
++
++#endif /* !CONFIG_XEN */
++}
++
++#ifdef CONFIG_XEN
++static int
++xen_panic_event(struct notifier_block *this, unsigned long event, void *ptr)
++{
++ HYPERVISOR_shutdown(SHUTDOWN_crash);
++ /* we're never actually going to get here... */
++ return NOTIFY_DONE;
++}
++#endif /* !CONFIG_XEN */
++
++
++static int __cpuinit get_model_name(struct cpuinfo_x86 *c)
++{
++ unsigned int *v;
++
++ if (c->extended_cpuid_level < 0x80000004)
++ return 0;
++
++ v = (unsigned int *) c->x86_model_id;
++ cpuid(0x80000002, &v[0], &v[1], &v[2], &v[3]);
++ cpuid(0x80000003, &v[4], &v[5], &v[6], &v[7]);
++ cpuid(0x80000004, &v[8], &v[9], &v[10], &v[11]);
++ c->x86_model_id[48] = 0;
++ return 1;
++}
++
++
++static void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c)
++{
++ unsigned int n, dummy, eax, ebx, ecx, edx;
++
++ n = c->extended_cpuid_level;
++
++ if (n >= 0x80000005) {
++ cpuid(0x80000005, &dummy, &ebx, &ecx, &edx);
++ printk(KERN_INFO "CPU: L1 I Cache: %dK (%d bytes/line), D cache %dK (%d bytes/line)\n",
++ edx>>24, edx&0xFF, ecx>>24, ecx&0xFF);
++ c->x86_cache_size=(ecx>>24)+(edx>>24);
++ /* On K8 L1 TLB is inclusive, so don't count it */
++ c->x86_tlbsize = 0;
++ }
++
++ if (n >= 0x80000006) {
++ cpuid(0x80000006, &dummy, &ebx, &ecx, &edx);
++ ecx = cpuid_ecx(0x80000006);
++ c->x86_cache_size = ecx >> 16;
++ c->x86_tlbsize += ((ebx >> 16) & 0xfff) + (ebx & 0xfff);
++
++ printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n",
++ c->x86_cache_size, ecx & 0xFF);
++ }
++
++ if (n >= 0x80000007)
++ cpuid(0x80000007, &dummy, &dummy, &dummy, &c->x86_power);
++ if (n >= 0x80000008) {
++ cpuid(0x80000008, &eax, &dummy, &dummy, &dummy);
++ c->x86_virt_bits = (eax >> 8) & 0xff;
++ c->x86_phys_bits = eax & 0xff;
++ }
++}
++
++#ifdef CONFIG_NUMA
++static int nearby_node(int apicid)
++{
++ int i;
++ for (i = apicid - 1; i >= 0; i--) {
++ int node = apicid_to_node[i];
++ if (node != NUMA_NO_NODE && node_online(node))
++ return node;
++ }
++ for (i = apicid + 1; i < MAX_LOCAL_APIC; i++) {
++ int node = apicid_to_node[i];
++ if (node != NUMA_NO_NODE && node_online(node))
++ return node;
++ }
++ return first_node(node_online_map); /* Shouldn't happen */
++}
++#endif
++
++/*
++ * On a AMD dual core setup the lower bits of the APIC id distingush the cores.
++ * Assumes number of cores is a power of two.
++ */
++static void __init amd_detect_cmp(struct cpuinfo_x86 *c)
++{
++#ifdef CONFIG_SMP
++ unsigned bits;
++#ifdef CONFIG_NUMA
++ int cpu = smp_processor_id();
++ int node = 0;
++ unsigned apicid = hard_smp_processor_id();
++#endif
++ unsigned ecx = cpuid_ecx(0x80000008);
++
++ c->x86_max_cores = (ecx & 0xff) + 1;
++
++ /* CPU telling us the core id bits shift? */
++ bits = (ecx >> 12) & 0xF;
++
++ /* Otherwise recompute */
++ if (bits == 0) {
++ while ((1 << bits) < c->x86_max_cores)
++ bits++;
++ }
++
++ /* Low order bits define the core id (index of core in socket) */
++ c->cpu_core_id = c->phys_proc_id & ((1 << bits)-1);
++ /* Convert the APIC ID into the socket ID */
++ c->phys_proc_id = phys_pkg_id(bits);
++
++#ifdef CONFIG_NUMA
++ node = c->phys_proc_id;
++ if (apicid_to_node[apicid] != NUMA_NO_NODE)
++ node = apicid_to_node[apicid];
++ if (!node_online(node)) {
++ /* Two possibilities here:
++ - The CPU is missing memory and no node was created.
++ In that case try picking one from a nearby CPU
++ - The APIC IDs differ from the HyperTransport node IDs
++ which the K8 northbridge parsing fills in.
++ Assume they are all increased by a constant offset,
++ but in the same order as the HT nodeids.
++ If that doesn't result in a usable node fall back to the
++ path for the previous case. */
++ int ht_nodeid = apicid - (cpu_data[0].phys_proc_id << bits);
++ if (ht_nodeid >= 0 &&
++ apicid_to_node[ht_nodeid] != NUMA_NO_NODE)
++ node = apicid_to_node[ht_nodeid];
++ /* Pick a nearby node */
++ if (!node_online(node))
++ node = nearby_node(apicid);
++ }
++ numa_set_node(cpu, node);
++
++ printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node);
++#endif
++#endif
++}
++
++static void __init init_amd(struct cpuinfo_x86 *c)
++{
++ unsigned level;
++
++#ifdef CONFIG_SMP
++ unsigned long value;
++
++ /*
++ * Disable TLB flush filter by setting HWCR.FFDIS on K8
++ * bit 6 of msr C001_0015
++ *
++ * Errata 63 for SH-B3 steppings
++ * Errata 122 for all steppings (F+ have it disabled by default)
++ */
++ if (c->x86 == 15) {
++ rdmsrl(MSR_K8_HWCR, value);
++ value |= 1 << 6;
++ wrmsrl(MSR_K8_HWCR, value);
++ }
++#endif
++
++ /* Bit 31 in normal CPUID used for nonstandard 3DNow ID;
++ 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway */
++ clear_bit(0*32+31, &c->x86_capability);
++
++ /* On C+ stepping K8 rep microcode works well for copy/memset */
++ level = cpuid_eax(1);
++ if (c->x86 == 15 && ((level >= 0x0f48 && level < 0x0f50) || level >= 0x0f58))
++ set_bit(X86_FEATURE_REP_GOOD, &c->x86_capability);
++
++ /* Enable workaround for FXSAVE leak */
++ if (c->x86 >= 6)
++ set_bit(X86_FEATURE_FXSAVE_LEAK, &c->x86_capability);
++
++ level = get_model_name(c);
++ if (!level) {
++ switch (c->x86) {
++ case 15:
++ /* Should distinguish Models here, but this is only
++ a fallback anyways. */
++ strcpy(c->x86_model_id, "Hammer");
++ break;
++ }
++ }
++ display_cacheinfo(c);
++
++ /* c->x86_power is 8000_0007 edx. Bit 8 is constant TSC */
++ if (c->x86_power & (1<<8))
++ set_bit(X86_FEATURE_CONSTANT_TSC, &c->x86_capability);
++
++ /* Multi core CPU? */
++ if (c->extended_cpuid_level >= 0x80000008)
++ amd_detect_cmp(c);
++
++ /* Fix cpuid4 emulation for more */
++ num_cache_leaves = 3;
++}
++
++static void __cpuinit detect_ht(struct cpuinfo_x86 *c)
++{
++#ifdef CONFIG_SMP
++ u32 eax, ebx, ecx, edx;
++ int index_msb, core_bits;
++
++ cpuid(1, &eax, &ebx, &ecx, &edx);
++
++
++ if (!cpu_has(c, X86_FEATURE_HT))
++ return;
++ if (cpu_has(c, X86_FEATURE_CMP_LEGACY))
++ goto out;
++
++ smp_num_siblings = (ebx & 0xff0000) >> 16;
++
++ if (smp_num_siblings == 1) {
++ printk(KERN_INFO "CPU: Hyper-Threading is disabled\n");
++ } else if (smp_num_siblings > 1 ) {
++
++ if (smp_num_siblings > NR_CPUS) {
++ printk(KERN_WARNING "CPU: Unsupported number of the siblings %d", smp_num_siblings);
++ smp_num_siblings = 1;
++ return;
++ }
++
++ index_msb = get_count_order(smp_num_siblings);
++ c->phys_proc_id = phys_pkg_id(index_msb);
++
++ smp_num_siblings = smp_num_siblings / c->x86_max_cores;
++
++ index_msb = get_count_order(smp_num_siblings) ;
++
++ core_bits = get_count_order(c->x86_max_cores);
++
++ c->cpu_core_id = phys_pkg_id(index_msb) &
++ ((1 << core_bits) - 1);
++ }
++out:
++ if ((c->x86_max_cores * smp_num_siblings) > 1) {
++ printk(KERN_INFO "CPU: Physical Processor ID: %d\n", c->phys_proc_id);
++ printk(KERN_INFO "CPU: Processor Core ID: %d\n", c->cpu_core_id);
++ }
++
++#endif
++}
++
++/*
++ * find out the number of processor cores on the die
++ */
++static int __cpuinit intel_num_cpu_cores(struct cpuinfo_x86 *c)
++{
++ unsigned int eax, t;
++
++ if (c->cpuid_level < 4)
++ return 1;
++
++ cpuid_count(4, 0, &eax, &t, &t, &t);
++
++ if (eax & 0x1f)
++ return ((eax >> 26) + 1);
++ else
++ return 1;
++}
++
++static void srat_detect_node(void)
++{
++#ifdef CONFIG_NUMA
++ unsigned node;
++ int cpu = smp_processor_id();
++ int apicid = hard_smp_processor_id();
++
++ /* Don't do the funky fallback heuristics the AMD version employs
++ for now. */
++ node = apicid_to_node[apicid];
++ if (node == NUMA_NO_NODE)
++ node = first_node(node_online_map);
++ numa_set_node(cpu, node);
++
++ if (acpi_numa > 0)
++ printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node);
++#endif
++}
++
++static void __cpuinit init_intel(struct cpuinfo_x86 *c)
++{
++ /* Cache sizes */
++ unsigned n;
++
++ init_intel_cacheinfo(c);
++ if (c->cpuid_level > 9 ) {
++ unsigned eax = cpuid_eax(10);
++ /* Check for version and the number of counters */
++ if ((eax & 0xff) && (((eax>>8) & 0xff) > 1))
++ set_bit(X86_FEATURE_ARCH_PERFMON, &c->x86_capability);
++ }
++
++ n = c->extended_cpuid_level;
++ if (n >= 0x80000008) {
++ unsigned eax = cpuid_eax(0x80000008);
++ c->x86_virt_bits = (eax >> 8) & 0xff;
++ c->x86_phys_bits = eax & 0xff;
++ /* CPUID workaround for Intel 0F34 CPU */
++ if (c->x86_vendor == X86_VENDOR_INTEL &&
++ c->x86 == 0xF && c->x86_model == 0x3 &&
++ c->x86_mask == 0x4)
++ c->x86_phys_bits = 36;
++ }
++
++ if (c->x86 == 15)
++ c->x86_cache_alignment = c->x86_clflush_size * 2;
++ if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
++ (c->x86 == 0x6 && c->x86_model >= 0x0e))
++ set_bit(X86_FEATURE_CONSTANT_TSC, &c->x86_capability);
++ set_bit(X86_FEATURE_SYNC_RDTSC, &c->x86_capability);
++ c->x86_max_cores = intel_num_cpu_cores(c);
++
++ srat_detect_node();
++}
++
++static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c)
++{
++ char *v = c->x86_vendor_id;
++
++ if (!strcmp(v, "AuthenticAMD"))
++ c->x86_vendor = X86_VENDOR_AMD;
++ else if (!strcmp(v, "GenuineIntel"))
++ c->x86_vendor = X86_VENDOR_INTEL;
++ else
++ c->x86_vendor = X86_VENDOR_UNKNOWN;
++}
++
++struct cpu_model_info {
++ int vendor;
++ int family;
++ char *model_names[16];
++};
++
++/* Do some early cpuid on the boot CPU to get some parameter that are
++ needed before check_bugs. Everything advanced is in identify_cpu
++ below. */
++void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c)
++{
++ u32 tfms;
++
++ c->loops_per_jiffy = loops_per_jiffy;
++ c->x86_cache_size = -1;
++ c->x86_vendor = X86_VENDOR_UNKNOWN;
++ c->x86_model = c->x86_mask = 0; /* So far unknown... */
++ c->x86_vendor_id[0] = '\0'; /* Unset */
++ c->x86_model_id[0] = '\0'; /* Unset */
++ c->x86_clflush_size = 64;
++ c->x86_cache_alignment = c->x86_clflush_size;
++ c->x86_max_cores = 1;
++ c->extended_cpuid_level = 0;
++ memset(&c->x86_capability, 0, sizeof c->x86_capability);
++
++ /* Get vendor name */
++ cpuid(0x00000000, (unsigned int *)&c->cpuid_level,
++ (unsigned int *)&c->x86_vendor_id[0],
++ (unsigned int *)&c->x86_vendor_id[8],
++ (unsigned int *)&c->x86_vendor_id[4]);
++
++ get_cpu_vendor(c);
++
++ /* Initialize the standard set of capabilities */
++ /* Note that the vendor-specific code below might override */
++
++ /* Intel-defined flags: level 0x00000001 */
++ if (c->cpuid_level >= 0x00000001) {
++ __u32 misc;
++ cpuid(0x00000001, &tfms, &misc, &c->x86_capability[4],
++ &c->x86_capability[0]);
++ c->x86 = (tfms >> 8) & 0xf;
++ c->x86_model = (tfms >> 4) & 0xf;
++ c->x86_mask = tfms & 0xf;
++ if (c->x86 == 0xf)
++ c->x86 += (tfms >> 20) & 0xff;
++ if (c->x86 >= 0x6)
++ c->x86_model += ((tfms >> 16) & 0xF) << 4;
++ if (c->x86_capability[0] & (1<<19))
++ c->x86_clflush_size = ((misc >> 8) & 0xff) * 8;
++ } else {
++ /* Have CPUID level 0 only - unheard of */
++ c->x86 = 4;
++ }
++
++#ifdef CONFIG_SMP
++ c->phys_proc_id = (cpuid_ebx(1) >> 24) & 0xff;
++#endif
++}
++
++/*
++ * This does the hard work of actually picking apart the CPU stuff...
++ */
++void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
++{
++ int i;
++ u32 xlvl;
++
++ early_identify_cpu(c);
++
++ /* AMD-defined flags: level 0x80000001 */
++ xlvl = cpuid_eax(0x80000000);
++ c->extended_cpuid_level = xlvl;
++ if ((xlvl & 0xffff0000) == 0x80000000) {
++ if (xlvl >= 0x80000001) {
++ c->x86_capability[1] = cpuid_edx(0x80000001);
++ c->x86_capability[6] = cpuid_ecx(0x80000001);
++ }
++ if (xlvl >= 0x80000004)
++ get_model_name(c); /* Default name */
++ }
++
++ /* Transmeta-defined flags: level 0x80860001 */
++ xlvl = cpuid_eax(0x80860000);
++ if ((xlvl & 0xffff0000) == 0x80860000) {
++ /* Don't set x86_cpuid_level here for now to not confuse. */
++ if (xlvl >= 0x80860001)
++ c->x86_capability[2] = cpuid_edx(0x80860001);
++ }
++
++ c->apicid = phys_pkg_id(0);
++
++ /*
++ * Vendor-specific initialization. In this section we
++ * canonicalize the feature flags, meaning if there are
++ * features a certain CPU supports which CPUID doesn't
++ * tell us, CPUID claiming incorrect flags, or other bugs,
++ * we handle them here.
++ *
++ * At the end of this section, c->x86_capability better
++ * indicate the features this CPU genuinely supports!
++ */
++ switch (c->x86_vendor) {
++ case X86_VENDOR_AMD:
++ init_amd(c);
++ break;
++
++ case X86_VENDOR_INTEL:
++ init_intel(c);
++ break;
++
++ case X86_VENDOR_UNKNOWN:
++ default:
++ display_cacheinfo(c);
++ break;
++ }
++
++ select_idle_routine(c);
++ detect_ht(c);
++
++ /*
++ * On SMP, boot_cpu_data holds the common feature set between
++ * all CPUs; so make sure that we indicate which features are
++ * common between the CPUs. The first time this routine gets
++ * executed, c == &boot_cpu_data.
++ */
++ if (c != &boot_cpu_data) {
++ /* AND the already accumulated flags with these */
++ for (i = 0 ; i < NCAPINTS ; i++)
++ boot_cpu_data.x86_capability[i] &= c->x86_capability[i];
++ }
++
++#ifdef CONFIG_X86_MCE
++ mcheck_init(c);
++#endif
++ if (c == &boot_cpu_data)
++ mtrr_bp_init();
++ else
++ mtrr_ap_init();
++#ifdef CONFIG_NUMA
++ numa_add_cpu(smp_processor_id());
++#endif
++}
++
++
++void __cpuinit print_cpu_info(struct cpuinfo_x86 *c)
++{
++ if (c->x86_model_id[0])
++ printk("%s", c->x86_model_id);
++
++ if (c->x86_mask || c->cpuid_level >= 0)
++ printk(" stepping %02x\n", c->x86_mask);
++ else
++ printk("\n");
++}
++
++/*
++ * Get CPU information for use by the procfs.
++ */
++
++static int show_cpuinfo(struct seq_file *m, void *v)
++{
++ struct cpuinfo_x86 *c = v;
++
++ /*
++ * These flag bits must match the definitions in <asm/cpufeature.h>.
++ * NULL means this bit is undefined or reserved; either way it doesn't
++ * have meaning as far as Linux is concerned. Note that it's important
++ * to realize there is a difference between this table and CPUID -- if
++ * applications want to get the raw CPUID data, they should access
++ * /dev/cpu/<cpu_nr>/cpuid instead.
++ */
++ static char *x86_cap_flags[] = {
++ /* Intel-defined */
++ "fpu", "vme", "de", "pse", "tsc", "msr", "pae", "mce",
++ "cx8", "apic", NULL, "sep", "mtrr", "pge", "mca", "cmov",
++ "pat", "pse36", "pn", "clflush", NULL, "dts", "acpi", "mmx",
++ "fxsr", "sse", "sse2", "ss", "ht", "tm", "ia64", NULL,
++
++ /* AMD-defined */
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ NULL, NULL, NULL, "syscall", NULL, NULL, NULL, NULL,
++ NULL, NULL, NULL, NULL, "nx", NULL, "mmxext", NULL,
++ NULL, "fxsr_opt", NULL, "rdtscp", NULL, "lm", "3dnowext", "3dnow",
++
++ /* Transmeta-defined */
++ "recovery", "longrun", NULL, "lrti", NULL, NULL, NULL, NULL,
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++
++ /* Other (Linux-defined) */
++ "cxmmx", NULL, "cyrix_arr", "centaur_mcr", NULL,
++ "constant_tsc", NULL, NULL,
++ "up", NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++
++ /* Intel-defined (#2) */
++ "pni", NULL, NULL, "monitor", "ds_cpl", "vmx", "smx", "est",
++ "tm2", NULL, "cid", NULL, NULL, "cx16", "xtpr", NULL,
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++
++ /* VIA/Cyrix/Centaur-defined */
++ NULL, NULL, "rng", "rng_en", NULL, NULL, "ace", "ace_en",
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++
++ /* AMD-defined (#2) */
++ "lahf_lm", "cmp_legacy", "svm", NULL, "cr8_legacy", NULL, NULL, NULL,
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ };
++ static char *x86_power_flags[] = {
++ "ts", /* temperature sensor */
++ "fid", /* frequency id control */
++ "vid", /* voltage id control */
++ "ttp", /* thermal trip */
++ "tm",
++ "stc",
++ NULL,
++ /* nothing */ /* constant_tsc - moved to flags */
++ };
++
++
++#ifdef CONFIG_SMP
++ if (!cpu_online(c-cpu_data))
++ return 0;
++#endif
++
++ seq_printf(m,"processor\t: %u\n"
++ "vendor_id\t: %s\n"
++ "cpu family\t: %d\n"
++ "model\t\t: %d\n"
++ "model name\t: %s\n",
++ (unsigned)(c-cpu_data),
++ c->x86_vendor_id[0] ? c->x86_vendor_id : "unknown",
++ c->x86,
++ (int)c->x86_model,
++ c->x86_model_id[0] ? c->x86_model_id : "unknown");
++
++ if (c->x86_mask || c->cpuid_level >= 0)
++ seq_printf(m, "stepping\t: %d\n", c->x86_mask);
++ else
++ seq_printf(m, "stepping\t: unknown\n");
++
++ if (cpu_has(c,X86_FEATURE_TSC)) {
++ unsigned int freq = cpufreq_quick_get((unsigned)(c-cpu_data));
++ if (!freq)
++ freq = cpu_khz;
++ seq_printf(m, "cpu MHz\t\t: %u.%03u\n",
++ freq / 1000, (freq % 1000));
++ }
++
++ /* Cache size */
++ if (c->x86_cache_size >= 0)
++ seq_printf(m, "cache size\t: %d KB\n", c->x86_cache_size);
++
++#ifdef CONFIG_SMP
++ if (smp_num_siblings * c->x86_max_cores > 1) {
++ int cpu = c - cpu_data;
++ seq_printf(m, "physical id\t: %d\n", c->phys_proc_id);
++ seq_printf(m, "siblings\t: %d\n", cpus_weight(cpu_core_map[cpu]));
++ seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id);
++ seq_printf(m, "cpu cores\t: %d\n", c->booted_cores);
++ }
++#endif
++
++ seq_printf(m,
++ "fpu\t\t: yes\n"
++ "fpu_exception\t: yes\n"
++ "cpuid level\t: %d\n"
++ "wp\t\t: yes\n"
++ "flags\t\t:",
++ c->cpuid_level);
++
++ {
++ int i;
++ for ( i = 0 ; i < 32*NCAPINTS ; i++ )
++ if (cpu_has(c, i) && x86_cap_flags[i] != NULL)
++ seq_printf(m, " %s", x86_cap_flags[i]);
++ }
++
++ seq_printf(m, "\nbogomips\t: %lu.%02lu\n",
++ c->loops_per_jiffy/(500000/HZ),
++ (c->loops_per_jiffy/(5000/HZ)) % 100);
++
++ if (c->x86_tlbsize > 0)
++ seq_printf(m, "TLB size\t: %d 4K pages\n", c->x86_tlbsize);
++ seq_printf(m, "clflush size\t: %d\n", c->x86_clflush_size);
++ seq_printf(m, "cache_alignment\t: %d\n", c->x86_cache_alignment);
++
++ seq_printf(m, "address sizes\t: %u bits physical, %u bits virtual\n",
++ c->x86_phys_bits, c->x86_virt_bits);
++
++ seq_printf(m, "power management:");
++ {
++ unsigned i;
++ for (i = 0; i < 32; i++)
++ if (c->x86_power & (1 << i)) {
++ if (i < ARRAY_SIZE(x86_power_flags) &&
++ x86_power_flags[i])
++ seq_printf(m, "%s%s",
++ x86_power_flags[i][0]?" ":"",
++ x86_power_flags[i]);
++ else
++ seq_printf(m, " [%d]", i);
++ }
++ }
++
++ seq_printf(m, "\n\n");
++
++ return 0;
++}
++
++static void *c_start(struct seq_file *m, loff_t *pos)
++{
++ return *pos < NR_CPUS ? cpu_data + *pos : NULL;
++}
++
++static void *c_next(struct seq_file *m, void *v, loff_t *pos)
++{
++ ++*pos;
++ return c_start(m, pos);
++}
++
++static void c_stop(struct seq_file *m, void *v)
++{
++}
++
++struct seq_operations cpuinfo_op = {
++ .start =c_start,
++ .next = c_next,
++ .stop = c_stop,
++ .show = show_cpuinfo,
++};
++
++#if defined(CONFIG_INPUT_PCSPKR) || defined(CONFIG_INPUT_PCSPKR_MODULE)
++#include <linux/platform_device.h>
++static __init int add_pcspkr(void)
++{
++ struct platform_device *pd;
++ int ret;
++
++ if (!is_initial_xendomain())
++ return 0;
++
++ pd = platform_device_alloc("pcspkr", -1);
++ if (!pd)
++ return -ENOMEM;
++
++ ret = platform_device_add(pd);
++ if (ret)
++ platform_device_put(pd);
++
++ return ret;
++}
++device_initcall(add_pcspkr);
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/setup.c
+--- a/arch/x86_64/kernel/setup.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/kernel/setup.c Wed Sep 10 10:54:08 2008 +0100
+@@ -122,7 +122,9 @@
+ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
+ { .name = "timer1", .start = 0x50, .end = 0x53,
+ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
+- { .name = "keyboard", .start = 0x60, .end = 0x6f,
++ { .name = "keyboard", .start = 0x60, .end = 0x60,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "keyboard", .start = 0x64, .end = 0x64,
+ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
+ { .name = "dma page reg", .start = 0x80, .end = 0x8f,
+ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/setup64-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/setup64-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,367 @@
++/*
++ * X86-64 specific CPU setup.
++ * Copyright (C) 1995 Linus Torvalds
++ * Copyright 2001, 2002, 2003 SuSE Labs / Andi Kleen.
++ * See setup.c for older changelog.
++ *
++ * Jun Nakajima <jun.nakajima at intel.com>
++ * Modified for Xen
++ *
++ */
++#include <linux/init.h>
++#include <linux/kernel.h>
++#include <linux/sched.h>
++#include <linux/string.h>
++#include <linux/bootmem.h>
++#include <linux/bitops.h>
++#include <linux/module.h>
++#include <asm/bootsetup.h>
++#include <asm/pda.h>
++#include <asm/pgtable.h>
++#include <asm/processor.h>
++#include <asm/desc.h>
++#include <asm/atomic.h>
++#include <asm/mmu_context.h>
++#include <asm/smp.h>
++#include <asm/i387.h>
++#include <asm/percpu.h>
++#include <asm/proto.h>
++#include <asm/sections.h>
++#ifdef CONFIG_XEN
++#include <asm/hypervisor.h>
++#endif
++
++char x86_boot_params[BOOT_PARAM_SIZE] __initdata = {0,};
++
++cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE;
++
++struct x8664_pda *_cpu_pda[NR_CPUS] __read_mostly;
++EXPORT_SYMBOL(_cpu_pda);
++struct x8664_pda boot_cpu_pda[NR_CPUS] __cacheline_aligned;
++
++#ifndef CONFIG_X86_NO_IDT
++struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table };
++#endif
++
++char boot_cpu_stack[IRQSTACKSIZE] __attribute__((section(".bss.page_aligned")));
++
++unsigned long __supported_pte_mask __read_mostly = ~0UL;
++EXPORT_SYMBOL(__supported_pte_mask);
++static int do_not_nx __cpuinitdata = 0;
++
++/* noexec=on|off
++Control non executable mappings for 64bit processes.
++
++on Enable(default)
++off Disable
++*/
++int __init nonx_setup(char *str)
++{
++ if (!strncmp(str, "on", 2)) {
++ __supported_pte_mask |= _PAGE_NX;
++ do_not_nx = 0;
++ } else if (!strncmp(str, "off", 3)) {
++ do_not_nx = 1;
++ __supported_pte_mask &= ~_PAGE_NX;
++ }
++ return 1;
++}
++__setup("noexec=", nonx_setup); /* parsed early actually */
++
++int force_personality32 = 0;
++
++/* noexec32=on|off
++Control non executable heap for 32bit processes.
++To control the stack too use noexec=off
++
++on PROT_READ does not imply PROT_EXEC for 32bit processes
++off PROT_READ implies PROT_EXEC (default)
++*/
++static int __init nonx32_setup(char *str)
++{
++ if (!strcmp(str, "on"))
++ force_personality32 &= ~READ_IMPLIES_EXEC;
++ else if (!strcmp(str, "off"))
++ force_personality32 |= READ_IMPLIES_EXEC;
++ return 1;
++}
++__setup("noexec32=", nonx32_setup);
++
++/*
++ * Great future plan:
++ * Declare PDA itself and support (irqstack,tss,pgd) as per cpu data.
++ * Always point %gs to its beginning
++ */
++void __init setup_per_cpu_areas(void)
++{
++ int i;
++ unsigned long size;
++
++#ifdef CONFIG_HOTPLUG_CPU
++ prefill_possible_map();
++#endif
++
++ /* Copy section for each CPU (we discard the original) */
++ size = ALIGN(__per_cpu_end - __per_cpu_start, SMP_CACHE_BYTES);
++#ifdef CONFIG_MODULES
++ if (size < PERCPU_ENOUGH_ROOM)
++ size = PERCPU_ENOUGH_ROOM;
++#endif
++
++ for_each_cpu_mask (i, cpu_possible_map) {
++ char *ptr;
++
++ if (!NODE_DATA(cpu_to_node(i))) {
++ printk("cpu with no node %d, num_online_nodes %d\n",
++ i, num_online_nodes());
++ ptr = alloc_bootmem(size);
++ } else {
++ ptr = alloc_bootmem_node(NODE_DATA(cpu_to_node(i)), size);
++ }
++ if (!ptr)
++ panic("Cannot allocate cpu data for CPU %d\n", i);
++ cpu_pda(i)->data_offset = ptr - __per_cpu_start;
++ memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start);
++ }
++}
++
++#ifdef CONFIG_XEN
++static void switch_pt(void)
++{
++ xen_pt_switch(__pa_symbol(init_level4_pgt));
++ xen_new_user_pt(__pa_symbol(__user_pgd(init_level4_pgt)));
++}
++
++static void __cpuinit cpu_gdt_init(const struct desc_ptr *gdt_descr)
++{
++ unsigned long frames[16];
++ unsigned long va;
++ int f;
++
++ for (va = gdt_descr->address, f = 0;
++ va < gdt_descr->address + gdt_descr->size;
++ va += PAGE_SIZE, f++) {
++ frames[f] = virt_to_mfn(va);
++ make_page_readonly(
++ (void *)va, XENFEAT_writable_descriptor_tables);
++ }
++ if (HYPERVISOR_set_gdt(frames, (gdt_descr->size + 1) /
++ sizeof (struct desc_struct)))
++ BUG();
++}
++#else
++static void switch_pt(void)
++{
++ asm volatile("movq %0,%%cr3" :: "r" (__pa_symbol(&init_level4_pgt)));
++}
++
++static void __cpuinit cpu_gdt_init(const struct desc_ptr *gdt_descr)
++{
++ asm volatile("lgdt %0" :: "m" (*gdt_descr));
++ asm volatile("lidt %0" :: "m" (idt_descr));
++}
++#endif
++
++void pda_init(int cpu)
++{
++ struct x8664_pda *pda = cpu_pda(cpu);
++
++ /* Setup up data that may be needed in __get_free_pages early */
++ asm volatile("movl %0,%%fs ; movl %0,%%gs" :: "r" (0));
++#ifndef CONFIG_XEN
++ wrmsrl(MSR_GS_BASE, pda);
++#else
++ if (HYPERVISOR_set_segment_base(SEGBASE_GS_KERNEL,
++ (unsigned long)pda))
++ BUG();
++#endif
++ pda->cpunumber = cpu;
++ pda->irqcount = -1;
++ pda->kernelstack =
++ (unsigned long)stack_thread_info() - PDA_STACKOFFSET + THREAD_SIZE;
++ pda->active_mm = &init_mm;
++ pda->mmu_state = 0;
++
++ if (cpu == 0) {
++#ifdef CONFIG_XEN
++ xen_init_pt();
++#endif
++ /* others are initialized in smpboot.c */
++ pda->pcurrent = &init_task;
++ pda->irqstackptr = boot_cpu_stack;
++ } else {
++ pda->irqstackptr = (char *)
++ __get_free_pages(GFP_ATOMIC, IRQSTACK_ORDER);
++ if (!pda->irqstackptr)
++ panic("cannot allocate irqstack for cpu %d", cpu);
++ }
++
++ switch_pt();
++
++ pda->irqstackptr += IRQSTACKSIZE-64;
++}
++
++#ifndef CONFIG_X86_NO_TSS
++char boot_exception_stacks[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + DEBUG_STKSZ]
++__attribute__((section(".bss.page_aligned")));
++#endif
++
++/* May not be marked __init: used by software suspend */
++void syscall_init(void)
++{
++#ifndef CONFIG_XEN
++ /*
++ * LSTAR and STAR live in a bit strange symbiosis.
++ * They both write to the same internal register. STAR allows to set CS/DS
++ * but only a 32bit target. LSTAR sets the 64bit rip.
++ */
++ wrmsrl(MSR_STAR, ((u64)__USER32_CS)<<48 | ((u64)__KERNEL_CS)<<32);
++ wrmsrl(MSR_LSTAR, system_call);
++
++ /* Flags to clear on syscall */
++ wrmsrl(MSR_SYSCALL_MASK, EF_TF|EF_DF|EF_IE|0x3000);
++#endif
++#ifdef CONFIG_IA32_EMULATION
++ syscall32_cpu_init ();
++#endif
++}
++
++void __cpuinit check_efer(void)
++{
++ unsigned long efer;
++
++ rdmsrl(MSR_EFER, efer);
++ if (!(efer & EFER_NX) || do_not_nx) {
++ __supported_pte_mask &= ~_PAGE_NX;
++ }
++}
++
++unsigned long kernel_eflags;
++
++/*
++ * cpu_init() initializes state that is per-CPU. Some data is already
++ * initialized (naturally) in the bootstrap process, such as the GDT
++ * and IDT. We reload them nevertheless, this function acts as a
++ * 'CPU state barrier', nothing should get across.
++ * A lot of state is already set up in PDA init.
++ */
++void __cpuinit cpu_init (void)
++{
++ int cpu = stack_smp_processor_id();
++#ifndef CONFIG_X86_NO_TSS
++ struct tss_struct *t = &per_cpu(init_tss, cpu);
++ struct orig_ist *orig_ist = &per_cpu(orig_ist, cpu);
++ unsigned long v;
++ char *estacks = NULL;
++ unsigned i;
++#endif
++ struct task_struct *me;
++
++ /* CPU 0 is initialised in head64.c */
++ if (cpu != 0) {
++ pda_init(cpu);
++ zap_low_mappings(cpu);
++ }
++#ifndef CONFIG_X86_NO_TSS
++ else
++ estacks = boot_exception_stacks;
++#endif
++
++ me = current;
++
++ if (cpu_test_and_set(cpu, cpu_initialized))
++ panic("CPU#%d already initialized!\n", cpu);
++
++ printk("Initializing CPU#%d\n", cpu);
++
++ clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE);
++
++ /*
++ * Initialize the per-CPU GDT with the boot GDT,
++ * and set up the GDT descriptor:
++ */
++#ifndef CONFIG_XEN
++ if (cpu)
++ memcpy(cpu_gdt(cpu), cpu_gdt_table, GDT_SIZE);
++#endif
++
++ cpu_gdt_descr[cpu].size = GDT_SIZE;
++ cpu_gdt_init(&cpu_gdt_descr[cpu]);
++
++ memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8);
++ syscall_init();
++
++ wrmsrl(MSR_FS_BASE, 0);
++ wrmsrl(MSR_KERNEL_GS_BASE, 0);
++ barrier();
++
++ check_efer();
++
++#ifndef CONFIG_X86_NO_TSS
++ /*
++ * set up and load the per-CPU TSS
++ */
++ for (v = 0; v < N_EXCEPTION_STACKS; v++) {
++ if (cpu) {
++ static const unsigned int order[N_EXCEPTION_STACKS] = {
++ [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STACK_ORDER,
++ [DEBUG_STACK - 1] = DEBUG_STACK_ORDER
++ };
++
++ estacks = (char *)__get_free_pages(GFP_ATOMIC, order[v]);
++ if (!estacks)
++ panic("Cannot allocate exception stack %ld %d\n",
++ v, cpu);
++ }
++ switch (v + 1) {
++#if DEBUG_STKSZ > EXCEPTION_STKSZ
++ case DEBUG_STACK:
++ cpu_pda(cpu)->debugstack = (unsigned long)estacks;
++ estacks += DEBUG_STKSZ;
++ break;
++#endif
++ default:
++ estacks += EXCEPTION_STKSZ;
++ break;
++ }
++ orig_ist->ist[v] = t->ist[v] = (unsigned long)estacks;
++ }
++
++ t->io_bitmap_base = offsetof(struct tss_struct, io_bitmap);
++ /*
++ * <= is required because the CPU will access up to
++ * 8 bits beyond the end of the IO permission bitmap.
++ */
++ for (i = 0; i <= IO_BITMAP_LONGS; i++)
++ t->io_bitmap[i] = ~0UL;
++#endif
++
++ atomic_inc(&init_mm.mm_count);
++ me->active_mm = &init_mm;
++ if (me->mm)
++ BUG();
++ enter_lazy_tlb(&init_mm, me);
++
++#ifndef CONFIG_X86_NO_TSS
++ set_tss_desc(cpu, t);
++#endif
++#ifndef CONFIG_XEN
++ load_TR_desc();
++#endif
++ load_LDT(&init_mm.context);
++
++ /*
++ * Clear all 6 debug registers:
++ */
++
++ set_debugreg(0UL, 0);
++ set_debugreg(0UL, 1);
++ set_debugreg(0UL, 2);
++ set_debugreg(0UL, 3);
++ set_debugreg(0UL, 6);
++ set_debugreg(0UL, 7);
++
++ fpu_init();
++
++ raw_local_save_flags(kernel_eflags);
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/signal.c
+--- a/arch/x86_64/kernel/signal.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/kernel/signal.c Wed Sep 10 10:54:08 2008 +0100
+@@ -36,37 +36,6 @@
+ sigset_t *set, struct pt_regs * regs);
+ int ia32_setup_frame(int sig, struct k_sigaction *ka,
+ sigset_t *set, struct pt_regs * regs);
+-
+-asmlinkage long
+-sys_rt_sigsuspend(sigset_t __user *unewset, size_t sigsetsize, struct pt_regs *regs)
+-{
+- sigset_t saveset, newset;
+-
+- /* XXX: Don't preclude handling different sized sigset_t's. */
+- if (sigsetsize != sizeof(sigset_t))
+- return -EINVAL;
+-
+- if (copy_from_user(&newset, unewset, sizeof(newset)))
+- return -EFAULT;
+- sigdelsetmask(&newset, ~_BLOCKABLE);
+-
+- spin_lock_irq(¤t->sighand->siglock);
+- saveset = current->blocked;
+- current->blocked = newset;
+- recalc_sigpending();
+- spin_unlock_irq(¤t->sighand->siglock);
+-#ifdef DEBUG_SIG
+- printk("rt_sigsuspend savset(%lx) newset(%lx) regs(%p) rip(%lx)\n",
+- saveset, newset, regs, regs->rip);
+-#endif
+- regs->rax = -EINTR;
+- while (1) {
+- current->state = TASK_INTERRUPTIBLE;
+- schedule();
+- if (do_signal(regs, &saveset))
+- return -EINTR;
+- }
+-}
+
+ asmlinkage long
+ sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss,
+@@ -341,11 +310,11 @@
+ current->comm, current->pid, frame, regs->rip, frame->pretcode);
+ #endif
+
+- return 1;
++ return 0;
+
+ give_sigsegv:
+ force_sigsegv(sig, current);
+- return 0;
++ return -EFAULT;
+ }
+
+ /*
+@@ -408,7 +377,7 @@
+ #endif
+ ret = setup_rt_frame(sig, ka, info, oldset, regs);
+
+- if (ret) {
++ if (ret == 0) {
+ spin_lock_irq(¤t->sighand->siglock);
+ sigorsets(¤t->blocked,¤t->blocked,&ka->sa.sa_mask);
+ if (!(ka->sa.sa_flags & SA_NODEFER))
+@@ -425,11 +394,12 @@
+ * want to handle. Thus you cannot kill init even with a SIGKILL even by
+ * mistake.
+ */
+-int do_signal(struct pt_regs *regs, sigset_t *oldset)
++static void do_signal(struct pt_regs *regs)
+ {
+ struct k_sigaction ka;
+ siginfo_t info;
+ int signr;
++ sigset_t *oldset;
+
+ /*
+ * We want the common case to go fast, which
+@@ -438,9 +408,11 @@
+ * if so.
+ */
+ if (!user_mode(regs))
+- return 1;
++ return;
+
+- if (!oldset)
++ if (test_thread_flag(TIF_RESTORE_SIGMASK))
++ oldset = ¤t->saved_sigmask;
++ else
+ oldset = ¤t->blocked;
+
+ signr = get_signal_to_deliver(&info, &ka, regs, NULL);
+@@ -454,30 +426,46 @@
+ set_debugreg(current->thread.debugreg7, 7);
+
+ /* Whee! Actually deliver the signal. */
+- return handle_signal(signr, &info, &ka, oldset, regs);
++ if (handle_signal(signr, &info, &ka, oldset, regs) == 0) {
++ /* a signal was successfully delivered; the saved
++ * sigmask will have been stored in the signal frame,
++ * and will be restored by sigreturn, so we can simply
++ * clear the TIF_RESTORE_SIGMASK flag */
++ clear_thread_flag(TIF_RESTORE_SIGMASK);
++ }
++ return;
+ }
+
+ /* Did we come from a system call? */
+ if ((long)regs->orig_rax >= 0) {
+ /* Restart the system call - no handlers present */
+ long res = regs->rax;
+- if (res == -ERESTARTNOHAND ||
+- res == -ERESTARTSYS ||
+- res == -ERESTARTNOINTR) {
++ switch (res) {
++ case -ERESTARTNOHAND:
++ case -ERESTARTSYS:
++ case -ERESTARTNOINTR:
+ regs->rax = regs->orig_rax;
+ regs->rip -= 2;
+- }
+- if (regs->rax == (unsigned long)-ERESTART_RESTARTBLOCK) {
++ break;
++ case -ERESTART_RESTARTBLOCK:
+ regs->rax = test_thread_flag(TIF_IA32) ?
+ __NR_ia32_restart_syscall :
+ __NR_restart_syscall;
+ regs->rip -= 2;
++ break;
+ }
+ }
+- return 0;
++
++ /* if there's no signal to deliver, we just put the saved sigmask
++ back. */
++ if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
++ clear_thread_flag(TIF_RESTORE_SIGMASK);
++ sigprocmask(SIG_SETMASK, ¤t->saved_sigmask, NULL);
++ }
+ }
+
+-void do_notify_resume(struct pt_regs *regs, sigset_t *oldset, __u32 thread_info_flags)
++void
++do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags)
+ {
+ #ifdef DEBUG_SIG
+ printk("do_notify_resume flags:%x rip:%lx rsp:%lx caller:%lx pending:%lx\n",
+@@ -491,8 +479,8 @@
+ }
+
+ /* deal with pending signal delivery */
+- if (thread_info_flags & _TIF_SIGPENDING)
+- do_signal(regs,oldset);
++ if (thread_info_flags & (_TIF_SIGPENDING|_TIF_RESTORE_SIGMASK))
++ do_signal(regs);
+ }
+
+ void signal_fault(struct pt_regs *regs, void __user *frame, char *where)
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/smp-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/smp-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,575 @@
++/*
++ * Intel SMP support routines.
++ *
++ * (c) 1995 Alan Cox, Building #3 <alan at redhat.com>
++ * (c) 1998-99, 2000 Ingo Molnar <mingo at redhat.com>
++ * (c) 2002,2003 Andi Kleen, SuSE Labs.
++ *
++ * This code is released under the GNU General Public License version 2 or
++ * later.
++ */
++
++#include <linux/init.h>
++
++#include <linux/mm.h>
++#include <linux/delay.h>
++#include <linux/spinlock.h>
++#include <linux/smp_lock.h>
++#include <linux/smp.h>
++#include <linux/kernel_stat.h>
++#include <linux/mc146818rtc.h>
++#include <linux/interrupt.h>
++
++#include <asm/mtrr.h>
++#include <asm/pgalloc.h>
++#include <asm/tlbflush.h>
++#include <asm/mach_apic.h>
++#include <asm/mmu_context.h>
++#include <asm/proto.h>
++#include <asm/apicdef.h>
++#include <asm/idle.h>
++#ifdef CONFIG_XEN
++#include <xen/evtchn.h>
++#endif
++
++#ifndef CONFIG_XEN
++/*
++ * Smarter SMP flushing macros.
++ * c/o Linus Torvalds.
++ *
++ * These mean you can really definitely utterly forget about
++ * writing to user space from interrupts. (Its not allowed anyway).
++ *
++ * Optimizations Manfred Spraul <manfred at colorfullife.com>
++ *
++ * More scalable flush, from Andi Kleen
++ *
++ * To avoid global state use 8 different call vectors.
++ * Each CPU uses a specific vector to trigger flushes on other
++ * CPUs. Depending on the received vector the target CPUs look into
++ * the right per cpu variable for the flush data.
++ *
++ * With more than 8 CPUs they are hashed to the 8 available
++ * vectors. The limited global vector space forces us to this right now.
++ * In future when interrupts are split into per CPU domains this could be
++ * fixed, at the cost of triggering multiple IPIs in some cases.
++ */
++
++union smp_flush_state {
++ struct {
++ cpumask_t flush_cpumask;
++ struct mm_struct *flush_mm;
++ unsigned long flush_va;
++#define FLUSH_ALL -1ULL
++ spinlock_t tlbstate_lock;
++ };
++ char pad[SMP_CACHE_BYTES];
++} ____cacheline_aligned;
++
++/* State is put into the per CPU data section, but padded
++ to a full cache line because other CPUs can access it and we don't
++ want false sharing in the per cpu data segment. */
++static DEFINE_PER_CPU(union smp_flush_state, flush_state);
++
++/*
++ * We cannot call mmdrop() because we are in interrupt context,
++ * instead update mm->cpu_vm_mask.
++ */
++static inline void leave_mm(unsigned long cpu)
++{
++ if (read_pda(mmu_state) == TLBSTATE_OK)
++ BUG();
++ cpu_clear(cpu, read_pda(active_mm)->cpu_vm_mask);
++ load_cr3(swapper_pg_dir);
++}
++
++/*
++ *
++ * The flush IPI assumes that a thread switch happens in this order:
++ * [cpu0: the cpu that switches]
++ * 1) switch_mm() either 1a) or 1b)
++ * 1a) thread switch to a different mm
++ * 1a1) cpu_clear(cpu, old_mm->cpu_vm_mask);
++ * Stop ipi delivery for the old mm. This is not synchronized with
++ * the other cpus, but smp_invalidate_interrupt ignore flush ipis
++ * for the wrong mm, and in the worst case we perform a superfluous
++ * tlb flush.
++ * 1a2) set cpu mmu_state to TLBSTATE_OK
++ * Now the smp_invalidate_interrupt won't call leave_mm if cpu0
++ * was in lazy tlb mode.
++ * 1a3) update cpu active_mm
++ * Now cpu0 accepts tlb flushes for the new mm.
++ * 1a4) cpu_set(cpu, new_mm->cpu_vm_mask);
++ * Now the other cpus will send tlb flush ipis.
++ * 1a4) change cr3.
++ * 1b) thread switch without mm change
++ * cpu active_mm is correct, cpu0 already handles
++ * flush ipis.
++ * 1b1) set cpu mmu_state to TLBSTATE_OK
++ * 1b2) test_and_set the cpu bit in cpu_vm_mask.
++ * Atomically set the bit [other cpus will start sending flush ipis],
++ * and test the bit.
++ * 1b3) if the bit was 0: leave_mm was called, flush the tlb.
++ * 2) switch %%esp, ie current
++ *
++ * The interrupt must handle 2 special cases:
++ * - cr3 is changed before %%esp, ie. it cannot use current->{active_,}mm.
++ * - the cpu performs speculative tlb reads, i.e. even if the cpu only
++ * runs in kernel space, the cpu could load tlb entries for user space
++ * pages.
++ *
++ * The good news is that cpu mmu_state is local to each cpu, no
++ * write/read ordering problems.
++ */
++
++/*
++ * TLB flush IPI:
++ *
++ * 1) Flush the tlb entries if the cpu uses the mm that's being flushed.
++ * 2) Leave the mm if we are in the lazy tlb mode.
++ *
++ * Interrupts are disabled.
++ */
++
++asmlinkage void smp_invalidate_interrupt(struct pt_regs *regs)
++{
++ int cpu;
++ int sender;
++ union smp_flush_state *f;
++
++ cpu = smp_processor_id();
++ /*
++ * orig_rax contains the negated interrupt vector.
++ * Use that to determine where the sender put the data.
++ */
++ sender = ~regs->orig_rax - INVALIDATE_TLB_VECTOR_START;
++ f = &per_cpu(flush_state, sender);
++
++ if (!cpu_isset(cpu, f->flush_cpumask))
++ goto out;
++ /*
++ * This was a BUG() but until someone can quote me the
++ * line from the intel manual that guarantees an IPI to
++ * multiple CPUs is retried _only_ on the erroring CPUs
++ * its staying as a return
++ *
++ * BUG();
++ */
++
++ if (f->flush_mm == read_pda(active_mm)) {
++ if (read_pda(mmu_state) == TLBSTATE_OK) {
++ if (f->flush_va == FLUSH_ALL)
++ local_flush_tlb();
++ else
++ __flush_tlb_one(f->flush_va);
++ } else
++ leave_mm(cpu);
++ }
++out:
++ ack_APIC_irq();
++ cpu_clear(cpu, f->flush_cpumask);
++}
++
++static void flush_tlb_others(cpumask_t cpumask, struct mm_struct *mm,
++ unsigned long va)
++{
++ int sender;
++ union smp_flush_state *f;
++
++ /* Caller has disabled preemption */
++ sender = smp_processor_id() % NUM_INVALIDATE_TLB_VECTORS;
++ f = &per_cpu(flush_state, sender);
++
++ /* Could avoid this lock when
++ num_online_cpus() <= NUM_INVALIDATE_TLB_VECTORS, but it is
++ probably not worth checking this for a cache-hot lock. */
++ spin_lock(&f->tlbstate_lock);
++
++ f->flush_mm = mm;
++ f->flush_va = va;
++ cpus_or(f->flush_cpumask, cpumask, f->flush_cpumask);
++
++ /*
++ * We have to send the IPI only to
++ * CPUs affected.
++ */
++ send_IPI_mask(cpumask, INVALIDATE_TLB_VECTOR_START + sender);
++
++ while (!cpus_empty(f->flush_cpumask))
++ cpu_relax();
++
++ f->flush_mm = NULL;
++ f->flush_va = 0;
++ spin_unlock(&f->tlbstate_lock);
++}
++
++int __cpuinit init_smp_flush(void)
++{
++ int i;
++ for_each_cpu_mask(i, cpu_possible_map) {
++ spin_lock_init(&per_cpu(flush_state, i).tlbstate_lock);
++ }
++ return 0;
++}
++
++core_initcall(init_smp_flush);
++
++void flush_tlb_current_task(void)
++{
++ struct mm_struct *mm = current->mm;
++ cpumask_t cpu_mask;
++
++ preempt_disable();
++ cpu_mask = mm->cpu_vm_mask;
++ cpu_clear(smp_processor_id(), cpu_mask);
++
++ local_flush_tlb();
++ if (!cpus_empty(cpu_mask))
++ flush_tlb_others(cpu_mask, mm, FLUSH_ALL);
++ preempt_enable();
++}
++EXPORT_SYMBOL(flush_tlb_current_task);
++
++void flush_tlb_mm (struct mm_struct * mm)
++{
++ cpumask_t cpu_mask;
++
++ preempt_disable();
++ cpu_mask = mm->cpu_vm_mask;
++ cpu_clear(smp_processor_id(), cpu_mask);
++
++ if (current->active_mm == mm) {
++ if (current->mm)
++ local_flush_tlb();
++ else
++ leave_mm(smp_processor_id());
++ }
++ if (!cpus_empty(cpu_mask))
++ flush_tlb_others(cpu_mask, mm, FLUSH_ALL);
++
++ preempt_enable();
++}
++EXPORT_SYMBOL(flush_tlb_mm);
++
++void flush_tlb_page(struct vm_area_struct * vma, unsigned long va)
++{
++ struct mm_struct *mm = vma->vm_mm;
++ cpumask_t cpu_mask;
++
++ preempt_disable();
++ cpu_mask = mm->cpu_vm_mask;
++ cpu_clear(smp_processor_id(), cpu_mask);
++
++ if (current->active_mm == mm) {
++ if(current->mm)
++ __flush_tlb_one(va);
++ else
++ leave_mm(smp_processor_id());
++ }
++
++ if (!cpus_empty(cpu_mask))
++ flush_tlb_others(cpu_mask, mm, va);
++
++ preempt_enable();
++}
++EXPORT_SYMBOL(flush_tlb_page);
++
++static void do_flush_tlb_all(void* info)
++{
++ unsigned long cpu = smp_processor_id();
++
++ __flush_tlb_all();
++ if (read_pda(mmu_state) == TLBSTATE_LAZY)
++ leave_mm(cpu);
++}
++
++void flush_tlb_all(void)
++{
++ on_each_cpu(do_flush_tlb_all, NULL, 1, 1);
++}
++#endif /* Xen */
++
++/*
++ * this function sends a 'reschedule' IPI to another CPU.
++ * it goes straight through and wastes no time serializing
++ * anything. Worst case is that we lose a reschedule ...
++ */
++
++void smp_send_reschedule(int cpu)
++{
++ send_IPI_mask(cpumask_of_cpu(cpu), RESCHEDULE_VECTOR);
++}
++
++/*
++ * Structure and data for smp_call_function(). This is designed to minimise
++ * static memory requirements. It also looks cleaner.
++ */
++static DEFINE_SPINLOCK(call_lock);
++
++struct call_data_struct {
++ void (*func) (void *info);
++ void *info;
++ atomic_t started;
++ atomic_t finished;
++ int wait;
++};
++
++static struct call_data_struct * call_data;
++
++void lock_ipi_call_lock(void)
++{
++ spin_lock_irq(&call_lock);
++}
++
++void unlock_ipi_call_lock(void)
++{
++ spin_unlock_irq(&call_lock);
++}
++
++/*
++ * this function sends a 'generic call function' IPI to one other CPU
++ * in the system.
++ *
++ * cpu is a standard Linux logical CPU number.
++ */
++static void
++__smp_call_function_single(int cpu, void (*func) (void *info), void *info,
++ int nonatomic, int wait)
++{
++ struct call_data_struct data;
++ int cpus = 1;
++
++ data.func = func;
++ data.info = info;
++ atomic_set(&data.started, 0);
++ data.wait = wait;
++ if (wait)
++ atomic_set(&data.finished, 0);
++
++ call_data = &data;
++ wmb();
++ /* Send a message to all other CPUs and wait for them to respond */
++ send_IPI_mask(cpumask_of_cpu(cpu), CALL_FUNCTION_VECTOR);
++
++ /* Wait for response */
++ while (atomic_read(&data.started) != cpus)
++ cpu_relax();
++
++ if (!wait)
++ return;
++
++ while (atomic_read(&data.finished) != cpus)
++ cpu_relax();
++}
++
++/*
++ * smp_call_function_single - Run a function on another CPU
++ * @func: The function to run. This must be fast and non-blocking.
++ * @info: An arbitrary pointer to pass to the function.
++ * @nonatomic: Currently unused.
++ * @wait: If true, wait until function has completed on other CPUs.
++ *
++ * Retrurns 0 on success, else a negative status code.
++ *
++ * Does not return until the remote CPU is nearly ready to execute <func>
++ * or is or has executed.
++ */
++
++int smp_call_function_single (int cpu, void (*func) (void *info), void *info,
++ int nonatomic, int wait)
++{
++ /* prevent preemption and reschedule on another processor */
++ int me = get_cpu();
++ if (cpu == me) {
++ WARN_ON(1);
++ put_cpu();
++ return -EBUSY;
++ }
++ spin_lock_bh(&call_lock);
++ __smp_call_function_single(cpu, func, info, nonatomic, wait);
++ spin_unlock_bh(&call_lock);
++ put_cpu();
++ return 0;
++}
++
++/*
++ * this function sends a 'generic call function' IPI to all other CPUs
++ * in the system.
++ */
++static void __smp_call_function (void (*func) (void *info), void *info,
++ int nonatomic, int wait)
++{
++ struct call_data_struct data;
++ int cpus = num_online_cpus()-1;
++
++ if (!cpus)
++ return;
++
++ data.func = func;
++ data.info = info;
++ atomic_set(&data.started, 0);
++ data.wait = wait;
++ if (wait)
++ atomic_set(&data.finished, 0);
++
++ call_data = &data;
++ wmb();
++ /* Send a message to all other CPUs and wait for them to respond */
++ send_IPI_allbutself(CALL_FUNCTION_VECTOR);
++
++ /* Wait for response */
++ while (atomic_read(&data.started) != cpus)
++ cpu_relax();
++
++ if (!wait)
++ return;
++
++ while (atomic_read(&data.finished) != cpus)
++ cpu_relax();
++}
++
++/*
++ * smp_call_function - run a function on all other CPUs.
++ * @func: The function to run. This must be fast and non-blocking.
++ * @info: An arbitrary pointer to pass to the function.
++ * @nonatomic: currently unused.
++ * @wait: If true, wait (atomically) until function has completed on other
++ * CPUs.
++ *
++ * Returns 0 on success, else a negative status code. Does not return until
++ * remote CPUs are nearly ready to execute func or are or have executed.
++ *
++ * You must not call this function with disabled interrupts or from a
++ * hardware interrupt handler or from a bottom half handler.
++ * Actually there are a few legal cases, like panic.
++ */
++int smp_call_function (void (*func) (void *info), void *info, int nonatomic,
++ int wait)
++{
++ spin_lock(&call_lock);
++ __smp_call_function(func,info,nonatomic,wait);
++ spin_unlock(&call_lock);
++ return 0;
++}
++EXPORT_SYMBOL(smp_call_function);
++
++void smp_stop_cpu(void)
++{
++ unsigned long flags;
++ /*
++ * Remove this CPU:
++ */
++ cpu_clear(smp_processor_id(), cpu_online_map);
++ local_irq_save(flags);
++ disable_all_local_evtchn();
++ local_irq_restore(flags);
++}
++
++static void smp_really_stop_cpu(void *dummy)
++{
++ smp_stop_cpu();
++ for (;;)
++ halt();
++}
++
++void smp_send_stop(void)
++{
++ int nolock = 0;
++#ifndef CONFIG_XEN
++ if (reboot_force)
++ return;
++#endif
++ /* Don't deadlock on the call lock in panic */
++ if (!spin_trylock(&call_lock)) {
++ /* ignore locking because we have panicked anyways */
++ nolock = 1;
++ }
++ __smp_call_function(smp_really_stop_cpu, NULL, 0, 0);
++ if (!nolock)
++ spin_unlock(&call_lock);
++
++ local_irq_disable();
++ disable_all_local_evtchn();
++ local_irq_enable();
++}
++
++/*
++ * Reschedule call back. Nothing to do,
++ * all the work is done automatically when
++ * we return from the interrupt.
++ */
++#ifndef CONFIG_XEN
++asmlinkage void smp_reschedule_interrupt(void)
++#else
++asmlinkage irqreturn_t smp_reschedule_interrupt(void)
++#endif
++{
++#ifndef CONFIG_XEN
++ ack_APIC_irq();
++#else
++ return IRQ_HANDLED;
++#endif
++}
++
++#ifndef CONFIG_XEN
++asmlinkage void smp_call_function_interrupt(void)
++#else
++asmlinkage irqreturn_t smp_call_function_interrupt(void)
++#endif
++{
++ void (*func) (void *info) = call_data->func;
++ void *info = call_data->info;
++ int wait = call_data->wait;
++
++#ifndef CONFIG_XEN
++ ack_APIC_irq();
++#endif
++ /*
++ * Notify initiating CPU that I've grabbed the data and am
++ * about to execute the function
++ */
++ mb();
++ atomic_inc(&call_data->started);
++ /*
++ * At this point the info structure may be out of scope unless wait==1
++ */
++ exit_idle();
++ irq_enter();
++ (*func)(info);
++ irq_exit();
++ if (wait) {
++ mb();
++ atomic_inc(&call_data->finished);
++ }
++#ifdef CONFIG_XEN
++ return IRQ_HANDLED;
++#endif
++}
++
++int safe_smp_processor_id(void)
++{
++#ifdef CONFIG_XEN
++ return smp_processor_id();
++#else
++ unsigned apicid, i;
++
++ if (disable_apic)
++ return 0;
++
++ apicid = hard_smp_processor_id();
++ if (apicid < NR_CPUS && x86_cpu_to_apicid[apicid] == apicid)
++ return apicid;
++
++ for (i = 0; i < NR_CPUS; ++i) {
++ if (x86_cpu_to_apicid[i] == apicid)
++ return i;
++ }
++
++ /* No entries in x86_cpu_to_apicid? Either no MPS|ACPI,
++ * or called too early. Either way, we must be CPU 0. */
++ if (x86_cpu_to_apicid[0] == BAD_APICID)
++ return 0;
++
++ return 0; /* Should not happen */
++#endif
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/suspend.c
+--- a/arch/x86_64/kernel/suspend.c Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/kernel/suspend.c Wed Sep 10 10:54:08 2008 +0100
+@@ -114,12 +114,14 @@
+
+ void fix_processor_context(void)
+ {
++#ifndef CONFIG_X86_NO_TSS
+ int cpu = smp_processor_id();
+ struct tss_struct *t = &per_cpu(init_tss, cpu);
+
+ set_tss_desc(cpu,t); /* This just modifies memory; should not be neccessary. But... This is neccessary, because 386 hardware has concept of busy TSS or some similar stupidity. */
+
+ cpu_gdt(cpu)[GDT_ENTRY_TSS].type = 9;
++#endif
+
+ syscall_init(); /* This sets MSR_*STAR and related */
+ load_TR_desc(); /* This does ltr */
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/traps-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/traps-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1173 @@
++/*
++ * linux/arch/x86-64/traps.c
++ *
++ * Copyright (C) 1991, 1992 Linus Torvalds
++ * Copyright (C) 2000, 2001, 2002 Andi Kleen, SuSE Labs
++ *
++ * Pentium III FXSR, SSE support
++ * Gareth Hughes <gareth at valinux.com>, May 2000
++ */
++
++/*
++ * 'Traps.c' handles hardware traps and faults after we have saved some
++ * state in 'entry.S'.
++ */
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/string.h>
++#include <linux/errno.h>
++#include <linux/ptrace.h>
++#include <linux/timer.h>
++#include <linux/mm.h>
++#include <linux/init.h>
++#include <linux/delay.h>
++#include <linux/spinlock.h>
++#include <linux/interrupt.h>
++#include <linux/module.h>
++#include <linux/moduleparam.h>
++#include <linux/nmi.h>
++#include <linux/kprobes.h>
++#include <linux/kexec.h>
++#include <linux/unwind.h>
++
++#include <asm/system.h>
++#include <asm/uaccess.h>
++#include <asm/io.h>
++#include <asm/atomic.h>
++#include <asm/debugreg.h>
++#include <asm/desc.h>
++#include <asm/i387.h>
++#include <asm/kdebug.h>
++#include <asm/processor.h>
++#include <asm/unwind.h>
++#include <asm/smp.h>
++#include <asm/pgalloc.h>
++#include <asm/pda.h>
++#include <asm/proto.h>
++#include <asm/nmi.h>
++
++asmlinkage void divide_error(void);
++asmlinkage void debug(void);
++asmlinkage void nmi(void);
++asmlinkage void int3(void);
++asmlinkage void overflow(void);
++asmlinkage void bounds(void);
++asmlinkage void invalid_op(void);
++asmlinkage void device_not_available(void);
++asmlinkage void double_fault(void);
++asmlinkage void coprocessor_segment_overrun(void);
++asmlinkage void invalid_TSS(void);
++asmlinkage void segment_not_present(void);
++asmlinkage void stack_segment(void);
++asmlinkage void general_protection(void);
++asmlinkage void page_fault(void);
++asmlinkage void coprocessor_error(void);
++asmlinkage void simd_coprocessor_error(void);
++asmlinkage void reserved(void);
++asmlinkage void alignment_check(void);
++asmlinkage void machine_check(void);
++asmlinkage void spurious_interrupt_bug(void);
++
++ATOMIC_NOTIFIER_HEAD(die_chain);
++EXPORT_SYMBOL(die_chain);
++
++int register_die_notifier(struct notifier_block *nb)
++{
++ vmalloc_sync_all();
++ return atomic_notifier_chain_register(&die_chain, nb);
++}
++EXPORT_SYMBOL(register_die_notifier); /* used modular by kdb */
++
++int unregister_die_notifier(struct notifier_block *nb)
++{
++ return atomic_notifier_chain_unregister(&die_chain, nb);
++}
++EXPORT_SYMBOL(unregister_die_notifier); /* used modular by kdb */
++
++static inline void conditional_sti(struct pt_regs *regs)
++{
++ if (regs->eflags & X86_EFLAGS_IF)
++ local_irq_enable();
++}
++
++static inline void preempt_conditional_sti(struct pt_regs *regs)
++{
++ preempt_disable();
++ if (regs->eflags & X86_EFLAGS_IF)
++ local_irq_enable();
++}
++
++static inline void preempt_conditional_cli(struct pt_regs *regs)
++{
++ if (regs->eflags & X86_EFLAGS_IF)
++ local_irq_disable();
++ /* Make sure to not schedule here because we could be running
++ on an exception stack. */
++ preempt_enable_no_resched();
++}
++
++static int kstack_depth_to_print = 12;
++#ifdef CONFIG_STACK_UNWIND
++static int call_trace = 1;
++#else
++#define call_trace (-1)
++#endif
++
++#ifdef CONFIG_KALLSYMS
++# include <linux/kallsyms.h>
++void printk_address(unsigned long address)
++{
++ unsigned long offset = 0, symsize;
++ const char *symname;
++ char *modname;
++ char *delim = ":";
++ char namebuf[128];
++
++ symname = kallsyms_lookup(address, &symsize, &offset,
++ &modname, namebuf);
++ if (!symname) {
++ printk(" [<%016lx>]\n", address);
++ return;
++ }
++ if (!modname)
++ modname = delim = "";
++ printk(" [<%016lx>] %s%s%s%s+0x%lx/0x%lx\n",
++ address, delim, modname, delim, symname, offset, symsize);
++}
++#else
++void printk_address(unsigned long address)
++{
++ printk(" [<%016lx>]\n", address);
++}
++#endif
++
++static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
++ unsigned *usedp, const char **idp)
++{
++#ifndef CONFIG_X86_NO_TSS
++ static char ids[][8] = {
++ [DEBUG_STACK - 1] = "#DB",
++ [NMI_STACK - 1] = "NMI",
++ [DOUBLEFAULT_STACK - 1] = "#DF",
++ [STACKFAULT_STACK - 1] = "#SS",
++ [MCE_STACK - 1] = "#MC",
++#if DEBUG_STKSZ > EXCEPTION_STKSZ
++ [N_EXCEPTION_STACKS ... N_EXCEPTION_STACKS + DEBUG_STKSZ / EXCEPTION_STKSZ - 2] = "#DB[?]"
++#endif
++ };
++ unsigned k;
++
++ /*
++ * Iterate over all exception stacks, and figure out whether
++ * 'stack' is in one of them:
++ */
++ for (k = 0; k < N_EXCEPTION_STACKS; k++) {
++ unsigned long end;
++
++ /*
++ * set 'end' to the end of the exception stack.
++ */
++ switch (k + 1) {
++ /*
++ * TODO: this block is not needed i think, because
++ * setup64.c:cpu_init() sets up t->ist[DEBUG_STACK]
++ * properly too.
++ */
++#if DEBUG_STKSZ > EXCEPTION_STKSZ
++ case DEBUG_STACK:
++ end = cpu_pda(cpu)->debugstack + DEBUG_STKSZ;
++ break;
++#endif
++ default:
++ end = per_cpu(orig_ist, cpu).ist[k];
++ break;
++ }
++ /*
++ * Is 'stack' above this exception frame's end?
++ * If yes then skip to the next frame.
++ */
++ if (stack >= end)
++ continue;
++ /*
++ * Is 'stack' above this exception frame's start address?
++ * If yes then we found the right frame.
++ */
++ if (stack >= end - EXCEPTION_STKSZ) {
++ /*
++ * Make sure we only iterate through an exception
++ * stack once. If it comes up for the second time
++ * then there's something wrong going on - just
++ * break out and return NULL:
++ */
++ if (*usedp & (1U << k))
++ break;
++ *usedp |= 1U << k;
++ *idp = ids[k];
++ return (unsigned long *)end;
++ }
++ /*
++ * If this is a debug stack, and if it has a larger size than
++ * the usual exception stacks, then 'stack' might still
++ * be within the lower portion of the debug stack:
++ */
++#if DEBUG_STKSZ > EXCEPTION_STKSZ
++ if (k == DEBUG_STACK - 1 && stack >= end - DEBUG_STKSZ) {
++ unsigned j = N_EXCEPTION_STACKS - 1;
++
++ /*
++ * Black magic. A large debug stack is composed of
++ * multiple exception stack entries, which we
++ * iterate through now. Dont look:
++ */
++ do {
++ ++j;
++ end -= EXCEPTION_STKSZ;
++ ids[j][4] = '1' + (j - N_EXCEPTION_STACKS);
++ } while (stack < end - EXCEPTION_STKSZ);
++ if (*usedp & (1U << j))
++ break;
++ *usedp |= 1U << j;
++ *idp = ids[j];
++ return (unsigned long *)end;
++ }
++#endif
++ }
++#endif
++ return NULL;
++}
++
++static int show_trace_unwind(struct unwind_frame_info *info, void *context)
++{
++ int n = 0;
++
++ while (unwind(info) == 0 && UNW_PC(info)) {
++ n++;
++ printk_address(UNW_PC(info));
++ if (arch_unw_user_mode(info))
++ break;
++ }
++ return n;
++}
++
++/*
++ * x86-64 can have upto three kernel stacks:
++ * process stack
++ * interrupt stack
++ * severe exception (double fault, nmi, stack fault, debug, mce) hardware stack
++ */
++
++void show_trace(struct task_struct *tsk, struct pt_regs *regs, unsigned long * stack)
++{
++ const unsigned cpu = safe_smp_processor_id();
++ unsigned long *irqstack_end = (unsigned long *)cpu_pda(cpu)->irqstackptr;
++ unsigned used = 0;
++
++ printk("\nCall Trace:\n");
++
++ if (!tsk)
++ tsk = current;
++
++ if (call_trace >= 0) {
++ int unw_ret = 0;
++ struct unwind_frame_info info;
++
++ if (regs) {
++ if (unwind_init_frame_info(&info, tsk, regs) == 0)
++ unw_ret = show_trace_unwind(&info, NULL);
++ } else if (tsk == current)
++ unw_ret = unwind_init_running(&info, show_trace_unwind, NULL);
++ else {
++ if (unwind_init_blocked(&info, tsk) == 0)
++ unw_ret = show_trace_unwind(&info, NULL);
++ }
++ if (unw_ret > 0) {
++ if (call_trace == 1 && !arch_unw_user_mode(&info)) {
++ print_symbol("DWARF2 unwinder stuck at %s\n",
++ UNW_PC(&info));
++ if ((long)UNW_SP(&info) < 0) {
++ printk("Leftover inexact backtrace:\n");
++ stack = (unsigned long *)UNW_SP(&info);
++ } else
++ printk("Full inexact backtrace again:\n");
++ } else if (call_trace >= 1)
++ return;
++ else
++ printk("Full inexact backtrace again:\n");
++ } else
++ printk("Inexact backtrace:\n");
++ }
++
++ /*
++ * Print function call entries within a stack. 'cond' is the
++ * "end of stackframe" condition, that the 'stack++'
++ * iteration will eventually trigger.
++ */
++#define HANDLE_STACK(cond) \
++ do while (cond) { \
++ unsigned long addr = *stack++; \
++ if (kernel_text_address(addr)) { \
++ /* \
++ * If the address is either in the text segment of the \
++ * kernel, or in the region which contains vmalloc'ed \
++ * memory, it *may* be the address of a calling \
++ * routine; if so, print it so that someone tracing \
++ * down the cause of the crash will be able to figure \
++ * out the call path that was taken. \
++ */ \
++ printk_address(addr); \
++ } \
++ } while (0)
++
++ /*
++ * Print function call entries in all stacks, starting at the
++ * current stack address. If the stacks consist of nested
++ * exceptions
++ */
++ for ( ; ; ) {
++ const char *id;
++ unsigned long *estack_end;
++ estack_end = in_exception_stack(cpu, (unsigned long)stack,
++ &used, &id);
++
++ if (estack_end) {
++ printk(" <%s>", id);
++ HANDLE_STACK (stack < estack_end);
++ printk(" <EOE>");
++ /*
++ * We link to the next stack via the
++ * second-to-last pointer (index -2 to end) in the
++ * exception stack:
++ */
++ stack = (unsigned long *) estack_end[-2];
++ continue;
++ }
++ if (irqstack_end) {
++ unsigned long *irqstack;
++ irqstack = irqstack_end -
++ (IRQSTACKSIZE - 64) / sizeof(*irqstack);
++
++ if (stack >= irqstack && stack < irqstack_end) {
++ printk(" <IRQ>");
++ HANDLE_STACK (stack < irqstack_end);
++ /*
++ * We link to the next stack (which would be
++ * the process stack normally) the last
++ * pointer (index -1 to end) in the IRQ stack:
++ */
++ stack = (unsigned long *) (irqstack_end[-1]);
++ irqstack_end = NULL;
++ printk(" <EOI>");
++ continue;
++ }
++ }
++ break;
++ }
++
++ /*
++ * This prints the process stack:
++ */
++ HANDLE_STACK (((long) stack & (THREAD_SIZE-1)) != 0);
++#undef HANDLE_STACK
++
++ printk("\n");
++}
++
++static void _show_stack(struct task_struct *tsk, struct pt_regs *regs, unsigned long * rsp)
++{
++ unsigned long *stack;
++ int i;
++ const int cpu = safe_smp_processor_id();
++ unsigned long *irqstack_end = (unsigned long *) (cpu_pda(cpu)->irqstackptr);
++ unsigned long *irqstack = (unsigned long *) (cpu_pda(cpu)->irqstackptr - IRQSTACKSIZE);
++
++ // debugging aid: "show_stack(NULL, NULL);" prints the
++ // back trace for this cpu.
++
++ if (rsp == NULL) {
++ if (tsk)
++ rsp = (unsigned long *)tsk->thread.rsp;
++ else
++ rsp = (unsigned long *)&rsp;
++ }
++
++ stack = rsp;
++ for(i=0; i < kstack_depth_to_print; i++) {
++ if (stack >= irqstack && stack <= irqstack_end) {
++ if (stack == irqstack_end) {
++ stack = (unsigned long *) (irqstack_end[-1]);
++ printk(" <EOI> ");
++ }
++ } else {
++ if (((long) stack & (THREAD_SIZE-1)) == 0)
++ break;
++ }
++ if (i && ((i % 4) == 0))
++ printk("\n");
++ printk(" %016lx", *stack++);
++ touch_nmi_watchdog();
++ }
++ show_trace(tsk, regs, rsp);
++}
++
++void show_stack(struct task_struct *tsk, unsigned long * rsp)
++{
++ _show_stack(tsk, NULL, rsp);
++}
++
++/*
++ * The architecture-independent dump_stack generator
++ */
++void dump_stack(void)
++{
++ unsigned long dummy;
++ show_trace(NULL, NULL, &dummy);
++}
++
++EXPORT_SYMBOL(dump_stack);
++
++void show_registers(struct pt_regs *regs)
++{
++ int i;
++ int in_kernel = !user_mode(regs);
++ unsigned long rsp;
++ const int cpu = safe_smp_processor_id();
++ struct task_struct *cur = cpu_pda(cpu)->pcurrent;
++
++ rsp = regs->rsp;
++
++ printk("CPU %d ", cpu);
++ __show_regs(regs);
++ printk("Process %s (pid: %d, threadinfo %p, task %p)\n",
++ cur->comm, cur->pid, task_thread_info(cur), cur);
++
++ /*
++ * When in-kernel, we also print out the stack and code at the
++ * time of the fault..
++ */
++ if (in_kernel) {
++
++ printk("Stack: ");
++ _show_stack(NULL, regs, (unsigned long*)rsp);
++
++ printk("\nCode: ");
++ if (regs->rip < PAGE_OFFSET)
++ goto bad;
++
++ for (i=0; i<20; i++) {
++ unsigned char c;
++ if (__get_user(c, &((unsigned char*)regs->rip)[i])) {
++bad:
++ printk(" Bad RIP value.");
++ break;
++ }
++ printk("%02x ", c);
++ }
++ }
++ printk("\n");
++}
++
++void handle_BUG(struct pt_regs *regs)
++{
++ struct bug_frame f;
++ long len;
++ const char *prefix = "";
++
++ if (user_mode(regs))
++ return;
++ if (__copy_from_user(&f, (const void __user *) regs->rip,
++ sizeof(struct bug_frame)))
++ return;
++ if (f.filename >= 0 ||
++ f.ud2[0] != 0x0f || f.ud2[1] != 0x0b)
++ return;
++ len = __strnlen_user((char *)(long)f.filename, PATH_MAX) - 1;
++ if (len < 0 || len >= PATH_MAX)
++ f.filename = (int)(long)"unmapped filename";
++ else if (len > 50) {
++ f.filename += len - 50;
++ prefix = "...";
++ }
++ printk("----------- [cut here ] --------- [please bite here ] ---------\n");
++ printk(KERN_ALERT "Kernel BUG at %s%.50s:%d\n", prefix, (char *)(long)f.filename, f.line);
++}
++
++#ifdef CONFIG_BUG
++void out_of_line_bug(void)
++{
++ BUG();
++}
++EXPORT_SYMBOL(out_of_line_bug);
++#endif
++
++static DEFINE_SPINLOCK(die_lock);
++static int die_owner = -1;
++static unsigned int die_nest_count;
++
++unsigned __kprobes long oops_begin(void)
++{
++ int cpu = safe_smp_processor_id();
++ unsigned long flags;
++
++ /* racy, but better than risking deadlock. */
++ local_irq_save(flags);
++ if (!spin_trylock(&die_lock)) {
++ if (cpu == die_owner)
++ /* nested oops. should stop eventually */;
++ else
++ spin_lock(&die_lock);
++ }
++ die_nest_count++;
++ die_owner = cpu;
++ console_verbose();
++ bust_spinlocks(1);
++ return flags;
++}
++
++void __kprobes oops_end(unsigned long flags)
++{
++ die_owner = -1;
++ bust_spinlocks(0);
++ die_nest_count--;
++ if (die_nest_count)
++ /* We still own the lock */
++ local_irq_restore(flags);
++ else
++ /* Nest count reaches zero, release the lock. */
++ spin_unlock_irqrestore(&die_lock, flags);
++ if (panic_on_oops)
++ panic("Fatal exception");
++}
++
++void __kprobes __die(const char * str, struct pt_regs * regs, long err)
++{
++ static int die_counter;
++ printk(KERN_EMERG "%s: %04lx [%u] ", str, err & 0xffff,++die_counter);
++#ifdef CONFIG_PREEMPT
++ printk("PREEMPT ");
++#endif
++#ifdef CONFIG_SMP
++ printk("SMP ");
++#endif
++#ifdef CONFIG_DEBUG_PAGEALLOC
++ printk("DEBUG_PAGEALLOC");
++#endif
++ printk("\n");
++ notify_die(DIE_OOPS, str, regs, err, current->thread.trap_no, SIGSEGV);
++ show_registers(regs);
++ /* Executive summary in case the oops scrolled away */
++ printk(KERN_ALERT "RIP ");
++ printk_address(regs->rip);
++ printk(" RSP <%016lx>\n", regs->rsp);
++ if (kexec_should_crash(current))
++ crash_kexec(regs);
++}
++
++void die(const char * str, struct pt_regs * regs, long err)
++{
++ unsigned long flags = oops_begin();
++
++ handle_BUG(regs);
++ __die(str, regs, err);
++ oops_end(flags);
++ do_exit(SIGSEGV);
++}
++
++#ifdef CONFIG_X86_LOCAL_APIC
++void __kprobes die_nmi(char *str, struct pt_regs *regs)
++{
++ unsigned long flags = oops_begin();
++
++ /*
++ * We are in trouble anyway, lets at least try
++ * to get a message out.
++ */
++ printk(str, safe_smp_processor_id());
++ show_registers(regs);
++ if (kexec_should_crash(current))
++ crash_kexec(regs);
++ if (panic_on_timeout || panic_on_oops)
++ panic("nmi watchdog");
++ printk("console shuts up ...\n");
++ oops_end(flags);
++ nmi_exit();
++ local_irq_enable();
++ do_exit(SIGSEGV);
++}
++#endif
++
++static void __kprobes do_trap(int trapnr, int signr, char *str,
++ struct pt_regs * regs, long error_code,
++ siginfo_t *info)
++{
++ struct task_struct *tsk = current;
++
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = trapnr;
++
++ if (user_mode(regs)) {
++ if (exception_trace && unhandled_signal(tsk, signr))
++ printk(KERN_INFO
++ "%s[%d] trap %s rip:%lx rsp:%lx error:%lx\n",
++ tsk->comm, tsk->pid, str,
++ regs->rip, regs->rsp, error_code);
++
++ if (info)
++ force_sig_info(signr, info, tsk);
++ else
++ force_sig(signr, tsk);
++ return;
++ }
++
++
++ /* kernel trap */
++ {
++ const struct exception_table_entry *fixup;
++ fixup = search_exception_tables(regs->rip);
++ if (fixup)
++ regs->rip = fixup->fixup;
++ else
++ die(str, regs, error_code);
++ return;
++ }
++}
++
++#define DO_ERROR(trapnr, signr, str, name) \
++asmlinkage void do_##name(struct pt_regs * regs, long error_code) \
++{ \
++ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
++ == NOTIFY_STOP) \
++ return; \
++ conditional_sti(regs); \
++ do_trap(trapnr, signr, str, regs, error_code, NULL); \
++}
++
++#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \
++asmlinkage void do_##name(struct pt_regs * regs, long error_code) \
++{ \
++ siginfo_t info; \
++ info.si_signo = signr; \
++ info.si_errno = 0; \
++ info.si_code = sicode; \
++ info.si_addr = (void __user *)siaddr; \
++ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
++ == NOTIFY_STOP) \
++ return; \
++ conditional_sti(regs); \
++ do_trap(trapnr, signr, str, regs, error_code, &info); \
++}
++
++DO_ERROR_INFO( 0, SIGFPE, "divide error", divide_error, FPE_INTDIV, regs->rip)
++DO_ERROR( 4, SIGSEGV, "overflow", overflow)
++DO_ERROR( 5, SIGSEGV, "bounds", bounds)
++DO_ERROR_INFO( 6, SIGILL, "invalid opcode", invalid_op, ILL_ILLOPN, regs->rip)
++DO_ERROR( 7, SIGSEGV, "device not available", device_not_available)
++DO_ERROR( 9, SIGFPE, "coprocessor segment overrun", coprocessor_segment_overrun)
++DO_ERROR(10, SIGSEGV, "invalid TSS", invalid_TSS)
++DO_ERROR(11, SIGBUS, "segment not present", segment_not_present)
++DO_ERROR_INFO(17, SIGBUS, "alignment check", alignment_check, BUS_ADRALN, 0)
++DO_ERROR(18, SIGSEGV, "reserved", reserved)
++
++/* Runs on IST stack */
++asmlinkage void do_stack_segment(struct pt_regs *regs, long error_code)
++{
++ if (notify_die(DIE_TRAP, "stack segment", regs, error_code,
++ 12, SIGBUS) == NOTIFY_STOP)
++ return;
++ preempt_conditional_sti(regs);
++ do_trap(12, SIGBUS, "stack segment", regs, error_code, NULL);
++ preempt_conditional_cli(regs);
++}
++
++asmlinkage void do_double_fault(struct pt_regs * regs, long error_code)
++{
++ static const char str[] = "double fault";
++ struct task_struct *tsk = current;
++
++ /* Return not checked because double check cannot be ignored */
++ notify_die(DIE_TRAP, str, regs, error_code, 8, SIGSEGV);
++
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = 8;
++
++ /* This is always a kernel trap and never fixable (and thus must
++ never return). */
++ for (;;)
++ die(str, regs, error_code);
++}
++
++asmlinkage void __kprobes do_general_protection(struct pt_regs * regs,
++ long error_code)
++{
++ struct task_struct *tsk = current;
++
++ conditional_sti(regs);
++
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = 13;
++
++ if (user_mode(regs)) {
++ if (exception_trace && unhandled_signal(tsk, SIGSEGV))
++ printk(KERN_INFO
++ "%s[%d] general protection rip:%lx rsp:%lx error:%lx\n",
++ tsk->comm, tsk->pid,
++ regs->rip, regs->rsp, error_code);
++
++ force_sig(SIGSEGV, tsk);
++ return;
++ }
++
++ /* kernel gp */
++ {
++ const struct exception_table_entry *fixup;
++ fixup = search_exception_tables(regs->rip);
++ if (fixup) {
++ regs->rip = fixup->fixup;
++ return;
++ }
++ if (notify_die(DIE_GPF, "general protection fault", regs,
++ error_code, 13, SIGSEGV) == NOTIFY_STOP)
++ return;
++ die("general protection fault", regs, error_code);
++ }
++}
++
++static __kprobes void
++mem_parity_error(unsigned char reason, struct pt_regs * regs)
++{
++ printk("Uhhuh. NMI received. Dazed and confused, but trying to continue\n");
++ printk("You probably have a hardware problem with your RAM chips\n");
++
++#if 0 /* XEN */
++ /* Clear and disable the memory parity error line. */
++ reason = (reason & 0xf) | 4;
++ outb(reason, 0x61);
++#endif /* XEN */
++}
++
++static __kprobes void
++io_check_error(unsigned char reason, struct pt_regs * regs)
++{
++ printk("NMI: IOCK error (debug interrupt?)\n");
++ show_registers(regs);
++
++#if 0 /* XEN */
++ /* Re-enable the IOCK line, wait for a few seconds */
++ reason = (reason & 0xf) | 8;
++ outb(reason, 0x61);
++ mdelay(2000);
++ reason &= ~8;
++ outb(reason, 0x61);
++#endif /* XEN */
++}
++
++static __kprobes void
++unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
++{ printk("Uhhuh. NMI received for unknown reason %02x.\n", reason);
++ printk("Dazed and confused, but trying to continue\n");
++ printk("Do you have a strange power saving mode enabled?\n");
++}
++
++/* Runs on IST stack. This code must keep interrupts off all the time.
++ Nested NMIs are prevented by the CPU. */
++asmlinkage __kprobes void default_do_nmi(struct pt_regs *regs)
++{
++ unsigned char reason = 0;
++ int cpu;
++
++ cpu = smp_processor_id();
++
++ /* Only the BSP gets external NMIs from the system. */
++ if (!cpu)
++ reason = get_nmi_reason();
++
++ if (!(reason & 0xc0)) {
++ if (notify_die(DIE_NMI_IPI, "nmi_ipi", regs, reason, 2, SIGINT)
++ == NOTIFY_STOP)
++ return;
++#ifdef CONFIG_X86_LOCAL_APIC
++ /*
++ * Ok, so this is none of the documented NMI sources,
++ * so it must be the NMI watchdog.
++ */
++ if (nmi_watchdog > 0) {
++ nmi_watchdog_tick(regs,reason);
++ return;
++ }
++#endif
++ unknown_nmi_error(reason, regs);
++ return;
++ }
++ if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP)
++ return;
++
++ /* AK: following checks seem to be broken on modern chipsets. FIXME */
++
++ if (reason & 0x80)
++ mem_parity_error(reason, regs);
++ if (reason & 0x40)
++ io_check_error(reason, regs);
++}
++
++/* runs on IST stack. */
++asmlinkage void __kprobes do_int3(struct pt_regs * regs, long error_code)
++{
++ if (notify_die(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP) == NOTIFY_STOP) {
++ return;
++ }
++ preempt_conditional_sti(regs);
++ do_trap(3, SIGTRAP, "int3", regs, error_code, NULL);
++ preempt_conditional_cli(regs);
++}
++
++/* Help handler running on IST stack to switch back to user stack
++ for scheduling or signal handling. The actual stack switch is done in
++ entry.S */
++asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs)
++{
++ struct pt_regs *regs = eregs;
++ /* Did already sync */
++ if (eregs == (struct pt_regs *)eregs->rsp)
++ ;
++ /* Exception from user space */
++ else if (user_mode(eregs))
++ regs = task_pt_regs(current);
++ /* Exception from kernel and interrupts are enabled. Move to
++ kernel process stack. */
++ else if (eregs->eflags & X86_EFLAGS_IF)
++ regs = (struct pt_regs *)(eregs->rsp -= sizeof(struct pt_regs));
++ if (eregs != regs)
++ *regs = *eregs;
++ return regs;
++}
++
++/* runs on IST stack. */
++asmlinkage void __kprobes do_debug(struct pt_regs * regs,
++ unsigned long error_code)
++{
++ unsigned long condition;
++ struct task_struct *tsk = current;
++ siginfo_t info;
++
++ get_debugreg(condition, 6);
++
++ if (notify_die(DIE_DEBUG, "debug", regs, condition, error_code,
++ SIGTRAP) == NOTIFY_STOP)
++ return;
++
++ preempt_conditional_sti(regs);
++
++ /* Mask out spurious debug traps due to lazy DR7 setting */
++ if (condition & (DR_TRAP0|DR_TRAP1|DR_TRAP2|DR_TRAP3)) {
++ if (!tsk->thread.debugreg7) {
++ goto clear_dr7;
++ }
++ }
++
++ tsk->thread.debugreg6 = condition;
++
++ /* Mask out spurious TF errors due to lazy TF clearing */
++ if (condition & DR_STEP) {
++ /*
++ * The TF error should be masked out only if the current
++ * process is not traced and if the TRAP flag has been set
++ * previously by a tracing process (condition detected by
++ * the PT_DTRACE flag); remember that the i386 TRAP flag
++ * can be modified by the process itself in user mode,
++ * allowing programs to debug themselves without the ptrace()
++ * interface.
++ */
++ if (!user_mode(regs))
++ goto clear_TF_reenable;
++ /*
++ * Was the TF flag set by a debugger? If so, clear it now,
++ * so that register information is correct.
++ */
++ if (tsk->ptrace & PT_DTRACE) {
++ regs->eflags &= ~TF_MASK;
++ tsk->ptrace &= ~PT_DTRACE;
++ }
++ }
++
++ /* Ok, finally something we can handle */
++ tsk->thread.trap_no = 1;
++ tsk->thread.error_code = error_code;
++ info.si_signo = SIGTRAP;
++ info.si_errno = 0;
++ info.si_code = TRAP_BRKPT;
++ info.si_addr = user_mode(regs) ? (void __user *)regs->rip : NULL;
++ force_sig_info(SIGTRAP, &info, tsk);
++
++clear_dr7:
++ set_debugreg(0UL, 7);
++ preempt_conditional_cli(regs);
++ return;
++
++clear_TF_reenable:
++ set_tsk_thread_flag(tsk, TIF_SINGLESTEP);
++ regs->eflags &= ~TF_MASK;
++ preempt_conditional_cli(regs);
++}
++
++static int kernel_math_error(struct pt_regs *regs, const char *str, int trapnr)
++{
++ const struct exception_table_entry *fixup;
++ fixup = search_exception_tables(regs->rip);
++ if (fixup) {
++ regs->rip = fixup->fixup;
++ return 1;
++ }
++ notify_die(DIE_GPF, str, regs, 0, trapnr, SIGFPE);
++ /* Illegal floating point operation in the kernel */
++ current->thread.trap_no = trapnr;
++ die(str, regs, 0);
++ return 0;
++}
++
++/*
++ * Note that we play around with the 'TS' bit in an attempt to get
++ * the correct behaviour even in the presence of the asynchronous
++ * IRQ13 behaviour
++ */
++asmlinkage void do_coprocessor_error(struct pt_regs *regs)
++{
++ void __user *rip = (void __user *)(regs->rip);
++ struct task_struct * task;
++ siginfo_t info;
++ unsigned short cwd, swd;
++
++ conditional_sti(regs);
++ if (!user_mode(regs) &&
++ kernel_math_error(regs, "kernel x87 math error", 16))
++ return;
++
++ /*
++ * Save the info for the exception handler and clear the error.
++ */
++ task = current;
++ save_init_fpu(task);
++ task->thread.trap_no = 16;
++ task->thread.error_code = 0;
++ info.si_signo = SIGFPE;
++ info.si_errno = 0;
++ info.si_code = __SI_FAULT;
++ info.si_addr = rip;
++ /*
++ * (~cwd & swd) will mask out exceptions that are not set to unmasked
++ * status. 0x3f is the exception bits in these regs, 0x200 is the
++ * C1 reg you need in case of a stack fault, 0x040 is the stack
++ * fault bit. We should only be taking one exception at a time,
++ * so if this combination doesn't produce any single exception,
++ * then we have a bad program that isn't synchronizing its FPU usage
++ * and it will suffer the consequences since we won't be able to
++ * fully reproduce the context of the exception
++ */
++ cwd = get_fpu_cwd(task);
++ swd = get_fpu_swd(task);
++ switch (swd & ~cwd & 0x3f) {
++ case 0x000:
++ default:
++ break;
++ case 0x001: /* Invalid Op */
++ /*
++ * swd & 0x240 == 0x040: Stack Underflow
++ * swd & 0x240 == 0x240: Stack Overflow
++ * User must clear the SF bit (0x40) if set
++ */
++ info.si_code = FPE_FLTINV;
++ break;
++ case 0x002: /* Denormalize */
++ case 0x010: /* Underflow */
++ info.si_code = FPE_FLTUND;
++ break;
++ case 0x004: /* Zero Divide */
++ info.si_code = FPE_FLTDIV;
++ break;
++ case 0x008: /* Overflow */
++ info.si_code = FPE_FLTOVF;
++ break;
++ case 0x020: /* Precision */
++ info.si_code = FPE_FLTRES;
++ break;
++ }
++ force_sig_info(SIGFPE, &info, task);
++}
++
++asmlinkage void bad_intr(void)
++{
++ printk("bad interrupt");
++}
++
++asmlinkage void do_simd_coprocessor_error(struct pt_regs *regs)
++{
++ void __user *rip = (void __user *)(regs->rip);
++ struct task_struct * task;
++ siginfo_t info;
++ unsigned short mxcsr;
++
++ conditional_sti(regs);
++ if (!user_mode(regs) &&
++ kernel_math_error(regs, "kernel simd math error", 19))
++ return;
++
++ /*
++ * Save the info for the exception handler and clear the error.
++ */
++ task = current;
++ save_init_fpu(task);
++ task->thread.trap_no = 19;
++ task->thread.error_code = 0;
++ info.si_signo = SIGFPE;
++ info.si_errno = 0;
++ info.si_code = __SI_FAULT;
++ info.si_addr = rip;
++ /*
++ * The SIMD FPU exceptions are handled a little differently, as there
++ * is only a single status/control register. Thus, to determine which
++ * unmasked exception was caught we must mask the exception mask bits
++ * at 0x1f80, and then use these to mask the exception bits at 0x3f.
++ */
++ mxcsr = get_fpu_mxcsr(task);
++ switch (~((mxcsr & 0x1f80) >> 7) & (mxcsr & 0x3f)) {
++ case 0x000:
++ default:
++ break;
++ case 0x001: /* Invalid Op */
++ info.si_code = FPE_FLTINV;
++ break;
++ case 0x002: /* Denormalize */
++ case 0x010: /* Underflow */
++ info.si_code = FPE_FLTUND;
++ break;
++ case 0x004: /* Zero Divide */
++ info.si_code = FPE_FLTDIV;
++ break;
++ case 0x008: /* Overflow */
++ info.si_code = FPE_FLTOVF;
++ break;
++ case 0x020: /* Precision */
++ info.si_code = FPE_FLTRES;
++ break;
++ }
++ force_sig_info(SIGFPE, &info, task);
++}
++
++asmlinkage void do_spurious_interrupt_bug(struct pt_regs * regs)
++{
++}
++
++#if 0
++asmlinkage void __attribute__((weak)) smp_thermal_interrupt(void)
++{
++}
++#endif
++
++asmlinkage void __attribute__((weak)) mce_threshold_interrupt(void)
++{
++}
++
++/*
++ * 'math_state_restore()' saves the current math information in the
++ * old math state array, and gets the new ones from the current task
++ *
++ * Careful.. There are problems with IBM-designed IRQ13 behaviour.
++ * Don't touch unless you *really* know how it works.
++ */
++asmlinkage void math_state_restore(void)
++{
++ struct task_struct *me = current;
++ /* clts(); */ /* 'clts' is done for us by Xen during virtual trap. */
++
++ if (!used_math())
++ init_fpu(me);
++ restore_fpu_checking(&me->thread.i387.fxsave);
++ task_thread_info(me)->status |= TS_USEDFPU;
++}
++
++
++/*
++ * NB. All these are "interrupt gates" (i.e. events_mask is set) because we
++ * specify <dpl>|4 in the second field.
++ */
++static trap_info_t __cpuinitdata trap_table[] = {
++ { 0, 0|4, __KERNEL_CS, (unsigned long)divide_error },
++ { 1, 0|4, __KERNEL_CS, (unsigned long)debug },
++ { 3, 3|4, __KERNEL_CS, (unsigned long)int3 },
++ { 4, 3|4, __KERNEL_CS, (unsigned long)overflow },
++ { 5, 0|4, __KERNEL_CS, (unsigned long)bounds },
++ { 6, 0|4, __KERNEL_CS, (unsigned long)invalid_op },
++ { 7, 0|4, __KERNEL_CS, (unsigned long)device_not_available },
++ { 9, 0|4, __KERNEL_CS, (unsigned long)coprocessor_segment_overrun},
++ { 10, 0|4, __KERNEL_CS, (unsigned long)invalid_TSS },
++ { 11, 0|4, __KERNEL_CS, (unsigned long)segment_not_present },
++ { 12, 0|4, __KERNEL_CS, (unsigned long)stack_segment },
++ { 13, 0|4, __KERNEL_CS, (unsigned long)general_protection },
++ { 14, 0|4, __KERNEL_CS, (unsigned long)page_fault },
++ { 15, 0|4, __KERNEL_CS, (unsigned long)spurious_interrupt_bug },
++ { 16, 0|4, __KERNEL_CS, (unsigned long)coprocessor_error },
++ { 17, 0|4, __KERNEL_CS, (unsigned long)alignment_check },
++#ifdef CONFIG_X86_MCE
++ { 18, 0|4, __KERNEL_CS, (unsigned long)machine_check },
++#endif
++ { 19, 0|4, __KERNEL_CS, (unsigned long)simd_coprocessor_error },
++#ifdef CONFIG_IA32_EMULATION
++ { IA32_SYSCALL_VECTOR, 3, __KERNEL_CS, (unsigned long)ia32_syscall},
++#endif
++ { 0, 0, 0, 0 }
++};
++
++void __init trap_init(void)
++{
++ int ret;
++
++ ret = HYPERVISOR_set_trap_table(trap_table);
++ if (ret)
++ printk("HYPERVISOR_set_trap_table failed: error %d\n", ret);
++
++ /*
++ * Should be a barrier for any external CPU state.
++ */
++ cpu_init();
++}
++
++void __cpuinit smp_trap_init(trap_info_t *trap_ctxt)
++{
++ const trap_info_t *t = trap_table;
++
++ for (t = trap_table; t->address; t++) {
++ trap_ctxt[t->vector].flags = t->flags;
++ trap_ctxt[t->vector].cs = t->cs;
++ trap_ctxt[t->vector].address = t->address;
++ }
++}
++
++
++/* Actual parsing is done early in setup.c. */
++static int __init oops_dummy(char *s)
++{
++ panic_on_oops = 1;
++ return 1;
++}
++__setup("oops=", oops_dummy);
++
++static int __init kstack_setup(char *s)
++{
++ kstack_depth_to_print = simple_strtoul(s,NULL,0);
++ return 1;
++}
++__setup("kstack=", kstack_setup);
++
++#ifdef CONFIG_STACK_UNWIND
++static int __init call_trace_setup(char *s)
++{
++ if (strcmp(s, "old") == 0)
++ call_trace = -1;
++ else if (strcmp(s, "both") == 0)
++ call_trace = 0;
++ else if (strcmp(s, "newfallback") == 0)
++ call_trace = 1;
++ else if (strcmp(s, "new") == 0)
++ call_trace = 2;
++ return 1;
++}
++__setup("call_trace=", call_trace_setup);
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/vmlinux.lds.S
+--- a/arch/x86_64/kernel/vmlinux.lds.S Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/kernel/vmlinux.lds.S Wed Sep 10 10:54:08 2008 +0100
+@@ -13,6 +13,13 @@
+ OUTPUT_ARCH(i386:x86-64)
+ ENTRY(phys_startup_64)
+ jiffies_64 = jiffies;
++PHDRS {
++ text PT_LOAD FLAGS(5); /* R_E */
++ data PT_LOAD FLAGS(7); /* RWE */
++ user PT_LOAD FLAGS(7); /* RWE */
++ data.init PT_LOAD FLAGS(7); /* RWE */
++ note PT_NOTE FLAGS(4); /* R__ */
++}
+ SECTIONS
+ {
+ . = __START_KERNEL;
+@@ -31,7 +38,7 @@
+ KPROBES_TEXT
+ *(.fixup)
+ *(.gnu.warning)
+- } = 0x9090
++ } :text = 0x9090
+ /* out-of-line lock text */
+ .text.lock : AT(ADDR(.text.lock) - LOAD_OFFSET) { *(.text.lock) }
+
+@@ -57,16 +64,9 @@
+ .data : AT(ADDR(.data) - LOAD_OFFSET) {
+ *(.data)
+ CONSTRUCTORS
+- }
++ } :data
+
+ _edata = .; /* End of data section */
+-
+- __bss_start = .; /* BSS */
+- .bss : AT(ADDR(.bss) - LOAD_OFFSET) {
+- *(.bss.page_aligned)
+- *(.bss)
+- }
+- __bss_stop = .;
+
+ . = ALIGN(PAGE_SIZE);
+ . = ALIGN(CONFIG_X86_L1_CACHE_BYTES);
+@@ -89,7 +89,7 @@
+ #define VVIRT(x) (ADDR(x) - VVIRT_OFFSET)
+
+ . = VSYSCALL_ADDR;
+- .vsyscall_0 : AT(VSYSCALL_PHYS_ADDR) { *(.vsyscall_0) }
++ .vsyscall_0 : AT(VSYSCALL_PHYS_ADDR) { *(.vsyscall_0) } :user
+ __vsyscall_0 = VSYSCALL_VIRT_ADDR;
+
+ . = ALIGN(CONFIG_X86_L1_CACHE_BYTES);
+@@ -132,7 +132,7 @@
+ . = ALIGN(8192); /* init_task */
+ .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) {
+ *(.data.init_task)
+- }
++ }:data.init
+
+ . = ALIGN(4096);
+ .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) {
+@@ -222,6 +222,14 @@
+ . = ALIGN(4096);
+ __nosave_end = .;
+
++ __bss_start = .; /* BSS */
++ . = ALIGN(4096);
++ .bss : AT(ADDR(.bss) - LOAD_OFFSET) {
++ *(.bss.page_aligned)
++ *(.bss)
++ }
++ __bss_stop = .;
++
+ _end = . ;
+
+ /* Sections to be discarded */
+@@ -235,4 +243,6 @@
+ STABS_DEBUG
+
+ DWARF_DEBUG
++
++ NOTES
+ }
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/vsyscall-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/vsyscall-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,227 @@
++/*
++ * linux/arch/x86_64/kernel/vsyscall.c
++ *
++ * Copyright (C) 2001 Andrea Arcangeli <andrea at suse.de> SuSE
++ * Copyright 2003 Andi Kleen, SuSE Labs.
++ *
++ * Thanks to hpa at transmeta.com for some useful hint.
++ * Special thanks to Ingo Molnar for his early experience with
++ * a different vsyscall implementation for Linux/IA32 and for the name.
++ *
++ * vsyscall 1 is located at -10Mbyte, vsyscall 2 is located
++ * at virtual address -10Mbyte+1024bytes etc... There are at max 4
++ * vsyscalls. One vsyscall can reserve more than 1 slot to avoid
++ * jumping out of line if necessary. We cannot add more with this
++ * mechanism because older kernels won't return -ENOSYS.
++ * If we want more than four we need a vDSO.
++ *
++ * Note: the concept clashes with user mode linux. If you use UML and
++ * want per guest time just set the kernel.vsyscall64 sysctl to 0.
++ */
++
++#include <linux/time.h>
++#include <linux/init.h>
++#include <linux/kernel.h>
++#include <linux/timer.h>
++#include <linux/seqlock.h>
++#include <linux/jiffies.h>
++#include <linux/sysctl.h>
++
++#include <asm/vsyscall.h>
++#include <asm/pgtable.h>
++#include <asm/page.h>
++#include <asm/fixmap.h>
++#include <asm/errno.h>
++#include <asm/io.h>
++
++#define __vsyscall(nr) __attribute__ ((unused,__section__(".vsyscall_" #nr)))
++
++int __sysctl_vsyscall __section_sysctl_vsyscall = 1;
++seqlock_t __xtime_lock __section_xtime_lock = SEQLOCK_UNLOCKED;
++
++#include <asm/unistd.h>
++
++static __always_inline void timeval_normalize(struct timeval * tv)
++{
++ time_t __sec;
++
++ __sec = tv->tv_usec / 1000000;
++ if (__sec) {
++ tv->tv_usec %= 1000000;
++ tv->tv_sec += __sec;
++ }
++}
++
++static __always_inline void do_vgettimeofday(struct timeval * tv)
++{
++ long sequence, t;
++ unsigned long sec, usec;
++
++ do {
++ sequence = read_seqbegin(&__xtime_lock);
++
++ sec = __xtime.tv_sec;
++ usec = (__xtime.tv_nsec / 1000) +
++ (__jiffies - __wall_jiffies) * (1000000 / HZ);
++
++ if (__vxtime.mode != VXTIME_HPET) {
++ t = get_cycles_sync();
++ if (t < __vxtime.last_tsc)
++ t = __vxtime.last_tsc;
++ usec += ((t - __vxtime.last_tsc) *
++ __vxtime.tsc_quot) >> 32;
++ /* See comment in x86_64 do_gettimeofday. */
++ } else {
++ usec += ((readl((void *)fix_to_virt(VSYSCALL_HPET) + 0xf0) -
++ __vxtime.last) * __vxtime.quot) >> 32;
++ }
++ } while (read_seqretry(&__xtime_lock, sequence));
++
++ tv->tv_sec = sec + usec / 1000000;
++ tv->tv_usec = usec % 1000000;
++}
++
++/* RED-PEN may want to readd seq locking, but then the variable should be write-once. */
++static __always_inline void do_get_tz(struct timezone * tz)
++{
++ *tz = __sys_tz;
++}
++
++static __always_inline int gettimeofday(struct timeval *tv, struct timezone *tz)
++{
++ int ret;
++ asm volatile("vsysc2: syscall"
++ : "=a" (ret)
++ : "0" (__NR_gettimeofday),"D" (tv),"S" (tz) : __syscall_clobber );
++ return ret;
++}
++
++static __always_inline long time_syscall(long *t)
++{
++ long secs;
++ asm volatile("vsysc1: syscall"
++ : "=a" (secs)
++ : "0" (__NR_time),"D" (t) : __syscall_clobber);
++ return secs;
++}
++
++int __vsyscall(0) vgettimeofday(struct timeval * tv, struct timezone * tz)
++{
++ if (!__sysctl_vsyscall)
++ return gettimeofday(tv,tz);
++ if (tv)
++ do_vgettimeofday(tv);
++ if (tz)
++ do_get_tz(tz);
++ return 0;
++}
++
++/* This will break when the xtime seconds get inaccurate, but that is
++ * unlikely */
++time_t __vsyscall(1) vtime(time_t *t)
++{
++ if (!__sysctl_vsyscall)
++ return time_syscall(t);
++ else if (t)
++ *t = __xtime.tv_sec;
++ return __xtime.tv_sec;
++}
++
++long __vsyscall(2) venosys_0(void)
++{
++ return -ENOSYS;
++}
++
++long __vsyscall(3) venosys_1(void)
++{
++ return -ENOSYS;
++}
++
++#ifdef CONFIG_SYSCTL
++
++#define SYSCALL 0x050f
++#define NOP2 0x9090
++
++/*
++ * NOP out syscall in vsyscall page when not needed.
++ */
++static int vsyscall_sysctl_change(ctl_table *ctl, int write, struct file * filp,
++ void __user *buffer, size_t *lenp, loff_t *ppos)
++{
++ extern u16 vsysc1, vsysc2;
++ u16 *map1, *map2;
++ int ret = proc_dointvec(ctl, write, filp, buffer, lenp, ppos);
++ if (!write)
++ return ret;
++ /* gcc has some trouble with __va(__pa()), so just do it this
++ way. */
++ map1 = ioremap(__pa_symbol(&vsysc1), 2);
++ if (!map1)
++ return -ENOMEM;
++ map2 = ioremap(__pa_symbol(&vsysc2), 2);
++ if (!map2) {
++ ret = -ENOMEM;
++ goto out;
++ }
++ if (!sysctl_vsyscall) {
++ *map1 = SYSCALL;
++ *map2 = SYSCALL;
++ } else {
++ *map1 = NOP2;
++ *map2 = NOP2;
++ }
++ iounmap(map2);
++out:
++ iounmap(map1);
++ return ret;
++}
++
++static int vsyscall_sysctl_nostrat(ctl_table *t, int __user *name, int nlen,
++ void __user *oldval, size_t __user *oldlenp,
++ void __user *newval, size_t newlen,
++ void **context)
++{
++ return -ENOSYS;
++}
++
++static ctl_table kernel_table2[] = {
++ { .ctl_name = 99, .procname = "vsyscall64",
++ .data = &sysctl_vsyscall, .maxlen = sizeof(int), .mode = 0644,
++ .strategy = vsyscall_sysctl_nostrat,
++ .proc_handler = vsyscall_sysctl_change },
++ { 0, }
++};
++
++static ctl_table kernel_root_table2[] = {
++ { .ctl_name = CTL_KERN, .procname = "kernel", .mode = 0555,
++ .child = kernel_table2 },
++ { 0 },
++};
++
++#endif
++
++static void __init map_vsyscall(void)
++{
++ extern char __vsyscall_0;
++ unsigned long physaddr_page0 = __pa_symbol(&__vsyscall_0);
++
++ __set_fixmap(VSYSCALL_FIRST_PAGE, physaddr_page0, PAGE_KERNEL_VSYSCALL);
++}
++
++static int __init vsyscall_init(void)
++{
++ BUG_ON(((unsigned long) &vgettimeofday !=
++ VSYSCALL_ADDR(__NR_vgettimeofday)));
++ BUG_ON((unsigned long) &vtime != VSYSCALL_ADDR(__NR_vtime));
++ BUG_ON((VSYSCALL_ADDR(0) != __fix_to_virt(VSYSCALL_FIRST_PAGE)));
++ map_vsyscall();
++#ifdef CONFIG_XEN
++ sysctl_vsyscall = 0; /* disable vgettimeofay() */
++#endif
++#ifdef CONFIG_SYSCTL
++ register_sysctl_table(kernel_root_table2, 0);
++#endif
++ return 0;
++}
++
++__initcall(vsyscall_init);
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/kernel/xen_entry.S
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/kernel/xen_entry.S Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,36 @@
++/*
++ * Copied from arch/xen/i386/kernel/entry.S
++ */
++/* Offsets into shared_info_t. */
++#define evtchn_upcall_pending /* 0 */
++#define evtchn_upcall_mask 1
++
++#define sizeof_vcpu_shift 6
++
++#ifdef CONFIG_SMP
++//#define preempt_disable(reg) incl threadinfo_preempt_count(reg)
++//#define preempt_enable(reg) decl threadinfo_preempt_count(reg)
++#define preempt_disable(reg)
++#define preempt_enable(reg)
++#define XEN_GET_VCPU_INFO(reg) preempt_disable(%rbp) ; \
++ movq %gs:pda_cpunumber,reg ; \
++ shl $32, reg ; \
++ shr $32-sizeof_vcpu_shift,reg ; \
++ addq HYPERVISOR_shared_info,reg
++#define XEN_PUT_VCPU_INFO(reg) preempt_enable(%rbp) ; \
++#define XEN_PUT_VCPU_INFO_fixup .byte 0xff,0xff,0xff
++#else
++#define XEN_GET_VCPU_INFO(reg) movq HYPERVISOR_shared_info,reg
++#define XEN_PUT_VCPU_INFO(reg)
++#define XEN_PUT_VCPU_INFO_fixup
++#endif
++
++#define XEN_LOCKED_BLOCK_EVENTS(reg) movb $1,evtchn_upcall_mask(reg)
++#define XEN_LOCKED_UNBLOCK_EVENTS(reg) movb $0,evtchn_upcall_mask(reg)
++#define XEN_BLOCK_EVENTS(reg) XEN_GET_VCPU_INFO(reg) ; \
++ XEN_LOCKED_BLOCK_EVENTS(reg) ; \
++ XEN_PUT_VCPU_INFO(reg)
++#define XEN_UNBLOCK_EVENTS(reg) XEN_GET_VCPU_INFO(reg) ; \
++ XEN_LOCKED_UNBLOCK_EVENTS(reg) ; \
++ XEN_PUT_VCPU_INFO(reg)
++#define XEN_TEST_PENDING(reg) testb $0xFF,evtchn_upcall_pending(reg)
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/lib/Makefile
+--- a/arch/x86_64/lib/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/lib/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -10,3 +10,4 @@
+ usercopy.o getuser.o putuser.o \
+ thunk.o clear_page.o copy_page.o bitstr.o bitops.o
+ lib-y += memcpy.o memmove.o memset.o copy_user.o
++lib-$(CONFIG_XEN_SCRUB_PAGES) += scrub.o
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/lib/scrub.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/lib/scrub.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1 @@
++#include "../../i386/lib/scrub.c"
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/mm/Makefile
+--- a/arch/x86_64/mm/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/mm/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -7,5 +7,8 @@
+ obj-$(CONFIG_NUMA) += numa.o
+ obj-$(CONFIG_K8_NUMA) += k8topology.o
+ obj-$(CONFIG_ACPI_NUMA) += srat.o
++obj-$(CONFIG_XEN) += hypervisor.o
+
+ hugetlbpage-y = ../../i386/mm/hugetlbpage.o
++ioremap-$(CONFIG_XEN) := ../../i386/mm/ioremap.o
++hypervisor-y := ../../i386/mm/hypervisor.o
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/mm/fault-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/mm/fault-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,724 @@
++/*
++ * linux/arch/x86-64/mm/fault.c
++ *
++ * Copyright (C) 1995 Linus Torvalds
++ * Copyright (C) 2001,2002 Andi Kleen, SuSE Labs.
++ */
++
++#include <linux/signal.h>
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/string.h>
++#include <linux/types.h>
++#include <linux/ptrace.h>
++#include <linux/mman.h>
++#include <linux/mm.h>
++#include <linux/smp.h>
++#include <linux/smp_lock.h>
++#include <linux/interrupt.h>
++#include <linux/init.h>
++#include <linux/tty.h>
++#include <linux/vt_kern.h> /* For unblank_screen() */
++#include <linux/compiler.h>
++#include <linux/module.h>
++#include <linux/kprobes.h>
++
++#include <asm/system.h>
++#include <asm/uaccess.h>
++#include <asm/pgalloc.h>
++#include <asm/smp.h>
++#include <asm/tlbflush.h>
++#include <asm/proto.h>
++#include <asm/kdebug.h>
++#include <asm-generic/sections.h>
++
++/* Page fault error code bits */
++#define PF_PROT (1<<0) /* or no page found */
++#define PF_WRITE (1<<1)
++#define PF_USER (1<<2)
++#define PF_RSVD (1<<3)
++#define PF_INSTR (1<<4)
++
++#ifdef CONFIG_KPROBES
++ATOMIC_NOTIFIER_HEAD(notify_page_fault_chain);
++
++/* Hook to register for page fault notifications */
++int register_page_fault_notifier(struct notifier_block *nb)
++{
++ vmalloc_sync_all();
++ return atomic_notifier_chain_register(¬ify_page_fault_chain, nb);
++}
++
++int unregister_page_fault_notifier(struct notifier_block *nb)
++{
++ return atomic_notifier_chain_unregister(¬ify_page_fault_chain, nb);
++}
++
++static inline int notify_page_fault(enum die_val val, const char *str,
++ struct pt_regs *regs, long err, int trap, int sig)
++{
++ struct die_args args = {
++ .regs = regs,
++ .str = str,
++ .err = err,
++ .trapnr = trap,
++ .signr = sig
++ };
++ return atomic_notifier_call_chain(¬ify_page_fault_chain, val, &args);
++}
++#else
++static inline int notify_page_fault(enum die_val val, const char *str,
++ struct pt_regs *regs, long err, int trap, int sig)
++{
++ return NOTIFY_DONE;
++}
++#endif
++
++void bust_spinlocks(int yes)
++{
++ int loglevel_save = console_loglevel;
++ if (yes) {
++ oops_in_progress = 1;
++ } else {
++#ifdef CONFIG_VT
++ unblank_screen();
++#endif
++ oops_in_progress = 0;
++ /*
++ * OK, the message is on the console. Now we call printk()
++ * without oops_in_progress set so that printk will give klogd
++ * a poke. Hold onto your hats...
++ */
++ console_loglevel = 15; /* NMI oopser may have shut the console up */
++ printk(" ");
++ console_loglevel = loglevel_save;
++ }
++}
++
++/* Sometimes the CPU reports invalid exceptions on prefetch.
++ Check that here and ignore.
++ Opcode checker based on code by Richard Brunner */
++static noinline int is_prefetch(struct pt_regs *regs, unsigned long addr,
++ unsigned long error_code)
++{
++ unsigned char *instr;
++ int scan_more = 1;
++ int prefetch = 0;
++ unsigned char *max_instr;
++
++ /* If it was a exec fault ignore */
++ if (error_code & PF_INSTR)
++ return 0;
++
++ instr = (unsigned char *)convert_rip_to_linear(current, regs);
++ max_instr = instr + 15;
++
++ if (user_mode(regs) && instr >= (unsigned char *)TASK_SIZE)
++ return 0;
++
++ while (scan_more && instr < max_instr) {
++ unsigned char opcode;
++ unsigned char instr_hi;
++ unsigned char instr_lo;
++
++ if (__get_user(opcode, instr))
++ break;
++
++ instr_hi = opcode & 0xf0;
++ instr_lo = opcode & 0x0f;
++ instr++;
++
++ switch (instr_hi) {
++ case 0x20:
++ case 0x30:
++ /* Values 0x26,0x2E,0x36,0x3E are valid x86
++ prefixes. In long mode, the CPU will signal
++ invalid opcode if some of these prefixes are
++ present so we will never get here anyway */
++ scan_more = ((instr_lo & 7) == 0x6);
++ break;
++
++ case 0x40:
++ /* In AMD64 long mode, 0x40 to 0x4F are valid REX prefixes
++ Need to figure out under what instruction mode the
++ instruction was issued ... */
++ /* Could check the LDT for lm, but for now it's good
++ enough to assume that long mode only uses well known
++ segments or kernel. */
++ scan_more = (!user_mode(regs)) || (regs->cs == __USER_CS);
++ break;
++
++ case 0x60:
++ /* 0x64 thru 0x67 are valid prefixes in all modes. */
++ scan_more = (instr_lo & 0xC) == 0x4;
++ break;
++ case 0xF0:
++ /* 0xF0, 0xF2, and 0xF3 are valid prefixes in all modes. */
++ scan_more = !instr_lo || (instr_lo>>1) == 1;
++ break;
++ case 0x00:
++ /* Prefetch instruction is 0x0F0D or 0x0F18 */
++ scan_more = 0;
++ if (__get_user(opcode, instr))
++ break;
++ prefetch = (instr_lo == 0xF) &&
++ (opcode == 0x0D || opcode == 0x18);
++ break;
++ default:
++ scan_more = 0;
++ break;
++ }
++ }
++ return prefetch;
++}
++
++static int bad_address(void *p)
++{
++ unsigned long dummy;
++ return __get_user(dummy, (unsigned long *)p);
++}
++
++void dump_pagetable(unsigned long address)
++{
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
++
++ pgd = __va(read_cr3() & PHYSICAL_PAGE_MASK);
++ pgd += pgd_index(address);
++ if (bad_address(pgd)) goto bad;
++ printk("PGD %lx ", pgd_val(*pgd));
++ if (!pgd_present(*pgd)) goto ret;
++
++ pud = pud_offset(pgd, address);
++ if (bad_address(pud)) goto bad;
++ printk("PUD %lx ", pud_val(*pud));
++ if (!pud_present(*pud)) goto ret;
++
++ pmd = pmd_offset(pud, address);
++ if (bad_address(pmd)) goto bad;
++ printk("PMD %lx ", pmd_val(*pmd));
++ if (!pmd_present(*pmd)) goto ret;
++
++ pte = pte_offset_kernel(pmd, address);
++ if (bad_address(pte)) goto bad;
++ printk("PTE %lx", pte_val(*pte));
++ret:
++ printk("\n");
++ return;
++bad:
++ printk("BAD\n");
++}
++
++static const char errata93_warning[] =
++KERN_ERR "******* Your BIOS seems to not contain a fix for K8 errata #93\n"
++KERN_ERR "******* Working around it, but it may cause SEGVs or burn power.\n"
++KERN_ERR "******* Please consider a BIOS update.\n"
++KERN_ERR "******* Disabling USB legacy in the BIOS may also help.\n";
++
++/* Workaround for K8 erratum #93 & buggy BIOS.
++ BIOS SMM functions are required to use a specific workaround
++ to avoid corruption of the 64bit RIP register on C stepping K8.
++ A lot of BIOS that didn't get tested properly miss this.
++ The OS sees this as a page fault with the upper 32bits of RIP cleared.
++ Try to work around it here.
++ Note we only handle faults in kernel here. */
++
++static int is_errata93(struct pt_regs *regs, unsigned long address)
++{
++ static int warned;
++ if (address != regs->rip)
++ return 0;
++ if ((address >> 32) != 0)
++ return 0;
++ address |= 0xffffffffUL << 32;
++ if ((address >= (u64)_stext && address <= (u64)_etext) ||
++ (address >= MODULES_VADDR && address <= MODULES_END)) {
++ if (!warned) {
++ printk(errata93_warning);
++ warned = 1;
++ }
++ regs->rip = address;
++ return 1;
++ }
++ return 0;
++}
++
++int unhandled_signal(struct task_struct *tsk, int sig)
++{
++ if (tsk->pid == 1)
++ return 1;
++ if (tsk->ptrace & PT_PTRACED)
++ return 0;
++ return (tsk->sighand->action[sig-1].sa.sa_handler == SIG_IGN) ||
++ (tsk->sighand->action[sig-1].sa.sa_handler == SIG_DFL);
++}
++
++static noinline void pgtable_bad(unsigned long address, struct pt_regs *regs,
++ unsigned long error_code)
++{
++ unsigned long flags = oops_begin();
++ struct task_struct *tsk;
++
++ printk(KERN_ALERT "%s: Corrupted page table at address %lx\n",
++ current->comm, address);
++ dump_pagetable(address);
++ tsk = current;
++ tsk->thread.cr2 = address;
++ tsk->thread.trap_no = 14;
++ tsk->thread.error_code = error_code;
++ __die("Bad pagetable", regs, error_code);
++ oops_end(flags);
++ do_exit(SIGKILL);
++}
++
++/*
++ * Handle a fault on the vmalloc area
++ *
++ * This assumes no large pages in there.
++ */
++static int vmalloc_fault(unsigned long address)
++{
++ pgd_t *pgd, *pgd_ref;
++ pud_t *pud, *pud_ref;
++ pmd_t *pmd, *pmd_ref;
++ pte_t *pte, *pte_ref;
++
++ /* Copy kernel mappings over when needed. This can also
++ happen within a race in page table update. In the later
++ case just flush. */
++
++ /* On Xen the line below does not always work. Needs investigating! */
++ /*pgd = pgd_offset(current->mm ?: &init_mm, address);*/
++ pgd = __va(read_cr3() & PHYSICAL_PAGE_MASK);
++ pgd += pgd_index(address);
++ pgd_ref = pgd_offset_k(address);
++ if (pgd_none(*pgd_ref))
++ return -1;
++ if (pgd_none(*pgd))
++ set_pgd(pgd, *pgd_ref);
++ else
++ BUG_ON(pgd_page(*pgd) != pgd_page(*pgd_ref));
++
++ /* Below here mismatches are bugs because these lower tables
++ are shared */
++
++ pud = pud_offset(pgd, address);
++ pud_ref = pud_offset(pgd_ref, address);
++ if (pud_none(*pud_ref))
++ return -1;
++ if (pud_none(*pud) || pud_page(*pud) != pud_page(*pud_ref))
++ BUG();
++ pmd = pmd_offset(pud, address);
++ pmd_ref = pmd_offset(pud_ref, address);
++ if (pmd_none(*pmd_ref))
++ return -1;
++ if (pmd_none(*pmd) || pmd_page(*pmd) != pmd_page(*pmd_ref))
++ BUG();
++ pte_ref = pte_offset_kernel(pmd_ref, address);
++ if (!pte_present(*pte_ref))
++ return -1;
++ pte = pte_offset_kernel(pmd, address);
++ /* Don't use pte_page here, because the mappings can point
++ outside mem_map, and the NUMA hash lookup cannot handle
++ that. */
++ if (!pte_present(*pte) || pte_pfn(*pte) != pte_pfn(*pte_ref))
++ BUG();
++ return 0;
++}
++
++int page_fault_trace = 0;
++int exception_trace = 1;
++
++
++#define MEM_VERBOSE 1
++
++#ifdef MEM_VERBOSE
++#define MEM_LOG(_f, _a...) \
++ printk("fault.c:[%d]-> " _f "\n", \
++ __LINE__ , ## _a )
++#else
++#define MEM_LOG(_f, _a...) ((void)0)
++#endif
++
++static int spurious_fault(struct pt_regs *regs,
++ unsigned long address,
++ unsigned long error_code)
++{
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
++
++#ifdef CONFIG_XEN
++ /* Faults in hypervisor area are never spurious. */
++ if ((address >= HYPERVISOR_VIRT_START) &&
++ (address < HYPERVISOR_VIRT_END))
++ return 0;
++#endif
++
++ /* Reserved-bit violation or user access to kernel space? */
++ if (error_code & (PF_RSVD|PF_USER))
++ return 0;
++
++ pgd = init_mm.pgd + pgd_index(address);
++ if (!pgd_present(*pgd))
++ return 0;
++
++ pud = pud_offset(pgd, address);
++ if (!pud_present(*pud))
++ return 0;
++
++ pmd = pmd_offset(pud, address);
++ if (!pmd_present(*pmd))
++ return 0;
++
++ pte = pte_offset_kernel(pmd, address);
++ if (!pte_present(*pte))
++ return 0;
++ if ((error_code & PF_WRITE) && !pte_write(*pte))
++ return 0;
++ if ((error_code & PF_INSTR) && (__pte_val(*pte) & _PAGE_NX))
++ return 0;
++
++ return 1;
++}
++
++/*
++ * This routine handles page faults. It determines the address,
++ * and the problem, and then passes it off to one of the appropriate
++ * routines.
++ */
++asmlinkage void __kprobes do_page_fault(struct pt_regs *regs,
++ unsigned long error_code)
++{
++ struct task_struct *tsk;
++ struct mm_struct *mm;
++ struct vm_area_struct * vma;
++ unsigned long address;
++ const struct exception_table_entry *fixup;
++ int write;
++ unsigned long flags;
++ siginfo_t info;
++
++ if (!user_mode(regs))
++ error_code &= ~PF_USER; /* means kernel */
++
++ tsk = current;
++ mm = tsk->mm;
++ prefetchw(&mm->mmap_sem);
++
++ /* get the address */
++ address = current_vcpu_info()->arch.cr2;
++
++ info.si_code = SEGV_MAPERR;
++
++
++ /*
++ * We fault-in kernel-space virtual memory on-demand. The
++ * 'reference' page table is init_mm.pgd.
++ *
++ * NOTE! We MUST NOT take any locks for this case. We may
++ * be in an interrupt or a critical region, and should
++ * only copy the information from the master page table,
++ * nothing more.
++ *
++ * This verifies that the fault happens in kernel space
++ * (error_code & 4) == 0, and that the fault was not a
++ * protection error (error_code & 9) == 0.
++ */
++ if (unlikely(address >= TASK_SIZE64)) {
++ /*
++ * Don't check for the module range here: its PML4
++ * is always initialized because it's shared with the main
++ * kernel text. Only vmalloc may need PML4 syncups.
++ */
++ if (!(error_code & (PF_RSVD|PF_USER|PF_PROT)) &&
++ ((address >= VMALLOC_START && address < VMALLOC_END))) {
++ if (vmalloc_fault(address) >= 0)
++ return;
++ }
++ /* Can take a spurious fault if mapping changes R/O -> R/W. */
++ if (spurious_fault(regs, address, error_code))
++ return;
++ if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
++ SIGSEGV) == NOTIFY_STOP)
++ return;
++ /*
++ * Don't take the mm semaphore here. If we fixup a prefetch
++ * fault we could otherwise deadlock.
++ */
++ goto bad_area_nosemaphore;
++ }
++
++ if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
++ SIGSEGV) == NOTIFY_STOP)
++ return;
++
++ if (likely(regs->eflags & X86_EFLAGS_IF))
++ local_irq_enable();
++
++ if (unlikely(page_fault_trace))
++ printk("pagefault rip:%lx rsp:%lx cs:%lu ss:%lu address %lx error %lx\n",
++ regs->rip,regs->rsp,regs->cs,regs->ss,address,error_code);
++
++ if (unlikely(error_code & PF_RSVD))
++ pgtable_bad(address, regs, error_code);
++
++ /*
++ * If we're in an interrupt or have no user
++ * context, we must not take the fault..
++ */
++ if (unlikely(in_atomic() || !mm))
++ goto bad_area_nosemaphore;
++
++ again:
++ /* When running in the kernel we expect faults to occur only to
++ * addresses in user space. All other faults represent errors in the
++ * kernel and should generate an OOPS. Unfortunatly, in the case of an
++ * erroneous fault occurring in a code path which already holds mmap_sem
++ * we will deadlock attempting to validate the fault against the
++ * address space. Luckily the kernel only validly references user
++ * space from well defined areas of code, which are listed in the
++ * exceptions table.
++ *
++ * As the vast majority of faults will be valid we will only perform
++ * the source reference check when there is a possibilty of a deadlock.
++ * Attempt to lock the address space, if we cannot we then validate the
++ * source. If this is invalid we can skip the address space check,
++ * thus avoiding the deadlock.
++ */
++ if (!down_read_trylock(&mm->mmap_sem)) {
++ if ((error_code & PF_USER) == 0 &&
++ !search_exception_tables(regs->rip))
++ goto bad_area_nosemaphore;
++ down_read(&mm->mmap_sem);
++ }
++
++ vma = find_vma(mm, address);
++ if (!vma)
++ goto bad_area;
++ if (likely(vma->vm_start <= address))
++ goto good_area;
++ if (!(vma->vm_flags & VM_GROWSDOWN))
++ goto bad_area;
++ if (error_code & 4) {
++ /* Allow userspace just enough access below the stack pointer
++ * to let the 'enter' instruction work.
++ */
++ if (address + 65536 + 32 * sizeof(unsigned long) < regs->rsp)
++ goto bad_area;
++ }
++ if (expand_stack(vma, address))
++ goto bad_area;
++/*
++ * Ok, we have a good vm_area for this memory access, so
++ * we can handle it..
++ */
++good_area:
++ info.si_code = SEGV_ACCERR;
++ write = 0;
++ switch (error_code & (PF_PROT|PF_WRITE)) {
++ default: /* 3: write, present */
++ /* fall through */
++ case PF_WRITE: /* write, not present */
++ if (!(vma->vm_flags & VM_WRITE))
++ goto bad_area;
++ write++;
++ break;
++ case PF_PROT: /* read, present */
++ goto bad_area;
++ case 0: /* read, not present */
++ if (!(vma->vm_flags & (VM_READ | VM_EXEC)))
++ goto bad_area;
++ }
++
++ /*
++ * If for any reason at all we couldn't handle the fault,
++ * make sure we exit gracefully rather than endlessly redo
++ * the fault.
++ */
++ switch (handle_mm_fault(mm, vma, address, write)) {
++ case VM_FAULT_MINOR:
++ tsk->min_flt++;
++ break;
++ case VM_FAULT_MAJOR:
++ tsk->maj_flt++;
++ break;
++ case VM_FAULT_SIGBUS:
++ goto do_sigbus;
++ default:
++ goto out_of_memory;
++ }
++
++ up_read(&mm->mmap_sem);
++ return;
++
++/*
++ * Something tried to access memory that isn't in our memory map..
++ * Fix it, but check if it's kernel or user first..
++ */
++bad_area:
++ up_read(&mm->mmap_sem);
++
++bad_area_nosemaphore:
++ /* User mode accesses just cause a SIGSEGV */
++ if (error_code & PF_USER) {
++ if (is_prefetch(regs, address, error_code))
++ return;
++
++ /* Work around K8 erratum #100 K8 in compat mode
++ occasionally jumps to illegal addresses >4GB. We
++ catch this here in the page fault handler because
++ these addresses are not reachable. Just detect this
++ case and return. Any code segment in LDT is
++ compatibility mode. */
++ if ((regs->cs == __USER32_CS || (regs->cs & (1<<2))) &&
++ (address >> 32))
++ return;
++
++ if (exception_trace && unhandled_signal(tsk, SIGSEGV)) {
++ printk(
++ "%s%s[%d]: segfault at %016lx rip %016lx rsp %016lx error %lx\n",
++ tsk->pid > 1 ? KERN_INFO : KERN_EMERG,
++ tsk->comm, tsk->pid, address, regs->rip,
++ regs->rsp, error_code);
++ }
++
++ tsk->thread.cr2 = address;
++ /* Kernel addresses are always protection faults */
++ tsk->thread.error_code = error_code | (address >= TASK_SIZE);
++ tsk->thread.trap_no = 14;
++ info.si_signo = SIGSEGV;
++ info.si_errno = 0;
++ /* info.si_code has been set above */
++ info.si_addr = (void __user *)address;
++ force_sig_info(SIGSEGV, &info, tsk);
++ return;
++ }
++
++no_context:
++
++ /* Are we prepared to handle this kernel fault? */
++ fixup = search_exception_tables(regs->rip);
++ if (fixup) {
++ regs->rip = fixup->fixup;
++ return;
++ }
++
++ /*
++ * Hall of shame of CPU/BIOS bugs.
++ */
++
++ if (is_prefetch(regs, address, error_code))
++ return;
++
++ if (is_errata93(regs, address))
++ return;
++
++/*
++ * Oops. The kernel tried to access some bad page. We'll have to
++ * terminate things with extreme prejudice.
++ */
++
++ flags = oops_begin();
++
++ if (address < PAGE_SIZE)
++ printk(KERN_ALERT "Unable to handle kernel NULL pointer dereference");
++ else
++ printk(KERN_ALERT "Unable to handle kernel paging request");
++ printk(" at %016lx RIP: \n" KERN_ALERT,address);
++ printk_address(regs->rip);
++ dump_pagetable(address);
++ tsk->thread.cr2 = address;
++ tsk->thread.trap_no = 14;
++ tsk->thread.error_code = error_code;
++ __die("Oops", regs, error_code);
++ /* Executive summary in case the body of the oops scrolled away */
++ printk(KERN_EMERG "CR2: %016lx\n", address);
++ oops_end(flags);
++ do_exit(SIGKILL);
++
++/*
++ * We ran out of memory, or some other thing happened to us that made
++ * us unable to handle the page fault gracefully.
++ */
++out_of_memory:
++ up_read(&mm->mmap_sem);
++ if (current->pid == 1) {
++ yield();
++ goto again;
++ }
++ printk("VM: killing process %s\n", tsk->comm);
++ if (error_code & 4)
++ do_exit(SIGKILL);
++ goto no_context;
++
++do_sigbus:
++ up_read(&mm->mmap_sem);
++
++ /* Kernel mode? Handle exceptions or die */
++ if (!(error_code & PF_USER))
++ goto no_context;
++
++ tsk->thread.cr2 = address;
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = 14;
++ info.si_signo = SIGBUS;
++ info.si_errno = 0;
++ info.si_code = BUS_ADRERR;
++ info.si_addr = (void __user *)address;
++ force_sig_info(SIGBUS, &info, tsk);
++ return;
++}
++
++DEFINE_SPINLOCK(pgd_lock);
++struct page *pgd_list;
++
++void vmalloc_sync_all(void)
++{
++ /* Note that races in the updates of insync and start aren't
++ problematic:
++ insync can only get set bits added, and updates to start are only
++ improving performance (without affecting correctness if undone). */
++ static DECLARE_BITMAP(insync, PTRS_PER_PGD);
++ static unsigned long start = VMALLOC_START & PGDIR_MASK;
++ unsigned long address;
++
++ for (address = start; address <= VMALLOC_END; address += PGDIR_SIZE) {
++ if (!test_bit(pgd_index(address), insync)) {
++ const pgd_t *pgd_ref = pgd_offset_k(address);
++ struct page *page;
++
++ if (pgd_none(*pgd_ref))
++ continue;
++ spin_lock(&pgd_lock);
++ for (page = pgd_list; page;
++ page = (struct page *)page->index) {
++ pgd_t *pgd;
++ pgd = (pgd_t *)page_address(page) + pgd_index(address);
++ if (pgd_none(*pgd))
++ set_pgd(pgd, *pgd_ref);
++ else
++ BUG_ON(pgd_page(*pgd) != pgd_page(*pgd_ref));
++ }
++ spin_unlock(&pgd_lock);
++ set_bit(pgd_index(address), insync);
++ }
++ if (address == start)
++ start = address + PGDIR_SIZE;
++ }
++ /* Check that there is no need to do the same for the modules area. */
++ BUILD_BUG_ON(!(MODULES_VADDR > __START_KERNEL));
++ BUILD_BUG_ON(!(((MODULES_END - 1) & PGDIR_MASK) ==
++ (__START_KERNEL & PGDIR_MASK)));
++}
++
++static int __init enable_pagefaulttrace(char *str)
++{
++ page_fault_trace = 1;
++ return 1;
++}
++__setup("pagefaulttrace", enable_pagefaulttrace);
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/mm/init-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/mm/init-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1213 @@
++/*
++ * linux/arch/x86_64/mm/init.c
++ *
++ * Copyright (C) 1995 Linus Torvalds
++ * Copyright (C) 2000 Pavel Machek <pavel at suse.cz>
++ * Copyright (C) 2002,2003 Andi Kleen <ak at suse.de>
++ *
++ * Jun Nakajima <jun.nakajima at intel.com>
++ * Modified for Xen.
++ */
++
++#include <linux/signal.h>
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/string.h>
++#include <linux/types.h>
++#include <linux/ptrace.h>
++#include <linux/mman.h>
++#include <linux/mm.h>
++#include <linux/swap.h>
++#include <linux/smp.h>
++#include <linux/init.h>
++#include <linux/pagemap.h>
++#include <linux/bootmem.h>
++#include <linux/proc_fs.h>
++#include <linux/pci.h>
++#include <linux/poison.h>
++#include <linux/dma-mapping.h>
++#include <linux/module.h>
++#include <linux/memory_hotplug.h>
++
++#include <asm/processor.h>
++#include <asm/system.h>
++#include <asm/uaccess.h>
++#include <asm/pgtable.h>
++#include <asm/pgalloc.h>
++#include <asm/dma.h>
++#include <asm/fixmap.h>
++#include <asm/e820.h>
++#include <asm/apic.h>
++#include <asm/tlb.h>
++#include <asm/mmu_context.h>
++#include <asm/proto.h>
++#include <asm/smp.h>
++#include <asm/sections.h>
++
++#include <xen/features.h>
++
++#ifndef Dprintk
++#define Dprintk(x...)
++#endif
++
++struct dma_mapping_ops* dma_ops;
++EXPORT_SYMBOL(dma_ops);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++unsigned int __kernel_page_user;
++EXPORT_SYMBOL(__kernel_page_user);
++#endif
++
++int after_bootmem;
++
++extern unsigned long *contiguous_bitmap;
++
++static unsigned long dma_reserve __initdata;
++
++DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
++extern unsigned long start_pfn;
++
++/*
++ * Use this until direct mapping is established, i.e. before __va() is
++ * available in init_memory_mapping().
++ */
++
++#define addr_to_page(addr, page) \
++ (addr) &= PHYSICAL_PAGE_MASK; \
++ (page) = ((unsigned long *) ((unsigned long) \
++ (((mfn_to_pfn((addr) >> PAGE_SHIFT)) << PAGE_SHIFT) + \
++ __START_KERNEL_map)))
++
++static void __meminit early_make_page_readonly(void *va, unsigned int feature)
++{
++ unsigned long addr, _va = (unsigned long)va;
++ pte_t pte, *ptep;
++ unsigned long *page = (unsigned long *) init_level4_pgt;
++
++ BUG_ON(after_bootmem);
++
++ if (xen_feature(feature))
++ return;
++
++ addr = (unsigned long) page[pgd_index(_va)];
++ addr_to_page(addr, page);
++
++ addr = page[pud_index(_va)];
++ addr_to_page(addr, page);
++
++ addr = page[pmd_index(_va)];
++ addr_to_page(addr, page);
++
++ ptep = (pte_t *) &page[pte_index(_va)];
++
++ pte.pte = ptep->pte & ~_PAGE_RW;
++ if (HYPERVISOR_update_va_mapping(_va, pte, 0))
++ BUG();
++}
++
++static void __make_page_readonly(void *va)
++{
++ pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t pte, *ptep;
++ unsigned long addr = (unsigned long) va;
++
++ pgd = pgd_offset_k(addr);
++ pud = pud_offset(pgd, addr);
++ pmd = pmd_offset(pud, addr);
++ ptep = pte_offset_kernel(pmd, addr);
++
++ pte.pte = ptep->pte & ~_PAGE_RW;
++ if (HYPERVISOR_update_va_mapping(addr, pte, 0))
++ xen_l1_entry_update(ptep, pte); /* fallback */
++
++ if ((addr >= VMALLOC_START) && (addr < VMALLOC_END))
++ __make_page_readonly(__va(pte_pfn(pte) << PAGE_SHIFT));
++}
++
++static void __make_page_writable(void *va)
++{
++ pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t pte, *ptep;
++ unsigned long addr = (unsigned long) va;
++
++ pgd = pgd_offset_k(addr);
++ pud = pud_offset(pgd, addr);
++ pmd = pmd_offset(pud, addr);
++ ptep = pte_offset_kernel(pmd, addr);
++
++ pte.pte = ptep->pte | _PAGE_RW;
++ if (HYPERVISOR_update_va_mapping(addr, pte, 0))
++ xen_l1_entry_update(ptep, pte); /* fallback */
++
++ if ((addr >= VMALLOC_START) && (addr < VMALLOC_END))
++ __make_page_writable(__va(pte_pfn(pte) << PAGE_SHIFT));
++}
++
++void make_page_readonly(void *va, unsigned int feature)
++{
++ if (!xen_feature(feature))
++ __make_page_readonly(va);
++}
++
++void make_page_writable(void *va, unsigned int feature)
++{
++ if (!xen_feature(feature))
++ __make_page_writable(va);
++}
++
++void make_pages_readonly(void *va, unsigned nr, unsigned int feature)
++{
++ if (xen_feature(feature))
++ return;
++
++ while (nr-- != 0) {
++ __make_page_readonly(va);
++ va = (void*)((unsigned long)va + PAGE_SIZE);
++ }
++}
++
++void make_pages_writable(void *va, unsigned nr, unsigned int feature)
++{
++ if (xen_feature(feature))
++ return;
++
++ while (nr-- != 0) {
++ __make_page_writable(va);
++ va = (void*)((unsigned long)va + PAGE_SIZE);
++ }
++}
++
++/*
++ * NOTE: pagetable_init alloc all the fixmap pagetables contiguous on the
++ * physical space so we can cache the place of the first one and move
++ * around without checking the pgd every time.
++ */
++
++void show_mem(void)
++{
++ long i, total = 0, reserved = 0;
++ long shared = 0, cached = 0;
++ pg_data_t *pgdat;
++ struct page *page;
++
++ printk(KERN_INFO "Mem-info:\n");
++ show_free_areas();
++ printk(KERN_INFO "Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10));
++
++ for_each_online_pgdat(pgdat) {
++ for (i = 0; i < pgdat->node_spanned_pages; ++i) {
++ page = pfn_to_page(pgdat->node_start_pfn + i);
++ total++;
++ if (PageReserved(page))
++ reserved++;
++ else if (PageSwapCache(page))
++ cached++;
++ else if (page_count(page))
++ shared += page_count(page) - 1;
++ }
++ }
++ printk(KERN_INFO "%lu pages of RAM\n", total);
++ printk(KERN_INFO "%lu reserved pages\n",reserved);
++ printk(KERN_INFO "%lu pages shared\n",shared);
++ printk(KERN_INFO "%lu pages swap cached\n",cached);
++}
++
++
++static __init void *spp_getpage(void)
++{
++ void *ptr;
++ if (after_bootmem)
++ ptr = (void *) get_zeroed_page(GFP_ATOMIC);
++ else if (start_pfn < table_end) {
++ ptr = __va(start_pfn << PAGE_SHIFT);
++ start_pfn++;
++ memset(ptr, 0, PAGE_SIZE);
++ } else
++ ptr = alloc_bootmem_pages(PAGE_SIZE);
++ if (!ptr || ((unsigned long)ptr & ~PAGE_MASK))
++ panic("set_pte_phys: cannot allocate page data %s\n", after_bootmem?"after bootmem":"");
++
++ Dprintk("spp_getpage %p\n", ptr);
++ return ptr;
++}
++
++#define pgd_offset_u(address) (__user_pgd(init_level4_pgt) + pgd_index(address))
++#define pud_offset_u(address) (level3_user_pgt + pud_index(address))
++
++static __init void set_pte_phys(unsigned long vaddr,
++ unsigned long phys, pgprot_t prot, int user_mode)
++{
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte, new_pte;
++
++ Dprintk("set_pte_phys %lx to %lx\n", vaddr, phys);
++
++ pgd = (user_mode ? pgd_offset_u(vaddr) : pgd_offset_k(vaddr));
++ if (pgd_none(*pgd)) {
++ printk("PGD FIXMAP MISSING, it should be setup in head.S!\n");
++ return;
++ }
++ pud = (user_mode ? pud_offset_u(vaddr) : pud_offset(pgd, vaddr));
++ if (pud_none(*pud)) {
++ pmd = (pmd_t *) spp_getpage();
++ make_page_readonly(pmd, XENFEAT_writable_page_tables);
++ set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE | _PAGE_USER));
++ if (pmd != pmd_offset(pud, 0)) {
++ printk("PAGETABLE BUG #01! %p <-> %p\n", pmd, pmd_offset(pud,0));
++ return;
++ }
++ }
++ pmd = pmd_offset(pud, vaddr);
++ if (pmd_none(*pmd)) {
++ pte = (pte_t *) spp_getpage();
++ make_page_readonly(pte, XENFEAT_writable_page_tables);
++ set_pmd(pmd, __pmd(__pa(pte) | _KERNPG_TABLE | _PAGE_USER));
++ if (pte != pte_offset_kernel(pmd, 0)) {
++ printk("PAGETABLE BUG #02!\n");
++ return;
++ }
++ }
++ if (pgprot_val(prot))
++ new_pte = pfn_pte(phys >> PAGE_SHIFT, prot);
++ else
++ new_pte = __pte(0);
++
++ pte = pte_offset_kernel(pmd, vaddr);
++ if (!pte_none(*pte) &&
++ __pte_val(*pte) != (__pte_val(new_pte) & __supported_pte_mask))
++ pte_ERROR(*pte);
++ set_pte(pte, new_pte);
++
++ /*
++ * It's enough to flush this one mapping.
++ * (PGE mappings get flushed as well)
++ */
++ __flush_tlb_one(vaddr);
++}
++
++static __init void set_pte_phys_ma(unsigned long vaddr,
++ unsigned long phys, pgprot_t prot)
++{
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte, new_pte;
++
++ Dprintk("set_pte_phys %lx to %lx\n", vaddr, phys);
++
++ pgd = pgd_offset_k(vaddr);
++ if (pgd_none(*pgd)) {
++ printk("PGD FIXMAP MISSING, it should be setup in head.S!\n");
++ return;
++ }
++ pud = pud_offset(pgd, vaddr);
++ if (pud_none(*pud)) {
++
++ pmd = (pmd_t *) spp_getpage();
++ make_page_readonly(pmd, XENFEAT_writable_page_tables);
++ set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE | _PAGE_USER));
++ if (pmd != pmd_offset(pud, 0)) {
++ printk("PAGETABLE BUG #01! %p <-> %p\n", pmd, pmd_offset(pud,0));
++ return;
++ }
++ }
++ pmd = pmd_offset(pud, vaddr);
++ if (pmd_none(*pmd)) {
++ pte = (pte_t *) spp_getpage();
++ make_page_readonly(pte, XENFEAT_writable_page_tables);
++ set_pmd(pmd, __pmd(__pa(pte) | _KERNPG_TABLE | _PAGE_USER));
++ if (pte != pte_offset_kernel(pmd, 0)) {
++ printk("PAGETABLE BUG #02!\n");
++ return;
++ }
++ }
++ new_pte = pfn_pte_ma(phys >> PAGE_SHIFT, prot);
++
++ pte = pte_offset_kernel(pmd, vaddr);
++ if (!pte_none(*pte) &&
++#ifdef CONFIG_ACPI
++ /* __acpi_map_table() fails to properly call clear_fixmap() */
++ (vaddr < __fix_to_virt(FIX_ACPI_END) ||
++ vaddr > __fix_to_virt(FIX_ACPI_BEGIN)) &&
++#endif
++ __pte_val(*pte) != (__pte_val(new_pte) & __supported_pte_mask))
++ pte_ERROR(*pte);
++ set_pte(pte, new_pte);
++
++ /*
++ * It's enough to flush this one mapping.
++ * (PGE mappings get flushed as well)
++ */
++ __flush_tlb_one(vaddr);
++}
++
++/* NOTE: this is meant to be run only at boot */
++void __init
++__set_fixmap (enum fixed_addresses idx, unsigned long phys, pgprot_t prot)
++{
++ unsigned long address = __fix_to_virt(idx);
++
++ if (idx >= __end_of_fixed_addresses) {
++ printk("Invalid __set_fixmap\n");
++ return;
++ }
++ switch (idx) {
++ case VSYSCALL_LAST_PAGE ... VSYSCALL_FIRST_PAGE:
++ set_pte_phys(address, phys, prot, 0);
++ set_pte_phys(address, phys, prot, 1);
++ break;
++ default:
++ set_pte_phys_ma(address, phys, prot);
++ break;
++ }
++}
++
++unsigned long __initdata table_start, table_end;
++
++static __meminit void *alloc_static_page(unsigned long *phys)
++{
++ unsigned long va = (start_pfn << PAGE_SHIFT) + __START_KERNEL_map;
++
++ if (after_bootmem) {
++ void *adr = (void *)get_zeroed_page(GFP_ATOMIC);
++
++ *phys = __pa(adr);
++ return adr;
++ }
++
++ *phys = start_pfn << PAGE_SHIFT;
++ start_pfn++;
++ memset((void *)va, 0, PAGE_SIZE);
++ return (void *)va;
++}
++
++#define PTE_SIZE PAGE_SIZE
++
++static inline int make_readonly(unsigned long paddr)
++{
++ extern char __vsyscall_0;
++ int readonly = 0;
++
++ /* Make new page tables read-only. */
++ if (!xen_feature(XENFEAT_writable_page_tables)
++ && (paddr >= (table_start << PAGE_SHIFT))
++ && (paddr < (table_end << PAGE_SHIFT)))
++ readonly = 1;
++ /* Make old page tables read-only. */
++ if (!xen_feature(XENFEAT_writable_page_tables)
++ && (paddr >= (xen_start_info->pt_base - __START_KERNEL_map))
++ && (paddr < (start_pfn << PAGE_SHIFT)))
++ readonly = 1;
++
++ /*
++ * No need for writable mapping of kernel image. This also ensures that
++ * page and descriptor tables embedded inside don't have writable
++ * mappings. Exclude the vsyscall area here, allowing alternative
++ * instruction patching to work.
++ */
++ if ((paddr >= __pa_symbol(&_text)) && (paddr < __pa_symbol(&_end))
++ && !(paddr >= __pa_symbol(&__vsyscall_0)
++ && paddr < __pa_symbol(&__vsyscall_0) + PAGE_SIZE))
++ readonly = 1;
++
++ return readonly;
++}
++
++#ifndef CONFIG_XEN
++/* Must run before zap_low_mappings */
++__init void *early_ioremap(unsigned long addr, unsigned long size)
++{
++ unsigned long map = round_down(addr, LARGE_PAGE_SIZE);
++
++ /* actually usually some more */
++ if (size >= LARGE_PAGE_SIZE) {
++ printk("SMBIOS area too long %lu\n", size);
++ return NULL;
++ }
++ set_pmd(temp_mappings[0].pmd, __pmd(map | _KERNPG_TABLE | _PAGE_PSE));
++ map += LARGE_PAGE_SIZE;
++ set_pmd(temp_mappings[1].pmd, __pmd(map | _KERNPG_TABLE | _PAGE_PSE));
++ __flush_tlb();
++ return temp_mappings[0].address + (addr & (LARGE_PAGE_SIZE-1));
++}
++
++/* To avoid virtual aliases later */
++__init void early_iounmap(void *addr, unsigned long size)
++{
++ if ((void *)round_down((unsigned long)addr, LARGE_PAGE_SIZE) != temp_mappings[0].address)
++ printk("early_iounmap: bad address %p\n", addr);
++ set_pmd(temp_mappings[0].pmd, __pmd(0));
++ set_pmd(temp_mappings[1].pmd, __pmd(0));
++ __flush_tlb();
++}
++#endif
++
++static void __meminit
++phys_pmd_init(pmd_t *pmd, unsigned long address, unsigned long end)
++{
++ int i, k;
++
++ for (i = 0; i < PTRS_PER_PMD; pmd++, i++) {
++ unsigned long pte_phys;
++ pte_t *pte, *pte_save;
++
++ if (address >= end)
++ break;
++ pte = alloc_static_page(&pte_phys);
++ pte_save = pte;
++ for (k = 0; k < PTRS_PER_PTE; pte++, k++, address += PTE_SIZE) {
++ unsigned long pteval = address | _PAGE_NX | _KERNPG_TABLE;
++
++ if (address >= (after_bootmem
++ ? end
++ : xen_start_info->nr_pages << PAGE_SHIFT))
++ pteval = 0;
++ else if (make_readonly(address))
++ pteval &= ~_PAGE_RW;
++ set_pte(pte, __pte(pteval & __supported_pte_mask));
++ }
++ if (!after_bootmem) {
++ early_make_page_readonly(pte_save, XENFEAT_writable_page_tables);
++ *pmd = __pmd(pte_phys | _KERNPG_TABLE);
++ } else {
++ make_page_readonly(pte_save, XENFEAT_writable_page_tables);
++ set_pmd(pmd, __pmd(pte_phys | _KERNPG_TABLE));
++ }
++ }
++}
++
++static void __meminit
++phys_pmd_update(pud_t *pud, unsigned long address, unsigned long end)
++{
++ pmd_t *pmd = pmd_offset(pud, (unsigned long)__va(address));
++
++ if (pmd_none(*pmd)) {
++ spin_lock(&init_mm.page_table_lock);
++ phys_pmd_init(pmd, address, end);
++ spin_unlock(&init_mm.page_table_lock);
++ __flush_tlb_all();
++ }
++}
++
++static void __meminit phys_pud_init(pud_t *pud, unsigned long address, unsigned long end)
++{
++ long i = pud_index(address);
++
++ pud = pud + i;
++
++ if (after_bootmem && pud_val(*pud)) {
++ phys_pmd_update(pud, address, end);
++ return;
++ }
++
++ for (; i < PTRS_PER_PUD; pud++, i++) {
++ unsigned long paddr, pmd_phys;
++ pmd_t *pmd;
++
++ paddr = (address & PGDIR_MASK) + i*PUD_SIZE;
++ if (paddr >= end)
++ break;
++
++ pmd = alloc_static_page(&pmd_phys);
++
++ spin_lock(&init_mm.page_table_lock);
++ *pud = __pud(pmd_phys | _KERNPG_TABLE);
++ phys_pmd_init(pmd, paddr, end);
++ spin_unlock(&init_mm.page_table_lock);
++
++ early_make_page_readonly(pmd, XENFEAT_writable_page_tables);
++ }
++ __flush_tlb();
++}
++
++void __init xen_init_pt(void)
++{
++ unsigned long addr, *page;
++
++ /* Find the initial pte page that was built for us. */
++ page = (unsigned long *)xen_start_info->pt_base;
++ addr = page[pgd_index(__START_KERNEL_map)];
++ addr_to_page(addr, page);
++ addr = page[pud_index(__START_KERNEL_map)];
++ addr_to_page(addr, page);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ /* On Xen 3.0.2 and older we may need to explicitly specify _PAGE_USER
++ in kernel PTEs. We check that here. */
++ if (HYPERVISOR_xen_version(XENVER_version, NULL) <= 0x30000) {
++ unsigned long *pg;
++ pte_t pte;
++
++ /* Mess with the initial mapping of page 0. It's not needed. */
++ BUILD_BUG_ON(__START_KERNEL <= __START_KERNEL_map);
++ addr = page[pmd_index(__START_KERNEL_map)];
++ addr_to_page(addr, pg);
++ pte.pte = pg[pte_index(__START_KERNEL_map)];
++ BUG_ON(!(pte.pte & _PAGE_PRESENT));
++
++ /* If _PAGE_USER isn't set, we obviously do not need it. */
++ if (pte.pte & _PAGE_USER) {
++ /* _PAGE_USER is needed, but is it set implicitly? */
++ pte.pte &= ~_PAGE_USER;
++ if ((HYPERVISOR_update_va_mapping(__START_KERNEL_map,
++ pte, 0) != 0) ||
++ !(pg[pte_index(__START_KERNEL_map)] & _PAGE_USER))
++ /* We need to explicitly specify _PAGE_USER. */
++ __kernel_page_user = _PAGE_USER;
++ }
++ }
++#endif
++
++ /* Construct mapping of initial pte page in our own directories. */
++ init_level4_pgt[pgd_index(__START_KERNEL_map)] =
++ __pgd(__pa_symbol(level3_kernel_pgt) | _PAGE_TABLE);
++ level3_kernel_pgt[pud_index(__START_KERNEL_map)] =
++ __pud(__pa_symbol(level2_kernel_pgt) | _PAGE_TABLE);
++ memcpy(level2_kernel_pgt, page, PAGE_SIZE);
++
++ __user_pgd(init_level4_pgt)[pgd_index(VSYSCALL_START)] =
++ __pgd(__pa_symbol(level3_user_pgt) | _PAGE_TABLE);
++
++ early_make_page_readonly(init_level4_pgt,
++ XENFEAT_writable_page_tables);
++ early_make_page_readonly(__user_pgd(init_level4_pgt),
++ XENFEAT_writable_page_tables);
++ early_make_page_readonly(level3_kernel_pgt,
++ XENFEAT_writable_page_tables);
++ early_make_page_readonly(level3_user_pgt,
++ XENFEAT_writable_page_tables);
++ early_make_page_readonly(level2_kernel_pgt,
++ XENFEAT_writable_page_tables);
++
++ if (!xen_feature(XENFEAT_writable_page_tables)) {
++ xen_pgd_pin(__pa_symbol(init_level4_pgt));
++ xen_pgd_pin(__pa_symbol(__user_pgd(init_level4_pgt)));
++ }
++}
++
++static void __init extend_init_mapping(unsigned long tables_space)
++{
++ unsigned long va = __START_KERNEL_map;
++ unsigned long phys, addr, *pte_page;
++ pmd_t *pmd;
++ pte_t *pte, new_pte;
++ unsigned long *page = (unsigned long *)init_level4_pgt;
++
++ addr = page[pgd_index(va)];
++ addr_to_page(addr, page);
++ addr = page[pud_index(va)];
++ addr_to_page(addr, page);
++
++ /* Kill mapping of low 1MB. */
++ while (va < (unsigned long)&_text) {
++ if (HYPERVISOR_update_va_mapping(va, __pte_ma(0), 0))
++ BUG();
++ va += PAGE_SIZE;
++ }
++
++ /* Ensure init mappings cover kernel text/data and initial tables. */
++ while (va < (__START_KERNEL_map
++ + (start_pfn << PAGE_SHIFT)
++ + tables_space)) {
++ pmd = (pmd_t *)&page[pmd_index(va)];
++ if (pmd_none(*pmd)) {
++ pte_page = alloc_static_page(&phys);
++ early_make_page_readonly(
++ pte_page, XENFEAT_writable_page_tables);
++ set_pmd(pmd, __pmd(phys | _KERNPG_TABLE));
++ } else {
++ addr = page[pmd_index(va)];
++ addr_to_page(addr, pte_page);
++ }
++ pte = (pte_t *)&pte_page[pte_index(va)];
++ if (pte_none(*pte)) {
++ new_pte = pfn_pte(
++ (va - __START_KERNEL_map) >> PAGE_SHIFT,
++ __pgprot(_KERNPG_TABLE));
++ xen_l1_entry_update(pte, new_pte);
++ }
++ va += PAGE_SIZE;
++ }
++
++ /* Finally, blow away any spurious initial mappings. */
++ while (1) {
++ pmd = (pmd_t *)&page[pmd_index(va)];
++ if (pmd_none(*pmd))
++ break;
++ if (HYPERVISOR_update_va_mapping(va, __pte_ma(0), 0))
++ BUG();
++ va += PAGE_SIZE;
++ }
++}
++
++static void __init find_early_table_space(unsigned long end)
++{
++ unsigned long puds, pmds, ptes, tables;
++
++ puds = (end + PUD_SIZE - 1) >> PUD_SHIFT;
++ pmds = (end + PMD_SIZE - 1) >> PMD_SHIFT;
++ ptes = (end + PTE_SIZE - 1) >> PAGE_SHIFT;
++
++ tables = round_up(puds * 8, PAGE_SIZE) +
++ round_up(pmds * 8, PAGE_SIZE) +
++ round_up(ptes * 8, PAGE_SIZE);
++
++ extend_init_mapping(tables);
++
++ table_start = start_pfn;
++ table_end = table_start + (tables>>PAGE_SHIFT);
++
++ early_printk("kernel direct mapping tables up to %lx @ %lx-%lx\n",
++ end, table_start << PAGE_SHIFT,
++ (table_start << PAGE_SHIFT) + tables);
++}
++
++static void xen_finish_init_mapping(void)
++{
++ unsigned long i, start, end;
++
++ /* Re-vector virtual addresses pointing into the initial
++ mapping to the just-established permanent ones. */
++ xen_start_info = __va(__pa(xen_start_info));
++ xen_start_info->pt_base = (unsigned long)
++ __va(__pa(xen_start_info->pt_base));
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ phys_to_machine_mapping =
++ __va(__pa(xen_start_info->mfn_list));
++ xen_start_info->mfn_list = (unsigned long)
++ phys_to_machine_mapping;
++ }
++ if (xen_start_info->mod_start)
++ xen_start_info->mod_start = (unsigned long)
++ __va(__pa(xen_start_info->mod_start));
++
++ /* Destroy the Xen-created mappings beyond the kernel image as
++ * well as the temporary mappings created above. Prevents
++ * overlap with modules area (if init mapping is very big).
++ */
++ start = PAGE_ALIGN((unsigned long)_end);
++ end = __START_KERNEL_map + (table_end << PAGE_SHIFT);
++ for (; start < end; start += PAGE_SIZE)
++ if (HYPERVISOR_update_va_mapping(start, __pte_ma(0), 0))
++ BUG();
++
++ /* Allocate pte's for initial fixmaps from 'start_pfn' allocator. */
++ table_end = ~0UL;
++
++ /*
++ * Prefetch pte's for the bt_ioremap() area. It gets used before the
++ * boot-time allocator is online, so allocate-on-demand would fail.
++ */
++ for (i = FIX_BTMAP_END; i <= FIX_BTMAP_BEGIN; i++)
++ __set_fixmap(i, 0, __pgprot(0));
++
++ /* Switch to the real shared_info page, and clear the dummy page. */
++ set_fixmap(FIX_SHARED_INFO, xen_start_info->shared_info);
++ HYPERVISOR_shared_info = (shared_info_t *)fix_to_virt(FIX_SHARED_INFO);
++ memset(empty_zero_page, 0, sizeof(empty_zero_page));
++
++ /* Set up mapping of lowest 1MB of physical memory. */
++ for (i = 0; i < NR_FIX_ISAMAPS; i++)
++ if (is_initial_xendomain())
++ set_fixmap(FIX_ISAMAP_BEGIN - i, i * PAGE_SIZE);
++ else
++ __set_fixmap(FIX_ISAMAP_BEGIN - i,
++ virt_to_mfn(empty_zero_page)
++ << PAGE_SHIFT,
++ PAGE_KERNEL_RO);
++
++ /* Disable the 'start_pfn' allocator. */
++ table_end = start_pfn;
++}
++
++/* Setup the direct mapping of the physical memory at PAGE_OFFSET.
++ This runs before bootmem is initialized and gets pages directly from the
++ physical memory. To access them they are temporarily mapped. */
++void __meminit init_memory_mapping(unsigned long start, unsigned long end)
++{
++ unsigned long next;
++
++ Dprintk("init_memory_mapping\n");
++
++ /*
++ * Find space for the kernel direct mapping tables.
++ * Later we should allocate these tables in the local node of the memory
++ * mapped. Unfortunately this is done currently before the nodes are
++ * discovered.
++ */
++ if (!after_bootmem)
++ find_early_table_space(end);
++
++ start = (unsigned long)__va(start);
++ end = (unsigned long)__va(end);
++
++ for (; start < end; start = next) {
++ unsigned long pud_phys;
++ pgd_t *pgd = pgd_offset_k(start);
++ pud_t *pud;
++
++ if (after_bootmem)
++ pud = pud_offset(pgd, start & PGDIR_MASK);
++ else
++ pud = alloc_static_page(&pud_phys);
++ next = start + PGDIR_SIZE;
++ if (next > end)
++ next = end;
++ phys_pud_init(pud, __pa(start), __pa(next));
++ if (!after_bootmem) {
++ early_make_page_readonly(pud, XENFEAT_writable_page_tables);
++ set_pgd(pgd_offset_k(start), mk_kernel_pgd(pud_phys));
++ }
++ }
++
++ if (!after_bootmem) {
++ BUG_ON(start_pfn != table_end);
++ xen_finish_init_mapping();
++ }
++
++ __flush_tlb_all();
++}
++
++void __cpuinit zap_low_mappings(int cpu)
++{
++ /* this is not required for Xen */
++#if 0
++ swap_low_mappings();
++#endif
++}
++
++/* Compute zone sizes for the DMA and DMA32 zones in a node. */
++__init void
++size_zones(unsigned long *z, unsigned long *h,
++ unsigned long start_pfn, unsigned long end_pfn)
++{
++ int i;
++ unsigned long w;
++
++ for (i = 0; i < MAX_NR_ZONES; i++)
++ z[i] = 0;
++
++ if (start_pfn < MAX_DMA_PFN)
++ z[ZONE_DMA] = MAX_DMA_PFN - start_pfn;
++ if (start_pfn < MAX_DMA32_PFN) {
++ unsigned long dma32_pfn = MAX_DMA32_PFN;
++ if (dma32_pfn > end_pfn)
++ dma32_pfn = end_pfn;
++ z[ZONE_DMA32] = dma32_pfn - start_pfn;
++ }
++ z[ZONE_NORMAL] = end_pfn - start_pfn;
++
++ /* Remove lower zones from higher ones. */
++ w = 0;
++ for (i = 0; i < MAX_NR_ZONES; i++) {
++ if (z[i])
++ z[i] -= w;
++ w += z[i];
++ }
++
++ /* Compute holes */
++ w = start_pfn;
++ for (i = 0; i < MAX_NR_ZONES; i++) {
++ unsigned long s = w;
++ w += z[i];
++ h[i] = e820_hole_size(s, w);
++ }
++
++ /* Add the space pace needed for mem_map to the holes too. */
++ for (i = 0; i < MAX_NR_ZONES; i++)
++ h[i] += (z[i] * sizeof(struct page)) / PAGE_SIZE;
++
++ /* The 16MB DMA zone has the kernel and other misc mappings.
++ Account them too */
++ if (h[ZONE_DMA]) {
++ h[ZONE_DMA] += dma_reserve;
++ if (h[ZONE_DMA] >= z[ZONE_DMA]) {
++ printk(KERN_WARNING
++ "Kernel too large and filling up ZONE_DMA?\n");
++ h[ZONE_DMA] = z[ZONE_DMA];
++ }
++ }
++}
++
++#ifndef CONFIG_NUMA
++void __init paging_init(void)
++{
++ unsigned long zones[MAX_NR_ZONES], holes[MAX_NR_ZONES];
++
++ memory_present(0, 0, end_pfn);
++ sparse_init();
++ size_zones(zones, holes, 0, end_pfn);
++ free_area_init_node(0, NODE_DATA(0), zones,
++ __pa(PAGE_OFFSET) >> PAGE_SHIFT, holes);
++
++ init_mm.context.pinned = 1;
++}
++#endif
++
++/* Unmap a kernel mapping if it exists. This is useful to avoid prefetches
++ from the CPU leading to inconsistent cache lines. address and size
++ must be aligned to 2MB boundaries.
++ Does nothing when the mapping doesn't exist. */
++void __init clear_kernel_mapping(unsigned long address, unsigned long size)
++{
++ unsigned long end = address + size;
++
++ BUG_ON(address & ~LARGE_PAGE_MASK);
++ BUG_ON(size & ~LARGE_PAGE_MASK);
++
++ for (; address < end; address += LARGE_PAGE_SIZE) {
++ pgd_t *pgd = pgd_offset_k(address);
++ pud_t *pud;
++ pmd_t *pmd;
++ if (pgd_none(*pgd))
++ continue;
++ pud = pud_offset(pgd, address);
++ if (pud_none(*pud))
++ continue;
++ pmd = pmd_offset(pud, address);
++ if (!pmd || pmd_none(*pmd))
++ continue;
++ if (0 == (__pmd_val(*pmd) & _PAGE_PSE)) {
++ /* Could handle this, but it should not happen currently. */
++ printk(KERN_ERR
++ "clear_kernel_mapping: mapping has been split. will leak memory\n");
++ pmd_ERROR(*pmd);
++ }
++ set_pmd(pmd, __pmd(0));
++ }
++ __flush_tlb_all();
++}
++
++/*
++ * Memory hotplug specific functions
++ */
++void online_page(struct page *page)
++{
++ ClearPageReserved(page);
++ init_page_count(page);
++ __free_page(page);
++ totalram_pages++;
++ num_physpages++;
++}
++
++#ifdef CONFIG_MEMORY_HOTPLUG
++/*
++ * XXX: memory_add_physaddr_to_nid() is to find node id from physical address
++ * via probe interface of sysfs. If acpi notifies hot-add event, then it
++ * can tell node id by searching dsdt. But, probe interface doesn't have
++ * node id. So, return 0 as node id at this time.
++ */
++#ifdef CONFIG_NUMA
++int memory_add_physaddr_to_nid(u64 start)
++{
++ return 0;
++}
++#endif
++
++/*
++ * Memory is added always to NORMAL zone. This means you will never get
++ * additional DMA/DMA32 memory.
++ */
++int arch_add_memory(int nid, u64 start, u64 size)
++{
++ struct pglist_data *pgdat = NODE_DATA(nid);
++ struct zone *zone = pgdat->node_zones + MAX_NR_ZONES-2;
++ unsigned long start_pfn = start >> PAGE_SHIFT;
++ unsigned long nr_pages = size >> PAGE_SHIFT;
++ int ret;
++
++ ret = __add_pages(zone, start_pfn, nr_pages);
++ if (ret)
++ goto error;
++
++ init_memory_mapping(start, (start + size -1));
++
++ return ret;
++error:
++ printk("%s: Problem encountered in __add_pages!\n", __func__);
++ return ret;
++}
++EXPORT_SYMBOL_GPL(arch_add_memory);
++
++int remove_memory(u64 start, u64 size)
++{
++ return -EINVAL;
++}
++EXPORT_SYMBOL_GPL(remove_memory);
++
++#else /* CONFIG_MEMORY_HOTPLUG */
++/*
++ * Memory Hotadd without sparsemem. The mem_maps have been allocated in advance,
++ * just online the pages.
++ */
++int __add_pages(struct zone *z, unsigned long start_pfn, unsigned long nr_pages)
++{
++ int err = -EIO;
++ unsigned long pfn;
++ unsigned long total = 0, mem = 0;
++ for (pfn = start_pfn; pfn < start_pfn + nr_pages; pfn++) {
++ if (pfn_valid(pfn)) {
++ online_page(pfn_to_page(pfn));
++ err = 0;
++ mem++;
++ }
++ total++;
++ }
++ if (!err) {
++ z->spanned_pages += total;
++ z->present_pages += mem;
++ z->zone_pgdat->node_spanned_pages += total;
++ z->zone_pgdat->node_present_pages += mem;
++ }
++ return err;
++}
++#endif /* CONFIG_MEMORY_HOTPLUG */
++
++static struct kcore_list kcore_mem, kcore_vmalloc, kcore_kernel, kcore_modules,
++ kcore_vsyscall;
++
++void __init mem_init(void)
++{
++ long codesize, reservedpages, datasize, initsize;
++ unsigned long pfn;
++
++ contiguous_bitmap = alloc_bootmem_low_pages(
++ (end_pfn + 2*BITS_PER_LONG) >> 3);
++ BUG_ON(!contiguous_bitmap);
++ memset(contiguous_bitmap, 0, (end_pfn + 2*BITS_PER_LONG) >> 3);
++
++ pci_iommu_alloc();
++
++ /* How many end-of-memory variables you have, grandma! */
++ max_low_pfn = end_pfn;
++ max_pfn = end_pfn;
++ num_physpages = end_pfn;
++ high_memory = (void *) __va(end_pfn * PAGE_SIZE);
++
++ /* clear the zero-page */
++ memset(empty_zero_page, 0, PAGE_SIZE);
++
++ reservedpages = 0;
++
++ /* this will put all low memory onto the freelists */
++#ifdef CONFIG_NUMA
++ totalram_pages = numa_free_all_bootmem();
++#else
++ totalram_pages = free_all_bootmem();
++#endif
++ /* XEN: init and count pages outside initial allocation. */
++ for (pfn = xen_start_info->nr_pages; pfn < max_pfn; pfn++) {
++ ClearPageReserved(pfn_to_page(pfn));
++ init_page_count(pfn_to_page(pfn));
++ totalram_pages++;
++ }
++ reservedpages = end_pfn - totalram_pages - e820_hole_size(0, end_pfn);
++
++ after_bootmem = 1;
++
++ codesize = (unsigned long) &_etext - (unsigned long) &_text;
++ datasize = (unsigned long) &_edata - (unsigned long) &_etext;
++ initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin;
++
++ /* Register memory areas for /proc/kcore */
++ kclist_add(&kcore_mem, __va(0), max_low_pfn << PAGE_SHIFT);
++ kclist_add(&kcore_vmalloc, (void *)VMALLOC_START,
++ VMALLOC_END-VMALLOC_START);
++ kclist_add(&kcore_kernel, &_stext, _end - _stext);
++ kclist_add(&kcore_modules, (void *)MODULES_VADDR, MODULES_LEN);
++ kclist_add(&kcore_vsyscall, (void *)VSYSCALL_START,
++ VSYSCALL_END - VSYSCALL_START);
++
++ printk("Memory: %luk/%luk available (%ldk kernel code, %ldk reserved, %ldk data, %ldk init)\n",
++ (unsigned long) nr_free_pages() << (PAGE_SHIFT-10),
++ end_pfn << (PAGE_SHIFT-10),
++ codesize >> 10,
++ reservedpages << (PAGE_SHIFT-10),
++ datasize >> 10,
++ initsize >> 10);
++
++#ifndef CONFIG_XEN
++#ifdef CONFIG_SMP
++ /*
++ * Sync boot_level4_pgt mappings with the init_level4_pgt
++ * except for the low identity mappings which are already zapped
++ * in init_level4_pgt. This sync-up is essential for AP's bringup
++ */
++ memcpy(boot_level4_pgt+1, init_level4_pgt+1, (PTRS_PER_PGD-1)*sizeof(pgd_t));
++#endif
++#endif
++}
++
++void free_init_pages(char *what, unsigned long begin, unsigned long end)
++{
++ unsigned long addr;
++
++ if (begin >= end)
++ return;
++
++ printk(KERN_INFO "Freeing %s: %ldk freed\n", what, (end - begin) >> 10);
++ for (addr = begin; addr < end; addr += PAGE_SIZE) {
++ ClearPageReserved(virt_to_page(addr));
++ init_page_count(virt_to_page(addr));
++ memset((void *)(addr & ~(PAGE_SIZE-1)),
++ POISON_FREE_INITMEM, PAGE_SIZE);
++ if (addr >= __START_KERNEL_map) {
++ /* make_readonly() reports all kernel addresses. */
++ __make_page_writable(__va(__pa(addr)));
++ if (HYPERVISOR_update_va_mapping(addr, __pte(0), 0)) {
++ pgd_t *pgd = pgd_offset_k(addr);
++ pud_t *pud = pud_offset(pgd, addr);
++ pmd_t *pmd = pmd_offset(pud, addr);
++ pte_t *pte = pte_offset_kernel(pmd, addr);
++
++ xen_l1_entry_update(pte, __pte(0)); /* fallback */
++ }
++ }
++ free_page(addr);
++ totalram_pages++;
++ }
++}
++
++void free_initmem(void)
++{
++ memset(__initdata_begin, POISON_FREE_INITDATA,
++ __initdata_end - __initdata_begin);
++ free_init_pages("unused kernel memory",
++ (unsigned long)(&__init_begin),
++ (unsigned long)(&__init_end));
++}
++
++#ifdef CONFIG_DEBUG_RODATA
++
++void mark_rodata_ro(void)
++{
++ unsigned long addr = (unsigned long)__start_rodata;
++
++ for (; addr < (unsigned long)__end_rodata; addr += PAGE_SIZE)
++ change_page_attr_addr(addr, 1, PAGE_KERNEL_RO);
++
++ printk ("Write protecting the kernel read-only data: %luk\n",
++ (__end_rodata - __start_rodata) >> 10);
++
++ /*
++ * change_page_attr_addr() requires a global_flush_tlb() call after it.
++ * We do this after the printk so that if something went wrong in the
++ * change, the printk gets out at least to give a better debug hint
++ * of who is the culprit.
++ */
++ global_flush_tlb();
++}
++#endif
++
++#ifdef CONFIG_BLK_DEV_INITRD
++void free_initrd_mem(unsigned long start, unsigned long end)
++{
++ free_init_pages("initrd memory", start, end);
++}
++#endif
++
++void __init reserve_bootmem_generic(unsigned long phys, unsigned len)
++{
++ /* Should check here against the e820 map to avoid double free */
++#ifdef CONFIG_NUMA
++ int nid = phys_to_nid(phys);
++ reserve_bootmem_node(NODE_DATA(nid), phys, len);
++#else
++ reserve_bootmem(phys, len);
++#endif
++ if (phys+len <= MAX_DMA_PFN*PAGE_SIZE)
++ dma_reserve += len / PAGE_SIZE;
++}
++
++int kern_addr_valid(unsigned long addr)
++{
++ unsigned long above = ((long)addr) >> __VIRTUAL_MASK_SHIFT;
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
++
++ if (above != 0 && above != -1UL)
++ return 0;
++
++ pgd = pgd_offset_k(addr);
++ if (pgd_none(*pgd))
++ return 0;
++
++ pud = pud_offset(pgd, addr);
++ if (pud_none(*pud))
++ return 0;
++
++ pmd = pmd_offset(pud, addr);
++ if (pmd_none(*pmd))
++ return 0;
++ if (pmd_large(*pmd))
++ return pfn_valid(pmd_pfn(*pmd));
++
++ pte = pte_offset_kernel(pmd, addr);
++ if (pte_none(*pte))
++ return 0;
++ return pfn_valid(pte_pfn(*pte));
++}
++
++#ifdef CONFIG_SYSCTL
++#include <linux/sysctl.h>
++
++extern int exception_trace, page_fault_trace;
++
++static ctl_table debug_table2[] = {
++ { 99, "exception-trace", &exception_trace, sizeof(int), 0644, NULL,
++ proc_dointvec },
++ { 0, }
++};
++
++static ctl_table debug_root_table2[] = {
++ { .ctl_name = CTL_DEBUG, .procname = "debug", .mode = 0555,
++ .child = debug_table2 },
++ { 0 },
++};
++
++static __init int x8664_sysctl_init(void)
++{
++ register_sysctl_table(debug_root_table2, 1);
++ return 0;
++}
++__initcall(x8664_sysctl_init);
++#endif
++
++/* A pseudo VMAs to allow ptrace access for the vsyscall page. This only
++ covers the 64bit vsyscall page now. 32bit has a real VMA now and does
++ not need special handling anymore. */
++
++static struct vm_area_struct gate_vma = {
++ .vm_start = VSYSCALL_START,
++ .vm_end = VSYSCALL_END,
++ .vm_page_prot = PAGE_READONLY
++};
++
++struct vm_area_struct *get_gate_vma(struct task_struct *tsk)
++{
++#ifdef CONFIG_IA32_EMULATION
++ if (test_tsk_thread_flag(tsk, TIF_IA32))
++ return NULL;
++#endif
++ return &gate_vma;
++}
++
++int in_gate_area(struct task_struct *task, unsigned long addr)
++{
++ struct vm_area_struct *vma = get_gate_vma(task);
++ if (!vma)
++ return 0;
++ return (addr >= vma->vm_start) && (addr < vma->vm_end);
++}
++
++/* Use this when you have no reliable task/vma, typically from interrupt
++ * context. It is less reliable than using the task's vma and may give
++ * false positives.
++ */
++int in_gate_area_no_task(unsigned long addr)
++{
++ return (addr >= VSYSCALL_START) && (addr < VSYSCALL_END);
++}
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/mm/pageattr-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/arch/x86_64/mm/pageattr-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,502 @@
++/*
++ * Copyright 2002 Andi Kleen, SuSE Labs.
++ * Thanks to Ben LaHaise for precious feedback.
++ */
++
++#include <linux/mm.h>
++#include <linux/sched.h>
++#include <linux/highmem.h>
++#include <linux/module.h>
++#include <linux/slab.h>
++#include <asm/uaccess.h>
++#include <asm/processor.h>
++#include <asm/tlbflush.h>
++#include <asm/io.h>
++
++#ifdef CONFIG_XEN
++#include <asm/pgalloc.h>
++#include <asm/mmu_context.h>
++
++LIST_HEAD(mm_unpinned);
++DEFINE_SPINLOCK(mm_unpinned_lock);
++
++static void _pin_lock(struct mm_struct *mm, int lock) {
++ if (lock)
++ spin_lock(&mm->page_table_lock);
++#if NR_CPUS >= CONFIG_SPLIT_PTLOCK_CPUS
++ /* While mm->page_table_lock protects us against insertions and
++ * removals of higher level page table pages, it doesn't protect
++ * against updates of pte-s. Such updates, however, require the
++ * pte pages to be in consistent state (unpinned+writable or
++ * pinned+readonly). The pinning and attribute changes, however
++ * cannot be done atomically, which is why such updates must be
++ * prevented from happening concurrently.
++ * Note that no pte lock can ever elsewhere be acquired nesting
++ * with an already acquired one in the same mm, or with the mm's
++ * page_table_lock already acquired, as that would break in the
++ * non-split case (where all these are actually resolving to the
++ * one page_table_lock). Thus acquiring all of them here is not
++ * going to result in dead locks, and the order of acquires
++ * doesn't matter.
++ */
++ {
++ pgd_t *pgd = mm->pgd;
++ unsigned g;
++
++ for (g = 0; g <= ((TASK_SIZE64-1) / PGDIR_SIZE); g++, pgd++) {
++ pud_t *pud;
++ unsigned u;
++
++ if (pgd_none(*pgd))
++ continue;
++ pud = pud_offset(pgd, 0);
++ for (u = 0; u < PTRS_PER_PUD; u++, pud++) {
++ pmd_t *pmd;
++ unsigned m;
++
++ if (pud_none(*pud))
++ continue;
++ pmd = pmd_offset(pud, 0);
++ for (m = 0; m < PTRS_PER_PMD; m++, pmd++) {
++ spinlock_t *ptl;
++
++ if (pmd_none(*pmd))
++ continue;
++ ptl = pte_lockptr(0, pmd);
++ if (lock)
++ spin_lock(ptl);
++ else
++ spin_unlock(ptl);
++ }
++ }
++ }
++ }
++#endif
++ if (!lock)
++ spin_unlock(&mm->page_table_lock);
++}
++#define pin_lock(mm) _pin_lock(mm, 1)
++#define pin_unlock(mm) _pin_lock(mm, 0)
++
++#define PIN_BATCH 8
++static DEFINE_PER_CPU(multicall_entry_t[PIN_BATCH], pb_mcl);
++
++static inline unsigned int mm_walk_set_prot(void *pt, pgprot_t flags,
++ unsigned int cpu, unsigned int seq)
++{
++ struct page *page = virt_to_page(pt);
++ unsigned long pfn = page_to_pfn(page);
++
++ MULTI_update_va_mapping(per_cpu(pb_mcl, cpu) + seq,
++ (unsigned long)__va(pfn << PAGE_SHIFT),
++ pfn_pte(pfn, flags), 0);
++ if (unlikely(++seq == PIN_BATCH)) {
++ if (unlikely(HYPERVISOR_multicall_check(per_cpu(pb_mcl, cpu),
++ PIN_BATCH, NULL)))
++ BUG();
++ seq = 0;
++ }
++
++ return seq;
++}
++
++static void mm_walk(struct mm_struct *mm, pgprot_t flags)
++{
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
++ int g,u,m;
++ unsigned int cpu, seq;
++ multicall_entry_t *mcl;
++
++ pgd = mm->pgd;
++ cpu = get_cpu();
++
++ /*
++ * Cannot iterate up to USER_PTRS_PER_PGD as these pagetables may not
++ * be the 'current' task's pagetables (e.g., current may be 32-bit,
++ * but the pagetables may be for a 64-bit task).
++ * Subtracting 1 from TASK_SIZE64 means the loop limit is correct
++ * regardless of whether TASK_SIZE64 is a multiple of PGDIR_SIZE.
++ */
++ for (g = 0, seq = 0; g <= ((TASK_SIZE64-1) / PGDIR_SIZE); g++, pgd++) {
++ if (pgd_none(*pgd))
++ continue;
++ pud = pud_offset(pgd, 0);
++ if (PTRS_PER_PUD > 1) /* not folded */
++ seq = mm_walk_set_prot(pud,flags,cpu,seq);
++ for (u = 0; u < PTRS_PER_PUD; u++, pud++) {
++ if (pud_none(*pud))
++ continue;
++ pmd = pmd_offset(pud, 0);
++ if (PTRS_PER_PMD > 1) /* not folded */
++ seq = mm_walk_set_prot(pmd,flags,cpu,seq);
++ for (m = 0; m < PTRS_PER_PMD; m++, pmd++) {
++ if (pmd_none(*pmd))
++ continue;
++ pte = pte_offset_kernel(pmd,0);
++ seq = mm_walk_set_prot(pte,flags,cpu,seq);
++ }
++ }
++ }
++
++ mcl = per_cpu(pb_mcl, cpu);
++ if (unlikely(seq > PIN_BATCH - 2)) {
++ if (unlikely(HYPERVISOR_multicall_check(mcl, seq, NULL)))
++ BUG();
++ seq = 0;
++ }
++ MULTI_update_va_mapping(mcl + seq,
++ (unsigned long)__user_pgd(mm->pgd),
++ pfn_pte(virt_to_phys(__user_pgd(mm->pgd))>>PAGE_SHIFT, flags),
++ 0);
++ MULTI_update_va_mapping(mcl + seq + 1,
++ (unsigned long)mm->pgd,
++ pfn_pte(virt_to_phys(mm->pgd)>>PAGE_SHIFT, flags),
++ UVMF_TLB_FLUSH);
++ if (unlikely(HYPERVISOR_multicall_check(mcl, seq + 2, NULL)))
++ BUG();
++
++ put_cpu();
++}
++
++void mm_pin(struct mm_struct *mm)
++{
++ if (xen_feature(XENFEAT_writable_page_tables))
++ return;
++
++ pin_lock(mm);
++
++ mm_walk(mm, PAGE_KERNEL_RO);
++ xen_pgd_pin(__pa(mm->pgd)); /* kernel */
++ xen_pgd_pin(__pa(__user_pgd(mm->pgd))); /* user */
++ mm->context.pinned = 1;
++ spin_lock(&mm_unpinned_lock);
++ list_del(&mm->context.unpinned);
++ spin_unlock(&mm_unpinned_lock);
++
++ pin_unlock(mm);
++}
++
++void mm_unpin(struct mm_struct *mm)
++{
++ if (xen_feature(XENFEAT_writable_page_tables))
++ return;
++
++ pin_lock(mm);
++
++ xen_pgd_unpin(__pa(mm->pgd));
++ xen_pgd_unpin(__pa(__user_pgd(mm->pgd)));
++ mm_walk(mm, PAGE_KERNEL);
++ mm->context.pinned = 0;
++ spin_lock(&mm_unpinned_lock);
++ list_add(&mm->context.unpinned, &mm_unpinned);
++ spin_unlock(&mm_unpinned_lock);
++
++ pin_unlock(mm);
++}
++
++void mm_pin_all(void)
++{
++ if (xen_feature(XENFEAT_writable_page_tables))
++ return;
++
++ /*
++ * Allow uninterrupted access to the mm_unpinned list. We don't
++ * actually take the mm_unpinned_lock as it is taken inside mm_pin().
++ * All other CPUs must be at a safe point (e.g., in stop_machine
++ * or offlined entirely).
++ */
++ preempt_disable();
++ while (!list_empty(&mm_unpinned))
++ mm_pin(list_entry(mm_unpinned.next, struct mm_struct,
++ context.unpinned));
++ preempt_enable();
++}
++
++void _arch_dup_mmap(struct mm_struct *mm)
++{
++ if (!mm->context.pinned)
++ mm_pin(mm);
++}
++
++void _arch_exit_mmap(struct mm_struct *mm)
++{
++ struct task_struct *tsk = current;
++
++ task_lock(tsk);
++
++ /*
++ * We aggressively remove defunct pgd from cr3. We execute unmap_vmas()
++ * *much* faster this way, as no tlb flushes means bigger wrpt batches.
++ */
++ if (tsk->active_mm == mm) {
++ tsk->active_mm = &init_mm;
++ atomic_inc(&init_mm.mm_count);
++
++ switch_mm(mm, &init_mm, tsk);
++
++ atomic_dec(&mm->mm_count);
++ BUG_ON(atomic_read(&mm->mm_count) == 0);
++ }
++
++ task_unlock(tsk);
++
++ if ( mm->context.pinned && (atomic_read(&mm->mm_count) == 1) &&
++ !mm->context.has_foreign_mappings )
++ mm_unpin(mm);
++}
++
++struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
++{
++ struct page *pte;
++
++ pte = alloc_pages(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO, 0);
++ if (pte) {
++ SetPageForeign(pte, pte_free);
++ init_page_count(pte);
++ }
++ return pte;
++}
++
++void pte_free(struct page *pte)
++{
++ unsigned long va = (unsigned long)__va(page_to_pfn(pte)<<PAGE_SHIFT);
++
++ if (!pte_write(*virt_to_ptep(va)))
++ if (HYPERVISOR_update_va_mapping(
++ va, pfn_pte(page_to_pfn(pte), PAGE_KERNEL), 0))
++ BUG();
++
++ ClearPageForeign(pte);
++ init_page_count(pte);
++
++ __free_page(pte);
++}
++#endif /* CONFIG_XEN */
++
++pte_t *lookup_address(unsigned long address)
++{
++ pgd_t *pgd = pgd_offset_k(address);
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
++ if (pgd_none(*pgd))
++ return NULL;
++ pud = pud_offset(pgd, address);
++ if (!pud_present(*pud))
++ return NULL;
++ pmd = pmd_offset(pud, address);
++ if (!pmd_present(*pmd))
++ return NULL;
++ if (pmd_large(*pmd))
++ return (pte_t *)pmd;
++ pte = pte_offset_kernel(pmd, address);
++ if (pte && !pte_present(*pte))
++ pte = NULL;
++ return pte;
++}
++
++static struct page *split_large_page(unsigned long address, pgprot_t prot,
++ pgprot_t ref_prot)
++{
++ int i;
++ unsigned long addr;
++ struct page *base = alloc_pages(GFP_KERNEL, 0);
++ pte_t *pbase;
++ if (!base)
++ return NULL;
++ /*
++ * page_private is used to track the number of entries in
++ * the page table page have non standard attributes.
++ */
++ SetPagePrivate(base);
++ page_private(base) = 0;
++
++ address = __pa(address);
++ addr = address & LARGE_PAGE_MASK;
++ pbase = (pte_t *)page_address(base);
++ for (i = 0; i < PTRS_PER_PTE; i++, addr += PAGE_SIZE) {
++ pbase[i] = pfn_pte(addr >> PAGE_SHIFT,
++ addr == address ? prot : ref_prot);
++ }
++ return base;
++}
++
++
++static void flush_kernel_map(void *address)
++{
++ if (0 && address && cpu_has_clflush) {
++ /* is this worth it? */
++ int i;
++ for (i = 0; i < PAGE_SIZE; i += boot_cpu_data.x86_clflush_size)
++ asm volatile("clflush (%0)" :: "r" (address + i));
++ } else
++ asm volatile("wbinvd":::"memory");
++ if (address)
++ __flush_tlb_one(address);
++ else
++ __flush_tlb_all();
++}
++
++
++static inline void flush_map(unsigned long address)
++{
++ on_each_cpu(flush_kernel_map, (void *)address, 1, 1);
++}
++
++static struct page *deferred_pages; /* protected by init_mm.mmap_sem */
++
++static inline void save_page(struct page *fpage)
++{
++ fpage->lru.next = (struct list_head *)deferred_pages;
++ deferred_pages = fpage;
++}
++
++/*
++ * No more special protections in this 2/4MB area - revert to a
++ * large page again.
++ */
++static void revert_page(unsigned long address, pgprot_t ref_prot)
++{
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t large_pte;
++
++ pgd = pgd_offset_k(address);
++ BUG_ON(pgd_none(*pgd));
++ pud = pud_offset(pgd,address);
++ BUG_ON(pud_none(*pud));
++ pmd = pmd_offset(pud, address);
++ BUG_ON(__pmd_val(*pmd) & _PAGE_PSE);
++ pgprot_val(ref_prot) |= _PAGE_PSE;
++ large_pte = mk_pte_phys(__pa(address) & LARGE_PAGE_MASK, ref_prot);
++ set_pte((pte_t *)pmd, large_pte);
++}
++
++static int
++__change_page_attr(unsigned long address, unsigned long pfn, pgprot_t prot,
++ pgprot_t ref_prot)
++{
++ pte_t *kpte;
++ struct page *kpte_page;
++ unsigned kpte_flags;
++ pgprot_t ref_prot2;
++ kpte = lookup_address(address);
++ if (!kpte) return 0;
++ kpte_page = virt_to_page(((unsigned long)kpte) & PAGE_MASK);
++ kpte_flags = pte_val(*kpte);
++ if (pgprot_val(prot) != pgprot_val(ref_prot)) {
++ if ((kpte_flags & _PAGE_PSE) == 0) {
++ set_pte(kpte, pfn_pte(pfn, prot));
++ } else {
++ /*
++ * split_large_page will take the reference for this
++ * change_page_attr on the split page.
++ */
++
++ struct page *split;
++ ref_prot2 = __pgprot(pgprot_val(pte_pgprot(*lookup_address(address))) & ~(1<<_PAGE_BIT_PSE));
++
++ split = split_large_page(address, prot, ref_prot2);
++ if (!split)
++ return -ENOMEM;
++ set_pte(kpte,mk_pte(split, ref_prot2));
++ kpte_page = split;
++ }
++ page_private(kpte_page)++;
++ } else if ((kpte_flags & _PAGE_PSE) == 0) {
++ set_pte(kpte, pfn_pte(pfn, ref_prot));
++ BUG_ON(page_private(kpte_page) == 0);
++ page_private(kpte_page)--;
++ } else
++ BUG();
++
++ /* on x86-64 the direct mapping set at boot is not using 4k pages */
++ /*
++ * ..., but the XEN guest kernels (currently) do:
++ * If the pte was reserved, it means it was created at boot
++ * time (not via split_large_page) and in turn we must not
++ * replace it with a large page.
++ */
++#ifndef CONFIG_XEN
++ BUG_ON(PageReserved(kpte_page));
++#else
++ if (PageReserved(kpte_page))
++ return 0;
++#endif
++
++ if (page_private(kpte_page) == 0) {
++ save_page(kpte_page);
++ revert_page(address, ref_prot);
++ }
++ return 0;
++}
++
++/*
++ * Change the page attributes of an page in the linear mapping.
++ *
++ * This should be used when a page is mapped with a different caching policy
++ * than write-back somewhere - some CPUs do not like it when mappings with
++ * different caching policies exist. This changes the page attributes of the
++ * in kernel linear mapping too.
++ *
++ * The caller needs to ensure that there are no conflicting mappings elsewhere.
++ * This function only deals with the kernel linear map.
++ *
++ * Caller must call global_flush_tlb() after this.
++ */
++int change_page_attr_addr(unsigned long address, int numpages, pgprot_t prot)
++{
++ int err = 0;
++ int i;
++
++ down_write(&init_mm.mmap_sem);
++ for (i = 0; i < numpages; i++, address += PAGE_SIZE) {
++ unsigned long pfn = __pa(address) >> PAGE_SHIFT;
++
++ err = __change_page_attr(address, pfn, prot, PAGE_KERNEL);
++ if (err)
++ break;
++ /* Handle kernel mapping too which aliases part of the
++ * lowmem */
++ if (__pa(address) < KERNEL_TEXT_SIZE) {
++ unsigned long addr2;
++ pgprot_t prot2 = prot;
++ addr2 = __START_KERNEL_map + __pa(address);
++ pgprot_val(prot2) &= ~_PAGE_NX;
++ err = __change_page_attr(addr2, pfn, prot2, PAGE_KERNEL_EXEC);
++ }
++ }
++ up_write(&init_mm.mmap_sem);
++ return err;
++}
++
++/* Don't call this for MMIO areas that may not have a mem_map entry */
++int change_page_attr(struct page *page, int numpages, pgprot_t prot)
++{
++ unsigned long addr = (unsigned long)page_address(page);
++ return change_page_attr_addr(addr, numpages, prot);
++}
++
++void global_flush_tlb(void)
++{
++ struct page *dpage;
++
++ down_read(&init_mm.mmap_sem);
++ dpage = xchg(&deferred_pages, NULL);
++ up_read(&init_mm.mmap_sem);
++
++ flush_map((dpage && !dpage->lru.next) ? (unsigned long)page_address(dpage) : 0);
++ while (dpage) {
++ struct page *tmp = dpage;
++ dpage = (struct page *)dpage->lru.next;
++ ClearPagePrivate(tmp);
++ __free_page(tmp);
++ }
++}
++
++EXPORT_SYMBOL(change_page_attr);
++EXPORT_SYMBOL(global_flush_tlb);
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/oprofile/Makefile
+--- a/arch/x86_64/oprofile/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/oprofile/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -11,9 +11,15 @@
+ oprofilefs.o oprofile_stats.o \
+ timer_int.o )
+
++ifdef CONFIG_XEN
++XENOPROF_COMMON_OBJS = $(addprefix ../../../drivers/xen/xenoprof/, \
++ xenoprofile.o)
++OPROFILE-y := xenoprof.o
++else
+ OPROFILE-y := init.o backtrace.o
+ OPROFILE-$(CONFIG_X86_LOCAL_APIC) += nmi_int.o op_model_athlon.o op_model_p4.o \
+ op_model_ppro.o
+ OPROFILE-$(CONFIG_X86_IO_APIC) += nmi_timer_int.o
+-
+-oprofile-y = $(DRIVER_OBJS) $(addprefix ../../i386/oprofile/, $(OPROFILE-y))
++endif
++oprofile-y = $(DRIVER_OBJS) $(XENOPROF_COMMON_OBJS) \
++ $(addprefix ../../i386/oprofile/, $(OPROFILE-y))
+diff -r d894e36cfc30 -r 0aa021803deb arch/x86_64/pci/Makefile
+--- a/arch/x86_64/pci/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/arch/x86_64/pci/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -15,8 +15,13 @@
+
+ obj-$(CONFIG_NUMA) += k8-bus.o
+
++# pcifront should be after mmconfig.o and direct.o as it should only
++# take over if direct access to the PCI bus is unavailable
++obj-$(CONFIG_XEN_PCIDEV_FRONTEND) += pcifront.o
++
+ direct-y += ../../i386/pci/direct.o
+ acpi-y += ../../i386/pci/acpi.o
++pcifront-y += ../../i386/pci/pcifront.o
+ legacy-y += ../../i386/pci/legacy.o
+ irq-y += ../../i386/pci/irq.o
+ common-y += ../../i386/pci/common.o
+diff -r d894e36cfc30 -r 0aa021803deb block/elevator.c
+--- a/block/elevator.c Tue Sep 09 11:37:38 2008 +0200
++++ b/block/elevator.c Wed Sep 10 10:54:08 2008 +0100
+@@ -493,6 +493,16 @@
+ int ret;
+
+ while ((rq = __elv_next_request(q)) != NULL) {
++ /*
++ * Kill the empty barrier place holder, the driver must
++ * not ever see it.
++ */
++ if (blk_empty_barrier(rq)) {
++ blkdev_dequeue_request(rq);
++ end_that_request_chunk(rq, 1, 0);
++ end_that_request_last(rq, 1);
++ continue;
++ }
+ if (!(rq->flags & REQ_STARTED)) {
+ elevator_t *e = q->elevator;
+
+diff -r d894e36cfc30 -r 0aa021803deb block/ll_rw_blk.c
+--- a/block/ll_rw_blk.c Tue Sep 09 11:37:38 2008 +0200
++++ b/block/ll_rw_blk.c Wed Sep 10 10:54:08 2008 +0100
+@@ -483,9 +483,12 @@
+ * Queue ordered sequence. As we stack them at the head, we
+ * need to queue in reverse order. Note that we rely on that
+ * no fs request uses ELEVATOR_INSERT_FRONT and thus no fs
+- * request gets inbetween ordered sequence.
++ * request gets inbetween ordered sequence. If this request is
++ * an empty barrier, we don't need to do a postflush ever since
++ * there will be no data written between the pre and post flush.
++ * Hence a single flush will suffice.
+ */
+- if (q->ordered & QUEUE_ORDERED_POSTFLUSH)
++ if ((q->ordered & QUEUE_ORDERED_POSTFLUSH) && !blk_empty_barrier(rq))
+ queue_flush(q, QUEUE_ORDERED_POSTFLUSH);
+ else
+ q->ordseq |= QUEUE_ORDSEQ_POSTFLUSH;
+@@ -2967,7 +2970,7 @@
+ {
+ struct block_device *bdev = bio->bi_bdev;
+
+- if (bdev != bdev->bd_contains) {
++ if (bio_sectors(bio) && bdev != bdev->bd_contains) {
+ struct hd_struct *p = bdev->bd_part;
+ const int rw = bio_data_dir(bio);
+
+@@ -3028,7 +3031,7 @@
+ might_sleep();
+ /* Test device or partition size, when known. */
+ maxsector = bio->bi_bdev->bd_inode->i_size >> 9;
+- if (maxsector) {
++ if (maxsector && nr_sectors) {
+ sector_t sector = bio->bi_sector;
+
+ if (maxsector < nr_sectors || maxsector - nr_sectors < sector) {
+@@ -3094,7 +3097,7 @@
+ old_dev = bio->bi_bdev->bd_dev;
+
+ maxsector = bio->bi_bdev->bd_inode->i_size >> 9;
+- if (maxsector) {
++ if (maxsector && nr_sectors) {
+ sector_t sector = bio->bi_sector;
+
+ if (maxsector < nr_sectors || maxsector - nr_sectors < sector) {
+@@ -3128,21 +3131,25 @@
+ {
+ int count = bio_sectors(bio);
+
+- BIO_BUG_ON(!bio->bi_size);
+- BIO_BUG_ON(!bio->bi_io_vec);
+ bio->bi_rw |= rw;
+- if (rw & WRITE)
+- count_vm_events(PGPGOUT, count);
+- else
+- count_vm_events(PGPGIN, count);
+
+- if (unlikely(block_dump)) {
+- char b[BDEVNAME_SIZE];
+- printk(KERN_DEBUG "%s(%d): %s block %Lu on %s\n",
+- current->comm, current->pid,
+- (rw & WRITE) ? "WRITE" : "READ",
+- (unsigned long long)bio->bi_sector,
+- bdevname(bio->bi_bdev,b));
++ if (!bio_empty_barrier(bio)) {
++ BIO_BUG_ON(!bio->bi_size);
++ BIO_BUG_ON(!bio->bi_io_vec);
++
++ if (rw & WRITE)
++ count_vm_events(PGPGOUT, count);
++ else
++ count_vm_events(PGPGIN, count);
++
++ if (unlikely(block_dump)) {
++ char b[BDEVNAME_SIZE];
++ printk(KERN_DEBUG "%s(%d): %s block %Lu on %s\n",
++ current->comm, current->pid,
++ (rw & WRITE) ? "WRITE" : "READ",
++ (unsigned long long)bio->bi_sector,
++ bdevname(bio->bi_bdev,b));
++ }
+ }
+
+ generic_make_request(bio);
+@@ -3259,6 +3266,13 @@
+ total_bytes = bio_nbytes = 0;
+ while ((bio = req->bio) != NULL) {
+ int nbytes;
++
++ /* For an empty barrier request, the low level driver must
++ * store a potential error location in ->sector. We pass
++ * that back up in ->bi_sector
++ */
++ if (blk_empty_barrier(req))
++ bio->bi_sector = req->sector;
+
+ if (nr_bytes >= bio->bi_size) {
+ req->bio = bio->bi_next;
+diff -r d894e36cfc30 -r 0aa021803deb buildconfigs/conf.linux-native/00_xen_to_native
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/buildconfigs/conf.linux-native/00_xen_to_native Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,86 @@
++# Linux kernel version: 2.6.16.13-native
++# Mon May 15 10:59:54 2006
++#
++CONFIG_X86_PC=y
++# CONFIG_X86_XEN is not set
++# CONFIG_HPET_TIMER is not set
++# CONFIG_SCHED_SMT is not set
++# CONFIG_X86_MCE is not set
++# CONFIG_X86_MSR is not set
++# CONFIG_SWIOTLB is not set
++# CONFIG_EDD is not set
++CONFIG_ARCH_FLATMEM_ENABLE=y
++CONFIG_ARCH_SPARSEMEM_ENABLE=y
++CONFIG_ARCH_SELECT_MEMORY_MODEL=y
++CONFIG_SPARSEMEM_STATIC=y
++CONFIG_SPLIT_PTLOCK_CPUS=4
++# CONFIG_HIGHPTE is not set
++# CONFIG_MATH_EMULATION is not set
++# CONFIG_EFI is not set
++CONFIG_IRQBALANCE=y
++# CONFIG_KEXEC is not set
++CONFIG_DOUBLEFAULT=y
++CONFIG_PM_LEGACY=y
++# CONFIG_PM_DEBUG is not set
++# CONFIG_SOFTWARE_SUSPEND is not set
++CONFIG_SUSPEND_SMP=y
++CONFIG_ACPI_SLEEP=y
++CONFIG_ACPI_SLEEP_PROC_FS=y
++# CONFIG_ACPI_SLEEP_PROC_SLEEP is not set
++CONFIG_X86_PM_TIMER=y
++# APM (Advanced Power Management) BIOS Support
++#
++# CONFIG_APM is not set
++
++#
++CONFIG_PCI_BIOS=y
++# CONFIG_XEN_PCIDEV_FRONTEND is not set
++# CONFIG_XEN_PCIDEV_FE_DEBUG is not set
++# CONFIG_PCI_MSI is not set
++# CONFIG_ISA is not set
++# CONFIG_MCA is not set
++# CONFIG_HOTPLUG_PCI_COMPAQ is not set
++# CONFIG_HOTPLUG_PCI_IBM is not set
++# CONFIG_I2O_EXT_ADAPTEC_DMA64 is not set
++CONFIG_SERIAL_8250=y
++CONFIG_SERIAL_8250_CONSOLE=y
++CONFIG_SERIAL_CORE=y
++CONFIG_SERIAL_CORE_CONSOLE=y
++# CONFIG_TCG_XEN is not set
++# CONFIG_HUGETLBFS is not set
++# CONFIG_XEN is not set
++# CONFIG_XEN_INTERFACE_VERSION is not set
++
++#
++# XEN
++#
++# CONFIG_XEN_PRIVILEGED_GUEST is not set
++# CONFIG_XEN_UNPRIVILEGED_GUEST is not set
++# CONFIG_XEN_BACKEND is not set
++# CONFIG_XEN_PCIDEV_BACKEND is not set
++# CONFIG_XEN_PCIDEV_BACKEND_VPCI is not set
++# CONFIG_XEN_PCIDEV_BACKEND_PASS is not set
++# CONFIG_XEN_PCIDEV_BE_DEBUG is not set
++# CONFIG_XEN_BLKDEV_BACKEND is not set
++# CONFIG_XEN_BLKDEV_TAP_BE is not set
++# CONFIG_XEN_NETDEV_BACKEND is not set
++# CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER is not set
++# CONFIG_XEN_NETDEV_LOOPBACK is not set
++# CONFIG_XEN_TPMDEV_BACKEND is not set
++# CONFIG_XEN_BLKDEV_FRONTEND is not set
++# CONFIG_XEN_NETDEV_FRONTEND is not set
++# CONFIG_XEN_BLKDEV_TAP is not set
++# CONFIG_XEN_SCRUB_PAGES is not set
++# CONFIG_XEN_DISABLE_SERIAL is not set
++# CONFIG_XEN_SYSFS is not set
++# CONFIG_XEN_COMPAT_030002_AND_LATER is not set
++# CONFIG_XEN_COMPAT_030004_AND_LATER is not set
++# CONFIG_XEN_COMPAT_LATEST_ONLY is not set
++# CONFIG_XEN_COMPAT_030002 is not set
++# CONFIG_XEN_COMPAT_030004 is not set
++# CONFIG_HAVE_ARCH_ALLOC_SKB is not set
++# CONFIG_HAVE_ARCH_DEV_ALLOC_SKB is not set
++# CONFIG_NO_IDLE_HZ is not set
++CONFIG_X86_HT=y
++# CONFIG_X86_NO_TSS is not set
++# CONFIG_X86_NO_IDT is not set
+diff -r d894e36cfc30 -r 0aa021803deb buildconfigs/create_config.sh
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/buildconfigs/create_config.sh Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,54 @@
++#!/bin/sh
++set -e
++
++
++# Parse arguments
++#
++if [ $# -lt 1 -o $# -gt 4 ]; then
++ echo "Usage: $0 config-file EXTRAVERSION XEN_TARGET_ARCH XEN_SYSTYPE"
++ exit 1
++fi
++
++config_file=$1
++extraversion=$2
++target_arch=$3
++systype=$4
++
++
++# Start with initial config skeleton file, if any.
++# Derive from linux-defconfig_xen_x86_32 otherwise.
++#
++skeleton=buildconfigs/linux-defconfig_${extraversion#-}_${target_arch}${systype}
++[ -r $skeleton ] || skeleton=buildconfigs/linux-defconfig_xen_x86_32
++[ -r $skeleton.local ] && skeleton=$skeleton.local
++cp $skeleton $config_file
++
++echo "Using $skeleton as base config"
++
++# Update
++#
++filter_template="s/^#\{0,1\} *\(CONFIG[^= ]*\).*/\/^#\\\{0,1\\\} *\1[= ].*\/d/p"
++config_dirs="buildconfigs/conf.linux buildconfigs/conf.linux-${target_arch} buildconfigs/conf.linux${extraversion} buildconfigs/conf.linux-${target_arch}${extraversion}"
++
++for config_dir in $config_dirs
++do
++ if [ -d $config_dir ]; then
++ echo "Processing $config_dir..." 1>&2
++ # processing is done in alphanumeric order
++ find $config_dir -type f | sort | while read update
++ do
++ echo " ... $update" 1>&2
++ # create the filter rules in a temp file
++ filter_rules=`mktemp -t xenupdateconf.XXXXXXXXXX`
++ sed -n "${filter_template}" < $update > $filter_rules
++
++ # filter the config file in place, removing any options that
++ # will be updated.
++ sed -f $filter_rules -i $config_file
++ cat $update >> $config_file
++
++ # clean up
++ rm -f $filter_rules
++ done
++ fi
++done
+diff -r d894e36cfc30 -r 0aa021803deb buildconfigs/linux-defconfig_xen0_ia64
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/buildconfigs/linux-defconfig_xen0_ia64 Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1703 @@
++#
++# Automatically generated make config: don't edit
++# Linux kernel version: 2.6.18.8
++# Tue Feb 19 11:20:09 2008
++#
++CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config"
++
++#
++# Code maturity level options
++#
++CONFIG_EXPERIMENTAL=y
++CONFIG_LOCK_KERNEL=y
++CONFIG_INIT_ENV_ARG_LIMIT=32
++
++#
++# General setup
++#
++CONFIG_LOCALVERSION=""
++CONFIG_LOCALVERSION_AUTO=y
++CONFIG_SWAP=y
++CONFIG_SYSVIPC=y
++CONFIG_POSIX_MQUEUE=y
++CONFIG_BSD_PROCESS_ACCT=y
++# CONFIG_BSD_PROCESS_ACCT_V3 is not set
++# CONFIG_TASKSTATS is not set
++# CONFIG_AUDIT is not set
++CONFIG_IKCONFIG=y
++CONFIG_IKCONFIG_PROC=y
++# CONFIG_CPUSETS is not set
++# CONFIG_RELAY is not set
++CONFIG_INITRAMFS_SOURCE=""
++CONFIG_CC_OPTIMIZE_FOR_SIZE=y
++# CONFIG_EMBEDDED is not set
++CONFIG_SYSCTL=y
++CONFIG_KALLSYMS=y
++CONFIG_KALLSYMS_ALL=y
++CONFIG_KALLSYMS_EXTRA_PASS=y
++CONFIG_HOTPLUG=y
++CONFIG_PRINTK=y
++CONFIG_BUG=y
++CONFIG_ELF_CORE=y
++CONFIG_BASE_FULL=y
++CONFIG_FUTEX=y
++CONFIG_EPOLL=y
++CONFIG_SHMEM=y
++CONFIG_SLAB=y
++CONFIG_VM_EVENT_COUNTERS=y
++CONFIG_RT_MUTEXES=y
++# CONFIG_TINY_SHMEM is not set
++CONFIG_BASE_SMALL=0
++# CONFIG_SLOB is not set
++
++#
++# Loadable module support
++#
++CONFIG_MODULES=y
++CONFIG_MODULE_UNLOAD=y
++# CONFIG_MODULE_FORCE_UNLOAD is not set
++CONFIG_MODVERSIONS=y
++CONFIG_MODULE_SRCVERSION_ALL=y
++CONFIG_KMOD=y
++CONFIG_STOP_MACHINE=y
++
++#
++# Block layer
++#
++# CONFIG_BLK_DEV_IO_TRACE is not set
++
++#
++# IO Schedulers
++#
++CONFIG_IOSCHED_NOOP=y
++CONFIG_IOSCHED_AS=y
++CONFIG_IOSCHED_DEADLINE=y
++CONFIG_IOSCHED_CFQ=y
++CONFIG_DEFAULT_AS=y
++# CONFIG_DEFAULT_DEADLINE is not set
++# CONFIG_DEFAULT_CFQ is not set
++# CONFIG_DEFAULT_NOOP is not set
++CONFIG_DEFAULT_IOSCHED="anticipatory"
++
++#
++# Processor type and features
++#
++CONFIG_IA64=y
++CONFIG_64BIT=y
++CONFIG_MMU=y
++CONFIG_SWIOTLB=y
++CONFIG_RWSEM_XCHGADD_ALGORITHM=y
++CONFIG_GENERIC_FIND_NEXT_BIT=y
++CONFIG_GENERIC_CALIBRATE_DELAY=y
++CONFIG_TIME_INTERPOLATION=y
++CONFIG_DMI=y
++CONFIG_EFI=y
++CONFIG_GENERIC_IOMAP=y
++CONFIG_XEN=y
++CONFIG_XEN_IA64_EXPOSE_P2M=y
++CONFIG_XEN_IA64_EXPOSE_P2M_USE_DTR=y
++CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER=y
++CONFIG_DMA_IS_DMA32=y
++CONFIG_AUDIT_ARCH=y
++CONFIG_IA64_GENERIC=y
++# CONFIG_IA64_DIG is not set
++# CONFIG_IA64_HP_ZX1 is not set
++# CONFIG_IA64_HP_ZX1_SWIOTLB is not set
++# CONFIG_IA64_SGI_SN2 is not set
++# CONFIG_IA64_HP_SIM is not set
++# CONFIG_IA64_XEN is not set
++# CONFIG_ITANIUM is not set
++CONFIG_MCKINLEY=y
++# CONFIG_IA64_PAGE_SIZE_4KB is not set
++# CONFIG_IA64_PAGE_SIZE_8KB is not set
++CONFIG_IA64_PAGE_SIZE_16KB=y
++# CONFIG_IA64_PAGE_SIZE_64KB is not set
++CONFIG_PGTABLE_3=y
++# CONFIG_PGTABLE_4 is not set
++CONFIG_HZ_100=y
++# CONFIG_HZ_250 is not set
++# CONFIG_HZ_1000 is not set
++CONFIG_HZ=100
++CONFIG_IA64_L1_CACHE_SHIFT=7
++CONFIG_IA64_CYCLONE=y
++CONFIG_IOSAPIC=y
++# CONFIG_IA64_SGI_SN_XP is not set
++CONFIG_FORCE_MAX_ZONEORDER=11
++CONFIG_SMP=y
++CONFIG_NR_CPUS=16
++CONFIG_HOTPLUG_CPU=y
++CONFIG_ARCH_ENABLE_MEMORY_HOTPLUG=y
++# CONFIG_SCHED_SMT is not set
++# CONFIG_PERMIT_BSP_REMOVE is not set
++# CONFIG_PREEMPT is not set
++CONFIG_SELECT_MEMORY_MODEL=y
++# CONFIG_FLATMEM_MANUAL is not set
++CONFIG_DISCONTIGMEM_MANUAL=y
++# CONFIG_SPARSEMEM_MANUAL is not set
++CONFIG_DISCONTIGMEM=y
++CONFIG_FLAT_NODE_MEM_MAP=y
++CONFIG_NEED_MULTIPLE_NODES=y
++# CONFIG_SPARSEMEM_STATIC is not set
++CONFIG_SPLIT_PTLOCK_CPUS=4
++# CONFIG_MIGRATION is not set
++CONFIG_RESOURCES_64BIT=y
++CONFIG_ARCH_SELECT_MEMORY_MODEL=y
++CONFIG_ARCH_DISCONTIGMEM_ENABLE=y
++CONFIG_ARCH_FLATMEM_ENABLE=y
++CONFIG_ARCH_SPARSEMEM_ENABLE=y
++CONFIG_ARCH_DISCONTIGMEM_DEFAULT=y
++CONFIG_NUMA=y
++CONFIG_NODES_SHIFT=10
++CONFIG_VIRTUAL_MEM_MAP=y
++CONFIG_HOLES_IN_ZONE=y
++CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID=y
++CONFIG_HAVE_ARCH_NODEDATA_EXTENSION=y
++# CONFIG_IA32_SUPPORT is not set
++# CONFIG_IA64_MCA_RECOVERY is not set
++CONFIG_PERFMON=y
++CONFIG_IA64_PALINFO=y
++CONFIG_SGI_SN=y
++
++#
++# SN Devices
++#
++# CONFIG_SGI_IOC3 is not set
++CONFIG_KEXEC=y
++
++#
++# Firmware Drivers
++#
++CONFIG_EFI_VARS=y
++CONFIG_EFI_PCDP=y
++CONFIG_BINFMT_ELF=y
++CONFIG_BINFMT_MISC=y
++
++#
++# Power management and ACPI
++#
++CONFIG_PM=y
++CONFIG_PM_LEGACY=y
++# CONFIG_PM_DEBUG is not set
++
++#
++# ACPI (Advanced Configuration and Power Interface) Support
++#
++CONFIG_ACPI=y
++CONFIG_ACPI_BUTTON=y
++CONFIG_ACPI_FAN=y
++# CONFIG_ACPI_DOCK is not set
++CONFIG_ACPI_PROCESSOR=y
++CONFIG_ACPI_HOTPLUG_CPU=y
++CONFIG_ACPI_THERMAL=y
++CONFIG_ACPI_NUMA=y
++CONFIG_ACPI_BLACKLIST_YEAR=0
++# CONFIG_ACPI_DEBUG is not set
++CONFIG_ACPI_EC=y
++CONFIG_ACPI_POWER=y
++CONFIG_ACPI_SYSTEM=y
++CONFIG_ACPI_CONTAINER=y
++
++#
++# CPU Frequency scaling
++#
++# CONFIG_CPU_FREQ is not set
++
++#
++# Bus options (PCI, PCMCIA)
++#
++CONFIG_PCI=y
++CONFIG_PCI_DOMAINS=y
++CONFIG_XEN_PCIDEV_FRONTEND=y
++# CONFIG_XEN_PCIDEV_FE_DEBUG is not set
++# CONFIG_PCIEPORTBUS is not set
++# CONFIG_PCI_DEBUG is not set
++
++#
++# PCI Hotplug Support
++#
++CONFIG_HOTPLUG_PCI=y
++# CONFIG_HOTPLUG_PCI_FAKE is not set
++CONFIG_HOTPLUG_PCI_ACPI=y
++# CONFIG_HOTPLUG_PCI_ACPI_IBM is not set
++# CONFIG_HOTPLUG_PCI_CPCI is not set
++# CONFIG_HOTPLUG_PCI_SHPC is not set
++# CONFIG_HOTPLUG_PCI_SGI is not set
++
++#
++# PCCARD (PCMCIA/CardBus) support
++#
++# CONFIG_PCCARD is not set
++
++#
++# Networking
++#
++CONFIG_NET=y
++
++#
++# Networking options
++#
++# CONFIG_NETDEBUG is not set
++CONFIG_PACKET=y
++# CONFIG_PACKET_MMAP is not set
++CONFIG_UNIX=y
++CONFIG_XFRM=y
++# CONFIG_XFRM_USER is not set
++# CONFIG_NET_KEY is not set
++CONFIG_INET=y
++CONFIG_IP_MULTICAST=y
++# CONFIG_IP_ADVANCED_ROUTER is not set
++CONFIG_IP_FIB_HASH=y
++CONFIG_IP_PNP=y
++CONFIG_IP_PNP_DHCP=y
++# CONFIG_IP_PNP_BOOTP is not set
++# CONFIG_IP_PNP_RARP is not set
++# CONFIG_NET_IPIP is not set
++# CONFIG_NET_IPGRE is not set
++# CONFIG_IP_MROUTE is not set
++CONFIG_ARPD=y
++CONFIG_SYN_COOKIES=y
++# CONFIG_INET_AH is not set
++# CONFIG_INET_ESP is not set
++# CONFIG_INET_IPCOMP is not set
++# CONFIG_INET_XFRM_TUNNEL is not set
++# CONFIG_INET_TUNNEL is not set
++CONFIG_INET_XFRM_MODE_TRANSPORT=y
++CONFIG_INET_XFRM_MODE_TUNNEL=y
++CONFIG_INET_DIAG=y
++CONFIG_INET_TCP_DIAG=y
++# CONFIG_TCP_CONG_ADVANCED is not set
++CONFIG_TCP_CONG_BIC=y
++
++#
++# IP: Virtual Server Configuration
++#
++# CONFIG_IP_VS is not set
++# CONFIG_IPV6 is not set
++# CONFIG_INET6_XFRM_TUNNEL is not set
++# CONFIG_INET6_TUNNEL is not set
++# CONFIG_NETWORK_SECMARK is not set
++CONFIG_NETFILTER=y
++# CONFIG_NETFILTER_DEBUG is not set
++CONFIG_BRIDGE_NETFILTER=y
++
++#
++# Core Netfilter Configuration
++#
++# CONFIG_NETFILTER_NETLINK is not set
++# CONFIG_NF_CONNTRACK is not set
++# CONFIG_NETFILTER_XTABLES is not set
++
++#
++# IP: Netfilter Configuration
++#
++# CONFIG_IP_NF_CONNTRACK is not set
++# CONFIG_IP_NF_QUEUE is not set
++
++#
++# Bridge: Netfilter Configuration
++#
++# CONFIG_BRIDGE_NF_EBTABLES is not set
++
++#
++# DCCP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_DCCP is not set
++
++#
++# SCTP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_SCTP is not set
++
++#
++# TIPC Configuration (EXPERIMENTAL)
++#
++# CONFIG_TIPC is not set
++# CONFIG_ATM is not set
++CONFIG_BRIDGE=y
++# CONFIG_VLAN_8021Q is not set
++# CONFIG_DECNET is not set
++CONFIG_LLC=y
++# CONFIG_LLC2 is not set
++# CONFIG_IPX is not set
++# CONFIG_ATALK is not set
++# CONFIG_X25 is not set
++# CONFIG_LAPB is not set
++# CONFIG_ECONET is not set
++# CONFIG_WAN_ROUTER is not set
++
++#
++# QoS and/or fair queueing
++#
++# CONFIG_NET_SCHED is not set
++
++#
++# Network testing
++#
++# CONFIG_NET_PKTGEN is not set
++# CONFIG_HAMRADIO is not set
++# CONFIG_IRDA is not set
++# CONFIG_BT is not set
++# CONFIG_IEEE80211 is not set
++
++#
++# Device Drivers
++#
++
++#
++# Generic Driver Options
++#
++CONFIG_STANDALONE=y
++CONFIG_PREVENT_FIRMWARE_BUILD=y
++CONFIG_FW_LOADER=y
++# CONFIG_DEBUG_DRIVER is not set
++CONFIG_SYS_HYPERVISOR=y
++
++#
++# Connector - unified userspace <-> kernelspace linker
++#
++# CONFIG_CONNECTOR is not set
++
++#
++# Memory Technology Devices (MTD)
++#
++# CONFIG_MTD is not set
++
++#
++# Parallel port support
++#
++# CONFIG_PARPORT is not set
++
++#
++# Plug and Play support
++#
++# CONFIG_PNP is not set
++
++#
++# Block devices
++#
++# CONFIG_BLK_CPQ_DA is not set
++CONFIG_BLK_CPQ_CISS_DA=y
++# CONFIG_CISS_SCSI_TAPE is not set
++# CONFIG_BLK_DEV_DAC960 is not set
++# CONFIG_BLK_DEV_UMEM is not set
++# CONFIG_BLK_DEV_COW_COMMON is not set
++CONFIG_BLK_DEV_LOOP=y
++CONFIG_BLK_DEV_CRYPTOLOOP=y
++CONFIG_BLK_DEV_NBD=m
++# CONFIG_BLK_DEV_SX8 is not set
++# CONFIG_BLK_DEV_UB is not set
++CONFIG_BLK_DEV_RAM=y
++CONFIG_BLK_DEV_RAM_COUNT=16
++CONFIG_BLK_DEV_RAM_SIZE=4096
++CONFIG_BLK_DEV_RAM_BLOCKSIZE=1024
++CONFIG_BLK_DEV_INITRD=y
++# CONFIG_CDROM_PKTCDVD is not set
++# CONFIG_ATA_OVER_ETH is not set
++
++#
++# ATA/ATAPI/MFM/RLL support
++#
++CONFIG_IDE=y
++CONFIG_IDE_MAX_HWIFS=4
++CONFIG_BLK_DEV_IDE=y
++
++#
++# Please see Documentation/ide.txt for help/info on IDE drives
++#
++# CONFIG_BLK_DEV_IDE_SATA is not set
++CONFIG_BLK_DEV_IDEDISK=y
++# CONFIG_IDEDISK_MULTI_MODE is not set
++CONFIG_BLK_DEV_IDECD=y
++# CONFIG_BLK_DEV_IDETAPE is not set
++CONFIG_BLK_DEV_IDEFLOPPY=y
++CONFIG_BLK_DEV_IDESCSI=y
++# CONFIG_IDE_TASK_IOCTL is not set
++
++#
++# IDE chipset support/bugfixes
++#
++# CONFIG_IDE_GENERIC is not set
++CONFIG_BLK_DEV_IDEPCI=y
++# CONFIG_IDEPCI_SHARE_IRQ is not set
++# CONFIG_BLK_DEV_OFFBOARD is not set
++CONFIG_BLK_DEV_GENERIC=y
++# CONFIG_BLK_DEV_OPTI621 is not set
++CONFIG_BLK_DEV_IDEDMA_PCI=y
++# CONFIG_BLK_DEV_IDEDMA_FORCED is not set
++CONFIG_IDEDMA_PCI_AUTO=y
++# CONFIG_IDEDMA_ONLYDISK is not set
++# CONFIG_BLK_DEV_AEC62XX is not set
++# CONFIG_BLK_DEV_ALI15X3 is not set
++# CONFIG_BLK_DEV_AMD74XX is not set
++CONFIG_BLK_DEV_CMD64X=y
++# CONFIG_BLK_DEV_TRIFLEX is not set
++# CONFIG_BLK_DEV_CY82C693 is not set
++# CONFIG_BLK_DEV_CS5520 is not set
++# CONFIG_BLK_DEV_CS5530 is not set
++# CONFIG_BLK_DEV_HPT34X is not set
++# CONFIG_BLK_DEV_HPT366 is not set
++# CONFIG_BLK_DEV_SC1200 is not set
++CONFIG_BLK_DEV_PIIX=y
++# CONFIG_BLK_DEV_IT821X is not set
++# CONFIG_BLK_DEV_NS87415 is not set
++# CONFIG_BLK_DEV_PDC202XX_OLD is not set
++# CONFIG_BLK_DEV_PDC202XX_NEW is not set
++# CONFIG_BLK_DEV_SVWKS is not set
++# CONFIG_BLK_DEV_SIIMAGE is not set
++# CONFIG_BLK_DEV_SLC90E66 is not set
++# CONFIG_BLK_DEV_TRM290 is not set
++# CONFIG_BLK_DEV_VIA82CXXX is not set
++# CONFIG_IDE_ARM is not set
++CONFIG_BLK_DEV_IDEDMA=y
++# CONFIG_IDEDMA_IVB is not set
++CONFIG_IDEDMA_AUTO=y
++# CONFIG_BLK_DEV_HD is not set
++
++#
++# SCSI device support
++#
++# CONFIG_RAID_ATTRS is not set
++CONFIG_SCSI=y
++CONFIG_SCSI_PROC_FS=y
++
++#
++# SCSI support type (disk, tape, CD-ROM)
++#
++CONFIG_BLK_DEV_SD=y
++CONFIG_CHR_DEV_ST=y
++CONFIG_CHR_DEV_OSST=y
++CONFIG_BLK_DEV_SR=y
++CONFIG_BLK_DEV_SR_VENDOR=y
++CONFIG_CHR_DEV_SG=y
++# CONFIG_CHR_DEV_SCH is not set
++
++#
++# Some SCSI devices (e.g. CD jukebox) support multiple LUNs
++#
++CONFIG_SCSI_MULTI_LUN=y
++CONFIG_SCSI_CONSTANTS=y
++CONFIG_SCSI_LOGGING=y
++
++#
++# SCSI Transport Attributes
++#
++CONFIG_SCSI_SPI_ATTRS=y
++CONFIG_SCSI_FC_ATTRS=y
++# CONFIG_SCSI_ISCSI_ATTRS is not set
++CONFIG_SCSI_SAS_ATTRS=y
++
++#
++# SCSI low-level drivers
++#
++# CONFIG_ISCSI_TCP is not set
++# CONFIG_BLK_DEV_3W_XXXX_RAID is not set
++# CONFIG_SCSI_3W_9XXX is not set
++# CONFIG_SCSI_ACARD is not set
++# CONFIG_SCSI_AACRAID is not set
++# CONFIG_SCSI_AIC7XXX is not set
++# CONFIG_SCSI_AIC7XXX_OLD is not set
++# CONFIG_SCSI_AIC79XX is not set
++# CONFIG_MEGARAID_NEWGEN is not set
++# CONFIG_MEGARAID_LEGACY is not set
++# CONFIG_MEGARAID_SAS is not set
++# CONFIG_SCSI_SATA is not set
++# CONFIG_SCSI_HPTIOP is not set
++# CONFIG_SCSI_DMX3191D is not set
++# CONFIG_SCSI_FUTURE_DOMAIN is not set
++# CONFIG_SCSI_IPS is not set
++# CONFIG_SCSI_INITIO is not set
++# CONFIG_SCSI_INIA100 is not set
++CONFIG_SCSI_SYM53C8XX_2=y
++CONFIG_SCSI_SYM53C8XX_DMA_ADDRESSING_MODE=1
++CONFIG_SCSI_SYM53C8XX_DEFAULT_TAGS=16
++CONFIG_SCSI_SYM53C8XX_MAX_TAGS=64
++CONFIG_SCSI_SYM53C8XX_MMIO=y
++# CONFIG_SCSI_IPR is not set
++CONFIG_SCSI_QLOGIC_1280=y
++# CONFIG_SCSI_QLA_FC is not set
++# CONFIG_SCSI_LPFC is not set
++# CONFIG_SCSI_DC395x is not set
++# CONFIG_SCSI_DC390T is not set
++# CONFIG_SCSI_DEBUG is not set
++
++#
++# Multi-device support (RAID and LVM)
++#
++CONFIG_MD=y
++# CONFIG_BLK_DEV_MD is not set
++CONFIG_BLK_DEV_DM=y
++CONFIG_DM_CRYPT=m
++CONFIG_DM_SNAPSHOT=y
++CONFIG_DM_MIRROR=m
++CONFIG_DM_ZERO=m
++CONFIG_DM_MULTIPATH=m
++CONFIG_DM_MULTIPATH_EMC=m
++
++#
++# Fusion MPT device support
++#
++CONFIG_FUSION=y
++CONFIG_FUSION_SPI=y
++# CONFIG_FUSION_FC is not set
++CONFIG_FUSION_SAS=y
++CONFIG_FUSION_MAX_SGE=128
++# CONFIG_FUSION_CTL is not set
++
++#
++# IEEE 1394 (FireWire) support
++#
++# CONFIG_IEEE1394 is not set
++
++#
++# I2O device support
++#
++# CONFIG_I2O is not set
++
++#
++# Network device support
++#
++CONFIG_NETDEVICES=y
++CONFIG_DUMMY=y
++# CONFIG_BONDING is not set
++# CONFIG_EQUALIZER is not set
++CONFIG_TUN=y
++
++#
++# ARCnet devices
++#
++CONFIG_ARCNET=y
++# CONFIG_ARCNET_1201 is not set
++# CONFIG_ARCNET_1051 is not set
++# CONFIG_ARCNET_RAW is not set
++# CONFIG_ARCNET_CAP is not set
++# CONFIG_ARCNET_COM90xx is not set
++# CONFIG_ARCNET_COM90xxIO is not set
++# CONFIG_ARCNET_RIM_I is not set
++# CONFIG_ARCNET_COM20020 is not set
++
++#
++# PHY device support
++#
++# CONFIG_PHYLIB is not set
++
++#
++# Ethernet (10 or 100Mbit)
++#
++CONFIG_NET_ETHERNET=y
++CONFIG_MII=y
++# CONFIG_HAPPYMEAL is not set
++# CONFIG_SUNGEM is not set
++# CONFIG_CASSINI is not set
++# CONFIG_NET_VENDOR_3COM is not set
++
++#
++# Tulip family network device support
++#
++CONFIG_NET_TULIP=y
++# CONFIG_DE2104X is not set
++CONFIG_TULIP=y
++CONFIG_TULIP_MWI=y
++CONFIG_TULIP_MMIO=y
++CONFIG_TULIP_NAPI=y
++CONFIG_TULIP_NAPI_HW_MITIGATION=y
++# CONFIG_DE4X5 is not set
++# CONFIG_WINBOND_840 is not set
++# CONFIG_DM9102 is not set
++# CONFIG_ULI526X is not set
++# CONFIG_HP100 is not set
++CONFIG_NET_PCI=y
++# CONFIG_PCNET32 is not set
++# CONFIG_AMD8111_ETH is not set
++# CONFIG_ADAPTEC_STARFIRE is not set
++# CONFIG_B44 is not set
++# CONFIG_FORCEDETH is not set
++# CONFIG_DGRS is not set
++CONFIG_EEPRO100=y
++CONFIG_E100=y
++# CONFIG_FEALNX is not set
++# CONFIG_NATSEMI is not set
++# CONFIG_NE2K_PCI is not set
++# CONFIG_8139CP is not set
++# CONFIG_8139TOO is not set
++# CONFIG_SIS900 is not set
++# CONFIG_EPIC100 is not set
++# CONFIG_SUNDANCE is not set
++# CONFIG_VIA_RHINE is not set
++
++#
++# Ethernet (1000 Mbit)
++#
++# CONFIG_ACENIC is not set
++# CONFIG_DL2K is not set
++CONFIG_E1000=y
++# CONFIG_E1000_NAPI is not set
++# CONFIG_E1000_DISABLE_PACKET_SPLIT is not set
++# CONFIG_NS83820 is not set
++# CONFIG_HAMACHI is not set
++# CONFIG_YELLOWFIN is not set
++# CONFIG_R8169 is not set
++# CONFIG_SIS190 is not set
++# CONFIG_SKGE is not set
++# CONFIG_SKY2 is not set
++# CONFIG_SK98LIN is not set
++# CONFIG_VIA_VELOCITY is not set
++CONFIG_TIGON3=y
++# CONFIG_BNX2 is not set
++
++#
++# Ethernet (10000 Mbit)
++#
++# CONFIG_CHELSIO_T1 is not set
++# CONFIG_IXGB is not set
++# CONFIG_S2IO is not set
++# CONFIG_MYRI10GE is not set
++# CONFIG_SFC is not set
++
++#
++# Token Ring devices
++#
++# CONFIG_TR is not set
++
++#
++# Wireless LAN (non-hamradio)
++#
++# CONFIG_NET_RADIO is not set
++
++#
++# Wan interfaces
++#
++# CONFIG_WAN is not set
++# CONFIG_FDDI is not set
++# CONFIG_HIPPI is not set
++# CONFIG_PPP is not set
++# CONFIG_SLIP is not set
++# CONFIG_NET_FC is not set
++# CONFIG_SHAPER is not set
++CONFIG_NETCONSOLE=y
++CONFIG_NETPOLL=y
++# CONFIG_NETPOLL_RX is not set
++# CONFIG_NETPOLL_TRAP is not set
++CONFIG_NET_POLL_CONTROLLER=y
++
++#
++# ISDN subsystem
++#
++CONFIG_ISDN=m
++
++#
++# Old ISDN4Linux
++#
++# CONFIG_ISDN_I4L is not set
++
++#
++# CAPI subsystem
++#
++# CONFIG_ISDN_CAPI is not set
++
++#
++# Telephony Support
++#
++# CONFIG_PHONE is not set
++
++#
++# Input device support
++#
++CONFIG_INPUT=y
++
++#
++# Userland interfaces
++#
++CONFIG_INPUT_MOUSEDEV=y
++CONFIG_INPUT_MOUSEDEV_PSAUX=y
++CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024
++CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768
++CONFIG_INPUT_JOYDEV=y
++# CONFIG_INPUT_TSDEV is not set
++CONFIG_INPUT_EVDEV=y
++# CONFIG_INPUT_EVBUG is not set
++
++#
++# Input Device Drivers
++#
++CONFIG_INPUT_KEYBOARD=y
++CONFIG_KEYBOARD_ATKBD=y
++# CONFIG_KEYBOARD_SUNKBD is not set
++# CONFIG_KEYBOARD_LKKBD is not set
++# CONFIG_KEYBOARD_XTKBD is not set
++# CONFIG_KEYBOARD_NEWTON is not set
++CONFIG_INPUT_MOUSE=y
++CONFIG_MOUSE_PS2=y
++# CONFIG_MOUSE_SERIAL is not set
++# CONFIG_MOUSE_VSXXXAA is not set
++# CONFIG_INPUT_JOYSTICK is not set
++# CONFIG_INPUT_TOUCHSCREEN is not set
++# CONFIG_INPUT_MISC is not set
++
++#
++# Hardware I/O ports
++#
++CONFIG_SERIO=y
++CONFIG_SERIO_I8042=y
++# CONFIG_SERIO_SERPORT is not set
++# CONFIG_SERIO_PCIPS2 is not set
++CONFIG_SERIO_LIBPS2=y
++# CONFIG_SERIO_RAW is not set
++CONFIG_GAMEPORT=y
++# CONFIG_GAMEPORT_NS558 is not set
++# CONFIG_GAMEPORT_L4 is not set
++# CONFIG_GAMEPORT_EMU10K1 is not set
++# CONFIG_GAMEPORT_FM801 is not set
++
++#
++# Character devices
++#
++CONFIG_VT=y
++CONFIG_VT_CONSOLE=y
++CONFIG_HW_CONSOLE=y
++# CONFIG_VT_HW_CONSOLE_BINDING is not set
++CONFIG_SERIAL_NONSTANDARD=y
++# CONFIG_COMPUTONE is not set
++# CONFIG_ROCKETPORT is not set
++# CONFIG_CYCLADES is not set
++# CONFIG_DIGIEPCA is not set
++# CONFIG_MOXA_INTELLIO is not set
++# CONFIG_MOXA_SMARTIO is not set
++# CONFIG_ISI is not set
++# CONFIG_SYNCLINKMP is not set
++# CONFIG_SYNCLINK_GT is not set
++# CONFIG_N_HDLC is not set
++# CONFIG_SPECIALIX is not set
++# CONFIG_SX is not set
++# CONFIG_RIO is not set
++# CONFIG_STALDRV is not set
++# CONFIG_SGI_SNSC is not set
++# CONFIG_SGI_TIOCX is not set
++
++#
++# Serial drivers
++#
++
++#
++# Non-8250 serial port support
++#
++# CONFIG_SERIAL_SGI_L1_CONSOLE is not set
++# CONFIG_SERIAL_JSM is not set
++CONFIG_UNIX98_PTYS=y
++CONFIG_LEGACY_PTYS=y
++CONFIG_LEGACY_PTY_COUNT=256
++
++#
++# IPMI
++#
++# CONFIG_IPMI_HANDLER is not set
++
++#
++# Watchdog Cards
++#
++# CONFIG_WATCHDOG is not set
++# CONFIG_HW_RANDOM is not set
++CONFIG_EFI_RTC=y
++# CONFIG_DTLK is not set
++# CONFIG_R3964 is not set
++# CONFIG_APPLICOM is not set
++
++#
++# Ftape, the floppy tape device driver
++#
++CONFIG_AGP=y
++# CONFIG_AGP_SIS is not set
++# CONFIG_AGP_VIA is not set
++CONFIG_AGP_I460=y
++# CONFIG_AGP_HP_ZX1 is not set
++# CONFIG_AGP_SGI_TIOCA is not set
++CONFIG_DRM=y
++# CONFIG_DRM_TDFX is not set
++# CONFIG_DRM_R128 is not set
++# CONFIG_DRM_RADEON is not set
++# CONFIG_DRM_MGA is not set
++# CONFIG_DRM_SIS is not set
++# CONFIG_DRM_VIA is not set
++# CONFIG_DRM_SAVAGE is not set
++# CONFIG_RAW_DRIVER is not set
++# CONFIG_HPET is not set
++# CONFIG_HANGCHECK_TIMER is not set
++# CONFIG_MMTIMER is not set
++
++#
++# TPM devices
++#
++# CONFIG_TCG_TPM is not set
++# CONFIG_TELCLOCK is not set
++
++#
++# I2C support
++#
++CONFIG_I2C=y
++CONFIG_I2C_CHARDEV=y
++
++#
++# I2C Algorithms
++#
++CONFIG_I2C_ALGOBIT=y
++CONFIG_I2C_ALGOPCF=y
++# CONFIG_I2C_ALGOPCA is not set
++
++#
++# I2C Hardware Bus support
++#
++# CONFIG_I2C_ALI1535 is not set
++# CONFIG_I2C_ALI1563 is not set
++# CONFIG_I2C_ALI15X3 is not set
++# CONFIG_I2C_AMD756 is not set
++# CONFIG_I2C_AMD8111 is not set
++# CONFIG_I2C_I801 is not set
++# CONFIG_I2C_I810 is not set
++# CONFIG_I2C_PIIX4 is not set
++# CONFIG_I2C_NFORCE2 is not set
++# CONFIG_I2C_OCORES is not set
++# CONFIG_I2C_PARPORT_LIGHT is not set
++# CONFIG_I2C_PROSAVAGE is not set
++# CONFIG_I2C_SAVAGE4 is not set
++# CONFIG_I2C_SIS5595 is not set
++# CONFIG_I2C_SIS630 is not set
++# CONFIG_I2C_SIS96X is not set
++# CONFIG_I2C_STUB is not set
++# CONFIG_I2C_VIA is not set
++# CONFIG_I2C_VIAPRO is not set
++# CONFIG_I2C_VOODOO3 is not set
++# CONFIG_I2C_PCA_ISA is not set
++
++#
++# Miscellaneous I2C Chip support
++#
++# CONFIG_SENSORS_DS1337 is not set
++# CONFIG_SENSORS_DS1374 is not set
++# CONFIG_SENSORS_EEPROM is not set
++# CONFIG_SENSORS_PCF8574 is not set
++# CONFIG_SENSORS_PCA9539 is not set
++# CONFIG_SENSORS_PCF8591 is not set
++# CONFIG_SENSORS_MAX6875 is not set
++# CONFIG_I2C_DEBUG_CORE is not set
++# CONFIG_I2C_DEBUG_ALGO is not set
++# CONFIG_I2C_DEBUG_BUS is not set
++# CONFIG_I2C_DEBUG_CHIP is not set
++
++#
++# SPI support
++#
++# CONFIG_SPI is not set
++# CONFIG_SPI_MASTER is not set
++
++#
++# Dallas's 1-wire bus
++#
++
++#
++# Hardware Monitoring support
++#
++CONFIG_HWMON=y
++# CONFIG_HWMON_VID is not set
++# CONFIG_SENSORS_ABITUGURU is not set
++# CONFIG_SENSORS_ADM1021 is not set
++# CONFIG_SENSORS_ADM1025 is not set
++# CONFIG_SENSORS_ADM1026 is not set
++# CONFIG_SENSORS_ADM1031 is not set
++# CONFIG_SENSORS_ADM9240 is not set
++# CONFIG_SENSORS_ASB100 is not set
++# CONFIG_SENSORS_ATXP1 is not set
++# CONFIG_SENSORS_DS1621 is not set
++# CONFIG_SENSORS_F71805F is not set
++# CONFIG_SENSORS_FSCHER is not set
++# CONFIG_SENSORS_FSCPOS is not set
++# CONFIG_SENSORS_GL518SM is not set
++# CONFIG_SENSORS_GL520SM is not set
++# CONFIG_SENSORS_IT87 is not set
++# CONFIG_SENSORS_LM63 is not set
++# CONFIG_SENSORS_LM75 is not set
++# CONFIG_SENSORS_LM77 is not set
++# CONFIG_SENSORS_LM78 is not set
++# CONFIG_SENSORS_LM80 is not set
++# CONFIG_SENSORS_LM83 is not set
++# CONFIG_SENSORS_LM85 is not set
++# CONFIG_SENSORS_LM87 is not set
++# CONFIG_SENSORS_LM90 is not set
++# CONFIG_SENSORS_LM92 is not set
++# CONFIG_SENSORS_MAX1619 is not set
++# CONFIG_SENSORS_PC87360 is not set
++# CONFIG_SENSORS_SIS5595 is not set
++# CONFIG_SENSORS_SMSC47M1 is not set
++# CONFIG_SENSORS_SMSC47M192 is not set
++# CONFIG_SENSORS_SMSC47B397 is not set
++# CONFIG_SENSORS_VIA686A is not set
++# CONFIG_SENSORS_VT8231 is not set
++# CONFIG_SENSORS_W83781D is not set
++# CONFIG_SENSORS_W83791D is not set
++# CONFIG_SENSORS_W83792D is not set
++# CONFIG_SENSORS_W83L785TS is not set
++# CONFIG_SENSORS_W83627HF is not set
++# CONFIG_SENSORS_W83627EHF is not set
++# CONFIG_HWMON_DEBUG_CHIP is not set
++
++#
++# Misc devices
++#
++
++#
++# Multimedia devices
++#
++CONFIG_VIDEO_DEV=y
++CONFIG_VIDEO_V4L1=y
++CONFIG_VIDEO_V4L1_COMPAT=y
++CONFIG_VIDEO_V4L2=y
++
++#
++# Video Capture Adapters
++#
++
++#
++# Video Capture Adapters
++#
++# CONFIG_VIDEO_ADV_DEBUG is not set
++# CONFIG_VIDEO_VIVI is not set
++# CONFIG_VIDEO_BT848 is not set
++# CONFIG_VIDEO_CPIA is not set
++# CONFIG_VIDEO_CPIA2 is not set
++# CONFIG_VIDEO_SAA5246A is not set
++# CONFIG_VIDEO_SAA5249 is not set
++# CONFIG_TUNER_3036 is not set
++# CONFIG_VIDEO_STRADIS is not set
++# CONFIG_VIDEO_ZORAN is not set
++# CONFIG_VIDEO_SAA7134 is not set
++# CONFIG_VIDEO_MXB is not set
++# CONFIG_VIDEO_DPC is not set
++# CONFIG_VIDEO_HEXIUM_ORION is not set
++# CONFIG_VIDEO_HEXIUM_GEMINI is not set
++# CONFIG_VIDEO_CX88 is not set
++
++#
++# Encoders and Decoders
++#
++# CONFIG_VIDEO_MSP3400 is not set
++# CONFIG_VIDEO_CS53L32A is not set
++# CONFIG_VIDEO_TLV320AIC23B is not set
++# CONFIG_VIDEO_WM8775 is not set
++# CONFIG_VIDEO_WM8739 is not set
++# CONFIG_VIDEO_CX2341X is not set
++# CONFIG_VIDEO_CX25840 is not set
++# CONFIG_VIDEO_SAA711X is not set
++# CONFIG_VIDEO_SAA7127 is not set
++# CONFIG_VIDEO_UPD64031A is not set
++# CONFIG_VIDEO_UPD64083 is not set
++
++#
++# V4L USB devices
++#
++# CONFIG_VIDEO_PVRUSB2 is not set
++# CONFIG_VIDEO_EM28XX is not set
++# CONFIG_USB_VICAM is not set
++# CONFIG_USB_IBMCAM is not set
++# CONFIG_USB_KONICAWC is not set
++# CONFIG_USB_QUICKCAM_MESSENGER is not set
++# CONFIG_USB_ET61X251 is not set
++# CONFIG_VIDEO_OVCAMCHIP is not set
++# CONFIG_USB_W9968CF is not set
++# CONFIG_USB_OV511 is not set
++# CONFIG_USB_SE401 is not set
++# CONFIG_USB_SN9C102 is not set
++# CONFIG_USB_STV680 is not set
++# CONFIG_USB_ZC0301 is not set
++# CONFIG_USB_PWC is not set
++
++#
++# Radio Adapters
++#
++# CONFIG_RADIO_GEMTEK_PCI is not set
++# CONFIG_RADIO_MAXIRADIO is not set
++# CONFIG_RADIO_MAESTRO is not set
++# CONFIG_USB_DSBR is not set
++
++#
++# Digital Video Broadcasting Devices
++#
++# CONFIG_DVB is not set
++# CONFIG_USB_DABUSB is not set
++
++#
++# Graphics support
++#
++CONFIG_FIRMWARE_EDID=y
++CONFIG_FB=y
++CONFIG_FB_CFB_FILLRECT=y
++CONFIG_FB_CFB_COPYAREA=y
++CONFIG_FB_CFB_IMAGEBLIT=y
++# CONFIG_FB_MACMODES is not set
++# CONFIG_FB_BACKLIGHT is not set
++CONFIG_FB_MODE_HELPERS=y
++# CONFIG_FB_TILEBLITTING is not set
++# CONFIG_FB_CIRRUS is not set
++# CONFIG_FB_PM2 is not set
++# CONFIG_FB_CYBER2000 is not set
++# CONFIG_FB_ASILIANT is not set
++# CONFIG_FB_IMSTT is not set
++# CONFIG_FB_S1D13XXX is not set
++# CONFIG_FB_NVIDIA is not set
++# CONFIG_FB_RIVA is not set
++# CONFIG_FB_MATROX is not set
++# CONFIG_FB_RADEON is not set
++# CONFIG_FB_ATY128 is not set
++# CONFIG_FB_ATY is not set
++# CONFIG_FB_SAVAGE is not set
++# CONFIG_FB_SIS is not set
++# CONFIG_FB_NEOMAGIC is not set
++# CONFIG_FB_KYRO is not set
++# CONFIG_FB_3DFX is not set
++# CONFIG_FB_VOODOO1 is not set
++# CONFIG_FB_TRIDENT is not set
++# CONFIG_FB_VIRTUAL is not set
++
++#
++# Console display driver support
++#
++CONFIG_VGA_CONSOLE=y
++# CONFIG_VGACON_SOFT_SCROLLBACK is not set
++CONFIG_DUMMY_CONSOLE=y
++CONFIG_FRAMEBUFFER_CONSOLE=y
++# CONFIG_FRAMEBUFFER_CONSOLE_ROTATION is not set
++# CONFIG_FONTS is not set
++CONFIG_FONT_8x8=y
++CONFIG_FONT_8x16=y
++
++#
++# Logo configuration
++#
++CONFIG_LOGO=y
++# CONFIG_LOGO_LINUX_MONO is not set
++# CONFIG_LOGO_LINUX_VGA16 is not set
++CONFIG_LOGO_LINUX_CLUT224=y
++# CONFIG_BACKLIGHT_LCD_SUPPORT is not set
++
++#
++# Sound
++#
++CONFIG_SOUND=y
++
++#
++# Advanced Linux Sound Architecture
++#
++CONFIG_SND=y
++CONFIG_SND_TIMER=y
++CONFIG_SND_PCM=y
++CONFIG_SND_HWDEP=y
++CONFIG_SND_RAWMIDI=y
++CONFIG_SND_SEQUENCER=y
++CONFIG_SND_SEQ_DUMMY=y
++CONFIG_SND_OSSEMUL=y
++CONFIG_SND_MIXER_OSS=y
++CONFIG_SND_PCM_OSS=y
++CONFIG_SND_PCM_OSS_PLUGINS=y
++CONFIG_SND_SEQUENCER_OSS=y
++# CONFIG_SND_DYNAMIC_MINORS is not set
++CONFIG_SND_SUPPORT_OLD_API=y
++CONFIG_SND_VERBOSE_PROCFS=y
++# CONFIG_SND_VERBOSE_PRINTK is not set
++# CONFIG_SND_DEBUG is not set
++
++#
++# Generic devices
++#
++CONFIG_SND_MPU401_UART=y
++CONFIG_SND_OPL3_LIB=y
++CONFIG_SND_AC97_CODEC=y
++CONFIG_SND_AC97_BUS=y
++CONFIG_SND_DUMMY=y
++CONFIG_SND_VIRMIDI=y
++# CONFIG_SND_MTPAV is not set
++CONFIG_SND_SERIAL_U16550=y
++CONFIG_SND_MPU401=y
++
++#
++# PCI devices
++#
++# CONFIG_SND_AD1889 is not set
++# CONFIG_SND_ALS300 is not set
++# CONFIG_SND_ALI5451 is not set
++CONFIG_SND_ATIIXP=y
++# CONFIG_SND_ATIIXP_MODEM is not set
++# CONFIG_SND_AU8810 is not set
++# CONFIG_SND_AU8820 is not set
++# CONFIG_SND_AU8830 is not set
++# CONFIG_SND_AZT3328 is not set
++# CONFIG_SND_BT87X is not set
++# CONFIG_SND_CA0106 is not set
++# CONFIG_SND_CMIPCI is not set
++# CONFIG_SND_CS4281 is not set
++# CONFIG_SND_CS46XX is not set
++# CONFIG_SND_DARLA20 is not set
++# CONFIG_SND_GINA20 is not set
++# CONFIG_SND_LAYLA20 is not set
++# CONFIG_SND_DARLA24 is not set
++# CONFIG_SND_GINA24 is not set
++# CONFIG_SND_LAYLA24 is not set
++# CONFIG_SND_MONA is not set
++# CONFIG_SND_MIA is not set
++# CONFIG_SND_ECHO3G is not set
++# CONFIG_SND_INDIGO is not set
++# CONFIG_SND_INDIGOIO is not set
++# CONFIG_SND_INDIGODJ is not set
++# CONFIG_SND_EMU10K1 is not set
++# CONFIG_SND_EMU10K1X is not set
++# CONFIG_SND_ENS1370 is not set
++# CONFIG_SND_ENS1371 is not set
++# CONFIG_SND_ES1938 is not set
++# CONFIG_SND_ES1968 is not set
++CONFIG_SND_FM801=y
++# CONFIG_SND_FM801_TEA575X_BOOL is not set
++# CONFIG_SND_HDA_INTEL is not set
++# CONFIG_SND_HDSP is not set
++# CONFIG_SND_HDSPM is not set
++# CONFIG_SND_ICE1712 is not set
++# CONFIG_SND_ICE1724 is not set
++# CONFIG_SND_INTEL8X0 is not set
++# CONFIG_SND_INTEL8X0M is not set
++# CONFIG_SND_KORG1212 is not set
++# CONFIG_SND_MAESTRO3 is not set
++# CONFIG_SND_MIXART is not set
++# CONFIG_SND_NM256 is not set
++# CONFIG_SND_PCXHR is not set
++# CONFIG_SND_RIPTIDE is not set
++# CONFIG_SND_RME32 is not set
++# CONFIG_SND_RME96 is not set
++# CONFIG_SND_RME9652 is not set
++# CONFIG_SND_SONICVIBES is not set
++# CONFIG_SND_TRIDENT is not set
++# CONFIG_SND_VIA82XX is not set
++# CONFIG_SND_VIA82XX_MODEM is not set
++# CONFIG_SND_VX222 is not set
++# CONFIG_SND_YMFPCI is not set
++
++#
++# USB devices
++#
++# CONFIG_SND_USB_AUDIO is not set
++
++#
++# Open Sound System
++#
++CONFIG_SOUND_PRIME=y
++# CONFIG_OSS_OBSOLETE_DRIVER is not set
++# CONFIG_SOUND_BT878 is not set
++# CONFIG_SOUND_ES1371 is not set
++# CONFIG_SOUND_ICH is not set
++# CONFIG_SOUND_TRIDENT is not set
++# CONFIG_SOUND_MSNDCLAS is not set
++# CONFIG_SOUND_MSNDPIN is not set
++# CONFIG_SOUND_VIA82CXXX is not set
++# CONFIG_SOUND_TVMIXER is not set
++
++#
++# USB support
++#
++CONFIG_USB_ARCH_HAS_HCD=y
++CONFIG_USB_ARCH_HAS_OHCI=y
++CONFIG_USB_ARCH_HAS_EHCI=y
++CONFIG_USB=y
++# CONFIG_USB_DEBUG is not set
++
++#
++# Miscellaneous USB options
++#
++CONFIG_USB_DEVICEFS=y
++CONFIG_USB_BANDWIDTH=y
++# CONFIG_USB_DYNAMIC_MINORS is not set
++# CONFIG_USB_SUSPEND is not set
++# CONFIG_USB_OTG is not set
++
++#
++# USB Host Controller Drivers
++#
++CONFIG_USB_EHCI_HCD=y
++# CONFIG_USB_EHCI_SPLIT_ISO is not set
++# CONFIG_USB_EHCI_ROOT_HUB_TT is not set
++# CONFIG_USB_EHCI_TT_NEWSCHED is not set
++# CONFIG_USB_ISP116X_HCD is not set
++CONFIG_USB_OHCI_HCD=y
++# CONFIG_USB_OHCI_BIG_ENDIAN is not set
++CONFIG_USB_OHCI_LITTLE_ENDIAN=y
++CONFIG_USB_UHCI_HCD=y
++# CONFIG_USB_SL811_HCD is not set
++
++#
++# USB Device Class drivers
++#
++# CONFIG_USB_ACM is not set
++# CONFIG_USB_PRINTER is not set
++
++#
++# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support'
++#
++
++#
++# may also be needed; see USB_STORAGE Help for more information
++#
++CONFIG_USB_STORAGE=y
++# CONFIG_USB_STORAGE_DEBUG is not set
++# CONFIG_USB_STORAGE_DATAFAB is not set
++# CONFIG_USB_STORAGE_FREECOM is not set
++# CONFIG_USB_STORAGE_ISD200 is not set
++# CONFIG_USB_STORAGE_DPCM is not set
++# CONFIG_USB_STORAGE_USBAT is not set
++# CONFIG_USB_STORAGE_SDDR09 is not set
++# CONFIG_USB_STORAGE_SDDR55 is not set
++# CONFIG_USB_STORAGE_JUMPSHOT is not set
++# CONFIG_USB_STORAGE_ALAUDA is not set
++# CONFIG_USB_LIBUSUAL is not set
++
++#
++# USB Input Devices
++#
++CONFIG_USB_HID=y
++CONFIG_USB_HIDINPUT=y
++# CONFIG_USB_HIDINPUT_POWERBOOK is not set
++# CONFIG_HID_FF is not set
++CONFIG_USB_HIDDEV=y
++# CONFIG_USB_AIPTEK is not set
++# CONFIG_USB_WACOM is not set
++# CONFIG_USB_ACECAD is not set
++# CONFIG_USB_KBTAB is not set
++# CONFIG_USB_POWERMATE is not set
++# CONFIG_USB_TOUCHSCREEN is not set
++# CONFIG_USB_YEALINK is not set
++# CONFIG_USB_XPAD is not set
++# CONFIG_USB_ATI_REMOTE is not set
++# CONFIG_USB_ATI_REMOTE2 is not set
++# CONFIG_USB_KEYSPAN_REMOTE is not set
++# CONFIG_USB_APPLETOUCH is not set
++
++#
++# USB Imaging devices
++#
++# CONFIG_USB_MDC800 is not set
++# CONFIG_USB_MICROTEK is not set
++
++#
++# USB Network Adapters
++#
++# CONFIG_USB_CATC is not set
++# CONFIG_USB_KAWETH is not set
++# CONFIG_USB_PEGASUS is not set
++# CONFIG_USB_RTL8150 is not set
++# CONFIG_USB_USBNET is not set
++CONFIG_USB_MON=y
++
++#
++# USB port drivers
++#
++
++#
++# USB Serial Converter support
++#
++# CONFIG_USB_SERIAL is not set
++
++#
++# USB Miscellaneous drivers
++#
++# CONFIG_USB_EMI62 is not set
++# CONFIG_USB_EMI26 is not set
++# CONFIG_USB_AUERSWALD is not set
++# CONFIG_USB_RIO500 is not set
++# CONFIG_USB_LEGOTOWER is not set
++# CONFIG_USB_LCD is not set
++# CONFIG_USB_LED is not set
++# CONFIG_USB_CYPRESS_CY7C63 is not set
++# CONFIG_USB_CYTHERM is not set
++# CONFIG_USB_PHIDGETKIT is not set
++# CONFIG_USB_PHIDGETSERVO is not set
++# CONFIG_USB_IDMOUSE is not set
++# CONFIG_USB_APPLEDISPLAY is not set
++# CONFIG_USB_SISUSBVGA is not set
++# CONFIG_USB_LD is not set
++# CONFIG_USB_TEST is not set
++
++#
++# USB DSL modem support
++#
++
++#
++# USB Gadget Support
++#
++# CONFIG_USB_GADGET is not set
++
++#
++# MMC/SD Card support
++#
++# CONFIG_MMC is not set
++
++#
++# LED devices
++#
++# CONFIG_NEW_LEDS is not set
++
++#
++# LED drivers
++#
++
++#
++# LED Triggers
++#
++
++#
++# InfiniBand support
++#
++# CONFIG_INFINIBAND is not set
++
++#
++# EDAC - error detection and reporting (RAS) (EXPERIMENTAL)
++#
++
++#
++# Real Time Clock
++#
++# CONFIG_RTC_CLASS is not set
++
++#
++# DMA Engine support
++#
++# CONFIG_DMA_ENGINE is not set
++
++#
++# DMA Clients
++#
++
++#
++# DMA Devices
++#
++
++#
++# File systems
++#
++CONFIG_EXT2_FS=y
++CONFIG_EXT2_FS_XATTR=y
++CONFIG_EXT2_FS_POSIX_ACL=y
++CONFIG_EXT2_FS_SECURITY=y
++# CONFIG_EXT2_FS_XIP is not set
++CONFIG_EXT3_FS=y
++CONFIG_EXT3_FS_XATTR=y
++CONFIG_EXT3_FS_POSIX_ACL=y
++CONFIG_EXT3_FS_SECURITY=y
++CONFIG_JBD=y
++# CONFIG_JBD_DEBUG is not set
++CONFIG_FS_MBCACHE=y
++CONFIG_REISERFS_FS=y
++# CONFIG_REISERFS_CHECK is not set
++# CONFIG_REISERFS_PROC_INFO is not set
++CONFIG_REISERFS_FS_XATTR=y
++CONFIG_REISERFS_FS_POSIX_ACL=y
++CONFIG_REISERFS_FS_SECURITY=y
++# CONFIG_JFS_FS is not set
++CONFIG_FS_POSIX_ACL=y
++CONFIG_XFS_FS=y
++# CONFIG_XFS_QUOTA is not set
++# CONFIG_XFS_SECURITY is not set
++# CONFIG_XFS_POSIX_ACL is not set
++# CONFIG_XFS_RT is not set
++# CONFIG_OCFS2_FS is not set
++# CONFIG_MINIX_FS is not set
++# CONFIG_ROMFS_FS is not set
++CONFIG_INOTIFY=y
++CONFIG_INOTIFY_USER=y
++# CONFIG_QUOTA is not set
++CONFIG_DNOTIFY=y
++CONFIG_AUTOFS_FS=y
++CONFIG_AUTOFS4_FS=y
++# CONFIG_FUSE_FS is not set
++
++#
++# CD-ROM/DVD Filesystems
++#
++CONFIG_ISO9660_FS=y
++CONFIG_JOLIET=y
++# CONFIG_ZISOFS is not set
++CONFIG_UDF_FS=y
++CONFIG_UDF_NLS=y
++
++#
++# DOS/FAT/NT Filesystems
++#
++CONFIG_FAT_FS=y
++CONFIG_MSDOS_FS=y
++CONFIG_VFAT_FS=y
++CONFIG_FAT_DEFAULT_CODEPAGE=437
++CONFIG_FAT_DEFAULT_IOCHARSET="iso8859-1"
++# CONFIG_NTFS_FS is not set
++
++#
++# Pseudo filesystems
++#
++CONFIG_PROC_FS=y
++CONFIG_PROC_KCORE=y
++CONFIG_PROC_IOMEM_MACHINE=y
++CONFIG_SYSFS=y
++CONFIG_TMPFS=y
++# CONFIG_HUGETLB_PAGE is not set
++CONFIG_RAMFS=y
++# CONFIG_CONFIGFS_FS is not set
++
++#
++# Miscellaneous filesystems
++#
++# CONFIG_ADFS_FS is not set
++# CONFIG_AFFS_FS is not set
++# CONFIG_HFS_FS is not set
++# CONFIG_HFSPLUS_FS is not set
++# CONFIG_BEFS_FS is not set
++# CONFIG_BFS_FS is not set
++# CONFIG_EFS_FS is not set
++# CONFIG_CRAMFS is not set
++# CONFIG_VXFS_FS is not set
++# CONFIG_HPFS_FS is not set
++# CONFIG_QNX4FS_FS is not set
++# CONFIG_SYSV_FS is not set
++# CONFIG_UFS_FS is not set
++
++#
++# Network File Systems
++#
++CONFIG_NFS_FS=y
++CONFIG_NFS_V3=y
++# CONFIG_NFS_V3_ACL is not set
++CONFIG_NFS_V4=y
++CONFIG_NFS_DIRECTIO=y
++CONFIG_NFSD=y
++CONFIG_NFSD_V3=y
++# CONFIG_NFSD_V3_ACL is not set
++CONFIG_NFSD_V4=y
++CONFIG_NFSD_TCP=y
++CONFIG_ROOT_NFS=y
++CONFIG_LOCKD=y
++CONFIG_LOCKD_V4=y
++CONFIG_EXPORTFS=y
++CONFIG_NFS_COMMON=y
++CONFIG_SUNRPC=y
++CONFIG_SUNRPC_GSS=y
++CONFIG_RPCSEC_GSS_KRB5=y
++# CONFIG_RPCSEC_GSS_SPKM3 is not set
++CONFIG_SMB_FS=y
++CONFIG_SMB_NLS_DEFAULT=y
++CONFIG_SMB_NLS_REMOTE="cp437"
++CONFIG_CIFS=y
++# CONFIG_CIFS_STATS is not set
++# CONFIG_CIFS_WEAK_PW_HASH is not set
++# CONFIG_CIFS_XATTR is not set
++# CONFIG_CIFS_DEBUG2 is not set
++# CONFIG_CIFS_EXPERIMENTAL is not set
++# CONFIG_NCP_FS is not set
++# CONFIG_CODA_FS is not set
++# CONFIG_AFS_FS is not set
++# CONFIG_9P_FS is not set
++
++#
++# Partition Types
++#
++CONFIG_PARTITION_ADVANCED=y
++# CONFIG_ACORN_PARTITION is not set
++# CONFIG_OSF_PARTITION is not set
++# CONFIG_AMIGA_PARTITION is not set
++# CONFIG_ATARI_PARTITION is not set
++# CONFIG_MAC_PARTITION is not set
++CONFIG_MSDOS_PARTITION=y
++# CONFIG_BSD_DISKLABEL is not set
++# CONFIG_MINIX_SUBPARTITION is not set
++# CONFIG_SOLARIS_X86_PARTITION is not set
++# CONFIG_UNIXWARE_DISKLABEL is not set
++# CONFIG_LDM_PARTITION is not set
++CONFIG_SGI_PARTITION=y
++# CONFIG_ULTRIX_PARTITION is not set
++# CONFIG_SUN_PARTITION is not set
++# CONFIG_KARMA_PARTITION is not set
++CONFIG_EFI_PARTITION=y
++
++#
++# Native Language Support
++#
++CONFIG_NLS=y
++CONFIG_NLS_DEFAULT="iso8859-1"
++CONFIG_NLS_CODEPAGE_437=y
++CONFIG_NLS_CODEPAGE_737=y
++CONFIG_NLS_CODEPAGE_775=y
++CONFIG_NLS_CODEPAGE_850=y
++CONFIG_NLS_CODEPAGE_852=y
++CONFIG_NLS_CODEPAGE_855=y
++CONFIG_NLS_CODEPAGE_857=y
++CONFIG_NLS_CODEPAGE_860=y
++CONFIG_NLS_CODEPAGE_861=y
++CONFIG_NLS_CODEPAGE_862=y
++CONFIG_NLS_CODEPAGE_863=y
++CONFIG_NLS_CODEPAGE_864=y
++CONFIG_NLS_CODEPAGE_865=y
++CONFIG_NLS_CODEPAGE_866=y
++CONFIG_NLS_CODEPAGE_869=y
++CONFIG_NLS_CODEPAGE_936=y
++CONFIG_NLS_CODEPAGE_950=y
++CONFIG_NLS_CODEPAGE_932=y
++CONFIG_NLS_CODEPAGE_949=y
++CONFIG_NLS_CODEPAGE_874=y
++CONFIG_NLS_ISO8859_8=y
++# CONFIG_NLS_CODEPAGE_1250 is not set
++CONFIG_NLS_CODEPAGE_1251=y
++# CONFIG_NLS_ASCII is not set
++CONFIG_NLS_ISO8859_1=y
++CONFIG_NLS_ISO8859_2=y
++CONFIG_NLS_ISO8859_3=y
++CONFIG_NLS_ISO8859_4=y
++CONFIG_NLS_ISO8859_5=y
++CONFIG_NLS_ISO8859_6=y
++CONFIG_NLS_ISO8859_7=y
++CONFIG_NLS_ISO8859_9=y
++CONFIG_NLS_ISO8859_13=y
++CONFIG_NLS_ISO8859_14=y
++CONFIG_NLS_ISO8859_15=y
++CONFIG_NLS_KOI8_R=y
++CONFIG_NLS_KOI8_U=y
++CONFIG_NLS_UTF8=y
++
++#
++# Library routines
++#
++# CONFIG_CRC_CCITT is not set
++# CONFIG_CRC16 is not set
++CONFIG_CRC32=y
++# CONFIG_LIBCRC32C is not set
++CONFIG_PLIST=y
++CONFIG_GENERIC_HARDIRQS=y
++CONFIG_GENERIC_IRQ_PROBE=y
++CONFIG_GENERIC_PENDING_IRQ=y
++CONFIG_IRQ_PER_CPU=y
++
++#
++# HP Simulator drivers
++#
++# CONFIG_HP_SIMETH is not set
++# CONFIG_HP_SIMSERIAL is not set
++# CONFIG_HP_SIMSCSI is not set
++
++#
++# Instrumentation Support
++#
++# CONFIG_PROFILING is not set
++# CONFIG_KPROBES is not set
++
++#
++# Kernel hacking
++#
++# CONFIG_PRINTK_TIME is not set
++CONFIG_MAGIC_SYSRQ=y
++CONFIG_UNUSED_SYMBOLS=y
++CONFIG_DEBUG_KERNEL=y
++CONFIG_LOG_BUF_SHIFT=20
++CONFIG_DETECT_SOFTLOCKUP=y
++# CONFIG_SCHEDSTATS is not set
++# CONFIG_DEBUG_SLAB is not set
++# CONFIG_DEBUG_RT_MUTEXES is not set
++# CONFIG_RT_MUTEX_TESTER is not set
++# CONFIG_DEBUG_SPINLOCK is not set
++CONFIG_DEBUG_MUTEXES=y
++# CONFIG_DEBUG_RWSEMS is not set
++# CONFIG_DEBUG_SPINLOCK_SLEEP is not set
++# CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set
++# CONFIG_DEBUG_KOBJECT is not set
++# CONFIG_DEBUG_INFO is not set
++# CONFIG_DEBUG_FS is not set
++# CONFIG_DEBUG_VM is not set
++CONFIG_FORCED_INLINING=y
++# CONFIG_RCU_TORTURE_TEST is not set
++CONFIG_IA64_GRANULE_16MB=y
++# CONFIG_IA64_GRANULE_64MB is not set
++CONFIG_IA64_PRINT_HAZARDS=y
++# CONFIG_DISABLE_VHPT is not set
++# CONFIG_IA64_DEBUG_CMPXCHG is not set
++# CONFIG_IA64_DEBUG_IRQ is not set
++
++#
++# Security options
++#
++# CONFIG_KEYS is not set
++# CONFIG_SECURITY is not set
++
++#
++# Cryptographic options
++#
++CONFIG_CRYPTO=y
++# CONFIG_CRYPTO_HMAC is not set
++# CONFIG_CRYPTO_NULL is not set
++# CONFIG_CRYPTO_MD4 is not set
++CONFIG_CRYPTO_MD5=y
++# CONFIG_CRYPTO_SHA1 is not set
++# CONFIG_CRYPTO_SHA256 is not set
++# CONFIG_CRYPTO_SHA512 is not set
++# CONFIG_CRYPTO_WP512 is not set
++# CONFIG_CRYPTO_TGR192 is not set
++CONFIG_CRYPTO_DES=y
++# CONFIG_CRYPTO_BLOWFISH is not set
++# CONFIG_CRYPTO_TWOFISH is not set
++# CONFIG_CRYPTO_SERPENT is not set
++# CONFIG_CRYPTO_AES is not set
++# CONFIG_CRYPTO_CAST5 is not set
++# CONFIG_CRYPTO_CAST6 is not set
++# CONFIG_CRYPTO_TEA is not set
++# CONFIG_CRYPTO_ARC4 is not set
++# CONFIG_CRYPTO_KHAZAD is not set
++# CONFIG_CRYPTO_ANUBIS is not set
++# CONFIG_CRYPTO_DEFLATE is not set
++# CONFIG_CRYPTO_MICHAEL_MIC is not set
++# CONFIG_CRYPTO_CRC32C is not set
++# CONFIG_CRYPTO_TEST is not set
++
++#
++# Hardware crypto devices
++#
++# CONFIG_XEN_SMPBOOT is not set
++# CONFIG_XEN_DEVMEM is not set
++CONFIG_XEN_INTERFACE_VERSION=0x00030207
++
++#
++# XEN
++#
++CONFIG_XEN_PRIVILEGED_GUEST=y
++# CONFIG_XEN_UNPRIVILEGED_GUEST is not set
++CONFIG_XEN_PRIVCMD=y
++CONFIG_XEN_XENBUS_DEV=y
++CONFIG_XEN_BACKEND=y
++CONFIG_XEN_BLKDEV_BACKEND=y
++CONFIG_XEN_BLKDEV_TAP=y
++CONFIG_XEN_NETDEV_BACKEND=y
++# CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_UTIL is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_BACKEND is not set
++# CONFIG_XEN_NETDEV_LOOPBACK is not set
++CONFIG_XEN_PCIDEV_BACKEND=y
++# CONFIG_XEN_PCIDEV_BACKEND_VPCI is not set
++# CONFIG_XEN_PCIDEV_BACKEND_PASS is not set
++# CONFIG_XEN_PCIDEV_BACKEND_SLOT is not set
++CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER=y
++# CONFIG_XEN_PCIDEV_BE_DEBUG is not set
++CONFIG_XEN_TPMDEV_BACKEND=m
++CONFIG_XEN_SCSI_BACKEND=m
++CONFIG_XEN_BLKDEV_FRONTEND=y
++CONFIG_XEN_NETDEV_FRONTEND=y
++CONFIG_XEN_SCSI_FRONTEND=m
++CONFIG_XEN_GRANT_DEV=y
++# CONFIG_XEN_NETDEV_ACCEL_SFC_FRONTEND is not set
++CONFIG_XEN_FRAMEBUFFER=y
++CONFIG_XEN_KEYBOARD=y
++# CONFIG_XEN_SCRUB_PAGES is not set
++CONFIG_XEN_DISABLE_SERIAL=y
++CONFIG_XEN_SYSFS=y
++CONFIG_XEN_COMPAT_030002_AND_LATER=y
++# CONFIG_XEN_COMPAT_030004_AND_LATER is not set
++# CONFIG_XEN_COMPAT_030100_AND_LATER is not set
++# CONFIG_XEN_COMPAT_LATEST_ONLY is not set
++CONFIG_XEN_COMPAT=0x030002
++CONFIG_HAVE_IRQ_IGNORE_UNHANDLED=y
++CONFIG_NO_IDLE_HZ=y
++CONFIG_XEN_BALLOON=y
++CONFIG_XEN_XENCOMM=y
+diff -r d894e36cfc30 -r 0aa021803deb buildconfigs/linux-defconfig_xen0_x86_32
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/buildconfigs/linux-defconfig_xen0_x86_32 Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1458 @@
++#
++# Automatically generated make config: don't edit
++# Linux kernel version: 2.6.18.8
++# Tue Oct 16 09:31:19 2007
++#
++CONFIG_X86_32=y
++CONFIG_LOCKDEP_SUPPORT=y
++CONFIG_STACKTRACE_SUPPORT=y
++CONFIG_SEMAPHORE_SLEEPERS=y
++CONFIG_X86=y
++CONFIG_MMU=y
++CONFIG_GENERIC_ISA_DMA=y
++CONFIG_GENERIC_IOMAP=y
++CONFIG_GENERIC_HWEIGHT=y
++CONFIG_ARCH_MAY_HAVE_PC_FDC=y
++CONFIG_DMI=y
++CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config"
++
++#
++# Code maturity level options
++#
++CONFIG_EXPERIMENTAL=y
++CONFIG_LOCK_KERNEL=y
++CONFIG_INIT_ENV_ARG_LIMIT=32
++
++#
++# General setup
++#
++CONFIG_LOCALVERSION=""
++# CONFIG_LOCALVERSION_AUTO is not set
++CONFIG_SWAP=y
++CONFIG_SYSVIPC=y
++# CONFIG_POSIX_MQUEUE is not set
++# CONFIG_BSD_PROCESS_ACCT is not set
++# CONFIG_TASKSTATS is not set
++# CONFIG_AUDIT is not set
++# CONFIG_IKCONFIG is not set
++# CONFIG_CPUSETS is not set
++# CONFIG_RELAY is not set
++CONFIG_INITRAMFS_SOURCE=""
++# CONFIG_CC_OPTIMIZE_FOR_SIZE is not set
++# CONFIG_EMBEDDED is not set
++CONFIG_UID16=y
++CONFIG_SYSCTL=y
++CONFIG_KALLSYMS=y
++# CONFIG_KALLSYMS_ALL is not set
++# CONFIG_KALLSYMS_EXTRA_PASS is not set
++CONFIG_HOTPLUG=y
++CONFIG_PRINTK=y
++CONFIG_BUG=y
++CONFIG_ELF_CORE=y
++CONFIG_BASE_FULL=y
++CONFIG_FUTEX=y
++CONFIG_EPOLL=y
++CONFIG_SHMEM=y
++CONFIG_SLAB=y
++CONFIG_VM_EVENT_COUNTERS=y
++CONFIG_RT_MUTEXES=y
++# CONFIG_TINY_SHMEM is not set
++CONFIG_BASE_SMALL=0
++# CONFIG_SLOB is not set
++
++#
++# Loadable module support
++#
++CONFIG_MODULES=y
++CONFIG_MODULE_UNLOAD=y
++# CONFIG_MODULE_FORCE_UNLOAD is not set
++# CONFIG_MODVERSIONS is not set
++# CONFIG_MODULE_SRCVERSION_ALL is not set
++CONFIG_KMOD=y
++CONFIG_STOP_MACHINE=y
++
++#
++# Block layer
++#
++# CONFIG_LBD is not set
++# CONFIG_BLK_DEV_IO_TRACE is not set
++CONFIG_LSF=y
++
++#
++# IO Schedulers
++#
++CONFIG_IOSCHED_NOOP=y
++CONFIG_IOSCHED_AS=y
++CONFIG_IOSCHED_DEADLINE=y
++CONFIG_IOSCHED_CFQ=y
++CONFIG_DEFAULT_AS=y
++# CONFIG_DEFAULT_DEADLINE is not set
++# CONFIG_DEFAULT_CFQ is not set
++# CONFIG_DEFAULT_NOOP is not set
++CONFIG_DEFAULT_IOSCHED="anticipatory"
++
++#
++# Processor type and features
++#
++CONFIG_SMP=y
++# CONFIG_X86_PC is not set
++CONFIG_X86_XEN=y
++# CONFIG_X86_ELAN is not set
++# CONFIG_X86_VOYAGER is not set
++# CONFIG_X86_NUMAQ is not set
++# CONFIG_X86_SUMMIT is not set
++# CONFIG_X86_BIGSMP is not set
++# CONFIG_X86_VISWS is not set
++# CONFIG_X86_GENERICARCH is not set
++# CONFIG_X86_ES7000 is not set
++# CONFIG_M386 is not set
++# CONFIG_M486 is not set
++# CONFIG_M586 is not set
++# CONFIG_M586TSC is not set
++# CONFIG_M586MMX is not set
++CONFIG_M686=y
++# CONFIG_MPENTIUMII is not set
++# CONFIG_MPENTIUMIII is not set
++# CONFIG_MPENTIUMM is not set
++# CONFIG_MPENTIUM4 is not set
++# CONFIG_MK6 is not set
++# CONFIG_MK7 is not set
++# CONFIG_MK8 is not set
++# CONFIG_MCRUSOE is not set
++# CONFIG_MEFFICEON is not set
++# CONFIG_MWINCHIPC6 is not set
++# CONFIG_MWINCHIP2 is not set
++# CONFIG_MWINCHIP3D is not set
++# CONFIG_MGEODEGX1 is not set
++# CONFIG_MGEODE_LX is not set
++# CONFIG_MCYRIXIII is not set
++# CONFIG_MVIAC3_2 is not set
++# CONFIG_X86_GENERIC is not set
++CONFIG_X86_CMPXCHG=y
++CONFIG_X86_XADD=y
++CONFIG_X86_L1_CACHE_SHIFT=5
++CONFIG_RWSEM_XCHGADD_ALGORITHM=y
++CONFIG_GENERIC_CALIBRATE_DELAY=y
++CONFIG_X86_PPRO_FENCE=y
++CONFIG_X86_WP_WORKS_OK=y
++CONFIG_X86_INVLPG=y
++CONFIG_X86_BSWAP=y
++CONFIG_X86_POPAD_OK=y
++CONFIG_X86_CMPXCHG64=y
++CONFIG_X86_GOOD_APIC=y
++CONFIG_X86_USE_PPRO_CHECKSUM=y
++CONFIG_NR_CPUS=8
++CONFIG_PREEMPT_NONE=y
++# CONFIG_PREEMPT_VOLUNTARY is not set
++# CONFIG_PREEMPT is not set
++CONFIG_PREEMPT_BKL=y
++CONFIG_X86_LOCAL_APIC=y
++CONFIG_X86_IO_APIC=y
++CONFIG_VM86=y
++# CONFIG_TOSHIBA is not set
++# CONFIG_I8K is not set
++# CONFIG_X86_REBOOTFIXUPS is not set
++CONFIG_MICROCODE=y
++CONFIG_X86_MSR=y
++CONFIG_X86_CPUID=y
++CONFIG_SWIOTLB=y
++
++#
++# Firmware Drivers
++#
++CONFIG_EDD=y
++# CONFIG_DELL_RBU is not set
++# CONFIG_DCDBAS is not set
++# CONFIG_NOHIGHMEM is not set
++CONFIG_HIGHMEM4G=y
++# CONFIG_HIGHMEM64G is not set
++CONFIG_PAGE_OFFSET=0xC0000000
++CONFIG_HIGHMEM=y
++CONFIG_SELECT_MEMORY_MODEL=y
++CONFIG_FLATMEM_MANUAL=y
++# CONFIG_DISCONTIGMEM_MANUAL is not set
++# CONFIG_SPARSEMEM_MANUAL is not set
++CONFIG_FLATMEM=y
++CONFIG_FLAT_NODE_MEM_MAP=y
++# CONFIG_SPARSEMEM_STATIC is not set
++CONFIG_SPLIT_PTLOCK_CPUS=4
++CONFIG_RESOURCES_64BIT=y
++# CONFIG_HIGHPTE is not set
++CONFIG_MTRR=y
++# CONFIG_REGPARM is not set
++CONFIG_SECCOMP=y
++CONFIG_HZ_100=y
++# CONFIG_HZ_250 is not set
++# CONFIG_HZ_1000 is not set
++CONFIG_HZ=100
++CONFIG_KEXEC=y
++# CONFIG_CRASH_DUMP is not set
++CONFIG_PHYSICAL_START=0x100000
++CONFIG_HOTPLUG_CPU=y
++CONFIG_ARCH_ENABLE_MEMORY_HOTPLUG=y
++
++#
++# Power management options (ACPI, APM)
++#
++CONFIG_PM=y
++CONFIG_PM_LEGACY=y
++# CONFIG_PM_DEBUG is not set
++# CONFIG_SOFTWARE_SUSPEND is not set
++CONFIG_SUSPEND_SMP=y
++
++#
++# ACPI (Advanced Configuration and Power Interface) Support
++#
++CONFIG_ACPI=y
++CONFIG_ACPI_SLEEP=y
++CONFIG_ACPI_SLEEP_PROC_FS=y
++# CONFIG_ACPI_SLEEP_PROC_SLEEP is not set
++CONFIG_ACPI_AC=m
++CONFIG_ACPI_BATTERY=m
++CONFIG_ACPI_BUTTON=m
++CONFIG_ACPI_VIDEO=m
++CONFIG_ACPI_HOTKEY=m
++CONFIG_ACPI_FAN=m
++CONFIG_ACPI_DOCK=m
++CONFIG_ACPI_PROCESSOR=m
++CONFIG_ACPI_HOTPLUG_CPU=y
++CONFIG_ACPI_THERMAL=m
++CONFIG_ACPI_ASUS=m
++CONFIG_ACPI_IBM=m
++CONFIG_ACPI_TOSHIBA=m
++# CONFIG_ACPI_CUSTOM_DSDT is not set
++CONFIG_ACPI_BLACKLIST_YEAR=0
++# CONFIG_ACPI_DEBUG is not set
++CONFIG_ACPI_EC=y
++CONFIG_ACPI_POWER=y
++CONFIG_ACPI_SYSTEM=y
++CONFIG_ACPI_CONTAINER=m
++CONFIG_ACPI_PV_SLEEP=y
++
++#
++# CPU Frequency scaling
++#
++# CONFIG_CPU_FREQ is not set
++
++#
++# Bus options (PCI, PCMCIA, EISA, MCA, ISA)
++#
++CONFIG_PCI=y
++# CONFIG_PCI_GOBIOS is not set
++# CONFIG_PCI_GOMMCONFIG is not set
++# CONFIG_PCI_GODIRECT is not set
++# CONFIG_PCI_GOXEN_FE is not set
++CONFIG_PCI_GOANY=y
++CONFIG_PCI_DIRECT=y
++CONFIG_PCI_MMCONFIG=y
++CONFIG_XEN_PCIDEV_FRONTEND=y
++# CONFIG_XEN_PCIDEV_FE_DEBUG is not set
++# CONFIG_PCIEPORTBUS is not set
++CONFIG_PCI_MSI=y
++# CONFIG_PCI_DEBUG is not set
++CONFIG_ISA_DMA_API=y
++# CONFIG_SCx200 is not set
++CONFIG_K8_NB=y
++
++#
++# PCCARD (PCMCIA/CardBus) support
++#
++# CONFIG_PCCARD is not set
++
++#
++# PCI Hotplug Support
++#
++# CONFIG_HOTPLUG_PCI is not set
++
++#
++# Executable file formats
++#
++CONFIG_BINFMT_ELF=y
++# CONFIG_BINFMT_AOUT is not set
++# CONFIG_BINFMT_MISC is not set
++
++#
++# Networking
++#
++CONFIG_NET=y
++
++#
++# Networking options
++#
++# CONFIG_NETDEBUG is not set
++CONFIG_PACKET=y
++# CONFIG_PACKET_MMAP is not set
++CONFIG_UNIX=y
++CONFIG_XFRM=y
++# CONFIG_XFRM_USER is not set
++# CONFIG_NET_KEY is not set
++CONFIG_INET=y
++# CONFIG_IP_MULTICAST is not set
++# CONFIG_IP_ADVANCED_ROUTER is not set
++CONFIG_IP_FIB_HASH=y
++CONFIG_IP_PNP=y
++CONFIG_IP_PNP_DHCP=y
++# CONFIG_IP_PNP_BOOTP is not set
++# CONFIG_IP_PNP_RARP is not set
++# CONFIG_NET_IPIP is not set
++# CONFIG_NET_IPGRE is not set
++# CONFIG_ARPD is not set
++# CONFIG_SYN_COOKIES is not set
++# CONFIG_INET_AH is not set
++# CONFIG_INET_ESP is not set
++# CONFIG_INET_IPCOMP is not set
++# CONFIG_INET_XFRM_TUNNEL is not set
++# CONFIG_INET_TUNNEL is not set
++CONFIG_INET_XFRM_MODE_TRANSPORT=y
++CONFIG_INET_XFRM_MODE_TUNNEL=y
++# CONFIG_INET_DIAG is not set
++# CONFIG_TCP_CONG_ADVANCED is not set
++CONFIG_TCP_CONG_BIC=y
++
++#
++# IP: Virtual Server Configuration
++#
++# CONFIG_IP_VS is not set
++# CONFIG_IPV6 is not set
++# CONFIG_INET6_XFRM_TUNNEL is not set
++# CONFIG_INET6_TUNNEL is not set
++# CONFIG_NETWORK_SECMARK is not set
++CONFIG_NETFILTER=y
++# CONFIG_NETFILTER_DEBUG is not set
++CONFIG_BRIDGE_NETFILTER=y
++
++#
++# Core Netfilter Configuration
++#
++# CONFIG_NETFILTER_NETLINK is not set
++# CONFIG_NETFILTER_XTABLES is not set
++
++#
++# IP: Netfilter Configuration
++#
++CONFIG_IP_NF_CONNTRACK=m
++CONFIG_IP_NF_CT_ACCT=y
++# CONFIG_IP_NF_CONNTRACK_MARK is not set
++# CONFIG_IP_NF_CONNTRACK_EVENTS is not set
++# CONFIG_IP_NF_CT_PROTO_SCTP is not set
++CONFIG_IP_NF_FTP=m
++# CONFIG_IP_NF_IRC is not set
++# CONFIG_IP_NF_NETBIOS_NS is not set
++# CONFIG_IP_NF_TFTP is not set
++# CONFIG_IP_NF_AMANDA is not set
++# CONFIG_IP_NF_PPTP is not set
++# CONFIG_IP_NF_H323 is not set
++# CONFIG_IP_NF_SIP is not set
++# CONFIG_IP_NF_QUEUE is not set
++
++#
++# Bridge: Netfilter Configuration
++#
++# CONFIG_BRIDGE_NF_EBTABLES is not set
++
++#
++# DCCP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_DCCP is not set
++
++#
++# SCTP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_SCTP is not set
++
++#
++# TIPC Configuration (EXPERIMENTAL)
++#
++# CONFIG_TIPC is not set
++# CONFIG_ATM is not set
++CONFIG_BRIDGE=y
++# CONFIG_VLAN_8021Q is not set
++# CONFIG_DECNET is not set
++CONFIG_LLC=y
++# CONFIG_LLC2 is not set
++# CONFIG_IPX is not set
++# CONFIG_ATALK is not set
++# CONFIG_X25 is not set
++# CONFIG_LAPB is not set
++# CONFIG_ECONET is not set
++# CONFIG_WAN_ROUTER is not set
++
++#
++# QoS and/or fair queueing
++#
++# CONFIG_NET_SCHED is not set
++
++#
++# Network testing
++#
++# CONFIG_NET_PKTGEN is not set
++# CONFIG_HAMRADIO is not set
++# CONFIG_IRDA is not set
++# CONFIG_BT is not set
++# CONFIG_IEEE80211 is not set
++
++#
++# Device Drivers
++#
++
++#
++# Generic Driver Options
++#
++# CONFIG_STANDALONE is not set
++CONFIG_PREVENT_FIRMWARE_BUILD=y
++# CONFIG_FW_LOADER is not set
++# CONFIG_DEBUG_DRIVER is not set
++CONFIG_SYS_HYPERVISOR=y
++
++#
++# Connector - unified userspace <-> kernelspace linker
++#
++# CONFIG_CONNECTOR is not set
++
++#
++# Memory Technology Devices (MTD)
++#
++# CONFIG_MTD is not set
++
++#
++# Parallel port support
++#
++# CONFIG_PARPORT is not set
++
++#
++# Plug and Play support
++#
++CONFIG_PNP=y
++CONFIG_PNP_DEBUG=y
++
++#
++# Protocols
++#
++CONFIG_PNPACPI=y
++
++#
++# Block devices
++#
++CONFIG_BLK_DEV_FD=y
++# CONFIG_BLK_CPQ_DA is not set
++CONFIG_BLK_CPQ_CISS_DA=y
++# CONFIG_CISS_SCSI_TAPE is not set
++# CONFIG_BLK_DEV_DAC960 is not set
++# CONFIG_BLK_DEV_UMEM is not set
++# CONFIG_BLK_DEV_COW_COMMON is not set
++CONFIG_BLK_DEV_LOOP=y
++# CONFIG_BLK_DEV_CRYPTOLOOP is not set
++# CONFIG_BLK_DEV_NBD is not set
++# CONFIG_BLK_DEV_SX8 is not set
++# CONFIG_BLK_DEV_UB is not set
++CONFIG_BLK_DEV_RAM=y
++CONFIG_BLK_DEV_RAM_COUNT=16
++CONFIG_BLK_DEV_RAM_SIZE=4096
++CONFIG_BLK_DEV_RAM_BLOCKSIZE=1024
++CONFIG_BLK_DEV_INITRD=y
++# CONFIG_CDROM_PKTCDVD is not set
++# CONFIG_ATA_OVER_ETH is not set
++
++#
++# ATA/ATAPI/MFM/RLL support
++#
++CONFIG_IDE=y
++CONFIG_BLK_DEV_IDE=y
++
++#
++# Please see Documentation/ide.txt for help/info on IDE drives
++#
++# CONFIG_BLK_DEV_IDE_SATA is not set
++# CONFIG_BLK_DEV_HD_IDE is not set
++CONFIG_BLK_DEV_IDEDISK=y
++# CONFIG_IDEDISK_MULTI_MODE is not set
++CONFIG_BLK_DEV_IDECD=y
++# CONFIG_BLK_DEV_IDETAPE is not set
++# CONFIG_BLK_DEV_IDEFLOPPY is not set
++# CONFIG_BLK_DEV_IDESCSI is not set
++# CONFIG_IDE_TASK_IOCTL is not set
++
++#
++# IDE chipset support/bugfixes
++#
++CONFIG_IDE_GENERIC=y
++# CONFIG_BLK_DEV_CMD640 is not set
++# CONFIG_BLK_DEV_IDEPNP is not set
++CONFIG_BLK_DEV_IDEPCI=y
++# CONFIG_IDEPCI_SHARE_IRQ is not set
++# CONFIG_BLK_DEV_OFFBOARD is not set
++CONFIG_BLK_DEV_GENERIC=y
++# CONFIG_BLK_DEV_OPTI621 is not set
++# CONFIG_BLK_DEV_RZ1000 is not set
++CONFIG_BLK_DEV_IDEDMA_PCI=y
++# CONFIG_BLK_DEV_IDEDMA_FORCED is not set
++CONFIG_IDEDMA_PCI_AUTO=y
++# CONFIG_IDEDMA_ONLYDISK is not set
++# CONFIG_BLK_DEV_AEC62XX is not set
++# CONFIG_BLK_DEV_ALI15X3 is not set
++# CONFIG_BLK_DEV_AMD74XX is not set
++# CONFIG_BLK_DEV_ATIIXP is not set
++# CONFIG_BLK_DEV_CMD64X is not set
++# CONFIG_BLK_DEV_TRIFLEX is not set
++# CONFIG_BLK_DEV_CY82C693 is not set
++# CONFIG_BLK_DEV_CS5520 is not set
++# CONFIG_BLK_DEV_CS5530 is not set
++# CONFIG_BLK_DEV_CS5535 is not set
++# CONFIG_BLK_DEV_HPT34X is not set
++# CONFIG_BLK_DEV_HPT366 is not set
++# CONFIG_BLK_DEV_SC1200 is not set
++CONFIG_BLK_DEV_PIIX=y
++# CONFIG_BLK_DEV_IT821X is not set
++# CONFIG_BLK_DEV_NS87415 is not set
++# CONFIG_BLK_DEV_PDC202XX_OLD is not set
++# CONFIG_BLK_DEV_PDC202XX_NEW is not set
++CONFIG_BLK_DEV_SVWKS=y
++# CONFIG_BLK_DEV_SIIMAGE is not set
++# CONFIG_BLK_DEV_SIS5513 is not set
++# CONFIG_BLK_DEV_SLC90E66 is not set
++# CONFIG_BLK_DEV_TRM290 is not set
++# CONFIG_BLK_DEV_VIA82CXXX is not set
++# CONFIG_IDE_ARM is not set
++CONFIG_BLK_DEV_IDEDMA=y
++# CONFIG_IDEDMA_IVB is not set
++CONFIG_IDEDMA_AUTO=y
++# CONFIG_BLK_DEV_HD is not set
++
++#
++# SCSI device support
++#
++# CONFIG_RAID_ATTRS is not set
++CONFIG_SCSI=y
++CONFIG_SCSI_PROC_FS=y
++
++#
++# SCSI support type (disk, tape, CD-ROM)
++#
++CONFIG_BLK_DEV_SD=y
++# CONFIG_CHR_DEV_ST is not set
++# CONFIG_CHR_DEV_OSST is not set
++# CONFIG_BLK_DEV_SR is not set
++# CONFIG_CHR_DEV_SG is not set
++# CONFIG_CHR_DEV_SCH is not set
++
++#
++# Some SCSI devices (e.g. CD jukebox) support multiple LUNs
++#
++# CONFIG_SCSI_MULTI_LUN is not set
++# CONFIG_SCSI_CONSTANTS is not set
++# CONFIG_SCSI_LOGGING is not set
++
++#
++# SCSI Transport Attributes
++#
++CONFIG_SCSI_SPI_ATTRS=y
++# CONFIG_SCSI_FC_ATTRS is not set
++# CONFIG_SCSI_ISCSI_ATTRS is not set
++# CONFIG_SCSI_SAS_ATTRS is not set
++
++#
++# SCSI low-level drivers
++#
++# CONFIG_ISCSI_TCP is not set
++CONFIG_BLK_DEV_3W_XXXX_RAID=y
++# CONFIG_SCSI_3W_9XXX is not set
++# CONFIG_SCSI_ACARD is not set
++CONFIG_SCSI_AACRAID=y
++CONFIG_SCSI_AIC7XXX=y
++CONFIG_AIC7XXX_CMDS_PER_DEVICE=32
++CONFIG_AIC7XXX_RESET_DELAY_MS=15000
++CONFIG_AIC7XXX_DEBUG_ENABLE=y
++CONFIG_AIC7XXX_DEBUG_MASK=0
++CONFIG_AIC7XXX_REG_PRETTY_PRINT=y
++# CONFIG_SCSI_AIC7XXX_OLD is not set
++CONFIG_SCSI_AIC79XX=y
++CONFIG_AIC79XX_CMDS_PER_DEVICE=32
++CONFIG_AIC79XX_RESET_DELAY_MS=15000
++# CONFIG_AIC79XX_ENABLE_RD_STRM is not set
++CONFIG_AIC79XX_DEBUG_ENABLE=y
++CONFIG_AIC79XX_DEBUG_MASK=0
++CONFIG_AIC79XX_REG_PRETTY_PRINT=y
++# CONFIG_SCSI_DPT_I2O is not set
++CONFIG_SCSI_ADVANSYS=y
++CONFIG_MEGARAID_NEWGEN=y
++# CONFIG_MEGARAID_MM is not set
++# CONFIG_MEGARAID_LEGACY is not set
++# CONFIG_MEGARAID_SAS is not set
++CONFIG_SCSI_SATA=y
++CONFIG_SCSI_SATA_AHCI=y
++# CONFIG_SCSI_SATA_SVW is not set
++CONFIG_SCSI_ATA_PIIX=y
++# CONFIG_SCSI_SATA_MV is not set
++# CONFIG_SCSI_SATA_NV is not set
++# CONFIG_SCSI_PDC_ADMA is not set
++# CONFIG_SCSI_HPTIOP is not set
++# CONFIG_SCSI_SATA_QSTOR is not set
++CONFIG_SCSI_SATA_PROMISE=y
++CONFIG_SCSI_SATA_SX4=y
++CONFIG_SCSI_SATA_SIL=y
++CONFIG_SCSI_SATA_SIL24=y
++# CONFIG_SCSI_SATA_SIS is not set
++# CONFIG_SCSI_SATA_ULI is not set
++# CONFIG_SCSI_SATA_VIA is not set
++# CONFIG_SCSI_SATA_VITESSE is not set
++CONFIG_SCSI_SATA_INTEL_COMBINED=y
++# CONFIG_SCSI_BUSLOGIC is not set
++# CONFIG_SCSI_DMX3191D is not set
++# CONFIG_SCSI_EATA is not set
++# CONFIG_SCSI_FUTURE_DOMAIN is not set
++# CONFIG_SCSI_GDTH is not set
++# CONFIG_SCSI_IPS is not set
++# CONFIG_SCSI_INITIO is not set
++# CONFIG_SCSI_INIA100 is not set
++# CONFIG_SCSI_SYM53C8XX_2 is not set
++# CONFIG_SCSI_IPR is not set
++# CONFIG_SCSI_QLOGIC_1280 is not set
++# CONFIG_SCSI_QLA_FC is not set
++# CONFIG_SCSI_LPFC is not set
++# CONFIG_SCSI_DC395x is not set
++# CONFIG_SCSI_DC390T is not set
++# CONFIG_SCSI_NSP32 is not set
++# CONFIG_SCSI_DEBUG is not set
++
++#
++# Multi-device support (RAID and LVM)
++#
++CONFIG_MD=y
++CONFIG_BLK_DEV_MD=y
++# CONFIG_MD_LINEAR is not set
++CONFIG_MD_RAID0=y
++CONFIG_MD_RAID1=y
++# CONFIG_MD_RAID10 is not set
++# CONFIG_MD_RAID456 is not set
++# CONFIG_MD_MULTIPATH is not set
++# CONFIG_MD_FAULTY is not set
++CONFIG_BLK_DEV_DM=y
++# CONFIG_DM_CRYPT is not set
++CONFIG_DM_SNAPSHOT=y
++CONFIG_DM_MIRROR=y
++# CONFIG_DM_ZERO is not set
++# CONFIG_DM_MULTIPATH is not set
++
++#
++# Fusion MPT device support
++#
++CONFIG_FUSION=y
++CONFIG_FUSION_SPI=y
++# CONFIG_FUSION_FC is not set
++# CONFIG_FUSION_SAS is not set
++CONFIG_FUSION_MAX_SGE=128
++# CONFIG_FUSION_CTL is not set
++
++#
++# IEEE 1394 (FireWire) support
++#
++# CONFIG_IEEE1394 is not set
++
++#
++# I2O device support
++#
++# CONFIG_I2O is not set
++
++#
++# Network device support
++#
++CONFIG_NETDEVICES=y
++# CONFIG_DUMMY is not set
++# CONFIG_BONDING is not set
++# CONFIG_EQUALIZER is not set
++CONFIG_TUN=y
++# CONFIG_NET_SB1000 is not set
++
++#
++# ARCnet devices
++#
++# CONFIG_ARCNET is not set
++
++#
++# PHY device support
++#
++# CONFIG_PHYLIB is not set
++
++#
++# Ethernet (10 or 100Mbit)
++#
++CONFIG_NET_ETHERNET=y
++CONFIG_MII=y
++# CONFIG_HAPPYMEAL is not set
++# CONFIG_SUNGEM is not set
++# CONFIG_CASSINI is not set
++CONFIG_NET_VENDOR_3COM=y
++CONFIG_VORTEX=y
++# CONFIG_TYPHOON is not set
++
++#
++# Tulip family network device support
++#
++CONFIG_NET_TULIP=y
++# CONFIG_DE2104X is not set
++CONFIG_TULIP=y
++# CONFIG_TULIP_MWI is not set
++# CONFIG_TULIP_MMIO is not set
++# CONFIG_TULIP_NAPI is not set
++# CONFIG_DE4X5 is not set
++# CONFIG_WINBOND_840 is not set
++# CONFIG_DM9102 is not set
++# CONFIG_ULI526X is not set
++# CONFIG_HP100 is not set
++CONFIG_NET_PCI=y
++CONFIG_PCNET32=y
++# CONFIG_AMD8111_ETH is not set
++# CONFIG_ADAPTEC_STARFIRE is not set
++# CONFIG_B44 is not set
++# CONFIG_FORCEDETH is not set
++# CONFIG_DGRS is not set
++# CONFIG_EEPRO100 is not set
++CONFIG_E100=y
++# CONFIG_FEALNX is not set
++# CONFIG_NATSEMI is not set
++CONFIG_NE2K_PCI=y
++# CONFIG_8139CP is not set
++CONFIG_8139TOO=y
++CONFIG_8139TOO_PIO=y
++# CONFIG_8139TOO_TUNE_TWISTER is not set
++# CONFIG_8139TOO_8129 is not set
++# CONFIG_8139_OLD_RX_RESET is not set
++# CONFIG_SIS900 is not set
++# CONFIG_EPIC100 is not set
++# CONFIG_SUNDANCE is not set
++# CONFIG_TLAN is not set
++CONFIG_VIA_RHINE=y
++# CONFIG_VIA_RHINE_MMIO is not set
++# CONFIG_VIA_RHINE_NAPI is not set
++
++#
++# Ethernet (1000 Mbit)
++#
++CONFIG_ACENIC=y
++# CONFIG_ACENIC_OMIT_TIGON_I is not set
++# CONFIG_DL2K is not set
++CONFIG_E1000=y
++# CONFIG_E1000_NAPI is not set
++# CONFIG_E1000_DISABLE_PACKET_SPLIT is not set
++# CONFIG_NS83820 is not set
++# CONFIG_HAMACHI is not set
++# CONFIG_YELLOWFIN is not set
++# CONFIG_R8169 is not set
++# CONFIG_SIS190 is not set
++# CONFIG_SKGE is not set
++# CONFIG_SKY2 is not set
++CONFIG_SK98LIN=y
++# CONFIG_VIA_VELOCITY is not set
++CONFIG_TIGON3=y
++CONFIG_BNX2=y
++
++#
++# Ethernet (10000 Mbit)
++#
++# CONFIG_CHELSIO_T1 is not set
++# CONFIG_IXGB is not set
++# CONFIG_S2IO is not set
++# CONFIG_MYRI10GE is not set
++# CONFIG_SFC is not set
++
++#
++# Token Ring devices
++#
++# CONFIG_TR is not set
++
++#
++# Wireless LAN (non-hamradio)
++#
++# CONFIG_NET_RADIO is not set
++
++#
++# Wan interfaces
++#
++# CONFIG_WAN is not set
++# CONFIG_FDDI is not set
++# CONFIG_HIPPI is not set
++# CONFIG_PPP is not set
++# CONFIG_SLIP is not set
++# CONFIG_NET_FC is not set
++# CONFIG_SHAPER is not set
++# CONFIG_NETCONSOLE is not set
++# CONFIG_NETPOLL is not set
++# CONFIG_NET_POLL_CONTROLLER is not set
++
++#
++# ISDN subsystem
++#
++# CONFIG_ISDN is not set
++
++#
++# Telephony Support
++#
++# CONFIG_PHONE is not set
++
++#
++# Input device support
++#
++CONFIG_INPUT=y
++
++#
++# Userland interfaces
++#
++CONFIG_INPUT_MOUSEDEV=y
++CONFIG_INPUT_MOUSEDEV_PSAUX=y
++CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024
++CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768
++# CONFIG_INPUT_JOYDEV is not set
++# CONFIG_INPUT_TSDEV is not set
++# CONFIG_INPUT_EVDEV is not set
++# CONFIG_INPUT_EVBUG is not set
++
++#
++# Input Device Drivers
++#
++CONFIG_INPUT_KEYBOARD=y
++CONFIG_KEYBOARD_ATKBD=y
++# CONFIG_KEYBOARD_SUNKBD is not set
++# CONFIG_KEYBOARD_LKKBD is not set
++# CONFIG_KEYBOARD_XTKBD is not set
++# CONFIG_KEYBOARD_NEWTON is not set
++CONFIG_INPUT_MOUSE=y
++CONFIG_MOUSE_PS2=y
++# CONFIG_MOUSE_SERIAL is not set
++# CONFIG_MOUSE_VSXXXAA is not set
++# CONFIG_INPUT_JOYSTICK is not set
++# CONFIG_INPUT_TOUCHSCREEN is not set
++# CONFIG_INPUT_MISC is not set
++
++#
++# Hardware I/O ports
++#
++CONFIG_SERIO=y
++CONFIG_SERIO_I8042=y
++CONFIG_SERIO_SERPORT=y
++# CONFIG_SERIO_CT82C710 is not set
++# CONFIG_SERIO_PCIPS2 is not set
++CONFIG_SERIO_LIBPS2=y
++# CONFIG_SERIO_RAW is not set
++# CONFIG_GAMEPORT is not set
++
++#
++# Character devices
++#
++CONFIG_VT=y
++CONFIG_VT_CONSOLE=y
++CONFIG_HW_CONSOLE=y
++CONFIG_VT_HW_CONSOLE_BINDING=y
++# CONFIG_SERIAL_NONSTANDARD is not set
++
++#
++# Serial drivers
++#
++
++#
++# Non-8250 serial port support
++#
++# CONFIG_SERIAL_JSM is not set
++CONFIG_UNIX98_PTYS=y
++CONFIG_LEGACY_PTYS=y
++CONFIG_LEGACY_PTY_COUNT=256
++
++#
++# IPMI
++#
++# CONFIG_IPMI_HANDLER is not set
++
++#
++# Watchdog Cards
++#
++# CONFIG_WATCHDOG is not set
++# CONFIG_HW_RANDOM is not set
++# CONFIG_NVRAM is not set
++# CONFIG_RTC is not set
++# CONFIG_GEN_RTC is not set
++# CONFIG_DTLK is not set
++# CONFIG_R3964 is not set
++# CONFIG_APPLICOM is not set
++# CONFIG_SONYPI is not set
++
++#
++# Ftape, the floppy tape device driver
++#
++CONFIG_AGP=m
++CONFIG_AGP_ALI=m
++CONFIG_AGP_ATI=m
++CONFIG_AGP_AMD=m
++CONFIG_AGP_AMD64=m
++CONFIG_AGP_INTEL=m
++CONFIG_AGP_NVIDIA=m
++CONFIG_AGP_SIS=m
++CONFIG_AGP_SWORKS=m
++CONFIG_AGP_VIA=m
++# CONFIG_AGP_EFFICEON is not set
++CONFIG_DRM=m
++CONFIG_DRM_TDFX=m
++CONFIG_DRM_R128=m
++CONFIG_DRM_RADEON=m
++CONFIG_DRM_I810=m
++CONFIG_DRM_I830=m
++CONFIG_DRM_I915=m
++CONFIG_DRM_MGA=m
++CONFIG_DRM_SIS=m
++# CONFIG_DRM_VIA is not set
++# CONFIG_DRM_SAVAGE is not set
++# CONFIG_MWAVE is not set
++# CONFIG_PC8736x_GPIO is not set
++# CONFIG_NSC_GPIO is not set
++# CONFIG_CS5535_GPIO is not set
++# CONFIG_RAW_DRIVER is not set
++# CONFIG_HPET is not set
++# CONFIG_HANGCHECK_TIMER is not set
++
++#
++# TPM devices
++#
++# CONFIG_TCG_TPM is not set
++# CONFIG_TELCLOCK is not set
++
++#
++# I2C support
++#
++# CONFIG_I2C is not set
++
++#
++# SPI support
++#
++# CONFIG_SPI is not set
++# CONFIG_SPI_MASTER is not set
++
++#
++# Dallas's 1-wire bus
++#
++
++#
++# Hardware Monitoring support
++#
++# CONFIG_HWMON is not set
++# CONFIG_HWMON_VID is not set
++
++#
++# Misc devices
++#
++# CONFIG_IBM_ASM is not set
++
++#
++# Multimedia devices
++#
++# CONFIG_VIDEO_DEV is not set
++
++#
++# Digital Video Broadcasting Devices
++#
++# CONFIG_DVB is not set
++# CONFIG_USB_DABUSB is not set
++
++#
++# Graphics support
++#
++CONFIG_FIRMWARE_EDID=y
++# CONFIG_FB is not set
++
++#
++# Console display driver support
++#
++CONFIG_VGA_CONSOLE=y
++# CONFIG_VGACON_SOFT_SCROLLBACK is not set
++CONFIG_DUMMY_CONSOLE=y
++# CONFIG_BACKLIGHT_LCD_SUPPORT is not set
++
++#
++# Sound
++#
++# CONFIG_SOUND is not set
++
++#
++# USB support
++#
++CONFIG_USB_ARCH_HAS_HCD=y
++CONFIG_USB_ARCH_HAS_OHCI=y
++CONFIG_USB_ARCH_HAS_EHCI=y
++CONFIG_USB=y
++# CONFIG_USB_DEBUG is not set
++
++#
++# Miscellaneous USB options
++#
++# CONFIG_USB_DEVICEFS is not set
++# CONFIG_USB_BANDWIDTH is not set
++# CONFIG_USB_DYNAMIC_MINORS is not set
++# CONFIG_USB_SUSPEND is not set
++# CONFIG_USB_OTG is not set
++
++#
++# USB Host Controller Drivers
++#
++# CONFIG_USB_EHCI_HCD is not set
++# CONFIG_USB_ISP116X_HCD is not set
++CONFIG_USB_OHCI_HCD=y
++# CONFIG_USB_OHCI_BIG_ENDIAN is not set
++CONFIG_USB_OHCI_LITTLE_ENDIAN=y
++CONFIG_USB_UHCI_HCD=y
++# CONFIG_USB_SL811_HCD is not set
++
++#
++# USB Device Class drivers
++#
++# CONFIG_USB_ACM is not set
++# CONFIG_USB_PRINTER is not set
++
++#
++# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support'
++#
++
++#
++# may also be needed; see USB_STORAGE Help for more information
++#
++# CONFIG_USB_STORAGE is not set
++# CONFIG_USB_LIBUSUAL is not set
++
++#
++# USB Input Devices
++#
++CONFIG_USB_HID=y
++CONFIG_USB_HIDINPUT=y
++# CONFIG_USB_HIDINPUT_POWERBOOK is not set
++# CONFIG_HID_FF is not set
++# CONFIG_USB_HIDDEV is not set
++# CONFIG_USB_AIPTEK is not set
++# CONFIG_USB_WACOM is not set
++# CONFIG_USB_ACECAD is not set
++# CONFIG_USB_KBTAB is not set
++# CONFIG_USB_POWERMATE is not set
++# CONFIG_USB_TOUCHSCREEN is not set
++# CONFIG_USB_YEALINK is not set
++# CONFIG_USB_XPAD is not set
++# CONFIG_USB_ATI_REMOTE is not set
++# CONFIG_USB_ATI_REMOTE2 is not set
++# CONFIG_USB_KEYSPAN_REMOTE is not set
++# CONFIG_USB_APPLETOUCH is not set
++
++#
++# USB Imaging devices
++#
++# CONFIG_USB_MDC800 is not set
++# CONFIG_USB_MICROTEK is not set
++
++#
++# USB Network Adapters
++#
++# CONFIG_USB_CATC is not set
++# CONFIG_USB_KAWETH is not set
++# CONFIG_USB_PEGASUS is not set
++# CONFIG_USB_RTL8150 is not set
++# CONFIG_USB_USBNET is not set
++CONFIG_USB_MON=y
++
++#
++# USB port drivers
++#
++
++#
++# USB Serial Converter support
++#
++# CONFIG_USB_SERIAL is not set
++
++#
++# USB Miscellaneous drivers
++#
++# CONFIG_USB_EMI62 is not set
++# CONFIG_USB_EMI26 is not set
++# CONFIG_USB_AUERSWALD is not set
++# CONFIG_USB_RIO500 is not set
++# CONFIG_USB_LEGOTOWER is not set
++# CONFIG_USB_LCD is not set
++# CONFIG_USB_LED is not set
++# CONFIG_USB_CYPRESS_CY7C63 is not set
++# CONFIG_USB_CYTHERM is not set
++# CONFIG_USB_PHIDGETKIT is not set
++# CONFIG_USB_PHIDGETSERVO is not set
++# CONFIG_USB_IDMOUSE is not set
++# CONFIG_USB_APPLEDISPLAY is not set
++# CONFIG_USB_LD is not set
++
++#
++# USB DSL modem support
++#
++
++#
++# USB Gadget Support
++#
++# CONFIG_USB_GADGET is not set
++
++#
++# MMC/SD Card support
++#
++# CONFIG_MMC is not set
++
++#
++# LED devices
++#
++# CONFIG_NEW_LEDS is not set
++
++#
++# LED drivers
++#
++
++#
++# LED Triggers
++#
++
++#
++# InfiniBand support
++#
++# CONFIG_INFINIBAND is not set
++
++#
++# EDAC - error detection and reporting (RAS) (EXPERIMENTAL)
++#
++# CONFIG_EDAC is not set
++
++#
++# Real Time Clock
++#
++CONFIG_RTC_LIB=m
++CONFIG_RTC_CLASS=m
++
++#
++# RTC interfaces
++#
++CONFIG_RTC_INTF_SYSFS=m
++CONFIG_RTC_INTF_PROC=m
++CONFIG_RTC_INTF_DEV=m
++# CONFIG_RTC_INTF_DEV_UIE_EMUL is not set
++
++#
++# RTC drivers
++#
++# CONFIG_RTC_DRV_DS1553 is not set
++# CONFIG_RTC_DRV_DS1742 is not set
++CONFIG_RTC_DRV_M48T86=m
++CONFIG_RTC_DRV_TEST=m
++# CONFIG_RTC_DRV_V3020 is not set
++
++#
++# DMA Engine support
++#
++# CONFIG_DMA_ENGINE is not set
++
++#
++# DMA Clients
++#
++
++#
++# DMA Devices
++#
++
++#
++# File systems
++#
++CONFIG_EXT2_FS=y
++# CONFIG_EXT2_FS_XATTR is not set
++# CONFIG_EXT2_FS_XIP is not set
++CONFIG_EXT3_FS=y
++CONFIG_EXT3_FS_XATTR=y
++# CONFIG_EXT3_FS_POSIX_ACL is not set
++# CONFIG_EXT3_FS_SECURITY is not set
++CONFIG_JBD=y
++# CONFIG_JBD_DEBUG is not set
++CONFIG_FS_MBCACHE=y
++CONFIG_REISERFS_FS=y
++# CONFIG_REISERFS_CHECK is not set
++# CONFIG_REISERFS_PROC_INFO is not set
++# CONFIG_REISERFS_FS_XATTR is not set
++# CONFIG_JFS_FS is not set
++# CONFIG_FS_POSIX_ACL is not set
++# CONFIG_XFS_FS is not set
++# CONFIG_OCFS2_FS is not set
++# CONFIG_MINIX_FS is not set
++# CONFIG_ROMFS_FS is not set
++CONFIG_INOTIFY=y
++CONFIG_INOTIFY_USER=y
++# CONFIG_QUOTA is not set
++CONFIG_DNOTIFY=y
++CONFIG_AUTOFS_FS=y
++CONFIG_AUTOFS4_FS=y
++# CONFIG_FUSE_FS is not set
++
++#
++# CD-ROM/DVD Filesystems
++#
++CONFIG_ISO9660_FS=y
++CONFIG_JOLIET=y
++CONFIG_ZISOFS=y
++CONFIG_ZISOFS_FS=y
++# CONFIG_UDF_FS is not set
++
++#
++# DOS/FAT/NT Filesystems
++#
++CONFIG_FAT_FS=m
++CONFIG_MSDOS_FS=m
++CONFIG_VFAT_FS=m
++CONFIG_FAT_DEFAULT_CODEPAGE=437
++CONFIG_FAT_DEFAULT_IOCHARSET="iso8859-1"
++# CONFIG_NTFS_FS is not set
++
++#
++# Pseudo filesystems
++#
++CONFIG_PROC_FS=y
++CONFIG_PROC_KCORE=y
++CONFIG_SYSFS=y
++CONFIG_TMPFS=y
++# CONFIG_HUGETLB_PAGE is not set
++CONFIG_RAMFS=y
++# CONFIG_CONFIGFS_FS is not set
++
++#
++# Miscellaneous filesystems
++#
++# CONFIG_ADFS_FS is not set
++# CONFIG_AFFS_FS is not set
++# CONFIG_HFS_FS is not set
++# CONFIG_HFSPLUS_FS is not set
++# CONFIG_BEFS_FS is not set
++# CONFIG_BFS_FS is not set
++# CONFIG_EFS_FS is not set
++CONFIG_CRAMFS=y
++# CONFIG_VXFS_FS is not set
++# CONFIG_HPFS_FS is not set
++# CONFIG_QNX4FS_FS is not set
++# CONFIG_SYSV_FS is not set
++# CONFIG_UFS_FS is not set
++
++#
++# Network File Systems
++#
++CONFIG_NFS_FS=y
++CONFIG_NFS_V3=y
++# CONFIG_NFS_V3_ACL is not set
++# CONFIG_NFS_V4 is not set
++# CONFIG_NFS_DIRECTIO is not set
++CONFIG_NFSD=m
++CONFIG_NFSD_V3=y
++# CONFIG_NFSD_V3_ACL is not set
++# CONFIG_NFSD_V4 is not set
++CONFIG_NFSD_TCP=y
++CONFIG_ROOT_NFS=y
++CONFIG_LOCKD=y
++CONFIG_LOCKD_V4=y
++CONFIG_EXPORTFS=m
++CONFIG_NFS_COMMON=y
++CONFIG_SUNRPC=y
++# CONFIG_RPCSEC_GSS_KRB5 is not set
++# CONFIG_RPCSEC_GSS_SPKM3 is not set
++# CONFIG_SMB_FS is not set
++# CONFIG_CIFS is not set
++# CONFIG_NCP_FS is not set
++# CONFIG_CODA_FS is not set
++# CONFIG_AFS_FS is not set
++# CONFIG_9P_FS is not set
++
++#
++# Partition Types
++#
++# CONFIG_PARTITION_ADVANCED is not set
++CONFIG_MSDOS_PARTITION=y
++
++#
++# Native Language Support
++#
++CONFIG_NLS=y
++CONFIG_NLS_DEFAULT="iso8859-1"
++CONFIG_NLS_CODEPAGE_437=y
++# CONFIG_NLS_CODEPAGE_737 is not set
++# CONFIG_NLS_CODEPAGE_775 is not set
++# CONFIG_NLS_CODEPAGE_850 is not set
++# CONFIG_NLS_CODEPAGE_852 is not set
++# CONFIG_NLS_CODEPAGE_855 is not set
++# CONFIG_NLS_CODEPAGE_857 is not set
++# CONFIG_NLS_CODEPAGE_860 is not set
++# CONFIG_NLS_CODEPAGE_861 is not set
++# CONFIG_NLS_CODEPAGE_862 is not set
++# CONFIG_NLS_CODEPAGE_863 is not set
++# CONFIG_NLS_CODEPAGE_864 is not set
++# CONFIG_NLS_CODEPAGE_865 is not set
++# CONFIG_NLS_CODEPAGE_866 is not set
++# CONFIG_NLS_CODEPAGE_869 is not set
++# CONFIG_NLS_CODEPAGE_936 is not set
++# CONFIG_NLS_CODEPAGE_950 is not set
++# CONFIG_NLS_CODEPAGE_932 is not set
++# CONFIG_NLS_CODEPAGE_949 is not set
++# CONFIG_NLS_CODEPAGE_874 is not set
++# CONFIG_NLS_ISO8859_8 is not set
++# CONFIG_NLS_CODEPAGE_1250 is not set
++# CONFIG_NLS_CODEPAGE_1251 is not set
++# CONFIG_NLS_ASCII is not set
++CONFIG_NLS_ISO8859_1=y
++# CONFIG_NLS_ISO8859_2 is not set
++# CONFIG_NLS_ISO8859_3 is not set
++# CONFIG_NLS_ISO8859_4 is not set
++# CONFIG_NLS_ISO8859_5 is not set
++# CONFIG_NLS_ISO8859_6 is not set
++# CONFIG_NLS_ISO8859_7 is not set
++# CONFIG_NLS_ISO8859_9 is not set
++# CONFIG_NLS_ISO8859_13 is not set
++# CONFIG_NLS_ISO8859_14 is not set
++# CONFIG_NLS_ISO8859_15 is not set
++# CONFIG_NLS_KOI8_R is not set
++# CONFIG_NLS_KOI8_U is not set
++# CONFIG_NLS_UTF8 is not set
++
++#
++# Instrumentation Support
++#
++# CONFIG_PROFILING is not set
++# CONFIG_KPROBES is not set
++
++#
++# Kernel hacking
++#
++CONFIG_TRACE_IRQFLAGS_SUPPORT=y
++# CONFIG_PRINTK_TIME is not set
++CONFIG_MAGIC_SYSRQ=y
++CONFIG_UNUSED_SYMBOLS=y
++CONFIG_DEBUG_KERNEL=y
++CONFIG_LOG_BUF_SHIFT=14
++CONFIG_DETECT_SOFTLOCKUP=y
++# CONFIG_SCHEDSTATS is not set
++# CONFIG_DEBUG_SLAB is not set
++# CONFIG_DEBUG_RT_MUTEXES is not set
++# CONFIG_RT_MUTEX_TESTER is not set
++# CONFIG_DEBUG_SPINLOCK is not set
++# CONFIG_DEBUG_MUTEXES is not set
++# CONFIG_DEBUG_RWSEMS is not set
++# CONFIG_DEBUG_LOCK_ALLOC is not set
++# CONFIG_PROVE_LOCKING is not set
++# CONFIG_DEBUG_SPINLOCK_SLEEP is not set
++# CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set
++# CONFIG_DEBUG_KOBJECT is not set
++# CONFIG_DEBUG_HIGHMEM is not set
++CONFIG_DEBUG_BUGVERBOSE=y
++CONFIG_DEBUG_INFO=y
++# CONFIG_DEBUG_FS is not set
++# CONFIG_DEBUG_VM is not set
++CONFIG_FRAME_POINTER=y
++# CONFIG_UNWIND_INFO is not set
++CONFIG_FORCED_INLINING=y
++# CONFIG_RCU_TORTURE_TEST is not set
++CONFIG_EARLY_PRINTK=y
++# CONFIG_DEBUG_STACKOVERFLOW is not set
++# CONFIG_DEBUG_STACK_USAGE is not set
++# CONFIG_DEBUG_PAGEALLOC is not set
++# CONFIG_DEBUG_RODATA is not set
++# CONFIG_4KSTACKS is not set
++CONFIG_X86_FIND_SMP_CONFIG=y
++CONFIG_X86_MPPARSE=y
++
++#
++# Security options
++#
++# CONFIG_KEYS is not set
++# CONFIG_SECURITY is not set
++
++#
++# Cryptographic options
++#
++CONFIG_CRYPTO=y
++CONFIG_CRYPTO_HMAC=y
++# CONFIG_CRYPTO_NULL is not set
++# CONFIG_CRYPTO_MD4 is not set
++CONFIG_CRYPTO_MD5=m
++CONFIG_CRYPTO_SHA1=m
++# CONFIG_CRYPTO_SHA256 is not set
++# CONFIG_CRYPTO_SHA512 is not set
++# CONFIG_CRYPTO_WP512 is not set
++# CONFIG_CRYPTO_TGR192 is not set
++CONFIG_CRYPTO_DES=m
++# CONFIG_CRYPTO_BLOWFISH is not set
++# CONFIG_CRYPTO_TWOFISH is not set
++# CONFIG_CRYPTO_SERPENT is not set
++# CONFIG_CRYPTO_AES is not set
++# CONFIG_CRYPTO_AES_586 is not set
++# CONFIG_CRYPTO_CAST5 is not set
++# CONFIG_CRYPTO_CAST6 is not set
++# CONFIG_CRYPTO_TEA is not set
++# CONFIG_CRYPTO_ARC4 is not set
++# CONFIG_CRYPTO_KHAZAD is not set
++# CONFIG_CRYPTO_ANUBIS is not set
++# CONFIG_CRYPTO_DEFLATE is not set
++# CONFIG_CRYPTO_MICHAEL_MIC is not set
++CONFIG_CRYPTO_CRC32C=m
++# CONFIG_CRYPTO_TEST is not set
++
++#
++# Hardware crypto devices
++#
++# CONFIG_CRYPTO_DEV_PADLOCK is not set
++CONFIG_XEN=y
++CONFIG_XEN_INTERFACE_VERSION=0x00030207
++
++#
++# XEN
++#
++CONFIG_XEN_PRIVILEGED_GUEST=y
++# CONFIG_XEN_UNPRIVILEGED_GUEST is not set
++CONFIG_XEN_PRIVCMD=y
++CONFIG_XEN_XENBUS_DEV=y
++CONFIG_XEN_BACKEND=y
++CONFIG_XEN_BLKDEV_BACKEND=y
++CONFIG_XEN_BLKDEV_TAP=y
++CONFIG_XEN_NETDEV_BACKEND=y
++# CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_UTIL is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_BACKEND is not set
++# CONFIG_XEN_NETDEV_LOOPBACK is not set
++CONFIG_XEN_PCIDEV_BACKEND=y
++# CONFIG_XEN_PCIDEV_BACKEND_VPCI is not set
++CONFIG_XEN_PCIDEV_BACKEND_PASS=y
++# CONFIG_XEN_PCIDEV_BACKEND_SLOT is not set
++# CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER is not set
++# CONFIG_XEN_PCIDEV_BE_DEBUG is not set
++CONFIG_XEN_TPMDEV_BACKEND=m
++CONFIG_XEN_SCSI_BACKEND=m
++CONFIG_XEN_BLKDEV_FRONTEND=y
++CONFIG_XEN_NETDEV_FRONTEND=y
++CONFIG_XEN_SCSI_FRONTEND=m
++CONFIG_XEN_GRANT_DEV=y
++# CONFIG_XEN_NETDEV_ACCEL_SFC_FRONTEND is not set
++CONFIG_XEN_SCRUB_PAGES=y
++CONFIG_XEN_DISABLE_SERIAL=y
++CONFIG_XEN_SYSFS=y
++CONFIG_XEN_COMPAT_030002_AND_LATER=y
++# CONFIG_XEN_COMPAT_030004_AND_LATER is not set
++# CONFIG_XEN_COMPAT_030100_AND_LATER is not set
++# CONFIG_XEN_COMPAT_LATEST_ONLY is not set
++CONFIG_XEN_COMPAT=0x030002
++CONFIG_HAVE_IRQ_IGNORE_UNHANDLED=y
++CONFIG_NO_IDLE_HZ=y
++CONFIG_XEN_SMPBOOT=y
++CONFIG_XEN_BALLOON=y
++CONFIG_XEN_DEVMEM=y
++
++#
++# Library routines
++#
++# CONFIG_CRC_CCITT is not set
++# CONFIG_CRC16 is not set
++CONFIG_CRC32=y
++CONFIG_LIBCRC32C=y
++CONFIG_ZLIB_INFLATE=y
++CONFIG_PLIST=y
++CONFIG_GENERIC_HARDIRQS=y
++CONFIG_GENERIC_IRQ_PROBE=y
++CONFIG_GENERIC_PENDING_IRQ=y
++CONFIG_X86_SMP=y
++CONFIG_X86_BIOS_REBOOT=y
++CONFIG_X86_TRAMPOLINE=y
++CONFIG_X86_NO_TSS=y
++CONFIG_X86_NO_IDT=y
++CONFIG_KTIME_SCALAR=y
+diff -r d894e36cfc30 -r 0aa021803deb buildconfigs/linux-defconfig_xen0_x86_64
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/buildconfigs/linux-defconfig_xen0_x86_64 Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1390 @@
++#
++# Automatically generated make config: don't edit
++# Linux kernel version: 2.6.18.8
++# Mon Feb 18 10:41:04 2008
++#
++CONFIG_X86_64=y
++CONFIG_64BIT=y
++CONFIG_X86=y
++CONFIG_LOCKDEP_SUPPORT=y
++CONFIG_STACKTRACE_SUPPORT=y
++CONFIG_SEMAPHORE_SLEEPERS=y
++CONFIG_MMU=y
++CONFIG_RWSEM_GENERIC_SPINLOCK=y
++CONFIG_GENERIC_HWEIGHT=y
++CONFIG_GENERIC_CALIBRATE_DELAY=y
++CONFIG_X86_CMPXCHG=y
++CONFIG_EARLY_PRINTK=y
++CONFIG_GENERIC_ISA_DMA=y
++CONFIG_GENERIC_IOMAP=y
++CONFIG_ARCH_MAY_HAVE_PC_FDC=y
++CONFIG_DMI=y
++CONFIG_AUDIT_ARCH=y
++CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config"
++
++#
++# Code maturity level options
++#
++CONFIG_EXPERIMENTAL=y
++CONFIG_LOCK_KERNEL=y
++CONFIG_INIT_ENV_ARG_LIMIT=32
++
++#
++# General setup
++#
++CONFIG_LOCALVERSION=""
++# CONFIG_LOCALVERSION_AUTO is not set
++CONFIG_SWAP=y
++CONFIG_SYSVIPC=y
++# CONFIG_POSIX_MQUEUE is not set
++# CONFIG_BSD_PROCESS_ACCT is not set
++# CONFIG_TASKSTATS is not set
++# CONFIG_AUDIT is not set
++# CONFIG_IKCONFIG is not set
++# CONFIG_CPUSETS is not set
++# CONFIG_RELAY is not set
++CONFIG_INITRAMFS_SOURCE=""
++# CONFIG_CC_OPTIMIZE_FOR_SIZE is not set
++# CONFIG_EMBEDDED is not set
++CONFIG_UID16=y
++CONFIG_SYSCTL=y
++CONFIG_KALLSYMS=y
++# CONFIG_KALLSYMS_ALL is not set
++# CONFIG_KALLSYMS_EXTRA_PASS is not set
++CONFIG_HOTPLUG=y
++CONFIG_PRINTK=y
++CONFIG_BUG=y
++CONFIG_ELF_CORE=y
++CONFIG_BASE_FULL=y
++CONFIG_FUTEX=y
++CONFIG_EPOLL=y
++CONFIG_SHMEM=y
++CONFIG_SLAB=y
++CONFIG_VM_EVENT_COUNTERS=y
++CONFIG_RT_MUTEXES=y
++# CONFIG_TINY_SHMEM is not set
++CONFIG_BASE_SMALL=0
++# CONFIG_SLOB is not set
++
++#
++# Loadable module support
++#
++CONFIG_MODULES=y
++CONFIG_MODULE_UNLOAD=y
++# CONFIG_MODULE_FORCE_UNLOAD is not set
++# CONFIG_MODVERSIONS is not set
++# CONFIG_MODULE_SRCVERSION_ALL is not set
++CONFIG_KMOD=y
++CONFIG_STOP_MACHINE=y
++
++#
++# Block layer
++#
++# CONFIG_LBD is not set
++# CONFIG_BLK_DEV_IO_TRACE is not set
++CONFIG_LSF=y
++
++#
++# IO Schedulers
++#
++CONFIG_IOSCHED_NOOP=y
++CONFIG_IOSCHED_AS=y
++CONFIG_IOSCHED_DEADLINE=y
++CONFIG_IOSCHED_CFQ=y
++CONFIG_DEFAULT_AS=y
++# CONFIG_DEFAULT_DEADLINE is not set
++# CONFIG_DEFAULT_CFQ is not set
++# CONFIG_DEFAULT_NOOP is not set
++CONFIG_DEFAULT_IOSCHED="anticipatory"
++
++#
++# Processor type and features
++#
++CONFIG_X86_PC=y
++# CONFIG_X86_VSMP is not set
++# CONFIG_MK8 is not set
++# CONFIG_MPSC is not set
++CONFIG_GENERIC_CPU=y
++CONFIG_X86_64_XEN=y
++CONFIG_X86_NO_TSS=y
++CONFIG_X86_NO_IDT=y
++CONFIG_X86_L1_CACHE_BYTES=128
++CONFIG_X86_L1_CACHE_SHIFT=7
++CONFIG_X86_INTERNODE_CACHE_BYTES=128
++CONFIG_X86_GOOD_APIC=y
++CONFIG_MICROCODE=y
++CONFIG_X86_MSR=y
++CONFIG_X86_CPUID=y
++CONFIG_X86_IO_APIC=y
++CONFIG_X86_XEN_GENAPIC=y
++CONFIG_X86_LOCAL_APIC=y
++CONFIG_MTRR=y
++CONFIG_SMP=y
++CONFIG_PREEMPT_NONE=y
++# CONFIG_PREEMPT_VOLUNTARY is not set
++# CONFIG_PREEMPT is not set
++CONFIG_PREEMPT_BKL=y
++CONFIG_ARCH_FLATMEM_ENABLE=y
++CONFIG_SELECT_MEMORY_MODEL=y
++CONFIG_FLATMEM_MANUAL=y
++# CONFIG_DISCONTIGMEM_MANUAL is not set
++# CONFIG_SPARSEMEM_MANUAL is not set
++CONFIG_FLATMEM=y
++CONFIG_FLAT_NODE_MEM_MAP=y
++# CONFIG_SPARSEMEM_STATIC is not set
++CONFIG_SPLIT_PTLOCK_CPUS=4
++CONFIG_RESOURCES_64BIT=y
++CONFIG_NR_CPUS=8
++CONFIG_HOTPLUG_CPU=y
++CONFIG_ARCH_ENABLE_MEMORY_HOTPLUG=y
++CONFIG_SWIOTLB=y
++CONFIG_KEXEC=y
++# CONFIG_CRASH_DUMP is not set
++CONFIG_PHYSICAL_START=0x200000
++CONFIG_SECCOMP=y
++CONFIG_HZ_100=y
++# CONFIG_HZ_250 is not set
++# CONFIG_HZ_1000 is not set
++CONFIG_HZ=100
++# CONFIG_REORDER is not set
++CONFIG_GENERIC_HARDIRQS=y
++CONFIG_GENERIC_IRQ_PROBE=y
++CONFIG_ISA_DMA_API=y
++CONFIG_GENERIC_PENDING_IRQ=y
++
++#
++# Power management options
++#
++CONFIG_PM=y
++CONFIG_PM_LEGACY=y
++# CONFIG_PM_DEBUG is not set
++# CONFIG_SOFTWARE_SUSPEND is not set
++CONFIG_SUSPEND_SMP=y
++
++#
++# ACPI (Advanced Configuration and Power Interface) Support
++#
++CONFIG_ACPI=y
++CONFIG_ACPI_SLEEP=y
++CONFIG_ACPI_SLEEP_PROC_FS=y
++# CONFIG_ACPI_SLEEP_PROC_SLEEP is not set
++CONFIG_ACPI_AC=m
++CONFIG_ACPI_BATTERY=m
++CONFIG_ACPI_BUTTON=m
++CONFIG_ACPI_VIDEO=m
++CONFIG_ACPI_HOTKEY=m
++CONFIG_ACPI_FAN=m
++CONFIG_ACPI_DOCK=m
++CONFIG_ACPI_PROCESSOR=m
++CONFIG_ACPI_HOTPLUG_CPU=y
++CONFIG_ACPI_THERMAL=m
++CONFIG_ACPI_ASUS=m
++CONFIG_ACPI_IBM=m
++CONFIG_ACPI_TOSHIBA=m
++# CONFIG_ACPI_CUSTOM_DSDT is not set
++CONFIG_ACPI_BLACKLIST_YEAR=0
++# CONFIG_ACPI_DEBUG is not set
++CONFIG_ACPI_EC=y
++CONFIG_ACPI_POWER=y
++CONFIG_ACPI_SYSTEM=y
++CONFIG_ACPI_CONTAINER=m
++CONFIG_ACPI_PV_SLEEP=y
++
++#
++# CPU Frequency scaling
++#
++# CONFIG_CPU_FREQ is not set
++
++#
++# Bus options (PCI etc.)
++#
++CONFIG_PCI=y
++CONFIG_PCI_DIRECT=y
++CONFIG_PCI_MMCONFIG=y
++CONFIG_XEN_PCIDEV_FRONTEND=y
++# CONFIG_XEN_PCIDEV_FE_DEBUG is not set
++# CONFIG_PCIEPORTBUS is not set
++CONFIG_PCI_MSI=y
++# CONFIG_PCI_DEBUG is not set
++
++#
++# PCCARD (PCMCIA/CardBus) support
++#
++# CONFIG_PCCARD is not set
++
++#
++# PCI Hotplug Support
++#
++# CONFIG_HOTPLUG_PCI is not set
++
++#
++# Executable file formats / Emulations
++#
++CONFIG_BINFMT_ELF=y
++# CONFIG_BINFMT_MISC is not set
++CONFIG_IA32_EMULATION=y
++CONFIG_IA32_AOUT=y
++CONFIG_COMPAT=y
++CONFIG_SYSVIPC_COMPAT=y
++
++#
++# Networking
++#
++CONFIG_NET=y
++
++#
++# Networking options
++#
++# CONFIG_NETDEBUG is not set
++CONFIG_PACKET=y
++# CONFIG_PACKET_MMAP is not set
++CONFIG_UNIX=y
++CONFIG_XFRM=y
++# CONFIG_XFRM_USER is not set
++# CONFIG_NET_KEY is not set
++CONFIG_INET=y
++# CONFIG_IP_MULTICAST is not set
++# CONFIG_IP_ADVANCED_ROUTER is not set
++CONFIG_IP_FIB_HASH=y
++CONFIG_IP_PNP=y
++CONFIG_IP_PNP_DHCP=y
++# CONFIG_IP_PNP_BOOTP is not set
++# CONFIG_IP_PNP_RARP is not set
++# CONFIG_NET_IPIP is not set
++# CONFIG_NET_IPGRE is not set
++# CONFIG_ARPD is not set
++# CONFIG_SYN_COOKIES is not set
++# CONFIG_INET_AH is not set
++# CONFIG_INET_ESP is not set
++# CONFIG_INET_IPCOMP is not set
++# CONFIG_INET_XFRM_TUNNEL is not set
++# CONFIG_INET_TUNNEL is not set
++CONFIG_INET_XFRM_MODE_TRANSPORT=y
++CONFIG_INET_XFRM_MODE_TUNNEL=y
++# CONFIG_INET_DIAG is not set
++# CONFIG_TCP_CONG_ADVANCED is not set
++CONFIG_TCP_CONG_BIC=y
++
++#
++# IP: Virtual Server Configuration
++#
++# CONFIG_IP_VS is not set
++# CONFIG_IPV6 is not set
++# CONFIG_INET6_XFRM_TUNNEL is not set
++# CONFIG_INET6_TUNNEL is not set
++# CONFIG_NETWORK_SECMARK is not set
++CONFIG_NETFILTER=y
++# CONFIG_NETFILTER_DEBUG is not set
++CONFIG_BRIDGE_NETFILTER=y
++
++#
++# Core Netfilter Configuration
++#
++# CONFIG_NETFILTER_NETLINK is not set
++# CONFIG_NETFILTER_XTABLES is not set
++
++#
++# IP: Netfilter Configuration
++#
++CONFIG_IP_NF_CONNTRACK=m
++CONFIG_IP_NF_CT_ACCT=y
++# CONFIG_IP_NF_CONNTRACK_MARK is not set
++# CONFIG_IP_NF_CONNTRACK_EVENTS is not set
++# CONFIG_IP_NF_CT_PROTO_SCTP is not set
++CONFIG_IP_NF_FTP=m
++# CONFIG_IP_NF_IRC is not set
++# CONFIG_IP_NF_NETBIOS_NS is not set
++# CONFIG_IP_NF_TFTP is not set
++# CONFIG_IP_NF_AMANDA is not set
++# CONFIG_IP_NF_PPTP is not set
++# CONFIG_IP_NF_H323 is not set
++# CONFIG_IP_NF_SIP is not set
++# CONFIG_IP_NF_QUEUE is not set
++
++#
++# Bridge: Netfilter Configuration
++#
++# CONFIG_BRIDGE_NF_EBTABLES is not set
++
++#
++# DCCP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_DCCP is not set
++
++#
++# SCTP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_SCTP is not set
++
++#
++# TIPC Configuration (EXPERIMENTAL)
++#
++# CONFIG_TIPC is not set
++# CONFIG_ATM is not set
++CONFIG_BRIDGE=y
++# CONFIG_VLAN_8021Q is not set
++# CONFIG_DECNET is not set
++CONFIG_LLC=y
++# CONFIG_LLC2 is not set
++# CONFIG_IPX is not set
++# CONFIG_ATALK is not set
++# CONFIG_X25 is not set
++# CONFIG_LAPB is not set
++# CONFIG_ECONET is not set
++# CONFIG_WAN_ROUTER is not set
++
++#
++# QoS and/or fair queueing
++#
++# CONFIG_NET_SCHED is not set
++
++#
++# Network testing
++#
++# CONFIG_NET_PKTGEN is not set
++# CONFIG_HAMRADIO is not set
++# CONFIG_IRDA is not set
++# CONFIG_BT is not set
++# CONFIG_IEEE80211 is not set
++
++#
++# Device Drivers
++#
++
++#
++# Generic Driver Options
++#
++# CONFIG_STANDALONE is not set
++CONFIG_PREVENT_FIRMWARE_BUILD=y
++# CONFIG_FW_LOADER is not set
++# CONFIG_DEBUG_DRIVER is not set
++CONFIG_SYS_HYPERVISOR=y
++
++#
++# Connector - unified userspace <-> kernelspace linker
++#
++# CONFIG_CONNECTOR is not set
++
++#
++# Memory Technology Devices (MTD)
++#
++# CONFIG_MTD is not set
++
++#
++# Parallel port support
++#
++# CONFIG_PARPORT is not set
++
++#
++# Plug and Play support
++#
++CONFIG_PNP=y
++CONFIG_PNP_DEBUG=y
++
++#
++# Protocols
++#
++CONFIG_PNPACPI=y
++
++#
++# Block devices
++#
++CONFIG_BLK_DEV_FD=y
++# CONFIG_BLK_CPQ_DA is not set
++CONFIG_BLK_CPQ_CISS_DA=y
++# CONFIG_CISS_SCSI_TAPE is not set
++# CONFIG_BLK_DEV_DAC960 is not set
++# CONFIG_BLK_DEV_UMEM is not set
++# CONFIG_BLK_DEV_COW_COMMON is not set
++CONFIG_BLK_DEV_LOOP=y
++# CONFIG_BLK_DEV_CRYPTOLOOP is not set
++# CONFIG_BLK_DEV_NBD is not set
++# CONFIG_BLK_DEV_SX8 is not set
++# CONFIG_BLK_DEV_UB is not set
++CONFIG_BLK_DEV_RAM=y
++CONFIG_BLK_DEV_RAM_COUNT=16
++CONFIG_BLK_DEV_RAM_SIZE=4096
++CONFIG_BLK_DEV_RAM_BLOCKSIZE=1024
++CONFIG_BLK_DEV_INITRD=y
++# CONFIG_CDROM_PKTCDVD is not set
++# CONFIG_ATA_OVER_ETH is not set
++
++#
++# ATA/ATAPI/MFM/RLL support
++#
++CONFIG_IDE=y
++CONFIG_BLK_DEV_IDE=y
++
++#
++# Please see Documentation/ide.txt for help/info on IDE drives
++#
++# CONFIG_BLK_DEV_IDE_SATA is not set
++# CONFIG_BLK_DEV_HD_IDE is not set
++CONFIG_BLK_DEV_IDEDISK=y
++# CONFIG_IDEDISK_MULTI_MODE is not set
++CONFIG_BLK_DEV_IDECD=y
++# CONFIG_BLK_DEV_IDETAPE is not set
++# CONFIG_BLK_DEV_IDEFLOPPY is not set
++# CONFIG_BLK_DEV_IDESCSI is not set
++# CONFIG_IDE_TASK_IOCTL is not set
++
++#
++# IDE chipset support/bugfixes
++#
++CONFIG_IDE_GENERIC=y
++# CONFIG_BLK_DEV_CMD640 is not set
++# CONFIG_BLK_DEV_IDEPNP is not set
++CONFIG_BLK_DEV_IDEPCI=y
++# CONFIG_IDEPCI_SHARE_IRQ is not set
++# CONFIG_BLK_DEV_OFFBOARD is not set
++CONFIG_BLK_DEV_GENERIC=y
++# CONFIG_BLK_DEV_OPTI621 is not set
++# CONFIG_BLK_DEV_RZ1000 is not set
++CONFIG_BLK_DEV_IDEDMA_PCI=y
++# CONFIG_BLK_DEV_IDEDMA_FORCED is not set
++CONFIG_IDEDMA_PCI_AUTO=y
++# CONFIG_IDEDMA_ONLYDISK is not set
++# CONFIG_BLK_DEV_AEC62XX is not set
++# CONFIG_BLK_DEV_ALI15X3 is not set
++# CONFIG_BLK_DEV_AMD74XX is not set
++# CONFIG_BLK_DEV_ATIIXP is not set
++# CONFIG_BLK_DEV_CMD64X is not set
++# CONFIG_BLK_DEV_TRIFLEX is not set
++# CONFIG_BLK_DEV_CY82C693 is not set
++# CONFIG_BLK_DEV_CS5520 is not set
++# CONFIG_BLK_DEV_CS5530 is not set
++# CONFIG_BLK_DEV_HPT34X is not set
++# CONFIG_BLK_DEV_HPT366 is not set
++# CONFIG_BLK_DEV_SC1200 is not set
++CONFIG_BLK_DEV_PIIX=y
++# CONFIG_BLK_DEV_IT821X is not set
++# CONFIG_BLK_DEV_NS87415 is not set
++# CONFIG_BLK_DEV_PDC202XX_OLD is not set
++# CONFIG_BLK_DEV_PDC202XX_NEW is not set
++CONFIG_BLK_DEV_SVWKS=y
++# CONFIG_BLK_DEV_SIIMAGE is not set
++# CONFIG_BLK_DEV_SIS5513 is not set
++# CONFIG_BLK_DEV_SLC90E66 is not set
++# CONFIG_BLK_DEV_TRM290 is not set
++# CONFIG_BLK_DEV_VIA82CXXX is not set
++# CONFIG_IDE_ARM is not set
++CONFIG_BLK_DEV_IDEDMA=y
++# CONFIG_IDEDMA_IVB is not set
++CONFIG_IDEDMA_AUTO=y
++# CONFIG_BLK_DEV_HD is not set
++
++#
++# SCSI device support
++#
++# CONFIG_RAID_ATTRS is not set
++CONFIG_SCSI=y
++CONFIG_SCSI_PROC_FS=y
++
++#
++# SCSI support type (disk, tape, CD-ROM)
++#
++CONFIG_BLK_DEV_SD=y
++# CONFIG_CHR_DEV_ST is not set
++# CONFIG_CHR_DEV_OSST is not set
++# CONFIG_BLK_DEV_SR is not set
++# CONFIG_CHR_DEV_SG is not set
++# CONFIG_CHR_DEV_SCH is not set
++
++#
++# Some SCSI devices (e.g. CD jukebox) support multiple LUNs
++#
++# CONFIG_SCSI_MULTI_LUN is not set
++# CONFIG_SCSI_CONSTANTS is not set
++# CONFIG_SCSI_LOGGING is not set
++
++#
++# SCSI Transport Attributes
++#
++CONFIG_SCSI_SPI_ATTRS=y
++# CONFIG_SCSI_FC_ATTRS is not set
++# CONFIG_SCSI_ISCSI_ATTRS is not set
++# CONFIG_SCSI_SAS_ATTRS is not set
++
++#
++# SCSI low-level drivers
++#
++# CONFIG_ISCSI_TCP is not set
++CONFIG_BLK_DEV_3W_XXXX_RAID=y
++# CONFIG_SCSI_3W_9XXX is not set
++# CONFIG_SCSI_ACARD is not set
++CONFIG_SCSI_AACRAID=y
++CONFIG_SCSI_AIC7XXX=y
++CONFIG_AIC7XXX_CMDS_PER_DEVICE=32
++CONFIG_AIC7XXX_RESET_DELAY_MS=15000
++CONFIG_AIC7XXX_DEBUG_ENABLE=y
++CONFIG_AIC7XXX_DEBUG_MASK=0
++CONFIG_AIC7XXX_REG_PRETTY_PRINT=y
++# CONFIG_SCSI_AIC7XXX_OLD is not set
++CONFIG_SCSI_AIC79XX=y
++CONFIG_AIC79XX_CMDS_PER_DEVICE=32
++CONFIG_AIC79XX_RESET_DELAY_MS=15000
++# CONFIG_AIC79XX_ENABLE_RD_STRM is not set
++CONFIG_AIC79XX_DEBUG_ENABLE=y
++CONFIG_AIC79XX_DEBUG_MASK=0
++CONFIG_AIC79XX_REG_PRETTY_PRINT=y
++CONFIG_MEGARAID_NEWGEN=y
++# CONFIG_MEGARAID_MM is not set
++# CONFIG_MEGARAID_LEGACY is not set
++# CONFIG_MEGARAID_SAS is not set
++CONFIG_SCSI_SATA=y
++CONFIG_SCSI_SATA_AHCI=y
++# CONFIG_SCSI_SATA_SVW is not set
++CONFIG_SCSI_ATA_PIIX=y
++# CONFIG_SCSI_SATA_MV is not set
++# CONFIG_SCSI_SATA_NV is not set
++# CONFIG_SCSI_PDC_ADMA is not set
++# CONFIG_SCSI_HPTIOP is not set
++# CONFIG_SCSI_SATA_QSTOR is not set
++CONFIG_SCSI_SATA_PROMISE=y
++CONFIG_SCSI_SATA_SX4=y
++CONFIG_SCSI_SATA_SIL=y
++CONFIG_SCSI_SATA_SIL24=y
++# CONFIG_SCSI_SATA_SIS is not set
++# CONFIG_SCSI_SATA_ULI is not set
++# CONFIG_SCSI_SATA_VIA is not set
++# CONFIG_SCSI_SATA_VITESSE is not set
++CONFIG_SCSI_SATA_INTEL_COMBINED=y
++# CONFIG_SCSI_BUSLOGIC is not set
++# CONFIG_SCSI_DMX3191D is not set
++# CONFIG_SCSI_EATA is not set
++# CONFIG_SCSI_FUTURE_DOMAIN is not set
++# CONFIG_SCSI_GDTH is not set
++# CONFIG_SCSI_IPS is not set
++# CONFIG_SCSI_INITIO is not set
++# CONFIG_SCSI_INIA100 is not set
++# CONFIG_SCSI_SYM53C8XX_2 is not set
++# CONFIG_SCSI_IPR is not set
++# CONFIG_SCSI_QLOGIC_1280 is not set
++# CONFIG_SCSI_QLA_FC is not set
++# CONFIG_SCSI_LPFC is not set
++# CONFIG_SCSI_DC395x is not set
++# CONFIG_SCSI_DC390T is not set
++# CONFIG_SCSI_DEBUG is not set
++
++#
++# Multi-device support (RAID and LVM)
++#
++CONFIG_MD=y
++CONFIG_BLK_DEV_MD=y
++# CONFIG_MD_LINEAR is not set
++CONFIG_MD_RAID0=y
++CONFIG_MD_RAID1=y
++# CONFIG_MD_RAID10 is not set
++# CONFIG_MD_RAID456 is not set
++# CONFIG_MD_MULTIPATH is not set
++# CONFIG_MD_FAULTY is not set
++CONFIG_BLK_DEV_DM=y
++# CONFIG_DM_CRYPT is not set
++CONFIG_DM_SNAPSHOT=y
++CONFIG_DM_MIRROR=y
++# CONFIG_DM_ZERO is not set
++# CONFIG_DM_MULTIPATH is not set
++
++#
++# Fusion MPT device support
++#
++CONFIG_FUSION=y
++CONFIG_FUSION_SPI=y
++# CONFIG_FUSION_FC is not set
++# CONFIG_FUSION_SAS is not set
++CONFIG_FUSION_MAX_SGE=128
++# CONFIG_FUSION_CTL is not set
++
++#
++# IEEE 1394 (FireWire) support
++#
++# CONFIG_IEEE1394 is not set
++
++#
++# I2O device support
++#
++# CONFIG_I2O is not set
++
++#
++# Network device support
++#
++CONFIG_NETDEVICES=y
++# CONFIG_DUMMY is not set
++# CONFIG_BONDING is not set
++# CONFIG_EQUALIZER is not set
++CONFIG_TUN=y
++# CONFIG_NET_SB1000 is not set
++
++#
++# ARCnet devices
++#
++# CONFIG_ARCNET is not set
++
++#
++# PHY device support
++#
++# CONFIG_PHYLIB is not set
++
++#
++# Ethernet (10 or 100Mbit)
++#
++CONFIG_NET_ETHERNET=y
++CONFIG_MII=y
++# CONFIG_HAPPYMEAL is not set
++# CONFIG_SUNGEM is not set
++# CONFIG_CASSINI is not set
++CONFIG_NET_VENDOR_3COM=y
++CONFIG_VORTEX=y
++# CONFIG_TYPHOON is not set
++
++#
++# Tulip family network device support
++#
++CONFIG_NET_TULIP=y
++# CONFIG_DE2104X is not set
++CONFIG_TULIP=y
++# CONFIG_TULIP_MWI is not set
++# CONFIG_TULIP_MMIO is not set
++# CONFIG_TULIP_NAPI is not set
++# CONFIG_DE4X5 is not set
++# CONFIG_WINBOND_840 is not set
++# CONFIG_DM9102 is not set
++# CONFIG_ULI526X is not set
++# CONFIG_HP100 is not set
++CONFIG_NET_PCI=y
++CONFIG_PCNET32=y
++# CONFIG_AMD8111_ETH is not set
++# CONFIG_ADAPTEC_STARFIRE is not set
++# CONFIG_B44 is not set
++# CONFIG_FORCEDETH is not set
++# CONFIG_DGRS is not set
++# CONFIG_EEPRO100 is not set
++CONFIG_E100=y
++# CONFIG_FEALNX is not set
++# CONFIG_NATSEMI is not set
++CONFIG_NE2K_PCI=y
++# CONFIG_8139CP is not set
++CONFIG_8139TOO=y
++CONFIG_8139TOO_PIO=y
++# CONFIG_8139TOO_TUNE_TWISTER is not set
++# CONFIG_8139TOO_8129 is not set
++# CONFIG_8139_OLD_RX_RESET is not set
++# CONFIG_SIS900 is not set
++# CONFIG_EPIC100 is not set
++# CONFIG_SUNDANCE is not set
++CONFIG_VIA_RHINE=y
++# CONFIG_VIA_RHINE_MMIO is not set
++# CONFIG_VIA_RHINE_NAPI is not set
++
++#
++# Ethernet (1000 Mbit)
++#
++CONFIG_ACENIC=y
++# CONFIG_ACENIC_OMIT_TIGON_I is not set
++# CONFIG_DL2K is not set
++CONFIG_E1000=y
++# CONFIG_E1000_NAPI is not set
++# CONFIG_E1000_DISABLE_PACKET_SPLIT is not set
++# CONFIG_NS83820 is not set
++# CONFIG_HAMACHI is not set
++# CONFIG_YELLOWFIN is not set
++# CONFIG_R8169 is not set
++# CONFIG_SIS190 is not set
++# CONFIG_SKGE is not set
++# CONFIG_SKY2 is not set
++CONFIG_SK98LIN=y
++# CONFIG_VIA_VELOCITY is not set
++CONFIG_TIGON3=y
++CONFIG_BNX2=y
++
++#
++# Ethernet (10000 Mbit)
++#
++# CONFIG_CHELSIO_T1 is not set
++# CONFIG_IXGB is not set
++# CONFIG_S2IO is not set
++# CONFIG_MYRI10GE is not set
++# CONFIG_SFC is not set
++
++#
++# Token Ring devices
++#
++# CONFIG_TR is not set
++
++#
++# Wireless LAN (non-hamradio)
++#
++# CONFIG_NET_RADIO is not set
++
++#
++# Wan interfaces
++#
++# CONFIG_WAN is not set
++# CONFIG_FDDI is not set
++# CONFIG_HIPPI is not set
++# CONFIG_PPP is not set
++# CONFIG_SLIP is not set
++# CONFIG_NET_FC is not set
++# CONFIG_SHAPER is not set
++# CONFIG_NETCONSOLE is not set
++# CONFIG_NETPOLL is not set
++# CONFIG_NET_POLL_CONTROLLER is not set
++
++#
++# ISDN subsystem
++#
++# CONFIG_ISDN is not set
++
++#
++# Telephony Support
++#
++# CONFIG_PHONE is not set
++
++#
++# Input device support
++#
++CONFIG_INPUT=y
++
++#
++# Userland interfaces
++#
++CONFIG_INPUT_MOUSEDEV=y
++CONFIG_INPUT_MOUSEDEV_PSAUX=y
++CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024
++CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768
++# CONFIG_INPUT_JOYDEV is not set
++# CONFIG_INPUT_TSDEV is not set
++# CONFIG_INPUT_EVDEV is not set
++# CONFIG_INPUT_EVBUG is not set
++
++#
++# Input Device Drivers
++#
++CONFIG_INPUT_KEYBOARD=y
++CONFIG_KEYBOARD_ATKBD=y
++# CONFIG_KEYBOARD_SUNKBD is not set
++# CONFIG_KEYBOARD_LKKBD is not set
++# CONFIG_KEYBOARD_XTKBD is not set
++# CONFIG_KEYBOARD_NEWTON is not set
++CONFIG_INPUT_MOUSE=y
++CONFIG_MOUSE_PS2=y
++# CONFIG_MOUSE_SERIAL is not set
++# CONFIG_MOUSE_VSXXXAA is not set
++# CONFIG_INPUT_JOYSTICK is not set
++# CONFIG_INPUT_TOUCHSCREEN is not set
++# CONFIG_INPUT_MISC is not set
++
++#
++# Hardware I/O ports
++#
++CONFIG_SERIO=y
++CONFIG_SERIO_I8042=y
++CONFIG_SERIO_SERPORT=y
++# CONFIG_SERIO_CT82C710 is not set
++# CONFIG_SERIO_PCIPS2 is not set
++CONFIG_SERIO_LIBPS2=y
++# CONFIG_SERIO_RAW is not set
++# CONFIG_GAMEPORT is not set
++
++#
++# Character devices
++#
++CONFIG_VT=y
++CONFIG_VT_CONSOLE=y
++CONFIG_HW_CONSOLE=y
++CONFIG_VT_HW_CONSOLE_BINDING=y
++# CONFIG_SERIAL_NONSTANDARD is not set
++
++#
++# Serial drivers
++#
++
++#
++# Non-8250 serial port support
++#
++# CONFIG_SERIAL_JSM is not set
++CONFIG_UNIX98_PTYS=y
++CONFIG_LEGACY_PTYS=y
++CONFIG_LEGACY_PTY_COUNT=256
++
++#
++# IPMI
++#
++# CONFIG_IPMI_HANDLER is not set
++
++#
++# Watchdog Cards
++#
++# CONFIG_WATCHDOG is not set
++# CONFIG_HW_RANDOM is not set
++# CONFIG_NVRAM is not set
++# CONFIG_RTC is not set
++# CONFIG_GEN_RTC is not set
++# CONFIG_DTLK is not set
++# CONFIG_R3964 is not set
++# CONFIG_APPLICOM is not set
++
++#
++# Ftape, the floppy tape device driver
++#
++CONFIG_AGP=m
++# CONFIG_AGP_AMD64 is not set
++CONFIG_AGP_INTEL=m
++CONFIG_AGP_SIS=m
++CONFIG_AGP_VIA=m
++CONFIG_DRM=m
++CONFIG_DRM_TDFX=m
++CONFIG_DRM_R128=m
++CONFIG_DRM_RADEON=m
++CONFIG_DRM_I810=m
++CONFIG_DRM_I830=m
++CONFIG_DRM_I915=m
++CONFIG_DRM_MGA=m
++CONFIG_DRM_SIS=m
++# CONFIG_DRM_VIA is not set
++# CONFIG_DRM_SAVAGE is not set
++# CONFIG_MWAVE is not set
++# CONFIG_PC8736x_GPIO is not set
++# CONFIG_RAW_DRIVER is not set
++# CONFIG_HPET is not set
++# CONFIG_HANGCHECK_TIMER is not set
++
++#
++# TPM devices
++#
++# CONFIG_TCG_TPM is not set
++# CONFIG_TELCLOCK is not set
++
++#
++# I2C support
++#
++# CONFIG_I2C is not set
++
++#
++# SPI support
++#
++# CONFIG_SPI is not set
++# CONFIG_SPI_MASTER is not set
++
++#
++# Dallas's 1-wire bus
++#
++
++#
++# Hardware Monitoring support
++#
++# CONFIG_HWMON is not set
++# CONFIG_HWMON_VID is not set
++
++#
++# Misc devices
++#
++# CONFIG_IBM_ASM is not set
++
++#
++# Multimedia devices
++#
++# CONFIG_VIDEO_DEV is not set
++
++#
++# Digital Video Broadcasting Devices
++#
++# CONFIG_DVB is not set
++# CONFIG_USB_DABUSB is not set
++
++#
++# Graphics support
++#
++CONFIG_FIRMWARE_EDID=y
++# CONFIG_FB is not set
++
++#
++# Console display driver support
++#
++CONFIG_VGA_CONSOLE=y
++# CONFIG_VGACON_SOFT_SCROLLBACK is not set
++CONFIG_DUMMY_CONSOLE=y
++# CONFIG_BACKLIGHT_LCD_SUPPORT is not set
++
++#
++# Sound
++#
++# CONFIG_SOUND is not set
++
++#
++# USB support
++#
++CONFIG_USB_ARCH_HAS_HCD=y
++CONFIG_USB_ARCH_HAS_OHCI=y
++CONFIG_USB_ARCH_HAS_EHCI=y
++CONFIG_USB=y
++# CONFIG_USB_DEBUG is not set
++
++#
++# Miscellaneous USB options
++#
++# CONFIG_USB_DEVICEFS is not set
++# CONFIG_USB_BANDWIDTH is not set
++# CONFIG_USB_DYNAMIC_MINORS is not set
++# CONFIG_USB_SUSPEND is not set
++# CONFIG_USB_OTG is not set
++
++#
++# USB Host Controller Drivers
++#
++# CONFIG_USB_EHCI_HCD is not set
++# CONFIG_USB_ISP116X_HCD is not set
++CONFIG_USB_OHCI_HCD=y
++# CONFIG_USB_OHCI_BIG_ENDIAN is not set
++CONFIG_USB_OHCI_LITTLE_ENDIAN=y
++CONFIG_USB_UHCI_HCD=y
++# CONFIG_USB_SL811_HCD is not set
++
++#
++# USB Device Class drivers
++#
++# CONFIG_USB_ACM is not set
++# CONFIG_USB_PRINTER is not set
++
++#
++# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support'
++#
++
++#
++# may also be needed; see USB_STORAGE Help for more information
++#
++# CONFIG_USB_STORAGE is not set
++# CONFIG_USB_LIBUSUAL is not set
++
++#
++# USB Input Devices
++#
++CONFIG_USB_HID=y
++CONFIG_USB_HIDINPUT=y
++# CONFIG_USB_HIDINPUT_POWERBOOK is not set
++# CONFIG_HID_FF is not set
++# CONFIG_USB_HIDDEV is not set
++# CONFIG_USB_AIPTEK is not set
++# CONFIG_USB_WACOM is not set
++# CONFIG_USB_ACECAD is not set
++# CONFIG_USB_KBTAB is not set
++# CONFIG_USB_POWERMATE is not set
++# CONFIG_USB_TOUCHSCREEN is not set
++# CONFIG_USB_YEALINK is not set
++# CONFIG_USB_XPAD is not set
++# CONFIG_USB_ATI_REMOTE is not set
++# CONFIG_USB_ATI_REMOTE2 is not set
++# CONFIG_USB_KEYSPAN_REMOTE is not set
++# CONFIG_USB_APPLETOUCH is not set
++
++#
++# USB Imaging devices
++#
++# CONFIG_USB_MDC800 is not set
++# CONFIG_USB_MICROTEK is not set
++
++#
++# USB Network Adapters
++#
++# CONFIG_USB_CATC is not set
++# CONFIG_USB_KAWETH is not set
++# CONFIG_USB_PEGASUS is not set
++# CONFIG_USB_RTL8150 is not set
++# CONFIG_USB_USBNET is not set
++CONFIG_USB_MON=y
++
++#
++# USB port drivers
++#
++
++#
++# USB Serial Converter support
++#
++# CONFIG_USB_SERIAL is not set
++
++#
++# USB Miscellaneous drivers
++#
++# CONFIG_USB_EMI62 is not set
++# CONFIG_USB_EMI26 is not set
++# CONFIG_USB_AUERSWALD is not set
++# CONFIG_USB_RIO500 is not set
++# CONFIG_USB_LEGOTOWER is not set
++# CONFIG_USB_LCD is not set
++# CONFIG_USB_LED is not set
++# CONFIG_USB_CYPRESS_CY7C63 is not set
++# CONFIG_USB_CYTHERM is not set
++# CONFIG_USB_PHIDGETKIT is not set
++# CONFIG_USB_PHIDGETSERVO is not set
++# CONFIG_USB_IDMOUSE is not set
++# CONFIG_USB_APPLEDISPLAY is not set
++# CONFIG_USB_LD is not set
++
++#
++# USB DSL modem support
++#
++
++#
++# USB Gadget Support
++#
++# CONFIG_USB_GADGET is not set
++
++#
++# MMC/SD Card support
++#
++# CONFIG_MMC is not set
++
++#
++# LED devices
++#
++# CONFIG_NEW_LEDS is not set
++
++#
++# LED drivers
++#
++
++#
++# LED Triggers
++#
++
++#
++# InfiniBand support
++#
++# CONFIG_INFINIBAND is not set
++
++#
++# EDAC - error detection and reporting (RAS) (EXPERIMENTAL)
++#
++# CONFIG_EDAC is not set
++
++#
++# Real Time Clock
++#
++CONFIG_RTC_LIB=m
++CONFIG_RTC_CLASS=m
++
++#
++# RTC interfaces
++#
++CONFIG_RTC_INTF_SYSFS=m
++CONFIG_RTC_INTF_PROC=m
++CONFIG_RTC_INTF_DEV=m
++# CONFIG_RTC_INTF_DEV_UIE_EMUL is not set
++
++#
++# RTC drivers
++#
++# CONFIG_RTC_DRV_DS1553 is not set
++# CONFIG_RTC_DRV_DS1742 is not set
++CONFIG_RTC_DRV_M48T86=m
++CONFIG_RTC_DRV_TEST=m
++# CONFIG_RTC_DRV_V3020 is not set
++
++#
++# DMA Engine support
++#
++# CONFIG_DMA_ENGINE is not set
++
++#
++# DMA Clients
++#
++
++#
++# DMA Devices
++#
++
++#
++# Firmware Drivers
++#
++CONFIG_EDD=y
++# CONFIG_DELL_RBU is not set
++# CONFIG_DCDBAS is not set
++
++#
++# File systems
++#
++CONFIG_EXT2_FS=y
++# CONFIG_EXT2_FS_XATTR is not set
++# CONFIG_EXT2_FS_XIP is not set
++CONFIG_EXT3_FS=y
++CONFIG_EXT3_FS_XATTR=y
++# CONFIG_EXT3_FS_POSIX_ACL is not set
++# CONFIG_EXT3_FS_SECURITY is not set
++CONFIG_JBD=y
++# CONFIG_JBD_DEBUG is not set
++CONFIG_FS_MBCACHE=y
++CONFIG_REISERFS_FS=y
++# CONFIG_REISERFS_CHECK is not set
++# CONFIG_REISERFS_PROC_INFO is not set
++# CONFIG_REISERFS_FS_XATTR is not set
++# CONFIG_JFS_FS is not set
++# CONFIG_FS_POSIX_ACL is not set
++# CONFIG_XFS_FS is not set
++# CONFIG_OCFS2_FS is not set
++# CONFIG_MINIX_FS is not set
++# CONFIG_ROMFS_FS is not set
++CONFIG_INOTIFY=y
++CONFIG_INOTIFY_USER=y
++# CONFIG_QUOTA is not set
++CONFIG_DNOTIFY=y
++CONFIG_AUTOFS_FS=y
++CONFIG_AUTOFS4_FS=y
++# CONFIG_FUSE_FS is not set
++
++#
++# CD-ROM/DVD Filesystems
++#
++CONFIG_ISO9660_FS=y
++CONFIG_JOLIET=y
++CONFIG_ZISOFS=y
++CONFIG_ZISOFS_FS=y
++# CONFIG_UDF_FS is not set
++
++#
++# DOS/FAT/NT Filesystems
++#
++CONFIG_FAT_FS=m
++CONFIG_MSDOS_FS=m
++CONFIG_VFAT_FS=m
++CONFIG_FAT_DEFAULT_CODEPAGE=437
++CONFIG_FAT_DEFAULT_IOCHARSET="iso8859-1"
++# CONFIG_NTFS_FS is not set
++
++#
++# Pseudo filesystems
++#
++CONFIG_PROC_FS=y
++CONFIG_PROC_KCORE=y
++CONFIG_SYSFS=y
++CONFIG_TMPFS=y
++# CONFIG_HUGETLB_PAGE is not set
++CONFIG_RAMFS=y
++# CONFIG_CONFIGFS_FS is not set
++
++#
++# Miscellaneous filesystems
++#
++# CONFIG_ADFS_FS is not set
++# CONFIG_AFFS_FS is not set
++# CONFIG_HFS_FS is not set
++# CONFIG_HFSPLUS_FS is not set
++# CONFIG_BEFS_FS is not set
++# CONFIG_BFS_FS is not set
++# CONFIG_EFS_FS is not set
++CONFIG_CRAMFS=y
++# CONFIG_VXFS_FS is not set
++# CONFIG_HPFS_FS is not set
++# CONFIG_QNX4FS_FS is not set
++# CONFIG_SYSV_FS is not set
++# CONFIG_UFS_FS is not set
++
++#
++# Network File Systems
++#
++CONFIG_NFS_FS=y
++CONFIG_NFS_V3=y
++# CONFIG_NFS_V3_ACL is not set
++# CONFIG_NFS_V4 is not set
++# CONFIG_NFS_DIRECTIO is not set
++CONFIG_NFSD=m
++CONFIG_NFSD_V3=y
++# CONFIG_NFSD_V3_ACL is not set
++# CONFIG_NFSD_V4 is not set
++CONFIG_NFSD_TCP=y
++CONFIG_ROOT_NFS=y
++CONFIG_LOCKD=y
++CONFIG_LOCKD_V4=y
++CONFIG_EXPORTFS=m
++CONFIG_NFS_COMMON=y
++CONFIG_SUNRPC=y
++# CONFIG_RPCSEC_GSS_KRB5 is not set
++# CONFIG_RPCSEC_GSS_SPKM3 is not set
++# CONFIG_SMB_FS is not set
++# CONFIG_CIFS is not set
++# CONFIG_NCP_FS is not set
++# CONFIG_CODA_FS is not set
++# CONFIG_AFS_FS is not set
++# CONFIG_9P_FS is not set
++
++#
++# Partition Types
++#
++# CONFIG_PARTITION_ADVANCED is not set
++CONFIG_MSDOS_PARTITION=y
++
++#
++# Native Language Support
++#
++CONFIG_NLS=y
++CONFIG_NLS_DEFAULT="iso8859-1"
++CONFIG_NLS_CODEPAGE_437=y
++# CONFIG_NLS_CODEPAGE_737 is not set
++# CONFIG_NLS_CODEPAGE_775 is not set
++# CONFIG_NLS_CODEPAGE_850 is not set
++# CONFIG_NLS_CODEPAGE_852 is not set
++# CONFIG_NLS_CODEPAGE_855 is not set
++# CONFIG_NLS_CODEPAGE_857 is not set
++# CONFIG_NLS_CODEPAGE_860 is not set
++# CONFIG_NLS_CODEPAGE_861 is not set
++# CONFIG_NLS_CODEPAGE_862 is not set
++# CONFIG_NLS_CODEPAGE_863 is not set
++# CONFIG_NLS_CODEPAGE_864 is not set
++# CONFIG_NLS_CODEPAGE_865 is not set
++# CONFIG_NLS_CODEPAGE_866 is not set
++# CONFIG_NLS_CODEPAGE_869 is not set
++# CONFIG_NLS_CODEPAGE_936 is not set
++# CONFIG_NLS_CODEPAGE_950 is not set
++# CONFIG_NLS_CODEPAGE_932 is not set
++# CONFIG_NLS_CODEPAGE_949 is not set
++# CONFIG_NLS_CODEPAGE_874 is not set
++# CONFIG_NLS_ISO8859_8 is not set
++# CONFIG_NLS_CODEPAGE_1250 is not set
++# CONFIG_NLS_CODEPAGE_1251 is not set
++# CONFIG_NLS_ASCII is not set
++CONFIG_NLS_ISO8859_1=y
++# CONFIG_NLS_ISO8859_2 is not set
++# CONFIG_NLS_ISO8859_3 is not set
++# CONFIG_NLS_ISO8859_4 is not set
++# CONFIG_NLS_ISO8859_5 is not set
++# CONFIG_NLS_ISO8859_6 is not set
++# CONFIG_NLS_ISO8859_7 is not set
++# CONFIG_NLS_ISO8859_9 is not set
++# CONFIG_NLS_ISO8859_13 is not set
++# CONFIG_NLS_ISO8859_14 is not set
++# CONFIG_NLS_ISO8859_15 is not set
++# CONFIG_NLS_KOI8_R is not set
++# CONFIG_NLS_KOI8_U is not set
++# CONFIG_NLS_UTF8 is not set
++
++#
++# Instrumentation Support
++#
++# CONFIG_PROFILING is not set
++# CONFIG_KPROBES is not set
++
++#
++# Kernel hacking
++#
++CONFIG_TRACE_IRQFLAGS_SUPPORT=y
++# CONFIG_PRINTK_TIME is not set
++CONFIG_MAGIC_SYSRQ=y
++CONFIG_UNUSED_SYMBOLS=y
++CONFIG_DEBUG_KERNEL=y
++CONFIG_LOG_BUF_SHIFT=14
++CONFIG_DETECT_SOFTLOCKUP=y
++# CONFIG_SCHEDSTATS is not set
++# CONFIG_DEBUG_SLAB is not set
++# CONFIG_DEBUG_RT_MUTEXES is not set
++# CONFIG_RT_MUTEX_TESTER is not set
++# CONFIG_DEBUG_SPINLOCK is not set
++# CONFIG_DEBUG_MUTEXES is not set
++# CONFIG_DEBUG_RWSEMS is not set
++# CONFIG_DEBUG_LOCK_ALLOC is not set
++# CONFIG_PROVE_LOCKING is not set
++# CONFIG_DEBUG_SPINLOCK_SLEEP is not set
++# CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set
++# CONFIG_DEBUG_KOBJECT is not set
++CONFIG_DEBUG_INFO=y
++# CONFIG_DEBUG_FS is not set
++# CONFIG_DEBUG_VM is not set
++CONFIG_FRAME_POINTER=y
++# CONFIG_UNWIND_INFO is not set
++CONFIG_FORCED_INLINING=y
++# CONFIG_RCU_TORTURE_TEST is not set
++# CONFIG_DEBUG_RODATA is not set
++# CONFIG_DEBUG_STACKOVERFLOW is not set
++# CONFIG_DEBUG_STACK_USAGE is not set
++
++#
++# Security options
++#
++# CONFIG_KEYS is not set
++# CONFIG_SECURITY is not set
++
++#
++# Cryptographic options
++#
++CONFIG_CRYPTO=y
++CONFIG_CRYPTO_HMAC=y
++# CONFIG_CRYPTO_NULL is not set
++# CONFIG_CRYPTO_MD4 is not set
++CONFIG_CRYPTO_MD5=m
++CONFIG_CRYPTO_SHA1=m
++# CONFIG_CRYPTO_SHA256 is not set
++# CONFIG_CRYPTO_SHA512 is not set
++# CONFIG_CRYPTO_WP512 is not set
++# CONFIG_CRYPTO_TGR192 is not set
++CONFIG_CRYPTO_DES=m
++# CONFIG_CRYPTO_BLOWFISH is not set
++# CONFIG_CRYPTO_TWOFISH is not set
++# CONFIG_CRYPTO_SERPENT is not set
++# CONFIG_CRYPTO_AES is not set
++# CONFIG_CRYPTO_AES_X86_64 is not set
++# CONFIG_CRYPTO_CAST5 is not set
++# CONFIG_CRYPTO_CAST6 is not set
++# CONFIG_CRYPTO_TEA is not set
++# CONFIG_CRYPTO_ARC4 is not set
++# CONFIG_CRYPTO_KHAZAD is not set
++# CONFIG_CRYPTO_ANUBIS is not set
++# CONFIG_CRYPTO_DEFLATE is not set
++# CONFIG_CRYPTO_MICHAEL_MIC is not set
++CONFIG_CRYPTO_CRC32C=m
++# CONFIG_CRYPTO_TEST is not set
++
++#
++# Hardware crypto devices
++#
++CONFIG_XEN=y
++CONFIG_XEN_INTERFACE_VERSION=0x00030207
++
++#
++# XEN
++#
++CONFIG_XEN_PRIVILEGED_GUEST=y
++# CONFIG_XEN_UNPRIVILEGED_GUEST is not set
++CONFIG_XEN_PRIVCMD=y
++CONFIG_XEN_XENBUS_DEV=y
++CONFIG_XEN_BACKEND=y
++CONFIG_XEN_BLKDEV_BACKEND=y
++CONFIG_XEN_BLKDEV_TAP=y
++CONFIG_XEN_NETDEV_BACKEND=y
++# CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_UTIL is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_BACKEND is not set
++# CONFIG_XEN_NETDEV_LOOPBACK is not set
++CONFIG_XEN_PCIDEV_BACKEND=y
++# CONFIG_XEN_PCIDEV_BACKEND_VPCI is not set
++CONFIG_XEN_PCIDEV_BACKEND_PASS=y
++# CONFIG_XEN_PCIDEV_BACKEND_SLOT is not set
++# CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER is not set
++# CONFIG_XEN_PCIDEV_BE_DEBUG is not set
++# CONFIG_XEN_TPMDEV_BACKEND is not set
++# CONFIG_XEN_SCSI_BACKEND is not set
++CONFIG_XEN_BLKDEV_FRONTEND=y
++CONFIG_XEN_NETDEV_FRONTEND=y
++CONFIG_XEN_SCSI_FRONTEND=m
++CONFIG_XEN_GRANT_DEV=y
++# CONFIG_XEN_NETDEV_ACCEL_SFC_FRONTEND is not set
++CONFIG_XEN_SCRUB_PAGES=y
++CONFIG_XEN_DISABLE_SERIAL=y
++CONFIG_XEN_SYSFS=y
++CONFIG_XEN_COMPAT_030002_AND_LATER=y
++# CONFIG_XEN_COMPAT_030004_AND_LATER is not set
++# CONFIG_XEN_COMPAT_030100_AND_LATER is not set
++# CONFIG_XEN_COMPAT_LATEST_ONLY is not set
++CONFIG_XEN_COMPAT=0x030002
++CONFIG_HAVE_IRQ_IGNORE_UNHANDLED=y
++CONFIG_NO_IDLE_HZ=y
++CONFIG_XEN_SMPBOOT=y
++CONFIG_XEN_BALLOON=y
++CONFIG_XEN_DEVMEM=y
++
++#
++# Library routines
++#
++# CONFIG_CRC_CCITT is not set
++# CONFIG_CRC16 is not set
++CONFIG_CRC32=y
++CONFIG_LIBCRC32C=y
++CONFIG_ZLIB_INFLATE=y
++CONFIG_PLIST=y
+diff -r d894e36cfc30 -r 0aa021803deb buildconfigs/linux-defconfig_xenU_ia64
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/buildconfigs/linux-defconfig_xenU_ia64 Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1511 @@
++#
++# Automatically generated make config: don't edit
++# Linux kernel version: 2.6.18.8
++# Tue Feb 19 11:20:14 2008
++#
++CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config"
++
++#
++# Code maturity level options
++#
++CONFIG_EXPERIMENTAL=y
++CONFIG_LOCK_KERNEL=y
++CONFIG_INIT_ENV_ARG_LIMIT=32
++
++#
++# General setup
++#
++CONFIG_LOCALVERSION=""
++CONFIG_LOCALVERSION_AUTO=y
++CONFIG_SWAP=y
++CONFIG_SYSVIPC=y
++# CONFIG_POSIX_MQUEUE is not set
++CONFIG_BSD_PROCESS_ACCT=y
++# CONFIG_BSD_PROCESS_ACCT_V3 is not set
++# CONFIG_TASKSTATS is not set
++# CONFIG_AUDIT is not set
++# CONFIG_IKCONFIG is not set
++# CONFIG_CPUSETS is not set
++# CONFIG_RELAY is not set
++CONFIG_INITRAMFS_SOURCE=""
++CONFIG_CC_OPTIMIZE_FOR_SIZE=y
++# CONFIG_EMBEDDED is not set
++CONFIG_SYSCTL=y
++CONFIG_KALLSYMS=y
++# CONFIG_KALLSYMS_ALL is not set
++# CONFIG_KALLSYMS_EXTRA_PASS is not set
++CONFIG_HOTPLUG=y
++CONFIG_PRINTK=y
++CONFIG_BUG=y
++CONFIG_ELF_CORE=y
++CONFIG_BASE_FULL=y
++CONFIG_FUTEX=y
++CONFIG_EPOLL=y
++CONFIG_SHMEM=y
++CONFIG_SLAB=y
++CONFIG_VM_EVENT_COUNTERS=y
++CONFIG_RT_MUTEXES=y
++# CONFIG_TINY_SHMEM is not set
++CONFIG_BASE_SMALL=0
++# CONFIG_SLOB is not set
++
++#
++# Loadable module support
++#
++CONFIG_MODULES=y
++# CONFIG_MODULE_UNLOAD is not set
++# CONFIG_MODVERSIONS is not set
++# CONFIG_MODULE_SRCVERSION_ALL is not set
++# CONFIG_KMOD is not set
++CONFIG_STOP_MACHINE=y
++
++#
++# Block layer
++#
++# CONFIG_BLK_DEV_IO_TRACE is not set
++
++#
++# IO Schedulers
++#
++CONFIG_IOSCHED_NOOP=y
++CONFIG_IOSCHED_AS=y
++CONFIG_IOSCHED_DEADLINE=y
++CONFIG_IOSCHED_CFQ=y
++CONFIG_DEFAULT_AS=y
++# CONFIG_DEFAULT_DEADLINE is not set
++# CONFIG_DEFAULT_CFQ is not set
++# CONFIG_DEFAULT_NOOP is not set
++CONFIG_DEFAULT_IOSCHED="anticipatory"
++
++#
++# Processor type and features
++#
++CONFIG_IA64=y
++CONFIG_64BIT=y
++CONFIG_MMU=y
++CONFIG_SWIOTLB=y
++CONFIG_RWSEM_XCHGADD_ALGORITHM=y
++CONFIG_GENERIC_FIND_NEXT_BIT=y
++CONFIG_GENERIC_CALIBRATE_DELAY=y
++CONFIG_TIME_INTERPOLATION=y
++CONFIG_DMI=y
++CONFIG_EFI=y
++CONFIG_GENERIC_IOMAP=y
++CONFIG_XEN=y
++CONFIG_XEN_IA64_EXPOSE_P2M=y
++CONFIG_XEN_IA64_EXPOSE_P2M_USE_DTR=y
++CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER=y
++CONFIG_DMA_IS_DMA32=y
++CONFIG_AUDIT_ARCH=y
++# CONFIG_IA64_GENERIC is not set
++# CONFIG_IA64_DIG is not set
++# CONFIG_IA64_HP_ZX1 is not set
++# CONFIG_IA64_HP_ZX1_SWIOTLB is not set
++# CONFIG_IA64_SGI_SN2 is not set
++# CONFIG_IA64_HP_SIM is not set
++CONFIG_IA64_XEN=y
++# CONFIG_ITANIUM is not set
++CONFIG_MCKINLEY=y
++# CONFIG_IA64_PAGE_SIZE_4KB is not set
++# CONFIG_IA64_PAGE_SIZE_8KB is not set
++CONFIG_IA64_PAGE_SIZE_16KB=y
++# CONFIG_IA64_PAGE_SIZE_64KB is not set
++CONFIG_PGTABLE_3=y
++# CONFIG_PGTABLE_4 is not set
++CONFIG_HZ_100=y
++# CONFIG_HZ_250 is not set
++# CONFIG_HZ_1000 is not set
++CONFIG_HZ=100
++CONFIG_IA64_L1_CACHE_SHIFT=7
++# CONFIG_IA64_CYCLONE is not set
++CONFIG_IOSAPIC=y
++CONFIG_FORCE_MAX_ZONEORDER=11
++CONFIG_SMP=y
++CONFIG_NR_CPUS=16
++CONFIG_HOTPLUG_CPU=y
++CONFIG_ARCH_ENABLE_MEMORY_HOTPLUG=y
++# CONFIG_SCHED_SMT is not set
++# CONFIG_PERMIT_BSP_REMOVE is not set
++# CONFIG_PREEMPT is not set
++CONFIG_SELECT_MEMORY_MODEL=y
++CONFIG_FLATMEM_MANUAL=y
++# CONFIG_DISCONTIGMEM_MANUAL is not set
++# CONFIG_SPARSEMEM_MANUAL is not set
++CONFIG_FLATMEM=y
++CONFIG_FLAT_NODE_MEM_MAP=y
++# CONFIG_SPARSEMEM_STATIC is not set
++CONFIG_SPLIT_PTLOCK_CPUS=4
++CONFIG_RESOURCES_64BIT=y
++CONFIG_ARCH_SELECT_MEMORY_MODEL=y
++CONFIG_ARCH_DISCONTIGMEM_ENABLE=y
++CONFIG_ARCH_FLATMEM_ENABLE=y
++CONFIG_ARCH_SPARSEMEM_ENABLE=y
++CONFIG_VIRTUAL_MEM_MAP=y
++CONFIG_HOLES_IN_ZONE=y
++# CONFIG_IA32_SUPPORT is not set
++CONFIG_IA64_MCA_RECOVERY=y
++CONFIG_PERFMON=y
++CONFIG_IA64_PALINFO=y
++# CONFIG_KEXEC is not set
++# CONFIG_CRASH_DUMP is not set
++
++#
++# Firmware Drivers
++#
++CONFIG_EFI_VARS=y
++CONFIG_EFI_PCDP=y
++CONFIG_BINFMT_ELF=y
++CONFIG_BINFMT_MISC=y
++
++#
++# Power management and ACPI
++#
++CONFIG_PM=y
++CONFIG_PM_LEGACY=y
++# CONFIG_PM_DEBUG is not set
++
++#
++# ACPI (Advanced Configuration and Power Interface) Support
++#
++CONFIG_ACPI=y
++CONFIG_ACPI_BUTTON=y
++CONFIG_ACPI_FAN=y
++# CONFIG_ACPI_DOCK is not set
++CONFIG_ACPI_PROCESSOR=y
++CONFIG_ACPI_HOTPLUG_CPU=y
++CONFIG_ACPI_THERMAL=y
++CONFIG_ACPI_BLACKLIST_YEAR=0
++# CONFIG_ACPI_DEBUG is not set
++CONFIG_ACPI_EC=y
++CONFIG_ACPI_POWER=y
++CONFIG_ACPI_SYSTEM=y
++CONFIG_ACPI_CONTAINER=y
++
++#
++# CPU Frequency scaling
++#
++# CONFIG_CPU_FREQ is not set
++
++#
++# Bus options (PCI, PCMCIA)
++#
++CONFIG_PCI=y
++CONFIG_PCI_DOMAINS=y
++CONFIG_XEN_PCIDEV_FRONTEND=y
++# CONFIG_XEN_PCIDEV_FE_DEBUG is not set
++# CONFIG_PCIEPORTBUS is not set
++# CONFIG_PCI_DEBUG is not set
++
++#
++# PCI Hotplug Support
++#
++CONFIG_HOTPLUG_PCI=y
++# CONFIG_HOTPLUG_PCI_FAKE is not set
++CONFIG_HOTPLUG_PCI_ACPI=y
++# CONFIG_HOTPLUG_PCI_ACPI_IBM is not set
++# CONFIG_HOTPLUG_PCI_CPCI is not set
++# CONFIG_HOTPLUG_PCI_SHPC is not set
++
++#
++# PCCARD (PCMCIA/CardBus) support
++#
++# CONFIG_PCCARD is not set
++
++#
++# Networking
++#
++CONFIG_NET=y
++
++#
++# Networking options
++#
++# CONFIG_NETDEBUG is not set
++CONFIG_PACKET=y
++# CONFIG_PACKET_MMAP is not set
++CONFIG_UNIX=y
++CONFIG_XFRM=y
++# CONFIG_XFRM_USER is not set
++# CONFIG_NET_KEY is not set
++CONFIG_INET=y
++CONFIG_IP_MULTICAST=y
++# CONFIG_IP_ADVANCED_ROUTER is not set
++CONFIG_IP_FIB_HASH=y
++CONFIG_IP_PNP=y
++CONFIG_IP_PNP_DHCP=y
++# CONFIG_IP_PNP_BOOTP is not set
++# CONFIG_IP_PNP_RARP is not set
++# CONFIG_NET_IPIP is not set
++# CONFIG_NET_IPGRE is not set
++# CONFIG_IP_MROUTE is not set
++# CONFIG_ARPD is not set
++# CONFIG_SYN_COOKIES is not set
++# CONFIG_INET_AH is not set
++# CONFIG_INET_ESP is not set
++# CONFIG_INET_IPCOMP is not set
++# CONFIG_INET_XFRM_TUNNEL is not set
++# CONFIG_INET_TUNNEL is not set
++CONFIG_INET_XFRM_MODE_TRANSPORT=y
++CONFIG_INET_XFRM_MODE_TUNNEL=y
++CONFIG_INET_DIAG=y
++CONFIG_INET_TCP_DIAG=y
++# CONFIG_TCP_CONG_ADVANCED is not set
++CONFIG_TCP_CONG_BIC=y
++
++#
++# IP: Virtual Server Configuration
++#
++# CONFIG_IP_VS is not set
++# CONFIG_IPV6 is not set
++# CONFIG_INET6_XFRM_TUNNEL is not set
++# CONFIG_INET6_TUNNEL is not set
++# CONFIG_NETWORK_SECMARK is not set
++CONFIG_NETFILTER=y
++# CONFIG_NETFILTER_DEBUG is not set
++
++#
++# Core Netfilter Configuration
++#
++# CONFIG_NETFILTER_NETLINK is not set
++# CONFIG_NF_CONNTRACK is not set
++# CONFIG_NETFILTER_XTABLES is not set
++
++#
++# IP: Netfilter Configuration
++#
++# CONFIG_IP_NF_CONNTRACK is not set
++# CONFIG_IP_NF_QUEUE is not set
++
++#
++# DCCP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_DCCP is not set
++
++#
++# SCTP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_SCTP is not set
++
++#
++# TIPC Configuration (EXPERIMENTAL)
++#
++# CONFIG_TIPC is not set
++# CONFIG_ATM is not set
++# CONFIG_BRIDGE is not set
++# CONFIG_VLAN_8021Q is not set
++# CONFIG_DECNET is not set
++# CONFIG_LLC2 is not set
++# CONFIG_IPX is not set
++# CONFIG_ATALK is not set
++# CONFIG_X25 is not set
++# CONFIG_LAPB is not set
++# CONFIG_ECONET is not set
++# CONFIG_WAN_ROUTER is not set
++
++#
++# QoS and/or fair queueing
++#
++# CONFIG_NET_SCHED is not set
++
++#
++# Network testing
++#
++# CONFIG_NET_PKTGEN is not set
++# CONFIG_HAMRADIO is not set
++# CONFIG_IRDA is not set
++# CONFIG_BT is not set
++# CONFIG_IEEE80211 is not set
++
++#
++# Device Drivers
++#
++
++#
++# Generic Driver Options
++#
++CONFIG_STANDALONE=y
++CONFIG_PREVENT_FIRMWARE_BUILD=y
++# CONFIG_FW_LOADER is not set
++# CONFIG_DEBUG_DRIVER is not set
++CONFIG_SYS_HYPERVISOR=y
++
++#
++# Connector - unified userspace <-> kernelspace linker
++#
++# CONFIG_CONNECTOR is not set
++
++#
++# Memory Technology Devices (MTD)
++#
++# CONFIG_MTD is not set
++
++#
++# Parallel port support
++#
++# CONFIG_PARPORT is not set
++
++#
++# Plug and Play support
++#
++# CONFIG_PNP is not set
++
++#
++# Block devices
++#
++# CONFIG_BLK_CPQ_DA is not set
++# CONFIG_BLK_CPQ_CISS_DA is not set
++# CONFIG_BLK_DEV_DAC960 is not set
++# CONFIG_BLK_DEV_UMEM is not set
++# CONFIG_BLK_DEV_COW_COMMON is not set
++CONFIG_BLK_DEV_LOOP=y
++# CONFIG_BLK_DEV_CRYPTOLOOP is not set
++# CONFIG_BLK_DEV_NBD is not set
++# CONFIG_BLK_DEV_SX8 is not set
++# CONFIG_BLK_DEV_UB is not set
++CONFIG_BLK_DEV_RAM=y
++CONFIG_BLK_DEV_RAM_COUNT=16
++CONFIG_BLK_DEV_RAM_SIZE=4096
++CONFIG_BLK_DEV_RAM_BLOCKSIZE=1024
++CONFIG_BLK_DEV_INITRD=y
++# CONFIG_CDROM_PKTCDVD is not set
++# CONFIG_ATA_OVER_ETH is not set
++
++#
++# ATA/ATAPI/MFM/RLL support
++#
++# CONFIG_IDE is not set
++
++#
++# SCSI device support
++#
++# CONFIG_RAID_ATTRS is not set
++CONFIG_SCSI=y
++CONFIG_SCSI_PROC_FS=y
++
++#
++# SCSI support type (disk, tape, CD-ROM)
++#
++CONFIG_BLK_DEV_SD=y
++CONFIG_CHR_DEV_ST=y
++CONFIG_CHR_DEV_OSST=y
++CONFIG_BLK_DEV_SR=y
++CONFIG_BLK_DEV_SR_VENDOR=y
++CONFIG_CHR_DEV_SG=y
++# CONFIG_CHR_DEV_SCH is not set
++
++#
++# Some SCSI devices (e.g. CD jukebox) support multiple LUNs
++#
++CONFIG_SCSI_MULTI_LUN=y
++CONFIG_SCSI_CONSTANTS=y
++CONFIG_SCSI_LOGGING=y
++
++#
++# SCSI Transport Attributes
++#
++CONFIG_SCSI_SPI_ATTRS=y
++# CONFIG_SCSI_FC_ATTRS is not set
++# CONFIG_SCSI_ISCSI_ATTRS is not set
++CONFIG_SCSI_SAS_ATTRS=y
++
++#
++# SCSI low-level drivers
++#
++# CONFIG_ISCSI_TCP is not set
++# CONFIG_BLK_DEV_3W_XXXX_RAID is not set
++# CONFIG_SCSI_3W_9XXX is not set
++# CONFIG_SCSI_ACARD is not set
++# CONFIG_SCSI_AACRAID is not set
++# CONFIG_SCSI_AIC7XXX is not set
++# CONFIG_SCSI_AIC7XXX_OLD is not set
++# CONFIG_SCSI_AIC79XX is not set
++# CONFIG_MEGARAID_NEWGEN is not set
++# CONFIG_MEGARAID_LEGACY is not set
++# CONFIG_MEGARAID_SAS is not set
++# CONFIG_SCSI_SATA is not set
++# CONFIG_SCSI_HPTIOP is not set
++# CONFIG_SCSI_DMX3191D is not set
++# CONFIG_SCSI_FUTURE_DOMAIN is not set
++# CONFIG_SCSI_IPS is not set
++# CONFIG_SCSI_INITIO is not set
++# CONFIG_SCSI_INIA100 is not set
++CONFIG_SCSI_SYM53C8XX_2=y
++CONFIG_SCSI_SYM53C8XX_DMA_ADDRESSING_MODE=1
++CONFIG_SCSI_SYM53C8XX_DEFAULT_TAGS=16
++CONFIG_SCSI_SYM53C8XX_MAX_TAGS=64
++CONFIG_SCSI_SYM53C8XX_MMIO=y
++# CONFIG_SCSI_IPR is not set
++CONFIG_SCSI_QLOGIC_1280=y
++# CONFIG_SCSI_QLA_FC is not set
++# CONFIG_SCSI_LPFC is not set
++# CONFIG_SCSI_DC395x is not set
++# CONFIG_SCSI_DC390T is not set
++# CONFIG_SCSI_DEBUG is not set
++
++#
++# Multi-device support (RAID and LVM)
++#
++# CONFIG_MD is not set
++
++#
++# Fusion MPT device support
++#
++CONFIG_FUSION=y
++CONFIG_FUSION_SPI=y
++# CONFIG_FUSION_FC is not set
++CONFIG_FUSION_SAS=y
++CONFIG_FUSION_MAX_SGE=128
++# CONFIG_FUSION_CTL is not set
++
++#
++# IEEE 1394 (FireWire) support
++#
++# CONFIG_IEEE1394 is not set
++
++#
++# I2O device support
++#
++# CONFIG_I2O is not set
++
++#
++# Network device support
++#
++CONFIG_NETDEVICES=y
++CONFIG_DUMMY=y
++# CONFIG_BONDING is not set
++# CONFIG_EQUALIZER is not set
++# CONFIG_TUN is not set
++
++#
++# ARCnet devices
++#
++# CONFIG_ARCNET is not set
++
++#
++# PHY device support
++#
++# CONFIG_PHYLIB is not set
++
++#
++# Ethernet (10 or 100Mbit)
++#
++CONFIG_NET_ETHERNET=y
++CONFIG_MII=y
++# CONFIG_HAPPYMEAL is not set
++# CONFIG_SUNGEM is not set
++# CONFIG_CASSINI is not set
++# CONFIG_NET_VENDOR_3COM is not set
++
++#
++# Tulip family network device support
++#
++CONFIG_NET_TULIP=y
++# CONFIG_DE2104X is not set
++CONFIG_TULIP=y
++CONFIG_TULIP_MWI=y
++CONFIG_TULIP_MMIO=y
++CONFIG_TULIP_NAPI=y
++CONFIG_TULIP_NAPI_HW_MITIGATION=y
++# CONFIG_DE4X5 is not set
++# CONFIG_WINBOND_840 is not set
++# CONFIG_DM9102 is not set
++# CONFIG_ULI526X is not set
++# CONFIG_HP100 is not set
++CONFIG_NET_PCI=y
++# CONFIG_PCNET32 is not set
++# CONFIG_AMD8111_ETH is not set
++# CONFIG_ADAPTEC_STARFIRE is not set
++# CONFIG_B44 is not set
++# CONFIG_FORCEDETH is not set
++# CONFIG_DGRS is not set
++# CONFIG_EEPRO100 is not set
++CONFIG_E100=y
++# CONFIG_FEALNX is not set
++# CONFIG_NATSEMI is not set
++# CONFIG_NE2K_PCI is not set
++# CONFIG_8139CP is not set
++# CONFIG_8139TOO is not set
++# CONFIG_SIS900 is not set
++# CONFIG_EPIC100 is not set
++# CONFIG_SUNDANCE is not set
++# CONFIG_VIA_RHINE is not set
++
++#
++# Ethernet (1000 Mbit)
++#
++# CONFIG_ACENIC is not set
++# CONFIG_DL2K is not set
++CONFIG_E1000=y
++# CONFIG_E1000_NAPI is not set
++# CONFIG_E1000_DISABLE_PACKET_SPLIT is not set
++# CONFIG_NS83820 is not set
++# CONFIG_HAMACHI is not set
++# CONFIG_YELLOWFIN is not set
++# CONFIG_R8169 is not set
++# CONFIG_SIS190 is not set
++# CONFIG_SKGE is not set
++# CONFIG_SKY2 is not set
++# CONFIG_SK98LIN is not set
++# CONFIG_VIA_VELOCITY is not set
++CONFIG_TIGON3=y
++# CONFIG_BNX2 is not set
++
++#
++# Ethernet (10000 Mbit)
++#
++# CONFIG_CHELSIO_T1 is not set
++# CONFIG_IXGB is not set
++# CONFIG_S2IO is not set
++# CONFIG_MYRI10GE is not set
++# CONFIG_SFC is not set
++
++#
++# Token Ring devices
++#
++# CONFIG_TR is not set
++
++#
++# Wireless LAN (non-hamradio)
++#
++# CONFIG_NET_RADIO is not set
++
++#
++# Wan interfaces
++#
++# CONFIG_WAN is not set
++# CONFIG_FDDI is not set
++# CONFIG_HIPPI is not set
++# CONFIG_PPP is not set
++# CONFIG_SLIP is not set
++# CONFIG_NET_FC is not set
++# CONFIG_SHAPER is not set
++# CONFIG_NETCONSOLE is not set
++# CONFIG_NETPOLL is not set
++# CONFIG_NET_POLL_CONTROLLER is not set
++
++#
++# ISDN subsystem
++#
++# CONFIG_ISDN is not set
++
++#
++# Telephony Support
++#
++# CONFIG_PHONE is not set
++
++#
++# Input device support
++#
++CONFIG_INPUT=y
++
++#
++# Userland interfaces
++#
++CONFIG_INPUT_MOUSEDEV=y
++CONFIG_INPUT_MOUSEDEV_PSAUX=y
++CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024
++CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768
++CONFIG_INPUT_JOYDEV=y
++# CONFIG_INPUT_TSDEV is not set
++CONFIG_INPUT_EVDEV=y
++# CONFIG_INPUT_EVBUG is not set
++
++#
++# Input Device Drivers
++#
++# CONFIG_INPUT_KEYBOARD is not set
++# CONFIG_INPUT_MOUSE is not set
++# CONFIG_INPUT_JOYSTICK is not set
++# CONFIG_INPUT_TOUCHSCREEN is not set
++# CONFIG_INPUT_MISC is not set
++
++#
++# Hardware I/O ports
++#
++CONFIG_SERIO=y
++# CONFIG_SERIO_I8042 is not set
++# CONFIG_SERIO_SERPORT is not set
++# CONFIG_SERIO_PCIPS2 is not set
++# CONFIG_SERIO_RAW is not set
++# CONFIG_GAMEPORT is not set
++
++#
++# Character devices
++#
++CONFIG_VT=y
++CONFIG_VT_CONSOLE=y
++CONFIG_HW_CONSOLE=y
++# CONFIG_VT_HW_CONSOLE_BINDING is not set
++# CONFIG_SERIAL_NONSTANDARD is not set
++
++#
++# Serial drivers
++#
++CONFIG_SERIAL_8250=y
++CONFIG_SERIAL_8250_CONSOLE=y
++CONFIG_SERIAL_8250_PCI=y
++CONFIG_SERIAL_8250_NR_UARTS=8
++CONFIG_SERIAL_8250_RUNTIME_UARTS=4
++CONFIG_SERIAL_8250_EXTENDED=y
++CONFIG_SERIAL_8250_SHARE_IRQ=y
++# CONFIG_SERIAL_8250_DETECT_IRQ is not set
++# CONFIG_SERIAL_8250_RSA is not set
++
++#
++# Non-8250 serial port support
++#
++CONFIG_SERIAL_CORE=y
++CONFIG_SERIAL_CORE_CONSOLE=y
++# CONFIG_SERIAL_JSM is not set
++CONFIG_UNIX98_PTYS=y
++CONFIG_LEGACY_PTYS=y
++CONFIG_LEGACY_PTY_COUNT=256
++
++#
++# IPMI
++#
++# CONFIG_IPMI_HANDLER is not set
++
++#
++# Watchdog Cards
++#
++# CONFIG_WATCHDOG is not set
++# CONFIG_HW_RANDOM is not set
++CONFIG_EFI_RTC=y
++# CONFIG_DTLK is not set
++# CONFIG_R3964 is not set
++# CONFIG_APPLICOM is not set
++
++#
++# Ftape, the floppy tape device driver
++#
++CONFIG_AGP=y
++# CONFIG_AGP_SIS is not set
++# CONFIG_AGP_VIA is not set
++CONFIG_DRM=y
++# CONFIG_DRM_TDFX is not set
++# CONFIG_DRM_R128 is not set
++CONFIG_DRM_RADEON=y
++# CONFIG_DRM_MGA is not set
++# CONFIG_DRM_SIS is not set
++# CONFIG_DRM_VIA is not set
++# CONFIG_DRM_SAVAGE is not set
++# CONFIG_RAW_DRIVER is not set
++# CONFIG_HPET is not set
++# CONFIG_HANGCHECK_TIMER is not set
++
++#
++# TPM devices
++#
++# CONFIG_TCG_TPM is not set
++# CONFIG_TELCLOCK is not set
++
++#
++# I2C support
++#
++CONFIG_I2C=y
++CONFIG_I2C_CHARDEV=y
++
++#
++# I2C Algorithms
++#
++CONFIG_I2C_ALGOBIT=y
++CONFIG_I2C_ALGOPCF=y
++# CONFIG_I2C_ALGOPCA is not set
++
++#
++# I2C Hardware Bus support
++#
++# CONFIG_I2C_ALI1535 is not set
++# CONFIG_I2C_ALI1563 is not set
++# CONFIG_I2C_ALI15X3 is not set
++# CONFIG_I2C_AMD756 is not set
++# CONFIG_I2C_AMD8111 is not set
++# CONFIG_I2C_I801 is not set
++# CONFIG_I2C_I810 is not set
++# CONFIG_I2C_PIIX4 is not set
++# CONFIG_I2C_NFORCE2 is not set
++# CONFIG_I2C_OCORES is not set
++# CONFIG_I2C_PARPORT_LIGHT is not set
++# CONFIG_I2C_PROSAVAGE is not set
++# CONFIG_I2C_SAVAGE4 is not set
++# CONFIG_I2C_SIS5595 is not set
++# CONFIG_I2C_SIS630 is not set
++# CONFIG_I2C_SIS96X is not set
++# CONFIG_I2C_STUB is not set
++# CONFIG_I2C_VIA is not set
++# CONFIG_I2C_VIAPRO is not set
++# CONFIG_I2C_VOODOO3 is not set
++# CONFIG_I2C_PCA_ISA is not set
++
++#
++# Miscellaneous I2C Chip support
++#
++# CONFIG_SENSORS_DS1337 is not set
++# CONFIG_SENSORS_DS1374 is not set
++# CONFIG_SENSORS_EEPROM is not set
++# CONFIG_SENSORS_PCF8574 is not set
++# CONFIG_SENSORS_PCA9539 is not set
++# CONFIG_SENSORS_PCF8591 is not set
++# CONFIG_SENSORS_MAX6875 is not set
++# CONFIG_I2C_DEBUG_CORE is not set
++# CONFIG_I2C_DEBUG_ALGO is not set
++# CONFIG_I2C_DEBUG_BUS is not set
++# CONFIG_I2C_DEBUG_CHIP is not set
++
++#
++# SPI support
++#
++# CONFIG_SPI is not set
++# CONFIG_SPI_MASTER is not set
++
++#
++# Dallas's 1-wire bus
++#
++
++#
++# Hardware Monitoring support
++#
++CONFIG_HWMON=y
++# CONFIG_HWMON_VID is not set
++# CONFIG_SENSORS_ABITUGURU is not set
++# CONFIG_SENSORS_ADM1021 is not set
++# CONFIG_SENSORS_ADM1025 is not set
++# CONFIG_SENSORS_ADM1026 is not set
++# CONFIG_SENSORS_ADM1031 is not set
++# CONFIG_SENSORS_ADM9240 is not set
++# CONFIG_SENSORS_ASB100 is not set
++# CONFIG_SENSORS_ATXP1 is not set
++# CONFIG_SENSORS_DS1621 is not set
++# CONFIG_SENSORS_F71805F is not set
++# CONFIG_SENSORS_FSCHER is not set
++# CONFIG_SENSORS_FSCPOS is not set
++# CONFIG_SENSORS_GL518SM is not set
++# CONFIG_SENSORS_GL520SM is not set
++# CONFIG_SENSORS_IT87 is not set
++# CONFIG_SENSORS_LM63 is not set
++# CONFIG_SENSORS_LM75 is not set
++# CONFIG_SENSORS_LM77 is not set
++# CONFIG_SENSORS_LM78 is not set
++# CONFIG_SENSORS_LM80 is not set
++# CONFIG_SENSORS_LM83 is not set
++# CONFIG_SENSORS_LM85 is not set
++# CONFIG_SENSORS_LM87 is not set
++# CONFIG_SENSORS_LM90 is not set
++# CONFIG_SENSORS_LM92 is not set
++# CONFIG_SENSORS_MAX1619 is not set
++# CONFIG_SENSORS_PC87360 is not set
++# CONFIG_SENSORS_SIS5595 is not set
++# CONFIG_SENSORS_SMSC47M1 is not set
++# CONFIG_SENSORS_SMSC47M192 is not set
++# CONFIG_SENSORS_SMSC47B397 is not set
++# CONFIG_SENSORS_VIA686A is not set
++# CONFIG_SENSORS_VT8231 is not set
++# CONFIG_SENSORS_W83781D is not set
++# CONFIG_SENSORS_W83791D is not set
++# CONFIG_SENSORS_W83792D is not set
++# CONFIG_SENSORS_W83L785TS is not set
++# CONFIG_SENSORS_W83627HF is not set
++# CONFIG_SENSORS_W83627EHF is not set
++# CONFIG_HWMON_DEBUG_CHIP is not set
++
++#
++# Misc devices
++#
++
++#
++# Multimedia devices
++#
++CONFIG_VIDEO_DEV=y
++CONFIG_VIDEO_V4L1=y
++CONFIG_VIDEO_V4L1_COMPAT=y
++CONFIG_VIDEO_V4L2=y
++
++#
++# Video Capture Adapters
++#
++
++#
++# Video Capture Adapters
++#
++# CONFIG_VIDEO_ADV_DEBUG is not set
++# CONFIG_VIDEO_VIVI is not set
++# CONFIG_VIDEO_BT848 is not set
++# CONFIG_VIDEO_CPIA is not set
++# CONFIG_VIDEO_CPIA2 is not set
++# CONFIG_VIDEO_SAA5246A is not set
++# CONFIG_VIDEO_SAA5249 is not set
++# CONFIG_TUNER_3036 is not set
++# CONFIG_VIDEO_STRADIS is not set
++# CONFIG_VIDEO_ZORAN is not set
++# CONFIG_VIDEO_SAA7134 is not set
++# CONFIG_VIDEO_MXB is not set
++# CONFIG_VIDEO_DPC is not set
++# CONFIG_VIDEO_HEXIUM_ORION is not set
++# CONFIG_VIDEO_HEXIUM_GEMINI is not set
++# CONFIG_VIDEO_CX88 is not set
++
++#
++# Encoders and Decoders
++#
++# CONFIG_VIDEO_MSP3400 is not set
++# CONFIG_VIDEO_CS53L32A is not set
++# CONFIG_VIDEO_TLV320AIC23B is not set
++# CONFIG_VIDEO_WM8775 is not set
++# CONFIG_VIDEO_WM8739 is not set
++# CONFIG_VIDEO_CX2341X is not set
++# CONFIG_VIDEO_CX25840 is not set
++# CONFIG_VIDEO_SAA711X is not set
++# CONFIG_VIDEO_SAA7127 is not set
++# CONFIG_VIDEO_UPD64031A is not set
++# CONFIG_VIDEO_UPD64083 is not set
++
++#
++# V4L USB devices
++#
++# CONFIG_VIDEO_PVRUSB2 is not set
++# CONFIG_VIDEO_EM28XX is not set
++# CONFIG_USB_VICAM is not set
++# CONFIG_USB_IBMCAM is not set
++# CONFIG_USB_KONICAWC is not set
++# CONFIG_USB_QUICKCAM_MESSENGER is not set
++# CONFIG_USB_ET61X251 is not set
++# CONFIG_VIDEO_OVCAMCHIP is not set
++# CONFIG_USB_W9968CF is not set
++# CONFIG_USB_OV511 is not set
++# CONFIG_USB_SE401 is not set
++# CONFIG_USB_SN9C102 is not set
++# CONFIG_USB_STV680 is not set
++# CONFIG_USB_ZC0301 is not set
++# CONFIG_USB_PWC is not set
++
++#
++# Radio Adapters
++#
++# CONFIG_RADIO_GEMTEK_PCI is not set
++# CONFIG_RADIO_MAXIRADIO is not set
++# CONFIG_RADIO_MAESTRO is not set
++# CONFIG_USB_DSBR is not set
++
++#
++# Digital Video Broadcasting Devices
++#
++# CONFIG_DVB is not set
++# CONFIG_USB_DABUSB is not set
++
++#
++# Graphics support
++#
++CONFIG_FIRMWARE_EDID=y
++CONFIG_FB=y
++CONFIG_FB_CFB_FILLRECT=y
++CONFIG_FB_CFB_COPYAREA=y
++CONFIG_FB_CFB_IMAGEBLIT=y
++# CONFIG_FB_MACMODES is not set
++# CONFIG_FB_BACKLIGHT is not set
++CONFIG_FB_MODE_HELPERS=y
++# CONFIG_FB_TILEBLITTING is not set
++# CONFIG_FB_CIRRUS is not set
++# CONFIG_FB_PM2 is not set
++# CONFIG_FB_CYBER2000 is not set
++# CONFIG_FB_ASILIANT is not set
++# CONFIG_FB_IMSTT is not set
++# CONFIG_FB_S1D13XXX is not set
++# CONFIG_FB_NVIDIA is not set
++# CONFIG_FB_RIVA is not set
++# CONFIG_FB_MATROX is not set
++# CONFIG_FB_RADEON is not set
++# CONFIG_FB_ATY128 is not set
++# CONFIG_FB_ATY is not set
++# CONFIG_FB_SAVAGE is not set
++# CONFIG_FB_SIS is not set
++# CONFIG_FB_NEOMAGIC is not set
++# CONFIG_FB_KYRO is not set
++# CONFIG_FB_3DFX is not set
++# CONFIG_FB_VOODOO1 is not set
++# CONFIG_FB_TRIDENT is not set
++# CONFIG_FB_VIRTUAL is not set
++
++#
++# Console display driver support
++#
++CONFIG_VGA_CONSOLE=y
++# CONFIG_VGACON_SOFT_SCROLLBACK is not set
++CONFIG_DUMMY_CONSOLE=y
++CONFIG_FRAMEBUFFER_CONSOLE=y
++# CONFIG_FRAMEBUFFER_CONSOLE_ROTATION is not set
++# CONFIG_FONTS is not set
++CONFIG_FONT_8x8=y
++CONFIG_FONT_8x16=y
++
++#
++# Logo configuration
++#
++CONFIG_LOGO=y
++# CONFIG_LOGO_LINUX_MONO is not set
++# CONFIG_LOGO_LINUX_VGA16 is not set
++CONFIG_LOGO_LINUX_CLUT224=y
++# CONFIG_BACKLIGHT_LCD_SUPPORT is not set
++
++#
++# Sound
++#
++CONFIG_SOUND=y
++
++#
++# Advanced Linux Sound Architecture
++#
++CONFIG_SND=y
++CONFIG_SND_TIMER=y
++CONFIG_SND_PCM=y
++CONFIG_SND_HWDEP=y
++CONFIG_SND_RAWMIDI=y
++CONFIG_SND_SEQUENCER=y
++# CONFIG_SND_SEQ_DUMMY is not set
++CONFIG_SND_OSSEMUL=y
++CONFIG_SND_MIXER_OSS=y
++CONFIG_SND_PCM_OSS=y
++CONFIG_SND_PCM_OSS_PLUGINS=y
++CONFIG_SND_SEQUENCER_OSS=y
++# CONFIG_SND_DYNAMIC_MINORS is not set
++CONFIG_SND_SUPPORT_OLD_API=y
++CONFIG_SND_VERBOSE_PROCFS=y
++# CONFIG_SND_VERBOSE_PRINTK is not set
++# CONFIG_SND_DEBUG is not set
++
++#
++# Generic devices
++#
++CONFIG_SND_MPU401_UART=y
++CONFIG_SND_OPL3_LIB=y
++CONFIG_SND_AC97_CODEC=y
++CONFIG_SND_AC97_BUS=y
++# CONFIG_SND_DUMMY is not set
++# CONFIG_SND_VIRMIDI is not set
++# CONFIG_SND_MTPAV is not set
++# CONFIG_SND_SERIAL_U16550 is not set
++# CONFIG_SND_MPU401 is not set
++
++#
++# PCI devices
++#
++# CONFIG_SND_AD1889 is not set
++# CONFIG_SND_ALS300 is not set
++# CONFIG_SND_ALI5451 is not set
++# CONFIG_SND_ATIIXP is not set
++# CONFIG_SND_ATIIXP_MODEM is not set
++# CONFIG_SND_AU8810 is not set
++# CONFIG_SND_AU8820 is not set
++# CONFIG_SND_AU8830 is not set
++# CONFIG_SND_AZT3328 is not set
++# CONFIG_SND_BT87X is not set
++# CONFIG_SND_CA0106 is not set
++# CONFIG_SND_CMIPCI is not set
++# CONFIG_SND_CS4281 is not set
++# CONFIG_SND_CS46XX is not set
++# CONFIG_SND_EMU10K1 is not set
++# CONFIG_SND_EMU10K1X is not set
++# CONFIG_SND_ENS1370 is not set
++# CONFIG_SND_ENS1371 is not set
++# CONFIG_SND_ES1938 is not set
++# CONFIG_SND_ES1968 is not set
++CONFIG_SND_FM801=y
++# CONFIG_SND_FM801_TEA575X_BOOL is not set
++# CONFIG_SND_HDA_INTEL is not set
++# CONFIG_SND_HDSP is not set
++# CONFIG_SND_HDSPM is not set
++# CONFIG_SND_ICE1712 is not set
++# CONFIG_SND_ICE1724 is not set
++# CONFIG_SND_INTEL8X0 is not set
++# CONFIG_SND_INTEL8X0M is not set
++# CONFIG_SND_KORG1212 is not set
++# CONFIG_SND_MAESTRO3 is not set
++# CONFIG_SND_MIXART is not set
++# CONFIG_SND_NM256 is not set
++# CONFIG_SND_PCXHR is not set
++# CONFIG_SND_RME32 is not set
++# CONFIG_SND_RME96 is not set
++# CONFIG_SND_RME9652 is not set
++# CONFIG_SND_SONICVIBES is not set
++# CONFIG_SND_TRIDENT is not set
++# CONFIG_SND_VIA82XX is not set
++# CONFIG_SND_VIA82XX_MODEM is not set
++# CONFIG_SND_VX222 is not set
++# CONFIG_SND_YMFPCI is not set
++
++#
++# USB devices
++#
++# CONFIG_SND_USB_AUDIO is not set
++
++#
++# Open Sound System
++#
++# CONFIG_SOUND_PRIME is not set
++
++#
++# USB support
++#
++CONFIG_USB_ARCH_HAS_HCD=y
++CONFIG_USB_ARCH_HAS_OHCI=y
++CONFIG_USB_ARCH_HAS_EHCI=y
++CONFIG_USB=y
++# CONFIG_USB_DEBUG is not set
++
++#
++# Miscellaneous USB options
++#
++# CONFIG_USB_DEVICEFS is not set
++CONFIG_USB_BANDWIDTH=y
++# CONFIG_USB_DYNAMIC_MINORS is not set
++# CONFIG_USB_SUSPEND is not set
++# CONFIG_USB_OTG is not set
++
++#
++# USB Host Controller Drivers
++#
++CONFIG_USB_EHCI_HCD=y
++# CONFIG_USB_EHCI_SPLIT_ISO is not set
++# CONFIG_USB_EHCI_ROOT_HUB_TT is not set
++# CONFIG_USB_EHCI_TT_NEWSCHED is not set
++# CONFIG_USB_ISP116X_HCD is not set
++CONFIG_USB_OHCI_HCD=y
++# CONFIG_USB_OHCI_BIG_ENDIAN is not set
++CONFIG_USB_OHCI_LITTLE_ENDIAN=y
++CONFIG_USB_UHCI_HCD=y
++# CONFIG_USB_SL811_HCD is not set
++
++#
++# USB Device Class drivers
++#
++# CONFIG_USB_ACM is not set
++# CONFIG_USB_PRINTER is not set
++
++#
++# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support'
++#
++
++#
++# may also be needed; see USB_STORAGE Help for more information
++#
++CONFIG_USB_STORAGE=y
++# CONFIG_USB_STORAGE_DEBUG is not set
++# CONFIG_USB_STORAGE_DATAFAB is not set
++# CONFIG_USB_STORAGE_FREECOM is not set
++# CONFIG_USB_STORAGE_DPCM is not set
++# CONFIG_USB_STORAGE_USBAT is not set
++# CONFIG_USB_STORAGE_SDDR09 is not set
++# CONFIG_USB_STORAGE_SDDR55 is not set
++# CONFIG_USB_STORAGE_JUMPSHOT is not set
++# CONFIG_USB_STORAGE_ALAUDA is not set
++# CONFIG_USB_LIBUSUAL is not set
++
++#
++# USB Input Devices
++#
++CONFIG_USB_HID=y
++CONFIG_USB_HIDINPUT=y
++# CONFIG_USB_HIDINPUT_POWERBOOK is not set
++# CONFIG_HID_FF is not set
++CONFIG_USB_HIDDEV=y
++# CONFIG_USB_AIPTEK is not set
++# CONFIG_USB_WACOM is not set
++# CONFIG_USB_ACECAD is not set
++# CONFIG_USB_KBTAB is not set
++# CONFIG_USB_POWERMATE is not set
++# CONFIG_USB_TOUCHSCREEN is not set
++# CONFIG_USB_YEALINK is not set
++# CONFIG_USB_XPAD is not set
++# CONFIG_USB_ATI_REMOTE is not set
++# CONFIG_USB_ATI_REMOTE2 is not set
++# CONFIG_USB_KEYSPAN_REMOTE is not set
++# CONFIG_USB_APPLETOUCH is not set
++
++#
++# USB Imaging devices
++#
++# CONFIG_USB_MDC800 is not set
++# CONFIG_USB_MICROTEK is not set
++
++#
++# USB Network Adapters
++#
++# CONFIG_USB_CATC is not set
++# CONFIG_USB_KAWETH is not set
++# CONFIG_USB_PEGASUS is not set
++# CONFIG_USB_RTL8150 is not set
++# CONFIG_USB_USBNET is not set
++CONFIG_USB_MON=y
++
++#
++# USB port drivers
++#
++
++#
++# USB Serial Converter support
++#
++# CONFIG_USB_SERIAL is not set
++
++#
++# USB Miscellaneous drivers
++#
++# CONFIG_USB_EMI62 is not set
++# CONFIG_USB_EMI26 is not set
++# CONFIG_USB_AUERSWALD is not set
++# CONFIG_USB_RIO500 is not set
++# CONFIG_USB_LEGOTOWER is not set
++# CONFIG_USB_LCD is not set
++# CONFIG_USB_LED is not set
++# CONFIG_USB_CYPRESS_CY7C63 is not set
++# CONFIG_USB_CYTHERM is not set
++# CONFIG_USB_PHIDGETKIT is not set
++# CONFIG_USB_PHIDGETSERVO is not set
++# CONFIG_USB_IDMOUSE is not set
++# CONFIG_USB_APPLEDISPLAY is not set
++# CONFIG_USB_SISUSBVGA is not set
++# CONFIG_USB_LD is not set
++
++#
++# USB DSL modem support
++#
++
++#
++# USB Gadget Support
++#
++# CONFIG_USB_GADGET is not set
++
++#
++# MMC/SD Card support
++#
++# CONFIG_MMC is not set
++
++#
++# LED devices
++#
++# CONFIG_NEW_LEDS is not set
++
++#
++# LED drivers
++#
++
++#
++# LED Triggers
++#
++
++#
++# InfiniBand support
++#
++# CONFIG_INFINIBAND is not set
++
++#
++# EDAC - error detection and reporting (RAS) (EXPERIMENTAL)
++#
++
++#
++# Real Time Clock
++#
++# CONFIG_RTC_CLASS is not set
++
++#
++# DMA Engine support
++#
++# CONFIG_DMA_ENGINE is not set
++
++#
++# DMA Clients
++#
++
++#
++# DMA Devices
++#
++
++#
++# File systems
++#
++CONFIG_EXT2_FS=y
++CONFIG_EXT2_FS_XATTR=y
++# CONFIG_EXT2_FS_POSIX_ACL is not set
++# CONFIG_EXT2_FS_SECURITY is not set
++# CONFIG_EXT2_FS_XIP is not set
++CONFIG_EXT3_FS=y
++CONFIG_EXT3_FS_XATTR=y
++# CONFIG_EXT3_FS_POSIX_ACL is not set
++# CONFIG_EXT3_FS_SECURITY is not set
++CONFIG_JBD=y
++# CONFIG_JBD_DEBUG is not set
++CONFIG_FS_MBCACHE=y
++# CONFIG_REISERFS_FS is not set
++# CONFIG_JFS_FS is not set
++# CONFIG_FS_POSIX_ACL is not set
++# CONFIG_XFS_FS is not set
++# CONFIG_OCFS2_FS is not set
++# CONFIG_MINIX_FS is not set
++# CONFIG_ROMFS_FS is not set
++CONFIG_INOTIFY=y
++CONFIG_INOTIFY_USER=y
++# CONFIG_QUOTA is not set
++CONFIG_DNOTIFY=y
++CONFIG_AUTOFS_FS=y
++# CONFIG_AUTOFS4_FS is not set
++# CONFIG_FUSE_FS is not set
++
++#
++# CD-ROM/DVD Filesystems
++#
++CONFIG_ISO9660_FS=y
++CONFIG_JOLIET=y
++# CONFIG_ZISOFS is not set
++CONFIG_UDF_FS=y
++CONFIG_UDF_NLS=y
++
++#
++# DOS/FAT/NT Filesystems
++#
++CONFIG_FAT_FS=y
++CONFIG_MSDOS_FS=y
++CONFIG_VFAT_FS=y
++CONFIG_FAT_DEFAULT_CODEPAGE=437
++CONFIG_FAT_DEFAULT_IOCHARSET="iso8859-1"
++# CONFIG_NTFS_FS is not set
++
++#
++# Pseudo filesystems
++#
++CONFIG_PROC_FS=y
++CONFIG_PROC_KCORE=y
++CONFIG_SYSFS=y
++CONFIG_TMPFS=y
++# CONFIG_HUGETLB_PAGE is not set
++CONFIG_RAMFS=y
++# CONFIG_CONFIGFS_FS is not set
++
++#
++# Miscellaneous filesystems
++#
++# CONFIG_ADFS_FS is not set
++# CONFIG_AFFS_FS is not set
++# CONFIG_HFS_FS is not set
++# CONFIG_HFSPLUS_FS is not set
++# CONFIG_BEFS_FS is not set
++# CONFIG_BFS_FS is not set
++# CONFIG_EFS_FS is not set
++# CONFIG_CRAMFS is not set
++# CONFIG_VXFS_FS is not set
++# CONFIG_HPFS_FS is not set
++# CONFIG_QNX4FS_FS is not set
++# CONFIG_SYSV_FS is not set
++# CONFIG_UFS_FS is not set
++
++#
++# Network File Systems
++#
++CONFIG_NFS_FS=y
++CONFIG_NFS_V3=y
++# CONFIG_NFS_V3_ACL is not set
++CONFIG_NFS_V4=y
++# CONFIG_NFS_DIRECTIO is not set
++CONFIG_NFSD=y
++CONFIG_NFSD_V3=y
++# CONFIG_NFSD_V3_ACL is not set
++# CONFIG_NFSD_V4 is not set
++# CONFIG_NFSD_TCP is not set
++CONFIG_ROOT_NFS=y
++CONFIG_LOCKD=y
++CONFIG_LOCKD_V4=y
++CONFIG_EXPORTFS=y
++CONFIG_NFS_COMMON=y
++CONFIG_SUNRPC=y
++CONFIG_SUNRPC_GSS=y
++CONFIG_RPCSEC_GSS_KRB5=y
++# CONFIG_RPCSEC_GSS_SPKM3 is not set
++# CONFIG_SMB_FS is not set
++# CONFIG_CIFS is not set
++# CONFIG_NCP_FS is not set
++# CONFIG_CODA_FS is not set
++# CONFIG_AFS_FS is not set
++# CONFIG_9P_FS is not set
++
++#
++# Partition Types
++#
++CONFIG_PARTITION_ADVANCED=y
++# CONFIG_ACORN_PARTITION is not set
++# CONFIG_OSF_PARTITION is not set
++# CONFIG_AMIGA_PARTITION is not set
++# CONFIG_ATARI_PARTITION is not set
++# CONFIG_MAC_PARTITION is not set
++CONFIG_MSDOS_PARTITION=y
++# CONFIG_BSD_DISKLABEL is not set
++# CONFIG_MINIX_SUBPARTITION is not set
++# CONFIG_SOLARIS_X86_PARTITION is not set
++# CONFIG_UNIXWARE_DISKLABEL is not set
++# CONFIG_LDM_PARTITION is not set
++# CONFIG_SGI_PARTITION is not set
++# CONFIG_ULTRIX_PARTITION is not set
++# CONFIG_SUN_PARTITION is not set
++# CONFIG_KARMA_PARTITION is not set
++CONFIG_EFI_PARTITION=y
++
++#
++# Native Language Support
++#
++CONFIG_NLS=y
++CONFIG_NLS_DEFAULT="iso8859-1"
++CONFIG_NLS_CODEPAGE_437=y
++CONFIG_NLS_CODEPAGE_737=y
++CONFIG_NLS_CODEPAGE_775=y
++CONFIG_NLS_CODEPAGE_850=y
++CONFIG_NLS_CODEPAGE_852=y
++CONFIG_NLS_CODEPAGE_855=y
++CONFIG_NLS_CODEPAGE_857=y
++CONFIG_NLS_CODEPAGE_860=y
++CONFIG_NLS_CODEPAGE_861=y
++CONFIG_NLS_CODEPAGE_862=y
++CONFIG_NLS_CODEPAGE_863=y
++CONFIG_NLS_CODEPAGE_864=y
++CONFIG_NLS_CODEPAGE_865=y
++CONFIG_NLS_CODEPAGE_866=y
++CONFIG_NLS_CODEPAGE_869=y
++CONFIG_NLS_CODEPAGE_936=y
++CONFIG_NLS_CODEPAGE_950=y
++CONFIG_NLS_CODEPAGE_932=y
++CONFIG_NLS_CODEPAGE_949=y
++CONFIG_NLS_CODEPAGE_874=y
++CONFIG_NLS_ISO8859_8=y
++# CONFIG_NLS_CODEPAGE_1250 is not set
++CONFIG_NLS_CODEPAGE_1251=y
++# CONFIG_NLS_ASCII is not set
++CONFIG_NLS_ISO8859_1=y
++CONFIG_NLS_ISO8859_2=y
++CONFIG_NLS_ISO8859_3=y
++CONFIG_NLS_ISO8859_4=y
++CONFIG_NLS_ISO8859_5=y
++CONFIG_NLS_ISO8859_6=y
++CONFIG_NLS_ISO8859_7=y
++CONFIG_NLS_ISO8859_9=y
++CONFIG_NLS_ISO8859_13=y
++CONFIG_NLS_ISO8859_14=y
++CONFIG_NLS_ISO8859_15=y
++CONFIG_NLS_KOI8_R=y
++CONFIG_NLS_KOI8_U=y
++CONFIG_NLS_UTF8=y
++
++#
++# Library routines
++#
++# CONFIG_CRC_CCITT is not set
++# CONFIG_CRC16 is not set
++CONFIG_CRC32=y
++# CONFIG_LIBCRC32C is not set
++CONFIG_PLIST=y
++CONFIG_GENERIC_HARDIRQS=y
++CONFIG_GENERIC_IRQ_PROBE=y
++CONFIG_GENERIC_PENDING_IRQ=y
++CONFIG_IRQ_PER_CPU=y
++
++#
++# Instrumentation Support
++#
++# CONFIG_PROFILING is not set
++# CONFIG_KPROBES is not set
++
++#
++# Kernel hacking
++#
++# CONFIG_PRINTK_TIME is not set
++CONFIG_MAGIC_SYSRQ=y
++CONFIG_UNUSED_SYMBOLS=y
++CONFIG_DEBUG_KERNEL=y
++CONFIG_LOG_BUF_SHIFT=17
++CONFIG_DETECT_SOFTLOCKUP=y
++# CONFIG_SCHEDSTATS is not set
++# CONFIG_DEBUG_SLAB is not set
++# CONFIG_DEBUG_RT_MUTEXES is not set
++# CONFIG_RT_MUTEX_TESTER is not set
++# CONFIG_DEBUG_SPINLOCK is not set
++CONFIG_DEBUG_MUTEXES=y
++# CONFIG_DEBUG_RWSEMS is not set
++# CONFIG_DEBUG_SPINLOCK_SLEEP is not set
++# CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set
++# CONFIG_DEBUG_KOBJECT is not set
++# CONFIG_DEBUG_INFO is not set
++# CONFIG_DEBUG_FS is not set
++# CONFIG_DEBUG_VM is not set
++CONFIG_FORCED_INLINING=y
++# CONFIG_RCU_TORTURE_TEST is not set
++CONFIG_IA64_GRANULE_16MB=y
++# CONFIG_IA64_GRANULE_64MB is not set
++CONFIG_IA64_PRINT_HAZARDS=y
++# CONFIG_DISABLE_VHPT is not set
++# CONFIG_IA64_DEBUG_CMPXCHG is not set
++# CONFIG_IA64_DEBUG_IRQ is not set
++
++#
++# Security options
++#
++# CONFIG_KEYS is not set
++# CONFIG_SECURITY is not set
++
++#
++# Cryptographic options
++#
++CONFIG_CRYPTO=y
++# CONFIG_CRYPTO_HMAC is not set
++# CONFIG_CRYPTO_NULL is not set
++# CONFIG_CRYPTO_MD4 is not set
++CONFIG_CRYPTO_MD5=y
++# CONFIG_CRYPTO_SHA1 is not set
++# CONFIG_CRYPTO_SHA256 is not set
++# CONFIG_CRYPTO_SHA512 is not set
++# CONFIG_CRYPTO_WP512 is not set
++# CONFIG_CRYPTO_TGR192 is not set
++CONFIG_CRYPTO_DES=y
++# CONFIG_CRYPTO_BLOWFISH is not set
++# CONFIG_CRYPTO_TWOFISH is not set
++# CONFIG_CRYPTO_SERPENT is not set
++# CONFIG_CRYPTO_AES is not set
++# CONFIG_CRYPTO_CAST5 is not set
++# CONFIG_CRYPTO_CAST6 is not set
++# CONFIG_CRYPTO_TEA is not set
++# CONFIG_CRYPTO_ARC4 is not set
++# CONFIG_CRYPTO_KHAZAD is not set
++# CONFIG_CRYPTO_ANUBIS is not set
++# CONFIG_CRYPTO_DEFLATE is not set
++# CONFIG_CRYPTO_MICHAEL_MIC is not set
++# CONFIG_CRYPTO_CRC32C is not set
++# CONFIG_CRYPTO_TEST is not set
++
++#
++# Hardware crypto devices
++#
++# CONFIG_XEN_SMPBOOT is not set
++# CONFIG_XEN_DEVMEM is not set
++CONFIG_XEN_INTERFACE_VERSION=0x00030207
++
++#
++# XEN
++#
++# CONFIG_XEN_PRIVILEGED_GUEST is not set
++CONFIG_XEN_UNPRIVILEGED_GUEST=y
++CONFIG_XEN_PRIVCMD=y
++CONFIG_XEN_XENBUS_DEV=y
++# CONFIG_XEN_BACKEND is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_UTIL is not set
++CONFIG_XEN_BLKDEV_FRONTEND=y
++CONFIG_XEN_NETDEV_FRONTEND=y
++CONFIG_XEN_SCSI_FRONTEND=m
++# CONFIG_XEN_GRANT_DEV is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_FRONTEND is not set
++CONFIG_XEN_FRAMEBUFFER=y
++CONFIG_XEN_KEYBOARD=y
++# CONFIG_XEN_SCRUB_PAGES is not set
++# CONFIG_XEN_DISABLE_SERIAL is not set
++CONFIG_XEN_SYSFS=y
++CONFIG_XEN_COMPAT_030002_AND_LATER=y
++# CONFIG_XEN_COMPAT_030004_AND_LATER is not set
++# CONFIG_XEN_COMPAT_030100_AND_LATER is not set
++# CONFIG_XEN_COMPAT_LATEST_ONLY is not set
++CONFIG_XEN_COMPAT=0x030002
++CONFIG_HAVE_IRQ_IGNORE_UNHANDLED=y
++CONFIG_NO_IDLE_HZ=y
++CONFIG_XEN_BALLOON=y
++CONFIG_XEN_XENCOMM=y
+diff -r d894e36cfc30 -r 0aa021803deb buildconfigs/linux-defconfig_xenU_x86_32
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/buildconfigs/linux-defconfig_xenU_x86_32 Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,953 @@
++#
++# Automatically generated make config: don't edit
++# Linux kernel version: 2.6.18.8
++# Tue Oct 16 09:31:29 2007
++#
++CONFIG_X86_32=y
++CONFIG_LOCKDEP_SUPPORT=y
++CONFIG_STACKTRACE_SUPPORT=y
++CONFIG_SEMAPHORE_SLEEPERS=y
++CONFIG_X86=y
++CONFIG_MMU=y
++CONFIG_GENERIC_ISA_DMA=y
++CONFIG_GENERIC_IOMAP=y
++CONFIG_GENERIC_HWEIGHT=y
++CONFIG_ARCH_MAY_HAVE_PC_FDC=y
++CONFIG_DMI=y
++CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config"
++
++#
++# Code maturity level options
++#
++CONFIG_EXPERIMENTAL=y
++CONFIG_LOCK_KERNEL=y
++CONFIG_INIT_ENV_ARG_LIMIT=32
++
++#
++# General setup
++#
++CONFIG_LOCALVERSION=""
++# CONFIG_LOCALVERSION_AUTO is not set
++CONFIG_SWAP=y
++CONFIG_SYSVIPC=y
++# CONFIG_POSIX_MQUEUE is not set
++# CONFIG_BSD_PROCESS_ACCT is not set
++# CONFIG_TASKSTATS is not set
++# CONFIG_AUDIT is not set
++# CONFIG_IKCONFIG is not set
++# CONFIG_CPUSETS is not set
++# CONFIG_RELAY is not set
++CONFIG_INITRAMFS_SOURCE=""
++# CONFIG_CC_OPTIMIZE_FOR_SIZE is not set
++# CONFIG_EMBEDDED is not set
++CONFIG_UID16=y
++CONFIG_SYSCTL=y
++CONFIG_KALLSYMS=y
++# CONFIG_KALLSYMS_ALL is not set
++# CONFIG_KALLSYMS_EXTRA_PASS is not set
++CONFIG_HOTPLUG=y
++CONFIG_PRINTK=y
++CONFIG_BUG=y
++CONFIG_ELF_CORE=y
++CONFIG_BASE_FULL=y
++CONFIG_FUTEX=y
++CONFIG_EPOLL=y
++CONFIG_SHMEM=y
++CONFIG_SLAB=y
++CONFIG_VM_EVENT_COUNTERS=y
++CONFIG_RT_MUTEXES=y
++# CONFIG_TINY_SHMEM is not set
++CONFIG_BASE_SMALL=0
++# CONFIG_SLOB is not set
++
++#
++# Loadable module support
++#
++CONFIG_MODULES=y
++CONFIG_MODULE_UNLOAD=y
++# CONFIG_MODULE_FORCE_UNLOAD is not set
++# CONFIG_MODVERSIONS is not set
++# CONFIG_MODULE_SRCVERSION_ALL is not set
++CONFIG_KMOD=y
++CONFIG_STOP_MACHINE=y
++
++#
++# Block layer
++#
++# CONFIG_LBD is not set
++# CONFIG_BLK_DEV_IO_TRACE is not set
++CONFIG_LSF=y
++
++#
++# IO Schedulers
++#
++CONFIG_IOSCHED_NOOP=y
++CONFIG_IOSCHED_AS=y
++CONFIG_IOSCHED_DEADLINE=y
++CONFIG_IOSCHED_CFQ=y
++CONFIG_DEFAULT_AS=y
++# CONFIG_DEFAULT_DEADLINE is not set
++# CONFIG_DEFAULT_CFQ is not set
++# CONFIG_DEFAULT_NOOP is not set
++CONFIG_DEFAULT_IOSCHED="anticipatory"
++
++#
++# Processor type and features
++#
++CONFIG_SMP=y
++# CONFIG_X86_PC is not set
++CONFIG_X86_XEN=y
++# CONFIG_X86_ELAN is not set
++# CONFIG_X86_VOYAGER is not set
++# CONFIG_X86_NUMAQ is not set
++# CONFIG_X86_SUMMIT is not set
++# CONFIG_X86_BIGSMP is not set
++# CONFIG_X86_VISWS is not set
++# CONFIG_X86_GENERICARCH is not set
++# CONFIG_X86_ES7000 is not set
++# CONFIG_M386 is not set
++# CONFIG_M486 is not set
++# CONFIG_M586 is not set
++# CONFIG_M586TSC is not set
++# CONFIG_M586MMX is not set
++CONFIG_M686=y
++# CONFIG_MPENTIUMII is not set
++# CONFIG_MPENTIUMIII is not set
++# CONFIG_MPENTIUMM is not set
++# CONFIG_MPENTIUM4 is not set
++# CONFIG_MK6 is not set
++# CONFIG_MK7 is not set
++# CONFIG_MK8 is not set
++# CONFIG_MCRUSOE is not set
++# CONFIG_MEFFICEON is not set
++# CONFIG_MWINCHIPC6 is not set
++# CONFIG_MWINCHIP2 is not set
++# CONFIG_MWINCHIP3D is not set
++# CONFIG_MGEODEGX1 is not set
++# CONFIG_MGEODE_LX is not set
++# CONFIG_MCYRIXIII is not set
++# CONFIG_MVIAC3_2 is not set
++# CONFIG_X86_GENERIC is not set
++CONFIG_X86_CMPXCHG=y
++CONFIG_X86_XADD=y
++CONFIG_X86_L1_CACHE_SHIFT=5
++CONFIG_RWSEM_XCHGADD_ALGORITHM=y
++CONFIG_GENERIC_CALIBRATE_DELAY=y
++CONFIG_X86_PPRO_FENCE=y
++CONFIG_X86_WP_WORKS_OK=y
++CONFIG_X86_INVLPG=y
++CONFIG_X86_BSWAP=y
++CONFIG_X86_POPAD_OK=y
++CONFIG_X86_CMPXCHG64=y
++CONFIG_X86_GOOD_APIC=y
++CONFIG_X86_USE_PPRO_CHECKSUM=y
++CONFIG_NR_CPUS=8
++CONFIG_PREEMPT_NONE=y
++# CONFIG_PREEMPT_VOLUNTARY is not set
++# CONFIG_PREEMPT is not set
++CONFIG_PREEMPT_BKL=y
++CONFIG_VM86=y
++# CONFIG_TOSHIBA is not set
++# CONFIG_I8K is not set
++# CONFIG_X86_REBOOTFIXUPS is not set
++# CONFIG_X86_MSR is not set
++CONFIG_X86_CPUID=y
++CONFIG_SWIOTLB=y
++
++#
++# Firmware Drivers
++#
++# CONFIG_EDD is not set
++# CONFIG_DELL_RBU is not set
++# CONFIG_DCDBAS is not set
++# CONFIG_NOHIGHMEM is not set
++CONFIG_HIGHMEM4G=y
++# CONFIG_HIGHMEM64G is not set
++CONFIG_PAGE_OFFSET=0xC0000000
++CONFIG_HIGHMEM=y
++CONFIG_SELECT_MEMORY_MODEL=y
++CONFIG_FLATMEM_MANUAL=y
++# CONFIG_DISCONTIGMEM_MANUAL is not set
++# CONFIG_SPARSEMEM_MANUAL is not set
++CONFIG_FLATMEM=y
++CONFIG_FLAT_NODE_MEM_MAP=y
++# CONFIG_SPARSEMEM_STATIC is not set
++CONFIG_SPLIT_PTLOCK_CPUS=4
++CONFIG_RESOURCES_64BIT=y
++# CONFIG_HIGHPTE is not set
++# CONFIG_REGPARM is not set
++CONFIG_SECCOMP=y
++CONFIG_HZ_100=y
++# CONFIG_HZ_250 is not set
++# CONFIG_HZ_1000 is not set
++CONFIG_HZ=100
++# CONFIG_CRASH_DUMP is not set
++CONFIG_PHYSICAL_START=0x100000
++CONFIG_HOTPLUG_CPU=y
++CONFIG_ARCH_ENABLE_MEMORY_HOTPLUG=y
++
++#
++# Bus options (PCI, PCMCIA, EISA, MCA, ISA)
++#
++# CONFIG_PCI is not set
++CONFIG_ISA_DMA_API=y
++# CONFIG_SCx200 is not set
++
++#
++# PCCARD (PCMCIA/CardBus) support
++#
++# CONFIG_PCCARD is not set
++
++#
++# PCI Hotplug Support
++#
++
++#
++# Executable file formats
++#
++CONFIG_BINFMT_ELF=y
++# CONFIG_BINFMT_AOUT is not set
++# CONFIG_BINFMT_MISC is not set
++
++#
++# Networking
++#
++CONFIG_NET=y
++
++#
++# Networking options
++#
++# CONFIG_NETDEBUG is not set
++CONFIG_PACKET=y
++# CONFIG_PACKET_MMAP is not set
++CONFIG_UNIX=y
++CONFIG_XFRM=y
++# CONFIG_XFRM_USER is not set
++# CONFIG_NET_KEY is not set
++CONFIG_INET=y
++# CONFIG_IP_MULTICAST is not set
++# CONFIG_IP_ADVANCED_ROUTER is not set
++CONFIG_IP_FIB_HASH=y
++CONFIG_IP_PNP=y
++# CONFIG_IP_PNP_DHCP is not set
++# CONFIG_IP_PNP_BOOTP is not set
++# CONFIG_IP_PNP_RARP is not set
++# CONFIG_NET_IPIP is not set
++# CONFIG_NET_IPGRE is not set
++# CONFIG_ARPD is not set
++# CONFIG_SYN_COOKIES is not set
++# CONFIG_INET_AH is not set
++# CONFIG_INET_ESP is not set
++# CONFIG_INET_IPCOMP is not set
++# CONFIG_INET_XFRM_TUNNEL is not set
++# CONFIG_INET_TUNNEL is not set
++CONFIG_INET_XFRM_MODE_TRANSPORT=y
++CONFIG_INET_XFRM_MODE_TUNNEL=y
++# CONFIG_INET_DIAG is not set
++# CONFIG_TCP_CONG_ADVANCED is not set
++CONFIG_TCP_CONG_BIC=y
++# CONFIG_IPV6 is not set
++# CONFIG_INET6_XFRM_TUNNEL is not set
++# CONFIG_INET6_TUNNEL is not set
++# CONFIG_NETWORK_SECMARK is not set
++# CONFIG_NETFILTER is not set
++
++#
++# DCCP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_DCCP is not set
++
++#
++# SCTP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_SCTP is not set
++
++#
++# TIPC Configuration (EXPERIMENTAL)
++#
++# CONFIG_TIPC is not set
++# CONFIG_ATM is not set
++# CONFIG_BRIDGE is not set
++# CONFIG_VLAN_8021Q is not set
++# CONFIG_DECNET is not set
++# CONFIG_LLC2 is not set
++# CONFIG_IPX is not set
++# CONFIG_ATALK is not set
++# CONFIG_X25 is not set
++# CONFIG_LAPB is not set
++# CONFIG_ECONET is not set
++# CONFIG_WAN_ROUTER is not set
++
++#
++# QoS and/or fair queueing
++#
++# CONFIG_NET_SCHED is not set
++
++#
++# Network testing
++#
++# CONFIG_NET_PKTGEN is not set
++# CONFIG_HAMRADIO is not set
++# CONFIG_IRDA is not set
++# CONFIG_BT is not set
++# CONFIG_IEEE80211 is not set
++
++#
++# Device Drivers
++#
++
++#
++# Generic Driver Options
++#
++CONFIG_STANDALONE=y
++CONFIG_PREVENT_FIRMWARE_BUILD=y
++# CONFIG_FW_LOADER is not set
++# CONFIG_DEBUG_DRIVER is not set
++CONFIG_SYS_HYPERVISOR=y
++
++#
++# Connector - unified userspace <-> kernelspace linker
++#
++# CONFIG_CONNECTOR is not set
++
++#
++# Memory Technology Devices (MTD)
++#
++# CONFIG_MTD is not set
++
++#
++# Parallel port support
++#
++# CONFIG_PARPORT is not set
++
++#
++# Plug and Play support
++#
++
++#
++# Block devices
++#
++# CONFIG_BLK_DEV_FD is not set
++# CONFIG_BLK_DEV_COW_COMMON is not set
++CONFIG_BLK_DEV_LOOP=m
++# CONFIG_BLK_DEV_CRYPTOLOOP is not set
++CONFIG_BLK_DEV_NBD=m
++CONFIG_BLK_DEV_RAM=y
++CONFIG_BLK_DEV_RAM_COUNT=16
++CONFIG_BLK_DEV_RAM_SIZE=4096
++CONFIG_BLK_DEV_RAM_BLOCKSIZE=1024
++CONFIG_BLK_DEV_INITRD=y
++# CONFIG_CDROM_PKTCDVD is not set
++# CONFIG_ATA_OVER_ETH is not set
++
++#
++# ATA/ATAPI/MFM/RLL support
++#
++# CONFIG_IDE is not set
++
++#
++# SCSI device support
++#
++# CONFIG_RAID_ATTRS is not set
++CONFIG_SCSI=m
++CONFIG_SCSI_PROC_FS=y
++
++#
++# SCSI support type (disk, tape, CD-ROM)
++#
++CONFIG_BLK_DEV_SD=m
++# CONFIG_CHR_DEV_ST is not set
++# CONFIG_CHR_DEV_OSST is not set
++# CONFIG_BLK_DEV_SR is not set
++# CONFIG_CHR_DEV_SG is not set
++# CONFIG_CHR_DEV_SCH is not set
++
++#
++# Some SCSI devices (e.g. CD jukebox) support multiple LUNs
++#
++# CONFIG_SCSI_MULTI_LUN is not set
++# CONFIG_SCSI_CONSTANTS is not set
++# CONFIG_SCSI_LOGGING is not set
++
++#
++# SCSI Transport Attributes
++#
++# CONFIG_SCSI_SPI_ATTRS is not set
++# CONFIG_SCSI_FC_ATTRS is not set
++# CONFIG_SCSI_ISCSI_ATTRS is not set
++# CONFIG_SCSI_SAS_ATTRS is not set
++
++#
++# SCSI low-level drivers
++#
++# CONFIG_ISCSI_TCP is not set
++# CONFIG_SCSI_SATA is not set
++# CONFIG_SCSI_DEBUG is not set
++
++#
++# Multi-device support (RAID and LVM)
++#
++# CONFIG_MD is not set
++
++#
++# Fusion MPT device support
++#
++# CONFIG_FUSION is not set
++
++#
++# IEEE 1394 (FireWire) support
++#
++
++#
++# I2O device support
++#
++
++#
++# Network device support
++#
++CONFIG_NETDEVICES=y
++# CONFIG_DUMMY is not set
++# CONFIG_BONDING is not set
++# CONFIG_EQUALIZER is not set
++# CONFIG_TUN is not set
++
++#
++# PHY device support
++#
++
++#
++# Ethernet (10 or 100Mbit)
++#
++# CONFIG_NET_ETHERNET is not set
++
++#
++# Ethernet (1000 Mbit)
++#
++
++#
++# Ethernet (10000 Mbit)
++#
++
++#
++# Token Ring devices
++#
++
++#
++# Wireless LAN (non-hamradio)
++#
++# CONFIG_NET_RADIO is not set
++
++#
++# Wan interfaces
++#
++# CONFIG_WAN is not set
++# CONFIG_PPP is not set
++# CONFIG_SLIP is not set
++# CONFIG_SHAPER is not set
++# CONFIG_NETCONSOLE is not set
++# CONFIG_NETPOLL is not set
++# CONFIG_NET_POLL_CONTROLLER is not set
++
++#
++# ISDN subsystem
++#
++# CONFIG_ISDN is not set
++
++#
++# Telephony Support
++#
++# CONFIG_PHONE is not set
++
++#
++# Input device support
++#
++CONFIG_INPUT=y
++
++#
++# Userland interfaces
++#
++CONFIG_INPUT_MOUSEDEV=y
++# CONFIG_INPUT_MOUSEDEV_PSAUX is not set
++CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024
++CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768
++# CONFIG_INPUT_JOYDEV is not set
++# CONFIG_INPUT_TSDEV is not set
++# CONFIG_INPUT_EVDEV is not set
++# CONFIG_INPUT_EVBUG is not set
++
++#
++# Input Device Drivers
++#
++CONFIG_INPUT_KEYBOARD=y
++# CONFIG_KEYBOARD_ATKBD is not set
++# CONFIG_KEYBOARD_SUNKBD is not set
++# CONFIG_KEYBOARD_LKKBD is not set
++# CONFIG_KEYBOARD_XTKBD is not set
++# CONFIG_KEYBOARD_NEWTON is not set
++# CONFIG_INPUT_MOUSE is not set
++# CONFIG_INPUT_JOYSTICK is not set
++# CONFIG_INPUT_TOUCHSCREEN is not set
++# CONFIG_INPUT_MISC is not set
++
++#
++# Hardware I/O ports
++#
++CONFIG_SERIO=y
++CONFIG_SERIO_I8042=y
++CONFIG_SERIO_SERPORT=y
++# CONFIG_SERIO_CT82C710 is not set
++# CONFIG_SERIO_RAW is not set
++# CONFIG_GAMEPORT is not set
++
++#
++# Character devices
++#
++CONFIG_VT=y
++CONFIG_VT_CONSOLE=y
++CONFIG_HW_CONSOLE=y
++# CONFIG_VT_HW_CONSOLE_BINDING is not set
++# CONFIG_SERIAL_NONSTANDARD is not set
++
++#
++# Serial drivers
++#
++
++#
++# Non-8250 serial port support
++#
++CONFIG_UNIX98_PTYS=y
++CONFIG_LEGACY_PTYS=y
++CONFIG_LEGACY_PTY_COUNT=256
++
++#
++# IPMI
++#
++# CONFIG_IPMI_HANDLER is not set
++
++#
++# Watchdog Cards
++#
++# CONFIG_WATCHDOG is not set
++CONFIG_HW_RANDOM=y
++CONFIG_HW_RANDOM_VIA=y
++# CONFIG_NVRAM is not set
++# CONFIG_RTC is not set
++# CONFIG_GEN_RTC is not set
++# CONFIG_DTLK is not set
++# CONFIG_R3964 is not set
++
++#
++# Ftape, the floppy tape device driver
++#
++# CONFIG_MWAVE is not set
++# CONFIG_PC8736x_GPIO is not set
++# CONFIG_NSC_GPIO is not set
++# CONFIG_CS5535_GPIO is not set
++# CONFIG_RAW_DRIVER is not set
++# CONFIG_HANGCHECK_TIMER is not set
++
++#
++# TPM devices
++#
++# CONFIG_TCG_TPM is not set
++# CONFIG_TELCLOCK is not set
++
++#
++# I2C support
++#
++# CONFIG_I2C is not set
++
++#
++# SPI support
++#
++# CONFIG_SPI is not set
++# CONFIG_SPI_MASTER is not set
++
++#
++# Dallas's 1-wire bus
++#
++
++#
++# Hardware Monitoring support
++#
++# CONFIG_HWMON is not set
++# CONFIG_HWMON_VID is not set
++
++#
++# Misc devices
++#
++
++#
++# Multimedia devices
++#
++# CONFIG_VIDEO_DEV is not set
++
++#
++# Digital Video Broadcasting Devices
++#
++# CONFIG_DVB is not set
++
++#
++# Graphics support
++#
++CONFIG_FIRMWARE_EDID=y
++# CONFIG_FB is not set
+
-+ifdef CONFIG_XEN
-+AFLAGS_vsyscall-int80.o = -m32 -Wa,-32 -Iarch/i386/kernel
-+CFLAGS_syscall32-xen.o += -DUSE_INT80
-+AFLAGS_syscall32_syscall-xen.o += -DUSE_INT80
++#
++# Console display driver support
++#
++CONFIG_VGA_CONSOLE=y
++# CONFIG_VGACON_SOFT_SCROLLBACK is not set
++CONFIG_DUMMY_CONSOLE=y
++# CONFIG_BACKLIGHT_LCD_SUPPORT is not set
+
-+$(obj)/syscall32_syscall-xen.o: \
-+ $(foreach F,int80 sysenter syscall,$(obj)/vsyscall-$F.so)
++#
++# Sound
++#
++# CONFIG_SOUND is not set
+
-+targets := $(foreach F,int80 sysenter syscall,vsyscall-$F.o vsyscall-$F.so)
++#
++# USB support
++#
++# CONFIG_USB_ARCH_HAS_HCD is not set
++# CONFIG_USB_ARCH_HAS_OHCI is not set
++# CONFIG_USB_ARCH_HAS_EHCI is not set
+
-+include $(srctree)/scripts/Makefile.xen
++#
++# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support'
++#
+
-+obj-y := $(call cherrypickxen, $(obj-y))
-+endif
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/ia32/syscall32_syscall-xen.S tmp-linux-2.6-xen.patch/arch/x86_64/ia32/syscall32_syscall-xen.S
---- pristine-linux-2.6.18.2/arch/x86_64/ia32/syscall32_syscall-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/ia32/syscall32_syscall-xen.S 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,28 @@
-+/* 32bit VDSOs mapped into user space. */
++#
++# USB Gadget Support
++#
++# CONFIG_USB_GADGET is not set
+
-+ .section ".init.data","aw"
++#
++# MMC/SD Card support
++#
++# CONFIG_MMC is not set
+
-+#ifdef USE_INT80
++#
++# LED devices
++#
++# CONFIG_NEW_LEDS is not set
+
-+ .globl syscall32_int80
-+ .globl syscall32_int80_end
++#
++# LED drivers
++#
+
-+syscall32_int80:
-+ .incbin "arch/x86_64/ia32/vsyscall-int80.so"
-+syscall32_int80_end:
++#
++# LED Triggers
++#
+
-+#endif
++#
++# InfiniBand support
++#
+
-+ .globl syscall32_syscall
-+ .globl syscall32_syscall_end
++#
++# EDAC - error detection and reporting (RAS) (EXPERIMENTAL)
++#
++# CONFIG_EDAC is not set
+
-+syscall32_syscall:
-+ .incbin "arch/x86_64/ia32/vsyscall-syscall.so"
-+syscall32_syscall_end:
++#
++# Real Time Clock
++#
++# CONFIG_RTC_CLASS is not set
+
-+ .globl syscall32_sysenter
-+ .globl syscall32_sysenter_end
++#
++# DMA Engine support
++#
++# CONFIG_DMA_ENGINE is not set
+
-+syscall32_sysenter:
-+ .incbin "arch/x86_64/ia32/vsyscall-sysenter.so"
-+syscall32_sysenter_end:
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/ia32/syscall32-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/ia32/syscall32-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/ia32/syscall32-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/ia32/syscall32-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,128 @@
-+/* Copyright 2002,2003 Andi Kleen, SuSE Labs */
++#
++# DMA Clients
++#
+
-+/* vsyscall handling for 32bit processes. Map a stub page into it
-+ on demand because 32bit cannot reach the kernel's fixmaps */
++#
++# DMA Devices
++#
+
-+#include <linux/mm.h>
-+#include <linux/string.h>
-+#include <linux/kernel.h>
-+#include <linux/gfp.h>
-+#include <linux/init.h>
-+#include <linux/stringify.h>
-+#include <linux/security.h>
-+#include <asm/proto.h>
-+#include <asm/tlbflush.h>
-+#include <asm/ia32_unistd.h>
++#
++# File systems
++#
++CONFIG_EXT2_FS=y
++# CONFIG_EXT2_FS_XATTR is not set
++# CONFIG_EXT2_FS_XIP is not set
++CONFIG_EXT3_FS=y
++CONFIG_EXT3_FS_XATTR=y
++# CONFIG_EXT3_FS_POSIX_ACL is not set
++# CONFIG_EXT3_FS_SECURITY is not set
++CONFIG_JBD=y
++# CONFIG_JBD_DEBUG is not set
++CONFIG_FS_MBCACHE=y
++CONFIG_REISERFS_FS=y
++# CONFIG_REISERFS_CHECK is not set
++# CONFIG_REISERFS_PROC_INFO is not set
++# CONFIG_REISERFS_FS_XATTR is not set
++# CONFIG_JFS_FS is not set
++# CONFIG_FS_POSIX_ACL is not set
++# CONFIG_XFS_FS is not set
++# CONFIG_OCFS2_FS is not set
++# CONFIG_MINIX_FS is not set
++# CONFIG_ROMFS_FS is not set
++CONFIG_INOTIFY=y
++CONFIG_INOTIFY_USER=y
++# CONFIG_QUOTA is not set
++CONFIG_DNOTIFY=y
++CONFIG_AUTOFS_FS=y
++CONFIG_AUTOFS4_FS=y
++# CONFIG_FUSE_FS is not set
+
-+#ifdef USE_INT80
-+extern unsigned char syscall32_int80[], syscall32_int80_end[];
-+#endif
-+extern unsigned char syscall32_syscall[], syscall32_syscall_end[];
-+extern unsigned char syscall32_sysenter[], syscall32_sysenter_end[];
-+extern int sysctl_vsyscall32;
++#
++# CD-ROM/DVD Filesystems
++#
++CONFIG_ISO9660_FS=y
++CONFIG_JOLIET=y
++CONFIG_ZISOFS=y
++CONFIG_ZISOFS_FS=y
++# CONFIG_UDF_FS is not set
+
-+char *syscall32_page;
-+#ifndef USE_INT80
-+static int use_sysenter = -1;
-+#endif
++#
++# DOS/FAT/NT Filesystems
++#
++CONFIG_FAT_FS=m
++CONFIG_MSDOS_FS=m
++CONFIG_VFAT_FS=m
++CONFIG_FAT_DEFAULT_CODEPAGE=437
++CONFIG_FAT_DEFAULT_IOCHARSET="iso8859-1"
++# CONFIG_NTFS_FS is not set
+
-+static struct page *
-+syscall32_nopage(struct vm_area_struct *vma, unsigned long adr, int *type)
-+{
-+ struct page *p = virt_to_page(adr - vma->vm_start + syscall32_page);
-+ get_page(p);
-+ return p;
-+}
++#
++# Pseudo filesystems
++#
++CONFIG_PROC_FS=y
++CONFIG_PROC_KCORE=y
++CONFIG_SYSFS=y
++CONFIG_TMPFS=y
++# CONFIG_HUGETLB_PAGE is not set
++CONFIG_RAMFS=y
++# CONFIG_CONFIGFS_FS is not set
+
-+/* Prevent VMA merging */
-+static void syscall32_vma_close(struct vm_area_struct *vma)
-+{
-+}
++#
++# Miscellaneous filesystems
++#
++# CONFIG_ADFS_FS is not set
++# CONFIG_AFFS_FS is not set
++# CONFIG_HFS_FS is not set
++# CONFIG_HFSPLUS_FS is not set
++# CONFIG_BEFS_FS is not set
++# CONFIG_BFS_FS is not set
++# CONFIG_EFS_FS is not set
++CONFIG_CRAMFS=y
++# CONFIG_VXFS_FS is not set
++# CONFIG_HPFS_FS is not set
++# CONFIG_QNX4FS_FS is not set
++# CONFIG_SYSV_FS is not set
++# CONFIG_UFS_FS is not set
+
-+static struct vm_operations_struct syscall32_vm_ops = {
-+ .close = syscall32_vma_close,
-+ .nopage = syscall32_nopage,
-+};
++#
++# Network File Systems
++#
++CONFIG_NFS_FS=y
++CONFIG_NFS_V3=y
++# CONFIG_NFS_V3_ACL is not set
++# CONFIG_NFS_V4 is not set
++# CONFIG_NFS_DIRECTIO is not set
++# CONFIG_NFSD is not set
++CONFIG_ROOT_NFS=y
++CONFIG_LOCKD=y
++CONFIG_LOCKD_V4=y
++CONFIG_NFS_COMMON=y
++CONFIG_SUNRPC=y
++# CONFIG_RPCSEC_GSS_KRB5 is not set
++# CONFIG_RPCSEC_GSS_SPKM3 is not set
++# CONFIG_SMB_FS is not set
++# CONFIG_CIFS is not set
++# CONFIG_NCP_FS is not set
++# CONFIG_CODA_FS is not set
++# CONFIG_AFS_FS is not set
++# CONFIG_9P_FS is not set
+
-+struct linux_binprm;
++#
++# Partition Types
++#
++# CONFIG_PARTITION_ADVANCED is not set
++CONFIG_MSDOS_PARTITION=y
+
-+/* Setup a VMA at program startup for the vsyscall page */
-+int syscall32_setup_pages(struct linux_binprm *bprm, int exstack)
-+{
-+ int npages = (VSYSCALL32_END - VSYSCALL32_BASE) >> PAGE_SHIFT;
-+ struct vm_area_struct *vma;
-+ struct mm_struct *mm = current->mm;
-+ int ret;
++#
++# Native Language Support
++#
++CONFIG_NLS=y
++CONFIG_NLS_DEFAULT="iso8859-1"
++CONFIG_NLS_CODEPAGE_437=y
++# CONFIG_NLS_CODEPAGE_737 is not set
++# CONFIG_NLS_CODEPAGE_775 is not set
++# CONFIG_NLS_CODEPAGE_850 is not set
++# CONFIG_NLS_CODEPAGE_852 is not set
++# CONFIG_NLS_CODEPAGE_855 is not set
++# CONFIG_NLS_CODEPAGE_857 is not set
++# CONFIG_NLS_CODEPAGE_860 is not set
++# CONFIG_NLS_CODEPAGE_861 is not set
++# CONFIG_NLS_CODEPAGE_862 is not set
++# CONFIG_NLS_CODEPAGE_863 is not set
++# CONFIG_NLS_CODEPAGE_864 is not set
++# CONFIG_NLS_CODEPAGE_865 is not set
++# CONFIG_NLS_CODEPAGE_866 is not set
++# CONFIG_NLS_CODEPAGE_869 is not set
++# CONFIG_NLS_CODEPAGE_936 is not set
++# CONFIG_NLS_CODEPAGE_950 is not set
++# CONFIG_NLS_CODEPAGE_932 is not set
++# CONFIG_NLS_CODEPAGE_949 is not set
++# CONFIG_NLS_CODEPAGE_874 is not set
++# CONFIG_NLS_ISO8859_8 is not set
++# CONFIG_NLS_CODEPAGE_1250 is not set
++# CONFIG_NLS_CODEPAGE_1251 is not set
++# CONFIG_NLS_ASCII is not set
++CONFIG_NLS_ISO8859_1=y
++# CONFIG_NLS_ISO8859_2 is not set
++# CONFIG_NLS_ISO8859_3 is not set
++# CONFIG_NLS_ISO8859_4 is not set
++# CONFIG_NLS_ISO8859_5 is not set
++# CONFIG_NLS_ISO8859_6 is not set
++# CONFIG_NLS_ISO8859_7 is not set
++# CONFIG_NLS_ISO8859_9 is not set
++# CONFIG_NLS_ISO8859_13 is not set
++# CONFIG_NLS_ISO8859_14 is not set
++# CONFIG_NLS_ISO8859_15 is not set
++# CONFIG_NLS_KOI8_R is not set
++# CONFIG_NLS_KOI8_U is not set
++# CONFIG_NLS_UTF8 is not set
+
-+ vma = kmem_cache_alloc(vm_area_cachep, SLAB_KERNEL);
-+ if (!vma)
-+ return -ENOMEM;
++#
++# Instrumentation Support
++#
++# CONFIG_PROFILING is not set
++# CONFIG_KPROBES is not set
+
-+ memset(vma, 0, sizeof(struct vm_area_struct));
-+ /* Could randomize here */
-+ vma->vm_start = VSYSCALL32_BASE;
-+ vma->vm_end = VSYSCALL32_END;
-+ /* MAYWRITE to allow gdb to COW and set breakpoints */
-+ vma->vm_flags = VM_READ|VM_EXEC|VM_MAYREAD|VM_MAYEXEC|VM_MAYWRITE;
-+ vma->vm_flags |= mm->def_flags;
-+ vma->vm_page_prot = protection_map[vma->vm_flags & 7];
-+ vma->vm_ops = &syscall32_vm_ops;
-+ vma->vm_mm = mm;
++#
++# Kernel hacking
++#
++CONFIG_TRACE_IRQFLAGS_SUPPORT=y
++# CONFIG_PRINTK_TIME is not set
++CONFIG_MAGIC_SYSRQ=y
++CONFIG_UNUSED_SYMBOLS=y
++CONFIG_DEBUG_KERNEL=y
++CONFIG_LOG_BUF_SHIFT=14
++CONFIG_DETECT_SOFTLOCKUP=y
++# CONFIG_SCHEDSTATS is not set
++# CONFIG_DEBUG_SLAB is not set
++# CONFIG_DEBUG_RT_MUTEXES is not set
++# CONFIG_RT_MUTEX_TESTER is not set
++# CONFIG_DEBUG_SPINLOCK is not set
++# CONFIG_DEBUG_MUTEXES is not set
++# CONFIG_DEBUG_RWSEMS is not set
++# CONFIG_DEBUG_LOCK_ALLOC is not set
++# CONFIG_PROVE_LOCKING is not set
++# CONFIG_DEBUG_SPINLOCK_SLEEP is not set
++# CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set
++# CONFIG_DEBUG_KOBJECT is not set
++# CONFIG_DEBUG_HIGHMEM is not set
++CONFIG_DEBUG_BUGVERBOSE=y
++CONFIG_DEBUG_INFO=y
++# CONFIG_DEBUG_FS is not set
++# CONFIG_DEBUG_VM is not set
++CONFIG_FRAME_POINTER=y
++# CONFIG_UNWIND_INFO is not set
++CONFIG_FORCED_INLINING=y
++# CONFIG_RCU_TORTURE_TEST is not set
++CONFIG_EARLY_PRINTK=y
++# CONFIG_DEBUG_STACKOVERFLOW is not set
++# CONFIG_DEBUG_STACK_USAGE is not set
++# CONFIG_DEBUG_PAGEALLOC is not set
++# CONFIG_DEBUG_RODATA is not set
++# CONFIG_4KSTACKS is not set
+
-+ down_write(&mm->mmap_sem);
-+ if ((ret = insert_vm_struct(mm, vma))) {
-+ up_write(&mm->mmap_sem);
-+ kmem_cache_free(vm_area_cachep, vma);
-+ return ret;
-+ }
-+ mm->total_vm += npages;
-+ up_write(&mm->mmap_sem);
-+ return 0;
-+}
++#
++# Security options
++#
++# CONFIG_KEYS is not set
++# CONFIG_SECURITY is not set
+
-+static int __init init_syscall32(void)
-+{
-+ syscall32_page = (void *)get_zeroed_page(GFP_KERNEL);
-+ if (!syscall32_page)
-+ panic("Cannot allocate syscall32 page");
++#
++# Cryptographic options
++#
++CONFIG_CRYPTO=y
++# CONFIG_CRYPTO_HMAC is not set
++# CONFIG_CRYPTO_NULL is not set
++# CONFIG_CRYPTO_MD4 is not set
++CONFIG_CRYPTO_MD5=m
++# CONFIG_CRYPTO_SHA1 is not set
++# CONFIG_CRYPTO_SHA256 is not set
++# CONFIG_CRYPTO_SHA512 is not set
++# CONFIG_CRYPTO_WP512 is not set
++# CONFIG_CRYPTO_TGR192 is not set
++# CONFIG_CRYPTO_DES is not set
++# CONFIG_CRYPTO_BLOWFISH is not set
++# CONFIG_CRYPTO_TWOFISH is not set
++# CONFIG_CRYPTO_SERPENT is not set
++# CONFIG_CRYPTO_AES is not set
++# CONFIG_CRYPTO_AES_586 is not set
++# CONFIG_CRYPTO_CAST5 is not set
++# CONFIG_CRYPTO_CAST6 is not set
++# CONFIG_CRYPTO_TEA is not set
++# CONFIG_CRYPTO_ARC4 is not set
++# CONFIG_CRYPTO_KHAZAD is not set
++# CONFIG_CRYPTO_ANUBIS is not set
++# CONFIG_CRYPTO_DEFLATE is not set
++# CONFIG_CRYPTO_MICHAEL_MIC is not set
++CONFIG_CRYPTO_CRC32C=m
++# CONFIG_CRYPTO_TEST is not set
+
-+#ifdef USE_INT80
-+ /*
-+ * At this point we use int 0x80.
-+ */
-+ memcpy(syscall32_page, syscall32_int80,
-+ syscall32_int80_end - syscall32_int80);
-+#else
-+ if (use_sysenter > 0) {
-+ memcpy(syscall32_page, syscall32_sysenter,
-+ syscall32_sysenter_end - syscall32_sysenter);
-+ } else {
-+ memcpy(syscall32_page, syscall32_syscall,
-+ syscall32_syscall_end - syscall32_syscall);
-+ }
-+#endif
-+ return 0;
-+}
++#
++# Hardware crypto devices
++#
++# CONFIG_CRYPTO_DEV_PADLOCK is not set
++CONFIG_XEN=y
++CONFIG_XEN_INTERFACE_VERSION=0x00030207
+
-+/*
-+ * This must be done early in case we have an initrd containing 32-bit
-+ * binaries (e.g., hotplug). This could be pushed upstream to arch/x86_64.
-+ */
-+core_initcall(init_syscall32);
++#
++# XEN
++#
++# CONFIG_XEN_PRIVILEGED_GUEST is not set
++CONFIG_XEN_UNPRIVILEGED_GUEST=y
++CONFIG_XEN_PRIVCMD=y
++CONFIG_XEN_XENBUS_DEV=y
++# CONFIG_XEN_BACKEND is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_UTIL is not set
++CONFIG_XEN_BLKDEV_FRONTEND=y
++CONFIG_XEN_NETDEV_FRONTEND=y
++CONFIG_XEN_SCSI_FRONTEND=m
++# CONFIG_XEN_GRANT_DEV is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_FRONTEND is not set
++CONFIG_XEN_SCRUB_PAGES=y
++CONFIG_XEN_DISABLE_SERIAL=y
++CONFIG_XEN_SYSFS=y
++CONFIG_XEN_COMPAT_030002_AND_LATER=y
++# CONFIG_XEN_COMPAT_030004_AND_LATER is not set
++# CONFIG_XEN_COMPAT_030100_AND_LATER is not set
++# CONFIG_XEN_COMPAT_LATEST_ONLY is not set
++CONFIG_XEN_COMPAT=0x030002
++CONFIG_HAVE_IRQ_IGNORE_UNHANDLED=y
++CONFIG_NO_IDLE_HZ=y
++CONFIG_XEN_SMPBOOT=y
++CONFIG_XEN_BALLOON=y
++CONFIG_XEN_DEVMEM=y
+
-+/* May not be __init: called during resume */
-+void syscall32_cpu_init(void)
-+{
-+#ifndef USE_INT80
-+ if (use_sysenter < 0)
-+ use_sysenter = (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL);
++#
++# Library routines
++#
++# CONFIG_CRC_CCITT is not set
++# CONFIG_CRC16 is not set
++# CONFIG_CRC32 is not set
++CONFIG_LIBCRC32C=m
++CONFIG_ZLIB_INFLATE=y
++CONFIG_PLIST=y
++CONFIG_GENERIC_HARDIRQS=y
++CONFIG_GENERIC_IRQ_PROBE=y
++CONFIG_GENERIC_PENDING_IRQ=y
++CONFIG_X86_SMP=y
++CONFIG_X86_BIOS_REBOOT=y
++CONFIG_X86_TRAMPOLINE=y
++CONFIG_X86_NO_TSS=y
++CONFIG_X86_NO_IDT=y
++CONFIG_KTIME_SCALAR=y
+diff -r d894e36cfc30 -r 0aa021803deb buildconfigs/linux-defconfig_xenU_x86_64
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/buildconfigs/linux-defconfig_xenU_x86_64 Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,906 @@
++#
++# Automatically generated make config: don't edit
++# Linux kernel version: 2.6.18.8
++# Mon Feb 18 10:42:42 2008
++#
++CONFIG_X86_64=y
++CONFIG_64BIT=y
++CONFIG_X86=y
++CONFIG_LOCKDEP_SUPPORT=y
++CONFIG_STACKTRACE_SUPPORT=y
++CONFIG_SEMAPHORE_SLEEPERS=y
++CONFIG_MMU=y
++CONFIG_RWSEM_GENERIC_SPINLOCK=y
++CONFIG_GENERIC_HWEIGHT=y
++CONFIG_GENERIC_CALIBRATE_DELAY=y
++CONFIG_X86_CMPXCHG=y
++CONFIG_EARLY_PRINTK=y
++CONFIG_GENERIC_ISA_DMA=y
++CONFIG_GENERIC_IOMAP=y
++CONFIG_ARCH_MAY_HAVE_PC_FDC=y
++CONFIG_DMI=y
++CONFIG_AUDIT_ARCH=y
++CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config"
+
-+ /* Load these always in case some future AMD CPU supports
-+ SYSENTER from compat mode too. */
-+ checking_wrmsrl(MSR_IA32_SYSENTER_CS, (u64)__KERNEL_CS);
-+ checking_wrmsrl(MSR_IA32_SYSENTER_ESP, 0ULL);
-+ checking_wrmsrl(MSR_IA32_SYSENTER_EIP, (u64)ia32_sysenter_target);
++#
++# Code maturity level options
++#
++CONFIG_EXPERIMENTAL=y
++CONFIG_LOCK_KERNEL=y
++CONFIG_INIT_ENV_ARG_LIMIT=32
+
-+ wrmsrl(MSR_CSTAR, ia32_cstar_target);
-+#endif
-+}
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/ia32/vsyscall-int80.S tmp-linux-2.6-xen.patch/arch/x86_64/ia32/vsyscall-int80.S
---- pristine-linux-2.6.18.2/arch/x86_64/ia32/vsyscall-int80.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/ia32/vsyscall-int80.S 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,58 @@
-+/*
-+ * Code for the vsyscall page. This version uses the old int $0x80 method.
-+ *
-+ * NOTE:
-+ * 1) __kernel_vsyscall _must_ be first in this page.
-+ * 2) there are alignment constraints on this stub, see vsyscall-sigreturn.S
-+ * for details.
-+ */
-+#include <asm/ia32_unistd.h>
-+#include <asm/asm-offsets.h>
++#
++# General setup
++#
++CONFIG_LOCALVERSION=""
++# CONFIG_LOCALVERSION_AUTO is not set
++CONFIG_SWAP=y
++CONFIG_SYSVIPC=y
++# CONFIG_POSIX_MQUEUE is not set
++# CONFIG_BSD_PROCESS_ACCT is not set
++# CONFIG_TASKSTATS is not set
++# CONFIG_AUDIT is not set
++# CONFIG_IKCONFIG is not set
++# CONFIG_CPUSETS is not set
++# CONFIG_RELAY is not set
++CONFIG_INITRAMFS_SOURCE=""
++# CONFIG_CC_OPTIMIZE_FOR_SIZE is not set
++# CONFIG_EMBEDDED is not set
++CONFIG_UID16=y
++CONFIG_SYSCTL=y
++CONFIG_KALLSYMS=y
++# CONFIG_KALLSYMS_ALL is not set
++# CONFIG_KALLSYMS_EXTRA_PASS is not set
++CONFIG_HOTPLUG=y
++CONFIG_PRINTK=y
++CONFIG_BUG=y
++CONFIG_ELF_CORE=y
++CONFIG_BASE_FULL=y
++CONFIG_FUTEX=y
++CONFIG_EPOLL=y
++CONFIG_SHMEM=y
++CONFIG_SLAB=y
++CONFIG_VM_EVENT_COUNTERS=y
++CONFIG_RT_MUTEXES=y
++# CONFIG_TINY_SHMEM is not set
++CONFIG_BASE_SMALL=0
++# CONFIG_SLOB is not set
++
++#
++# Loadable module support
++#
++CONFIG_MODULES=y
++CONFIG_MODULE_UNLOAD=y
++# CONFIG_MODULE_FORCE_UNLOAD is not set
++# CONFIG_MODVERSIONS is not set
++# CONFIG_MODULE_SRCVERSION_ALL is not set
++CONFIG_KMOD=y
++CONFIG_STOP_MACHINE=y
++
++#
++# Block layer
++#
++# CONFIG_LBD is not set
++# CONFIG_BLK_DEV_IO_TRACE is not set
++CONFIG_LSF=y
++
++#
++# IO Schedulers
++#
++CONFIG_IOSCHED_NOOP=y
++CONFIG_IOSCHED_AS=y
++CONFIG_IOSCHED_DEADLINE=y
++CONFIG_IOSCHED_CFQ=y
++CONFIG_DEFAULT_AS=y
++# CONFIG_DEFAULT_DEADLINE is not set
++# CONFIG_DEFAULT_CFQ is not set
++# CONFIG_DEFAULT_NOOP is not set
++CONFIG_DEFAULT_IOSCHED="anticipatory"
++
++#
++# Processor type and features
++#
++CONFIG_X86_PC=y
++# CONFIG_X86_VSMP is not set
++# CONFIG_MK8 is not set
++# CONFIG_MPSC is not set
++CONFIG_GENERIC_CPU=y
++CONFIG_X86_64_XEN=y
++CONFIG_X86_NO_TSS=y
++CONFIG_X86_NO_IDT=y
++CONFIG_X86_L1_CACHE_BYTES=128
++CONFIG_X86_L1_CACHE_SHIFT=7
++CONFIG_X86_INTERNODE_CACHE_BYTES=128
++CONFIG_X86_GOOD_APIC=y
++# CONFIG_MICROCODE is not set
++# CONFIG_X86_MSR is not set
++CONFIG_X86_CPUID=y
++CONFIG_X86_XEN_GENAPIC=y
++CONFIG_SMP=y
++CONFIG_PREEMPT_NONE=y
++# CONFIG_PREEMPT_VOLUNTARY is not set
++# CONFIG_PREEMPT is not set
++CONFIG_PREEMPT_BKL=y
++CONFIG_ARCH_FLATMEM_ENABLE=y
++CONFIG_SELECT_MEMORY_MODEL=y
++CONFIG_FLATMEM_MANUAL=y
++# CONFIG_DISCONTIGMEM_MANUAL is not set
++# CONFIG_SPARSEMEM_MANUAL is not set
++CONFIG_FLATMEM=y
++CONFIG_FLAT_NODE_MEM_MAP=y
++# CONFIG_SPARSEMEM_STATIC is not set
++CONFIG_SPLIT_PTLOCK_CPUS=4
++CONFIG_RESOURCES_64BIT=y
++CONFIG_NR_CPUS=8
++CONFIG_HOTPLUG_CPU=y
++CONFIG_ARCH_ENABLE_MEMORY_HOTPLUG=y
++CONFIG_SWIOTLB=y
++# CONFIG_CRASH_DUMP is not set
++CONFIG_PHYSICAL_START=0x200000
++CONFIG_SECCOMP=y
++CONFIG_HZ_100=y
++# CONFIG_HZ_250 is not set
++# CONFIG_HZ_1000 is not set
++CONFIG_HZ=100
++# CONFIG_REORDER is not set
++CONFIG_GENERIC_HARDIRQS=y
++CONFIG_GENERIC_IRQ_PROBE=y
++CONFIG_ISA_DMA_API=y
++CONFIG_GENERIC_PENDING_IRQ=y
++
++#
++# Bus options (PCI etc.)
++#
++# CONFIG_PCI is not set
++
++#
++# PCCARD (PCMCIA/CardBus) support
++#
++# CONFIG_PCCARD is not set
++
++#
++# PCI Hotplug Support
++#
++
++#
++# Executable file formats / Emulations
++#
++CONFIG_BINFMT_ELF=y
++# CONFIG_BINFMT_MISC is not set
++CONFIG_IA32_EMULATION=y
++CONFIG_IA32_AOUT=y
++CONFIG_COMPAT=y
++CONFIG_SYSVIPC_COMPAT=y
++
++#
++# Networking
++#
++CONFIG_NET=y
++
++#
++# Networking options
++#
++# CONFIG_NETDEBUG is not set
++CONFIG_PACKET=y
++# CONFIG_PACKET_MMAP is not set
++CONFIG_UNIX=y
++CONFIG_XFRM=y
++# CONFIG_XFRM_USER is not set
++# CONFIG_NET_KEY is not set
++CONFIG_INET=y
++# CONFIG_IP_MULTICAST is not set
++# CONFIG_IP_ADVANCED_ROUTER is not set
++CONFIG_IP_FIB_HASH=y
++CONFIG_IP_PNP=y
++# CONFIG_IP_PNP_DHCP is not set
++# CONFIG_IP_PNP_BOOTP is not set
++# CONFIG_IP_PNP_RARP is not set
++# CONFIG_NET_IPIP is not set
++# CONFIG_NET_IPGRE is not set
++# CONFIG_ARPD is not set
++# CONFIG_SYN_COOKIES is not set
++# CONFIG_INET_AH is not set
++# CONFIG_INET_ESP is not set
++# CONFIG_INET_IPCOMP is not set
++# CONFIG_INET_XFRM_TUNNEL is not set
++# CONFIG_INET_TUNNEL is not set
++CONFIG_INET_XFRM_MODE_TRANSPORT=y
++CONFIG_INET_XFRM_MODE_TUNNEL=y
++# CONFIG_INET_DIAG is not set
++# CONFIG_TCP_CONG_ADVANCED is not set
++CONFIG_TCP_CONG_BIC=y
++# CONFIG_IPV6 is not set
++# CONFIG_INET6_XFRM_TUNNEL is not set
++# CONFIG_INET6_TUNNEL is not set
++# CONFIG_NETWORK_SECMARK is not set
++# CONFIG_NETFILTER is not set
++
++#
++# DCCP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_DCCP is not set
++
++#
++# SCTP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_SCTP is not set
++
++#
++# TIPC Configuration (EXPERIMENTAL)
++#
++# CONFIG_TIPC is not set
++# CONFIG_ATM is not set
++# CONFIG_BRIDGE is not set
++# CONFIG_VLAN_8021Q is not set
++# CONFIG_DECNET is not set
++# CONFIG_LLC2 is not set
++# CONFIG_IPX is not set
++# CONFIG_ATALK is not set
++# CONFIG_X25 is not set
++# CONFIG_LAPB is not set
++# CONFIG_ECONET is not set
++# CONFIG_WAN_ROUTER is not set
++
++#
++# QoS and/or fair queueing
++#
++# CONFIG_NET_SCHED is not set
++
++#
++# Network testing
++#
++# CONFIG_NET_PKTGEN is not set
++# CONFIG_HAMRADIO is not set
++# CONFIG_IRDA is not set
++# CONFIG_BT is not set
++# CONFIG_IEEE80211 is not set
++
++#
++# Device Drivers
++#
++
++#
++# Generic Driver Options
++#
++CONFIG_STANDALONE=y
++CONFIG_PREVENT_FIRMWARE_BUILD=y
++# CONFIG_FW_LOADER is not set
++# CONFIG_DEBUG_DRIVER is not set
++CONFIG_SYS_HYPERVISOR=y
++
++#
++# Connector - unified userspace <-> kernelspace linker
++#
++# CONFIG_CONNECTOR is not set
++
++#
++# Memory Technology Devices (MTD)
++#
++# CONFIG_MTD is not set
++
++#
++# Parallel port support
++#
++# CONFIG_PARPORT is not set
++
++#
++# Plug and Play support
++#
++
++#
++# Block devices
++#
++# CONFIG_BLK_DEV_FD is not set
++# CONFIG_BLK_DEV_COW_COMMON is not set
++CONFIG_BLK_DEV_LOOP=m
++# CONFIG_BLK_DEV_CRYPTOLOOP is not set
++CONFIG_BLK_DEV_NBD=m
++CONFIG_BLK_DEV_RAM=y
++CONFIG_BLK_DEV_RAM_COUNT=16
++CONFIG_BLK_DEV_RAM_SIZE=4096
++CONFIG_BLK_DEV_RAM_BLOCKSIZE=1024
++CONFIG_BLK_DEV_INITRD=y
++# CONFIG_CDROM_PKTCDVD is not set
++# CONFIG_ATA_OVER_ETH is not set
++
++#
++# ATA/ATAPI/MFM/RLL support
++#
++# CONFIG_IDE is not set
++
++#
++# SCSI device support
++#
++# CONFIG_RAID_ATTRS is not set
++CONFIG_SCSI=m
++CONFIG_SCSI_PROC_FS=y
++
++#
++# SCSI support type (disk, tape, CD-ROM)
++#
++CONFIG_BLK_DEV_SD=m
++# CONFIG_CHR_DEV_ST is not set
++# CONFIG_CHR_DEV_OSST is not set
++# CONFIG_BLK_DEV_SR is not set
++# CONFIG_CHR_DEV_SG is not set
++# CONFIG_CHR_DEV_SCH is not set
++
++#
++# Some SCSI devices (e.g. CD jukebox) support multiple LUNs
++#
++# CONFIG_SCSI_MULTI_LUN is not set
++# CONFIG_SCSI_CONSTANTS is not set
++# CONFIG_SCSI_LOGGING is not set
++
++#
++# SCSI Transport Attributes
++#
++# CONFIG_SCSI_SPI_ATTRS is not set
++# CONFIG_SCSI_FC_ATTRS is not set
++# CONFIG_SCSI_ISCSI_ATTRS is not set
++# CONFIG_SCSI_SAS_ATTRS is not set
++
++#
++# SCSI low-level drivers
++#
++# CONFIG_ISCSI_TCP is not set
++# CONFIG_SCSI_SATA is not set
++# CONFIG_SCSI_DEBUG is not set
++
++#
++# Multi-device support (RAID and LVM)
++#
++# CONFIG_MD is not set
++
++#
++# Fusion MPT device support
++#
++# CONFIG_FUSION is not set
++
++#
++# IEEE 1394 (FireWire) support
++#
++
++#
++# I2O device support
++#
++
++#
++# Network device support
++#
++CONFIG_NETDEVICES=y
++# CONFIG_DUMMY is not set
++# CONFIG_BONDING is not set
++# CONFIG_EQUALIZER is not set
++# CONFIG_TUN is not set
++
++#
++# PHY device support
++#
++
++#
++# Ethernet (10 or 100Mbit)
++#
++# CONFIG_NET_ETHERNET is not set
++
++#
++# Ethernet (1000 Mbit)
++#
++
++#
++# Ethernet (10000 Mbit)
++#
++
++#
++# Token Ring devices
++#
++
++#
++# Wireless LAN (non-hamradio)
++#
++# CONFIG_NET_RADIO is not set
++
++#
++# Wan interfaces
++#
++# CONFIG_WAN is not set
++# CONFIG_PPP is not set
++# CONFIG_SLIP is not set
++# CONFIG_SHAPER is not set
++# CONFIG_NETCONSOLE is not set
++# CONFIG_NETPOLL is not set
++# CONFIG_NET_POLL_CONTROLLER is not set
++
++#
++# ISDN subsystem
++#
++# CONFIG_ISDN is not set
++
++#
++# Telephony Support
++#
++# CONFIG_PHONE is not set
++
++#
++# Input device support
++#
++CONFIG_INPUT=y
++
++#
++# Userland interfaces
++#
++CONFIG_INPUT_MOUSEDEV=y
++# CONFIG_INPUT_MOUSEDEV_PSAUX is not set
++CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024
++CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768
++# CONFIG_INPUT_JOYDEV is not set
++# CONFIG_INPUT_TSDEV is not set
++# CONFIG_INPUT_EVDEV is not set
++# CONFIG_INPUT_EVBUG is not set
++
++#
++# Input Device Drivers
++#
++CONFIG_INPUT_KEYBOARD=y
++CONFIG_KEYBOARD_ATKBD=y
++# CONFIG_KEYBOARD_SUNKBD is not set
++# CONFIG_KEYBOARD_LKKBD is not set
++# CONFIG_KEYBOARD_XTKBD is not set
++# CONFIG_KEYBOARD_NEWTON is not set
++# CONFIG_INPUT_MOUSE is not set
++# CONFIG_INPUT_JOYSTICK is not set
++# CONFIG_INPUT_TOUCHSCREEN is not set
++# CONFIG_INPUT_MISC is not set
++
++#
++# Hardware I/O ports
++#
++CONFIG_SERIO=y
++CONFIG_SERIO_I8042=y
++CONFIG_SERIO_SERPORT=y
++# CONFIG_SERIO_CT82C710 is not set
++CONFIG_SERIO_LIBPS2=y
++# CONFIG_SERIO_RAW is not set
++# CONFIG_GAMEPORT is not set
++
++#
++# Character devices
++#
++CONFIG_VT=y
++CONFIG_VT_CONSOLE=y
++CONFIG_HW_CONSOLE=y
++# CONFIG_VT_HW_CONSOLE_BINDING is not set
++# CONFIG_SERIAL_NONSTANDARD is not set
++
++#
++# Serial drivers
++#
++
++#
++# Non-8250 serial port support
++#
++CONFIG_UNIX98_PTYS=y
++CONFIG_LEGACY_PTYS=y
++CONFIG_LEGACY_PTY_COUNT=256
++
++#
++# IPMI
++#
++# CONFIG_IPMI_HANDLER is not set
++
++#
++# Watchdog Cards
++#
++# CONFIG_WATCHDOG is not set
++CONFIG_HW_RANDOM=y
++# CONFIG_NVRAM is not set
++# CONFIG_RTC is not set
++# CONFIG_GEN_RTC is not set
++# CONFIG_DTLK is not set
++# CONFIG_R3964 is not set
++
++#
++# Ftape, the floppy tape device driver
++#
++# CONFIG_MWAVE is not set
++# CONFIG_PC8736x_GPIO is not set
++# CONFIG_RAW_DRIVER is not set
++# CONFIG_HANGCHECK_TIMER is not set
++
++#
++# TPM devices
++#
++# CONFIG_TCG_TPM is not set
++# CONFIG_TELCLOCK is not set
++
++#
++# I2C support
++#
++# CONFIG_I2C is not set
++
++#
++# SPI support
++#
++# CONFIG_SPI is not set
++# CONFIG_SPI_MASTER is not set
++
++#
++# Dallas's 1-wire bus
++#
++
++#
++# Hardware Monitoring support
++#
++# CONFIG_HWMON is not set
++# CONFIG_HWMON_VID is not set
++
++#
++# Misc devices
++#
++
++#
++# Multimedia devices
++#
++# CONFIG_VIDEO_DEV is not set
++
++#
++# Digital Video Broadcasting Devices
++#
++# CONFIG_DVB is not set
++
++#
++# Graphics support
++#
++CONFIG_FIRMWARE_EDID=y
++# CONFIG_FB is not set
++
++#
++# Console display driver support
++#
++CONFIG_VGA_CONSOLE=y
++# CONFIG_VGACON_SOFT_SCROLLBACK is not set
++CONFIG_DUMMY_CONSOLE=y
++# CONFIG_BACKLIGHT_LCD_SUPPORT is not set
++
++#
++# Sound
++#
++# CONFIG_SOUND is not set
++
++#
++# USB support
++#
++# CONFIG_USB_ARCH_HAS_HCD is not set
++# CONFIG_USB_ARCH_HAS_OHCI is not set
++# CONFIG_USB_ARCH_HAS_EHCI is not set
++
++#
++# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support'
++#
++
++#
++# USB Gadget Support
++#
++# CONFIG_USB_GADGET is not set
++
++#
++# MMC/SD Card support
++#
++# CONFIG_MMC is not set
++
++#
++# LED devices
++#
++# CONFIG_NEW_LEDS is not set
++
++#
++# LED drivers
++#
++
++#
++# LED Triggers
++#
++
++#
++# InfiniBand support
++#
++
++#
++# EDAC - error detection and reporting (RAS) (EXPERIMENTAL)
++#
++# CONFIG_EDAC is not set
++
++#
++# Real Time Clock
++#
++# CONFIG_RTC_CLASS is not set
++
++#
++# DMA Engine support
++#
++# CONFIG_DMA_ENGINE is not set
++
++#
++# DMA Clients
++#
++
++#
++# DMA Devices
++#
++
++#
++# Firmware Drivers
++#
++# CONFIG_EDD is not set
++# CONFIG_DELL_RBU is not set
++# CONFIG_DCDBAS is not set
++
++#
++# File systems
++#
++CONFIG_EXT2_FS=y
++# CONFIG_EXT2_FS_XATTR is not set
++# CONFIG_EXT2_FS_XIP is not set
++CONFIG_EXT3_FS=y
++CONFIG_EXT3_FS_XATTR=y
++# CONFIG_EXT3_FS_POSIX_ACL is not set
++# CONFIG_EXT3_FS_SECURITY is not set
++CONFIG_JBD=y
++# CONFIG_JBD_DEBUG is not set
++CONFIG_FS_MBCACHE=y
++CONFIG_REISERFS_FS=y
++# CONFIG_REISERFS_CHECK is not set
++# CONFIG_REISERFS_PROC_INFO is not set
++# CONFIG_REISERFS_FS_XATTR is not set
++# CONFIG_JFS_FS is not set
++# CONFIG_FS_POSIX_ACL is not set
++# CONFIG_XFS_FS is not set
++# CONFIG_OCFS2_FS is not set
++# CONFIG_MINIX_FS is not set
++# CONFIG_ROMFS_FS is not set
++CONFIG_INOTIFY=y
++CONFIG_INOTIFY_USER=y
++# CONFIG_QUOTA is not set
++CONFIG_DNOTIFY=y
++CONFIG_AUTOFS_FS=y
++CONFIG_AUTOFS4_FS=y
++# CONFIG_FUSE_FS is not set
++
++#
++# CD-ROM/DVD Filesystems
++#
++CONFIG_ISO9660_FS=y
++CONFIG_JOLIET=y
++CONFIG_ZISOFS=y
++CONFIG_ZISOFS_FS=y
++# CONFIG_UDF_FS is not set
++
++#
++# DOS/FAT/NT Filesystems
++#
++CONFIG_FAT_FS=m
++CONFIG_MSDOS_FS=m
++CONFIG_VFAT_FS=m
++CONFIG_FAT_DEFAULT_CODEPAGE=437
++CONFIG_FAT_DEFAULT_IOCHARSET="iso8859-1"
++# CONFIG_NTFS_FS is not set
++
++#
++# Pseudo filesystems
++#
++CONFIG_PROC_FS=y
++CONFIG_PROC_KCORE=y
++CONFIG_SYSFS=y
++CONFIG_TMPFS=y
++# CONFIG_HUGETLB_PAGE is not set
++CONFIG_RAMFS=y
++# CONFIG_CONFIGFS_FS is not set
++
++#
++# Miscellaneous filesystems
++#
++# CONFIG_ADFS_FS is not set
++# CONFIG_AFFS_FS is not set
++# CONFIG_HFS_FS is not set
++# CONFIG_HFSPLUS_FS is not set
++# CONFIG_BEFS_FS is not set
++# CONFIG_BFS_FS is not set
++# CONFIG_EFS_FS is not set
++CONFIG_CRAMFS=y
++# CONFIG_VXFS_FS is not set
++# CONFIG_HPFS_FS is not set
++# CONFIG_QNX4FS_FS is not set
++# CONFIG_SYSV_FS is not set
++# CONFIG_UFS_FS is not set
++
++#
++# Network File Systems
++#
++CONFIG_NFS_FS=y
++CONFIG_NFS_V3=y
++# CONFIG_NFS_V3_ACL is not set
++# CONFIG_NFS_V4 is not set
++# CONFIG_NFS_DIRECTIO is not set
++# CONFIG_NFSD is not set
++CONFIG_ROOT_NFS=y
++CONFIG_LOCKD=y
++CONFIG_LOCKD_V4=y
++CONFIG_NFS_COMMON=y
++CONFIG_SUNRPC=y
++# CONFIG_RPCSEC_GSS_KRB5 is not set
++# CONFIG_RPCSEC_GSS_SPKM3 is not set
++# CONFIG_SMB_FS is not set
++# CONFIG_CIFS is not set
++# CONFIG_NCP_FS is not set
++# CONFIG_CODA_FS is not set
++# CONFIG_AFS_FS is not set
++# CONFIG_9P_FS is not set
++
++#
++# Partition Types
++#
++# CONFIG_PARTITION_ADVANCED is not set
++CONFIG_MSDOS_PARTITION=y
++
++#
++# Native Language Support
++#
++CONFIG_NLS=y
++CONFIG_NLS_DEFAULT="iso8859-1"
++CONFIG_NLS_CODEPAGE_437=y
++# CONFIG_NLS_CODEPAGE_737 is not set
++# CONFIG_NLS_CODEPAGE_775 is not set
++# CONFIG_NLS_CODEPAGE_850 is not set
++# CONFIG_NLS_CODEPAGE_852 is not set
++# CONFIG_NLS_CODEPAGE_855 is not set
++# CONFIG_NLS_CODEPAGE_857 is not set
++# CONFIG_NLS_CODEPAGE_860 is not set
++# CONFIG_NLS_CODEPAGE_861 is not set
++# CONFIG_NLS_CODEPAGE_862 is not set
++# CONFIG_NLS_CODEPAGE_863 is not set
++# CONFIG_NLS_CODEPAGE_864 is not set
++# CONFIG_NLS_CODEPAGE_865 is not set
++# CONFIG_NLS_CODEPAGE_866 is not set
++# CONFIG_NLS_CODEPAGE_869 is not set
++# CONFIG_NLS_CODEPAGE_936 is not set
++# CONFIG_NLS_CODEPAGE_950 is not set
++# CONFIG_NLS_CODEPAGE_932 is not set
++# CONFIG_NLS_CODEPAGE_949 is not set
++# CONFIG_NLS_CODEPAGE_874 is not set
++# CONFIG_NLS_ISO8859_8 is not set
++# CONFIG_NLS_CODEPAGE_1250 is not set
++# CONFIG_NLS_CODEPAGE_1251 is not set
++# CONFIG_NLS_ASCII is not set
++CONFIG_NLS_ISO8859_1=y
++# CONFIG_NLS_ISO8859_2 is not set
++# CONFIG_NLS_ISO8859_3 is not set
++# CONFIG_NLS_ISO8859_4 is not set
++# CONFIG_NLS_ISO8859_5 is not set
++# CONFIG_NLS_ISO8859_6 is not set
++# CONFIG_NLS_ISO8859_7 is not set
++# CONFIG_NLS_ISO8859_9 is not set
++# CONFIG_NLS_ISO8859_13 is not set
++# CONFIG_NLS_ISO8859_14 is not set
++# CONFIG_NLS_ISO8859_15 is not set
++# CONFIG_NLS_KOI8_R is not set
++# CONFIG_NLS_KOI8_U is not set
++# CONFIG_NLS_UTF8 is not set
++
++#
++# Instrumentation Support
++#
++# CONFIG_PROFILING is not set
++# CONFIG_KPROBES is not set
++
++#
++# Kernel hacking
++#
++CONFIG_TRACE_IRQFLAGS_SUPPORT=y
++# CONFIG_PRINTK_TIME is not set
++CONFIG_MAGIC_SYSRQ=y
++CONFIG_UNUSED_SYMBOLS=y
++CONFIG_DEBUG_KERNEL=y
++CONFIG_LOG_BUF_SHIFT=14
++CONFIG_DETECT_SOFTLOCKUP=y
++# CONFIG_SCHEDSTATS is not set
++# CONFIG_DEBUG_SLAB is not set
++# CONFIG_DEBUG_RT_MUTEXES is not set
++# CONFIG_RT_MUTEX_TESTER is not set
++# CONFIG_DEBUG_SPINLOCK is not set
++# CONFIG_DEBUG_MUTEXES is not set
++# CONFIG_DEBUG_RWSEMS is not set
++# CONFIG_DEBUG_LOCK_ALLOC is not set
++# CONFIG_PROVE_LOCKING is not set
++# CONFIG_DEBUG_SPINLOCK_SLEEP is not set
++# CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set
++# CONFIG_DEBUG_KOBJECT is not set
++CONFIG_DEBUG_INFO=y
++# CONFIG_DEBUG_FS is not set
++# CONFIG_DEBUG_VM is not set
++CONFIG_FRAME_POINTER=y
++# CONFIG_UNWIND_INFO is not set
++CONFIG_FORCED_INLINING=y
++# CONFIG_RCU_TORTURE_TEST is not set
++# CONFIG_DEBUG_RODATA is not set
++# CONFIG_DEBUG_STACKOVERFLOW is not set
++# CONFIG_DEBUG_STACK_USAGE is not set
++
++#
++# Security options
++#
++# CONFIG_KEYS is not set
++# CONFIG_SECURITY is not set
++
++#
++# Cryptographic options
++#
++CONFIG_CRYPTO=y
++# CONFIG_CRYPTO_HMAC is not set
++# CONFIG_CRYPTO_NULL is not set
++# CONFIG_CRYPTO_MD4 is not set
++CONFIG_CRYPTO_MD5=m
++# CONFIG_CRYPTO_SHA1 is not set
++# CONFIG_CRYPTO_SHA256 is not set
++# CONFIG_CRYPTO_SHA512 is not set
++# CONFIG_CRYPTO_WP512 is not set
++# CONFIG_CRYPTO_TGR192 is not set
++# CONFIG_CRYPTO_DES is not set
++# CONFIG_CRYPTO_BLOWFISH is not set
++# CONFIG_CRYPTO_TWOFISH is not set
++# CONFIG_CRYPTO_SERPENT is not set
++# CONFIG_CRYPTO_AES is not set
++# CONFIG_CRYPTO_AES_X86_64 is not set
++# CONFIG_CRYPTO_CAST5 is not set
++# CONFIG_CRYPTO_CAST6 is not set
++# CONFIG_CRYPTO_TEA is not set
++# CONFIG_CRYPTO_ARC4 is not set
++# CONFIG_CRYPTO_KHAZAD is not set
++# CONFIG_CRYPTO_ANUBIS is not set
++# CONFIG_CRYPTO_DEFLATE is not set
++# CONFIG_CRYPTO_MICHAEL_MIC is not set
++CONFIG_CRYPTO_CRC32C=m
++# CONFIG_CRYPTO_TEST is not set
++
++#
++# Hardware crypto devices
++#
++CONFIG_XEN=y
++CONFIG_XEN_INTERFACE_VERSION=0x00030207
++
++#
++# XEN
++#
++# CONFIG_XEN_PRIVILEGED_GUEST is not set
++CONFIG_XEN_UNPRIVILEGED_GUEST=y
++CONFIG_XEN_PRIVCMD=y
++CONFIG_XEN_XENBUS_DEV=y
++# CONFIG_XEN_BACKEND is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_UTIL is not set
++CONFIG_XEN_BLKDEV_FRONTEND=y
++CONFIG_XEN_NETDEV_FRONTEND=y
++CONFIG_XEN_SCSI_FRONTEND=m
++CONFIG_XEN_GRANT_DEV=y
++# CONFIG_XEN_NETDEV_ACCEL_SFC_FRONTEND is not set
++CONFIG_XEN_SCRUB_PAGES=y
++CONFIG_XEN_DISABLE_SERIAL=y
++CONFIG_XEN_SYSFS=y
++CONFIG_XEN_COMPAT_030002_AND_LATER=y
++# CONFIG_XEN_COMPAT_030004_AND_LATER is not set
++# CONFIG_XEN_COMPAT_030100_AND_LATER is not set
++# CONFIG_XEN_COMPAT_LATEST_ONLY is not set
++CONFIG_XEN_COMPAT=0x030002
++CONFIG_HAVE_IRQ_IGNORE_UNHANDLED=y
++CONFIG_NO_IDLE_HZ=y
++CONFIG_XEN_SMPBOOT=y
++CONFIG_XEN_BALLOON=y
++CONFIG_XEN_DEVMEM=y
++
++#
++# Library routines
++#
++# CONFIG_CRC_CCITT is not set
++# CONFIG_CRC16 is not set
++# CONFIG_CRC32 is not set
++CONFIG_LIBCRC32C=m
++CONFIG_ZLIB_INFLATE=y
++CONFIG_PLIST=y
+diff -r d894e36cfc30 -r 0aa021803deb buildconfigs/linux-defconfig_xen_ia64
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/buildconfigs/linux-defconfig_xen_ia64 Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1703 @@
++#
++# Automatically generated make config: don't edit
++# Linux kernel version: 2.6.18.8
++# Tue Feb 19 11:20:00 2008
++#
++CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config"
++
++#
++# Code maturity level options
++#
++CONFIG_EXPERIMENTAL=y
++CONFIG_LOCK_KERNEL=y
++CONFIG_INIT_ENV_ARG_LIMIT=32
++
++#
++# General setup
++#
++CONFIG_LOCALVERSION=""
++CONFIG_LOCALVERSION_AUTO=y
++CONFIG_SWAP=y
++CONFIG_SYSVIPC=y
++CONFIG_POSIX_MQUEUE=y
++CONFIG_BSD_PROCESS_ACCT=y
++# CONFIG_BSD_PROCESS_ACCT_V3 is not set
++# CONFIG_TASKSTATS is not set
++# CONFIG_AUDIT is not set
++CONFIG_IKCONFIG=y
++CONFIG_IKCONFIG_PROC=y
++# CONFIG_CPUSETS is not set
++# CONFIG_RELAY is not set
++CONFIG_INITRAMFS_SOURCE=""
++CONFIG_CC_OPTIMIZE_FOR_SIZE=y
++# CONFIG_EMBEDDED is not set
++CONFIG_SYSCTL=y
++CONFIG_KALLSYMS=y
++CONFIG_KALLSYMS_ALL=y
++CONFIG_KALLSYMS_EXTRA_PASS=y
++CONFIG_HOTPLUG=y
++CONFIG_PRINTK=y
++CONFIG_BUG=y
++CONFIG_ELF_CORE=y
++CONFIG_BASE_FULL=y
++CONFIG_FUTEX=y
++CONFIG_EPOLL=y
++CONFIG_SHMEM=y
++CONFIG_SLAB=y
++CONFIG_VM_EVENT_COUNTERS=y
++CONFIG_RT_MUTEXES=y
++# CONFIG_TINY_SHMEM is not set
++CONFIG_BASE_SMALL=0
++# CONFIG_SLOB is not set
++
++#
++# Loadable module support
++#
++CONFIG_MODULES=y
++CONFIG_MODULE_UNLOAD=y
++# CONFIG_MODULE_FORCE_UNLOAD is not set
++CONFIG_MODVERSIONS=y
++CONFIG_MODULE_SRCVERSION_ALL=y
++CONFIG_KMOD=y
++CONFIG_STOP_MACHINE=y
++
++#
++# Block layer
++#
++# CONFIG_BLK_DEV_IO_TRACE is not set
++
++#
++# IO Schedulers
++#
++CONFIG_IOSCHED_NOOP=y
++CONFIG_IOSCHED_AS=y
++CONFIG_IOSCHED_DEADLINE=y
++CONFIG_IOSCHED_CFQ=y
++CONFIG_DEFAULT_AS=y
++# CONFIG_DEFAULT_DEADLINE is not set
++# CONFIG_DEFAULT_CFQ is not set
++# CONFIG_DEFAULT_NOOP is not set
++CONFIG_DEFAULT_IOSCHED="anticipatory"
++
++#
++# Processor type and features
++#
++CONFIG_IA64=y
++CONFIG_64BIT=y
++CONFIG_MMU=y
++CONFIG_SWIOTLB=y
++CONFIG_RWSEM_XCHGADD_ALGORITHM=y
++CONFIG_GENERIC_FIND_NEXT_BIT=y
++CONFIG_GENERIC_CALIBRATE_DELAY=y
++CONFIG_TIME_INTERPOLATION=y
++CONFIG_DMI=y
++CONFIG_EFI=y
++CONFIG_GENERIC_IOMAP=y
++CONFIG_XEN=y
++CONFIG_XEN_IA64_EXPOSE_P2M=y
++CONFIG_XEN_IA64_EXPOSE_P2M_USE_DTR=y
++CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER=y
++CONFIG_DMA_IS_DMA32=y
++CONFIG_AUDIT_ARCH=y
++CONFIG_IA64_GENERIC=y
++# CONFIG_IA64_DIG is not set
++# CONFIG_IA64_HP_ZX1 is not set
++# CONFIG_IA64_HP_ZX1_SWIOTLB is not set
++# CONFIG_IA64_SGI_SN2 is not set
++# CONFIG_IA64_HP_SIM is not set
++# CONFIG_IA64_XEN is not set
++# CONFIG_ITANIUM is not set
++CONFIG_MCKINLEY=y
++# CONFIG_IA64_PAGE_SIZE_4KB is not set
++# CONFIG_IA64_PAGE_SIZE_8KB is not set
++CONFIG_IA64_PAGE_SIZE_16KB=y
++# CONFIG_IA64_PAGE_SIZE_64KB is not set
++CONFIG_PGTABLE_3=y
++# CONFIG_PGTABLE_4 is not set
++CONFIG_HZ_100=y
++# CONFIG_HZ_250 is not set
++# CONFIG_HZ_1000 is not set
++CONFIG_HZ=100
++CONFIG_IA64_L1_CACHE_SHIFT=7
++CONFIG_IA64_CYCLONE=y
++CONFIG_IOSAPIC=y
++# CONFIG_IA64_SGI_SN_XP is not set
++CONFIG_FORCE_MAX_ZONEORDER=11
++CONFIG_SMP=y
++CONFIG_NR_CPUS=16
++CONFIG_HOTPLUG_CPU=y
++CONFIG_ARCH_ENABLE_MEMORY_HOTPLUG=y
++# CONFIG_SCHED_SMT is not set
++# CONFIG_PERMIT_BSP_REMOVE is not set
++# CONFIG_PREEMPT is not set
++CONFIG_SELECT_MEMORY_MODEL=y
++# CONFIG_FLATMEM_MANUAL is not set
++CONFIG_DISCONTIGMEM_MANUAL=y
++# CONFIG_SPARSEMEM_MANUAL is not set
++CONFIG_DISCONTIGMEM=y
++CONFIG_FLAT_NODE_MEM_MAP=y
++CONFIG_NEED_MULTIPLE_NODES=y
++# CONFIG_SPARSEMEM_STATIC is not set
++CONFIG_SPLIT_PTLOCK_CPUS=4
++# CONFIG_MIGRATION is not set
++CONFIG_RESOURCES_64BIT=y
++CONFIG_ARCH_SELECT_MEMORY_MODEL=y
++CONFIG_ARCH_DISCONTIGMEM_ENABLE=y
++CONFIG_ARCH_FLATMEM_ENABLE=y
++CONFIG_ARCH_SPARSEMEM_ENABLE=y
++CONFIG_ARCH_DISCONTIGMEM_DEFAULT=y
++CONFIG_NUMA=y
++CONFIG_NODES_SHIFT=10
++CONFIG_VIRTUAL_MEM_MAP=y
++CONFIG_HOLES_IN_ZONE=y
++CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID=y
++CONFIG_HAVE_ARCH_NODEDATA_EXTENSION=y
++# CONFIG_IA32_SUPPORT is not set
++# CONFIG_IA64_MCA_RECOVERY is not set
++CONFIG_PERFMON=y
++CONFIG_IA64_PALINFO=y
++CONFIG_SGI_SN=y
++
++#
++# SN Devices
++#
++# CONFIG_SGI_IOC3 is not set
++CONFIG_KEXEC=y
++
++#
++# Firmware Drivers
++#
++CONFIG_EFI_VARS=y
++CONFIG_EFI_PCDP=y
++CONFIG_BINFMT_ELF=y
++CONFIG_BINFMT_MISC=y
++
++#
++# Power management and ACPI
++#
++CONFIG_PM=y
++CONFIG_PM_LEGACY=y
++# CONFIG_PM_DEBUG is not set
++
++#
++# ACPI (Advanced Configuration and Power Interface) Support
++#
++CONFIG_ACPI=y
++CONFIG_ACPI_BUTTON=y
++CONFIG_ACPI_FAN=y
++# CONFIG_ACPI_DOCK is not set
++CONFIG_ACPI_PROCESSOR=y
++CONFIG_ACPI_HOTPLUG_CPU=y
++CONFIG_ACPI_THERMAL=y
++CONFIG_ACPI_NUMA=y
++CONFIG_ACPI_BLACKLIST_YEAR=0
++# CONFIG_ACPI_DEBUG is not set
++CONFIG_ACPI_EC=y
++CONFIG_ACPI_POWER=y
++CONFIG_ACPI_SYSTEM=y
++CONFIG_ACPI_CONTAINER=y
++
++#
++# CPU Frequency scaling
++#
++# CONFIG_CPU_FREQ is not set
++
++#
++# Bus options (PCI, PCMCIA)
++#
++CONFIG_PCI=y
++CONFIG_PCI_DOMAINS=y
++CONFIG_XEN_PCIDEV_FRONTEND=y
++# CONFIG_XEN_PCIDEV_FE_DEBUG is not set
++# CONFIG_PCIEPORTBUS is not set
++# CONFIG_PCI_DEBUG is not set
+
-+ .code32
-+ .text
-+ .section .text.vsyscall,"ax"
-+ .globl __kernel_vsyscall
-+ .type __kernel_vsyscall, at function
-+__kernel_vsyscall:
-+.LSTART_vsyscall:
-+ int $0x80
-+ ret
-+.LEND_vsyscall:
-+ .size __kernel_vsyscall,.-.LSTART_vsyscall
-+ .previous
++#
++# PCI Hotplug Support
++#
++CONFIG_HOTPLUG_PCI=y
++# CONFIG_HOTPLUG_PCI_FAKE is not set
++CONFIG_HOTPLUG_PCI_ACPI=y
++# CONFIG_HOTPLUG_PCI_ACPI_IBM is not set
++# CONFIG_HOTPLUG_PCI_CPCI is not set
++# CONFIG_HOTPLUG_PCI_SHPC is not set
++# CONFIG_HOTPLUG_PCI_SGI is not set
+
-+ .section .eh_frame,"a", at progbits
-+.LSTARTFRAME:
-+ .long .LENDCIE-.LSTARTCIE
-+.LSTARTCIE:
-+ .long 0 /* CIE ID */
-+ .byte 1 /* Version number */
-+ .string "zR" /* NUL-terminated augmentation string */
-+ .uleb128 1 /* Code alignment factor */
-+ .sleb128 -4 /* Data alignment factor */
-+ .byte 8 /* Return address register column */
-+ .uleb128 1 /* Augmentation value length */
-+ .byte 0x1b /* DW_EH_PE_pcrel|DW_EH_PE_sdata4. */
-+ .byte 0x0c /* DW_CFA_def_cfa */
-+ .uleb128 4
-+ .uleb128 4
-+ .byte 0x88 /* DW_CFA_offset, column 0x8 */
-+ .uleb128 1
-+ .align 4
-+.LENDCIE:
++#
++# PCCARD (PCMCIA/CardBus) support
++#
++# CONFIG_PCCARD is not set
+
-+ .long .LENDFDE1-.LSTARTFDE1 /* Length FDE */
-+.LSTARTFDE1:
-+ .long .LSTARTFDE1-.LSTARTFRAME /* CIE pointer */
-+ .long .LSTART_vsyscall-. /* PC-relative start address */
-+ .long .LEND_vsyscall-.LSTART_vsyscall
-+ .uleb128 0 /* Augmentation length */
-+ .align 4
-+.LENDFDE1:
-+
-+/*
-+ * Get the common code for the sigreturn entry points.
-+ */
-+#define SYSCALL_ENTER_KERNEL int $0x80
-+#include "vsyscall-sigreturn.S"
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/ia32/vsyscall-sigreturn.S tmp-linux-2.6-xen.patch/arch/x86_64/ia32/vsyscall-sigreturn.S
---- pristine-linux-2.6.18.2/arch/x86_64/ia32/vsyscall-sigreturn.S 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/ia32/vsyscall-sigreturn.S 2007-07-30 16:35:11.000000000 +0200
-@@ -139,5 +139,5 @@ __kernel_rt_sigreturn:
- .align 4
- .LENDFDE3:
-
--#include "../../i386/kernel/vsyscall-note.S"
-+#include <vsyscall-note.S>
-
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/Kconfig tmp-linux-2.6-xen.patch/arch/x86_64/Kconfig
---- pristine-linux-2.6.18.2/arch/x86_64/Kconfig 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/Kconfig 2007-07-30 16:35:11.000000000 +0200
-@@ -135,6 +135,22 @@ config GENERIC_CPU
-
- endchoice
-
-+config X86_64_XEN
-+ bool "Enable Xen compatible kernel"
-+ select SWIOTLB
-+ help
-+ This option will compile a kernel compatible with Xen hypervisor
++#
++# Networking
++#
++CONFIG_NET=y
+
-+config X86_NO_TSS
-+ bool
-+ depends on X86_64_XEN
-+ default y
++#
++# Networking options
++#
++# CONFIG_NETDEBUG is not set
++CONFIG_PACKET=y
++# CONFIG_PACKET_MMAP is not set
++CONFIG_UNIX=y
++CONFIG_XFRM=y
++# CONFIG_XFRM_USER is not set
++# CONFIG_NET_KEY is not set
++CONFIG_INET=y
++CONFIG_IP_MULTICAST=y
++# CONFIG_IP_ADVANCED_ROUTER is not set
++CONFIG_IP_FIB_HASH=y
++CONFIG_IP_PNP=y
++CONFIG_IP_PNP_DHCP=y
++# CONFIG_IP_PNP_BOOTP is not set
++# CONFIG_IP_PNP_RARP is not set
++# CONFIG_NET_IPIP is not set
++# CONFIG_NET_IPGRE is not set
++# CONFIG_IP_MROUTE is not set
++CONFIG_ARPD=y
++CONFIG_SYN_COOKIES=y
++# CONFIG_INET_AH is not set
++# CONFIG_INET_ESP is not set
++# CONFIG_INET_IPCOMP is not set
++# CONFIG_INET_XFRM_TUNNEL is not set
++# CONFIG_INET_TUNNEL is not set
++CONFIG_INET_XFRM_MODE_TRANSPORT=y
++CONFIG_INET_XFRM_MODE_TUNNEL=y
++CONFIG_INET_DIAG=y
++CONFIG_INET_TCP_DIAG=y
++# CONFIG_TCP_CONG_ADVANCED is not set
++CONFIG_TCP_CONG_BIC=y
+
-+config X86_NO_IDT
-+ bool
-+ depends on X86_64_XEN
-+ default y
++#
++# IP: Virtual Server Configuration
++#
++# CONFIG_IP_VS is not set
++# CONFIG_IPV6 is not set
++# CONFIG_INET6_XFRM_TUNNEL is not set
++# CONFIG_INET6_TUNNEL is not set
++# CONFIG_NETWORK_SECMARK is not set
++CONFIG_NETFILTER=y
++# CONFIG_NETFILTER_DEBUG is not set
++CONFIG_BRIDGE_NETFILTER=y
+
- #
- # Define implied options from the CPU selection here
- #
-@@ -155,6 +171,7 @@ config X86_INTERNODE_CACHE_BYTES
-
- config X86_TSC
- bool
-+ depends on !X86_64_XEN
- default y
-
- config X86_GOOD_APIC
-@@ -197,7 +214,7 @@ config X86_CPUID
-
- config X86_HT
- bool
-- depends on SMP && !MK8
-+ depends on SMP && !MK8 && !X86_64_XEN
- default y
-
- config MATH_EMULATION
-@@ -211,14 +228,22 @@ config EISA
-
- config X86_IO_APIC
- bool
-+ depends !XEN_UNPRIVILEGED_GUEST
- default y
-
-+config X86_XEN_GENAPIC
-+ bool
-+ depends X86_64_XEN
-+ default XEN_PRIVILEGED_GUEST || SMP
++#
++# Core Netfilter Configuration
++#
++# CONFIG_NETFILTER_NETLINK is not set
++# CONFIG_NF_CONNTRACK is not set
++# CONFIG_NETFILTER_XTABLES is not set
+
- config X86_LOCAL_APIC
- bool
-+ depends !XEN_UNPRIVILEGED_GUEST
- default y
-
- config MTRR
- bool "MTRR (Memory Type Range Register) support"
-+ depends on !XEN_UNPRIVILEGED_GUEST
- ---help---
- On Intel P6 family processors (Pentium Pro, Pentium II and later)
- the Memory Type Range Registers (MTRRs) may be used to control
-@@ -259,7 +284,7 @@ config SMP
-
- config SCHED_SMT
- bool "SMT (Hyperthreading) scheduler support"
-- depends on SMP
-+ depends on SMP && !X86_64_XEN
- default n
- help
- SMT scheduler support improves the CPU scheduler's decision making
-@@ -269,7 +294,7 @@ config SCHED_SMT
-
- config SCHED_MC
- bool "Multi-core scheduler support"
-- depends on SMP
-+ depends on SMP && !X86_64_XEN
- default y
- help
- Multi-core scheduler support improves the CPU scheduler's decision
-@@ -280,7 +305,7 @@ source "kernel/Kconfig.preempt"
-
- config NUMA
- bool "Non Uniform Memory Access (NUMA) Support"
-- depends on SMP
-+ depends on SMP && !X86_64_XEN
- help
- Enable NUMA (Non Uniform Memory Access) support. The kernel
- will try to allocate memory used by a CPU on the local memory
-@@ -341,7 +366,7 @@ config ARCH_DISCONTIGMEM_DEFAULT
-
- config ARCH_SPARSEMEM_ENABLE
- def_bool y
-- depends on (NUMA || EXPERIMENTAL)
-+ depends on (NUMA || EXPERIMENTAL) && !X86_64_XEN
-
- config ARCH_MEMORY_PROBE
- def_bool y
-@@ -365,6 +390,7 @@ config NR_CPUS
- int "Maximum number of CPUs (2-256)"
- range 2 255
- depends on SMP
-+ default "16" if X86_64_XEN
- default "8"
- help
- This allows you to specify the maximum number of CPUs which this
-@@ -387,6 +413,7 @@ config ARCH_ENABLE_MEMORY_HOTPLUG
-
- config HPET_TIMER
- bool
-+ depends on !X86_64_XEN
- default y
- help
- Use the IA-PC HPET (High Precision Event Timer) to manage
-@@ -407,7 +434,7 @@ config IOMMU
- default y
- select SWIOTLB
- select AGP
-- depends on PCI
-+ depends on PCI && !X86_64_XEN
- help
- Support for full DMA access of devices with 32bit memory access only
- on systems with more than 3GB. This is usually needed for USB,
-@@ -423,7 +450,7 @@ config CALGARY_IOMMU
- bool "IBM Calgary IOMMU support"
- default y
- select SWIOTLB
-- depends on PCI && EXPERIMENTAL
-+ depends on PCI && !X86_64_XEN && EXPERIMENTAL
- help
- Support for hardware IOMMUs in IBM's xSeries x366 and x460
- systems. Needed to run systems with more than 3GB of memory
-@@ -444,6 +471,7 @@ config SWIOTLB
-
- config X86_MCE
- bool "Machine check support" if EMBEDDED
-+ depends on !X86_64_XEN
- default y
- help
- Include a machine check error handler to report hardware errors.
-@@ -469,7 +497,7 @@ config X86_MCE_AMD
-
- config KEXEC
- bool "kexec system call (EXPERIMENTAL)"
-- depends on EXPERIMENTAL
-+ depends on EXPERIMENTAL && !XEN_UNPRIVILEGED_GUEST
- help
- kexec is a system call that implements the ability to shutdown your
- current kernel, and to start another kernel. It is like a reboot
-@@ -564,8 +592,11 @@ config GENERIC_PENDING_IRQ
- default y
-
- menu "Power management options"
-+ depends on !XEN_UNPRIVILEGED_GUEST
-
-+if !X86_64_XEN
- source kernel/power/Kconfig
-+endif
-
- source "drivers/acpi/Kconfig"
-
-@@ -588,6 +619,21 @@ config PCI_MMCONFIG
- bool "Support mmconfig PCI config space access"
- depends on PCI && ACPI
-
-+config XEN_PCIDEV_FRONTEND
-+ bool "Xen PCI Frontend"
-+ depends on PCI && X86_64_XEN
-+ default y
-+ help
-+ The PCI device frontend driver allows the kernel to import arbitrary
-+ PCI devices from a PCI backend to support PCI driver domains.
++#
++# IP: Netfilter Configuration
++#
++# CONFIG_IP_NF_CONNTRACK is not set
++# CONFIG_IP_NF_QUEUE is not set
+
-+config XEN_PCIDEV_FE_DEBUG
-+ bool "Xen PCI Frontend Debugging"
-+ depends on XEN_PCIDEV_FRONTEND
-+ default n
-+ help
-+ Enables some debug statements within the PCI Frontend.
++#
++# Bridge: Netfilter Configuration
++#
++# CONFIG_BRIDGE_NF_EBTABLES is not set
+
- source "drivers/pci/pcie/Kconfig"
-
- source "drivers/pci/Kconfig"
-@@ -658,4 +704,6 @@ source "security/Kconfig"
-
- source "crypto/Kconfig"
-
-+source "drivers/xen/Kconfig"
++#
++# DCCP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_DCCP is not set
+
- source "lib/Kconfig"
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/acpi/Makefile tmp-linux-2.6-xen.patch/arch/x86_64/kernel/acpi/Makefile
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/acpi/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/acpi/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -7,3 +7,4 @@ obj-y += processor.o
- processor-y := ../../../i386/kernel/acpi/processor.o ../../../i386/kernel/acpi/cstate.o
- endif
-
-+boot-$(CONFIG_XEN) := ../../../i386/kernel/acpi/boot-xen.o
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/apic-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/apic-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/apic-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/apic-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,197 @@
-+/*
-+ * Local APIC handling, local APIC timers
-+ *
-+ * (c) 1999, 2000 Ingo Molnar <mingo at redhat.com>
-+ *
-+ * Fixes
-+ * Maciej W. Rozycki : Bits for genuine 82489DX APICs;
-+ * thanks to Eric Gilmore
-+ * and Rolf G. Tews
-+ * for testing these extensively.
-+ * Maciej W. Rozycki : Various updates and fixes.
-+ * Mikael Pettersson : Power Management for UP-APIC.
-+ * Pavel Machek and
-+ * Mikael Pettersson : PM converted to driver model.
-+ */
++#
++# SCTP Configuration (EXPERIMENTAL)
++#
++# CONFIG_IP_SCTP is not set
+
-+#include <linux/init.h>
++#
++# TIPC Configuration (EXPERIMENTAL)
++#
++# CONFIG_TIPC is not set
++# CONFIG_ATM is not set
++CONFIG_BRIDGE=y
++# CONFIG_VLAN_8021Q is not set
++# CONFIG_DECNET is not set
++CONFIG_LLC=y
++# CONFIG_LLC2 is not set
++# CONFIG_IPX is not set
++# CONFIG_ATALK is not set
++# CONFIG_X25 is not set
++# CONFIG_LAPB is not set
++# CONFIG_ECONET is not set
++# CONFIG_WAN_ROUTER is not set
+
-+#include <linux/mm.h>
-+#include <linux/delay.h>
-+#include <linux/bootmem.h>
-+#include <linux/smp_lock.h>
-+#include <linux/interrupt.h>
-+#include <linux/mc146818rtc.h>
-+#include <linux/kernel_stat.h>
-+#include <linux/sysdev.h>
-+#include <linux/module.h>
++#
++# QoS and/or fair queueing
++#
++# CONFIG_NET_SCHED is not set
+
-+#include <asm/atomic.h>
-+#include <asm/smp.h>
-+#include <asm/mtrr.h>
-+#include <asm/mpspec.h>
-+#include <asm/desc.h>
-+#include <asm/arch_hooks.h>
-+#include <asm/hpet.h>
-+#include <asm/idle.h>
++#
++# Network testing
++#
++# CONFIG_NET_PKTGEN is not set
++# CONFIG_HAMRADIO is not set
++# CONFIG_IRDA is not set
++# CONFIG_BT is not set
++# CONFIG_IEEE80211 is not set
+
-+int apic_verbosity;
++#
++# Device Drivers
++#
+
-+/*
-+ * 'what should we do if we get a hw irq event on an illegal vector'.
-+ * each architecture has to answer this themselves.
-+ */
-+void ack_bad_irq(unsigned int irq)
-+{
-+ printk("unexpected IRQ trap at vector %02x\n", irq);
-+ /*
-+ * Currently unexpected vectors happen only on SMP and APIC.
-+ * We _must_ ack these because every local APIC has only N
-+ * irq slots per priority level, and a 'hanging, unacked' IRQ
-+ * holds up an irq slot - in excessive cases (when multiple
-+ * unexpected vectors occur) that might lock up the APIC
-+ * completely.
-+ * But don't ack when the APIC is disabled. -AK
-+ */
-+ if (!disable_apic)
-+ ack_APIC_irq();
-+}
++#
++# Generic Driver Options
++#
++CONFIG_STANDALONE=y
++CONFIG_PREVENT_FIRMWARE_BUILD=y
++CONFIG_FW_LOADER=y
++# CONFIG_DEBUG_DRIVER is not set
++CONFIG_SYS_HYPERVISOR=y
+
-+int setup_profiling_timer(unsigned int multiplier)
-+{
-+ return -EINVAL;
-+}
++#
++# Connector - unified userspace <-> kernelspace linker
++#
++# CONFIG_CONNECTOR is not set
+
-+void smp_local_timer_interrupt(struct pt_regs *regs)
-+{
-+ profile_tick(CPU_PROFILING, regs);
-+#ifndef CONFIG_XEN
-+#ifdef CONFIG_SMP
-+ update_process_times(user_mode(regs));
-+#endif
-+#endif
-+ /*
-+ * We take the 'long' return path, and there every subsystem
-+ * grabs the appropriate locks (kernel lock/ irq lock).
-+ *
-+ * we might want to decouple profiling from the 'long path',
-+ * and do the profiling totally in assembly.
-+ *
-+ * Currently this isn't too much of an issue (performance wise),
-+ * we can take more than 100K local irqs per second on a 100 MHz P5.
-+ */
-+}
++#
++# Memory Technology Devices (MTD)
++#
++# CONFIG_MTD is not set
+
-+/*
-+ * Local APIC timer interrupt. This is the most natural way for doing
-+ * local interrupts, but local timer interrupts can be emulated by
-+ * broadcast interrupts too. [in case the hw doesn't support APIC timers]
-+ *
-+ * [ if a single-CPU system runs an SMP kernel then we call the local
-+ * interrupt as well. Thus we cannot inline the local irq ... ]
-+ */
-+void smp_apic_timer_interrupt(struct pt_regs *regs)
-+{
-+ /*
-+ * the NMI deadlock-detector uses this.
-+ */
-+ add_pda(apic_timer_irqs, 1);
++#
++# Parallel port support
++#
++# CONFIG_PARPORT is not set
+
-+ /*
-+ * NOTE! We'd better ACK the irq immediately,
-+ * because timer handling can be slow.
-+ */
-+ ack_APIC_irq();
-+ /*
-+ * update_process_times() expects us to have done irq_enter().
-+ * Besides, if we don't timer interrupts ignore the global
-+ * interrupt lock, which is the WrongThing (tm) to do.
-+ */
-+ exit_idle();
-+ irq_enter();
-+ smp_local_timer_interrupt(regs);
-+ irq_exit();
-+}
++#
++# Plug and Play support
++#
++# CONFIG_PNP is not set
+
-+/*
-+ * This interrupt should _never_ happen with our APIC/SMP architecture
-+ */
-+asmlinkage void smp_spurious_interrupt(void)
-+{
-+ unsigned int v;
-+ exit_idle();
-+ irq_enter();
-+ /*
-+ * Check if this really is a spurious interrupt and ACK it
-+ * if it is a vectored one. Just in case...
-+ * Spurious interrupts should not be ACKed.
-+ */
-+ v = apic_read(APIC_ISR + ((SPURIOUS_APIC_VECTOR & ~0x1f) >> 1));
-+ if (v & (1 << (SPURIOUS_APIC_VECTOR & 0x1f)))
-+ ack_APIC_irq();
++#
++# Block devices
++#
++# CONFIG_BLK_CPQ_DA is not set
++CONFIG_BLK_CPQ_CISS_DA=y
++# CONFIG_CISS_SCSI_TAPE is not set
++# CONFIG_BLK_DEV_DAC960 is not set
++# CONFIG_BLK_DEV_UMEM is not set
++# CONFIG_BLK_DEV_COW_COMMON is not set
++CONFIG_BLK_DEV_LOOP=y
++CONFIG_BLK_DEV_CRYPTOLOOP=y
++CONFIG_BLK_DEV_NBD=m
++# CONFIG_BLK_DEV_SX8 is not set
++# CONFIG_BLK_DEV_UB is not set
++CONFIG_BLK_DEV_RAM=y
++CONFIG_BLK_DEV_RAM_COUNT=16
++CONFIG_BLK_DEV_RAM_SIZE=4096
++CONFIG_BLK_DEV_RAM_BLOCKSIZE=1024
++CONFIG_BLK_DEV_INITRD=y
++# CONFIG_CDROM_PKTCDVD is not set
++# CONFIG_ATA_OVER_ETH is not set
+
-+#if 0
-+ static unsigned long last_warning;
-+ static unsigned long skipped;
++#
++# ATA/ATAPI/MFM/RLL support
++#
++CONFIG_IDE=y
++CONFIG_IDE_MAX_HWIFS=4
++CONFIG_BLK_DEV_IDE=y
+
-+ /* see sw-dev-man vol 3, chapter 7.4.13.5 */
-+ if (time_before(last_warning+30*HZ,jiffies)) {
-+ printk(KERN_INFO "spurious APIC interrupt on CPU#%d, %ld skipped.\n",
-+ smp_processor_id(), skipped);
-+ last_warning = jiffies;
-+ skipped = 0;
-+ } else {
-+ skipped++;
-+ }
-+#endif
-+ irq_exit();
-+}
++#
++# Please see Documentation/ide.txt for help/info on IDE drives
++#
++# CONFIG_BLK_DEV_IDE_SATA is not set
++CONFIG_BLK_DEV_IDEDISK=y
++# CONFIG_IDEDISK_MULTI_MODE is not set
++CONFIG_BLK_DEV_IDECD=y
++# CONFIG_BLK_DEV_IDETAPE is not set
++CONFIG_BLK_DEV_IDEFLOPPY=y
++CONFIG_BLK_DEV_IDESCSI=y
++# CONFIG_IDE_TASK_IOCTL is not set
+
-+/*
-+ * This interrupt should never happen with our APIC/SMP architecture
-+ */
++#
++# IDE chipset support/bugfixes
++#
++# CONFIG_IDE_GENERIC is not set
++CONFIG_BLK_DEV_IDEPCI=y
++# CONFIG_IDEPCI_SHARE_IRQ is not set
++# CONFIG_BLK_DEV_OFFBOARD is not set
++CONFIG_BLK_DEV_GENERIC=y
++# CONFIG_BLK_DEV_OPTI621 is not set
++CONFIG_BLK_DEV_IDEDMA_PCI=y
++# CONFIG_BLK_DEV_IDEDMA_FORCED is not set
++CONFIG_IDEDMA_PCI_AUTO=y
++# CONFIG_IDEDMA_ONLYDISK is not set
++# CONFIG_BLK_DEV_AEC62XX is not set
++# CONFIG_BLK_DEV_ALI15X3 is not set
++# CONFIG_BLK_DEV_AMD74XX is not set
++CONFIG_BLK_DEV_CMD64X=y
++# CONFIG_BLK_DEV_TRIFLEX is not set
++# CONFIG_BLK_DEV_CY82C693 is not set
++# CONFIG_BLK_DEV_CS5520 is not set
++# CONFIG_BLK_DEV_CS5530 is not set
++# CONFIG_BLK_DEV_HPT34X is not set
++# CONFIG_BLK_DEV_HPT366 is not set
++# CONFIG_BLK_DEV_SC1200 is not set
++CONFIG_BLK_DEV_PIIX=y
++# CONFIG_BLK_DEV_IT821X is not set
++# CONFIG_BLK_DEV_NS87415 is not set
++# CONFIG_BLK_DEV_PDC202XX_OLD is not set
++# CONFIG_BLK_DEV_PDC202XX_NEW is not set
++# CONFIG_BLK_DEV_SVWKS is not set
++# CONFIG_BLK_DEV_SIIMAGE is not set
++# CONFIG_BLK_DEV_SLC90E66 is not set
++# CONFIG_BLK_DEV_TRM290 is not set
++# CONFIG_BLK_DEV_VIA82CXXX is not set
++# CONFIG_IDE_ARM is not set
++CONFIG_BLK_DEV_IDEDMA=y
++# CONFIG_IDEDMA_IVB is not set
++CONFIG_IDEDMA_AUTO=y
++# CONFIG_BLK_DEV_HD is not set
+
-+asmlinkage void smp_error_interrupt(void)
-+{
-+ unsigned int v, v1;
++#
++# SCSI device support
++#
++# CONFIG_RAID_ATTRS is not set
++CONFIG_SCSI=y
++CONFIG_SCSI_PROC_FS=y
+
-+ exit_idle();
-+ irq_enter();
-+ /* First tickle the hardware, only then report what went on. -- REW */
-+ v = apic_read(APIC_ESR);
-+ apic_write(APIC_ESR, 0);
-+ v1 = apic_read(APIC_ESR);
-+ ack_APIC_irq();
-+ atomic_inc(&irq_err_count);
++#
++# SCSI support type (disk, tape, CD-ROM)
++#
++CONFIG_BLK_DEV_SD=y
++CONFIG_CHR_DEV_ST=y
++CONFIG_CHR_DEV_OSST=y
++CONFIG_BLK_DEV_SR=y
++CONFIG_BLK_DEV_SR_VENDOR=y
++CONFIG_CHR_DEV_SG=y
++# CONFIG_CHR_DEV_SCH is not set
+
-+ /* Here is what the APIC error bits mean:
-+ 0: Send CS error
-+ 1: Receive CS error
-+ 2: Send accept error
-+ 3: Receive accept error
-+ 4: Reserved
-+ 5: Send illegal vector
-+ 6: Received illegal vector
-+ 7: Illegal register address
-+ */
-+ printk (KERN_DEBUG "APIC error on CPU%d: %02x(%02x)\n",
-+ smp_processor_id(), v , v1);
-+ irq_exit();
-+}
++#
++# Some SCSI devices (e.g. CD jukebox) support multiple LUNs
++#
++CONFIG_SCSI_MULTI_LUN=y
++CONFIG_SCSI_CONSTANTS=y
++CONFIG_SCSI_LOGGING=y
+
-+int disable_apic;
++#
++# SCSI Transport Attributes
++#
++CONFIG_SCSI_SPI_ATTRS=y
++CONFIG_SCSI_FC_ATTRS=y
++# CONFIG_SCSI_ISCSI_ATTRS is not set
++CONFIG_SCSI_SAS_ATTRS=y
+
-+/*
-+ * This initializes the IO-APIC and APIC hardware if this is
-+ * a UP kernel.
-+ */
-+int __init APIC_init_uniprocessor (void)
-+{
-+#ifdef CONFIG_X86_IO_APIC
-+ if (smp_found_config)
-+ if (!skip_ioapic_setup && nr_ioapics)
-+ setup_IO_APIC();
-+#endif
++#
++# SCSI low-level drivers
++#
++# CONFIG_ISCSI_TCP is not set
++# CONFIG_BLK_DEV_3W_XXXX_RAID is not set
++# CONFIG_SCSI_3W_9XXX is not set
++# CONFIG_SCSI_ACARD is not set
++# CONFIG_SCSI_AACRAID is not set
++# CONFIG_SCSI_AIC7XXX is not set
++# CONFIG_SCSI_AIC7XXX_OLD is not set
++# CONFIG_SCSI_AIC79XX is not set
++# CONFIG_MEGARAID_NEWGEN is not set
++# CONFIG_MEGARAID_LEGACY is not set
++# CONFIG_MEGARAID_SAS is not set
++# CONFIG_SCSI_SATA is not set
++# CONFIG_SCSI_HPTIOP is not set
++# CONFIG_SCSI_DMX3191D is not set
++# CONFIG_SCSI_FUTURE_DOMAIN is not set
++# CONFIG_SCSI_IPS is not set
++# CONFIG_SCSI_INITIO is not set
++# CONFIG_SCSI_INIA100 is not set
++CONFIG_SCSI_SYM53C8XX_2=y
++CONFIG_SCSI_SYM53C8XX_DMA_ADDRESSING_MODE=1
++CONFIG_SCSI_SYM53C8XX_DEFAULT_TAGS=16
++CONFIG_SCSI_SYM53C8XX_MAX_TAGS=64
++CONFIG_SCSI_SYM53C8XX_MMIO=y
++# CONFIG_SCSI_IPR is not set
++CONFIG_SCSI_QLOGIC_1280=y
++# CONFIG_SCSI_QLA_FC is not set
++# CONFIG_SCSI_LPFC is not set
++# CONFIG_SCSI_DC395x is not set
++# CONFIG_SCSI_DC390T is not set
++# CONFIG_SCSI_DEBUG is not set
+
-+ return 1;
-+}
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/asm-offsets.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/asm-offsets.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/asm-offsets.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/asm-offsets.c 2007-07-30 16:35:11.000000000 +0200
-@@ -67,8 +67,10 @@ int main(void)
- DEFINE(pbe_address, offsetof(struct pbe, address));
- DEFINE(pbe_orig_address, offsetof(struct pbe, orig_address));
- DEFINE(pbe_next, offsetof(struct pbe, next));
-+#ifndef CONFIG_X86_NO_TSS
- BLANK();
- DEFINE(TSS_ist, offsetof(struct tss_struct, ist));
-+#endif
- BLANK();
- DEFINE(crypto_tfm_ctx_offset, offsetof(struct crypto_tfm, __crt_ctx));
- return 0;
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/crash.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/crash.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/crash.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/crash.c 2007-07-30 16:35:11.000000000 +0200
-@@ -92,6 +92,7 @@ static void crash_save_self(struct pt_re
- crash_save_this_cpu(regs, cpu);
- }
-
-+#ifndef CONFIG_XEN
- #ifdef CONFIG_SMP
- static atomic_t waiting_for_crash_ipi;
-
-@@ -156,6 +157,7 @@ static void nmi_shootdown_cpus(void)
- /* There are no cpus to shootdown */
- }
- #endif
-+#endif /* CONFIG_XEN */
-
- void machine_crash_shutdown(struct pt_regs *regs)
- {
-@@ -173,6 +175,8 @@ void machine_crash_shutdown(struct pt_re
-
- /* Make a note of crashing cpu. Will be used in NMI callback.*/
- crashing_cpu = smp_processor_id();
++#
++# Multi-device support (RAID and LVM)
++#
++CONFIG_MD=y
++# CONFIG_BLK_DEV_MD is not set
++CONFIG_BLK_DEV_DM=y
++CONFIG_DM_CRYPT=m
++CONFIG_DM_SNAPSHOT=y
++CONFIG_DM_MIRROR=m
++CONFIG_DM_ZERO=m
++CONFIG_DM_MULTIPATH=m
++CONFIG_DM_MULTIPATH_EMC=m
+
-+#ifndef CONFIG_XEN
- nmi_shootdown_cpus();
-
- if(cpu_has_apic)
-@@ -181,6 +185,6 @@ void machine_crash_shutdown(struct pt_re
- #if defined(CONFIG_X86_IO_APIC)
- disable_IO_APIC();
- #endif
--
-+#endif /* CONFIG_XEN */
- crash_save_self(regs);
- }
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/e820-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/e820-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/e820-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/e820-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,774 @@
-+/*
-+ * Handle the memory map.
-+ * The functions here do the job until bootmem takes over.
-+ *
-+ * Getting sanitize_e820_map() in sync with i386 version by applying change:
-+ * - Provisions for empty E820 memory regions (reported by certain BIOSes).
-+ * Alex Achenbach <xela at slit.de>, December 2002.
-+ * Venkatesh Pallipadi <venkatesh.pallipadi at intel.com>
-+ *
-+ */
-+#include <linux/kernel.h>
-+#include <linux/types.h>
-+#include <linux/init.h>
-+#include <linux/bootmem.h>
-+#include <linux/ioport.h>
-+#include <linux/string.h>
-+#include <linux/kexec.h>
-+#include <linux/module.h>
++#
++# Fusion MPT device support
++#
++CONFIG_FUSION=y
++CONFIG_FUSION_SPI=y
++# CONFIG_FUSION_FC is not set
++CONFIG_FUSION_SAS=y
++CONFIG_FUSION_MAX_SGE=128
++# CONFIG_FUSION_CTL is not set
+
-+#include <asm/pgtable.h>
-+#include <asm/page.h>
-+#include <asm/e820.h>
-+#include <asm/proto.h>
-+#include <asm/bootsetup.h>
-+#include <asm/sections.h>
-+#include <xen/interface/memory.h>
++#
++# IEEE 1394 (FireWire) support
++#
++# CONFIG_IEEE1394 is not set
+
-+/*
-+ * PFN of last memory page.
-+ */
-+unsigned long end_pfn;
-+EXPORT_SYMBOL(end_pfn);
++#
++# I2O device support
++#
++# CONFIG_I2O is not set
+
-+/*
-+ * end_pfn only includes RAM, while end_pfn_map includes all e820 entries.
-+ * The direct mapping extends to end_pfn_map, so that we can directly access
-+ * apertures, ACPI and other tables without having to play with fixmaps.
-+ */
-+unsigned long end_pfn_map;
++#
++# Network device support
++#
++CONFIG_NETDEVICES=y
++CONFIG_DUMMY=y
++# CONFIG_BONDING is not set
++# CONFIG_EQUALIZER is not set
++CONFIG_TUN=y
+
-+/*
-+ * Last pfn which the user wants to use.
-+ */
-+unsigned long end_user_pfn = MAXMEM>>PAGE_SHIFT;
++#
++# ARCnet devices
++#
++CONFIG_ARCNET=y
++# CONFIG_ARCNET_1201 is not set
++# CONFIG_ARCNET_1051 is not set
++# CONFIG_ARCNET_RAW is not set
++# CONFIG_ARCNET_CAP is not set
++# CONFIG_ARCNET_COM90xx is not set
++# CONFIG_ARCNET_COM90xxIO is not set
++# CONFIG_ARCNET_RIM_I is not set
++# CONFIG_ARCNET_COM20020 is not set
+
-+extern struct resource code_resource, data_resource;
++#
++# PHY device support
++#
++# CONFIG_PHYLIB is not set
+
-+/* Check for some hardcoded bad areas that early boot is not allowed to touch */
-+static inline int bad_addr(unsigned long *addrp, unsigned long size)
-+{
-+ unsigned long addr = *addrp, last = addr + size;
++#
++# Ethernet (10 or 100Mbit)
++#
++CONFIG_NET_ETHERNET=y
++CONFIG_MII=y
++# CONFIG_HAPPYMEAL is not set
++# CONFIG_SUNGEM is not set
++# CONFIG_CASSINI is not set
++# CONFIG_NET_VENDOR_3COM is not set
+
-+#ifndef CONFIG_XEN
-+ /* various gunk below that needed for SMP startup */
-+ if (addr < 0x8000) {
-+ *addrp = 0x8000;
-+ return 1;
-+ }
++#
++# Tulip family network device support
++#
++CONFIG_NET_TULIP=y
++# CONFIG_DE2104X is not set
++CONFIG_TULIP=y
++CONFIG_TULIP_MWI=y
++CONFIG_TULIP_MMIO=y
++CONFIG_TULIP_NAPI=y
++CONFIG_TULIP_NAPI_HW_MITIGATION=y
++# CONFIG_DE4X5 is not set
++# CONFIG_WINBOND_840 is not set
++# CONFIG_DM9102 is not set
++# CONFIG_ULI526X is not set
++# CONFIG_HP100 is not set
++CONFIG_NET_PCI=y
++# CONFIG_PCNET32 is not set
++# CONFIG_AMD8111_ETH is not set
++# CONFIG_ADAPTEC_STARFIRE is not set
++# CONFIG_B44 is not set
++# CONFIG_FORCEDETH is not set
++# CONFIG_DGRS is not set
++CONFIG_EEPRO100=y
++CONFIG_E100=y
++# CONFIG_FEALNX is not set
++# CONFIG_NATSEMI is not set
++# CONFIG_NE2K_PCI is not set
++# CONFIG_8139CP is not set
++# CONFIG_8139TOO is not set
++# CONFIG_SIS900 is not set
++# CONFIG_EPIC100 is not set
++# CONFIG_SUNDANCE is not set
++# CONFIG_VIA_RHINE is not set
+
-+ /* direct mapping tables of the kernel */
-+ if (last >= table_start<<PAGE_SHIFT && addr < table_end<<PAGE_SHIFT) {
-+ *addrp = table_end << PAGE_SHIFT;
-+ return 1;
-+ }
++#
++# Ethernet (1000 Mbit)
++#
++# CONFIG_ACENIC is not set
++# CONFIG_DL2K is not set
++CONFIG_E1000=y
++# CONFIG_E1000_NAPI is not set
++# CONFIG_E1000_DISABLE_PACKET_SPLIT is not set
++# CONFIG_NS83820 is not set
++# CONFIG_HAMACHI is not set
++# CONFIG_YELLOWFIN is not set
++# CONFIG_R8169 is not set
++# CONFIG_SIS190 is not set
++# CONFIG_SKGE is not set
++# CONFIG_SKY2 is not set
++# CONFIG_SK98LIN is not set
++# CONFIG_VIA_VELOCITY is not set
++CONFIG_TIGON3=y
++# CONFIG_BNX2 is not set
+
-+ /* initrd */
-+#ifdef CONFIG_BLK_DEV_INITRD
-+ if (LOADER_TYPE && INITRD_START && last >= INITRD_START &&
-+ addr < INITRD_START+INITRD_SIZE) {
-+ *addrp = INITRD_START + INITRD_SIZE;
-+ return 1;
-+ }
-+#endif
-+ /* kernel code + 640k memory hole (later should not be needed, but
-+ be paranoid for now) */
-+ if (last >= 640*1024 && addr < 1024*1024) {
-+ *addrp = 1024*1024;
-+ return 1;
-+ }
-+ if (last >= __pa_symbol(&_text) && last < __pa_symbol(&_end)) {
-+ *addrp = __pa_symbol(&_end);
-+ return 1;
-+ }
++#
++# Ethernet (10000 Mbit)
++#
++# CONFIG_CHELSIO_T1 is not set
++# CONFIG_IXGB is not set
++# CONFIG_S2IO is not set
++# CONFIG_MYRI10GE is not set
++# CONFIG_SFC is not set
+
-+ if (last >= ebda_addr && addr < ebda_addr + ebda_size) {
-+ *addrp = ebda_addr + ebda_size;
-+ return 1;
-+ }
++#
++# Token Ring devices
++#
++# CONFIG_TR is not set
+
-+ /* XXX ramdisk image here? */
-+#else
-+ if (last < (table_end<<PAGE_SHIFT)) {
-+ *addrp = table_end << PAGE_SHIFT;
-+ return 1;
-+ }
-+#endif
-+ return 0;
-+}
++#
++# Wireless LAN (non-hamradio)
++#
++# CONFIG_NET_RADIO is not set
+
-+#ifndef CONFIG_XEN
-+/*
-+ * This function checks if any part of the range <start,end> is mapped
-+ * with type.
-+ */
-+int __meminit
-+e820_any_mapped(unsigned long start, unsigned long end, unsigned type)
-+{
-+ int i;
-+ for (i = 0; i < e820.nr_map; i++) {
-+ struct e820entry *ei = &e820.map[i];
-+ if (type && ei->type != type)
-+ continue;
-+ if (ei->addr >= end || ei->addr + ei->size <= start)
-+ continue;
-+ return 1;
-+ }
-+ return 0;
-+}
-+#endif
++#
++# Wan interfaces
++#
++# CONFIG_WAN is not set
++# CONFIG_FDDI is not set
++# CONFIG_HIPPI is not set
++# CONFIG_PPP is not set
++# CONFIG_SLIP is not set
++# CONFIG_NET_FC is not set
++# CONFIG_SHAPER is not set
++CONFIG_NETCONSOLE=y
++CONFIG_NETPOLL=y
++# CONFIG_NETPOLL_RX is not set
++# CONFIG_NETPOLL_TRAP is not set
++CONFIG_NET_POLL_CONTROLLER=y
+
-+/*
-+ * This function checks if the entire range <start,end> is mapped with type.
-+ *
-+ * Note: this function only works correct if the e820 table is sorted and
-+ * not-overlapping, which is the case
-+ */
-+int __init e820_all_mapped(unsigned long start, unsigned long end, unsigned type)
-+{
-+ int i;
++#
++# ISDN subsystem
++#
++CONFIG_ISDN=m
+
-+#ifndef CONFIG_XEN
-+ for (i = 0; i < e820.nr_map; i++) {
-+ struct e820entry *ei = &e820.map[i];
-+#else
-+ extern struct e820map machine_e820;
++#
++# Old ISDN4Linux
++#
++# CONFIG_ISDN_I4L is not set
+
-+ if (!is_initial_xendomain())
-+ return 0;
-+ for (i = 0; i < machine_e820.nr_map; i++) {
-+ const struct e820entry *ei = &machine_e820.map[i];
-+#endif
++#
++# CAPI subsystem
++#
++# CONFIG_ISDN_CAPI is not set
+
-+ if (type && ei->type != type)
-+ continue;
-+ /* is the region (part) in overlap with the current region ?*/
-+ if (ei->addr >= end || ei->addr + ei->size <= start)
-+ continue;
++#
++# Telephony Support
++#
++# CONFIG_PHONE is not set
+
-+ /* if the region is at the beginning of <start,end> we move
-+ * start to the end of the region since it's ok until there
-+ */
-+ if (ei->addr <= start)
-+ start = ei->addr + ei->size;
-+ /* if start is now at or beyond end, we're done, full coverage */
-+ if (start >= end)
-+ return 1; /* we're done */
-+ }
-+ return 0;
-+}
++#
++# Input device support
++#
++CONFIG_INPUT=y
+
-+/*
-+ * Find a free area in a specific range.
-+ */
-+unsigned long __init find_e820_area(unsigned long start, unsigned long end, unsigned size)
-+{
-+ int i;
-+ for (i = 0; i < e820.nr_map; i++) {
-+ struct e820entry *ei = &e820.map[i];
-+ unsigned long addr = ei->addr, last;
-+ if (ei->type != E820_RAM)
-+ continue;
-+ if (addr < start)
-+ addr = start;
-+ if (addr > ei->addr + ei->size)
-+ continue;
-+ while (bad_addr(&addr, size) && addr+size <= ei->addr+ei->size)
-+ ;
-+ last = addr + size;
-+ if (last > ei->addr + ei->size)
-+ continue;
-+ if (last > end)
-+ continue;
-+ return addr;
-+ }
-+ return -1UL;
-+}
++#
++# Userland interfaces
++#
++CONFIG_INPUT_MOUSEDEV=y
++CONFIG_INPUT_MOUSEDEV_PSAUX=y
++CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024
++CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768
++CONFIG_INPUT_JOYDEV=y
++# CONFIG_INPUT_TSDEV is not set
++CONFIG_INPUT_EVDEV=y
++# CONFIG_INPUT_EVBUG is not set
+
-+/*
-+ * Free bootmem based on the e820 table for a node.
-+ */
-+void __init e820_bootmem_free(pg_data_t *pgdat, unsigned long start,unsigned long end)
-+{
-+ int i;
-+ for (i = 0; i < e820.nr_map; i++) {
-+ struct e820entry *ei = &e820.map[i];
-+ unsigned long last, addr;
++#
++# Input Device Drivers
++#
++CONFIG_INPUT_KEYBOARD=y
++CONFIG_KEYBOARD_ATKBD=y
++# CONFIG_KEYBOARD_SUNKBD is not set
++# CONFIG_KEYBOARD_LKKBD is not set
++# CONFIG_KEYBOARD_XTKBD is not set
++# CONFIG_KEYBOARD_NEWTON is not set
++CONFIG_INPUT_MOUSE=y
++CONFIG_MOUSE_PS2=y
++# CONFIG_MOUSE_SERIAL is not set
++# CONFIG_MOUSE_VSXXXAA is not set
++# CONFIG_INPUT_JOYSTICK is not set
++# CONFIG_INPUT_TOUCHSCREEN is not set
++# CONFIG_INPUT_MISC is not set
+
-+ if (ei->type != E820_RAM ||
-+ ei->addr+ei->size <= start ||
-+ ei->addr >= end)
-+ continue;
++#
++# Hardware I/O ports
++#
++CONFIG_SERIO=y
++CONFIG_SERIO_I8042=y
++# CONFIG_SERIO_SERPORT is not set
++# CONFIG_SERIO_PCIPS2 is not set
++CONFIG_SERIO_LIBPS2=y
++# CONFIG_SERIO_RAW is not set
++CONFIG_GAMEPORT=y
++# CONFIG_GAMEPORT_NS558 is not set
++# CONFIG_GAMEPORT_L4 is not set
++# CONFIG_GAMEPORT_EMU10K1 is not set
++# CONFIG_GAMEPORT_FM801 is not set
+
-+ addr = round_up(ei->addr, PAGE_SIZE);
-+ if (addr < start)
-+ addr = start;
++#
++# Character devices
++#
++CONFIG_VT=y
++CONFIG_VT_CONSOLE=y
++CONFIG_HW_CONSOLE=y
++# CONFIG_VT_HW_CONSOLE_BINDING is not set
++CONFIG_SERIAL_NONSTANDARD=y
++# CONFIG_COMPUTONE is not set
++# CONFIG_ROCKETPORT is not set
++# CONFIG_CYCLADES is not set
++# CONFIG_DIGIEPCA is not set
++# CONFIG_MOXA_INTELLIO is not set
++# CONFIG_MOXA_SMARTIO is not set
++# CONFIG_ISI is not set
++# CONFIG_SYNCLINKMP is not set
++# CONFIG_SYNCLINK_GT is not set
++# CONFIG_N_HDLC is not set
++# CONFIG_SPECIALIX is not set
++# CONFIG_SX is not set
++# CONFIG_RIO is not set
++# CONFIG_STALDRV is not set
++# CONFIG_SGI_SNSC is not set
++# CONFIG_SGI_TIOCX is not set
+
-+ last = round_down(ei->addr + ei->size, PAGE_SIZE);
-+ if (last >= end)
-+ last = end;
++#
++# Serial drivers
++#
+
-+ if (last > addr && last-addr >= PAGE_SIZE)
-+ free_bootmem_node(pgdat, addr, last-addr);
-+ }
-+}
++#
++# Non-8250 serial port support
++#
++# CONFIG_SERIAL_SGI_L1_CONSOLE is not set
++# CONFIG_SERIAL_JSM is not set
++CONFIG_UNIX98_PTYS=y
++CONFIG_LEGACY_PTYS=y
++CONFIG_LEGACY_PTY_COUNT=256
+
-+/*
-+ * Find the highest page frame number we have available
-+ */
-+unsigned long __init e820_end_of_ram(void)
-+{
-+ int i;
-+ unsigned long end_pfn = 0;
-+
-+ for (i = 0; i < e820.nr_map; i++) {
-+ struct e820entry *ei = &e820.map[i];
-+ unsigned long start, end;
++#
++# IPMI
++#
++# CONFIG_IPMI_HANDLER is not set
+
-+ start = round_up(ei->addr, PAGE_SIZE);
-+ end = round_down(ei->addr + ei->size, PAGE_SIZE);
-+ if (start >= end)
-+ continue;
-+ if (ei->type == E820_RAM) {
-+ if (end > end_pfn<<PAGE_SHIFT)
-+ end_pfn = end>>PAGE_SHIFT;
-+ } else {
-+ if (end > end_pfn_map<<PAGE_SHIFT)
-+ end_pfn_map = end>>PAGE_SHIFT;
-+ }
-+ }
++#
++# Watchdog Cards
++#
++# CONFIG_WATCHDOG is not set
++# CONFIG_HW_RANDOM is not set
++CONFIG_EFI_RTC=y
++# CONFIG_DTLK is not set
++# CONFIG_R3964 is not set
++# CONFIG_APPLICOM is not set
+
-+ if (end_pfn > end_pfn_map)
-+ end_pfn_map = end_pfn;
-+ if (end_pfn_map > MAXMEM>>PAGE_SHIFT)
-+ end_pfn_map = MAXMEM>>PAGE_SHIFT;
-+ if (end_pfn > end_user_pfn)
-+ end_pfn = end_user_pfn;
-+ if (end_pfn > end_pfn_map)
-+ end_pfn = end_pfn_map;
++#
++# Ftape, the floppy tape device driver
++#
++CONFIG_AGP=y
++# CONFIG_AGP_SIS is not set
++# CONFIG_AGP_VIA is not set
++CONFIG_AGP_I460=y
++# CONFIG_AGP_HP_ZX1 is not set
++# CONFIG_AGP_SGI_TIOCA is not set
++CONFIG_DRM=y
++# CONFIG_DRM_TDFX is not set
++# CONFIG_DRM_R128 is not set
++# CONFIG_DRM_RADEON is not set
++# CONFIG_DRM_MGA is not set
++# CONFIG_DRM_SIS is not set
++# CONFIG_DRM_VIA is not set
++# CONFIG_DRM_SAVAGE is not set
++# CONFIG_RAW_DRIVER is not set
++# CONFIG_HPET is not set
++# CONFIG_HANGCHECK_TIMER is not set
++# CONFIG_MMTIMER is not set
+
-+ return end_pfn;
-+}
++#
++# TPM devices
++#
++# CONFIG_TCG_TPM is not set
++# CONFIG_TELCLOCK is not set
+
-+/*
-+ * Compute how much memory is missing in a range.
-+ * Unlike the other functions in this file the arguments are in page numbers.
-+ */
-+unsigned long __init
-+e820_hole_size(unsigned long start_pfn, unsigned long end_pfn)
-+{
-+ unsigned long ram = 0;
-+ unsigned long start = start_pfn << PAGE_SHIFT;
-+ unsigned long end = end_pfn << PAGE_SHIFT;
-+ int i;
-+ for (i = 0; i < e820.nr_map; i++) {
-+ struct e820entry *ei = &e820.map[i];
-+ unsigned long last, addr;
++#
++# I2C support
++#
++CONFIG_I2C=y
++CONFIG_I2C_CHARDEV=y
+
-+ if (ei->type != E820_RAM ||
-+ ei->addr+ei->size <= start ||
-+ ei->addr >= end)
-+ continue;
++#
++# I2C Algorithms
++#
++CONFIG_I2C_ALGOBIT=y
++CONFIG_I2C_ALGOPCF=y
++# CONFIG_I2C_ALGOPCA is not set
+
-+ addr = round_up(ei->addr, PAGE_SIZE);
-+ if (addr < start)
-+ addr = start;
++#
++# I2C Hardware Bus support
++#
++# CONFIG_I2C_ALI1535 is not set
++# CONFIG_I2C_ALI1563 is not set
++# CONFIG_I2C_ALI15X3 is not set
++# CONFIG_I2C_AMD756 is not set
++# CONFIG_I2C_AMD8111 is not set
++# CONFIG_I2C_I801 is not set
++# CONFIG_I2C_I810 is not set
++# CONFIG_I2C_PIIX4 is not set
++# CONFIG_I2C_NFORCE2 is not set
++# CONFIG_I2C_OCORES is not set
++# CONFIG_I2C_PARPORT_LIGHT is not set
++# CONFIG_I2C_PROSAVAGE is not set
++# CONFIG_I2C_SAVAGE4 is not set
++# CONFIG_I2C_SIS5595 is not set
++# CONFIG_I2C_SIS630 is not set
++# CONFIG_I2C_SIS96X is not set
++# CONFIG_I2C_STUB is not set
++# CONFIG_I2C_VIA is not set
++# CONFIG_I2C_VIAPRO is not set
++# CONFIG_I2C_VOODOO3 is not set
++# CONFIG_I2C_PCA_ISA is not set
+
-+ last = round_down(ei->addr + ei->size, PAGE_SIZE);
-+ if (last >= end)
-+ last = end;
++#
++# Miscellaneous I2C Chip support
++#
++# CONFIG_SENSORS_DS1337 is not set
++# CONFIG_SENSORS_DS1374 is not set
++# CONFIG_SENSORS_EEPROM is not set
++# CONFIG_SENSORS_PCF8574 is not set
++# CONFIG_SENSORS_PCA9539 is not set
++# CONFIG_SENSORS_PCF8591 is not set
++# CONFIG_SENSORS_MAX6875 is not set
++# CONFIG_I2C_DEBUG_CORE is not set
++# CONFIG_I2C_DEBUG_ALGO is not set
++# CONFIG_I2C_DEBUG_BUS is not set
++# CONFIG_I2C_DEBUG_CHIP is not set
+
-+ if (last > addr)
-+ ram += last - addr;
-+ }
-+ return ((end - start) - ram) >> PAGE_SHIFT;
-+}
++#
++# SPI support
++#
++# CONFIG_SPI is not set
++# CONFIG_SPI_MASTER is not set
+
-+/*
-+ * Mark e820 reserved areas as busy for the resource manager.
-+ */
-+void __init e820_reserve_resources(struct e820entry *e820, int nr_map)
-+{
-+ int i;
-+ for (i = 0; i < nr_map; i++) {
-+ struct resource *res;
-+ res = alloc_bootmem_low(sizeof(struct resource));
-+ switch (e820[i].type) {
-+ case E820_RAM: res->name = "System RAM"; break;
-+ case E820_ACPI: res->name = "ACPI Tables"; break;
-+ case E820_NVS: res->name = "ACPI Non-volatile Storage"; break;
-+ default: res->name = "reserved";
-+ }
-+ res->start = e820[i].addr;
-+ res->end = res->start + e820[i].size - 1;
-+ res->flags = IORESOURCE_MEM | IORESOURCE_BUSY;
-+ request_resource(&iomem_resource, res);
-+ if (e820[i].type == E820_RAM) {
-+ /*
-+ * We don't know which RAM region contains kernel data,
-+ * so we try it repeatedly and let the resource manager
-+ * test it.
-+ */
-+#ifndef CONFIG_XEN
-+ request_resource(res, &code_resource);
-+ request_resource(res, &data_resource);
-+#endif
-+#ifdef CONFIG_KEXEC
-+ if (crashk_res.start != crashk_res.end)
-+ request_resource(res, &crashk_res);
-+#ifdef CONFIG_XEN
-+ xen_machine_kexec_register_resources(res);
-+#endif
-+#endif
-+ }
-+ }
-+}
++#
++# Dallas's 1-wire bus
++#
+
-+/*
-+ * Add a memory region to the kernel e820 map.
-+ */
-+void __init add_memory_region(unsigned long start, unsigned long size, int type)
-+{
-+ int x = e820.nr_map;
++#
++# Hardware Monitoring support
++#
++CONFIG_HWMON=y
++# CONFIG_HWMON_VID is not set
++# CONFIG_SENSORS_ABITUGURU is not set
++# CONFIG_SENSORS_ADM1021 is not set
++# CONFIG_SENSORS_ADM1025 is not set
++# CONFIG_SENSORS_ADM1026 is not set
++# CONFIG_SENSORS_ADM1031 is not set
++# CONFIG_SENSORS_ADM9240 is not set
++# CONFIG_SENSORS_ASB100 is not set
++# CONFIG_SENSORS_ATXP1 is not set
++# CONFIG_SENSORS_DS1621 is not set
++# CONFIG_SENSORS_F71805F is not set
++# CONFIG_SENSORS_FSCHER is not set
++# CONFIG_SENSORS_FSCPOS is not set
++# CONFIG_SENSORS_GL518SM is not set
++# CONFIG_SENSORS_GL520SM is not set
++# CONFIG_SENSORS_IT87 is not set
++# CONFIG_SENSORS_LM63 is not set
++# CONFIG_SENSORS_LM75 is not set
++# CONFIG_SENSORS_LM77 is not set
++# CONFIG_SENSORS_LM78 is not set
++# CONFIG_SENSORS_LM80 is not set
++# CONFIG_SENSORS_LM83 is not set
++# CONFIG_SENSORS_LM85 is not set
++# CONFIG_SENSORS_LM87 is not set
++# CONFIG_SENSORS_LM90 is not set
++# CONFIG_SENSORS_LM92 is not set
++# CONFIG_SENSORS_MAX1619 is not set
++# CONFIG_SENSORS_PC87360 is not set
++# CONFIG_SENSORS_SIS5595 is not set
++# CONFIG_SENSORS_SMSC47M1 is not set
++# CONFIG_SENSORS_SMSC47M192 is not set
++# CONFIG_SENSORS_SMSC47B397 is not set
++# CONFIG_SENSORS_VIA686A is not set
++# CONFIG_SENSORS_VT8231 is not set
++# CONFIG_SENSORS_W83781D is not set
++# CONFIG_SENSORS_W83791D is not set
++# CONFIG_SENSORS_W83792D is not set
++# CONFIG_SENSORS_W83L785TS is not set
++# CONFIG_SENSORS_W83627HF is not set
++# CONFIG_SENSORS_W83627EHF is not set
++# CONFIG_HWMON_DEBUG_CHIP is not set
+
-+ if (x == E820MAX) {
-+ printk(KERN_ERR "Ooops! Too many entries in the memory map!\n");
-+ return;
-+ }
++#
++# Misc devices
++#
+
-+ e820.map[x].addr = start;
-+ e820.map[x].size = size;
-+ e820.map[x].type = type;
-+ e820.nr_map++;
-+}
++#
++# Multimedia devices
++#
++CONFIG_VIDEO_DEV=y
++CONFIG_VIDEO_V4L1=y
++CONFIG_VIDEO_V4L1_COMPAT=y
++CONFIG_VIDEO_V4L2=y
+
-+void __init e820_print_map(char *who)
-+{
-+ int i;
++#
++# Video Capture Adapters
++#
+
-+ for (i = 0; i < e820.nr_map; i++) {
-+ printk(" %s: %016Lx - %016Lx ", who,
-+ (unsigned long long) e820.map[i].addr,
-+ (unsigned long long) (e820.map[i].addr + e820.map[i].size));
-+ switch (e820.map[i].type) {
-+ case E820_RAM: printk("(usable)\n");
-+ break;
-+ case E820_RESERVED:
-+ printk("(reserved)\n");
-+ break;
-+ case E820_ACPI:
-+ printk("(ACPI data)\n");
-+ break;
-+ case E820_NVS:
-+ printk("(ACPI NVS)\n");
-+ break;
-+ default: printk("type %u\n", e820.map[i].type);
-+ break;
-+ }
-+ }
-+}
++#
++# Video Capture Adapters
++#
++# CONFIG_VIDEO_ADV_DEBUG is not set
++# CONFIG_VIDEO_VIVI is not set
++# CONFIG_VIDEO_BT848 is not set
++# CONFIG_VIDEO_CPIA is not set
++# CONFIG_VIDEO_CPIA2 is not set
++# CONFIG_VIDEO_SAA5246A is not set
++# CONFIG_VIDEO_SAA5249 is not set
++# CONFIG_TUNER_3036 is not set
++# CONFIG_VIDEO_STRADIS is not set
++# CONFIG_VIDEO_ZORAN is not set
++# CONFIG_VIDEO_SAA7134 is not set
++# CONFIG_VIDEO_MXB is not set
++# CONFIG_VIDEO_DPC is not set
++# CONFIG_VIDEO_HEXIUM_ORION is not set
++# CONFIG_VIDEO_HEXIUM_GEMINI is not set
++# CONFIG_VIDEO_CX88 is not set
+
-+/*
-+ * Sanitize the BIOS e820 map.
-+ *
-+ * Some e820 responses include overlapping entries. The following
-+ * replaces the original e820 map with a new one, removing overlaps.
-+ *
-+ */
-+static int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map)
-+{
-+ struct change_member {
-+ struct e820entry *pbios; /* pointer to original bios entry */
-+ unsigned long long addr; /* address for this change point */
-+ };
-+ static struct change_member change_point_list[2*E820MAX] __initdata;
-+ static struct change_member *change_point[2*E820MAX] __initdata;
-+ static struct e820entry *overlap_list[E820MAX] __initdata;
-+ static struct e820entry new_bios[E820MAX] __initdata;
-+ struct change_member *change_tmp;
-+ unsigned long current_type, last_type;
-+ unsigned long long last_addr;
-+ int chgidx, still_changing;
-+ int overlap_entries;
-+ int new_bios_entry;
-+ int old_nr, new_nr, chg_nr;
-+ int i;
++#
++# Encoders and Decoders
++#
++# CONFIG_VIDEO_MSP3400 is not set
++# CONFIG_VIDEO_CS53L32A is not set
++# CONFIG_VIDEO_TLV320AIC23B is not set
++# CONFIG_VIDEO_WM8775 is not set
++# CONFIG_VIDEO_WM8739 is not set
++# CONFIG_VIDEO_CX2341X is not set
++# CONFIG_VIDEO_CX25840 is not set
++# CONFIG_VIDEO_SAA711X is not set
++# CONFIG_VIDEO_SAA7127 is not set
++# CONFIG_VIDEO_UPD64031A is not set
++# CONFIG_VIDEO_UPD64083 is not set
+
-+ /*
-+ Visually we're performing the following (1,2,3,4 = memory types)...
++#
++# V4L USB devices
++#
++# CONFIG_VIDEO_PVRUSB2 is not set
++# CONFIG_VIDEO_EM28XX is not set
++# CONFIG_USB_VICAM is not set
++# CONFIG_USB_IBMCAM is not set
++# CONFIG_USB_KONICAWC is not set
++# CONFIG_USB_QUICKCAM_MESSENGER is not set
++# CONFIG_USB_ET61X251 is not set
++# CONFIG_VIDEO_OVCAMCHIP is not set
++# CONFIG_USB_W9968CF is not set
++# CONFIG_USB_OV511 is not set
++# CONFIG_USB_SE401 is not set
++# CONFIG_USB_SN9C102 is not set
++# CONFIG_USB_STV680 is not set
++# CONFIG_USB_ZC0301 is not set
++# CONFIG_USB_PWC is not set
+
-+ Sample memory map (w/overlaps):
-+ ____22__________________
-+ ______________________4_
-+ ____1111________________
-+ _44_____________________
-+ 11111111________________
-+ ____________________33__
-+ ___________44___________
-+ __________33333_________
-+ ______________22________
-+ ___________________2222_
-+ _________111111111______
-+ _____________________11_
-+ _________________4______
++#
++# Radio Adapters
++#
++# CONFIG_RADIO_GEMTEK_PCI is not set
++# CONFIG_RADIO_MAXIRADIO is not set
++# CONFIG_RADIO_MAESTRO is not set
++# CONFIG_USB_DSBR is not set
+
-+ Sanitized equivalent (no overlap):
-+ 1_______________________
-+ _44_____________________
-+ ___1____________________
-+ ____22__________________
-+ ______11________________
-+ _________1______________
-+ __________3_____________
-+ ___________44___________
-+ _____________33_________
-+ _______________2________
-+ ________________1_______
-+ _________________4______
-+ ___________________2____
-+ ____________________33__
-+ ______________________4_
-+ */
++#
++# Digital Video Broadcasting Devices
++#
++# CONFIG_DVB is not set
++# CONFIG_USB_DABUSB is not set
+
-+ /* if there's only one memory region, don't bother */
-+ if (*pnr_map < 2)
-+ return -1;
++#
++# Graphics support
++#
++CONFIG_FIRMWARE_EDID=y
++CONFIG_FB=y
++CONFIG_FB_CFB_FILLRECT=y
++CONFIG_FB_CFB_COPYAREA=y
++CONFIG_FB_CFB_IMAGEBLIT=y
++# CONFIG_FB_MACMODES is not set
++# CONFIG_FB_BACKLIGHT is not set
++CONFIG_FB_MODE_HELPERS=y
++# CONFIG_FB_TILEBLITTING is not set
++# CONFIG_FB_CIRRUS is not set
++# CONFIG_FB_PM2 is not set
++# CONFIG_FB_CYBER2000 is not set
++# CONFIG_FB_ASILIANT is not set
++# CONFIG_FB_IMSTT is not set
++# CONFIG_FB_S1D13XXX is not set
++# CONFIG_FB_NVIDIA is not set
++# CONFIG_FB_RIVA is not set
++# CONFIG_FB_MATROX is not set
++# CONFIG_FB_RADEON is not set
++# CONFIG_FB_ATY128 is not set
++# CONFIG_FB_ATY is not set
++# CONFIG_FB_SAVAGE is not set
++# CONFIG_FB_SIS is not set
++# CONFIG_FB_NEOMAGIC is not set
++# CONFIG_FB_KYRO is not set
++# CONFIG_FB_3DFX is not set
++# CONFIG_FB_VOODOO1 is not set
++# CONFIG_FB_TRIDENT is not set
++# CONFIG_FB_VIRTUAL is not set
+
-+ old_nr = *pnr_map;
++#
++# Console display driver support
++#
++CONFIG_VGA_CONSOLE=y
++# CONFIG_VGACON_SOFT_SCROLLBACK is not set
++CONFIG_DUMMY_CONSOLE=y
++CONFIG_FRAMEBUFFER_CONSOLE=y
++# CONFIG_FRAMEBUFFER_CONSOLE_ROTATION is not set
++# CONFIG_FONTS is not set
++CONFIG_FONT_8x8=y
++CONFIG_FONT_8x16=y
+
-+ /* bail out if we find any unreasonable addresses in bios map */
-+ for (i=0; i<old_nr; i++)
-+ if (biosmap[i].addr + biosmap[i].size < biosmap[i].addr)
-+ return -1;
++#
++# Logo configuration
++#
++CONFIG_LOGO=y
++# CONFIG_LOGO_LINUX_MONO is not set
++# CONFIG_LOGO_LINUX_VGA16 is not set
++CONFIG_LOGO_LINUX_CLUT224=y
++# CONFIG_BACKLIGHT_LCD_SUPPORT is not set
+
-+ /* create pointers for initial change-point information (for sorting) */
-+ for (i=0; i < 2*old_nr; i++)
-+ change_point[i] = &change_point_list[i];
++#
++# Sound
++#
++CONFIG_SOUND=y
+
-+ /* record all known change-points (starting and ending addresses),
-+ omitting those that are for empty memory regions */
-+ chgidx = 0;
-+ for (i=0; i < old_nr; i++) {
-+ if (biosmap[i].size != 0) {
-+ change_point[chgidx]->addr = biosmap[i].addr;
-+ change_point[chgidx++]->pbios = &biosmap[i];
-+ change_point[chgidx]->addr = biosmap[i].addr + biosmap[i].size;
-+ change_point[chgidx++]->pbios = &biosmap[i];
-+ }
-+ }
-+ chg_nr = chgidx;
++#
++# Advanced Linux Sound Architecture
++#
++CONFIG_SND=y
++CONFIG_SND_TIMER=y
++CONFIG_SND_PCM=y
++CONFIG_SND_HWDEP=y
++CONFIG_SND_RAWMIDI=y
++CONFIG_SND_SEQUENCER=y
++CONFIG_SND_SEQ_DUMMY=y
++CONFIG_SND_OSSEMUL=y
++CONFIG_SND_MIXER_OSS=y
++CONFIG_SND_PCM_OSS=y
++CONFIG_SND_PCM_OSS_PLUGINS=y
++CONFIG_SND_SEQUENCER_OSS=y
++# CONFIG_SND_DYNAMIC_MINORS is not set
++CONFIG_SND_SUPPORT_OLD_API=y
++CONFIG_SND_VERBOSE_PROCFS=y
++# CONFIG_SND_VERBOSE_PRINTK is not set
++# CONFIG_SND_DEBUG is not set
+
-+ /* sort change-point list by memory addresses (low -> high) */
-+ still_changing = 1;
-+ while (still_changing) {
-+ still_changing = 0;
-+ for (i=1; i < chg_nr; i++) {
-+ /* if <current_addr> > <last_addr>, swap */
-+ /* or, if current=<start_addr> & last=<end_addr>, swap */
-+ if ((change_point[i]->addr < change_point[i-1]->addr) ||
-+ ((change_point[i]->addr == change_point[i-1]->addr) &&
-+ (change_point[i]->addr == change_point[i]->pbios->addr) &&
-+ (change_point[i-1]->addr != change_point[i-1]->pbios->addr))
-+ )
-+ {
-+ change_tmp = change_point[i];
-+ change_point[i] = change_point[i-1];
-+ change_point[i-1] = change_tmp;
-+ still_changing=1;
-+ }
-+ }
-+ }
++#
++# Generic devices
++#
++CONFIG_SND_MPU401_UART=y
++CONFIG_SND_OPL3_LIB=y
++CONFIG_SND_AC97_CODEC=y
++CONFIG_SND_AC97_BUS=y
++CONFIG_SND_DUMMY=y
++CONFIG_SND_VIRMIDI=y
++# CONFIG_SND_MTPAV is not set
++CONFIG_SND_SERIAL_U16550=y
++CONFIG_SND_MPU401=y
++
++#
++# PCI devices
++#
++# CONFIG_SND_AD1889 is not set
++# CONFIG_SND_ALS300 is not set
++# CONFIG_SND_ALI5451 is not set
++CONFIG_SND_ATIIXP=y
++# CONFIG_SND_ATIIXP_MODEM is not set
++# CONFIG_SND_AU8810 is not set
++# CONFIG_SND_AU8820 is not set
++# CONFIG_SND_AU8830 is not set
++# CONFIG_SND_AZT3328 is not set
++# CONFIG_SND_BT87X is not set
++# CONFIG_SND_CA0106 is not set
++# CONFIG_SND_CMIPCI is not set
++# CONFIG_SND_CS4281 is not set
++# CONFIG_SND_CS46XX is not set
++# CONFIG_SND_DARLA20 is not set
++# CONFIG_SND_GINA20 is not set
++# CONFIG_SND_LAYLA20 is not set
++# CONFIG_SND_DARLA24 is not set
++# CONFIG_SND_GINA24 is not set
++# CONFIG_SND_LAYLA24 is not set
++# CONFIG_SND_MONA is not set
++# CONFIG_SND_MIA is not set
++# CONFIG_SND_ECHO3G is not set
++# CONFIG_SND_INDIGO is not set
++# CONFIG_SND_INDIGOIO is not set
++# CONFIG_SND_INDIGODJ is not set
++# CONFIG_SND_EMU10K1 is not set
++# CONFIG_SND_EMU10K1X is not set
++# CONFIG_SND_ENS1370 is not set
++# CONFIG_SND_ENS1371 is not set
++# CONFIG_SND_ES1938 is not set
++# CONFIG_SND_ES1968 is not set
++CONFIG_SND_FM801=y
++# CONFIG_SND_FM801_TEA575X_BOOL is not set
++# CONFIG_SND_HDA_INTEL is not set
++# CONFIG_SND_HDSP is not set
++# CONFIG_SND_HDSPM is not set
++# CONFIG_SND_ICE1712 is not set
++# CONFIG_SND_ICE1724 is not set
++# CONFIG_SND_INTEL8X0 is not set
++# CONFIG_SND_INTEL8X0M is not set
++# CONFIG_SND_KORG1212 is not set
++# CONFIG_SND_MAESTRO3 is not set
++# CONFIG_SND_MIXART is not set
++# CONFIG_SND_NM256 is not set
++# CONFIG_SND_PCXHR is not set
++# CONFIG_SND_RIPTIDE is not set
++# CONFIG_SND_RME32 is not set
++# CONFIG_SND_RME96 is not set
++# CONFIG_SND_RME9652 is not set
++# CONFIG_SND_SONICVIBES is not set
++# CONFIG_SND_TRIDENT is not set
++# CONFIG_SND_VIA82XX is not set
++# CONFIG_SND_VIA82XX_MODEM is not set
++# CONFIG_SND_VX222 is not set
++# CONFIG_SND_YMFPCI is not set
++
++#
++# USB devices
++#
++# CONFIG_SND_USB_AUDIO is not set
+
-+ /* create a new bios memory map, removing overlaps */
-+ overlap_entries=0; /* number of entries in the overlap table */
-+ new_bios_entry=0; /* index for creating new bios map entries */
-+ last_type = 0; /* start with undefined memory type */
-+ last_addr = 0; /* start with 0 as last starting address */
-+ /* loop through change-points, determining affect on the new bios map */
-+ for (chgidx=0; chgidx < chg_nr; chgidx++)
-+ {
-+ /* keep track of all overlapping bios entries */
-+ if (change_point[chgidx]->addr == change_point[chgidx]->pbios->addr)
-+ {
-+ /* add map entry to overlap list (> 1 entry implies an overlap) */
-+ overlap_list[overlap_entries++]=change_point[chgidx]->pbios;
-+ }
-+ else
-+ {
-+ /* remove entry from list (order independent, so swap with last) */
-+ for (i=0; i<overlap_entries; i++)
-+ {
-+ if (overlap_list[i] == change_point[chgidx]->pbios)
-+ overlap_list[i] = overlap_list[overlap_entries-1];
-+ }
-+ overlap_entries--;
-+ }
-+ /* if there are overlapping entries, decide which "type" to use */
-+ /* (larger value takes precedence -- 1=usable, 2,3,4,4+=unusable) */
-+ current_type = 0;
-+ for (i=0; i<overlap_entries; i++)
-+ if (overlap_list[i]->type > current_type)
-+ current_type = overlap_list[i]->type;
-+ /* continue building up new bios map based on this information */
-+ if (current_type != last_type) {
-+ if (last_type != 0) {
-+ new_bios[new_bios_entry].size =
-+ change_point[chgidx]->addr - last_addr;
-+ /* move forward only if the new size was non-zero */
-+ if (new_bios[new_bios_entry].size != 0)
-+ if (++new_bios_entry >= E820MAX)
-+ break; /* no more space left for new bios entries */
-+ }
-+ if (current_type != 0) {
-+ new_bios[new_bios_entry].addr = change_point[chgidx]->addr;
-+ new_bios[new_bios_entry].type = current_type;
-+ last_addr=change_point[chgidx]->addr;
-+ }
-+ last_type = current_type;
-+ }
-+ }
-+ new_nr = new_bios_entry; /* retain count for new bios entries */
++#
++# Open Sound System
++#
++CONFIG_SOUND_PRIME=y
++# CONFIG_OSS_OBSOLETE_DRIVER is not set
++# CONFIG_SOUND_BT878 is not set
++# CONFIG_SOUND_ES1371 is not set
++# CONFIG_SOUND_ICH is not set
++# CONFIG_SOUND_TRIDENT is not set
++# CONFIG_SOUND_MSNDCLAS is not set
++# CONFIG_SOUND_MSNDPIN is not set
++# CONFIG_SOUND_VIA82CXXX is not set
++# CONFIG_SOUND_TVMIXER is not set
+
-+ /* copy new bios mapping into original location */
-+ memcpy(biosmap, new_bios, new_nr*sizeof(struct e820entry));
-+ *pnr_map = new_nr;
++#
++# USB support
++#
++CONFIG_USB_ARCH_HAS_HCD=y
++CONFIG_USB_ARCH_HAS_OHCI=y
++CONFIG_USB_ARCH_HAS_EHCI=y
++CONFIG_USB=y
++# CONFIG_USB_DEBUG is not set
+
-+ return 0;
-+}
++#
++# Miscellaneous USB options
++#
++CONFIG_USB_DEVICEFS=y
++CONFIG_USB_BANDWIDTH=y
++# CONFIG_USB_DYNAMIC_MINORS is not set
++# CONFIG_USB_SUSPEND is not set
++# CONFIG_USB_OTG is not set
+
-+/*
-+ * Copy the BIOS e820 map into a safe place.
-+ *
-+ * Sanity-check it while we're at it..
-+ *
-+ * If we're lucky and live on a modern system, the setup code
-+ * will have given us a memory map that we can use to properly
-+ * set up memory. If we aren't, we'll fake a memory map.
-+ *
-+ * We check to see that the memory map contains at least 2 elements
-+ * before we'll use it, because the detection code in setup.S may
-+ * not be perfect and most every PC known to man has two memory
-+ * regions: one from 0 to 640k, and one from 1mb up. (The IBM
-+ * thinkpad 560x, for example, does not cooperate with the memory
-+ * detection code.)
-+ */
-+static int __init copy_e820_map(struct e820entry * biosmap, int nr_map)
-+{
-+#ifndef CONFIG_XEN
-+ /* Only one memory region (or negative)? Ignore it */
-+ if (nr_map < 2)
-+ return -1;
-+#else
-+ BUG_ON(nr_map < 1);
-+#endif
++#
++# USB Host Controller Drivers
++#
++CONFIG_USB_EHCI_HCD=y
++# CONFIG_USB_EHCI_SPLIT_ISO is not set
++# CONFIG_USB_EHCI_ROOT_HUB_TT is not set
++# CONFIG_USB_EHCI_TT_NEWSCHED is not set
++# CONFIG_USB_ISP116X_HCD is not set
++CONFIG_USB_OHCI_HCD=y
++# CONFIG_USB_OHCI_BIG_ENDIAN is not set
++CONFIG_USB_OHCI_LITTLE_ENDIAN=y
++CONFIG_USB_UHCI_HCD=y
++# CONFIG_USB_SL811_HCD is not set
+
-+ do {
-+ unsigned long start = biosmap->addr;
-+ unsigned long size = biosmap->size;
-+ unsigned long end = start + size;
-+ unsigned long type = biosmap->type;
++#
++# USB Device Class drivers
++#
++# CONFIG_USB_ACM is not set
++# CONFIG_USB_PRINTER is not set
+
-+ /* Overflow in 64 bits? Ignore the memory map. */
-+ if (start > end)
-+ return -1;
++#
++# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support'
++#
+
-+#ifndef CONFIG_XEN
-+ /*
-+ * Some BIOSes claim RAM in the 640k - 1M region.
-+ * Not right. Fix it up.
-+ *
-+ * This should be removed on Hammer which is supposed to not
-+ * have non e820 covered ISA mappings there, but I had some strange
-+ * problems so it stays for now. -AK
-+ */
-+ if (type == E820_RAM) {
-+ if (start < 0x100000ULL && end > 0xA0000ULL) {
-+ if (start < 0xA0000ULL)
-+ add_memory_region(start, 0xA0000ULL-start, type);
-+ if (end <= 0x100000ULL)
-+ continue;
-+ start = 0x100000ULL;
-+ size = end - start;
-+ }
-+ }
-+#endif
++#
++# may also be needed; see USB_STORAGE Help for more information
++#
++CONFIG_USB_STORAGE=y
++# CONFIG_USB_STORAGE_DEBUG is not set
++# CONFIG_USB_STORAGE_DATAFAB is not set
++# CONFIG_USB_STORAGE_FREECOM is not set
++# CONFIG_USB_STORAGE_ISD200 is not set
++# CONFIG_USB_STORAGE_DPCM is not set
++# CONFIG_USB_STORAGE_USBAT is not set
++# CONFIG_USB_STORAGE_SDDR09 is not set
++# CONFIG_USB_STORAGE_SDDR55 is not set
++# CONFIG_USB_STORAGE_JUMPSHOT is not set
++# CONFIG_USB_STORAGE_ALAUDA is not set
++# CONFIG_USB_LIBUSUAL is not set
+
-+ add_memory_region(start, size, type);
-+ } while (biosmap++,--nr_map);
-+ return 0;
-+}
++#
++# USB Input Devices
++#
++CONFIG_USB_HID=y
++CONFIG_USB_HIDINPUT=y
++# CONFIG_USB_HIDINPUT_POWERBOOK is not set
++# CONFIG_HID_FF is not set
++CONFIG_USB_HIDDEV=y
++# CONFIG_USB_AIPTEK is not set
++# CONFIG_USB_WACOM is not set
++# CONFIG_USB_ACECAD is not set
++# CONFIG_USB_KBTAB is not set
++# CONFIG_USB_POWERMATE is not set
++# CONFIG_USB_TOUCHSCREEN is not set
++# CONFIG_USB_YEALINK is not set
++# CONFIG_USB_XPAD is not set
++# CONFIG_USB_ATI_REMOTE is not set
++# CONFIG_USB_ATI_REMOTE2 is not set
++# CONFIG_USB_KEYSPAN_REMOTE is not set
++# CONFIG_USB_APPLETOUCH is not set
+
-+#ifndef CONFIG_XEN
-+void __init setup_memory_region(void)
-+{
-+ char *who = "BIOS-e820";
++#
++# USB Imaging devices
++#
++# CONFIG_USB_MDC800 is not set
++# CONFIG_USB_MICROTEK is not set
+
-+ /*
-+ * Try to copy the BIOS-supplied E820-map.
-+ *
-+ * Otherwise fake a memory map; one section from 0k->640k,
-+ * the next section from 1mb->appropriate_mem_k
-+ */
-+ sanitize_e820_map(E820_MAP, &E820_MAP_NR);
-+ if (copy_e820_map(E820_MAP, E820_MAP_NR) < 0) {
-+ unsigned long mem_size;
++#
++# USB Network Adapters
++#
++# CONFIG_USB_CATC is not set
++# CONFIG_USB_KAWETH is not set
++# CONFIG_USB_PEGASUS is not set
++# CONFIG_USB_RTL8150 is not set
++# CONFIG_USB_USBNET is not set
++CONFIG_USB_MON=y
+
-+ /* compare results from other methods and take the greater */
-+ if (ALT_MEM_K < EXT_MEM_K) {
-+ mem_size = EXT_MEM_K;
-+ who = "BIOS-88";
-+ } else {
-+ mem_size = ALT_MEM_K;
-+ who = "BIOS-e801";
-+ }
++#
++# USB port drivers
++#
+
-+ e820.nr_map = 0;
-+ add_memory_region(0, LOWMEMSIZE(), E820_RAM);
-+ add_memory_region(HIGH_MEMORY, mem_size << 10, E820_RAM);
-+ }
-+ printk(KERN_INFO "BIOS-provided physical RAM map:\n");
-+ e820_print_map(who);
-+}
++#
++# USB Serial Converter support
++#
++# CONFIG_USB_SERIAL is not set
+
-+#else /* CONFIG_XEN */
++#
++# USB Miscellaneous drivers
++#
++# CONFIG_USB_EMI62 is not set
++# CONFIG_USB_EMI26 is not set
++# CONFIG_USB_AUERSWALD is not set
++# CONFIG_USB_RIO500 is not set
++# CONFIG_USB_LEGOTOWER is not set
++# CONFIG_USB_LCD is not set
++# CONFIG_USB_LED is not set
++# CONFIG_USB_CYPRESS_CY7C63 is not set
++# CONFIG_USB_CYTHERM is not set
++# CONFIG_USB_PHIDGETKIT is not set
++# CONFIG_USB_PHIDGETSERVO is not set
++# CONFIG_USB_IDMOUSE is not set
++# CONFIG_USB_APPLEDISPLAY is not set
++# CONFIG_USB_SISUSBVGA is not set
++# CONFIG_USB_LD is not set
++# CONFIG_USB_TEST is not set
+
-+void __init setup_memory_region(void)
-+{
-+ int rc;
-+ struct xen_memory_map memmap;
-+ /*
-+ * This is rather large for a stack variable but this early in
-+ * the boot process we know we have plenty slack space.
-+ */
-+ struct e820entry map[E820MAX];
++#
++# USB DSL modem support
++#
+
-+ memmap.nr_entries = E820MAX;
-+ set_xen_guest_handle(memmap.buffer, map);
++#
++# USB Gadget Support
++#
++# CONFIG_USB_GADGET is not set
+
-+ rc = HYPERVISOR_memory_op(XENMEM_memory_map, &memmap);
-+ if ( rc == -ENOSYS ) {
-+ memmap.nr_entries = 1;
-+ map[0].addr = 0ULL;
-+ map[0].size = xen_start_info->nr_pages << PAGE_SHIFT;
-+ /* 8MB slack (to balance backend allocations). */
-+ map[0].size += 8 << 20;
-+ map[0].type = E820_RAM;
-+ rc = 0;
-+ }
-+ BUG_ON(rc);
++#
++# MMC/SD Card support
++#
++# CONFIG_MMC is not set
+
-+ sanitize_e820_map(map, (char *)&memmap.nr_entries);
++#
++# LED devices
++#
++# CONFIG_NEW_LEDS is not set
+
-+ BUG_ON(copy_e820_map(map, (char)memmap.nr_entries) < 0);
++#
++# LED drivers
++#
+
-+ printk(KERN_INFO "BIOS-provided physical RAM map:\n");
-+ e820_print_map("Xen");
-+}
-+#endif
++#
++# LED Triggers
++#
+
-+void __init parse_memopt(char *p, char **from)
-+{
-+ int i;
-+ unsigned long current_end;
-+ unsigned long end;
++#
++# InfiniBand support
++#
++# CONFIG_INFINIBAND is not set
+
-+ end_user_pfn = memparse(p, from);
-+ end_user_pfn >>= PAGE_SHIFT;
++#
++# EDAC - error detection and reporting (RAS) (EXPERIMENTAL)
++#
+
-+ end = end_user_pfn<<PAGE_SHIFT;
-+ i = e820.nr_map-1;
-+ current_end = e820.map[i].addr + e820.map[i].size;
++#
++# Real Time Clock
++#
++# CONFIG_RTC_CLASS is not set
+
-+ if (current_end < end) {
-+ /*
-+ * The e820 map ends before our requested size so
-+ * extend the final entry to the requested address.
-+ */
-+ if (e820.map[i].type == E820_RAM)
-+ e820.map[i].size = end - e820.map[i].addr;
-+ else
-+ add_memory_region(current_end, end - current_end, E820_RAM);
-+ }
-+}
++#
++# DMA Engine support
++#
++# CONFIG_DMA_ENGINE is not set
+
-+void __init parse_memmapopt(char *p, char **from)
-+{
-+ unsigned long long start_at, mem_size;
++#
++# DMA Clients
++#
+
-+ mem_size = memparse(p, from);
-+ p = *from;
-+ if (*p == '@') {
-+ start_at = memparse(p+1, from);
-+ add_memory_region(start_at, mem_size, E820_RAM);
-+ } else if (*p == '#') {
-+ start_at = memparse(p+1, from);
-+ add_memory_region(start_at, mem_size, E820_ACPI);
-+ } else if (*p == '$') {
-+ start_at = memparse(p+1, from);
-+ add_memory_region(start_at, mem_size, E820_RESERVED);
-+ } else {
-+ end_user_pfn = (mem_size >> PAGE_SHIFT);
-+ }
-+ p = *from;
-+}
++#
++# DMA Devices
++#
+
-+unsigned long pci_mem_start = 0xaeedbabe;
-+EXPORT_SYMBOL(pci_mem_start);
++#
++# File systems
++#
++CONFIG_EXT2_FS=y
++CONFIG_EXT2_FS_XATTR=y
++CONFIG_EXT2_FS_POSIX_ACL=y
++CONFIG_EXT2_FS_SECURITY=y
++# CONFIG_EXT2_FS_XIP is not set
++CONFIG_EXT3_FS=y
++CONFIG_EXT3_FS_XATTR=y
++CONFIG_EXT3_FS_POSIX_ACL=y
++CONFIG_EXT3_FS_SECURITY=y
++CONFIG_JBD=y
++# CONFIG_JBD_DEBUG is not set
++CONFIG_FS_MBCACHE=y
++CONFIG_REISERFS_FS=y
++# CONFIG_REISERFS_CHECK is not set
++# CONFIG_REISERFS_PROC_INFO is not set
++CONFIG_REISERFS_FS_XATTR=y
++CONFIG_REISERFS_FS_POSIX_ACL=y
++CONFIG_REISERFS_FS_SECURITY=y
++# CONFIG_JFS_FS is not set
++CONFIG_FS_POSIX_ACL=y
++CONFIG_XFS_FS=y
++# CONFIG_XFS_QUOTA is not set
++# CONFIG_XFS_SECURITY is not set
++# CONFIG_XFS_POSIX_ACL is not set
++# CONFIG_XFS_RT is not set
++# CONFIG_OCFS2_FS is not set
++# CONFIG_MINIX_FS is not set
++# CONFIG_ROMFS_FS is not set
++CONFIG_INOTIFY=y
++CONFIG_INOTIFY_USER=y
++# CONFIG_QUOTA is not set
++CONFIG_DNOTIFY=y
++CONFIG_AUTOFS_FS=y
++CONFIG_AUTOFS4_FS=y
++# CONFIG_FUSE_FS is not set
+
-+/*
-+ * Search for the biggest gap in the low 32 bits of the e820
-+ * memory space. We pass this space to PCI to assign MMIO resources
-+ * for hotplug or unconfigured devices in.
-+ * Hopefully the BIOS let enough space left.
-+ */
-+__init void e820_setup_gap(struct e820entry *e820, int nr_map)
-+{
-+ unsigned long gapstart, gapsize, round;
-+ unsigned long last;
-+ int i;
-+ int found = 0;
++#
++# CD-ROM/DVD Filesystems
++#
++CONFIG_ISO9660_FS=y
++CONFIG_JOLIET=y
++# CONFIG_ZISOFS is not set
++CONFIG_UDF_FS=y
++CONFIG_UDF_NLS=y
+
-+ last = 0x100000000ull;
-+ gapstart = 0x10000000;
-+ gapsize = 0x400000;
-+ i = nr_map;
-+ while (--i >= 0) {
-+ unsigned long long start = e820[i].addr;
-+ unsigned long long end = start + e820[i].size;
++#
++# DOS/FAT/NT Filesystems
++#
++CONFIG_FAT_FS=y
++CONFIG_MSDOS_FS=y
++CONFIG_VFAT_FS=y
++CONFIG_FAT_DEFAULT_CODEPAGE=437
++CONFIG_FAT_DEFAULT_IOCHARSET="iso8859-1"
++# CONFIG_NTFS_FS is not set
+
-+ /*
-+ * Since "last" is at most 4GB, we know we'll
-+ * fit in 32 bits if this condition is true
-+ */
-+ if (last > end) {
-+ unsigned long gap = last - end;
++#
++# Pseudo filesystems
++#
++CONFIG_PROC_FS=y
++CONFIG_PROC_KCORE=y
++CONFIG_PROC_IOMEM_MACHINE=y
++CONFIG_SYSFS=y
++CONFIG_TMPFS=y
++# CONFIG_HUGETLB_PAGE is not set
++CONFIG_RAMFS=y
++# CONFIG_CONFIGFS_FS is not set
+
-+ if (gap > gapsize) {
-+ gapsize = gap;
-+ gapstart = end;
-+ found = 1;
-+ }
-+ }
-+ if (start < last)
-+ last = start;
-+ }
++#
++# Miscellaneous filesystems
++#
++# CONFIG_ADFS_FS is not set
++# CONFIG_AFFS_FS is not set
++# CONFIG_HFS_FS is not set
++# CONFIG_HFSPLUS_FS is not set
++# CONFIG_BEFS_FS is not set
++# CONFIG_BFS_FS is not set
++# CONFIG_EFS_FS is not set
++# CONFIG_CRAMFS is not set
++# CONFIG_VXFS_FS is not set
++# CONFIG_HPFS_FS is not set
++# CONFIG_QNX4FS_FS is not set
++# CONFIG_SYSV_FS is not set
++# CONFIG_UFS_FS is not set
+
-+ if (!found) {
-+ gapstart = (end_pfn << PAGE_SHIFT) + 1024*1024;
-+ printk(KERN_ERR "PCI: Warning: Cannot find a gap in the 32bit address range\n"
-+ KERN_ERR "PCI: Unassigned devices with 32bit resource registers may break!\n");
-+ }
++#
++# Network File Systems
++#
++CONFIG_NFS_FS=y
++CONFIG_NFS_V3=y
++# CONFIG_NFS_V3_ACL is not set
++CONFIG_NFS_V4=y
++CONFIG_NFS_DIRECTIO=y
++CONFIG_NFSD=y
++CONFIG_NFSD_V3=y
++# CONFIG_NFSD_V3_ACL is not set
++CONFIG_NFSD_V4=y
++CONFIG_NFSD_TCP=y
++CONFIG_ROOT_NFS=y
++CONFIG_LOCKD=y
++CONFIG_LOCKD_V4=y
++CONFIG_EXPORTFS=y
++CONFIG_NFS_COMMON=y
++CONFIG_SUNRPC=y
++CONFIG_SUNRPC_GSS=y
++CONFIG_RPCSEC_GSS_KRB5=y
++# CONFIG_RPCSEC_GSS_SPKM3 is not set
++CONFIG_SMB_FS=y
++CONFIG_SMB_NLS_DEFAULT=y
++CONFIG_SMB_NLS_REMOTE="cp437"
++CONFIG_CIFS=y
++# CONFIG_CIFS_STATS is not set
++# CONFIG_CIFS_WEAK_PW_HASH is not set
++# CONFIG_CIFS_XATTR is not set
++# CONFIG_CIFS_DEBUG2 is not set
++# CONFIG_CIFS_EXPERIMENTAL is not set
++# CONFIG_NCP_FS is not set
++# CONFIG_CODA_FS is not set
++# CONFIG_AFS_FS is not set
++# CONFIG_9P_FS is not set
+
-+ /*
-+ * See how much we want to round up: start off with
-+ * rounding to the next 1MB area.
-+ */
-+ round = 0x100000;
-+ while ((gapsize >> 4) > round)
-+ round += round;
-+ /* Fun with two's complement */
-+ pci_mem_start = (gapstart + round) & -round;
++#
++# Partition Types
++#
++CONFIG_PARTITION_ADVANCED=y
++# CONFIG_ACORN_PARTITION is not set
++# CONFIG_OSF_PARTITION is not set
++# CONFIG_AMIGA_PARTITION is not set
++# CONFIG_ATARI_PARTITION is not set
++# CONFIG_MAC_PARTITION is not set
++CONFIG_MSDOS_PARTITION=y
++# CONFIG_BSD_DISKLABEL is not set
++# CONFIG_MINIX_SUBPARTITION is not set
++# CONFIG_SOLARIS_X86_PARTITION is not set
++# CONFIG_UNIXWARE_DISKLABEL is not set
++# CONFIG_LDM_PARTITION is not set
++CONFIG_SGI_PARTITION=y
++# CONFIG_ULTRIX_PARTITION is not set
++# CONFIG_SUN_PARTITION is not set
++# CONFIG_KARMA_PARTITION is not set
++CONFIG_EFI_PARTITION=y
+
-+ printk(KERN_INFO "Allocating PCI resources starting at %lx (gap: %lx:%lx)\n",
-+ pci_mem_start, gapstart, gapsize);
-+}
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/early_printk-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/early_printk-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/early_printk-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/early_printk-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,302 @@
-+#include <linux/console.h>
-+#include <linux/kernel.h>
-+#include <linux/init.h>
-+#include <linux/string.h>
-+#include <linux/screen_info.h>
-+#include <asm/io.h>
-+#include <asm/processor.h>
-+#include <asm/fcntl.h>
++#
++# Native Language Support
++#
++CONFIG_NLS=y
++CONFIG_NLS_DEFAULT="iso8859-1"
++CONFIG_NLS_CODEPAGE_437=y
++CONFIG_NLS_CODEPAGE_737=y
++CONFIG_NLS_CODEPAGE_775=y
++CONFIG_NLS_CODEPAGE_850=y
++CONFIG_NLS_CODEPAGE_852=y
++CONFIG_NLS_CODEPAGE_855=y
++CONFIG_NLS_CODEPAGE_857=y
++CONFIG_NLS_CODEPAGE_860=y
++CONFIG_NLS_CODEPAGE_861=y
++CONFIG_NLS_CODEPAGE_862=y
++CONFIG_NLS_CODEPAGE_863=y
++CONFIG_NLS_CODEPAGE_864=y
++CONFIG_NLS_CODEPAGE_865=y
++CONFIG_NLS_CODEPAGE_866=y
++CONFIG_NLS_CODEPAGE_869=y
++CONFIG_NLS_CODEPAGE_936=y
++CONFIG_NLS_CODEPAGE_950=y
++CONFIG_NLS_CODEPAGE_932=y
++CONFIG_NLS_CODEPAGE_949=y
++CONFIG_NLS_CODEPAGE_874=y
++CONFIG_NLS_ISO8859_8=y
++# CONFIG_NLS_CODEPAGE_1250 is not set
++CONFIG_NLS_CODEPAGE_1251=y
++# CONFIG_NLS_ASCII is not set
++CONFIG_NLS_ISO8859_1=y
++CONFIG_NLS_ISO8859_2=y
++CONFIG_NLS_ISO8859_3=y
++CONFIG_NLS_ISO8859_4=y
++CONFIG_NLS_ISO8859_5=y
++CONFIG_NLS_ISO8859_6=y
++CONFIG_NLS_ISO8859_7=y
++CONFIG_NLS_ISO8859_9=y
++CONFIG_NLS_ISO8859_13=y
++CONFIG_NLS_ISO8859_14=y
++CONFIG_NLS_ISO8859_15=y
++CONFIG_NLS_KOI8_R=y
++CONFIG_NLS_KOI8_U=y
++CONFIG_NLS_UTF8=y
+
-+/* Simple VGA output */
++#
++# Library routines
++#
++# CONFIG_CRC_CCITT is not set
++# CONFIG_CRC16 is not set
++CONFIG_CRC32=y
++# CONFIG_LIBCRC32C is not set
++CONFIG_PLIST=y
++CONFIG_GENERIC_HARDIRQS=y
++CONFIG_GENERIC_IRQ_PROBE=y
++CONFIG_GENERIC_PENDING_IRQ=y
++CONFIG_IRQ_PER_CPU=y
+
-+#ifdef __i386__
-+#include <asm/setup.h>
-+#define VGABASE (__ISA_IO_base + 0xb8000)
-+#else
-+#include <asm/bootsetup.h>
-+#define VGABASE ((void __iomem *)0xffffffff800b8000UL)
-+#endif
++#
++# HP Simulator drivers
++#
++# CONFIG_HP_SIMETH is not set
++# CONFIG_HP_SIMSERIAL is not set
++# CONFIG_HP_SIMSCSI is not set
+
-+#ifndef CONFIG_XEN
-+static int max_ypos = 25, max_xpos = 80;
-+static int current_ypos = 25, current_xpos = 0;
++#
++# Instrumentation Support
++#
++# CONFIG_PROFILING is not set
++# CONFIG_KPROBES is not set
+
-+static void early_vga_write(struct console *con, const char *str, unsigned n)
-+{
-+ char c;
-+ int i, k, j;
++#
++# Kernel hacking
++#
++# CONFIG_PRINTK_TIME is not set
++CONFIG_MAGIC_SYSRQ=y
++CONFIG_UNUSED_SYMBOLS=y
++CONFIG_DEBUG_KERNEL=y
++CONFIG_LOG_BUF_SHIFT=20
++CONFIG_DETECT_SOFTLOCKUP=y
++# CONFIG_SCHEDSTATS is not set
++# CONFIG_DEBUG_SLAB is not set
++# CONFIG_DEBUG_RT_MUTEXES is not set
++# CONFIG_RT_MUTEX_TESTER is not set
++# CONFIG_DEBUG_SPINLOCK is not set
++CONFIG_DEBUG_MUTEXES=y
++# CONFIG_DEBUG_RWSEMS is not set
++# CONFIG_DEBUG_SPINLOCK_SLEEP is not set
++# CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set
++# CONFIG_DEBUG_KOBJECT is not set
++# CONFIG_DEBUG_INFO is not set
++# CONFIG_DEBUG_FS is not set
++# CONFIG_DEBUG_VM is not set
++CONFIG_FORCED_INLINING=y
++# CONFIG_RCU_TORTURE_TEST is not set
++CONFIG_IA64_GRANULE_16MB=y
++# CONFIG_IA64_GRANULE_64MB is not set
++CONFIG_IA64_PRINT_HAZARDS=y
++# CONFIG_DISABLE_VHPT is not set
++# CONFIG_IA64_DEBUG_CMPXCHG is not set
++# CONFIG_IA64_DEBUG_IRQ is not set
+
-+ while ((c = *str++) != '\0' && n-- > 0) {
-+ if (current_ypos >= max_ypos) {
-+ /* scroll 1 line up */
-+ for (k = 1, j = 0; k < max_ypos; k++, j++) {
-+ for (i = 0; i < max_xpos; i++) {
-+ writew(readw(VGABASE+2*(max_xpos*k+i)),
-+ VGABASE + 2*(max_xpos*j + i));
-+ }
-+ }
-+ for (i = 0; i < max_xpos; i++)
-+ writew(0x720, VGABASE + 2*(max_xpos*j + i));
-+ current_ypos = max_ypos-1;
-+ }
-+ if (c == '\n') {
-+ current_xpos = 0;
-+ current_ypos++;
-+ } else if (c != '\r') {
-+ writew(((0x7 << 8) | (unsigned short) c),
-+ VGABASE + 2*(max_xpos*current_ypos +
-+ current_xpos++));
-+ if (current_xpos >= max_xpos) {
-+ current_xpos = 0;
-+ current_ypos++;
-+ }
-+ }
-+ }
-+}
++#
++# Security options
++#
++# CONFIG_KEYS is not set
++# CONFIG_SECURITY is not set
+
-+static struct console early_vga_console = {
-+ .name = "earlyvga",
-+ .write = early_vga_write,
-+ .flags = CON_PRINTBUFFER,
-+ .index = -1,
-+};
++#
++# Cryptographic options
++#
++CONFIG_CRYPTO=y
++# CONFIG_CRYPTO_HMAC is not set
++# CONFIG_CRYPTO_NULL is not set
++# CONFIG_CRYPTO_MD4 is not set
++CONFIG_CRYPTO_MD5=y
++# CONFIG_CRYPTO_SHA1 is not set
++# CONFIG_CRYPTO_SHA256 is not set
++# CONFIG_CRYPTO_SHA512 is not set
++# CONFIG_CRYPTO_WP512 is not set
++# CONFIG_CRYPTO_TGR192 is not set
++CONFIG_CRYPTO_DES=y
++# CONFIG_CRYPTO_BLOWFISH is not set
++# CONFIG_CRYPTO_TWOFISH is not set
++# CONFIG_CRYPTO_SERPENT is not set
++# CONFIG_CRYPTO_AES is not set
++# CONFIG_CRYPTO_CAST5 is not set
++# CONFIG_CRYPTO_CAST6 is not set
++# CONFIG_CRYPTO_TEA is not set
++# CONFIG_CRYPTO_ARC4 is not set
++# CONFIG_CRYPTO_KHAZAD is not set
++# CONFIG_CRYPTO_ANUBIS is not set
++# CONFIG_CRYPTO_DEFLATE is not set
++# CONFIG_CRYPTO_MICHAEL_MIC is not set
++# CONFIG_CRYPTO_CRC32C is not set
++# CONFIG_CRYPTO_TEST is not set
+
-+/* Serial functions loosely based on a similar package from Klaus P. Gerlicher */
++#
++# Hardware crypto devices
++#
++# CONFIG_XEN_SMPBOOT is not set
++# CONFIG_XEN_DEVMEM is not set
++CONFIG_XEN_INTERFACE_VERSION=0x00030207
+
-+static int early_serial_base = 0x3f8; /* ttyS0 */
++#
++# XEN
++#
++CONFIG_XEN_PRIVILEGED_GUEST=y
++# CONFIG_XEN_UNPRIVILEGED_GUEST is not set
++CONFIG_XEN_PRIVCMD=y
++CONFIG_XEN_XENBUS_DEV=y
++CONFIG_XEN_BACKEND=y
++CONFIG_XEN_BLKDEV_BACKEND=y
++CONFIG_XEN_BLKDEV_TAP=y
++CONFIG_XEN_NETDEV_BACKEND=y
++# CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_UTIL is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_BACKEND is not set
++# CONFIG_XEN_NETDEV_LOOPBACK is not set
++CONFIG_XEN_PCIDEV_BACKEND=y
++# CONFIG_XEN_PCIDEV_BACKEND_VPCI is not set
++# CONFIG_XEN_PCIDEV_BACKEND_PASS is not set
++# CONFIG_XEN_PCIDEV_BACKEND_SLOT is not set
++CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER=y
++# CONFIG_XEN_PCIDEV_BE_DEBUG is not set
++CONFIG_XEN_TPMDEV_BACKEND=m
++CONFIG_XEN_SCSI_BACKEND=m
++CONFIG_XEN_BLKDEV_FRONTEND=y
++CONFIG_XEN_NETDEV_FRONTEND=y
++CONFIG_XEN_SCSI_FRONTEND=m
++CONFIG_XEN_GRANT_DEV=y
++# CONFIG_XEN_NETDEV_ACCEL_SFC_FRONTEND is not set
++CONFIG_XEN_FRAMEBUFFER=y
++CONFIG_XEN_KEYBOARD=y
++# CONFIG_XEN_SCRUB_PAGES is not set
++CONFIG_XEN_DISABLE_SERIAL=y
++CONFIG_XEN_SYSFS=y
++CONFIG_XEN_COMPAT_030002_AND_LATER=y
++# CONFIG_XEN_COMPAT_030004_AND_LATER is not set
++# CONFIG_XEN_COMPAT_030100_AND_LATER is not set
++# CONFIG_XEN_COMPAT_LATEST_ONLY is not set
++CONFIG_XEN_COMPAT=0x030002
++CONFIG_HAVE_IRQ_IGNORE_UNHANDLED=y
++CONFIG_NO_IDLE_HZ=y
++CONFIG_XEN_BALLOON=y
++CONFIG_XEN_XENCOMM=y
+diff -r d894e36cfc30 -r 0aa021803deb buildconfigs/linux-defconfig_xen_x86_32
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/buildconfigs/linux-defconfig_xen_x86_32 Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,3326 @@
++#
++# Automatically generated make config: don't edit
++# Linux kernel version: 2.6.18.8
++# Mon Oct 15 14:37:47 2007
++#
++CONFIG_X86_32=y
++CONFIG_LOCKDEP_SUPPORT=y
++CONFIG_STACKTRACE_SUPPORT=y
++CONFIG_SEMAPHORE_SLEEPERS=y
++CONFIG_X86=y
++CONFIG_MMU=y
++CONFIG_GENERIC_ISA_DMA=y
++CONFIG_GENERIC_IOMAP=y
++CONFIG_GENERIC_HWEIGHT=y
++CONFIG_ARCH_MAY_HAVE_PC_FDC=y
++CONFIG_DMI=y
++CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config"
+
-+#define XMTRDY 0x20
++#
++# Code maturity level options
++#
++CONFIG_EXPERIMENTAL=y
++CONFIG_LOCK_KERNEL=y
++CONFIG_INIT_ENV_ARG_LIMIT=32
+
-+#define DLAB 0x80
++#
++# General setup
++#
++CONFIG_LOCALVERSION=""
++# CONFIG_LOCALVERSION_AUTO is not set
++CONFIG_SWAP=y
++CONFIG_SYSVIPC=y
++CONFIG_POSIX_MQUEUE=y
++CONFIG_BSD_PROCESS_ACCT=y
++CONFIG_BSD_PROCESS_ACCT_V3=y
++CONFIG_TASKSTATS=y
++CONFIG_TASK_DELAY_ACCT=y
++CONFIG_AUDIT=y
++CONFIG_AUDITSYSCALL=y
++CONFIG_IKCONFIG=y
++CONFIG_IKCONFIG_PROC=y
++CONFIG_CPUSETS=y
++# CONFIG_RELAY is not set
++CONFIG_INITRAMFS_SOURCE=""
++# CONFIG_CC_OPTIMIZE_FOR_SIZE is not set
++# CONFIG_EMBEDDED is not set
++CONFIG_UID16=y
++CONFIG_SYSCTL=y
++CONFIG_KALLSYMS=y
++# CONFIG_KALLSYMS_ALL is not set
++# CONFIG_KALLSYMS_EXTRA_PASS is not set
++CONFIG_HOTPLUG=y
++CONFIG_PRINTK=y
++CONFIG_BUG=y
++CONFIG_ELF_CORE=y
++CONFIG_BASE_FULL=y
++CONFIG_FUTEX=y
++CONFIG_EPOLL=y
++CONFIG_SHMEM=y
++CONFIG_SLAB=y
++CONFIG_VM_EVENT_COUNTERS=y
++CONFIG_RT_MUTEXES=y
++# CONFIG_TINY_SHMEM is not set
++CONFIG_BASE_SMALL=0
++# CONFIG_SLOB is not set
+
-+#define TXR 0 /* Transmit register (WRITE) */
-+#define RXR 0 /* Receive register (READ) */
-+#define IER 1 /* Interrupt Enable */
-+#define IIR 2 /* Interrupt ID */
-+#define FCR 2 /* FIFO control */
-+#define LCR 3 /* Line control */
-+#define MCR 4 /* Modem control */
-+#define LSR 5 /* Line Status */
-+#define MSR 6 /* Modem Status */
-+#define DLL 0 /* Divisor Latch Low */
-+#define DLH 1 /* Divisor latch High */
++#
++# Loadable module support
++#
++CONFIG_MODULES=y
++CONFIG_MODULE_UNLOAD=y
++CONFIG_MODULE_FORCE_UNLOAD=y
++CONFIG_MODVERSIONS=y
++CONFIG_MODULE_SRCVERSION_ALL=y
++CONFIG_KMOD=y
++CONFIG_STOP_MACHINE=y
+
-+static int early_serial_putc(unsigned char ch)
-+{
-+ unsigned timeout = 0xffff;
-+ while ((inb(early_serial_base + LSR) & XMTRDY) == 0 && --timeout)
-+ cpu_relax();
-+ outb(ch, early_serial_base + TXR);
-+ return timeout ? 0 : -1;
-+}
++#
++# Block layer
++#
++CONFIG_LBD=y
++# CONFIG_BLK_DEV_IO_TRACE is not set
++CONFIG_LSF=y
+
-+static void early_serial_write(struct console *con, const char *s, unsigned n)
-+{
-+ while (*s && n-- > 0) {
-+ early_serial_putc(*s);
-+ if (*s == '\n')
-+ early_serial_putc('\r');
-+ s++;
-+ }
-+}
++#
++# IO Schedulers
++#
++CONFIG_IOSCHED_NOOP=y
++CONFIG_IOSCHED_AS=y
++CONFIG_IOSCHED_DEADLINE=y
++CONFIG_IOSCHED_CFQ=y
++# CONFIG_DEFAULT_AS is not set
++# CONFIG_DEFAULT_DEADLINE is not set
++CONFIG_DEFAULT_CFQ=y
++# CONFIG_DEFAULT_NOOP is not set
++CONFIG_DEFAULT_IOSCHED="cfq"
+
-+#define DEFAULT_BAUD 9600
++#
++# Processor type and features
++#
++CONFIG_SMP=y
++# CONFIG_X86_PC is not set
++CONFIG_X86_XEN=y
++# CONFIG_X86_ELAN is not set
++# CONFIG_X86_VOYAGER is not set
++# CONFIG_X86_NUMAQ is not set
++# CONFIG_X86_SUMMIT is not set
++# CONFIG_X86_BIGSMP is not set
++# CONFIG_X86_VISWS is not set
++# CONFIG_X86_GENERICARCH is not set
++# CONFIG_X86_ES7000 is not set
++# CONFIG_M386 is not set
++# CONFIG_M486 is not set
++# CONFIG_M586 is not set
++# CONFIG_M586TSC is not set
++# CONFIG_M586MMX is not set
++CONFIG_M686=y
++# CONFIG_MPENTIUMII is not set
++# CONFIG_MPENTIUMIII is not set
++# CONFIG_MPENTIUMM is not set
++# CONFIG_MPENTIUM4 is not set
++# CONFIG_MK6 is not set
++# CONFIG_MK7 is not set
++# CONFIG_MK8 is not set
++# CONFIG_MCRUSOE is not set
++# CONFIG_MEFFICEON is not set
++# CONFIG_MWINCHIPC6 is not set
++# CONFIG_MWINCHIP2 is not set
++# CONFIG_MWINCHIP3D is not set
++# CONFIG_MGEODEGX1 is not set
++# CONFIG_MGEODE_LX is not set
++# CONFIG_MCYRIXIII is not set
++# CONFIG_MVIAC3_2 is not set
++CONFIG_X86_GENERIC=y
++CONFIG_X86_CMPXCHG=y
++CONFIG_X86_XADD=y
++CONFIG_X86_L1_CACHE_SHIFT=7
++CONFIG_RWSEM_XCHGADD_ALGORITHM=y
++CONFIG_GENERIC_CALIBRATE_DELAY=y
++CONFIG_X86_PPRO_FENCE=y
++CONFIG_X86_WP_WORKS_OK=y
++CONFIG_X86_INVLPG=y
++CONFIG_X86_BSWAP=y
++CONFIG_X86_POPAD_OK=y
++CONFIG_X86_CMPXCHG64=y
++CONFIG_X86_GOOD_APIC=y
++CONFIG_X86_INTEL_USERCOPY=y
++CONFIG_X86_USE_PPRO_CHECKSUM=y
++CONFIG_NR_CPUS=32
++# CONFIG_PREEMPT_NONE is not set
++CONFIG_PREEMPT_VOLUNTARY=y
++# CONFIG_PREEMPT is not set
++CONFIG_PREEMPT_BKL=y
++CONFIG_X86_LOCAL_APIC=y
++CONFIG_X86_IO_APIC=y
++CONFIG_VM86=y
++# CONFIG_TOSHIBA is not set
++# CONFIG_I8K is not set
++# CONFIG_X86_REBOOTFIXUPS is not set
++CONFIG_MICROCODE=y
++CONFIG_X86_MSR=y
++CONFIG_X86_CPUID=m
++CONFIG_SWIOTLB=y
+
-+static __init void early_serial_init(char *s)
-+{
-+ unsigned char c;
-+ unsigned divisor;
-+ unsigned baud = DEFAULT_BAUD;
-+ char *e;
++#
++# Firmware Drivers
++#
++CONFIG_EDD=y
++CONFIG_DELL_RBU=m
++CONFIG_DCDBAS=m
++# CONFIG_NOHIGHMEM is not set
++CONFIG_HIGHMEM4G=y
++# CONFIG_HIGHMEM64G is not set
++CONFIG_PAGE_OFFSET=0xC0000000
++CONFIG_HIGHMEM=y
++CONFIG_SELECT_MEMORY_MODEL=y
++CONFIG_FLATMEM_MANUAL=y
++# CONFIG_DISCONTIGMEM_MANUAL is not set
++# CONFIG_SPARSEMEM_MANUAL is not set
++CONFIG_FLATMEM=y
++CONFIG_FLAT_NODE_MEM_MAP=y
++# CONFIG_SPARSEMEM_STATIC is not set
++CONFIG_SPLIT_PTLOCK_CPUS=4
++CONFIG_RESOURCES_64BIT=y
++# CONFIG_HIGHPTE is not set
++CONFIG_MTRR=y
++CONFIG_REGPARM=y
++CONFIG_SECCOMP=y
++CONFIG_HZ_100=y
++# CONFIG_HZ_250 is not set
++# CONFIG_HZ_1000 is not set
++CONFIG_HZ=100
++CONFIG_KEXEC=y
++# CONFIG_CRASH_DUMP is not set
++CONFIG_PHYSICAL_START=0x100000
++CONFIG_HOTPLUG_CPU=y
++CONFIG_ARCH_ENABLE_MEMORY_HOTPLUG=y
+
-+ if (*s == ',')
-+ ++s;
++#
++# Power management options (ACPI, APM)
++#
++CONFIG_PM=y
++# CONFIG_PM_LEGACY is not set
++CONFIG_PM_DEBUG=y
++# CONFIG_PM_TRACE is not set
++# CONFIG_SOFTWARE_SUSPEND is not set
++CONFIG_SUSPEND_SMP=y
+
-+ if (*s) {
-+ unsigned port;
-+ if (!strncmp(s,"0x",2)) {
-+ early_serial_base = simple_strtoul(s, &e, 16);
-+ } else {
-+ static int bases[] = { 0x3f8, 0x2f8 };
++#
++# ACPI (Advanced Configuration and Power Interface) Support
++#
++CONFIG_ACPI=y
++CONFIG_ACPI_SLEEP=y
++CONFIG_ACPI_SLEEP_PROC_FS=y
++# CONFIG_ACPI_SLEEP_PROC_SLEEP is not set
++CONFIG_ACPI_AC=m
++CONFIG_ACPI_BATTERY=m
++CONFIG_ACPI_BUTTON=m
++CONFIG_ACPI_VIDEO=m
++CONFIG_ACPI_HOTKEY=m
++CONFIG_ACPI_FAN=m
++CONFIG_ACPI_DOCK=m
++CONFIG_ACPI_PROCESSOR=m
++CONFIG_ACPI_HOTPLUG_CPU=y
++CONFIG_ACPI_THERMAL=m
++CONFIG_ACPI_ASUS=m
++CONFIG_ACPI_IBM=m
++CONFIG_ACPI_TOSHIBA=m
++CONFIG_ACPI_BLACKLIST_YEAR=0
++# CONFIG_ACPI_DEBUG is not set
++CONFIG_ACPI_EC=y
++CONFIG_ACPI_POWER=y
++CONFIG_ACPI_SYSTEM=y
++CONFIG_ACPI_CONTAINER=m
++CONFIG_ACPI_SBS=m
++CONFIG_ACPI_PV_SLEEP=y
+
-+ if (!strncmp(s,"ttyS",4))
-+ s += 4;
-+ port = simple_strtoul(s, &e, 10);
-+ if (port > 1 || s == e)
-+ port = 0;
-+ early_serial_base = bases[port];
-+ }
-+ s += strcspn(s, ",");
-+ if (*s == ',')
-+ s++;
-+ }
++#
++# CPU Frequency scaling
++#
++# CONFIG_CPU_FREQ is not set
+
-+ outb(0x3, early_serial_base + LCR); /* 8n1 */
-+ outb(0, early_serial_base + IER); /* no interrupt */
-+ outb(0, early_serial_base + FCR); /* no fifo */
-+ outb(0x3, early_serial_base + MCR); /* DTR + RTS */
++#
++# Bus options (PCI, PCMCIA, EISA, MCA, ISA)
++#
++CONFIG_PCI=y
++# CONFIG_PCI_GOBIOS is not set
++# CONFIG_PCI_GOMMCONFIG is not set
++# CONFIG_PCI_GODIRECT is not set
++# CONFIG_PCI_GOXEN_FE is not set
++CONFIG_PCI_GOANY=y
++CONFIG_PCI_DIRECT=y
++CONFIG_PCI_MMCONFIG=y
++CONFIG_XEN_PCIDEV_FRONTEND=y
++# CONFIG_XEN_PCIDEV_FE_DEBUG is not set
++# CONFIG_PCIEPORTBUS is not set
++CONFIG_PCI_MSI=y
++# CONFIG_PCI_DEBUG is not set
++CONFIG_ISA_DMA_API=y
++CONFIG_SCx200=m
++CONFIG_K8_NB=y
+
-+ if (*s) {
-+ baud = simple_strtoul(s, &e, 0);
-+ if (baud == 0 || s == e)
-+ baud = DEFAULT_BAUD;
-+ }
++#
++# PCCARD (PCMCIA/CardBus) support
++#
++CONFIG_PCCARD=m
++# CONFIG_PCMCIA_DEBUG is not set
++CONFIG_PCMCIA=m
++CONFIG_PCMCIA_LOAD_CIS=y
++CONFIG_PCMCIA_IOCTL=y
++CONFIG_CARDBUS=y
+
-+ divisor = 115200 / baud;
-+ c = inb(early_serial_base + LCR);
-+ outb(c | DLAB, early_serial_base + LCR);
-+ outb(divisor & 0xff, early_serial_base + DLL);
-+ outb((divisor >> 8) & 0xff, early_serial_base + DLH);
-+ outb(c & ~DLAB, early_serial_base + LCR);
-+}
++#
++# PC-card bridges
++#
++CONFIG_YENTA=m
++CONFIG_YENTA_O2=y
++CONFIG_YENTA_RICOH=y
++CONFIG_YENTA_TI=y
++CONFIG_YENTA_ENE_TUNE=y
++CONFIG_YENTA_TOSHIBA=y
++CONFIG_PD6729=m
++CONFIG_I82092=m
++CONFIG_PCCARD_NONSTATIC=m
+
-+#else /* CONFIG_XEN */
++#
++# PCI Hotplug Support
++#
++CONFIG_HOTPLUG_PCI=m
++CONFIG_HOTPLUG_PCI_FAKE=m
++CONFIG_HOTPLUG_PCI_ACPI=m
++CONFIG_HOTPLUG_PCI_ACPI_IBM=m
++CONFIG_HOTPLUG_PCI_CPCI=y
++CONFIG_HOTPLUG_PCI_CPCI_ZT5550=m
++CONFIG_HOTPLUG_PCI_CPCI_GENERIC=m
++CONFIG_HOTPLUG_PCI_SHPC=m
++# CONFIG_HOTPLUG_PCI_SHPC_POLL_EVENT_MODE is not set
+
-+static void
-+early_serial_write(struct console *con, const char *s, unsigned count)
-+{
-+ int n;
++#
++# Executable file formats
++#
++CONFIG_BINFMT_ELF=y
++CONFIG_BINFMT_AOUT=m
++CONFIG_BINFMT_MISC=m
+
-+ while (count > 0) {
-+ n = HYPERVISOR_console_io(CONSOLEIO_write, count, (char *)s);
-+ if (n <= 0)
-+ break;
-+ count -= n;
-+ s += n;
-+ }
-+}
++#
++# Networking
++#
++CONFIG_NET=y
+
-+static __init void early_serial_init(char *s)
-+{
-+}
++#
++# Networking options
++#
++# CONFIG_NETDEBUG is not set
++CONFIG_PACKET=y
++CONFIG_PACKET_MMAP=y
++CONFIG_UNIX=y
++CONFIG_XFRM=y
++CONFIG_XFRM_USER=m
++CONFIG_NET_KEY=m
++CONFIG_INET=y
++CONFIG_IP_MULTICAST=y
++CONFIG_IP_ADVANCED_ROUTER=y
++CONFIG_ASK_IP_FIB_HASH=y
++# CONFIG_IP_FIB_TRIE is not set
++CONFIG_IP_FIB_HASH=y
++CONFIG_IP_MULTIPLE_TABLES=y
++CONFIG_IP_ROUTE_FWMARK=y
++CONFIG_IP_ROUTE_MULTIPATH=y
++# CONFIG_IP_ROUTE_MULTIPATH_CACHED is not set
++CONFIG_IP_ROUTE_VERBOSE=y
++# CONFIG_IP_PNP is not set
++CONFIG_NET_IPIP=m
++CONFIG_NET_IPGRE=m
++CONFIG_NET_IPGRE_BROADCAST=y
++CONFIG_IP_MROUTE=y
++CONFIG_IP_PIMSM_V1=y
++CONFIG_IP_PIMSM_V2=y
++# CONFIG_ARPD is not set
++CONFIG_SYN_COOKIES=y
++CONFIG_INET_AH=m
++CONFIG_INET_ESP=m
++CONFIG_INET_IPCOMP=m
++CONFIG_INET_XFRM_TUNNEL=m
++CONFIG_INET_TUNNEL=m
++CONFIG_INET_XFRM_MODE_TRANSPORT=m
++CONFIG_INET_XFRM_MODE_TUNNEL=m
++CONFIG_INET_DIAG=m
++CONFIG_INET_TCP_DIAG=m
++CONFIG_TCP_CONG_ADVANCED=y
+
-+/*
-+ * No early VGA console on Xen, as we do not have convenient ISA-space
-+ * mappings. Someone should fix this for domain 0. For now, use fake serial.
-+ */
-+#define early_vga_console early_serial_console
++#
++# TCP congestion control
++#
++CONFIG_TCP_CONG_BIC=m
++CONFIG_TCP_CONG_CUBIC=m
++CONFIG_TCP_CONG_WESTWOOD=m
++CONFIG_TCP_CONG_HTCP=m
++CONFIG_TCP_CONG_HSTCP=m
++CONFIG_TCP_CONG_HYBLA=m
++CONFIG_TCP_CONG_VEGAS=m
++CONFIG_TCP_CONG_SCALABLE=m
++CONFIG_TCP_CONG_LP=m
++CONFIG_TCP_CONG_VENO=m
+
-+#endif
++#
++# IP: Virtual Server Configuration
++#
++CONFIG_IP_VS=m
++# CONFIG_IP_VS_DEBUG is not set
++CONFIG_IP_VS_TAB_BITS=12
+
-+static struct console early_serial_console = {
-+ .name = "earlyser",
-+ .write = early_serial_write,
-+ .flags = CON_PRINTBUFFER,
-+ .index = -1,
-+};
++#
++# IPVS transport protocol load balancing support
++#
++CONFIG_IP_VS_PROTO_TCP=y
++CONFIG_IP_VS_PROTO_UDP=y
++CONFIG_IP_VS_PROTO_ESP=y
++CONFIG_IP_VS_PROTO_AH=y
+
-+/* Console interface to a host file on AMD's SimNow! */
++#
++# IPVS scheduler
++#
++CONFIG_IP_VS_RR=m
++CONFIG_IP_VS_WRR=m
++CONFIG_IP_VS_LC=m
++CONFIG_IP_VS_WLC=m
++CONFIG_IP_VS_LBLC=m
++CONFIG_IP_VS_LBLCR=m
++CONFIG_IP_VS_DH=m
++CONFIG_IP_VS_SH=m
++CONFIG_IP_VS_SED=m
++CONFIG_IP_VS_NQ=m
+
-+static int simnow_fd;
++#
++# IPVS application helper
++#
++CONFIG_IP_VS_FTP=m
++CONFIG_IPV6=m
++CONFIG_IPV6_PRIVACY=y
++# CONFIG_IPV6_ROUTER_PREF is not set
++CONFIG_INET6_AH=m
++CONFIG_INET6_ESP=m
++CONFIG_INET6_IPCOMP=m
++CONFIG_INET6_XFRM_TUNNEL=m
++CONFIG_INET6_TUNNEL=m
++CONFIG_INET6_XFRM_MODE_TRANSPORT=m
++CONFIG_INET6_XFRM_MODE_TUNNEL=m
++CONFIG_IPV6_TUNNEL=m
++CONFIG_NETWORK_SECMARK=y
++CONFIG_NETFILTER=y
++# CONFIG_NETFILTER_DEBUG is not set
++CONFIG_BRIDGE_NETFILTER=y
+
-+enum {
-+ MAGIC1 = 0xBACCD00A,
-+ MAGIC2 = 0xCA110000,
-+ XOPEN = 5,
-+ XWRITE = 4,
-+};
++#
++# Core Netfilter Configuration
++#
++CONFIG_NETFILTER_NETLINK=m
++CONFIG_NETFILTER_NETLINK_QUEUE=m
++CONFIG_NETFILTER_NETLINK_LOG=m
++CONFIG_NETFILTER_XTABLES=m
++CONFIG_NETFILTER_XT_TARGET_CLASSIFY=m
++CONFIG_NETFILTER_XT_TARGET_CONNMARK=m
++CONFIG_NETFILTER_XT_TARGET_MARK=m
++CONFIG_NETFILTER_XT_TARGET_NFQUEUE=m
++CONFIG_NETFILTER_XT_TARGET_NOTRACK=m
++CONFIG_NETFILTER_XT_TARGET_SECMARK=m
++# CONFIG_NETFILTER_XT_TARGET_CONNSECMARK is not set
++CONFIG_NETFILTER_XT_MATCH_COMMENT=m
++CONFIG_NETFILTER_XT_MATCH_CONNBYTES=m
++CONFIG_NETFILTER_XT_MATCH_CONNMARK=m
++CONFIG_NETFILTER_XT_MATCH_CONNTRACK=m
++CONFIG_NETFILTER_XT_MATCH_DCCP=m
++CONFIG_NETFILTER_XT_MATCH_ESP=m
++CONFIG_NETFILTER_XT_MATCH_HELPER=m
++CONFIG_NETFILTER_XT_MATCH_LENGTH=m
++CONFIG_NETFILTER_XT_MATCH_LIMIT=m
++CONFIG_NETFILTER_XT_MATCH_MAC=m
++CONFIG_NETFILTER_XT_MATCH_MARK=m
++CONFIG_NETFILTER_XT_MATCH_POLICY=m
++CONFIG_NETFILTER_XT_MATCH_MULTIPORT=m
++CONFIG_NETFILTER_XT_MATCH_PHYSDEV=m
++CONFIG_NETFILTER_XT_MATCH_PKTTYPE=m
++CONFIG_NETFILTER_XT_MATCH_QUOTA=m
++CONFIG_NETFILTER_XT_MATCH_REALM=m
++CONFIG_NETFILTER_XT_MATCH_SCTP=m
++CONFIG_NETFILTER_XT_MATCH_STATE=m
++CONFIG_NETFILTER_XT_MATCH_STATISTIC=m
++CONFIG_NETFILTER_XT_MATCH_STRING=m
++CONFIG_NETFILTER_XT_MATCH_TCPMSS=m
+
-+static noinline long simnow(long cmd, long a, long b, long c)
-+{
-+ long ret;
-+ asm volatile("cpuid" :
-+ "=a" (ret) :
-+ "b" (a), "c" (b), "d" (c), "0" (MAGIC1), "D" (cmd + MAGIC2));
-+ return ret;
-+}
++#
++# IP: Netfilter Configuration
++#
++CONFIG_IP_NF_CONNTRACK=m
++CONFIG_IP_NF_CT_ACCT=y
++CONFIG_IP_NF_CONNTRACK_MARK=y
++CONFIG_IP_NF_CONNTRACK_SECMARK=y
++CONFIG_IP_NF_CONNTRACK_EVENTS=y
++CONFIG_IP_NF_CONNTRACK_NETLINK=m
++CONFIG_IP_NF_CT_PROTO_SCTP=m
++CONFIG_IP_NF_FTP=m
++CONFIG_IP_NF_IRC=m
++CONFIG_IP_NF_NETBIOS_NS=m
++CONFIG_IP_NF_TFTP=m
++CONFIG_IP_NF_AMANDA=m
++CONFIG_IP_NF_PPTP=m
++CONFIG_IP_NF_H323=m
++CONFIG_IP_NF_SIP=m
++CONFIG_IP_NF_QUEUE=m
++CONFIG_IP_NF_IPTABLES=m
++CONFIG_IP_NF_MATCH_IPRANGE=m
++CONFIG_IP_NF_MATCH_TOS=m
++CONFIG_IP_NF_MATCH_RECENT=m
++CONFIG_IP_NF_MATCH_ECN=m
++CONFIG_IP_NF_MATCH_DSCP=m
++CONFIG_IP_NF_MATCH_AH=m
++CONFIG_IP_NF_MATCH_TTL=m
++CONFIG_IP_NF_MATCH_OWNER=m
++CONFIG_IP_NF_MATCH_ADDRTYPE=m
++CONFIG_IP_NF_MATCH_HASHLIMIT=m
++CONFIG_IP_NF_FILTER=m
++CONFIG_IP_NF_TARGET_REJECT=m
++CONFIG_IP_NF_TARGET_LOG=m
++CONFIG_IP_NF_TARGET_ULOG=m
++CONFIG_IP_NF_TARGET_TCPMSS=m
++CONFIG_IP_NF_NAT=m
++CONFIG_IP_NF_NAT_NEEDED=y
++CONFIG_IP_NF_TARGET_MASQUERADE=m
++CONFIG_IP_NF_TARGET_REDIRECT=m
++CONFIG_IP_NF_TARGET_NETMAP=m
++CONFIG_IP_NF_TARGET_SAME=m
++CONFIG_IP_NF_NAT_SNMP_BASIC=m
++CONFIG_IP_NF_NAT_IRC=m
++CONFIG_IP_NF_NAT_FTP=m
++CONFIG_IP_NF_NAT_TFTP=m
++CONFIG_IP_NF_NAT_AMANDA=m
++CONFIG_IP_NF_NAT_PPTP=m
++CONFIG_IP_NF_NAT_H323=m
++CONFIG_IP_NF_NAT_SIP=m
++CONFIG_IP_NF_MANGLE=m
++CONFIG_IP_NF_TARGET_TOS=m
++CONFIG_IP_NF_TARGET_ECN=m
++CONFIG_IP_NF_TARGET_DSCP=m
++CONFIG_IP_NF_TARGET_TTL=m
++CONFIG_IP_NF_TARGET_CLUSTERIP=m
++CONFIG_IP_NF_RAW=m
++CONFIG_IP_NF_ARPTABLES=m
++CONFIG_IP_NF_ARPFILTER=m
++CONFIG_IP_NF_ARP_MANGLE=m
+
-+void __init simnow_init(char *str)
-+{
-+ char *fn = "klog";
-+ if (*str == '=')
-+ fn = ++str;
-+ /* error ignored */
-+ simnow_fd = simnow(XOPEN, (unsigned long)fn, O_WRONLY|O_APPEND|O_CREAT, 0644);
-+}
++#
++# IPv6: Netfilter Configuration (EXPERIMENTAL)
++#
++CONFIG_IP6_NF_QUEUE=m
++CONFIG_IP6_NF_IPTABLES=m
++CONFIG_IP6_NF_MATCH_RT=m
++CONFIG_IP6_NF_MATCH_OPTS=m
++CONFIG_IP6_NF_MATCH_FRAG=m
++CONFIG_IP6_NF_MATCH_HL=m
++CONFIG_IP6_NF_MATCH_OWNER=m
++CONFIG_IP6_NF_MATCH_IPV6HEADER=m
++CONFIG_IP6_NF_MATCH_AH=m
++CONFIG_IP6_NF_MATCH_EUI64=m
++CONFIG_IP6_NF_FILTER=m
++CONFIG_IP6_NF_TARGET_LOG=m
++CONFIG_IP6_NF_TARGET_REJECT=m
++CONFIG_IP6_NF_MANGLE=m
++CONFIG_IP6_NF_TARGET_HL=m
++CONFIG_IP6_NF_RAW=m
+
-+static void simnow_write(struct console *con, const char *s, unsigned n)
-+{
-+ simnow(XWRITE, simnow_fd, (unsigned long)s, n);
-+}
++#
++# DECnet: Netfilter Configuration
++#
++CONFIG_DECNET_NF_GRABULATOR=m
+
-+static struct console simnow_console = {
-+ .name = "simnow",
-+ .write = simnow_write,
-+ .flags = CON_PRINTBUFFER,
-+ .index = -1,
-+};
++#
++# Bridge: Netfilter Configuration
++#
++CONFIG_BRIDGE_NF_EBTABLES=m
++CONFIG_BRIDGE_EBT_BROUTE=m
++CONFIG_BRIDGE_EBT_T_FILTER=m
++CONFIG_BRIDGE_EBT_T_NAT=m
++CONFIG_BRIDGE_EBT_802_3=m
++CONFIG_BRIDGE_EBT_AMONG=m
++CONFIG_BRIDGE_EBT_ARP=m
++CONFIG_BRIDGE_EBT_IP=m
++CONFIG_BRIDGE_EBT_LIMIT=m
++CONFIG_BRIDGE_EBT_MARK=m
++CONFIG_BRIDGE_EBT_PKTTYPE=m
++CONFIG_BRIDGE_EBT_STP=m
++CONFIG_BRIDGE_EBT_VLAN=m
++CONFIG_BRIDGE_EBT_ARPREPLY=m
++CONFIG_BRIDGE_EBT_DNAT=m
++CONFIG_BRIDGE_EBT_MARK_T=m
++CONFIG_BRIDGE_EBT_REDIRECT=m
++CONFIG_BRIDGE_EBT_SNAT=m
++CONFIG_BRIDGE_EBT_LOG=m
++CONFIG_BRIDGE_EBT_ULOG=m
+
-+/* Direct interface for emergencies */
-+struct console *early_console = &early_vga_console;
-+static int early_console_initialized = 0;
++#
++# DCCP Configuration (EXPERIMENTAL)
++#
++CONFIG_IP_DCCP=m
++CONFIG_INET_DCCP_DIAG=m
++CONFIG_IP_DCCP_ACKVEC=y
+
-+void early_printk(const char *fmt, ...)
-+{
-+ char buf[512];
-+ int n;
-+ va_list ap;
++#
++# DCCP CCIDs Configuration (EXPERIMENTAL)
++#
++CONFIG_IP_DCCP_CCID2=m
++CONFIG_IP_DCCP_CCID3=m
++CONFIG_IP_DCCP_TFRC_LIB=m
+
-+ va_start(ap,fmt);
-+ n = vscnprintf(buf,512,fmt,ap);
-+ early_console->write(early_console,buf,n);
-+ va_end(ap);
-+}
++#
++# DCCP Kernel Hacking
++#
++# CONFIG_IP_DCCP_DEBUG is not set
+
-+static int __initdata keep_early;
++#
++# SCTP Configuration (EXPERIMENTAL)
++#
++CONFIG_IP_SCTP=m
++# CONFIG_SCTP_DBG_MSG is not set
++# CONFIG_SCTP_DBG_OBJCNT is not set
++# CONFIG_SCTP_HMAC_NONE is not set
++# CONFIG_SCTP_HMAC_SHA1 is not set
++CONFIG_SCTP_HMAC_MD5=y
+
-+int __init setup_early_printk(char *opt)
-+{
-+ char *space;
-+ char buf[256];
++#
++# TIPC Configuration (EXPERIMENTAL)
++#
++# CONFIG_TIPC is not set
++CONFIG_ATM=m
++CONFIG_ATM_CLIP=m
++CONFIG_ATM_CLIP_NO_ICMP=y
++CONFIG_ATM_LANE=m
++CONFIG_ATM_MPOA=m
++CONFIG_ATM_BR2684=m
++# CONFIG_ATM_BR2684_IPFILTER is not set
++CONFIG_BRIDGE=m
++CONFIG_VLAN_8021Q=m
++CONFIG_DECNET=m
++CONFIG_DECNET_ROUTER=y
++CONFIG_DECNET_ROUTE_FWMARK=y
++CONFIG_LLC=y
++CONFIG_LLC2=m
++CONFIG_IPX=m
++# CONFIG_IPX_INTERN is not set
++CONFIG_ATALK=m
++CONFIG_DEV_APPLETALK=m
++CONFIG_IPDDP=m
++CONFIG_IPDDP_ENCAP=y
++CONFIG_IPDDP_DECAP=y
++CONFIG_X25=m
++CONFIG_LAPB=m
++CONFIG_ECONET=m
++CONFIG_ECONET_AUNUDP=y
++CONFIG_ECONET_NATIVE=y
++CONFIG_WAN_ROUTER=m
+
-+ if (early_console_initialized)
-+ return 1;
++#
++# QoS and/or fair queueing
++#
++CONFIG_NET_SCHED=y
++CONFIG_NET_SCH_CLK_JIFFIES=y
++# CONFIG_NET_SCH_CLK_GETTIMEOFDAY is not set
++# CONFIG_NET_SCH_CLK_CPU is not set
+
-+ strlcpy(buf,opt,sizeof(buf));
-+ space = strchr(buf, ' ');
-+ if (space)
-+ *space = 0;
++#
++# Queueing/Scheduling
++#
++CONFIG_NET_SCH_CBQ=m
++CONFIG_NET_SCH_HTB=m
++CONFIG_NET_SCH_HFSC=m
++CONFIG_NET_SCH_ATM=m
++CONFIG_NET_SCH_PRIO=m
++CONFIG_NET_SCH_RED=m
++CONFIG_NET_SCH_SFQ=m
++CONFIG_NET_SCH_TEQL=m
++CONFIG_NET_SCH_TBF=m
++CONFIG_NET_SCH_GRED=m
++CONFIG_NET_SCH_DSMARK=m
++CONFIG_NET_SCH_NETEM=m
++CONFIG_NET_SCH_INGRESS=m
+
-+ if (strstr(buf,"keep"))
-+ keep_early = 1;
++#
++# Classification
++#
++CONFIG_NET_CLS=y
++CONFIG_NET_CLS_BASIC=m
++CONFIG_NET_CLS_TCINDEX=m
++CONFIG_NET_CLS_ROUTE4=m
++CONFIG_NET_CLS_ROUTE=y
++CONFIG_NET_CLS_FW=m
++CONFIG_NET_CLS_U32=m
++CONFIG_CLS_U32_PERF=y
++CONFIG_CLS_U32_MARK=y
++CONFIG_NET_CLS_RSVP=m
++CONFIG_NET_CLS_RSVP6=m
++CONFIG_NET_EMATCH=y
++CONFIG_NET_EMATCH_STACK=32
++CONFIG_NET_EMATCH_CMP=m
++CONFIG_NET_EMATCH_NBYTE=m
++CONFIG_NET_EMATCH_U32=m
++CONFIG_NET_EMATCH_META=m
++CONFIG_NET_EMATCH_TEXT=m
++# CONFIG_NET_CLS_ACT is not set
++CONFIG_NET_CLS_POLICE=y
++# CONFIG_NET_CLS_IND is not set
++CONFIG_NET_ESTIMATOR=y
+
-+ if (!strncmp(buf, "serial", 6)) {
-+ early_serial_init(buf + 6);
-+ early_console = &early_serial_console;
-+ } else if (!strncmp(buf, "ttyS", 4)) {
-+ early_serial_init(buf);
-+ early_console = &early_serial_console;
-+ } else if (!strncmp(buf, "vga", 3)
-+#ifndef CONFIG_XEN
-+ && SCREEN_INFO.orig_video_isVGA == 1) {
-+ max_xpos = SCREEN_INFO.orig_video_cols;
-+ max_ypos = SCREEN_INFO.orig_video_lines;
-+ current_ypos = SCREEN_INFO.orig_y;
-+#else
-+ || !strncmp(buf, "xen", 3)) {
-+#endif
-+ early_console = &early_vga_console;
-+ } else if (!strncmp(buf, "simnow", 6)) {
-+ simnow_init(buf + 6);
-+ early_console = &simnow_console;
-+ keep_early = 1;
-+ }
-+ early_console_initialized = 1;
-+ register_console(early_console);
-+ return 0;
-+}
++#
++# Network testing
++#
++CONFIG_NET_PKTGEN=m
++CONFIG_HAMRADIO=y
+
-+void __init disable_early_printk(void)
-+{
-+ if (!early_console_initialized || !early_console)
-+ return;
-+ if (!keep_early) {
-+ printk("disabling early console\n");
-+ unregister_console(early_console);
-+ early_console_initialized = 0;
-+ } else {
-+ printk("keeping early console\n");
-+ }
-+}
++#
++# Packet Radio protocols
++#
++CONFIG_AX25=m
++# CONFIG_AX25_DAMA_SLAVE is not set
++CONFIG_NETROM=m
++CONFIG_ROSE=m
+
-+__setup("earlyprintk=", setup_early_printk);
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/entry-xen.S tmp-linux-2.6-xen.patch/arch/x86_64/kernel/entry-xen.S
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/entry-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/entry-xen.S 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,1325 @@
-+/*
-+ * linux/arch/x86_64/entry.S
-+ *
-+ * Copyright (C) 1991, 1992 Linus Torvalds
-+ * Copyright (C) 2000, 2001, 2002 Andi Kleen SuSE Labs
-+ * Copyright (C) 2000 Pavel Machek <pavel at suse.cz>
-+ *
-+ * $Id$
-+ *
-+ * Jun Nakajima <jun.nakajima at intel.com>
-+ * Asit Mallick <asit.k.mallick at intel.com>
-+ * Modified for Xen
-+ */
++#
++# AX.25 network device drivers
++#
++CONFIG_MKISS=m
++CONFIG_6PACK=m
++CONFIG_BPQETHER=m
++CONFIG_BAYCOM_SER_FDX=m
++CONFIG_BAYCOM_SER_HDX=m
++CONFIG_BAYCOM_PAR=m
++CONFIG_BAYCOM_EPP=m
++CONFIG_YAM=m
++CONFIG_IRDA=m
+
-+/*
-+ * entry.S contains the system-call and fault low-level handling routines.
-+ *
-+ * NOTE: This code handles signal-recognition, which happens every time
-+ * after an interrupt and after each system call.
-+ *
-+ * Normal syscalls and interrupts don't save a full stack frame, this is
-+ * only done for syscall tracing, signals or fork/exec et.al.
-+ *
-+ * A note on terminology:
-+ * - top of stack: Architecture defined interrupt frame from SS to RIP
-+ * at the top of the kernel process stack.
-+ * - partial stack frame: partially saved registers upto R11.
-+ * - full stack frame: Like partial stack frame, but all register saved.
-+ *
-+ * TODO:
-+ * - schedule it carefully for the final hardware.
-+ */
++#
++# IrDA protocols
++#
++CONFIG_IRLAN=m
++CONFIG_IRNET=m
++CONFIG_IRCOMM=m
++# CONFIG_IRDA_ULTRA is not set
+
-+#define ASSEMBLY 1
-+#include <linux/linkage.h>
-+#include <asm/segment.h>
-+#include <asm/smp.h>
-+#include <asm/cache.h>
-+#include <asm/errno.h>
-+#include <asm/dwarf2.h>
-+#include <asm/calling.h>
-+#include <asm/asm-offsets.h>
-+#include <asm/msr.h>
-+#include <asm/unistd.h>
-+#include <asm/thread_info.h>
-+#include <asm/hw_irq.h>
-+#include <asm/page.h>
-+#include <asm/irqflags.h>
-+#include <asm/errno.h>
-+#include <xen/interface/arch-x86_64.h>
-+#include <xen/interface/features.h>
++#
++# IrDA options
++#
++CONFIG_IRDA_CACHE_LAST_LSAP=y
++CONFIG_IRDA_FAST_RR=y
++CONFIG_IRDA_DEBUG=y
+
-+#include "irq_vectors.h"
++#
++# Infrared-port device drivers
++#
+
-+#include "xen_entry.S"
-+
-+ .code64
++#
++# SIR device drivers
++#
++CONFIG_IRTTY_SIR=m
+
-+#ifndef CONFIG_PREEMPT
-+#define retint_kernel retint_restore_args
-+#endif
++#
++# Dongle support
++#
++CONFIG_DONGLE=y
++CONFIG_ESI_DONGLE=m
++CONFIG_ACTISYS_DONGLE=m
++CONFIG_TEKRAM_DONGLE=m
++CONFIG_TOIM3232_DONGLE=m
++CONFIG_LITELINK_DONGLE=m
++CONFIG_MA600_DONGLE=m
++CONFIG_GIRBIL_DONGLE=m
++CONFIG_MCP2120_DONGLE=m
++CONFIG_OLD_BELKIN_DONGLE=m
++CONFIG_ACT200L_DONGLE=m
+
++#
++# Old SIR device drivers
++#
+
-+.macro TRACE_IRQS_IRETQ offset=ARGOFFSET
-+#ifdef CONFIG_TRACE_IRQFLAGS
-+ bt $9,EFLAGS-\offset(%rsp) /* interrupts off? */
-+ jnc 1f
-+ TRACE_IRQS_ON
-+1:
-+#endif
-+.endm
++#
++# Old Serial dongle support
++#
+
-+NMI_MASK = 0x80000000
-+
-+/*
-+ * C code is not supposed to know about undefined top of stack. Every time
-+ * a C function with an pt_regs argument is called from the SYSCALL based
-+ * fast path FIXUP_TOP_OF_STACK is needed.
-+ * RESTORE_TOP_OF_STACK syncs the syscall state after any possible ptregs
-+ * manipulation.
-+ */
-+
-+ /* %rsp:at FRAMEEND */
-+ .macro FIXUP_TOP_OF_STACK tmp
-+ movq $__USER_CS,CS(%rsp)
-+ movq $-1,RCX(%rsp)
-+ .endm
++#
++# FIR device drivers
++#
++CONFIG_USB_IRDA=m
++CONFIG_SIGMATEL_FIR=m
++CONFIG_NSC_FIR=m
++CONFIG_WINBOND_FIR=m
++CONFIG_TOSHIBA_FIR=m
++CONFIG_SMC_IRCC_FIR=m
++CONFIG_ALI_FIR=m
++CONFIG_VLSI_FIR=m
++CONFIG_VIA_FIR=m
++CONFIG_MCS_FIR=m
++CONFIG_BT=m
++CONFIG_BT_L2CAP=m
++CONFIG_BT_SCO=m
++CONFIG_BT_RFCOMM=m
++CONFIG_BT_RFCOMM_TTY=y
++CONFIG_BT_BNEP=m
++CONFIG_BT_BNEP_MC_FILTER=y
++CONFIG_BT_BNEP_PROTO_FILTER=y
++CONFIG_BT_CMTP=m
++CONFIG_BT_HIDP=m
+
-+ .macro RESTORE_TOP_OF_STACK tmp,offset=0
-+ .endm
++#
++# Bluetooth device drivers
++#
++CONFIG_BT_HCIUSB=m
++CONFIG_BT_HCIUSB_SCO=y
++CONFIG_BT_HCIUART=m
++CONFIG_BT_HCIUART_H4=y
++CONFIG_BT_HCIUART_BCSP=y
++CONFIG_BT_HCIBCM203X=m
++CONFIG_BT_HCIBPA10X=m
++CONFIG_BT_HCIBFUSB=m
++CONFIG_BT_HCIDTL1=m
++CONFIG_BT_HCIBT3C=m
++CONFIG_BT_HCIBLUECARD=m
++CONFIG_BT_HCIBTUART=m
++CONFIG_BT_HCIVHCI=m
++CONFIG_IEEE80211=m
++# CONFIG_IEEE80211_DEBUG is not set
++CONFIG_IEEE80211_CRYPT_WEP=m
++CONFIG_IEEE80211_CRYPT_CCMP=m
++CONFIG_IEEE80211_CRYPT_TKIP=m
++CONFIG_IEEE80211_SOFTMAC=m
++# CONFIG_IEEE80211_SOFTMAC_DEBUG is not set
++CONFIG_WIRELESS_EXT=y
+
-+ .macro FAKE_STACK_FRAME child_rip
-+ /* push in order ss, rsp, eflags, cs, rip */
-+ xorl %eax, %eax
-+ pushq %rax /* ss */
-+ CFI_ADJUST_CFA_OFFSET 8
-+ /*CFI_REL_OFFSET ss,0*/
-+ pushq %rax /* rsp */
-+ CFI_ADJUST_CFA_OFFSET 8
-+ CFI_REL_OFFSET rsp,0
-+ pushq $(1<<9) /* eflags - interrupts on */
-+ CFI_ADJUST_CFA_OFFSET 8
-+ /*CFI_REL_OFFSET rflags,0*/
-+ pushq $__KERNEL_CS /* cs */
-+ CFI_ADJUST_CFA_OFFSET 8
-+ /*CFI_REL_OFFSET cs,0*/
-+ pushq \child_rip /* rip */
-+ CFI_ADJUST_CFA_OFFSET 8
-+ CFI_REL_OFFSET rip,0
-+ pushq %rax /* orig rax */
-+ CFI_ADJUST_CFA_OFFSET 8
-+ .endm
++#
++# Device Drivers
++#
+
-+ .macro UNFAKE_STACK_FRAME
-+ addq $8*6, %rsp
-+ CFI_ADJUST_CFA_OFFSET -(6*8)
-+ .endm
++#
++# Generic Driver Options
++#
++CONFIG_STANDALONE=y
++CONFIG_PREVENT_FIRMWARE_BUILD=y
++CONFIG_FW_LOADER=m
++# CONFIG_DEBUG_DRIVER is not set
++CONFIG_SYS_HYPERVISOR=y
+
-+ .macro CFI_DEFAULT_STACK start=1,adj=0
-+ .if \start
-+ CFI_STARTPROC simple
-+ CFI_DEF_CFA rsp,SS+8-(\adj*ARGOFFSET)
-+ .else
-+ CFI_DEF_CFA_OFFSET SS+8-(\adj*ARGOFFSET)
-+ .endif
-+ .if \adj == 0
-+ CFI_REL_OFFSET r15,R15
-+ CFI_REL_OFFSET r14,R14
-+ CFI_REL_OFFSET r13,R13
-+ CFI_REL_OFFSET r12,R12
-+ CFI_REL_OFFSET rbp,RBP
-+ CFI_REL_OFFSET rbx,RBX
-+ .endif
-+ CFI_REL_OFFSET r11,R11
-+ CFI_REL_OFFSET r10,R10
-+ CFI_REL_OFFSET r9,R9
-+ CFI_REL_OFFSET r8,R8
-+ CFI_REL_OFFSET rax,RAX
-+ CFI_REL_OFFSET rcx,RCX
-+ CFI_REL_OFFSET rdx,RDX
-+ CFI_REL_OFFSET rsi,RSI
-+ CFI_REL_OFFSET rdi,RDI
-+ CFI_REL_OFFSET rip,RIP
-+ /*CFI_REL_OFFSET cs,CS*/
-+ /*CFI_REL_OFFSET rflags,EFLAGS*/
-+ CFI_REL_OFFSET rsp,RSP
-+ /*CFI_REL_OFFSET ss,SS*/
-+ .endm
++#
++# Connector - unified userspace <-> kernelspace linker
++#
++CONFIG_CONNECTOR=m
+
-+ /*
-+ * Must be consistent with the definition in arch-x86/xen-x86_64.h:
-+ * struct iret_context {
-+ * u64 rax, r11, rcx, flags, rip, cs, rflags, rsp, ss;
-+ * };
-+ * with rax, r11, and rcx being taken care of in the hypercall stub.
-+ */
-+ .macro HYPERVISOR_IRET flag
-+ testb $3,1*8(%rsp)
-+ jnz 2f
-+ testl $NMI_MASK,2*8(%rsp)
-+ jnz 2f
++#
++# Memory Technology Devices (MTD)
++#
++CONFIG_MTD=m
++# CONFIG_MTD_DEBUG is not set
++CONFIG_MTD_CONCAT=m
++CONFIG_MTD_PARTITIONS=y
++CONFIG_MTD_REDBOOT_PARTS=m
++CONFIG_MTD_REDBOOT_DIRECTORY_BLOCK=-1
++# CONFIG_MTD_REDBOOT_PARTS_UNALLOCATED is not set
++# CONFIG_MTD_REDBOOT_PARTS_READONLY is not set
++# CONFIG_MTD_CMDLINE_PARTS is not set
+
-+ cmpb $0,(xen_features+XENFEAT_supervisor_mode_kernel)(%rip)
-+ jne 1f
++#
++# User Modules And Translation Layers
++#
++CONFIG_MTD_CHAR=m
++CONFIG_MTD_BLOCK=m
++CONFIG_MTD_BLOCK_RO=m
++CONFIG_FTL=m
++CONFIG_NFTL=m
++CONFIG_NFTL_RW=y
++CONFIG_INFTL=m
++CONFIG_RFD_FTL=m
+
-+ /* Direct iret to kernel space. Correct CS and SS. */
-+ orl $3,1*8(%rsp)
-+ orl $3,4*8(%rsp)
-+1: iretq
++#
++# RAM/ROM/Flash chip drivers
++#
++CONFIG_MTD_CFI=m
++CONFIG_MTD_JEDECPROBE=m
++CONFIG_MTD_GEN_PROBE=m
++CONFIG_MTD_CFI_ADV_OPTIONS=y
++CONFIG_MTD_CFI_NOSWAP=y
++# CONFIG_MTD_CFI_BE_BYTE_SWAP is not set
++# CONFIG_MTD_CFI_LE_BYTE_SWAP is not set
++# CONFIG_MTD_CFI_GEOMETRY is not set
++CONFIG_MTD_MAP_BANK_WIDTH_1=y
++CONFIG_MTD_MAP_BANK_WIDTH_2=y
++CONFIG_MTD_MAP_BANK_WIDTH_4=y
++# CONFIG_MTD_MAP_BANK_WIDTH_8 is not set
++# CONFIG_MTD_MAP_BANK_WIDTH_16 is not set
++# CONFIG_MTD_MAP_BANK_WIDTH_32 is not set
++CONFIG_MTD_CFI_I1=y
++CONFIG_MTD_CFI_I2=y
++# CONFIG_MTD_CFI_I4 is not set
++# CONFIG_MTD_CFI_I8 is not set
++# CONFIG_MTD_OTP is not set
++CONFIG_MTD_CFI_INTELEXT=m
++CONFIG_MTD_CFI_AMDSTD=m
++CONFIG_MTD_CFI_STAA=m
++CONFIG_MTD_CFI_UTIL=m
++CONFIG_MTD_RAM=m
++CONFIG_MTD_ROM=m
++CONFIG_MTD_ABSENT=m
++# CONFIG_MTD_OBSOLETE_CHIPS is not set
+
-+2: /* Slow iret via hypervisor. */
-+ andl $~NMI_MASK, 2*8(%rsp)
-+ pushq $\flag
-+ jmp hypercall_page + (__HYPERVISOR_iret * 32)
-+ .endm
++#
++# Mapping drivers for chip access
++#
++CONFIG_MTD_COMPLEX_MAPPINGS=y
++CONFIG_MTD_PHYSMAP=m
++CONFIG_MTD_PHYSMAP_START=0x8000000
++CONFIG_MTD_PHYSMAP_LEN=0x4000000
++CONFIG_MTD_PHYSMAP_BANKWIDTH=2
++CONFIG_MTD_PNC2000=m
++CONFIG_MTD_SC520CDP=m
++CONFIG_MTD_NETSC520=m
++CONFIG_MTD_TS5500=m
++CONFIG_MTD_SBC_GXX=m
++CONFIG_MTD_SCx200_DOCFLASH=m
++CONFIG_MTD_AMD76XROM=m
++CONFIG_MTD_ICHXROM=m
++CONFIG_MTD_SCB2_FLASH=m
++CONFIG_MTD_NETtel=m
++CONFIG_MTD_DILNETPC=m
++CONFIG_MTD_DILNETPC_BOOTSIZE=0x80000
++CONFIG_MTD_L440GX=m
++CONFIG_MTD_PCI=m
++CONFIG_MTD_PLATRAM=m
+
-+/*
-+ * A newly forked process directly context switches into this.
-+ */
-+/* rdi: prev */
-+ENTRY(ret_from_fork)
-+ CFI_DEFAULT_STACK
-+ call schedule_tail
-+ GET_THREAD_INFO(%rcx)
-+ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%rcx)
-+ jnz rff_trace
-+rff_action:
-+ RESTORE_REST
-+ testl $3,CS-ARGOFFSET(%rsp) # from kernel_thread?
-+ je int_ret_from_sys_call
-+ testl $_TIF_IA32,threadinfo_flags(%rcx)
-+ jnz int_ret_from_sys_call
-+ RESTORE_TOP_OF_STACK %rdi,ARGOFFSET
-+ jmp ret_from_sys_call
-+rff_trace:
-+ movq %rsp,%rdi
-+ call syscall_trace_leave
-+ GET_THREAD_INFO(%rcx)
-+ jmp rff_action
-+ CFI_ENDPROC
-+END(ret_from_fork)
++#
++# Self-contained MTD device drivers
++#
++CONFIG_MTD_PMC551=m
++# CONFIG_MTD_PMC551_BUGFIX is not set
++# CONFIG_MTD_PMC551_DEBUG is not set
++CONFIG_MTD_DATAFLASH=m
++CONFIG_MTD_M25P80=m
++CONFIG_MTD_SLRAM=m
++CONFIG_MTD_PHRAM=m
++CONFIG_MTD_MTDRAM=m
++CONFIG_MTDRAM_TOTAL_SIZE=4096
++CONFIG_MTDRAM_ERASE_SIZE=128
++CONFIG_MTD_BLOCK2MTD=m
+
-+/*
-+ * initial frame state for interrupts and exceptions
-+ */
-+ .macro _frame ref
-+ CFI_STARTPROC simple
-+ CFI_DEF_CFA rsp,SS+8-\ref
-+ /*CFI_REL_OFFSET ss,SS-\ref*/
-+ CFI_REL_OFFSET rsp,RSP-\ref
-+ /*CFI_REL_OFFSET rflags,EFLAGS-\ref*/
-+ /*CFI_REL_OFFSET cs,CS-\ref*/
-+ CFI_REL_OFFSET rip,RIP-\ref
-+ .endm
++#
++# Disk-On-Chip Device Drivers
++#
++CONFIG_MTD_DOC2000=m
++CONFIG_MTD_DOC2001=m
++CONFIG_MTD_DOC2001PLUS=m
++CONFIG_MTD_DOCPROBE=m
++CONFIG_MTD_DOCECC=m
++# CONFIG_MTD_DOCPROBE_ADVANCED is not set
++CONFIG_MTD_DOCPROBE_ADDRESS=0
+
-+/*
-+ * System call entry. Upto 6 arguments in registers are supported.
-+ *
-+ * SYSCALL does not save anything on the stack and does not change the
-+ * stack pointer.
-+ */
-+
-+/*
-+ * Register setup:
-+ * rax system call number
-+ * rdi arg0
-+ * rcx return address for syscall/sysret, C arg3
-+ * rsi arg1
-+ * rdx arg2
-+ * r10 arg3 (--> moved to rcx for C)
-+ * r8 arg4
-+ * r9 arg5
-+ * r11 eflags for syscall/sysret, temporary for C
-+ * r12-r15,rbp,rbx saved by C code, not touched.
-+ *
-+ * Interrupts are off on entry.
-+ * Only called from user space.
-+ *
-+ * XXX if we had a free scratch register we could save the RSP into the stack frame
-+ * and report it properly in ps. Unfortunately we haven't.
-+ *
-+ * When user can change the frames always force IRET. That is because
-+ * it deals with uncanonical addresses better. SYSRET has trouble
-+ * with them due to bugs in both AMD and Intel CPUs.
-+ */
++#
++# NAND Flash Device Drivers
++#
++CONFIG_MTD_NAND=m
++# CONFIG_MTD_NAND_VERIFY_WRITE is not set
++# CONFIG_MTD_NAND_ECC_SMC is not set
++CONFIG_MTD_NAND_IDS=m
++CONFIG_MTD_NAND_DISKONCHIP=m
++# CONFIG_MTD_NAND_DISKONCHIP_PROBE_ADVANCED is not set
++CONFIG_MTD_NAND_DISKONCHIP_PROBE_ADDRESS=0
++CONFIG_MTD_NAND_DISKONCHIP_BBTWRITE=y
++CONFIG_MTD_NAND_NANDSIM=m
+
-+ENTRY(system_call)
-+ _frame (RIP-0x10)
-+ SAVE_ARGS -8,0
-+ movq %rax,ORIG_RAX-ARGOFFSET(%rsp)
-+ /*
-+ * No need to follow this irqs off/on section - it's straight
-+ * and short:
-+ */
-+ XEN_UNBLOCK_EVENTS(%r11)
-+ GET_THREAD_INFO(%rcx)
-+ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%rcx)
-+ CFI_REMEMBER_STATE
-+ jnz tracesys
-+ cmpq $__NR_syscall_max,%rax
-+ ja badsys
-+ movq %r10,%rcx
-+ call *sys_call_table(,%rax,8) # XXX: rip relative
-+ movq %rax,RAX-ARGOFFSET(%rsp)
-+/*
-+ * Syscall return path ending with SYSRET (fast path)
-+ * Has incomplete stack frame and undefined top of stack.
-+ */
-+ .globl ret_from_sys_call
-+ret_from_sys_call:
-+ movl $_TIF_ALLWORK_MASK,%edi
-+ /* edi: flagmask */
-+sysret_check:
-+ GET_THREAD_INFO(%rcx)
-+ XEN_BLOCK_EVENTS(%rsi)
-+ TRACE_IRQS_OFF
-+ movl threadinfo_flags(%rcx),%edx
-+ andl %edi,%edx
-+ CFI_REMEMBER_STATE
-+ jnz sysret_careful
-+ /*
-+ * sysretq will re-enable interrupts:
-+ */
-+ TRACE_IRQS_ON
-+ XEN_UNBLOCK_EVENTS(%rsi)
-+ RESTORE_ARGS 0,8,0
-+ HYPERVISOR_IRET VGCF_IN_SYSCALL
++#
++# OneNAND Flash Device Drivers
++#
++CONFIG_MTD_ONENAND=m
++# CONFIG_MTD_ONENAND_VERIFY_WRITE is not set
++# CONFIG_MTD_ONENAND_OTP is not set
+
-+ /* Handle reschedules */
-+ /* edx: work, edi: workmask */
-+sysret_careful:
-+ CFI_RESTORE_STATE
-+ bt $TIF_NEED_RESCHED,%edx
-+ jnc sysret_signal
-+ TRACE_IRQS_ON
-+ XEN_UNBLOCK_EVENTS(%rsi)
-+ pushq %rdi
-+ CFI_ADJUST_CFA_OFFSET 8
-+ call schedule
-+ popq %rdi
-+ CFI_ADJUST_CFA_OFFSET -8
-+ jmp sysret_check
++#
++# Parallel port support
++#
++CONFIG_PARPORT=m
++CONFIG_PARPORT_PC=m
++CONFIG_PARPORT_SERIAL=m
++CONFIG_PARPORT_PC_FIFO=y
++CONFIG_PARPORT_PC_SUPERIO=y
++CONFIG_PARPORT_PC_PCMCIA=m
++CONFIG_PARPORT_NOT_PC=y
++# CONFIG_PARPORT_GSC is not set
++CONFIG_PARPORT_AX88796=m
++CONFIG_PARPORT_1284=y
+
-+ /* Handle a signal */
-+sysret_signal:
-+ TRACE_IRQS_ON
-+/* sti */
-+ XEN_UNBLOCK_EVENTS(%rsi)
-+ testl $(_TIF_SIGPENDING|_TIF_NOTIFY_RESUME|_TIF_SINGLESTEP),%edx
-+ jz 1f
++#
++# Plug and Play support
++#
++CONFIG_PNP=y
++# CONFIG_PNP_DEBUG is not set
+
-+ /* Really a signal */
-+ /* edx: work flags (arg3) */
-+ leaq do_notify_resume(%rip),%rax
-+ leaq -ARGOFFSET(%rsp),%rdi # &pt_regs -> arg1
-+ xorl %esi,%esi # oldset -> arg2
-+ call ptregscall_common
-+1: movl $_TIF_NEED_RESCHED,%edi
-+ /* Use IRET because user could have changed frame. This
-+ works because ptregscall_common has called FIXUP_TOP_OF_STACK. */
-+ XEN_BLOCK_EVENTS(%rsi)
-+ TRACE_IRQS_OFF
-+ jmp int_with_check
-+
-+badsys:
-+ movq $-ENOSYS,RAX-ARGOFFSET(%rsp)
-+ jmp ret_from_sys_call
++#
++# Protocols
++#
++CONFIG_PNPACPI=y
+
-+ /* Do syscall tracing */
-+tracesys:
-+ CFI_RESTORE_STATE
-+ SAVE_REST
-+ movq $-ENOSYS,RAX(%rsp)
-+ FIXUP_TOP_OF_STACK %rdi
-+ movq %rsp,%rdi
-+ call syscall_trace_enter
-+ LOAD_ARGS ARGOFFSET /* reload args from stack in case ptrace changed it */
-+ RESTORE_REST
-+ cmpq $__NR_syscall_max,%rax
-+ ja 1f
-+ movq %r10,%rcx /* fixup for C */
-+ call *sys_call_table(,%rax,8)
-+1: movq %rax,RAX-ARGOFFSET(%rsp)
-+ /* Use IRET because user could have changed frame */
-+ jmp int_ret_from_sys_call
-+ CFI_ENDPROC
-+END(system_call)
-+
-+/*
-+ * Syscall return path ending with IRET.
-+ * Has correct top of stack, but partial stack frame.
-+ */
-+ENTRY(int_ret_from_sys_call)
-+ CFI_STARTPROC simple
-+ CFI_DEF_CFA rsp,SS+8-ARGOFFSET
-+ /*CFI_REL_OFFSET ss,SS-ARGOFFSET*/
-+ CFI_REL_OFFSET rsp,RSP-ARGOFFSET
-+ /*CFI_REL_OFFSET rflags,EFLAGS-ARGOFFSET*/
-+ /*CFI_REL_OFFSET cs,CS-ARGOFFSET*/
-+ CFI_REL_OFFSET rip,RIP-ARGOFFSET
-+ CFI_REL_OFFSET rdx,RDX-ARGOFFSET
-+ CFI_REL_OFFSET rcx,RCX-ARGOFFSET
-+ CFI_REL_OFFSET rax,RAX-ARGOFFSET
-+ CFI_REL_OFFSET rdi,RDI-ARGOFFSET
-+ CFI_REL_OFFSET rsi,RSI-ARGOFFSET
-+ CFI_REL_OFFSET r8,R8-ARGOFFSET
-+ CFI_REL_OFFSET r9,R9-ARGOFFSET
-+ CFI_REL_OFFSET r10,R10-ARGOFFSET
-+ CFI_REL_OFFSET r11,R11-ARGOFFSET
-+ XEN_BLOCK_EVENTS(%rsi)
-+ TRACE_IRQS_OFF
-+ testb $3,CS-ARGOFFSET(%rsp)
-+ jnz 1f
-+ /* Need to set the proper %ss (not NULL) for ring 3 iretq */
-+ movl $__KERNEL_DS,SS-ARGOFFSET(%rsp)
-+ jmp retint_restore_args # retrun from ring3 kernel
-+1:
-+ movl $_TIF_ALLWORK_MASK,%edi
-+ /* edi: mask to check */
-+int_with_check:
-+ GET_THREAD_INFO(%rcx)
-+ movl threadinfo_flags(%rcx),%edx
-+ andl %edi,%edx
-+ jnz int_careful
-+ andl $~TS_COMPAT,threadinfo_status(%rcx)
-+ jmp retint_restore_args
++#
++# Block devices
++#
++CONFIG_BLK_DEV_FD=y
++CONFIG_PARIDE=m
++CONFIG_PARIDE_PARPORT=m
+
-+ /* Either reschedule or signal or syscall exit tracking needed. */
-+ /* First do a reschedule test. */
-+ /* edx: work, edi: workmask */
-+int_careful:
-+ bt $TIF_NEED_RESCHED,%edx
-+ jnc int_very_careful
-+ TRACE_IRQS_ON
-+/* sti */
-+ XEN_UNBLOCK_EVENTS(%rsi)
-+ pushq %rdi
-+ CFI_ADJUST_CFA_OFFSET 8
-+ call schedule
-+ popq %rdi
-+ CFI_ADJUST_CFA_OFFSET -8
-+ XEN_BLOCK_EVENTS(%rsi)
-+ TRACE_IRQS_OFF
-+ jmp int_with_check
++#
++# Parallel IDE high-level drivers
++#
++CONFIG_PARIDE_PD=m
++CONFIG_PARIDE_PCD=m
++CONFIG_PARIDE_PF=m
++CONFIG_PARIDE_PT=m
++CONFIG_PARIDE_PG=m
+
-+ /* handle signals and tracing -- both require a full stack frame */
-+int_very_careful:
-+ TRACE_IRQS_ON
-+/* sti */
-+ XEN_UNBLOCK_EVENTS(%rsi)
-+ SAVE_REST
-+ /* Check for syscall exit trace */
-+ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP),%edx
-+ jz int_signal
-+ pushq %rdi
-+ CFI_ADJUST_CFA_OFFSET 8
-+ leaq 8(%rsp),%rdi # &ptregs -> arg1
-+ call syscall_trace_leave
-+ popq %rdi
-+ CFI_ADJUST_CFA_OFFSET -8
-+ andl $~(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP),%edi
-+ XEN_BLOCK_EVENTS(%rsi)
-+ TRACE_IRQS_OFF
-+ jmp int_restore_rest
-+
-+int_signal:
-+ testl $(_TIF_NOTIFY_RESUME|_TIF_SIGPENDING|_TIF_SINGLESTEP),%edx
-+ jz 1f
-+ movq %rsp,%rdi # &ptregs -> arg1
-+ xorl %esi,%esi # oldset -> arg2
-+ call do_notify_resume
-+1: movl $_TIF_NEED_RESCHED,%edi
-+int_restore_rest:
-+ RESTORE_REST
-+ XEN_BLOCK_EVENTS(%rsi)
-+ TRACE_IRQS_OFF
-+ jmp int_with_check
-+ CFI_ENDPROC
-+END(int_ret_from_sys_call)
-+
-+/*
-+ * Certain special system calls that need to save a complete full stack frame.
-+ */
-+
-+ .macro PTREGSCALL label,func,arg
-+ .globl \label
-+\label:
-+ leaq \func(%rip),%rax
-+ leaq -ARGOFFSET+8(%rsp),\arg /* 8 for return address */
-+ jmp ptregscall_common
-+END(\label)
-+ .endm
++#
++# Parallel IDE protocol modules
++#
++CONFIG_PARIDE_ATEN=m
++CONFIG_PARIDE_BPCK=m
++CONFIG_PARIDE_BPCK6=m
++CONFIG_PARIDE_COMM=m
++CONFIG_PARIDE_DSTR=m
++CONFIG_PARIDE_FIT2=m
++CONFIG_PARIDE_FIT3=m
++CONFIG_PARIDE_EPAT=m
++CONFIG_PARIDE_EPATC8=y
++CONFIG_PARIDE_EPIA=m
++CONFIG_PARIDE_FRIQ=m
++CONFIG_PARIDE_FRPW=m
++CONFIG_PARIDE_KBIC=m
++CONFIG_PARIDE_KTTI=m
++CONFIG_PARIDE_ON20=m
++CONFIG_PARIDE_ON26=m
++CONFIG_BLK_CPQ_DA=m
++CONFIG_BLK_CPQ_CISS_DA=m
++CONFIG_CISS_SCSI_TAPE=y
++CONFIG_BLK_DEV_DAC960=m
++CONFIG_BLK_DEV_UMEM=m
++# CONFIG_BLK_DEV_COW_COMMON is not set
++CONFIG_BLK_DEV_LOOP=y
++CONFIG_BLK_DEV_CRYPTOLOOP=m
++CONFIG_BLK_DEV_NBD=m
++CONFIG_BLK_DEV_SX8=m
++# CONFIG_BLK_DEV_UB is not set
++CONFIG_BLK_DEV_RAM=y
++CONFIG_BLK_DEV_RAM_COUNT=16
++CONFIG_BLK_DEV_RAM_SIZE=16384
++CONFIG_BLK_DEV_RAM_BLOCKSIZE=1024
++CONFIG_BLK_DEV_INITRD=y
++CONFIG_CDROM_PKTCDVD=m
++CONFIG_CDROM_PKTCDVD_BUFFERS=8
++CONFIG_CDROM_PKTCDVD_WCACHE=y
++CONFIG_ATA_OVER_ETH=m
+
-+ CFI_STARTPROC
++#
++# ATA/ATAPI/MFM/RLL support
++#
++CONFIG_IDE=y
++CONFIG_BLK_DEV_IDE=y
+
-+ PTREGSCALL stub_clone, sys_clone, %r8
-+ PTREGSCALL stub_fork, sys_fork, %rdi
-+ PTREGSCALL stub_vfork, sys_vfork, %rdi
-+ PTREGSCALL stub_rt_sigsuspend, sys_rt_sigsuspend, %rdx
-+ PTREGSCALL stub_sigaltstack, sys_sigaltstack, %rdx
-+ PTREGSCALL stub_iopl, sys_iopl, %rsi
++#
++# Please see Documentation/ide.txt for help/info on IDE drives
++#
++# CONFIG_BLK_DEV_IDE_SATA is not set
++# CONFIG_BLK_DEV_HD_IDE is not set
++CONFIG_BLK_DEV_IDEDISK=m
++CONFIG_IDEDISK_MULTI_MODE=y
++CONFIG_BLK_DEV_IDECS=m
++CONFIG_BLK_DEV_IDECD=m
++CONFIG_BLK_DEV_IDETAPE=m
++CONFIG_BLK_DEV_IDEFLOPPY=m
++CONFIG_BLK_DEV_IDESCSI=m
++# CONFIG_IDE_TASK_IOCTL is not set
+
-+ENTRY(ptregscall_common)
-+ popq %r11
-+ CFI_ADJUST_CFA_OFFSET -8
-+ CFI_REGISTER rip, r11
-+ SAVE_REST
-+ movq %r11, %r15
-+ CFI_REGISTER rip, r15
-+ FIXUP_TOP_OF_STACK %r11
-+ call *%rax
-+ RESTORE_TOP_OF_STACK %r11
-+ movq %r15, %r11
-+ CFI_REGISTER rip, r11
-+ RESTORE_REST
-+ pushq %r11
-+ CFI_ADJUST_CFA_OFFSET 8
-+ CFI_REL_OFFSET rip, 0
-+ ret
-+ CFI_ENDPROC
-+END(ptregscall_common)
-+
-+ENTRY(stub_execve)
-+ CFI_STARTPROC
-+ popq %r11
-+ CFI_ADJUST_CFA_OFFSET -8
-+ CFI_REGISTER rip, r11
-+ SAVE_REST
-+ FIXUP_TOP_OF_STACK %r11
-+ call sys_execve
-+ RESTORE_TOP_OF_STACK %r11
-+ movq %rax,RAX(%rsp)
-+ RESTORE_REST
-+ jmp int_ret_from_sys_call
-+ CFI_ENDPROC
-+END(stub_execve)
-+
-+/*
-+ * sigreturn is special because it needs to restore all registers on return.
-+ * This cannot be done with SYSRET, so use the IRET return path instead.
-+ */
-+ENTRY(stub_rt_sigreturn)
-+ CFI_STARTPROC
-+ addq $8, %rsp
-+ CFI_ADJUST_CFA_OFFSET -8
-+ SAVE_REST
-+ movq %rsp,%rdi
-+ FIXUP_TOP_OF_STACK %r11
-+ call sys_rt_sigreturn
-+ movq %rax,RAX(%rsp) # fixme, this could be done at the higher layer
-+ RESTORE_REST
-+ jmp int_ret_from_sys_call
-+ CFI_ENDPROC
-+END(stub_rt_sigreturn)
++#
++# IDE chipset support/bugfixes
++#
++CONFIG_IDE_GENERIC=m
++CONFIG_BLK_DEV_CMD640=y
++CONFIG_BLK_DEV_CMD640_ENHANCED=y
++CONFIG_BLK_DEV_IDEPNP=y
++CONFIG_BLK_DEV_IDEPCI=y
++CONFIG_IDEPCI_SHARE_IRQ=y
++CONFIG_BLK_DEV_OFFBOARD=y
++CONFIG_BLK_DEV_GENERIC=y
++CONFIG_BLK_DEV_OPTI621=m
++CONFIG_BLK_DEV_RZ1000=m
++CONFIG_BLK_DEV_IDEDMA_PCI=y
++# CONFIG_BLK_DEV_IDEDMA_FORCED is not set
++CONFIG_IDEDMA_PCI_AUTO=y
++# CONFIG_IDEDMA_ONLYDISK is not set
++CONFIG_BLK_DEV_AEC62XX=m
++CONFIG_BLK_DEV_ALI15X3=m
++# CONFIG_WDC_ALI15X3 is not set
++CONFIG_BLK_DEV_AMD74XX=m
++CONFIG_BLK_DEV_ATIIXP=m
++CONFIG_BLK_DEV_CMD64X=m
++CONFIG_BLK_DEV_TRIFLEX=m
++CONFIG_BLK_DEV_CY82C693=m
++CONFIG_BLK_DEV_CS5520=m
++CONFIG_BLK_DEV_CS5530=m
++CONFIG_BLK_DEV_CS5535=m
++CONFIG_BLK_DEV_HPT34X=m
++CONFIG_HPT34X_AUTODMA=y
++CONFIG_BLK_DEV_HPT366=m
++CONFIG_BLK_DEV_SC1200=m
++CONFIG_BLK_DEV_PIIX=m
++CONFIG_BLK_DEV_IT821X=m
++CONFIG_BLK_DEV_NS87415=m
++CONFIG_BLK_DEV_PDC202XX_OLD=m
++CONFIG_PDC202XX_BURST=y
++CONFIG_BLK_DEV_PDC202XX_NEW=m
++CONFIG_BLK_DEV_SVWKS=m
++CONFIG_BLK_DEV_SIIMAGE=m
++CONFIG_BLK_DEV_SIS5513=m
++CONFIG_BLK_DEV_SLC90E66=m
++CONFIG_BLK_DEV_TRM290=m
++CONFIG_BLK_DEV_VIA82CXXX=m
++# CONFIG_IDE_ARM is not set
++CONFIG_BLK_DEV_IDEDMA=y
++# CONFIG_IDEDMA_IVB is not set
++CONFIG_IDEDMA_AUTO=y
++# CONFIG_BLK_DEV_HD is not set
+
-+/* initial frame state for interrupts (and exceptions without error code) */
-+#define INTR_FRAME _frame (RIP-0x10); \
-+ CFI_REL_OFFSET rcx,0; \
-+ CFI_REL_OFFSET r11,8
++#
++# SCSI device support
++#
++CONFIG_RAID_ATTRS=m
++CONFIG_SCSI=m
++CONFIG_SCSI_PROC_FS=y
+
-+/* initial frame state for exceptions with error code (and interrupts with
-+ vector already pushed) */
-+#define XCPT_FRAME _frame (RIP-0x18); \
-+ CFI_REL_OFFSET rcx,0; \
-+ CFI_REL_OFFSET r11,8
++#
++# SCSI support type (disk, tape, CD-ROM)
++#
++CONFIG_BLK_DEV_SD=m
++CONFIG_CHR_DEV_ST=m
++CONFIG_CHR_DEV_OSST=m
++CONFIG_BLK_DEV_SR=m
++# CONFIG_BLK_DEV_SR_VENDOR is not set
++CONFIG_CHR_DEV_SG=m
++CONFIG_CHR_DEV_SCH=m
+
-+/*
-+ * Interrupt exit.
-+ *
-+ */
++#
++# Some SCSI devices (e.g. CD jukebox) support multiple LUNs
++#
++CONFIG_SCSI_MULTI_LUN=y
++CONFIG_SCSI_CONSTANTS=y
++CONFIG_SCSI_LOGGING=y
+
-+retint_check:
-+ CFI_DEFAULT_STACK adj=1
-+ movl threadinfo_flags(%rcx),%edx
-+ andl %edi,%edx
-+ CFI_REMEMBER_STATE
-+ jnz retint_careful
-+retint_restore_args:
-+ movl EFLAGS-REST_SKIP(%rsp), %eax
-+ shr $9, %eax # EAX[0] == IRET_EFLAGS.IF
-+ XEN_GET_VCPU_INFO(%rsi)
-+ andb evtchn_upcall_mask(%rsi),%al
-+ andb $1,%al # EAX[0] == IRET_EFLAGS.IF & event_mask
-+ jnz restore_all_enable_events # != 0 => enable event delivery
-+ XEN_PUT_VCPU_INFO(%rsi)
-+
-+ RESTORE_ARGS 0,8,0
-+ HYPERVISOR_IRET 0
-+
-+ /* edi: workmask, edx: work */
-+retint_careful:
-+ CFI_RESTORE_STATE
-+ bt $TIF_NEED_RESCHED,%edx
-+ jnc retint_signal
-+ TRACE_IRQS_ON
-+ XEN_UNBLOCK_EVENTS(%rsi)
-+/* sti */
-+ pushq %rdi
-+ CFI_ADJUST_CFA_OFFSET 8
-+ call schedule
-+ popq %rdi
-+ CFI_ADJUST_CFA_OFFSET -8
-+ GET_THREAD_INFO(%rcx)
-+ XEN_BLOCK_EVENTS(%rsi)
-+/* cli */
-+ TRACE_IRQS_OFF
-+ jmp retint_check
-+
-+retint_signal:
-+ testl $(_TIF_SIGPENDING|_TIF_NOTIFY_RESUME|_TIF_SINGLESTEP),%edx
-+ jz retint_restore_args
-+ TRACE_IRQS_ON
-+ XEN_UNBLOCK_EVENTS(%rsi)
-+ SAVE_REST
-+ movq $-1,ORIG_RAX(%rsp)
-+ xorl %esi,%esi # oldset
-+ movq %rsp,%rdi # &pt_regs
-+ call do_notify_resume
-+ RESTORE_REST
-+ XEN_BLOCK_EVENTS(%rsi)
-+ TRACE_IRQS_OFF
-+ movl $_TIF_NEED_RESCHED,%edi
-+ GET_THREAD_INFO(%rcx)
-+ jmp retint_check
++#
++# SCSI Transport Attributes
++#
++CONFIG_SCSI_SPI_ATTRS=m
++CONFIG_SCSI_FC_ATTRS=m
++CONFIG_SCSI_ISCSI_ATTRS=m
++CONFIG_SCSI_SAS_ATTRS=m
+
-+#ifdef CONFIG_PREEMPT
-+ /* Returning to kernel space. Check if we need preemption */
-+ /* rcx: threadinfo. interrupts off. */
-+ .p2align
-+retint_kernel:
-+ cmpl $0,threadinfo_preempt_count(%rcx)
-+ jnz retint_restore_args
-+ bt $TIF_NEED_RESCHED,threadinfo_flags(%rcx)
-+ jnc retint_restore_args
-+ bt $9,EFLAGS-ARGOFFSET(%rsp) /* interrupts off? */
-+ jnc retint_restore_args
-+ call preempt_schedule_irq
-+ jmp retint_kernel /* check again */
-+#endif
++#
++# SCSI low-level drivers
++#
++CONFIG_ISCSI_TCP=m
++CONFIG_BLK_DEV_3W_XXXX_RAID=m
++CONFIG_SCSI_3W_9XXX=m
++CONFIG_SCSI_ACARD=m
++CONFIG_SCSI_AACRAID=m
++CONFIG_SCSI_AIC7XXX=m
++CONFIG_AIC7XXX_CMDS_PER_DEVICE=8
++CONFIG_AIC7XXX_RESET_DELAY_MS=15000
++CONFIG_AIC7XXX_DEBUG_ENABLE=y
++CONFIG_AIC7XXX_DEBUG_MASK=0
++CONFIG_AIC7XXX_REG_PRETTY_PRINT=y
++CONFIG_SCSI_AIC7XXX_OLD=m
++CONFIG_SCSI_AIC79XX=m
++CONFIG_AIC79XX_CMDS_PER_DEVICE=32
++CONFIG_AIC79XX_RESET_DELAY_MS=15000
++CONFIG_AIC79XX_ENABLE_RD_STRM=y
++CONFIG_AIC79XX_DEBUG_ENABLE=y
++CONFIG_AIC79XX_DEBUG_MASK=0
++CONFIG_AIC79XX_REG_PRETTY_PRINT=y
++CONFIG_SCSI_DPT_I2O=m
++CONFIG_SCSI_ADVANSYS=m
++CONFIG_MEGARAID_NEWGEN=y
++CONFIG_MEGARAID_MM=m
++CONFIG_MEGARAID_MAILBOX=m
++CONFIG_MEGARAID_LEGACY=m
++CONFIG_MEGARAID_SAS=m
++CONFIG_SCSI_SATA=m
++CONFIG_SCSI_SATA_AHCI=m
++CONFIG_SCSI_SATA_SVW=m
++CONFIG_SCSI_ATA_PIIX=m
++CONFIG_SCSI_SATA_MV=m
++CONFIG_SCSI_SATA_NV=m
++CONFIG_SCSI_PDC_ADMA=m
++CONFIG_SCSI_HPTIOP=m
++CONFIG_SCSI_SATA_QSTOR=m
++CONFIG_SCSI_SATA_PROMISE=m
++CONFIG_SCSI_SATA_SX4=m
++CONFIG_SCSI_SATA_SIL=m
++CONFIG_SCSI_SATA_SIL24=m
++CONFIG_SCSI_SATA_SIS=m
++CONFIG_SCSI_SATA_ULI=m
++CONFIG_SCSI_SATA_VIA=m
++CONFIG_SCSI_SATA_VITESSE=m
++CONFIG_SCSI_SATA_INTEL_COMBINED=y
++CONFIG_SCSI_BUSLOGIC=m
++# CONFIG_SCSI_OMIT_FLASHPOINT is not set
++CONFIG_SCSI_DMX3191D=m
++CONFIG_SCSI_EATA=m
++CONFIG_SCSI_EATA_TAGGED_QUEUE=y
++CONFIG_SCSI_EATA_LINKED_COMMANDS=y
++CONFIG_SCSI_EATA_MAX_TAGS=16
++CONFIG_SCSI_FUTURE_DOMAIN=m
++CONFIG_SCSI_GDTH=m
++CONFIG_SCSI_IPS=m
++CONFIG_SCSI_INITIO=m
++CONFIG_SCSI_INIA100=m
++CONFIG_SCSI_PPA=m
++CONFIG_SCSI_IMM=m
++# CONFIG_SCSI_IZIP_EPP16 is not set
++# CONFIG_SCSI_IZIP_SLOW_CTR is not set
++CONFIG_SCSI_SYM53C8XX_2=m
++CONFIG_SCSI_SYM53C8XX_DMA_ADDRESSING_MODE=1
++CONFIG_SCSI_SYM53C8XX_DEFAULT_TAGS=16
++CONFIG_SCSI_SYM53C8XX_MAX_TAGS=64
++CONFIG_SCSI_SYM53C8XX_MMIO=y
++CONFIG_SCSI_IPR=m
++CONFIG_SCSI_IPR_TRACE=y
++CONFIG_SCSI_IPR_DUMP=y
++CONFIG_SCSI_QLOGIC_1280=m
++CONFIG_SCSI_QLA_FC=m
++CONFIG_SCSI_LPFC=m
++CONFIG_SCSI_DC395x=m
++CONFIG_SCSI_DC390T=m
++CONFIG_SCSI_NSP32=m
++CONFIG_SCSI_DEBUG=m
+
-+ CFI_ENDPROC
-+END(retint_check)
-+
-+#ifndef CONFIG_XEN
-+/*
-+ * APIC interrupts.
-+ */
-+ .macro apicinterrupt num,func
-+ INTR_FRAME
-+ pushq $~(\num)
-+ CFI_ADJUST_CFA_OFFSET 8
-+ interrupt \func
-+ jmp error_entry
-+ CFI_ENDPROC
-+ .endm
++#
++# PCMCIA SCSI adapter support
++#
++CONFIG_PCMCIA_AHA152X=m
++CONFIG_PCMCIA_FDOMAIN=m
++CONFIG_PCMCIA_NINJA_SCSI=m
++CONFIG_PCMCIA_QLOGIC=m
++CONFIG_PCMCIA_SYM53C500=m
+
-+ENTRY(thermal_interrupt)
-+ apicinterrupt THERMAL_APIC_VECTOR,smp_thermal_interrupt
-+END(thermal_interrupt)
++#
++# Multi-device support (RAID and LVM)
++#
++CONFIG_MD=y
++CONFIG_BLK_DEV_MD=y
++CONFIG_MD_LINEAR=m
++CONFIG_MD_RAID0=m
++CONFIG_MD_RAID1=m
++CONFIG_MD_RAID10=m
++CONFIG_MD_RAID456=m
++CONFIG_MD_RAID5_RESHAPE=y
++CONFIG_MD_MULTIPATH=m
++CONFIG_MD_FAULTY=m
++CONFIG_BLK_DEV_DM=m
++CONFIG_DM_CRYPT=m
++CONFIG_DM_SNAPSHOT=m
++CONFIG_DM_MIRROR=m
++CONFIG_DM_ZERO=m
++CONFIG_DM_MULTIPATH=m
++CONFIG_DM_MULTIPATH_EMC=m
+
-+ENTRY(threshold_interrupt)
-+ apicinterrupt THRESHOLD_APIC_VECTOR,mce_threshold_interrupt
-+END(threshold_interrupt)
++#
++# Fusion MPT device support
++#
++CONFIG_FUSION=y
++CONFIG_FUSION_SPI=m
++CONFIG_FUSION_FC=m
++CONFIG_FUSION_SAS=m
++CONFIG_FUSION_MAX_SGE=128
++CONFIG_FUSION_CTL=m
++CONFIG_FUSION_LAN=m
+
-+#ifdef CONFIG_SMP
-+ENTRY(reschedule_interrupt)
-+ apicinterrupt RESCHEDULE_VECTOR,smp_reschedule_interrupt
-+END(reschedule_interrupt)
++#
++# IEEE 1394 (FireWire) support
++#
++CONFIG_IEEE1394=m
+
-+ .macro INVALIDATE_ENTRY num
-+ENTRY(invalidate_interrupt\num)
-+ apicinterrupt INVALIDATE_TLB_VECTOR_START+\num,smp_invalidate_interrupt
-+END(invalidate_interrupt\num)
-+ .endm
++#
++# Subsystem Options
++#
++# CONFIG_IEEE1394_VERBOSEDEBUG is not set
++# CONFIG_IEEE1394_OUI_DB is not set
++CONFIG_IEEE1394_EXTRA_CONFIG_ROMS=y
++CONFIG_IEEE1394_CONFIG_ROM_IP1394=y
++CONFIG_IEEE1394_EXPORT_FULL_API=y
+
-+ INVALIDATE_ENTRY 0
-+ INVALIDATE_ENTRY 1
-+ INVALIDATE_ENTRY 2
-+ INVALIDATE_ENTRY 3
-+ INVALIDATE_ENTRY 4
-+ INVALIDATE_ENTRY 5
-+ INVALIDATE_ENTRY 6
-+ INVALIDATE_ENTRY 7
++#
++# Device Drivers
++#
++CONFIG_IEEE1394_PCILYNX=m
++CONFIG_IEEE1394_OHCI1394=m
+
-+ENTRY(call_function_interrupt)
-+ apicinterrupt CALL_FUNCTION_VECTOR,smp_call_function_interrupt
-+END(call_function_interrupt)
-+#endif
++#
++# Protocol Drivers
++#
++CONFIG_IEEE1394_VIDEO1394=m
++CONFIG_IEEE1394_SBP2=m
++# CONFIG_IEEE1394_SBP2_PHYS_DMA is not set
++CONFIG_IEEE1394_ETH1394=m
++CONFIG_IEEE1394_DV1394=m
++CONFIG_IEEE1394_RAWIO=m
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ENTRY(apic_timer_interrupt)
-+ apicinterrupt LOCAL_TIMER_VECTOR,smp_apic_timer_interrupt
-+END(apic_timer_interrupt)
++#
++# I2O device support
++#
++CONFIG_I2O=m
++CONFIG_I2O_LCT_NOTIFY_ON_CHANGES=y
++CONFIG_I2O_EXT_ADAPTEC=y
++CONFIG_I2O_EXT_ADAPTEC_DMA64=y
++CONFIG_I2O_CONFIG=m
++CONFIG_I2O_CONFIG_OLD_IOCTL=y
++CONFIG_I2O_BUS=m
++CONFIG_I2O_BLOCK=m
++CONFIG_I2O_SCSI=m
++CONFIG_I2O_PROC=m
+
-+ENTRY(error_interrupt)
-+ apicinterrupt ERROR_APIC_VECTOR,smp_error_interrupt
-+END(error_interrupt)
++#
++# Network device support
++#
++CONFIG_NETDEVICES=y
++CONFIG_DUMMY=m
++CONFIG_BONDING=m
++CONFIG_EQUALIZER=m
++CONFIG_TUN=m
++CONFIG_NET_SB1000=m
+
-+ENTRY(spurious_interrupt)
-+ apicinterrupt SPURIOUS_APIC_VECTOR,smp_spurious_interrupt
-+END(spurious_interrupt)
-+#endif
-+#endif /* !CONFIG_XEN */
-+
-+/*
-+ * Exception entry points.
-+ */
-+ .macro zeroentry sym
-+ INTR_FRAME
-+ movq (%rsp),%rcx
-+ CFI_RESTORE rcx
-+ movq 8(%rsp),%r11
-+ CFI_RESTORE r11
-+ addq $0x10,%rsp /* skip rcx and r11 */
-+ CFI_ADJUST_CFA_OFFSET -0x10
-+ pushq $0 /* push error code/oldrax */
-+ CFI_ADJUST_CFA_OFFSET 8
-+ pushq %rax /* push real oldrax to the rdi slot */
-+ CFI_ADJUST_CFA_OFFSET 8
-+ CFI_REL_OFFSET rax,0
-+ leaq \sym(%rip),%rax
-+ jmp error_entry
-+ CFI_ENDPROC
-+ .endm
++#
++# ARCnet devices
++#
++CONFIG_ARCNET=m
++CONFIG_ARCNET_1201=m
++CONFIG_ARCNET_1051=m
++CONFIG_ARCNET_RAW=m
++CONFIG_ARCNET_CAP=m
++CONFIG_ARCNET_COM90xx=m
++CONFIG_ARCNET_COM90xxIO=m
++CONFIG_ARCNET_RIM_I=m
++# CONFIG_ARCNET_COM20020 is not set
+
-+ .macro errorentry sym
-+ XCPT_FRAME
-+ movq (%rsp),%rcx
-+ CFI_RESTORE rcx
-+ movq 8(%rsp),%r11
-+ CFI_RESTORE r11
-+ addq $0x10,%rsp /* rsp points to the error code */
-+ CFI_ADJUST_CFA_OFFSET -0x10
-+ pushq %rax
-+ CFI_ADJUST_CFA_OFFSET 8
-+ CFI_REL_OFFSET rax,0
-+ leaq \sym(%rip),%rax
-+ jmp error_entry
-+ CFI_ENDPROC
-+ .endm
++#
++# PHY device support
++#
++CONFIG_PHYLIB=m
+
-+#if 0 /* not XEN */
-+ /* error code is on the stack already */
-+ /* handle NMI like exceptions that can happen everywhere */
-+ .macro paranoidentry sym, ist=0, irqtrace=1
-+ movq (%rsp),%rcx
-+ movq 8(%rsp),%r11
-+ addq $0x10,%rsp /* skip rcx and r11 */
-+ SAVE_ALL
-+ cld
-+#if 0 /* not XEN */
-+ movl $1,%ebx
-+ movl $MSR_GS_BASE,%ecx
-+ rdmsr
-+ testl %edx,%edx
-+ js 1f
-+ swapgs
-+ xorl %ebx,%ebx
-+1:
-+#endif
-+ .if \ist
-+ movq %gs:pda_data_offset, %rbp
-+ .endif
-+ movq %rsp,%rdi
-+ movq ORIG_RAX(%rsp),%rsi
-+ movq $-1,ORIG_RAX(%rsp)
-+ .if \ist
-+ subq $EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp)
-+ .endif
-+ call \sym
-+ .if \ist
-+ addq $EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp)
-+ .endif
-+/* cli */
-+ XEN_BLOCK_EVENTS(%rsi)
-+ .if \irqtrace
-+ TRACE_IRQS_OFF
-+ .endif
-+ .endm
++#
++# MII PHY device drivers
++#
++CONFIG_MARVELL_PHY=m
++CONFIG_DAVICOM_PHY=m
++CONFIG_QSEMI_PHY=m
++CONFIG_LXT_PHY=m
++CONFIG_CICADA_PHY=m
++CONFIG_VITESSE_PHY=m
++CONFIG_SMSC_PHY=m
++CONFIG_FIXED_PHY=m
++# CONFIG_FIXED_MII_10_FDX is not set
++# CONFIG_FIXED_MII_100_FDX is not set
+
-+ /*
-+ * "Paranoid" exit path from exception stack.
-+ * Paranoid because this is used by NMIs and cannot take
-+ * any kernel state for granted.
-+ * We don't do kernel preemption checks here, because only
-+ * NMI should be common and it does not enable IRQs and
-+ * cannot get reschedule ticks.
-+ *
-+ * "trace" is 0 for the NMI handler only, because irq-tracing
-+ * is fundamentally NMI-unsafe. (we cannot change the soft and
-+ * hard flags at once, atomically)
-+ */
-+ .macro paranoidexit trace=1
-+ /* ebx: no swapgs flag */
-+paranoid_exit\trace:
-+ testl %ebx,%ebx /* swapgs needed? */
-+ jnz paranoid_restore\trace
-+ testl $3,CS(%rsp)
-+ jnz paranoid_userspace\trace
-+paranoid_swapgs\trace:
-+ TRACE_IRQS_IRETQ 0
-+ swapgs
-+paranoid_restore\trace:
-+ RESTORE_ALL 8
-+ iretq
-+paranoid_userspace\trace:
-+ GET_THREAD_INFO(%rcx)
-+ movl threadinfo_flags(%rcx),%ebx
-+ andl $_TIF_WORK_MASK,%ebx
-+ jz paranoid_swapgs\trace
-+ movq %rsp,%rdi /* &pt_regs */
-+ call sync_regs
-+ movq %rax,%rsp /* switch stack for scheduling */
-+ testl $_TIF_NEED_RESCHED,%ebx
-+ jnz paranoid_schedule\trace
-+ movl %ebx,%edx /* arg3: thread flags */
-+ .if \trace
-+ TRACE_IRQS_ON
-+ .endif
-+ sti
-+ xorl %esi,%esi /* arg2: oldset */
-+ movq %rsp,%rdi /* arg1: &pt_regs */
-+ call do_notify_resume
-+ cli
-+ .if \trace
-+ TRACE_IRQS_OFF
-+ .endif
-+ jmp paranoid_userspace\trace
-+paranoid_schedule\trace:
-+ .if \trace
-+ TRACE_IRQS_ON
-+ .endif
-+ sti
-+ call schedule
-+ cli
-+ .if \trace
-+ TRACE_IRQS_OFF
-+ .endif
-+ jmp paranoid_userspace\trace
-+ CFI_ENDPROC
-+ .endm
-+#endif
++#
++# Ethernet (10 or 100Mbit)
++#
++CONFIG_NET_ETHERNET=y
++CONFIG_MII=m
++CONFIG_HAPPYMEAL=m
++CONFIG_SUNGEM=m
++CONFIG_CASSINI=m
++CONFIG_NET_VENDOR_3COM=y
++CONFIG_VORTEX=m
++CONFIG_TYPHOON=m
++
++#
++# Tulip family network device support
++#
++CONFIG_NET_TULIP=y
++CONFIG_DE2104X=m
++CONFIG_TULIP=m
++# CONFIG_TULIP_MWI is not set
++# CONFIG_TULIP_MMIO is not set
++CONFIG_TULIP_NAPI=y
++CONFIG_TULIP_NAPI_HW_MITIGATION=y
++CONFIG_DE4X5=m
++CONFIG_WINBOND_840=m
++CONFIG_DM9102=m
++CONFIG_ULI526X=m
++CONFIG_PCMCIA_XIRCOM=m
++CONFIG_HP100=m
++CONFIG_NET_PCI=y
++CONFIG_PCNET32=m
++CONFIG_AMD8111_ETH=m
++# CONFIG_AMD8111E_NAPI is not set
++CONFIG_ADAPTEC_STARFIRE=m
++CONFIG_ADAPTEC_STARFIRE_NAPI=y
++CONFIG_B44=m
++CONFIG_FORCEDETH=m
++CONFIG_DGRS=m
++CONFIG_EEPRO100=m
++CONFIG_E100=m
++CONFIG_FEALNX=m
++CONFIG_NATSEMI=m
++CONFIG_NE2K_PCI=m
++CONFIG_8139CP=m
++CONFIG_8139TOO=m
++# CONFIG_8139TOO_PIO is not set
++# CONFIG_8139TOO_TUNE_TWISTER is not set
++CONFIG_8139TOO_8129=y
++# CONFIG_8139_OLD_RX_RESET is not set
++CONFIG_SIS900=m
++CONFIG_EPIC100=m
++CONFIG_SUNDANCE=m
++# CONFIG_SUNDANCE_MMIO is not set
++CONFIG_TLAN=m
++CONFIG_VIA_RHINE=m
++# CONFIG_VIA_RHINE_MMIO is not set
++# CONFIG_VIA_RHINE_NAPI is not set
++CONFIG_NET_POCKET=y
++CONFIG_ATP=m
++CONFIG_DE600=m
++CONFIG_DE620=m
++
++#
++# Ethernet (1000 Mbit)
++#
++CONFIG_ACENIC=m
++# CONFIG_ACENIC_OMIT_TIGON_I is not set
++CONFIG_DL2K=m
++CONFIG_E1000=m
++CONFIG_E1000_NAPI=y
++# CONFIG_E1000_DISABLE_PACKET_SPLIT is not set
++CONFIG_NS83820=m
++CONFIG_HAMACHI=m
++CONFIG_YELLOWFIN=m
++CONFIG_R8169=m
++# CONFIG_R8169_NAPI is not set
++CONFIG_R8169_VLAN=y
++CONFIG_SIS190=m
++CONFIG_SKGE=m
++CONFIG_SKY2=m
++CONFIG_SK98LIN=m
++CONFIG_VIA_VELOCITY=m
++CONFIG_TIGON3=m
++CONFIG_BNX2=m
++
++#
++# Ethernet (10000 Mbit)
++#
++CONFIG_CHELSIO_T1=m
++CONFIG_IXGB=m
++CONFIG_IXGB_NAPI=y
++CONFIG_S2IO=m
++CONFIG_S2IO_NAPI=y
++CONFIG_MYRI10GE=m
++# CONFIG_SFC is not set
++
++#
++# Token Ring devices
++#
++CONFIG_TR=y
++CONFIG_IBMOL=m
++CONFIG_IBMLS=m
++CONFIG_3C359=m
++CONFIG_TMS380TR=m
++CONFIG_TMSPCI=m
++CONFIG_ABYSS=m
++
++#
++# Wireless LAN (non-hamradio)
++#
++CONFIG_NET_RADIO=y
++CONFIG_NET_WIRELESS_RTNETLINK=y
+
-+/*
-+ * Exception entry point. This expects an error code/orig_rax on the stack
-+ * and the exception handler in %rax.
-+ */
-+ENTRY(error_entry)
-+ _frame RDI
-+ CFI_REL_OFFSET rax,0
-+ /* rdi slot contains rax, oldrax contains error code */
-+ cld
-+ subq $14*8,%rsp
-+ CFI_ADJUST_CFA_OFFSET (14*8)
-+ movq %rsi,13*8(%rsp)
-+ CFI_REL_OFFSET rsi,RSI
-+ movq 14*8(%rsp),%rsi /* load rax from rdi slot */
-+ CFI_REGISTER rax,rsi
-+ movq %rdx,12*8(%rsp)
-+ CFI_REL_OFFSET rdx,RDX
-+ movq %rcx,11*8(%rsp)
-+ CFI_REL_OFFSET rcx,RCX
-+ movq %rsi,10*8(%rsp) /* store rax */
-+ CFI_REL_OFFSET rax,RAX
-+ movq %r8, 9*8(%rsp)
-+ CFI_REL_OFFSET r8,R8
-+ movq %r9, 8*8(%rsp)
-+ CFI_REL_OFFSET r9,R9
-+ movq %r10,7*8(%rsp)
-+ CFI_REL_OFFSET r10,R10
-+ movq %r11,6*8(%rsp)
-+ CFI_REL_OFFSET r11,R11
-+ movq %rbx,5*8(%rsp)
-+ CFI_REL_OFFSET rbx,RBX
-+ movq %rbp,4*8(%rsp)
-+ CFI_REL_OFFSET rbp,RBP
-+ movq %r12,3*8(%rsp)
-+ CFI_REL_OFFSET r12,R12
-+ movq %r13,2*8(%rsp)
-+ CFI_REL_OFFSET r13,R13
-+ movq %r14,1*8(%rsp)
-+ CFI_REL_OFFSET r14,R14
-+ movq %r15,(%rsp)
-+ CFI_REL_OFFSET r15,R15
-+#if 0
-+ cmpl $__KERNEL_CS,CS(%rsp)
-+ CFI_REMEMBER_STATE
-+ je error_kernelspace
-+#endif
-+error_call_handler:
-+ movq %rdi, RDI(%rsp)
-+ CFI_REL_OFFSET rdi,RDI
-+ movq %rsp,%rdi
-+ movq ORIG_RAX(%rsp),%rsi # get error code
-+ movq $-1,ORIG_RAX(%rsp)
-+ call *%rax
-+error_exit:
-+ RESTORE_REST
-+/* cli */
-+ XEN_BLOCK_EVENTS(%rsi)
-+ TRACE_IRQS_OFF
-+ GET_THREAD_INFO(%rcx)
-+ testb $3,CS-ARGOFFSET(%rsp)
-+ jz retint_kernel
-+ movl threadinfo_flags(%rcx),%edx
-+ movl $_TIF_WORK_MASK,%edi
-+ andl %edi,%edx
-+ jnz retint_careful
-+ /*
-+ * The iret might restore flags:
-+ */
-+ TRACE_IRQS_IRETQ
-+ jmp retint_restore_args
++#
++# Obsolete Wireless cards support (pre-802.11)
++#
++CONFIG_STRIP=m
++CONFIG_PCMCIA_WAVELAN=m
++CONFIG_PCMCIA_NETWAVE=m
+
-+#if 0
-+ /*
-+ * We need to re-write the logic here because we don't do iretq to
-+ * to return to user mode. It's still possible that we get trap/fault
-+ * in the kernel (when accessing buffers pointed to by system calls,
-+ * for example).
-+ *
-+ */
-+ CFI_RESTORE_STATE
-+error_kernelspace:
-+ incl %ebx
-+ /* There are two places in the kernel that can potentially fault with
-+ usergs. Handle them here. The exception handlers after
-+ iret run with kernel gs again, so don't set the user space flag.
-+ B stepping K8s sometimes report an truncated RIP for IRET
-+ exceptions returning to compat mode. Check for these here too. */
-+ leaq iret_label(%rip),%rbp
-+ cmpq %rbp,RIP(%rsp)
-+ je error_swapgs
-+ movl %ebp,%ebp /* zero extend */
-+ cmpq %rbp,RIP(%rsp)
-+ je error_swapgs
-+ cmpq $gs_change,RIP(%rsp)
-+ je error_swapgs
-+ jmp error_sti
-+#endif
-+ CFI_ENDPROC
-+END(error_entry)
-+
-+ENTRY(hypervisor_callback)
-+ zeroentry do_hypervisor_callback
-+END(hypervisor_callback)
-+
-+/*
-+ * Copied from arch/xen/i386/kernel/entry.S
-+ */
-+# A note on the "critical region" in our callback handler.
-+# We want to avoid stacking callback handlers due to events occurring
-+# during handling of the last event. To do this, we keep events disabled
-+# until we've done all processing. HOWEVER, we must enable events before
-+# popping the stack frame (can't be done atomically) and so it would still
-+# be possible to get enough handler activations to overflow the stack.
-+# Although unlikely, bugs of that kind are hard to track down, so we'd
-+# like to avoid the possibility.
-+# So, on entry to the handler we detect whether we interrupted an
-+# existing activation in its critical region -- if so, we pop the current
-+# activation and restart the handler using the previous one.
-+ENTRY(do_hypervisor_callback) # do_hypervisor_callback(struct *pt_regs)
-+ CFI_STARTPROC
-+# Since we don't modify %rdi, evtchn_do_upall(struct *pt_regs) will
-+# see the correct pointer to the pt_regs
-+ movq %rdi, %rsp # we don't return, adjust the stack frame
-+ CFI_ENDPROC
-+ CFI_DEFAULT_STACK
-+11: incl %gs:pda_irqcount
-+ movq %rsp,%rbp
-+ CFI_DEF_CFA_REGISTER rbp
-+ cmovzq %gs:pda_irqstackptr,%rsp
-+ pushq %rbp # backlink for old unwinder
-+ call evtchn_do_upcall
-+ popq %rsp
-+ CFI_DEF_CFA_REGISTER rsp
-+ decl %gs:pda_irqcount
-+ jmp error_exit
-+ CFI_ENDPROC
-+END(do_hypervisor_callback)
++#
++# Wireless 802.11 Frequency Hopping cards support
++#
++CONFIG_PCMCIA_RAYCS=m
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+KPROBE_ENTRY(nmi)
-+ zeroentry do_nmi_callback
-+ENTRY(do_nmi_callback)
-+ CFI_STARTPROC
-+ addq $8, %rsp
-+ CFI_ENDPROC
-+ CFI_DEFAULT_STACK
-+ call do_nmi
-+ orl $NMI_MASK,EFLAGS(%rsp)
-+ RESTORE_REST
-+ XEN_BLOCK_EVENTS(%rsi)
-+ TRACE_IRQS_OFF
-+ GET_THREAD_INFO(%rcx)
-+ jmp retint_restore_args
-+ CFI_ENDPROC
-+ .previous .text
-+END(nmi)
-+#endif
++#
++# Wireless 802.11b ISA/PCI cards support
++#
++CONFIG_IPW2100=m
++CONFIG_IPW2100_MONITOR=y
++# CONFIG_IPW2100_DEBUG is not set
++CONFIG_IPW2200=m
++CONFIG_IPW2200_MONITOR=y
++CONFIG_IPW2200_RADIOTAP=y
++CONFIG_IPW2200_PROMISCUOUS=y
++CONFIG_IPW2200_QOS=y
++# CONFIG_IPW2200_DEBUG is not set
++CONFIG_AIRO=m
++CONFIG_HERMES=m
++CONFIG_PLX_HERMES=m
++CONFIG_TMD_HERMES=m
++CONFIG_NORTEL_HERMES=m
++CONFIG_PCI_HERMES=m
++CONFIG_ATMEL=m
++CONFIG_PCI_ATMEL=m
+
-+ ALIGN
-+restore_all_enable_events:
-+ CFI_DEFAULT_STACK adj=1
-+ TRACE_IRQS_ON
-+ XEN_UNBLOCK_EVENTS(%rsi) # %rsi is already set up...
++#
++# Wireless 802.11b Pcmcia/Cardbus cards support
++#
++CONFIG_PCMCIA_HERMES=m
++CONFIG_PCMCIA_SPECTRUM=m
++CONFIG_AIRO_CS=m
++CONFIG_PCMCIA_ATMEL=m
++CONFIG_PCMCIA_WL3501=m
+
-+scrit: /**** START OF CRITICAL REGION ****/
-+ XEN_TEST_PENDING(%rsi)
-+ CFI_REMEMBER_STATE
-+ jnz 14f # process more events if necessary...
-+ XEN_PUT_VCPU_INFO(%rsi)
-+ RESTORE_ARGS 0,8,0
-+ HYPERVISOR_IRET 0
-+
-+ CFI_RESTORE_STATE
-+14: XEN_LOCKED_BLOCK_EVENTS(%rsi)
-+ XEN_PUT_VCPU_INFO(%rsi)
-+ SAVE_REST
-+ movq %rsp,%rdi # set the argument again
-+ jmp 11b
-+ CFI_ENDPROC
-+ecrit: /**** END OF CRITICAL REGION ****/
-+# At this point, unlike on x86-32, we don't do the fixup to simplify the
-+# code and the stack frame is more complex on x86-64.
-+# When the kernel is interrupted in the critical section, the kernel
-+# will do IRET in that case, and everything will be restored at that point,
-+# i.e. it just resumes from the next instruction interrupted with the same context.
++#
++# Prism GT/Duette 802.11(a/b/g) PCI/Cardbus support
++#
++CONFIG_PRISM54=m
++CONFIG_USB_ZD1201=m
++CONFIG_HOSTAP=m
++CONFIG_HOSTAP_FIRMWARE=y
++CONFIG_HOSTAP_FIRMWARE_NVRAM=y
++CONFIG_HOSTAP_PLX=m
++CONFIG_HOSTAP_PCI=m
++CONFIG_HOSTAP_CS=m
++CONFIG_BCM43XX=m
++CONFIG_BCM43XX_DEBUG=y
++CONFIG_BCM43XX_DMA=y
++CONFIG_BCM43XX_PIO=y
++CONFIG_BCM43XX_DMA_AND_PIO_MODE=y
++# CONFIG_BCM43XX_DMA_MODE is not set
++# CONFIG_BCM43XX_PIO_MODE is not set
++CONFIG_ZD1211RW=m
++# CONFIG_ZD1211RW_DEBUG is not set
++CONFIG_NET_WIRELESS=y
+
-+# Hypervisor uses this for application faults while it executes.
-+# We get here for two reasons:
-+# 1. Fault while reloading DS, ES, FS or GS
-+# 2. Fault while executing IRET
-+# Category 1 we do not need to fix up as Xen has already reloaded all segment
-+# registers that could be reloaded and zeroed the others.
-+# Category 2 we fix up by killing the current process. We cannot use the
-+# normal Linux return path in this case because if we use the IRET hypercall
-+# to pop the stack frame we end up in an infinite loop of failsafe callbacks.
-+# We distinguish between categories by comparing each saved segment register
-+# with its current contents: any discrepancy means we in category 1.
-+ENTRY(failsafe_callback)
-+ _frame (RIP-0x30)
-+ CFI_REL_OFFSET rcx, 0
-+ CFI_REL_OFFSET r11, 8
-+ movw %ds,%cx
-+ cmpw %cx,0x10(%rsp)
-+ CFI_REMEMBER_STATE
-+ jne 1f
-+ movw %es,%cx
-+ cmpw %cx,0x18(%rsp)
-+ jne 1f
-+ movw %fs,%cx
-+ cmpw %cx,0x20(%rsp)
-+ jne 1f
-+ movw %gs,%cx
-+ cmpw %cx,0x28(%rsp)
-+ jne 1f
-+ /* All segments match their saved values => Category 2 (Bad IRET). */
-+ movq (%rsp),%rcx
-+ CFI_RESTORE rcx
-+ movq 8(%rsp),%r11
-+ CFI_RESTORE r11
-+ addq $0x30,%rsp
-+ CFI_ADJUST_CFA_OFFSET -0x30
-+ movq $11,%rdi /* SIGSEGV */
-+ jmp do_exit
-+ CFI_RESTORE_STATE
-+1: /* Segment mismatch => Category 1 (Bad segment). Retry the IRET. */
-+ movq (%rsp),%rcx
-+ CFI_RESTORE rcx
-+ movq 8(%rsp),%r11
-+ CFI_RESTORE r11
-+ addq $0x30,%rsp
-+ CFI_ADJUST_CFA_OFFSET -0x30
-+ pushq $0
-+ CFI_ADJUST_CFA_OFFSET 8
-+ SAVE_ALL
-+ jmp error_exit
-+ CFI_ENDPROC
-+#if 0
-+ .section __ex_table,"a"
-+ .align 8
-+ .quad gs_change,bad_gs
-+ .previous
-+ .section .fixup,"ax"
-+ /* running with kernelgs */
-+bad_gs:
-+/* swapgs */ /* switch back to user gs */
-+ xorl %eax,%eax
-+ movl %eax,%gs
-+ jmp 2b
-+ .previous
-+#endif
-+
-+/*
-+ * Create a kernel thread.
-+ *
-+ * C extern interface:
-+ * extern long kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
-+ *
-+ * asm input arguments:
-+ * rdi: fn, rsi: arg, rdx: flags
-+ */
-+ENTRY(kernel_thread)
-+ CFI_STARTPROC
-+ FAKE_STACK_FRAME $child_rip
-+ SAVE_ALL
++#
++# PCMCIA network device support
++#
++CONFIG_NET_PCMCIA=y
++CONFIG_PCMCIA_3C589=m
++CONFIG_PCMCIA_3C574=m
++CONFIG_PCMCIA_FMVJ18X=m
++CONFIG_PCMCIA_PCNET=m
++CONFIG_PCMCIA_NMCLAN=m
++CONFIG_PCMCIA_SMC91C92=m
++CONFIG_PCMCIA_XIRC2PS=m
++CONFIG_PCMCIA_AXNET=m
++CONFIG_PCMCIA_IBMTR=m
+
-+ # rdi: flags, rsi: usp, rdx: will be &pt_regs
-+ movq %rdx,%rdi
-+ orq kernel_thread_flags(%rip),%rdi
-+ movq $-1, %rsi
-+ movq %rsp, %rdx
++#
++# Wan interfaces
++#
++CONFIG_WAN=y
++CONFIG_DSCC4=m
++CONFIG_DSCC4_PCISYNC=y
++CONFIG_DSCC4_PCI_RST=y
++CONFIG_LANMEDIA=m
++CONFIG_HDLC=m
++CONFIG_HDLC_RAW=y
++CONFIG_HDLC_RAW_ETH=y
++CONFIG_HDLC_CISCO=y
++CONFIG_HDLC_FR=y
++CONFIG_HDLC_PPP=y
++CONFIG_HDLC_X25=y
++CONFIG_PCI200SYN=m
++CONFIG_WANXL=m
++CONFIG_PC300=m
++CONFIG_PC300_MLPPP=y
++CONFIG_FARSYNC=m
++CONFIG_DLCI=m
++CONFIG_DLCI_COUNT=24
++CONFIG_DLCI_MAX=8
++CONFIG_WAN_ROUTER_DRIVERS=y
++CONFIG_CYCLADES_SYNC=m
++CONFIG_CYCLOMX_X25=y
++CONFIG_LAPBETHER=m
++CONFIG_X25_ASY=m
++CONFIG_SBNI=m
++# CONFIG_SBNI_MULTILINE is not set
+
-+ xorl %r8d,%r8d
-+ xorl %r9d,%r9d
-+
-+ # clone now
-+ call do_fork
-+ movq %rax,RAX(%rsp)
-+ xorl %edi,%edi
++#
++# ATM drivers
++#
++CONFIG_ATM_DUMMY=m
++CONFIG_ATM_TCP=m
++CONFIG_ATM_LANAI=m
++CONFIG_ATM_ENI=m
++# CONFIG_ATM_ENI_DEBUG is not set
++# CONFIG_ATM_ENI_TUNE_BURST is not set
++CONFIG_ATM_FIRESTREAM=m
++CONFIG_ATM_ZATM=m
++# CONFIG_ATM_ZATM_DEBUG is not set
++CONFIG_ATM_NICSTAR=m
++CONFIG_ATM_NICSTAR_USE_SUNI=y
++CONFIG_ATM_NICSTAR_USE_IDT77105=y
++CONFIG_ATM_IDT77252=m
++# CONFIG_ATM_IDT77252_DEBUG is not set
++CONFIG_ATM_IDT77252_RCV_ALL=y
++CONFIG_ATM_IDT77252_USE_SUNI=y
++CONFIG_ATM_AMBASSADOR=m
++# CONFIG_ATM_AMBASSADOR_DEBUG is not set
++CONFIG_ATM_HORIZON=m
++# CONFIG_ATM_HORIZON_DEBUG is not set
++CONFIG_ATM_IA=m
++# CONFIG_ATM_IA_DEBUG is not set
++CONFIG_ATM_FORE200E_MAYBE=m
++CONFIG_ATM_FORE200E_PCA=y
++CONFIG_ATM_FORE200E_PCA_DEFAULT_FW=y
++CONFIG_ATM_FORE200E_USE_TASKLET=y
++CONFIG_ATM_FORE200E_TX_RETRY=16
++CONFIG_ATM_FORE200E_DEBUG=0
++CONFIG_ATM_FORE200E=m
++CONFIG_ATM_HE=m
++CONFIG_ATM_HE_USE_SUNI=y
++CONFIG_FDDI=y
++# CONFIG_DEFXX is not set
++CONFIG_SKFP=m
++CONFIG_HIPPI=y
++CONFIG_ROADRUNNER=m
++CONFIG_ROADRUNNER_LARGE_RINGS=y
++CONFIG_PLIP=m
++CONFIG_PPP=m
++CONFIG_PPP_MULTILINK=y
++CONFIG_PPP_FILTER=y
++CONFIG_PPP_ASYNC=m
++CONFIG_PPP_SYNC_TTY=m
++CONFIG_PPP_DEFLATE=m
++CONFIG_PPP_BSDCOMP=m
++CONFIG_PPP_MPPE=m
++CONFIG_PPPOE=m
++CONFIG_PPPOATM=m
++CONFIG_SLIP=m
++CONFIG_SLIP_COMPRESSED=y
++CONFIG_SLIP_SMART=y
++CONFIG_SLIP_MODE_SLIP6=y
++CONFIG_NET_FC=y
++CONFIG_SHAPER=m
++CONFIG_NETCONSOLE=m
++CONFIG_NETPOLL=y
++CONFIG_NETPOLL_RX=y
++CONFIG_NETPOLL_TRAP=y
++CONFIG_NET_POLL_CONTROLLER=y
+
-+ /*
-+ * It isn't worth to check for reschedule here,
-+ * so internally to the x86_64 port you can rely on kernel_thread()
-+ * not to reschedule the child before returning, this avoids the need
-+ * of hacks for example to fork off the per-CPU idle tasks.
-+ * [Hopefully no generic code relies on the reschedule -AK]
-+ */
-+ RESTORE_ALL
-+ UNFAKE_STACK_FRAME
-+ ret
-+ CFI_ENDPROC
-+ENDPROC(kernel_thread)
-+
-+child_rip:
-+ pushq $0 # fake return address
-+ CFI_STARTPROC
-+ /*
-+ * Here we are in the child and the registers are set as they were
-+ * at kernel_thread() invocation in the parent.
-+ */
-+ movq %rdi, %rax
-+ movq %rsi, %rdi
-+ call *%rax
-+ # exit
-+ xorl %edi, %edi
-+ call do_exit
-+ CFI_ENDPROC
-+ENDPROC(child_rip)
++#
++# ISDN subsystem
++#
++CONFIG_ISDN=m
+
-+/*
-+ * execve(). This function needs to use IRET, not SYSRET, to set up all state properly.
-+ *
-+ * C extern interface:
-+ * extern long execve(char *name, char **argv, char **envp)
-+ *
-+ * asm input arguments:
-+ * rdi: name, rsi: argv, rdx: envp
-+ *
-+ * We want to fallback into:
-+ * extern long sys_execve(char *name, char **argv,char **envp, struct pt_regs regs)
-+ *
-+ * do_sys_execve asm fallback arguments:
-+ * rdi: name, rsi: argv, rdx: envp, fake frame on the stack
-+ */
-+ENTRY(execve)
-+ CFI_STARTPROC
-+ FAKE_STACK_FRAME $0
-+ SAVE_ALL
-+ call sys_execve
-+ movq %rax, RAX(%rsp)
-+ RESTORE_REST
-+ testq %rax,%rax
-+ jne 1f
-+ jmp int_ret_from_sys_call
-+1: RESTORE_ARGS
-+ UNFAKE_STACK_FRAME
-+ ret
-+ CFI_ENDPROC
-+ENDPROC(execve)
++#
++# Old ISDN4Linux
++#
++CONFIG_ISDN_I4L=m
++CONFIG_ISDN_PPP=y
++CONFIG_ISDN_PPP_VJ=y
++CONFIG_ISDN_MPP=y
++CONFIG_IPPP_FILTER=y
++CONFIG_ISDN_PPP_BSDCOMP=m
++CONFIG_ISDN_AUDIO=y
++CONFIG_ISDN_TTY_FAX=y
++CONFIG_ISDN_X25=y
+
-+KPROBE_ENTRY(page_fault)
-+ errorentry do_page_fault
-+END(page_fault)
-+ .previous .text
++#
++# ISDN feature submodules
++#
++CONFIG_ISDN_DIVERSION=m
+
-+ENTRY(coprocessor_error)
-+ zeroentry do_coprocessor_error
-+END(coprocessor_error)
++#
++# ISDN4Linux hardware drivers
++#
+
-+ENTRY(simd_coprocessor_error)
-+ zeroentry do_simd_coprocessor_error
-+END(simd_coprocessor_error)
++#
++# Passive cards
++#
++CONFIG_ISDN_DRV_HISAX=m
+
-+ENTRY(device_not_available)
-+ zeroentry math_state_restore
-+END(device_not_available)
++#
++# D-channel protocol features
++#
++CONFIG_HISAX_EURO=y
++CONFIG_DE_AOC=y
++# CONFIG_HISAX_NO_SENDCOMPLETE is not set
++# CONFIG_HISAX_NO_LLC is not set
++# CONFIG_HISAX_NO_KEYPAD is not set
++CONFIG_HISAX_1TR6=y
++CONFIG_HISAX_NI1=y
++CONFIG_HISAX_MAX_CARDS=8
+
-+ /* runs on exception stack */
-+KPROBE_ENTRY(debug)
-+/* INTR_FRAME
-+ pushq $0
-+ CFI_ADJUST_CFA_OFFSET 8 */
-+ zeroentry do_debug
-+/* paranoidexit
-+ CFI_ENDPROC */
-+END(debug)
-+ .previous .text
++#
++# HiSax supported cards
++#
++CONFIG_HISAX_16_3=y
++CONFIG_HISAX_TELESPCI=y
++CONFIG_HISAX_S0BOX=y
++CONFIG_HISAX_FRITZPCI=y
++CONFIG_HISAX_AVM_A1_PCMCIA=y
++CONFIG_HISAX_ELSA=y
++CONFIG_HISAX_DIEHLDIVA=y
++CONFIG_HISAX_SEDLBAUER=y
++CONFIG_HISAX_NETJET=y
++CONFIG_HISAX_NETJET_U=y
++CONFIG_HISAX_NICCY=y
++CONFIG_HISAX_BKM_A4T=y
++CONFIG_HISAX_SCT_QUADRO=y
++CONFIG_HISAX_GAZEL=y
++CONFIG_HISAX_HFC_PCI=y
++CONFIG_HISAX_W6692=y
++CONFIG_HISAX_HFC_SX=y
++CONFIG_HISAX_ENTERNOW_PCI=y
++# CONFIG_HISAX_DEBUG is not set
+
-+#if 0
-+ /* runs on exception stack */
-+KPROBE_ENTRY(nmi)
-+ INTR_FRAME
-+ pushq $-1
-+ CFI_ADJUST_CFA_OFFSET 8
-+ paranoidentry do_nmi, 0, 0
-+#ifdef CONFIG_TRACE_IRQFLAGS
-+ paranoidexit 0
-+#else
-+ jmp paranoid_exit1
-+ CFI_ENDPROC
-+#endif
-+END(nmi)
-+ .previous .text
-+#endif
++#
++# HiSax PCMCIA card service modules
++#
++CONFIG_HISAX_SEDLBAUER_CS=m
++CONFIG_HISAX_ELSA_CS=m
++CONFIG_HISAX_AVM_A1_CS=m
++CONFIG_HISAX_TELES_CS=m
+
-+KPROBE_ENTRY(int3)
-+/* INTR_FRAME
-+ pushq $0
-+ CFI_ADJUST_CFA_OFFSET 8 */
-+ zeroentry do_int3
-+/* jmp paranoid_exit1
-+ CFI_ENDPROC */
-+END(int3)
-+ .previous .text
++#
++# HiSax sub driver modules
++#
++CONFIG_HISAX_ST5481=m
++CONFIG_HISAX_HFCUSB=m
++CONFIG_HISAX_HFC4S8S=m
++CONFIG_HISAX_FRITZ_PCIPNP=m
++CONFIG_HISAX_HDLC=y
+
-+ENTRY(overflow)
-+ zeroentry do_overflow
-+END(overflow)
++#
++# Active cards
++#
+
-+ENTRY(bounds)
-+ zeroentry do_bounds
-+END(bounds)
++#
++# Siemens Gigaset
++#
++CONFIG_ISDN_DRV_GIGASET=m
++CONFIG_GIGASET_BASE=m
++CONFIG_GIGASET_M105=m
++# CONFIG_GIGASET_DEBUG is not set
++# CONFIG_GIGASET_UNDOCREQ is not set
+
-+ENTRY(invalid_op)
-+ zeroentry do_invalid_op
-+END(invalid_op)
++#
++# CAPI subsystem
++#
++CONFIG_ISDN_CAPI=m
++CONFIG_ISDN_DRV_AVMB1_VERBOSE_REASON=y
++CONFIG_ISDN_CAPI_MIDDLEWARE=y
++CONFIG_ISDN_CAPI_CAPI20=m
++CONFIG_ISDN_CAPI_CAPIFS_BOOL=y
++CONFIG_ISDN_CAPI_CAPIFS=m
++CONFIG_ISDN_CAPI_CAPIDRV=m
+
-+ENTRY(coprocessor_segment_overrun)
-+ zeroentry do_coprocessor_segment_overrun
-+END(coprocessor_segment_overrun)
++#
++# CAPI hardware drivers
++#
+
-+ENTRY(reserved)
-+ zeroentry do_reserved
-+END(reserved)
++#
++# Active AVM cards
++#
++CONFIG_CAPI_AVM=y
++CONFIG_ISDN_DRV_AVMB1_B1PCI=m
++CONFIG_ISDN_DRV_AVMB1_B1PCIV4=y
++CONFIG_ISDN_DRV_AVMB1_B1PCMCIA=m
++CONFIG_ISDN_DRV_AVMB1_AVM_CS=m
++CONFIG_ISDN_DRV_AVMB1_T1PCI=m
++CONFIG_ISDN_DRV_AVMB1_C4=m
+
-+#if 0
-+ /* runs on exception stack */
-+ENTRY(double_fault)
-+ XCPT_FRAME
-+ paranoidentry do_double_fault
-+ jmp paranoid_exit1
-+ CFI_ENDPROC
-+END(double_fault)
-+#endif
++#
++# Active Eicon DIVA Server cards
++#
++CONFIG_CAPI_EICON=y
++CONFIG_ISDN_DIVAS=m
++CONFIG_ISDN_DIVAS_BRIPCI=y
++CONFIG_ISDN_DIVAS_PRIPCI=y
++CONFIG_ISDN_DIVAS_DIVACAPI=m
++CONFIG_ISDN_DIVAS_USERIDI=m
++CONFIG_ISDN_DIVAS_MAINT=m
+
-+ENTRY(invalid_TSS)
-+ errorentry do_invalid_TSS
-+END(invalid_TSS)
++#
++# Telephony Support
++#
++CONFIG_PHONE=m
++CONFIG_PHONE_IXJ=m
++CONFIG_PHONE_IXJ_PCMCIA=m
+
-+ENTRY(segment_not_present)
-+ errorentry do_segment_not_present
-+END(segment_not_present)
++#
++# Input device support
++#
++CONFIG_INPUT=y
+
-+ /* runs on exception stack */
-+ENTRY(stack_segment)
-+/* XCPT_FRAME
-+ paranoidentry do_stack_segment */
-+ errorentry do_stack_segment
-+/* jmp paranoid_exit1
-+ CFI_ENDPROC */
-+END(stack_segment)
++#
++# Userland interfaces
++#
++CONFIG_INPUT_MOUSEDEV=y
++CONFIG_INPUT_MOUSEDEV_PSAUX=y
++CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024
++CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768
++CONFIG_INPUT_JOYDEV=m
++CONFIG_INPUT_TSDEV=m
++CONFIG_INPUT_TSDEV_SCREEN_X=240
++CONFIG_INPUT_TSDEV_SCREEN_Y=320
++CONFIG_INPUT_EVDEV=m
++# CONFIG_INPUT_EVBUG is not set
++
++#
++# Input Device Drivers
++#
++CONFIG_INPUT_KEYBOARD=y
++CONFIG_KEYBOARD_ATKBD=y
++CONFIG_KEYBOARD_SUNKBD=m
++CONFIG_KEYBOARD_LKKBD=m
++CONFIG_KEYBOARD_XTKBD=m
++CONFIG_KEYBOARD_NEWTON=m
++CONFIG_INPUT_MOUSE=y
++CONFIG_MOUSE_PS2=y
++CONFIG_MOUSE_SERIAL=m
++CONFIG_MOUSE_VSXXXAA=m
++CONFIG_INPUT_JOYSTICK=y
++CONFIG_JOYSTICK_ANALOG=m
++CONFIG_JOYSTICK_A3D=m
++CONFIG_JOYSTICK_ADI=m
++CONFIG_JOYSTICK_COBRA=m
++CONFIG_JOYSTICK_GF2K=m
++CONFIG_JOYSTICK_GRIP=m
++CONFIG_JOYSTICK_GRIP_MP=m
++CONFIG_JOYSTICK_GUILLEMOT=m
++CONFIG_JOYSTICK_INTERACT=m
++CONFIG_JOYSTICK_SIDEWINDER=m
++CONFIG_JOYSTICK_TMDC=m
++CONFIG_JOYSTICK_IFORCE=m
++CONFIG_JOYSTICK_IFORCE_USB=y
++CONFIG_JOYSTICK_IFORCE_232=y
++CONFIG_JOYSTICK_WARRIOR=m
++CONFIG_JOYSTICK_MAGELLAN=m
++CONFIG_JOYSTICK_SPACEORB=m
++CONFIG_JOYSTICK_SPACEBALL=m
++CONFIG_JOYSTICK_STINGER=m
++CONFIG_JOYSTICK_TWIDJOY=m
++CONFIG_JOYSTICK_DB9=m
++CONFIG_JOYSTICK_GAMECON=m
++CONFIG_JOYSTICK_TURBOGRAFX=m
++CONFIG_JOYSTICK_JOYDUMP=m
++CONFIG_INPUT_TOUCHSCREEN=y
++CONFIG_TOUCHSCREEN_ADS7846=m
++CONFIG_TOUCHSCREEN_GUNZE=m
++CONFIG_TOUCHSCREEN_ELO=m
++CONFIG_TOUCHSCREEN_MTOUCH=m
++CONFIG_TOUCHSCREEN_MK712=m
++CONFIG_INPUT_MISC=y
++CONFIG_INPUT_PCSPKR=m
++CONFIG_INPUT_WISTRON_BTNS=m
++CONFIG_INPUT_UINPUT=m
++
++#
++# Hardware I/O ports
++#
++CONFIG_SERIO=y
++CONFIG_SERIO_I8042=y
++CONFIG_SERIO_SERPORT=m
++CONFIG_SERIO_CT82C710=m
++CONFIG_SERIO_PARKBD=m
++CONFIG_SERIO_PCIPS2=m
++CONFIG_SERIO_LIBPS2=y
++CONFIG_SERIO_RAW=m
++CONFIG_GAMEPORT=m
++CONFIG_GAMEPORT_NS558=m
++CONFIG_GAMEPORT_L4=m
++CONFIG_GAMEPORT_EMU10K1=m
++CONFIG_GAMEPORT_FM801=m
++
++#
++# Character devices
++#
++CONFIG_VT=y
++CONFIG_VT_CONSOLE=y
++CONFIG_HW_CONSOLE=y
++CONFIG_VT_HW_CONSOLE_BINDING=y
++# CONFIG_SERIAL_NONSTANDARD is not set
++
++#
++# Serial drivers
++#
++CONFIG_SERIAL_8250=m
++CONFIG_SERIAL_8250_PCI=m
++CONFIG_SERIAL_8250_PNP=m
++# CONFIG_SERIAL_8250_CS is not set
++CONFIG_SERIAL_8250_NR_UARTS=4
++CONFIG_SERIAL_8250_RUNTIME_UARTS=4
++# CONFIG_SERIAL_8250_EXTENDED is not set
++
++#
++# Non-8250 serial port support
++#
++CONFIG_SERIAL_CORE=m
++CONFIG_SERIAL_JSM=m
++CONFIG_UNIX98_PTYS=y
++CONFIG_LEGACY_PTYS=y
++CONFIG_LEGACY_PTY_COUNT=256
++CONFIG_PRINTER=m
++# CONFIG_LP_CONSOLE is not set
++CONFIG_PPDEV=m
++CONFIG_TIPAR=m
++
++#
++# IPMI
++#
++CONFIG_IPMI_HANDLER=m
++# CONFIG_IPMI_PANIC_EVENT is not set
++CONFIG_IPMI_DEVICE_INTERFACE=m
++CONFIG_IPMI_SI=m
++CONFIG_IPMI_WATCHDOG=m
++CONFIG_IPMI_POWEROFF=m
++
++#
++# Watchdog Cards
++#
++CONFIG_WATCHDOG=y
++# CONFIG_WATCHDOG_NOWAYOUT is not set
++
++#
++# Watchdog Device Drivers
++#
++CONFIG_SOFT_WATCHDOG=m
++CONFIG_ACQUIRE_WDT=m
++CONFIG_ADVANTECH_WDT=m
++CONFIG_ALIM1535_WDT=m
++CONFIG_ALIM7101_WDT=m
++CONFIG_SC520_WDT=m
++CONFIG_EUROTECH_WDT=m
++CONFIG_IB700_WDT=m
++CONFIG_IBMASR=m
++CONFIG_WAFER_WDT=m
++CONFIG_I6300ESB_WDT=m
++CONFIG_I8XX_TCO=m
++CONFIG_SC1200_WDT=m
++CONFIG_SCx200_WDT=m
++CONFIG_60XX_WDT=m
++CONFIG_SBC8360_WDT=m
++CONFIG_CPU5_WDT=m
++CONFIG_W83627HF_WDT=m
++CONFIG_W83877F_WDT=m
++CONFIG_W83977F_WDT=m
++CONFIG_MACHZ_WDT=m
++CONFIG_SBC_EPX_C3_WATCHDOG=m
++
++#
++# PCI-based Watchdog Cards
++#
++CONFIG_PCIPCWATCHDOG=m
++CONFIG_WDTPCI=m
++CONFIG_WDT_501_PCI=y
++
++#
++# USB-based Watchdog Cards
++#
++CONFIG_USBPCWATCHDOG=m
++CONFIG_HW_RANDOM=y
++CONFIG_HW_RANDOM_INTEL=m
++CONFIG_HW_RANDOM_AMD=m
++CONFIG_HW_RANDOM_GEODE=m
++CONFIG_HW_RANDOM_VIA=m
++CONFIG_NVRAM=m
++CONFIG_RTC=m
++CONFIG_GEN_RTC=m
++CONFIG_GEN_RTC_X=y
++CONFIG_DTLK=m
++CONFIG_R3964=m
++CONFIG_APPLICOM=m
++CONFIG_SONYPI=m
++
++#
++# Ftape, the floppy tape device driver
++#
++CONFIG_AGP=m
++CONFIG_AGP_ALI=m
++CONFIG_AGP_ATI=m
++CONFIG_AGP_AMD=m
++CONFIG_AGP_AMD64=m
++CONFIG_AGP_INTEL=m
++CONFIG_AGP_NVIDIA=m
++CONFIG_AGP_SIS=m
++CONFIG_AGP_SWORKS=m
++CONFIG_AGP_VIA=m
++CONFIG_AGP_EFFICEON=m
++CONFIG_DRM=m
++CONFIG_DRM_TDFX=m
++CONFIG_DRM_R128=m
++CONFIG_DRM_RADEON=m
++CONFIG_DRM_I810=m
++CONFIG_DRM_I830=m
++CONFIG_DRM_I915=m
++CONFIG_DRM_MGA=m
++CONFIG_DRM_SIS=m
++CONFIG_DRM_VIA=m
++CONFIG_DRM_SAVAGE=m
++
++#
++# PCMCIA character devices
++#
++CONFIG_SYNCLINK_CS=m
++CONFIG_CARDMAN_4000=m
++CONFIG_CARDMAN_4040=m
++CONFIG_MWAVE=m
++CONFIG_SCx200_GPIO=m
++CONFIG_PC8736x_GPIO=m
++CONFIG_NSC_GPIO=m
++CONFIG_CS5535_GPIO=m
++CONFIG_RAW_DRIVER=m
++CONFIG_MAX_RAW_DEVS=256
++# CONFIG_HPET is not set
++CONFIG_HANGCHECK_TIMER=m
++
++#
++# TPM devices
++#
++CONFIG_TCG_TPM=m
++CONFIG_TCG_TIS=m
++CONFIG_TCG_NSC=m
++CONFIG_TCG_ATMEL=m
++CONFIG_TCG_INFINEON=m
++CONFIG_TCG_XEN=m
++CONFIG_TELCLOCK=m
++
++#
++# I2C support
++#
++CONFIG_I2C=m
++CONFIG_I2C_CHARDEV=m
++
++#
++# I2C Algorithms
++#
++CONFIG_I2C_ALGOBIT=m
++CONFIG_I2C_ALGOPCF=m
++CONFIG_I2C_ALGOPCA=m
+
-+KPROBE_ENTRY(general_protection)
-+ errorentry do_general_protection
-+END(general_protection)
-+ .previous .text
++#
++# I2C Hardware Bus support
++#
++CONFIG_I2C_ALI1535=m
++CONFIG_I2C_ALI1563=m
++CONFIG_I2C_ALI15X3=m
++CONFIG_I2C_AMD756=m
++CONFIG_I2C_AMD756_S4882=m
++CONFIG_I2C_AMD8111=m
++CONFIG_I2C_I801=m
++CONFIG_I2C_I810=m
++CONFIG_I2C_PIIX4=m
++CONFIG_I2C_ISA=m
++CONFIG_I2C_NFORCE2=m
++CONFIG_I2C_OCORES=m
++CONFIG_I2C_PARPORT=m
++CONFIG_I2C_PARPORT_LIGHT=m
++CONFIG_I2C_PROSAVAGE=m
++CONFIG_I2C_SAVAGE4=m
++CONFIG_SCx200_I2C=m
++CONFIG_SCx200_I2C_SCL=12
++CONFIG_SCx200_I2C_SDA=13
++CONFIG_SCx200_ACB=m
++CONFIG_I2C_SIS5595=m
++CONFIG_I2C_SIS630=m
++CONFIG_I2C_SIS96X=m
++CONFIG_I2C_STUB=m
++CONFIG_I2C_VIA=m
++CONFIG_I2C_VIAPRO=m
++CONFIG_I2C_VOODOO3=m
++CONFIG_I2C_PCA_ISA=m
+
-+ENTRY(alignment_check)
-+ errorentry do_alignment_check
-+END(alignment_check)
++#
++# Miscellaneous I2C Chip support
++#
++CONFIG_SENSORS_DS1337=m
++CONFIG_SENSORS_DS1374=m
++CONFIG_SENSORS_EEPROM=m
++CONFIG_SENSORS_PCF8574=m
++CONFIG_SENSORS_PCA9539=m
++CONFIG_SENSORS_PCF8591=m
++CONFIG_SENSORS_MAX6875=m
++# CONFIG_I2C_DEBUG_CORE is not set
++# CONFIG_I2C_DEBUG_ALGO is not set
++# CONFIG_I2C_DEBUG_BUS is not set
++# CONFIG_I2C_DEBUG_CHIP is not set
+
-+ENTRY(divide_error)
-+ zeroentry do_divide_error
-+END(divide_error)
++#
++# SPI support
++#
++CONFIG_SPI=y
++# CONFIG_SPI_DEBUG is not set
++CONFIG_SPI_MASTER=y
+
-+ENTRY(spurious_interrupt_bug)
-+ zeroentry do_spurious_interrupt_bug
-+END(spurious_interrupt_bug)
++#
++# SPI Master Controller Drivers
++#
++CONFIG_SPI_BITBANG=m
++CONFIG_SPI_BUTTERFLY=m
+
-+#ifdef CONFIG_X86_MCE
-+ /* runs on exception stack */
-+ENTRY(machine_check)
-+ INTR_FRAME
-+ pushq $0
-+ CFI_ADJUST_CFA_OFFSET 8
-+ paranoidentry do_machine_check
-+ jmp paranoid_exit1
-+ CFI_ENDPROC
-+END(machine_check)
-+#endif
++#
++# SPI Protocol Masters
++#
+
-+/* Call softirq on interrupt stack. Interrupts are off. */
-+ENTRY(call_softirq)
-+ CFI_STARTPROC
-+ push %rbp
-+ CFI_ADJUST_CFA_OFFSET 8
-+ CFI_REL_OFFSET rbp,0
-+ mov %rsp,%rbp
-+ CFI_DEF_CFA_REGISTER rbp
-+ incl %gs:pda_irqcount
-+ cmove %gs:pda_irqstackptr,%rsp
-+ push %rbp # backlink for old unwinder
-+ call __do_softirq
-+ leaveq
-+ CFI_DEF_CFA_REGISTER rsp
-+ CFI_ADJUST_CFA_OFFSET -8
-+ decl %gs:pda_irqcount
-+ ret
-+ CFI_ENDPROC
-+ENDPROC(call_softirq)
++#
++# Dallas's 1-wire bus
++#
++CONFIG_W1=m
++CONFIG_W1_CON=y
+
-+#ifdef CONFIG_STACK_UNWIND
-+ENTRY(arch_unwind_init_running)
-+ CFI_STARTPROC
-+ movq %r15, R15(%rdi)
-+ movq %r14, R14(%rdi)
-+ xchgq %rsi, %rdx
-+ movq %r13, R13(%rdi)
-+ movq %r12, R12(%rdi)
-+ xorl %eax, %eax
-+ movq %rbp, RBP(%rdi)
-+ movq %rbx, RBX(%rdi)
-+ movq (%rsp), %rcx
-+ movq %rax, R11(%rdi)
-+ movq %rax, R10(%rdi)
-+ movq %rax, R9(%rdi)
-+ movq %rax, R8(%rdi)
-+ movq %rax, RAX(%rdi)
-+ movq %rax, RCX(%rdi)
-+ movq %rax, RDX(%rdi)
-+ movq %rax, RSI(%rdi)
-+ movq %rax, RDI(%rdi)
-+ movq %rax, ORIG_RAX(%rdi)
-+ movq %rcx, RIP(%rdi)
-+ leaq 8(%rsp), %rcx
-+ movq $__KERNEL_CS, CS(%rdi)
-+ movq %rax, EFLAGS(%rdi)
-+ movq %rcx, RSP(%rdi)
-+ movq $__KERNEL_DS, SS(%rdi)
-+ jmpq *%rdx
-+ CFI_ENDPROC
-+ENDPROC(arch_unwind_init_running)
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/genapic_xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/genapic_xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/genapic_xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/genapic_xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,161 @@
-+/*
-+ * Copyright 2004 James Cleverdon, IBM.
-+ * Subject to the GNU Public License, v.2
-+ *
-+ * Xen APIC subarch code. Maximum 8 CPUs, logical delivery.
-+ *
-+ * Hacked for x86-64 by James Cleverdon from i386 architecture code by
-+ * Martin Bligh, Andi Kleen, James Bottomley, John Stultz, and
-+ * James Cleverdon.
-+ *
-+ * Hacked to pieces for Xen by Chris Wright.
-+ */
-+#include <linux/threads.h>
-+#include <linux/cpumask.h>
-+#include <linux/string.h>
-+#include <linux/kernel.h>
-+#include <linux/ctype.h>
-+#include <linux/init.h>
-+#ifdef CONFIG_XEN_PRIVILEGED_GUEST
-+#include <asm/smp.h>
-+#include <asm/ipi.h>
-+#else
-+#include <asm/apic.h>
-+#include <asm/apicdef.h>
-+#include <asm/genapic.h>
-+#endif
-+#include <xen/evtchn.h>
++#
++# 1-wire Bus Masters
++#
++CONFIG_W1_MASTER_MATROX=m
++CONFIG_W1_MASTER_DS2490=m
++CONFIG_W1_MASTER_DS2482=m
+
-+DECLARE_PER_CPU(int, ipi_to_irq[NR_IPIS]);
++#
++# 1-wire Slaves
++#
++CONFIG_W1_SLAVE_THERM=m
++CONFIG_W1_SLAVE_SMEM=m
++CONFIG_W1_SLAVE_DS2433=m
++# CONFIG_W1_SLAVE_DS2433_CRC is not set
+
-+static inline void __send_IPI_one(unsigned int cpu, int vector)
-+{
-+ int irq = per_cpu(ipi_to_irq, cpu)[vector];
-+ BUG_ON(irq < 0);
-+ notify_remote_via_irq(irq);
-+}
++#
++# Hardware Monitoring support
++#
++CONFIG_HWMON=m
++CONFIG_HWMON_VID=m
++CONFIG_SENSORS_ABITUGURU=m
++CONFIG_SENSORS_ADM1021=m
++CONFIG_SENSORS_ADM1025=m
++CONFIG_SENSORS_ADM1026=m
++CONFIG_SENSORS_ADM1031=m
++CONFIG_SENSORS_ADM9240=m
++CONFIG_SENSORS_ASB100=m
++CONFIG_SENSORS_ATXP1=m
++CONFIG_SENSORS_DS1621=m
++CONFIG_SENSORS_F71805F=m
++CONFIG_SENSORS_FSCHER=m
++CONFIG_SENSORS_FSCPOS=m
++CONFIG_SENSORS_GL518SM=m
++CONFIG_SENSORS_GL520SM=m
++CONFIG_SENSORS_IT87=m
++CONFIG_SENSORS_LM63=m
++CONFIG_SENSORS_LM70=m
++CONFIG_SENSORS_LM75=m
++CONFIG_SENSORS_LM77=m
++CONFIG_SENSORS_LM78=m
++CONFIG_SENSORS_LM80=m
++CONFIG_SENSORS_LM83=m
++CONFIG_SENSORS_LM85=m
++CONFIG_SENSORS_LM87=m
++CONFIG_SENSORS_LM90=m
++CONFIG_SENSORS_LM92=m
++CONFIG_SENSORS_MAX1619=m
++CONFIG_SENSORS_PC87360=m
++CONFIG_SENSORS_SIS5595=m
++CONFIG_SENSORS_SMSC47M1=m
++CONFIG_SENSORS_SMSC47M192=m
++CONFIG_SENSORS_SMSC47B397=m
++CONFIG_SENSORS_VIA686A=m
++CONFIG_SENSORS_VT8231=m
++CONFIG_SENSORS_W83781D=m
++CONFIG_SENSORS_W83791D=m
++CONFIG_SENSORS_W83792D=m
++CONFIG_SENSORS_W83L785TS=m
++CONFIG_SENSORS_W83627HF=m
++CONFIG_SENSORS_W83627EHF=m
++CONFIG_SENSORS_HDAPS=m
++# CONFIG_HWMON_DEBUG_CHIP is not set
+
-+void xen_send_IPI_shortcut(unsigned int shortcut, int vector, unsigned int dest)
-+{
-+ int cpu;
++#
++# Misc devices
++#
++CONFIG_IBM_ASM=m
+
-+ switch (shortcut) {
-+ case APIC_DEST_SELF:
-+ __send_IPI_one(smp_processor_id(), vector);
-+ break;
-+ case APIC_DEST_ALLBUT:
-+ for (cpu = 0; cpu < NR_CPUS; ++cpu) {
-+ if (cpu == smp_processor_id())
-+ continue;
-+ if (cpu_isset(cpu, cpu_online_map)) {
-+ __send_IPI_one(cpu, vector);
-+ }
-+ }
-+ break;
-+ case APIC_DEST_ALLINC:
-+ for (cpu = 0; cpu < NR_CPUS; ++cpu) {
-+ if (cpu_isset(cpu, cpu_online_map)) {
-+ __send_IPI_one(cpu, vector);
-+ }
-+ }
-+ break;
-+ default:
-+ printk("XXXXXX __send_IPI_shortcut %08x vector %d\n", shortcut,
-+ vector);
-+ break;
-+ }
-+}
++#
++# Multimedia devices
++#
++CONFIG_VIDEO_DEV=m
++CONFIG_VIDEO_V4L1=y
++CONFIG_VIDEO_V4L1_COMPAT=y
++CONFIG_VIDEO_V4L2=y
+
-+static cpumask_t xen_target_cpus(void)
-+{
-+ return cpu_online_map;
-+}
++#
++# Video Capture Adapters
++#
+
-+/*
-+ * Set up the logical destination ID.
-+ * Do nothing, not called now.
-+ */
-+static void xen_init_apic_ldr(void)
-+{
-+ Dprintk("%s\n", __FUNCTION__);
-+ return;
-+}
++#
++# Video Capture Adapters
++#
++# CONFIG_VIDEO_ADV_DEBUG is not set
++CONFIG_VIDEO_VIVI=m
++CONFIG_VIDEO_BT848=m
++CONFIG_VIDEO_BT848_DVB=y
++CONFIG_VIDEO_SAA6588=m
++CONFIG_VIDEO_BWQCAM=m
++CONFIG_VIDEO_CQCAM=m
++CONFIG_VIDEO_W9966=m
++CONFIG_VIDEO_CPIA=m
++CONFIG_VIDEO_CPIA_PP=m
++CONFIG_VIDEO_CPIA_USB=m
++CONFIG_VIDEO_CPIA2=m
++CONFIG_VIDEO_SAA5246A=m
++CONFIG_VIDEO_SAA5249=m
++CONFIG_TUNER_3036=m
++CONFIG_VIDEO_STRADIS=m
++CONFIG_VIDEO_ZORAN=m
++CONFIG_VIDEO_ZORAN_BUZ=m
++CONFIG_VIDEO_ZORAN_DC10=m
++CONFIG_VIDEO_ZORAN_DC30=m
++CONFIG_VIDEO_ZORAN_LML33=m
++CONFIG_VIDEO_ZORAN_LML33R10=m
++CONFIG_VIDEO_ZORAN_AVS6EYES=m
++CONFIG_VIDEO_MEYE=m
++CONFIG_VIDEO_SAA7134=m
++CONFIG_VIDEO_SAA7134_ALSA=m
++# CONFIG_VIDEO_SAA7134_OSS is not set
++CONFIG_VIDEO_SAA7134_DVB=m
++CONFIG_VIDEO_SAA7134_DVB_ALL_FRONTENDS=y
++CONFIG_VIDEO_MXB=m
++CONFIG_VIDEO_DPC=m
++CONFIG_VIDEO_HEXIUM_ORION=m
++CONFIG_VIDEO_HEXIUM_GEMINI=m
++CONFIG_VIDEO_CX88_VP3054=m
++CONFIG_VIDEO_CX88=m
++CONFIG_VIDEO_CX88_ALSA=m
++CONFIG_VIDEO_CX88_BLACKBIRD=m
++CONFIG_VIDEO_CX88_DVB=m
++CONFIG_VIDEO_CX88_DVB_ALL_FRONTENDS=y
+
-+static void xen_send_IPI_allbutself(int vector)
-+{
-+ /*
-+ * if there are no other CPUs in the system then
-+ * we get an APIC send error if we try to broadcast.
-+ * thus we have to avoid sending IPIs in this case.
-+ */
-+ Dprintk("%s\n", __FUNCTION__);
-+ if (num_online_cpus() > 1)
-+ xen_send_IPI_shortcut(APIC_DEST_ALLBUT, vector, APIC_DEST_LOGICAL);
-+}
++#
++# Encoders and Decoders
++#
++CONFIG_VIDEO_MSP3400=m
++CONFIG_VIDEO_CS53L32A=m
++CONFIG_VIDEO_TLV320AIC23B=m
++CONFIG_VIDEO_WM8775=m
++CONFIG_VIDEO_WM8739=m
++CONFIG_VIDEO_CX2341X=m
++CONFIG_VIDEO_CX25840=m
++CONFIG_VIDEO_SAA711X=m
++CONFIG_VIDEO_SAA7127=m
++CONFIG_VIDEO_UPD64031A=m
++CONFIG_VIDEO_UPD64083=m
+
-+static void xen_send_IPI_all(int vector)
-+{
-+ Dprintk("%s\n", __FUNCTION__);
-+ xen_send_IPI_shortcut(APIC_DEST_ALLINC, vector, APIC_DEST_LOGICAL);
-+}
++#
++# V4L USB devices
++#
++CONFIG_VIDEO_PVRUSB2=m
++CONFIG_VIDEO_PVRUSB2_24XXX=y
++CONFIG_VIDEO_PVRUSB2_SYSFS=y
++# CONFIG_VIDEO_PVRUSB2_DEBUGIFC is not set
++CONFIG_VIDEO_EM28XX=m
++CONFIG_VIDEO_USBVIDEO=m
++CONFIG_USB_VICAM=m
++CONFIG_USB_IBMCAM=m
++CONFIG_USB_KONICAWC=m
++CONFIG_USB_QUICKCAM_MESSENGER=m
++CONFIG_USB_ET61X251=m
++CONFIG_VIDEO_OVCAMCHIP=m
++CONFIG_USB_W9968CF=m
++CONFIG_USB_OV511=m
++CONFIG_USB_SE401=m
++CONFIG_USB_SN9C102=m
++CONFIG_USB_STV680=m
++CONFIG_USB_ZC0301=m
++CONFIG_USB_PWC=m
++# CONFIG_USB_PWC_DEBUG is not set
+
-+static void xen_send_IPI_mask(cpumask_t cpumask, int vector)
-+{
-+ unsigned long mask = cpus_addr(cpumask)[0];
-+ unsigned int cpu;
-+ unsigned long flags;
++#
++# Radio Adapters
++#
++CONFIG_RADIO_GEMTEK_PCI=m
++CONFIG_RADIO_MAXIRADIO=m
++CONFIG_RADIO_MAESTRO=m
++CONFIG_USB_DSBR=m
+
-+ Dprintk("%s\n", __FUNCTION__);
-+ local_irq_save(flags);
-+ WARN_ON(mask & ~cpus_addr(cpu_online_map)[0]);
++#
++# Digital Video Broadcasting Devices
++#
++CONFIG_DVB=y
++CONFIG_DVB_CORE=m
+
-+ for (cpu = 0; cpu < NR_CPUS; ++cpu) {
-+ if (cpu_isset(cpu, cpumask)) {
-+ __send_IPI_one(cpu, vector);
-+ }
-+ }
-+ local_irq_restore(flags);
-+}
++#
++# Supported SAA7146 based PCI Adapters
++#
++CONFIG_DVB_AV7110=m
++CONFIG_DVB_AV7110_OSD=y
++CONFIG_DVB_BUDGET=m
++CONFIG_DVB_BUDGET_CI=m
++CONFIG_DVB_BUDGET_AV=m
++CONFIG_DVB_BUDGET_PATCH=m
+
-+#ifdef CONFIG_XEN_PRIVILEGED_GUEST
-+static int xen_apic_id_registered(void)
-+{
-+ /* better be set */
-+ Dprintk("%s\n", __FUNCTION__);
-+ return physid_isset(smp_processor_id(), phys_cpu_present_map);
-+}
-+#endif
++#
++# Supported USB Adapters
++#
++CONFIG_DVB_USB=m
++# CONFIG_DVB_USB_DEBUG is not set
++CONFIG_DVB_USB_A800=m
++CONFIG_DVB_USB_DIBUSB_MB=m
++# CONFIG_DVB_USB_DIBUSB_MB_FAULTY is not set
++CONFIG_DVB_USB_DIBUSB_MC=m
++CONFIG_DVB_USB_UMT_010=m
++CONFIG_DVB_USB_CXUSB=m
++CONFIG_DVB_USB_DIGITV=m
++CONFIG_DVB_USB_VP7045=m
++CONFIG_DVB_USB_VP702X=m
++CONFIG_DVB_USB_GP8PSK=m
++CONFIG_DVB_USB_NOVA_T_USB2=m
++CONFIG_DVB_USB_DTT200U=m
++CONFIG_DVB_TTUSB_BUDGET=m
++CONFIG_DVB_TTUSB_DEC=m
++CONFIG_DVB_CINERGYT2=m
++# CONFIG_DVB_CINERGYT2_TUNING is not set
+
-+static unsigned int xen_cpu_mask_to_apicid(cpumask_t cpumask)
-+{
-+ Dprintk("%s\n", __FUNCTION__);
-+ return cpus_addr(cpumask)[0] & APIC_ALL_CPUS;
-+}
++#
++# Supported FlexCopII (B2C2) Adapters
++#
++CONFIG_DVB_B2C2_FLEXCOP=m
++CONFIG_DVB_B2C2_FLEXCOP_PCI=m
++CONFIG_DVB_B2C2_FLEXCOP_USB=m
++# CONFIG_DVB_B2C2_FLEXCOP_DEBUG is not set
+
-+static unsigned int phys_pkg_id(int index_msb)
-+{
-+ u32 ebx;
++#
++# Supported BT878 Adapters
++#
++CONFIG_DVB_BT8XX=m
+
-+ Dprintk("%s\n", __FUNCTION__);
-+ ebx = cpuid_ebx(1);
-+ return ((ebx >> 24) & 0xFF) >> index_msb;
-+}
++#
++# Supported Pluto2 Adapters
++#
++CONFIG_DVB_PLUTO2=m
+
-+struct genapic apic_xen = {
-+ .name = "xen",
-+#ifdef CONFIG_XEN_PRIVILEGED_GUEST
-+ .int_delivery_mode = dest_LowestPrio,
-+#endif
-+ .int_dest_mode = (APIC_DEST_LOGICAL != 0),
-+ .int_delivery_dest = APIC_DEST_LOGICAL | APIC_DM_LOWEST,
-+ .target_cpus = xen_target_cpus,
-+#ifdef CONFIG_XEN_PRIVILEGED_GUEST
-+ .apic_id_registered = xen_apic_id_registered,
-+#endif
-+ .init_apic_ldr = xen_init_apic_ldr,
-+ .send_IPI_all = xen_send_IPI_all,
-+ .send_IPI_allbutself = xen_send_IPI_allbutself,
-+ .send_IPI_mask = xen_send_IPI_mask,
-+ .cpu_mask_to_apicid = xen_cpu_mask_to_apicid,
-+ .phys_pkg_id = phys_pkg_id,
-+};
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/genapic-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/genapic-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/genapic-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/genapic-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,143 @@
-+/*
-+ * Copyright 2004 James Cleverdon, IBM.
-+ * Subject to the GNU Public License, v.2
-+ *
-+ * Generic APIC sub-arch probe layer.
-+ *
-+ * Hacked for x86-64 by James Cleverdon from i386 architecture code by
-+ * Martin Bligh, Andi Kleen, James Bottomley, John Stultz, and
-+ * James Cleverdon.
-+ */
-+#include <linux/threads.h>
-+#include <linux/cpumask.h>
-+#include <linux/string.h>
-+#include <linux/kernel.h>
-+#include <linux/ctype.h>
-+#include <linux/init.h>
-+#include <linux/module.h>
++#
++# Supported DVB Frontends
++#
+
-+#include <asm/smp.h>
-+#include <asm/ipi.h>
++#
++# Customise DVB Frontends
++#
+
-+#if defined(CONFIG_ACPI)
-+#include <acpi/acpi_bus.h>
-+#endif
++#
++# DVB-S (satellite) frontends
++#
++CONFIG_DVB_STV0299=m
++CONFIG_DVB_CX24110=m
++CONFIG_DVB_CX24123=m
++CONFIG_DVB_TDA8083=m
++CONFIG_DVB_MT312=m
++CONFIG_DVB_VES1X93=m
++CONFIG_DVB_S5H1420=m
+
-+/* which logical CPU number maps to which CPU (physical APIC ID) */
-+u8 x86_cpu_to_apicid[NR_CPUS] __read_mostly = { [0 ... NR_CPUS-1] = BAD_APICID };
-+EXPORT_SYMBOL(x86_cpu_to_apicid);
-+u8 x86_cpu_to_log_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = BAD_APICID };
++#
++# DVB-T (terrestrial) frontends
++#
++CONFIG_DVB_SP8870=m
++CONFIG_DVB_SP887X=m
++CONFIG_DVB_CX22700=m
++CONFIG_DVB_CX22702=m
++CONFIG_DVB_L64781=m
++CONFIG_DVB_TDA1004X=m
++CONFIG_DVB_NXT6000=m
++CONFIG_DVB_MT352=m
++CONFIG_DVB_ZL10353=m
++CONFIG_DVB_DIB3000MB=m
++CONFIG_DVB_DIB3000MC=m
+
-+extern struct genapic apic_cluster;
-+extern struct genapic apic_flat;
-+extern struct genapic apic_physflat;
++#
++# DVB-C (cable) frontends
++#
++CONFIG_DVB_VES1820=m
++CONFIG_DVB_TDA10021=m
++CONFIG_DVB_STV0297=m
+
-+#ifndef CONFIG_XEN
-+struct genapic *genapic = &apic_flat;
-+#else
-+extern struct genapic apic_xen;
-+struct genapic *genapic = &apic_xen;
-+#endif
++#
++# ATSC (North American/Korean Terrestrial/Cable DTV) frontends
++#
++CONFIG_DVB_NXT200X=m
++CONFIG_DVB_OR51211=m
++CONFIG_DVB_OR51132=m
++CONFIG_DVB_BCM3510=m
++CONFIG_DVB_LGDT330X=m
+
++#
++# Miscellaneous devices
++#
++CONFIG_DVB_PLL=m
++CONFIG_DVB_LNBP21=m
++CONFIG_DVB_ISL6421=m
++CONFIG_VIDEO_SAA7146=m
++CONFIG_VIDEO_SAA7146_VV=m
++CONFIG_VIDEO_VIDEOBUF=m
++CONFIG_VIDEO_TUNER=m
++CONFIG_VIDEO_BUF=m
++CONFIG_VIDEO_BUF_DVB=m
++CONFIG_VIDEO_BTCX=m
++CONFIG_VIDEO_IR=m
++CONFIG_VIDEO_TVEEPROM=m
++CONFIG_USB_DABUSB=m
+
-+/*
-+ * Check the APIC IDs in bios_cpu_apicid and choose the APIC mode.
-+ */
-+void __init clustered_apic_check(void)
-+{
-+#ifndef CONFIG_XEN
-+ long i;
-+ u8 clusters, max_cluster;
-+ u8 id;
-+ u8 cluster_cnt[NUM_APIC_CLUSTERS];
-+ int max_apic = 0;
++#
++# Graphics support
++#
++CONFIG_FIRMWARE_EDID=y
++CONFIG_FB=y
++CONFIG_FB_CFB_FILLRECT=y
++CONFIG_FB_CFB_COPYAREA=y
++CONFIG_FB_CFB_IMAGEBLIT=y
++# CONFIG_FB_MACMODES is not set
++# CONFIG_FB_BACKLIGHT is not set
++CONFIG_FB_MODE_HELPERS=y
++CONFIG_FB_TILEBLITTING=y
++CONFIG_FB_CIRRUS=m
++CONFIG_FB_PM2=m
++CONFIG_FB_PM2_FIFO_DISCONNECT=y
++CONFIG_FB_CYBER2000=m
++CONFIG_FB_ARC=m
++# CONFIG_FB_ASILIANT is not set
++# CONFIG_FB_IMSTT is not set
++CONFIG_FB_VGA16=m
++CONFIG_FB_VESA=y
++CONFIG_FB_HGA=m
++# CONFIG_FB_HGA_ACCEL is not set
++CONFIG_FB_S1D13XXX=m
++CONFIG_FB_NVIDIA=m
++CONFIG_FB_NVIDIA_I2C=y
++CONFIG_FB_RIVA=m
++CONFIG_FB_RIVA_I2C=y
++CONFIG_FB_RIVA_DEBUG=y
++CONFIG_FB_I810=m
++# CONFIG_FB_I810_GTF is not set
++CONFIG_FB_INTEL=m
++# CONFIG_FB_INTEL_DEBUG is not set
++CONFIG_FB_MATROX=m
++CONFIG_FB_MATROX_MILLENIUM=y
++CONFIG_FB_MATROX_MYSTIQUE=y
++CONFIG_FB_MATROX_G=y
++# CONFIG_FB_MATROX_I2C is not set
++CONFIG_FB_MATROX_MULTIHEAD=y
++CONFIG_FB_RADEON=m
++CONFIG_FB_RADEON_I2C=y
++# CONFIG_FB_RADEON_DEBUG is not set
++CONFIG_FB_ATY128=m
++CONFIG_FB_ATY=m
++CONFIG_FB_ATY_CT=y
++CONFIG_FB_ATY_GENERIC_LCD=y
++CONFIG_FB_ATY_GX=y
++CONFIG_FB_SAVAGE=m
++CONFIG_FB_SAVAGE_I2C=y
++CONFIG_FB_SAVAGE_ACCEL=y
++CONFIG_FB_SIS=m
++CONFIG_FB_SIS_300=y
++CONFIG_FB_SIS_315=y
++CONFIG_FB_NEOMAGIC=m
++CONFIG_FB_KYRO=m
++CONFIG_FB_3DFX=m
++# CONFIG_FB_3DFX_ACCEL is not set
++CONFIG_FB_VOODOO1=m
++CONFIG_FB_CYBLA=m
++CONFIG_FB_TRIDENT=m
++# CONFIG_FB_TRIDENT_ACCEL is not set
++CONFIG_FB_GEODE=y
++CONFIG_FB_GEODE_GX=m
++CONFIG_FB_GEODE_GX1=m
++CONFIG_FB_VIRTUAL=m
+
-+#if defined(CONFIG_ACPI)
-+ /*
-+ * Some x86_64 machines use physical APIC mode regardless of how many
-+ * procs/clusters are present (x86_64 ES7000 is an example).
-+ */
-+ if (acpi_fadt.revision > FADT2_REVISION_ID)
-+ if (acpi_fadt.force_apic_physical_destination_mode) {
-+ genapic = &apic_cluster;
-+ goto print;
-+ }
-+#endif
++#
++# Console display driver support
++#
++CONFIG_VGA_CONSOLE=y
++# CONFIG_VGACON_SOFT_SCROLLBACK is not set
++CONFIG_VIDEO_SELECT=y
++CONFIG_DUMMY_CONSOLE=y
++CONFIG_FRAMEBUFFER_CONSOLE=m
++# CONFIG_FRAMEBUFFER_CONSOLE_ROTATION is not set
++# CONFIG_FONTS is not set
++CONFIG_FONT_8x8=y
++CONFIG_FONT_8x16=y
+
-+ memset(cluster_cnt, 0, sizeof(cluster_cnt));
-+ for (i = 0; i < NR_CPUS; i++) {
-+ id = bios_cpu_apicid[i];
-+ if (id == BAD_APICID)
-+ continue;
-+ if (id > max_apic)
-+ max_apic = id;
-+ cluster_cnt[APIC_CLUSTERID(id)]++;
-+ }
++#
++# Logo configuration
++#
++# CONFIG_LOGO is not set
++CONFIG_BACKLIGHT_LCD_SUPPORT=y
++CONFIG_BACKLIGHT_CLASS_DEVICE=m
++CONFIG_BACKLIGHT_DEVICE=y
++CONFIG_LCD_CLASS_DEVICE=m
++CONFIG_LCD_DEVICE=y
+
-+ /* Don't use clustered mode on AMD platforms. */
-+ if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD) {
-+ genapic = &apic_physflat;
-+#ifndef CONFIG_HOTPLUG_CPU
-+ /* In the CPU hotplug case we cannot use broadcast mode
-+ because that opens a race when a CPU is removed.
-+ Stay at physflat mode in this case.
-+ It is bad to do this unconditionally though. Once
-+ we have ACPI platform support for CPU hotplug
-+ we should detect hotplug capablity from ACPI tables and
-+ only do this when really needed. -AK */
-+ if (max_apic <= 8)
-+ genapic = &apic_flat;
-+#endif
-+ goto print;
-+ }
++#
++# Sound
++#
++CONFIG_SOUND=m
+
-+ clusters = 0;
-+ max_cluster = 0;
++#
++# Advanced Linux Sound Architecture
++#
++CONFIG_SND=m
++CONFIG_SND_TIMER=m
++CONFIG_SND_PCM=m
++CONFIG_SND_HWDEP=m
++CONFIG_SND_RAWMIDI=m
++CONFIG_SND_SEQUENCER=m
++CONFIG_SND_SEQ_DUMMY=m
++CONFIG_SND_OSSEMUL=y
++CONFIG_SND_MIXER_OSS=m
++CONFIG_SND_PCM_OSS=m
++CONFIG_SND_PCM_OSS_PLUGINS=y
++CONFIG_SND_SEQUENCER_OSS=y
++CONFIG_SND_RTCTIMER=m
++CONFIG_SND_SEQ_RTCTIMER_DEFAULT=y
++# CONFIG_SND_DYNAMIC_MINORS is not set
++CONFIG_SND_SUPPORT_OLD_API=y
++CONFIG_SND_VERBOSE_PROCFS=y
++# CONFIG_SND_VERBOSE_PRINTK is not set
++# CONFIG_SND_DEBUG is not set
+
-+ for (i = 0; i < NUM_APIC_CLUSTERS; i++) {
-+ if (cluster_cnt[i] > 0) {
-+ ++clusters;
-+ if (cluster_cnt[i] > max_cluster)
-+ max_cluster = cluster_cnt[i];
-+ }
-+ }
++#
++# Generic devices
++#
++CONFIG_SND_MPU401_UART=m
++CONFIG_SND_OPL3_LIB=m
++CONFIG_SND_VX_LIB=m
++CONFIG_SND_AC97_CODEC=m
++CONFIG_SND_AC97_BUS=m
++CONFIG_SND_DUMMY=m
++CONFIG_SND_VIRMIDI=m
++CONFIG_SND_MTPAV=m
++CONFIG_SND_SERIAL_U16550=m
++CONFIG_SND_MPU401=m
+
-+ /*
-+ * If we have clusters <= 1 and CPUs <= 8 in cluster 0, then flat mode,
-+ * else if max_cluster <= 4 and cluster_cnt[15] == 0, clustered logical
-+ * else physical mode.
-+ * (We don't use lowest priority delivery + HW APIC IRQ steering, so
-+ * can ignore the clustered logical case and go straight to physical.)
-+ */
-+ if (clusters <= 1 && max_cluster <= 8 && cluster_cnt[0] == max_cluster) {
-+#ifdef CONFIG_HOTPLUG_CPU
-+ /* Don't use APIC shortcuts in CPU hotplug to avoid races */
-+ genapic = &apic_physflat;
-+#else
-+ genapic = &apic_flat;
-+#endif
-+ } else
-+ genapic = &apic_cluster;
++#
++# PCI devices
++#
++CONFIG_SND_AD1889=m
++CONFIG_SND_ALS300=m
++CONFIG_SND_ALS4000=m
++CONFIG_SND_ALI5451=m
++CONFIG_SND_ATIIXP=m
++CONFIG_SND_ATIIXP_MODEM=m
++CONFIG_SND_AU8810=m
++CONFIG_SND_AU8820=m
++CONFIG_SND_AU8830=m
++CONFIG_SND_AZT3328=m
++CONFIG_SND_BT87X=m
++# CONFIG_SND_BT87X_OVERCLOCK is not set
++CONFIG_SND_CA0106=m
++CONFIG_SND_CMIPCI=m
++CONFIG_SND_CS4281=m
++CONFIG_SND_CS46XX=m
++CONFIG_SND_CS46XX_NEW_DSP=y
++CONFIG_SND_CS5535AUDIO=m
++CONFIG_SND_DARLA20=m
++CONFIG_SND_GINA20=m
++CONFIG_SND_LAYLA20=m
++CONFIG_SND_DARLA24=m
++CONFIG_SND_GINA24=m
++CONFIG_SND_LAYLA24=m
++CONFIG_SND_MONA=m
++CONFIG_SND_MIA=m
++CONFIG_SND_ECHO3G=m
++CONFIG_SND_INDIGO=m
++CONFIG_SND_INDIGOIO=m
++CONFIG_SND_INDIGODJ=m
++CONFIG_SND_EMU10K1=m
++CONFIG_SND_EMU10K1X=m
++CONFIG_SND_ENS1370=m
++CONFIG_SND_ENS1371=m
++CONFIG_SND_ES1938=m
++CONFIG_SND_ES1968=m
++CONFIG_SND_FM801=m
++# CONFIG_SND_FM801_TEA575X_BOOL is not set
++CONFIG_SND_HDA_INTEL=m
++CONFIG_SND_HDSP=m
++CONFIG_SND_HDSPM=m
++CONFIG_SND_ICE1712=m
++CONFIG_SND_ICE1724=m
++CONFIG_SND_INTEL8X0=m
++CONFIG_SND_INTEL8X0M=m
++CONFIG_SND_KORG1212=m
++CONFIG_SND_MAESTRO3=m
++CONFIG_SND_MIXART=m
++CONFIG_SND_NM256=m
++CONFIG_SND_PCXHR=m
++CONFIG_SND_RIPTIDE=m
++CONFIG_SND_RME32=m
++CONFIG_SND_RME96=m
++CONFIG_SND_RME9652=m
++CONFIG_SND_SONICVIBES=m
++CONFIG_SND_TRIDENT=m
++CONFIG_SND_VIA82XX=m
++# CONFIG_SND_VIA82XX_MODEM is not set
++CONFIG_SND_VX222=m
++CONFIG_SND_YMFPCI=m
+
-+print:
-+#else
-+ /* hardcode to xen apic functions */
-+ genapic = &apic_xen;
-+#endif
-+ printk(KERN_INFO "Setting APIC routing to %s\n", genapic->name);
-+}
++#
++# USB devices
++#
++CONFIG_SND_USB_AUDIO=m
++CONFIG_SND_USB_USX2Y=m
+
-+/* Same for both flat and clustered. */
++#
++# PCMCIA devices
++#
++CONFIG_SND_VXPOCKET=m
++CONFIG_SND_PDAUDIOCF=m
+
-+#ifdef CONFIG_XEN
-+extern void xen_send_IPI_shortcut(unsigned int shortcut, int vector, unsigned int dest);
-+#endif
++#
++# Open Sound System
++#
++CONFIG_SOUND_PRIME=m
++# CONFIG_OSS_OBSOLETE_DRIVER is not set
++CONFIG_SOUND_BT878=m
++CONFIG_SOUND_ES1371=m
++CONFIG_SOUND_ICH=m
++CONFIG_SOUND_TRIDENT=m
++# CONFIG_SOUND_MSNDCLAS is not set
++# CONFIG_SOUND_MSNDPIN is not set
++CONFIG_SOUND_VIA82CXXX=m
++# CONFIG_MIDI_VIA82CXXX is not set
++# CONFIG_SOUND_OSS is not set
++CONFIG_SOUND_TVMIXER=m
+
-+void send_IPI_self(int vector)
-+{
-+#ifndef CONFIG_XEN
-+ __send_IPI_shortcut(APIC_DEST_SELF, vector, APIC_DEST_PHYSICAL);
-+#else
-+ xen_send_IPI_shortcut(APIC_DEST_SELF, vector, APIC_DEST_PHYSICAL);
-+#endif
-+}
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/head64-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/head64-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/head64-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/head64-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,162 @@
-+/*
-+ * linux/arch/x86_64/kernel/head64.c -- prepare to run common code
-+ *
-+ * Copyright (C) 2000 Andrea Arcangeli <andrea at suse.de> SuSE
-+ *
-+ * Jun Nakajima <jun.nakajima at intel.com>
-+ * Modified for Xen.
-+ */
++#
++# USB support
++#
++CONFIG_USB_ARCH_HAS_HCD=y
++CONFIG_USB_ARCH_HAS_OHCI=y
++CONFIG_USB_ARCH_HAS_EHCI=y
++CONFIG_USB=m
++# CONFIG_USB_DEBUG is not set
+
-+#include <linux/init.h>
-+#include <linux/linkage.h>
-+#include <linux/types.h>
-+#include <linux/kernel.h>
-+#include <linux/string.h>
-+#include <linux/percpu.h>
-+#include <linux/module.h>
++#
++# Miscellaneous USB options
++#
++CONFIG_USB_DEVICEFS=y
++CONFIG_USB_BANDWIDTH=y
++# CONFIG_USB_DYNAMIC_MINORS is not set
++# CONFIG_USB_SUSPEND is not set
++# CONFIG_USB_OTG is not set
+
-+#include <asm/processor.h>
-+#include <asm/proto.h>
-+#include <asm/smp.h>
-+#include <asm/bootsetup.h>
-+#include <asm/setup.h>
-+#include <asm/desc.h>
-+#include <asm/pgtable.h>
-+#include <asm/sections.h>
++#
++# USB Host Controller Drivers
++#
++CONFIG_USB_EHCI_HCD=m
++CONFIG_USB_EHCI_SPLIT_ISO=y
++CONFIG_USB_EHCI_ROOT_HUB_TT=y
++CONFIG_USB_EHCI_TT_NEWSCHED=y
++CONFIG_USB_ISP116X_HCD=m
++CONFIG_USB_OHCI_HCD=m
++# CONFIG_USB_OHCI_BIG_ENDIAN is not set
++CONFIG_USB_OHCI_LITTLE_ENDIAN=y
++CONFIG_USB_UHCI_HCD=m
++CONFIG_USB_SL811_HCD=m
++CONFIG_USB_SL811_CS=m
+
-+unsigned long start_pfn;
++#
++# USB Device Class drivers
++#
++CONFIG_USB_ACM=m
++CONFIG_USB_PRINTER=m
+
-+/* Don't add a printk in there. printk relies on the PDA which is not initialized
-+ yet. */
-+#if 0
-+static void __init clear_bss(void)
-+{
-+ memset(__bss_start, 0,
-+ (unsigned long) __bss_stop - (unsigned long) __bss_start);
-+}
-+#endif
++#
++# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support'
++#
+
-+#define NEW_CL_POINTER 0x228 /* Relative to real mode data */
-+#define OLD_CL_MAGIC_ADDR 0x90020
-+#define OLD_CL_MAGIC 0xA33F
-+#define OLD_CL_BASE_ADDR 0x90000
-+#define OLD_CL_OFFSET 0x90022
++#
++# may also be needed; see USB_STORAGE Help for more information
++#
++CONFIG_USB_STORAGE=m
++# CONFIG_USB_STORAGE_DEBUG is not set
++CONFIG_USB_STORAGE_DATAFAB=y
++CONFIG_USB_STORAGE_FREECOM=y
++CONFIG_USB_STORAGE_ISD200=y
++CONFIG_USB_STORAGE_DPCM=y
++CONFIG_USB_STORAGE_USBAT=y
++CONFIG_USB_STORAGE_SDDR09=y
++CONFIG_USB_STORAGE_SDDR55=y
++CONFIG_USB_STORAGE_JUMPSHOT=y
++CONFIG_USB_STORAGE_ALAUDA=y
++CONFIG_USB_LIBUSUAL=y
+
-+extern char saved_command_line[];
++#
++# USB Input Devices
++#
++CONFIG_USB_HID=m
++CONFIG_USB_HIDINPUT=y
++# CONFIG_USB_HIDINPUT_POWERBOOK is not set
++CONFIG_HID_FF=y
++CONFIG_HID_PID=y
++CONFIG_LOGITECH_FF=y
++CONFIG_THRUSTMASTER_FF=y
++CONFIG_USB_HIDDEV=y
+
-+static void __init copy_bootdata(char *real_mode_data)
-+{
-+#ifndef CONFIG_XEN
-+ int new_data;
-+ char * command_line;
++#
++# USB HID Boot Protocol drivers
++#
++# CONFIG_USB_KBD is not set
++# CONFIG_USB_MOUSE is not set
++CONFIG_USB_AIPTEK=m
++CONFIG_USB_WACOM=m
++CONFIG_USB_ACECAD=m
++CONFIG_USB_KBTAB=m
++CONFIG_USB_POWERMATE=m
++CONFIG_USB_TOUCHSCREEN=m
++CONFIG_USB_TOUCHSCREEN_EGALAX=y
++CONFIG_USB_TOUCHSCREEN_PANJIT=y
++CONFIG_USB_TOUCHSCREEN_3M=y
++CONFIG_USB_TOUCHSCREEN_ITM=y
++CONFIG_USB_YEALINK=m
++CONFIG_USB_XPAD=m
++CONFIG_USB_ATI_REMOTE=m
++CONFIG_USB_ATI_REMOTE2=m
++CONFIG_USB_KEYSPAN_REMOTE=m
++CONFIG_USB_APPLETOUCH=m
+
-+ memcpy(x86_boot_params, real_mode_data, BOOT_PARAM_SIZE);
-+ new_data = *(int *) (x86_boot_params + NEW_CL_POINTER);
-+ if (!new_data) {
-+ if (OLD_CL_MAGIC != * (u16 *) OLD_CL_MAGIC_ADDR) {
-+ printk("so old bootloader that it does not support commandline?!\n");
-+ return;
-+ }
-+ new_data = OLD_CL_BASE_ADDR + * (u16 *) OLD_CL_OFFSET;
-+ printk("old bootloader convention, maybe loadlin?\n");
-+ }
-+ command_line = (char *) ((u64)(new_data));
-+ memcpy(saved_command_line, command_line, COMMAND_LINE_SIZE);
-+#else
-+ int max_cmdline;
-+
-+ if ((max_cmdline = MAX_GUEST_CMDLINE) > COMMAND_LINE_SIZE)
-+ max_cmdline = COMMAND_LINE_SIZE;
-+ memcpy(saved_command_line, xen_start_info->cmd_line, max_cmdline);
-+ saved_command_line[max_cmdline-1] = '\0';
-+#endif
-+ printk("Bootdata ok (command line is %s)\n", saved_command_line);
-+}
++#
++# USB Imaging devices
++#
++CONFIG_USB_MDC800=m
++CONFIG_USB_MICROTEK=m
+
-+static void __init setup_boot_cpu_data(void)
-+{
-+ unsigned int dummy, eax;
++#
++# USB Network Adapters
++#
++CONFIG_USB_CATC=m
++CONFIG_USB_KAWETH=m
++CONFIG_USB_PEGASUS=m
++CONFIG_USB_RTL8150=m
++CONFIG_USB_USBNET=m
++CONFIG_USB_NET_AX8817X=m
++CONFIG_USB_NET_CDCETHER=m
++CONFIG_USB_NET_GL620A=m
++CONFIG_USB_NET_NET1080=m
++CONFIG_USB_NET_PLUSB=m
++CONFIG_USB_NET_RNDIS_HOST=m
++CONFIG_USB_NET_CDC_SUBSET=m
++CONFIG_USB_ALI_M5632=y
++CONFIG_USB_AN2720=y
++CONFIG_USB_BELKIN=y
++CONFIG_USB_ARMLINUX=y
++# CONFIG_USB_EPSON2888 is not set
++CONFIG_USB_NET_ZAURUS=m
++CONFIG_USB_MON=y
+
-+ /* get vendor info */
-+ cpuid(0, (unsigned int *)&boot_cpu_data.cpuid_level,
-+ (unsigned int *)&boot_cpu_data.x86_vendor_id[0],
-+ (unsigned int *)&boot_cpu_data.x86_vendor_id[8],
-+ (unsigned int *)&boot_cpu_data.x86_vendor_id[4]);
++#
++# USB port drivers
++#
++CONFIG_USB_USS720=m
+
-+ /* get cpu type */
-+ cpuid(1, &eax, &dummy, &dummy,
-+ (unsigned int *) &boot_cpu_data.x86_capability);
-+ boot_cpu_data.x86 = (eax >> 8) & 0xf;
-+ boot_cpu_data.x86_model = (eax >> 4) & 0xf;
-+ boot_cpu_data.x86_mask = eax & 0xf;
-+}
++#
++# USB Serial Converter support
++#
++CONFIG_USB_SERIAL=m
++CONFIG_USB_SERIAL_GENERIC=y
++CONFIG_USB_SERIAL_AIRPRIME=m
++CONFIG_USB_SERIAL_ARK3116=m
++CONFIG_USB_SERIAL_BELKIN=m
++CONFIG_USB_SERIAL_WHITEHEAT=m
++CONFIG_USB_SERIAL_DIGI_ACCELEPORT=m
++CONFIG_USB_SERIAL_CP2101=m
++CONFIG_USB_SERIAL_CYPRESS_M8=m
++CONFIG_USB_SERIAL_EMPEG=m
++CONFIG_USB_SERIAL_FTDI_SIO=m
++CONFIG_USB_SERIAL_FUNSOFT=m
++CONFIG_USB_SERIAL_VISOR=m
++CONFIG_USB_SERIAL_IPAQ=m
++CONFIG_USB_SERIAL_IR=m
++CONFIG_USB_SERIAL_EDGEPORT=m
++CONFIG_USB_SERIAL_EDGEPORT_TI=m
++CONFIG_USB_SERIAL_GARMIN=m
++CONFIG_USB_SERIAL_IPW=m
++CONFIG_USB_SERIAL_KEYSPAN_PDA=m
++CONFIG_USB_SERIAL_KEYSPAN=m
++CONFIG_USB_SERIAL_KEYSPAN_MPR=y
++CONFIG_USB_SERIAL_KEYSPAN_USA28=y
++CONFIG_USB_SERIAL_KEYSPAN_USA28X=y
++CONFIG_USB_SERIAL_KEYSPAN_USA28XA=y
++CONFIG_USB_SERIAL_KEYSPAN_USA28XB=y
++CONFIG_USB_SERIAL_KEYSPAN_USA19=y
++CONFIG_USB_SERIAL_KEYSPAN_USA18X=y
++CONFIG_USB_SERIAL_KEYSPAN_USA19W=y
++CONFIG_USB_SERIAL_KEYSPAN_USA19QW=y
++CONFIG_USB_SERIAL_KEYSPAN_USA19QI=y
++CONFIG_USB_SERIAL_KEYSPAN_USA49W=y
++CONFIG_USB_SERIAL_KEYSPAN_USA49WLC=y
++CONFIG_USB_SERIAL_KLSI=m
++CONFIG_USB_SERIAL_KOBIL_SCT=m
++CONFIG_USB_SERIAL_MCT_U232=m
++CONFIG_USB_SERIAL_NAVMAN=m
++CONFIG_USB_SERIAL_PL2303=m
++CONFIG_USB_SERIAL_HP4X=m
++CONFIG_USB_SERIAL_SAFE=m
++CONFIG_USB_SERIAL_SAFE_PADDED=y
++CONFIG_USB_SERIAL_SIERRAWIRELESS=m
++CONFIG_USB_SERIAL_TI=m
++CONFIG_USB_SERIAL_CYBERJACK=m
++CONFIG_USB_SERIAL_XIRCOM=m
++CONFIG_USB_SERIAL_OPTION=m
++CONFIG_USB_SERIAL_OMNINET=m
++CONFIG_USB_EZUSB=y
+
-+#include <xen/interface/memory.h>
-+unsigned long *machine_to_phys_mapping;
-+EXPORT_SYMBOL(machine_to_phys_mapping);
-+unsigned int machine_to_phys_order;
-+EXPORT_SYMBOL(machine_to_phys_order);
++#
++# USB Miscellaneous drivers
++#
++CONFIG_USB_EMI62=m
++CONFIG_USB_EMI26=m
++CONFIG_USB_AUERSWALD=m
++CONFIG_USB_RIO500=m
++CONFIG_USB_LEGOTOWER=m
++CONFIG_USB_LCD=m
++CONFIG_USB_LED=m
++CONFIG_USB_CYPRESS_CY7C63=m
++CONFIG_USB_CYTHERM=m
++CONFIG_USB_PHIDGETKIT=m
++CONFIG_USB_PHIDGETSERVO=m
++CONFIG_USB_IDMOUSE=m
++CONFIG_USB_APPLEDISPLAY=m
++CONFIG_USB_SISUSBVGA=m
++CONFIG_USB_SISUSBVGA_CON=y
++CONFIG_USB_LD=m
++CONFIG_USB_TEST=m
+
-+void __init x86_64_start_kernel(char * real_mode_data)
-+{
-+ struct xen_machphys_mapping mapping;
-+ unsigned long machine_to_phys_nr_ents;
-+ char *s;
-+ int i;
++#
++# USB DSL modem support
++#
++CONFIG_USB_ATM=m
++CONFIG_USB_SPEEDTOUCH=m
++CONFIG_USB_CXACRU=m
++CONFIG_USB_UEAGLEATM=m
++CONFIG_USB_XUSBATM=m
+
-+ setup_xen_features();
++#
++# USB Gadget Support
++#
++CONFIG_USB_GADGET=m
++# CONFIG_USB_GADGET_DEBUG_FILES is not set
++CONFIG_USB_GADGET_SELECTED=y
++CONFIG_USB_GADGET_NET2280=y
++CONFIG_USB_NET2280=m
++# CONFIG_USB_GADGET_PXA2XX is not set
++# CONFIG_USB_GADGET_GOKU is not set
++# CONFIG_USB_GADGET_LH7A40X is not set
++# CONFIG_USB_GADGET_OMAP is not set
++# CONFIG_USB_GADGET_AT91 is not set
++# CONFIG_USB_GADGET_DUMMY_HCD is not set
++CONFIG_USB_GADGET_DUALSPEED=y
++CONFIG_USB_ZERO=m
++CONFIG_USB_ETH=m
++CONFIG_USB_ETH_RNDIS=y
++CONFIG_USB_GADGETFS=m
++CONFIG_USB_FILE_STORAGE=m
++# CONFIG_USB_FILE_STORAGE_TEST is not set
++CONFIG_USB_G_SERIAL=m
+
-+ xen_start_info = (struct start_info *)real_mode_data;
-+ if (!xen_feature(XENFEAT_auto_translated_physmap))
-+ phys_to_machine_mapping =
-+ (unsigned long *)xen_start_info->mfn_list;
-+ start_pfn = (__pa(xen_start_info->pt_base) >> PAGE_SHIFT) +
-+ xen_start_info->nr_pt_frames;
++#
++# MMC/SD Card support
++#
++CONFIG_MMC=m
++# CONFIG_MMC_DEBUG is not set
++CONFIG_MMC_BLOCK=m
++CONFIG_MMC_SDHCI=m
++CONFIG_MMC_WBSD=m
+
-+ machine_to_phys_mapping = (unsigned long *)MACH2PHYS_VIRT_START;
-+ machine_to_phys_nr_ents = MACH2PHYS_NR_ENTRIES;
-+ if (HYPERVISOR_memory_op(XENMEM_machphys_mapping, &mapping) == 0) {
-+ machine_to_phys_mapping = (unsigned long *)mapping.v_start;
-+ machine_to_phys_nr_ents = mapping.max_mfn + 1;
-+ }
-+ while ((1UL << machine_to_phys_order) < machine_to_phys_nr_ents )
-+ machine_to_phys_order++;
++#
++# LED devices
++#
++CONFIG_NEW_LEDS=y
++CONFIG_LEDS_CLASS=m
+
-+#if 0
-+ for (i = 0; i < 256; i++)
-+ set_intr_gate(i, early_idt_handler);
-+ asm volatile("lidt %0" :: "m" (idt_descr));
-+#endif
++#
++# LED drivers
++#
++CONFIG_LEDS_NET48XX=m
+
-+ /*
-+ * This must be called really, really early:
-+ */
-+ lockdep_init();
++#
++# LED Triggers
++#
++CONFIG_LEDS_TRIGGERS=y
++CONFIG_LEDS_TRIGGER_TIMER=m
++CONFIG_LEDS_TRIGGER_IDE_DISK=y
++CONFIG_LEDS_TRIGGER_HEARTBEAT=m
+
-+ for (i = 0; i < NR_CPUS; i++)
-+ cpu_pda(i) = &boot_cpu_pda[i];
++#
++# InfiniBand support
++#
++CONFIG_INFINIBAND=m
++CONFIG_INFINIBAND_USER_MAD=m
++CONFIG_INFINIBAND_USER_ACCESS=m
++CONFIG_INFINIBAND_ADDR_TRANS=y
++CONFIG_INFINIBAND_MTHCA=m
++CONFIG_INFINIBAND_MTHCA_DEBUG=y
++CONFIG_INFINIBAND_IPOIB=m
++CONFIG_INFINIBAND_IPOIB_DEBUG=y
++# CONFIG_INFINIBAND_IPOIB_DEBUG_DATA is not set
++CONFIG_INFINIBAND_SRP=m
++CONFIG_INFINIBAND_ISER=m
+
-+ pda_init(0);
-+ copy_bootdata(real_mode_data);
-+#ifdef CONFIG_SMP
-+ cpu_set(0, cpu_online_map);
-+#endif
-+ s = strstr(saved_command_line, "earlyprintk=");
-+ if (s != NULL)
-+ setup_early_printk(strchr(s, '=') + 1);
-+#ifdef CONFIG_NUMA
-+ s = strstr(saved_command_line, "numa=");
-+ if (s != NULL)
-+ numa_setup(s+5);
-+#endif
-+#ifdef CONFIG_X86_IO_APIC
-+ if (strstr(saved_command_line, "disableapic"))
-+ disable_apic = 1;
-+#endif
-+ /* You need early console to see that */
-+ if (__pa_symbol(&_end) >= KERNEL_TEXT_SIZE)
-+ panic("Kernel too big for kernel mapping\n");
++#
++# EDAC - error detection and reporting (RAS) (EXPERIMENTAL)
++#
++CONFIG_EDAC=m
+
-+ setup_boot_cpu_data();
-+ start_kernel();
-+}
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/head-xen.S tmp-linux-2.6-xen.patch/arch/x86_64/kernel/head-xen.S
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/head-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/head-xen.S 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,203 @@
-+/*
-+ * linux/arch/x86_64/kernel/head.S -- start in 32bit and switch to 64bit
-+ *
-+ * Copyright (C) 2000 Andrea Arcangeli <andrea at suse.de> SuSE
-+ * Copyright (C) 2000 Pavel Machek <pavel at suse.cz>
-+ * Copyright (C) 2000 Karsten Keil <kkeil at suse.de>
-+ * Copyright (C) 2001,2002 Andi Kleen <ak at suse.de>
-+ *
-+ * $Id: head.S,v 1.49 2002/03/19 17:39:25 ak Exp $
-+ *
-+ * Jun Nakajima <jun.nakajima at intel.com>
-+ * Modified for Xen
-+ */
++#
++# Reporting subsystems
++#
++# CONFIG_EDAC_DEBUG is not set
++CONFIG_EDAC_MM_EDAC=m
++CONFIG_EDAC_AMD76X=m
++CONFIG_EDAC_E7XXX=m
++CONFIG_EDAC_E752X=m
++CONFIG_EDAC_I82875P=m
++CONFIG_EDAC_I82860=m
++CONFIG_EDAC_R82600=m
++CONFIG_EDAC_POLL=y
+
++#
++# Real Time Clock
++#
++CONFIG_RTC_LIB=m
++CONFIG_RTC_CLASS=m
+
-+#include <linux/linkage.h>
-+#include <linux/threads.h>
-+#include <linux/init.h>
-+#include <linux/elfnote.h>
-+#include <asm/desc.h>
-+#include <asm/segment.h>
-+#include <asm/page.h>
-+#include <asm/msr.h>
-+#include <asm/cache.h>
-+#include <asm/dwarf2.h>
-+#include <xen/interface/elfnote.h>
++#
++# RTC interfaces
++#
++CONFIG_RTC_INTF_SYSFS=m
++CONFIG_RTC_INTF_PROC=m
++CONFIG_RTC_INTF_DEV=m
++CONFIG_RTC_INTF_DEV_UIE_EMUL=y
+
-+ .section .bootstrap.text, "ax", @progbits
-+ .code64
-+#define VIRT_ENTRY_OFFSET 0x0
-+.org VIRT_ENTRY_OFFSET
-+ .globl startup_64
-+startup_64:
-+ENTRY(_start)
-+ movq $(init_thread_union+THREAD_SIZE-8),%rsp
++#
++# RTC drivers
++#
++CONFIG_RTC_DRV_X1205=m
++CONFIG_RTC_DRV_DS1307=m
++CONFIG_RTC_DRV_DS1553=m
++CONFIG_RTC_DRV_ISL1208=m
++CONFIG_RTC_DRV_DS1672=m
++CONFIG_RTC_DRV_DS1742=m
++CONFIG_RTC_DRV_PCF8563=m
++CONFIG_RTC_DRV_PCF8583=m
++CONFIG_RTC_DRV_RS5C348=m
++CONFIG_RTC_DRV_RS5C372=m
++CONFIG_RTC_DRV_M48T86=m
++CONFIG_RTC_DRV_TEST=m
++CONFIG_RTC_DRV_MAX6902=m
++CONFIG_RTC_DRV_V3020=m
+
-+ /* rsi is pointer to startup info structure.
-+ pass it to C */
-+ movq %rsi,%rdi
-+ pushq $0 # fake return address
-+ jmp x86_64_start_kernel
++#
++# DMA Engine support
++#
++CONFIG_DMA_ENGINE=y
+
-+ENTRY(stext)
-+ENTRY(_stext)
++#
++# DMA Clients
++#
++CONFIG_NET_DMA=y
+
-+ $page = 0
-+#define NEXT_PAGE(name) \
-+ $page = $page + 1; \
-+ .org $page * 0x1000; \
-+ phys_##name = $page * 0x1000 + __PHYSICAL_START; \
-+ENTRY(name)
++#
++# DMA Devices
++#
++CONFIG_INTEL_IOATDMA=m
+
-+NEXT_PAGE(init_level4_pgt)
-+ /* This gets initialized in x86_64_start_kernel */
-+ .fill 512,8,0
++#
++# File systems
++#
++CONFIG_EXT2_FS=y
++CONFIG_EXT2_FS_XATTR=y
++CONFIG_EXT2_FS_POSIX_ACL=y
++CONFIG_EXT2_FS_SECURITY=y
++CONFIG_EXT2_FS_XIP=y
++CONFIG_FS_XIP=y
++CONFIG_EXT3_FS=m
++CONFIG_EXT3_FS_XATTR=y
++CONFIG_EXT3_FS_POSIX_ACL=y
++CONFIG_EXT3_FS_SECURITY=y
++CONFIG_JBD=m
++# CONFIG_JBD_DEBUG is not set
++CONFIG_FS_MBCACHE=y
++CONFIG_REISERFS_FS=m
++# CONFIG_REISERFS_CHECK is not set
++# CONFIG_REISERFS_PROC_INFO is not set
++CONFIG_REISERFS_FS_XATTR=y
++CONFIG_REISERFS_FS_POSIX_ACL=y
++CONFIG_REISERFS_FS_SECURITY=y
++CONFIG_JFS_FS=m
++CONFIG_JFS_POSIX_ACL=y
++# CONFIG_JFS_SECURITY is not set
++# CONFIG_JFS_DEBUG is not set
++CONFIG_JFS_STATISTICS=y
++CONFIG_FS_POSIX_ACL=y
++CONFIG_XFS_FS=m
++CONFIG_XFS_QUOTA=y
++CONFIG_XFS_SECURITY=y
++CONFIG_XFS_POSIX_ACL=y
++CONFIG_XFS_RT=y
++CONFIG_OCFS2_FS=m
++CONFIG_OCFS2_DEBUG_MASKLOG=y
++CONFIG_MINIX_FS=m
++CONFIG_ROMFS_FS=m
++CONFIG_INOTIFY=y
++CONFIG_INOTIFY_USER=y
++CONFIG_QUOTA=y
++CONFIG_QFMT_V1=m
++CONFIG_QFMT_V2=m
++CONFIG_QUOTACTL=y
++CONFIG_DNOTIFY=y
++CONFIG_AUTOFS_FS=m
++CONFIG_AUTOFS4_FS=m
++CONFIG_FUSE_FS=m
+
-+ /*
-+ * We update two pgd entries to make kernel and user pgd consistent
-+ * at pgd_populate(). It can be used for kernel modules. So we place
-+ * this page here for those cases to avoid memory corruption.
-+ * We also use this page to establish the initiali mapping for
-+ * vsyscall area.
-+ */
-+NEXT_PAGE(init_level4_user_pgt)
-+ .fill 512,8,0
++#
++# CD-ROM/DVD Filesystems
++#
++CONFIG_ISO9660_FS=m
++CONFIG_JOLIET=y
++CONFIG_ZISOFS=y
++CONFIG_ZISOFS_FS=m
++CONFIG_UDF_FS=m
++CONFIG_UDF_NLS=y
+
-+NEXT_PAGE(level3_kernel_pgt)
-+ .fill 512,8,0
++#
++# DOS/FAT/NT Filesystems
++#
++CONFIG_FAT_FS=m
++CONFIG_MSDOS_FS=m
++CONFIG_VFAT_FS=m
++CONFIG_FAT_DEFAULT_CODEPAGE=437
++CONFIG_FAT_DEFAULT_IOCHARSET="iso8859-1"
++CONFIG_NTFS_FS=m
++# CONFIG_NTFS_DEBUG is not set
++# CONFIG_NTFS_RW is not set
+
-+ /*
-+ * This is used for vsyscall area mapping as we have a different
-+ * level4 page table for user.
-+ */
-+NEXT_PAGE(level3_user_pgt)
-+ .fill 512,8,0
++#
++# Pseudo filesystems
++#
++CONFIG_PROC_FS=y
++CONFIG_PROC_KCORE=y
++CONFIG_SYSFS=y
++CONFIG_TMPFS=y
++# CONFIG_HUGETLB_PAGE is not set
++CONFIG_RAMFS=y
++CONFIG_CONFIGFS_FS=m
+
-+NEXT_PAGE(level2_kernel_pgt)
-+ .fill 512,8,0
++#
++# Miscellaneous filesystems
++#
++CONFIG_ADFS_FS=m
++# CONFIG_ADFS_FS_RW is not set
++CONFIG_AFFS_FS=m
++CONFIG_HFS_FS=m
++CONFIG_HFSPLUS_FS=m
++CONFIG_BEFS_FS=m
++# CONFIG_BEFS_DEBUG is not set
++CONFIG_BFS_FS=m
++CONFIG_EFS_FS=m
++CONFIG_JFFS_FS=m
++CONFIG_JFFS_FS_VERBOSE=0
++CONFIG_JFFS_PROC_FS=y
++CONFIG_JFFS2_FS=m
++CONFIG_JFFS2_FS_DEBUG=0
++CONFIG_JFFS2_FS_WRITEBUFFER=y
++# CONFIG_JFFS2_SUMMARY is not set
++# CONFIG_JFFS2_FS_XATTR is not set
++# CONFIG_JFFS2_COMPRESSION_OPTIONS is not set
++CONFIG_JFFS2_ZLIB=y
++CONFIG_JFFS2_RTIME=y
++# CONFIG_JFFS2_RUBIN is not set
++CONFIG_CRAMFS=m
++CONFIG_VXFS_FS=m
++CONFIG_HPFS_FS=m
++CONFIG_QNX4FS_FS=m
++CONFIG_SYSV_FS=m
++CONFIG_UFS_FS=m
++# CONFIG_UFS_FS_WRITE is not set
++# CONFIG_UFS_DEBUG is not set
+
-+NEXT_PAGE(hypercall_page)
-+ CFI_STARTPROC
-+ .rept 0x1000 / 0x20
-+ .skip 1 /* push %rcx */
-+ CFI_ADJUST_CFA_OFFSET 8
-+ CFI_REL_OFFSET rcx,0
-+ .skip 2 /* push %r11 */
-+ CFI_ADJUST_CFA_OFFSET 8
-+ CFI_REL_OFFSET rcx,0
-+ .skip 5 /* mov $#,%eax */
-+ .skip 2 /* syscall */
-+ .skip 2 /* pop %r11 */
-+ CFI_ADJUST_CFA_OFFSET -8
-+ CFI_RESTORE r11
-+ .skip 1 /* pop %rcx */
-+ CFI_ADJUST_CFA_OFFSET -8
-+ CFI_RESTORE rcx
-+ .align 0x20,0 /* ret */
-+ .endr
-+ CFI_ENDPROC
++#
++# Network File Systems
++#
++CONFIG_NFS_FS=m
++CONFIG_NFS_V3=y
++CONFIG_NFS_V3_ACL=y
++CONFIG_NFS_V4=y
++CONFIG_NFS_DIRECTIO=y
++CONFIG_NFSD=m
++CONFIG_NFSD_V2_ACL=y
++CONFIG_NFSD_V3=y
++CONFIG_NFSD_V3_ACL=y
++CONFIG_NFSD_V4=y
++CONFIG_NFSD_TCP=y
++CONFIG_LOCKD=m
++CONFIG_LOCKD_V4=y
++CONFIG_EXPORTFS=m
++CONFIG_NFS_ACL_SUPPORT=m
++CONFIG_NFS_COMMON=y
++CONFIG_SUNRPC=m
++CONFIG_SUNRPC_GSS=m
++CONFIG_RPCSEC_GSS_KRB5=m
++CONFIG_RPCSEC_GSS_SPKM3=m
++CONFIG_SMB_FS=m
++CONFIG_SMB_NLS_DEFAULT=y
++CONFIG_SMB_NLS_REMOTE="cp850"
++CONFIG_CIFS=m
++CONFIG_CIFS_STATS=y
++CONFIG_CIFS_STATS2=y
++# CONFIG_CIFS_WEAK_PW_HASH is not set
++CONFIG_CIFS_XATTR=y
++# CONFIG_CIFS_POSIX is not set
++# CONFIG_CIFS_DEBUG2 is not set
++# CONFIG_CIFS_EXPERIMENTAL is not set
++CONFIG_NCP_FS=m
++CONFIG_NCPFS_PACKET_SIGNING=y
++CONFIG_NCPFS_IOCTL_LOCKING=y
++CONFIG_NCPFS_STRONG=y
++CONFIG_NCPFS_NFS_NS=y
++CONFIG_NCPFS_OS2_NS=y
++# CONFIG_NCPFS_SMALLDOS is not set
++CONFIG_NCPFS_NLS=y
++CONFIG_NCPFS_EXTRAS=y
++CONFIG_CODA_FS=m
++# CONFIG_CODA_FS_OLD_API is not set
++CONFIG_AFS_FS=m
++CONFIG_RXRPC=m
++CONFIG_9P_FS=m
+
-+#undef NEXT_PAGE
++#
++# Partition Types
++#
++CONFIG_PARTITION_ADVANCED=y
++# CONFIG_ACORN_PARTITION is not set
++CONFIG_OSF_PARTITION=y
++# CONFIG_AMIGA_PARTITION is not set
++CONFIG_ATARI_PARTITION=y
++CONFIG_MAC_PARTITION=y
++CONFIG_MSDOS_PARTITION=y
++CONFIG_BSD_DISKLABEL=y
++# CONFIG_MINIX_SUBPARTITION is not set
++CONFIG_SOLARIS_X86_PARTITION=y
++CONFIG_UNIXWARE_DISKLABEL=y
++CONFIG_LDM_PARTITION=y
++# CONFIG_LDM_DEBUG is not set
++CONFIG_SGI_PARTITION=y
++CONFIG_ULTRIX_PARTITION=y
++CONFIG_SUN_PARTITION=y
++CONFIG_KARMA_PARTITION=y
++CONFIG_EFI_PARTITION=y
+
-+ .data
++#
++# Native Language Support
++#
++CONFIG_NLS=y
++CONFIG_NLS_DEFAULT="utf8"
++CONFIG_NLS_CODEPAGE_437=m
++CONFIG_NLS_CODEPAGE_737=m
++CONFIG_NLS_CODEPAGE_775=m
++CONFIG_NLS_CODEPAGE_850=m
++CONFIG_NLS_CODEPAGE_852=m
++CONFIG_NLS_CODEPAGE_855=m
++CONFIG_NLS_CODEPAGE_857=m
++CONFIG_NLS_CODEPAGE_860=m
++CONFIG_NLS_CODEPAGE_861=m
++CONFIG_NLS_CODEPAGE_862=m
++CONFIG_NLS_CODEPAGE_863=m
++CONFIG_NLS_CODEPAGE_864=m
++CONFIG_NLS_CODEPAGE_865=m
++CONFIG_NLS_CODEPAGE_866=m
++CONFIG_NLS_CODEPAGE_869=m
++CONFIG_NLS_CODEPAGE_936=m
++CONFIG_NLS_CODEPAGE_950=m
++CONFIG_NLS_CODEPAGE_932=m
++CONFIG_NLS_CODEPAGE_949=m
++CONFIG_NLS_CODEPAGE_874=m
++CONFIG_NLS_ISO8859_8=m
++CONFIG_NLS_CODEPAGE_1250=m
++CONFIG_NLS_CODEPAGE_1251=m
++CONFIG_NLS_ASCII=m
++CONFIG_NLS_ISO8859_1=m
++CONFIG_NLS_ISO8859_2=m
++CONFIG_NLS_ISO8859_3=m
++CONFIG_NLS_ISO8859_4=m
++CONFIG_NLS_ISO8859_5=m
++CONFIG_NLS_ISO8859_6=m
++CONFIG_NLS_ISO8859_7=m
++CONFIG_NLS_ISO8859_9=m
++CONFIG_NLS_ISO8859_13=m
++CONFIG_NLS_ISO8859_14=m
++CONFIG_NLS_ISO8859_15=m
++CONFIG_NLS_KOI8_R=m
++CONFIG_NLS_KOI8_U=m
++CONFIG_NLS_UTF8=m
+
-+ .align 16
-+ .globl cpu_gdt_descr
-+cpu_gdt_descr:
-+ .word gdt_end-cpu_gdt_table-1
-+gdt:
-+ .quad cpu_gdt_table
-+#ifdef CONFIG_SMP
-+ .rept NR_CPUS-1
-+ .word 0
-+ .quad 0
-+ .endr
-+#endif
++#
++# Instrumentation Support
++#
++# CONFIG_PROFILING is not set
++# CONFIG_KPROBES is not set
+
-+/* We need valid kernel segments for data and code in long mode too
-+ * IRET will check the segment types kkeil 2000/10/28
-+ * Also sysret mandates a special GDT layout
-+ */
-+
-+ .section .data.page_aligned, "aw"
-+ .align PAGE_SIZE
++#
++# Kernel hacking
++#
++CONFIG_TRACE_IRQFLAGS_SUPPORT=y
++# CONFIG_PRINTK_TIME is not set
++CONFIG_MAGIC_SYSRQ=y
++CONFIG_UNUSED_SYMBOLS=y
++CONFIG_DEBUG_KERNEL=y
++CONFIG_LOG_BUF_SHIFT=14
++CONFIG_DETECT_SOFTLOCKUP=y
++# CONFIG_SCHEDSTATS is not set
++# CONFIG_DEBUG_SLAB is not set
++# CONFIG_DEBUG_RT_MUTEXES is not set
++# CONFIG_RT_MUTEX_TESTER is not set
++# CONFIG_DEBUG_SPINLOCK is not set
++# CONFIG_DEBUG_MUTEXES is not set
++# CONFIG_DEBUG_RWSEMS is not set
++# CONFIG_DEBUG_LOCK_ALLOC is not set
++# CONFIG_PROVE_LOCKING is not set
++# CONFIG_DEBUG_SPINLOCK_SLEEP is not set
++# CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set
++# CONFIG_DEBUG_KOBJECT is not set
++# CONFIG_DEBUG_HIGHMEM is not set
++CONFIG_DEBUG_BUGVERBOSE=y
++# CONFIG_DEBUG_INFO is not set
++# CONFIG_DEBUG_FS is not set
++# CONFIG_DEBUG_VM is not set
++# CONFIG_FRAME_POINTER is not set
++# CONFIG_UNWIND_INFO is not set
++CONFIG_FORCED_INLINING=y
++# CONFIG_RCU_TORTURE_TEST is not set
++CONFIG_EARLY_PRINTK=y
++CONFIG_DEBUG_STACKOVERFLOW=y
++# CONFIG_DEBUG_STACK_USAGE is not set
++# CONFIG_DEBUG_PAGEALLOC is not set
++# CONFIG_DEBUG_RODATA is not set
++# CONFIG_4KSTACKS is not set
++CONFIG_X86_FIND_SMP_CONFIG=y
++CONFIG_X86_MPPARSE=y
+
-+/* The TLS descriptors are currently at a different place compared to i386.
-+ Hopefully nobody expects them at a fixed place (Wine?) */
++#
++# Security options
++#
++CONFIG_KEYS=y
++# CONFIG_KEYS_DEBUG_PROC_KEYS is not set
++CONFIG_SECURITY=y
++CONFIG_SECURITY_NETWORK=y
++# CONFIG_SECURITY_NETWORK_XFRM is not set
++CONFIG_SECURITY_CAPABILITIES=y
++CONFIG_SECURITY_ROOTPLUG=m
++CONFIG_SECURITY_SECLVL=m
++# CONFIG_SECURITY_SELINUX is not set
+
-+ENTRY(cpu_gdt_table)
-+ .quad 0x0000000000000000 /* NULL descriptor */
-+ .quad 0x0 /* unused */
-+ .quad 0x00af9a000000ffff /* __KERNEL_CS */
-+ .quad 0x00cf92000000ffff /* __KERNEL_DS */
-+ .quad 0x00cffa000000ffff /* __USER32_CS */
-+ .quad 0x00cff2000000ffff /* __USER_DS, __USER32_DS */
-+ .quad 0x00affa000000ffff /* __USER_CS */
-+ .quad 0x00cf9a000000ffff /* __KERNEL32_CS */
-+ .quad 0,0 /* TSS */
-+ .quad 0,0 /* LDT */
-+ .quad 0,0,0 /* three TLS descriptors */
-+ .quad 0 /* unused */
-+gdt_end:
-+ /* asm/segment.h:GDT_ENTRIES must match this */
-+ /* This should be a multiple of the cache line size */
-+ /* GDTs of other CPUs are now dynamically allocated */
++#
++# Cryptographic options
++#
++CONFIG_CRYPTO=y
++CONFIG_CRYPTO_HMAC=y
++CONFIG_CRYPTO_NULL=m
++CONFIG_CRYPTO_MD4=m
++CONFIG_CRYPTO_MD5=y
++CONFIG_CRYPTO_SHA1=m
++CONFIG_CRYPTO_SHA256=m
++CONFIG_CRYPTO_SHA512=m
++CONFIG_CRYPTO_WP512=m
++CONFIG_CRYPTO_TGR192=m
++CONFIG_CRYPTO_DES=m
++CONFIG_CRYPTO_BLOWFISH=m
++CONFIG_CRYPTO_TWOFISH=m
++CONFIG_CRYPTO_SERPENT=m
++CONFIG_CRYPTO_AES=m
++CONFIG_CRYPTO_AES_586=m
++CONFIG_CRYPTO_CAST5=m
++CONFIG_CRYPTO_CAST6=m
++CONFIG_CRYPTO_TEA=m
++CONFIG_CRYPTO_ARC4=m
++CONFIG_CRYPTO_KHAZAD=m
++CONFIG_CRYPTO_ANUBIS=m
++CONFIG_CRYPTO_DEFLATE=m
++CONFIG_CRYPTO_MICHAEL_MIC=m
++CONFIG_CRYPTO_CRC32C=m
++CONFIG_CRYPTO_TEST=m
+
-+ /* zero the remaining page */
-+ .fill PAGE_SIZE / 8 - GDT_ENTRIES,8,0
++#
++# Hardware crypto devices
++#
++# CONFIG_CRYPTO_DEV_PADLOCK is not set
++CONFIG_XEN=y
++CONFIG_XEN_INTERFACE_VERSION=0x00030207
+
-+ .section .bss.page_aligned, "aw", @nobits
-+ .align PAGE_SIZE
-+ENTRY(empty_zero_page)
-+ .skip PAGE_SIZE
++#
++# XEN
++#
++CONFIG_XEN_PRIVILEGED_GUEST=y
++# CONFIG_XEN_UNPRIVILEGED_GUEST is not set
++CONFIG_XEN_PRIVCMD=y
++CONFIG_XEN_XENBUS_DEV=y
++CONFIG_XEN_BACKEND=y
++CONFIG_XEN_BLKDEV_BACKEND=y
++CONFIG_XEN_BLKDEV_TAP=y
++CONFIG_XEN_NETDEV_BACKEND=y
++# CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_UTIL is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_BACKEND is not set
++# CONFIG_XEN_NETDEV_LOOPBACK is not set
++CONFIG_XEN_PCIDEV_BACKEND=m
++CONFIG_XEN_PCIDEV_BACKEND_VPCI=y
++# CONFIG_XEN_PCIDEV_BACKEND_PASS is not set
++# CONFIG_XEN_PCIDEV_BACKEND_SLOT is not set
++# CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER is not set
++# CONFIG_XEN_PCIDEV_BE_DEBUG is not set
++CONFIG_XEN_TPMDEV_BACKEND=m
++CONFIG_XEN_SCSI_BACKEND=m
++CONFIG_XEN_BLKDEV_FRONTEND=y
++CONFIG_XEN_NETDEV_FRONTEND=y
++CONFIG_XEN_SCSI_FRONTEND=m
++CONFIG_XEN_GRANT_DEV=y
++# CONFIG_XEN_NETDEV_ACCEL_SFC_FRONTEND is not set
++CONFIG_XEN_FRAMEBUFFER=y
++CONFIG_XEN_KEYBOARD=y
++CONFIG_XEN_SCRUB_PAGES=y
++CONFIG_XEN_DISABLE_SERIAL=y
++CONFIG_XEN_SYSFS=y
++CONFIG_XEN_COMPAT_030002_AND_LATER=y
++# CONFIG_XEN_COMPAT_030004_AND_LATER is not set
++# CONFIG_XEN_COMPAT_030100_AND_LATER is not set
++# CONFIG_XEN_COMPAT_LATEST_ONLY is not set
++CONFIG_XEN_COMPAT=0x030002
++CONFIG_HAVE_IRQ_IGNORE_UNHANDLED=y
++CONFIG_NO_IDLE_HZ=y
++CONFIG_XEN_SMPBOOT=y
++CONFIG_XEN_BALLOON=y
++CONFIG_XEN_DEVMEM=y
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+/*
-+ * __xen_guest information
-+ */
-+.macro utoh value
-+ .if (\value) < 0 || (\value) >= 0x10
-+ utoh (((\value)>>4)&0x0fffffffffffffff)
-+ .endif
-+ .if ((\value) & 0xf) < 10
-+ .byte '0' + ((\value) & 0xf)
-+ .else
-+ .byte 'A' + ((\value) & 0xf) - 10
-+ .endif
-+.endm
++#
++# Library routines
++#
++CONFIG_CRC_CCITT=m
++CONFIG_CRC16=m
++CONFIG_CRC32=y
++CONFIG_LIBCRC32C=m
++CONFIG_AUDIT_GENERIC=y
++CONFIG_ZLIB_INFLATE=m
++CONFIG_ZLIB_DEFLATE=m
++CONFIG_REED_SOLOMON=m
++CONFIG_REED_SOLOMON_DEC16=y
++CONFIG_TEXTSEARCH=y
++CONFIG_TEXTSEARCH_KMP=m
++CONFIG_TEXTSEARCH_BM=m
++CONFIG_TEXTSEARCH_FSM=m
++CONFIG_PLIST=y
++CONFIG_GENERIC_HARDIRQS=y
++CONFIG_GENERIC_IRQ_PROBE=y
++CONFIG_GENERIC_PENDING_IRQ=y
++CONFIG_X86_SMP=y
++CONFIG_X86_BIOS_REBOOT=y
++CONFIG_X86_TRAMPOLINE=y
++CONFIG_X86_NO_TSS=y
++CONFIG_X86_NO_IDT=y
++CONFIG_KTIME_SCALAR=y
+diff -r d894e36cfc30 -r 0aa021803deb buildconfigs/linux-defconfig_xen_x86_64
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/buildconfigs/linux-defconfig_xen_x86_64 Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,3146 @@
++#
++# Automatically generated make config: don't edit
++# Linux kernel version: 2.6.18.8
++# Mon Feb 18 10:38:24 2008
++#
++CONFIG_X86_64=y
++CONFIG_64BIT=y
++CONFIG_X86=y
++CONFIG_LOCKDEP_SUPPORT=y
++CONFIG_STACKTRACE_SUPPORT=y
++CONFIG_SEMAPHORE_SLEEPERS=y
++CONFIG_MMU=y
++CONFIG_RWSEM_GENERIC_SPINLOCK=y
++CONFIG_GENERIC_HWEIGHT=y
++CONFIG_GENERIC_CALIBRATE_DELAY=y
++CONFIG_X86_CMPXCHG=y
++CONFIG_EARLY_PRINTK=y
++CONFIG_GENERIC_ISA_DMA=y
++CONFIG_GENERIC_IOMAP=y
++CONFIG_ARCH_MAY_HAVE_PC_FDC=y
++CONFIG_DMI=y
++CONFIG_AUDIT_ARCH=y
++CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config"
+
-+.section __xen_guest
-+ .ascii "GUEST_OS=linux,GUEST_VER=2.6"
-+ .ascii ",XEN_VER=xen-3.0"
-+ .ascii ",VIRT_BASE=0x"
-+ utoh __START_KERNEL_map
-+ .ascii ",ELF_PADDR_OFFSET=0x"
-+ utoh __START_KERNEL_map
-+ .ascii ",VIRT_ENTRY=0x"
-+ utoh (__START_KERNEL_map + __PHYSICAL_START + VIRT_ENTRY_OFFSET)
-+ .ascii ",HYPERCALL_PAGE=0x"
-+ utoh (phys_hypercall_page >> PAGE_SHIFT)
-+ .ascii ",FEATURES=writable_page_tables"
-+ .ascii "|writable_descriptor_tables"
-+ .ascii "|auto_translated_physmap"
-+ .ascii "|supervisor_mode_kernel"
-+ .ascii ",LOADER=generic"
-+ .byte 0
-+#endif /* CONFIG_XEN_COMPAT <= 0x030002 */
-+
-+ ELFNOTE(Xen, XEN_ELFNOTE_GUEST_OS, .asciz, "linux")
-+ ELFNOTE(Xen, XEN_ELFNOTE_GUEST_VERSION, .asciz, "2.6")
-+ ELFNOTE(Xen, XEN_ELFNOTE_XEN_VERSION, .asciz, "xen-3.0")
-+ ELFNOTE(Xen, XEN_ELFNOTE_VIRT_BASE, .quad, __START_KERNEL_map)
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET, .quad, __START_KERNEL_map)
-+#else
-+ ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET, .quad, 0)
-+#endif
-+ ELFNOTE(Xen, XEN_ELFNOTE_ENTRY, .quad, startup_64)
-+ ELFNOTE(Xen, XEN_ELFNOTE_HYPERCALL_PAGE, .quad, hypercall_page)
-+ ELFNOTE(Xen, XEN_ELFNOTE_L1_MFN_VALID, .quad, _PAGE_PRESENT,_PAGE_PRESENT)
-+ ELFNOTE(Xen, XEN_ELFNOTE_FEATURES, .asciz, "writable_page_tables|writable_descriptor_tables|auto_translated_physmap|pae_pgdir_above_4gb|supervisor_mode_kernel")
-+ ELFNOTE(Xen, XEN_ELFNOTE_LOADER, .asciz, "generic")
-+ ELFNOTE(Xen, XEN_ELFNOTE_SUSPEND_CANCEL, .long, 1)
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/init_task.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/init_task.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/init_task.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/init_task.c 2007-07-30 16:35:11.000000000 +0200
-@@ -37,6 +37,8 @@ union thread_union init_thread_union
- struct task_struct init_task = INIT_TASK(init_task);
-
- EXPORT_SYMBOL(init_task);
++#
++# Code maturity level options
++#
++CONFIG_EXPERIMENTAL=y
++CONFIG_LOCK_KERNEL=y
++CONFIG_INIT_ENV_ARG_LIMIT=32
+
-+#ifndef CONFIG_X86_NO_TSS
- /*
- * per-CPU TSS segments. Threads are completely 'soft' on Linux,
- * no more per-task TSS's. The TSS size is kept cacheline-aligned
-@@ -50,5 +52,6 @@ DEFINE_PER_CPU(struct tss_struct, init_t
- * debugging, no special alignment required.
- */
- DEFINE_PER_CPU(struct orig_ist, orig_ist);
-+#endif
-
- #define ALIGN_TO_4K __attribute__((section(".data.init_task")))
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/io_apic-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/io_apic-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/io_apic-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/io_apic-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,2269 @@
-+/*
-+ * Intel IO-APIC support for multi-Pentium hosts.
-+ *
-+ * Copyright (C) 1997, 1998, 1999, 2000 Ingo Molnar, Hajnalka Szabo
-+ *
-+ * Many thanks to Stig Venaas for trying out countless experimental
-+ * patches and reporting/debugging problems patiently!
-+ *
-+ * (c) 1999, Multiple IO-APIC support, developed by
-+ * Ken-ichi Yaku <yaku at css1.kbnes.nec.co.jp> and
-+ * Hidemi Kishimoto <kisimoto at css1.kbnes.nec.co.jp>,
-+ * further tested and cleaned up by Zach Brown <zab at redhat.com>
-+ * and Ingo Molnar <mingo at redhat.com>
-+ *
-+ * Fixes
-+ * Maciej W. Rozycki : Bits for genuine 82489DX APICs;
-+ * thanks to Eric Gilmore
-+ * and Rolf G. Tews
-+ * for testing these extensively
-+ * Paul Diefenbaugh : Added full ACPI support
-+ */
++#
++# General setup
++#
++CONFIG_LOCALVERSION=""
++# CONFIG_LOCALVERSION_AUTO is not set
++CONFIG_SWAP=y
++CONFIG_SYSVIPC=y
++CONFIG_POSIX_MQUEUE=y
++CONFIG_BSD_PROCESS_ACCT=y
++CONFIG_BSD_PROCESS_ACCT_V3=y
++CONFIG_TASKSTATS=y
++CONFIG_TASK_DELAY_ACCT=y
++CONFIG_AUDIT=y
++CONFIG_AUDITSYSCALL=y
++CONFIG_IKCONFIG=y
++CONFIG_IKCONFIG_PROC=y
++CONFIG_CPUSETS=y
++# CONFIG_RELAY is not set
++CONFIG_INITRAMFS_SOURCE=""
++CONFIG_CC_OPTIMIZE_FOR_SIZE=y
++# CONFIG_EMBEDDED is not set
++CONFIG_UID16=y
++CONFIG_SYSCTL=y
++CONFIG_KALLSYMS=y
++# CONFIG_KALLSYMS_ALL is not set
++CONFIG_KALLSYMS_EXTRA_PASS=y
++CONFIG_HOTPLUG=y
++CONFIG_PRINTK=y
++CONFIG_BUG=y
++CONFIG_ELF_CORE=y
++CONFIG_BASE_FULL=y
++CONFIG_FUTEX=y
++CONFIG_EPOLL=y
++CONFIG_SHMEM=y
++CONFIG_SLAB=y
++CONFIG_VM_EVENT_COUNTERS=y
++CONFIG_RT_MUTEXES=y
++# CONFIG_TINY_SHMEM is not set
++CONFIG_BASE_SMALL=0
++# CONFIG_SLOB is not set
+
-+#include <linux/mm.h>
-+#include <linux/interrupt.h>
-+#include <linux/init.h>
-+#include <linux/delay.h>
-+#include <linux/sched.h>
-+#include <linux/smp_lock.h>
-+#include <linux/mc146818rtc.h>
-+#include <linux/acpi.h>
-+#include <linux/sysdev.h>
-+#ifdef CONFIG_ACPI
-+#include <acpi/acpi_bus.h>
-+#endif
++#
++# Loadable module support
++#
++CONFIG_MODULES=y
++CONFIG_MODULE_UNLOAD=y
++CONFIG_MODULE_FORCE_UNLOAD=y
++CONFIG_MODVERSIONS=y
++CONFIG_MODULE_SRCVERSION_ALL=y
++CONFIG_KMOD=y
++CONFIG_STOP_MACHINE=y
+
-+#include <asm/io.h>
-+#include <asm/smp.h>
-+#include <asm/desc.h>
-+#include <asm/proto.h>
-+#include <asm/mach_apic.h>
-+#include <asm/acpi.h>
-+#include <asm/dma.h>
-+#include <asm/nmi.h>
++#
++# Block layer
++#
++CONFIG_LBD=y
++# CONFIG_BLK_DEV_IO_TRACE is not set
++CONFIG_LSF=y
+
-+#define __apicdebuginit __init
++#
++# IO Schedulers
++#
++CONFIG_IOSCHED_NOOP=y
++CONFIG_IOSCHED_AS=y
++CONFIG_IOSCHED_DEADLINE=y
++CONFIG_IOSCHED_CFQ=y
++# CONFIG_DEFAULT_AS is not set
++# CONFIG_DEFAULT_DEADLINE is not set
++CONFIG_DEFAULT_CFQ=y
++# CONFIG_DEFAULT_NOOP is not set
++CONFIG_DEFAULT_IOSCHED="cfq"
+
-+int sis_apic_bug; /* not actually supported, dummy for compile */
++#
++# Processor type and features
++#
++CONFIG_X86_PC=y
++# CONFIG_X86_VSMP is not set
++# CONFIG_MK8 is not set
++# CONFIG_MPSC is not set
++CONFIG_GENERIC_CPU=y
++CONFIG_X86_64_XEN=y
++CONFIG_X86_NO_TSS=y
++CONFIG_X86_NO_IDT=y
++CONFIG_X86_L1_CACHE_BYTES=128
++CONFIG_X86_L1_CACHE_SHIFT=7
++CONFIG_X86_INTERNODE_CACHE_BYTES=128
++CONFIG_X86_GOOD_APIC=y
++CONFIG_MICROCODE=y
++CONFIG_X86_MSR=m
++CONFIG_X86_CPUID=m
++CONFIG_X86_IO_APIC=y
++CONFIG_X86_XEN_GENAPIC=y
++CONFIG_X86_LOCAL_APIC=y
++CONFIG_MTRR=y
++CONFIG_SMP=y
++# CONFIG_PREEMPT_NONE is not set
++CONFIG_PREEMPT_VOLUNTARY=y
++# CONFIG_PREEMPT is not set
++CONFIG_PREEMPT_BKL=y
++CONFIG_ARCH_FLATMEM_ENABLE=y
++CONFIG_SELECT_MEMORY_MODEL=y
++CONFIG_FLATMEM_MANUAL=y
++# CONFIG_DISCONTIGMEM_MANUAL is not set
++# CONFIG_SPARSEMEM_MANUAL is not set
++CONFIG_FLATMEM=y
++CONFIG_FLAT_NODE_MEM_MAP=y
++# CONFIG_SPARSEMEM_STATIC is not set
++CONFIG_SPLIT_PTLOCK_CPUS=4
++CONFIG_RESOURCES_64BIT=y
++CONFIG_NR_CPUS=32
++CONFIG_HOTPLUG_CPU=y
++CONFIG_ARCH_ENABLE_MEMORY_HOTPLUG=y
++CONFIG_SWIOTLB=y
++CONFIG_KEXEC=y
++# CONFIG_CRASH_DUMP is not set
++CONFIG_PHYSICAL_START=0x200000
++CONFIG_SECCOMP=y
++CONFIG_HZ_100=y
++# CONFIG_HZ_250 is not set
++# CONFIG_HZ_1000 is not set
++CONFIG_HZ=100
++# CONFIG_REORDER is not set
++CONFIG_K8_NB=y
++CONFIG_GENERIC_HARDIRQS=y
++CONFIG_GENERIC_IRQ_PROBE=y
++CONFIG_ISA_DMA_API=y
++CONFIG_GENERIC_PENDING_IRQ=y
+
-+static int no_timer_check;
++#
++# Power management options
++#
++CONFIG_PM=y
++# CONFIG_PM_LEGACY is not set
++CONFIG_PM_DEBUG=y
++# CONFIG_SOFTWARE_SUSPEND is not set
++CONFIG_SUSPEND_SMP=y
+
-+int disable_timer_pin_1 __initdata;
++#
++# ACPI (Advanced Configuration and Power Interface) Support
++#
++CONFIG_ACPI=y
++CONFIG_ACPI_SLEEP=y
++CONFIG_ACPI_SLEEP_PROC_FS=y
++# CONFIG_ACPI_SLEEP_PROC_SLEEP is not set
++CONFIG_ACPI_AC=m
++CONFIG_ACPI_BATTERY=m
++CONFIG_ACPI_BUTTON=m
++CONFIG_ACPI_VIDEO=m
++CONFIG_ACPI_HOTKEY=m
++CONFIG_ACPI_FAN=m
++CONFIG_ACPI_DOCK=m
++CONFIG_ACPI_PROCESSOR=m
++CONFIG_ACPI_HOTPLUG_CPU=y
++CONFIG_ACPI_THERMAL=m
++CONFIG_ACPI_ASUS=m
++CONFIG_ACPI_IBM=m
++CONFIG_ACPI_TOSHIBA=m
++CONFIG_ACPI_BLACKLIST_YEAR=0
++# CONFIG_ACPI_DEBUG is not set
++CONFIG_ACPI_EC=y
++CONFIG_ACPI_POWER=y
++CONFIG_ACPI_SYSTEM=y
++CONFIG_ACPI_CONTAINER=m
++CONFIG_ACPI_SBS=m
++CONFIG_ACPI_PV_SLEEP=y
+
-+#ifndef CONFIG_XEN
-+int timer_over_8254 __initdata = 0;
++#
++# CPU Frequency scaling
++#
++# CONFIG_CPU_FREQ is not set
+
-+/* Where if anywhere is the i8259 connect in external int mode */
-+static struct { int pin, apic; } ioapic_i8259 = { -1, -1 };
-+#endif
++#
++# Bus options (PCI etc.)
++#
++CONFIG_PCI=y
++CONFIG_PCI_DIRECT=y
++# CONFIG_PCI_MMCONFIG is not set
++CONFIG_XEN_PCIDEV_FRONTEND=y
++# CONFIG_XEN_PCIDEV_FE_DEBUG is not set
++# CONFIG_PCIEPORTBUS is not set
++CONFIG_PCI_MSI=y
++# CONFIG_PCI_DEBUG is not set
+
-+static DEFINE_SPINLOCK(ioapic_lock);
-+static DEFINE_SPINLOCK(vector_lock);
++#
++# PCCARD (PCMCIA/CardBus) support
++#
++CONFIG_PCCARD=m
++# CONFIG_PCMCIA_DEBUG is not set
++CONFIG_PCMCIA=m
++CONFIG_PCMCIA_LOAD_CIS=y
++CONFIG_PCMCIA_IOCTL=y
++CONFIG_CARDBUS=y
+
-+/*
-+ * # of IRQ routing registers
-+ */
-+int nr_ioapic_registers[MAX_IO_APICS];
++#
++# PC-card bridges
++#
++CONFIG_YENTA=m
++CONFIG_YENTA_O2=y
++CONFIG_YENTA_RICOH=y
++CONFIG_YENTA_TI=y
++CONFIG_YENTA_ENE_TUNE=y
++CONFIG_YENTA_TOSHIBA=y
++CONFIG_PD6729=m
++CONFIG_I82092=m
++CONFIG_PCCARD_NONSTATIC=m
+
-+/*
-+ * Rough estimation of how many shared IRQs there are, can
-+ * be changed anytime.
-+ */
-+#define MAX_PLUS_SHARED_IRQS NR_IRQ_VECTORS
-+#define PIN_MAP_SIZE (MAX_PLUS_SHARED_IRQS + NR_IRQS)
++#
++# PCI Hotplug Support
++#
++CONFIG_HOTPLUG_PCI=m
++# CONFIG_HOTPLUG_PCI_FAKE is not set
++# CONFIG_HOTPLUG_PCI_ACPI is not set
++# CONFIG_HOTPLUG_PCI_CPCI is not set
++# CONFIG_HOTPLUG_PCI_SHPC is not set
+
-+/*
-+ * This is performance-critical, we want to do it O(1)
-+ *
-+ * the indexing order of this array favors 1:1 mappings
-+ * between pins and IRQs.
-+ */
++#
++# Executable file formats / Emulations
++#
++CONFIG_BINFMT_ELF=y
++CONFIG_BINFMT_MISC=m
++CONFIG_IA32_EMULATION=y
++CONFIG_IA32_AOUT=y
++CONFIG_COMPAT=y
++CONFIG_SYSVIPC_COMPAT=y
+
-+static struct irq_pin_list {
-+ short apic, pin, next;
-+} irq_2_pin[PIN_MAP_SIZE];
++#
++# Networking
++#
++CONFIG_NET=y
+
-+int vector_irq[NR_VECTORS] __read_mostly = { [0 ... NR_VECTORS - 1] = -1};
-+#ifdef CONFIG_PCI_MSI
-+#define vector_to_irq(vector) \
-+ (platform_legacy_irq(vector) ? vector : vector_irq[vector])
-+#else
-+#define vector_to_irq(vector) (vector)
-+#endif
++#
++# Networking options
++#
++# CONFIG_NETDEBUG is not set
++CONFIG_PACKET=y
++CONFIG_PACKET_MMAP=y
++CONFIG_UNIX=y
++CONFIG_XFRM=y
++CONFIG_XFRM_USER=m
++CONFIG_NET_KEY=m
++CONFIG_INET=y
++CONFIG_IP_MULTICAST=y
++CONFIG_IP_ADVANCED_ROUTER=y
++CONFIG_ASK_IP_FIB_HASH=y
++# CONFIG_IP_FIB_TRIE is not set
++CONFIG_IP_FIB_HASH=y
++CONFIG_IP_MULTIPLE_TABLES=y
++CONFIG_IP_ROUTE_FWMARK=y
++CONFIG_IP_ROUTE_MULTIPATH=y
++# CONFIG_IP_ROUTE_MULTIPATH_CACHED is not set
++CONFIG_IP_ROUTE_VERBOSE=y
++CONFIG_IP_PNP=y
++CONFIG_IP_PNP_DHCP=y
++CONFIG_IP_PNP_BOOTP=y
++CONFIG_IP_PNP_RARP=y
++CONFIG_NET_IPIP=m
++CONFIG_NET_IPGRE=m
++CONFIG_NET_IPGRE_BROADCAST=y
++CONFIG_IP_MROUTE=y
++CONFIG_IP_PIMSM_V1=y
++CONFIG_IP_PIMSM_V2=y
++# CONFIG_ARPD is not set
++CONFIG_SYN_COOKIES=y
++CONFIG_INET_AH=m
++CONFIG_INET_ESP=m
++CONFIG_INET_IPCOMP=m
++CONFIG_INET_XFRM_TUNNEL=m
++CONFIG_INET_TUNNEL=m
++CONFIG_INET_XFRM_MODE_TRANSPORT=m
++CONFIG_INET_XFRM_MODE_TUNNEL=m
++CONFIG_INET_DIAG=m
++CONFIG_INET_TCP_DIAG=m
++CONFIG_TCP_CONG_ADVANCED=y
+
-+#ifdef CONFIG_XEN
++#
++# TCP congestion control
++#
++CONFIG_TCP_CONG_BIC=m
++CONFIG_TCP_CONG_CUBIC=m
++CONFIG_TCP_CONG_WESTWOOD=m
++CONFIG_TCP_CONG_HTCP=m
++CONFIG_TCP_CONG_HSTCP=m
++CONFIG_TCP_CONG_HYBLA=m
++CONFIG_TCP_CONG_VEGAS=m
++CONFIG_TCP_CONG_SCALABLE=m
++CONFIG_TCP_CONG_LP=m
++CONFIG_TCP_CONG_VENO=m
+
-+#include <xen/interface/xen.h>
-+#include <xen/interface/physdev.h>
++#
++# IP: Virtual Server Configuration
++#
++CONFIG_IP_VS=m
++# CONFIG_IP_VS_DEBUG is not set
++CONFIG_IP_VS_TAB_BITS=12
+
-+/* Fake i8259 */
-+#define make_8259A_irq(_irq) (io_apic_irqs &= ~(1UL<<(_irq)))
-+#define disable_8259A_irq(_irq) ((void)0)
-+#define i8259A_irq_pending(_irq) (0)
++#
++# IPVS transport protocol load balancing support
++#
++CONFIG_IP_VS_PROTO_TCP=y
++CONFIG_IP_VS_PROTO_UDP=y
++CONFIG_IP_VS_PROTO_ESP=y
++CONFIG_IP_VS_PROTO_AH=y
+
-+unsigned long io_apic_irqs;
++#
++# IPVS scheduler
++#
++CONFIG_IP_VS_RR=m
++CONFIG_IP_VS_WRR=m
++CONFIG_IP_VS_LC=m
++CONFIG_IP_VS_WLC=m
++CONFIG_IP_VS_LBLC=m
++CONFIG_IP_VS_LBLCR=m
++CONFIG_IP_VS_DH=m
++CONFIG_IP_VS_SH=m
++CONFIG_IP_VS_SED=m
++CONFIG_IP_VS_NQ=m
+
-+static inline unsigned int xen_io_apic_read(unsigned int apic, unsigned int reg)
-+{
-+ struct physdev_apic apic_op;
-+ int ret;
++#
++# IPVS application helper
++#
++CONFIG_IP_VS_FTP=m
++CONFIG_IPV6=m
++CONFIG_IPV6_PRIVACY=y
++# CONFIG_IPV6_ROUTER_PREF is not set
++CONFIG_INET6_AH=m
++CONFIG_INET6_ESP=m
++CONFIG_INET6_IPCOMP=m
++CONFIG_INET6_XFRM_TUNNEL=m
++CONFIG_INET6_TUNNEL=m
++CONFIG_INET6_XFRM_MODE_TRANSPORT=m
++CONFIG_INET6_XFRM_MODE_TUNNEL=m
++CONFIG_IPV6_TUNNEL=m
++CONFIG_NETWORK_SECMARK=y
++CONFIG_NETFILTER=y
++# CONFIG_NETFILTER_DEBUG is not set
++CONFIG_BRIDGE_NETFILTER=y
+
-+ apic_op.apic_physbase = mp_ioapics[apic].mpc_apicaddr;
-+ apic_op.reg = reg;
-+ ret = HYPERVISOR_physdev_op(PHYSDEVOP_apic_read, &apic_op);
-+ if (ret)
-+ return ret;
-+ return apic_op.value;
-+}
++#
++# Core Netfilter Configuration
++#
++CONFIG_NETFILTER_NETLINK=m
++CONFIG_NETFILTER_NETLINK_QUEUE=m
++CONFIG_NETFILTER_NETLINK_LOG=m
++CONFIG_NETFILTER_XTABLES=m
++CONFIG_NETFILTER_XT_TARGET_CLASSIFY=m
++CONFIG_NETFILTER_XT_TARGET_CONNMARK=m
++CONFIG_NETFILTER_XT_TARGET_MARK=m
++CONFIG_NETFILTER_XT_TARGET_NFQUEUE=m
++CONFIG_NETFILTER_XT_TARGET_NOTRACK=m
++CONFIG_NETFILTER_XT_TARGET_SECMARK=m
++CONFIG_NETFILTER_XT_TARGET_CONNSECMARK=m
++CONFIG_NETFILTER_XT_MATCH_COMMENT=m
++CONFIG_NETFILTER_XT_MATCH_CONNBYTES=m
++CONFIG_NETFILTER_XT_MATCH_CONNMARK=m
++CONFIG_NETFILTER_XT_MATCH_CONNTRACK=m
++CONFIG_NETFILTER_XT_MATCH_DCCP=m
++CONFIG_NETFILTER_XT_MATCH_ESP=m
++CONFIG_NETFILTER_XT_MATCH_HELPER=m
++CONFIG_NETFILTER_XT_MATCH_LENGTH=m
++CONFIG_NETFILTER_XT_MATCH_LIMIT=m
++CONFIG_NETFILTER_XT_MATCH_MAC=m
++CONFIG_NETFILTER_XT_MATCH_MARK=m
++CONFIG_NETFILTER_XT_MATCH_POLICY=m
++CONFIG_NETFILTER_XT_MATCH_MULTIPORT=m
++CONFIG_NETFILTER_XT_MATCH_PHYSDEV=m
++CONFIG_NETFILTER_XT_MATCH_PKTTYPE=m
++CONFIG_NETFILTER_XT_MATCH_QUOTA=m
++CONFIG_NETFILTER_XT_MATCH_REALM=m
++CONFIG_NETFILTER_XT_MATCH_SCTP=m
++CONFIG_NETFILTER_XT_MATCH_STATE=m
++CONFIG_NETFILTER_XT_MATCH_STATISTIC=m
++CONFIG_NETFILTER_XT_MATCH_STRING=m
++CONFIG_NETFILTER_XT_MATCH_TCPMSS=m
+
-+static inline void xen_io_apic_write(unsigned int apic, unsigned int reg, unsigned int value)
-+{
-+ struct physdev_apic apic_op;
++#
++# IP: Netfilter Configuration
++#
++CONFIG_IP_NF_CONNTRACK=m
++CONFIG_IP_NF_CT_ACCT=y
++CONFIG_IP_NF_CONNTRACK_MARK=y
++CONFIG_IP_NF_CONNTRACK_SECMARK=y
++CONFIG_IP_NF_CONNTRACK_EVENTS=y
++CONFIG_IP_NF_CONNTRACK_NETLINK=m
++CONFIG_IP_NF_CT_PROTO_SCTP=m
++CONFIG_IP_NF_FTP=m
++CONFIG_IP_NF_IRC=m
++CONFIG_IP_NF_NETBIOS_NS=m
++CONFIG_IP_NF_TFTP=m
++CONFIG_IP_NF_AMANDA=m
++CONFIG_IP_NF_PPTP=m
++CONFIG_IP_NF_H323=m
++CONFIG_IP_NF_SIP=m
++CONFIG_IP_NF_QUEUE=m
++CONFIG_IP_NF_IPTABLES=m
++CONFIG_IP_NF_MATCH_IPRANGE=m
++CONFIG_IP_NF_MATCH_TOS=m
++CONFIG_IP_NF_MATCH_RECENT=m
++CONFIG_IP_NF_MATCH_ECN=m
++CONFIG_IP_NF_MATCH_DSCP=m
++CONFIG_IP_NF_MATCH_AH=m
++CONFIG_IP_NF_MATCH_TTL=m
++CONFIG_IP_NF_MATCH_OWNER=m
++CONFIG_IP_NF_MATCH_ADDRTYPE=m
++CONFIG_IP_NF_MATCH_HASHLIMIT=m
++CONFIG_IP_NF_FILTER=m
++CONFIG_IP_NF_TARGET_REJECT=m
++CONFIG_IP_NF_TARGET_LOG=m
++CONFIG_IP_NF_TARGET_ULOG=m
++CONFIG_IP_NF_TARGET_TCPMSS=m
++CONFIG_IP_NF_NAT=m
++CONFIG_IP_NF_NAT_NEEDED=y
++CONFIG_IP_NF_TARGET_MASQUERADE=m
++CONFIG_IP_NF_TARGET_REDIRECT=m
++CONFIG_IP_NF_TARGET_NETMAP=m
++CONFIG_IP_NF_TARGET_SAME=m
++CONFIG_IP_NF_NAT_SNMP_BASIC=m
++CONFIG_IP_NF_NAT_IRC=m
++CONFIG_IP_NF_NAT_FTP=m
++CONFIG_IP_NF_NAT_TFTP=m
++CONFIG_IP_NF_NAT_AMANDA=m
++CONFIG_IP_NF_NAT_PPTP=m
++CONFIG_IP_NF_NAT_H323=m
++CONFIG_IP_NF_NAT_SIP=m
++CONFIG_IP_NF_MANGLE=m
++CONFIG_IP_NF_TARGET_TOS=m
++CONFIG_IP_NF_TARGET_ECN=m
++CONFIG_IP_NF_TARGET_DSCP=m
++CONFIG_IP_NF_TARGET_TTL=m
++CONFIG_IP_NF_TARGET_CLUSTERIP=m
++CONFIG_IP_NF_RAW=m
++CONFIG_IP_NF_ARPTABLES=m
++CONFIG_IP_NF_ARPFILTER=m
++CONFIG_IP_NF_ARP_MANGLE=m
+
-+ apic_op.apic_physbase = mp_ioapics[apic].mpc_apicaddr;
-+ apic_op.reg = reg;
-+ apic_op.value = value;
-+ HYPERVISOR_physdev_op(PHYSDEVOP_apic_write, &apic_op);
-+}
++#
++# IPv6: Netfilter Configuration (EXPERIMENTAL)
++#
++CONFIG_IP6_NF_QUEUE=m
++CONFIG_IP6_NF_IPTABLES=m
++CONFIG_IP6_NF_MATCH_RT=m
++CONFIG_IP6_NF_MATCH_OPTS=m
++CONFIG_IP6_NF_MATCH_FRAG=m
++CONFIG_IP6_NF_MATCH_HL=m
++CONFIG_IP6_NF_MATCH_OWNER=m
++CONFIG_IP6_NF_MATCH_IPV6HEADER=m
++CONFIG_IP6_NF_MATCH_AH=m
++CONFIG_IP6_NF_MATCH_EUI64=m
++CONFIG_IP6_NF_FILTER=m
++CONFIG_IP6_NF_TARGET_LOG=m
++CONFIG_IP6_NF_TARGET_REJECT=m
++CONFIG_IP6_NF_MANGLE=m
++CONFIG_IP6_NF_TARGET_HL=m
++CONFIG_IP6_NF_RAW=m
+
-+#define io_apic_read(a,r) xen_io_apic_read(a,r)
-+#define io_apic_write(a,r,v) xen_io_apic_write(a,r,v)
++#
++# DECnet: Netfilter Configuration
++#
++# CONFIG_DECNET_NF_GRABULATOR is not set
+
-+#define clear_IO_APIC() ((void)0)
++#
++# Bridge: Netfilter Configuration
++#
++CONFIG_BRIDGE_NF_EBTABLES=m
++CONFIG_BRIDGE_EBT_BROUTE=m
++CONFIG_BRIDGE_EBT_T_FILTER=m
++CONFIG_BRIDGE_EBT_T_NAT=m
++CONFIG_BRIDGE_EBT_802_3=m
++CONFIG_BRIDGE_EBT_AMONG=m
++CONFIG_BRIDGE_EBT_ARP=m
++CONFIG_BRIDGE_EBT_IP=m
++CONFIG_BRIDGE_EBT_LIMIT=m
++CONFIG_BRIDGE_EBT_MARK=m
++CONFIG_BRIDGE_EBT_PKTTYPE=m
++CONFIG_BRIDGE_EBT_STP=m
++CONFIG_BRIDGE_EBT_VLAN=m
++CONFIG_BRIDGE_EBT_ARPREPLY=m
++CONFIG_BRIDGE_EBT_DNAT=m
++CONFIG_BRIDGE_EBT_MARK_T=m
++CONFIG_BRIDGE_EBT_REDIRECT=m
++CONFIG_BRIDGE_EBT_SNAT=m
++CONFIG_BRIDGE_EBT_LOG=m
++CONFIG_BRIDGE_EBT_ULOG=m
+
-+#else
++#
++# DCCP Configuration (EXPERIMENTAL)
++#
++CONFIG_IP_DCCP=m
++CONFIG_INET_DCCP_DIAG=m
++CONFIG_IP_DCCP_ACKVEC=y
++
++#
++# DCCP CCIDs Configuration (EXPERIMENTAL)
++#
++CONFIG_IP_DCCP_CCID2=m
++CONFIG_IP_DCCP_CCID3=m
++CONFIG_IP_DCCP_TFRC_LIB=m
++
++#
++# DCCP Kernel Hacking
++#
++# CONFIG_IP_DCCP_DEBUG is not set
++
++#
++# SCTP Configuration (EXPERIMENTAL)
++#
++CONFIG_IP_SCTP=m
++# CONFIG_SCTP_DBG_MSG is not set
++# CONFIG_SCTP_DBG_OBJCNT is not set
++# CONFIG_SCTP_HMAC_NONE is not set
++# CONFIG_SCTP_HMAC_SHA1 is not set
++CONFIG_SCTP_HMAC_MD5=y
++
++#
++# TIPC Configuration (EXPERIMENTAL)
++#
++CONFIG_TIPC=m
++# CONFIG_TIPC_ADVANCED is not set
++# CONFIG_TIPC_DEBUG is not set
++CONFIG_ATM=m
++CONFIG_ATM_CLIP=m
++# CONFIG_ATM_CLIP_NO_ICMP is not set
++CONFIG_ATM_LANE=m
++# CONFIG_ATM_MPOA is not set
++CONFIG_ATM_BR2684=m
++# CONFIG_ATM_BR2684_IPFILTER is not set
++CONFIG_BRIDGE=m
++CONFIG_VLAN_8021Q=m
++CONFIG_DECNET=m
++# CONFIG_DECNET_ROUTER is not set
++CONFIG_LLC=y
++CONFIG_LLC2=m
++CONFIG_IPX=m
++CONFIG_IPX_INTERN=y
++CONFIG_ATALK=m
++CONFIG_DEV_APPLETALK=m
++CONFIG_IPDDP=m
++CONFIG_IPDDP_ENCAP=y
++CONFIG_IPDDP_DECAP=y
++CONFIG_X25=m
++CONFIG_LAPB=m
++CONFIG_ECONET=m
++# CONFIG_ECONET_AUNUDP is not set
++# CONFIG_ECONET_NATIVE is not set
++CONFIG_WAN_ROUTER=m
+
-+#ifdef CONFIG_SMP
-+static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t mask)
-+{
-+ unsigned long flags;
-+ unsigned int dest;
-+ cpumask_t tmp;
++#
++# QoS and/or fair queueing
++#
++CONFIG_NET_SCHED=y
++CONFIG_NET_SCH_CLK_JIFFIES=y
++# CONFIG_NET_SCH_CLK_GETTIMEOFDAY is not set
++# CONFIG_NET_SCH_CLK_CPU is not set
+
-+ cpus_and(tmp, mask, cpu_online_map);
-+ if (cpus_empty(tmp))
-+ tmp = TARGET_CPUS;
++#
++# Queueing/Scheduling
++#
++CONFIG_NET_SCH_CBQ=m
++CONFIG_NET_SCH_HTB=m
++CONFIG_NET_SCH_HFSC=m
++CONFIG_NET_SCH_ATM=m
++CONFIG_NET_SCH_PRIO=m
++CONFIG_NET_SCH_RED=m
++CONFIG_NET_SCH_SFQ=m
++CONFIG_NET_SCH_TEQL=m
++CONFIG_NET_SCH_TBF=m
++CONFIG_NET_SCH_GRED=m
++CONFIG_NET_SCH_DSMARK=m
++CONFIG_NET_SCH_NETEM=m
++CONFIG_NET_SCH_INGRESS=m
+
-+ cpus_and(mask, tmp, CPU_MASK_ALL);
++#
++# Classification
++#
++CONFIG_NET_CLS=y
++CONFIG_NET_CLS_BASIC=m
++CONFIG_NET_CLS_TCINDEX=m
++CONFIG_NET_CLS_ROUTE4=m
++CONFIG_NET_CLS_ROUTE=y
++CONFIG_NET_CLS_FW=m
++CONFIG_NET_CLS_U32=m
++CONFIG_CLS_U32_PERF=y
++CONFIG_CLS_U32_MARK=y
++CONFIG_NET_CLS_RSVP=m
++CONFIG_NET_CLS_RSVP6=m
++CONFIG_NET_EMATCH=y
++CONFIG_NET_EMATCH_STACK=32
++CONFIG_NET_EMATCH_CMP=m
++CONFIG_NET_EMATCH_NBYTE=m
++CONFIG_NET_EMATCH_U32=m
++CONFIG_NET_EMATCH_META=m
++CONFIG_NET_EMATCH_TEXT=m
++# CONFIG_NET_CLS_ACT is not set
++CONFIG_NET_CLS_POLICE=y
++CONFIG_NET_CLS_IND=y
++CONFIG_NET_ESTIMATOR=y
+
-+ dest = cpu_mask_to_apicid(mask);
++#
++# Network testing
++#
++CONFIG_NET_PKTGEN=m
++CONFIG_HAMRADIO=y
+
-+ /*
-+ * Only the high 8 bits are valid.
-+ */
-+ dest = SET_APIC_LOGICAL_ID(dest);
++#
++# Packet Radio protocols
++#
++CONFIG_AX25=m
++CONFIG_AX25_DAMA_SLAVE=y
++CONFIG_NETROM=m
++CONFIG_ROSE=m
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ __DO_ACTION(1, = dest, )
-+ set_irq_info(irq, mask);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+}
-+#endif
++#
++# AX.25 network device drivers
++#
++CONFIG_MKISS=m
++CONFIG_6PACK=m
++CONFIG_BPQETHER=m
++CONFIG_BAYCOM_SER_FDX=m
++CONFIG_BAYCOM_SER_HDX=m
++CONFIG_BAYCOM_PAR=m
++CONFIG_YAM=m
++CONFIG_IRDA=m
+
-+#endif /* !CONFIG_XEN */
++#
++# IrDA protocols
++#
++CONFIG_IRLAN=m
++CONFIG_IRNET=m
++CONFIG_IRCOMM=m
++CONFIG_IRDA_ULTRA=y
+
-+/*
-+ * The common case is 1:1 IRQ<->pin mappings. Sometimes there are
-+ * shared ISA-space IRQs, so we have to support them. We are super
-+ * fast in the common case, and fast for shared ISA-space IRQs.
-+ */
-+static void add_pin_to_irq(unsigned int irq, int apic, int pin)
-+{
-+ static int first_free_entry = NR_IRQS;
-+ struct irq_pin_list *entry = irq_2_pin + irq;
++#
++# IrDA options
++#
++CONFIG_IRDA_CACHE_LAST_LSAP=y
++CONFIG_IRDA_FAST_RR=y
++# CONFIG_IRDA_DEBUG is not set
+
-+ BUG_ON(irq >= NR_IRQS);
-+ while (entry->next)
-+ entry = irq_2_pin + entry->next;
++#
++# Infrared-port device drivers
++#
+
-+ if (entry->pin != -1) {
-+ entry->next = first_free_entry;
-+ entry = irq_2_pin + entry->next;
-+ if (++first_free_entry >= PIN_MAP_SIZE)
-+ panic("io_apic.c: ran out of irq_2_pin entries!");
-+ }
-+ entry->apic = apic;
-+ entry->pin = pin;
-+}
++#
++# SIR device drivers
++#
++CONFIG_IRTTY_SIR=m
+
-+#ifndef CONFIG_XEN
-+#define __DO_ACTION(R, ACTION, FINAL) \
-+ \
-+{ \
-+ int pin; \
-+ struct irq_pin_list *entry = irq_2_pin + irq; \
-+ \
-+ BUG_ON(irq >= NR_IRQS); \
-+ for (;;) { \
-+ unsigned int reg; \
-+ pin = entry->pin; \
-+ if (pin == -1) \
-+ break; \
-+ reg = io_apic_read(entry->apic, 0x10 + R + pin*2); \
-+ reg ACTION; \
-+ io_apic_modify(entry->apic, reg); \
-+ if (!entry->next) \
-+ break; \
-+ entry = irq_2_pin + entry->next; \
-+ } \
-+ FINAL; \
-+}
++#
++# Dongle support
++#
++CONFIG_DONGLE=y
++CONFIG_ESI_DONGLE=m
++CONFIG_ACTISYS_DONGLE=m
++CONFIG_TEKRAM_DONGLE=m
++CONFIG_TOIM3232_DONGLE=m
++CONFIG_LITELINK_DONGLE=m
++CONFIG_MA600_DONGLE=m
++CONFIG_GIRBIL_DONGLE=m
++CONFIG_MCP2120_DONGLE=m
++CONFIG_OLD_BELKIN_DONGLE=m
++CONFIG_ACT200L_DONGLE=m
+
-+#define DO_ACTION(name,R,ACTION, FINAL) \
-+ \
-+ static void name##_IO_APIC_irq (unsigned int irq) \
-+ __DO_ACTION(R, ACTION, FINAL)
++#
++# Old SIR device drivers
++#
+
-+DO_ACTION( __mask, 0, |= 0x00010000, io_apic_sync(entry->apic) )
-+ /* mask = 1 */
-+DO_ACTION( __unmask, 0, &= 0xfffeffff, )
-+ /* mask = 0 */
++#
++# Old Serial dongle support
++#
+
-+static void mask_IO_APIC_irq (unsigned int irq)
-+{
-+ unsigned long flags;
++#
++# FIR device drivers
++#
++CONFIG_USB_IRDA=m
++CONFIG_SIGMATEL_FIR=m
++CONFIG_NSC_FIR=m
++CONFIG_WINBOND_FIR=m
++CONFIG_SMC_IRCC_FIR=m
++CONFIG_ALI_FIR=m
++CONFIG_VLSI_FIR=m
++CONFIG_VIA_FIR=m
++CONFIG_MCS_FIR=m
++CONFIG_BT=m
++CONFIG_BT_L2CAP=m
++CONFIG_BT_SCO=m
++CONFIG_BT_RFCOMM=m
++CONFIG_BT_RFCOMM_TTY=y
++CONFIG_BT_BNEP=m
++CONFIG_BT_BNEP_MC_FILTER=y
++CONFIG_BT_BNEP_PROTO_FILTER=y
++CONFIG_BT_CMTP=m
++CONFIG_BT_HIDP=m
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ __mask_IO_APIC_irq(irq);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+}
++#
++# Bluetooth device drivers
++#
++CONFIG_BT_HCIUSB=m
++CONFIG_BT_HCIUSB_SCO=y
++CONFIG_BT_HCIUART=m
++CONFIG_BT_HCIUART_H4=y
++CONFIG_BT_HCIUART_BCSP=y
++CONFIG_BT_HCIBCM203X=m
++CONFIG_BT_HCIBPA10X=m
++CONFIG_BT_HCIBFUSB=m
++# CONFIG_BT_HCIDTL1 is not set
++# CONFIG_BT_HCIBT3C is not set
++# CONFIG_BT_HCIBLUECARD is not set
++# CONFIG_BT_HCIBTUART is not set
++CONFIG_BT_HCIVHCI=m
++CONFIG_IEEE80211=m
++# CONFIG_IEEE80211_DEBUG is not set
++CONFIG_IEEE80211_CRYPT_WEP=m
++CONFIG_IEEE80211_CRYPT_CCMP=m
++CONFIG_IEEE80211_CRYPT_TKIP=m
++CONFIG_IEEE80211_SOFTMAC=m
++# CONFIG_IEEE80211_SOFTMAC_DEBUG is not set
++CONFIG_WIRELESS_EXT=y
+
-+static void unmask_IO_APIC_irq (unsigned int irq)
-+{
-+ unsigned long flags;
++#
++# Device Drivers
++#
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ __unmask_IO_APIC_irq(irq);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+}
++#
++# Generic Driver Options
++#
++CONFIG_STANDALONE=y
++CONFIG_PREVENT_FIRMWARE_BUILD=y
++CONFIG_FW_LOADER=m
++# CONFIG_DEBUG_DRIVER is not set
++CONFIG_SYS_HYPERVISOR=y
+
-+static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin)
-+{
-+ struct IO_APIC_route_entry entry;
-+ unsigned long flags;
++#
++# Connector - unified userspace <-> kernelspace linker
++#
++CONFIG_CONNECTOR=y
++CONFIG_PROC_EVENTS=y
+
-+ /* Check delivery_mode to be sure we're not clearing an SMI pin */
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ *(((int*)&entry) + 0) = io_apic_read(apic, 0x10 + 2 * pin);
-+ *(((int*)&entry) + 1) = io_apic_read(apic, 0x11 + 2 * pin);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+ if (entry.delivery_mode == dest_SMI)
-+ return;
-+ /*
-+ * Disable it in the IO-APIC irq-routing table:
-+ */
-+ memset(&entry, 0, sizeof(entry));
-+ entry.mask = 1;
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ io_apic_write(apic, 0x10 + 2 * pin, *(((int *)&entry) + 0));
-+ io_apic_write(apic, 0x11 + 2 * pin, *(((int *)&entry) + 1));
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+}
++#
++# Memory Technology Devices (MTD)
++#
++CONFIG_MTD=m
++# CONFIG_MTD_DEBUG is not set
++CONFIG_MTD_CONCAT=m
++CONFIG_MTD_PARTITIONS=y
++CONFIG_MTD_REDBOOT_PARTS=m
++CONFIG_MTD_REDBOOT_DIRECTORY_BLOCK=-1
++# CONFIG_MTD_REDBOOT_PARTS_UNALLOCATED is not set
++# CONFIG_MTD_REDBOOT_PARTS_READONLY is not set
++CONFIG_MTD_CMDLINE_PARTS=y
+
-+static void clear_IO_APIC (void)
-+{
-+ int apic, pin;
++#
++# User Modules And Translation Layers
++#
++CONFIG_MTD_CHAR=m
++CONFIG_MTD_BLOCK=m
++CONFIG_MTD_BLOCK_RO=m
++CONFIG_FTL=m
++CONFIG_NFTL=m
++CONFIG_NFTL_RW=y
++CONFIG_INFTL=m
++CONFIG_RFD_FTL=m
+
-+ for (apic = 0; apic < nr_ioapics; apic++)
-+ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++)
-+ clear_IO_APIC_pin(apic, pin);
-+}
++#
++# RAM/ROM/Flash chip drivers
++#
++CONFIG_MTD_CFI=m
++CONFIG_MTD_JEDECPROBE=m
++CONFIG_MTD_GEN_PROBE=m
++CONFIG_MTD_CFI_ADV_OPTIONS=y
++CONFIG_MTD_CFI_NOSWAP=y
++# CONFIG_MTD_CFI_BE_BYTE_SWAP is not set
++# CONFIG_MTD_CFI_LE_BYTE_SWAP is not set
++# CONFIG_MTD_CFI_GEOMETRY is not set
++CONFIG_MTD_MAP_BANK_WIDTH_1=y
++CONFIG_MTD_MAP_BANK_WIDTH_2=y
++CONFIG_MTD_MAP_BANK_WIDTH_4=y
++# CONFIG_MTD_MAP_BANK_WIDTH_8 is not set
++# CONFIG_MTD_MAP_BANK_WIDTH_16 is not set
++# CONFIG_MTD_MAP_BANK_WIDTH_32 is not set
++CONFIG_MTD_CFI_I1=y
++CONFIG_MTD_CFI_I2=y
++# CONFIG_MTD_CFI_I4 is not set
++# CONFIG_MTD_CFI_I8 is not set
++# CONFIG_MTD_OTP is not set
++CONFIG_MTD_CFI_INTELEXT=m
++CONFIG_MTD_CFI_AMDSTD=m
++CONFIG_MTD_CFI_STAA=m
++CONFIG_MTD_CFI_UTIL=m
++CONFIG_MTD_RAM=m
++# CONFIG_MTD_ROM is not set
++CONFIG_MTD_ABSENT=m
++# CONFIG_MTD_OBSOLETE_CHIPS is not set
+
-+#endif /* !CONFIG_XEN */
++#
++# Mapping drivers for chip access
++#
++CONFIG_MTD_COMPLEX_MAPPINGS=y
++CONFIG_MTD_PHYSMAP=m
++CONFIG_MTD_PHYSMAP_START=0x8000000
++CONFIG_MTD_PHYSMAP_LEN=0x4000000
++CONFIG_MTD_PHYSMAP_BANKWIDTH=2
++# CONFIG_MTD_PNC2000 is not set
++CONFIG_MTD_SC520CDP=m
++CONFIG_MTD_NETSC520=m
++CONFIG_MTD_TS5500=m
++CONFIG_MTD_SBC_GXX=m
++# CONFIG_MTD_AMD76XROM is not set
++# CONFIG_MTD_ICHXROM is not set
++CONFIG_MTD_SCB2_FLASH=m
++# CONFIG_MTD_NETtel is not set
++# CONFIG_MTD_DILNETPC is not set
++# CONFIG_MTD_L440GX is not set
++CONFIG_MTD_PCI=m
++CONFIG_MTD_PLATRAM=m
+
-+static u8 gsi_2_irq[NR_IRQ_VECTORS] = { [0 ... NR_IRQ_VECTORS-1] = 0xFF };
++#
++# Self-contained MTD device drivers
++#
++CONFIG_MTD_PMC551=m
++# CONFIG_MTD_PMC551_BUGFIX is not set
++# CONFIG_MTD_PMC551_DEBUG is not set
++CONFIG_MTD_DATAFLASH=m
++CONFIG_MTD_M25P80=m
++# CONFIG_MTD_SLRAM is not set
++# CONFIG_MTD_PHRAM is not set
++CONFIG_MTD_MTDRAM=m
++CONFIG_MTDRAM_TOTAL_SIZE=4096
++CONFIG_MTDRAM_ERASE_SIZE=128
++CONFIG_MTD_BLOCK2MTD=m
+
-+/*
-+ * support for broken MP BIOSs, enables hand-redirection of PIRQ0-7 to
-+ * specific CPU-side IRQs.
-+ */
++#
++# Disk-On-Chip Device Drivers
++#
++CONFIG_MTD_DOC2000=m
++CONFIG_MTD_DOC2001=m
++CONFIG_MTD_DOC2001PLUS=m
++CONFIG_MTD_DOCPROBE=m
++CONFIG_MTD_DOCECC=m
++CONFIG_MTD_DOCPROBE_ADVANCED=y
++CONFIG_MTD_DOCPROBE_ADDRESS=0x0000
++CONFIG_MTD_DOCPROBE_HIGH=y
++CONFIG_MTD_DOCPROBE_55AA=y
+
-+#define MAX_PIRQS 8
-+static int pirq_entries [MAX_PIRQS];
-+static int pirqs_enabled;
-+int skip_ioapic_setup;
-+int ioapic_force;
++#
++# NAND Flash Device Drivers
++#
++CONFIG_MTD_NAND=m
++# CONFIG_MTD_NAND_VERIFY_WRITE is not set
++# CONFIG_MTD_NAND_ECC_SMC is not set
++CONFIG_MTD_NAND_IDS=m
++CONFIG_MTD_NAND_DISKONCHIP=m
++# CONFIG_MTD_NAND_DISKONCHIP_PROBE_ADVANCED is not set
++CONFIG_MTD_NAND_DISKONCHIP_PROBE_ADDRESS=0
++CONFIG_MTD_NAND_DISKONCHIP_BBTWRITE=y
++CONFIG_MTD_NAND_NANDSIM=m
+
-+/* dummy parsing: see setup.c */
++#
++# OneNAND Flash Device Drivers
++#
++CONFIG_MTD_ONENAND=m
++# CONFIG_MTD_ONENAND_VERIFY_WRITE is not set
++# CONFIG_MTD_ONENAND_OTP is not set
+
-+static int __init disable_ioapic_setup(char *str)
-+{
-+ skip_ioapic_setup = 1;
-+ return 1;
-+}
++#
++# Parallel port support
++#
++CONFIG_PARPORT=m
++CONFIG_PARPORT_PC=m
++# CONFIG_PARPORT_PC_FIFO is not set
++# CONFIG_PARPORT_PC_SUPERIO is not set
++# CONFIG_PARPORT_PC_PCMCIA is not set
++CONFIG_PARPORT_NOT_PC=y
++# CONFIG_PARPORT_GSC is not set
++CONFIG_PARPORT_AX88796=m
++CONFIG_PARPORT_1284=y
+
-+static int __init enable_ioapic_setup(char *str)
-+{
-+ ioapic_force = 1;
-+ skip_ioapic_setup = 0;
-+ return 1;
-+}
++#
++# Plug and Play support
++#
++CONFIG_PNP=y
++CONFIG_PNP_DEBUG=y
+
-+__setup("noapic", disable_ioapic_setup);
-+__setup("apic", enable_ioapic_setup);
++#
++# Protocols
++#
++CONFIG_PNPACPI=y
+
-+#ifndef CONFIG_XEN
-+static int __init setup_disable_8254_timer(char *s)
-+{
-+ timer_over_8254 = -1;
-+ return 1;
-+}
-+static int __init setup_enable_8254_timer(char *s)
-+{
-+ timer_over_8254 = 2;
-+ return 1;
-+}
++#
++# Block devices
++#
++CONFIG_BLK_DEV_FD=m
++CONFIG_PARIDE=m
++CONFIG_PARIDE_PARPORT=m
+
-+__setup("disable_8254_timer", setup_disable_8254_timer);
-+__setup("enable_8254_timer", setup_enable_8254_timer);
-+#endif /* !CONFIG_XEN */
++#
++# Parallel IDE high-level drivers
++#
++CONFIG_PARIDE_PD=m
++CONFIG_PARIDE_PCD=m
++CONFIG_PARIDE_PF=m
++CONFIG_PARIDE_PT=m
++CONFIG_PARIDE_PG=m
+
-+#include <asm/pci-direct.h>
-+#include <linux/pci_ids.h>
-+#include <linux/pci.h>
++#
++# Parallel IDE protocol modules
++#
++CONFIG_PARIDE_ATEN=m
++CONFIG_PARIDE_BPCK=m
++CONFIG_PARIDE_COMM=m
++CONFIG_PARIDE_DSTR=m
++CONFIG_PARIDE_FIT2=m
++CONFIG_PARIDE_FIT3=m
++CONFIG_PARIDE_EPAT=m
++CONFIG_PARIDE_EPATC8=y
++CONFIG_PARIDE_EPIA=m
++CONFIG_PARIDE_FRIQ=m
++CONFIG_PARIDE_FRPW=m
++CONFIG_PARIDE_KBIC=m
++CONFIG_PARIDE_KTTI=m
++CONFIG_PARIDE_ON20=m
++CONFIG_PARIDE_ON26=m
++CONFIG_BLK_CPQ_DA=m
++CONFIG_BLK_CPQ_CISS_DA=m
++CONFIG_CISS_SCSI_TAPE=y
++CONFIG_BLK_DEV_DAC960=m
++CONFIG_BLK_DEV_UMEM=m
++# CONFIG_BLK_DEV_COW_COMMON is not set
++CONFIG_BLK_DEV_LOOP=y
++CONFIG_BLK_DEV_CRYPTOLOOP=m
++CONFIG_BLK_DEV_NBD=m
++CONFIG_BLK_DEV_SX8=m
++# CONFIG_BLK_DEV_UB is not set
++CONFIG_BLK_DEV_RAM=y
++CONFIG_BLK_DEV_RAM_COUNT=16
++CONFIG_BLK_DEV_RAM_SIZE=16384
++CONFIG_BLK_DEV_RAM_BLOCKSIZE=1024
++CONFIG_BLK_DEV_INITRD=y
++CONFIG_CDROM_PKTCDVD=m
++CONFIG_CDROM_PKTCDVD_BUFFERS=8
++# CONFIG_CDROM_PKTCDVD_WCACHE is not set
++CONFIG_ATA_OVER_ETH=m
+
++#
++# ATA/ATAPI/MFM/RLL support
++#
++CONFIG_IDE=y
++CONFIG_BLK_DEV_IDE=y
+
-+#ifdef CONFIG_ACPI
++#
++# Please see Documentation/ide.txt for help/info on IDE drives
++#
++# CONFIG_BLK_DEV_IDE_SATA is not set
++# CONFIG_BLK_DEV_HD_IDE is not set
++CONFIG_BLK_DEV_IDEDISK=m
++CONFIG_IDEDISK_MULTI_MODE=y
++# CONFIG_BLK_DEV_IDECS is not set
++CONFIG_BLK_DEV_IDECD=m
++CONFIG_BLK_DEV_IDETAPE=m
++CONFIG_BLK_DEV_IDEFLOPPY=m
++CONFIG_BLK_DEV_IDESCSI=m
++# CONFIG_IDE_TASK_IOCTL is not set
+
-+static int nvidia_hpet_detected __initdata;
++#
++# IDE chipset support/bugfixes
++#
++CONFIG_IDE_GENERIC=m
++CONFIG_BLK_DEV_CMD640=y
++CONFIG_BLK_DEV_CMD640_ENHANCED=y
++CONFIG_BLK_DEV_IDEPNP=y
++CONFIG_BLK_DEV_IDEPCI=y
++CONFIG_IDEPCI_SHARE_IRQ=y
++CONFIG_BLK_DEV_OFFBOARD=y
++CONFIG_BLK_DEV_GENERIC=y
++# CONFIG_BLK_DEV_OPTI621 is not set
++CONFIG_BLK_DEV_RZ1000=m
++CONFIG_BLK_DEV_IDEDMA_PCI=y
++# CONFIG_BLK_DEV_IDEDMA_FORCED is not set
++CONFIG_IDEDMA_PCI_AUTO=y
++# CONFIG_IDEDMA_ONLYDISK is not set
++CONFIG_BLK_DEV_AEC62XX=m
++CONFIG_BLK_DEV_ALI15X3=m
++# CONFIG_WDC_ALI15X3 is not set
++CONFIG_BLK_DEV_AMD74XX=m
++CONFIG_BLK_DEV_ATIIXP=m
++CONFIG_BLK_DEV_CMD64X=m
++CONFIG_BLK_DEV_TRIFLEX=m
++CONFIG_BLK_DEV_CY82C693=m
++CONFIG_BLK_DEV_CS5520=m
++CONFIG_BLK_DEV_CS5530=m
++CONFIG_BLK_DEV_HPT34X=m
++CONFIG_HPT34X_AUTODMA=y
++CONFIG_BLK_DEV_HPT366=m
++CONFIG_BLK_DEV_SC1200=m
++CONFIG_BLK_DEV_PIIX=m
++CONFIG_BLK_DEV_IT821X=m
++CONFIG_BLK_DEV_NS87415=m
++CONFIG_BLK_DEV_PDC202XX_OLD=m
++CONFIG_PDC202XX_BURST=y
++CONFIG_BLK_DEV_PDC202XX_NEW=m
++CONFIG_BLK_DEV_SVWKS=m
++CONFIG_BLK_DEV_SIIMAGE=m
++CONFIG_BLK_DEV_SIS5513=m
++CONFIG_BLK_DEV_SLC90E66=m
++# CONFIG_BLK_DEV_TRM290 is not set
++CONFIG_BLK_DEV_VIA82CXXX=m
++# CONFIG_IDE_ARM is not set
++CONFIG_BLK_DEV_IDEDMA=y
++# CONFIG_IDEDMA_IVB is not set
++CONFIG_IDEDMA_AUTO=y
++# CONFIG_BLK_DEV_HD is not set
+
-+static int __init nvidia_hpet_check(unsigned long phys, unsigned long size)
-+{
-+ nvidia_hpet_detected = 1;
-+ return 0;
-+}
-+#endif
++#
++# SCSI device support
++#
++CONFIG_RAID_ATTRS=m
++CONFIG_SCSI=m
++CONFIG_SCSI_PROC_FS=y
+
-+/* Temporary Hack. Nvidia and VIA boards currently only work with IO-APIC
-+ off. Check for an Nvidia or VIA PCI bridge and turn it off.
-+ Use pci direct infrastructure because this runs before the PCI subsystem.
++#
++# SCSI support type (disk, tape, CD-ROM)
++#
++CONFIG_BLK_DEV_SD=m
++CONFIG_CHR_DEV_ST=m
++CONFIG_CHR_DEV_OSST=m
++CONFIG_BLK_DEV_SR=m
++CONFIG_BLK_DEV_SR_VENDOR=y
++CONFIG_CHR_DEV_SG=m
++CONFIG_CHR_DEV_SCH=m
+
-+ Can be overwritten with "apic"
++#
++# Some SCSI devices (e.g. CD jukebox) support multiple LUNs
++#
++CONFIG_SCSI_MULTI_LUN=y
++CONFIG_SCSI_CONSTANTS=y
++CONFIG_SCSI_LOGGING=y
+
-+ And another hack to disable the IOMMU on VIA chipsets.
++#
++# SCSI Transport Attributes
++#
++CONFIG_SCSI_SPI_ATTRS=m
++CONFIG_SCSI_FC_ATTRS=m
++CONFIG_SCSI_ISCSI_ATTRS=m
++CONFIG_SCSI_SAS_ATTRS=m
+
-+ ... and others. Really should move this somewhere else.
++#
++# SCSI low-level drivers
++#
++CONFIG_ISCSI_TCP=m
++CONFIG_BLK_DEV_3W_XXXX_RAID=m
++CONFIG_SCSI_3W_9XXX=m
++CONFIG_SCSI_ACARD=m
++CONFIG_SCSI_AACRAID=m
++CONFIG_SCSI_AIC7XXX=m
++CONFIG_AIC7XXX_CMDS_PER_DEVICE=4
++CONFIG_AIC7XXX_RESET_DELAY_MS=15000
++# CONFIG_AIC7XXX_DEBUG_ENABLE is not set
++CONFIG_AIC7XXX_DEBUG_MASK=0
++# CONFIG_AIC7XXX_REG_PRETTY_PRINT is not set
++CONFIG_SCSI_AIC7XXX_OLD=m
++CONFIG_SCSI_AIC79XX=m
++CONFIG_AIC79XX_CMDS_PER_DEVICE=4
++CONFIG_AIC79XX_RESET_DELAY_MS=15000
++# CONFIG_AIC79XX_ENABLE_RD_STRM is not set
++# CONFIG_AIC79XX_DEBUG_ENABLE is not set
++CONFIG_AIC79XX_DEBUG_MASK=0
++# CONFIG_AIC79XX_REG_PRETTY_PRINT is not set
++CONFIG_MEGARAID_NEWGEN=y
++CONFIG_MEGARAID_MM=m
++CONFIG_MEGARAID_MAILBOX=m
++CONFIG_MEGARAID_LEGACY=m
++CONFIG_MEGARAID_SAS=m
++CONFIG_SCSI_SATA=m
++CONFIG_SCSI_SATA_AHCI=m
++CONFIG_SCSI_SATA_SVW=m
++CONFIG_SCSI_ATA_PIIX=m
++CONFIG_SCSI_SATA_MV=m
++CONFIG_SCSI_SATA_NV=m
++CONFIG_SCSI_PDC_ADMA=m
++CONFIG_SCSI_HPTIOP=m
++CONFIG_SCSI_SATA_QSTOR=m
++CONFIG_SCSI_SATA_PROMISE=m
++CONFIG_SCSI_SATA_SX4=m
++CONFIG_SCSI_SATA_SIL=m
++CONFIG_SCSI_SATA_SIL24=m
++CONFIG_SCSI_SATA_SIS=m
++CONFIG_SCSI_SATA_ULI=m
++CONFIG_SCSI_SATA_VIA=m
++CONFIG_SCSI_SATA_VITESSE=m
++CONFIG_SCSI_SATA_INTEL_COMBINED=y
++CONFIG_SCSI_BUSLOGIC=m
++# CONFIG_SCSI_OMIT_FLASHPOINT is not set
++CONFIG_SCSI_DMX3191D=m
++CONFIG_SCSI_EATA=m
++CONFIG_SCSI_EATA_TAGGED_QUEUE=y
++CONFIG_SCSI_EATA_LINKED_COMMANDS=y
++CONFIG_SCSI_EATA_MAX_TAGS=16
++CONFIG_SCSI_FUTURE_DOMAIN=m
++CONFIG_SCSI_GDTH=m
++CONFIG_SCSI_IPS=m
++CONFIG_SCSI_INITIO=m
++CONFIG_SCSI_INIA100=m
++CONFIG_SCSI_PPA=m
++CONFIG_SCSI_IMM=m
++# CONFIG_SCSI_IZIP_EPP16 is not set
++# CONFIG_SCSI_IZIP_SLOW_CTR is not set
++CONFIG_SCSI_SYM53C8XX_2=m
++CONFIG_SCSI_SYM53C8XX_DMA_ADDRESSING_MODE=1
++CONFIG_SCSI_SYM53C8XX_DEFAULT_TAGS=16
++CONFIG_SCSI_SYM53C8XX_MAX_TAGS=64
++CONFIG_SCSI_SYM53C8XX_MMIO=y
++# CONFIG_SCSI_IPR is not set
++CONFIG_SCSI_QLOGIC_1280=m
++CONFIG_SCSI_QLA_FC=m
++CONFIG_SCSI_LPFC=m
++CONFIG_SCSI_DC395x=m
++CONFIG_SCSI_DC390T=m
++# CONFIG_SCSI_DEBUG is not set
+
-+ Kludge-O-Rama. */
-+void __init check_ioapic(void)
-+{
-+ int num,slot,func;
-+ /* Poor man's PCI discovery */
-+ for (num = 0; num < 32; num++) {
-+ for (slot = 0; slot < 32; slot++) {
-+ for (func = 0; func < 8; func++) {
-+ u32 class;
-+ u32 vendor;
-+ u8 type;
-+ class = read_pci_config(num,slot,func,
-+ PCI_CLASS_REVISION);
-+ if (class == 0xffffffff)
-+ break;
++#
++# PCMCIA SCSI adapter support
++#
++# CONFIG_PCMCIA_FDOMAIN is not set
++# CONFIG_PCMCIA_QLOGIC is not set
++# CONFIG_PCMCIA_SYM53C500 is not set
+
-+ if ((class >> 16) != PCI_CLASS_BRIDGE_PCI)
-+ continue;
++#
++# Multi-device support (RAID and LVM)
++#
++CONFIG_MD=y
++CONFIG_BLK_DEV_MD=y
++CONFIG_MD_LINEAR=m
++CONFIG_MD_RAID0=m
++CONFIG_MD_RAID1=m
++CONFIG_MD_RAID10=m
++CONFIG_MD_RAID456=m
++CONFIG_MD_RAID5_RESHAPE=y
++CONFIG_MD_MULTIPATH=m
++CONFIG_MD_FAULTY=m
++CONFIG_BLK_DEV_DM=m
++CONFIG_DM_CRYPT=m
++CONFIG_DM_SNAPSHOT=m
++CONFIG_DM_MIRROR=m
++CONFIG_DM_ZERO=m
++CONFIG_DM_MULTIPATH=m
++CONFIG_DM_MULTIPATH_EMC=m
+
-+ vendor = read_pci_config(num, slot, func,
-+ PCI_VENDOR_ID);
-+ vendor &= 0xffff;
-+ switch (vendor) {
-+ case PCI_VENDOR_ID_VIA:
-+#ifdef CONFIG_IOMMU
-+ if ((end_pfn > MAX_DMA32_PFN ||
-+ force_iommu) &&
-+ !iommu_aperture_allowed) {
-+ printk(KERN_INFO
-+ "Looks like a VIA chipset. Disabling IOMMU. Override with \"iommu=allowed\"\n");
-+ iommu_aperture_disabled = 1;
-+ }
-+#endif
-+ return;
-+ case PCI_VENDOR_ID_NVIDIA:
-+#ifdef CONFIG_ACPI
-+ /*
-+ * All timer overrides on Nvidia are
-+ * wrong unless HPET is enabled.
-+ */
-+ nvidia_hpet_detected = 0;
-+ acpi_table_parse(ACPI_HPET,
-+ nvidia_hpet_check);
-+ if (nvidia_hpet_detected == 0) {
-+ acpi_skip_timer_override = 1;
-+ printk(KERN_INFO "Nvidia board "
-+ "detected. Ignoring ACPI "
-+ "timer override.\n");
-+ }
-+#endif
-+ /* RED-PEN skip them on mptables too? */
-+ return;
-+ case PCI_VENDOR_ID_ATI:
++#
++# Fusion MPT device support
++#
++CONFIG_FUSION=y
++CONFIG_FUSION_SPI=m
++CONFIG_FUSION_FC=m
++CONFIG_FUSION_SAS=m
++CONFIG_FUSION_MAX_SGE=128
++CONFIG_FUSION_CTL=m
++CONFIG_FUSION_LAN=m
+
-+ /* This should be actually default, but
-+ for 2.6.16 let's do it for ATI only where
-+ it's really needed. */
-+#ifndef CONFIG_XEN
-+ if (timer_over_8254 == 1) {
-+ timer_over_8254 = 0;
-+ printk(KERN_INFO
-+ "ATI board detected. Disabling timer routing over 8254.\n");
-+ }
-+#endif
-+ return;
-+ }
++#
++# IEEE 1394 (FireWire) support
++#
++CONFIG_IEEE1394=m
+
++#
++# Subsystem Options
++#
++# CONFIG_IEEE1394_VERBOSEDEBUG is not set
++CONFIG_IEEE1394_OUI_DB=y
++CONFIG_IEEE1394_EXTRA_CONFIG_ROMS=y
++CONFIG_IEEE1394_CONFIG_ROM_IP1394=y
++# CONFIG_IEEE1394_EXPORT_FULL_API is not set
+
-+ /* No multi-function device? */
-+ type = read_pci_config_byte(num,slot,func,
-+ PCI_HEADER_TYPE);
-+ if (!(type & 0x80))
-+ break;
-+ }
-+ }
-+ }
-+}
++#
++# Device Drivers
++#
++CONFIG_IEEE1394_PCILYNX=m
++CONFIG_IEEE1394_OHCI1394=m
+
-+static int __init ioapic_pirq_setup(char *str)
-+{
-+ int i, max;
-+ int ints[MAX_PIRQS+1];
++#
++# Protocol Drivers
++#
++CONFIG_IEEE1394_VIDEO1394=m
++CONFIG_IEEE1394_SBP2=m
++CONFIG_IEEE1394_ETH1394=m
++CONFIG_IEEE1394_DV1394=m
++CONFIG_IEEE1394_RAWIO=m
+
-+ get_options(str, ARRAY_SIZE(ints), ints);
++#
++# I2O device support
++#
++CONFIG_I2O=m
++CONFIG_I2O_LCT_NOTIFY_ON_CHANGES=y
++CONFIG_I2O_EXT_ADAPTEC=y
++CONFIG_I2O_EXT_ADAPTEC_DMA64=y
++CONFIG_I2O_CONFIG=m
++CONFIG_I2O_CONFIG_OLD_IOCTL=y
++CONFIG_I2O_BUS=m
++CONFIG_I2O_BLOCK=m
++CONFIG_I2O_SCSI=m
++CONFIG_I2O_PROC=m
+
-+ for (i = 0; i < MAX_PIRQS; i++)
-+ pirq_entries[i] = -1;
++#
++# Network device support
++#
++CONFIG_NETDEVICES=y
++CONFIG_DUMMY=m
++CONFIG_BONDING=m
++CONFIG_EQUALIZER=m
++CONFIG_TUN=m
++CONFIG_NET_SB1000=m
+
-+ pirqs_enabled = 1;
-+ apic_printk(APIC_VERBOSE, "PIRQ redirection, working around broken MP-BIOS.\n");
-+ max = MAX_PIRQS;
-+ if (ints[0] < MAX_PIRQS)
-+ max = ints[0];
++#
++# ARCnet devices
++#
++CONFIG_ARCNET=m
++CONFIG_ARCNET_1201=m
++CONFIG_ARCNET_1051=m
++CONFIG_ARCNET_RAW=m
++CONFIG_ARCNET_CAP=m
++CONFIG_ARCNET_COM90xx=m
++CONFIG_ARCNET_COM90xxIO=m
++CONFIG_ARCNET_RIM_I=m
++# CONFIG_ARCNET_COM20020 is not set
+
-+ for (i = 0; i < max; i++) {
-+ apic_printk(APIC_VERBOSE, "... PIRQ%d -> IRQ %d\n", i, ints[i+1]);
-+ /*
-+ * PIRQs are mapped upside down, usually.
-+ */
-+ pirq_entries[MAX_PIRQS-i-1] = ints[i+1];
-+ }
-+ return 1;
-+}
++#
++# PHY device support
++#
++CONFIG_PHYLIB=m
+
-+__setup("pirq=", ioapic_pirq_setup);
++#
++# MII PHY device drivers
++#
++CONFIG_MARVELL_PHY=m
++CONFIG_DAVICOM_PHY=m
++CONFIG_QSEMI_PHY=m
++CONFIG_LXT_PHY=m
++CONFIG_CICADA_PHY=m
++CONFIG_VITESSE_PHY=m
++CONFIG_SMSC_PHY=m
++CONFIG_FIXED_PHY=m
++# CONFIG_FIXED_MII_10_FDX is not set
++# CONFIG_FIXED_MII_100_FDX is not set
+
-+/*
-+ * Find the IRQ entry number of a certain pin.
-+ */
-+static int find_irq_entry(int apic, int pin, int type)
-+{
-+ int i;
++#
++# Ethernet (10 or 100Mbit)
++#
++CONFIG_NET_ETHERNET=y
++CONFIG_MII=m
++CONFIG_HAPPYMEAL=m
++CONFIG_SUNGEM=m
++CONFIG_CASSINI=m
++CONFIG_NET_VENDOR_3COM=y
++CONFIG_VORTEX=m
++CONFIG_TYPHOON=m
+
-+ for (i = 0; i < mp_irq_entries; i++)
-+ if (mp_irqs[i].mpc_irqtype == type &&
-+ (mp_irqs[i].mpc_dstapic == mp_ioapics[apic].mpc_apicid ||
-+ mp_irqs[i].mpc_dstapic == MP_APIC_ALL) &&
-+ mp_irqs[i].mpc_dstirq == pin)
-+ return i;
++#
++# Tulip family network device support
++#
++CONFIG_NET_TULIP=y
++CONFIG_DE2104X=m
++CONFIG_TULIP=m
++# CONFIG_TULIP_MWI is not set
++# CONFIG_TULIP_MMIO is not set
++CONFIG_TULIP_NAPI=y
++CONFIG_TULIP_NAPI_HW_MITIGATION=y
++CONFIG_DE4X5=m
++CONFIG_WINBOND_840=m
++CONFIG_DM9102=m
++CONFIG_ULI526X=m
++CONFIG_PCMCIA_XIRCOM=m
++CONFIG_HP100=m
++CONFIG_NET_PCI=y
++CONFIG_PCNET32=m
++CONFIG_AMD8111_ETH=m
++CONFIG_AMD8111E_NAPI=y
++CONFIG_ADAPTEC_STARFIRE=m
++CONFIG_ADAPTEC_STARFIRE_NAPI=y
++CONFIG_B44=m
++CONFIG_FORCEDETH=m
++CONFIG_DGRS=m
++CONFIG_EEPRO100=m
++CONFIG_E100=m
++CONFIG_FEALNX=m
++CONFIG_NATSEMI=m
++CONFIG_NE2K_PCI=m
++CONFIG_8139CP=m
++CONFIG_8139TOO=m
++# CONFIG_8139TOO_PIO is not set
++# CONFIG_8139TOO_TUNE_TWISTER is not set
++CONFIG_8139TOO_8129=y
++# CONFIG_8139_OLD_RX_RESET is not set
++CONFIG_SIS900=m
++CONFIG_EPIC100=m
++CONFIG_SUNDANCE=m
++# CONFIG_SUNDANCE_MMIO is not set
++CONFIG_VIA_RHINE=m
++# CONFIG_VIA_RHINE_MMIO is not set
++# CONFIG_VIA_RHINE_NAPI is not set
++# CONFIG_NET_POCKET is not set
+
-+ return -1;
-+}
++#
++# Ethernet (1000 Mbit)
++#
++CONFIG_ACENIC=m
++# CONFIG_ACENIC_OMIT_TIGON_I is not set
++CONFIG_DL2K=m
++CONFIG_E1000=m
++CONFIG_E1000_NAPI=y
++# CONFIG_E1000_DISABLE_PACKET_SPLIT is not set
++CONFIG_NS83820=m
++CONFIG_HAMACHI=m
++CONFIG_YELLOWFIN=m
++CONFIG_R8169=m
++CONFIG_R8169_NAPI=y
++CONFIG_R8169_VLAN=y
++CONFIG_SIS190=m
++CONFIG_SKGE=m
++CONFIG_SKY2=m
++CONFIG_SK98LIN=m
++CONFIG_VIA_VELOCITY=m
++CONFIG_TIGON3=m
++CONFIG_BNX2=m
+
-+#ifndef CONFIG_XEN
-+/*
-+ * Find the pin to which IRQ[irq] (ISA) is connected
-+ */
-+static int __init find_isa_irq_pin(int irq, int type)
-+{
-+ int i;
++#
++# Ethernet (10000 Mbit)
++#
++CONFIG_CHELSIO_T1=m
++CONFIG_IXGB=m
++CONFIG_IXGB_NAPI=y
++CONFIG_S2IO=m
++CONFIG_S2IO_NAPI=y
++CONFIG_MYRI10GE=m
++# CONFIG_SFC is not set
+
-+ for (i = 0; i < mp_irq_entries; i++) {
-+ int lbus = mp_irqs[i].mpc_srcbus;
++#
++# Token Ring devices
++#
++CONFIG_TR=y
++CONFIG_IBMOL=m
++CONFIG_3C359=m
++CONFIG_TMS380TR=m
++CONFIG_TMSPCI=m
++CONFIG_ABYSS=m
+
-+ if ((mp_bus_id_to_type[lbus] == MP_BUS_ISA ||
-+ mp_bus_id_to_type[lbus] == MP_BUS_EISA ||
-+ mp_bus_id_to_type[lbus] == MP_BUS_MCA) &&
-+ (mp_irqs[i].mpc_irqtype == type) &&
-+ (mp_irqs[i].mpc_srcbusirq == irq))
++#
++# Wireless LAN (non-hamradio)
++#
++CONFIG_NET_RADIO=y
++CONFIG_NET_WIRELESS_RTNETLINK=y
+
-+ return mp_irqs[i].mpc_dstirq;
-+ }
-+ return -1;
-+}
++#
++# Obsolete Wireless cards support (pre-802.11)
++#
++CONFIG_STRIP=m
++CONFIG_PCMCIA_WAVELAN=m
++CONFIG_PCMCIA_NETWAVE=m
+
-+static int __init find_isa_irq_apic(int irq, int type)
-+{
-+ int i;
++#
++# Wireless 802.11 Frequency Hopping cards support
++#
++CONFIG_PCMCIA_RAYCS=m
+
-+ for (i = 0; i < mp_irq_entries; i++) {
-+ int lbus = mp_irqs[i].mpc_srcbus;
++#
++# Wireless 802.11b ISA/PCI cards support
++#
++CONFIG_IPW2100=m
++CONFIG_IPW2100_MONITOR=y
++# CONFIG_IPW2100_DEBUG is not set
++CONFIG_IPW2200=m
++CONFIG_IPW2200_MONITOR=y
++# CONFIG_IPW2200_RADIOTAP is not set
++# CONFIG_IPW2200_PROMISCUOUS is not set
++CONFIG_IPW2200_QOS=y
++# CONFIG_IPW2200_DEBUG is not set
++CONFIG_AIRO=m
++CONFIG_HERMES=m
++CONFIG_PLX_HERMES=m
++CONFIG_TMD_HERMES=m
++CONFIG_NORTEL_HERMES=m
++CONFIG_PCI_HERMES=m
++CONFIG_ATMEL=m
++CONFIG_PCI_ATMEL=m
+
-+ if ((mp_bus_id_to_type[lbus] == MP_BUS_ISA ||
-+ mp_bus_id_to_type[lbus] == MP_BUS_EISA ||
-+ mp_bus_id_to_type[lbus] == MP_BUS_MCA) &&
-+ (mp_irqs[i].mpc_irqtype == type) &&
-+ (mp_irqs[i].mpc_srcbusirq == irq))
-+ break;
-+ }
-+ if (i < mp_irq_entries) {
-+ int apic;
-+ for(apic = 0; apic < nr_ioapics; apic++) {
-+ if (mp_ioapics[apic].mpc_apicid == mp_irqs[i].mpc_dstapic)
-+ return apic;
-+ }
-+ }
++#
++# Wireless 802.11b Pcmcia/Cardbus cards support
++#
++# CONFIG_PCMCIA_HERMES is not set
++# CONFIG_PCMCIA_SPECTRUM is not set
++# CONFIG_AIRO_CS is not set
++# CONFIG_PCMCIA_ATMEL is not set
++# CONFIG_PCMCIA_WL3501 is not set
+
-+ return -1;
-+}
-+#endif
++#
++# Prism GT/Duette 802.11(a/b/g) PCI/Cardbus support
++#
++CONFIG_PRISM54=m
++CONFIG_USB_ZD1201=m
++CONFIG_HOSTAP=m
++CONFIG_HOSTAP_FIRMWARE=y
++CONFIG_HOSTAP_FIRMWARE_NVRAM=y
++CONFIG_HOSTAP_PLX=m
++CONFIG_HOSTAP_PCI=m
++# CONFIG_HOSTAP_CS is not set
++CONFIG_BCM43XX=m
++CONFIG_BCM43XX_DEBUG=y
++CONFIG_BCM43XX_DMA=y
++CONFIG_BCM43XX_PIO=y
++CONFIG_BCM43XX_DMA_AND_PIO_MODE=y
++# CONFIG_BCM43XX_DMA_MODE is not set
++# CONFIG_BCM43XX_PIO_MODE is not set
++CONFIG_ZD1211RW=m
++# CONFIG_ZD1211RW_DEBUG is not set
++CONFIG_NET_WIRELESS=y
+
-+/*
-+ * Find a specific PCI IRQ entry.
-+ * Not an __init, possibly needed by modules
-+ */
-+static int pin_2_irq(int idx, int apic, int pin);
++#
++# PCMCIA network device support
++#
++# CONFIG_NET_PCMCIA is not set
+
-+int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin)
-+{
-+ int apic, i, best_guess = -1;
++#
++# Wan interfaces
++#
++# CONFIG_WAN is not set
+
-+ apic_printk(APIC_DEBUG, "querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n",
-+ bus, slot, pin);
-+ if (mp_bus_id_to_pci_bus[bus] == -1) {
-+ apic_printk(APIC_VERBOSE, "PCI BIOS passed nonexistent PCI bus %d!\n", bus);
-+ return -1;
-+ }
-+ for (i = 0; i < mp_irq_entries; i++) {
-+ int lbus = mp_irqs[i].mpc_srcbus;
++#
++# ATM drivers
++#
++CONFIG_ATM_DUMMY=m
++CONFIG_ATM_TCP=m
++CONFIG_ATM_LANAI=m
++CONFIG_ATM_ENI=m
++# CONFIG_ATM_ENI_DEBUG is not set
++# CONFIG_ATM_ENI_TUNE_BURST is not set
++CONFIG_ATM_FIRESTREAM=m
++# CONFIG_ATM_ZATM is not set
++CONFIG_ATM_IDT77252=m
++# CONFIG_ATM_IDT77252_DEBUG is not set
++# CONFIG_ATM_IDT77252_RCV_ALL is not set
++CONFIG_ATM_IDT77252_USE_SUNI=y
++CONFIG_ATM_AMBASSADOR=m
++# CONFIG_ATM_AMBASSADOR_DEBUG is not set
++CONFIG_ATM_HORIZON=m
++# CONFIG_ATM_HORIZON_DEBUG is not set
++CONFIG_ATM_FORE200E_MAYBE=m
++# CONFIG_ATM_FORE200E_PCA is not set
++CONFIG_ATM_HE=m
++# CONFIG_ATM_HE_USE_SUNI is not set
++CONFIG_FDDI=y
++# CONFIG_DEFXX is not set
++CONFIG_SKFP=m
++CONFIG_HIPPI=y
++CONFIG_ROADRUNNER=m
++CONFIG_ROADRUNNER_LARGE_RINGS=y
++CONFIG_PLIP=m
++CONFIG_PPP=m
++CONFIG_PPP_MULTILINK=y
++CONFIG_PPP_FILTER=y
++CONFIG_PPP_ASYNC=m
++CONFIG_PPP_SYNC_TTY=m
++CONFIG_PPP_DEFLATE=m
++CONFIG_PPP_BSDCOMP=m
++CONFIG_PPP_MPPE=m
++CONFIG_PPPOE=m
++CONFIG_PPPOATM=m
++CONFIG_SLIP=m
++CONFIG_SLIP_COMPRESSED=y
++CONFIG_SLIP_SMART=y
++CONFIG_SLIP_MODE_SLIP6=y
++CONFIG_NET_FC=y
++CONFIG_SHAPER=m
++CONFIG_NETCONSOLE=m
++CONFIG_NETPOLL=y
++CONFIG_NETPOLL_RX=y
++CONFIG_NETPOLL_TRAP=y
++CONFIG_NET_POLL_CONTROLLER=y
+
-+ for (apic = 0; apic < nr_ioapics; apic++)
-+ if (mp_ioapics[apic].mpc_apicid == mp_irqs[i].mpc_dstapic ||
-+ mp_irqs[i].mpc_dstapic == MP_APIC_ALL)
-+ break;
++#
++# ISDN subsystem
++#
++CONFIG_ISDN=m
+
-+ if ((mp_bus_id_to_type[lbus] == MP_BUS_PCI) &&
-+ !mp_irqs[i].mpc_irqtype &&
-+ (bus == lbus) &&
-+ (slot == ((mp_irqs[i].mpc_srcbusirq >> 2) & 0x1f))) {
-+ int irq = pin_2_irq(i,apic,mp_irqs[i].mpc_dstirq);
++#
++# Old ISDN4Linux
++#
++CONFIG_ISDN_I4L=m
++CONFIG_ISDN_PPP=y
++CONFIG_ISDN_PPP_VJ=y
++CONFIG_ISDN_MPP=y
++CONFIG_IPPP_FILTER=y
++CONFIG_ISDN_PPP_BSDCOMP=m
++CONFIG_ISDN_AUDIO=y
++CONFIG_ISDN_TTY_FAX=y
++CONFIG_ISDN_X25=y
+
-+ if (!(apic || IO_APIC_IRQ(irq)))
-+ continue;
++#
++# ISDN feature submodules
++#
++CONFIG_ISDN_DIVERSION=m
+
-+ if (pin == (mp_irqs[i].mpc_srcbusirq & 3))
-+ return irq;
-+ /*
-+ * Use the first all-but-pin matching entry as a
-+ * best-guess fuzzy result for broken mptables.
-+ */
-+ if (best_guess < 0)
-+ best_guess = irq;
-+ }
-+ }
-+ BUG_ON(best_guess >= NR_IRQS);
-+ return best_guess;
-+}
++#
++# ISDN4Linux hardware drivers
++#
+
-+/*
-+ * EISA Edge/Level control register, ELCR
-+ */
-+static int EISA_ELCR(unsigned int irq)
-+{
-+ if (irq < 16) {
-+ unsigned int port = 0x4d0 + (irq >> 3);
-+ return (inb(port) >> (irq & 7)) & 1;
-+ }
-+ apic_printk(APIC_VERBOSE, "Broken MPtable reports ISA irq %d\n", irq);
-+ return 0;
-+}
++#
++# Passive cards
++#
++CONFIG_ISDN_DRV_HISAX=m
+
-+/* EISA interrupts are always polarity zero and can be edge or level
-+ * trigger depending on the ELCR value. If an interrupt is listed as
-+ * EISA conforming in the MP table, that means its trigger type must
-+ * be read in from the ELCR */
++#
++# D-channel protocol features
++#
++CONFIG_HISAX_EURO=y
++CONFIG_DE_AOC=y
++CONFIG_HISAX_NO_SENDCOMPLETE=y
++CONFIG_HISAX_NO_LLC=y
++CONFIG_HISAX_NO_KEYPAD=y
++CONFIG_HISAX_1TR6=y
++CONFIG_HISAX_NI1=y
++CONFIG_HISAX_MAX_CARDS=8
+
-+#define default_EISA_trigger(idx) (EISA_ELCR(mp_irqs[idx].mpc_srcbusirq))
-+#define default_EISA_polarity(idx) (0)
++#
++# HiSax supported cards
++#
++CONFIG_HISAX_16_3=y
++CONFIG_HISAX_TELESPCI=y
++CONFIG_HISAX_S0BOX=y
++CONFIG_HISAX_FRITZPCI=y
++CONFIG_HISAX_AVM_A1_PCMCIA=y
++CONFIG_HISAX_ELSA=y
++CONFIG_HISAX_DIEHLDIVA=y
++CONFIG_HISAX_SEDLBAUER=y
++CONFIG_HISAX_NETJET=y
++CONFIG_HISAX_NETJET_U=y
++CONFIG_HISAX_NICCY=y
++CONFIG_HISAX_BKM_A4T=y
++CONFIG_HISAX_SCT_QUADRO=y
++CONFIG_HISAX_GAZEL=y
++CONFIG_HISAX_HFC_PCI=y
++CONFIG_HISAX_W6692=y
++CONFIG_HISAX_HFC_SX=y
++CONFIG_HISAX_ENTERNOW_PCI=y
++# CONFIG_HISAX_DEBUG is not set
+
-+/* ISA interrupts are always polarity zero edge triggered,
-+ * when listed as conforming in the MP table. */
++#
++# HiSax PCMCIA card service modules
++#
++# CONFIG_HISAX_SEDLBAUER_CS is not set
++# CONFIG_HISAX_ELSA_CS is not set
++# CONFIG_HISAX_AVM_A1_CS is not set
++# CONFIG_HISAX_TELES_CS is not set
+
-+#define default_ISA_trigger(idx) (0)
-+#define default_ISA_polarity(idx) (0)
++#
++# HiSax sub driver modules
++#
++CONFIG_HISAX_ST5481=m
++CONFIG_HISAX_HFCUSB=m
++CONFIG_HISAX_HFC4S8S=m
++CONFIG_HISAX_FRITZ_PCIPNP=m
++CONFIG_HISAX_HDLC=y
+
-+/* PCI interrupts are always polarity one level triggered,
-+ * when listed as conforming in the MP table. */
++#
++# Active cards
++#
+
-+#define default_PCI_trigger(idx) (1)
-+#define default_PCI_polarity(idx) (1)
++#
++# Siemens Gigaset
++#
++CONFIG_ISDN_DRV_GIGASET=m
++CONFIG_GIGASET_BASE=m
++CONFIG_GIGASET_M105=m
++# CONFIG_GIGASET_DEBUG is not set
++# CONFIG_GIGASET_UNDOCREQ is not set
+
-+/* MCA interrupts are always polarity zero level triggered,
-+ * when listed as conforming in the MP table. */
++#
++# CAPI subsystem
++#
++CONFIG_ISDN_CAPI=m
++CONFIG_ISDN_DRV_AVMB1_VERBOSE_REASON=y
++CONFIG_ISDN_CAPI_MIDDLEWARE=y
++CONFIG_ISDN_CAPI_CAPI20=m
++CONFIG_ISDN_CAPI_CAPIFS_BOOL=y
++CONFIG_ISDN_CAPI_CAPIFS=m
++CONFIG_ISDN_CAPI_CAPIDRV=m
+
-+#define default_MCA_trigger(idx) (1)
-+#define default_MCA_polarity(idx) (0)
++#
++# CAPI hardware drivers
++#
+
-+static int __init MPBIOS_polarity(int idx)
-+{
-+ int bus = mp_irqs[idx].mpc_srcbus;
-+ int polarity;
++#
++# Active AVM cards
++#
++CONFIG_CAPI_AVM=y
++CONFIG_ISDN_DRV_AVMB1_B1PCI=m
++CONFIG_ISDN_DRV_AVMB1_B1PCIV4=y
++CONFIG_ISDN_DRV_AVMB1_B1PCMCIA=m
++# CONFIG_ISDN_DRV_AVMB1_AVM_CS is not set
++CONFIG_ISDN_DRV_AVMB1_T1PCI=m
++CONFIG_ISDN_DRV_AVMB1_C4=m
+
-+ /*
-+ * Determine IRQ line polarity (high active or low active):
-+ */
-+ switch (mp_irqs[idx].mpc_irqflag & 3)
-+ {
-+ case 0: /* conforms, ie. bus-type dependent polarity */
-+ {
-+ switch (mp_bus_id_to_type[bus])
-+ {
-+ case MP_BUS_ISA: /* ISA pin */
-+ {
-+ polarity = default_ISA_polarity(idx);
-+ break;
-+ }
-+ case MP_BUS_EISA: /* EISA pin */
-+ {
-+ polarity = default_EISA_polarity(idx);
-+ break;
-+ }
-+ case MP_BUS_PCI: /* PCI pin */
-+ {
-+ polarity = default_PCI_polarity(idx);
-+ break;
-+ }
-+ case MP_BUS_MCA: /* MCA pin */
-+ {
-+ polarity = default_MCA_polarity(idx);
-+ break;
-+ }
-+ default:
-+ {
-+ printk(KERN_WARNING "broken BIOS!!\n");
-+ polarity = 1;
-+ break;
-+ }
-+ }
-+ break;
-+ }
-+ case 1: /* high active */
-+ {
-+ polarity = 0;
-+ break;
-+ }
-+ case 2: /* reserved */
-+ {
-+ printk(KERN_WARNING "broken BIOS!!\n");
-+ polarity = 1;
-+ break;
-+ }
-+ case 3: /* low active */
-+ {
-+ polarity = 1;
-+ break;
-+ }
-+ default: /* invalid */
-+ {
-+ printk(KERN_WARNING "broken BIOS!!\n");
-+ polarity = 1;
-+ break;
-+ }
-+ }
-+ return polarity;
-+}
++#
++# Active Eicon DIVA Server cards
++#
++# CONFIG_CAPI_EICON is not set
+
-+static int MPBIOS_trigger(int idx)
-+{
-+ int bus = mp_irqs[idx].mpc_srcbus;
-+ int trigger;
++#
++# Telephony Support
++#
++CONFIG_PHONE=m
++CONFIG_PHONE_IXJ=m
++CONFIG_PHONE_IXJ_PCMCIA=m
+
-+ /*
-+ * Determine IRQ trigger mode (edge or level sensitive):
-+ */
-+ switch ((mp_irqs[idx].mpc_irqflag>>2) & 3)
-+ {
-+ case 0: /* conforms, ie. bus-type dependent */
-+ {
-+ switch (mp_bus_id_to_type[bus])
-+ {
-+ case MP_BUS_ISA: /* ISA pin */
-+ {
-+ trigger = default_ISA_trigger(idx);
-+ break;
-+ }
-+ case MP_BUS_EISA: /* EISA pin */
-+ {
-+ trigger = default_EISA_trigger(idx);
-+ break;
-+ }
-+ case MP_BUS_PCI: /* PCI pin */
-+ {
-+ trigger = default_PCI_trigger(idx);
-+ break;
-+ }
-+ case MP_BUS_MCA: /* MCA pin */
-+ {
-+ trigger = default_MCA_trigger(idx);
-+ break;
-+ }
-+ default:
-+ {
-+ printk(KERN_WARNING "broken BIOS!!\n");
-+ trigger = 1;
-+ break;
-+ }
-+ }
-+ break;
-+ }
-+ case 1: /* edge */
-+ {
-+ trigger = 0;
-+ break;
-+ }
-+ case 2: /* reserved */
-+ {
-+ printk(KERN_WARNING "broken BIOS!!\n");
-+ trigger = 1;
-+ break;
-+ }
-+ case 3: /* level */
-+ {
-+ trigger = 1;
-+ break;
-+ }
-+ default: /* invalid */
-+ {
-+ printk(KERN_WARNING "broken BIOS!!\n");
-+ trigger = 0;
-+ break;
-+ }
-+ }
-+ return trigger;
-+}
++#
++# Input device support
++#
++CONFIG_INPUT=y
+
-+static inline int irq_polarity(int idx)
-+{
-+ return MPBIOS_polarity(idx);
-+}
++#
++# Userland interfaces
++#
++CONFIG_INPUT_MOUSEDEV=y
++CONFIG_INPUT_MOUSEDEV_PSAUX=y
++CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024
++CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768
++CONFIG_INPUT_JOYDEV=m
++CONFIG_INPUT_TSDEV=m
++CONFIG_INPUT_TSDEV_SCREEN_X=240
++CONFIG_INPUT_TSDEV_SCREEN_Y=320
++CONFIG_INPUT_EVDEV=y
++# CONFIG_INPUT_EVBUG is not set
+
-+static inline int irq_trigger(int idx)
-+{
-+ return MPBIOS_trigger(idx);
-+}
++#
++# Input Device Drivers
++#
++CONFIG_INPUT_KEYBOARD=y
++CONFIG_KEYBOARD_ATKBD=y
++CONFIG_KEYBOARD_SUNKBD=m
++# CONFIG_KEYBOARD_LKKBD is not set
++CONFIG_KEYBOARD_XTKBD=m
++CONFIG_KEYBOARD_NEWTON=m
++CONFIG_INPUT_MOUSE=y
++CONFIG_MOUSE_PS2=y
++CONFIG_MOUSE_SERIAL=m
++# CONFIG_MOUSE_VSXXXAA is not set
++CONFIG_INPUT_JOYSTICK=y
++CONFIG_JOYSTICK_ANALOG=m
++CONFIG_JOYSTICK_A3D=m
++CONFIG_JOYSTICK_ADI=m
++CONFIG_JOYSTICK_COBRA=m
++CONFIG_JOYSTICK_GF2K=m
++CONFIG_JOYSTICK_GRIP=m
++CONFIG_JOYSTICK_GRIP_MP=m
++CONFIG_JOYSTICK_GUILLEMOT=m
++CONFIG_JOYSTICK_INTERACT=m
++CONFIG_JOYSTICK_SIDEWINDER=m
++CONFIG_JOYSTICK_TMDC=m
++CONFIG_JOYSTICK_IFORCE=m
++CONFIG_JOYSTICK_IFORCE_USB=y
++CONFIG_JOYSTICK_IFORCE_232=y
++CONFIG_JOYSTICK_WARRIOR=m
++CONFIG_JOYSTICK_MAGELLAN=m
++CONFIG_JOYSTICK_SPACEORB=m
++CONFIG_JOYSTICK_SPACEBALL=m
++CONFIG_JOYSTICK_STINGER=m
++CONFIG_JOYSTICK_TWIDJOY=m
++CONFIG_JOYSTICK_DB9=m
++CONFIG_JOYSTICK_GAMECON=m
++CONFIG_JOYSTICK_TURBOGRAFX=m
++CONFIG_JOYSTICK_JOYDUMP=m
++CONFIG_INPUT_TOUCHSCREEN=y
++CONFIG_TOUCHSCREEN_ADS7846=m
++CONFIG_TOUCHSCREEN_GUNZE=m
++CONFIG_TOUCHSCREEN_ELO=m
++CONFIG_TOUCHSCREEN_MTOUCH=m
++CONFIG_TOUCHSCREEN_MK712=m
++CONFIG_INPUT_MISC=y
++CONFIG_INPUT_PCSPKR=m
++CONFIG_INPUT_UINPUT=m
+
-+static int next_irq = 16;
++#
++# Hardware I/O ports
++#
++CONFIG_SERIO=y
++CONFIG_SERIO_I8042=y
++CONFIG_SERIO_SERPORT=m
++CONFIG_SERIO_CT82C710=m
++CONFIG_SERIO_PARKBD=m
++CONFIG_SERIO_PCIPS2=m
++CONFIG_SERIO_LIBPS2=y
++CONFIG_SERIO_RAW=m
++CONFIG_GAMEPORT=m
++CONFIG_GAMEPORT_NS558=m
++CONFIG_GAMEPORT_L4=m
++CONFIG_GAMEPORT_EMU10K1=m
++CONFIG_GAMEPORT_FM801=m
+
-+/*
-+ * gsi_irq_sharing -- Name overload! "irq" can be either a legacy IRQ
-+ * in the range 0-15, a linux IRQ in the range 0-223, or a GSI number
-+ * from ACPI, which can reach 800 in large boxen.
-+ *
-+ * Compact the sparse GSI space into a sequential IRQ series and reuse
-+ * vectors if possible.
-+ */
-+int gsi_irq_sharing(int gsi)
-+{
-+ int i, tries, vector;
++#
++# Character devices
++#
++CONFIG_VT=y
++CONFIG_VT_CONSOLE=y
++CONFIG_HW_CONSOLE=y
++CONFIG_VT_HW_CONSOLE_BINDING=y
++# CONFIG_SERIAL_NONSTANDARD is not set
+
-+ BUG_ON(gsi >= NR_IRQ_VECTORS);
++#
++# Serial drivers
++#
+
-+ if (platform_legacy_irq(gsi))
-+ return gsi;
++#
++# Non-8250 serial port support
++#
++# CONFIG_SERIAL_JSM is not set
++CONFIG_UNIX98_PTYS=y
++CONFIG_LEGACY_PTYS=y
++CONFIG_LEGACY_PTY_COUNT=64
++CONFIG_PRINTER=m
++CONFIG_LP_CONSOLE=y
++CONFIG_PPDEV=m
++CONFIG_TIPAR=m
+
-+ if (gsi_2_irq[gsi] != 0xFF)
-+ return (int)gsi_2_irq[gsi];
++#
++# IPMI
++#
++CONFIG_IPMI_HANDLER=m
++CONFIG_IPMI_PANIC_EVENT=y
++# CONFIG_IPMI_PANIC_STRING is not set
++CONFIG_IPMI_DEVICE_INTERFACE=m
++CONFIG_IPMI_SI=m
++CONFIG_IPMI_WATCHDOG=m
++CONFIG_IPMI_POWEROFF=m
+
-+ tries = NR_IRQS;
-+ try_again:
-+ vector = assign_irq_vector(gsi);
++#
++# Watchdog Cards
++#
++CONFIG_WATCHDOG=y
++# CONFIG_WATCHDOG_NOWAYOUT is not set
+
-+ /*
-+ * Sharing vectors means sharing IRQs, so scan irq_vectors for previous
-+ * use of vector and if found, return that IRQ. However, we never want
-+ * to share legacy IRQs, which usually have a different trigger mode
-+ * than PCI.
-+ */
-+ for (i = 0; i < NR_IRQS; i++)
-+ if (IO_APIC_VECTOR(i) == vector)
-+ break;
-+ if (platform_legacy_irq(i)) {
-+ if (--tries >= 0) {
-+ IO_APIC_VECTOR(i) = 0;
-+ goto try_again;
-+ }
-+ panic("gsi_irq_sharing: didn't find an IRQ using vector 0x%02X for GSI %d", vector, gsi);
-+ }
-+ if (i < NR_IRQS) {
-+ gsi_2_irq[gsi] = i;
-+ printk(KERN_INFO "GSI %d sharing vector 0x%02X and IRQ %d\n",
-+ gsi, vector, i);
-+ return i;
-+ }
++#
++# Watchdog Device Drivers
++#
++CONFIG_SOFT_WATCHDOG=m
++CONFIG_ACQUIRE_WDT=m
++CONFIG_ADVANTECH_WDT=m
++CONFIG_ALIM1535_WDT=m
++CONFIG_ALIM7101_WDT=m
++CONFIG_SC520_WDT=m
++CONFIG_EUROTECH_WDT=m
++CONFIG_IB700_WDT=m
++CONFIG_IBMASR=m
++CONFIG_WAFER_WDT=m
++CONFIG_I6300ESB_WDT=m
++CONFIG_I8XX_TCO=m
++CONFIG_SC1200_WDT=m
++CONFIG_60XX_WDT=m
++CONFIG_SBC8360_WDT=m
++CONFIG_CPU5_WDT=m
++CONFIG_W83627HF_WDT=m
++CONFIG_W83877F_WDT=m
++CONFIG_W83977F_WDT=m
++CONFIG_MACHZ_WDT=m
++CONFIG_SBC_EPX_C3_WATCHDOG=m
+
-+ i = next_irq++;
-+ BUG_ON(i >= NR_IRQS);
-+ gsi_2_irq[gsi] = i;
-+ IO_APIC_VECTOR(i) = vector;
-+ printk(KERN_INFO "GSI %d assigned vector 0x%02X and IRQ %d\n",
-+ gsi, vector, i);
-+ return i;
-+}
++#
++# PCI-based Watchdog Cards
++#
++CONFIG_PCIPCWATCHDOG=m
++CONFIG_WDTPCI=m
++CONFIG_WDT_501_PCI=y
+
-+static int pin_2_irq(int idx, int apic, int pin)
-+{
-+ int irq, i;
-+ int bus = mp_irqs[idx].mpc_srcbus;
++#
++# USB-based Watchdog Cards
++#
++CONFIG_USBPCWATCHDOG=m
++CONFIG_HW_RANDOM=y
++CONFIG_HW_RANDOM_INTEL=m
++CONFIG_HW_RANDOM_AMD=m
++CONFIG_HW_RANDOM_GEODE=m
++CONFIG_NVRAM=y
++CONFIG_RTC=y
++CONFIG_DTLK=m
++CONFIG_R3964=m
++CONFIG_APPLICOM=m
+
-+ /*
-+ * Debugging check, we are in big trouble if this message pops up!
-+ */
-+ if (mp_irqs[idx].mpc_dstirq != pin)
-+ printk(KERN_ERR "broken BIOS or MPTABLE parser, ayiee!!\n");
++#
++# Ftape, the floppy tape device driver
++#
++CONFIG_AGP=m
++CONFIG_AGP_AMD64=m
++CONFIG_AGP_INTEL=m
++CONFIG_AGP_SIS=m
++CONFIG_AGP_VIA=m
++CONFIG_DRM=m
++CONFIG_DRM_TDFX=m
++CONFIG_DRM_R128=m
++CONFIG_DRM_RADEON=m
++CONFIG_DRM_I810=m
++# CONFIG_DRM_I830 is not set
++CONFIG_DRM_I915=m
++CONFIG_DRM_MGA=m
++# CONFIG_DRM_SIS is not set
++CONFIG_DRM_VIA=m
++CONFIG_DRM_SAVAGE=m
+
-+ switch (mp_bus_id_to_type[bus])
-+ {
-+ case MP_BUS_ISA: /* ISA pin */
-+ case MP_BUS_EISA:
-+ case MP_BUS_MCA:
-+ {
-+ irq = mp_irqs[idx].mpc_srcbusirq;
-+ break;
-+ }
-+ case MP_BUS_PCI: /* PCI pin */
-+ {
-+ /*
-+ * PCI IRQs are mapped in order
-+ */
-+ i = irq = 0;
-+ while (i < apic)
-+ irq += nr_ioapic_registers[i++];
-+ irq += pin;
-+ irq = gsi_irq_sharing(irq);
-+ break;
-+ }
-+ default:
-+ {
-+ printk(KERN_ERR "unknown bus type %d.\n",bus);
-+ irq = 0;
-+ break;
-+ }
-+ }
-+ BUG_ON(irq >= NR_IRQS);
++#
++# PCMCIA character devices
++#
++# CONFIG_SYNCLINK_CS is not set
++# CONFIG_CARDMAN_4000 is not set
++# CONFIG_CARDMAN_4040 is not set
++# CONFIG_MWAVE is not set
++CONFIG_PC8736x_GPIO=m
++CONFIG_NSC_GPIO=m
++# CONFIG_RAW_DRIVER is not set
++# CONFIG_HPET is not set
++CONFIG_HANGCHECK_TIMER=m
+
-+ /*
-+ * PCI IRQ command line redirection. Yes, limits are hardcoded.
-+ */
-+ if ((pin >= 16) && (pin <= 23)) {
-+ if (pirq_entries[pin-16] != -1) {
-+ if (!pirq_entries[pin-16]) {
-+ apic_printk(APIC_VERBOSE, "disabling PIRQ%d\n", pin-16);
-+ } else {
-+ irq = pirq_entries[pin-16];
-+ apic_printk(APIC_VERBOSE, "using PIRQ%d -> IRQ %d\n",
-+ pin-16, irq);
-+ }
-+ }
-+ }
-+ BUG_ON(irq >= NR_IRQS);
-+ return irq;
-+}
++#
++# TPM devices
++#
++CONFIG_TCG_TPM=m
++CONFIG_TCG_TIS=m
++CONFIG_TCG_NSC=m
++CONFIG_TCG_ATMEL=m
++CONFIG_TCG_INFINEON=m
++CONFIG_TCG_XEN=m
++CONFIG_TELCLOCK=m
+
-+static inline int IO_APIC_irq_trigger(int irq)
-+{
-+ int apic, idx, pin;
++#
++# I2C support
++#
++CONFIG_I2C=m
++CONFIG_I2C_CHARDEV=m
+
-+ for (apic = 0; apic < nr_ioapics; apic++) {
-+ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
-+ idx = find_irq_entry(apic,pin,mp_INT);
-+ if ((idx != -1) && (irq == pin_2_irq(idx,apic,pin)))
-+ return irq_trigger(idx);
-+ }
-+ }
-+ /*
-+ * nonexistent IRQs are edge default
-+ */
-+ return 0;
-+}
++#
++# I2C Algorithms
++#
++CONFIG_I2C_ALGOBIT=m
++CONFIG_I2C_ALGOPCF=m
++CONFIG_I2C_ALGOPCA=m
+
-+/* irq_vectors is indexed by the sum of all RTEs in all I/O APICs. */
-+u8 irq_vector[NR_IRQ_VECTORS] __read_mostly;
++#
++# I2C Hardware Bus support
++#
++CONFIG_I2C_ALI1535=m
++CONFIG_I2C_ALI1563=m
++CONFIG_I2C_ALI15X3=m
++CONFIG_I2C_AMD756=m
++CONFIG_I2C_AMD756_S4882=m
++CONFIG_I2C_AMD8111=m
++CONFIG_I2C_I801=m
++CONFIG_I2C_I810=m
++CONFIG_I2C_PIIX4=m
++CONFIG_I2C_ISA=m
++CONFIG_I2C_NFORCE2=m
++CONFIG_I2C_OCORES=m
++CONFIG_I2C_PARPORT=m
++CONFIG_I2C_PARPORT_LIGHT=m
++CONFIG_I2C_PROSAVAGE=m
++CONFIG_I2C_SAVAGE4=m
++CONFIG_I2C_SIS5595=m
++CONFIG_I2C_SIS630=m
++CONFIG_I2C_SIS96X=m
++CONFIG_I2C_STUB=m
++CONFIG_I2C_VIA=m
++CONFIG_I2C_VIAPRO=m
++CONFIG_I2C_VOODOO3=m
++CONFIG_I2C_PCA_ISA=m
+
-+int assign_irq_vector(int irq)
-+{
-+ unsigned long flags;
-+ int vector;
-+ struct physdev_irq irq_op;
-+
-+ BUG_ON(irq != AUTO_ASSIGN && (unsigned)irq >= NR_IRQ_VECTORS);
++#
++# Miscellaneous I2C Chip support
++#
++CONFIG_SENSORS_DS1337=m
++CONFIG_SENSORS_DS1374=m
++CONFIG_SENSORS_EEPROM=m
++CONFIG_SENSORS_PCF8574=m
++CONFIG_SENSORS_PCA9539=m
++CONFIG_SENSORS_PCF8591=m
++CONFIG_SENSORS_MAX6875=m
++# CONFIG_I2C_DEBUG_CORE is not set
++# CONFIG_I2C_DEBUG_ALGO is not set
++# CONFIG_I2C_DEBUG_BUS is not set
++# CONFIG_I2C_DEBUG_CHIP is not set
+
-+ spin_lock_irqsave(&vector_lock, flags);
++#
++# SPI support
++#
++CONFIG_SPI=y
++# CONFIG_SPI_DEBUG is not set
++CONFIG_SPI_MASTER=y
+
-+ if (irq != AUTO_ASSIGN && IO_APIC_VECTOR(irq) > 0) {
-+ spin_unlock_irqrestore(&vector_lock, flags);
-+ return IO_APIC_VECTOR(irq);
-+ }
++#
++# SPI Master Controller Drivers
++#
++CONFIG_SPI_BITBANG=m
++CONFIG_SPI_BUTTERFLY=m
+
-+ irq_op.irq = irq;
-+ if (HYPERVISOR_physdev_op(PHYSDEVOP_alloc_irq_vector, &irq_op)) {
-+ spin_unlock_irqrestore(&vector_lock, flags);
-+ return -ENOSPC;
-+ }
++#
++# SPI Protocol Masters
++#
+
-+ vector = irq_op.vector;
-+ vector_irq[vector] = irq;
-+ if (irq != AUTO_ASSIGN)
-+ IO_APIC_VECTOR(irq) = vector;
++#
++# Dallas's 1-wire bus
++#
++CONFIG_W1=m
++CONFIG_W1_CON=y
+
-+ spin_unlock_irqrestore(&vector_lock, flags);
++#
++# 1-wire Bus Masters
++#
++CONFIG_W1_MASTER_MATROX=m
++CONFIG_W1_MASTER_DS2490=m
++CONFIG_W1_MASTER_DS2482=m
+
-+ return vector;
-+}
++#
++# 1-wire Slaves
++#
++CONFIG_W1_SLAVE_THERM=m
++CONFIG_W1_SLAVE_SMEM=m
++CONFIG_W1_SLAVE_DS2433=m
++# CONFIG_W1_SLAVE_DS2433_CRC is not set
+
-+extern void (*interrupt[NR_IRQS])(void);
-+#ifndef CONFIG_XEN
-+static struct hw_interrupt_type ioapic_level_type;
-+static struct hw_interrupt_type ioapic_edge_type;
++#
++# Hardware Monitoring support
++#
++CONFIG_HWMON=m
++CONFIG_HWMON_VID=m
++CONFIG_SENSORS_ABITUGURU=m
++CONFIG_SENSORS_ADM1021=m
++CONFIG_SENSORS_ADM1025=m
++CONFIG_SENSORS_ADM1026=m
++CONFIG_SENSORS_ADM1031=m
++CONFIG_SENSORS_ADM9240=m
++CONFIG_SENSORS_ASB100=m
++CONFIG_SENSORS_ATXP1=m
++CONFIG_SENSORS_DS1621=m
++CONFIG_SENSORS_F71805F=m
++CONFIG_SENSORS_FSCHER=m
++CONFIG_SENSORS_FSCPOS=m
++CONFIG_SENSORS_GL518SM=m
++CONFIG_SENSORS_GL520SM=m
++CONFIG_SENSORS_IT87=m
++CONFIG_SENSORS_LM63=m
++CONFIG_SENSORS_LM70=m
++CONFIG_SENSORS_LM75=m
++CONFIG_SENSORS_LM77=m
++CONFIG_SENSORS_LM78=m
++CONFIG_SENSORS_LM80=m
++CONFIG_SENSORS_LM83=m
++CONFIG_SENSORS_LM85=m
++CONFIG_SENSORS_LM87=m
++CONFIG_SENSORS_LM90=m
++CONFIG_SENSORS_LM92=m
++CONFIG_SENSORS_MAX1619=m
++CONFIG_SENSORS_PC87360=m
++CONFIG_SENSORS_SIS5595=m
++CONFIG_SENSORS_SMSC47M1=m
++CONFIG_SENSORS_SMSC47M192=m
++CONFIG_SENSORS_SMSC47B397=m
++CONFIG_SENSORS_VIA686A=m
++CONFIG_SENSORS_VT8231=m
++CONFIG_SENSORS_W83781D=m
++CONFIG_SENSORS_W83791D=m
++CONFIG_SENSORS_W83792D=m
++CONFIG_SENSORS_W83L785TS=m
++CONFIG_SENSORS_W83627HF=m
++CONFIG_SENSORS_W83627EHF=m
++CONFIG_SENSORS_HDAPS=m
++# CONFIG_HWMON_DEBUG_CHIP is not set
+
-+#define IOAPIC_AUTO -1
-+#define IOAPIC_EDGE 0
-+#define IOAPIC_LEVEL 1
++#
++# Misc devices
++#
++CONFIG_IBM_ASM=m
+
-+static void ioapic_register_intr(int irq, int vector, unsigned long trigger)
-+{
-+ unsigned idx;
++#
++# Multimedia devices
++#
++CONFIG_VIDEO_DEV=m
++CONFIG_VIDEO_V4L1=y
++CONFIG_VIDEO_V4L1_COMPAT=y
++CONFIG_VIDEO_V4L2=y
+
-+ idx = use_pci_vector() && !platform_legacy_irq(irq) ? vector : irq;
++#
++# Video Capture Adapters
++#
+
-+ if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
-+ trigger == IOAPIC_LEVEL)
-+ irq_desc[idx].chip = &ioapic_level_type;
-+ else
-+ irq_desc[idx].chip = &ioapic_edge_type;
-+ set_intr_gate(vector, interrupt[idx]);
-+}
-+#else
-+#define ioapic_register_intr(_irq,_vector,_trigger) ((void)0)
-+#endif /* !CONFIG_XEN */
++#
++# Video Capture Adapters
++#
++# CONFIG_VIDEO_ADV_DEBUG is not set
++CONFIG_VIDEO_VIVI=m
++CONFIG_VIDEO_BT848=m
++CONFIG_VIDEO_BT848_DVB=y
++CONFIG_VIDEO_SAA6588=m
++CONFIG_VIDEO_BWQCAM=m
++CONFIG_VIDEO_CQCAM=m
++CONFIG_VIDEO_W9966=m
++CONFIG_VIDEO_CPIA=m
++CONFIG_VIDEO_CPIA_PP=m
++CONFIG_VIDEO_CPIA_USB=m
++CONFIG_VIDEO_CPIA2=m
++CONFIG_VIDEO_SAA5246A=m
++CONFIG_VIDEO_SAA5249=m
++CONFIG_TUNER_3036=m
++CONFIG_VIDEO_STRADIS=m
++CONFIG_VIDEO_ZORAN=m
++CONFIG_VIDEO_ZORAN_BUZ=m
++CONFIG_VIDEO_ZORAN_DC10=m
++CONFIG_VIDEO_ZORAN_DC30=m
++CONFIG_VIDEO_ZORAN_LML33=m
++CONFIG_VIDEO_ZORAN_LML33R10=m
++CONFIG_VIDEO_ZORAN_AVS6EYES=m
++CONFIG_VIDEO_SAA7134=m
++CONFIG_VIDEO_SAA7134_ALSA=m
++CONFIG_VIDEO_SAA7134_DVB=m
++CONFIG_VIDEO_SAA7134_DVB_ALL_FRONTENDS=y
++CONFIG_VIDEO_MXB=m
++CONFIG_VIDEO_DPC=m
++CONFIG_VIDEO_HEXIUM_ORION=m
++CONFIG_VIDEO_HEXIUM_GEMINI=m
++CONFIG_VIDEO_CX88_VP3054=m
++CONFIG_VIDEO_CX88=m
++CONFIG_VIDEO_CX88_ALSA=m
++CONFIG_VIDEO_CX88_BLACKBIRD=m
++CONFIG_VIDEO_CX88_DVB=m
++CONFIG_VIDEO_CX88_DVB_ALL_FRONTENDS=y
+
-+static void __init setup_IO_APIC_irqs(void)
-+{
-+ struct IO_APIC_route_entry entry;
-+ int apic, pin, idx, irq, first_notcon = 1, vector;
-+ unsigned long flags;
++#
++# Encoders and Decoders
++#
++CONFIG_VIDEO_MSP3400=m
++CONFIG_VIDEO_CS53L32A=m
++CONFIG_VIDEO_TLV320AIC23B=m
++CONFIG_VIDEO_WM8775=m
++CONFIG_VIDEO_WM8739=m
++CONFIG_VIDEO_CX2341X=m
++CONFIG_VIDEO_CX25840=m
++CONFIG_VIDEO_SAA711X=m
++CONFIG_VIDEO_SAA7127=m
++CONFIG_VIDEO_UPD64031A=m
++CONFIG_VIDEO_UPD64083=m
+
-+ apic_printk(APIC_VERBOSE, KERN_DEBUG "init IO_APIC IRQs\n");
++#
++# V4L USB devices
++#
++CONFIG_VIDEO_PVRUSB2=m
++CONFIG_VIDEO_PVRUSB2_24XXX=y
++CONFIG_VIDEO_PVRUSB2_SYSFS=y
++# CONFIG_VIDEO_PVRUSB2_DEBUGIFC is not set
++CONFIG_VIDEO_EM28XX=m
++CONFIG_VIDEO_USBVIDEO=m
++CONFIG_USB_VICAM=m
++CONFIG_USB_IBMCAM=m
++CONFIG_USB_KONICAWC=m
++CONFIG_USB_QUICKCAM_MESSENGER=m
++CONFIG_USB_ET61X251=m
++CONFIG_VIDEO_OVCAMCHIP=m
++CONFIG_USB_W9968CF=m
++CONFIG_USB_OV511=m
++CONFIG_USB_SE401=m
++CONFIG_USB_SN9C102=m
++CONFIG_USB_STV680=m
++CONFIG_USB_ZC0301=m
++CONFIG_USB_PWC=m
++# CONFIG_USB_PWC_DEBUG is not set
+
-+ for (apic = 0; apic < nr_ioapics; apic++) {
-+ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
++#
++# Radio Adapters
++#
++CONFIG_RADIO_GEMTEK_PCI=m
++CONFIG_RADIO_MAXIRADIO=m
++CONFIG_RADIO_MAESTRO=m
++CONFIG_USB_DSBR=m
+
-+ /*
-+ * add it to the IO-APIC irq-routing table:
-+ */
-+ memset(&entry,0,sizeof(entry));
++#
++# Digital Video Broadcasting Devices
++#
++CONFIG_DVB=y
++CONFIG_DVB_CORE=m
+
-+ entry.delivery_mode = INT_DELIVERY_MODE;
-+ entry.dest_mode = INT_DEST_MODE;
-+ entry.mask = 0; /* enable IRQ */
-+ entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS);
++#
++# Supported SAA7146 based PCI Adapters
++#
++CONFIG_DVB_AV7110=m
++CONFIG_DVB_AV7110_OSD=y
++CONFIG_DVB_BUDGET=m
++CONFIG_DVB_BUDGET_CI=m
++CONFIG_DVB_BUDGET_AV=m
++CONFIG_DVB_BUDGET_PATCH=m
+
-+ idx = find_irq_entry(apic,pin,mp_INT);
-+ if (idx == -1) {
-+ if (first_notcon) {
-+ apic_printk(APIC_VERBOSE, KERN_DEBUG " IO-APIC (apicid-pin) %d-%d", mp_ioapics[apic].mpc_apicid, pin);
-+ first_notcon = 0;
-+ } else
-+ apic_printk(APIC_VERBOSE, ", %d-%d", mp_ioapics[apic].mpc_apicid, pin);
-+ continue;
-+ }
++#
++# Supported USB Adapters
++#
++CONFIG_DVB_USB=m
++# CONFIG_DVB_USB_DEBUG is not set
++CONFIG_DVB_USB_A800=m
++CONFIG_DVB_USB_DIBUSB_MB=m
++# CONFIG_DVB_USB_DIBUSB_MB_FAULTY is not set
++CONFIG_DVB_USB_DIBUSB_MC=m
++CONFIG_DVB_USB_UMT_010=m
++CONFIG_DVB_USB_CXUSB=m
++CONFIG_DVB_USB_DIGITV=m
++CONFIG_DVB_USB_VP7045=m
++CONFIG_DVB_USB_VP702X=m
++CONFIG_DVB_USB_GP8PSK=m
++CONFIG_DVB_USB_NOVA_T_USB2=m
++CONFIG_DVB_USB_DTT200U=m
++CONFIG_DVB_TTUSB_BUDGET=m
++CONFIG_DVB_TTUSB_DEC=m
++CONFIG_DVB_CINERGYT2=m
++CONFIG_DVB_CINERGYT2_TUNING=y
++CONFIG_DVB_CINERGYT2_STREAM_URB_COUNT=32
++CONFIG_DVB_CINERGYT2_STREAM_BUF_SIZE=512
++CONFIG_DVB_CINERGYT2_QUERY_INTERVAL=250
++CONFIG_DVB_CINERGYT2_ENABLE_RC_INPUT_DEVICE=y
++CONFIG_DVB_CINERGYT2_RC_QUERY_INTERVAL=100
+
-+ entry.trigger = irq_trigger(idx);
-+ entry.polarity = irq_polarity(idx);
++#
++# Supported FlexCopII (B2C2) Adapters
++#
++CONFIG_DVB_B2C2_FLEXCOP=m
++CONFIG_DVB_B2C2_FLEXCOP_PCI=m
++CONFIG_DVB_B2C2_FLEXCOP_USB=m
++# CONFIG_DVB_B2C2_FLEXCOP_DEBUG is not set
+
-+ if (irq_trigger(idx)) {
-+ entry.trigger = 1;
-+ entry.mask = 1;
-+ entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS);
-+ }
++#
++# Supported BT878 Adapters
++#
++CONFIG_DVB_BT8XX=m
+
-+ irq = pin_2_irq(idx, apic, pin);
-+ add_pin_to_irq(irq, apic, pin);
++#
++# Supported Pluto2 Adapters
++#
++CONFIG_DVB_PLUTO2=m
+
-+ if (/* !apic && */ !IO_APIC_IRQ(irq))
-+ continue;
++#
++# Supported DVB Frontends
++#
+
-+ if (IO_APIC_IRQ(irq)) {
-+ vector = assign_irq_vector(irq);
-+ entry.vector = vector;
++#
++# Customise DVB Frontends
++#
+
-+ ioapic_register_intr(irq, vector, IOAPIC_AUTO);
-+ if (!apic && (irq < 16))
-+ disable_8259A_irq(irq);
-+ }
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ io_apic_write(apic, 0x11+2*pin, *(((int *)&entry)+1));
-+ io_apic_write(apic, 0x10+2*pin, *(((int *)&entry)+0));
-+ set_native_irq_info(irq, TARGET_CPUS);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+ }
-+ }
++#
++# DVB-S (satellite) frontends
++#
++CONFIG_DVB_STV0299=m
++CONFIG_DVB_CX24110=m
++CONFIG_DVB_CX24123=m
++CONFIG_DVB_TDA8083=m
++CONFIG_DVB_MT312=m
++CONFIG_DVB_VES1X93=m
++CONFIG_DVB_S5H1420=m
+
-+ if (!first_notcon)
-+ apic_printk(APIC_VERBOSE," not connected.\n");
-+}
++#
++# DVB-T (terrestrial) frontends
++#
++CONFIG_DVB_SP8870=m
++CONFIG_DVB_SP887X=m
++CONFIG_DVB_CX22700=m
++CONFIG_DVB_CX22702=m
++CONFIG_DVB_L64781=m
++CONFIG_DVB_TDA1004X=m
++CONFIG_DVB_NXT6000=m
++CONFIG_DVB_MT352=m
++CONFIG_DVB_ZL10353=m
++CONFIG_DVB_DIB3000MB=m
++CONFIG_DVB_DIB3000MC=m
+
-+#ifndef CONFIG_XEN
-+/*
-+ * Set up the 8259A-master output pin as broadcast to all
-+ * CPUs.
-+ */
-+static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, int vector)
-+{
-+ struct IO_APIC_route_entry entry;
-+ unsigned long flags;
++#
++# DVB-C (cable) frontends
++#
++CONFIG_DVB_VES1820=m
++CONFIG_DVB_TDA10021=m
++CONFIG_DVB_STV0297=m
+
-+ memset(&entry,0,sizeof(entry));
++#
++# ATSC (North American/Korean Terrestrial/Cable DTV) frontends
++#
++CONFIG_DVB_NXT200X=m
++CONFIG_DVB_OR51211=m
++CONFIG_DVB_OR51132=m
++CONFIG_DVB_BCM3510=m
++CONFIG_DVB_LGDT330X=m
+
-+ disable_8259A_irq(0);
++#
++# Miscellaneous devices
++#
++CONFIG_DVB_PLL=m
++CONFIG_DVB_LNBP21=m
++CONFIG_DVB_ISL6421=m
++CONFIG_VIDEO_SAA7146=m
++CONFIG_VIDEO_SAA7146_VV=m
++CONFIG_VIDEO_VIDEOBUF=m
++CONFIG_VIDEO_TUNER=m
++CONFIG_VIDEO_BUF=m
++CONFIG_VIDEO_BUF_DVB=m
++CONFIG_VIDEO_BTCX=m
++CONFIG_VIDEO_IR=m
++CONFIG_VIDEO_TVEEPROM=m
++CONFIG_USB_DABUSB=m
+
-+ /* mask LVT0 */
-+ apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
++#
++# Graphics support
++#
++CONFIG_FIRMWARE_EDID=y
++CONFIG_FB=y
++CONFIG_FB_CFB_FILLRECT=y
++CONFIG_FB_CFB_COPYAREA=y
++CONFIG_FB_CFB_IMAGEBLIT=y
++# CONFIG_FB_MACMODES is not set
++# CONFIG_FB_BACKLIGHT is not set
++CONFIG_FB_MODE_HELPERS=y
++CONFIG_FB_TILEBLITTING=y
++CONFIG_FB_CIRRUS=m
++# CONFIG_FB_PM2 is not set
++CONFIG_FB_CYBER2000=m
++CONFIG_FB_ARC=m
++# CONFIG_FB_ASILIANT is not set
++# CONFIG_FB_IMSTT is not set
++CONFIG_FB_VGA16=m
++CONFIG_FB_VESA=y
++CONFIG_FB_HGA=m
++CONFIG_FB_HGA_ACCEL=y
++CONFIG_FB_S1D13XXX=m
++CONFIG_FB_NVIDIA=m
++CONFIG_FB_NVIDIA_I2C=y
++CONFIG_FB_RIVA=m
++CONFIG_FB_RIVA_I2C=y
++# CONFIG_FB_RIVA_DEBUG is not set
++CONFIG_FB_INTEL=m
++# CONFIG_FB_INTEL_DEBUG is not set
++CONFIG_FB_MATROX=m
++CONFIG_FB_MATROX_MILLENIUM=y
++CONFIG_FB_MATROX_MYSTIQUE=y
++CONFIG_FB_MATROX_G=y
++CONFIG_FB_MATROX_I2C=m
++CONFIG_FB_MATROX_MAVEN=m
++CONFIG_FB_MATROX_MULTIHEAD=y
++CONFIG_FB_RADEON=m
++CONFIG_FB_RADEON_I2C=y
++# CONFIG_FB_RADEON_DEBUG is not set
++CONFIG_FB_ATY128=m
++CONFIG_FB_ATY=m
++CONFIG_FB_ATY_CT=y
++CONFIG_FB_ATY_GENERIC_LCD=y
++CONFIG_FB_ATY_GX=y
++CONFIG_FB_SAVAGE=m
++CONFIG_FB_SAVAGE_I2C=y
++CONFIG_FB_SAVAGE_ACCEL=y
++CONFIG_FB_SIS=m
++CONFIG_FB_SIS_300=y
++CONFIG_FB_SIS_315=y
++CONFIG_FB_NEOMAGIC=m
++CONFIG_FB_KYRO=m
++CONFIG_FB_3DFX=m
++CONFIG_FB_3DFX_ACCEL=y
++CONFIG_FB_VOODOO1=m
++CONFIG_FB_TRIDENT=m
++CONFIG_FB_TRIDENT_ACCEL=y
++CONFIG_FB_GEODE=y
++CONFIG_FB_GEODE_GX=m
++CONFIG_FB_GEODE_GX1=m
++CONFIG_FB_VIRTUAL=m
++
++#
++# Console display driver support
++#
++CONFIG_VGA_CONSOLE=y
++# CONFIG_VGACON_SOFT_SCROLLBACK is not set
++CONFIG_DUMMY_CONSOLE=y
++CONFIG_FRAMEBUFFER_CONSOLE=y
++CONFIG_FRAMEBUFFER_CONSOLE_ROTATION=y
++# CONFIG_FONTS is not set
++CONFIG_FONT_8x8=y
++CONFIG_FONT_8x16=y
+
-+ /*
-+ * We use logical delivery to get the timer IRQ
-+ * to the first CPU.
-+ */
-+ entry.dest_mode = INT_DEST_MODE;
-+ entry.mask = 0; /* unmask IRQ now */
-+ entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS);
-+ entry.delivery_mode = INT_DELIVERY_MODE;
-+ entry.polarity = 0;
-+ entry.trigger = 0;
-+ entry.vector = vector;
++#
++# Logo configuration
++#
++CONFIG_LOGO=y
++# CONFIG_LOGO_LINUX_MONO is not set
++# CONFIG_LOGO_LINUX_VGA16 is not set
++CONFIG_LOGO_LINUX_CLUT224=y
++CONFIG_BACKLIGHT_LCD_SUPPORT=y
++CONFIG_BACKLIGHT_CLASS_DEVICE=m
++CONFIG_BACKLIGHT_DEVICE=y
++CONFIG_LCD_CLASS_DEVICE=m
++CONFIG_LCD_DEVICE=y
+
-+ /*
-+ * The timer IRQ doesn't have to know that behind the
-+ * scene we have a 8259A-master in AEOI mode ...
-+ */
-+ irq_desc[0].chip = &ioapic_edge_type;
++#
++# Sound
++#
++CONFIG_SOUND=m
+
-+ /*
-+ * Add it to the IO-APIC irq-routing table:
-+ */
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ io_apic_write(apic, 0x11+2*pin, *(((int *)&entry)+1));
-+ io_apic_write(apic, 0x10+2*pin, *(((int *)&entry)+0));
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
++#
++# Advanced Linux Sound Architecture
++#
++CONFIG_SND=m
++CONFIG_SND_TIMER=m
++CONFIG_SND_PCM=m
++CONFIG_SND_HWDEP=m
++CONFIG_SND_RAWMIDI=m
++CONFIG_SND_SEQUENCER=m
++CONFIG_SND_SEQ_DUMMY=m
++CONFIG_SND_OSSEMUL=y
++CONFIG_SND_MIXER_OSS=m
++CONFIG_SND_PCM_OSS=m
++CONFIG_SND_PCM_OSS_PLUGINS=y
++CONFIG_SND_SEQUENCER_OSS=y
++CONFIG_SND_RTCTIMER=m
++CONFIG_SND_SEQ_RTCTIMER_DEFAULT=y
++CONFIG_SND_DYNAMIC_MINORS=y
++CONFIG_SND_SUPPORT_OLD_API=y
++CONFIG_SND_VERBOSE_PROCFS=y
++# CONFIG_SND_VERBOSE_PRINTK is not set
++# CONFIG_SND_DEBUG is not set
+
-+ enable_8259A_irq(0);
-+}
++#
++# Generic devices
++#
++CONFIG_SND_MPU401_UART=m
++CONFIG_SND_OPL3_LIB=m
++CONFIG_SND_VX_LIB=m
++CONFIG_SND_AC97_CODEC=m
++CONFIG_SND_AC97_BUS=m
++CONFIG_SND_DUMMY=m
++CONFIG_SND_VIRMIDI=m
++CONFIG_SND_MTPAV=m
++CONFIG_SND_SERIAL_U16550=m
++CONFIG_SND_MPU401=m
+
-+void __init UNEXPECTED_IO_APIC(void)
-+{
-+}
++#
++# PCI devices
++#
++CONFIG_SND_AD1889=m
++CONFIG_SND_ALS300=m
++CONFIG_SND_ALS4000=m
++CONFIG_SND_ALI5451=m
++CONFIG_SND_ATIIXP=m
++CONFIG_SND_ATIIXP_MODEM=m
++CONFIG_SND_AU8810=m
++CONFIG_SND_AU8820=m
++CONFIG_SND_AU8830=m
++CONFIG_SND_AZT3328=m
++CONFIG_SND_BT87X=m
++# CONFIG_SND_BT87X_OVERCLOCK is not set
++CONFIG_SND_CA0106=m
++CONFIG_SND_CMIPCI=m
++CONFIG_SND_CS4281=m
++CONFIG_SND_CS46XX=m
++CONFIG_SND_CS46XX_NEW_DSP=y
++CONFIG_SND_DARLA20=m
++CONFIG_SND_GINA20=m
++CONFIG_SND_LAYLA20=m
++CONFIG_SND_DARLA24=m
++CONFIG_SND_GINA24=m
++CONFIG_SND_LAYLA24=m
++CONFIG_SND_MONA=m
++CONFIG_SND_MIA=m
++CONFIG_SND_ECHO3G=m
++CONFIG_SND_INDIGO=m
++CONFIG_SND_INDIGOIO=m
++CONFIG_SND_INDIGODJ=m
++CONFIG_SND_EMU10K1=m
++CONFIG_SND_EMU10K1X=m
++CONFIG_SND_ENS1370=m
++CONFIG_SND_ENS1371=m
++CONFIG_SND_ES1938=m
++CONFIG_SND_ES1968=m
++CONFIG_SND_FM801=m
++CONFIG_SND_FM801_TEA575X_BOOL=y
++CONFIG_SND_FM801_TEA575X=m
++CONFIG_SND_HDA_INTEL=m
++CONFIG_SND_HDSP=m
++CONFIG_SND_HDSPM=m
++CONFIG_SND_ICE1712=m
++CONFIG_SND_ICE1724=m
++CONFIG_SND_INTEL8X0=m
++CONFIG_SND_INTEL8X0M=m
++CONFIG_SND_KORG1212=m
++CONFIG_SND_MAESTRO3=m
++CONFIG_SND_MIXART=m
++CONFIG_SND_NM256=m
++CONFIG_SND_PCXHR=m
++CONFIG_SND_RIPTIDE=m
++CONFIG_SND_RME32=m
++CONFIG_SND_RME96=m
++CONFIG_SND_RME9652=m
++CONFIG_SND_SONICVIBES=m
++CONFIG_SND_TRIDENT=m
++CONFIG_SND_VIA82XX=m
++CONFIG_SND_VIA82XX_MODEM=m
++CONFIG_SND_VX222=m
++CONFIG_SND_YMFPCI=m
+
-+void __apicdebuginit print_IO_APIC(void)
-+{
-+ int apic, i;
-+ union IO_APIC_reg_00 reg_00;
-+ union IO_APIC_reg_01 reg_01;
-+ union IO_APIC_reg_02 reg_02;
-+ unsigned long flags;
++#
++# USB devices
++#
++CONFIG_SND_USB_AUDIO=m
++CONFIG_SND_USB_USX2Y=m
+
-+ if (apic_verbosity == APIC_QUIET)
-+ return;
++#
++# PCMCIA devices
++#
++CONFIG_SND_VXPOCKET=m
++CONFIG_SND_PDAUDIOCF=m
+
-+ printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries);
-+ for (i = 0; i < nr_ioapics; i++)
-+ printk(KERN_DEBUG "number of IO-APIC #%d registers: %d.\n",
-+ mp_ioapics[i].mpc_apicid, nr_ioapic_registers[i]);
++#
++# Open Sound System
++#
++# CONFIG_SOUND_PRIME is not set
+
-+ /*
-+ * We are a bit conservative about what we expect. We have to
-+ * know about every hardware change ASAP.
-+ */
-+ printk(KERN_INFO "testing the IO APIC.......................\n");
++#
++# USB support
++#
++CONFIG_USB_ARCH_HAS_HCD=y
++CONFIG_USB_ARCH_HAS_OHCI=y
++CONFIG_USB_ARCH_HAS_EHCI=y
++CONFIG_USB=m
++# CONFIG_USB_DEBUG is not set
+
-+ for (apic = 0; apic < nr_ioapics; apic++) {
++#
++# Miscellaneous USB options
++#
++CONFIG_USB_DEVICEFS=y
++# CONFIG_USB_BANDWIDTH is not set
++# CONFIG_USB_DYNAMIC_MINORS is not set
++# CONFIG_USB_SUSPEND is not set
++# CONFIG_USB_OTG is not set
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ reg_00.raw = io_apic_read(apic, 0);
-+ reg_01.raw = io_apic_read(apic, 1);
-+ if (reg_01.bits.version >= 0x10)
-+ reg_02.raw = io_apic_read(apic, 2);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
++#
++# USB Host Controller Drivers
++#
++CONFIG_USB_EHCI_HCD=m
++CONFIG_USB_EHCI_SPLIT_ISO=y
++CONFIG_USB_EHCI_ROOT_HUB_TT=y
++CONFIG_USB_EHCI_TT_NEWSCHED=y
++CONFIG_USB_ISP116X_HCD=m
++CONFIG_USB_OHCI_HCD=m
++# CONFIG_USB_OHCI_BIG_ENDIAN is not set
++CONFIG_USB_OHCI_LITTLE_ENDIAN=y
++CONFIG_USB_UHCI_HCD=m
++CONFIG_USB_SL811_HCD=m
++# CONFIG_USB_SL811_CS is not set
+
-+ printk("\n");
-+ printk(KERN_DEBUG "IO APIC #%d......\n", mp_ioapics[apic].mpc_apicid);
-+ printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw);
-+ printk(KERN_DEBUG "....... : physical APIC id: %02X\n", reg_00.bits.ID);
-+ if (reg_00.bits.__reserved_1 || reg_00.bits.__reserved_2)
-+ UNEXPECTED_IO_APIC();
++#
++# USB Device Class drivers
++#
++CONFIG_USB_ACM=m
++CONFIG_USB_PRINTER=m
+
-+ printk(KERN_DEBUG ".... register #01: %08X\n", *(int *)®_01);
-+ printk(KERN_DEBUG "....... : max redirection entries: %04X\n", reg_01.bits.entries);
-+ if ( (reg_01.bits.entries != 0x0f) && /* older (Neptune) boards */
-+ (reg_01.bits.entries != 0x17) && /* typical ISA+PCI boards */
-+ (reg_01.bits.entries != 0x1b) && /* Compaq Proliant boards */
-+ (reg_01.bits.entries != 0x1f) && /* dual Xeon boards */
-+ (reg_01.bits.entries != 0x22) && /* bigger Xeon boards */
-+ (reg_01.bits.entries != 0x2E) &&
-+ (reg_01.bits.entries != 0x3F) &&
-+ (reg_01.bits.entries != 0x03)
-+ )
-+ UNEXPECTED_IO_APIC();
++#
++# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support'
++#
+
-+ printk(KERN_DEBUG "....... : PRQ implemented: %X\n", reg_01.bits.PRQ);
-+ printk(KERN_DEBUG "....... : IO APIC version: %04X\n", reg_01.bits.version);
-+ if ( (reg_01.bits.version != 0x01) && /* 82489DX IO-APICs */
-+ (reg_01.bits.version != 0x02) && /* 82801BA IO-APICs (ICH2) */
-+ (reg_01.bits.version != 0x10) && /* oldest IO-APICs */
-+ (reg_01.bits.version != 0x11) && /* Pentium/Pro IO-APICs */
-+ (reg_01.bits.version != 0x13) && /* Xeon IO-APICs */
-+ (reg_01.bits.version != 0x20) /* Intel P64H (82806 AA) */
-+ )
-+ UNEXPECTED_IO_APIC();
-+ if (reg_01.bits.__reserved_1 || reg_01.bits.__reserved_2)
-+ UNEXPECTED_IO_APIC();
++#
++# may also be needed; see USB_STORAGE Help for more information
++#
++CONFIG_USB_STORAGE=m
++# CONFIG_USB_STORAGE_DEBUG is not set
++CONFIG_USB_STORAGE_DATAFAB=y
++CONFIG_USB_STORAGE_FREECOM=y
++CONFIG_USB_STORAGE_ISD200=y
++CONFIG_USB_STORAGE_DPCM=y
++CONFIG_USB_STORAGE_USBAT=y
++CONFIG_USB_STORAGE_SDDR09=y
++CONFIG_USB_STORAGE_SDDR55=y
++CONFIG_USB_STORAGE_JUMPSHOT=y
++# CONFIG_USB_STORAGE_ALAUDA is not set
++# CONFIG_USB_LIBUSUAL is not set
+
-+ if (reg_01.bits.version >= 0x10) {
-+ printk(KERN_DEBUG ".... register #02: %08X\n", reg_02.raw);
-+ printk(KERN_DEBUG "....... : arbitration: %02X\n", reg_02.bits.arbitration);
-+ if (reg_02.bits.__reserved_1 || reg_02.bits.__reserved_2)
-+ UNEXPECTED_IO_APIC();
-+ }
++#
++# USB Input Devices
++#
++CONFIG_USB_HID=m
++CONFIG_USB_HIDINPUT=y
++# CONFIG_USB_HIDINPUT_POWERBOOK is not set
++CONFIG_HID_FF=y
++CONFIG_HID_PID=y
++CONFIG_LOGITECH_FF=y
++CONFIG_THRUSTMASTER_FF=y
++CONFIG_USB_HIDDEV=y
+
-+ printk(KERN_DEBUG ".... IRQ redirection table:\n");
++#
++# USB HID Boot Protocol drivers
++#
++CONFIG_USB_KBD=m
++CONFIG_USB_MOUSE=m
++CONFIG_USB_AIPTEK=m
++CONFIG_USB_WACOM=m
++CONFIG_USB_ACECAD=m
++CONFIG_USB_KBTAB=m
++CONFIG_USB_POWERMATE=m
++CONFIG_USB_TOUCHSCREEN=m
++CONFIG_USB_TOUCHSCREEN_EGALAX=y
++CONFIG_USB_TOUCHSCREEN_PANJIT=y
++CONFIG_USB_TOUCHSCREEN_3M=y
++CONFIG_USB_TOUCHSCREEN_ITM=y
++CONFIG_USB_YEALINK=m
++CONFIG_USB_XPAD=m
++CONFIG_USB_ATI_REMOTE=m
++CONFIG_USB_ATI_REMOTE2=m
++CONFIG_USB_KEYSPAN_REMOTE=m
++CONFIG_USB_APPLETOUCH=m
+
-+ printk(KERN_DEBUG " NR Log Phy Mask Trig IRR Pol"
-+ " Stat Dest Deli Vect: \n");
++#
++# USB Imaging devices
++#
++CONFIG_USB_MDC800=m
++CONFIG_USB_MICROTEK=m
+
-+ for (i = 0; i <= reg_01.bits.entries; i++) {
-+ struct IO_APIC_route_entry entry;
++#
++# USB Network Adapters
++#
++CONFIG_USB_CATC=m
++CONFIG_USB_KAWETH=m
++CONFIG_USB_PEGASUS=m
++CONFIG_USB_RTL8150=m
++CONFIG_USB_USBNET=m
++CONFIG_USB_NET_AX8817X=m
++CONFIG_USB_NET_CDCETHER=m
++CONFIG_USB_NET_GL620A=m
++CONFIG_USB_NET_NET1080=m
++CONFIG_USB_NET_PLUSB=m
++CONFIG_USB_NET_RNDIS_HOST=m
++CONFIG_USB_NET_CDC_SUBSET=m
++CONFIG_USB_ALI_M5632=y
++CONFIG_USB_AN2720=y
++CONFIG_USB_BELKIN=y
++CONFIG_USB_ARMLINUX=y
++CONFIG_USB_EPSON2888=y
++CONFIG_USB_NET_ZAURUS=m
++CONFIG_USB_MON=y
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ *(((int *)&entry)+0) = io_apic_read(apic, 0x10+i*2);
-+ *(((int *)&entry)+1) = io_apic_read(apic, 0x11+i*2);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
++#
++# USB port drivers
++#
++CONFIG_USB_USS720=m
+
-+ printk(KERN_DEBUG " %02x %03X %02X ",
-+ i,
-+ entry.dest.logical.logical_dest,
-+ entry.dest.physical.physical_dest
-+ );
++#
++# USB Serial Converter support
++#
++CONFIG_USB_SERIAL=m
++CONFIG_USB_SERIAL_GENERIC=y
++CONFIG_USB_SERIAL_AIRPRIME=m
++CONFIG_USB_SERIAL_ARK3116=m
++CONFIG_USB_SERIAL_BELKIN=m
++CONFIG_USB_SERIAL_WHITEHEAT=m
++CONFIG_USB_SERIAL_DIGI_ACCELEPORT=m
++CONFIG_USB_SERIAL_CP2101=m
++CONFIG_USB_SERIAL_CYPRESS_M8=m
++CONFIG_USB_SERIAL_EMPEG=m
++CONFIG_USB_SERIAL_FTDI_SIO=m
++CONFIG_USB_SERIAL_FUNSOFT=m
++CONFIG_USB_SERIAL_VISOR=m
++CONFIG_USB_SERIAL_IPAQ=m
++CONFIG_USB_SERIAL_IR=m
++CONFIG_USB_SERIAL_EDGEPORT=m
++CONFIG_USB_SERIAL_EDGEPORT_TI=m
++CONFIG_USB_SERIAL_GARMIN=m
++CONFIG_USB_SERIAL_IPW=m
++CONFIG_USB_SERIAL_KEYSPAN_PDA=m
++CONFIG_USB_SERIAL_KEYSPAN=m
++CONFIG_USB_SERIAL_KEYSPAN_MPR=y
++CONFIG_USB_SERIAL_KEYSPAN_USA28=y
++CONFIG_USB_SERIAL_KEYSPAN_USA28X=y
++CONFIG_USB_SERIAL_KEYSPAN_USA28XA=y
++CONFIG_USB_SERIAL_KEYSPAN_USA28XB=y
++CONFIG_USB_SERIAL_KEYSPAN_USA19=y
++CONFIG_USB_SERIAL_KEYSPAN_USA18X=y
++CONFIG_USB_SERIAL_KEYSPAN_USA19W=y
++CONFIG_USB_SERIAL_KEYSPAN_USA19QW=y
++CONFIG_USB_SERIAL_KEYSPAN_USA19QI=y
++CONFIG_USB_SERIAL_KEYSPAN_USA49W=y
++CONFIG_USB_SERIAL_KEYSPAN_USA49WLC=y
++CONFIG_USB_SERIAL_KLSI=m
++CONFIG_USB_SERIAL_KOBIL_SCT=m
++CONFIG_USB_SERIAL_MCT_U232=m
++CONFIG_USB_SERIAL_NAVMAN=m
++CONFIG_USB_SERIAL_PL2303=m
++CONFIG_USB_SERIAL_HP4X=m
++CONFIG_USB_SERIAL_SAFE=m
++CONFIG_USB_SERIAL_SAFE_PADDED=y
++CONFIG_USB_SERIAL_SIERRAWIRELESS=m
++CONFIG_USB_SERIAL_TI=m
++CONFIG_USB_SERIAL_CYBERJACK=m
++CONFIG_USB_SERIAL_XIRCOM=m
++# CONFIG_USB_SERIAL_OPTION is not set
++CONFIG_USB_SERIAL_OMNINET=m
++CONFIG_USB_EZUSB=y
+
-+ printk("%1d %1d %1d %1d %1d %1d %1d %02X\n",
-+ entry.mask,
-+ entry.trigger,
-+ entry.irr,
-+ entry.polarity,
-+ entry.delivery_status,
-+ entry.dest_mode,
-+ entry.delivery_mode,
-+ entry.vector
-+ );
-+ }
-+ }
-+ if (use_pci_vector())
-+ printk(KERN_INFO "Using vector-based indexing\n");
-+ printk(KERN_DEBUG "IRQ to pin mappings:\n");
-+ for (i = 0; i < NR_IRQS; i++) {
-+ struct irq_pin_list *entry = irq_2_pin + i;
-+ if (entry->pin < 0)
-+ continue;
-+ if (use_pci_vector() && !platform_legacy_irq(i))
-+ printk(KERN_DEBUG "IRQ%d ", IO_APIC_VECTOR(i));
-+ else
-+ printk(KERN_DEBUG "IRQ%d ", i);
-+ for (;;) {
-+ printk("-> %d:%d", entry->apic, entry->pin);
-+ if (!entry->next)
-+ break;
-+ entry = irq_2_pin + entry->next;
-+ }
-+ printk("\n");
-+ }
++#
++# USB Miscellaneous drivers
++#
++CONFIG_USB_EMI62=m
++CONFIG_USB_EMI26=m
++CONFIG_USB_AUERSWALD=m
++CONFIG_USB_RIO500=m
++CONFIG_USB_LEGOTOWER=m
++CONFIG_USB_LCD=m
++CONFIG_USB_LED=m
++CONFIG_USB_CYPRESS_CY7C63=m
++CONFIG_USB_CYTHERM=m
++CONFIG_USB_PHIDGETKIT=m
++CONFIG_USB_PHIDGETSERVO=m
++CONFIG_USB_IDMOUSE=m
++CONFIG_USB_APPLEDISPLAY=m
++CONFIG_USB_SISUSBVGA=m
++CONFIG_USB_SISUSBVGA_CON=y
++CONFIG_USB_LD=m
++CONFIG_USB_TEST=m
+
-+ printk(KERN_INFO ".................................... done.\n");
++#
++# USB DSL modem support
++#
++CONFIG_USB_ATM=m
++CONFIG_USB_SPEEDTOUCH=m
++CONFIG_USB_CXACRU=m
++CONFIG_USB_UEAGLEATM=m
++CONFIG_USB_XUSBATM=m
+
-+ return;
-+}
++#
++# USB Gadget Support
++#
++# CONFIG_USB_GADGET is not set
+
-+#if 0
++#
++# MMC/SD Card support
++#
++CONFIG_MMC=m
++# CONFIG_MMC_DEBUG is not set
++CONFIG_MMC_BLOCK=m
++CONFIG_MMC_SDHCI=m
++CONFIG_MMC_WBSD=m
+
-+static __apicdebuginit void print_APIC_bitfield (int base)
-+{
-+ unsigned int v;
-+ int i, j;
++#
++# LED devices
++#
++CONFIG_NEW_LEDS=y
++CONFIG_LEDS_CLASS=m
+
-+ if (apic_verbosity == APIC_QUIET)
-+ return;
++#
++# LED drivers
++#
+
-+ printk(KERN_DEBUG "0123456789abcdef0123456789abcdef\n" KERN_DEBUG);
-+ for (i = 0; i < 8; i++) {
-+ v = apic_read(base + i*0x10);
-+ for (j = 0; j < 32; j++) {
-+ if (v & (1<<j))
-+ printk("1");
-+ else
-+ printk("0");
-+ }
-+ printk("\n");
-+ }
-+}
++#
++# LED Triggers
++#
++CONFIG_LEDS_TRIGGERS=y
++CONFIG_LEDS_TRIGGER_TIMER=m
++CONFIG_LEDS_TRIGGER_IDE_DISK=y
++CONFIG_LEDS_TRIGGER_HEARTBEAT=m
+
-+void __apicdebuginit print_local_APIC(void * dummy)
-+{
-+ unsigned int v, ver, maxlvt;
++#
++# InfiniBand support
++#
++CONFIG_INFINIBAND=m
++CONFIG_INFINIBAND_USER_MAD=m
++CONFIG_INFINIBAND_USER_ACCESS=m
++CONFIG_INFINIBAND_ADDR_TRANS=y
++CONFIG_INFINIBAND_MTHCA=m
++CONFIG_INFINIBAND_MTHCA_DEBUG=y
++CONFIG_INFINIBAND_IPOIB=m
++CONFIG_INFINIBAND_IPOIB_DEBUG=y
++# CONFIG_INFINIBAND_IPOIB_DEBUG_DATA is not set
++CONFIG_INFINIBAND_SRP=m
++CONFIG_INFINIBAND_ISER=m
+
-+ if (apic_verbosity == APIC_QUIET)
-+ return;
++#
++# EDAC - error detection and reporting (RAS) (EXPERIMENTAL)
++#
++CONFIG_EDAC=m
+
-+ printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n",
-+ smp_processor_id(), hard_smp_processor_id());
-+ v = apic_read(APIC_ID);
-+ printk(KERN_INFO "... APIC ID: %08x (%01x)\n", v, GET_APIC_ID(v));
-+ v = apic_read(APIC_LVR);
-+ printk(KERN_INFO "... APIC VERSION: %08x\n", v);
-+ ver = GET_APIC_VERSION(v);
-+ maxlvt = get_maxlvt();
++#
++# Reporting subsystems
++#
++# CONFIG_EDAC_DEBUG is not set
++CONFIG_EDAC_MM_EDAC=m
++CONFIG_EDAC_E752X=m
++CONFIG_EDAC_POLL=y
+
-+ v = apic_read(APIC_TASKPRI);
-+ printk(KERN_DEBUG "... APIC TASKPRI: %08x (%02x)\n", v, v & APIC_TPRI_MASK);
++#
++# Real Time Clock
++#
++CONFIG_RTC_LIB=m
++CONFIG_RTC_CLASS=m
+
-+ v = apic_read(APIC_ARBPRI);
-+ printk(KERN_DEBUG "... APIC ARBPRI: %08x (%02x)\n", v,
-+ v & APIC_ARBPRI_MASK);
-+ v = apic_read(APIC_PROCPRI);
-+ printk(KERN_DEBUG "... APIC PROCPRI: %08x\n", v);
++#
++# RTC interfaces
++#
++CONFIG_RTC_INTF_SYSFS=m
++CONFIG_RTC_INTF_PROC=m
++CONFIG_RTC_INTF_DEV=m
++CONFIG_RTC_INTF_DEV_UIE_EMUL=y
+
-+ v = apic_read(APIC_EOI);
-+ printk(KERN_DEBUG "... APIC EOI: %08x\n", v);
-+ v = apic_read(APIC_RRR);
-+ printk(KERN_DEBUG "... APIC RRR: %08x\n", v);
-+ v = apic_read(APIC_LDR);
-+ printk(KERN_DEBUG "... APIC LDR: %08x\n", v);
-+ v = apic_read(APIC_DFR);
-+ printk(KERN_DEBUG "... APIC DFR: %08x\n", v);
-+ v = apic_read(APIC_SPIV);
-+ printk(KERN_DEBUG "... APIC SPIV: %08x\n", v);
++#
++# RTC drivers
++#
++CONFIG_RTC_DRV_X1205=m
++CONFIG_RTC_DRV_DS1307=m
++CONFIG_RTC_DRV_DS1553=m
++CONFIG_RTC_DRV_ISL1208=m
++CONFIG_RTC_DRV_DS1672=m
++CONFIG_RTC_DRV_DS1742=m
++CONFIG_RTC_DRV_PCF8563=m
++CONFIG_RTC_DRV_PCF8583=m
++CONFIG_RTC_DRV_RS5C348=m
++CONFIG_RTC_DRV_RS5C372=m
++CONFIG_RTC_DRV_M48T86=m
++CONFIG_RTC_DRV_TEST=m
++CONFIG_RTC_DRV_MAX6902=m
++CONFIG_RTC_DRV_V3020=m
+
-+ printk(KERN_DEBUG "... APIC ISR field:\n");
-+ print_APIC_bitfield(APIC_ISR);
-+ printk(KERN_DEBUG "... APIC TMR field:\n");
-+ print_APIC_bitfield(APIC_TMR);
-+ printk(KERN_DEBUG "... APIC IRR field:\n");
-+ print_APIC_bitfield(APIC_IRR);
++#
++# DMA Engine support
++#
++CONFIG_DMA_ENGINE=y
+
-+ v = apic_read(APIC_ESR);
-+ printk(KERN_DEBUG "... APIC ESR: %08x\n", v);
++#
++# DMA Clients
++#
++CONFIG_NET_DMA=y
+
-+ v = apic_read(APIC_ICR);
-+ printk(KERN_DEBUG "... APIC ICR: %08x\n", v);
-+ v = apic_read(APIC_ICR2);
-+ printk(KERN_DEBUG "... APIC ICR2: %08x\n", v);
++#
++# DMA Devices
++#
++CONFIG_INTEL_IOATDMA=m
+
-+ v = apic_read(APIC_LVTT);
-+ printk(KERN_DEBUG "... APIC LVTT: %08x\n", v);
++#
++# Firmware Drivers
++#
++CONFIG_EDD=y
++CONFIG_DELL_RBU=m
++CONFIG_DCDBAS=m
+
-+ if (maxlvt > 3) { /* PC is LVT#4. */
-+ v = apic_read(APIC_LVTPC);
-+ printk(KERN_DEBUG "... APIC LVTPC: %08x\n", v);
-+ }
-+ v = apic_read(APIC_LVT0);
-+ printk(KERN_DEBUG "... APIC LVT0: %08x\n", v);
-+ v = apic_read(APIC_LVT1);
-+ printk(KERN_DEBUG "... APIC LVT1: %08x\n", v);
++#
++# File systems
++#
++CONFIG_EXT2_FS=y
++CONFIG_EXT2_FS_XATTR=y
++CONFIG_EXT2_FS_POSIX_ACL=y
++CONFIG_EXT2_FS_SECURITY=y
++CONFIG_EXT2_FS_XIP=y
++CONFIG_FS_XIP=y
++CONFIG_EXT3_FS=m
++CONFIG_EXT3_FS_XATTR=y
++CONFIG_EXT3_FS_POSIX_ACL=y
++CONFIG_EXT3_FS_SECURITY=y
++CONFIG_JBD=m
++# CONFIG_JBD_DEBUG is not set
++CONFIG_FS_MBCACHE=y
++CONFIG_REISERFS_FS=m
++# CONFIG_REISERFS_CHECK is not set
++CONFIG_REISERFS_PROC_INFO=y
++CONFIG_REISERFS_FS_XATTR=y
++CONFIG_REISERFS_FS_POSIX_ACL=y
++CONFIG_REISERFS_FS_SECURITY=y
++CONFIG_JFS_FS=m
++CONFIG_JFS_POSIX_ACL=y
++CONFIG_JFS_SECURITY=y
++# CONFIG_JFS_DEBUG is not set
++# CONFIG_JFS_STATISTICS is not set
++CONFIG_FS_POSIX_ACL=y
++CONFIG_XFS_FS=m
++CONFIG_XFS_QUOTA=y
++CONFIG_XFS_SECURITY=y
++CONFIG_XFS_POSIX_ACL=y
++CONFIG_XFS_RT=y
++CONFIG_OCFS2_FS=m
++CONFIG_OCFS2_DEBUG_MASKLOG=y
++CONFIG_MINIX_FS=m
++CONFIG_ROMFS_FS=m
++CONFIG_INOTIFY=y
++CONFIG_INOTIFY_USER=y
++CONFIG_QUOTA=y
++# CONFIG_QFMT_V1 is not set
++CONFIG_QFMT_V2=m
++CONFIG_QUOTACTL=y
++CONFIG_DNOTIFY=y
++CONFIG_AUTOFS_FS=m
++CONFIG_AUTOFS4_FS=m
++CONFIG_FUSE_FS=m
+
-+ if (maxlvt > 2) { /* ERR is LVT#3. */
-+ v = apic_read(APIC_LVTERR);
-+ printk(KERN_DEBUG "... APIC LVTERR: %08x\n", v);
-+ }
++#
++# CD-ROM/DVD Filesystems
++#
++CONFIG_ISO9660_FS=y
++CONFIG_JOLIET=y
++CONFIG_ZISOFS=y
++CONFIG_ZISOFS_FS=y
++CONFIG_UDF_FS=m
++CONFIG_UDF_NLS=y
+
-+ v = apic_read(APIC_TMICT);
-+ printk(KERN_DEBUG "... APIC TMICT: %08x\n", v);
-+ v = apic_read(APIC_TMCCT);
-+ printk(KERN_DEBUG "... APIC TMCCT: %08x\n", v);
-+ v = apic_read(APIC_TDCR);
-+ printk(KERN_DEBUG "... APIC TDCR: %08x\n", v);
-+ printk("\n");
-+}
++#
++# DOS/FAT/NT Filesystems
++#
++CONFIG_FAT_FS=m
++CONFIG_MSDOS_FS=m
++CONFIG_VFAT_FS=m
++CONFIG_FAT_DEFAULT_CODEPAGE=437
++CONFIG_FAT_DEFAULT_IOCHARSET="iso8859-1"
++CONFIG_NTFS_FS=m
++# CONFIG_NTFS_DEBUG is not set
++# CONFIG_NTFS_RW is not set
+
-+void print_all_local_APICs (void)
-+{
-+ on_each_cpu(print_local_APIC, NULL, 1, 1);
-+}
++#
++# Pseudo filesystems
++#
++CONFIG_PROC_FS=y
++CONFIG_PROC_KCORE=y
++CONFIG_SYSFS=y
++CONFIG_TMPFS=y
++# CONFIG_HUGETLB_PAGE is not set
++CONFIG_RAMFS=y
++CONFIG_CONFIGFS_FS=m
+
-+void __apicdebuginit print_PIC(void)
-+{
-+ unsigned int v;
-+ unsigned long flags;
++#
++# Miscellaneous filesystems
++#
++CONFIG_ADFS_FS=m
++# CONFIG_ADFS_FS_RW is not set
++CONFIG_AFFS_FS=m
++CONFIG_HFS_FS=m
++CONFIG_HFSPLUS_FS=m
++CONFIG_BEFS_FS=m
++# CONFIG_BEFS_DEBUG is not set
++CONFIG_BFS_FS=m
++CONFIG_EFS_FS=m
++CONFIG_JFFS_FS=m
++CONFIG_JFFS_FS_VERBOSE=0
++CONFIG_JFFS_PROC_FS=y
++CONFIG_JFFS2_FS=m
++CONFIG_JFFS2_FS_DEBUG=0
++CONFIG_JFFS2_FS_WRITEBUFFER=y
++CONFIG_JFFS2_SUMMARY=y
++CONFIG_JFFS2_FS_XATTR=y
++CONFIG_JFFS2_FS_POSIX_ACL=y
++CONFIG_JFFS2_FS_SECURITY=y
++CONFIG_JFFS2_COMPRESSION_OPTIONS=y
++CONFIG_JFFS2_ZLIB=y
++CONFIG_JFFS2_RTIME=y
++# CONFIG_JFFS2_RUBIN is not set
++# CONFIG_JFFS2_CMODE_NONE is not set
++CONFIG_JFFS2_CMODE_PRIORITY=y
++# CONFIG_JFFS2_CMODE_SIZE is not set
++CONFIG_CRAMFS=m
++CONFIG_VXFS_FS=m
++CONFIG_HPFS_FS=m
++CONFIG_QNX4FS_FS=m
++CONFIG_SYSV_FS=m
++CONFIG_UFS_FS=m
++# CONFIG_UFS_FS_WRITE is not set
++# CONFIG_UFS_DEBUG is not set
+
-+ if (apic_verbosity == APIC_QUIET)
-+ return;
++#
++# Network File Systems
++#
++CONFIG_NFS_FS=m
++CONFIG_NFS_V3=y
++CONFIG_NFS_V3_ACL=y
++CONFIG_NFS_V4=y
++CONFIG_NFS_DIRECTIO=y
++CONFIG_NFSD=m
++CONFIG_NFSD_V2_ACL=y
++CONFIG_NFSD_V3=y
++CONFIG_NFSD_V3_ACL=y
++CONFIG_NFSD_V4=y
++CONFIG_NFSD_TCP=y
++CONFIG_LOCKD=m
++CONFIG_LOCKD_V4=y
++CONFIG_EXPORTFS=m
++CONFIG_NFS_ACL_SUPPORT=m
++CONFIG_NFS_COMMON=y
++CONFIG_SUNRPC=m
++CONFIG_SUNRPC_GSS=m
++CONFIG_RPCSEC_GSS_KRB5=m
++CONFIG_RPCSEC_GSS_SPKM3=m
++CONFIG_SMB_FS=m
++CONFIG_SMB_NLS_DEFAULT=y
++CONFIG_SMB_NLS_REMOTE="cp850"
++CONFIG_CIFS=m
++CONFIG_CIFS_STATS=y
++# CONFIG_CIFS_STATS2 is not set
++# CONFIG_CIFS_WEAK_PW_HASH is not set
++CONFIG_CIFS_XATTR=y
++CONFIG_CIFS_POSIX=y
++# CONFIG_CIFS_DEBUG2 is not set
++# CONFIG_CIFS_EXPERIMENTAL is not set
++CONFIG_NCP_FS=m
++CONFIG_NCPFS_PACKET_SIGNING=y
++CONFIG_NCPFS_IOCTL_LOCKING=y
++CONFIG_NCPFS_STRONG=y
++CONFIG_NCPFS_NFS_NS=y
++CONFIG_NCPFS_OS2_NS=y
++CONFIG_NCPFS_SMALLDOS=y
++CONFIG_NCPFS_NLS=y
++CONFIG_NCPFS_EXTRAS=y
++CONFIG_CODA_FS=m
++# CONFIG_CODA_FS_OLD_API is not set
++CONFIG_AFS_FS=m
++CONFIG_RXRPC=m
++CONFIG_9P_FS=m
+
-+ printk(KERN_DEBUG "\nprinting PIC contents\n");
++#
++# Partition Types
++#
++CONFIG_PARTITION_ADVANCED=y
++# CONFIG_ACORN_PARTITION is not set
++CONFIG_OSF_PARTITION=y
++CONFIG_AMIGA_PARTITION=y
++# CONFIG_ATARI_PARTITION is not set
++CONFIG_MAC_PARTITION=y
++CONFIG_MSDOS_PARTITION=y
++CONFIG_BSD_DISKLABEL=y
++CONFIG_MINIX_SUBPARTITION=y
++CONFIG_SOLARIS_X86_PARTITION=y
++CONFIG_UNIXWARE_DISKLABEL=y
++# CONFIG_LDM_PARTITION is not set
++CONFIG_SGI_PARTITION=y
++# CONFIG_ULTRIX_PARTITION is not set
++CONFIG_SUN_PARTITION=y
++# CONFIG_KARMA_PARTITION is not set
++CONFIG_EFI_PARTITION=y
+
-+ spin_lock_irqsave(&i8259A_lock, flags);
++#
++# Native Language Support
++#
++CONFIG_NLS=y
++CONFIG_NLS_DEFAULT="utf8"
++CONFIG_NLS_CODEPAGE_437=m
++CONFIG_NLS_CODEPAGE_737=m
++CONFIG_NLS_CODEPAGE_775=m
++CONFIG_NLS_CODEPAGE_850=m
++CONFIG_NLS_CODEPAGE_852=m
++CONFIG_NLS_CODEPAGE_855=m
++CONFIG_NLS_CODEPAGE_857=m
++CONFIG_NLS_CODEPAGE_860=m
++CONFIG_NLS_CODEPAGE_861=m
++CONFIG_NLS_CODEPAGE_862=m
++CONFIG_NLS_CODEPAGE_863=m
++CONFIG_NLS_CODEPAGE_864=m
++CONFIG_NLS_CODEPAGE_865=m
++CONFIG_NLS_CODEPAGE_866=m
++CONFIG_NLS_CODEPAGE_869=m
++CONFIG_NLS_CODEPAGE_936=m
++CONFIG_NLS_CODEPAGE_950=m
++CONFIG_NLS_CODEPAGE_932=m
++CONFIG_NLS_CODEPAGE_949=m
++CONFIG_NLS_CODEPAGE_874=m
++CONFIG_NLS_ISO8859_8=m
++CONFIG_NLS_CODEPAGE_1250=m
++CONFIG_NLS_CODEPAGE_1251=m
++CONFIG_NLS_ASCII=m
++CONFIG_NLS_ISO8859_1=m
++CONFIG_NLS_ISO8859_2=m
++CONFIG_NLS_ISO8859_3=m
++CONFIG_NLS_ISO8859_4=m
++CONFIG_NLS_ISO8859_5=m
++CONFIG_NLS_ISO8859_6=m
++CONFIG_NLS_ISO8859_7=m
++CONFIG_NLS_ISO8859_9=m
++CONFIG_NLS_ISO8859_13=m
++CONFIG_NLS_ISO8859_14=m
++CONFIG_NLS_ISO8859_15=m
++CONFIG_NLS_KOI8_R=m
++CONFIG_NLS_KOI8_U=m
++CONFIG_NLS_UTF8=m
+
-+ v = inb(0xa1) << 8 | inb(0x21);
-+ printk(KERN_DEBUG "... PIC IMR: %04x\n", v);
++#
++# Instrumentation Support
++#
++# CONFIG_PROFILING is not set
++# CONFIG_KPROBES is not set
+
-+ v = inb(0xa0) << 8 | inb(0x20);
-+ printk(KERN_DEBUG "... PIC IRR: %04x\n", v);
++#
++# Kernel hacking
++#
++CONFIG_TRACE_IRQFLAGS_SUPPORT=y
++# CONFIG_PRINTK_TIME is not set
++CONFIG_MAGIC_SYSRQ=y
++CONFIG_UNUSED_SYMBOLS=y
++CONFIG_DEBUG_KERNEL=y
++CONFIG_LOG_BUF_SHIFT=15
++CONFIG_DETECT_SOFTLOCKUP=y
++# CONFIG_SCHEDSTATS is not set
++# CONFIG_DEBUG_SLAB is not set
++# CONFIG_DEBUG_RT_MUTEXES is not set
++# CONFIG_RT_MUTEX_TESTER is not set
++# CONFIG_DEBUG_SPINLOCK is not set
++# CONFIG_DEBUG_MUTEXES is not set
++# CONFIG_DEBUG_RWSEMS is not set
++# CONFIG_DEBUG_LOCK_ALLOC is not set
++# CONFIG_PROVE_LOCKING is not set
++# CONFIG_DEBUG_SPINLOCK_SLEEP is not set
++# CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set
++# CONFIG_DEBUG_KOBJECT is not set
++# CONFIG_DEBUG_INFO is not set
++# CONFIG_DEBUG_FS is not set
++# CONFIG_DEBUG_VM is not set
++# CONFIG_FRAME_POINTER is not set
++# CONFIG_UNWIND_INFO is not set
++CONFIG_FORCED_INLINING=y
++# CONFIG_RCU_TORTURE_TEST is not set
++# CONFIG_DEBUG_RODATA is not set
++# CONFIG_DEBUG_STACKOVERFLOW is not set
++# CONFIG_DEBUG_STACK_USAGE is not set
+
-+ outb(0x0b,0xa0);
-+ outb(0x0b,0x20);
-+ v = inb(0xa0) << 8 | inb(0x20);
-+ outb(0x0a,0xa0);
-+ outb(0x0a,0x20);
++#
++# Security options
++#
++CONFIG_KEYS=y
++CONFIG_KEYS_DEBUG_PROC_KEYS=y
++CONFIG_SECURITY=y
++CONFIG_SECURITY_NETWORK=y
++# CONFIG_SECURITY_NETWORK_XFRM is not set
++CONFIG_SECURITY_CAPABILITIES=y
++CONFIG_SECURITY_ROOTPLUG=m
++CONFIG_SECURITY_SECLVL=m
++# CONFIG_SECURITY_SELINUX is not set
+
-+ spin_unlock_irqrestore(&i8259A_lock, flags);
++#
++# Cryptographic options
++#
++CONFIG_CRYPTO=y
++CONFIG_CRYPTO_HMAC=y
++CONFIG_CRYPTO_NULL=m
++CONFIG_CRYPTO_MD4=m
++CONFIG_CRYPTO_MD5=y
++CONFIG_CRYPTO_SHA1=m
++CONFIG_CRYPTO_SHA256=m
++CONFIG_CRYPTO_SHA512=m
++CONFIG_CRYPTO_WP512=m
++CONFIG_CRYPTO_TGR192=m
++CONFIG_CRYPTO_DES=m
++CONFIG_CRYPTO_BLOWFISH=m
++CONFIG_CRYPTO_TWOFISH=m
++CONFIG_CRYPTO_SERPENT=m
++CONFIG_CRYPTO_AES=m
++CONFIG_CRYPTO_AES_X86_64=m
++CONFIG_CRYPTO_CAST5=m
++CONFIG_CRYPTO_CAST6=m
++CONFIG_CRYPTO_TEA=m
++CONFIG_CRYPTO_ARC4=m
++CONFIG_CRYPTO_KHAZAD=m
++CONFIG_CRYPTO_ANUBIS=m
++CONFIG_CRYPTO_DEFLATE=m
++CONFIG_CRYPTO_MICHAEL_MIC=m
++CONFIG_CRYPTO_CRC32C=m
++CONFIG_CRYPTO_TEST=m
+
-+ printk(KERN_DEBUG "... PIC ISR: %04x\n", v);
++#
++# Hardware crypto devices
++#
++CONFIG_XEN=y
++CONFIG_XEN_INTERFACE_VERSION=0x00030207
+
-+ v = inb(0x4d1) << 8 | inb(0x4d0);
-+ printk(KERN_DEBUG "... PIC ELCR: %04x\n", v);
-+}
++#
++# XEN
++#
++CONFIG_XEN_PRIVILEGED_GUEST=y
++# CONFIG_XEN_UNPRIVILEGED_GUEST is not set
++CONFIG_XEN_PRIVCMD=y
++CONFIG_XEN_XENBUS_DEV=y
++CONFIG_XEN_BACKEND=y
++CONFIG_XEN_BLKDEV_BACKEND=y
++CONFIG_XEN_BLKDEV_TAP=y
++CONFIG_XEN_NETDEV_BACKEND=y
++# CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_UTIL is not set
++# CONFIG_XEN_NETDEV_ACCEL_SFC_BACKEND is not set
++# CONFIG_XEN_NETDEV_LOOPBACK is not set
++CONFIG_XEN_PCIDEV_BACKEND=m
++# CONFIG_XEN_PCIDEV_BACKEND_VPCI is not set
++CONFIG_XEN_PCIDEV_BACKEND_PASS=y
++# CONFIG_XEN_PCIDEV_BACKEND_SLOT is not set
++# CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER is not set
++# CONFIG_XEN_PCIDEV_BE_DEBUG is not set
++CONFIG_XEN_TPMDEV_BACKEND=m
++CONFIG_XEN_SCSI_BACKEND=m
++CONFIG_XEN_BLKDEV_FRONTEND=y
++CONFIG_XEN_NETDEV_FRONTEND=y
++CONFIG_XEN_SCSI_FRONTEND=m
++CONFIG_XEN_GRANT_DEV=y
++# CONFIG_XEN_NETDEV_ACCEL_SFC_FRONTEND is not set
++CONFIG_XEN_FRAMEBUFFER=y
++CONFIG_XEN_KEYBOARD=y
++CONFIG_XEN_SCRUB_PAGES=y
++CONFIG_XEN_DISABLE_SERIAL=y
++CONFIG_XEN_SYSFS=y
++CONFIG_XEN_COMPAT_030002_AND_LATER=y
++# CONFIG_XEN_COMPAT_030004_AND_LATER is not set
++# CONFIG_XEN_COMPAT_030100_AND_LATER is not set
++# CONFIG_XEN_COMPAT_LATEST_ONLY is not set
++CONFIG_XEN_COMPAT=0x030002
++CONFIG_HAVE_IRQ_IGNORE_UNHANDLED=y
++CONFIG_NO_IDLE_HZ=y
++CONFIG_XEN_SMPBOOT=y
++CONFIG_XEN_BALLOON=y
++CONFIG_XEN_DEVMEM=y
+
-+#endif /* 0 */
++#
++# Library routines
++#
++CONFIG_CRC_CCITT=m
++CONFIG_CRC16=m
++CONFIG_CRC32=y
++CONFIG_LIBCRC32C=m
++CONFIG_ZLIB_INFLATE=y
++CONFIG_ZLIB_DEFLATE=m
++CONFIG_REED_SOLOMON=m
++CONFIG_REED_SOLOMON_DEC16=y
++CONFIG_TEXTSEARCH=y
++CONFIG_TEXTSEARCH_KMP=m
++CONFIG_TEXTSEARCH_BM=m
++CONFIG_TEXTSEARCH_FSM=m
++CONFIG_PLIST=y
+diff -r d894e36cfc30 -r 0aa021803deb drivers/Makefile
+--- a/drivers/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -31,6 +31,7 @@
+ obj-$(CONFIG_NUBUS) += nubus/
+ obj-$(CONFIG_ATM) += atm/
+ obj-$(CONFIG_PPC_PMAC) += macintosh/
++obj-$(CONFIG_XEN) += xen/
+ obj-$(CONFIG_IDE) += ide/
+ obj-$(CONFIG_FC4) += fc4/
+ obj-$(CONFIG_SCSI) += scsi/
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/Kconfig
+--- a/drivers/acpi/Kconfig Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/acpi/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -305,6 +305,7 @@
+ config X86_PM_TIMER
+ bool "Power Management Timer Support" if EMBEDDED
+ depends on X86
++ depends on !XEN
+ default y
+ help
+ The Power Management Timer is available on all ACPI-capable,
+@@ -362,6 +363,15 @@
+ A "Smart Battery" is quite old and quite rare compared
+ to today's ACPI "Control Method" battery.
+
++config ACPI_PV_SLEEP
++ bool
++ depends on X86 && XEN && ACPI_SLEEP
++ default y
+
-+#else
-+void __init print_IO_APIC(void) { }
-+#endif /* !CONFIG_XEN */
++config PROCESSOR_EXTERNAL_CONTROL
++ bool
++ depends on X86 && XEN
++ default y
+ endif # ACPI
+
+ endmenu
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/Makefile
+--- a/drivers/acpi/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/acpi/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -34,6 +34,9 @@
+ ifdef CONFIG_CPU_FREQ
+ processor-objs += processor_perflib.o
+ endif
++ifdef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++processor-objs += processor_perflib.o processor_extcntl.o
++endif
+
+ obj-y += sleep/
+ obj-y += bus.o glue.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/dispatcher/dsobject.c
+--- a/drivers/acpi/dispatcher/dsobject.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/acpi/dispatcher/dsobject.c Wed Sep 10 10:54:08 2008 +0100
+@@ -137,6 +137,71 @@
+ return_ACPI_STATUS(status);
+ }
+ }
+
-+static void __init enable_IO_APIC(void)
-+{
-+ union IO_APIC_reg_01 reg_01;
-+#ifndef CONFIG_XEN
-+ int i8259_apic, i8259_pin;
-+#endif
-+ int i, apic;
-+ unsigned long flags;
++ /* Special object resolution for elements of a package */
+
-+ for (i = 0; i < PIN_MAP_SIZE; i++) {
-+ irq_2_pin[i].pin = -1;
-+ irq_2_pin[i].next = 0;
-+ }
-+ if (!pirqs_enabled)
-+ for (i = 0; i < MAX_PIRQS; i++)
-+ pirq_entries[i] = -1;
++ if ((op->common.parent->common.aml_opcode == AML_PACKAGE_OP) ||
++ (op->common.parent->common.aml_opcode ==
++ AML_VAR_PACKAGE_OP)) {
++ /*
++ * Attempt to resolve the node to a value before we insert it into
++ * the package. If this is a reference to a common data type,
++ * resolve it immediately. According to the ACPI spec, package
++ * elements can only be "data objects" or method references.
++ * Attempt to resolve to an Integer, Buffer, String or Package.
++ * If cannot, return the named reference (for things like Devices,
++ * Methods, etc.) Buffer Fields and Fields will resolve to simple
++ * objects (int/buf/str/pkg).
++ *
++ * NOTE: References to things like Devices, Methods, Mutexes, etc.
++ * will remain as named references. This behavior is not described
++ * in the ACPI spec, but it appears to be an oversight.
++ */
++ obj_desc = (union acpi_operand_object *)op->common.node;
+
-+ /*
-+ * The number of IO-APIC IRQ registers (== #pins):
-+ */
-+ for (apic = 0; apic < nr_ioapics; apic++) {
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ reg_01.raw = io_apic_read(apic, 1);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+ nr_ioapic_registers[apic] = reg_01.bits.entries+1;
-+ }
-+#ifndef CONFIG_XEN
-+ for(apic = 0; apic < nr_ioapics; apic++) {
-+ int pin;
-+ /* See if any of the pins is in ExtINT mode */
-+ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
-+ struct IO_APIC_route_entry entry;
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ *(((int *)&entry) + 0) = io_apic_read(apic, 0x10 + 2 * pin);
-+ *(((int *)&entry) + 1) = io_apic_read(apic, 0x11 + 2 * pin);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
++ status =
++ acpi_ex_resolve_node_to_value(ACPI_CAST_INDIRECT_PTR
++ (struct
++ acpi_namespace_node,
++ &obj_desc),
++ walk_state);
++ if (ACPI_FAILURE(status)) {
++ return_ACPI_STATUS(status);
++ }
++
++ switch (op->common.node->type) {
++ /*
++ * For these types, we need the actual node, not the subobject.
++ * However, the subobject got an extra reference count above.
++ */
++ case ACPI_TYPE_MUTEX:
++ case ACPI_TYPE_METHOD:
++ case ACPI_TYPE_POWER:
++ case ACPI_TYPE_PROCESSOR:
++ case ACPI_TYPE_EVENT:
++ case ACPI_TYPE_REGION:
++ case ACPI_TYPE_DEVICE:
++ case ACPI_TYPE_THERMAL:
++
++ obj_desc =
++ (union acpi_operand_object *)op->common.
++ node;
++ break;
+
++ default:
++ break;
++ }
+
-+ /* If the interrupt line is enabled and in ExtInt mode
-+ * I have found the pin where the i8259 is connected.
++ /*
++ * If above resolved to an operand object, we are done. Otherwise,
++ * we have a NS node, we must create the package entry as a named
++ * reference.
+ */
-+ if ((entry.mask == 0) && (entry.delivery_mode == dest_ExtINT)) {
-+ ioapic_i8259.apic = apic;
-+ ioapic_i8259.pin = pin;
-+ goto found_i8259;
++ if (ACPI_GET_DESCRIPTOR_TYPE(obj_desc) !=
++ ACPI_DESC_TYPE_NAMED) {
++ goto exit;
+ }
+ }
-+ }
-+ found_i8259:
-+ /* Look to see what if the MP table has reported the ExtINT */
-+ i8259_pin = find_isa_irq_pin(0, mp_ExtINT);
-+ i8259_apic = find_isa_irq_apic(0, mp_ExtINT);
-+ /* Trust the MP table if nothing is setup in the hardware */
-+ if ((ioapic_i8259.pin == -1) && (i8259_pin >= 0)) {
-+ printk(KERN_WARNING "ExtINT not setup in hardware but reported by MP table\n");
-+ ioapic_i8259.pin = i8259_pin;
-+ ioapic_i8259.apic = i8259_apic;
-+ }
-+ /* Complain if the MP table and the hardware disagree */
-+ if (((ioapic_i8259.apic != i8259_apic) || (ioapic_i8259.pin != i8259_pin)) &&
-+ (i8259_pin >= 0) && (ioapic_i8259.pin >= 0))
-+ {
-+ printk(KERN_WARNING "ExtINT in hardware and MP table differ\n");
+ }
+
+ /* Create and init a new internal ACPI object */
+@@ -156,6 +221,7 @@
+ return_ACPI_STATUS(status);
+ }
+
++ exit:
+ *obj_desc_ptr = obj_desc;
+ return_ACPI_STATUS(AE_OK);
+ }
+@@ -358,12 +424,25 @@
+ arg = arg->common.next;
+ for (i = 0; arg; i++) {
+ if (arg->common.aml_opcode == AML_INT_RETURN_VALUE_OP) {
++ if (arg->common.node->type == ACPI_TYPE_METHOD) {
++ /*
++ * A method reference "looks" to the parser to be a method
++ * invocation, so we special case it here
++ */
++ arg->common.aml_opcode = AML_INT_NAMEPATH_OP;
++ status =
++ acpi_ds_build_internal_object(walk_state,
++ arg,
++ &obj_desc->
++ package.
++ elements[i]);
++ } else {
++ /* This package element is already built, just get it */
+
+- /* Object (package or buffer) is already built */
+-
+- obj_desc->package.elements[i] =
+- ACPI_CAST_PTR(union acpi_operand_object,
+- arg->common.node);
++ obj_desc->package.elements[i] =
++ ACPI_CAST_PTR(union acpi_operand_object,
++ arg->common.node);
++ }
+ } else {
+ status = acpi_ds_build_internal_object(walk_state, arg,
+ &obj_desc->
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/hardware/hwregs.c
+--- a/drivers/acpi/hardware/hwregs.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/acpi/hardware/hwregs.c Wed Sep 10 10:54:08 2008 +0100
+@@ -665,8 +665,6 @@
+
+ /*
+ * Perform a read first to preserve certain bits (per ACPI spec)
+- *
+- * Note: This includes SCI_EN, we never want to change this bit
+ */
+ status = acpi_hw_register_read(ACPI_MTX_DO_NOT_LOCK,
+ ACPI_REGISTER_PM1_CONTROL,
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/hardware/hwsleep.c
+--- a/drivers/acpi/hardware/hwsleep.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/acpi/hardware/hwsleep.c Wed Sep 10 10:54:08 2008 +0100
+@@ -227,7 +227,11 @@
+ u32 PM1Bcontrol;
+ struct acpi_bit_register_info *sleep_type_reg_info;
+ struct acpi_bit_register_info *sleep_enable_reg_info;
++#if !(defined(CONFIG_XEN) && defined(CONFIG_X86))
+ u32 in_value;
++#else
++ int err;
++#endif
+ acpi_status status;
+
+ ACPI_FUNCTION_TRACE(acpi_enter_sleep_state);
+@@ -327,6 +331,7 @@
+
+ ACPI_FLUSH_CPU_CACHE();
+
++#if !(defined(CONFIG_XEN) && defined(CONFIG_X86))
+ status = acpi_hw_register_write(ACPI_MTX_DO_NOT_LOCK,
+ ACPI_REGISTER_PM1A_CONTROL,
+ PM1Acontrol);
+@@ -376,6 +381,16 @@
+ /* Spin until we wake */
+
+ } while (!in_value);
++#else
++ /* PV ACPI just need check hypercall return value */
++ err = acpi_notify_hypervisor_state(sleep_state,
++ PM1Acontrol, PM1Bcontrol);
++ if (err) {
++ ACPI_DEBUG_PRINT((ACPI_DB_ERROR,
++ "Hypervisor failure [%d]\n", err));
++ return_ACPI_STATUS(AE_ERROR);
+ }
+#endif
+
+ return_ACPI_STATUS(AE_OK);
+ }
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/processor_core.c
+--- a/drivers/acpi/processor_core.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/acpi/processor_core.c Wed Sep 10 10:54:08 2008 +0100
+@@ -475,7 +475,8 @@
+ */
+ if (cpu_index == -1) {
+ if (ACPI_FAILURE
+- (acpi_processor_hotadd_init(pr->handle, &pr->id))) {
++ (acpi_processor_hotadd_init(pr->handle, &pr->id)) &&
++ !processor_cntl_external()) {
+ printk(KERN_ERR PREFIX
+ "Getting cpuindex for acpiid 0x%x\n",
+ pr->acpi_id);
+@@ -508,7 +509,7 @@
+ request_region(pr->throttling.address, 6, "ACPI CPU throttle");
+ }
+
+-#ifdef CONFIG_CPU_FREQ
++#if defined(CONFIG_CPU_FREQ) || defined(CONFIG_PROCESSOR_EXTERNAL_CONTROL)
+ acpi_processor_ppc_has_changed(pr);
+ #endif
+ acpi_processor_get_throttling_info(pr);
+@@ -517,7 +518,7 @@
+ return 0;
+ }
+
+-static void *processor_device_array[NR_CPUS];
++static void *processor_device_array[NR_ACPI_CPUS];
+
+ static int acpi_processor_start(struct acpi_device *device)
+ {
+@@ -525,31 +526,46 @@
+ acpi_status status = AE_OK;
+ struct acpi_processor *pr;
+
++ processor_extcntl_init();
+
+ pr = acpi_driver_data(device);
+
+ result = acpi_processor_get_info(pr);
+- if (result) {
++ if (result ||
++ ((pr->id == -1) && !processor_cntl_external())) {
+ /* Processor is physically not present */
+ return 0;
+ }
+
+- BUG_ON((pr->id >= NR_CPUS) || (pr->id < 0));
++ BUG_ON(!processor_cntl_external() &&
++ ((pr->id >= NR_CPUS) || (pr->id < 0)));
+
+ /*
+ * Buggy BIOS check
+ * ACPI id of processors can be reported wrongly by the BIOS.
+ * Don't trust it blindly
+ */
++#ifdef CONFIG_XEN
++ BUG_ON(pr->acpi_id >= NR_ACPI_CPUS);
++ if (processor_device_array[pr->acpi_id] != NULL &&
++ processor_device_array[pr->acpi_id] != (void *)device) {
++#else
+ if (processor_device_array[pr->id] != NULL &&
+ processor_device_array[pr->id] != (void *)device) {
++#endif /* CONFIG_XEN */
+ printk(KERN_WARNING "BIOS reported wrong ACPI id"
+ "for the processor\n");
+ return -ENODEV;
+ }
++#ifdef CONFIG_XEN
++ processor_device_array[pr->acpi_id] = (void *)device;
++ if (pr->id != -1)
++ processors[pr->id] = pr;
++#else
+ processor_device_array[pr->id] = (void *)device;
+
+ processors[pr->id] = pr;
++#endif /* CONFIG_XEN */
+
+ result = acpi_processor_add_fs(device);
+ if (result)
+@@ -563,6 +579,10 @@
+ acpi_processor_set_pdc(pr);
+
+ acpi_processor_power_init(pr, device);
+
-+ /*
-+ * Do not trust the IO-APIC being empty at bootup
-+ */
-+ clear_IO_APIC();
-+}
-+
-+/*
-+ * Not an __init, needed by the reboot code
-+ */
-+void disable_IO_APIC(void)
-+{
-+ /*
-+ * Clear the IO-APIC before rebooting:
-+ */
-+ clear_IO_APIC();
-+
-+#ifndef CONFIG_XEN
-+ /*
-+ * If the i8259 is routed through an IOAPIC
-+ * Put that IOAPIC in virtual wire mode
-+ * so legacy interrupts can be delivered.
-+ */
-+ if (ioapic_i8259.pin != -1) {
-+ struct IO_APIC_route_entry entry;
-+ unsigned long flags;
++ result = processor_extcntl_prepare(pr);
++ if (result)
++ goto end;
+
+ if (pr->flags.throttling) {
+ printk(KERN_INFO PREFIX "%s [%s] (supports",
+@@ -656,7 +676,13 @@
+
+ acpi_processor_remove_fs(device);
+
++#ifdef CONFIG_XEN
++ if (pr->id != -1)
++ processors[pr->id] = NULL;
++#else
+ processors[pr->id] = NULL;
++#endif /* CONFIG_XEN */
+
-+ memset(&entry, 0, sizeof(entry));
-+ entry.mask = 0; /* Enabled */
-+ entry.trigger = 0; /* Edge */
-+ entry.irr = 0;
-+ entry.polarity = 0; /* High */
-+ entry.delivery_status = 0;
-+ entry.dest_mode = 0; /* Physical */
-+ entry.delivery_mode = dest_ExtINT; /* ExtInt */
-+ entry.vector = 0;
-+ entry.dest.physical.physical_dest =
-+ GET_APIC_ID(apic_read(APIC_ID));
+
+ kfree(pr);
+
+@@ -709,6 +735,10 @@
+ pr = acpi_driver_data(*device);
+ if (!pr)
+ return -ENODEV;
+
-+ /*
-+ * Add it to the IO-APIC irq-routing table:
-+ */
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ io_apic_write(ioapic_i8259.apic, 0x11+2*ioapic_i8259.pin,
-+ *(((int *)&entry)+1));
-+ io_apic_write(ioapic_i8259.apic, 0x10+2*ioapic_i8259.pin,
-+ *(((int *)&entry)+0));
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+ }
++ if (processor_cntl_external())
++ processor_notify_external(pr,
++ PROCESSOR_HOTPLUG, HOTPLUG_TYPE_ADD);
+
+ if ((pr->id >= 0) && (pr->id < NR_CPUS)) {
+ kobject_uevent(&(*device)->kobj, KOBJ_ONLINE);
+@@ -747,6 +777,10 @@
+ printk(KERN_ERR PREFIX "Driver data is NULL\n");
+ break;
+ }
+
-+ disconnect_bsp_APIC(ioapic_i8259.pin != -1);
-+#endif
-+}
++ if (processor_cntl_external())
++ processor_notify_external(pr,
++ PROCESSOR_HOTPLUG, HOTPLUG_TYPE_ADD);
+
+ if (pr->id >= 0 && (pr->id < NR_CPUS)) {
+ kobject_uevent(&device->kobj, KOBJ_OFFLINE);
+@@ -777,8 +811,18 @@
+ return;
+ }
+
++#ifdef CONFIG_XEN
++ if ((pr->id >= 0) && (pr->id < NR_CPUS)
++ && (cpu_present(pr->id)))
++#else
+ if ((pr->id < NR_CPUS) && (cpu_present(pr->id)))
++#endif /* CONFIG_XEN */
+ kobject_uevent(&device->kobj, KOBJ_OFFLINE);
++
++ if (processor_cntl_external())
++ processor_notify_external(pr, PROCESSOR_HOTPLUG,
++ HOTPLUG_TYPE_REMOVE);
++
+ break;
+ default:
+ ACPI_DEBUG_PRINT((ACPI_DB_INFO,
+@@ -843,6 +887,11 @@
+
+ static int acpi_processor_handle_eject(struct acpi_processor *pr)
+ {
++#ifdef CONFIG_XEN
++ if (pr->id == -1)
++ return (0);
++#endif /* CONFIG_XEN */
+
+ if (cpu_online(pr->id)) {
+ return (-EINVAL);
+ }
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/processor_extcntl.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/acpi/processor_extcntl.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,241 @@
+/*
-+ * function to set the IO-APIC physical IDs based on the
-+ * values stored in the MPC table.
++ * processor_extcntl.c - channel to external control logic
++ *
++ * Copyright (C) 2008, Intel corporation
++ *
++ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or (at
++ * your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful, but
++ * WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
++ * General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License along
++ * with this program; if not, write to the Free Software Foundation, Inc.,
++ * 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.
+ *
-+ * by Matt Domsch <Matt_Domsch at dell.com> Tue Dec 21 12:25:05 CST 1999
+ */
+
-+#ifndef CONFIG_XEN
-+static void __init setup_ioapic_ids_from_mpc (void)
-+{
-+ union IO_APIC_reg_00 reg_00;
-+ int apic;
-+ int i;
-+ unsigned char old_id;
-+ unsigned long flags;
-+
-+ /*
-+ * Set the IOAPIC ID to the value stored in the MPC table.
-+ */
-+ for (apic = 0; apic < nr_ioapics; apic++) {
++#include <linux/kernel.h>
++#include <linux/init.h>
++#include <linux/types.h>
++#include <linux/acpi.h>
++#include <linux/pm.h>
++#include <linux/cpu.h>
+
-+ /* Read the register 0 value */
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ reg_00.raw = io_apic_read(apic, 0);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+
-+ old_id = mp_ioapics[apic].mpc_apicid;
++#include <acpi/processor.h>
+
++#define ACPI_PROCESSOR_COMPONENT 0x01000000
++#define ACPI_PROCESSOR_CLASS "processor"
++#define ACPI_PROCESSOR_DRIVER_NAME "ACPI Processor Driver"
++#define _COMPONENT ACPI_PROCESSOR_COMPONENT
++ACPI_MODULE_NAME("acpi_processor")
+
-+ printk(KERN_INFO "Using IO-APIC %d\n", mp_ioapics[apic].mpc_apicid);
++static int processor_extcntl_parse_csd(struct acpi_processor *pr);
++static int processor_extcntl_get_performance(struct acpi_processor *pr);
++/*
++ * External processor control logic may register with its own set of
++ * ops to get ACPI related notification. One example is like VMM.
++ */
++const struct processor_extcntl_ops *processor_extcntl_ops;
++EXPORT_SYMBOL(processor_extcntl_ops);
+
++static int processor_notify_smm(void)
++{
++ acpi_status status;
++ static int is_done = 0;
+
-+ /*
-+ * We need to adjust the IRQ routing table
-+ * if the ID changed.
-+ */
-+ if (old_id != mp_ioapics[apic].mpc_apicid)
-+ for (i = 0; i < mp_irq_entries; i++)
-+ if (mp_irqs[i].mpc_dstapic == old_id)
-+ mp_irqs[i].mpc_dstapic
-+ = mp_ioapics[apic].mpc_apicid;
++ /* only need successfully notify BIOS once */
++ /* avoid double notification which may lead to unexpected result */
++ if (is_done)
++ return 0;
+
-+ /*
-+ * Read the right value from the MPC table and
-+ * write it into the ID register.
-+ */
-+ apic_printk(APIC_VERBOSE,KERN_INFO "...changing IO-APIC physical APIC ID to %d ...",
-+ mp_ioapics[apic].mpc_apicid);
++ /* Can't write pstate_cnt to smi_cmd if either value is zero */
++ if ((!acpi_fadt.smi_cmd) || (!acpi_fadt.pstate_cnt)) {
++ ACPI_DEBUG_PRINT((ACPI_DB_INFO,"No SMI port or pstate_cnt\n"));
++ return 0;
++ }
+
-+ reg_00.bits.ID = mp_ioapics[apic].mpc_apicid;
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ io_apic_write(apic, 0, reg_00.raw);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
++ ACPI_DEBUG_PRINT((ACPI_DB_INFO,
++ "Writing pstate_cnt [0x%x] to smi_cmd [0x%x]\n",
++ acpi_fadt.pstate_cnt, acpi_fadt.smi_cmd));
+
-+ /*
-+ * Sanity check
-+ */
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ reg_00.raw = io_apic_read(apic, 0);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+ if (reg_00.bits.ID != mp_ioapics[apic].mpc_apicid)
-+ printk("could not set ID!\n");
-+ else
-+ apic_printk(APIC_VERBOSE," ok.\n");
++ /* FADT v1 doesn't support pstate_cnt, many BIOS vendors use
++ * it anyway, so we need to support it... */
++ if (acpi_fadt_is_v1) {
++ ACPI_DEBUG_PRINT((ACPI_DB_INFO,
++ "Using v1.0 FADT reserved value for pstate_cnt\n"));
+ }
-+}
-+#else
-+static void __init setup_ioapic_ids_from_mpc(void) { }
-+#endif
-+
-+/*
-+ * There is a nasty bug in some older SMP boards, their mptable lies
-+ * about the timer IRQ. We do the following to work around the situation:
-+ *
-+ * - timer IRQ defaults to IO-APIC IRQ
-+ * - if this function detects that timer IRQs are defunct, then we fall
-+ * back to ISA timer IRQs
-+ */
-+#ifndef CONFIG_XEN
-+static int __init timer_irq_works(void)
-+{
-+ unsigned long t1 = jiffies;
+
-+ local_irq_enable();
-+ /* Let ten ticks pass... */
-+ mdelay((10 * 1000) / HZ);
++ status = acpi_os_write_port(acpi_fadt.smi_cmd,
++ (u32) acpi_fadt.pstate_cnt, 8);
++ if (ACPI_FAILURE(status))
++ return status;
+
-+ /*
-+ * Expect a few ticks at least, to be sure some possible
-+ * glue logic does not lock up after one or two first
-+ * ticks in a non-ExtINT mode. Also the local APIC
-+ * might have cached one ExtINT interrupt. Finally, at
-+ * least one tick may be lost due to delays.
-+ */
++ is_done = 1;
+
-+ /* jiffies wrap? */
-+ if (jiffies - t1 > 4)
-+ return 1;
+ return 0;
+}
+
-+/*
-+ * In the SMP+IOAPIC case it might happen that there are an unspecified
-+ * number of pending IRQ events unhandled. These cases are very rare,
-+ * so we 'resend' these IRQs via IPIs, to the same CPU. It's much
-+ * better to do it this way as thus we do not have to be aware of
-+ * 'pending' interrupts in the IRQ path, except at this point.
-+ */
-+/*
-+ * Edge triggered needs to resend any interrupt
-+ * that was delayed but this is now handled in the device
-+ * independent code.
-+ */
++int processor_notify_external(struct acpi_processor *pr, int event, int type)
++{
++ int ret = -EINVAL;
+
-+/*
-+ * Starting up a edge-triggered IO-APIC interrupt is
-+ * nasty - we need to make sure that we get the edge.
-+ * If it is already asserted for some reason, we need
-+ * return 1 to indicate that is was pending.
-+ *
-+ * This is not complete - we should be able to fake
-+ * an edge even if it isn't on the 8259A...
-+ */
++ if (!processor_cntl_external())
++ return -EINVAL;
+
-+static unsigned int startup_edge_ioapic_irq(unsigned int irq)
-+{
-+ int was_pending = 0;
-+ unsigned long flags;
++ switch (event) {
++ case PROCESSOR_PM_INIT:
++ case PROCESSOR_PM_CHANGE:
++ if ((type >= PM_TYPE_MAX) ||
++ !processor_extcntl_ops->pm_ops[type])
++ break;
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ if (irq < 16) {
-+ disable_8259A_irq(irq);
-+ if (i8259A_irq_pending(irq))
-+ was_pending = 1;
++ ret = processor_extcntl_ops->pm_ops[type](pr, event);
++ break;
++ case PROCESSOR_HOTPLUG:
++ if (processor_extcntl_ops->hotplug)
++ ret = processor_extcntl_ops->hotplug(pr, type);
++ break;
++ default:
++ printk(KERN_ERR "Unsupport processor events %d.\n", event);
++ break;
+ }
-+ __unmask_IO_APIC_irq(irq);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
+
-+ return was_pending;
++ return ret;
+}
+
+/*
-+ * Once we have recorded IRQ_PENDING already, we can mask the
-+ * interrupt for real. This prevents IRQ storms from unhandled
-+ * devices.
++ * External control logic can decide to grab full or part of physical
++ * processor control bits. Take a VMM for example, physical processors
++ * are owned by VMM and thus existence information like hotplug is
++ * always required to be notified to VMM. Similar is processor idle
++ * state which is also necessarily controlled by VMM. But for other
++ * control bits like performance/throttle states, VMM may choose to
++ * control or not upon its own policy.
+ */
-+static void ack_edge_ioapic_irq(unsigned int irq)
++void processor_extcntl_init(void)
+{
-+ move_irq(irq);
-+ if ((irq_desc[irq].status & (IRQ_PENDING | IRQ_DISABLED))
-+ == (IRQ_PENDING | IRQ_DISABLED))
-+ mask_IO_APIC_irq(irq);
-+ ack_APIC_irq();
++ if (!processor_extcntl_ops)
++ arch_acpi_processor_init_extcntl(&processor_extcntl_ops);
+}
+
+/*
-+ * Level triggered interrupts can just be masked,
-+ * and shutting down and starting up the interrupt
-+ * is the same as enabling and disabling them -- except
-+ * with a startup need to return a "was pending" value.
-+ *
-+ * Level triggered interrupts are special because we
-+ * do not touch any IO-APIC register while handling
-+ * them. We ack the APIC in the end-IRQ handler, not
-+ * in the start-IRQ-handler. Protection against reentrance
-+ * from the same interrupt is still provided, both by the
-+ * generic IRQ layer and by the fact that an unacked local
-+ * APIC does not accept IRQs.
++ * This is called from ACPI processor init, and targeted to hold
++ * some tricky housekeeping jobs to satisfy external control model.
++ * For example, we may put dependency parse stub here for idle
++ * and performance state. Those information may be not available
++ * if splitting from dom0 control logic like cpufreq driver.
+ */
-+static unsigned int startup_level_ioapic_irq (unsigned int irq)
-+{
-+ unmask_IO_APIC_irq(irq);
-+
-+ return 0; /* don't check for pending */
-+}
-+
-+static void end_level_ioapic_irq (unsigned int irq)
++int processor_extcntl_prepare(struct acpi_processor *pr)
+{
-+ move_irq(irq);
-+ ack_APIC_irq();
-+}
++ /* parse cstate dependency information */
++ if (processor_pm_external())
++ processor_extcntl_parse_csd(pr);
+
-+#ifdef CONFIG_PCI_MSI
-+static unsigned int startup_edge_ioapic_vector(unsigned int vector)
-+{
-+ int irq = vector_to_irq(vector);
++ /* Initialize performance states */
++ if (processor_pmperf_external())
++ processor_extcntl_get_performance(pr);
+
-+ return startup_edge_ioapic_irq(irq);
++ return 0;
+}
+
-+static void ack_edge_ioapic_vector(unsigned int vector)
++/*
++ * Currently no _CSD is implemented which is why existing ACPI code
++ * doesn't parse _CSD at all. But to keep interface complete with
++ * external control logic, we put a placeholder here for future
++ * compatibility.
++ */
++static int processor_extcntl_parse_csd(struct acpi_processor *pr)
+{
-+ int irq = vector_to_irq(vector);
++ int i;
+
-+ move_native_irq(vector);
-+ ack_edge_ioapic_irq(irq);
-+}
++ for (i = 0; i < pr->power.count; i++) {
++ if (!pr->power.states[i].valid)
++ continue;
+
-+static unsigned int startup_level_ioapic_vector (unsigned int vector)
-+{
-+ int irq = vector_to_irq(vector);
++ /* No dependency by default */
++ pr->power.states[i].domain_info = NULL;
++ pr->power.states[i].csd_count = 0;
++ }
+
-+ return startup_level_ioapic_irq (irq);
++ return 0;
+}
+
-+static void end_level_ioapic_vector (unsigned int vector)
++/*
++ * Existing ACPI module does parse performance states at some point,
++ * when acpi-cpufreq driver is loaded which however is something
++ * we'd like to disable to avoid confliction with external control
++ * logic. So we have to collect raw performance information here
++ * when ACPI processor object is found and started.
++ */
++static int processor_extcntl_get_performance(struct acpi_processor *pr)
+{
-+ int irq = vector_to_irq(vector);
-+
-+ move_native_irq(vector);
-+ end_level_ioapic_irq(irq);
-+}
++ int ret;
++ struct acpi_processor_performance *perf;
++ struct acpi_psd_package *pdomain;
+
-+static void mask_IO_APIC_vector (unsigned int vector)
-+{
-+ int irq = vector_to_irq(vector);
++ if (pr->performance)
++ return -EBUSY;
+
-+ mask_IO_APIC_irq(irq);
-+}
++ perf = kzalloc(sizeof(struct acpi_processor_performance), GFP_KERNEL);
++ if (!perf)
++ return -ENOMEM;
+
-+static void unmask_IO_APIC_vector (unsigned int vector)
-+{
-+ int irq = vector_to_irq(vector);
++ pr->performance = perf;
++ /* Get basic performance state information */
++ ret = acpi_processor_get_performance_info(pr);
++ if (ret < 0)
++ goto err_out;
+
-+ unmask_IO_APIC_irq(irq);
-+}
++ /*
++ * Well, here we need retrieve performance dependency information
++ * from _PSD object. The reason why existing interface is not used
++ * is due to the reason that existing interface sticks to Linux cpu
++ * id to construct some bitmap, however we want to split ACPI
++ * processor objects from Linux cpu id logic. For example, even
++ * when Linux is configured as UP, we still want to parse all ACPI
++ * processor objects to external logic. In this case, it's preferred
++ * to use ACPI ID instead.
++ */
++ pdomain = &pr->performance->domain_info;
++ pdomain->num_processors = 0;
++ ret = acpi_processor_get_psd(pr);
++ if (ret < 0) {
++ /*
++ * _PSD is optional - assume no coordination if absent (or
++ * broken), matching native kernels' behavior.
++ */
++ pdomain->num_entries = ACPI_PSD_REV0_ENTRIES;
++ pdomain->revision = ACPI_PSD_REV0_REVISION;
++ pdomain->domain = pr->acpi_id;
++ pdomain->coord_type = DOMAIN_COORD_TYPE_SW_ALL;
++ pdomain->num_processors = 1;
++ }
+
-+#ifdef CONFIG_SMP
-+static void set_ioapic_affinity_vector (unsigned int vector,
-+ cpumask_t cpu_mask)
-+{
-+ int irq = vector_to_irq(vector);
++ /* Some sanity check */
++ if ((pdomain->revision != ACPI_PSD_REV0_REVISION) ||
++ (pdomain->num_entries != ACPI_PSD_REV0_ENTRIES) ||
++ ((pdomain->coord_type != DOMAIN_COORD_TYPE_SW_ALL) &&
++ (pdomain->coord_type != DOMAIN_COORD_TYPE_SW_ANY) &&
++ (pdomain->coord_type != DOMAIN_COORD_TYPE_HW_ALL))) {
++ ret = -EINVAL;
++ goto err_out;
++ }
+
-+ set_native_irq_info(vector, cpu_mask);
-+ set_ioapic_affinity_irq(irq, cpu_mask);
-+}
-+#endif // CONFIG_SMP
-+#endif // CONFIG_PCI_MSI
++ /* Last step is to notify BIOS that external logic exists */
++ processor_notify_smm();
+
-+static int ioapic_retrigger(unsigned int irq)
-+{
-+ send_IPI_self(IO_APIC_VECTOR(irq));
++ processor_notify_external(pr, PROCESSOR_PM_INIT, PM_TYPE_PERF);
+
-+ return 1;
++ return 0;
++err_out:
++ pr->performance = NULL;
++ kfree(perf);
++ return ret;
+}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/processor_idle.c
+--- a/drivers/acpi/processor_idle.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/acpi/processor_idle.c Wed Sep 10 10:54:08 2008 +0100
+@@ -714,8 +714,12 @@
+ (reg->space_id != ACPI_ADR_SPACE_FIXED_HARDWARE))
+ continue;
+
+- cx.address = (reg->space_id == ACPI_ADR_SPACE_FIXED_HARDWARE) ?
+- 0 : reg->address;
++ if (!processor_pm_external())
++ cx.address = (reg->space_id ==
++ ACPI_ADR_SPACE_FIXED_HARDWARE) ?
++ 0 : reg->address;
++ else
++ cx.address = reg->address;
+
+ /* There should be an easy way to extract an integer... */
+ obj = (union acpi_object *)&(element->package.elements[1]);
+@@ -724,7 +728,9 @@
+
+ cx.type = obj->integer.value;
+
+- if ((cx.type != ACPI_STATE_C1) &&
++ /* Following check doesn't apply to external control case */
++ if (!processor_pm_external() &&
++ (cx.type != ACPI_STATE_C1) &&
+ (reg->space_id != ACPI_ADR_SPACE_SYSTEM_IO))
+ continue;
+
+@@ -742,6 +748,12 @@
+ continue;
+
+ cx.power = obj->integer.value;
+
-+/*
-+ * Level and edge triggered IO-APIC interrupts need different handling,
-+ * so we use two separate IRQ descriptors. Edge triggered IRQs can be
-+ * handled with the level-triggered descriptor, but that one has slightly
-+ * more overhead. Level-triggered interrupts cannot be handled with the
-+ * edge-triggered handler, without risking IRQ storms and other ugly
-+ * races.
-+ */
-+
-+static struct hw_interrupt_type ioapic_edge_type __read_mostly = {
-+ .typename = "IO-APIC-edge",
-+ .startup = startup_edge_ioapic,
-+ .shutdown = shutdown_edge_ioapic,
-+ .enable = enable_edge_ioapic,
-+ .disable = disable_edge_ioapic,
-+ .ack = ack_edge_ioapic,
-+ .end = end_edge_ioapic,
-+#ifdef CONFIG_SMP
-+ .set_affinity = set_ioapic_affinity,
++#ifdef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++ /* cache control methods to notify external logic */
++ if (processor_pm_external())
++ memcpy(&cx.reg, reg, sizeof(*reg));
+#endif
-+ .retrigger = ioapic_retrigger,
-+};
-+
-+static struct hw_interrupt_type ioapic_level_type __read_mostly = {
-+ .typename = "IO-APIC-level",
-+ .startup = startup_level_ioapic,
-+ .shutdown = shutdown_level_ioapic,
-+ .enable = enable_level_ioapic,
-+ .disable = disable_level_ioapic,
-+ .ack = mask_and_ack_level_ioapic,
-+ .end = end_level_ioapic,
-+#ifdef CONFIG_SMP
-+ .set_affinity = set_ioapic_affinity,
+
+ current_count++;
+ memcpy(&(pr->power.states[current_count]), &cx, sizeof(cx));
+@@ -985,12 +997,16 @@
+ return -ENODEV;
+
+ /* Fall back to the default idle loop */
+- pm_idle = pm_idle_save;
++ if (!processor_pm_external())
++ pm_idle = pm_idle_save;
+ synchronize_sched(); /* Relies on interrupts forcing exit from idle. */
+
+ pr->flags.power = 0;
+ result = acpi_processor_get_power_info(pr);
+- if ((pr->flags.power == 1) && (pr->flags.power_setup_done))
++ if (processor_pm_external())
++ processor_notify_external(pr,
++ PROCESSOR_PM_CHANGE, PM_TYPE_IDLE);
++ else if ((pr->flags.power == 1) && (pr->flags.power_setup_done))
+ pm_idle = acpi_processor_idle;
+
+ return result;
+@@ -1122,7 +1138,7 @@
+ pr->power.states[i].type);
+ printk(")\n");
+
+- if (pr->id == 0) {
++ if (!processor_pm_external() && (pr->id == 0)) {
+ pm_idle_save = pm_idle;
+ pm_idle = acpi_processor_idle;
+ }
+@@ -1141,6 +1157,9 @@
+
+ pr->flags.power_setup_done = 1;
+
++ if (processor_pm_external())
++ processor_notify_external(pr,
++ PROCESSOR_PM_INIT, PM_TYPE_IDLE);
+ return 0;
+ }
+
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/processor_perflib.c
+--- a/drivers/acpi/processor_perflib.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/acpi/processor_perflib.c Wed Sep 10 10:54:08 2008 +0100
+@@ -66,6 +66,7 @@
+
+ static int acpi_processor_ppc_status = 0;
+
++#ifdef CONFIG_CPU_FREQ
+ static int acpi_processor_ppc_notifier(struct notifier_block *nb,
+ unsigned long event, void *data)
+ {
+@@ -102,6 +103,7 @@
+ static struct notifier_block acpi_ppc_notifier_block = {
+ .notifier_call = acpi_processor_ppc_notifier,
+ };
++#endif /* CONFIG_CPU_FREQ */
+
+ static int acpi_processor_get_platform_limit(struct acpi_processor *pr)
+ {
+@@ -137,9 +139,15 @@
+ if (ret < 0)
+ return (ret);
+ else
++#ifdef CONFIG_CPU_FREQ
+ return cpufreq_update_policy(pr->id);
++#elif CONFIG_PROCESSOR_EXTERNAL_CONTROL
++ return processor_notify_external(pr,
++ PROCESSOR_PM_CHANGE, PM_TYPE_PERF);
+#endif
-+ .retrigger = ioapic_retrigger,
-+};
-+#endif /* !CONFIG_XEN */
-+
-+static inline void init_IO_APIC_traps(void)
-+{
-+ int irq;
-+
-+ /*
-+ * NOTE! The local APIC isn't very good at handling
-+ * multiple interrupts at the same interrupt level.
-+ * As the interrupt level is determined by taking the
-+ * vector number and shifting that right by 4, we
-+ * want to spread these out a bit so that they don't
-+ * all fall in the same interrupt level.
-+ *
-+ * Also, we've got to be careful not to trash gate
-+ * 0x80, because int 0x80 is hm, kind of importantish. ;)
-+ */
-+ for (irq = 0; irq < NR_IRQS ; irq++) {
-+ int tmp = irq;
-+ if (use_pci_vector()) {
-+ if (!platform_legacy_irq(tmp))
-+ if ((tmp = vector_to_irq(tmp)) == -1)
-+ continue;
-+ }
-+ if (IO_APIC_IRQ(tmp) && !IO_APIC_VECTOR(tmp)) {
-+ /*
-+ * Hmm.. We don't have an entry for this,
-+ * so default to an old-fashioned 8259
-+ * interrupt if we can..
-+ */
-+ if (irq < 16)
-+ make_8259A_irq(irq);
-+#ifndef CONFIG_XEN
-+ else
-+ /* Strange. Oh, well.. */
-+ irq_desc[irq].chip = &no_irq_type;
+ }
+
++#ifdef CONFIG_CPU_FREQ
+ void acpi_processor_ppc_init(void)
+ {
+ if (!cpufreq_register_notifier
+@@ -158,6 +166,7 @@
+
+ acpi_processor_ppc_status &= ~PPC_REGISTERED;
+ }
++#endif /* CONFIG_CPU_FREQ */
+
+ static int acpi_processor_get_performance_control(struct acpi_processor *pr)
+ {
+@@ -299,7 +308,10 @@
+ return result;
+ }
+
+-static int acpi_processor_get_performance_info(struct acpi_processor *pr)
++#ifndef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++static
++#endif
++int acpi_processor_get_performance_info(struct acpi_processor *pr)
+ {
+ int result = 0;
+ acpi_status status = AE_OK;
+@@ -331,6 +343,7 @@
+ return 0;
+ }
+
++#ifdef CONFIG_CPU_FREQ
+ int acpi_processor_notify_smm(struct module *calling_module)
+ {
+ acpi_status status;
+@@ -398,6 +411,7 @@
+ }
+
+ EXPORT_SYMBOL(acpi_processor_notify_smm);
++#endif /* CONFIG_CPU_FREQ */
+
+ #ifdef CONFIG_X86_ACPI_CPUFREQ_PROC_INTF
+ /* /proc/acpi/processor/../performance interface (DEPRECATED) */
+@@ -538,7 +552,10 @@
+ }
+ #endif /* CONFIG_X86_ACPI_CPUFREQ_PROC_INTF */
+
+-static int acpi_processor_get_psd(struct acpi_processor *pr)
++#ifndef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++static
++#endif
++int acpi_processor_get_psd(struct acpi_processor *pr)
+ {
+ int result = 0;
+ acpi_status status = AE_OK;
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/resources/rsxface.c
+--- a/drivers/acpi/resources/rsxface.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/acpi/resources/rsxface.c Wed Sep 10 10:54:08 2008 +0100
+@@ -476,8 +476,6 @@
+ return (AE_CTRL_TERMINATE);
+ }
+
+-ACPI_EXPORT_SYMBOL(acpi_rs_match_vendor_resource)
+-
+ /*******************************************************************************
+ *
+ * FUNCTION: acpi_walk_resources
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/sleep/main.c
+--- a/drivers/acpi/sleep/main.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/acpi/sleep/main.c Wed Sep 10 10:54:08 2008 +0100
+@@ -91,7 +91,14 @@
+ break;
+
+ case PM_SUSPEND_MEM:
++#ifdef CONFIG_ACPI_PV_SLEEP
++ /* Hyperviosr will save and restore CPU context
++ * and then we can skip low level housekeeping here.
++ */
++ acpi_enter_sleep_state(acpi_state);
++#else
+ do_suspend_lowlevel();
++#endif
+ break;
+
+ case PM_SUSPEND_DISK:
+@@ -145,10 +152,12 @@
+ /* reset firmware waking vector */
+ acpi_set_firmware_waking_vector((acpi_physical_address) 0);
+
++#ifndef CONFIG_ACPI_PV_SLEEP
+ if (init_8259A_after_S1) {
+ printk("Broken toshiba laptop -> kicking interrupts\n");
+ init_8259A(0);
+ }
++#endif
+ return 0;
+ }
+
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/sleep/poweroff.c
+--- a/drivers/acpi/sleep/poweroff.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/acpi/sleep/poweroff.c Wed Sep 10 10:54:08 2008 +0100
+@@ -20,6 +20,7 @@
+ int acpi_sleep_prepare(u32 acpi_state)
+ {
+ #ifdef CONFIG_ACPI_SLEEP
++#ifndef CONFIG_ACPI_PV_SLEEP
+ /* do we have a wakeup address for S2 and S3? */
+ if (acpi_state == ACPI_STATE_S3) {
+ if (!acpi_wakeup_address) {
+@@ -30,6 +31,7 @@
+ acpi_wakeup_address));
+
+ }
+#endif
+ ACPI_FLUSH_CPU_CACHE();
+ acpi_enable_wakeup_device_prep(acpi_state);
+ #endif
+diff -r d894e36cfc30 -r 0aa021803deb drivers/acpi/sleep/proc.c
+--- a/drivers/acpi/sleep/proc.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/acpi/sleep/proc.c Wed Sep 10 10:54:08 2008 +0100
+@@ -251,6 +251,8 @@
+
+ if (adjust) {
+ yr += CMOS_READ(RTC_YEAR);
++ if (acpi_gbl_FADT->century)
++ yr += CMOS_READ(acpi_gbl_FADT->century) * 100;
+ mo += CMOS_READ(RTC_MONTH);
+ day += CMOS_READ(RTC_DAY_OF_MONTH);
+ hr += CMOS_READ(RTC_HOURS);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/block/floppy.c
+--- a/drivers/block/floppy.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/block/floppy.c Wed Sep 10 10:54:08 2008 +0100
+@@ -4392,11 +4392,15 @@
+ if (fd_request_dma()) {
+ DPRINT("Unable to grab DMA%d for the floppy driver\n",
+ FLOPPY_DMA);
+- fd_free_irq();
+- spin_lock_irqsave(&floppy_usage_lock, flags);
+- usage_count--;
+- spin_unlock_irqrestore(&floppy_usage_lock, flags);
+- return -1;
++ if (can_use_virtual_dma & 2)
++ use_virtual_dma = can_use_virtual_dma = 1;
++ if (!(can_use_virtual_dma & 1)) {
++ fd_free_irq();
++ spin_lock_irqsave(&floppy_usage_lock, flags);
++ usage_count--;
++ spin_unlock_irqrestore(&floppy_usage_lock, flags);
++ return -1;
+ }
+ }
+
+ for (fdc = 0; fdc < N_FDC; fdc++) {
+diff -r d894e36cfc30 -r 0aa021803deb drivers/char/Kconfig
+--- a/drivers/char/Kconfig Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/char/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -1013,7 +1013,7 @@
+ config HPET
+ bool "HPET - High Precision Event Timer" if (X86 || IA64)
+ default n
+- depends on ACPI
++ depends on ACPI && !XEN
+ help
+ If you say Y here, you will have a miscdevice named "/dev/hpet/". Each
+ open selects one of the timers supported by the HPET. The timers are
+diff -r d894e36cfc30 -r 0aa021803deb drivers/char/agp/amd64-agp.c
+--- a/drivers/char/agp/amd64-agp.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/char/agp/amd64-agp.c Wed Sep 10 10:54:08 2008 +0100
+@@ -15,6 +15,7 @@
+ #include <linux/mmzone.h>
+ #include <asm/page.h> /* PAGE_SIZE */
+ #include <asm/k8.h>
++#include <asm/e820.h>
+ #include "agp.h"
+
+ /* PTE bits. */
+@@ -252,7 +253,6 @@
+ /* Some basic sanity checks for the aperture. */
+ static int __devinit aperture_valid(u64 aper, u32 size)
+ {
+- u32 pfn, c;
+ if (aper == 0) {
+ printk(KERN_ERR PFX "No aperture\n");
+ return 0;
+@@ -265,14 +265,9 @@
+ printk(KERN_ERR PFX "Aperture out of bounds\n");
+ return 0;
+ }
+- pfn = aper >> PAGE_SHIFT;
+- for (c = 0; c < size/PAGE_SIZE; c++) {
+- if (!pfn_valid(pfn + c))
+- break;
+- if (!PageReserved(pfn_to_page(pfn + c))) {
+- printk(KERN_ERR PFX "Aperture pointing to RAM\n");
+- return 0;
+- }
++ if (e820_any_mapped(aper, aper + size, E820_RAM)) {
++ printk(KERN_ERR PFX "Aperture pointing to RAM\n");
++ return 0;
+ }
+
+ /* Request the Aperture. This catches cases when someone else
+diff -r d894e36cfc30 -r 0aa021803deb drivers/char/agp/generic.c
+--- a/drivers/char/agp/generic.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/char/agp/generic.c Wed Sep 10 10:54:08 2008 +0100
+@@ -50,28 +50,6 @@
+ * nice to do this some other way instead of needing this export.
+ */
+ EXPORT_SYMBOL_GPL(agp_memory_reserved);
+-
+-#if defined(CONFIG_X86)
+-int map_page_into_agp(struct page *page)
+-{
+- int i;
+- i = change_page_attr(page, 1, PAGE_KERNEL_NOCACHE);
+- /* Caller's responsibility to call global_flush_tlb() for
+- * performance reasons */
+- return i;
+-}
+-EXPORT_SYMBOL_GPL(map_page_into_agp);
+-
+-int unmap_page_from_agp(struct page *page)
+-{
+- int i;
+- i = change_page_attr(page, 1, PAGE_KERNEL);
+- /* Caller's responsibility to call global_flush_tlb() for
+- * performance reasons */
+- return i;
+-}
+-EXPORT_SYMBOL_GPL(unmap_page_from_agp);
+-#endif
+
+ /*
+ * Generic routines for handling agp_memory structures -
+diff -r d894e36cfc30 -r 0aa021803deb drivers/char/agp/intel-agp.c
+--- a/drivers/char/agp/intel-agp.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/char/agp/intel-agp.c Wed Sep 10 10:54:08 2008 +0100
+@@ -173,9 +173,17 @@
+ if (page == NULL)
+ return NULL;
+
++#ifdef CONFIG_XEN
++ if (xen_create_contiguous_region((unsigned long)page_address(page), 2, 32)) {
++ __free_pages(page, 2);
++ return NULL;
+ }
-+}
++#endif
+
-+#ifndef CONFIG_XEN
-+static void enable_lapic_irq (unsigned int irq)
-+{
-+ unsigned long v;
+ if (change_page_attr(page, 4, PAGE_KERNEL_NOCACHE) < 0) {
++ change_page_attr(page, 4, PAGE_KERNEL);
+ global_flush_tlb();
+- __free_page(page);
++ __free_pages(page, 2);
+ return NULL;
+ }
+ global_flush_tlb();
+@@ -195,9 +203,12 @@
+ page = virt_to_page(addr);
+ change_page_attr(page, 4, PAGE_KERNEL);
+ global_flush_tlb();
++#ifdef CONFIG_XEN
++ xen_destroy_contiguous_region((unsigned long)page_address(page), 2);
++#endif
+ put_page(page);
+ unlock_page(page);
+- free_pages((unsigned long)addr, 2);
++ __free_pages(page, 2);
+ atomic_dec(&agp_bridge->current_memory_agp);
+ }
+
+diff -r d894e36cfc30 -r 0aa021803deb drivers/char/mem.c
+--- a/drivers/char/mem.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/char/mem.c Wed Sep 10 10:54:08 2008 +0100
+@@ -101,6 +101,7 @@
+ }
+ #endif
+
++#ifndef ARCH_HAS_DEV_MEM
+ /*
+ * This funcion reads the *physical* memory. The f_pos points directly to the
+ * memory location.
+@@ -223,6 +224,7 @@
+ *ppos += written;
+ return written;
+ }
++#endif
+
+ #ifndef __HAVE_PHYS_MEM_ACCESS_PROT
+ static pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
+@@ -262,6 +264,9 @@
+ static int mmap_kmem(struct file * file, struct vm_area_struct * vma)
+ {
+ unsigned long pfn;
++#ifdef CONFIG_XEN
++ unsigned long i, count;
++#endif
+
+ /* Turn a kernel-virtual address into a physical page frame */
+ pfn = __pa((u64)vma->vm_pgoff << PAGE_SHIFT) >> PAGE_SHIFT;
+@@ -275,6 +280,13 @@
+ */
+ if (!pfn_valid(pfn))
+ return -EIO;
++
++#ifdef CONFIG_XEN
++ count = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
++ for (i = 0; i < count; i++)
++ if ((pfn + i) != mfn_to_local_pfn(pfn_to_mfn(pfn + i)))
++ return -EIO;
++#endif
+
+ vma->vm_pgoff = pfn;
+ return mmap_mem(file, vma);
+@@ -780,6 +792,7 @@
+ #define open_kmem open_mem
+ #define open_oldmem open_mem
+
++#ifndef ARCH_HAS_DEV_MEM
+ static const struct file_operations mem_fops = {
+ .llseek = memory_lseek,
+ .read = read_mem,
+@@ -787,6 +800,9 @@
+ .mmap = mmap_mem,
+ .open = open_mem,
+ };
++#else
++extern const struct file_operations mem_fops;
++#endif
+
+ static const struct file_operations kmem_fops = {
+ .llseek = memory_lseek,
+diff -r d894e36cfc30 -r 0aa021803deb drivers/char/tpm/Kconfig
+--- a/drivers/char/tpm/Kconfig Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/char/tpm/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -58,5 +58,13 @@
+ Further information on this driver and the supported hardware
+ can be found at http://www.prosec.rub.de/tpm
+
++config TCG_XEN
++ tristate "XEN TPM Interface"
++ depends on TCG_TPM && XEN
++ ---help---
++ If you want to make TPM support available to a Xen user domain,
++ say Yes and it will be accessible from within Linux.
++ To compile this driver as a module, choose M here; the module
++ will be called tpm_xenu.
+
-+ v = apic_read(APIC_LVT0);
-+ apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED);
-+}
+ endmenu
+-
+diff -r d894e36cfc30 -r 0aa021803deb drivers/char/tpm/Makefile
+--- a/drivers/char/tpm/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/char/tpm/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -9,3 +9,5 @@
+ obj-$(CONFIG_TCG_NSC) += tpm_nsc.o
+ obj-$(CONFIG_TCG_ATMEL) += tpm_atmel.o
+ obj-$(CONFIG_TCG_INFINEON) += tpm_infineon.o
++obj-$(CONFIG_TCG_XEN) += tpm_xenu.o
++tpm_xenu-y = tpm_xen.o tpm_vtpm.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/char/tpm/tpm.h
+--- a/drivers/char/tpm/tpm.h Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/char/tpm/tpm.h Wed Sep 10 10:54:08 2008 +0100
+@@ -105,6 +105,9 @@
+ struct dentry **bios_dir;
+
+ struct list_head list;
++#ifdef CONFIG_XEN
++ void *priv;
++#endif
+ };
+
+ #define to_tpm_chip(n) container_of(n, struct tpm_chip, vendor)
+@@ -120,6 +123,18 @@
+ outb(index, base);
+ outb(value & 0xFF, base+1);
+ }
+
-+static void disable_lapic_irq (unsigned int irq)
++#ifdef CONFIG_XEN
++static inline void *chip_get_private(const struct tpm_chip *chip)
+{
-+ unsigned long v;
-+
-+ v = apic_read(APIC_LVT0);
-+ apic_write(APIC_LVT0, v | APIC_LVT_MASKED);
++ return chip->priv;
+}
+
-+static void ack_lapic_irq (unsigned int irq)
++static inline void chip_set_private(struct tpm_chip *chip, void *priv)
+{
-+ ack_APIC_irq();
++ chip->priv = priv;
+}
++#endif
+
+ extern void tpm_get_timeouts(struct tpm_chip *);
+ extern void tpm_gen_interrupt(struct tpm_chip *);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/char/tpm/tpm_vtpm.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/char/tpm/tpm_vtpm.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,542 @@
++/*
++ * Copyright (C) 2006 IBM Corporation
++ *
++ * Authors:
++ * Stefan Berger <stefanb at us.ibm.com>
++ *
++ * Generic device driver part for device drivers in a virtualized
++ * environment.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License as
++ * published by the Free Software Foundation, version 2 of the
++ * License.
++ *
++ */
+
-+static void end_lapic_irq (unsigned int i) { /* nothing */ }
++#include <asm/uaccess.h>
++#include <linux/list.h>
++#include <linux/device.h>
++#include <linux/interrupt.h>
++#include <linux/platform_device.h>
++#include "tpm.h"
++#include "tpm_vtpm.h"
+
-+static struct hw_interrupt_type lapic_irq_type __read_mostly = {
-+ .typename = "local-APIC-edge",
-+ .startup = NULL, /* startup_irq() not used for IRQ0 */
-+ .shutdown = NULL, /* shutdown_irq() not used for IRQ0 */
-+ .enable = enable_lapic_irq,
-+ .disable = disable_lapic_irq,
-+ .ack = ack_lapic_irq,
-+ .end = end_lapic_irq,
++/* read status bits */
++enum {
++ STATUS_BUSY = 0x01,
++ STATUS_DATA_AVAIL = 0x02,
++ STATUS_READY = 0x04
+};
+
-+static void setup_nmi (void)
-+{
-+ /*
-+ * Dirty trick to enable the NMI watchdog ...
-+ * We put the 8259A master into AEOI mode and
-+ * unmask on all local APICs LVT0 as NMI.
-+ *
-+ * The idea to use the 8259A in AEOI mode ('8259A Virtual Wire')
-+ * is from Maciej W. Rozycki - so we do not have to EOI from
-+ * the NMI handler or the timer interrupt.
-+ */
-+ printk(KERN_INFO "activating NMI Watchdog ...");
-+
-+ enable_NMI_through_LVT0(NULL);
-+
-+ printk(" done.\n");
-+}
-+
-+/*
-+ * This looks a bit hackish but it's about the only one way of sending
-+ * a few INTA cycles to 8259As and any associated glue logic. ICR does
-+ * not support the ExtINT mode, unfortunately. We need to send these
-+ * cycles as some i82489DX-based boards have glue logic that keeps the
-+ * 8259A interrupt line asserted until INTA. --macro
-+ */
-+static inline void unlock_ExtINT_logic(void)
-+{
-+ int apic, pin, i;
-+ struct IO_APIC_route_entry entry0, entry1;
-+ unsigned char save_control, save_freq_select;
-+ unsigned long flags;
++struct transmission {
++ struct list_head next;
+
-+ pin = find_isa_irq_pin(8, mp_INT);
-+ apic = find_isa_irq_apic(8, mp_INT);
-+ if (pin == -1)
-+ return;
++ unsigned char *request;
++ size_t request_len;
++ size_t request_buflen;
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ *(((int *)&entry0) + 1) = io_apic_read(apic, 0x11 + 2 * pin);
-+ *(((int *)&entry0) + 0) = io_apic_read(apic, 0x10 + 2 * pin);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+ clear_IO_APIC_pin(apic, pin);
++ unsigned char *response;
++ size_t response_len;
++ size_t response_buflen;
+
-+ memset(&entry1, 0, sizeof(entry1));
++ unsigned int flags;
++};
+
-+ entry1.dest_mode = 0; /* physical delivery */
-+ entry1.mask = 0; /* unmask IRQ now */
-+ entry1.dest.physical.physical_dest = hard_smp_processor_id();
-+ entry1.delivery_mode = dest_ExtINT;
-+ entry1.polarity = entry0.polarity;
-+ entry1.trigger = 0;
-+ entry1.vector = 0;
++enum {
++ TRANSMISSION_FLAG_WAS_QUEUED = 0x1
++};
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ io_apic_write(apic, 0x11 + 2 * pin, *(((int *)&entry1) + 1));
-+ io_apic_write(apic, 0x10 + 2 * pin, *(((int *)&entry1) + 0));
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
+
-+ save_control = CMOS_READ(RTC_CONTROL);
-+ save_freq_select = CMOS_READ(RTC_FREQ_SELECT);
-+ CMOS_WRITE((save_freq_select & ~RTC_RATE_SELECT) | 0x6,
-+ RTC_FREQ_SELECT);
-+ CMOS_WRITE(save_control | RTC_PIE, RTC_CONTROL);
++enum {
++ DATAEX_FLAG_QUEUED_ONLY = 0x1
++};
+
-+ i = 100;
-+ while (i-- > 0) {
-+ mdelay(10);
-+ if ((CMOS_READ(RTC_INTR_FLAGS) & RTC_PF) == RTC_PF)
-+ i -= 10;
-+ }
+
-+ CMOS_WRITE(save_control, RTC_CONTROL);
-+ CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT);
-+ clear_IO_APIC_pin(apic, pin);
++/* local variables */
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ io_apic_write(apic, 0x11 + 2 * pin, *(((int *)&entry0) + 1));
-+ io_apic_write(apic, 0x10 + 2 * pin, *(((int *)&entry0) + 0));
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+}
++/* local function prototypes */
++static int _vtpm_send_queued(struct tpm_chip *chip);
+
-+int timer_uses_ioapic_pin_0;
+
-+/*
-+ * This code may look a bit paranoid, but it's supposed to cooperate with
-+ * a wide range of boards and BIOS bugs. Fortunately only the timer IRQ
-+ * is so screwy. Thanks to Brian Perkins for testing/hacking this beast
-+ * fanatically on his truly buggy board.
-+ *
-+ * FIXME: really need to revamp this for modern platforms only.
++/* =============================================================
++ * Some utility functions
++ * =============================================================
+ */
-+static inline void check_timer(void)
++static void vtpm_state_init(struct vtpm_state *vtpms)
+{
-+ int apic1, pin1, apic2, pin2;
-+ int vector;
-+
-+ /*
-+ * get/set the timer IRQ vector:
-+ */
-+ disable_8259A_irq(0);
-+ vector = assign_irq_vector(0);
-+ set_intr_gate(vector, interrupt[0]);
-+
-+ /*
-+ * Subtle, code in do_timer_interrupt() expects an AEOI
-+ * mode for the 8259A whenever interrupts are routed
-+ * through I/O APICs. Also IRQ0 has to be enabled in
-+ * the 8259A which implies the virtual wire has to be
-+ * disabled in the local APIC.
-+ */
-+ apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
-+ init_8259A(1);
-+ if (timer_over_8254 > 0)
-+ enable_8259A_irq(0);
++ vtpms->current_request = NULL;
++ spin_lock_init(&vtpms->req_list_lock);
++ init_waitqueue_head(&vtpms->req_wait_queue);
++ INIT_LIST_HEAD(&vtpms->queued_requests);
+
-+ pin1 = find_isa_irq_pin(0, mp_INT);
-+ apic1 = find_isa_irq_apic(0, mp_INT);
-+ pin2 = ioapic_i8259.pin;
-+ apic2 = ioapic_i8259.apic;
++ vtpms->current_response = NULL;
++ spin_lock_init(&vtpms->resp_list_lock);
++ init_waitqueue_head(&vtpms->resp_wait_queue);
+
-+ if (pin1 == 0)
-+ timer_uses_ioapic_pin_0 = 1;
++ vtpms->disconnect_time = jiffies;
++}
+
-+ apic_printk(APIC_VERBOSE,KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n",
-+ vector, apic1, pin1, apic2, pin2);
+
-+ if (pin1 != -1) {
-+ /*
-+ * Ok, does IRQ0 through the IOAPIC work?
-+ */
-+ unmask_IO_APIC_irq(0);
-+ if (!no_timer_check && timer_irq_works()) {
-+ nmi_watchdog_default();
-+ if (nmi_watchdog == NMI_IO_APIC) {
-+ disable_8259A_irq(0);
-+ setup_nmi();
-+ enable_8259A_irq(0);
-+ }
-+ if (disable_timer_pin_1 > 0)
-+ clear_IO_APIC_pin(0, pin1);
-+ return;
-+ }
-+ clear_IO_APIC_pin(apic1, pin1);
-+ apic_printk(APIC_QUIET,KERN_ERR "..MP-BIOS bug: 8254 timer not "
-+ "connected to IO-APIC\n");
-+ }
++static inline struct transmission *transmission_alloc(void)
++{
++ return kzalloc(sizeof(struct transmission), GFP_ATOMIC);
++}
+
-+ apic_printk(APIC_VERBOSE,KERN_INFO "...trying to set up timer (IRQ0) "
-+ "through the 8259A ... ");
-+ if (pin2 != -1) {
-+ apic_printk(APIC_VERBOSE,"\n..... (found apic %d pin %d) ...",
-+ apic2, pin2);
-+ /*
-+ * legacy devices should be connected to IO APIC #0
-+ */
-+ setup_ExtINT_IRQ0_pin(apic2, pin2, vector);
-+ if (timer_irq_works()) {
-+ apic_printk(APIC_VERBOSE," works.\n");
-+ nmi_watchdog_default();
-+ if (nmi_watchdog == NMI_IO_APIC) {
-+ setup_nmi();
-+ }
-+ return;
++static unsigned char *
++transmission_set_req_buffer(struct transmission *t,
++ unsigned char *buffer, size_t len)
++{
++ if (t->request_buflen < len) {
++ kfree(t->request);
++ t->request = kmalloc(len, GFP_KERNEL);
++ if (!t->request) {
++ t->request_buflen = 0;
++ return NULL;
+ }
-+ /*
-+ * Cleanup, just in case ...
-+ */
-+ clear_IO_APIC_pin(apic2, pin2);
-+ }
-+ apic_printk(APIC_VERBOSE," failed.\n");
-+
-+ if (nmi_watchdog == NMI_IO_APIC) {
-+ printk(KERN_WARNING "timer doesn't work through the IO-APIC - disabling NMI Watchdog!\n");
-+ nmi_watchdog = 0;
++ t->request_buflen = len;
+ }
+
-+ apic_printk(APIC_VERBOSE, KERN_INFO "...trying to set up timer as Virtual Wire IRQ...");
++ memcpy(t->request, buffer, len);
++ t->request_len = len;
+
-+ disable_8259A_irq(0);
-+ irq_desc[0].chip = &lapic_irq_type;
-+ apic_write(APIC_LVT0, APIC_DM_FIXED | vector); /* Fixed mode */
-+ enable_8259A_irq(0);
++ return t->request;
++}
+
-+ if (timer_irq_works()) {
-+ apic_printk(APIC_VERBOSE," works.\n");
-+ return;
++static unsigned char *
++transmission_set_res_buffer(struct transmission *t,
++ const unsigned char *buffer, size_t len)
++{
++ if (t->response_buflen < len) {
++ kfree(t->response);
++ t->response = kmalloc(len, GFP_ATOMIC);
++ if (!t->response) {
++ t->response_buflen = 0;
++ return NULL;
++ }
++ t->response_buflen = len;
+ }
-+ apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | vector);
-+ apic_printk(APIC_VERBOSE," failed.\n");
-+
-+ apic_printk(APIC_VERBOSE, KERN_INFO "...trying to set up timer as ExtINT IRQ...");
-+
-+ init_8259A(0);
-+ make_8259A_irq(0);
-+ apic_write(APIC_LVT0, APIC_DM_EXTINT);
+
-+ unlock_ExtINT_logic();
++ memcpy(t->response, buffer, len);
++ t->response_len = len;
+
-+ if (timer_irq_works()) {
-+ apic_printk(APIC_VERBOSE," works.\n");
-+ return;
-+ }
-+ apic_printk(APIC_VERBOSE," failed :(.\n");
-+ panic("IO-APIC + timer doesn't work! Try using the 'noapic' kernel parameter\n");
++ return t->response;
+}
-+#else
-+#define check_timer() ((void)0)
-+int timer_uses_ioapic_pin_0 = 0;
-+#endif /* !CONFIG_XEN */
+
-+static int __init notimercheck(char *s)
++static inline void transmission_free(struct transmission *t)
+{
-+ no_timer_check = 1;
-+ return 1;
++ kfree(t->request);
++ kfree(t->response);
++ kfree(t);
+}
-+__setup("no_timer_check", notimercheck);
+
++/* =============================================================
++ * Interface with the lower layer driver
++ * =============================================================
++ */
+/*
-+ *
-+ * IRQ's that are handled by the PIC in the MPS IOAPIC case.
-+ * - IRQ2 is the cascade IRQ, and cannot be a io-apic IRQ.
-+ * Linux doesn't really care, as it's not actually used
-+ * for any interrupt handling anyway.
++ * Lower layer uses this function to make a response available.
+ */
-+#define PIC_IRQS (1<<2)
-+
-+void __init setup_IO_APIC(void)
++int vtpm_vd_recv(const struct tpm_chip *chip,
++ const unsigned char *buffer, size_t count,
++ void *ptr)
+{
-+ enable_IO_APIC();
-+
-+ if (acpi_ioapic)
-+ io_apic_irqs = ~0; /* all IRQs go through IOAPIC */
-+ else
-+ io_apic_irqs = ~PIC_IRQS;
++ unsigned long flags;
++ int ret_size = 0;
++ struct transmission *t;
++ struct vtpm_state *vtpms;
+
-+ apic_printk(APIC_VERBOSE, "ENABLING IO-APIC IRQs\n");
++ vtpms = (struct vtpm_state *)chip_get_private(chip);
+
+ /*
-+ * Set up the IO-APIC IRQ routing table.
++ * The list with requests must contain one request
++ * only and the element there must be the one that
++ * was passed to me from the front-end.
+ */
-+ if (!acpi_ioapic)
-+ setup_ioapic_ids_from_mpc();
-+#ifndef CONFIG_XEN
-+ sync_Arb_IDs();
-+#endif /* !CONFIG_XEN */
-+ setup_IO_APIC_irqs();
-+ init_IO_APIC_traps();
-+ check_timer();
-+ if (!acpi_ioapic)
-+ print_IO_APIC();
-+}
-+
-+struct sysfs_ioapic_data {
-+ struct sys_device dev;
-+ struct IO_APIC_route_entry entry[0];
-+};
-+static struct sysfs_ioapic_data * mp_ioapic_data[MAX_IO_APICS];
-+
-+static int ioapic_suspend(struct sys_device *dev, pm_message_t state)
-+{
-+ struct IO_APIC_route_entry *entry;
-+ struct sysfs_ioapic_data *data;
-+ unsigned long flags;
-+ int i;
-+
-+ data = container_of(dev, struct sysfs_ioapic_data, dev);
-+ entry = data->entry;
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ for (i = 0; i < nr_ioapic_registers[dev->id]; i ++, entry ++ ) {
-+ *(((int *)entry) + 1) = io_apic_read(dev->id, 0x11 + 2 * i);
-+ *(((int *)entry) + 0) = io_apic_read(dev->id, 0x10 + 2 * i);
++ spin_lock_irqsave(&vtpms->resp_list_lock, flags);
++ if (vtpms->current_request != ptr) {
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ return 0;
+ }
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+
-+ return 0;
-+}
-+
-+static int ioapic_resume(struct sys_device *dev)
-+{
-+ struct IO_APIC_route_entry *entry;
-+ struct sysfs_ioapic_data *data;
-+ unsigned long flags;
-+ union IO_APIC_reg_00 reg_00;
-+ int i;
-+
-+ data = container_of(dev, struct sysfs_ioapic_data, dev);
-+ entry = data->entry;
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ reg_00.raw = io_apic_read(dev->id, 0);
-+ if (reg_00.bits.ID != mp_ioapics[dev->id].mpc_apicid) {
-+ reg_00.bits.ID = mp_ioapics[dev->id].mpc_apicid;
-+ io_apic_write(dev->id, 0, reg_00.raw);
-+ }
-+ for (i = 0; i < nr_ioapic_registers[dev->id]; i ++, entry ++ ) {
-+ io_apic_write(dev->id, 0x11+2*i, *(((int *)entry)+1));
-+ io_apic_write(dev->id, 0x10+2*i, *(((int *)entry)+0));
++ if ((t = vtpms->current_request)) {
++ transmission_free(t);
++ vtpms->current_request = NULL;
+ }
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+
-+ return 0;
-+}
-+
-+static struct sysdev_class ioapic_sysdev_class = {
-+ set_kset_name("ioapic"),
-+ .suspend = ioapic_suspend,
-+ .resume = ioapic_resume,
-+};
+
-+static int __init ioapic_init_sysfs(void)
-+{
-+ struct sys_device * dev;
-+ int i, size, error = 0;
-+
-+ error = sysdev_class_register(&ioapic_sysdev_class);
-+ if (error)
-+ return error;
-+
-+ for (i = 0; i < nr_ioapics; i++ ) {
-+ size = sizeof(struct sys_device) + nr_ioapic_registers[i]
-+ * sizeof(struct IO_APIC_route_entry);
-+ mp_ioapic_data[i] = kmalloc(size, GFP_KERNEL);
-+ if (!mp_ioapic_data[i]) {
-+ printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
-+ continue;
-+ }
-+ memset(mp_ioapic_data[i], 0, size);
-+ dev = &mp_ioapic_data[i]->dev;
-+ dev->id = i;
-+ dev->cls = &ioapic_sysdev_class;
-+ error = sysdev_register(dev);
-+ if (error) {
-+ kfree(mp_ioapic_data[i]);
-+ mp_ioapic_data[i] = NULL;
-+ printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
-+ continue;
++ t = transmission_alloc();
++ if (t) {
++ if (!transmission_set_res_buffer(t, buffer, count)) {
++ transmission_free(t);
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ return -ENOMEM;
+ }
++ ret_size = count;
++ vtpms->current_response = t;
++ wake_up_interruptible(&vtpms->resp_wait_queue);
+ }
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
+
-+ return 0;
-+}
-+
-+device_initcall(ioapic_init_sysfs);
-+
-+/* --------------------------------------------------------------------------
-+ ACPI-based IOAPIC Configuration
-+ -------------------------------------------------------------------------- */
-+
-+#ifdef CONFIG_ACPI
-+
-+#define IO_APIC_MAX_ID 0xFE
-+
-+int __init io_apic_get_version (int ioapic)
-+{
-+ union IO_APIC_reg_01 reg_01;
-+ unsigned long flags;
-+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ reg_01.raw = io_apic_read(ioapic, 1);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+
-+ return reg_01.bits.version;
++ return ret_size;
+}
+
+
-+int __init io_apic_get_redir_entries (int ioapic)
++/*
++ * Lower layer indicates its status (connected/disconnected)
++ */
++void vtpm_vd_status(const struct tpm_chip *chip, u8 vd_status)
+{
-+ union IO_APIC_reg_01 reg_01;
-+ unsigned long flags;
++ struct vtpm_state *vtpms;
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ reg_01.raw = io_apic_read(ioapic, 1);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
++ vtpms = (struct vtpm_state *)chip_get_private(chip);
+
-+ return reg_01.bits.entries;
++ vtpms->vd_status = vd_status;
++ if ((vtpms->vd_status & TPM_VD_STATUS_CONNECTED) == 0) {
++ vtpms->disconnect_time = jiffies;
++ }
+}
+
-+
-+int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int active_high_low)
++/* =============================================================
++ * Interface with the generic TPM driver
++ * =============================================================
++ */
++static int vtpm_recv(struct tpm_chip *chip, u8 *buf, size_t count)
+{
-+ struct IO_APIC_route_entry entry;
++ int rc = 0;
+ unsigned long flags;
++ struct vtpm_state *vtpms;
+
-+ if (!IO_APIC_IRQ(irq)) {
-+ apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
-+ ioapic);
-+ return -EINVAL;
-+ }
++ vtpms = (struct vtpm_state *)chip_get_private(chip);
+
+ /*
-+ * Generate a PCI IRQ routing entry and program the IOAPIC accordingly.
-+ * Note that we mask (disable) IRQs now -- these get enabled when the
-+ * corresponding device driver registers for this IRQ.
++ * Check if the previous operation only queued the command
++ * In this case there won't be a response, so I just
++ * return from here and reset that flag. In any other
++ * case I should receive a response from the back-end.
+ */
++ spin_lock_irqsave(&vtpms->resp_list_lock, flags);
++ if ((vtpms->flags & DATAEX_FLAG_QUEUED_ONLY) != 0) {
++ vtpms->flags &= ~DATAEX_FLAG_QUEUED_ONLY;
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ /*
++ * The first few commands (measurements) must be
++ * queued since it might not be possible to talk to the
++ * TPM, yet.
++ * Return a response of up to 30 '0's.
++ */
+
-+ memset(&entry,0,sizeof(entry));
-+
-+ entry.delivery_mode = INT_DELIVERY_MODE;
-+ entry.dest_mode = INT_DEST_MODE;
-+ entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS);
-+ entry.trigger = edge_level;
-+ entry.polarity = active_high_low;
-+ entry.mask = 1; /* Disabled (masked) */
-+
-+ irq = gsi_irq_sharing(irq);
++ count = min_t(size_t, count, 30);
++ memset(buf, 0x0, count);
++ return count;
++ }
+ /*
-+ * IRQs < 16 are already in the irq_2_pin[] map
++ * Check whether something is in the responselist and if
++ * there's nothing in the list wait for something to appear.
+ */
-+ if (irq >= 16)
-+ add_pin_to_irq(irq, ioapic, pin);
-+
-+ entry.vector = assign_irq_vector(irq);
-+
-+ apic_printk(APIC_VERBOSE,KERN_DEBUG "IOAPIC[%d]: Set PCI routing entry (%d-%d -> 0x%x -> "
-+ "IRQ %d Mode:%i Active:%i)\n", ioapic,
-+ mp_ioapics[ioapic].mpc_apicid, pin, entry.vector, irq,
-+ edge_level, active_high_low);
-+
-+ ioapic_register_intr(irq, entry.vector, edge_level);
+
-+ if (!ioapic && (irq < 16))
-+ disable_8259A_irq(irq);
++ if (!vtpms->current_response) {
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ interruptible_sleep_on_timeout(&vtpms->resp_wait_queue,
++ 1000);
++ spin_lock_irqsave(&vtpms->resp_list_lock ,flags);
++ }
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ io_apic_write(ioapic, 0x11+2*pin, *(((int *)&entry)+1));
-+ io_apic_write(ioapic, 0x10+2*pin, *(((int *)&entry)+0));
-+ set_native_irq_info(use_pci_vector() ? entry.vector : irq, TARGET_CPUS);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
++ if (vtpms->current_response) {
++ struct transmission *t = vtpms->current_response;
++ vtpms->current_response = NULL;
++ rc = min(count, t->response_len);
++ memcpy(buf, t->response, rc);
++ transmission_free(t);
++ }
+
-+ return 0;
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ return rc;
+}
+
-+#endif /* CONFIG_ACPI */
-+
-+
-+#ifndef CONFIG_XEN
-+/*
-+ * This function currently is only a helper for the i386 smp boot process where
-+ * we need to reprogram the ioredtbls to cater for the cpus which have come online
-+ * so mask in all cases should simply be TARGET_CPUS
-+ */
-+#ifdef CONFIG_SMP
-+void __init setup_ioapic_dest(void)
++static int vtpm_send(struct tpm_chip *chip, u8 *buf, size_t count)
+{
-+ int pin, ioapic, irq, irq_entry;
-+
-+ if (skip_ioapic_setup == 1)
-+ return;
++ int rc = 0;
++ unsigned long flags;
++ struct transmission *t = transmission_alloc();
++ struct vtpm_state *vtpms;
+
-+ for (ioapic = 0; ioapic < nr_ioapics; ioapic++) {
-+ for (pin = 0; pin < nr_ioapic_registers[ioapic]; pin++) {
-+ irq_entry = find_irq_entry(ioapic, pin, mp_INT);
-+ if (irq_entry == -1)
-+ continue;
-+ irq = pin_2_irq(irq_entry, ioapic, pin);
-+ set_ioapic_affinity_irq(irq, TARGET_CPUS);
-+ }
++ vtpms = (struct vtpm_state *)chip_get_private(chip);
+
++ if (!t)
++ return -ENOMEM;
++ /*
++ * If there's a current request, it must be the
++ * previous request that has timed out.
++ */
++ spin_lock_irqsave(&vtpms->req_list_lock, flags);
++ if (vtpms->current_request != NULL) {
++ printk("WARNING: Sending although there is a request outstanding.\n"
++ " Previous request must have timed out.\n");
++ transmission_free(vtpms->current_request);
++ vtpms->current_request = NULL;
+ }
-+}
-+#endif
-+#endif /* !CONFIG_XEN */
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/ioport-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/ioport-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/ioport-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/ioport-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,99 @@
-+/*
-+ * linux/arch/x86_64/kernel/ioport.c
-+ *
-+ * This contains the io-permission bitmap code - written by obz, with changes
-+ * by Linus.
-+ */
++ spin_unlock_irqrestore(&vtpms->req_list_lock, flags);
+
-+#include <linux/sched.h>
-+#include <linux/kernel.h>
-+#include <linux/capability.h>
-+#include <linux/errno.h>
-+#include <linux/types.h>
-+#include <linux/ioport.h>
-+#include <linux/mm.h>
-+#include <linux/smp.h>
-+#include <linux/smp_lock.h>
-+#include <linux/stddef.h>
-+#include <linux/slab.h>
-+#include <linux/thread_info.h>
-+#include <xen/interface/physdev.h>
++ /*
++ * Queue the packet if the driver below is not
++ * ready, yet, or there is any packet already
++ * in the queue.
++ * If the driver below is ready, unqueue all
++ * packets first before sending our current
++ * packet.
++ * For each unqueued packet, except for the
++ * last (=current) packet, call the function
++ * tpm_xen_recv to wait for the response to come
++ * back.
++ */
++ if ((vtpms->vd_status & TPM_VD_STATUS_CONNECTED) == 0) {
++ if (time_after(jiffies,
++ vtpms->disconnect_time + HZ * 10)) {
++ rc = -ENOENT;
++ } else {
++ goto queue_it;
++ }
++ } else {
++ /*
++ * Send all queued packets.
++ */
++ if (_vtpm_send_queued(chip) == 0) {
+
-+/* Set EXTENT bits starting at BASE in BITMAP to value TURN_ON. */
-+static void set_bitmap(unsigned long *bitmap, unsigned int base, unsigned int extent, int new_value)
-+{
-+ int i;
++ vtpms->current_request = t;
+
-+ if (new_value)
-+ for (i = base; i < base + extent; i++)
-+ __set_bit(i, bitmap);
-+ else
-+ for (i = base; i < base + extent; i++)
-+ clear_bit(i, bitmap);
++ rc = vtpm_vd_send(vtpms->tpm_private,
++ buf,
++ count,
++ t);
++ /*
++ * The generic TPM driver will call
++ * the function to receive the response.
++ */
++ if (rc < 0) {
++ vtpms->current_request = NULL;
++ goto queue_it;
++ }
++ } else {
++queue_it:
++ if (!transmission_set_req_buffer(t, buf, count)) {
++ transmission_free(t);
++ rc = -ENOMEM;
++ goto exit;
++ }
++ /*
++ * An error occurred. Don't event try
++ * to send the current request. Just
++ * queue it.
++ */
++ spin_lock_irqsave(&vtpms->req_list_lock, flags);
++ vtpms->flags |= DATAEX_FLAG_QUEUED_ONLY;
++ list_add_tail(&t->next, &vtpms->queued_requests);
++ spin_unlock_irqrestore(&vtpms->req_list_lock, flags);
++ }
++ }
++
++exit:
++ return rc;
+}
+
++
+/*
-+ * this changes the io permissions bitmap in the current task.
++ * Send all queued requests.
+ */
-+asmlinkage long sys_ioperm(unsigned long from, unsigned long num, int turn_on)
++static int _vtpm_send_queued(struct tpm_chip *chip)
+{
-+ struct thread_struct * t = ¤t->thread;
-+ unsigned long *bitmap;
-+ struct physdev_set_iobitmap set_iobitmap;
++ int rc;
++ int error = 0;
++ long flags;
++ unsigned char buffer[1];
++ struct vtpm_state *vtpms;
++ vtpms = (struct vtpm_state *)chip_get_private(chip);
+
-+ if ((from + num <= from) || (from + num > IO_BITMAP_BITS))
-+ return -EINVAL;
-+ if (turn_on && !capable(CAP_SYS_RAWIO))
-+ return -EPERM;
++ spin_lock_irqsave(&vtpms->req_list_lock, flags);
+
-+ /*
-+ * If it's the first ioperm() call in this thread's lifetime, set the
-+ * IO bitmap up. ioperm() is much less timing critical than clone(),
-+ * this is why we delay this operation until now:
-+ */
-+ if (!t->io_bitmap_ptr) {
-+ bitmap = kmalloc(IO_BITMAP_BYTES, GFP_KERNEL);
-+ if (!bitmap)
-+ return -ENOMEM;
++ while (!list_empty(&vtpms->queued_requests)) {
++ /*
++ * Need to dequeue them.
++ * Read the result into a dummy buffer.
++ */
++ struct transmission *qt = (struct transmission *)
++ vtpms->queued_requests.next;
++ list_del(&qt->next);
++ vtpms->current_request = qt;
++ spin_unlock_irqrestore(&vtpms->req_list_lock, flags);
+
-+ memset(bitmap, 0xff, IO_BITMAP_BYTES);
-+ t->io_bitmap_ptr = bitmap;
++ rc = vtpm_vd_send(vtpms->tpm_private,
++ qt->request,
++ qt->request_len,
++ qt);
+
-+ set_xen_guest_handle(set_iobitmap.bitmap, (char *)bitmap);
-+ set_iobitmap.nr_ports = IO_BITMAP_BITS;
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap, &set_iobitmap);
++ if (rc < 0) {
++ spin_lock_irqsave(&vtpms->req_list_lock, flags);
++ if ((qt = vtpms->current_request) != NULL) {
++ /*
++ * requeue it at the beginning
++ * of the list
++ */
++ list_add(&qt->next,
++ &vtpms->queued_requests);
++ }
++ vtpms->current_request = NULL;
++ error = 1;
++ break;
++ }
++ /*
++ * After this point qt is not valid anymore!
++ * It is freed when the front-end is delivering
++ * the data by calling tpm_recv
++ */
++ /*
++ * Receive response into provided dummy buffer
++ */
++ rc = vtpm_recv(chip, buffer, sizeof(buffer));
++ spin_lock_irqsave(&vtpms->req_list_lock, flags);
+ }
+
-+ set_bitmap(t->io_bitmap_ptr, from, num, !turn_on);
++ spin_unlock_irqrestore(&vtpms->req_list_lock, flags);
+
-+ return 0;
++ return error;
+}
+
-+/*
-+ * sys_iopl has to be used when you want to access the IO ports
-+ * beyond the 0x3ff range: to get the full 65536 ports bitmapped
-+ * you'd need 8kB of bitmaps/process, which is a bit excessive.
-+ *
-+ */
-+
-+asmlinkage long sys_iopl(unsigned int new_iopl, struct pt_regs *regs)
++static void vtpm_cancel(struct tpm_chip *chip)
+{
-+ unsigned int old_iopl = current->thread.iopl;
-+ struct physdev_set_iopl set_iopl;
-+
-+ if (new_iopl > 3)
-+ return -EINVAL;
++ unsigned long flags;
++ struct vtpm_state *vtpms = (struct vtpm_state *)chip_get_private(chip);
+
-+ /* Need "raw I/O" privileges for direct port access. */
-+ if ((new_iopl > old_iopl) && !capable(CAP_SYS_RAWIO))
-+ return -EPERM;
++ spin_lock_irqsave(&vtpms->resp_list_lock,flags);
+
-+ /* Change our version of the privilege levels. */
-+ current->thread.iopl = new_iopl;
++ if (!vtpms->current_response && vtpms->current_request) {
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ interruptible_sleep_on(&vtpms->resp_wait_queue);
++ spin_lock_irqsave(&vtpms->resp_list_lock,flags);
++ }
+
-+ /* Force the change at ring 0. */
-+ set_iopl.iopl = (new_iopl == 0) ? 1 : new_iopl;
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
++ if (vtpms->current_response) {
++ struct transmission *t = vtpms->current_response;
++ vtpms->current_response = NULL;
++ transmission_free(t);
++ }
+
-+ return 0;
++ spin_unlock_irqrestore(&vtpms->resp_list_lock,flags);
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/irq-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/irq-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/irq-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/irq-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,197 @@
-+/*
-+ * linux/arch/x86_64/kernel/irq.c
-+ *
-+ * Copyright (C) 1992, 1998 Linus Torvalds, Ingo Molnar
-+ *
-+ * This file contains the lowest level x86_64-specific interrupt
-+ * entry and irq statistics code. All the remaining irq logic is
-+ * done by the generic kernel/irq/ code and in the
-+ * x86_64-specific irq controller code. (e.g. i8259.c and
-+ * io_apic.c.)
-+ */
+
-+#include <linux/kernel_stat.h>
-+#include <linux/interrupt.h>
-+#include <linux/seq_file.h>
-+#include <linux/module.h>
-+#include <linux/delay.h>
-+#include <asm/uaccess.h>
-+#include <asm/io_apic.h>
-+#include <asm/idle.h>
-+
-+atomic_t irq_err_count;
-+#ifdef CONFIG_X86_IO_APIC
-+#ifdef APIC_MISMATCH_DEBUG
-+atomic_t irq_mis_count;
-+#endif
-+#endif
-+
-+#ifdef CONFIG_DEBUG_STACKOVERFLOW
-+/*
-+ * Probabilistic stack overflow check:
-+ *
-+ * Only check the stack in process context, because everything else
-+ * runs on the big interrupt stacks. Checking reliably is too expensive,
-+ * so we just check from interrupts.
-+ */
-+static inline void stack_overflow_check(struct pt_regs *regs)
++static u8 vtpm_status(struct tpm_chip *chip)
+{
-+ u64 curbase = (u64) current->thread_info;
-+ static unsigned long warned = -60*HZ;
++ u8 rc = 0;
++ unsigned long flags;
++ struct vtpm_state *vtpms;
+
-+ if (regs->rsp >= curbase && regs->rsp <= curbase + THREAD_SIZE &&
-+ regs->rsp < curbase + sizeof(struct thread_info) + 128 &&
-+ time_after(jiffies, warned + 60*HZ)) {
-+ printk("do_IRQ: %s near stack overflow (cur:%Lx,rsp:%lx)\n",
-+ current->comm, curbase, regs->rsp);
-+ show_stack(NULL,NULL);
-+ warned = jiffies;
++ vtpms = (struct vtpm_state *)chip_get_private(chip);
++
++ spin_lock_irqsave(&vtpms->resp_list_lock, flags);
++ /*
++ * Data are available if:
++ * - there's a current response
++ * - the last packet was queued only (this is fake, but necessary to
++ * get the generic TPM layer to call the receive function.)
++ */
++ if (vtpms->current_response ||
++ 0 != (vtpms->flags & DATAEX_FLAG_QUEUED_ONLY)) {
++ rc = STATUS_DATA_AVAIL;
++ } else if (!vtpms->current_response && !vtpms->current_request) {
++ rc = STATUS_READY;
+ }
++
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ return rc;
+}
-+#endif
+
-+/*
-+ * Generic, controller-independent functions:
-+ */
++static struct file_operations vtpm_ops = {
++ .owner = THIS_MODULE,
++ .llseek = no_llseek,
++ .open = tpm_open,
++ .read = tpm_read,
++ .write = tpm_write,
++ .release = tpm_release,
++};
+
-+int show_interrupts(struct seq_file *p, void *v)
-+{
-+ int i = *(loff_t *) v, j;
-+ struct irqaction * action;
-+ unsigned long flags;
++static DEVICE_ATTR(pubek, S_IRUGO, tpm_show_pubek, NULL);
++static DEVICE_ATTR(pcrs, S_IRUGO, tpm_show_pcrs, NULL);
++static DEVICE_ATTR(enabled, S_IRUGO, tpm_show_enabled, NULL);
++static DEVICE_ATTR(active, S_IRUGO, tpm_show_active, NULL);
++static DEVICE_ATTR(owned, S_IRUGO, tpm_show_owned, NULL);
++static DEVICE_ATTR(temp_deactivated, S_IRUGO, tpm_show_temp_deactivated,
++ NULL);
++static DEVICE_ATTR(caps, S_IRUGO, tpm_show_caps, NULL);
++static DEVICE_ATTR(cancel, S_IWUSR |S_IWGRP, NULL, tpm_store_cancel);
+
-+ if (i == 0) {
-+ seq_printf(p, " ");
-+ for_each_online_cpu(j)
-+ seq_printf(p, "CPU%-8d",j);
-+ seq_putc(p, '\n');
-+ }
++static struct attribute *vtpm_attrs[] = {
++ &dev_attr_pubek.attr,
++ &dev_attr_pcrs.attr,
++ &dev_attr_enabled.attr,
++ &dev_attr_active.attr,
++ &dev_attr_owned.attr,
++ &dev_attr_temp_deactivated.attr,
++ &dev_attr_caps.attr,
++ &dev_attr_cancel.attr,
++ NULL,
++};
+
-+ if (i < NR_IRQS) {
-+ spin_lock_irqsave(&irq_desc[i].lock, flags);
-+ action = irq_desc[i].action;
-+ if (!action)
-+ goto skip;
-+ seq_printf(p, "%3d: ",i);
-+#ifndef CONFIG_SMP
-+ seq_printf(p, "%10u ", kstat_irqs(i));
-+#else
-+ for_each_online_cpu(j)
-+ seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
-+#endif
-+ seq_printf(p, " %14s", irq_desc[i].chip->typename);
++static struct attribute_group vtpm_attr_grp = { .attrs = vtpm_attrs };
+
-+ seq_printf(p, " %s", action->name);
-+ for (action=action->next; action; action = action->next)
-+ seq_printf(p, ", %s", action->name);
-+ seq_putc(p, '\n');
-+skip:
-+ spin_unlock_irqrestore(&irq_desc[i].lock, flags);
-+ } else if (i == NR_IRQS) {
-+ seq_printf(p, "NMI: ");
-+ for_each_online_cpu(j)
-+ seq_printf(p, "%10u ", cpu_pda(j)->__nmi_count);
-+ seq_putc(p, '\n');
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ seq_printf(p, "LOC: ");
-+ for_each_online_cpu(j)
-+ seq_printf(p, "%10u ", cpu_pda(j)->apic_timer_irqs);
-+ seq_putc(p, '\n');
-+#endif
-+ seq_printf(p, "ERR: %10u\n", atomic_read(&irq_err_count));
-+#ifdef CONFIG_X86_IO_APIC
-+#ifdef APIC_MISMATCH_DEBUG
-+ seq_printf(p, "MIS: %10u\n", atomic_read(&irq_mis_count));
-+#endif
-+#endif
-+ }
-+ return 0;
-+}
++#define TPM_LONG_TIMEOUT (10 * 60 * HZ)
++
++static struct tpm_vendor_specific tpm_vtpm = {
++ .recv = vtpm_recv,
++ .send = vtpm_send,
++ .cancel = vtpm_cancel,
++ .status = vtpm_status,
++ .req_complete_mask = STATUS_BUSY | STATUS_DATA_AVAIL,
++ .req_complete_val = STATUS_DATA_AVAIL,
++ .req_canceled = STATUS_READY,
++ .attr_group = &vtpm_attr_grp,
++ .miscdev = {
++ .fops = &vtpm_ops,
++ },
++ .duration = {
++ TPM_LONG_TIMEOUT,
++ TPM_LONG_TIMEOUT,
++ TPM_LONG_TIMEOUT,
++ },
++};
++
++struct tpm_chip *init_vtpm(struct device *dev,
++ struct tpm_private *tp)
++{
++ long rc;
++ struct tpm_chip *chip;
++ struct vtpm_state *vtpms;
+
-+/*
-+ * do_IRQ handles all normal device IRQ's (the special
-+ * SMP cross-CPU interrupts have their own specific
-+ * handlers).
-+ */
-+asmlinkage unsigned int do_IRQ(struct pt_regs *regs)
-+{
-+ /* high bit used in ret_from_ code */
-+ unsigned irq = ~regs->orig_rax;
++ vtpms = kzalloc(sizeof(struct vtpm_state), GFP_KERNEL);
++ if (!vtpms)
++ return ERR_PTR(-ENOMEM);
+
-+ if (unlikely(irq >= NR_IRQS)) {
-+ printk(KERN_EMERG "%s: cannot handle IRQ %d\n",
-+ __FUNCTION__, irq);
-+ BUG();
++ vtpm_state_init(vtpms);
++ vtpms->tpm_private = tp;
++
++ chip = tpm_register_hardware(dev, &tpm_vtpm);
++ if (!chip) {
++ rc = -ENODEV;
++ goto err_free_mem;
+ }
+
-+ exit_idle();
-+ irq_enter();
-+#ifdef CONFIG_DEBUG_STACKOVERFLOW
-+ stack_overflow_check(regs);
-+#endif
-+ __do_IRQ(irq, regs);
-+ irq_exit();
++ chip_set_private(chip, vtpms);
+
-+ return 1;
++ return chip;
++
++err_free_mem:
++ kfree(vtpms);
++
++ return ERR_PTR(rc);
+}
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+void fixup_irqs(cpumask_t map)
++void cleanup_vtpm(struct device *dev)
+{
-+ unsigned int irq;
-+ static int warned;
++ struct tpm_chip *chip = dev_get_drvdata(dev);
++ struct vtpm_state *vtpms = (struct vtpm_state*)chip_get_private(chip);
++ tpm_remove_hardware(dev);
++ kfree(vtpms);
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/char/tpm/tpm_vtpm.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/char/tpm/tpm_vtpm.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,55 @@
++#ifndef TPM_VTPM_H
++#define TPM_VTPM_H
+
-+ for (irq = 0; irq < NR_IRQS; irq++) {
-+ cpumask_t mask;
-+ if (irq == 2)
-+ continue;
++struct tpm_chip;
++struct tpm_private;
+
-+ cpus_and(mask, irq_desc[irq].affinity, map);
-+ if (any_online_cpu(mask) == NR_CPUS) {
-+ printk("Breaking affinity for irq %i\n", irq);
-+ mask = map;
-+ }
-+ if (irq_desc[irq].chip->set_affinity)
-+ irq_desc[irq].chip->set_affinity(irq, mask);
-+ else if (irq_desc[irq].action && !(warned++))
-+ printk("Cannot set affinity for irq %i\n", irq);
-+ }
++struct vtpm_state {
++ struct transmission *current_request;
++ spinlock_t req_list_lock;
++ wait_queue_head_t req_wait_queue;
+
-+ /* That doesn't seem sufficient. Give it 1ms. */
-+ local_irq_enable();
-+ mdelay(1);
-+ local_irq_disable();
-+}
-+#endif
++ struct list_head queued_requests;
+
-+extern void call_softirq(void);
++ struct transmission *current_response;
++ spinlock_t resp_list_lock;
++ wait_queue_head_t resp_wait_queue; // processes waiting for responses
+
-+asmlinkage void do_softirq(void)
-+{
-+ __u32 pending;
-+ unsigned long flags;
++ u8 vd_status;
++ u8 flags;
+
-+ if (in_interrupt())
-+ return;
++ unsigned long disconnect_time;
+
-+ local_irq_save(flags);
-+ pending = local_softirq_pending();
-+ /* Switch to interrupt stack */
-+ if (pending) {
-+ call_softirq();
-+ WARN_ON_ONCE(softirq_count());
-+ }
-+ local_irq_restore(flags);
-+}
-+EXPORT_SYMBOL(do_softirq);
++ /*
++ * The following is a private structure of the underlying
++ * driver. It is passed as parameter in the send function.
++ */
++ struct tpm_private *tpm_private;
++};
+
-+#ifndef CONFIG_X86_LOCAL_APIC
-+/*
-+ * 'what should we do if we get a hw irq event on an illegal vector'.
-+ * each architecture has to answer this themselves.
-+ */
-+void ack_bad_irq(unsigned int irq)
++
++enum vdev_status {
++ TPM_VD_STATUS_DISCONNECTED = 0x0,
++ TPM_VD_STATUS_CONNECTED = 0x1
++};
++
++/* this function is called from tpm_vtpm.c */
++int vtpm_vd_send(struct tpm_private * tp,
++ const u8 * buf, size_t count, void *ptr);
++
++/* these functions are offered by tpm_vtpm.c */
++struct tpm_chip *init_vtpm(struct device *,
++ struct tpm_private *);
++void cleanup_vtpm(struct device *);
++int vtpm_vd_recv(const struct tpm_chip* chip,
++ const unsigned char *buffer, size_t count, void *ptr);
++void vtpm_vd_status(const struct tpm_chip *, u8 status);
++
++static inline struct tpm_private *tpm_private_from_dev(struct device *dev)
+{
-+ printk("unexpected IRQ trap at vector %02x\n", irq);
++ struct tpm_chip *chip = dev_get_drvdata(dev);
++ struct vtpm_state *vtpms = chip_get_private(chip);
++ return vtpms->tpm_private;
+}
++
+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/ldt-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/ldt-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/ldt-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/ldt-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,282 @@
+diff -r d894e36cfc30 -r 0aa021803deb drivers/char/tpm/tpm_xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/char/tpm/tpm_xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,722 @@
+/*
-+ * linux/arch/x86_64/kernel/ldt.c
++ * Copyright (c) 2005, IBM Corporation
+ *
-+ * Copyright (C) 1992 Krishna Balasubramanian and Linus Torvalds
-+ * Copyright (C) 1999 Ingo Molnar <mingo at redhat.com>
-+ * Copyright (C) 2002 Andi Kleen
-+ *
-+ * This handles calls from both 32bit and 64bit mode.
++ * Author: Stefan Berger, stefanb at us.ibm.com
++ * Grant table support: Mahadevan Gomathisankaran
++ *
++ * This code has been derived from drivers/xen/netfront/netfront.c
++ *
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
+
+#include <linux/errno.h>
-+#include <linux/sched.h>
-+#include <linux/string.h>
-+#include <linux/mm.h>
-+#include <linux/smp.h>
-+#include <linux/smp_lock.h>
-+#include <linux/vmalloc.h>
-+#include <linux/slab.h>
-+
++#include <linux/err.h>
++#include <linux/interrupt.h>
++#include <linux/mutex.h>
+#include <asm/uaccess.h>
-+#include <asm/system.h>
-+#include <asm/ldt.h>
-+#include <asm/desc.h>
-+#include <asm/proto.h>
-+#include <asm/pgalloc.h>
++#include <xen/evtchn.h>
++#include <xen/interface/grant_table.h>
++#include <xen/interface/io/tpmif.h>
++#include <xen/gnttab.h>
++#include <xen/xenbus.h>
++#include "tpm.h"
++#include "tpm_vtpm.h"
+
-+#ifdef CONFIG_SMP /* avoids "defined but not used" warnig */
-+static void flush_ldt(void *null)
-+{
-+ if (current->active_mm)
-+ load_LDT(¤t->active_mm->context);
-+}
-+#endif
++#undef DEBUG
+
-+static int alloc_ldt(mm_context_t *pc, unsigned mincount, int reload)
-+{
-+ void *oldldt;
-+ void *newldt;
-+ unsigned oldsize;
++/* local structures */
++struct tpm_private {
++ struct tpm_chip *chip;
+
-+ if (mincount <= (unsigned)pc->size)
-+ return 0;
-+ oldsize = pc->size;
-+ mincount = (mincount+511)&(~511);
-+ if (mincount*LDT_ENTRY_SIZE > PAGE_SIZE)
-+ newldt = vmalloc(mincount*LDT_ENTRY_SIZE);
-+ else
-+ newldt = kmalloc(mincount*LDT_ENTRY_SIZE, GFP_KERNEL);
++ tpmif_tx_interface_t *tx;
++ atomic_t refcnt;
++ unsigned int irq;
++ u8 is_connected;
++ u8 is_suspended;
+
-+ if (!newldt)
-+ return -ENOMEM;
++ spinlock_t tx_lock;
+
-+ if (oldsize)
-+ memcpy(newldt, pc->ldt, oldsize*LDT_ENTRY_SIZE);
-+ oldldt = pc->ldt;
-+ memset(newldt+oldsize*LDT_ENTRY_SIZE, 0, (mincount-oldsize)*LDT_ENTRY_SIZE);
-+ wmb();
-+ pc->ldt = newldt;
-+ wmb();
-+ pc->size = mincount;
-+ wmb();
-+ if (reload) {
-+#ifdef CONFIG_SMP
-+ cpumask_t mask;
++ struct tx_buffer *tx_buffers[TPMIF_TX_RING_SIZE];
+
-+ preempt_disable();
-+#endif
-+ make_pages_readonly(
-+ pc->ldt,
-+ (pc->size * LDT_ENTRY_SIZE) / PAGE_SIZE,
-+ XENFEAT_writable_descriptor_tables);
-+ load_LDT(pc);
-+#ifdef CONFIG_SMP
-+ mask = cpumask_of_cpu(smp_processor_id());
-+ if (!cpus_equal(current->mm->cpu_vm_mask, mask))
-+ smp_call_function(flush_ldt, NULL, 1, 1);
-+ preempt_enable();
-+#endif
-+ }
-+ if (oldsize) {
-+ make_pages_writable(
-+ oldldt,
-+ (oldsize * LDT_ENTRY_SIZE) / PAGE_SIZE,
-+ XENFEAT_writable_descriptor_tables);
-+ if (oldsize*LDT_ENTRY_SIZE > PAGE_SIZE)
-+ vfree(oldldt);
-+ else
-+ kfree(oldldt);
-+ }
-+ return 0;
-+}
++ atomic_t tx_busy;
++ void *tx_remember;
+
-+static inline int copy_ldt(mm_context_t *new, mm_context_t *old)
-+{
-+ int err = alloc_ldt(new, old->size, 0);
-+ if (err < 0)
-+ return err;
-+ memcpy(new->ldt, old->ldt, old->size*LDT_ENTRY_SIZE);
-+ make_pages_readonly(
-+ new->ldt,
-+ (new->size * LDT_ENTRY_SIZE) / PAGE_SIZE,
-+ XENFEAT_writable_descriptor_tables);
-+ return 0;
-+}
++ domid_t backend_id;
++ wait_queue_head_t wait_q;
+
-+/*
-+ * we do not have to muck with descriptors here, that is
-+ * done in switch_mm() as needed.
-+ */
-+int init_new_context(struct task_struct *tsk, struct mm_struct *mm)
-+{
-+ struct mm_struct * old_mm;
-+ int retval = 0;
++ struct xenbus_device *dev;
++ int ring_ref;
++};
+
-+ memset(&mm->context, 0, sizeof(mm->context));
-+ init_MUTEX(&mm->context.sem);
-+ old_mm = current->mm;
-+ if (old_mm && old_mm->context.size > 0) {
-+ down(&old_mm->context.sem);
-+ retval = copy_ldt(&mm->context, &old_mm->context);
-+ up(&old_mm->context.sem);
-+ }
-+ if (retval == 0) {
-+ spin_lock(&mm_unpinned_lock);
-+ list_add(&mm->context.unpinned, &mm_unpinned);
-+ spin_unlock(&mm_unpinned_lock);
-+ }
-+ return retval;
-+}
++struct tx_buffer {
++ unsigned int size; // available space in data
++ unsigned int len; // used space in data
++ unsigned char *data; // pointer to a page
++};
+
-+/*
-+ *
-+ * Don't touch the LDT register - we're already in the next thread.
-+ */
-+void destroy_context(struct mm_struct *mm)
++
++/* locally visible variables */
++static grant_ref_t gref_head;
++static struct tpm_private *my_priv;
++
++/* local function prototypes */
++static irqreturn_t tpmif_int(int irq,
++ void *tpm_priv,
++ struct pt_regs *ptregs);
++static void tpmif_rx_action(unsigned long unused);
++static int tpmif_connect(struct xenbus_device *dev,
++ struct tpm_private *tp,
++ domid_t domid);
++static DECLARE_TASKLET(tpmif_rx_tasklet, tpmif_rx_action, 0);
++static int tpmif_allocate_tx_buffers(struct tpm_private *tp);
++static void tpmif_free_tx_buffers(struct tpm_private *tp);
++static void tpmif_set_connected_state(struct tpm_private *tp,
++ u8 newstate);
++static int tpm_xmit(struct tpm_private *tp,
++ const u8 * buf, size_t count, int userbuffer,
++ void *remember);
++static void destroy_tpmring(struct tpm_private *tp);
++void __exit tpmif_exit(void);
++
++#define DPRINTK(fmt, args...) \
++ pr_debug("xen_tpm_fr (%s:%d) " fmt, __FUNCTION__, __LINE__, ##args)
++#define IPRINTK(fmt, args...) \
++ printk(KERN_INFO "xen_tpm_fr: " fmt, ##args)
++#define WPRINTK(fmt, args...) \
++ printk(KERN_WARNING "xen_tpm_fr: " fmt, ##args)
++
++#define GRANT_INVALID_REF 0
++
++
++static inline int
++tx_buffer_copy(struct tx_buffer *txb, const u8 *src, int len,
++ int isuserbuffer)
+{
-+ if (mm->context.size) {
-+ if (mm == current->active_mm)
-+ clear_LDT();
-+ make_pages_writable(
-+ mm->context.ldt,
-+ (mm->context.size * LDT_ENTRY_SIZE) / PAGE_SIZE,
-+ XENFEAT_writable_descriptor_tables);
-+ if (mm->context.size*LDT_ENTRY_SIZE > PAGE_SIZE)
-+ vfree(mm->context.ldt);
-+ else
-+ kfree(mm->context.ldt);
-+ mm->context.size = 0;
-+ }
-+ if (!mm->context.pinned) {
-+ spin_lock(&mm_unpinned_lock);
-+ list_del(&mm->context.unpinned);
-+ spin_unlock(&mm_unpinned_lock);
++ int copied = len;
++
++ if (len > txb->size)
++ copied = txb->size;
++ if (isuserbuffer) {
++ if (copy_from_user(txb->data, src, copied))
++ return -EFAULT;
++ } else {
++ memcpy(txb->data, src, copied);
+ }
++ txb->len = len;
++ return copied;
+}
+
-+static int read_ldt(void __user * ptr, unsigned long bytecount)
++static inline struct tx_buffer *tx_buffer_alloc(void)
+{
-+ int err;
-+ unsigned long size;
-+ struct mm_struct * mm = current->mm;
++ struct tx_buffer *txb;
+
-+ if (!mm->context.size)
-+ return 0;
-+ if (bytecount > LDT_ENTRY_SIZE*LDT_ENTRIES)
-+ bytecount = LDT_ENTRY_SIZE*LDT_ENTRIES;
++ txb = kzalloc(sizeof(struct tx_buffer), GFP_KERNEL);
++ if (!txb)
++ return NULL;
+
-+ down(&mm->context.sem);
-+ size = mm->context.size*LDT_ENTRY_SIZE;
-+ if (size > bytecount)
-+ size = bytecount;
++ txb->len = 0;
++ txb->size = PAGE_SIZE;
++ txb->data = (unsigned char *)__get_free_page(GFP_KERNEL);
++ if (txb->data == NULL) {
++ kfree(txb);
++ txb = NULL;
++ }
+
-+ err = 0;
-+ if (copy_to_user(ptr, mm->context.ldt, size))
-+ err = -EFAULT;
-+ up(&mm->context.sem);
-+ if (err < 0)
-+ goto error_return;
-+ if (size != bytecount) {
-+ /* zero-fill the rest */
-+ if (clear_user(ptr+size, bytecount-size) != 0) {
-+ err = -EFAULT;
-+ goto error_return;
-+ }
++ return txb;
++}
++
++
++static inline void tx_buffer_free(struct tx_buffer *txb)
++{
++ if (txb) {
++ free_page((long)txb->data);
++ kfree(txb);
+ }
-+ return bytecount;
-+error_return:
-+ return err;
+}
+
-+static int read_default_ldt(void __user * ptr, unsigned long bytecount)
++/**************************************************************
++ Utility function for the tpm_private structure
++**************************************************************/
++static void tpm_private_init(struct tpm_private *tp)
+{
-+ /* Arbitrary number */
-+ /* x86-64 default LDT is all zeros */
-+ if (bytecount > 128)
-+ bytecount = 128;
-+ if (clear_user(ptr, bytecount))
-+ return -EFAULT;
-+ return bytecount;
++ spin_lock_init(&tp->tx_lock);
++ init_waitqueue_head(&tp->wait_q);
++ atomic_set(&tp->refcnt, 1);
+}
+
-+static int write_ldt(void __user * ptr, unsigned long bytecount, int oldmode)
++static void tpm_private_put(void)
+{
-+ struct task_struct *me = current;
-+ struct mm_struct * mm = me->mm;
-+ __u32 entry_1, entry_2, *lp;
-+ unsigned long mach_lp;
-+ int error;
-+ struct user_desc ldt_info;
++ if (!atomic_dec_and_test(&my_priv->refcnt))
++ return;
+
-+ error = -EINVAL;
++ tpmif_free_tx_buffers(my_priv);
++ kfree(my_priv);
++ my_priv = NULL;
++}
+
-+ if (bytecount != sizeof(ldt_info))
-+ goto out;
-+ error = -EFAULT;
-+ if (copy_from_user(&ldt_info, ptr, bytecount))
-+ goto out;
++static struct tpm_private *tpm_private_get(void)
++{
++ int err;
+
-+ error = -EINVAL;
-+ if (ldt_info.entry_number >= LDT_ENTRIES)
-+ goto out;
-+ if (ldt_info.contents == 3) {
-+ if (oldmode)
-+ goto out;
-+ if (ldt_info.seg_not_present == 0)
-+ goto out;
++ if (my_priv) {
++ atomic_inc(&my_priv->refcnt);
++ return my_priv;
+ }
+
-+ down(&mm->context.sem);
-+ if (ldt_info.entry_number >= (unsigned)mm->context.size) {
-+ error = alloc_ldt(¤t->mm->context, ldt_info.entry_number+1, 1);
-+ if (error < 0)
-+ goto out_unlock;
-+ }
++ my_priv = kzalloc(sizeof(struct tpm_private), GFP_KERNEL);
++ if (!my_priv)
++ return NULL;
+
-+ lp = (__u32 *) ((ldt_info.entry_number << 3) + (char *) mm->context.ldt);
-+ mach_lp = arbitrary_virt_to_machine(lp);
++ tpm_private_init(my_priv);
++ err = tpmif_allocate_tx_buffers(my_priv);
++ if (err < 0)
++ tpm_private_put();
+
-+ /* Allow LDTs to be cleared by the user. */
-+ if (ldt_info.base_addr == 0 && ldt_info.limit == 0) {
-+ if (oldmode || LDT_empty(&ldt_info)) {
-+ entry_1 = 0;
-+ entry_2 = 0;
-+ goto install;
-+ }
-+ }
++ return my_priv;
++}
+
-+ entry_1 = LDT_entry_a(&ldt_info);
-+ entry_2 = LDT_entry_b(&ldt_info);
-+ if (oldmode)
-+ entry_2 &= ~(1 << 20);
++/**************************************************************
+
-+ /* Install the new entry ... */
-+install:
-+ error = HYPERVISOR_update_descriptor(mach_lp, (unsigned long)((entry_1 | (unsigned long) entry_2 << 32)));
++ The interface to let the tpm plugin register its callback
++ function and send data to another partition using this module
+
-+out_unlock:
-+ up(&mm->context.sem);
-+out:
-+ return error;
-+}
++**************************************************************/
+
-+asmlinkage int sys_modify_ldt(int func, void __user *ptr, unsigned long bytecount)
++static DEFINE_MUTEX(suspend_lock);
++/*
++ * Send data via this module by calling this function
++ */
++int vtpm_vd_send(struct tpm_private *tp,
++ const u8 * buf, size_t count, void *ptr)
+{
-+ int ret = -ENOSYS;
++ int sent;
+
-+ switch (func) {
-+ case 0:
-+ ret = read_ldt(ptr, bytecount);
-+ break;
-+ case 1:
-+ ret = write_ldt(ptr, bytecount, 1);
-+ break;
-+ case 2:
-+ ret = read_default_ldt(ptr, bytecount);
-+ break;
-+ case 0x11:
-+ ret = write_ldt(ptr, bytecount, 0);
-+ break;
-+ }
-+ return ret;
++ mutex_lock(&suspend_lock);
++ sent = tpm_xmit(tp, buf, count, 0, ptr);
++ mutex_unlock(&suspend_lock);
++
++ return sent;
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/machine_kexec.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/machine_kexec.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/machine_kexec.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/machine_kexec.c 2007-07-30 16:35:11.000000000 +0200
-@@ -15,6 +15,113 @@
- #include <asm/mmu_context.h>
- #include <asm/io.h>
-
-+#define PAGE_ALIGNED __attribute__ ((__aligned__(PAGE_SIZE)))
-+static u64 kexec_pgd[512] PAGE_ALIGNED;
-+static u64 kexec_pud0[512] PAGE_ALIGNED;
-+static u64 kexec_pmd0[512] PAGE_ALIGNED;
-+static u64 kexec_pte0[512] PAGE_ALIGNED;
-+static u64 kexec_pud1[512] PAGE_ALIGNED;
-+static u64 kexec_pmd1[512] PAGE_ALIGNED;
-+static u64 kexec_pte1[512] PAGE_ALIGNED;
+
-+#ifdef CONFIG_XEN
++/**************************************************************
++ XENBUS support code
++**************************************************************/
+
-+/* In the case of Xen, override hypervisor functions to be able to create
-+ * a regular identity mapping page table...
-+ */
++static int setup_tpmring(struct xenbus_device *dev,
++ struct tpm_private *tp)
++{
++ tpmif_tx_interface_t *sring;
++ int err;
+
-+#include <xen/interface/kexec.h>
-+#include <xen/interface/memory.h>
++ tp->ring_ref = GRANT_INVALID_REF;
+
-+#define x__pmd(x) ((pmd_t) { (x) } )
-+#define x__pud(x) ((pud_t) { (x) } )
-+#define x__pgd(x) ((pgd_t) { (x) } )
++ sring = (void *)__get_free_page(GFP_KERNEL);
++ if (!sring) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating shared ring");
++ return -ENOMEM;
++ }
++ tp->tx = sring;
+
-+#define x_pmd_val(x) ((x).pmd)
-+#define x_pud_val(x) ((x).pud)
-+#define x_pgd_val(x) ((x).pgd)
++ err = xenbus_grant_ring(dev, virt_to_mfn(tp->tx));
++ if (err < 0) {
++ free_page((unsigned long)sring);
++ tp->tx = NULL;
++ xenbus_dev_fatal(dev, err, "allocating grant reference");
++ goto fail;
++ }
++ tp->ring_ref = err;
+
-+static inline void x_set_pmd(pmd_t *dst, pmd_t val)
-+{
-+ x_pmd_val(*dst) = x_pmd_val(val);
-+}
++ err = tpmif_connect(dev, tp, dev->otherend_id);
++ if (err)
++ goto fail;
+
-+static inline void x_set_pud(pud_t *dst, pud_t val)
-+{
-+ x_pud_val(*dst) = phys_to_machine(x_pud_val(val));
++ return 0;
++fail:
++ destroy_tpmring(tp);
++ return err;
+}
+
-+static inline void x_pud_clear (pud_t *pud)
-+{
-+ x_pud_val(*pud) = 0;
-+}
+
-+static inline void x_set_pgd(pgd_t *dst, pgd_t val)
++static void destroy_tpmring(struct tpm_private *tp)
+{
-+ x_pgd_val(*dst) = phys_to_machine(x_pgd_val(val));
-+}
++ tpmif_set_connected_state(tp, 0);
+
-+static inline void x_pgd_clear (pgd_t * pgd)
-+{
-+ x_pgd_val(*pgd) = 0;
++ if (tp->ring_ref != GRANT_INVALID_REF) {
++ gnttab_end_foreign_access(tp->ring_ref, (unsigned long)tp->tx);
++ tp->ring_ref = GRANT_INVALID_REF;
++ tp->tx = NULL;
++ }
++
++ if (tp->irq)
++ unbind_from_irqhandler(tp->irq, tp);
++
++ tp->irq = 0;
+}
+
-+#define X__PAGE_KERNEL_LARGE_EXEC \
-+ _PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_PSE
-+#define X_KERNPG_TABLE _PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY
+
-+#define __ma(x) (pfn_to_mfn(__pa((x)) >> PAGE_SHIFT) << PAGE_SHIFT)
++static int talk_to_backend(struct xenbus_device *dev,
++ struct tpm_private *tp)
++{
++ const char *message = NULL;
++ int err;
++ struct xenbus_transaction xbt;
+
-+#if PAGES_NR > KEXEC_XEN_NO_PAGES
-+#error PAGES_NR is greater than KEXEC_XEN_NO_PAGES - Xen support will break
-+#endif
++ err = setup_tpmring(dev, tp);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "setting up ring");
++ goto out;
++ }
+
-+#if PA_CONTROL_PAGE != 0
-+#error PA_CONTROL_PAGE is non zero - Xen support will break
-+#endif
++again:
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "starting transaction");
++ goto destroy_tpmring;
++ }
+
-+void machine_kexec_setup_load_arg(xen_kexec_image_t *xki, struct kimage *image)
-+{
-+ void *control_page;
-+ void *table_page;
++ err = xenbus_printf(xbt, dev->nodename,
++ "ring-ref","%u", tp->ring_ref);
++ if (err) {
++ message = "writing ring-ref";
++ goto abort_transaction;
++ }
+
-+ memset(xki->page_list, 0, sizeof(xki->page_list));
++ err = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
++ irq_to_evtchn_port(tp->irq));
++ if (err) {
++ message = "writing event-channel";
++ goto abort_transaction;
++ }
+
-+ control_page = page_address(image->control_code_page) + PAGE_SIZE;
-+ memcpy(control_page, relocate_kernel, PAGE_SIZE);
++ err = xenbus_transaction_end(xbt, 0);
++ if (err == -EAGAIN)
++ goto again;
++ if (err) {
++ xenbus_dev_fatal(dev, err, "completing transaction");
++ goto destroy_tpmring;
++ }
+
-+ table_page = page_address(image->control_code_page);
++ xenbus_switch_state(dev, XenbusStateConnected);
+
-+ xki->page_list[PA_CONTROL_PAGE] = __ma(control_page);
-+ xki->page_list[PA_TABLE_PAGE] = __ma(table_page);
++ return 0;
+
-+ xki->page_list[PA_PGD] = __ma(kexec_pgd);
-+ xki->page_list[PA_PUD_0] = __ma(kexec_pud0);
-+ xki->page_list[PA_PUD_1] = __ma(kexec_pud1);
-+ xki->page_list[PA_PMD_0] = __ma(kexec_pmd0);
-+ xki->page_list[PA_PMD_1] = __ma(kexec_pmd1);
-+ xki->page_list[PA_PTE_0] = __ma(kexec_pte0);
-+ xki->page_list[PA_PTE_1] = __ma(kexec_pte1);
++abort_transaction:
++ xenbus_transaction_end(xbt, 1);
++ if (message)
++ xenbus_dev_error(dev, err, "%s", message);
++destroy_tpmring:
++ destroy_tpmring(tp);
++out:
++ return err;
+}
+
-+#else /* CONFIG_XEN */
++/**
++ * Callback received when the backend's state changes.
++ */
++static void backend_changed(struct xenbus_device *dev,
++ enum xenbus_state backend_state)
++{
++ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
++ DPRINTK("\n");
+
-+#define x__pmd(x) __pmd(x)
-+#define x__pud(x) __pud(x)
-+#define x__pgd(x) __pgd(x)
++ switch (backend_state) {
++ case XenbusStateInitialising:
++ case XenbusStateInitWait:
++ case XenbusStateInitialised:
++ case XenbusStateReconfiguring:
++ case XenbusStateReconfigured:
++ case XenbusStateUnknown:
++ break;
+
-+#define x_set_pmd(x, y) set_pmd(x, y)
-+#define x_set_pud(x, y) set_pud(x, y)
-+#define x_set_pgd(x, y) set_pgd(x, y)
++ case XenbusStateConnected:
++ tpmif_set_connected_state(tp, 1);
++ break;
+
-+#define x_pud_clear(x) pud_clear(x)
-+#define x_pgd_clear(x) pgd_clear(x)
++ case XenbusStateClosing:
++ tpmif_set_connected_state(tp, 0);
++ xenbus_frontend_closed(dev);
++ break;
+
-+#define X__PAGE_KERNEL_LARGE_EXEC __PAGE_KERNEL_LARGE_EXEC
-+#define X_KERNPG_TABLE _KERNPG_TABLE
++ case XenbusStateClosed:
++ tpmif_set_connected_state(tp, 0);
++ if (tp->is_suspended == 0)
++ device_unregister(&dev->dev);
++ xenbus_frontend_closed(dev);
++ break;
++ }
++}
+
-+#endif /* CONFIG_XEN */
++static int tpmfront_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ int err;
++ int handle;
++ struct tpm_private *tp = tpm_private_get();
+
- static void init_level2_page(pmd_t *level2p, unsigned long addr)
- {
- unsigned long end_addr;
-@@ -22,7 +129,7 @@ static void init_level2_page(pmd_t *leve
- addr &= PAGE_MASK;
- end_addr = addr + PUD_SIZE;
- while (addr < end_addr) {
-- set_pmd(level2p++, __pmd(addr | __PAGE_KERNEL_LARGE_EXEC));
-+ x_set_pmd(level2p++, x__pmd(addr | X__PAGE_KERNEL_LARGE_EXEC));
- addr += PMD_SIZE;
- }
- }
-@@ -47,12 +154,12 @@ static int init_level3_page(struct kimag
- }
- level2p = (pmd_t *)page_address(page);
- init_level2_page(level2p, addr);
-- set_pud(level3p++, __pud(__pa(level2p) | _KERNPG_TABLE));
-+ x_set_pud(level3p++, x__pud(__pa(level2p) | X_KERNPG_TABLE));
- addr += PUD_SIZE;
- }
- /* clear the unused entries */
- while (addr < end_addr) {
-- pud_clear(level3p++);
-+ x_pud_clear(level3p++);
- addr += PUD_SIZE;
- }
- out:
-@@ -83,12 +190,12 @@ static int init_level4_page(struct kimag
- if (result) {
- goto out;
- }
-- set_pgd(level4p++, __pgd(__pa(level3p) | _KERNPG_TABLE));
-+ x_set_pgd(level4p++, x__pgd(__pa(level3p) | X_KERNPG_TABLE));
- addr += PGDIR_SIZE;
- }
- /* clear the unused entries */
- while (addr < end_addr) {
-- pgd_clear(level4p++);
-+ x_pgd_clear(level4p++);
- addr += PGDIR_SIZE;
- }
- out:
-@@ -99,77 +206,29 @@ out:
- static int init_pgtable(struct kimage *image, unsigned long start_pgtable)
- {
- pgd_t *level4p;
-- level4p = (pgd_t *)__va(start_pgtable);
-- return init_level4_page(image, level4p, 0, end_pfn << PAGE_SHIFT);
--}
--
--static void set_idt(void *newidt, u16 limit)
--{
-- struct desc_ptr curidt;
-+ unsigned long x_end_pfn = end_pfn;
-
-- /* x86-64 supports unaliged loads & stores */
-- curidt.size = limit;
-- curidt.address = (unsigned long)newidt;
-+#ifdef CONFIG_XEN
-+ x_end_pfn = HYPERVISOR_memory_op(XENMEM_maximum_ram_page, NULL);
-+#endif
-
-- __asm__ __volatile__ (
-- "lidtq %0\n"
-- : : "m" (curidt)
-- );
--};
--
--
--static void set_gdt(void *newgdt, u16 limit)
--{
-- struct desc_ptr curgdt;
--
-- /* x86-64 supports unaligned loads & stores */
-- curgdt.size = limit;
-- curgdt.address = (unsigned long)newgdt;
--
-- __asm__ __volatile__ (
-- "lgdtq %0\n"
-- : : "m" (curgdt)
-- );
--};
--
--static void load_segments(void)
--{
-- __asm__ __volatile__ (
-- "\tmovl %0,%%ds\n"
-- "\tmovl %0,%%es\n"
-- "\tmovl %0,%%ss\n"
-- "\tmovl %0,%%fs\n"
-- "\tmovl %0,%%gs\n"
-- : : "a" (__KERNEL_DS) : "memory"
-- );
-+ level4p = (pgd_t *)__va(start_pgtable);
-+ return init_level4_page(image, level4p, 0, x_end_pfn << PAGE_SHIFT);
- }
-
--typedef NORET_TYPE void (*relocate_new_kernel_t)(unsigned long indirection_page,
-- unsigned long control_code_buffer,
-- unsigned long start_address,
-- unsigned long pgtable) ATTRIB_NORET;
--
--extern const unsigned char relocate_new_kernel[];
--extern const unsigned long relocate_new_kernel_size;
--
- int machine_kexec_prepare(struct kimage *image)
- {
-- unsigned long start_pgtable, control_code_buffer;
-+ unsigned long start_pgtable;
- int result;
-
- /* Calculate the offsets */
- start_pgtable = page_to_pfn(image->control_code_page) << PAGE_SHIFT;
-- control_code_buffer = start_pgtable + PAGE_SIZE;
-
- /* Setup the identity mapped 64bit page table */
- result = init_pgtable(image, start_pgtable);
- if (result)
- return result;
-
-- /* Place the code in the reboot code buffer */
-- memcpy(__va(control_code_buffer), relocate_new_kernel,
-- relocate_new_kernel_size);
--
- return 0;
- }
-
-@@ -178,51 +237,43 @@ void machine_kexec_cleanup(struct kimage
- return;
- }
-
-+#ifndef CONFIG_XEN
- /*
- * Do not allocate memory (or fail in any way) in machine_kexec().
- * We are past the point of no return, committed to rebooting now.
- */
- NORET_TYPE void machine_kexec(struct kimage *image)
- {
-- unsigned long page_list;
-- unsigned long control_code_buffer;
-- unsigned long start_pgtable;
-- relocate_new_kernel_t rnk;
-+ unsigned long page_list[PAGES_NR];
-+ void *control_page;
-
- /* Interrupts aren't acceptable while we reboot */
- local_irq_disable();
-
-- /* Calculate the offsets */
-- page_list = image->head;
-- start_pgtable = page_to_pfn(image->control_code_page) << PAGE_SHIFT;
-- control_code_buffer = start_pgtable + PAGE_SIZE;
-+ control_page = page_address(image->control_code_page) + PAGE_SIZE;
-+ memcpy(control_page, relocate_kernel, PAGE_SIZE);
++ if (!tp)
++ return -ENOMEM;
+
-+ page_list[PA_CONTROL_PAGE] = __pa(control_page);
-+ page_list[VA_CONTROL_PAGE] = (unsigned long)relocate_kernel;
-+ page_list[PA_PGD] = __pa(kexec_pgd);
-+ page_list[VA_PGD] = (unsigned long)kexec_pgd;
-+ page_list[PA_PUD_0] = __pa(kexec_pud0);
-+ page_list[VA_PUD_0] = (unsigned long)kexec_pud0;
-+ page_list[PA_PMD_0] = __pa(kexec_pmd0);
-+ page_list[VA_PMD_0] = (unsigned long)kexec_pmd0;
-+ page_list[PA_PTE_0] = __pa(kexec_pte0);
-+ page_list[VA_PTE_0] = (unsigned long)kexec_pte0;
-+ page_list[PA_PUD_1] = __pa(kexec_pud1);
-+ page_list[VA_PUD_1] = (unsigned long)kexec_pud1;
-+ page_list[PA_PMD_1] = __pa(kexec_pmd1);
-+ page_list[VA_PMD_1] = (unsigned long)kexec_pmd1;
-+ page_list[PA_PTE_1] = __pa(kexec_pte1);
-+ page_list[VA_PTE_1] = (unsigned long)kexec_pte1;
++ tp->chip = init_vtpm(&dev->dev, tp);
++ if (IS_ERR(tp->chip))
++ return PTR_ERR(tp->chip);
+
-+ page_list[PA_TABLE_PAGE] =
-+ (unsigned long)__pa(page_address(image->control_code_page));
-
-- /* Set the low half of the page table to my identity mapped
-- * page table for kexec. Leave the high half pointing at the
-- * kernel pages. Don't bother to flush the global pages
-- * as that will happen when I fully switch to my identity mapped
-- * page table anyway.
-- */
-- memcpy(__va(read_cr3()), __va(start_pgtable), PAGE_SIZE/2);
-- __flush_tlb();
--
--
-- /* The segment registers are funny things, they have both a
-- * visible and an invisible part. Whenever the visible part is
-- * set to a specific selector, the invisible part is loaded
-- * with from a table in memory. At no other time is the
-- * descriptor table in memory accessed.
-- *
-- * I take advantage of this here by force loading the
-- * segments, before I zap the gdt with an invalid value.
-- */
-- load_segments();
-- /* The gdt & idt are now invalid.
-- * If you want to load them you must set up your own idt & gdt.
-- */
-- set_gdt(phys_to_virt(0),0);
-- set_idt(phys_to_virt(0),0);
-- /* now call it */
-- rnk = (relocate_new_kernel_t) control_code_buffer;
-- (*rnk)(page_list, control_code_buffer, image->start, start_pgtable);
-+ relocate_kernel((unsigned long)image->head, (unsigned long)page_list,
-+ image->start);
- }
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/Makefile tmp-linux-2.6-xen.patch/arch/x86_64/kernel/Makefile
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -21,11 +21,13 @@ obj-$(CONFIG_MICROCODE) += microcode.o
- obj-$(CONFIG_X86_CPUID) += cpuid.o
- obj-$(CONFIG_SMP) += smp.o smpboot.o trampoline.o
- obj-$(CONFIG_X86_LOCAL_APIC) += apic.o nmi.o
-+obj-$(CONFIG_X86_XEN_GENAPIC) += genapic.o genapic_xen.o
- obj-$(CONFIG_X86_IO_APIC) += io_apic.o mpparse.o \
- genapic.o genapic_cluster.o genapic_flat.o
- obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o crash.o
- obj-$(CONFIG_CRASH_DUMP) += crash_dump.o
--obj-$(CONFIG_PM) += suspend.o
-+obj-$(CONFIG_SOFTWARE_SUSPEND) += suspend.o
-+obj-$(CONFIG_ACPI_SLEEP) += suspend.o
- obj-$(CONFIG_SOFTWARE_SUSPEND) += suspend_asm.o
- obj-$(CONFIG_CPU_FREQ) += cpufreq/
- obj-$(CONFIG_EARLY_PRINTK) += early_printk.o
-@@ -55,3 +57,18 @@ i8237-y += ../../i386/kernel/i8237.o
- msr-$(subst m,y,$(CONFIG_X86_MSR)) += ../../i386/kernel/msr.o
- alternative-y += ../../i386/kernel/alternative.o
-
-+ifdef CONFIG_XEN
-+time-y += ../../i386/kernel/time-xen.o
-+pci-dma-y += ../../i386/kernel/pci-dma-xen.o
-+microcode-$(subst m,y,$(CONFIG_MICROCODE)) := ../../i386/kernel/microcode-xen.o
-+quirks-y := ../../i386/kernel/quirks-xen.o
-+
-+n-obj-xen := i8259.o reboot.o i8237.o smpboot.o trampoline.o
-+
-+include $(srctree)/scripts/Makefile.xen
-+
-+obj-y := $(call filterxen, $(obj-y), $(n-obj-xen))
-+obj-y := $(call cherrypickxen, $(obj-y))
-+extra-y := $(call cherrypickxen, $(extra-y))
-+%/head-xen.o %/head-xen.s: EXTRA_AFLAGS :=
-+endif
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/mpparse-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/mpparse-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/mpparse-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/mpparse-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,1011 @@
-+/*
-+ * Intel Multiprocessor Specification 1.1 and 1.4
-+ * compliant MP-table parsing routines.
-+ *
-+ * (c) 1995 Alan Cox, Building #3 <alan at redhat.com>
-+ * (c) 1998, 1999, 2000 Ingo Molnar <mingo at redhat.com>
-+ *
-+ * Fixes
-+ * Erich Boleyn : MP v1.4 and additional changes.
-+ * Alan Cox : Added EBDA scanning
-+ * Ingo Molnar : various cleanups and rewrites
-+ * Maciej W. Rozycki: Bits for default MP configurations
-+ * Paul Diefenbaugh: Added full ACPI support
-+ */
++ err = xenbus_scanf(XBT_NIL, dev->nodename,
++ "handle", "%i", &handle);
++ if (XENBUS_EXIST_ERR(err))
++ return err;
+
-+#include <linux/mm.h>
-+#include <linux/init.h>
-+#include <linux/delay.h>
-+#include <linux/bootmem.h>
-+#include <linux/smp_lock.h>
-+#include <linux/kernel_stat.h>
-+#include <linux/mc146818rtc.h>
-+#include <linux/acpi.h>
-+#include <linux/module.h>
++ if (err < 0) {
++ xenbus_dev_fatal(dev,err,"reading virtual-device");
++ return err;
++ }
+
-+#include <asm/smp.h>
-+#include <asm/mtrr.h>
-+#include <asm/mpspec.h>
-+#include <asm/pgalloc.h>
-+#include <asm/io_apic.h>
-+#include <asm/proto.h>
-+#include <asm/acpi.h>
++ tp->dev = dev;
++
++ err = talk_to_backend(dev, tp);
++ if (err) {
++ tpm_private_put();
++ return err;
++ }
+
-+/* Have we found an MP table */
-+int smp_found_config;
-+unsigned int __initdata maxcpus = NR_CPUS;
++ return 0;
++}
+
-+int acpi_found_madt;
+
-+/*
-+ * Various Linux-internal data structures created from the
-+ * MP-table.
-+ */
-+unsigned char apic_version [MAX_APICS];
-+unsigned char mp_bus_id_to_type [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 };
-+int mp_bus_id_to_pci_bus [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 };
++static int tpmfront_remove(struct xenbus_device *dev)
++{
++ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
++ destroy_tpmring(tp);
++ cleanup_vtpm(&dev->dev);
++ return 0;
++}
+
-+static int mp_current_pci_id = 0;
-+/* I/O APIC entries */
-+struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS];
++static int tpmfront_suspend(struct xenbus_device *dev)
++{
++ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
++ u32 ctr;
+
-+/* # of MP IRQ source entries */
-+struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES];
++ /* Take the lock, preventing any application from sending. */
++ mutex_lock(&suspend_lock);
++ tp->is_suspended = 1;
+
-+/* MP IRQ source entries */
-+int mp_irq_entries;
++ for (ctr = 0; atomic_read(&tp->tx_busy); ctr++) {
++ if ((ctr % 10) == 0)
++ printk("TPM-FE [INFO]: Waiting for outstanding "
++ "request.\n");
++ /* Wait for a request to be responded to. */
++ interruptible_sleep_on_timeout(&tp->wait_q, 100);
++ }
+
-+int nr_ioapics;
-+int pic_mode;
-+unsigned long mp_lapic_addr = 0;
++ return 0;
++}
+
++static int tpmfront_suspend_finish(struct tpm_private *tp)
++{
++ tp->is_suspended = 0;
++ /* Allow applications to send again. */
++ mutex_unlock(&suspend_lock);
++ return 0;
++}
+
++static int tpmfront_suspend_cancel(struct xenbus_device *dev)
++{
++ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
++ return tpmfront_suspend_finish(tp);
++}
+
-+/* Processor that is doing the boot up */
-+unsigned int boot_cpu_id = -1U;
-+/* Internal processor count */
-+unsigned int num_processors __initdata = 0;
++static int tpmfront_resume(struct xenbus_device *dev)
++{
++ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
++ destroy_tpmring(tp);
++ return talk_to_backend(dev, tp);
++}
+
-+unsigned disabled_cpus __initdata;
++static int tpmif_connect(struct xenbus_device *dev,
++ struct tpm_private *tp,
++ domid_t domid)
++{
++ int err;
+
-+/* Bitmask of physically existing CPUs */
-+physid_mask_t phys_cpu_present_map = PHYSID_MASK_NONE;
++ tp->backend_id = domid;
+
-+/* ACPI MADT entry parsing functions */
-+#ifdef CONFIG_ACPI
-+extern struct acpi_boot_flags acpi_boot;
-+#ifdef CONFIG_X86_LOCAL_APIC
-+extern int acpi_parse_lapic (acpi_table_entry_header *header);
-+extern int acpi_parse_lapic_addr_ovr (acpi_table_entry_header *header);
-+extern int acpi_parse_lapic_nmi (acpi_table_entry_header *header);
-+#endif /*CONFIG_X86_LOCAL_APIC*/
-+#ifdef CONFIG_X86_IO_APIC
-+extern int acpi_parse_ioapic (acpi_table_entry_header *header);
-+#endif /*CONFIG_X86_IO_APIC*/
-+#endif /*CONFIG_ACPI*/
++ err = bind_listening_port_to_irqhandler(
++ domid, tpmif_int, SA_SAMPLE_RANDOM, "tpmif", tp);
++ if (err <= 0) {
++ WPRINTK("bind_listening_port_to_irqhandler failed "
++ "(err=%d)\n", err);
++ return err;
++ }
++ tp->irq = err;
+
-+u8 bios_cpu_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = BAD_APICID };
++ return 0;
++}
+
++static struct xenbus_device_id tpmfront_ids[] = {
++ { "vtpm" },
++ { "" }
++};
+
-+/*
-+ * Intel MP BIOS table parsing routines:
-+ */
++static struct xenbus_driver tpmfront = {
++ .name = "vtpm",
++ .owner = THIS_MODULE,
++ .ids = tpmfront_ids,
++ .probe = tpmfront_probe,
++ .remove = tpmfront_remove,
++ .resume = tpmfront_resume,
++ .otherend_changed = backend_changed,
++ .suspend = tpmfront_suspend,
++ .suspend_cancel = tpmfront_suspend_cancel,
++};
+
-+/*
-+ * Checksum an MP configuration block.
-+ */
++static void __init init_tpm_xenbus(void)
++{
++ xenbus_register_frontend(&tpmfront);
++}
+
-+static int __init mpf_checksum(unsigned char *mp, int len)
++static int tpmif_allocate_tx_buffers(struct tpm_private *tp)
+{
-+ int sum = 0;
++ unsigned int i;
+
-+ while (len--)
-+ sum += *mp++;
++ for (i = 0; i < TPMIF_TX_RING_SIZE; i++) {
++ tp->tx_buffers[i] = tx_buffer_alloc();
++ if (!tp->tx_buffers[i]) {
++ tpmif_free_tx_buffers(tp);
++ return -ENOMEM;
++ }
++ }
++ return 0;
++}
+
-+ return sum & 0xFF;
++static void tpmif_free_tx_buffers(struct tpm_private *tp)
++{
++ unsigned int i;
++
++ for (i = 0; i < TPMIF_TX_RING_SIZE; i++)
++ tx_buffer_free(tp->tx_buffers[i]);
+}
+
-+#ifndef CONFIG_XEN
-+static void __cpuinit MP_processor_info (struct mpc_config_processor *m)
++static void tpmif_rx_action(unsigned long priv)
+{
-+ int cpu;
-+ unsigned char ver;
-+ cpumask_t tmp_map;
++ struct tpm_private *tp = (struct tpm_private *)priv;
++ int i = 0;
++ unsigned int received;
++ unsigned int offset = 0;
++ u8 *buffer;
++ tpmif_tx_request_t *tx = &tp->tx->ring[i].req;
+
-+ if (!(m->mpc_cpuflag & CPU_ENABLED)) {
-+ disabled_cpus++;
-+ return;
-+ }
++ atomic_set(&tp->tx_busy, 0);
++ wake_up_interruptible(&tp->wait_q);
+
-+ printk(KERN_INFO "Processor #%d %d:%d APIC version %d\n",
-+ m->mpc_apicid,
-+ (m->mpc_cpufeature & CPU_FAMILY_MASK)>>8,
-+ (m->mpc_cpufeature & CPU_MODEL_MASK)>>4,
-+ m->mpc_apicver);
++ received = tx->size;
+
-+ if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) {
-+ Dprintk(" Bootup CPU\n");
-+ boot_cpu_id = m->mpc_apicid;
-+ }
-+ if (num_processors >= NR_CPUS) {
-+ printk(KERN_WARNING "WARNING: NR_CPUS limit of %i reached."
-+ " Processor ignored.\n", NR_CPUS);
++ buffer = kmalloc(received, GFP_ATOMIC);
++ if (!buffer)
+ return;
-+ }
+
-+ num_processors++;
-+ cpus_complement(tmp_map, cpu_present_map);
-+ cpu = first_cpu(tmp_map);
++ for (i = 0; i < TPMIF_TX_RING_SIZE && offset < received; i++) {
++ struct tx_buffer *txb = tp->tx_buffers[i];
++ tpmif_tx_request_t *tx;
++ unsigned int tocopy;
+
-+#if MAX_APICS < 255
-+ if ((int)m->mpc_apicid > MAX_APICS) {
-+ printk(KERN_ERR "Processor #%d INVALID. (Max ID: %d).\n",
-+ m->mpc_apicid, MAX_APICS);
-+ return;
-+ }
-+#endif
-+ ver = m->mpc_apicver;
++ tx = &tp->tx->ring[i].req;
++ tocopy = tx->size;
++ if (tocopy > PAGE_SIZE)
++ tocopy = PAGE_SIZE;
+
-+ physid_set(m->mpc_apicid, phys_cpu_present_map);
-+ /*
-+ * Validate version
-+ */
-+ if (ver == 0x0) {
-+ printk(KERN_ERR "BIOS bug, APIC version is 0 for CPU#%d! fixing up to 0x10. (tell your hw vendor)\n", m->mpc_apicid);
-+ ver = 0x10;
++ memcpy(&buffer[offset], txb->data, tocopy);
++
++ gnttab_release_grant_reference(&gref_head, tx->ref);
++
++ offset += tocopy;
+ }
-+ apic_version[m->mpc_apicid] = ver;
-+ if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) {
-+ /*
-+ * bios_cpu_apicid is required to have processors listed
-+ * in same order as logical cpu numbers. Hence the first
-+ * entry is BSP, and so on.
-+ */
-+ cpu = 0;
-+ }
-+ bios_cpu_apicid[cpu] = m->mpc_apicid;
-+ x86_cpu_to_apicid[cpu] = m->mpc_apicid;
+
-+ cpu_set(cpu, cpu_possible_map);
-+ cpu_set(cpu, cpu_present_map);
-+}
-+#else
-+static void __cpuinit MP_processor_info (struct mpc_config_processor *m)
-+{
-+ num_processors++;
++ vtpm_vd_recv(tp->chip, buffer, received, tp->tx_remember);
++ kfree(buffer);
+}
-+#endif /* CONFIG_XEN */
+
-+static void __init MP_bus_info (struct mpc_config_bus *m)
++
++static irqreturn_t tpmif_int(int irq, void *tpm_priv, struct pt_regs *ptregs)
+{
-+ char str[7];
++ struct tpm_private *tp = tpm_priv;
++ unsigned long flags;
+
-+ memcpy(str, m->mpc_bustype, 6);
-+ str[6] = 0;
-+ Dprintk("Bus #%d is %s\n", m->mpc_busid, str);
++ spin_lock_irqsave(&tp->tx_lock, flags);
++ tpmif_rx_tasklet.data = (unsigned long)tp;
++ tasklet_schedule(&tpmif_rx_tasklet);
++ spin_unlock_irqrestore(&tp->tx_lock, flags);
+
-+ if (strncmp(str, "ISA", 3) == 0) {
-+ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_ISA;
-+ } else if (strncmp(str, "EISA", 4) == 0) {
-+ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_EISA;
-+ } else if (strncmp(str, "PCI", 3) == 0) {
-+ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_PCI;
-+ mp_bus_id_to_pci_bus[m->mpc_busid] = mp_current_pci_id;
-+ mp_current_pci_id++;
-+ } else if (strncmp(str, "MCA", 3) == 0) {
-+ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_MCA;
-+ } else {
-+ printk(KERN_ERR "Unknown bustype %s\n", str);
-+ }
++ return IRQ_HANDLED;
+}
+
-+static void __init MP_ioapic_info (struct mpc_config_ioapic *m)
++
++static int tpm_xmit(struct tpm_private *tp,
++ const u8 * buf, size_t count, int isuserbuffer,
++ void *remember)
+{
-+ if (!(m->mpc_flags & MPC_APIC_USABLE))
-+ return;
++ tpmif_tx_request_t *tx;
++ TPMIF_RING_IDX i;
++ unsigned int offset = 0;
+
-+ printk("I/O APIC #%d Version %d at 0x%X.\n",
-+ m->mpc_apicid, m->mpc_apicver, m->mpc_apicaddr);
-+ if (nr_ioapics >= MAX_IO_APICS) {
-+ printk(KERN_ERR "Max # of I/O APICs (%d) exceeded (found %d).\n",
-+ MAX_IO_APICS, nr_ioapics);
-+ panic("Recompile kernel with bigger MAX_IO_APICS!.\n");
++ spin_lock_irq(&tp->tx_lock);
++
++ if (unlikely(atomic_read(&tp->tx_busy))) {
++ printk("tpm_xmit: There's an outstanding request/response "
++ "on the way!\n");
++ spin_unlock_irq(&tp->tx_lock);
++ return -EBUSY;
+ }
-+ if (!m->mpc_apicaddr) {
-+ printk(KERN_ERR "WARNING: bogus zero I/O APIC address"
-+ " found in MP table, skipping!\n");
-+ return;
++
++ if (tp->is_connected != 1) {
++ spin_unlock_irq(&tp->tx_lock);
++ return -EIO;
+ }
-+ mp_ioapics[nr_ioapics] = *m;
-+ nr_ioapics++;
++
++ for (i = 0; count > 0 && i < TPMIF_TX_RING_SIZE; i++) {
++ struct tx_buffer *txb = tp->tx_buffers[i];
++ int copied;
++
++ if (!txb) {
++ DPRINTK("txb (i=%d) is NULL. buffers initilized?\n"
++ "Not transmitting anything!\n", i);
++ spin_unlock_irq(&tp->tx_lock);
++ return -EFAULT;
++ }
++
++ copied = tx_buffer_copy(txb, &buf[offset], count,
++ isuserbuffer);
++ if (copied < 0) {
++ /* An error occurred */
++ spin_unlock_irq(&tp->tx_lock);
++ return copied;
++ }
++ count -= copied;
++ offset += copied;
++
++ tx = &tp->tx->ring[i].req;
++ tx->addr = virt_to_machine(txb->data);
++ tx->size = txb->len;
++ tx->unused = 0;
++
++ DPRINTK("First 4 characters sent by TPM-FE are "
++ "0x%02x 0x%02x 0x%02x 0x%02x\n",
++ txb->data[0],txb->data[1],txb->data[2],txb->data[3]);
++
++ /* Get the granttable reference for this page. */
++ tx->ref = gnttab_claim_grant_reference(&gref_head);
++ if (tx->ref == -ENOSPC) {
++ spin_unlock_irq(&tp->tx_lock);
++ DPRINTK("Grant table claim reference failed in "
++ "func:%s line:%d file:%s\n",
++ __FUNCTION__, __LINE__, __FILE__);
++ return -ENOSPC;
++ }
++ gnttab_grant_foreign_access_ref(tx->ref,
++ tp->backend_id,
++ virt_to_mfn(txb->data),
++ 0 /*RW*/);
++ wmb();
++ }
++
++ atomic_set(&tp->tx_busy, 1);
++ tp->tx_remember = remember;
++
++ mb();
++
++ notify_remote_via_irq(tp->irq);
++
++ spin_unlock_irq(&tp->tx_lock);
++ return offset;
+}
+
-+static void __init MP_intsrc_info (struct mpc_config_intsrc *m)
++
++static void tpmif_notify_upperlayer(struct tpm_private *tp)
+{
-+ mp_irqs [mp_irq_entries] = *m;
-+ Dprintk("Int: type %d, pol %d, trig %d, bus %d,"
-+ " IRQ %02x, APIC ID %x, APIC INT %02x\n",
-+ m->mpc_irqtype, m->mpc_irqflag & 3,
-+ (m->mpc_irqflag >> 2) & 3, m->mpc_srcbus,
-+ m->mpc_srcbusirq, m->mpc_dstapic, m->mpc_dstirq);
-+ if (++mp_irq_entries >= MAX_IRQ_SOURCES)
-+ panic("Max # of irq sources exceeded!!\n");
++ /* Notify upper layer about the state of the connection to the BE. */
++ vtpm_vd_status(tp->chip, (tp->is_connected
++ ? TPM_VD_STATUS_CONNECTED
++ : TPM_VD_STATUS_DISCONNECTED));
+}
+
-+static void __init MP_lintsrc_info (struct mpc_config_lintsrc *m)
++
++static void tpmif_set_connected_state(struct tpm_private *tp, u8 is_connected)
+{
-+ Dprintk("Lint: type %d, pol %d, trig %d, bus %d,"
-+ " IRQ %02x, APIC ID %x, APIC LINT %02x\n",
-+ m->mpc_irqtype, m->mpc_irqflag & 3,
-+ (m->mpc_irqflag >> 2) &3, m->mpc_srcbusid,
-+ m->mpc_srcbusirq, m->mpc_destapic, m->mpc_destapiclint);
+ /*
-+ * Well it seems all SMP boards in existence
-+ * use ExtINT/LVT1 == LINT0 and
-+ * NMI/LVT2 == LINT1 - the following check
-+ * will show us if this assumptions is false.
-+ * Until then we do not have to add baggage.
++ * Don't notify upper layer if we are in suspend mode and
++ * should disconnect - assumption is that we will resume
++ * The mutex keeps apps from sending.
+ */
-+ if ((m->mpc_irqtype == mp_ExtINT) &&
-+ (m->mpc_destapiclint != 0))
-+ BUG();
-+ if ((m->mpc_irqtype == mp_NMI) &&
-+ (m->mpc_destapiclint != 1))
-+ BUG();
++ if (is_connected == 0 && tp->is_suspended == 1)
++ return;
++
++ /*
++ * Unlock the mutex if we are connected again
++ * after being suspended - now resuming.
++ * This also removes the suspend state.
++ */
++ if (is_connected == 1 && tp->is_suspended == 1)
++ tpmfront_suspend_finish(tp);
++
++ if (is_connected != tp->is_connected) {
++ tp->is_connected = is_connected;
++ tpmif_notify_upperlayer(tp);
++ }
+}
+
-+/*
-+ * Read/parse the MPC
++
++
++/* =================================================================
++ * Initialization function.
++ * =================================================================
+ */
+
-+static int __init smp_read_mpc(struct mp_config_table *mpc)
++
++static int __init tpmif_init(void)
+{
-+ char str[16];
-+ int count=sizeof(*mpc);
-+ unsigned char *mpt=((unsigned char *)mpc)+count;
++ struct tpm_private *tp;
+
-+ if (memcmp(mpc->mpc_signature,MPC_SIGNATURE,4)) {
-+ printk("SMP mptable: bad signature [%c%c%c%c]!\n",
-+ mpc->mpc_signature[0],
-+ mpc->mpc_signature[1],
-+ mpc->mpc_signature[2],
-+ mpc->mpc_signature[3]);
-+ return 0;
-+ }
-+ if (mpf_checksum((unsigned char *)mpc,mpc->mpc_length)) {
-+ printk("SMP mptable: checksum error!\n");
-+ return 0;
-+ }
-+ if (mpc->mpc_spec!=0x01 && mpc->mpc_spec!=0x04) {
-+ printk(KERN_ERR "SMP mptable: bad table version (%d)!!\n",
-+ mpc->mpc_spec);
-+ return 0;
-+ }
-+ if (!mpc->mpc_lapic) {
-+ printk(KERN_ERR "SMP mptable: null local APIC address!\n");
-+ return 0;
++ if (is_initial_xendomain())
++ return -EPERM;
++
++ tp = tpm_private_get();
++ if (!tp)
++ return -ENOMEM;
++
++ IPRINTK("Initialising the vTPM driver.\n");
++ if (gnttab_alloc_grant_references(TPMIF_TX_RING_SIZE,
++ &gref_head) < 0) {
++ tpm_private_put();
++ return -EFAULT;
+ }
-+ memcpy(str,mpc->mpc_oem,8);
-+ str[8]=0;
-+ printk(KERN_INFO "OEM ID: %s ",str);
+
-+ memcpy(str,mpc->mpc_productid,12);
-+ str[12]=0;
-+ printk("Product ID: %s ",str);
++ init_tpm_xenbus();
++ return 0;
++}
+
-+ printk("APIC at: 0x%X\n",mpc->mpc_lapic);
+
-+ /* save the local APIC address, it might be non-default */
-+ if (!acpi_lapic)
-+ mp_lapic_addr = mpc->mpc_lapic;
++module_init(tpmif_init);
+
-+ /*
-+ * Now process the configuration blocks.
-+ */
-+ while (count < mpc->mpc_length) {
-+ switch(*mpt) {
-+ case MP_PROCESSOR:
-+ {
-+ struct mpc_config_processor *m=
-+ (struct mpc_config_processor *)mpt;
-+ if (!acpi_lapic)
-+ MP_processor_info(m);
-+ mpt += sizeof(*m);
-+ count += sizeof(*m);
-+ break;
-+ }
-+ case MP_BUS:
-+ {
-+ struct mpc_config_bus *m=
-+ (struct mpc_config_bus *)mpt;
-+ MP_bus_info(m);
-+ mpt += sizeof(*m);
-+ count += sizeof(*m);
-+ break;
-+ }
-+ case MP_IOAPIC:
-+ {
-+ struct mpc_config_ioapic *m=
-+ (struct mpc_config_ioapic *)mpt;
-+ MP_ioapic_info(m);
-+ mpt+=sizeof(*m);
-+ count+=sizeof(*m);
-+ break;
-+ }
-+ case MP_INTSRC:
-+ {
-+ struct mpc_config_intsrc *m=
-+ (struct mpc_config_intsrc *)mpt;
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/char/tty_io.c
+--- a/drivers/char/tty_io.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/char/tty_io.c Wed Sep 10 10:54:08 2008 +0100
+@@ -129,6 +129,8 @@
+ /* Semaphore to protect creating and releasing a tty. This is shared with
+ vt.c for deeply disgusting hack reasons */
+ DEFINE_MUTEX(tty_mutex);
++
++int console_use_vt = 1;
+
+ #ifdef CONFIG_UNIX98_PTYS
+ extern struct tty_driver *ptm_driver; /* Unix98 pty masters; for /dev/ptmx */
+@@ -2483,7 +2485,7 @@
+ goto got_driver;
+ }
+ #ifdef CONFIG_VT
+- if (device == MKDEV(TTY_MAJOR,0)) {
++ if (console_use_vt && (device == MKDEV(TTY_MAJOR,0))) {
+ extern struct tty_driver *console_driver;
+ driver = console_driver;
+ index = fg_console;
+@@ -3909,6 +3911,8 @@
+ #endif
+
+ #ifdef CONFIG_VT
++ if (!console_use_vt)
++ goto out_vt;
+ cdev_init(&vc0_cdev, &console_fops);
+ if (cdev_add(&vc0_cdev, MKDEV(TTY_MAJOR, 0), 1) ||
+ register_chrdev_region(MKDEV(TTY_MAJOR, 0), 1, "/dev/vc/0") < 0)
+@@ -3916,6 +3920,7 @@
+ class_device_create(tty_class, NULL, MKDEV(TTY_MAJOR, 0), NULL, "tty0");
+
+ vty_init();
++ out_vt:
+ #endif
+ return 0;
+ }
+diff -r d894e36cfc30 -r 0aa021803deb drivers/cpufreq/Kconfig
+--- a/drivers/cpufreq/Kconfig Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/cpufreq/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -1,5 +1,6 @@
+ config CPU_FREQ
+ bool "CPU Frequency scaling"
++ depends on !PROCESSOR_EXTERNAL_CONTROL
+ help
+ CPU Frequency scaling allows you to change the clock speed of
+ CPUs on the fly. This is a nice method to save power, because
+diff -r d894e36cfc30 -r 0aa021803deb drivers/cpufreq/cpufreq_ondemand.c
+--- a/drivers/cpufreq/cpufreq_ondemand.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/cpufreq/cpufreq_ondemand.c Wed Sep 10 10:54:08 2008 +0100
+@@ -95,6 +95,7 @@
+
+ return retval;
+ }
+
-+ MP_intsrc_info(m);
-+ mpt+=sizeof(*m);
-+ count+=sizeof(*m);
-+ break;
-+ }
-+ case MP_LINTSRC:
-+ {
-+ struct mpc_config_lintsrc *m=
-+ (struct mpc_config_lintsrc *)mpt;
-+ MP_lintsrc_info(m);
-+ mpt+=sizeof(*m);
-+ count+=sizeof(*m);
-+ break;
-+ }
-+ }
-+ }
-+ clustered_apic_check();
-+ if (!num_processors)
-+ printk(KERN_ERR "SMP mptable: no processors registered!\n");
-+ return num_processors;
+
+ /************************** sysfs interface ************************/
+ static ssize_t show_sampling_rate_max(struct cpufreq_policy *policy, char *buf)
+@@ -222,17 +223,14 @@
+
+ /************************** sysfs end ************************/
+
+-static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info)
++#ifndef CONFIG_XEN
++static int dbs_calc_load(struct cpu_dbs_info_s *this_dbs_info)
+ {
+- unsigned int idle_ticks, total_ticks;
+- unsigned int load;
++ struct cpufreq_policy *policy;
+ cputime64_t cur_jiffies;
+-
+- struct cpufreq_policy *policy;
++ cputime64_t total_ticks, idle_ticks;
+ unsigned int j;
+-
+- if (!this_dbs_info->enable)
+- return;
++ int load;
+
+ policy = this_dbs_info->cur_policy;
+ cur_jiffies = jiffies64_to_cputime64(get_jiffies_64());
+@@ -240,7 +238,7 @@
+ this_dbs_info->prev_cpu_wall);
+ this_dbs_info->prev_cpu_wall = cur_jiffies;
+ if (!total_ticks)
+- return;
++ return 200;
+ /*
+ * Every sampling_rate, we check, if current idle time is less
+ * than 20% (default), then we try to increase frequency
+@@ -270,6 +268,81 @@
+ idle_ticks = tmp_idle_ticks;
+ }
+ load = (100 * (total_ticks - idle_ticks)) / total_ticks;
++ return load;
+}
++#else
+
-+static int __init ELCR_trigger(unsigned int irq)
++#include <xen/interface/platform.h>
++static int dbs_calc_load(struct cpu_dbs_info_s *this_dbs_info)
+{
-+ unsigned int port;
-+
-+ port = 0x4d0 + (irq >> 3);
-+ return (inb(port) >> (irq & 7)) & 1;
++ int load = 0;
++ struct xen_platform_op op;
++ uint64_t idletime[NR_CPUS];
++ struct cpufreq_policy *policy;
++ unsigned int j;
++ cpumask_t cpumap;
++
++ policy = this_dbs_info->cur_policy;
++ cpumap = policy->cpus;
++
++ op.cmd = XENPF_getidletime;
++ set_xen_guest_handle(op.u.getidletime.cpumap_bitmap, (uint8_t *) cpus_addr(cpumap));
++ op.u.getidletime.cpumap_nr_cpus = NR_CPUS;
++ set_xen_guest_handle(op.u.getidletime.idletime, idletime);
++ if (HYPERVISOR_platform_op(&op))
++ return 200;
++
++ for_each_cpu_mask(j, cpumap) {
++ cputime64_t total_idle_nsecs, tmp_idle_nsecs;
++ cputime64_t total_wall_nsecs, tmp_wall_nsecs;
++ struct cpu_dbs_info_s *j_dbs_info;
++ unsigned long tmp_load, tmp_wall_msecs, tmp_idle_msecs;
++
++ j_dbs_info = &per_cpu(cpu_dbs_info, j);
++ total_idle_nsecs = idletime[j];
++ tmp_idle_nsecs = cputime64_sub(total_idle_nsecs,
++ j_dbs_info->prev_cpu_idle);
++ total_wall_nsecs = op.u.getidletime.now;
++ tmp_wall_nsecs = cputime64_sub(total_wall_nsecs,
++ j_dbs_info->prev_cpu_wall);
++
++ if (tmp_wall_nsecs == 0)
++ return 200;
++
++ j_dbs_info->prev_cpu_wall = total_wall_nsecs;
++ j_dbs_info->prev_cpu_idle = total_idle_nsecs;
++
++ /* Convert nsecs to msecs and clamp times to sane values. */
++ do_div(tmp_wall_nsecs, 1000000);
++ tmp_wall_msecs = tmp_wall_nsecs;
++ do_div(tmp_idle_nsecs, 1000000);
++ tmp_idle_msecs = tmp_idle_nsecs;
++ if (tmp_wall_msecs == 0)
++ tmp_wall_msecs = 1;
++ if (tmp_idle_msecs > tmp_wall_msecs)
++ tmp_idle_msecs = tmp_wall_msecs;
++
++ tmp_load = (100 * (tmp_wall_msecs - tmp_idle_msecs)) /
++ tmp_wall_msecs;
++ load = max(load, min(100, (int) tmp_load));
++ }
++ return load;
+}
++#endif
+
-+static void __init construct_default_ioirq_mptable(int mpc_default_type)
++static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info)
+{
-+ struct mpc_config_intsrc intsrc;
-+ int i;
-+ int ELCR_fallback = 0;
++ int load;
+
-+ intsrc.mpc_type = MP_INTSRC;
-+ intsrc.mpc_irqflag = 0; /* conforming */
-+ intsrc.mpc_srcbus = 0;
-+ intsrc.mpc_dstapic = mp_ioapics[0].mpc_apicid;
++ struct cpufreq_policy *policy;
+
-+ intsrc.mpc_irqtype = mp_INT;
++ if (!this_dbs_info->enable)
++ return;
+
-+ /*
-+ * If true, we have an ISA/PCI system with no IRQ entries
-+ * in the MP table. To prevent the PCI interrupts from being set up
-+ * incorrectly, we try to use the ELCR. The sanity check to see if
-+ * there is good ELCR data is very simple - IRQ0, 1, 2 and 13 can
-+ * never be level sensitive, so we simply see if the ELCR agrees.
-+ * If it does, we assume it's valid.
-+ */
-+ if (mpc_default_type == 5) {
-+ printk(KERN_INFO "ISA/PCI bus type with no IRQ information... falling back to ELCR\n");
++ policy = this_dbs_info->cur_policy;
++ load = dbs_calc_load(this_dbs_info);
++ if (load > 100)
++ return;
+
+ /* Check for frequency increase */
+ if (load > dbs_tuners_ins.up_threshold) {
+diff -r d894e36cfc30 -r 0aa021803deb drivers/cpufreq/cpufreq_stats.c
+--- a/drivers/cpufreq/cpufreq_stats.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/cpufreq/cpufreq_stats.c Wed Sep 10 10:54:08 2008 +0100
+@@ -292,6 +292,9 @@
+ if (old_index == new_index)
+ return 0;
+
++ if ((old_index < 0) || (new_index < 0))
++ return 0;
+
-+ if (ELCR_trigger(0) || ELCR_trigger(1) || ELCR_trigger(2) || ELCR_trigger(13))
-+ printk(KERN_ERR "ELCR contains invalid data... not using ELCR\n");
-+ else {
-+ printk(KERN_INFO "Using ELCR to identify PCI interrupts\n");
-+ ELCR_fallback = 1;
-+ }
-+ }
+ spin_lock(&cpufreq_stats_lock);
+ stat->last_index = new_index;
+ #ifdef CONFIG_CPU_FREQ_STAT_DETAILS
+diff -r d894e36cfc30 -r 0aa021803deb drivers/ide/ide-lib.c
+--- a/drivers/ide/ide-lib.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/ide/ide-lib.c Wed Sep 10 10:54:08 2008 +0100
+@@ -408,10 +408,10 @@
+ {
+ u64 addr = BLK_BOUNCE_HIGH; /* dma64_addr_t */
+
+- if (!PCI_DMA_BUS_IS_PHYS) {
+- addr = BLK_BOUNCE_ANY;
+- } else if (on && drive->media == ide_disk) {
+- if (HWIF(drive)->pci_dev)
++ if (on && drive->media == ide_disk) {
++ if (!PCI_DMA_BUS_IS_PHYS)
++ addr = BLK_BOUNCE_ANY;
++ else if (HWIF(drive)->pci_dev)
+ addr = HWIF(drive)->pci_dev->dma_mask;
+ }
+
+diff -r d894e36cfc30 -r 0aa021803deb drivers/oprofile/buffer_sync.c
+--- a/drivers/oprofile/buffer_sync.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/oprofile/buffer_sync.c Wed Sep 10 10:54:08 2008 +0100
+@@ -5,6 +5,10 @@
+ * @remark Read the file COPYING
+ *
+ * @author John Levon <levon at movementarian.org>
++ *
++ * Modified by Aravind Menon for Xen
++ * These modifications are:
++ * Copyright (C) 2005 Hewlett-Packard Co.
+ *
+ * This is the core of the buffer management. Each
+ * CPU buffer is processed and entered into the
+@@ -38,6 +42,7 @@
+ static DEFINE_SPINLOCK(task_mortuary);
+ static void process_task_mortuary(void);
+
++static int cpu_current_domain[NR_CPUS];
+
+ /* Take ownership of the task struct and place it on the
+ * list for processing. Only after two full buffer syncs
+@@ -146,6 +151,11 @@
+ int sync_start(void)
+ {
+ int err;
++ int i;
+
-+ for (i = 0; i < 16; i++) {
-+ switch (mpc_default_type) {
-+ case 2:
-+ if (i == 0 || i == 13)
-+ continue; /* IRQ0 & IRQ13 not connected */
-+ /* fall through */
-+ default:
-+ if (i == 2)
-+ continue; /* IRQ2 is never connected */
-+ }
++ for (i = 0; i < NR_CPUS; i++) {
++ cpu_current_domain[i] = COORDINATOR_DOMAIN;
++ }
+
+ start_cpu_work();
+
+@@ -275,15 +285,31 @@
+ last_cookie = INVALID_COOKIE;
+ }
+
+-static void add_kernel_ctx_switch(unsigned int in_kernel)
++static void add_cpu_mode_switch(unsigned int cpu_mode)
+ {
+ add_event_entry(ESCAPE_CODE);
+- if (in_kernel)
+- add_event_entry(KERNEL_ENTER_SWITCH_CODE);
+- else
+- add_event_entry(KERNEL_EXIT_SWITCH_CODE);
++ switch (cpu_mode) {
++ case CPU_MODE_USER:
++ add_event_entry(USER_ENTER_SWITCH_CODE);
++ break;
++ case CPU_MODE_KERNEL:
++ add_event_entry(KERNEL_ENTER_SWITCH_CODE);
++ break;
++ case CPU_MODE_XEN:
++ add_event_entry(XEN_ENTER_SWITCH_CODE);
++ break;
++ default:
++ break;
++ }
+ }
+-
+
-+ if (ELCR_fallback) {
-+ /*
-+ * If the ELCR indicates a level-sensitive interrupt, we
-+ * copy that information over to the MP table in the
-+ * irqflag field (level sensitive, active high polarity).
-+ */
-+ if (ELCR_trigger(i))
-+ intsrc.mpc_irqflag = 13;
-+ else
-+ intsrc.mpc_irqflag = 0;
-+ }
++static void add_domain_switch(unsigned long domain_id)
++{
++ add_event_entry(ESCAPE_CODE);
++ add_event_entry(DOMAIN_SWITCH_CODE);
++ add_event_entry(domain_id);
++}
+
-+ intsrc.mpc_srcbusirq = i;
-+ intsrc.mpc_dstirq = i ? i : 2; /* IRQ0 to INTIN2 */
-+ MP_intsrc_info(&intsrc);
+ static void
+ add_user_ctx_switch(struct task_struct const * task, unsigned long cookie)
+ {
+@@ -348,9 +374,9 @@
+ * for later lookup from userspace.
+ */
+ static int
+-add_sample(struct mm_struct * mm, struct op_sample * s, int in_kernel)
++add_sample(struct mm_struct * mm, struct op_sample * s, int cpu_mode)
+ {
+- if (in_kernel) {
++ if (cpu_mode >= CPU_MODE_KERNEL) {
+ add_sample_entry(s->eip, s->event);
+ return 1;
+ } else if (mm) {
+@@ -496,15 +522,21 @@
+ struct mm_struct *mm = NULL;
+ struct task_struct * new;
+ unsigned long cookie = 0;
+- int in_kernel = 1;
++ int cpu_mode = 1;
+ unsigned int i;
+ sync_buffer_state state = sb_buffer_start;
+ unsigned long available;
++ int domain_switch = 0;
+
+ mutex_lock(&buffer_mutex);
+
+ add_cpu_switch(cpu);
+
++ /* We need to assign the first samples in this CPU buffer to the
++ same domain that we were processing at the last sync_buffer */
++ if (cpu_current_domain[cpu] != COORDINATOR_DOMAIN) {
++ add_domain_switch(cpu_current_domain[cpu]);
++ }
+ /* Remember, only we can modify tail_pos */
+
+ available = get_slots(cpu_buf);
+@@ -512,16 +544,18 @@
+ for (i = 0; i < available; ++i) {
+ struct op_sample * s = &cpu_buf->buffer[cpu_buf->tail_pos];
+
+- if (is_code(s->eip)) {
+- if (s->event <= CPU_IS_KERNEL) {
+- /* kernel/userspace switch */
+- in_kernel = s->event;
++ if (is_code(s->eip) && !domain_switch) {
++ if (s->event <= CPU_MODE_XEN) {
++ /* xen/kernel/userspace switch */
++ cpu_mode = s->event;
+ if (state == sb_buffer_start)
+ state = sb_sample_start;
+- add_kernel_ctx_switch(s->event);
++ add_cpu_mode_switch(s->event);
+ } else if (s->event == CPU_TRACE_BEGIN) {
+ state = sb_bt_start;
+ add_trace_begin();
++ } else if (s->event == CPU_DOMAIN_SWITCH) {
++ domain_switch = 1;
+ } else {
+ struct mm_struct * oldmm = mm;
+
+@@ -535,11 +569,21 @@
+ add_user_ctx_switch(new, cookie);
+ }
+ } else {
+- if (state >= sb_bt_start &&
+- !add_sample(mm, s, in_kernel)) {
+- if (state == sb_bt_start) {
+- state = sb_bt_ignore;
+- atomic_inc(&oprofile_stats.bt_lost_no_mapping);
++ if (domain_switch) {
++ cpu_current_domain[cpu] = s->eip;
++ add_domain_switch(s->eip);
++ domain_switch = 0;
++ } else {
++ if (cpu_current_domain[cpu] !=
++ COORDINATOR_DOMAIN) {
++ add_sample_entry(s->eip, s->event);
++ }
++ else if (state >= sb_bt_start &&
++ !add_sample(mm, s, cpu_mode)) {
++ if (state == sb_bt_start) {
++ state = sb_bt_ignore;
++ atomic_inc(&oprofile_stats.bt_lost_no_mapping);
++ }
+ }
+ }
+ }
+@@ -548,6 +592,11 @@
+ }
+ release_mm(mm);
+
++ /* We reset domain to COORDINATOR at each CPU switch */
++ if (cpu_current_domain[cpu] != COORDINATOR_DOMAIN) {
++ add_domain_switch(COORDINATOR_DOMAIN);
+ }
+
-+ intsrc.mpc_irqtype = mp_ExtINT;
-+ intsrc.mpc_srcbusirq = 0;
-+ intsrc.mpc_dstirq = 0; /* 8259A to INTIN0 */
-+ MP_intsrc_info(&intsrc);
-+}
+ mark_done(cpu);
+
+ mutex_unlock(&buffer_mutex);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/oprofile/cpu_buffer.c
+--- a/drivers/oprofile/cpu_buffer.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/oprofile/cpu_buffer.c Wed Sep 10 10:54:08 2008 +0100
+@@ -5,6 +5,10 @@
+ * @remark Read the file COPYING
+ *
+ * @author John Levon <levon at movementarian.org>
++ *
++ * Modified by Aravind Menon for Xen
++ * These modifications are:
++ * Copyright (C) 2005 Hewlett-Packard Co.
+ *
+ * Each CPU has a local buffer that stores PC value/event
+ * pairs. We also log context switches when we notice them.
+@@ -34,6 +38,8 @@
+ #define DEFAULT_TIMER_EXPIRE (HZ / 10)
+ static int work_enabled;
+
++static int32_t current_domain = COORDINATOR_DOMAIN;
+
-+static inline void __init construct_default_ISA_mptable(int mpc_default_type)
+ void free_cpu_buffers(void)
+ {
+ int i;
+@@ -57,7 +63,7 @@
+ goto fail;
+
+ b->last_task = NULL;
+- b->last_is_kernel = -1;
++ b->last_cpu_mode = -1;
+ b->tracing = 0;
+ b->buffer_size = buffer_size;
+ b->tail_pos = 0;
+@@ -113,7 +119,7 @@
+ * collected will populate the buffer with proper
+ * values to initialize the buffer
+ */
+- cpu_buf->last_is_kernel = -1;
++ cpu_buf->last_cpu_mode = -1;
+ cpu_buf->last_task = NULL;
+ }
+
+@@ -163,13 +169,13 @@
+ * because of the head/tail separation of the writer and reader
+ * of the CPU buffer.
+ *
+- * is_kernel is needed because on some architectures you cannot
++ * cpu_mode is needed because on some architectures you cannot
+ * tell if you are in kernel or user space simply by looking at
+- * pc. We tag this in the buffer by generating kernel enter/exit
+- * events whenever is_kernel changes
++ * pc. We tag this in the buffer by generating kernel/user (and xen)
++ * enter events whenever cpu_mode changes
+ */
+ static int log_sample(struct oprofile_cpu_buffer * cpu_buf, unsigned long pc,
+- int is_kernel, unsigned long event)
++ int cpu_mode, unsigned long event)
+ {
+ struct task_struct * task;
+
+@@ -180,18 +186,18 @@
+ return 0;
+ }
+
+- is_kernel = !!is_kernel;
+-
+ task = current;
+
+ /* notice a switch from user->kernel or vice versa */
+- if (cpu_buf->last_is_kernel != is_kernel) {
+- cpu_buf->last_is_kernel = is_kernel;
+- add_code(cpu_buf, is_kernel);
++ if (cpu_buf->last_cpu_mode != cpu_mode) {
++ cpu_buf->last_cpu_mode = cpu_mode;
++ add_code(cpu_buf, cpu_mode);
+ }
+-
++
+ /* notice a task switch */
+- if (cpu_buf->last_task != task) {
++ /* if not processing other domain samples */
++ if ((cpu_buf->last_task != task) &&
++ (current_domain == COORDINATOR_DOMAIN)) {
+ cpu_buf->last_task = task;
+ add_code(cpu_buf, (unsigned long)task);
+ }
+@@ -275,6 +281,25 @@
+ add_sample(cpu_buf, pc, 0);
+ }
+
++int oprofile_add_domain_switch(int32_t domain_id)
+{
-+ struct mpc_config_processor processor;
-+ struct mpc_config_bus bus;
-+ struct mpc_config_ioapic ioapic;
-+ struct mpc_config_lintsrc lintsrc;
-+ int linttypes[2] = { mp_ExtINT, mp_NMI };
-+ int i;
++ struct oprofile_cpu_buffer * cpu_buf = &cpu_buffer[smp_processor_id()];
+
-+ /*
-+ * local APIC has default address
-+ */
-+ mp_lapic_addr = APIC_DEFAULT_PHYS_BASE;
++ /* should have space for switching into and out of domain
++ (2 slots each) plus one sample and one cpu mode switch */
++ if (((nr_available_slots(cpu_buf) < 6) &&
++ (domain_id != COORDINATOR_DOMAIN)) ||
++ (nr_available_slots(cpu_buf) < 2))
++ return 0;
+
-+ /*
-+ * 2 CPUs, numbered 0 & 1.
-+ */
-+ processor.mpc_type = MP_PROCESSOR;
-+ /* Either an integrated APIC or a discrete 82489DX. */
-+ processor.mpc_apicver = mpc_default_type > 4 ? 0x10 : 0x01;
-+ processor.mpc_cpuflag = CPU_ENABLED;
-+ processor.mpc_cpufeature = (boot_cpu_data.x86 << 8) |
-+ (boot_cpu_data.x86_model << 4) |
-+ boot_cpu_data.x86_mask;
-+ processor.mpc_featureflag = boot_cpu_data.x86_capability[0];
-+ processor.mpc_reserved[0] = 0;
-+ processor.mpc_reserved[1] = 0;
-+ for (i = 0; i < 2; i++) {
-+ processor.mpc_apicid = i;
-+ MP_processor_info(&processor);
-+ }
++ add_code(cpu_buf, CPU_DOMAIN_SWITCH);
++ add_sample(cpu_buf, domain_id, 0);
+
-+ bus.mpc_type = MP_BUS;
-+ bus.mpc_busid = 0;
-+ switch (mpc_default_type) {
-+ default:
-+ printk(KERN_ERR "???\nUnknown standard configuration %d\n",
-+ mpc_default_type);
-+ /* fall through */
-+ case 1:
-+ case 5:
-+ memcpy(bus.mpc_bustype, "ISA ", 6);
-+ break;
-+ case 2:
-+ case 6:
-+ case 3:
-+ memcpy(bus.mpc_bustype, "EISA ", 6);
-+ break;
-+ case 4:
-+ case 7:
-+ memcpy(bus.mpc_bustype, "MCA ", 6);
-+ }
-+ MP_bus_info(&bus);
-+ if (mpc_default_type > 4) {
-+ bus.mpc_busid = 1;
-+ memcpy(bus.mpc_bustype, "PCI ", 6);
-+ MP_bus_info(&bus);
-+ }
++ current_domain = domain_id;
+
-+ ioapic.mpc_type = MP_IOAPIC;
-+ ioapic.mpc_apicid = 2;
-+ ioapic.mpc_apicver = mpc_default_type > 4 ? 0x10 : 0x01;
-+ ioapic.mpc_flags = MPC_APIC_USABLE;
-+ ioapic.mpc_apicaddr = 0xFEC00000;
-+ MP_ioapic_info(&ioapic);
++ return 1;
++}
+
-+ /*
-+ * We set up most of the low 16 IO-APIC pins according to MPS rules.
-+ */
-+ construct_default_ioirq_mptable(mpc_default_type);
+ /*
+ * This serves to avoid cpu buffer overflow, and makes sure
+ * the task mortuary progresses
+diff -r d894e36cfc30 -r 0aa021803deb drivers/oprofile/cpu_buffer.h
+--- a/drivers/oprofile/cpu_buffer.h Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/oprofile/cpu_buffer.h Wed Sep 10 10:54:08 2008 +0100
+@@ -36,7 +36,7 @@
+ volatile unsigned long tail_pos;
+ unsigned long buffer_size;
+ struct task_struct * last_task;
+- int last_is_kernel;
++ int last_cpu_mode;
+ int tracing;
+ struct op_sample * buffer;
+ unsigned long sample_received;
+@@ -51,7 +51,10 @@
+ void cpu_buffer_reset(struct oprofile_cpu_buffer * cpu_buf);
+
+ /* transient events for the CPU buffer -> event buffer */
+-#define CPU_IS_KERNEL 1
+-#define CPU_TRACE_BEGIN 2
++#define CPU_MODE_USER 0
++#define CPU_MODE_KERNEL 1
++#define CPU_MODE_XEN 2
++#define CPU_TRACE_BEGIN 3
++#define CPU_DOMAIN_SWITCH 4
+
+ #endif /* OPROFILE_CPU_BUFFER_H */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/oprofile/event_buffer.h
+--- a/drivers/oprofile/event_buffer.h Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/oprofile/event_buffer.h Wed Sep 10 10:54:08 2008 +0100
+@@ -29,14 +29,19 @@
+ #define CPU_SWITCH_CODE 2
+ #define COOKIE_SWITCH_CODE 3
+ #define KERNEL_ENTER_SWITCH_CODE 4
+-#define KERNEL_EXIT_SWITCH_CODE 5
++#define USER_ENTER_SWITCH_CODE 5
+ #define MODULE_LOADED_CODE 6
+ #define CTX_TGID_CODE 7
+ #define TRACE_BEGIN_CODE 8
+ #define TRACE_END_CODE 9
++#define XEN_ENTER_SWITCH_CODE 10
++#define DOMAIN_SWITCH_CODE 11
+
+ #define INVALID_COOKIE ~0UL
+ #define NO_COOKIE 0UL
+
-+ lintsrc.mpc_type = MP_LINTSRC;
-+ lintsrc.mpc_irqflag = 0; /* conforming */
-+ lintsrc.mpc_srcbusid = 0;
-+ lintsrc.mpc_srcbusirq = 0;
-+ lintsrc.mpc_destapic = MP_APIC_ALL;
-+ for (i = 0; i < 2; i++) {
-+ lintsrc.mpc_irqtype = linttypes[i];
-+ lintsrc.mpc_destapiclint = i;
-+ MP_lintsrc_info(&lintsrc);
-+ }
-+}
++/* Constant used to refer to coordinator domain (Xen) */
++#define COORDINATOR_DOMAIN -1
+
+ /* add data to the event buffer */
+ void add_event_entry(unsigned long data);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/oprofile/oprof.c
+--- a/drivers/oprofile/oprof.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/oprofile/oprof.c Wed Sep 10 10:54:08 2008 +0100
+@@ -5,6 +5,10 @@
+ * @remark Read the file COPYING
+ *
+ * @author John Levon <levon at movementarian.org>
++ *
++ * Modified by Aravind Menon for Xen
++ * These modifications are:
++ * Copyright (C) 2005 Hewlett-Packard Co.
+ */
+
+ #include <linux/kernel.h>
+@@ -19,7 +23,7 @@
+ #include "cpu_buffer.h"
+ #include "buffer_sync.h"
+ #include "oprofile_stats.h"
+-
+
-+static struct intel_mp_floating *mpf_found;
+ struct oprofile_operations oprofile_ops;
+
+ unsigned long oprofile_started;
+@@ -32,6 +36,32 @@
+ 1 - use the timer int mechanism regardless
+ */
+ static int timer = 0;
+
-+/*
-+ * Scan the memory blocks for an SMP configuration block.
-+ */
-+void __init get_smp_config (void)
++int oprofile_set_active(int active_domains[], unsigned int adomains)
+{
-+ struct intel_mp_floating *mpf = mpf_found;
-+
-+ /*
-+ * ACPI supports both logical (e.g. Hyper-Threading) and physical
-+ * processors, where MPS only supports physical.
-+ */
-+ if (acpi_lapic && acpi_ioapic) {
-+ printk(KERN_INFO "Using ACPI (MADT) for SMP configuration information\n");
-+ return;
-+ }
-+ else if (acpi_lapic)
-+ printk(KERN_INFO "Using ACPI for processor (LAPIC) configuration information\n");
-+
-+ printk("Intel MultiProcessor Specification v1.%d\n", mpf->mpf_specification);
-+ if (mpf->mpf_feature2 & (1<<7)) {
-+ printk(KERN_INFO " IMCR and PIC compatibility mode.\n");
-+ pic_mode = 1;
-+ } else {
-+ printk(KERN_INFO " Virtual Wire compatibility mode.\n");
-+ pic_mode = 0;
-+ }
++ int err;
+
-+ /*
-+ * Now see if we need to read further.
-+ */
-+ if (mpf->mpf_feature1 != 0) {
++ if (!oprofile_ops.set_active)
++ return -EINVAL;
+
-+ printk(KERN_INFO "Default MP configuration #%d\n", mpf->mpf_feature1);
-+ construct_default_ISA_mptable(mpf->mpf_feature1);
++ mutex_lock(&start_mutex);
++ err = oprofile_ops.set_active(active_domains, adomains);
++ mutex_unlock(&start_mutex);
++ return err;
++}
+
-+ } else if (mpf->mpf_physptr) {
++int oprofile_set_passive(int passive_domains[], unsigned int pdomains)
++{
++ int err;
+
-+ /*
-+ * Read the physical hardware table. Anything here will
-+ * override the defaults.
-+ */
-+ if (!smp_read_mpc(isa_bus_to_virt(mpf->mpf_physptr))) {
-+ smp_found_config = 0;
-+ printk(KERN_ERR "BIOS bug, MP table errors detected!...\n");
-+ printk(KERN_ERR "... disabling SMP support. (tell your hw vendor)\n");
-+ return;
-+ }
-+ /*
-+ * If there are no explicit MP IRQ entries, then we are
-+ * broken. We set up most of the low 16 IO-APIC pins to
-+ * ISA defaults and hope it will work.
-+ */
-+ if (!mp_irq_entries) {
-+ struct mpc_config_bus bus;
++ if (!oprofile_ops.set_passive)
++ return -EINVAL;
+
-+ printk(KERN_ERR "BIOS bug, no explicit IRQ entries, using default mptable. (tell your hw vendor)\n");
++ mutex_lock(&start_mutex);
++ err = oprofile_ops.set_passive(passive_domains, pdomains);
++ mutex_unlock(&start_mutex);
++ return err;
++}
+
+ int oprofile_setup(void)
+ {
+diff -r d894e36cfc30 -r 0aa021803deb drivers/oprofile/oprof.h
+--- a/drivers/oprofile/oprof.h Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/oprofile/oprof.h Wed Sep 10 10:54:08 2008 +0100
+@@ -35,5 +35,8 @@
+ void oprofile_timer_init(struct oprofile_operations * ops);
+
+ int oprofile_set_backtrace(unsigned long depth);
+
-+ bus.mpc_type = MP_BUS;
-+ bus.mpc_busid = 0;
-+ memcpy(bus.mpc_bustype, "ISA ", 6);
-+ MP_bus_info(&bus);
++int oprofile_set_active(int active_domains[], unsigned int adomains);
++int oprofile_set_passive(int passive_domains[], unsigned int pdomains);
+
+ #endif /* OPROF_H */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/oprofile/oprofile_files.c
+--- a/drivers/oprofile/oprofile_files.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/oprofile/oprofile_files.c Wed Sep 10 10:54:08 2008 +0100
+@@ -5,15 +5,21 @@
+ * @remark Read the file COPYING
+ *
+ * @author John Levon <levon at movementarian.org>
++ *
++ * Modified by Aravind Menon for Xen
++ * These modifications are:
++ * Copyright (C) 2005 Hewlett-Packard Co.
+ */
+
+ #include <linux/fs.h>
+ #include <linux/oprofile.h>
++#include <asm/uaccess.h>
++#include <linux/ctype.h>
+
+ #include "event_buffer.h"
+ #include "oprofile_stats.h"
+ #include "oprof.h"
+-
+
-+ construct_default_ioirq_mptable(0);
-+ }
+ unsigned long fs_buffer_size = 131072;
+ unsigned long fs_cpu_buffer_size = 8192;
+ unsigned long fs_buffer_watershed = 32768; /* FIXME: tune */
+@@ -117,11 +123,202 @@
+ static struct file_operations dump_fops = {
+ .write = dump_write,
+ };
+-
+
-+ } else
-+ BUG();
++#define TMPBUFSIZE 512
+
-+ printk(KERN_INFO "Processors: %d\n", num_processors);
-+ /*
-+ * Only use the first configuration found.
-+ */
-+}
++static unsigned int adomains = 0;
++static int active_domains[MAX_OPROF_DOMAINS + 1];
++static DEFINE_MUTEX(adom_mutex);
+
-+static int __init smp_scan_config (unsigned long base, unsigned long length)
++static ssize_t adomain_write(struct file * file, char const __user * buf,
++ size_t count, loff_t * offset)
+{
-+ extern void __bad_mpf_size(void);
-+ unsigned int *bp = isa_bus_to_virt(base);
-+ struct intel_mp_floating *mpf;
-+
-+ Dprintk("Scan SMP from %p for %ld bytes.\n", bp,length);
-+ if (sizeof(*mpf) != 16)
-+ __bad_mpf_size();
++ char *tmpbuf;
++ char *startp, *endp;
++ int i;
++ unsigned long val;
++ ssize_t retval = count;
++
++ if (*offset)
++ return -EINVAL;
++ if (count > TMPBUFSIZE - 1)
++ return -EINVAL;
+
-+ while (length > 0) {
-+ mpf = (struct intel_mp_floating *)bp;
-+ if ((*bp == SMP_MAGIC_IDENT) &&
-+ (mpf->mpf_length == 1) &&
-+ !mpf_checksum((unsigned char *)bp, 16) &&
-+ ((mpf->mpf_specification == 1)
-+ || (mpf->mpf_specification == 4)) ) {
++ if (!(tmpbuf = kmalloc(TMPBUFSIZE, GFP_KERNEL)))
++ return -ENOMEM;
+
-+ smp_found_config = 1;
-+ mpf_found = mpf;
-+ return 1;
-+ }
-+ bp += 4;
-+ length -= 16;
++ if (copy_from_user(tmpbuf, buf, count)) {
++ kfree(tmpbuf);
++ return -EFAULT;
+ }
-+ return 0;
-+}
++ tmpbuf[count] = 0;
+
-+void __init find_intel_smp (void)
-+{
-+ unsigned int address;
++ mutex_lock(&adom_mutex);
+
-+ /*
-+ * FIXME: Linux assumes you have 640K of base ram..
-+ * this continues the error...
-+ *
-+ * 1) Scan the bottom 1K for a signature
-+ * 2) Scan the top 1K of base RAM
-+ * 3) Scan the 64K of bios
-+ */
-+ if (smp_scan_config(0x0,0x400) ||
-+ smp_scan_config(639*0x400,0x400) ||
-+ smp_scan_config(0xF0000,0x10000))
-+ return;
-+ /*
-+ * If it is an SMP machine we should know now, unless the
-+ * configuration is in an EISA/MCA bus machine with an
-+ * extended bios data area.
-+ *
-+ * there is a real-mode segmented pointer pointing to the
-+ * 4K EBDA area at 0x40E, calculate and scan it here.
-+ *
-+ * NOTE! There are Linux loaders that will corrupt the EBDA
-+ * area, and as such this kind of SMP config may be less
-+ * trustworthy, simply because the SMP table may have been
-+ * stomped on during early boot. These loaders are buggy and
-+ * should be fixed.
-+ */
++ startp = tmpbuf;
++ /* Parse one more than MAX_OPROF_DOMAINS, for easy error checking */
++ for (i = 0; i <= MAX_OPROF_DOMAINS; i++) {
++ val = simple_strtoul(startp, &endp, 0);
++ if (endp == startp)
++ break;
++ while (ispunct(*endp) || isspace(*endp))
++ endp++;
++ active_domains[i] = val;
++ if (active_domains[i] != val)
++ /* Overflow, force error below */
++ i = MAX_OPROF_DOMAINS + 1;
++ startp = endp;
++ }
++ /* Force error on trailing junk */
++ adomains = *startp ? MAX_OPROF_DOMAINS + 1 : i;
+
-+ address = *(unsigned short *)phys_to_virt(0x40E);
-+ address <<= 4;
-+ if (smp_scan_config(address, 0x1000))
-+ return;
++ kfree(tmpbuf);
+
-+ /* If we have come this far, we did not find an MP table */
-+ printk(KERN_INFO "No mptable found.\n");
-+}
++ if (adomains > MAX_OPROF_DOMAINS
++ || oprofile_set_active(active_domains, adomains)) {
++ adomains = 0;
++ retval = -EINVAL;
++ }
+
-+/*
-+ * - Intel MP Configuration Table
-+ */
-+void __init find_smp_config (void)
-+{
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ find_intel_smp();
-+#endif
++ mutex_unlock(&adom_mutex);
++ return retval;
+}
+
++static ssize_t adomain_read(struct file * file, char __user * buf,
++ size_t count, loff_t * offset)
++{
++ char * tmpbuf;
++ size_t len;
++ int i;
++ ssize_t retval;
+
-+/* --------------------------------------------------------------------------
-+ ACPI-based MP Configuration
-+ -------------------------------------------------------------------------- */
++ if (!(tmpbuf = kmalloc(TMPBUFSIZE, GFP_KERNEL)))
++ return -ENOMEM;
+
-+#ifdef CONFIG_ACPI
++ mutex_lock(&adom_mutex);
+
-+void __init mp_register_lapic_address (
-+ u64 address)
-+{
-+#ifndef CONFIG_XEN
-+ mp_lapic_addr = (unsigned long) address;
++ len = 0;
++ for (i = 0; i < adomains; i++)
++ len += snprintf(tmpbuf + len,
++ len < TMPBUFSIZE ? TMPBUFSIZE - len : 0,
++ "%u ", active_domains[i]);
++ WARN_ON(len > TMPBUFSIZE);
++ if (len != 0 && len <= TMPBUFSIZE)
++ tmpbuf[len-1] = '\n';
+
-+ set_fixmap_nocache(FIX_APIC_BASE, mp_lapic_addr);
++ mutex_unlock(&adom_mutex);
+
-+ if (boot_cpu_id == -1U)
-+ boot_cpu_id = GET_APIC_ID(apic_read(APIC_ID));
++ retval = simple_read_from_buffer(buf, count, offset, tmpbuf, len);
+
-+ Dprintk("Boot CPU = %d\n", boot_cpu_physical_apicid);
-+#endif
++ kfree(tmpbuf);
++ return retval;
+}
+
+
-+void __cpuinit mp_register_lapic (
-+ u8 id,
-+ u8 enabled)
-+{
-+ struct mpc_config_processor processor;
-+ int boot_cpu = 0;
-+
-+ if (id >= MAX_APICS) {
-+ printk(KERN_WARNING "Processor #%d invalid (max %d)\n",
-+ id, MAX_APICS);
-+ return;
-+ }
-+
-+ if (id == boot_cpu_physical_apicid)
-+ boot_cpu = 1;
++static struct file_operations active_domain_ops = {
++ .read = adomain_read,
++ .write = adomain_write,
++};
+
-+#ifndef CONFIG_XEN
-+ processor.mpc_type = MP_PROCESSOR;
-+ processor.mpc_apicid = id;
-+ processor.mpc_apicver = GET_APIC_VERSION(apic_read(APIC_LVR));
-+ processor.mpc_cpuflag = (enabled ? CPU_ENABLED : 0);
-+ processor.mpc_cpuflag |= (boot_cpu ? CPU_BOOTPROCESSOR : 0);
-+ processor.mpc_cpufeature = (boot_cpu_data.x86 << 8) |
-+ (boot_cpu_data.x86_model << 4) | boot_cpu_data.x86_mask;
-+ processor.mpc_featureflag = boot_cpu_data.x86_capability[0];
-+ processor.mpc_reserved[0] = 0;
-+ processor.mpc_reserved[1] = 0;
-+#endif
++static unsigned int pdomains = 0;
++static int passive_domains[MAX_OPROF_DOMAINS];
++static DEFINE_MUTEX(pdom_mutex);
+
-+ MP_processor_info(&processor);
-+}
++static ssize_t pdomain_write(struct file * file, char const __user * buf,
++ size_t count, loff_t * offset)
++{
++ char *tmpbuf;
++ char *startp, *endp;
++ int i;
++ unsigned long val;
++ ssize_t retval = count;
++
++ if (*offset)
++ return -EINVAL;
++ if (count > TMPBUFSIZE - 1)
++ return -EINVAL;
+
-+#ifdef CONFIG_X86_IO_APIC
++ if (!(tmpbuf = kmalloc(TMPBUFSIZE, GFP_KERNEL)))
++ return -ENOMEM;
+
-+#define MP_ISA_BUS 0
-+#define MP_MAX_IOAPIC_PIN 127
++ if (copy_from_user(tmpbuf, buf, count)) {
++ kfree(tmpbuf);
++ return -EFAULT;
++ }
++ tmpbuf[count] = 0;
+
-+static struct mp_ioapic_routing {
-+ int apic_id;
-+ int gsi_start;
-+ int gsi_end;
-+ u32 pin_programmed[4];
-+} mp_ioapic_routing[MAX_IO_APICS];
++ mutex_lock(&pdom_mutex);
+
++ startp = tmpbuf;
++ /* Parse one more than MAX_OPROF_DOMAINS, for easy error checking */
++ for (i = 0; i <= MAX_OPROF_DOMAINS; i++) {
++ val = simple_strtoul(startp, &endp, 0);
++ if (endp == startp)
++ break;
++ while (ispunct(*endp) || isspace(*endp))
++ endp++;
++ passive_domains[i] = val;
++ if (passive_domains[i] != val)
++ /* Overflow, force error below */
++ i = MAX_OPROF_DOMAINS + 1;
++ startp = endp;
++ }
++ /* Force error on trailing junk */
++ pdomains = *startp ? MAX_OPROF_DOMAINS + 1 : i;
+
-+static int mp_find_ioapic (
-+ int gsi)
-+{
-+ int i = 0;
++ kfree(tmpbuf);
+
-+ /* Find the IOAPIC that manages this GSI. */
-+ for (i = 0; i < nr_ioapics; i++) {
-+ if ((gsi >= mp_ioapic_routing[i].gsi_start)
-+ && (gsi <= mp_ioapic_routing[i].gsi_end))
-+ return i;
++ if (pdomains > MAX_OPROF_DOMAINS
++ || oprofile_set_passive(passive_domains, pdomains)) {
++ pdomains = 0;
++ retval = -EINVAL;
+ }
+
-+ printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi);
-+
-+ return -1;
++ mutex_unlock(&pdom_mutex);
++ return retval;
+}
-+
+
-+void __init mp_register_ioapic (
-+ u8 id,
-+ u32 address,
-+ u32 gsi_base)
++static ssize_t pdomain_read(struct file * file, char __user * buf,
++ size_t count, loff_t * offset)
+{
-+ int idx = 0;
++ char * tmpbuf;
++ size_t len;
++ int i;
++ ssize_t retval;
+
-+ if (nr_ioapics >= MAX_IO_APICS) {
-+ printk(KERN_ERR "ERROR: Max # of I/O APICs (%d) exceeded "
-+ "(found %d)\n", MAX_IO_APICS, nr_ioapics);
-+ panic("Recompile kernel with bigger MAX_IO_APICS!\n");
-+ }
-+ if (!address) {
-+ printk(KERN_ERR "WARNING: Bogus (zero) I/O APIC address"
-+ " found in MADT table, skipping!\n");
-+ return;
-+ }
++ if (!(tmpbuf = kmalloc(TMPBUFSIZE, GFP_KERNEL)))
++ return -ENOMEM;
+
-+ idx = nr_ioapics++;
++ mutex_lock(&pdom_mutex);
+
-+ mp_ioapics[idx].mpc_type = MP_IOAPIC;
-+ mp_ioapics[idx].mpc_flags = MPC_APIC_USABLE;
-+ mp_ioapics[idx].mpc_apicaddr = address;
++ len = 0;
++ for (i = 0; i < pdomains; i++)
++ len += snprintf(tmpbuf + len,
++ len < TMPBUFSIZE ? TMPBUFSIZE - len : 0,
++ "%u ", passive_domains[i]);
++ WARN_ON(len > TMPBUFSIZE);
++ if (len != 0 && len <= TMPBUFSIZE)
++ tmpbuf[len-1] = '\n';
+
-+#ifndef CONFIG_XEN
-+ set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address);
-+#endif
-+ mp_ioapics[idx].mpc_apicid = id;
-+ mp_ioapics[idx].mpc_apicver = io_apic_get_version(idx);
-+
-+ /*
-+ * Build basic IRQ lookup table to facilitate gsi->io_apic lookups
-+ * and to prevent reprogramming of IOAPIC pins (PCI IRQs).
-+ */
-+ mp_ioapic_routing[idx].apic_id = mp_ioapics[idx].mpc_apicid;
-+ mp_ioapic_routing[idx].gsi_start = gsi_base;
-+ mp_ioapic_routing[idx].gsi_end = gsi_base +
-+ io_apic_get_redir_entries(idx);
++ mutex_unlock(&pdom_mutex);
+
-+ printk(KERN_INFO "IOAPIC[%d]: apic_id %d, version %d, address 0x%x, "
-+ "GSI %d-%d\n", idx, mp_ioapics[idx].mpc_apicid,
-+ mp_ioapics[idx].mpc_apicver, mp_ioapics[idx].mpc_apicaddr,
-+ mp_ioapic_routing[idx].gsi_start,
-+ mp_ioapic_routing[idx].gsi_end);
++ retval = simple_read_from_buffer(buf, count, offset, tmpbuf, len);
+
-+ return;
++ kfree(tmpbuf);
++ return retval;
+}
+
++static struct file_operations passive_domain_ops = {
++ .read = pdomain_read,
++ .write = pdomain_write,
++};
+
-+void __init mp_override_legacy_irq (
-+ u8 bus_irq,
-+ u8 polarity,
-+ u8 trigger,
-+ u32 gsi)
-+{
-+ struct mpc_config_intsrc intsrc;
-+ int ioapic = -1;
-+ int pin = -1;
-+
-+ /*
-+ * Convert 'gsi' to 'ioapic.pin'.
-+ */
-+ ioapic = mp_find_ioapic(gsi);
-+ if (ioapic < 0)
-+ return;
-+ pin = gsi - mp_ioapic_routing[ioapic].gsi_start;
-+
-+ /*
-+ * TBD: This check is for faulty timer entries, where the override
-+ * erroneously sets the trigger to level, resulting in a HUGE
-+ * increase of timer interrupts!
-+ */
-+ if ((bus_irq == 0) && (trigger == 3))
-+ trigger = 1;
+ void oprofile_create_files(struct super_block * sb, struct dentry * root)
+ {
+ oprofilefs_create_file(sb, root, "enable", &enable_fops);
+ oprofilefs_create_file_perm(sb, root, "dump", &dump_fops, 0666);
++ oprofilefs_create_file(sb, root, "active_domains", &active_domain_ops);
++ oprofilefs_create_file(sb, root, "passive_domains", &passive_domain_ops);
+ oprofilefs_create_file(sb, root, "buffer", &event_buffer_fops);
+ oprofilefs_create_ulong(sb, root, "buffer_size", &fs_buffer_size);
+ oprofilefs_create_ulong(sb, root, "buffer_watershed", &fs_buffer_watershed);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/pci/bus.c
+--- a/drivers/pci/bus.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/pci/bus.c Wed Sep 10 10:54:08 2008 +0100
+@@ -16,6 +16,8 @@
+ #include <linux/init.h>
+
+ #include "pci.h"
+
-+ intsrc.mpc_type = MP_INTSRC;
-+ intsrc.mpc_irqtype = mp_INT;
-+ intsrc.mpc_irqflag = (trigger << 2) | polarity;
-+ intsrc.mpc_srcbus = MP_ISA_BUS;
-+ intsrc.mpc_srcbusirq = bus_irq; /* IRQ */
-+ intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid; /* APIC ID */
-+ intsrc.mpc_dstirq = pin; /* INTIN# */
++extern int pci_mem_align;
+
+ /**
+ * pci_bus_alloc_resource - allocate a resource from a parent bus
+@@ -43,6 +45,11 @@
+ int i, ret = -ENOMEM;
+
+ type_mask |= IORESOURCE_IO | IORESOURCE_MEM;
++
++ /* If the boot parameter 'pci-mem-align' was specified then we need to
++ align the memory addresses, at page size alignment. */
++ if (pci_mem_align && (align < (PAGE_SIZE-1)))
++ align = PAGE_SIZE - 1;
+
+ for (i = 0; i < PCI_BUS_NUM_RESOURCES; i++) {
+ struct resource *r = bus->resource[i];
+diff -r d894e36cfc30 -r 0aa021803deb drivers/pci/msi-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/pci/msi-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,760 @@
++/*
++ * File: msi.c
++ * Purpose: PCI Message Signaled Interrupt (MSI)
++ *
++ * Copyright (C) 2003-2004 Intel
++ * Copyright (C) Tom Long Nguyen (tom.l.nguyen at intel.com)
++ */
+
-+ Dprintk("Int: type %d, pol %d, trig %d, bus %d, irq %d, %d-%d\n",
-+ intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3,
-+ (intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus,
-+ intsrc.mpc_srcbusirq, intsrc.mpc_dstapic, intsrc.mpc_dstirq);
++#include <linux/mm.h>
++#include <linux/irq.h>
++#include <linux/interrupt.h>
++#include <linux/init.h>
++#include <linux/ioport.h>
++#include <linux/smp_lock.h>
++#include <linux/pci.h>
++#include <linux/proc_fs.h>
+
-+ mp_irqs[mp_irq_entries] = intsrc;
-+ if (++mp_irq_entries == MAX_IRQ_SOURCES)
-+ panic("Max # of irq sources exceeded!\n");
++#include <xen/evtchn.h>
+
-+ return;
-+}
++#include <asm/errno.h>
++#include <asm/io.h>
++#include <asm/smp.h>
+
++#include "pci.h"
++#include "msi.h"
+
-+void __init mp_config_acpi_legacy_irqs (void)
-+{
-+ struct mpc_config_intsrc intsrc;
-+ int i = 0;
-+ int ioapic = -1;
++static int pci_msi_enable = 1;
+
-+ /*
-+ * Fabricate the legacy ISA bus (bus #31).
-+ */
-+ mp_bus_id_to_type[MP_ISA_BUS] = MP_BUS_ISA;
-+ Dprintk("Bus #%d is ISA\n", MP_ISA_BUS);
++static struct msi_ops *msi_ops;
+
-+ /*
-+ * Locate the IOAPIC that manages the ISA IRQs (0-15).
-+ */
-+ ioapic = mp_find_ioapic(0);
-+ if (ioapic < 0)
-+ return;
++int msi_register(struct msi_ops *ops)
++{
++ msi_ops = ops;
++ return 0;
++}
+
-+ intsrc.mpc_type = MP_INTSRC;
-+ intsrc.mpc_irqflag = 0; /* Conforming */
-+ intsrc.mpc_srcbus = MP_ISA_BUS;
-+ intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid;
++static LIST_HEAD(msi_dev_head);
++DEFINE_SPINLOCK(msi_dev_lock);
+
-+ /*
-+ * Use the default configuration for the IRQs 0-15. Unless
-+ * overridden by (MADT) interrupt source override entries.
-+ */
-+ for (i = 0; i < 16; i++) {
-+ int idx;
++struct msi_dev_list {
++ struct pci_dev *dev;
++ struct list_head list;
++ spinlock_t pirq_list_lock;
++ struct list_head pirq_list_head;
++};
+
-+ for (idx = 0; idx < mp_irq_entries; idx++) {
-+ struct mpc_config_intsrc *irq = mp_irqs + idx;
++struct msi_pirq_entry {
++ struct list_head list;
++ int pirq;
++ int entry_nr;
++};
+
-+ /* Do we already have a mapping for this ISA IRQ? */
-+ if (irq->mpc_srcbus == MP_ISA_BUS && irq->mpc_srcbusirq == i)
-+ break;
++static struct msi_dev_list *get_msi_dev_pirq_list(struct pci_dev *dev)
++{
++ struct msi_dev_list *msi_dev_list, *ret = NULL;
++ unsigned long flags;
+
-+ /* Do we already have a mapping for this IOAPIC pin */
-+ if ((irq->mpc_dstapic == intsrc.mpc_dstapic) &&
-+ (irq->mpc_dstirq == i))
-+ break;
-+ }
++ spin_lock_irqsave(&msi_dev_lock, flags);
+
-+ if (idx != mp_irq_entries) {
-+ printk(KERN_DEBUG "ACPI: IRQ%d used by override.\n", i);
-+ continue; /* IRQ already used */
-+ }
++ list_for_each_entry(msi_dev_list, &msi_dev_head, list)
++ if ( msi_dev_list->dev == dev )
++ ret = msi_dev_list;
+
-+ intsrc.mpc_irqtype = mp_INT;
-+ intsrc.mpc_srcbusirq = i; /* Identity mapped */
-+ intsrc.mpc_dstirq = i;
++ if ( ret ) {
++ spin_unlock_irqrestore(&msi_dev_lock, flags);
++ return ret;
++ }
+
-+ Dprintk("Int: type %d, pol %d, trig %d, bus %d, irq %d, "
-+ "%d-%d\n", intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3,
-+ (intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus,
-+ intsrc.mpc_srcbusirq, intsrc.mpc_dstapic,
-+ intsrc.mpc_dstirq);
++ /* Has not allocate msi_dev until now. */
++ ret = kmalloc(sizeof(struct msi_dev_list), GFP_ATOMIC);
+
-+ mp_irqs[mp_irq_entries] = intsrc;
-+ if (++mp_irq_entries == MAX_IRQ_SOURCES)
-+ panic("Max # of irq sources exceeded!\n");
++ /* Failed to allocate msi_dev structure */
++ if ( !ret ) {
++ spin_unlock_irqrestore(&msi_dev_lock, flags);
++ return NULL;
+ }
+
-+ return;
++ spin_lock_init(&ret->pirq_list_lock);
++ INIT_LIST_HEAD(&ret->pirq_list_head);
++ list_add_tail(&ret->list, &msi_dev_head);
++ spin_unlock_irqrestore(&msi_dev_lock, flags);
++ return ret;
+}
+
-+#define MAX_GSI_NUM 4096
-+
-+int mp_register_gsi(u32 gsi, int triggering, int polarity)
++static int attach_pirq_entry(int pirq, int entry_nr,
++ struct msi_dev_list *msi_dev_entry)
+{
-+ int ioapic = -1;
-+ int ioapic_pin = 0;
-+ int idx, bit = 0;
-+ static int pci_irq = 16;
-+ /*
-+ * Mapping between Global System Interrupts, which
-+ * represent all possible interrupts, to the IRQs
-+ * assigned to actual devices.
-+ */
-+ static int gsi_to_irq[MAX_GSI_NUM];
++ struct msi_pirq_entry *entry = kmalloc(sizeof(*entry), GFP_ATOMIC);
++ unsigned long flags;
+
-+ if (acpi_irq_model != ACPI_IRQ_MODEL_IOAPIC)
-+ return gsi;
++ if (!entry)
++ return -ENOMEM;
++ entry->pirq = pirq;
++ entry->entry_nr = entry_nr;
++ spin_lock_irqsave(&msi_dev_entry->pirq_list_lock, flags);
++ list_add_tail(&entry->list, &msi_dev_entry->pirq_list_head);
++ spin_unlock_irqrestore(&msi_dev_entry->pirq_list_lock, flags);
++ return 0;
++}
+
-+ /* Don't set up the ACPI SCI because it's already set up */
-+ if (acpi_fadt.sci_int == gsi)
-+ return gsi;
++static void detach_pirq_entry(int entry_nr,
++ struct msi_dev_list *msi_dev_entry)
++{
++ unsigned long flags;
++ struct msi_pirq_entry *pirq_entry;
+
-+ ioapic = mp_find_ioapic(gsi);
-+ if (ioapic < 0) {
-+ printk(KERN_WARNING "No IOAPIC for GSI %u\n", gsi);
-+ return gsi;
++ list_for_each_entry(pirq_entry, &msi_dev_entry->pirq_list_head, list) {
++ if (pirq_entry->entry_nr == entry_nr) {
++ spin_lock_irqsave(&msi_dev_entry->pirq_list_lock, flags);
++ list_del(&pirq_entry->list);
++ spin_unlock_irqrestore(&msi_dev_entry->pirq_list_lock, flags);
++ kfree(pirq_entry);
++ return;
++ }
+ }
++}
+
-+ ioapic_pin = gsi - mp_ioapic_routing[ioapic].gsi_start;
++/*
++ * pciback will provide device's owner
++ */
++static int (*get_owner)(struct pci_dev *dev);
+
-+ /*
-+ * Avoid pin reprogramming. PRTs typically include entries
-+ * with redundant pin->gsi mappings (but unique PCI devices);
-+ * we only program the IOAPIC on the first.
-+ */
-+ bit = ioapic_pin % 32;
-+ idx = (ioapic_pin < 32) ? 0 : (ioapic_pin / 32);
-+ if (idx > 3) {
-+ printk(KERN_ERR "Invalid reference to IOAPIC pin "
-+ "%d-%d\n", mp_ioapic_routing[ioapic].apic_id,
-+ ioapic_pin);
-+ return gsi;
-+ }
-+ if ((1<<bit) & mp_ioapic_routing[ioapic].pin_programmed[idx]) {
-+ Dprintk(KERN_DEBUG "Pin %d-%d already programmed\n",
-+ mp_ioapic_routing[ioapic].apic_id, ioapic_pin);
-+ return gsi_to_irq[gsi];
++int register_msi_get_owner(int (*func)(struct pci_dev *dev))
++{
++ if (get_owner) {
++ printk(KERN_WARNING "register msi_get_owner again\n");
++ return -EEXIST;
+ }
++ get_owner = func;
++ return 0;
++}
+
-+ mp_ioapic_routing[ioapic].pin_programmed[idx] |= (1<<bit);
++int unregister_msi_get_owner(int (*func)(struct pci_dev *dev))
++{
++ if (get_owner != func)
++ return -EINVAL;
++ get_owner = NULL;
++ return 0;
++}
+
-+ if (triggering == ACPI_LEVEL_SENSITIVE) {
-+ /*
-+ * For PCI devices assign IRQs in order, avoiding gaps
-+ * due to unused I/O APIC pins.
-+ */
-+ int irq = gsi;
-+ if (gsi < MAX_GSI_NUM) {
-+ /*
-+ * Retain the VIA chipset work-around (gsi > 15), but
-+ * avoid a problem where the 8254 timer (IRQ0) is setup
-+ * via an override (so it's not on pin 0 of the ioapic),
-+ * and at the same time, the pin 0 interrupt is a PCI
-+ * type. The gsi > 15 test could cause these two pins
-+ * to be shared as IRQ0, and they are not shareable.
-+ * So test for this condition, and if necessary, avoid
-+ * the pin collision.
-+ */
-+ if (gsi > 15 || (gsi == 0 && !timer_uses_ioapic_pin_0))
-+ gsi = pci_irq++;
-+ /*
-+ * Don't assign IRQ used by ACPI SCI
-+ */
-+ if (gsi == acpi_fadt.sci_int)
-+ gsi = pci_irq++;
-+ gsi_to_irq[irq] = gsi;
-+ } else {
-+ printk(KERN_ERR "GSI %u is too high\n", gsi);
-+ return gsi;
-+ }
++static int msi_get_dev_owner(struct pci_dev *dev)
++{
++ int owner;
++
++ BUG_ON(!is_initial_xendomain());
++ if (get_owner && (owner = get_owner(dev)) >= 0) {
++ printk(KERN_INFO "get owner for dev %x get %x \n",
++ dev->devfn, owner);
++ return owner;
+ }
+
-+ io_apic_set_pci_routing(ioapic, ioapic_pin, gsi,
-+ triggering == ACPI_EDGE_SENSITIVE ? 0 : 1,
-+ polarity == ACPI_ACTIVE_HIGH ? 0 : 1);
-+ return gsi;
++ return DOMID_SELF;
+}
+
-+#endif /*CONFIG_X86_IO_APIC*/
-+#endif /*CONFIG_ACPI*/
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/pci-swiotlb-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/pci-swiotlb-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/pci-swiotlb-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/pci-swiotlb-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,55 @@
-+/* Glue code to lib/swiotlb.c */
++static int msi_unmap_pirq(struct pci_dev *dev, int pirq)
++{
++ struct physdev_unmap_pirq unmap;
++ int rc;
+
-+#include <linux/pci.h>
-+#include <linux/cache.h>
-+#include <linux/module.h>
-+#include <asm/dma-mapping.h>
-+#include <asm/proto.h>
-+#include <asm/swiotlb.h>
-+#include <asm/dma.h>
++ unmap.domid = msi_get_dev_owner(dev);
++ unmap.pirq = evtchn_get_xen_pirq(pirq);
+
-+#if 0
-+int swiotlb __read_mostly;
-+EXPORT_SYMBOL(swiotlb);
-+#endif
++ if ((rc = HYPERVISOR_physdev_op(PHYSDEVOP_unmap_pirq, &unmap)))
++ printk(KERN_WARNING "unmap irq %x failed\n", pirq);
+
-+struct dma_mapping_ops swiotlb_dma_ops = {
-+#if 0
-+ .mapping_error = swiotlb_dma_mapping_error,
-+ .alloc_coherent = swiotlb_alloc_coherent,
-+ .free_coherent = swiotlb_free_coherent,
-+ .map_single = swiotlb_map_single,
-+ .unmap_single = swiotlb_unmap_single,
-+ .sync_single_for_cpu = swiotlb_sync_single_for_cpu,
-+ .sync_single_for_device = swiotlb_sync_single_for_device,
-+ .sync_single_range_for_cpu = swiotlb_sync_single_range_for_cpu,
-+ .sync_single_range_for_device = swiotlb_sync_single_range_for_device,
-+ .sync_sg_for_cpu = swiotlb_sync_sg_for_cpu,
-+ .sync_sg_for_device = swiotlb_sync_sg_for_device,
-+ .map_sg = swiotlb_map_sg,
-+ .unmap_sg = swiotlb_unmap_sg,
-+ .dma_supported = NULL,
-+#endif
-+};
++ if (rc < 0)
++ return rc;
+
-+void pci_swiotlb_init(void)
++ evtchn_map_pirq(pirq, 0);
++ return 0;
++}
++
++static u64 find_table_base(struct pci_dev *dev, int pos)
+{
-+#if 0
-+ /* don't initialize swiotlb if iommu=off (no_iommu=1) */
-+ if (!iommu_detected && !no_iommu && end_pfn > MAX_DMA32_PFN)
-+ swiotlb = 1;
-+ if (swiotlb_force)
-+ swiotlb = 1;
-+ if (swiotlb) {
-+ printk(KERN_INFO "PCI-DMA: Using software bounce buffering for IO (SWIOTLB)\n");
-+ swiotlb_init();
-+ dma_ops = &swiotlb_dma_ops;
-+ }
-+#else
-+ swiotlb_init();
-+ if (swiotlb) {
-+ printk(KERN_INFO "PCI-DMA: Using software bounce buffering for IO (SWIOTLB)\n");
-+ dma_ops = &swiotlb_dma_ops;
-+ }
-+#endif
++ u8 bar;
++ u32 reg;
++ unsigned long flags;
++
++ pci_read_config_dword(dev, msix_table_offset_reg(pos), ®);
++ bar = reg & PCI_MSIX_FLAGS_BIRMASK;
++
++ flags = pci_resource_flags(dev, bar);
++ if (flags & (IORESOURCE_DISABLED | IORESOURCE_UNSET | IORESOURCE_BUSY))
++ return 0;
++
++ return pci_resource_start(dev, bar);
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/process-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/process-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/process-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/process-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,829 @@
-+/*
-+ * linux/arch/x86-64/kernel/process.c
-+ *
-+ * Copyright (C) 1995 Linus Torvalds
-+ *
-+ * Pentium III FXSR, SSE support
-+ * Gareth Hughes <gareth at valinux.com>, May 2000
-+ *
-+ * X86-64 port
-+ * Andi Kleen.
-+ *
-+ * CPU hotplug support - ashok.raj at intel.com
-+ *
-+ * Jun Nakajima <jun.nakajima at intel.com>
-+ * Modified for Xen
-+ */
+
+/*
-+ * This file handles the architecture-dependent parts of process handling..
++ * Protected by msi_lock
+ */
++static int msi_map_pirq_to_vector(struct pci_dev *dev, int pirq,
++ int entry_nr, u64 table_base)
++{
++ struct physdev_map_pirq map_irq;
++ int rc;
++ domid_t domid = DOMID_SELF;
+
-+#include <stdarg.h>
++ domid = msi_get_dev_owner(dev);
+
-+#include <linux/cpu.h>
-+#include <linux/errno.h>
-+#include <linux/sched.h>
-+#include <linux/kernel.h>
-+#include <linux/mm.h>
-+#include <linux/elfcore.h>
-+#include <linux/smp.h>
-+#include <linux/slab.h>
-+#include <linux/user.h>
-+#include <linux/module.h>
-+#include <linux/a.out.h>
-+#include <linux/interrupt.h>
-+#include <linux/delay.h>
-+#include <linux/ptrace.h>
-+#include <linux/utsname.h>
-+#include <linux/random.h>
-+#include <linux/notifier.h>
-+#include <linux/kprobes.h>
++ map_irq.domid = domid;
++ map_irq.type = MAP_PIRQ_TYPE_MSI;
++ map_irq.index = -1;
++ map_irq.pirq = pirq < 0 ? -1 : evtchn_get_xen_pirq(pirq);
++ map_irq.bus = dev->bus->number;
++ map_irq.devfn = dev->devfn;
++ map_irq.entry_nr = entry_nr;
++ map_irq.table_base = table_base;
+
-+#include <asm/uaccess.h>
-+#include <asm/pgtable.h>
-+#include <asm/system.h>
-+#include <asm/io.h>
-+#include <asm/processor.h>
-+#include <asm/i387.h>
-+#include <asm/mmu_context.h>
-+#include <asm/pda.h>
-+#include <asm/prctl.h>
-+#include <asm/kdebug.h>
-+#include <xen/interface/platform.h>
-+#include <xen/interface/physdev.h>
-+#include <xen/interface/vcpu.h>
-+#include <asm/desc.h>
-+#include <asm/proto.h>
-+#include <asm/hardirq.h>
-+#include <asm/ia32.h>
-+#include <asm/idle.h>
++ if ((rc = HYPERVISOR_physdev_op(PHYSDEVOP_map_pirq, &map_irq)))
++ printk(KERN_WARNING "map irq failed\n");
+
-+#include <xen/cpu_hotplug.h>
++ if (rc < 0)
++ return rc;
++ /* This happens when MSI support is not enabled in Xen. */
++ if (rc == 0 && map_irq.pirq < 0)
++ return -ENOSYS;
+
-+asmlinkage extern void ret_from_fork(void);
++ BUG_ON(map_irq.pirq <= 0);
++ return evtchn_map_pirq(pirq, map_irq.pirq);
++}
+
-+unsigned long kernel_thread_flags = CLONE_VM | CLONE_UNTRACED;
++static int msi_map_vector(struct pci_dev *dev, int entry_nr, u64 table_base)
++{
++ return msi_map_pirq_to_vector(dev, -1, entry_nr, table_base);
++}
+
-+unsigned long boot_option_idle_override = 0;
-+EXPORT_SYMBOL(boot_option_idle_override);
++static int msi_init(void)
++{
++ static int status = 0;
+
-+/*
-+ * Powermanagement idle function, if any..
-+ */
-+void (*pm_idle)(void);
-+EXPORT_SYMBOL(pm_idle);
-+static DEFINE_PER_CPU(unsigned int, cpu_idle_state);
++ if (pci_msi_quirk) {
++ pci_msi_enable = 0;
++ printk(KERN_WARNING "PCI: MSI quirk detected. MSI disabled.\n");
++ status = -EINVAL;
++ }
+
-+static ATOMIC_NOTIFIER_HEAD(idle_notifier);
++ return status;
++}
+
-+void idle_notifier_register(struct notifier_block *n)
++void pci_scan_msi_device(struct pci_dev *dev) { }
++
++void disable_msi_mode(struct pci_dev *dev, int pos, int type)
+{
-+ atomic_notifier_chain_register(&idle_notifier, n);
++ u16 control;
++
++ pci_read_config_word(dev, msi_control_reg(pos), &control);
++ if (type == PCI_CAP_ID_MSI) {
++ /* Set enabled bits to single MSI & enable MSI_enable bit */
++ msi_disable(control);
++ pci_write_config_word(dev, msi_control_reg(pos), control);
++ dev->msi_enabled = 0;
++ } else {
++ msix_disable(control);
++ pci_write_config_word(dev, msi_control_reg(pos), control);
++ dev->msix_enabled = 0;
++ }
++ if (pci_find_capability(dev, PCI_CAP_ID_EXP)) {
++ /* PCI Express Endpoint device detected */
++ pci_intx(dev, 1); /* enable intx */
++ }
+}
-+EXPORT_SYMBOL_GPL(idle_notifier_register);
+
-+void idle_notifier_unregister(struct notifier_block *n)
++static void enable_msi_mode(struct pci_dev *dev, int pos, int type)
+{
-+ atomic_notifier_chain_unregister(&idle_notifier, n);
++ u16 control;
++
++ pci_read_config_word(dev, msi_control_reg(pos), &control);
++ if (type == PCI_CAP_ID_MSI) {
++ /* Set enabled bits to single MSI & enable MSI_enable bit */
++ msi_enable(control, 1);
++ pci_write_config_word(dev, msi_control_reg(pos), control);
++ dev->msi_enabled = 1;
++ } else {
++ msix_enable(control);
++ pci_write_config_word(dev, msi_control_reg(pos), control);
++ dev->msix_enabled = 1;
++ }
++ if (pci_find_capability(dev, PCI_CAP_ID_EXP)) {
++ /* PCI Express Endpoint device detected */
++ pci_intx(dev, 0); /* disable intx */
++ }
+}
-+EXPORT_SYMBOL(idle_notifier_unregister);
+
-+enum idle_state { CPU_IDLE, CPU_NOT_IDLE };
-+static DEFINE_PER_CPU(enum idle_state, idle_state) = CPU_NOT_IDLE;
++#ifdef CONFIG_PM
++int pci_save_msi_state(struct pci_dev *dev)
++{
++ int pos;
+
-+void enter_idle(void)
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
++ if (pos <= 0 || dev->no_msi)
++ return 0;
++
++ if (!dev->msi_enabled)
++ return 0;
++
++ /* Restore dev->irq to its default pin-assertion vector */
++ msi_unmap_pirq(dev, dev->irq);
++ /* Disable MSI mode */
++ disable_msi_mode(dev, pos, PCI_CAP_ID_MSI);
++ /* Set the flags for use of restore */
++ dev->msi_enabled = 1;
++ return 0;
++}
++
++void pci_restore_msi_state(struct pci_dev *dev)
+{
-+ __get_cpu_var(idle_state) = CPU_IDLE;
-+ atomic_notifier_call_chain(&idle_notifier, IDLE_START, NULL);
++ int pos, pirq;
++
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
++ if (pos <= 0)
++ return;
++
++ if (!dev->msi_enabled)
++ return;
++
++ pirq = msi_map_pirq_to_vector(dev, dev->irq, 0, 0);
++ if (pirq < 0)
++ return;
++ enable_msi_mode(dev, pos, PCI_CAP_ID_MSI);
+}
+
-+static void __exit_idle(void)
++int pci_save_msix_state(struct pci_dev *dev)
+{
-+ __get_cpu_var(idle_state) = CPU_NOT_IDLE;
-+ atomic_notifier_call_chain(&idle_notifier, IDLE_END, NULL);
++ int pos;
++ unsigned long flags;
++ struct msi_dev_list *msi_dev_entry;
++ struct msi_pirq_entry *pirq_entry, *tmp;
++
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
++ if (pos <= 0 || dev->no_msi)
++ return 0;
++
++ /* save the capability */
++ if (!dev->msix_enabled)
++ return 0;
++
++ msi_dev_entry = get_msi_dev_pirq_list(dev);
++
++ spin_lock_irqsave(&msi_dev_entry->pirq_list_lock, flags);
++ list_for_each_entry_safe(pirq_entry, tmp,
++ &msi_dev_entry->pirq_list_head, list)
++ msi_unmap_pirq(dev, pirq_entry->pirq);
++ spin_unlock_irqrestore(&msi_dev_entry->pirq_list_lock, flags);
++
++ disable_msi_mode(dev, pos, PCI_CAP_ID_MSIX);
++ /* Set the flags for use of restore */
++ dev->msix_enabled = 1;
++
++ return 0;
+}
+
-+/* Called from interrupts to signify idle end */
-+void exit_idle(void)
++void pci_restore_msix_state(struct pci_dev *dev)
+{
-+ if (current->pid | read_pda(irqcount))
++ int pos;
++ unsigned long flags;
++ u64 table_base;
++ struct msi_dev_list *msi_dev_entry;
++ struct msi_pirq_entry *pirq_entry, *tmp;
++
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
++ if (pos <= 0)
+ return;
-+ __exit_idle();
++
++ if (!dev->msix_enabled)
++ return;
++
++ msi_dev_entry = get_msi_dev_pirq_list(dev);
++ table_base = find_table_base(dev, pos);
++ if (!table_base)
++ return;
++
++ spin_lock_irqsave(&msi_dev_entry->pirq_list_lock, flags);
++ list_for_each_entry_safe(pirq_entry, tmp,
++ &msi_dev_entry->pirq_list_head, list) {
++ int rc = msi_map_pirq_to_vector(dev, pirq_entry->pirq,
++ pirq_entry->entry_nr, table_base);
++ if (rc < 0)
++ printk(KERN_WARNING
++ "%s: re-mapping irq #%d (pirq%d) failed: %d\n",
++ pci_name(dev), pirq_entry->entry_nr,
++ pirq_entry->pirq, rc);
++ }
++ spin_unlock_irqrestore(&msi_dev_entry->pirq_list_lock, flags);
++
++ enable_msi_mode(dev, pos, PCI_CAP_ID_MSIX);
+}
++#endif
+
-+/*
-+ * On SMP it's slightly faster (but much more power-consuming!)
-+ * to poll the ->need_resched flag instead of waiting for the
-+ * cross-CPU IPI to arrive. Use this option with caution.
-+ */
-+static void poll_idle (void)
++/**
++ * msi_capability_init - configure device's MSI capability structure
++ * @dev: pointer to the pci_dev data structure of MSI device function
++ *
++ * Setup the MSI capability structure of device function with a single
++ * MSI vector, regardless of device function is capable of handling
++ * multiple messages. A return of zero indicates the successful setup
++ * of an entry zero with the new MSI vector or non-zero for otherwise.
++ **/
++static int msi_capability_init(struct pci_dev *dev)
+{
-+ local_irq_enable();
++ int pos, pirq;
++ u16 control;
+
-+ asm volatile(
-+ "2:"
-+ "testl %0,%1;"
-+ "rep; nop;"
-+ "je 2b;"
-+ : :
-+ "i" (_TIF_NEED_RESCHED),
-+ "m" (current_thread_info()->flags));
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
++ pci_read_config_word(dev, msi_control_reg(pos), &control);
++
++ pirq = msi_map_vector(dev, 0, 0);
++ if (pirq < 0)
++ return -EBUSY;
++
++ dev->irq = pirq;
++ /* Set MSI enabled bits */
++ enable_msi_mode(dev, pos, PCI_CAP_ID_MSI);
++ dev->msi_enabled = 1;
++
++ return 0;
+}
+
-+static void xen_idle(void)
++/**
++ * msix_capability_init - configure device's MSI-X capability
++ * @dev: pointer to the pci_dev data structure of MSI-X device function
++ * @entries: pointer to an array of struct msix_entry entries
++ * @nvec: number of @entries
++ *
++ * Setup the MSI-X capability structure of device function with a
++ * single MSI-X vector. A return of zero indicates the successful setup of
++ * requested MSI-X entries with allocated vectors or non-zero for otherwise.
++ **/
++static int msix_capability_init(struct pci_dev *dev,
++ struct msix_entry *entries, int nvec)
+{
-+ local_irq_disable();
++ u64 table_base;
++ int pirq, i, j, mapped, pos;
++ struct msi_dev_list *msi_dev_entry = get_msi_dev_pirq_list(dev);
++ struct msi_pirq_entry *pirq_entry;
+
-+ if (need_resched())
-+ local_irq_enable();
-+ else {
-+ current_thread_info()->status &= ~TS_POLLING;
-+ smp_mb__after_clear_bit();
-+ safe_halt();
-+ current_thread_info()->status |= TS_POLLING;
++ if (!msi_dev_entry)
++ return -ENOMEM;
++
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
++ table_base = find_table_base(dev, pos);
++ if (!table_base)
++ return -ENODEV;
++
++ /* MSI-X Table Initialization */
++ for (i = 0; i < nvec; i++) {
++ mapped = 0;
++ list_for_each_entry(pirq_entry, &msi_dev_entry->pirq_list_head, list) {
++ if (pirq_entry->entry_nr == entries[i].entry) {
++ printk(KERN_WARNING "msix entry %d for dev %02x:%02x:%01x are \
++ not freed before acquire again.\n", entries[i].entry,
++ dev->bus->number, PCI_SLOT(dev->devfn),
++ PCI_FUNC(dev->devfn));
++ (entries + i)->vector = pirq_entry->pirq;
++ mapped = 1;
++ break;
++ }
++ }
++ if (mapped)
++ continue;
++ pirq = msi_map_vector(dev, entries[i].entry, table_base);
++ if (pirq < 0)
++ break;
++ attach_pirq_entry(pirq, entries[i].entry, msi_dev_entry);
++ (entries + i)->vector = pirq;
++ }
++
++ if (i != nvec) {
++ for (j = --i; j >= 0; j--) {
++ msi_unmap_pirq(dev, entries[j].vector);
++ detach_pirq_entry(entries[j].entry, msi_dev_entry);
++ entries[j].vector = 0;
++ }
++ return -EBUSY;
+ }
++
++ enable_msi_mode(dev, pos, PCI_CAP_ID_MSIX);
++ dev->msix_enabled = 1;
++
++ return 0;
+}
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+static inline void play_dead(void)
++/**
++ * pci_enable_msi - configure device's MSI capability structure
++ * @dev: pointer to the pci_dev data structure of MSI device function
++ *
++ * Setup the MSI capability structure of device function with
++ * a single MSI vector upon its software driver call to request for
++ * MSI mode enabled on its hardware device function. A return of zero
++ * indicates the successful setup of an entry zero with the new MSI
++ * vector or non-zero for otherwise.
++ **/
++extern int pci_frontend_enable_msi(struct pci_dev *dev);
++int pci_enable_msi(struct pci_dev* dev)
+{
-+ idle_task_exit();
-+ local_irq_disable();
-+ cpu_clear(smp_processor_id(), cpu_initialized);
-+ preempt_enable_no_resched();
-+ HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL);
-+ cpu_bringup();
++ struct pci_bus *bus;
++ int pos, temp, status = -EINVAL;
++
++ if (!pci_msi_enable || !dev)
++ return status;
++
++ if (dev->no_msi)
++ return status;
++
++ for (bus = dev->bus; bus; bus = bus->parent)
++ if (bus->bus_flags & PCI_BUS_FLAGS_NO_MSI)
++ return -EINVAL;
++
++ status = msi_init();
++ if (status < 0)
++ return status;
++
++#ifdef CONFIG_XEN_PCIDEV_FRONTEND
++ if (!is_initial_xendomain())
++ {
++ int ret;
++
++ temp = dev->irq;
++ ret = pci_frontend_enable_msi(dev);
++ if (ret)
++ return ret;
++
++ dev->irq_old = temp;
++
++ return ret;
++ }
++#endif
++
++ temp = dev->irq;
++
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
++ if (!pos)
++ return -EINVAL;
++
++ /* Check whether driver already requested for MSI-X vectors */
++ if (dev->msix_enabled) {
++ printk(KERN_INFO "PCI: %s: Can't enable MSI. "
++ "Device already has MSI-X vectors assigned\n",
++ pci_name(dev));
++ dev->irq = temp;
++ return -EINVAL;
++ }
++
++ status = msi_capability_init(dev);
++ if ( !status )
++ dev->irq_old = temp;
++ else
++ dev->irq = temp;
++
++ return status;
+}
-+#else
-+static inline void play_dead(void)
++
++extern void pci_frontend_disable_msi(struct pci_dev* dev);
++void pci_disable_msi(struct pci_dev* dev)
+{
-+ BUG();
++ int pos;
++ int pirq;
++
++ if (!pci_msi_enable)
++ return;
++ if (!dev)
++ return;
++
++#ifdef CONFIG_XEN_PCIDEV_FRONTEND
++ if (!is_initial_xendomain()) {
++ pci_frontend_disable_msi(dev);
++ dev->irq = dev->irq_old;
++ return;
++ }
++#endif
++
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
++ if (!pos)
++ return;
++
++ pirq = dev->irq;
++ /* Restore dev->irq to its default pin-assertion vector */
++ dev->irq = dev->irq_old;
++ msi_unmap_pirq(dev, pirq);
++
++ /* Disable MSI mode */
++ disable_msi_mode(dev, pos, PCI_CAP_ID_MSI);
+}
-+#endif /* CONFIG_HOTPLUG_CPU */
+
-+/*
-+ * The idle thread. There's no useful work to be
-+ * done, so just try to conserve power and have a
-+ * low exit latency (ie sit in a loop waiting for
-+ * somebody to say that they'd like to reschedule)
-+ */
-+void cpu_idle (void)
++/**
++ * pci_enable_msix - configure device's MSI-X capability structure
++ * @dev: pointer to the pci_dev data structure of MSI-X device function
++ * @entries: pointer to an array of MSI-X entries
++ * @nvec: number of MSI-X vectors requested for allocation by device driver
++ *
++ * Setup the MSI-X capability structure of device function with the number
++ * of requested vectors upon its software driver call to request for
++ * MSI-X mode enabled on its hardware device function. A return of zero
++ * indicates the successful configuration of MSI-X capability structure
++ * with new allocated MSI-X vectors. A return of < 0 indicates a failure.
++ * Or a return of > 0 indicates that driver request is exceeding the number
++ * of vectors available. Driver should use the returned value to re-send
++ * its request.
++ **/
++extern int pci_frontend_enable_msix(struct pci_dev *dev,
++ struct msix_entry *entries, int nvec);
++int pci_enable_msix(struct pci_dev* dev, struct msix_entry *entries, int nvec)
+{
-+ current_thread_info()->status |= TS_POLLING;
-+ /* endless idle loop with no priority at all */
-+ while (1) {
-+ while (!need_resched()) {
-+ void (*idle)(void);
++ struct pci_bus *bus;
++ int status, pos, nr_entries;
++ int i, j, temp;
++ u16 control;
+
-+ if (__get_cpu_var(cpu_idle_state))
-+ __get_cpu_var(cpu_idle_state) = 0;
-+ rmb();
-+ idle = xen_idle; /* no alternatives */
-+ if (cpu_is_offline(smp_processor_id()))
-+ play_dead();
-+ enter_idle();
-+ idle();
-+ __exit_idle();
++ if (!pci_msi_enable || !dev || !entries)
++ return -EINVAL;
++
++ if (dev->no_msi)
++ return -EINVAL;
++
++ for (bus = dev->bus; bus; bus = bus->parent)
++ if (bus->bus_flags & PCI_BUS_FLAGS_NO_MSI)
++ return -EINVAL;
++
++#ifdef CONFIG_XEN_PCIDEV_FRONTEND
++ if (!is_initial_xendomain()) {
++ int ret;
++
++ ret = pci_frontend_enable_msix(dev, entries, nvec);
++ if (ret) {
++ printk("get %x from pci_frontend_enable_msix\n", ret);
++ return ret;
+ }
+
-+ preempt_enable_no_resched();
-+ schedule();
-+ preempt_disable();
++ return 0;
+ }
-+}
++#endif
+
-+void cpu_idle_wait(void)
-+{
-+ unsigned int cpu, this_cpu = get_cpu();
-+ cpumask_t map;
++ status = msi_init();
++ if (status < 0)
++ return status;
++
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
++ if (!pos)
++ return -EINVAL;
++
++ pci_read_config_word(dev, msi_control_reg(pos), &control);
++ nr_entries = multi_msix_capable(control);
++ if (nvec > nr_entries)
++ return -EINVAL;
+
-+ set_cpus_allowed(current, cpumask_of_cpu(this_cpu));
-+ put_cpu();
++ /* Check for any invalid entries */
++ for (i = 0; i < nvec; i++) {
++ if (entries[i].entry >= nr_entries)
++ return -EINVAL; /* invalid entry */
++ for (j = i + 1; j < nvec; j++) {
++ if (entries[i].entry == entries[j].entry)
++ return -EINVAL; /* duplicate entry */
++ }
++ }
+
-+ cpus_clear(map);
-+ for_each_online_cpu(cpu) {
-+ per_cpu(cpu_idle_state, cpu) = 1;
-+ cpu_set(cpu, map);
++ temp = dev->irq;
++ /* Check whether driver already requested for MSI vector */
++ if (dev->msi_enabled) {
++ printk(KERN_INFO "PCI: %s: Can't enable MSI-X. "
++ "Device already has an MSI vector assigned\n",
++ pci_name(dev));
++ dev->irq = temp;
++ return -EINVAL;
+ }
+
-+ __get_cpu_var(cpu_idle_state) = 0;
++ status = msix_capability_init(dev, entries, nvec);
+
-+ wmb();
-+ do {
-+ ssleep(1);
-+ for_each_online_cpu(cpu) {
-+ if (cpu_isset(cpu, map) &&
-+ !per_cpu(cpu_idle_state, cpu))
-+ cpu_clear(cpu, map);
-+ }
-+ cpus_and(map, map, cpu_online_map);
-+ } while (!cpus_empty(map));
-+}
-+EXPORT_SYMBOL_GPL(cpu_idle_wait);
++ if ( !status )
++ dev->irq_old = temp;
++ else
++ dev->irq = temp;
+
-+void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c)
-+{
++ return status;
+}
+
-+static int __init idle_setup (char *str)
++extern void pci_frontend_disable_msix(struct pci_dev* dev);
++void pci_disable_msix(struct pci_dev* dev)
+{
-+ if (!strncmp(str, "poll", 4)) {
-+ printk("using polling idle threads.\n");
-+ pm_idle = poll_idle;
-+ }
-+
-+ boot_option_idle_override = 1;
-+ return 1;
-+}
++ int pos;
++ u16 control;
+
-+__setup("idle=", idle_setup);
+
-+/* Prints also some state that isn't saved in the pt_regs */
-+void __show_regs(struct pt_regs * regs)
-+{
-+ unsigned long fs, gs, shadowgs;
-+ unsigned int fsindex,gsindex;
-+ unsigned int ds,cs,es;
++ if (!pci_msi_enable)
++ return;
++ if (!dev)
++ return;
+
-+ printk("\n");
-+ print_modules();
-+ printk("Pid: %d, comm: %.20s %s %s %.*s\n",
-+ current->pid, current->comm, print_tainted(),
-+ system_utsname.release,
-+ (int)strcspn(system_utsname.version, " "),
-+ system_utsname.version);
-+ printk("RIP: %04lx:[<%016lx>] ", regs->cs & 0xffff, regs->rip);
-+ printk_address(regs->rip);
-+ printk("RSP: %04lx:%016lx EFLAGS: %08lx\n", regs->ss, regs->rsp,
-+ regs->eflags);
-+ printk("RAX: %016lx RBX: %016lx RCX: %016lx\n",
-+ regs->rax, regs->rbx, regs->rcx);
-+ printk("RDX: %016lx RSI: %016lx RDI: %016lx\n",
-+ regs->rdx, regs->rsi, regs->rdi);
-+ printk("RBP: %016lx R08: %016lx R09: %016lx\n",
-+ regs->rbp, regs->r8, regs->r9);
-+ printk("R10: %016lx R11: %016lx R12: %016lx\n",
-+ regs->r10, regs->r11, regs->r12);
-+ printk("R13: %016lx R14: %016lx R15: %016lx\n",
-+ regs->r13, regs->r14, regs->r15);
++#ifdef CONFIG_XEN_PCIDEV_FRONTEND
++ if (!is_initial_xendomain()) {
++ pci_frontend_disable_msix(dev);
++ dev->irq = dev->irq_old;
++ return;
++ }
++#endif
+
-+ asm("mov %%ds,%0" : "=r" (ds));
-+ asm("mov %%cs,%0" : "=r" (cs));
-+ asm("mov %%es,%0" : "=r" (es));
-+ asm("mov %%fs,%0" : "=r" (fsindex));
-+ asm("mov %%gs,%0" : "=r" (gsindex));
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
++ if (!pos)
++ return;
+
-+ rdmsrl(MSR_FS_BASE, fs);
-+ rdmsrl(MSR_GS_BASE, gs);
-+ rdmsrl(MSR_KERNEL_GS_BASE, shadowgs);
++ pci_read_config_word(dev, msi_control_reg(pos), &control);
++ if (!(control & PCI_MSIX_FLAGS_ENABLE))
++ return;
+
-+ printk("FS: %016lx(%04x) GS:%016lx(%04x) knlGS:%016lx\n",
-+ fs,fsindex,gs,gsindex,shadowgs);
-+ printk("CS: %04x DS: %04x ES: %04x\n", cs, ds, es);
++ msi_remove_pci_irq_vectors(dev);
+
++ /* Disable MSI mode */
++ disable_msi_mode(dev, pos, PCI_CAP_ID_MSIX);
+}
+
-+void show_regs(struct pt_regs *regs)
++/**
++ * msi_remove_pci_irq_vectors - reclaim MSI(X) vectors to unused state
++ * @dev: pointer to the pci_dev data structure of MSI(X) device function
++ *
++ * Being called during hotplug remove, from which the device function
++ * is hot-removed. All previous assigned MSI/MSI-X vectors, if
++ * allocated for this device function, are reclaimed to unused state,
++ * which may be used later on.
++ **/
++void msi_remove_pci_irq_vectors(struct pci_dev* dev)
+{
-+ printk("CPU %d:", smp_processor_id());
-+ __show_regs(regs);
-+ show_trace(NULL, regs, (void *)(regs + 1));
++ unsigned long flags;
++ struct msi_dev_list *msi_dev_entry;
++ struct msi_pirq_entry *pirq_entry, *tmp;
++
++ if (!pci_msi_enable || !dev)
++ return;
++
++ msi_dev_entry = get_msi_dev_pirq_list(dev);
++
++ spin_lock_irqsave(&msi_dev_entry->pirq_list_lock, flags);
++ if (!list_empty(&msi_dev_entry->pirq_list_head))
++ {
++ printk(KERN_WARNING "msix pirqs for dev %02x:%02x:%01x are not freed \
++ before acquire again.\n", dev->bus->number, PCI_SLOT(dev->devfn),
++ PCI_FUNC(dev->devfn));
++ list_for_each_entry_safe(pirq_entry, tmp,
++ &msi_dev_entry->pirq_list_head, list) {
++ msi_unmap_pirq(dev, pirq_entry->pirq);
++ list_del(&pirq_entry->list);
++ kfree(pirq_entry);
++ }
++ }
++ spin_unlock_irqrestore(&msi_dev_entry->pirq_list_lock, flags);
++ dev->irq = dev->irq_old;
+}
+
-+/*
-+ * Free current thread data structures etc..
-+ */
-+void exit_thread(void)
++void pci_no_msi(void)
+{
-+ struct task_struct *me = current;
-+ struct thread_struct *t = &me->thread;
++ pci_msi_enable = 0;
++}
+
-+ if (me->thread.io_bitmap_ptr) {
-+#ifndef CONFIG_X86_NO_TSS
-+ struct tss_struct *tss = &per_cpu(init_tss, get_cpu());
-+#endif
++EXPORT_SYMBOL(pci_enable_msi);
++EXPORT_SYMBOL(pci_disable_msi);
++EXPORT_SYMBOL(pci_enable_msix);
++EXPORT_SYMBOL(pci_disable_msix);
+#ifdef CONFIG_XEN
-+ struct physdev_set_iobitmap iobmp_op;
-+ memset(&iobmp_op, 0, sizeof(iobmp_op));
++EXPORT_SYMBOL(register_msi_get_owner);
++EXPORT_SYMBOL(unregister_msi_get_owner);
+#endif
+
-+ kfree(t->io_bitmap_ptr);
-+ t->io_bitmap_ptr = NULL;
-+ /*
-+ * Careful, clear this in the TSS too:
-+ */
-+#ifndef CONFIG_X86_NO_TSS
-+ memset(tss->io_bitmap, 0xff, t->io_bitmap_max);
-+ put_cpu();
-+#endif
-+#ifdef CONFIG_XEN
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap, &iobmp_op);
-+#endif
-+ t->io_bitmap_max = 0;
-+ }
-+}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/pci/quirks.c
+--- a/drivers/pci/quirks.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/pci/quirks.c Wed Sep 10 10:54:08 2008 +0100
+@@ -22,6 +22,40 @@
+ #include <linux/delay.h>
+ #include <linux/acpi.h>
+ #include "pci.h"
+
-+void load_gs_index(unsigned gs)
++/* A global flag which signals if we should page-align PCI mem windows. */
++int pci_mem_align = 0;
++
++static int __init set_pci_mem_align(char *str)
+{
-+ HYPERVISOR_set_segment_base(SEGBASE_GS_USER_SEL, gs);
++ pci_mem_align = 1;
++ return 1;
+}
++__setup("pci-mem-align", set_pci_mem_align);
+
-+void flush_thread(void)
++/* This quirk function enables us to force all memory resources which are
++ * assigned to PCI devices, to be page-aligned.
++ */
++static void __devinit quirk_align_mem_resources(struct pci_dev *dev)
+{
-+ struct task_struct *tsk = current;
-+ struct thread_info *t = current_thread_info();
++ int i;
++ struct resource *r;
++ resource_size_t old_start;
+
-+ if (t->flags & _TIF_ABI_PENDING) {
-+ t->flags ^= (_TIF_ABI_PENDING | _TIF_IA32);
-+ if (t->flags & _TIF_IA32)
-+ current_thread_info()->status |= TS_COMPAT;
-+ }
++ if (!pci_mem_align)
++ return;
+
-+ tsk->thread.debugreg0 = 0;
-+ tsk->thread.debugreg1 = 0;
-+ tsk->thread.debugreg2 = 0;
-+ tsk->thread.debugreg3 = 0;
-+ tsk->thread.debugreg6 = 0;
-+ tsk->thread.debugreg7 = 0;
-+ memset(tsk->thread.tls_array, 0, sizeof(tsk->thread.tls_array));
-+ /*
-+ * Forget coprocessor state..
-+ */
-+ clear_fpu(tsk);
-+ clear_used_math();
-+}
++ for (i=0; i < DEVICE_COUNT_RESOURCE; i++) {
++ r = &dev->resource[i];
++ if ((r == NULL) || !(r->flags & IORESOURCE_MEM))
++ continue;
+
-+void release_thread(struct task_struct *dead_task)
-+{
-+ if (dead_task->mm) {
-+ if (dead_task->mm->context.size) {
-+ printk("WARNING: dead process %8s still has LDT? <%p/%d>\n",
-+ dead_task->comm,
-+ dead_task->mm->context.ldt,
-+ dead_task->mm->context.size);
-+ BUG();
-+ }
++ old_start = r->start;
++ r->start = (r->start + PAGE_SIZE - 1) & ~(PAGE_SIZE - 1);
++ r->end = r->end - (old_start - r->start);
+ }
+}
++DECLARE_PCI_FIXUP_HEADER(PCI_ANY_ID, PCI_ANY_ID, quirk_align_mem_resources);
+
+ /* The Mellanox Tavor device gives false positive parity errors
+ * Mark this device with a broken_parity_status, to allow
+@@ -1512,10 +1546,11 @@
+
+ static void __devinit quirk_e100_interrupt(struct pci_dev *dev)
+ {
+- u16 command;
++ u16 command, pmcsr;
+ u32 bar;
+ u8 __iomem *csr;
+ u8 cmd_hi;
++ int pm;
+
+ switch (dev->device) {
+ /* PCI IDs taken from drivers/net/e100.c */
+@@ -1549,6 +1584,17 @@
+
+ if (!(command & PCI_COMMAND_MEMORY) || !bar)
+ return;
+
-+static inline void set_32bit_tls(struct task_struct *t, int tls, u32 addr)
-+{
-+ struct user_desc ud = {
-+ .base_addr = addr,
-+ .limit = 0xfffff,
-+ .seg_32bit = 1,
-+ .limit_in_pages = 1,
-+ .useable = 1,
-+ };
-+ struct n_desc_struct *desc = (void *)t->thread.tls_array;
-+ desc += tls;
-+ desc->a = LDT_entry_a(&ud);
-+ desc->b = LDT_entry_b(&ud);
-+}
++ /*
++ * Check that the device is in the D0 power state. If it's not,
++ * there is no point to look any further.
++ */
++ pm = pci_find_capability(dev, PCI_CAP_ID_PM);
++ if (pm) {
++ pci_read_config_word(dev, pm + PCI_PM_CTRL, &pmcsr);
++ if ((pmcsr & PCI_PM_CTRL_STATE_MASK) != PCI_D0)
++ return;
++ }
+
+ csr = ioremap(bar, 8);
+ if (!csr) {
+diff -r d894e36cfc30 -r 0aa021803deb drivers/pnp/manager.c
+--- a/drivers/pnp/manager.c Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/pnp/manager.c Wed Sep 10 10:54:08 2008 +0100
+@@ -168,7 +168,7 @@
+ return 0;
+ }
+
+-static int pnp_assign_dma(struct pnp_dev *dev, struct pnp_dma *rule, int idx)
++static void pnp_assign_dma(struct pnp_dev *dev, struct pnp_dma *rule, int idx)
+ {
+ resource_size_t *start, *end;
+ unsigned long *flags;
+@@ -179,18 +179,14 @@
+ 1, 3, 5, 6, 7, 0, 2, 4
+ };
+
+- if (!dev || !rule)
+- return -EINVAL;
+-
+ if (idx >= PNP_MAX_DMA) {
+ pnp_err("More than 2 dmas is incompatible with pnp specifications.");
+- /* pretend we were successful so at least the manager won't try again */
+- return 1;
++ return;
+ }
+
+ /* check if this resource has been manually set, if so skip */
+ if (!(dev->res.dma_resource[idx].flags & IORESOURCE_AUTO))
+- return 1;
++ return;
+
+ start = &dev->res.dma_resource[idx].start;
+ end = &dev->res.dma_resource[idx].end;
+@@ -200,19 +196,17 @@
+ *flags |= rule->flags | IORESOURCE_DMA;
+ *flags &= ~IORESOURCE_UNSET;
+
+- if (!rule->map) {
+- *flags |= IORESOURCE_DISABLED;
+- return 1; /* skip disabled resource requests */
+- }
+-
+ for (i = 0; i < 8; i++) {
+ if(rule->map & (1<<xtab[i])) {
+ *start = *end = xtab[i];
+ if(pnp_check_dma(dev, idx))
+- return 1;
++ return;
+ }
+ }
+- return 0;
++#ifdef MAX_DMA_CHANNELS
++ *start = *end = MAX_DMA_CHANNELS;
++#endif
++ *flags |= IORESOURCE_UNSET | IORESOURCE_DISABLED;
+ }
+
+ /**
+@@ -331,8 +325,7 @@
+ irq = irq->next;
+ }
+ while (dma) {
+- if (!pnp_assign_dma(dev, dma, ndma))
+- goto fail;
++ pnp_assign_dma(dev, dma, ndma);
+ ndma++;
+ dma = dma->next;
+ }
+@@ -367,8 +360,7 @@
+ irq = irq->next;
+ }
+ while (dma) {
+- if (!pnp_assign_dma(dev, dma, ndma))
+- goto fail;
++ pnp_assign_dma(dev, dma, ndma);
+ ndma++;
+ dma = dma->next;
+ }
+diff -r d894e36cfc30 -r 0aa021803deb drivers/serial/Kconfig
+--- a/drivers/serial/Kconfig Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/serial/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -11,6 +11,7 @@
+ config SERIAL_8250
+ tristate "8250/16550 and compatible serial support"
+ depends on (BROKEN || !SPARC)
++ depends on !XEN_DISABLE_SERIAL
+ select SERIAL_CORE
+ ---help---
+ This selects whether you want to include the driver for the standard
+diff -r d894e36cfc30 -r 0aa021803deb drivers/video/Kconfig
+--- a/drivers/video/Kconfig Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/video/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -1254,7 +1254,7 @@
+ tristate "Cyberblade/i1 support"
+ depends on FB && PCI && X86_32 && !64BIT
+ select FB_CFB_IMAGEBLIT
+- select VIDEO_SELECT
++ select VIDEO_SELECT if !XEN
+ ---help---
+ This driver is supposed to support the Trident Cyberblade/i1
+ graphics core integrated in the VIA VT8601A North Bridge,
+diff -r d894e36cfc30 -r 0aa021803deb drivers/video/console/Kconfig
+--- a/drivers/video/console/Kconfig Tue Sep 09 11:37:38 2008 +0200
++++ b/drivers/video/console/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -53,6 +53,7 @@
+ config VIDEO_SELECT
+ bool "Video mode selection support"
+ depends on X86 && VGA_CONSOLE
++ depends on !XEN
+ ---help---
+ This enables support for text mode selection on kernel startup. If
+ you want to take advantage of some high-resolution text mode your
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/Kconfig
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,298 @@
++#
++# This Kconfig describe xen options
++#
+
-+static inline u32 read_32bit_tls(struct task_struct *t, int tls)
-+{
-+ struct desc_struct *desc = (void *)t->thread.tls_array;
-+ desc += tls;
-+ return desc->base0 |
-+ (((u32)desc->base1) << 16) |
-+ (((u32)desc->base2) << 24);
-+}
++mainmenu "Xen Configuration"
+
-+/*
-+ * This gets called before we allocate a new thread and copy
-+ * the current task into it.
-+ */
-+void prepare_to_copy(struct task_struct *tsk)
-+{
-+ unlazy_fpu(tsk);
-+}
++config XEN
++ bool
+
-+int copy_thread(int nr, unsigned long clone_flags, unsigned long rsp,
-+ unsigned long unused,
-+ struct task_struct * p, struct pt_regs * regs)
-+{
-+ int err;
-+ struct pt_regs * childregs;
-+ struct task_struct *me = current;
++if XEN
++config XEN_INTERFACE_VERSION
++ hex
++ default 0x00030207
+
-+ childregs = ((struct pt_regs *)
-+ (THREAD_SIZE + task_stack_page(p))) - 1;
-+ *childregs = *regs;
++menu "XEN"
+
-+ childregs->rax = 0;
-+ childregs->rsp = rsp;
-+ if (rsp == ~0UL)
-+ childregs->rsp = (unsigned long)childregs;
++config XEN_PRIVILEGED_GUEST
++ bool "Privileged Guest (domain 0)"
++ help
++ Support for privileged operation (domain 0)
++
++config XEN_UNPRIVILEGED_GUEST
++ def_bool !XEN_PRIVILEGED_GUEST
++
++config XEN_PRIVCMD
++ def_bool y
++ depends on PROC_FS
++
++config XEN_XENBUS_DEV
++ def_bool y
++ depends on PROC_FS
++
++config XEN_BACKEND
++ tristate "Backend driver support"
++ default XEN_PRIVILEGED_GUEST
++ help
++ Support for backend device drivers that provide I/O services
++ to other virtual machines.
++
++config XEN_BLKDEV_BACKEND
++ tristate "Block-device backend driver"
++ depends on XEN_BACKEND
++ default XEN_BACKEND
++ help
++ The block-device backend driver allows the kernel to export its
++ block devices to other guests via a high-performance shared-memory
++ interface.
++
++config XEN_BLKDEV_TAP
++ tristate "Block-device tap backend driver"
++ depends on XEN_BACKEND
++ default XEN_BACKEND
++ help
++ The block tap driver is an alternative to the block back driver
++ and allows VM block requests to be redirected to userspace through
++ a device interface. The tap allows user-space development of
++ high-performance block backends, where disk images may be implemented
++ as files, in memory, or on other hosts across the network. This
++ driver can safely coexist with the existing blockback driver.
++
++config XEN_NETDEV_BACKEND
++ tristate "Network-device backend driver"
++ depends on XEN_BACKEND && NET
++ default XEN_BACKEND
++ help
++ The network-device backend driver allows the kernel to export its
++ network devices to other guests via a high-performance shared-memory
++ interface.
++
++config XEN_NETDEV_PIPELINED_TRANSMITTER
++ bool "Pipelined transmitter (DANGEROUS)"
++ depends on XEN_NETDEV_BACKEND
++ help
++ If the net backend is a dumb domain, such as a transparent Ethernet
++ bridge with no local IP interface, it is safe to say Y here to get
++ slightly lower network overhead.
++ If the backend has a local IP interface; or may be doing smart things
++ like reassembling packets to perform firewall filtering; or if you
++ are unsure; or if you experience network hangs when this option is
++ enabled; then you must say N here.
++
++config XEN_NETDEV_LOOPBACK
++ tristate "Network-device loopback driver"
++ depends on XEN_NETDEV_BACKEND
++ help
++ A two-interface loopback device to emulate a local netfront-netback
++ connection. If unsure, it is probably safe to say N here.
++
++config XEN_PCIDEV_BACKEND
++ tristate "PCI-device backend driver"
++ depends on PCI && XEN_BACKEND
++ default XEN_BACKEND
++ help
++ The PCI device backend driver allows the kernel to export arbitrary
++ PCI devices to other guests. If you select this to be a module, you
++ will need to make sure no other driver has bound to the device(s)
++ you want to make visible to other guests.
++
++choice
++ prompt "PCI Backend Mode"
++ depends on XEN_PCIDEV_BACKEND
++ default XEN_PCIDEV_BACKEND_VPCI if !IA64
++ default XEN_PCIDEV_BACKEND_CONTROLLER if IA64
++
++config XEN_PCIDEV_BACKEND_VPCI
++ bool "Virtual PCI"
++ ---help---
++ This PCI Backend hides the true PCI topology and makes the frontend
++ think there is a single PCI bus with only the exported devices on it.
++ For example, a device at 03:05.0 will be re-assigned to 00:00.0. A
++ second device at 02:1a.1 will be re-assigned to 00:01.1.
++
++config XEN_PCIDEV_BACKEND_PASS
++ bool "Passthrough"
++ ---help---
++ This PCI Backend provides a real view of the PCI topology to the
++ frontend (for example, a device at 06:01.b will still appear at
++ 06:01.b to the frontend). This is similar to how Xen 2.0.x exposed
++ PCI devices to its driver domains. This may be required for drivers
++ which depend on finding their hardward in certain bus/slot
++ locations.
++
++config XEN_PCIDEV_BACKEND_SLOT
++ bool "Slot"
++ ---help---
++ This PCI Backend hides the true PCI topology and makes the frontend
++ think there is a single PCI bus with only the exported devices on it.
++ Contrary to the virtual PCI backend, a function becomes a new slot.
++ For example, a device at 03:05.2 will be re-assigned to 00:00.0. A
++ second device at 02:1a.1 will be re-assigned to 00:01.0.
++
++config XEN_PCIDEV_BACKEND_CONTROLLER
++ bool "Controller"
++ depends on IA64
++ ---help---
++ This PCI backend virtualizes the PCI bus topology by providing a
++ virtual bus per PCI root device. Devices which are physically under
++ the same root bus will appear on the same virtual bus. For systems
++ with complex I/O addressing, this is the only backend which supports
++ extended I/O port spaces and MMIO translation offsets. This backend
++ also supports slot virtualization. For example, a device at
++ 0000:01:02.1 will be re-assigned to 0000:00:00.0. A second device
++ at 0000:02:05.0 (behind a P2P bridge on bus 0000:01) will be
++ re-assigned to 0000:00:01.0. A third device at 0000:16:05.0 (under
++ a different PCI root bus) will be re-assigned to 0000:01:00.0.
++
++endchoice
++
++config XEN_PCIDEV_BE_DEBUG
++ bool "PCI Backend Debugging"
++ depends on XEN_PCIDEV_BACKEND
++
++config XEN_TPMDEV_BACKEND
++ tristate "TPM-device backend driver"
++ depends on XEN_BACKEND
++ help
++ The TPM-device backend driver
++
++config XEN_SCSI_BACKEND
++ tristate "SCSI backend driver"
++ depends on SCSI && XEN_BACKEND
++ default m
++ help
++ The SCSI backend driver allows the kernel to export its SCSI Devices
++ to other guests via a high-performance shared-memory interface.
++
++config XEN_BLKDEV_FRONTEND
++ tristate "Block-device frontend driver"
++ default y
++ help
++ The block-device frontend driver allows the kernel to access block
++ devices mounted within another guest OS. Unless you are building a
++ dedicated device-driver domain, or your master control domain
++ (domain 0), then you almost certainly want to say Y here.
++
++config XEN_NETDEV_FRONTEND
++ tristate "Network-device frontend driver"
++ depends on NET
++ default y
++ help
++ The network-device frontend driver allows the kernel to access
++ network interfaces within another guest OS. Unless you are building a
++ dedicated device-driver domain, or your master control domain
++ (domain 0), then you almost certainly want to say Y here.
++
++config XEN_SCSI_FRONTEND
++ tristate "SCSI frontend driver"
++ depends on SCSI
++ default m
++ help
++ The SCSI frontend driver allows the kernel to access SCSI Devices
++ within another guest OS.
++
++config XEN_GRANT_DEV
++ tristate "User-space granted page access driver"
++ default XEN_PRIVILEGED_GUEST
++ help
++ Device for accessing (in user-space) pages that have been granted
++ by other domains.
++
++config XEN_FRAMEBUFFER
++ tristate "Framebuffer-device frontend driver"
++ depends on FB
++ select FB_CFB_FILLRECT
++ select FB_CFB_COPYAREA
++ select FB_CFB_IMAGEBLIT
++ default y
++ help
++ The framebuffer-device frontend drivers allows the kernel to create a
++ virtual framebuffer. This framebuffer can be viewed in another
++ domain. Unless this domain has access to a real video card, you
++ probably want to say Y here.
++
++config XEN_KEYBOARD
++ tristate "Keyboard-device frontend driver"
++ depends on XEN_FRAMEBUFFER && INPUT
++ default y
++ help
++ The keyboard-device frontend driver allows the kernel to create a
++ virtual keyboard. This keyboard can then be driven by another
++ domain. If you've said Y to CONFIG_XEN_FRAMEBUFFER, you probably
++ want to say Y here.
++
++config XEN_SCRUB_PAGES
++ bool "Scrub memory before freeing it to Xen"
++ default y
++ help
++ Erase memory contents before freeing it back to Xen's global
++ pool. This ensures that any secrets contained within that
++ memory (e.g., private keys) cannot be found by other guests that
++ may be running on the machine. Most people will want to say Y here.
++ If security is not a concern then you may increase performance by
++ saying N.
++
++config XEN_DISABLE_SERIAL
++ bool "Disable serial port drivers"
++ default y
++ help
++ Disable serial port drivers, allowing the Xen console driver
++ to provide a serial console at ttyS0.
++
++config XEN_SYSFS
++ tristate "Export Xen attributes in sysfs"
++ depends on SYSFS
++ select SYS_HYPERVISOR
++ default y
++ help
++ Xen hypervisor attributes will show up under /sys/hypervisor/.
++
++choice
++ prompt "Xen version compatibility"
++ default XEN_COMPAT_030002_AND_LATER
++
++ config XEN_COMPAT_030002_AND_LATER
++ bool "3.0.2 and later"
++
++ config XEN_COMPAT_030004_AND_LATER
++ bool "3.0.4 and later"
++
++ config XEN_COMPAT_030100_AND_LATER
++ bool "3.1.0 and later"
++
++ config XEN_COMPAT_LATEST_ONLY
++ bool "no compatibility code"
++
++endchoice
++
++config XEN_COMPAT
++ hex
++ default 0xffffff if XEN_COMPAT_LATEST_ONLY
++ default 0x030100 if XEN_COMPAT_030100_AND_LATER
++ default 0x030004 if XEN_COMPAT_030004_AND_LATER
++ default 0x030002 if XEN_COMPAT_030002_AND_LATER
++ default 0
++
++endmenu
+
-+ p->thread.rsp = (unsigned long) childregs;
-+ p->thread.rsp0 = (unsigned long) (childregs+1);
-+ p->thread.userrsp = me->thread.userrsp;
++config HAVE_IRQ_IGNORE_UNHANDLED
++ def_bool y
+
-+ set_tsk_thread_flag(p, TIF_FORK);
++config NO_IDLE_HZ
++ def_bool y
+
-+ p->thread.fs = me->thread.fs;
-+ p->thread.gs = me->thread.gs;
++config XEN_SMPBOOT
++ def_bool y
++ depends on SMP && !PPC_XEN
+
-+ asm("mov %%gs,%0" : "=m" (p->thread.gsindex));
-+ asm("mov %%fs,%0" : "=m" (p->thread.fsindex));
-+ asm("mov %%es,%0" : "=m" (p->thread.es));
-+ asm("mov %%ds,%0" : "=m" (p->thread.ds));
++config XEN_BALLOON
++ def_bool y
++ depends on !PPC_XEN
+
-+ if (unlikely(me->thread.io_bitmap_ptr != NULL)) {
-+ p->thread.io_bitmap_ptr = kmalloc(IO_BITMAP_BYTES, GFP_KERNEL);
-+ if (!p->thread.io_bitmap_ptr) {
-+ p->thread.io_bitmap_max = 0;
-+ return -ENOMEM;
-+ }
-+ memcpy(p->thread.io_bitmap_ptr, me->thread.io_bitmap_ptr,
-+ IO_BITMAP_BYTES);
-+ }
++config XEN_XENCOMM
++ bool
+
-+ /*
-+ * Set a new TLS for the child thread?
-+ */
-+ if (clone_flags & CLONE_SETTLS) {
-+#ifdef CONFIG_IA32_EMULATION
-+ if (test_thread_flag(TIF_IA32))
-+ err = ia32_child_tls(p, childregs);
-+ else
-+#endif
-+ err = do_arch_prctl(p, ARCH_SET_FS, childregs->r8);
-+ if (err)
-+ goto out;
-+ }
-+ p->thread.iopl = current->thread.iopl;
++config XEN_DEVMEM
++ def_bool y
+
-+ err = 0;
-+out:
-+ if (err && p->thread.io_bitmap_ptr) {
-+ kfree(p->thread.io_bitmap_ptr);
-+ p->thread.io_bitmap_max = 0;
-+ }
-+ return err;
-+}
++endif
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,22 @@
++obj-y += core/
++obj-y += console/
++obj-y += evtchn/
++obj-y += xenbus/
++obj-y += char/
+
-+static inline void __save_init_fpu( struct task_struct *tsk )
-+{
-+ asm volatile( "rex64 ; fxsave %0 ; fnclex"
-+ : "=m" (tsk->thread.i387.fxsave));
-+ tsk->thread_info->status &= ~TS_USEDFPU;
-+}
++obj-y += util.o
++obj-$(CONFIG_XEN_BALLOON) += balloon/
++obj-$(CONFIG_XEN_BLKDEV_BACKEND) += blkback/
++obj-$(CONFIG_XEN_BLKDEV_TAP) += blktap/
++obj-$(CONFIG_XEN_NETDEV_BACKEND) += netback/
++obj-$(CONFIG_XEN_TPMDEV_BACKEND) += tpmback/
++obj-$(CONFIG_XEN_BLKDEV_FRONTEND) += blkfront/
++obj-$(CONFIG_XEN_NETDEV_FRONTEND) += netfront/
++obj-$(CONFIG_XEN_PCIDEV_BACKEND) += pciback/
++obj-$(CONFIG_XEN_PCIDEV_FRONTEND) += pcifront/
++obj-$(CONFIG_XEN_FRAMEBUFFER) += fbfront/
++obj-$(CONFIG_XEN_KEYBOARD) += fbfront/
++obj-$(CONFIG_XEN_SCSI_BACKEND) += scsiback/
++obj-$(CONFIG_XEN_SCSI_FRONTEND) += scsifront/
++obj-$(CONFIG_XEN_PRIVCMD) += privcmd/
++obj-$(CONFIG_XEN_GRANT_DEV) += gntdev/
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/balloon/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/balloon/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,2 @@
+
-+/*
-+ * switch_to(x,y) should switch tasks from x to y.
++obj-y := balloon.o sysfs.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/balloon/balloon.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/balloon/balloon.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,724 @@
++/******************************************************************************
++ * balloon.c
+ *
-+ * This could still be optimized:
-+ * - fold all the options into a flag word and test it with a single test.
-+ * - could test fs/gs bitsliced
++ * Xen balloon driver - enables returning/claiming memory to/from Xen.
+ *
-+ * Kprobes not supported here. Set the probe on schedule instead.
++ * Copyright (c) 2003, B Dragovic
++ * Copyright (c) 2003-2004, M Williamson, K Fraser
++ * Copyright (c) 2005 Dan M. Smith, IBM Corporation
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
-+__kprobes struct task_struct *
-+__switch_to(struct task_struct *prev_p, struct task_struct *next_p)
-+{
-+ struct thread_struct *prev = &prev_p->thread,
-+ *next = &next_p->thread;
-+ int cpu = smp_processor_id();
-+#ifndef CONFIG_X86_NO_TSS
-+ struct tss_struct *tss = &per_cpu(init_tss, cpu);
++
++#include <linux/kernel.h>
++#include <linux/module.h>
++#include <linux/sched.h>
++#include <linux/errno.h>
++#include <linux/mm.h>
++#include <linux/mman.h>
++#include <linux/smp_lock.h>
++#include <linux/pagemap.h>
++#include <linux/bootmem.h>
++#include <linux/highmem.h>
++#include <linux/vmalloc.h>
++#include <linux/mutex.h>
++#include <xen/xen_proc.h>
++#include <asm/hypervisor.h>
++#include <xen/balloon.h>
++#include <xen/interface/memory.h>
++#include <asm/maddr.h>
++#include <asm/page.h>
++#include <asm/pgalloc.h>
++#include <asm/pgtable.h>
++#include <asm/uaccess.h>
++#include <asm/tlb.h>
++#include <linux/highmem.h>
++#include <linux/list.h>
++#include <xen/xenbus.h>
++#include "common.h"
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
+#endif
-+ struct physdev_set_iopl iopl_op;
-+ struct physdev_set_iobitmap iobmp_op;
-+ multicall_entry_t _mcl[8], *mcl = _mcl;
+
-+ /*
-+ * This is basically '__unlazy_fpu', except that we queue a
-+ * multicall to indicate FPU task switch, rather than
-+ * synchronously trapping to Xen.
-+ * The AMD workaround requires it to be after DS reload, or
-+ * after DS has been cleared, which we do in __prepare_arch_switch.
-+ */
-+ if (prev_p->thread_info->status & TS_USEDFPU) {
-+ __save_init_fpu(prev_p); /* _not_ save_init_fpu() */
-+ mcl->op = __HYPERVISOR_fpu_taskswitch;
-+ mcl->args[0] = 1;
-+ mcl++;
-+ }
++#ifdef CONFIG_PROC_FS
++static struct proc_dir_entry *balloon_pde;
++#endif
+
-+ /*
-+ * Reload esp0, LDT and the page table pointer:
-+ */
-+ mcl->op = __HYPERVISOR_stack_switch;
-+ mcl->args[0] = __KERNEL_DS;
-+ mcl->args[1] = next->rsp0;
-+ mcl++;
++static DEFINE_MUTEX(balloon_mutex);
+
-+ /*
-+ * Load the per-thread Thread-Local Storage descriptor.
-+ * This is load_TLS(next, cpu) with multicalls.
-+ */
-+#define C(i) do { \
-+ if (unlikely(next->tls_array[i] != prev->tls_array[i])) { \
-+ mcl->op = __HYPERVISOR_update_descriptor; \
-+ mcl->args[0] = virt_to_machine( \
-+ &cpu_gdt(cpu)[GDT_ENTRY_TLS_MIN + i]); \
-+ mcl->args[1] = next->tls_array[i]; \
-+ mcl++; \
-+ } \
-+} while (0)
-+ C(0); C(1); C(2);
-+#undef C
++/*
++ * Protects atomic reservation decrease/increase against concurrent increases.
++ * Also protects non-atomic updates of current_pages and driver_pages, and
++ * balloon lists.
++ */
++DEFINE_SPINLOCK(balloon_lock);
+
-+ if (unlikely(prev->iopl != next->iopl)) {
-+ iopl_op.iopl = (next->iopl == 0) ? 1 : next->iopl;
-+ mcl->op = __HYPERVISOR_physdev_op;
-+ mcl->args[0] = PHYSDEVOP_set_iopl;
-+ mcl->args[1] = (unsigned long)&iopl_op;
-+ mcl++;
-+ }
++struct balloon_stats balloon_stats;
+
-+ if (unlikely(prev->io_bitmap_ptr || next->io_bitmap_ptr)) {
-+ set_xen_guest_handle(iobmp_op.bitmap,
-+ (char *)next->io_bitmap_ptr);
-+ iobmp_op.nr_ports = next->io_bitmap_ptr ? IO_BITMAP_BITS : 0;
-+ mcl->op = __HYPERVISOR_physdev_op;
-+ mcl->args[0] = PHYSDEVOP_set_iobitmap;
-+ mcl->args[1] = (unsigned long)&iobmp_op;
-+ mcl++;
-+ }
++/* We increase/decrease in batches which fit in a page */
++static unsigned long frame_list[PAGE_SIZE / sizeof(unsigned long)];
+
-+ (void)HYPERVISOR_multicall(_mcl, mcl - _mcl);
-+ /*
-+ * Switch DS and ES.
-+ * This won't pick up thread selector changes, but I guess that is ok.
-+ */
-+ if (unlikely(next->es))
-+ loadsegment(es, next->es);
-+
-+ if (unlikely(next->ds))
-+ loadsegment(ds, next->ds);
++/* VM /proc information for memory */
++extern unsigned long totalram_pages;
+
-+ /*
-+ * Switch FS and GS.
-+ */
-+ if (unlikely(next->fsindex))
-+ loadsegment(fs, next->fsindex);
++#ifndef MODULE
++extern unsigned long totalhigh_pages;
++#define inc_totalhigh_pages() (totalhigh_pages++)
++#define dec_totalhigh_pages() (totalhigh_pages--)
++#else
++#define inc_totalhigh_pages() ((void)0)
++#define dec_totalhigh_pages() ((void)0)
++#endif
+
-+ if (next->fs)
-+ HYPERVISOR_set_segment_base(SEGBASE_FS, next->fs);
-+
-+ if (unlikely(next->gsindex))
-+ load_gs_index(next->gsindex);
++/* List of ballooned pages, threaded through the mem_map array. */
++static LIST_HEAD(ballooned_pages);
+
-+ if (next->gs)
-+ HYPERVISOR_set_segment_base(SEGBASE_GS_USER, next->gs);
++/* Main work function, always executed in process context. */
++static void balloon_process(void *unused);
++static DECLARE_WORK(balloon_worker, balloon_process, NULL);
++static struct timer_list balloon_timer;
+
-+ /*
-+ * Switch the PDA context.
-+ */
-+ prev->userrsp = read_pda(oldrsp);
-+ write_pda(oldrsp, next->userrsp);
-+ write_pda(pcurrent, next_p);
-+ write_pda(kernelstack,
-+ task_stack_page(next_p) + THREAD_SIZE - PDA_STACKOFFSET);
++/* When ballooning out (allocating memory to return to Xen) we don't really
++ want the kernel to try too hard since that can trigger the oom killer. */
++#define GFP_BALLOON \
++ (GFP_HIGHUSER|__GFP_NOWARN|__GFP_NORETRY|__GFP_NOMEMALLOC|__GFP_COLD)
+
-+ /*
-+ * Now maybe reload the debug registers
-+ */
-+ if (unlikely(next->debugreg7)) {
-+ set_debugreg(next->debugreg0, 0);
-+ set_debugreg(next->debugreg1, 1);
-+ set_debugreg(next->debugreg2, 2);
-+ set_debugreg(next->debugreg3, 3);
-+ /* no 4 and 5 */
-+ set_debugreg(next->debugreg6, 6);
-+ set_debugreg(next->debugreg7, 7);
-+ }
++#define PAGE_TO_LIST(p) (&(p)->lru)
++#define LIST_TO_PAGE(l) list_entry((l), struct page, lru)
++#define UNLIST_PAGE(p) \
++ do { \
++ list_del(PAGE_TO_LIST(p)); \
++ PAGE_TO_LIST(p)->next = NULL; \
++ PAGE_TO_LIST(p)->prev = NULL; \
++ } while(0)
+
-+ return prev_p;
++#define IPRINTK(fmt, args...) \
++ printk(KERN_INFO "xen_mem: " fmt, ##args)
++#define WPRINTK(fmt, args...) \
++ printk(KERN_WARNING "xen_mem: " fmt, ##args)
++
++/* balloon_append: add the given page to the balloon. */
++static void balloon_append(struct page *page)
++{
++ /* Lowmem is re-populated first, so highmem pages go at list tail. */
++ if (PageHighMem(page)) {
++ list_add_tail(PAGE_TO_LIST(page), &ballooned_pages);
++ bs.balloon_high++;
++ dec_totalhigh_pages();
++ } else {
++ list_add(PAGE_TO_LIST(page), &ballooned_pages);
++ bs.balloon_low++;
++ }
+}
+
-+/*
-+ * sys_execve() executes a new program.
-+ */
-+asmlinkage
-+long sys_execve(char __user *name, char __user * __user *argv,
-+ char __user * __user *envp, struct pt_regs regs)
++/* balloon_retrieve: rescue a page from the balloon, if it is not empty. */
++static struct page *balloon_retrieve(void)
+{
-+ long error;
-+ char * filename;
++ struct page *page;
+
-+ filename = getname(name);
-+ error = PTR_ERR(filename);
-+ if (IS_ERR(filename))
-+ return error;
-+ error = do_execve(filename, argv, envp, ®s);
-+ if (error == 0) {
-+ task_lock(current);
-+ current->ptrace &= ~PT_DTRACE;
-+ task_unlock(current);
++ if (list_empty(&ballooned_pages))
++ return NULL;
++
++ page = LIST_TO_PAGE(ballooned_pages.next);
++ UNLIST_PAGE(page);
++
++ if (PageHighMem(page)) {
++ bs.balloon_high--;
++ inc_totalhigh_pages();
+ }
-+ putname(filename);
-+ return error;
++ else
++ bs.balloon_low--;
++
++ return page;
+}
+
-+void set_personality_64bit(void)
++static struct page *balloon_first_page(void)
+{
-+ /* inherit personality from parent */
-+
-+ /* Make sure to be in 64bit mode */
-+ clear_thread_flag(TIF_IA32);
++ if (list_empty(&ballooned_pages))
++ return NULL;
++ return LIST_TO_PAGE(ballooned_pages.next);
++}
+
-+ /* TBD: overwrites user setup. Should have two bits.
-+ But 64bit processes have always behaved this way,
-+ so it's not too bad. The main problem is just that
-+ 32bit childs are affected again. */
-+ current->personality &= ~READ_IMPLIES_EXEC;
++static struct page *balloon_next_page(struct page *page)
++{
++ struct list_head *next = PAGE_TO_LIST(page)->next;
++ if (next == &ballooned_pages)
++ return NULL;
++ return LIST_TO_PAGE(next);
+}
+
-+asmlinkage long sys_fork(struct pt_regs *regs)
++static inline void balloon_free_page(struct page *page)
+{
-+ return do_fork(SIGCHLD, regs->rsp, regs, 0, NULL, NULL);
++#ifndef MODULE
++ if (put_page_testzero(page))
++ free_cold_page(page);
++#else
++ /* free_cold_page() is not being exported. */
++ __free_page(page);
++#endif
+}
+
-+asmlinkage long
-+sys_clone(unsigned long clone_flags, unsigned long newsp,
-+ void __user *parent_tid, void __user *child_tid, struct pt_regs *regs)
++static void balloon_alarm(unsigned long unused)
+{
-+ if (!newsp)
-+ newsp = regs->rsp;
-+ return do_fork(clone_flags, newsp, regs, 0, parent_tid, child_tid);
++ schedule_work(&balloon_worker);
+}
+
-+/*
-+ * This is trivial, and on the face of it looks like it
-+ * could equally well be done in user mode.
-+ *
-+ * Not so, for quite unobvious reasons - register pressure.
-+ * In user mode vfork() cannot have a stack frame, and if
-+ * done by calling the "clone()" system call directly, you
-+ * do not have enough call-clobbered registers to hold all
-+ * the information you need.
-+ */
-+asmlinkage long sys_vfork(struct pt_regs *regs)
++static unsigned long current_target(void)
+{
-+ return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, regs->rsp, regs, 0,
-+ NULL, NULL);
++ unsigned long target = min(bs.target_pages, bs.hard_limit);
++ if (target > (bs.current_pages + bs.balloon_low + bs.balloon_high))
++ target = bs.current_pages + bs.balloon_low + bs.balloon_high;
++ return target;
+}
+
-+unsigned long get_wchan(struct task_struct *p)
++static unsigned long minimum_target(void)
+{
-+ unsigned long stack;
-+ u64 fp,rip;
-+ int count = 0;
++#ifndef CONFIG_XEN
++#define max_pfn num_physpages
++#endif
++ unsigned long min_pages, curr_pages = current_target();
+
-+ if (!p || p == current || p->state==TASK_RUNNING)
-+ return 0;
-+ stack = (unsigned long)task_stack_page(p);
-+ if (p->thread.rsp < stack || p->thread.rsp > stack+THREAD_SIZE)
-+ return 0;
-+ fp = *(u64 *)(p->thread.rsp);
-+ do {
-+ if (fp < (unsigned long)stack ||
-+ fp > (unsigned long)stack+THREAD_SIZE)
-+ return 0;
-+ rip = *(u64 *)(fp+8);
-+ if (!in_sched_functions(rip))
-+ return rip;
-+ fp = *(u64 *)fp;
-+ } while (count++ < 16);
-+ return 0;
++#define MB2PAGES(mb) ((mb) << (20 - PAGE_SHIFT))
++ /* Simple continuous piecewiese linear function:
++ * max MiB -> min MiB gradient
++ * 0 0
++ * 16 16
++ * 32 24
++ * 128 72 (1/2)
++ * 512 168 (1/4)
++ * 2048 360 (1/8)
++ * 8192 552 (1/32)
++ * 32768 1320
++ * 131072 4392
++ */
++ if (max_pfn < MB2PAGES(128))
++ min_pages = MB2PAGES(8) + (max_pfn >> 1);
++ else if (max_pfn < MB2PAGES(512))
++ min_pages = MB2PAGES(40) + (max_pfn >> 2);
++ else if (max_pfn < MB2PAGES(2048))
++ min_pages = MB2PAGES(104) + (max_pfn >> 3);
++ else
++ min_pages = MB2PAGES(296) + (max_pfn >> 5);
++#undef MB2PAGES
++
++ /* Don't enforce growth */
++ return min(min_pages, curr_pages);
++#ifndef CONFIG_XEN
++#undef max_pfn
++#endif
+}
+
-+long do_arch_prctl(struct task_struct *task, int code, unsigned long addr)
-+{
-+ int ret = 0;
-+ int doit = task == current;
-+ int cpu;
++static int increase_reservation(unsigned long nr_pages)
++{
++ unsigned long pfn, i, flags;
++ struct page *page;
++ long rc;
++ struct xen_memory_reservation reservation = {
++ .address_bits = 0,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
+
-+ switch (code) {
-+ case ARCH_SET_GS:
-+ if (addr >= TASK_SIZE_OF(task))
-+ return -EPERM;
-+ cpu = get_cpu();
-+ /* handle small bases via the GDT because that's faster to
-+ switch. */
-+ if (addr <= 0xffffffff) {
-+ set_32bit_tls(task, GS_TLS, addr);
-+ if (doit) {
-+ load_TLS(&task->thread, cpu);
-+ load_gs_index(GS_TLS_SEL);
-+ }
-+ task->thread.gsindex = GS_TLS_SEL;
-+ task->thread.gs = 0;
-+ } else {
-+ task->thread.gsindex = 0;
-+ task->thread.gs = addr;
-+ if (doit) {
-+ load_gs_index(0);
-+ ret = HYPERVISOR_set_segment_base(
-+ SEGBASE_GS_USER, addr);
-+ }
-+ }
-+ put_cpu();
-+ break;
-+ case ARCH_SET_FS:
-+ /* Not strictly needed for fs, but do it for symmetry
-+ with gs */
-+ if (addr >= TASK_SIZE_OF(task))
-+ return -EPERM;
-+ cpu = get_cpu();
-+ /* handle small bases via the GDT because that's faster to
-+ switch. */
-+ if (addr <= 0xffffffff) {
-+ set_32bit_tls(task, FS_TLS, addr);
-+ if (doit) {
-+ load_TLS(&task->thread, cpu);
-+ asm volatile("movl %0,%%fs" :: "r"(FS_TLS_SEL));
-+ }
-+ task->thread.fsindex = FS_TLS_SEL;
-+ task->thread.fs = 0;
-+ } else {
-+ task->thread.fsindex = 0;
-+ task->thread.fs = addr;
-+ if (doit) {
-+ /* set the selector to 0 to not confuse
-+ __switch_to */
-+ asm volatile("movl %0,%%fs" :: "r" (0));
-+ ret = HYPERVISOR_set_segment_base(SEGBASE_FS,
-+ addr);
-+ }
-+ }
-+ put_cpu();
-+ break;
-+ case ARCH_GET_FS: {
-+ unsigned long base;
-+ if (task->thread.fsindex == FS_TLS_SEL)
-+ base = read_32bit_tls(task, FS_TLS);
-+ else if (doit)
-+ rdmsrl(MSR_FS_BASE, base);
-+ else
-+ base = task->thread.fs;
-+ ret = put_user(base, (unsigned long __user *)addr);
-+ break;
++ if (nr_pages > ARRAY_SIZE(frame_list))
++ nr_pages = ARRAY_SIZE(frame_list);
++
++ balloon_lock(flags);
++
++ page = balloon_first_page();
++ for (i = 0; i < nr_pages; i++) {
++ BUG_ON(page == NULL);
++ frame_list[i] = page_to_pfn(page);;
++ page = balloon_next_page(page);
+ }
-+ case ARCH_GET_GS: {
-+ unsigned long base;
-+ unsigned gsindex;
-+ if (task->thread.gsindex == GS_TLS_SEL)
-+ base = read_32bit_tls(task, GS_TLS);
-+ else if (doit) {
-+ asm("movl %%gs,%0" : "=r" (gsindex));
-+ if (gsindex)
-+ rdmsrl(MSR_KERNEL_GS_BASE, base);
-+ else
-+ base = task->thread.gs;
++
++ set_xen_guest_handle(reservation.extent_start, frame_list);
++ reservation.nr_extents = nr_pages;
++ rc = HYPERVISOR_memory_op(
++ XENMEM_populate_physmap, &reservation);
++ if (rc < nr_pages) {
++ if (rc > 0) {
++ int ret;
++
++ /* We hit the Xen hard limit: reprobe. */
++ reservation.nr_extents = rc;
++ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &reservation);
++ BUG_ON(ret != rc);
+ }
-+ else
-+ base = task->thread.gs;
-+ ret = put_user(base, (unsigned long __user *)addr);
-+ break;
++ if (rc >= 0)
++ bs.hard_limit = (bs.current_pages + rc -
++ bs.driver_pages);
++ goto out;
+ }
+
-+ default:
-+ ret = -EINVAL;
-+ break;
-+ }
++ for (i = 0; i < nr_pages; i++) {
++ page = balloon_retrieve();
++ BUG_ON(page == NULL);
+
-+ return ret;
-+}
++ pfn = page_to_pfn(page);
++ BUG_ON(!xen_feature(XENFEAT_auto_translated_physmap) &&
++ phys_to_machine_mapping_valid(pfn));
+
-+long sys_arch_prctl(int code, unsigned long addr)
-+{
-+ return do_arch_prctl(current, code, addr);
-+}
++ set_phys_to_machine(pfn, frame_list[i]);
+
-+/*
-+ * Capture the user space registers if the task is not running (in user space)
-+ */
-+int dump_task_regs(struct task_struct *tsk, elf_gregset_t *regs)
-+{
-+ struct pt_regs *pp, ptregs;
++#ifdef CONFIG_XEN
++ /* Link back into the page tables if not highmem. */
++ if (pfn < max_low_pfn) {
++ int ret;
++ ret = HYPERVISOR_update_va_mapping(
++ (unsigned long)__va(pfn << PAGE_SHIFT),
++ pfn_pte_ma(frame_list[i], PAGE_KERNEL),
++ 0);
++ BUG_ON(ret);
++ }
++#endif
+
-+ pp = task_pt_regs(tsk);
++ /* Relinquish the page back to the allocator. */
++ ClearPageReserved(page);
++ init_page_count(page);
++ balloon_free_page(page);
++ }
+
-+ ptregs = *pp;
-+ ptregs.cs &= 0xffff;
-+ ptregs.ss &= 0xffff;
++ bs.current_pages += nr_pages;
++ totalram_pages = bs.current_pages;
+
-+ elf_core_copy_regs(regs, &ptregs);
-+
-+ boot_option_idle_override = 1;
-+ return 1;
-+}
++ out:
++ balloon_unlock(flags);
+
-+unsigned long arch_align_stack(unsigned long sp)
-+{
-+ if (randomize_va_space)
-+ sp -= get_random_int() % 8192;
-+ return sp & ~0xf;
++ return 0;
+}
+
-+#ifndef CONFIG_SMP
-+void _restore_vcpu(void)
++static int decrease_reservation(unsigned long nr_pages)
+{
-+}
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/relocate_kernel.S tmp-linux-2.6-xen.patch/arch/x86_64/kernel/relocate_kernel.S
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/relocate_kernel.S 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/relocate_kernel.S 2007-10-14 01:51:15.000000000 +0200
-@@ -7,31 +7,195 @@
- */
-
- #include <linux/linkage.h>
-+#include <asm/page.h>
-+#include <asm/kexec.h>
-
-- /*
-- * Must be relocatable PIC code callable as a C function, that once
-- * it starts can not use the previous processes stack.
-- */
-- .globl relocate_new_kernel
-+/*
-+ * Must be relocatable PIC code callable as a C function
-+ */
++ unsigned long pfn, i, flags;
++ struct page *page;
++ void *v;
++ int need_sleep = 0;
++ int ret;
++ struct xen_memory_reservation reservation = {
++ .address_bits = 0,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
+
-+#define PTR(x) (x << 3)
-+#define PAGE_ALIGNED (1 << PAGE_SHIFT)
-+#define PAGE_ATTR 0x63 /* _PAGE_PRESENT|_PAGE_RW|_PAGE_ACCESSED|_PAGE_DIRTY */
++ if (nr_pages > ARRAY_SIZE(frame_list))
++ nr_pages = ARRAY_SIZE(frame_list);
+
-+ .text
-+ .align PAGE_ALIGNED
- .code64
-+ .globl relocate_kernel
-+relocate_kernel:
-+ /* %rdi indirection_page
-+ * %rsi page_list
-+ * %rdx start address
-+ */
++ for (i = 0; i < nr_pages; i++) {
++ if ((page = alloc_page(GFP_BALLOON)) == NULL) {
++ nr_pages = i;
++ need_sleep = 1;
++ break;
++ }
+
-+ /* map the control page at its virtual address */
++ pfn = page_to_pfn(page);
++ frame_list[i] = pfn_to_mfn(pfn);
+
-+ movq $0x0000ff8000000000, %r10 /* mask */
-+ mov $(39 - 3), %cl /* bits to shift */
-+ movq PTR(VA_CONTROL_PAGE)(%rsi), %r11 /* address to map */
++ if (!PageHighMem(page)) {
++ v = phys_to_virt(pfn << PAGE_SHIFT);
++ scrub_pages(v, 1);
++#ifdef CONFIG_XEN
++ ret = HYPERVISOR_update_va_mapping(
++ (unsigned long)v, __pte_ma(0), 0);
++ BUG_ON(ret);
++#endif
++ }
++#ifdef CONFIG_XEN_SCRUB_PAGES
++ else {
++ v = kmap(page);
++ scrub_pages(v, 1);
++ kunmap(page);
++ }
++#endif
++ }
+
-+ movq %r11, %r9
-+ andq %r10, %r9
-+ shrq %cl, %r9
++#ifdef CONFIG_XEN
++ /* Ensure that ballooned highmem pages don't have kmaps. */
++ kmap_flush_unused();
++ flush_tlb_all();
++#endif
+
-+ movq PTR(VA_PGD)(%rsi), %r8
-+ addq %r8, %r9
-+ movq PTR(PA_PUD_0)(%rsi), %r8
-+ orq $PAGE_ATTR, %r8
-+ movq %r8, (%r9)
++ balloon_lock(flags);
+
-+ shrq $9, %r10
-+ sub $9, %cl
++ /* No more mappings: invalidate P2M and add to balloon. */
++ for (i = 0; i < nr_pages; i++) {
++ pfn = mfn_to_pfn(frame_list[i]);
++ set_phys_to_machine(pfn, INVALID_P2M_ENTRY);
++ balloon_append(pfn_to_page(pfn));
++ }
+
-+ movq %r11, %r9
-+ andq %r10, %r9
-+ shrq %cl, %r9
++ set_xen_guest_handle(reservation.extent_start, frame_list);
++ reservation.nr_extents = nr_pages;
++ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation, &reservation);
++ BUG_ON(ret != nr_pages);
+
-+ movq PTR(VA_PUD_0)(%rsi), %r8
-+ addq %r8, %r9
-+ movq PTR(PA_PMD_0)(%rsi), %r8
-+ orq $PAGE_ATTR, %r8
-+ movq %r8, (%r9)
++ bs.current_pages -= nr_pages;
++ totalram_pages = bs.current_pages;
+
-+ shrq $9, %r10
-+ sub $9, %cl
++ balloon_unlock(flags);
+
-+ movq %r11, %r9
-+ andq %r10, %r9
-+ shrq %cl, %r9
++ return need_sleep;
++}
+
-+ movq PTR(VA_PMD_0)(%rsi), %r8
-+ addq %r8, %r9
-+ movq PTR(PA_PTE_0)(%rsi), %r8
-+ orq $PAGE_ATTR, %r8
-+ movq %r8, (%r9)
++/*
++ * We avoid multiple worker processes conflicting via the balloon mutex.
++ * We may of course race updates of the target counts (which are protected
++ * by the balloon lock), or with changes to the Xen hard limit, but we will
++ * recover from these in time.
++ */
++static void balloon_process(void *unused)
++{
++ int need_sleep = 0;
++ long credit;
+
-+ shrq $9, %r10
-+ sub $9, %cl
++ mutex_lock(&balloon_mutex);
+
-+ movq %r11, %r9
-+ andq %r10, %r9
-+ shrq %cl, %r9
++ do {
++ credit = current_target() - bs.current_pages;
++ if (credit > 0)
++ need_sleep = (increase_reservation(credit) != 0);
++ if (credit < 0)
++ need_sleep = (decrease_reservation(-credit) != 0);
+
-+ movq PTR(VA_PTE_0)(%rsi), %r8
-+ addq %r8, %r9
-+ movq PTR(PA_CONTROL_PAGE)(%rsi), %r8
-+ orq $PAGE_ATTR, %r8
-+ movq %r8, (%r9)
++#ifndef CONFIG_PREEMPT
++ if (need_resched())
++ schedule();
++#endif
++ } while ((credit != 0) && !need_sleep);
+
-+ /* identity map the control page at its physical address */
++ /* Schedule more work if there is some still to be done. */
++ if (current_target() != bs.current_pages)
++ mod_timer(&balloon_timer, jiffies + HZ);
+
-+ movq $0x0000ff8000000000, %r10 /* mask */
-+ mov $(39 - 3), %cl /* bits to shift */
-+ movq PTR(PA_CONTROL_PAGE)(%rsi), %r11 /* address to map */
++ mutex_unlock(&balloon_mutex);
++}
+
-+ movq %r11, %r9
-+ andq %r10, %r9
-+ shrq %cl, %r9
++/* Resets the Xen limit, sets new target, and kicks off processing. */
++void balloon_set_new_target(unsigned long target)
++{
++ /* No need for lock. Not read-modify-write updates. */
++ bs.hard_limit = ~0UL;
++ bs.target_pages = max(target, minimum_target());
++ schedule_work(&balloon_worker);
++}
+
-+ movq PTR(VA_PGD)(%rsi), %r8
-+ addq %r8, %r9
-+ movq PTR(PA_PUD_1)(%rsi), %r8
-+ orq $PAGE_ATTR, %r8
-+ movq %r8, (%r9)
++static struct xenbus_watch target_watch =
++{
++ .node = "memory/target"
++};
+
-+ shrq $9, %r10
-+ sub $9, %cl
++/* React to a change in the target key */
++static void watch_target(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ unsigned long long new_target;
++ int err;
+
-+ movq %r11, %r9
-+ andq %r10, %r9
-+ shrq %cl, %r9
++ err = xenbus_scanf(XBT_NIL, "memory", "target", "%llu", &new_target);
++ if (err != 1) {
++ /* This is ok (for domain0 at least) - so just return */
++ return;
++ }
+
-+ movq PTR(VA_PUD_1)(%rsi), %r8
-+ addq %r8, %r9
-+ movq PTR(PA_PMD_1)(%rsi), %r8
-+ orq $PAGE_ATTR, %r8
-+ movq %r8, (%r9)
++ /* The given memory/target value is in KiB, so it needs converting to
++ * pages. PAGE_SHIFT converts bytes to pages, hence PAGE_SHIFT - 10.
++ */
++ balloon_set_new_target(new_target >> (PAGE_SHIFT - 10));
++}
+
-+ shrq $9, %r10
-+ sub $9, %cl
++static int balloon_init_watcher(struct notifier_block *notifier,
++ unsigned long event,
++ void *data)
++{
++ int err;
+
-+ movq %r11, %r9
-+ andq %r10, %r9
-+ shrq %cl, %r9
++ err = register_xenbus_watch(&target_watch);
++ if (err)
++ printk(KERN_ERR "Failed to set balloon watcher\n");
+
-+ movq PTR(VA_PMD_1)(%rsi), %r8
-+ addq %r8, %r9
-+ movq PTR(PA_PTE_1)(%rsi), %r8
-+ orq $PAGE_ATTR, %r8
-+ movq %r8, (%r9)
++ return NOTIFY_DONE;
++}
+
-+ shrq $9, %r10
-+ sub $9, %cl
++#ifdef CONFIG_PROC_FS
++static int balloon_write(struct file *file, const char __user *buffer,
++ unsigned long count, void *data)
++{
++ char memstring[64], *endchar;
++ unsigned long long target_bytes;
+
-+ movq %r11, %r9
-+ andq %r10, %r9
-+ shrq %cl, %r9
++ if (!capable(CAP_SYS_ADMIN))
++ return -EPERM;
+
-+ movq PTR(VA_PTE_1)(%rsi), %r8
-+ addq %r8, %r9
-+ movq PTR(PA_CONTROL_PAGE)(%rsi), %r8
-+ orq $PAGE_ATTR, %r8
-+ movq %r8, (%r9)
++ if (count <= 1)
++ return -EBADMSG; /* runt */
++ if (count > sizeof(memstring))
++ return -EFBIG; /* too long */
+
- relocate_new_kernel:
-- /* %rdi page_list
-- * %rsi reboot_code_buffer
-+ /* %rdi indirection_page
-+ * %rsi page_list
- * %rdx start address
-- * %rcx page_table
-- * %r8 arg5
-- * %r9 arg6
- */
-
- /* zero out flags, and disable interrupts */
- pushq $0
- popfq
-
-- /* set a new stack at the bottom of our page... */
-- lea 4096(%rsi), %rsp
--
-- /* store the parameters back on the stack */
-- pushq %rdx /* store the start address */
-+ /* get physical address of control page now */
-+ /* this is impossible after page table switch */
-+ movq PTR(PA_CONTROL_PAGE)(%rsi), %r8
++ if (copy_from_user(memstring, buffer, count))
++ return -EFAULT;
++ memstring[sizeof(memstring)-1] = '\0';
+
-+ /* get physical address of page table now too */
-+ movq PTR(PA_TABLE_PAGE)(%rsi), %rcx
++ target_bytes = memparse(memstring, &endchar);
++ balloon_set_new_target(target_bytes >> PAGE_SHIFT);
+
-+ /* switch to new set of page tables */
-+ movq PTR(PA_PGD)(%rsi), %r9
-+ movq %r9, %cr3
++ return count;
++}
+
-+ /* setup idt */
-+ movq %r8, %rax
-+ addq $(idt_80 - relocate_kernel), %rax
-+ lidtq (%rax)
++static int balloon_read(char *page, char **start, off_t off,
++ int count, int *eof, void *data)
++{
++ int len;
+
-+ /* setup gdt */
-+ movq %r8, %rax
-+ addq $(gdt - relocate_kernel), %rax
-+ movq %r8, %r9
-+ addq $((gdt_80 - relocate_kernel) + 2), %r9
-+ movq %rax, (%r9)
++ len = sprintf(
++ page,
++ "Current allocation: %8lu kB\n"
++ "Requested target: %8lu kB\n"
++ "Low-mem balloon: %8lu kB\n"
++ "High-mem balloon: %8lu kB\n"
++ "Driver pages: %8lu kB\n"
++ "Xen hard limit: ",
++ PAGES2KB(bs.current_pages), PAGES2KB(bs.target_pages),
++ PAGES2KB(bs.balloon_low), PAGES2KB(bs.balloon_high),
++ PAGES2KB(bs.driver_pages));
+
-+ movq %r8, %rax
-+ addq $(gdt_80 - relocate_kernel), %rax
-+ lgdtq (%rax)
++ if (bs.hard_limit != ~0UL)
++ len += sprintf(page + len, "%8lu kB\n",
++ PAGES2KB(bs.hard_limit));
++ else
++ len += sprintf(page + len, " ??? kB\n");
+
-+ /* setup data segment registers */
-+ xorl %eax, %eax
-+ movl %eax, %ds
-+ movl %eax, %es
-+ movl %eax, %fs
-+ movl %eax, %gs
-+ movl %eax, %ss
-+
-+ /* setup a new stack at the end of the physical control page */
-+ lea 4096(%r8), %rsp
++ *eof = 1;
++ return len;
++}
++#endif
+
-+ /* load new code segment and jump to identity mapped page */
-+ movq %r8, %rax
-+ addq $(identity_mapped - relocate_kernel), %rax
-+ pushq $(gdt_cs - gdt)
-+ pushq %rax
-+ lretq
++static struct notifier_block xenstore_notifier;
+
-+identity_mapped:
-+ /* store the start address on the stack */
-+ pushq %rdx
-
- /* Set cr0 to a known state:
- * 31 1 == Paging enabled
-@@ -134,10 +298,19 @@ relocate_new_kernel:
- xorq %r13, %r13
- xorq %r14, %r14
- xorq %r15, %r15
--
- ret
--relocate_new_kernel_end:
-
-- .globl relocate_new_kernel_size
--relocate_new_kernel_size:
-- .quad relocate_new_kernel_end - relocate_new_kernel
-+ .align 16
-+gdt:
-+ .quad 0x0000000000000000 /* NULL descriptor */
-+gdt_cs:
-+ .quad 0x00af9a000000ffff
-+gdt_end:
++static int __init balloon_init(void)
++{
++#if defined(CONFIG_X86) && defined(CONFIG_XEN)
++ unsigned long pfn;
++ struct page *page;
++#endif
+
-+gdt_80:
-+ .word gdt_end - gdt - 1 /* limit */
-+ .quad 0 /* base - filled in by code above */
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ IPRINTK("Initialising balloon driver.\n");
+
-+idt_80:
-+ .word 0 /* limit */
-+ .quad 0 /* base */
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/setup64-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/setup64-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/setup64-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/setup64-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,361 @@
-+/*
-+ * X86-64 specific CPU setup.
-+ * Copyright (C) 1995 Linus Torvalds
-+ * Copyright 2001, 2002, 2003 SuSE Labs / Andi Kleen.
-+ * See setup.c for older changelog.
-+ *
-+ * Jun Nakajima <jun.nakajima at intel.com>
-+ * Modified for Xen
-+ *
-+ */
-+#include <linux/init.h>
-+#include <linux/kernel.h>
-+#include <linux/sched.h>
-+#include <linux/string.h>
-+#include <linux/bootmem.h>
-+#include <linux/bitops.h>
-+#include <linux/module.h>
-+#include <asm/bootsetup.h>
-+#include <asm/pda.h>
-+#include <asm/pgtable.h>
-+#include <asm/processor.h>
-+#include <asm/desc.h>
-+#include <asm/atomic.h>
-+#include <asm/mmu_context.h>
-+#include <asm/smp.h>
-+#include <asm/i387.h>
-+#include <asm/percpu.h>
-+#include <asm/proto.h>
-+#include <asm/sections.h>
+#ifdef CONFIG_XEN
-+#include <asm/hypervisor.h>
++ bs.current_pages = min(xen_start_info->nr_pages, max_pfn);
++ totalram_pages = bs.current_pages;
++#else
++ bs.current_pages = totalram_pages;
++#endif
++ bs.target_pages = bs.current_pages;
++ bs.balloon_low = 0;
++ bs.balloon_high = 0;
++ bs.driver_pages = 0UL;
++ bs.hard_limit = ~0UL;
++
++ init_timer(&balloon_timer);
++ balloon_timer.data = 0;
++ balloon_timer.function = balloon_alarm;
++
++#ifdef CONFIG_PROC_FS
++ if ((balloon_pde = create_xen_proc_entry("balloon", 0644)) == NULL) {
++ WPRINTK("Unable to create /proc/xen/balloon.\n");
++ return -1;
++ }
++
++ balloon_pde->read_proc = balloon_read;
++ balloon_pde->write_proc = balloon_write;
++#endif
++ balloon_sysfs_init();
++
++#if defined(CONFIG_X86) && defined(CONFIG_XEN)
++ /* Initialise the balloon with excess memory space. */
++ for (pfn = xen_start_info->nr_pages; pfn < max_pfn; pfn++) {
++ page = pfn_to_page(pfn);
++ if (!PageReserved(page))
++ balloon_append(page);
++ }
+#endif
+
-+char x86_boot_params[BOOT_PARAM_SIZE] __initdata = {0,};
++ target_watch.callback = watch_target;
++ xenstore_notifier.notifier_call = balloon_init_watcher;
++
++ register_xenstore_notifier(&xenstore_notifier);
++
++ return 0;
++}
+
-+cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE;
++subsys_initcall(balloon_init);
+
-+struct x8664_pda *_cpu_pda[NR_CPUS] __read_mostly;
-+EXPORT_SYMBOL(_cpu_pda);
-+struct x8664_pda boot_cpu_pda[NR_CPUS] __cacheline_aligned;
++static void __exit balloon_exit(void)
++{
++ /* XXX - release balloon here */
++ return;
++}
+
-+#ifndef CONFIG_X86_NO_IDT
-+struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table };
-+#endif
++module_exit(balloon_exit);
+
-+char boot_cpu_stack[IRQSTACKSIZE] __attribute__((section(".bss.page_aligned")));
++void balloon_update_driver_allowance(long delta)
++{
++ unsigned long flags;
+
-+unsigned long __supported_pte_mask __read_mostly = ~0UL;
-+EXPORT_SYMBOL(__supported_pte_mask);
-+static int do_not_nx __cpuinitdata = 0;
++ balloon_lock(flags);
++ bs.driver_pages += delta;
++ balloon_unlock(flags);
++}
+
-+/* noexec=on|off
-+Control non executable mappings for 64bit processes.
++#ifdef CONFIG_XEN
++static int dealloc_pte_fn(
++ pte_t *pte, struct page *pmd_page, unsigned long addr, void *data)
++{
++ unsigned long mfn = pte_mfn(*pte);
++ int ret;
++ struct xen_memory_reservation reservation = {
++ .nr_extents = 1,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
++ set_xen_guest_handle(reservation.extent_start, &mfn);
++ set_pte_at(&init_mm, addr, pte, __pte_ma(0));
++ set_phys_to_machine(__pa(addr) >> PAGE_SHIFT, INVALID_P2M_ENTRY);
++ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation, &reservation);
++ BUG_ON(ret != 1);
++ return 0;
++}
++#endif
+
-+on Enable(default)
-+off Disable
-+*/
-+int __init nonx_setup(char *str)
++struct page **alloc_empty_pages_and_pagevec(int nr_pages)
+{
-+ if (!strncmp(str, "on", 2)) {
-+ __supported_pte_mask |= _PAGE_NX;
-+ do_not_nx = 0;
-+ } else if (!strncmp(str, "off", 3)) {
-+ do_not_nx = 1;
-+ __supported_pte_mask &= ~_PAGE_NX;
-+ }
-+ return 1;
-+}
-+__setup("noexec=", nonx_setup); /* parsed early actually */
++ unsigned long flags;
++ void *v;
++ struct page *page, **pagevec;
++ int i, ret;
+
-+int force_personality32 = 0;
++ pagevec = kmalloc(sizeof(page) * nr_pages, GFP_KERNEL);
++ if (pagevec == NULL)
++ return NULL;
+
-+/* noexec32=on|off
-+Control non executable heap for 32bit processes.
-+To control the stack too use noexec=off
++ for (i = 0; i < nr_pages; i++) {
++ page = pagevec[i] = alloc_page(GFP_KERNEL|__GFP_COLD);
++ if (page == NULL)
++ goto err;
+
-+on PROT_READ does not imply PROT_EXEC for 32bit processes
-+off PROT_READ implies PROT_EXEC (default)
-+*/
-+static int __init nonx32_setup(char *str)
-+{
-+ if (!strcmp(str, "on"))
-+ force_personality32 &= ~READ_IMPLIES_EXEC;
-+ else if (!strcmp(str, "off"))
-+ force_personality32 |= READ_IMPLIES_EXEC;
-+ return 1;
-+}
-+__setup("noexec32=", nonx32_setup);
++ v = page_address(page);
++ scrub_pages(v, 1);
+
-+/*
-+ * Great future plan:
-+ * Declare PDA itself and support (irqstack,tss,pgd) as per cpu data.
-+ * Always point %gs to its beginning
-+ */
-+void __init setup_per_cpu_areas(void)
-+{
-+ int i;
-+ unsigned long size;
++ balloon_lock(flags);
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+ prefill_possible_map();
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ unsigned long gmfn = page_to_pfn(page);
++ struct xen_memory_reservation reservation = {
++ .nr_extents = 1,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
++ set_xen_guest_handle(reservation.extent_start, &gmfn);
++ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &reservation);
++ if (ret == 1)
++ ret = 0; /* success */
++ } else {
++#ifdef CONFIG_XEN
++ ret = apply_to_page_range(&init_mm, (unsigned long)v,
++ PAGE_SIZE, dealloc_pte_fn,
++ NULL);
++#else
++ /* Cannot handle non-auto translate mode. */
++ ret = 1;
+#endif
++ }
+
-+ /* Copy section for each CPU (we discard the original) */
-+ size = ALIGN(__per_cpu_end - __per_cpu_start, SMP_CACHE_BYTES);
-+#ifdef CONFIG_MODULES
-+ if (size < PERCPU_ENOUGH_ROOM)
-+ size = PERCPU_ENOUGH_ROOM;
-+#endif
++ if (ret != 0) {
++ balloon_unlock(flags);
++ balloon_free_page(page);
++ goto err;
++ }
+
-+ for_each_cpu_mask (i, cpu_possible_map) {
-+ char *ptr;
++ totalram_pages = --bs.current_pages;
+
-+ if (!NODE_DATA(cpu_to_node(i))) {
-+ printk("cpu with no node %d, num_online_nodes %d\n",
-+ i, num_online_nodes());
-+ ptr = alloc_bootmem(size);
-+ } else {
-+ ptr = alloc_bootmem_node(NODE_DATA(cpu_to_node(i)), size);
-+ }
-+ if (!ptr)
-+ panic("Cannot allocate cpu data for CPU %d\n", i);
-+ cpu_pda(i)->data_offset = ptr - __per_cpu_start;
-+ memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start);
++ balloon_unlock(flags);
+ }
-+}
+
++ out:
++ schedule_work(&balloon_worker);
+#ifdef CONFIG_XEN
-+static void switch_pt(void)
-+{
-+ xen_pt_switch(__pa(init_level4_pgt));
-+ xen_new_user_pt(__pa(init_level4_user_pgt));
++ flush_tlb_all();
++#endif
++ return pagevec;
++
++ err:
++ balloon_lock(flags);
++ while (--i >= 0)
++ balloon_append(pagevec[i]);
++ balloon_unlock(flags);
++ kfree(pagevec);
++ pagevec = NULL;
++ goto out;
+}
+
-+void __cpuinit cpu_gdt_init(struct desc_ptr *gdt_descr)
++void free_empty_pages_and_pagevec(struct page **pagevec, int nr_pages)
+{
-+ unsigned long frames[16];
-+ unsigned long va;
-+ int f;
++ unsigned long flags;
++ int i;
+
-+ for (va = gdt_descr->address, f = 0;
-+ va < gdt_descr->address + gdt_descr->size;
-+ va += PAGE_SIZE, f++) {
-+ frames[f] = virt_to_mfn(va);
-+ make_page_readonly(
-+ (void *)va, XENFEAT_writable_descriptor_tables);
++ if (pagevec == NULL)
++ return;
++
++ balloon_lock(flags);
++ for (i = 0; i < nr_pages; i++) {
++ BUG_ON(page_count(pagevec[i]) != 1);
++ balloon_append(pagevec[i]);
+ }
-+ if (HYPERVISOR_set_gdt(frames, gdt_descr->size /
-+ sizeof (struct desc_struct)))
-+ BUG();
-+}
-+#else
-+static void switch_pt(void)
-+{
-+ asm volatile("movq %0,%%cr3" :: "r" (__pa_symbol(&init_level4_pgt)));
++ balloon_unlock(flags);
++
++ kfree(pagevec);
++
++ schedule_work(&balloon_worker);
+}
+
-+void __init cpu_gdt_init(struct desc_ptr *gdt_descr)
++void balloon_release_driver_page(struct page *page)
+{
-+ asm volatile("lgdt %0" :: "m" (*gdt_descr));
-+ asm volatile("lidt %0" :: "m" (idt_descr));
++ unsigned long flags;
++
++ balloon_lock(flags);
++ balloon_append(page);
++ bs.driver_pages--;
++ balloon_unlock(flags);
++
++ schedule_work(&balloon_worker);
+}
-+#endif
+
-+void pda_init(int cpu)
-+{
-+ struct x8664_pda *pda = cpu_pda(cpu);
++EXPORT_SYMBOL_GPL(balloon_update_driver_allowance);
++EXPORT_SYMBOL_GPL(alloc_empty_pages_and_pagevec);
++EXPORT_SYMBOL_GPL(free_empty_pages_and_pagevec);
++EXPORT_SYMBOL_GPL(balloon_release_driver_page);
+
-+ /* Setup up data that may be needed in __get_free_pages early */
-+ asm volatile("movl %0,%%fs ; movl %0,%%gs" :: "r" (0));
-+#ifndef CONFIG_XEN
-+ wrmsrl(MSR_GS_BASE, pda);
-+#else
-+ HYPERVISOR_set_segment_base(SEGBASE_GS_KERNEL, (unsigned long)pda);
-+#endif
-+ pda->cpunumber = cpu;
-+ pda->irqcount = -1;
-+ pda->kernelstack =
-+ (unsigned long)stack_thread_info() - PDA_STACKOFFSET + THREAD_SIZE;
-+ pda->active_mm = &init_mm;
-+ pda->mmu_state = 0;
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/balloon/common.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/balloon/common.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,58 @@
++/******************************************************************************
++ * balloon/common.h
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ if (cpu == 0) {
-+#ifdef CONFIG_XEN
-+ xen_init_pt();
-+#endif
-+ /* others are initialized in smpboot.c */
-+ pda->pcurrent = &init_task;
-+ pda->irqstackptr = boot_cpu_stack;
-+ } else {
-+ pda->irqstackptr = (char *)
-+ __get_free_pages(GFP_ATOMIC, IRQSTACK_ORDER);
-+ if (!pda->irqstackptr)
-+ panic("cannot allocate irqstack for cpu %d", cpu);
-+ }
++#ifndef __XEN_BALLOON_COMMON_H__
++#define __XEN_BALLOON_COMMON_H__
+
-+ switch_pt();
++#define PAGES2KB(_p) ((_p)<<(PAGE_SHIFT-10))
+
-+ pda->irqstackptr += IRQSTACKSIZE-64;
-+}
++struct balloon_stats {
++ /* We aim for 'current allocation' == 'target allocation'. */
++ unsigned long current_pages;
++ unsigned long target_pages;
++ /* We may hit the hard limit in Xen. If we do then we remember it. */
++ unsigned long hard_limit;
++ /*
++ * Drivers may alter the memory reservation independently, but they
++ * must inform the balloon driver so we avoid hitting the hard limit.
++ */
++ unsigned long driver_pages;
++ /* Number of pages in high- and low-memory balloons. */
++ unsigned long balloon_low;
++ unsigned long balloon_high;
++};
+
-+#ifndef CONFIG_X86_NO_TSS
-+char boot_exception_stacks[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + DEBUG_STKSZ]
-+__attribute__((section(".bss.page_aligned")));
-+#endif
++extern struct balloon_stats balloon_stats;
++#define bs balloon_stats
+
-+/* May not be marked __init: used by software suspend */
-+void syscall_init(void)
-+{
-+#ifndef CONFIG_XEN
-+ /*
-+ * LSTAR and STAR live in a bit strange symbiosis.
-+ * They both write to the same internal register. STAR allows to set CS/DS
-+ * but only a 32bit target. LSTAR sets the 64bit rip.
-+ */
-+ wrmsrl(MSR_STAR, ((u64)__USER32_CS)<<48 | ((u64)__KERNEL_CS)<<32);
-+ wrmsrl(MSR_LSTAR, system_call);
++int balloon_sysfs_init(void);
++void balloon_sysfs_exit(void);
+
-+ /* Flags to clear on syscall */
-+ wrmsrl(MSR_SYSCALL_MASK, EF_TF|EF_DF|EF_IE|0x3000);
-+#endif
-+#ifdef CONFIG_IA32_EMULATION
-+ syscall32_cpu_init ();
++void balloon_set_new_target(unsigned long target);
++
++#endif /* __XEN_BALLOON_COMMON_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/balloon/sysfs.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/balloon/sysfs.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,170 @@
++/******************************************************************************
++ * balloon/sysfs.c
++ *
++ * Xen balloon driver - sysfs interfaces.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/capability.h>
++#include <linux/errno.h>
++#include <linux/stat.h>
++#include <linux/string.h>
++#include <linux/sysdev.h>
++#include "common.h"
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
+#endif
-+}
+
-+void __cpuinit check_efer(void)
-+{
-+ unsigned long efer;
++#define BALLOON_CLASS_NAME "xen_memory"
+
-+ rdmsrl(MSR_EFER, efer);
-+ if (!(efer & EFER_NX) || do_not_nx) {
-+ __supported_pte_mask &= ~_PAGE_NX;
-+ }
++#define BALLOON_SHOW(name, format, args...) \
++ static ssize_t show_##name(struct sys_device *dev, \
++ char *buf) \
++ { \
++ return sprintf(buf, format, ##args); \
++ } \
++ static SYSDEV_ATTR(name, S_IRUGO, show_##name, NULL)
++
++BALLOON_SHOW(current_kb, "%lu\n", PAGES2KB(bs.current_pages));
++BALLOON_SHOW(low_kb, "%lu\n", PAGES2KB(bs.balloon_low));
++BALLOON_SHOW(high_kb, "%lu\n", PAGES2KB(bs.balloon_high));
++BALLOON_SHOW(hard_limit_kb,
++ (bs.hard_limit!=~0UL) ? "%lu\n" : "???\n",
++ (bs.hard_limit!=~0UL) ? PAGES2KB(bs.hard_limit) : 0);
++BALLOON_SHOW(driver_kb, "%lu\n", PAGES2KB(bs.driver_pages));
++
++static ssize_t show_target_kb(struct sys_device *dev, char *buf)
++{
++ return sprintf(buf, "%lu\n", PAGES2KB(bs.target_pages));
+}
+
-+/*
-+ * cpu_init() initializes state that is per-CPU. Some data is already
-+ * initialized (naturally) in the bootstrap process, such as the GDT
-+ * and IDT. We reload them nevertheless, this function acts as a
-+ * 'CPU state barrier', nothing should get across.
-+ * A lot of state is already set up in PDA init.
-+ */
-+void __cpuinit cpu_init (void)
++static ssize_t store_target_kb(struct sys_device *dev,
++ const char *buf,
++ size_t count)
+{
-+ int cpu = stack_smp_processor_id();
-+#ifndef CONFIG_X86_NO_TSS
-+ struct tss_struct *t = &per_cpu(init_tss, cpu);
-+ struct orig_ist *orig_ist = &per_cpu(orig_ist, cpu);
-+ unsigned long v;
-+ char *estacks = NULL;
-+ unsigned i;
-+#endif
-+ struct task_struct *me;
++ char memstring[64], *endchar;
++ unsigned long long target_bytes;
+
-+ /* CPU 0 is initialised in head64.c */
-+ if (cpu != 0) {
-+ pda_init(cpu);
-+ zap_low_mappings(cpu);
-+ }
-+#ifndef CONFIG_X86_NO_TSS
-+ else
-+ estacks = boot_exception_stacks;
-+#endif
++ if (!capable(CAP_SYS_ADMIN))
++ return -EPERM;
++
++ if (count <= 1)
++ return -EBADMSG; /* runt */
++ if (count > sizeof(memstring))
++ return -EFBIG; /* too long */
++ strcpy(memstring, buf);
++
++ target_bytes = memparse(memstring, &endchar);
++ balloon_set_new_target(target_bytes >> PAGE_SHIFT);
++
++ return count;
++}
+
-+ me = current;
++static SYSDEV_ATTR(target_kb, S_IRUGO | S_IWUSR,
++ show_target_kb, store_target_kb);
+
-+ if (cpu_test_and_set(cpu, cpu_initialized))
-+ panic("CPU#%d already initialized!\n", cpu);
++static struct sysdev_attribute *balloon_attrs[] = {
++ &attr_target_kb,
++};
+
-+ printk("Initializing CPU#%d\n", cpu);
++static struct attribute *balloon_info_attrs[] = {
++ &attr_current_kb.attr,
++ &attr_low_kb.attr,
++ &attr_high_kb.attr,
++ &attr_hard_limit_kb.attr,
++ &attr_driver_kb.attr,
++ NULL
++};
+
-+ clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE);
++static struct attribute_group balloon_info_group = {
++ .name = "info",
++ .attrs = balloon_info_attrs,
++};
+
-+ /*
-+ * Initialize the per-CPU GDT with the boot GDT,
-+ * and set up the GDT descriptor:
-+ */
-+#ifndef CONFIG_XEN
-+ if (cpu)
-+ memcpy(cpu_gdt(cpu), cpu_gdt_table, GDT_SIZE);
-+#endif
++static struct sysdev_class balloon_sysdev_class = {
++ set_kset_name(BALLOON_CLASS_NAME),
++};
+
-+ cpu_gdt_descr[cpu].size = GDT_SIZE;
-+ cpu_gdt_init(&cpu_gdt_descr[cpu]);
++static struct sys_device balloon_sysdev;
+
-+ memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8);
-+ syscall_init();
++static int register_balloon(struct sys_device *sysdev)
++{
++ int i, error;
+
-+ wrmsrl(MSR_FS_BASE, 0);
-+ wrmsrl(MSR_KERNEL_GS_BASE, 0);
-+ barrier();
++ error = sysdev_class_register(&balloon_sysdev_class);
++ if (error)
++ return error;
+
-+ check_efer();
++ sysdev->id = 0;
++ sysdev->cls = &balloon_sysdev_class;
+
-+#ifndef CONFIG_X86_NO_TSS
-+ /*
-+ * set up and load the per-CPU TSS
-+ */
-+ for (v = 0; v < N_EXCEPTION_STACKS; v++) {
-+ if (cpu) {
-+ static const unsigned int order[N_EXCEPTION_STACKS] = {
-+ [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STACK_ORDER,
-+ [DEBUG_STACK - 1] = DEBUG_STACK_ORDER
-+ };
++ error = sysdev_register(sysdev);
++ if (error) {
++ sysdev_class_unregister(&balloon_sysdev_class);
++ return error;
++ }
+
-+ estacks = (char *)__get_free_pages(GFP_ATOMIC, order[v]);
-+ if (!estacks)
-+ panic("Cannot allocate exception stack %ld %d\n",
-+ v, cpu);
-+ }
-+ switch (v + 1) {
-+#if DEBUG_STKSZ > EXCEPTION_STKSZ
-+ case DEBUG_STACK:
-+ cpu_pda(cpu)->debugstack = (unsigned long)estacks;
-+ estacks += DEBUG_STKSZ;
-+ break;
-+#endif
-+ default:
-+ estacks += EXCEPTION_STKSZ;
-+ break;
-+ }
-+ orig_ist->ist[v] = t->ist[v] = (unsigned long)estacks;
++ for (i = 0; i < ARRAY_SIZE(balloon_attrs); i++) {
++ error = sysdev_create_file(sysdev, balloon_attrs[i]);
++ if (error)
++ goto fail;
+ }
+
-+ t->io_bitmap_base = offsetof(struct tss_struct, io_bitmap);
-+ /*
-+ * <= is required because the CPU will access up to
-+ * 8 bits beyond the end of the IO permission bitmap.
-+ */
-+ for (i = 0; i <= IO_BITMAP_LONGS; i++)
-+ t->io_bitmap[i] = ~0UL;
-+#endif
++ error = sysfs_create_group(&sysdev->kobj, &balloon_info_group);
++ if (error)
++ goto fail;
++
++ return 0;
+
-+ atomic_inc(&init_mm.mm_count);
-+ me->active_mm = &init_mm;
-+ if (me->mm)
-+ BUG();
-+ enter_lazy_tlb(&init_mm, me);
++ fail:
++ while (--i >= 0)
++ sysdev_remove_file(sysdev, balloon_attrs[i]);
++ sysdev_unregister(sysdev);
++ sysdev_class_unregister(&balloon_sysdev_class);
++ return error;
++}
+
-+#ifndef CONFIG_X86_NO_TSS
-+ set_tss_desc(cpu, t);
-+#endif
-+#ifndef CONFIG_XEN
-+ load_TR_desc();
-+#endif
-+ load_LDT(&init_mm.context);
++static void unregister_balloon(struct sys_device *sysdev)
++{
++ int i;
+
-+ /*
-+ * Clear all 6 debug registers:
-+ */
++ sysfs_remove_group(&sysdev->kobj, &balloon_info_group);
++ for (i = 0; i < ARRAY_SIZE(balloon_attrs); i++)
++ sysdev_remove_file(sysdev, balloon_attrs[i]);
++ sysdev_unregister(sysdev);
++ sysdev_class_unregister(&balloon_sysdev_class);
++}
+
-+ set_debugreg(0UL, 0);
-+ set_debugreg(0UL, 1);
-+ set_debugreg(0UL, 2);
-+ set_debugreg(0UL, 3);
-+ set_debugreg(0UL, 6);
-+ set_debugreg(0UL, 7);
++int balloon_sysfs_init(void)
++{
++ return register_balloon(&balloon_sysdev);
++}
+
-+ fpu_init();
++void balloon_sysfs_exit(void)
++{
++ unregister_balloon(&balloon_sysdev);
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/setup-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/setup-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/setup-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/setup-xen.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,1677 @@
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blkback/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blkback/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,3 @@
++obj-$(CONFIG_XEN_BLKDEV_BACKEND) := blkbk.o
++
++blkbk-y := blkback.o xenbus.o interface.o vbd.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blkback/blkback.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blkback/blkback.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,646 @@
++/******************************************************************************
++ * arch/xen/drivers/blkif/backend/main.c
++ *
++ * Back-end of the driver for virtual block devices. This portion of the
++ * driver exports a 'unified' block-device interface that can be accessed
++ * by any operating system that implements a compatible front end. A
++ * reference front-end implementation can be found in:
++ * arch/xen/drivers/blkif/frontend
++ *
++ * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
++ * Copyright (c) 2005, Christopher Clark
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/spinlock.h>
++#include <linux/kthread.h>
++#include <linux/list.h>
++#include <linux/delay.h>
++#include <xen/balloon.h>
++#include <asm/hypervisor.h>
++#include "common.h"
++
+/*
-+ * linux/arch/x86-64/kernel/setup.c
-+ *
-+ * Copyright (C) 1995 Linus Torvalds
-+ *
-+ * Nov 2001 Dave Jones <davej at suse.de>
-+ * Forked from i386 setup code.
++ * These are rather arbitrary. They are fairly large because adjacent requests
++ * pulled from a communication ring are quite likely to end up being part of
++ * the same scatter/gather request at the disc.
++ *
++ * ** TRY INCREASING 'blkif_reqs' IF WRITE SPEEDS SEEM TOO LOW **
++ *
++ * This will increase the chances of being able to write whole tracks.
++ * 64 should be enough to keep us competitive with Linux.
+ */
++static int blkif_reqs = 64;
++module_param_named(reqs, blkif_reqs, int, 0);
++MODULE_PARM_DESC(reqs, "Number of blkback requests to allocate");
++
++/* Run-time switchable: /sys/module/blkback/parameters/ */
++static unsigned int log_stats = 0;
++static unsigned int debug_lvl = 0;
++module_param(log_stats, int, 0644);
++module_param(debug_lvl, int, 0644);
+
+/*
-+ * This file handles the architecture-dependent parts of initialization
++ * Each outstanding request that we've passed to the lower device layers has a
++ * 'pending_req' allocated to it. Each buffer_head that completes decrements
++ * the pendcnt towards zero. When it hits zero, the specified domain has a
++ * response queued for it, with the saved 'id' passed back.
+ */
++typedef struct {
++ blkif_t *blkif;
++ u64 id;
++ int nr_pages;
++ atomic_t pendcnt;
++ unsigned short operation;
++ int status;
++ struct list_head free_list;
++} pending_req_t;
+
-+#include <linux/errno.h>
-+#include <linux/sched.h>
-+#include <linux/kernel.h>
-+#include <linux/mm.h>
-+#include <linux/stddef.h>
-+#include <linux/unistd.h>
-+#include <linux/ptrace.h>
-+#include <linux/slab.h>
-+#include <linux/user.h>
-+#include <linux/a.out.h>
-+#include <linux/screen_info.h>
-+#include <linux/ioport.h>
-+#include <linux/delay.h>
-+#include <linux/init.h>
-+#include <linux/initrd.h>
-+#include <linux/highmem.h>
-+#include <linux/bootmem.h>
-+#include <linux/module.h>
-+#include <asm/processor.h>
-+#include <linux/console.h>
-+#include <linux/seq_file.h>
-+#include <linux/crash_dump.h>
-+#include <linux/root_dev.h>
-+#include <linux/pci.h>
-+#include <linux/acpi.h>
-+#include <linux/kallsyms.h>
-+#include <linux/edd.h>
-+#include <linux/mmzone.h>
-+#include <linux/kexec.h>
-+#include <linux/cpufreq.h>
-+#include <linux/dmi.h>
-+#include <linux/dma-mapping.h>
-+#include <linux/ctype.h>
++static pending_req_t *pending_reqs;
++static struct list_head pending_free;
++static DEFINE_SPINLOCK(pending_free_lock);
++static DECLARE_WAIT_QUEUE_HEAD(pending_free_wq);
+
-+#include <asm/mtrr.h>
-+#include <asm/uaccess.h>
-+#include <asm/system.h>
-+#include <asm/io.h>
-+#include <asm/smp.h>
-+#include <asm/msr.h>
-+#include <asm/desc.h>
-+#include <video/edid.h>
-+#include <asm/e820.h>
-+#include <asm/dma.h>
-+#include <asm/mpspec.h>
-+#include <asm/mmu_context.h>
-+#include <asm/bootsetup.h>
-+#include <asm/proto.h>
-+#include <asm/setup.h>
-+#include <asm/mach_apic.h>
-+#include <asm/numa.h>
-+#include <asm/sections.h>
-+#include <asm/dmi.h>
-+#ifdef CONFIG_XEN
-+#include <linux/percpu.h>
-+#include <xen/interface/physdev.h>
-+#include "setup_arch_pre.h"
-+#include <asm/hypervisor.h>
-+#include <xen/interface/nmi.h>
-+#include <xen/features.h>
-+#include <xen/xencons.h>
-+#define PFN_UP(x) (((x) + PAGE_SIZE-1) >> PAGE_SHIFT)
-+#define PFN_PHYS(x) ((x) << PAGE_SHIFT)
-+#include <asm/mach-xen/setup_arch_post.h>
-+#include <xen/interface/memory.h>
++#define BLKBACK_INVALID_HANDLE (~0)
+
-+#ifdef CONFIG_XEN
-+#include <xen/interface/kexec.h>
-+#endif
++static struct page **pending_pages;
++static grant_handle_t *pending_grant_handles;
+
-+extern unsigned long start_pfn;
-+extern struct edid_info edid_info;
++static inline int vaddr_pagenr(pending_req_t *req, int seg)
++{
++ return (req - pending_reqs) * BLKIF_MAX_SEGMENTS_PER_REQUEST + seg;
++}
+
-+shared_info_t *HYPERVISOR_shared_info = (shared_info_t *)empty_zero_page;
-+EXPORT_SYMBOL(HYPERVISOR_shared_info);
++static inline unsigned long vaddr(pending_req_t *req, int seg)
++{
++ unsigned long pfn = page_to_pfn(pending_pages[vaddr_pagenr(req, seg)]);
++ return (unsigned long)pfn_to_kaddr(pfn);
++}
+
-+extern char hypercall_page[PAGE_SIZE];
-+EXPORT_SYMBOL(hypercall_page);
++#define pending_handle(_req, _seg) \
++ (pending_grant_handles[vaddr_pagenr(_req, _seg)])
+
-+static int xen_panic_event(struct notifier_block *, unsigned long, void *);
-+static struct notifier_block xen_panic_block = {
-+ xen_panic_event, NULL, 0 /* try to go last */
-+};
+
-+unsigned long *phys_to_machine_mapping;
-+unsigned long *pfn_to_mfn_frame_list_list, *pfn_to_mfn_frame_list[512];
++static int do_block_io_op(blkif_t *blkif);
++static void dispatch_rw_block_io(blkif_t *blkif,
++ blkif_request_t *req,
++ pending_req_t *pending_req);
++static void make_response(blkif_t *blkif, u64 id,
++ unsigned short op, int st);
+
-+EXPORT_SYMBOL(phys_to_machine_mapping);
++/******************************************************************
++ * misc small helpers
++ */
++static pending_req_t* alloc_req(void)
++{
++ pending_req_t *req = NULL;
++ unsigned long flags;
+
-+DEFINE_PER_CPU(multicall_entry_t, multicall_list[8]);
-+DEFINE_PER_CPU(int, nr_multicall_ents);
++ spin_lock_irqsave(&pending_free_lock, flags);
++ if (!list_empty(&pending_free)) {
++ req = list_entry(pending_free.next, pending_req_t, free_list);
++ list_del(&req->free_list);
++ }
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++ return req;
++}
+
-+/* Raw start-of-day parameters from the hypervisor. */
-+start_info_t *xen_start_info;
-+EXPORT_SYMBOL(xen_start_info);
-+#endif
++static void free_req(pending_req_t *req)
++{
++ unsigned long flags;
++ int was_empty;
+
-+/*
-+ * Machine setup..
-+ */
++ spin_lock_irqsave(&pending_free_lock, flags);
++ was_empty = list_empty(&pending_free);
++ list_add(&req->free_list, &pending_free);
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++ if (was_empty)
++ wake_up(&pending_free_wq);
++}
+
-+struct cpuinfo_x86 boot_cpu_data __read_mostly;
-+EXPORT_SYMBOL(boot_cpu_data);
++static void unplug_queue(blkif_t *blkif)
++{
++ if (blkif->plug == NULL)
++ return;
++ if (blkif->plug->unplug_fn)
++ blkif->plug->unplug_fn(blkif->plug);
++ blk_put_queue(blkif->plug);
++ blkif->plug = NULL;
++}
+
-+unsigned long mmu_cr4_features;
++static void plug_queue(blkif_t *blkif, struct bio *bio)
++{
++ request_queue_t *q = bdev_get_queue(bio->bi_bdev);
+
-+int acpi_disabled;
-+EXPORT_SYMBOL(acpi_disabled);
-+#ifdef CONFIG_ACPI
-+extern int __initdata acpi_ht;
-+extern acpi_interrupt_flags acpi_sci_flags;
-+int __initdata acpi_force = 0;
-+#endif
++ if (q == blkif->plug)
++ return;
++ unplug_queue(blkif);
++ blk_get_queue(q);
++ blkif->plug = q;
++}
+
-+int acpi_numa __initdata;
++static void fast_flush_area(pending_req_t *req)
++{
++ struct gnttab_unmap_grant_ref unmap[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++ unsigned int i, invcount = 0;
++ grant_handle_t handle;
++ int ret;
+
-+/* Boot loader ID as an integer, for the benefit of proc_dointvec */
-+int bootloader_type;
++ for (i = 0; i < req->nr_pages; i++) {
++ handle = pending_handle(req, i);
++ if (handle == BLKBACK_INVALID_HANDLE)
++ continue;
++ gnttab_set_unmap_op(&unmap[invcount], vaddr(req, i),
++ GNTMAP_host_map, handle);
++ pending_handle(req, i) = BLKBACK_INVALID_HANDLE;
++ invcount++;
++ }
+
-+unsigned long saved_video_mode;
++ ret = HYPERVISOR_grant_table_op(
++ GNTTABOP_unmap_grant_ref, unmap, invcount);
++ BUG_ON(ret);
++}
+
-+/*
-+ * Early DMI memory
++/******************************************************************
++ * SCHEDULER FUNCTIONS
+ */
-+int dmi_alloc_index;
-+char dmi_alloc_data[DMI_MAX_DATA];
+
-+/*
-+ * Setup options
-+ */
-+struct screen_info screen_info;
-+EXPORT_SYMBOL(screen_info);
-+struct sys_desc_table_struct {
-+ unsigned short length;
-+ unsigned char table[0];
-+};
++static void print_stats(blkif_t *blkif)
++{
++ printk(KERN_DEBUG "%s: oo %3d | rd %4d | wr %4d | br %4d\n",
++ current->comm, blkif->st_oo_req,
++ blkif->st_rd_req, blkif->st_wr_req, blkif->st_br_req);
++ blkif->st_print = jiffies + msecs_to_jiffies(10 * 1000);
++ blkif->st_rd_req = 0;
++ blkif->st_wr_req = 0;
++ blkif->st_oo_req = 0;
++}
+
-+struct edid_info edid_info;
-+EXPORT_SYMBOL_GPL(edid_info);
-+struct e820map e820;
-+#ifdef CONFIG_XEN
-+struct e820map machine_e820;
-+#endif
++int blkif_schedule(void *arg)
++{
++ blkif_t *blkif = arg;
+
-+extern int root_mountflags;
++ blkif_get(blkif);
+
-+char command_line[COMMAND_LINE_SIZE];
++ if (debug_lvl)
++ printk(KERN_DEBUG "%s: started\n", current->comm);
+
-+struct resource standard_io_resources[] = {
-+ { .name = "dma1", .start = 0x00, .end = 0x1f,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
-+ { .name = "pic1", .start = 0x20, .end = 0x21,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
-+ { .name = "timer0", .start = 0x40, .end = 0x43,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
-+ { .name = "timer1", .start = 0x50, .end = 0x53,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
-+ { .name = "keyboard", .start = 0x60, .end = 0x6f,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
-+ { .name = "dma page reg", .start = 0x80, .end = 0x8f,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
-+ { .name = "pic2", .start = 0xa0, .end = 0xa1,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
-+ { .name = "dma2", .start = 0xc0, .end = 0xdf,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
-+ { .name = "fpu", .start = 0xf0, .end = 0xff,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO }
-+};
++ while (!kthread_should_stop()) {
++ if (try_to_freeze())
++ continue;
+
-+#define STANDARD_IO_RESOURCES \
-+ (sizeof standard_io_resources / sizeof standard_io_resources[0])
++ wait_event_interruptible(
++ blkif->wq,
++ blkif->waiting_reqs || kthread_should_stop());
++ wait_event_interruptible(
++ pending_free_wq,
++ !list_empty(&pending_free) || kthread_should_stop());
+
-+#define IORESOURCE_RAM (IORESOURCE_BUSY | IORESOURCE_MEM)
++ blkif->waiting_reqs = 0;
++ smp_mb(); /* clear flag *before* checking for work */
+
-+struct resource data_resource = {
-+ .name = "Kernel data",
-+ .start = 0,
-+ .end = 0,
-+ .flags = IORESOURCE_RAM,
-+};
-+struct resource code_resource = {
-+ .name = "Kernel code",
-+ .start = 0,
-+ .end = 0,
-+ .flags = IORESOURCE_RAM,
-+};
++ if (do_block_io_op(blkif))
++ blkif->waiting_reqs = 1;
++ unplug_queue(blkif);
+
-+#define IORESOURCE_ROM (IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM)
++ if (log_stats && time_after(jiffies, blkif->st_print))
++ print_stats(blkif);
++ }
+
-+static struct resource system_rom_resource = {
-+ .name = "System ROM",
-+ .start = 0xf0000,
-+ .end = 0xfffff,
-+ .flags = IORESOURCE_ROM,
-+};
++ if (log_stats)
++ print_stats(blkif);
++ if (debug_lvl)
++ printk(KERN_DEBUG "%s: exiting\n", current->comm);
+
-+static struct resource extension_rom_resource = {
-+ .name = "Extension ROM",
-+ .start = 0xe0000,
-+ .end = 0xeffff,
-+ .flags = IORESOURCE_ROM,
-+};
++ blkif->xenblkd = NULL;
++ blkif_put(blkif);
+
-+static struct resource adapter_rom_resources[] = {
-+ { .name = "Adapter ROM", .start = 0xc8000, .end = 0,
-+ .flags = IORESOURCE_ROM },
-+ { .name = "Adapter ROM", .start = 0, .end = 0,
-+ .flags = IORESOURCE_ROM },
-+ { .name = "Adapter ROM", .start = 0, .end = 0,
-+ .flags = IORESOURCE_ROM },
-+ { .name = "Adapter ROM", .start = 0, .end = 0,
-+ .flags = IORESOURCE_ROM },
-+ { .name = "Adapter ROM", .start = 0, .end = 0,
-+ .flags = IORESOURCE_ROM },
-+ { .name = "Adapter ROM", .start = 0, .end = 0,
-+ .flags = IORESOURCE_ROM }
-+};
++ return 0;
++}
+
-+#define ADAPTER_ROM_RESOURCES \
-+ (sizeof adapter_rom_resources / sizeof adapter_rom_resources[0])
++/******************************************************************
++ * COMPLETION CALLBACK -- Called as bh->b_end_io()
++ */
+
-+static struct resource video_rom_resource = {
-+ .name = "Video ROM",
-+ .start = 0xc0000,
-+ .end = 0xc7fff,
-+ .flags = IORESOURCE_ROM,
-+};
++static void __end_block_io_op(pending_req_t *pending_req, int error)
++{
++ /* An error fails the entire request. */
++ if ((pending_req->operation == BLKIF_OP_WRITE_BARRIER) &&
++ (error == -EOPNOTSUPP)) {
++ DPRINTK("blkback: write barrier op failed, not supported\n");
++ blkback_barrier(XBT_NIL, pending_req->blkif->be, 0);
++ pending_req->status = BLKIF_RSP_EOPNOTSUPP;
++ } else if (error) {
++ DPRINTK("Buffer not up-to-date at end of operation, "
++ "error=%d\n", error);
++ pending_req->status = BLKIF_RSP_ERROR;
++ }
+
-+static struct resource video_ram_resource = {
-+ .name = "Video RAM area",
-+ .start = 0xa0000,
-+ .end = 0xbffff,
-+ .flags = IORESOURCE_RAM,
-+};
++ if (atomic_dec_and_test(&pending_req->pendcnt)) {
++ fast_flush_area(pending_req);
++ make_response(pending_req->blkif, pending_req->id,
++ pending_req->operation, pending_req->status);
++ blkif_put(pending_req->blkif);
++ free_req(pending_req);
++ }
++}
+
-+#define romsignature(x) (*(unsigned short *)(x) == 0xaa55)
++static int end_block_io_op(struct bio *bio, unsigned int done, int error)
++{
++ if (bio->bi_size != 0)
++ return 1;
++ __end_block_io_op(bio->bi_private, error);
++ bio_put(bio);
++ return error;
++}
++
++
++/******************************************************************************
++ * NOTIFICATION FROM GUEST OS.
++ */
+
-+static int __init romchecksum(unsigned char *rom, unsigned long length)
++static void blkif_notify_work(blkif_t *blkif)
+{
-+ unsigned char *p, sum = 0;
-+
-+ for (p = rom; p < rom + length; p++)
-+ sum += *p;
-+ return sum == 0;
++ blkif->waiting_reqs = 1;
++ wake_up(&blkif->wq);
+}
+
-+static void __init probe_roms(void)
++irqreturn_t blkif_be_int(int irq, void *dev_id, struct pt_regs *regs)
+{
-+ unsigned long start, length, upper;
-+ unsigned char *rom;
-+ int i;
-+
-+#ifdef CONFIG_XEN
-+ /* Nothing to do if not running in dom0. */
-+ if (!is_initial_xendomain())
-+ return;
-+#endif
++ blkif_notify_work(dev_id);
++ return IRQ_HANDLED;
++}
+
-+ /* video rom */
-+ upper = adapter_rom_resources[0].start;
-+ for (start = video_rom_resource.start; start < upper; start += 2048) {
-+ rom = isa_bus_to_virt(start);
-+ if (!romsignature(rom))
-+ continue;
+
-+ video_rom_resource.start = start;
+
-+ /* 0 < length <= 0x7f * 512, historically */
-+ length = rom[2] * 512;
++/******************************************************************
++ * DOWNWARD CALLS -- These interface with the block-device layer proper.
++ */
+
-+ /* if checksum okay, trust length byte */
-+ if (length && romchecksum(rom, length))
-+ video_rom_resource.end = start + length - 1;
++static int do_block_io_op(blkif_t *blkif)
++{
++ blkif_back_rings_t *blk_rings = &blkif->blk_rings;
++ blkif_request_t req;
++ pending_req_t *pending_req;
++ RING_IDX rc, rp;
++ int more_to_do = 0;
+
-+ request_resource(&iomem_resource, &video_rom_resource);
-+ break;
-+ }
++ rc = blk_rings->common.req_cons;
++ rp = blk_rings->common.sring->req_prod;
++ rmb(); /* Ensure we see queued requests up to 'rp'. */
+
-+ start = (video_rom_resource.end + 1 + 2047) & ~2047UL;
-+ if (start < upper)
-+ start = upper;
++ while (rc != rp) {
+
-+ /* system rom */
-+ request_resource(&iomem_resource, &system_rom_resource);
-+ upper = system_rom_resource.start;
++ if (RING_REQUEST_CONS_OVERFLOW(&blk_rings->common, rc))
++ break;
+
-+ /* check for extension rom (ignore length byte!) */
-+ rom = isa_bus_to_virt(extension_rom_resource.start);
-+ if (romsignature(rom)) {
-+ length = extension_rom_resource.end - extension_rom_resource.start + 1;
-+ if (romchecksum(rom, length)) {
-+ request_resource(&iomem_resource, &extension_rom_resource);
-+ upper = extension_rom_resource.start;
++ pending_req = alloc_req();
++ if (NULL == pending_req) {
++ blkif->st_oo_req++;
++ more_to_do = 1;
++ break;
+ }
-+ }
+
-+ /* check for adapter roms on 2k boundaries */
-+ for (i = 0; i < ADAPTER_ROM_RESOURCES && start < upper; start += 2048) {
-+ rom = isa_bus_to_virt(start);
-+ if (!romsignature(rom))
-+ continue;
++ if (kthread_should_stop()) {
++ more_to_do = 1;
++ break;
++ }
+
-+ /* 0 < length <= 0x7f * 512, historically */
-+ length = rom[2] * 512;
++ switch (blkif->blk_protocol) {
++ case BLKIF_PROTOCOL_NATIVE:
++ memcpy(&req, RING_GET_REQUEST(&blk_rings->native, rc), sizeof(req));
++ break;
++ case BLKIF_PROTOCOL_X86_32:
++ blkif_get_x86_32_req(&req, RING_GET_REQUEST(&blk_rings->x86_32, rc));
++ break;
++ case BLKIF_PROTOCOL_X86_64:
++ blkif_get_x86_64_req(&req, RING_GET_REQUEST(&blk_rings->x86_64, rc));
++ break;
++ default:
++ BUG();
++ }
++ blk_rings->common.req_cons = ++rc; /* before make_response() */
+
-+ /* but accept any length that fits if checksum okay */
-+ if (!length || start + length > upper || !romchecksum(rom, length))
-+ continue;
++ /* Apply all sanity checks to /private copy/ of request. */
++ barrier();
+
-+ adapter_rom_resources[i].start = start;
-+ adapter_rom_resources[i].end = start + length - 1;
-+ request_resource(&iomem_resource, &adapter_rom_resources[i]);
++ switch (req.operation) {
++ case BLKIF_OP_READ:
++ blkif->st_rd_req++;
++ dispatch_rw_block_io(blkif, &req, pending_req);
++ break;
++ case BLKIF_OP_WRITE_BARRIER:
++ blkif->st_br_req++;
++ /* fall through */
++ case BLKIF_OP_WRITE:
++ blkif->st_wr_req++;
++ dispatch_rw_block_io(blkif, &req, pending_req);
++ break;
++ default:
++ /* A good sign something is wrong: sleep for a while to
++ * avoid excessive CPU consumption by a bad guest. */
++ msleep(1);
++ DPRINTK("error: unknown block io operation [%d]\n",
++ req.operation);
++ make_response(blkif, req.id, req.operation,
++ BLKIF_RSP_ERROR);
++ free_req(pending_req);
++ break;
++ }
+
-+ start = adapter_rom_resources[i++].end & ~2047UL;
++ /* Yield point for this unbounded loop. */
++ cond_resched();
+ }
-+}
+
-+/* Check for full argument with no trailing characters */
-+static int fullarg(char *p, char *arg)
-+{
-+ int l = strlen(arg);
-+ return !memcmp(p, arg, l) && (p[l] == 0 || isspace(p[l]));
++ return more_to_do;
+}
+
-+static __init void parse_cmdline_early (char ** cmdline_p)
++static void dispatch_rw_block_io(blkif_t *blkif,
++ blkif_request_t *req,
++ pending_req_t *pending_req)
+{
-+ char c = ' ', *to = command_line, *from = COMMAND_LINE;
-+ int len = 0;
-+ int userdef = 0;
++ extern void ll_rw_block(int rw, int nr, struct buffer_head * bhs[]);
++ struct gnttab_map_grant_ref map[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++ struct phys_req preq;
++ struct {
++ unsigned long buf; unsigned int nsec;
++ } seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++ unsigned int nseg;
++ struct bio *bio = NULL, *biolist[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++ int ret, i, nbio = 0;
++ int operation;
+
-+ for (;;) {
-+ if (c != ' ')
-+ goto next_char;
++ switch (req->operation) {
++ case BLKIF_OP_READ:
++ operation = READ;
++ break;
++ case BLKIF_OP_WRITE:
++ operation = WRITE;
++ break;
++ case BLKIF_OP_WRITE_BARRIER:
++ operation = WRITE_BARRIER;
++ break;
++ default:
++ operation = 0; /* make gcc happy */
++ BUG();
++ }
+
-+#ifdef CONFIG_SMP
-+ /*
-+ * If the BIOS enumerates physical processors before logical,
-+ * maxcpus=N at enumeration-time can be used to disable HT.
-+ */
-+ else if (!memcmp(from, "maxcpus=", 8)) {
-+ extern unsigned int maxcpus;
++ /* Check that number of segments is sane. */
++ nseg = req->nr_segments;
++ if (unlikely(nseg == 0 && operation != WRITE_BARRIER) ||
++ unlikely(nseg > BLKIF_MAX_SEGMENTS_PER_REQUEST)) {
++ DPRINTK("Bad number of segments in request (%d)\n", nseg);
++ goto fail_response;
++ }
+
-+ maxcpus = simple_strtoul(from + 8, NULL, 0);
-+ }
-+#endif
-+#ifdef CONFIG_ACPI
-+ /* "acpi=off" disables both ACPI table parsing and interpreter init */
-+ if (fullarg(from,"acpi=off"))
-+ disable_acpi();
++ preq.dev = req->handle;
++ preq.sector_number = req->sector_number;
++ preq.nr_sects = 0;
+
-+ if (fullarg(from, "acpi=force")) {
-+ /* add later when we do DMI horrors: */
-+ acpi_force = 1;
-+ acpi_disabled = 0;
-+ }
++ pending_req->blkif = blkif;
++ pending_req->id = req->id;
++ pending_req->operation = req->operation;
++ pending_req->status = BLKIF_RSP_OKAY;
++ pending_req->nr_pages = nseg;
+
-+ /* acpi=ht just means: do ACPI MADT parsing
-+ at bootup, but don't enable the full ACPI interpreter */
-+ if (fullarg(from, "acpi=ht")) {
-+ if (!acpi_force)
-+ disable_acpi();
-+ acpi_ht = 1;
-+ }
-+ else if (fullarg(from, "pci=noacpi"))
-+ acpi_disable_pci();
-+ else if (fullarg(from, "acpi=noirq"))
-+ acpi_noirq_set();
++ for (i = 0; i < nseg; i++) {
++ uint32_t flags;
+
-+ else if (fullarg(from, "acpi_sci=edge"))
-+ acpi_sci_flags.trigger = 1;
-+ else if (fullarg(from, "acpi_sci=level"))
-+ acpi_sci_flags.trigger = 3;
-+ else if (fullarg(from, "acpi_sci=high"))
-+ acpi_sci_flags.polarity = 1;
-+ else if (fullarg(from, "acpi_sci=low"))
-+ acpi_sci_flags.polarity = 3;
++ seg[i].nsec = req->seg[i].last_sect -
++ req->seg[i].first_sect + 1;
+
-+ /* acpi=strict disables out-of-spec workarounds */
-+ else if (fullarg(from, "acpi=strict")) {
-+ acpi_strict = 1;
-+ }
-+#ifdef CONFIG_X86_IO_APIC
-+ else if (fullarg(from, "acpi_skip_timer_override"))
-+ acpi_skip_timer_override = 1;
-+#endif
-+#endif
++ if ((req->seg[i].last_sect >= (PAGE_SIZE >> 9)) ||
++ (req->seg[i].last_sect < req->seg[i].first_sect))
++ goto fail_response;
++ preq.nr_sects += seg[i].nsec;
+
-+#ifndef CONFIG_XEN
-+ if (fullarg(from, "nolapic") || fullarg(from, "disableapic")) {
-+ clear_bit(X86_FEATURE_APIC, boot_cpu_data.x86_capability);
-+ disable_apic = 1;
-+ }
++ flags = GNTMAP_host_map;
++ if (operation != READ)
++ flags |= GNTMAP_readonly;
++ gnttab_set_map_op(&map[i], vaddr(pending_req, i), flags,
++ req->seg[i].gref, blkif->domid);
++ }
+
-+ if (fullarg(from, "noapic"))
-+ skip_ioapic_setup = 1;
++ ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, map, nseg);
++ BUG_ON(ret);
+
-+ if (fullarg(from,"apic")) {
-+ skip_ioapic_setup = 0;
-+ ioapic_force = 1;
++ for (i = 0; i < nseg; i++) {
++ if (unlikely(map[i].status != 0)) {
++ DPRINTK("invalid buffer -- could not remap it\n");
++ map[i].handle = BLKBACK_INVALID_HANDLE;
++ ret |= 1;
+ }
-+#endif
-+
-+ if (!memcmp(from, "mem=", 4))
-+ parse_memopt(from+4, &from);
+
-+ if (!memcmp(from, "memmap=", 7)) {
-+ /* exactmap option is for used defined memory */
-+ if (!memcmp(from+7, "exactmap", 8)) {
-+#ifdef CONFIG_CRASH_DUMP
-+ /* If we are doing a crash dump, we
-+ * still need to know the real mem
-+ * size before original memory map is
-+ * reset.
-+ */
-+ saved_max_pfn = e820_end_of_ram();
-+#endif
-+ from += 8+7;
-+ end_pfn_map = 0;
-+ e820.nr_map = 0;
-+ userdef = 1;
-+ }
-+ else {
-+ parse_memmapopt(from+7, &from);
-+ userdef = 1;
-+ }
-+ }
++ pending_handle(pending_req, i) = map[i].handle;
+
-+#ifdef CONFIG_NUMA
-+ if (!memcmp(from, "numa=", 5))
-+ numa_setup(from+5);
-+#endif
++ if (ret)
++ continue;
+
-+ if (!memcmp(from,"iommu=",6)) {
-+ iommu_setup(from+6);
-+ }
++ set_phys_to_machine(__pa(vaddr(
++ pending_req, i)) >> PAGE_SHIFT,
++ FOREIGN_FRAME(map[i].dev_bus_addr >> PAGE_SHIFT));
++ seg[i].buf = map[i].dev_bus_addr |
++ (req->seg[i].first_sect << 9);
++ }
+
-+ if (fullarg(from,"oops=panic"))
-+ panic_on_oops = 1;
++ if (ret)
++ goto fail_flush;
+
-+ if (!memcmp(from, "noexec=", 7))
-+ nonx_setup(from + 7);
++ if (vbd_translate(&preq, blkif, operation) != 0) {
++ DPRINTK("access denied: %s of [%llu,%llu] on dev=%04x\n",
++ operation == READ ? "read" : "write",
++ preq.sector_number,
++ preq.sector_number + preq.nr_sects, preq.dev);
++ goto fail_flush;
++ }
+
-+#ifdef CONFIG_KEXEC
-+ /* crashkernel=size at addr specifies the location to reserve for
-+ * a crash kernel. By reserving this memory we guarantee
-+ * that linux never set's it up as a DMA target.
-+ * Useful for holding code to do something appropriate
-+ * after a kernel panic.
-+ */
-+ else if (!memcmp(from, "crashkernel=", 12)) {
-+#ifndef CONFIG_XEN
-+ unsigned long size, base;
-+ size = memparse(from+12, &from);
-+ if (*from == '@') {
-+ base = memparse(from+1, &from);
-+ /* FIXME: Do I want a sanity check
-+ * to validate the memory range?
-+ */
-+ crashk_res.start = base;
-+ crashk_res.end = base + size - 1;
-+ }
-+#else
-+ printk("Ignoring crashkernel command line, "
-+ "parameter will be supplied by xen\n");
-+#endif
++ for (i = 0; i < nseg; i++) {
++ if (((int)preq.sector_number|(int)seg[i].nsec) &
++ ((bdev_hardsect_size(preq.bdev) >> 9) - 1)) {
++ DPRINTK("Misaligned I/O request from domain %d",
++ blkif->domid);
++ goto fail_put_bio;
+ }
-+#endif
+
-+#ifdef CONFIG_PROC_VMCORE
-+ /* elfcorehdr= specifies the location of elf core header
-+ * stored by the crashed kernel. This option will be passed
-+ * by kexec loader to the capture kernel.
-+ */
-+ else if(!memcmp(from, "elfcorehdr=", 11))
-+ elfcorehdr_addr = memparse(from+11, &from);
-+#endif
++ while ((bio == NULL) ||
++ (bio_add_page(bio,
++ virt_to_page(vaddr(pending_req, i)),
++ seg[i].nsec << 9,
++ seg[i].buf & ~PAGE_MASK) == 0)) {
++ bio = biolist[nbio++] = bio_alloc(GFP_KERNEL, nseg-i);
++ if (unlikely(bio == NULL))
++ goto fail_put_bio;
+
-+#if defined(CONFIG_HOTPLUG_CPU) && !defined(CONFIG_XEN)
-+ else if (!memcmp(from, "additional_cpus=", 16))
-+ setup_additional_cpus(from+16);
-+#endif
++ bio->bi_bdev = preq.bdev;
++ bio->bi_private = pending_req;
++ bio->bi_end_io = end_block_io_op;
++ bio->bi_sector = preq.sector_number;
++ }
+
-+ next_char:
-+ c = *(from++);
-+ if (!c)
-+ break;
-+ if (COMMAND_LINE_SIZE <= ++len)
-+ break;
-+ *(to++) = c;
++ preq.sector_number += seg[i].nsec;
+ }
-+ if (userdef) {
-+ printk(KERN_INFO "user-defined physical RAM map:\n");
-+ e820_print_map("user");
++
++ if (!bio) {
++ BUG_ON(operation != WRITE_BARRIER);
++ bio = biolist[nbio++] = bio_alloc(GFP_KERNEL, 0);
++ if (unlikely(bio == NULL))
++ goto fail_put_bio;
++
++ bio->bi_bdev = preq.bdev;
++ bio->bi_private = pending_req;
++ bio->bi_end_io = end_block_io_op;
++ bio->bi_sector = -1;
+ }
-+ *to = '\0';
-+ *cmdline_p = command_line;
-+}
+
-+#ifndef CONFIG_NUMA
-+static void __init
-+contig_initmem_init(unsigned long start_pfn, unsigned long end_pfn)
-+{
-+ unsigned long bootmap_size, bootmap;
++ plug_queue(blkif, bio);
++ atomic_set(&pending_req->pendcnt, nbio);
++ blkif_get(blkif);
+
-+ bootmap_size = bootmem_bootmap_pages(end_pfn)<<PAGE_SHIFT;
-+ bootmap = find_e820_area(0, end_pfn<<PAGE_SHIFT, bootmap_size);
-+ if (bootmap == -1L)
-+ panic("Cannot find bootmem map of size %ld\n",bootmap_size);
-+ bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, end_pfn);
-+#ifdef CONFIG_XEN
-+ e820_bootmem_free(NODE_DATA(0), 0, xen_start_info->nr_pages<<PAGE_SHIFT);
-+#else
-+ e820_bootmem_free(NODE_DATA(0), 0, end_pfn << PAGE_SHIFT);
-+#endif
-+ reserve_bootmem(bootmap, bootmap_size);
-+}
-+#endif
++ for (i = 0; i < nbio; i++)
++ submit_bio(operation, biolist[i]);
+
-+#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE)
-+struct edd edd;
-+#ifdef CONFIG_EDD_MODULE
-+EXPORT_SYMBOL(edd);
-+#endif
-+/**
-+ * copy_edd() - Copy the BIOS EDD information
-+ * from boot_params into a safe place.
-+ *
-+ */
-+static inline void copy_edd(void)
-+{
-+ memcpy(edd.mbr_signature, EDD_MBR_SIGNATURE, sizeof(edd.mbr_signature));
-+ memcpy(edd.edd_info, EDD_BUF, sizeof(edd.edd_info));
-+ edd.mbr_signature_nr = EDD_MBR_SIG_NR;
-+ edd.edd_info_nr = EDD_NR;
-+}
-+#else
-+static inline void copy_edd(void)
-+{
-+}
-+#endif
++ if (operation == READ)
++ blkif->st_rd_sect += preq.nr_sects;
++ else if (operation == WRITE || operation == WRITE_BARRIER)
++ blkif->st_wr_sect += preq.nr_sects;
+
-+#ifndef CONFIG_XEN
-+#define EBDA_ADDR_POINTER 0x40E
++ return;
+
-+unsigned __initdata ebda_addr;
-+unsigned __initdata ebda_size;
++ fail_put_bio:
++ for (i = 0; i < (nbio-1); i++)
++ bio_put(biolist[i]);
++ fail_flush:
++ fast_flush_area(pending_req);
++ fail_response:
++ make_response(blkif, req->id, req->operation, BLKIF_RSP_ERROR);
++ free_req(pending_req);
++ msleep(1); /* back off a bit */
++}
+
-+static void discover_ebda(void)
-+{
-+ /*
-+ * there is a real-mode segmented pointer pointing to the
-+ * 4K EBDA area at 0x40E
-+ */
-+ ebda_addr = *(unsigned short *)EBDA_ADDR_POINTER;
-+ ebda_addr <<= 4;
+
-+ ebda_size = *(unsigned short *)(unsigned long)ebda_addr;
+
-+ /* Round EBDA up to pages */
-+ if (ebda_size == 0)
-+ ebda_size = 1;
-+ ebda_size <<= 10;
-+ ebda_size = round_up(ebda_size + (ebda_addr & ~PAGE_MASK), PAGE_SIZE);
-+ if (ebda_size > 64*1024)
-+ ebda_size = 64*1024;
-+}
-+#else
-+#define discover_ebda() ((void)0)
-+#endif
++/******************************************************************
++ * MISCELLANEOUS SETUP / TEARDOWN / DEBUGGING
++ */
+
-+void __init setup_arch(char **cmdline_p)
-+{
-+#ifdef CONFIG_XEN
-+ /* Register a call for panic conditions. */
-+ atomic_notifier_chain_register(&panic_notifier_list, &xen_panic_block);
+
-+ ROOT_DEV = MKDEV(RAMDISK_MAJOR,0);
-+ screen_info = SCREEN_INFO;
++static void make_response(blkif_t *blkif, u64 id,
++ unsigned short op, int st)
++{
++ blkif_response_t resp;
++ unsigned long flags;
++ blkif_back_rings_t *blk_rings = &blkif->blk_rings;
++ int more_to_do = 0;
++ int notify;
+
-+ if (is_initial_xendomain()) {
-+ /* This is drawn from a dump from vgacon:startup in
-+ * standard Linux. */
-+ screen_info.orig_video_mode = 3;
-+ screen_info.orig_video_isVGA = 1;
-+ screen_info.orig_video_lines = 25;
-+ screen_info.orig_video_cols = 80;
-+ screen_info.orig_video_ega_bx = 3;
-+ screen_info.orig_video_points = 16;
-+ screen_info.orig_y = screen_info.orig_video_lines - 1;
-+ if (xen_start_info->console.dom0.info_size >=
-+ sizeof(struct dom0_vga_console_info)) {
-+ const struct dom0_vga_console_info *info =
-+ (struct dom0_vga_console_info *)(
-+ (char *)xen_start_info +
-+ xen_start_info->console.dom0.info_off);
-+ dom0_init_screen_info(info);
-+ }
-+ xen_start_info->console.domU.mfn = 0;
-+ xen_start_info->console.domU.evtchn = 0;
-+ } else
-+ screen_info.orig_video_isVGA = 0;
++ resp.id = id;
++ resp.operation = op;
++ resp.status = st;
+
-+ edid_info = EDID_INFO;
-+ saved_video_mode = SAVED_VIDEO_MODE;
-+ bootloader_type = LOADER_TYPE;
++ spin_lock_irqsave(&blkif->blk_ring_lock, flags);
++ /* Place on the response ring for the relevant domain. */
++ switch (blkif->blk_protocol) {
++ case BLKIF_PROTOCOL_NATIVE:
++ memcpy(RING_GET_RESPONSE(&blk_rings->native, blk_rings->native.rsp_prod_pvt),
++ &resp, sizeof(resp));
++ break;
++ case BLKIF_PROTOCOL_X86_32:
++ memcpy(RING_GET_RESPONSE(&blk_rings->x86_32, blk_rings->x86_32.rsp_prod_pvt),
++ &resp, sizeof(resp));
++ break;
++ case BLKIF_PROTOCOL_X86_64:
++ memcpy(RING_GET_RESPONSE(&blk_rings->x86_64, blk_rings->x86_64.rsp_prod_pvt),
++ &resp, sizeof(resp));
++ break;
++ default:
++ BUG();
++ }
++ blk_rings->common.rsp_prod_pvt++;
++ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&blk_rings->common, notify);
++ if (blk_rings->common.rsp_prod_pvt == blk_rings->common.req_cons) {
++ /*
++ * Tail check for pending requests. Allows frontend to avoid
++ * notifications if requests are already in flight (lower
++ * overheads and promotes batching).
++ */
++ RING_FINAL_CHECK_FOR_REQUESTS(&blk_rings->common, more_to_do);
+
-+#ifdef CONFIG_BLK_DEV_RAM
-+ rd_image_start = RAMDISK_FLAGS & RAMDISK_IMAGE_START_MASK;
-+ rd_prompt = ((RAMDISK_FLAGS & RAMDISK_PROMPT_FLAG) != 0);
-+ rd_doload = ((RAMDISK_FLAGS & RAMDISK_LOAD_FLAG) != 0);
++ } else if (RING_HAS_UNCONSUMED_REQUESTS(&blk_rings->common)) {
++ more_to_do = 1;
++ }
+
++ spin_unlock_irqrestore(&blkif->blk_ring_lock, flags);
+
-+#endif
++ if (more_to_do)
++ blkif_notify_work(blkif);
++ if (notify)
++ notify_remote_via_irq(blkif->irq);
++}
+
-+ HYPERVISOR_vm_assist(VMASST_CMD_enable,
-+ VMASST_TYPE_writable_pagetables);
++static int __init blkif_init(void)
++{
++ int i, mmap_pages;
+
-+ ARCH_SETUP
-+#else
-+ ROOT_DEV = old_decode_dev(ORIG_ROOT_DEV);
-+ screen_info = SCREEN_INFO;
-+ edid_info = EDID_INFO;
-+ saved_video_mode = SAVED_VIDEO_MODE;
-+ bootloader_type = LOADER_TYPE;
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+#ifdef CONFIG_BLK_DEV_RAM
-+ rd_image_start = RAMDISK_FLAGS & RAMDISK_IMAGE_START_MASK;
-+ rd_prompt = ((RAMDISK_FLAGS & RAMDISK_PROMPT_FLAG) != 0);
-+ rd_doload = ((RAMDISK_FLAGS & RAMDISK_LOAD_FLAG) != 0);
-+#endif
-+#endif /* !CONFIG_XEN */
-+ setup_memory_region();
-+ copy_edd();
++ mmap_pages = blkif_reqs * BLKIF_MAX_SEGMENTS_PER_REQUEST;
+
-+ if (!MOUNT_ROOT_RDONLY)
-+ root_mountflags &= ~MS_RDONLY;
-+ init_mm.start_code = (unsigned long) &_text;
-+ init_mm.end_code = (unsigned long) &_etext;
-+ init_mm.end_data = (unsigned long) &_edata;
-+ init_mm.brk = (unsigned long) &_end;
++ pending_reqs = kmalloc(sizeof(pending_reqs[0]) *
++ blkif_reqs, GFP_KERNEL);
++ pending_grant_handles = kmalloc(sizeof(pending_grant_handles[0]) *
++ mmap_pages, GFP_KERNEL);
++ pending_pages = alloc_empty_pages_and_pagevec(mmap_pages);
+
-+ code_resource.start = virt_to_phys(&_text);
-+ code_resource.end = virt_to_phys(&_etext)-1;
-+ data_resource.start = virt_to_phys(&_etext);
-+ data_resource.end = virt_to_phys(&_edata)-1;
++ if (!pending_reqs || !pending_grant_handles || !pending_pages)
++ goto out_of_memory;
+
-+ parse_cmdline_early(cmdline_p);
++ for (i = 0; i < mmap_pages; i++)
++ pending_grant_handles[i] = BLKBACK_INVALID_HANDLE;
+
-+ early_identify_cpu(&boot_cpu_data);
++ blkif_interface_init();
+
-+ /*
-+ * partially used pages are not usable - thus
-+ * we are rounding upwards:
-+ */
-+ end_pfn = e820_end_of_ram();
-+ num_physpages = end_pfn; /* for pfn_valid */
++ memset(pending_reqs, 0, sizeof(pending_reqs));
++ INIT_LIST_HEAD(&pending_free);
+
-+ check_efer();
++ for (i = 0; i < blkif_reqs; i++)
++ list_add_tail(&pending_reqs[i].free_list, &pending_free);
+
-+ discover_ebda();
++ blkif_xenbus_init();
+
-+ init_memory_mapping(0, (end_pfn_map << PAGE_SHIFT));
++ return 0;
+
-+ if (is_initial_xendomain())
-+ dmi_scan_machine();
++ out_of_memory:
++ kfree(pending_reqs);
++ kfree(pending_grant_handles);
++ free_empty_pages_and_pagevec(pending_pages, mmap_pages);
++ printk("%s: out of memory\n", __FUNCTION__);
++ return -ENOMEM;
++}
+
-+#ifdef CONFIG_ACPI_NUMA
-+ /*
-+ * Parse SRAT to discover nodes.
-+ */
-+ acpi_numa_init();
-+#endif
++module_init(blkif_init);
+
-+#ifdef CONFIG_NUMA
-+ numa_initmem_init(0, end_pfn);
-+#else
-+ contig_initmem_init(0, end_pfn);
-+#endif
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blkback/common.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blkback/common.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,139 @@
++/*
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+#ifdef CONFIG_XEN
-+ /*
-+ * Reserve kernel, physmap, start info, initial page tables, and
-+ * direct mapping.
-+ */
-+ reserve_bootmem_generic(__pa_symbol(&_text),
-+ (table_end << PAGE_SHIFT) - __pa_symbol(&_text));
-+#else
-+ /* Reserve direct mapping */
-+ reserve_bootmem_generic(table_start << PAGE_SHIFT,
-+ (table_end - table_start) << PAGE_SHIFT);
++#ifndef __BLKIF__BACKEND__COMMON_H__
++#define __BLKIF__BACKEND__COMMON_H__
+
-+ /* reserve kernel */
-+ reserve_bootmem_generic(__pa_symbol(&_text),
-+ __pa_symbol(&_end) - __pa_symbol(&_text));
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/interrupt.h>
++#include <linux/slab.h>
++#include <linux/blkdev.h>
++#include <linux/vmalloc.h>
++#include <linux/wait.h>
++#include <asm/io.h>
++#include <asm/setup.h>
++#include <asm/pgalloc.h>
++#include <xen/evtchn.h>
++#include <asm/hypervisor.h>
++#include <xen/blkif.h>
++#include <xen/gnttab.h>
++#include <xen/driver_util.h>
++#include <xen/xenbus.h>
+
-+ /*
-+ * reserve physical page 0 - it's a special BIOS page on many boxes,
-+ * enabling clean reboots, SMP operation, laptop functions.
-+ */
-+ reserve_bootmem_generic(0, PAGE_SIZE);
++#define DPRINTK(_f, _a...) \
++ pr_debug("(file=%s, line=%d) " _f, \
++ __FILE__ , __LINE__ , ## _a )
+
-+ /* reserve ebda region */
-+ if (ebda_addr)
-+ reserve_bootmem_generic(ebda_addr, ebda_size);
++struct vbd {
++ blkif_vdev_t handle; /* what the domain refers to this vbd as */
++ unsigned char readonly; /* Non-zero -> read-only */
++ unsigned char type; /* VDISK_xxx */
++ u32 pdevice; /* phys device that this vbd maps to */
++ struct block_device *bdev;
++};
+
-+#ifdef CONFIG_SMP
-+ /*
-+ * But first pinch a few for the stack/trampoline stuff
-+ * FIXME: Don't need the extra page at 4K, but need to fix
-+ * trampoline before removing it. (see the GDT stuff)
-+ */
-+ reserve_bootmem_generic(PAGE_SIZE, PAGE_SIZE);
++struct backend_info;
+
-+ /* Reserve SMP trampoline */
-+ reserve_bootmem_generic(SMP_TRAMPOLINE_BASE, PAGE_SIZE);
-+#endif
-+#endif
++typedef struct blkif_st {
++ /* Unique identifier for this interface. */
++ domid_t domid;
++ unsigned int handle;
++ /* Physical parameters of the comms window. */
++ unsigned int irq;
++ /* Comms information. */
++ enum blkif_protocol blk_protocol;
++ blkif_back_rings_t blk_rings;
++ struct vm_struct *blk_ring_area;
++ /* The VBD attached to this interface. */
++ struct vbd vbd;
++ /* Back pointer to the backend_info. */
++ struct backend_info *be;
++ /* Private fields. */
++ spinlock_t blk_ring_lock;
++ atomic_t refcnt;
+
-+#ifdef CONFIG_ACPI_SLEEP
-+ /*
-+ * Reserve low memory region for sleep support.
-+ */
-+ acpi_reserve_bootmem();
-+#endif
-+#ifdef CONFIG_XEN
-+#ifdef CONFIG_BLK_DEV_INITRD
-+ if (xen_start_info->mod_start) {
-+ if (INITRD_START + INITRD_SIZE <= (end_pfn << PAGE_SHIFT)) {
-+ /*reserve_bootmem_generic(INITRD_START, INITRD_SIZE);*/
-+ initrd_start = INITRD_START + PAGE_OFFSET;
-+ initrd_end = initrd_start+INITRD_SIZE;
-+ initrd_below_start_ok = 1;
-+ } else {
-+ printk(KERN_ERR "initrd extends beyond end of memory "
-+ "(0x%08lx > 0x%08lx)\ndisabling initrd\n",
-+ (unsigned long)(INITRD_START + INITRD_SIZE),
-+ (unsigned long)(end_pfn << PAGE_SHIFT));
-+ initrd_start = 0;
-+ }
-+ }
-+#endif
-+#else /* CONFIG_XEN */
-+#ifdef CONFIG_BLK_DEV_INITRD
-+ if (LOADER_TYPE && INITRD_START) {
-+ if (INITRD_START + INITRD_SIZE <= (end_pfn << PAGE_SHIFT)) {
-+ reserve_bootmem_generic(INITRD_START, INITRD_SIZE);
-+ initrd_start =
-+ INITRD_START ? INITRD_START + PAGE_OFFSET : 0;
-+ initrd_end = initrd_start+INITRD_SIZE;
-+ }
-+ else {
-+ printk(KERN_ERR "initrd extends beyond end of memory "
-+ "(0x%08lx > 0x%08lx)\ndisabling initrd\n",
-+ (unsigned long)(INITRD_START + INITRD_SIZE),
-+ (unsigned long)(end_pfn << PAGE_SHIFT));
-+ initrd_start = 0;
-+ }
-+ }
-+#endif
-+#endif /* !CONFIG_XEN */
-+#ifdef CONFIG_KEXEC
-+#ifdef CONFIG_XEN
-+ xen_machine_kexec_setup_resources();
-+#else
-+ if (crashk_res.start != crashk_res.end) {
-+ reserve_bootmem_generic(crashk_res.start,
-+ crashk_res.end - crashk_res.start + 1);
-+ }
-+#endif
-+#endif
++ wait_queue_head_t wq;
++ struct task_struct *xenblkd;
++ unsigned int waiting_reqs;
++ request_queue_t *plug;
+
-+ paging_init();
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ /*
-+ * Find and reserve possible boot-time SMP configuration:
-+ */
-+ find_smp_config();
-+#endif
-+#ifdef CONFIG_XEN
-+ {
-+ int i, j, k, fpp;
-+ unsigned long p2m_pages;
++ /* statistics */
++ unsigned long st_print;
++ int st_rd_req;
++ int st_wr_req;
++ int st_oo_req;
++ int st_br_req;
++ int st_rd_sect;
++ int st_wr_sect;
+
-+ p2m_pages = end_pfn;
-+ if (xen_start_info->nr_pages > end_pfn) {
-+ /*
-+ * the end_pfn was shrunk (probably by mem= or highmem=
-+ * kernel parameter); shrink reservation with the HV
-+ */
-+ struct xen_memory_reservation reservation = {
-+ .address_bits = 0,
-+ .extent_order = 0,
-+ .domid = DOMID_SELF
-+ };
-+ unsigned int difference;
-+ int ret;
-+
-+ difference = xen_start_info->nr_pages - end_pfn;
-+
-+ set_xen_guest_handle(reservation.extent_start,
-+ ((unsigned long *)xen_start_info->mfn_list) + end_pfn);
-+ reservation.nr_extents = difference;
-+ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
-+ &reservation);
-+ BUG_ON (ret != difference);
-+ }
-+ else if (end_pfn > xen_start_info->nr_pages)
-+ p2m_pages = xen_start_info->nr_pages;
++ wait_queue_head_t waiting_to_free;
+
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ /* Make sure we have a large enough P->M table. */
-+ phys_to_machine_mapping = alloc_bootmem_pages(
-+ end_pfn * sizeof(unsigned long));
-+ memset(phys_to_machine_mapping, ~0,
-+ end_pfn * sizeof(unsigned long));
-+ memcpy(phys_to_machine_mapping,
-+ (unsigned long *)xen_start_info->mfn_list,
-+ p2m_pages * sizeof(unsigned long));
-+ free_bootmem(
-+ __pa(xen_start_info->mfn_list),
-+ PFN_PHYS(PFN_UP(xen_start_info->nr_pages *
-+ sizeof(unsigned long))));
++ grant_handle_t shmem_handle;
++ grant_ref_t shmem_ref;
++} blkif_t;
+
-+ /*
-+ * Initialise the list of the frames that specify the
-+ * list of frames that make up the p2m table. Used by
-+ * save/restore.
-+ */
-+ pfn_to_mfn_frame_list_list = alloc_bootmem_pages(PAGE_SIZE);
++blkif_t *blkif_alloc(domid_t domid);
++void blkif_disconnect(blkif_t *blkif);
++void blkif_free(blkif_t *blkif);
++int blkif_map(blkif_t *blkif, unsigned long shared_page, unsigned int evtchn);
+
-+ fpp = PAGE_SIZE/sizeof(unsigned long);
-+ for (i=0, j=0, k=-1; i< end_pfn; i+=fpp, j++) {
-+ if ((j % fpp) == 0) {
-+ k++;
-+ BUG_ON(k>=fpp);
-+ pfn_to_mfn_frame_list[k] =
-+ alloc_bootmem_pages(PAGE_SIZE);
-+ pfn_to_mfn_frame_list_list[k] =
-+ virt_to_mfn(pfn_to_mfn_frame_list[k]);
-+ j=0;
-+ }
-+ pfn_to_mfn_frame_list[k][j] =
-+ virt_to_mfn(&phys_to_machine_mapping[i]);
-+ }
-+ HYPERVISOR_shared_info->arch.max_pfn = end_pfn;
-+ HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list =
-+ virt_to_mfn(pfn_to_mfn_frame_list_list);
-+ }
++#define blkif_get(_b) (atomic_inc(&(_b)->refcnt))
++#define blkif_put(_b) \
++ do { \
++ if (atomic_dec_and_test(&(_b)->refcnt)) \
++ wake_up(&(_b)->waiting_to_free);\
++ } while (0)
+
-+ }
++/* Create a vbd. */
++int vbd_create(blkif_t *blkif, blkif_vdev_t vdevice, unsigned major,
++ unsigned minor, int readonly, int cdrom);
++void vbd_free(struct vbd *vbd);
+
-+ if (!is_initial_xendomain()) {
-+ acpi_disabled = 1;
-+#ifdef CONFIG_ACPI
-+ acpi_ht = 0;
-+#endif
-+ }
-+#endif
++unsigned long long vbd_size(struct vbd *vbd);
++unsigned int vbd_info(struct vbd *vbd);
++unsigned long vbd_secsize(struct vbd *vbd);
+
-+#ifndef CONFIG_XEN
-+ check_ioapic();
-+#endif
++struct phys_req {
++ unsigned short dev;
++ unsigned short nr_sects;
++ struct block_device *bdev;
++ blkif_sector_t sector_number;
++};
+
-+ zap_low_mappings(0);
++int vbd_translate(struct phys_req *req, blkif_t *blkif, int operation);
+
-+ /*
-+ * set this early, so we dont allocate cpu0
-+ * if MADT list doesnt list BSP first
-+ * mpparse.c/MP_processor_info() allocates logical cpu numbers.
-+ */
-+ cpu_set(0, cpu_present_map);
-+#ifdef CONFIG_ACPI
-+ /*
-+ * Initialize the ACPI boot-time table parser (gets the RSDP and SDT).
-+ * Call this early for SRAT node setup.
-+ */
-+ acpi_boot_table_init();
++void blkif_interface_init(void);
+
-+ /*
-+ * Read APIC and some other early information from ACPI tables.
-+ */
-+ acpi_boot_init();
-+#endif
++void blkif_xenbus_init(void);
+
-+ init_cpu_to_node();
++irqreturn_t blkif_be_int(int irq, void *dev_id, struct pt_regs *regs);
++int blkif_schedule(void *arg);
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ /*
-+ * get boot-time SMP configuration:
-+ */
-+ if (smp_found_config)
-+ get_smp_config();
-+#ifndef CONFIG_XEN
-+ init_apic_mappings();
-+#endif
-+#endif
-+#if defined(CONFIG_XEN) && defined(CONFIG_SMP) && !defined(CONFIG_HOTPLUG_CPU)
-+ prefill_possible_map();
-+#endif
++int blkback_barrier(struct xenbus_transaction xbt,
++ struct backend_info *be, int state);
+
-+ /*
-+ * Request address space for all standard RAM and ROM resources
-+ * and also for regions reported as reserved by the e820.
-+ */
-+ probe_roms();
-+#ifdef CONFIG_XEN
-+ if (is_initial_xendomain()) {
-+ struct xen_memory_map memmap;
++#endif /* __BLKIF__BACKEND__COMMON_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blkback/interface.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blkback/interface.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,181 @@
++/******************************************************************************
++ * arch/xen/drivers/blkif/backend/interface.c
++ *
++ * Block-device interface management.
++ *
++ * Copyright (c) 2004, Keir Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ memmap.nr_entries = E820MAX;
-+ set_xen_guest_handle(memmap.buffer, machine_e820.map);
++#include "common.h"
++#include <xen/evtchn.h>
++#include <linux/kthread.h>
+
-+ if (HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap))
-+ BUG();
-+ machine_e820.nr_map = memmap.nr_entries;
++static kmem_cache_t *blkif_cachep;
+
-+ e820_reserve_resources(machine_e820.map, machine_e820.nr_map);
-+ }
-+#else
-+ e820_reserve_resources(e820.map, e820.nr_map);
-+#endif
++blkif_t *blkif_alloc(domid_t domid)
++{
++ blkif_t *blkif;
+
-+ request_resource(&iomem_resource, &video_ram_resource);
++ blkif = kmem_cache_alloc(blkif_cachep, GFP_KERNEL);
++ if (!blkif)
++ return ERR_PTR(-ENOMEM);
+
-+ {
-+ unsigned i;
-+ /* request I/O space for devices used on all i[345]86 PCs */
-+ for (i = 0; i < STANDARD_IO_RESOURCES; i++)
-+ request_resource(&ioport_resource, &standard_io_resources[i]);
-+ }
++ memset(blkif, 0, sizeof(*blkif));
++ blkif->domid = domid;
++ spin_lock_init(&blkif->blk_ring_lock);
++ atomic_set(&blkif->refcnt, 1);
++ init_waitqueue_head(&blkif->wq);
++ blkif->st_print = jiffies;
++ init_waitqueue_head(&blkif->waiting_to_free);
+
-+#ifdef CONFIG_XEN
-+ if (is_initial_xendomain())
-+ e820_setup_gap(machine_e820.map, machine_e820.nr_map);
-+#else
-+ e820_setup_gap(e820.map, e820.nr_map);
-+#endif
++ return blkif;
++}
+
-+#ifdef CONFIG_XEN
-+ {
-+ struct physdev_set_iopl set_iopl;
++static int map_frontend_page(blkif_t *blkif, unsigned long shared_page)
++{
++ struct gnttab_map_grant_ref op;
+
-+ set_iopl.iopl = 1;
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
++ gnttab_set_map_op(&op, (unsigned long)blkif->blk_ring_area->addr,
++ GNTMAP_host_map, shared_page, blkif->domid);
+
-+ if (is_initial_xendomain()) {
-+#ifdef CONFIG_VT
-+#if defined(CONFIG_VGA_CONSOLE)
-+ conswitchp = &vga_con;
-+#elif defined(CONFIG_DUMMY_CONSOLE)
-+ conswitchp = &dummy_con;
-+#endif
-+#endif
-+ } else {
-+#if defined(CONFIG_VT) && defined(CONFIG_DUMMY_CONSOLE)
-+ conswitchp = &dummy_con;
-+#endif
-+ }
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
++
++ if (op.status) {
++ DPRINTK(" Grant table operation failure !\n");
++ return op.status;
+ }
-+ xencons_early_setup();
-+#else /* CONFIG_XEN */
+
-+#ifdef CONFIG_VT
-+#if defined(CONFIG_VGA_CONSOLE)
-+ conswitchp = &vga_con;
-+#elif defined(CONFIG_DUMMY_CONSOLE)
-+ conswitchp = &dummy_con;
-+#endif
-+#endif
++ blkif->shmem_ref = shared_page;
++ blkif->shmem_handle = op.handle;
+
-+#endif /* !CONFIG_XEN */
++ return 0;
+}
+
-+#ifdef CONFIG_XEN
-+static int
-+xen_panic_event(struct notifier_block *this, unsigned long event, void *ptr)
++static void unmap_frontend_page(blkif_t *blkif)
+{
-+ HYPERVISOR_shutdown(SHUTDOWN_crash);
-+ /* we're never actually going to get here... */
-+ return NOTIFY_DONE;
-+}
-+#endif /* !CONFIG_XEN */
++ struct gnttab_unmap_grant_ref op;
+
++ gnttab_set_unmap_op(&op, (unsigned long)blkif->blk_ring_area->addr,
++ GNTMAP_host_map, blkif->shmem_handle);
+
-+static int __cpuinit get_model_name(struct cpuinfo_x86 *c)
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
++}
++
++int blkif_map(blkif_t *blkif, unsigned long shared_page, unsigned int evtchn)
+{
-+ unsigned int *v;
++ int err;
+
-+ if (c->extended_cpuid_level < 0x80000004)
++ /* Already connected through? */
++ if (blkif->irq)
+ return 0;
+
-+ v = (unsigned int *) c->x86_model_id;
-+ cpuid(0x80000002, &v[0], &v[1], &v[2], &v[3]);
-+ cpuid(0x80000003, &v[4], &v[5], &v[6], &v[7]);
-+ cpuid(0x80000004, &v[8], &v[9], &v[10], &v[11]);
-+ c->x86_model_id[48] = 0;
-+ return 1;
-+}
++ if ( (blkif->blk_ring_area = alloc_vm_area(PAGE_SIZE)) == NULL )
++ return -ENOMEM;
+
++ err = map_frontend_page(blkif, shared_page);
++ if (err) {
++ free_vm_area(blkif->blk_ring_area);
++ return err;
++ }
+
-+static void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c)
-+{
-+ unsigned int n, dummy, eax, ebx, ecx, edx;
++ switch (blkif->blk_protocol) {
++ case BLKIF_PROTOCOL_NATIVE:
++ {
++ blkif_sring_t *sring;
++ sring = (blkif_sring_t *)blkif->blk_ring_area->addr;
++ BACK_RING_INIT(&blkif->blk_rings.native, sring, PAGE_SIZE);
++ break;
++ }
++ case BLKIF_PROTOCOL_X86_32:
++ {
++ blkif_x86_32_sring_t *sring_x86_32;
++ sring_x86_32 = (blkif_x86_32_sring_t *)blkif->blk_ring_area->addr;
++ BACK_RING_INIT(&blkif->blk_rings.x86_32, sring_x86_32, PAGE_SIZE);
++ break;
++ }
++ case BLKIF_PROTOCOL_X86_64:
++ {
++ blkif_x86_64_sring_t *sring_x86_64;
++ sring_x86_64 = (blkif_x86_64_sring_t *)blkif->blk_ring_area->addr;
++ BACK_RING_INIT(&blkif->blk_rings.x86_64, sring_x86_64, PAGE_SIZE);
++ break;
++ }
++ default:
++ BUG();
++ }
+
-+ n = c->extended_cpuid_level;
++ err = bind_interdomain_evtchn_to_irqhandler(
++ blkif->domid, evtchn, blkif_be_int, 0, "blkif-backend", blkif);
++ if (err < 0)
++ {
++ unmap_frontend_page(blkif);
++ free_vm_area(blkif->blk_ring_area);
++ blkif->blk_rings.common.sring = NULL;
++ return err;
++ }
++ blkif->irq = err;
+
-+ if (n >= 0x80000005) {
-+ cpuid(0x80000005, &dummy, &ebx, &ecx, &edx);
-+ printk(KERN_INFO "CPU: L1 I Cache: %dK (%d bytes/line), D cache %dK (%d bytes/line)\n",
-+ edx>>24, edx&0xFF, ecx>>24, ecx&0xFF);
-+ c->x86_cache_size=(ecx>>24)+(edx>>24);
-+ /* On K8 L1 TLB is inclusive, so don't count it */
-+ c->x86_tlbsize = 0;
++ return 0;
++}
++
++void blkif_disconnect(blkif_t *blkif)
++{
++ if (blkif->xenblkd) {
++ kthread_stop(blkif->xenblkd);
++ blkif->xenblkd = NULL;
+ }
+
-+ if (n >= 0x80000006) {
-+ cpuid(0x80000006, &dummy, &ebx, &ecx, &edx);
-+ ecx = cpuid_ecx(0x80000006);
-+ c->x86_cache_size = ecx >> 16;
-+ c->x86_tlbsize += ((ebx >> 16) & 0xfff) + (ebx & 0xfff);
++ atomic_dec(&blkif->refcnt);
++ wait_event(blkif->waiting_to_free, atomic_read(&blkif->refcnt) == 0);
++ atomic_inc(&blkif->refcnt);
+
-+ printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n",
-+ c->x86_cache_size, ecx & 0xFF);
++ if (blkif->irq) {
++ unbind_from_irqhandler(blkif->irq, blkif);
++ blkif->irq = 0;
+ }
+
-+ if (n >= 0x80000007)
-+ cpuid(0x80000007, &dummy, &dummy, &dummy, &c->x86_power);
-+ if (n >= 0x80000008) {
-+ cpuid(0x80000008, &eax, &dummy, &dummy, &dummy);
-+ c->x86_virt_bits = (eax >> 8) & 0xff;
-+ c->x86_phys_bits = eax & 0xff;
++ if (blkif->blk_rings.common.sring) {
++ unmap_frontend_page(blkif);
++ free_vm_area(blkif->blk_ring_area);
++ blkif->blk_rings.common.sring = NULL;
+ }
+}
+
-+#ifdef CONFIG_NUMA
-+static int nearby_node(int apicid)
++void blkif_free(blkif_t *blkif)
+{
-+ int i;
-+ for (i = apicid - 1; i >= 0; i--) {
-+ int node = apicid_to_node[i];
-+ if (node != NUMA_NO_NODE && node_online(node))
-+ return node;
-+ }
-+ for (i = apicid + 1; i < MAX_LOCAL_APIC; i++) {
-+ int node = apicid_to_node[i];
-+ if (node != NUMA_NO_NODE && node_online(node))
-+ return node;
-+ }
-+ return first_node(node_online_map); /* Shouldn't happen */
++ if (!atomic_dec_and_test(&blkif->refcnt))
++ BUG();
++ kmem_cache_free(blkif_cachep, blkif);
+}
-+#endif
+
-+/*
-+ * On a AMD dual core setup the lower bits of the APIC id distingush the cores.
-+ * Assumes number of cores is a power of two.
-+ */
-+static void __init amd_detect_cmp(struct cpuinfo_x86 *c)
++void __init blkif_interface_init(void)
+{
-+#ifdef CONFIG_SMP
-+ unsigned bits;
-+#ifdef CONFIG_NUMA
-+ int cpu = smp_processor_id();
-+ int node = 0;
-+ unsigned apicid = hard_smp_processor_id();
-+#endif
-+ unsigned ecx = cpuid_ecx(0x80000008);
-+
-+ c->x86_max_cores = (ecx & 0xff) + 1;
++ blkif_cachep = kmem_cache_create("blkif_cache", sizeof(blkif_t),
++ 0, 0, NULL, NULL);
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blkback/vbd.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blkback/vbd.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,118 @@
++/******************************************************************************
++ * blkback/vbd.c
++ *
++ * Routines for managing virtual block devices (VBDs).
++ *
++ * Copyright (c) 2003-2005, Keir Fraser & Steve Hand
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ /* CPU telling us the core id bits shift? */
-+ bits = (ecx >> 12) & 0xF;
++#include "common.h"
+
-+ /* Otherwise recompute */
-+ if (bits == 0) {
-+ while ((1 << bits) < c->x86_max_cores)
-+ bits++;
-+ }
++#define vbd_sz(_v) ((_v)->bdev->bd_part ? \
++ (_v)->bdev->bd_part->nr_sects : (_v)->bdev->bd_disk->capacity)
+
-+ /* Low order bits define the core id (index of core in socket) */
-+ c->cpu_core_id = c->phys_proc_id & ((1 << bits)-1);
-+ /* Convert the APIC ID into the socket ID */
-+ c->phys_proc_id = phys_pkg_id(bits);
++unsigned long long vbd_size(struct vbd *vbd)
++{
++ return vbd_sz(vbd);
++}
+
-+#ifdef CONFIG_NUMA
-+ node = c->phys_proc_id;
-+ if (apicid_to_node[apicid] != NUMA_NO_NODE)
-+ node = apicid_to_node[apicid];
-+ if (!node_online(node)) {
-+ /* Two possibilities here:
-+ - The CPU is missing memory and no node was created.
-+ In that case try picking one from a nearby CPU
-+ - The APIC IDs differ from the HyperTransport node IDs
-+ which the K8 northbridge parsing fills in.
-+ Assume they are all increased by a constant offset,
-+ but in the same order as the HT nodeids.
-+ If that doesn't result in a usable node fall back to the
-+ path for the previous case. */
-+ int ht_nodeid = apicid - (cpu_data[0].phys_proc_id << bits);
-+ if (ht_nodeid >= 0 &&
-+ apicid_to_node[ht_nodeid] != NUMA_NO_NODE)
-+ node = apicid_to_node[ht_nodeid];
-+ /* Pick a nearby node */
-+ if (!node_online(node))
-+ node = nearby_node(apicid);
-+ }
-+ numa_set_node(cpu, node);
++unsigned int vbd_info(struct vbd *vbd)
++{
++ return vbd->type | (vbd->readonly?VDISK_READONLY:0);
++}
+
-+ printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node);
-+#endif
-+#endif
++unsigned long vbd_secsize(struct vbd *vbd)
++{
++ return bdev_hardsect_size(vbd->bdev);
+}
+
-+static void __init init_amd(struct cpuinfo_x86 *c)
++int vbd_create(blkif_t *blkif, blkif_vdev_t handle, unsigned major,
++ unsigned minor, int readonly, int cdrom)
+{
-+ unsigned level;
++ struct vbd *vbd;
++ struct block_device *bdev;
+
-+#ifdef CONFIG_SMP
-+ unsigned long value;
++ vbd = &blkif->vbd;
++ vbd->handle = handle;
++ vbd->readonly = readonly;
++ vbd->type = 0;
+
-+ /*
-+ * Disable TLB flush filter by setting HWCR.FFDIS on K8
-+ * bit 6 of msr C001_0015
-+ *
-+ * Errata 63 for SH-B3 steppings
-+ * Errata 122 for all steppings (F+ have it disabled by default)
-+ */
-+ if (c->x86 == 15) {
-+ rdmsrl(MSR_K8_HWCR, value);
-+ value |= 1 << 6;
-+ wrmsrl(MSR_K8_HWCR, value);
-+ }
-+#endif
++ vbd->pdevice = MKDEV(major, minor);
+
-+ /* Bit 31 in normal CPUID used for nonstandard 3DNow ID;
-+ 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway */
-+ clear_bit(0*32+31, &c->x86_capability);
-+
-+ /* On C+ stepping K8 rep microcode works well for copy/memset */
-+ level = cpuid_eax(1);
-+ if (c->x86 == 15 && ((level >= 0x0f48 && level < 0x0f50) || level >= 0x0f58))
-+ set_bit(X86_FEATURE_REP_GOOD, &c->x86_capability);
++ bdev = open_by_devnum(vbd->pdevice,
++ vbd->readonly ? FMODE_READ : FMODE_WRITE);
+
-+ /* Enable workaround for FXSAVE leak */
-+ if (c->x86 >= 6)
-+ set_bit(X86_FEATURE_FXSAVE_LEAK, &c->x86_capability);
++ if (IS_ERR(bdev)) {
++ DPRINTK("vbd_creat: device %08x could not be opened.\n",
++ vbd->pdevice);
++ return -ENOENT;
++ }
+
-+ level = get_model_name(c);
-+ if (!level) {
-+ switch (c->x86) {
-+ case 15:
-+ /* Should distinguish Models here, but this is only
-+ a fallback anyways. */
-+ strcpy(c->x86_model_id, "Hammer");
-+ break;
-+ }
-+ }
-+ display_cacheinfo(c);
++ vbd->bdev = bdev;
+
-+ /* c->x86_power is 8000_0007 edx. Bit 8 is constant TSC */
-+ if (c->x86_power & (1<<8))
-+ set_bit(X86_FEATURE_CONSTANT_TSC, &c->x86_capability);
++ if (vbd->bdev->bd_disk == NULL) {
++ DPRINTK("vbd_creat: device %08x doesn't exist.\n",
++ vbd->pdevice);
++ vbd_free(vbd);
++ return -ENOENT;
++ }
+
-+ /* Multi core CPU? */
-+ if (c->extended_cpuid_level >= 0x80000008)
-+ amd_detect_cmp(c);
++ if (vbd->bdev->bd_disk->flags & GENHD_FL_CD || cdrom)
++ vbd->type |= VDISK_CDROM;
++ if (vbd->bdev->bd_disk->flags & GENHD_FL_REMOVABLE)
++ vbd->type |= VDISK_REMOVABLE;
+
-+ /* Fix cpuid4 emulation for more */
-+ num_cache_leaves = 3;
++ DPRINTK("Successful creation of handle=%04x (dom=%u)\n",
++ handle, blkif->domid);
++ return 0;
+}
+
-+static void __cpuinit detect_ht(struct cpuinfo_x86 *c)
++void vbd_free(struct vbd *vbd)
+{
-+#ifdef CONFIG_SMP
-+ u32 eax, ebx, ecx, edx;
-+ int index_msb, core_bits;
++ if (vbd->bdev)
++ blkdev_put(vbd->bdev);
++ vbd->bdev = NULL;
++}
+
-+ cpuid(1, &eax, &ebx, &ecx, &edx);
++int vbd_translate(struct phys_req *req, blkif_t *blkif, int operation)
++{
++ struct vbd *vbd = &blkif->vbd;
++ int rc = -EACCES;
+
++ if ((operation != READ) && vbd->readonly)
++ goto out;
+
-+ if (!cpu_has(c, X86_FEATURE_HT))
-+ return;
-+ if (cpu_has(c, X86_FEATURE_CMP_LEGACY))
++ if (unlikely((req->sector_number + req->nr_sects) > vbd_sz(vbd)))
+ goto out;
+
-+ smp_num_siblings = (ebx & 0xff0000) >> 16;
++ req->dev = vbd->pdevice;
++ req->bdev = vbd->bdev;
++ rc = 0;
+
-+ if (smp_num_siblings == 1) {
-+ printk(KERN_INFO "CPU: Hyper-Threading is disabled\n");
-+ } else if (smp_num_siblings > 1 ) {
++ out:
++ return rc;
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blkback/xenbus.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blkback/xenbus.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,541 @@
++/* Xenbus code for blkif backend
++ Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
++ Copyright (C) 2005 XenSource Ltd
+
-+ if (smp_num_siblings > NR_CPUS) {
-+ printk(KERN_WARNING "CPU: Unsupported number of the siblings %d", smp_num_siblings);
-+ smp_num_siblings = 1;
-+ return;
-+ }
++ This program is free software; you can redistribute it and/or modify
++ it under the terms of the GNU General Public License as published by
++ the Free Software Foundation; either version 2 of the License, or
++ (at your option) any later version.
+
-+ index_msb = get_count_order(smp_num_siblings);
-+ c->phys_proc_id = phys_pkg_id(index_msb);
++ This program is distributed in the hope that it will be useful,
++ but WITHOUT ANY WARRANTY; without even the implied warranty of
++ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ GNU General Public License for more details.
+
-+ smp_num_siblings = smp_num_siblings / c->x86_max_cores;
++ You should have received a copy of the GNU General Public License
++ along with this program; if not, write to the Free Software
++ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++*/
+
-+ index_msb = get_count_order(smp_num_siblings) ;
++#include <stdarg.h>
++#include <linux/module.h>
++#include <linux/kthread.h>
++#include "common.h"
+
-+ core_bits = get_count_order(c->x86_max_cores);
++#undef DPRINTK
++#define DPRINTK(fmt, args...) \
++ pr_debug("blkback/xenbus (%s:%d) " fmt ".\n", \
++ __FUNCTION__, __LINE__, ##args)
+
-+ c->cpu_core_id = phys_pkg_id(index_msb) &
-+ ((1 << core_bits) - 1);
-+ }
-+out:
-+ if ((c->x86_max_cores * smp_num_siblings) > 1) {
-+ printk(KERN_INFO "CPU: Physical Processor ID: %d\n", c->phys_proc_id);
-+ printk(KERN_INFO "CPU: Processor Core ID: %d\n", c->cpu_core_id);
-+ }
++struct backend_info
++{
++ struct xenbus_device *dev;
++ blkif_t *blkif;
++ struct xenbus_watch backend_watch;
++ unsigned major;
++ unsigned minor;
++ char *mode;
++};
+
-+#endif
-+}
++static void connect(struct backend_info *);
++static int connect_ring(struct backend_info *);
++static void backend_changed(struct xenbus_watch *, const char **,
++ unsigned int);
+
-+/*
-+ * find out the number of processor cores on the die
-+ */
-+static int __cpuinit intel_num_cpu_cores(struct cpuinfo_x86 *c)
++static int blkback_name(blkif_t *blkif, char *buf)
+{
-+ unsigned int eax, t;
++ char *devpath, *devname;
++ struct xenbus_device *dev = blkif->be->dev;
+
-+ if (c->cpuid_level < 4)
-+ return 1;
++ devpath = xenbus_read(XBT_NIL, dev->nodename, "dev", NULL);
++ if (IS_ERR(devpath))
++ return PTR_ERR(devpath);
++
++ if ((devname = strstr(devpath, "/dev/")) != NULL)
++ devname += strlen("/dev/");
++ else
++ devname = devpath;
+
-+ cpuid_count(4, 0, &eax, &t, &t, &t);
++ snprintf(buf, TASK_COMM_LEN, "blkback.%d.%s", blkif->domid, devname);
++ kfree(devpath);
++
++ return 0;
++}
+
-+ if (eax & 0x1f)
-+ return ((eax >> 26) + 1);
-+ else
-+ return 1;
++static void update_blkif_status(blkif_t *blkif)
++{
++ int err;
++ char name[TASK_COMM_LEN];
++
++ /* Not ready to connect? */
++ if (!blkif->irq || !blkif->vbd.bdev)
++ return;
++
++ /* Already connected? */
++ if (blkif->be->dev->state == XenbusStateConnected)
++ return;
++
++ /* Attempt to connect: exit if we fail to. */
++ connect(blkif->be);
++ if (blkif->be->dev->state != XenbusStateConnected)
++ return;
++
++ err = blkback_name(blkif, name);
++ if (err) {
++ xenbus_dev_error(blkif->be->dev, err, "get blkback dev name");
++ return;
++ }
++
++ blkif->xenblkd = kthread_run(blkif_schedule, blkif, name);
++ if (IS_ERR(blkif->xenblkd)) {
++ err = PTR_ERR(blkif->xenblkd);
++ blkif->xenblkd = NULL;
++ xenbus_dev_error(blkif->be->dev, err, "start xenblkd");
++ }
+}
+
-+static void srat_detect_node(void)
-+{
-+#ifdef CONFIG_NUMA
-+ unsigned node;
-+ int cpu = smp_processor_id();
-+ int apicid = hard_smp_processor_id();
+
-+ /* Don't do the funky fallback heuristics the AMD version employs
-+ for now. */
-+ node = apicid_to_node[apicid];
-+ if (node == NUMA_NO_NODE)
-+ node = first_node(node_online_map);
-+ numa_set_node(cpu, node);
++/****************************************************************
++ * sysfs interface for VBD I/O requests
++ */
++
++#define VBD_SHOW(name, format, args...) \
++ static ssize_t show_##name(struct device *_dev, \
++ struct device_attribute *attr, \
++ char *buf) \
++ { \
++ struct xenbus_device *dev = to_xenbus_device(_dev); \
++ struct backend_info *be = dev->dev.driver_data; \
++ \
++ return sprintf(buf, format, ##args); \
++ } \
++ static DEVICE_ATTR(name, S_IRUGO, show_##name, NULL)
++
++VBD_SHOW(oo_req, "%d\n", be->blkif->st_oo_req);
++VBD_SHOW(rd_req, "%d\n", be->blkif->st_rd_req);
++VBD_SHOW(wr_req, "%d\n", be->blkif->st_wr_req);
++VBD_SHOW(br_req, "%d\n", be->blkif->st_br_req);
++VBD_SHOW(rd_sect, "%d\n", be->blkif->st_rd_sect);
++VBD_SHOW(wr_sect, "%d\n", be->blkif->st_wr_sect);
++
++static struct attribute *vbdstat_attrs[] = {
++ &dev_attr_oo_req.attr,
++ &dev_attr_rd_req.attr,
++ &dev_attr_wr_req.attr,
++ &dev_attr_br_req.attr,
++ &dev_attr_rd_sect.attr,
++ &dev_attr_wr_sect.attr,
++ NULL
++};
++
++static struct attribute_group vbdstat_group = {
++ .name = "statistics",
++ .attrs = vbdstat_attrs,
++};
+
-+ if (acpi_numa > 0)
-+ printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node);
-+#endif
-+}
++VBD_SHOW(physical_device, "%x:%x\n", be->major, be->minor);
++VBD_SHOW(mode, "%s\n", be->mode);
+
-+static void __cpuinit init_intel(struct cpuinfo_x86 *c)
++int xenvbd_sysfs_addif(struct xenbus_device *dev)
+{
-+ /* Cache sizes */
-+ unsigned n;
++ int error;
++
++ error = device_create_file(&dev->dev, &dev_attr_physical_device);
++ if (error)
++ goto fail1;
+
-+ init_intel_cacheinfo(c);
-+ if (c->cpuid_level > 9 ) {
-+ unsigned eax = cpuid_eax(10);
-+ /* Check for version and the number of counters */
-+ if ((eax & 0xff) && (((eax>>8) & 0xff) > 1))
-+ set_bit(X86_FEATURE_ARCH_PERFMON, &c->x86_capability);
-+ }
++ error = device_create_file(&dev->dev, &dev_attr_mode);
++ if (error)
++ goto fail2;
+
-+ n = c->extended_cpuid_level;
-+ if (n >= 0x80000008) {
-+ unsigned eax = cpuid_eax(0x80000008);
-+ c->x86_virt_bits = (eax >> 8) & 0xff;
-+ c->x86_phys_bits = eax & 0xff;
-+ /* CPUID workaround for Intel 0F34 CPU */
-+ if (c->x86_vendor == X86_VENDOR_INTEL &&
-+ c->x86 == 0xF && c->x86_model == 0x3 &&
-+ c->x86_mask == 0x4)
-+ c->x86_phys_bits = 36;
-+ }
++ error = sysfs_create_group(&dev->dev.kobj, &vbdstat_group);
++ if (error)
++ goto fail3;
+
-+ if (c->x86 == 15)
-+ c->x86_cache_alignment = c->x86_clflush_size * 2;
-+ if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
-+ (c->x86 == 0x6 && c->x86_model >= 0x0e))
-+ set_bit(X86_FEATURE_CONSTANT_TSC, &c->x86_capability);
-+ set_bit(X86_FEATURE_SYNC_RDTSC, &c->x86_capability);
-+ c->x86_max_cores = intel_num_cpu_cores(c);
++ return 0;
+
-+ srat_detect_node();
++fail3: sysfs_remove_group(&dev->dev.kobj, &vbdstat_group);
++fail2: device_remove_file(&dev->dev, &dev_attr_mode);
++fail1: device_remove_file(&dev->dev, &dev_attr_physical_device);
++ return error;
+}
+
-+static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c)
++void xenvbd_sysfs_delif(struct xenbus_device *dev)
+{
-+ char *v = c->x86_vendor_id;
-+
-+ if (!strcmp(v, "AuthenticAMD"))
-+ c->x86_vendor = X86_VENDOR_AMD;
-+ else if (!strcmp(v, "GenuineIntel"))
-+ c->x86_vendor = X86_VENDOR_INTEL;
-+ else
-+ c->x86_vendor = X86_VENDOR_UNKNOWN;
++ sysfs_remove_group(&dev->dev.kobj, &vbdstat_group);
++ device_remove_file(&dev->dev, &dev_attr_mode);
++ device_remove_file(&dev->dev, &dev_attr_physical_device);
+}
+
-+struct cpu_model_info {
-+ int vendor;
-+ int family;
-+ char *model_names[16];
-+};
-+
-+/* Do some early cpuid on the boot CPU to get some parameter that are
-+ needed before check_bugs. Everything advanced is in identify_cpu
-+ below. */
-+void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c)
++static int blkback_remove(struct xenbus_device *dev)
+{
-+ u32 tfms;
++ struct backend_info *be = dev->dev.driver_data;
+
-+ c->loops_per_jiffy = loops_per_jiffy;
-+ c->x86_cache_size = -1;
-+ c->x86_vendor = X86_VENDOR_UNKNOWN;
-+ c->x86_model = c->x86_mask = 0; /* So far unknown... */
-+ c->x86_vendor_id[0] = '\0'; /* Unset */
-+ c->x86_model_id[0] = '\0'; /* Unset */
-+ c->x86_clflush_size = 64;
-+ c->x86_cache_alignment = c->x86_clflush_size;
-+ c->x86_max_cores = 1;
-+ c->extended_cpuid_level = 0;
-+ memset(&c->x86_capability, 0, sizeof c->x86_capability);
++ DPRINTK("");
+
-+ /* Get vendor name */
-+ cpuid(0x00000000, (unsigned int *)&c->cpuid_level,
-+ (unsigned int *)&c->x86_vendor_id[0],
-+ (unsigned int *)&c->x86_vendor_id[8],
-+ (unsigned int *)&c->x86_vendor_id[4]);
-+
-+ get_cpu_vendor(c);
++ if (be->major || be->minor)
++ xenvbd_sysfs_delif(dev);
+
-+ /* Initialize the standard set of capabilities */
-+ /* Note that the vendor-specific code below might override */
++ if (be->backend_watch.node) {
++ unregister_xenbus_watch(&be->backend_watch);
++ kfree(be->backend_watch.node);
++ be->backend_watch.node = NULL;
++ }
+
-+ /* Intel-defined flags: level 0x00000001 */
-+ if (c->cpuid_level >= 0x00000001) {
-+ __u32 misc;
-+ cpuid(0x00000001, &tfms, &misc, &c->x86_capability[4],
-+ &c->x86_capability[0]);
-+ c->x86 = (tfms >> 8) & 0xf;
-+ c->x86_model = (tfms >> 4) & 0xf;
-+ c->x86_mask = tfms & 0xf;
-+ if (c->x86 == 0xf)
-+ c->x86 += (tfms >> 20) & 0xff;
-+ if (c->x86 >= 0x6)
-+ c->x86_model += ((tfms >> 16) & 0xF) << 4;
-+ if (c->x86_capability[0] & (1<<19))
-+ c->x86_clflush_size = ((misc >> 8) & 0xff) * 8;
-+ } else {
-+ /* Have CPUID level 0 only - unheard of */
-+ c->x86 = 4;
++ if (be->blkif) {
++ blkif_disconnect(be->blkif);
++ vbd_free(&be->blkif->vbd);
++ blkif_free(be->blkif);
++ be->blkif = NULL;
+ }
+
-+#ifdef CONFIG_SMP
-+ c->phys_proc_id = (cpuid_ebx(1) >> 24) & 0xff;
-+#endif
++ kfree(be);
++ dev->dev.driver_data = NULL;
++ return 0;
+}
+
-+/*
-+ * This does the hard work of actually picking apart the CPU stuff...
-+ */
-+void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
++int blkback_barrier(struct xenbus_transaction xbt,
++ struct backend_info *be, int state)
+{
-+ int i;
-+ u32 xlvl;
++ struct xenbus_device *dev = be->dev;
++ int err;
+
-+ early_identify_cpu(c);
++ err = xenbus_printf(xbt, dev->nodename, "feature-barrier",
++ "%d", state);
++ if (err)
++ xenbus_dev_fatal(dev, err, "writing feature-barrier");
+
-+ /* AMD-defined flags: level 0x80000001 */
-+ xlvl = cpuid_eax(0x80000000);
-+ c->extended_cpuid_level = xlvl;
-+ if ((xlvl & 0xffff0000) == 0x80000000) {
-+ if (xlvl >= 0x80000001) {
-+ c->x86_capability[1] = cpuid_edx(0x80000001);
-+ c->x86_capability[6] = cpuid_ecx(0x80000001);
-+ }
-+ if (xlvl >= 0x80000004)
-+ get_model_name(c); /* Default name */
-+ }
++ return err;
++}
+
-+ /* Transmeta-defined flags: level 0x80860001 */
-+ xlvl = cpuid_eax(0x80860000);
-+ if ((xlvl & 0xffff0000) == 0x80860000) {
-+ /* Don't set x86_cpuid_level here for now to not confuse. */
-+ if (xlvl >= 0x80860001)
-+ c->x86_capability[2] = cpuid_edx(0x80860001);
++/**
++ * Entry point to this code when a new device is created. Allocate the basic
++ * structures, and watch the store waiting for the hotplug scripts to tell us
++ * the device's physical major and minor numbers. Switch to InitWait.
++ */
++static int blkback_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ int err;
++ struct backend_info *be = kzalloc(sizeof(struct backend_info),
++ GFP_KERNEL);
++ if (!be) {
++ xenbus_dev_fatal(dev, -ENOMEM,
++ "allocating backend structure");
++ return -ENOMEM;
+ }
++ be->dev = dev;
++ dev->dev.driver_data = be;
+
-+ c->apicid = phys_pkg_id(0);
-+
-+ /*
-+ * Vendor-specific initialization. In this section we
-+ * canonicalize the feature flags, meaning if there are
-+ * features a certain CPU supports which CPUID doesn't
-+ * tell us, CPUID claiming incorrect flags, or other bugs,
-+ * we handle them here.
-+ *
-+ * At the end of this section, c->x86_capability better
-+ * indicate the features this CPU genuinely supports!
-+ */
-+ switch (c->x86_vendor) {
-+ case X86_VENDOR_AMD:
-+ init_amd(c);
-+ break;
-+
-+ case X86_VENDOR_INTEL:
-+ init_intel(c);
-+ break;
-+
-+ case X86_VENDOR_UNKNOWN:
-+ default:
-+ display_cacheinfo(c);
-+ break;
++ be->blkif = blkif_alloc(dev->otherend_id);
++ if (IS_ERR(be->blkif)) {
++ err = PTR_ERR(be->blkif);
++ be->blkif = NULL;
++ xenbus_dev_fatal(dev, err, "creating block interface");
++ goto fail;
+ }
+
-+ select_idle_routine(c);
-+ detect_ht(c);
++ /* setup back pointer */
++ be->blkif->be = be;
+
-+ /*
-+ * On SMP, boot_cpu_data holds the common feature set between
-+ * all CPUs; so make sure that we indicate which features are
-+ * common between the CPUs. The first time this routine gets
-+ * executed, c == &boot_cpu_data.
-+ */
-+ if (c != &boot_cpu_data) {
-+ /* AND the already accumulated flags with these */
-+ for (i = 0 ; i < NCAPINTS ; i++)
-+ boot_cpu_data.x86_capability[i] &= c->x86_capability[i];
-+ }
++ err = xenbus_watch_path2(dev, dev->nodename, "physical-device",
++ &be->backend_watch, backend_changed);
++ if (err)
++ goto fail;
+
-+#ifdef CONFIG_X86_MCE
-+ mcheck_init(c);
-+#endif
-+ if (c == &boot_cpu_data)
-+ mtrr_bp_init();
-+ else
-+ mtrr_ap_init();
-+#ifdef CONFIG_NUMA
-+ numa_add_cpu(smp_processor_id());
-+#endif
-+}
-+
++ err = xenbus_switch_state(dev, XenbusStateInitWait);
++ if (err)
++ goto fail;
+
-+void __cpuinit print_cpu_info(struct cpuinfo_x86 *c)
-+{
-+ if (c->x86_model_id[0])
-+ printk("%s", c->x86_model_id);
++ return 0;
+
-+ if (c->x86_mask || c->cpuid_level >= 0)
-+ printk(" stepping %02x\n", c->x86_mask);
-+ else
-+ printk("\n");
++fail:
++ DPRINTK("failed");
++ blkback_remove(dev);
++ return err;
+}
+
-+/*
-+ * Get CPU information for use by the procfs.
-+ */
+
-+static int show_cpuinfo(struct seq_file *m, void *v)
++/**
++ * Callback received when the hotplug scripts have placed the physical-device
++ * node. Read it and the mode node, and create a vbd. If the frontend is
++ * ready, connect.
++ */
++static void backend_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
+{
-+ struct cpuinfo_x86 *c = v;
++ int err;
++ unsigned major;
++ unsigned minor;
++ struct backend_info *be
++ = container_of(watch, struct backend_info, backend_watch);
++ struct xenbus_device *dev = be->dev;
++ int cdrom = 0;
++ char *device_type;
+
-+ /*
-+ * These flag bits must match the definitions in <asm/cpufeature.h>.
-+ * NULL means this bit is undefined or reserved; either way it doesn't
-+ * have meaning as far as Linux is concerned. Note that it's important
-+ * to realize there is a difference between this table and CPUID -- if
-+ * applications want to get the raw CPUID data, they should access
-+ * /dev/cpu/<cpu_nr>/cpuid instead.
-+ */
-+ static char *x86_cap_flags[] = {
-+ /* Intel-defined */
-+ "fpu", "vme", "de", "pse", "tsc", "msr", "pae", "mce",
-+ "cx8", "apic", NULL, "sep", "mtrr", "pge", "mca", "cmov",
-+ "pat", "pse36", "pn", "clflush", NULL, "dts", "acpi", "mmx",
-+ "fxsr", "sse", "sse2", "ss", "ht", "tm", "ia64", NULL,
++ DPRINTK("");
+
-+ /* AMD-defined */
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, "syscall", NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, "nx", NULL, "mmxext", NULL,
-+ NULL, "fxsr_opt", NULL, "rdtscp", NULL, "lm", "3dnowext", "3dnow",
++ err = xenbus_scanf(XBT_NIL, dev->nodename, "physical-device", "%x:%x",
++ &major, &minor);
++ if (XENBUS_EXIST_ERR(err)) {
++ /* Since this watch will fire once immediately after it is
++ registered, we expect this. Ignore it, and wait for the
++ hotplug scripts. */
++ return;
++ }
++ if (err != 2) {
++ xenbus_dev_fatal(dev, err, "reading physical-device");
++ return;
++ }
+
-+ /* Transmeta-defined */
-+ "recovery", "longrun", NULL, "lrti", NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ if ((be->major || be->minor) &&
++ ((be->major != major) || (be->minor != minor))) {
++ printk(KERN_WARNING
++ "blkback: changing physical device (from %x:%x to "
++ "%x:%x) not supported.\n", be->major, be->minor,
++ major, minor);
++ return;
++ }
+
-+ /* Other (Linux-defined) */
-+ "cxmmx", NULL, "cyrix_arr", "centaur_mcr", NULL,
-+ "constant_tsc", NULL, NULL,
-+ "up", NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ be->mode = xenbus_read(XBT_NIL, dev->nodename, "mode", NULL);
++ if (IS_ERR(be->mode)) {
++ err = PTR_ERR(be->mode);
++ be->mode = NULL;
++ xenbus_dev_fatal(dev, err, "reading mode");
++ return;
++ }
+
-+ /* Intel-defined (#2) */
-+ "pni", NULL, NULL, "monitor", "ds_cpl", "vmx", "smx", "est",
-+ "tm2", NULL, "cid", NULL, NULL, "cx16", "xtpr", NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ device_type = xenbus_read(XBT_NIL, dev->otherend, "device-type", NULL);
++ if (!IS_ERR(device_type)) {
++ cdrom = strcmp(device_type, "cdrom") == 0;
++ kfree(device_type);
++ }
+
-+ /* VIA/Cyrix/Centaur-defined */
-+ NULL, NULL, "rng", "rng_en", NULL, NULL, "ace", "ace_en",
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ if (be->major == 0 && be->minor == 0) {
++ /* Front end dir is a number, which is used as the handle. */
+
-+ /* AMD-defined (#2) */
-+ "lahf_lm", "cmp_legacy", "svm", NULL, "cr8_legacy", NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ };
-+ static char *x86_power_flags[] = {
-+ "ts", /* temperature sensor */
-+ "fid", /* frequency id control */
-+ "vid", /* voltage id control */
-+ "ttp", /* thermal trip */
-+ "tm",
-+ "stc",
-+ NULL,
-+ /* nothing */ /* constant_tsc - moved to flags */
-+ };
++ char *p = strrchr(dev->otherend, '/') + 1;
++ long handle = simple_strtoul(p, NULL, 0);
+
++ be->major = major;
++ be->minor = minor;
+
-+#ifdef CONFIG_SMP
-+ if (!cpu_online(c-cpu_data))
-+ return 0;
-+#endif
++ err = vbd_create(be->blkif, handle, major, minor,
++ (NULL == strchr(be->mode, 'w')), cdrom);
++ if (err) {
++ be->major = be->minor = 0;
++ xenbus_dev_fatal(dev, err, "creating vbd structure");
++ return;
++ }
+
-+ seq_printf(m,"processor\t: %u\n"
-+ "vendor_id\t: %s\n"
-+ "cpu family\t: %d\n"
-+ "model\t\t: %d\n"
-+ "model name\t: %s\n",
-+ (unsigned)(c-cpu_data),
-+ c->x86_vendor_id[0] ? c->x86_vendor_id : "unknown",
-+ c->x86,
-+ (int)c->x86_model,
-+ c->x86_model_id[0] ? c->x86_model_id : "unknown");
-+
-+ if (c->x86_mask || c->cpuid_level >= 0)
-+ seq_printf(m, "stepping\t: %d\n", c->x86_mask);
-+ else
-+ seq_printf(m, "stepping\t: unknown\n");
-+
-+ if (cpu_has(c,X86_FEATURE_TSC)) {
-+ unsigned int freq = cpufreq_quick_get((unsigned)(c-cpu_data));
-+ if (!freq)
-+ freq = cpu_khz;
-+ seq_printf(m, "cpu MHz\t\t: %u.%03u\n",
-+ freq / 1000, (freq % 1000));
-+ }
++ err = xenvbd_sysfs_addif(dev);
++ if (err) {
++ vbd_free(&be->blkif->vbd);
++ be->major = be->minor = 0;
++ xenbus_dev_fatal(dev, err, "creating sysfs entries");
++ return;
++ }
+
-+ /* Cache size */
-+ if (c->x86_cache_size >= 0)
-+ seq_printf(m, "cache size\t: %d KB\n", c->x86_cache_size);
-+
-+#ifdef CONFIG_SMP
-+ if (smp_num_siblings * c->x86_max_cores > 1) {
-+ int cpu = c - cpu_data;
-+ seq_printf(m, "physical id\t: %d\n", c->phys_proc_id);
-+ seq_printf(m, "siblings\t: %d\n", cpus_weight(cpu_core_map[cpu]));
-+ seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id);
-+ seq_printf(m, "cpu cores\t: %d\n", c->booted_cores);
++ /* We're potentially connected now */
++ update_blkif_status(be->blkif);
+ }
-+#endif
++}
+
-+ seq_printf(m,
-+ "fpu\t\t: yes\n"
-+ "fpu_exception\t: yes\n"
-+ "cpuid level\t: %d\n"
-+ "wp\t\t: yes\n"
-+ "flags\t\t:",
-+ c->cpuid_level);
+
-+ {
-+ int i;
-+ for ( i = 0 ; i < 32*NCAPINTS ; i++ )
-+ if (cpu_has(c, i) && x86_cap_flags[i] != NULL)
-+ seq_printf(m, " %s", x86_cap_flags[i]);
-+ }
-+
-+ seq_printf(m, "\nbogomips\t: %lu.%02lu\n",
-+ c->loops_per_jiffy/(500000/HZ),
-+ (c->loops_per_jiffy/(5000/HZ)) % 100);
++/**
++ * Callback received when the frontend's state changes.
++ */
++static void frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state)
++{
++ struct backend_info *be = dev->dev.driver_data;
++ int err;
+
-+ if (c->x86_tlbsize > 0)
-+ seq_printf(m, "TLB size\t: %d 4K pages\n", c->x86_tlbsize);
-+ seq_printf(m, "clflush size\t: %d\n", c->x86_clflush_size);
-+ seq_printf(m, "cache_alignment\t: %d\n", c->x86_cache_alignment);
++ DPRINTK("%s", xenbus_strstate(frontend_state));
+
-+ seq_printf(m, "address sizes\t: %u bits physical, %u bits virtual\n",
-+ c->x86_phys_bits, c->x86_virt_bits);
++ switch (frontend_state) {
++ case XenbusStateInitialising:
++ if (dev->state == XenbusStateClosed) {
++ printk(KERN_INFO "%s: %s: prepare for reconnect\n",
++ __FUNCTION__, dev->nodename);
++ xenbus_switch_state(dev, XenbusStateInitWait);
++ }
++ break;
+
-+ seq_printf(m, "power management:");
-+ {
-+ unsigned i;
-+ for (i = 0; i < 32; i++)
-+ if (c->x86_power & (1 << i)) {
-+ if (i < ARRAY_SIZE(x86_power_flags) &&
-+ x86_power_flags[i])
-+ seq_printf(m, "%s%s",
-+ x86_power_flags[i][0]?" ":"",
-+ x86_power_flags[i]);
-+ else
-+ seq_printf(m, " [%d]", i);
-+ }
-+ }
++ case XenbusStateInitialised:
++ case XenbusStateConnected:
++ /* Ensure we connect even when two watches fire in
++ close successsion and we miss the intermediate value
++ of frontend_state. */
++ if (dev->state == XenbusStateConnected)
++ break;
+
-+ seq_printf(m, "\n\n");
++ err = connect_ring(be);
++ if (err)
++ break;
++ update_blkif_status(be->blkif);
++ break;
+
-+ return 0;
-+}
++ case XenbusStateClosing:
++ blkif_disconnect(be->blkif);
++ xenbus_switch_state(dev, XenbusStateClosing);
++ break;
+
-+static void *c_start(struct seq_file *m, loff_t *pos)
-+{
-+ return *pos < NR_CPUS ? cpu_data + *pos : NULL;
-+}
++ case XenbusStateClosed:
++ xenbus_switch_state(dev, XenbusStateClosed);
++ if (xenbus_dev_is_online(dev))
++ break;
++ /* fall through if not online */
++ case XenbusStateUnknown:
++ device_unregister(&dev->dev);
++ break;
+
-+static void *c_next(struct seq_file *m, void *v, loff_t *pos)
-+{
-+ ++*pos;
-+ return c_start(m, pos);
++ default:
++ xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
++ frontend_state);
++ break;
++ }
+}
+
-+static void c_stop(struct seq_file *m, void *v)
-+{
-+}
+
-+struct seq_operations cpuinfo_op = {
-+ .start =c_start,
-+ .next = c_next,
-+ .stop = c_stop,
-+ .show = show_cpuinfo,
-+};
++/* ** Connection ** */
+
-+#if defined(CONFIG_INPUT_PCSPKR) || defined(CONFIG_INPUT_PCSPKR_MODULE)
-+#include <linux/platform_device.h>
-+static __init int add_pcspkr(void)
-+{
-+ struct platform_device *pd;
-+ int ret;
+
-+ if (!is_initial_xendomain())
-+ return 0;
++/**
++ * Write the physical details regarding the block device to the store, and
++ * switch to Connected state.
++ */
++static void connect(struct backend_info *be)
++{
++ struct xenbus_transaction xbt;
++ int err;
++ struct xenbus_device *dev = be->dev;
+
-+ pd = platform_device_alloc("pcspkr", -1);
-+ if (!pd)
-+ return -ENOMEM;
++ DPRINTK("%s", dev->otherend);
+
-+ ret = platform_device_add(pd);
-+ if (ret)
-+ platform_device_put(pd);
++ /* Supply the information about the device the frontend needs */
++again:
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "starting transaction");
++ return;
++ }
+
-+ return ret;
-+}
-+device_initcall(add_pcspkr);
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/smp-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/smp-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/smp-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/smp-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,600 @@
-+/*
-+ * Intel SMP support routines.
-+ *
-+ * (c) 1995 Alan Cox, Building #3 <alan at redhat.com>
-+ * (c) 1998-99, 2000 Ingo Molnar <mingo at redhat.com>
-+ * (c) 2002,2003 Andi Kleen, SuSE Labs.
-+ *
-+ * This code is released under the GNU General Public License version 2 or
-+ * later.
-+ */
++ err = blkback_barrier(xbt, be, 1);
++ if (err)
++ goto abort;
+
-+#include <linux/init.h>
++ err = xenbus_printf(xbt, dev->nodename, "sectors", "%llu",
++ vbd_size(&be->blkif->vbd));
++ if (err) {
++ xenbus_dev_fatal(dev, err, "writing %s/sectors",
++ dev->nodename);
++ goto abort;
++ }
+
-+#include <linux/mm.h>
-+#include <linux/delay.h>
-+#include <linux/spinlock.h>
-+#include <linux/smp_lock.h>
-+#include <linux/smp.h>
-+#include <linux/kernel_stat.h>
-+#include <linux/mc146818rtc.h>
-+#include <linux/interrupt.h>
++ /* FIXME: use a typename instead */
++ err = xenbus_printf(xbt, dev->nodename, "info", "%u",
++ vbd_info(&be->blkif->vbd));
++ if (err) {
++ xenbus_dev_fatal(dev, err, "writing %s/info",
++ dev->nodename);
++ goto abort;
++ }
++ err = xenbus_printf(xbt, dev->nodename, "sector-size", "%lu",
++ vbd_secsize(&be->blkif->vbd));
++ if (err) {
++ xenbus_dev_fatal(dev, err, "writing %s/sector-size",
++ dev->nodename);
++ goto abort;
++ }
+
-+#include <asm/mtrr.h>
-+#include <asm/pgalloc.h>
-+#include <asm/tlbflush.h>
-+#include <asm/mach_apic.h>
-+#include <asm/mmu_context.h>
-+#include <asm/proto.h>
-+#include <asm/apicdef.h>
-+#include <asm/idle.h>
-+#ifdef CONFIG_XEN
-+#include <xen/evtchn.h>
-+#endif
++ err = xenbus_transaction_end(xbt, 0);
++ if (err == -EAGAIN)
++ goto again;
++ if (err)
++ xenbus_dev_fatal(dev, err, "ending transaction");
+
-+#ifndef CONFIG_XEN
-+/*
-+ * Smarter SMP flushing macros.
-+ * c/o Linus Torvalds.
-+ *
-+ * These mean you can really definitely utterly forget about
-+ * writing to user space from interrupts. (Its not allowed anyway).
-+ *
-+ * Optimizations Manfred Spraul <manfred at colorfullife.com>
-+ *
-+ * More scalable flush, from Andi Kleen
-+ *
-+ * To avoid global state use 8 different call vectors.
-+ * Each CPU uses a specific vector to trigger flushes on other
-+ * CPUs. Depending on the received vector the target CPUs look into
-+ * the right per cpu variable for the flush data.
-+ *
-+ * With more than 8 CPUs they are hashed to the 8 available
-+ * vectors. The limited global vector space forces us to this right now.
-+ * In future when interrupts are split into per CPU domains this could be
-+ * fixed, at the cost of triggering multiple IPIs in some cases.
-+ */
++ err = xenbus_switch_state(dev, XenbusStateConnected);
++ if (err)
++ xenbus_dev_fatal(dev, err, "switching to Connected state",
++ dev->nodename);
+
-+union smp_flush_state {
-+ struct {
-+ cpumask_t flush_cpumask;
-+ struct mm_struct *flush_mm;
-+ unsigned long flush_va;
-+#define FLUSH_ALL -1ULL
-+ spinlock_t tlbstate_lock;
-+ };
-+ char pad[SMP_CACHE_BYTES];
-+} ____cacheline_aligned;
++ return;
++ abort:
++ xenbus_transaction_end(xbt, 1);
++}
+
-+/* State is put into the per CPU data section, but padded
-+ to a full cache line because other CPUs can access it and we don't
-+ want false sharing in the per cpu data segment. */
-+static DEFINE_PER_CPU(union smp_flush_state, flush_state);
-+#endif
+
-+/*
-+ * We cannot call mmdrop() because we are in interrupt context,
-+ * instead update mm->cpu_vm_mask.
-+ */
-+static inline void leave_mm(unsigned long cpu)
++static int connect_ring(struct backend_info *be)
+{
-+ if (read_pda(mmu_state) == TLBSTATE_OK)
-+ BUG();
-+ cpu_clear(cpu, read_pda(active_mm)->cpu_vm_mask);
-+ load_cr3(swapper_pg_dir);
-+}
-+
-+#ifndef CONFIG_XEN
-+/*
-+ *
-+ * The flush IPI assumes that a thread switch happens in this order:
-+ * [cpu0: the cpu that switches]
-+ * 1) switch_mm() either 1a) or 1b)
-+ * 1a) thread switch to a different mm
-+ * 1a1) cpu_clear(cpu, old_mm->cpu_vm_mask);
-+ * Stop ipi delivery for the old mm. This is not synchronized with
-+ * the other cpus, but smp_invalidate_interrupt ignore flush ipis
-+ * for the wrong mm, and in the worst case we perform a superfluous
-+ * tlb flush.
-+ * 1a2) set cpu mmu_state to TLBSTATE_OK
-+ * Now the smp_invalidate_interrupt won't call leave_mm if cpu0
-+ * was in lazy tlb mode.
-+ * 1a3) update cpu active_mm
-+ * Now cpu0 accepts tlb flushes for the new mm.
-+ * 1a4) cpu_set(cpu, new_mm->cpu_vm_mask);
-+ * Now the other cpus will send tlb flush ipis.
-+ * 1a4) change cr3.
-+ * 1b) thread switch without mm change
-+ * cpu active_mm is correct, cpu0 already handles
-+ * flush ipis.
-+ * 1b1) set cpu mmu_state to TLBSTATE_OK
-+ * 1b2) test_and_set the cpu bit in cpu_vm_mask.
-+ * Atomically set the bit [other cpus will start sending flush ipis],
-+ * and test the bit.
-+ * 1b3) if the bit was 0: leave_mm was called, flush the tlb.
-+ * 2) switch %%esp, ie current
-+ *
-+ * The interrupt must handle 2 special cases:
-+ * - cr3 is changed before %%esp, ie. it cannot use current->{active_,}mm.
-+ * - the cpu performs speculative tlb reads, i.e. even if the cpu only
-+ * runs in kernel space, the cpu could load tlb entries for user space
-+ * pages.
-+ *
-+ * The good news is that cpu mmu_state is local to each cpu, no
-+ * write/read ordering problems.
-+ */
++ struct xenbus_device *dev = be->dev;
++ unsigned long ring_ref;
++ unsigned int evtchn;
++ char protocol[64] = "";
++ int err;
+
-+/*
-+ * TLB flush IPI:
-+ *
-+ * 1) Flush the tlb entries if the cpu uses the mm that's being flushed.
-+ * 2) Leave the mm if we are in the lazy tlb mode.
-+ *
-+ * Interrupts are disabled.
-+ */
++ DPRINTK("%s", dev->otherend);
+
-+asmlinkage void smp_invalidate_interrupt(struct pt_regs *regs)
-+{
-+ int cpu;
-+ int sender;
-+ union smp_flush_state *f;
++ err = xenbus_gather(XBT_NIL, dev->otherend, "ring-ref", "%lu", &ring_ref,
++ "event-channel", "%u", &evtchn, NULL);
++ if (err) {
++ xenbus_dev_fatal(dev, err,
++ "reading %s/ring-ref and event-channel",
++ dev->otherend);
++ return err;
++ }
+
-+ cpu = smp_processor_id();
-+ /*
-+ * orig_rax contains the negated interrupt vector.
-+ * Use that to determine where the sender put the data.
-+ */
-+ sender = ~regs->orig_rax - INVALIDATE_TLB_VECTOR_START;
-+ f = &per_cpu(flush_state, sender);
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_NATIVE;
++ err = xenbus_gather(XBT_NIL, dev->otherend, "protocol",
++ "%63s", protocol, NULL);
++ if (err)
++ strcpy(protocol, "unspecified, assuming native");
++ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_NATIVE))
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_NATIVE;
++ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_X86_32))
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_X86_32;
++ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_X86_64))
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_X86_64;
++ else {
++ xenbus_dev_fatal(dev, err, "unknown fe protocol %s", protocol);
++ return -1;
++ }
++ printk(KERN_INFO
++ "blkback: ring-ref %ld, event-channel %d, protocol %d (%s)\n",
++ ring_ref, evtchn, be->blkif->blk_protocol, protocol);
+
-+ if (!cpu_isset(cpu, f->flush_cpumask))
-+ goto out;
-+ /*
-+ * This was a BUG() but until someone can quote me the
-+ * line from the intel manual that guarantees an IPI to
-+ * multiple CPUs is retried _only_ on the erroring CPUs
-+ * its staying as a return
-+ *
-+ * BUG();
-+ */
-+
-+ if (f->flush_mm == read_pda(active_mm)) {
-+ if (read_pda(mmu_state) == TLBSTATE_OK) {
-+ if (f->flush_va == FLUSH_ALL)
-+ local_flush_tlb();
-+ else
-+ __flush_tlb_one(f->flush_va);
-+ } else
-+ leave_mm(cpu);
++ /* Map the shared frame, irq etc. */
++ err = blkif_map(be->blkif, ring_ref, evtchn);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "mapping ring-ref %lu port %u",
++ ring_ref, evtchn);
++ return err;
+ }
-+out:
-+ ack_APIC_irq();
-+ cpu_clear(cpu, f->flush_cpumask);
++
++ return 0;
+}
+
-+static void flush_tlb_others(cpumask_t cpumask, struct mm_struct *mm,
-+ unsigned long va)
-+{
-+ int sender;
-+ union smp_flush_state *f;
+
-+ /* Caller has disabled preemption */
-+ sender = smp_processor_id() % NUM_INVALIDATE_TLB_VECTORS;
-+ f = &per_cpu(flush_state, sender);
++/* ** Driver Registration ** */
+
-+ /* Could avoid this lock when
-+ num_online_cpus() <= NUM_INVALIDATE_TLB_VECTORS, but it is
-+ probably not worth checking this for a cache-hot lock. */
-+ spin_lock(&f->tlbstate_lock);
+
-+ f->flush_mm = mm;
-+ f->flush_va = va;
-+ cpus_or(f->flush_cpumask, cpumask, f->flush_cpumask);
++static const struct xenbus_device_id blkback_ids[] = {
++ { "vbd" },
++ { "" }
++};
+
-+ /*
-+ * We have to send the IPI only to
-+ * CPUs affected.
-+ */
-+ send_IPI_mask(cpumask, INVALIDATE_TLB_VECTOR_START + sender);
+
-+ while (!cpus_empty(f->flush_cpumask))
-+ cpu_relax();
++static struct xenbus_driver blkback = {
++ .name = "vbd",
++ .owner = THIS_MODULE,
++ .ids = blkback_ids,
++ .probe = blkback_probe,
++ .remove = blkback_remove,
++ .otherend_changed = frontend_changed
++};
+
-+ f->flush_mm = NULL;
-+ f->flush_va = 0;
-+ spin_unlock(&f->tlbstate_lock);
-+}
+
-+int __cpuinit init_smp_flush(void)
++void blkif_xenbus_init(void)
+{
-+ int i;
-+ for_each_cpu_mask(i, cpu_possible_map) {
-+ spin_lock_init(&per_cpu(flush_state, i).tlbstate_lock);
-+ }
-+ return 0;
++ xenbus_register_backend(&blkback);
+}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blkfront/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blkfront/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,5 @@
+
-+core_initcall(init_smp_flush);
-+
-+void flush_tlb_current_task(void)
-+{
-+ struct mm_struct *mm = current->mm;
-+ cpumask_t cpu_mask;
++obj-$(CONFIG_XEN_BLKDEV_FRONTEND) := xenblk.o
+
-+ preempt_disable();
-+ cpu_mask = mm->cpu_vm_mask;
-+ cpu_clear(smp_processor_id(), cpu_mask);
++xenblk-objs := blkfront.o vbd.o
+
-+ local_flush_tlb();
-+ if (!cpus_empty(cpu_mask))
-+ flush_tlb_others(cpu_mask, mm, FLUSH_ALL);
-+ preempt_enable();
-+}
-+EXPORT_SYMBOL(flush_tlb_current_task);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blkfront/blkfront.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blkfront/blkfront.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,936 @@
++/******************************************************************************
++ * blkfront.c
++ *
++ * XenLinux virtual block-device driver.
++ *
++ * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
++ * Modifications by Mark A. Williamson are (c) Intel Research Cambridge
++ * Copyright (c) 2004, Christian Limpach
++ * Copyright (c) 2004, Andrew Warfield
++ * Copyright (c) 2005, Christopher Clark
++ * Copyright (c) 2005, XenSource Ltd
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+void flush_tlb_mm (struct mm_struct * mm)
-+{
-+ cpumask_t cpu_mask;
++#include <linux/version.h>
++#include "block.h"
++#include <linux/cdrom.h>
++#include <linux/sched.h>
++#include <linux/interrupt.h>
++#include <scsi/scsi.h>
++#include <xen/evtchn.h>
++#include <xen/xenbus.h>
++#include <xen/interface/grant_table.h>
++#include <xen/interface/io/protocols.h>
++#include <xen/gnttab.h>
++#include <asm/hypervisor.h>
++#include <asm/maddr.h>
+
-+ preempt_disable();
-+ cpu_mask = mm->cpu_vm_mask;
-+ cpu_clear(smp_processor_id(), cpu_mask);
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
+
-+ if (current->active_mm == mm) {
-+ if (current->mm)
-+ local_flush_tlb();
-+ else
-+ leave_mm(smp_processor_id());
-+ }
-+ if (!cpus_empty(cpu_mask))
-+ flush_tlb_others(cpu_mask, mm, FLUSH_ALL);
++#define BLKIF_STATE_DISCONNECTED 0
++#define BLKIF_STATE_CONNECTED 1
++#define BLKIF_STATE_SUSPENDED 2
+
-+ preempt_enable();
-+}
-+EXPORT_SYMBOL(flush_tlb_mm);
++#define MAXIMUM_OUTSTANDING_BLOCK_REQS \
++ (BLKIF_MAX_SEGMENTS_PER_REQUEST * BLK_RING_SIZE)
++#define GRANT_INVALID_REF 0
+
-+void flush_tlb_page(struct vm_area_struct * vma, unsigned long va)
++static void connect(struct blkfront_info *);
++static void blkfront_closing(struct xenbus_device *);
++static int blkfront_remove(struct xenbus_device *);
++static int talk_to_backend(struct xenbus_device *, struct blkfront_info *);
++static int setup_blkring(struct xenbus_device *, struct blkfront_info *);
++
++static void kick_pending_request_queues(struct blkfront_info *);
++
++static irqreturn_t blkif_int(int irq, void *dev_id, struct pt_regs *ptregs);
++static void blkif_restart_queue(void *arg);
++static void blkif_recover(struct blkfront_info *);
++static void blkif_completion(struct blk_shadow *);
++static void blkif_free(struct blkfront_info *, int);
++
++
++/**
++ * Entry point to this code when a new device is created. Allocate the basic
++ * structures and the ring buffer for communication with the backend, and
++ * inform the backend of the appropriate details for those. Switch to
++ * Initialised state.
++ */
++static int blkfront_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
+{
-+ struct mm_struct *mm = vma->vm_mm;
-+ cpumask_t cpu_mask;
++ int err, vdevice, i;
++ struct blkfront_info *info;
+
-+ preempt_disable();
-+ cpu_mask = mm->cpu_vm_mask;
-+ cpu_clear(smp_processor_id(), cpu_mask);
++ /* FIXME: Use dynamic device id if this is not set. */
++ err = xenbus_scanf(XBT_NIL, dev->nodename,
++ "virtual-device", "%i", &vdevice);
++ if (err != 1) {
++ /* go looking in the extended area instead */
++ err = xenbus_scanf(XBT_NIL, dev->nodename, "virtual-device-ext",
++ "%i", &vdevice);
++ if (err != 1) {
++ xenbus_dev_fatal(dev, err, "reading virtual-device");
++ return err;
++ }
++ }
+
-+ if (current->active_mm == mm) {
-+ if(current->mm)
-+ __flush_tlb_one(va);
-+ else
-+ leave_mm(smp_processor_id());
++ info = kzalloc(sizeof(*info), GFP_KERNEL);
++ if (!info) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
++ return -ENOMEM;
+ }
+
-+ if (!cpus_empty(cpu_mask))
-+ flush_tlb_others(cpu_mask, mm, va);
++ info->xbdev = dev;
++ info->vdevice = vdevice;
++ info->connected = BLKIF_STATE_DISCONNECTED;
++ INIT_WORK(&info->work, blkif_restart_queue, (void *)info);
++
++ for (i = 0; i < BLK_RING_SIZE; i++)
++ info->shadow[i].req.id = i+1;
++ info->shadow[BLK_RING_SIZE-1].req.id = 0x0fffffff;
+
-+ preempt_enable();
-+}
-+EXPORT_SYMBOL(flush_tlb_page);
++ /* Front end dir is a number, which is used as the id. */
++ info->handle = simple_strtoul(strrchr(dev->nodename,'/')+1, NULL, 0);
++ dev->dev.driver_data = info;
+
-+static void do_flush_tlb_all(void* info)
-+{
-+ unsigned long cpu = smp_processor_id();
++ err = talk_to_backend(dev, info);
++ if (err) {
++ kfree(info);
++ dev->dev.driver_data = NULL;
++ return err;
++ }
+
-+ __flush_tlb_all();
-+ if (read_pda(mmu_state) == TLBSTATE_LAZY)
-+ leave_mm(cpu);
++ return 0;
+}
+
-+void flush_tlb_all(void)
-+{
-+ on_each_cpu(do_flush_tlb_all, NULL, 1, 1);
-+}
-+#else
-+asmlinkage void smp_invalidate_interrupt (void)
-+{ return; }
-+void flush_tlb_current_task(void)
-+{ xen_tlb_flush_mask(¤t->mm->cpu_vm_mask); }
-+void flush_tlb_mm (struct mm_struct * mm)
-+{ xen_tlb_flush_mask(&mm->cpu_vm_mask); }
-+void flush_tlb_page(struct vm_area_struct * vma, unsigned long va)
-+{ xen_invlpg_mask(&vma->vm_mm->cpu_vm_mask, va); }
-+void flush_tlb_all(void)
-+{ xen_tlb_flush_all(); }
-+#endif /* Xen */
+
-+/*
-+ * this function sends a 'reschedule' IPI to another CPU.
-+ * it goes straight through and wastes no time serializing
-+ * anything. Worst case is that we lose a reschedule ...
++/**
++ * We are reconnecting to the backend, due to a suspend/resume, or a backend
++ * driver restart. We tear down our blkif structure and recreate it, but
++ * leave the device-layer structures intact so that this is transparent to the
++ * rest of the kernel.
+ */
-+
-+void smp_send_reschedule(int cpu)
++static int blkfront_resume(struct xenbus_device *dev)
+{
-+ send_IPI_mask(cpumask_of_cpu(cpu), RESCHEDULE_VECTOR);
-+}
++ struct blkfront_info *info = dev->dev.driver_data;
++ int err;
+
-+/*
-+ * Structure and data for smp_call_function(). This is designed to minimise
-+ * static memory requirements. It also looks cleaner.
-+ */
-+static DEFINE_SPINLOCK(call_lock);
++ DPRINTK("blkfront_resume: %s\n", dev->nodename);
+
-+struct call_data_struct {
-+ void (*func) (void *info);
-+ void *info;
-+ atomic_t started;
-+ atomic_t finished;
-+ int wait;
-+};
++ blkif_free(info, info->connected == BLKIF_STATE_CONNECTED);
+
-+static struct call_data_struct * call_data;
++ err = talk_to_backend(dev, info);
++ if (info->connected == BLKIF_STATE_SUSPENDED && !err)
++ blkif_recover(info);
+
-+void lock_ipi_call_lock(void)
-+{
-+ spin_lock_irq(&call_lock);
++ return err;
+}
+
-+void unlock_ipi_call_lock(void)
-+{
-+ spin_unlock_irq(&call_lock);
-+}
+
-+/*
-+ * this function sends a 'generic call function' IPI to one other CPU
-+ * in the system.
-+ *
-+ * cpu is a standard Linux logical CPU number.
-+ */
-+static void
-+__smp_call_function_single(int cpu, void (*func) (void *info), void *info,
-+ int nonatomic, int wait)
++/* Common code used when first setting up, and when resuming. */
++static int talk_to_backend(struct xenbus_device *dev,
++ struct blkfront_info *info)
+{
-+ struct call_data_struct data;
-+ int cpus = 1;
++ const char *message = NULL;
++ struct xenbus_transaction xbt;
++ int err;
+
-+ data.func = func;
-+ data.info = info;
-+ atomic_set(&data.started, 0);
-+ data.wait = wait;
-+ if (wait)
-+ atomic_set(&data.finished, 0);
++ /* Create shared ring, alloc event channel. */
++ err = setup_blkring(dev, info);
++ if (err)
++ goto out;
+
-+ call_data = &data;
-+ wmb();
-+ /* Send a message to all other CPUs and wait for them to respond */
-+ send_IPI_mask(cpumask_of_cpu(cpu), CALL_FUNCTION_VECTOR);
++again:
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "starting transaction");
++ goto destroy_blkring;
++ }
+
-+ /* Wait for response */
-+ while (atomic_read(&data.started) != cpus)
-+ cpu_relax();
++ err = xenbus_printf(xbt, dev->nodename,
++ "ring-ref","%u", info->ring_ref);
++ if (err) {
++ message = "writing ring-ref";
++ goto abort_transaction;
++ }
++ err = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
++ irq_to_evtchn_port(info->irq));
++ if (err) {
++ message = "writing event-channel";
++ goto abort_transaction;
++ }
++ err = xenbus_printf(xbt, dev->nodename, "protocol", "%s",
++ XEN_IO_PROTO_ABI_NATIVE);
++ if (err) {
++ message = "writing protocol";
++ goto abort_transaction;
++ }
+
-+ if (!wait)
-+ return;
++ err = xenbus_transaction_end(xbt, 0);
++ if (err) {
++ if (err == -EAGAIN)
++ goto again;
++ xenbus_dev_fatal(dev, err, "completing transaction");
++ goto destroy_blkring;
++ }
+
-+ while (atomic_read(&data.finished) != cpus)
-+ cpu_relax();
++ xenbus_switch_state(dev, XenbusStateInitialised);
++
++ return 0;
++
++ abort_transaction:
++ xenbus_transaction_end(xbt, 1);
++ if (message)
++ xenbus_dev_fatal(dev, err, "%s", message);
++ destroy_blkring:
++ blkif_free(info, 0);
++ out:
++ return err;
+}
+
-+/*
-+ * smp_call_function_single - Run a function on another CPU
-+ * @func: The function to run. This must be fast and non-blocking.
-+ * @info: An arbitrary pointer to pass to the function.
-+ * @nonatomic: Currently unused.
-+ * @wait: If true, wait until function has completed on other CPUs.
-+ *
-+ * Retrurns 0 on success, else a negative status code.
-+ *
-+ * Does not return until the remote CPU is nearly ready to execute <func>
-+ * or is or has executed.
-+ */
+
-+int smp_call_function_single (int cpu, void (*func) (void *info), void *info,
-+ int nonatomic, int wait)
++static int setup_blkring(struct xenbus_device *dev,
++ struct blkfront_info *info)
+{
-+ /* prevent preemption and reschedule on another processor */
-+ int me = get_cpu();
-+ if (cpu == me) {
-+ WARN_ON(1);
-+ put_cpu();
-+ return -EBUSY;
++ blkif_sring_t *sring;
++ int err;
++
++ info->ring_ref = GRANT_INVALID_REF;
++
++ sring = (blkif_sring_t *)__get_free_page(GFP_NOIO | __GFP_HIGH);
++ if (!sring) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating shared ring");
++ return -ENOMEM;
+ }
-+ spin_lock_bh(&call_lock);
-+ __smp_call_function_single(cpu, func, info, nonatomic, wait);
-+ spin_unlock_bh(&call_lock);
-+ put_cpu();
++ SHARED_RING_INIT(sring);
++ FRONT_RING_INIT(&info->ring, sring, PAGE_SIZE);
++
++ err = xenbus_grant_ring(dev, virt_to_mfn(info->ring.sring));
++ if (err < 0) {
++ free_page((unsigned long)sring);
++ info->ring.sring = NULL;
++ goto fail;
++ }
++ info->ring_ref = err;
++
++ err = bind_listening_port_to_irqhandler(
++ dev->otherend_id, blkif_int, SA_SAMPLE_RANDOM, "blkif", info);
++ if (err <= 0) {
++ xenbus_dev_fatal(dev, err,
++ "bind_listening_port_to_irqhandler");
++ goto fail;
++ }
++ info->irq = err;
++
+ return 0;
++fail:
++ blkif_free(info, 0);
++ return err;
+}
+
-+/*
-+ * this function sends a 'generic call function' IPI to all other CPUs
-+ * in the system.
++
++/**
++ * Callback received when the backend's state changes.
+ */
-+static void __smp_call_function (void (*func) (void *info), void *info,
-+ int nonatomic, int wait)
++static void backend_changed(struct xenbus_device *dev,
++ enum xenbus_state backend_state)
+{
-+ struct call_data_struct data;
-+ int cpus = num_online_cpus()-1;
++ struct blkfront_info *info = dev->dev.driver_data;
++ struct block_device *bd;
+
-+ if (!cpus)
-+ return;
++ DPRINTK("blkfront:backend_changed.\n");
+
-+ data.func = func;
-+ data.info = info;
-+ atomic_set(&data.started, 0);
-+ data.wait = wait;
-+ if (wait)
-+ atomic_set(&data.finished, 0);
++ switch (backend_state) {
++ case XenbusStateInitialising:
++ case XenbusStateInitWait:
++ case XenbusStateInitialised:
++ case XenbusStateReconfiguring:
++ case XenbusStateReconfigured:
++ case XenbusStateUnknown:
++ case XenbusStateClosed:
++ break;
+
-+ call_data = &data;
-+ wmb();
-+ /* Send a message to all other CPUs and wait for them to respond */
-+ send_IPI_allbutself(CALL_FUNCTION_VECTOR);
++ case XenbusStateConnected:
++ connect(info);
++ break;
+
-+ /* Wait for response */
-+ while (atomic_read(&data.started) != cpus)
-+#ifndef CONFIG_XEN
-+ cpu_relax();
++ case XenbusStateClosing:
++ bd = bdget(info->dev);
++ if (bd == NULL)
++ xenbus_dev_fatal(dev, -ENODEV, "bdget failed");
++
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
++ down(&bd->bd_sem);
+#else
-+ barrier();
++ mutex_lock(&bd->bd_mutex);
+#endif
-+
-+ if (!wait)
-+ return;
-+
-+ while (atomic_read(&data.finished) != cpus)
-+#ifndef CONFIG_XEN
-+ cpu_relax();
++ if (info->users > 0)
++ xenbus_dev_error(dev, -EBUSY,
++ "Device in use; refusing to close");
++ else
++ blkfront_closing(dev);
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
++ up(&bd->bd_sem);
+#else
-+ barrier();
++ mutex_unlock(&bd->bd_mutex);
+#endif
++ bdput(bd);
++ break;
++ }
+}
+
++
++/* ** Connection ** */
++
++
+/*
-+ * smp_call_function - run a function on all other CPUs.
-+ * @func: The function to run. This must be fast and non-blocking.
-+ * @info: An arbitrary pointer to pass to the function.
-+ * @nonatomic: currently unused.
-+ * @wait: If true, wait (atomically) until function has completed on other
-+ * CPUs.
-+ *
-+ * Returns 0 on success, else a negative status code. Does not return until
-+ * remote CPUs are nearly ready to execute func or are or have executed.
-+ *
-+ * You must not call this function with disabled interrupts or from a
-+ * hardware interrupt handler or from a bottom half handler.
-+ * Actually there are a few legal cases, like panic.
++ * Invoked when the backend is finally 'ready' (and has told produced
++ * the details about the physical device - #sectors, size, etc).
+ */
-+int smp_call_function (void (*func) (void *info), void *info, int nonatomic,
-+ int wait)
++static void connect(struct blkfront_info *info)
+{
-+ spin_lock(&call_lock);
-+ __smp_call_function(func,info,nonatomic,wait);
-+ spin_unlock(&call_lock);
-+ return 0;
-+}
-+EXPORT_SYMBOL(smp_call_function);
++ unsigned long long sectors;
++ unsigned long sector_size;
++ unsigned int binfo;
++ int err;
+
-+void smp_stop_cpu(void)
-+{
-+ unsigned long flags;
-+ /*
-+ * Remove this CPU:
-+ */
-+ cpu_clear(smp_processor_id(), cpu_online_map);
-+ local_irq_save(flags);
-+#ifndef CONFIG_XEN
-+ disable_local_APIC();
-+#endif
-+ local_irq_restore(flags);
-+}
++ if ((info->connected == BLKIF_STATE_CONNECTED) ||
++ (info->connected == BLKIF_STATE_SUSPENDED) )
++ return;
+
-+static void smp_really_stop_cpu(void *dummy)
-+{
-+ smp_stop_cpu();
-+ for (;;)
-+ halt();
-+}
++ DPRINTK("blkfront.c:connect:%s.\n", info->xbdev->otherend);
+
-+void smp_send_stop(void)
-+{
-+ int nolock = 0;
-+#ifndef CONFIG_XEN
-+ if (reboot_force)
++ err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
++ "sectors", "%Lu", §ors,
++ "info", "%u", &binfo,
++ "sector-size", "%lu", §or_size,
++ NULL);
++ if (err) {
++ xenbus_dev_fatal(info->xbdev, err,
++ "reading backend fields at %s",
++ info->xbdev->otherend);
+ return;
-+#endif
-+ /* Don't deadlock on the call lock in panic */
-+ if (!spin_trylock(&call_lock)) {
-+ /* ignore locking because we have panicked anyways */
-+ nolock = 1;
+ }
-+ __smp_call_function(smp_really_stop_cpu, NULL, 0, 0);
-+ if (!nolock)
-+ spin_unlock(&call_lock);
-+
-+ local_irq_disable();
-+#ifndef CONFIG_XEN
-+ disable_local_APIC();
-+#endif
-+ local_irq_enable();
-+}
-+
-+/*
-+ * Reschedule call back. Nothing to do,
-+ * all the work is done automatically when
-+ * we return from the interrupt.
-+ */
-+#ifndef CONFIG_XEN
-+asmlinkage void smp_reschedule_interrupt(void)
-+#else
-+asmlinkage irqreturn_t smp_reschedule_interrupt(void)
-+#endif
-+{
-+#ifndef CONFIG_XEN
-+ ack_APIC_irq();
-+#else
-+ return IRQ_HANDLED;
-+#endif
-+}
+
-+#ifndef CONFIG_XEN
-+asmlinkage void smp_call_function_interrupt(void)
-+#else
-+asmlinkage irqreturn_t smp_call_function_interrupt(void)
-+#endif
-+{
-+ void (*func) (void *info) = call_data->func;
-+ void *info = call_data->info;
-+ int wait = call_data->wait;
++ err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
++ "feature-barrier", "%lu", &info->feature_barrier,
++ NULL);
++ if (err)
++ info->feature_barrier = 0;
+
-+#ifndef CONFIG_XEN
-+ ack_APIC_irq();
-+#endif
-+ /*
-+ * Notify initiating CPU that I've grabbed the data and am
-+ * about to execute the function
-+ */
-+ mb();
-+ atomic_inc(&call_data->started);
-+ /*
-+ * At this point the info structure may be out of scope unless wait==1
-+ */
-+ exit_idle();
-+ irq_enter();
-+ (*func)(info);
-+ irq_exit();
-+ if (wait) {
-+ mb();
-+ atomic_inc(&call_data->finished);
++ err = xlvbd_add(sectors, info->vdevice, binfo, sector_size, info);
++ if (err) {
++ xenbus_dev_fatal(info->xbdev, err, "xlvbd_add at %s",
++ info->xbdev->otherend);
++ return;
+ }
-+#ifdef CONFIG_XEN
-+ return IRQ_HANDLED;
-+#endif
-+}
-+
-+int safe_smp_processor_id(void)
-+{
-+#ifdef CONFIG_XEN
-+ return smp_processor_id();
-+#else
-+ unsigned apicid, i;
+
-+ if (disable_apic)
-+ return 0;
++ err = xlvbd_sysfs_addif(info);
++ if (err) {
++ xenbus_dev_fatal(info->xbdev, err, "xlvbd_sysfs_addif at %s",
++ info->xbdev->otherend);
++ return;
++ }
+
-+ apicid = hard_smp_processor_id();
-+ if (apicid < NR_CPUS && x86_cpu_to_apicid[apicid] == apicid)
-+ return apicid;
++ (void)xenbus_switch_state(info->xbdev, XenbusStateConnected);
+
-+ for (i = 0; i < NR_CPUS; ++i) {
-+ if (x86_cpu_to_apicid[i] == apicid)
-+ return i;
-+ }
++ /* Kick pending requests. */
++ spin_lock_irq(&blkif_io_lock);
++ info->connected = BLKIF_STATE_CONNECTED;
++ kick_pending_request_queues(info);
++ spin_unlock_irq(&blkif_io_lock);
+
-+ /* No entries in x86_cpu_to_apicid? Either no MPS|ACPI,
-+ * or called too early. Either way, we must be CPU 0. */
-+ if (x86_cpu_to_apicid[0] == BAD_APICID)
-+ return 0;
++ add_disk(info->gd);
+
-+ return 0; /* Should not happen */
-+#endif
++ info->is_ready = 1;
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/traps-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/traps-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/traps-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/traps-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,1175 @@
-+/*
-+ * linux/arch/x86-64/traps.c
-+ *
-+ * Copyright (C) 1991, 1992 Linus Torvalds
-+ * Copyright (C) 2000, 2001, 2002 Andi Kleen, SuSE Labs
-+ *
-+ * Pentium III FXSR, SSE support
-+ * Gareth Hughes <gareth at valinux.com>, May 2000
-+ */
+
-+/*
-+ * 'Traps.c' handles hardware traps and faults after we have saved some
-+ * state in 'entry.S'.
++/**
++ * Handle the change of state of the backend to Closing. We must delete our
++ * device-layer structures now, to ensure that writes are flushed through to
++ * the backend. Once is this done, we can switch to Closed in
++ * acknowledgement.
+ */
-+#include <linux/sched.h>
-+#include <linux/kernel.h>
-+#include <linux/string.h>
-+#include <linux/errno.h>
-+#include <linux/ptrace.h>
-+#include <linux/timer.h>
-+#include <linux/mm.h>
-+#include <linux/init.h>
-+#include <linux/delay.h>
-+#include <linux/spinlock.h>
-+#include <linux/interrupt.h>
-+#include <linux/module.h>
-+#include <linux/moduleparam.h>
-+#include <linux/nmi.h>
-+#include <linux/kprobes.h>
-+#include <linux/kexec.h>
-+#include <linux/unwind.h>
++static void blkfront_closing(struct xenbus_device *dev)
++{
++ struct blkfront_info *info = dev->dev.driver_data;
++ unsigned long flags;
+
-+#include <asm/system.h>
-+#include <asm/uaccess.h>
-+#include <asm/io.h>
-+#include <asm/atomic.h>
-+#include <asm/debugreg.h>
-+#include <asm/desc.h>
-+#include <asm/i387.h>
-+#include <asm/kdebug.h>
-+#include <asm/processor.h>
-+#include <asm/unwind.h>
-+#include <asm/smp.h>
-+#include <asm/pgalloc.h>
-+#include <asm/pda.h>
-+#include <asm/proto.h>
-+#include <asm/nmi.h>
++ DPRINTK("blkfront_closing: %s removed\n", dev->nodename);
++
++ if (info->rq == NULL)
++ goto out;
+
-+asmlinkage void divide_error(void);
-+asmlinkage void debug(void);
-+asmlinkage void nmi(void);
-+asmlinkage void int3(void);
-+asmlinkage void overflow(void);
-+asmlinkage void bounds(void);
-+asmlinkage void invalid_op(void);
-+asmlinkage void device_not_available(void);
-+asmlinkage void double_fault(void);
-+asmlinkage void coprocessor_segment_overrun(void);
-+asmlinkage void invalid_TSS(void);
-+asmlinkage void segment_not_present(void);
-+asmlinkage void stack_segment(void);
-+asmlinkage void general_protection(void);
-+asmlinkage void page_fault(void);
-+asmlinkage void coprocessor_error(void);
-+asmlinkage void simd_coprocessor_error(void);
-+asmlinkage void reserved(void);
-+asmlinkage void alignment_check(void);
-+asmlinkage void machine_check(void);
-+asmlinkage void spurious_interrupt_bug(void);
++ spin_lock_irqsave(&blkif_io_lock, flags);
++ /* No more blkif_request(). */
++ blk_stop_queue(info->rq);
++ /* No more gnttab callback work. */
++ gnttab_cancel_free_callback(&info->callback);
++ spin_unlock_irqrestore(&blkif_io_lock, flags);
+
-+ATOMIC_NOTIFIER_HEAD(die_chain);
-+EXPORT_SYMBOL(die_chain);
++ /* Flush gnttab callback work. Must be done with no locks held. */
++ flush_scheduled_work();
+
-+int register_die_notifier(struct notifier_block *nb)
++ xlvbd_sysfs_delif(info);
++
++ xlvbd_del(info);
++
++ out:
++ xenbus_frontend_closed(dev);
++}
++
++
++static int blkfront_remove(struct xenbus_device *dev)
+{
-+ vmalloc_sync_all();
-+ return atomic_notifier_chain_register(&die_chain, nb);
++ struct blkfront_info *info = dev->dev.driver_data;
++
++ DPRINTK("blkfront_remove: %s removed\n", dev->nodename);
++
++ blkif_free(info, 0);
++
++ kfree(info);
++
++ return 0;
+}
-+EXPORT_SYMBOL(register_die_notifier); /* used modular by kdb */
+
-+int unregister_die_notifier(struct notifier_block *nb)
++
++static inline int GET_ID_FROM_FREELIST(
++ struct blkfront_info *info)
+{
-+ return atomic_notifier_chain_unregister(&die_chain, nb);
++ unsigned long free = info->shadow_free;
++ BUG_ON(free > BLK_RING_SIZE);
++ info->shadow_free = info->shadow[free].req.id;
++ info->shadow[free].req.id = 0x0fffffee; /* debug */
++ return free;
+}
-+EXPORT_SYMBOL(unregister_die_notifier); /* used modular by kdb */
+
-+static inline void conditional_sti(struct pt_regs *regs)
++static inline void ADD_ID_TO_FREELIST(
++ struct blkfront_info *info, unsigned long id)
+{
-+ if (regs->eflags & X86_EFLAGS_IF)
-+ local_irq_enable();
++ info->shadow[id].req.id = info->shadow_free;
++ info->shadow[id].request = 0;
++ info->shadow_free = id;
+}
+
-+static inline void preempt_conditional_sti(struct pt_regs *regs)
++static inline void flush_requests(struct blkfront_info *info)
+{
-+ preempt_disable();
-+ if (regs->eflags & X86_EFLAGS_IF)
-+ local_irq_enable();
++ int notify;
++
++ RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&info->ring, notify);
++
++ if (notify)
++ notify_remote_via_irq(info->irq);
+}
+
-+static inline void preempt_conditional_cli(struct pt_regs *regs)
++static void kick_pending_request_queues(struct blkfront_info *info)
+{
-+ if (regs->eflags & X86_EFLAGS_IF)
-+ local_irq_disable();
-+ /* Make sure to not schedule here because we could be running
-+ on an exception stack. */
-+ preempt_enable_no_resched();
++ if (!RING_FULL(&info->ring)) {
++ /* Re-enable calldowns. */
++ blk_start_queue(info->rq);
++ /* Kick things off immediately. */
++ do_blkif_request(info->rq);
++ }
+}
+
-+static int kstack_depth_to_print = 12;
-+#ifdef CONFIG_STACK_UNWIND
-+static int call_trace = 1;
-+#else
-+#define call_trace (-1)
-+#endif
++static void blkif_restart_queue(void *arg)
++{
++ struct blkfront_info *info = (struct blkfront_info *)arg;
++ spin_lock_irq(&blkif_io_lock);
++ if (info->connected == BLKIF_STATE_CONNECTED)
++ kick_pending_request_queues(info);
++ spin_unlock_irq(&blkif_io_lock);
++}
+
-+#ifdef CONFIG_KALLSYMS
-+# include <linux/kallsyms.h>
-+void printk_address(unsigned long address)
++static void blkif_restart_queue_callback(void *arg)
+{
-+ unsigned long offset = 0, symsize;
-+ const char *symname;
-+ char *modname;
-+ char *delim = ":";
-+ char namebuf[128];
++ struct blkfront_info *info = (struct blkfront_info *)arg;
++ schedule_work(&info->work);
++}
+
-+ symname = kallsyms_lookup(address, &symsize, &offset,
-+ &modname, namebuf);
-+ if (!symname) {
-+ printk(" [<%016lx>]\n", address);
-+ return;
-+ }
-+ if (!modname)
-+ modname = delim = "";
-+ printk(" [<%016lx>] %s%s%s%s+0x%lx/0x%lx\n",
-+ address, delim, modname, delim, symname, offset, symsize);
++int blkif_open(struct inode *inode, struct file *filep)
++{
++ struct blkfront_info *info = inode->i_bdev->bd_disk->private_data;
++ info->users++;
++ return 0;
+}
-+#else
-+void printk_address(unsigned long address)
++
++
++int blkif_release(struct inode *inode, struct file *filep)
+{
-+ printk(" [<%016lx>]\n", address);
++ struct blkfront_info *info = inode->i_bdev->bd_disk->private_data;
++ info->users--;
++ if (info->users == 0) {
++ /* Check whether we have been instructed to close. We will
++ have ignored this request initially, as the device was
++ still mounted. */
++ struct xenbus_device * dev = info->xbdev;
++ enum xenbus_state state = xenbus_read_driver_state(dev->otherend);
++
++ if (state == XenbusStateClosing && info->is_ready)
++ blkfront_closing(dev);
++ }
++ return 0;
+}
-+#endif
+
-+static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
-+ unsigned *usedp, const char **idp)
++
++int blkif_ioctl(struct inode *inode, struct file *filep,
++ unsigned command, unsigned long argument)
+{
-+#ifndef CONFIG_X86_NO_TSS
-+ static char ids[][8] = {
-+ [DEBUG_STACK - 1] = "#DB",
-+ [NMI_STACK - 1] = "NMI",
-+ [DOUBLEFAULT_STACK - 1] = "#DF",
-+ [STACKFAULT_STACK - 1] = "#SS",
-+ [MCE_STACK - 1] = "#MC",
-+#if DEBUG_STKSZ > EXCEPTION_STKSZ
-+ [N_EXCEPTION_STACKS ... N_EXCEPTION_STACKS + DEBUG_STKSZ / EXCEPTION_STKSZ - 2] = "#DB[?]"
-+#endif
-+ };
-+ unsigned k;
++ int i;
+
-+ /*
-+ * Iterate over all exception stacks, and figure out whether
-+ * 'stack' is in one of them:
-+ */
-+ for (k = 0; k < N_EXCEPTION_STACKS; k++) {
-+ unsigned long end;
++ DPRINTK_IOCTL("command: 0x%x, argument: 0x%lx, dev: 0x%04x\n",
++ command, (long)argument, inode->i_rdev);
+
-+ /*
-+ * set 'end' to the end of the exception stack.
-+ */
-+ switch (k + 1) {
-+ /*
-+ * TODO: this block is not needed i think, because
-+ * setup64.c:cpu_init() sets up t->ist[DEBUG_STACK]
-+ * properly too.
-+ */
-+#if DEBUG_STKSZ > EXCEPTION_STKSZ
-+ case DEBUG_STACK:
-+ end = cpu_pda(cpu)->debugstack + DEBUG_STKSZ;
-+ break;
-+#endif
-+ default:
-+ end = per_cpu(orig_ist, cpu).ist[k];
-+ break;
-+ }
-+ /*
-+ * Is 'stack' above this exception frame's end?
-+ * If yes then skip to the next frame.
-+ */
-+ if (stack >= end)
-+ continue;
-+ /*
-+ * Is 'stack' above this exception frame's start address?
-+ * If yes then we found the right frame.
-+ */
-+ if (stack >= end - EXCEPTION_STKSZ) {
-+ /*
-+ * Make sure we only iterate through an exception
-+ * stack once. If it comes up for the second time
-+ * then there's something wrong going on - just
-+ * break out and return NULL:
-+ */
-+ if (*usedp & (1U << k))
-+ break;
-+ *usedp |= 1U << k;
-+ *idp = ids[k];
-+ return (unsigned long *)end;
-+ }
-+ /*
-+ * If this is a debug stack, and if it has a larger size than
-+ * the usual exception stacks, then 'stack' might still
-+ * be within the lower portion of the debug stack:
-+ */
-+#if DEBUG_STKSZ > EXCEPTION_STKSZ
-+ if (k == DEBUG_STACK - 1 && stack >= end - DEBUG_STKSZ) {
-+ unsigned j = N_EXCEPTION_STACKS - 1;
++ switch (command) {
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,16)
++ case HDIO_GETGEO: {
++ struct block_device *bd = inode->i_bdev;
++ struct hd_geometry geo;
++ int ret;
+
-+ /*
-+ * Black magic. A large debug stack is composed of
-+ * multiple exception stack entries, which we
-+ * iterate through now. Dont look:
-+ */
-+ do {
-+ ++j;
-+ end -= EXCEPTION_STKSZ;
-+ ids[j][4] = '1' + (j - N_EXCEPTION_STACKS);
-+ } while (stack < end - EXCEPTION_STKSZ);
-+ if (*usedp & (1U << j))
-+ break;
-+ *usedp |= 1U << j;
-+ *idp = ids[j];
-+ return (unsigned long *)end;
-+ }
-+#endif
++ if (!argument)
++ return -EINVAL;
++
++ geo.start = get_start_sect(bd);
++ ret = blkif_getgeo(bd, &geo);
++ if (ret)
++ return ret;
++
++ if (copy_to_user((struct hd_geometry __user *)argument, &geo,
++ sizeof(geo)))
++ return -EFAULT;
++
++ return 0;
+ }
+#endif
-+ return NULL;
++ case CDROMMULTISESSION:
++ DPRINTK("FIXME: support multisession CDs later\n");
++ for (i = 0; i < sizeof(struct cdrom_multisession); i++)
++ if (put_user(0, (char __user *)(argument + i)))
++ return -EFAULT;
++ return 0;
++
++ case CDROM_GET_CAPABILITY: {
++ struct blkfront_info *info =
++ inode->i_bdev->bd_disk->private_data;
++ struct gendisk *gd = info->gd;
++ if (gd->flags & GENHD_FL_CD)
++ return 0;
++ return -EINVAL;
++ }
++ default:
++ /*printk(KERN_ALERT "ioctl %08x not supported by Xen blkdev\n",
++ command);*/
++ return -EINVAL; /* same return as native Linux */
++ }
++
++ return 0;
+}
+
-+static int show_trace_unwind(struct unwind_frame_info *info, void *context)
++
++int blkif_getgeo(struct block_device *bd, struct hd_geometry *hg)
+{
-+ int n = 0;
++ /* We don't have real geometry info, but let's at least return
++ values consistent with the size of the device */
++ sector_t nsect = get_capacity(bd->bd_disk);
++ sector_t cylinders = nsect;
+
-+ while (unwind(info) == 0 && UNW_PC(info)) {
-+ n++;
-+ printk_address(UNW_PC(info));
-+ if (arch_unw_user_mode(info))
-+ break;
-+ }
-+ return n;
++ hg->heads = 0xff;
++ hg->sectors = 0x3f;
++ sector_div(cylinders, hg->heads * hg->sectors);
++ hg->cylinders = cylinders;
++ if ((sector_t)(hg->cylinders + 1) * hg->heads * hg->sectors < nsect)
++ hg->cylinders = 0xffff;
++ return 0;
+}
+
++
+/*
-+ * x86-64 can have upto three kernel stacks:
-+ * process stack
-+ * interrupt stack
-+ * severe exception (double fault, nmi, stack fault, debug, mce) hardware stack
++ * blkif_queue_request
++ *
++ * request block io
++ *
++ * id: for guest use only.
++ * operation: BLKIF_OP_{READ,WRITE,PROBE}
++ * buffer: buffer to read/write into. this should be a
++ * virtual address in the guest os.
+ */
-+
-+void show_trace(struct task_struct *tsk, struct pt_regs *regs, unsigned long * stack)
++static int blkif_queue_request(struct request *req)
+{
-+ const unsigned cpu = safe_smp_processor_id();
-+ unsigned long *irqstack_end = (unsigned long *)cpu_pda(cpu)->irqstackptr;
-+ unsigned used = 0;
++ struct blkfront_info *info = req->rq_disk->private_data;
++ unsigned long buffer_mfn;
++ blkif_request_t *ring_req;
++ struct bio *bio;
++ struct bio_vec *bvec;
++ int idx;
++ unsigned long id;
++ unsigned int fsect, lsect;
++ int ref;
++ grant_ref_t gref_head;
+
-+ printk("\nCall Trace:\n");
++ if (unlikely(info->connected != BLKIF_STATE_CONNECTED))
++ return 1;
+
-+ if (!tsk)
-+ tsk = current;
++ if (gnttab_alloc_grant_references(
++ BLKIF_MAX_SEGMENTS_PER_REQUEST, &gref_head) < 0) {
++ gnttab_request_free_callback(
++ &info->callback,
++ blkif_restart_queue_callback,
++ info,
++ BLKIF_MAX_SEGMENTS_PER_REQUEST);
++ return 1;
++ }
+
-+ if (call_trace >= 0) {
-+ int unw_ret = 0;
-+ struct unwind_frame_info info;
++ /* Fill out a communications ring structure. */
++ ring_req = RING_GET_REQUEST(&info->ring, info->ring.req_prod_pvt);
++ id = GET_ID_FROM_FREELIST(info);
++ info->shadow[id].request = (unsigned long)req;
+
-+ if (regs) {
-+ if (unwind_init_frame_info(&info, tsk, regs) == 0)
-+ unw_ret = show_trace_unwind(&info, NULL);
-+ } else if (tsk == current)
-+ unw_ret = unwind_init_running(&info, show_trace_unwind, NULL);
-+ else {
-+ if (unwind_init_blocked(&info, tsk) == 0)
-+ unw_ret = show_trace_unwind(&info, NULL);
++ ring_req->id = id;
++ ring_req->sector_number = (blkif_sector_t)req->sector;
++ ring_req->handle = info->handle;
++
++ ring_req->operation = rq_data_dir(req) ?
++ BLKIF_OP_WRITE : BLKIF_OP_READ;
++ if (blk_barrier_rq(req))
++ ring_req->operation = BLKIF_OP_WRITE_BARRIER;
++
++ ring_req->nr_segments = 0;
++ rq_for_each_bio (bio, req) {
++ bio_for_each_segment (bvec, bio, idx) {
++ BUG_ON(ring_req->nr_segments
++ == BLKIF_MAX_SEGMENTS_PER_REQUEST);
++ buffer_mfn = page_to_phys(bvec->bv_page) >> PAGE_SHIFT;
++ fsect = bvec->bv_offset >> 9;
++ lsect = fsect + (bvec->bv_len >> 9) - 1;
++ /* install a grant reference. */
++ ref = gnttab_claim_grant_reference(&gref_head);
++ BUG_ON(ref == -ENOSPC);
++
++ gnttab_grant_foreign_access_ref(
++ ref,
++ info->xbdev->otherend_id,
++ buffer_mfn,
++ rq_data_dir(req) ? GTF_readonly : 0 );
++
++ info->shadow[id].frame[ring_req->nr_segments] =
++ mfn_to_pfn(buffer_mfn);
++
++ ring_req->seg[ring_req->nr_segments] =
++ (struct blkif_request_segment) {
++ .gref = ref,
++ .first_sect = fsect,
++ .last_sect = lsect };
++
++ ring_req->nr_segments++;
+ }
-+ if (unw_ret > 0) {
-+ if (call_trace == 1 && !arch_unw_user_mode(&info)) {
-+ print_symbol("DWARF2 unwinder stuck at %s\n",
-+ UNW_PC(&info));
-+ if ((long)UNW_SP(&info) < 0) {
-+ printk("Leftover inexact backtrace:\n");
-+ stack = (unsigned long *)UNW_SP(&info);
-+ } else
-+ printk("Full inexact backtrace again:\n");
-+ } else if (call_trace >= 1)
-+ return;
-+ else
-+ printk("Full inexact backtrace again:\n");
-+ } else
-+ printk("Inexact backtrace:\n");
+ }
+
-+ /*
-+ * Print function call entries within a stack. 'cond' is the
-+ * "end of stackframe" condition, that the 'stack++'
-+ * iteration will eventually trigger.
-+ */
-+#define HANDLE_STACK(cond) \
-+ do while (cond) { \
-+ unsigned long addr = *stack++; \
-+ if (kernel_text_address(addr)) { \
-+ /* \
-+ * If the address is either in the text segment of the \
-+ * kernel, or in the region which contains vmalloc'ed \
-+ * memory, it *may* be the address of a calling \
-+ * routine; if so, print it so that someone tracing \
-+ * down the cause of the crash will be able to figure \
-+ * out the call path that was taken. \
-+ */ \
-+ printk_address(addr); \
-+ } \
-+ } while (0)
++ info->ring.req_prod_pvt++;
+
-+ /*
-+ * Print function call entries in all stacks, starting at the
-+ * current stack address. If the stacks consist of nested
-+ * exceptions
-+ */
-+ for ( ; ; ) {
-+ const char *id;
-+ unsigned long *estack_end;
-+ estack_end = in_exception_stack(cpu, (unsigned long)stack,
-+ &used, &id);
++ /* Keep a private copy so we can reissue requests when recovering. */
++ info->shadow[id].req = *ring_req;
+
-+ if (estack_end) {
-+ printk(" <%s>", id);
-+ HANDLE_STACK (stack < estack_end);
-+ printk(" <EOE>");
-+ /*
-+ * We link to the next stack via the
-+ * second-to-last pointer (index -2 to end) in the
-+ * exception stack:
-+ */
-+ stack = (unsigned long *) estack_end[-2];
++ gnttab_free_grant_references(gref_head);
++
++ return 0;
++}
++
++/*
++ * do_blkif_request
++ * read a block; request is in a request queue
++ */
++void do_blkif_request(request_queue_t *rq)
++{
++ struct blkfront_info *info = NULL;
++ struct request *req;
++ int queued;
++
++ DPRINTK("Entered do_blkif_request\n");
++
++ queued = 0;
++
++ while ((req = elv_next_request(rq)) != NULL) {
++ info = req->rq_disk->private_data;
++ if (!blk_fs_request(req)) {
++ end_request(req, 0);
+ continue;
+ }
-+ if (irqstack_end) {
-+ unsigned long *irqstack;
-+ irqstack = irqstack_end -
-+ (IRQSTACKSIZE - 64) / sizeof(*irqstack);
+
-+ if (stack >= irqstack && stack < irqstack_end) {
-+ printk(" <IRQ>");
-+ HANDLE_STACK (stack < irqstack_end);
-+ /*
-+ * We link to the next stack (which would be
-+ * the process stack normally) the last
-+ * pointer (index -1 to end) in the IRQ stack:
-+ */
-+ stack = (unsigned long *) (irqstack_end[-1]);
-+ irqstack_end = NULL;
-+ printk(" <EOI>");
-+ continue;
-+ }
++ if (RING_FULL(&info->ring))
++ goto wait;
++
++ DPRINTK("do_blk_req %p: cmd %p, sec %llx, "
++ "(%u/%li) buffer:%p [%s]\n",
++ req, req->cmd, (long long)req->sector,
++ req->current_nr_sectors,
++ req->nr_sectors, req->buffer,
++ rq_data_dir(req) ? "write" : "read");
++
++
++ blkdev_dequeue_request(req);
++ if (blkif_queue_request(req)) {
++ blk_requeue_request(rq, req);
++ wait:
++ /* Avoid pointless unplugs. */
++ blk_stop_queue(rq);
++ break;
+ }
-+ break;
-+ }
+
-+ /*
-+ * This prints the process stack:
-+ */
-+ HANDLE_STACK (((long) stack & (THREAD_SIZE-1)) != 0);
-+#undef HANDLE_STACK
++ queued++;
++ }
+
-+ printk("\n");
++ if (queued != 0)
++ flush_requests(info);
+}
+
-+static void _show_stack(struct task_struct *tsk, struct pt_regs *regs, unsigned long * rsp)
++
++static irqreturn_t blkif_int(int irq, void *dev_id, struct pt_regs *ptregs)
+{
-+ unsigned long *stack;
-+ int i;
-+ const int cpu = safe_smp_processor_id();
-+ unsigned long *irqstack_end = (unsigned long *) (cpu_pda(cpu)->irqstackptr);
-+ unsigned long *irqstack = (unsigned long *) (cpu_pda(cpu)->irqstackptr - IRQSTACKSIZE);
++ struct request *req;
++ blkif_response_t *bret;
++ RING_IDX i, rp;
++ unsigned long flags;
++ struct blkfront_info *info = (struct blkfront_info *)dev_id;
++ int uptodate;
+
-+ // debugging aid: "show_stack(NULL, NULL);" prints the
-+ // back trace for this cpu.
++ spin_lock_irqsave(&blkif_io_lock, flags);
+
-+ if (rsp == NULL) {
-+ if (tsk)
-+ rsp = (unsigned long *)tsk->thread.rsp;
-+ else
-+ rsp = (unsigned long *)&rsp;
++ if (unlikely(info->connected != BLKIF_STATE_CONNECTED)) {
++ spin_unlock_irqrestore(&blkif_io_lock, flags);
++ return IRQ_HANDLED;
+ }
+
-+ stack = rsp;
-+ for(i=0; i < kstack_depth_to_print; i++) {
-+ if (stack >= irqstack && stack <= irqstack_end) {
-+ if (stack == irqstack_end) {
-+ stack = (unsigned long *) (irqstack_end[-1]);
-+ printk(" <EOI> ");
++ again:
++ rp = info->ring.sring->rsp_prod;
++ rmb(); /* Ensure we see queued responses up to 'rp'. */
++
++ for (i = info->ring.rsp_cons; i != rp; i++) {
++ unsigned long id;
++ int ret;
++
++ bret = RING_GET_RESPONSE(&info->ring, i);
++ id = bret->id;
++ req = (struct request *)info->shadow[id].request;
++
++ blkif_completion(&info->shadow[id]);
++
++ ADD_ID_TO_FREELIST(info, id);
++
++ uptodate = (bret->status == BLKIF_RSP_OKAY);
++ switch (bret->operation) {
++ case BLKIF_OP_WRITE_BARRIER:
++ if (unlikely(bret->status == BLKIF_RSP_EOPNOTSUPP)) {
++ printk("blkfront: %s: write barrier op failed\n",
++ info->gd->disk_name);
++ uptodate = -EOPNOTSUPP;
++ info->feature_barrier = 0;
++ xlvbd_barrier(info);
+ }
-+ } else {
-+ if (((long) stack & (THREAD_SIZE-1)) == 0)
++ /* fall through */
++ case BLKIF_OP_READ:
++ case BLKIF_OP_WRITE:
++ if (unlikely(bret->status != BLKIF_RSP_OKAY))
++ DPRINTK("Bad return from blkdev data "
++ "request: %x\n", bret->status);
++
++ ret = end_that_request_first(req, uptodate,
++ req->hard_nr_sectors);
++ BUG_ON(ret);
++ end_that_request_last(req, uptodate);
+ break;
++ default:
++ BUG();
+ }
-+ if (i && ((i % 4) == 0))
-+ printk("\n");
-+ printk(" %016lx", *stack++);
-+ touch_nmi_watchdog();
+ }
-+ show_trace(tsk, regs, rsp);
++
++ info->ring.rsp_cons = i;
++
++ if (i != info->ring.req_prod_pvt) {
++ int more_to_do;
++ RING_FINAL_CHECK_FOR_RESPONSES(&info->ring, more_to_do);
++ if (more_to_do)
++ goto again;
++ } else
++ info->ring.sring->rsp_event = i + 1;
++
++ kick_pending_request_queues(info);
++
++ spin_unlock_irqrestore(&blkif_io_lock, flags);
++
++ return IRQ_HANDLED;
+}
+
-+void show_stack(struct task_struct *tsk, unsigned long * rsp)
++static void blkif_free(struct blkfront_info *info, int suspend)
+{
-+ _show_stack(tsk, NULL, rsp);
++ /* Prevent new requests being issued until we fix things up. */
++ spin_lock_irq(&blkif_io_lock);
++ info->connected = suspend ?
++ BLKIF_STATE_SUSPENDED : BLKIF_STATE_DISCONNECTED;
++ /* No more blkif_request(). */
++ if (info->rq)
++ blk_stop_queue(info->rq);
++ /* No more gnttab callback work. */
++ gnttab_cancel_free_callback(&info->callback);
++ spin_unlock_irq(&blkif_io_lock);
++
++ /* Flush gnttab callback work. Must be done with no locks held. */
++ flush_scheduled_work();
++
++ /* Free resources associated with old device channel. */
++ if (info->ring_ref != GRANT_INVALID_REF) {
++ gnttab_end_foreign_access(info->ring_ref,
++ (unsigned long)info->ring.sring);
++ info->ring_ref = GRANT_INVALID_REF;
++ info->ring.sring = NULL;
++ }
++ if (info->irq)
++ unbind_from_irqhandler(info->irq, info);
++ info->irq = 0;
+}
+
-+/*
-+ * The architecture-independent dump_stack generator
-+ */
-+void dump_stack(void)
++static void blkif_completion(struct blk_shadow *s)
+{
-+ unsigned long dummy;
-+ show_trace(NULL, NULL, &dummy);
++ int i;
++ for (i = 0; i < s->req.nr_segments; i++)
++ gnttab_end_foreign_access(s->req.seg[i].gref, 0UL);
+}
+
-+EXPORT_SYMBOL(dump_stack);
-+
-+void show_registers(struct pt_regs *regs)
++static void blkif_recover(struct blkfront_info *info)
+{
+ int i;
-+ int in_kernel = !user_mode(regs);
-+ unsigned long rsp;
-+ const int cpu = safe_smp_processor_id();
-+ struct task_struct *cur = cpu_pda(cpu)->pcurrent;
-+
-+ rsp = regs->rsp;
++ blkif_request_t *req;
++ struct blk_shadow *copy;
++ int j;
+
-+ printk("CPU %d ", cpu);
-+ __show_regs(regs);
-+ printk("Process %s (pid: %d, threadinfo %p, task %p)\n",
-+ cur->comm, cur->pid, task_thread_info(cur), cur);
++ /* Stage 1: Make a safe copy of the shadow state. */
++ copy = kmalloc(sizeof(info->shadow), GFP_NOIO | __GFP_NOFAIL | __GFP_HIGH);
++ memcpy(copy, info->shadow, sizeof(info->shadow));
+
-+ /*
-+ * When in-kernel, we also print out the stack and code at the
-+ * time of the fault..
-+ */
-+ if (in_kernel) {
++ /* Stage 2: Set up free list. */
++ memset(&info->shadow, 0, sizeof(info->shadow));
++ for (i = 0; i < BLK_RING_SIZE; i++)
++ info->shadow[i].req.id = i+1;
++ info->shadow_free = info->ring.req_prod_pvt;
++ info->shadow[BLK_RING_SIZE-1].req.id = 0x0fffffff;
+
-+ printk("Stack: ");
-+ _show_stack(NULL, regs, (unsigned long*)rsp);
++ /* Stage 3: Find pending requests and requeue them. */
++ for (i = 0; i < BLK_RING_SIZE; i++) {
++ /* Not in use? */
++ if (copy[i].request == 0)
++ continue;
+
-+ printk("\nCode: ");
-+ if (regs->rip < PAGE_OFFSET)
-+ goto bad;
++ /* Grab a request slot and copy shadow state into it. */
++ req = RING_GET_REQUEST(
++ &info->ring, info->ring.req_prod_pvt);
++ *req = copy[i].req;
+
-+ for (i=0; i<20; i++) {
-+ unsigned char c;
-+ if (__get_user(c, &((unsigned char*)regs->rip)[i])) {
-+bad:
-+ printk(" Bad RIP value.");
-+ break;
-+ }
-+ printk("%02x ", c);
-+ }
-+ }
-+ printk("\n");
-+}
++ /* We get a new request id, and must reset the shadow state. */
++ req->id = GET_ID_FROM_FREELIST(info);
++ memcpy(&info->shadow[req->id], ©[i], sizeof(copy[i]));
+
-+void handle_BUG(struct pt_regs *regs)
-+{
-+ struct bug_frame f;
-+ long len;
-+ const char *prefix = "";
++ /* Rewrite any grant references invalidated by susp/resume. */
++ for (j = 0; j < req->nr_segments; j++)
++ gnttab_grant_foreign_access_ref(
++ req->seg[j].gref,
++ info->xbdev->otherend_id,
++ pfn_to_mfn(info->shadow[req->id].frame[j]),
++ rq_data_dir((struct request *)
++ info->shadow[req->id].request) ?
++ GTF_readonly : 0);
++ info->shadow[req->id].req = *req;
+
-+ if (user_mode(regs))
-+ return;
-+ if (__copy_from_user(&f, (const void __user *) regs->rip,
-+ sizeof(struct bug_frame)))
-+ return;
-+ if (f.filename >= 0 ||
-+ f.ud2[0] != 0x0f || f.ud2[1] != 0x0b)
-+ return;
-+ len = __strnlen_user((char *)(long)f.filename, PATH_MAX) - 1;
-+ if (len < 0 || len >= PATH_MAX)
-+ f.filename = (int)(long)"unmapped filename";
-+ else if (len > 50) {
-+ f.filename += len - 50;
-+ prefix = "...";
++ info->ring.req_prod_pvt++;
+ }
-+ printk("----------- [cut here ] --------- [please bite here ] ---------\n");
-+ printk(KERN_ALERT "Kernel BUG at %s%.50s:%d\n", prefix, (char *)(long)f.filename, f.line);
-+}
-+
-+#ifdef CONFIG_BUG
-+void out_of_line_bug(void)
-+{
-+ BUG();
-+}
-+EXPORT_SYMBOL(out_of_line_bug);
-+#endif
+
-+static DEFINE_SPINLOCK(die_lock);
-+static int die_owner = -1;
-+static unsigned int die_nest_count;
++ kfree(copy);
+
-+unsigned __kprobes long oops_begin(void)
-+{
-+ int cpu = safe_smp_processor_id();
-+ unsigned long flags;
++ (void)xenbus_switch_state(info->xbdev, XenbusStateConnected);
+
-+ /* racy, but better than risking deadlock. */
-+ local_irq_save(flags);
-+ if (!spin_trylock(&die_lock)) {
-+ if (cpu == die_owner)
-+ /* nested oops. should stop eventually */;
-+ else
-+ spin_lock(&die_lock);
-+ }
-+ die_nest_count++;
-+ die_owner = cpu;
-+ console_verbose();
-+ bust_spinlocks(1);
-+ return flags;
-+}
++ spin_lock_irq(&blkif_io_lock);
+
-+void __kprobes oops_end(unsigned long flags)
-+{
-+ die_owner = -1;
-+ bust_spinlocks(0);
-+ die_nest_count--;
-+ if (die_nest_count)
-+ /* We still own the lock */
-+ local_irq_restore(flags);
-+ else
-+ /* Nest count reaches zero, release the lock. */
-+ spin_unlock_irqrestore(&die_lock, flags);
-+ if (panic_on_oops)
-+ panic("Fatal exception");
-+}
++ /* Now safe for us to use the shared ring */
++ info->connected = BLKIF_STATE_CONNECTED;
+
-+void __kprobes __die(const char * str, struct pt_regs * regs, long err)
-+{
-+ static int die_counter;
-+ printk(KERN_EMERG "%s: %04lx [%u] ", str, err & 0xffff,++die_counter);
-+#ifdef CONFIG_PREEMPT
-+ printk("PREEMPT ");
-+#endif
-+#ifdef CONFIG_SMP
-+ printk("SMP ");
-+#endif
-+#ifdef CONFIG_DEBUG_PAGEALLOC
-+ printk("DEBUG_PAGEALLOC");
-+#endif
-+ printk("\n");
-+ notify_die(DIE_OOPS, str, regs, err, current->thread.trap_no, SIGSEGV);
-+ show_registers(regs);
-+ /* Executive summary in case the oops scrolled away */
-+ printk(KERN_ALERT "RIP ");
-+ printk_address(regs->rip);
-+ printk(" RSP <%016lx>\n", regs->rsp);
-+ if (kexec_should_crash(current))
-+ crash_kexec(regs);
-+}
++ /* Send off requeued requests */
++ flush_requests(info);
+
-+void die(const char * str, struct pt_regs * regs, long err)
-+{
-+ unsigned long flags = oops_begin();
++ /* Kick any other new requests queued since we resumed */
++ kick_pending_request_queues(info);
+
-+ handle_BUG(regs);
-+ __die(str, regs, err);
-+ oops_end(flags);
-+ do_exit(SIGSEGV);
++ spin_unlock_irq(&blkif_io_lock);
+}
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+void __kprobes die_nmi(char *str, struct pt_regs *regs)
++int blkfront_is_ready(struct xenbus_device *dev)
+{
-+ unsigned long flags = oops_begin();
++ struct blkfront_info *info = dev->dev.driver_data;
+
-+ /*
-+ * We are in trouble anyway, lets at least try
-+ * to get a message out.
-+ */
-+ printk(str, safe_smp_processor_id());
-+ show_registers(regs);
-+ if (kexec_should_crash(current))
-+ crash_kexec(regs);
-+ if (panic_on_timeout || panic_on_oops)
-+ panic("nmi watchdog");
-+ printk("console shuts up ...\n");
-+ oops_end(flags);
-+ nmi_exit();
-+ local_irq_enable();
-+ do_exit(SIGSEGV);
++ return info->is_ready;
+}
-+#endif
+
-+static void __kprobes do_trap(int trapnr, int signr, char *str,
-+ struct pt_regs * regs, long error_code,
-+ siginfo_t *info)
-+{
-+ struct task_struct *tsk = current;
+
-+ tsk->thread.error_code = error_code;
-+ tsk->thread.trap_no = trapnr;
++/* ** Driver Registration ** */
+
-+ if (user_mode(regs)) {
-+ if (exception_trace && unhandled_signal(tsk, signr))
-+ printk(KERN_INFO
-+ "%s[%d] trap %s rip:%lx rsp:%lx error:%lx\n",
-+ tsk->comm, tsk->pid, str,
-+ regs->rip, regs->rsp, error_code);
+
-+ if (info)
-+ force_sig_info(signr, info, tsk);
-+ else
-+ force_sig(signr, tsk);
-+ return;
-+ }
++static const struct xenbus_device_id blkfront_ids[] = {
++ { "vbd" },
++ { "" }
++};
++MODULE_ALIAS("xen:vbd");
+
++static struct xenbus_driver blkfront = {
++ .name = "vbd",
++ .owner = THIS_MODULE,
++ .ids = blkfront_ids,
++ .probe = blkfront_probe,
++ .remove = blkfront_remove,
++ .resume = blkfront_resume,
++ .otherend_changed = backend_changed,
++ .is_ready = blkfront_is_ready,
++};
+
-+ /* kernel trap */
-+ {
-+ const struct exception_table_entry *fixup;
-+ fixup = search_exception_tables(regs->rip);
-+ if (fixup)
-+ regs->rip = fixup->fixup;
-+ else
-+ die(str, regs, error_code);
-+ return;
-+ }
-+}
+
-+#define DO_ERROR(trapnr, signr, str, name) \
-+asmlinkage void do_##name(struct pt_regs * regs, long error_code) \
-+{ \
-+ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
-+ == NOTIFY_STOP) \
-+ return; \
-+ conditional_sti(regs); \
-+ do_trap(trapnr, signr, str, regs, error_code, NULL); \
-+}
++static int __init xlblk_init(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \
-+asmlinkage void do_##name(struct pt_regs * regs, long error_code) \
-+{ \
-+ siginfo_t info; \
-+ info.si_signo = signr; \
-+ info.si_errno = 0; \
-+ info.si_code = sicode; \
-+ info.si_addr = (void __user *)siaddr; \
-+ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
-+ == NOTIFY_STOP) \
-+ return; \
-+ conditional_sti(regs); \
-+ do_trap(trapnr, signr, str, regs, error_code, &info); \
++ return xenbus_register_frontend(&blkfront);
+}
++module_init(xlblk_init);
+
-+DO_ERROR_INFO( 0, SIGFPE, "divide error", divide_error, FPE_INTDIV, regs->rip)
-+DO_ERROR( 4, SIGSEGV, "overflow", overflow)
-+DO_ERROR( 5, SIGSEGV, "bounds", bounds)
-+DO_ERROR_INFO( 6, SIGILL, "invalid opcode", invalid_op, ILL_ILLOPN, regs->rip)
-+DO_ERROR( 7, SIGSEGV, "device not available", device_not_available)
-+DO_ERROR( 9, SIGFPE, "coprocessor segment overrun", coprocessor_segment_overrun)
-+DO_ERROR(10, SIGSEGV, "invalid TSS", invalid_TSS)
-+DO_ERROR(11, SIGBUS, "segment not present", segment_not_present)
-+DO_ERROR_INFO(17, SIGBUS, "alignment check", alignment_check, BUS_ADRALN, 0)
-+DO_ERROR(18, SIGSEGV, "reserved", reserved)
+
-+/* Runs on IST stack */
-+asmlinkage void do_stack_segment(struct pt_regs *regs, long error_code)
++static void __exit xlblk_exit(void)
+{
-+ if (notify_die(DIE_TRAP, "stack segment", regs, error_code,
-+ 12, SIGBUS) == NOTIFY_STOP)
-+ return;
-+ preempt_conditional_sti(regs);
-+ do_trap(12, SIGBUS, "stack segment", regs, error_code, NULL);
-+ preempt_conditional_cli(regs);
++ return xenbus_unregister_driver(&blkfront);
+}
++module_exit(xlblk_exit);
+
-+asmlinkage void do_double_fault(struct pt_regs * regs, long error_code)
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blkfront/block.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blkfront/block.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,158 @@
++/******************************************************************************
++ * block.h
++ *
++ * Shared definitions between all levels of XenLinux Virtual block devices.
++ *
++ * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
++ * Modifications by Mark A. Williamson are (c) Intel Research Cambridge
++ * Copyright (c) 2004-2005, Christian Limpach
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_DRIVERS_BLOCK_H__
++#define __XEN_DRIVERS_BLOCK_H__
++
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/kernel.h>
++#include <linux/sched.h>
++#include <linux/slab.h>
++#include <linux/string.h>
++#include <linux/errno.h>
++#include <linux/fs.h>
++#include <linux/hdreg.h>
++#include <linux/blkdev.h>
++#include <linux/major.h>
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
++#include <xen/gnttab.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/io/blkif.h>
++#include <xen/interface/io/ring.h>
++#include <asm/io.h>
++#include <asm/atomic.h>
++#include <asm/uaccess.h>
++
++#define DPRINTK(_f, _a...) pr_debug(_f, ## _a)
++
++#if 0
++#define DPRINTK_IOCTL(_f, _a...) printk(KERN_ALERT _f, ## _a)
++#else
++#define DPRINTK_IOCTL(_f, _a...) ((void)0)
++#endif
++
++struct xlbd_type_info
+{
-+ static const char str[] = "double fault";
-+ struct task_struct *tsk = current;
++ int partn_shift;
++ int disks_per_major;
++ char *devname;
++ char *diskname;
++};
+
-+ /* Return not checked because double check cannot be ignored */
-+ notify_die(DIE_TRAP, str, regs, error_code, 8, SIGSEGV);
++struct xlbd_major_info
++{
++ int major;
++ int index;
++ int usage;
++ struct xlbd_type_info *type;
++};
+
-+ tsk->thread.error_code = error_code;
-+ tsk->thread.trap_no = 8;
++struct blk_shadow {
++ blkif_request_t req;
++ unsigned long request;
++ unsigned long frame[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++};
+
-+ /* This is always a kernel trap and never fixable (and thus must
-+ never return). */
-+ for (;;)
-+ die(str, regs, error_code);
-+}
++#define BLK_RING_SIZE __RING_SIZE((blkif_sring_t *)0, PAGE_SIZE)
+
-+asmlinkage void __kprobes do_general_protection(struct pt_regs * regs,
-+ long error_code)
++/*
++ * We have one of these per vbd, whether ide, scsi or 'other'. They
++ * hang in private_data off the gendisk structure. We may end up
++ * putting all kinds of interesting stuff here :-)
++ */
++struct blkfront_info
+{
-+ struct task_struct *tsk = current;
-+
-+ conditional_sti(regs);
++ struct xenbus_device *xbdev;
++ dev_t dev;
++ struct gendisk *gd;
++ int vdevice;
++ blkif_vdev_t handle;
++ int connected;
++ int ring_ref;
++ blkif_front_ring_t ring;
++ unsigned int irq;
++ struct xlbd_major_info *mi;
++ request_queue_t *rq;
++ struct work_struct work;
++ struct gnttab_free_callback callback;
++ struct blk_shadow shadow[BLK_RING_SIZE];
++ unsigned long shadow_free;
++ int feature_barrier;
++ int is_ready;
+
-+ tsk->thread.error_code = error_code;
-+ tsk->thread.trap_no = 13;
++ /**
++ * The number of people holding this device open. We won't allow a
++ * hot-unplug unless this is 0.
++ */
++ int users;
++};
+
-+ if (user_mode(regs)) {
-+ if (exception_trace && unhandled_signal(tsk, SIGSEGV))
-+ printk(KERN_INFO
-+ "%s[%d] general protection rip:%lx rsp:%lx error:%lx\n",
-+ tsk->comm, tsk->pid,
-+ regs->rip, regs->rsp, error_code);
++extern spinlock_t blkif_io_lock;
+
-+ force_sig(SIGSEGV, tsk);
-+ return;
-+ }
++extern int blkif_open(struct inode *inode, struct file *filep);
++extern int blkif_release(struct inode *inode, struct file *filep);
++extern int blkif_ioctl(struct inode *inode, struct file *filep,
++ unsigned command, unsigned long argument);
++extern int blkif_getgeo(struct block_device *, struct hd_geometry *);
++extern int blkif_check(dev_t dev);
++extern int blkif_revalidate(dev_t dev);
++extern void do_blkif_request (request_queue_t *rq);
+
-+ /* kernel gp */
-+ {
-+ const struct exception_table_entry *fixup;
-+ fixup = search_exception_tables(regs->rip);
-+ if (fixup) {
-+ regs->rip = fixup->fixup;
-+ return;
-+ }
-+ if (notify_die(DIE_GPF, "general protection fault", regs,
-+ error_code, 13, SIGSEGV) == NOTIFY_STOP)
-+ return;
-+ die("general protection fault", regs, error_code);
-+ }
-+}
++/* Virtual block-device subsystem. */
++/* Note that xlvbd_add doesn't call add_disk for you: you're expected
++ to call add_disk on info->gd once the disk is properly connected
++ up. */
++int xlvbd_add(blkif_sector_t capacity, int device,
++ u16 vdisk_info, u16 sector_size, struct blkfront_info *info);
++void xlvbd_del(struct blkfront_info *info);
++int xlvbd_barrier(struct blkfront_info *info);
+
-+static __kprobes void
-+mem_parity_error(unsigned char reason, struct pt_regs * regs)
++#ifdef CONFIG_SYSFS
++int xlvbd_sysfs_addif(struct blkfront_info *info);
++void xlvbd_sysfs_delif(struct blkfront_info *info);
++#else
++static inline int xlvbd_sysfs_addif(struct blkfront_info *info)
+{
-+ printk("Uhhuh. NMI received. Dazed and confused, but trying to continue\n");
-+ printk("You probably have a hardware problem with your RAM chips\n");
-+
-+#if 0 /* XEN */
-+ /* Clear and disable the memory parity error line. */
-+ reason = (reason & 0xf) | 4;
-+ outb(reason, 0x61);
-+#endif /* XEN */
++ return 0;
+}
+
-+static __kprobes void
-+io_check_error(unsigned char reason, struct pt_regs * regs)
++static inline void xlvbd_sysfs_delif(struct blkfront_info *info)
+{
-+ printk("NMI: IOCK error (debug interrupt?)\n");
-+ show_registers(regs);
-+
-+#if 0 /* XEN */
-+ /* Re-enable the IOCK line, wait for a few seconds */
-+ reason = (reason & 0xf) | 8;
-+ outb(reason, 0x61);
-+ mdelay(2000);
-+ reason &= ~8;
-+ outb(reason, 0x61);
-+#endif /* XEN */
++ ;
+}
++#endif
+
-+static __kprobes void
-+unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
-+{ printk("Uhhuh. NMI received for unknown reason %02x.\n", reason);
-+ printk("Dazed and confused, but trying to continue\n");
-+ printk("Do you have a strange power saving mode enabled?\n");
-+}
++#endif /* __XEN_DRIVERS_BLOCK_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blkfront/vbd.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blkfront/vbd.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,460 @@
++/******************************************************************************
++ * vbd.c
++ *
++ * XenLinux virtual block-device driver (xvd).
++ *
++ * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
++ * Modifications by Mark A. Williamson are (c) Intel Research Cambridge
++ * Copyright (c) 2004-2005, Christian Limpach
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+/* Runs on IST stack. This code must keep interrupts off all the time.
-+ Nested NMIs are prevented by the CPU. */
-+asmlinkage __kprobes void default_do_nmi(struct pt_regs *regs)
-+{
-+ unsigned char reason = 0;
-+ int cpu;
++#include "block.h"
++#include <linux/blkdev.h>
++#include <linux/list.h>
+
-+ cpu = smp_processor_id();
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
+
-+ /* Only the BSP gets external NMIs from the system. */
-+ if (!cpu)
-+ reason = get_nmi_reason();
++#define BLKIF_MAJOR(dev) ((dev)>>8)
++#define BLKIF_MINOR(dev) ((dev) & 0xff)
+
-+ if (!(reason & 0xc0)) {
-+ if (notify_die(DIE_NMI_IPI, "nmi_ipi", regs, reason, 2, SIGINT)
-+ == NOTIFY_STOP)
-+ return;
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ /*
-+ * Ok, so this is none of the documented NMI sources,
-+ * so it must be the NMI watchdog.
-+ */
-+ if (nmi_watchdog > 0) {
-+ nmi_watchdog_tick(regs,reason);
-+ return;
-+ }
-+#endif
-+ unknown_nmi_error(reason, regs);
-+ return;
-+ }
-+ if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP)
-+ return;
++#define EXT_SHIFT 28
++#define EXTENDED (1<<EXT_SHIFT)
++#define VDEV_IS_EXTENDED(dev) ((dev)&(EXTENDED))
++#define BLKIF_MINOR_EXT(dev) ((dev)&(~EXTENDED))
+
-+ /* AK: following checks seem to be broken on modern chipsets. FIXME */
++/*
++ * For convenience we distinguish between ide, scsi and 'other' (i.e.,
++ * potentially combinations of the two) in the naming scheme and in a few other
++ * places.
++ */
+
-+ if (reason & 0x80)
-+ mem_parity_error(reason, regs);
-+ if (reason & 0x40)
-+ io_check_error(reason, regs);
-+}
++#define NUM_IDE_MAJORS 10
++#define NUM_SCSI_MAJORS 17
++#define NUM_VBD_MAJORS 2
+
-+/* runs on IST stack. */
-+asmlinkage void __kprobes do_int3(struct pt_regs * regs, long error_code)
-+{
-+ if (notify_die(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP) == NOTIFY_STOP) {
-+ return;
-+ }
-+ preempt_conditional_sti(regs);
-+ do_trap(3, SIGTRAP, "int3", regs, error_code, NULL);
-+ preempt_conditional_cli(regs);
-+}
++static struct xlbd_type_info xlbd_ide_type = {
++ .partn_shift = 6,
++ .disks_per_major = 2,
++ .devname = "ide",
++ .diskname = "hd",
++};
+
-+/* Help handler running on IST stack to switch back to user stack
-+ for scheduling or signal handling. The actual stack switch is done in
-+ entry.S */
-+asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs)
-+{
-+ struct pt_regs *regs = eregs;
-+ /* Did already sync */
-+ if (eregs == (struct pt_regs *)eregs->rsp)
-+ ;
-+ /* Exception from user space */
-+ else if (user_mode(eregs))
-+ regs = task_pt_regs(current);
-+ /* Exception from kernel and interrupts are enabled. Move to
-+ kernel process stack. */
-+ else if (eregs->eflags & X86_EFLAGS_IF)
-+ regs = (struct pt_regs *)(eregs->rsp -= sizeof(struct pt_regs));
-+ if (eregs != regs)
-+ *regs = *eregs;
-+ return regs;
-+}
++static struct xlbd_type_info xlbd_scsi_type = {
++ .partn_shift = 4,
++ .disks_per_major = 16,
++ .devname = "sd",
++ .diskname = "sd",
++};
+
-+/* runs on IST stack. */
-+asmlinkage void __kprobes do_debug(struct pt_regs * regs,
-+ unsigned long error_code)
-+{
-+ unsigned long condition;
-+ struct task_struct *tsk = current;
-+ siginfo_t info;
++static struct xlbd_type_info xlbd_vbd_type = {
++ .partn_shift = 4,
++ .disks_per_major = 16,
++ .devname = "xvd",
++ .diskname = "xvd",
++};
+
-+ get_debugreg(condition, 6);
++static struct xlbd_type_info xlbd_vbd_type_ext = {
++ .partn_shift = 8,
++ .disks_per_major = 256,
++ .devname = "xvd",
++ .diskname = "xvd",
++};
+
-+ if (notify_die(DIE_DEBUG, "debug", regs, condition, error_code,
-+ SIGTRAP) == NOTIFY_STOP)
-+ return;
++static struct xlbd_major_info *major_info[NUM_IDE_MAJORS + NUM_SCSI_MAJORS +
++ NUM_VBD_MAJORS];
+
-+ preempt_conditional_sti(regs);
++#define XLBD_MAJOR_IDE_START 0
++#define XLBD_MAJOR_SCSI_START (NUM_IDE_MAJORS)
++#define XLBD_MAJOR_VBD_START (NUM_IDE_MAJORS + NUM_SCSI_MAJORS)
+
-+ /* Mask out spurious debug traps due to lazy DR7 setting */
-+ if (condition & (DR_TRAP0|DR_TRAP1|DR_TRAP2|DR_TRAP3)) {
-+ if (!tsk->thread.debugreg7) {
-+ goto clear_dr7;
-+ }
-+ }
++#define XLBD_MAJOR_IDE_RANGE XLBD_MAJOR_IDE_START ... XLBD_MAJOR_SCSI_START - 1
++#define XLBD_MAJOR_SCSI_RANGE XLBD_MAJOR_SCSI_START ... XLBD_MAJOR_VBD_START - 1
++#define XLBD_MAJOR_VBD_RANGE XLBD_MAJOR_VBD_START ... XLBD_MAJOR_VBD_START + NUM_VBD_MAJORS - 1
+
-+ tsk->thread.debugreg6 = condition;
++static struct block_device_operations xlvbd_block_fops =
++{
++ .owner = THIS_MODULE,
++ .open = blkif_open,
++ .release = blkif_release,
++ .ioctl = blkif_ioctl,
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
++ .getgeo = blkif_getgeo
++#endif
++};
+
-+ /* Mask out spurious TF errors due to lazy TF clearing */
-+ if (condition & DR_STEP) {
-+ /*
-+ * The TF error should be masked out only if the current
-+ * process is not traced and if the TRAP flag has been set
-+ * previously by a tracing process (condition detected by
-+ * the PT_DTRACE flag); remember that the i386 TRAP flag
-+ * can be modified by the process itself in user mode,
-+ * allowing programs to debug themselves without the ptrace()
-+ * interface.
-+ */
-+ if (!user_mode(regs))
-+ goto clear_TF_reenable;
-+ /*
-+ * Was the TF flag set by a debugger? If so, clear it now,
-+ * so that register information is correct.
-+ */
-+ if (tsk->ptrace & PT_DTRACE) {
-+ regs->eflags &= ~TF_MASK;
-+ tsk->ptrace &= ~PT_DTRACE;
-+ }
-+ }
++DEFINE_SPINLOCK(blkif_io_lock);
+
-+ /* Ok, finally something we can handle */
-+ tsk->thread.trap_no = 1;
-+ tsk->thread.error_code = error_code;
-+ info.si_signo = SIGTRAP;
-+ info.si_errno = 0;
-+ info.si_code = TRAP_BRKPT;
-+ info.si_addr = user_mode(regs) ? (void __user *)regs->rip : NULL;
-+ force_sig_info(SIGTRAP, &info, tsk);
++static struct xlbd_major_info *
++xlbd_alloc_major_info(int major, int minor, int index)
++{
++ struct xlbd_major_info *ptr;
++ int do_register;
+
-+clear_dr7:
-+ set_debugreg(0UL, 7);
-+ preempt_conditional_cli(regs);
-+ return;
++ ptr = kzalloc(sizeof(struct xlbd_major_info), GFP_KERNEL);
++ if (ptr == NULL)
++ return NULL;
+
-+clear_TF_reenable:
-+ set_tsk_thread_flag(tsk, TIF_SINGLESTEP);
-+ regs->eflags &= ~TF_MASK;
-+ preempt_conditional_cli(regs);
-+}
++ ptr->major = major;
++ do_register = 1;
+
-+static int kernel_math_error(struct pt_regs *regs, const char *str, int trapnr)
-+{
-+ const struct exception_table_entry *fixup;
-+ fixup = search_exception_tables(regs->rip);
-+ if (fixup) {
-+ regs->rip = fixup->fixup;
-+ return 1;
-+ }
-+ notify_die(DIE_GPF, str, regs, 0, trapnr, SIGFPE);
-+ /* Illegal floating point operation in the kernel */
-+ current->thread.trap_no = trapnr;
-+ die(str, regs, 0);
-+ return 0;
-+}
++ switch (index) {
++ case XLBD_MAJOR_IDE_RANGE:
++ ptr->type = &xlbd_ide_type;
++ ptr->index = index - XLBD_MAJOR_IDE_START;
++ break;
++ case XLBD_MAJOR_SCSI_RANGE:
++ ptr->type = &xlbd_scsi_type;
++ ptr->index = index - XLBD_MAJOR_SCSI_START;
++ break;
++ case XLBD_MAJOR_VBD_RANGE:
++ ptr->index = 0;
++ if ((index - XLBD_MAJOR_VBD_START) == 0)
++ ptr->type = &xlbd_vbd_type;
++ else
++ ptr->type = &xlbd_vbd_type_ext;
+
-+/*
-+ * Note that we play around with the 'TS' bit in an attempt to get
-+ * the correct behaviour even in the presence of the asynchronous
-+ * IRQ13 behaviour
-+ */
-+asmlinkage void do_coprocessor_error(struct pt_regs *regs)
-+{
-+ void __user *rip = (void __user *)(regs->rip);
-+ struct task_struct * task;
-+ siginfo_t info;
-+ unsigned short cwd, swd;
++ /*
++ * if someone already registered block major 202,
++ * don't try to register it again
++ */
++ if (major_info[XLBD_MAJOR_VBD_START] != NULL)
++ do_register = 0;
++ break;
++ }
+
-+ conditional_sti(regs);
-+ if (!user_mode(regs) &&
-+ kernel_math_error(regs, "kernel x87 math error", 16))
-+ return;
++ if (do_register) {
++ if (register_blkdev(ptr->major, ptr->type->devname)) {
++ kfree(ptr);
++ return NULL;
++ }
+
-+ /*
-+ * Save the info for the exception handler and clear the error.
-+ */
-+ task = current;
-+ save_init_fpu(task);
-+ task->thread.trap_no = 16;
-+ task->thread.error_code = 0;
-+ info.si_signo = SIGFPE;
-+ info.si_errno = 0;
-+ info.si_code = __SI_FAULT;
-+ info.si_addr = rip;
-+ /*
-+ * (~cwd & swd) will mask out exceptions that are not set to unmasked
-+ * status. 0x3f is the exception bits in these regs, 0x200 is the
-+ * C1 reg you need in case of a stack fault, 0x040 is the stack
-+ * fault bit. We should only be taking one exception at a time,
-+ * so if this combination doesn't produce any single exception,
-+ * then we have a bad program that isn't synchronizing its FPU usage
-+ * and it will suffer the consequences since we won't be able to
-+ * fully reproduce the context of the exception
-+ */
-+ cwd = get_fpu_cwd(task);
-+ swd = get_fpu_swd(task);
-+ switch (swd & ~cwd & 0x3f) {
-+ case 0x000:
-+ default:
-+ break;
-+ case 0x001: /* Invalid Op */
-+ /*
-+ * swd & 0x240 == 0x040: Stack Underflow
-+ * swd & 0x240 == 0x240: Stack Overflow
-+ * User must clear the SF bit (0x40) if set
-+ */
-+ info.si_code = FPE_FLTINV;
-+ break;
-+ case 0x002: /* Denormalize */
-+ case 0x010: /* Underflow */
-+ info.si_code = FPE_FLTUND;
-+ break;
-+ case 0x004: /* Zero Divide */
-+ info.si_code = FPE_FLTDIV;
-+ break;
-+ case 0x008: /* Overflow */
-+ info.si_code = FPE_FLTOVF;
-+ break;
-+ case 0x020: /* Precision */
-+ info.si_code = FPE_FLTRES;
-+ break;
++ printk("xen-vbd: registered block device major %i\n", ptr->major);
+ }
-+ force_sig_info(SIGFPE, &info, task);
-+}
+
-+asmlinkage void bad_intr(void)
-+{
-+ printk("bad interrupt");
++ major_info[index] = ptr;
++ return ptr;
+}
+
-+asmlinkage void do_simd_coprocessor_error(struct pt_regs *regs)
++static struct xlbd_major_info *
++xlbd_get_major_info(int major, int minor, int vdevice)
+{
-+ void __user *rip = (void __user *)(regs->rip);
-+ struct task_struct * task;
-+ siginfo_t info;
-+ unsigned short mxcsr;
-+
-+ conditional_sti(regs);
-+ if (!user_mode(regs) &&
-+ kernel_math_error(regs, "kernel simd math error", 19))
-+ return;
++ struct xlbd_major_info *mi;
++ int index;
+
-+ /*
-+ * Save the info for the exception handler and clear the error.
-+ */
-+ task = current;
-+ save_init_fpu(task);
-+ task->thread.trap_no = 19;
-+ task->thread.error_code = 0;
-+ info.si_signo = SIGFPE;
-+ info.si_errno = 0;
-+ info.si_code = __SI_FAULT;
-+ info.si_addr = rip;
-+ /*
-+ * The SIMD FPU exceptions are handled a little differently, as there
-+ * is only a single status/control register. Thus, to determine which
-+ * unmasked exception was caught we must mask the exception mask bits
-+ * at 0x1f80, and then use these to mask the exception bits at 0x3f.
-+ */
-+ mxcsr = get_fpu_mxcsr(task);
-+ switch (~((mxcsr & 0x1f80) >> 7) & (mxcsr & 0x3f)) {
-+ case 0x000:
-+ default:
-+ break;
-+ case 0x001: /* Invalid Op */
-+ info.si_code = FPE_FLTINV;
-+ break;
-+ case 0x002: /* Denormalize */
-+ case 0x010: /* Underflow */
-+ info.si_code = FPE_FLTUND;
-+ break;
-+ case 0x004: /* Zero Divide */
-+ info.si_code = FPE_FLTDIV;
-+ break;
-+ case 0x008: /* Overflow */
-+ info.si_code = FPE_FLTOVF;
-+ break;
-+ case 0x020: /* Precision */
-+ info.si_code = FPE_FLTRES;
-+ break;
++ switch (major) {
++ case IDE0_MAJOR: index = 0; break;
++ case IDE1_MAJOR: index = 1; break;
++ case IDE2_MAJOR: index = 2; break;
++ case IDE3_MAJOR: index = 3; break;
++ case IDE4_MAJOR: index = 4; break;
++ case IDE5_MAJOR: index = 5; break;
++ case IDE6_MAJOR: index = 6; break;
++ case IDE7_MAJOR: index = 7; break;
++ case IDE8_MAJOR: index = 8; break;
++ case IDE9_MAJOR: index = 9; break;
++ case SCSI_DISK0_MAJOR: index = 10; break;
++ case SCSI_DISK1_MAJOR ... SCSI_DISK7_MAJOR:
++ index = 11 + major - SCSI_DISK1_MAJOR;
++ break;
++ case SCSI_DISK8_MAJOR ... SCSI_DISK15_MAJOR:
++ index = 18 + major - SCSI_DISK8_MAJOR;
++ break;
++ case SCSI_CDROM_MAJOR: index = 26; break;
++ default:
++ if (!VDEV_IS_EXTENDED(vdevice))
++ index = 27;
++ else
++ index = 28;
++ break;
+ }
-+ force_sig_info(SIGFPE, &info, task);
-+}
+
-+asmlinkage void do_spurious_interrupt_bug(struct pt_regs * regs)
-+{
++ mi = ((major_info[index] != NULL) ? major_info[index] :
++ xlbd_alloc_major_info(major, minor, index));
++ if (mi)
++ mi->usage++;
++ return mi;
+}
+
-+#if 0
-+asmlinkage void __attribute__((weak)) smp_thermal_interrupt(void)
++static void
++xlbd_put_major_info(struct xlbd_major_info *mi)
+{
++ mi->usage--;
++ /* XXX: release major if 0 */
+}
-+#endif
+
-+asmlinkage void __attribute__((weak)) mce_threshold_interrupt(void)
++static int
++xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size)
+{
-+}
++ request_queue_t *rq;
+
-+/*
-+ * 'math_state_restore()' saves the current math information in the
-+ * old math state array, and gets the new ones from the current task
-+ *
-+ * Careful.. There are problems with IBM-designed IRQ13 behaviour.
-+ * Don't touch unless you *really* know how it works.
-+ */
-+asmlinkage void math_state_restore(void)
-+{
-+ struct task_struct *me = current;
-+ /* clts(); */ /* 'clts' is done for us by Xen during virtual trap. */
++ rq = blk_init_queue(do_blkif_request, &blkif_io_lock);
++ if (rq == NULL)
++ return -1;
+
-+ if (!used_math())
-+ init_fpu(me);
-+ restore_fpu_checking(&me->thread.i387.fxsave);
-+ task_thread_info(me)->status |= TS_USEDFPU;
-+}
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,10)
++ elevator_init(rq, "noop");
++#else
++ elevator_init(rq, &elevator_noop);
++#endif
+
++ /* Hard sector size and max sectors impersonate the equiv. hardware. */
++ blk_queue_hardsect_size(rq, sector_size);
++ blk_queue_max_sectors(rq, 512);
+
-+/*
-+ * NB. All these are "interrupt gates" (i.e. events_mask is set) because we
-+ * specify <dpl>|4 in the second field.
-+ */
-+static trap_info_t trap_table[] = {
-+ { 0, 0|4, __KERNEL_CS, (unsigned long)divide_error },
-+ { 1, 0|4, __KERNEL_CS, (unsigned long)debug },
-+ { 3, 3|4, __KERNEL_CS, (unsigned long)int3 },
-+ { 4, 3|4, __KERNEL_CS, (unsigned long)overflow },
-+ { 5, 0|4, __KERNEL_CS, (unsigned long)bounds },
-+ { 6, 0|4, __KERNEL_CS, (unsigned long)invalid_op },
-+ { 7, 0|4, __KERNEL_CS, (unsigned long)device_not_available },
-+ { 9, 0|4, __KERNEL_CS, (unsigned long)coprocessor_segment_overrun},
-+ { 10, 0|4, __KERNEL_CS, (unsigned long)invalid_TSS },
-+ { 11, 0|4, __KERNEL_CS, (unsigned long)segment_not_present },
-+ { 12, 0|4, __KERNEL_CS, (unsigned long)stack_segment },
-+ { 13, 0|4, __KERNEL_CS, (unsigned long)general_protection },
-+ { 14, 0|4, __KERNEL_CS, (unsigned long)page_fault },
-+ { 15, 0|4, __KERNEL_CS, (unsigned long)spurious_interrupt_bug },
-+ { 16, 0|4, __KERNEL_CS, (unsigned long)coprocessor_error },
-+ { 17, 0|4, __KERNEL_CS, (unsigned long)alignment_check },
-+#ifdef CONFIG_X86_MCE
-+ { 18, 0|4, __KERNEL_CS, (unsigned long)machine_check },
-+#endif
-+ { 19, 0|4, __KERNEL_CS, (unsigned long)simd_coprocessor_error },
-+#ifdef CONFIG_IA32_EMULATION
-+ { IA32_SYSCALL_VECTOR, 3|4, __KERNEL_CS, (unsigned long)ia32_syscall},
-+#endif
-+ { 0, 0, 0, 0 }
-+};
++ /* Each segment in a request is up to an aligned page in size. */
++ blk_queue_segment_boundary(rq, PAGE_SIZE - 1);
++ blk_queue_max_segment_size(rq, PAGE_SIZE);
+
-+void __init trap_init(void)
-+{
-+ int ret;
++ /* Ensure a merged request will fit in a single I/O ring slot. */
++ blk_queue_max_phys_segments(rq, BLKIF_MAX_SEGMENTS_PER_REQUEST);
++ blk_queue_max_hw_segments(rq, BLKIF_MAX_SEGMENTS_PER_REQUEST);
+
-+ ret = HYPERVISOR_set_trap_table(trap_table);
-+
-+ if (ret)
-+ printk("HYPERVISOR_set_trap_table faild: error %d\n",
-+ ret);
++ /* Make sure buffer addresses are sector-aligned. */
++ blk_queue_dma_alignment(rq, 511);
+
-+ /*
-+ * Should be a barrier for any external CPU state.
-+ */
-+ cpu_init();
++ /* Make sure we don't use bounce buffers. */
++ blk_queue_bounce_limit(rq, BLK_BOUNCE_ANY);
++
++ gd->queue = rq;
++
++ return 0;
+}
+
-+void smp_trap_init(trap_info_t *trap_ctxt)
++static int
++xlvbd_alloc_gendisk(int major, int minor, blkif_sector_t capacity, int vdevice,
++ u16 vdisk_info, u16 sector_size,
++ struct blkfront_info *info)
+{
-+ trap_info_t *t = trap_table;
++ struct gendisk *gd;
++ struct xlbd_major_info *mi;
++ int nr_minors = 1;
++ int err = -ENODEV;
++ unsigned int offset;
+
-+ for (t = trap_table; t->address; t++) {
-+ trap_ctxt[t->vector].flags = t->flags;
-+ trap_ctxt[t->vector].cs = t->cs;
-+ trap_ctxt[t->vector].address = t->address;
-+ }
-+}
++ BUG_ON(info->gd != NULL);
++ BUG_ON(info->mi != NULL);
++ BUG_ON(info->rq != NULL);
+
++ mi = xlbd_get_major_info(major, minor, vdevice);
++ if (mi == NULL)
++ goto out;
++ info->mi = mi;
+
-+/* Actual parsing is done early in setup.c. */
-+static int __init oops_dummy(char *s)
-+{
-+ panic_on_oops = 1;
-+ return 1;
-+}
-+__setup("oops=", oops_dummy);
++ if ((minor & ((1 << mi->type->partn_shift) - 1)) == 0)
++ nr_minors = 1 << mi->type->partn_shift;
+
-+static int __init kstack_setup(char *s)
-+{
-+ kstack_depth_to_print = simple_strtoul(s,NULL,0);
-+ return 1;
-+}
-+__setup("kstack=", kstack_setup);
++ gd = alloc_disk(nr_minors);
++ if (gd == NULL)
++ goto out;
+
-+#ifdef CONFIG_STACK_UNWIND
-+static int __init call_trace_setup(char *s)
-+{
-+ if (strcmp(s, "old") == 0)
-+ call_trace = -1;
-+ else if (strcmp(s, "both") == 0)
-+ call_trace = 0;
-+ else if (strcmp(s, "newfallback") == 0)
-+ call_trace = 1;
-+ else if (strcmp(s, "new") == 0)
-+ call_trace = 2;
-+ return 1;
-+}
-+__setup("call_trace=", call_trace_setup);
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/vmlinux.lds.S tmp-linux-2.6-xen.patch/arch/x86_64/kernel/vmlinux.lds.S
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/vmlinux.lds.S 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/vmlinux.lds.S 2007-10-14 01:51:15.000000000 +0200
-@@ -13,6 +13,13 @@ OUTPUT_FORMAT("elf64-x86-64", "elf64-x86
- OUTPUT_ARCH(i386:x86-64)
- ENTRY(phys_startup_64)
- jiffies_64 = jiffies;
-+PHDRS {
-+ text PT_LOAD FLAGS(5); /* R_E */
-+ data PT_LOAD FLAGS(7); /* RWE */
-+ user PT_LOAD FLAGS(7); /* RWE */
-+ data.init PT_LOAD FLAGS(7); /* RWE */
-+ note PT_NOTE FLAGS(4); /* R__ */
-+}
- SECTIONS
- {
- . = __START_KERNEL;
-@@ -31,7 +38,7 @@ SECTIONS
- KPROBES_TEXT
- *(.fixup)
- *(.gnu.warning)
-- } = 0x9090
-+ } :text = 0x9090
- /* out-of-line lock text */
- .text.lock : AT(ADDR(.text.lock) - LOAD_OFFSET) { *(.text.lock) }
-
-@@ -57,17 +64,10 @@ SECTIONS
- .data : AT(ADDR(.data) - LOAD_OFFSET) {
- *(.data)
- CONSTRUCTORS
-- }
-+ } :data
-
- _edata = .; /* End of data section */
-
-- __bss_start = .; /* BSS */
-- .bss : AT(ADDR(.bss) - LOAD_OFFSET) {
-- *(.bss.page_aligned)
-- *(.bss)
-- }
-- __bss_stop = .;
--
- . = ALIGN(PAGE_SIZE);
- . = ALIGN(CONFIG_X86_L1_CACHE_BYTES);
- .data.cacheline_aligned : AT(ADDR(.data.cacheline_aligned) - LOAD_OFFSET) {
-@@ -89,7 +89,7 @@ SECTIONS
- #define VVIRT(x) (ADDR(x) - VVIRT_OFFSET)
-
- . = VSYSCALL_ADDR;
-- .vsyscall_0 : AT(VSYSCALL_PHYS_ADDR) { *(.vsyscall_0) }
-+ .vsyscall_0 : AT(VSYSCALL_PHYS_ADDR) { *(.vsyscall_0) } :user
- __vsyscall_0 = VSYSCALL_VIRT_ADDR;
-
- . = ALIGN(CONFIG_X86_L1_CACHE_BYTES);
-@@ -132,7 +132,7 @@ SECTIONS
- . = ALIGN(8192); /* init_task */
- .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) {
- *(.data.init_task)
-- }
-+ }:data.init
-
- . = ALIGN(4096);
- .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) {
-@@ -222,6 +222,14 @@ SECTIONS
- . = ALIGN(4096);
- __nosave_end = .;
-
-+ __bss_start = .; /* BSS */
-+ . = ALIGN(4096);
-+ .bss : AT(ADDR(.bss) - LOAD_OFFSET) {
-+ *(.bss.page_aligned)
-+ *(.bss)
++ offset = mi->index * mi->type->disks_per_major +
++ (minor >> mi->type->partn_shift);
++ if (nr_minors > 1) {
++ if (offset < 26) {
++ sprintf(gd->disk_name, "%s%c",
++ mi->type->diskname, 'a' + offset );
++ }
++ else {
++ sprintf(gd->disk_name, "%s%c%c",
++ mi->type->diskname,
++ 'a' + ((offset/26)-1), 'a' + (offset%26) );
++ }
++ }
++ else {
++ if (offset < 26) {
++ sprintf(gd->disk_name, "%s%c%d",
++ mi->type->diskname,
++ 'a' + offset,
++ minor & ((1 << mi->type->partn_shift) - 1));
++ }
++ else {
++ sprintf(gd->disk_name, "%s%c%c%d",
++ mi->type->diskname,
++ 'a' + ((offset/26)-1), 'a' + (offset%26),
++ minor & ((1 << mi->type->partn_shift) - 1));
++ }
+ }
-+ __bss_stop = .;
+
- _end = . ;
-
- /* Sections to be discarded */
-@@ -235,4 +243,6 @@ SECTIONS
- STABS_DEBUG
-
- DWARF_DEBUG
++ gd->major = mi->major;
++ gd->first_minor = minor;
++ gd->fops = &xlvbd_block_fops;
++ gd->private_data = info;
++ gd->driverfs_dev = &(info->xbdev->dev);
++ set_capacity(gd, capacity);
+
-+ NOTES
- }
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/vsyscall-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/kernel/vsyscall-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/vsyscall-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/vsyscall-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,239 @@
-+/*
-+ * linux/arch/x86_64/kernel/vsyscall.c
-+ *
-+ * Copyright (C) 2001 Andrea Arcangeli <andrea at suse.de> SuSE
-+ * Copyright 2003 Andi Kleen, SuSE Labs.
-+ *
-+ * Thanks to hpa at transmeta.com for some useful hint.
-+ * Special thanks to Ingo Molnar for his early experience with
-+ * a different vsyscall implementation for Linux/IA32 and for the name.
-+ *
-+ * vsyscall 1 is located at -10Mbyte, vsyscall 2 is located
-+ * at virtual address -10Mbyte+1024bytes etc... There are at max 4
-+ * vsyscalls. One vsyscall can reserve more than 1 slot to avoid
-+ * jumping out of line if necessary. We cannot add more with this
-+ * mechanism because older kernels won't return -ENOSYS.
-+ * If we want more than four we need a vDSO.
-+ *
-+ * Note: the concept clashes with user mode linux. If you use UML and
-+ * want per guest time just set the kernel.vsyscall64 sysctl to 0.
-+ */
++ if (xlvbd_init_blk_queue(gd, sector_size)) {
++ del_gendisk(gd);
++ goto out;
++ }
+
-+#include <linux/time.h>
-+#include <linux/init.h>
-+#include <linux/kernel.h>
-+#include <linux/timer.h>
-+#include <linux/seqlock.h>
-+#include <linux/jiffies.h>
-+#include <linux/sysctl.h>
++ info->rq = gd->queue;
++ info->gd = gd;
+
-+#include <asm/vsyscall.h>
-+#include <asm/pgtable.h>
-+#include <asm/page.h>
-+#include <asm/fixmap.h>
-+#include <asm/errno.h>
-+#include <asm/io.h>
++ if (info->feature_barrier)
++ xlvbd_barrier(info);
+
-+#define __vsyscall(nr) __attribute__ ((unused,__section__(".vsyscall_" #nr)))
++ if (vdisk_info & VDISK_READONLY)
++ set_disk_ro(gd, 1);
+
-+int __sysctl_vsyscall __section_sysctl_vsyscall = 1;
-+seqlock_t __xtime_lock __section_xtime_lock = SEQLOCK_UNLOCKED;
++ if (vdisk_info & VDISK_REMOVABLE)
++ gd->flags |= GENHD_FL_REMOVABLE;
+
-+#include <asm/unistd.h>
++ if (vdisk_info & VDISK_CDROM)
++ gd->flags |= GENHD_FL_CD;
+
-+static __always_inline void timeval_normalize(struct timeval * tv)
-+{
-+ time_t __sec;
++ return 0;
+
-+ __sec = tv->tv_usec / 1000000;
-+ if (__sec) {
-+ tv->tv_usec %= 1000000;
-+ tv->tv_sec += __sec;
-+ }
++ out:
++ if (mi)
++ xlbd_put_major_info(mi);
++ info->mi = NULL;
++ return err;
+}
+
-+static __always_inline void do_vgettimeofday(struct timeval * tv)
++int
++xlvbd_add(blkif_sector_t capacity, int vdevice, u16 vdisk_info,
++ u16 sector_size, struct blkfront_info *info)
+{
-+ long sequence, t;
-+ unsigned long sec, usec;
++ struct block_device *bd;
++ int err = 0;
++ int major, minor;
+
-+ do {
-+ sequence = read_seqbegin(&__xtime_lock);
-+
-+ sec = __xtime.tv_sec;
-+ usec = (__xtime.tv_nsec / 1000) +
-+ (__jiffies - __wall_jiffies) * (1000000 / HZ);
++ if ((vdevice>>EXT_SHIFT) > 1) {
++ /* this is above the extended range; something is wrong */
++ printk(KERN_WARNING "blkfront: vdevice 0x%x is above the extended range; ignoring\n", vdevice);
++ return -ENODEV;
++ }
+
-+ if (__vxtime.mode != VXTIME_HPET) {
-+ t = get_cycles_sync();
-+ if (t < __vxtime.last_tsc)
-+ t = __vxtime.last_tsc;
-+ usec += ((t - __vxtime.last_tsc) *
-+ __vxtime.tsc_quot) >> 32;
-+ /* See comment in x86_64 do_gettimeofday. */
-+ } else {
-+ usec += ((readl((void *)fix_to_virt(VSYSCALL_HPET) + 0xf0) -
-+ __vxtime.last) * __vxtime.quot) >> 32;
-+ }
-+ } while (read_seqretry(&__xtime_lock, sequence));
++ if (!VDEV_IS_EXTENDED(vdevice)) {
++ major = BLKIF_MAJOR(vdevice);
++ minor = BLKIF_MINOR(vdevice);
++ }
++ else {
++ major = 202;
++ minor = BLKIF_MINOR_EXT(vdevice);
++ }
+
-+ tv->tv_sec = sec + usec / 1000000;
-+ tv->tv_usec = usec % 1000000;
-+}
++ info->dev = MKDEV(major, minor);
++ bd = bdget(info->dev);
++ if (bd == NULL)
++ return -ENODEV;
+
-+/* RED-PEN may want to readd seq locking, but then the variable should be write-once. */
-+static __always_inline void do_get_tz(struct timezone * tz)
-+{
-+ *tz = __sys_tz;
-+}
++ err = xlvbd_alloc_gendisk(major, minor, capacity, vdevice, vdisk_info,
++ sector_size, info);
+
-+static __always_inline int gettimeofday(struct timeval *tv, struct timezone *tz)
-+{
-+ int ret;
-+ asm volatile("vsysc2: syscall"
-+ : "=a" (ret)
-+ : "0" (__NR_gettimeofday),"D" (tv),"S" (tz) : __syscall_clobber );
-+ return ret;
++ bdput(bd);
++ return err;
+}
+
-+static __always_inline long time_syscall(long *t)
++void
++xlvbd_del(struct blkfront_info *info)
+{
-+ long secs;
-+ asm volatile("vsysc1: syscall"
-+ : "=a" (secs)
-+ : "0" (__NR_time),"D" (t) : __syscall_clobber);
-+ return secs;
-+}
++ if (info->mi == NULL)
++ return;
+
-+int __vsyscall(0) vgettimeofday(struct timeval * tv, struct timezone * tz)
-+{
-+ if (!__sysctl_vsyscall)
-+ return gettimeofday(tv,tz);
-+ if (tv)
-+ do_vgettimeofday(tv);
-+ if (tz)
-+ do_get_tz(tz);
-+ return 0;
-+}
++ BUG_ON(info->gd == NULL);
++ del_gendisk(info->gd);
++ put_disk(info->gd);
++ info->gd = NULL;
+
-+/* This will break when the xtime seconds get inaccurate, but that is
-+ * unlikely */
-+time_t __vsyscall(1) vtime(time_t *t)
-+{
-+ if (!__sysctl_vsyscall)
-+ return time_syscall(t);
-+ else if (t)
-+ *t = __xtime.tv_sec;
-+ return __xtime.tv_sec;
++ xlbd_put_major_info(info->mi);
++ info->mi = NULL;
++
++ BUG_ON(info->rq == NULL);
++ blk_cleanup_queue(info->rq);
++ info->rq = NULL;
+}
+
-+long __vsyscall(2) venosys_0(void)
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
++int
++xlvbd_barrier(struct blkfront_info *info)
+{
-+ return -ENOSYS;
-+}
++ int err;
+
-+long __vsyscall(3) venosys_1(void)
++ err = blk_queue_ordered(info->rq,
++ info->feature_barrier ? QUEUE_ORDERED_DRAIN : QUEUE_ORDERED_NONE, NULL);
++ if (err)
++ return err;
++ printk(KERN_INFO "blkfront: %s: barriers %s\n",
++ info->gd->disk_name, info->feature_barrier ? "enabled" : "disabled");
++ return 0;
++}
++#else
++int
++xlvbd_barrier(struct blkfront_info *info)
+{
++ printk(KERN_INFO "blkfront: %s: barriers disabled\n", info->gd->disk_name);
+ return -ENOSYS;
+}
++#endif
+
-+#ifdef CONFIG_SYSCTL
-+
-+#define SYSCALL 0x050f
-+#define NOP2 0x9090
-+
-+/*
-+ * NOP out syscall in vsyscall page when not needed.
-+ */
-+static int vsyscall_sysctl_change(ctl_table *ctl, int write, struct file * filp,
-+ void __user *buffer, size_t *lenp, loff_t *ppos)
++#ifdef CONFIG_SYSFS
++static ssize_t show_media(struct device *dev,
++ struct device_attribute *attr, char *buf)
+{
-+ extern u16 vsysc1, vsysc2;
-+ u16 *map1, *map2;
-+ int ret = proc_dointvec(ctl, write, filp, buffer, lenp, ppos);
-+ if (!write)
-+ return ret;
-+ /* gcc has some trouble with __va(__pa()), so just do it this
-+ way. */
-+ map1 = ioremap(__pa_symbol(&vsysc1), 2);
-+ if (!map1)
-+ return -ENOMEM;
-+ map2 = ioremap(__pa_symbol(&vsysc2), 2);
-+ if (!map2) {
-+ ret = -ENOMEM;
-+ goto out;
-+ }
-+ if (!sysctl_vsyscall) {
-+ *map1 = SYSCALL;
-+ *map2 = SYSCALL;
-+ } else {
-+ *map1 = NOP2;
-+ *map2 = NOP2;
-+ }
-+ iounmap(map2);
-+out:
-+ iounmap(map1);
-+ return ret;
-+}
++ struct xenbus_device *xendev = to_xenbus_device(dev);
++ struct blkfront_info *info = xendev->dev.driver_data;
+
-+static int vsyscall_sysctl_nostrat(ctl_table *t, int __user *name, int nlen,
-+ void __user *oldval, size_t __user *oldlenp,
-+ void __user *newval, size_t newlen,
-+ void **context)
-+{
-+ return -ENOSYS;
++ if (info->gd->flags & GENHD_FL_CD)
++ return sprintf(buf, "cdrom\n");
++ return sprintf(buf, "disk\n");
+}
+
-+static ctl_table kernel_table2[] = {
-+ { .ctl_name = 99, .procname = "vsyscall64",
-+ .data = &sysctl_vsyscall, .maxlen = sizeof(int), .mode = 0644,
-+ .strategy = vsyscall_sysctl_nostrat,
-+ .proc_handler = vsyscall_sysctl_change },
-+ { 0, }
-+};
-+
-+static ctl_table kernel_root_table2[] = {
-+ { .ctl_name = CTL_KERN, .procname = "kernel", .mode = 0555,
-+ .child = kernel_table2 },
-+ { 0 },
++static struct device_attribute xlvbd_attrs[] = {
++ __ATTR(media, S_IRUGO, show_media, NULL),
+};
+
-+#endif
-+
-+static void __init map_vsyscall(void)
++int xlvbd_sysfs_addif(struct blkfront_info *info)
+{
-+ extern char __vsyscall_0;
-+ unsigned long physaddr_page0 = __pa_symbol(&__vsyscall_0);
++ int i;
++ int error = 0;
+
-+ __set_fixmap(VSYSCALL_FIRST_PAGE, physaddr_page0, PAGE_KERNEL_VSYSCALL);
++ for (i = 0; i < ARRAY_SIZE(xlvbd_attrs); i++) {
++ error = device_create_file(info->gd->driverfs_dev,
++ &xlvbd_attrs[i]);
++ if (error)
++ goto fail;
++ }
++ return 0;
++
++fail:
++ while (--i >= 0)
++ device_remove_file(info->gd->driverfs_dev, &xlvbd_attrs[i]);
++ return error;
+}
+
-+#ifdef CONFIG_XEN
-+static void __init map_vsyscall_user(void)
++void xlvbd_sysfs_delif(struct blkfront_info *info)
+{
-+ extern void __set_fixmap_user(enum fixed_addresses, unsigned long, pgprot_t);
-+ extern char __vsyscall_0;
-+ unsigned long physaddr_page0 = __pa_symbol(&__vsyscall_0);
++ int i;
+
-+ __set_fixmap_user(VSYSCALL_FIRST_PAGE, physaddr_page0, PAGE_KERNEL_VSYSCALL);
++ for (i = 0; i < ARRAY_SIZE(xlvbd_attrs); i++)
++ device_remove_file(info->gd->driverfs_dev, &xlvbd_attrs[i]);
+}
-+#endif
+
-+static int __init vsyscall_init(void)
-+{
-+ BUG_ON(((unsigned long) &vgettimeofday !=
-+ VSYSCALL_ADDR(__NR_vgettimeofday)));
-+ BUG_ON((unsigned long) &vtime != VSYSCALL_ADDR(__NR_vtime));
-+ BUG_ON((VSYSCALL_ADDR(0) != __fix_to_virt(VSYSCALL_FIRST_PAGE)));
-+ map_vsyscall();
-+#ifdef CONFIG_XEN
-+ map_vsyscall_user();
-+ sysctl_vsyscall = 0; /* disable vgettimeofay() */
-+#endif
-+#ifdef CONFIG_SYSCTL
-+ register_sysctl_table(kernel_root_table2, 0);
-+#endif
-+ return 0;
-+}
++#endif /* CONFIG_SYSFS */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blktap/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blktap/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,5 @@
++LINUXINCLUDE += -I../xen/include/public/io
++
++obj-$(CONFIG_XEN_BLKDEV_TAP) := xenblktap.o
++
++xenblktap-y := xenbus.o interface.o blktap.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blktap/blktap.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blktap/blktap.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1681 @@
++/******************************************************************************
++ * drivers/xen/blktap/blktap.c
++ *
++ * Back-end driver for user level virtual block devices. This portion of the
++ * driver exports a 'unified' block-device interface that can be accessed
++ * by any operating system that implements a compatible front end. Requests
++ * are remapped to a user-space memory region.
++ *
++ * Based on the blkback driver code.
++ *
++ * Copyright (c) 2004-2005, Andrew Warfield and Julian Chesterfield
++ *
++ * Clean ups and fix ups:
++ * Copyright (c) 2006, Steven Rostedt - Red Hat, Inc.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/spinlock.h>
++#include <linux/kthread.h>
++#include <linux/list.h>
++#include <asm/hypervisor.h>
++#include "common.h"
++#include <xen/balloon.h>
++#include <xen/driver_util.h>
++#include <linux/kernel.h>
++#include <linux/fs.h>
++#include <linux/mm.h>
++#include <linux/errno.h>
++#include <linux/major.h>
++#include <linux/gfp.h>
++#include <linux/poll.h>
++#include <linux/delay.h>
++#include <asm/tlbflush.h>
++
++#define MAX_TAP_DEV 256 /*the maximum number of tapdisk ring devices */
++#define MAX_DEV_NAME 100 /*the max tapdisk ring device name e.g. blktap0 */
+
-+__initcall(vsyscall_init);
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/kernel/xen_entry.S tmp-linux-2.6-xen.patch/arch/x86_64/kernel/xen_entry.S
---- pristine-linux-2.6.18.2/arch/x86_64/kernel/xen_entry.S 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/kernel/xen_entry.S 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,40 @@
+/*
-+ * Copied from arch/xen/i386/kernel/entry.S
-+ */
-+/* Offsets into shared_info_t. */
-+#define evtchn_upcall_pending /* 0 */
-+#define evtchn_upcall_mask 1
++ * The maximum number of requests that can be outstanding at any time
++ * is determined by
++ *
++ * [mmap_alloc * MAX_PENDING_REQS * BLKIF_MAX_SEGMENTS_PER_REQUEST]
++ *
++ * where mmap_alloc < MAX_DYNAMIC_MEM.
++ *
++ * TODO:
++ * mmap_alloc is initialised to 2 and should be adjustable on the fly via
++ * sysfs.
++ */
++#define BLK_RING_SIZE __RING_SIZE((blkif_sring_t *)0, PAGE_SIZE)
++#define MAX_DYNAMIC_MEM BLK_RING_SIZE
++#define MAX_PENDING_REQS BLK_RING_SIZE
++#define MMAP_PAGES (MAX_PENDING_REQS * BLKIF_MAX_SEGMENTS_PER_REQUEST)
++#define MMAP_VADDR(_start, _req,_seg) \
++ (_start + \
++ ((_req) * BLKIF_MAX_SEGMENTS_PER_REQUEST * PAGE_SIZE) + \
++ ((_seg) * PAGE_SIZE))
++static int blkif_reqs = MAX_PENDING_REQS;
++static int mmap_pages = MMAP_PAGES;
+
-+#define sizeof_vcpu_shift 6
++#define RING_PAGES 1 /* BLKTAP - immediately before the mmap area, we
++ * have a bunch of pages reserved for shared
++ * memory rings.
++ */
+
-+#ifdef CONFIG_SMP
-+//#define preempt_disable(reg) incl threadinfo_preempt_count(reg)
-+//#define preempt_enable(reg) decl threadinfo_preempt_count(reg)
-+#define preempt_disable(reg)
-+#define preempt_enable(reg)
-+#define XEN_GET_VCPU_INFO(reg) preempt_disable(%rbp) ; \
-+ movq %gs:pda_cpunumber,reg ; \
-+ shl $32, reg ; \
-+ shr $32-sizeof_vcpu_shift,reg ; \
-+ addq HYPERVISOR_shared_info,reg
-+#define XEN_PUT_VCPU_INFO(reg) preempt_enable(%rbp) ; \
-+#define XEN_PUT_VCPU_INFO_fixup .byte 0xff,0xff,0xff
-+#else
-+#define XEN_GET_VCPU_INFO(reg) movq HYPERVISOR_shared_info,reg
-+#define XEN_PUT_VCPU_INFO(reg)
-+#define XEN_PUT_VCPU_INFO_fixup
-+#endif
++/*Data struct handed back to userspace for tapdisk device to VBD mapping*/
++typedef struct domid_translate {
++ unsigned short domid;
++ unsigned short busid;
++} domid_translate_t ;
+
-+#define XEN_LOCKED_BLOCK_EVENTS(reg) movb $1,evtchn_upcall_mask(reg)
-+#define XEN_LOCKED_UNBLOCK_EVENTS(reg) movb $0,evtchn_upcall_mask(reg)
-+#define XEN_BLOCK_EVENTS(reg) XEN_GET_VCPU_INFO(reg) ; \
-+ XEN_LOCKED_BLOCK_EVENTS(reg) ; \
-+ XEN_PUT_VCPU_INFO(reg)
-+#define XEN_UNBLOCK_EVENTS(reg) XEN_GET_VCPU_INFO(reg) ; \
-+ XEN_LOCKED_UNBLOCK_EVENTS(reg) ; \
-+ XEN_PUT_VCPU_INFO(reg)
-+#define XEN_TEST_PENDING(reg) testb $0xFF,evtchn_upcall_pending(reg)
++typedef struct domid_translate_ext {
++ unsigned short domid;
++ u32 busid;
++} domid_translate_ext_t ;
+
-+VGCF_IN_SYSCALL = (1<<8)
-+
-+
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/Makefile tmp-linux-2.6-xen.patch/arch/x86_64/Makefile
---- pristine-linux-2.6.18.2/arch/x86_64/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -32,6 +32,10 @@ cflags-$(CONFIG_MK8) += $(call cc-option
- cflags-$(CONFIG_MPSC) += $(call cc-option,-march=nocona)
- cflags-$(CONFIG_GENERIC_CPU) += $(call cc-option,-mtune=generic)
-
-+cppflags-$(CONFIG_XEN) += \
-+ -D__XEN_INTERFACE_VERSION__=$(CONFIG_XEN_INTERFACE_VERSION)
-+CPPFLAGS += $(cppflags-y)
++/*Data struct associated with each of the tapdisk devices*/
++typedef struct tap_blkif {
++ struct vm_area_struct *vma; /*Shared memory area */
++ unsigned long rings_vstart; /*Kernel memory mapping */
++ unsigned long user_vstart; /*User memory mapping */
++ unsigned long dev_inuse; /*One process opens device at a time. */
++ unsigned long dev_pending; /*In process of being opened */
++ unsigned long ring_ok; /*make this ring->state */
++ blkif_front_ring_t ufe_ring; /*Rings up to user space. */
++ wait_queue_head_t wait; /*for poll */
++ unsigned long mode; /*current switching mode */
++ int minor; /*Minor number for tapdisk device */
++ pid_t pid; /*tapdisk process id */
++ enum { RUNNING, CLEANSHUTDOWN } status; /*Detect a clean userspace
++ shutdown */
++ unsigned long *idx_map; /*Record the user ring id to kern
++ [req id, idx] tuple */
++ blkif_t *blkif; /*Associate blkif with tapdev */
++ struct domid_translate_ext trans; /*Translation from domid to bus. */
++} tap_blkif_t;
+
- cflags-y += -m64
- cflags-y += -mno-red-zone
- cflags-y += -mcmodel=kernel
-@@ -74,6 +78,21 @@ boot := arch/x86_64/boot
- PHONY += bzImage bzlilo install archmrproper \
- fdimage fdimage144 fdimage288 isoimage archclean
-
-+ifdef CONFIG_XEN
-+CPPFLAGS := -Iinclude$(if $(KBUILD_SRC),2)/asm/mach-xen $(CPPFLAGS)
-+head-y := arch/x86_64/kernel/head-xen.o arch/x86_64/kernel/head64-xen.o arch/x86_64/kernel/init_task.o
-+LDFLAGS_vmlinux := -e _start
-+boot := arch/i386/boot-xen
-+.PHONY: vmlinuz
-+#Default target when executing "make"
-+all: vmlinuz
++static struct tap_blkif *tapfds[MAX_TAP_DEV];
++static int blktap_next_minor;
+
-+vmlinuz: vmlinux
-+ $(Q)$(MAKE) $(build)=$(boot) $@
++module_param(blkif_reqs, int, 0);
++/* Run-time switchable: /sys/module/blktap/parameters/ */
++static unsigned int log_stats = 0;
++static unsigned int debug_lvl = 0;
++module_param(log_stats, int, 0644);
++module_param(debug_lvl, int, 0644);
+
-+install:
-+ $(Q)$(MAKE) $(build)=$(boot) XENGUEST=$(XENGUEST) $@
-+else
- #Default target when executing "make"
- all: bzImage
-
-@@ -94,6 +113,7 @@ fdimage fdimage144 fdimage288 isoimage:
-
- install:
- $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(BOOTIMAGE) $@
-+endif
-
- archclean:
- $(Q)$(MAKE) $(clean)=$(boot)
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/mm/fault-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/mm/fault-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/mm/fault-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/mm/fault-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,724 @@
+/*
-+ * linux/arch/x86-64/mm/fault.c
-+ *
-+ * Copyright (C) 1995 Linus Torvalds
-+ * Copyright (C) 2001,2002 Andi Kleen, SuSE Labs.
++ * Each outstanding request that we've passed to the lower device layers has a
++ * 'pending_req' allocated to it. Each buffer_head that completes decrements
++ * the pendcnt towards zero. When it hits zero, the specified domain has a
++ * response queued for it, with the saved 'id' passed back.
+ */
++typedef struct {
++ blkif_t *blkif;
++ u64 id;
++ unsigned short mem_idx;
++ int nr_pages;
++ atomic_t pendcnt;
++ unsigned short operation;
++ int status;
++ struct list_head free_list;
++ int inuse;
++} pending_req_t;
+
-+#include <linux/signal.h>
-+#include <linux/sched.h>
-+#include <linux/kernel.h>
-+#include <linux/errno.h>
-+#include <linux/string.h>
-+#include <linux/types.h>
-+#include <linux/ptrace.h>
-+#include <linux/mman.h>
-+#include <linux/mm.h>
-+#include <linux/smp.h>
-+#include <linux/smp_lock.h>
-+#include <linux/interrupt.h>
-+#include <linux/init.h>
-+#include <linux/tty.h>
-+#include <linux/vt_kern.h> /* For unblank_screen() */
-+#include <linux/compiler.h>
-+#include <linux/module.h>
-+#include <linux/kprobes.h>
-+
-+#include <asm/system.h>
-+#include <asm/uaccess.h>
-+#include <asm/pgalloc.h>
-+#include <asm/smp.h>
-+#include <asm/tlbflush.h>
-+#include <asm/proto.h>
-+#include <asm/kdebug.h>
-+#include <asm-generic/sections.h>
-+
-+/* Page fault error code bits */
-+#define PF_PROT (1<<0) /* or no page found */
-+#define PF_WRITE (1<<1)
-+#define PF_USER (1<<2)
-+#define PF_RSVD (1<<3)
-+#define PF_INSTR (1<<4)
++static pending_req_t *pending_reqs[MAX_PENDING_REQS];
++static struct list_head pending_free;
++static DEFINE_SPINLOCK(pending_free_lock);
++static DECLARE_WAIT_QUEUE_HEAD (pending_free_wq);
++static int alloc_pending_reqs;
+
-+#ifdef CONFIG_KPROBES
-+ATOMIC_NOTIFIER_HEAD(notify_page_fault_chain);
++typedef unsigned int PEND_RING_IDX;
+
-+/* Hook to register for page fault notifications */
-+int register_page_fault_notifier(struct notifier_block *nb)
-+{
-+ vmalloc_sync_all();
-+ return atomic_notifier_chain_register(¬ify_page_fault_chain, nb);
++static inline int MASK_PEND_IDX(int i) {
++ return (i & (MAX_PENDING_REQS-1));
+}
+
-+int unregister_page_fault_notifier(struct notifier_block *nb)
-+{
-+ return atomic_notifier_chain_unregister(¬ify_page_fault_chain, nb);
++static inline unsigned int RTN_PEND_IDX(pending_req_t *req, int idx) {
++ return (req - pending_reqs[idx]);
+}
+
-+static inline int notify_page_fault(enum die_val val, const char *str,
-+ struct pt_regs *regs, long err, int trap, int sig)
-+{
-+ struct die_args args = {
-+ .regs = regs,
-+ .str = str,
-+ .err = err,
-+ .trapnr = trap,
-+ .signr = sig
-+ };
-+ return atomic_notifier_call_chain(¬ify_page_fault_chain, val, &args);
-+}
-+#else
-+static inline int notify_page_fault(enum die_val val, const char *str,
-+ struct pt_regs *regs, long err, int trap, int sig)
-+{
-+ return NOTIFY_DONE;
-+}
-+#endif
++#define NR_PENDING_REQS (MAX_PENDING_REQS - pending_prod + pending_cons)
+
-+void bust_spinlocks(int yes)
++#define BLKBACK_INVALID_HANDLE (~0)
++
++static struct page **foreign_pages[MAX_DYNAMIC_MEM];
++static inline unsigned long idx_to_kaddr(
++ unsigned int mmap_idx, unsigned int req_idx, unsigned int sg_idx)
+{
-+ int loglevel_save = console_loglevel;
-+ if (yes) {
-+ oops_in_progress = 1;
-+ } else {
-+#ifdef CONFIG_VT
-+ unblank_screen();
-+#endif
-+ oops_in_progress = 0;
-+ /*
-+ * OK, the message is on the console. Now we call printk()
-+ * without oops_in_progress set so that printk will give klogd
-+ * a poke. Hold onto your hats...
-+ */
-+ console_loglevel = 15; /* NMI oopser may have shut the console up */
-+ printk(" ");
-+ console_loglevel = loglevel_save;
-+ }
++ unsigned int arr_idx = req_idx*BLKIF_MAX_SEGMENTS_PER_REQUEST + sg_idx;
++ unsigned long pfn = page_to_pfn(foreign_pages[mmap_idx][arr_idx]);
++ return (unsigned long)pfn_to_kaddr(pfn);
+}
+
-+/* Sometimes the CPU reports invalid exceptions on prefetch.
-+ Check that here and ignore.
-+ Opcode checker based on code by Richard Brunner */
-+static noinline int is_prefetch(struct pt_regs *regs, unsigned long addr,
-+ unsigned long error_code)
-+{
-+ unsigned char *instr;
-+ int scan_more = 1;
-+ int prefetch = 0;
-+ unsigned char *max_instr;
++static unsigned short mmap_alloc = 0;
++static unsigned short mmap_lock = 0;
++static unsigned short mmap_inuse = 0;
+
-+ /* If it was a exec fault ignore */
-+ if (error_code & PF_INSTR)
-+ return 0;
-+
-+ instr = (unsigned char *)convert_rip_to_linear(current, regs);
-+ max_instr = instr + 15;
++/******************************************************************
++ * GRANT HANDLES
++ */
+
-+ if (user_mode(regs) && instr >= (unsigned char *)TASK_SIZE)
-+ return 0;
++/* When using grant tables to map a frame for device access then the
++ * handle returned must be used to unmap the frame. This is needed to
++ * drop the ref count on the frame.
++ */
++struct grant_handle_pair
++{
++ grant_handle_t kernel;
++ grant_handle_t user;
++};
++#define INVALID_GRANT_HANDLE 0xFFFF
+
-+ while (scan_more && instr < max_instr) {
-+ unsigned char opcode;
-+ unsigned char instr_hi;
-+ unsigned char instr_lo;
++static struct grant_handle_pair
++ pending_grant_handles[MAX_DYNAMIC_MEM][MMAP_PAGES];
++#define pending_handle(_id, _idx, _i) \
++ (pending_grant_handles[_id][((_idx) * BLKIF_MAX_SEGMENTS_PER_REQUEST) \
++ + (_i)])
+
-+ if (__get_user(opcode, instr))
-+ break;
+
-+ instr_hi = opcode & 0xf0;
-+ instr_lo = opcode & 0x0f;
-+ instr++;
++static int blktap_read_ufe_ring(tap_blkif_t *info); /*local prototypes*/
+
-+ switch (instr_hi) {
-+ case 0x20:
-+ case 0x30:
-+ /* Values 0x26,0x2E,0x36,0x3E are valid x86
-+ prefixes. In long mode, the CPU will signal
-+ invalid opcode if some of these prefixes are
-+ present so we will never get here anyway */
-+ scan_more = ((instr_lo & 7) == 0x6);
-+ break;
-+
-+ case 0x40:
-+ /* In AMD64 long mode, 0x40 to 0x4F are valid REX prefixes
-+ Need to figure out under what instruction mode the
-+ instruction was issued ... */
-+ /* Could check the LDT for lm, but for now it's good
-+ enough to assume that long mode only uses well known
-+ segments or kernel. */
-+ scan_more = (!user_mode(regs)) || (regs->cs == __USER_CS);
-+ break;
-+
-+ case 0x60:
-+ /* 0x64 thru 0x67 are valid prefixes in all modes. */
-+ scan_more = (instr_lo & 0xC) == 0x4;
-+ break;
-+ case 0xF0:
-+ /* 0xF0, 0xF2, and 0xF3 are valid prefixes in all modes. */
-+ scan_more = !instr_lo || (instr_lo>>1) == 1;
-+ break;
-+ case 0x00:
-+ /* Prefetch instruction is 0x0F0D or 0x0F18 */
-+ scan_more = 0;
-+ if (__get_user(opcode, instr))
-+ break;
-+ prefetch = (instr_lo == 0xF) &&
-+ (opcode == 0x0D || opcode == 0x18);
-+ break;
-+ default:
-+ scan_more = 0;
-+ break;
-+ }
-+ }
-+ return prefetch;
-+}
++#define BLKTAP_MINOR 0 /*/dev/xen/blktap has a dynamic major */
++#define BLKTAP_DEV_DIR "/dev/xen"
+
-+static int bad_address(void *p)
-+{
-+ unsigned long dummy;
-+ return __get_user(dummy, (unsigned long *)p);
-+}
++static int blktap_major;
+
-+void dump_pagetable(unsigned long address)
-+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte;
++/* blktap IOCTLs: */
++#define BLKTAP_IOCTL_KICK_FE 1
++#define BLKTAP_IOCTL_KICK_BE 2 /* currently unused */
++#define BLKTAP_IOCTL_SETMODE 3
++#define BLKTAP_IOCTL_SENDPID 4
++#define BLKTAP_IOCTL_NEWINTF 5
++#define BLKTAP_IOCTL_MINOR 6
++#define BLKTAP_IOCTL_MAJOR 7
++#define BLKTAP_QUERY_ALLOC_REQS 8
++#define BLKTAP_IOCTL_FREEINTF 9
++#define BLKTAP_IOCTL_NEWINTF_EXT 50
++#define BLKTAP_IOCTL_PRINT_IDXS 100
+
-+ pgd = __va(read_cr3() & PHYSICAL_PAGE_MASK);
-+ pgd += pgd_index(address);
-+ if (bad_address(pgd)) goto bad;
-+ printk("PGD %lx ", pgd_val(*pgd));
-+ if (!pgd_present(*pgd)) goto ret;
++/* blktap switching modes: (Set with BLKTAP_IOCTL_SETMODE) */
++#define BLKTAP_MODE_PASSTHROUGH 0x00000000 /* default */
++#define BLKTAP_MODE_INTERCEPT_FE 0x00000001
++#define BLKTAP_MODE_INTERCEPT_BE 0x00000002 /* unimp. */
+
-+ pud = pud_offset(pgd, address);
-+ if (bad_address(pud)) goto bad;
-+ printk("PUD %lx ", pud_val(*pud));
-+ if (!pud_present(*pud)) goto ret;
++#define BLKTAP_MODE_INTERPOSE \
++ (BLKTAP_MODE_INTERCEPT_FE | BLKTAP_MODE_INTERCEPT_BE)
+
-+ pmd = pmd_offset(pud, address);
-+ if (bad_address(pmd)) goto bad;
-+ printk("PMD %lx ", pmd_val(*pmd));
-+ if (!pmd_present(*pmd)) goto ret;
+
-+ pte = pte_offset_kernel(pmd, address);
-+ if (bad_address(pte)) goto bad;
-+ printk("PTE %lx", pte_val(*pte));
-+ret:
-+ printk("\n");
-+ return;
-+bad:
-+ printk("BAD\n");
++static inline int BLKTAP_MODE_VALID(unsigned long arg)
++{
++ return ((arg == BLKTAP_MODE_PASSTHROUGH ) ||
++ (arg == BLKTAP_MODE_INTERCEPT_FE) ||
++ (arg == BLKTAP_MODE_INTERPOSE ));
+}
+
-+static const char errata93_warning[] =
-+KERN_ERR "******* Your BIOS seems to not contain a fix for K8 errata #93\n"
-+KERN_ERR "******* Working around it, but it may cause SEGVs or burn power.\n"
-+KERN_ERR "******* Please consider a BIOS update.\n"
-+KERN_ERR "******* Disabling USB legacy in the BIOS may also help.\n";
++/* Requests passing through the tap to userspace are re-assigned an ID.
++ * We must record a mapping between the BE [IDX,ID] tuple and the userspace
++ * ring ID.
++ */
+
-+/* Workaround for K8 erratum #93 & buggy BIOS.
-+ BIOS SMM functions are required to use a specific workaround
-+ to avoid corruption of the 64bit RIP register on C stepping K8.
-+ A lot of BIOS that didn't get tested properly miss this.
-+ The OS sees this as a page fault with the upper 32bits of RIP cleared.
-+ Try to work around it here.
-+ Note we only handle faults in kernel here. */
++static inline unsigned long MAKE_ID(domid_t fe_dom, PEND_RING_IDX idx)
++{
++ return ((fe_dom << 16) | MASK_PEND_IDX(idx));
++}
+
-+static int is_errata93(struct pt_regs *regs, unsigned long address)
++extern inline PEND_RING_IDX ID_TO_IDX(unsigned long id)
+{
-+ static int warned;
-+ if (address != regs->rip)
-+ return 0;
-+ if ((address >> 32) != 0)
-+ return 0;
-+ address |= 0xffffffffUL << 32;
-+ if ((address >= (u64)_stext && address <= (u64)_etext) ||
-+ (address >= MODULES_VADDR && address <= MODULES_END)) {
-+ if (!warned) {
-+ printk(errata93_warning);
-+ warned = 1;
-+ }
-+ regs->rip = address;
-+ return 1;
-+ }
-+ return 0;
-+}
++ return (PEND_RING_IDX)(id & 0x0000ffff);
++}
++
++extern inline int ID_TO_MIDX(unsigned long id)
++{
++ return (int)(id >> 16);
++}
+
-+int unhandled_signal(struct task_struct *tsk, int sig)
++#define INVALID_REQ 0xdead0000
++
++/*TODO: Convert to a free list*/
++static inline int GET_NEXT_REQ(unsigned long *idx_map)
+{
-+ if (tsk->pid == 1)
-+ return 1;
-+ if (tsk->ptrace & PT_PTRACED)
-+ return 0;
-+ return (tsk->sighand->action[sig-1].sa.sa_handler == SIG_IGN) ||
-+ (tsk->sighand->action[sig-1].sa.sa_handler == SIG_DFL);
++ int i;
++ for (i = 0; i < MAX_PENDING_REQS; i++)
++ if (idx_map[i] == INVALID_REQ)
++ return i;
++
++ return INVALID_REQ;
+}
+
-+static noinline void pgtable_bad(unsigned long address, struct pt_regs *regs,
-+ unsigned long error_code)
++static inline int OFFSET_TO_USR_IDX(int offset)
+{
-+ unsigned long flags = oops_begin();
-+ struct task_struct *tsk;
++ return offset / BLKIF_MAX_SEGMENTS_PER_REQUEST;
++}
+
-+ printk(KERN_ALERT "%s: Corrupted page table at address %lx\n",
-+ current->comm, address);
-+ dump_pagetable(address);
-+ tsk = current;
-+ tsk->thread.cr2 = address;
-+ tsk->thread.trap_no = 14;
-+ tsk->thread.error_code = error_code;
-+ __die("Bad pagetable", regs, error_code);
-+ oops_end(flags);
-+ do_exit(SIGKILL);
++static inline int OFFSET_TO_SEG(int offset)
++{
++ return offset % BLKIF_MAX_SEGMENTS_PER_REQUEST;
+}
+
-+/*
-+ * Handle a fault on the vmalloc area
-+ *
-+ * This assumes no large pages in there.
++
++#define BLKTAP_INVALID_HANDLE(_g) \
++ (((_g->kernel) == INVALID_GRANT_HANDLE) && \
++ ((_g->user) == INVALID_GRANT_HANDLE))
++
++#define BLKTAP_INVALIDATE_HANDLE(_g) do { \
++ (_g)->kernel = INVALID_GRANT_HANDLE; (_g)->user = INVALID_GRANT_HANDLE; \
++ } while(0)
++
++
++/******************************************************************
++ * BLKTAP VM OPS
+ */
-+static int vmalloc_fault(unsigned long address)
++
++static struct page *blktap_nopage(struct vm_area_struct *vma,
++ unsigned long address,
++ int *type)
+{
-+ pgd_t *pgd, *pgd_ref;
-+ pud_t *pud, *pud_ref;
-+ pmd_t *pmd, *pmd_ref;
-+ pte_t *pte, *pte_ref;
++ /*
++ * if the page has not been mapped in by the driver then return
++ * NOPAGE_SIGBUS to the domain.
++ */
+
-+ /* Copy kernel mappings over when needed. This can also
-+ happen within a race in page table update. In the later
-+ case just flush. */
++ return NOPAGE_SIGBUS;
++}
+
-+ /* On Xen the line below does not always work. Needs investigating! */
-+ /*pgd = pgd_offset(current->mm ?: &init_mm, address);*/
-+ pgd = __va(read_cr3() & PHYSICAL_PAGE_MASK);
-+ pgd += pgd_index(address);
-+ pgd_ref = pgd_offset_k(address);
-+ if (pgd_none(*pgd_ref))
-+ return -1;
-+ if (pgd_none(*pgd))
-+ set_pgd(pgd, *pgd_ref);
-+ else
-+ BUG_ON(pgd_page(*pgd) != pgd_page(*pgd_ref));
++static pte_t blktap_clear_pte(struct vm_area_struct *vma,
++ unsigned long uvaddr,
++ pte_t *ptep, int is_fullmm)
++{
++ pte_t copy;
++ tap_blkif_t *info;
++ int offset, seg, usr_idx, pending_idx, mmap_idx;
++ unsigned long uvstart = vma->vm_start + (RING_PAGES << PAGE_SHIFT);
++ unsigned long kvaddr;
++ struct page **map;
++ struct page *pg;
++ struct grant_handle_pair *khandle;
++ struct gnttab_unmap_grant_ref unmap[2];
++ int count = 0;
+
-+ /* Below here mismatches are bugs because these lower tables
-+ are shared */
++ /*
++ * If the address is before the start of the grant mapped region or
++ * if vm_file is NULL (meaning mmap failed and we have nothing to do)
++ */
++ if (uvaddr < uvstart || vma->vm_file == NULL)
++ return ptep_get_and_clear_full(vma->vm_mm, uvaddr,
++ ptep, is_fullmm);
+
-+ pud = pud_offset(pgd, address);
-+ pud_ref = pud_offset(pgd_ref, address);
-+ if (pud_none(*pud_ref))
-+ return -1;
-+ if (pud_none(*pud) || pud_page(*pud) != pud_page(*pud_ref))
-+ BUG();
-+ pmd = pmd_offset(pud, address);
-+ pmd_ref = pmd_offset(pud_ref, address);
-+ if (pmd_none(*pmd_ref))
-+ return -1;
-+ if (pmd_none(*pmd) || pmd_page(*pmd) != pmd_page(*pmd_ref))
-+ BUG();
-+ pte_ref = pte_offset_kernel(pmd_ref, address);
-+ if (!pte_present(*pte_ref))
-+ return -1;
-+ pte = pte_offset_kernel(pmd, address);
-+ /* Don't use pte_page here, because the mappings can point
-+ outside mem_map, and the NUMA hash lookup cannot handle
-+ that. */
-+ if (!pte_present(*pte) || pte_pfn(*pte) != pte_pfn(*pte_ref))
-+ BUG();
-+ return 0;
-+}
++ info = vma->vm_file->private_data;
++ map = vma->vm_private_data;
+
-+int page_fault_trace = 0;
-+int exception_trace = 1;
++ /* TODO Should these be changed to if statements? */
++ BUG_ON(!info);
++ BUG_ON(!info->idx_map);
++ BUG_ON(!map);
+
++ offset = (int) ((uvaddr - uvstart) >> PAGE_SHIFT);
++ usr_idx = OFFSET_TO_USR_IDX(offset);
++ seg = OFFSET_TO_SEG(offset);
+
-+#define MEM_VERBOSE 1
++ pending_idx = MASK_PEND_IDX(ID_TO_IDX(info->idx_map[usr_idx]));
++ mmap_idx = ID_TO_MIDX(info->idx_map[usr_idx]);
+
-+#ifdef MEM_VERBOSE
-+#define MEM_LOG(_f, _a...) \
-+ printk("fault.c:[%d]-> " _f "\n", \
-+ __LINE__ , ## _a )
-+#else
-+#define MEM_LOG(_f, _a...) ((void)0)
-+#endif
++ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, seg);
++ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
++ ClearPageReserved(pg);
++ map[offset + RING_PAGES] = NULL;
+
-+static int spurious_fault(struct pt_regs *regs,
-+ unsigned long address,
-+ unsigned long error_code)
-+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte;
++ khandle = &pending_handle(mmap_idx, pending_idx, seg);
+
-+#ifdef CONFIG_XEN
-+ /* Faults in hypervisor area are never spurious. */
-+ if ((address >= HYPERVISOR_VIRT_START) &&
-+ (address < HYPERVISOR_VIRT_END))
-+ return 0;
-+#endif
++ if (khandle->kernel != INVALID_GRANT_HANDLE) {
++ gnttab_set_unmap_op(&unmap[count], kvaddr,
++ GNTMAP_host_map, khandle->kernel);
++ count++;
+
-+ /* Reserved-bit violation or user access to kernel space? */
-+ if (error_code & (PF_RSVD|PF_USER))
-+ return 0;
++ set_phys_to_machine(__pa(kvaddr) >> PAGE_SHIFT,
++ INVALID_P2M_ENTRY);
++ }
+
-+ pgd = init_mm.pgd + pgd_index(address);
-+ if (!pgd_present(*pgd))
-+ return 0;
++ if (khandle->user != INVALID_GRANT_HANDLE) {
++ BUG_ON(xen_feature(XENFEAT_auto_translated_physmap));
+
-+ pud = pud_offset(pgd, address);
-+ if (!pud_present(*pud))
-+ return 0;
++ copy = *ptep;
++ gnttab_set_unmap_op(&unmap[count], virt_to_machine(ptep),
++ GNTMAP_host_map
++ | GNTMAP_application_map
++ | GNTMAP_contains_pte,
++ khandle->user);
++ count++;
++ } else {
++ BUG_ON(!xen_feature(XENFEAT_auto_translated_physmap));
+
-+ pmd = pmd_offset(pud, address);
-+ if (!pmd_present(*pmd))
-+ return 0;
++ /* USING SHADOW PAGE TABLES. */
++ copy = ptep_get_and_clear_full(vma->vm_mm, uvaddr, ptep,
++ is_fullmm);
++ }
+
-+ pte = pte_offset_kernel(pmd, address);
-+ if (!pte_present(*pte))
-+ return 0;
-+ if ((error_code & PF_WRITE) && !pte_write(*pte))
-+ return 0;
-+ if ((error_code & PF_INSTR) && (pte_val(*pte) & _PAGE_NX))
-+ return 0;
++ if (count) {
++ BLKTAP_INVALIDATE_HANDLE(khandle);
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref,
++ unmap, count))
++ BUG();
++ }
+
-+ return 1;
++ return copy;
+}
+
-+/*
-+ * This routine handles page faults. It determines the address,
-+ * and the problem, and then passes it off to one of the appropriate
-+ * routines.
++struct vm_operations_struct blktap_vm_ops = {
++ nopage: blktap_nopage,
++ zap_pte: blktap_clear_pte,
++};
++
++/******************************************************************
++ * BLKTAP FILE OPS
+ */
-+asmlinkage void __kprobes do_page_fault(struct pt_regs *regs,
-+ unsigned long error_code)
-+{
-+ struct task_struct *tsk;
-+ struct mm_struct *mm;
-+ struct vm_area_struct * vma;
-+ unsigned long address;
-+ const struct exception_table_entry *fixup;
-+ int write;
-+ unsigned long flags;
-+ siginfo_t info;
++
++/*Function Declarations*/
++static tap_blkif_t *get_next_free_dev(void);
++static int blktap_open(struct inode *inode, struct file *filp);
++static int blktap_release(struct inode *inode, struct file *filp);
++static int blktap_mmap(struct file *filp, struct vm_area_struct *vma);
++static int blktap_ioctl(struct inode *inode, struct file *filp,
++ unsigned int cmd, unsigned long arg);
++static unsigned int blktap_poll(struct file *file, poll_table *wait);
+
-+ if (!user_mode(regs))
-+ error_code &= ~PF_USER; /* means kernel */
++static const struct file_operations blktap_fops = {
++ .owner = THIS_MODULE,
++ .poll = blktap_poll,
++ .ioctl = blktap_ioctl,
++ .open = blktap_open,
++ .release = blktap_release,
++ .mmap = blktap_mmap,
++};
+
-+ tsk = current;
-+ mm = tsk->mm;
-+ prefetchw(&mm->mmap_sem);
+
-+ /* get the address */
-+ address = current_vcpu_info()->arch.cr2;
++static tap_blkif_t *get_next_free_dev(void)
++{
++ struct class *class;
++ tap_blkif_t *info;
++ int minor;
+
-+ info.si_code = SEGV_MAPERR;
++ /*
++ * This is called only from the ioctl, which
++ * means we should always have interrupts enabled.
++ */
++ BUG_ON(irqs_disabled());
++
++ spin_lock_irq(&pending_free_lock);
++
++ /* tapfds[0] is always NULL */
+
++ for (minor = 1; minor < blktap_next_minor; minor++) {
++ info = tapfds[minor];
++ /* we could have failed a previous attempt. */
++ if (!info ||
++ ((info->dev_inuse == 0) &&
++ (info->dev_pending == 0)) ) {
++ info->dev_pending = 1;
++ goto found;
++ }
++ }
++ info = NULL;
++ minor = -1;
+
+ /*
-+ * We fault-in kernel-space virtual memory on-demand. The
-+ * 'reference' page table is init_mm.pgd.
-+ *
-+ * NOTE! We MUST NOT take any locks for this case. We may
-+ * be in an interrupt or a critical region, and should
-+ * only copy the information from the master page table,
-+ * nothing more.
-+ *
-+ * This verifies that the fault happens in kernel space
-+ * (error_code & 4) == 0, and that the fault was not a
-+ * protection error (error_code & 9) == 0.
++ * We didn't find free device. If we can still allocate
++ * more, then we grab the next device minor that is
++ * available. This is done while we are still under
++ * the protection of the pending_free_lock.
+ */
-+ if (unlikely(address >= TASK_SIZE64)) {
-+ /*
-+ * Don't check for the module range here: its PML4
-+ * is always initialized because it's shared with the main
-+ * kernel text. Only vmalloc may need PML4 syncups.
-+ */
-+ if (!(error_code & (PF_RSVD|PF_USER|PF_PROT)) &&
-+ ((address >= VMALLOC_START && address < VMALLOC_END))) {
-+ if (vmalloc_fault(address) >= 0)
-+ return;
++ if (blktap_next_minor < MAX_TAP_DEV)
++ minor = blktap_next_minor++;
++found:
++ spin_unlock_irq(&pending_free_lock);
++
++ if (!info && minor > 0) {
++ info = kzalloc(sizeof(*info), GFP_KERNEL);
++ if (unlikely(!info)) {
++ /*
++ * If we failed here, try to put back
++ * the next minor number. But if one
++ * was just taken, then we just lose this
++ * minor. We can try to allocate this
++ * minor again later.
++ */
++ spin_lock_irq(&pending_free_lock);
++ if (blktap_next_minor == minor+1)
++ blktap_next_minor--;
++ spin_unlock_irq(&pending_free_lock);
++ goto out;
+ }
-+ /* Can take a spurious fault if mapping changes R/O -> R/W. */
-+ if (spurious_fault(regs, address, error_code))
-+ return;
-+ if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
-+ SIGSEGV) == NOTIFY_STOP)
-+ return;
++
++ info->minor = minor;
+ /*
-+ * Don't take the mm semaphore here. If we fixup a prefetch
-+ * fault we could otherwise deadlock.
++ * Make sure that we have a minor before others can
++ * see us.
+ */
-+ goto bad_area_nosemaphore;
++ wmb();
++ tapfds[minor] = info;
++
++ if ((class = get_xen_class()) != NULL)
++ class_device_create(class, NULL,
++ MKDEV(blktap_major, minor), NULL,
++ "blktap%d", minor);
+ }
+
-+ if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
-+ SIGSEGV) == NOTIFY_STOP)
-+ return;
++out:
++ return info;
++}
+
-+ if (likely(regs->eflags & X86_EFLAGS_IF))
-+ local_irq_enable();
++int dom_to_devid(domid_t domid, int xenbus_id, blkif_t *blkif)
++{
++ tap_blkif_t *info;
++ int i;
+
-+ if (unlikely(page_fault_trace))
-+ printk("pagefault rip:%lx rsp:%lx cs:%lu ss:%lu address %lx error %lx\n",
-+ regs->rip,regs->rsp,regs->cs,regs->ss,address,error_code);
++ for (i = 1; i < blktap_next_minor; i++) {
++ info = tapfds[i];
++ if ( info &&
++ (info->trans.domid == domid) &&
++ (info->trans.busid == xenbus_id) ) {
++ info->blkif = blkif;
++ info->status = RUNNING;
++ return i;
++ }
++ }
++ return -1;
++}
+
-+ if (unlikely(error_code & PF_RSVD))
-+ pgtable_bad(address, regs, error_code);
++void signal_tapdisk(int idx)
++{
++ tap_blkif_t *info;
++ struct task_struct *ptask;
+
+ /*
-+ * If we're in an interrupt or have no user
-+ * context, we must not take the fault..
++ * if the userland tools set things up wrong, this could be negative;
++ * just don't try to signal in this case
+ */
-+ if (unlikely(in_atomic() || !mm))
-+ goto bad_area_nosemaphore;
++ if (idx < 0)
++ return;
+
-+ again:
-+ /* When running in the kernel we expect faults to occur only to
-+ * addresses in user space. All other faults represent errors in the
-+ * kernel and should generate an OOPS. Unfortunatly, in the case of an
-+ * erroneous fault occurring in a code path which already holds mmap_sem
-+ * we will deadlock attempting to validate the fault against the
-+ * address space. Luckily the kernel only validly references user
-+ * space from well defined areas of code, which are listed in the
-+ * exceptions table.
-+ *
-+ * As the vast majority of faults will be valid we will only perform
-+ * the source reference check when there is a possibilty of a deadlock.
-+ * Attempt to lock the address space, if we cannot we then validate the
-+ * source. If this is invalid we can skip the address space check,
-+ * thus avoiding the deadlock.
-+ */
-+ if (!down_read_trylock(&mm->mmap_sem)) {
-+ if ((error_code & PF_USER) == 0 &&
-+ !search_exception_tables(regs->rip))
-+ goto bad_area_nosemaphore;
-+ down_read(&mm->mmap_sem);
-+ }
++ info = tapfds[idx];
++ if ((idx < 0) || (idx > MAX_TAP_DEV) || !info)
++ return;
+
-+ vma = find_vma(mm, address);
-+ if (!vma)
-+ goto bad_area;
-+ if (likely(vma->vm_start <= address))
-+ goto good_area;
-+ if (!(vma->vm_flags & VM_GROWSDOWN))
-+ goto bad_area;
-+ if (error_code & 4) {
-+ /* Allow userspace just enough access below the stack pointer
-+ * to let the 'enter' instruction work.
-+ */
-+ if (address + 65536 + 32 * sizeof(unsigned long) < regs->rsp)
-+ goto bad_area;
++ if (info->pid > 0) {
++ ptask = find_task_by_pid(info->pid);
++ if (ptask)
++ info->status = CLEANSHUTDOWN;
+ }
-+ if (expand_stack(vma, address))
-+ goto bad_area;
-+/*
-+ * Ok, we have a good vm_area for this memory access, so
-+ * we can handle it..
-+ */
-+good_area:
-+ info.si_code = SEGV_ACCERR;
-+ write = 0;
-+ switch (error_code & (PF_PROT|PF_WRITE)) {
-+ default: /* 3: write, present */
-+ /* fall through */
-+ case PF_WRITE: /* write, not present */
-+ if (!(vma->vm_flags & VM_WRITE))
-+ goto bad_area;
-+ write++;
-+ break;
-+ case PF_PROT: /* read, present */
-+ goto bad_area;
-+ case 0: /* read, not present */
-+ if (!(vma->vm_flags & (VM_READ | VM_EXEC)))
-+ goto bad_area;
++ info->blkif = NULL;
++
++ return;
++}
++
++static int blktap_open(struct inode *inode, struct file *filp)
++{
++ blkif_sring_t *sring;
++ int idx = iminor(inode) - BLKTAP_MINOR;
++ tap_blkif_t *info;
++ int i;
++
++ /* ctrl device, treat differently */
++ if (!idx)
++ return 0;
++
++ info = tapfds[idx];
++
++ if ((idx < 0) || (idx > MAX_TAP_DEV) || !info) {
++ WPRINTK("Unable to open device /dev/xen/blktap%d\n",
++ idx);
++ return -ENODEV;
+ }
+
-+ /*
-+ * If for any reason at all we couldn't handle the fault,
-+ * make sure we exit gracefully rather than endlessly redo
-+ * the fault.
-+ */
-+ switch (handle_mm_fault(mm, vma, address, write)) {
-+ case VM_FAULT_MINOR:
-+ tsk->min_flt++;
-+ break;
-+ case VM_FAULT_MAJOR:
-+ tsk->maj_flt++;
-+ break;
-+ case VM_FAULT_SIGBUS:
-+ goto do_sigbus;
-+ default:
-+ goto out_of_memory;
++ DPRINTK("Opening device /dev/xen/blktap%d\n",idx);
++
++ /*Only one process can access device at a time*/
++ if (test_and_set_bit(0, &info->dev_inuse))
++ return -EBUSY;
++
++ info->dev_pending = 0;
++
++ /* Allocate the fe ring. */
++ sring = (blkif_sring_t *)get_zeroed_page(GFP_KERNEL);
++ if (sring == NULL)
++ goto fail_nomem;
++
++ SetPageReserved(virt_to_page(sring));
++
++ SHARED_RING_INIT(sring);
++ FRONT_RING_INIT(&info->ufe_ring, sring, PAGE_SIZE);
++
++ filp->private_data = info;
++ info->vma = NULL;
++
++ info->idx_map = kmalloc(sizeof(unsigned long) * MAX_PENDING_REQS,
++ GFP_KERNEL);
++
++ if (info->idx_map == NULL)
++ goto fail_nomem;
++
++ if (idx > 0) {
++ init_waitqueue_head(&info->wait);
++ for (i = 0; i < MAX_PENDING_REQS; i++)
++ info->idx_map[i] = INVALID_REQ;
+ }
+
-+ up_read(&mm->mmap_sem);
-+ return;
++ DPRINTK("Tap open: device /dev/xen/blktap%d\n",idx);
++ return 0;
+
-+/*
-+ * Something tried to access memory that isn't in our memory map..
-+ * Fix it, but check if it's kernel or user first..
-+ */
-+bad_area:
-+ up_read(&mm->mmap_sem);
++ fail_nomem:
++ return -ENOMEM;
++}
+
-+bad_area_nosemaphore:
-+ /* User mode accesses just cause a SIGSEGV */
-+ if (error_code & PF_USER) {
-+ if (is_prefetch(regs, address, error_code))
-+ return;
++static int blktap_release(struct inode *inode, struct file *filp)
++{
++ tap_blkif_t *info = filp->private_data;
++
++ /* check for control device */
++ if (!info)
++ return 0;
+
-+ /* Work around K8 erratum #100 K8 in compat mode
-+ occasionally jumps to illegal addresses >4GB. We
-+ catch this here in the page fault handler because
-+ these addresses are not reachable. Just detect this
-+ case and return. Any code segment in LDT is
-+ compatibility mode. */
-+ if ((regs->cs == __USER32_CS || (regs->cs & (1<<2))) &&
-+ (address >> 32))
-+ return;
++ info->dev_inuse = 0;
++ DPRINTK("Freeing device [/dev/xen/blktap%d]\n",info->minor);
+
-+ if (exception_trace && unhandled_signal(tsk, SIGSEGV)) {
-+ printk(
-+ "%s%s[%d]: segfault at %016lx rip %016lx rsp %016lx error %lx\n",
-+ tsk->pid > 1 ? KERN_INFO : KERN_EMERG,
-+ tsk->comm, tsk->pid, address, regs->rip,
-+ regs->rsp, error_code);
-+ }
-+
-+ tsk->thread.cr2 = address;
-+ /* Kernel addresses are always protection faults */
-+ tsk->thread.error_code = error_code | (address >= TASK_SIZE);
-+ tsk->thread.trap_no = 14;
-+ info.si_signo = SIGSEGV;
-+ info.si_errno = 0;
-+ /* info.si_code has been set above */
-+ info.si_addr = (void __user *)address;
-+ force_sig_info(SIGSEGV, &info, tsk);
-+ return;
++ /* Free the ring page. */
++ ClearPageReserved(virt_to_page(info->ufe_ring.sring));
++ free_page((unsigned long) info->ufe_ring.sring);
++
++ /* Clear any active mappings and free foreign map table */
++ if (info->vma) {
++ zap_page_range(
++ info->vma, info->vma->vm_start,
++ info->vma->vm_end - info->vma->vm_start, NULL);
++
++ kfree(info->vma->vm_private_data);
++
++ info->vma = NULL;
+ }
+
-+no_context:
-+
-+ /* Are we prepared to handle this kernel fault? */
-+ fixup = search_exception_tables(regs->rip);
-+ if (fixup) {
-+ regs->rip = fixup->fixup;
-+ return;
++ if (info->idx_map) {
++ kfree(info->idx_map);
++ info->idx_map = NULL;
+ }
+
-+ /*
-+ * Hall of shame of CPU/BIOS bugs.
-+ */
++ if ( (info->status != CLEANSHUTDOWN) && (info->blkif != NULL) ) {
++ if (info->blkif->xenblkd != NULL) {
++ kthread_stop(info->blkif->xenblkd);
++ info->blkif->xenblkd = NULL;
++ }
++ info->status = CLEANSHUTDOWN;
++ }
+
-+ if (is_prefetch(regs, address, error_code))
-+ return;
++ return 0;
++}
+
-+ if (is_errata93(regs, address))
-+ return;
+
-+/*
-+ * Oops. The kernel tried to access some bad page. We'll have to
-+ * terminate things with extreme prejudice.
++/* Note on mmap:
++ * We need to map pages to user space in a way that will allow the block
++ * subsystem set up direct IO to them. This couldn't be done before, because
++ * there isn't really a sane way to translate a user virtual address down to a
++ * physical address when the page belongs to another domain.
++ *
++ * My first approach was to map the page in to kernel memory, add an entry
++ * for it in the physical frame list (using alloc_lomem_region as in blkback)
++ * and then attempt to map that page up to user space. This is disallowed
++ * by xen though, which realizes that we don't really own the machine frame
++ * underlying the physical page.
++ *
++ * The new approach is to provide explicit support for this in xen linux.
++ * The VMA now has a flag, VM_FOREIGN, to indicate that it contains pages
++ * mapped from other vms. vma->vm_private_data is set up as a mapping
++ * from pages to actual page structs. There is a new clause in get_user_pages
++ * that does the right thing for this sort of mapping.
+ */
++static int blktap_mmap(struct file *filp, struct vm_area_struct *vma)
++{
++ int size;
++ struct page **map;
++ int i;
++ tap_blkif_t *info = filp->private_data;
++ int ret;
+
-+ flags = oops_begin();
++ if (info == NULL) {
++ WPRINTK("blktap: mmap, retrieving idx failed\n");
++ return -ENOMEM;
++ }
++
++ vma->vm_flags |= VM_RESERVED;
++ vma->vm_ops = &blktap_vm_ops;
+
-+ if (address < PAGE_SIZE)
-+ printk(KERN_ALERT "Unable to handle kernel NULL pointer dereference");
++ size = vma->vm_end - vma->vm_start;
++ if (size != ((mmap_pages + RING_PAGES) << PAGE_SHIFT)) {
++ WPRINTK("you _must_ map exactly %d pages!\n",
++ mmap_pages + RING_PAGES);
++ return -EAGAIN;
++ }
++
++ size >>= PAGE_SHIFT;
++ info->rings_vstart = vma->vm_start;
++ info->user_vstart = info->rings_vstart + (RING_PAGES << PAGE_SHIFT);
++
++ /* Map the ring pages to the start of the region and reserve it. */
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ ret = vm_insert_page(vma, vma->vm_start,
++ virt_to_page(info->ufe_ring.sring));
+ else
-+ printk(KERN_ALERT "Unable to handle kernel paging request");
-+ printk(" at %016lx RIP: \n" KERN_ALERT,address);
-+ printk_address(regs->rip);
-+ dump_pagetable(address);
-+ tsk->thread.cr2 = address;
-+ tsk->thread.trap_no = 14;
-+ tsk->thread.error_code = error_code;
-+ __die("Oops", regs, error_code);
-+ /* Executive summary in case the body of the oops scrolled away */
-+ printk(KERN_EMERG "CR2: %016lx\n", address);
-+ oops_end(flags);
-+ do_exit(SIGKILL);
++ ret = remap_pfn_range(vma, vma->vm_start,
++ __pa(info->ufe_ring.sring) >> PAGE_SHIFT,
++ PAGE_SIZE, vma->vm_page_prot);
++ if (ret) {
++ WPRINTK("Mapping user ring failed!\n");
++ goto fail;
++ }
+
-+/*
-+ * We ran out of memory, or some other thing happened to us that made
-+ * us unable to handle the page fault gracefully.
-+ */
-+out_of_memory:
-+ up_read(&mm->mmap_sem);
-+ if (current->pid == 1) {
-+ yield();
-+ goto again;
++ /* Mark this VM as containing foreign pages, and set up mappings. */
++ map = kzalloc(((vma->vm_end - vma->vm_start) >> PAGE_SHIFT)
++ * sizeof(struct page *),
++ GFP_KERNEL);
++ if (map == NULL) {
++ WPRINTK("Couldn't alloc VM_FOREIGN map.\n");
++ goto fail;
+ }
-+ printk("VM: killing process %s\n", tsk->comm);
-+ if (error_code & 4)
-+ do_exit(SIGKILL);
-+ goto no_context;
+
-+do_sigbus:
-+ up_read(&mm->mmap_sem);
++ for (i = 0; i < ((vma->vm_end - vma->vm_start) >> PAGE_SHIFT); i++)
++ map[i] = NULL;
++
++ vma->vm_private_data = map;
++ vma->vm_flags |= VM_FOREIGN;
++ vma->vm_flags |= VM_DONTCOPY;
+
-+ /* Kernel mode? Handle exceptions or die */
-+ if (!(error_code & PF_USER))
-+ goto no_context;
++#ifdef CONFIG_X86
++ vma->vm_mm->context.has_foreign_mappings = 1;
++#endif
+
-+ tsk->thread.cr2 = address;
-+ tsk->thread.error_code = error_code;
-+ tsk->thread.trap_no = 14;
-+ info.si_signo = SIGBUS;
-+ info.si_errno = 0;
-+ info.si_code = BUS_ADRERR;
-+ info.si_addr = (void __user *)address;
-+ force_sig_info(SIGBUS, &info, tsk);
-+ return;
++ info->vma = vma;
++ info->ring_ok = 1;
++ return 0;
++ fail:
++ /* Clear any active mappings. */
++ zap_page_range(vma, vma->vm_start,
++ vma->vm_end - vma->vm_start, NULL);
++
++ return -ENOMEM;
+}
+
-+DEFINE_SPINLOCK(pgd_lock);
-+struct page *pgd_list;
+
-+void vmalloc_sync_all(void)
++static int blktap_ioctl(struct inode *inode, struct file *filp,
++ unsigned int cmd, unsigned long arg)
+{
-+ /* Note that races in the updates of insync and start aren't
-+ problematic:
-+ insync can only get set bits added, and updates to start are only
-+ improving performance (without affecting correctness if undone). */
-+ static DECLARE_BITMAP(insync, PTRS_PER_PGD);
-+ static unsigned long start = VMALLOC_START & PGDIR_MASK;
-+ unsigned long address;
-+
-+ for (address = start; address <= VMALLOC_END; address += PGDIR_SIZE) {
-+ if (!test_bit(pgd_index(address), insync)) {
-+ const pgd_t *pgd_ref = pgd_offset_k(address);
-+ struct page *page;
++ tap_blkif_t *info = filp->private_data;
+
-+ if (pgd_none(*pgd_ref))
-+ continue;
-+ spin_lock(&pgd_lock);
-+ for (page = pgd_list; page;
-+ page = (struct page *)page->index) {
-+ pgd_t *pgd;
-+ pgd = (pgd_t *)page_address(page) + pgd_index(address);
-+ if (pgd_none(*pgd))
-+ set_pgd(pgd, *pgd_ref);
-+ else
-+ BUG_ON(pgd_page(*pgd) != pgd_page(*pgd_ref));
++ switch(cmd) {
++ case BLKTAP_IOCTL_KICK_FE:
++ {
++ /* There are fe messages to process. */
++ return blktap_read_ufe_ring(info);
++ }
++ case BLKTAP_IOCTL_SETMODE:
++ {
++ if (info) {
++ if (BLKTAP_MODE_VALID(arg)) {
++ info->mode = arg;
++ /* XXX: may need to flush rings here. */
++ DPRINTK("blktap: set mode to %lx\n",
++ arg);
++ return 0;
+ }
-+ spin_unlock(&pgd_lock);
-+ set_bit(pgd_index(address), insync);
+ }
-+ if (address == start)
-+ start = address + PGDIR_SIZE;
++ return 0;
+ }
-+ /* Check that there is no need to do the same for the modules area. */
-+ BUILD_BUG_ON(!(MODULES_VADDR > __START_KERNEL));
-+ BUILD_BUG_ON(!(((MODULES_END - 1) & PGDIR_MASK) ==
-+ (__START_KERNEL & PGDIR_MASK)));
-+}
++ case BLKTAP_IOCTL_PRINT_IDXS:
++ {
++ if (info) {
++ printk("User Rings: \n-----------\n");
++ printk("UF: rsp_cons: %2d, req_prod_prv: %2d "
++ "| req_prod: %2d, rsp_prod: %2d\n",
++ info->ufe_ring.rsp_cons,
++ info->ufe_ring.req_prod_pvt,
++ info->ufe_ring.sring->req_prod,
++ info->ufe_ring.sring->rsp_prod);
++ }
++ return 0;
++ }
++ case BLKTAP_IOCTL_SENDPID:
++ {
++ if (info) {
++ info->pid = (pid_t)arg;
++ DPRINTK("blktap: pid received %d\n",
++ info->pid);
++ }
++ return 0;
++ }
++ case BLKTAP_IOCTL_NEWINTF:
++ {
++ uint64_t val = (uint64_t)arg;
++ domid_translate_t *tr = (domid_translate_t *)&val;
+
-+static int __init enable_pagefaulttrace(char *str)
-+{
-+ page_fault_trace = 1;
-+ return 1;
-+}
-+__setup("pagefaulttrace", enable_pagefaulttrace);
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/mm/init-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/mm/init-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/mm/init-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/mm/init-xen.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,1238 @@
-+/*
-+ * linux/arch/x86_64/mm/init.c
-+ *
-+ * Copyright (C) 1995 Linus Torvalds
-+ * Copyright (C) 2000 Pavel Machek <pavel at suse.cz>
-+ * Copyright (C) 2002,2003 Andi Kleen <ak at suse.de>
-+ *
-+ * Jun Nakajima <jun.nakajima at intel.com>
-+ * Modified for Xen.
-+ */
++ DPRINTK("NEWINTF Req for domid %d and bus id %d\n",
++ tr->domid, tr->busid);
++ info = get_next_free_dev();
++ if (!info) {
++ WPRINTK("Error initialising /dev/xen/blktap - "
++ "No more devices\n");
++ return -1;
++ }
++ info->trans.domid = tr->domid;
++ info->trans.busid = tr->busid;
++ return info->minor;
++ }
++ case BLKTAP_IOCTL_NEWINTF_EXT:
++ {
++ void __user *udata = (void __user *) arg;
++ domid_translate_ext_t tr;
+
-+#include <linux/signal.h>
-+#include <linux/sched.h>
-+#include <linux/kernel.h>
-+#include <linux/errno.h>
-+#include <linux/string.h>
-+#include <linux/types.h>
-+#include <linux/ptrace.h>
-+#include <linux/mman.h>
-+#include <linux/mm.h>
-+#include <linux/swap.h>
-+#include <linux/smp.h>
-+#include <linux/init.h>
-+#include <linux/pagemap.h>
-+#include <linux/bootmem.h>
-+#include <linux/proc_fs.h>
-+#include <linux/pci.h>
-+#include <linux/poison.h>
-+#include <linux/dma-mapping.h>
-+#include <linux/module.h>
-+#include <linux/memory_hotplug.h>
++ if (copy_from_user(&tr, udata, sizeof(domid_translate_ext_t)))
++ return -EFAULT;
+
-+#include <asm/processor.h>
-+#include <asm/system.h>
-+#include <asm/uaccess.h>
-+#include <asm/pgtable.h>
-+#include <asm/pgalloc.h>
-+#include <asm/dma.h>
-+#include <asm/fixmap.h>
-+#include <asm/e820.h>
-+#include <asm/apic.h>
-+#include <asm/tlb.h>
-+#include <asm/mmu_context.h>
-+#include <asm/proto.h>
-+#include <asm/smp.h>
-+#include <asm/sections.h>
++ DPRINTK("NEWINTF_EXT Req for domid %d and bus id %d\n",
++ tr.domid, tr.busid);
++ info = get_next_free_dev();
++ if (!info) {
++ WPRINTK("Error initialising /dev/xen/blktap - "
++ "No more devices\n");
++ return -1;
++ }
++ info->trans.domid = tr.domid;
++ info->trans.busid = tr.busid;
++ return info->minor;
++ }
++ case BLKTAP_IOCTL_FREEINTF:
++ {
++ unsigned long dev = arg;
++ unsigned long flags;
+
-+#include <xen/features.h>
++ info = tapfds[dev];
+
-+#ifndef Dprintk
-+#define Dprintk(x...)
-+#endif
++ if ((dev > MAX_TAP_DEV) || !info)
++ return 0; /* should this be an error? */
+
-+struct dma_mapping_ops* dma_ops;
-+EXPORT_SYMBOL(dma_ops);
++ spin_lock_irqsave(&pending_free_lock, flags);
++ if (info->dev_pending)
++ info->dev_pending = 0;
++ spin_unlock_irqrestore(&pending_free_lock, flags);
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+unsigned int __kernel_page_user;
-+EXPORT_SYMBOL(__kernel_page_user);
-+#endif
++ return 0;
++ }
++ case BLKTAP_IOCTL_MINOR:
++ {
++ unsigned long dev = arg;
+
-+extern unsigned long *contiguous_bitmap;
++ info = tapfds[dev];
+
-+static unsigned long dma_reserve __initdata;
++ if ((dev > MAX_TAP_DEV) || !info)
++ return -EINVAL;
+
-+DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
-+extern unsigned long start_pfn;
++ return info->minor;
++ }
++ case BLKTAP_IOCTL_MAJOR:
++ return blktap_major;
+
-+/*
-+ * Use this until direct mapping is established, i.e. before __va() is
-+ * available in init_memory_mapping().
-+ */
++ case BLKTAP_QUERY_ALLOC_REQS:
++ {
++ WPRINTK("BLKTAP_QUERY_ALLOC_REQS ioctl: %d/%d\n",
++ alloc_pending_reqs, blkif_reqs);
++ return (alloc_pending_reqs/blkif_reqs) * 100;
++ }
++ }
++ return -ENOIOCTLCMD;
++}
+
-+#define addr_to_page(addr, page) \
-+ (addr) &= PHYSICAL_PAGE_MASK; \
-+ (page) = ((unsigned long *) ((unsigned long) \
-+ (((mfn_to_pfn((addr) >> PAGE_SHIFT)) << PAGE_SHIFT) + \
-+ __START_KERNEL_map)))
++static unsigned int blktap_poll(struct file *filp, poll_table *wait)
++{
++ tap_blkif_t *info = filp->private_data;
++
++ /* do not work on the control device */
++ if (!info)
++ return 0;
+
-+static void __meminit early_make_page_readonly(void *va, unsigned int feature)
++ poll_wait(filp, &info->wait, wait);
++ if (info->ufe_ring.req_prod_pvt != info->ufe_ring.sring->req_prod) {
++ RING_PUSH_REQUESTS(&info->ufe_ring);
++ return POLLIN | POLLRDNORM;
++ }
++ return 0;
++}
++
++void blktap_kick_user(int idx)
+{
-+ unsigned long addr, _va = (unsigned long)va;
-+ pte_t pte, *ptep;
-+ unsigned long *page = (unsigned long *) init_level4_pgt;
++ tap_blkif_t *info;
+
-+ if (xen_feature(feature))
++ info = tapfds[idx];
++
++ if ((idx < 0) || (idx > MAX_TAP_DEV) || !info)
+ return;
+
-+ addr = (unsigned long) page[pgd_index(_va)];
-+ addr_to_page(addr, page);
++ wake_up_interruptible(&info->wait);
+
-+ addr = page[pud_index(_va)];
-+ addr_to_page(addr, page);
++ return;
++}
+
-+ addr = page[pmd_index(_va)];
-+ addr_to_page(addr, page);
++static int do_block_io_op(blkif_t *blkif);
++static void dispatch_rw_block_io(blkif_t *blkif,
++ blkif_request_t *req,
++ pending_req_t *pending_req);
++static void make_response(blkif_t *blkif, u64 id,
++ unsigned short op, int st);
+
-+ ptep = (pte_t *) &page[pte_index(_va)];
++/******************************************************************
++ * misc small helpers
++ */
++static int req_increase(void)
++{
++ int i, j;
+
-+ pte.pte = ptep->pte & ~_PAGE_RW;
-+ if (HYPERVISOR_update_va_mapping(_va, pte, 0))
-+ BUG();
++ if (mmap_alloc >= MAX_PENDING_REQS || mmap_lock)
++ return -EINVAL;
++
++ pending_reqs[mmap_alloc] = kzalloc(sizeof(pending_req_t)
++ * blkif_reqs, GFP_KERNEL);
++ foreign_pages[mmap_alloc] = alloc_empty_pages_and_pagevec(mmap_pages);
++
++ if (!pending_reqs[mmap_alloc] || !foreign_pages[mmap_alloc])
++ goto out_of_memory;
++
++ DPRINTK("%s: reqs=%d, pages=%d\n",
++ __FUNCTION__, blkif_reqs, mmap_pages);
++
++ for (i = 0; i < MAX_PENDING_REQS; i++) {
++ list_add_tail(&pending_reqs[mmap_alloc][i].free_list,
++ &pending_free);
++ pending_reqs[mmap_alloc][i].mem_idx = mmap_alloc;
++ for (j = 0; j < BLKIF_MAX_SEGMENTS_PER_REQUEST; j++)
++ BLKTAP_INVALIDATE_HANDLE(&pending_handle(mmap_alloc,
++ i, j));
++ }
++
++ mmap_alloc++;
++ DPRINTK("# MMAPs increased to %d\n",mmap_alloc);
++ return 0;
++
++ out_of_memory:
++ free_empty_pages_and_pagevec(foreign_pages[mmap_alloc], mmap_pages);
++ kfree(pending_reqs[mmap_alloc]);
++ WPRINTK("%s: out of memory\n", __FUNCTION__);
++ return -ENOMEM;
+}
+
-+static void __make_page_readonly(void *va)
++static void mmap_req_del(int mmap)
+{
-+ pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t pte, *ptep;
-+ unsigned long addr = (unsigned long) va;
++ BUG_ON(!spin_is_locked(&pending_free_lock));
+
-+ pgd = pgd_offset_k(addr);
-+ pud = pud_offset(pgd, addr);
-+ pmd = pmd_offset(pud, addr);
-+ ptep = pte_offset_kernel(pmd, addr);
++ kfree(pending_reqs[mmap]);
++ pending_reqs[mmap] = NULL;
+
-+ pte.pte = ptep->pte & ~_PAGE_RW;
-+ if (HYPERVISOR_update_va_mapping(addr, pte, 0))
-+ xen_l1_entry_update(ptep, pte); /* fallback */
++ free_empty_pages_and_pagevec(foreign_pages[mmap_alloc], mmap_pages);
++ foreign_pages[mmap] = NULL;
+
-+ if ((addr >= VMALLOC_START) && (addr < VMALLOC_END))
-+ __make_page_readonly(__va(pte_pfn(pte) << PAGE_SHIFT));
++ mmap_lock = 0;
++ DPRINTK("# MMAPs decreased to %d\n",mmap_alloc);
++ mmap_alloc--;
+}
+
-+static void __make_page_writable(void *va)
++static pending_req_t* alloc_req(void)
+{
-+ pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t pte, *ptep;
-+ unsigned long addr = (unsigned long) va;
++ pending_req_t *req = NULL;
++ unsigned long flags;
+
-+ pgd = pgd_offset_k(addr);
-+ pud = pud_offset(pgd, addr);
-+ pmd = pmd_offset(pud, addr);
-+ ptep = pte_offset_kernel(pmd, addr);
++ spin_lock_irqsave(&pending_free_lock, flags);
+
-+ pte.pte = ptep->pte | _PAGE_RW;
-+ if (HYPERVISOR_update_va_mapping(addr, pte, 0))
-+ xen_l1_entry_update(ptep, pte); /* fallback */
++ if (!list_empty(&pending_free)) {
++ req = list_entry(pending_free.next, pending_req_t, free_list);
++ list_del(&req->free_list);
++ }
+
-+ if ((addr >= VMALLOC_START) && (addr < VMALLOC_END))
-+ __make_page_writable(__va(pte_pfn(pte) << PAGE_SHIFT));
-+}
++ if (req) {
++ req->inuse = 1;
++ alloc_pending_reqs++;
++ }
++ spin_unlock_irqrestore(&pending_free_lock, flags);
+
-+void make_page_readonly(void *va, unsigned int feature)
-+{
-+ if (!xen_feature(feature))
-+ __make_page_readonly(va);
++ return req;
+}
+
-+void make_page_writable(void *va, unsigned int feature)
++static void free_req(pending_req_t *req)
+{
-+ if (!xen_feature(feature))
-+ __make_page_writable(va);
-+}
++ unsigned long flags;
++ int was_empty;
+
-+void make_pages_readonly(void *va, unsigned nr, unsigned int feature)
-+{
-+ if (xen_feature(feature))
-+ return;
++ spin_lock_irqsave(&pending_free_lock, flags);
+
-+ while (nr-- != 0) {
-+ __make_page_readonly(va);
-+ va = (void*)((unsigned long)va + PAGE_SIZE);
++ alloc_pending_reqs--;
++ req->inuse = 0;
++ if (mmap_lock && (req->mem_idx == mmap_alloc-1)) {
++ mmap_inuse--;
++ if (mmap_inuse == 0) mmap_req_del(mmap_alloc-1);
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++ return;
+ }
++ was_empty = list_empty(&pending_free);
++ list_add(&req->free_list, &pending_free);
++
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++
++ if (was_empty)
++ wake_up(&pending_free_wq);
+}
+
-+void make_pages_writable(void *va, unsigned nr, unsigned int feature)
++static void fast_flush_area(pending_req_t *req, int k_idx, int u_idx,
++ int tapidx)
+{
-+ if (xen_feature(feature))
++ struct gnttab_unmap_grant_ref unmap[BLKIF_MAX_SEGMENTS_PER_REQUEST*2];
++ unsigned int i, invcount = 0;
++ struct grant_handle_pair *khandle;
++ uint64_t ptep;
++ int ret, mmap_idx;
++ unsigned long kvaddr, uvaddr;
++ tap_blkif_t *info;
++
++
++ info = tapfds[tapidx];
++
++ if ((tapidx < 0) || (tapidx > MAX_TAP_DEV) || !info) {
++ WPRINTK("fast_flush: Couldn't get info!\n");
+ return;
++ }
+
-+ while (nr-- != 0) {
-+ __make_page_writable(va);
-+ va = (void*)((unsigned long)va + PAGE_SIZE);
++ if (info->vma != NULL &&
++ xen_feature(XENFEAT_auto_translated_physmap)) {
++ down_write(&info->vma->vm_mm->mmap_sem);
++ zap_page_range(info->vma,
++ MMAP_VADDR(info->user_vstart, u_idx, 0),
++ req->nr_pages << PAGE_SHIFT, NULL);
++ up_write(&info->vma->vm_mm->mmap_sem);
++ return;
+ }
-+}
+
-+/*
-+ * NOTE: pagetable_init alloc all the fixmap pagetables contiguous on the
-+ * physical space so we can cache the place of the first one and move
-+ * around without checking the pgd every time.
-+ */
++ mmap_idx = req->mem_idx;
+
-+void show_mem(void)
-+{
-+ long i, total = 0, reserved = 0;
-+ long shared = 0, cached = 0;
-+ pg_data_t *pgdat;
-+ struct page *page;
++ for (i = 0; i < req->nr_pages; i++) {
++ kvaddr = idx_to_kaddr(mmap_idx, k_idx, i);
++ uvaddr = MMAP_VADDR(info->user_vstart, u_idx, i);
+
-+ printk(KERN_INFO "Mem-info:\n");
-+ show_free_areas();
-+ printk(KERN_INFO "Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10));
++ khandle = &pending_handle(mmap_idx, k_idx, i);
+
-+ for_each_online_pgdat(pgdat) {
-+ for (i = 0; i < pgdat->node_spanned_pages; ++i) {
-+ page = pfn_to_page(pgdat->node_start_pfn + i);
-+ total++;
-+ if (PageReserved(page))
-+ reserved++;
-+ else if (PageSwapCache(page))
-+ cached++;
-+ else if (page_count(page))
-+ shared += page_count(page) - 1;
-+ }
-+ }
-+ printk(KERN_INFO "%lu pages of RAM\n", total);
-+ printk(KERN_INFO "%lu reserved pages\n",reserved);
-+ printk(KERN_INFO "%lu pages shared\n",shared);
-+ printk(KERN_INFO "%lu pages swap cached\n",cached);
-+}
++ if (khandle->kernel != INVALID_GRANT_HANDLE) {
++ gnttab_set_unmap_op(&unmap[invcount],
++ idx_to_kaddr(mmap_idx, k_idx, i),
++ GNTMAP_host_map, khandle->kernel);
++ invcount++;
+
-+int after_bootmem;
++ set_phys_to_machine(
++ __pa(idx_to_kaddr(mmap_idx, k_idx, i))
++ >> PAGE_SHIFT, INVALID_P2M_ENTRY);
++ }
+
-+static __init void *spp_getpage(void)
-+{
-+ void *ptr;
-+ if (after_bootmem)
-+ ptr = (void *) get_zeroed_page(GFP_ATOMIC);
-+ else if (start_pfn < table_end) {
-+ ptr = __va(start_pfn << PAGE_SHIFT);
-+ start_pfn++;
-+ memset(ptr, 0, PAGE_SIZE);
-+ } else
-+ ptr = alloc_bootmem_pages(PAGE_SIZE);
-+ if (!ptr || ((unsigned long)ptr & ~PAGE_MASK))
-+ panic("set_pte_phys: cannot allocate page data %s\n", after_bootmem?"after bootmem":"");
++ if (khandle->user != INVALID_GRANT_HANDLE) {
++ BUG_ON(xen_feature(XENFEAT_auto_translated_physmap));
++ if (create_lookup_pte_addr(
++ info->vma->vm_mm,
++ MMAP_VADDR(info->user_vstart, u_idx, i),
++ &ptep) !=0) {
++ WPRINTK("Couldn't get a pte addr!\n");
++ return;
++ }
+
-+ Dprintk("spp_getpage %p\n", ptr);
-+ return ptr;
-+}
++ gnttab_set_unmap_op(&unmap[invcount], ptep,
++ GNTMAP_host_map
++ | GNTMAP_application_map
++ | GNTMAP_contains_pte,
++ khandle->user);
++ invcount++;
++ }
++
++ BLKTAP_INVALIDATE_HANDLE(khandle);
++ }
++ ret = HYPERVISOR_grant_table_op(
++ GNTTABOP_unmap_grant_ref, unmap, invcount);
++ BUG_ON(ret);
++
++ if (info->vma != NULL && !xen_feature(XENFEAT_auto_translated_physmap))
++ zap_page_range(info->vma,
++ MMAP_VADDR(info->user_vstart, u_idx, 0),
++ req->nr_pages << PAGE_SHIFT, NULL);
++}
+
-+#define pgd_offset_u(address) (pgd_t *)(init_level4_user_pgt + pgd_index(address))
++/******************************************************************
++ * SCHEDULER FUNCTIONS
++ */
+
-+static inline pud_t *pud_offset_u(unsigned long address)
++static void print_stats(blkif_t *blkif)
+{
-+ pud_t *pud = level3_user_pgt;
-+
-+ return pud + pud_index(address);
++ printk(KERN_DEBUG "%s: oo %3d | rd %4d | wr %4d\n",
++ current->comm, blkif->st_oo_req,
++ blkif->st_rd_req, blkif->st_wr_req);
++ blkif->st_print = jiffies + msecs_to_jiffies(10 * 1000);
++ blkif->st_rd_req = 0;
++ blkif->st_wr_req = 0;
++ blkif->st_oo_req = 0;
+}
+
-+static __init void set_pte_phys(unsigned long vaddr,
-+ unsigned long phys, pgprot_t prot, int user_mode)
++int tap_blkif_schedule(void *arg)
+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte, new_pte;
++ blkif_t *blkif = arg;
+
-+ Dprintk("set_pte_phys %lx to %lx\n", vaddr, phys);
++ blkif_get(blkif);
+
-+ pgd = (user_mode ? pgd_offset_u(vaddr) : pgd_offset_k(vaddr));
-+ if (pgd_none(*pgd)) {
-+ printk("PGD FIXMAP MISSING, it should be setup in head.S!\n");
-+ return;
-+ }
-+ pud = (user_mode ? pud_offset_u(vaddr) : pud_offset(pgd, vaddr));
-+ if (pud_none(*pud)) {
-+ pmd = (pmd_t *) spp_getpage();
-+ make_page_readonly(pmd, XENFEAT_writable_page_tables);
-+ set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE | _PAGE_USER));
-+ if (pmd != pmd_offset(pud, 0)) {
-+ printk("PAGETABLE BUG #01! %p <-> %p\n", pmd, pmd_offset(pud,0));
-+ return;
-+ }
-+ }
-+ pmd = pmd_offset(pud, vaddr);
-+ if (pmd_none(*pmd)) {
-+ pte = (pte_t *) spp_getpage();
-+ make_page_readonly(pte, XENFEAT_writable_page_tables);
-+ set_pmd(pmd, __pmd(__pa(pte) | _KERNPG_TABLE | _PAGE_USER));
-+ if (pte != pte_offset_kernel(pmd, 0)) {
-+ printk("PAGETABLE BUG #02!\n");
-+ return;
-+ }
-+ }
-+ if (pgprot_val(prot))
-+ new_pte = pfn_pte(phys >> PAGE_SHIFT, prot);
-+ else
-+ new_pte = __pte(0);
++ if (debug_lvl)
++ printk(KERN_DEBUG "%s: started\n", current->comm);
+
-+ pte = pte_offset_kernel(pmd, vaddr);
-+ if (!pte_none(*pte) &&
-+ pte_val(*pte) != (pte_val(new_pte) & __supported_pte_mask))
-+ pte_ERROR(*pte);
-+ set_pte(pte, new_pte);
++ while (!kthread_should_stop()) {
++ if (try_to_freeze())
++ continue;
+
-+ /*
-+ * It's enough to flush this one mapping.
-+ * (PGE mappings get flushed as well)
-+ */
-+ __flush_tlb_one(vaddr);
-+}
++ wait_event_interruptible(
++ blkif->wq,
++ blkif->waiting_reqs || kthread_should_stop());
++ wait_event_interruptible(
++ pending_free_wq,
++ !list_empty(&pending_free) || kthread_should_stop());
+
-+static __init void set_pte_phys_ma(unsigned long vaddr,
-+ unsigned long phys, pgprot_t prot)
-+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte, new_pte;
++ blkif->waiting_reqs = 0;
++ smp_mb(); /* clear flag *before* checking for work */
+
-+ Dprintk("set_pte_phys %lx to %lx\n", vaddr, phys);
++ if (do_block_io_op(blkif))
++ blkif->waiting_reqs = 1;
+
-+ pgd = pgd_offset_k(vaddr);
-+ if (pgd_none(*pgd)) {
-+ printk("PGD FIXMAP MISSING, it should be setup in head.S!\n");
-+ return;
++ if (log_stats && time_after(jiffies, blkif->st_print))
++ print_stats(blkif);
+ }
-+ pud = pud_offset(pgd, vaddr);
-+ if (pud_none(*pud)) {
+
-+ pmd = (pmd_t *) spp_getpage();
-+ make_page_readonly(pmd, XENFEAT_writable_page_tables);
-+ set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE | _PAGE_USER));
-+ if (pmd != pmd_offset(pud, 0)) {
-+ printk("PAGETABLE BUG #01! %p <-> %p\n", pmd, pmd_offset(pud,0));
-+ return;
-+ }
-+ }
-+ pmd = pmd_offset(pud, vaddr);
-+ if (pmd_none(*pmd)) {
-+ pte = (pte_t *) spp_getpage();
-+ make_page_readonly(pte, XENFEAT_writable_page_tables);
-+ set_pmd(pmd, __pmd(__pa(pte) | _KERNPG_TABLE | _PAGE_USER));
-+ if (pte != pte_offset_kernel(pmd, 0)) {
-+ printk("PAGETABLE BUG #02!\n");
-+ return;
-+ }
-+ }
-+ new_pte = pfn_pte_ma(phys >> PAGE_SHIFT, prot);
++ if (log_stats)
++ print_stats(blkif);
++ if (debug_lvl)
++ printk(KERN_DEBUG "%s: exiting\n", current->comm);
+
-+ pte = pte_offset_kernel(pmd, vaddr);
-+ set_pte(pte, new_pte);
++ blkif->xenblkd = NULL;
++ blkif_put(blkif);
+
-+ /*
-+ * It's enough to flush this one mapping.
-+ * (PGE mappings get flushed as well)
-+ */
-+ __flush_tlb_one(vaddr);
++ return 0;
+}
+
-+#define SET_FIXMAP_KERNEL 0
-+#define SET_FIXMAP_USER 1
++/******************************************************************
++ * COMPLETION CALLBACK -- Called by user level ioctl()
++ */
+
-+/* NOTE: this is meant to be run only at boot */
-+void __init
-+__set_fixmap (enum fixed_addresses idx, unsigned long phys, pgprot_t prot)
++static int blktap_read_ufe_ring(tap_blkif_t *info)
+{
-+ unsigned long address = __fix_to_virt(idx);
++ /* This is called to read responses from the UFE ring. */
++ RING_IDX i, j, rp;
++ blkif_response_t *resp;
++ blkif_t *blkif=NULL;
++ int pending_idx, usr_idx, mmap_idx;
++ pending_req_t *pending_req;
++
++ if (!info)
++ return 0;
+
-+ if (idx >= __end_of_fixed_addresses) {
-+ printk("Invalid __set_fixmap\n");
-+ return;
-+ }
-+ switch (idx) {
-+ case VSYSCALL_LAST_PAGE ... VSYSCALL_FIRST_PAGE:
-+ set_pte_phys(address, phys, prot, SET_FIXMAP_KERNEL);
-+ break;
-+ default:
-+ set_pte_phys_ma(address, phys, prot);
-+ break;
-+ }
-+}
++ /* We currently only forward packets in INTERCEPT_FE mode. */
++ if (!(info->mode & BLKTAP_MODE_INTERCEPT_FE))
++ return 0;
+
-+/*
-+ * This only supports vsyscall area.
-+ */
-+void __init
-+__set_fixmap_user (enum fixed_addresses idx, unsigned long phys, pgprot_t prot)
-+{
-+ unsigned long address = __fix_to_virt(idx);
++ /* for each outstanding message on the UFEring */
++ rp = info->ufe_ring.sring->rsp_prod;
++ rmb();
++
++ for (i = info->ufe_ring.rsp_cons; i != rp; i++) {
++ blkif_response_t res;
++ resp = RING_GET_RESPONSE(&info->ufe_ring, i);
++ memcpy(&res, resp, sizeof(res));
++ mb(); /* rsp_cons read by RING_FULL() in do_block_io_op(). */
++ ++info->ufe_ring.rsp_cons;
+
-+ if (idx >= __end_of_fixed_addresses) {
-+ printk("Invalid __set_fixmap\n");
-+ return;
-+ }
++ /*retrieve [usr_idx] to [mmap_idx,pending_idx] mapping*/
++ usr_idx = (int)res.id;
++ pending_idx = MASK_PEND_IDX(ID_TO_IDX(info->idx_map[usr_idx]));
++ mmap_idx = ID_TO_MIDX(info->idx_map[usr_idx]);
+
-+ set_pte_phys(address, phys, prot, SET_FIXMAP_USER);
-+}
++ if ( (mmap_idx >= mmap_alloc) ||
++ (ID_TO_IDX(info->idx_map[usr_idx]) >= MAX_PENDING_REQS) )
++ WPRINTK("Incorrect req map"
++ "[%d], internal map [%d,%d (%d)]\n",
++ usr_idx, mmap_idx,
++ ID_TO_IDX(info->idx_map[usr_idx]),
++ MASK_PEND_IDX(
++ ID_TO_IDX(info->idx_map[usr_idx])));
+
-+unsigned long __initdata table_start, table_end;
++ pending_req = &pending_reqs[mmap_idx][pending_idx];
++ blkif = pending_req->blkif;
+
-+static __meminit void *alloc_static_page(unsigned long *phys)
-+{
-+ unsigned long va = (start_pfn << PAGE_SHIFT) + __START_KERNEL_map;
++ for (j = 0; j < pending_req->nr_pages; j++) {
+
-+ if (after_bootmem) {
-+ void *adr = (void *)get_zeroed_page(GFP_ATOMIC);
++ unsigned long kvaddr, uvaddr;
++ struct page **map = info->vma->vm_private_data;
++ struct page *pg;
++ int offset;
+
-+ *phys = __pa(adr);
-+ return adr;
++ uvaddr = MMAP_VADDR(info->user_vstart, usr_idx, j);
++ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, j);
++
++ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
++ ClearPageReserved(pg);
++ offset = (uvaddr - info->vma->vm_start)
++ >> PAGE_SHIFT;
++ map[offset] = NULL;
++ }
++ fast_flush_area(pending_req, pending_idx, usr_idx, info->minor);
++ info->idx_map[usr_idx] = INVALID_REQ;
++ make_response(blkif, pending_req->id, res.operation,
++ res.status);
++ blkif_put(pending_req->blkif);
++ free_req(pending_req);
+ }
++
++ return 0;
++}
+
-+ *phys = start_pfn << PAGE_SHIFT;
-+ start_pfn++;
-+ memset((void *)va, 0, PAGE_SIZE);
-+ return (void *)va;
-+}
+
-+#define PTE_SIZE PAGE_SIZE
++/******************************************************************************
++ * NOTIFICATION FROM GUEST OS.
++ */
+
-+static inline void __set_pte(pte_t *dst, pte_t val)
++static void blkif_notify_work(blkif_t *blkif)
+{
-+ *dst = val;
++ blkif->waiting_reqs = 1;
++ wake_up(&blkif->wq);
+}
+
-+static inline int make_readonly(unsigned long paddr)
++irqreturn_t tap_blkif_be_int(int irq, void *dev_id, struct pt_regs *regs)
+{
-+ extern char __vsyscall_0;
-+ int readonly = 0;
++ blkif_notify_work(dev_id);
++ return IRQ_HANDLED;
++}
+
-+ /* Make new page tables read-only. */
-+ if (!xen_feature(XENFEAT_writable_page_tables)
-+ && (paddr >= (table_start << PAGE_SHIFT))
-+ && (paddr < (table_end << PAGE_SHIFT)))
-+ readonly = 1;
-+ /* Make old page tables read-only. */
-+ if (!xen_feature(XENFEAT_writable_page_tables)
-+ && (paddr >= (xen_start_info->pt_base - __START_KERNEL_map))
-+ && (paddr < (start_pfn << PAGE_SHIFT)))
-+ readonly = 1;
+
-+ /*
-+ * No need for writable mapping of kernel image. This also ensures that
-+ * page and descriptor tables embedded inside don't have writable
-+ * mappings. Exclude the vsyscall area here, allowing alternative
-+ * instruction patching to work.
-+ */
-+ if ((paddr >= __pa_symbol(&_text)) && (paddr < __pa_symbol(&_end))
-+ && !(paddr >= __pa_symbol(&__vsyscall_0)
-+ && paddr < __pa_symbol(&__vsyscall_0) + PAGE_SIZE))
-+ readonly = 1;
+
-+ return readonly;
-+}
++/******************************************************************
++ * DOWNWARD CALLS -- These interface with the block-device layer proper.
++ */
++static int print_dbug = 1;
++static int do_block_io_op(blkif_t *blkif)
++{
++ blkif_back_rings_t *blk_rings = &blkif->blk_rings;
++ blkif_request_t req;
++ pending_req_t *pending_req;
++ RING_IDX rc, rp;
++ int more_to_do = 0;
++ tap_blkif_t *info;
++
++ rc = blk_rings->common.req_cons;
++ rp = blk_rings->common.sring->req_prod;
++ rmb(); /* Ensure we see queued requests up to 'rp'. */
++
++ /*Check blkif has corresponding UE ring*/
++ if (blkif->dev_num < 0) {
++ /*oops*/
++ if (print_dbug) {
++ WPRINTK("Corresponding UE "
++ "ring does not exist!\n");
++ print_dbug = 0; /*We only print this message once*/
++ }
++ return 0;
++ }
++
++ info = tapfds[blkif->dev_num];
++
++ if (blkif->dev_num > MAX_TAP_DEV || !info || !info->dev_inuse) {
++ if (print_dbug) {
++ WPRINTK("Can't get UE info!\n");
++ print_dbug = 0;
++ }
++ return 0;
++ }
++
++ while (rc != rp) {
++
++ if (RING_FULL(&info->ufe_ring)) {
++ WPRINTK("RING_FULL! More to do\n");
++ more_to_do = 1;
++ break;
++ }
++
++ if (RING_REQUEST_CONS_OVERFLOW(&blk_rings->common, rc)) {
++ WPRINTK("RING_REQUEST_CONS_OVERFLOW!"
++ " More to do\n");
++ more_to_do = 1;
++ break;
++ }
++
++ pending_req = alloc_req();
++ if (NULL == pending_req) {
++ blkif->st_oo_req++;
++ more_to_do = 1;
++ break;
++ }
+
-+#ifndef CONFIG_XEN
-+/* Must run before zap_low_mappings */
-+__init void *early_ioremap(unsigned long addr, unsigned long size)
-+{
-+ unsigned long map = round_down(addr, LARGE_PAGE_SIZE);
++ if (kthread_should_stop()) {
++ more_to_do = 1;
++ break;
++ }
+
-+ /* actually usually some more */
-+ if (size >= LARGE_PAGE_SIZE) {
-+ printk("SMBIOS area too long %lu\n", size);
-+ return NULL;
-+ }
-+ set_pmd(temp_mappings[0].pmd, __pmd(map | _KERNPG_TABLE | _PAGE_PSE));
-+ map += LARGE_PAGE_SIZE;
-+ set_pmd(temp_mappings[1].pmd, __pmd(map | _KERNPG_TABLE | _PAGE_PSE));
-+ __flush_tlb();
-+ return temp_mappings[0].address + (addr & (LARGE_PAGE_SIZE-1));
-+}
++ switch (blkif->blk_protocol) {
++ case BLKIF_PROTOCOL_NATIVE:
++ memcpy(&req, RING_GET_REQUEST(&blk_rings->native, rc),
++ sizeof(req));
++ break;
++ case BLKIF_PROTOCOL_X86_32:
++ blkif_get_x86_32_req(&req, RING_GET_REQUEST(&blk_rings->x86_32, rc));
++ break;
++ case BLKIF_PROTOCOL_X86_64:
++ blkif_get_x86_64_req(&req, RING_GET_REQUEST(&blk_rings->x86_64, rc));
++ break;
++ default:
++ BUG();
++ }
++ blk_rings->common.req_cons = ++rc; /* before make_response() */
+
-+/* To avoid virtual aliases later */
-+__init void early_iounmap(void *addr, unsigned long size)
-+{
-+ if ((void *)round_down((unsigned long)addr, LARGE_PAGE_SIZE) != temp_mappings[0].address)
-+ printk("early_iounmap: bad address %p\n", addr);
-+ set_pmd(temp_mappings[0].pmd, __pmd(0));
-+ set_pmd(temp_mappings[1].pmd, __pmd(0));
-+ __flush_tlb();
-+}
-+#endif
++ /* Apply all sanity checks to /private copy/ of request. */
++ barrier();
+
-+static void __meminit
-+phys_pmd_init(pmd_t *pmd, unsigned long address, unsigned long end)
-+{
-+ int i, k;
++ switch (req.operation) {
++ case BLKIF_OP_READ:
++ blkif->st_rd_req++;
++ dispatch_rw_block_io(blkif, &req, pending_req);
++ break;
+
-+ for (i = 0; i < PTRS_PER_PMD; pmd++, i++) {
-+ unsigned long pte_phys;
-+ pte_t *pte, *pte_save;
++ case BLKIF_OP_WRITE:
++ blkif->st_wr_req++;
++ dispatch_rw_block_io(blkif, &req, pending_req);
++ break;
+
-+ if (address >= end) {
-+ if (!after_bootmem)
-+ for (; i < PTRS_PER_PMD; i++, pmd++)
-+ set_pmd(pmd, __pmd(0));
++ default:
++ /* A good sign something is wrong: sleep for a while to
++ * avoid excessive CPU consumption by a bad guest. */
++ msleep(1);
++ WPRINTK("unknown operation [%d]\n",
++ req.operation);
++ make_response(blkif, req.id, req.operation,
++ BLKIF_RSP_ERROR);
++ free_req(pending_req);
+ break;
+ }
-+ pte = alloc_static_page(&pte_phys);
-+ pte_save = pte;
-+ for (k = 0; k < PTRS_PER_PTE; pte++, k++, address += PTE_SIZE) {
-+ unsigned long pteval = address | _PAGE_NX | _KERNPG_TABLE;
+
-+ if ((address >= end) ||
-+ ((address >> PAGE_SHIFT) >=
-+ xen_start_info->nr_pages))
-+ pteval = 0;
-+ else if (make_readonly(address))
-+ pteval &= ~_PAGE_RW;
-+ __set_pte(pte, __pte(pteval & __supported_pte_mask));
-+ }
-+ pte = pte_save;
-+ early_make_page_readonly(pte, XENFEAT_writable_page_tables);
-+ set_pmd(pmd, __pmd(pte_phys | _KERNPG_TABLE));
++ /* Yield point for this unbounded loop. */
++ cond_resched();
+ }
++
++ blktap_kick_user(blkif->dev_num);
++
++ return more_to_do;
+}
+
-+static void __meminit
-+phys_pmd_update(pud_t *pud, unsigned long address, unsigned long end)
++static void dispatch_rw_block_io(blkif_t *blkif,
++ blkif_request_t *req,
++ pending_req_t *pending_req)
+{
-+ pmd_t *pmd = pmd_offset(pud, (unsigned long)__va(address));
-+
-+ if (pmd_none(*pmd)) {
-+ spin_lock(&init_mm.page_table_lock);
-+ phys_pmd_init(pmd, address, end);
-+ spin_unlock(&init_mm.page_table_lock);
-+ __flush_tlb_all();
-+ }
-+}
++ extern void ll_rw_block(int rw, int nr, struct buffer_head * bhs[]);
++ int op, operation = (req->operation == BLKIF_OP_WRITE) ? WRITE : READ;
++ struct gnttab_map_grant_ref map[BLKIF_MAX_SEGMENTS_PER_REQUEST*2];
++ unsigned int nseg;
++ int ret, i, nr_sects = 0;
++ tap_blkif_t *info;
++ blkif_request_t *target;
++ int pending_idx = RTN_PEND_IDX(pending_req,pending_req->mem_idx);
++ int usr_idx;
++ uint16_t mmap_idx = pending_req->mem_idx;
+
-+static void __meminit phys_pud_init(pud_t *pud, unsigned long address, unsigned long end)
-+{
-+ long i = pud_index(address);
++ if (blkif->dev_num < 0 || blkif->dev_num > MAX_TAP_DEV)
++ goto fail_response;
+
-+ pud = pud + i;
++ info = tapfds[blkif->dev_num];
++ if (info == NULL)
++ goto fail_response;
+
-+ if (after_bootmem && pud_val(*pud)) {
-+ phys_pmd_update(pud, address, end);
-+ return;
++ /* Check we have space on user ring - should never fail. */
++ usr_idx = GET_NEXT_REQ(info->idx_map);
++ if (usr_idx == INVALID_REQ) {
++ BUG();
++ goto fail_response;
+ }
+
-+ for (; i < PTRS_PER_PUD; pud++, i++) {
-+ unsigned long paddr, pmd_phys;
-+ pmd_t *pmd;
-+
-+ paddr = (address & PGDIR_MASK) + i*PUD_SIZE;
-+ if (paddr >= end)
-+ break;
++ /* Check that number of segments is sane. */
++ nseg = req->nr_segments;
++ if ( unlikely(nseg == 0) ||
++ unlikely(nseg > BLKIF_MAX_SEGMENTS_PER_REQUEST) ) {
++ WPRINTK("Bad number of segments in request (%d)\n", nseg);
++ goto fail_response;
++ }
++
++ /* Make sure userspace is ready. */
++ if (!info->ring_ok) {
++ WPRINTK("blktap: ring not ready for requests!\n");
++ goto fail_response;
++ }
+
-+ pmd = alloc_static_page(&pmd_phys);
-+ early_make_page_readonly(pmd, XENFEAT_writable_page_tables);
-+ spin_lock(&init_mm.page_table_lock);
-+ set_pud(pud, __pud(pmd_phys | _KERNPG_TABLE));
-+ phys_pmd_init(pmd, paddr, end);
-+ spin_unlock(&init_mm.page_table_lock);
++ if (RING_FULL(&info->ufe_ring)) {
++ WPRINTK("blktap: fe_ring is full, can't add "
++ "IO Request will be dropped. %d %d\n",
++ RING_SIZE(&info->ufe_ring),
++ RING_SIZE(&blkif->blk_rings.common));
++ goto fail_response;
+ }
-+ __flush_tlb();
-+}
+
-+void __init xen_init_pt(void)
-+{
-+ unsigned long addr, *page;
++ pending_req->blkif = blkif;
++ pending_req->id = req->id;
++ pending_req->operation = operation;
++ pending_req->status = BLKIF_RSP_OKAY;
++ pending_req->nr_pages = nseg;
++ op = 0;
++ for (i = 0; i < nseg; i++) {
++ unsigned long uvaddr;
++ unsigned long kvaddr;
++ uint64_t ptep;
++ uint32_t flags;
+
-+ /* Find the initial pte page that was built for us. */
-+ page = (unsigned long *)xen_start_info->pt_base;
-+ addr = page[pgd_index(__START_KERNEL_map)];
-+ addr_to_page(addr, page);
-+ addr = page[pud_index(__START_KERNEL_map)];
-+ addr_to_page(addr, page);
++ uvaddr = MMAP_VADDR(info->user_vstart, usr_idx, i);
++ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, i);
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ /* On Xen 3.0.2 and older we may need to explicitly specify _PAGE_USER
-+ in kernel PTEs. We check that here. */
-+ if (HYPERVISOR_xen_version(XENVER_version, NULL) <= 0x30000) {
-+ unsigned long *pg;
-+ pte_t pte;
++ flags = GNTMAP_host_map;
++ if (operation == WRITE)
++ flags |= GNTMAP_readonly;
++ gnttab_set_map_op(&map[op], kvaddr, flags,
++ req->seg[i].gref, blkif->domid);
++ op++;
+
-+ /* Mess with the initial mapping of page 0. It's not needed. */
-+ BUILD_BUG_ON(__START_KERNEL <= __START_KERNEL_map);
-+ addr = page[pmd_index(__START_KERNEL_map)];
-+ addr_to_page(addr, pg);
-+ pte.pte = pg[pte_index(__START_KERNEL_map)];
-+ BUG_ON(!(pte.pte & _PAGE_PRESENT));
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* Now map it to user. */
++ ret = create_lookup_pte_addr(info->vma->vm_mm,
++ uvaddr, &ptep);
++ if (ret) {
++ WPRINTK("Couldn't get a pte addr!\n");
++ goto fail_flush;
++ }
+
-+ /* If _PAGE_USER isn't set, we obviously do not need it. */
-+ if (pte.pte & _PAGE_USER) {
-+ /* _PAGE_USER is needed, but is it set implicitly? */
-+ pte.pte &= ~_PAGE_USER;
-+ if ((HYPERVISOR_update_va_mapping(__START_KERNEL_map,
-+ pte, 0) != 0) ||
-+ !(pg[pte_index(__START_KERNEL_map)] & _PAGE_USER))
-+ /* We need to explicitly specify _PAGE_USER. */
-+ __kernel_page_user = _PAGE_USER;
++ flags = GNTMAP_host_map | GNTMAP_application_map
++ | GNTMAP_contains_pte;
++ if (operation == WRITE)
++ flags |= GNTMAP_readonly;
++ gnttab_set_map_op(&map[op], ptep, flags,
++ req->seg[i].gref, blkif->domid);
++ op++;
+ }
++
++ nr_sects += (req->seg[i].last_sect -
++ req->seg[i].first_sect + 1);
+ }
-+#endif
+
-+ /* Construct mapping of initial pte page in our own directories. */
-+ init_level4_pgt[pgd_index(__START_KERNEL_map)] =
-+ __pgd(__pa_symbol(level3_kernel_pgt) | _PAGE_TABLE);
-+ level3_kernel_pgt[pud_index(__START_KERNEL_map)] =
-+ __pud(__pa_symbol(level2_kernel_pgt) |
-+ _KERNPG_TABLE);
-+ memcpy((void *)level2_kernel_pgt, page, PAGE_SIZE);
++ ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, map, op);
++ BUG_ON(ret);
+
-+ early_make_page_readonly(init_level4_pgt,
-+ XENFEAT_writable_page_tables);
-+ early_make_page_readonly(init_level4_user_pgt,
-+ XENFEAT_writable_page_tables);
-+ early_make_page_readonly(level3_kernel_pgt,
-+ XENFEAT_writable_page_tables);
-+ early_make_page_readonly(level3_user_pgt,
-+ XENFEAT_writable_page_tables);
-+ early_make_page_readonly(level2_kernel_pgt,
-+ XENFEAT_writable_page_tables);
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ for (i = 0; i < (nseg*2); i+=2) {
++ unsigned long uvaddr;
++ unsigned long kvaddr;
++ unsigned long offset;
++ struct page *pg;
+
-+ if (!xen_feature(XENFEAT_writable_page_tables)) {
-+ xen_pgd_pin(__pa_symbol(init_level4_pgt));
-+ xen_pgd_pin(__pa_symbol(init_level4_user_pgt));
-+ }
++ uvaddr = MMAP_VADDR(info->user_vstart, usr_idx, i/2);
++ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, i/2);
+
-+ set_pgd((pgd_t *)(init_level4_user_pgt + 511),
-+ mk_kernel_pgd(__pa_symbol(level3_user_pgt)));
-+}
++ if (unlikely(map[i].status != 0)) {
++ WPRINTK("invalid kernel buffer -- "
++ "could not remap it\n");
++ ret |= 1;
++ map[i].handle = INVALID_GRANT_HANDLE;
++ }
+
-+static void __init extend_init_mapping(unsigned long tables_space)
-+{
-+ unsigned long va = __START_KERNEL_map;
-+ unsigned long phys, addr, *pte_page;
-+ pmd_t *pmd;
-+ pte_t *pte, new_pte;
-+ unsigned long *page = (unsigned long *)init_level4_pgt;
++ if (unlikely(map[i+1].status != 0)) {
++ WPRINTK("invalid user buffer -- "
++ "could not remap it\n");
++ ret |= 1;
++ map[i+1].handle = INVALID_GRANT_HANDLE;
++ }
+
-+ addr = page[pgd_index(va)];
-+ addr_to_page(addr, page);
-+ addr = page[pud_index(va)];
-+ addr_to_page(addr, page);
++ pending_handle(mmap_idx, pending_idx, i/2).kernel
++ = map[i].handle;
++ pending_handle(mmap_idx, pending_idx, i/2).user
++ = map[i+1].handle;
+
-+ /* Kill mapping of low 1MB. */
-+ while (va < (unsigned long)&_text) {
-+ HYPERVISOR_update_va_mapping(va, __pte_ma(0), 0);
-+ va += PAGE_SIZE;
-+ }
++ if (ret)
++ continue;
+
-+ /* Ensure init mappings cover kernel text/data and initial tables. */
-+ while (va < (__START_KERNEL_map
-+ + (start_pfn << PAGE_SHIFT)
-+ + tables_space)) {
-+ pmd = (pmd_t *)&page[pmd_index(va)];
-+ if (pmd_none(*pmd)) {
-+ pte_page = alloc_static_page(&phys);
-+ early_make_page_readonly(
-+ pte_page, XENFEAT_writable_page_tables);
-+ set_pmd(pmd, __pmd(phys | _KERNPG_TABLE));
-+ } else {
-+ addr = page[pmd_index(va)];
-+ addr_to_page(addr, pte_page);
-+ }
-+ pte = (pte_t *)&pte_page[pte_index(va)];
-+ if (pte_none(*pte)) {
-+ new_pte = pfn_pte(
-+ (va - __START_KERNEL_map) >> PAGE_SHIFT,
-+ __pgprot(_KERNPG_TABLE));
-+ xen_l1_entry_update(pte, new_pte);
++ set_phys_to_machine(__pa(kvaddr) >> PAGE_SHIFT,
++ FOREIGN_FRAME(map[i].dev_bus_addr
++ >> PAGE_SHIFT));
++ offset = (uvaddr - info->vma->vm_start) >> PAGE_SHIFT;
++ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
++ ((struct page **)info->vma->vm_private_data)[offset] =
++ pg;
+ }
-+ va += PAGE_SIZE;
-+ }
-+
-+ /* Finally, blow away any spurious initial mappings. */
-+ while (1) {
-+ pmd = (pmd_t *)&page[pmd_index(va)];
-+ if (pmd_none(*pmd))
-+ break;
-+ HYPERVISOR_update_va_mapping(va, __pte_ma(0), 0);
-+ va += PAGE_SIZE;
-+ }
-+}
++ } else {
++ for (i = 0; i < nseg; i++) {
++ unsigned long uvaddr;
++ unsigned long kvaddr;
++ unsigned long offset;
++ struct page *pg;
+
-+static void __init find_early_table_space(unsigned long end)
-+{
-+ unsigned long puds, pmds, ptes, tables;
++ uvaddr = MMAP_VADDR(info->user_vstart, usr_idx, i);
++ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, i);
+
-+ puds = (end + PUD_SIZE - 1) >> PUD_SHIFT;
-+ pmds = (end + PMD_SIZE - 1) >> PMD_SHIFT;
-+ ptes = (end + PTE_SIZE - 1) >> PAGE_SHIFT;
++ if (unlikely(map[i].status != 0)) {
++ WPRINTK("invalid kernel buffer -- "
++ "could not remap it\n");
++ ret |= 1;
++ map[i].handle = INVALID_GRANT_HANDLE;
++ }
+
-+ tables = round_up(puds * 8, PAGE_SIZE) +
-+ round_up(pmds * 8, PAGE_SIZE) +
-+ round_up(ptes * 8, PAGE_SIZE);
++ pending_handle(mmap_idx, pending_idx, i).kernel
++ = map[i].handle;
+
-+ extend_init_mapping(tables);
++ if (ret)
++ continue;
+
-+ table_start = start_pfn;
-+ table_end = table_start + (tables>>PAGE_SHIFT);
++ offset = (uvaddr - info->vma->vm_start) >> PAGE_SHIFT;
++ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
++ ((struct page **)info->vma->vm_private_data)[offset] =
++ pg;
++ }
++ }
+
-+ early_printk("kernel direct mapping tables up to %lx @ %lx-%lx\n",
-+ end, table_start << PAGE_SHIFT,
-+ (table_start << PAGE_SHIFT) + tables);
-+}
++ if (ret)
++ goto fail_flush;
+
-+static void xen_finish_init_mapping(void)
-+{
-+ unsigned long i, start, end;
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ down_write(&info->vma->vm_mm->mmap_sem);
++ /* Mark mapped pages as reserved: */
++ for (i = 0; i < req->nr_segments; i++) {
++ unsigned long kvaddr;
++ struct page *pg;
+
-+ /* Re-vector virtual addresses pointing into the initial
-+ mapping to the just-established permanent ones. */
-+ xen_start_info = __va(__pa(xen_start_info));
-+ xen_start_info->pt_base = (unsigned long)
-+ __va(__pa(xen_start_info->pt_base));
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ phys_to_machine_mapping =
-+ __va(__pa(xen_start_info->mfn_list));
-+ xen_start_info->mfn_list = (unsigned long)
-+ phys_to_machine_mapping;
++ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, i);
++ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
++ SetPageReserved(pg);
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ ret = vm_insert_page(info->vma,
++ MMAP_VADDR(info->user_vstart,
++ usr_idx, i), pg);
++ if (ret) {
++ up_write(&info->vma->vm_mm->mmap_sem);
++ goto fail_flush;
++ }
++ }
+ }
-+ if (xen_start_info->mod_start)
-+ xen_start_info->mod_start = (unsigned long)
-+ __va(__pa(xen_start_info->mod_start));
-+
-+ /* Destroy the Xen-created mappings beyond the kernel image as
-+ * well as the temporary mappings created above. Prevents
-+ * overlap with modules area (if init mapping is very big).
-+ */
-+ start = PAGE_ALIGN((unsigned long)_end);
-+ end = __START_KERNEL_map + (table_end << PAGE_SHIFT);
-+ for (; start < end; start += PAGE_SIZE)
-+ WARN_ON(HYPERVISOR_update_va_mapping(
-+ start, __pte_ma(0), 0));
-+
-+ /* Allocate pte's for initial fixmaps from 'start_pfn' allocator. */
-+ table_end = ~0UL;
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ up_write(&info->vma->vm_mm->mmap_sem);
++
++ /*record [mmap_idx,pending_idx] to [usr_idx] mapping*/
++ info->idx_map[usr_idx] = MAKE_ID(mmap_idx, pending_idx);
+
-+ /*
-+ * Prefetch pte's for the bt_ioremap() area. It gets used before the
-+ * boot-time allocator is online, so allocate-on-demand would fail.
-+ */
-+ for (i = FIX_BTMAP_END; i <= FIX_BTMAP_BEGIN; i++)
-+ __set_fixmap(i, 0, __pgprot(0));
++ blkif_get(blkif);
++ /* Finally, write the request message to the user ring. */
++ target = RING_GET_REQUEST(&info->ufe_ring,
++ info->ufe_ring.req_prod_pvt);
++ memcpy(target, req, sizeof(*req));
++ target->id = usr_idx;
++ wmb(); /* blktap_poll() reads req_prod_pvt asynchronously */
++ info->ufe_ring.req_prod_pvt++;
+
-+ /* Switch to the real shared_info page, and clear the dummy page. */
-+ set_fixmap(FIX_SHARED_INFO, xen_start_info->shared_info);
-+ HYPERVISOR_shared_info = (shared_info_t *)fix_to_virt(FIX_SHARED_INFO);
-+ memset(empty_zero_page, 0, sizeof(empty_zero_page));
++ if (operation == READ)
++ blkif->st_rd_sect += nr_sects;
++ else if (operation == WRITE)
++ blkif->st_wr_sect += nr_sects;
+
-+ /* Set up mapping of lowest 1MB of physical memory. */
-+ for (i = 0; i < NR_FIX_ISAMAPS; i++)
-+ if (is_initial_xendomain())
-+ set_fixmap(FIX_ISAMAP_BEGIN - i, i * PAGE_SIZE);
-+ else
-+ __set_fixmap(FIX_ISAMAP_BEGIN - i,
-+ virt_to_mfn(empty_zero_page)
-+ << PAGE_SHIFT,
-+ PAGE_KERNEL_RO);
++ return;
+
-+ /* Disable the 'start_pfn' allocator. */
-+ table_end = start_pfn;
++ fail_flush:
++ WPRINTK("Reached Fail_flush\n");
++ fast_flush_area(pending_req, pending_idx, usr_idx, blkif->dev_num);
++ fail_response:
++ make_response(blkif, req->id, req->operation, BLKIF_RSP_ERROR);
++ free_req(pending_req);
++ msleep(1); /* back off a bit */
+}
+
-+/* Setup the direct mapping of the physical memory at PAGE_OFFSET.
-+ This runs before bootmem is initialized and gets pages directly from the
-+ physical memory. To access them they are temporarily mapped. */
-+void __meminit init_memory_mapping(unsigned long start, unsigned long end)
-+{
-+ unsigned long next;
+
-+ Dprintk("init_memory_mapping\n");
+
-+ /*
-+ * Find space for the kernel direct mapping tables.
-+ * Later we should allocate these tables in the local node of the memory
-+ * mapped. Unfortunately this is done currently before the nodes are
-+ * discovered.
-+ */
-+ if (!after_bootmem)
-+ find_early_table_space(end);
++/******************************************************************
++ * MISCELLANEOUS SETUP / TEARDOWN / DEBUGGING
++ */
+
-+ start = (unsigned long)__va(start);
-+ end = (unsigned long)__va(end);
+
-+ for (; start < end; start = next) {
-+ unsigned long pud_phys;
-+ pgd_t *pgd = pgd_offset_k(start);
-+ pud_t *pud;
++static void make_response(blkif_t *blkif, u64 id,
++ unsigned short op, int st)
++{
++ blkif_response_t resp;
++ unsigned long flags;
++ blkif_back_rings_t *blk_rings = &blkif->blk_rings;
++ int more_to_do = 0;
++ int notify;
+
-+ if (after_bootmem) {
-+ pud = pud_offset(pgd, start & PGDIR_MASK);
-+ make_page_readonly(pud, XENFEAT_writable_page_tables);
-+ pud_phys = __pa(pud);
-+ } else {
-+ pud = alloc_static_page(&pud_phys);
-+ early_make_page_readonly(pud, XENFEAT_writable_page_tables);
-+ }
-+ next = start + PGDIR_SIZE;
-+ if (next > end)
-+ next = end;
-+ phys_pud_init(pud, __pa(start), __pa(next));
-+ if (!after_bootmem)
-+ set_pgd(pgd_offset_k(start), mk_kernel_pgd(pud_phys));
++ resp.id = id;
++ resp.operation = op;
++ resp.status = st;
++
++ spin_lock_irqsave(&blkif->blk_ring_lock, flags);
++ /* Place on the response ring for the relevant domain. */
++ switch (blkif->blk_protocol) {
++ case BLKIF_PROTOCOL_NATIVE:
++ memcpy(RING_GET_RESPONSE(&blk_rings->native,
++ blk_rings->native.rsp_prod_pvt),
++ &resp, sizeof(resp));
++ break;
++ case BLKIF_PROTOCOL_X86_32:
++ memcpy(RING_GET_RESPONSE(&blk_rings->x86_32,
++ blk_rings->x86_32.rsp_prod_pvt),
++ &resp, sizeof(resp));
++ break;
++ case BLKIF_PROTOCOL_X86_64:
++ memcpy(RING_GET_RESPONSE(&blk_rings->x86_64,
++ blk_rings->x86_64.rsp_prod_pvt),
++ &resp, sizeof(resp));
++ break;
++ default:
++ BUG();
+ }
++ blk_rings->common.rsp_prod_pvt++;
++ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&blk_rings->common, notify);
+
-+ if (!after_bootmem) {
-+ BUG_ON(start_pfn != table_end);
-+ xen_finish_init_mapping();
++ if (blk_rings->common.rsp_prod_pvt == blk_rings->common.req_cons) {
++ /*
++ * Tail check for pending requests. Allows frontend to avoid
++ * notifications if requests are already in flight (lower
++ * overheads and promotes batching).
++ */
++ RING_FINAL_CHECK_FOR_REQUESTS(&blk_rings->common, more_to_do);
++ } else if (RING_HAS_UNCONSUMED_REQUESTS(&blk_rings->common)) {
++ more_to_do = 1;
+ }
+
-+ __flush_tlb_all();
-+}
-+
-+void __cpuinit zap_low_mappings(int cpu)
-+{
-+ /* this is not required for Xen */
-+#if 0
-+ swap_low_mappings();
-+#endif
++ spin_unlock_irqrestore(&blkif->blk_ring_lock, flags);
++ if (more_to_do)
++ blkif_notify_work(blkif);
++ if (notify)
++ notify_remote_via_irq(blkif->irq);
+}
+
-+/* Compute zone sizes for the DMA and DMA32 zones in a node. */
-+__init void
-+size_zones(unsigned long *z, unsigned long *h,
-+ unsigned long start_pfn, unsigned long end_pfn)
++static int __init blkif_init(void)
+{
-+ int i;
-+#ifndef CONFIG_XEN
-+ unsigned long w;
-+#endif
-+
-+ for (i = 0; i < MAX_NR_ZONES; i++)
-+ z[i] = 0;
++ int i, ret;
++ struct class *class;
+
-+#ifndef CONFIG_XEN
-+ if (start_pfn < MAX_DMA_PFN)
-+ z[ZONE_DMA] = MAX_DMA_PFN - start_pfn;
-+ if (start_pfn < MAX_DMA32_PFN) {
-+ unsigned long dma32_pfn = MAX_DMA32_PFN;
-+ if (dma32_pfn > end_pfn)
-+ dma32_pfn = end_pfn;
-+ z[ZONE_DMA32] = dma32_pfn - start_pfn;
-+ }
-+ z[ZONE_NORMAL] = end_pfn - start_pfn;
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+ /* Remove lower zones from higher ones. */
-+ w = 0;
-+ for (i = 0; i < MAX_NR_ZONES; i++) {
-+ if (z[i])
-+ z[i] -= w;
-+ w += z[i];
++ INIT_LIST_HEAD(&pending_free);
++ for(i = 0; i < 2; i++) {
++ ret = req_increase();
++ if (ret)
++ break;
+ }
++ if (i == 0)
++ return ret;
+
-+ /* Compute holes */
-+ w = start_pfn;
-+ for (i = 0; i < MAX_NR_ZONES; i++) {
-+ unsigned long s = w;
-+ w += z[i];
-+ h[i] = e820_hole_size(s, w);
-+ }
++ tap_blkif_interface_init();
+
-+ /* Add the space pace needed for mem_map to the holes too. */
-+ for (i = 0; i < MAX_NR_ZONES; i++)
-+ h[i] += (z[i] * sizeof(struct page)) / PAGE_SIZE;
++ alloc_pending_reqs = 0;
+
-+ /* The 16MB DMA zone has the kernel and other misc mappings.
-+ Account them too */
-+ if (h[ZONE_DMA]) {
-+ h[ZONE_DMA] += dma_reserve;
-+ if (h[ZONE_DMA] >= z[ZONE_DMA]) {
-+ printk(KERN_WARNING
-+ "Kernel too large and filling up ZONE_DMA?\n");
-+ h[ZONE_DMA] = z[ZONE_DMA];
-+ }
-+ }
-+#else
-+ z[ZONE_DMA] = end_pfn;
-+ for (i = 0; i < MAX_NR_ZONES; i++)
-+ h[i] = 0;
-+#endif
-+}
++ tap_blkif_xenbus_init();
+
-+#ifndef CONFIG_NUMA
-+void __init paging_init(void)
-+{
-+ unsigned long zones[MAX_NR_ZONES], holes[MAX_NR_ZONES];
++ /* Dynamically allocate a major for this device */
++ ret = register_chrdev(0, "blktap", &blktap_fops);
+
-+ memory_present(0, 0, end_pfn);
-+ sparse_init();
-+ size_zones(zones, holes, 0, end_pfn);
-+ free_area_init_node(0, NODE_DATA(0), zones,
-+ __pa(PAGE_OFFSET) >> PAGE_SHIFT, holes);
++ if (ret < 0) {
++ WPRINTK("Couldn't register /dev/xen/blktap\n");
++ return -ENOMEM;
++ }
++
++ blktap_major = ret;
+
-+ init_mm.context.pinned = 1;
-+}
-+#endif
++ /* tapfds[0] is always NULL */
++ blktap_next_minor++;
+
-+/* Unmap a kernel mapping if it exists. This is useful to avoid prefetches
-+ from the CPU leading to inconsistent cache lines. address and size
-+ must be aligned to 2MB boundaries.
-+ Does nothing when the mapping doesn't exist. */
-+void __init clear_kernel_mapping(unsigned long address, unsigned long size)
-+{
-+ unsigned long end = address + size;
++ DPRINTK("Created misc_dev [/dev/xen/blktap%d]\n",i);
+
-+ BUG_ON(address & ~LARGE_PAGE_MASK);
-+ BUG_ON(size & ~LARGE_PAGE_MASK);
-+
-+ for (; address < end; address += LARGE_PAGE_SIZE) {
-+ pgd_t *pgd = pgd_offset_k(address);
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ if (pgd_none(*pgd))
-+ continue;
-+ pud = pud_offset(pgd, address);
-+ if (pud_none(*pud))
-+ continue;
-+ pmd = pmd_offset(pud, address);
-+ if (!pmd || pmd_none(*pmd))
-+ continue;
-+ if (0 == (pmd_val(*pmd) & _PAGE_PSE)) {
-+ /* Could handle this, but it should not happen currently. */
-+ printk(KERN_ERR
-+ "clear_kernel_mapping: mapping has been split. will leak memory\n");
-+ pmd_ERROR(*pmd);
-+ }
-+ set_pmd(pmd, __pmd(0));
++ /* Make sure the xen class exists */
++ if ((class = get_xen_class()) != NULL) {
++ /*
++ * This will allow udev to create the blktap ctrl device.
++ * We only want to create blktap0 first. We don't want
++ * to flood the sysfs system with needless blktap devices.
++ * We only create the device when a request of a new device is
++ * made.
++ */
++ class_device_create(class, NULL,
++ MKDEV(blktap_major, 0), NULL,
++ "blktap0");
++ } else {
++ /* this is bad, but not fatal */
++ WPRINTK("blktap: sysfs xen_class not created\n");
+ }
-+ __flush_tlb_all();
-+}
+
-+/*
-+ * Memory hotplug specific functions
-+ */
-+void online_page(struct page *page)
-+{
-+ ClearPageReserved(page);
-+ init_page_count(page);
-+ __free_page(page);
-+ totalram_pages++;
-+ num_physpages++;
-+}
++ DPRINTK("Blktap device successfully created\n");
+
-+#ifdef CONFIG_MEMORY_HOTPLUG
-+/*
-+ * XXX: memory_add_physaddr_to_nid() is to find node id from physical address
-+ * via probe interface of sysfs. If acpi notifies hot-add event, then it
-+ * can tell node id by searching dsdt. But, probe interface doesn't have
-+ * node id. So, return 0 as node id at this time.
-+ */
-+#ifdef CONFIG_NUMA
-+int memory_add_physaddr_to_nid(u64 start)
-+{
+ return 0;
+}
-+#endif
+
-+/*
-+ * Memory is added always to NORMAL zone. This means you will never get
-+ * additional DMA/DMA32 memory.
++module_init(blkif_init);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blktap/common.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blktap/common.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,122 @@
++/*
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
-+int arch_add_memory(int nid, u64 start, u64 size)
-+{
-+ struct pglist_data *pgdat = NODE_DATA(nid);
-+ struct zone *zone = pgdat->node_zones + MAX_NR_ZONES-2;
-+ unsigned long start_pfn = start >> PAGE_SHIFT;
-+ unsigned long nr_pages = size >> PAGE_SHIFT;
-+ int ret;
+
-+ ret = __add_pages(zone, start_pfn, nr_pages);
-+ if (ret)
-+ goto error;
++#ifndef __BLKIF__BACKEND__COMMON_H__
++#define __BLKIF__BACKEND__COMMON_H__
+
-+ init_memory_mapping(start, (start + size -1));
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/interrupt.h>
++#include <linux/slab.h>
++#include <linux/blkdev.h>
++#include <linux/vmalloc.h>
++#include <asm/io.h>
++#include <asm/setup.h>
++#include <asm/pgalloc.h>
++#include <xen/evtchn.h>
++#include <asm/hypervisor.h>
++#include <xen/blkif.h>
++#include <xen/gnttab.h>
++#include <xen/driver_util.h>
+
-+ return ret;
-+error:
-+ printk("%s: Problem encountered in __add_pages!\n", __func__);
-+ return ret;
-+}
-+EXPORT_SYMBOL_GPL(arch_add_memory);
++#define DPRINTK(_f, _a...) pr_debug("(file=%s, line=%d) " _f, \
++ __FILE__ , __LINE__ , ## _a )
+
-+int remove_memory(u64 start, u64 size)
-+{
-+ return -EINVAL;
-+}
-+EXPORT_SYMBOL_GPL(remove_memory);
++#define WPRINTK(fmt, args...) printk(KERN_WARNING "blk_tap: " fmt, ##args)
+
-+#else /* CONFIG_MEMORY_HOTPLUG */
-+/*
-+ * Memory Hotadd without sparsemem. The mem_maps have been allocated in advance,
-+ * just online the pages.
-+ */
-+int __add_pages(struct zone *z, unsigned long start_pfn, unsigned long nr_pages)
-+{
-+ int err = -EIO;
-+ unsigned long pfn;
-+ unsigned long total = 0, mem = 0;
-+ for (pfn = start_pfn; pfn < start_pfn + nr_pages; pfn++) {
-+ if (pfn_valid(pfn)) {
-+ online_page(pfn_to_page(pfn));
-+ err = 0;
-+ mem++;
-+ }
-+ total++;
-+ }
-+ if (!err) {
-+ z->spanned_pages += total;
-+ z->present_pages += mem;
-+ z->zone_pgdat->node_spanned_pages += total;
-+ z->zone_pgdat->node_present_pages += mem;
-+ }
-+ return err;
-+}
-+#endif /* CONFIG_MEMORY_HOTPLUG */
++struct backend_info;
+
-+static struct kcore_list kcore_mem, kcore_vmalloc, kcore_kernel, kcore_modules,
-+ kcore_vsyscall;
++typedef struct blkif_st {
++ /* Unique identifier for this interface. */
++ domid_t domid;
++ unsigned int handle;
++ /* Physical parameters of the comms window. */
++ unsigned int irq;
++ /* Comms information. */
++ enum blkif_protocol blk_protocol;
++ blkif_back_rings_t blk_rings;
++ struct vm_struct *blk_ring_area;
++ /* Back pointer to the backend_info. */
++ struct backend_info *be;
++ /* Private fields. */
++ spinlock_t blk_ring_lock;
++ atomic_t refcnt;
+
-+void __init mem_init(void)
-+{
-+ long codesize, reservedpages, datasize, initsize;
-+ unsigned long pfn;
++ wait_queue_head_t wq;
++ struct task_struct *xenblkd;
++ unsigned int waiting_reqs;
++ request_queue_t *plug;
+
-+ contiguous_bitmap = alloc_bootmem_low_pages(
-+ (end_pfn + 2*BITS_PER_LONG) >> 3);
-+ BUG_ON(!contiguous_bitmap);
-+ memset(contiguous_bitmap, 0, (end_pfn + 2*BITS_PER_LONG) >> 3);
++ /* statistics */
++ unsigned long st_print;
++ int st_rd_req;
++ int st_wr_req;
++ int st_oo_req;
++ int st_rd_sect;
++ int st_wr_sect;
+
-+ pci_iommu_alloc();
++ wait_queue_head_t waiting_to_free;
+
-+ /* How many end-of-memory variables you have, grandma! */
-+ max_low_pfn = end_pfn;
-+ max_pfn = end_pfn;
-+ num_physpages = end_pfn;
-+ high_memory = (void *) __va(end_pfn * PAGE_SIZE);
++ grant_handle_t shmem_handle;
++ grant_ref_t shmem_ref;
++
++ int dev_num;
++ uint64_t sectors;
++} blkif_t;
+
-+ /* clear the zero-page */
-+ memset(empty_zero_page, 0, PAGE_SIZE);
++blkif_t *tap_alloc_blkif(domid_t domid);
++void tap_blkif_free(blkif_t *blkif);
++void tap_blkif_kmem_cache_free(blkif_t *blkif);
++int tap_blkif_map(blkif_t *blkif, unsigned long shared_page,
++ unsigned int evtchn);
++void tap_blkif_unmap(blkif_t *blkif);
+
-+ reservedpages = 0;
++#define blkif_get(_b) (atomic_inc(&(_b)->refcnt))
++#define blkif_put(_b) \
++ do { \
++ if (atomic_dec_and_test(&(_b)->refcnt)) \
++ wake_up(&(_b)->waiting_to_free);\
++ } while (0)
+
-+ /* this will put all low memory onto the freelists */
-+#ifdef CONFIG_NUMA
-+ totalram_pages = numa_free_all_bootmem();
-+#else
-+ totalram_pages = free_all_bootmem();
-+#endif
-+ /* XEN: init and count pages outside initial allocation. */
-+ for (pfn = xen_start_info->nr_pages; pfn < max_pfn; pfn++) {
-+ ClearPageReserved(pfn_to_page(pfn));
-+ init_page_count(pfn_to_page(pfn));
-+ totalram_pages++;
-+ }
-+ reservedpages = end_pfn - totalram_pages - e820_hole_size(0, end_pfn);
+
-+ after_bootmem = 1;
++struct phys_req {
++ unsigned short dev;
++ unsigned short nr_sects;
++ struct block_device *bdev;
++ blkif_sector_t sector_number;
++};
+
-+ codesize = (unsigned long) &_etext - (unsigned long) &_text;
-+ datasize = (unsigned long) &_edata - (unsigned long) &_etext;
-+ initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin;
++void tap_blkif_interface_init(void);
+
-+ /* Register memory areas for /proc/kcore */
-+ kclist_add(&kcore_mem, __va(0), max_low_pfn << PAGE_SHIFT);
-+ kclist_add(&kcore_vmalloc, (void *)VMALLOC_START,
-+ VMALLOC_END-VMALLOC_START);
-+ kclist_add(&kcore_kernel, &_stext, _end - _stext);
-+ kclist_add(&kcore_modules, (void *)MODULES_VADDR, MODULES_LEN);
-+ kclist_add(&kcore_vsyscall, (void *)VSYSCALL_START,
-+ VSYSCALL_END - VSYSCALL_START);
++void tap_blkif_xenbus_init(void);
+
-+ printk("Memory: %luk/%luk available (%ldk kernel code, %ldk reserved, %ldk data, %ldk init)\n",
-+ (unsigned long) nr_free_pages() << (PAGE_SHIFT-10),
-+ end_pfn << (PAGE_SHIFT-10),
-+ codesize >> 10,
-+ reservedpages << (PAGE_SHIFT-10),
-+ datasize >> 10,
-+ initsize >> 10);
++irqreturn_t tap_blkif_be_int(int irq, void *dev_id, struct pt_regs *regs);
++int tap_blkif_schedule(void *arg);
+
-+#ifndef CONFIG_XEN
-+#ifdef CONFIG_SMP
-+ /*
-+ * Sync boot_level4_pgt mappings with the init_level4_pgt
-+ * except for the low identity mappings which are already zapped
-+ * in init_level4_pgt. This sync-up is essential for AP's bringup
-+ */
-+ memcpy(boot_level4_pgt+1, init_level4_pgt+1, (PTRS_PER_PGD-1)*sizeof(pgd_t));
-+#endif
-+#endif
-+}
++int dom_to_devid(domid_t domid, int xenbus_id, blkif_t *blkif);
++void signal_tapdisk(int idx);
+
-+void free_init_pages(char *what, unsigned long begin, unsigned long end)
-+{
-+ unsigned long addr;
++#endif /* __BLKIF__BACKEND__COMMON_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blktap/interface.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blktap/interface.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,181 @@
++/******************************************************************************
++ * drivers/xen/blktap/interface.c
++ *
++ * Block-device interface management.
++ *
++ * Copyright (c) 2004, Keir Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+
-+ if (begin >= end)
-+ return;
++ */
+
-+ printk(KERN_INFO "Freeing %s: %ldk freed\n", what, (end - begin) >> 10);
-+ for (addr = begin; addr < end; addr += PAGE_SIZE) {
-+ ClearPageReserved(virt_to_page(addr));
-+ init_page_count(virt_to_page(addr));
-+ memset((void *)(addr & ~(PAGE_SIZE-1)),
-+ POISON_FREE_INITMEM, PAGE_SIZE);
-+ if (addr >= __START_KERNEL_map) {
-+ /* make_readonly() reports all kernel addresses. */
-+ __make_page_writable(__va(__pa(addr)));
-+ if (HYPERVISOR_update_va_mapping(addr, __pte(0), 0)) {
-+ pgd_t *pgd = pgd_offset_k(addr);
-+ pud_t *pud = pud_offset(pgd, addr);
-+ pmd_t *pmd = pmd_offset(pud, addr);
-+ pte_t *pte = pte_offset_kernel(pmd, addr);
++#include "common.h"
++#include <xen/evtchn.h>
+
-+ xen_l1_entry_update(pte, __pte(0)); /* fallback */
-+ }
-+ }
-+ free_page(addr);
-+ totalram_pages++;
-+ }
-+}
++static kmem_cache_t *blkif_cachep;
+
-+void free_initmem(void)
++blkif_t *tap_alloc_blkif(domid_t domid)
+{
-+ memset(__initdata_begin, POISON_FREE_INITDATA,
-+ __initdata_end - __initdata_begin);
-+ free_init_pages("unused kernel memory",
-+ (unsigned long)(&__init_begin),
-+ (unsigned long)(&__init_end));
-+}
++ blkif_t *blkif;
+
-+#ifdef CONFIG_DEBUG_RODATA
++ blkif = kmem_cache_alloc(blkif_cachep, GFP_KERNEL);
++ if (!blkif)
++ return ERR_PTR(-ENOMEM);
+
-+void mark_rodata_ro(void)
++ memset(blkif, 0, sizeof(*blkif));
++ blkif->domid = domid;
++ spin_lock_init(&blkif->blk_ring_lock);
++ atomic_set(&blkif->refcnt, 1);
++ init_waitqueue_head(&blkif->wq);
++ blkif->st_print = jiffies;
++ init_waitqueue_head(&blkif->waiting_to_free);
++
++ return blkif;
++}
++
++static int map_frontend_page(blkif_t *blkif, unsigned long shared_page)
+{
-+ unsigned long addr = (unsigned long)__start_rodata;
++ struct gnttab_map_grant_ref op;
+
-+ for (; addr < (unsigned long)__end_rodata; addr += PAGE_SIZE)
-+ change_page_attr_addr(addr, 1, PAGE_KERNEL_RO);
++ gnttab_set_map_op(&op, (unsigned long)blkif->blk_ring_area->addr,
++ GNTMAP_host_map, shared_page, blkif->domid);
+
-+ printk ("Write protecting the kernel read-only data: %luk\n",
-+ (__end_rodata - __start_rodata) >> 10);
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
+
-+ /*
-+ * change_page_attr_addr() requires a global_flush_tlb() call after it.
-+ * We do this after the printk so that if something went wrong in the
-+ * change, the printk gets out at least to give a better debug hint
-+ * of who is the culprit.
-+ */
-+ global_flush_tlb();
-+}
-+#endif
++ if (op.status) {
++ DPRINTK(" Grant table operation failure !\n");
++ return op.status;
++ }
+
-+#ifdef CONFIG_BLK_DEV_INITRD
-+void free_initrd_mem(unsigned long start, unsigned long end)
-+{
-+ free_init_pages("initrd memory", start, end);
-+}
-+#endif
++ blkif->shmem_ref = shared_page;
++ blkif->shmem_handle = op.handle;
+
-+void __init reserve_bootmem_generic(unsigned long phys, unsigned len)
-+{
-+ /* Should check here against the e820 map to avoid double free */
-+#ifdef CONFIG_NUMA
-+ int nid = phys_to_nid(phys);
-+ reserve_bootmem_node(NODE_DATA(nid), phys, len);
-+#else
-+ reserve_bootmem(phys, len);
-+#endif
-+ if (phys+len <= MAX_DMA_PFN*PAGE_SIZE)
-+ dma_reserve += len / PAGE_SIZE;
++ return 0;
+}
+
-+int kern_addr_valid(unsigned long addr)
-+{
-+ unsigned long above = ((long)addr) >> __VIRTUAL_MASK_SHIFT;
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte;
++static void unmap_frontend_page(blkif_t *blkif)
++{
++ struct gnttab_unmap_grant_ref op;
+
-+ if (above != 0 && above != -1UL)
-+ return 0;
-+
-+ pgd = pgd_offset_k(addr);
-+ if (pgd_none(*pgd))
-+ return 0;
++ gnttab_set_unmap_op(&op, (unsigned long)blkif->blk_ring_area->addr,
++ GNTMAP_host_map, blkif->shmem_handle);
+
-+ pud = pud_offset(pgd, addr);
-+ if (pud_none(*pud))
-+ return 0;
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
++}
+
-+ pmd = pmd_offset(pud, addr);
-+ if (pmd_none(*pmd))
-+ return 0;
-+ if (pmd_large(*pmd))
-+ return pfn_valid(pmd_pfn(*pmd));
++int tap_blkif_map(blkif_t *blkif, unsigned long shared_page,
++ unsigned int evtchn)
++{
++ int err;
+
-+ pte = pte_offset_kernel(pmd, addr);
-+ if (pte_none(*pte))
++ /* Already connected through? */
++ if (blkif->irq)
+ return 0;
-+ return pfn_valid(pte_pfn(*pte));
-+}
+
-+#ifdef CONFIG_SYSCTL
-+#include <linux/sysctl.h>
++ if ( (blkif->blk_ring_area = alloc_vm_area(PAGE_SIZE)) == NULL )
++ return -ENOMEM;
+
-+extern int exception_trace, page_fault_trace;
++ err = map_frontend_page(blkif, shared_page);
++ if (err) {
++ free_vm_area(blkif->blk_ring_area);
++ return err;
++ }
+
-+static ctl_table debug_table2[] = {
-+ { 99, "exception-trace", &exception_trace, sizeof(int), 0644, NULL,
-+ proc_dointvec },
-+ { 0, }
-+};
++ switch (blkif->blk_protocol) {
++ case BLKIF_PROTOCOL_NATIVE:
++ {
++ blkif_sring_t *sring;
++ sring = (blkif_sring_t *)blkif->blk_ring_area->addr;
++ BACK_RING_INIT(&blkif->blk_rings.native, sring, PAGE_SIZE);
++ break;
++ }
++ case BLKIF_PROTOCOL_X86_32:
++ {
++ blkif_x86_32_sring_t *sring_x86_32;
++ sring_x86_32 = (blkif_x86_32_sring_t *)blkif->blk_ring_area->addr;
++ BACK_RING_INIT(&blkif->blk_rings.x86_32, sring_x86_32, PAGE_SIZE);
++ break;
++ }
++ case BLKIF_PROTOCOL_X86_64:
++ {
++ blkif_x86_64_sring_t *sring_x86_64;
++ sring_x86_64 = (blkif_x86_64_sring_t *)blkif->blk_ring_area->addr;
++ BACK_RING_INIT(&blkif->blk_rings.x86_64, sring_x86_64, PAGE_SIZE);
++ break;
++ }
++ default:
++ BUG();
++ }
+
-+static ctl_table debug_root_table2[] = {
-+ { .ctl_name = CTL_DEBUG, .procname = "debug", .mode = 0555,
-+ .child = debug_table2 },
-+ { 0 },
-+};
++ err = bind_interdomain_evtchn_to_irqhandler(
++ blkif->domid, evtchn, tap_blkif_be_int,
++ 0, "blkif-backend", blkif);
++ if (err < 0) {
++ unmap_frontend_page(blkif);
++ free_vm_area(blkif->blk_ring_area);
++ blkif->blk_rings.common.sring = NULL;
++ return err;
++ }
++ blkif->irq = err;
+
-+static __init int x8664_sysctl_init(void)
-+{
-+ register_sysctl_table(debug_root_table2, 1);
+ return 0;
+}
-+__initcall(x8664_sysctl_init);
-+#endif
-+
-+/* A pseudo VMAs to allow ptrace access for the vsyscall page. This only
-+ covers the 64bit vsyscall page now. 32bit has a real VMA now and does
-+ not need special handling anymore. */
+
-+static struct vm_area_struct gate_vma = {
-+ .vm_start = VSYSCALL_START,
-+ .vm_end = VSYSCALL_END,
-+ .vm_page_prot = PAGE_READONLY
-+};
-+
-+struct vm_area_struct *get_gate_vma(struct task_struct *tsk)
++void tap_blkif_unmap(blkif_t *blkif)
+{
-+#ifdef CONFIG_IA32_EMULATION
-+ if (test_tsk_thread_flag(tsk, TIF_IA32))
-+ return NULL;
-+#endif
-+ return &gate_vma;
++ if (blkif->irq) {
++ unbind_from_irqhandler(blkif->irq, blkif);
++ blkif->irq = 0;
++ }
++ if (blkif->blk_rings.common.sring) {
++ unmap_frontend_page(blkif);
++ free_vm_area(blkif->blk_ring_area);
++ blkif->blk_rings.common.sring = NULL;
++ }
+}
+
-+int in_gate_area(struct task_struct *task, unsigned long addr)
++void tap_blkif_free(blkif_t *blkif)
+{
-+ struct vm_area_struct *vma = get_gate_vma(task);
-+ if (!vma)
-+ return 0;
-+ return (addr >= vma->vm_start) && (addr < vma->vm_end);
++ atomic_dec(&blkif->refcnt);
++ wait_event(blkif->waiting_to_free, atomic_read(&blkif->refcnt) == 0);
++ atomic_inc(&blkif->refcnt);
++
++ tap_blkif_unmap(blkif);
+}
+
-+/* Use this when you have no reliable task/vma, typically from interrupt
-+ * context. It is less reliable than using the task's vma and may give
-+ * false positives.
-+ */
-+int in_gate_area_no_task(unsigned long addr)
++void tap_blkif_kmem_cache_free(blkif_t *blkif)
+{
-+ return (addr >= VSYSCALL_START) && (addr < VSYSCALL_END);
++ if (!atomic_dec_and_test(&blkif->refcnt))
++ BUG();
++ kmem_cache_free(blkif_cachep, blkif);
+}
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/mm/Makefile tmp-linux-2.6-xen.patch/arch/x86_64/mm/Makefile
---- pristine-linux-2.6.18.2/arch/x86_64/mm/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/mm/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -9,3 +9,13 @@ obj-$(CONFIG_K8_NUMA) += k8topology.o
- obj-$(CONFIG_ACPI_NUMA) += srat.o
-
- hugetlbpage-y = ../../i386/mm/hugetlbpage.o
-+
-+ifdef CONFIG_XEN
-+include $(srctree)/scripts/Makefile.xen
-+
-+ioremap-y += ../../i386/mm/ioremap-xen.o
-+hypervisor-y += ../../i386/mm/hypervisor.o
-+obj-y += hypervisor.o
+
-+obj-y := $(call cherrypickxen, $(obj-y))
-+endif
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/mm/pageattr-xen.c tmp-linux-2.6-xen.patch/arch/x86_64/mm/pageattr-xen.c
---- pristine-linux-2.6.18.2/arch/x86_64/mm/pageattr-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/mm/pageattr-xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,433 @@
-+/*
-+ * Copyright 2002 Andi Kleen, SuSE Labs.
-+ * Thanks to Ben LaHaise for precious feedback.
-+ */
++void __init tap_blkif_interface_init(void)
++{
++ blkif_cachep = kmem_cache_create("blktapif_cache", sizeof(blkif_t),
++ 0, 0, NULL, NULL);
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/blktap/xenbus.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/blktap/xenbus.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,479 @@
++/* drivers/xen/blktap/xenbus.c
++ *
++ * Xenbus code for blktap
++ *
++ * Copyright (c) 2004-2005, Andrew Warfield and Julian Chesterfield
++ *
++ * Based on the blkback xenbus code:
++ *
++ * Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
++ * Copyright (C) 2005 XenSource Ltd
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+#include <linux/mm.h>
-+#include <linux/sched.h>
-+#include <linux/highmem.h>
++#include <stdarg.h>
+#include <linux/module.h>
-+#include <linux/slab.h>
-+#include <asm/uaccess.h>
-+#include <asm/processor.h>
-+#include <asm/tlbflush.h>
-+#include <asm/io.h>
-+
-+#ifdef CONFIG_XEN
-+#include <asm/pgalloc.h>
-+#include <asm/mmu_context.h>
++#include <linux/kthread.h>
++#include <xen/xenbus.h>
++#include "common.h"
+
-+LIST_HEAD(mm_unpinned);
-+DEFINE_SPINLOCK(mm_unpinned_lock);
+
-+static inline void mm_walk_set_prot(void *pt, pgprot_t flags)
++struct backend_info
+{
-+ struct page *page = virt_to_page(pt);
-+ unsigned long pfn = page_to_pfn(page);
-+ int rc;
++ struct xenbus_device *dev;
++ blkif_t *blkif;
++ struct xenbus_watch backend_watch;
++ int xenbus_id;
++ int group_added;
++};
+
-+ rc = HYPERVISOR_update_va_mapping(
-+ (unsigned long)__va(pfn << PAGE_SHIFT),
-+ pfn_pte(pfn, flags), 0);
-+ if (rc)
-+ BUG();
-+}
+
-+static void mm_walk(struct mm_struct *mm, pgprot_t flags)
++static void connect(struct backend_info *);
++static int connect_ring(struct backend_info *);
++static int blktap_remove(struct xenbus_device *dev);
++static int blktap_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id);
++static void tap_backend_changed(struct xenbus_watch *, const char **,
++ unsigned int);
++static void tap_frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state);
++
++static int strsep_len(const char *str, char c, unsigned int len)
+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte;
-+ int g,u,m;
++ unsigned int i;
+
-+ pgd = mm->pgd;
-+ /*
-+ * Cannot iterate up to USER_PTRS_PER_PGD as these pagetables may not
-+ * be the 'current' task's pagetables (e.g., current may be 32-bit,
-+ * but the pagetables may be for a 64-bit task).
-+ * Subtracting 1 from TASK_SIZE64 means the loop limit is correct
-+ * regardless of whether TASK_SIZE64 is a multiple of PGDIR_SIZE.
-+ */
-+ for (g = 0; g <= ((TASK_SIZE64-1) / PGDIR_SIZE); g++, pgd++) {
-+ if (pgd_none(*pgd))
-+ continue;
-+ pud = pud_offset(pgd, 0);
-+ if (PTRS_PER_PUD > 1) /* not folded */
-+ mm_walk_set_prot(pud,flags);
-+ for (u = 0; u < PTRS_PER_PUD; u++, pud++) {
-+ if (pud_none(*pud))
-+ continue;
-+ pmd = pmd_offset(pud, 0);
-+ if (PTRS_PER_PMD > 1) /* not folded */
-+ mm_walk_set_prot(pmd,flags);
-+ for (m = 0; m < PTRS_PER_PMD; m++, pmd++) {
-+ if (pmd_none(*pmd))
-+ continue;
-+ pte = pte_offset_kernel(pmd,0);
-+ mm_walk_set_prot(pte,flags);
-+ }
-+ }
-+ }
++ for (i = 0; str[i]; i++)
++ if (str[i] == c) {
++ if (len == 0)
++ return i;
++ len--;
++ }
++ return (len == 0) ? i : -ERANGE;
+}
+
-+void mm_pin(struct mm_struct *mm)
++static long get_id(const char *str)
+{
-+ if (xen_feature(XENFEAT_writable_page_tables))
-+ return;
++ int len,end;
++ const char *ptr;
++ char *tptr, num[10];
++
++ len = strsep_len(str, '/', 2);
++ end = strlen(str);
++ if ( (len < 0) || (end < 0) ) return -1;
++
++ ptr = str + len + 1;
++ strncpy(num,ptr,end - len);
++ tptr = num + (end - (len + 1));
++ *tptr = '\0';
++ DPRINTK("Get_id called for %s (%s)\n",str,num);
++
++ return simple_strtol(num, NULL, 10);
++}
+
-+ spin_lock(&mm->page_table_lock);
++static int blktap_name(blkif_t *blkif, char *buf)
++{
++ char *devpath, *devname;
++ struct xenbus_device *dev = blkif->be->dev;
+
-+ mm_walk(mm, PAGE_KERNEL_RO);
-+ if (HYPERVISOR_update_va_mapping(
-+ (unsigned long)mm->pgd,
-+ pfn_pte(virt_to_phys(mm->pgd)>>PAGE_SHIFT, PAGE_KERNEL_RO),
-+ UVMF_TLB_FLUSH))
-+ BUG();
-+ if (HYPERVISOR_update_va_mapping(
-+ (unsigned long)__user_pgd(mm->pgd),
-+ pfn_pte(virt_to_phys(__user_pgd(mm->pgd))>>PAGE_SHIFT,
-+ PAGE_KERNEL_RO),
-+ UVMF_TLB_FLUSH))
-+ BUG();
-+ xen_pgd_pin(__pa(mm->pgd)); /* kernel */
-+ xen_pgd_pin(__pa(__user_pgd(mm->pgd))); /* user */
-+ mm->context.pinned = 1;
-+ spin_lock(&mm_unpinned_lock);
-+ list_del(&mm->context.unpinned);
-+ spin_unlock(&mm_unpinned_lock);
++ devpath = xenbus_read(XBT_NIL, dev->nodename, "dev", NULL);
++ if (IS_ERR(devpath))
++ return PTR_ERR(devpath);
++
++ if ((devname = strstr(devpath, "/dev/")) != NULL)
++ devname += strlen("/dev/");
++ else
++ devname = devpath;
+
-+ spin_unlock(&mm->page_table_lock);
++ snprintf(buf, TASK_COMM_LEN, "blktap.%d.%s", blkif->domid, devname);
++ kfree(devpath);
++
++ return 0;
+}
+
-+void mm_unpin(struct mm_struct *mm)
-+{
-+ if (xen_feature(XENFEAT_writable_page_tables))
-+ return;
++/****************************************************************
++ * sysfs interface for I/O requests of blktap device
++ */
+
-+ spin_lock(&mm->page_table_lock);
++#define VBD_SHOW(name, format, args...) \
++ static ssize_t show_##name(struct device *_dev, \
++ struct device_attribute *attr, \
++ char *buf) \
++ { \
++ struct xenbus_device *dev = to_xenbus_device(_dev); \
++ struct backend_info *be = dev->dev.driver_data; \
++ \
++ return sprintf(buf, format, ##args); \
++ } \
++ static DEVICE_ATTR(name, S_IRUGO, show_##name, NULL)
+
-+ xen_pgd_unpin(__pa(mm->pgd));
-+ xen_pgd_unpin(__pa(__user_pgd(mm->pgd)));
-+ if (HYPERVISOR_update_va_mapping(
-+ (unsigned long)mm->pgd,
-+ pfn_pte(virt_to_phys(mm->pgd)>>PAGE_SHIFT, PAGE_KERNEL), 0))
-+ BUG();
-+ if (HYPERVISOR_update_va_mapping(
-+ (unsigned long)__user_pgd(mm->pgd),
-+ pfn_pte(virt_to_phys(__user_pgd(mm->pgd))>>PAGE_SHIFT,
-+ PAGE_KERNEL), 0))
-+ BUG();
-+ mm_walk(mm, PAGE_KERNEL);
-+ xen_tlb_flush();
-+ mm->context.pinned = 0;
-+ spin_lock(&mm_unpinned_lock);
-+ list_add(&mm->context.unpinned, &mm_unpinned);
-+ spin_unlock(&mm_unpinned_lock);
++VBD_SHOW(oo_req, "%d\n", be->blkif->st_oo_req);
++VBD_SHOW(rd_req, "%d\n", be->blkif->st_rd_req);
++VBD_SHOW(wr_req, "%d\n", be->blkif->st_wr_req);
++VBD_SHOW(rd_sect, "%d\n", be->blkif->st_rd_sect);
++VBD_SHOW(wr_sect, "%d\n", be->blkif->st_wr_sect);
+
-+ spin_unlock(&mm->page_table_lock);
-+}
++static struct attribute *tapstat_attrs[] = {
++ &dev_attr_oo_req.attr,
++ &dev_attr_rd_req.attr,
++ &dev_attr_wr_req.attr,
++ &dev_attr_rd_sect.attr,
++ &dev_attr_wr_sect.attr,
++ NULL
++};
+
-+void mm_pin_all(void)
-+{
-+ if (xen_feature(XENFEAT_writable_page_tables))
-+ return;
++static struct attribute_group tapstat_group = {
++ .name = "statistics",
++ .attrs = tapstat_attrs,
++};
+
-+ /*
-+ * Allow uninterrupted access to the mm_unpinned list. We don't
-+ * actually take the mm_unpinned_lock as it is taken inside mm_pin().
-+ * All other CPUs must be at a safe point (e.g., in stop_machine
-+ * or offlined entirely).
-+ */
-+ preempt_disable();
-+ while (!list_empty(&mm_unpinned))
-+ mm_pin(list_entry(mm_unpinned.next, struct mm_struct,
-+ context.unpinned));
-+ preempt_enable();
++int xentap_sysfs_addif(struct xenbus_device *dev)
++{
++ int err;
++ struct backend_info *be = dev->dev.driver_data;
++ err = sysfs_create_group(&dev->dev.kobj, &tapstat_group);
++ if (!err)
++ be->group_added = 1;
++ return err;
+}
+
-+void _arch_dup_mmap(struct mm_struct *mm)
++void xentap_sysfs_delif(struct xenbus_device *dev)
+{
-+ if (!mm->context.pinned)
-+ mm_pin(mm);
++ struct backend_info *be = dev->dev.driver_data;
++ sysfs_remove_group(&dev->dev.kobj, &tapstat_group);
++ be->group_added = 0;
+}
+
-+void _arch_exit_mmap(struct mm_struct *mm)
++static int blktap_remove(struct xenbus_device *dev)
+{
-+ struct task_struct *tsk = current;
++ struct backend_info *be = dev->dev.driver_data;
+
-+ task_lock(tsk);
++ if (be->group_added)
++ xentap_sysfs_delif(be->dev);
++ if (be->backend_watch.node) {
++ unregister_xenbus_watch(&be->backend_watch);
++ kfree(be->backend_watch.node);
++ be->backend_watch.node = NULL;
++ }
++ if (be->blkif) {
++ if (be->blkif->xenblkd)
++ kthread_stop(be->blkif->xenblkd);
++ signal_tapdisk(be->blkif->dev_num);
++ tap_blkif_free(be->blkif);
++ tap_blkif_kmem_cache_free(be->blkif);
++ be->blkif = NULL;
++ }
++ kfree(be);
++ dev->dev.driver_data = NULL;
++ return 0;
++}
+
-+ /*
-+ * We aggressively remove defunct pgd from cr3. We execute unmap_vmas()
-+ * *much* faster this way, as no tlb flushes means bigger wrpt batches.
-+ */
-+ if (tsk->active_mm == mm) {
-+ tsk->active_mm = &init_mm;
-+ atomic_inc(&init_mm.mm_count);
++static void tap_update_blkif_status(blkif_t *blkif)
++{
++ int err;
++ char name[TASK_COMM_LEN];
+
-+ switch_mm(mm, &init_mm, tsk);
++ /* Not ready to connect? */
++ if(!blkif->irq || !blkif->sectors) {
++ return;
++ }
+
-+ atomic_dec(&mm->mm_count);
-+ BUG_ON(atomic_read(&mm->mm_count) == 0);
++ /* Already connected? */
++ if (blkif->be->dev->state == XenbusStateConnected)
++ return;
++
++ /* Attempt to connect: exit if we fail to. */
++ connect(blkif->be);
++ if (blkif->be->dev->state != XenbusStateConnected)
++ return;
++
++ err = blktap_name(blkif, name);
++ if (err) {
++ xenbus_dev_error(blkif->be->dev, err, "get blktap dev name");
++ return;
+ }
+
-+ task_unlock(tsk);
++ if (!blkif->be->group_added) {
++ err = xentap_sysfs_addif(blkif->be->dev);
++ if (err) {
++ xenbus_dev_fatal(blkif->be->dev, err,
++ "creating sysfs entries");
++ return;
++ }
++ }
+
-+ if ( mm->context.pinned && (atomic_read(&mm->mm_count) == 1) &&
-+ !mm->context.has_foreign_mappings )
-+ mm_unpin(mm);
++ blkif->xenblkd = kthread_run(tap_blkif_schedule, blkif, name);
++ if (IS_ERR(blkif->xenblkd)) {
++ err = PTR_ERR(blkif->xenblkd);
++ blkif->xenblkd = NULL;
++ xenbus_dev_fatal(blkif->be->dev, err, "start xenblkd");
++ WPRINTK("Error starting thread\n");
++ }
+}
+
-+struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
++/**
++ * Entry point to this code when a new device is created. Allocate
++ * the basic structures, and watch the store waiting for the
++ * user-space program to tell us the physical device info. Switch to
++ * InitWait.
++ */
++static int blktap_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
+{
-+ struct page *pte;
-+
-+ pte = alloc_pages(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO, 0);
-+ if (pte) {
-+ SetPageForeign(pte, pte_free);
-+ init_page_count(pte);
++ int err;
++ struct backend_info *be = kzalloc(sizeof(struct backend_info),
++ GFP_KERNEL);
++ if (!be) {
++ xenbus_dev_fatal(dev, -ENOMEM,
++ "allocating backend structure");
++ return -ENOMEM;
+ }
-+ return pte;
-+}
+
-+void pte_free(struct page *pte)
-+{
-+ unsigned long va = (unsigned long)__va(page_to_pfn(pte)<<PAGE_SHIFT);
++ be->dev = dev;
++ dev->dev.driver_data = be;
++ be->xenbus_id = get_id(dev->nodename);
+
-+ if (!pte_write(*virt_to_ptep(va)))
-+ if (HYPERVISOR_update_va_mapping(
-+ va, pfn_pte(page_to_pfn(pte), PAGE_KERNEL), 0))
-+ BUG();
++ be->blkif = tap_alloc_blkif(dev->otherend_id);
++ if (IS_ERR(be->blkif)) {
++ err = PTR_ERR(be->blkif);
++ be->blkif = NULL;
++ xenbus_dev_fatal(dev, err, "creating block interface");
++ goto fail;
++ }
+
-+ ClearPageForeign(pte);
-+ init_page_count(pte);
++ /* setup back pointer */
++ be->blkif->be = be;
++ be->blkif->sectors = 0;
+
-+ __free_page(pte);
++ /* set a watch on disk info, waiting for userspace to update details*/
++ err = xenbus_watch_path2(dev, dev->nodename, "info",
++ &be->backend_watch, tap_backend_changed);
++ if (err)
++ goto fail;
++
++ err = xenbus_switch_state(dev, XenbusStateInitWait);
++ if (err)
++ goto fail;
++ return 0;
++
++fail:
++ DPRINTK("blktap probe failed\n");
++ blktap_remove(dev);
++ return err;
+}
-+#endif /* CONFIG_XEN */
+
-+static inline pte_t *lookup_address(unsigned long address)
-+{
-+ pgd_t *pgd = pgd_offset_k(address);
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte;
-+ if (pgd_none(*pgd))
-+ return NULL;
-+ pud = pud_offset(pgd, address);
-+ if (!pud_present(*pud))
-+ return NULL;
-+ pmd = pmd_offset(pud, address);
-+ if (!pmd_present(*pmd))
-+ return NULL;
-+ if (pmd_large(*pmd))
-+ return (pte_t *)pmd;
-+ pte = pte_offset_kernel(pmd, address);
-+ if (pte && !pte_present(*pte))
-+ pte = NULL;
-+ return pte;
-+}
+
-+static struct page *split_large_page(unsigned long address, pgprot_t prot,
-+ pgprot_t ref_prot)
-+{
-+ int i;
-+ unsigned long addr;
-+ struct page *base = alloc_pages(GFP_KERNEL, 0);
-+ pte_t *pbase;
-+ if (!base)
-+ return NULL;
-+ /*
-+ * page_private is used to track the number of entries in
-+ * the page table page have non standard attributes.
++/**
++ * Callback received when the user space code has placed the device
++ * information in xenstore.
++ */
++static void tap_backend_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ int err;
++ unsigned long info;
++ struct backend_info *be
++ = container_of(watch, struct backend_info, backend_watch);
++ struct xenbus_device *dev = be->dev;
++
++ /**
++ * Check to see whether userspace code has opened the image
++ * and written sector
++ * and disk info to xenstore
+ */
-+ SetPagePrivate(base);
-+ page_private(base) = 0;
-+
-+ address = __pa(address);
-+ addr = address & LARGE_PAGE_MASK;
-+ pbase = (pte_t *)page_address(base);
-+ for (i = 0; i < PTRS_PER_PTE; i++, addr += PAGE_SIZE) {
-+ pbase[i] = pfn_pte(addr >> PAGE_SHIFT,
-+ addr == address ? prot : ref_prot);
++ err = xenbus_gather(XBT_NIL, dev->nodename, "info", "%lu", &info,
++ NULL);
++ if (XENBUS_EXIST_ERR(err))
++ return;
++ if (err) {
++ xenbus_dev_error(dev, err, "getting info");
++ return;
+ }
-+ return base;
-+}
-+
-+
-+static void flush_kernel_map(void *address)
-+{
-+ if (0 && address && cpu_has_clflush) {
-+ /* is this worth it? */
-+ int i;
-+ for (i = 0; i < PAGE_SIZE; i += boot_cpu_data.x86_clflush_size)
-+ asm volatile("clflush (%0)" :: "r" (address + i));
-+ } else
-+ asm volatile("wbinvd":::"memory");
-+ if (address)
-+ __flush_tlb_one(address);
-+ else
-+ __flush_tlb_all();
-+}
+
++ DPRINTK("Userspace update on disk info, %lu\n",info);
+
-+static inline void flush_map(unsigned long address)
-+{
-+ on_each_cpu(flush_kernel_map, (void *)address, 1, 1);
-+}
++ err = xenbus_gather(XBT_NIL, dev->nodename, "sectors", "%llu",
++ &be->blkif->sectors, NULL);
+
-+static struct page *deferred_pages; /* protected by init_mm.mmap_sem */
++ /* Associate tap dev with domid*/
++ be->blkif->dev_num = dom_to_devid(be->blkif->domid, be->xenbus_id,
++ be->blkif);
++ DPRINTK("Thread started for domid [%d], connecting disk\n",
++ be->blkif->dev_num);
+
-+static inline void save_page(struct page *fpage)
-+{
-+ fpage->lru.next = (struct list_head *)deferred_pages;
-+ deferred_pages = fpage;
++ tap_update_blkif_status(be->blkif);
+}
+
-+/*
-+ * No more special protections in this 2/4MB area - revert to a
-+ * large page again.
++/**
++ * Callback received when the frontend's state changes.
+ */
-+static void revert_page(unsigned long address, pgprot_t ref_prot)
++static void tap_frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state)
+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t large_pte;
++ struct backend_info *be = dev->dev.driver_data;
++ int err;
+
-+ pgd = pgd_offset_k(address);
-+ BUG_ON(pgd_none(*pgd));
-+ pud = pud_offset(pgd,address);
-+ BUG_ON(pud_none(*pud));
-+ pmd = pmd_offset(pud, address);
-+ BUG_ON(pmd_val(*pmd) & _PAGE_PSE);
-+ pgprot_val(ref_prot) |= _PAGE_PSE;
-+ large_pte = mk_pte_phys(__pa(address) & LARGE_PAGE_MASK, ref_prot);
-+ set_pte((pte_t *)pmd, large_pte);
-+}
++ DPRINTK("\n");
+
-+static int
-+__change_page_attr(unsigned long address, unsigned long pfn, pgprot_t prot,
-+ pgprot_t ref_prot)
-+{
-+ pte_t *kpte;
-+ struct page *kpte_page;
-+ unsigned kpte_flags;
-+ pgprot_t ref_prot2;
-+ kpte = lookup_address(address);
-+ if (!kpte) return 0;
-+ kpte_page = virt_to_page(((unsigned long)kpte) & PAGE_MASK);
-+ kpte_flags = pte_val(*kpte);
-+ if (pgprot_val(prot) != pgprot_val(ref_prot)) {
-+ if ((kpte_flags & _PAGE_PSE) == 0) {
-+ set_pte(kpte, pfn_pte(pfn, prot));
-+ } else {
-+ /*
-+ * split_large_page will take the reference for this
-+ * change_page_attr on the split page.
-+ */
++ switch (frontend_state) {
++ case XenbusStateInitialising:
++ if (dev->state == XenbusStateClosed) {
++ printk(KERN_INFO "%s: %s: prepare for reconnect\n",
++ __FUNCTION__, dev->nodename);
++ xenbus_switch_state(dev, XenbusStateInitWait);
++ }
++ break;
+
-+ struct page *split;
-+ ref_prot2 = __pgprot(pgprot_val(pte_pgprot(*lookup_address(address))) & ~(1<<_PAGE_BIT_PSE));
++ case XenbusStateInitialised:
++ case XenbusStateConnected:
++ /* Ensure we connect even when two watches fire in
++ close successsion and we miss the intermediate value
++ of frontend_state. */
++ if (dev->state == XenbusStateConnected)
++ break;
+
-+ split = split_large_page(address, prot, ref_prot2);
-+ if (!split)
-+ return -ENOMEM;
-+ set_pte(kpte,mk_pte(split, ref_prot2));
-+ kpte_page = split;
-+ }
-+ page_private(kpte_page)++;
-+ } else if ((kpte_flags & _PAGE_PSE) == 0) {
-+ set_pte(kpte, pfn_pte(pfn, ref_prot));
-+ BUG_ON(page_private(kpte_page) == 0);
-+ page_private(kpte_page)--;
-+ } else
-+ BUG();
++ err = connect_ring(be);
++ if (err)
++ break;
++ tap_update_blkif_status(be->blkif);
++ break;
+
-+ /* on x86-64 the direct mapping set at boot is not using 4k pages */
-+ /*
-+ * ..., but the XEN guest kernels (currently) do:
-+ * If the pte was reserved, it means it was created at boot
-+ * time (not via split_large_page) and in turn we must not
-+ * replace it with a large page.
-+ */
-+#ifndef CONFIG_XEN
-+ BUG_ON(PageReserved(kpte_page));
-+#else
-+ if (PageReserved(kpte_page))
-+ return 0;
-+#endif
++ case XenbusStateClosing:
++ if (be->blkif->xenblkd) {
++ kthread_stop(be->blkif->xenblkd);
++ be->blkif->xenblkd = NULL;
++ }
++ tap_blkif_free(be->blkif);
++ xenbus_switch_state(dev, XenbusStateClosing);
++ break;
+
-+ if (page_private(kpte_page) == 0) {
-+ save_page(kpte_page);
-+ revert_page(address, ref_prot);
++ case XenbusStateClosed:
++ xenbus_switch_state(dev, XenbusStateClosed);
++ if (xenbus_dev_is_online(dev))
++ break;
++ /* fall through if not online */
++ case XenbusStateUnknown:
++ device_unregister(&dev->dev);
++ break;
++
++ default:
++ xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
++ frontend_state);
++ break;
+ }
-+ return 0;
-+}
++}
+
-+/*
-+ * Change the page attributes of an page in the linear mapping.
-+ *
-+ * This should be used when a page is mapped with a different caching policy
-+ * than write-back somewhere - some CPUs do not like it when mappings with
-+ * different caching policies exist. This changes the page attributes of the
-+ * in kernel linear mapping too.
-+ *
-+ * The caller needs to ensure that there are no conflicting mappings elsewhere.
-+ * This function only deals with the kernel linear map.
-+ *
-+ * Caller must call global_flush_tlb() after this.
++
++/**
++ * Switch to Connected state.
+ */
-+int change_page_attr_addr(unsigned long address, int numpages, pgprot_t prot)
++static void connect(struct backend_info *be)
+{
-+ int err = 0;
-+ int i;
++ int err;
+
-+ down_write(&init_mm.mmap_sem);
-+ for (i = 0; i < numpages; i++, address += PAGE_SIZE) {
-+ unsigned long pfn = __pa(address) >> PAGE_SHIFT;
++ struct xenbus_device *dev = be->dev;
+
-+ err = __change_page_attr(address, pfn, prot, PAGE_KERNEL);
-+ if (err)
-+ break;
-+ /* Handle kernel mapping too which aliases part of the
-+ * lowmem */
-+ if (__pa(address) < KERNEL_TEXT_SIZE) {
-+ unsigned long addr2;
-+ pgprot_t prot2 = prot;
-+ addr2 = __START_KERNEL_map + __pa(address);
-+ pgprot_val(prot2) &= ~_PAGE_NX;
-+ err = __change_page_attr(addr2, pfn, prot2, PAGE_KERNEL_EXEC);
-+ }
-+ }
-+ up_write(&init_mm.mmap_sem);
-+ return err;
-+}
++ err = xenbus_switch_state(dev, XenbusStateConnected);
++ if (err)
++ xenbus_dev_fatal(dev, err, "switching to Connected state",
++ dev->nodename);
+
-+/* Don't call this for MMIO areas that may not have a mem_map entry */
-+int change_page_attr(struct page *page, int numpages, pgprot_t prot)
-+{
-+ unsigned long addr = (unsigned long)page_address(page);
-+ return change_page_attr_addr(addr, numpages, prot);
++ return;
+}
+
-+void global_flush_tlb(void)
-+{
-+ struct page *dpage;
-+
-+ down_read(&init_mm.mmap_sem);
-+ dpage = xchg(&deferred_pages, NULL);
-+ up_read(&init_mm.mmap_sem);
-+
-+ flush_map((dpage && !dpage->lru.next) ? (unsigned long)page_address(dpage) : 0);
-+ while (dpage) {
-+ struct page *tmp = dpage;
-+ dpage = (struct page *)dpage->lru.next;
-+ ClearPagePrivate(tmp);
-+ __free_page(tmp);
-+ }
-+}
-+
-+EXPORT_SYMBOL(change_page_attr);
-+EXPORT_SYMBOL(global_flush_tlb);
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/oprofile/Makefile tmp-linux-2.6-xen.patch/arch/x86_64/oprofile/Makefile
---- pristine-linux-2.6.18.2/arch/x86_64/oprofile/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/oprofile/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -11,9 +11,15 @@ DRIVER_OBJS = $(addprefix ../../../drive
- oprofilefs.o oprofile_stats.o \
- timer_int.o )
-
-+ifdef CONFIG_XEN
-+XENOPROF_COMMON_OBJS = $(addprefix ../../../drivers/xen/xenoprof/, \
-+ xenoprofile.o)
-+OPROFILE-y := xenoprof.o
-+else
- OPROFILE-y := init.o backtrace.o
- OPROFILE-$(CONFIG_X86_LOCAL_APIC) += nmi_int.o op_model_athlon.o op_model_p4.o \
- op_model_ppro.o
- OPROFILE-$(CONFIG_X86_IO_APIC) += nmi_timer_int.o
--
--oprofile-y = $(DRIVER_OBJS) $(addprefix ../../i386/oprofile/, $(OPROFILE-y))
-+endif
-+oprofile-y = $(DRIVER_OBJS) $(XENOPROF_COMMON_OBJS) \
-+ $(addprefix ../../i386/oprofile/, $(OPROFILE-y))
-diff -Nurp pristine-linux-2.6.18.2/arch/x86_64/pci/Makefile tmp-linux-2.6-xen.patch/arch/x86_64/pci/Makefile
---- pristine-linux-2.6.18.2/arch/x86_64/pci/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/arch/x86_64/pci/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -15,11 +15,23 @@ obj-$(CONFIG_PCI_MMCONFIG) += mmconfig.o
-
- obj-$(CONFIG_NUMA) += k8-bus.o
-
-+# pcifront should be after mmconfig.o and direct.o as it should only
-+# take over if direct access to the PCI bus is unavailable
-+obj-$(CONFIG_XEN_PCIDEV_FRONTEND) += pcifront.o
-+
- direct-y += ../../i386/pci/direct.o
- acpi-y += ../../i386/pci/acpi.o
-+pcifront-y += ../../i386/pci/pcifront.o
- legacy-y += ../../i386/pci/legacy.o
- irq-y += ../../i386/pci/irq.o
- common-y += ../../i386/pci/common.o
- fixup-y += ../../i386/pci/fixup.o
- i386-y += ../../i386/pci/i386.o
- init-y += ../../i386/pci/init.o
-+
-+ifdef CONFIG_XEN
-+irq-y := ../../i386/pci/irq-xen.o
-+include $(srctree)/scripts/Makefile.xen
+
-+obj-y := $(call cherrypickxen, $(obj-y))
-+endif
-diff -Nurp pristine-linux-2.6.18.2/drivers/acpi/Kconfig tmp-linux-2.6-xen.patch/drivers/acpi/Kconfig
---- pristine-linux-2.6.18.2/drivers/acpi/Kconfig 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/acpi/Kconfig 2007-07-30 16:35:11.000000000 +0200
-@@ -45,7 +45,7 @@ if ACPI
-
- config ACPI_SLEEP
- bool "Sleep States"
-- depends on X86 && (!SMP || SUSPEND_SMP)
-+ depends on X86 && (!SMP || SUSPEND_SMP) && !XEN
- depends on PM
- default y
- ---help---
-@@ -305,6 +305,7 @@ config ACPI_SYSTEM
- config X86_PM_TIMER
- bool "Power Management Timer Support" if EMBEDDED
- depends on X86
-+ depends on !XEN
- default y
- help
- The Power Management Timer is available on all ACPI-capable,
-diff -Nurp pristine-linux-2.6.18.2/drivers/char/mem.c tmp-linux-2.6-xen.patch/drivers/char/mem.c
---- pristine-linux-2.6.18.2/drivers/char/mem.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/char/mem.c 2007-07-30 16:35:11.000000000 +0200
-@@ -101,6 +101,7 @@ static inline int valid_mmap_phys_addr_r
- }
- #endif
-
-+#ifndef ARCH_HAS_DEV_MEM
- /*
- * This funcion reads the *physical* memory. The f_pos points directly to the
- * memory location.
-@@ -223,6 +224,7 @@ static ssize_t write_mem(struct file * f
- *ppos += written;
- return written;
- }
-+#endif
-
- #ifndef __HAVE_PHYS_MEM_ACCESS_PROT
- static pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
-@@ -776,6 +778,7 @@ static int open_port(struct inode * inod
- #define open_kmem open_mem
- #define open_oldmem open_mem
-
-+#ifndef ARCH_HAS_DEV_MEM
- static const struct file_operations mem_fops = {
- .llseek = memory_lseek,
- .read = read_mem,
-@@ -783,6 +786,9 @@ static const struct file_operations mem_
- .mmap = mmap_mem,
- .open = open_mem,
- };
-+#else
-+extern const struct file_operations mem_fops;
-+#endif
-
- static const struct file_operations kmem_fops = {
- .llseek = memory_lseek,
-diff -Nurp pristine-linux-2.6.18.2/drivers/char/tpm/Kconfig tmp-linux-2.6-xen.patch/drivers/char/tpm/Kconfig
---- pristine-linux-2.6.18.2/drivers/char/tpm/Kconfig 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/char/tpm/Kconfig 2007-07-30 16:35:11.000000000 +0200
-@@ -58,5 +58,13 @@ config TCG_INFINEON
- Further information on this driver and the supported hardware
- can be found at http://www.prosec.rub.de/tpm
-
--endmenu
-+config TCG_XEN
-+ tristate "XEN TPM Interface"
-+ depends on TCG_TPM && XEN
-+ ---help---
-+ If you want to make TPM support available to a Xen user domain,
-+ say Yes and it will be accessible from within Linux.
-+ To compile this driver as a module, choose M here; the module
-+ will be called tpm_xenu.
-
-+endmenu
-diff -Nurp pristine-linux-2.6.18.2/drivers/char/tpm/Makefile tmp-linux-2.6-xen.patch/drivers/char/tpm/Makefile
---- pristine-linux-2.6.18.2/drivers/char/tpm/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/char/tpm/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -9,3 +9,5 @@ obj-$(CONFIG_TCG_TIS) += tpm_tis.o
- obj-$(CONFIG_TCG_NSC) += tpm_nsc.o
- obj-$(CONFIG_TCG_ATMEL) += tpm_atmel.o
- obj-$(CONFIG_TCG_INFINEON) += tpm_infineon.o
-+obj-$(CONFIG_TCG_XEN) += tpm_xenu.o
-+tpm_xenu-y = tpm_xen.o tpm_vtpm.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/char/tpm/tpm.h tmp-linux-2.6-xen.patch/drivers/char/tpm/tpm.h
---- pristine-linux-2.6.18.2/drivers/char/tpm/tpm.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/char/tpm/tpm.h 2007-07-30 16:35:11.000000000 +0200
-@@ -105,6 +105,9 @@ struct tpm_chip {
- struct dentry **bios_dir;
-
- struct list_head list;
-+#ifdef CONFIG_XEN
-+ void *priv;
-+#endif
- };
-
- #define to_tpm_chip(n) container_of(n, struct tpm_chip, vendor)
-@@ -121,6 +124,18 @@ static inline void tpm_write_index(int b
- outb(value & 0xFF, base+1);
- }
-
-+#ifdef CONFIG_XEN
-+static inline void *chip_get_private(const struct tpm_chip *chip)
++static int connect_ring(struct backend_info *be)
+{
-+ return chip->priv;
-+}
++ struct xenbus_device *dev = be->dev;
++ unsigned long ring_ref;
++ unsigned int evtchn;
++ char protocol[64];
++ int err;
+
-+static inline void chip_set_private(struct tpm_chip *chip, void *priv)
-+{
-+ chip->priv = priv;
-+}
-+#endif
++ DPRINTK("%s\n", dev->otherend);
+
- extern void tpm_get_timeouts(struct tpm_chip *);
- extern void tpm_gen_interrupt(struct tpm_chip *);
- extern void tpm_continue_selftest(struct tpm_chip *);
-diff -Nurp pristine-linux-2.6.18.2/drivers/char/tpm/tpm_vtpm.c tmp-linux-2.6-xen.patch/drivers/char/tpm/tpm_vtpm.c
---- pristine-linux-2.6.18.2/drivers/char/tpm/tpm_vtpm.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/char/tpm/tpm_vtpm.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,542 @@
-+/*
-+ * Copyright (C) 2006 IBM Corporation
-+ *
-+ * Authors:
-+ * Stefan Berger <stefanb at us.ibm.com>
-+ *
-+ * Generic device driver part for device drivers in a virtualized
-+ * environment.
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License as
-+ * published by the Free Software Foundation, version 2 of the
-+ * License.
-+ *
-+ */
++ err = xenbus_gather(XBT_NIL, dev->otherend, "ring-ref", "%lu",
++ &ring_ref, "event-channel", "%u", &evtchn, NULL);
++ if (err) {
++ xenbus_dev_fatal(dev, err,
++ "reading %s/ring-ref and event-channel",
++ dev->otherend);
++ return err;
++ }
+
-+#include <asm/uaccess.h>
-+#include <linux/list.h>
-+#include <linux/device.h>
-+#include <linux/interrupt.h>
-+#include <linux/platform_device.h>
-+#include "tpm.h"
-+#include "tpm_vtpm.h"
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_NATIVE;
++ err = xenbus_gather(XBT_NIL, dev->otherend, "protocol",
++ "%63s", protocol, NULL);
++ if (err)
++ strcpy(protocol, "unspecified, assuming native");
++ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_NATIVE))
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_NATIVE;
++ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_X86_32))
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_X86_32;
++ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_X86_64))
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_X86_64;
++ else {
++ xenbus_dev_fatal(dev, err, "unknown fe protocol %s", protocol);
++ return -1;
++ }
++ printk(KERN_INFO
++ "blktap: ring-ref %ld, event-channel %d, protocol %d (%s)\n",
++ ring_ref, evtchn, be->blkif->blk_protocol, protocol);
+
-+/* read status bits */
-+enum {
-+ STATUS_BUSY = 0x01,
-+ STATUS_DATA_AVAIL = 0x02,
-+ STATUS_READY = 0x04
-+};
++ /* Map the shared frame, irq etc. */
++ err = tap_blkif_map(be->blkif, ring_ref, evtchn);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "mapping ring-ref %lu port %u",
++ ring_ref, evtchn);
++ return err;
++ }
+
-+struct transmission {
-+ struct list_head next;
++ return 0;
++}
+
-+ unsigned char *request;
-+ size_t request_len;
-+ size_t request_buflen;
+
-+ unsigned char *response;
-+ size_t response_len;
-+ size_t response_buflen;
++/* ** Driver Registration ** */
+
-+ unsigned int flags;
-+};
+
-+enum {
-+ TRANSMISSION_FLAG_WAS_QUEUED = 0x1
++static const struct xenbus_device_id blktap_ids[] = {
++ { "tap" },
++ { "" }
+};
+
+
-+enum {
-+ DATAEX_FLAG_QUEUED_ONLY = 0x1
++static struct xenbus_driver blktap = {
++ .name = "tap",
++ .owner = THIS_MODULE,
++ .ids = blktap_ids,
++ .probe = blktap_probe,
++ .remove = blktap_remove,
++ .otherend_changed = tap_frontend_changed
+};
+
+
-+/* local variables */
-+
-+/* local function prototypes */
-+static int _vtpm_send_queued(struct tpm_chip *chip);
-+
-+
-+/* =============================================================
-+ * Some utility functions
-+ * =============================================================
-+ */
-+static void vtpm_state_init(struct vtpm_state *vtpms)
++void tap_blkif_xenbus_init(void)
+{
-+ vtpms->current_request = NULL;
-+ spin_lock_init(&vtpms->req_list_lock);
-+ init_waitqueue_head(&vtpms->req_wait_queue);
-+ INIT_LIST_HEAD(&vtpms->queued_requests);
-+
-+ vtpms->current_response = NULL;
-+ spin_lock_init(&vtpms->resp_list_lock);
-+ init_waitqueue_head(&vtpms->resp_wait_queue);
-+
-+ vtpms->disconnect_time = jiffies;
++ xenbus_register_backend(&blktap);
+}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/char/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/char/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1 @@
++obj-$(CONFIG_XEN_DEVMEM) := mem.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/char/mem.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/char/mem.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,190 @@
++/*
++ * Originally from linux/drivers/char/mem.c
++ *
++ * Copyright (C) 1991, 1992 Linus Torvalds
++ *
++ * Added devfs support.
++ * Jan-11-1998, C. Scott Ananian <cananian at alumni.princeton.edu>
++ * Shared /dev/zero mmaping support, Feb 2000, Kanoj Sarcar <kanoj at sgi.com>
++ */
+
++#include <linux/mm.h>
++#include <linux/miscdevice.h>
++#include <linux/slab.h>
++#include <linux/vmalloc.h>
++#include <linux/mman.h>
++#include <linux/random.h>
++#include <linux/init.h>
++#include <linux/raw.h>
++#include <linux/tty.h>
++#include <linux/capability.h>
++#include <linux/smp_lock.h>
++#include <linux/ptrace.h>
++#include <linux/device.h>
++#include <asm/pgalloc.h>
++#include <asm/uaccess.h>
++#include <asm/io.h>
++#include <asm/hypervisor.h>
+
-+static inline struct transmission *transmission_alloc(void)
++static inline int uncached_access(struct file *file)
+{
-+ return kzalloc(sizeof(struct transmission), GFP_ATOMIC);
++ if (file->f_flags & O_SYNC)
++ return 1;
++ /* Xen sets correct MTRR type on non-RAM for us. */
++ return 0;
+}
+
-+static unsigned char *
-+transmission_set_req_buffer(struct transmission *t,
-+ unsigned char *buffer, size_t len)
++/*
++ * This funcion reads the *physical* memory. The f_pos points directly to the
++ * memory location.
++ */
++static ssize_t read_mem(struct file * file, char __user * buf,
++ size_t count, loff_t *ppos)
+{
-+ if (t->request_buflen < len) {
-+ kfree(t->request);
-+ t->request = kmalloc(len, GFP_KERNEL);
-+ if (!t->request) {
-+ t->request_buflen = 0;
-+ return NULL;
-+ }
-+ t->request_buflen = len;
-+ }
++ unsigned long p = *ppos, ignored;
++ ssize_t read = 0, sz;
++ void __iomem *v;
+
-+ memcpy(t->request, buffer, len);
-+ t->request_len = len;
++ while (count > 0) {
++ /*
++ * Handle first page in case it's not aligned
++ */
++ if (-p & (PAGE_SIZE - 1))
++ sz = -p & (PAGE_SIZE - 1);
++ else
++ sz = PAGE_SIZE;
+
-+ return t->request;
-+}
++ sz = min_t(unsigned long, sz, count);
+
-+static unsigned char *
-+transmission_set_res_buffer(struct transmission *t,
-+ const unsigned char *buffer, size_t len)
-+{
-+ if (t->response_buflen < len) {
-+ kfree(t->response);
-+ t->response = kmalloc(len, GFP_ATOMIC);
-+ if (!t->response) {
-+ t->response_buflen = 0;
-+ return NULL;
++ v = ioremap(p, sz);
++ if (IS_ERR(v) || v == NULL) {
++ /*
++ * Some programs (e.g., dmidecode) groove off into
++ * weird RAM areas where no tables can possibly exist
++ * (because Xen will have stomped on them!). These
++ * programs get rather upset if we let them know that
++ * Xen failed their access, so we fake out a read of
++ * all zeroes.
++ */
++ if (clear_user(buf, count))
++ return -EFAULT;
++ read += count;
++ break;
+ }
-+ t->response_buflen = len;
-+ }
+
-+ memcpy(t->response, buffer, len);
-+ t->response_len = len;
-+
-+ return t->response;
-+}
++ ignored = copy_to_user(buf, v, sz);
++ iounmap(v);
++ if (ignored)
++ return -EFAULT;
++ buf += sz;
++ p += sz;
++ count -= sz;
++ read += sz;
++ }
+
-+static inline void transmission_free(struct transmission *t)
-+{
-+ kfree(t->request);
-+ kfree(t->response);
-+ kfree(t);
++ *ppos += read;
++ return read;
+}
+
-+/* =============================================================
-+ * Interface with the lower layer driver
-+ * =============================================================
-+ */
-+/*
-+ * Lower layer uses this function to make a response available.
-+ */
-+int vtpm_vd_recv(const struct tpm_chip *chip,
-+ const unsigned char *buffer, size_t count,
-+ void *ptr)
++static ssize_t write_mem(struct file * file, const char __user * buf,
++ size_t count, loff_t *ppos)
+{
-+ unsigned long flags;
-+ int ret_size = 0;
-+ struct transmission *t;
-+ struct vtpm_state *vtpms;
++ unsigned long p = *ppos, ignored;
++ ssize_t written = 0, sz;
++ void __iomem *v;
+
-+ vtpms = (struct vtpm_state *)chip_get_private(chip);
++ while (count > 0) {
++ /*
++ * Handle first page in case it's not aligned
++ */
++ if (-p & (PAGE_SIZE - 1))
++ sz = -p & (PAGE_SIZE - 1);
++ else
++ sz = PAGE_SIZE;
+
-+ /*
-+ * The list with requests must contain one request
-+ * only and the element there must be the one that
-+ * was passed to me from the front-end.
-+ */
-+ spin_lock_irqsave(&vtpms->resp_list_lock, flags);
-+ if (vtpms->current_request != ptr) {
-+ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
-+ return 0;
-+ }
++ sz = min_t(unsigned long, sz, count);
+
-+ if ((t = vtpms->current_request)) {
-+ transmission_free(t);
-+ vtpms->current_request = NULL;
-+ }
++ v = ioremap(p, sz);
++ if (v == NULL)
++ break;
++ if (IS_ERR(v)) {
++ if (written == 0)
++ return PTR_ERR(v);
++ break;
++ }
+
-+ t = transmission_alloc();
-+ if (t) {
-+ if (!transmission_set_res_buffer(t, buffer, count)) {
-+ transmission_free(t);
-+ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
-+ return -ENOMEM;
++ ignored = copy_from_user(v, buf, sz);
++ iounmap(v);
++ if (ignored) {
++ written += sz - ignored;
++ if (written)
++ break;
++ return -EFAULT;
+ }
-+ ret_size = count;
-+ vtpms->current_response = t;
-+ wake_up_interruptible(&vtpms->resp_wait_queue);
++ buf += sz;
++ p += sz;
++ count -= sz;
++ written += sz;
+ }
-+ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
+
-+ return ret_size;
++ *ppos += written;
++ return written;
+}
+
-+
-+/*
-+ * Lower layer indicates its status (connected/disconnected)
-+ */
-+void vtpm_vd_status(const struct tpm_chip *chip, u8 vd_status)
++#ifndef ARCH_HAS_DEV_MEM_MMAP_MEM
++static int xen_mmap_mem(struct file * file, struct vm_area_struct * vma)
+{
-+ struct vtpm_state *vtpms;
++ size_t size = vma->vm_end - vma->vm_start;
+
-+ vtpms = (struct vtpm_state *)chip_get_private(chip);
++ if (uncached_access(file))
++ vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
+
-+ vtpms->vd_status = vd_status;
-+ if ((vtpms->vd_status & TPM_VD_STATUS_CONNECTED) == 0) {
-+ vtpms->disconnect_time = jiffies;
-+ }
++ /* We want to return the real error code, not EAGAIN. */
++ return direct_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff,
++ size, vma->vm_page_prot, DOMID_IO);
+}
++#endif
+
-+/* =============================================================
-+ * Interface with the generic TPM driver
-+ * =============================================================
++/*
++ * The memory devices use the full 32/64 bits of the offset, and so we cannot
++ * check against negative addresses: they are ok. The return value is weird,
++ * though, in that case (0).
++ *
++ * also note that seeking relative to the "end of file" isn't supported:
++ * it has no meaning, so it returns -EINVAL.
+ */
-+static int vtpm_recv(struct tpm_chip *chip, u8 *buf, size_t count)
++static loff_t memory_lseek(struct file * file, loff_t offset, int orig)
+{
-+ int rc = 0;
-+ unsigned long flags;
-+ struct vtpm_state *vtpms;
-+
-+ vtpms = (struct vtpm_state *)chip_get_private(chip);
-+
-+ /*
-+ * Check if the previous operation only queued the command
-+ * In this case there won't be a response, so I just
-+ * return from here and reset that flag. In any other
-+ * case I should receive a response from the back-end.
-+ */
-+ spin_lock_irqsave(&vtpms->resp_list_lock, flags);
-+ if ((vtpms->flags & DATAEX_FLAG_QUEUED_ONLY) != 0) {
-+ vtpms->flags &= ~DATAEX_FLAG_QUEUED_ONLY;
-+ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
-+ /*
-+ * The first few commands (measurements) must be
-+ * queued since it might not be possible to talk to the
-+ * TPM, yet.
-+ * Return a response of up to 30 '0's.
-+ */
-+
-+ count = min_t(size_t, count, 30);
-+ memset(buf, 0x0, count);
-+ return count;
-+ }
-+ /*
-+ * Check whether something is in the responselist and if
-+ * there's nothing in the list wait for something to appear.
-+ */
-+
-+ if (!vtpms->current_response) {
-+ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
-+ interruptible_sleep_on_timeout(&vtpms->resp_wait_queue,
-+ 1000);
-+ spin_lock_irqsave(&vtpms->resp_list_lock ,flags);
-+ }
++ loff_t ret;
+
-+ if (vtpms->current_response) {
-+ struct transmission *t = vtpms->current_response;
-+ vtpms->current_response = NULL;
-+ rc = min(count, t->response_len);
-+ memcpy(buf, t->response, rc);
-+ transmission_free(t);
++ mutex_lock(&file->f_dentry->d_inode->i_mutex);
++ switch (orig) {
++ case 0:
++ file->f_pos = offset;
++ ret = file->f_pos;
++ force_successful_syscall_return();
++ break;
++ case 1:
++ file->f_pos += offset;
++ ret = file->f_pos;
++ force_successful_syscall_return();
++ break;
++ default:
++ ret = -EINVAL;
+ }
-+
-+ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
-+ return rc;
++ mutex_unlock(&file->f_dentry->d_inode->i_mutex);
++ return ret;
+}
+
-+static int vtpm_send(struct tpm_chip *chip, u8 *buf, size_t count)
++static int open_mem(struct inode * inode, struct file * filp)
+{
-+ int rc = 0;
-+ unsigned long flags;
-+ struct transmission *t = transmission_alloc();
-+ struct vtpm_state *vtpms;
-+
-+ vtpms = (struct vtpm_state *)chip_get_private(chip);
++ return capable(CAP_SYS_RAWIO) ? 0 : -EPERM;
++}
+
-+ if (!t)
-+ return -ENOMEM;
-+ /*
-+ * If there's a current request, it must be the
-+ * previous request that has timed out.
-+ */
-+ spin_lock_irqsave(&vtpms->req_list_lock, flags);
-+ if (vtpms->current_request != NULL) {
-+ printk("WARNING: Sending although there is a request outstanding.\n"
-+ " Previous request must have timed out.\n");
-+ transmission_free(vtpms->current_request);
-+ vtpms->current_request = NULL;
-+ }
-+ spin_unlock_irqrestore(&vtpms->req_list_lock, flags);
++const struct file_operations mem_fops = {
++ .llseek = memory_lseek,
++ .read = read_mem,
++ .write = write_mem,
++ .mmap = xen_mmap_mem,
++ .open = open_mem,
++};
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/console/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/console/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,2 @@
+
-+ /*
-+ * Queue the packet if the driver below is not
-+ * ready, yet, or there is any packet already
-+ * in the queue.
-+ * If the driver below is ready, unqueue all
-+ * packets first before sending our current
-+ * packet.
-+ * For each unqueued packet, except for the
-+ * last (=current) packet, call the function
-+ * tpm_xen_recv to wait for the response to come
-+ * back.
-+ */
-+ if ((vtpms->vd_status & TPM_VD_STATUS_CONNECTED) == 0) {
-+ if (time_after(jiffies,
-+ vtpms->disconnect_time + HZ * 10)) {
-+ rc = -ENOENT;
-+ } else {
-+ goto queue_it;
-+ }
-+ } else {
-+ /*
-+ * Send all queued packets.
-+ */
-+ if (_vtpm_send_queued(chip) == 0) {
++obj-y := console.o xencons_ring.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/console/console.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/console/console.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,731 @@
++/******************************************************************************
++ * console.c
++ *
++ * Virtual console driver.
++ *
++ * Copyright (c) 2002-2004, K A Fraser.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ vtpms->current_request = t;
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/errno.h>
++#include <linux/signal.h>
++#include <linux/sched.h>
++#include <linux/interrupt.h>
++#include <linux/tty.h>
++#include <linux/tty_flip.h>
++#include <linux/serial.h>
++#include <linux/major.h>
++#include <linux/ptrace.h>
++#include <linux/ioport.h>
++#include <linux/mm.h>
++#include <linux/slab.h>
++#include <linux/init.h>
++#include <linux/console.h>
++#include <linux/bootmem.h>
++#include <linux/sysrq.h>
++#include <linux/screen_info.h>
++#include <linux/vt.h>
++#include <asm/io.h>
++#include <asm/irq.h>
++#include <asm/uaccess.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/event_channel.h>
++#include <asm/hypervisor.h>
++#include <xen/evtchn.h>
++#include <xen/xenbus.h>
++#include <xen/xencons.h>
+
-+ rc = vtpm_vd_send(vtpms->tpm_private,
-+ buf,
-+ count,
-+ t);
-+ /*
-+ * The generic TPM driver will call
-+ * the function to receive the response.
-+ */
-+ if (rc < 0) {
-+ vtpms->current_request = NULL;
-+ goto queue_it;
-+ }
-+ } else {
-+queue_it:
-+ if (!transmission_set_req_buffer(t, buf, count)) {
-+ transmission_free(t);
-+ rc = -ENOMEM;
-+ goto exit;
-+ }
-+ /*
-+ * An error occurred. Don't event try
-+ * to send the current request. Just
-+ * queue it.
-+ */
-+ spin_lock_irqsave(&vtpms->req_list_lock, flags);
-+ vtpms->flags |= DATAEX_FLAG_QUEUED_ONLY;
-+ list_add_tail(&t->next, &vtpms->queued_requests);
-+ spin_unlock_irqrestore(&vtpms->req_list_lock, flags);
-+ }
-+ }
++/*
++ * Modes:
++ * 'xencons=off' [XC_OFF]: Console is disabled.
++ * 'xencons=tty' [XC_TTY]: Console attached to '/dev/tty[0-9]+'.
++ * 'xencons=ttyS' [XC_SERIAL]: Console attached to '/dev/ttyS[0-9]+'.
++ * 'xencons=xvc' [XC_XVC]: Console attached to '/dev/xvc0'.
++ * default: XC_XVC
++ *
++ * NB. In mode XC_TTY, we create dummy consoles for tty2-63. This suppresses
++ * warnings from standard distro startup scripts.
++ */
++static enum {
++ XC_OFF, XC_TTY, XC_SERIAL, XC_XVC
++} xc_mode = XC_XVC;
++static int xc_num = -1;
+
-+exit:
-+ return rc;
-+}
++/* /dev/xvc0 device number allocated by lanana.org. */
++#define XEN_XVC_MAJOR 204
++#define XEN_XVC_MINOR 191
+
++#ifdef CONFIG_MAGIC_SYSRQ
++static unsigned long sysrq_requested;
++extern int sysrq_enabled;
++#endif
+
-+/*
-+ * Send all queued requests.
-+ */
-+static int _vtpm_send_queued(struct tpm_chip *chip)
++static int __init xencons_setup(char *str)
+{
-+ int rc;
-+ int error = 0;
-+ long flags;
-+ unsigned char buffer[1];
-+ struct vtpm_state *vtpms;
-+ vtpms = (struct vtpm_state *)chip_get_private(chip);
++ char *q;
++ int n;
++ extern int console_use_vt;
+
-+ spin_lock_irqsave(&vtpms->req_list_lock, flags);
++ console_use_vt = 1;
++ if (!strncmp(str, "ttyS", 4)) {
++ xc_mode = XC_SERIAL;
++ str += 4;
++ } else if (!strncmp(str, "tty", 3)) {
++ xc_mode = XC_TTY;
++ str += 3;
++ console_use_vt = 0;
++ } else if (!strncmp(str, "xvc", 3)) {
++ xc_mode = XC_XVC;
++ str += 3;
++ } else if (!strncmp(str, "off", 3)) {
++ xc_mode = XC_OFF;
++ str += 3;
++ }
+
-+ while (!list_empty(&vtpms->queued_requests)) {
-+ /*
-+ * Need to dequeue them.
-+ * Read the result into a dummy buffer.
-+ */
-+ struct transmission *qt = (struct transmission *)
-+ vtpms->queued_requests.next;
-+ list_del(&qt->next);
-+ vtpms->current_request = qt;
-+ spin_unlock_irqrestore(&vtpms->req_list_lock, flags);
++ n = simple_strtol(str, &q, 10);
++ if (q != str)
++ xc_num = n;
+
-+ rc = vtpm_vd_send(vtpms->tpm_private,
-+ qt->request,
-+ qt->request_len,
-+ qt);
++ return 1;
++}
++__setup("xencons=", xencons_setup);
+
-+ if (rc < 0) {
-+ spin_lock_irqsave(&vtpms->req_list_lock, flags);
-+ if ((qt = vtpms->current_request) != NULL) {
-+ /*
-+ * requeue it at the beginning
-+ * of the list
-+ */
-+ list_add(&qt->next,
-+ &vtpms->queued_requests);
-+ }
-+ vtpms->current_request = NULL;
-+ error = 1;
-+ break;
-+ }
-+ /*
-+ * After this point qt is not valid anymore!
-+ * It is freed when the front-end is delivering
-+ * the data by calling tpm_recv
-+ */
-+ /*
-+ * Receive response into provided dummy buffer
-+ */
-+ rc = vtpm_recv(chip, buffer, sizeof(buffer));
-+ spin_lock_irqsave(&vtpms->req_list_lock, flags);
++/* The kernel and user-land drivers share a common transmit buffer. */
++static unsigned int wbuf_size = 4096;
++#define WBUF_MASK(_i) ((_i)&(wbuf_size-1))
++static char *wbuf;
++static unsigned int wc, wp; /* write_cons, write_prod */
++
++static int __init xencons_bufsz_setup(char *str)
++{
++ unsigned int goal;
++ goal = simple_strtoul(str, NULL, 0);
++ if (goal) {
++ goal = roundup_pow_of_two(goal);
++ if (wbuf_size < goal)
++ wbuf_size = goal;
+ }
++ return 1;
++}
++__setup("xencons_bufsz=", xencons_bufsz_setup);
+
-+ spin_unlock_irqrestore(&vtpms->req_list_lock, flags);
++/* This lock protects accesses to the common transmit buffer. */
++static DEFINE_SPINLOCK(xencons_lock);
+
-+ return error;
-+}
++/* Common transmit-kick routine. */
++static void __xencons_tx_flush(void);
+
-+static void vtpm_cancel(struct tpm_chip *chip)
++static struct tty_driver *xencons_driver;
++
++/******************** Kernel console driver ********************************/
++
++static void kcons_write(struct console *c, const char *s, unsigned int count)
+{
++ int i = 0;
+ unsigned long flags;
-+ struct vtpm_state *vtpms = (struct vtpm_state *)chip_get_private(chip);
+
-+ spin_lock_irqsave(&vtpms->resp_list_lock,flags);
++ spin_lock_irqsave(&xencons_lock, flags);
+
-+ if (!vtpms->current_response && vtpms->current_request) {
-+ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
-+ interruptible_sleep_on(&vtpms->resp_wait_queue);
-+ spin_lock_irqsave(&vtpms->resp_list_lock,flags);
-+ }
++ while (i < count) {
++ for (; i < count; i++) {
++ if ((wp - wc) >= (wbuf_size - 1))
++ break;
++ if ((wbuf[WBUF_MASK(wp++)] = s[i]) == '\n')
++ wbuf[WBUF_MASK(wp++)] = '\r';
++ }
+
-+ if (vtpms->current_response) {
-+ struct transmission *t = vtpms->current_response;
-+ vtpms->current_response = NULL;
-+ transmission_free(t);
++ __xencons_tx_flush();
+ }
+
-+ spin_unlock_irqrestore(&vtpms->resp_list_lock,flags);
++ spin_unlock_irqrestore(&xencons_lock, flags);
+}
+
-+static u8 vtpm_status(struct tpm_chip *chip)
++static void kcons_write_dom0(struct console *c, const char *s, unsigned int count)
+{
-+ u8 rc = 0;
-+ unsigned long flags;
-+ struct vtpm_state *vtpms;
-+
-+ vtpms = (struct vtpm_state *)chip_get_private(chip);
+
-+ spin_lock_irqsave(&vtpms->resp_list_lock, flags);
-+ /*
-+ * Data are available if:
-+ * - there's a current response
-+ * - the last packet was queued only (this is fake, but necessary to
-+ * get the generic TPM layer to call the receive function.)
-+ */
-+ if (vtpms->current_response ||
-+ 0 != (vtpms->flags & DATAEX_FLAG_QUEUED_ONLY)) {
-+ rc = STATUS_DATA_AVAIL;
-+ } else if (!vtpms->current_response && !vtpms->current_request) {
-+ rc = STATUS_READY;
++ while (count > 0) {
++ int rc;
++ rc = HYPERVISOR_console_io( CONSOLEIO_write, count, (char *)s);
++ if (rc <= 0)
++ break;
++ count -= rc;
++ s += rc;
+ }
-+
-+ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
-+ return rc;
+}
+
-+static struct file_operations vtpm_ops = {
-+ .owner = THIS_MODULE,
-+ .llseek = no_llseek,
-+ .open = tpm_open,
-+ .read = tpm_read,
-+ .write = tpm_write,
-+ .release = tpm_release,
-+};
-+
-+static DEVICE_ATTR(pubek, S_IRUGO, tpm_show_pubek, NULL);
-+static DEVICE_ATTR(pcrs, S_IRUGO, tpm_show_pcrs, NULL);
-+static DEVICE_ATTR(enabled, S_IRUGO, tpm_show_enabled, NULL);
-+static DEVICE_ATTR(active, S_IRUGO, tpm_show_active, NULL);
-+static DEVICE_ATTR(owned, S_IRUGO, tpm_show_owned, NULL);
-+static DEVICE_ATTR(temp_deactivated, S_IRUGO, tpm_show_temp_deactivated,
-+ NULL);
-+static DEVICE_ATTR(caps, S_IRUGO, tpm_show_caps, NULL);
-+static DEVICE_ATTR(cancel, S_IWUSR |S_IWGRP, NULL, tpm_store_cancel);
++static struct tty_driver *kcons_device(struct console *c, int *index)
++{
++ *index = 0;
++ return xencons_driver;
++}
+
-+static struct attribute *vtpm_attrs[] = {
-+ &dev_attr_pubek.attr,
-+ &dev_attr_pcrs.attr,
-+ &dev_attr_enabled.attr,
-+ &dev_attr_active.attr,
-+ &dev_attr_owned.attr,
-+ &dev_attr_temp_deactivated.attr,
-+ &dev_attr_caps.attr,
-+ &dev_attr_cancel.attr,
-+ NULL,
++static struct console kcons_info = {
++ .device = kcons_device,
++ .flags = CON_PRINTBUFFER | CON_ENABLED,
++ .index = -1,
+};
+
-+static struct attribute_group vtpm_attr_grp = { .attrs = vtpm_attrs };
-+
-+#define TPM_LONG_TIMEOUT (10 * 60 * HZ)
++static int __init xen_console_init(void)
++{
++ if (!is_running_on_xen())
++ goto out;
+
-+static struct tpm_vendor_specific tpm_vtpm = {
-+ .recv = vtpm_recv,
-+ .send = vtpm_send,
-+ .cancel = vtpm_cancel,
-+ .status = vtpm_status,
-+ .req_complete_mask = STATUS_BUSY | STATUS_DATA_AVAIL,
-+ .req_complete_val = STATUS_DATA_AVAIL,
-+ .req_canceled = STATUS_READY,
-+ .attr_group = &vtpm_attr_grp,
-+ .miscdev = {
-+ .fops = &vtpm_ops,
-+ },
-+ .duration = {
-+ TPM_LONG_TIMEOUT,
-+ TPM_LONG_TIMEOUT,
-+ TPM_LONG_TIMEOUT,
-+ },
-+};
++ if (is_initial_xendomain()) {
++ kcons_info.write = kcons_write_dom0;
++ } else {
++ if (!xen_start_info->console.domU.evtchn)
++ goto out;
++ kcons_info.write = kcons_write;
++ }
+
-+struct tpm_chip *init_vtpm(struct device *dev,
-+ struct tpm_private *tp)
-+{
-+ long rc;
-+ struct tpm_chip *chip;
-+ struct vtpm_state *vtpms;
++ switch (xc_mode) {
++ case XC_XVC:
++ strcpy(kcons_info.name, "xvc");
++ if (xc_num == -1)
++ xc_num = 0;
++ break;
+
-+ vtpms = kzalloc(sizeof(struct vtpm_state), GFP_KERNEL);
-+ if (!vtpms)
-+ return ERR_PTR(-ENOMEM);
++ case XC_SERIAL:
++ strcpy(kcons_info.name, "ttyS");
++ if (xc_num == -1)
++ xc_num = 0;
++ break;
+
-+ vtpm_state_init(vtpms);
-+ vtpms->tpm_private = tp;
++ case XC_TTY:
++ strcpy(kcons_info.name, "tty");
++ if (xc_num == -1)
++ xc_num = 1;
++ break;
+
-+ chip = tpm_register_hardware(dev, &tpm_vtpm);
-+ if (!chip) {
-+ rc = -ENODEV;
-+ goto err_free_mem;
++ default:
++ goto out;
+ }
+
-+ chip_set_private(chip, vtpms);
-+
-+ return chip;
++ wbuf = alloc_bootmem(wbuf_size);
+
-+err_free_mem:
-+ kfree(vtpms);
++ register_console(&kcons_info);
+
-+ return ERR_PTR(rc);
++ out:
++ return 0;
+}
++console_initcall(xen_console_init);
+
-+void cleanup_vtpm(struct device *dev)
++/*** Useful function for console debugging -- goes straight to Xen. ***/
++asmlinkage int xprintk(const char *fmt, ...)
+{
-+ struct tpm_chip *chip = dev_get_drvdata(dev);
-+ struct vtpm_state *vtpms = (struct vtpm_state*)chip_get_private(chip);
-+ tpm_remove_hardware(dev);
-+ kfree(vtpms);
-+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/char/tpm/tpm_vtpm.h tmp-linux-2.6-xen.patch/drivers/char/tpm/tpm_vtpm.h
---- pristine-linux-2.6.18.2/drivers/char/tpm/tpm_vtpm.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/char/tpm/tpm_vtpm.h 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,55 @@
-+#ifndef TPM_VTPM_H
-+#define TPM_VTPM_H
-+
-+struct tpm_chip;
-+struct tpm_private;
-+
-+struct vtpm_state {
-+ struct transmission *current_request;
-+ spinlock_t req_list_lock;
-+ wait_queue_head_t req_wait_queue;
-+
-+ struct list_head queued_requests;
-+
-+ struct transmission *current_response;
-+ spinlock_t resp_list_lock;
-+ wait_queue_head_t resp_wait_queue; // processes waiting for responses
-+
-+ u8 vd_status;
-+ u8 flags;
-+
-+ unsigned long disconnect_time;
++ va_list args;
++ int printk_len;
++ static char printk_buf[1024];
+
-+ /*
-+ * The following is a private structure of the underlying
-+ * driver. It is passed as parameter in the send function.
-+ */
-+ struct tpm_private *tpm_private;
-+};
++ /* Emit the output into the temporary buffer */
++ va_start(args, fmt);
++ printk_len = vsnprintf(printk_buf, sizeof(printk_buf), fmt, args);
++ va_end(args);
+
++ /* Send the processed output directly to Xen. */
++ kcons_write_dom0(NULL, printk_buf, printk_len);
+
-+enum vdev_status {
-+ TPM_VD_STATUS_DISCONNECTED = 0x0,
-+ TPM_VD_STATUS_CONNECTED = 0x1
-+};
++ return 0;
++}
+
-+/* this function is called from tpm_vtpm.c */
-+int vtpm_vd_send(struct tpm_private * tp,
-+ const u8 * buf, size_t count, void *ptr);
++/*** Forcibly flush console data before dying. ***/
++void xencons_force_flush(void)
++{
++ int sz;
+
-+/* these functions are offered by tpm_vtpm.c */
-+struct tpm_chip *init_vtpm(struct device *,
-+ struct tpm_private *);
-+void cleanup_vtpm(struct device *);
-+int vtpm_vd_recv(const struct tpm_chip* chip,
-+ const unsigned char *buffer, size_t count, void *ptr);
-+void vtpm_vd_status(const struct tpm_chip *, u8 status);
++ /* Emergency console is synchronous, so there's nothing to flush. */
++ if (!is_running_on_xen() ||
++ is_initial_xendomain() ||
++ !xen_start_info->console.domU.evtchn)
++ return;
+
-+static inline struct tpm_private *tpm_private_from_dev(struct device *dev)
-+{
-+ struct tpm_chip *chip = dev_get_drvdata(dev);
-+ struct vtpm_state *vtpms = chip_get_private(chip);
-+ return vtpms->tpm_private;
++ /* Spin until console data is flushed through to the daemon. */
++ while (wc != wp) {
++ int sent = 0;
++ if ((sz = wp - wc) == 0)
++ continue;
++ sent = xencons_ring_send(&wbuf[WBUF_MASK(wc)], sz);
++ if (sent > 0)
++ wc += sent;
++ }
+}
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/drivers/char/tpm/tpm_xen.c tmp-linux-2.6-xen.patch/drivers/char/tpm/tpm_xen.c
---- pristine-linux-2.6.18.2/drivers/char/tpm/tpm_xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/char/tpm/tpm_xen.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,720 @@
-+/*
-+ * Copyright (c) 2005, IBM Corporation
-+ *
-+ * Author: Stefan Berger, stefanb at us.ibm.com
-+ * Grant table support: Mahadevan Gomathisankaran
-+ *
-+ * This code has been derived from drivers/xen/netfront/netfront.c
-+ *
-+ * Copyright (c) 2002-2004, K A Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
-+
-+#include <linux/errno.h>
-+#include <linux/err.h>
-+#include <linux/interrupt.h>
-+#include <linux/mutex.h>
-+#include <asm/uaccess.h>
-+#include <xen/evtchn.h>
-+#include <xen/interface/grant_table.h>
-+#include <xen/interface/io/tpmif.h>
-+#include <xen/gnttab.h>
-+#include <xen/xenbus.h>
-+#include "tpm.h"
-+#include "tpm_vtpm.h"
+
-+#undef DEBUG
++void __init dom0_init_screen_info(const struct dom0_vga_console_info *info, size_t size)
++{
++ /* This is drawn from a dump from vgacon:startup in
++ * standard Linux. */
++ screen_info.orig_video_mode = 3;
++ screen_info.orig_video_isVGA = 1;
++ screen_info.orig_video_lines = 25;
++ screen_info.orig_video_cols = 80;
++ screen_info.orig_video_ega_bx = 3;
++ screen_info.orig_video_points = 16;
++ screen_info.orig_y = screen_info.orig_video_lines - 1;
+
-+/* local structures */
-+struct tpm_private {
-+ struct tpm_chip *chip;
++ switch (info->video_type) {
++ case XEN_VGATYPE_TEXT_MODE_3:
++ if (size < offsetof(struct dom0_vga_console_info, u.text_mode_3)
++ + sizeof(info->u.text_mode_3))
++ break;
++ screen_info.orig_video_lines = info->u.text_mode_3.rows;
++ screen_info.orig_video_cols = info->u.text_mode_3.columns;
++ screen_info.orig_x = info->u.text_mode_3.cursor_x;
++ screen_info.orig_y = info->u.text_mode_3.cursor_y;
++ screen_info.orig_video_points =
++ info->u.text_mode_3.font_height;
++ break;
+
-+ tpmif_tx_interface_t *tx;
-+ atomic_t refcnt;
-+ unsigned int irq;
-+ u8 is_connected;
-+ u8 is_suspended;
++ case XEN_VGATYPE_VESA_LFB:
++ if (size < offsetof(struct dom0_vga_console_info,
++ u.vesa_lfb.gbl_caps))
++ break;
++ screen_info.orig_video_isVGA = VIDEO_TYPE_VLFB;
++ screen_info.lfb_width = info->u.vesa_lfb.width;
++ screen_info.lfb_height = info->u.vesa_lfb.height;
++ screen_info.lfb_depth = info->u.vesa_lfb.bits_per_pixel;
++ screen_info.lfb_base = info->u.vesa_lfb.lfb_base;
++ screen_info.lfb_size = info->u.vesa_lfb.lfb_size;
++ screen_info.lfb_linelength = info->u.vesa_lfb.bytes_per_line;
++ screen_info.red_size = info->u.vesa_lfb.red_size;
++ screen_info.red_pos = info->u.vesa_lfb.red_pos;
++ screen_info.green_size = info->u.vesa_lfb.green_size;
++ screen_info.green_pos = info->u.vesa_lfb.green_pos;
++ screen_info.blue_size = info->u.vesa_lfb.blue_size;
++ screen_info.blue_pos = info->u.vesa_lfb.blue_pos;
++ screen_info.rsvd_size = info->u.vesa_lfb.rsvd_size;
++ screen_info.rsvd_pos = info->u.vesa_lfb.rsvd_pos;
++ if (size >= offsetof(struct dom0_vga_console_info,
++ u.vesa_lfb.gbl_caps)
++ + sizeof(info->u.vesa_lfb.gbl_caps))
++ screen_info.capabilities = info->u.vesa_lfb.gbl_caps;
++ if (size >= offsetof(struct dom0_vga_console_info,
++ u.vesa_lfb.mode_attrs)
++ + sizeof(info->u.vesa_lfb.mode_attrs))
++ screen_info.vesa_attributes = info->u.vesa_lfb.mode_attrs;
++ break;
++ }
++}
+
-+ spinlock_t tx_lock;
+
-+ struct tx_buffer *tx_buffers[TPMIF_TX_RING_SIZE];
++/******************** User-space console driver (/dev/console) ************/
+
-+ atomic_t tx_busy;
-+ void *tx_remember;
++#define DRV(_d) (_d)
++#define DUMMY_TTY(_tty) ((xc_mode == XC_TTY) && \
++ ((_tty)->index != (xc_num - 1)))
+
-+ domid_t backend_id;
-+ wait_queue_head_t wait_q;
++static struct termios *xencons_termios[MAX_NR_CONSOLES];
++static struct termios *xencons_termios_locked[MAX_NR_CONSOLES];
++static struct tty_struct *xencons_tty;
++static int xencons_priv_irq;
++static char x_char;
+
-+ struct xenbus_device *dev;
-+ int ring_ref;
-+};
++void xencons_rx(char *buf, unsigned len, struct pt_regs *regs)
++{
++ int i;
++ unsigned long flags;
+
-+struct tx_buffer {
-+ unsigned int size; // available space in data
-+ unsigned int len; // used space in data
-+ unsigned char *data; // pointer to a page
-+};
++ spin_lock_irqsave(&xencons_lock, flags);
++ if (xencons_tty == NULL)
++ goto out;
+
++ for (i = 0; i < len; i++) {
++#ifdef CONFIG_MAGIC_SYSRQ
++ if (sysrq_enabled) {
++ if (buf[i] == '\x0f') { /* ^O */
++ if (!sysrq_requested) {
++ sysrq_requested = jiffies;
++ continue; /* don't print sysrq key */
++ }
++ sysrq_requested = 0;
++ } else if (sysrq_requested) {
++ unsigned long sysrq_timeout =
++ sysrq_requested + HZ*2;
++ sysrq_requested = 0;
++ if (time_before(jiffies, sysrq_timeout)) {
++ spin_unlock_irqrestore(
++ &xencons_lock, flags);
++ handle_sysrq(
++ buf[i], regs, xencons_tty);
++ spin_lock_irqsave(
++ &xencons_lock, flags);
++ continue;
++ }
++ }
++ }
++#endif
++ tty_insert_flip_char(xencons_tty, buf[i], 0);
++ }
++ tty_flip_buffer_push(xencons_tty);
+
-+/* locally visible variables */
-+static grant_ref_t gref_head;
-+static struct tpm_private *my_priv;
++ out:
++ spin_unlock_irqrestore(&xencons_lock, flags);
++}
+
-+/* local function prototypes */
-+static irqreturn_t tpmif_int(int irq,
-+ void *tpm_priv,
-+ struct pt_regs *ptregs);
-+static void tpmif_rx_action(unsigned long unused);
-+static int tpmif_connect(struct xenbus_device *dev,
-+ struct tpm_private *tp,
-+ domid_t domid);
-+static DECLARE_TASKLET(tpmif_rx_tasklet, tpmif_rx_action, 0);
-+static int tpmif_allocate_tx_buffers(struct tpm_private *tp);
-+static void tpmif_free_tx_buffers(struct tpm_private *tp);
-+static void tpmif_set_connected_state(struct tpm_private *tp,
-+ u8 newstate);
-+static int tpm_xmit(struct tpm_private *tp,
-+ const u8 * buf, size_t count, int userbuffer,
-+ void *remember);
-+static void destroy_tpmring(struct tpm_private *tp);
-+void __exit tpmif_exit(void);
++static void __xencons_tx_flush(void)
++{
++ int sent, sz, work_done = 0;
+
-+#define DPRINTK(fmt, args...) \
-+ pr_debug("xen_tpm_fr (%s:%d) " fmt, __FUNCTION__, __LINE__, ##args)
-+#define IPRINTK(fmt, args...) \
-+ printk(KERN_INFO "xen_tpm_fr: " fmt, ##args)
-+#define WPRINTK(fmt, args...) \
-+ printk(KERN_WARNING "xen_tpm_fr: " fmt, ##args)
++ if (x_char) {
++ if (is_initial_xendomain())
++ kcons_write_dom0(NULL, &x_char, 1);
++ else
++ while (x_char)
++ if (xencons_ring_send(&x_char, 1) == 1)
++ break;
++ x_char = 0;
++ work_done = 1;
++ }
+
-+#define GRANT_INVALID_REF 0
++ while (wc != wp) {
++ sz = wp - wc;
++ if (sz > (wbuf_size - WBUF_MASK(wc)))
++ sz = wbuf_size - WBUF_MASK(wc);
++ if (is_initial_xendomain()) {
++ kcons_write_dom0(NULL, &wbuf[WBUF_MASK(wc)], sz);
++ wc += sz;
++ } else {
++ sent = xencons_ring_send(&wbuf[WBUF_MASK(wc)], sz);
++ if (sent == 0)
++ break;
++ wc += sent;
++ }
++ work_done = 1;
++ }
+
++ if (work_done && (xencons_tty != NULL)) {
++ wake_up_interruptible(&xencons_tty->write_wait);
++ if ((xencons_tty->flags & (1 << TTY_DO_WRITE_WAKEUP)) &&
++ (xencons_tty->ldisc.write_wakeup != NULL))
++ (xencons_tty->ldisc.write_wakeup)(xencons_tty);
++ }
++}
+
-+static inline int
-+tx_buffer_copy(struct tx_buffer *txb, const u8 *src, int len,
-+ int isuserbuffer)
++void xencons_tx(void)
+{
-+ int copied = len;
++ unsigned long flags;
+
-+ if (len > txb->size)
-+ copied = txb->size;
-+ if (isuserbuffer) {
-+ if (copy_from_user(txb->data, src, copied))
-+ return -EFAULT;
-+ } else {
-+ memcpy(txb->data, src, copied);
-+ }
-+ txb->len = len;
-+ return copied;
++ spin_lock_irqsave(&xencons_lock, flags);
++ __xencons_tx_flush();
++ spin_unlock_irqrestore(&xencons_lock, flags);
+}
+
-+static inline struct tx_buffer *tx_buffer_alloc(void)
++/* Privileged receive callback and transmit kicker. */
++static irqreturn_t xencons_priv_interrupt(int irq, void *dev_id,
++ struct pt_regs *regs)
+{
-+ struct tx_buffer *txb;
++ static char rbuf[16];
++ int l;
+
-+ txb = kzalloc(sizeof(struct tx_buffer), GFP_KERNEL);
-+ if (!txb)
-+ return NULL;
++ while ((l = HYPERVISOR_console_io(CONSOLEIO_read, 16, rbuf)) > 0)
++ xencons_rx(rbuf, l, regs);
+
-+ txb->len = 0;
-+ txb->size = PAGE_SIZE;
-+ txb->data = (unsigned char *)__get_free_page(GFP_KERNEL);
-+ if (txb->data == NULL) {
-+ kfree(txb);
-+ txb = NULL;
-+ }
++ xencons_tx();
+
-+ return txb;
++ return IRQ_HANDLED;
+}
+
++static int xencons_write_room(struct tty_struct *tty)
++{
++ return wbuf_size - (wp - wc);
++}
+
-+static inline void tx_buffer_free(struct tx_buffer *txb)
++static int xencons_chars_in_buffer(struct tty_struct *tty)
+{
-+ if (txb) {
-+ free_page((long)txb->data);
-+ kfree(txb);
-+ }
++ return wp - wc;
+}
+
-+/**************************************************************
-+ Utility function for the tpm_private structure
-+**************************************************************/
-+static void tpm_private_init(struct tpm_private *tp)
++static void xencons_send_xchar(struct tty_struct *tty, char ch)
+{
-+ spin_lock_init(&tp->tx_lock);
-+ init_waitqueue_head(&tp->wait_q);
-+ atomic_set(&tp->refcnt, 1);
++ unsigned long flags;
++
++ if (DUMMY_TTY(tty))
++ return;
++
++ spin_lock_irqsave(&xencons_lock, flags);
++ x_char = ch;
++ __xencons_tx_flush();
++ spin_unlock_irqrestore(&xencons_lock, flags);
+}
+
-+static void tpm_private_put(void)
++static void xencons_throttle(struct tty_struct *tty)
+{
-+ if (!atomic_dec_and_test(&my_priv->refcnt))
++ if (DUMMY_TTY(tty))
+ return;
+
-+ tpmif_free_tx_buffers(my_priv);
-+ kfree(my_priv);
-+ my_priv = NULL;
++ if (I_IXOFF(tty))
++ xencons_send_xchar(tty, STOP_CHAR(tty));
+}
+
-+static struct tpm_private *tpm_private_get(void)
++static void xencons_unthrottle(struct tty_struct *tty)
+{
-+ int err;
++ if (DUMMY_TTY(tty))
++ return;
+
-+ if (my_priv) {
-+ atomic_inc(&my_priv->refcnt);
-+ return my_priv;
++ if (I_IXOFF(tty)) {
++ if (x_char != 0)
++ x_char = 0;
++ else
++ xencons_send_xchar(tty, START_CHAR(tty));
+ }
++}
+
-+ my_priv = kzalloc(sizeof(struct tpm_private), GFP_KERNEL);
-+ if (!my_priv)
-+ return NULL;
++static void xencons_flush_buffer(struct tty_struct *tty)
++{
++ unsigned long flags;
+
-+ tpm_private_init(my_priv);
-+ err = tpmif_allocate_tx_buffers(my_priv);
-+ if (err < 0)
-+ tpm_private_put();
++ if (DUMMY_TTY(tty))
++ return;
+
-+ return my_priv;
++ spin_lock_irqsave(&xencons_lock, flags);
++ wc = wp = 0;
++ spin_unlock_irqrestore(&xencons_lock, flags);
+}
+
-+/**************************************************************
++static inline int __xencons_put_char(int ch)
++{
++ char _ch = (char)ch;
++ if ((wp - wc) == wbuf_size)
++ return 0;
++ wbuf[WBUF_MASK(wp++)] = _ch;
++ return 1;
++}
+
-+ The interface to let the tpm plugin register its callback
-+ function and send data to another partition using this module
++static int xencons_write(
++ struct tty_struct *tty,
++ const unsigned char *buf,
++ int count)
++{
++ int i;
++ unsigned long flags;
+
-+**************************************************************/
++ if (DUMMY_TTY(tty))
++ return count;
+
-+static DEFINE_MUTEX(suspend_lock);
-+/*
-+ * Send data via this module by calling this function
-+ */
-+int vtpm_vd_send(struct tpm_private *tp,
-+ const u8 * buf, size_t count, void *ptr)
++ spin_lock_irqsave(&xencons_lock, flags);
++
++ for (i = 0; i < count; i++)
++ if (!__xencons_put_char(buf[i]))
++ break;
++
++ if (i != 0)
++ __xencons_tx_flush();
++
++ spin_unlock_irqrestore(&xencons_lock, flags);
++
++ return i;
++}
++
++static void xencons_put_char(struct tty_struct *tty, u_char ch)
+{
-+ int sent;
++ unsigned long flags;
+
-+ mutex_lock(&suspend_lock);
-+ sent = tpm_xmit(tp, buf, count, 0, ptr);
-+ mutex_unlock(&suspend_lock);
++ if (DUMMY_TTY(tty))
++ return;
+
-+ return sent;
++ spin_lock_irqsave(&xencons_lock, flags);
++ (void)__xencons_put_char(ch);
++ spin_unlock_irqrestore(&xencons_lock, flags);
+}
+
-+/**************************************************************
-+ XENBUS support code
-+**************************************************************/
++static void xencons_flush_chars(struct tty_struct *tty)
++{
++ unsigned long flags;
+
-+static int setup_tpmring(struct xenbus_device *dev,
-+ struct tpm_private *tp)
++ if (DUMMY_TTY(tty))
++ return;
++
++ spin_lock_irqsave(&xencons_lock, flags);
++ __xencons_tx_flush();
++ spin_unlock_irqrestore(&xencons_lock, flags);
++}
++
++static void xencons_wait_until_sent(struct tty_struct *tty, int timeout)
+{
-+ tpmif_tx_interface_t *sring;
-+ int err;
++ unsigned long orig_jiffies = jiffies;
+
-+ tp->ring_ref = GRANT_INVALID_REF;
++ if (DUMMY_TTY(tty))
++ return;
+
-+ sring = (void *)__get_free_page(GFP_KERNEL);
-+ if (!sring) {
-+ xenbus_dev_fatal(dev, -ENOMEM, "allocating shared ring");
-+ return -ENOMEM;
++ while (DRV(tty->driver)->chars_in_buffer(tty)) {
++ set_current_state(TASK_INTERRUPTIBLE);
++ schedule_timeout(1);
++ if (signal_pending(current))
++ break;
++ if (timeout && time_after(jiffies, orig_jiffies + timeout))
++ break;
+ }
-+ tp->tx = sring;
+
-+ err = xenbus_grant_ring(dev, virt_to_mfn(tp->tx));
-+ if (err < 0) {
-+ free_page((unsigned long)sring);
-+ tp->tx = NULL;
-+ xenbus_dev_fatal(dev, err, "allocating grant reference");
-+ goto fail;
-+ }
-+ tp->ring_ref = err;
++ set_current_state(TASK_RUNNING);
++}
+
-+ err = tpmif_connect(dev, tp, dev->otherend_id);
-+ if (err)
-+ goto fail;
++static int xencons_open(struct tty_struct *tty, struct file *filp)
++{
++ unsigned long flags;
++
++ if (DUMMY_TTY(tty))
++ return 0;
++
++ spin_lock_irqsave(&xencons_lock, flags);
++ tty->driver_data = NULL;
++ if (xencons_tty == NULL)
++ xencons_tty = tty;
++ __xencons_tx_flush();
++ spin_unlock_irqrestore(&xencons_lock, flags);
+
+ return 0;
-+fail:
-+ destroy_tpmring(tp);
-+ return err;
+}
+
-+
-+static void destroy_tpmring(struct tpm_private *tp)
++static void xencons_close(struct tty_struct *tty, struct file *filp)
+{
-+ tpmif_set_connected_state(tp, 0);
++ unsigned long flags;
+
-+ if (tp->ring_ref != GRANT_INVALID_REF) {
-+ gnttab_end_foreign_access(tp->ring_ref, 0,
-+ (unsigned long)tp->tx);
-+ tp->ring_ref = GRANT_INVALID_REF;
-+ tp->tx = NULL;
++ if (DUMMY_TTY(tty))
++ return;
++
++ mutex_lock(&tty_mutex);
++
++ if (tty->count != 1) {
++ mutex_unlock(&tty_mutex);
++ return;
+ }
+
-+ if (tp->irq)
-+ unbind_from_irqhandler(tp->irq, tp);
++ /* Prevent other threads from re-opening this tty. */
++ set_bit(TTY_CLOSING, &tty->flags);
++ mutex_unlock(&tty_mutex);
+
-+ tp->irq = 0;
++ tty->closing = 1;
++ tty_wait_until_sent(tty, 0);
++ if (DRV(tty->driver)->flush_buffer != NULL)
++ DRV(tty->driver)->flush_buffer(tty);
++ if (tty->ldisc.flush_buffer != NULL)
++ tty->ldisc.flush_buffer(tty);
++ tty->closing = 0;
++ spin_lock_irqsave(&xencons_lock, flags);
++ xencons_tty = NULL;
++ spin_unlock_irqrestore(&xencons_lock, flags);
+}
+
++static struct tty_operations xencons_ops = {
++ .open = xencons_open,
++ .close = xencons_close,
++ .write = xencons_write,
++ .write_room = xencons_write_room,
++ .put_char = xencons_put_char,
++ .flush_chars = xencons_flush_chars,
++ .chars_in_buffer = xencons_chars_in_buffer,
++ .send_xchar = xencons_send_xchar,
++ .flush_buffer = xencons_flush_buffer,
++ .throttle = xencons_throttle,
++ .unthrottle = xencons_unthrottle,
++ .wait_until_sent = xencons_wait_until_sent,
++};
+
-+static int talk_to_backend(struct xenbus_device *dev,
-+ struct tpm_private *tp)
++static int __init xencons_init(void)
+{
-+ const char *message = NULL;
-+ int err;
-+ struct xenbus_transaction xbt;
++ int rc;
+
-+ err = setup_tpmring(dev, tp);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "setting up ring");
-+ goto out;
-+ }
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+again:
-+ err = xenbus_transaction_start(&xbt);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "starting transaction");
-+ goto destroy_tpmring;
++ if (xc_mode == XC_OFF)
++ return 0;
++
++ if (!is_initial_xendomain()) {
++ rc = xencons_ring_init();
++ if (rc)
++ return rc;
+ }
+
-+ err = xenbus_printf(xbt, dev->nodename,
-+ "ring-ref","%u", tp->ring_ref);
-+ if (err) {
-+ message = "writing ring-ref";
-+ goto abort_transaction;
++ xencons_driver = alloc_tty_driver((xc_mode == XC_TTY) ?
++ MAX_NR_CONSOLES : 1);
++ if (xencons_driver == NULL)
++ return -ENOMEM;
++
++ DRV(xencons_driver)->name = "xencons";
++ DRV(xencons_driver)->major = TTY_MAJOR;
++ DRV(xencons_driver)->type = TTY_DRIVER_TYPE_SERIAL;
++ DRV(xencons_driver)->subtype = SERIAL_TYPE_NORMAL;
++ DRV(xencons_driver)->init_termios = tty_std_termios;
++ DRV(xencons_driver)->flags =
++ TTY_DRIVER_REAL_RAW |
++ TTY_DRIVER_RESET_TERMIOS;
++ DRV(xencons_driver)->termios = xencons_termios;
++ DRV(xencons_driver)->termios_locked = xencons_termios_locked;
++
++ switch (xc_mode) {
++ case XC_XVC:
++ DRV(xencons_driver)->name = "xvc";
++ DRV(xencons_driver)->major = XEN_XVC_MAJOR;
++ DRV(xencons_driver)->minor_start = XEN_XVC_MINOR;
++ DRV(xencons_driver)->name_base = xc_num;
++ break;
++ case XC_SERIAL:
++ DRV(xencons_driver)->name = "ttyS";
++ DRV(xencons_driver)->minor_start = 64 + xc_num;
++ DRV(xencons_driver)->name_base = xc_num;
++ break;
++ default:
++ DRV(xencons_driver)->name = "tty";
++ DRV(xencons_driver)->minor_start = 1;
++ DRV(xencons_driver)->name_base = 1;
++ break;
+ }
+
-+ err = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
-+ irq_to_evtchn_port(tp->irq));
-+ if (err) {
-+ message = "writing event-channel";
-+ goto abort_transaction;
++ tty_set_operations(xencons_driver, &xencons_ops);
++
++ if ((rc = tty_register_driver(DRV(xencons_driver))) != 0) {
++ printk("WARNING: Failed to register Xen virtual "
++ "console driver as '%s%d'\n",
++ DRV(xencons_driver)->name,
++ DRV(xencons_driver)->name_base);
++ put_tty_driver(xencons_driver);
++ xencons_driver = NULL;
++ return rc;
+ }
+
-+ err = xenbus_transaction_end(xbt, 0);
-+ if (err == -EAGAIN)
-+ goto again;
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "completing transaction");
-+ goto destroy_tpmring;
++ if (is_initial_xendomain()) {
++ xencons_priv_irq = bind_virq_to_irqhandler(
++ VIRQ_CONSOLE,
++ 0,
++ xencons_priv_interrupt,
++ 0,
++ "console",
++ NULL);
++ BUG_ON(xencons_priv_irq < 0);
+ }
+
-+ xenbus_switch_state(dev, XenbusStateConnected);
++ printk("Xen virtual console successfully installed as %s%d\n",
++ DRV(xencons_driver)->name, xc_num);
+
+ return 0;
-+
-+abort_transaction:
-+ xenbus_transaction_end(xbt, 1);
-+ if (message)
-+ xenbus_dev_error(dev, err, "%s", message);
-+destroy_tpmring:
-+ destroy_tpmring(tp);
-+out:
-+ return err;
+}
+
-+/**
-+ * Callback received when the backend's state changes.
++module_init(xencons_init);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/console/xencons_ring.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/console/xencons_ring.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,143 @@
++/*
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
-+static void backend_changed(struct xenbus_device *dev,
-+ enum xenbus_state backend_state)
-+{
-+ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
-+ DPRINTK("\n");
+
-+ switch (backend_state) {
-+ case XenbusStateInitialising:
-+ case XenbusStateInitWait:
-+ case XenbusStateInitialised:
-+ case XenbusStateUnknown:
-+ break;
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/errno.h>
++#include <linux/signal.h>
++#include <linux/sched.h>
++#include <linux/interrupt.h>
++#include <linux/tty.h>
++#include <linux/tty_flip.h>
++#include <linux/serial.h>
++#include <linux/major.h>
++#include <linux/ptrace.h>
++#include <linux/ioport.h>
++#include <linux/mm.h>
++#include <linux/slab.h>
+
-+ case XenbusStateConnected:
-+ tpmif_set_connected_state(tp, 1);
-+ break;
++#include <asm/hypervisor.h>
++#include <xen/evtchn.h>
++#include <xen/xencons.h>
++#include <linux/wait.h>
++#include <linux/interrupt.h>
++#include <linux/sched.h>
++#include <linux/err.h>
++#include <xen/interface/io/console.h>
++
++static int xencons_irq;
+
-+ case XenbusStateClosing:
-+ tpmif_set_connected_state(tp, 0);
-+ xenbus_frontend_closed(dev);
-+ break;
++static inline struct xencons_interface *xencons_interface(void)
++{
++ return mfn_to_virt(xen_start_info->console.domU.mfn);
++}
+
-+ case XenbusStateClosed:
-+ tpmif_set_connected_state(tp, 0);
-+ if (tp->is_suspended == 0)
-+ device_unregister(&dev->dev);
-+ xenbus_frontend_closed(dev);
-+ break;
-+ }
++static inline void notify_daemon(void)
++{
++ /* Use evtchn: this is called early, before irq is set up. */
++ notify_remote_via_evtchn(xen_start_info->console.domU.evtchn);
+}
+
-+static int tpmfront_probe(struct xenbus_device *dev,
-+ const struct xenbus_device_id *id)
++int xencons_ring_send(const char *data, unsigned len)
+{
-+ int err;
-+ int handle;
-+ struct tpm_private *tp = tpm_private_get();
++ int sent = 0;
++ struct xencons_interface *intf = xencons_interface();
++ XENCONS_RING_IDX cons, prod;
+
-+ if (!tp)
-+ return -ENOMEM;
++ cons = intf->out_cons;
++ prod = intf->out_prod;
++ mb();
++ BUG_ON((prod - cons) > sizeof(intf->out));
+
-+ tp->chip = init_vtpm(&dev->dev, tp);
-+ if (IS_ERR(tp->chip))
-+ return PTR_ERR(tp->chip);
++ while ((sent < len) && ((prod - cons) < sizeof(intf->out)))
++ intf->out[MASK_XENCONS_IDX(prod++, intf->out)] = data[sent++];
+
-+ err = xenbus_scanf(XBT_NIL, dev->nodename,
-+ "handle", "%i", &handle);
-+ if (XENBUS_EXIST_ERR(err))
-+ return err;
++ wmb();
++ intf->out_prod = prod;
+
-+ if (err < 0) {
-+ xenbus_dev_fatal(dev,err,"reading virtual-device");
-+ return err;
-+ }
++ notify_daemon();
+
-+ tp->dev = dev;
++ return sent;
++}
+
-+ err = talk_to_backend(dev, tp);
-+ if (err) {
-+ tpm_private_put();
-+ return err;
++static irqreturn_t handle_input(int irq, void *unused, struct pt_regs *regs)
++{
++ struct xencons_interface *intf = xencons_interface();
++ XENCONS_RING_IDX cons, prod;
++
++ cons = intf->in_cons;
++ prod = intf->in_prod;
++ mb();
++ BUG_ON((prod - cons) > sizeof(intf->in));
++
++ while (cons != prod) {
++ xencons_rx(intf->in+MASK_XENCONS_IDX(cons,intf->in), 1, regs);
++ cons++;
+ }
+
-+ return 0;
-+}
++ mb();
++ intf->in_cons = cons;
+
++ notify_daemon();
+
-+static int tpmfront_remove(struct xenbus_device *dev)
-+{
-+ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
-+ destroy_tpmring(tp);
-+ cleanup_vtpm(&dev->dev);
-+ return 0;
++ xencons_tx();
++
++ return IRQ_HANDLED;
+}
+
-+static int tpmfront_suspend(struct xenbus_device *dev)
++int xencons_ring_init(void)
+{
-+ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
-+ u32 ctr;
++ int irq;
+
-+ /* Take the lock, preventing any application from sending. */
-+ mutex_lock(&suspend_lock);
-+ tp->is_suspended = 1;
++ if (xencons_irq)
++ unbind_from_irqhandler(xencons_irq, NULL);
++ xencons_irq = 0;
+
-+ for (ctr = 0; atomic_read(&tp->tx_busy); ctr++) {
-+ if ((ctr % 10) == 0)
-+ printk("TPM-FE [INFO]: Waiting for outstanding "
-+ "request.\n");
-+ /* Wait for a request to be responded to. */
-+ interruptible_sleep_on_timeout(&tp->wait_q, 100);
++ if (!is_running_on_xen() ||
++ is_initial_xendomain() ||
++ !xen_start_info->console.domU.evtchn)
++ return -ENODEV;
++
++ irq = bind_caller_port_to_irqhandler(
++ xen_start_info->console.domU.evtchn,
++ handle_input, 0, "xencons", NULL);
++ if (irq < 0) {
++ printk(KERN_ERR "XEN console request irq failed %i\n", irq);
++ return irq;
+ }
+
-+ return 0;
-+}
++ xencons_irq = irq;
++
++ /* In case we have in-flight data after save/restore... */
++ notify_daemon();
+
-+static int tpmfront_suspend_finish(struct tpm_private *tp)
-+{
-+ tp->is_suspended = 0;
-+ /* Allow applications to send again. */
-+ mutex_unlock(&suspend_lock);
+ return 0;
+}
+
-+static int tpmfront_suspend_cancel(struct xenbus_device *dev)
++void xencons_resume(void)
+{
-+ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
-+ return tpmfront_suspend_finish(tp);
++ (void)xencons_ring_init();
+}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,14 @@
++#
++# Makefile for the linux kernel.
++#
+
-+static int tpmfront_resume(struct xenbus_device *dev)
++obj-y := evtchn.o gnttab.o features.o reboot.o machine_reboot.o firmware.o
++
++obj-$(CONFIG_PCI) += pci.o
++obj-$(CONFIG_PROC_FS) += xen_proc.o
++obj-$(CONFIG_SYS_HYPERVISOR) += hypervisor_sysfs.o
++obj-$(CONFIG_HOTPLUG_CPU) += cpu_hotplug.o
++obj-$(CONFIG_XEN_SYSFS) += xen_sysfs.o
++obj-$(CONFIG_XEN_SMPBOOT) += smpboot.o
++obj-$(CONFIG_KEXEC) += machine_kexec.o
++obj-$(CONFIG_XEN_XENCOMM) += xencomm.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/cpu_hotplug.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/cpu_hotplug.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,173 @@
++#include <linux/init.h>
++#include <linux/kernel.h>
++#include <linux/sched.h>
++#include <linux/notifier.h>
++#include <linux/cpu.h>
++#include <xen/cpu_hotplug.h>
++#include <xen/xenbus.h>
++
++/*
++ * Set of CPUs that remote admin software will allow us to bring online.
++ * Notified to us via xenbus.
++ */
++static cpumask_t xenbus_allowed_cpumask;
++
++/* Set of CPUs that local admin will allow us to bring online. */
++static cpumask_t local_allowed_cpumask = CPU_MASK_ALL;
++
++static int local_cpu_hotplug_request(void)
+{
-+ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
-+ destroy_tpmring(tp);
-+ return talk_to_backend(dev, tp);
++ /*
++ * We assume a CPU hotplug request comes from local admin if it is made
++ * via a userspace process (i.e., one with a real mm_struct).
++ */
++ return (current->mm != NULL);
+}
+
-+static int tpmif_connect(struct xenbus_device *dev,
-+ struct tpm_private *tp,
-+ domid_t domid)
++static void vcpu_hotplug(unsigned int cpu)
+{
+ int err;
++ char dir[32], state[32];
+
-+ tp->backend_id = domid;
++ if ((cpu >= NR_CPUS) || !cpu_possible(cpu))
++ return;
+
-+ err = bind_listening_port_to_irqhandler(
-+ domid, tpmif_int, SA_SAMPLE_RANDOM, "tpmif", tp);
-+ if (err <= 0) {
-+ WPRINTK("bind_listening_port_to_irqhandler failed "
-+ "(err=%d)\n", err);
-+ return err;
++ sprintf(dir, "cpu/%u", cpu);
++ err = xenbus_scanf(XBT_NIL, dir, "availability", "%s", state);
++ if (err != 1) {
++ printk(KERN_ERR "XENBUS: Unable to read cpu state\n");
++ return;
+ }
-+ tp->irq = err;
-+
-+ return 0;
-+}
-+
-+static struct xenbus_device_id tpmfront_ids[] = {
-+ { "vtpm" },
-+ { "" }
-+};
-+
-+static struct xenbus_driver tpmfront = {
-+ .name = "vtpm",
-+ .owner = THIS_MODULE,
-+ .ids = tpmfront_ids,
-+ .probe = tpmfront_probe,
-+ .remove = tpmfront_remove,
-+ .resume = tpmfront_resume,
-+ .otherend_changed = backend_changed,
-+ .suspend = tpmfront_suspend,
-+ .suspend_cancel = tpmfront_suspend_cancel,
-+};
+
-+static void __init init_tpm_xenbus(void)
-+{
-+ xenbus_register_frontend(&tpmfront);
++ if (strcmp(state, "online") == 0) {
++ cpu_set(cpu, xenbus_allowed_cpumask);
++ (void)cpu_up(cpu);
++ } else if (strcmp(state, "offline") == 0) {
++ cpu_clear(cpu, xenbus_allowed_cpumask);
++ (void)cpu_down(cpu);
++ } else {
++ printk(KERN_ERR "XENBUS: unknown state(%s) on CPU%d\n",
++ state, cpu);
++ }
+}
+
-+static int tpmif_allocate_tx_buffers(struct tpm_private *tp)
++static void handle_vcpu_hotplug_event(
++ struct xenbus_watch *watch, const char **vec, unsigned int len)
+{
-+ unsigned int i;
++ unsigned int cpu;
++ char *cpustr;
++ const char *node = vec[XS_WATCH_PATH];
+
-+ for (i = 0; i < TPMIF_TX_RING_SIZE; i++) {
-+ tp->tx_buffers[i] = tx_buffer_alloc();
-+ if (!tp->tx_buffers[i]) {
-+ tpmif_free_tx_buffers(tp);
-+ return -ENOMEM;
-+ }
++ if ((cpustr = strstr(node, "cpu/")) != NULL) {
++ sscanf(cpustr, "cpu/%u", &cpu);
++ vcpu_hotplug(cpu);
+ }
-+ return 0;
+}
+
-+static void tpmif_free_tx_buffers(struct tpm_private *tp)
++static int smpboot_cpu_notify(struct notifier_block *notifier,
++ unsigned long action, void *hcpu)
+{
-+ unsigned int i;
++ unsigned int cpu = (long)hcpu;
+
-+ for (i = 0; i < TPMIF_TX_RING_SIZE; i++)
-+ tx_buffer_free(tp->tx_buffers[i]);
++ /*
++ * We do this in a callback notifier rather than __cpu_disable()
++ * because local_cpu_hotplug_request() does not work in the latter
++ * as it's always executed from within a stopmachine kthread.
++ */
++ if ((action == CPU_DOWN_PREPARE) && local_cpu_hotplug_request())
++ cpu_clear(cpu, local_allowed_cpumask);
++
++ return NOTIFY_OK;
+}
+
-+static void tpmif_rx_action(unsigned long priv)
++static int setup_cpu_watcher(struct notifier_block *notifier,
++ unsigned long event, void *data)
+{
-+ struct tpm_private *tp = (struct tpm_private *)priv;
-+ int i = 0;
-+ unsigned int received;
-+ unsigned int offset = 0;
-+ u8 *buffer;
-+ tpmif_tx_request_t *tx = &tp->tx->ring[i].req;
-+
-+ atomic_set(&tp->tx_busy, 0);
-+ wake_up_interruptible(&tp->wait_q);
-+
-+ received = tx->size;
++ unsigned int i;
+
-+ buffer = kmalloc(received, GFP_ATOMIC);
-+ if (!buffer)
-+ return;
++ static struct xenbus_watch cpu_watch = {
++ .node = "cpu",
++ .callback = handle_vcpu_hotplug_event,
++ .flags = XBWF_new_thread };
++ (void)register_xenbus_watch(&cpu_watch);
+
-+ for (i = 0; i < TPMIF_TX_RING_SIZE && offset < received; i++) {
-+ struct tx_buffer *txb = tp->tx_buffers[i];
-+ tpmif_tx_request_t *tx;
-+ unsigned int tocopy;
++ if (!is_initial_xendomain()) {
++ for_each_possible_cpu(i)
++ vcpu_hotplug(i);
++ printk(KERN_INFO "Brought up %ld CPUs\n",
++ (long)num_online_cpus());
++ }
+
-+ tx = &tp->tx->ring[i].req;
-+ tocopy = tx->size;
-+ if (tocopy > PAGE_SIZE)
-+ tocopy = PAGE_SIZE;
++ return NOTIFY_DONE;
++}
+
-+ memcpy(&buffer[offset], txb->data, tocopy);
++static int __init setup_vcpu_hotplug_event(void)
++{
++ static struct notifier_block hotplug_cpu = {
++ .notifier_call = smpboot_cpu_notify };
++ static struct notifier_block xsn_cpu = {
++ .notifier_call = setup_cpu_watcher };
+
-+ gnttab_release_grant_reference(&gref_head, tx->ref);
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+ offset += tocopy;
-+ }
++ register_cpu_notifier(&hotplug_cpu);
++ register_xenstore_notifier(&xsn_cpu);
+
-+ vtpm_vd_recv(tp->chip, buffer, received, tp->tx_remember);
-+ kfree(buffer);
++ return 0;
+}
+
++arch_initcall(setup_vcpu_hotplug_event);
+
-+static irqreturn_t tpmif_int(int irq, void *tpm_priv, struct pt_regs *ptregs)
++int smp_suspend(void)
+{
-+ struct tpm_private *tp = tpm_priv;
-+ unsigned long flags;
++ unsigned int cpu;
++ int err;
+
-+ spin_lock_irqsave(&tp->tx_lock, flags);
-+ tpmif_rx_tasklet.data = (unsigned long)tp;
-+ tasklet_schedule(&tpmif_rx_tasklet);
-+ spin_unlock_irqrestore(&tp->tx_lock, flags);
++ for_each_online_cpu(cpu) {
++ if (cpu == 0)
++ continue;
++ err = cpu_down(cpu);
++ if (err) {
++ printk(KERN_CRIT "Failed to take all CPUs "
++ "down: %d.\n", err);
++ for_each_possible_cpu(cpu)
++ vcpu_hotplug(cpu);
++ return err;
++ }
++ }
+
-+ return IRQ_HANDLED;
++ return 0;
+}
+
-+
-+static int tpm_xmit(struct tpm_private *tp,
-+ const u8 * buf, size_t count, int isuserbuffer,
-+ void *remember)
++void smp_resume(void)
+{
-+ tpmif_tx_request_t *tx;
-+ TPMIF_RING_IDX i;
-+ unsigned int offset = 0;
++ unsigned int cpu;
+
-+ spin_lock_irq(&tp->tx_lock);
++ for_each_possible_cpu(cpu)
++ vcpu_hotplug(cpu);
++}
+
-+ if (unlikely(atomic_read(&tp->tx_busy))) {
-+ printk("tpm_xmit: There's an outstanding request/response "
-+ "on the way!\n");
-+ spin_unlock_irq(&tp->tx_lock);
-+ return -EBUSY;
-+ }
++int cpu_up_check(unsigned int cpu)
++{
++ int rc = 0;
+
-+ if (tp->is_connected != 1) {
-+ spin_unlock_irq(&tp->tx_lock);
-+ return -EIO;
++ if (local_cpu_hotplug_request()) {
++ cpu_set(cpu, local_allowed_cpumask);
++ if (!cpu_isset(cpu, xenbus_allowed_cpumask)) {
++ printk("%s: attempt to bring up CPU %u disallowed by "
++ "remote admin.\n", __FUNCTION__, cpu);
++ rc = -EBUSY;
++ }
++ } else if (!cpu_isset(cpu, local_allowed_cpumask) ||
++ !cpu_isset(cpu, xenbus_allowed_cpumask)) {
++ rc = -EBUSY;
+ }
+
-+ for (i = 0; count > 0 && i < TPMIF_TX_RING_SIZE; i++) {
-+ struct tx_buffer *txb = tp->tx_buffers[i];
-+ int copied;
++ return rc;
++}
+
-+ if (!txb) {
-+ DPRINTK("txb (i=%d) is NULL. buffers initilized?\n"
-+ "Not transmitting anything!\n", i);
-+ spin_unlock_irq(&tp->tx_lock);
-+ return -EFAULT;
-+ }
++void init_xenbus_allowed_cpumask(void)
++{
++ xenbus_allowed_cpumask = cpu_present_map;
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/evtchn.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/evtchn.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1129 @@
++/******************************************************************************
++ * evtchn.c
++ *
++ * Communication via Xen event channels.
++ *
++ * Copyright (c) 2002-2005, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ copied = tx_buffer_copy(txb, &buf[offset], count,
-+ isuserbuffer);
-+ if (copied < 0) {
-+ /* An error occurred */
-+ spin_unlock_irq(&tp->tx_lock);
-+ return copied;
-+ }
-+ count -= copied;
-+ offset += copied;
++#include <linux/module.h>
++#include <linux/irq.h>
++#include <linux/interrupt.h>
++#include <linux/sched.h>
++#include <linux/kernel_stat.h>
++#include <linux/version.h>
++#include <asm/atomic.h>
++#include <asm/system.h>
++#include <asm/ptrace.h>
++#include <asm/synch_bitops.h>
++#include <xen/evtchn.h>
++#include <xen/interface/event_channel.h>
++#include <xen/interface/physdev.h>
++#include <asm/hypervisor.h>
++#include <linux/mc146818rtc.h> /* RTC_IRQ */
+
-+ tx = &tp->tx->ring[i].req;
-+ tx->addr = virt_to_machine(txb->data);
-+ tx->size = txb->len;
++/*
++ * This lock protects updates to the following mapping and reference-count
++ * arrays. The lock does not need to be acquired to read the mapping tables.
++ */
++static DEFINE_SPINLOCK(irq_mapping_update_lock);
+
-+ DPRINTK("First 4 characters sent by TPM-FE are "
-+ "0x%02x 0x%02x 0x%02x 0x%02x\n",
-+ txb->data[0],txb->data[1],txb->data[2],txb->data[3]);
++/* IRQ <-> event-channel mappings. */
++static int evtchn_to_irq[NR_EVENT_CHANNELS] = {
++ [0 ... NR_EVENT_CHANNELS-1] = -1 };
+
-+ /* Get the granttable reference for this page. */
-+ tx->ref = gnttab_claim_grant_reference(&gref_head);
-+ if (tx->ref == -ENOSPC) {
-+ spin_unlock_irq(&tp->tx_lock);
-+ DPRINTK("Grant table claim reference failed in "
-+ "func:%s line:%d file:%s\n",
-+ __FUNCTION__, __LINE__, __FILE__);
-+ return -ENOSPC;
-+ }
-+ gnttab_grant_foreign_access_ref(tx->ref,
-+ tp->backend_id,
-+ virt_to_mfn(txb->data),
-+ 0 /*RW*/);
-+ wmb();
-+ }
++/* Packed IRQ information: binding type, sub-type index, and event channel. */
++static u32 irq_info[NR_IRQS];
+
-+ atomic_set(&tp->tx_busy, 1);
-+ tp->tx_remember = remember;
++/* Binding types. */
++enum {
++ IRQT_UNBOUND,
++ IRQT_PIRQ,
++ IRQT_VIRQ,
++ IRQT_IPI,
++ IRQT_LOCAL_PORT,
++ IRQT_CALLER_PORT,
++ _IRQT_COUNT
++};
+
-+ mb();
++#define _IRQT_BITS 4
++#define _EVTCHN_BITS 12
++#define _INDEX_BITS (32 - _IRQT_BITS - _EVTCHN_BITS)
+
-+ notify_remote_via_irq(tp->irq);
++/* Constructor for packed IRQ information. */
++static inline u32 mk_irq_info(u32 type, u32 index, u32 evtchn)
++{
++ BUILD_BUG_ON(_IRQT_COUNT > (1U << _IRQT_BITS));
+
-+ spin_unlock_irq(&tp->tx_lock);
-+ return offset;
-+}
++ BUILD_BUG_ON(NR_PIRQS > (1U << _INDEX_BITS));
++ BUILD_BUG_ON(NR_VIRQS > (1U << _INDEX_BITS));
++ BUILD_BUG_ON(NR_IPIS > (1U << _INDEX_BITS));
++ BUG_ON(index >> _INDEX_BITS);
+
++ BUILD_BUG_ON(NR_EVENT_CHANNELS > (1U << _EVTCHN_BITS));
+
-+static void tpmif_notify_upperlayer(struct tpm_private *tp)
-+{
-+ /* Notify upper layer about the state of the connection to the BE. */
-+ vtpm_vd_status(tp->chip, (tp->is_connected
-+ ? TPM_VD_STATUS_CONNECTED
-+ : TPM_VD_STATUS_DISCONNECTED));
++ return ((type << (32 - _IRQT_BITS)) | (index << _EVTCHN_BITS) | evtchn);
+}
+
++/* Convenient shorthand for packed representation of an unbound IRQ. */
++#define IRQ_UNBOUND mk_irq_info(IRQT_UNBOUND, 0, 0)
+
-+static void tpmif_set_connected_state(struct tpm_private *tp, u8 is_connected)
-+{
-+ /*
-+ * Don't notify upper layer if we are in suspend mode and
-+ * should disconnect - assumption is that we will resume
-+ * The mutex keeps apps from sending.
-+ */
-+ if (is_connected == 0 && tp->is_suspended == 1)
-+ return;
-+
-+ /*
-+ * Unlock the mutex if we are connected again
-+ * after being suspended - now resuming.
-+ * This also removes the suspend state.
-+ */
-+ if (is_connected == 1 && tp->is_suspended == 1)
-+ tpmfront_suspend_finish(tp);
++/*
++ * Accessors for packed IRQ information.
++ */
+
-+ if (is_connected != tp->is_connected) {
-+ tp->is_connected = is_connected;
-+ tpmif_notify_upperlayer(tp);
-+ }
++static inline unsigned int evtchn_from_irq(int irq)
++{
++ return irq_info[irq] & ((1U << _EVTCHN_BITS) - 1);
+}
+
++static inline unsigned int index_from_irq(int irq)
++{
++ return (irq_info[irq] >> _EVTCHN_BITS) & ((1U << _INDEX_BITS) - 1);
++}
+
++static inline unsigned int type_from_irq(int irq)
++{
++ return irq_info[irq] >> (32 - _IRQT_BITS);
++}
+
-+/* =================================================================
-+ * Initialization function.
-+ * =================================================================
-+ */
++/* IRQ <-> VIRQ mapping. */
++DEFINE_PER_CPU(int, virq_to_irq[NR_VIRQS]) = {[0 ... NR_VIRQS-1] = -1};
+
++/* IRQ <-> IPI mapping. */
++#ifndef NR_IPIS
++#define NR_IPIS 1
++#endif
++DEFINE_PER_CPU(int, ipi_to_irq[NR_IPIS]) = {[0 ... NR_IPIS-1] = -1};
+
-+static int __init tpmif_init(void)
-+{
-+ struct tpm_private *tp;
++/* Reference counts for bindings to IRQs. */
++static int irq_bindcount[NR_IRQS];
+
-+ if (is_initial_xendomain())
-+ return -EPERM;
++/* Bitmap indicating which PIRQs require Xen to be notified on unmask. */
++static DECLARE_BITMAP(pirq_needs_eoi, NR_PIRQS);
+
-+ tp = tpm_private_get();
-+ if (!tp)
-+ return -ENOMEM;
++#ifdef CONFIG_SMP
+
-+ IPRINTK("Initialising the vTPM driver.\n");
-+ if (gnttab_alloc_grant_references(TPMIF_TX_RING_SIZE,
-+ &gref_head) < 0) {
-+ tpm_private_put();
-+ return -EFAULT;
-+ }
++static u8 cpu_evtchn[NR_EVENT_CHANNELS];
++static unsigned long cpu_evtchn_mask[NR_CPUS][NR_EVENT_CHANNELS/BITS_PER_LONG];
+
-+ init_tpm_xenbus();
-+ return 0;
++static inline unsigned long active_evtchns(unsigned int cpu, shared_info_t *sh,
++ unsigned int idx)
++{
++ return (sh->evtchn_pending[idx] &
++ cpu_evtchn_mask[cpu][idx] &
++ ~sh->evtchn_mask[idx]);
+}
+
++static void bind_evtchn_to_cpu(unsigned int chn, unsigned int cpu)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ int irq = evtchn_to_irq[chn];
+
-+module_init(tpmif_init);
++ BUG_ON(!test_bit(chn, s->evtchn_mask));
+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/char/tty_io.c tmp-linux-2.6-xen.patch/drivers/char/tty_io.c
---- pristine-linux-2.6.18.2/drivers/char/tty_io.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/char/tty_io.c 2007-07-30 16:35:11.000000000 +0200
-@@ -130,6 +130,8 @@ LIST_HEAD(tty_drivers); /* linked list
- vt.c for deeply disgusting hack reasons */
- DEFINE_MUTEX(tty_mutex);
-
-+int console_use_vt = 1;
++ if (irq != -1)
++ set_native_irq_info(irq, cpumask_of_cpu(cpu));
+
- #ifdef CONFIG_UNIX98_PTYS
- extern struct tty_driver *ptm_driver; /* Unix98 pty masters; for /dev/ptmx */
- extern int pty_limit; /* Config limit on Unix98 ptys */
-@@ -2483,7 +2485,7 @@ retry_open:
- goto got_driver;
- }
- #ifdef CONFIG_VT
-- if (device == MKDEV(TTY_MAJOR,0)) {
-+ if (console_use_vt && (device == MKDEV(TTY_MAJOR,0))) {
- extern struct tty_driver *console_driver;
- driver = console_driver;
- index = fg_console;
-@@ -3909,6 +3911,8 @@ static int __init tty_init(void)
- #endif
-
- #ifdef CONFIG_VT
-+ if (!console_use_vt)
-+ goto out_vt;
- cdev_init(&vc0_cdev, &console_fops);
- if (cdev_add(&vc0_cdev, MKDEV(TTY_MAJOR, 0), 1) ||
- register_chrdev_region(MKDEV(TTY_MAJOR, 0), 1, "/dev/vc/0") < 0)
-@@ -3916,6 +3920,7 @@ static int __init tty_init(void)
- class_device_create(tty_class, NULL, MKDEV(TTY_MAJOR, 0), NULL, "tty0");
-
- vty_init();
-+ out_vt:
- #endif
- return 0;
- }
-diff -Nurp pristine-linux-2.6.18.2/drivers/firmware/Kconfig tmp-linux-2.6-xen.patch/drivers/firmware/Kconfig
---- pristine-linux-2.6.18.2/drivers/firmware/Kconfig 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/firmware/Kconfig 2007-07-30 16:35:11.000000000 +0200
-@@ -8,6 +8,7 @@ menu "Firmware Drivers"
- config EDD
- tristate "BIOS Enhanced Disk Drive calls determine boot disk"
- depends on !IA64
-+ depends on !XEN
- help
- Say Y or M here if you want to enable BIOS Enhanced Disk Drive
- Services real mode BIOS calls to determine which disk
-diff -Nurp pristine-linux-2.6.18.2/drivers/ide/ide-lib.c tmp-linux-2.6-xen.patch/drivers/ide/ide-lib.c
---- pristine-linux-2.6.18.2/drivers/ide/ide-lib.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/ide/ide-lib.c 2007-10-14 01:51:15.000000000 +0200
-@@ -408,10 +408,10 @@ void ide_toggle_bounce(ide_drive_t *driv
- {
- u64 addr = BLK_BOUNCE_HIGH; /* dma64_addr_t */
-
-- if (!PCI_DMA_BUS_IS_PHYS) {
-- addr = BLK_BOUNCE_ANY;
-- } else if (on && drive->media == ide_disk) {
-- if (HWIF(drive)->pci_dev)
-+ if (on && drive->media == ide_disk) {
-+ if (!PCI_DMA_BUS_IS_PHYS)
-+ addr = BLK_BOUNCE_ANY;
-+ else if (HWIF(drive)->pci_dev)
- addr = HWIF(drive)->pci_dev->dma_mask;
- }
-
-diff -Nurp pristine-linux-2.6.18.2/drivers/Makefile tmp-linux-2.6-xen.patch/drivers/Makefile
---- pristine-linux-2.6.18.2/drivers/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -31,6 +31,7 @@ obj-y += base/ block/ misc/ mfd/ net/
- obj-$(CONFIG_NUBUS) += nubus/
- obj-$(CONFIG_ATM) += atm/
- obj-$(CONFIG_PPC_PMAC) += macintosh/
-+obj-$(CONFIG_XEN) += xen/
- obj-$(CONFIG_IDE) += ide/
- obj-$(CONFIG_FC4) += fc4/
- obj-$(CONFIG_SCSI) += scsi/
-diff -Nurp pristine-linux-2.6.18.2/drivers/oprofile/buffer_sync.c tmp-linux-2.6-xen.patch/drivers/oprofile/buffer_sync.c
---- pristine-linux-2.6.18.2/drivers/oprofile/buffer_sync.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/oprofile/buffer_sync.c 2007-10-14 01:51:15.000000000 +0200
-@@ -6,6 +6,10 @@
- *
- * @author John Levon <levon at movementarian.org>
- *
-+ * Modified by Aravind Menon for Xen
-+ * These modifications are:
-+ * Copyright (C) 2005 Hewlett-Packard Co.
-+ *
- * This is the core of the buffer management. Each
- * CPU buffer is processed and entered into the
- * global event buffer. Such processing is necessary
-@@ -38,6 +42,7 @@ static cpumask_t marked_cpus = CPU_MASK_
- static DEFINE_SPINLOCK(task_mortuary);
- static void process_task_mortuary(void);
-
-+static int cpu_current_domain[NR_CPUS];
-
- /* Take ownership of the task struct and place it on the
- * list for processing. Only after two full buffer syncs
-@@ -146,6 +151,11 @@ static void end_sync(void)
- int sync_start(void)
- {
- int err;
++ clear_bit(chn, (unsigned long *)cpu_evtchn_mask[cpu_evtchn[chn]]);
++ set_bit(chn, (unsigned long *)cpu_evtchn_mask[cpu]);
++ cpu_evtchn[chn] = cpu;
++}
++
++static void init_evtchn_cpu_bindings(void)
++{
+ int i;
+
-+ for (i = 0; i < NR_CPUS; i++) {
-+ cpu_current_domain[i] = COORDINATOR_DOMAIN;
-+ }
-
- start_cpu_work();
-
-@@ -275,15 +285,31 @@ static void add_cpu_switch(int i)
- last_cookie = INVALID_COOKIE;
- }
-
--static void add_kernel_ctx_switch(unsigned int in_kernel)
-+static void add_cpu_mode_switch(unsigned int cpu_mode)
- {
- add_event_entry(ESCAPE_CODE);
-- if (in_kernel)
-- add_event_entry(KERNEL_ENTER_SWITCH_CODE);
-- else
-- add_event_entry(KERNEL_EXIT_SWITCH_CODE);
-+ switch (cpu_mode) {
-+ case CPU_MODE_USER:
-+ add_event_entry(USER_ENTER_SWITCH_CODE);
-+ break;
-+ case CPU_MODE_KERNEL:
-+ add_event_entry(KERNEL_ENTER_SWITCH_CODE);
-+ break;
-+ case CPU_MODE_XEN:
-+ add_event_entry(XEN_ENTER_SWITCH_CODE);
-+ break;
-+ default:
-+ break;
-+ }
- }
--
++ /* By default all event channels notify CPU#0. */
++ for (i = 0; i < NR_IRQS; i++)
++ set_native_irq_info(i, cpumask_of_cpu(0));
+
-+static void add_domain_switch(unsigned long domain_id)
++ memset(cpu_evtchn, 0, sizeof(cpu_evtchn));
++ memset(cpu_evtchn_mask[0], ~0, sizeof(cpu_evtchn_mask[0]));
++}
++
++static inline unsigned int cpu_from_evtchn(unsigned int evtchn)
+{
-+ add_event_entry(ESCAPE_CODE);
-+ add_event_entry(DOMAIN_SWITCH_CODE);
-+ add_event_entry(domain_id);
++ return cpu_evtchn[evtchn];
+}
+
- static void
- add_user_ctx_switch(struct task_struct const * task, unsigned long cookie)
- {
-@@ -348,9 +374,9 @@ static int add_us_sample(struct mm_struc
- * for later lookup from userspace.
- */
- static int
--add_sample(struct mm_struct * mm, struct op_sample * s, int in_kernel)
-+add_sample(struct mm_struct * mm, struct op_sample * s, int cpu_mode)
- {
-- if (in_kernel) {
-+ if (cpu_mode >= CPU_MODE_KERNEL) {
- add_sample_entry(s->eip, s->event);
- return 1;
- } else if (mm) {
-@@ -496,15 +522,21 @@ void sync_buffer(int cpu)
- struct mm_struct *mm = NULL;
- struct task_struct * new;
- unsigned long cookie = 0;
-- int in_kernel = 1;
-+ int cpu_mode = 1;
- unsigned int i;
- sync_buffer_state state = sb_buffer_start;
- unsigned long available;
-+ int domain_switch = 0;
-
- mutex_lock(&buffer_mutex);
-
- add_cpu_switch(cpu);
-
-+ /* We need to assign the first samples in this CPU buffer to the
-+ same domain that we were processing at the last sync_buffer */
-+ if (cpu_current_domain[cpu] != COORDINATOR_DOMAIN) {
-+ add_domain_switch(cpu_current_domain[cpu]);
-+ }
- /* Remember, only we can modify tail_pos */
-
- available = get_slots(cpu_buf);
-@@ -512,16 +544,18 @@ void sync_buffer(int cpu)
- for (i = 0; i < available; ++i) {
- struct op_sample * s = &cpu_buf->buffer[cpu_buf->tail_pos];
-
-- if (is_code(s->eip)) {
-- if (s->event <= CPU_IS_KERNEL) {
-- /* kernel/userspace switch */
-- in_kernel = s->event;
-+ if (is_code(s->eip) && !domain_switch) {
-+ if (s->event <= CPU_MODE_XEN) {
-+ /* xen/kernel/userspace switch */
-+ cpu_mode = s->event;
- if (state == sb_buffer_start)
- state = sb_sample_start;
-- add_kernel_ctx_switch(s->event);
-+ add_cpu_mode_switch(s->event);
- } else if (s->event == CPU_TRACE_BEGIN) {
- state = sb_bt_start;
- add_trace_begin();
-+ } else if (s->event == CPU_DOMAIN_SWITCH) {
-+ domain_switch = 1;
- } else {
- struct mm_struct * oldmm = mm;
-
-@@ -535,11 +569,21 @@ void sync_buffer(int cpu)
- add_user_ctx_switch(new, cookie);
- }
- } else {
-- if (state >= sb_bt_start &&
-- !add_sample(mm, s, in_kernel)) {
-- if (state == sb_bt_start) {
-- state = sb_bt_ignore;
-- atomic_inc(&oprofile_stats.bt_lost_no_mapping);
-+ if (domain_switch) {
-+ cpu_current_domain[cpu] = s->eip;
-+ add_domain_switch(s->eip);
-+ domain_switch = 0;
-+ } else {
-+ if (cpu_current_domain[cpu] !=
-+ COORDINATOR_DOMAIN) {
-+ add_sample_entry(s->eip, s->event);
-+ }
-+ else if (state >= sb_bt_start &&
-+ !add_sample(mm, s, cpu_mode)) {
-+ if (state == sb_bt_start) {
-+ state = sb_bt_ignore;
-+ atomic_inc(&oprofile_stats.bt_lost_no_mapping);
-+ }
- }
- }
- }
-@@ -548,6 +592,11 @@ void sync_buffer(int cpu)
- }
- release_mm(mm);
-
-+ /* We reset domain to COORDINATOR at each CPU switch */
-+ if (cpu_current_domain[cpu] != COORDINATOR_DOMAIN) {
-+ add_domain_switch(COORDINATOR_DOMAIN);
-+ }
++#else
+
- mark_done(cpu);
-
- mutex_unlock(&buffer_mutex);
-diff -Nurp pristine-linux-2.6.18.2/drivers/oprofile/cpu_buffer.c tmp-linux-2.6-xen.patch/drivers/oprofile/cpu_buffer.c
---- pristine-linux-2.6.18.2/drivers/oprofile/cpu_buffer.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/oprofile/cpu_buffer.c 2007-10-14 01:51:15.000000000 +0200
-@@ -6,6 +6,10 @@
- *
- * @author John Levon <levon at movementarian.org>
- *
-+ * Modified by Aravind Menon for Xen
-+ * These modifications are:
-+ * Copyright (C) 2005 Hewlett-Packard Co.
-+ *
- * Each CPU has a local buffer that stores PC value/event
- * pairs. We also log context switches when we notice them.
- * Eventually each CPU's buffer is processed into the global
-@@ -34,6 +38,8 @@ static void wq_sync_buffer(void *);
- #define DEFAULT_TIMER_EXPIRE (HZ / 10)
- static int work_enabled;
-
-+static int32_t current_domain = COORDINATOR_DOMAIN;
++static inline unsigned long active_evtchns(unsigned int cpu, shared_info_t *sh,
++ unsigned int idx)
++{
++ return (sh->evtchn_pending[idx] & ~sh->evtchn_mask[idx]);
++}
+
- void free_cpu_buffers(void)
- {
- int i;
-@@ -57,7 +63,7 @@ int alloc_cpu_buffers(void)
- goto fail;
-
- b->last_task = NULL;
-- b->last_is_kernel = -1;
-+ b->last_cpu_mode = -1;
- b->tracing = 0;
- b->buffer_size = buffer_size;
- b->tail_pos = 0;
-@@ -113,7 +119,7 @@ void cpu_buffer_reset(struct oprofile_cp
- * collected will populate the buffer with proper
- * values to initialize the buffer
- */
-- cpu_buf->last_is_kernel = -1;
-+ cpu_buf->last_cpu_mode = -1;
- cpu_buf->last_task = NULL;
- }
-
-@@ -163,13 +169,13 @@ add_code(struct oprofile_cpu_buffer * bu
- * because of the head/tail separation of the writer and reader
- * of the CPU buffer.
- *
-- * is_kernel is needed because on some architectures you cannot
-+ * cpu_mode is needed because on some architectures you cannot
- * tell if you are in kernel or user space simply by looking at
-- * pc. We tag this in the buffer by generating kernel enter/exit
-- * events whenever is_kernel changes
-+ * pc. We tag this in the buffer by generating kernel/user (and xen)
-+ * enter events whenever cpu_mode changes
- */
- static int log_sample(struct oprofile_cpu_buffer * cpu_buf, unsigned long pc,
-- int is_kernel, unsigned long event)
-+ int cpu_mode, unsigned long event)
- {
- struct task_struct * task;
-
-@@ -180,18 +186,18 @@ static int log_sample(struct oprofile_cp
- return 0;
- }
-
-- is_kernel = !!is_kernel;
--
- task = current;
-
- /* notice a switch from user->kernel or vice versa */
-- if (cpu_buf->last_is_kernel != is_kernel) {
-- cpu_buf->last_is_kernel = is_kernel;
-- add_code(cpu_buf, is_kernel);
-+ if (cpu_buf->last_cpu_mode != cpu_mode) {
-+ cpu_buf->last_cpu_mode = cpu_mode;
-+ add_code(cpu_buf, cpu_mode);
- }
--
-+
- /* notice a task switch */
-- if (cpu_buf->last_task != task) {
-+ /* if not processing other domain samples */
-+ if ((cpu_buf->last_task != task) &&
-+ (current_domain == COORDINATOR_DOMAIN)) {
- cpu_buf->last_task = task;
- add_code(cpu_buf, (unsigned long)task);
- }
-@@ -275,6 +281,25 @@ void oprofile_add_trace(unsigned long pc
- add_sample(cpu_buf, pc, 0);
- }
-
-+int oprofile_add_domain_switch(int32_t domain_id)
++static void bind_evtchn_to_cpu(unsigned int chn, unsigned int cpu)
+{
-+ struct oprofile_cpu_buffer * cpu_buf = &cpu_buffer[smp_processor_id()];
++}
+
-+ /* should have space for switching into and out of domain
-+ (2 slots each) plus one sample and one cpu mode switch */
-+ if (((nr_available_slots(cpu_buf) < 6) &&
-+ (domain_id != COORDINATOR_DOMAIN)) ||
-+ (nr_available_slots(cpu_buf) < 2))
-+ return 0;
++static void init_evtchn_cpu_bindings(void)
++{
++}
+
-+ add_code(cpu_buf, CPU_DOMAIN_SWITCH);
-+ add_sample(cpu_buf, domain_id, 0);
++static inline unsigned int cpu_from_evtchn(unsigned int evtchn)
++{
++ return 0;
++}
+
-+ current_domain = domain_id;
++#endif
+
-+ return 1;
++/* Upcall to generic IRQ layer. */
++#ifdef CONFIG_X86
++extern fastcall unsigned int do_IRQ(struct pt_regs *regs);
++void __init xen_init_IRQ(void);
++void __init init_IRQ(void)
++{
++ irq_ctx_init(0);
++ xen_init_IRQ();
+}
++#if defined (__i386__)
++static inline void exit_idle(void) {}
++#define IRQ_REG orig_eax
++#elif defined (__x86_64__)
++#include <asm/idle.h>
++#define IRQ_REG orig_rax
++#endif
++#define do_IRQ(irq, regs) do { \
++ (regs)->IRQ_REG = ~(irq); \
++ do_IRQ((regs)); \
++} while (0)
++#endif
+
- /*
- * This serves to avoid cpu buffer overflow, and makes sure
- * the task mortuary progresses
-diff -Nurp pristine-linux-2.6.18.2/drivers/oprofile/cpu_buffer.h tmp-linux-2.6-xen.patch/drivers/oprofile/cpu_buffer.h
---- pristine-linux-2.6.18.2/drivers/oprofile/cpu_buffer.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/oprofile/cpu_buffer.h 2007-10-14 01:51:15.000000000 +0200
-@@ -36,7 +36,7 @@ struct oprofile_cpu_buffer {
- volatile unsigned long tail_pos;
- unsigned long buffer_size;
- struct task_struct * last_task;
-- int last_is_kernel;
-+ int last_cpu_mode;
- int tracing;
- struct op_sample * buffer;
- unsigned long sample_received;
-@@ -51,7 +51,10 @@ extern struct oprofile_cpu_buffer cpu_bu
- void cpu_buffer_reset(struct oprofile_cpu_buffer * cpu_buf);
-
- /* transient events for the CPU buffer -> event buffer */
--#define CPU_IS_KERNEL 1
--#define CPU_TRACE_BEGIN 2
-+#define CPU_MODE_USER 0
-+#define CPU_MODE_KERNEL 1
-+#define CPU_MODE_XEN 2
-+#define CPU_TRACE_BEGIN 3
-+#define CPU_DOMAIN_SWITCH 4
-
- #endif /* OPROFILE_CPU_BUFFER_H */
-diff -Nurp pristine-linux-2.6.18.2/drivers/oprofile/event_buffer.h tmp-linux-2.6-xen.patch/drivers/oprofile/event_buffer.h
---- pristine-linux-2.6.18.2/drivers/oprofile/event_buffer.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/oprofile/event_buffer.h 2007-10-14 01:51:15.000000000 +0200
-@@ -29,15 +29,20 @@ void wake_up_buffer_waiter(void);
- #define CPU_SWITCH_CODE 2
- #define COOKIE_SWITCH_CODE 3
- #define KERNEL_ENTER_SWITCH_CODE 4
--#define KERNEL_EXIT_SWITCH_CODE 5
-+#define USER_ENTER_SWITCH_CODE 5
- #define MODULE_LOADED_CODE 6
- #define CTX_TGID_CODE 7
- #define TRACE_BEGIN_CODE 8
- #define TRACE_END_CODE 9
-+#define XEN_ENTER_SWITCH_CODE 10
-+#define DOMAIN_SWITCH_CODE 11
-
- #define INVALID_COOKIE ~0UL
- #define NO_COOKIE 0UL
-
-+/* Constant used to refer to coordinator domain (Xen) */
-+#define COORDINATOR_DOMAIN -1
++/* Xen will never allocate port zero for any purpose. */
++#define VALID_EVTCHN(chn) ((chn) != 0)
+
- /* add data to the event buffer */
- void add_event_entry(unsigned long data);
-
-diff -Nurp pristine-linux-2.6.18.2/drivers/oprofile/oprof.c tmp-linux-2.6-xen.patch/drivers/oprofile/oprof.c
---- pristine-linux-2.6.18.2/drivers/oprofile/oprof.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/oprofile/oprof.c 2007-10-14 01:51:15.000000000 +0200
-@@ -5,6 +5,10 @@
- * @remark Read the file COPYING
- *
- * @author John Levon <levon at movementarian.org>
-+ *
-+ * Modified by Aravind Menon for Xen
-+ * These modifications are:
-+ * Copyright (C) 2005 Hewlett-Packard Co.
- */
-
- #include <linux/kernel.h>
-@@ -19,7 +23,7 @@
- #include "cpu_buffer.h"
- #include "buffer_sync.h"
- #include "oprofile_stats.h"
--
++/*
++ * Force a proper event-channel callback from Xen after clearing the
++ * callback mask. We do this in a very simple manner, by making a call
++ * down into Xen. The pending flag will be checked by Xen on return.
++ */
++void force_evtchn_callback(void)
++{
++ VOID(HYPERVISOR_xen_version(0, NULL));
++}
++/* Not a GPL symbol: used in ubiquitous macros, so too restrictive. */
++EXPORT_SYMBOL(force_evtchn_callback);
+
- struct oprofile_operations oprofile_ops;
-
- unsigned long oprofile_started;
-@@ -33,6 +37,32 @@ static DEFINE_MUTEX(start_mutex);
- */
- static int timer = 0;
-
-+int oprofile_set_active(int active_domains[], unsigned int adomains)
++static DEFINE_PER_CPU(unsigned int, upcall_count) = { 0 };
++static DEFINE_PER_CPU(unsigned int, last_processed_l1i) = { BITS_PER_LONG - 1 };
++static DEFINE_PER_CPU(unsigned int, last_processed_l2i) = { BITS_PER_LONG - 1 };
++
++/* NB. Interrupts are disabled on entry. */
++asmlinkage void evtchn_do_upcall(struct pt_regs *regs)
+{
-+ int err;
++ unsigned long l1, l2;
++ unsigned long masked_l1, masked_l2;
++ unsigned int l1i, l2i, port, count;
++ int irq;
++ unsigned int cpu = smp_processor_id();
++ shared_info_t *s = HYPERVISOR_shared_info;
++ vcpu_info_t *vcpu_info = &s->vcpu_info[cpu];
+
-+ if (!oprofile_ops.set_active)
-+ return -EINVAL;
+
-+ mutex_lock(&start_mutex);
-+ err = oprofile_ops.set_active(active_domains, adomains);
-+ mutex_unlock(&start_mutex);
-+ return err;
-+}
++ do {
++ /* Avoid a callback storm when we reenable delivery. */
++ vcpu_info->evtchn_upcall_pending = 0;
+
-+int oprofile_set_passive(int passive_domains[], unsigned int pdomains)
-+{
-+ int err;
++ /* Nested invocations bail immediately. */
++ if (unlikely(per_cpu(upcall_count, cpu)++))
++ return;
++
++#ifndef CONFIG_X86 /* No need for a barrier -- XCHG is a barrier on x86. */
++ /* Clear master flag /before/ clearing selector flag. */
++ wmb();
++#endif
++ l1 = xchg(&vcpu_info->evtchn_pending_sel, 0);
++
++ l1i = per_cpu(last_processed_l1i, cpu);
++ l2i = per_cpu(last_processed_l2i, cpu);
++
++ while (l1 != 0) {
++
++ l1i = (l1i + 1) % BITS_PER_LONG;
++ masked_l1 = l1 & ((~0UL) << l1i);
++
++ if (masked_l1 == 0) { /* if we masked out all events, wrap around to the beginning */
++ l1i = BITS_PER_LONG - 1;
++ l2i = BITS_PER_LONG - 1;
++ continue;
++ }
++ l1i = __ffs(masked_l1);
+
-+ if (!oprofile_ops.set_passive)
-+ return -EINVAL;
++ do {
++ l2 = active_evtchns(cpu, s, l1i);
+
-+ mutex_lock(&start_mutex);
-+ err = oprofile_ops.set_passive(passive_domains, pdomains);
-+ mutex_unlock(&start_mutex);
-+ return err;
-+}
++ l2i = (l2i + 1) % BITS_PER_LONG;
++ masked_l2 = l2 & ((~0UL) << l2i);
+
- int oprofile_setup(void)
- {
- int err;
-diff -Nurp pristine-linux-2.6.18.2/drivers/oprofile/oprof.h tmp-linux-2.6-xen.patch/drivers/oprofile/oprof.h
---- pristine-linux-2.6.18.2/drivers/oprofile/oprof.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/oprofile/oprof.h 2007-10-14 01:51:15.000000000 +0200
-@@ -35,5 +35,8 @@ void oprofile_create_files(struct super_
- void oprofile_timer_init(struct oprofile_operations * ops);
-
- int oprofile_set_backtrace(unsigned long depth);
++ if (masked_l2 == 0) { /* if we masked out all events, move on */
++ l2i = BITS_PER_LONG - 1;
++ break;
++ }
+
-+int oprofile_set_active(int active_domains[], unsigned int adomains);
-+int oprofile_set_passive(int passive_domains[], unsigned int pdomains);
-
- #endif /* OPROF_H */
-diff -Nurp pristine-linux-2.6.18.2/drivers/oprofile/oprofile_files.c tmp-linux-2.6-xen.patch/drivers/oprofile/oprofile_files.c
---- pristine-linux-2.6.18.2/drivers/oprofile/oprofile_files.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/oprofile/oprofile_files.c 2007-10-14 01:51:15.000000000 +0200
-@@ -5,15 +5,21 @@
- * @remark Read the file COPYING
- *
- * @author John Levon <levon at movementarian.org>
-+ *
-+ * Modified by Aravind Menon for Xen
-+ * These modifications are:
-+ * Copyright (C) 2005 Hewlett-Packard Co.
- */
-
- #include <linux/fs.h>
- #include <linux/oprofile.h>
-+#include <asm/uaccess.h>
-+#include <linux/ctype.h>
-
- #include "event_buffer.h"
- #include "oprofile_stats.h"
- #include "oprof.h"
--
++ l2i = __ffs(masked_l2);
+
- unsigned long fs_buffer_size = 131072;
- unsigned long fs_cpu_buffer_size = 8192;
- unsigned long fs_buffer_watershed = 32768; /* FIXME: tune */
-@@ -117,11 +123,202 @@ static ssize_t dump_write(struct file *
- static struct file_operations dump_fops = {
- .write = dump_write,
- };
--
++ /* process port */
++ port = (l1i * BITS_PER_LONG) + l2i;
++ if ((irq = evtchn_to_irq[port]) != -1)
++ do_IRQ(irq, regs);
++ else {
++ exit_idle();
++ evtchn_device_upcall(port);
++ }
+
-+#define TMPBUFSIZE 512
++ /* if this is the final port processed, we'll pick up here+1 next time */
++ per_cpu(last_processed_l1i, cpu) = l1i;
++ per_cpu(last_processed_l2i, cpu) = l2i;
++
++ } while (l2i != BITS_PER_LONG - 1);
++
++ l2 = active_evtchns(cpu, s, l1i);
++ if (l2 == 0) /* we handled all ports, so we can clear the selector bit */
++ l1 &= ~(1UL << l1i);
+
-+static unsigned int adomains = 0;
-+static int active_domains[MAX_OPROF_DOMAINS + 1];
-+static DEFINE_MUTEX(adom_mutex);
++ }
+
-+static ssize_t adomain_write(struct file * file, char const __user * buf,
-+ size_t count, loff_t * offset)
++ /* If there were nested callbacks then we have more to do. */
++ count = per_cpu(upcall_count, cpu);
++ per_cpu(upcall_count, cpu) = 0;
++ } while (unlikely(count != 1));
++}
++
++static int find_unbound_irq(void)
+{
-+ char *tmpbuf;
-+ char *startp, *endp;
-+ int i;
-+ unsigned long val;
-+ ssize_t retval = count;
-+
-+ if (*offset)
-+ return -EINVAL;
-+ if (count > TMPBUFSIZE - 1)
-+ return -EINVAL;
++ static int warned;
++ int dynirq, irq;
+
-+ if (!(tmpbuf = kmalloc(TMPBUFSIZE, GFP_KERNEL)))
-+ return -ENOMEM;
++ for (dynirq = 0; dynirq < NR_DYNIRQS; dynirq++) {
++ irq = dynirq_to_irq(dynirq);
++ if (irq_bindcount[irq] == 0)
++ return irq;
++ }
+
-+ if (copy_from_user(tmpbuf, buf, count)) {
-+ kfree(tmpbuf);
-+ return -EFAULT;
++ if (!warned) {
++ warned = 1;
++ printk(KERN_WARNING "No available IRQ to bind to: "
++ "increase NR_DYNIRQS.\n");
+ }
-+ tmpbuf[count] = 0;
+
-+ mutex_lock(&adom_mutex);
++ return -ENOSPC;
++}
+
-+ startp = tmpbuf;
-+ /* Parse one more than MAX_OPROF_DOMAINS, for easy error checking */
-+ for (i = 0; i <= MAX_OPROF_DOMAINS; i++) {
-+ val = simple_strtoul(startp, &endp, 0);
-+ if (endp == startp)
-+ break;
-+ while (ispunct(*endp) || isspace(*endp))
-+ endp++;
-+ active_domains[i] = val;
-+ if (active_domains[i] != val)
-+ /* Overflow, force error below */
-+ i = MAX_OPROF_DOMAINS + 1;
-+ startp = endp;
-+ }
-+ /* Force error on trailing junk */
-+ adomains = *startp ? MAX_OPROF_DOMAINS + 1 : i;
++static int bind_caller_port_to_irq(unsigned int caller_port)
++{
++ int irq;
+
-+ kfree(tmpbuf);
++ spin_lock(&irq_mapping_update_lock);
+
-+ if (adomains > MAX_OPROF_DOMAINS
-+ || oprofile_set_active(active_domains, adomains)) {
-+ adomains = 0;
-+ retval = -EINVAL;
++ if ((irq = evtchn_to_irq[caller_port]) == -1) {
++ if ((irq = find_unbound_irq()) < 0)
++ goto out;
++
++ evtchn_to_irq[caller_port] = irq;
++ irq_info[irq] = mk_irq_info(IRQT_CALLER_PORT, 0, caller_port);
+ }
+
-+ mutex_unlock(&adom_mutex);
-+ return retval;
++ irq_bindcount[irq]++;
++
++ out:
++ spin_unlock(&irq_mapping_update_lock);
++ return irq;
+}
+
-+static ssize_t adomain_read(struct file * file, char __user * buf,
-+ size_t count, loff_t * offset)
++static int bind_local_port_to_irq(unsigned int local_port)
+{
-+ char * tmpbuf;
-+ size_t len;
-+ int i;
-+ ssize_t retval;
++ int irq;
+
-+ if (!(tmpbuf = kmalloc(TMPBUFSIZE, GFP_KERNEL)))
-+ return -ENOMEM;
++ spin_lock(&irq_mapping_update_lock);
+
-+ mutex_lock(&adom_mutex);
++ BUG_ON(evtchn_to_irq[local_port] != -1);
+
-+ len = 0;
-+ for (i = 0; i < adomains; i++)
-+ len += snprintf(tmpbuf + len,
-+ len < TMPBUFSIZE ? TMPBUFSIZE - len : 0,
-+ "%u ", active_domains[i]);
-+ WARN_ON(len > TMPBUFSIZE);
-+ if (len != 0 && len <= TMPBUFSIZE)
-+ tmpbuf[len-1] = '\n';
++ if ((irq = find_unbound_irq()) < 0) {
++ struct evtchn_close close = { .port = local_port };
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_close, &close))
++ BUG();
++ goto out;
++ }
+
-+ mutex_unlock(&adom_mutex);
++ evtchn_to_irq[local_port] = irq;
++ irq_info[irq] = mk_irq_info(IRQT_LOCAL_PORT, 0, local_port);
++ irq_bindcount[irq]++;
+
-+ retval = simple_read_from_buffer(buf, count, offset, tmpbuf, len);
++ out:
++ spin_unlock(&irq_mapping_update_lock);
++ return irq;
++}
+
-+ kfree(tmpbuf);
-+ return retval;
++static int bind_listening_port_to_irq(unsigned int remote_domain)
++{
++ struct evtchn_alloc_unbound alloc_unbound;
++ int err;
++
++ alloc_unbound.dom = DOMID_SELF;
++ alloc_unbound.remote_dom = remote_domain;
++
++ err = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
++ &alloc_unbound);
++
++ return err ? : bind_local_port_to_irq(alloc_unbound.port);
+}
+
++static int bind_interdomain_evtchn_to_irq(unsigned int remote_domain,
++ unsigned int remote_port)
++{
++ struct evtchn_bind_interdomain bind_interdomain;
++ int err;
+
-+static struct file_operations active_domain_ops = {
-+ .read = adomain_read,
-+ .write = adomain_write,
-+};
++ bind_interdomain.remote_dom = remote_domain;
++ bind_interdomain.remote_port = remote_port;
+
-+static unsigned int pdomains = 0;
-+static int passive_domains[MAX_OPROF_DOMAINS];
-+static DEFINE_MUTEX(pdom_mutex);
++ err = HYPERVISOR_event_channel_op(EVTCHNOP_bind_interdomain,
++ &bind_interdomain);
+
-+static ssize_t pdomain_write(struct file * file, char const __user * buf,
-+ size_t count, loff_t * offset)
++ return err ? : bind_local_port_to_irq(bind_interdomain.local_port);
++}
++
++static int bind_virq_to_irq(unsigned int virq, unsigned int cpu)
+{
-+ char *tmpbuf;
-+ char *startp, *endp;
-+ int i;
-+ unsigned long val;
-+ ssize_t retval = count;
-+
-+ if (*offset)
-+ return -EINVAL;
-+ if (count > TMPBUFSIZE - 1)
-+ return -EINVAL;
++ struct evtchn_bind_virq bind_virq;
++ int evtchn, irq;
+
-+ if (!(tmpbuf = kmalloc(TMPBUFSIZE, GFP_KERNEL)))
-+ return -ENOMEM;
++ spin_lock(&irq_mapping_update_lock);
+
-+ if (copy_from_user(tmpbuf, buf, count)) {
-+ kfree(tmpbuf);
-+ return -EFAULT;
-+ }
-+ tmpbuf[count] = 0;
++ if ((irq = per_cpu(virq_to_irq, cpu)[virq]) == -1) {
++ if ((irq = find_unbound_irq()) < 0)
++ goto out;
+
-+ mutex_lock(&pdom_mutex);
++ bind_virq.virq = virq;
++ bind_virq.vcpu = cpu;
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
++ &bind_virq) != 0)
++ BUG();
++ evtchn = bind_virq.port;
+
-+ startp = tmpbuf;
-+ /* Parse one more than MAX_OPROF_DOMAINS, for easy error checking */
-+ for (i = 0; i <= MAX_OPROF_DOMAINS; i++) {
-+ val = simple_strtoul(startp, &endp, 0);
-+ if (endp == startp)
-+ break;
-+ while (ispunct(*endp) || isspace(*endp))
-+ endp++;
-+ passive_domains[i] = val;
-+ if (passive_domains[i] != val)
-+ /* Overflow, force error below */
-+ i = MAX_OPROF_DOMAINS + 1;
-+ startp = endp;
-+ }
-+ /* Force error on trailing junk */
-+ pdomains = *startp ? MAX_OPROF_DOMAINS + 1 : i;
++ evtchn_to_irq[evtchn] = irq;
++ irq_info[irq] = mk_irq_info(IRQT_VIRQ, virq, evtchn);
+
-+ kfree(tmpbuf);
++ per_cpu(virq_to_irq, cpu)[virq] = irq;
+
-+ if (pdomains > MAX_OPROF_DOMAINS
-+ || oprofile_set_passive(passive_domains, pdomains)) {
-+ pdomains = 0;
-+ retval = -EINVAL;
++ bind_evtchn_to_cpu(evtchn, cpu);
+ }
+
-+ mutex_unlock(&pdom_mutex);
-+ return retval;
++ irq_bindcount[irq]++;
++
++ out:
++ spin_unlock(&irq_mapping_update_lock);
++ return irq;
+}
+
-+static ssize_t pdomain_read(struct file * file, char __user * buf,
-+ size_t count, loff_t * offset)
++static int bind_ipi_to_irq(unsigned int ipi, unsigned int cpu)
+{
-+ char * tmpbuf;
-+ size_t len;
-+ int i;
-+ ssize_t retval;
++ struct evtchn_bind_ipi bind_ipi;
++ int evtchn, irq;
+
-+ if (!(tmpbuf = kmalloc(TMPBUFSIZE, GFP_KERNEL)))
-+ return -ENOMEM;
++ spin_lock(&irq_mapping_update_lock);
+
-+ mutex_lock(&pdom_mutex);
++ if ((irq = per_cpu(ipi_to_irq, cpu)[ipi]) == -1) {
++ if ((irq = find_unbound_irq()) < 0)
++ goto out;
+
-+ len = 0;
-+ for (i = 0; i < pdomains; i++)
-+ len += snprintf(tmpbuf + len,
-+ len < TMPBUFSIZE ? TMPBUFSIZE - len : 0,
-+ "%u ", passive_domains[i]);
-+ WARN_ON(len > TMPBUFSIZE);
-+ if (len != 0 && len <= TMPBUFSIZE)
-+ tmpbuf[len-1] = '\n';
++ bind_ipi.vcpu = cpu;
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_ipi,
++ &bind_ipi) != 0)
++ BUG();
++ evtchn = bind_ipi.port;
+
-+ mutex_unlock(&pdom_mutex);
++ evtchn_to_irq[evtchn] = irq;
++ irq_info[irq] = mk_irq_info(IRQT_IPI, ipi, evtchn);
+
-+ retval = simple_read_from_buffer(buf, count, offset, tmpbuf, len);
++ per_cpu(ipi_to_irq, cpu)[ipi] = irq;
+
-+ kfree(tmpbuf);
-+ return retval;
++ bind_evtchn_to_cpu(evtchn, cpu);
++ }
++
++ irq_bindcount[irq]++;
++
++ out:
++ spin_unlock(&irq_mapping_update_lock);
++ return irq;
+}
+
-+static struct file_operations passive_domain_ops = {
-+ .read = pdomain_read,
-+ .write = pdomain_write,
-+};
++static void unbind_from_irq(unsigned int irq)
++{
++ struct evtchn_close close;
++ unsigned int cpu;
++ int evtchn = evtchn_from_irq(irq);
+
- void oprofile_create_files(struct super_block * sb, struct dentry * root)
- {
- oprofilefs_create_file(sb, root, "enable", &enable_fops);
- oprofilefs_create_file_perm(sb, root, "dump", &dump_fops, 0666);
-+ oprofilefs_create_file(sb, root, "active_domains", &active_domain_ops);
-+ oprofilefs_create_file(sb, root, "passive_domains", &passive_domain_ops);
- oprofilefs_create_file(sb, root, "buffer", &event_buffer_fops);
- oprofilefs_create_ulong(sb, root, "buffer_size", &fs_buffer_size);
- oprofilefs_create_ulong(sb, root, "buffer_watershed", &fs_buffer_watershed);
-diff -Nurp pristine-linux-2.6.18.2/drivers/pci/Kconfig tmp-linux-2.6-xen.patch/drivers/pci/Kconfig
---- pristine-linux-2.6.18.2/drivers/pci/Kconfig 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/pci/Kconfig 2007-07-30 16:35:11.000000000 +0200
-@@ -5,6 +5,7 @@ config PCI_MSI
- bool "Message Signaled Interrupts (MSI and MSI-X)"
- depends on PCI
- depends on (X86_LOCAL_APIC && X86_IO_APIC) || IA64
-+ depends on !XEN
- help
- This allows device drivers to enable MSI (Message Signaled
- Interrupts). Message Signaled Interrupts enable a device to
-diff -Nurp pristine-linux-2.6.18.2/drivers/serial/Kconfig tmp-linux-2.6-xen.patch/drivers/serial/Kconfig
---- pristine-linux-2.6.18.2/drivers/serial/Kconfig 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/serial/Kconfig 2007-07-30 16:35:11.000000000 +0200
-@@ -11,6 +11,7 @@ menu "Serial drivers"
- config SERIAL_8250
- tristate "8250/16550 and compatible serial support"
- depends on (BROKEN || !SPARC)
-+ depends on !XEN_DISABLE_SERIAL
- select SERIAL_CORE
- ---help---
- This selects whether you want to include the driver for the standard
-diff -Nurp pristine-linux-2.6.18.2/drivers/video/console/Kconfig tmp-linux-2.6-xen.patch/drivers/video/console/Kconfig
---- pristine-linux-2.6.18.2/drivers/video/console/Kconfig 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/video/console/Kconfig 2007-07-30 16:35:11.000000000 +0200
-@@ -53,6 +53,7 @@ config VGACON_SOFT_SCROLLBACK_SIZE
- config VIDEO_SELECT
- bool "Video mode selection support"
- depends on X86 && VGA_CONSOLE
-+ depends on !XEN
- ---help---
- This enables support for text mode selection on kernel startup. If
- you want to take advantage of some high-resolution text mode your
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/balloon/balloon.c tmp-linux-2.6-xen.patch/drivers/xen/balloon/balloon.c
---- pristine-linux-2.6.18.2/drivers/xen/balloon/balloon.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/balloon/balloon.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,663 @@
-+/******************************************************************************
-+ * balloon.c
-+ *
-+ * Xen balloon driver - enables returning/claiming memory to/from Xen.
-+ *
-+ * Copyright (c) 2003, B Dragovic
-+ * Copyright (c) 2003-2004, M Williamson, K Fraser
-+ * Copyright (c) 2005 Dan M. Smith, IBM Corporation
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ spin_lock(&irq_mapping_update_lock);
+
-+#include <linux/kernel.h>
-+#include <linux/module.h>
-+#include <linux/sched.h>
-+#include <linux/errno.h>
-+#include <linux/mm.h>
-+#include <linux/mman.h>
-+#include <linux/smp_lock.h>
-+#include <linux/pagemap.h>
-+#include <linux/bootmem.h>
-+#include <linux/highmem.h>
-+#include <linux/vmalloc.h>
-+#include <linux/mutex.h>
-+#include <xen/xen_proc.h>
-+#include <asm/hypervisor.h>
-+#include <xen/balloon.h>
-+#include <xen/interface/memory.h>
-+#include <asm/maddr.h>
-+#include <asm/page.h>
-+#include <asm/pgalloc.h>
-+#include <asm/pgtable.h>
-+#include <asm/uaccess.h>
-+#include <asm/tlb.h>
-+#include <linux/highmem.h>
-+#include <linux/list.h>
-+#include <xen/xenbus.h>
-+#include "common.h"
++ if ((--irq_bindcount[irq] == 0) && VALID_EVTCHN(evtchn)) {
++ close.port = evtchn;
++ if ((type_from_irq(irq) != IRQT_CALLER_PORT) &&
++ HYPERVISOR_event_channel_op(EVTCHNOP_close, &close))
++ BUG();
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
-+#endif
++ switch (type_from_irq(irq)) {
++ case IRQT_VIRQ:
++ per_cpu(virq_to_irq, cpu_from_evtchn(evtchn))
++ [index_from_irq(irq)] = -1;
++ break;
++ case IRQT_IPI:
++ per_cpu(ipi_to_irq, cpu_from_evtchn(evtchn))
++ [index_from_irq(irq)] = -1;
++ break;
++ default:
++ break;
++ }
+
-+#ifdef CONFIG_PROC_FS
-+static struct proc_dir_entry *balloon_pde;
-+#endif
++ /* Closed ports are implicitly re-bound to VCPU0. */
++ bind_evtchn_to_cpu(evtchn, 0);
+
-+static DEFINE_MUTEX(balloon_mutex);
++ evtchn_to_irq[evtchn] = -1;
++ irq_info[irq] = IRQ_UNBOUND;
+
-+/*
-+ * Protects atomic reservation decrease/increase against concurrent increases.
-+ * Also protects non-atomic updates of current_pages and driver_pages, and
-+ * balloon lists.
-+ */
-+DEFINE_SPINLOCK(balloon_lock);
++ /* Zap stats across IRQ changes of use. */
++ for_each_possible_cpu(cpu)
++ kstat_cpu(cpu).irqs[irq] = 0;
++ }
+
-+struct balloon_stats balloon_stats;
++ spin_unlock(&irq_mapping_update_lock);
++}
+
-+/* We increase/decrease in batches which fit in a page */
-+static unsigned long frame_list[PAGE_SIZE / sizeof(unsigned long)];
++int bind_caller_port_to_irqhandler(
++ unsigned int caller_port,
++ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id)
++{
++ int irq, retval;
+
-+/* VM /proc information for memory */
-+extern unsigned long totalram_pages;
++ irq = bind_caller_port_to_irq(caller_port);
++ if (irq < 0)
++ return irq;
+
-+/* List of ballooned pages, threaded through the mem_map array. */
-+static LIST_HEAD(ballooned_pages);
++ retval = request_irq(irq, handler, irqflags, devname, dev_id);
++ if (retval != 0) {
++ unbind_from_irq(irq);
++ return retval;
++ }
+
-+/* Main work function, always executed in process context. */
-+static void balloon_process(void *unused);
-+static DECLARE_WORK(balloon_worker, balloon_process, NULL);
-+static struct timer_list balloon_timer;
++ return irq;
++}
++EXPORT_SYMBOL_GPL(bind_caller_port_to_irqhandler);
+
-+/* When ballooning out (allocating memory to return to Xen) we don't really
-+ want the kernel to try too hard since that can trigger the oom killer. */
-+#define GFP_BALLOON \
-+ (GFP_HIGHUSER | __GFP_NOWARN | __GFP_NORETRY | __GFP_NOMEMALLOC)
++int bind_listening_port_to_irqhandler(
++ unsigned int remote_domain,
++ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id)
++{
++ int irq, retval;
++
++ irq = bind_listening_port_to_irq(remote_domain);
++ if (irq < 0)
++ return irq;
++
++ retval = request_irq(irq, handler, irqflags, devname, dev_id);
++ if (retval != 0) {
++ unbind_from_irq(irq);
++ return retval;
++ }
++
++ return irq;
++}
++EXPORT_SYMBOL_GPL(bind_listening_port_to_irqhandler);
++
++int bind_interdomain_evtchn_to_irqhandler(
++ unsigned int remote_domain,
++ unsigned int remote_port,
++ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id)
++{
++ int irq, retval;
+
-+#define PAGE_TO_LIST(p) (&(p)->lru)
-+#define LIST_TO_PAGE(l) list_entry((l), struct page, lru)
-+#define UNLIST_PAGE(p) \
-+ do { \
-+ list_del(PAGE_TO_LIST(p)); \
-+ PAGE_TO_LIST(p)->next = NULL; \
-+ PAGE_TO_LIST(p)->prev = NULL; \
-+ } while(0)
++ irq = bind_interdomain_evtchn_to_irq(remote_domain, remote_port);
++ if (irq < 0)
++ return irq;
+
-+#define IPRINTK(fmt, args...) \
-+ printk(KERN_INFO "xen_mem: " fmt, ##args)
-+#define WPRINTK(fmt, args...) \
-+ printk(KERN_WARNING "xen_mem: " fmt, ##args)
++ retval = request_irq(irq, handler, irqflags, devname, dev_id);
++ if (retval != 0) {
++ unbind_from_irq(irq);
++ return retval;
++ }
+
-+/* balloon_append: add the given page to the balloon. */
-+static void balloon_append(struct page *page)
++ return irq;
++}
++EXPORT_SYMBOL_GPL(bind_interdomain_evtchn_to_irqhandler);
++
++int bind_virq_to_irqhandler(
++ unsigned int virq,
++ unsigned int cpu,
++ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id)
+{
-+ /* Lowmem is re-populated first, so highmem pages go at list tail. */
-+ if (PageHighMem(page)) {
-+ list_add_tail(PAGE_TO_LIST(page), &ballooned_pages);
-+ bs.balloon_high++;
-+ } else {
-+ list_add(PAGE_TO_LIST(page), &ballooned_pages);
-+ bs.balloon_low++;
++ int irq, retval;
++
++ irq = bind_virq_to_irq(virq, cpu);
++ if (irq < 0)
++ return irq;
++
++ retval = request_irq(irq, handler, irqflags, devname, dev_id);
++ if (retval != 0) {
++ unbind_from_irq(irq);
++ return retval;
+ }
++
++ return irq;
+}
++EXPORT_SYMBOL_GPL(bind_virq_to_irqhandler);
+
-+/* balloon_retrieve: rescue a page from the balloon, if it is not empty. */
-+static struct page *balloon_retrieve(void)
++int bind_ipi_to_irqhandler(
++ unsigned int ipi,
++ unsigned int cpu,
++ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id)
+{
-+ struct page *page;
-+
-+ if (list_empty(&ballooned_pages))
-+ return NULL;
++ int irq, retval;
+
-+ page = LIST_TO_PAGE(ballooned_pages.next);
-+ UNLIST_PAGE(page);
++ irq = bind_ipi_to_irq(ipi, cpu);
++ if (irq < 0)
++ return irq;
+
-+ if (PageHighMem(page))
-+ bs.balloon_high--;
-+ else
-+ bs.balloon_low--;
++ retval = request_irq(irq, handler, irqflags, devname, dev_id);
++ if (retval != 0) {
++ unbind_from_irq(irq);
++ return retval;
++ }
+
-+ return page;
++ return irq;
+}
++EXPORT_SYMBOL_GPL(bind_ipi_to_irqhandler);
+
-+static struct page *balloon_first_page(void)
++void unbind_from_irqhandler(unsigned int irq, void *dev_id)
+{
-+ if (list_empty(&ballooned_pages))
-+ return NULL;
-+ return LIST_TO_PAGE(ballooned_pages.next);
++ free_irq(irq, dev_id);
++ unbind_from_irq(irq);
+}
++EXPORT_SYMBOL_GPL(unbind_from_irqhandler);
+
-+static struct page *balloon_next_page(struct page *page)
++#ifdef CONFIG_SMP
++void rebind_evtchn_to_cpu(int port, unsigned int cpu)
+{
-+ struct list_head *next = PAGE_TO_LIST(page)->next;
-+ if (next == &ballooned_pages)
-+ return NULL;
-+ return LIST_TO_PAGE(next);
++ struct evtchn_bind_vcpu ebv = { .port = port, .vcpu = cpu };
++ int masked;
++
++ masked = test_and_set_evtchn_mask(port);
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_vcpu, &ebv) == 0)
++ bind_evtchn_to_cpu(port, cpu);
++ if (!masked)
++ unmask_evtchn(port);
+}
+
-+static void balloon_alarm(unsigned long unused)
++static void rebind_irq_to_cpu(unsigned int irq, unsigned int tcpu)
+{
-+ schedule_work(&balloon_worker);
++ int evtchn = evtchn_from_irq(irq);
++
++ if (VALID_EVTCHN(evtchn))
++ rebind_evtchn_to_cpu(evtchn, tcpu);
+}
+
-+static unsigned long current_target(void)
++static void set_affinity_irq(unsigned int irq, cpumask_t dest)
+{
-+ unsigned long target = min(bs.target_pages, bs.hard_limit);
-+ if (target > (bs.current_pages + bs.balloon_low + bs.balloon_high))
-+ target = bs.current_pages + bs.balloon_low + bs.balloon_high;
-+ return target;
++ unsigned tcpu = first_cpu(dest);
++ rebind_irq_to_cpu(irq, tcpu);
+}
++#endif
+
-+static int increase_reservation(unsigned long nr_pages)
++int resend_irq_on_evtchn(unsigned int irq)
+{
-+ unsigned long pfn, i, flags;
-+ struct page *page;
-+ long rc;
-+ struct xen_memory_reservation reservation = {
-+ .address_bits = 0,
-+ .extent_order = 0,
-+ .domid = DOMID_SELF
-+ };
-+
-+ if (nr_pages > ARRAY_SIZE(frame_list))
-+ nr_pages = ARRAY_SIZE(frame_list);
-+
-+ balloon_lock(flags);
-+
-+ page = balloon_first_page();
-+ for (i = 0; i < nr_pages; i++) {
-+ BUG_ON(page == NULL);
-+ frame_list[i] = page_to_pfn(page);;
-+ page = balloon_next_page(page);
-+ }
-+
-+ set_xen_guest_handle(reservation.extent_start, frame_list);
-+ reservation.nr_extents = nr_pages;
-+ rc = HYPERVISOR_memory_op(
-+ XENMEM_populate_physmap, &reservation);
-+ if (rc < nr_pages) {
-+ if (rc > 0) {
-+ int ret;
-+
-+ /* We hit the Xen hard limit: reprobe. */
-+ reservation.nr_extents = rc;
-+ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
-+ &reservation);
-+ BUG_ON(ret != rc);
-+ }
-+ if (rc >= 0)
-+ bs.hard_limit = (bs.current_pages + rc -
-+ bs.driver_pages);
-+ goto out;
-+ }
-+
-+ for (i = 0; i < nr_pages; i++) {
-+ page = balloon_retrieve();
-+ BUG_ON(page == NULL);
-+
-+ pfn = page_to_pfn(page);
-+ BUG_ON(!xen_feature(XENFEAT_auto_translated_physmap) &&
-+ phys_to_machine_mapping_valid(pfn));
++ int masked, evtchn = evtchn_from_irq(irq);
++ shared_info_t *s = HYPERVISOR_shared_info;
+
-+ set_phys_to_machine(pfn, frame_list[i]);
++ if (!VALID_EVTCHN(evtchn))
++ return 1;
+
-+#ifdef CONFIG_XEN
-+ /* Link back into the page tables if not highmem. */
-+ if (pfn < max_low_pfn) {
-+ int ret;
-+ ret = HYPERVISOR_update_va_mapping(
-+ (unsigned long)__va(pfn << PAGE_SHIFT),
-+ pfn_pte_ma(frame_list[i], PAGE_KERNEL),
-+ 0);
-+ BUG_ON(ret);
-+ }
-+#endif
++ masked = test_and_set_evtchn_mask(evtchn);
++ synch_set_bit(evtchn, s->evtchn_pending);
++ if (!masked)
++ unmask_evtchn(evtchn);
+
-+ /* Relinquish the page back to the allocator. */
-+ ClearPageReserved(page);
-+ init_page_count(page);
-+ __free_page(page);
-+ }
++ return 1;
++}
+
-+ bs.current_pages += nr_pages;
-+ totalram_pages = bs.current_pages;
++/*
++ * Interface to generic handling in irq.c
++ */
+
-+ out:
-+ balloon_unlock(flags);
++static unsigned int startup_dynirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
+
++ if (VALID_EVTCHN(evtchn))
++ unmask_evtchn(evtchn);
+ return 0;
+}
+
-+static int decrease_reservation(unsigned long nr_pages)
++static void shutdown_dynirq(unsigned int irq)
+{
-+ unsigned long pfn, i, flags;
-+ struct page *page;
-+ void *v;
-+ int need_sleep = 0;
-+ int ret;
-+ struct xen_memory_reservation reservation = {
-+ .address_bits = 0,
-+ .extent_order = 0,
-+ .domid = DOMID_SELF
-+ };
-+
-+ if (nr_pages > ARRAY_SIZE(frame_list))
-+ nr_pages = ARRAY_SIZE(frame_list);
-+
-+ for (i = 0; i < nr_pages; i++) {
-+ if ((page = alloc_page(GFP_BALLOON)) == NULL) {
-+ nr_pages = i;
-+ need_sleep = 1;
-+ break;
-+ }
-+
-+ pfn = page_to_pfn(page);
-+ frame_list[i] = pfn_to_mfn(pfn);
++ int evtchn = evtchn_from_irq(irq);
+
-+ if (!PageHighMem(page)) {
-+ v = phys_to_virt(pfn << PAGE_SHIFT);
-+ scrub_pages(v, 1);
-+#ifdef CONFIG_XEN
-+ ret = HYPERVISOR_update_va_mapping(
-+ (unsigned long)v, __pte_ma(0), 0);
-+ BUG_ON(ret);
-+#endif
-+ }
-+#ifdef CONFIG_XEN_SCRUB_PAGES
-+ else {
-+ v = kmap(page);
-+ scrub_pages(v, 1);
-+ kunmap(page);
-+ }
-+#endif
-+ }
++ if (VALID_EVTCHN(evtchn))
++ mask_evtchn(evtchn);
++}
+
-+#ifdef CONFIG_XEN
-+ /* Ensure that ballooned highmem pages don't have kmaps. */
-+ kmap_flush_unused();
-+ flush_tlb_all();
-+#endif
++static void enable_dynirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
+
-+ balloon_lock(flags);
++ if (VALID_EVTCHN(evtchn))
++ unmask_evtchn(evtchn);
++}
+
-+ /* No more mappings: invalidate P2M and add to balloon. */
-+ for (i = 0; i < nr_pages; i++) {
-+ pfn = mfn_to_pfn(frame_list[i]);
-+ set_phys_to_machine(pfn, INVALID_P2M_ENTRY);
-+ balloon_append(pfn_to_page(pfn));
-+ }
++static void disable_dynirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
+
-+ set_xen_guest_handle(reservation.extent_start, frame_list);
-+ reservation.nr_extents = nr_pages;
-+ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation, &reservation);
-+ BUG_ON(ret != nr_pages);
++ if (VALID_EVTCHN(evtchn))
++ mask_evtchn(evtchn);
++}
+
-+ bs.current_pages -= nr_pages;
-+ totalram_pages = bs.current_pages;
++static void ack_dynirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
+
-+ balloon_unlock(flags);
++ move_native_irq(irq);
+
-+ return need_sleep;
++ if (VALID_EVTCHN(evtchn)) {
++ mask_evtchn(evtchn);
++ clear_evtchn(evtchn);
++ }
+}
+
-+/*
-+ * We avoid multiple worker processes conflicting via the balloon mutex.
-+ * We may of course race updates of the target counts (which are protected
-+ * by the balloon lock), or with changes to the Xen hard limit, but we will
-+ * recover from these in time.
-+ */
-+static void balloon_process(void *unused)
++static void end_dynirq(unsigned int irq)
+{
-+ int need_sleep = 0;
-+ long credit;
-+
-+ mutex_lock(&balloon_mutex);
++ int evtchn = evtchn_from_irq(irq);
+
-+ do {
-+ credit = current_target() - bs.current_pages;
-+ if (credit > 0)
-+ need_sleep = (increase_reservation(credit) != 0);
-+ if (credit < 0)
-+ need_sleep = (decrease_reservation(-credit) != 0);
++ if (VALID_EVTCHN(evtchn) && !(irq_desc[irq].status & IRQ_DISABLED))
++ unmask_evtchn(evtchn);
++}
+
-+#ifndef CONFIG_PREEMPT
-+ if (need_resched())
-+ schedule();
++static struct hw_interrupt_type dynirq_type = {
++ .typename = "Dynamic-irq",
++ .startup = startup_dynirq,
++ .shutdown = shutdown_dynirq,
++ .enable = enable_dynirq,
++ .disable = disable_dynirq,
++ .ack = ack_dynirq,
++ .end = end_dynirq,
++#ifdef CONFIG_SMP
++ .set_affinity = set_affinity_irq,
+#endif
-+ } while ((credit != 0) && !need_sleep);
-+
-+ /* Schedule more work if there is some still to be done. */
-+ if (current_target() != bs.current_pages)
-+ mod_timer(&balloon_timer, jiffies + HZ);
-+
-+ mutex_unlock(&balloon_mutex);
-+}
++ .retrigger = resend_irq_on_evtchn,
++};
+
-+/* Resets the Xen limit, sets new target, and kicks off processing. */
-+void balloon_set_new_target(unsigned long target)
++void evtchn_register_pirq(int irq)
+{
-+ /* No need for lock. Not read-modify-write updates. */
-+ bs.hard_limit = ~0UL;
-+ bs.target_pages = target;
-+ schedule_work(&balloon_worker);
++ irq_info[irq] = mk_irq_info(IRQT_PIRQ, irq, 0);
+}
+
-+static struct xenbus_watch target_watch =
-+{
-+ .node = "memory/target"
-+};
++#ifndef CONFIG_X86_IO_APIC
++#undef IO_APIC_IRQ
++#define IO_APIC_IRQ(irq) ((irq) >= pirq_to_irq(16))
++#endif
+
-+/* React to a change in the target key */
-+static void watch_target(struct xenbus_watch *watch,
-+ const char **vec, unsigned int len)
++int evtchn_map_pirq(int irq, int xen_pirq)
+{
-+ unsigned long long new_target;
-+ int err;
++ if (irq < 0) {
++ static DEFINE_SPINLOCK(irq_alloc_lock);
+
-+ err = xenbus_scanf(XBT_NIL, "memory", "target", "%llu", &new_target);
-+ if (err != 1) {
-+ /* This is ok (for domain0 at least) - so just return */
-+ return;
++ irq = pirq_to_irq(NR_PIRQS - 1);
++ spin_lock(&irq_alloc_lock);
++ do {
++ if (!IO_APIC_IRQ(irq))
++ continue;
++ if (!index_from_irq(irq)) {
++ BUG_ON(type_from_irq(irq) != IRQT_UNBOUND);
++ irq_info[irq] = mk_irq_info(IRQT_PIRQ,
++ xen_pirq, 0);
++ break;
++ }
++ } while (--irq);
++ spin_unlock(&irq_alloc_lock);
++ if (irq < pirq_to_irq(16))
++ return -ENOSPC;
++ } else if (!xen_pirq) {
++ if (unlikely(type_from_irq(irq) != IRQT_PIRQ))
++ return -EINVAL;
++ irq_info[irq] = IRQ_UNBOUND;
++ return 0;
++ } else if (type_from_irq(irq) != IRQT_PIRQ
++ || index_from_irq(irq) != xen_pirq) {
++ printk(KERN_ERR "IRQ#%d is already mapped to %d:%u - "
++ "cannot map to PIRQ#%u\n",
++ irq, type_from_irq(irq), index_from_irq(irq), xen_pirq);
++ return -EINVAL;
+ }
-+
-+ /* The given memory/target value is in KiB, so it needs converting to
-+ * pages. PAGE_SHIFT converts bytes to pages, hence PAGE_SHIFT - 10.
-+ */
-+ balloon_set_new_target(new_target >> (PAGE_SHIFT - 10));
++ return index_from_irq(irq) ? irq : -EINVAL;
+}
+
-+static int balloon_init_watcher(struct notifier_block *notifier,
-+ unsigned long event,
-+ void *data)
++int evtchn_get_xen_pirq(int irq)
+{
-+ int err;
++ if (!IO_APIC_IRQ(irq))
++ return irq;
++ if (unlikely(type_from_irq(irq) != IRQT_PIRQ))
++ return 0;
++ return index_from_irq(irq);
++}
+
-+ err = register_xenbus_watch(&target_watch);
-+ if (err)
-+ printk(KERN_ERR "Failed to set balloon watcher\n");
++static inline void pirq_unmask_notify(int pirq)
++{
++ struct physdev_eoi eoi = { .irq = pirq };
++ if (unlikely(test_bit(pirq, pirq_needs_eoi)))
++ VOID(HYPERVISOR_physdev_op(PHYSDEVOP_eoi, &eoi));
++}
+
-+ return NOTIFY_DONE;
++static inline void pirq_query_unmask(int pirq)
++{
++ struct physdev_irq_status_query irq_status;
++ irq_status.irq = pirq;
++ if (HYPERVISOR_physdev_op(PHYSDEVOP_irq_status_query, &irq_status))
++ irq_status.flags = 0;
++ clear_bit(pirq, pirq_needs_eoi);
++ if (irq_status.flags & XENIRQSTAT_needs_eoi)
++ set_bit(pirq, pirq_needs_eoi);
+}
+
-+#ifdef CONFIG_PROC_FS
-+static int balloon_write(struct file *file, const char __user *buffer,
-+ unsigned long count, void *data)
++/*
++ * On startup, if there is no action associated with the IRQ then we are
++ * probing. In this case we should not share with others as it will confuse us.
++ */
++#define probing_irq(_irq) (irq_desc[(_irq)].action == NULL)
++
++static unsigned int startup_pirq(unsigned int irq)
+{
-+ char memstring[64], *endchar;
-+ unsigned long long target_bytes;
++ struct evtchn_bind_pirq bind_pirq;
++ int evtchn = evtchn_from_irq(irq);
+
-+ if (!capable(CAP_SYS_ADMIN))
-+ return -EPERM;
++ if (VALID_EVTCHN(evtchn))
++ goto out;
+
-+ if (count <= 1)
-+ return -EBADMSG; /* runt */
-+ if (count > sizeof(memstring))
-+ return -EFBIG; /* too long */
++ bind_pirq.pirq = evtchn_get_xen_pirq(irq);
++ /* NB. We are happy to share unless we are probing. */
++ bind_pirq.flags = probing_irq(irq) ? 0 : BIND_PIRQ__WILL_SHARE;
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_pirq, &bind_pirq) != 0) {
++ if (!probing_irq(irq))
++ printk(KERN_INFO "Failed to obtain physical IRQ %d\n",
++ irq);
++ return 0;
++ }
++ evtchn = bind_pirq.port;
+
-+ if (copy_from_user(memstring, buffer, count))
-+ return -EFAULT;
-+ memstring[sizeof(memstring)-1] = '\0';
++ pirq_query_unmask(irq_to_pirq(irq));
+
-+ target_bytes = memparse(memstring, &endchar);
-+ balloon_set_new_target(target_bytes >> PAGE_SHIFT);
++ evtchn_to_irq[evtchn] = irq;
++ bind_evtchn_to_cpu(evtchn, 0);
++ irq_info[irq] = mk_irq_info(IRQT_PIRQ, bind_pirq.pirq, evtchn);
+
-+ return count;
++ out:
++ unmask_evtchn(evtchn);
++ pirq_unmask_notify(irq_to_pirq(irq));
++
++ return 0;
+}
+
-+static int balloon_read(char *page, char **start, off_t off,
-+ int count, int *eof, void *data)
++static void shutdown_pirq(unsigned int irq)
+{
-+ int len;
++ struct evtchn_close close;
++ int evtchn = evtchn_from_irq(irq);
+
-+ len = sprintf(
-+ page,
-+ "Current allocation: %8lu kB\n"
-+ "Requested target: %8lu kB\n"
-+ "Low-mem balloon: %8lu kB\n"
-+ "High-mem balloon: %8lu kB\n"
-+ "Driver pages: %8lu kB\n"
-+ "Xen hard limit: ",
-+ PAGES2KB(bs.current_pages), PAGES2KB(bs.target_pages),
-+ PAGES2KB(bs.balloon_low), PAGES2KB(bs.balloon_high),
-+ PAGES2KB(bs.driver_pages));
++ if (!VALID_EVTCHN(evtchn))
++ return;
+
-+ if (bs.hard_limit != ~0UL)
-+ len += sprintf(page + len, "%8lu kB\n",
-+ PAGES2KB(bs.hard_limit));
-+ else
-+ len += sprintf(page + len, " ??? kB\n");
++ mask_evtchn(evtchn);
+
-+ *eof = 1;
-+ return len;
++ close.port = evtchn;
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_close, &close) != 0)
++ BUG();
++
++ bind_evtchn_to_cpu(evtchn, 0);
++ evtchn_to_irq[evtchn] = -1;
++ irq_info[irq] = mk_irq_info(IRQT_PIRQ, index_from_irq(irq), 0);
+}
-+#endif
+
-+static struct notifier_block xenstore_notifier;
++static void enable_pirq(unsigned int irq)
++{
++ startup_pirq(irq);
++}
+
-+static int __init balloon_init(void)
++static void disable_pirq(unsigned int irq)
+{
-+#if defined(CONFIG_X86) && defined(CONFIG_XEN)
-+ unsigned long pfn;
-+ struct page *page;
-+#endif
++}
+
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++static void ack_pirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
+
-+ IPRINTK("Initialising balloon driver.\n");
++ move_native_irq(irq);
+
-+#ifdef CONFIG_XEN
-+ bs.current_pages = min(xen_start_info->nr_pages, max_pfn);
-+ totalram_pages = bs.current_pages;
-+#else
-+ bs.current_pages = totalram_pages;
-+#endif
-+ bs.target_pages = bs.current_pages;
-+ bs.balloon_low = 0;
-+ bs.balloon_high = 0;
-+ bs.driver_pages = 0UL;
-+ bs.hard_limit = ~0UL;
++ if (VALID_EVTCHN(evtchn)) {
++ mask_evtchn(evtchn);
++ clear_evtchn(evtchn);
++ }
++}
+
-+ init_timer(&balloon_timer);
-+ balloon_timer.data = 0;
-+ balloon_timer.function = balloon_alarm;
-+
-+#ifdef CONFIG_PROC_FS
-+ if ((balloon_pde = create_xen_proc_entry("balloon", 0644)) == NULL) {
-+ WPRINTK("Unable to create /proc/xen/balloon.\n");
-+ return -1;
++static void end_pirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ if ((irq_desc[irq].status & (IRQ_DISABLED|IRQ_PENDING)) ==
++ (IRQ_DISABLED|IRQ_PENDING)) {
++ shutdown_pirq(irq);
++ } else if (VALID_EVTCHN(evtchn)) {
++ unmask_evtchn(evtchn);
++ pirq_unmask_notify(irq_to_pirq(irq));
+ }
++}
+
-+ balloon_pde->read_proc = balloon_read;
-+ balloon_pde->write_proc = balloon_write;
++static struct hw_interrupt_type pirq_type = {
++ .typename = "Phys-irq",
++ .startup = startup_pirq,
++ .shutdown = shutdown_pirq,
++ .enable = enable_pirq,
++ .disable = disable_pirq,
++ .ack = ack_pirq,
++ .end = end_pirq,
++#ifdef CONFIG_SMP
++ .set_affinity = set_affinity_irq,
+#endif
-+ balloon_sysfs_init();
++ .retrigger = resend_irq_on_evtchn,
++};
+
-+#if defined(CONFIG_X86) && defined(CONFIG_XEN)
-+ /* Initialise the balloon with excess memory space. */
-+ for (pfn = xen_start_info->nr_pages; pfn < max_pfn; pfn++) {
-+ page = pfn_to_page(pfn);
-+ if (!PageReserved(page))
-+ balloon_append(page);
-+ }
-+#endif
++int irq_ignore_unhandled(unsigned int irq)
++{
++ struct physdev_irq_status_query irq_status = { .irq = irq };
+
-+ target_watch.callback = watch_target;
-+ xenstore_notifier.notifier_call = balloon_init_watcher;
++ if (!is_running_on_xen())
++ return 0;
+
-+ register_xenstore_notifier(&xenstore_notifier);
-+
-+ return 0;
++ if (HYPERVISOR_physdev_op(PHYSDEVOP_irq_status_query, &irq_status))
++ return 0;
++ return !!(irq_status.flags & XENIRQSTAT_shared);
+}
+
-+subsys_initcall(balloon_init);
-+
-+static void balloon_exit(void)
++void notify_remote_via_irq(int irq)
+{
-+ /* XXX - release balloon here */
-+ return;
-+}
++ int evtchn = evtchn_from_irq(irq);
+
-+module_exit(balloon_exit);
++ if (VALID_EVTCHN(evtchn))
++ notify_remote_via_evtchn(evtchn);
++}
++EXPORT_SYMBOL_GPL(notify_remote_via_irq);
+
-+void balloon_update_driver_allowance(long delta)
++int irq_to_evtchn_port(int irq)
+{
-+ unsigned long flags;
-+
-+ balloon_lock(flags);
-+ bs.driver_pages += delta;
-+ balloon_unlock(flags);
++ return evtchn_from_irq(irq);
+}
++EXPORT_SYMBOL_GPL(irq_to_evtchn_port);
+
-+#ifdef CONFIG_XEN
-+static int dealloc_pte_fn(
-+ pte_t *pte, struct page *pmd_page, unsigned long addr, void *data)
++void mask_evtchn(int port)
+{
-+ unsigned long mfn = pte_mfn(*pte);
-+ int ret;
-+ struct xen_memory_reservation reservation = {
-+ .nr_extents = 1,
-+ .extent_order = 0,
-+ .domid = DOMID_SELF
-+ };
-+ set_xen_guest_handle(reservation.extent_start, &mfn);
-+ set_pte_at(&init_mm, addr, pte, __pte_ma(0));
-+ set_phys_to_machine(__pa(addr) >> PAGE_SHIFT, INVALID_P2M_ENTRY);
-+ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation, &reservation);
-+ BUG_ON(ret != 1);
-+ return 0;
++ shared_info_t *s = HYPERVISOR_shared_info;
++ synch_set_bit(port, s->evtchn_mask);
+}
-+#endif
++EXPORT_SYMBOL_GPL(mask_evtchn);
+
-+struct page **alloc_empty_pages_and_pagevec(int nr_pages)
++void unmask_evtchn(int port)
+{
-+ unsigned long vaddr, flags;
-+ struct page *page, **pagevec;
-+ int i, ret;
++ shared_info_t *s = HYPERVISOR_shared_info;
++ unsigned int cpu = smp_processor_id();
++ vcpu_info_t *vcpu_info = &s->vcpu_info[cpu];
+
-+ pagevec = kmalloc(sizeof(page) * nr_pages, GFP_KERNEL);
-+ if (pagevec == NULL)
-+ return NULL;
++ BUG_ON(!irqs_disabled());
+
-+ for (i = 0; i < nr_pages; i++) {
-+ page = pagevec[i] = alloc_page(GFP_KERNEL);
-+ if (page == NULL)
-+ goto err;
++ /* Slow path (hypercall) if this is a non-local port. */
++ if (unlikely(cpu != cpu_from_evtchn(port))) {
++ struct evtchn_unmask unmask = { .port = port };
++ VOID(HYPERVISOR_event_channel_op(EVTCHNOP_unmask, &unmask));
++ return;
++ }
+
-+ vaddr = (unsigned long)page_address(page);
++ synch_clear_bit(port, s->evtchn_mask);
+
-+ scrub_pages(vaddr, 1);
++ /* Did we miss an interrupt 'edge'? Re-fire if so. */
++ if (synch_test_bit(port, s->evtchn_pending) &&
++ !synch_test_and_set_bit(port / BITS_PER_LONG,
++ &vcpu_info->evtchn_pending_sel))
++ vcpu_info->evtchn_upcall_pending = 1;
++}
++EXPORT_SYMBOL_GPL(unmask_evtchn);
+
-+ balloon_lock(flags);
++void disable_all_local_evtchn(void)
++{
++ unsigned i, cpu = smp_processor_id();
++ shared_info_t *s = HYPERVISOR_shared_info;
+
-+ if (xen_feature(XENFEAT_auto_translated_physmap)) {
-+ unsigned long gmfn = page_to_pfn(page);
-+ struct xen_memory_reservation reservation = {
-+ .nr_extents = 1,
-+ .extent_order = 0,
-+ .domid = DOMID_SELF
-+ };
-+ set_xen_guest_handle(reservation.extent_start, &gmfn);
-+ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
-+ &reservation);
-+ if (ret == 1)
-+ ret = 0; /* success */
-+ } else {
-+#ifdef CONFIG_XEN
-+ ret = apply_to_page_range(&init_mm, vaddr, PAGE_SIZE,
-+ dealloc_pte_fn, NULL);
-+#else
-+ /* Cannot handle non-auto translate mode. */
-+ ret = 1;
-+#endif
-+ }
++ for (i = 0; i < NR_EVENT_CHANNELS; ++i)
++ if (cpu_from_evtchn(i) == cpu)
++ synch_set_bit(i, &s->evtchn_mask[0]);
++}
+
-+ if (ret != 0) {
-+ balloon_unlock(flags);
-+ __free_page(page);
-+ goto err;
-+ }
++static void restore_cpu_virqs(unsigned int cpu)
++{
++ struct evtchn_bind_virq bind_virq;
++ int virq, irq, evtchn;
+
-+ totalram_pages = --bs.current_pages;
++ for (virq = 0; virq < NR_VIRQS; virq++) {
++ if ((irq = per_cpu(virq_to_irq, cpu)[virq]) == -1)
++ continue;
+
-+ balloon_unlock(flags);
-+ }
++ BUG_ON(irq_info[irq] != mk_irq_info(IRQT_VIRQ, virq, 0));
+
-+ out:
-+ schedule_work(&balloon_worker);
-+#ifdef CONFIG_XEN
-+ flush_tlb_all();
-+#endif
-+ return pagevec;
++ /* Get a new binding from Xen. */
++ bind_virq.virq = virq;
++ bind_virq.vcpu = cpu;
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
++ &bind_virq) != 0)
++ BUG();
++ evtchn = bind_virq.port;
+
-+ err:
-+ balloon_lock(flags);
-+ while (--i >= 0)
-+ balloon_append(pagevec[i]);
-+ balloon_unlock(flags);
-+ kfree(pagevec);
-+ pagevec = NULL;
-+ goto out;
++ /* Record the new mapping. */
++ evtchn_to_irq[evtchn] = irq;
++ irq_info[irq] = mk_irq_info(IRQT_VIRQ, virq, evtchn);
++ bind_evtchn_to_cpu(evtchn, cpu);
++
++ /* Ready for use. */
++ unmask_evtchn(evtchn);
++ }
+}
+
-+void free_empty_pages_and_pagevec(struct page **pagevec, int nr_pages)
++static void restore_cpu_ipis(unsigned int cpu)
+{
-+ unsigned long flags;
-+ int i;
++ struct evtchn_bind_ipi bind_ipi;
++ int ipi, irq, evtchn;
+
-+ if (pagevec == NULL)
-+ return;
++ for (ipi = 0; ipi < NR_IPIS; ipi++) {
++ if ((irq = per_cpu(ipi_to_irq, cpu)[ipi]) == -1)
++ continue;
+
-+ balloon_lock(flags);
-+ for (i = 0; i < nr_pages; i++) {
-+ BUG_ON(page_count(pagevec[i]) != 1);
-+ balloon_append(pagevec[i]);
-+ }
-+ balloon_unlock(flags);
++ BUG_ON(irq_info[irq] != mk_irq_info(IRQT_IPI, ipi, 0));
+
-+ kfree(pagevec);
++ /* Get a new binding from Xen. */
++ bind_ipi.vcpu = cpu;
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_ipi,
++ &bind_ipi) != 0)
++ BUG();
++ evtchn = bind_ipi.port;
+
-+ schedule_work(&balloon_worker);
++ /* Record the new mapping. */
++ evtchn_to_irq[evtchn] = irq;
++ irq_info[irq] = mk_irq_info(IRQT_IPI, ipi, evtchn);
++ bind_evtchn_to_cpu(evtchn, cpu);
++
++ /* Ready for use. */
++ unmask_evtchn(evtchn);
++
++ }
+}
+
-+void balloon_release_driver_page(struct page *page)
++void irq_resume(void)
+{
-+ unsigned long flags;
++ unsigned int cpu, pirq, irq, evtchn;
+
-+ balloon_lock(flags);
-+ balloon_append(page);
-+ bs.driver_pages--;
-+ balloon_unlock(flags);
++ init_evtchn_cpu_bindings();
+
-+ schedule_work(&balloon_worker);
-+}
++ /* New event-channel space is not 'live' yet. */
++ for (evtchn = 0; evtchn < NR_EVENT_CHANNELS; evtchn++)
++ mask_evtchn(evtchn);
+
-+EXPORT_SYMBOL_GPL(balloon_update_driver_allowance);
-+EXPORT_SYMBOL_GPL(alloc_empty_pages_and_pagevec);
-+EXPORT_SYMBOL_GPL(free_empty_pages_and_pagevec);
-+EXPORT_SYMBOL_GPL(balloon_release_driver_page);
++ /* Check that no PIRQs are still bound. */
++ for (pirq = 0; pirq < NR_PIRQS; pirq++)
++ BUG_ON(irq_info[pirq_to_irq(pirq)] != IRQ_UNBOUND);
+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/balloon/common.h tmp-linux-2.6-xen.patch/drivers/xen/balloon/common.h
---- pristine-linux-2.6.18.2/drivers/xen/balloon/common.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/balloon/common.h 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,58 @@
-+/******************************************************************************
-+ * balloon/common.h
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ /* No IRQ <-> event-channel mappings. */
++ for (irq = 0; irq < NR_IRQS; irq++)
++ irq_info[irq] &= ~((1U << _EVTCHN_BITS) - 1);
++ for (evtchn = 0; evtchn < NR_EVENT_CHANNELS; evtchn++)
++ evtchn_to_irq[evtchn] = -1;
+
-+#ifndef __XEN_BALLOON_COMMON_H__
-+#define __XEN_BALLOON_COMMON_H__
++ for_each_possible_cpu(cpu) {
++ restore_cpu_virqs(cpu);
++ restore_cpu_ipis(cpu);
++ }
+
-+#define PAGES2KB(_p) ((_p)<<(PAGE_SHIFT-10))
++}
+
-+struct balloon_stats {
-+ /* We aim for 'current allocation' == 'target allocation'. */
-+ unsigned long current_pages;
-+ unsigned long target_pages;
-+ /* We may hit the hard limit in Xen. If we do then we remember it. */
-+ unsigned long hard_limit;
-+ /*
-+ * Drivers may alter the memory reservation independently, but they
-+ * must inform the balloon driver so we avoid hitting the hard limit.
-+ */
-+ unsigned long driver_pages;
-+ /* Number of pages in high- and low-memory balloons. */
-+ unsigned long balloon_low;
-+ unsigned long balloon_high;
-+};
++void __init xen_init_IRQ(void)
++{
++ unsigned int i;
+
-+extern struct balloon_stats balloon_stats;
-+#define bs balloon_stats
++ init_evtchn_cpu_bindings();
+
-+int balloon_sysfs_init(void);
-+void balloon_sysfs_exit(void);
++ /* No event channels are 'live' right now. */
++ for (i = 0; i < NR_EVENT_CHANNELS; i++)
++ mask_evtchn(i);
+
-+void balloon_set_new_target(unsigned long target);
++ /* No IRQ -> event-channel mappings. */
++ for (i = 0; i < NR_IRQS; i++)
++ irq_info[i] = IRQ_UNBOUND;
+
-+#endif /* __XEN_BALLOON_COMMON_H__ */
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/balloon/Makefile tmp-linux-2.6-xen.patch/drivers/xen/balloon/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/balloon/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/balloon/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,2 @@
++ /* Dynamic IRQ space is currently unbound. Zero the refcnts. */
++ for (i = 0; i < NR_DYNIRQS; i++) {
++ irq_bindcount[dynirq_to_irq(i)] = 0;
+
-+obj-y := balloon.o sysfs.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/balloon/sysfs.c tmp-linux-2.6-xen.patch/drivers/xen/balloon/sysfs.c
---- pristine-linux-2.6.18.2/drivers/xen/balloon/sysfs.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/balloon/sysfs.c 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,170 @@
++ irq_desc[dynirq_to_irq(i)].status = IRQ_DISABLED;
++ irq_desc[dynirq_to_irq(i)].action = NULL;
++ irq_desc[dynirq_to_irq(i)].depth = 1;
++ irq_desc[dynirq_to_irq(i)].chip = &dynirq_type;
++ }
++
++ /* Phys IRQ space is statically bound (1:1 mapping). Nail refcnts. */
++ for (i = 0; i < NR_PIRQS; i++) {
++ irq_bindcount[pirq_to_irq(i)] = 1;
++
++#ifdef RTC_IRQ
++ /* If not domain 0, force our RTC driver to fail its probe. */
++ if ((i == RTC_IRQ) && !is_initial_xendomain())
++ continue;
++#endif
++
++ irq_desc[pirq_to_irq(i)].status = IRQ_DISABLED;
++ irq_desc[pirq_to_irq(i)].action = NULL;
++ irq_desc[pirq_to_irq(i)].depth = 1;
++ irq_desc[pirq_to_irq(i)].chip = &pirq_type;
++ }
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/features.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/features.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,34 @@
+/******************************************************************************
-+ * balloon/sysfs.c
++ * features.c
+ *
-+ * Xen balloon driver - sysfs interfaces.
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
++ * Xen feature flags.
++ *
++ * Copyright (c) 2006, Ian Campbell, XenSource Inc.
+ */
-+
-+#include <linux/capability.h>
-+#include <linux/errno.h>
-+#include <linux/stat.h>
-+#include <linux/string.h>
-+#include <linux/sysdev.h>
-+#include "common.h"
++#include <linux/types.h>
++#include <linux/cache.h>
++#include <linux/module.h>
++#include <asm/hypervisor.h>
++#include <xen/features.h>
+
+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
+#include <xen/platform-compat.h>
+#endif
+
-+#define BALLOON_CLASS_NAME "memory"
-+
-+#define BALLOON_SHOW(name, format, args...) \
-+ static ssize_t show_##name(struct sys_device *dev, \
-+ char *buf) \
-+ { \
-+ return sprintf(buf, format, ##args); \
-+ } \
-+ static SYSDEV_ATTR(name, S_IRUGO, show_##name, NULL)
-+
-+BALLOON_SHOW(current_kb, "%lu\n", PAGES2KB(bs.current_pages));
-+BALLOON_SHOW(low_kb, "%lu\n", PAGES2KB(bs.balloon_low));
-+BALLOON_SHOW(high_kb, "%lu\n", PAGES2KB(bs.balloon_high));
-+BALLOON_SHOW(hard_limit_kb,
-+ (bs.hard_limit!=~0UL) ? "%lu\n" : "???\n",
-+ (bs.hard_limit!=~0UL) ? PAGES2KB(bs.hard_limit) : 0);
-+BALLOON_SHOW(driver_kb, "%lu\n", PAGES2KB(bs.driver_pages));
-+
-+static ssize_t show_target_kb(struct sys_device *dev, char *buf)
-+{
-+ return sprintf(buf, "%lu\n", PAGES2KB(bs.target_pages));
-+}
++u8 xen_features[XENFEAT_NR_SUBMAPS * 32] __read_mostly;
++/* Not a GPL symbol: used in ubiquitous macros, so too restrictive. */
++EXPORT_SYMBOL(xen_features);
+
-+static ssize_t store_target_kb(struct sys_device *dev,
-+ const char *buf,
-+ size_t count)
++void setup_xen_features(void)
+{
-+ char memstring[64], *endchar;
-+ unsigned long long target_bytes;
++ xen_feature_info_t fi;
++ int i, j;
+
-+ if (!capable(CAP_SYS_ADMIN))
-+ return -EPERM;
-+
-+ if (count <= 1)
-+ return -EBADMSG; /* runt */
-+ if (count > sizeof(memstring))
-+ return -EFBIG; /* too long */
-+ strcpy(memstring, buf);
-+
-+ target_bytes = memparse(memstring, &endchar);
-+ balloon_set_new_target(target_bytes >> PAGE_SHIFT);
-+
-+ return count;
++ for (i = 0; i < XENFEAT_NR_SUBMAPS; i++) {
++ fi.submap_idx = i;
++ if (HYPERVISOR_xen_version(XENVER_get_features, &fi) < 0)
++ break;
++ for (j=0; j<32; j++)
++ xen_features[i*32+j] = !!(fi.submap & 1<<j);
++ }
+}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/firmware.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/firmware.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,74 @@
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/init.h>
++#include <linux/edd.h>
++#include <video/edid.h>
++#include <xen/interface/platform.h>
++#include <asm/hypervisor.h>
+
-+static SYSDEV_ATTR(target_kb, S_IRUGO | S_IWUSR,
-+ show_target_kb, store_target_kb);
-+
-+static struct sysdev_attribute *balloon_attrs[] = {
-+ &attr_target_kb,
-+};
-+
-+static struct attribute *balloon_info_attrs[] = {
-+ &attr_current_kb.attr,
-+ &attr_low_kb.attr,
-+ &attr_high_kb.attr,
-+ &attr_hard_limit_kb.attr,
-+ &attr_driver_kb.attr,
-+ NULL
-+};
-+
-+static struct attribute_group balloon_info_group = {
-+ .name = "info",
-+ .attrs = balloon_info_attrs,
-+};
-+
-+static struct sysdev_class balloon_sysdev_class = {
-+ set_kset_name(BALLOON_CLASS_NAME),
-+};
++#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE)
++void __init copy_edd(void)
++{
++ int ret;
++ struct xen_platform_op op;
+
-+static struct sys_device balloon_sysdev;
++ if (!is_initial_xendomain())
++ return;
+
-+static int register_balloon(struct sys_device *sysdev)
-+{
-+ int i, error;
++ op.cmd = XENPF_firmware_info;
+
-+ error = sysdev_class_register(&balloon_sysdev_class);
-+ if (error)
-+ return error;
++ op.u.firmware_info.type = XEN_FW_DISK_INFO;
++ for (op.u.firmware_info.index = 0;
++ edd.edd_info_nr < EDDMAXNR;
++ op.u.firmware_info.index++) {
++ struct edd_info *info = edd.edd_info + edd.edd_info_nr;
++
++ info->params.length = sizeof(info->params);
++ set_xen_guest_handle(op.u.firmware_info.u.disk_info.edd_params,
++ &info->params);
++ ret = HYPERVISOR_platform_op(&op);
++ if (ret)
++ break;
+
-+ sysdev->id = 0;
-+ sysdev->cls = &balloon_sysdev_class;
++#define C(x) info->x = op.u.firmware_info.u.disk_info.x
++ C(device);
++ C(version);
++ C(interface_support);
++ C(legacy_max_cylinder);
++ C(legacy_max_head);
++ C(legacy_sectors_per_track);
++#undef C
+
-+ error = sysdev_register(sysdev);
-+ if (error) {
-+ sysdev_class_unregister(&balloon_sysdev_class);
-+ return error;
++ edd.edd_info_nr++;
+ }
+
-+ for (i = 0; i < ARRAY_SIZE(balloon_attrs); i++) {
-+ error = sysdev_create_file(sysdev, balloon_attrs[i]);
-+ if (error)
-+ goto fail;
++ op.u.firmware_info.type = XEN_FW_DISK_MBR_SIGNATURE;
++ for (op.u.firmware_info.index = 0;
++ edd.mbr_signature_nr < EDD_MBR_SIG_MAX;
++ op.u.firmware_info.index++) {
++ ret = HYPERVISOR_platform_op(&op);
++ if (ret)
++ break;
++ edd.mbr_signature[edd.mbr_signature_nr++] =
++ op.u.firmware_info.u.disk_mbr_signature.mbr_signature;
+ }
-+
-+ error = sysfs_create_group(&sysdev->kobj, &balloon_info_group);
-+ if (error)
-+ goto fail;
-+
-+ return 0;
-+
-+ fail:
-+ while (--i >= 0)
-+ sysdev_remove_file(sysdev, balloon_attrs[i]);
-+ sysdev_unregister(sysdev);
-+ sysdev_class_unregister(&balloon_sysdev_class);
-+ return error;
+}
++#endif
+
-+static void unregister_balloon(struct sys_device *sysdev)
++void __init copy_edid(void)
+{
-+ int i;
-+
-+ sysfs_remove_group(&sysdev->kobj, &balloon_info_group);
-+ for (i = 0; i < ARRAY_SIZE(balloon_attrs); i++)
-+ sysdev_remove_file(sysdev, balloon_attrs[i]);
-+ sysdev_unregister(sysdev);
-+ sysdev_class_unregister(&balloon_sysdev_class);
-+}
++#if defined(CONFIG_FIRMWARE_EDID) && defined(CONFIG_X86)
++ struct xen_platform_op op;
+
-+int balloon_sysfs_init(void)
-+{
-+ return register_balloon(&balloon_sysdev);
-+}
++ if (!is_initial_xendomain())
++ return;
+
-+void balloon_sysfs_exit(void)
-+{
-+ unregister_balloon(&balloon_sysdev);
++ op.cmd = XENPF_firmware_info;
++ op.u.firmware_info.index = 0;
++ op.u.firmware_info.type = XEN_FW_VBEDDC_INFO;
++ set_xen_guest_handle(op.u.firmware_info.u.vbeddc_info.edid,
++ edid_info.dummy);
++ if (HYPERVISOR_platform_op(&op) != 0)
++ memset(edid_info.dummy, 0x13, sizeof(edid_info.dummy));
++#endif
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blkback/blkback.c tmp-linux-2.6-xen.patch/drivers/xen/blkback/blkback.c
---- pristine-linux-2.6.18.2/drivers/xen/blkback/blkback.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blkback/blkback.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,617 @@
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/gnttab.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/gnttab.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,769 @@
+/******************************************************************************
-+ * arch/xen/drivers/blkif/backend/main.c
-+ *
-+ * Back-end of the driver for virtual block devices. This portion of the
-+ * driver exports a 'unified' block-device interface that can be accessed
-+ * by any operating system that implements a compatible front end. A
-+ * reference front-end implementation can be found in:
-+ * arch/xen/drivers/blkif/frontend
-+ *
-+ * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
-+ * Copyright (c) 2005, Christopher Clark
-+ *
++ * gnttab.c
++ *
++ * Granting foreign access to our memory reservation.
++ *
++ * Copyright (c) 2005-2006, Christopher Clark
++ * Copyright (c) 2004-2005, K A Fraser
++ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation; or, when distributed
+ * separately from the Linux kernel or incorporated into other
+ * software packages, subject to the following license:
-+ *
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this source file (the "Software"), to deal in the Software without
+ * restriction, including without limitation the rights to use, copy, modify,
+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
+ * and to permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
-+ *
++ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
-+ *
++ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
@@ -51276,3050 +87244,2845 @@
+ * IN THE SOFTWARE.
+ */
+
-+#include <linux/spinlock.h>
-+#include <linux/kthread.h>
-+#include <linux/list.h>
-+#include <xen/balloon.h>
-+#include <asm/hypervisor.h>
-+#include "common.h"
-+
-+/*
-+ * These are rather arbitrary. They are fairly large because adjacent requests
-+ * pulled from a communication ring are quite likely to end up being part of
-+ * the same scatter/gather request at the disc.
-+ *
-+ * ** TRY INCREASING 'blkif_reqs' IF WRITE SPEEDS SEEM TOO LOW **
-+ *
-+ * This will increase the chances of being able to write whole tracks.
-+ * 64 should be enough to keep us competitive with Linux.
-+ */
-+static int blkif_reqs = 64;
-+module_param_named(reqs, blkif_reqs, int, 0);
-+MODULE_PARM_DESC(reqs, "Number of blkback requests to allocate");
-+
-+/* Run-time switchable: /sys/module/blkback/parameters/ */
-+static unsigned int log_stats = 0;
-+static unsigned int debug_lvl = 0;
-+module_param(log_stats, int, 0644);
-+module_param(debug_lvl, int, 0644);
-+
-+/*
-+ * Each outstanding request that we've passed to the lower device layers has a
-+ * 'pending_req' allocated to it. Each buffer_head that completes decrements
-+ * the pendcnt towards zero. When it hits zero, the specified domain has a
-+ * response queued for it, with the saved 'id' passed back.
-+ */
-+typedef struct {
-+ blkif_t *blkif;
-+ u64 id;
-+ int nr_pages;
-+ atomic_t pendcnt;
-+ unsigned short operation;
-+ int status;
-+ struct list_head free_list;
-+} pending_req_t;
++#include <linux/module.h>
++#include <linux/sched.h>
++#include <linux/mm.h>
++#include <linux/seqlock.h>
++#include <xen/interface/xen.h>
++#include <xen/gnttab.h>
++#include <asm/pgtable.h>
++#include <asm/uaccess.h>
++#include <asm/synch_bitops.h>
++#include <asm/io.h>
++#include <xen/interface/memory.h>
++#include <xen/driver_util.h>
++#include <asm/gnttab_dma.h>
+
-+static pending_req_t *pending_reqs;
-+static struct list_head pending_free;
-+static DEFINE_SPINLOCK(pending_free_lock);
-+static DECLARE_WAIT_QUEUE_HEAD(pending_free_wq);
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
+
-+#define BLKBACK_INVALID_HANDLE (~0)
++/* External tools reserve first few grant table entries. */
++#define NR_RESERVED_ENTRIES 8
++#define GNTTAB_LIST_END 0xffffffff
++#define ENTRIES_PER_GRANT_FRAME (PAGE_SIZE / sizeof(grant_entry_t))
+
-+static struct page **pending_pages;
-+static grant_handle_t *pending_grant_handles;
++static grant_ref_t **gnttab_list;
++static unsigned int nr_grant_frames;
++static unsigned int boot_max_nr_grant_frames;
++static int gnttab_free_count;
++static grant_ref_t gnttab_free_head;
++static DEFINE_SPINLOCK(gnttab_list_lock);
+
-+static inline int vaddr_pagenr(pending_req_t *req, int seg)
-+{
-+ return (req - pending_reqs) * BLKIF_MAX_SEGMENTS_PER_REQUEST + seg;
-+}
++static struct grant_entry *shared;
+
-+static inline unsigned long vaddr(pending_req_t *req, int seg)
-+{
-+ unsigned long pfn = page_to_pfn(pending_pages[vaddr_pagenr(req, seg)]);
-+ return (unsigned long)pfn_to_kaddr(pfn);
-+}
++static struct gnttab_free_callback *gnttab_free_callback_list;
+
-+#define pending_handle(_req, _seg) \
-+ (pending_grant_handles[vaddr_pagenr(_req, _seg)])
++static int gnttab_expand(unsigned int req_entries);
+
++#define RPP (PAGE_SIZE / sizeof(grant_ref_t))
++#define gnttab_entry(entry) (gnttab_list[(entry) / RPP][(entry) % RPP])
+
-+static int do_block_io_op(blkif_t *blkif);
-+static void dispatch_rw_block_io(blkif_t *blkif,
-+ blkif_request_t *req,
-+ pending_req_t *pending_req);
-+static void make_response(blkif_t *blkif, u64 id,
-+ unsigned short op, int st);
++#define nr_freelist_frames(grant_frames) \
++ (((grant_frames) * ENTRIES_PER_GRANT_FRAME + RPP - 1) / RPP)
+
-+/******************************************************************
-+ * misc small helpers
-+ */
-+static pending_req_t* alloc_req(void)
++static int get_free_entries(int count)
+{
-+ pending_req_t *req = NULL;
+ unsigned long flags;
++ int ref, rc;
++ grant_ref_t head;
+
-+ spin_lock_irqsave(&pending_free_lock, flags);
-+ if (!list_empty(&pending_free)) {
-+ req = list_entry(pending_free.next, pending_req_t, free_list);
-+ list_del(&req->free_list);
++ spin_lock_irqsave(&gnttab_list_lock, flags);
++
++ if ((gnttab_free_count < count) &&
++ ((rc = gnttab_expand(count - gnttab_free_count)) < 0)) {
++ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++ return rc;
+ }
-+ spin_unlock_irqrestore(&pending_free_lock, flags);
-+ return req;
-+}
+
-+static void free_req(pending_req_t *req)
-+{
-+ unsigned long flags;
-+ int was_empty;
++ ref = head = gnttab_free_head;
++ gnttab_free_count -= count;
++ while (count-- > 1)
++ head = gnttab_entry(head);
++ gnttab_free_head = gnttab_entry(head);
++ gnttab_entry(head) = GNTTAB_LIST_END;
+
-+ spin_lock_irqsave(&pending_free_lock, flags);
-+ was_empty = list_empty(&pending_free);
-+ list_add(&req->free_list, &pending_free);
-+ spin_unlock_irqrestore(&pending_free_lock, flags);
-+ if (was_empty)
-+ wake_up(&pending_free_wq);
-+}
++ spin_unlock_irqrestore(&gnttab_list_lock, flags);
+
-+static void unplug_queue(blkif_t *blkif)
-+{
-+ if (blkif->plug == NULL)
-+ return;
-+ if (blkif->plug->unplug_fn)
-+ blkif->plug->unplug_fn(blkif->plug);
-+ blk_put_queue(blkif->plug);
-+ blkif->plug = NULL;
++ return ref;
+}
+
-+static void plug_queue(blkif_t *blkif, struct bio *bio)
-+{
-+ request_queue_t *q = bdev_get_queue(bio->bi_bdev);
-+
-+ if (q == blkif->plug)
-+ return;
-+ unplug_queue(blkif);
-+ blk_get_queue(q);
-+ blkif->plug = q;
-+}
++#define get_free_entry() get_free_entries(1)
+
-+static void fast_flush_area(pending_req_t *req)
++static void do_free_callbacks(void)
+{
-+ struct gnttab_unmap_grant_ref unmap[BLKIF_MAX_SEGMENTS_PER_REQUEST];
-+ unsigned int i, invcount = 0;
-+ grant_handle_t handle;
-+ int ret;
++ struct gnttab_free_callback *callback, *next;
+
-+ for (i = 0; i < req->nr_pages; i++) {
-+ handle = pending_handle(req, i);
-+ if (handle == BLKBACK_INVALID_HANDLE)
-+ continue;
-+ gnttab_set_unmap_op(&unmap[i], vaddr(req, i), GNTMAP_host_map,
-+ handle);
-+ pending_handle(req, i) = BLKBACK_INVALID_HANDLE;
-+ invcount++;
-+ }
++ callback = gnttab_free_callback_list;
++ gnttab_free_callback_list = NULL;
+
-+ ret = HYPERVISOR_grant_table_op(
-+ GNTTABOP_unmap_grant_ref, unmap, invcount);
-+ BUG_ON(ret);
++ while (callback != NULL) {
++ next = callback->next;
++ if (gnttab_free_count >= callback->count) {
++ callback->next = NULL;
++ callback->fn(callback->arg);
++ } else {
++ callback->next = gnttab_free_callback_list;
++ gnttab_free_callback_list = callback;
++ }
++ callback = next;
++ }
+}
+
-+/******************************************************************
-+ * SCHEDULER FUNCTIONS
-+ */
-+
-+static void print_stats(blkif_t *blkif)
++static inline void check_free_callbacks(void)
+{
-+ printk(KERN_DEBUG "%s: oo %3d | rd %4d | wr %4d | br %4d\n",
-+ current->comm, blkif->st_oo_req,
-+ blkif->st_rd_req, blkif->st_wr_req, blkif->st_br_req);
-+ blkif->st_print = jiffies + msecs_to_jiffies(10 * 1000);
-+ blkif->st_rd_req = 0;
-+ blkif->st_wr_req = 0;
-+ blkif->st_oo_req = 0;
++ if (unlikely(gnttab_free_callback_list))
++ do_free_callbacks();
+}
+
-+int blkif_schedule(void *arg)
++static void put_free_entry(grant_ref_t ref)
+{
-+ blkif_t *blkif = arg;
-+
-+ blkif_get(blkif);
-+
-+ if (debug_lvl)
-+ printk(KERN_DEBUG "%s: started\n", current->comm);
-+
-+ while (!kthread_should_stop()) {
-+ if (try_to_freeze())
-+ continue;
-+
-+ wait_event_interruptible(
-+ blkif->wq,
-+ blkif->waiting_reqs || kthread_should_stop());
-+ wait_event_interruptible(
-+ pending_free_wq,
-+ !list_empty(&pending_free) || kthread_should_stop());
-+
-+ blkif->waiting_reqs = 0;
-+ smp_mb(); /* clear flag *before* checking for work */
-+
-+ if (do_block_io_op(blkif))
-+ blkif->waiting_reqs = 1;
-+ unplug_queue(blkif);
-+
-+ if (log_stats && time_after(jiffies, blkif->st_print))
-+ print_stats(blkif);
-+ }
-+
-+ if (log_stats)
-+ print_stats(blkif);
-+ if (debug_lvl)
-+ printk(KERN_DEBUG "%s: exiting\n", current->comm);
-+
-+ blkif->xenblkd = NULL;
-+ blkif_put(blkif);
-+
-+ return 0;
++ unsigned long flags;
++ spin_lock_irqsave(&gnttab_list_lock, flags);
++ gnttab_entry(ref) = gnttab_free_head;
++ gnttab_free_head = ref;
++ gnttab_free_count++;
++ check_free_callbacks();
++ spin_unlock_irqrestore(&gnttab_list_lock, flags);
+}
+
-+/******************************************************************
-+ * COMPLETION CALLBACK -- Called as bh->b_end_io()
++/*
++ * Public grant-issuing interface functions
+ */
+
-+static void __end_block_io_op(pending_req_t *pending_req, int error)
-+{
-+ /* An error fails the entire request. */
-+ if ((pending_req->operation == BLKIF_OP_WRITE_BARRIER) &&
-+ (error == -EOPNOTSUPP)) {
-+ DPRINTK("blkback: write barrier op failed, not supported\n");
-+ blkback_barrier(XBT_NIL, pending_req->blkif->be, 0);
-+ pending_req->status = BLKIF_RSP_EOPNOTSUPP;
-+ } else if (error) {
-+ DPRINTK("Buffer not up-to-date at end of operation, "
-+ "error=%d\n", error);
-+ pending_req->status = BLKIF_RSP_ERROR;
-+ }
-+
-+ if (atomic_dec_and_test(&pending_req->pendcnt)) {
-+ fast_flush_area(pending_req);
-+ make_response(pending_req->blkif, pending_req->id,
-+ pending_req->operation, pending_req->status);
-+ blkif_put(pending_req->blkif);
-+ free_req(pending_req);
-+ }
-+}
-+
-+static int end_block_io_op(struct bio *bio, unsigned int done, int error)
++int gnttab_grant_foreign_access(domid_t domid, unsigned long frame,
++ int flags)
+{
-+ if (bio->bi_size != 0)
-+ return 1;
-+ __end_block_io_op(bio->bi_private, error);
-+ bio_put(bio);
-+ return error;
-+}
++ int ref;
+
++ if (unlikely((ref = get_free_entry()) < 0))
++ return -ENOSPC;
+
-+/******************************************************************************
-+ * NOTIFICATION FROM GUEST OS.
-+ */
++ shared[ref].frame = frame;
++ shared[ref].domid = domid;
++ wmb();
++ BUG_ON(flags & (GTF_accept_transfer | GTF_reading | GTF_writing));
++ shared[ref].flags = GTF_permit_access | flags;
+
-+static void blkif_notify_work(blkif_t *blkif)
-+{
-+ blkif->waiting_reqs = 1;
-+ wake_up(&blkif->wq);
++ return ref;
+}
++EXPORT_SYMBOL_GPL(gnttab_grant_foreign_access);
+
-+irqreturn_t blkif_be_int(int irq, void *dev_id, struct pt_regs *regs)
++void gnttab_grant_foreign_access_ref(grant_ref_t ref, domid_t domid,
++ unsigned long frame, int flags)
+{
-+ blkif_notify_work(dev_id);
-+ return IRQ_HANDLED;
++ shared[ref].frame = frame;
++ shared[ref].domid = domid;
++ wmb();
++ BUG_ON(flags & (GTF_accept_transfer | GTF_reading | GTF_writing));
++ shared[ref].flags = GTF_permit_access | flags;
+}
++EXPORT_SYMBOL_GPL(gnttab_grant_foreign_access_ref);
+
+
-+
-+/******************************************************************
-+ * DOWNWARD CALLS -- These interface with the block-device layer proper.
-+ */
-+
-+static int do_block_io_op(blkif_t *blkif)
++int gnttab_query_foreign_access(grant_ref_t ref)
+{
-+ blkif_back_rings_t *blk_rings = &blkif->blk_rings;
-+ blkif_request_t req;
-+ pending_req_t *pending_req;
-+ RING_IDX rc, rp;
-+ int more_to_do = 0;
-+
-+ rc = blk_rings->common.req_cons;
-+ rp = blk_rings->common.sring->req_prod;
-+ rmb(); /* Ensure we see queued requests up to 'rp'. */
++ u16 nflags;
+
-+ while ((rc != rp)) {
++ nflags = shared[ref].flags;
+
-+ if (RING_REQUEST_CONS_OVERFLOW(&blk_rings->common, rc))
-+ break;
++ return (nflags & (GTF_reading|GTF_writing));
++}
++EXPORT_SYMBOL_GPL(gnttab_query_foreign_access);
+
-+ pending_req = alloc_req();
-+ if (NULL == pending_req) {
-+ blkif->st_oo_req++;
-+ more_to_do = 1;
-+ break;
-+ }
++int gnttab_end_foreign_access_ref(grant_ref_t ref)
++{
++ u16 flags, nflags;
+
-+ switch (blkif->blk_protocol) {
-+ case BLKIF_PROTOCOL_NATIVE:
-+ memcpy(&req, RING_GET_REQUEST(&blk_rings->native, rc), sizeof(req));
-+ break;
-+ case BLKIF_PROTOCOL_X86_32:
-+ blkif_get_x86_32_req(&req, RING_GET_REQUEST(&blk_rings->x86_32, rc));
-+ break;
-+ case BLKIF_PROTOCOL_X86_64:
-+ blkif_get_x86_64_req(&req, RING_GET_REQUEST(&blk_rings->x86_64, rc));
-+ break;
-+ default:
-+ BUG();
++ nflags = shared[ref].flags;
++ do {
++ if ((flags = nflags) & (GTF_reading|GTF_writing)) {
++ printk(KERN_DEBUG "WARNING: g.e. still in use!\n");
++ return 0;
+ }
-+ blk_rings->common.req_cons = ++rc; /* before make_response() */
++ } while ((nflags = synch_cmpxchg_subword(&shared[ref].flags, flags, 0)) !=
++ flags);
+
-+ switch (req.operation) {
-+ case BLKIF_OP_READ:
-+ blkif->st_rd_req++;
-+ dispatch_rw_block_io(blkif, &req, pending_req);
-+ break;
-+ case BLKIF_OP_WRITE_BARRIER:
-+ blkif->st_br_req++;
-+ /* fall through */
-+ case BLKIF_OP_WRITE:
-+ blkif->st_wr_req++;
-+ dispatch_rw_block_io(blkif, &req, pending_req);
-+ break;
-+ default:
-+ DPRINTK("error: unknown block io operation [%d]\n",
-+ req.operation);
-+ make_response(blkif, req.id, req.operation,
-+ BLKIF_RSP_ERROR);
-+ free_req(pending_req);
-+ break;
-+ }
-+ }
-+ return more_to_do;
++ return 1;
+}
++EXPORT_SYMBOL_GPL(gnttab_end_foreign_access_ref);
+
-+static void dispatch_rw_block_io(blkif_t *blkif,
-+ blkif_request_t *req,
-+ pending_req_t *pending_req)
++void gnttab_end_foreign_access(grant_ref_t ref, unsigned long page)
+{
-+ extern void ll_rw_block(int rw, int nr, struct buffer_head * bhs[]);
-+ struct gnttab_map_grant_ref map[BLKIF_MAX_SEGMENTS_PER_REQUEST];
-+ struct phys_req preq;
-+ struct {
-+ unsigned long buf; unsigned int nsec;
-+ } seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
-+ unsigned int nseg;
-+ struct bio *bio = NULL, *biolist[BLKIF_MAX_SEGMENTS_PER_REQUEST];
-+ int ret, i, nbio = 0;
-+ int operation;
-+
-+ switch (req->operation) {
-+ case BLKIF_OP_READ:
-+ operation = READ;
-+ break;
-+ case BLKIF_OP_WRITE:
-+ operation = WRITE;
-+ break;
-+ case BLKIF_OP_WRITE_BARRIER:
-+ operation = WRITE_BARRIER;
-+ break;
-+ default:
-+ operation = 0; /* make gcc happy */
-+ BUG();
-+ }
-+
-+ /* Check that number of segments is sane. */
-+ nseg = req->nr_segments;
-+ if (unlikely(nseg == 0) ||
-+ unlikely(nseg > BLKIF_MAX_SEGMENTS_PER_REQUEST)) {
-+ DPRINTK("Bad number of segments in request (%d)\n", nseg);
-+ goto fail_response;
++ if (gnttab_end_foreign_access_ref(ref)) {
++ put_free_entry(ref);
++ if (page != 0)
++ free_page(page);
++ } else {
++ /* XXX This needs to be fixed so that the ref and page are
++ placed on a list to be freed up later. */
++ printk(KERN_DEBUG
++ "WARNING: leaking g.e. and page still in use!\n");
+ }
++}
++EXPORT_SYMBOL_GPL(gnttab_end_foreign_access);
+
-+ preq.dev = req->handle;
-+ preq.sector_number = req->sector_number;
-+ preq.nr_sects = 0;
-+
-+ pending_req->blkif = blkif;
-+ pending_req->id = req->id;
-+ pending_req->operation = req->operation;
-+ pending_req->status = BLKIF_RSP_OKAY;
-+ pending_req->nr_pages = nseg;
-+
-+ for (i = 0; i < nseg; i++) {
-+ uint32_t flags;
-+
-+ seg[i].nsec = req->seg[i].last_sect -
-+ req->seg[i].first_sect + 1;
-+
-+ if ((req->seg[i].last_sect >= (PAGE_SIZE >> 9)) ||
-+ (req->seg[i].last_sect < req->seg[i].first_sect))
-+ goto fail_response;
-+ preq.nr_sects += seg[i].nsec;
-+
-+ flags = GNTMAP_host_map;
-+ if (operation != READ)
-+ flags |= GNTMAP_readonly;
-+ gnttab_set_map_op(&map[i], vaddr(pending_req, i), flags,
-+ req->seg[i].gref, blkif->domid);
-+ }
++int gnttab_grant_foreign_transfer(domid_t domid, unsigned long pfn)
++{
++ int ref;
+
-+ ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, map, nseg);
-+ BUG_ON(ret);
++ if (unlikely((ref = get_free_entry()) < 0))
++ return -ENOSPC;
++ gnttab_grant_foreign_transfer_ref(ref, domid, pfn);
+
-+ for (i = 0; i < nseg; i++) {
-+ if (unlikely(map[i].status != 0)) {
-+ DPRINTK("invalid buffer -- could not remap it\n");
-+ map[i].handle = BLKBACK_INVALID_HANDLE;
-+ ret |= 1;
-+ }
++ return ref;
++}
++EXPORT_SYMBOL_GPL(gnttab_grant_foreign_transfer);
+
-+ pending_handle(pending_req, i) = map[i].handle;
++void gnttab_grant_foreign_transfer_ref(grant_ref_t ref, domid_t domid,
++ unsigned long pfn)
++{
++ shared[ref].frame = pfn;
++ shared[ref].domid = domid;
++ wmb();
++ shared[ref].flags = GTF_accept_transfer;
++}
++EXPORT_SYMBOL_GPL(gnttab_grant_foreign_transfer_ref);
+
-+ if (ret)
-+ continue;
++unsigned long gnttab_end_foreign_transfer_ref(grant_ref_t ref)
++{
++ unsigned long frame;
++ u16 flags;
+
-+ set_phys_to_machine(__pa(vaddr(
-+ pending_req, i)) >> PAGE_SHIFT,
-+ FOREIGN_FRAME(map[i].dev_bus_addr >> PAGE_SHIFT));
-+ seg[i].buf = map[i].dev_bus_addr |
-+ (req->seg[i].first_sect << 9);
++ /*
++ * If a transfer is not even yet started, try to reclaim the grant
++ * reference and return failure (== 0).
++ */
++ while (!((flags = shared[ref].flags) & GTF_transfer_committed)) {
++ if (synch_cmpxchg_subword(&shared[ref].flags, flags, 0) == flags)
++ return 0;
++ cpu_relax();
+ }
+
-+ if (ret)
-+ goto fail_flush;
-+
-+ if (vbd_translate(&preq, blkif, operation) != 0) {
-+ DPRINTK("access denied: %s of [%llu,%llu] on dev=%04x\n",
-+ operation == READ ? "read" : "write",
-+ preq.sector_number,
-+ preq.sector_number + preq.nr_sects, preq.dev);
-+ goto fail_flush;
++ /* If a transfer is in progress then wait until it is completed. */
++ while (!(flags & GTF_transfer_completed)) {
++ flags = shared[ref].flags;
++ cpu_relax();
+ }
+
-+ for (i = 0; i < nseg; i++) {
-+ if (((int)preq.sector_number|(int)seg[i].nsec) &
-+ ((bdev_hardsect_size(preq.bdev) >> 9) - 1)) {
-+ DPRINTK("Misaligned I/O request from domain %d",
-+ blkif->domid);
-+ goto fail_put_bio;
-+ }
++ /* Read the frame number /after/ reading completion status. */
++ rmb();
++ frame = shared[ref].frame;
++ BUG_ON(frame == 0);
+
-+ while ((bio == NULL) ||
-+ (bio_add_page(bio,
-+ virt_to_page(vaddr(pending_req, i)),
-+ seg[i].nsec << 9,
-+ seg[i].buf & ~PAGE_MASK) == 0)) {
-+ bio = biolist[nbio++] = bio_alloc(GFP_KERNEL, nseg-i);
-+ if (unlikely(bio == NULL))
-+ goto fail_put_bio;
++ return frame;
++}
++EXPORT_SYMBOL_GPL(gnttab_end_foreign_transfer_ref);
+
-+ bio->bi_bdev = preq.bdev;
-+ bio->bi_private = pending_req;
-+ bio->bi_end_io = end_block_io_op;
-+ bio->bi_sector = preq.sector_number;
-+ }
++unsigned long gnttab_end_foreign_transfer(grant_ref_t ref)
++{
++ unsigned long frame = gnttab_end_foreign_transfer_ref(ref);
++ put_free_entry(ref);
++ return frame;
++}
++EXPORT_SYMBOL_GPL(gnttab_end_foreign_transfer);
+
-+ preq.sector_number += seg[i].nsec;
++void gnttab_free_grant_reference(grant_ref_t ref)
++{
++ put_free_entry(ref);
++}
++EXPORT_SYMBOL_GPL(gnttab_free_grant_reference);
++
++void gnttab_free_grant_references(grant_ref_t head)
++{
++ grant_ref_t ref;
++ unsigned long flags;
++ int count = 1;
++ if (head == GNTTAB_LIST_END)
++ return;
++ spin_lock_irqsave(&gnttab_list_lock, flags);
++ ref = head;
++ while (gnttab_entry(ref) != GNTTAB_LIST_END) {
++ ref = gnttab_entry(ref);
++ count++;
+ }
++ gnttab_entry(ref) = gnttab_free_head;
++ gnttab_free_head = head;
++ gnttab_free_count += count;
++ check_free_callbacks();
++ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++}
++EXPORT_SYMBOL_GPL(gnttab_free_grant_references);
+
-+ plug_queue(blkif, bio);
-+ atomic_set(&pending_req->pendcnt, nbio);
-+ blkif_get(blkif);
++int gnttab_alloc_grant_references(u16 count, grant_ref_t *head)
++{
++ int h = get_free_entries(count);
+
-+ for (i = 0; i < nbio; i++)
-+ submit_bio(operation, biolist[i]);
++ if (h < 0)
++ return -ENOSPC;
+
-+ if (operation == READ)
-+ blkif->st_rd_sect += preq.nr_sects;
-+ else if (operation == WRITE)
-+ blkif->st_wr_sect += preq.nr_sects;
++ *head = h;
+
-+ return;
++ return 0;
++}
++EXPORT_SYMBOL_GPL(gnttab_alloc_grant_references);
+
-+ fail_put_bio:
-+ for (i = 0; i < (nbio-1); i++)
-+ bio_put(biolist[i]);
-+ fail_flush:
-+ fast_flush_area(pending_req);
-+ fail_response:
-+ make_response(blkif, req->id, req->operation, BLKIF_RSP_ERROR);
-+ free_req(pending_req);
-+}
++int gnttab_empty_grant_references(const grant_ref_t *private_head)
++{
++ return (*private_head == GNTTAB_LIST_END);
++}
++EXPORT_SYMBOL_GPL(gnttab_empty_grant_references);
++
++int gnttab_claim_grant_reference(grant_ref_t *private_head)
++{
++ grant_ref_t g = *private_head;
++ if (unlikely(g == GNTTAB_LIST_END))
++ return -ENOSPC;
++ *private_head = gnttab_entry(g);
++ return g;
++}
++EXPORT_SYMBOL_GPL(gnttab_claim_grant_reference);
+
++void gnttab_release_grant_reference(grant_ref_t *private_head,
++ grant_ref_t release)
++{
++ gnttab_entry(release) = *private_head;
++ *private_head = release;
++}
++EXPORT_SYMBOL_GPL(gnttab_release_grant_reference);
+
++void gnttab_request_free_callback(struct gnttab_free_callback *callback,
++ void (*fn)(void *), void *arg, u16 count)
++{
++ unsigned long flags;
++ spin_lock_irqsave(&gnttab_list_lock, flags);
++ if (callback->next)
++ goto out;
++ callback->fn = fn;
++ callback->arg = arg;
++ callback->count = count;
++ callback->next = gnttab_free_callback_list;
++ gnttab_free_callback_list = callback;
++ check_free_callbacks();
++out:
++ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++}
++EXPORT_SYMBOL_GPL(gnttab_request_free_callback);
+
-+/******************************************************************
-+ * MISCELLANEOUS SETUP / TEARDOWN / DEBUGGING
-+ */
++void gnttab_cancel_free_callback(struct gnttab_free_callback *callback)
++{
++ struct gnttab_free_callback **pcb;
++ unsigned long flags;
+
++ spin_lock_irqsave(&gnttab_list_lock, flags);
++ for (pcb = &gnttab_free_callback_list; *pcb; pcb = &(*pcb)->next) {
++ if (*pcb == callback) {
++ *pcb = callback->next;
++ break;
++ }
++ }
++ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++}
++EXPORT_SYMBOL_GPL(gnttab_cancel_free_callback);
+
-+static void make_response(blkif_t *blkif, u64 id,
-+ unsigned short op, int st)
++static int grow_gnttab_list(unsigned int more_frames)
+{
-+ blkif_response_t resp;
-+ unsigned long flags;
-+ blkif_back_rings_t *blk_rings = &blkif->blk_rings;
-+ int more_to_do = 0;
-+ int notify;
++ unsigned int new_nr_grant_frames, extra_entries, i;
++ unsigned int nr_glist_frames, new_nr_glist_frames;
+
-+ resp.id = id;
-+ resp.operation = op;
-+ resp.status = st;
++ new_nr_grant_frames = nr_grant_frames + more_frames;
++ extra_entries = more_frames * ENTRIES_PER_GRANT_FRAME;
+
-+ spin_lock_irqsave(&blkif->blk_ring_lock, flags);
-+ /* Place on the response ring for the relevant domain. */
-+ switch (blkif->blk_protocol) {
-+ case BLKIF_PROTOCOL_NATIVE:
-+ memcpy(RING_GET_RESPONSE(&blk_rings->native, blk_rings->native.rsp_prod_pvt),
-+ &resp, sizeof(resp));
-+ break;
-+ case BLKIF_PROTOCOL_X86_32:
-+ memcpy(RING_GET_RESPONSE(&blk_rings->x86_32, blk_rings->x86_32.rsp_prod_pvt),
-+ &resp, sizeof(resp));
-+ break;
-+ case BLKIF_PROTOCOL_X86_64:
-+ memcpy(RING_GET_RESPONSE(&blk_rings->x86_64, blk_rings->x86_64.rsp_prod_pvt),
-+ &resp, sizeof(resp));
-+ break;
-+ default:
-+ BUG();
++ nr_glist_frames = nr_freelist_frames(nr_grant_frames);
++ new_nr_glist_frames = nr_freelist_frames(new_nr_grant_frames);
++ for (i = nr_glist_frames; i < new_nr_glist_frames; i++) {
++ gnttab_list[i] = (grant_ref_t *)__get_free_page(GFP_ATOMIC);
++ if (!gnttab_list[i])
++ goto grow_nomem;
+ }
-+ blk_rings->common.rsp_prod_pvt++;
-+ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&blk_rings->common, notify);
-+ if (blk_rings->common.rsp_prod_pvt == blk_rings->common.req_cons) {
-+ /*
-+ * Tail check for pending requests. Allows frontend to avoid
-+ * notifications if requests are already in flight (lower
-+ * overheads and promotes batching).
-+ */
-+ RING_FINAL_CHECK_FOR_REQUESTS(&blk_rings->common, more_to_do);
+
-+ } else if (RING_HAS_UNCONSUMED_REQUESTS(&blk_rings->common)) {
-+ more_to_do = 1;
-+ }
++ for (i = ENTRIES_PER_GRANT_FRAME * nr_grant_frames;
++ i < ENTRIES_PER_GRANT_FRAME * new_nr_grant_frames - 1; i++)
++ gnttab_entry(i) = i + 1;
+
-+ spin_unlock_irqrestore(&blkif->blk_ring_lock, flags);
++ gnttab_entry(i) = gnttab_free_head;
++ gnttab_free_head = ENTRIES_PER_GRANT_FRAME * nr_grant_frames;
++ gnttab_free_count += extra_entries;
+
-+ if (more_to_do)
-+ blkif_notify_work(blkif);
-+ if (notify)
-+ notify_remote_via_irq(blkif->irq);
++ nr_grant_frames = new_nr_grant_frames;
++
++ check_free_callbacks();
++
++ return 0;
++
++grow_nomem:
++ for ( ; i >= nr_glist_frames; i--)
++ free_page((unsigned long) gnttab_list[i]);
++ return -ENOMEM;
+}
+
-+static int __init blkif_init(void)
++static unsigned int __max_nr_grant_frames(void)
+{
-+ int i, mmap_pages;
-+
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++ struct gnttab_query_size query;
++ int rc;
+
-+ mmap_pages = blkif_reqs * BLKIF_MAX_SEGMENTS_PER_REQUEST;
++ query.dom = DOMID_SELF;
+
-+ pending_reqs = kmalloc(sizeof(pending_reqs[0]) *
-+ blkif_reqs, GFP_KERNEL);
-+ pending_grant_handles = kmalloc(sizeof(pending_grant_handles[0]) *
-+ mmap_pages, GFP_KERNEL);
-+ pending_pages = alloc_empty_pages_and_pagevec(mmap_pages);
++ rc = HYPERVISOR_grant_table_op(GNTTABOP_query_size, &query, 1);
++ if ((rc < 0) || (query.status != GNTST_okay))
++ return 4; /* Legacy max supported number of frames */
+
-+ if (!pending_reqs || !pending_grant_handles || !pending_pages)
-+ goto out_of_memory;
++ return query.max_nr_frames;
++}
+
-+ for (i = 0; i < mmap_pages; i++)
-+ pending_grant_handles[i] = BLKBACK_INVALID_HANDLE;
++static inline unsigned int max_nr_grant_frames(void)
++{
++ unsigned int xen_max = __max_nr_grant_frames();
+
-+ blkif_interface_init();
++ if (xen_max > boot_max_nr_grant_frames)
++ return boot_max_nr_grant_frames;
++ return xen_max;
++}
+
-+ memset(pending_reqs, 0, sizeof(pending_reqs));
-+ INIT_LIST_HEAD(&pending_free);
++#ifdef CONFIG_XEN
+
-+ for (i = 0; i < blkif_reqs; i++)
-+ list_add_tail(&pending_reqs[i].free_list, &pending_free);
++static DEFINE_SEQLOCK(gnttab_dma_lock);
+
-+ blkif_xenbus_init();
++#ifdef CONFIG_X86
++static int map_pte_fn(pte_t *pte, struct page *pmd_page,
++ unsigned long addr, void *data)
++{
++ unsigned long **frames = (unsigned long **)data;
+
++ set_pte_at(&init_mm, addr, pte, pfn_pte_ma((*frames)[0], PAGE_KERNEL));
++ (*frames)++;
+ return 0;
++}
+
-+ out_of_memory:
-+ kfree(pending_reqs);
-+ kfree(pending_grant_handles);
-+ free_empty_pages_and_pagevec(pending_pages, mmap_pages);
-+ printk("%s: out of memory\n", __FUNCTION__);
-+ return -ENOMEM;
++static int unmap_pte_fn(pte_t *pte, struct page *pmd_page,
++ unsigned long addr, void *data)
++{
++
++ set_pte_at(&init_mm, addr, pte, __pte(0));
++ return 0;
+}
+
-+module_init(blkif_init);
++void *arch_gnttab_alloc_shared(unsigned long *frames)
++{
++ struct vm_struct *area;
++ area = alloc_vm_area(PAGE_SIZE * max_nr_grant_frames());
++ BUG_ON(area == NULL);
++ return area->addr;
++}
++#endif /* CONFIG_X86 */
+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blkback/common.h tmp-linux-2.6-xen.patch/drivers/xen/blkback/common.h
---- pristine-linux-2.6.18.2/drivers/xen/blkback/common.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blkback/common.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,139 @@
-+/*
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++static int gnttab_map(unsigned int start_idx, unsigned int end_idx)
++{
++ struct gnttab_setup_table setup;
++ unsigned long *frames;
++ unsigned int nr_gframes = end_idx + 1;
++ int rc;
+
-+#ifndef __BLKIF__BACKEND__COMMON_H__
-+#define __BLKIF__BACKEND__COMMON_H__
++ frames = kmalloc(nr_gframes * sizeof(unsigned long), GFP_ATOMIC);
++ if (!frames)
++ return -ENOMEM;
+
-+#include <linux/version.h>
-+#include <linux/module.h>
-+#include <linux/interrupt.h>
-+#include <linux/slab.h>
-+#include <linux/blkdev.h>
-+#include <linux/vmalloc.h>
-+#include <linux/wait.h>
-+#include <asm/io.h>
-+#include <asm/setup.h>
-+#include <asm/pgalloc.h>
-+#include <xen/evtchn.h>
-+#include <asm/hypervisor.h>
-+#include <xen/blkif.h>
-+#include <xen/gnttab.h>
-+#include <xen/driver_util.h>
-+#include <xen/xenbus.h>
++ setup.dom = DOMID_SELF;
++ setup.nr_frames = nr_gframes;
++ set_xen_guest_handle(setup.frame_list, frames);
+
-+#define DPRINTK(_f, _a...) \
-+ pr_debug("(file=%s, line=%d) " _f, \
-+ __FILE__ , __LINE__ , ## _a )
++ rc = HYPERVISOR_grant_table_op(GNTTABOP_setup_table, &setup, 1);
++ if (rc == -ENOSYS) {
++ kfree(frames);
++ return -ENOSYS;
++ }
+
-+struct vbd {
-+ blkif_vdev_t handle; /* what the domain refers to this vbd as */
-+ unsigned char readonly; /* Non-zero -> read-only */
-+ unsigned char type; /* VDISK_xxx */
-+ u32 pdevice; /* phys device that this vbd maps to */
-+ struct block_device *bdev;
-+};
++ BUG_ON(rc || setup.status);
+
-+struct backend_info;
++ if (shared == NULL)
++ shared = arch_gnttab_alloc_shared(frames);
+
-+typedef struct blkif_st {
-+ /* Unique identifier for this interface. */
-+ domid_t domid;
-+ unsigned int handle;
-+ /* Physical parameters of the comms window. */
-+ unsigned int irq;
-+ /* Comms information. */
-+ enum blkif_protocol blk_protocol;
-+ blkif_back_rings_t blk_rings;
-+ struct vm_struct *blk_ring_area;
-+ /* The VBD attached to this interface. */
-+ struct vbd vbd;
-+ /* Back pointer to the backend_info. */
-+ struct backend_info *be;
-+ /* Private fields. */
-+ spinlock_t blk_ring_lock;
-+ atomic_t refcnt;
++#ifdef CONFIG_X86
++ rc = apply_to_page_range(&init_mm, (unsigned long)shared,
++ PAGE_SIZE * nr_gframes,
++ map_pte_fn, &frames);
++ BUG_ON(rc);
++ frames -= nr_gframes; /* adjust after map_pte_fn() */
++#endif /* CONFIG_X86 */
+
-+ wait_queue_head_t wq;
-+ struct task_struct *xenblkd;
-+ unsigned int waiting_reqs;
-+ request_queue_t *plug;
++ kfree(frames);
+
-+ /* statistics */
-+ unsigned long st_print;
-+ int st_rd_req;
-+ int st_wr_req;
-+ int st_oo_req;
-+ int st_br_req;
-+ int st_rd_sect;
-+ int st_wr_sect;
++ return 0;
++}
+
-+ wait_queue_head_t waiting_to_free;
++static void gnttab_page_free(struct page *page)
++{
++ ClearPageForeign(page);
++ gnttab_reset_grant_page(page);
++ put_page(page);
++}
+
-+ grant_handle_t shmem_handle;
-+ grant_ref_t shmem_ref;
-+} blkif_t;
++/*
++ * Must not be called with IRQs off. This should only be used on the
++ * slow path.
++ *
++ * Copy a foreign granted page to local memory.
++ */
++int gnttab_copy_grant_page(grant_ref_t ref, struct page **pagep)
++{
++ struct gnttab_unmap_and_replace unmap;
++ mmu_update_t mmu;
++ struct page *page;
++ struct page *new_page;
++ void *new_addr;
++ void *addr;
++ paddr_t pfn;
++ maddr_t mfn;
++ maddr_t new_mfn;
++ int err;
+
-+blkif_t *blkif_alloc(domid_t domid);
-+void blkif_disconnect(blkif_t *blkif);
-+void blkif_free(blkif_t *blkif);
-+int blkif_map(blkif_t *blkif, unsigned long shared_page, unsigned int evtchn);
++ page = *pagep;
++ if (!get_page_unless_zero(page))
++ return -ENOENT;
+
-+#define blkif_get(_b) (atomic_inc(&(_b)->refcnt))
-+#define blkif_put(_b) \
-+ do { \
-+ if (atomic_dec_and_test(&(_b)->refcnt)) \
-+ wake_up(&(_b)->waiting_to_free);\
-+ } while (0)
++ err = -ENOMEM;
++ new_page = alloc_page(GFP_ATOMIC | __GFP_NOWARN);
++ if (!new_page)
++ goto out;
+
-+/* Create a vbd. */
-+int vbd_create(blkif_t *blkif, blkif_vdev_t vdevice, unsigned major,
-+ unsigned minor, int readonly);
-+void vbd_free(struct vbd *vbd);
++ new_addr = page_address(new_page);
++ addr = page_address(page);
++ memcpy(new_addr, addr, PAGE_SIZE);
+
-+unsigned long long vbd_size(struct vbd *vbd);
-+unsigned int vbd_info(struct vbd *vbd);
-+unsigned long vbd_secsize(struct vbd *vbd);
++ pfn = page_to_pfn(page);
++ mfn = pfn_to_mfn(pfn);
++ new_mfn = virt_to_mfn(new_addr);
+
-+struct phys_req {
-+ unsigned short dev;
-+ unsigned short nr_sects;
-+ struct block_device *bdev;
-+ blkif_sector_t sector_number;
-+};
++ write_seqlock(&gnttab_dma_lock);
+
-+int vbd_translate(struct phys_req *req, blkif_t *blkif, int operation);
++ /* Make seq visible before checking page_mapped. */
++ smp_mb();
+
-+void blkif_interface_init(void);
++ /* Has the page been DMA-mapped? */
++ if (unlikely(page_mapped(page))) {
++ write_sequnlock(&gnttab_dma_lock);
++ put_page(new_page);
++ err = -EBUSY;
++ goto out;
++ }
+
-+void blkif_xenbus_init(void);
++ if (!xen_feature(XENFEAT_auto_translated_physmap))
++ set_phys_to_machine(pfn, new_mfn);
+
-+irqreturn_t blkif_be_int(int irq, void *dev_id, struct pt_regs *regs);
-+int blkif_schedule(void *arg);
++ gnttab_set_replace_op(&unmap, (unsigned long)addr,
++ (unsigned long)new_addr, ref);
+
-+int blkback_barrier(struct xenbus_transaction xbt,
-+ struct backend_info *be, int state);
++ err = HYPERVISOR_grant_table_op(GNTTABOP_unmap_and_replace,
++ &unmap, 1);
++ BUG_ON(err);
++ BUG_ON(unmap.status);
+
-+#endif /* __BLKIF__BACKEND__COMMON_H__ */
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blkback/interface.c tmp-linux-2.6-xen.patch/drivers/xen/blkback/interface.c
---- pristine-linux-2.6.18.2/drivers/xen/blkback/interface.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blkback/interface.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,181 @@
-+/******************************************************************************
-+ * arch/xen/drivers/blkif/backend/interface.c
-+ *
-+ * Block-device interface management.
-+ *
-+ * Copyright (c) 2004, Keir Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ write_sequnlock(&gnttab_dma_lock);
+
-+#include "common.h"
-+#include <xen/evtchn.h>
-+#include <linux/kthread.h>
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ set_phys_to_machine(page_to_pfn(new_page), INVALID_P2M_ENTRY);
+
-+static kmem_cache_t *blkif_cachep;
++ mmu.ptr = (new_mfn << PAGE_SHIFT) | MMU_MACHPHYS_UPDATE;
++ mmu.val = pfn;
++ err = HYPERVISOR_mmu_update(&mmu, 1, NULL, DOMID_SELF);
++ BUG_ON(err);
++ }
+
-+blkif_t *blkif_alloc(domid_t domid)
-+{
-+ blkif_t *blkif;
++ new_page->mapping = page->mapping;
++ new_page->index = page->index;
++ set_bit(PG_foreign, &new_page->flags);
++ *pagep = new_page;
+
-+ blkif = kmem_cache_alloc(blkif_cachep, GFP_KERNEL);
-+ if (!blkif)
-+ return ERR_PTR(-ENOMEM);
++ SetPageForeign(page, gnttab_page_free);
++ page->mapping = NULL;
+
-+ memset(blkif, 0, sizeof(*blkif));
-+ blkif->domid = domid;
-+ spin_lock_init(&blkif->blk_ring_lock);
-+ atomic_set(&blkif->refcnt, 1);
-+ init_waitqueue_head(&blkif->wq);
-+ blkif->st_print = jiffies;
-+ init_waitqueue_head(&blkif->waiting_to_free);
++out:
++ put_page(page);
++ return err;
++}
++EXPORT_SYMBOL_GPL(gnttab_copy_grant_page);
+
-+ return blkif;
++void gnttab_reset_grant_page(struct page *page)
++{
++ init_page_count(page);
++ reset_page_mapcount(page);
+}
++EXPORT_SYMBOL_GPL(gnttab_reset_grant_page);
+
-+static int map_frontend_page(blkif_t *blkif, unsigned long shared_page)
++/*
++ * Keep track of foreign pages marked as PageForeign so that we don't
++ * return them to the remote domain prematurely.
++ *
++ * PageForeign pages are pinned down by increasing their mapcount.
++ *
++ * All other pages are simply returned as is.
++ */
++void __gnttab_dma_map_page(struct page *page)
+{
-+ struct gnttab_map_grant_ref op;
++ unsigned int seq;
+
-+ gnttab_set_map_op(&op, (unsigned long)blkif->blk_ring_area->addr,
-+ GNTMAP_host_map, shared_page, blkif->domid);
++ if (!is_running_on_xen() || !PageForeign(page))
++ return;
+
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
-+ BUG();
++ do {
++ seq = read_seqbegin(&gnttab_dma_lock);
+
-+ if (op.status) {
-+ DPRINTK(" Grant table operation failure !\n");
-+ return op.status;
-+ }
++ if (gnttab_dma_local_pfn(page))
++ break;
+
-+ blkif->shmem_ref = shared_page;
-+ blkif->shmem_handle = op.handle;
++ atomic_set(&page->_mapcount, 0);
+
-+ return 0;
++ /* Make _mapcount visible before read_seqretry. */
++ smp_mb();
++ } while (unlikely(read_seqretry(&gnttab_dma_lock, seq)));
+}
+
-+static void unmap_frontend_page(blkif_t *blkif)
++int gnttab_resume(void)
+{
-+ struct gnttab_unmap_grant_ref op;
-+
-+ gnttab_set_unmap_op(&op, (unsigned long)blkif->blk_ring_area->addr,
-+ GNTMAP_host_map, blkif->shmem_handle);
-+
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
-+ BUG();
++ if (max_nr_grant_frames() < nr_grant_frames)
++ return -ENOSYS;
++ return gnttab_map(0, nr_grant_frames - 1);
+}
+
-+int blkif_map(blkif_t *blkif, unsigned long shared_page, unsigned int evtchn)
++int gnttab_suspend(void)
+{
-+ int err;
++#ifdef CONFIG_X86
++ apply_to_page_range(&init_mm, (unsigned long)shared,
++ PAGE_SIZE * nr_grant_frames,
++ unmap_pte_fn, NULL);
++#endif
++ return 0;
++}
+
-+ /* Already connected through? */
-+ if (blkif->irq)
-+ return 0;
++#else /* !CONFIG_XEN */
+
-+ if ( (blkif->blk_ring_area = alloc_vm_area(PAGE_SIZE)) == NULL )
-+ return -ENOMEM;
++#include <platform-pci.h>
+
-+ err = map_frontend_page(blkif, shared_page);
-+ if (err) {
-+ free_vm_area(blkif->blk_ring_area);
-+ return err;
-+ }
++static unsigned long resume_frames;
+
-+ switch (blkif->blk_protocol) {
-+ case BLKIF_PROTOCOL_NATIVE:
-+ {
-+ blkif_sring_t *sring;
-+ sring = (blkif_sring_t *)blkif->blk_ring_area->addr;
-+ BACK_RING_INIT(&blkif->blk_rings.native, sring, PAGE_SIZE);
-+ break;
-+ }
-+ case BLKIF_PROTOCOL_X86_32:
-+ {
-+ blkif_x86_32_sring_t *sring_x86_32;
-+ sring_x86_32 = (blkif_x86_32_sring_t *)blkif->blk_ring_area->addr;
-+ BACK_RING_INIT(&blkif->blk_rings.x86_32, sring_x86_32, PAGE_SIZE);
-+ break;
-+ }
-+ case BLKIF_PROTOCOL_X86_64:
-+ {
-+ blkif_x86_64_sring_t *sring_x86_64;
-+ sring_x86_64 = (blkif_x86_64_sring_t *)blkif->blk_ring_area->addr;
-+ BACK_RING_INIT(&blkif->blk_rings.x86_64, sring_x86_64, PAGE_SIZE);
-+ break;
-+ }
-+ default:
-+ BUG();
-+ }
++static int gnttab_map(unsigned int start_idx, unsigned int end_idx)
++{
++ struct xen_add_to_physmap xatp;
++ unsigned int i = end_idx;
+
-+ err = bind_interdomain_evtchn_to_irqhandler(
-+ blkif->domid, evtchn, blkif_be_int, 0, "blkif-backend", blkif);
-+ if (err < 0)
-+ {
-+ unmap_frontend_page(blkif);
-+ free_vm_area(blkif->blk_ring_area);
-+ blkif->blk_rings.common.sring = NULL;
-+ return err;
-+ }
-+ blkif->irq = err;
++ /* Loop backwards, so that the first hypercall has the largest index,
++ * ensuring that the table will grow only once.
++ */
++ do {
++ xatp.domid = DOMID_SELF;
++ xatp.idx = i;
++ xatp.space = XENMAPSPACE_grant_table;
++ xatp.gpfn = (resume_frames >> PAGE_SHIFT) + i;
++ if (HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp))
++ BUG();
++ } while (i-- > start_idx);
+
+ return 0;
+}
+
-+void blkif_disconnect(blkif_t *blkif)
++int gnttab_resume(void)
+{
-+ if (blkif->xenblkd) {
-+ kthread_stop(blkif->xenblkd);
-+ blkif->xenblkd = NULL;
-+ }
-+
-+ atomic_dec(&blkif->refcnt);
-+ wait_event(blkif->waiting_to_free, atomic_read(&blkif->refcnt) == 0);
-+ atomic_inc(&blkif->refcnt);
++ unsigned int max_nr_gframes, nr_gframes;
+
-+ if (blkif->irq) {
-+ unbind_from_irqhandler(blkif->irq, blkif);
-+ blkif->irq = 0;
-+ }
++ nr_gframes = nr_grant_frames;
++ max_nr_gframes = max_nr_grant_frames();
++ if (max_nr_gframes < nr_gframes)
++ return -ENOSYS;
+
-+ if (blkif->blk_rings.common.sring) {
-+ unmap_frontend_page(blkif);
-+ free_vm_area(blkif->blk_ring_area);
-+ blkif->blk_rings.common.sring = NULL;
++ if (!resume_frames) {
++ resume_frames = alloc_xen_mmio(PAGE_SIZE * max_nr_gframes);
++ shared = ioremap(resume_frames, PAGE_SIZE * max_nr_gframes);
++ if (shared == NULL) {
++ printk("error to ioremap gnttab share frames\n");
++ return -1;
++ }
+ }
-+}
+
-+void blkif_free(blkif_t *blkif)
-+{
-+ if (!atomic_dec_and_test(&blkif->refcnt))
-+ BUG();
-+ kmem_cache_free(blkif_cachep, blkif);
-+}
++ gnttab_map(0, nr_gframes - 1);
+
-+void __init blkif_interface_init(void)
-+{
-+ blkif_cachep = kmem_cache_create("blkif_cache", sizeof(blkif_t),
-+ 0, 0, NULL, NULL);
++ return 0;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blkback/Makefile tmp-linux-2.6-xen.patch/drivers/xen/blkback/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/blkback/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blkback/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,3 @@
-+obj-$(CONFIG_XEN_BLKDEV_BACKEND) := blkbk.o
+
-+blkbk-y := blkback.o xenbus.o interface.o vbd.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blkback/vbd.c tmp-linux-2.6-xen.patch/drivers/xen/blkback/vbd.c
---- pristine-linux-2.6.18.2/drivers/xen/blkback/vbd.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blkback/vbd.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,118 @@
-+/******************************************************************************
-+ * blkback/vbd.c
-+ *
-+ * Routines for managing virtual block devices (VBDs).
-+ *
-+ * Copyright (c) 2003-2005, Keir Fraser & Steve Hand
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++#endif /* !CONFIG_XEN */
+
-+#include "common.h"
++static int gnttab_expand(unsigned int req_entries)
++{
++ int rc;
++ unsigned int cur, extra;
+
-+#define vbd_sz(_v) ((_v)->bdev->bd_part ? \
-+ (_v)->bdev->bd_part->nr_sects : (_v)->bdev->bd_disk->capacity)
++ cur = nr_grant_frames;
++ extra = ((req_entries + (ENTRIES_PER_GRANT_FRAME-1)) /
++ ENTRIES_PER_GRANT_FRAME);
++ if (cur + extra > max_nr_grant_frames())
++ return -ENOSPC;
+
-+unsigned long long vbd_size(struct vbd *vbd)
-+{
-+ return vbd_sz(vbd);
-+}
++ if ((rc = gnttab_map(cur, cur + extra - 1)) == 0)
++ rc = grow_gnttab_list(extra);
+
-+unsigned int vbd_info(struct vbd *vbd)
-+{
-+ return vbd->type | (vbd->readonly?VDISK_READONLY:0);
++ return rc;
+}
+
-+unsigned long vbd_secsize(struct vbd *vbd)
++int __devinit gnttab_init(void)
+{
-+ return bdev_hardsect_size(vbd->bdev);
-+}
++ int i;
++ unsigned int max_nr_glist_frames, nr_glist_frames;
++ unsigned int nr_init_grefs;
+
-+int vbd_create(blkif_t *blkif, blkif_vdev_t handle, unsigned major,
-+ unsigned minor, int readonly)
-+{
-+ struct vbd *vbd;
-+ struct block_device *bdev;
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+ vbd = &blkif->vbd;
-+ vbd->handle = handle;
-+ vbd->readonly = readonly;
-+ vbd->type = 0;
++ nr_grant_frames = 1;
++ boot_max_nr_grant_frames = __max_nr_grant_frames();
+
-+ vbd->pdevice = MKDEV(major, minor);
++ /* Determine the maximum number of frames required for the
++ * grant reference free list on the current hypervisor.
++ */
++ max_nr_glist_frames = nr_freelist_frames(boot_max_nr_grant_frames);
+
-+ bdev = open_by_devnum(vbd->pdevice,
-+ vbd->readonly ? FMODE_READ : FMODE_WRITE);
++ gnttab_list = kmalloc(max_nr_glist_frames * sizeof(grant_ref_t *),
++ GFP_KERNEL);
++ if (gnttab_list == NULL)
++ return -ENOMEM;
+
-+ if (IS_ERR(bdev)) {
-+ DPRINTK("vbd_creat: device %08x could not be opened.\n",
-+ vbd->pdevice);
-+ return -ENOENT;
++ nr_glist_frames = nr_freelist_frames(nr_grant_frames);
++ for (i = 0; i < nr_glist_frames; i++) {
++ gnttab_list[i] = (grant_ref_t *)__get_free_page(GFP_KERNEL);
++ if (gnttab_list[i] == NULL)
++ goto ini_nomem;
+ }
+
-+ vbd->bdev = bdev;
++ if (gnttab_resume() < 0)
++ return -ENODEV;
+
-+ if (vbd->bdev->bd_disk == NULL) {
-+ DPRINTK("vbd_creat: device %08x doesn't exist.\n",
-+ vbd->pdevice);
-+ vbd_free(vbd);
-+ return -ENOENT;
-+ }
++ nr_init_grefs = nr_grant_frames * ENTRIES_PER_GRANT_FRAME;
+
-+ if (vbd->bdev->bd_disk->flags & GENHD_FL_CD)
-+ vbd->type |= VDISK_CDROM;
-+ if (vbd->bdev->bd_disk->flags & GENHD_FL_REMOVABLE)
-+ vbd->type |= VDISK_REMOVABLE;
++ for (i = NR_RESERVED_ENTRIES; i < nr_init_grefs - 1; i++)
++ gnttab_entry(i) = i + 1;
++
++ gnttab_entry(nr_init_grefs - 1) = GNTTAB_LIST_END;
++ gnttab_free_count = nr_init_grefs - NR_RESERVED_ENTRIES;
++ gnttab_free_head = NR_RESERVED_ENTRIES;
+
-+ DPRINTK("Successful creation of handle=%04x (dom=%u)\n",
-+ handle, blkif->domid);
+ return 0;
++
++ ini_nomem:
++ for (i--; i >= 0; i--)
++ free_page((unsigned long)gnttab_list[i]);
++ kfree(gnttab_list);
++ return -ENOMEM;
+}
+
-+void vbd_free(struct vbd *vbd)
++#ifdef CONFIG_XEN
++core_initcall(gnttab_init);
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/hypervisor_sysfs.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/hypervisor_sysfs.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,57 @@
++/*
++ * copyright (c) 2006 IBM Corporation
++ * Authored by: Mike D. Day <ncmike at us.ibm.com>
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License version 2 as
++ * published by the Free Software Foundation.
++ */
++
++#include <linux/kernel.h>
++#include <linux/module.h>
++#include <linux/kobject.h>
++#include <xen/hypervisor_sysfs.h>
++#include <asm/hypervisor.h>
++
++static ssize_t hyp_sysfs_show(struct kobject *kobj,
++ struct attribute *attr,
++ char *buffer)
+{
-+ if (vbd->bdev)
-+ blkdev_put(vbd->bdev);
-+ vbd->bdev = NULL;
++ struct hyp_sysfs_attr *hyp_attr;
++ hyp_attr = container_of(attr, struct hyp_sysfs_attr, attr);
++ if (hyp_attr->show)
++ return hyp_attr->show(hyp_attr, buffer);
++ return 0;
+}
+
-+int vbd_translate(struct phys_req *req, blkif_t *blkif, int operation)
++static ssize_t hyp_sysfs_store(struct kobject *kobj,
++ struct attribute *attr,
++ const char *buffer,
++ size_t len)
+{
-+ struct vbd *vbd = &blkif->vbd;
-+ int rc = -EACCES;
++ struct hyp_sysfs_attr *hyp_attr;
++ hyp_attr = container_of(attr, struct hyp_sysfs_attr, attr);
++ if (hyp_attr->store)
++ return hyp_attr->store(hyp_attr, buffer, len);
++ return 0;
++}
+
-+ if ((operation != READ) && vbd->readonly)
-+ goto out;
++static struct sysfs_ops hyp_sysfs_ops = {
++ .show = hyp_sysfs_show,
++ .store = hyp_sysfs_store,
++};
+
-+ if (unlikely((req->sector_number + req->nr_sects) > vbd_sz(vbd)))
-+ goto out;
++static struct kobj_type hyp_sysfs_kobj_type = {
++ .sysfs_ops = &hyp_sysfs_ops,
++};
+
-+ req->dev = vbd->pdevice;
-+ req->bdev = vbd->bdev;
-+ rc = 0;
++static int __init hypervisor_subsys_init(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+ out:
-+ return rc;
++ hypervisor_subsys.kset.kobj.ktype = &hyp_sysfs_kobj_type;
++ return 0;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blkback/xenbus.c tmp-linux-2.6-xen.patch/drivers/xen/blkback/xenbus.c
---- pristine-linux-2.6.18.2/drivers/xen/blkback/xenbus.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blkback/xenbus.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,533 @@
-+/* Xenbus code for blkif backend
-+ Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
-+ Copyright (C) 2005 XenSource Ltd
+
-+ This program is free software; you can redistribute it and/or modify
-+ it under the terms of the GNU General Public License as published by
-+ the Free Software Foundation; either version 2 of the License, or
-+ (at your option) any later version.
++device_initcall(hypervisor_subsys_init);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/machine_kexec.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/machine_kexec.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,207 @@
++/*
++ * drivers/xen/core/machine_kexec.c
++ * handle transition of Linux booting another kernel
++ */
+
-+ This program is distributed in the hope that it will be useful,
-+ but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ GNU General Public License for more details.
++#include <linux/kexec.h>
++#include <xen/interface/kexec.h>
++#include <linux/mm.h>
++#include <linux/bootmem.h>
+
-+ You should have received a copy of the GNU General Public License
-+ along with this program; if not, write to the Free Software
-+ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
-+*/
++extern void machine_kexec_setup_load_arg(xen_kexec_image_t *xki,
++ struct kimage *image);
++extern int machine_kexec_setup_resources(struct resource *hypervisor,
++ struct resource *phys_cpus,
++ int nr_phys_cpus);
++extern void machine_kexec_register_resources(struct resource *res);
+
-+#include <stdarg.h>
-+#include <linux/module.h>
-+#include <linux/kthread.h>
-+#include "common.h"
++static int __initdata xen_max_nr_phys_cpus;
++static struct resource xen_hypervisor_res;
++static struct resource *xen_phys_cpus;
+
-+#undef DPRINTK
-+#define DPRINTK(fmt, args...) \
-+ pr_debug("blkback/xenbus (%s:%d) " fmt ".\n", \
-+ __FUNCTION__, __LINE__, ##args)
++size_t vmcoreinfo_size_xen;
++unsigned long paddr_vmcoreinfo_xen;
+
-+struct backend_info
++void __init xen_machine_kexec_setup_resources(void)
+{
-+ struct xenbus_device *dev;
-+ blkif_t *blkif;
-+ struct xenbus_watch backend_watch;
-+ unsigned major;
-+ unsigned minor;
-+ char *mode;
-+};
++ xen_kexec_range_t range;
++ struct resource *res;
++ int k = 0;
+
-+static void connect(struct backend_info *);
-+static int connect_ring(struct backend_info *);
-+static void backend_changed(struct xenbus_watch *, const char **,
-+ unsigned int);
++ if (!is_initial_xendomain())
++ return;
+
-+static int blkback_name(blkif_t *blkif, char *buf)
-+{
-+ char *devpath, *devname;
-+ struct xenbus_device *dev = blkif->be->dev;
++ /* determine maximum number of physical cpus */
+
-+ devpath = xenbus_read(XBT_NIL, dev->nodename, "dev", NULL);
-+ if (IS_ERR(devpath))
-+ return PTR_ERR(devpath);
-+
-+ if ((devname = strstr(devpath, "/dev/")) != NULL)
-+ devname += strlen("/dev/");
-+ else
-+ devname = devpath;
++ while (1) {
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_CPU;
++ range.nr = k;
+
-+ snprintf(buf, TASK_COMM_LEN, "blkback.%d.%s", blkif->domid, devname);
-+ kfree(devpath);
-+
-+ return 0;
-+}
++ if(HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ break;
+
-+static void update_blkif_status(blkif_t *blkif)
-+{
-+ int err;
-+ char name[TASK_COMM_LEN];
++ k++;
++ }
+
-+ /* Not ready to connect? */
-+ if (!blkif->irq || !blkif->vbd.bdev)
++ if (k == 0)
+ return;
+
-+ /* Already connected? */
-+ if (blkif->be->dev->state == XenbusStateConnected)
-+ return;
++ xen_max_nr_phys_cpus = k;
+
-+ /* Attempt to connect: exit if we fail to. */
-+ connect(blkif->be);
-+ if (blkif->be->dev->state != XenbusStateConnected)
-+ return;
++ /* allocate xen_phys_cpus */
+
-+ err = blkback_name(blkif, name);
-+ if (err) {
-+ xenbus_dev_error(blkif->be->dev, err, "get blkback dev name");
-+ return;
-+ }
++ xen_phys_cpus = alloc_bootmem_low(k * sizeof(struct resource));
++ BUG_ON(xen_phys_cpus == NULL);
+
-+ blkif->xenblkd = kthread_run(blkif_schedule, blkif, name);
-+ if (IS_ERR(blkif->xenblkd)) {
-+ err = PTR_ERR(blkif->xenblkd);
-+ blkif->xenblkd = NULL;
-+ xenbus_dev_error(blkif->be->dev, err, "start xenblkd");
++ /* fill in xen_phys_cpus with per-cpu crash note information */
++
++ for (k = 0; k < xen_max_nr_phys_cpus; k++) {
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_CPU;
++ range.nr = k;
++
++ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ goto err;
++
++ res = xen_phys_cpus + k;
++
++ memset(res, 0, sizeof(*res));
++ res->name = "Crash note";
++ res->start = range.start;
++ res->end = range.start + range.size - 1;
++ res->flags = IORESOURCE_BUSY | IORESOURCE_MEM;
+ }
-+}
+
++ /* fill in xen_hypervisor_res with hypervisor machine address range */
+
-+/****************************************************************
-+ * sysfs interface for VBD I/O requests
-+ */
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_XEN;
+
-+#define VBD_SHOW(name, format, args...) \
-+ static ssize_t show_##name(struct device *_dev, \
-+ struct device_attribute *attr, \
-+ char *buf) \
-+ { \
-+ struct xenbus_device *dev = to_xenbus_device(_dev); \
-+ struct backend_info *be = dev->dev.driver_data; \
-+ \
-+ return sprintf(buf, format, ##args); \
-+ } \
-+ static DEVICE_ATTR(name, S_IRUGO, show_##name, NULL)
++ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ goto err;
+
-+VBD_SHOW(oo_req, "%d\n", be->blkif->st_oo_req);
-+VBD_SHOW(rd_req, "%d\n", be->blkif->st_rd_req);
-+VBD_SHOW(wr_req, "%d\n", be->blkif->st_wr_req);
-+VBD_SHOW(br_req, "%d\n", be->blkif->st_br_req);
-+VBD_SHOW(rd_sect, "%d\n", be->blkif->st_rd_sect);
-+VBD_SHOW(wr_sect, "%d\n", be->blkif->st_wr_sect);
++ xen_hypervisor_res.name = "Hypervisor code and data";
++ xen_hypervisor_res.start = range.start;
++ xen_hypervisor_res.end = range.start + range.size - 1;
++ xen_hypervisor_res.flags = IORESOURCE_BUSY | IORESOURCE_MEM;
+
-+static struct attribute *vbdstat_attrs[] = {
-+ &dev_attr_oo_req.attr,
-+ &dev_attr_rd_req.attr,
-+ &dev_attr_wr_req.attr,
-+ &dev_attr_br_req.attr,
-+ &dev_attr_rd_sect.attr,
-+ &dev_attr_wr_sect.attr,
-+ NULL
-+};
++ /* fill in crashk_res if range is reserved by hypervisor */
+
-+static struct attribute_group vbdstat_group = {
-+ .name = "statistics",
-+ .attrs = vbdstat_attrs,
-+};
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_CRASH;
++
++ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ goto err;
++
++ if (range.size) {
++ crashk_res.start = range.start;
++ crashk_res.end = range.start + range.size - 1;
++ }
+
-+VBD_SHOW(physical_device, "%x:%x\n", be->major, be->minor);
-+VBD_SHOW(mode, "%s\n", be->mode);
++ /* get physical address of vmcoreinfo */
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_VMCOREINFO;
+
-+int xenvbd_sysfs_addif(struct xenbus_device *dev)
-+{
-+ int error;
-+
-+ error = device_create_file(&dev->dev, &dev_attr_physical_device);
-+ if (error)
-+ goto fail1;
++ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ return;
+
-+ error = device_create_file(&dev->dev, &dev_attr_mode);
-+ if (error)
-+ goto fail2;
++ if (range.size) {
++ paddr_vmcoreinfo_xen = range.start;
++ vmcoreinfo_size_xen = range.size;
++ }
+
-+ error = sysfs_create_group(&dev->dev.kobj, &vbdstat_group);
-+ if (error)
-+ goto fail3;
++ if (machine_kexec_setup_resources(&xen_hypervisor_res, xen_phys_cpus,
++ xen_max_nr_phys_cpus))
++ goto err;
+
-+ return 0;
++ return;
+
-+fail3: sysfs_remove_group(&dev->dev.kobj, &vbdstat_group);
-+fail2: device_remove_file(&dev->dev, &dev_attr_mode);
-+fail1: device_remove_file(&dev->dev, &dev_attr_physical_device);
-+ return error;
++ err:
++ /*
++ * It isn't possible to free xen_phys_cpus this early in the
++ * boot. Failure at this stage is unexpected and the amount of
++ * memory is small therefore we tolerate the potential leak.
++ */
++ xen_max_nr_phys_cpus = 0;
++ return;
+}
+
-+void xenvbd_sysfs_delif(struct xenbus_device *dev)
++void __init xen_machine_kexec_register_resources(struct resource *res)
+{
-+ sysfs_remove_group(&dev->dev.kobj, &vbdstat_group);
-+ device_remove_file(&dev->dev, &dev_attr_mode);
-+ device_remove_file(&dev->dev, &dev_attr_physical_device);
++ request_resource(res, &xen_hypervisor_res);
++ machine_kexec_register_resources(res);
+}
+
-+static int blkback_remove(struct xenbus_device *dev)
++static void setup_load_arg(xen_kexec_image_t *xki, struct kimage *image)
+{
-+ struct backend_info *be = dev->dev.driver_data;
-+
-+ DPRINTK("");
-+
-+ if (be->major || be->minor)
-+ xenvbd_sysfs_delif(dev);
++ machine_kexec_setup_load_arg(xki, image);
+
-+ if (be->backend_watch.node) {
-+ unregister_xenbus_watch(&be->backend_watch);
-+ kfree(be->backend_watch.node);
-+ be->backend_watch.node = NULL;
-+ }
++ xki->indirection_page = image->head;
++ xki->start_address = image->start;
++}
+
-+ if (be->blkif) {
-+ blkif_disconnect(be->blkif);
-+ vbd_free(&be->blkif->vbd);
-+ blkif_free(be->blkif);
-+ be->blkif = NULL;
-+ }
++/*
++ * Load the image into xen so xen can kdump itself
++ * This might have been done in prepare, but prepare
++ * is currently called too early. It might make sense
++ * to move prepare, but for now, just add an extra hook.
++ */
++int xen_machine_kexec_load(struct kimage *image)
++{
++ xen_kexec_load_t xkl;
+
-+ kfree(be);
-+ dev->dev.driver_data = NULL;
-+ return 0;
++ memset(&xkl, 0, sizeof(xkl));
++ xkl.type = image->type;
++ setup_load_arg(&xkl.image, image);
++ return HYPERVISOR_kexec_op(KEXEC_CMD_kexec_load, &xkl);
+}
+
-+int blkback_barrier(struct xenbus_transaction xbt,
-+ struct backend_info *be, int state)
++/*
++ * Unload the image that was stored by machine_kexec_load()
++ * This might have been done in machine_kexec_cleanup() but it
++ * is called too late, and its possible xen could try and kdump
++ * using resources that have been freed.
++ */
++void xen_machine_kexec_unload(struct kimage *image)
+{
-+ struct xenbus_device *dev = be->dev;
-+ int err;
-+
-+ err = xenbus_printf(xbt, dev->nodename, "feature-barrier",
-+ "%d", state);
-+ if (err)
-+ xenbus_dev_fatal(dev, err, "writing feature-barrier");
++ xen_kexec_load_t xkl;
+
-+ return err;
++ memset(&xkl, 0, sizeof(xkl));
++ xkl.type = image->type;
++ WARN_ON(HYPERVISOR_kexec_op(KEXEC_CMD_kexec_unload, &xkl));
+}
+
-+/**
-+ * Entry point to this code when a new device is created. Allocate the basic
-+ * structures, and watch the store waiting for the hotplug scripts to tell us
-+ * the device's physical major and minor numbers. Switch to InitWait.
++/*
++ * Do not allocate memory (or fail in any way) in machine_kexec().
++ * We are past the point of no return, committed to rebooting now.
++ *
++ * This has the hypervisor move to the prefered reboot CPU,
++ * stop all CPUs and kexec. That is it combines machine_shutdown()
++ * and machine_kexec() in Linux kexec terms.
+ */
-+static int blkback_probe(struct xenbus_device *dev,
-+ const struct xenbus_device_id *id)
++NORET_TYPE void machine_kexec(struct kimage *image)
+{
-+ int err;
-+ struct backend_info *be = kzalloc(sizeof(struct backend_info),
-+ GFP_KERNEL);
-+ if (!be) {
-+ xenbus_dev_fatal(dev, -ENOMEM,
-+ "allocating backend structure");
-+ return -ENOMEM;
-+ }
-+ be->dev = dev;
-+ dev->dev.driver_data = be;
++ xen_kexec_exec_t xke;
+
-+ be->blkif = blkif_alloc(dev->otherend_id);
-+ if (IS_ERR(be->blkif)) {
-+ err = PTR_ERR(be->blkif);
-+ be->blkif = NULL;
-+ xenbus_dev_fatal(dev, err, "creating block interface");
-+ goto fail;
-+ }
++ memset(&xke, 0, sizeof(xke));
++ xke.type = image->type;
++ VOID(HYPERVISOR_kexec_op(KEXEC_CMD_kexec, &xke));
++ panic("KEXEC_CMD_kexec hypercall should not return\n");
++}
+
-+ /* setup back pointer */
-+ be->blkif->be = be;
++void machine_shutdown(void)
++{
++ /* do nothing */
++}
+
-+ err = xenbus_watch_path2(dev, dev->nodename, "physical-device",
-+ &be->backend_watch, backend_changed);
-+ if (err)
-+ goto fail;
+
-+ err = xenbus_switch_state(dev, XenbusStateInitWait);
-+ if (err)
-+ goto fail;
++/*
++ * Local variables:
++ * c-file-style: "linux"
++ * indent-tabs-mode: t
++ * c-indent-level: 8
++ * c-basic-offset: 8
++ * tab-width: 8
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/machine_reboot.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/machine_reboot.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,247 @@
++#include <linux/version.h>
++#include <linux/kernel.h>
++#include <linux/mm.h>
++#include <linux/unistd.h>
++#include <linux/module.h>
++#include <linux/reboot.h>
++#include <linux/sysrq.h>
++#include <linux/stringify.h>
++#include <linux/stop_machine.h>
++#include <asm/irq.h>
++#include <asm/mmu_context.h>
++#include <xen/evtchn.h>
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
++#include <linux/cpu.h>
++#include <xen/gnttab.h>
++#include <xen/xencons.h>
++#include <xen/cpu_hotplug.h>
++#include <xen/interface/vcpu.h>
+
-+ return 0;
++#if defined(__i386__) || defined(__x86_64__)
+
-+fail:
-+ DPRINTK("failed");
-+ blkback_remove(dev);
-+ return err;
++/*
++ * Power off function, if any
++ */
++void (*pm_power_off)(void);
++EXPORT_SYMBOL(pm_power_off);
++
++void machine_emergency_restart(void)
++{
++ /* We really want to get pending console data out before we die. */
++ xencons_force_flush();
++ HYPERVISOR_shutdown(SHUTDOWN_reboot);
+}
+
++void machine_restart(char * __unused)
++{
++ machine_emergency_restart();
++}
+
-+/**
-+ * Callback received when the hotplug scripts have placed the physical-device
-+ * node. Read it and the mode node, and create a vbd. If the frontend is
-+ * ready, connect.
-+ */
-+static void backend_changed(struct xenbus_watch *watch,
-+ const char **vec, unsigned int len)
++void machine_halt(void)
+{
-+ int err;
-+ unsigned major;
-+ unsigned minor;
-+ struct backend_info *be
-+ = container_of(watch, struct backend_info, backend_watch);
-+ struct xenbus_device *dev = be->dev;
++ machine_power_off();
++}
+
-+ DPRINTK("");
++void machine_power_off(void)
++{
++ /* We really want to get pending console data out before we die. */
++ xencons_force_flush();
++ if (pm_power_off)
++ pm_power_off();
++ HYPERVISOR_shutdown(SHUTDOWN_poweroff);
++}
+
-+ err = xenbus_scanf(XBT_NIL, dev->nodename, "physical-device", "%x:%x",
-+ &major, &minor);
-+ if (XENBUS_EXIST_ERR(err)) {
-+ /* Since this watch will fire once immediately after it is
-+ registered, we expect this. Ignore it, and wait for the
-+ hotplug scripts. */
-+ return;
-+ }
-+ if (err != 2) {
-+ xenbus_dev_fatal(dev, err, "reading physical-device");
-+ return;
-+ }
++int reboot_thru_bios = 0; /* for dmi_scan.c */
++EXPORT_SYMBOL(machine_restart);
++EXPORT_SYMBOL(machine_halt);
++EXPORT_SYMBOL(machine_power_off);
+
-+ if ((be->major || be->minor) &&
-+ ((be->major != major) || (be->minor != minor))) {
-+ printk(KERN_WARNING
-+ "blkback: changing physical device (from %x:%x to "
-+ "%x:%x) not supported.\n", be->major, be->minor,
-+ major, minor);
-+ return;
-+ }
++static void pre_suspend(void)
++{
++ HYPERVISOR_shared_info = (shared_info_t *)empty_zero_page;
++ WARN_ON(HYPERVISOR_update_va_mapping(fix_to_virt(FIX_SHARED_INFO),
++ __pte_ma(0), 0));
+
-+ be->mode = xenbus_read(XBT_NIL, dev->nodename, "mode", NULL);
-+ if (IS_ERR(be->mode)) {
-+ err = PTR_ERR(be->mode);
-+ be->mode = NULL;
-+ xenbus_dev_fatal(dev, err, "reading mode");
-+ return;
-+ }
++ xen_start_info->store_mfn = mfn_to_pfn(xen_start_info->store_mfn);
++ xen_start_info->console.domU.mfn =
++ mfn_to_pfn(xen_start_info->console.domU.mfn);
++}
+
-+ if (be->major == 0 && be->minor == 0) {
-+ /* Front end dir is a number, which is used as the handle. */
++static void post_suspend(int suspend_cancelled)
++{
++ int i, j, k, fpp;
++ unsigned long shinfo_mfn;
++ extern unsigned long max_pfn;
++ extern unsigned long *pfn_to_mfn_frame_list_list;
++ extern unsigned long *pfn_to_mfn_frame_list[];
+
-+ char *p = strrchr(dev->otherend, '/') + 1;
-+ long handle = simple_strtoul(p, NULL, 0);
++ if (suspend_cancelled) {
++ xen_start_info->store_mfn =
++ pfn_to_mfn(xen_start_info->store_mfn);
++ xen_start_info->console.domU.mfn =
++ pfn_to_mfn(xen_start_info->console.domU.mfn);
++ } else {
++#ifdef CONFIG_SMP
++ cpu_initialized_map = cpu_online_map;
++#endif
++ }
+
-+ be->major = major;
-+ be->minor = minor;
++ shinfo_mfn = xen_start_info->shared_info >> PAGE_SHIFT;
++ if (HYPERVISOR_update_va_mapping(fix_to_virt(FIX_SHARED_INFO),
++ pfn_pte_ma(shinfo_mfn, PAGE_KERNEL),
++ 0))
++ BUG();
++ HYPERVISOR_shared_info = (shared_info_t *)fix_to_virt(FIX_SHARED_INFO);
+
-+ err = vbd_create(be->blkif, handle, major, minor,
-+ (NULL == strchr(be->mode, 'w')));
-+ if (err) {
-+ be->major = be->minor = 0;
-+ xenbus_dev_fatal(dev, err, "creating vbd structure");
-+ return;
-+ }
++ memset(empty_zero_page, 0, PAGE_SIZE);
+
-+ err = xenvbd_sysfs_addif(dev);
-+ if (err) {
-+ vbd_free(&be->blkif->vbd);
-+ be->major = be->minor = 0;
-+ xenbus_dev_fatal(dev, err, "creating sysfs entries");
-+ return;
++ fpp = PAGE_SIZE/sizeof(unsigned long);
++ for (i = 0, j = 0, k = -1; i < max_pfn; i += fpp, j++) {
++ if ((j % fpp) == 0) {
++ k++;
++ pfn_to_mfn_frame_list_list[k] =
++ virt_to_mfn(pfn_to_mfn_frame_list[k]);
++ j = 0;
+ }
-+
-+ /* We're potentially connected now */
-+ update_blkif_status(be->blkif);
++ pfn_to_mfn_frame_list[k][j] =
++ virt_to_mfn(&phys_to_machine_mapping[i]);
+ }
++ HYPERVISOR_shared_info->arch.max_pfn = max_pfn;
++ HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list =
++ virt_to_mfn(pfn_to_mfn_frame_list_list);
+}
+
++#else /* !(defined(__i386__) || defined(__x86_64__)) */
+
-+/**
-+ * Callback received when the frontend's state changes.
-+ */
-+static void frontend_changed(struct xenbus_device *dev,
-+ enum xenbus_state frontend_state)
-+{
-+ struct backend_info *be = dev->dev.driver_data;
-+ int err;
-+
-+ DPRINTK("%s", xenbus_strstate(frontend_state));
-+
-+ switch (frontend_state) {
-+ case XenbusStateInitialising:
-+ if (dev->state == XenbusStateClosed) {
-+ printk(KERN_INFO "%s: %s: prepare for reconnect\n",
-+ __FUNCTION__, dev->nodename);
-+ xenbus_switch_state(dev, XenbusStateInitWait);
-+ }
-+ break;
++#ifndef HAVE_XEN_PRE_SUSPEND
++#define xen_pre_suspend() ((void)0)
++#endif
+
-+ case XenbusStateInitialised:
-+ case XenbusStateConnected:
-+ /* Ensure we connect even when two watches fire in
-+ close successsion and we miss the intermediate value
-+ of frontend_state. */
-+ if (dev->state == XenbusStateConnected)
-+ break;
++#ifndef HAVE_XEN_POST_SUSPEND
++#define xen_post_suspend(x) ((void)0)
++#endif
+
-+ err = connect_ring(be);
-+ if (err)
-+ break;
-+ update_blkif_status(be->blkif);
-+ break;
++#define switch_idle_mm() ((void)0)
++#define mm_pin_all() ((void)0)
++#define pre_suspend() xen_pre_suspend()
++#define post_suspend(x) xen_post_suspend(x)
+
-+ case XenbusStateClosing:
-+ blkif_disconnect(be->blkif);
-+ xenbus_switch_state(dev, XenbusStateClosing);
-+ break;
++#endif
+
-+ case XenbusStateClosed:
-+ xenbus_switch_state(dev, XenbusStateClosed);
-+ if (xenbus_dev_is_online(dev))
-+ break;
-+ /* fall through if not online */
-+ case XenbusStateUnknown:
-+ device_unregister(&dev->dev);
-+ break;
++struct suspend {
++ int fast_suspend;
++ void (*resume_notifier)(int);
++};
+
-+ default:
-+ xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
-+ frontend_state);
-+ break;
-+ }
-+}
++static int take_machine_down(void *_suspend)
++{
++ struct suspend *suspend = _suspend;
++ int suspend_cancelled, err;
++ extern void time_resume(void);
+
++ if (suspend->fast_suspend) {
++ BUG_ON(!irqs_disabled());
++ } else {
++ BUG_ON(irqs_disabled());
+
-+/* ** Connection ** */
++ for (;;) {
++ err = smp_suspend();
++ if (err)
++ return err;
+
++ xenbus_suspend();
++ preempt_disable();
+
-+/**
-+ * Write the physical details regarding the block device to the store, and
-+ * switch to Connected state.
-+ */
-+static void connect(struct backend_info *be)
-+{
-+ struct xenbus_transaction xbt;
-+ int err;
-+ struct xenbus_device *dev = be->dev;
++ if (num_online_cpus() == 1)
++ break;
+
-+ DPRINTK("%s", dev->otherend);
++ preempt_enable();
++ xenbus_suspend_cancel();
++ }
+
-+ /* Supply the information about the device the frontend needs */
-+again:
-+ err = xenbus_transaction_start(&xbt);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "starting transaction");
-+ return;
++ local_irq_disable();
+ }
+
-+ err = blkback_barrier(xbt, be, 1);
-+ if (err)
-+ goto abort;
++ mm_pin_all();
++ gnttab_suspend();
++ pre_suspend();
+
-+ err = xenbus_printf(xbt, dev->nodename, "sectors", "%llu",
-+ vbd_size(&be->blkif->vbd));
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "writing %s/sectors",
-+ dev->nodename);
-+ goto abort;
-+ }
++ /*
++ * This hypercall returns 1 if suspend was cancelled or the domain was
++ * merely checkpointed, and 0 if it is resuming in a new domain.
++ */
++ suspend_cancelled = HYPERVISOR_suspend(virt_to_mfn(xen_start_info));
+
-+ /* FIXME: use a typename instead */
-+ err = xenbus_printf(xbt, dev->nodename, "info", "%u",
-+ vbd_info(&be->blkif->vbd));
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "writing %s/info",
-+ dev->nodename);
-+ goto abort;
-+ }
-+ err = xenbus_printf(xbt, dev->nodename, "sector-size", "%lu",
-+ vbd_secsize(&be->blkif->vbd));
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "writing %s/sector-size",
-+ dev->nodename);
-+ goto abort;
++ suspend->resume_notifier(suspend_cancelled);
++ post_suspend(suspend_cancelled);
++ gnttab_resume();
++ if (!suspend_cancelled) {
++ irq_resume();
++#ifdef __x86_64__
++ /*
++ * Older versions of Xen do not save/restore the user %cr3.
++ * We do it here just in case, but there's no need if we are
++ * in fast-suspend mode as that implies a new enough Xen.
++ */
++ if (!suspend->fast_suspend)
++ xen_new_user_pt(__pa(__user_pgd(
++ current->active_mm->pgd)));
++#endif
+ }
++ time_resume();
+
-+ err = xenbus_transaction_end(xbt, 0);
-+ if (err == -EAGAIN)
-+ goto again;
-+ if (err)
-+ xenbus_dev_fatal(dev, err, "ending transaction");
-+
-+ err = xenbus_switch_state(dev, XenbusStateConnected);
-+ if (err)
-+ xenbus_dev_fatal(dev, err, "switching to Connected state",
-+ dev->nodename);
++ if (!suspend->fast_suspend)
++ local_irq_enable();
+
-+ return;
-+ abort:
-+ xenbus_transaction_end(xbt, 1);
++ return suspend_cancelled;
+}
+
-+
-+static int connect_ring(struct backend_info *be)
++int __xen_suspend(int fast_suspend, void (*resume_notifier)(int))
+{
-+ struct xenbus_device *dev = be->dev;
-+ unsigned long ring_ref;
-+ unsigned int evtchn;
-+ char protocol[64] = "";
-+ int err;
++ int err, suspend_cancelled;
++ struct suspend suspend;
+
-+ DPRINTK("%s", dev->otherend);
++ BUG_ON(smp_processor_id() != 0);
++ BUG_ON(in_interrupt());
+
-+ err = xenbus_gather(XBT_NIL, dev->otherend, "ring-ref", "%lu", &ring_ref,
-+ "event-channel", "%u", &evtchn, NULL);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err,
-+ "reading %s/ring-ref and event-channel",
-+ dev->otherend);
-+ return err;
++#if defined(__i386__) || defined(__x86_64__)
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ printk(KERN_WARNING "Cannot suspend in "
++ "auto_translated_physmap mode.\n");
++ return -EOPNOTSUPP;
+ }
++#endif
+
-+ be->blkif->blk_protocol = BLKIF_PROTOCOL_NATIVE;
-+ err = xenbus_gather(XBT_NIL, dev->otherend, "protocol",
-+ "%63s", protocol, NULL);
-+ if (err)
-+ strcpy(protocol, "unspecified, assuming native");
-+ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_NATIVE))
-+ be->blkif->blk_protocol = BLKIF_PROTOCOL_NATIVE;
-+ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_X86_32))
-+ be->blkif->blk_protocol = BLKIF_PROTOCOL_X86_32;
-+ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_X86_64))
-+ be->blkif->blk_protocol = BLKIF_PROTOCOL_X86_64;
-+ else {
-+ xenbus_dev_fatal(dev, err, "unknown fe protocol %s", protocol);
-+ return -1;
++ /* If we are definitely UP then 'slow mode' is actually faster. */
++ if (num_possible_cpus() == 1)
++ fast_suspend = 0;
++
++ suspend.fast_suspend = fast_suspend;
++ suspend.resume_notifier = resume_notifier;
++
++ if (fast_suspend) {
++ xenbus_suspend();
++ err = stop_machine_run(take_machine_down, &suspend, 0);
++ if (err < 0)
++ xenbus_suspend_cancel();
++ } else {
++ err = take_machine_down(&suspend);
+ }
-+ printk(KERN_INFO
-+ "blkback: ring-ref %ld, event-channel %d, protocol %d (%s)\n",
-+ ring_ref, evtchn, be->blkif->blk_protocol, protocol);
+
-+ /* Map the shared frame, irq etc. */
-+ err = blkif_map(be->blkif, ring_ref, evtchn);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "mapping ring-ref %lu port %u",
-+ ring_ref, evtchn);
++ if (err < 0)
+ return err;
++
++ suspend_cancelled = err;
++ if (!suspend_cancelled) {
++ xencons_resume();
++ xenbus_resume();
++ } else {
++ xenbus_suspend_cancel();
+ }
+
++ if (!fast_suspend)
++ smp_resume();
++
+ return 0;
+}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/pci.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/pci.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,67 @@
++/*
++ * vim:shiftwidth=8:noexpandtab
++ */
+
++#include <linux/kernel.h>
++#include <linux/init.h>
++#include <linux/pci.h>
++#include <xen/interface/physdev.h>
+
-+/* ** Driver Registration ** */
++static int (*pci_bus_probe)(struct device *dev);
++static int (*pci_bus_remove)(struct device *dev);
+
++static int pci_bus_probe_wrapper(struct device *dev)
++{
++ int r;
++ struct pci_dev *pci_dev = to_pci_dev(dev);
++ struct physdev_manage_pci manage_pci;
++ manage_pci.bus = pci_dev->bus->number;
++ manage_pci.devfn = pci_dev->devfn;
++
++ r = HYPERVISOR_physdev_op(PHYSDEVOP_manage_pci_add, &manage_pci);
++ if (r && r != -ENOSYS)
++ return r;
+
-+static struct xenbus_device_id blkback_ids[] = {
-+ { "vbd" },
-+ { "" }
-+};
++ r = pci_bus_probe(dev);
++ if (r) {
++ int ret;
+
++ ret = HYPERVISOR_physdev_op(PHYSDEVOP_manage_pci_remove,
++ &manage_pci);
++ WARN_ON(ret && ret != -ENOSYS);
++ }
+
-+static struct xenbus_driver blkback = {
-+ .name = "vbd",
-+ .owner = THIS_MODULE,
-+ .ids = blkback_ids,
-+ .probe = blkback_probe,
-+ .remove = blkback_remove,
-+ .otherend_changed = frontend_changed
-+};
++ return r;
++}
+
++static int pci_bus_remove_wrapper(struct device *dev)
++{
++ int r;
++ struct pci_dev *pci_dev = to_pci_dev(dev);
++ struct physdev_manage_pci manage_pci;
++ manage_pci.bus = pci_dev->bus->number;
++ manage_pci.devfn = pci_dev->devfn;
+
-+void blkif_xenbus_init(void)
++ r = pci_bus_remove(dev);
++ /* dev and pci_dev are no longer valid!! */
++
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_manage_pci_remove,
++ &manage_pci));
++ return r;
++}
++
++static int __init hook_pci_bus(void)
+{
-+ xenbus_register_backend(&blkback);
++ if (!is_running_on_xen() || !is_initial_xendomain())
++ return 0;
++
++ pci_bus_probe = pci_bus_type.probe;
++ pci_bus_type.probe = pci_bus_probe_wrapper;
++
++ pci_bus_remove = pci_bus_type.remove;
++ pci_bus_type.remove = pci_bus_remove_wrapper;
++
++ return 0;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blkfront/blkfront.c tmp-linux-2.6-xen.patch/drivers/xen/blkfront/blkfront.c
---- pristine-linux-2.6.18.2/drivers/xen/blkfront/blkfront.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blkfront/blkfront.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,912 @@
-+/******************************************************************************
-+ * blkfront.c
-+ *
-+ * XenLinux virtual block-device driver.
-+ *
-+ * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
-+ * Modifications by Mark A. Williamson are (c) Intel Research Cambridge
-+ * Copyright (c) 2004, Christian Limpach
-+ * Copyright (c) 2004, Andrew Warfield
-+ * Copyright (c) 2005, Christopher Clark
-+ * Copyright (c) 2005, XenSource Ltd
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
+
++core_initcall(hook_pci_bus);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/reboot.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/reboot.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,335 @@
++#define __KERNEL_SYSCALLS__
+#include <linux/version.h>
-+#include "block.h"
-+#include <linux/cdrom.h>
-+#include <linux/sched.h>
-+#include <linux/interrupt.h>
-+#include <scsi/scsi.h>
-+#include <xen/evtchn.h>
-+#include <xen/xenbus.h>
-+#include <xen/interface/grant_table.h>
-+#include <xen/interface/io/protocols.h>
-+#include <xen/gnttab.h>
++#include <linux/kernel.h>
++#include <linux/unistd.h>
++#include <linux/module.h>
++#include <linux/reboot.h>
++#include <linux/sysrq.h>
+#include <asm/hypervisor.h>
-+#include <asm/maddr.h>
++#include <xen/xenbus.h>
++#include <xen/evtchn.h>
++#include <linux/kmod.h>
++#include <linux/slab.h>
++#include <linux/workqueue.h>
+
+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
+#include <xen/platform-compat.h>
+#endif
+
-+#define BLKIF_STATE_DISCONNECTED 0
-+#define BLKIF_STATE_CONNECTED 1
-+#define BLKIF_STATE_SUSPENDED 2
++MODULE_LICENSE("Dual BSD/GPL");
+
-+#define MAXIMUM_OUTSTANDING_BLOCK_REQS \
-+ (BLKIF_MAX_SEGMENTS_PER_REQUEST * BLK_RING_SIZE)
-+#define GRANT_INVALID_REF 0
++#define SHUTDOWN_INVALID -1
++#define SHUTDOWN_POWEROFF 0
++#define SHUTDOWN_SUSPEND 2
++#define SHUTDOWN_RESUMING 3
++#define SHUTDOWN_HALT 4
+
-+static void connect(struct blkfront_info *);
-+static void blkfront_closing(struct xenbus_device *);
-+static int blkfront_remove(struct xenbus_device *);
-+static int talk_to_backend(struct xenbus_device *, struct blkfront_info *);
-+static int setup_blkring(struct xenbus_device *, struct blkfront_info *);
++/* Ignore multiple shutdown requests. */
++static int shutting_down = SHUTDOWN_INVALID;
+
-+static void kick_pending_request_queues(struct blkfront_info *);
++/* Was last suspend request cancelled? */
++static int suspend_cancelled;
+
-+static irqreturn_t blkif_int(int irq, void *dev_id, struct pt_regs *ptregs);
-+static void blkif_restart_queue(void *arg);
-+static void blkif_recover(struct blkfront_info *);
-+static void blkif_completion(struct blk_shadow *);
-+static void blkif_free(struct blkfront_info *, int);
++/* Can we leave APs online when we suspend? */
++static int fast_suspend;
+
++static void __shutdown_handler(void *unused);
++static DECLARE_WORK(shutdown_work, __shutdown_handler, NULL);
+
-+/**
-+ * Entry point to this code when a new device is created. Allocate the basic
-+ * structures and the ring buffer for communication with the backend, and
-+ * inform the backend of the appropriate details for those. Switch to
-+ * Initialised state.
-+ */
-+static int blkfront_probe(struct xenbus_device *dev,
-+ const struct xenbus_device_id *id)
++static int setup_suspend_evtchn(void);
++
++int __xen_suspend(int fast_suspend, void (*resume_notifier)(int));
++
++static int shutdown_process(void *__unused)
+{
-+ int err, vdevice, i;
-+ struct blkfront_info *info;
++ static char *envp[] = { "HOME=/", "TERM=linux",
++ "PATH=/sbin:/usr/sbin:/bin:/usr/bin", NULL };
++ static char *poweroff_argv[] = { "/sbin/poweroff", NULL };
+
-+ /* FIXME: Use dynamic device id if this is not set. */
-+ err = xenbus_scanf(XBT_NIL, dev->nodename,
-+ "virtual-device", "%i", &vdevice);
-+ if (err != 1) {
-+ xenbus_dev_fatal(dev, err, "reading virtual-device");
-+ return err;
-+ }
++ extern asmlinkage long sys_reboot(int magic1, int magic2,
++ unsigned int cmd, void *arg);
+
-+ info = kzalloc(sizeof(*info), GFP_KERNEL);
-+ if (!info) {
-+ xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
-+ return -ENOMEM;
++ if ((shutting_down == SHUTDOWN_POWEROFF) ||
++ (shutting_down == SHUTDOWN_HALT)) {
++ if (call_usermodehelper("/sbin/poweroff", poweroff_argv,
++ envp, 0) < 0) {
++#ifdef CONFIG_XEN
++ sys_reboot(LINUX_REBOOT_MAGIC1,
++ LINUX_REBOOT_MAGIC2,
++ LINUX_REBOOT_CMD_POWER_OFF,
++ NULL);
++#endif /* CONFIG_XEN */
++ }
+ }
+
-+ info->xbdev = dev;
-+ info->vdevice = vdevice;
-+ info->connected = BLKIF_STATE_DISCONNECTED;
-+ INIT_WORK(&info->work, blkif_restart_queue, (void *)info);
++ shutting_down = SHUTDOWN_INVALID; /* could try again */
+
-+ for (i = 0; i < BLK_RING_SIZE; i++)
-+ info->shadow[i].req.id = i+1;
-+ info->shadow[BLK_RING_SIZE-1].req.id = 0x0fffffff;
++ return 0;
++}
+
-+ /* Front end dir is a number, which is used as the id. */
-+ info->handle = simple_strtoul(strrchr(dev->nodename,'/')+1, NULL, 0);
-+ dev->dev.driver_data = info;
++static void xen_resume_notifier(int _suspend_cancelled)
++{
++ int old_state = xchg(&shutting_down, SHUTDOWN_RESUMING);
++ BUG_ON(old_state != SHUTDOWN_SUSPEND);
++ suspend_cancelled = _suspend_cancelled;
++}
+
-+ err = talk_to_backend(dev, info);
++static int xen_suspend(void *__unused)
++{
++ int err, old_state;
++
++ daemonize("suspend");
++ err = set_cpus_allowed(current, cpumask_of_cpu(0));
+ if (err) {
-+ kfree(info);
-+ dev->dev.driver_data = NULL;
-+ return err;
++ printk(KERN_ERR "Xen suspend can't run on CPU0 (%d)\n", err);
++ goto fail;
++ }
++
++ do {
++ err = __xen_suspend(fast_suspend, xen_resume_notifier);
++ if (err) {
++ printk(KERN_ERR "Xen suspend failed (%d)\n", err);
++ goto fail;
++ }
++ if (!suspend_cancelled)
++ setup_suspend_evtchn();
++ old_state = cmpxchg(
++ &shutting_down, SHUTDOWN_RESUMING, SHUTDOWN_INVALID);
++ } while (old_state == SHUTDOWN_SUSPEND);
++
++ switch (old_state) {
++ case SHUTDOWN_INVALID:
++ case SHUTDOWN_SUSPEND:
++ BUG();
++ case SHUTDOWN_RESUMING:
++ break;
++ default:
++ schedule_work(&shutdown_work);
++ break;
+ }
+
+ return 0;
-+}
+
++ fail:
++ old_state = xchg(&shutting_down, SHUTDOWN_INVALID);
++ BUG_ON(old_state != SHUTDOWN_SUSPEND);
++ return 0;
++}
+
-+/**
-+ * We are reconnecting to the backend, due to a suspend/resume, or a backend
-+ * driver restart. We tear down our blkif structure and recreate it, but
-+ * leave the device-layer structures intact so that this is transparent to the
-+ * rest of the kernel.
-+ */
-+static int blkfront_resume(struct xenbus_device *dev)
++static void switch_shutdown_state(int new_state)
+{
-+ struct blkfront_info *info = dev->dev.driver_data;
-+ int err;
++ int prev_state, old_state = SHUTDOWN_INVALID;
+
-+ DPRINTK("blkfront_resume: %s\n", dev->nodename);
-+
-+ blkif_free(info, info->connected == BLKIF_STATE_CONNECTED);
++ /* We only drive shutdown_state into an active state. */
++ if (new_state == SHUTDOWN_INVALID)
++ return;
+
-+ err = talk_to_backend(dev, info);
-+ if (info->connected == BLKIF_STATE_SUSPENDED && !err)
-+ blkif_recover(info);
++ do {
++ /* We drop this transition if already in an active state. */
++ if ((old_state != SHUTDOWN_INVALID) &&
++ (old_state != SHUTDOWN_RESUMING))
++ return;
++ /* Attempt to transition. */
++ prev_state = old_state;
++ old_state = cmpxchg(&shutting_down, old_state, new_state);
++ } while (old_state != prev_state);
+
-+ return err;
++ /* Either we kick off the work, or we leave it to xen_suspend(). */
++ if (old_state == SHUTDOWN_INVALID)
++ schedule_work(&shutdown_work);
++ else
++ BUG_ON(old_state != SHUTDOWN_RESUMING);
+}
+
++static void __shutdown_handler(void *unused)
++{
++ int err;
+
-+/* Common code used when first setting up, and when resuming. */
-+static int talk_to_backend(struct xenbus_device *dev,
-+ struct blkfront_info *info)
++ err = kernel_thread((shutting_down == SHUTDOWN_SUSPEND) ?
++ xen_suspend : shutdown_process,
++ NULL, CLONE_FS | CLONE_FILES);
++
++ if (err < 0) {
++ printk(KERN_WARNING "Error creating shutdown process (%d): "
++ "retrying...\n", -err);
++ schedule_delayed_work(&shutdown_work, HZ/2);
++ }
++}
++
++static void shutdown_handler(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
+{
-+ const char *message = NULL;
++ extern void ctrl_alt_del(void);
++ char *str;
+ struct xenbus_transaction xbt;
-+ int err;
++ int err, new_state = SHUTDOWN_INVALID;
+
-+ /* Create shared ring, alloc event channel. */
-+ err = setup_blkring(dev, info);
-+ if (err)
-+ goto out;
++ if ((shutting_down != SHUTDOWN_INVALID) &&
++ (shutting_down != SHUTDOWN_RESUMING))
++ return;
+
-+again:
++ again:
+ err = xenbus_transaction_start(&xbt);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "starting transaction");
-+ goto destroy_blkring;
-+ }
++ if (err)
++ return;
+
-+ err = xenbus_printf(xbt, dev->nodename,
-+ "ring-ref","%u", info->ring_ref);
-+ if (err) {
-+ message = "writing ring-ref";
-+ goto abort_transaction;
++ str = (char *)xenbus_read(xbt, "control", "shutdown", NULL);
++ /* Ignore read errors and empty reads. */
++ if (XENBUS_IS_ERR_READ(str)) {
++ xenbus_transaction_end(xbt, 1);
++ return;
+ }
-+ err = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
-+ irq_to_evtchn_port(info->irq));
-+ if (err) {
-+ message = "writing event-channel";
-+ goto abort_transaction;
++
++ xenbus_write(xbt, "control", "shutdown", "");
++
++ err = xenbus_transaction_end(xbt, 0);
++ if (err == -EAGAIN) {
++ kfree(str);
++ goto again;
+ }
-+ err = xenbus_printf(xbt, dev->nodename, "protocol", "%s",
-+ XEN_IO_PROTO_ABI_NATIVE);
-+ if (err) {
-+ message = "writing protocol";
-+ goto abort_transaction;
++
++ if (strcmp(str, "poweroff") == 0)
++ new_state = SHUTDOWN_POWEROFF;
++ else if (strcmp(str, "reboot") == 0)
++ ctrl_alt_del();
++ else if (strcmp(str, "suspend") == 0)
++ new_state = SHUTDOWN_SUSPEND;
++ else if (strcmp(str, "halt") == 0)
++ new_state = SHUTDOWN_HALT;
++ else
++ printk("Ignoring shutdown request: %s\n", str);
++
++ switch_shutdown_state(new_state);
++
++ kfree(str);
++}
++
++static void sysrq_handler(struct xenbus_watch *watch, const char **vec,
++ unsigned int len)
++{
++ char sysrq_key = '\0';
++ struct xenbus_transaction xbt;
++ int err;
++
++ again:
++ err = xenbus_transaction_start(&xbt);
++ if (err)
++ return;
++ if (!xenbus_scanf(xbt, "control", "sysrq", "%c", &sysrq_key)) {
++ printk(KERN_ERR "Unable to read sysrq code in "
++ "control/sysrq\n");
++ xenbus_transaction_end(xbt, 1);
++ return;
+ }
+
++ if (sysrq_key != '\0')
++ xenbus_printf(xbt, "control", "sysrq", "%c", '\0');
++
+ err = xenbus_transaction_end(xbt, 0);
-+ if (err) {
-+ if (err == -EAGAIN)
-+ goto again;
-+ xenbus_dev_fatal(dev, err, "completing transaction");
-+ goto destroy_blkring;
-+ }
++ if (err == -EAGAIN)
++ goto again;
+
-+ xenbus_switch_state(dev, XenbusStateInitialised);
++#ifdef CONFIG_MAGIC_SYSRQ
++ if (sysrq_key != '\0')
++ handle_sysrq(sysrq_key, NULL, NULL);
++#endif
++}
+
-+ return 0;
++static struct xenbus_watch shutdown_watch = {
++ .node = "control/shutdown",
++ .callback = shutdown_handler
++};
+
-+ abort_transaction:
-+ xenbus_transaction_end(xbt, 1);
-+ if (message)
-+ xenbus_dev_fatal(dev, err, "%s", message);
-+ destroy_blkring:
-+ blkif_free(info, 0);
-+ out:
-+ return err;
++static struct xenbus_watch sysrq_watch = {
++ .node = "control/sysrq",
++ .callback = sysrq_handler
++};
++
++static irqreturn_t suspend_int(int irq, void* dev_id, struct pt_regs *ptregs)
++{
++ switch_shutdown_state(SHUTDOWN_SUSPEND);
++ return IRQ_HANDLED;
+}
+
++static int setup_suspend_evtchn(void)
++{
++ static int irq;
++ int port;
++ char portstr[16];
+
-+static int setup_blkring(struct xenbus_device *dev,
-+ struct blkfront_info *info)
++ if (irq > 0)
++ unbind_from_irqhandler(irq, NULL);
++
++ irq = bind_listening_port_to_irqhandler(0, suspend_int, 0, "suspend",
++ NULL);
++ if (irq <= 0)
++ return -1;
++
++ port = irq_to_evtchn_port(irq);
++ printk(KERN_INFO "suspend: event channel %d\n", port);
++ sprintf(portstr, "%d", port);
++ xenbus_write(XBT_NIL, "device/suspend", "event-channel", portstr);
++
++ return 0;
++}
++
++static int setup_shutdown_watcher(void)
+{
-+ blkif_sring_t *sring;
+ int err;
+
-+ info->ring_ref = GRANT_INVALID_REF;
++ xenbus_scanf(XBT_NIL, "control",
++ "platform-feature-multiprocessor-suspend",
++ "%d", &fast_suspend);
+
-+ sring = (blkif_sring_t *)__get_free_page(GFP_KERNEL);
-+ if (!sring) {
-+ xenbus_dev_fatal(dev, -ENOMEM, "allocating shared ring");
-+ return -ENOMEM;
++ err = register_xenbus_watch(&shutdown_watch);
++ if (err) {
++ printk(KERN_ERR "Failed to set shutdown watcher\n");
++ return err;
+ }
-+ SHARED_RING_INIT(sring);
-+ FRONT_RING_INIT(&info->ring, sring, PAGE_SIZE);
+
-+ err = xenbus_grant_ring(dev, virt_to_mfn(info->ring.sring));
-+ if (err < 0) {
-+ free_page((unsigned long)sring);
-+ info->ring.sring = NULL;
-+ goto fail;
++ err = register_xenbus_watch(&sysrq_watch);
++ if (err) {
++ printk(KERN_ERR "Failed to set sysrq watcher\n");
++ return err;
+ }
-+ info->ring_ref = err;
+
-+ err = bind_listening_port_to_irqhandler(
-+ dev->otherend_id, blkif_int, SA_SAMPLE_RANDOM, "blkif", info);
-+ if (err <= 0) {
-+ xenbus_dev_fatal(dev, err,
-+ "bind_listening_port_to_irqhandler");
-+ goto fail;
++ /* suspend event channel */
++ err = setup_suspend_evtchn();
++ if (err) {
++ printk(KERN_ERR "Failed to register suspend event channel\n");
++ return err;
+ }
-+ info->irq = err;
+
+ return 0;
-+fail:
-+ blkif_free(info, 0);
-+ return err;
+}
+
++#ifdef CONFIG_XEN
+
-+/**
-+ * Callback received when the backend's state changes.
-+ */
-+static void backend_changed(struct xenbus_device *dev,
-+ enum xenbus_state backend_state)
++static int shutdown_event(struct notifier_block *notifier,
++ unsigned long event,
++ void *data)
+{
-+ struct blkfront_info *info = dev->dev.driver_data;
-+ struct block_device *bd;
++ setup_shutdown_watcher();
++ return NOTIFY_DONE;
++}
+
-+ DPRINTK("blkfront:backend_changed.\n");
++static int __init setup_shutdown_event(void)
++{
++ static struct notifier_block xenstore_notifier = {
++ .notifier_call = shutdown_event
++ };
++ register_xenstore_notifier(&xenstore_notifier);
+
-+ switch (backend_state) {
-+ case XenbusStateInitialising:
-+ case XenbusStateInitWait:
-+ case XenbusStateInitialised:
-+ case XenbusStateUnknown:
-+ case XenbusStateClosed:
-+ break;
++ return 0;
++}
+
-+ case XenbusStateConnected:
-+ connect(info);
-+ break;
++subsys_initcall(setup_shutdown_event);
+
-+ case XenbusStateClosing:
-+ bd = bdget(info->dev);
-+ if (bd == NULL)
-+ xenbus_dev_fatal(dev, -ENODEV, "bdget failed");
++#else /* !defined(CONFIG_XEN) */
+
-+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-+ down(&bd->bd_sem);
-+#else
-+ mutex_lock(&bd->bd_mutex);
-+#endif
-+ if (info->users > 0)
-+ xenbus_dev_error(dev, -EBUSY,
-+ "Device in use; refusing to close");
-+ else
-+ blkfront_closing(dev);
-+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-+ up(&bd->bd_sem);
-+#else
-+ mutex_unlock(&bd->bd_mutex);
-+#endif
-+ bdput(bd);
-+ break;
-+ }
++int xen_reboot_init(void)
++{
++ return setup_shutdown_watcher();
+}
+
++#endif /* !defined(CONFIG_XEN) */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/smpboot.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/smpboot.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,464 @@
++/*
++ * Xen SMP booting functions
++ *
++ * See arch/i386/kernel/smpboot.c for copyright and credits for derived
++ * portions of this file.
++ */
++
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/kernel.h>
++#include <linux/mm.h>
++#include <linux/sched.h>
++#include <linux/kernel_stat.h>
++#include <linux/smp_lock.h>
++#include <linux/irq.h>
++#include <linux/bootmem.h>
++#include <linux/notifier.h>
++#include <linux/cpu.h>
++#include <linux/percpu.h>
++#include <asm/desc.h>
++#include <asm/arch_hooks.h>
++#include <asm/pgalloc.h>
++#include <xen/evtchn.h>
++#include <xen/interface/vcpu.h>
++#include <xen/cpu_hotplug.h>
++#include <xen/xenbus.h>
++
++extern irqreturn_t smp_reschedule_interrupt(int, void *, struct pt_regs *);
++extern irqreturn_t smp_call_function_interrupt(int, void *, struct pt_regs *);
+
-+/* ** Connection ** */
++extern int local_setup_timer(unsigned int cpu);
++extern void local_teardown_timer(unsigned int cpu);
+
++extern void hypervisor_callback(void);
++extern void failsafe_callback(void);
++extern void system_call(void);
++extern void smp_trap_init(trap_info_t *);
+
-+/*
-+ * Invoked when the backend is finally 'ready' (and has told produced
-+ * the details about the physical device - #sectors, size, etc).
-+ */
-+static void connect(struct blkfront_info *info)
-+{
-+ unsigned long long sectors;
-+ unsigned long sector_size;
-+ unsigned int binfo;
-+ int err;
++/* Number of siblings per CPU package */
++int smp_num_siblings = 1;
+
-+ if ((info->connected == BLKIF_STATE_CONNECTED) ||
-+ (info->connected == BLKIF_STATE_SUSPENDED) )
-+ return;
++cpumask_t cpu_online_map;
++EXPORT_SYMBOL(cpu_online_map);
++cpumask_t cpu_possible_map;
++EXPORT_SYMBOL(cpu_possible_map);
++cpumask_t cpu_initialized_map;
+
-+ DPRINTK("blkfront.c:connect:%s.\n", info->xbdev->otherend);
++struct cpuinfo_x86 cpu_data[NR_CPUS] __cacheline_aligned;
++EXPORT_SYMBOL(cpu_data);
+
-+ err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
-+ "sectors", "%Lu", §ors,
-+ "info", "%u", &binfo,
-+ "sector-size", "%lu", §or_size,
-+ NULL);
-+ if (err) {
-+ xenbus_dev_fatal(info->xbdev, err,
-+ "reading backend fields at %s",
-+ info->xbdev->otherend);
-+ return;
-+ }
++#ifdef CONFIG_HOTPLUG_CPU
++DEFINE_PER_CPU(int, cpu_state) = { 0 };
++#endif
+
-+ err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
-+ "feature-barrier", "%lu", &info->feature_barrier,
-+ NULL);
-+ if (err)
-+ info->feature_barrier = 0;
++static DEFINE_PER_CPU(int, resched_irq);
++static DEFINE_PER_CPU(int, callfunc_irq);
++static char resched_name[NR_CPUS][15];
++static char callfunc_name[NR_CPUS][15];
+
-+ err = xlvbd_add(sectors, info->vdevice, binfo, sector_size, info);
-+ if (err) {
-+ xenbus_dev_fatal(info->xbdev, err, "xlvbd_add at %s",
-+ info->xbdev->otherend);
-+ return;
-+ }
++u8 cpu_2_logical_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = BAD_APICID };
+
-+ (void)xenbus_switch_state(info->xbdev, XenbusStateConnected);
++cpumask_t cpu_sibling_map[NR_CPUS] __cacheline_aligned;
++cpumask_t cpu_core_map[NR_CPUS] __cacheline_aligned;
++EXPORT_SYMBOL(cpu_core_map);
+
-+ /* Kick pending requests. */
-+ spin_lock_irq(&blkif_io_lock);
-+ info->connected = BLKIF_STATE_CONNECTED;
-+ kick_pending_request_queues(info);
-+ spin_unlock_irq(&blkif_io_lock);
++#if defined(__i386__)
++u8 x86_cpu_to_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = 0xff };
++EXPORT_SYMBOL(x86_cpu_to_apicid);
++#elif !defined(CONFIG_X86_IO_APIC)
++unsigned int maxcpus = NR_CPUS;
++#endif
+
-+ add_disk(info->gd);
++void __init prefill_possible_map(void)
++{
++ int i, rc;
+
-+ info->is_ready = 1;
++ for_each_possible_cpu(i)
++ if (i != smp_processor_id())
++ return;
++
++ for (i = 0; i < NR_CPUS; i++) {
++ rc = HYPERVISOR_vcpu_op(VCPUOP_is_up, i, NULL);
++ if (rc >= 0)
++ cpu_set(i, cpu_possible_map);
++ }
+}
+
-+/**
-+ * Handle the change of state of the backend to Closing. We must delete our
-+ * device-layer structures now, to ensure that writes are flushed through to
-+ * the backend. Once is this done, we can switch to Closed in
-+ * acknowledgement.
-+ */
-+static void blkfront_closing(struct xenbus_device *dev)
++void __init smp_alloc_memory(void)
+{
-+ struct blkfront_info *info = dev->dev.driver_data;
-+ unsigned long flags;
++}
+
-+ DPRINTK("blkfront_closing: %s removed\n", dev->nodename);
++static inline void
++set_cpu_sibling_map(unsigned int cpu)
++{
++ cpu_data[cpu].phys_proc_id = cpu;
++ cpu_data[cpu].cpu_core_id = 0;
+
-+ if (info->rq == NULL)
-+ goto out;
++ cpu_sibling_map[cpu] = cpumask_of_cpu(cpu);
++ cpu_core_map[cpu] = cpumask_of_cpu(cpu);
+
-+ spin_lock_irqsave(&blkif_io_lock, flags);
-+ /* No more blkif_request(). */
-+ blk_stop_queue(info->rq);
-+ /* No more gnttab callback work. */
-+ gnttab_cancel_free_callback(&info->callback);
-+ spin_unlock_irqrestore(&blkif_io_lock, flags);
++ cpu_data[cpu].booted_cores = 1;
++}
+
-+ /* Flush gnttab callback work. Must be done with no locks held. */
-+ flush_scheduled_work();
++static void
++remove_siblinginfo(unsigned int cpu)
++{
++ cpu_data[cpu].phys_proc_id = BAD_APICID;
++ cpu_data[cpu].cpu_core_id = BAD_APICID;
+
-+ xlvbd_del(info);
++ cpus_clear(cpu_sibling_map[cpu]);
++ cpus_clear(cpu_core_map[cpu]);
+
-+ out:
-+ xenbus_frontend_closed(dev);
++ cpu_data[cpu].booted_cores = 0;
+}
+
-+
-+static int blkfront_remove(struct xenbus_device *dev)
++static int __cpuinit xen_smp_intr_init(unsigned int cpu)
+{
-+ struct blkfront_info *info = dev->dev.driver_data;
++ int rc;
+
-+ DPRINTK("blkfront_remove: %s removed\n", dev->nodename);
++ per_cpu(resched_irq, cpu) = per_cpu(callfunc_irq, cpu) = -1;
+
-+ blkif_free(info, 0);
++ sprintf(resched_name[cpu], "resched%u", cpu);
++ rc = bind_ipi_to_irqhandler(RESCHEDULE_VECTOR,
++ cpu,
++ smp_reschedule_interrupt,
++ SA_INTERRUPT,
++ resched_name[cpu],
++ NULL);
++ if (rc < 0)
++ goto fail;
++ per_cpu(resched_irq, cpu) = rc;
+
-+ kfree(info);
++ sprintf(callfunc_name[cpu], "callfunc%u", cpu);
++ rc = bind_ipi_to_irqhandler(CALL_FUNCTION_VECTOR,
++ cpu,
++ smp_call_function_interrupt,
++ SA_INTERRUPT,
++ callfunc_name[cpu],
++ NULL);
++ if (rc < 0)
++ goto fail;
++ per_cpu(callfunc_irq, cpu) = rc;
++
++ if ((cpu != 0) && ((rc = local_setup_timer(cpu)) != 0))
++ goto fail;
+
+ return 0;
++
++ fail:
++ if (per_cpu(resched_irq, cpu) >= 0)
++ unbind_from_irqhandler(per_cpu(resched_irq, cpu), NULL);
++ if (per_cpu(callfunc_irq, cpu) >= 0)
++ unbind_from_irqhandler(per_cpu(callfunc_irq, cpu), NULL);
++ return rc;
+}
+
++#ifdef CONFIG_HOTPLUG_CPU
++static void xen_smp_intr_exit(unsigned int cpu)
++{
++ if (cpu != 0)
++ local_teardown_timer(cpu);
+
-+static inline int GET_ID_FROM_FREELIST(
-+ struct blkfront_info *info)
++ unbind_from_irqhandler(per_cpu(resched_irq, cpu), NULL);
++ unbind_from_irqhandler(per_cpu(callfunc_irq, cpu), NULL);
++}
++#endif
++
++void __cpuinit cpu_bringup(void)
+{
-+ unsigned long free = info->shadow_free;
-+ BUG_ON(free > BLK_RING_SIZE);
-+ info->shadow_free = info->shadow[free].req.id;
-+ info->shadow[free].req.id = 0x0fffffee; /* debug */
-+ return free;
++ cpu_init();
++ identify_cpu(cpu_data + smp_processor_id());
++ touch_softlockup_watchdog();
++ preempt_disable();
++ local_irq_enable();
+}
+
-+static inline void ADD_ID_TO_FREELIST(
-+ struct blkfront_info *info, unsigned long id)
++static void __cpuinit cpu_bringup_and_idle(void)
+{
-+ info->shadow[id].req.id = info->shadow_free;
-+ info->shadow[id].request = 0;
-+ info->shadow_free = id;
++ cpu_bringup();
++ cpu_idle();
+}
+
-+static inline void flush_requests(struct blkfront_info *info)
++static void __cpuinit cpu_initialize_context(unsigned int cpu)
+{
-+ int notify;
++ /* vcpu_guest_context_t is too large to allocate on the stack.
++ * Hence we allocate statically and protect it with a lock */
++ static vcpu_guest_context_t ctxt;
++ static DEFINE_SPINLOCK(ctxt_lock);
+
-+ RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&info->ring, notify);
++ struct task_struct *idle = idle_task(cpu);
++#ifdef __x86_64__
++ struct desc_ptr *gdt_descr = &cpu_gdt_descr[cpu];
++#else
++ struct Xgt_desc_struct *gdt_descr = &per_cpu(cpu_gdt_descr, cpu);
++#endif
+
-+ if (notify)
-+ notify_remote_via_irq(info->irq);
-+}
++ if (cpu_test_and_set(cpu, cpu_initialized_map))
++ return;
+
-+static void kick_pending_request_queues(struct blkfront_info *info)
-+{
-+ if (!RING_FULL(&info->ring)) {
-+ /* Re-enable calldowns. */
-+ blk_start_queue(info->rq);
-+ /* Kick things off immediately. */
-+ do_blkif_request(info->rq);
-+ }
-+}
++ spin_lock(&ctxt_lock);
+
-+static void blkif_restart_queue(void *arg)
-+{
-+ struct blkfront_info *info = (struct blkfront_info *)arg;
-+ spin_lock_irq(&blkif_io_lock);
-+ if (info->connected == BLKIF_STATE_CONNECTED)
-+ kick_pending_request_queues(info);
-+ spin_unlock_irq(&blkif_io_lock);
-+}
++ memset(&ctxt, 0, sizeof(ctxt));
+
-+static void blkif_restart_queue_callback(void *arg)
-+{
-+ struct blkfront_info *info = (struct blkfront_info *)arg;
-+ schedule_work(&info->work);
-+}
++ ctxt.flags = VGCF_IN_KERNEL;
++ ctxt.user_regs.ds = __USER_DS;
++ ctxt.user_regs.es = __USER_DS;
++ ctxt.user_regs.fs = 0;
++ ctxt.user_regs.gs = 0;
++ ctxt.user_regs.ss = __KERNEL_DS;
++ ctxt.user_regs.eip = (unsigned long)cpu_bringup_and_idle;
++ ctxt.user_regs.eflags = X86_EFLAGS_IF | 0x1000; /* IOPL_RING1 */
+
-+int blkif_open(struct inode *inode, struct file *filep)
-+{
-+ struct blkfront_info *info = inode->i_bdev->bd_disk->private_data;
-+ info->users++;
-+ return 0;
-+}
++ memset(&ctxt.fpu_ctxt, 0, sizeof(ctxt.fpu_ctxt));
+
++ smp_trap_init(ctxt.trap_ctxt);
+
-+int blkif_release(struct inode *inode, struct file *filep)
-+{
-+ struct blkfront_info *info = inode->i_bdev->bd_disk->private_data;
-+ info->users--;
-+ if (info->users == 0) {
-+ /* Check whether we have been instructed to close. We will
-+ have ignored this request initially, as the device was
-+ still mounted. */
-+ struct xenbus_device * dev = info->xbdev;
-+ enum xenbus_state state = xenbus_read_driver_state(dev->otherend);
++ ctxt.ldt_ents = 0;
+
-+ if (state == XenbusStateClosing)
-+ blkfront_closing(dev);
-+ }
-+ return 0;
-+}
++ ctxt.gdt_frames[0] = virt_to_mfn(gdt_descr->address);
++ ctxt.gdt_ents = gdt_descr->size / 8;
+
++#ifdef __i386__
++ ctxt.user_regs.cs = __KERNEL_CS;
++ ctxt.user_regs.esp = idle->thread.esp0 - sizeof(struct pt_regs);
+
-+int blkif_ioctl(struct inode *inode, struct file *filep,
-+ unsigned command, unsigned long argument)
-+{
-+ int i;
++ ctxt.kernel_ss = __KERNEL_DS;
++ ctxt.kernel_sp = idle->thread.esp0;
+
-+ DPRINTK_IOCTL("command: 0x%x, argument: 0x%lx, dev: 0x%04x\n",
-+ command, (long)argument, inode->i_rdev);
++ ctxt.event_callback_cs = __KERNEL_CS;
++ ctxt.event_callback_eip = (unsigned long)hypervisor_callback;
++ ctxt.failsafe_callback_cs = __KERNEL_CS;
++ ctxt.failsafe_callback_eip = (unsigned long)failsafe_callback;
+
-+ switch (command) {
-+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,16)
-+ case HDIO_GETGEO: {
-+ struct block_device *bd = inode->i_bdev;
-+ struct hd_geometry geo;
-+ int ret;
++ ctxt.ctrlreg[3] = xen_pfn_to_cr3(virt_to_mfn(swapper_pg_dir));
++#else /* __x86_64__ */
++ ctxt.user_regs.cs = __KERNEL_CS;
++ ctxt.user_regs.esp = idle->thread.rsp0 - sizeof(struct pt_regs);
+
-+ if (!argument)
-+ return -EINVAL;
++ ctxt.kernel_ss = __KERNEL_DS;
++ ctxt.kernel_sp = idle->thread.rsp0;
+
-+ geo.start = get_start_sect(bd);
-+ ret = blkif_getgeo(bd, &geo);
-+ if (ret)
-+ return ret;
++ ctxt.event_callback_eip = (unsigned long)hypervisor_callback;
++ ctxt.failsafe_callback_eip = (unsigned long)failsafe_callback;
++ ctxt.syscall_callback_eip = (unsigned long)system_call;
+
-+ if (copy_to_user((struct hd_geometry __user *)argument, &geo,
-+ sizeof(geo)))
-+ return -EFAULT;
++ ctxt.ctrlreg[3] = xen_pfn_to_cr3(virt_to_mfn(init_level4_pgt));
+
-+ return 0;
-+ }
++ ctxt.gs_base_kernel = (unsigned long)(cpu_pda(cpu));
+#endif
-+ case CDROMMULTISESSION:
-+ DPRINTK("FIXME: support multisession CDs later\n");
-+ for (i = 0; i < sizeof(struct cdrom_multisession); i++)
-+ if (put_user(0, (char __user *)(argument + i)))
-+ return -EFAULT;
-+ return 0;
+
-+ default:
-+ /*printk(KERN_ALERT "ioctl %08x not supported by Xen blkdev\n",
-+ command);*/
-+ return -EINVAL; /* same return as native Linux */
-+ }
++ if (HYPERVISOR_vcpu_op(VCPUOP_initialise, cpu, &ctxt))
++ BUG();
+
-+ return 0;
++ spin_unlock(&ctxt_lock);
+}
+
-+
-+int blkif_getgeo(struct block_device *bd, struct hd_geometry *hg)
++void __init smp_prepare_cpus(unsigned int max_cpus)
+{
-+ /* We don't have real geometry info, but let's at least return
-+ values consistent with the size of the device */
-+ sector_t nsect = get_capacity(bd->bd_disk);
-+ sector_t cylinders = nsect;
++ unsigned int cpu;
++ struct task_struct *idle;
++ int apicid, acpiid;
++ struct vcpu_get_physid cpu_id;
++#ifdef __x86_64__
++ struct desc_ptr *gdt_descr;
++#else
++ struct Xgt_desc_struct *gdt_descr;
++#endif
+
-+ hg->heads = 0xff;
-+ hg->sectors = 0x3f;
-+ sector_div(cylinders, hg->heads * hg->sectors);
-+ hg->cylinders = cylinders;
-+ if ((sector_t)(hg->cylinders + 1) * hg->heads * hg->sectors < nsect)
-+ hg->cylinders = 0xffff;
-+ return 0;
-+}
++ apicid = 0;
++ if (HYPERVISOR_vcpu_op(VCPUOP_get_physid, 0, &cpu_id) == 0) {
++ apicid = xen_vcpu_physid_to_x86_apicid(cpu_id.phys_id);
++ acpiid = xen_vcpu_physid_to_x86_acpiid(cpu_id.phys_id);
++#ifdef CONFIG_ACPI
++ if (acpiid != 0xff)
++ x86_acpiid_to_apicid[acpiid] = apicid;
++#endif
++ }
++ boot_cpu_data.apicid = apicid;
++ cpu_data[0] = boot_cpu_data;
+
++ cpu_2_logical_apicid[0] = apicid;
++ x86_cpu_to_apicid[0] = apicid;
+
-+/*
-+ * blkif_queue_request
-+ *
-+ * request block io
-+ *
-+ * id: for guest use only.
-+ * operation: BLKIF_OP_{READ,WRITE,PROBE}
-+ * buffer: buffer to read/write into. this should be a
-+ * virtual address in the guest os.
-+ */
-+static int blkif_queue_request(struct request *req)
-+{
-+ struct blkfront_info *info = req->rq_disk->private_data;
-+ unsigned long buffer_mfn;
-+ blkif_request_t *ring_req;
-+ struct bio *bio;
-+ struct bio_vec *bvec;
-+ int idx;
-+ unsigned long id;
-+ unsigned int fsect, lsect;
-+ int ref;
-+ grant_ref_t gref_head;
++ current_thread_info()->cpu = 0;
+
-+ if (unlikely(info->connected != BLKIF_STATE_CONNECTED))
-+ return 1;
++ for (cpu = 0; cpu < NR_CPUS; cpu++) {
++ cpus_clear(cpu_sibling_map[cpu]);
++ cpus_clear(cpu_core_map[cpu]);
++ }
+
-+ if (gnttab_alloc_grant_references(
-+ BLKIF_MAX_SEGMENTS_PER_REQUEST, &gref_head) < 0) {
-+ gnttab_request_free_callback(
-+ &info->callback,
-+ blkif_restart_queue_callback,
-+ info,
-+ BLKIF_MAX_SEGMENTS_PER_REQUEST);
-+ return 1;
++ set_cpu_sibling_map(0);
++
++ if (xen_smp_intr_init(0))
++ BUG();
++
++ cpu_initialized_map = cpumask_of_cpu(0);
++
++ /* Restrict the possible_map according to max_cpus. */
++ while ((num_possible_cpus() > 1) && (num_possible_cpus() > max_cpus)) {
++ for (cpu = NR_CPUS-1; !cpu_isset(cpu, cpu_possible_map); cpu--)
++ continue;
++ cpu_clear(cpu, cpu_possible_map);
+ }
+
-+ /* Fill out a communications ring structure. */
-+ ring_req = RING_GET_REQUEST(&info->ring, info->ring.req_prod_pvt);
-+ id = GET_ID_FROM_FREELIST(info);
-+ info->shadow[id].request = (unsigned long)req;
++ for_each_possible_cpu (cpu) {
++ if (cpu == 0)
++ continue;
+
-+ ring_req->id = id;
-+ ring_req->sector_number = (blkif_sector_t)req->sector;
-+ ring_req->handle = info->handle;
++#ifdef __x86_64__
++ gdt_descr = &cpu_gdt_descr[cpu];
++#else
++ gdt_descr = &per_cpu(cpu_gdt_descr, cpu);
++#endif
++ gdt_descr->address = get_zeroed_page(GFP_KERNEL);
++ if (unlikely(!gdt_descr->address)) {
++ printk(KERN_CRIT "CPU%d failed to allocate GDT\n",
++ cpu);
++ continue;
++ }
++ gdt_descr->size = GDT_SIZE;
++ memcpy((void *)gdt_descr->address, cpu_gdt_table, GDT_SIZE);
++ make_page_readonly(
++ (void *)gdt_descr->address,
++ XENFEAT_writable_descriptor_tables);
+
-+ ring_req->operation = rq_data_dir(req) ?
-+ BLKIF_OP_WRITE : BLKIF_OP_READ;
-+ if (blk_barrier_rq(req))
-+ ring_req->operation = BLKIF_OP_WRITE_BARRIER;
++ apicid = cpu;
++ if (HYPERVISOR_vcpu_op(VCPUOP_get_physid, cpu, &cpu_id) == 0) {
++ apicid = xen_vcpu_physid_to_x86_apicid(cpu_id.phys_id);
++ acpiid = xen_vcpu_physid_to_x86_acpiid(cpu_id.phys_id);
++#ifdef CONFIG_ACPI
++ if (acpiid != 0xff)
++ x86_acpiid_to_apicid[acpiid] = apicid;
++#endif
++ }
++ cpu_data[cpu] = boot_cpu_data;
++ cpu_data[cpu].apicid = apicid;
+
-+ ring_req->nr_segments = 0;
-+ rq_for_each_bio (bio, req) {
-+ bio_for_each_segment (bvec, bio, idx) {
-+ BUG_ON(ring_req->nr_segments
-+ == BLKIF_MAX_SEGMENTS_PER_REQUEST);
-+ buffer_mfn = page_to_phys(bvec->bv_page) >> PAGE_SHIFT;
-+ fsect = bvec->bv_offset >> 9;
-+ lsect = fsect + (bvec->bv_len >> 9) - 1;
-+ /* install a grant reference. */
-+ ref = gnttab_claim_grant_reference(&gref_head);
-+ BUG_ON(ref == -ENOSPC);
++ cpu_2_logical_apicid[cpu] = apicid;
++ x86_cpu_to_apicid[cpu] = apicid;
+
-+ gnttab_grant_foreign_access_ref(
-+ ref,
-+ info->xbdev->otherend_id,
-+ buffer_mfn,
-+ rq_data_dir(req) );
++ idle = fork_idle(cpu);
++ if (IS_ERR(idle))
++ panic("failed fork for CPU %d", cpu);
+
-+ info->shadow[id].frame[ring_req->nr_segments] =
-+ mfn_to_pfn(buffer_mfn);
++#ifdef __x86_64__
++ cpu_pda(cpu)->pcurrent = idle;
++ cpu_pda(cpu)->cpunumber = cpu;
++ clear_ti_thread_flag(idle->thread_info, TIF_FORK);
++#endif
+
-+ ring_req->seg[ring_req->nr_segments] =
-+ (struct blkif_request_segment) {
-+ .gref = ref,
-+ .first_sect = fsect,
-+ .last_sect = lsect };
++ irq_ctx_init(cpu);
+
-+ ring_req->nr_segments++;
-+ }
++#ifdef CONFIG_HOTPLUG_CPU
++ if (is_initial_xendomain())
++ cpu_set(cpu, cpu_present_map);
++#else
++ cpu_set(cpu, cpu_present_map);
++#endif
+ }
+
-+ info->ring.req_prod_pvt++;
++ init_xenbus_allowed_cpumask();
+
-+ /* Keep a private copy so we can reissue requests when recovering. */
-+ info->shadow[id].req = *ring_req;
++#ifdef CONFIG_X86_IO_APIC
++ /*
++ * Here we can be sure that there is an IO-APIC in the system. Let's
++ * go and set it up:
++ */
++ if (!skip_ioapic_setup && nr_ioapics)
++ setup_IO_APIC();
++#endif
++}
++
++void __devinit smp_prepare_boot_cpu(void)
++{
++ prefill_possible_map();
++}
++
++#ifdef CONFIG_HOTPLUG_CPU
++
++/*
++ * Initialize cpu_present_map late to skip SMP boot code in init/main.c.
++ * But do it early enough to catch critical for_each_present_cpu() loops
++ * in i386-specific code.
++ */
++static int __init initialize_cpu_present_map(void)
++{
++ cpu_present_map = cpu_possible_map;
++ return 0;
++}
++core_initcall(initialize_cpu_present_map);
++
++int __cpu_disable(void)
++{
++ cpumask_t map = cpu_online_map;
++ unsigned int cpu = smp_processor_id();
++
++ if (cpu == 0)
++ return -EBUSY;
++
++ remove_siblinginfo(cpu);
+
-+ gnttab_free_grant_references(gref_head);
++ cpu_clear(cpu, map);
++ fixup_irqs(map);
++ cpu_clear(cpu, cpu_online_map);
+
+ return 0;
+}
+
-+/*
-+ * do_blkif_request
-+ * read a block; request is in a request queue
-+ */
-+void do_blkif_request(request_queue_t *rq)
++void __cpu_die(unsigned int cpu)
+{
-+ struct blkfront_info *info = NULL;
-+ struct request *req;
-+ int queued;
++ while (HYPERVISOR_vcpu_op(VCPUOP_is_up, cpu, NULL)) {
++ current->state = TASK_UNINTERRUPTIBLE;
++ schedule_timeout(HZ/10);
++ }
+
-+ DPRINTK("Entered do_blkif_request\n");
++ xen_smp_intr_exit(cpu);
+
-+ queued = 0;
++ if (num_online_cpus() == 1)
++ alternatives_smp_switch(0);
++}
+
-+ while ((req = elv_next_request(rq)) != NULL) {
-+ info = req->rq_disk->private_data;
-+ if (!blk_fs_request(req)) {
-+ end_request(req, 0);
-+ continue;
-+ }
++#endif /* CONFIG_HOTPLUG_CPU */
+
-+ if (RING_FULL(&info->ring))
-+ goto wait;
++int __cpuinit __cpu_up(unsigned int cpu)
++{
++ int rc;
+
-+ DPRINTK("do_blk_req %p: cmd %p, sec %llx, "
-+ "(%u/%li) buffer:%p [%s]\n",
-+ req, req->cmd, (long long)req->sector,
-+ req->current_nr_sectors,
-+ req->nr_sectors, req->buffer,
-+ rq_data_dir(req) ? "write" : "read");
++ rc = cpu_up_check(cpu);
++ if (rc)
++ return rc;
+
++ cpu_initialize_context(cpu);
+
-+ blkdev_dequeue_request(req);
-+ if (blkif_queue_request(req)) {
-+ blk_requeue_request(rq, req);
-+ wait:
-+ /* Avoid pointless unplugs. */
-+ blk_stop_queue(rq);
-+ break;
-+ }
++ if (num_online_cpus() == 1)
++ alternatives_smp_switch(1);
+
-+ queued++;
-+ }
++ /* This must be done before setting cpu_online_map */
++ set_cpu_sibling_map(cpu);
++ wmb();
+
-+ if (queued != 0)
-+ flush_requests(info);
-+}
++ rc = xen_smp_intr_init(cpu);
++ if (rc) {
++ remove_siblinginfo(cpu);
++ return rc;
++ }
+
++ cpu_set(cpu, cpu_online_map);
+
-+static irqreturn_t blkif_int(int irq, void *dev_id, struct pt_regs *ptregs)
-+{
-+ struct request *req;
-+ blkif_response_t *bret;
-+ RING_IDX i, rp;
-+ unsigned long flags;
-+ struct blkfront_info *info = (struct blkfront_info *)dev_id;
-+ int uptodate;
++ rc = HYPERVISOR_vcpu_op(VCPUOP_up, cpu, NULL);
++ BUG_ON(rc);
+
-+ spin_lock_irqsave(&blkif_io_lock, flags);
++ return 0;
++}
+
-+ if (unlikely(info->connected != BLKIF_STATE_CONNECTED)) {
-+ spin_unlock_irqrestore(&blkif_io_lock, flags);
-+ return IRQ_HANDLED;
-+ }
++void __init smp_cpus_done(unsigned int max_cpus)
++{
++}
+
-+ again:
-+ rp = info->ring.sring->rsp_prod;
-+ rmb(); /* Ensure we see queued responses up to 'rp'. */
++#ifndef CONFIG_X86_LOCAL_APIC
++int setup_profiling_timer(unsigned int multiplier)
++{
++ return -EINVAL;
++}
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/xen_proc.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/xen_proc.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,23 @@
+
-+ for (i = info->ring.rsp_cons; i != rp; i++) {
-+ unsigned long id;
-+ int ret;
++#include <linux/module.h>
++#include <linux/proc_fs.h>
++#include <xen/xen_proc.h>
+
-+ bret = RING_GET_RESPONSE(&info->ring, i);
-+ id = bret->id;
-+ req = (struct request *)info->shadow[id].request;
++static struct proc_dir_entry *xen_base;
+
-+ blkif_completion(&info->shadow[id]);
++struct proc_dir_entry *create_xen_proc_entry(const char *name, mode_t mode)
++{
++ if ( xen_base == NULL )
++ if ( (xen_base = proc_mkdir("xen", &proc_root)) == NULL )
++ panic("Couldn't create /proc/xen");
++ return create_proc_entry(name, mode, xen_base);
++}
+
-+ ADD_ID_TO_FREELIST(info, id);
++EXPORT_SYMBOL_GPL(create_xen_proc_entry);
+
-+ uptodate = (bret->status == BLKIF_RSP_OKAY);
-+ switch (bret->operation) {
-+ case BLKIF_OP_WRITE_BARRIER:
-+ if (unlikely(bret->status == BLKIF_RSP_EOPNOTSUPP)) {
-+ printk("blkfront: %s: write barrier op failed\n",
-+ info->gd->disk_name);
-+ uptodate = -EOPNOTSUPP;
-+ info->feature_barrier = 0;
-+ xlvbd_barrier(info);
-+ }
-+ /* fall through */
-+ case BLKIF_OP_READ:
-+ case BLKIF_OP_WRITE:
-+ if (unlikely(bret->status != BLKIF_RSP_OKAY))
-+ DPRINTK("Bad return from blkdev data "
-+ "request: %x\n", bret->status);
++void remove_xen_proc_entry(const char *name)
++{
++ remove_proc_entry(name, xen_base);
++}
+
-+ ret = end_that_request_first(req, uptodate,
-+ req->hard_nr_sectors);
-+ BUG_ON(ret);
-+ end_that_request_last(req, uptodate);
-+ break;
-+ default:
-+ BUG();
-+ }
-+ }
++EXPORT_SYMBOL_GPL(remove_xen_proc_entry);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/xen_sysfs.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/xen_sysfs.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,420 @@
++/*
++ * copyright (c) 2006 IBM Corporation
++ * Authored by: Mike D. Day <ncmike at us.ibm.com>
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License version 2 as
++ * published by the Free Software Foundation.
++ */
+
-+ info->ring.rsp_cons = i;
++#include <linux/err.h>
++#include <linux/kernel.h>
++#include <linux/module.h>
++#include <linux/init.h>
++#include <asm/hypervisor.h>
++#include <xen/features.h>
++#include <xen/hypervisor_sysfs.h>
++#include <xen/xenbus.h>
++#include <xen/interface/kexec.h>
+
-+ if (i != info->ring.req_prod_pvt) {
-+ int more_to_do;
-+ RING_FINAL_CHECK_FOR_RESPONSES(&info->ring, more_to_do);
-+ if (more_to_do)
-+ goto again;
-+ } else
-+ info->ring.sring->rsp_event = i + 1;
++MODULE_LICENSE("GPL");
++MODULE_AUTHOR("Mike D. Day <ncmike at us.ibm.com>");
+
-+ kick_pending_request_queues(info);
++static ssize_t type_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ return sprintf(buffer, "xen\n");
++}
+
-+ spin_unlock_irqrestore(&blkif_io_lock, flags);
++HYPERVISOR_ATTR_RO(type);
+
-+ return IRQ_HANDLED;
++static int __init xen_sysfs_type_init(void)
++{
++ return sysfs_create_file(&hypervisor_subsys.kset.kobj, &type_attr.attr);
+}
+
-+static void blkif_free(struct blkfront_info *info, int suspend)
++static void xen_sysfs_type_destroy(void)
+{
-+ /* Prevent new requests being issued until we fix things up. */
-+ spin_lock_irq(&blkif_io_lock);
-+ info->connected = suspend ?
-+ BLKIF_STATE_SUSPENDED : BLKIF_STATE_DISCONNECTED;
-+ /* No more blkif_request(). */
-+ if (info->rq)
-+ blk_stop_queue(info->rq);
-+ /* No more gnttab callback work. */
-+ gnttab_cancel_free_callback(&info->callback);
-+ spin_unlock_irq(&blkif_io_lock);
-+
-+ /* Flush gnttab callback work. Must be done with no locks held. */
-+ flush_scheduled_work();
-+
-+ /* Free resources associated with old device channel. */
-+ if (info->ring_ref != GRANT_INVALID_REF) {
-+ gnttab_end_foreign_access(info->ring_ref, 0,
-+ (unsigned long)info->ring.sring);
-+ info->ring_ref = GRANT_INVALID_REF;
-+ info->ring.sring = NULL;
-+ }
-+ if (info->irq)
-+ unbind_from_irqhandler(info->irq, info);
-+ info->irq = 0;
++ sysfs_remove_file(&hypervisor_subsys.kset.kobj, &type_attr.attr);
+}
+
-+static void blkif_completion(struct blk_shadow *s)
++/* xen version attributes */
++static ssize_t major_show(struct hyp_sysfs_attr *attr, char *buffer)
+{
-+ int i;
-+ for (i = 0; i < s->req.nr_segments; i++)
-+ gnttab_end_foreign_access(s->req.seg[i].gref, 0, 0UL);
++ int version = HYPERVISOR_xen_version(XENVER_version, NULL);
++ if (version)
++ return sprintf(buffer, "%d\n", version >> 16);
++ return -ENODEV;
+}
+
-+static void blkif_recover(struct blkfront_info *info)
-+{
-+ int i;
-+ blkif_request_t *req;
-+ struct blk_shadow *copy;
-+ int j;
++HYPERVISOR_ATTR_RO(major);
+
-+ /* Stage 1: Make a safe copy of the shadow state. */
-+ copy = kmalloc(sizeof(info->shadow), GFP_KERNEL | __GFP_NOFAIL);
-+ memcpy(copy, info->shadow, sizeof(info->shadow));
++static ssize_t minor_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int version = HYPERVISOR_xen_version(XENVER_version, NULL);
++ if (version)
++ return sprintf(buffer, "%d\n", version & 0xff);
++ return -ENODEV;
++}
+
-+ /* Stage 2: Set up free list. */
-+ memset(&info->shadow, 0, sizeof(info->shadow));
-+ for (i = 0; i < BLK_RING_SIZE; i++)
-+ info->shadow[i].req.id = i+1;
-+ info->shadow_free = info->ring.req_prod_pvt;
-+ info->shadow[BLK_RING_SIZE-1].req.id = 0x0fffffff;
++HYPERVISOR_ATTR_RO(minor);
+
-+ /* Stage 3: Find pending requests and requeue them. */
-+ for (i = 0; i < BLK_RING_SIZE; i++) {
-+ /* Not in use? */
-+ if (copy[i].request == 0)
-+ continue;
++static ssize_t extra_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret = -ENOMEM;
++ char *extra;
+
-+ /* Grab a request slot and copy shadow state into it. */
-+ req = RING_GET_REQUEST(
-+ &info->ring, info->ring.req_prod_pvt);
-+ *req = copy[i].req;
++ extra = kmalloc(XEN_EXTRAVERSION_LEN, GFP_KERNEL);
++ if (extra) {
++ ret = HYPERVISOR_xen_version(XENVER_extraversion, extra);
++ if (!ret)
++ ret = sprintf(buffer, "%s\n", extra);
++ kfree(extra);
++ }
+
-+ /* We get a new request id, and must reset the shadow state. */
-+ req->id = GET_ID_FROM_FREELIST(info);
-+ memcpy(&info->shadow[req->id], ©[i], sizeof(copy[i]));
++ return ret;
++}
+
-+ /* Rewrite any grant references invalidated by susp/resume. */
-+ for (j = 0; j < req->nr_segments; j++)
-+ gnttab_grant_foreign_access_ref(
-+ req->seg[j].gref,
-+ info->xbdev->otherend_id,
-+ pfn_to_mfn(info->shadow[req->id].frame[j]),
-+ rq_data_dir(
-+ (struct request *)
-+ info->shadow[req->id].request));
-+ info->shadow[req->id].req = *req;
++HYPERVISOR_ATTR_RO(extra);
+
-+ info->ring.req_prod_pvt++;
-+ }
++static struct attribute *version_attrs[] = {
++ &major_attr.attr,
++ &minor_attr.attr,
++ &extra_attr.attr,
++ NULL
++};
+
-+ kfree(copy);
++static struct attribute_group version_group = {
++ .name = "version",
++ .attrs = version_attrs,
++};
+
-+ (void)xenbus_switch_state(info->xbdev, XenbusStateConnected);
++static int __init xen_sysfs_version_init(void)
++{
++ return sysfs_create_group(&hypervisor_subsys.kset.kobj,
++ &version_group);
++}
+
-+ spin_lock_irq(&blkif_io_lock);
++static void xen_sysfs_version_destroy(void)
++{
++ sysfs_remove_group(&hypervisor_subsys.kset.kobj, &version_group);
++}
+
-+ /* Now safe for us to use the shared ring */
-+ info->connected = BLKIF_STATE_CONNECTED;
++/* UUID */
+
-+ /* Send off requeued requests */
-+ flush_requests(info);
++static ssize_t uuid_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ char *vm, *val;
++ int ret;
++ extern int xenstored_ready;
+
-+ /* Kick any other new requests queued since we resumed */
-+ kick_pending_request_queues(info);
++ if (!xenstored_ready)
++ return -EBUSY;
+
-+ spin_unlock_irq(&blkif_io_lock);
++ vm = xenbus_read(XBT_NIL, "vm", "", NULL);
++ if (IS_ERR(vm))
++ return PTR_ERR(vm);
++ val = xenbus_read(XBT_NIL, vm, "uuid", NULL);
++ kfree(vm);
++ if (IS_ERR(val))
++ return PTR_ERR(val);
++ ret = sprintf(buffer, "%s\n", val);
++ kfree(val);
++ return ret;
+}
+
-+int blkfront_is_ready(struct xenbus_device *dev)
-+{
-+ struct blkfront_info *info = dev->dev.driver_data;
++HYPERVISOR_ATTR_RO(uuid);
+
-+ return info->is_ready;
++static int __init xen_sysfs_uuid_init(void)
++{
++ return sysfs_create_file(&hypervisor_subsys.kset.kobj, &uuid_attr.attr);
+}
+
++static void xen_sysfs_uuid_destroy(void)
++{
++ sysfs_remove_file(&hypervisor_subsys.kset.kobj, &uuid_attr.attr);
++}
+
-+/* ** Driver Registration ** */
++/* xen compilation attributes */
+
++static ssize_t compiler_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret = -ENOMEM;
++ struct xen_compile_info *info;
+
-+static struct xenbus_device_id blkfront_ids[] = {
-+ { "vbd" },
-+ { "" }
-+};
-+MODULE_ALIAS("xen:vbd");
++ info = kmalloc(sizeof(struct xen_compile_info), GFP_KERNEL);
++ if (info) {
++ ret = HYPERVISOR_xen_version(XENVER_compile_info, info);
++ if (!ret)
++ ret = sprintf(buffer, "%s\n", info->compiler);
++ kfree(info);
++ }
+
-+static struct xenbus_driver blkfront = {
-+ .name = "vbd",
-+ .owner = THIS_MODULE,
-+ .ids = blkfront_ids,
-+ .probe = blkfront_probe,
-+ .remove = blkfront_remove,
-+ .resume = blkfront_resume,
-+ .otherend_changed = backend_changed,
-+ .is_ready = blkfront_is_ready,
-+};
++ return ret;
++}
+
++HYPERVISOR_ATTR_RO(compiler);
+
-+static int __init xlblk_init(void)
++static ssize_t compiled_by_show(struct hyp_sysfs_attr *attr, char *buffer)
+{
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++ int ret = -ENOMEM;
++ struct xen_compile_info *info;
+
-+ return xenbus_register_frontend(&blkfront);
++ info = kmalloc(sizeof(struct xen_compile_info), GFP_KERNEL);
++ if (info) {
++ ret = HYPERVISOR_xen_version(XENVER_compile_info, info);
++ if (!ret)
++ ret = sprintf(buffer, "%s\n", info->compile_by);
++ kfree(info);
++ }
++
++ return ret;
+}
-+module_init(xlblk_init);
+
++HYPERVISOR_ATTR_RO(compiled_by);
+
-+static void xlblk_exit(void)
++static ssize_t compile_date_show(struct hyp_sysfs_attr *attr, char *buffer)
+{
-+ return xenbus_unregister_driver(&blkfront);
-+}
-+module_exit(xlblk_exit);
++ int ret = -ENOMEM;
++ struct xen_compile_info *info;
+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blkfront/block.h tmp-linux-2.6-xen.patch/drivers/xen/blkfront/block.h
---- pristine-linux-2.6.18.2/drivers/xen/blkfront/block.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blkfront/block.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,143 @@
-+/******************************************************************************
-+ * block.h
-+ *
-+ * Shared definitions between all levels of XenLinux Virtual block devices.
-+ *
-+ * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
-+ * Modifications by Mark A. Williamson are (c) Intel Research Cambridge
-+ * Copyright (c) 2004-2005, Christian Limpach
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ info = kmalloc(sizeof(struct xen_compile_info), GFP_KERNEL);
++ if (info) {
++ ret = HYPERVISOR_xen_version(XENVER_compile_info, info);
++ if (!ret)
++ ret = sprintf(buffer, "%s\n", info->compile_date);
++ kfree(info);
++ }
+
-+#ifndef __XEN_DRIVERS_BLOCK_H__
-+#define __XEN_DRIVERS_BLOCK_H__
++ return ret;
++}
+
-+#include <linux/version.h>
-+#include <linux/module.h>
-+#include <linux/kernel.h>
-+#include <linux/sched.h>
-+#include <linux/slab.h>
-+#include <linux/string.h>
-+#include <linux/errno.h>
-+#include <linux/fs.h>
-+#include <linux/hdreg.h>
-+#include <linux/blkdev.h>
-+#include <linux/major.h>
-+#include <asm/hypervisor.h>
-+#include <xen/xenbus.h>
-+#include <xen/gnttab.h>
-+#include <xen/interface/xen.h>
-+#include <xen/interface/io/blkif.h>
-+#include <xen/interface/io/ring.h>
-+#include <asm/io.h>
-+#include <asm/atomic.h>
-+#include <asm/uaccess.h>
++HYPERVISOR_ATTR_RO(compile_date);
+
-+#define DPRINTK(_f, _a...) pr_debug(_f, ## _a)
++static struct attribute *xen_compile_attrs[] = {
++ &compiler_attr.attr,
++ &compiled_by_attr.attr,
++ &compile_date_attr.attr,
++ NULL
++};
+
-+#if 0
-+#define DPRINTK_IOCTL(_f, _a...) printk(KERN_ALERT _f, ## _a)
-+#else
-+#define DPRINTK_IOCTL(_f, _a...) ((void)0)
-+#endif
++static struct attribute_group xen_compilation_group = {
++ .name = "compilation",
++ .attrs = xen_compile_attrs,
++};
+
-+struct xlbd_type_info
++int __init static xen_compilation_init(void)
+{
-+ int partn_shift;
-+ int disks_per_major;
-+ char *devname;
-+ char *diskname;
-+};
++ return sysfs_create_group(&hypervisor_subsys.kset.kobj,
++ &xen_compilation_group);
++}
+
-+struct xlbd_major_info
++static void xen_compilation_destroy(void)
+{
-+ int major;
-+ int index;
-+ int usage;
-+ struct xlbd_type_info *type;
-+};
++ sysfs_remove_group(&hypervisor_subsys.kset.kobj,
++ &xen_compilation_group);
++}
+
-+struct blk_shadow {
-+ blkif_request_t req;
-+ unsigned long request;
-+ unsigned long frame[BLKIF_MAX_SEGMENTS_PER_REQUEST];
-+};
++/* xen properties info */
++
++static ssize_t capabilities_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret = -ENOMEM;
++ char *caps;
++
++ caps = kmalloc(XEN_CAPABILITIES_INFO_LEN, GFP_KERNEL);
++ if (caps) {
++ ret = HYPERVISOR_xen_version(XENVER_capabilities, caps);
++ if (!ret)
++ ret = sprintf(buffer, "%s\n", caps);
++ kfree(caps);
++ }
+
-+#define BLK_RING_SIZE __RING_SIZE((blkif_sring_t *)0, PAGE_SIZE)
++ return ret;
++}
+
-+/*
-+ * We have one of these per vbd, whether ide, scsi or 'other'. They
-+ * hang in private_data off the gendisk structure. We may end up
-+ * putting all kinds of interesting stuff here :-)
-+ */
-+struct blkfront_info
++HYPERVISOR_ATTR_RO(capabilities);
++
++static ssize_t changeset_show(struct hyp_sysfs_attr *attr, char *buffer)
+{
-+ struct xenbus_device *xbdev;
-+ dev_t dev;
-+ struct gendisk *gd;
-+ int vdevice;
-+ blkif_vdev_t handle;
-+ int connected;
-+ int ring_ref;
-+ blkif_front_ring_t ring;
-+ unsigned int irq;
-+ struct xlbd_major_info *mi;
-+ request_queue_t *rq;
-+ struct work_struct work;
-+ struct gnttab_free_callback callback;
-+ struct blk_shadow shadow[BLK_RING_SIZE];
-+ unsigned long shadow_free;
-+ int feature_barrier;
-+ int is_ready;
++ int ret = -ENOMEM;
++ char *cset;
+
-+ /**
-+ * The number of people holding this device open. We won't allow a
-+ * hot-unplug unless this is 0.
-+ */
-+ int users;
-+};
++ cset = kmalloc(XEN_CHANGESET_INFO_LEN, GFP_KERNEL);
++ if (cset) {
++ ret = HYPERVISOR_xen_version(XENVER_changeset, cset);
++ if (!ret)
++ ret = sprintf(buffer, "%s\n", cset);
++ kfree(cset);
++ }
+
-+extern spinlock_t blkif_io_lock;
++ return ret;
++}
+
-+extern int blkif_open(struct inode *inode, struct file *filep);
-+extern int blkif_release(struct inode *inode, struct file *filep);
-+extern int blkif_ioctl(struct inode *inode, struct file *filep,
-+ unsigned command, unsigned long argument);
-+extern int blkif_getgeo(struct block_device *, struct hd_geometry *);
-+extern int blkif_check(dev_t dev);
-+extern int blkif_revalidate(dev_t dev);
-+extern void do_blkif_request (request_queue_t *rq);
++HYPERVISOR_ATTR_RO(changeset);
+
-+/* Virtual block-device subsystem. */
-+/* Note that xlvbd_add doesn't call add_disk for you: you're expected
-+ to call add_disk on info->gd once the disk is properly connected
-+ up. */
-+int xlvbd_add(blkif_sector_t capacity, int device,
-+ u16 vdisk_info, u16 sector_size, struct blkfront_info *info);
-+void xlvbd_del(struct blkfront_info *info);
-+int xlvbd_barrier(struct blkfront_info *info);
++static ssize_t virtual_start_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret = -ENOMEM;
++ struct xen_platform_parameters *parms;
+
-+#endif /* __XEN_DRIVERS_BLOCK_H__ */
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blkfront/Makefile tmp-linux-2.6-xen.patch/drivers/xen/blkfront/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/blkfront/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blkfront/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,5 @@
++ parms = kmalloc(sizeof(struct xen_platform_parameters), GFP_KERNEL);
++ if (parms) {
++ ret = HYPERVISOR_xen_version(XENVER_platform_parameters,
++ parms);
++ if (!ret)
++ ret = sprintf(buffer, "%lx\n", parms->virt_start);
++ kfree(parms);
++ }
+
-+obj-$(CONFIG_XEN_BLKDEV_FRONTEND) := xenblk.o
++ return ret;
++}
+
-+xenblk-objs := blkfront.o vbd.o
++HYPERVISOR_ATTR_RO(virtual_start);
+
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blkfront/vbd.c tmp-linux-2.6-xen.patch/drivers/xen/blkfront/vbd.c
---- pristine-linux-2.6.18.2/drivers/xen/blkfront/vbd.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blkfront/vbd.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,375 @@
-+/******************************************************************************
-+ * vbd.c
-+ *
-+ * XenLinux virtual block-device driver (xvd).
-+ *
-+ * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
-+ * Modifications by Mark A. Williamson are (c) Intel Research Cambridge
-+ * Copyright (c) 2004-2005, Christian Limpach
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++static ssize_t pagesize_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret;
+
-+#include "block.h"
-+#include <linux/blkdev.h>
-+#include <linux/list.h>
++ ret = HYPERVISOR_xen_version(XENVER_pagesize, NULL);
++ if (ret > 0)
++ ret = sprintf(buffer, "%x\n", ret);
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
-+#endif
++ return ret;
++}
+
-+#define BLKIF_MAJOR(dev) ((dev)>>8)
-+#define BLKIF_MINOR(dev) ((dev) & 0xff)
++HYPERVISOR_ATTR_RO(pagesize);
+
-+/*
-+ * For convenience we distinguish between ide, scsi and 'other' (i.e.,
-+ * potentially combinations of the two) in the naming scheme and in a few other
-+ * places.
-+ */
++/* eventually there will be several more features to export */
++static ssize_t xen_feature_show(int index, char *buffer)
++{
++ int ret = -ENOMEM;
++ struct xen_feature_info *info;
+
-+#define NUM_IDE_MAJORS 10
-+#define NUM_SCSI_MAJORS 17
-+#define NUM_VBD_MAJORS 1
++ info = kmalloc(sizeof(struct xen_feature_info), GFP_KERNEL);
++ if (info) {
++ info->submap_idx = index;
++ ret = HYPERVISOR_xen_version(XENVER_get_features, info);
++ if (!ret)
++ ret = sprintf(buffer, "%d\n", info->submap);
++ kfree(info);
++ }
+
-+static struct xlbd_type_info xlbd_ide_type = {
-+ .partn_shift = 6,
-+ .disks_per_major = 2,
-+ .devname = "ide",
-+ .diskname = "hd",
-+};
++ return ret;
++}
+
-+static struct xlbd_type_info xlbd_scsi_type = {
-+ .partn_shift = 4,
-+ .disks_per_major = 16,
-+ .devname = "sd",
-+ .diskname = "sd",
-+};
++static ssize_t writable_pt_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ return xen_feature_show(XENFEAT_writable_page_tables, buffer);
++}
+
-+static struct xlbd_type_info xlbd_vbd_type = {
-+ .partn_shift = 4,
-+ .disks_per_major = 16,
-+ .devname = "xvd",
-+ .diskname = "xvd",
++HYPERVISOR_ATTR_RO(writable_pt);
++
++static struct attribute *xen_properties_attrs[] = {
++ &capabilities_attr.attr,
++ &changeset_attr.attr,
++ &virtual_start_attr.attr,
++ &pagesize_attr.attr,
++ &writable_pt_attr.attr,
++ NULL
+};
+
-+static struct xlbd_major_info *major_info[NUM_IDE_MAJORS + NUM_SCSI_MAJORS +
-+ NUM_VBD_MAJORS];
++static struct attribute_group xen_properties_group = {
++ .name = "properties",
++ .attrs = xen_properties_attrs,
++};
+
-+#define XLBD_MAJOR_IDE_START 0
-+#define XLBD_MAJOR_SCSI_START (NUM_IDE_MAJORS)
-+#define XLBD_MAJOR_VBD_START (NUM_IDE_MAJORS + NUM_SCSI_MAJORS)
++static int __init xen_properties_init(void)
++{
++ return sysfs_create_group(&hypervisor_subsys.kset.kobj,
++ &xen_properties_group);
++}
+
-+#define XLBD_MAJOR_IDE_RANGE XLBD_MAJOR_IDE_START ... XLBD_MAJOR_SCSI_START - 1
-+#define XLBD_MAJOR_SCSI_RANGE XLBD_MAJOR_SCSI_START ... XLBD_MAJOR_VBD_START - 1
-+#define XLBD_MAJOR_VBD_RANGE XLBD_MAJOR_VBD_START ... XLBD_MAJOR_VBD_START + NUM_VBD_MAJORS - 1
++static void xen_properties_destroy(void)
++{
++ sysfs_remove_group(&hypervisor_subsys.kset.kobj,
++ &xen_properties_group);
++}
+
-+/* Information about our VBDs. */
-+#define MAX_VBDS 64
-+static LIST_HEAD(vbds_list);
++#ifdef CONFIG_KEXEC
+
-+static struct block_device_operations xlvbd_block_fops =
++static ssize_t vmcoreinfo_show(struct hyp_sysfs_attr *attr, char *page)
+{
-+ .owner = THIS_MODULE,
-+ .open = blkif_open,
-+ .release = blkif_release,
-+ .ioctl = blkif_ioctl,
-+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
-+ .getgeo = blkif_getgeo
-+#endif
-+};
++ extern size_t vmcoreinfo_size_xen;
++ extern unsigned long paddr_vmcoreinfo_xen;
+
-+DEFINE_SPINLOCK(blkif_io_lock);
++ return sprintf(page, "%lx %zx\n",
++ paddr_vmcoreinfo_xen, vmcoreinfo_size_xen);
++}
+
-+static struct xlbd_major_info *
-+xlbd_alloc_major_info(int major, int minor, int index)
-+{
-+ struct xlbd_major_info *ptr;
++HYPERVISOR_ATTR_RO(vmcoreinfo);
+
-+ ptr = kzalloc(sizeof(struct xlbd_major_info), GFP_KERNEL);
-+ if (ptr == NULL)
-+ return NULL;
++static int __init xen_sysfs_vmcoreinfo_init(void)
++{
++ return sysfs_create_file(&hypervisor_subsys.kset.kobj,
++ &vmcoreinfo_attr.attr);
++}
+
-+ ptr->major = major;
++static void xen_sysfs_vmcoreinfo_destroy(void)
++{
++ sysfs_remove_file(&hypervisor_subsys.kset.kobj, &vmcoreinfo_attr.attr);
++}
+
-+ switch (index) {
-+ case XLBD_MAJOR_IDE_RANGE:
-+ ptr->type = &xlbd_ide_type;
-+ ptr->index = index - XLBD_MAJOR_IDE_START;
-+ break;
-+ case XLBD_MAJOR_SCSI_RANGE:
-+ ptr->type = &xlbd_scsi_type;
-+ ptr->index = index - XLBD_MAJOR_SCSI_START;
-+ break;
-+ case XLBD_MAJOR_VBD_RANGE:
-+ ptr->type = &xlbd_vbd_type;
-+ ptr->index = index - XLBD_MAJOR_VBD_START;
-+ break;
-+ }
++#else
+
-+ if (register_blkdev(ptr->major, ptr->type->devname)) {
-+ kfree(ptr);
-+ return NULL;
-+ }
++#define xen_sysfs_vmcoreinfo_init() 0
++#define xen_sysfs_vmcoreinfo_destroy() ((void)0)
+
-+ printk("xen-vbd: registered block device major %i\n", ptr->major);
-+ major_info[index] = ptr;
-+ return ptr;
-+}
++#endif
+
-+static struct xlbd_major_info *
-+xlbd_get_major_info(int vdevice)
++static int __init hyper_sysfs_init(void)
+{
-+ struct xlbd_major_info *mi;
-+ int major, minor, index;
++ int ret;
+
-+ major = BLKIF_MAJOR(vdevice);
-+ minor = BLKIF_MINOR(vdevice);
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+ switch (major) {
-+ case IDE0_MAJOR: index = 0; break;
-+ case IDE1_MAJOR: index = 1; break;
-+ case IDE2_MAJOR: index = 2; break;
-+ case IDE3_MAJOR: index = 3; break;
-+ case IDE4_MAJOR: index = 4; break;
-+ case IDE5_MAJOR: index = 5; break;
-+ case IDE6_MAJOR: index = 6; break;
-+ case IDE7_MAJOR: index = 7; break;
-+ case IDE8_MAJOR: index = 8; break;
-+ case IDE9_MAJOR: index = 9; break;
-+ case SCSI_DISK0_MAJOR: index = 10; break;
-+ case SCSI_DISK1_MAJOR ... SCSI_DISK7_MAJOR:
-+ index = 11 + major - SCSI_DISK1_MAJOR;
-+ break;
-+ case SCSI_DISK8_MAJOR ... SCSI_DISK15_MAJOR:
-+ index = 18 + major - SCSI_DISK8_MAJOR;
-+ break;
-+ case SCSI_CDROM_MAJOR: index = 26; break;
-+ default: index = 27; break;
-+ }
++ ret = xen_sysfs_type_init();
++ if (ret)
++ goto out;
++ ret = xen_sysfs_version_init();
++ if (ret)
++ goto version_out;
++ ret = xen_compilation_init();
++ if (ret)
++ goto comp_out;
++ ret = xen_sysfs_uuid_init();
++ if (ret)
++ goto uuid_out;
++ ret = xen_properties_init();
++ if (ret)
++ goto prop_out;
++ ret = xen_sysfs_vmcoreinfo_init();
++ if (!ret)
++ goto out;
+
-+ mi = ((major_info[index] != NULL) ? major_info[index] :
-+ xlbd_alloc_major_info(major, minor, index));
-+ if (mi)
-+ mi->usage++;
-+ return mi;
++ xen_properties_destroy();
++prop_out:
++ xen_sysfs_uuid_destroy();
++uuid_out:
++ xen_compilation_destroy();
++comp_out:
++ xen_sysfs_version_destroy();
++version_out:
++ xen_sysfs_type_destroy();
++out:
++ return ret;
+}
+
-+static void
-+xlbd_put_major_info(struct xlbd_major_info *mi)
++static void __exit hyper_sysfs_exit(void)
+{
-+ mi->usage--;
-+ /* XXX: release major if 0 */
++ xen_sysfs_vmcoreinfo_destroy();
++ xen_properties_destroy();
++ xen_compilation_destroy();
++ xen_sysfs_uuid_destroy();
++ xen_sysfs_version_destroy();
++ xen_sysfs_type_destroy();
++
+}
+
-+static int
-+xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size)
-+{
-+ request_queue_t *rq;
++module_init(hyper_sysfs_init);
++module_exit(hyper_sysfs_exit);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/core/xencomm.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/core/xencomm.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,229 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Hollis Blanchard <hollisb at us.ibm.com>
++ */
+
-+ rq = blk_init_queue(do_blkif_request, &blkif_io_lock);
-+ if (rq == NULL)
-+ return -1;
++#include <linux/gfp.h>
++#include <linux/mm.h>
++#include <asm/page.h>
++#include <xen/xencomm.h>
++#include <xen/interface/xen.h>
++#ifdef __ia64__
++#include <asm/xen/xencomm.h> /* for is_kern_addr() */
++#endif
+
-+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,10)
-+ elevator_init(rq, "noop");
-+#else
-+ elevator_init(rq, &elevator_noop);
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
+#endif
+
-+ /* Hard sector size and max sectors impersonate the equiv. hardware. */
-+ blk_queue_hardsect_size(rq, sector_size);
-+ blk_queue_max_sectors(rq, 512);
++static int xencomm_init(struct xencomm_desc *desc,
++ void *buffer, unsigned long bytes)
++{
++ unsigned long recorded = 0;
++ int i = 0;
+
-+ /* Each segment in a request is up to an aligned page in size. */
-+ blk_queue_segment_boundary(rq, PAGE_SIZE - 1);
-+ blk_queue_max_segment_size(rq, PAGE_SIZE);
++ while ((recorded < bytes) && (i < desc->nr_addrs)) {
++ unsigned long vaddr = (unsigned long)buffer + recorded;
++ unsigned long paddr;
++ int offset;
++ int chunksz;
+
-+ /* Ensure a merged request will fit in a single I/O ring slot. */
-+ blk_queue_max_phys_segments(rq, BLKIF_MAX_SEGMENTS_PER_REQUEST);
-+ blk_queue_max_hw_segments(rq, BLKIF_MAX_SEGMENTS_PER_REQUEST);
++ offset = vaddr % PAGE_SIZE; /* handle partial pages */
++ chunksz = min(PAGE_SIZE - offset, bytes - recorded);
+
-+ /* Make sure buffer addresses are sector-aligned. */
-+ blk_queue_dma_alignment(rq, 511);
++ paddr = xencomm_vtop(vaddr);
++ if (paddr == ~0UL) {
++ printk("%s: couldn't translate vaddr %lx\n",
++ __func__, vaddr);
++ return -EINVAL;
++ }
+
-+ /* Make sure we don't use bounce buffers. */
-+ blk_queue_bounce_limit(rq, BLK_BOUNCE_ANY);
++ desc->address[i++] = paddr;
++ recorded += chunksz;
++ }
+
-+ gd->queue = rq;
++ if (recorded < bytes) {
++ printk("%s: could only translate %ld of %ld bytes\n",
++ __func__, recorded, bytes);
++ return -ENOSPC;
++ }
++
++ /* mark remaining addresses invalid (just for safety) */
++ while (i < desc->nr_addrs)
++ desc->address[i++] = XENCOMM_INVALID;
++
++ desc->magic = XENCOMM_MAGIC;
+
+ return 0;
+}
+
-+static int
-+xlvbd_alloc_gendisk(int minor, blkif_sector_t capacity, int vdevice,
-+ u16 vdisk_info, u16 sector_size,
-+ struct blkfront_info *info)
++static struct xencomm_desc *xencomm_alloc(gfp_t gfp_mask,
++ void *buffer, unsigned long bytes)
+{
-+ struct gendisk *gd;
-+ struct xlbd_major_info *mi;
-+ int nr_minors = 1;
-+ int err = -ENODEV;
-+ unsigned int offset;
++ struct xencomm_desc *desc;
++ unsigned long buffer_ulong = (unsigned long)buffer;
++ unsigned long start = buffer_ulong & PAGE_MASK;
++ unsigned long end = (buffer_ulong + bytes) | ~PAGE_MASK;
++ unsigned long nr_addrs = (end - start + 1) >> PAGE_SHIFT;
++ unsigned long size = sizeof(*desc) +
++ sizeof(desc->address[0]) * nr_addrs;
++
++ /*
++ * slab allocator returns at least sizeof(void*) aligned pointer.
++ * When sizeof(*desc) > sizeof(void*), struct xencomm_desc might
++ * cross page boundary.
++ */
++ if (sizeof(*desc) > sizeof(void*)) {
++ unsigned long order = get_order(size);
++ desc = (struct xencomm_desc *)__get_free_pages(gfp_mask,
++ order);
++ if (desc == NULL)
++ return NULL;
+
-+ BUG_ON(info->gd != NULL);
-+ BUG_ON(info->mi != NULL);
-+ BUG_ON(info->rq != NULL);
++ desc->nr_addrs =
++ ((PAGE_SIZE << order) - sizeof(struct xencomm_desc)) /
++ sizeof(*desc->address);
++ } else {
++ desc = kmalloc(size, gfp_mask);
++ if (desc == NULL)
++ return NULL;
+
-+ mi = xlbd_get_major_info(vdevice);
-+ if (mi == NULL)
-+ goto out;
-+ info->mi = mi;
++ desc->nr_addrs = nr_addrs;
++ }
++ return desc;
++}
+
-+ if ((minor & ((1 << mi->type->partn_shift) - 1)) == 0)
-+ nr_minors = 1 << mi->type->partn_shift;
++void xencomm_free(struct xencomm_handle *desc)
++{
++ if (desc && !((ulong)desc & XENCOMM_INLINE_FLAG)) {
++ struct xencomm_desc *desc__ = (struct xencomm_desc*)desc;
++ if (sizeof(*desc__) > sizeof(void*)) {
++ unsigned long size = sizeof(*desc__) +
++ sizeof(desc__->address[0]) * desc__->nr_addrs;
++ unsigned long order = get_order(size);
++ free_pages((unsigned long)__va(desc), order);
++ } else
++ kfree(__va(desc));
++ }
++}
+
-+ gd = alloc_disk(nr_minors);
-+ if (gd == NULL)
-+ goto out;
++static int xencomm_create(void *buffer, unsigned long bytes, struct xencomm_desc **ret, gfp_t gfp_mask)
++{
++ struct xencomm_desc *desc;
++ int rc;
+
-+ offset = mi->index * mi->type->disks_per_major +
-+ (minor >> mi->type->partn_shift);
-+ if (nr_minors > 1) {
-+ if (offset < 26) {
-+ sprintf(gd->disk_name, "%s%c",
-+ mi->type->diskname, 'a' + offset );
-+ }
-+ else {
-+ sprintf(gd->disk_name, "%s%c%c",
-+ mi->type->diskname,
-+ 'a' + ((offset/26)-1), 'a' + (offset%26) );
-+ }
-+ }
-+ else {
-+ if (offset < 26) {
-+ sprintf(gd->disk_name, "%s%c%d",
-+ mi->type->diskname,
-+ 'a' + offset,
-+ minor & ((1 << mi->type->partn_shift) - 1));
-+ }
-+ else {
-+ sprintf(gd->disk_name, "%s%c%c%d",
-+ mi->type->diskname,
-+ 'a' + ((offset/26)-1), 'a' + (offset%26),
-+ minor & ((1 << mi->type->partn_shift) - 1));
-+ }
++ pr_debug("%s: %p[%ld]\n", __func__, buffer, bytes);
++
++ if (bytes == 0) {
++ /* don't create a descriptor; Xen recognizes NULL. */
++ BUG_ON(buffer != NULL);
++ *ret = NULL;
++ return 0;
+ }
+
-+ gd->major = mi->major;
-+ gd->first_minor = minor;
-+ gd->fops = &xlvbd_block_fops;
-+ gd->private_data = info;
-+ gd->driverfs_dev = &(info->xbdev->dev);
-+ set_capacity(gd, capacity);
++ BUG_ON(buffer == NULL); /* 'bytes' is non-zero */
+
-+ if (xlvbd_init_blk_queue(gd, sector_size)) {
-+ del_gendisk(gd);
-+ goto out;
++ desc = xencomm_alloc(gfp_mask, buffer, bytes);
++ if (!desc) {
++ printk("%s failure\n", "xencomm_alloc");
++ return -ENOMEM;
+ }
+
-+ info->rq = gd->queue;
-+ info->gd = gd;
++ rc = xencomm_init(desc, buffer, bytes);
++ if (rc) {
++ printk("%s failure: %d\n", "xencomm_init", rc);
++ xencomm_free((struct xencomm_handle *)__pa(desc));
++ return rc;
++ }
+
-+ if (info->feature_barrier)
-+ xlvbd_barrier(info);
++ *ret = desc;
++ return 0;
++}
+
-+ if (vdisk_info & VDISK_READONLY)
-+ set_disk_ro(gd, 1);
++/* check if memory address is within VMALLOC region */
++static int is_phys_contiguous(unsigned long addr)
++{
++ if (!is_kernel_addr(addr))
++ return 0;
+
-+ if (vdisk_info & VDISK_REMOVABLE)
-+ gd->flags |= GENHD_FL_REMOVABLE;
++ return (addr < VMALLOC_START) || (addr >= VMALLOC_END);
++}
+
-+ if (vdisk_info & VDISK_CDROM)
-+ gd->flags |= GENHD_FL_CD;
++static struct xencomm_handle *xencomm_create_inline(void *ptr)
++{
++ unsigned long paddr;
+
-+ return 0;
++ BUG_ON(!is_phys_contiguous((unsigned long)ptr));
+
-+ out:
-+ if (mi)
-+ xlbd_put_major_info(mi);
-+ info->mi = NULL;
-+ return err;
++ paddr = (unsigned long)xencomm_pa(ptr);
++ BUG_ON(paddr & XENCOMM_INLINE_FLAG);
++ return (struct xencomm_handle *)(paddr | XENCOMM_INLINE_FLAG);
+}
+
-+int
-+xlvbd_add(blkif_sector_t capacity, int vdevice, u16 vdisk_info,
-+ u16 sector_size, struct blkfront_info *info)
++/* "mini" routine, for stack-based communications: */
++static int xencomm_create_mini(void *buffer,
++ unsigned long bytes, struct xencomm_mini *xc_desc,
++ struct xencomm_desc **ret)
+{
-+ struct block_device *bd;
-+ int err = 0;
++ int rc = 0;
++ struct xencomm_desc *desc;
++ BUG_ON(((unsigned long)xc_desc) % sizeof(*xc_desc) != 0);
+
-+ info->dev = MKDEV(BLKIF_MAJOR(vdevice), BLKIF_MINOR(vdevice));
++ desc = (void *)xc_desc;
+
-+ bd = bdget(info->dev);
-+ if (bd == NULL)
-+ return -ENODEV;
++ desc->nr_addrs = XENCOMM_MINI_ADDRS;
+
-+ err = xlvbd_alloc_gendisk(BLKIF_MINOR(vdevice), capacity, vdevice,
-+ vdisk_info, sector_size, info);
++ if (!(rc = xencomm_init(desc, buffer, bytes)))
++ *ret = desc;
+
-+ bdput(bd);
-+ return err;
++ return rc;
+}
+
-+void
-+xlvbd_del(struct blkfront_info *info)
++struct xencomm_handle *xencomm_map(void *ptr, unsigned long bytes)
+{
-+ if (info->mi == NULL)
-+ return;
++ int rc;
++ struct xencomm_desc *desc;
+
-+ BUG_ON(info->gd == NULL);
-+ del_gendisk(info->gd);
-+ put_disk(info->gd);
-+ info->gd = NULL;
++ if (is_phys_contiguous((unsigned long)ptr))
++ return xencomm_create_inline(ptr);
+
-+ xlbd_put_major_info(info->mi);
-+ info->mi = NULL;
++ rc = xencomm_create(ptr, bytes, &desc, GFP_KERNEL);
+
-+ BUG_ON(info->rq == NULL);
-+ blk_cleanup_queue(info->rq);
-+ info->rq = NULL;
++ if (rc || desc == NULL)
++ return NULL;
++
++ return xencomm_pa(desc);
+}
+
-+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
-+int
-+xlvbd_barrier(struct blkfront_info *info)
++struct xencomm_handle *__xencomm_map_no_alloc(void *ptr, unsigned long bytes,
++ struct xencomm_mini *xc_desc)
+{
-+ int err;
++ int rc;
++ struct xencomm_desc *desc = NULL;
+
-+ err = blk_queue_ordered(info->rq,
-+ info->feature_barrier ? QUEUE_ORDERED_DRAIN : QUEUE_ORDERED_NONE, NULL);
-+ if (err)
-+ return err;
-+ printk("blkfront: %s: barriers %s\n",
-+ info->gd->disk_name, info->feature_barrier ? "enabled" : "disabled");
-+ return 0;
-+}
-+#else
-+int
-+xlvbd_barrier(struct blkfront_info *info)
-+{
-+ printk("blkfront: %s: barriers disabled\n", info->gd->disk_name);
-+ return -ENOSYS;
++ if (is_phys_contiguous((unsigned long)ptr))
++ return xencomm_create_inline(ptr);
++
++ rc = xencomm_create_mini(ptr, bytes, xc_desc,
++ &desc);
++
++ if (rc)
++ return NULL;
++
++ return xencomm_pa(desc);
+}
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blktap/blktap.c tmp-linux-2.6-xen.patch/drivers/xen/blktap/blktap.c
---- pristine-linux-2.6.18.2/drivers/xen/blktap/blktap.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blktap/blktap.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,1632 @@
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/evtchn/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/evtchn/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,2 @@
++
++obj-y := evtchn.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/evtchn/evtchn.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/evtchn/evtchn.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,560 @@
+/******************************************************************************
-+ * drivers/xen/blktap/blktap.c
++ * evtchn.c
+ *
-+ * Back-end driver for user level virtual block devices. This portion of the
-+ * driver exports a 'unified' block-device interface that can be accessed
-+ * by any operating system that implements a compatible front end. Requests
-+ * are remapped to a user-space memory region.
-+ *
-+ * Based on the blkback driver code.
++ * Driver for receiving and demuxing event-channel signals.
++ *
++ * Copyright (c) 2004-2005, K A Fraser
++ * Multi-process extensions Copyright (c) 2004, Steven Smith
+ *
-+ * Copyright (c) 2004-2005, Andrew Warfield and Julian Chesterfield
-+ *
-+ * Clean ups and fix ups:
-+ * Copyright (c) 2006, Steven Rostedt - Red Hat, Inc.
-+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation; or, when distributed
@@ -54345,2614 +90108,2890 @@
+ * IN THE SOFTWARE.
+ */
+
-+#include <linux/spinlock.h>
-+#include <linux/kthread.h>
-+#include <linux/list.h>
-+#include <asm/hypervisor.h>
-+#include "common.h"
-+#include <xen/balloon.h>
-+#include <xen/driver_util.h>
++#include <linux/module.h>
+#include <linux/kernel.h>
++#include <linux/sched.h>
++#include <linux/slab.h>
++#include <linux/string.h>
++#include <linux/errno.h>
+#include <linux/fs.h>
-+#include <linux/mm.h>
+#include <linux/errno.h>
++#include <linux/miscdevice.h>
+#include <linux/major.h>
-+#include <linux/gfp.h>
++#include <linux/proc_fs.h>
++#include <linux/stat.h>
+#include <linux/poll.h>
-+#include <asm/tlbflush.h>
++#include <linux/irq.h>
++#include <linux/init.h>
++#include <linux/gfp.h>
++#include <linux/mutex.h>
++#include <linux/cpu.h>
++#include <xen/evtchn.h>
++#include <xen/public/evtchn.h>
+
-+#define MAX_TAP_DEV 256 /*the maximum number of tapdisk ring devices */
-+#define MAX_DEV_NAME 100 /*the max tapdisk ring device name e.g. blktap0 */
++struct per_user_data {
++ /* Notification ring, accessed via /dev/xen/evtchn. */
++#define EVTCHN_RING_SIZE (PAGE_SIZE / sizeof(evtchn_port_t))
++#define EVTCHN_RING_MASK(_i) ((_i)&(EVTCHN_RING_SIZE-1))
++ evtchn_port_t *ring;
++ unsigned int ring_cons, ring_prod, ring_overflow;
++ struct mutex ring_cons_mutex; /* protect against concurrent readers */
+
-+/*
-+ * The maximum number of requests that can be outstanding at any time
-+ * is determined by
-+ *
-+ * [mmap_alloc * MAX_PENDING_REQS * BLKIF_MAX_SEGMENTS_PER_REQUEST]
-+ *
-+ * where mmap_alloc < MAX_DYNAMIC_MEM.
-+ *
-+ * TODO:
-+ * mmap_alloc is initialised to 2 and should be adjustable on the fly via
-+ * sysfs.
-+ */
-+#define BLK_RING_SIZE __RING_SIZE((blkif_sring_t *)0, PAGE_SIZE)
-+#define MAX_DYNAMIC_MEM BLK_RING_SIZE
-+#define MAX_PENDING_REQS BLK_RING_SIZE
-+#define MMAP_PAGES (MAX_PENDING_REQS * BLKIF_MAX_SEGMENTS_PER_REQUEST)
-+#define MMAP_VADDR(_start, _req,_seg) \
-+ (_start + \
-+ ((_req) * BLKIF_MAX_SEGMENTS_PER_REQUEST * PAGE_SIZE) + \
-+ ((_seg) * PAGE_SIZE))
-+static int blkif_reqs = MAX_PENDING_REQS;
-+static int mmap_pages = MMAP_PAGES;
++ /* Processes wait on this queue when ring is empty. */
++ wait_queue_head_t evtchn_wait;
++ struct fasync_struct *evtchn_async_queue;
+
-+#define RING_PAGES 1 /* BLKTAP - immediately before the mmap area, we
-+ * have a bunch of pages reserved for shared
-+ * memory rings.
-+ */
++ int bind_cpu;
++ int nr_event_wrong_delivery;
++};
+
-+/*Data struct handed back to userspace for tapdisk device to VBD mapping*/
-+typedef struct domid_translate {
-+ unsigned short domid;
-+ unsigned short busid;
-+} domid_translate_t ;
++/* Who's bound to each port? */
++static struct per_user_data *port_user[NR_EVENT_CHANNELS];
++static spinlock_t port_user_lock;
+
-+/*Data struct associated with each of the tapdisk devices*/
-+typedef struct tap_blkif {
-+ struct vm_area_struct *vma; /*Shared memory area */
-+ unsigned long rings_vstart; /*Kernel memory mapping */
-+ unsigned long user_vstart; /*User memory mapping */
-+ unsigned long dev_inuse; /*One process opens device at a time. */
-+ unsigned long dev_pending; /*In process of being opened */
-+ unsigned long ring_ok; /*make this ring->state */
-+ blkif_front_ring_t ufe_ring; /*Rings up to user space. */
-+ wait_queue_head_t wait; /*for poll */
-+ unsigned long mode; /*current switching mode */
-+ int minor; /*Minor number for tapdisk device */
-+ pid_t pid; /*tapdisk process id */
-+ enum { RUNNING, CLEANSHUTDOWN } status; /*Detect a clean userspace
-+ shutdown */
-+ unsigned long *idx_map; /*Record the user ring id to kern
-+ [req id, idx] tuple */
-+ blkif_t *blkif; /*Associate blkif with tapdev */
-+ struct domid_translate trans; /*Translation from domid to bus. */
-+} tap_blkif_t;
++void evtchn_device_upcall(int port)
++{
++ struct per_user_data *u;
+
-+static struct tap_blkif *tapfds[MAX_TAP_DEV];
-+static int blktap_next_minor;
++ spin_lock(&port_user_lock);
+
-+static int __init set_blkif_reqs(char *str)
-+{
-+ get_option(&str, &blkif_reqs);
-+ return 1;
++ mask_evtchn(port);
++ clear_evtchn(port);
++
++ if ((u = port_user[port]) != NULL) {
++ if ((u->ring_prod - u->ring_cons) < EVTCHN_RING_SIZE) {
++ u->ring[EVTCHN_RING_MASK(u->ring_prod)] = port;
++ wmb(); /* Ensure ring contents visible */
++ if (u->ring_cons == u->ring_prod++) {
++ wake_up_interruptible(&u->evtchn_wait);
++ kill_fasync(&u->evtchn_async_queue,
++ SIGIO, POLL_IN);
++ }
++ } else {
++ u->ring_overflow = 1;
++ }
++ }
++
++ spin_unlock(&port_user_lock);
+}
-+__setup("blkif_reqs=", set_blkif_reqs);
+
-+/* Run-time switchable: /sys/module/blktap/parameters/ */
-+static unsigned int log_stats = 0;
-+static unsigned int debug_lvl = 0;
-+module_param(log_stats, int, 0644);
-+module_param(debug_lvl, int, 0644);
++static void evtchn_check_wrong_delivery(struct per_user_data *u)
++{
++ evtchn_port_t port;
++ unsigned int current_cpu = smp_processor_id();
+
-+/*
-+ * Each outstanding request that we've passed to the lower device layers has a
-+ * 'pending_req' allocated to it. Each buffer_head that completes decrements
-+ * the pendcnt towards zero. When it hits zero, the specified domain has a
-+ * response queued for it, with the saved 'id' passed back.
-+ */
-+typedef struct {
-+ blkif_t *blkif;
-+ u64 id;
-+ unsigned short mem_idx;
-+ int nr_pages;
-+ atomic_t pendcnt;
-+ unsigned short operation;
-+ int status;
-+ struct list_head free_list;
-+ int inuse;
-+} pending_req_t;
++ /* Delivered to correct CPU? All is good. */
++ if (u->bind_cpu == current_cpu) {
++ u->nr_event_wrong_delivery = 0;
++ return;
++ }
+
-+static pending_req_t *pending_reqs[MAX_PENDING_REQS];
-+static struct list_head pending_free;
-+static DEFINE_SPINLOCK(pending_free_lock);
-+static DECLARE_WAIT_QUEUE_HEAD (pending_free_wq);
-+static int alloc_pending_reqs;
++ /* Tolerate up to 100 consecutive misdeliveries. */
++ if (++u->nr_event_wrong_delivery < 100)
++ return;
+
-+typedef unsigned int PEND_RING_IDX;
++ spin_lock_irq(&port_user_lock);
+
-+static inline int MASK_PEND_IDX(int i) {
-+ return (i & (MAX_PENDING_REQS-1));
-+}
++ for (port = 0; port < NR_EVENT_CHANNELS; port++)
++ if (port_user[port] == u)
++ rebind_evtchn_to_cpu(port, current_cpu);
+
-+static inline unsigned int RTN_PEND_IDX(pending_req_t *req, int idx) {
-+ return (req - pending_reqs[idx]);
++ u->bind_cpu = current_cpu;
++ u->nr_event_wrong_delivery = 0;
++
++ spin_unlock_irq(&port_user_lock);
+}
+
-+#define NR_PENDING_REQS (MAX_PENDING_REQS - pending_prod + pending_cons)
++static ssize_t evtchn_read(struct file *file, char __user *buf,
++ size_t count, loff_t *ppos)
++{
++ int rc;
++ unsigned int c, p, bytes1 = 0, bytes2 = 0;
++ struct per_user_data *u = file->private_data;
+
-+#define BLKBACK_INVALID_HANDLE (~0)
++ /* Whole number of ports. */
++ count &= ~(sizeof(evtchn_port_t)-1);
+
-+static struct page **foreign_pages[MAX_DYNAMIC_MEM];
-+static inline unsigned long idx_to_kaddr(
-+ unsigned int mmap_idx, unsigned int req_idx, unsigned int sg_idx)
-+{
-+ unsigned int arr_idx = req_idx*BLKIF_MAX_SEGMENTS_PER_REQUEST + sg_idx;
-+ unsigned long pfn = page_to_pfn(foreign_pages[mmap_idx][arr_idx]);
-+ return (unsigned long)pfn_to_kaddr(pfn);
-+}
++ if (count == 0)
++ return 0;
+
-+static unsigned short mmap_alloc = 0;
-+static unsigned short mmap_lock = 0;
-+static unsigned short mmap_inuse = 0;
++ if (count > PAGE_SIZE)
++ count = PAGE_SIZE;
+
-+/******************************************************************
-+ * GRANT HANDLES
-+ */
++ for (;;) {
++ mutex_lock(&u->ring_cons_mutex);
+
-+/* When using grant tables to map a frame for device access then the
-+ * handle returned must be used to unmap the frame. This is needed to
-+ * drop the ref count on the frame.
-+ */
-+struct grant_handle_pair
-+{
-+ grant_handle_t kernel;
-+ grant_handle_t user;
-+};
-+#define INVALID_GRANT_HANDLE 0xFFFF
++ rc = -EFBIG;
++ if (u->ring_overflow)
++ goto unlock_out;
+
-+static struct grant_handle_pair
-+ pending_grant_handles[MAX_DYNAMIC_MEM][MMAP_PAGES];
-+#define pending_handle(_id, _idx, _i) \
-+ (pending_grant_handles[_id][((_idx) * BLKIF_MAX_SEGMENTS_PER_REQUEST) \
-+ + (_i)])
++ if ((c = u->ring_cons) != (p = u->ring_prod))
++ break;
+
++ mutex_unlock(&u->ring_cons_mutex);
+
-+static int blktap_read_ufe_ring(tap_blkif_t *info); /*local prototypes*/
++ if (file->f_flags & O_NONBLOCK)
++ return -EAGAIN;
+
-+#define BLKTAP_MINOR 0 /*/dev/xen/blktap has a dynamic major */
-+#define BLKTAP_DEV_DIR "/dev/xen"
++ rc = wait_event_interruptible(
++ u->evtchn_wait, u->ring_cons != u->ring_prod);
++ if (rc)
++ return rc;
++ }
+
-+static int blktap_major;
++ /* Byte lengths of two chunks. Chunk split (if any) is at ring wrap. */
++ if (((c ^ p) & EVTCHN_RING_SIZE) != 0) {
++ bytes1 = (EVTCHN_RING_SIZE - EVTCHN_RING_MASK(c)) *
++ sizeof(evtchn_port_t);
++ bytes2 = EVTCHN_RING_MASK(p) * sizeof(evtchn_port_t);
++ } else {
++ bytes1 = (p - c) * sizeof(evtchn_port_t);
++ bytes2 = 0;
++ }
+
-+/* blktap IOCTLs: */
-+#define BLKTAP_IOCTL_KICK_FE 1
-+#define BLKTAP_IOCTL_KICK_BE 2 /* currently unused */
-+#define BLKTAP_IOCTL_SETMODE 3
-+#define BLKTAP_IOCTL_SENDPID 4
-+#define BLKTAP_IOCTL_NEWINTF 5
-+#define BLKTAP_IOCTL_MINOR 6
-+#define BLKTAP_IOCTL_MAJOR 7
-+#define BLKTAP_QUERY_ALLOC_REQS 8
-+#define BLKTAP_IOCTL_FREEINTF 9
-+#define BLKTAP_IOCTL_PRINT_IDXS 100
++ /* Truncate chunks according to caller's maximum byte count. */
++ if (bytes1 > count) {
++ bytes1 = count;
++ bytes2 = 0;
++ } else if ((bytes1 + bytes2) > count) {
++ bytes2 = count - bytes1;
++ }
+
-+/* blktap switching modes: (Set with BLKTAP_IOCTL_SETMODE) */
-+#define BLKTAP_MODE_PASSTHROUGH 0x00000000 /* default */
-+#define BLKTAP_MODE_INTERCEPT_FE 0x00000001
-+#define BLKTAP_MODE_INTERCEPT_BE 0x00000002 /* unimp. */
++ rc = -EFAULT;
++ rmb(); /* Ensure that we see the port before we copy it. */
++ if (copy_to_user(buf, &u->ring[EVTCHN_RING_MASK(c)], bytes1) ||
++ ((bytes2 != 0) &&
++ copy_to_user(&buf[bytes1], &u->ring[0], bytes2)))
++ goto unlock_out;
++
++ evtchn_check_wrong_delivery(u);
+
-+#define BLKTAP_MODE_INTERPOSE \
-+ (BLKTAP_MODE_INTERCEPT_FE | BLKTAP_MODE_INTERCEPT_BE)
++ u->ring_cons += (bytes1 + bytes2) / sizeof(evtchn_port_t);
++ rc = bytes1 + bytes2;
+
++ unlock_out:
++ mutex_unlock(&u->ring_cons_mutex);
++ return rc;
++}
+
-+static inline int BLKTAP_MODE_VALID(unsigned long arg)
++static ssize_t evtchn_write(struct file *file, const char __user *buf,
++ size_t count, loff_t *ppos)
+{
-+ return ((arg == BLKTAP_MODE_PASSTHROUGH ) ||
-+ (arg == BLKTAP_MODE_INTERCEPT_FE) ||
-+ (arg == BLKTAP_MODE_INTERPOSE ));
-+}
++ int rc, i;
++ evtchn_port_t *kbuf = (evtchn_port_t *)__get_free_page(GFP_KERNEL);
++ struct per_user_data *u = file->private_data;
+
-+/* Requests passing through the tap to userspace are re-assigned an ID.
-+ * We must record a mapping between the BE [IDX,ID] tuple and the userspace
-+ * ring ID.
-+ */
++ if (kbuf == NULL)
++ return -ENOMEM;
+
-+static inline unsigned long MAKE_ID(domid_t fe_dom, PEND_RING_IDX idx)
-+{
-+ return ((fe_dom << 16) | MASK_PEND_IDX(idx));
++ /* Whole number of ports. */
++ count &= ~(sizeof(evtchn_port_t)-1);
++
++ rc = 0;
++ if (count == 0)
++ goto out;
++
++ if (count > PAGE_SIZE)
++ count = PAGE_SIZE;
++
++ rc = -EFAULT;
++ if (copy_from_user(kbuf, buf, count) != 0)
++ goto out;
++
++ spin_lock_irq(&port_user_lock);
++ for (i = 0; i < (count/sizeof(evtchn_port_t)); i++)
++ if ((kbuf[i] < NR_EVENT_CHANNELS) && (port_user[kbuf[i]] == u))
++ unmask_evtchn(kbuf[i]);
++ spin_unlock_irq(&port_user_lock);
++
++ rc = count;
++
++ out:
++ free_page((unsigned long)kbuf);
++ return rc;
+}
+
-+extern inline PEND_RING_IDX ID_TO_IDX(unsigned long id)
++static unsigned int next_bind_cpu(cpumask_t map)
+{
-+ return (PEND_RING_IDX)(id & 0x0000ffff);
++ static unsigned int bind_cpu;
++ bind_cpu = next_cpu(bind_cpu, map);
++ if (bind_cpu >= NR_CPUS)
++ bind_cpu = first_cpu(map);
++ return bind_cpu;
+}
+
-+extern inline int ID_TO_MIDX(unsigned long id)
++static void evtchn_bind_to_user(struct per_user_data *u, int port)
+{
-+ return (int)(id >> 16);
-+}
++ spin_lock_irq(&port_user_lock);
+
-+#define INVALID_REQ 0xdead0000
++ BUG_ON(port_user[port] != NULL);
++ port_user[port] = u;
+
-+/*TODO: Convert to a free list*/
-+static inline int GET_NEXT_REQ(unsigned long *idx_map)
-+{
-+ int i;
-+ for (i = 0; i < MAX_PENDING_REQS; i++)
-+ if (idx_map[i] == INVALID_REQ)
-+ return i;
++ if (u->bind_cpu == -1)
++ u->bind_cpu = next_bind_cpu(cpu_online_map);
+
-+ return INVALID_REQ;
-+}
++ rebind_evtchn_to_cpu(port, u->bind_cpu);
+
-+static inline int OFFSET_TO_USR_IDX(int offset)
-+{
-+ return offset / BLKIF_MAX_SEGMENTS_PER_REQUEST;
++ unmask_evtchn(port);
++
++ spin_unlock_irq(&port_user_lock);
+}
+
-+static inline int OFFSET_TO_SEG(int offset)
++static long evtchn_ioctl(struct file *file,
++ unsigned int cmd, unsigned long arg)
+{
-+ return offset % BLKIF_MAX_SEGMENTS_PER_REQUEST;
-+}
++ int rc;
++ struct per_user_data *u = file->private_data;
++ void __user *uarg = (void __user *) arg;
+
++ switch (cmd) {
++ case IOCTL_EVTCHN_BIND_VIRQ: {
++ struct ioctl_evtchn_bind_virq bind;
++ struct evtchn_bind_virq bind_virq;
+
-+#define BLKTAP_INVALID_HANDLE(_g) \
-+ (((_g->kernel) == INVALID_GRANT_HANDLE) && \
-+ ((_g->user) == INVALID_GRANT_HANDLE))
++ rc = -EFAULT;
++ if (copy_from_user(&bind, uarg, sizeof(bind)))
++ break;
+
-+#define BLKTAP_INVALIDATE_HANDLE(_g) do { \
-+ (_g)->kernel = INVALID_GRANT_HANDLE; (_g)->user = INVALID_GRANT_HANDLE; \
-+ } while(0)
++ bind_virq.virq = bind.virq;
++ bind_virq.vcpu = 0;
++ rc = HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
++ &bind_virq);
++ if (rc != 0)
++ break;
+
++ rc = bind_virq.port;
++ evtchn_bind_to_user(u, rc);
++ break;
++ }
+
-+/******************************************************************
-+ * BLKTAP VM OPS
-+ */
++ case IOCTL_EVTCHN_BIND_INTERDOMAIN: {
++ struct ioctl_evtchn_bind_interdomain bind;
++ struct evtchn_bind_interdomain bind_interdomain;
+
-+static struct page *blktap_nopage(struct vm_area_struct *vma,
-+ unsigned long address,
-+ int *type)
-+{
-+ /*
-+ * if the page has not been mapped in by the driver then return
-+ * NOPAGE_SIGBUS to the domain.
-+ */
++ rc = -EFAULT;
++ if (copy_from_user(&bind, uarg, sizeof(bind)))
++ break;
+
-+ return NOPAGE_SIGBUS;
-+}
++ bind_interdomain.remote_dom = bind.remote_domain;
++ bind_interdomain.remote_port = bind.remote_port;
++ rc = HYPERVISOR_event_channel_op(EVTCHNOP_bind_interdomain,
++ &bind_interdomain);
++ if (rc != 0)
++ break;
+
-+static pte_t blktap_clear_pte(struct vm_area_struct *vma,
-+ unsigned long uvaddr,
-+ pte_t *ptep, int is_fullmm)
-+{
-+ pte_t copy;
-+ tap_blkif_t *info;
-+ int offset, seg, usr_idx, pending_idx, mmap_idx;
-+ unsigned long uvstart = vma->vm_start + (RING_PAGES << PAGE_SHIFT);
-+ unsigned long kvaddr;
-+ struct page **map;
-+ struct page *pg;
-+ struct grant_handle_pair *khandle;
-+ struct gnttab_unmap_grant_ref unmap[2];
-+ int count = 0;
++ rc = bind_interdomain.local_port;
++ evtchn_bind_to_user(u, rc);
++ break;
++ }
+
-+ /*
-+ * If the address is before the start of the grant mapped region or
-+ * if vm_file is NULL (meaning mmap failed and we have nothing to do)
-+ */
-+ if (uvaddr < uvstart || vma->vm_file == NULL)
-+ return ptep_get_and_clear_full(vma->vm_mm, uvaddr,
-+ ptep, is_fullmm);
++ case IOCTL_EVTCHN_BIND_UNBOUND_PORT: {
++ struct ioctl_evtchn_bind_unbound_port bind;
++ struct evtchn_alloc_unbound alloc_unbound;
+
-+ info = vma->vm_file->private_data;
-+ map = vma->vm_private_data;
++ rc = -EFAULT;
++ if (copy_from_user(&bind, uarg, sizeof(bind)))
++ break;
+
-+ /* TODO Should these be changed to if statements? */
-+ BUG_ON(!info);
-+ BUG_ON(!info->idx_map);
-+ BUG_ON(!map);
++ alloc_unbound.dom = DOMID_SELF;
++ alloc_unbound.remote_dom = bind.remote_domain;
++ rc = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
++ &alloc_unbound);
++ if (rc != 0)
++ break;
+
-+ offset = (int) ((uvaddr - uvstart) >> PAGE_SHIFT);
-+ usr_idx = OFFSET_TO_USR_IDX(offset);
-+ seg = OFFSET_TO_SEG(offset);
++ rc = alloc_unbound.port;
++ evtchn_bind_to_user(u, rc);
++ break;
++ }
++
++ case IOCTL_EVTCHN_UNBIND: {
++ struct ioctl_evtchn_unbind unbind;
++ struct evtchn_close close;
++ int ret;
++
++ rc = -EFAULT;
++ if (copy_from_user(&unbind, uarg, sizeof(unbind)))
++ break;
++
++ rc = -EINVAL;
++ if (unbind.port >= NR_EVENT_CHANNELS)
++ break;
+
-+ pending_idx = MASK_PEND_IDX(ID_TO_IDX(info->idx_map[usr_idx]));
-+ mmap_idx = ID_TO_MIDX(info->idx_map[usr_idx]);
++ spin_lock_irq(&port_user_lock);
++
++ rc = -ENOTCONN;
++ if (port_user[unbind.port] != u) {
++ spin_unlock_irq(&port_user_lock);
++ break;
++ }
+
-+ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, seg);
-+ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
-+ ClearPageReserved(pg);
-+ map[offset + RING_PAGES] = NULL;
++ port_user[unbind.port] = NULL;
++ mask_evtchn(unbind.port);
++ rebind_evtchn_to_cpu(unbind.port, 0);
+
-+ khandle = &pending_handle(mmap_idx, pending_idx, seg);
++ spin_unlock_irq(&port_user_lock);
+
-+ if (khandle->kernel != INVALID_GRANT_HANDLE) {
-+ gnttab_set_unmap_op(&unmap[count], kvaddr,
-+ GNTMAP_host_map, khandle->kernel);
-+ count++;
++ close.port = unbind.port;
++ ret = HYPERVISOR_event_channel_op(EVTCHNOP_close, &close);
++ BUG_ON(ret);
+
-+ set_phys_to_machine(__pa(kvaddr) >> PAGE_SHIFT,
-+ INVALID_P2M_ENTRY);
++ rc = 0;
++ break;
+ }
+
-+ if (khandle->user != INVALID_GRANT_HANDLE) {
-+ BUG_ON(xen_feature(XENFEAT_auto_translated_physmap));
++ case IOCTL_EVTCHN_NOTIFY: {
++ struct ioctl_evtchn_notify notify;
+
-+ copy = *ptep;
-+ gnttab_set_unmap_op(&unmap[count], virt_to_machine(ptep),
-+ GNTMAP_host_map
-+ | GNTMAP_application_map
-+ | GNTMAP_contains_pte,
-+ khandle->user);
-+ count++;
-+ } else {
-+ BUG_ON(!xen_feature(XENFEAT_auto_translated_physmap));
++ rc = -EFAULT;
++ if (copy_from_user(¬ify, uarg, sizeof(notify)))
++ break;
+
-+ /* USING SHADOW PAGE TABLES. */
-+ copy = ptep_get_and_clear_full(vma->vm_mm, uvaddr, ptep,
-+ is_fullmm);
++ if (notify.port >= NR_EVENT_CHANNELS) {
++ rc = -EINVAL;
++ } else if (port_user[notify.port] != u) {
++ rc = -ENOTCONN;
++ } else {
++ notify_remote_via_evtchn(notify.port);
++ rc = 0;
++ }
++ break;
+ }
+
-+ if (count) {
-+ BLKTAP_INVALIDATE_HANDLE(khandle);
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref,
-+ unmap, count))
-+ BUG();
++ case IOCTL_EVTCHN_RESET: {
++ /* Initialise the ring to empty. Clear errors. */
++ mutex_lock(&u->ring_cons_mutex);
++ spin_lock_irq(&port_user_lock);
++ u->ring_cons = u->ring_prod = u->ring_overflow = 0;
++ spin_unlock_irq(&port_user_lock);
++ mutex_unlock(&u->ring_cons_mutex);
++ rc = 0;
++ break;
+ }
+
-+ return copy;
-+}
-+
-+struct vm_operations_struct blktap_vm_ops = {
-+ nopage: blktap_nopage,
-+ zap_pte: blktap_clear_pte,
-+};
++ default:
++ rc = -ENOSYS;
++ break;
++ }
+
-+/******************************************************************
-+ * BLKTAP FILE OPS
-+ */
-+
-+/*Function Declarations*/
-+static tap_blkif_t *get_next_free_dev(void);
-+static int blktap_open(struct inode *inode, struct file *filp);
-+static int blktap_release(struct inode *inode, struct file *filp);
-+static int blktap_mmap(struct file *filp, struct vm_area_struct *vma);
-+static int blktap_ioctl(struct inode *inode, struct file *filp,
-+ unsigned int cmd, unsigned long arg);
-+static unsigned int blktap_poll(struct file *file, poll_table *wait);
++ return rc;
++}
+
-+static const struct file_operations blktap_fops = {
-+ .owner = THIS_MODULE,
-+ .poll = blktap_poll,
-+ .ioctl = blktap_ioctl,
-+ .open = blktap_open,
-+ .release = blktap_release,
-+ .mmap = blktap_mmap,
-+};
++static unsigned int evtchn_poll(struct file *file, poll_table *wait)
++{
++ unsigned int mask = POLLOUT | POLLWRNORM;
++ struct per_user_data *u = file->private_data;
+
++ poll_wait(file, &u->evtchn_wait, wait);
++ if (u->ring_cons != u->ring_prod)
++ mask |= POLLIN | POLLRDNORM;
++ if (u->ring_overflow)
++ mask = POLLERR;
++ return mask;
++}
+
-+static tap_blkif_t *get_next_free_dev(void)
++static int evtchn_fasync(int fd, struct file *filp, int on)
+{
-+ struct class *class;
-+ tap_blkif_t *info;
-+ int minor;
++ struct per_user_data *u = filp->private_data;
++ return fasync_helper(fd, filp, on, &u->evtchn_async_queue);
++}
+
-+ /*
-+ * This is called only from the ioctl, which
-+ * means we should always have interrupts enabled.
-+ */
-+ BUG_ON(irqs_disabled());
++static int evtchn_open(struct inode *inode, struct file *filp)
++{
++ struct per_user_data *u;
+
-+ spin_lock_irq(&pending_free_lock);
++ if ((u = kmalloc(sizeof(*u), GFP_KERNEL)) == NULL)
++ return -ENOMEM;
+
-+ /* tapfds[0] is always NULL */
++ memset(u, 0, sizeof(*u));
++ init_waitqueue_head(&u->evtchn_wait);
+
-+ for (minor = 1; minor < blktap_next_minor; minor++) {
-+ info = tapfds[minor];
-+ /* we could have failed a previous attempt. */
-+ if (!info ||
-+ ((info->dev_inuse == 0) &&
-+ (info->dev_pending == 0)) ) {
-+ info->dev_pending = 1;
-+ goto found;
-+ }
++ u->ring = (evtchn_port_t *)__get_free_page(GFP_KERNEL);
++ if (u->ring == NULL) {
++ kfree(u);
++ return -ENOMEM;
+ }
-+ info = NULL;
-+ minor = -1;
-+
-+ /*
-+ * We didn't find free device. If we can still allocate
-+ * more, then we grab the next device minor that is
-+ * available. This is done while we are still under
-+ * the protection of the pending_free_lock.
-+ */
-+ if (blktap_next_minor < MAX_TAP_DEV)
-+ minor = blktap_next_minor++;
-+found:
-+ spin_unlock_irq(&pending_free_lock);
+
-+ if (!info && minor > 0) {
-+ info = kzalloc(sizeof(*info), GFP_KERNEL);
-+ if (unlikely(!info)) {
-+ /*
-+ * If we failed here, try to put back
-+ * the next minor number. But if one
-+ * was just taken, then we just lose this
-+ * minor. We can try to allocate this
-+ * minor again later.
-+ */
-+ spin_lock_irq(&pending_free_lock);
-+ if (blktap_next_minor == minor+1)
-+ blktap_next_minor--;
-+ spin_unlock_irq(&pending_free_lock);
-+ goto out;
-+ }
++ mutex_init(&u->ring_cons_mutex);
+
-+ info->minor = minor;
-+ /*
-+ * Make sure that we have a minor before others can
-+ * see us.
-+ */
-+ wmb();
-+ tapfds[minor] = info;
++ filp->private_data = u;
+
-+ if ((class = get_xen_class()) != NULL)
-+ class_device_create(class, NULL,
-+ MKDEV(blktap_major, minor), NULL,
-+ "blktap%d", minor);
-+ }
++ u->bind_cpu = -1;
+
-+out:
-+ return info;
++ return 0;
+}
+
-+int dom_to_devid(domid_t domid, int xenbus_id, blkif_t *blkif)
++static int evtchn_release(struct inode *inode, struct file *filp)
+{
-+ tap_blkif_t *info;
+ int i;
++ struct per_user_data *u = filp->private_data;
++ struct evtchn_close close;
+
-+ for (i = 1; i < blktap_next_minor; i++) {
-+ info = tapfds[i];
-+ if ( info &&
-+ (info->trans.domid == domid) &&
-+ (info->trans.busid == xenbus_id) ) {
-+ info->blkif = blkif;
-+ info->status = RUNNING;
-+ return i;
-+ }
-+ }
-+ return -1;
-+}
++ spin_lock_irq(&port_user_lock);
+
-+void signal_tapdisk(int idx)
-+{
-+ tap_blkif_t *info;
-+ struct task_struct *ptask;
++ free_page((unsigned long)u->ring);
+
-+ info = tapfds[idx];
-+ if ((idx < 0) || (idx > MAX_TAP_DEV) || !info)
-+ return;
++ for (i = 0; i < NR_EVENT_CHANNELS; i++) {
++ int ret;
++ if (port_user[i] != u)
++ continue;
+
-+ if (info->pid > 0) {
-+ ptask = find_task_by_pid(info->pid);
-+ if (ptask)
-+ info->status = CLEANSHUTDOWN;
++ port_user[i] = NULL;
++ mask_evtchn(i);
++ rebind_evtchn_to_cpu(i, 0);
++
++ close.port = i;
++ ret = HYPERVISOR_event_channel_op(EVTCHNOP_close, &close);
++ BUG_ON(ret);
+ }
-+ info->blkif = NULL;
+
-+ return;
++ spin_unlock_irq(&port_user_lock);
++
++ kfree(u);
++
++ return 0;
+}
+
-+static int blktap_open(struct inode *inode, struct file *filp)
-+{
-+ blkif_sring_t *sring;
-+ int idx = iminor(inode) - BLKTAP_MINOR;
-+ tap_blkif_t *info;
-+ int i;
-+
-+ /* ctrl device, treat differently */
-+ if (!idx)
-+ return 0;
++static const struct file_operations evtchn_fops = {
++ .owner = THIS_MODULE,
++ .read = evtchn_read,
++ .write = evtchn_write,
++ .unlocked_ioctl = evtchn_ioctl,
++ .poll = evtchn_poll,
++ .fasync = evtchn_fasync,
++ .open = evtchn_open,
++ .release = evtchn_release,
++};
+
-+ info = tapfds[idx];
++static struct miscdevice evtchn_miscdev = {
++ .minor = MISC_DYNAMIC_MINOR,
++ .name = "evtchn",
++ .fops = &evtchn_fops,
++};
+
-+ if ((idx < 0) || (idx > MAX_TAP_DEV) || !info) {
-+ WPRINTK("Unable to open device /dev/xen/blktap%d\n",
-+ idx);
-+ return -ENODEV;
++static int __cpuinit evtchn_cpu_notify(struct notifier_block *nfb,
++ unsigned long action, void *hcpu)
++{
++ int hotcpu = (unsigned long)hcpu;
++ cpumask_t map = cpu_online_map;
++ int port, newcpu;
++ struct per_user_data *u;
++
++ switch (action) {
++ case CPU_DOWN_PREPARE:
++ cpu_clear(hotcpu, map);
++ spin_lock_irq(&port_user_lock);
++ for (port = 0; port < NR_EVENT_CHANNELS; port++) {
++ if ((u = port_user[port]) != NULL &&
++ u->bind_cpu == hotcpu &&
++ (newcpu = next_bind_cpu(map)) < NR_CPUS) {
++ rebind_evtchn_to_cpu(port, newcpu);
++ u->bind_cpu = newcpu;
++ }
++ }
++ spin_unlock_irq(&port_user_lock);
++ break;
++ default:
++ return NOTIFY_DONE;
+ }
++ return NOTIFY_OK;
++}
+
-+ DPRINTK("Opening device /dev/xen/blktap%d\n",idx);
-+
-+ /*Only one process can access device at a time*/
-+ if (test_and_set_bit(0, &info->dev_inuse))
-+ return -EBUSY;
++static struct notifier_block __cpuinitdata evtchn_cpu_nfb = {
++ .notifier_call = evtchn_cpu_notify
++};
+
-+ info->dev_pending = 0;
-+
-+ /* Allocate the fe ring. */
-+ sring = (blkif_sring_t *)get_zeroed_page(GFP_KERNEL);
-+ if (sring == NULL)
-+ goto fail_nomem;
++static int __init evtchn_init(void)
++{
++ int err;
+
-+ SetPageReserved(virt_to_page(sring));
-+
-+ SHARED_RING_INIT(sring);
-+ FRONT_RING_INIT(&info->ufe_ring, sring, PAGE_SIZE);
-+
-+ filp->private_data = info;
-+ info->vma = NULL;
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+ info->idx_map = kmalloc(sizeof(unsigned long) * MAX_PENDING_REQS,
-+ GFP_KERNEL);
-+
-+ if (info->idx_map == NULL)
-+ goto fail_nomem;
++ spin_lock_init(&port_user_lock);
++ memset(port_user, 0, sizeof(port_user));
+
-+ if (idx > 0) {
-+ init_waitqueue_head(&info->wait);
-+ for (i = 0; i < MAX_PENDING_REQS; i++)
-+ info->idx_map[i] = INVALID_REQ;
++ /* Create '/dev/misc/evtchn'. */
++ err = misc_register(&evtchn_miscdev);
++ if (err != 0) {
++ printk(KERN_ALERT "Could not register /dev/misc/evtchn\n");
++ return err;
+ }
+
-+ DPRINTK("Tap open: device /dev/xen/blktap%d\n",idx);
-+ return 0;
++ register_cpu_notifier(&evtchn_cpu_nfb);
+
-+ fail_nomem:
-+ return -ENOMEM;
++ printk("Event-channel device installed.\n");
++
++ return 0;
+}
+
-+static int blktap_release(struct inode *inode, struct file *filp)
++static void __exit evtchn_cleanup(void)
+{
-+ tap_blkif_t *info = filp->private_data;
-+
-+ /* check for control device */
-+ if (!info)
-+ return 0;
++ misc_deregister(&evtchn_miscdev);
++ unregister_cpu_notifier(&evtchn_cpu_nfb);
++}
+
-+ info->dev_inuse = 0;
-+ DPRINTK("Freeing device [/dev/xen/blktap%d]\n",info->minor);
++module_init(evtchn_init);
++module_exit(evtchn_cleanup);
+
-+ /* Free the ring page. */
-+ ClearPageReserved(virt_to_page(info->ufe_ring.sring));
-+ free_page((unsigned long) info->ufe_ring.sring);
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/fbfront/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/fbfront/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,2 @@
++obj-$(CONFIG_XEN_FRAMEBUFFER) := xenfb.o
++obj-$(CONFIG_XEN_KEYBOARD) += xenkbd.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/fbfront/xenfb.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/fbfront/xenfb.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,887 @@
++/*
++ * linux/drivers/video/xenfb.c -- Xen para-virtual frame buffer device
++ *
++ * Copyright (C) 2005-2006 Anthony Liguori <aliguori at us.ibm.com>
++ * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster <armbru at redhat.com>
++ *
++ * Based on linux/drivers/video/q40fb.c
++ *
++ * This file is subject to the terms and conditions of the GNU General Public
++ * License. See the file COPYING in the main directory of this archive for
++ * more details.
++ */
+
-+ /* Clear any active mappings and free foreign map table */
-+ if (info->vma) {
-+ zap_page_range(
-+ info->vma, info->vma->vm_start,
-+ info->vma->vm_end - info->vma->vm_start, NULL);
++/*
++ * TODO:
++ *
++ * Switch to grant tables when they become capable of dealing with the
++ * frame buffer.
++ */
+
-+ kfree(info->vma->vm_private_data);
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/fb.h>
++#include <linux/module.h>
++#include <linux/vmalloc.h>
++#include <linux/mm.h>
++#include <linux/mutex.h>
++#include <asm/hypervisor.h>
++#include <xen/evtchn.h>
++#include <xen/interface/io/fbif.h>
++#include <xen/interface/io/protocols.h>
++#include <xen/xenbus.h>
++#include <linux/kthread.h>
+
-+ info->vma = NULL;
-+ }
++struct xenfb_mapping
++{
++ struct list_head link;
++ struct vm_area_struct *vma;
++ atomic_t map_refs;
++ int faults;
++ struct xenfb_info *info;
++};
+
-+ if (info->idx_map) {
-+ kfree(info->idx_map);
-+ info->idx_map = NULL;
-+ }
++struct xenfb_info
++{
++ struct task_struct *kthread;
++ wait_queue_head_t wq;
+
-+ if ( (info->status != CLEANSHUTDOWN) && (info->blkif != NULL) ) {
-+ if (info->blkif->xenblkd != NULL) {
-+ kthread_stop(info->blkif->xenblkd);
-+ info->blkif->xenblkd = NULL;
-+ }
-+ info->status = CLEANSHUTDOWN;
-+ }
++ unsigned char *fb;
++ struct fb_info *fb_info;
++ struct timer_list refresh;
++ int dirty;
++ int x1, y1, x2, y2; /* dirty rectangle,
++ protected by dirty_lock */
++ spinlock_t dirty_lock;
++ struct mutex mm_lock;
++ int nr_pages;
++ struct page **pages;
++ struct list_head mappings; /* protected by mm_lock */
+
-+ return 0;
-+}
++ int irq;
++ struct xenfb_page *page;
++ unsigned long *mfns;
++ int update_wanted; /* XENFB_TYPE_UPDATE wanted */
++ int feature_resize; /* Backend has resize feature */
++ struct xenfb_resize resize;
++ int resize_dpy;
++ spinlock_t resize_lock;
+
++ struct xenbus_device *xbdev;
++};
+
-+/* Note on mmap:
-+ * We need to map pages to user space in a way that will allow the block
-+ * subsystem set up direct IO to them. This couldn't be done before, because
-+ * there isn't really a sane way to translate a user virtual address down to a
-+ * physical address when the page belongs to another domain.
++/*
++ * There are three locks:
++ * spinlock resize_lock protecting resize_dpy and resize
++ * spinlock dirty_lock protecting the dirty rectangle
++ * mutex mm_lock protecting mappings.
+ *
-+ * My first approach was to map the page in to kernel memory, add an entry
-+ * for it in the physical frame list (using alloc_lomem_region as in blkback)
-+ * and then attempt to map that page up to user space. This is disallowed
-+ * by xen though, which realizes that we don't really own the machine frame
-+ * underlying the physical page.
++ * How the dirty and mapping locks work together
+ *
-+ * The new approach is to provide explicit support for this in xen linux.
-+ * The VMA now has a flag, VM_FOREIGN, to indicate that it contains pages
-+ * mapped from other vms. vma->vm_private_data is set up as a mapping
-+ * from pages to actual page structs. There is a new clause in get_user_pages
-+ * that does the right thing for this sort of mapping.
++ * The problem is that dirty rectangle and mappings aren't
++ * independent: the dirty rectangle must cover all faulted pages in
++ * mappings. We need to prove that our locking maintains this
++ * invariant.
++ *
++ * There are several kinds of critical regions:
++ *
++ * 1. Holding only dirty_lock: xenfb_refresh(). May run in
++ * interrupts. Extends the dirty rectangle. Trivially preserves
++ * invariant.
++ *
++ * 2. Holding only mm_lock: xenfb_mmap() and xenfb_vm_close(). Touch
++ * only mappings. The former creates unfaulted pages. Preserves
++ * invariant. The latter removes pages. Preserves invariant.
++ *
++ * 3. Holding both locks: xenfb_vm_nopage(). Extends the dirty
++ * rectangle and updates mappings consistently. Preserves
++ * invariant.
++ *
++ * 4. The ugliest one: xenfb_update_screen(). Clear the dirty
++ * rectangle and update mappings consistently.
++ *
++ * We can't simply hold both locks, because zap_page_range() cannot
++ * be called with a spinlock held.
++ *
++ * Therefore, we first clear the dirty rectangle with both locks
++ * held. Then we unlock dirty_lock and update the mappings.
++ * Critical regions that hold only dirty_lock may interfere with
++ * that. This can only be region 1: xenfb_refresh(). But that
++ * just extends the dirty rectangle, which can't harm the
++ * invariant.
++ *
++ * But FIXME: the invariant is too weak. It misses that the fault
++ * record in mappings must be consistent with the mapping of pages in
++ * the associated address space! do_no_page() updates the PTE after
++ * xenfb_vm_nopage() returns, i.e. outside the critical region. This
++ * allows the following race:
++ *
++ * X writes to some address in the Xen frame buffer
++ * Fault - call do_no_page()
++ * call xenfb_vm_nopage()
++ * grab mm_lock
++ * map->faults++;
++ * release mm_lock
++ * return back to do_no_page()
++ * (preempted, or SMP)
++ * Xen worker thread runs.
++ * grab mm_lock
++ * look at mappings
++ * find this mapping, zaps its pages (but page not in pte yet)
++ * clear map->faults
++ * releases mm_lock
++ * (back to X process)
++ * put page in X's pte
++ *
++ * Oh well, we wont be updating the writes to this page anytime soon.
+ */
-+static int blktap_mmap(struct file *filp, struct vm_area_struct *vma)
-+{
-+ int size;
-+ struct page **map;
-+ int i;
-+ tap_blkif_t *info = filp->private_data;
++#define MB_ (1024*1024)
++#define XENFB_DEFAULT_FB_LEN (XENFB_WIDTH * XENFB_HEIGHT * XENFB_DEPTH / 8)
++
++enum {KPARAM_MEM, KPARAM_WIDTH, KPARAM_HEIGHT, KPARAM_CNT};
++static int video[KPARAM_CNT] = {2, XENFB_WIDTH, XENFB_HEIGHT};
++module_param_array(video, int, NULL, 0);
++MODULE_PARM_DESC(video,
++ "Size of video memory in MB and width,height in pixels, default = (2,800,600)");
+
-+ if (info == NULL) {
-+ WPRINTK("blktap: mmap, retrieving idx failed\n");
-+ return -ENOMEM;
-+ }
-+
-+ vma->vm_flags |= VM_RESERVED;
-+ vma->vm_ops = &blktap_vm_ops;
++static int xenfb_fps = 20;
+
-+ size = vma->vm_end - vma->vm_start;
-+ if (size != ((mmap_pages + RING_PAGES) << PAGE_SHIFT)) {
-+ WPRINTK("you _must_ map exactly %d pages!\n",
-+ mmap_pages + RING_PAGES);
-+ return -EAGAIN;
-+ }
++static int xenfb_remove(struct xenbus_device *);
++static void xenfb_init_shared_page(struct xenfb_info *, struct fb_info *);
++static int xenfb_connect_backend(struct xenbus_device *, struct xenfb_info *);
++static void xenfb_disconnect_backend(struct xenfb_info *);
+
-+ size >>= PAGE_SHIFT;
-+ info->rings_vstart = vma->vm_start;
-+ info->user_vstart = info->rings_vstart + (RING_PAGES << PAGE_SHIFT);
-+
-+ /* Map the ring pages to the start of the region and reserve it. */
-+ if (remap_pfn_range(vma, vma->vm_start,
-+ __pa(info->ufe_ring.sring) >> PAGE_SHIFT,
-+ PAGE_SIZE, vma->vm_page_prot)) {
-+ WPRINTK("Mapping user ring failed!\n");
-+ goto fail;
-+ }
++static void xenfb_send_event(struct xenfb_info *info,
++ union xenfb_out_event *event)
++{
++ __u32 prod;
+
-+ /* Mark this VM as containing foreign pages, and set up mappings. */
-+ map = kzalloc(((vma->vm_end - vma->vm_start) >> PAGE_SHIFT)
-+ * sizeof(struct page_struct*),
-+ GFP_KERNEL);
-+ if (map == NULL) {
-+ WPRINTK("Couldn't alloc VM_FOREIGN map.\n");
-+ goto fail;
-+ }
++ prod = info->page->out_prod;
++ /* caller ensures !xenfb_queue_full() */
++ mb(); /* ensure ring space available */
++ XENFB_OUT_RING_REF(info->page, prod) = *event;
++ wmb(); /* ensure ring contents visible */
++ info->page->out_prod = prod + 1;
+
-+ for (i = 0; i < ((vma->vm_end - vma->vm_start) >> PAGE_SHIFT); i++)
-+ map[i] = NULL;
-+
-+ vma->vm_private_data = map;
-+ vma->vm_flags |= VM_FOREIGN;
-+ vma->vm_flags |= VM_DONTCOPY;
++ notify_remote_via_irq(info->irq);
++}
+
-+#ifdef CONFIG_X86
-+ vma->vm_mm->context.has_foreign_mappings = 1;
-+#endif
++static void xenfb_do_update(struct xenfb_info *info,
++ int x, int y, int w, int h)
++{
++ union xenfb_out_event event;
+
-+ info->vma = vma;
-+ info->ring_ok = 1;
-+ return 0;
-+ fail:
-+ /* Clear any active mappings. */
-+ zap_page_range(vma, vma->vm_start,
-+ vma->vm_end - vma->vm_start, NULL);
++ memset(&event, 0, sizeof(event));
++ event.type = XENFB_TYPE_UPDATE;
++ event.update.x = x;
++ event.update.y = y;
++ event.update.width = w;
++ event.update.height = h;
+
-+ return -ENOMEM;
++ /* caller ensures !xenfb_queue_full() */
++ xenfb_send_event(info, &event);
+}
+
-+
-+static int blktap_ioctl(struct inode *inode, struct file *filp,
-+ unsigned int cmd, unsigned long arg)
++static void xenfb_do_resize(struct xenfb_info *info)
+{
-+ tap_blkif_t *info = filp->private_data;
++ union xenfb_out_event event;
+
-+ switch(cmd) {
-+ case BLKTAP_IOCTL_KICK_FE:
-+ {
-+ /* There are fe messages to process. */
-+ return blktap_read_ufe_ring(info);
-+ }
-+ case BLKTAP_IOCTL_SETMODE:
-+ {
-+ if (info) {
-+ if (BLKTAP_MODE_VALID(arg)) {
-+ info->mode = arg;
-+ /* XXX: may need to flush rings here. */
-+ DPRINTK("blktap: set mode to %lx\n",
-+ arg);
-+ return 0;
-+ }
-+ }
-+ return 0;
-+ }
-+ case BLKTAP_IOCTL_PRINT_IDXS:
-+ {
-+ if (info) {
-+ printk("User Rings: \n-----------\n");
-+ printk("UF: rsp_cons: %2d, req_prod_prv: %2d "
-+ "| req_prod: %2d, rsp_prod: %2d\n",
-+ info->ufe_ring.rsp_cons,
-+ info->ufe_ring.req_prod_pvt,
-+ info->ufe_ring.sring->req_prod,
-+ info->ufe_ring.sring->rsp_prod);
-+ }
-+ return 0;
-+ }
-+ case BLKTAP_IOCTL_SENDPID:
-+ {
-+ if (info) {
-+ info->pid = (pid_t)arg;
-+ DPRINTK("blktap: pid received %d\n",
-+ info->pid);
-+ }
-+ return 0;
-+ }
-+ case BLKTAP_IOCTL_NEWINTF:
-+ {
-+ uint64_t val = (uint64_t)arg;
-+ domid_translate_t *tr = (domid_translate_t *)&val;
++ memset(&event, 0, sizeof(event));
++ event.resize = info->resize;
+
-+ DPRINTK("NEWINTF Req for domid %d and bus id %d\n",
-+ tr->domid, tr->busid);
-+ info = get_next_free_dev();
-+ if (!info) {
-+ WPRINTK("Error initialising /dev/xen/blktap - "
-+ "No more devices\n");
-+ return -1;
-+ }
-+ info->trans.domid = tr->domid;
-+ info->trans.busid = tr->busid;
-+ return info->minor;
-+ }
-+ case BLKTAP_IOCTL_FREEINTF:
-+ {
-+ unsigned long dev = arg;
-+ unsigned long flags;
++ /* caller ensures !xenfb_queue_full() */
++ xenfb_send_event(info, &event);
++}
+
-+ info = tapfds[dev];
++static int xenfb_queue_full(struct xenfb_info *info)
++{
++ __u32 cons, prod;
+
-+ if ((dev > MAX_TAP_DEV) || !info)
-+ return 0; /* should this be an error? */
++ prod = info->page->out_prod;
++ cons = info->page->out_cons;
++ return prod - cons == XENFB_OUT_RING_LEN;
++}
+
-+ spin_lock_irqsave(&pending_free_lock, flags);
-+ if (info->dev_pending)
-+ info->dev_pending = 0;
-+ spin_unlock_irqrestore(&pending_free_lock, flags);
++static void xenfb_update_screen(struct xenfb_info *info)
++{
++ unsigned long flags;
++ int y1, y2, x1, x2;
++ struct xenfb_mapping *map;
+
-+ return 0;
-+ }
-+ case BLKTAP_IOCTL_MINOR:
-+ {
-+ unsigned long dev = arg;
++ if (!info->update_wanted)
++ return;
++ if (xenfb_queue_full(info))
++ return;
+
-+ info = tapfds[dev];
++ mutex_lock(&info->mm_lock);
+
-+ if ((dev > MAX_TAP_DEV) || !info)
-+ return -EINVAL;
++ spin_lock_irqsave(&info->dirty_lock, flags);
++ y1 = info->y1;
++ y2 = info->y2;
++ x1 = info->x1;
++ x2 = info->x2;
++ info->x1 = info->y1 = INT_MAX;
++ info->x2 = info->y2 = 0;
++ spin_unlock_irqrestore(&info->dirty_lock, flags);
+
-+ return info->minor;
++ list_for_each_entry(map, &info->mappings, link) {
++ if (!map->faults)
++ continue;
++ zap_page_range(map->vma, map->vma->vm_start,
++ map->vma->vm_end - map->vma->vm_start, NULL);
++ map->faults = 0;
+ }
-+ case BLKTAP_IOCTL_MAJOR:
-+ return blktap_major;
+
-+ case BLKTAP_QUERY_ALLOC_REQS:
-+ {
-+ WPRINTK("BLKTAP_QUERY_ALLOC_REQS ioctl: %d/%d\n",
-+ alloc_pending_reqs, blkif_reqs);
-+ return (alloc_pending_reqs/blkif_reqs) * 100;
-+ }
++ mutex_unlock(&info->mm_lock);
++
++ if (x2 < x1 || y2 < y1) {
++ printk("xenfb_update_screen bogus rect %d %d %d %d\n",
++ x1, x2, y1, y2);
++ WARN_ON(1);
+ }
-+ return -ENOIOCTLCMD;
++ xenfb_do_update(info, x1, y1, x2 - x1, y2 - y1);
+}
+
-+static unsigned int blktap_poll(struct file *filp, poll_table *wait)
++static void xenfb_handle_resize_dpy(struct xenfb_info *info)
+{
-+ tap_blkif_t *info = filp->private_data;
-+
-+ /* do not work on the control device */
-+ if (!info)
-+ return 0;
++ unsigned long flags;
+
-+ poll_wait(filp, &info->wait, wait);
-+ if (info->ufe_ring.req_prod_pvt != info->ufe_ring.sring->req_prod) {
-+ RING_PUSH_REQUESTS(&info->ufe_ring);
-+ return POLLIN | POLLRDNORM;
++ spin_lock_irqsave(&info->resize_lock, flags);
++ if (info->resize_dpy) {
++ if (!xenfb_queue_full(info)) {
++ info->resize_dpy = 0;
++ xenfb_do_resize(info);
++ }
+ }
-+ return 0;
++ spin_unlock_irqrestore(&info->resize_lock, flags);
+}
+
-+void blktap_kick_user(int idx)
++static int xenfb_thread(void *data)
+{
-+ tap_blkif_t *info;
-+
-+ info = tapfds[idx];
-+
-+ if ((idx < 0) || (idx > MAX_TAP_DEV) || !info)
-+ return;
-+
-+ wake_up_interruptible(&info->wait);
++ struct xenfb_info *info = data;
+
-+ return;
++ while (!kthread_should_stop()) {
++ xenfb_handle_resize_dpy(info);
++ if (info->dirty) {
++ info->dirty = 0;
++ xenfb_update_screen(info);
++ }
++ wait_event_interruptible(info->wq,
++ kthread_should_stop() || info->dirty);
++ try_to_freeze();
++ }
++ return 0;
+}
+
-+static int do_block_io_op(blkif_t *blkif);
-+static void dispatch_rw_block_io(blkif_t *blkif,
-+ blkif_request_t *req,
-+ pending_req_t *pending_req);
-+static void make_response(blkif_t *blkif, u64 id,
-+ unsigned short op, int st);
-+
-+/******************************************************************
-+ * misc small helpers
-+ */
-+static int req_increase(void)
++static int xenfb_setcolreg(unsigned regno, unsigned red, unsigned green,
++ unsigned blue, unsigned transp,
++ struct fb_info *info)
+{
-+ int i, j;
-+
-+ if (mmap_alloc >= MAX_PENDING_REQS || mmap_lock)
-+ return -EINVAL;
++ u32 v;
+
-+ pending_reqs[mmap_alloc] = kzalloc(sizeof(pending_req_t)
-+ * blkif_reqs, GFP_KERNEL);
-+ foreign_pages[mmap_alloc] = alloc_empty_pages_and_pagevec(mmap_pages);
++ if (regno > info->cmap.len)
++ return 1;
+
-+ if (!pending_reqs[mmap_alloc] || !foreign_pages[mmap_alloc])
-+ goto out_of_memory;
++ red >>= (16 - info->var.red.length);
++ green >>= (16 - info->var.green.length);
++ blue >>= (16 - info->var.blue.length);
+
-+ DPRINTK("%s: reqs=%d, pages=%d\n",
-+ __FUNCTION__, blkif_reqs, mmap_pages);
++ v = (red << info->var.red.offset) |
++ (green << info->var.green.offset) |
++ (blue << info->var.blue.offset);
+
-+ for (i = 0; i < MAX_PENDING_REQS; i++) {
-+ list_add_tail(&pending_reqs[mmap_alloc][i].free_list,
-+ &pending_free);
-+ pending_reqs[mmap_alloc][i].mem_idx = mmap_alloc;
-+ for (j = 0; j < BLKIF_MAX_SEGMENTS_PER_REQUEST; j++)
-+ BLKTAP_INVALIDATE_HANDLE(&pending_handle(mmap_alloc,
-+ i, j));
++ /* FIXME is this sane? check against xxxfb_setcolreg()! */
++ switch (info->var.bits_per_pixel) {
++ case 16:
++ case 24:
++ case 32:
++ ((u32 *)info->pseudo_palette)[regno] = v;
++ break;
+ }
-+
-+ mmap_alloc++;
-+ DPRINTK("# MMAPs increased to %d\n",mmap_alloc);
++
+ return 0;
++}
+
-+ out_of_memory:
-+ free_empty_pages_and_pagevec(foreign_pages[mmap_alloc], mmap_pages);
-+ kfree(pending_reqs[mmap_alloc]);
-+ WPRINTK("%s: out of memory\n", __FUNCTION__);
-+ return -ENOMEM;
++static void xenfb_timer(unsigned long data)
++{
++ struct xenfb_info *info = (struct xenfb_info *)data;
++ wake_up(&info->wq);
+}
+
-+static void mmap_req_del(int mmap)
++static void __xenfb_refresh(struct xenfb_info *info,
++ int x1, int y1, int w, int h)
+{
-+ BUG_ON(!spin_is_locked(&pending_free_lock));
++ int y2, x2;
+
-+ kfree(pending_reqs[mmap]);
-+ pending_reqs[mmap] = NULL;
++ y2 = y1 + h;
++ x2 = x1 + w;
+
-+ free_empty_pages_and_pagevec(foreign_pages[mmap_alloc], mmap_pages);
-+ foreign_pages[mmap] = NULL;
++ if (info->y1 > y1)
++ info->y1 = y1;
++ if (info->y2 < y2)
++ info->y2 = y2;
++ if (info->x1 > x1)
++ info->x1 = x1;
++ if (info->x2 < x2)
++ info->x2 = x2;
++ info->dirty = 1;
+
-+ mmap_lock = 0;
-+ DPRINTK("# MMAPs decreased to %d\n",mmap_alloc);
-+ mmap_alloc--;
++ if (timer_pending(&info->refresh))
++ return;
++
++ mod_timer(&info->refresh, jiffies + HZ/xenfb_fps);
+}
+
-+static pending_req_t* alloc_req(void)
++static void xenfb_refresh(struct xenfb_info *info,
++ int x1, int y1, int w, int h)
+{
-+ pending_req_t *req = NULL;
+ unsigned long flags;
+
-+ spin_lock_irqsave(&pending_free_lock, flags);
-+
-+ if (!list_empty(&pending_free)) {
-+ req = list_entry(pending_free.next, pending_req_t, free_list);
-+ list_del(&req->free_list);
-+ }
++ spin_lock_irqsave(&info->dirty_lock, flags);
++ __xenfb_refresh(info, x1, y1, w, h);
++ spin_unlock_irqrestore(&info->dirty_lock, flags);
++}
+
-+ if (req) {
-+ req->inuse = 1;
-+ alloc_pending_reqs++;
-+ }
-+ spin_unlock_irqrestore(&pending_free_lock, flags);
++static void xenfb_fillrect(struct fb_info *p, const struct fb_fillrect *rect)
++{
++ struct xenfb_info *info = p->par;
+
-+ return req;
++ cfb_fillrect(p, rect);
++ xenfb_refresh(info, rect->dx, rect->dy, rect->width, rect->height);
+}
+
-+static void free_req(pending_req_t *req)
++static void xenfb_imageblit(struct fb_info *p, const struct fb_image *image)
+{
-+ unsigned long flags;
-+ int was_empty;
-+
-+ spin_lock_irqsave(&pending_free_lock, flags);
++ struct xenfb_info *info = p->par;
+
-+ alloc_pending_reqs--;
-+ req->inuse = 0;
-+ if (mmap_lock && (req->mem_idx == mmap_alloc-1)) {
-+ mmap_inuse--;
-+ if (mmap_inuse == 0) mmap_req_del(mmap_alloc-1);
-+ spin_unlock_irqrestore(&pending_free_lock, flags);
-+ return;
-+ }
-+ was_empty = list_empty(&pending_free);
-+ list_add(&req->free_list, &pending_free);
++ cfb_imageblit(p, image);
++ xenfb_refresh(info, image->dx, image->dy, image->width, image->height);
++}
+
-+ spin_unlock_irqrestore(&pending_free_lock, flags);
++static void xenfb_copyarea(struct fb_info *p, const struct fb_copyarea *area)
++{
++ struct xenfb_info *info = p->par;
+
-+ if (was_empty)
-+ wake_up(&pending_free_wq);
++ cfb_copyarea(p, area);
++ xenfb_refresh(info, area->dx, area->dy, area->width, area->height);
+}
+
-+static void fast_flush_area(pending_req_t *req, int k_idx, int u_idx,
-+ int tapidx)
++static void xenfb_vm_open(struct vm_area_struct *vma)
+{
-+ struct gnttab_unmap_grant_ref unmap[BLKIF_MAX_SEGMENTS_PER_REQUEST*2];
-+ unsigned int i, invcount = 0;
-+ struct grant_handle_pair *khandle;
-+ uint64_t ptep;
-+ int ret, mmap_idx;
-+ unsigned long kvaddr, uvaddr;
-+ tap_blkif_t *info;
-+
++ struct xenfb_mapping *map = vma->vm_private_data;
++ atomic_inc(&map->map_refs);
++}
+
-+ info = tapfds[tapidx];
++static void xenfb_vm_close(struct vm_area_struct *vma)
++{
++ struct xenfb_mapping *map = vma->vm_private_data;
++ struct xenfb_info *info = map->info;
+
-+ if ((tapidx < 0) || (tapidx > MAX_TAP_DEV) || !info) {
-+ WPRINTK("fast_flush: Couldn't get info!\n");
-+ return;
++ mutex_lock(&info->mm_lock);
++ if (atomic_dec_and_test(&map->map_refs)) {
++ list_del(&map->link);
++ kfree(map);
+ }
++ mutex_unlock(&info->mm_lock);
++}
+
-+ if (info->vma != NULL &&
-+ xen_feature(XENFEAT_auto_translated_physmap)) {
-+ down_write(&info->vma->vm_mm->mmap_sem);
-+ zap_page_range(info->vma,
-+ MMAP_VADDR(info->user_vstart, u_idx, 0),
-+ req->nr_pages << PAGE_SHIFT, NULL);
-+ up_write(&info->vma->vm_mm->mmap_sem);
-+ return;
-+ }
++static struct page *xenfb_vm_nopage(struct vm_area_struct *vma,
++ unsigned long vaddr, int *type)
++{
++ struct xenfb_mapping *map = vma->vm_private_data;
++ struct xenfb_info *info = map->info;
++ int pgnr = (vaddr - vma->vm_start) >> PAGE_SHIFT;
++ unsigned long flags;
++ struct page *page;
++ int y1, y2;
+
-+ mmap_idx = req->mem_idx;
++ if (pgnr >= info->nr_pages)
++ return NOPAGE_SIGBUS;
+
-+ for (i = 0; i < req->nr_pages; i++) {
-+ kvaddr = idx_to_kaddr(mmap_idx, k_idx, i);
-+ uvaddr = MMAP_VADDR(info->user_vstart, u_idx, i);
++ mutex_lock(&info->mm_lock);
++ spin_lock_irqsave(&info->dirty_lock, flags);
++ page = info->pages[pgnr];
++ get_page(page);
++ map->faults++;
+
-+ khandle = &pending_handle(mmap_idx, k_idx, i);
++ y1 = pgnr * PAGE_SIZE / info->fb_info->fix.line_length;
++ y2 = (pgnr * PAGE_SIZE + PAGE_SIZE - 1) / info->fb_info->fix.line_length;
++ if (y2 > info->fb_info->var.yres)
++ y2 = info->fb_info->var.yres;
++ __xenfb_refresh(info, 0, y1, info->fb_info->var.xres, y2 - y1);
++ spin_unlock_irqrestore(&info->dirty_lock, flags);
++ mutex_unlock(&info->mm_lock);
+
-+ if (khandle->kernel != INVALID_GRANT_HANDLE) {
-+ gnttab_set_unmap_op(&unmap[invcount],
-+ idx_to_kaddr(mmap_idx, k_idx, i),
-+ GNTMAP_host_map, khandle->kernel);
-+ invcount++;
++ if (type)
++ *type = VM_FAULT_MINOR;
+
-+ set_phys_to_machine(
-+ __pa(idx_to_kaddr(mmap_idx, k_idx, i))
-+ >> PAGE_SHIFT, INVALID_P2M_ENTRY);
-+ }
++ return page;
++}
+
-+ if (khandle->user != INVALID_GRANT_HANDLE) {
-+ BUG_ON(xen_feature(XENFEAT_auto_translated_physmap));
-+ if (create_lookup_pte_addr(
-+ info->vma->vm_mm,
-+ MMAP_VADDR(info->user_vstart, u_idx, i),
-+ &ptep) !=0) {
-+ WPRINTK("Couldn't get a pte addr!\n");
-+ return;
-+ }
++static struct vm_operations_struct xenfb_vm_ops = {
++ .open = xenfb_vm_open,
++ .close = xenfb_vm_close,
++ .nopage = xenfb_vm_nopage,
++};
+
-+ gnttab_set_unmap_op(&unmap[invcount], ptep,
-+ GNTMAP_host_map
-+ | GNTMAP_application_map
-+ | GNTMAP_contains_pte,
-+ khandle->user);
-+ invcount++;
-+ }
++static int xenfb_mmap(struct fb_info *fb_info, struct vm_area_struct *vma)
++{
++ struct xenfb_info *info = fb_info->par;
++ struct xenfb_mapping *map;
++ int map_pages;
+
-+ BLKTAP_INVALIDATE_HANDLE(khandle);
-+ }
-+ ret = HYPERVISOR_grant_table_op(
-+ GNTTABOP_unmap_grant_ref, unmap, invcount);
-+ BUG_ON(ret);
-+
-+ if (info->vma != NULL && !xen_feature(XENFEAT_auto_translated_physmap))
-+ zap_page_range(info->vma,
-+ MMAP_VADDR(info->user_vstart, u_idx, 0),
-+ req->nr_pages << PAGE_SHIFT, NULL);
-+}
++ if (!(vma->vm_flags & VM_WRITE))
++ return -EINVAL;
++ if (!(vma->vm_flags & VM_SHARED))
++ return -EINVAL;
++ if (vma->vm_pgoff != 0)
++ return -EINVAL;
+
-+/******************************************************************
-+ * SCHEDULER FUNCTIONS
-+ */
++ map_pages = (vma->vm_end - vma->vm_start + PAGE_SIZE-1) >> PAGE_SHIFT;
++ if (map_pages > info->nr_pages)
++ return -EINVAL;
+
-+static void print_stats(blkif_t *blkif)
-+{
-+ printk(KERN_DEBUG "%s: oo %3d | rd %4d | wr %4d\n",
-+ current->comm, blkif->st_oo_req,
-+ blkif->st_rd_req, blkif->st_wr_req);
-+ blkif->st_print = jiffies + msecs_to_jiffies(10 * 1000);
-+ blkif->st_rd_req = 0;
-+ blkif->st_wr_req = 0;
-+ blkif->st_oo_req = 0;
-+}
++ map = kzalloc(sizeof(*map), GFP_KERNEL);
++ if (map == NULL)
++ return -ENOMEM;
+
-+int tap_blkif_schedule(void *arg)
-+{
-+ blkif_t *blkif = arg;
++ map->vma = vma;
++ map->faults = 0;
++ map->info = info;
++ atomic_set(&map->map_refs, 1);
+
-+ blkif_get(blkif);
++ mutex_lock(&info->mm_lock);
++ list_add(&map->link, &info->mappings);
++ mutex_unlock(&info->mm_lock);
+
-+ if (debug_lvl)
-+ printk(KERN_DEBUG "%s: started\n", current->comm);
++ vma->vm_ops = &xenfb_vm_ops;
++ vma->vm_flags |= (VM_DONTEXPAND | VM_RESERVED);
++ vma->vm_private_data = map;
+
-+ while (!kthread_should_stop()) {
-+ if (try_to_freeze())
-+ continue;
++ return 0;
++}
+
-+ wait_event_interruptible(
-+ blkif->wq,
-+ blkif->waiting_reqs || kthread_should_stop());
-+ wait_event_interruptible(
-+ pending_free_wq,
-+ !list_empty(&pending_free) || kthread_should_stop());
++static int
++xenfb_check_var(struct fb_var_screeninfo *var, struct fb_info *info)
++{
++ struct xenfb_info *xenfb_info;
++ int required_mem_len;
+
-+ blkif->waiting_reqs = 0;
-+ smp_mb(); /* clear flag *before* checking for work */
++ xenfb_info = info->par;
+
-+ if (do_block_io_op(blkif))
-+ blkif->waiting_reqs = 1;
++ if (!xenfb_info->feature_resize) {
++ if (var->xres == video[KPARAM_WIDTH] &&
++ var->yres == video[KPARAM_HEIGHT] &&
++ var->bits_per_pixel == xenfb_info->page->depth) {
++ return 0;
++ }
++ return -EINVAL;
++ }
+
-+ if (log_stats && time_after(jiffies, blkif->st_print))
-+ print_stats(blkif);
++ /* Can't resize past initial width and height */
++ if (var->xres > video[KPARAM_WIDTH] || var->yres > video[KPARAM_HEIGHT])
++ return -EINVAL;
++
++ required_mem_len = var->xres * var->yres * (xenfb_info->page->depth / 8);
++ if (var->bits_per_pixel == xenfb_info->page->depth &&
++ var->xres <= info->fix.line_length / (XENFB_DEPTH / 8) &&
++ required_mem_len <= info->fix.smem_len) {
++ var->xres_virtual = var->xres;
++ var->yres_virtual = var->yres;
++ return 0;
+ }
++ return -EINVAL;
++}
+
-+ if (log_stats)
-+ print_stats(blkif);
-+ if (debug_lvl)
-+ printk(KERN_DEBUG "%s: exiting\n", current->comm);
++static int xenfb_set_par(struct fb_info *info)
++{
++ struct xenfb_info *xenfb_info;
++ unsigned long flags;
+
-+ blkif->xenblkd = NULL;
-+ blkif_put(blkif);
++ xenfb_info = info->par;
+
++ spin_lock_irqsave(&xenfb_info->resize_lock, flags);
++ xenfb_info->resize.type = XENFB_TYPE_RESIZE;
++ xenfb_info->resize.width = info->var.xres;
++ xenfb_info->resize.height = info->var.yres;
++ xenfb_info->resize.stride = info->fix.line_length;
++ xenfb_info->resize.depth = info->var.bits_per_pixel;
++ xenfb_info->resize.offset = 0;
++ xenfb_info->resize_dpy = 1;
++ spin_unlock_irqrestore(&xenfb_info->resize_lock, flags);
+ return 0;
+}
+
-+/******************************************************************
-+ * COMPLETION CALLBACK -- Called by user level ioctl()
-+ */
++static struct fb_ops xenfb_fb_ops = {
++ .owner = THIS_MODULE,
++ .fb_setcolreg = xenfb_setcolreg,
++ .fb_fillrect = xenfb_fillrect,
++ .fb_copyarea = xenfb_copyarea,
++ .fb_imageblit = xenfb_imageblit,
++ .fb_mmap = xenfb_mmap,
++ .fb_check_var = xenfb_check_var,
++ .fb_set_par = xenfb_set_par,
++};
+
-+static int blktap_read_ufe_ring(tap_blkif_t *info)
++static irqreturn_t xenfb_event_handler(int rq, void *dev_id,
++ struct pt_regs *regs)
+{
-+ /* This is called to read responses from the UFE ring. */
-+ RING_IDX i, j, rp;
-+ blkif_response_t *resp;
-+ blkif_t *blkif=NULL;
-+ int pending_idx, usr_idx, mmap_idx;
-+ pending_req_t *pending_req;
-+
-+ if (!info)
-+ return 0;
++ /*
++ * No in events recognized, simply ignore them all.
++ * If you need to recognize some, see xenbkd's input_handler()
++ * for how to do that.
++ */
++ struct xenfb_info *info = dev_id;
++ struct xenfb_page *page = info->page;
+
-+ /* We currently only forward packets in INTERCEPT_FE mode. */
-+ if (!(info->mode & BLKTAP_MODE_INTERCEPT_FE))
-+ return 0;
++ if (page->in_cons != page->in_prod) {
++ info->page->in_cons = info->page->in_prod;
++ notify_remote_via_irq(info->irq);
++ }
++ return IRQ_HANDLED;
++}
+
-+ /* for each outstanding message on the UFEring */
-+ rp = info->ufe_ring.sring->rsp_prod;
-+ rmb();
-+
-+ for (i = info->ufe_ring.rsp_cons; i != rp; i++) {
-+ blkif_response_t res;
-+ resp = RING_GET_RESPONSE(&info->ufe_ring, i);
-+ memcpy(&res, resp, sizeof(res));
-+ mb(); /* rsp_cons read by RING_FULL() in do_block_io_op(). */
-+ ++info->ufe_ring.rsp_cons;
++static unsigned long vmalloc_to_mfn(void *address)
++{
++ return pfn_to_mfn(vmalloc_to_pfn(address));
++}
+
-+ /*retrieve [usr_idx] to [mmap_idx,pending_idx] mapping*/
-+ usr_idx = (int)res.id;
-+ pending_idx = MASK_PEND_IDX(ID_TO_IDX(info->idx_map[usr_idx]));
-+ mmap_idx = ID_TO_MIDX(info->idx_map[usr_idx]);
++static int __devinit xenfb_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ struct xenfb_info *info;
++ struct fb_info *fb_info;
++ int fb_size;
++ int val;
++ int ret;
+
-+ if ( (mmap_idx >= mmap_alloc) ||
-+ (ID_TO_IDX(info->idx_map[usr_idx]) >= MAX_PENDING_REQS) )
-+ WPRINTK("Incorrect req map"
-+ "[%d], internal map [%d,%d (%d)]\n",
-+ usr_idx, mmap_idx,
-+ ID_TO_IDX(info->idx_map[usr_idx]),
-+ MASK_PEND_IDX(
-+ ID_TO_IDX(info->idx_map[usr_idx])));
++ info = kzalloc(sizeof(*info), GFP_KERNEL);
++ if (info == NULL) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
++ return -ENOMEM;
++ }
+
-+ pending_req = &pending_reqs[mmap_idx][pending_idx];
-+ blkif = pending_req->blkif;
++ /* Limit kernel param videoram amount to what is in xenstore */
++ if (xenbus_scanf(XBT_NIL, dev->otherend, "videoram", "%d", &val) == 1) {
++ if (val < video[KPARAM_MEM])
++ video[KPARAM_MEM] = val;
++ }
+
-+ for (j = 0; j < pending_req->nr_pages; j++) {
++ /* If requested res does not fit in available memory, use default */
++ fb_size = video[KPARAM_MEM] * MB_;
++ if (video[KPARAM_WIDTH] * video[KPARAM_HEIGHT] * XENFB_DEPTH/8 > fb_size) {
++ video[KPARAM_WIDTH] = XENFB_WIDTH;
++ video[KPARAM_HEIGHT] = XENFB_HEIGHT;
++ fb_size = XENFB_DEFAULT_FB_LEN;
++ }
+
-+ unsigned long kvaddr, uvaddr;
-+ struct page **map = info->vma->vm_private_data;
-+ struct page *pg;
-+ int offset;
++ dev->dev.driver_data = info;
++ info->xbdev = dev;
++ info->irq = -1;
++ info->x1 = info->y1 = INT_MAX;
++ spin_lock_init(&info->dirty_lock);
++ spin_lock_init(&info->resize_lock);
++ mutex_init(&info->mm_lock);
++ init_waitqueue_head(&info->wq);
++ init_timer(&info->refresh);
++ info->refresh.function = xenfb_timer;
++ info->refresh.data = (unsigned long)info;
++ INIT_LIST_HEAD(&info->mappings);
+
-+ uvaddr = MMAP_VADDR(info->user_vstart, usr_idx, j);
-+ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, j);
++ info->fb = vmalloc(fb_size);
++ if (info->fb == NULL)
++ goto error_nomem;
++ memset(info->fb, 0, fb_size);
+
-+ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
-+ ClearPageReserved(pg);
-+ offset = (uvaddr - info->vma->vm_start)
-+ >> PAGE_SHIFT;
-+ map[offset] = NULL;
-+ }
-+ fast_flush_area(pending_req, pending_idx, usr_idx, info->minor);
-+ info->idx_map[usr_idx] = INVALID_REQ;
-+ make_response(blkif, pending_req->id, res.operation,
-+ res.status);
-+ blkif_put(pending_req->blkif);
-+ free_req(pending_req);
-+ }
-+
-+ return 0;
-+}
++ info->nr_pages = (fb_size + PAGE_SIZE - 1) >> PAGE_SHIFT;
+
++ info->pages = kmalloc(sizeof(struct page *) * info->nr_pages,
++ GFP_KERNEL);
++ if (info->pages == NULL)
++ goto error_nomem;
+
-+/******************************************************************************
-+ * NOTIFICATION FROM GUEST OS.
-+ */
++ info->mfns = vmalloc(sizeof(unsigned long) * info->nr_pages);
++ if (!info->mfns)
++ goto error_nomem;
+
-+static void blkif_notify_work(blkif_t *blkif)
-+{
-+ blkif->waiting_reqs = 1;
-+ wake_up(&blkif->wq);
-+}
++ /* set up shared page */
++ info->page = (void *)__get_free_page(GFP_KERNEL | __GFP_ZERO);
++ if (!info->page)
++ goto error_nomem;
+
-+irqreturn_t tap_blkif_be_int(int irq, void *dev_id, struct pt_regs *regs)
-+{
-+ blkif_notify_work(dev_id);
-+ return IRQ_HANDLED;
-+}
++ fb_info = framebuffer_alloc(sizeof(u32) * 256, NULL);
++ /* see fishy hackery below */
++ if (fb_info == NULL)
++ goto error_nomem;
+
++ /* FIXME fishy hackery */
++ fb_info->pseudo_palette = fb_info->par;
++ fb_info->par = info;
++ /* /FIXME */
++ fb_info->screen_base = info->fb;
+
++ fb_info->fbops = &xenfb_fb_ops;
++ fb_info->var.xres_virtual = fb_info->var.xres = video[KPARAM_WIDTH];
++ fb_info->var.yres_virtual = fb_info->var.yres = video[KPARAM_HEIGHT];
++ fb_info->var.bits_per_pixel = XENFB_DEPTH;
+
-+/******************************************************************
-+ * DOWNWARD CALLS -- These interface with the block-device layer proper.
-+ */
-+static int print_dbug = 1;
-+static int do_block_io_op(blkif_t *blkif)
-+{
-+ blkif_back_rings_t *blk_rings = &blkif->blk_rings;
-+ blkif_request_t req;
-+ pending_req_t *pending_req;
-+ RING_IDX rc, rp;
-+ int more_to_do = 0;
-+ tap_blkif_t *info;
++ fb_info->var.red = (struct fb_bitfield){16, 8, 0};
++ fb_info->var.green = (struct fb_bitfield){8, 8, 0};
++ fb_info->var.blue = (struct fb_bitfield){0, 8, 0};
+
-+ rc = blk_rings->common.req_cons;
-+ rp = blk_rings->common.sring->req_prod;
-+ rmb(); /* Ensure we see queued requests up to 'rp'. */
++ fb_info->var.activate = FB_ACTIVATE_NOW;
++ fb_info->var.height = -1;
++ fb_info->var.width = -1;
++ fb_info->var.vmode = FB_VMODE_NONINTERLACED;
+
-+ /*Check blkif has corresponding UE ring*/
-+ if (blkif->dev_num < 0) {
-+ /*oops*/
-+ if (print_dbug) {
-+ WPRINTK("Corresponding UE "
-+ "ring does not exist!\n");
-+ print_dbug = 0; /*We only print this message once*/
-+ }
-+ return 0;
++ fb_info->fix.visual = FB_VISUAL_TRUECOLOR;
++ fb_info->fix.line_length = fb_info->var.xres * (XENFB_DEPTH / 8);
++ fb_info->fix.smem_start = 0;
++ fb_info->fix.smem_len = fb_size;
++ strcpy(fb_info->fix.id, "xen");
++ fb_info->fix.type = FB_TYPE_PACKED_PIXELS;
++ fb_info->fix.accel = FB_ACCEL_NONE;
++
++ fb_info->flags = FBINFO_FLAG_DEFAULT;
++
++ ret = fb_alloc_cmap(&fb_info->cmap, 256, 0);
++ if (ret < 0) {
++ framebuffer_release(fb_info);
++ xenbus_dev_fatal(dev, ret, "fb_alloc_cmap");
++ goto error;
+ }
+
-+ info = tapfds[blkif->dev_num];
++ xenfb_init_shared_page(info, fb_info);
+
-+ if (blkif->dev_num > MAX_TAP_DEV || !info || !info->dev_inuse) {
-+ if (print_dbug) {
-+ WPRINTK("Can't get UE info!\n");
-+ print_dbug = 0;
-+ }
-+ return 0;
++ ret = register_framebuffer(fb_info);
++ if (ret) {
++ fb_dealloc_cmap(&info->fb_info->cmap);
++ framebuffer_release(fb_info);
++ xenbus_dev_fatal(dev, ret, "register_framebuffer");
++ goto error;
+ }
++ info->fb_info = fb_info;
+
-+ while (rc != rp) {
-+
-+ if (RING_FULL(&info->ufe_ring)) {
-+ WPRINTK("RING_FULL! More to do\n");
-+ more_to_do = 1;
-+ break;
-+ }
++ /* FIXME should this be delayed until backend XenbusStateConnected? */
++ info->kthread = kthread_run(xenfb_thread, info, "xenfb thread");
++ if (IS_ERR(info->kthread)) {
++ ret = PTR_ERR(info->kthread);
++ info->kthread = NULL;
++ xenbus_dev_fatal(dev, ret, "register_framebuffer");
++ goto error;
++ }
+
-+ if (RING_REQUEST_CONS_OVERFLOW(&blk_rings->common, rc)) {
-+ WPRINTK("RING_REQUEST_CONS_OVERFLOW!"
-+ " More to do\n");
-+ more_to_do = 1;
-+ break;
-+ }
++ ret = xenfb_connect_backend(dev, info);
++ if (ret < 0)
++ goto error;
+
-+ pending_req = alloc_req();
-+ if (NULL == pending_req) {
-+ blkif->st_oo_req++;
-+ more_to_do = 1;
-+ break;
-+ }
++ return 0;
+
-+ switch (blkif->blk_protocol) {
-+ case BLKIF_PROTOCOL_NATIVE:
-+ memcpy(&req, RING_GET_REQUEST(&blk_rings->native, rc),
-+ sizeof(req));
-+ break;
-+ case BLKIF_PROTOCOL_X86_32:
-+ blkif_get_x86_32_req(&req, RING_GET_REQUEST(&blk_rings->x86_32, rc));
-+ break;
-+ case BLKIF_PROTOCOL_X86_64:
-+ blkif_get_x86_64_req(&req, RING_GET_REQUEST(&blk_rings->x86_64, rc));
-+ break;
-+ default:
-+ BUG();
-+ }
-+ blk_rings->common.req_cons = ++rc; /* before make_response() */
++ error_nomem:
++ ret = -ENOMEM;
++ xenbus_dev_fatal(dev, ret, "allocating device memory");
++ error:
++ xenfb_remove(dev);
++ return ret;
++}
+
-+ switch (req.operation) {
-+ case BLKIF_OP_READ:
-+ blkif->st_rd_req++;
-+ dispatch_rw_block_io(blkif, &req, pending_req);
-+ break;
++static int xenfb_resume(struct xenbus_device *dev)
++{
++ struct xenfb_info *info = dev->dev.driver_data;
+
-+ case BLKIF_OP_WRITE:
-+ blkif->st_wr_req++;
-+ dispatch_rw_block_io(blkif, &req, pending_req);
-+ break;
++ xenfb_disconnect_backend(info);
++ xenfb_init_shared_page(info, info->fb_info);
++ return xenfb_connect_backend(dev, info);
++}
+
-+ default:
-+ WPRINTK("unknown operation [%d]\n",
-+ req.operation);
-+ make_response(blkif, req.id, req.operation,
-+ BLKIF_RSP_ERROR);
-+ free_req(pending_req);
-+ break;
-+ }
++static int xenfb_remove(struct xenbus_device *dev)
++{
++ struct xenfb_info *info = dev->dev.driver_data;
++
++ del_timer(&info->refresh);
++ if (info->kthread)
++ kthread_stop(info->kthread);
++ xenfb_disconnect_backend(info);
++ if (info->fb_info) {
++ unregister_framebuffer(info->fb_info);
++ fb_dealloc_cmap(&info->fb_info->cmap);
++ framebuffer_release(info->fb_info);
+ }
-+
-+ blktap_kick_user(blkif->dev_num);
++ free_page((unsigned long)info->page);
++ vfree(info->mfns);
++ kfree(info->pages);
++ vfree(info->fb);
++ kfree(info);
+
-+ return more_to_do;
++ return 0;
+}
+
-+static void dispatch_rw_block_io(blkif_t *blkif,
-+ blkif_request_t *req,
-+ pending_req_t *pending_req)
++static void xenfb_init_shared_page(struct xenfb_info *info,
++ struct fb_info * fb_info)
+{
-+ extern void ll_rw_block(int rw, int nr, struct buffer_head * bhs[]);
-+ int op, operation = (req->operation == BLKIF_OP_WRITE) ? WRITE : READ;
-+ struct gnttab_map_grant_ref map[BLKIF_MAX_SEGMENTS_PER_REQUEST*2];
-+ unsigned int nseg;
-+ int ret, i, nr_sects = 0;
-+ tap_blkif_t *info;
-+ blkif_request_t *target;
-+ int pending_idx = RTN_PEND_IDX(pending_req,pending_req->mem_idx);
-+ int usr_idx;
-+ uint16_t mmap_idx = pending_req->mem_idx;
++ int i;
++ int epd = PAGE_SIZE / sizeof(info->mfns[0]);
+
-+ if (blkif->dev_num < 0 || blkif->dev_num > MAX_TAP_DEV)
-+ goto fail_response;
++ for (i = 0; i < info->nr_pages; i++)
++ info->pages[i] = vmalloc_to_page(info->fb + i * PAGE_SIZE);
+
-+ info = tapfds[blkif->dev_num];
-+ if (info == NULL)
-+ goto fail_response;
++ for (i = 0; i < info->nr_pages; i++)
++ info->mfns[i] = vmalloc_to_mfn(info->fb + i * PAGE_SIZE);
+
-+ /* Check we have space on user ring - should never fail. */
-+ usr_idx = GET_NEXT_REQ(info->idx_map);
-+ if (usr_idx == INVALID_REQ) {
-+ BUG();
-+ goto fail_response;
++ for (i = 0; i * epd < info->nr_pages; i++)
++ info->page->pd[i] = vmalloc_to_mfn(&info->mfns[i * epd]);
++
++ info->page->width = fb_info->var.xres;
++ info->page->height = fb_info->var.yres;
++ info->page->depth = fb_info->var.bits_per_pixel;
++ info->page->line_length = fb_info->fix.line_length;
++ info->page->mem_length = fb_info->fix.smem_len;
++ info->page->in_cons = info->page->in_prod = 0;
++ info->page->out_cons = info->page->out_prod = 0;
++}
++
++static int xenfb_connect_backend(struct xenbus_device *dev,
++ struct xenfb_info *info)
++{
++ int ret;
++ struct xenbus_transaction xbt;
++
++ ret = bind_listening_port_to_irqhandler(
++ dev->otherend_id, xenfb_event_handler, 0, "xenfb", info);
++ if (ret < 0) {
++ xenbus_dev_fatal(dev, ret,
++ "bind_listening_port_to_irqhandler");
++ return ret;
+ }
++ info->irq = ret;
+
-+ /* Check that number of segments is sane. */
-+ nseg = req->nr_segments;
-+ if ( unlikely(nseg == 0) ||
-+ unlikely(nseg > BLKIF_MAX_SEGMENTS_PER_REQUEST) ) {
-+ WPRINTK("Bad number of segments in request (%d)\n", nseg);
-+ goto fail_response;
++ again:
++ ret = xenbus_transaction_start(&xbt);
++ if (ret) {
++ xenbus_dev_fatal(dev, ret, "starting transaction");
++ return ret;
+ }
-+
-+ /* Make sure userspace is ready. */
-+ if (!info->ring_ok) {
-+ WPRINTK("blktap: ring not ready for requests!\n");
-+ goto fail_response;
++ ret = xenbus_printf(xbt, dev->nodename, "page-ref", "%lu",
++ virt_to_mfn(info->page));
++ if (ret)
++ goto error_xenbus;
++ ret = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
++ irq_to_evtchn_port(info->irq));
++ if (ret)
++ goto error_xenbus;
++ ret = xenbus_printf(xbt, dev->nodename, "protocol", "%s",
++ XEN_IO_PROTO_ABI_NATIVE);
++ if (ret)
++ goto error_xenbus;
++ ret = xenbus_printf(xbt, dev->nodename, "feature-update", "1");
++ if (ret)
++ goto error_xenbus;
++ ret = xenbus_transaction_end(xbt, 0);
++ if (ret) {
++ if (ret == -EAGAIN)
++ goto again;
++ xenbus_dev_fatal(dev, ret, "completing transaction");
++ return ret;
+ }
+
-+ if (RING_FULL(&info->ufe_ring)) {
-+ WPRINTK("blktap: fe_ring is full, can't add "
-+ "IO Request will be dropped. %d %d\n",
-+ RING_SIZE(&info->ufe_ring),
-+ RING_SIZE(&blkif->blk_rings.common));
-+ goto fail_response;
-+ }
++ xenbus_switch_state(dev, XenbusStateInitialised);
++ return 0;
+
-+ pending_req->blkif = blkif;
-+ pending_req->id = req->id;
-+ pending_req->operation = operation;
-+ pending_req->status = BLKIF_RSP_OKAY;
-+ pending_req->nr_pages = nseg;
-+ op = 0;
-+ for (i = 0; i < nseg; i++) {
-+ unsigned long uvaddr;
-+ unsigned long kvaddr;
-+ uint64_t ptep;
-+ uint32_t flags;
++ error_xenbus:
++ xenbus_transaction_end(xbt, 1);
++ xenbus_dev_fatal(dev, ret, "writing xenstore");
++ return ret;
++}
+
-+ uvaddr = MMAP_VADDR(info->user_vstart, usr_idx, i);
-+ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, i);
++static void xenfb_disconnect_backend(struct xenfb_info *info)
++{
++ if (info->irq >= 0)
++ unbind_from_irqhandler(info->irq, info);
++ info->irq = -1;
++}
+
-+ flags = GNTMAP_host_map;
-+ if (operation == WRITE)
-+ flags |= GNTMAP_readonly;
-+ gnttab_set_map_op(&map[op], kvaddr, flags,
-+ req->seg[i].gref, blkif->domid);
-+ op++;
++static void xenfb_backend_changed(struct xenbus_device *dev,
++ enum xenbus_state backend_state)
++{
++ struct xenfb_info *info = dev->dev.driver_data;
++ int val;
+
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ /* Now map it to user. */
-+ ret = create_lookup_pte_addr(info->vma->vm_mm,
-+ uvaddr, &ptep);
-+ if (ret) {
-+ WPRINTK("Couldn't get a pte addr!\n");
-+ goto fail_flush;
-+ }
++ switch (backend_state) {
++ case XenbusStateInitialising:
++ case XenbusStateInitialised:
++ case XenbusStateReconfiguring:
++ case XenbusStateReconfigured:
++ case XenbusStateUnknown:
++ case XenbusStateClosed:
++ break;
++
++ case XenbusStateInitWait:
++ InitWait:
++ xenbus_switch_state(dev, XenbusStateConnected);
++ break;
++
++ case XenbusStateConnected:
++ /*
++ * Work around xenbus race condition: If backend goes
++ * through InitWait to Connected fast enough, we can
++ * get Connected twice here.
++ */
++ if (dev->state != XenbusStateConnected)
++ goto InitWait; /* no InitWait seen yet, fudge it */
+
-+ flags = GNTMAP_host_map | GNTMAP_application_map
-+ | GNTMAP_contains_pte;
-+ if (operation == WRITE)
-+ flags |= GNTMAP_readonly;
-+ gnttab_set_map_op(&map[op], ptep, flags,
-+ req->seg[i].gref, blkif->domid);
-+ op++;
-+ }
++ if (xenbus_scanf(XBT_NIL, info->xbdev->otherend,
++ "request-update", "%d", &val) < 0)
++ val = 0;
++ if (val)
++ info->update_wanted = 1;
+
-+ nr_sects += (req->seg[i].last_sect -
-+ req->seg[i].first_sect + 1);
++ if (xenbus_scanf(XBT_NIL, dev->otherend,
++ "feature-resize", "%d", &val) < 0)
++ val = 0;
++ info->feature_resize = val;
++ break;
++
++ case XenbusStateClosing:
++ // FIXME is this safe in any dev->state?
++ xenbus_frontend_closed(dev);
++ break;
+ }
++}
+
-+ ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, map, op);
-+ BUG_ON(ret);
++static const struct xenbus_device_id xenfb_ids[] = {
++ { "vfb" },
++ { "" }
++};
++MODULE_ALIAS("xen:vfb");
+
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ for (i = 0; i < (nseg*2); i+=2) {
-+ unsigned long uvaddr;
-+ unsigned long kvaddr;
-+ unsigned long offset;
-+ struct page *pg;
++static struct xenbus_driver xenfb_driver = {
++ .name = "vfb",
++ .owner = THIS_MODULE,
++ .ids = xenfb_ids,
++ .probe = xenfb_probe,
++ .remove = xenfb_remove,
++ .resume = xenfb_resume,
++ .otherend_changed = xenfb_backend_changed,
++};
+
-+ uvaddr = MMAP_VADDR(info->user_vstart, usr_idx, i/2);
-+ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, i/2);
++static int __init xenfb_init(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+ if (unlikely(map[i].status != 0)) {
-+ WPRINTK("invalid kernel buffer -- "
-+ "could not remap it\n");
-+ ret |= 1;
-+ map[i].handle = INVALID_GRANT_HANDLE;
-+ }
++ /* Nothing to do if running in dom0. */
++ if (is_initial_xendomain())
++ return -ENODEV;
+
-+ if (unlikely(map[i+1].status != 0)) {
-+ WPRINTK("invalid user buffer -- "
-+ "could not remap it\n");
-+ ret |= 1;
-+ map[i+1].handle = INVALID_GRANT_HANDLE;
-+ }
++ return xenbus_register_frontend(&xenfb_driver);
++}
+
-+ pending_handle(mmap_idx, pending_idx, i/2).kernel
-+ = map[i].handle;
-+ pending_handle(mmap_idx, pending_idx, i/2).user
-+ = map[i+1].handle;
++static void __exit xenfb_cleanup(void)
++{
++ return xenbus_unregister_driver(&xenfb_driver);
++}
+
-+ if (ret)
-+ continue;
++module_init(xenfb_init);
++module_exit(xenfb_cleanup);
+
-+ set_phys_to_machine(__pa(kvaddr) >> PAGE_SHIFT,
-+ FOREIGN_FRAME(map[i].dev_bus_addr
-+ >> PAGE_SHIFT));
-+ offset = (uvaddr - info->vma->vm_start) >> PAGE_SHIFT;
-+ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
-+ ((struct page **)info->vma->vm_private_data)[offset] =
-+ pg;
-+ }
-+ } else {
-+ for (i = 0; i < nseg; i++) {
-+ unsigned long uvaddr;
-+ unsigned long kvaddr;
-+ unsigned long offset;
-+ struct page *pg;
++MODULE_LICENSE("GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/fbfront/xenkbd.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/fbfront/xenkbd.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,354 @@
++/*
++ * linux/drivers/input/keyboard/xenkbd.c -- Xen para-virtual input device
++ *
++ * Copyright (C) 2005 Anthony Liguori <aliguori at us.ibm.com>
++ * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster <armbru at redhat.com>
++ *
++ * Based on linux/drivers/input/mouse/sermouse.c
++ *
++ * This file is subject to the terms and conditions of the GNU General Public
++ * License. See the file COPYING in the main directory of this archive for
++ * more details.
++ */
+
-+ uvaddr = MMAP_VADDR(info->user_vstart, usr_idx, i);
-+ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, i);
++/*
++ * TODO:
++ *
++ * Switch to grant tables together with xenfb.c.
++ */
+
-+ if (unlikely(map[i].status != 0)) {
-+ WPRINTK("invalid kernel buffer -- "
-+ "could not remap it\n");
-+ ret |= 1;
-+ map[i].handle = INVALID_GRANT_HANDLE;
-+ }
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/module.h>
++#include <linux/input.h>
++#include <asm/hypervisor.h>
++#include <xen/evtchn.h>
++#include <xen/interface/io/fbif.h>
++#include <xen/interface/io/kbdif.h>
++#include <xen/xenbus.h>
+
-+ pending_handle(mmap_idx, pending_idx, i).kernel
-+ = map[i].handle;
++struct xenkbd_info
++{
++ struct input_dev *kbd;
++ struct input_dev *ptr;
++ struct xenkbd_page *page;
++ int irq;
++ struct xenbus_device *xbdev;
++ char phys[32];
++};
+
-+ if (ret)
-+ continue;
++static int xenkbd_remove(struct xenbus_device *);
++static int xenkbd_connect_backend(struct xenbus_device *, struct xenkbd_info *);
++static void xenkbd_disconnect_backend(struct xenkbd_info *);
+
-+ offset = (uvaddr - info->vma->vm_start) >> PAGE_SHIFT;
-+ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
-+ ((struct page **)info->vma->vm_private_data)[offset] =
-+ pg;
++/*
++ * Note: if you need to send out events, see xenfb_do_update() for how
++ * to do that.
++ */
++
++static irqreturn_t input_handler(int rq, void *dev_id, struct pt_regs *regs)
++{
++ struct xenkbd_info *info = dev_id;
++ struct xenkbd_page *page = info->page;
++ __u32 cons, prod;
++
++ prod = page->in_prod;
++ if (prod == page->in_cons)
++ return IRQ_HANDLED;
++ rmb(); /* ensure we see ring contents up to prod */
++ for (cons = page->in_cons; cons != prod; cons++) {
++ union xenkbd_in_event *event;
++ struct input_dev *dev;
++ event = &XENKBD_IN_RING_REF(page, cons);
++
++ dev = info->ptr;
++ switch (event->type) {
++ case XENKBD_TYPE_MOTION:
++ if (event->motion.rel_z)
++ input_report_rel(dev, REL_WHEEL,
++ -event->motion.rel_z);
++ input_report_rel(dev, REL_X, event->motion.rel_x);
++ input_report_rel(dev, REL_Y, event->motion.rel_y);
++ break;
++ case XENKBD_TYPE_KEY:
++ dev = NULL;
++ if (test_bit(event->key.keycode, info->kbd->keybit))
++ dev = info->kbd;
++ if (test_bit(event->key.keycode, info->ptr->keybit))
++ dev = info->ptr;
++ if (dev)
++ input_report_key(dev, event->key.keycode,
++ event->key.pressed);
++ else
++ printk("xenkbd: unhandled keycode 0x%x\n",
++ event->key.keycode);
++ break;
++ case XENKBD_TYPE_POS:
++ if (event->pos.rel_z)
++ input_report_rel(dev, REL_WHEEL,
++ -event->pos.rel_z);
++ input_report_abs(dev, ABS_X, event->pos.abs_x);
++ input_report_abs(dev, ABS_Y, event->pos.abs_y);
++ break;
+ }
++ if (dev)
++ input_sync(dev);
+ }
++ mb(); /* ensure we got ring contents */
++ page->in_cons = cons;
++ notify_remote_via_irq(info->irq);
+
-+ if (ret)
-+ goto fail_flush;
++ return IRQ_HANDLED;
++}
+
-+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ down_write(&info->vma->vm_mm->mmap_sem);
-+ /* Mark mapped pages as reserved: */
-+ for (i = 0; i < req->nr_segments; i++) {
-+ unsigned long kvaddr;
-+ struct page *pg;
++int __devinit xenkbd_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ int ret, i;
++ struct xenkbd_info *info;
++ struct input_dev *kbd, *ptr;
+
-+ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, i);
-+ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
-+ SetPageReserved(pg);
-+ if (xen_feature(XENFEAT_auto_translated_physmap)) {
-+ ret = vm_insert_page(info->vma,
-+ MMAP_VADDR(info->user_vstart,
-+ usr_idx, i), pg);
-+ if (ret) {
-+ up_write(&info->vma->vm_mm->mmap_sem);
-+ goto fail_flush;
-+ }
-+ }
++ info = kzalloc(sizeof(*info), GFP_KERNEL);
++ if (!info) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
++ return -ENOMEM;
+ }
-+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ up_write(&info->vma->vm_mm->mmap_sem);
-+
-+ /*record [mmap_idx,pending_idx] to [usr_idx] mapping*/
-+ info->idx_map[usr_idx] = MAKE_ID(mmap_idx, pending_idx);
++ dev->dev.driver_data = info;
++ info->xbdev = dev;
++ snprintf(info->phys, sizeof(info->phys), "xenbus/%s", dev->nodename);
+
-+ blkif_get(blkif);
-+ /* Finally, write the request message to the user ring. */
-+ target = RING_GET_REQUEST(&info->ufe_ring,
-+ info->ufe_ring.req_prod_pvt);
-+ memcpy(target, req, sizeof(*req));
-+ target->id = usr_idx;
-+ wmb(); /* blktap_poll() reads req_prod_pvt asynchronously */
-+ info->ufe_ring.req_prod_pvt++;
++ info->page = (void *)__get_free_page(GFP_KERNEL);
++ if (!info->page)
++ goto error_nomem;
++ info->page->in_cons = info->page->in_prod = 0;
++ info->page->out_cons = info->page->out_prod = 0;
+
-+ if (operation == READ)
-+ blkif->st_rd_sect += nr_sects;
-+ else if (operation == WRITE)
-+ blkif->st_wr_sect += nr_sects;
++ /* keyboard */
++ kbd = input_allocate_device();
++ if (!kbd)
++ goto error_nomem;
++ kbd->name = "Xen Virtual Keyboard";
++ kbd->phys = info->phys;
++ kbd->id.bustype = BUS_PCI;
++ kbd->id.vendor = 0x5853;
++ kbd->id.product = 0xffff;
++ kbd->evbit[0] = BIT(EV_KEY);
++ for (i = KEY_ESC; i < KEY_UNKNOWN; i++)
++ set_bit(i, kbd->keybit);
++ for (i = KEY_OK; i < KEY_MAX; i++)
++ set_bit(i, kbd->keybit);
+
-+ return;
++ ret = input_register_device(kbd);
++ if (ret) {
++ input_free_device(kbd);
++ xenbus_dev_fatal(dev, ret, "input_register_device(kbd)");
++ goto error;
++ }
++ info->kbd = kbd;
+
-+ fail_flush:
-+ WPRINTK("Reached Fail_flush\n");
-+ fast_flush_area(pending_req, pending_idx, usr_idx, blkif->dev_num);
-+ fail_response:
-+ make_response(blkif, req->id, req->operation, BLKIF_RSP_ERROR);
-+ free_req(pending_req);
-+}
++ /* pointing device */
++ ptr = input_allocate_device();
++ if (!ptr)
++ goto error_nomem;
++ ptr->name = "Xen Virtual Pointer";
++ ptr->phys = info->phys;
++ ptr->id.bustype = BUS_PCI;
++ ptr->id.vendor = 0x5853;
++ ptr->id.product = 0xfffe;
++ ptr->evbit[0] = BIT(EV_KEY) | BIT(EV_REL) | BIT(EV_ABS);
++ for (i = BTN_LEFT; i <= BTN_TASK; i++)
++ set_bit(i, ptr->keybit);
++ ptr->relbit[0] = BIT(REL_X) | BIT(REL_Y) | BIT(REL_WHEEL);
++ input_set_abs_params(ptr, ABS_X, 0, XENFB_WIDTH, 0, 0);
++ input_set_abs_params(ptr, ABS_Y, 0, XENFB_HEIGHT, 0, 0);
+
++ ret = input_register_device(ptr);
++ if (ret) {
++ input_free_device(ptr);
++ xenbus_dev_fatal(dev, ret, "input_register_device(ptr)");
++ goto error;
++ }
++ info->ptr = ptr;
+
++ ret = xenkbd_connect_backend(dev, info);
++ if (ret < 0)
++ goto error;
+
-+/******************************************************************
-+ * MISCELLANEOUS SETUP / TEARDOWN / DEBUGGING
-+ */
++ return 0;
+
++ error_nomem:
++ ret = -ENOMEM;
++ xenbus_dev_fatal(dev, ret, "allocating device memory");
++ error:
++ xenkbd_remove(dev);
++ return ret;
++}
+
-+static void make_response(blkif_t *blkif, u64 id,
-+ unsigned short op, int st)
++static int xenkbd_resume(struct xenbus_device *dev)
+{
-+ blkif_response_t resp;
-+ unsigned long flags;
-+ blkif_back_rings_t *blk_rings = &blkif->blk_rings;
-+ int more_to_do = 0;
-+ int notify;
-+
-+ resp.id = id;
-+ resp.operation = op;
-+ resp.status = st;
++ struct xenkbd_info *info = dev->dev.driver_data;
+
-+ spin_lock_irqsave(&blkif->blk_ring_lock, flags);
-+ /* Place on the response ring for the relevant domain. */
-+ switch (blkif->blk_protocol) {
-+ case BLKIF_PROTOCOL_NATIVE:
-+ memcpy(RING_GET_RESPONSE(&blk_rings->native,
-+ blk_rings->native.rsp_prod_pvt),
-+ &resp, sizeof(resp));
-+ break;
-+ case BLKIF_PROTOCOL_X86_32:
-+ memcpy(RING_GET_RESPONSE(&blk_rings->x86_32,
-+ blk_rings->x86_32.rsp_prod_pvt),
-+ &resp, sizeof(resp));
-+ break;
-+ case BLKIF_PROTOCOL_X86_64:
-+ memcpy(RING_GET_RESPONSE(&blk_rings->x86_64,
-+ blk_rings->x86_64.rsp_prod_pvt),
-+ &resp, sizeof(resp));
-+ break;
-+ default:
-+ BUG();
-+ }
-+ blk_rings->common.rsp_prod_pvt++;
-+ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&blk_rings->common, notify);
++ xenkbd_disconnect_backend(info);
++ info->page->in_cons = info->page->in_prod = 0;
++ info->page->out_cons = info->page->out_prod = 0;
++ return xenkbd_connect_backend(dev, info);
++}
+
-+ if (blk_rings->common.rsp_prod_pvt == blk_rings->common.req_cons) {
-+ /*
-+ * Tail check for pending requests. Allows frontend to avoid
-+ * notifications if requests are already in flight (lower
-+ * overheads and promotes batching).
-+ */
-+ RING_FINAL_CHECK_FOR_REQUESTS(&blk_rings->common, more_to_do);
-+ } else if (RING_HAS_UNCONSUMED_REQUESTS(&blk_rings->common)) {
-+ more_to_do = 1;
-+ }
++static int xenkbd_remove(struct xenbus_device *dev)
++{
++ struct xenkbd_info *info = dev->dev.driver_data;
+
-+ spin_unlock_irqrestore(&blkif->blk_ring_lock, flags);
-+ if (more_to_do)
-+ blkif_notify_work(blkif);
-+ if (notify)
-+ notify_remote_via_irq(blkif->irq);
++ xenkbd_disconnect_backend(info);
++ input_unregister_device(info->kbd);
++ input_unregister_device(info->ptr);
++ free_page((unsigned long)info->page);
++ kfree(info);
++ return 0;
+}
+
-+static int __init blkif_init(void)
++static int xenkbd_connect_backend(struct xenbus_device *dev,
++ struct xenkbd_info *info)
+{
-+ int i, ret;
-+ struct class *class;
++ int ret;
++ struct xenbus_transaction xbt;
+
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++ ret = bind_listening_port_to_irqhandler(
++ dev->otherend_id, input_handler, 0, "xenkbd", info);
++ if (ret < 0) {
++ xenbus_dev_fatal(dev, ret,
++ "bind_listening_port_to_irqhandler");
++ return ret;
++ }
++ info->irq = ret;
+
-+ INIT_LIST_HEAD(&pending_free);
-+ for(i = 0; i < 2; i++) {
-+ ret = req_increase();
-+ if (ret)
-+ break;
++ again:
++ ret = xenbus_transaction_start(&xbt);
++ if (ret) {
++ xenbus_dev_fatal(dev, ret, "starting transaction");
++ return ret;
+ }
-+ if (i == 0)
++ ret = xenbus_printf(xbt, dev->nodename, "page-ref", "%lu",
++ virt_to_mfn(info->page));
++ if (ret)
++ goto error_xenbus;
++ ret = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
++ irq_to_evtchn_port(info->irq));
++ if (ret)
++ goto error_xenbus;
++ ret = xenbus_transaction_end(xbt, 0);
++ if (ret) {
++ if (ret == -EAGAIN)
++ goto again;
++ xenbus_dev_fatal(dev, ret, "completing transaction");
+ return ret;
++ }
+
-+ tap_blkif_interface_init();
-+
-+ alloc_pending_reqs = 0;
++ xenbus_switch_state(dev, XenbusStateInitialised);
++ return 0;
+
-+ tap_blkif_xenbus_init();
++ error_xenbus:
++ xenbus_transaction_end(xbt, 1);
++ xenbus_dev_fatal(dev, ret, "writing xenstore");
++ return ret;
++}
+
-+ /* Dynamically allocate a major for this device */
-+ ret = register_chrdev(0, "blktap", &blktap_fops);
++static void xenkbd_disconnect_backend(struct xenkbd_info *info)
++{
++ if (info->irq >= 0)
++ unbind_from_irqhandler(info->irq, info);
++ info->irq = -1;
++}
+
-+ if (ret < 0) {
-+ WPRINTK("Couldn't register /dev/xen/blktap\n");
-+ return -ENOMEM;
-+ }
-+
-+ blktap_major = ret;
++static void xenkbd_backend_changed(struct xenbus_device *dev,
++ enum xenbus_state backend_state)
++{
++ struct xenkbd_info *info = dev->dev.driver_data;
++ int ret, val;
+
-+ /* tapfds[0] is always NULL */
-+ blktap_next_minor++;
++ switch (backend_state) {
++ case XenbusStateInitialising:
++ case XenbusStateInitialised:
++ case XenbusStateReconfiguring:
++ case XenbusStateReconfigured:
++ case XenbusStateUnknown:
++ case XenbusStateClosed:
++ break;
+
-+ DPRINTK("Created misc_dev [/dev/xen/blktap%d]\n",i);
++ case XenbusStateInitWait:
++ InitWait:
++ ret = xenbus_scanf(XBT_NIL, info->xbdev->otherend,
++ "feature-abs-pointer", "%d", &val);
++ if (ret < 0)
++ val = 0;
++ if (val) {
++ ret = xenbus_printf(XBT_NIL, info->xbdev->nodename,
++ "request-abs-pointer", "1");
++ if (ret)
++ ; /* FIXME */
++ }
++ xenbus_switch_state(dev, XenbusStateConnected);
++ break;
+
-+ /* Make sure the xen class exists */
-+ if ((class = get_xen_class()) != NULL) {
++ case XenbusStateConnected:
+ /*
-+ * This will allow udev to create the blktap ctrl device.
-+ * We only want to create blktap0 first. We don't want
-+ * to flood the sysfs system with needless blktap devices.
-+ * We only create the device when a request of a new device is
-+ * made.
++ * Work around xenbus race condition: If backend goes
++ * through InitWait to Connected fast enough, we can
++ * get Connected twice here.
+ */
-+ class_device_create(class, NULL,
-+ MKDEV(blktap_major, 0), NULL,
-+ "blktap0");
-+ } else {
-+ /* this is bad, but not fatal */
-+ WPRINTK("blktap: sysfs xen_class not created\n");
++ if (dev->state != XenbusStateConnected)
++ goto InitWait; /* no InitWait seen yet, fudge it */
++
++ /* Set input abs params to match backend screen res */
++ if (xenbus_scanf(XBT_NIL, info->xbdev->otherend,
++ "width", "%d", &val) > 0 )
++ input_set_abs_params(info->ptr, ABS_X, 0, val, 0, 0);
++
++ if (xenbus_scanf(XBT_NIL, info->xbdev->otherend,
++ "height", "%d", &val) > 0 )
++ input_set_abs_params(info->ptr, ABS_Y, 0, val, 0, 0);
++
++ break;
++
++ case XenbusStateClosing:
++ xenbus_frontend_closed(dev);
++ break;
+ }
++}
++
++static const struct xenbus_device_id xenkbd_ids[] = {
++ { "vkbd" },
++ { "" }
++};
++MODULE_ALIAS("xen:vkbd");
++
++static struct xenbus_driver xenkbd_driver = {
++ .name = "vkbd",
++ .owner = THIS_MODULE,
++ .ids = xenkbd_ids,
++ .probe = xenkbd_probe,
++ .remove = xenkbd_remove,
++ .resume = xenkbd_resume,
++ .otherend_changed = xenkbd_backend_changed,
++};
+
-+ DPRINTK("Blktap device successfully created\n");
++static int __init xenkbd_init(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+ return 0;
++ /* Nothing to do if running in dom0. */
++ if (is_initial_xendomain())
++ return -ENODEV;
++
++ return xenbus_register_frontend(&xenkbd_driver);
+}
+
-+module_init(blkif_init);
++static void __exit xenkbd_cleanup(void)
++{
++ return xenbus_unregister_driver(&xenkbd_driver);
++}
+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blktap/common.h tmp-linux-2.6-xen.patch/drivers/xen/blktap/common.h
---- pristine-linux-2.6.18.2/drivers/xen/blktap/common.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blktap/common.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,121 @@
-+/*
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
++module_init(xenkbd_init);
++module_exit(xenkbd_cleanup);
++
++MODULE_LICENSE("GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/gntdev/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/gntdev/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1 @@
++obj-$(CONFIG_XEN_GRANT_DEV) := gntdev.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/gntdev/gntdev.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/gntdev/gntdev.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1074 @@
++/******************************************************************************
++ * gntdev.c
+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
++ * Device for accessing (in user-space) pages that have been granted by other
++ * domains.
++ *
++ * Copyright (c) 2006-2007, D G Murray.
+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ */
+
-+#ifndef __BLKIF__BACKEND__COMMON_H__
-+#define __BLKIF__BACKEND__COMMON_H__
-+
-+#include <linux/version.h>
++#include <asm/atomic.h>
+#include <linux/module.h>
-+#include <linux/interrupt.h>
-+#include <linux/slab.h>
-+#include <linux/blkdev.h>
-+#include <linux/vmalloc.h>
++#include <linux/kernel.h>
++#include <linux/init.h>
++#include <linux/fs.h>
++#include <linux/device.h>
++#include <linux/mm.h>
++#include <linux/mman.h>
++#include <asm/uaccess.h>
+#include <asm/io.h>
-+#include <asm/setup.h>
-+#include <asm/pgalloc.h>
-+#include <xen/evtchn.h>
-+#include <asm/hypervisor.h>
-+#include <xen/blkif.h>
+#include <xen/gnttab.h>
++#include <asm/hypervisor.h>
++#include <xen/balloon.h>
++#include <xen/evtchn.h>
+#include <xen/driver_util.h>
+
-+#define DPRINTK(_f, _a...) pr_debug("(file=%s, line=%d) " _f, \
-+ __FILE__ , __LINE__ , ## _a )
++#include <linux/types.h>
++#include <xen/public/gntdev.h>
+
-+#define WPRINTK(fmt, args...) printk(KERN_WARNING "blk_tap: " fmt, ##args)
+
-+struct backend_info;
++#define DRIVER_AUTHOR "Derek G. Murray <Derek.Murray at cl.cam.ac.uk>"
++#define DRIVER_DESC "User-space granted page access driver"
+
-+typedef struct blkif_st {
-+ /* Unique identifier for this interface. */
-+ domid_t domid;
-+ unsigned int handle;
-+ /* Physical parameters of the comms window. */
-+ unsigned int irq;
-+ /* Comms information. */
-+ enum blkif_protocol blk_protocol;
-+ blkif_back_rings_t blk_rings;
-+ struct vm_struct *blk_ring_area;
-+ /* Back pointer to the backend_info. */
-+ struct backend_info *be;
-+ /* Private fields. */
-+ spinlock_t blk_ring_lock;
-+ atomic_t refcnt;
++MODULE_LICENSE("GPL");
++MODULE_AUTHOR(DRIVER_AUTHOR);
++MODULE_DESCRIPTION(DRIVER_DESC);
+
-+ wait_queue_head_t wq;
-+ struct task_struct *xenblkd;
-+ unsigned int waiting_reqs;
-+ request_queue_t *plug;
++#define MAX_GRANTS_LIMIT 1024
++#define DEFAULT_MAX_GRANTS 128
+
-+ /* statistics */
-+ unsigned long st_print;
-+ int st_rd_req;
-+ int st_wr_req;
-+ int st_oo_req;
-+ int st_rd_sect;
-+ int st_wr_sect;
++/* A slot can be in one of three states:
++ *
++ * 0. GNTDEV_SLOT_INVALID:
++ * This slot is not associated with a grant reference, and is therefore free
++ * to be overwritten by a new grant reference.
++ *
++ * 1. GNTDEV_SLOT_NOT_YET_MAPPED:
++ * This slot is associated with a grant reference (via the
++ * IOCTL_GNTDEV_MAP_GRANT_REF ioctl), but it has not yet been mmap()-ed.
++ *
++ * 2. GNTDEV_SLOT_MAPPED:
++ * This slot is associated with a grant reference, and has been mmap()-ed.
++ */
++typedef enum gntdev_slot_state {
++ GNTDEV_SLOT_INVALID = 0,
++ GNTDEV_SLOT_NOT_YET_MAPPED,
++ GNTDEV_SLOT_MAPPED
++} gntdev_slot_state_t;
+
-+ wait_queue_head_t waiting_to_free;
++#define GNTDEV_INVALID_HANDLE -1
++#define GNTDEV_FREE_LIST_INVALID -1
++/* Each opened instance of gntdev is associated with a list of grants,
++ * represented by an array of elements of the following type,
++ * gntdev_grant_info_t.
++ */
++typedef struct gntdev_grant_info {
++ gntdev_slot_state_t state;
++ union {
++ uint32_t free_list_index;
++ struct {
++ domid_t domid;
++ grant_ref_t ref;
++ grant_handle_t kernel_handle;
++ grant_handle_t user_handle;
++ uint64_t dev_bus_addr;
++ } valid;
++ } u;
++} gntdev_grant_info_t;
+
-+ grant_handle_t shmem_handle;
-+ grant_ref_t shmem_ref;
-+
-+ int dev_num;
-+ uint64_t sectors;
-+} blkif_t;
++/* Private data structure, which is stored in the file pointer for files
++ * associated with this device.
++ */
++typedef struct gntdev_file_private_data {
++
++ /* Array of grant information. */
++ gntdev_grant_info_t *grants;
++ uint32_t grants_size;
+
-+blkif_t *tap_alloc_blkif(domid_t domid);
-+void tap_blkif_free(blkif_t *blkif);
-+int tap_blkif_map(blkif_t *blkif, unsigned long shared_page,
-+ unsigned int evtchn);
-+void tap_blkif_unmap(blkif_t *blkif);
++ /* Read/write semaphore used to protect the grants array. */
++ struct rw_semaphore grants_sem;
+
-+#define blkif_get(_b) (atomic_inc(&(_b)->refcnt))
-+#define blkif_put(_b) \
-+ do { \
-+ if (atomic_dec_and_test(&(_b)->refcnt)) \
-+ wake_up(&(_b)->waiting_to_free);\
-+ } while (0)
++ /* An array of indices of free slots in the grants array.
++ * N.B. An entry in this list may temporarily have the value
++ * GNTDEV_FREE_LIST_INVALID if the corresponding slot has been removed
++ * from the list by the contiguous allocator, but the list has not yet
++ * been compressed. However, this is not visible across invocations of
++ * the device.
++ */
++ int32_t *free_list;
++
++ /* The number of free slots in the grants array. */
++ uint32_t free_list_size;
+
++ /* Read/write semaphore used to protect the free list. */
++ struct rw_semaphore free_list_sem;
++
++ /* Index of the next slot after the most recent contiguous allocation,
++ * for use in a next-fit allocator.
++ */
++ uint32_t next_fit_index;
+
-+struct phys_req {
-+ unsigned short dev;
-+ unsigned short nr_sects;
-+ struct block_device *bdev;
-+ blkif_sector_t sector_number;
-+};
++ /* Used to map grants into the kernel, before mapping them into user
++ * space.
++ */
++ struct page **foreign_pages;
+
-+void tap_blkif_interface_init(void);
++} gntdev_file_private_data_t;
+
-+void tap_blkif_xenbus_init(void);
++/* Module lifecycle operations. */
++static int __init gntdev_init(void);
++static void __exit gntdev_exit(void);
+
-+irqreturn_t tap_blkif_be_int(int irq, void *dev_id, struct pt_regs *regs);
-+int tap_blkif_schedule(void *arg);
++module_init(gntdev_init);
++module_exit(gntdev_exit);
+
-+int dom_to_devid(domid_t domid, int xenbus_id, blkif_t *blkif);
-+void signal_tapdisk(int idx);
++/* File operations. */
++static int gntdev_open(struct inode *inode, struct file *flip);
++static int gntdev_release(struct inode *inode, struct file *flip);
++static int gntdev_mmap(struct file *flip, struct vm_area_struct *vma);
++static long gntdev_ioctl(struct file *flip,
++ unsigned int cmd, unsigned long arg);
+
-+#endif /* __BLKIF__BACKEND__COMMON_H__ */
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blktap/interface.c tmp-linux-2.6-xen.patch/drivers/xen/blktap/interface.c
---- pristine-linux-2.6.18.2/drivers/xen/blktap/interface.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blktap/interface.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,174 @@
-+/******************************************************************************
-+ * drivers/xen/blktap/interface.c
-+ *
-+ * Block-device interface management.
-+ *
-+ * Copyright (c) 2004, Keir Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
++static const struct file_operations gntdev_fops = {
++ .owner = THIS_MODULE,
++ .open = gntdev_open,
++ .release = gntdev_release,
++ .mmap = gntdev_mmap,
++ .unlocked_ioctl = gntdev_ioctl
++};
+
-+ */
++/* VM operations. */
++static void gntdev_vma_close(struct vm_area_struct *vma);
++static pte_t gntdev_clear_pte(struct vm_area_struct *vma, unsigned long addr,
++ pte_t *ptep, int is_fullmm);
+
-+#include "common.h"
-+#include <xen/evtchn.h>
++static struct vm_operations_struct gntdev_vmops = {
++ .close = gntdev_vma_close,
++ .zap_pte = gntdev_clear_pte
++};
+
-+static kmem_cache_t *blkif_cachep;
++/* Global variables. */
+
-+blkif_t *tap_alloc_blkif(domid_t domid)
-+{
-+ blkif_t *blkif;
++/* The driver major number, for use when unregistering the driver. */
++static int gntdev_major;
+
-+ blkif = kmem_cache_alloc(blkif_cachep, GFP_KERNEL);
-+ if (!blkif)
-+ return ERR_PTR(-ENOMEM);
++#define GNTDEV_NAME "gntdev"
+
-+ memset(blkif, 0, sizeof(*blkif));
-+ blkif->domid = domid;
-+ spin_lock_init(&blkif->blk_ring_lock);
-+ atomic_set(&blkif->refcnt, 1);
-+ init_waitqueue_head(&blkif->wq);
-+ blkif->st_print = jiffies;
-+ init_waitqueue_head(&blkif->waiting_to_free);
++/* Memory mapping functions
++ * ------------------------
++ *
++ * Every granted page is mapped into both kernel and user space, and the two
++ * following functions return the respective virtual addresses of these pages.
++ *
++ * When shadow paging is disabled, the granted page is mapped directly into
++ * user space; when it is enabled, it is mapped into the kernel and remapped
++ * into user space using vm_insert_page() (see gntdev_mmap(), below).
++ */
+
-+ return blkif;
++/* Returns the virtual address (in user space) of the @page_index'th page
++ * in the given VM area.
++ */
++static inline unsigned long get_user_vaddr (struct vm_area_struct *vma,
++ int page_index)
++{
++ return (unsigned long) vma->vm_start + (page_index << PAGE_SHIFT);
+}
+
-+static int map_frontend_page(blkif_t *blkif, unsigned long shared_page)
++/* Returns the virtual address (in kernel space) of the @slot_index'th page
++ * mapped by the gntdev instance that owns the given private data struct.
++ */
++static inline unsigned long get_kernel_vaddr (gntdev_file_private_data_t *priv,
++ int slot_index)
+{
-+ struct gnttab_map_grant_ref op;
++ unsigned long pfn;
++ void *kaddr;
++ pfn = page_to_pfn(priv->foreign_pages[slot_index]);
++ kaddr = pfn_to_kaddr(pfn);
++ return (unsigned long) kaddr;
++}
+
-+ gnttab_set_map_op(&op, (unsigned long)blkif->blk_ring_area->addr,
-+ GNTMAP_host_map, shared_page, blkif->domid);
++/* Helper functions. */
+
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
-+ BUG();
++/* Adds information about a grant reference to the list of grants in the file's
++ * private data structure. Returns non-zero on failure. On success, sets the
++ * value of *offset to the offset that should be mmap()-ed in order to map the
++ * grant reference.
++ */
++static int add_grant_reference(struct file *flip,
++ struct ioctl_gntdev_grant_ref *op,
++ uint64_t *offset)
++{
++ gntdev_file_private_data_t *private_data
++ = (gntdev_file_private_data_t *) flip->private_data;
+
-+ if (op.status) {
-+ DPRINTK(" Grant table operation failure !\n");
-+ return op.status;
-+ }
++ uint32_t slot_index;
+
-+ blkif->shmem_ref = shared_page;
-+ blkif->shmem_handle = op.handle;
++ if (unlikely(private_data->free_list_size == 0)) {
++ return -ENOMEM;
++ }
+
-+ return 0;
-+}
++ slot_index = private_data->free_list[--private_data->free_list_size];
++ private_data->free_list[private_data->free_list_size]
++ = GNTDEV_FREE_LIST_INVALID;
+
-+static void unmap_frontend_page(blkif_t *blkif)
-+{
-+ struct gnttab_unmap_grant_ref op;
++ /* Copy the grant information into file's private data. */
++ private_data->grants[slot_index].state = GNTDEV_SLOT_NOT_YET_MAPPED;
++ private_data->grants[slot_index].u.valid.domid = op->domid;
++ private_data->grants[slot_index].u.valid.ref = op->ref;
+
-+ gnttab_set_unmap_op(&op, (unsigned long)blkif->blk_ring_area->addr,
-+ GNTMAP_host_map, blkif->shmem_handle);
++ /* The offset is calculated as the index of the chosen entry in the
++ * file's private data's array of grant information. This is then
++ * shifted to give an offset into the virtual "file address space".
++ */
++ *offset = slot_index << PAGE_SHIFT;
+
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
-+ BUG();
++ return 0;
+}
+
-+int tap_blkif_map(blkif_t *blkif, unsigned long shared_page,
-+ unsigned int evtchn)
++/* Adds the @count grant references to the contiguous range in the slot array
++ * beginning at @first_slot. It is assumed that @first_slot was returned by a
++ * previous invocation of find_contiguous_free_range(), during the same
++ * invocation of the driver.
++ */
++static int add_grant_references(struct file *flip,
++ int count,
++ struct ioctl_gntdev_grant_ref *ops,
++ uint32_t first_slot)
+{
-+ int err;
-+
-+ /* Already connected through? */
-+ if (blkif->irq)
-+ return 0;
++ gntdev_file_private_data_t *private_data
++ = (gntdev_file_private_data_t *) flip->private_data;
++ int i;
++
++ for (i = 0; i < count; ++i) {
+
-+ if ( (blkif->blk_ring_area = alloc_vm_area(PAGE_SIZE)) == NULL )
-+ return -ENOMEM;
++ /* First, mark the slot's entry in the free list as invalid. */
++ int free_list_index =
++ private_data->grants[first_slot+i].u.free_list_index;
++ private_data->free_list[free_list_index] =
++ GNTDEV_FREE_LIST_INVALID;
+
-+ err = map_frontend_page(blkif, shared_page);
-+ if (err) {
-+ free_vm_area(blkif->blk_ring_area);
-+ return err;
++ /* Now, update the slot. */
++ private_data->grants[first_slot+i].state =
++ GNTDEV_SLOT_NOT_YET_MAPPED;
++ private_data->grants[first_slot+i].u.valid.domid =
++ ops[i].domid;
++ private_data->grants[first_slot+i].u.valid.ref = ops[i].ref;
+ }
+
-+ switch (blkif->blk_protocol) {
-+ case BLKIF_PROTOCOL_NATIVE:
-+ {
-+ blkif_sring_t *sring;
-+ sring = (blkif_sring_t *)blkif->blk_ring_area->addr;
-+ BACK_RING_INIT(&blkif->blk_rings.native, sring, PAGE_SIZE);
-+ break;
-+ }
-+ case BLKIF_PROTOCOL_X86_32:
-+ {
-+ blkif_x86_32_sring_t *sring_x86_32;
-+ sring_x86_32 = (blkif_x86_32_sring_t *)blkif->blk_ring_area->addr;
-+ BACK_RING_INIT(&blkif->blk_rings.x86_32, sring_x86_32, PAGE_SIZE);
-+ break;
-+ }
-+ case BLKIF_PROTOCOL_X86_64:
-+ {
-+ blkif_x86_64_sring_t *sring_x86_64;
-+ sring_x86_64 = (blkif_x86_64_sring_t *)blkif->blk_ring_area->addr;
-+ BACK_RING_INIT(&blkif->blk_rings.x86_64, sring_x86_64, PAGE_SIZE);
-+ break;
-+ }
-+ default:
-+ BUG();
-+ }
++ return 0;
++}
+
-+ err = bind_interdomain_evtchn_to_irqhandler(
-+ blkif->domid, evtchn, tap_blkif_be_int,
-+ 0, "blkif-backend", blkif);
-+ if (err < 0) {
-+ unmap_frontend_page(blkif);
-+ free_vm_area(blkif->blk_ring_area);
-+ blkif->blk_rings.common.sring = NULL;
-+ return err;
++/* Scans through the free list for @flip, removing entries that are marked as
++ * GNTDEV_SLOT_INVALID. This will reduce the recorded size of the free list to
++ * the number of valid entries.
++ */
++static void compress_free_list(struct file *flip)
++{
++ gntdev_file_private_data_t *private_data
++ = (gntdev_file_private_data_t *) flip->private_data;
++ int i, j = 0, old_size, slot_index;
++
++ old_size = private_data->free_list_size;
++ for (i = 0; i < old_size; ++i) {
++ if (private_data->free_list[i] != GNTDEV_FREE_LIST_INVALID) {
++ if (i > j) {
++ slot_index = private_data->free_list[i];
++ private_data->free_list[j] = slot_index;
++ private_data->grants[slot_index].u
++ .free_list_index = j;
++ private_data->free_list[i]
++ = GNTDEV_FREE_LIST_INVALID;
++ }
++ ++j;
++ } else {
++ --private_data->free_list_size;
++ }
+ }
-+ blkif->irq = err;
-+
-+ return 0;
+}
+
-+void tap_blkif_unmap(blkif_t *blkif)
++/* Searches the grant array in the private data of @flip for a range of
++ * @num_slots contiguous slots in the GNTDEV_SLOT_INVALID state.
++ *
++ * Returns the index of the first slot if a range is found, otherwise -ENOMEM.
++ */
++static int find_contiguous_free_range(struct file *flip,
++ uint32_t num_slots)
+{
-+ if (blkif->irq) {
-+ unbind_from_irqhandler(blkif->irq, blkif);
-+ blkif->irq = 0;
++ gntdev_file_private_data_t *private_data
++ = (gntdev_file_private_data_t *) flip->private_data;
++
++ int i;
++ int start_index = private_data->next_fit_index;
++ int range_start = 0, range_length;
++
++ if (private_data->free_list_size < num_slots) {
++ return -ENOMEM;
+ }
-+ if (blkif->blk_rings.common.sring) {
-+ unmap_frontend_page(blkif);
-+ free_vm_area(blkif->blk_ring_area);
-+ blkif->blk_rings.common.sring = NULL;
++
++ /* First search from the start_index to the end of the array. */
++ range_length = 0;
++ for (i = start_index; i < private_data->grants_size; ++i) {
++ if (private_data->grants[i].state == GNTDEV_SLOT_INVALID) {
++ if (range_length == 0) {
++ range_start = i;
++ }
++ ++range_length;
++ if (range_length == num_slots) {
++ return range_start;
++ }
++ }
++ }
++
++ /* Now search from the start of the array to the start_index. */
++ range_length = 0;
++ for (i = 0; i < start_index; ++i) {
++ if (private_data->grants[i].state == GNTDEV_SLOT_INVALID) {
++ if (range_length == 0) {
++ range_start = i;
++ }
++ ++range_length;
++ if (range_length == num_slots) {
++ return range_start;
++ }
++ }
+ }
++
++ return -ENOMEM;
+}
+
-+void tap_blkif_free(blkif_t *blkif)
-+{
-+ atomic_dec(&blkif->refcnt);
-+ wait_event(blkif->waiting_to_free, atomic_read(&blkif->refcnt) == 0);
-+
-+ tap_blkif_unmap(blkif);
-+ kmem_cache_free(blkif_cachep, blkif);
-+}
-+
-+void __init tap_blkif_interface_init(void)
++static int init_private_data(gntdev_file_private_data_t *priv,
++ uint32_t max_grants)
+{
-+ blkif_cachep = kmem_cache_create("blktapif_cache", sizeof(blkif_t),
-+ 0, 0, NULL, NULL);
-+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blktap/Makefile tmp-linux-2.6-xen.patch/drivers/xen/blktap/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/blktap/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blktap/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,5 @@
-+LINUXINCLUDE += -I../xen/include/public/io
++ int i;
+
-+obj-$(CONFIG_XEN_BLKDEV_TAP) := xenblktap.o
++ /* Allocate space for the kernel-mapping of granted pages. */
++ priv->foreign_pages =
++ alloc_empty_pages_and_pagevec(max_grants);
++ if (!priv->foreign_pages)
++ goto nomem_out;
+
-+xenblktap-y := xenbus.o interface.o blktap.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/blktap/xenbus.c tmp-linux-2.6-xen.patch/drivers/xen/blktap/xenbus.c
---- pristine-linux-2.6.18.2/drivers/xen/blktap/xenbus.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/blktap/xenbus.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,477 @@
-+/* drivers/xen/blktap/xenbus.c
-+ *
-+ * Xenbus code for blktap
-+ *
-+ * Copyright (c) 2004-2005, Andrew Warfield and Julian Chesterfield
-+ *
-+ * Based on the blkback xenbus code:
-+ *
-+ * Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
-+ * Copyright (C) 2005 XenSource Ltd
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ /* Allocate the grant list and free-list. */
++ priv->grants = kmalloc(max_grants * sizeof(gntdev_grant_info_t),
++ GFP_KERNEL);
++ if (!priv->grants)
++ goto nomem_out2;
++ priv->free_list = kmalloc(max_grants * sizeof(int32_t), GFP_KERNEL);
++ if (!priv->free_list)
++ goto nomem_out3;
+
-+#include <stdarg.h>
-+#include <linux/module.h>
-+#include <linux/kthread.h>
-+#include <xen/xenbus.h>
-+#include "common.h"
++ /* Initialise the free-list, which contains all slots at first. */
++ for (i = 0; i < max_grants; ++i) {
++ priv->free_list[max_grants - i - 1] = i;
++ priv->grants[i].state = GNTDEV_SLOT_INVALID;
++ priv->grants[i].u.free_list_index = max_grants - i - 1;
++ }
++ priv->grants_size = max_grants;
++ priv->free_list_size = max_grants;
++ priv->next_fit_index = 0;
+
++ return 0;
+
-+struct backend_info
-+{
-+ struct xenbus_device *dev;
-+ blkif_t *blkif;
-+ struct xenbus_watch backend_watch;
-+ int xenbus_id;
-+ int group_added;
-+};
++nomem_out3:
++ kfree(priv->grants);
++nomem_out2:
++ free_empty_pages_and_pagevec(priv->foreign_pages, max_grants);
++nomem_out:
++ return -ENOMEM;
+
++}
+
-+static void connect(struct backend_info *);
-+static int connect_ring(struct backend_info *);
-+static int blktap_remove(struct xenbus_device *dev);
-+static int blktap_probe(struct xenbus_device *dev,
-+ const struct xenbus_device_id *id);
-+static void tap_backend_changed(struct xenbus_watch *, const char **,
-+ unsigned int);
-+static void tap_frontend_changed(struct xenbus_device *dev,
-+ enum xenbus_state frontend_state);
++/* Interface functions. */
+
-+static int strsep_len(const char *str, char c, unsigned int len)
++/* Initialises the driver. Called when the module is loaded. */
++static int __init gntdev_init(void)
+{
-+ unsigned int i;
++ struct class *class;
++ struct class_device *device;
+
-+ for (i = 0; str[i]; i++)
-+ if (str[i] == c) {
-+ if (len == 0)
-+ return i;
-+ len--;
-+ }
-+ return (len == 0) ? i : -ERANGE;
-+}
++ if (!is_running_on_xen()) {
++ printk(KERN_ERR "You must be running Xen to use gntdev\n");
++ return -ENODEV;
++ }
+
-+static long get_id(const char *str)
-+{
-+ int len,end;
-+ const char *ptr;
-+ char *tptr, num[10];
-+
-+ len = strsep_len(str, '/', 2);
-+ end = strlen(str);
-+ if ( (len < 0) || (end < 0) ) return -1;
-+
-+ ptr = str + len + 1;
-+ strncpy(num,ptr,end - len);
-+ tptr = num + (end - (len + 1));
-+ *tptr = '\0';
-+ DPRINTK("Get_id called for %s (%s)\n",str,num);
-+
-+ return simple_strtol(num, NULL, 10);
-+}
++ gntdev_major = register_chrdev(0, GNTDEV_NAME, &gntdev_fops);
++ if (gntdev_major < 0)
++ {
++ printk(KERN_ERR "Could not register gntdev device\n");
++ return -ENOMEM;
++ }
+
-+static int blktap_name(blkif_t *blkif, char *buf)
-+{
-+ char *devpath, *devname;
-+ struct xenbus_device *dev = blkif->be->dev;
++ /* Note that if the sysfs code fails, we will still initialise the
++ * device, and output the major number so that the device can be
++ * created manually using mknod.
++ */
++ if ((class = get_xen_class()) == NULL) {
++ printk(KERN_ERR "Error setting up xen_class\n");
++ printk(KERN_ERR "gntdev created with major number = %d\n",
++ gntdev_major);
++ return 0;
++ }
+
-+ devpath = xenbus_read(XBT_NIL, dev->nodename, "dev", NULL);
-+ if (IS_ERR(devpath))
-+ return PTR_ERR(devpath);
-+
-+ if ((devname = strstr(devpath, "/dev/")) != NULL)
-+ devname += strlen("/dev/");
-+ else
-+ devname = devpath;
++ device = class_device_create(class, NULL, MKDEV(gntdev_major, 0),
++ NULL, GNTDEV_NAME);
++ if (IS_ERR(device)) {
++ printk(KERN_ERR "Error creating gntdev device in xen_class\n");
++ printk(KERN_ERR "gntdev created with major number = %d\n",
++ gntdev_major);
++ return 0;
++ }
+
-+ snprintf(buf, TASK_COMM_LEN, "blktap.%d.%s", blkif->domid, devname);
-+ kfree(devpath);
-+
+ return 0;
+}
+
-+/****************************************************************
-+ * sysfs interface for I/O requests of blktap device
++/* Cleans up and unregisters the driver. Called when the driver is unloaded.
+ */
++static void __exit gntdev_exit(void)
++{
++ struct class *class;
++ if ((class = get_xen_class()) != NULL)
++ class_device_destroy(class, MKDEV(gntdev_major, 0));
++ unregister_chrdev(gntdev_major, GNTDEV_NAME);
++}
+
-+#define VBD_SHOW(name, format, args...) \
-+ static ssize_t show_##name(struct device *_dev, \
-+ struct device_attribute *attr, \
-+ char *buf) \
-+ { \
-+ struct xenbus_device *dev = to_xenbus_device(_dev); \
-+ struct backend_info *be = dev->dev.driver_data; \
-+ \
-+ return sprintf(buf, format, ##args); \
-+ } \
-+ static DEVICE_ATTR(name, S_IRUGO, show_##name, NULL)
++/* Called when the device is opened. */
++static int gntdev_open(struct inode *inode, struct file *flip)
++{
++ gntdev_file_private_data_t *private_data;
+
-+VBD_SHOW(oo_req, "%d\n", be->blkif->st_oo_req);
-+VBD_SHOW(rd_req, "%d\n", be->blkif->st_rd_req);
-+VBD_SHOW(wr_req, "%d\n", be->blkif->st_wr_req);
-+VBD_SHOW(rd_sect, "%d\n", be->blkif->st_rd_sect);
-+VBD_SHOW(wr_sect, "%d\n", be->blkif->st_wr_sect);
++ try_module_get(THIS_MODULE);
+
-+static struct attribute *tapstat_attrs[] = {
-+ &dev_attr_oo_req.attr,
-+ &dev_attr_rd_req.attr,
-+ &dev_attr_wr_req.attr,
-+ &dev_attr_rd_sect.attr,
-+ &dev_attr_wr_sect.attr,
-+ NULL
-+};
++ /* Allocate space for the per-instance private data. */
++ private_data = kmalloc(sizeof(*private_data), GFP_KERNEL);
++ if (!private_data)
++ goto nomem_out;
+
-+static struct attribute_group tapstat_group = {
-+ .name = "statistics",
-+ .attrs = tapstat_attrs,
-+};
++ /* These will be lazily initialised by init_private_data. */
++ private_data->grants = NULL;
++ private_data->free_list = NULL;
++ private_data->foreign_pages = NULL;
+
-+int xentap_sysfs_addif(struct xenbus_device *dev)
-+{
-+ int err;
-+ struct backend_info *be = dev->dev.driver_data;
-+ err = sysfs_create_group(&dev->dev.kobj, &tapstat_group);
-+ if (!err)
-+ be->group_added = 1;
-+ return err;
-+}
++ init_rwsem(&private_data->grants_sem);
++ init_rwsem(&private_data->free_list_sem);
+
-+void xentap_sysfs_delif(struct xenbus_device *dev)
-+{
-+ struct backend_info *be = dev->dev.driver_data;
-+ sysfs_remove_group(&dev->dev.kobj, &tapstat_group);
-+ be->group_added = 0;
++ flip->private_data = private_data;
++
++ return 0;
++
++nomem_out:
++ return -ENOMEM;
+}
+
-+static int blktap_remove(struct xenbus_device *dev)
++/* Called when the device is closed.
++ */
++static int gntdev_release(struct inode *inode, struct file *flip)
+{
-+ struct backend_info *be = dev->dev.driver_data;
-+
-+ if (be->group_added)
-+ xentap_sysfs_delif(be->dev);
-+ if (be->backend_watch.node) {
-+ unregister_xenbus_watch(&be->backend_watch);
-+ kfree(be->backend_watch.node);
-+ be->backend_watch.node = NULL;
-+ }
-+ if (be->blkif) {
-+ if (be->blkif->xenblkd)
-+ kthread_stop(be->blkif->xenblkd);
-+ signal_tapdisk(be->blkif->dev_num);
-+ tap_blkif_free(be->blkif);
-+ be->blkif = NULL;
++ if (flip->private_data) {
++ gntdev_file_private_data_t *private_data =
++ (gntdev_file_private_data_t *) flip->private_data;
++ if (private_data->foreign_pages)
++ free_empty_pages_and_pagevec
++ (private_data->foreign_pages,
++ private_data->grants_size);
++ if (private_data->grants)
++ kfree(private_data->grants);
++ if (private_data->free_list)
++ kfree(private_data->free_list);
++ kfree(private_data);
+ }
-+ kfree(be);
-+ dev->dev.driver_data = NULL;
++ module_put(THIS_MODULE);
+ return 0;
+}
+
-+static void tap_update_blkif_status(blkif_t *blkif)
-+{
-+ int err;
-+ char name[TASK_COMM_LEN];
-+
-+ /* Not ready to connect? */
-+ if(!blkif->irq || !blkif->sectors) {
-+ return;
-+ }
++/* Called when an attempt is made to mmap() the device. The private data from
++ * @flip contains the list of grant references that can be mapped. The vm_pgoff
++ * field of @vma contains the index into that list that refers to the grant
++ * reference that will be mapped. Only mappings that are a multiple of
++ * PAGE_SIZE are handled.
++ */
++static int gntdev_mmap (struct file *flip, struct vm_area_struct *vma)
++{
++ struct gnttab_map_grant_ref op;
++ unsigned long slot_index = vma->vm_pgoff;
++ unsigned long kernel_vaddr, user_vaddr;
++ uint32_t size = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
++ uint64_t ptep;
++ int ret;
++ int flags;
++ int i;
++ struct page *page;
++ gntdev_file_private_data_t *private_data = flip->private_data;
+
-+ /* Already connected? */
-+ if (blkif->be->dev->state == XenbusStateConnected)
-+ return;
++ if (unlikely(!private_data)) {
++ printk(KERN_ERR "File's private data is NULL.\n");
++ return -EINVAL;
++ }
+
-+ /* Attempt to connect: exit if we fail to. */
-+ connect(blkif->be);
-+ if (blkif->be->dev->state != XenbusStateConnected)
-+ return;
++ /* Test to make sure that the grants array has been initialised. */
++ down_read(&private_data->grants_sem);
++ if (unlikely(!private_data->grants)) {
++ up_read(&private_data->grants_sem);
++ printk(KERN_ERR "Attempted to mmap before ioctl.\n");
++ return -EINVAL;
++ }
++ up_read(&private_data->grants_sem);
+
-+ err = blktap_name(blkif, name);
-+ if (err) {
-+ xenbus_dev_error(blkif->be->dev, err, "get blktap dev name");
-+ return;
++ if (unlikely((size <= 0) ||
++ (size + slot_index) > private_data->grants_size)) {
++ printk(KERN_ERR "Invalid number of pages or offset"
++ "(num_pages = %d, first_slot = %ld).\n",
++ size, slot_index);
++ return -ENXIO;
+ }
+
-+ if (!blkif->be->group_added) {
-+ err = xentap_sysfs_addif(blkif->be->dev);
-+ if (err) {
-+ xenbus_dev_fatal(blkif->be->dev, err,
-+ "creating sysfs entries");
-+ return;
-+ }
++ if ((vma->vm_flags & VM_WRITE) && !(vma->vm_flags & VM_SHARED)) {
++ printk(KERN_ERR "Writable mappings must be shared.\n");
++ return -EINVAL;
+ }
+
-+ blkif->xenblkd = kthread_run(tap_blkif_schedule, blkif, name);
-+ if (IS_ERR(blkif->xenblkd)) {
-+ err = PTR_ERR(blkif->xenblkd);
-+ blkif->xenblkd = NULL;
-+ xenbus_dev_fatal(blkif->be->dev, err, "start xenblkd");
-+ WPRINTK("Error starting thread\n");
++ /* Slots must be in the NOT_YET_MAPPED state. */
++ down_write(&private_data->grants_sem);
++ for (i = 0; i < size; ++i) {
++ if (private_data->grants[slot_index + i].state !=
++ GNTDEV_SLOT_NOT_YET_MAPPED) {
++ printk(KERN_ERR "Slot (index = %ld) is in the wrong "
++ "state (%d).\n", slot_index + i,
++ private_data->grants[slot_index + i].state);
++ up_write(&private_data->grants_sem);
++ return -EINVAL;
++ }
+ }
-+}
+
-+/**
-+ * Entry point to this code when a new device is created. Allocate
-+ * the basic structures, and watch the store waiting for the
-+ * user-space program to tell us the physical device info. Switch to
-+ * InitWait.
-+ */
-+static int blktap_probe(struct xenbus_device *dev,
-+ const struct xenbus_device_id *id)
-+{
-+ int err;
-+ struct backend_info *be = kzalloc(sizeof(struct backend_info),
-+ GFP_KERNEL);
-+ if (!be) {
-+ xenbus_dev_fatal(dev, -ENOMEM,
-+ "allocating backend structure");
++ /* Install the hook for unmapping. */
++ vma->vm_ops = &gntdev_vmops;
++
++ /* The VM area contains pages from another VM. */
++ vma->vm_flags |= VM_FOREIGN;
++ vma->vm_private_data = kzalloc(size * sizeof(struct page *),
++ GFP_KERNEL);
++ if (vma->vm_private_data == NULL) {
++ printk(KERN_ERR "Couldn't allocate mapping structure for VM "
++ "area.\n");
+ return -ENOMEM;
+ }
+
-+ be->dev = dev;
-+ dev->dev.driver_data = be;
-+ be->xenbus_id = get_id(dev->nodename);
++ /* This flag prevents Bad PTE errors when the memory is unmapped. */
++ vma->vm_flags |= VM_RESERVED;
+
-+ be->blkif = tap_alloc_blkif(dev->otherend_id);
-+ if (IS_ERR(be->blkif)) {
-+ err = PTR_ERR(be->blkif);
-+ be->blkif = NULL;
-+ xenbus_dev_fatal(dev, err, "creating block interface");
-+ goto fail;
-+ }
++ /* This flag prevents this VM area being copied on a fork(). A better
++ * behaviour might be to explicitly carry out the appropriate mappings
++ * on fork(), but I don't know if there's a hook for this.
++ */
++ vma->vm_flags |= VM_DONTCOPY;
+
-+ /* setup back pointer */
-+ be->blkif->be = be;
-+ be->blkif->sectors = 0;
++#ifdef CONFIG_X86
++ /* This flag ensures that the page tables are not unpinned before the
++ * VM area is unmapped. Therefore Xen still recognises the PTE as
++ * belonging to an L1 pagetable, and the grant unmap operation will
++ * succeed, even if the process does not exit cleanly.
++ */
++ vma->vm_mm->context.has_foreign_mappings = 1;
++#endif
+
-+ /* set a watch on disk info, waiting for userspace to update details*/
-+ err = xenbus_watch_path2(dev, dev->nodename, "info",
-+ &be->backend_watch, tap_backend_changed);
-+ if (err)
-+ goto fail;
-+
-+ err = xenbus_switch_state(dev, XenbusStateInitWait);
-+ if (err)
-+ goto fail;
-+ return 0;
++ for (i = 0; i < size; ++i) {
+
-+fail:
-+ DPRINTK("blktap probe failed\n");
-+ blktap_remove(dev);
-+ return err;
-+}
++ flags = GNTMAP_host_map;
++ if (!(vma->vm_flags & VM_WRITE))
++ flags |= GNTMAP_readonly;
+
++ kernel_vaddr = get_kernel_vaddr(private_data, slot_index + i);
++ user_vaddr = get_user_vaddr(vma, i);
++ page = pfn_to_page(__pa(kernel_vaddr) >> PAGE_SHIFT);
+
-+/**
-+ * Callback received when the user space code has placed the device
-+ * information in xenstore.
-+ */
-+static void tap_backend_changed(struct xenbus_watch *watch,
-+ const char **vec, unsigned int len)
-+{
-+ int err;
-+ unsigned long info;
-+ struct backend_info *be
-+ = container_of(watch, struct backend_info, backend_watch);
-+ struct xenbus_device *dev = be->dev;
-+
-+ /**
-+ * Check to see whether userspace code has opened the image
-+ * and written sector
-+ * and disk info to xenstore
-+ */
-+ err = xenbus_gather(XBT_NIL, dev->nodename, "info", "%lu", &info,
-+ NULL);
-+ if (XENBUS_EXIST_ERR(err))
-+ return;
-+ if (err) {
-+ xenbus_dev_error(dev, err, "getting info");
-+ return;
-+ }
++ gnttab_set_map_op(&op, kernel_vaddr, flags,
++ private_data->grants[slot_index+i]
++ .u.valid.ref,
++ private_data->grants[slot_index+i]
++ .u.valid.domid);
+
-+ DPRINTK("Userspace update on disk info, %lu\n",info);
++ /* Carry out the mapping of the grant reference. */
++ ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref,
++ &op, 1);
++ BUG_ON(ret);
++ if (op.status) {
++ printk(KERN_ERR "Error mapping the grant reference "
++ "into the kernel (%d). domid = %d; ref = %d\n",
++ op.status,
++ private_data->grants[slot_index+i]
++ .u.valid.domid,
++ private_data->grants[slot_index+i]
++ .u.valid.ref);
++ goto undo_map_out;
++ }
+
-+ err = xenbus_gather(XBT_NIL, dev->nodename, "sectors", "%llu",
-+ &be->blkif->sectors, NULL);
++ /* Store a reference to the page that will be mapped into user
++ * space.
++ */
++ ((struct page **) vma->vm_private_data)[i] = page;
+
-+ /* Associate tap dev with domid*/
-+ be->blkif->dev_num = dom_to_devid(be->blkif->domid, be->xenbus_id,
-+ be->blkif);
-+ DPRINTK("Thread started for domid [%d], connecting disk\n",
-+ be->blkif->dev_num);
++ /* Mark mapped page as reserved. */
++ SetPageReserved(page);
+
-+ tap_update_blkif_status(be->blkif);
-+}
++ /* Record the grant handle, for use in the unmap operation. */
++ private_data->grants[slot_index+i].u.valid.kernel_handle =
++ op.handle;
++ private_data->grants[slot_index+i].u.valid.dev_bus_addr =
++ op.dev_bus_addr;
++
++ private_data->grants[slot_index+i].state = GNTDEV_SLOT_MAPPED;
++ private_data->grants[slot_index+i].u.valid.user_handle =
++ GNTDEV_INVALID_HANDLE;
+
-+/**
-+ * Callback received when the frontend's state changes.
-+ */
-+static void tap_frontend_changed(struct xenbus_device *dev,
-+ enum xenbus_state frontend_state)
-+{
-+ struct backend_info *be = dev->dev.driver_data;
-+ int err;
++ /* Now perform the mapping to user space. */
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
+
-+ DPRINTK("\n");
++ /* NOT USING SHADOW PAGE TABLES. */
++ /* In this case, we map the grant(s) straight into user
++ * space.
++ */
+
-+ switch (frontend_state) {
-+ case XenbusStateInitialising:
-+ if (dev->state == XenbusStateClosed) {
-+ printk(KERN_INFO "%s: %s: prepare for reconnect\n",
-+ __FUNCTION__, dev->nodename);
-+ xenbus_switch_state(dev, XenbusStateInitWait);
-+ }
-+ break;
++ /* Get the machine address of the PTE for the user
++ * page.
++ */
++ if ((ret = create_lookup_pte_addr(vma->vm_mm,
++ vma->vm_start
++ + (i << PAGE_SHIFT),
++ &ptep)))
++ {
++ printk(KERN_ERR "Error obtaining PTE pointer "
++ "(%d).\n", ret);
++ goto undo_map_out;
++ }
++
++ /* Configure the map operation. */
++
++ /* The reference is to be used by host CPUs. */
++ flags = GNTMAP_host_map;
++
++ /* Specifies a user space mapping. */
++ flags |= GNTMAP_application_map;
++
++ /* The map request contains the machine address of the
++ * PTE to update.
++ */
++ flags |= GNTMAP_contains_pte;
++
++ if (!(vma->vm_flags & VM_WRITE))
++ flags |= GNTMAP_readonly;
+
-+ case XenbusStateInitialised:
-+ case XenbusStateConnected:
-+ /* Ensure we connect even when two watches fire in
-+ close successsion and we miss the intermediate value
-+ of frontend_state. */
-+ if (dev->state == XenbusStateConnected)
-+ break;
++ gnttab_set_map_op(&op, ptep, flags,
++ private_data->grants[slot_index+i]
++ .u.valid.ref,
++ private_data->grants[slot_index+i]
++ .u.valid.domid);
+
-+ err = connect_ring(be);
-+ if (err)
-+ break;
-+ tap_update_blkif_status(be->blkif);
-+ break;
++ /* Carry out the mapping of the grant reference. */
++ ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref,
++ &op, 1);
++ BUG_ON(ret);
++ if (op.status) {
++ printk(KERN_ERR "Error mapping the grant "
++ "reference into user space (%d). domid "
++ "= %d; ref = %d\n", op.status,
++ private_data->grants[slot_index+i].u
++ .valid.domid,
++ private_data->grants[slot_index+i].u
++ .valid.ref);
++ goto undo_map_out;
++ }
++
++ /* Record the grant handle, for use in the unmap
++ * operation.
++ */
++ private_data->grants[slot_index+i].u.
++ valid.user_handle = op.handle;
+
-+ case XenbusStateClosing:
-+ if (be->blkif->xenblkd) {
-+ kthread_stop(be->blkif->xenblkd);
-+ be->blkif->xenblkd = NULL;
++ /* Update p2m structure with the new mapping. */
++ set_phys_to_machine(__pa(kernel_vaddr) >> PAGE_SHIFT,
++ FOREIGN_FRAME(private_data->
++ grants[slot_index+i]
++ .u.valid.dev_bus_addr
++ >> PAGE_SHIFT));
++ } else {
++ /* USING SHADOW PAGE TABLES. */
++ /* In this case, we simply insert the page into the VM
++ * area. */
++ ret = vm_insert_page(vma, user_vaddr, page);
+ }
-+ xenbus_switch_state(dev, XenbusStateClosing);
-+ break;
-+
-+ case XenbusStateClosed:
-+ xenbus_switch_state(dev, XenbusStateClosed);
-+ if (xenbus_dev_is_online(dev))
-+ break;
-+ /* fall through if not online */
-+ case XenbusStateUnknown:
-+ device_unregister(&dev->dev);
-+ break;
+
-+ default:
-+ xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
-+ frontend_state);
-+ break;
+ }
-+}
+
++ up_write(&private_data->grants_sem);
++ return 0;
+
-+/**
-+ * Switch to Connected state.
-+ */
-+static void connect(struct backend_info *be)
-+{
-+ int err;
-+
-+ struct xenbus_device *dev = be->dev;
++undo_map_out:
++ /* If we have a mapping failure, the unmapping will be taken care of
++ * by do_mmap_pgoff(), which will eventually call gntdev_clear_pte().
++ * All we need to do here is free the vma_private_data.
++ */
++ kfree(vma->vm_private_data);
+
-+ err = xenbus_switch_state(dev, XenbusStateConnected);
-+ if (err)
-+ xenbus_dev_fatal(dev, err, "switching to Connected state",
-+ dev->nodename);
++ /* THIS IS VERY UNPLEASANT: do_mmap_pgoff() will set the vma->vm_file
++ * to NULL on failure. However, we need this in gntdev_clear_pte() to
++ * unmap the grants. Therefore, we smuggle a reference to the file's
++ * private data in the VM area's private data pointer.
++ */
++ vma->vm_private_data = private_data;
++
++ up_write(&private_data->grants_sem);
+
-+ return;
++ return -ENOMEM;
+}
+
-+
-+static int connect_ring(struct backend_info *be)
++static pte_t gntdev_clear_pte(struct vm_area_struct *vma, unsigned long addr,
++ pte_t *ptep, int is_fullmm)
+{
-+ struct xenbus_device *dev = be->dev;
-+ unsigned long ring_ref;
-+ unsigned int evtchn;
-+ char protocol[64];
-+ int err;
-+
-+ DPRINTK("%s\n", dev->otherend);
-+
-+ err = xenbus_gather(XBT_NIL, dev->otherend, "ring-ref", "%lu",
-+ &ring_ref, "event-channel", "%u", &evtchn, NULL);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err,
-+ "reading %s/ring-ref and event-channel",
-+ dev->otherend);
-+ return err;
-+ }
++ int slot_index, ret;
++ pte_t copy;
++ struct gnttab_unmap_grant_ref op;
++ gntdev_file_private_data_t *private_data;
+
-+ be->blkif->blk_protocol = BLKIF_PROTOCOL_NATIVE;
-+ err = xenbus_gather(XBT_NIL, dev->otherend, "protocol",
-+ "%63s", protocol, NULL);
-+ if (err)
-+ strcpy(protocol, "unspecified, assuming native");
-+ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_NATIVE))
-+ be->blkif->blk_protocol = BLKIF_PROTOCOL_NATIVE;
-+ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_X86_32))
-+ be->blkif->blk_protocol = BLKIF_PROTOCOL_X86_32;
-+ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_X86_64))
-+ be->blkif->blk_protocol = BLKIF_PROTOCOL_X86_64;
-+ else {
-+ xenbus_dev_fatal(dev, err, "unknown fe protocol %s", protocol);
-+ return -1;
++ /* THIS IS VERY UNPLEASANT: do_mmap_pgoff() will set the vma->vm_file
++ * to NULL on failure. However, we need this in gntdev_clear_pte() to
++ * unmap the grants. Therefore, we smuggle a reference to the file's
++ * private data in the VM area's private data pointer.
++ */
++ if (vma->vm_file) {
++ private_data = (gntdev_file_private_data_t *)
++ vma->vm_file->private_data;
++ } else if (vma->vm_private_data) {
++ private_data = (gntdev_file_private_data_t *)
++ vma->vm_private_data;
++ } else {
++ private_data = NULL; /* gcc warning */
++ BUG();
+ }
-+ printk(KERN_INFO
-+ "blktap: ring-ref %ld, event-channel %d, protocol %d (%s)\n",
-+ ring_ref, evtchn, be->blkif->blk_protocol, protocol);
+
-+ /* Map the shared frame, irq etc. */
-+ err = tap_blkif_map(be->blkif, ring_ref, evtchn);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "mapping ring-ref %lu port %u",
-+ ring_ref, evtchn);
-+ return err;
-+ }
++ /* Copy the existing value of the PTE for returning. */
++ copy = *ptep;
+
-+ return 0;
-+}
++ /* Calculate the grant relating to this PTE. */
++ slot_index = vma->vm_pgoff + ((addr - vma->vm_start) >> PAGE_SHIFT);
+
++ /* Only unmap grants if the slot has been mapped. This could be being
++ * called from a failing mmap().
++ */
++ if (private_data->grants[slot_index].state == GNTDEV_SLOT_MAPPED) {
+
-+/* ** Driver Registration ** */
++ /* First, we clear the user space mapping, if it has been made.
++ */
++ if (private_data->grants[slot_index].u.valid.user_handle !=
++ GNTDEV_INVALID_HANDLE &&
++ !xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* NOT USING SHADOW PAGE TABLES. */
++ gnttab_set_unmap_op(&op, virt_to_machine(ptep),
++ GNTMAP_contains_pte,
++ private_data->grants[slot_index]
++ .u.valid.user_handle);
++ ret = HYPERVISOR_grant_table_op(
++ GNTTABOP_unmap_grant_ref, &op, 1);
++ BUG_ON(ret);
++ if (op.status)
++ printk("User unmap grant status = %d\n",
++ op.status);
++ } else {
++ /* USING SHADOW PAGE TABLES. */
++ pte_clear_full(vma->vm_mm, addr, ptep, is_fullmm);
++ }
+
++ /* Finally, we unmap the grant from kernel space. */
++ gnttab_set_unmap_op(&op,
++ get_kernel_vaddr(private_data, slot_index),
++ GNTMAP_host_map,
++ private_data->grants[slot_index].u.valid
++ .kernel_handle);
++ ret = HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref,
++ &op, 1);
++ BUG_ON(ret);
++ if (op.status)
++ printk("Kernel unmap grant status = %d\n", op.status);
+
-+static struct xenbus_device_id blktap_ids[] = {
-+ { "tap" },
-+ { "" }
-+};
+
++ /* Return slot to the not-yet-mapped state, so that it may be
++ * mapped again, or removed by a subsequent ioctl.
++ */
++ private_data->grants[slot_index].state =
++ GNTDEV_SLOT_NOT_YET_MAPPED;
+
-+static struct xenbus_driver blktap = {
-+ .name = "tap",
-+ .owner = THIS_MODULE,
-+ .ids = blktap_ids,
-+ .probe = blktap_probe,
-+ .remove = blktap_remove,
-+ .otherend_changed = tap_frontend_changed
-+};
++ /* Invalidate the physical to machine mapping for this page. */
++ set_phys_to_machine(__pa(get_kernel_vaddr(private_data,
++ slot_index))
++ >> PAGE_SHIFT, INVALID_P2M_ENTRY);
+
++ } else {
++ pte_clear_full(vma->vm_mm, addr, ptep, is_fullmm);
++ }
+
-+void tap_blkif_xenbus_init(void)
-+{
-+ xenbus_register_backend(&blktap);
++ return copy;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/char/Makefile tmp-linux-2.6-xen.patch/drivers/xen/char/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/char/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/char/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,2 @@
+
-+obj-y := mem.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/char/mem.c tmp-linux-2.6-xen.patch/drivers/xen/char/mem.c
---- pristine-linux-2.6.18.2/drivers/xen/char/mem.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/char/mem.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,203 @@
-+/*
-+ * Originally from linux/drivers/char/mem.c
-+ *
-+ * Copyright (C) 1991, 1992 Linus Torvalds
-+ *
-+ * Added devfs support.
-+ * Jan-11-1998, C. Scott Ananian <cananian at alumni.princeton.edu>
-+ * Shared /dev/zero mmaping support, Feb 2000, Kanoj Sarcar <kanoj at sgi.com>
++/* "Destructor" for a VM area.
+ */
-+
-+#include <linux/mm.h>
-+#include <linux/miscdevice.h>
-+#include <linux/slab.h>
-+#include <linux/vmalloc.h>
-+#include <linux/mman.h>
-+#include <linux/random.h>
-+#include <linux/init.h>
-+#include <linux/raw.h>
-+#include <linux/tty.h>
-+#include <linux/capability.h>
-+#include <linux/smp_lock.h>
-+#include <linux/ptrace.h>
-+#include <linux/device.h>
-+#include <asm/pgalloc.h>
-+#include <asm/uaccess.h>
-+#include <asm/io.h>
-+#include <asm/hypervisor.h>
-+
-+#ifndef ARCH_HAS_VALID_PHYS_ADDR_RANGE
-+static inline int valid_phys_addr_range(unsigned long addr, size_t count)
-+{
-+ return 1;
++static void gntdev_vma_close(struct vm_area_struct *vma) {
++ if (vma->vm_private_data) {
++ kfree(vma->vm_private_data);
++ }
+}
-+#endif
+
-+/*
-+ * This funcion reads the *physical* memory. The f_pos points directly to the
-+ * memory location.
++/* Called when an ioctl is made on the device.
+ */
-+static ssize_t read_mem(struct file * file, char __user * buf,
-+ size_t count, loff_t *ppos)
++static long gntdev_ioctl(struct file *flip,
++ unsigned int cmd, unsigned long arg)
+{
-+ unsigned long p = *ppos, ignored;
-+ ssize_t read = 0, sz;
-+ void __iomem *v;
-+
-+ if (!valid_phys_addr_range(p, count))
-+ return -EFAULT;
++ int rc = 0;
++ gntdev_file_private_data_t *private_data =
++ (gntdev_file_private_data_t *) flip->private_data;
+
-+ while (count > 0) {
-+ /*
-+ * Handle first page in case it's not aligned
++ /* On the first invocation, we will lazily initialise the grant array
++ * and free-list.
++ */
++ if (unlikely(!private_data->grants)
++ && likely(cmd != IOCTL_GNTDEV_SET_MAX_GRANTS)) {
++ down_write(&private_data->grants_sem);
++
++ if (unlikely(private_data->grants)) {
++ up_write(&private_data->grants_sem);
++ goto private_data_initialised;
++ }
++
++ /* Just use the default. Setting to a non-default is handled
++ * in the ioctl switch.
+ */
-+ if (-p & (PAGE_SIZE - 1))
-+ sz = -p & (PAGE_SIZE - 1);
-+ else
-+ sz = PAGE_SIZE;
-+
-+ sz = min_t(unsigned long, sz, count);
++ rc = init_private_data(private_data, DEFAULT_MAX_GRANTS);
++
++ up_write(&private_data->grants_sem);
+
-+ v = xlate_dev_mem_ptr(p, sz);
-+ if (IS_ERR(v) || v == NULL) {
-+ /*
-+ * Some programs (e.g., dmidecode) groove off into
-+ * weird RAM areas where no tables can possibly exist
-+ * (because Xen will have stomped on them!). These
-+ * programs get rather upset if we let them know that
-+ * Xen failed their access, so we fake out a read of
-+ * all zeroes.
-+ */
-+ if (clear_user(buf, count))
-+ return -EFAULT;
-+ read += count;
-+ break;
++ if (rc) {
++ printk (KERN_ERR "Initialising gntdev private data "
++ "failed.\n");
++ return rc;
+ }
-+
-+ ignored = copy_to_user(buf, v, sz);
-+ xlate_dev_mem_ptr_unmap(v);
-+ if (ignored)
-+ return -EFAULT;
-+ buf += sz;
-+ p += sz;
-+ count -= sz;
-+ read += sz;
+ }
++
++private_data_initialised:
++ switch (cmd) {
++ case IOCTL_GNTDEV_MAP_GRANT_REF:
++ {
++ struct ioctl_gntdev_map_grant_ref op;
++ down_write(&private_data->grants_sem);
++ down_write(&private_data->free_list_sem);
+
-+ *ppos += read;
-+ return read;
-+}
-+
-+static ssize_t write_mem(struct file * file, const char __user * buf,
-+ size_t count, loff_t *ppos)
-+{
-+ unsigned long p = *ppos, ignored;
-+ ssize_t written = 0, sz;
-+ void __iomem *v;
-+
-+ if (!valid_phys_addr_range(p, count))
-+ return -EFAULT;
-+
-+ while (count > 0) {
-+ /*
-+ * Handle first page in case it's not aligned
-+ */
-+ if (-p & (PAGE_SIZE - 1))
-+ sz = -p & (PAGE_SIZE - 1);
-+ else
-+ sz = PAGE_SIZE;
-+
-+ sz = min_t(unsigned long, sz, count);
-+
-+ v = xlate_dev_mem_ptr(p, sz);
-+ if (v == NULL)
-+ break;
-+ if (IS_ERR(v)) {
-+ if (written == 0)
-+ return PTR_ERR(v);
-+ break;
++ if ((rc = copy_from_user(&op, (void __user *) arg,
++ sizeof(op)))) {
++ rc = -EFAULT;
++ goto map_out;
++ }
++ if (unlikely(op.count <= 0)) {
++ rc = -EINVAL;
++ goto map_out;
+ }
+
-+ ignored = copy_from_user(v, buf, sz);
-+ xlate_dev_mem_ptr_unmap(v);
-+ if (ignored) {
-+ written += sz - ignored;
-+ if (written)
-+ break;
-+ return -EFAULT;
++ if (op.count == 1) {
++ if ((rc = add_grant_reference(flip, &op.refs[0],
++ &op.index)) < 0) {
++ printk(KERN_ERR "Adding grant reference "
++ "failed (%d).\n", rc);
++ goto map_out;
++ }
++ } else {
++ struct ioctl_gntdev_grant_ref *refs, *u;
++ refs = kmalloc(op.count * sizeof(*refs), GFP_KERNEL);
++ if (!refs) {
++ rc = -ENOMEM;
++ goto map_out;
++ }
++ u = ((struct ioctl_gntdev_map_grant_ref *)arg)->refs;
++ if ((rc = copy_from_user(refs,
++ (void __user *)u,
++ sizeof(*refs) * op.count))) {
++ printk(KERN_ERR "Copying refs from user failed"
++ " (%d).\n", rc);
++ rc = -EINVAL;
++ goto map_out;
++ }
++ if ((rc = find_contiguous_free_range(flip, op.count))
++ < 0) {
++ printk(KERN_ERR "Finding contiguous range "
++ "failed (%d).\n", rc);
++ kfree(refs);
++ goto map_out;
++ }
++ op.index = rc << PAGE_SHIFT;
++ if ((rc = add_grant_references(flip, op.count,
++ refs, rc))) {
++ printk(KERN_ERR "Adding grant references "
++ "failed (%d).\n", rc);
++ kfree(refs);
++ goto map_out;
++ }
++ compress_free_list(flip);
++ kfree(refs);
+ }
-+ buf += sz;
-+ p += sz;
-+ count -= sz;
-+ written += sz;
++ if ((rc = copy_to_user((void __user *) arg,
++ &op,
++ sizeof(op)))) {
++ printk(KERN_ERR "Copying result back to user failed "
++ "(%d)\n", rc);
++ rc = -EFAULT;
++ goto map_out;
++ }
++ map_out:
++ up_write(&private_data->grants_sem);
++ up_write(&private_data->free_list_sem);
++ return rc;
+ }
++ case IOCTL_GNTDEV_UNMAP_GRANT_REF:
++ {
++ struct ioctl_gntdev_unmap_grant_ref op;
++ int i, start_index;
+
-+ *ppos += written;
-+ return written;
-+}
++ down_write(&private_data->grants_sem);
++ down_write(&private_data->free_list_sem);
+
-+#ifndef ARCH_HAS_DEV_MEM_MMAP_MEM
-+static inline int uncached_access(struct file *file)
-+{
-+ if (file->f_flags & O_SYNC)
-+ return 1;
-+ /* Xen sets correct MTRR type on non-RAM for us. */
-+ return 0;
-+}
++ if ((rc = copy_from_user(&op,
++ (void __user *) arg,
++ sizeof(op)))) {
++ rc = -EFAULT;
++ goto unmap_out;
++ }
+
-+static int xen_mmap_mem(struct file * file, struct vm_area_struct * vma)
-+{
-+ size_t size = vma->vm_end - vma->vm_start;
++ start_index = op.index >> PAGE_SHIFT;
+
-+ if (uncached_access(file))
-+ vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
++ /* First, check that all pages are in the NOT_YET_MAPPED
++ * state.
++ */
++ for (i = 0; i < op.count; ++i) {
++ if (unlikely
++ (private_data->grants[start_index + i].state
++ != GNTDEV_SLOT_NOT_YET_MAPPED)) {
++ if (private_data->grants[start_index + i].state
++ == GNTDEV_SLOT_INVALID) {
++ printk(KERN_ERR
++ "Tried to remove an invalid "
++ "grant at offset 0x%x.",
++ (start_index + i)
++ << PAGE_SHIFT);
++ rc = -EINVAL;
++ } else {
++ printk(KERN_ERR
++ "Tried to remove a grant which "
++ "is currently mmap()-ed at "
++ "offset 0x%x.",
++ (start_index + i)
++ << PAGE_SHIFT);
++ rc = -EBUSY;
++ }
++ goto unmap_out;
++ }
++ }
++
++ /* Unmap pages and add them to the free list.
++ */
++ for (i = 0; i < op.count; ++i) {
++ private_data->grants[start_index+i].state =
++ GNTDEV_SLOT_INVALID;
++ private_data->grants[start_index+i].u.free_list_index =
++ private_data->free_list_size;
++ private_data->free_list[private_data->free_list_size] =
++ start_index + i;
++ ++private_data->free_list_size;
++ }
+
-+ /* We want to return the real error code, not EAGAIN. */
-+ return direct_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff,
-+ size, vma->vm_page_prot, DOMID_IO);
-+}
-+#endif
++ unmap_out:
++ up_write(&private_data->grants_sem);
++ up_write(&private_data->free_list_sem);
++ return rc;
++ }
++ case IOCTL_GNTDEV_GET_OFFSET_FOR_VADDR:
++ {
++ struct ioctl_gntdev_get_offset_for_vaddr op;
++ struct vm_area_struct *vma;
++ unsigned long vaddr;
+
-+/*
-+ * The memory devices use the full 32/64 bits of the offset, and so we cannot
-+ * check against negative addresses: they are ok. The return value is weird,
-+ * though, in that case (0).
-+ *
-+ * also note that seeking relative to the "end of file" isn't supported:
-+ * it has no meaning, so it returns -EINVAL.
-+ */
-+static loff_t memory_lseek(struct file * file, loff_t offset, int orig)
-+{
-+ loff_t ret;
++ if ((rc = copy_from_user(&op,
++ (void __user *) arg,
++ sizeof(op)))) {
++ rc = -EFAULT;
++ goto get_offset_out;
++ }
++ vaddr = (unsigned long)op.vaddr;
+
-+ mutex_lock(&file->f_dentry->d_inode->i_mutex);
-+ switch (orig) {
-+ case 0:
-+ file->f_pos = offset;
-+ ret = file->f_pos;
-+ force_successful_syscall_return();
-+ break;
-+ case 1:
-+ file->f_pos += offset;
-+ ret = file->f_pos;
-+ force_successful_syscall_return();
-+ break;
-+ default:
-+ ret = -EINVAL;
++ down_read(¤t->mm->mmap_sem);
++ vma = find_vma(current->mm, vaddr);
++ if (vma == NULL) {
++ rc = -EFAULT;
++ goto get_offset_unlock_out;
++ }
++ if ((!vma->vm_ops) || (vma->vm_ops != &gntdev_vmops)) {
++ printk(KERN_ERR "The vaddr specified does not belong "
++ "to a gntdev instance: %#lx\n", vaddr);
++ rc = -EFAULT;
++ goto get_offset_unlock_out;
++ }
++ if (vma->vm_start != vaddr) {
++ printk(KERN_ERR "The vaddr specified in an "
++ "IOCTL_GNTDEV_GET_OFFSET_FOR_VADDR must be at "
++ "the start of the VM area. vma->vm_start = "
++ "%#lx; vaddr = %#lx\n",
++ vma->vm_start, vaddr);
++ rc = -EFAULT;
++ goto get_offset_unlock_out;
++ }
++ op.offset = vma->vm_pgoff << PAGE_SHIFT;
++ op.count = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
++ up_read(¤t->mm->mmap_sem);
++ if ((rc = copy_to_user((void __user *) arg,
++ &op,
++ sizeof(op)))) {
++ rc = -EFAULT;
++ goto get_offset_out;
++ }
++ goto get_offset_out;
++ get_offset_unlock_out:
++ up_read(¤t->mm->mmap_sem);
++ get_offset_out:
++ return rc;
++ }
++ case IOCTL_GNTDEV_SET_MAX_GRANTS:
++ {
++ struct ioctl_gntdev_set_max_grants op;
++ if ((rc = copy_from_user(&op,
++ (void __user *) arg,
++ sizeof(op)))) {
++ rc = -EFAULT;
++ goto set_max_out;
++ }
++ down_write(&private_data->grants_sem);
++ if (private_data->grants) {
++ rc = -EBUSY;
++ goto set_max_unlock_out;
++ }
++ if (op.count > MAX_GRANTS_LIMIT) {
++ rc = -EINVAL;
++ goto set_max_unlock_out;
++ }
++ rc = init_private_data(private_data, op.count);
++ set_max_unlock_out:
++ up_write(&private_data->grants_sem);
++ set_max_out:
++ return rc;
++ }
++ default:
++ return -ENOIOCTLCMD;
+ }
-+ mutex_unlock(&file->f_dentry->d_inode->i_mutex);
-+ return ret;
-+}
+
-+static int open_mem(struct inode * inode, struct file * filp)
-+{
-+ return capable(CAP_SYS_RAWIO) ? 0 : -EPERM;
++ return 0;
+}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/netback/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/netback/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,5 @@
++obj-$(CONFIG_XEN_NETDEV_BACKEND) := netbk.o
++obj-$(CONFIG_XEN_NETDEV_LOOPBACK) += netloop.o
+
-+const struct file_operations mem_fops = {
-+ .llseek = memory_lseek,
-+ .read = read_mem,
-+ .write = write_mem,
-+ .mmap = xen_mmap_mem,
-+ .open = open_mem,
-+};
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/console/console.c tmp-linux-2.6-xen.patch/drivers/xen/console/console.c
---- pristine-linux-2.6.18.2/drivers/xen/console/console.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/console/console.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,721 @@
++netbk-y := netback.o xenbus.o interface.o accel.o
++netloop-y := loopback.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/netback/accel.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/netback/accel.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,269 @@
+/******************************************************************************
-+ * console.c
-+ *
-+ * Virtual console driver.
++ * drivers/xen/netback/accel.c
++ *
++ * Interface between backend virtual network device and accelerated plugin.
+ *
-+ * Copyright (c) 2002-2004, K A Fraser.
++ * Copyright (C) 2007 Solarflare Communications, Inc
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
@@ -56979,706 +93018,830 @@
+ * IN THE SOFTWARE.
+ */
+
-+#include <linux/version.h>
-+#include <linux/module.h>
-+#include <linux/errno.h>
-+#include <linux/signal.h>
-+#include <linux/sched.h>
-+#include <linux/interrupt.h>
-+#include <linux/tty.h>
-+#include <linux/tty_flip.h>
-+#include <linux/serial.h>
-+#include <linux/major.h>
-+#include <linux/ptrace.h>
-+#include <linux/ioport.h>
-+#include <linux/mm.h>
-+#include <linux/slab.h>
-+#include <linux/init.h>
-+#include <linux/console.h>
-+#include <linux/bootmem.h>
-+#include <linux/sysrq.h>
-+#include <linux/screen_info.h>
-+#include <linux/vt.h>
-+#include <asm/io.h>
-+#include <asm/irq.h>
-+#include <asm/uaccess.h>
-+#include <xen/interface/xen.h>
-+#include <xen/interface/event_channel.h>
-+#include <asm/hypervisor.h>
-+#include <xen/evtchn.h>
++#include <linux/list.h>
++#include <asm/atomic.h>
+#include <xen/xenbus.h>
-+#include <xen/xencons.h>
-+
-+/*
-+ * Modes:
-+ * 'xencons=off' [XC_OFF]: Console is disabled.
-+ * 'xencons=tty' [XC_TTY]: Console attached to '/dev/tty[0-9]+'.
-+ * 'xencons=ttyS' [XC_SERIAL]: Console attached to '/dev/ttyS[0-9]+'.
-+ * 'xencons=xvc' [XC_XVC]: Console attached to '/dev/xvc0'.
-+ * default: DOM0 -> XC_SERIAL ; all others -> XC_TTY.
-+ *
-+ * NB. In mode XC_TTY, we create dummy consoles for tty2-63. This suppresses
-+ * warnings from standard distro startup scripts.
-+ */
-+static enum {
-+ XC_OFF, XC_TTY, XC_SERIAL, XC_XVC
-+} xc_mode;
-+static int xc_num = -1;
++#include <linux/mutex.h>
+
-+/* /dev/xvc0 device number allocated by lanana.org. */
-+#define XEN_XVC_MAJOR 204
-+#define XEN_XVC_MINOR 191
++#include "common.h"
+
-+#ifdef CONFIG_MAGIC_SYSRQ
-+static unsigned long sysrq_requested;
-+extern int sysrq_enabled;
++#if 0
++#undef DPRINTK
++#define DPRINTK(fmt, args...) \
++ printk("netback/accel (%s:%d) " fmt ".\n", __FUNCTION__, __LINE__, ##args)
+#endif
+
-+void xencons_early_setup(void)
-+{
-+ extern int console_use_vt;
++/*
++ * A list of available netback accelerator plugin modules (each list
++ * entry is of type struct netback_accelerator)
++ */
++static struct list_head accelerators_list;
++/* Lock used to protect access to accelerators_list */
++DEFINE_MUTEX(accelerators_mutex);
+
-+ if (is_initial_xendomain()) {
-+ xc_mode = XC_SERIAL;
++/*
++ * Compare a backend to an accelerator, and decide if they are
++ * compatible (i.e. if the accelerator should be used by the
++ * backend)
++ */
++static int match_accelerator(struct xenbus_device *xendev,
++ struct backend_info *be,
++ struct netback_accelerator *accelerator)
++{
++ int rc = 0;
++ char *eth_name = xenbus_read(XBT_NIL, xendev->nodename, "accel", NULL);
++
++ if (IS_ERR(eth_name)) {
++ /* Probably means not present */
++ DPRINTK("%s: no match due to xenbus_read accel error %d\n",
++ __FUNCTION__, PTR_ERR(eth_name));
++ return 0;
+ } else {
-+ xc_mode = XC_TTY;
-+ console_use_vt = 0;
++ if (!strcmp(eth_name, accelerator->eth_name))
++ rc = 1;
++ kfree(eth_name);
++ return rc;
+ }
+}
+
-+static int __init xencons_setup(char *str)
-+{
-+ char *q;
-+ int n;
-+ extern int console_use_vt;
+
-+ console_use_vt = 1;
-+ if (!strncmp(str, "ttyS", 4)) {
-+ xc_mode = XC_SERIAL;
-+ str += 4;
-+ } else if (!strncmp(str, "tty", 3)) {
-+ xc_mode = XC_TTY;
-+ str += 3;
-+ console_use_vt = 0;
-+ } else if (!strncmp(str, "xvc", 3)) {
-+ xc_mode = XC_XVC;
-+ str += 3;
-+ } else if (!strncmp(str, "off", 3)) {
-+ xc_mode = XC_OFF;
-+ str += 3;
++static void do_probe(struct backend_info *be,
++ struct netback_accelerator *accelerator,
++ struct xenbus_device *xendev)
++{
++ be->accelerator = accelerator;
++ atomic_inc(&be->accelerator->use_count);
++ if (be->accelerator->hooks->probe(xendev) != 0) {
++ atomic_dec(&be->accelerator->use_count);
++ module_put(be->accelerator->hooks->owner);
++ be->accelerator = NULL;
+ }
-+
-+ n = simple_strtol(str, &q, 10);
-+ if (q != str)
-+ xc_num = n;
-+
-+ return 1;
+}
-+__setup("xencons=", xencons_setup);
+
-+/* The kernel and user-land drivers share a common transmit buffer. */
-+static unsigned int wbuf_size = 4096;
-+#define WBUF_MASK(_i) ((_i)&(wbuf_size-1))
-+static char *wbuf;
-+static unsigned int wc, wp; /* write_cons, write_prod */
+
-+static int __init xencons_bufsz_setup(char *str)
++/*
++ * Notify suitable backends that a new accelerator is available and
++ * connected. This will also notify the accelerator plugin module
++ * that it is being used for a device through the probe hook.
++ */
++static int netback_accelerator_probe_backend(struct device *dev, void *arg)
+{
-+ unsigned int goal;
-+ goal = simple_strtoul(str, NULL, 0);
-+ if (goal) {
-+ goal = roundup_pow_of_two(goal);
-+ if (wbuf_size < goal)
-+ wbuf_size = goal;
++ struct netback_accelerator *accelerator =
++ (struct netback_accelerator *)arg;
++ struct xenbus_device *xendev = to_xenbus_device(dev);
++
++ if (!strcmp("vif", xendev->devicetype)) {
++ struct backend_info *be = xendev->dev.driver_data;
++
++ if (match_accelerator(xendev, be, accelerator) &&
++ try_module_get(accelerator->hooks->owner)) {
++ do_probe(be, accelerator, xendev);
++ }
+ }
-+ return 1;
++ return 0;
+}
-+__setup("xencons_bufsz=", xencons_bufsz_setup);
+
-+/* This lock protects accesses to the common transmit buffer. */
-+static DEFINE_SPINLOCK(xencons_lock);
+
-+/* Common transmit-kick routine. */
-+static void __xencons_tx_flush(void);
++/*
++ * Notify suitable backends that an accelerator is unavailable.
++ */
++static int netback_accelerator_remove_backend(struct device *dev, void *arg)
++{
++ struct xenbus_device *xendev = to_xenbus_device(dev);
++ struct netback_accelerator *accelerator =
++ (struct netback_accelerator *)arg;
++
++ if (!strcmp("vif", xendev->devicetype)) {
++ struct backend_info *be = xendev->dev.driver_data;
+
-+static struct tty_driver *xencons_driver;
++ if (be->accelerator == accelerator) {
++ be->accelerator->hooks->remove(xendev);
++ atomic_dec(&be->accelerator->use_count);
++ module_put(be->accelerator->hooks->owner);
++ be->accelerator = NULL;
++ }
++ }
++ return 0;
++}
+
-+/******************** Kernel console driver ********************************/
+
-+static void kcons_write(struct console *c, const char *s, unsigned int count)
-+{
-+ int i = 0;
-+ unsigned long flags;
+
-+ spin_lock_irqsave(&xencons_lock, flags);
++/*
++ * Entry point for an netback accelerator plugin module. Called to
++ * advertise its presence, and connect to any suitable backends.
++ */
++int netback_connect_accelerator(unsigned version, int id, const char *eth_name,
++ struct netback_accel_hooks *hooks)
++{
++ struct netback_accelerator *new_accelerator;
++ unsigned eth_name_len;
+
-+ while (i < count) {
-+ for (; i < count; i++) {
-+ if ((wp - wc) >= (wbuf_size - 1))
-+ break;
-+ if ((wbuf[WBUF_MASK(wp++)] = s[i]) == '\n')
-+ wbuf[WBUF_MASK(wp++)] = '\r';
++ if (version != NETBACK_ACCEL_VERSION) {
++ if (version > NETBACK_ACCEL_VERSION) {
++ /* Caller has higher version number, leave it
++ up to them to decide whether to continue.
++ They can recall with a lower number if
++ they're happy to be compatible with us */
++ return NETBACK_ACCEL_VERSION;
++ } else {
++ /* We have a more recent version than caller.
++ Currently reject, but may in future be able
++ to be backwardly compatible */
++ return -EPROTO;
+ }
++ }
+
-+ __xencons_tx_flush();
++ new_accelerator =
++ kmalloc(sizeof(struct netback_accelerator), GFP_KERNEL);
++ if (!new_accelerator) {
++ DPRINTK("%s: failed to allocate memory for accelerator\n",
++ __FUNCTION__);
++ return -ENOMEM;
+ }
+
-+ spin_unlock_irqrestore(&xencons_lock, flags);
++ new_accelerator->id = id;
++
++ eth_name_len = strlen(eth_name)+1;
++ new_accelerator->eth_name = kmalloc(eth_name_len, GFP_KERNEL);
++ if (!new_accelerator->eth_name) {
++ DPRINTK("%s: failed to allocate memory for eth_name string\n",
++ __FUNCTION__);
++ kfree(new_accelerator);
++ return -ENOMEM;
++ }
++ strlcpy(new_accelerator->eth_name, eth_name, eth_name_len);
++
++ new_accelerator->hooks = hooks;
++
++ atomic_set(&new_accelerator->use_count, 0);
++
++ mutex_lock(&accelerators_mutex);
++ list_add(&new_accelerator->link, &accelerators_list);
++
++ /* tell existing backends about new plugin */
++ xenbus_for_each_backend(new_accelerator,
++ netback_accelerator_probe_backend);
++
++ mutex_unlock(&accelerators_mutex);
++
++ return 0;
++
+}
++EXPORT_SYMBOL_GPL(netback_connect_accelerator);
+
-+static void kcons_write_dom0(struct console *c, const char *s, unsigned int count)
++
++/*
++ * Disconnect an accelerator plugin module that has previously been
++ * connected.
++ */
++void netback_disconnect_accelerator(int id, const char *eth_name)
+{
++ struct netback_accelerator *accelerator, *next;
+
-+ while (count > 0) {
-+ int rc;
-+ rc = HYPERVISOR_console_io( CONSOLEIO_write, count, (char *)s);
-+ if (rc <= 0)
++ mutex_lock(&accelerators_mutex);
++ list_for_each_entry_safe(accelerator, next, &accelerators_list, link) {
++ if (!strcmp(eth_name, accelerator->eth_name)) {
++ xenbus_for_each_backend
++ (accelerator, netback_accelerator_remove_backend);
++ BUG_ON(atomic_read(&accelerator->use_count) != 0);
++ list_del(&accelerator->link);
++ kfree(accelerator->eth_name);
++ kfree(accelerator);
+ break;
-+ count -= rc;
-+ s += rc;
++ }
+ }
++ mutex_unlock(&accelerators_mutex);
+}
++EXPORT_SYMBOL_GPL(netback_disconnect_accelerator);
+
-+static struct tty_driver *kcons_device(struct console *c, int *index)
++
++void netback_probe_accelerators(struct backend_info *be,
++ struct xenbus_device *dev)
+{
-+ *index = 0;
-+ return xencons_driver;
++ struct netback_accelerator *accelerator;
++
++ /*
++ * Check list of accelerators to see if any is suitable, and
++ * use it if it is.
++ */
++ mutex_lock(&accelerators_mutex);
++ list_for_each_entry(accelerator, &accelerators_list, link) {
++ if (match_accelerator(dev, be, accelerator) &&
++ try_module_get(accelerator->hooks->owner)) {
++ do_probe(be, accelerator, dev);
++ break;
++ }
++ }
++ mutex_unlock(&accelerators_mutex);
+}
+
-+static struct console kcons_info = {
-+ .device = kcons_device,
-+ .flags = CON_PRINTBUFFER | CON_ENABLED,
-+ .index = -1,
-+};
+
-+static int __init xen_console_init(void)
++void netback_remove_accelerators(struct backend_info *be,
++ struct xenbus_device *dev)
+{
-+ if (!is_running_on_xen())
-+ goto out;
-+
-+ if (is_initial_xendomain()) {
-+ kcons_info.write = kcons_write_dom0;
-+ } else {
-+ if (!xen_start_info->console.domU.evtchn)
-+ goto out;
-+ kcons_info.write = kcons_write;
++ mutex_lock(&accelerators_mutex);
++ /* Notify the accelerator (if any) of this device's removal */
++ if (be->accelerator != NULL) {
++ be->accelerator->hooks->remove(dev);
++ atomic_dec(&be->accelerator->use_count);
++ module_put(be->accelerator->hooks->owner);
++ be->accelerator = NULL;
+ }
++ mutex_unlock(&accelerators_mutex);
++}
+
-+ switch (xc_mode) {
-+ case XC_XVC:
-+ strcpy(kcons_info.name, "xvc");
-+ if (xc_num == -1)
-+ xc_num = 0;
-+ break;
+
-+ case XC_SERIAL:
-+ strcpy(kcons_info.name, "ttyS");
-+ if (xc_num == -1)
-+ xc_num = 0;
-+ break;
++void netif_accel_init(void)
++{
++ INIT_LIST_HEAD(&accelerators_list);
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/netback/common.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/netback/common.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,217 @@
++/******************************************************************************
++ * arch/xen/drivers/netif/backend/common.h
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ case XC_TTY:
-+ strcpy(kcons_info.name, "tty");
-+ if (xc_num == -1)
-+ xc_num = 1;
-+ break;
++#ifndef __NETIF__BACKEND__COMMON_H__
++#define __NETIF__BACKEND__COMMON_H__
+
-+ default:
-+ goto out;
-+ }
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/interrupt.h>
++#include <linux/slab.h>
++#include <linux/ip.h>
++#include <linux/in.h>
++#include <linux/netdevice.h>
++#include <linux/etherdevice.h>
++#include <linux/wait.h>
++#include <xen/evtchn.h>
++#include <xen/interface/io/netif.h>
++#include <asm/io.h>
++#include <asm/pgalloc.h>
++#include <xen/interface/grant_table.h>
++#include <xen/gnttab.h>
++#include <xen/driver_util.h>
++#include <xen/xenbus.h>
+
-+ wbuf = alloc_bootmem(wbuf_size);
++#define DPRINTK(_f, _a...) \
++ pr_debug("(file=%s, line=%d) " _f, \
++ __FILE__ , __LINE__ , ## _a )
++#define IPRINTK(fmt, args...) \
++ printk(KERN_INFO "xen_net: " fmt, ##args)
++#define WPRINTK(fmt, args...) \
++ printk(KERN_WARNING "xen_net: " fmt, ##args)
+
-+ register_console(&kcons_info);
++typedef struct netif_st {
++ /* Unique identifier for this interface. */
++ domid_t domid;
++ unsigned int handle;
+
-+ out:
-+ return 0;
-+}
-+console_initcall(xen_console_init);
++ u8 fe_dev_addr[6];
+
-+/*** Useful function for console debugging -- goes straight to Xen. ***/
-+asmlinkage int xprintk(const char *fmt, ...)
-+{
-+ va_list args;
-+ int printk_len;
-+ static char printk_buf[1024];
++ /* Physical parameters of the comms window. */
++ grant_handle_t tx_shmem_handle;
++ grant_ref_t tx_shmem_ref;
++ grant_handle_t rx_shmem_handle;
++ grant_ref_t rx_shmem_ref;
++ unsigned int irq;
+
-+ /* Emit the output into the temporary buffer */
-+ va_start(args, fmt);
-+ printk_len = vsnprintf(printk_buf, sizeof(printk_buf), fmt, args);
-+ va_end(args);
++ /* The shared rings and indexes. */
++ netif_tx_back_ring_t tx;
++ netif_rx_back_ring_t rx;
++ struct vm_struct *tx_comms_area;
++ struct vm_struct *rx_comms_area;
+
-+ /* Send the processed output directly to Xen. */
-+ kcons_write_dom0(NULL, printk_buf, printk_len);
++ /* Set of features that can be turned on in dev->features. */
++ int features;
+
-+ return 0;
-+}
++ /* Internal feature information. */
++ u8 can_queue:1; /* can queue packets for receiver? */
++ u8 copying_receiver:1; /* copy packets to receiver? */
+
-+/*** Forcibly flush console data before dying. ***/
-+void xencons_force_flush(void)
-+{
-+ int sz;
++ /* Allow netif_be_start_xmit() to peek ahead in the rx request ring. */
++ RING_IDX rx_req_cons_peek;
+
-+ /* Emergency console is synchronous, so there's nothing to flush. */
-+ if (!is_running_on_xen() ||
-+ is_initial_xendomain() ||
-+ !xen_start_info->console.domU.evtchn)
-+ return;
++ /* Transmit shaping: allow 'credit_bytes' every 'credit_usec'. */
++ unsigned long credit_bytes;
++ unsigned long credit_usec;
++ unsigned long remaining_credit;
++ struct timer_list credit_timeout;
+
-+ /* Spin until console data is flushed through to the daemon. */
-+ while (wc != wp) {
-+ int sent = 0;
-+ if ((sz = wp - wc) == 0)
-+ continue;
-+ sent = xencons_ring_send(&wbuf[WBUF_MASK(wc)], sz);
-+ if (sent > 0)
-+ wc += sent;
-+ }
-+}
++ /* Enforce draining of the transmit queue. */
++ struct timer_list tx_queue_timeout;
+
++ /* Miscellaneous private stuff. */
++ struct list_head list; /* scheduling list */
++ atomic_t refcnt;
++ struct net_device *dev;
++ struct net_device_stats stats;
+
-+void dom0_init_screen_info(const struct dom0_vga_console_info *info)
-+{
-+ switch (info->video_type) {
-+ case XEN_VGATYPE_TEXT_MODE_3:
-+ screen_info.orig_video_mode = 3;
-+ screen_info.orig_video_ega_bx = 3;
-+ screen_info.orig_video_isVGA = 1;
-+ screen_info.orig_video_lines = info->u.text_mode_3.rows;
-+ screen_info.orig_video_cols = info->u.text_mode_3.columns;
-+ screen_info.orig_x = info->u.text_mode_3.cursor_x;
-+ screen_info.orig_y = info->u.text_mode_3.cursor_y;
-+ screen_info.orig_video_points =
-+ info->u.text_mode_3.font_height;
-+ break;
-+ case XEN_VGATYPE_VESA_LFB:
-+ screen_info.orig_video_isVGA = VIDEO_TYPE_VLFB;
-+ screen_info.lfb_width = info->u.vesa_lfb.width;
-+ screen_info.lfb_height = info->u.vesa_lfb.height;
-+ screen_info.lfb_depth = info->u.vesa_lfb.bits_per_pixel;
-+ screen_info.lfb_base = info->u.vesa_lfb.lfb_base;
-+ screen_info.lfb_size = info->u.vesa_lfb.lfb_size;
-+ screen_info.lfb_linelength = info->u.vesa_lfb.bytes_per_line;
-+ screen_info.red_size = info->u.vesa_lfb.red_size;
-+ screen_info.red_pos = info->u.vesa_lfb.red_pos;
-+ screen_info.green_size = info->u.vesa_lfb.green_size;
-+ screen_info.green_pos = info->u.vesa_lfb.green_pos;
-+ screen_info.blue_size = info->u.vesa_lfb.blue_size;
-+ screen_info.blue_pos = info->u.vesa_lfb.blue_pos;
-+ screen_info.rsvd_size = info->u.vesa_lfb.rsvd_size;
-+ screen_info.rsvd_pos = info->u.vesa_lfb.rsvd_pos;
-+ break;
-+ }
-+}
++ unsigned int carrier;
++
++ wait_queue_head_t waiting_to_free;
++} netif_t;
++
++/*
++ * Implement our own carrier flag: the network stack's version causes delays
++ * when the carrier is re-enabled (in particular, dev_activate() may not
++ * immediately be called, which can cause packet loss; also the etherbridge
++ * can be rather lazy in activating its port).
++ */
++#define netback_carrier_on(netif) ((netif)->carrier = 1)
++#define netback_carrier_off(netif) ((netif)->carrier = 0)
++#define netback_carrier_ok(netif) ((netif)->carrier)
+
++enum {
++ NETBK_DONT_COPY_SKB,
++ NETBK_DELAYED_COPY_SKB,
++ NETBK_ALWAYS_COPY_SKB,
++};
+
-+/******************** User-space console driver (/dev/console) ************/
++extern int netbk_copy_skb_mode;
+
-+#define DRV(_d) (_d)
-+#define DUMMY_TTY(_tty) ((xc_mode == XC_TTY) && \
-+ ((_tty)->index != (xc_num - 1)))
++/* Function pointers into netback accelerator plugin modules */
++struct netback_accel_hooks {
++ struct module *owner;
++ int (*probe)(struct xenbus_device *dev);
++ int (*remove)(struct xenbus_device *dev);
++};
+
-+static struct termios *xencons_termios[MAX_NR_CONSOLES];
-+static struct termios *xencons_termios_locked[MAX_NR_CONSOLES];
-+static struct tty_struct *xencons_tty;
-+static int xencons_priv_irq;
-+static char x_char;
++/* Structure to track the state of a netback accelerator plugin */
++struct netback_accelerator {
++ struct list_head link;
++ int id;
++ char *eth_name;
++ atomic_t use_count;
++ struct netback_accel_hooks *hooks;
++};
+
-+void xencons_rx(char *buf, unsigned len, struct pt_regs *regs)
-+{
-+ int i;
-+ unsigned long flags;
++struct backend_info {
++ struct xenbus_device *dev;
++ netif_t *netif;
++ enum xenbus_state frontend_state;
+
-+ spin_lock_irqsave(&xencons_lock, flags);
-+ if (xencons_tty == NULL)
-+ goto out;
++ /* State relating to the netback accelerator */
++ void *netback_accel_priv;
++ /* The accelerator that this backend is currently using */
++ struct netback_accelerator *accelerator;
++};
+
-+ for (i = 0; i < len; i++) {
-+#ifdef CONFIG_MAGIC_SYSRQ
-+ if (sysrq_enabled) {
-+ if (buf[i] == '\x0f') { /* ^O */
-+ if (!sysrq_requested) {
-+ sysrq_requested = jiffies;
-+ continue; /* don't print sysrq key */
-+ }
-+ sysrq_requested = 0;
-+ } else if (sysrq_requested) {
-+ unsigned long sysrq_timeout =
-+ sysrq_requested + HZ*2;
-+ sysrq_requested = 0;
-+ if (time_before(jiffies, sysrq_timeout)) {
-+ spin_unlock_irqrestore(
-+ &xencons_lock, flags);
-+ handle_sysrq(
-+ buf[i], regs, xencons_tty);
-+ spin_lock_irqsave(
-+ &xencons_lock, flags);
-+ continue;
-+ }
-+ }
-+ }
-+#endif
-+ tty_insert_flip_char(xencons_tty, buf[i], 0);
-+ }
-+ tty_flip_buffer_push(xencons_tty);
++#define NETBACK_ACCEL_VERSION 0x00010001
+
-+ out:
-+ spin_unlock_irqrestore(&xencons_lock, flags);
-+}
++/*
++ * Connect an accelerator plugin module to netback. Returns zero on
++ * success, < 0 on error, > 0 (with highest version number supported)
++ * if version mismatch.
++ */
++extern int netback_connect_accelerator(unsigned version,
++ int id, const char *eth_name,
++ struct netback_accel_hooks *hooks);
++/* Disconnect a previously connected accelerator plugin module */
++extern void netback_disconnect_accelerator(int id, const char *eth_name);
++
++
++extern
++void netback_probe_accelerators(struct backend_info *be,
++ struct xenbus_device *dev);
++extern
++void netback_remove_accelerators(struct backend_info *be,
++ struct xenbus_device *dev);
++extern
++void netif_accel_init(void);
+
-+static void __xencons_tx_flush(void)
-+{
-+ int sent, sz, work_done = 0;
+
-+ if (x_char) {
-+ if (is_initial_xendomain())
-+ kcons_write_dom0(NULL, &x_char, 1);
-+ else
-+ while (x_char)
-+ if (xencons_ring_send(&x_char, 1) == 1)
-+ break;
-+ x_char = 0;
-+ work_done = 1;
-+ }
++#define NET_TX_RING_SIZE __RING_SIZE((netif_tx_sring_t *)0, PAGE_SIZE)
++#define NET_RX_RING_SIZE __RING_SIZE((netif_rx_sring_t *)0, PAGE_SIZE)
+
-+ while (wc != wp) {
-+ sz = wp - wc;
-+ if (sz > (wbuf_size - WBUF_MASK(wc)))
-+ sz = wbuf_size - WBUF_MASK(wc);
-+ if (is_initial_xendomain()) {
-+ kcons_write_dom0(NULL, &wbuf[WBUF_MASK(wc)], sz);
-+ wc += sz;
-+ } else {
-+ sent = xencons_ring_send(&wbuf[WBUF_MASK(wc)], sz);
-+ if (sent == 0)
-+ break;
-+ wc += sent;
-+ }
-+ work_done = 1;
-+ }
++void netif_disconnect(netif_t *netif);
+
-+ if (work_done && (xencons_tty != NULL)) {
-+ wake_up_interruptible(&xencons_tty->write_wait);
-+ if ((xencons_tty->flags & (1 << TTY_DO_WRITE_WAKEUP)) &&
-+ (xencons_tty->ldisc.write_wakeup != NULL))
-+ (xencons_tty->ldisc.write_wakeup)(xencons_tty);
-+ }
-+}
++netif_t *netif_alloc(domid_t domid, unsigned int handle);
++int netif_map(netif_t *netif, unsigned long tx_ring_ref,
++ unsigned long rx_ring_ref, unsigned int evtchn);
+
-+void xencons_tx(void)
-+{
-+ unsigned long flags;
++#define netif_get(_b) (atomic_inc(&(_b)->refcnt))
++#define netif_put(_b) \
++ do { \
++ if ( atomic_dec_and_test(&(_b)->refcnt) ) \
++ wake_up(&(_b)->waiting_to_free); \
++ } while (0)
+
-+ spin_lock_irqsave(&xencons_lock, flags);
-+ __xencons_tx_flush();
-+ spin_unlock_irqrestore(&xencons_lock, flags);
-+}
++void netif_xenbus_init(void);
+
-+/* Privileged receive callback and transmit kicker. */
-+static irqreturn_t xencons_priv_interrupt(int irq, void *dev_id,
-+ struct pt_regs *regs)
-+{
-+ static char rbuf[16];
-+ int l;
++#define netif_schedulable(netif) \
++ (netif_running((netif)->dev) && netback_carrier_ok(netif))
+
-+ while ((l = HYPERVISOR_console_io(CONSOLEIO_read, 16, rbuf)) > 0)
-+ xencons_rx(rbuf, l, regs);
++void netif_schedule_work(netif_t *netif);
++void netif_deschedule_work(netif_t *netif);
+
-+ xencons_tx();
++int netif_be_start_xmit(struct sk_buff *skb, struct net_device *dev);
++struct net_device_stats *netif_be_get_stats(struct net_device *dev);
++irqreturn_t netif_be_int(int irq, void *dev_id, struct pt_regs *regs);
+
-+ return IRQ_HANDLED;
++static inline int netbk_can_queue(struct net_device *dev)
++{
++ netif_t *netif = netdev_priv(dev);
++ return netif->can_queue;
+}
+
-+static int xencons_write_room(struct tty_struct *tty)
++static inline int netbk_can_sg(struct net_device *dev)
+{
-+ return wbuf_size - (wp - wc);
++ netif_t *netif = netdev_priv(dev);
++ return netif->features & NETIF_F_SG;
+}
+
-+static int xencons_chars_in_buffer(struct tty_struct *tty)
++#endif /* __NETIF__BACKEND__COMMON_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/netback/interface.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/netback/interface.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,336 @@
++/******************************************************************************
++ * arch/xen/drivers/netif/backend/interface.c
++ *
++ * Network-device interface management.
++ *
++ * Copyright (c) 2004-2005, Keir Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include "common.h"
++#include <linux/ethtool.h>
++#include <linux/rtnetlink.h>
++
++/*
++ * Module parameter 'queue_length':
++ *
++ * Enables queuing in the network stack when a client has run out of receive
++ * descriptors. Although this feature can improve receive bandwidth by avoiding
++ * packet loss, it can also result in packets sitting in the 'tx_queue' for
++ * unbounded time. This is bad if those packets hold onto foreign resources.
++ * For example, consider a packet that holds onto resources belonging to the
++ * guest for which it is queued (e.g., packet received on vif1.0, destined for
++ * vif1.1 which is not activated in the guest): in this situation the guest
++ * will never be destroyed, unless vif1.1 is taken down. To avoid this, we
++ * run a timer (tx_queue_timeout) to drain the queue when the interface is
++ * blocked.
++ */
++static unsigned long netbk_queue_length = 32;
++module_param_named(queue_length, netbk_queue_length, ulong, 0);
++
++static void __netif_up(netif_t *netif)
+{
-+ return wp - wc;
++ enable_irq(netif->irq);
++ netif_schedule_work(netif);
+}
+
-+static void xencons_send_xchar(struct tty_struct *tty, char ch)
++static void __netif_down(netif_t *netif)
+{
-+ unsigned long flags;
++ disable_irq(netif->irq);
++ netif_deschedule_work(netif);
++}
+
-+ if (DUMMY_TTY(tty))
-+ return;
++static int net_open(struct net_device *dev)
++{
++ netif_t *netif = netdev_priv(dev);
++ if (netback_carrier_ok(netif)) {
++ __netif_up(netif);
++ netif_start_queue(dev);
++ }
++ return 0;
++}
+
-+ spin_lock_irqsave(&xencons_lock, flags);
-+ x_char = ch;
-+ __xencons_tx_flush();
-+ spin_unlock_irqrestore(&xencons_lock, flags);
++static int net_close(struct net_device *dev)
++{
++ netif_t *netif = netdev_priv(dev);
++ if (netback_carrier_ok(netif))
++ __netif_down(netif);
++ netif_stop_queue(dev);
++ return 0;
+}
+
-+static void xencons_throttle(struct tty_struct *tty)
++static int netbk_change_mtu(struct net_device *dev, int mtu)
+{
-+ if (DUMMY_TTY(tty))
-+ return;
++ int max = netbk_can_sg(dev) ? 65535 - ETH_HLEN : ETH_DATA_LEN;
+
-+ if (I_IXOFF(tty))
-+ xencons_send_xchar(tty, STOP_CHAR(tty));
++ if (mtu > max)
++ return -EINVAL;
++ dev->mtu = mtu;
++ return 0;
+}
+
-+static void xencons_unthrottle(struct tty_struct *tty)
++static int netbk_set_sg(struct net_device *dev, u32 data)
+{
-+ if (DUMMY_TTY(tty))
-+ return;
++ if (data) {
++ netif_t *netif = netdev_priv(dev);
+
-+ if (I_IXOFF(tty)) {
-+ if (x_char != 0)
-+ x_char = 0;
-+ else
-+ xencons_send_xchar(tty, START_CHAR(tty));
++ if (!(netif->features & NETIF_F_SG))
++ return -ENOSYS;
+ }
++
++ return ethtool_op_set_sg(dev, data);
+}
+
-+static void xencons_flush_buffer(struct tty_struct *tty)
++static int netbk_set_tso(struct net_device *dev, u32 data)
+{
-+ unsigned long flags;
++ if (data) {
++ netif_t *netif = netdev_priv(dev);
+
-+ if (DUMMY_TTY(tty))
-+ return;
++ if (!(netif->features & NETIF_F_TSO))
++ return -ENOSYS;
++ }
+
-+ spin_lock_irqsave(&xencons_lock, flags);
-+ wc = wp = 0;
-+ spin_unlock_irqrestore(&xencons_lock, flags);
++ return ethtool_op_set_tso(dev, data);
+}
+
-+static inline int __xencons_put_char(int ch)
++static struct ethtool_ops network_ethtool_ops =
+{
-+ char _ch = (char)ch;
-+ if ((wp - wc) == wbuf_size)
-+ return 0;
-+ wbuf[WBUF_MASK(wp++)] = _ch;
-+ return 1;
-+}
++ .get_tx_csum = ethtool_op_get_tx_csum,
++ .set_tx_csum = ethtool_op_set_tx_csum,
++ .get_sg = ethtool_op_get_sg,
++ .set_sg = netbk_set_sg,
++ .get_tso = ethtool_op_get_tso,
++ .set_tso = netbk_set_tso,
++ .get_link = ethtool_op_get_link,
++};
+
-+static int xencons_write(
-+ struct tty_struct *tty,
-+ const unsigned char *buf,
-+ int count)
++netif_t *netif_alloc(domid_t domid, unsigned int handle)
+{
-+ int i;
-+ unsigned long flags;
-+
-+ if (DUMMY_TTY(tty))
-+ return count;
++ int err = 0;
++ struct net_device *dev;
++ netif_t *netif;
++ char name[IFNAMSIZ] = {};
+
-+ spin_lock_irqsave(&xencons_lock, flags);
++ snprintf(name, IFNAMSIZ - 1, "vif%u.%u", domid, handle);
++ dev = alloc_netdev(sizeof(netif_t), name, ether_setup);
++ if (dev == NULL) {
++ DPRINTK("Could not create netif: out of memory\n");
++ return ERR_PTR(-ENOMEM);
++ }
+
-+ for (i = 0; i < count; i++)
-+ if (!__xencons_put_char(buf[i]))
-+ break;
++ netif = netdev_priv(dev);
++ memset(netif, 0, sizeof(*netif));
++ netif->domid = domid;
++ netif->handle = handle;
++ atomic_set(&netif->refcnt, 1);
++ init_waitqueue_head(&netif->waiting_to_free);
++ netif->dev = dev;
+
-+ if (i != 0)
-+ __xencons_tx_flush();
++ netback_carrier_off(netif);
+
-+ spin_unlock_irqrestore(&xencons_lock, flags);
++ netif->credit_bytes = netif->remaining_credit = ~0UL;
++ netif->credit_usec = 0UL;
++ init_timer(&netif->credit_timeout);
++ /* Initialize 'expires' now: it's used to track the credit window. */
++ netif->credit_timeout.expires = jiffies;
+
-+ return i;
-+}
++ init_timer(&netif->tx_queue_timeout);
+
-+static void xencons_put_char(struct tty_struct *tty, u_char ch)
-+{
-+ unsigned long flags;
++ dev->hard_start_xmit = netif_be_start_xmit;
++ dev->get_stats = netif_be_get_stats;
++ dev->open = net_open;
++ dev->stop = net_close;
++ dev->change_mtu = netbk_change_mtu;
++ dev->features = NETIF_F_IP_CSUM;
+
-+ if (DUMMY_TTY(tty))
-+ return;
++ SET_ETHTOOL_OPS(dev, &network_ethtool_ops);
+
-+ spin_lock_irqsave(&xencons_lock, flags);
-+ (void)__xencons_put_char(ch);
-+ spin_unlock_irqrestore(&xencons_lock, flags);
-+}
++ dev->tx_queue_len = netbk_queue_length;
+
-+static void xencons_flush_chars(struct tty_struct *tty)
-+{
-+ unsigned long flags;
++ /*
++ * Initialise a dummy MAC address. We choose the numerically
++ * largest non-broadcast address to prevent the address getting
++ * stolen by an Ethernet bridge for STP purposes.
++ * (FE:FF:FF:FF:FF:FF)
++ */
++ memset(dev->dev_addr, 0xFF, ETH_ALEN);
++ dev->dev_addr[0] &= ~0x01;
+
-+ if (DUMMY_TTY(tty))
-+ return;
++ rtnl_lock();
++ err = register_netdevice(dev);
++ rtnl_unlock();
++ if (err) {
++ DPRINTK("Could not register new net device %s: err=%d\n",
++ dev->name, err);
++ free_netdev(dev);
++ return ERR_PTR(err);
++ }
+
-+ spin_lock_irqsave(&xencons_lock, flags);
-+ __xencons_tx_flush();
-+ spin_unlock_irqrestore(&xencons_lock, flags);
++ DPRINTK("Successfully created netif\n");
++ return netif;
+}
+
-+static void xencons_wait_until_sent(struct tty_struct *tty, int timeout)
++static int map_frontend_pages(
++ netif_t *netif, grant_ref_t tx_ring_ref, grant_ref_t rx_ring_ref)
+{
-+ unsigned long orig_jiffies = jiffies;
++ struct gnttab_map_grant_ref op;
+
-+ if (DUMMY_TTY(tty))
-+ return;
++ gnttab_set_map_op(&op, (unsigned long)netif->tx_comms_area->addr,
++ GNTMAP_host_map, tx_ring_ref, netif->domid);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
+
-+ while (DRV(tty->driver)->chars_in_buffer(tty)) {
-+ set_current_state(TASK_INTERRUPTIBLE);
-+ schedule_timeout(1);
-+ if (signal_pending(current))
-+ break;
-+ if (timeout && time_after(jiffies, orig_jiffies + timeout))
-+ break;
++ if (op.status) {
++ DPRINTK(" Gnttab failure mapping tx_ring_ref!\n");
++ return op.status;
+ }
+
-+ set_current_state(TASK_RUNNING);
-+}
++ netif->tx_shmem_ref = tx_ring_ref;
++ netif->tx_shmem_handle = op.handle;
+
-+static int xencons_open(struct tty_struct *tty, struct file *filp)
-+{
-+ unsigned long flags;
++ gnttab_set_map_op(&op, (unsigned long)netif->rx_comms_area->addr,
++ GNTMAP_host_map, rx_ring_ref, netif->domid);
+
-+ if (DUMMY_TTY(tty))
-+ return 0;
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
+
-+ spin_lock_irqsave(&xencons_lock, flags);
-+ tty->driver_data = NULL;
-+ if (xencons_tty == NULL)
-+ xencons_tty = tty;
-+ __xencons_tx_flush();
-+ spin_unlock_irqrestore(&xencons_lock, flags);
++ if (op.status) {
++ DPRINTK(" Gnttab failure mapping rx_ring_ref!\n");
++ return op.status;
++ }
++
++ netif->rx_shmem_ref = rx_ring_ref;
++ netif->rx_shmem_handle = op.handle;
+
+ return 0;
+}
+
-+static void xencons_close(struct tty_struct *tty, struct file *filp)
++static void unmap_frontend_pages(netif_t *netif)
+{
-+ unsigned long flags;
-+
-+ if (DUMMY_TTY(tty))
-+ return;
++ struct gnttab_unmap_grant_ref op;
+
-+ mutex_lock(&tty_mutex);
++ gnttab_set_unmap_op(&op, (unsigned long)netif->tx_comms_area->addr,
++ GNTMAP_host_map, netif->tx_shmem_handle);
+
-+ if (tty->count != 1) {
-+ mutex_unlock(&tty_mutex);
-+ return;
-+ }
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
+
-+ /* Prevent other threads from re-opening this tty. */
-+ set_bit(TTY_CLOSING, &tty->flags);
-+ mutex_unlock(&tty_mutex);
++ gnttab_set_unmap_op(&op, (unsigned long)netif->rx_comms_area->addr,
++ GNTMAP_host_map, netif->rx_shmem_handle);
+
-+ tty->closing = 1;
-+ tty_wait_until_sent(tty, 0);
-+ if (DRV(tty->driver)->flush_buffer != NULL)
-+ DRV(tty->driver)->flush_buffer(tty);
-+ if (tty->ldisc.flush_buffer != NULL)
-+ tty->ldisc.flush_buffer(tty);
-+ tty->closing = 0;
-+ spin_lock_irqsave(&xencons_lock, flags);
-+ xencons_tty = NULL;
-+ spin_unlock_irqrestore(&xencons_lock, flags);
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
+}
+
-+static struct tty_operations xencons_ops = {
-+ .open = xencons_open,
-+ .close = xencons_close,
-+ .write = xencons_write,
-+ .write_room = xencons_write_room,
-+ .put_char = xencons_put_char,
-+ .flush_chars = xencons_flush_chars,
-+ .chars_in_buffer = xencons_chars_in_buffer,
-+ .send_xchar = xencons_send_xchar,
-+ .flush_buffer = xencons_flush_buffer,
-+ .throttle = xencons_throttle,
-+ .unthrottle = xencons_unthrottle,
-+ .wait_until_sent = xencons_wait_until_sent,
-+};
-+
-+static int __init xencons_init(void)
++int netif_map(netif_t *netif, unsigned long tx_ring_ref,
++ unsigned long rx_ring_ref, unsigned int evtchn)
+{
-+ int rc;
-+
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++ int err = -ENOMEM;
++ netif_tx_sring_t *txs;
++ netif_rx_sring_t *rxs;
+
-+ if (xc_mode == XC_OFF)
++ /* Already connected through? */
++ if (netif->irq)
+ return 0;
+
-+ if (!is_initial_xendomain()) {
-+ rc = xencons_ring_init();
-+ if (rc)
-+ return rc;
-+ }
-+
-+ xencons_driver = alloc_tty_driver((xc_mode == XC_TTY) ?
-+ MAX_NR_CONSOLES : 1);
-+ if (xencons_driver == NULL)
++ netif->tx_comms_area = alloc_vm_area(PAGE_SIZE);
++ if (netif->tx_comms_area == NULL)
+ return -ENOMEM;
++ netif->rx_comms_area = alloc_vm_area(PAGE_SIZE);
++ if (netif->rx_comms_area == NULL)
++ goto err_rx;
+
-+ DRV(xencons_driver)->name = "xencons";
-+ DRV(xencons_driver)->major = TTY_MAJOR;
-+ DRV(xencons_driver)->type = TTY_DRIVER_TYPE_SERIAL;
-+ DRV(xencons_driver)->subtype = SERIAL_TYPE_NORMAL;
-+ DRV(xencons_driver)->init_termios = tty_std_termios;
-+ DRV(xencons_driver)->flags =
-+ TTY_DRIVER_REAL_RAW |
-+ TTY_DRIVER_RESET_TERMIOS;
-+ DRV(xencons_driver)->termios = xencons_termios;
-+ DRV(xencons_driver)->termios_locked = xencons_termios_locked;
++ err = map_frontend_pages(netif, tx_ring_ref, rx_ring_ref);
++ if (err)
++ goto err_map;
+
-+ switch (xc_mode) {
-+ case XC_XVC:
-+ DRV(xencons_driver)->name = "xvc";
-+ DRV(xencons_driver)->major = XEN_XVC_MAJOR;
-+ DRV(xencons_driver)->minor_start = XEN_XVC_MINOR;
-+ DRV(xencons_driver)->name_base = xc_num;
-+ break;
-+ case XC_SERIAL:
-+ DRV(xencons_driver)->name = "ttyS";
-+ DRV(xencons_driver)->minor_start = 64 + xc_num;
-+ DRV(xencons_driver)->name_base = xc_num;
-+ break;
-+ default:
-+ DRV(xencons_driver)->name = "tty";
-+ DRV(xencons_driver)->minor_start = 1;
-+ DRV(xencons_driver)->name_base = 1;
-+ break;
-+ }
++ err = bind_interdomain_evtchn_to_irqhandler(
++ netif->domid, evtchn, netif_be_int, 0,
++ netif->dev->name, netif);
++ if (err < 0)
++ goto err_hypervisor;
++ netif->irq = err;
++ disable_irq(netif->irq);
+
-+ tty_set_operations(xencons_driver, &xencons_ops);
++ txs = (netif_tx_sring_t *)netif->tx_comms_area->addr;
++ BACK_RING_INIT(&netif->tx, txs, PAGE_SIZE);
+
-+ if ((rc = tty_register_driver(DRV(xencons_driver))) != 0) {
-+ printk("WARNING: Failed to register Xen virtual "
-+ "console driver as '%s%d'\n",
-+ DRV(xencons_driver)->name,
-+ DRV(xencons_driver)->name_base);
-+ put_tty_driver(xencons_driver);
-+ xencons_driver = NULL;
-+ return rc;
-+ }
++ rxs = (netif_rx_sring_t *)
++ ((char *)netif->rx_comms_area->addr);
++ BACK_RING_INIT(&netif->rx, rxs, PAGE_SIZE);
+
-+ if (is_initial_xendomain()) {
-+ xencons_priv_irq = bind_virq_to_irqhandler(
-+ VIRQ_CONSOLE,
-+ 0,
-+ xencons_priv_interrupt,
-+ 0,
-+ "console",
-+ NULL);
-+ BUG_ON(xencons_priv_irq < 0);
-+ }
++ netif->rx_req_cons_peek = 0;
+
-+ printk("Xen virtual console successfully installed as %s%d\n",
-+ DRV(xencons_driver)->name, xc_num);
++ netif_get(netif);
++
++ rtnl_lock();
++ netback_carrier_on(netif);
++ if (netif_running(netif->dev))
++ __netif_up(netif);
++ rtnl_unlock();
+
+ return 0;
++err_hypervisor:
++ unmap_frontend_pages(netif);
++err_map:
++ free_vm_area(netif->rx_comms_area);
++err_rx:
++ free_vm_area(netif->tx_comms_area);
++ return err;
+}
+
-+module_init(xencons_init);
++void netif_disconnect(netif_t *netif)
++{
++ if (netback_carrier_ok(netif)) {
++ rtnl_lock();
++ netback_carrier_off(netif);
++ netif_carrier_off(netif->dev); /* discard queued packets */
++ if (netif_running(netif->dev))
++ __netif_down(netif);
++ rtnl_unlock();
++ netif_put(netif);
++ }
+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/console/Makefile tmp-linux-2.6-xen.patch/drivers/xen/console/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/console/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/console/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,2 @@
++ atomic_dec(&netif->refcnt);
++ wait_event(netif->waiting_to_free, atomic_read(&netif->refcnt) == 0);
+
-+obj-y := console.o xencons_ring.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/console/xencons_ring.c tmp-linux-2.6-xen.patch/drivers/xen/console/xencons_ring.c
---- pristine-linux-2.6.18.2/drivers/xen/console/xencons_ring.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/console/xencons_ring.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,143 @@
-+/*
++ del_timer_sync(&netif->credit_timeout);
++ del_timer_sync(&netif->tx_queue_timeout);
++
++ if (netif->irq)
++ unbind_from_irqhandler(netif->irq, netif);
++
++ unregister_netdev(netif->dev);
++
++ if (netif->tx.sring) {
++ unmap_frontend_pages(netif);
++ free_vm_area(netif->tx_comms_area);
++ free_vm_area(netif->rx_comms_area);
++ }
++
++ free_netdev(netif->dev);
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/netback/loopback.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/netback/loopback.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,324 @@
++/******************************************************************************
++ * netback/loopback.c
++ *
++ * A two-interface loopback device to emulate a local netfront-netback
++ * connection. This ensures that local packet delivery looks identical
++ * to inter-domain delivery. Most importantly, packets delivered locally
++ * originating from other domains will get *copied* when they traverse this
++ * driver. This prevents unbounded delays in socket-buffer queues from
++ * causing the netback driver to "seize up".
++ *
++ * This driver creates a symmetric pair of loopback interfaces with names
++ * vif0.0 and veth0. The intention is that 'vif0.0' is bound to an Ethernet
++ * bridge, just like a proper netback interface, while a local IP interface
++ * is configured on 'veth0'.
++ *
++ * As with a real netback interface, vif0.0 is configured with a suitable
++ * dummy MAC address. No default is provided for veth0: a reasonable strategy
++ * is to transfer eth0's MAC address to veth0, and give eth0 a dummy address
++ * (to avoid confusing the Etherbridge).
++ *
++ * Copyright (c) 2005 K A Fraser
++ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation; or, when distributed
@@ -57704,307 +93867,295 @@
+ * IN THE SOFTWARE.
+ */
+
-+#include <linux/version.h>
+#include <linux/module.h>
-+#include <linux/errno.h>
-+#include <linux/signal.h>
-+#include <linux/sched.h>
-+#include <linux/interrupt.h>
-+#include <linux/tty.h>
-+#include <linux/tty_flip.h>
-+#include <linux/serial.h>
-+#include <linux/major.h>
-+#include <linux/ptrace.h>
-+#include <linux/ioport.h>
-+#include <linux/mm.h>
-+#include <linux/slab.h>
++#include <linux/netdevice.h>
++#include <linux/inetdevice.h>
++#include <linux/etherdevice.h>
++#include <linux/skbuff.h>
++#include <linux/ethtool.h>
++#include <net/dst.h>
++#include <net/xfrm.h> /* secpath_reset() */
++#include <asm/hypervisor.h> /* is_initial_xendomain() */
+
-+#include <asm/hypervisor.h>
-+#include <xen/evtchn.h>
-+#include <xen/xencons.h>
-+#include <linux/wait.h>
-+#include <linux/interrupt.h>
-+#include <linux/sched.h>
-+#include <linux/err.h>
-+#include <xen/interface/io/console.h>
++static int nloopbacks = -1;
++module_param(nloopbacks, int, 0);
++MODULE_PARM_DESC(nloopbacks, "Number of netback-loopback devices to create");
+
-+static int xencons_irq;
++struct net_private {
++ struct net_device *loopback_dev;
++ struct net_device_stats stats;
++};
+
-+static inline struct xencons_interface *xencons_interface(void)
++static int loopback_open(struct net_device *dev)
+{
-+ return mfn_to_virt(xen_start_info->console.domU.mfn);
++ struct net_private *np = netdev_priv(dev);
++ memset(&np->stats, 0, sizeof(np->stats));
++ netif_start_queue(dev);
++ return 0;
+}
+
-+static inline void notify_daemon(void)
++static int loopback_close(struct net_device *dev)
+{
-+ /* Use evtchn: this is called early, before irq is set up. */
-+ notify_remote_via_evtchn(xen_start_info->console.domU.evtchn);
++ netif_stop_queue(dev);
++ return 0;
+}
+
-+int xencons_ring_send(const char *data, unsigned len)
++#ifdef CONFIG_X86
++static int is_foreign(unsigned long pfn)
+{
-+ int sent = 0;
-+ struct xencons_interface *intf = xencons_interface();
-+ XENCONS_RING_IDX cons, prod;
++ /* NB. Play it safe for auto-translation mode. */
++ return (xen_feature(XENFEAT_auto_translated_physmap) ||
++ (phys_to_machine_mapping[pfn] & FOREIGN_FRAME_BIT));
++}
++#else
++/* How to detect a foreign mapping? Play it safe. */
++#define is_foreign(pfn) (1)
++#endif
+
-+ cons = intf->out_cons;
-+ prod = intf->out_prod;
-+ mb();
-+ BUG_ON((prod - cons) > sizeof(intf->out));
++static int skb_remove_foreign_references(struct sk_buff *skb)
++{
++ struct page *page;
++ unsigned long pfn;
++ int i, off;
++ char *vaddr;
+
-+ while ((sent < len) && ((prod - cons) < sizeof(intf->out)))
-+ intf->out[MASK_XENCONS_IDX(prod++, intf->out)] = data[sent++];
++ BUG_ON(skb_shinfo(skb)->frag_list);
+
-+ wmb();
-+ intf->out_prod = prod;
++ if (skb_cloned(skb) &&
++ unlikely(pskb_expand_head(skb, 0, 0, GFP_ATOMIC)))
++ return 0;
+
-+ notify_daemon();
++ for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
++ pfn = page_to_pfn(skb_shinfo(skb)->frags[i].page);
++ if (!is_foreign(pfn))
++ continue;
++
++ page = alloc_page(GFP_ATOMIC | __GFP_NOWARN);
++ if (unlikely(!page))
++ return 0;
+
-+ return sent;
++ vaddr = kmap_skb_frag(&skb_shinfo(skb)->frags[i]);
++ off = skb_shinfo(skb)->frags[i].page_offset;
++ memcpy(page_address(page) + off,
++ vaddr + off,
++ skb_shinfo(skb)->frags[i].size);
++ kunmap_skb_frag(vaddr);
++
++ put_page(skb_shinfo(skb)->frags[i].page);
++ skb_shinfo(skb)->frags[i].page = page;
++ }
++
++ return 1;
+}
+
-+static irqreturn_t handle_input(int irq, void *unused, struct pt_regs *regs)
++static int loopback_start_xmit(struct sk_buff *skb, struct net_device *dev)
+{
-+ struct xencons_interface *intf = xencons_interface();
-+ XENCONS_RING_IDX cons, prod;
-+
-+ cons = intf->in_cons;
-+ prod = intf->in_prod;
-+ mb();
-+ BUG_ON((prod - cons) > sizeof(intf->in));
++ struct net_private *np = netdev_priv(dev);
+
-+ while (cons != prod) {
-+ xencons_rx(intf->in+MASK_XENCONS_IDX(cons,intf->in), 1, regs);
-+ cons++;
++ if (!skb_remove_foreign_references(skb)) {
++ np->stats.tx_dropped++;
++ dev_kfree_skb(skb);
++ return 0;
+ }
+
-+ mb();
-+ intf->in_cons = cons;
++ dst_release(skb->dst);
++ skb->dst = NULL;
+
-+ notify_daemon();
++ skb_orphan(skb);
+
-+ xencons_tx();
++ np->stats.tx_bytes += skb->len;
++ np->stats.tx_packets++;
+
-+ return IRQ_HANDLED;
-+}
++ /* Switch to loopback context. */
++ dev = np->loopback_dev;
++ np = netdev_priv(dev);
+
-+int xencons_ring_init(void)
-+{
-+ int irq;
++ np->stats.rx_bytes += skb->len;
++ np->stats.rx_packets++;
+
-+ if (xencons_irq)
-+ unbind_from_irqhandler(xencons_irq, NULL);
-+ xencons_irq = 0;
++ if (skb->ip_summed == CHECKSUM_HW) {
++ /* Defer checksum calculation. */
++ skb->proto_csum_blank = 1;
++ /* Must be a local packet: assert its integrity. */
++ skb->proto_data_valid = 1;
++ }
+
-+ if (!is_running_on_xen() ||
-+ is_initial_xendomain() ||
-+ !xen_start_info->console.domU.evtchn)
-+ return -ENODEV;
++ skb->ip_summed = skb->proto_data_valid ?
++ CHECKSUM_UNNECESSARY : CHECKSUM_NONE;
+
-+ irq = bind_caller_port_to_irqhandler(
-+ xen_start_info->console.domU.evtchn,
-+ handle_input, 0, "xencons", NULL);
-+ if (irq < 0) {
-+ printk(KERN_ERR "XEN console request irq failed %i\n", irq);
-+ return irq;
-+ }
++ skb->pkt_type = PACKET_HOST; /* overridden by eth_type_trans() */
++ skb->protocol = eth_type_trans(skb, dev);
++ skb->dev = dev;
++ dev->last_rx = jiffies;
+
-+ xencons_irq = irq;
++ /* Flush netfilter context: rx'ed skbuffs not expected to have any. */
++ nf_reset(skb);
++ secpath_reset(skb);
+
-+ /* In case we have in-flight data after save/restore... */
-+ notify_daemon();
++ netif_rx(skb);
+
+ return 0;
+}
+
-+void xencons_resume(void)
++static struct net_device_stats *loopback_get_stats(struct net_device *dev)
+{
-+ (void)xencons_ring_init();
++ struct net_private *np = netdev_priv(dev);
++ return &np->stats;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/core/cpu_hotplug.c tmp-linux-2.6-xen.patch/drivers/xen/core/cpu_hotplug.c
---- pristine-linux-2.6.18.2/drivers/xen/core/cpu_hotplug.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/core/cpu_hotplug.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,172 @@
-+#include <linux/init.h>
-+#include <linux/kernel.h>
-+#include <linux/sched.h>
-+#include <linux/notifier.h>
-+#include <linux/cpu.h>
-+#include <xen/cpu_hotplug.h>
-+#include <xen/xenbus.h>
++
++static struct ethtool_ops network_ethtool_ops =
++{
++ .get_tx_csum = ethtool_op_get_tx_csum,
++ .set_tx_csum = ethtool_op_set_tx_csum,
++ .get_sg = ethtool_op_get_sg,
++ .set_sg = ethtool_op_set_sg,
++ .get_tso = ethtool_op_get_tso,
++ .set_tso = ethtool_op_set_tso,
++ .get_link = ethtool_op_get_link,
++};
+
+/*
-+ * Set of CPUs that remote admin software will allow us to bring online.
-+ * Notified to us via xenbus.
++ * Nothing to do here. Virtual interface is point-to-point and the
++ * physical interface is probably promiscuous anyway.
+ */
-+static cpumask_t xenbus_allowed_cpumask;
-+
-+/* Set of CPUs that local admin will allow us to bring online. */
-+static cpumask_t local_allowed_cpumask = CPU_MASK_ALL;
-+
-+static int local_cpu_hotplug_request(void)
++static void loopback_set_multicast_list(struct net_device *dev)
+{
-+ /*
-+ * We assume a CPU hotplug request comes from local admin if it is made
-+ * via a userspace process (i.e., one with a real mm_struct).
-+ */
-+ return (current->mm != NULL);
+}
+
-+static void vcpu_hotplug(unsigned int cpu)
++static void loopback_construct(struct net_device *dev, struct net_device *lo)
+{
-+ int err;
-+ char dir[32], state[32];
-+
-+ if ((cpu >= NR_CPUS) || !cpu_possible(cpu))
-+ return;
++ struct net_private *np = netdev_priv(dev);
+
-+ sprintf(dir, "cpu/%d", cpu);
-+ err = xenbus_scanf(XBT_NIL, dir, "availability", "%s", state);
-+ if (err != 1) {
-+ printk(KERN_ERR "XENBUS: Unable to read cpu state\n");
-+ return;
-+ }
++ np->loopback_dev = lo;
+
-+ if (strcmp(state, "online") == 0) {
-+ cpu_set(cpu, xenbus_allowed_cpumask);
-+ (void)cpu_up(cpu);
-+ } else if (strcmp(state, "offline") == 0) {
-+ cpu_clear(cpu, xenbus_allowed_cpumask);
-+ (void)cpu_down(cpu);
-+ } else {
-+ printk(KERN_ERR "XENBUS: unknown state(%s) on CPU%d\n",
-+ state, cpu);
-+ }
-+}
++ dev->open = loopback_open;
++ dev->stop = loopback_close;
++ dev->hard_start_xmit = loopback_start_xmit;
++ dev->get_stats = loopback_get_stats;
++ dev->set_multicast_list = loopback_set_multicast_list;
++ dev->change_mtu = NULL; /* allow arbitrary mtu */
+
-+static void handle_vcpu_hotplug_event(
-+ struct xenbus_watch *watch, const char **vec, unsigned int len)
-+{
-+ int cpu;
-+ char *cpustr;
-+ const char *node = vec[XS_WATCH_PATH];
++ dev->tx_queue_len = 0;
+
-+ if ((cpustr = strstr(node, "cpu/")) != NULL) {
-+ sscanf(cpustr, "cpu/%d", &cpu);
-+ vcpu_hotplug(cpu);
-+ }
-+}
++ dev->features = (NETIF_F_HIGHDMA |
++ NETIF_F_LLTX |
++ NETIF_F_TSO |
++ NETIF_F_SG |
++ NETIF_F_IP_CSUM);
+
-+static int smpboot_cpu_notify(struct notifier_block *notifier,
-+ unsigned long action, void *hcpu)
-+{
-+ int cpu = (long)hcpu;
++ SET_ETHTOOL_OPS(dev, &network_ethtool_ops);
+
+ /*
-+ * We do this in a callback notifier rather than __cpu_disable()
-+ * because local_cpu_hotplug_request() does not work in the latter
-+ * as it's always executed from within a stopmachine kthread.
++ * We do not set a jumbo MTU on the interface. Otherwise the network
++ * stack will try to send large packets that will get dropped by the
++ * Ethernet bridge (unless the physical Ethernet interface is
++ * configured to transfer jumbo packets). If a larger MTU is desired
++ * then the system administrator can specify it using the 'ifconfig'
++ * command.
+ */
-+ if ((action == CPU_DOWN_PREPARE) && local_cpu_hotplug_request())
-+ cpu_clear(cpu, local_allowed_cpumask);
-+
-+ return NOTIFY_OK;
-+}
-+
-+static int setup_cpu_watcher(struct notifier_block *notifier,
-+ unsigned long event, void *data)
-+{
-+ int i;
-+
-+ static struct xenbus_watch cpu_watch = {
-+ .node = "cpu",
-+ .callback = handle_vcpu_hotplug_event,
-+ .flags = XBWF_new_thread };
-+ (void)register_xenbus_watch(&cpu_watch);
-+
-+ if (!is_initial_xendomain()) {
-+ for_each_possible_cpu(i)
-+ vcpu_hotplug(i);
-+ printk(KERN_INFO "Brought up %ld CPUs\n",
-+ (long)num_online_cpus());
-+ }
-+
-+ return NOTIFY_DONE;
++ /*dev->mtu = 16*1024;*/
+}
+
-+static int __init setup_vcpu_hotplug_event(void)
++static int __init make_loopback(int i)
+{
-+ static struct notifier_block hotplug_cpu = {
-+ .notifier_call = smpboot_cpu_notify };
-+ static struct notifier_block xsn_cpu = {
-+ .notifier_call = setup_cpu_watcher };
++ struct net_device *dev1, *dev2;
++ char dev_name[IFNAMSIZ];
++ int err = -ENOMEM;
+
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++ sprintf(dev_name, "vif0.%d", i);
++ dev1 = alloc_netdev(sizeof(struct net_private), dev_name, ether_setup);
++ if (!dev1)
++ return err;
+
-+ register_cpu_notifier(&hotplug_cpu);
-+ register_xenstore_notifier(&xsn_cpu);
++ sprintf(dev_name, "veth%d", i);
++ dev2 = alloc_netdev(sizeof(struct net_private), dev_name, ether_setup);
++ if (!dev2)
++ goto fail_netdev2;
+
-+ return 0;
-+}
++ loopback_construct(dev1, dev2);
++ loopback_construct(dev2, dev1);
+
-+arch_initcall(setup_vcpu_hotplug_event);
++ /*
++ * Initialise a dummy MAC address for the 'dummy backend' interface. We
++ * choose the numerically largest non-broadcast address to prevent the
++ * address getting stolen by an Ethernet bridge for STP purposes.
++ */
++ memset(dev1->dev_addr, 0xFF, ETH_ALEN);
++ dev1->dev_addr[0] &= ~0x01;
+
-+int smp_suspend(void)
-+{
-+ int cpu, err;
++ if ((err = register_netdev(dev1)) != 0)
++ goto fail;
+
-+ for_each_online_cpu(cpu) {
-+ if (cpu == 0)
-+ continue;
-+ err = cpu_down(cpu);
-+ if (err) {
-+ printk(KERN_CRIT "Failed to take all CPUs "
-+ "down: %d.\n", err);
-+ for_each_possible_cpu(cpu)
-+ vcpu_hotplug(cpu);
-+ return err;
-+ }
++ if ((err = register_netdev(dev2)) != 0) {
++ unregister_netdev(dev1);
++ goto fail;
+ }
+
+ return 0;
-+}
-+
-+void smp_resume(void)
-+{
-+ int cpu;
+
-+ for_each_possible_cpu(cpu)
-+ vcpu_hotplug(cpu);
++ fail:
++ free_netdev(dev2);
++ fail_netdev2:
++ free_netdev(dev1);
++ return err;
+}
+
-+int cpu_up_check(unsigned int cpu)
++static void __exit clean_loopback(int i)
+{
-+ int rc = 0;
++ struct net_device *dev1, *dev2;
++ char dev_name[IFNAMSIZ];
+
-+ if (local_cpu_hotplug_request()) {
-+ cpu_set(cpu, local_allowed_cpumask);
-+ if (!cpu_isset(cpu, xenbus_allowed_cpumask)) {
-+ printk("%s: attempt to bring up CPU %u disallowed by "
-+ "remote admin.\n", __FUNCTION__, cpu);
-+ rc = -EBUSY;
-+ }
-+ } else if (!cpu_isset(cpu, local_allowed_cpumask) ||
-+ !cpu_isset(cpu, xenbus_allowed_cpumask)) {
-+ rc = -EBUSY;
++ sprintf(dev_name, "vif0.%d", i);
++ dev1 = dev_get_by_name(dev_name);
++ sprintf(dev_name, "veth%d", i);
++ dev2 = dev_get_by_name(dev_name);
++ if (dev1 && dev2) {
++ unregister_netdev(dev2);
++ unregister_netdev(dev1);
++ free_netdev(dev2);
++ free_netdev(dev1);
+ }
++}
+
-+ return rc;
++static int __init loopback_init(void)
++{
++ int i, err = 0;
++
++ if (nloopbacks == -1)
++ nloopbacks = is_initial_xendomain() ? 4 : 0;
++
++ for (i = 0; i < nloopbacks; i++)
++ if ((err = make_loopback(i)) != 0)
++ break;
++
++ return err;
+}
+
-+void init_xenbus_allowed_cpumask(void)
++module_init(loopback_init);
++
++static void __exit loopback_exit(void)
+{
-+ xenbus_allowed_cpumask = cpu_present_map;
++ int i;
++
++ for (i = nloopbacks; i-- > 0; )
++ clean_loopback(i);
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/core/evtchn.c tmp-linux-2.6-xen.patch/drivers/xen/core/evtchn.c
---- pristine-linux-2.6.18.2/drivers/xen/core/evtchn.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/core/evtchn.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,1015 @@
++
++module_exit(loopback_exit);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/netback/netback.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/netback/netback.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1614 @@
+/******************************************************************************
-+ * evtchn.c
++ * drivers/xen/netback/netback.c
+ *
-+ * Communication via Xen event channels.
++ * Back-end of the driver for virtual network devices. This portion of the
++ * driver exports a 'unified' network-device interface that can be accessed
++ * by any operating system that implements a compatible front end. A
++ * reference front-end implementation can be found in:
++ * drivers/xen/netfront/netfront.c
+ *
+ * Copyright (c) 2002-2005, K A Fraser
+ *
@@ -58033,28282 +94184,31418 @@
+ * IN THE SOFTWARE.
+ */
+
-+#include <linux/module.h>
-+#include <linux/irq.h>
-+#include <linux/interrupt.h>
-+#include <linux/sched.h>
-+#include <linux/kernel_stat.h>
-+#include <linux/version.h>
-+#include <asm/atomic.h>
-+#include <asm/system.h>
-+#include <asm/ptrace.h>
-+#include <asm/synch_bitops.h>
-+#include <xen/evtchn.h>
-+#include <xen/interface/event_channel.h>
-+#include <xen/interface/physdev.h>
-+#include <asm/hypervisor.h>
-+#include <linux/mc146818rtc.h> /* RTC_IRQ */
++#include "common.h"
++#include <xen/balloon.h>
++#include <xen/interface/memory.h>
+
-+/*
-+ * This lock protects updates to the following mapping and reference-count
-+ * arrays. The lock does not need to be acquired to read the mapping tables.
-+ */
-+static DEFINE_SPINLOCK(irq_mapping_update_lock);
++/*define NETBE_DEBUG_INTERRUPT*/
+
-+/* IRQ <-> event-channel mappings. */
-+static int evtchn_to_irq[NR_EVENT_CHANNELS] = {
-+ [0 ... NR_EVENT_CHANNELS-1] = -1 };
++/* extra field used in struct page */
++#define netif_page_index(pg) (*(long *)&(pg)->mapping)
+
-+/* Packed IRQ information: binding type, sub-type index, and event channel. */
-+static u32 irq_info[NR_IRQS];
++struct netbk_rx_meta {
++ skb_frag_t frag;
++ int id;
++ u8 copy:1;
++};
+
-+/* Binding types. */
-+enum {
-+ IRQT_UNBOUND,
-+ IRQT_PIRQ,
-+ IRQT_VIRQ,
-+ IRQT_IPI,
-+ IRQT_LOCAL_PORT,
-+ IRQT_CALLER_PORT
++struct netbk_tx_pending_inuse {
++ struct list_head list;
++ unsigned long alloc_time;
+};
+
-+/* Constructor for packed IRQ information. */
-+static inline u32 mk_irq_info(u32 type, u32 index, u32 evtchn)
-+{
-+ return ((type << 24) | (index << 16) | evtchn);
-+}
++static void netif_idx_release(u16 pending_idx);
++static void netif_page_release(struct page *page);
++static void make_tx_response(netif_t *netif,
++ netif_tx_request_t *txp,
++ s8 st);
++static netif_rx_response_t *make_rx_response(netif_t *netif,
++ u16 id,
++ s8 st,
++ u16 offset,
++ u16 size,
++ u16 flags);
+
-+/* Convenient shorthand for packed representation of an unbound IRQ. */
-+#define IRQ_UNBOUND mk_irq_info(IRQT_UNBOUND, 0, 0)
++static void net_tx_action(unsigned long unused);
++static DECLARE_TASKLET(net_tx_tasklet, net_tx_action, 0);
+
-+/*
-+ * Accessors for packed IRQ information.
-+ */
++static void net_rx_action(unsigned long unused);
++static DECLARE_TASKLET(net_rx_tasklet, net_rx_action, 0);
+
-+static inline unsigned int evtchn_from_irq(int irq)
-+{
-+ return (u16)(irq_info[irq]);
-+}
++static struct timer_list net_timer;
++static struct timer_list netbk_tx_pending_timer;
+
-+static inline unsigned int index_from_irq(int irq)
++#define MAX_PENDING_REQS 256
++
++static struct sk_buff_head rx_queue;
++
++static struct page **mmap_pages;
++static inline unsigned long idx_to_pfn(unsigned int idx)
+{
-+ return (u8)(irq_info[irq] >> 16);
++ return page_to_pfn(mmap_pages[idx]);
+}
+
-+static inline unsigned int type_from_irq(int irq)
++static inline unsigned long idx_to_kaddr(unsigned int idx)
+{
-+ return (u8)(irq_info[irq] >> 24);
++ return (unsigned long)pfn_to_kaddr(idx_to_pfn(idx));
+}
+
-+/* IRQ <-> VIRQ mapping. */
-+DEFINE_PER_CPU(int, virq_to_irq[NR_VIRQS]) = {[0 ... NR_VIRQS-1] = -1};
++#define PKT_PROT_LEN 64
+
-+/* IRQ <-> IPI mapping. */
-+#ifndef NR_IPIS
-+#define NR_IPIS 1
-+#endif
-+DEFINE_PER_CPU(int, ipi_to_irq[NR_IPIS]) = {[0 ... NR_IPIS-1] = -1};
++static struct pending_tx_info {
++ netif_tx_request_t req;
++ netif_t *netif;
++} pending_tx_info[MAX_PENDING_REQS];
++static u16 pending_ring[MAX_PENDING_REQS];
++typedef unsigned int PEND_RING_IDX;
++#define MASK_PEND_IDX(_i) ((_i)&(MAX_PENDING_REQS-1))
++static PEND_RING_IDX pending_prod, pending_cons;
++#define NR_PENDING_REQS (MAX_PENDING_REQS - pending_prod + pending_cons)
+
-+/* Reference counts for bindings to IRQs. */
-+static int irq_bindcount[NR_IRQS];
++/* Freed TX SKBs get batched on this ring before return to pending_ring. */
++static u16 dealloc_ring[MAX_PENDING_REQS];
++static PEND_RING_IDX dealloc_prod, dealloc_cons;
+
-+/* Bitmap indicating which PIRQs require Xen to be notified on unmask. */
-+static DECLARE_BITMAP(pirq_needs_eoi, NR_PIRQS);
++/* Doubly-linked list of in-use pending entries. */
++static struct netbk_tx_pending_inuse pending_inuse[MAX_PENDING_REQS];
++static LIST_HEAD(pending_inuse_head);
+
-+#ifdef CONFIG_SMP
++static struct sk_buff_head tx_queue;
+
-+static u8 cpu_evtchn[NR_EVENT_CHANNELS];
-+static unsigned long cpu_evtchn_mask[NR_CPUS][NR_EVENT_CHANNELS/BITS_PER_LONG];
++static grant_handle_t grant_tx_handle[MAX_PENDING_REQS];
++static gnttab_unmap_grant_ref_t tx_unmap_ops[MAX_PENDING_REQS];
++static gnttab_map_grant_ref_t tx_map_ops[MAX_PENDING_REQS];
+
-+static inline unsigned long active_evtchns(unsigned int cpu, shared_info_t *sh,
-+ unsigned int idx)
-+{
-+ return (sh->evtchn_pending[idx] &
-+ cpu_evtchn_mask[cpu][idx] &
-+ ~sh->evtchn_mask[idx]);
-+}
++static struct list_head net_schedule_list;
++static spinlock_t net_schedule_list_lock;
+
-+static void bind_evtchn_to_cpu(unsigned int chn, unsigned int cpu)
-+{
-+ int irq = evtchn_to_irq[chn];
++#define MAX_MFN_ALLOC 64
++static unsigned long mfn_list[MAX_MFN_ALLOC];
++static unsigned int alloc_index = 0;
+
-+ BUG_ON(irq == -1);
-+ set_native_irq_info(irq, cpumask_of_cpu(cpu));
++/* Setting this allows the safe use of this driver without netloop. */
++static int MODPARM_copy_skb = 1;
++module_param_named(copy_skb, MODPARM_copy_skb, bool, 0);
++MODULE_PARM_DESC(copy_skb, "Copy data received from netfront without netloop");
+
-+ clear_bit(chn, (unsigned long *)cpu_evtchn_mask[cpu_evtchn[chn]]);
-+ set_bit(chn, (unsigned long *)cpu_evtchn_mask[cpu]);
-+ cpu_evtchn[chn] = cpu;
-+}
++int netbk_copy_skb_mode;
+
-+static void init_evtchn_cpu_bindings(void)
++static inline unsigned long alloc_mfn(void)
+{
-+ int i;
-+
-+ /* By default all event channels notify CPU#0. */
-+ for (i = 0; i < NR_IRQS; i++)
-+ set_native_irq_info(i, cpumask_of_cpu(0));
-+
-+ memset(cpu_evtchn, 0, sizeof(cpu_evtchn));
-+ memset(cpu_evtchn_mask[0], ~0, sizeof(cpu_evtchn_mask[0]));
++ BUG_ON(alloc_index == 0);
++ return mfn_list[--alloc_index];
+}
+
-+static inline unsigned int cpu_from_evtchn(unsigned int evtchn)
++static int check_mfn(int nr)
+{
-+ return cpu_evtchn[evtchn];
-+}
++ struct xen_memory_reservation reservation = {
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
++ int rc;
+
-+#else
++ if (likely(alloc_index >= nr))
++ return 0;
+
-+static inline unsigned long active_evtchns(unsigned int cpu, shared_info_t *sh,
-+ unsigned int idx)
-+{
-+ return (sh->evtchn_pending[idx] & ~sh->evtchn_mask[idx]);
++ set_xen_guest_handle(reservation.extent_start, mfn_list + alloc_index);
++ reservation.nr_extents = MAX_MFN_ALLOC - alloc_index;
++ rc = HYPERVISOR_memory_op(XENMEM_increase_reservation, &reservation);
++ if (likely(rc > 0))
++ alloc_index += rc;
++
++ return alloc_index >= nr ? 0 : -ENOMEM;
+}
+
-+static void bind_evtchn_to_cpu(unsigned int chn, unsigned int cpu)
++static inline void maybe_schedule_tx_action(void)
+{
++ smp_mb();
++ if ((NR_PENDING_REQS < (MAX_PENDING_REQS/2)) &&
++ !list_empty(&net_schedule_list))
++ tasklet_schedule(&net_tx_tasklet);
+}
+
-+static void init_evtchn_cpu_bindings(void)
++static struct sk_buff *netbk_copy_skb(struct sk_buff *skb)
+{
++ struct skb_shared_info *ninfo;
++ struct sk_buff *nskb;
++ unsigned long offset;
++ int ret;
++ int len;
++ int headlen;
++
++ BUG_ON(skb_shinfo(skb)->frag_list != NULL);
++
++ nskb = alloc_skb(SKB_MAX_HEAD(0), GFP_ATOMIC | __GFP_NOWARN);
++ if (unlikely(!nskb))
++ goto err;
++
++ skb_reserve(nskb, 16 + NET_IP_ALIGN);
++ headlen = nskb->end - nskb->data;
++ if (headlen > skb_headlen(skb))
++ headlen = skb_headlen(skb);
++ ret = skb_copy_bits(skb, 0, __skb_put(nskb, headlen), headlen);
++ BUG_ON(ret);
++
++ ninfo = skb_shinfo(nskb);
++ ninfo->gso_size = skb_shinfo(skb)->gso_size;
++ ninfo->gso_type = skb_shinfo(skb)->gso_type;
++
++ offset = headlen;
++ len = skb->len - headlen;
++
++ nskb->len = skb->len;
++ nskb->data_len = len;
++ nskb->truesize += len;
++
++ while (len) {
++ struct page *page;
++ int copy;
++ int zero;
++
++ if (unlikely(ninfo->nr_frags >= MAX_SKB_FRAGS)) {
++ dump_stack();
++ goto err_free;
++ }
++
++ copy = len >= PAGE_SIZE ? PAGE_SIZE : len;
++ zero = len >= PAGE_SIZE ? 0 : __GFP_ZERO;
++
++ page = alloc_page(GFP_ATOMIC | __GFP_NOWARN | zero);
++ if (unlikely(!page))
++ goto err_free;
++
++ ret = skb_copy_bits(skb, offset, page_address(page), copy);
++ BUG_ON(ret);
++
++ ninfo->frags[ninfo->nr_frags].page = page;
++ ninfo->frags[ninfo->nr_frags].page_offset = 0;
++ ninfo->frags[ninfo->nr_frags].size = copy;
++ ninfo->nr_frags++;
++
++ offset += copy;
++ len -= copy;
++ }
++
++ offset = nskb->data - skb->data;
++
++ nskb->h.raw = skb->h.raw + offset;
++ nskb->nh.raw = skb->nh.raw + offset;
++ nskb->mac.raw = skb->mac.raw + offset;
++
++ return nskb;
++
++ err_free:
++ kfree_skb(nskb);
++ err:
++ return NULL;
+}
+
-+static inline unsigned int cpu_from_evtchn(unsigned int evtchn)
++static inline int netbk_max_required_rx_slots(netif_t *netif)
+{
-+ return 0;
++ if (netif->features & (NETIF_F_SG|NETIF_F_TSO))
++ return MAX_SKB_FRAGS + 2; /* header + extra_info + frags */
++ return 1; /* all in one */
+}
+
-+#endif
-+
-+/* Upcall to generic IRQ layer. */
-+#ifdef CONFIG_X86
-+extern fastcall unsigned int do_IRQ(struct pt_regs *regs);
-+void __init xen_init_IRQ(void);
-+void __init init_IRQ(void)
++static inline int netbk_queue_full(netif_t *netif)
+{
-+ irq_ctx_init(0);
-+ xen_init_IRQ();
-+}
-+#if defined (__i386__)
-+static inline void exit_idle(void) {}
-+#define IRQ_REG orig_eax
-+#elif defined (__x86_64__)
-+#include <asm/idle.h>
-+#define IRQ_REG orig_rax
-+#endif
-+#define do_IRQ(irq, regs) do { \
-+ (regs)->IRQ_REG = ~(irq); \
-+ do_IRQ((regs)); \
-+} while (0)
-+#endif
++ RING_IDX peek = netif->rx_req_cons_peek;
++ RING_IDX needed = netbk_max_required_rx_slots(netif);
+
-+/* Xen will never allocate port zero for any purpose. */
-+#define VALID_EVTCHN(chn) ((chn) != 0)
++ return ((netif->rx.sring->req_prod - peek) < needed) ||
++ ((netif->rx.rsp_prod_pvt + NET_RX_RING_SIZE - peek) < needed);
++}
+
-+/*
-+ * Force a proper event-channel callback from Xen after clearing the
-+ * callback mask. We do this in a very simple manner, by making a call
-+ * down into Xen. The pending flag will be checked by Xen on return.
-+ */
-+void force_evtchn_callback(void)
++static void tx_queue_callback(unsigned long data)
+{
-+ (void)HYPERVISOR_xen_version(0, NULL);
++ netif_t *netif = (netif_t *)data;
++ if (netif_schedulable(netif))
++ netif_wake_queue(netif->dev);
+}
-+/* Not a GPL symbol: used in ubiquitous macros, so too restrictive. */
-+EXPORT_SYMBOL(force_evtchn_callback);
-+
-+static DEFINE_PER_CPU(unsigned int, upcall_count) = { 0 };
+
-+/* NB. Interrupts are disabled on entry. */
-+asmlinkage void evtchn_do_upcall(struct pt_regs *regs)
++int netif_be_start_xmit(struct sk_buff *skb, struct net_device *dev)
+{
-+ unsigned long l1, l2;
-+ unsigned int l1i, l2i, port, count;
-+ int irq, cpu = smp_processor_id();
-+ shared_info_t *s = HYPERVISOR_shared_info;
-+ vcpu_info_t *vcpu_info = &s->vcpu_info[cpu];
++ netif_t *netif = netdev_priv(dev);
+
-+ do {
-+ /* Avoid a callback storm when we reenable delivery. */
-+ vcpu_info->evtchn_upcall_pending = 0;
++ BUG_ON(skb->dev != dev);
+
-+ /* Nested invocations bail immediately. */
-+ if (unlikely(per_cpu(upcall_count, cpu)++))
-+ return;
++ /* Drop the packet if the target domain has no receive buffers. */
++ if (unlikely(!netif_schedulable(netif) || netbk_queue_full(netif)))
++ goto drop;
+
-+#ifndef CONFIG_X86 /* No need for a barrier -- XCHG is a barrier on x86. */
-+ /* Clear master flag /before/ clearing selector flag. */
-+ rmb();
-+#endif
-+ l1 = xchg(&vcpu_info->evtchn_pending_sel, 0);
-+ while (l1 != 0) {
-+ l1i = __ffs(l1);
-+ l1 &= ~(1UL << l1i);
++ /*
++ * Copy the packet here if it's destined for a flipping interface
++ * but isn't flippable (e.g. extra references to data).
++ * XXX For now we also copy skbuffs whose head crosses a page
++ * boundary, because netbk_gop_skb can't handle them.
++ */
++ if (!netif->copying_receiver ||
++ ((skb_headlen(skb) + offset_in_page(skb->data)) >= PAGE_SIZE)) {
++ struct sk_buff *nskb = netbk_copy_skb(skb);
++ if ( unlikely(nskb == NULL) )
++ goto drop;
++ /* Copy only the header fields we use in this driver. */
++ nskb->dev = skb->dev;
++ nskb->ip_summed = skb->ip_summed;
++ nskb->proto_data_valid = skb->proto_data_valid;
++ dev_kfree_skb(skb);
++ skb = nskb;
++ }
+
-+ while ((l2 = active_evtchns(cpu, s, l1i)) != 0) {
-+ l2i = __ffs(l2);
++ netif->rx_req_cons_peek += skb_shinfo(skb)->nr_frags + 1 +
++ !!skb_shinfo(skb)->gso_size;
++ netif_get(netif);
+
-+ port = (l1i * BITS_PER_LONG) + l2i;
-+ if ((irq = evtchn_to_irq[port]) != -1)
-+ do_IRQ(irq, regs);
-+ else {
-+ exit_idle();
-+ evtchn_device_upcall(port);
-+ }
-+ }
++ if (netbk_can_queue(dev) && netbk_queue_full(netif)) {
++ netif->rx.sring->req_event = netif->rx_req_cons_peek +
++ netbk_max_required_rx_slots(netif);
++ mb(); /* request notification /then/ check & stop the queue */
++ if (netbk_queue_full(netif)) {
++ netif_stop_queue(dev);
++ /*
++ * Schedule 500ms timeout to restart the queue, thus
++ * ensuring that an inactive queue will be drained.
++ * Packets will be immediately be dropped until more
++ * receive buffers become available (see
++ * netbk_queue_full() check above).
++ */
++ netif->tx_queue_timeout.data = (unsigned long)netif;
++ netif->tx_queue_timeout.function = tx_queue_callback;
++ __mod_timer(&netif->tx_queue_timeout, jiffies + HZ/2);
+ }
++ }
+
-+ /* If there were nested callbacks then we have more to do. */
-+ count = per_cpu(upcall_count, cpu);
-+ per_cpu(upcall_count, cpu) = 0;
-+ } while (unlikely(count != 1));
-+}
-+
-+static int find_unbound_irq(void)
-+{
-+ static int warned;
-+ int dynirq, irq;
++ skb_queue_tail(&rx_queue, skb);
++ tasklet_schedule(&net_rx_tasklet);
+
-+ for (dynirq = 0; dynirq < NR_DYNIRQS; dynirq++) {
-+ irq = dynirq_to_irq(dynirq);
-+ if (irq_bindcount[irq] == 0)
-+ return irq;
-+ }
++ return 0;
+
-+ if (!warned) {
-+ warned = 1;
-+ printk(KERN_WARNING "No available IRQ to bind to: "
-+ "increase NR_DYNIRQS.\n");
-+ }
++ drop:
++ netif->stats.tx_dropped++;
++ dev_kfree_skb(skb);
++ return 0;
++}
+
-+ return -ENOSPC;
++#if 0
++static void xen_network_done_notify(void)
++{
++ static struct net_device *eth0_dev = NULL;
++ if (unlikely(eth0_dev == NULL))
++ eth0_dev = __dev_get_by_name("eth0");
++ netif_rx_schedule(eth0_dev);
++}
++/*
++ * Add following to poll() function in NAPI driver (Tigon3 is example):
++ * if ( xen_network_done() )
++ * tg3_enable_ints(tp);
++ */
++int xen_network_done(void)
++{
++ return skb_queue_empty(&rx_queue);
+}
++#endif
+
-+static int bind_caller_port_to_irq(unsigned int caller_port)
++struct netrx_pending_operations {
++ unsigned trans_prod, trans_cons;
++ unsigned mmu_prod, mmu_mcl;
++ unsigned mcl_prod, mcl_cons;
++ unsigned copy_prod, copy_cons;
++ unsigned meta_prod, meta_cons;
++ mmu_update_t *mmu;
++ gnttab_transfer_t *trans;
++ gnttab_copy_t *copy;
++ multicall_entry_t *mcl;
++ struct netbk_rx_meta *meta;
++};
++
++/* Set up the grant operations for this fragment. If it's a flipping
++ interface, we also set up the unmap request from here. */
++static u16 netbk_gop_frag(netif_t *netif, struct netbk_rx_meta *meta,
++ int i, struct netrx_pending_operations *npo,
++ struct page *page, unsigned long size,
++ unsigned long offset)
+{
-+ int irq;
++ mmu_update_t *mmu;
++ gnttab_transfer_t *gop;
++ gnttab_copy_t *copy_gop;
++ multicall_entry_t *mcl;
++ netif_rx_request_t *req;
++ unsigned long old_mfn, new_mfn;
+
-+ spin_lock(&irq_mapping_update_lock);
++ old_mfn = virt_to_mfn(page_address(page));
+
-+ if ((irq = evtchn_to_irq[caller_port]) == -1) {
-+ if ((irq = find_unbound_irq()) < 0)
-+ goto out;
++ req = RING_GET_REQUEST(&netif->rx, netif->rx.req_cons + i);
++ if (netif->copying_receiver) {
++ /* The fragment needs to be copied rather than
++ flipped. */
++ meta->copy = 1;
++ copy_gop = npo->copy + npo->copy_prod++;
++ copy_gop->flags = GNTCOPY_dest_gref;
++ if (PageForeign(page)) {
++ struct pending_tx_info *src_pend =
++ &pending_tx_info[netif_page_index(page)];
++ copy_gop->source.domid = src_pend->netif->domid;
++ copy_gop->source.u.ref = src_pend->req.gref;
++ copy_gop->flags |= GNTCOPY_source_gref;
++ } else {
++ copy_gop->source.domid = DOMID_SELF;
++ copy_gop->source.u.gmfn = old_mfn;
++ }
++ copy_gop->source.offset = offset;
++ copy_gop->dest.domid = netif->domid;
++ copy_gop->dest.offset = 0;
++ copy_gop->dest.u.ref = req->gref;
++ copy_gop->len = size;
++ } else {
++ meta->copy = 0;
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ new_mfn = alloc_mfn();
+
-+ evtchn_to_irq[caller_port] = irq;
-+ irq_info[irq] = mk_irq_info(IRQT_CALLER_PORT, 0, caller_port);
-+ }
++ /*
++ * Set the new P2M table entry before
++ * reassigning the old data page. Heed the
++ * comment in pgtable-2level.h:pte_page(). :-)
++ */
++ set_phys_to_machine(page_to_pfn(page), new_mfn);
++
++ mcl = npo->mcl + npo->mcl_prod++;
++ MULTI_update_va_mapping(mcl,
++ (unsigned long)page_address(page),
++ pfn_pte_ma(new_mfn, PAGE_KERNEL),
++ 0);
+
-+ irq_bindcount[irq]++;
++ mmu = npo->mmu + npo->mmu_prod++;
++ mmu->ptr = ((maddr_t)new_mfn << PAGE_SHIFT) |
++ MMU_MACHPHYS_UPDATE;
++ mmu->val = page_to_pfn(page);
++ }
+
-+ out:
-+ spin_unlock(&irq_mapping_update_lock);
-+ return irq;
++ gop = npo->trans + npo->trans_prod++;
++ gop->mfn = old_mfn;
++ gop->domid = netif->domid;
++ gop->ref = req->gref;
++ }
++ return req->id;
+}
+
-+static int bind_local_port_to_irq(unsigned int local_port)
++static void netbk_gop_skb(struct sk_buff *skb,
++ struct netrx_pending_operations *npo)
+{
-+ int irq;
-+
-+ spin_lock(&irq_mapping_update_lock);
++ netif_t *netif = netdev_priv(skb->dev);
++ int nr_frags = skb_shinfo(skb)->nr_frags;
++ int i;
++ int extra;
++ struct netbk_rx_meta *head_meta, *meta;
+
-+ BUG_ON(evtchn_to_irq[local_port] != -1);
++ head_meta = npo->meta + npo->meta_prod++;
++ head_meta->frag.page_offset = skb_shinfo(skb)->gso_type;
++ head_meta->frag.size = skb_shinfo(skb)->gso_size;
++ extra = !!head_meta->frag.size + 1;
+
-+ if ((irq = find_unbound_irq()) < 0) {
-+ struct evtchn_close close = { .port = local_port };
-+ if (HYPERVISOR_event_channel_op(EVTCHNOP_close, &close))
-+ BUG();
-+ goto out;
++ for (i = 0; i < nr_frags; i++) {
++ meta = npo->meta + npo->meta_prod++;
++ meta->frag = skb_shinfo(skb)->frags[i];
++ meta->id = netbk_gop_frag(netif, meta, i + extra, npo,
++ meta->frag.page,
++ meta->frag.size,
++ meta->frag.page_offset);
+ }
+
-+ evtchn_to_irq[local_port] = irq;
-+ irq_info[irq] = mk_irq_info(IRQT_LOCAL_PORT, 0, local_port);
-+ irq_bindcount[irq]++;
++ /*
++ * This must occur at the end to ensure that we don't trash skb_shinfo
++ * until we're done. We know that the head doesn't cross a page
++ * boundary because such packets get copied in netif_be_start_xmit.
++ */
++ head_meta->id = netbk_gop_frag(netif, head_meta, 0, npo,
++ virt_to_page(skb->data),
++ skb_headlen(skb),
++ offset_in_page(skb->data));
+
-+ out:
-+ spin_unlock(&irq_mapping_update_lock);
-+ return irq;
++ netif->rx.req_cons += nr_frags + extra;
+}
+
-+static int bind_listening_port_to_irq(unsigned int remote_domain)
++static inline void netbk_free_pages(int nr_frags, struct netbk_rx_meta *meta)
+{
-+ struct evtchn_alloc_unbound alloc_unbound;
-+ int err;
-+
-+ alloc_unbound.dom = DOMID_SELF;
-+ alloc_unbound.remote_dom = remote_domain;
-+
-+ err = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
-+ &alloc_unbound);
++ int i;
+
-+ return err ? : bind_local_port_to_irq(alloc_unbound.port);
++ for (i = 0; i < nr_frags; i++)
++ put_page(meta[i].frag.page);
+}
+
-+static int bind_interdomain_evtchn_to_irq(unsigned int remote_domain,
-+ unsigned int remote_port)
++/* This is a twin to netbk_gop_skb. Assume that netbk_gop_skb was
++ used to set up the operations on the top of
++ netrx_pending_operations, which have since been done. Check that
++ they didn't give any errors and advance over them. */
++static int netbk_check_gop(int nr_frags, domid_t domid,
++ struct netrx_pending_operations *npo)
+{
-+ struct evtchn_bind_interdomain bind_interdomain;
-+ int err;
++ multicall_entry_t *mcl;
++ gnttab_transfer_t *gop;
++ gnttab_copy_t *copy_op;
++ int status = NETIF_RSP_OKAY;
++ int i;
+
-+ bind_interdomain.remote_dom = remote_domain;
-+ bind_interdomain.remote_port = remote_port;
++ for (i = 0; i <= nr_frags; i++) {
++ if (npo->meta[npo->meta_cons + i].copy) {
++ copy_op = npo->copy + npo->copy_cons++;
++ if (copy_op->status != GNTST_okay) {
++ DPRINTK("Bad status %d from copy to DOM%d.\n",
++ copy_op->status, domid);
++ status = NETIF_RSP_ERROR;
++ }
++ } else {
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ mcl = npo->mcl + npo->mcl_cons++;
++ /* The update_va_mapping() must not fail. */
++ BUG_ON(mcl->result != 0);
++ }
+
-+ err = HYPERVISOR_event_channel_op(EVTCHNOP_bind_interdomain,
-+ &bind_interdomain);
++ gop = npo->trans + npo->trans_cons++;
++ /* Check the reassignment error code. */
++ if (gop->status != 0) {
++ DPRINTK("Bad status %d from grant transfer to DOM%u\n",
++ gop->status, domid);
++ /*
++ * Page no longer belongs to us unless
++ * GNTST_bad_page, but that should be
++ * a fatal error anyway.
++ */
++ BUG_ON(gop->status == GNTST_bad_page);
++ status = NETIF_RSP_ERROR;
++ }
++ }
++ }
+
-+ return err ? : bind_local_port_to_irq(bind_interdomain.local_port);
++ return status;
+}
+
-+static int bind_virq_to_irq(unsigned int virq, unsigned int cpu)
++static void netbk_add_frag_responses(netif_t *netif, int status,
++ struct netbk_rx_meta *meta, int nr_frags)
+{
-+ struct evtchn_bind_virq bind_virq;
-+ int evtchn, irq;
-+
-+ spin_lock(&irq_mapping_update_lock);
-+
-+ if ((irq = per_cpu(virq_to_irq, cpu)[virq]) == -1) {
-+ if ((irq = find_unbound_irq()) < 0)
-+ goto out;
-+
-+ bind_virq.virq = virq;
-+ bind_virq.vcpu = cpu;
-+ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
-+ &bind_virq) != 0)
-+ BUG();
-+ evtchn = bind_virq.port;
-+
-+ evtchn_to_irq[evtchn] = irq;
-+ irq_info[irq] = mk_irq_info(IRQT_VIRQ, virq, evtchn);
++ int i;
++ unsigned long offset;
+
-+ per_cpu(virq_to_irq, cpu)[virq] = irq;
++ for (i = 0; i < nr_frags; i++) {
++ int id = meta[i].id;
++ int flags = (i == nr_frags - 1) ? 0 : NETRXF_more_data;
+
-+ bind_evtchn_to_cpu(evtchn, cpu);
++ if (meta[i].copy)
++ offset = 0;
++ else
++ offset = meta[i].frag.page_offset;
++ make_rx_response(netif, id, status, offset,
++ meta[i].frag.size, flags);
+ }
-+
-+ irq_bindcount[irq]++;
-+
-+ out:
-+ spin_unlock(&irq_mapping_update_lock);
-+ return irq;
+}
+
-+static int bind_ipi_to_irq(unsigned int ipi, unsigned int cpu)
++static void net_rx_action(unsigned long unused)
+{
-+ struct evtchn_bind_ipi bind_ipi;
-+ int evtchn, irq;
-+
-+ spin_lock(&irq_mapping_update_lock);
-+
-+ if ((irq = per_cpu(ipi_to_irq, cpu)[ipi]) == -1) {
-+ if ((irq = find_unbound_irq()) < 0)
-+ goto out;
-+
-+ bind_ipi.vcpu = cpu;
-+ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_ipi,
-+ &bind_ipi) != 0)
-+ BUG();
-+ evtchn = bind_ipi.port;
-+
-+ evtchn_to_irq[evtchn] = irq;
-+ irq_info[irq] = mk_irq_info(IRQT_IPI, ipi, evtchn);
-+
-+ per_cpu(ipi_to_irq, cpu)[ipi] = irq;
-+
-+ bind_evtchn_to_cpu(evtchn, cpu);
-+ }
++ netif_t *netif = NULL;
++ s8 status;
++ u16 id, irq, flags;
++ netif_rx_response_t *resp;
++ multicall_entry_t *mcl;
++ struct sk_buff_head rxq;
++ struct sk_buff *skb;
++ int notify_nr = 0;
++ int ret;
++ int nr_frags;
++ int count;
++ unsigned long offset;
+
-+ irq_bindcount[irq]++;
++ /*
++ * Putting hundreds of bytes on the stack is considered rude.
++ * Static works because a tasklet can only be on one CPU at any time.
++ */
++ static multicall_entry_t rx_mcl[NET_RX_RING_SIZE+3];
++ static mmu_update_t rx_mmu[NET_RX_RING_SIZE];
++ static gnttab_transfer_t grant_trans_op[NET_RX_RING_SIZE];
++ static gnttab_copy_t grant_copy_op[NET_RX_RING_SIZE];
++ static unsigned char rx_notify[NR_IRQS];
++ static u16 notify_list[NET_RX_RING_SIZE];
++ static struct netbk_rx_meta meta[NET_RX_RING_SIZE];
+
-+ out:
-+ spin_unlock(&irq_mapping_update_lock);
-+ return irq;
-+}
++ struct netrx_pending_operations npo = {
++ mmu: rx_mmu,
++ trans: grant_trans_op,
++ copy: grant_copy_op,
++ mcl: rx_mcl,
++ meta: meta};
+
-+static void unbind_from_irq(unsigned int irq)
-+{
-+ struct evtchn_close close;
-+ int cpu, evtchn = evtchn_from_irq(irq);
++ skb_queue_head_init(&rxq);
+
-+ spin_lock(&irq_mapping_update_lock);
++ count = 0;
+
-+ if ((--irq_bindcount[irq] == 0) && VALID_EVTCHN(evtchn)) {
-+ close.port = evtchn;
-+ if ((type_from_irq(irq) != IRQT_CALLER_PORT) &&
-+ HYPERVISOR_event_channel_op(EVTCHNOP_close, &close))
-+ BUG();
++ while ((skb = skb_dequeue(&rx_queue)) != NULL) {
++ nr_frags = skb_shinfo(skb)->nr_frags;
++ *(int *)skb->cb = nr_frags;
+
-+ switch (type_from_irq(irq)) {
-+ case IRQT_VIRQ:
-+ per_cpu(virq_to_irq, cpu_from_evtchn(evtchn))
-+ [index_from_irq(irq)] = -1;
-+ break;
-+ case IRQT_IPI:
-+ per_cpu(ipi_to_irq, cpu_from_evtchn(evtchn))
-+ [index_from_irq(irq)] = -1;
-+ break;
-+ default:
++ if (!xen_feature(XENFEAT_auto_translated_physmap) &&
++ !((netif_t *)netdev_priv(skb->dev))->copying_receiver &&
++ check_mfn(nr_frags + 1)) {
++ /* Memory squeeze? Back off for an arbitrary while. */
++ if ( net_ratelimit() )
++ WPRINTK("Memory squeeze in netback "
++ "driver.\n");
++ mod_timer(&net_timer, jiffies + HZ);
++ skb_queue_head(&rx_queue, skb);
+ break;
+ }
+
-+ /* Closed ports are implicitly re-bound to VCPU0. */
-+ bind_evtchn_to_cpu(evtchn, 0);
-+
-+ evtchn_to_irq[evtchn] = -1;
-+ irq_info[irq] = IRQ_UNBOUND;
-+
-+ /* Zap stats across IRQ changes of use. */
-+ for_each_possible_cpu(cpu)
-+ kstat_cpu(cpu).irqs[irq] = 0;
-+ }
-+
-+ spin_unlock(&irq_mapping_update_lock);
-+}
++ netbk_gop_skb(skb, &npo);
+
-+int bind_caller_port_to_irqhandler(
-+ unsigned int caller_port,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
-+ unsigned long irqflags,
-+ const char *devname,
-+ void *dev_id)
-+{
-+ int irq, retval;
++ count += nr_frags + 1;
+
-+ irq = bind_caller_port_to_irq(caller_port);
-+ if (irq < 0)
-+ return irq;
++ __skb_queue_tail(&rxq, skb);
+
-+ retval = request_irq(irq, handler, irqflags, devname, dev_id);
-+ if (retval != 0) {
-+ unbind_from_irq(irq);
-+ return retval;
++ /* Filled the batch queue? */
++ if (count + MAX_SKB_FRAGS >= NET_RX_RING_SIZE)
++ break;
+ }
+
-+ return irq;
-+}
-+EXPORT_SYMBOL_GPL(bind_caller_port_to_irqhandler);
++ BUG_ON(npo.meta_prod > ARRAY_SIZE(meta));
+
-+int bind_listening_port_to_irqhandler(
-+ unsigned int remote_domain,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
-+ unsigned long irqflags,
-+ const char *devname,
-+ void *dev_id)
-+{
-+ int irq, retval;
++ npo.mmu_mcl = npo.mcl_prod;
++ if (npo.mcl_prod) {
++ BUG_ON(xen_feature(XENFEAT_auto_translated_physmap));
++ BUG_ON(npo.mmu_prod > ARRAY_SIZE(rx_mmu));
++ mcl = npo.mcl + npo.mcl_prod++;
+
-+ irq = bind_listening_port_to_irq(remote_domain);
-+ if (irq < 0)
-+ return irq;
++ BUG_ON(mcl[-1].op != __HYPERVISOR_update_va_mapping);
++ mcl[-1].args[MULTI_UVMFLAGS_INDEX] = UVMF_TLB_FLUSH|UVMF_ALL;
+
-+ retval = request_irq(irq, handler, irqflags, devname, dev_id);
-+ if (retval != 0) {
-+ unbind_from_irq(irq);
-+ return retval;
++ mcl->op = __HYPERVISOR_mmu_update;
++ mcl->args[0] = (unsigned long)rx_mmu;
++ mcl->args[1] = npo.mmu_prod;
++ mcl->args[2] = 0;
++ mcl->args[3] = DOMID_SELF;
+ }
+
-+ return irq;
-+}
-+EXPORT_SYMBOL_GPL(bind_listening_port_to_irqhandler);
-+
-+int bind_interdomain_evtchn_to_irqhandler(
-+ unsigned int remote_domain,
-+ unsigned int remote_port,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
-+ unsigned long irqflags,
-+ const char *devname,
-+ void *dev_id)
-+{
-+ int irq, retval;
-+
-+ irq = bind_interdomain_evtchn_to_irq(remote_domain, remote_port);
-+ if (irq < 0)
-+ return irq;
++ if (npo.trans_prod) {
++ BUG_ON(npo.trans_prod > ARRAY_SIZE(grant_trans_op));
++ mcl = npo.mcl + npo.mcl_prod++;
++ mcl->op = __HYPERVISOR_grant_table_op;
++ mcl->args[0] = GNTTABOP_transfer;
++ mcl->args[1] = (unsigned long)grant_trans_op;
++ mcl->args[2] = npo.trans_prod;
++ }
+
-+ retval = request_irq(irq, handler, irqflags, devname, dev_id);
-+ if (retval != 0) {
-+ unbind_from_irq(irq);
-+ return retval;
++ if (npo.copy_prod) {
++ BUG_ON(npo.copy_prod > ARRAY_SIZE(grant_copy_op));
++ mcl = npo.mcl + npo.mcl_prod++;
++ mcl->op = __HYPERVISOR_grant_table_op;
++ mcl->args[0] = GNTTABOP_copy;
++ mcl->args[1] = (unsigned long)grant_copy_op;
++ mcl->args[2] = npo.copy_prod;
+ }
+
-+ return irq;
-+}
-+EXPORT_SYMBOL_GPL(bind_interdomain_evtchn_to_irqhandler);
++ /* Nothing to do? */
++ if (!npo.mcl_prod)
++ return;
+
-+int bind_virq_to_irqhandler(
-+ unsigned int virq,
-+ unsigned int cpu,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
-+ unsigned long irqflags,
-+ const char *devname,
-+ void *dev_id)
-+{
-+ int irq, retval;
++ BUG_ON(npo.mcl_prod > ARRAY_SIZE(rx_mcl));
+
-+ irq = bind_virq_to_irq(virq, cpu);
-+ if (irq < 0)
-+ return irq;
++ ret = HYPERVISOR_multicall(npo.mcl, npo.mcl_prod);
++ BUG_ON(ret != 0);
++ /* The mmu_machphys_update() must not fail. */
++ BUG_ON(npo.mmu_mcl && npo.mcl[npo.mmu_mcl].result != 0);
+
-+ retval = request_irq(irq, handler, irqflags, devname, dev_id);
-+ if (retval != 0) {
-+ unbind_from_irq(irq);
-+ return retval;
-+ }
++ while ((skb = __skb_dequeue(&rxq)) != NULL) {
++ nr_frags = *(int *)skb->cb;
+
-+ return irq;
-+}
-+EXPORT_SYMBOL_GPL(bind_virq_to_irqhandler);
++ netif = netdev_priv(skb->dev);
++ /* We can't rely on skb_release_data to release the
++ pages used by fragments for us, since it tries to
++ touch the pages in the fraglist. If we're in
++ flipping mode, that doesn't work. In copying mode,
++ we still have access to all of the pages, and so
++ it's safe to let release_data deal with it. */
++ /* (Freeing the fragments is safe since we copy
++ non-linear skbs destined for flipping interfaces) */
++ if (!netif->copying_receiver) {
++ atomic_set(&(skb_shinfo(skb)->dataref), 1);
++ skb_shinfo(skb)->frag_list = NULL;
++ skb_shinfo(skb)->nr_frags = 0;
++ netbk_free_pages(nr_frags, meta + npo.meta_cons + 1);
++ }
+
-+int bind_ipi_to_irqhandler(
-+ unsigned int ipi,
-+ unsigned int cpu,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
-+ unsigned long irqflags,
-+ const char *devname,
-+ void *dev_id)
-+{
-+ int irq, retval;
++ netif->stats.tx_bytes += skb->len;
++ netif->stats.tx_packets++;
+
-+ irq = bind_ipi_to_irq(ipi, cpu);
-+ if (irq < 0)
-+ return irq;
++ status = netbk_check_gop(nr_frags, netif->domid, &npo);
+
-+ retval = request_irq(irq, handler, irqflags, devname, dev_id);
-+ if (retval != 0) {
-+ unbind_from_irq(irq);
-+ return retval;
-+ }
++ id = meta[npo.meta_cons].id;
++ flags = nr_frags ? NETRXF_more_data : 0;
+
-+ return irq;
-+}
-+EXPORT_SYMBOL_GPL(bind_ipi_to_irqhandler);
++ if (skb->ip_summed == CHECKSUM_HW) /* local packet? */
++ flags |= NETRXF_csum_blank | NETRXF_data_validated;
++ else if (skb->proto_data_valid) /* remote but checksummed? */
++ flags |= NETRXF_data_validated;
+
-+void unbind_from_irqhandler(unsigned int irq, void *dev_id)
-+{
-+ free_irq(irq, dev_id);
-+ unbind_from_irq(irq);
-+}
-+EXPORT_SYMBOL_GPL(unbind_from_irqhandler);
++ if (meta[npo.meta_cons].copy)
++ offset = 0;
++ else
++ offset = offset_in_page(skb->data);
++ resp = make_rx_response(netif, id, status, offset,
++ skb_headlen(skb), flags);
+
-+#ifdef CONFIG_SMP
-+/* Rebind an evtchn so that it gets delivered to a specific cpu */
-+static void rebind_irq_to_cpu(unsigned irq, unsigned tcpu)
-+{
-+ struct evtchn_bind_vcpu bind_vcpu;
-+ int evtchn = evtchn_from_irq(irq);
++ if (meta[npo.meta_cons].frag.size) {
++ struct netif_extra_info *gso =
++ (struct netif_extra_info *)
++ RING_GET_RESPONSE(&netif->rx,
++ netif->rx.rsp_prod_pvt++);
+
-+ if (!VALID_EVTCHN(evtchn))
-+ return;
++ resp->flags |= NETRXF_extra_info;
++
++ gso->u.gso.size = meta[npo.meta_cons].frag.size;
++ gso->u.gso.type = XEN_NETIF_GSO_TYPE_TCPV4;
++ gso->u.gso.pad = 0;
++ gso->u.gso.features = 0;
+
-+ /* Send future instances of this interrupt to other vcpu. */
-+ bind_vcpu.port = evtchn;
-+ bind_vcpu.vcpu = tcpu;
++ gso->type = XEN_NETIF_EXTRA_TYPE_GSO;
++ gso->flags = 0;
++ }
+
-+ /*
-+ * If this fails, it usually just indicates that we're dealing with a
-+ * virq or IPI channel, which don't actually need to be rebound. Ignore
-+ * it, but don't do the xenlinux-level rebind in that case.
-+ */
-+ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_vcpu, &bind_vcpu) >= 0)
-+ bind_evtchn_to_cpu(evtchn, tcpu);
-+}
++ netbk_add_frag_responses(netif, status,
++ meta + npo.meta_cons + 1,
++ nr_frags);
+
-+static void set_affinity_irq(unsigned irq, cpumask_t dest)
-+{
-+ unsigned tcpu = first_cpu(dest);
-+ rebind_irq_to_cpu(irq, tcpu);
-+}
-+#endif
++ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&netif->rx, ret);
++ irq = netif->irq;
++ if (ret && !rx_notify[irq]) {
++ rx_notify[irq] = 1;
++ notify_list[notify_nr++] = irq;
++ }
+
-+int resend_irq_on_evtchn(unsigned int irq)
-+{
-+ int masked, evtchn = evtchn_from_irq(irq);
-+ shared_info_t *s = HYPERVISOR_shared_info;
++ if (netif_queue_stopped(netif->dev) &&
++ netif_schedulable(netif) &&
++ !netbk_queue_full(netif))
++ netif_wake_queue(netif->dev);
+
-+ if (!VALID_EVTCHN(evtchn))
-+ return 1;
++ netif_put(netif);
++ dev_kfree_skb(skb);
++ npo.meta_cons += nr_frags + 1;
++ }
+
-+ masked = synch_test_and_set_bit(evtchn, s->evtchn_mask);
-+ synch_set_bit(evtchn, s->evtchn_pending);
-+ if (!masked)
-+ unmask_evtchn(evtchn);
++ while (notify_nr != 0) {
++ irq = notify_list[--notify_nr];
++ rx_notify[irq] = 0;
++ notify_remote_via_irq(irq);
++ }
+
-+ return 1;
++ /* More work to do? */
++ if (!skb_queue_empty(&rx_queue) && !timer_pending(&net_timer))
++ tasklet_schedule(&net_rx_tasklet);
++#if 0
++ else
++ xen_network_done_notify();
++#endif
+}
+
-+/*
-+ * Interface to generic handling in irq.c
-+ */
-+
-+static unsigned int startup_dynirq(unsigned int irq)
++static void net_alarm(unsigned long unused)
+{
-+ int evtchn = evtchn_from_irq(irq);
-+
-+ if (VALID_EVTCHN(evtchn))
-+ unmask_evtchn(evtchn);
-+ return 0;
++ tasklet_schedule(&net_rx_tasklet);
+}
+
-+static void shutdown_dynirq(unsigned int irq)
++static void netbk_tx_pending_timeout(unsigned long unused)
+{
-+ int evtchn = evtchn_from_irq(irq);
++ tasklet_schedule(&net_tx_tasklet);
++}
+
-+ if (VALID_EVTCHN(evtchn))
-+ mask_evtchn(evtchn);
++struct net_device_stats *netif_be_get_stats(struct net_device *dev)
++{
++ netif_t *netif = netdev_priv(dev);
++ return &netif->stats;
+}
+
-+static void enable_dynirq(unsigned int irq)
++static int __on_net_schedule_list(netif_t *netif)
+{
-+ int evtchn = evtchn_from_irq(irq);
++ return netif->list.next != NULL;
++}
+
-+ if (VALID_EVTCHN(evtchn))
-+ unmask_evtchn(evtchn);
++static void remove_from_net_schedule_list(netif_t *netif)
++{
++ spin_lock_irq(&net_schedule_list_lock);
++ if (likely(__on_net_schedule_list(netif))) {
++ list_del(&netif->list);
++ netif->list.next = NULL;
++ netif_put(netif);
++ }
++ spin_unlock_irq(&net_schedule_list_lock);
+}
+
-+static void disable_dynirq(unsigned int irq)
++static void add_to_net_schedule_list_tail(netif_t *netif)
+{
-+ int evtchn = evtchn_from_irq(irq);
++ if (__on_net_schedule_list(netif))
++ return;
+
-+ if (VALID_EVTCHN(evtchn))
-+ mask_evtchn(evtchn);
++ spin_lock_irq(&net_schedule_list_lock);
++ if (!__on_net_schedule_list(netif) &&
++ likely(netif_schedulable(netif))) {
++ list_add_tail(&netif->list, &net_schedule_list);
++ netif_get(netif);
++ }
++ spin_unlock_irq(&net_schedule_list_lock);
+}
+
-+static void ack_dynirq(unsigned int irq)
++/*
++ * Note on CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER:
++ * If this driver is pipelining transmit requests then we can be very
++ * aggressive in avoiding new-packet notifications -- frontend only needs to
++ * send a notification if there are no outstanding unreceived responses.
++ * If we may be buffer transmit buffers for any reason then we must be rather
++ * more conservative and treat this as the final check for pending work.
++ */
++void netif_schedule_work(netif_t *netif)
+{
-+ int evtchn = evtchn_from_irq(irq);
++ int more_to_do;
+
-+ move_native_irq(irq);
++#ifdef CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER
++ more_to_do = RING_HAS_UNCONSUMED_REQUESTS(&netif->tx);
++#else
++ RING_FINAL_CHECK_FOR_REQUESTS(&netif->tx, more_to_do);
++#endif
+
-+ if (VALID_EVTCHN(evtchn)) {
-+ mask_evtchn(evtchn);
-+ clear_evtchn(evtchn);
++ if (more_to_do) {
++ add_to_net_schedule_list_tail(netif);
++ maybe_schedule_tx_action();
+ }
+}
+
-+static void end_dynirq(unsigned int irq)
++void netif_deschedule_work(netif_t *netif)
+{
-+ int evtchn = evtchn_from_irq(irq);
-+
-+ if (VALID_EVTCHN(evtchn) && !(irq_desc[irq].status & IRQ_DISABLED))
-+ unmask_evtchn(evtchn);
++ remove_from_net_schedule_list(netif);
+}
+
-+static struct hw_interrupt_type dynirq_type = {
-+ .typename = "Dynamic-irq",
-+ .startup = startup_dynirq,
-+ .shutdown = shutdown_dynirq,
-+ .enable = enable_dynirq,
-+ .disable = disable_dynirq,
-+ .ack = ack_dynirq,
-+ .end = end_dynirq,
-+#ifdef CONFIG_SMP
-+ .set_affinity = set_affinity_irq,
-+#endif
-+ .retrigger = resend_irq_on_evtchn,
-+};
+
-+static inline void pirq_unmask_notify(int pirq)
++static void tx_add_credit(netif_t *netif)
+{
-+ struct physdev_eoi eoi = { .irq = pirq };
-+ if (unlikely(test_bit(pirq, pirq_needs_eoi)))
-+ (void)HYPERVISOR_physdev_op(PHYSDEVOP_eoi, &eoi);
++ unsigned long max_burst, max_credit;
++
++ /*
++ * Allow a burst big enough to transmit a jumbo packet of up to 128kB.
++ * Otherwise the interface can seize up due to insufficient credit.
++ */
++ max_burst = RING_GET_REQUEST(&netif->tx, netif->tx.req_cons)->size;
++ max_burst = min(max_burst, 131072UL);
++ max_burst = max(max_burst, netif->credit_bytes);
++
++ /* Take care that adding a new chunk of credit doesn't wrap to zero. */
++ max_credit = netif->remaining_credit + netif->credit_bytes;
++ if (max_credit < netif->remaining_credit)
++ max_credit = ULONG_MAX; /* wrapped: clamp to ULONG_MAX */
++
++ netif->remaining_credit = min(max_credit, max_burst);
+}
+
-+static inline void pirq_query_unmask(int pirq)
++static void tx_credit_callback(unsigned long data)
+{
-+ struct physdev_irq_status_query irq_status;
-+ irq_status.irq = pirq;
-+ (void)HYPERVISOR_physdev_op(PHYSDEVOP_irq_status_query, &irq_status);
-+ clear_bit(pirq, pirq_needs_eoi);
-+ if (irq_status.flags & XENIRQSTAT_needs_eoi)
-+ set_bit(pirq, pirq_needs_eoi);
++ netif_t *netif = (netif_t *)data;
++ tx_add_credit(netif);
++ netif_schedule_work(netif);
+}
+
-+/*
-+ * On startup, if there is no action associated with the IRQ then we are
-+ * probing. In this case we should not share with others as it will confuse us.
-+ */
-+#define probing_irq(_irq) (irq_desc[(_irq)].action == NULL)
++static inline int copy_pending_req(PEND_RING_IDX pending_idx)
++{
++ return gnttab_copy_grant_page(grant_tx_handle[pending_idx],
++ &mmap_pages[pending_idx]);
++}
+
-+static unsigned int startup_pirq(unsigned int irq)
++inline static void net_tx_action_dealloc(void)
+{
-+ struct evtchn_bind_pirq bind_pirq;
-+ int evtchn = evtchn_from_irq(irq);
++ struct netbk_tx_pending_inuse *inuse, *n;
++ gnttab_unmap_grant_ref_t *gop;
++ u16 pending_idx;
++ PEND_RING_IDX dc, dp;
++ netif_t *netif;
++ int ret;
++ LIST_HEAD(list);
+
-+ if (VALID_EVTCHN(evtchn))
-+ goto out;
++ dc = dealloc_cons;
++ gop = tx_unmap_ops;
+
-+ bind_pirq.pirq = irq;
-+ /* NB. We are happy to share unless we are probing. */
-+ bind_pirq.flags = probing_irq(irq) ? 0 : BIND_PIRQ__WILL_SHARE;
-+ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_pirq, &bind_pirq) != 0) {
-+ if (!probing_irq(irq))
-+ printk(KERN_INFO "Failed to obtain physical IRQ %d\n",
-+ irq);
-+ return 0;
-+ }
-+ evtchn = bind_pirq.port;
++ /*
++ * Free up any grants we have finished using
++ */
++ do {
++ dp = dealloc_prod;
+
-+ pirq_query_unmask(irq_to_pirq(irq));
++ /* Ensure we see all indices enqueued by netif_idx_release(). */
++ smp_rmb();
+
-+ evtchn_to_irq[evtchn] = irq;
-+ bind_evtchn_to_cpu(evtchn, 0);
-+ irq_info[irq] = mk_irq_info(IRQT_PIRQ, irq, evtchn);
++ while (dc != dp) {
++ unsigned long pfn;
+
-+ out:
-+ unmask_evtchn(evtchn);
-+ pirq_unmask_notify(irq_to_pirq(irq));
++ pending_idx = dealloc_ring[MASK_PEND_IDX(dc++)];
++ list_move_tail(&pending_inuse[pending_idx].list, &list);
+
-+ return 0;
-+}
++ pfn = idx_to_pfn(pending_idx);
++ /* Already unmapped? */
++ if (!phys_to_machine_mapping_valid(pfn))
++ continue;
+
-+static void shutdown_pirq(unsigned int irq)
-+{
-+ struct evtchn_close close;
-+ int evtchn = evtchn_from_irq(irq);
++ gnttab_set_unmap_op(gop, idx_to_kaddr(pending_idx),
++ GNTMAP_host_map,
++ grant_tx_handle[pending_idx]);
++ gop++;
++ }
+
-+ if (!VALID_EVTCHN(evtchn))
-+ return;
++ if (netbk_copy_skb_mode != NETBK_DELAYED_COPY_SKB ||
++ list_empty(&pending_inuse_head))
++ break;
+
-+ mask_evtchn(evtchn);
++ /* Copy any entries that have been pending for too long. */
++ list_for_each_entry_safe(inuse, n, &pending_inuse_head, list) {
++ if (time_after(inuse->alloc_time + HZ / 2, jiffies))
++ break;
+
-+ close.port = evtchn;
-+ if (HYPERVISOR_event_channel_op(EVTCHNOP_close, &close) != 0)
-+ BUG();
++ switch (copy_pending_req(inuse - pending_inuse)) {
++ case 0:
++ list_move_tail(&inuse->list, &list);
++ continue;
++ case -EBUSY:
++ list_del_init(&inuse->list);
++ continue;
++ case -ENOENT:
++ continue;
++ }
+
-+ bind_evtchn_to_cpu(evtchn, 0);
-+ evtchn_to_irq[evtchn] = -1;
-+ irq_info[irq] = IRQ_UNBOUND;
-+}
++ break;
++ }
++ } while (dp != dealloc_prod);
+
-+static void enable_pirq(unsigned int irq)
-+{
-+ int evtchn = evtchn_from_irq(irq);
++ dealloc_cons = dc;
+
-+ if (VALID_EVTCHN(evtchn)) {
-+ unmask_evtchn(evtchn);
-+ pirq_unmask_notify(irq_to_pirq(irq));
-+ }
-+}
++ ret = HYPERVISOR_grant_table_op(
++ GNTTABOP_unmap_grant_ref, tx_unmap_ops, gop - tx_unmap_ops);
++ BUG_ON(ret);
+
-+static void disable_pirq(unsigned int irq)
-+{
-+ int evtchn = evtchn_from_irq(irq);
++ list_for_each_entry_safe(inuse, n, &list, list) {
++ pending_idx = inuse - pending_inuse;
+
-+ if (VALID_EVTCHN(evtchn))
-+ mask_evtchn(evtchn);
-+}
++ netif = pending_tx_info[pending_idx].netif;
+
-+static void ack_pirq(unsigned int irq)
-+{
-+ int evtchn = evtchn_from_irq(irq);
++ make_tx_response(netif, &pending_tx_info[pending_idx].req,
++ NETIF_RSP_OKAY);
+
-+ move_native_irq(irq);
++ /* Ready for next use. */
++ gnttab_reset_grant_page(mmap_pages[pending_idx]);
+
-+ if (VALID_EVTCHN(evtchn)) {
-+ mask_evtchn(evtchn);
-+ clear_evtchn(evtchn);
++ pending_ring[MASK_PEND_IDX(pending_prod++)] = pending_idx;
++
++ netif_put(netif);
++
++ list_del_init(&inuse->list);
+ }
+}
+
-+static void end_pirq(unsigned int irq)
++static void netbk_tx_err(netif_t *netif, netif_tx_request_t *txp, RING_IDX end)
+{
-+ int evtchn = evtchn_from_irq(irq);
++ RING_IDX cons = netif->tx.req_cons;
+
-+ if (VALID_EVTCHN(evtchn) && !(irq_desc[irq].status & IRQ_DISABLED)) {
-+ unmask_evtchn(evtchn);
-+ pirq_unmask_notify(irq_to_pirq(irq));
-+ }
++ do {
++ make_tx_response(netif, txp, NETIF_RSP_ERROR);
++ if (cons >= end)
++ break;
++ txp = RING_GET_REQUEST(&netif->tx, cons++);
++ } while (1);
++ netif->tx.req_cons = cons;
++ netif_schedule_work(netif);
++ netif_put(netif);
+}
+
-+static struct hw_interrupt_type pirq_type = {
-+ .typename = "Phys-irq",
-+ .startup = startup_pirq,
-+ .shutdown = shutdown_pirq,
-+ .enable = enable_pirq,
-+ .disable = disable_pirq,
-+ .ack = ack_pirq,
-+ .end = end_pirq,
-+#ifdef CONFIG_SMP
-+ .set_affinity = set_affinity_irq,
-+#endif
-+ .retrigger = resend_irq_on_evtchn,
-+};
-+
-+int irq_ignore_unhandled(unsigned int irq)
++static int netbk_count_requests(netif_t *netif, netif_tx_request_t *first,
++ netif_tx_request_t *txp, int work_to_do)
+{
-+ struct physdev_irq_status_query irq_status = { .irq = irq };
++ RING_IDX cons = netif->tx.req_cons;
++ int frags = 0;
+
-+ if (!is_running_on_xen())
++ if (!(first->flags & NETTXF_more_data))
+ return 0;
+
-+ (void)HYPERVISOR_physdev_op(PHYSDEVOP_irq_status_query, &irq_status);
-+ return !!(irq_status.flags & XENIRQSTAT_shared);
-+}
++ do {
++ if (frags >= work_to_do) {
++ DPRINTK("Need more frags\n");
++ return -frags;
++ }
+
-+void notify_remote_via_irq(int irq)
-+{
-+ int evtchn = evtchn_from_irq(irq);
++ if (unlikely(frags >= MAX_SKB_FRAGS)) {
++ DPRINTK("Too many frags\n");
++ return -frags;
++ }
+
-+ if (VALID_EVTCHN(evtchn))
-+ notify_remote_via_evtchn(evtchn);
-+}
-+EXPORT_SYMBOL_GPL(notify_remote_via_irq);
++ memcpy(txp, RING_GET_REQUEST(&netif->tx, cons + frags),
++ sizeof(*txp));
++ if (txp->size > first->size) {
++ DPRINTK("Frags galore\n");
++ return -frags;
++ }
+
-+int irq_to_evtchn_port(int irq)
-+{
-+ return evtchn_from_irq(irq);
-+}
-+EXPORT_SYMBOL_GPL(irq_to_evtchn_port);
++ first->size -= txp->size;
++ frags++;
+
-+void mask_evtchn(int port)
-+{
-+ shared_info_t *s = HYPERVISOR_shared_info;
-+ synch_set_bit(port, s->evtchn_mask);
++ if (unlikely((txp->offset + txp->size) > PAGE_SIZE)) {
++ DPRINTK("txp->offset: %x, size: %u\n",
++ txp->offset, txp->size);
++ return -frags;
++ }
++ } while ((txp++)->flags & NETTXF_more_data);
++
++ return frags;
+}
-+EXPORT_SYMBOL_GPL(mask_evtchn);
+
-+void unmask_evtchn(int port)
++static gnttab_map_grant_ref_t *netbk_get_requests(netif_t *netif,
++ struct sk_buff *skb,
++ netif_tx_request_t *txp,
++ gnttab_map_grant_ref_t *mop)
+{
-+ shared_info_t *s = HYPERVISOR_shared_info;
-+ unsigned int cpu = smp_processor_id();
-+ vcpu_info_t *vcpu_info = &s->vcpu_info[cpu];
++ struct skb_shared_info *shinfo = skb_shinfo(skb);
++ skb_frag_t *frags = shinfo->frags;
++ unsigned long pending_idx = *((u16 *)skb->data);
++ int i, start;
+
-+ BUG_ON(!irqs_disabled());
++ /* Skip first skb fragment if it is on same page as header fragment. */
++ start = ((unsigned long)shinfo->frags[0].page == pending_idx);
+
-+ /* Slow path (hypercall) if this is a non-local port. */
-+ if (unlikely(cpu != cpu_from_evtchn(port))) {
-+ struct evtchn_unmask unmask = { .port = port };
-+ (void)HYPERVISOR_event_channel_op(EVTCHNOP_unmask, &unmask);
-+ return;
-+ }
++ for (i = start; i < shinfo->nr_frags; i++, txp++) {
++ pending_idx = pending_ring[MASK_PEND_IDX(pending_cons++)];
+
-+ synch_clear_bit(port, s->evtchn_mask);
++ gnttab_set_map_op(mop++, idx_to_kaddr(pending_idx),
++ GNTMAP_host_map | GNTMAP_readonly,
++ txp->gref, netif->domid);
+
-+ /* Did we miss an interrupt 'edge'? Re-fire if so. */
-+ if (synch_test_bit(port, s->evtchn_pending) &&
-+ !synch_test_and_set_bit(port / BITS_PER_LONG,
-+ &vcpu_info->evtchn_pending_sel))
-+ vcpu_info->evtchn_upcall_pending = 1;
++ memcpy(&pending_tx_info[pending_idx].req, txp, sizeof(*txp));
++ netif_get(netif);
++ pending_tx_info[pending_idx].netif = netif;
++ frags[i].page = (void *)pending_idx;
++ }
++
++ return mop;
+}
-+EXPORT_SYMBOL_GPL(unmask_evtchn);
+
-+static void restore_cpu_virqs(int cpu)
++static int netbk_tx_check_mop(struct sk_buff *skb,
++ gnttab_map_grant_ref_t **mopp)
+{
-+ struct evtchn_bind_virq bind_virq;
-+ int virq, irq, evtchn;
-+
-+ for (virq = 0; virq < NR_VIRQS; virq++) {
-+ if ((irq = per_cpu(virq_to_irq, cpu)[virq]) == -1)
-+ continue;
-+
-+ BUG_ON(irq_info[irq] != mk_irq_info(IRQT_VIRQ, virq, 0));
++ gnttab_map_grant_ref_t *mop = *mopp;
++ int pending_idx = *((u16 *)skb->data);
++ netif_t *netif = pending_tx_info[pending_idx].netif;
++ netif_tx_request_t *txp;
++ struct skb_shared_info *shinfo = skb_shinfo(skb);
++ int nr_frags = shinfo->nr_frags;
++ int i, err, start;
+
-+ /* Get a new binding from Xen. */
-+ bind_virq.virq = virq;
-+ bind_virq.vcpu = cpu;
-+ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
-+ &bind_virq) != 0)
-+ BUG();
-+ evtchn = bind_virq.port;
++ /* Check status of header. */
++ err = mop->status;
++ if (unlikely(err)) {
++ txp = &pending_tx_info[pending_idx].req;
++ make_tx_response(netif, txp, NETIF_RSP_ERROR);
++ pending_ring[MASK_PEND_IDX(pending_prod++)] = pending_idx;
++ netif_put(netif);
++ } else {
++ set_phys_to_machine(
++ __pa(idx_to_kaddr(pending_idx)) >> PAGE_SHIFT,
++ FOREIGN_FRAME(mop->dev_bus_addr >> PAGE_SHIFT));
++ grant_tx_handle[pending_idx] = mop->handle;
++ }
+
-+ /* Record the new mapping. */
-+ evtchn_to_irq[evtchn] = irq;
-+ irq_info[irq] = mk_irq_info(IRQT_VIRQ, virq, evtchn);
-+ bind_evtchn_to_cpu(evtchn, cpu);
++ /* Skip first skb fragment if it is on same page as header fragment. */
++ start = ((unsigned long)shinfo->frags[0].page == pending_idx);
+
-+ /* Ready for use. */
-+ unmask_evtchn(evtchn);
-+ }
-+}
++ for (i = start; i < nr_frags; i++) {
++ int j, newerr;
+
-+static void restore_cpu_ipis(int cpu)
-+{
-+ struct evtchn_bind_ipi bind_ipi;
-+ int ipi, irq, evtchn;
++ pending_idx = (unsigned long)shinfo->frags[i].page;
+
-+ for (ipi = 0; ipi < NR_IPIS; ipi++) {
-+ if ((irq = per_cpu(ipi_to_irq, cpu)[ipi]) == -1)
++ /* Check error status: if okay then remember grant handle. */
++ newerr = (++mop)->status;
++ if (likely(!newerr)) {
++ set_phys_to_machine(
++ __pa(idx_to_kaddr(pending_idx))>>PAGE_SHIFT,
++ FOREIGN_FRAME(mop->dev_bus_addr>>PAGE_SHIFT));
++ grant_tx_handle[pending_idx] = mop->handle;
++ /* Had a previous error? Invalidate this fragment. */
++ if (unlikely(err))
++ netif_idx_release(pending_idx);
+ continue;
++ }
+
-+ BUG_ON(irq_info[irq] != mk_irq_info(IRQT_IPI, ipi, 0));
-+
-+ /* Get a new binding from Xen. */
-+ bind_ipi.vcpu = cpu;
-+ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_ipi,
-+ &bind_ipi) != 0)
-+ BUG();
-+ evtchn = bind_ipi.port;
++ /* Error on this fragment: respond to client with an error. */
++ txp = &pending_tx_info[pending_idx].req;
++ make_tx_response(netif, txp, NETIF_RSP_ERROR);
++ pending_ring[MASK_PEND_IDX(pending_prod++)] = pending_idx;
++ netif_put(netif);
+
-+ /* Record the new mapping. */
-+ evtchn_to_irq[evtchn] = irq;
-+ irq_info[irq] = mk_irq_info(IRQT_IPI, ipi, evtchn);
-+ bind_evtchn_to_cpu(evtchn, cpu);
++ /* Not the first error? Preceding frags already invalidated. */
++ if (err)
++ continue;
+
-+ /* Ready for use. */
-+ unmask_evtchn(evtchn);
++ /* First error: invalidate header and preceding fragments. */
++ pending_idx = *((u16 *)skb->data);
++ netif_idx_release(pending_idx);
++ for (j = start; j < i; j++) {
++ pending_idx = (unsigned long)shinfo->frags[i].page;
++ netif_idx_release(pending_idx);
++ }
+
++ /* Remember the error: invalidate all subsequent fragments. */
++ err = newerr;
+ }
++
++ *mopp = mop + 1;
++ return err;
+}
+
-+void irq_resume(void)
++static void netbk_fill_frags(struct sk_buff *skb)
+{
-+ int cpu, pirq, irq, evtchn;
++ struct skb_shared_info *shinfo = skb_shinfo(skb);
++ int nr_frags = shinfo->nr_frags;
++ int i;
+
-+ init_evtchn_cpu_bindings();
++ for (i = 0; i < nr_frags; i++) {
++ skb_frag_t *frag = shinfo->frags + i;
++ netif_tx_request_t *txp;
++ unsigned long pending_idx;
+
-+ /* New event-channel space is not 'live' yet. */
-+ for (evtchn = 0; evtchn < NR_EVENT_CHANNELS; evtchn++)
-+ mask_evtchn(evtchn);
++ pending_idx = (unsigned long)frag->page;
+
-+ /* Check that no PIRQs are still bound. */
-+ for (pirq = 0; pirq < NR_PIRQS; pirq++)
-+ BUG_ON(irq_info[pirq_to_irq(pirq)] != IRQ_UNBOUND);
++ pending_inuse[pending_idx].alloc_time = jiffies;
++ list_add_tail(&pending_inuse[pending_idx].list,
++ &pending_inuse_head);
+
-+ /* No IRQ <-> event-channel mappings. */
-+ for (irq = 0; irq < NR_IRQS; irq++)
-+ irq_info[irq] &= ~0xFFFF; /* zap event-channel binding */
-+ for (evtchn = 0; evtchn < NR_EVENT_CHANNELS; evtchn++)
-+ evtchn_to_irq[evtchn] = -1;
++ txp = &pending_tx_info[pending_idx].req;
++ frag->page = virt_to_page(idx_to_kaddr(pending_idx));
++ frag->size = txp->size;
++ frag->page_offset = txp->offset;
+
-+ for_each_possible_cpu(cpu) {
-+ restore_cpu_virqs(cpu);
-+ restore_cpu_ipis(cpu);
++ skb->len += txp->size;
++ skb->data_len += txp->size;
++ skb->truesize += txp->size;
+ }
-+
+}
+
-+void __init xen_init_IRQ(void)
++int netbk_get_extras(netif_t *netif, struct netif_extra_info *extras,
++ int work_to_do)
+{
-+ int i;
++ struct netif_extra_info extra;
++ RING_IDX cons = netif->tx.req_cons;
+
-+ init_evtchn_cpu_bindings();
++ do {
++ if (unlikely(work_to_do-- <= 0)) {
++ DPRINTK("Missing extra info\n");
++ return -EBADR;
++ }
+
-+ /* No event channels are 'live' right now. */
-+ for (i = 0; i < NR_EVENT_CHANNELS; i++)
-+ mask_evtchn(i);
++ memcpy(&extra, RING_GET_REQUEST(&netif->tx, cons),
++ sizeof(extra));
++ if (unlikely(!extra.type ||
++ extra.type >= XEN_NETIF_EXTRA_TYPE_MAX)) {
++ netif->tx.req_cons = ++cons;
++ DPRINTK("Invalid extra type: %d\n", extra.type);
++ return -EINVAL;
++ }
+
-+ /* No IRQ -> event-channel mappings. */
-+ for (i = 0; i < NR_IRQS; i++)
-+ irq_info[i] = IRQ_UNBOUND;
++ memcpy(&extras[extra.type - 1], &extra, sizeof(extra));
++ netif->tx.req_cons = ++cons;
++ } while (extra.flags & XEN_NETIF_EXTRA_FLAG_MORE);
+
-+ /* Dynamic IRQ space is currently unbound. Zero the refcnts. */
-+ for (i = 0; i < NR_DYNIRQS; i++) {
-+ irq_bindcount[dynirq_to_irq(i)] = 0;
++ return work_to_do;
++}
+
-+ irq_desc[dynirq_to_irq(i)].status = IRQ_DISABLED;
-+ irq_desc[dynirq_to_irq(i)].action = NULL;
-+ irq_desc[dynirq_to_irq(i)].depth = 1;
-+ irq_desc[dynirq_to_irq(i)].chip = &dynirq_type;
++static int netbk_set_skb_gso(struct sk_buff *skb, struct netif_extra_info *gso)
++{
++ if (!gso->u.gso.size) {
++ DPRINTK("GSO size must not be zero.\n");
++ return -EINVAL;
++ }
++
++ /* Currently only TCPv4 S.O. is supported. */
++ if (gso->u.gso.type != XEN_NETIF_GSO_TYPE_TCPV4) {
++ DPRINTK("Bad GSO type %d.\n", gso->u.gso.type);
++ return -EINVAL;
+ }
+
-+ /* Phys IRQ space is statically bound (1:1 mapping). Nail refcnts. */
-+ for (i = 0; i < NR_PIRQS; i++) {
-+ irq_bindcount[pirq_to_irq(i)] = 1;
++ skb_shinfo(skb)->gso_size = gso->u.gso.size;
++ skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4;
+
-+#ifdef RTC_IRQ
-+ /* If not domain 0, force our RTC driver to fail its probe. */
-+ if ((i == RTC_IRQ) && !is_initial_xendomain())
-+ continue;
-+#endif
++ /* Header must be checked, and gso_segs computed. */
++ skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY;
++ skb_shinfo(skb)->gso_segs = 0;
+
-+ irq_desc[pirq_to_irq(i)].status = IRQ_DISABLED;
-+ irq_desc[pirq_to_irq(i)].action = NULL;
-+ irq_desc[pirq_to_irq(i)].depth = 1;
-+ irq_desc[pirq_to_irq(i)].chip = &pirq_type;
-+ }
++ return 0;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/core/features.c tmp-linux-2.6-xen.patch/drivers/xen/core/features.c
---- pristine-linux-2.6.18.2/drivers/xen/core/features.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/core/features.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,34 @@
-+/******************************************************************************
-+ * features.c
-+ *
-+ * Xen feature flags.
-+ *
-+ * Copyright (c) 2006, Ian Campbell, XenSource Inc.
-+ */
-+#include <linux/types.h>
-+#include <linux/cache.h>
-+#include <linux/module.h>
-+#include <asm/hypervisor.h>
-+#include <xen/features.h>
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
-+#endif
++/* Called after netfront has transmitted */
++static void net_tx_action(unsigned long unused)
++{
++ struct list_head *ent;
++ struct sk_buff *skb;
++ netif_t *netif;
++ netif_tx_request_t txreq;
++ netif_tx_request_t txfrags[MAX_SKB_FRAGS];
++ struct netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX - 1];
++ u16 pending_idx;
++ RING_IDX i;
++ gnttab_map_grant_ref_t *mop;
++ unsigned int data_len;
++ int ret, work_to_do;
+
-+u8 xen_features[XENFEAT_NR_SUBMAPS * 32] __read_mostly;
-+/* Not a GPL symbol: used in ubiquitous macros, so too restrictive. */
-+EXPORT_SYMBOL(xen_features);
++ if (dealloc_cons != dealloc_prod)
++ net_tx_action_dealloc();
+
-+void setup_xen_features(void)
-+{
-+ xen_feature_info_t fi;
-+ int i, j;
++ mop = tx_map_ops;
++ while (((NR_PENDING_REQS + MAX_SKB_FRAGS) < MAX_PENDING_REQS) &&
++ !list_empty(&net_schedule_list)) {
++ /* Get a netif from the list with work to do. */
++ ent = net_schedule_list.next;
++ netif = list_entry(ent, netif_t, list);
++ netif_get(netif);
++ remove_from_net_schedule_list(netif);
+
-+ for (i = 0; i < XENFEAT_NR_SUBMAPS; i++) {
-+ fi.submap_idx = i;
-+ if (HYPERVISOR_xen_version(XENVER_get_features, &fi) < 0)
-+ break;
-+ for (j=0; j<32; j++)
-+ xen_features[i*32+j] = !!(fi.submap & 1<<j);
-+ }
-+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/core/gnttab.c tmp-linux-2.6-xen.patch/drivers/xen/core/gnttab.c
---- pristine-linux-2.6.18.2/drivers/xen/core/gnttab.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/core/gnttab.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,631 @@
-+/******************************************************************************
-+ * gnttab.c
-+ *
-+ * Granting foreign access to our memory reservation.
-+ *
-+ * Copyright (c) 2005-2006, Christopher Clark
-+ * Copyright (c) 2004-2005, K A Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ RING_FINAL_CHECK_FOR_REQUESTS(&netif->tx, work_to_do);
++ if (!work_to_do) {
++ netif_put(netif);
++ continue;
++ }
+
-+#include <linux/module.h>
-+#include <linux/sched.h>
-+#include <linux/mm.h>
-+#include <xen/interface/xen.h>
-+#include <xen/gnttab.h>
-+#include <asm/pgtable.h>
-+#include <asm/uaccess.h>
-+#include <asm/synch_bitops.h>
-+#include <asm/io.h>
-+#include <xen/interface/memory.h>
-+#include <xen/driver_util.h>
++ i = netif->tx.req_cons;
++ rmb(); /* Ensure that we see the request before we copy it. */
++ memcpy(&txreq, RING_GET_REQUEST(&netif->tx, i), sizeof(txreq));
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
-+#endif
++ /* Credit-based scheduling. */
++ if (txreq.size > netif->remaining_credit) {
++ unsigned long now = jiffies;
++ unsigned long next_credit =
++ netif->credit_timeout.expires +
++ msecs_to_jiffies(netif->credit_usec / 1000);
+
-+/* External tools reserve first few grant table entries. */
-+#define NR_RESERVED_ENTRIES 8
-+#define GNTTAB_LIST_END 0xffffffff
-+#define GREFS_PER_GRANT_FRAME (PAGE_SIZE / sizeof(grant_entry_t))
++ /* Timer could already be pending in rare cases. */
++ if (timer_pending(&netif->credit_timeout)) {
++ netif_put(netif);
++ continue;
++ }
+
-+static grant_ref_t **gnttab_list;
-+static unsigned int nr_grant_frames;
-+static unsigned int boot_max_nr_grant_frames;
-+static int gnttab_free_count;
-+static grant_ref_t gnttab_free_head;
-+static DEFINE_SPINLOCK(gnttab_list_lock);
++ /* Passed the point where we can replenish credit? */
++ if (time_after_eq(now, next_credit)) {
++ netif->credit_timeout.expires = now;
++ tx_add_credit(netif);
++ }
+
-+static struct grant_entry *shared;
++ /* Still too big to send right now? Set a callback. */
++ if (txreq.size > netif->remaining_credit) {
++ netif->credit_timeout.data =
++ (unsigned long)netif;
++ netif->credit_timeout.function =
++ tx_credit_callback;
++ __mod_timer(&netif->credit_timeout,
++ next_credit);
++ netif_put(netif);
++ continue;
++ }
++ }
++ netif->remaining_credit -= txreq.size;
+
-+static struct gnttab_free_callback *gnttab_free_callback_list;
++ work_to_do--;
++ netif->tx.req_cons = ++i;
+
-+static int gnttab_expand(unsigned int req_entries);
++ memset(extras, 0, sizeof(extras));
++ if (txreq.flags & NETTXF_extra_info) {
++ work_to_do = netbk_get_extras(netif, extras,
++ work_to_do);
++ i = netif->tx.req_cons;
++ if (unlikely(work_to_do < 0)) {
++ netbk_tx_err(netif, &txreq, i);
++ continue;
++ }
++ }
+
-+#define RPP (PAGE_SIZE / sizeof(grant_ref_t))
-+#define gnttab_entry(entry) (gnttab_list[(entry) / RPP][(entry) % RPP])
++ ret = netbk_count_requests(netif, &txreq, txfrags, work_to_do);
++ if (unlikely(ret < 0)) {
++ netbk_tx_err(netif, &txreq, i - ret);
++ continue;
++ }
++ i += ret;
+
-+static int get_free_entries(int count)
-+{
-+ unsigned long flags;
-+ int ref, rc;
-+ grant_ref_t head;
++ if (unlikely(txreq.size < ETH_HLEN)) {
++ DPRINTK("Bad packet size: %d\n", txreq.size);
++ netbk_tx_err(netif, &txreq, i);
++ continue;
++ }
+
-+ spin_lock_irqsave(&gnttab_list_lock, flags);
++ /* No crossing a page as the payload mustn't fragment. */
++ if (unlikely((txreq.offset + txreq.size) > PAGE_SIZE)) {
++ DPRINTK("txreq.offset: %x, size: %u, end: %lu\n",
++ txreq.offset, txreq.size,
++ (txreq.offset &~PAGE_MASK) + txreq.size);
++ netbk_tx_err(netif, &txreq, i);
++ continue;
++ }
+
-+ if ((gnttab_free_count < count) &&
-+ ((rc = gnttab_expand(count - gnttab_free_count)) < 0)) {
-+ spin_unlock_irqrestore(&gnttab_list_lock, flags);
-+ return rc;
-+ }
++ pending_idx = pending_ring[MASK_PEND_IDX(pending_cons)];
+
-+ ref = head = gnttab_free_head;
-+ gnttab_free_count -= count;
-+ while (count-- > 1)
-+ head = gnttab_entry(head);
-+ gnttab_free_head = gnttab_entry(head);
-+ gnttab_entry(head) = GNTTAB_LIST_END;
++ data_len = (txreq.size > PKT_PROT_LEN &&
++ ret < MAX_SKB_FRAGS) ?
++ PKT_PROT_LEN : txreq.size;
+
-+ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++ skb = alloc_skb(data_len + 16 + NET_IP_ALIGN,
++ GFP_ATOMIC | __GFP_NOWARN);
++ if (unlikely(skb == NULL)) {
++ DPRINTK("Can't allocate a skb in start_xmit.\n");
++ netbk_tx_err(netif, &txreq, i);
++ break;
++ }
+
-+ return ref;
-+}
++ /* Packets passed to netif_rx() must have some headroom. */
++ skb_reserve(skb, 16 + NET_IP_ALIGN);
+
-+#define get_free_entry() get_free_entries(1)
++ if (extras[XEN_NETIF_EXTRA_TYPE_GSO - 1].type) {
++ struct netif_extra_info *gso;
++ gso = &extras[XEN_NETIF_EXTRA_TYPE_GSO - 1];
+
-+static void do_free_callbacks(void)
-+{
-+ struct gnttab_free_callback *callback, *next;
++ if (netbk_set_skb_gso(skb, gso)) {
++ kfree_skb(skb);
++ netbk_tx_err(netif, &txreq, i);
++ continue;
++ }
++ }
+
-+ callback = gnttab_free_callback_list;
-+ gnttab_free_callback_list = NULL;
++ gnttab_set_map_op(mop, idx_to_kaddr(pending_idx),
++ GNTMAP_host_map | GNTMAP_readonly,
++ txreq.gref, netif->domid);
++ mop++;
+
-+ while (callback != NULL) {
-+ next = callback->next;
-+ if (gnttab_free_count >= callback->count) {
-+ callback->next = NULL;
-+ callback->fn(callback->arg);
++ memcpy(&pending_tx_info[pending_idx].req,
++ &txreq, sizeof(txreq));
++ pending_tx_info[pending_idx].netif = netif;
++ *((u16 *)skb->data) = pending_idx;
++
++ __skb_put(skb, data_len);
++
++ skb_shinfo(skb)->nr_frags = ret;
++ if (data_len < txreq.size) {
++ skb_shinfo(skb)->nr_frags++;
++ skb_shinfo(skb)->frags[0].page =
++ (void *)(unsigned long)pending_idx;
+ } else {
-+ callback->next = gnttab_free_callback_list;
-+ gnttab_free_callback_list = callback;
++ /* Discriminate from any valid pending_idx value. */
++ skb_shinfo(skb)->frags[0].page = (void *)~0UL;
+ }
-+ callback = next;
-+ }
-+}
+
-+static inline void check_free_callbacks(void)
-+{
-+ if (unlikely(gnttab_free_callback_list))
-+ do_free_callbacks();
-+}
++ __skb_queue_tail(&tx_queue, skb);
+
-+static void put_free_entry(grant_ref_t ref)
-+{
-+ unsigned long flags;
-+ spin_lock_irqsave(&gnttab_list_lock, flags);
-+ gnttab_entry(ref) = gnttab_free_head;
-+ gnttab_free_head = ref;
-+ gnttab_free_count++;
-+ check_free_callbacks();
-+ spin_unlock_irqrestore(&gnttab_list_lock, flags);
-+}
++ pending_cons++;
+
-+/*
-+ * Public grant-issuing interface functions
-+ */
++ mop = netbk_get_requests(netif, skb, txfrags, mop);
+
-+int gnttab_grant_foreign_access(domid_t domid, unsigned long frame,
-+ int readonly)
-+{
-+ int ref;
++ netif->tx.req_cons = i;
++ netif_schedule_work(netif);
+
-+ if (unlikely((ref = get_free_entry()) < 0))
-+ return -ENOSPC;
++ if ((mop - tx_map_ops) >= ARRAY_SIZE(tx_map_ops))
++ break;
++ }
+
-+ shared[ref].frame = frame;
-+ shared[ref].domid = domid;
-+ wmb();
-+ shared[ref].flags = GTF_permit_access | (readonly ? GTF_readonly : 0);
++ if (mop == tx_map_ops)
++ return;
+
-+ return ref;
-+}
-+EXPORT_SYMBOL_GPL(gnttab_grant_foreign_access);
++ ret = HYPERVISOR_grant_table_op(
++ GNTTABOP_map_grant_ref, tx_map_ops, mop - tx_map_ops);
++ BUG_ON(ret);
+
-+void gnttab_grant_foreign_access_ref(grant_ref_t ref, domid_t domid,
-+ unsigned long frame, int readonly)
-+{
-+ shared[ref].frame = frame;
-+ shared[ref].domid = domid;
-+ wmb();
-+ shared[ref].flags = GTF_permit_access | (readonly ? GTF_readonly : 0);
-+}
-+EXPORT_SYMBOL_GPL(gnttab_grant_foreign_access_ref);
++ mop = tx_map_ops;
++ while ((skb = __skb_dequeue(&tx_queue)) != NULL) {
++ netif_tx_request_t *txp;
+
++ pending_idx = *((u16 *)skb->data);
++ netif = pending_tx_info[pending_idx].netif;
++ txp = &pending_tx_info[pending_idx].req;
+
-+int gnttab_query_foreign_access(grant_ref_t ref)
-+{
-+ u16 nflags;
++ /* Check the remap error code. */
++ if (unlikely(netbk_tx_check_mop(skb, &mop))) {
++ DPRINTK("netback grant failed.\n");
++ skb_shinfo(skb)->nr_frags = 0;
++ kfree_skb(skb);
++ continue;
++ }
+
-+ nflags = shared[ref].flags;
++ data_len = skb->len;
++ memcpy(skb->data,
++ (void *)(idx_to_kaddr(pending_idx)|txp->offset),
++ data_len);
++ if (data_len < txp->size) {
++ /* Append the packet payload as a fragment. */
++ txp->offset += data_len;
++ txp->size -= data_len;
++ } else {
++ /* Schedule a response immediately. */
++ netif_idx_release(pending_idx);
++ }
+
-+ return (nflags & (GTF_reading|GTF_writing));
-+}
-+EXPORT_SYMBOL_GPL(gnttab_query_foreign_access);
++ /*
++ * Old frontends do not assert data_validated but we
++ * can infer it from csum_blank so test both flags.
++ */
++ if (txp->flags & (NETTXF_data_validated|NETTXF_csum_blank)) {
++ skb->ip_summed = CHECKSUM_UNNECESSARY;
++ skb->proto_data_valid = 1;
++ } else {
++ skb->ip_summed = CHECKSUM_NONE;
++ skb->proto_data_valid = 0;
++ }
++ skb->proto_csum_blank = !!(txp->flags & NETTXF_csum_blank);
+
-+int gnttab_end_foreign_access_ref(grant_ref_t ref, int readonly)
-+{
-+ u16 flags, nflags;
++ netbk_fill_frags(skb);
+
-+ nflags = shared[ref].flags;
-+ do {
-+ if ((flags = nflags) & (GTF_reading|GTF_writing)) {
-+ printk(KERN_DEBUG "WARNING: g.e. still in use!\n");
-+ return 0;
++ skb->dev = netif->dev;
++ skb->protocol = eth_type_trans(skb, skb->dev);
++
++ netif->stats.rx_bytes += skb->len;
++ netif->stats.rx_packets++;
++
++ if (unlikely(netbk_copy_skb_mode == NETBK_ALWAYS_COPY_SKB) &&
++ unlikely(skb_linearize(skb))) {
++ DPRINTK("Can't linearize skb in net_tx_action.\n");
++ kfree_skb(skb);
++ continue;
+ }
-+ } while ((nflags = synch_cmpxchg_subword(&shared[ref].flags, flags, 0)) !=
-+ flags);
+
-+ return 1;
-+}
-+EXPORT_SYMBOL_GPL(gnttab_end_foreign_access_ref);
++ netif_rx(skb);
++ netif->dev->last_rx = jiffies;
++ }
+
-+void gnttab_end_foreign_access(grant_ref_t ref, int readonly,
-+ unsigned long page)
-+{
-+ if (gnttab_end_foreign_access_ref(ref, readonly)) {
-+ put_free_entry(ref);
-+ if (page != 0)
-+ free_page(page);
-+ } else {
-+ /* XXX This needs to be fixed so that the ref and page are
-+ placed on a list to be freed up later. */
-+ printk(KERN_DEBUG
-+ "WARNING: leaking g.e. and page still in use!\n");
++ if (netbk_copy_skb_mode == NETBK_DELAYED_COPY_SKB &&
++ !list_empty(&pending_inuse_head)) {
++ struct netbk_tx_pending_inuse *oldest;
++
++ oldest = list_entry(pending_inuse_head.next,
++ struct netbk_tx_pending_inuse, list);
++ mod_timer(&netbk_tx_pending_timer, oldest->alloc_time + HZ);
+ }
+}
-+EXPORT_SYMBOL_GPL(gnttab_end_foreign_access);
+
-+int gnttab_grant_foreign_transfer(domid_t domid, unsigned long pfn)
++static void netif_idx_release(u16 pending_idx)
+{
-+ int ref;
++ static DEFINE_SPINLOCK(_lock);
++ unsigned long flags;
+
-+ if (unlikely((ref = get_free_entry()) < 0))
-+ return -ENOSPC;
-+ gnttab_grant_foreign_transfer_ref(ref, domid, pfn);
++ spin_lock_irqsave(&_lock, flags);
++ dealloc_ring[MASK_PEND_IDX(dealloc_prod)] = pending_idx;
++ /* Sync with net_tx_action_dealloc: insert idx /then/ incr producer. */
++ smp_wmb();
++ dealloc_prod++;
++ spin_unlock_irqrestore(&_lock, flags);
+
-+ return ref;
++ tasklet_schedule(&net_tx_tasklet);
+}
-+EXPORT_SYMBOL_GPL(gnttab_grant_foreign_transfer);
+
-+void gnttab_grant_foreign_transfer_ref(grant_ref_t ref, domid_t domid,
-+ unsigned long pfn)
++static void netif_page_release(struct page *page)
+{
-+ shared[ref].frame = pfn;
-+ shared[ref].domid = domid;
-+ wmb();
-+ shared[ref].flags = GTF_accept_transfer;
++ netif_idx_release(netif_page_index(page));
+}
-+EXPORT_SYMBOL_GPL(gnttab_grant_foreign_transfer_ref);
+
-+unsigned long gnttab_end_foreign_transfer_ref(grant_ref_t ref)
++irqreturn_t netif_be_int(int irq, void *dev_id, struct pt_regs *regs)
+{
-+ unsigned long frame;
-+ u16 flags;
-+
-+ /*
-+ * If a transfer is not even yet started, try to reclaim the grant
-+ * reference and return failure (== 0).
-+ */
-+ while (!((flags = shared[ref].flags) & GTF_transfer_committed)) {
-+ if (synch_cmpxchg_subword(&shared[ref].flags, flags, 0) == flags)
-+ return 0;
-+ cpu_relax();
-+ }
++ netif_t *netif = dev_id;
+
-+ /* If a transfer is in progress then wait until it is completed. */
-+ while (!(flags & GTF_transfer_completed)) {
-+ flags = shared[ref].flags;
-+ cpu_relax();
-+ }
++ add_to_net_schedule_list_tail(netif);
++ maybe_schedule_tx_action();
+
-+ /* Read the frame number /after/ reading completion status. */
-+ rmb();
-+ frame = shared[ref].frame;
-+ BUG_ON(frame == 0);
++ if (netif_schedulable(netif) && !netbk_queue_full(netif))
++ netif_wake_queue(netif->dev);
+
-+ return frame;
++ return IRQ_HANDLED;
+}
-+EXPORT_SYMBOL_GPL(gnttab_end_foreign_transfer_ref);
+
-+unsigned long gnttab_end_foreign_transfer(grant_ref_t ref)
++static void make_tx_response(netif_t *netif,
++ netif_tx_request_t *txp,
++ s8 st)
+{
-+ unsigned long frame = gnttab_end_foreign_transfer_ref(ref);
-+ put_free_entry(ref);
-+ return frame;
-+}
-+EXPORT_SYMBOL_GPL(gnttab_end_foreign_transfer);
++ RING_IDX i = netif->tx.rsp_prod_pvt;
++ netif_tx_response_t *resp;
++ int notify;
+
-+void gnttab_free_grant_reference(grant_ref_t ref)
-+{
-+ put_free_entry(ref);
-+}
-+EXPORT_SYMBOL_GPL(gnttab_free_grant_reference);
++ resp = RING_GET_RESPONSE(&netif->tx, i);
++ resp->id = txp->id;
++ resp->status = st;
+
-+void gnttab_free_grant_references(grant_ref_t head)
-+{
-+ grant_ref_t ref;
-+ unsigned long flags;
-+ int count = 1;
-+ if (head == GNTTAB_LIST_END)
-+ return;
-+ spin_lock_irqsave(&gnttab_list_lock, flags);
-+ ref = head;
-+ while (gnttab_entry(ref) != GNTTAB_LIST_END) {
-+ ref = gnttab_entry(ref);
-+ count++;
-+ }
-+ gnttab_entry(ref) = gnttab_free_head;
-+ gnttab_free_head = head;
-+ gnttab_free_count += count;
-+ check_free_callbacks();
-+ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++ if (txp->flags & NETTXF_extra_info)
++ RING_GET_RESPONSE(&netif->tx, ++i)->status = NETIF_RSP_NULL;
++
++ netif->tx.rsp_prod_pvt = ++i;
++ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&netif->tx, notify);
++ if (notify)
++ notify_remote_via_irq(netif->irq);
++
++#ifdef CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER
++ if (i == netif->tx.req_cons) {
++ int more_to_do;
++ RING_FINAL_CHECK_FOR_REQUESTS(&netif->tx, more_to_do);
++ if (more_to_do)
++ add_to_net_schedule_list_tail(netif);
++ }
++#endif
+}
-+EXPORT_SYMBOL_GPL(gnttab_free_grant_references);
+
-+int gnttab_alloc_grant_references(u16 count, grant_ref_t *head)
++static netif_rx_response_t *make_rx_response(netif_t *netif,
++ u16 id,
++ s8 st,
++ u16 offset,
++ u16 size,
++ u16 flags)
+{
-+ int h = get_free_entries(count);
++ RING_IDX i = netif->rx.rsp_prod_pvt;
++ netif_rx_response_t *resp;
+
-+ if (h < 0)
-+ return -ENOSPC;
++ resp = RING_GET_RESPONSE(&netif->rx, i);
++ resp->offset = offset;
++ resp->flags = flags;
++ resp->id = id;
++ resp->status = (s16)size;
++ if (st < 0)
++ resp->status = (s16)st;
+
-+ *head = h;
++ netif->rx.rsp_prod_pvt = ++i;
+
-+ return 0;
++ return resp;
+}
-+EXPORT_SYMBOL_GPL(gnttab_alloc_grant_references);
+
-+int gnttab_empty_grant_references(const grant_ref_t *private_head)
++#ifdef NETBE_DEBUG_INTERRUPT
++static irqreturn_t netif_be_dbg(int irq, void *dev_id, struct pt_regs *regs)
+{
-+ return (*private_head == GNTTAB_LIST_END);
-+}
-+EXPORT_SYMBOL_GPL(gnttab_empty_grant_references);
++ struct list_head *ent;
++ netif_t *netif;
++ int i = 0;
+
-+int gnttab_claim_grant_reference(grant_ref_t *private_head)
-+{
-+ grant_ref_t g = *private_head;
-+ if (unlikely(g == GNTTAB_LIST_END))
-+ return -ENOSPC;
-+ *private_head = gnttab_entry(g);
-+ return g;
-+}
-+EXPORT_SYMBOL_GPL(gnttab_claim_grant_reference);
++ printk(KERN_ALERT "netif_schedule_list:\n");
++ spin_lock_irq(&net_schedule_list_lock);
+
-+void gnttab_release_grant_reference(grant_ref_t *private_head,
-+ grant_ref_t release)
-+{
-+ gnttab_entry(release) = *private_head;
-+ *private_head = release;
-+}
-+EXPORT_SYMBOL_GPL(gnttab_release_grant_reference);
++ list_for_each (ent, &net_schedule_list) {
++ netif = list_entry(ent, netif_t, list);
++ printk(KERN_ALERT " %d: private(rx_req_cons=%08x "
++ "rx_resp_prod=%08x\n",
++ i, netif->rx.req_cons, netif->rx.rsp_prod_pvt);
++ printk(KERN_ALERT " tx_req_cons=%08x tx_resp_prod=%08x)\n",
++ netif->tx.req_cons, netif->tx.rsp_prod_pvt);
++ printk(KERN_ALERT " shared(rx_req_prod=%08x "
++ "rx_resp_prod=%08x\n",
++ netif->rx.sring->req_prod, netif->rx.sring->rsp_prod);
++ printk(KERN_ALERT " rx_event=%08x tx_req_prod=%08x\n",
++ netif->rx.sring->rsp_event, netif->tx.sring->req_prod);
++ printk(KERN_ALERT " tx_resp_prod=%08x, tx_event=%08x)\n",
++ netif->tx.sring->rsp_prod, netif->tx.sring->rsp_event);
++ i++;
++ }
+
-+void gnttab_request_free_callback(struct gnttab_free_callback *callback,
-+ void (*fn)(void *), void *arg, u16 count)
-+{
-+ unsigned long flags;
-+ spin_lock_irqsave(&gnttab_list_lock, flags);
-+ if (callback->next)
-+ goto out;
-+ callback->fn = fn;
-+ callback->arg = arg;
-+ callback->count = count;
-+ callback->next = gnttab_free_callback_list;
-+ gnttab_free_callback_list = callback;
-+ check_free_callbacks();
-+out:
-+ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++ spin_unlock_irq(&net_schedule_list_lock);
++ printk(KERN_ALERT " ** End of netif_schedule_list **\n");
++
++ return IRQ_HANDLED;
+}
-+EXPORT_SYMBOL_GPL(gnttab_request_free_callback);
++#endif
+
-+void gnttab_cancel_free_callback(struct gnttab_free_callback *callback)
++static int __init netback_init(void)
+{
-+ struct gnttab_free_callback **pcb;
-+ unsigned long flags;
++ int i;
++ struct page *page;
+
-+ spin_lock_irqsave(&gnttab_list_lock, flags);
-+ for (pcb = &gnttab_free_callback_list; *pcb; pcb = &(*pcb)->next) {
-+ if (*pcb == callback) {
-+ *pcb = callback->next;
-+ break;
-+ }
-+ }
-+ spin_unlock_irqrestore(&gnttab_list_lock, flags);
-+}
-+EXPORT_SYMBOL_GPL(gnttab_cancel_free_callback);
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+static int grow_gnttab_list(unsigned int more_frames)
-+{
-+ unsigned int new_nr_grant_frames, extra_entries, i;
++ /* We can increase reservation by this much in net_rx_action(). */
++ balloon_update_driver_allowance(NET_RX_RING_SIZE);
+
-+ new_nr_grant_frames = nr_grant_frames + more_frames;
-+ extra_entries = more_frames * GREFS_PER_GRANT_FRAME;
++ skb_queue_head_init(&rx_queue);
++ skb_queue_head_init(&tx_queue);
+
-+ for (i = nr_grant_frames; i < new_nr_grant_frames; i++)
-+ {
-+ gnttab_list[i] = (grant_ref_t *)__get_free_page(GFP_ATOMIC);
-+ if (!gnttab_list[i])
-+ goto grow_nomem;
++ init_timer(&net_timer);
++ net_timer.data = 0;
++ net_timer.function = net_alarm;
++
++ init_timer(&netbk_tx_pending_timer);
++ netbk_tx_pending_timer.data = 0;
++ netbk_tx_pending_timer.function = netbk_tx_pending_timeout;
++
++ mmap_pages = alloc_empty_pages_and_pagevec(MAX_PENDING_REQS);
++ if (mmap_pages == NULL) {
++ printk("%s: out of memory\n", __FUNCTION__);
++ return -ENOMEM;
+ }
+
++ for (i = 0; i < MAX_PENDING_REQS; i++) {
++ page = mmap_pages[i];
++ SetPageForeign(page, netif_page_release);
++ netif_page_index(page) = i;
++ INIT_LIST_HEAD(&pending_inuse[i].list);
++ }
+
-+ for (i = GREFS_PER_GRANT_FRAME * nr_grant_frames;
-+ i < GREFS_PER_GRANT_FRAME * new_nr_grant_frames - 1; i++)
-+ gnttab_entry(i) = i + 1;
++ pending_cons = 0;
++ pending_prod = MAX_PENDING_REQS;
++ for (i = 0; i < MAX_PENDING_REQS; i++)
++ pending_ring[i] = i;
+
-+ gnttab_entry(i) = gnttab_free_head;
-+ gnttab_free_head = GREFS_PER_GRANT_FRAME * nr_grant_frames;
-+ gnttab_free_count += extra_entries;
++ spin_lock_init(&net_schedule_list_lock);
++ INIT_LIST_HEAD(&net_schedule_list);
+
-+ nr_grant_frames = new_nr_grant_frames;
++ netbk_copy_skb_mode = NETBK_DONT_COPY_SKB;
++ if (MODPARM_copy_skb) {
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_and_replace,
++ NULL, 0))
++ netbk_copy_skb_mode = NETBK_ALWAYS_COPY_SKB;
++ else
++ netbk_copy_skb_mode = NETBK_DELAYED_COPY_SKB;
++ }
+
-+ check_free_callbacks();
++ netif_accel_init();
++
++ netif_xenbus_init();
++
++#ifdef NETBE_DEBUG_INTERRUPT
++ (void)bind_virq_to_irqhandler(VIRQ_DEBUG,
++ 0,
++ netif_be_dbg,
++ SA_SHIRQ,
++ "net-be-dbg",
++ &netif_be_dbg);
++#endif
+
+ return 0;
-+
-+grow_nomem:
-+ for ( ; i >= nr_grant_frames; i--)
-+ free_page((unsigned long) gnttab_list[i]);
-+ return -ENOMEM;
+}
+
-+static unsigned int __max_nr_grant_frames(void)
-+{
-+ struct gnttab_query_size query;
-+ int rc;
++module_init(netback_init);
+
-+ query.dom = DOMID_SELF;
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/netback/xenbus.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/netback/xenbus.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,454 @@
++/* Xenbus code for netif backend
++ Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
++ Copyright (C) 2005 XenSource Ltd
+
-+ rc = HYPERVISOR_grant_table_op(GNTTABOP_query_size, &query, 1);
-+ if ((rc < 0) || (query.status != GNTST_okay))
-+ return 4; /* Legacy max supported number of frames */
++ This program is free software; you can redistribute it and/or modify
++ it under the terms of the GNU General Public License as published by
++ the Free Software Foundation; either version 2 of the License, or
++ (at your option) any later version.
+
-+ return query.max_nr_frames;
-+}
++ This program is distributed in the hope that it will be useful,
++ but WITHOUT ANY WARRANTY; without even the implied warranty of
++ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ GNU General Public License for more details.
+
-+static inline unsigned int max_nr_grant_frames(void)
-+{
-+ unsigned int xen_max = __max_nr_grant_frames();
++ You should have received a copy of the GNU General Public License
++ along with this program; if not, write to the Free Software
++ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++*/
+
-+ if (xen_max > boot_max_nr_grant_frames)
-+ return boot_max_nr_grant_frames;
-+ return xen_max;
-+}
++#include <stdarg.h>
++#include <linux/module.h>
++#include <xen/xenbus.h>
++#include "common.h"
+
-+#ifdef CONFIG_XEN
++#if 0
++#undef DPRINTK
++#define DPRINTK(fmt, args...) \
++ printk("netback/xenbus (%s:%d) " fmt ".\n", __FUNCTION__, __LINE__, ##args)
++#endif
+
-+#ifndef __ia64__
-+static int map_pte_fn(pte_t *pte, struct page *pmd_page,
-+ unsigned long addr, void *data)
-+{
-+ unsigned long **frames = (unsigned long **)data;
+
-+ set_pte_at(&init_mm, addr, pte, pfn_pte_ma((*frames)[0], PAGE_KERNEL));
-+ (*frames)++;
-+ return 0;
-+}
++static int connect_rings(struct backend_info *);
++static void connect(struct backend_info *);
++static void backend_create_netif(struct backend_info *be);
+
-+static int unmap_pte_fn(pte_t *pte, struct page *pmd_page,
-+ unsigned long addr, void *data)
++static int netback_remove(struct xenbus_device *dev)
+{
++ struct backend_info *be = dev->dev.driver_data;
+
-+ set_pte_at(&init_mm, addr, pte, __pte(0));
++ netback_remove_accelerators(be, dev);
++
++ if (be->netif) {
++ kobject_uevent(&dev->dev.kobj, KOBJ_OFFLINE);
++ netif_disconnect(be->netif);
++ be->netif = NULL;
++ }
++ kfree(be);
++ dev->dev.driver_data = NULL;
+ return 0;
+}
-+#endif
+
-+static int gnttab_map(unsigned int start_idx, unsigned int end_idx)
-+{
-+ struct gnttab_setup_table setup;
-+ unsigned long *frames;
-+ unsigned int nr_gframes = end_idx + 1;
-+ int rc;
+
-+ frames = kmalloc(nr_gframes * sizeof(unsigned long), GFP_ATOMIC);
-+ if (!frames)
++/**
++ * Entry point to this code when a new device is created. Allocate the basic
++ * structures and switch to InitWait.
++ */
++static int netback_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ const char *message;
++ struct xenbus_transaction xbt;
++ int err;
++ int sg;
++ struct backend_info *be = kzalloc(sizeof(struct backend_info),
++ GFP_KERNEL);
++ if (!be) {
++ xenbus_dev_fatal(dev, -ENOMEM,
++ "allocating backend structure");
+ return -ENOMEM;
-+
-+ setup.dom = DOMID_SELF;
-+ setup.nr_frames = nr_gframes;
-+ set_xen_guest_handle(setup.frame_list, frames);
-+
-+ rc = HYPERVISOR_grant_table_op(GNTTABOP_setup_table, &setup, 1);
-+ if (rc == -ENOSYS) {
-+ kfree(frames);
-+ return -ENOSYS;
+ }
+
-+ BUG_ON(rc || setup.status);
++ be->dev = dev;
++ dev->dev.driver_data = be;
+
-+#ifndef __ia64__
-+ if (shared == NULL) {
-+ struct vm_struct *area;
-+ area = alloc_vm_area(PAGE_SIZE * max_nr_grant_frames());
-+ BUG_ON(area == NULL);
-+ shared = area->addr;
-+ }
-+ rc = apply_to_page_range(&init_mm, (unsigned long)shared,
-+ PAGE_SIZE * nr_gframes,
-+ map_pte_fn, &frames);
-+ BUG_ON(rc);
-+ frames -= nr_gframes; /* adjust after map_pte_fn() */
-+#else
-+ shared = __va(frames[0] << PAGE_SHIFT);
-+#endif
++ sg = 1;
++ if (netbk_copy_skb_mode == NETBK_ALWAYS_COPY_SKB)
++ sg = 0;
+
-+ kfree(frames);
++ do {
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "starting transaction");
++ goto fail;
++ }
+
-+ return 0;
-+}
++ err = xenbus_printf(xbt, dev->nodename, "feature-sg", "%d", sg);
++ if (err) {
++ message = "writing feature-sg";
++ goto abort_transaction;
++ }
+
-+int gnttab_resume(void)
-+{
-+ if (max_nr_grant_frames() < nr_grant_frames)
-+ return -ENOSYS;
-+ return gnttab_map(0, nr_grant_frames - 1);
-+}
++ err = xenbus_printf(xbt, dev->nodename, "feature-gso-tcpv4",
++ "%d", sg);
++ if (err) {
++ message = "writing feature-gso-tcpv4";
++ goto abort_transaction;
++ }
+
-+int gnttab_suspend(void)
-+{
-+#ifndef __ia64__
-+ apply_to_page_range(&init_mm, (unsigned long)shared,
-+ PAGE_SIZE * nr_grant_frames,
-+ unmap_pte_fn, NULL);
-+#endif
-+ return 0;
-+}
++ /* We support rx-copy path. */
++ err = xenbus_printf(xbt, dev->nodename,
++ "feature-rx-copy", "%d", 1);
++ if (err) {
++ message = "writing feature-rx-copy";
++ goto abort_transaction;
++ }
+
-+#else /* !CONFIG_XEN */
++ /*
++ * We don't support rx-flip path (except old guests who don't
++ * grok this feature flag).
++ */
++ err = xenbus_printf(xbt, dev->nodename,
++ "feature-rx-flip", "%d", 0);
++ if (err) {
++ message = "writing feature-rx-flip";
++ goto abort_transaction;
++ }
+
-+#include <platform-pci.h>
++ err = xenbus_transaction_end(xbt, 0);
++ } while (err == -EAGAIN);
+
-+static unsigned long resume_frames;
++ if (err) {
++ xenbus_dev_fatal(dev, err, "completing transaction");
++ goto fail;
++ }
+
-+static int gnttab_map(unsigned int start_idx, unsigned int end_idx)
-+{
-+ struct xen_add_to_physmap xatp;
-+ unsigned int i = end_idx;
++ netback_probe_accelerators(be, dev);
+
-+ /* Loop backwards, so that the first hypercall has the largest index,
-+ * ensuring that the table will grow only once.
-+ */
-+ do {
-+ xatp.domid = DOMID_SELF;
-+ xatp.idx = i;
-+ xatp.space = XENMAPSPACE_grant_table;
-+ xatp.gpfn = (resume_frames >> PAGE_SHIFT) + i;
-+ if (HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp))
-+ BUG();
-+ } while (i-- > start_idx);
++ err = xenbus_switch_state(dev, XenbusStateInitWait);
++ if (err)
++ goto fail;
++
++ /* This kicks hotplug scripts, so do it immediately. */
++ backend_create_netif(be);
+
+ return 0;
++
++abort_transaction:
++ xenbus_transaction_end(xbt, 1);
++ xenbus_dev_fatal(dev, err, "%s", message);
++fail:
++ DPRINTK("failed");
++ netback_remove(dev);
++ return err;
+}
+
-+int gnttab_resume(void)
++
++/**
++ * Handle the creation of the hotplug script environment. We add the script
++ * and vif variables to the environment, for the benefit of the vif-* hotplug
++ * scripts.
++ */
++static int netback_uevent(struct xenbus_device *xdev, char **envp,
++ int num_envp, char *buffer, int buffer_size)
+{
-+ unsigned int max_nr_gframes, nr_gframes;
++ struct backend_info *be = xdev->dev.driver_data;
++ netif_t *netif = be->netif;
++ int i = 0, length = 0;
++ char *val;
+
-+ nr_gframes = nr_grant_frames;
-+ max_nr_gframes = max_nr_grant_frames();
-+ if (max_nr_gframes < nr_gframes)
-+ return -ENOSYS;
++ DPRINTK("netback_uevent");
+
-+ if (!resume_frames) {
-+ resume_frames = alloc_xen_mmio(PAGE_SIZE * max_nr_gframes);
-+ shared = ioremap(resume_frames, PAGE_SIZE * max_nr_gframes);
-+ if (shared == NULL) {
-+ printk("error to ioremap gnttab share frames\n");
-+ return -1;
-+ }
++ val = xenbus_read(XBT_NIL, xdev->nodename, "script", NULL);
++ if (IS_ERR(val)) {
++ int err = PTR_ERR(val);
++ xenbus_dev_fatal(xdev, err, "reading script");
++ return err;
++ }
++ else {
++ add_uevent_var(envp, num_envp, &i, buffer, buffer_size,
++ &length, "script=%s", val);
++ kfree(val);
+ }
+
-+ gnttab_map(0, nr_gframes - 1);
++ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
++ "vif=%s", netif->dev->name);
++
++ envp[i] = NULL;
+
+ return 0;
+}
+
-+#endif /* !CONFIG_XEN */
+
-+static int gnttab_expand(unsigned int req_entries)
++static void backend_create_netif(struct backend_info *be)
+{
-+ int rc;
-+ unsigned int cur, extra;
++ int err;
++ long handle;
++ struct xenbus_device *dev = be->dev;
+
-+ cur = nr_grant_frames;
-+ extra = ((req_entries + (GREFS_PER_GRANT_FRAME-1)) /
-+ GREFS_PER_GRANT_FRAME);
-+ if (cur + extra > max_nr_grant_frames())
-+ return -ENOSPC;
++ if (be->netif != NULL)
++ return;
+
-+ if ((rc = gnttab_map(cur, cur + extra - 1)) == 0)
-+ rc = grow_gnttab_list(extra);
++ err = xenbus_scanf(XBT_NIL, dev->nodename, "handle", "%li", &handle);
++ if (err != 1) {
++ xenbus_dev_fatal(dev, err, "reading handle");
++ return;
++ }
+
-+ return rc;
++ be->netif = netif_alloc(dev->otherend_id, handle);
++ if (IS_ERR(be->netif)) {
++ err = PTR_ERR(be->netif);
++ be->netif = NULL;
++ xenbus_dev_fatal(dev, err, "creating interface");
++ return;
++ }
++
++ kobject_uevent(&dev->dev.kobj, KOBJ_ONLINE);
+}
+
-+int __devinit gnttab_init(void)
++
++/**
++ * Callback received when the frontend's state changes.
++ */
++static void frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state)
+{
-+ int i;
-+ unsigned int max_nr_glist_frames;
-+ unsigned int nr_init_grefs;
++ struct backend_info *be = dev->dev.driver_data;
+
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++ DPRINTK("%s", xenbus_strstate(frontend_state));
+
-+ nr_grant_frames = 1;
-+ boot_max_nr_grant_frames = __max_nr_grant_frames();
++ be->frontend_state = frontend_state;
+
-+ /* Determine the maximum number of frames required for the
-+ * grant reference free list on the current hypervisor.
-+ */
-+ max_nr_glist_frames = (boot_max_nr_grant_frames *
-+ GREFS_PER_GRANT_FRAME /
-+ (PAGE_SIZE / sizeof(grant_ref_t)));
++ switch (frontend_state) {
++ case XenbusStateInitialising:
++ if (dev->state == XenbusStateClosed) {
++ printk(KERN_INFO "%s: %s: prepare for reconnect\n",
++ __FUNCTION__, dev->nodename);
++ xenbus_switch_state(dev, XenbusStateInitWait);
++ }
++ break;
+
-+ gnttab_list = kmalloc(max_nr_glist_frames * sizeof(grant_ref_t *),
-+ GFP_KERNEL);
-+ if (gnttab_list == NULL)
-+ return -ENOMEM;
++ case XenbusStateInitialised:
++ break;
+
-+ for (i = 0; i < nr_grant_frames; i++) {
-+ gnttab_list[i] = (grant_ref_t *)__get_free_page(GFP_KERNEL);
-+ if (gnttab_list[i] == NULL)
-+ goto ini_nomem;
++ case XenbusStateConnected:
++ if (dev->state == XenbusStateConnected)
++ break;
++ backend_create_netif(be);
++ if (be->netif)
++ connect(be);
++ break;
++
++ case XenbusStateClosing:
++ if (be->netif) {
++ kobject_uevent(&dev->dev.kobj, KOBJ_OFFLINE);
++ netif_disconnect(be->netif);
++ be->netif = NULL;
++ }
++ xenbus_switch_state(dev, XenbusStateClosing);
++ break;
++
++ case XenbusStateClosed:
++ xenbus_switch_state(dev, XenbusStateClosed);
++ if (xenbus_dev_is_online(dev))
++ break;
++ /* fall through if not online */
++ case XenbusStateUnknown:
++ device_unregister(&dev->dev);
++ break;
++
++ default:
++ xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
++ frontend_state);
++ break;
+ }
++}
++
++
++static void xen_net_read_rate(struct xenbus_device *dev,
++ unsigned long *bytes, unsigned long *usec)
++{
++ char *s, *e;
++ unsigned long b, u;
++ char *ratestr;
++
++ /* Default to unlimited bandwidth. */
++ *bytes = ~0UL;
++ *usec = 0;
+
-+ if (gnttab_resume() < 0)
-+ return -ENODEV;
++ ratestr = xenbus_read(XBT_NIL, dev->nodename, "rate", NULL);
++ if (IS_ERR(ratestr))
++ return;
+
-+ nr_init_grefs = nr_grant_frames * GREFS_PER_GRANT_FRAME;
++ s = ratestr;
++ b = simple_strtoul(s, &e, 10);
++ if ((s == e) || (*e != ','))
++ goto fail;
+
-+ for (i = NR_RESERVED_ENTRIES; i < nr_init_grefs - 1; i++)
-+ gnttab_entry(i) = i + 1;
++ s = e + 1;
++ u = simple_strtoul(s, &e, 10);
++ if ((s == e) || (*e != '\0'))
++ goto fail;
+
-+ gnttab_entry(nr_init_grefs - 1) = GNTTAB_LIST_END;
-+ gnttab_free_count = nr_init_grefs - NR_RESERVED_ENTRIES;
-+ gnttab_free_head = NR_RESERVED_ENTRIES;
++ *bytes = b;
++ *usec = u;
+
-+ return 0;
++ kfree(ratestr);
++ return;
+
-+ ini_nomem:
-+ for (i--; i >= 0; i--)
-+ free_page((unsigned long)gnttab_list[i]);
-+ kfree(gnttab_list);
-+ return -ENOMEM;
++ fail:
++ WPRINTK("Failed to parse network rate limit. Traffic unlimited.\n");
++ kfree(ratestr);
+}
+
-+#ifdef CONFIG_XEN
-+core_initcall(gnttab_init);
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/core/hypervisor_sysfs.c tmp-linux-2.6-xen.patch/drivers/xen/core/hypervisor_sysfs.c
---- pristine-linux-2.6.18.2/drivers/xen/core/hypervisor_sysfs.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/core/hypervisor_sysfs.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,56 @@
-+/*
-+ * copyright (c) 2006 IBM Corporation
-+ * Authored by: Mike D. Day <ncmike at us.ibm.com>
-+ *
-+ * This program is free software; you can redistribute it and/or modify
-+ * it under the terms of the GNU General Public License version 2 as
-+ * published by the Free Software Foundation.
-+ */
++static int xen_net_read_mac(struct xenbus_device *dev, u8 mac[])
++{
++ char *s, *e, *macstr;
++ int i;
+
-+#include <linux/kernel.h>
-+#include <linux/module.h>
-+#include <linux/kobject.h>
-+#include <xen/hypervisor_sysfs.h>
++ macstr = s = xenbus_read(XBT_NIL, dev->nodename, "mac", NULL);
++ if (IS_ERR(macstr))
++ return PTR_ERR(macstr);
+
-+static ssize_t hyp_sysfs_show(struct kobject *kobj,
-+ struct attribute *attr,
-+ char *buffer)
-+{
-+ struct hyp_sysfs_attr *hyp_attr;
-+ hyp_attr = container_of(attr, struct hyp_sysfs_attr, attr);
-+ if (hyp_attr->show)
-+ return hyp_attr->show(hyp_attr, buffer);
++ for (i = 0; i < ETH_ALEN; i++) {
++ mac[i] = simple_strtoul(s, &e, 16);
++ if ((s == e) || (*e != ((i == ETH_ALEN-1) ? '\0' : ':'))) {
++ kfree(macstr);
++ return -ENOENT;
++ }
++ s = e+1;
++ }
++
++ kfree(macstr);
+ return 0;
+}
+
-+static ssize_t hyp_sysfs_store(struct kobject *kobj,
-+ struct attribute *attr,
-+ const char *buffer,
-+ size_t len)
++static void connect(struct backend_info *be)
+{
-+ struct hyp_sysfs_attr *hyp_attr;
-+ hyp_attr = container_of(attr, struct hyp_sysfs_attr, attr);
-+ if (hyp_attr->store)
-+ return hyp_attr->store(hyp_attr, buffer, len);
-+ return 0;
-+}
++ int err;
++ struct xenbus_device *dev = be->dev;
+
-+static struct sysfs_ops hyp_sysfs_ops = {
-+ .show = hyp_sysfs_show,
-+ .store = hyp_sysfs_store,
-+};
++ err = connect_rings(be);
++ if (err)
++ return;
+
-+static struct kobj_type hyp_sysfs_kobj_type = {
-+ .sysfs_ops = &hyp_sysfs_ops,
-+};
++ err = xen_net_read_mac(dev, be->netif->fe_dev_addr);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "parsing %s/mac", dev->nodename);
++ return;
++ }
+
-+static int __init hypervisor_subsys_init(void)
-+{
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++ xen_net_read_rate(dev, &be->netif->credit_bytes,
++ &be->netif->credit_usec);
++ be->netif->remaining_credit = be->netif->credit_bytes;
+
-+ hypervisor_subsys.kset.kobj.ktype = &hyp_sysfs_kobj_type;
-+ return 0;
++ xenbus_switch_state(dev, XenbusStateConnected);
++
++ netif_wake_queue(be->netif->dev);
+}
+
-+device_initcall(hypervisor_subsys_init);
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/core/machine_kexec.c tmp-linux-2.6-xen.patch/drivers/xen/core/machine_kexec.c
---- pristine-linux-2.6.18.2/drivers/xen/core/machine_kexec.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/core/machine_kexec.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,189 @@
-+/*
-+ * drivers/xen/core/machine_kexec.c
-+ * handle transition of Linux booting another kernel
-+ */
+
-+#include <linux/kexec.h>
-+#include <xen/interface/kexec.h>
-+#include <linux/mm.h>
-+#include <linux/bootmem.h>
++static int connect_rings(struct backend_info *be)
++{
++ struct xenbus_device *dev = be->dev;
++ unsigned long tx_ring_ref, rx_ring_ref;
++ unsigned int evtchn, rx_copy;
++ int err;
++ int val;
+
-+extern void machine_kexec_setup_load_arg(xen_kexec_image_t *xki,
-+ struct kimage *image);
++ DPRINTK("");
+
-+static int __initdata xen_max_nr_phys_cpus;
-+static struct resource xen_hypervisor_res;
-+static struct resource *xen_phys_cpus;
++ err = xenbus_gather(XBT_NIL, dev->otherend,
++ "tx-ring-ref", "%lu", &tx_ring_ref,
++ "rx-ring-ref", "%lu", &rx_ring_ref,
++ "event-channel", "%u", &evtchn, NULL);
++ if (err) {
++ xenbus_dev_fatal(dev, err,
++ "reading %s/ring-ref and event-channel",
++ dev->otherend);
++ return err;
++ }
+
-+void __init xen_machine_kexec_setup_resources(void)
-+{
-+ xen_kexec_range_t range;
-+ struct resource *res;
-+ int k = 0;
++ err = xenbus_scanf(XBT_NIL, dev->otherend, "request-rx-copy", "%u",
++ &rx_copy);
++ if (err == -ENOENT) {
++ err = 0;
++ rx_copy = 0;
++ }
++ if (err < 0) {
++ xenbus_dev_fatal(dev, err, "reading %s/request-rx-copy",
++ dev->otherend);
++ return err;
++ }
++ be->netif->copying_receiver = !!rx_copy;
+
-+ if (!is_initial_xendomain())
-+ return;
++ if (be->netif->dev->tx_queue_len != 0) {
++ if (xenbus_scanf(XBT_NIL, dev->otherend,
++ "feature-rx-notify", "%d", &val) < 0)
++ val = 0;
++ if (val)
++ be->netif->can_queue = 1;
++ else
++ /* Must be non-zero for pfifo_fast to work. */
++ be->netif->dev->tx_queue_len = 1;
++ }
+
-+ /* determine maximum number of physical cpus */
++ if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-sg", "%d", &val) < 0)
++ val = 0;
++ if (val) {
++ be->netif->features |= NETIF_F_SG;
++ be->netif->dev->features |= NETIF_F_SG;
++ }
+
-+ while (1) {
-+ memset(&range, 0, sizeof(range));
-+ range.range = KEXEC_RANGE_MA_CPU;
-+ range.nr = k;
++ if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-gso-tcpv4", "%d",
++ &val) < 0)
++ val = 0;
++ if (val) {
++ be->netif->features |= NETIF_F_TSO;
++ be->netif->dev->features |= NETIF_F_TSO;
++ }
+
-+ if(HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
-+ break;
++ if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-no-csum-offload",
++ "%d", &val) < 0)
++ val = 0;
++ if (val) {
++ be->netif->features &= ~NETIF_F_IP_CSUM;
++ be->netif->dev->features &= ~NETIF_F_IP_CSUM;
++ }
+
-+ k++;
++ /* Map the shared frame, irq etc. */
++ err = netif_map(be->netif, tx_ring_ref, rx_ring_ref, evtchn);
++ if (err) {
++ xenbus_dev_fatal(dev, err,
++ "mapping shared-frames %lu/%lu port %u",
++ tx_ring_ref, rx_ring_ref, evtchn);
++ return err;
+ }
++ return 0;
++}
+
-+ if (k == 0)
-+ return;
+
-+ xen_max_nr_phys_cpus = k;
++/* ** Driver Registration ** */
+
-+ /* allocate xen_phys_cpus */
+
-+ xen_phys_cpus = alloc_bootmem_low(k * sizeof(struct resource));
-+ BUG_ON(xen_phys_cpus == NULL);
++static const struct xenbus_device_id netback_ids[] = {
++ { "vif" },
++ { "" }
++};
+
-+ /* fill in xen_phys_cpus with per-cpu crash note information */
+
-+ for (k = 0; k < xen_max_nr_phys_cpus; k++) {
-+ memset(&range, 0, sizeof(range));
-+ range.range = KEXEC_RANGE_MA_CPU;
-+ range.nr = k;
++static struct xenbus_driver netback = {
++ .name = "vif",
++ .owner = THIS_MODULE,
++ .ids = netback_ids,
++ .probe = netback_probe,
++ .remove = netback_remove,
++ .uevent = netback_uevent,
++ .otherend_changed = frontend_changed,
++};
+
-+ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
-+ goto err;
+
-+ res = xen_phys_cpus + k;
++void netif_xenbus_init(void)
++{
++ xenbus_register_backend(&netback);
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/netfront/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/netfront/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,4 @@
+
-+ memset(res, 0, sizeof(*res));
-+ res->name = "Crash note";
-+ res->start = range.start;
-+ res->end = range.start + range.size - 1;
-+ res->flags = IORESOURCE_BUSY | IORESOURCE_MEM;
-+ }
++obj-$(CONFIG_XEN_NETDEV_FRONTEND) := xennet.o
+
-+ /* fill in xen_hypervisor_res with hypervisor machine address range */
++xennet-objs := netfront.o accel.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/netfront/accel.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/netfront/accel.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,824 @@
++/******************************************************************************
++ * Virtual network driver for conversing with remote driver backends.
++ *
++ * Copyright (C) 2007 Solarflare Communications, Inc.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ memset(&range, 0, sizeof(range));
-+ range.range = KEXEC_RANGE_MA_XEN;
++#include <linux/netdevice.h>
++#include <linux/skbuff.h>
++#include <linux/list.h>
++#include <linux/mutex.h>
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
+
-+ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
-+ goto err;
++#include "netfront.h"
+
-+ xen_hypervisor_res.name = "Hypervisor code and data";
-+ xen_hypervisor_res.start = range.start;
-+ xen_hypervisor_res.end = range.start + range.size - 1;
-+ xen_hypervisor_res.flags = IORESOURCE_BUSY | IORESOURCE_MEM;
++#define DPRINTK(fmt, args...) \
++ pr_debug("netfront/accel (%s:%d) " fmt, \
++ __FUNCTION__, __LINE__, ##args)
++#define IPRINTK(fmt, args...) \
++ printk(KERN_INFO "netfront/accel: " fmt, ##args)
++#define WPRINTK(fmt, args...) \
++ printk(KERN_WARNING "netfront/accel: " fmt, ##args)
+
-+ /* fill in crashk_res if range is reserved by hypervisor */
++static int netfront_remove_accelerator(struct netfront_info *np,
++ struct xenbus_device *dev);
++static int netfront_load_accelerator(struct netfront_info *np,
++ struct xenbus_device *dev,
++ const char *frontend);
+
-+ memset(&range, 0, sizeof(range));
-+ range.range = KEXEC_RANGE_MA_CRASH;
++/*
++ * List of all netfront accelerator plugin modules available. Each
++ * list entry is of type struct netfront_accelerator.
++ */
++static struct list_head accelerators_list;
+
-+ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
-+ return;
++/* Lock to protect access to accelerators_list */
++static spinlock_t accelerators_lock;
+
-+ if (range.size) {
-+ crashk_res.start = range.start;
-+ crashk_res.end = range.start + range.size - 1;
-+ }
++/* Workqueue to process acceleration configuration changes */
++struct workqueue_struct *accel_watch_workqueue;
+
-+ return;
++/* Mutex to prevent concurrent loads and suspends, etc. */
++DEFINE_MUTEX(accelerator_mutex);
+
-+ err:
-+ /*
-+ * It isn't possible to free xen_phys_cpus this early in the
-+ * boot. Failure at this stage is unexpected and the amount of
-+ * memory is small therefore we tolerate the potential leak.
-+ */
-+ xen_max_nr_phys_cpus = 0;
-+ return;
++void netif_init_accel(void)
++{
++ INIT_LIST_HEAD(&accelerators_list);
++ spin_lock_init(&accelerators_lock);
++
++ accel_watch_workqueue = create_workqueue("net_accel");
+}
+
-+void __init xen_machine_kexec_register_resources(struct resource *res)
++void netif_exit_accel(void)
+{
-+ int k;
++ struct netfront_accelerator *accelerator, *tmp;
++ unsigned long flags;
+
-+ request_resource(res, &xen_hypervisor_res);
++ flush_workqueue(accel_watch_workqueue);
++ destroy_workqueue(accel_watch_workqueue);
+
-+ for (k = 0; k < xen_max_nr_phys_cpus; k++)
-+ request_resource(&xen_hypervisor_res, xen_phys_cpus + k);
++ spin_lock_irqsave(&accelerators_lock, flags);
+
-+}
++ list_for_each_entry_safe(accelerator, tmp, &accelerators_list, link) {
++ BUG_ON(!list_empty(&accelerator->vif_states));
+
-+static void setup_load_arg(xen_kexec_image_t *xki, struct kimage *image)
-+{
-+ machine_kexec_setup_load_arg(xki, image);
++ list_del(&accelerator->link);
++ kfree(accelerator->frontend);
++ kfree(accelerator);
++ }
+
-+ xki->indirection_page = image->head;
-+ xki->start_address = image->start;
++ spin_unlock_irqrestore(&accelerators_lock, flags);
+}
+
-+/*
-+ * Load the image into xen so xen can kdump itself
-+ * This might have been done in prepare, but prepare
-+ * is currently called too early. It might make sense
-+ * to move prepare, but for now, just add an extra hook.
-+ */
-+int xen_machine_kexec_load(struct kimage *image)
-+{
-+ xen_kexec_load_t xkl;
-+
-+ memset(&xkl, 0, sizeof(xkl));
-+ xkl.type = image->type;
-+ setup_load_arg(&xkl.image, image);
-+ return HYPERVISOR_kexec_op(KEXEC_CMD_kexec_load, &xkl);
-+}
+
-+/*
-+ * Unload the image that was stored by machine_kexec_load()
-+ * This might have been done in machine_kexec_cleanup() but it
-+ * is called too late, and its possible xen could try and kdump
-+ * using resources that have been freed.
++/*
++ * Watch the configured accelerator and change plugin if it's modified
+ */
-+void xen_machine_kexec_unload(struct kimage *image)
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++static void accel_watch_work(struct work_struct *context)
++#else
++static void accel_watch_work(void *context)
++#endif
+{
-+ xen_kexec_load_t xkl;
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++ struct netfront_accel_vif_state *vif_state =
++ container_of(context, struct netfront_accel_vif_state,
++ accel_work);
++#else
++ struct netfront_accel_vif_state *vif_state =
++ (struct netfront_accel_vif_state *)context;
++#endif
++ struct netfront_info *np = vif_state->np;
++ char *accel_frontend;
++ int accel_len, rc = -1;
+
-+ memset(&xkl, 0, sizeof(xkl));
-+ xkl.type = image->type;
-+ HYPERVISOR_kexec_op(KEXEC_CMD_kexec_unload, &xkl);
-+}
++ mutex_lock(&accelerator_mutex);
+
-+/*
-+ * Do not allocate memory (or fail in any way) in machine_kexec().
-+ * We are past the point of no return, committed to rebooting now.
-+ *
-+ * This has the hypervisor move to the prefered reboot CPU,
-+ * stop all CPUs and kexec. That is it combines machine_shutdown()
-+ * and machine_kexec() in Linux kexec terms.
-+ */
-+NORET_TYPE void machine_kexec(struct kimage *image)
-+{
-+ xen_kexec_exec_t xke;
++ accel_frontend = xenbus_read(XBT_NIL, np->xbdev->otherend,
++ "accel-frontend", &accel_len);
++ if (IS_ERR(accel_frontend)) {
++ accel_frontend = NULL;
++ netfront_remove_accelerator(np, np->xbdev);
++ } else {
++ /* If this is the first time, request the accelerator,
++ otherwise only request one if it has changed */
++ if (vif_state->accel_frontend == NULL) {
++ rc = netfront_load_accelerator(np, np->xbdev,
++ accel_frontend);
++ } else {
++ if (strncmp(vif_state->accel_frontend, accel_frontend,
++ accel_len)) {
++ netfront_remove_accelerator(np, np->xbdev);
++ rc = netfront_load_accelerator(np, np->xbdev,
++ accel_frontend);
++ }
++ }
++ }
+
-+ memset(&xke, 0, sizeof(xke));
-+ xke.type = image->type;
-+ HYPERVISOR_kexec_op(KEXEC_CMD_kexec, &xke);
-+ panic("KEXEC_CMD_kexec hypercall should not return\n");
++ /* Get rid of previous state and replace with the new name */
++ if (vif_state->accel_frontend != NULL)
++ kfree(vif_state->accel_frontend);
++ vif_state->accel_frontend = accel_frontend;
++
++ mutex_unlock(&accelerator_mutex);
++
++ if (rc == 0) {
++ DPRINTK("requesting module %s\n", accel_frontend);
++ request_module("%s", accel_frontend);
++ /*
++ * Module should now call netfront_accelerator_loaded() once
++ * it's up and running, and we can continue from there
++ */
++ }
+}
+
-+void machine_shutdown(void)
++
++static void accel_watch_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
+{
-+ /* do nothing */
++ struct netfront_accel_vif_state *vif_state =
++ container_of(watch, struct netfront_accel_vif_state,
++ accel_watch);
++ queue_work(accel_watch_workqueue, &vif_state->accel_work);
+}
+
+
-+/*
-+ * Local variables:
-+ * c-file-style: "linux"
-+ * indent-tabs-mode: t
-+ * c-indent-level: 8
-+ * c-basic-offset: 8
-+ * tab-width: 8
-+ * End:
-+ */
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/core/machine_reboot.c tmp-linux-2.6-xen.patch/drivers/xen/core/machine_reboot.c
---- pristine-linux-2.6.18.2/drivers/xen/core/machine_reboot.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/core/machine_reboot.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,241 @@
-+#include <linux/version.h>
-+#include <linux/kernel.h>
-+#include <linux/mm.h>
-+#include <linux/unistd.h>
-+#include <linux/module.h>
-+#include <linux/reboot.h>
-+#include <linux/sysrq.h>
-+#include <linux/stringify.h>
-+#include <linux/stop_machine.h>
-+#include <asm/irq.h>
-+#include <asm/mmu_context.h>
-+#include <xen/evtchn.h>
-+#include <asm/hypervisor.h>
-+#include <xen/xenbus.h>
-+#include <linux/cpu.h>
-+#include <linux/kthread.h>
-+#include <xen/gnttab.h>
-+#include <xen/xencons.h>
-+#include <xen/cpu_hotplug.h>
-+#include <xen/interface/vcpu.h>
++void netfront_accelerator_add_watch(struct netfront_info *np)
++{
++ int err;
++
++ /* Check we're not trying to overwrite an existing watch */
++ BUG_ON(np->accel_vif_state.accel_watch.node != NULL);
+
-+#if defined(__i386__) || defined(__x86_64__)
++ /* Get a watch on the accelerator plugin */
++ err = xenbus_watch_path2(np->xbdev, np->xbdev->otherend,
++ "accel-frontend",
++ &np->accel_vif_state.accel_watch,
++ accel_watch_changed);
++ if (err) {
++ DPRINTK("%s: Failed to register accel watch: %d\n",
++ __FUNCTION__, err);
++ np->accel_vif_state.accel_watch.node = NULL;
++ }
++}
+
-+/*
-+ * Power off function, if any
-+ */
-+void (*pm_power_off)(void);
-+EXPORT_SYMBOL(pm_power_off);
+
-+void machine_emergency_restart(void)
++static
++void netfront_accelerator_remove_watch(struct netfront_info *np)
+{
-+ /* We really want to get pending console data out before we die. */
-+ xencons_force_flush();
-+ HYPERVISOR_shutdown(SHUTDOWN_reboot);
-+}
++ struct netfront_accel_vif_state *vif_state = &np->accel_vif_state;
+
-+void machine_restart(char * __unused)
-+{
-+ machine_emergency_restart();
++ /* Get rid of watch on accelerator plugin */
++ if (vif_state->accel_watch.node != NULL) {
++ unregister_xenbus_watch(&vif_state->accel_watch);
++ kfree(vif_state->accel_watch.node);
++ vif_state->accel_watch.node = NULL;
++
++ flush_workqueue(accel_watch_workqueue);
++
++ /* Clean up any state left from watch */
++ if (vif_state->accel_frontend != NULL) {
++ kfree(vif_state->accel_frontend);
++ vif_state->accel_frontend = NULL;
++ }
++ }
+}
+
-+void machine_halt(void)
-+{
-+ machine_power_off();
-+}
+
-+void machine_power_off(void)
++/*
++ * Initialise the accel_vif_state field in the netfront state
++ */
++void init_accelerator_vif(struct netfront_info *np,
++ struct xenbus_device *dev)
+{
-+ /* We really want to get pending console data out before we die. */
-+ xencons_force_flush();
-+ if (pm_power_off)
-+ pm_power_off();
-+ HYPERVISOR_shutdown(SHUTDOWN_poweroff);
++ np->accelerator = NULL;
++
++ /* It's assumed that these things don't change */
++ np->accel_vif_state.np = np;
++ np->accel_vif_state.dev = dev;
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++ INIT_WORK(&np->accel_vif_state.accel_work, accel_watch_work);
++#else
++ INIT_WORK(&np->accel_vif_state.accel_work, accel_watch_work,
++ &np->accel_vif_state);
++#endif
+}
+
-+int reboot_thru_bios = 0; /* for dmi_scan.c */
-+EXPORT_SYMBOL(machine_restart);
-+EXPORT_SYMBOL(machine_halt);
-+EXPORT_SYMBOL(machine_power_off);
+
-+static void pre_suspend(void)
++/*
++ * Compare a frontend description string against an accelerator to see
++ * if they match. Would ultimately be nice to replace the string with
++ * a unique numeric identifier for each accelerator.
++ */
++static int match_accelerator(const char *frontend,
++ struct netfront_accelerator *accelerator)
+{
-+ HYPERVISOR_shared_info = (shared_info_t *)empty_zero_page;
-+ HYPERVISOR_update_va_mapping(fix_to_virt(FIX_SHARED_INFO),
-+ __pte_ma(0), 0);
-+
-+ xen_start_info->store_mfn = mfn_to_pfn(xen_start_info->store_mfn);
-+ xen_start_info->console.domU.mfn =
-+ mfn_to_pfn(xen_start_info->console.domU.mfn);
++ return strcmp(frontend, accelerator->frontend) == 0;
+}
+
-+static void post_suspend(int suspend_cancelled)
++
++/*
++ * Add a frontend vif to the list of vifs that is using a netfront
++ * accelerator plugin module.
++ */
++static void add_accelerator_vif(struct netfront_accelerator *accelerator,
++ struct netfront_info *np)
+{
-+ int i, j, k, fpp;
-+ unsigned long shinfo_mfn;
-+ extern unsigned long max_pfn;
-+ extern unsigned long *pfn_to_mfn_frame_list_list;
-+ extern unsigned long *pfn_to_mfn_frame_list[];
++ unsigned long flags;
+
-+ if (suspend_cancelled) {
-+ xen_start_info->store_mfn =
-+ pfn_to_mfn(xen_start_info->store_mfn);
-+ xen_start_info->console.domU.mfn =
-+ pfn_to_mfn(xen_start_info->console.domU.mfn);
++ /* Need lock to write list */
++ spin_lock_irqsave(&accelerator->vif_states_lock, flags);
++
++ if (np->accelerator == NULL) {
++ np->accelerator = accelerator;
++
++ list_add(&np->accel_vif_state.link, &accelerator->vif_states);
+ } else {
-+#ifdef CONFIG_SMP
-+ cpu_initialized_map = cpu_online_map;
-+#endif
++ /*
++ * May get here legitimately if suspend_cancel is
++ * called, but in that case configuration should not
++ * have changed
++ */
++ BUG_ON(np->accelerator != accelerator);
+ }
+
-+ shinfo_mfn = xen_start_info->shared_info >> PAGE_SHIFT;
-+ HYPERVISOR_update_va_mapping(fix_to_virt(FIX_SHARED_INFO),
-+ pfn_pte_ma(shinfo_mfn, PAGE_KERNEL), 0);
-+ HYPERVISOR_shared_info = (shared_info_t *)fix_to_virt(FIX_SHARED_INFO);
++ spin_unlock_irqrestore(&accelerator->vif_states_lock, flags);
++}
+
-+ memset(empty_zero_page, 0, PAGE_SIZE);
+
-+ fpp = PAGE_SIZE/sizeof(unsigned long);
-+ for (i = 0, j = 0, k = -1; i < max_pfn; i += fpp, j++) {
-+ if ((j % fpp) == 0) {
-+ k++;
-+ pfn_to_mfn_frame_list_list[k] =
-+ virt_to_mfn(pfn_to_mfn_frame_list[k]);
-+ j = 0;
-+ }
-+ pfn_to_mfn_frame_list[k][j] =
-+ virt_to_mfn(&phys_to_machine_mapping[i]);
++/*
++ * Initialise the state to track an accelerator plugin module.
++ */
++static int init_accelerator(const char *frontend,
++ struct netfront_accelerator **result,
++ struct netfront_accel_hooks *hooks)
++{
++ struct netfront_accelerator *accelerator =
++ kmalloc(sizeof(struct netfront_accelerator), GFP_KERNEL);
++ unsigned long flags;
++ int frontend_len;
++
++ if (!accelerator) {
++ DPRINTK("no memory for accelerator\n");
++ return -ENOMEM;
+ }
-+ HYPERVISOR_shared_info->arch.max_pfn = max_pfn;
-+ HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list =
-+ virt_to_mfn(pfn_to_mfn_frame_list_list);
-+}
+
-+#else /* !(defined(__i386__) || defined(__x86_64__)) */
++ frontend_len = strlen(frontend) + 1;
++ accelerator->frontend = kmalloc(frontend_len, GFP_KERNEL);
++ if (!accelerator->frontend) {
++ DPRINTK("no memory for accelerator\n");
++ kfree(accelerator);
++ return -ENOMEM;
++ }
++ strlcpy(accelerator->frontend, frontend, frontend_len);
++
++ INIT_LIST_HEAD(&accelerator->vif_states);
++ spin_lock_init(&accelerator->vif_states_lock);
++
++ accelerator->hooks = hooks;
+
-+#ifndef HAVE_XEN_PRE_SUSPEND
-+#define xen_pre_suspend() ((void)0)
-+#endif
++ spin_lock_irqsave(&accelerators_lock, flags);
++ list_add(&accelerator->link, &accelerators_list);
++ spin_unlock_irqrestore(&accelerators_lock, flags);
+
-+#ifndef HAVE_XEN_POST_SUSPEND
-+#define xen_post_suspend(x) ((void)0)
-+#endif
++ *result = accelerator;
+
-+#define switch_idle_mm() ((void)0)
-+#define mm_pin_all() ((void)0)
-+#define pre_suspend() xen_pre_suspend()
-+#define post_suspend(x) xen_post_suspend(x)
++ return 0;
++}
+
-+#endif
+
-+static int take_machine_down(void *p_fast_suspend)
++/*
++ * Modify the hooks stored in the per-vif state to match that in the
++ * netfront accelerator's state.
++ */
++static void
++accelerator_set_vif_state_hooks(struct netfront_accel_vif_state *vif_state)
+{
-+ int fast_suspend = *(int *)p_fast_suspend;
-+ int suspend_cancelled, err;
-+ extern void time_resume(void);
++ /* This function must be called with the vif_states_lock held */
+
-+ if (fast_suspend) {
-+ BUG_ON(!irqs_disabled());
-+ } else {
-+ BUG_ON(irqs_disabled());
++ DPRINTK("%p\n",vif_state);
+
-+ for (;;) {
-+ err = smp_suspend();
-+ if (err)
-+ return err;
++ /* Make sure there are no data path operations going on */
++ netif_poll_disable(vif_state->np->netdev);
++ netif_tx_lock_bh(vif_state->np->netdev);
+
-+ xenbus_suspend();
-+ preempt_disable();
++ vif_state->hooks = vif_state->np->accelerator->hooks;
+
-+ if (num_online_cpus() == 1)
-+ break;
++ netif_tx_unlock_bh(vif_state->np->netdev);
++ netif_poll_enable(vif_state->np->netdev);
++}
+
-+ preempt_enable();
-+ xenbus_suspend_cancel();
-+ }
+
-+ local_irq_disable();
++static void accelerator_probe_new_vif(struct netfront_info *np,
++ struct xenbus_device *dev,
++ struct netfront_accelerator *accelerator)
++{
++ struct netfront_accel_hooks *hooks;
++ unsigned long flags;
++
++ DPRINTK("\n");
++
++ /* Include this frontend device on the accelerator's list */
++ add_accelerator_vif(accelerator, np);
++
++ hooks = accelerator->hooks;
++
++ if (hooks) {
++ if (hooks->new_device(np->netdev, dev) == 0) {
++ spin_lock_irqsave
++ (&accelerator->vif_states_lock, flags);
++
++ accelerator_set_vif_state_hooks(&np->accel_vif_state);
++
++ spin_unlock_irqrestore
++ (&accelerator->vif_states_lock, flags);
++ }
+ }
+
-+ mm_pin_all();
-+ gnttab_suspend();
-+ pre_suspend();
++ return;
++}
+
-+ /*
-+ * This hypercall returns 1 if suspend was cancelled or the domain was
-+ * merely checkpointed, and 0 if it is resuming in a new domain.
-+ */
-+ suspend_cancelled = HYPERVISOR_suspend(virt_to_mfn(xen_start_info));
+
-+ post_suspend(suspend_cancelled);
-+ gnttab_resume();
-+ if (!suspend_cancelled) {
-+ irq_resume();
-+#ifdef __x86_64__
-+ /*
-+ * Older versions of Xen do not save/restore the user %cr3.
-+ * We do it here just in case, but there's no need if we are
-+ * in fast-suspend mode as that implies a new enough Xen.
-+ */
-+ if (!fast_suspend) {
-+ struct mmuext_op op;
-+ op.cmd = MMUEXT_NEW_USER_BASEPTR;
-+ op.arg1.mfn = pfn_to_mfn(__pa(__user_pgd(
-+ current->active_mm->pgd)) >> PAGE_SHIFT);
-+ if (HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF))
-+ BUG();
++/*
++ * Request that a particular netfront accelerator plugin is loaded.
++ * Usually called as a result of the vif configuration specifying
++ * which one to use. Must be called with accelerator_mutex held
++ */
++static int netfront_load_accelerator(struct netfront_info *np,
++ struct xenbus_device *dev,
++ const char *frontend)
++{
++ struct netfront_accelerator *accelerator;
++ int rc = 0;
++
++ DPRINTK(" %s\n", frontend);
++
++ /*
++ * Look at list of loaded accelerators to see if the requested
++ * one is already there
++ */
++ list_for_each_entry(accelerator, &accelerators_list, link) {
++ if (match_accelerator(frontend, accelerator)) {
++ accelerator_probe_new_vif(np, dev, accelerator);
++ return 0;
+ }
-+#endif
+ }
-+ time_resume();
+
-+ if (!fast_suspend)
-+ local_irq_enable();
++ /* Couldn't find it, so create a new one and load the module */
++ if ((rc = init_accelerator(frontend, &accelerator, NULL)) < 0) {
++ return rc;
++ }
+
-+ return suspend_cancelled;
++ /* Include this frontend device on the accelerator's list */
++ add_accelerator_vif(accelerator, np);
++
++ return rc;
+}
+
-+int __xen_suspend(int fast_suspend)
++
++/*
++ * Go through all the netfront vifs and see if they have requested
++ * this accelerator. Notify the accelerator plugin of the relevant
++ * device if so. Called when an accelerator plugin module is first
++ * loaded and connects to netfront.
++ */
++static void
++accelerator_probe_vifs(struct netfront_accelerator *accelerator,
++ struct netfront_accel_hooks *hooks)
+{
-+ int err, suspend_cancelled;
++ struct netfront_accel_vif_state *vif_state, *tmp;
++ unsigned long flags;
+
-+ BUG_ON(smp_processor_id() != 0);
-+ BUG_ON(in_interrupt());
++ DPRINTK("%p\n", accelerator);
+
-+#if defined(__i386__) || defined(__x86_64__)
-+ if (xen_feature(XENFEAT_auto_translated_physmap)) {
-+ printk(KERN_WARNING "Cannot suspend in "
-+ "auto_translated_physmap mode.\n");
-+ return -EOPNOTSUPP;
++ /*
++ * Store the hooks for future calls to probe a new device, and
++ * to wire into the vif_state once the accelerator plugin is
++ * ready to accelerate each vif
++ */
++ BUG_ON(hooks == NULL);
++ accelerator->hooks = hooks;
++
++ /*
++ * currently hold accelerator_mutex, so don't need
++ * vif_states_lock to read the list
++ */
++ list_for_each_entry_safe(vif_state, tmp, &accelerator->vif_states,
++ link) {
++ struct netfront_info *np = vif_state->np;
++
++ if (hooks->new_device(np->netdev, vif_state->dev) == 0) {
++ spin_lock_irqsave
++ (&accelerator->vif_states_lock, flags);
++
++ accelerator_set_vif_state_hooks(vif_state);
++
++ spin_unlock_irqrestore
++ (&accelerator->vif_states_lock, flags);
++ }
+ }
-+#endif
++}
+
-+ /* If we are definitely UP then 'slow mode' is actually faster. */
-+ if (num_possible_cpus() == 1)
-+ fast_suspend = 0;
+
-+ if (fast_suspend) {
-+ xenbus_suspend();
-+ err = stop_machine_run(take_machine_down, &fast_suspend, 0);
-+ if (err < 0)
-+ xenbus_suspend_cancel();
-+ } else {
-+ err = take_machine_down(&fast_suspend);
++/*
++ * Called by the netfront accelerator plugin module when it has loaded
++ */
++int netfront_accelerator_loaded(int version, const char *frontend,
++ struct netfront_accel_hooks *hooks)
++{
++ struct netfront_accelerator *accelerator;
++
++ if (is_initial_xendomain())
++ return -EINVAL;
++
++ if (version != NETFRONT_ACCEL_VERSION) {
++ if (version > NETFRONT_ACCEL_VERSION) {
++ /* Caller has higher version number, leave it
++ up to them to decide whether to continue.
++ They can re-call with a lower number if
++ they're happy to be compatible with us */
++ return NETFRONT_ACCEL_VERSION;
++ } else {
++ /* We have a more recent version than caller.
++ Currently reject, but may in future be able
++ to be backwardly compatible */
++ return -EPROTO;
++ }
+ }
+
-+ if (err < 0)
-+ return err;
++ mutex_lock(&accelerator_mutex);
+
-+ suspend_cancelled = err;
-+ if (!suspend_cancelled) {
-+ xencons_resume();
-+ xenbus_resume();
-+ } else {
-+ xenbus_suspend_cancel();
++ /*
++ * Look through list of accelerators to see if it has already
++ * been requested
++ */
++ list_for_each_entry(accelerator, &accelerators_list, link) {
++ if (match_accelerator(frontend, accelerator)) {
++ accelerator_probe_vifs(accelerator, hooks);
++ goto out;
++ }
+ }
+
-+ if (!fast_suspend)
-+ smp_resume();
++ /*
++ * If it wasn't in the list, add it now so that when it is
++ * requested the caller will find it
++ */
++ DPRINTK("Couldn't find matching accelerator (%s)\n",
++ frontend);
+
++ init_accelerator(frontend, &accelerator, hooks);
++
++ out:
++ mutex_unlock(&accelerator_mutex);
+ return 0;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/core/Makefile tmp-linux-2.6-xen.patch/drivers/xen/core/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/core/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/core/Makefile 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,12 @@
-+#
-+# Makefile for the linux kernel.
-+#
++EXPORT_SYMBOL_GPL(netfront_accelerator_loaded);
+
-+obj-y := evtchn.o gnttab.o features.o reboot.o machine_reboot.o
+
-+obj-$(CONFIG_PROC_FS) += xen_proc.o
-+obj-$(CONFIG_SYS_HYPERVISOR) += hypervisor_sysfs.o
-+obj-$(CONFIG_HOTPLUG_CPU) += cpu_hotplug.o
-+obj-$(CONFIG_XEN_SYSFS) += xen_sysfs.o
-+obj-$(CONFIG_XEN_SMPBOOT) += smpboot.o
-+obj-$(CONFIG_KEXEC) += machine_kexec.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/core/reboot.c tmp-linux-2.6-xen.patch/drivers/xen/core/reboot.c
---- pristine-linux-2.6.18.2/drivers/xen/core/reboot.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/core/reboot.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,243 @@
-+#define __KERNEL_SYSCALLS__
-+#include <linux/version.h>
-+#include <linux/kernel.h>
-+#include <linux/unistd.h>
-+#include <linux/module.h>
-+#include <linux/reboot.h>
-+#include <linux/sysrq.h>
-+#include <asm/hypervisor.h>
-+#include <xen/xenbus.h>
++/*
++ * Remove the hooks from a single vif state.
++ */
++static void
++accelerator_remove_single_hook(struct netfront_accelerator *accelerator,
++ struct netfront_accel_vif_state *vif_state)
++{
++ /* Make sure there are no data path operations going on */
++ netif_poll_disable(vif_state->np->netdev);
++ netif_tx_lock_bh(vif_state->np->netdev);
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
-+#endif
++ /*
++ * Remove the hooks, but leave the vif_state on the
++ * accelerator's list as that signifies this vif is
++ * interested in using that accelerator if it becomes
++ * available again
++ */
++ vif_state->hooks = NULL;
++
++ netif_tx_unlock_bh(vif_state->np->netdev);
++ netif_poll_enable(vif_state->np->netdev);
++}
+
-+MODULE_LICENSE("Dual BSD/GPL");
+
-+#define SHUTDOWN_INVALID -1
-+#define SHUTDOWN_POWEROFF 0
-+#define SHUTDOWN_SUSPEND 2
-+/* Code 3 is SHUTDOWN_CRASH, which we don't use because the domain can only
-+ * report a crash, not be instructed to crash!
-+ * HALT is the same as POWEROFF, as far as we're concerned. The tools use
-+ * the distinction when we return the reason code to them.
++/*
++ * Safely remove the accelerator function hooks from a netfront state.
+ */
-+#define SHUTDOWN_HALT 4
-+
-+/* Ignore multiple shutdown requests. */
-+static int shutting_down = SHUTDOWN_INVALID;
-+
-+/* Can we leave APs online when we suspend? */
-+static int fast_suspend;
++static void accelerator_remove_hooks(struct netfront_accelerator *accelerator)
++{
++ struct netfront_accel_hooks *hooks;
++ struct netfront_accel_vif_state *vif_state, *tmp;
++ unsigned long flags;
+
-+static void __shutdown_handler(void *unused);
-+static DECLARE_WORK(shutdown_work, __shutdown_handler, NULL);
++ /* Mutex is held so don't need vif_states_lock to iterate list */
++ list_for_each_entry_safe(vif_state, tmp,
++ &accelerator->vif_states,
++ link) {
++ spin_lock_irqsave(&accelerator->vif_states_lock, flags);
+
-+int __xen_suspend(int fast_suspend);
++ if(vif_state->hooks) {
++ hooks = vif_state->hooks;
++
++ /* Last chance to get statistics from the accelerator */
++ hooks->get_stats(vif_state->np->netdev,
++ &vif_state->np->stats);
+
-+static int shutdown_process(void *__unused)
-+{
-+ static char *envp[] = { "HOME=/", "TERM=linux",
-+ "PATH=/sbin:/usr/sbin:/bin:/usr/bin", NULL };
-+ static char *poweroff_argv[] = { "/sbin/poweroff", NULL };
++ spin_unlock_irqrestore(&accelerator->vif_states_lock,
++ flags);
+
-+ extern asmlinkage long sys_reboot(int magic1, int magic2,
-+ unsigned int cmd, void *arg);
++ accelerator_remove_single_hook(accelerator, vif_state);
+
-+ if ((shutting_down == SHUTDOWN_POWEROFF) ||
-+ (shutting_down == SHUTDOWN_HALT)) {
-+ if (call_usermodehelper("/sbin/poweroff", poweroff_argv,
-+ envp, 0) < 0) {
-+#ifdef CONFIG_XEN
-+ sys_reboot(LINUX_REBOOT_MAGIC1,
-+ LINUX_REBOOT_MAGIC2,
-+ LINUX_REBOOT_CMD_POWER_OFF,
-+ NULL);
-+#endif /* CONFIG_XEN */
++ accelerator->hooks->remove(vif_state->dev);
++ } else {
++ spin_unlock_irqrestore(&accelerator->vif_states_lock,
++ flags);
+ }
+ }
-+
-+ shutting_down = SHUTDOWN_INVALID; /* could try again */
-+
-+ return 0;
++
++ accelerator->hooks = NULL;
+}
+
-+static int xen_suspend(void *__unused)
++
++/*
++ * Called by a netfront accelerator when it is unloaded. This safely
++ * removes the hooks into the plugin and blocks until all devices have
++ * finished using it, so on return it is safe to unload.
++ */
++void netfront_accelerator_stop(const char *frontend)
+{
-+ int err;
++ struct netfront_accelerator *accelerator;
++ unsigned long flags;
+
-+ daemonize("suspend");
-+ err = set_cpus_allowed(current, cpumask_of_cpu(0));
-+ if (err) {
-+ printk(KERN_ERR "Xen suspend can't run on CPU0 (%d)\n", err);
-+ goto out;
-+ }
++ mutex_lock(&accelerator_mutex);
++ spin_lock_irqsave(&accelerators_lock, flags);
+
-+ err = __xen_suspend(fast_suspend);
-+ if (err)
-+ printk(KERN_ERR "Xen suspend failed (%d)\n", err);
++ list_for_each_entry(accelerator, &accelerators_list, link) {
++ if (match_accelerator(frontend, accelerator)) {
++ spin_unlock_irqrestore(&accelerators_lock, flags);
++
++ accelerator_remove_hooks(accelerator);
+
++ goto out;
++ }
++ }
++ spin_unlock_irqrestore(&accelerators_lock, flags);
+ out:
-+ shutting_down = SHUTDOWN_INVALID;
-+ return 0;
++ mutex_unlock(&accelerator_mutex);
+}
++EXPORT_SYMBOL_GPL(netfront_accelerator_stop);
+
-+static void __shutdown_handler(void *unused)
++
++/* Helper for call_remove and do_suspend */
++static int do_remove(struct netfront_info *np, struct xenbus_device *dev,
++ unsigned long *lock_flags)
+{
-+ int err;
++ struct netfront_accelerator *accelerator = np->accelerator;
++ struct netfront_accel_hooks *hooks;
++ int rc = 0;
++
++ if (np->accel_vif_state.hooks) {
++ hooks = np->accel_vif_state.hooks;
+
-+ err = kernel_thread((shutting_down == SHUTDOWN_SUSPEND) ?
-+ xen_suspend : shutdown_process,
-+ NULL, CLONE_FS | CLONE_FILES);
++ /* Last chance to get statistics from the accelerator */
++ hooks->get_stats(np->netdev, &np->stats);
+
-+ if (err < 0) {
-+ printk(KERN_WARNING "Error creating shutdown process (%d): "
-+ "retrying...\n", -err);
-+ schedule_delayed_work(&shutdown_work, HZ/2);
++ spin_unlock_irqrestore(&accelerator->vif_states_lock,
++ *lock_flags);
++
++ /*
++ * Try and do the opposite of accelerator_probe_new_vif
++ * to ensure there's no state pointing back at the
++ * netdev
++ */
++ accelerator_remove_single_hook(accelerator,
++ &np->accel_vif_state);
++
++ rc = accelerator->hooks->remove(dev);
++
++ spin_lock_irqsave(&accelerator->vif_states_lock, *lock_flags);
+ }
++
++ return rc;
+}
+
-+static void shutdown_handler(struct xenbus_watch *watch,
-+ const char **vec, unsigned int len)
++
++static int netfront_remove_accelerator(struct netfront_info *np,
++ struct xenbus_device *dev)
+{
-+ extern void ctrl_alt_del(void);
-+ char *str;
-+ struct xenbus_transaction xbt;
-+ int err;
++ struct netfront_accelerator *accelerator;
++ struct netfront_accel_vif_state *tmp_vif_state;
++ unsigned long flags;
++ int rc = 0;
+
-+ if (shutting_down != SHUTDOWN_INVALID)
-+ return;
++ /* Check that we've got a device that was accelerated */
++ if (np->accelerator == NULL)
++ return rc;
+
-+ again:
-+ err = xenbus_transaction_start(&xbt);
-+ if (err)
-+ return;
++ accelerator = np->accelerator;
+
-+ str = (char *)xenbus_read(xbt, "control", "shutdown", NULL);
-+ /* Ignore read errors and empty reads. */
-+ if (XENBUS_IS_ERR_READ(str)) {
-+ xenbus_transaction_end(xbt, 1);
-+ return;
++ spin_lock_irqsave(&accelerator->vif_states_lock, flags);
++
++ list_for_each_entry(tmp_vif_state, &accelerator->vif_states,
++ link) {
++ if (tmp_vif_state == &np->accel_vif_state) {
++ list_del(&np->accel_vif_state.link);
++ break;
++ }
+ }
+
-+ xenbus_write(xbt, "control", "shutdown", "");
++ rc = do_remove(np, dev, &flags);
+
-+ err = xenbus_transaction_end(xbt, 0);
-+ if (err == -EAGAIN) {
-+ kfree(str);
-+ goto again;
-+ }
++ np->accelerator = NULL;
+
-+ if (strcmp(str, "poweroff") == 0)
-+ shutting_down = SHUTDOWN_POWEROFF;
-+ else if (strcmp(str, "reboot") == 0)
-+ ctrl_alt_del();
-+ else if (strcmp(str, "suspend") == 0)
-+ shutting_down = SHUTDOWN_SUSPEND;
-+ else if (strcmp(str, "halt") == 0)
-+ shutting_down = SHUTDOWN_HALT;
-+ else {
-+ printk("Ignoring shutdown request: %s\n", str);
-+ shutting_down = SHUTDOWN_INVALID;
-+ }
++ spin_unlock_irqrestore(&accelerator->vif_states_lock, flags);
+
-+ if (shutting_down != SHUTDOWN_INVALID)
-+ schedule_work(&shutdown_work);
++ return rc;
++}
+
-+ kfree(str);
++
++int netfront_accelerator_call_remove(struct netfront_info *np,
++ struct xenbus_device *dev)
++{
++ int rc;
++ netfront_accelerator_remove_watch(np);
++ mutex_lock(&accelerator_mutex);
++ rc = netfront_remove_accelerator(np, dev);
++ mutex_unlock(&accelerator_mutex);
++ return rc;
+}
+
-+static void sysrq_handler(struct xenbus_watch *watch, const char **vec,
-+ unsigned int len)
++
++int netfront_accelerator_suspend(struct netfront_info *np,
++ struct xenbus_device *dev)
+{
-+ char sysrq_key = '\0';
-+ struct xenbus_transaction xbt;
-+ int err;
++ unsigned long flags;
++ int rc = 0;
++
++ netfront_accelerator_remove_watch(np);
+
-+ again:
-+ err = xenbus_transaction_start(&xbt);
-+ if (err)
-+ return;
-+ if (!xenbus_scanf(xbt, "control", "sysrq", "%c", &sysrq_key)) {
-+ printk(KERN_ERR "Unable to read sysrq code in "
-+ "control/sysrq\n");
-+ xenbus_transaction_end(xbt, 1);
-+ return;
-+ }
++ mutex_lock(&accelerator_mutex);
+
-+ if (sysrq_key != '\0')
-+ xenbus_printf(xbt, "control", "sysrq", "%c", '\0');
++ /* Check that we've got a device that was accelerated */
++ if (np->accelerator == NULL)
++ goto out;
+
-+ err = xenbus_transaction_end(xbt, 0);
-+ if (err == -EAGAIN)
-+ goto again;
++ /*
++ * Call the remove accelerator hook, but leave the vif_state
++ * on the accelerator's list in case there is a suspend_cancel.
++ */
++ spin_lock_irqsave(&np->accelerator->vif_states_lock, flags);
++
++ rc = do_remove(np, dev, &flags);
+
-+#ifdef CONFIG_MAGIC_SYSRQ
-+ if (sysrq_key != '\0')
-+ handle_sysrq(sysrq_key, NULL, NULL);
-+#endif
++ spin_unlock_irqrestore(&np->accelerator->vif_states_lock, flags);
++ out:
++ mutex_unlock(&accelerator_mutex);
++ return rc;
++}
++
++
++int netfront_accelerator_suspend_cancel(struct netfront_info *np,
++ struct xenbus_device *dev)
++{
++ /*
++ * Setting the watch will cause it to fire and probe the
++ * accelerator, so no need to call accelerator_probe_new_vif()
++ * directly here
++ */
++ if (dev->state == XenbusStateConnected)
++ netfront_accelerator_add_watch(np);
++ return 0;
+}
++
++
++void netfront_accelerator_resume(struct netfront_info *np,
++ struct xenbus_device *dev)
++{
++ struct netfront_accel_vif_state *accel_vif_state = NULL;
++ spinlock_t *vif_states_lock;
++ unsigned long flags;
++
++ mutex_lock(&accelerator_mutex);
+
-+static struct xenbus_watch shutdown_watch = {
-+ .node = "control/shutdown",
-+ .callback = shutdown_handler
-+};
++ /* Check that we've got a device that was accelerated */
++ if(np->accelerator == NULL)
++ goto out;
+
-+static struct xenbus_watch sysrq_watch = {
-+ .node = "control/sysrq",
-+ .callback = sysrq_handler
-+};
++ /* Find the vif_state from the accelerator's list */
++ list_for_each_entry(accel_vif_state, &np->accelerator->vif_states,
++ link) {
++ if (accel_vif_state->dev == dev) {
++ BUG_ON(accel_vif_state != &np->accel_vif_state);
++
++ vif_states_lock = &np->accelerator->vif_states_lock;
++ spin_lock_irqsave(vif_states_lock, flags);
++
++ /*
++ * Remove it from the accelerator's list so
++ * state is consistent for probing new vifs
++ * when they get connected
++ */
++ list_del(&accel_vif_state->link);
++ np->accelerator = NULL;
++
++ spin_unlock_irqrestore(vif_states_lock, flags);
++
++ break;
++ }
++ }
+
-+static int setup_shutdown_watcher(void)
-+{
-+ int err;
++ out:
++ mutex_unlock(&accelerator_mutex);
++ return;
++}
+
-+ xenbus_scanf(XBT_NIL, "control",
-+ "platform-feature-multiprocessor-suspend",
-+ "%d", &fast_suspend);
+
-+ err = register_xenbus_watch(&shutdown_watch);
-+ if (err) {
-+ printk(KERN_ERR "Failed to set shutdown watcher\n");
-+ return err;
-+ }
++int netfront_check_accelerator_queue_ready(struct net_device *dev,
++ struct netfront_info *np)
++{
++ struct netfront_accelerator *accelerator;
++ struct netfront_accel_hooks *hooks;
++ int rc = 1;
++ unsigned long flags;
+
-+ err = register_xenbus_watch(&sysrq_watch);
-+ if (err) {
-+ printk(KERN_ERR "Failed to set sysrq watcher\n");
-+ return err;
++ accelerator = np->accelerator;
++
++ /* Call the check_ready accelerator hook. */
++ if (np->accel_vif_state.hooks && accelerator) {
++ spin_lock_irqsave(&accelerator->vif_states_lock, flags);
++ hooks = np->accel_vif_state.hooks;
++ if (hooks && np->accelerator == accelerator)
++ rc = np->accel_vif_state.hooks->check_ready(dev);
++ spin_unlock_irqrestore(&accelerator->vif_states_lock, flags);
+ }
+
-+ return 0;
++ return rc;
+}
+
-+#ifdef CONFIG_XEN
+
-+static int shutdown_event(struct notifier_block *notifier,
-+ unsigned long event,
-+ void *data)
++void netfront_accelerator_call_stop_napi_irq(struct netfront_info *np,
++ struct net_device *dev)
+{
-+ setup_shutdown_watcher();
-+ return NOTIFY_DONE;
-+}
++ struct netfront_accelerator *accelerator;
++ struct netfront_accel_hooks *hooks;
++ unsigned long flags;
+
-+static int __init setup_shutdown_event(void)
-+{
-+ static struct notifier_block xenstore_notifier = {
-+ .notifier_call = shutdown_event
-+ };
-+ register_xenstore_notifier(&xenstore_notifier);
++ accelerator = np->accelerator;
+
-+ return 0;
++ /* Call the stop_napi_interrupts accelerator hook. */
++ if (np->accel_vif_state.hooks && accelerator != NULL) {
++ spin_lock_irqsave(&accelerator->vif_states_lock, flags);
++ hooks = np->accel_vif_state.hooks;
++ if (hooks && np->accelerator == accelerator)
++ np->accel_vif_state.hooks->stop_napi_irq(dev);
++ spin_unlock_irqrestore(&accelerator->vif_states_lock, flags);
++ }
+}
+
-+subsys_initcall(setup_shutdown_event);
-+
-+#else /* !defined(CONFIG_XEN) */
+
-+int xen_reboot_init(void)
++int netfront_accelerator_call_get_stats(struct netfront_info *np,
++ struct net_device *dev)
+{
-+ return setup_shutdown_watcher();
++ struct netfront_accelerator *accelerator;
++ struct netfront_accel_hooks *hooks;
++ unsigned long flags;
++ int rc = 0;
++
++ accelerator = np->accelerator;
++
++ /* Call the get_stats accelerator hook. */
++ if (np->accel_vif_state.hooks && accelerator != NULL) {
++ spin_lock_irqsave(&accelerator->vif_states_lock, flags);
++ hooks = np->accel_vif_state.hooks;
++ if (hooks && np->accelerator == accelerator)
++ rc = np->accel_vif_state.hooks->get_stats(dev,
++ &np->stats);
++ spin_unlock_irqrestore(&accelerator->vif_states_lock, flags);
++ }
++ return rc;
+}
+
-+#endif /* !defined(CONFIG_XEN) */
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/core/smpboot.c tmp-linux-2.6-xen.patch/drivers/xen/core/smpboot.c
---- pristine-linux-2.6.18.2/drivers/xen/core/smpboot.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/core/smpboot.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,452 @@
-+/*
-+ * Xen SMP booting functions
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/netfront/netfront.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/netfront/netfront.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,2240 @@
++/******************************************************************************
++ * Virtual network driver for conversing with remote driver backends.
+ *
-+ * See arch/i386/kernel/smpboot.c for copyright and credits for derived
-+ * portions of this file.
++ * Copyright (c) 2002-2005, K A Fraser
++ * Copyright (c) 2005, XenSource Ltd
++ * Copyright (C) 2007 Solarflare Communications, Inc.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
+
+#include <linux/module.h>
-+#include <linux/init.h>
++#include <linux/version.h>
+#include <linux/kernel.h>
-+#include <linux/mm.h>
+#include <linux/sched.h>
-+#include <linux/kernel_stat.h>
-+#include <linux/smp_lock.h>
-+#include <linux/irq.h>
-+#include <linux/bootmem.h>
-+#include <linux/notifier.h>
-+#include <linux/cpu.h>
-+#include <linux/percpu.h>
-+#include <asm/desc.h>
-+#include <asm/arch_hooks.h>
-+#include <asm/pgalloc.h>
++#include <linux/slab.h>
++#include <linux/string.h>
++#include <linux/errno.h>
++#include <linux/netdevice.h>
++#include <linux/inetdevice.h>
++#include <linux/etherdevice.h>
++#include <linux/skbuff.h>
++#include <linux/init.h>
++#include <linux/bitops.h>
++#include <linux/ethtool.h>
++#include <linux/in.h>
++#include <linux/if_ether.h>
++#include <linux/io.h>
++#include <linux/moduleparam.h>
++#include <net/sock.h>
++#include <net/pkt_sched.h>
++#include <net/arp.h>
++#include <net/route.h>
++#include <asm/uaccess.h>
+#include <xen/evtchn.h>
-+#include <xen/interface/vcpu.h>
-+#include <xen/cpu_hotplug.h>
+#include <xen/xenbus.h>
++#include <xen/interface/io/netif.h>
++#include <xen/interface/memory.h>
++#include <xen/balloon.h>
++#include <asm/page.h>
++#include <asm/maddr.h>
++#include <asm/uaccess.h>
++#include <xen/interface/grant_table.h>
++#include <xen/gnttab.h>
+
-+extern irqreturn_t smp_reschedule_interrupt(int, void *, struct pt_regs *);
-+extern irqreturn_t smp_call_function_interrupt(int, void *, struct pt_regs *);
++struct netfront_cb {
++ struct page *page;
++ unsigned offset;
++};
+
-+extern int local_setup_timer(unsigned int cpu);
-+extern void local_teardown_timer(unsigned int cpu);
++#define NETFRONT_SKB_CB(skb) ((struct netfront_cb *)((skb)->cb))
+
-+extern void hypervisor_callback(void);
-+extern void failsafe_callback(void);
-+extern void system_call(void);
-+extern void smp_trap_init(trap_info_t *);
++#include "netfront.h"
+
-+/* Number of siblings per CPU package */
-+int smp_num_siblings = 1;
-+int phys_proc_id[NR_CPUS]; /* Package ID of each logical CPU */
-+EXPORT_SYMBOL(phys_proc_id);
-+int cpu_core_id[NR_CPUS]; /* Core ID of each logical CPU */
-+EXPORT_SYMBOL(cpu_core_id);
++/*
++ * Mutually-exclusive module options to select receive data path:
++ * rx_copy : Packets are copied by network backend into local memory
++ * rx_flip : Page containing packet data is transferred to our ownership
++ * For fully-virtualised guests there is no option - copying must be used.
++ * For paravirtualised guests, flipping is the default.
++ */
++#ifdef CONFIG_XEN
++static int MODPARM_rx_copy = 0;
++module_param_named(rx_copy, MODPARM_rx_copy, bool, 0);
++MODULE_PARM_DESC(rx_copy, "Copy packets from network card (rather than flip)");
++static int MODPARM_rx_flip = 0;
++module_param_named(rx_flip, MODPARM_rx_flip, bool, 0);
++MODULE_PARM_DESC(rx_flip, "Flip packets from network card (rather than copy)");
++#else
++static const int MODPARM_rx_copy = 1;
++static const int MODPARM_rx_flip = 0;
++#endif
++
++#define RX_COPY_THRESHOLD 256
++
++/* If we don't have GSO, fake things up so that we never try to use it. */
++#if defined(NETIF_F_GSO)
++#define HAVE_GSO 1
++#define HAVE_TSO 1 /* TSO is a subset of GSO */
++#define HAVE_CSUM_OFFLOAD 1
++static inline void dev_disable_gso_features(struct net_device *dev)
++{
++ /* Turn off all GSO bits except ROBUST. */
++ dev->features &= (1 << NETIF_F_GSO_SHIFT) - 1;
++ dev->features |= NETIF_F_GSO_ROBUST;
++}
++#elif defined(NETIF_F_TSO)
++#define HAVE_GSO 0
++#define HAVE_TSO 1
+
-+cpumask_t cpu_online_map;
-+EXPORT_SYMBOL(cpu_online_map);
-+cpumask_t cpu_possible_map;
-+EXPORT_SYMBOL(cpu_possible_map);
-+cpumask_t cpu_initialized_map;
++/* Some older kernels cannot cope with incorrect checksums,
++ * particularly in netfilter. I'm not sure there is 100% correlation
++ * with the presence of NETIF_F_TSO but it appears to be a good first
++ * approximiation.
++ */
++#define HAVE_CSUM_OFFLOAD 0
+
-+struct cpuinfo_x86 cpu_data[NR_CPUS] __cacheline_aligned;
-+EXPORT_SYMBOL(cpu_data);
++#define gso_size tso_size
++#define gso_segs tso_segs
++static inline void dev_disable_gso_features(struct net_device *dev)
++{
++ /* Turn off all TSO bits. */
++ dev->features &= ~NETIF_F_TSO;
++}
++static inline int skb_is_gso(const struct sk_buff *skb)
++{
++ return skb_shinfo(skb)->tso_size;
++}
++static inline int skb_gso_ok(struct sk_buff *skb, int features)
++{
++ return (features & NETIF_F_TSO);
++}
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+DEFINE_PER_CPU(int, cpu_state) = { 0 };
++static inline int netif_needs_gso(struct net_device *dev, struct sk_buff *skb)
++{
++ return skb_is_gso(skb) &&
++ (!skb_gso_ok(skb, dev->features) ||
++ unlikely(skb->ip_summed != CHECKSUM_HW));
++}
++#else
++#define HAVE_GSO 0
++#define HAVE_TSO 0
++#define HAVE_CSUM_OFFLOAD 0
++#define netif_needs_gso(dev, skb) 0
++#define dev_disable_gso_features(dev) ((void)0)
++#define ethtool_op_set_tso(dev, data) (-ENOSYS)
+#endif
+
-+static DEFINE_PER_CPU(int, resched_irq);
-+static DEFINE_PER_CPU(int, callfunc_irq);
-+static char resched_name[NR_CPUS][15];
-+static char callfunc_name[NR_CPUS][15];
-+
-+u8 cpu_2_logical_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = BAD_APICID };
++#define GRANT_INVALID_REF 0
+
-+void *xquad_portio;
++struct netfront_rx_info {
++ struct netif_rx_response rx;
++ struct netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX - 1];
++};
+
-+cpumask_t cpu_sibling_map[NR_CPUS] __cacheline_aligned;
-+cpumask_t cpu_core_map[NR_CPUS] __cacheline_aligned;
-+EXPORT_SYMBOL(cpu_core_map);
++/*
++ * Implement our own carrier flag: the network stack's version causes delays
++ * when the carrier is re-enabled (in particular, dev_activate() may not
++ * immediately be called, which can cause packet loss).
++ */
++#define netfront_carrier_on(netif) ((netif)->carrier = 1)
++#define netfront_carrier_off(netif) ((netif)->carrier = 0)
++#define netfront_carrier_ok(netif) ((netif)->carrier)
+
-+#if defined(__i386__)
-+u8 x86_cpu_to_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = 0xff };
-+EXPORT_SYMBOL(x86_cpu_to_apicid);
-+#elif !defined(CONFIG_X86_IO_APIC)
-+unsigned int maxcpus = NR_CPUS;
-+#endif
++/*
++ * Access macros for acquiring freeing slots in tx_skbs[].
++ */
+
-+void __init prefill_possible_map(void)
++static inline void add_id_to_freelist(struct sk_buff **list, unsigned short id)
+{
-+ int i, rc;
++ list[id] = list[0];
++ list[0] = (void *)(unsigned long)id;
++}
+
-+ for_each_possible_cpu(i)
-+ if (i != smp_processor_id())
-+ return;
++static inline unsigned short get_id_from_freelist(struct sk_buff **list)
++{
++ unsigned int id = (unsigned int)(unsigned long)list[0];
++ list[0] = list[id];
++ return id;
++}
+
-+ for (i = 0; i < NR_CPUS; i++) {
-+ rc = HYPERVISOR_vcpu_op(VCPUOP_is_up, i, NULL);
-+ if (rc >= 0)
-+ cpu_set(i, cpu_possible_map);
-+ }
++static inline int xennet_rxidx(RING_IDX idx)
++{
++ return idx & (NET_RX_RING_SIZE - 1);
+}
+
-+void __init smp_alloc_memory(void)
++static inline struct sk_buff *xennet_get_rx_skb(struct netfront_info *np,
++ RING_IDX ri)
+{
++ int i = xennet_rxidx(ri);
++ struct sk_buff *skb = np->rx_skbs[i];
++ np->rx_skbs[i] = NULL;
++ return skb;
+}
+
-+static inline void
-+set_cpu_sibling_map(int cpu)
++static inline grant_ref_t xennet_get_rx_ref(struct netfront_info *np,
++ RING_IDX ri)
+{
-+ phys_proc_id[cpu] = cpu;
-+ cpu_core_id[cpu] = 0;
++ int i = xennet_rxidx(ri);
++ grant_ref_t ref = np->grant_rx_ref[i];
++ np->grant_rx_ref[i] = GRANT_INVALID_REF;
++ return ref;
++}
+
-+ cpu_sibling_map[cpu] = cpumask_of_cpu(cpu);
-+ cpu_core_map[cpu] = cpumask_of_cpu(cpu);
++#define DPRINTK(fmt, args...) \
++ pr_debug("netfront (%s:%d) " fmt, \
++ __FUNCTION__, __LINE__, ##args)
++#define IPRINTK(fmt, args...) \
++ printk(KERN_INFO "netfront: " fmt, ##args)
++#define WPRINTK(fmt, args...) \
++ printk(KERN_WARNING "netfront: " fmt, ##args)
+
-+ cpu_data[cpu].booted_cores = 1;
-+}
++static int setup_device(struct xenbus_device *, struct netfront_info *);
++static struct net_device *create_netdev(struct xenbus_device *);
+
-+static void
-+remove_siblinginfo(int cpu)
-+{
-+ phys_proc_id[cpu] = BAD_APICID;
-+ cpu_core_id[cpu] = BAD_APICID;
++static void end_access(int, void *);
++static void netif_disconnect_backend(struct netfront_info *);
+
-+ cpus_clear(cpu_sibling_map[cpu]);
-+ cpus_clear(cpu_core_map[cpu]);
++static int network_connect(struct net_device *);
++static void network_tx_buf_gc(struct net_device *);
++static void network_alloc_rx_buffers(struct net_device *);
++static void send_fake_arp(struct net_device *);
+
-+ cpu_data[cpu].booted_cores = 0;
++static irqreturn_t netif_int(int irq, void *dev_id, struct pt_regs *ptregs);
++
++#ifdef CONFIG_SYSFS
++static int xennet_sysfs_addif(struct net_device *netdev);
++static void xennet_sysfs_delif(struct net_device *netdev);
++#else /* !CONFIG_SYSFS */
++#define xennet_sysfs_addif(dev) (0)
++#define xennet_sysfs_delif(dev) do { } while(0)
++#endif
++
++static inline int xennet_can_sg(struct net_device *dev)
++{
++ return dev->features & NETIF_F_SG;
+}
+
-+static int xen_smp_intr_init(unsigned int cpu)
++/**
++ * Entry point to this code when a new device is created. Allocate the basic
++ * structures and the ring buffers for communication with the backend, and
++ * inform the backend of the appropriate details for those.
++ */
++static int __devinit netfront_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
+{
-+ int rc;
++ int err;
++ struct net_device *netdev;
++ struct netfront_info *info;
+
-+ per_cpu(resched_irq, cpu) = per_cpu(callfunc_irq, cpu) = -1;
++ netdev = create_netdev(dev);
++ if (IS_ERR(netdev)) {
++ err = PTR_ERR(netdev);
++ xenbus_dev_fatal(dev, err, "creating netdev");
++ return err;
++ }
+
-+ sprintf(resched_name[cpu], "resched%d", cpu);
-+ rc = bind_ipi_to_irqhandler(RESCHEDULE_VECTOR,
-+ cpu,
-+ smp_reschedule_interrupt,
-+ SA_INTERRUPT,
-+ resched_name[cpu],
-+ NULL);
-+ if (rc < 0)
-+ goto fail;
-+ per_cpu(resched_irq, cpu) = rc;
++ info = netdev_priv(netdev);
++ dev->dev.driver_data = info;
+
-+ sprintf(callfunc_name[cpu], "callfunc%d", cpu);
-+ rc = bind_ipi_to_irqhandler(CALL_FUNCTION_VECTOR,
-+ cpu,
-+ smp_call_function_interrupt,
-+ SA_INTERRUPT,
-+ callfunc_name[cpu],
-+ NULL);
-+ if (rc < 0)
++ err = register_netdev(info->netdev);
++ if (err) {
++ printk(KERN_WARNING "%s: register_netdev err=%d\n",
++ __FUNCTION__, err);
+ goto fail;
-+ per_cpu(callfunc_irq, cpu) = rc;
++ }
+
-+ if ((cpu != 0) && ((rc = local_setup_timer(cpu)) != 0))
++ err = xennet_sysfs_addif(info->netdev);
++ if (err) {
++ unregister_netdev(info->netdev);
++ printk(KERN_WARNING "%s: add sysfs failed err=%d\n",
++ __FUNCTION__, err);
+ goto fail;
++ }
+
+ return 0;
+
+ fail:
-+ if (per_cpu(resched_irq, cpu) >= 0)
-+ unbind_from_irqhandler(per_cpu(resched_irq, cpu), NULL);
-+ if (per_cpu(callfunc_irq, cpu) >= 0)
-+ unbind_from_irqhandler(per_cpu(callfunc_irq, cpu), NULL);
-+ return rc;
++ free_netdev(netdev);
++ dev->dev.driver_data = NULL;
++ return err;
+}
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+static void xen_smp_intr_exit(unsigned int cpu)
++static int __devexit netfront_remove(struct xenbus_device *dev)
+{
-+ if (cpu != 0)
-+ local_teardown_timer(cpu);
++ struct netfront_info *info = dev->dev.driver_data;
+
-+ unbind_from_irqhandler(per_cpu(resched_irq, cpu), NULL);
-+ unbind_from_irqhandler(per_cpu(callfunc_irq, cpu), NULL);
-+}
-+#endif
++ DPRINTK("%s\n", dev->nodename);
+
-+void cpu_bringup(void)
-+{
-+ cpu_init();
-+ touch_softlockup_watchdog();
-+ preempt_disable();
-+ local_irq_enable();
-+}
++ netfront_accelerator_call_remove(info, dev);
+
-+static void cpu_bringup_and_idle(void)
-+{
-+ cpu_bringup();
-+ cpu_idle();
-+}
++ netif_disconnect_backend(info);
+
-+static void cpu_initialize_context(unsigned int cpu)
-+{
-+ vcpu_guest_context_t ctxt;
-+ struct task_struct *idle = idle_task(cpu);
-+#ifdef __x86_64__
-+ struct desc_ptr *gdt_descr = &cpu_gdt_descr[cpu];
-+#else
-+ struct Xgt_desc_struct *gdt_descr = &per_cpu(cpu_gdt_descr, cpu);
-+#endif
++ del_timer_sync(&info->rx_refill_timer);
+
-+ if (cpu_test_and_set(cpu, cpu_initialized_map))
-+ return;
++ xennet_sysfs_delif(info->netdev);
+
-+ memset(&ctxt, 0, sizeof(ctxt));
++ unregister_netdev(info->netdev);
+
-+ ctxt.flags = VGCF_IN_KERNEL;
-+ ctxt.user_regs.ds = __USER_DS;
-+ ctxt.user_regs.es = __USER_DS;
-+ ctxt.user_regs.fs = 0;
-+ ctxt.user_regs.gs = 0;
-+ ctxt.user_regs.ss = __KERNEL_DS;
-+ ctxt.user_regs.eip = (unsigned long)cpu_bringup_and_idle;
-+ ctxt.user_regs.eflags = X86_EFLAGS_IF | 0x1000; /* IOPL_RING1 */
++ free_netdev(info->netdev);
+
-+ memset(&ctxt.fpu_ctxt, 0, sizeof(ctxt.fpu_ctxt));
++ return 0;
++}
+
-+ smp_trap_init(ctxt.trap_ctxt);
+
-+ ctxt.ldt_ents = 0;
++static int netfront_suspend(struct xenbus_device *dev)
++{
++ struct netfront_info *info = dev->dev.driver_data;
++ return netfront_accelerator_suspend(info, dev);
++}
+
-+ ctxt.gdt_frames[0] = virt_to_mfn(gdt_descr->address);
-+ ctxt.gdt_ents = gdt_descr->size / 8;
+
-+#ifdef __i386__
-+ ctxt.user_regs.cs = __KERNEL_CS;
-+ ctxt.user_regs.esp = idle->thread.esp0 - sizeof(struct pt_regs);
++static int netfront_suspend_cancel(struct xenbus_device *dev)
++{
++ struct netfront_info *info = dev->dev.driver_data;
++ return netfront_accelerator_suspend_cancel(info, dev);
++}
+
-+ ctxt.kernel_ss = __KERNEL_DS;
-+ ctxt.kernel_sp = idle->thread.esp0;
+
-+ ctxt.event_callback_cs = __KERNEL_CS;
-+ ctxt.event_callback_eip = (unsigned long)hypervisor_callback;
-+ ctxt.failsafe_callback_cs = __KERNEL_CS;
-+ ctxt.failsafe_callback_eip = (unsigned long)failsafe_callback;
++/**
++ * We are reconnecting to the backend, due to a suspend/resume, or a backend
++ * driver restart. We tear down our netif structure and recreate it, but
++ * leave the device-layer structures intact so that this is transparent to the
++ * rest of the kernel.
++ */
++static int netfront_resume(struct xenbus_device *dev)
++{
++ struct netfront_info *info = dev->dev.driver_data;
+
-+ ctxt.ctrlreg[3] = xen_pfn_to_cr3(virt_to_mfn(swapper_pg_dir));
-+#else /* __x86_64__ */
-+ ctxt.user_regs.cs = __KERNEL_CS;
-+ ctxt.user_regs.esp = idle->thread.rsp0 - sizeof(struct pt_regs);
++ DPRINTK("%s\n", dev->nodename);
+
-+ ctxt.kernel_ss = __KERNEL_DS;
-+ ctxt.kernel_sp = idle->thread.rsp0;
++ netfront_accelerator_resume(info, dev);
+
-+ ctxt.event_callback_eip = (unsigned long)hypervisor_callback;
-+ ctxt.failsafe_callback_eip = (unsigned long)failsafe_callback;
-+ ctxt.syscall_callback_eip = (unsigned long)system_call;
++ netif_disconnect_backend(info);
++ return 0;
++}
+
-+ ctxt.ctrlreg[3] = xen_pfn_to_cr3(virt_to_mfn(init_level4_pgt));
++static int xen_net_read_mac(struct xenbus_device *dev, u8 mac[])
++{
++ char *s, *e, *macstr;
++ int i;
+
-+ ctxt.gs_base_kernel = (unsigned long)(cpu_pda(cpu));
-+#endif
++ macstr = s = xenbus_read(XBT_NIL, dev->nodename, "mac", NULL);
++ if (IS_ERR(macstr))
++ return PTR_ERR(macstr);
++
++ for (i = 0; i < ETH_ALEN; i++) {
++ mac[i] = simple_strtoul(s, &e, 16);
++ if ((s == e) || (*e != ((i == ETH_ALEN-1) ? '\0' : ':'))) {
++ kfree(macstr);
++ return -ENOENT;
++ }
++ s = e+1;
++ }
+
-+ BUG_ON(HYPERVISOR_vcpu_op(VCPUOP_initialise, cpu, &ctxt));
++ kfree(macstr);
++ return 0;
+}
+
-+void __init smp_prepare_cpus(unsigned int max_cpus)
++/* Common code used when first setting up, and when resuming. */
++static int talk_to_backend(struct xenbus_device *dev,
++ struct netfront_info *info)
+{
-+ int cpu;
-+ struct task_struct *idle;
-+#ifdef __x86_64__
-+ struct desc_ptr *gdt_descr;
-+#else
-+ struct Xgt_desc_struct *gdt_descr;
-+#endif
++ const char *message;
++ struct xenbus_transaction xbt;
++ int err;
+
-+ boot_cpu_data.apicid = 0;
-+ cpu_data[0] = boot_cpu_data;
++ /* Read mac only in the first setup. */
++ if (!is_valid_ether_addr(info->mac)) {
++ err = xen_net_read_mac(dev, info->mac);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "parsing %s/mac",
++ dev->nodename);
++ goto out;
++ }
++ }
+
-+ cpu_2_logical_apicid[0] = 0;
-+ x86_cpu_to_apicid[0] = 0;
++ /* Create shared ring, alloc event channel. */
++ err = setup_device(dev, info);
++ if (err)
++ goto out;
+
-+ current_thread_info()->cpu = 0;
++ /* This will load an accelerator if one is configured when the
++ * watch fires */
++ netfront_accelerator_add_watch(info);
+
-+ for (cpu = 0; cpu < NR_CPUS; cpu++) {
-+ cpus_clear(cpu_sibling_map[cpu]);
-+ cpus_clear(cpu_core_map[cpu]);
++again:
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "starting transaction");
++ goto destroy_ring;
+ }
+
-+ set_cpu_sibling_map(0);
++ err = xenbus_printf(xbt, dev->nodename, "tx-ring-ref","%u",
++ info->tx_ring_ref);
++ if (err) {
++ message = "writing tx ring-ref";
++ goto abort_transaction;
++ }
++ err = xenbus_printf(xbt, dev->nodename, "rx-ring-ref","%u",
++ info->rx_ring_ref);
++ if (err) {
++ message = "writing rx ring-ref";
++ goto abort_transaction;
++ }
++ err = xenbus_printf(xbt, dev->nodename,
++ "event-channel", "%u",
++ irq_to_evtchn_port(info->irq));
++ if (err) {
++ message = "writing event-channel";
++ goto abort_transaction;
++ }
+
-+ if (xen_smp_intr_init(0))
-+ BUG();
++ err = xenbus_printf(xbt, dev->nodename, "request-rx-copy", "%u",
++ info->copying_receiver);
++ if (err) {
++ message = "writing request-rx-copy";
++ goto abort_transaction;
++ }
+
-+ cpu_initialized_map = cpumask_of_cpu(0);
++ err = xenbus_printf(xbt, dev->nodename, "feature-rx-notify", "%d", 1);
++ if (err) {
++ message = "writing feature-rx-notify";
++ goto abort_transaction;
++ }
+
-+ /* Restrict the possible_map according to max_cpus. */
-+ while ((num_possible_cpus() > 1) && (num_possible_cpus() > max_cpus)) {
-+ for (cpu = NR_CPUS-1; !cpu_isset(cpu, cpu_possible_map); cpu--)
-+ continue;
-+ cpu_clear(cpu, cpu_possible_map);
++ err = xenbus_printf(xbt, dev->nodename, "feature-no-csum-offload",
++ "%d", !HAVE_CSUM_OFFLOAD);
++ if (err) {
++ message = "writing feature-no-csum-offload";
++ goto abort_transaction;
+ }
+
-+ for_each_possible_cpu (cpu) {
-+ if (cpu == 0)
-+ continue;
++ err = xenbus_printf(xbt, dev->nodename, "feature-sg", "%d", 1);
++ if (err) {
++ message = "writing feature-sg";
++ goto abort_transaction;
++ }
+
-+#ifdef __x86_64__
-+ gdt_descr = &cpu_gdt_descr[cpu];
-+#else
-+ gdt_descr = &per_cpu(cpu_gdt_descr, cpu);
-+#endif
-+ gdt_descr->address = get_zeroed_page(GFP_KERNEL);
-+ if (unlikely(!gdt_descr->address)) {
-+ printk(KERN_CRIT "CPU%d failed to allocate GDT\n",
-+ cpu);
-+ continue;
-+ }
-+ gdt_descr->size = GDT_SIZE;
-+ memcpy((void *)gdt_descr->address, cpu_gdt_table, GDT_SIZE);
-+ make_page_readonly(
-+ (void *)gdt_descr->address,
-+ XENFEAT_writable_descriptor_tables);
++ err = xenbus_printf(xbt, dev->nodename, "feature-gso-tcpv4", "%d",
++ HAVE_TSO);
++ if (err) {
++ message = "writing feature-gso-tcpv4";
++ goto abort_transaction;
++ }
+
-+ cpu_data[cpu] = boot_cpu_data;
-+ cpu_data[cpu].apicid = cpu;
++ err = xenbus_transaction_end(xbt, 0);
++ if (err) {
++ if (err == -EAGAIN)
++ goto again;
++ xenbus_dev_fatal(dev, err, "completing transaction");
++ goto destroy_ring;
++ }
+
-+ cpu_2_logical_apicid[cpu] = cpu;
-+ x86_cpu_to_apicid[cpu] = cpu;
++ return 0;
+
-+ idle = fork_idle(cpu);
-+ if (IS_ERR(idle))
-+ panic("failed fork for CPU %d", cpu);
++ abort_transaction:
++ xenbus_transaction_end(xbt, 1);
++ xenbus_dev_fatal(dev, err, "%s", message);
++ destroy_ring:
++ netfront_accelerator_call_remove(info, dev);
++ netif_disconnect_backend(info);
++ out:
++ return err;
++}
+
-+#ifdef __x86_64__
-+ cpu_pda(cpu)->pcurrent = idle;
-+ cpu_pda(cpu)->cpunumber = cpu;
-+ clear_ti_thread_flag(idle->thread_info, TIF_FORK);
-+#endif
++static int setup_device(struct xenbus_device *dev, struct netfront_info *info)
++{
++ struct netif_tx_sring *txs;
++ struct netif_rx_sring *rxs;
++ int err;
++ struct net_device *netdev = info->netdev;
+
-+ irq_ctx_init(cpu);
++ info->tx_ring_ref = GRANT_INVALID_REF;
++ info->rx_ring_ref = GRANT_INVALID_REF;
++ info->rx.sring = NULL;
++ info->tx.sring = NULL;
++ info->irq = 0;
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+ if (is_initial_xendomain())
-+ cpu_set(cpu, cpu_present_map);
-+#else
-+ cpu_set(cpu, cpu_present_map);
-+#endif
++ txs = (struct netif_tx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH);
++ if (!txs) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(dev, err, "allocating tx ring page");
++ goto fail;
+ }
++ SHARED_RING_INIT(txs);
++ FRONT_RING_INIT(&info->tx, txs, PAGE_SIZE);
+
-+ init_xenbus_allowed_cpumask();
++ err = xenbus_grant_ring(dev, virt_to_mfn(txs));
++ if (err < 0) {
++ free_page((unsigned long)txs);
++ goto fail;
++ }
++ info->tx_ring_ref = err;
+
-+#ifdef CONFIG_X86_IO_APIC
-+ /*
-+ * Here we can be sure that there is an IO-APIC in the system. Let's
-+ * go and set it up:
-+ */
-+ if (!skip_ioapic_setup && nr_ioapics)
-+ setup_IO_APIC();
-+#endif
-+}
++ rxs = (struct netif_rx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH);
++ if (!rxs) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(dev, err, "allocating rx ring page");
++ goto fail;
++ }
++ SHARED_RING_INIT(rxs);
++ FRONT_RING_INIT(&info->rx, rxs, PAGE_SIZE);
+
-+void __devinit smp_prepare_boot_cpu(void)
-+{
-+ prefill_possible_map();
-+}
++ err = xenbus_grant_ring(dev, virt_to_mfn(rxs));
++ if (err < 0) {
++ free_page((unsigned long)rxs);
++ goto fail;
++ }
++ info->rx_ring_ref = err;
+
-+#ifdef CONFIG_HOTPLUG_CPU
++ memcpy(netdev->dev_addr, info->mac, ETH_ALEN);
++
++ err = bind_listening_port_to_irqhandler(
++ dev->otherend_id, netif_int, SA_SAMPLE_RANDOM, netdev->name,
++ netdev);
++ if (err < 0)
++ goto fail;
++ info->irq = err;
+
-+/*
-+ * Initialize cpu_present_map late to skip SMP boot code in init/main.c.
-+ * But do it early enough to catch critical for_each_present_cpu() loops
-+ * in i386-specific code.
-+ */
-+static int __init initialize_cpu_present_map(void)
-+{
-+ cpu_present_map = cpu_possible_map;
+ return 0;
++
++ fail:
++ return err;
+}
-+core_initcall(initialize_cpu_present_map);
+
-+int __cpu_disable(void)
++/**
++ * Callback received when the backend's state changes.
++ */
++static void backend_changed(struct xenbus_device *dev,
++ enum xenbus_state backend_state)
+{
-+ cpumask_t map = cpu_online_map;
-+ int cpu = smp_processor_id();
++ struct netfront_info *np = dev->dev.driver_data;
++ struct net_device *netdev = np->netdev;
+
-+ if (cpu == 0)
-+ return -EBUSY;
++ DPRINTK("%s\n", xenbus_strstate(backend_state));
+
-+ remove_siblinginfo(cpu);
++ switch (backend_state) {
++ case XenbusStateInitialising:
++ case XenbusStateInitialised:
++ case XenbusStateConnected:
++ case XenbusStateReconfiguring:
++ case XenbusStateReconfigured:
++ case XenbusStateUnknown:
++ case XenbusStateClosed:
++ break;
+
-+ cpu_clear(cpu, map);
-+ fixup_irqs(map);
-+ cpu_clear(cpu, cpu_online_map);
++ case XenbusStateInitWait:
++ if (dev->state != XenbusStateInitialising)
++ break;
++ if (network_connect(netdev) != 0)
++ break;
++ xenbus_switch_state(dev, XenbusStateConnected);
++ send_fake_arp(netdev);
++ break;
+
-+ return 0;
++ case XenbusStateClosing:
++ xenbus_frontend_closed(dev);
++ break;
++ }
+}
+
-+void __cpu_die(unsigned int cpu)
++/** Send a packet on a net device to encourage switches to learn the
++ * MAC. We send a fake ARP request.
++ *
++ * @param dev device
++ * @return 0 on success, error code otherwise
++ */
++static void send_fake_arp(struct net_device *dev)
+{
-+ while (HYPERVISOR_vcpu_op(VCPUOP_is_up, cpu, NULL)) {
-+ current->state = TASK_UNINTERRUPTIBLE;
-+ schedule_timeout(HZ/10);
-+ }
++#ifdef CONFIG_INET
++ struct sk_buff *skb;
++ u32 src_ip, dst_ip;
+
-+ xen_smp_intr_exit(cpu);
++ dst_ip = INADDR_BROADCAST;
++ src_ip = inet_select_addr(dev, dst_ip, RT_SCOPE_LINK);
+
-+ if (num_online_cpus() == 1)
-+ alternatives_smp_switch(0);
-+}
++ /* No IP? Then nothing to do. */
++ if (src_ip == 0)
++ return;
++
++ skb = arp_create(ARPOP_REPLY, ETH_P_ARP,
++ dst_ip, dev, src_ip,
++ /*dst_hw*/ NULL, /*src_hw*/ NULL,
++ /*target_hw*/ dev->dev_addr);
++ if (skb == NULL)
++ return;
+
-+#else /* !CONFIG_HOTPLUG_CPU */
++ dev_queue_xmit(skb);
++#endif
++}
+
-+int __cpu_disable(void)
++static inline int netfront_tx_slot_available(struct netfront_info *np)
+{
-+ return -ENOSYS;
++ return ((np->tx.req_prod_pvt - np->tx.rsp_cons) <
++ (TX_MAX_TARGET - MAX_SKB_FRAGS - 2));
+}
+
-+void __cpu_die(unsigned int cpu)
++
++static inline void network_maybe_wake_tx(struct net_device *dev)
+{
-+ BUG();
++ struct netfront_info *np = netdev_priv(dev);
++
++ if (unlikely(netif_queue_stopped(dev)) &&
++ netfront_tx_slot_available(np) &&
++ likely(netif_running(dev)) &&
++ netfront_check_accelerator_queue_ready(dev, np))
++ netif_wake_queue(dev);
+}
+
-+#endif /* CONFIG_HOTPLUG_CPU */
+
-+int __devinit __cpu_up(unsigned int cpu)
++int netfront_check_queue_ready(struct net_device *dev)
+{
-+ int rc;
++ struct netfront_info *np = netdev_priv(dev);
+
-+ rc = cpu_up_check(cpu);
-+ if (rc)
-+ return rc;
++ return unlikely(netif_queue_stopped(dev)) &&
++ netfront_tx_slot_available(np) &&
++ likely(netif_running(dev));
++}
++EXPORT_SYMBOL(netfront_check_queue_ready);
+
-+ cpu_initialize_context(cpu);
+
-+ if (num_online_cpus() == 1)
-+ alternatives_smp_switch(1);
++static int network_open(struct net_device *dev)
++{
++ struct netfront_info *np = netdev_priv(dev);
+
-+ /* This must be done before setting cpu_online_map */
-+ set_cpu_sibling_map(cpu);
-+ wmb();
++ memset(&np->stats, 0, sizeof(np->stats));
+
-+ rc = xen_smp_intr_init(cpu);
-+ if (rc) {
-+ remove_siblinginfo(cpu);
-+ return rc;
-+ }
++ spin_lock_bh(&np->rx_lock);
++ if (netfront_carrier_ok(np)) {
++ network_alloc_rx_buffers(dev);
++ np->rx.sring->rsp_event = np->rx.rsp_cons + 1;
++ if (RING_HAS_UNCONSUMED_RESPONSES(&np->rx)){
++ netfront_accelerator_call_stop_napi_irq(np, dev);
+
-+ cpu_set(cpu, cpu_online_map);
++ netif_rx_schedule(dev);
++ }
++ }
++ spin_unlock_bh(&np->rx_lock);
+
-+ rc = HYPERVISOR_vcpu_op(VCPUOP_up, cpu, NULL);
-+ BUG_ON(rc);
++ network_maybe_wake_tx(dev);
+
+ return 0;
+}
+
-+void __init smp_cpus_done(unsigned int max_cpus)
-+{
-+}
-+
-+#ifndef CONFIG_X86_LOCAL_APIC
-+int setup_profiling_timer(unsigned int multiplier)
++static void network_tx_buf_gc(struct net_device *dev)
+{
-+ return -EINVAL;
-+}
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/core/xen_proc.c tmp-linux-2.6-xen.patch/drivers/xen/core/xen_proc.c
---- pristine-linux-2.6.18.2/drivers/xen/core/xen_proc.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/core/xen_proc.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,23 @@
-+
-+#include <linux/module.h>
-+#include <linux/proc_fs.h>
-+#include <xen/xen_proc.h>
++ RING_IDX cons, prod;
++ unsigned short id;
++ struct netfront_info *np = netdev_priv(dev);
++ struct sk_buff *skb;
+
-+static struct proc_dir_entry *xen_base;
++ BUG_ON(!netfront_carrier_ok(np));
+
-+struct proc_dir_entry *create_xen_proc_entry(const char *name, mode_t mode)
-+{
-+ if ( xen_base == NULL )
-+ if ( (xen_base = proc_mkdir("xen", &proc_root)) == NULL )
-+ panic("Couldn't create /proc/xen");
-+ return create_proc_entry(name, mode, xen_base);
-+}
++ do {
++ prod = np->tx.sring->rsp_prod;
++ rmb(); /* Ensure we see responses up to 'rp'. */
+
-+EXPORT_SYMBOL_GPL(create_xen_proc_entry);
++ for (cons = np->tx.rsp_cons; cons != prod; cons++) {
++ struct netif_tx_response *txrsp;
+
-+void remove_xen_proc_entry(const char *name)
-+{
-+ remove_proc_entry(name, xen_base);
-+}
++ txrsp = RING_GET_RESPONSE(&np->tx, cons);
++ if (txrsp->status == NETIF_RSP_NULL)
++ continue;
+
-+EXPORT_SYMBOL_GPL(remove_xen_proc_entry);
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/core/xen_sysfs.c tmp-linux-2.6-xen.patch/drivers/xen/core/xen_sysfs.c
---- pristine-linux-2.6.18.2/drivers/xen/core/xen_sysfs.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/core/xen_sysfs.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,378 @@
-+/*
-+ * copyright (c) 2006 IBM Corporation
-+ * Authored by: Mike D. Day <ncmike at us.ibm.com>
-+ *
-+ * This program is free software; you can redistribute it and/or modify
-+ * it under the terms of the GNU General Public License version 2 as
-+ * published by the Free Software Foundation.
-+ */
++ id = txrsp->id;
++ skb = np->tx_skbs[id];
++ if (unlikely(gnttab_query_foreign_access(
++ np->grant_tx_ref[id]) != 0)) {
++ printk(KERN_ALERT "network_tx_buf_gc: warning "
++ "-- grant still in use by backend "
++ "domain.\n");
++ BUG();
++ }
++ gnttab_end_foreign_access_ref(np->grant_tx_ref[id]);
++ gnttab_release_grant_reference(
++ &np->gref_tx_head, np->grant_tx_ref[id]);
++ np->grant_tx_ref[id] = GRANT_INVALID_REF;
++ add_id_to_freelist(np->tx_skbs, id);
++ dev_kfree_skb_irq(skb);
++ }
+
-+#include <linux/err.h>
-+#include <linux/kernel.h>
-+#include <linux/module.h>
-+#include <linux/init.h>
-+#include <asm/hypervisor.h>
-+#include <xen/features.h>
-+#include <xen/hypervisor_sysfs.h>
-+#include <xen/xenbus.h>
++ np->tx.rsp_cons = prod;
+
-+MODULE_LICENSE("GPL");
-+MODULE_AUTHOR("Mike D. Day <ncmike at us.ibm.com>");
++ /*
++ * Set a new event, then check for race with update of tx_cons.
++ * Note that it is essential to schedule a callback, no matter
++ * how few buffers are pending. Even if there is space in the
++ * transmit ring, higher layers may be blocked because too much
++ * data is outstanding: in such cases notification from Xen is
++ * likely to be the only kick that we'll get.
++ */
++ np->tx.sring->rsp_event =
++ prod + ((np->tx.sring->req_prod - prod) >> 1) + 1;
++ mb();
++ } while ((cons == prod) && (prod != np->tx.sring->rsp_prod));
+
-+static ssize_t type_show(struct hyp_sysfs_attr *attr, char *buffer)
-+{
-+ return sprintf(buffer, "xen\n");
++ network_maybe_wake_tx(dev);
+}
+
-+HYPERVISOR_ATTR_RO(type);
-+
-+static int __init xen_sysfs_type_init(void)
++static void rx_refill_timeout(unsigned long data)
+{
-+ return sysfs_create_file(&hypervisor_subsys.kset.kobj, &type_attr.attr);
-+}
++ struct net_device *dev = (struct net_device *)data;
++ struct netfront_info *np = netdev_priv(dev);
+
-+static void xen_sysfs_type_destroy(void)
-+{
-+ sysfs_remove_file(&hypervisor_subsys.kset.kobj, &type_attr.attr);
++ netfront_accelerator_call_stop_napi_irq(np, dev);
++
++ netif_rx_schedule(dev);
+}
+
-+/* xen version attributes */
-+static ssize_t major_show(struct hyp_sysfs_attr *attr, char *buffer)
++static void network_alloc_rx_buffers(struct net_device *dev)
+{
-+ int version = HYPERVISOR_xen_version(XENVER_version, NULL);
-+ if (version)
-+ return sprintf(buffer, "%d\n", version >> 16);
-+ return -ENODEV;
-+}
++ unsigned short id;
++ struct netfront_info *np = netdev_priv(dev);
++ struct sk_buff *skb;
++ struct page *page;
++ int i, batch_target, notify;
++ RING_IDX req_prod = np->rx.req_prod_pvt;
++ struct xen_memory_reservation reservation;
++ grant_ref_t ref;
++ unsigned long pfn;
++ void *vaddr;
++ int nr_flips;
++ netif_rx_request_t *req;
+
-+HYPERVISOR_ATTR_RO(major);
++ if (unlikely(!netfront_carrier_ok(np)))
++ return;
+
-+static ssize_t minor_show(struct hyp_sysfs_attr *attr, char *buffer)
-+{
-+ int version = HYPERVISOR_xen_version(XENVER_version, NULL);
-+ if (version)
-+ return sprintf(buffer, "%d\n", version & 0xff);
-+ return -ENODEV;
-+}
++ /*
++ * Allocate skbuffs greedily, even though we batch updates to the
++ * receive ring. This creates a less bursty demand on the memory
++ * allocator, so should reduce the chance of failed allocation requests
++ * both for ourself and for other kernel subsystems.
++ */
++ batch_target = np->rx_target - (req_prod - np->rx.rsp_cons);
++ for (i = skb_queue_len(&np->rx_batch); i < batch_target; i++) {
++ /*
++ * Allocate an skb and a page. Do not use __dev_alloc_skb as
++ * that will allocate page-sized buffers which is not
++ * necessary here.
++ * 16 bytes added as necessary headroom for netif_receive_skb.
++ */
++ skb = alloc_skb(RX_COPY_THRESHOLD + 16 + NET_IP_ALIGN,
++ GFP_ATOMIC | __GFP_NOWARN);
++ if (unlikely(!skb))
++ goto no_skb;
+
-+HYPERVISOR_ATTR_RO(minor);
++ page = alloc_page(GFP_ATOMIC | __GFP_NOWARN);
++ if (!page) {
++ kfree_skb(skb);
++no_skb:
++ /* Any skbuffs queued for refill? Force them out. */
++ if (i != 0)
++ goto refill;
++ /* Could not allocate any skbuffs. Try again later. */
++ mod_timer(&np->rx_refill_timer,
++ jiffies + (HZ/10));
++ break;
++ }
+
-+static ssize_t extra_show(struct hyp_sysfs_attr *attr, char *buffer)
-+{
-+ int ret = -ENOMEM;
-+ char *extra;
++ skb_reserve(skb, 16 + NET_IP_ALIGN); /* mimic dev_alloc_skb() */
++ skb_shinfo(skb)->frags[0].page = page;
++ skb_shinfo(skb)->nr_frags = 1;
++ __skb_queue_tail(&np->rx_batch, skb);
++ }
+
-+ extra = kmalloc(XEN_EXTRAVERSION_LEN, GFP_KERNEL);
-+ if (extra) {
-+ ret = HYPERVISOR_xen_version(XENVER_extraversion, extra);
-+ if (!ret)
-+ ret = sprintf(buffer, "%s\n", extra);
-+ kfree(extra);
++ /* Is the batch large enough to be worthwhile? */
++ if (i < (np->rx_target/2)) {
++ if (req_prod > np->rx.sring->req_prod)
++ goto push;
++ return;
+ }
+
-+ return ret;
-+}
++ /* Adjust our fill target if we risked running out of buffers. */
++ if (((req_prod - np->rx.sring->rsp_prod) < (np->rx_target / 4)) &&
++ ((np->rx_target *= 2) > np->rx_max_target))
++ np->rx_target = np->rx_max_target;
+
-+HYPERVISOR_ATTR_RO(extra);
++ refill:
++ for (nr_flips = i = 0; ; i++) {
++ if ((skb = __skb_dequeue(&np->rx_batch)) == NULL)
++ break;
+
-+static struct attribute *version_attrs[] = {
-+ &major_attr.attr,
-+ &minor_attr.attr,
-+ &extra_attr.attr,
-+ NULL
-+};
++ skb->dev = dev;
+
-+static struct attribute_group version_group = {
-+ .name = "version",
-+ .attrs = version_attrs,
-+};
++ id = xennet_rxidx(req_prod + i);
+
-+static int __init xen_sysfs_version_init(void)
-+{
-+ return sysfs_create_group(&hypervisor_subsys.kset.kobj,
-+ &version_group);
-+}
++ BUG_ON(np->rx_skbs[id]);
++ np->rx_skbs[id] = skb;
+
-+static void xen_sysfs_version_destroy(void)
-+{
-+ sysfs_remove_group(&hypervisor_subsys.kset.kobj, &version_group);
-+}
++ ref = gnttab_claim_grant_reference(&np->gref_rx_head);
++ BUG_ON((signed short)ref < 0);
++ np->grant_rx_ref[id] = ref;
+
-+/* UUID */
++ pfn = page_to_pfn(skb_shinfo(skb)->frags[0].page);
++ vaddr = page_address(skb_shinfo(skb)->frags[0].page);
+
-+static ssize_t uuid_show(struct hyp_sysfs_attr *attr, char *buffer)
-+{
-+ char *vm, *val;
-+ int ret;
++ req = RING_GET_REQUEST(&np->rx, req_prod + i);
++ if (!np->copying_receiver) {
++ gnttab_grant_foreign_transfer_ref(ref,
++ np->xbdev->otherend_id,
++ pfn);
++ np->rx_pfn_array[nr_flips] = pfn_to_mfn(pfn);
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* Remove this page before passing
++ * back to Xen. */
++ set_phys_to_machine(pfn, INVALID_P2M_ENTRY);
++ MULTI_update_va_mapping(np->rx_mcl+i,
++ (unsigned long)vaddr,
++ __pte(0), 0);
++ }
++ nr_flips++;
++ } else {
++ gnttab_grant_foreign_access_ref(ref,
++ np->xbdev->otherend_id,
++ pfn_to_mfn(pfn),
++ 0);
++ }
+
-+ vm = xenbus_read(XBT_NIL, "vm", "", NULL);
-+ if (IS_ERR(vm))
-+ return PTR_ERR(vm);
-+ val = xenbus_read(XBT_NIL, vm, "uuid", NULL);
-+ kfree(vm);
-+ if (IS_ERR(val))
-+ return PTR_ERR(val);
-+ ret = sprintf(buffer, "%s\n", val);
-+ kfree(val);
-+ return ret;
-+}
++ req->id = id;
++ req->gref = ref;
++ }
+
-+HYPERVISOR_ATTR_RO(uuid);
++ if ( nr_flips != 0 ) {
++ /* Tell the ballon driver what is going on. */
++ balloon_update_driver_allowance(i);
+
-+static int __init xen_sysfs_uuid_init(void)
-+{
-+ return sysfs_create_file(&hypervisor_subsys.kset.kobj, &uuid_attr.attr);
-+}
++ set_xen_guest_handle(reservation.extent_start,
++ np->rx_pfn_array);
++ reservation.nr_extents = nr_flips;
++ reservation.extent_order = 0;
++ reservation.address_bits = 0;
++ reservation.domid = DOMID_SELF;
+
-+static void xen_sysfs_uuid_destroy(void)
-+{
-+ sysfs_remove_file(&hypervisor_subsys.kset.kobj, &uuid_attr.attr);
-+}
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* After all PTEs have been zapped, flush the TLB. */
++ np->rx_mcl[i-1].args[MULTI_UVMFLAGS_INDEX] =
++ UVMF_TLB_FLUSH|UVMF_ALL;
+
-+/* xen compilation attributes */
++ /* Give away a batch of pages. */
++ np->rx_mcl[i].op = __HYPERVISOR_memory_op;
++ np->rx_mcl[i].args[0] = XENMEM_decrease_reservation;
++ np->rx_mcl[i].args[1] = (unsigned long)&reservation;
+
-+static ssize_t compiler_show(struct hyp_sysfs_attr *attr, char *buffer)
-+{
-+ int ret = -ENOMEM;
-+ struct xen_compile_info *info;
++ /* Zap PTEs and give away pages in one big
++ * multicall. */
++ if (unlikely(HYPERVISOR_multicall(np->rx_mcl, i+1)))
++ BUG();
+
-+ info = kmalloc(sizeof(struct xen_compile_info), GFP_KERNEL);
-+ if (info) {
-+ ret = HYPERVISOR_xen_version(XENVER_compile_info, info);
-+ if (!ret)
-+ ret = sprintf(buffer, "%s\n", info->compiler);
-+ kfree(info);
++ /* Check return status of HYPERVISOR_memory_op(). */
++ if (unlikely(np->rx_mcl[i].result != i))
++ panic("Unable to reduce memory reservation\n");
++ while (nr_flips--)
++ BUG_ON(np->rx_mcl[nr_flips].result);
++ } else {
++ if (HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &reservation) != i)
++ panic("Unable to reduce memory reservation\n");
++ }
++ } else {
++ wmb();
+ }
+
-+ return ret;
++ /* Above is a suitable barrier to ensure backend will see requests. */
++ np->rx.req_prod_pvt = req_prod + i;
++ push:
++ RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&np->rx, notify);
++ if (notify)
++ notify_remote_via_irq(np->irq);
+}
+
-+HYPERVISOR_ATTR_RO(compiler);
-+
-+static ssize_t compiled_by_show(struct hyp_sysfs_attr *attr, char *buffer)
++static void xennet_make_frags(struct sk_buff *skb, struct net_device *dev,
++ struct netif_tx_request *tx)
+{
-+ int ret = -ENOMEM;
-+ struct xen_compile_info *info;
-+
-+ info = kmalloc(sizeof(struct xen_compile_info), GFP_KERNEL);
-+ if (info) {
-+ ret = HYPERVISOR_xen_version(XENVER_compile_info, info);
-+ if (!ret)
-+ ret = sprintf(buffer, "%s\n", info->compile_by);
-+ kfree(info);
-+ }
++ struct netfront_info *np = netdev_priv(dev);
++ char *data = skb->data;
++ unsigned long mfn;
++ RING_IDX prod = np->tx.req_prod_pvt;
++ int frags = skb_shinfo(skb)->nr_frags;
++ unsigned int offset = offset_in_page(data);
++ unsigned int len = skb_headlen(skb);
++ unsigned int id;
++ grant_ref_t ref;
++ int i;
+
-+ return ret;
-+}
++ while (len > PAGE_SIZE - offset) {
++ tx->size = PAGE_SIZE - offset;
++ tx->flags |= NETTXF_more_data;
++ len -= tx->size;
++ data += tx->size;
++ offset = 0;
+
-+HYPERVISOR_ATTR_RO(compiled_by);
++ id = get_id_from_freelist(np->tx_skbs);
++ np->tx_skbs[id] = skb_get(skb);
++ tx = RING_GET_REQUEST(&np->tx, prod++);
++ tx->id = id;
++ ref = gnttab_claim_grant_reference(&np->gref_tx_head);
++ BUG_ON((signed short)ref < 0);
+
-+static ssize_t compile_date_show(struct hyp_sysfs_attr *attr, char *buffer)
-+{
-+ int ret = -ENOMEM;
-+ struct xen_compile_info *info;
++ mfn = virt_to_mfn(data);
++ gnttab_grant_foreign_access_ref(ref, np->xbdev->otherend_id,
++ mfn, GTF_readonly);
+
-+ info = kmalloc(sizeof(struct xen_compile_info), GFP_KERNEL);
-+ if (info) {
-+ ret = HYPERVISOR_xen_version(XENVER_compile_info, info);
-+ if (!ret)
-+ ret = sprintf(buffer, "%s\n", info->compile_date);
-+ kfree(info);
++ tx->gref = np->grant_tx_ref[id] = ref;
++ tx->offset = offset;
++ tx->size = len;
++ tx->flags = 0;
+ }
+
-+ return ret;
-+}
++ for (i = 0; i < frags; i++) {
++ skb_frag_t *frag = skb_shinfo(skb)->frags + i;
+
-+HYPERVISOR_ATTR_RO(compile_date);
++ tx->flags |= NETTXF_more_data;
++
++ id = get_id_from_freelist(np->tx_skbs);
++ np->tx_skbs[id] = skb_get(skb);
++ tx = RING_GET_REQUEST(&np->tx, prod++);
++ tx->id = id;
++ ref = gnttab_claim_grant_reference(&np->gref_tx_head);
++ BUG_ON((signed short)ref < 0);
+
-+static struct attribute *xen_compile_attrs[] = {
-+ &compiler_attr.attr,
-+ &compiled_by_attr.attr,
-+ &compile_date_attr.attr,
-+ NULL
-+};
++ mfn = pfn_to_mfn(page_to_pfn(frag->page));
++ gnttab_grant_foreign_access_ref(ref, np->xbdev->otherend_id,
++ mfn, GTF_readonly);
+
-+static struct attribute_group xen_compilation_group = {
-+ .name = "compilation",
-+ .attrs = xen_compile_attrs,
-+};
++ tx->gref = np->grant_tx_ref[id] = ref;
++ tx->offset = frag->page_offset;
++ tx->size = frag->size;
++ tx->flags = 0;
++ }
+
-+int __init static xen_compilation_init(void)
-+{
-+ return sysfs_create_group(&hypervisor_subsys.kset.kobj,
-+ &xen_compilation_group);
++ np->tx.req_prod_pvt = prod;
+}
+
-+static void xen_compilation_destroy(void)
++static int network_start_xmit(struct sk_buff *skb, struct net_device *dev)
+{
-+ sysfs_remove_group(&hypervisor_subsys.kset.kobj,
-+ &xen_compilation_group);
-+}
++ unsigned short id;
++ struct netfront_info *np = netdev_priv(dev);
++ struct netif_tx_request *tx;
++ struct netif_extra_info *extra;
++ char *data = skb->data;
++ RING_IDX i;
++ grant_ref_t ref;
++ unsigned long mfn;
++ int notify;
++ int frags = skb_shinfo(skb)->nr_frags;
++ unsigned int offset = offset_in_page(data);
++ unsigned int len = skb_headlen(skb);
+
-+/* xen properties info */
++ /* Check the fast path, if hooks are available */
++ if (np->accel_vif_state.hooks &&
++ np->accel_vif_state.hooks->start_xmit(skb, dev)) {
++ /* Fast path has sent this packet */
++ return 0;
++ }
+
-+static ssize_t capabilities_show(struct hyp_sysfs_attr *attr, char *buffer)
-+{
-+ int ret = -ENOMEM;
-+ char *caps;
++ frags += (offset + len + PAGE_SIZE - 1) / PAGE_SIZE;
++ if (unlikely(frags > MAX_SKB_FRAGS + 1)) {
++ printk(KERN_ALERT "xennet: skb rides the rocket: %d frags\n",
++ frags);
++ dump_stack();
++ goto drop;
++ }
+
-+ caps = kmalloc(XEN_CAPABILITIES_INFO_LEN, GFP_KERNEL);
-+ if (caps) {
-+ ret = HYPERVISOR_xen_version(XENVER_capabilities, caps);
-+ if (!ret)
-+ ret = sprintf(buffer, "%s\n", caps);
-+ kfree(caps);
++ spin_lock_irq(&np->tx_lock);
++
++ if (unlikely(!netfront_carrier_ok(np) ||
++ (frags > 1 && !xennet_can_sg(dev)) ||
++ netif_needs_gso(dev, skb))) {
++ spin_unlock_irq(&np->tx_lock);
++ goto drop;
+ }
+
-+ return ret;
-+}
++ i = np->tx.req_prod_pvt;
+
-+HYPERVISOR_ATTR_RO(capabilities);
++ id = get_id_from_freelist(np->tx_skbs);
++ np->tx_skbs[id] = skb;
+
-+static ssize_t changeset_show(struct hyp_sysfs_attr *attr, char *buffer)
-+{
-+ int ret = -ENOMEM;
-+ char *cset;
++ tx = RING_GET_REQUEST(&np->tx, i);
+
-+ cset = kmalloc(XEN_CHANGESET_INFO_LEN, GFP_KERNEL);
-+ if (cset) {
-+ ret = HYPERVISOR_xen_version(XENVER_changeset, cset);
-+ if (!ret)
-+ ret = sprintf(buffer, "%s\n", cset);
-+ kfree(cset);
-+ }
++ tx->id = id;
++ ref = gnttab_claim_grant_reference(&np->gref_tx_head);
++ BUG_ON((signed short)ref < 0);
++ mfn = virt_to_mfn(data);
++ gnttab_grant_foreign_access_ref(
++ ref, np->xbdev->otherend_id, mfn, GTF_readonly);
++ tx->gref = np->grant_tx_ref[id] = ref;
++ tx->offset = offset;
++ tx->size = len;
+
-+ return ret;
-+}
++ tx->flags = 0;
++ extra = NULL;
+
-+HYPERVISOR_ATTR_RO(changeset);
++ if (skb->ip_summed == CHECKSUM_HW) /* local packet? */
++ tx->flags |= NETTXF_csum_blank | NETTXF_data_validated;
++#ifdef CONFIG_XEN
++ if (skb->proto_data_valid) /* remote but checksummed? */
++ tx->flags |= NETTXF_data_validated;
++#endif
+
-+static ssize_t virtual_start_show(struct hyp_sysfs_attr *attr, char *buffer)
-+{
-+ int ret = -ENOMEM;
-+ struct xen_platform_parameters *parms;
++#if HAVE_TSO
++ if (skb_shinfo(skb)->gso_size) {
++ struct netif_extra_info *gso = (struct netif_extra_info *)
++ RING_GET_REQUEST(&np->tx, ++i);
+
-+ parms = kmalloc(sizeof(struct xen_platform_parameters), GFP_KERNEL);
-+ if (parms) {
-+ ret = HYPERVISOR_xen_version(XENVER_platform_parameters,
-+ parms);
-+ if (!ret)
-+ ret = sprintf(buffer, "%lx\n", parms->virt_start);
-+ kfree(parms);
++ if (extra)
++ extra->flags |= XEN_NETIF_EXTRA_FLAG_MORE;
++ else
++ tx->flags |= NETTXF_extra_info;
++
++ gso->u.gso.size = skb_shinfo(skb)->gso_size;
++ gso->u.gso.type = XEN_NETIF_GSO_TYPE_TCPV4;
++ gso->u.gso.pad = 0;
++ gso->u.gso.features = 0;
++
++ gso->type = XEN_NETIF_EXTRA_TYPE_GSO;
++ gso->flags = 0;
++ extra = gso;
+ }
++#endif
+
-+ return ret;
-+}
++ np->tx.req_prod_pvt = i + 1;
+
-+HYPERVISOR_ATTR_RO(virtual_start);
++ xennet_make_frags(skb, dev, tx);
++ tx->size = skb->len;
+
-+static ssize_t pagesize_show(struct hyp_sysfs_attr *attr, char *buffer)
-+{
-+ int ret;
++ RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&np->tx, notify);
++ if (notify)
++ notify_remote_via_irq(np->irq);
+
-+ ret = HYPERVISOR_xen_version(XENVER_pagesize, NULL);
-+ if (ret > 0)
-+ ret = sprintf(buffer, "%x\n", ret);
++ np->stats.tx_bytes += skb->len;
++ np->stats.tx_packets++;
++ dev->trans_start = jiffies;
+
-+ return ret;
-+}
++ /* Note: It is not safe to access skb after network_tx_buf_gc()! */
++ network_tx_buf_gc(dev);
+
-+HYPERVISOR_ATTR_RO(pagesize);
++ if (!netfront_tx_slot_available(np))
++ netif_stop_queue(dev);
+
-+/* eventually there will be several more features to export */
-+static ssize_t xen_feature_show(int index, char *buffer)
-+{
-+ int ret = -ENOMEM;
-+ struct xen_feature_info *info;
++ spin_unlock_irq(&np->tx_lock);
+
-+ info = kmalloc(sizeof(struct xen_feature_info), GFP_KERNEL);
-+ if (info) {
-+ info->submap_idx = index;
-+ ret = HYPERVISOR_xen_version(XENVER_get_features, info);
-+ if (!ret)
-+ ret = sprintf(buffer, "%d\n", info->submap);
-+ kfree(info);
-+ }
++ return 0;
+
-+ return ret;
++ drop:
++ np->stats.tx_dropped++;
++ dev_kfree_skb(skb);
++ return 0;
+}
+
-+static ssize_t writable_pt_show(struct hyp_sysfs_attr *attr, char *buffer)
++static irqreturn_t netif_int(int irq, void *dev_id, struct pt_regs *ptregs)
+{
-+ return xen_feature_show(XENFEAT_writable_page_tables, buffer);
-+}
++ struct net_device *dev = dev_id;
++ struct netfront_info *np = netdev_priv(dev);
++ unsigned long flags;
+
-+HYPERVISOR_ATTR_RO(writable_pt);
++ spin_lock_irqsave(&np->tx_lock, flags);
+
-+static struct attribute *xen_properties_attrs[] = {
-+ &capabilities_attr.attr,
-+ &changeset_attr.attr,
-+ &virtual_start_attr.attr,
-+ &pagesize_attr.attr,
-+ &writable_pt_attr.attr,
-+ NULL
-+};
++ if (likely(netfront_carrier_ok(np))) {
++ network_tx_buf_gc(dev);
++ /* Under tx_lock: protects access to rx shared-ring indexes. */
++ if (RING_HAS_UNCONSUMED_RESPONSES(&np->rx)) {
++ netfront_accelerator_call_stop_napi_irq(np, dev);
+
-+static struct attribute_group xen_properties_group = {
-+ .name = "properties",
-+ .attrs = xen_properties_attrs,
-+};
++ netif_rx_schedule(dev);
++ dev->last_rx = jiffies;
++ }
++ }
+
-+static int __init xen_properties_init(void)
-+{
-+ return sysfs_create_group(&hypervisor_subsys.kset.kobj,
-+ &xen_properties_group);
-+}
++ spin_unlock_irqrestore(&np->tx_lock, flags);
+
-+static void xen_properties_destroy(void)
-+{
-+ sysfs_remove_group(&hypervisor_subsys.kset.kobj,
-+ &xen_properties_group);
++ return IRQ_HANDLED;
+}
+
-+static int __init hyper_sysfs_init(void)
++static void xennet_move_rx_slot(struct netfront_info *np, struct sk_buff *skb,
++ grant_ref_t ref)
+{
-+ int ret;
-+
-+ if (!is_running_on_xen())
-+ return -ENODEV;
-+
-+ ret = xen_sysfs_type_init();
-+ if (ret)
-+ goto out;
-+ ret = xen_sysfs_version_init();
-+ if (ret)
-+ goto version_out;
-+ ret = xen_compilation_init();
-+ if (ret)
-+ goto comp_out;
-+ ret = xen_sysfs_uuid_init();
-+ if (ret)
-+ goto uuid_out;
-+ ret = xen_properties_init();
-+ if (!ret)
-+ goto out;
++ int new = xennet_rxidx(np->rx.req_prod_pvt);
+
-+ xen_sysfs_uuid_destroy();
-+uuid_out:
-+ xen_compilation_destroy();
-+comp_out:
-+ xen_sysfs_version_destroy();
-+version_out:
-+ xen_sysfs_type_destroy();
-+out:
-+ return ret;
++ BUG_ON(np->rx_skbs[new]);
++ np->rx_skbs[new] = skb;
++ np->grant_rx_ref[new] = ref;
++ RING_GET_REQUEST(&np->rx, np->rx.req_prod_pvt)->id = new;
++ RING_GET_REQUEST(&np->rx, np->rx.req_prod_pvt)->gref = ref;
++ np->rx.req_prod_pvt++;
+}
+
-+static void hyper_sysfs_exit(void)
++int xennet_get_extras(struct netfront_info *np,
++ struct netif_extra_info *extras, RING_IDX rp)
++
+{
-+ xen_properties_destroy();
-+ xen_compilation_destroy();
-+ xen_sysfs_uuid_destroy();
-+ xen_sysfs_version_destroy();
-+ xen_sysfs_type_destroy();
++ struct netif_extra_info *extra;
++ RING_IDX cons = np->rx.rsp_cons;
++ int err = 0;
+
-+}
++ do {
++ struct sk_buff *skb;
++ grant_ref_t ref;
+
-+module_init(hyper_sysfs_init);
-+module_exit(hyper_sysfs_exit);
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/evtchn/evtchn.c tmp-linux-2.6-xen.patch/drivers/xen/evtchn/evtchn.c
---- pristine-linux-2.6.18.2/drivers/xen/evtchn/evtchn.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/evtchn/evtchn.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,469 @@
-+/******************************************************************************
-+ * evtchn.c
-+ *
-+ * Driver for receiving and demuxing event-channel signals.
-+ *
-+ * Copyright (c) 2004-2005, K A Fraser
-+ * Multi-process extensions Copyright (c) 2004, Steven Smith
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ if (unlikely(cons + 1 == rp)) {
++ if (net_ratelimit())
++ WPRINTK("Missing extra info\n");
++ err = -EBADR;
++ break;
++ }
+
-+#include <linux/module.h>
-+#include <linux/kernel.h>
-+#include <linux/sched.h>
-+#include <linux/slab.h>
-+#include <linux/string.h>
-+#include <linux/errno.h>
-+#include <linux/fs.h>
-+#include <linux/errno.h>
-+#include <linux/miscdevice.h>
-+#include <linux/major.h>
-+#include <linux/proc_fs.h>
-+#include <linux/stat.h>
-+#include <linux/poll.h>
-+#include <linux/irq.h>
-+#include <linux/init.h>
-+#include <linux/gfp.h>
-+#include <linux/mutex.h>
-+#include <xen/evtchn.h>
-+#include <xen/public/evtchn.h>
++ extra = (struct netif_extra_info *)
++ RING_GET_RESPONSE(&np->rx, ++cons);
+
-+struct per_user_data {
-+ /* Notification ring, accessed via /dev/xen/evtchn. */
-+#define EVTCHN_RING_SIZE (PAGE_SIZE / sizeof(evtchn_port_t))
-+#define EVTCHN_RING_MASK(_i) ((_i)&(EVTCHN_RING_SIZE-1))
-+ evtchn_port_t *ring;
-+ unsigned int ring_cons, ring_prod, ring_overflow;
-+ struct mutex ring_cons_mutex; /* protect against concurrent readers */
++ if (unlikely(!extra->type ||
++ extra->type >= XEN_NETIF_EXTRA_TYPE_MAX)) {
++ if (net_ratelimit())
++ WPRINTK("Invalid extra type: %d\n",
++ extra->type);
++ err = -EINVAL;
++ } else {
++ memcpy(&extras[extra->type - 1], extra,
++ sizeof(*extra));
++ }
+
-+ /* Processes wait on this queue when ring is empty. */
-+ wait_queue_head_t evtchn_wait;
-+ struct fasync_struct *evtchn_async_queue;
-+};
++ skb = xennet_get_rx_skb(np, cons);
++ ref = xennet_get_rx_ref(np, cons);
++ xennet_move_rx_slot(np, skb, ref);
++ } while (extra->flags & XEN_NETIF_EXTRA_FLAG_MORE);
+
-+/* Who's bound to each port? */
-+static struct per_user_data *port_user[NR_EVENT_CHANNELS];
-+static spinlock_t port_user_lock;
++ np->rx.rsp_cons = cons;
++ return err;
++}
+
-+void evtchn_device_upcall(int port)
++static int xennet_get_responses(struct netfront_info *np,
++ struct netfront_rx_info *rinfo, RING_IDX rp,
++ struct sk_buff_head *list,
++ int *pages_flipped_p)
+{
-+ struct per_user_data *u;
++ int pages_flipped = *pages_flipped_p;
++ struct mmu_update *mmu;
++ struct multicall_entry *mcl;
++ struct netif_rx_response *rx = &rinfo->rx;
++ struct netif_extra_info *extras = rinfo->extras;
++ RING_IDX cons = np->rx.rsp_cons;
++ struct sk_buff *skb = xennet_get_rx_skb(np, cons);
++ grant_ref_t ref = xennet_get_rx_ref(np, cons);
++ int max = MAX_SKB_FRAGS + (rx->status <= RX_COPY_THRESHOLD);
++ int frags = 1;
++ int err = 0;
++ unsigned long ret;
+
-+ spin_lock(&port_user_lock);
++ if (rx->flags & NETRXF_extra_info) {
++ err = xennet_get_extras(np, extras, rp);
++ cons = np->rx.rsp_cons;
++ }
+
-+ mask_evtchn(port);
-+ clear_evtchn(port);
++ for (;;) {
++ unsigned long mfn;
+
-+ if ((u = port_user[port]) != NULL) {
-+ if ((u->ring_prod - u->ring_cons) < EVTCHN_RING_SIZE) {
-+ u->ring[EVTCHN_RING_MASK(u->ring_prod)] = port;
-+ if (u->ring_cons == u->ring_prod++) {
-+ wake_up_interruptible(&u->evtchn_wait);
-+ kill_fasync(&u->evtchn_async_queue,
-+ SIGIO, POLL_IN);
-+ }
-+ } else {
-+ u->ring_overflow = 1;
++ if (unlikely(rx->status < 0 ||
++ rx->offset + rx->status > PAGE_SIZE)) {
++ if (net_ratelimit())
++ WPRINTK("rx->offset: %x, size: %u\n",
++ rx->offset, rx->status);
++ xennet_move_rx_slot(np, skb, ref);
++ err = -EINVAL;
++ goto next;
+ }
-+ }
-+
-+ spin_unlock(&port_user_lock);
-+}
+
-+static ssize_t evtchn_read(struct file *file, char __user *buf,
-+ size_t count, loff_t *ppos)
-+{
-+ int rc;
-+ unsigned int c, p, bytes1 = 0, bytes2 = 0;
-+ struct per_user_data *u = file->private_data;
++ /*
++ * This definitely indicates a bug, either in this driver or in
++ * the backend driver. In future this should flag the bad
++ * situation to the system controller to reboot the backed.
++ */
++ if (ref == GRANT_INVALID_REF) {
++ if (net_ratelimit())
++ WPRINTK("Bad rx response id %d.\n", rx->id);
++ err = -EINVAL;
++ goto next;
++ }
+
-+ /* Whole number of ports. */
-+ count &= ~(sizeof(evtchn_port_t)-1);
++ if (!np->copying_receiver) {
++ /* Memory pressure, insufficient buffer
++ * headroom, ... */
++ if (!(mfn = gnttab_end_foreign_transfer_ref(ref))) {
++ if (net_ratelimit())
++ WPRINTK("Unfulfilled rx req "
++ "(id=%d, st=%d).\n",
++ rx->id, rx->status);
++ xennet_move_rx_slot(np, skb, ref);
++ err = -ENOMEM;
++ goto next;
++ }
+
-+ if (count == 0)
-+ return 0;
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* Remap the page. */
++ struct page *page =
++ skb_shinfo(skb)->frags[0].page;
++ unsigned long pfn = page_to_pfn(page);
++ void *vaddr = page_address(page);
+
-+ if (count > PAGE_SIZE)
-+ count = PAGE_SIZE;
++ mcl = np->rx_mcl + pages_flipped;
++ mmu = np->rx_mmu + pages_flipped;
+
-+ for (;;) {
-+ mutex_lock(&u->ring_cons_mutex);
++ MULTI_update_va_mapping(mcl,
++ (unsigned long)vaddr,
++ pfn_pte_ma(mfn,
++ PAGE_KERNEL),
++ 0);
++ mmu->ptr = ((maddr_t)mfn << PAGE_SHIFT)
++ | MMU_MACHPHYS_UPDATE;
++ mmu->val = pfn;
+
-+ rc = -EFBIG;
-+ if (u->ring_overflow)
-+ goto unlock_out;
++ set_phys_to_machine(pfn, mfn);
++ }
++ pages_flipped++;
++ } else {
++ ret = gnttab_end_foreign_access_ref(ref);
++ BUG_ON(!ret);
++ }
+
-+ if ((c = u->ring_cons) != (p = u->ring_prod))
-+ break;
++ gnttab_release_grant_reference(&np->gref_rx_head, ref);
+
-+ mutex_unlock(&u->ring_cons_mutex);
++ __skb_queue_tail(list, skb);
+
-+ if (file->f_flags & O_NONBLOCK)
-+ return -EAGAIN;
++next:
++ if (!(rx->flags & NETRXF_more_data))
++ break;
+
-+ rc = wait_event_interruptible(
-+ u->evtchn_wait, u->ring_cons != u->ring_prod);
-+ if (rc)
-+ return rc;
-+ }
++ if (cons + frags == rp) {
++ if (net_ratelimit())
++ WPRINTK("Need more frags\n");
++ err = -ENOENT;
++ break;
++ }
+
-+ /* Byte lengths of two chunks. Chunk split (if any) is at ring wrap. */
-+ if (((c ^ p) & EVTCHN_RING_SIZE) != 0) {
-+ bytes1 = (EVTCHN_RING_SIZE - EVTCHN_RING_MASK(c)) *
-+ sizeof(evtchn_port_t);
-+ bytes2 = EVTCHN_RING_MASK(p) * sizeof(evtchn_port_t);
-+ } else {
-+ bytes1 = (p - c) * sizeof(evtchn_port_t);
-+ bytes2 = 0;
++ rx = RING_GET_RESPONSE(&np->rx, cons + frags);
++ skb = xennet_get_rx_skb(np, cons + frags);
++ ref = xennet_get_rx_ref(np, cons + frags);
++ frags++;
+ }
-+
-+ /* Truncate chunks according to caller's maximum byte count. */
-+ if (bytes1 > count) {
-+ bytes1 = count;
-+ bytes2 = 0;
-+ } else if ((bytes1 + bytes2) > count) {
-+ bytes2 = count - bytes1;
++
++ if (unlikely(frags > max)) {
++ if (net_ratelimit())
++ WPRINTK("Too many frags\n");
++ err = -E2BIG;
+ }
+
-+ rc = -EFAULT;
-+ if (copy_to_user(buf, &u->ring[EVTCHN_RING_MASK(c)], bytes1) ||
-+ ((bytes2 != 0) &&
-+ copy_to_user(&buf[bytes1], &u->ring[0], bytes2)))
-+ goto unlock_out;
++ if (unlikely(err))
++ np->rx.rsp_cons = cons + frags;
+
-+ u->ring_cons += (bytes1 + bytes2) / sizeof(evtchn_port_t);
-+ rc = bytes1 + bytes2;
++ *pages_flipped_p = pages_flipped;
+
-+ unlock_out:
-+ mutex_unlock(&u->ring_cons_mutex);
-+ return rc;
++ return err;
+}
+
-+static ssize_t evtchn_write(struct file *file, const char __user *buf,
-+ size_t count, loff_t *ppos)
++static RING_IDX xennet_fill_frags(struct netfront_info *np,
++ struct sk_buff *skb,
++ struct sk_buff_head *list)
+{
-+ int rc, i;
-+ evtchn_port_t *kbuf = (evtchn_port_t *)__get_free_page(GFP_KERNEL);
-+ struct per_user_data *u = file->private_data;
-+
-+ if (kbuf == NULL)
-+ return -ENOMEM;
-+
-+ /* Whole number of ports. */
-+ count &= ~(sizeof(evtchn_port_t)-1);
-+
-+ rc = 0;
-+ if (count == 0)
-+ goto out;
++ struct skb_shared_info *shinfo = skb_shinfo(skb);
++ int nr_frags = shinfo->nr_frags;
++ RING_IDX cons = np->rx.rsp_cons;
++ skb_frag_t *frag = shinfo->frags + nr_frags;
++ struct sk_buff *nskb;
+
-+ if (count > PAGE_SIZE)
-+ count = PAGE_SIZE;
++ while ((nskb = __skb_dequeue(list))) {
++ struct netif_rx_response *rx =
++ RING_GET_RESPONSE(&np->rx, ++cons);
+
-+ rc = -EFAULT;
-+ if (copy_from_user(kbuf, buf, count) != 0)
-+ goto out;
++ frag->page = skb_shinfo(nskb)->frags[0].page;
++ frag->page_offset = rx->offset;
++ frag->size = rx->status;
+
-+ spin_lock_irq(&port_user_lock);
-+ for (i = 0; i < (count/sizeof(evtchn_port_t)); i++)
-+ if ((kbuf[i] < NR_EVENT_CHANNELS) && (port_user[kbuf[i]] == u))
-+ unmask_evtchn(kbuf[i]);
-+ spin_unlock_irq(&port_user_lock);
++ skb->data_len += rx->status;
+
-+ rc = count;
++ skb_shinfo(nskb)->nr_frags = 0;
++ kfree_skb(nskb);
+
-+ out:
-+ free_page((unsigned long)kbuf);
-+ return rc;
-+}
++ frag++;
++ nr_frags++;
++ }
+
-+static void evtchn_bind_to_user(struct per_user_data *u, int port)
-+{
-+ spin_lock_irq(&port_user_lock);
-+ BUG_ON(port_user[port] != NULL);
-+ port_user[port] = u;
-+ unmask_evtchn(port);
-+ spin_unlock_irq(&port_user_lock);
++ shinfo->nr_frags = nr_frags;
++ return cons;
+}
+
-+static long evtchn_ioctl(struct file *file,
-+ unsigned int cmd, unsigned long arg)
++static int xennet_set_skb_gso(struct sk_buff *skb,
++ struct netif_extra_info *gso)
+{
-+ int rc;
-+ struct per_user_data *u = file->private_data;
-+ void __user *uarg = (void __user *) arg;
-+
-+ switch (cmd) {
-+ case IOCTL_EVTCHN_BIND_VIRQ: {
-+ struct ioctl_evtchn_bind_virq bind;
-+ struct evtchn_bind_virq bind_virq;
++ if (!gso->u.gso.size) {
++ if (net_ratelimit())
++ WPRINTK("GSO size must not be zero.\n");
++ return -EINVAL;
++ }
+
-+ rc = -EFAULT;
-+ if (copy_from_user(&bind, uarg, sizeof(bind)))
-+ break;
++ /* Currently only TCPv4 S.O. is supported. */
++ if (gso->u.gso.type != XEN_NETIF_GSO_TYPE_TCPV4) {
++ if (net_ratelimit())
++ WPRINTK("Bad GSO type %d.\n", gso->u.gso.type);
++ return -EINVAL;
++ }
+
-+ bind_virq.virq = bind.virq;
-+ bind_virq.vcpu = 0;
-+ rc = HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
-+ &bind_virq);
-+ if (rc != 0)
-+ break;
++#if HAVE_TSO
++ skb_shinfo(skb)->gso_size = gso->u.gso.size;
++#if HAVE_GSO
++ skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4;
+
-+ rc = bind_virq.port;
-+ evtchn_bind_to_user(u, rc);
-+ break;
-+ }
++ /* Header must be checked, and gso_segs computed. */
++ skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY;
++#endif
++ skb_shinfo(skb)->gso_segs = 0;
+
-+ case IOCTL_EVTCHN_BIND_INTERDOMAIN: {
-+ struct ioctl_evtchn_bind_interdomain bind;
-+ struct evtchn_bind_interdomain bind_interdomain;
++ return 0;
++#else
++ if (net_ratelimit())
++ WPRINTK("GSO unsupported by this kernel.\n");
++ return -EINVAL;
++#endif
++}
+
-+ rc = -EFAULT;
-+ if (copy_from_user(&bind, uarg, sizeof(bind)))
-+ break;
++static int netif_poll(struct net_device *dev, int *pbudget)
++{
++ struct netfront_info *np = netdev_priv(dev);
++ struct sk_buff *skb;
++ struct netfront_rx_info rinfo;
++ struct netif_rx_response *rx = &rinfo.rx;
++ struct netif_extra_info *extras = rinfo.extras;
++ RING_IDX i, rp;
++ struct multicall_entry *mcl;
++ int work_done, budget, more_to_do = 1, accel_more_to_do = 1;
++ struct sk_buff_head rxq;
++ struct sk_buff_head errq;
++ struct sk_buff_head tmpq;
++ unsigned long flags;
++ unsigned int len;
++ int pages_flipped = 0;
++ int err;
+
-+ bind_interdomain.remote_dom = bind.remote_domain;
-+ bind_interdomain.remote_port = bind.remote_port;
-+ rc = HYPERVISOR_event_channel_op(EVTCHNOP_bind_interdomain,
-+ &bind_interdomain);
-+ if (rc != 0)
-+ break;
++ spin_lock(&np->rx_lock); /* no need for spin_lock_bh() in ->poll() */
+
-+ rc = bind_interdomain.local_port;
-+ evtchn_bind_to_user(u, rc);
-+ break;
++ if (unlikely(!netfront_carrier_ok(np))) {
++ spin_unlock(&np->rx_lock);
++ return 0;
+ }
+
-+ case IOCTL_EVTCHN_BIND_UNBOUND_PORT: {
-+ struct ioctl_evtchn_bind_unbound_port bind;
-+ struct evtchn_alloc_unbound alloc_unbound;
++ skb_queue_head_init(&rxq);
++ skb_queue_head_init(&errq);
++ skb_queue_head_init(&tmpq);
+
-+ rc = -EFAULT;
-+ if (copy_from_user(&bind, uarg, sizeof(bind)))
-+ break;
++ if ((budget = *pbudget) > dev->quota)
++ budget = dev->quota;
++ rp = np->rx.sring->rsp_prod;
++ rmb(); /* Ensure we see queued responses up to 'rp'. */
+
-+ alloc_unbound.dom = DOMID_SELF;
-+ alloc_unbound.remote_dom = bind.remote_domain;
-+ rc = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
-+ &alloc_unbound);
-+ if (rc != 0)
-+ break;
++ i = np->rx.rsp_cons;
++ work_done = 0;
++ while ((i != rp) && (work_done < budget)) {
++ memcpy(rx, RING_GET_RESPONSE(&np->rx, i), sizeof(*rx));
++ memset(extras, 0, sizeof(rinfo.extras));
+
-+ rc = alloc_unbound.port;
-+ evtchn_bind_to_user(u, rc);
-+ break;
-+ }
++ err = xennet_get_responses(np, &rinfo, rp, &tmpq,
++ &pages_flipped);
+
-+ case IOCTL_EVTCHN_UNBIND: {
-+ struct ioctl_evtchn_unbind unbind;
-+ struct evtchn_close close;
-+ int ret;
++ if (unlikely(err)) {
++err:
++ while ((skb = __skb_dequeue(&tmpq)))
++ __skb_queue_tail(&errq, skb);
++ np->stats.rx_errors++;
++ i = np->rx.rsp_cons;
++ continue;
++ }
+
-+ rc = -EFAULT;
-+ if (copy_from_user(&unbind, uarg, sizeof(unbind)))
-+ break;
++ skb = __skb_dequeue(&tmpq);
+
-+ rc = -EINVAL;
-+ if (unbind.port >= NR_EVENT_CHANNELS)
-+ break;
++ if (extras[XEN_NETIF_EXTRA_TYPE_GSO - 1].type) {
++ struct netif_extra_info *gso;
++ gso = &extras[XEN_NETIF_EXTRA_TYPE_GSO - 1];
+
-+ spin_lock_irq(&port_user_lock);
-+
-+ rc = -ENOTCONN;
-+ if (port_user[unbind.port] != u) {
-+ spin_unlock_irq(&port_user_lock);
-+ break;
++ if (unlikely(xennet_set_skb_gso(skb, gso))) {
++ __skb_queue_head(&tmpq, skb);
++ np->rx.rsp_cons += skb_queue_len(&tmpq);
++ goto err;
++ }
+ }
+
-+ port_user[unbind.port] = NULL;
-+ mask_evtchn(unbind.port);
++ NETFRONT_SKB_CB(skb)->page = skb_shinfo(skb)->frags[0].page;
++ NETFRONT_SKB_CB(skb)->offset = rx->offset;
+
-+ spin_unlock_irq(&port_user_lock);
++ len = rx->status;
++ if (len > RX_COPY_THRESHOLD)
++ len = RX_COPY_THRESHOLD;
++ skb_put(skb, len);
+
-+ close.port = unbind.port;
-+ ret = HYPERVISOR_event_channel_op(EVTCHNOP_close, &close);
-+ BUG_ON(ret);
++ if (rx->status > len) {
++ skb_shinfo(skb)->frags[0].page_offset =
++ rx->offset + len;
++ skb_shinfo(skb)->frags[0].size = rx->status - len;
++ skb->data_len = rx->status - len;
++ } else {
++ skb_shinfo(skb)->frags[0].page = NULL;
++ skb_shinfo(skb)->nr_frags = 0;
++ }
+
-+ rc = 0;
-+ break;
-+ }
++ i = xennet_fill_frags(np, skb, &tmpq);
+
-+ case IOCTL_EVTCHN_NOTIFY: {
-+ struct ioctl_evtchn_notify notify;
++ /*
++ * Truesize must approximates the size of true data plus
++ * any supervisor overheads. Adding hypervisor overheads
++ * has been shown to significantly reduce achievable
++ * bandwidth with the default receive buffer size. It is
++ * therefore not wise to account for it here.
++ *
++ * After alloc_skb(RX_COPY_THRESHOLD), truesize is set to
++ * RX_COPY_THRESHOLD + the supervisor overheads. Here, we
++ * add the size of the data pulled in xennet_fill_frags().
++ *
++ * We also adjust for any unused space in the main data
++ * area by subtracting (RX_COPY_THRESHOLD - len). This is
++ * especially important with drivers which split incoming
++ * packets into header and data, using only 66 bytes of
++ * the main data area (see the e1000 driver for example.)
++ * On such systems, without this last adjustement, our
++ * achievable receive throughout using the standard receive
++ * buffer size was cut by 25%(!!!).
++ */
++ skb->truesize += skb->data_len - (RX_COPY_THRESHOLD - len);
++ skb->len += skb->data_len;
+
-+ rc = -EFAULT;
-+ if (copy_from_user(¬ify, uarg, sizeof(notify)))
-+ break;
++ /*
++ * Old backends do not assert data_validated but we
++ * can infer it from csum_blank so test both flags.
++ */
++ if (rx->flags & (NETRXF_data_validated|NETRXF_csum_blank))
++ skb->ip_summed = CHECKSUM_UNNECESSARY;
++ else
++ skb->ip_summed = CHECKSUM_NONE;
++#ifdef CONFIG_XEN
++ skb->proto_data_valid = (skb->ip_summed != CHECKSUM_NONE);
++ skb->proto_csum_blank = !!(rx->flags & NETRXF_csum_blank);
++#endif
++ np->stats.rx_packets++;
++ np->stats.rx_bytes += skb->len;
+
-+ if (notify.port >= NR_EVENT_CHANNELS) {
-+ rc = -EINVAL;
-+ } else if (port_user[notify.port] != u) {
-+ rc = -ENOTCONN;
-+ } else {
-+ notify_remote_via_evtchn(notify.port);
-+ rc = 0;
-+ }
-+ break;
-+ }
++ __skb_queue_tail(&rxq, skb);
+
-+ case IOCTL_EVTCHN_RESET: {
-+ /* Initialise the ring to empty. Clear errors. */
-+ mutex_lock(&u->ring_cons_mutex);
-+ spin_lock_irq(&port_user_lock);
-+ u->ring_cons = u->ring_prod = u->ring_overflow = 0;
-+ spin_unlock_irq(&port_user_lock);
-+ mutex_unlock(&u->ring_cons_mutex);
-+ rc = 0;
-+ break;
++ np->rx.rsp_cons = ++i;
++ work_done++;
+ }
+
-+ default:
-+ rc = -ENOSYS;
-+ break;
++ if (pages_flipped) {
++ /* Some pages are no longer absent... */
++ balloon_update_driver_allowance(-pages_flipped);
++
++ /* Do all the remapping work and M2P updates. */
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ mcl = np->rx_mcl + pages_flipped;
++ mcl->op = __HYPERVISOR_mmu_update;
++ mcl->args[0] = (unsigned long)np->rx_mmu;
++ mcl->args[1] = pages_flipped;
++ mcl->args[2] = 0;
++ mcl->args[3] = DOMID_SELF;
++ err = HYPERVISOR_multicall_check(np->rx_mcl,
++ pages_flipped + 1,
++ NULL);
++ BUG_ON(err);
++ }
+ }
+
-+ return rc;
-+}
++ while ((skb = __skb_dequeue(&errq)))
++ kfree_skb(skb);
+
-+static unsigned int evtchn_poll(struct file *file, poll_table *wait)
-+{
-+ unsigned int mask = POLLOUT | POLLWRNORM;
-+ struct per_user_data *u = file->private_data;
++ while ((skb = __skb_dequeue(&rxq)) != NULL) {
++ struct page *page = NETFRONT_SKB_CB(skb)->page;
++ void *vaddr = page_address(page);
++ unsigned offset = NETFRONT_SKB_CB(skb)->offset;
+
-+ poll_wait(file, &u->evtchn_wait, wait);
-+ if (u->ring_cons != u->ring_prod)
-+ mask |= POLLIN | POLLRDNORM;
-+ if (u->ring_overflow)
-+ mask = POLLERR;
-+ return mask;
-+}
++ memcpy(skb->data, vaddr + offset, skb_headlen(skb));
+
-+static int evtchn_fasync(int fd, struct file *filp, int on)
-+{
-+ struct per_user_data *u = filp->private_data;
-+ return fasync_helper(fd, filp, on, &u->evtchn_async_queue);
-+}
++ if (page != skb_shinfo(skb)->frags[0].page)
++ __free_page(page);
+
-+static int evtchn_open(struct inode *inode, struct file *filp)
-+{
-+ struct per_user_data *u;
++ /* Ethernet work: Delayed to here as it peeks the header. */
++ skb->protocol = eth_type_trans(skb, dev);
+
-+ if ((u = kmalloc(sizeof(*u), GFP_KERNEL)) == NULL)
-+ return -ENOMEM;
++ /* Pass it up. */
++ netif_receive_skb(skb);
++ dev->last_rx = jiffies;
++ }
+
-+ memset(u, 0, sizeof(*u));
-+ init_waitqueue_head(&u->evtchn_wait);
++ /* If we get a callback with very few responses, reduce fill target. */
++ /* NB. Note exponential increase, linear decrease. */
++ if (((np->rx.req_prod_pvt - np->rx.sring->rsp_prod) >
++ ((3*np->rx_target) / 4)) &&
++ (--np->rx_target < np->rx_min_target))
++ np->rx_target = np->rx_min_target;
+
-+ u->ring = (evtchn_port_t *)__get_free_page(GFP_KERNEL);
-+ if (u->ring == NULL) {
-+ kfree(u);
-+ return -ENOMEM;
++ network_alloc_rx_buffers(dev);
++
++ if (work_done < budget) {
++ /* there's some spare capacity, try the accelerated path */
++ int accel_budget = budget - work_done;
++ int accel_budget_start = accel_budget;
++
++ if (np->accel_vif_state.hooks) {
++ accel_more_to_do =
++ np->accel_vif_state.hooks->netdev_poll
++ (dev, &accel_budget);
++ work_done += (accel_budget_start - accel_budget);
++ } else
++ accel_more_to_do = 0;
+ }
+
-+ mutex_init(&u->ring_cons_mutex);
++ *pbudget -= work_done;
++ dev->quota -= work_done;
+
-+ filp->private_data = u;
++ if (work_done < budget) {
++ local_irq_save(flags);
+
-+ return 0;
++ RING_FINAL_CHECK_FOR_RESPONSES(&np->rx, more_to_do);
++
++ if (!more_to_do && !accel_more_to_do &&
++ np->accel_vif_state.hooks) {
++ /*
++ * Slow path has nothing more to do, see if
++ * fast path is likewise
++ */
++ accel_more_to_do =
++ np->accel_vif_state.hooks->start_napi_irq(dev);
++ }
++
++ if (!more_to_do && !accel_more_to_do)
++ __netif_rx_complete(dev);
++
++ local_irq_restore(flags);
++ }
++
++ spin_unlock(&np->rx_lock);
++
++ return more_to_do | accel_more_to_do;
+}
+
-+static int evtchn_release(struct inode *inode, struct file *filp)
++static void netif_release_tx_bufs(struct netfront_info *np)
+{
++ struct sk_buff *skb;
+ int i;
-+ struct per_user_data *u = filp->private_data;
-+ struct evtchn_close close;
+
-+ spin_lock_irq(&port_user_lock);
++ for (i = 1; i <= NET_TX_RING_SIZE; i++) {
++ if ((unsigned long)np->tx_skbs[i] < PAGE_OFFSET)
++ continue;
+
-+ free_page((unsigned long)u->ring);
++ skb = np->tx_skbs[i];
++ gnttab_end_foreign_access_ref(np->grant_tx_ref[i]);
++ gnttab_release_grant_reference(
++ &np->gref_tx_head, np->grant_tx_ref[i]);
++ np->grant_tx_ref[i] = GRANT_INVALID_REF;
++ add_id_to_freelist(np->tx_skbs, i);
++ dev_kfree_skb_irq(skb);
++ }
++}
+
-+ for (i = 0; i < NR_EVENT_CHANNELS; i++) {
-+ int ret;
-+ if (port_user[i] != u)
-+ continue;
++static void netif_release_rx_bufs_flip(struct netfront_info *np)
++{
++ struct mmu_update *mmu = np->rx_mmu;
++ struct multicall_entry *mcl = np->rx_mcl;
++ struct sk_buff_head free_list;
++ struct sk_buff *skb;
++ unsigned long mfn;
++ int xfer = 0, noxfer = 0, unused = 0;
++ int id, ref, rc;
+
-+ port_user[i] = NULL;
-+ mask_evtchn(i);
++ skb_queue_head_init(&free_list);
+
-+ close.port = i;
-+ ret = HYPERVISOR_event_channel_op(EVTCHNOP_close, &close);
-+ BUG_ON(ret);
-+ }
++ spin_lock_bh(&np->rx_lock);
+
-+ spin_unlock_irq(&port_user_lock);
++ for (id = 0; id < NET_RX_RING_SIZE; id++) {
++ if ((ref = np->grant_rx_ref[id]) == GRANT_INVALID_REF) {
++ unused++;
++ continue;
++ }
+
-+ kfree(u);
++ skb = np->rx_skbs[id];
++ mfn = gnttab_end_foreign_transfer_ref(ref);
++ gnttab_release_grant_reference(&np->gref_rx_head, ref);
++ np->grant_rx_ref[id] = GRANT_INVALID_REF;
++ add_id_to_freelist(np->rx_skbs, id);
+
-+ return 0;
-+}
++ if (0 == mfn) {
++ struct page *page = skb_shinfo(skb)->frags[0].page;
++ balloon_release_driver_page(page);
++ skb_shinfo(skb)->nr_frags = 0;
++ dev_kfree_skb(skb);
++ noxfer++;
++ continue;
++ }
+
-+static const struct file_operations evtchn_fops = {
-+ .owner = THIS_MODULE,
-+ .read = evtchn_read,
-+ .write = evtchn_write,
-+ .unlocked_ioctl = evtchn_ioctl,
-+ .poll = evtchn_poll,
-+ .fasync = evtchn_fasync,
-+ .open = evtchn_open,
-+ .release = evtchn_release,
-+};
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* Remap the page. */
++ struct page *page = skb_shinfo(skb)->frags[0].page;
++ unsigned long pfn = page_to_pfn(page);
++ void *vaddr = page_address(page);
+
-+static struct miscdevice evtchn_miscdev = {
-+ .minor = MISC_DYNAMIC_MINOR,
-+ .name = "evtchn",
-+ .fops = &evtchn_fops,
-+};
++ MULTI_update_va_mapping(mcl, (unsigned long)vaddr,
++ pfn_pte_ma(mfn, PAGE_KERNEL),
++ 0);
++ mcl++;
++ mmu->ptr = ((maddr_t)mfn << PAGE_SHIFT)
++ | MMU_MACHPHYS_UPDATE;
++ mmu->val = pfn;
++ mmu++;
+
-+static int __init evtchn_init(void)
-+{
-+ int err;
++ set_phys_to_machine(pfn, mfn);
++ }
++ __skb_queue_tail(&free_list, skb);
++ xfer++;
++ }
+
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++ DPRINTK("%s: %d xfer, %d noxfer, %d unused\n",
++ __FUNCTION__, xfer, noxfer, unused);
+
-+ spin_lock_init(&port_user_lock);
-+ memset(port_user, 0, sizeof(port_user));
++ if (xfer) {
++ /* Some pages are no longer absent... */
++ balloon_update_driver_allowance(-xfer);
+
-+ /* Create '/dev/misc/evtchn'. */
-+ err = misc_register(&evtchn_miscdev);
-+ if (err != 0) {
-+ printk(KERN_ALERT "Could not register /dev/misc/evtchn\n");
-+ return err;
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* Do all the remapping work and M2P updates. */
++ mcl->op = __HYPERVISOR_mmu_update;
++ mcl->args[0] = (unsigned long)np->rx_mmu;
++ mcl->args[1] = mmu - np->rx_mmu;
++ mcl->args[2] = 0;
++ mcl->args[3] = DOMID_SELF;
++ mcl++;
++ rc = HYPERVISOR_multicall_check(
++ np->rx_mcl, mcl - np->rx_mcl, NULL);
++ BUG_ON(rc);
++ }
+ }
+
-+ printk("Event-channel device installed.\n");
++ while ((skb = __skb_dequeue(&free_list)) != NULL)
++ dev_kfree_skb(skb);
+
-+ return 0;
++ spin_unlock_bh(&np->rx_lock);
+}
+
-+static void evtchn_cleanup(void)
++static void netif_release_rx_bufs_copy(struct netfront_info *np)
+{
-+ misc_deregister(&evtchn_miscdev);
-+}
++ struct sk_buff *skb;
++ int i, ref;
++ int busy = 0, inuse = 0;
+
-+module_init(evtchn_init);
-+module_exit(evtchn_cleanup);
++ spin_lock_bh(&np->rx_lock);
+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/evtchn/Makefile tmp-linux-2.6-xen.patch/drivers/xen/evtchn/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/evtchn/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/evtchn/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,2 @@
++ for (i = 0; i < NET_RX_RING_SIZE; i++) {
++ ref = np->grant_rx_ref[i];
+
-+obj-y := evtchn.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/fbfront/Makefile tmp-linux-2.6-xen.patch/drivers/xen/fbfront/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/fbfront/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/fbfront/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,2 @@
-+obj-$(CONFIG_XEN_FRAMEBUFFER) := xenfb.o
-+obj-$(CONFIG_XEN_KEYBOARD) += xenkbd.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/fbfront/xenfb.c tmp-linux-2.6-xen.patch/drivers/xen/fbfront/xenfb.c
---- pristine-linux-2.6.18.2/drivers/xen/fbfront/xenfb.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/fbfront/xenfb.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,753 @@
-+/*
-+ * linux/drivers/video/xenfb.c -- Xen para-virtual frame buffer device
-+ *
-+ * Copyright (C) 2005-2006 Anthony Liguori <aliguori at us.ibm.com>
-+ * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster <armbru at redhat.com>
-+ *
-+ * Based on linux/drivers/video/q40fb.c
-+ *
-+ * This file is subject to the terms and conditions of the GNU General Public
-+ * License. See the file COPYING in the main directory of this archive for
-+ * more details.
-+ */
++ if (ref == GRANT_INVALID_REF)
++ continue;
+
-+/*
-+ * TODO:
-+ *
-+ * Switch to grant tables when they become capable of dealing with the
-+ * frame buffer.
-+ */
++ inuse++;
+
-+#include <linux/kernel.h>
-+#include <linux/errno.h>
-+#include <linux/fb.h>
-+#include <linux/module.h>
-+#include <linux/vmalloc.h>
-+#include <linux/mm.h>
-+#include <linux/mutex.h>
-+#include <asm/hypervisor.h>
-+#include <xen/evtchn.h>
-+#include <xen/interface/io/fbif.h>
-+#include <xen/interface/io/protocols.h>
-+#include <xen/xenbus.h>
-+#include <linux/kthread.h>
++ skb = np->rx_skbs[i];
+
-+struct xenfb_mapping
-+{
-+ struct list_head link;
-+ struct vm_area_struct *vma;
-+ atomic_t map_refs;
-+ int faults;
-+ struct xenfb_info *info;
-+};
++ if (!gnttab_end_foreign_access_ref(ref))
++ {
++ busy++;
++ continue;
++ }
+
-+struct xenfb_info
-+{
-+ struct task_struct *kthread;
-+ wait_queue_head_t wq;
++ gnttab_release_grant_reference(&np->gref_rx_head, ref);
++ np->grant_rx_ref[i] = GRANT_INVALID_REF;
++ add_id_to_freelist(np->rx_skbs, i);
+
-+ unsigned char *fb;
-+ struct fb_info *fb_info;
-+ struct timer_list refresh;
-+ int dirty;
-+ int x1, y1, x2, y2; /* dirty rectangle,
-+ protected by dirty_lock */
-+ spinlock_t dirty_lock;
-+ struct mutex mm_lock;
-+ int nr_pages;
-+ struct page **pages;
-+ struct list_head mappings; /* protected by mm_lock */
++ dev_kfree_skb(skb);
++ }
+
-+ int irq;
-+ struct xenfb_page *page;
-+ unsigned long *mfns;
-+ int update_wanted; /* XENFB_TYPE_UPDATE wanted */
++ if (busy)
++ DPRINTK("%s: Unable to release %d of %d inuse grant references out of %ld total.\n",
++ __FUNCTION__, busy, inuse, NET_RX_RING_SIZE);
+
-+ struct xenbus_device *xbdev;
-+};
++ spin_unlock_bh(&np->rx_lock);
++}
+
-+/*
-+ * How the locks work together
-+ *
-+ * There are two locks: spinlock dirty_lock protecting the dirty
-+ * rectangle, and mutex mm_lock protecting mappings.
-+ *
-+ * The problem is that dirty rectangle and mappings aren't
-+ * independent: the dirty rectangle must cover all faulted pages in
-+ * mappings. We need to prove that our locking maintains this
-+ * invariant.
-+ *
-+ * There are several kinds of critical regions:
-+ *
-+ * 1. Holding only dirty_lock: xenfb_refresh(). May run in
-+ * interrupts. Extends the dirty rectangle. Trivially preserves
-+ * invariant.
-+ *
-+ * 2. Holding only mm_lock: xenfb_mmap() and xenfb_vm_close(). Touch
-+ * only mappings. The former creates unfaulted pages. Preserves
-+ * invariant. The latter removes pages. Preserves invariant.
-+ *
-+ * 3. Holding both locks: xenfb_vm_nopage(). Extends the dirty
-+ * rectangle and updates mappings consistently. Preserves
-+ * invariant.
-+ *
-+ * 4. The ugliest one: xenfb_update_screen(). Clear the dirty
-+ * rectangle and update mappings consistently.
-+ *
-+ * We can't simply hold both locks, because zap_page_range() cannot
-+ * be called with a spinlock held.
-+ *
-+ * Therefore, we first clear the dirty rectangle with both locks
-+ * held. Then we unlock dirty_lock and update the mappings.
-+ * Critical regions that hold only dirty_lock may interfere with
-+ * that. This can only be region 1: xenfb_refresh(). But that
-+ * just extends the dirty rectangle, which can't harm the
-+ * invariant.
-+ *
-+ * But FIXME: the invariant is too weak. It misses that the fault
-+ * record in mappings must be consistent with the mapping of pages in
-+ * the associated address space! do_no_page() updates the PTE after
-+ * xenfb_vm_nopage() returns, i.e. outside the critical region. This
-+ * allows the following race:
-+ *
-+ * X writes to some address in the Xen frame buffer
-+ * Fault - call do_no_page()
-+ * call xenfb_vm_nopage()
-+ * grab mm_lock
-+ * map->faults++;
-+ * release mm_lock
-+ * return back to do_no_page()
-+ * (preempted, or SMP)
-+ * Xen worker thread runs.
-+ * grab mm_lock
-+ * look at mappings
-+ * find this mapping, zaps its pages (but page not in pte yet)
-+ * clear map->faults
-+ * releases mm_lock
-+ * (back to X process)
-+ * put page in X's pte
-+ *
-+ * Oh well, we wont be updating the writes to this page anytime soon.
-+ */
++static int network_close(struct net_device *dev)
++{
++ struct netfront_info *np = netdev_priv(dev);
++ netif_stop_queue(np->netdev);
++ return 0;
++}
+
-+static int xenfb_fps = 20;
-+static unsigned long xenfb_mem_len = XENFB_WIDTH * XENFB_HEIGHT * XENFB_DEPTH / 8;
+
-+static int xenfb_remove(struct xenbus_device *);
-+static void xenfb_init_shared_page(struct xenfb_info *);
-+static int xenfb_connect_backend(struct xenbus_device *, struct xenfb_info *);
-+static void xenfb_disconnect_backend(struct xenfb_info *);
++static struct net_device_stats *network_get_stats(struct net_device *dev)
++{
++ struct netfront_info *np = netdev_priv(dev);
+
-+static void xenfb_do_update(struct xenfb_info *info,
-+ int x, int y, int w, int h)
++ netfront_accelerator_call_get_stats(np, dev);
++ return &np->stats;
++}
++
++static int xennet_set_mac_address(struct net_device *dev, void *p)
+{
-+ union xenfb_out_event event;
-+ __u32 prod;
++ struct netfront_info *np = netdev_priv(dev);
++ struct sockaddr *addr = p;
+
-+ event.type = XENFB_TYPE_UPDATE;
-+ event.update.x = x;
-+ event.update.y = y;
-+ event.update.width = w;
-+ event.update.height = h;
++ if (netif_running(dev))
++ return -EBUSY;
+
-+ prod = info->page->out_prod;
-+ /* caller ensures !xenfb_queue_full() */
-+ mb(); /* ensure ring space available */
-+ XENFB_OUT_RING_REF(info->page, prod) = event;
-+ wmb(); /* ensure ring contents visible */
-+ info->page->out_prod = prod + 1;
++ if (!is_valid_ether_addr(addr->sa_data))
++ return -EADDRNOTAVAIL;
+
-+ notify_remote_via_irq(info->irq);
++ memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
++ memcpy(np->mac, addr->sa_data, ETH_ALEN);
++
++ return 0;
+}
+
-+static int xenfb_queue_full(struct xenfb_info *info)
++static int xennet_change_mtu(struct net_device *dev, int mtu)
+{
-+ __u32 cons, prod;
++ int max = xennet_can_sg(dev) ? 65535 - ETH_HLEN : ETH_DATA_LEN;
+
-+ prod = info->page->out_prod;
-+ cons = info->page->out_cons;
-+ return prod - cons == XENFB_OUT_RING_LEN;
++ if (mtu > max)
++ return -EINVAL;
++ dev->mtu = mtu;
++ return 0;
+}
+
-+static void xenfb_update_screen(struct xenfb_info *info)
++static int xennet_set_sg(struct net_device *dev, u32 data)
+{
-+ unsigned long flags;
-+ int y1, y2, x1, x2;
-+ struct xenfb_mapping *map;
++ if (data) {
++ struct netfront_info *np = netdev_priv(dev);
++ int val;
+
-+ if (!info->update_wanted)
-+ return;
-+ if (xenfb_queue_full(info))
-+ return;
++ if (xenbus_scanf(XBT_NIL, np->xbdev->otherend, "feature-sg",
++ "%d", &val) < 0)
++ val = 0;
++ if (!val)
++ return -ENOSYS;
++ } else if (dev->mtu > ETH_DATA_LEN)
++ dev->mtu = ETH_DATA_LEN;
+
-+ mutex_lock(&info->mm_lock);
++ return ethtool_op_set_sg(dev, data);
++}
+
-+ spin_lock_irqsave(&info->dirty_lock, flags);
-+ y1 = info->y1;
-+ y2 = info->y2;
-+ x1 = info->x1;
-+ x2 = info->x2;
-+ info->x1 = info->y1 = INT_MAX;
-+ info->x2 = info->y2 = 0;
-+ spin_unlock_irqrestore(&info->dirty_lock, flags);
++static int xennet_set_tso(struct net_device *dev, u32 data)
++{
++ if (data) {
++ struct netfront_info *np = netdev_priv(dev);
++ int val;
+
-+ list_for_each_entry(map, &info->mappings, link) {
-+ if (!map->faults)
-+ continue;
-+ zap_page_range(map->vma, map->vma->vm_start,
-+ map->vma->vm_end - map->vma->vm_start, NULL);
-+ map->faults = 0;
++ if (xenbus_scanf(XBT_NIL, np->xbdev->otherend,
++ "feature-gso-tcpv4", "%d", &val) < 0)
++ val = 0;
++ if (!val)
++ return -ENOSYS;
+ }
+
-+ mutex_unlock(&info->mm_lock);
-+
-+ xenfb_do_update(info, x1, y1, x2 - x1, y2 - y1);
++ return ethtool_op_set_tso(dev, data);
+}
+
-+static int xenfb_thread(void *data)
++static void xennet_set_features(struct net_device *dev)
+{
-+ struct xenfb_info *info = data;
++ dev_disable_gso_features(dev);
++ xennet_set_sg(dev, 0);
+
-+ while (!kthread_should_stop()) {
-+ if (info->dirty) {
-+ info->dirty = 0;
-+ xenfb_update_screen(info);
-+ }
-+ wait_event_interruptible(info->wq,
-+ kthread_should_stop() || info->dirty);
-+ try_to_freeze();
-+ }
-+ return 0;
++ /* We need checksum offload to enable scatter/gather and TSO. */
++ if (!(dev->features & NETIF_F_IP_CSUM))
++ return;
++
++ if (xennet_set_sg(dev, 1))
++ return;
++
++ /* Before 2.6.9 TSO seems to be unreliable so do not enable it
++ * on older kernels.
++ */
++ if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,9))
++ xennet_set_tso(dev, 1);
+}
+
-+static int xenfb_setcolreg(unsigned regno, unsigned red, unsigned green,
-+ unsigned blue, unsigned transp,
-+ struct fb_info *info)
++static int network_connect(struct net_device *dev)
+{
-+ u32 v;
++ struct netfront_info *np = netdev_priv(dev);
++ int i, requeue_idx, err;
++ struct sk_buff *skb;
++ grant_ref_t ref;
++ netif_rx_request_t *req;
++ unsigned int feature_rx_copy, feature_rx_flip;
+
-+ if (regno > info->cmap.len)
-+ return 1;
++ err = xenbus_scanf(XBT_NIL, np->xbdev->otherend,
++ "feature-rx-copy", "%u", &feature_rx_copy);
++ if (err != 1)
++ feature_rx_copy = 0;
++ err = xenbus_scanf(XBT_NIL, np->xbdev->otherend,
++ "feature-rx-flip", "%u", &feature_rx_flip);
++ if (err != 1)
++ feature_rx_flip = 1;
+
-+ red >>= (16 - info->var.red.length);
-+ green >>= (16 - info->var.green.length);
-+ blue >>= (16 - info->var.blue.length);
++ /*
++ * Copy packets on receive path if:
++ * (a) This was requested by user, and the backend supports it; or
++ * (b) Flipping was requested, but this is unsupported by the backend.
++ */
++ np->copying_receiver = ((MODPARM_rx_copy && feature_rx_copy) ||
++ (MODPARM_rx_flip && !feature_rx_flip));
+
-+ v = (red << info->var.red.offset) |
-+ (green << info->var.green.offset) |
-+ (blue << info->var.blue.offset);
++ err = talk_to_backend(np->xbdev, np);
++ if (err)
++ return err;
+
-+ /* FIXME is this sane? check against xxxfb_setcolreg()! */
-+ switch (info->var.bits_per_pixel) {
-+ case 16:
-+ case 24:
-+ case 32:
-+ ((u32 *)info->pseudo_palette)[regno] = v;
-+ break;
-+ }
-+
-+ return 0;
-+}
++ xennet_set_features(dev);
+
-+static void xenfb_timer(unsigned long data)
-+{
-+ struct xenfb_info *info = (struct xenfb_info *)data;
-+ info->dirty = 1;
-+ wake_up(&info->wq);
-+}
++ DPRINTK("device %s has %sing receive path.\n",
++ dev->name, np->copying_receiver ? "copy" : "flipp");
+
-+static void __xenfb_refresh(struct xenfb_info *info,
-+ int x1, int y1, int w, int h)
-+{
-+ int y2, x2;
++ spin_lock_bh(&np->rx_lock);
++ spin_lock_irq(&np->tx_lock);
+
-+ y2 = y1 + h;
-+ x2 = x1 + w;
++ /*
++ * Recovery procedure:
++ * NB. Freelist index entries are always going to be less than
++ * PAGE_OFFSET, whereas pointers to skbs will always be equal or
++ * greater than PAGE_OFFSET: we use this property to distinguish
++ * them.
++ */
+
-+ if (info->y1 > y1)
-+ info->y1 = y1;
-+ if (info->y2 < y2)
-+ info->y2 = y2;
-+ if (info->x1 > x1)
-+ info->x1 = x1;
-+ if (info->x2 < x2)
-+ info->x2 = x2;
++ /* Step 1: Discard all pending TX packet fragments. */
++ netif_release_tx_bufs(np);
+
-+ if (timer_pending(&info->refresh))
-+ return;
++ /* Step 2: Rebuild the RX buffer freelist and the RX ring itself. */
++ for (requeue_idx = 0, i = 0; i < NET_RX_RING_SIZE; i++) {
++ if (!np->rx_skbs[i])
++ continue;
+
-+ mod_timer(&info->refresh, jiffies + HZ/xenfb_fps);
-+}
++ skb = np->rx_skbs[requeue_idx] = xennet_get_rx_skb(np, i);
++ ref = np->grant_rx_ref[requeue_idx] = xennet_get_rx_ref(np, i);
++ req = RING_GET_REQUEST(&np->rx, requeue_idx);
+
-+static void xenfb_refresh(struct xenfb_info *info,
-+ int x1, int y1, int w, int h)
-+{
-+ unsigned long flags;
++ if (!np->copying_receiver) {
++ gnttab_grant_foreign_transfer_ref(
++ ref, np->xbdev->otherend_id,
++ page_to_pfn(skb_shinfo(skb)->frags->page));
++ } else {
++ gnttab_grant_foreign_access_ref(
++ ref, np->xbdev->otherend_id,
++ pfn_to_mfn(page_to_pfn(skb_shinfo(skb)->
++ frags->page)),
++ 0);
++ }
++ req->gref = ref;
++ req->id = requeue_idx;
+
-+ spin_lock_irqsave(&info->dirty_lock, flags);
-+ __xenfb_refresh(info, x1, y1, w, h);
-+ spin_unlock_irqrestore(&info->dirty_lock, flags);
-+}
++ requeue_idx++;
++ }
+
-+static void xenfb_fillrect(struct fb_info *p, const struct fb_fillrect *rect)
-+{
-+ struct xenfb_info *info = p->par;
++ np->rx.req_prod_pvt = requeue_idx;
+
-+ cfb_fillrect(p, rect);
-+ xenfb_refresh(info, rect->dx, rect->dy, rect->width, rect->height);
-+}
++ /*
++ * Step 3: All public and private state should now be sane. Get
++ * ready to start sending and receiving packets and give the driver
++ * domain a kick because we've probably just requeued some
++ * packets.
++ */
++ netfront_carrier_on(np);
++ notify_remote_via_irq(np->irq);
++ network_tx_buf_gc(dev);
++ network_alloc_rx_buffers(dev);
+
-+static void xenfb_imageblit(struct fb_info *p, const struct fb_image *image)
-+{
-+ struct xenfb_info *info = p->par;
++ spin_unlock_irq(&np->tx_lock);
++ spin_unlock_bh(&np->rx_lock);
+
-+ cfb_imageblit(p, image);
-+ xenfb_refresh(info, image->dx, image->dy, image->width, image->height);
++ return 0;
+}
+
-+static void xenfb_copyarea(struct fb_info *p, const struct fb_copyarea *area)
++static void netif_uninit(struct net_device *dev)
+{
-+ struct xenfb_info *info = p->par;
-+
-+ cfb_copyarea(p, area);
-+ xenfb_refresh(info, area->dx, area->dy, area->width, area->height);
++ struct netfront_info *np = netdev_priv(dev);
++ netif_release_tx_bufs(np);
++ if (np->copying_receiver)
++ netif_release_rx_bufs_copy(np);
++ else
++ netif_release_rx_bufs_flip(np);
++ gnttab_free_grant_references(np->gref_tx_head);
++ gnttab_free_grant_references(np->gref_rx_head);
+}
+
-+static void xenfb_vm_open(struct vm_area_struct *vma)
++static struct ethtool_ops network_ethtool_ops =
+{
-+ struct xenfb_mapping *map = vma->vm_private_data;
-+ atomic_inc(&map->map_refs);
-+}
++ .get_tx_csum = ethtool_op_get_tx_csum,
++ .set_tx_csum = ethtool_op_set_tx_csum,
++ .get_sg = ethtool_op_get_sg,
++ .set_sg = xennet_set_sg,
++#if HAVE_TSO
++ .get_tso = ethtool_op_get_tso,
++ .set_tso = xennet_set_tso,
++#endif
++ .get_link = ethtool_op_get_link,
++};
+
-+static void xenfb_vm_close(struct vm_area_struct *vma)
++#ifdef CONFIG_SYSFS
++static ssize_t show_rxbuf_min(struct class_device *cd, char *buf)
+{
-+ struct xenfb_mapping *map = vma->vm_private_data;
-+ struct xenfb_info *info = map->info;
++ struct net_device *netdev = container_of(cd, struct net_device,
++ class_dev);
++ struct netfront_info *info = netdev_priv(netdev);
+
-+ mutex_lock(&info->mm_lock);
-+ if (atomic_dec_and_test(&map->map_refs)) {
-+ list_del(&map->link);
-+ kfree(map);
-+ }
-+ mutex_unlock(&info->mm_lock);
++ return sprintf(buf, "%u\n", info->rx_min_target);
+}
+
-+static struct page *xenfb_vm_nopage(struct vm_area_struct *vma,
-+ unsigned long vaddr, int *type)
++static ssize_t store_rxbuf_min(struct class_device *cd,
++ const char *buf, size_t len)
+{
-+ struct xenfb_mapping *map = vma->vm_private_data;
-+ struct xenfb_info *info = map->info;
-+ int pgnr = (vaddr - vma->vm_start) >> PAGE_SHIFT;
-+ unsigned long flags;
-+ struct page *page;
-+ int y1, y2;
++ struct net_device *netdev = container_of(cd, struct net_device,
++ class_dev);
++ struct netfront_info *np = netdev_priv(netdev);
++ char *endp;
++ unsigned long target;
+
-+ if (pgnr >= info->nr_pages)
-+ return NOPAGE_SIGBUS;
++ if (!capable(CAP_NET_ADMIN))
++ return -EPERM;
++
++ target = simple_strtoul(buf, &endp, 0);
++ if (endp == buf)
++ return -EBADMSG;
++
++ if (target < RX_MIN_TARGET)
++ target = RX_MIN_TARGET;
++ if (target > RX_MAX_TARGET)
++ target = RX_MAX_TARGET;
++
++ spin_lock_bh(&np->rx_lock);
++ if (target > np->rx_max_target)
++ np->rx_max_target = target;
++ np->rx_min_target = target;
++ if (target > np->rx_target)
++ np->rx_target = target;
+
-+ mutex_lock(&info->mm_lock);
-+ spin_lock_irqsave(&info->dirty_lock, flags);
-+ page = info->pages[pgnr];
-+ get_page(page);
-+ map->faults++;
++ network_alloc_rx_buffers(netdev);
+
-+ y1 = pgnr * PAGE_SIZE / info->fb_info->fix.line_length;
-+ y2 = (pgnr * PAGE_SIZE + PAGE_SIZE - 1) / info->fb_info->fix.line_length;
-+ if (y2 > info->fb_info->var.yres)
-+ y2 = info->fb_info->var.yres;
-+ __xenfb_refresh(info, 0, y1, info->fb_info->var.xres, y2 - y1);
-+ spin_unlock_irqrestore(&info->dirty_lock, flags);
-+ mutex_unlock(&info->mm_lock);
++ spin_unlock_bh(&np->rx_lock);
++ return len;
++}
+
-+ if (type)
-+ *type = VM_FAULT_MINOR;
++static ssize_t show_rxbuf_max(struct class_device *cd, char *buf)
++{
++ struct net_device *netdev = container_of(cd, struct net_device,
++ class_dev);
++ struct netfront_info *info = netdev_priv(netdev);
+
-+ return page;
++ return sprintf(buf, "%u\n", info->rx_max_target);
+}
+
-+static struct vm_operations_struct xenfb_vm_ops = {
-+ .open = xenfb_vm_open,
-+ .close = xenfb_vm_close,
-+ .nopage = xenfb_vm_nopage,
-+};
-+
-+static int xenfb_mmap(struct fb_info *fb_info, struct vm_area_struct *vma)
++static ssize_t store_rxbuf_max(struct class_device *cd,
++ const char *buf, size_t len)
+{
-+ struct xenfb_info *info = fb_info->par;
-+ struct xenfb_mapping *map;
-+ int map_pages;
++ struct net_device *netdev = container_of(cd, struct net_device,
++ class_dev);
++ struct netfront_info *np = netdev_priv(netdev);
++ char *endp;
++ unsigned long target;
+
-+ if (!(vma->vm_flags & VM_WRITE))
-+ return -EINVAL;
-+ if (!(vma->vm_flags & VM_SHARED))
-+ return -EINVAL;
-+ if (vma->vm_pgoff != 0)
-+ return -EINVAL;
++ if (!capable(CAP_NET_ADMIN))
++ return -EPERM;
+
-+ map_pages = (vma->vm_end - vma->vm_start + PAGE_SIZE-1) >> PAGE_SHIFT;
-+ if (map_pages > info->nr_pages)
-+ return -EINVAL;
++ target = simple_strtoul(buf, &endp, 0);
++ if (endp == buf)
++ return -EBADMSG;
+
-+ map = kzalloc(sizeof(*map), GFP_KERNEL);
-+ if (map == NULL)
-+ return -ENOMEM;
++ if (target < RX_MIN_TARGET)
++ target = RX_MIN_TARGET;
++ if (target > RX_MAX_TARGET)
++ target = RX_MAX_TARGET;
+
-+ map->vma = vma;
-+ map->faults = 0;
-+ map->info = info;
-+ atomic_set(&map->map_refs, 1);
++ spin_lock_bh(&np->rx_lock);
++ if (target < np->rx_min_target)
++ np->rx_min_target = target;
++ np->rx_max_target = target;
++ if (target < np->rx_target)
++ np->rx_target = target;
+
-+ mutex_lock(&info->mm_lock);
-+ list_add(&map->link, &info->mappings);
-+ mutex_unlock(&info->mm_lock);
++ network_alloc_rx_buffers(netdev);
+
-+ vma->vm_ops = &xenfb_vm_ops;
-+ vma->vm_flags |= (VM_DONTEXPAND | VM_RESERVED);
-+ vma->vm_private_data = map;
++ spin_unlock_bh(&np->rx_lock);
++ return len;
++}
+
-+ return 0;
++static ssize_t show_rxbuf_cur(struct class_device *cd, char *buf)
++{
++ struct net_device *netdev = container_of(cd, struct net_device,
++ class_dev);
++ struct netfront_info *info = netdev_priv(netdev);
++
++ return sprintf(buf, "%u\n", info->rx_target);
+}
+
-+static struct fb_ops xenfb_fb_ops = {
-+ .owner = THIS_MODULE,
-+ .fb_setcolreg = xenfb_setcolreg,
-+ .fb_fillrect = xenfb_fillrect,
-+ .fb_copyarea = xenfb_copyarea,
-+ .fb_imageblit = xenfb_imageblit,
-+ .fb_mmap = xenfb_mmap,
++static const struct class_device_attribute xennet_attrs[] = {
++ __ATTR(rxbuf_min, S_IRUGO|S_IWUSR, show_rxbuf_min, store_rxbuf_min),
++ __ATTR(rxbuf_max, S_IRUGO|S_IWUSR, show_rxbuf_max, store_rxbuf_max),
++ __ATTR(rxbuf_cur, S_IRUGO, show_rxbuf_cur, NULL),
+};
+
-+static irqreturn_t xenfb_event_handler(int rq, void *dev_id,
-+ struct pt_regs *regs)
++static int xennet_sysfs_addif(struct net_device *netdev)
+{
-+ /*
-+ * No in events recognized, simply ignore them all.
-+ * If you need to recognize some, see xenbkd's input_handler()
-+ * for how to do that.
-+ */
-+ struct xenfb_info *info = dev_id;
-+ struct xenfb_page *page = info->page;
++ int i;
++ int error = 0;
+
-+ if (page->in_cons != page->in_prod) {
-+ info->page->in_cons = info->page->in_prod;
-+ notify_remote_via_irq(info->irq);
++ for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++) {
++ error = class_device_create_file(&netdev->class_dev,
++ &xennet_attrs[i]);
++ if (error)
++ goto fail;
+ }
-+ return IRQ_HANDLED;
-+}
++ return 0;
+
-+static unsigned long vmalloc_to_mfn(void *address)
-+{
-+ return pfn_to_mfn(vmalloc_to_pfn(address));
++ fail:
++ while (--i >= 0)
++ class_device_remove_file(&netdev->class_dev,
++ &xennet_attrs[i]);
++ return error;
+}
+
-+static int __devinit xenfb_probe(struct xenbus_device *dev,
-+ const struct xenbus_device_id *id)
++static void xennet_sysfs_delif(struct net_device *netdev)
+{
-+ struct xenfb_info *info;
-+ struct fb_info *fb_info;
-+ int ret;
++ int i;
+
-+ info = kzalloc(sizeof(*info), GFP_KERNEL);
-+ if (info == NULL) {
-+ xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
-+ return -ENOMEM;
++ for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++) {
++ class_device_remove_file(&netdev->class_dev,
++ &xennet_attrs[i]);
+ }
-+ dev->dev.driver_data = info;
-+ info->xbdev = dev;
-+ info->irq = -1;
-+ info->x1 = info->y1 = INT_MAX;
-+ spin_lock_init(&info->dirty_lock);
-+ mutex_init(&info->mm_lock);
-+ init_waitqueue_head(&info->wq);
-+ init_timer(&info->refresh);
-+ info->refresh.function = xenfb_timer;
-+ info->refresh.data = (unsigned long)info;
-+ INIT_LIST_HEAD(&info->mappings);
-+
-+ info->fb = vmalloc(xenfb_mem_len);
-+ if (info->fb == NULL)
-+ goto error_nomem;
-+ memset(info->fb, 0, xenfb_mem_len);
-+
-+ info->nr_pages = (xenfb_mem_len + PAGE_SIZE - 1) >> PAGE_SHIFT;
-+
-+ info->pages = kmalloc(sizeof(struct page *) * info->nr_pages,
-+ GFP_KERNEL);
-+ if (info->pages == NULL)
-+ goto error_nomem;
++}
+
-+ info->mfns = vmalloc(sizeof(unsigned long) * info->nr_pages);
-+ if (!info->mfns)
-+ goto error_nomem;
++#endif /* CONFIG_SYSFS */
+
-+ /* set up shared page */
-+ info->page = (void *)__get_free_page(GFP_KERNEL | __GFP_ZERO);
-+ if (!info->page)
-+ goto error_nomem;
+
-+ xenfb_init_shared_page(info);
++/*
++ * Nothing to do here. Virtual interface is point-to-point and the
++ * physical interface is probably promiscuous anyway.
++ */
++static void network_set_multicast_list(struct net_device *dev)
++{
++}
+
-+ fb_info = framebuffer_alloc(sizeof(u32) * 256, NULL);
-+ /* see fishy hackery below */
-+ if (fb_info == NULL)
-+ goto error_nomem;
++static struct net_device * __devinit create_netdev(struct xenbus_device *dev)
++{
++ int i, err = 0;
++ struct net_device *netdev = NULL;
++ struct netfront_info *np = NULL;
+
-+ /* FIXME fishy hackery */
-+ fb_info->pseudo_palette = fb_info->par;
-+ fb_info->par = info;
-+ /* /FIXME */
-+ fb_info->screen_base = info->fb;
++ netdev = alloc_etherdev(sizeof(struct netfront_info));
++ if (!netdev) {
++ printk(KERN_WARNING "%s> alloc_etherdev failed.\n",
++ __FUNCTION__);
++ return ERR_PTR(-ENOMEM);
++ }
+
-+ fb_info->fbops = &xenfb_fb_ops;
-+ fb_info->var.xres_virtual = fb_info->var.xres = info->page->width;
-+ fb_info->var.yres_virtual = fb_info->var.yres = info->page->height;
-+ fb_info->var.bits_per_pixel = info->page->depth;
++ np = netdev_priv(netdev);
++ np->xbdev = dev;
+
-+ fb_info->var.red = (struct fb_bitfield){16, 8, 0};
-+ fb_info->var.green = (struct fb_bitfield){8, 8, 0};
-+ fb_info->var.blue = (struct fb_bitfield){0, 8, 0};
++ spin_lock_init(&np->tx_lock);
++ spin_lock_init(&np->rx_lock);
+
-+ fb_info->var.activate = FB_ACTIVATE_NOW;
-+ fb_info->var.height = -1;
-+ fb_info->var.width = -1;
-+ fb_info->var.vmode = FB_VMODE_NONINTERLACED;
++ init_accelerator_vif(np, dev);
+
-+ fb_info->fix.visual = FB_VISUAL_TRUECOLOR;
-+ fb_info->fix.line_length = info->page->line_length;
-+ fb_info->fix.smem_start = 0;
-+ fb_info->fix.smem_len = xenfb_mem_len;
-+ strcpy(fb_info->fix.id, "xen");
-+ fb_info->fix.type = FB_TYPE_PACKED_PIXELS;
-+ fb_info->fix.accel = FB_ACCEL_NONE;
++ skb_queue_head_init(&np->rx_batch);
++ np->rx_target = RX_DFL_MIN_TARGET;
++ np->rx_min_target = RX_DFL_MIN_TARGET;
++ np->rx_max_target = RX_MAX_TARGET;
+
-+ fb_info->flags = FBINFO_FLAG_DEFAULT;
++ init_timer(&np->rx_refill_timer);
++ np->rx_refill_timer.data = (unsigned long)netdev;
++ np->rx_refill_timer.function = rx_refill_timeout;
+
-+ ret = fb_alloc_cmap(&fb_info->cmap, 256, 0);
-+ if (ret < 0) {
-+ framebuffer_release(fb_info);
-+ xenbus_dev_fatal(dev, ret, "fb_alloc_cmap");
-+ goto error;
++ /* Initialise {tx,rx}_skbs as a free chain containing every entry. */
++ for (i = 0; i <= NET_TX_RING_SIZE; i++) {
++ np->tx_skbs[i] = (void *)((unsigned long) i+1);
++ np->grant_tx_ref[i] = GRANT_INVALID_REF;
+ }
+
-+ ret = register_framebuffer(fb_info);
-+ if (ret) {
-+ fb_dealloc_cmap(&info->fb_info->cmap);
-+ framebuffer_release(fb_info);
-+ xenbus_dev_fatal(dev, ret, "register_framebuffer");
-+ goto error;
++ for (i = 0; i < NET_RX_RING_SIZE; i++) {
++ np->rx_skbs[i] = NULL;
++ np->grant_rx_ref[i] = GRANT_INVALID_REF;
+ }
-+ info->fb_info = fb_info;
+
-+ /* FIXME should this be delayed until backend XenbusStateConnected? */
-+ info->kthread = kthread_run(xenfb_thread, info, "xenfb thread");
-+ if (IS_ERR(info->kthread)) {
-+ ret = PTR_ERR(info->kthread);
-+ info->kthread = NULL;
-+ xenbus_dev_fatal(dev, ret, "register_framebuffer");
-+ goto error;
++ /* A grant for every tx ring slot */
++ if (gnttab_alloc_grant_references(TX_MAX_TARGET,
++ &np->gref_tx_head) < 0) {
++ printk(KERN_ALERT "#### netfront can't alloc tx grant refs\n");
++ err = -ENOMEM;
++ goto exit;
++ }
++ /* A grant for every rx ring slot */
++ if (gnttab_alloc_grant_references(RX_MAX_TARGET,
++ &np->gref_rx_head) < 0) {
++ printk(KERN_ALERT "#### netfront can't alloc rx grant refs\n");
++ err = -ENOMEM;
++ goto exit_free_tx;
+ }
+
-+ ret = xenfb_connect_backend(dev, info);
-+ if (ret < 0)
-+ goto error;
-+
-+ return 0;
-+
-+ error_nomem:
-+ ret = -ENOMEM;
-+ xenbus_dev_fatal(dev, ret, "allocating device memory");
-+ error:
-+ xenfb_remove(dev);
-+ return ret;
-+}
++ netdev->open = network_open;
++ netdev->hard_start_xmit = network_start_xmit;
++ netdev->stop = network_close;
++ netdev->get_stats = network_get_stats;
++ netdev->poll = netif_poll;
++ netdev->set_multicast_list = network_set_multicast_list;
++ netdev->uninit = netif_uninit;
++ netdev->set_mac_address = xennet_set_mac_address;
++ netdev->change_mtu = xennet_change_mtu;
++ netdev->weight = 64;
++ netdev->features = NETIF_F_IP_CSUM;
+
-+static int xenfb_resume(struct xenbus_device *dev)
-+{
-+ struct xenfb_info *info = dev->dev.driver_data;
++ SET_ETHTOOL_OPS(netdev, &network_ethtool_ops);
++ SET_MODULE_OWNER(netdev);
++ SET_NETDEV_DEV(netdev, &dev->dev);
+
-+ xenfb_disconnect_backend(info);
-+ xenfb_init_shared_page(info);
-+ return xenfb_connect_backend(dev, info);
-+}
++ np->netdev = netdev;
+
-+static int xenfb_remove(struct xenbus_device *dev)
-+{
-+ struct xenfb_info *info = dev->dev.driver_data;
++ netfront_carrier_off(np);
+
-+ del_timer(&info->refresh);
-+ if (info->kthread)
-+ kthread_stop(info->kthread);
-+ xenfb_disconnect_backend(info);
-+ if (info->fb_info) {
-+ unregister_framebuffer(info->fb_info);
-+ fb_dealloc_cmap(&info->fb_info->cmap);
-+ framebuffer_release(info->fb_info);
-+ }
-+ free_page((unsigned long)info->page);
-+ vfree(info->mfns);
-+ kfree(info->pages);
-+ vfree(info->fb);
-+ kfree(info);
++ return netdev;
+
-+ return 0;
++ exit_free_tx:
++ gnttab_free_grant_references(np->gref_tx_head);
++ exit:
++ free_netdev(netdev);
++ return ERR_PTR(err);
+}
+
-+static void xenfb_init_shared_page(struct xenfb_info *info)
++#ifdef CONFIG_INET
++/*
++ * We use this notifier to send out a fake ARP reply to reset switches and
++ * router ARP caches when an IP interface is brought up on a VIF.
++ */
++static int
++inetdev_notify(struct notifier_block *this, unsigned long event, void *ptr)
+{
-+ int i;
-+
-+ for (i = 0; i < info->nr_pages; i++)
-+ info->pages[i] = vmalloc_to_page(info->fb + i * PAGE_SIZE);
++ struct in_ifaddr *ifa = (struct in_ifaddr *)ptr;
++ struct net_device *dev = ifa->ifa_dev->dev;
+
-+ for (i = 0; i < info->nr_pages; i++)
-+ info->mfns[i] = vmalloc_to_mfn(info->fb + i * PAGE_SIZE);
++ /* UP event and is it one of our devices? */
++ if (event == NETDEV_UP && dev->open == network_open)
++ send_fake_arp(dev);
+
-+ info->page->pd[0] = vmalloc_to_mfn(info->mfns);
-+ info->page->pd[1] = 0;
-+ info->page->width = XENFB_WIDTH;
-+ info->page->height = XENFB_HEIGHT;
-+ info->page->depth = XENFB_DEPTH;
-+ info->page->line_length = (info->page->depth / 8) * info->page->width;
-+ info->page->mem_length = xenfb_mem_len;
-+ info->page->in_cons = info->page->in_prod = 0;
-+ info->page->out_cons = info->page->out_prod = 0;
++ return NOTIFY_DONE;
+}
+
-+static int xenfb_connect_backend(struct xenbus_device *dev,
-+ struct xenfb_info *info)
-+{
-+ int ret;
-+ struct xenbus_transaction xbt;
++static struct notifier_block notifier_inetdev = {
++ .notifier_call = inetdev_notify,
++ .next = NULL,
++ .priority = 0
++};
++#endif
+
-+ ret = bind_listening_port_to_irqhandler(
-+ dev->otherend_id, xenfb_event_handler, 0, "xenfb", info);
-+ if (ret < 0) {
-+ xenbus_dev_fatal(dev, ret,
-+ "bind_listening_port_to_irqhandler");
-+ return ret;
-+ }
-+ info->irq = ret;
+
-+ again:
-+ ret = xenbus_transaction_start(&xbt);
-+ if (ret) {
-+ xenbus_dev_fatal(dev, ret, "starting transaction");
-+ return ret;
-+ }
-+ ret = xenbus_printf(xbt, dev->nodename, "page-ref", "%lu",
-+ virt_to_mfn(info->page));
-+ if (ret)
-+ goto error_xenbus;
-+ ret = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
-+ irq_to_evtchn_port(info->irq));
-+ if (ret)
-+ goto error_xenbus;
-+ ret = xenbus_printf(xbt, dev->nodename, "protocol", "%s",
-+ XEN_IO_PROTO_ABI_NATIVE);
-+ if (ret)
-+ goto error_xenbus;
-+ ret = xenbus_printf(xbt, dev->nodename, "feature-update", "1");
-+ if (ret)
-+ goto error_xenbus;
-+ ret = xenbus_transaction_end(xbt, 0);
-+ if (ret) {
-+ if (ret == -EAGAIN)
-+ goto again;
-+ xenbus_dev_fatal(dev, ret, "completing transaction");
-+ return ret;
-+ }
++static void netif_disconnect_backend(struct netfront_info *info)
++{
++ /* Stop old i/f to prevent errors whilst we rebuild the state. */
++ spin_lock_bh(&info->rx_lock);
++ spin_lock_irq(&info->tx_lock);
++ netfront_carrier_off(info);
++ spin_unlock_irq(&info->tx_lock);
++ spin_unlock_bh(&info->rx_lock);
+
-+ xenbus_switch_state(dev, XenbusStateInitialised);
-+ return 0;
++ if (info->irq)
++ unbind_from_irqhandler(info->irq, info->netdev);
++ info->irq = 0;
+
-+ error_xenbus:
-+ xenbus_transaction_end(xbt, 1);
-+ xenbus_dev_fatal(dev, ret, "writing xenstore");
-+ return ret;
++ end_access(info->tx_ring_ref, info->tx.sring);
++ end_access(info->rx_ring_ref, info->rx.sring);
++ info->tx_ring_ref = GRANT_INVALID_REF;
++ info->rx_ring_ref = GRANT_INVALID_REF;
++ info->tx.sring = NULL;
++ info->rx.sring = NULL;
+}
+
-+static void xenfb_disconnect_backend(struct xenfb_info *info)
-+{
-+ if (info->irq >= 0)
-+ unbind_from_irqhandler(info->irq, info);
-+ info->irq = -1;
-+}
+
-+static void xenfb_backend_changed(struct xenbus_device *dev,
-+ enum xenbus_state backend_state)
++static void end_access(int ref, void *page)
+{
-+ struct xenfb_info *info = dev->dev.driver_data;
-+ int val;
-+
-+ switch (backend_state) {
-+ case XenbusStateInitialising:
-+ case XenbusStateInitialised:
-+ case XenbusStateUnknown:
-+ case XenbusStateClosed:
-+ break;
-+
-+ case XenbusStateInitWait:
-+ InitWait:
-+ xenbus_switch_state(dev, XenbusStateConnected);
-+ break;
++ if (ref != GRANT_INVALID_REF)
++ gnttab_end_foreign_access(ref, (unsigned long)page);
++}
+
-+ case XenbusStateConnected:
-+ /*
-+ * Work around xenbus race condition: If backend goes
-+ * through InitWait to Connected fast enough, we can
-+ * get Connected twice here.
-+ */
-+ if (dev->state != XenbusStateConnected)
-+ goto InitWait; /* no InitWait seen yet, fudge it */
+
-+ if (xenbus_scanf(XBT_NIL, info->xbdev->otherend,
-+ "request-update", "%d", &val) < 0)
-+ val = 0;
-+ if (val)
-+ info->update_wanted = 1;
-+ break;
++/* ** Driver registration ** */
+
-+ case XenbusStateClosing:
-+ // FIXME is this safe in any dev->state?
-+ xenbus_frontend_closed(dev);
-+ break;
-+ }
-+}
+
-+static struct xenbus_device_id xenfb_ids[] = {
-+ { "vfb" },
++static const struct xenbus_device_id netfront_ids[] = {
++ { "vif" },
+ { "" }
+};
-+MODULE_ALIAS("xen:vfb");
++MODULE_ALIAS("xen:vif");
+
-+static struct xenbus_driver xenfb = {
-+ .name = "vfb",
++
++static struct xenbus_driver netfront_driver = {
++ .name = "vif",
+ .owner = THIS_MODULE,
-+ .ids = xenfb_ids,
-+ .probe = xenfb_probe,
-+ .remove = xenfb_remove,
-+ .resume = xenfb_resume,
-+ .otherend_changed = xenfb_backend_changed,
++ .ids = netfront_ids,
++ .probe = netfront_probe,
++ .remove = __devexit_p(netfront_remove),
++ .suspend = netfront_suspend,
++ .suspend_cancel = netfront_suspend_cancel,
++ .resume = netfront_resume,
++ .otherend_changed = backend_changed,
+};
+
-+static int __init xenfb_init(void)
++
++static int __init netif_init(void)
+{
+ if (!is_running_on_xen())
+ return -ENODEV;
+
-+ /* Nothing to do if running in dom0. */
-+ if (is_initial_xendomain())
-+ return -ENODEV;
++#ifdef CONFIG_XEN
++ if (MODPARM_rx_flip && MODPARM_rx_copy) {
++ WPRINTK("Cannot specify both rx_copy and rx_flip.\n");
++ return -EINVAL;
++ }
++
++ if (!MODPARM_rx_flip && !MODPARM_rx_copy)
++ MODPARM_rx_flip = 1; /* Default is to flip. */
++#endif
++
++ netif_init_accel();
++
++ IPRINTK("Initialising virtual ethernet driver.\n");
++
++#ifdef CONFIG_INET
++ (void)register_inetaddr_notifier(¬ifier_inetdev);
++#endif
+
-+ return xenbus_register_frontend(&xenfb);
++ return xenbus_register_frontend(&netfront_driver);
+}
++module_init(netif_init);
+
-+static void __exit xenfb_cleanup(void)
++
++static void __exit netif_exit(void)
+{
-+ return xenbus_unregister_driver(&xenfb);
-+}
++#ifdef CONFIG_INET
++ unregister_inetaddr_notifier(¬ifier_inetdev);
++#endif
+
-+module_init(xenfb_init);
-+module_exit(xenfb_cleanup);
++ netif_exit_accel();
+
-+MODULE_LICENSE("GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/fbfront/xenkbd.c tmp-linux-2.6-xen.patch/drivers/xen/fbfront/xenkbd.c
---- pristine-linux-2.6.18.2/drivers/xen/fbfront/xenkbd.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/fbfront/xenkbd.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,334 @@
-+/*
-+ * linux/drivers/input/keyboard/xenkbd.c -- Xen para-virtual input device
++ return xenbus_unregister_driver(&netfront_driver);
++}
++module_exit(netif_exit);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/netfront/netfront.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/netfront/netfront.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,274 @@
++/******************************************************************************
++ * Virtual network driver for conversing with remote driver backends.
+ *
-+ * Copyright (C) 2005 Anthony Liguori <aliguori at us.ibm.com>
-+ * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster <armbru at redhat.com>
++ * Copyright (c) 2002-2005, K A Fraser
++ * Copyright (c) 2005, XenSource Ltd
++ * Copyright (C) 2007 Solarflare Communications, Inc.
+ *
-+ * Based on linux/drivers/input/mouse/sermouse.c
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
+ *
-+ * This file is subject to the terms and conditions of the GNU General Public
-+ * License. See the file COPYING in the main directory of this archive for
-+ * more details.
-+ */
-+
-+/*
-+ * TODO:
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
+ *
-+ * Switch to grant tables together with xenfb.c.
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
+
-+#include <linux/kernel.h>
-+#include <linux/errno.h>
-+#include <linux/module.h>
-+#include <linux/input.h>
-+#include <asm/hypervisor.h>
-+#include <xen/evtchn.h>
-+#include <xen/interface/io/fbif.h>
-+#include <xen/interface/io/kbdif.h>
++#ifndef NETFRONT_H
++#define NETFRONT_H
++
++#include <xen/interface/io/netif.h>
++#include <linux/netdevice.h>
++#include <linux/skbuff.h>
++#include <linux/list.h>
++
++#define NET_TX_RING_SIZE __RING_SIZE((struct netif_tx_sring *)0, PAGE_SIZE)
++#define NET_RX_RING_SIZE __RING_SIZE((struct netif_rx_sring *)0, PAGE_SIZE)
++
+#include <xen/xenbus.h>
+
-+struct xenkbd_info
-+{
-+ struct input_dev *kbd;
-+ struct input_dev *ptr;
-+ struct xenkbd_page *page;
-+ int irq;
-+ struct xenbus_device *xbdev;
-+ char phys[32];
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++/*
++ * Function pointer table for hooks into a network acceleration
++ * plugin. These are called at appropriate points from the netfront
++ * driver
++ */
++struct netfront_accel_hooks {
++ /*
++ * new_device: Accelerator hook to ask the plugin to support a
++ * new network interface
++ */
++ int (*new_device)(struct net_device *net_dev, struct xenbus_device *dev);
++ /*
++ * remove: Opposite of new_device
++ */
++ int (*remove)(struct xenbus_device *dev);
++ /*
++ * The net_device is being polled, check the accelerated
++ * hardware for any pending packets
++ */
++ int (*netdev_poll)(struct net_device *dev, int *pbudget);
++ /*
++ * start_xmit: Used to give the accelerated plugin the option
++ * of sending a packet. Returns non-zero if has done so, or
++ * zero to decline and force the packet onto normal send
++ * path
++ */
++ int (*start_xmit)(struct sk_buff *skb, struct net_device *dev);
++ /*
++ * start/stop_napi_interrupts Used by netfront to indicate
++ * when napi interrupts should be enabled or disabled
++ */
++ int (*start_napi_irq)(struct net_device *dev);
++ void (*stop_napi_irq)(struct net_device *dev);
++ /*
++ * Called before re-enabling the TX queue to check the fast
++ * path has slots too
++ */
++ int (*check_ready)(struct net_device *dev);
++ /*
++ * Get the fastpath network statistics
++ */
++ int (*get_stats)(struct net_device *dev,
++ struct net_device_stats *stats);
+};
+
-+static int xenkbd_remove(struct xenbus_device *);
-+static int xenkbd_connect_backend(struct xenbus_device *, struct xenkbd_info *);
-+static void xenkbd_disconnect_backend(struct xenkbd_info *);
+
-+/*
-+ * Note: if you need to send out events, see xenfb_do_update() for how
-+ * to do that.
-+ */
++/* Version of API/protocol for communication between netfront and
++ acceleration plugin supported */
++#define NETFRONT_ACCEL_VERSION 0x00010003
+
-+static irqreturn_t input_handler(int rq, void *dev_id, struct pt_regs *regs)
-+{
-+ struct xenkbd_info *info = dev_id;
-+ struct xenkbd_page *page = info->page;
-+ __u32 cons, prod;
++/*
++ * Per-netfront device state for the accelerator. This is used to
++ * allow efficient per-netfront device access to the accelerator
++ * hooks
++ */
++struct netfront_accel_vif_state {
++ struct list_head link;
+
-+ prod = page->in_prod;
-+ if (prod == page->out_cons)
-+ return IRQ_HANDLED;
-+ rmb(); /* ensure we see ring contents up to prod */
-+ for (cons = page->in_cons; cons != prod; cons++) {
-+ union xenkbd_in_event *event;
-+ struct input_dev *dev;
-+ event = &XENKBD_IN_RING_REF(page, cons);
++ struct xenbus_device *dev;
++ struct netfront_info *np;
++ struct netfront_accel_hooks *hooks;
+
-+ dev = info->ptr;
-+ switch (event->type) {
-+ case XENKBD_TYPE_MOTION:
-+ input_report_rel(dev, REL_X, event->motion.rel_x);
-+ input_report_rel(dev, REL_Y, event->motion.rel_y);
-+ break;
-+ case XENKBD_TYPE_KEY:
-+ dev = NULL;
-+ if (test_bit(event->key.keycode, info->kbd->keybit))
-+ dev = info->kbd;
-+ if (test_bit(event->key.keycode, info->ptr->keybit))
-+ dev = info->ptr;
-+ if (dev)
-+ input_report_key(dev, event->key.keycode,
-+ event->key.pressed);
-+ else
-+ printk("xenkbd: unhandled keycode 0x%x\n",
-+ event->key.keycode);
-+ break;
-+ case XENKBD_TYPE_POS:
-+ input_report_abs(dev, ABS_X, event->pos.abs_x);
-+ input_report_abs(dev, ABS_Y, event->pos.abs_y);
-+ break;
-+ }
-+ if (dev)
-+ input_sync(dev);
-+ }
-+ mb(); /* ensure we got ring contents */
-+ page->in_cons = cons;
-+ notify_remote_via_irq(info->irq);
++ /* Watch on the accelerator configuration value */
++ struct xenbus_watch accel_watch;
++ /* Work item to process change in accelerator */
++ struct work_struct accel_work;
++ /* The string from xenbus last time accel_watch fired */
++ char *accel_frontend;
++};
+
-+ return IRQ_HANDLED;
-+}
++/*
++ * Per-accelerator state stored in netfront. These form a list that
++ * is used to track which devices are accelerated by which plugins,
++ * and what plugins are available/have been requested
++ */
++struct netfront_accelerator {
++ /* Used to make a list */
++ struct list_head link;
++ /* ID of the accelerator */
++ int id;
++ /*
++ * String describing the accelerator. Currently this is the
++ * name of the accelerator module. This is provided by the
++ * backend accelerator through xenstore
++ */
++ char *frontend;
++ /* The hooks into the accelerator plugin module */
++ struct netfront_accel_hooks *hooks;
+
-+int __devinit xenkbd_probe(struct xenbus_device *dev,
-+ const struct xenbus_device_id *id)
-+{
-+ int ret, i;
-+ struct xenkbd_info *info;
-+ struct input_dev *kbd, *ptr;
++ /*
++ * List of per-netfront device state (struct
++ * netfront_accel_vif_state) for each netfront device that is
++ * using this accelerator
++ */
++ struct list_head vif_states;
++ spinlock_t vif_states_lock;
++};
+
-+ info = kzalloc(sizeof(*info), GFP_KERNEL);
-+ if (!info) {
-+ xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
-+ return -ENOMEM;
-+ }
-+ dev->dev.driver_data = info;
-+ info->xbdev = dev;
-+ snprintf(info->phys, sizeof(info->phys), "xenbus/%s", dev->nodename);
++struct netfront_info {
++ struct list_head list;
++ struct net_device *netdev;
+
-+ info->page = (void *)__get_free_page(GFP_KERNEL);
-+ if (!info->page)
-+ goto error_nomem;
-+ info->page->in_cons = info->page->in_prod = 0;
-+ info->page->out_cons = info->page->out_prod = 0;
++ struct net_device_stats stats;
+
-+ /* keyboard */
-+ kbd = input_allocate_device();
-+ if (!kbd)
-+ goto error_nomem;
-+ kbd->name = "Xen Virtual Keyboard";
-+ kbd->phys = info->phys;
-+ kbd->id.bustype = BUS_PCI;
-+ kbd->id.vendor = 0x5853;
-+ kbd->id.product = 0xffff;
-+ kbd->evbit[0] = BIT(EV_KEY);
-+ for (i = KEY_ESC; i < KEY_UNKNOWN; i++)
-+ set_bit(i, kbd->keybit);
-+ for (i = KEY_OK; i < KEY_MAX; i++)
-+ set_bit(i, kbd->keybit);
++ struct netif_tx_front_ring tx;
++ struct netif_rx_front_ring rx;
+
-+ ret = input_register_device(kbd);
-+ if (ret) {
-+ input_free_device(kbd);
-+ xenbus_dev_fatal(dev, ret, "input_register_device(kbd)");
-+ goto error;
-+ }
-+ info->kbd = kbd;
++ spinlock_t tx_lock;
++ spinlock_t rx_lock;
+
-+ /* pointing device */
-+ ptr = input_allocate_device();
-+ if (!ptr)
-+ goto error_nomem;
-+ ptr->name = "Xen Virtual Pointer";
-+ ptr->phys = info->phys;
-+ ptr->id.bustype = BUS_PCI;
-+ ptr->id.vendor = 0x5853;
-+ ptr->id.product = 0xfffe;
-+ ptr->evbit[0] = BIT(EV_KEY) | BIT(EV_REL) | BIT(EV_ABS);
-+ for (i = BTN_LEFT; i <= BTN_TASK; i++)
-+ set_bit(i, ptr->keybit);
-+ ptr->relbit[0] = BIT(REL_X) | BIT(REL_Y);
-+ input_set_abs_params(ptr, ABS_X, 0, XENFB_WIDTH, 0, 0);
-+ input_set_abs_params(ptr, ABS_Y, 0, XENFB_HEIGHT, 0, 0);
++ unsigned int irq;
++ unsigned int copying_receiver;
++ unsigned int carrier;
+
-+ ret = input_register_device(ptr);
-+ if (ret) {
-+ input_free_device(ptr);
-+ xenbus_dev_fatal(dev, ret, "input_register_device(ptr)");
-+ goto error;
-+ }
-+ info->ptr = ptr;
++ /* Receive-ring batched refills. */
++#define RX_MIN_TARGET 8
++#define RX_DFL_MIN_TARGET 64
++#define RX_MAX_TARGET min_t(int, NET_RX_RING_SIZE, 256)
++ unsigned rx_min_target, rx_max_target, rx_target;
++ struct sk_buff_head rx_batch;
+
-+ ret = xenkbd_connect_backend(dev, info);
-+ if (ret < 0)
-+ goto error;
++ struct timer_list rx_refill_timer;
+
-+ return 0;
++ /*
++ * {tx,rx}_skbs store outstanding skbuffs. The first entry in tx_skbs
++ * is an index into a chain of free entries.
++ */
++ struct sk_buff *tx_skbs[NET_TX_RING_SIZE+1];
++ struct sk_buff *rx_skbs[NET_RX_RING_SIZE];
+
-+ error_nomem:
-+ ret = -ENOMEM;
-+ xenbus_dev_fatal(dev, ret, "allocating device memory");
-+ error:
-+ xenkbd_remove(dev);
-+ return ret;
-+}
++#define TX_MAX_TARGET min_t(int, NET_RX_RING_SIZE, 256)
++ grant_ref_t gref_tx_head;
++ grant_ref_t grant_tx_ref[NET_TX_RING_SIZE + 1];
++ grant_ref_t gref_rx_head;
++ grant_ref_t grant_rx_ref[NET_RX_RING_SIZE];
+
-+static int xenkbd_resume(struct xenbus_device *dev)
-+{
-+ struct xenkbd_info *info = dev->dev.driver_data;
++ struct xenbus_device *xbdev;
++ int tx_ring_ref;
++ int rx_ring_ref;
++ u8 mac[ETH_ALEN];
+
-+ xenkbd_disconnect_backend(info);
-+ return xenkbd_connect_backend(dev, info);
-+}
++ unsigned long rx_pfn_array[NET_RX_RING_SIZE];
++ struct multicall_entry rx_mcl[NET_RX_RING_SIZE+1];
++ struct mmu_update rx_mmu[NET_RX_RING_SIZE];
+
-+static int xenkbd_remove(struct xenbus_device *dev)
-+{
-+ struct xenkbd_info *info = dev->dev.driver_data;
++ /* Private pointer to state internal to accelerator module */
++ void *accel_priv;
++ /* The accelerator used by this netfront device */
++ struct netfront_accelerator *accelerator;
++ /* The accelerator state for this netfront device */
++ struct netfront_accel_vif_state accel_vif_state;
++};
+
-+ xenkbd_disconnect_backend(info);
-+ input_unregister_device(info->kbd);
-+ input_unregister_device(info->ptr);
-+ free_page((unsigned long)info->page);
-+ kfree(info);
-+ return 0;
-+}
+
-+static int xenkbd_connect_backend(struct xenbus_device *dev,
-+ struct xenkbd_info *info)
-+{
-+ int ret;
-+ struct xenbus_transaction xbt;
++/* Exported Functions */
+
-+ ret = bind_listening_port_to_irqhandler(
-+ dev->otherend_id, input_handler, 0, "xenkbd", info);
-+ if (ret < 0) {
-+ xenbus_dev_fatal(dev, ret,
-+ "bind_listening_port_to_irqhandler");
-+ return ret;
-+ }
-+ info->irq = ret;
++/*
++ * Called by an accelerator plugin module when it has loaded.
++ *
++ * frontend: the string describing the accelerator, currently the module name
++ * hooks: the hooks for netfront to use to call into the accelerator
++ * version: the version of API between frontend and plugin requested
++ *
++ * return: 0 on success, <0 on error, >0 (with version supported) on
++ * version mismatch
++ */
++extern int netfront_accelerator_loaded(int version, const char *frontend,
++ struct netfront_accel_hooks *hooks);
+
-+ again:
-+ ret = xenbus_transaction_start(&xbt);
-+ if (ret) {
-+ xenbus_dev_fatal(dev, ret, "starting transaction");
-+ return ret;
-+ }
-+ ret = xenbus_printf(xbt, dev->nodename, "page-ref", "%lu",
-+ virt_to_mfn(info->page));
-+ if (ret)
-+ goto error_xenbus;
-+ ret = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
-+ irq_to_evtchn_port(info->irq));
-+ if (ret)
-+ goto error_xenbus;
-+ ret = xenbus_transaction_end(xbt, 0);
-+ if (ret) {
-+ if (ret == -EAGAIN)
-+ goto again;
-+ xenbus_dev_fatal(dev, ret, "completing transaction");
-+ return ret;
-+ }
++/*
++ * Called by an accelerator plugin module when it is about to unload.
++ *
++ * frontend: the string describing the accelerator. Must match the
++ * one passed to netfront_accelerator_loaded()
++ */
++extern void netfront_accelerator_stop(const char *frontend);
+
-+ xenbus_switch_state(dev, XenbusStateInitialised);
-+ return 0;
++/*
++ * Called by an accelerator before waking the net device's TX queue to
++ * ensure the slow path has available slots. Returns true if OK to
++ * wake, false if still busy
++ */
++extern int netfront_check_queue_ready(struct net_device *net_dev);
+
-+ error_xenbus:
-+ xenbus_transaction_end(xbt, 1);
-+ xenbus_dev_fatal(dev, ret, "writing xenstore");
-+ return ret;
-+}
+
-+static void xenkbd_disconnect_backend(struct xenkbd_info *info)
-+{
-+ if (info->irq >= 0)
-+ unbind_from_irqhandler(info->irq, info);
-+ info->irq = -1;
-+}
++/* Internal-to-netfront Functions */
+
-+static void xenkbd_backend_changed(struct xenbus_device *dev,
-+ enum xenbus_state backend_state)
-+{
-+ struct xenkbd_info *info = dev->dev.driver_data;
-+ int ret, val;
++/*
++ * Call into accelerator and check to see if it has tx space before we
++ * wake the net device's TX queue. Returns true if OK to wake, false
++ * if still busy
++ */
++extern
++int netfront_check_accelerator_queue_ready(struct net_device *dev,
++ struct netfront_info *np);
++extern
++int netfront_accelerator_call_remove(struct netfront_info *np,
++ struct xenbus_device *dev);
++extern
++int netfront_accelerator_suspend(struct netfront_info *np,
++ struct xenbus_device *dev);
++extern
++int netfront_accelerator_suspend_cancel(struct netfront_info *np,
++ struct xenbus_device *dev);
++extern
++void netfront_accelerator_resume(struct netfront_info *np,
++ struct xenbus_device *dev);
++extern
++void netfront_accelerator_call_stop_napi_irq(struct netfront_info *np,
++ struct net_device *dev);
++extern
++int netfront_accelerator_call_get_stats(struct netfront_info *np,
++ struct net_device *dev);
++extern
++void netfront_accelerator_add_watch(struct netfront_info *np);
++
++extern
++void netif_init_accel(void);
++extern
++void netif_exit_accel(void);
++
++extern
++void init_accelerator_vif(struct netfront_info *np,
++ struct xenbus_device *dev);
++#endif /* NETFRONT_H */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,17 @@
++obj-$(CONFIG_XEN_PCIDEV_BACKEND) += pciback.o
+
-+ switch (backend_state) {
-+ case XenbusStateInitialising:
-+ case XenbusStateInitialised:
-+ case XenbusStateUnknown:
-+ case XenbusStateClosed:
-+ break;
++pciback-y := pci_stub.o pciback_ops.o xenbus.o
++pciback-y += conf_space.o conf_space_header.o \
++ conf_space_capability.o \
++ conf_space_capability_vpd.o \
++ conf_space_capability_pm.o \
++ conf_space_quirks.o
++pciback-$(CONFIG_PCI_MSI) += conf_space_capability_msi.o
++pciback-$(CONFIG_XEN_PCIDEV_BACKEND_VPCI) += vpci.o
++pciback-$(CONFIG_XEN_PCIDEV_BACKEND_SLOT) += slot.o
++pciback-$(CONFIG_XEN_PCIDEV_BACKEND_PASS) += passthrough.o
++pciback-$(CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER) += controller.o
+
-+ case XenbusStateInitWait:
-+ InitWait:
-+ ret = xenbus_scanf(XBT_NIL, info->xbdev->otherend,
-+ "feature-abs-pointer", "%d", &val);
-+ if (ret < 0)
-+ val = 0;
-+ if (val) {
-+ ret = xenbus_printf(XBT_NIL, info->xbdev->nodename,
-+ "request-abs-pointer", "1");
-+ if (ret)
-+ ; /* FIXME */
-+ }
-+ xenbus_switch_state(dev, XenbusStateConnected);
-+ break;
++ifeq ($(CONFIG_XEN_PCIDEV_BE_DEBUG),y)
++EXTRA_CFLAGS += -DDEBUG
++endif
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/conf_space.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/conf_space.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,426 @@
++/*
++ * PCI Backend - Functions for creating a virtual configuration space for
++ * exported PCI Devices.
++ * It's dangerous to allow PCI Driver Domains to change their
++ * device's resources (memory, i/o ports, interrupts). We need to
++ * restrict changes to certain PCI Configuration registers:
++ * BARs, INTERRUPT_PIN, most registers in the header...
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
+
-+ case XenbusStateConnected:
-+ /*
-+ * Work around xenbus race condition: If backend goes
-+ * through InitWait to Connected fast enough, we can
-+ * get Connected twice here.
-+ */
-+ if (dev->state != XenbusStateConnected)
-+ goto InitWait; /* no InitWait seen yet, fudge it */
-+ break;
++#include <linux/kernel.h>
++#include <linux/pci.h>
++#include "pciback.h"
++#include "conf_space.h"
++#include "conf_space_quirks.h"
+
-+ case XenbusStateClosing:
-+ xenbus_frontend_closed(dev);
-+ break;
-+ }
++#define DEFINE_PCI_CONFIG(op,size,type) \
++int pciback_##op##_config_##size \
++(struct pci_dev *dev, int offset, type value, void *data) \
++{ \
++ return pci_##op##_config_##size (dev, offset, value); \
+}
+
-+static struct xenbus_device_id xenkbd_ids[] = {
-+ { "vkbd" },
-+ { "" }
-+};
-+MODULE_ALIAS("xen:vkbd");
++DEFINE_PCI_CONFIG(read, byte, u8 *)
++DEFINE_PCI_CONFIG(read, word, u16 *)
++DEFINE_PCI_CONFIG(read, dword, u32 *)
+
-+static struct xenbus_driver xenkbd = {
-+ .name = "vkbd",
-+ .owner = THIS_MODULE,
-+ .ids = xenkbd_ids,
-+ .probe = xenkbd_probe,
-+ .remove = xenkbd_remove,
-+ .resume = xenkbd_resume,
-+ .otherend_changed = xenkbd_backend_changed,
-+};
++DEFINE_PCI_CONFIG(write, byte, u8)
++DEFINE_PCI_CONFIG(write, word, u16)
++DEFINE_PCI_CONFIG(write, dword, u32)
+
-+static int __init xenkbd_init(void)
++static int conf_space_read(struct pci_dev *dev,
++ struct config_field_entry *entry, int offset,
++ u32 * value)
+{
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++ int ret = 0;
++ struct config_field *field = entry->field;
+
-+ /* Nothing to do if running in dom0. */
-+ if (is_initial_xendomain())
-+ return -ENODEV;
++ *value = 0;
+
-+ return xenbus_register_frontend(&xenkbd);
++ switch (field->size) {
++ case 1:
++ if (field->u.b.read)
++ ret = field->u.b.read(dev, offset, (u8 *) value,
++ entry->data);
++ break;
++ case 2:
++ if (field->u.w.read)
++ ret = field->u.w.read(dev, offset, (u16 *) value,
++ entry->data);
++ break;
++ case 4:
++ if (field->u.dw.read)
++ ret = field->u.dw.read(dev, offset, value, entry->data);
++ break;
++ }
++ return ret;
+}
+
-+static void __exit xenkbd_cleanup(void)
++static int conf_space_write(struct pci_dev *dev,
++ struct config_field_entry *entry, int offset,
++ u32 value)
+{
-+ return xenbus_unregister_driver(&xenkbd);
-+}
-+
-+module_init(xenkbd_init);
-+module_exit(xenkbd_cleanup);
-+
-+MODULE_LICENSE("GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/gntdev/gntdev.c tmp-linux-2.6-xen.patch/drivers/xen/gntdev/gntdev.c
---- pristine-linux-2.6.18.2/drivers/xen/gntdev/gntdev.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/gntdev/gntdev.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,973 @@
-+/******************************************************************************
-+ * gntdev.c
-+ *
-+ * Device for accessing (in user-space) pages that have been granted by other
-+ * domains.
-+ *
-+ * Copyright (c) 2006-2007, D G Murray.
-+ *
-+ * This program is distributed in the hope that it will be useful,
-+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ * GNU General Public License for more details.
-+ *
-+ * You should have received a copy of the GNU General Public License
-+ * along with this program; if not, write to the Free Software
-+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
-+ */
++ int ret = 0;
++ struct config_field *field = entry->field;
+
-+#include <asm/atomic.h>
-+#include <linux/module.h>
-+#include <linux/kernel.h>
-+#include <linux/init.h>
-+#include <linux/fs.h>
-+#include <linux/device.h>
-+#include <linux/mm.h>
-+#include <linux/mman.h>
-+#include <asm/uaccess.h>
-+#include <asm/io.h>
-+#include <xen/gnttab.h>
-+#include <asm/hypervisor.h>
-+#include <xen/balloon.h>
-+#include <xen/evtchn.h>
-+#include <xen/driver_util.h>
++ switch (field->size) {
++ case 1:
++ if (field->u.b.write)
++ ret = field->u.b.write(dev, offset, (u8) value,
++ entry->data);
++ break;
++ case 2:
++ if (field->u.w.write)
++ ret = field->u.w.write(dev, offset, (u16) value,
++ entry->data);
++ break;
++ case 4:
++ if (field->u.dw.write)
++ ret = field->u.dw.write(dev, offset, value,
++ entry->data);
++ break;
++ }
++ return ret;
++}
+
-+#include <linux/types.h>
-+#include <xen/public/gntdev.h>
++static inline u32 get_mask(int size)
++{
++ if (size == 1)
++ return 0xff;
++ else if (size == 2)
++ return 0xffff;
++ else
++ return 0xffffffff;
++}
+
++static inline int valid_request(int offset, int size)
++{
++ /* Validate request (no un-aligned requests) */
++ if ((size == 1 || size == 2 || size == 4) && (offset % size) == 0)
++ return 1;
++ return 0;
++}
+
-+#define DRIVER_AUTHOR "Derek G. Murray <Derek.Murray at cl.cam.ac.uk>"
-+#define DRIVER_DESC "User-space granted page access driver"
++static inline u32 merge_value(u32 val, u32 new_val, u32 new_val_mask,
++ int offset)
++{
++ if (offset >= 0) {
++ new_val_mask <<= (offset * 8);
++ new_val <<= (offset * 8);
++ } else {
++ new_val_mask >>= (offset * -8);
++ new_val >>= (offset * -8);
++ }
++ val = (val & ~new_val_mask) | (new_val & new_val_mask);
+
-+MODULE_LICENSE("GPL");
-+MODULE_AUTHOR(DRIVER_AUTHOR);
-+MODULE_DESCRIPTION(DRIVER_DESC);
++ return val;
++}
+
-+#define MAX_GRANTS 128
++static int pcibios_err_to_errno(int err)
++{
++ switch (err) {
++ case PCIBIOS_SUCCESSFUL:
++ return XEN_PCI_ERR_success;
++ case PCIBIOS_DEVICE_NOT_FOUND:
++ return XEN_PCI_ERR_dev_not_found;
++ case PCIBIOS_BAD_REGISTER_NUMBER:
++ return XEN_PCI_ERR_invalid_offset;
++ case PCIBIOS_FUNC_NOT_SUPPORTED:
++ return XEN_PCI_ERR_not_implemented;
++ case PCIBIOS_SET_FAILED:
++ return XEN_PCI_ERR_access_denied;
++ }
++ return err;
++}
+
-+/* A slot can be in one of three states:
-+ *
-+ * 0. GNTDEV_SLOT_INVALID:
-+ * This slot is not associated with a grant reference, and is therefore free
-+ * to be overwritten by a new grant reference.
-+ *
-+ * 1. GNTDEV_SLOT_NOT_YET_MAPPED:
-+ * This slot is associated with a grant reference (via the
-+ * IOCTL_GNTDEV_MAP_GRANT_REF ioctl), but it has not yet been mmap()-ed.
-+ *
-+ * 2. GNTDEV_SLOT_MAPPED:
-+ * This slot is associated with a grant reference, and has been mmap()-ed.
-+ */
-+typedef enum gntdev_slot_state {
-+ GNTDEV_SLOT_INVALID = 0,
-+ GNTDEV_SLOT_NOT_YET_MAPPED,
-+ GNTDEV_SLOT_MAPPED
-+} gntdev_slot_state_t;
++int pciback_config_read(struct pci_dev *dev, int offset, int size,
++ u32 * ret_val)
++{
++ int err = 0;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry;
++ struct config_field *field;
++ int req_start, req_end, field_start, field_end;
++ /* if read fails for any reason, return 0 (as if device didn't respond) */
++ u32 value = 0, tmp_val;
+
-+#define GNTDEV_INVALID_HANDLE -1
-+#define GNTDEV_FREE_LIST_INVALID -1
-+/* Each opened instance of gntdev is associated with a list of grants,
-+ * represented by an array of elements of the following type,
-+ * gntdev_grant_info_t.
-+ */
-+typedef struct gntdev_grant_info {
-+ gntdev_slot_state_t state;
-+ union {
-+ uint32_t free_list_index;
-+ struct {
-+ domid_t domid;
-+ grant_ref_t ref;
-+ grant_handle_t kernel_handle;
-+ grant_handle_t user_handle;
-+ uint64_t dev_bus_addr;
-+ } valid;
-+ } u;
-+} gntdev_grant_info_t;
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG "pciback: %s: read %d bytes at 0x%x\n",
++ pci_name(dev), size, offset);
+
-+/* Private data structure, which is stored in the file pointer for files
-+ * associated with this device.
-+ */
-+typedef struct gntdev_file_private_data {
-+
-+ /* Array of grant information. */
-+ gntdev_grant_info_t grants[MAX_GRANTS];
++ if (!valid_request(offset, size)) {
++ err = XEN_PCI_ERR_invalid_offset;
++ goto out;
++ }
+
-+ /* Read/write semaphore used to protect the grants array. */
-+ struct rw_semaphore grants_sem;
++ /* Get the real value first, then modify as appropriate */
++ switch (size) {
++ case 1:
++ err = pci_read_config_byte(dev, offset, (u8 *) & value);
++ break;
++ case 2:
++ err = pci_read_config_word(dev, offset, (u16 *) & value);
++ break;
++ case 4:
++ err = pci_read_config_dword(dev, offset, &value);
++ break;
++ }
+
-+ /* An array of indices of free slots in the grants array.
-+ * N.B. An entry in this list may temporarily have the value
-+ * GNTDEV_FREE_LIST_INVALID if the corresponding slot has been removed
-+ * from the list by the contiguous allocator, but the list has not yet
-+ * been compressed. However, this is not visible across invocations of
-+ * the device.
-+ */
-+ int32_t free_list[MAX_GRANTS];
-+
-+ /* The number of free slots in the grants array. */
-+ uint32_t free_list_size;
++ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
++ field = cfg_entry->field;
+
-+ /* Read/write semaphore used to protect the free list. */
-+ struct rw_semaphore free_list_sem;
-+
-+ /* Index of the next slot after the most recent contiguous allocation,
-+ * for use in a next-fit allocator.
-+ */
-+ uint32_t next_fit_index;
++ req_start = offset;
++ req_end = offset + size;
++ field_start = OFFSET(cfg_entry);
++ field_end = OFFSET(cfg_entry) + field->size;
+
-+ /* Used to map grants into the kernel, before mapping them into user
-+ * space.
-+ */
-+ struct page **foreign_pages;
++ if ((req_start >= field_start && req_start < field_end)
++ || (req_end > field_start && req_end <= field_end)) {
++ err = conf_space_read(dev, cfg_entry, field_start,
++ &tmp_val);
++ if (err)
++ goto out;
+
-+} gntdev_file_private_data_t;
++ value = merge_value(value, tmp_val,
++ get_mask(field->size),
++ field_start - req_start);
++ }
++ }
+
-+/* Module lifecycle operations. */
-+static int __init gntdev_init(void);
-+static void __exit gntdev_exit(void);
++ out:
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG "pciback: %s: read %d bytes at 0x%x = %x\n",
++ pci_name(dev), size, offset, value);
+
-+module_init(gntdev_init);
-+module_exit(gntdev_exit);
++ *ret_val = value;
++ return pcibios_err_to_errno(err);
++}
+
-+/* File operations. */
-+static int gntdev_open(struct inode *inode, struct file *flip);
-+static int gntdev_release(struct inode *inode, struct file *flip);
-+static int gntdev_mmap(struct file *flip, struct vm_area_struct *vma);
-+static long gntdev_ioctl(struct file *flip,
-+ unsigned int cmd, unsigned long arg);
++int pciback_config_write(struct pci_dev *dev, int offset, int size, u32 value)
++{
++ int err = 0, handled = 0;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry;
++ struct config_field *field;
++ u32 tmp_val;
++ int req_start, req_end, field_start, field_end;
+
-+static struct file_operations gntdev_fops = {
-+ .owner = THIS_MODULE,
-+ .open = gntdev_open,
-+ .release = gntdev_release,
-+ .mmap = gntdev_mmap,
-+ .unlocked_ioctl = gntdev_ioctl
-+};
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG
++ "pciback: %s: write request %d bytes at 0x%x = %x\n",
++ pci_name(dev), size, offset, value);
+
-+/* VM operations. */
-+static void gntdev_vma_close(struct vm_area_struct *vma);
-+static pte_t gntdev_clear_pte(struct vm_area_struct *vma, unsigned long addr,
-+ pte_t *ptep, int is_fullmm);
++ if (!valid_request(offset, size))
++ return XEN_PCI_ERR_invalid_offset;
+
-+static struct vm_operations_struct gntdev_vmops = {
-+ .close = gntdev_vma_close,
-+ .zap_pte = gntdev_clear_pte
-+};
++ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
++ field = cfg_entry->field;
+
-+/* Global variables. */
++ req_start = offset;
++ req_end = offset + size;
++ field_start = OFFSET(cfg_entry);
++ field_end = OFFSET(cfg_entry) + field->size;
+
-+/* The driver major number, for use when unregistering the driver. */
-+static int gntdev_major;
++ if ((req_start >= field_start && req_start < field_end)
++ || (req_end > field_start && req_end <= field_end)) {
++ tmp_val = 0;
+
-+#define GNTDEV_NAME "gntdev"
++ err = pciback_config_read(dev, field_start,
++ field->size, &tmp_val);
++ if (err)
++ break;
+
-+/* Memory mapping functions
-+ * ------------------------
-+ *
-+ * Every granted page is mapped into both kernel and user space, and the two
-+ * following functions return the respective virtual addresses of these pages.
-+ *
-+ * When shadow paging is disabled, the granted page is mapped directly into
-+ * user space; when it is enabled, it is mapped into the kernel and remapped
-+ * into user space using vm_insert_page() (see gntdev_mmap(), below).
-+ */
++ tmp_val = merge_value(tmp_val, value, get_mask(size),
++ req_start - field_start);
+
-+/* Returns the virtual address (in user space) of the @page_index'th page
-+ * in the given VM area.
-+ */
-+static inline unsigned long get_user_vaddr (struct vm_area_struct *vma,
-+ int page_index)
-+{
-+ return (unsigned long) vma->vm_start + (page_index << PAGE_SHIFT);
-+}
++ err = conf_space_write(dev, cfg_entry, field_start,
++ tmp_val);
+
-+/* Returns the virtual address (in kernel space) of the @slot_index'th page
-+ * mapped by the gntdev instance that owns the given private data struct.
-+ */
-+static inline unsigned long get_kernel_vaddr (gntdev_file_private_data_t *priv,
-+ int slot_index)
-+{
-+ unsigned long pfn;
-+ void *kaddr;
-+ pfn = page_to_pfn(priv->foreign_pages[slot_index]);
-+ kaddr = pfn_to_kaddr(pfn);
-+ return (unsigned long) kaddr;
-+}
++ /* handled is set true here, but not every byte
++ * may have been written! Properly detecting if
++ * every byte is handled is unnecessary as the
++ * flag is used to detect devices that need
++ * special helpers to work correctly.
++ */
++ handled = 1;
++ }
++ }
+
-+/* Helper functions. */
++ if (!handled && !err) {
++ /* By default, anything not specificially handled above is
++ * read-only. The permissive flag changes this behavior so
++ * that anything not specifically handled above is writable.
++ * This means that some fields may still be read-only because
++ * they have entries in the config_field list that intercept
++ * the write and do nothing. */
++ if (dev_data->permissive) {
++ switch (size) {
++ case 1:
++ err = pci_write_config_byte(dev, offset,
++ (u8) value);
++ break;
++ case 2:
++ err = pci_write_config_word(dev, offset,
++ (u16) value);
++ break;
++ case 4:
++ err = pci_write_config_dword(dev, offset,
++ (u32) value);
++ break;
++ }
++ } else if (!dev_data->warned_on_write) {
++ dev_data->warned_on_write = 1;
++ dev_warn(&dev->dev, "Driver tried to write to a "
++ "read-only configuration space field at offset "
++ "0x%x, size %d. This may be harmless, but if "
++ "you have problems with your device:\n"
++ "1) see permissive attribute in sysfs\n"
++ "2) report problems to the xen-devel "
++ "mailing list along with details of your "
++ "device obtained from lspci.\n", offset, size);
++ }
++ }
+
-+/* Adds information about a grant reference to the list of grants in the file's
-+ * private data structure. Returns non-zero on failure. On success, sets the
-+ * value of *offset to the offset that should be mmap()-ed in order to map the
-+ * grant reference.
-+ */
-+static int add_grant_reference(struct file *flip,
-+ struct ioctl_gntdev_grant_ref *op,
-+ uint64_t *offset)
++ return pcibios_err_to_errno(err);
++}
++
++void pciback_config_free_dyn_fields(struct pci_dev *dev)
+{
-+ gntdev_file_private_data_t *private_data
-+ = (gntdev_file_private_data_t *) flip->private_data;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry, *t;
++ struct config_field *field;
+
-+ uint32_t slot_index;
++ dev_dbg(&dev->dev,
++ "free-ing dynamically allocated virtual configuration space fields\n");
+
-+ if (unlikely(private_data->free_list_size == 0)) {
-+ return -ENOMEM;
-+ }
++ list_for_each_entry_safe(cfg_entry, t, &dev_data->config_fields, list) {
++ field = cfg_entry->field;
+
-+ slot_index = private_data->free_list[--private_data->free_list_size];
++ if (field->clean) {
++ field->clean(field);
+
-+ /* Copy the grant information into file's private data. */
-+ private_data->grants[slot_index].state = GNTDEV_SLOT_NOT_YET_MAPPED;
-+ private_data->grants[slot_index].u.valid.domid = op->domid;
-+ private_data->grants[slot_index].u.valid.ref = op->ref;
++ if (cfg_entry->data)
++ kfree(cfg_entry->data);
+
-+ /* The offset is calculated as the index of the chosen entry in the
-+ * file's private data's array of grant information. This is then
-+ * shifted to give an offset into the virtual "file address space".
-+ */
-+ *offset = slot_index << PAGE_SHIFT;
++ list_del(&cfg_entry->list);
++ kfree(cfg_entry);
++ }
+
-+ return 0;
++ }
+}
+
-+/* Adds the @count grant references to the contiguous range in the slot array
-+ * beginning at @first_slot. It is assumed that @first_slot was returned by a
-+ * previous invocation of find_contiguous_free_range(), during the same
-+ * invocation of the driver.
-+ */
-+static int add_grant_references(struct file *flip,
-+ int count,
-+ struct ioctl_gntdev_grant_ref *ops,
-+ uint32_t first_slot)
++void pciback_config_reset_dev(struct pci_dev *dev)
+{
-+ gntdev_file_private_data_t *private_data
-+ = (gntdev_file_private_data_t *) flip->private_data;
-+ int i;
-+
-+ for (i = 0; i < count; ++i) {
-+
-+ /* First, mark the slot's entry in the free list as invalid. */
-+ int free_list_index =
-+ private_data->grants[first_slot+i].u.free_list_index;
-+ private_data->free_list[free_list_index] =
-+ GNTDEV_FREE_LIST_INVALID;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry;
++ struct config_field *field;
+
-+ /* Now, update the slot. */
-+ private_data->grants[first_slot+i].state =
-+ GNTDEV_SLOT_NOT_YET_MAPPED;
-+ private_data->grants[first_slot+i].u.valid.domid =
-+ ops[i].domid;
-+ private_data->grants[first_slot+i].u.valid.ref = ops[i].ref;
-+ }
++ dev_dbg(&dev->dev, "resetting virtual configuration space\n");
+
-+ return 0;
-+}
++ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
++ field = cfg_entry->field;
+
-+/* Scans through the free list for @flip, removing entries that are marked as
-+ * GNTDEV_SLOT_INVALID. This will reduce the recorded size of the free list to
-+ * the number of valid entries.
-+ */
-+static void compress_free_list(struct file *flip)
-+{
-+ gntdev_file_private_data_t *private_data
-+ = (gntdev_file_private_data_t *) flip->private_data;
-+ int i, j = 0, old_size;
-+
-+ old_size = private_data->free_list_size;
-+ for (i = 0; i < old_size; ++i) {
-+ if (private_data->free_list[i] != GNTDEV_FREE_LIST_INVALID) {
-+ private_data->free_list[j] =
-+ private_data->free_list[i];
-+ ++j;
-+ } else {
-+ --private_data->free_list_size;
-+ }
++ if (field->reset)
++ field->reset(dev, OFFSET(cfg_entry), cfg_entry->data);
+ }
+}
+
-+/* Searches the grant array in the private data of @flip for a range of
-+ * @num_slots contiguous slots in the GNTDEV_SLOT_INVALID state.
-+ *
-+ * Returns the index of the first slot if a range is found, otherwise -ENOMEM.
-+ */
-+static int find_contiguous_free_range(struct file *flip,
-+ uint32_t num_slots)
++void pciback_config_free_dev(struct pci_dev *dev)
+{
-+ gntdev_file_private_data_t *private_data
-+ = (gntdev_file_private_data_t *) flip->private_data;
-+
-+ int i;
-+ int start_index = private_data->next_fit_index;
-+ int range_start = 0, range_length;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry, *t;
++ struct config_field *field;
+
-+ if (private_data->free_list_size < num_slots) {
-+ return -ENOMEM;
-+ }
++ dev_dbg(&dev->dev, "free-ing virtual configuration space fields\n");
+
-+ /* First search from the start_index to the end of the array. */
-+ range_length = 0;
-+ for (i = start_index; i < MAX_GRANTS; ++i) {
-+ if (private_data->grants[i].state == GNTDEV_SLOT_INVALID) {
-+ if (range_length == 0) {
-+ range_start = i;
-+ }
-+ ++range_length;
-+ if (range_length == num_slots) {
-+ return range_start;
-+ }
-+ }
-+ }
-+
-+ /* Now search from the start of the array to the start_index. */
-+ range_length = 0;
-+ for (i = 0; i < start_index; ++i) {
-+ if (private_data->grants[i].state == GNTDEV_SLOT_INVALID) {
-+ if (range_length == 0) {
-+ range_start = i;
-+ }
-+ ++range_length;
-+ if (range_length == num_slots) {
-+ return range_start;
-+ }
-+ }
++ list_for_each_entry_safe(cfg_entry, t, &dev_data->config_fields, list) {
++ list_del(&cfg_entry->list);
++
++ field = cfg_entry->field;
++
++ if (field->release)
++ field->release(dev, OFFSET(cfg_entry), cfg_entry->data);
++
++ kfree(cfg_entry);
+ }
-+
-+ return -ENOMEM;
+}
+
-+/* Interface functions. */
-+
-+/* Initialises the driver. Called when the module is loaded. */
-+static int __init gntdev_init(void)
++int pciback_config_add_field_offset(struct pci_dev *dev,
++ struct config_field *field,
++ unsigned int base_offset)
+{
-+ struct class *class;
-+ struct class_device *device;
++ int err = 0;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry;
++ void *tmp;
+
-+ if (!is_running_on_xen()) {
-+ printk(KERN_ERR "You must be running Xen to use gntdev\n");
-+ return -ENODEV;
++ cfg_entry = kmalloc(sizeof(*cfg_entry), GFP_KERNEL);
++ if (!cfg_entry) {
++ err = -ENOMEM;
++ goto out;
+ }
+
-+ gntdev_major = register_chrdev(0, GNTDEV_NAME, &gntdev_fops);
-+ if (gntdev_major < 0)
-+ {
-+ printk(KERN_ERR "Could not register gntdev device\n");
-+ return -ENOMEM;
-+ }
++ cfg_entry->data = NULL;
++ cfg_entry->field = field;
++ cfg_entry->base_offset = base_offset;
+
-+ /* Note that if the sysfs code fails, we will still initialise the
-+ * device, and output the major number so that the device can be
-+ * created manually using mknod.
-+ */
-+ if ((class = get_xen_class()) == NULL) {
-+ printk(KERN_ERR "Error setting up xen_class\n");
-+ printk(KERN_ERR "gntdev created with major number = %d\n",
-+ gntdev_major);
-+ return 0;
-+ }
++ /* silently ignore duplicate fields */
++ err = pciback_field_is_dup(dev,OFFSET(cfg_entry));
++ if (err)
++ goto out;
++
++ if (field->init) {
++ tmp = field->init(dev, OFFSET(cfg_entry));
++
++ if (IS_ERR(tmp)) {
++ err = PTR_ERR(tmp);
++ goto out;
++ }
+
-+ device = class_device_create(class, NULL, MKDEV(gntdev_major, 0),
-+ NULL, GNTDEV_NAME);
-+ if (IS_ERR(device)) {
-+ printk(KERN_ERR "Error creating gntdev device in xen_class\n");
-+ printk(KERN_ERR "gntdev created with major number = %d\n",
-+ gntdev_major);
-+ return 0;
++ cfg_entry->data = tmp;
+ }
+
-+ return 0;
-+}
++ dev_dbg(&dev->dev, "added config field at offset 0x%02x\n",
++ OFFSET(cfg_entry));
++ list_add_tail(&cfg_entry->list, &dev_data->config_fields);
+
-+/* Cleans up and unregisters the driver. Called when the driver is unloaded.
-+ */
-+static void __exit gntdev_exit(void)
-+{
-+ struct class *class;
-+ if ((class = get_xen_class()) != NULL)
-+ class_device_destroy(class, MKDEV(gntdev_major, 0));
-+ unregister_chrdev(gntdev_major, GNTDEV_NAME);
++ out:
++ if (err)
++ kfree(cfg_entry);
++
++ return err;
+}
+
-+/* Called when the device is opened. */
-+static int gntdev_open(struct inode *inode, struct file *flip)
++/* This sets up the device's virtual configuration space to keep track of
++ * certain registers (like the base address registers (BARs) so that we can
++ * keep the client from manipulating them directly.
++ */
++int pciback_config_init_dev(struct pci_dev *dev)
+{
-+ gntdev_file_private_data_t *private_data;
-+ int i;
-+
-+ try_module_get(THIS_MODULE);
-+
-+ /* Allocate space for the per-instance private data. */
-+ private_data = kmalloc(sizeof(*private_data), GFP_KERNEL);
-+ if (!private_data)
-+ goto nomem_out;
++ int err = 0;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
+
-+ /* Allocate space for the kernel-mapping of granted pages. */
-+ private_data->foreign_pages =
-+ alloc_empty_pages_and_pagevec(MAX_GRANTS);
-+ if (!private_data->foreign_pages)
-+ goto nomem_out2;
++ dev_dbg(&dev->dev, "initializing virtual configuration space\n");
+
-+ /* Initialise the free-list, which contains all slots at first.
-+ */
-+ for (i = 0; i < MAX_GRANTS; ++i) {
-+ private_data->free_list[MAX_GRANTS - i - 1] = i;
-+ private_data->grants[i].state = GNTDEV_SLOT_INVALID;
-+ private_data->grants[i].u.free_list_index = MAX_GRANTS - i - 1;
-+ }
-+ private_data->free_list_size = MAX_GRANTS;
-+ private_data->next_fit_index = 0;
++ INIT_LIST_HEAD(&dev_data->config_fields);
+
-+ init_rwsem(&private_data->grants_sem);
-+ init_rwsem(&private_data->free_list_sem);
++ err = pciback_config_header_add_fields(dev);
++ if (err)
++ goto out;
+
-+ flip->private_data = private_data;
++ err = pciback_config_capability_add_fields(dev);
++ if (err)
++ goto out;
+
-+ return 0;
++ err = pciback_config_quirks_init(dev);
+
-+nomem_out2:
-+ kfree(private_data);
-+nomem_out:
-+ return -ENOMEM;
++ out:
++ return err;
+}
+
-+/* Called when the device is closed.
-+ */
-+static int gntdev_release(struct inode *inode, struct file *flip)
++int pciback_config_init(void)
+{
-+ if (flip->private_data) {
-+ gntdev_file_private_data_t *private_data =
-+ (gntdev_file_private_data_t *) flip->private_data;
-+ if (private_data->foreign_pages) {
-+ free_empty_pages_and_pagevec
-+ (private_data->foreign_pages, MAX_GRANTS);
-+ }
-+ kfree(private_data);
-+ }
-+ module_put(THIS_MODULE);
-+ return 0;
++ return pciback_config_capability_init();
+}
-+
-+/* Called when an attempt is made to mmap() the device. The private data from
-+ * @flip contains the list of grant references that can be mapped. The vm_pgoff
-+ * field of @vma contains the index into that list that refers to the grant
-+ * reference that will be mapped. Only mappings that are a multiple of
-+ * PAGE_SIZE are handled.
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/conf_space.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/conf_space.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,126 @@
++/*
++ * PCI Backend - Common data structures for overriding the configuration space
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
+ */
-+static int gntdev_mmap (struct file *flip, struct vm_area_struct *vma)
-+{
-+ struct gnttab_map_grant_ref op;
-+ unsigned long slot_index = vma->vm_pgoff;
-+ unsigned long kernel_vaddr, user_vaddr;
-+ uint32_t size = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
-+ uint64_t ptep;
-+ int ret;
-+ int flags;
-+ int i;
-+ struct page *page;
-+ gntdev_file_private_data_t *private_data = flip->private_data;
+
-+ if (unlikely(!private_data)) {
-+ printk(KERN_ERR "File's private data is NULL.\n");
-+ return -EINVAL;
-+ }
++#ifndef __XEN_PCIBACK_CONF_SPACE_H__
++#define __XEN_PCIBACK_CONF_SPACE_H__
+
-+ if (unlikely((size <= 0) || (size + slot_index) > MAX_GRANTS)) {
-+ printk(KERN_ERR "Invalid number of pages or offset"
-+ "(num_pages = %d, first_slot = %ld).\n",
-+ size, slot_index);
-+ return -ENXIO;
-+ }
++#include <linux/list.h>
++#include <linux/err.h>
+
-+ if ((vma->vm_flags & VM_WRITE) && !(vma->vm_flags & VM_SHARED)) {
-+ printk(KERN_ERR "Writable mappings must be shared.\n");
-+ return -EINVAL;
-+ }
++/* conf_field_init can return an errno in a ptr with ERR_PTR() */
++typedef void *(*conf_field_init) (struct pci_dev * dev, int offset);
++typedef void (*conf_field_reset) (struct pci_dev * dev, int offset, void *data);
++typedef void (*conf_field_free) (struct pci_dev * dev, int offset, void *data);
+
-+ /* Slots must be in the NOT_YET_MAPPED state. */
-+ down_write(&private_data->grants_sem);
-+ for (i = 0; i < size; ++i) {
-+ if (private_data->grants[slot_index + i].state !=
-+ GNTDEV_SLOT_NOT_YET_MAPPED) {
-+ printk(KERN_ERR "Slot (index = %ld) is in the wrong "
-+ "state (%d).\n", slot_index + i,
-+ private_data->grants[slot_index + i].state);
-+ up_write(&private_data->grants_sem);
-+ return -EINVAL;
-+ }
-+ }
++typedef int (*conf_dword_write) (struct pci_dev * dev, int offset, u32 value,
++ void *data);
++typedef int (*conf_word_write) (struct pci_dev * dev, int offset, u16 value,
++ void *data);
++typedef int (*conf_byte_write) (struct pci_dev * dev, int offset, u8 value,
++ void *data);
++typedef int (*conf_dword_read) (struct pci_dev * dev, int offset, u32 * value,
++ void *data);
++typedef int (*conf_word_read) (struct pci_dev * dev, int offset, u16 * value,
++ void *data);
++typedef int (*conf_byte_read) (struct pci_dev * dev, int offset, u8 * value,
++ void *data);
+
-+ /* Install the hook for unmapping. */
-+ vma->vm_ops = &gntdev_vmops;
-+
-+ /* The VM area contains pages from another VM. */
-+ vma->vm_flags |= VM_FOREIGN;
-+ vma->vm_private_data = kzalloc(size * sizeof(struct page_struct *),
-+ GFP_KERNEL);
-+ if (vma->vm_private_data == NULL) {
-+ printk(KERN_ERR "Couldn't allocate mapping structure for VM "
-+ "area.\n");
-+ return -ENOMEM;
-+ }
++/* These are the fields within the configuration space which we
++ * are interested in intercepting reads/writes to and changing their
++ * values.
++ */
++struct config_field {
++ unsigned int offset;
++ unsigned int size;
++ unsigned int mask;
++ conf_field_init init;
++ conf_field_reset reset;
++ conf_field_free release;
++ void (*clean) (struct config_field * field);
++ union {
++ struct {
++ conf_dword_write write;
++ conf_dword_read read;
++ } dw;
++ struct {
++ conf_word_write write;
++ conf_word_read read;
++ } w;
++ struct {
++ conf_byte_write write;
++ conf_byte_read read;
++ } b;
++ } u;
++ struct list_head list;
++};
+
-+ /* This flag prevents Bad PTE errors when the memory is unmapped. */
-+ vma->vm_flags |= VM_RESERVED;
++struct config_field_entry {
++ struct list_head list;
++ struct config_field *field;
++ unsigned int base_offset;
++ void *data;
++};
+
-+ /* This flag prevents this VM area being copied on a fork(). A better
-+ * behaviour might be to explicitly carry out the appropriate mappings
-+ * on fork(), but I don't know if there's a hook for this.
-+ */
-+ vma->vm_flags |= VM_DONTCOPY;
++#define OFFSET(cfg_entry) ((cfg_entry)->base_offset+(cfg_entry)->field->offset)
+
-+#ifdef CONFIG_X86
-+ /* This flag ensures that the page tables are not unpinned before the
-+ * VM area is unmapped. Therefore Xen still recognises the PTE as
-+ * belonging to an L1 pagetable, and the grant unmap operation will
-+ * succeed, even if the process does not exit cleanly.
-+ */
-+ vma->vm_mm->context.has_foreign_mappings = 1;
-+#endif
++/* Add fields to a device - the add_fields macro expects to get a pointer to
++ * the first entry in an array (of which the ending is marked by size==0)
++ */
++int pciback_config_add_field_offset(struct pci_dev *dev,
++ struct config_field *field,
++ unsigned int offset);
+
-+ for (i = 0; i < size; ++i) {
++static inline int pciback_config_add_field(struct pci_dev *dev,
++ struct config_field *field)
++{
++ return pciback_config_add_field_offset(dev, field, 0);
++}
+
-+ flags = GNTMAP_host_map;
-+ if (!(vma->vm_flags & VM_WRITE))
-+ flags |= GNTMAP_readonly;
++static inline int pciback_config_add_fields(struct pci_dev *dev,
++ struct config_field *field)
++{
++ int i, err = 0;
++ for (i = 0; field[i].size != 0; i++) {
++ err = pciback_config_add_field(dev, &field[i]);
++ if (err)
++ break;
++ }
++ return err;
++}
+
-+ kernel_vaddr = get_kernel_vaddr(private_data, slot_index + i);
-+ user_vaddr = get_user_vaddr(vma, i);
-+ page = pfn_to_page(__pa(kernel_vaddr) >> PAGE_SHIFT);
++static inline int pciback_config_add_fields_offset(struct pci_dev *dev,
++ struct config_field *field,
++ unsigned int offset)
++{
++ int i, err = 0;
++ for (i = 0; field[i].size != 0; i++) {
++ err = pciback_config_add_field_offset(dev, &field[i], offset);
++ if (err)
++ break;
++ }
++ return err;
++}
+
-+ gnttab_set_map_op(&op, kernel_vaddr, flags,
-+ private_data->grants[slot_index+i]
-+ .u.valid.ref,
-+ private_data->grants[slot_index+i]
-+ .u.valid.domid);
++/* Read/Write the real configuration space */
++int pciback_read_config_byte(struct pci_dev *dev, int offset, u8 * value,
++ void *data);
++int pciback_read_config_word(struct pci_dev *dev, int offset, u16 * value,
++ void *data);
++int pciback_read_config_dword(struct pci_dev *dev, int offset, u32 * value,
++ void *data);
++int pciback_write_config_byte(struct pci_dev *dev, int offset, u8 value,
++ void *data);
++int pciback_write_config_word(struct pci_dev *dev, int offset, u16 value,
++ void *data);
++int pciback_write_config_dword(struct pci_dev *dev, int offset, u32 value,
++ void *data);
+
-+ /* Carry out the mapping of the grant reference. */
-+ ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref,
-+ &op, 1);
-+ BUG_ON(ret);
-+ if (op.status) {
-+ printk(KERN_ERR "Error mapping the grant reference "
-+ "into the kernel (%d). domid = %d; ref = %d\n",
-+ op.status,
-+ private_data->grants[slot_index+i]
-+ .u.valid.domid,
-+ private_data->grants[slot_index+i]
-+ .u.valid.ref);
-+ goto undo_map_out;
-+ }
++int pciback_config_capability_init(void);
+
-+ /* Store a reference to the page that will be mapped into user
-+ * space.
-+ */
-+ ((struct page **) vma->vm_private_data)[i] = page;
++int pciback_config_header_add_fields(struct pci_dev *dev);
++int pciback_config_capability_add_fields(struct pci_dev *dev);
+
-+ /* Mark mapped page as reserved. */
-+ SetPageReserved(page);
++#endif /* __XEN_PCIBACK_CONF_SPACE_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/conf_space_capability.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/conf_space_capability.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,71 @@
++/*
++ * PCI Backend - Handles the virtual fields found on the capability lists
++ * in the configuration space.
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
+
-+ /* Record the grant handle, for use in the unmap operation. */
-+ private_data->grants[slot_index+i].u.valid.kernel_handle =
-+ op.handle;
-+ private_data->grants[slot_index+i].u.valid.dev_bus_addr =
-+ op.dev_bus_addr;
-+
-+ private_data->grants[slot_index+i].state = GNTDEV_SLOT_MAPPED;
-+ private_data->grants[slot_index+i].u.valid.user_handle =
-+ GNTDEV_INVALID_HANDLE;
++#include <linux/kernel.h>
++#include <linux/pci.h>
++#include "pciback.h"
++#include "conf_space.h"
++#include "conf_space_capability.h"
+
-+ /* Now perform the mapping to user space. */
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++static LIST_HEAD(capabilities);
+
-+ /* NOT USING SHADOW PAGE TABLES. */
-+ /* In this case, we map the grant(s) straight into user
-+ * space.
-+ */
++static struct config_field caplist_header[] = {
++ {
++ .offset = PCI_CAP_LIST_ID,
++ .size = 2, /* encompass PCI_CAP_LIST_ID & PCI_CAP_LIST_NEXT */
++ .u.w.read = pciback_read_config_word,
++ .u.w.write = NULL,
++ },
++ {
++ .size = 0,
++ },
++};
+
-+ /* Get the machine address of the PTE for the user
-+ * page.
-+ */
-+ if ((ret = create_lookup_pte_addr(vma->vm_mm,
-+ vma->vm_start
-+ + (i << PAGE_SHIFT),
-+ &ptep)))
-+ {
-+ printk(KERN_ERR "Error obtaining PTE pointer "
-+ "(%d).\n", ret);
-+ goto undo_map_out;
-+ }
-+
-+ /* Configure the map operation. */
-+
-+ /* The reference is to be used by host CPUs. */
-+ flags = GNTMAP_host_map;
-+
-+ /* Specifies a user space mapping. */
-+ flags |= GNTMAP_application_map;
-+
-+ /* The map request contains the machine address of the
-+ * PTE to update.
-+ */
-+ flags |= GNTMAP_contains_pte;
-+
-+ if (!(vma->vm_flags & VM_WRITE))
-+ flags |= GNTMAP_readonly;
++static inline void register_capability(struct pciback_config_capability *cap)
++{
++ list_add_tail(&cap->cap_list, &capabilities);
++}
+
-+ gnttab_set_map_op(&op, ptep, flags,
-+ private_data->grants[slot_index+i]
-+ .u.valid.ref,
-+ private_data->grants[slot_index+i]
-+ .u.valid.domid);
++int pciback_config_capability_add_fields(struct pci_dev *dev)
++{
++ int err = 0;
++ struct pciback_config_capability *cap;
++ int cap_offset;
+
-+ /* Carry out the mapping of the grant reference. */
-+ ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref,
-+ &op, 1);
-+ BUG_ON(ret);
-+ if (op.status) {
-+ printk(KERN_ERR "Error mapping the grant "
-+ "reference into user space (%d). domid "
-+ "= %d; ref = %d\n", op.status,
-+ private_data->grants[slot_index+i].u
-+ .valid.domid,
-+ private_data->grants[slot_index+i].u
-+ .valid.ref);
-+ goto undo_map_out;
-+ }
-+
-+ /* Record the grant handle, for use in the unmap
-+ * operation.
-+ */
-+ private_data->grants[slot_index+i].u.
-+ valid.user_handle = op.handle;
++ list_for_each_entry(cap, &capabilities, cap_list) {
++ cap_offset = pci_find_capability(dev, cap->capability);
++ if (cap_offset) {
++ dev_dbg(&dev->dev, "Found capability 0x%x at 0x%x\n",
++ cap->capability, cap_offset);
+
-+ /* Update p2m structure with the new mapping. */
-+ set_phys_to_machine(__pa(kernel_vaddr) >> PAGE_SHIFT,
-+ FOREIGN_FRAME(private_data->
-+ grants[slot_index+i]
-+ .u.valid.dev_bus_addr
-+ >> PAGE_SHIFT));
-+ } else {
-+ /* USING SHADOW PAGE TABLES. */
-+ /* In this case, we simply insert the page into the VM
-+ * area. */
-+ ret = vm_insert_page(vma, user_vaddr, page);
++ err = pciback_config_add_fields_offset(dev,
++ caplist_header,
++ cap_offset);
++ if (err)
++ goto out;
++ err = pciback_config_add_fields_offset(dev,
++ cap->fields,
++ cap_offset);
++ if (err)
++ goto out;
+ }
-+
+ }
+
-+ up_write(&private_data->grants_sem);
-+ return 0;
-+
-+undo_map_out:
-+ /* If we have a mapping failure, the unmapping will be taken care of
-+ * by do_mmap_pgoff(), which will eventually call gntdev_clear_pte().
-+ * All we need to do here is free the vma_private_data.
-+ */
-+ kfree(vma->vm_private_data);
-+
-+ /* THIS IS VERY UNPLEASANT: do_mmap_pgoff() will set the vma->vm_file
-+ * to NULL on failure. However, we need this in gntdev_clear_pte() to
-+ * unmap the grants. Therefore, we smuggle a reference to the file's
-+ * private data in the VM area's private data pointer.
-+ */
-+ vma->vm_private_data = private_data;
-+
-+ up_write(&private_data->grants_sem);
-+
-+ return -ENOMEM;
++ out:
++ return err;
+}
+
-+static pte_t gntdev_clear_pte(struct vm_area_struct *vma, unsigned long addr,
-+ pte_t *ptep, int is_fullmm)
-+{
-+ int slot_index, ret;
-+ pte_t copy;
-+ struct gnttab_unmap_grant_ref op;
-+ gntdev_file_private_data_t *private_data;
-+
-+ /* THIS IS VERY UNPLEASANT: do_mmap_pgoff() will set the vma->vm_file
-+ * to NULL on failure. However, we need this in gntdev_clear_pte() to
-+ * unmap the grants. Therefore, we smuggle a reference to the file's
-+ * private data in the VM area's private data pointer.
-+ */
-+ if (vma->vm_file) {
-+ private_data = (gntdev_file_private_data_t *)
-+ vma->vm_file->private_data;
-+ } else if (vma->vm_private_data) {
-+ private_data = (gntdev_file_private_data_t *)
-+ vma->vm_private_data;
-+ } else {
-+ private_data = NULL; /* gcc warning */
-+ BUG();
-+ }
++extern struct pciback_config_capability pciback_config_capability_vpd;
++extern struct pciback_config_capability pciback_config_capability_pm;
+
-+ /* Copy the existing value of the PTE for returning. */
-+ copy = *ptep;
++int pciback_config_capability_init(void)
++{
++ register_capability(&pciback_config_capability_vpd);
++ register_capability(&pciback_config_capability_pm);
+
-+ /* Calculate the grant relating to this PTE. */
-+ slot_index = vma->vm_pgoff + ((addr - vma->vm_start) >> PAGE_SHIFT);
++ return 0;
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/conf_space_capability.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/conf_space_capability.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,23 @@
++/*
++ * PCI Backend - Data structures for special overlays for structures on
++ * the capability list.
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
+
-+ /* Only unmap grants if the slot has been mapped. This could be being
-+ * called from a failing mmap().
-+ */
-+ if (private_data->grants[slot_index].state == GNTDEV_SLOT_MAPPED) {
++#ifndef __PCIBACK_CONFIG_CAPABILITY_H__
++#define __PCIBACK_CONFIG_CAPABILITY_H__
+
-+ /* First, we clear the user space mapping, if it has been made.
-+ */
-+ if (private_data->grants[slot_index].u.valid.user_handle !=
-+ GNTDEV_INVALID_HANDLE &&
-+ !xen_feature(XENFEAT_auto_translated_physmap)) {
-+ /* NOT USING SHADOW PAGE TABLES. */
-+ gnttab_set_unmap_op(&op, virt_to_machine(ptep),
-+ GNTMAP_contains_pte,
-+ private_data->grants[slot_index]
-+ .u.valid.user_handle);
-+ ret = HYPERVISOR_grant_table_op(
-+ GNTTABOP_unmap_grant_ref, &op, 1);
-+ BUG_ON(ret);
-+ if (op.status)
-+ printk("User unmap grant status = %d\n",
-+ op.status);
-+ } else {
-+ /* USING SHADOW PAGE TABLES. */
-+ pte_clear_full(vma->vm_mm, addr, ptep, is_fullmm);
-+ }
++#include <linux/pci.h>
++#include <linux/list.h>
+
-+ /* Finally, we unmap the grant from kernel space. */
-+ gnttab_set_unmap_op(&op,
-+ get_kernel_vaddr(private_data, slot_index),
-+ GNTMAP_host_map,
-+ private_data->grants[slot_index].u.valid
-+ .kernel_handle);
-+ ret = HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref,
-+ &op, 1);
-+ BUG_ON(ret);
-+ if (op.status)
-+ printk("Kernel unmap grant status = %d\n", op.status);
++struct pciback_config_capability {
++ struct list_head cap_list;
+
++ int capability;
+
-+ /* Return slot to the not-yet-mapped state, so that it may be
-+ * mapped again, or removed by a subsequent ioctl.
-+ */
-+ private_data->grants[slot_index].state =
-+ GNTDEV_SLOT_NOT_YET_MAPPED;
++ /* If the device has the capability found above, add these fields */
++ struct config_field *fields;
++};
+
-+ /* Invalidate the physical to machine mapping for this page. */
-+ set_phys_to_machine(__pa(get_kernel_vaddr(private_data,
-+ slot_index))
-+ >> PAGE_SHIFT, INVALID_P2M_ENTRY);
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/conf_space_capability_msi.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/conf_space_capability_msi.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,79 @@
++/*
++ * PCI Backend -- Configuration overlay for MSI capability
++ */
++#include <linux/pci.h>
++#include <linux/slab.h>
++#include "conf_space.h"
++#include "conf_space_capability.h"
++#include <xen/interface/io/pciif.h>
++#include "pciback.h"
+
-+ } else {
-+ pte_clear_full(vma->vm_mm, addr, ptep, is_fullmm);
++int pciback_enable_msi(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op)
++{
++ int otherend = pdev->xdev->otherend_id;
++ int status;
++
++ status = pci_enable_msi(dev);
++
++ if (status) {
++ printk("error enable msi for guest %x status %x\n", otherend, status);
++ op->value = 0;
++ return XEN_PCI_ERR_op_failed;
+ }
+
-+ return copy;
++ op->value = dev->irq;
++ return 0;
+}
+
-+/* "Destructor" for a VM area.
-+ */
-+static void gntdev_vma_close(struct vm_area_struct *vma) {
-+ if (vma->vm_private_data) {
-+ kfree(vma->vm_private_data);
-+ }
++int pciback_disable_msi(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op)
++{
++ pci_disable_msi(dev);
++
++ op->value = dev->irq;
++ return 0;
+}
+
-+/* Called when an ioctl is made on the device.
-+ */
-+static long gntdev_ioctl(struct file *flip,
-+ unsigned int cmd, unsigned long arg)
++int pciback_enable_msix(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op)
+{
-+ int rc = 0;
-+ gntdev_file_private_data_t *private_data =
-+ (gntdev_file_private_data_t *) flip->private_data;
++ int i, result;
++ struct msix_entry *entries;
+
-+ switch (cmd) {
-+ case IOCTL_GNTDEV_MAP_GRANT_REF:
-+ {
-+ struct ioctl_gntdev_map_grant_ref op;
-+ down_write(&private_data->grants_sem);
-+ down_write(&private_data->free_list_sem);
++ if (op->value > SH_INFO_MAX_VEC)
++ return -EINVAL;
+
-+ if ((rc = copy_from_user(&op, (void __user *) arg,
-+ sizeof(op)))) {
-+ rc = -EFAULT;
-+ goto map_out;
-+ }
-+ if (unlikely(op.count <= 0)) {
-+ rc = -EINVAL;
-+ goto map_out;
-+ }
++ entries = kmalloc(op->value * sizeof(*entries), GFP_KERNEL);
++ if (entries == NULL)
++ return -ENOMEM;
+
-+ if (op.count == 1) {
-+ if ((rc = add_grant_reference(flip, &op.refs[0],
-+ &op.index)) < 0) {
-+ printk(KERN_ERR "Adding grant reference "
-+ "failed (%d).\n", rc);
-+ goto map_out;
-+ }
-+ } else {
-+ struct ioctl_gntdev_grant_ref *refs, *u;
-+ refs = kmalloc(op.count * sizeof(*refs), GFP_KERNEL);
-+ if (!refs) {
-+ rc = -ENOMEM;
-+ goto map_out;
-+ }
-+ u = ((struct ioctl_gntdev_map_grant_ref *)arg)->refs;
-+ if ((rc = copy_from_user(refs,
-+ (void __user *)u,
-+ sizeof(*refs) * op.count))) {
-+ printk(KERN_ERR "Copying refs from user failed"
-+ " (%d).\n", rc);
-+ rc = -EINVAL;
-+ goto map_out;
-+ }
-+ if ((rc = find_contiguous_free_range(flip, op.count))
-+ < 0) {
-+ printk(KERN_ERR "Finding contiguous range "
-+ "failed (%d).\n", rc);
-+ kfree(refs);
-+ goto map_out;
-+ }
-+ op.index = rc << PAGE_SHIFT;
-+ if ((rc = add_grant_references(flip, op.count,
-+ refs, rc))) {
-+ printk(KERN_ERR "Adding grant references "
-+ "failed (%d).\n", rc);
-+ kfree(refs);
-+ goto map_out;
-+ }
-+ compress_free_list(flip);
-+ kfree(refs);
-+ }
-+ if ((rc = copy_to_user((void __user *) arg,
-+ &op,
-+ sizeof(op)))) {
-+ printk(KERN_ERR "Copying result back to user failed "
-+ "(%d)\n", rc);
-+ rc = -EFAULT;
-+ goto map_out;
-+ }
-+ map_out:
-+ up_write(&private_data->grants_sem);
-+ up_write(&private_data->free_list_sem);
-+ return rc;
++ for (i = 0; i < op->value; i++) {
++ entries[i].entry = op->msix_entries[i].entry;
++ entries[i].vector = op->msix_entries[i].vector;
+ }
-+ case IOCTL_GNTDEV_UNMAP_GRANT_REF:
-+ {
-+ struct ioctl_gntdev_unmap_grant_ref op;
-+ int i, start_index;
-+
-+ down_write(&private_data->grants_sem);
-+ down_write(&private_data->free_list_sem);
+
-+ if ((rc = copy_from_user(&op,
-+ (void __user *) arg,
-+ sizeof(op)))) {
-+ rc = -EFAULT;
-+ goto unmap_out;
-+ }
++ result = pci_enable_msix(dev, entries, op->value);
+
-+ start_index = op.index >> PAGE_SHIFT;
++ for (i = 0; i < op->value; i++) {
++ op->msix_entries[i].entry = entries[i].entry;
++ op->msix_entries[i].vector = entries[i].vector;
++ }
+
-+ /* First, check that all pages are in the NOT_YET_MAPPED
-+ * state.
-+ */
-+ for (i = 0; i < op.count; ++i) {
-+ if (unlikely
-+ (private_data->grants[start_index + i].state
-+ != GNTDEV_SLOT_NOT_YET_MAPPED)) {
-+ if (private_data->grants[start_index + i].state
-+ == GNTDEV_SLOT_INVALID) {
-+ printk(KERN_ERR
-+ "Tried to remove an invalid "
-+ "grant at offset 0x%x.",
-+ (start_index + i)
-+ << PAGE_SHIFT);
-+ rc = -EINVAL;
-+ } else {
-+ printk(KERN_ERR
-+ "Tried to remove a grant which "
-+ "is currently mmap()-ed at "
-+ "offset 0x%x.",
-+ (start_index + i)
-+ << PAGE_SHIFT);
-+ rc = -EBUSY;
-+ }
-+ goto unmap_out;
-+ }
-+ }
++ kfree(entries);
+
-+ /* Unmap pages and add them to the free list.
-+ */
-+ for (i = 0; i < op.count; ++i) {
-+ private_data->grants[start_index+i].state =
-+ GNTDEV_SLOT_INVALID;
-+ private_data->grants[start_index+i].u.free_list_index =
-+ private_data->free_list_size;
-+ private_data->free_list[private_data->free_list_size] =
-+ start_index + i;
-+ ++private_data->free_list_size;
-+ }
-+ compress_free_list(flip);
++ op->value = result;
+
-+ unmap_out:
-+ up_write(&private_data->grants_sem);
-+ up_write(&private_data->free_list_sem);
-+ return rc;
-+ }
-+ case IOCTL_GNTDEV_GET_OFFSET_FOR_VADDR:
-+ {
-+ struct ioctl_gntdev_get_offset_for_vaddr op;
-+ struct vm_area_struct *vma;
-+ unsigned long vaddr;
++ return result;
++}
+
-+ if ((rc = copy_from_user(&op,
-+ (void __user *) arg,
-+ sizeof(op)))) {
-+ rc = -EFAULT;
-+ goto get_offset_out;
-+ }
-+ vaddr = (unsigned long)op.vaddr;
++int pciback_disable_msix(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op)
++{
+
-+ down_read(¤t->mm->mmap_sem);
-+ vma = find_vma(current->mm, vaddr);
-+ if (vma == NULL) {
-+ rc = -EFAULT;
-+ goto get_offset_unlock_out;
-+ }
-+ if ((!vma->vm_ops) || (vma->vm_ops != &gntdev_vmops)) {
-+ printk(KERN_ERR "The vaddr specified does not belong "
-+ "to a gntdev instance: %#lx\n", vaddr);
-+ rc = -EFAULT;
-+ goto get_offset_unlock_out;
-+ }
-+ if (vma->vm_start != vaddr) {
-+ printk(KERN_ERR "The vaddr specified in an "
-+ "IOCTL_GNTDEV_GET_OFFSET_FOR_VADDR must be at "
-+ "the start of the VM area. vma->vm_start = "
-+ "%#lx; vaddr = %#lx\n",
-+ vma->vm_start, vaddr);
-+ rc = -EFAULT;
-+ goto get_offset_unlock_out;
-+ }
-+ op.offset = vma->vm_pgoff << PAGE_SHIFT;
-+ op.count = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
-+ up_read(¤t->mm->mmap_sem);
-+ if ((rc = copy_to_user((void __user *) arg,
-+ &op,
-+ sizeof(op)))) {
-+ rc = -EFAULT;
-+ goto get_offset_out;
-+ }
-+ goto get_offset_out;
-+ get_offset_unlock_out:
-+ up_read(¤t->mm->mmap_sem);
-+ get_offset_out:
-+ return rc;
-+ }
-+ default:
-+ return -ENOIOCTLCMD;
-+ }
++ pci_disable_msix(dev);
+
++ op->value = dev->irq;
+ return 0;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/gntdev/Makefile tmp-linux-2.6-xen.patch/drivers/xen/gntdev/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/gntdev/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/gntdev/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1 @@
-+obj-y := gntdev.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/Kconfig tmp-linux-2.6-xen.patch/drivers/xen/Kconfig
---- pristine-linux-2.6.18.2/drivers/xen/Kconfig 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/Kconfig 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,277 @@
-+#
-+# This Kconfig describe xen options
-+#
+
-+mainmenu "Xen Configuration"
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/conf_space_capability_pm.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/conf_space_capability_pm.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,128 @@
++/*
++ * PCI Backend - Configuration space overlay for power management
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
+
-+config XEN
-+ bool
-+ default y if X86_XEN || X86_64_XEN
-+ help
-+ This is the Linux Xen port.
++#include <linux/pci.h>
++#include "conf_space.h"
++#include "conf_space_capability.h"
+
-+if XEN
-+config XEN_INTERFACE_VERSION
-+ hex
-+ default 0x00030205
++static int pm_caps_read(struct pci_dev *dev, int offset, u16 *value,
++ void *data)
++{
++ int err;
++ u16 real_value;
+
-+menu "XEN"
++ err = pci_read_config_word(dev, offset, &real_value);
++ if (err)
++ goto out;
+
-+config XEN_PRIVILEGED_GUEST
-+ bool "Privileged Guest (domain 0)"
-+ depends XEN
-+ default n
-+ help
-+ Support for privileged operation (domain 0)
++ *value = real_value & ~PCI_PM_CAP_PME_MASK;
+
-+config XEN_UNPRIVILEGED_GUEST
-+ bool
-+ default !XEN_PRIVILEGED_GUEST
++ out:
++ return err;
++}
+
-+config XEN_PRIVCMD
-+ bool
-+ depends on PROC_FS
-+ default y
++/* PM_OK_BITS specifies the bits that the driver domain is allowed to change.
++ * Can't allow driver domain to enable PMEs - they're shared */
++#define PM_OK_BITS (PCI_PM_CTRL_PME_STATUS|PCI_PM_CTRL_DATA_SEL_MASK)
+
-+config XEN_XENBUS_DEV
-+ bool
-+ depends on PROC_FS
-+ default y
++static int pm_ctrl_write(struct pci_dev *dev, int offset, u16 new_value,
++ void *data)
++{
++ int err;
++ u16 old_value;
++ pci_power_t new_state, old_state;
+
-+config XEN_BACKEND
-+ tristate "Backend driver support"
-+ default y
-+ help
-+ Support for backend device drivers that provide I/O services
-+ to other virtual machines.
++ err = pci_read_config_word(dev, offset, &old_value);
++ if (err)
++ goto out;
+
-+config XEN_BLKDEV_BACKEND
-+ tristate "Block-device backend driver"
-+ depends on XEN_BACKEND
-+ default y
-+ help
-+ The block-device backend driver allows the kernel to export its
-+ block devices to other guests via a high-performance shared-memory
-+ interface.
++ old_state = (pci_power_t)(old_value & PCI_PM_CTRL_STATE_MASK);
++ new_state = (pci_power_t)(new_value & PCI_PM_CTRL_STATE_MASK);
+
-+config XEN_BLKDEV_TAP
-+ tristate "Block-device tap backend driver"
-+ depends on XEN_BACKEND
-+ default XEN_PRIVILEGED_GUEST
-+ help
-+ The block tap driver is an alternative to the block back driver
-+ and allows VM block requests to be redirected to userspace through
-+ a device interface. The tap allows user-space development of
-+ high-performance block backends, where disk images may be implemented
-+ as files, in memory, or on other hosts across the network. This
-+ driver can safely coexist with the existing blockback driver.
++ new_value &= PM_OK_BITS;
++ if ((old_value & PM_OK_BITS) != new_value) {
++ new_value = (old_value & ~PM_OK_BITS) | new_value;
++ err = pci_write_config_word(dev, offset, new_value);
++ if (err)
++ goto out;
++ }
+
-+config XEN_NETDEV_BACKEND
-+ tristate "Network-device backend driver"
-+ depends on XEN_BACKEND && NET
-+ default y
-+ help
-+ The network-device backend driver allows the kernel to export its
-+ network devices to other guests via a high-performance shared-memory
-+ interface.
++ /* Let pci core handle the power management change */
++ dev_dbg(&dev->dev, "set power state to %x\n", new_state);
++ err = pci_set_power_state(dev, new_state);
++ if (err) {
++ err = PCIBIOS_SET_FAILED;
++ goto out;
++ }
+
-+config XEN_NETDEV_PIPELINED_TRANSMITTER
-+ bool "Pipelined transmitter (DANGEROUS)"
-+ depends on XEN_NETDEV_BACKEND
-+ default n
-+ help
-+ If the net backend is a dumb domain, such as a transparent Ethernet
-+ bridge with no local IP interface, it is safe to say Y here to get
-+ slightly lower network overhead.
-+ If the backend has a local IP interface; or may be doing smart things
-+ like reassembling packets to perform firewall filtering; or if you
-+ are unsure; or if you experience network hangs when this option is
-+ enabled; then you must say N here.
++ /*
++ * Device may lose PCI config info on D3->D0 transition. This
++ * is a problem for some guests which will not reset BARs. Even
++ * those that have a go will be foiled by our BAR-write handler
++ * which will discard the write! Since Linux won't re-init
++ * the config space automatically in all cases, we do it here.
++ * Future: Should we re-initialise all first 64 bytes of config space?
++ */
++ if (new_state == PCI_D0 &&
++ (old_state == PCI_D3hot || old_state == PCI_D3cold) &&
++ !(old_value & PCI_PM_CTRL_NO_SOFT_RESET))
++ pci_restore_bars(dev);
+
-+config XEN_NETDEV_LOOPBACK
-+ tristate "Network-device loopback driver"
-+ depends on XEN_NETDEV_BACKEND
-+ default y
-+ help
-+ A two-interface loopback device to emulate a local netfront-netback
-+ connection.
++ out:
++ return err;
++}
+
-+config XEN_PCIDEV_BACKEND
-+ tristate "PCI-device backend driver"
-+ depends on PCI && XEN_BACKEND
-+ default XEN_PRIVILEGED_GUEST
-+ help
-+ The PCI device backend driver allows the kernel to export arbitrary
-+ PCI devices to other guests. If you select this to be a module, you
-+ will need to make sure no other driver has bound to the device(s)
-+ you want to make visible to other guests.
++/* Ensure PMEs are disabled */
++static void *pm_ctrl_init(struct pci_dev *dev, int offset)
++{
++ int err;
++ u16 value;
+
-+choice
-+ prompt "PCI Backend Mode"
-+ depends on XEN_PCIDEV_BACKEND
-+ default XEN_PCIDEV_BACKEND_VPCI if !IA64
-+ default XEN_PCIDEV_BACKEND_CONTROLLER if IA64
++ err = pci_read_config_word(dev, offset, &value);
++ if (err)
++ goto out;
+
-+config XEN_PCIDEV_BACKEND_VPCI
-+ bool "Virtual PCI"
-+ ---help---
-+ This PCI Backend hides the true PCI topology and makes the frontend
-+ think there is a single PCI bus with only the exported devices on it.
-+ For example, a device at 03:05.0 will be re-assigned to 00:00.0. A
-+ second device at 02:1a.1 will be re-assigned to 00:01.1.
++ if (value & PCI_PM_CTRL_PME_ENABLE) {
++ value &= ~PCI_PM_CTRL_PME_ENABLE;
++ err = pci_write_config_word(dev, offset, value);
++ }
+
-+config XEN_PCIDEV_BACKEND_PASS
-+ bool "Passthrough"
-+ ---help---
-+ This PCI Backend provides a real view of the PCI topology to the
-+ frontend (for example, a device at 06:01.b will still appear at
-+ 06:01.b to the frontend). This is similar to how Xen 2.0.x exposed
-+ PCI devices to its driver domains. This may be required for drivers
-+ which depend on finding their hardward in certain bus/slot
-+ locations.
++ out:
++ return ERR_PTR(err);
++}
+
-+config XEN_PCIDEV_BACKEND_SLOT
-+ bool "Slot"
-+ ---help---
-+ This PCI Backend hides the true PCI topology and makes the frontend
-+ think there is a single PCI bus with only the exported devices on it.
-+ Contrary to the virtual PCI backend, a function becomes a new slot.
-+ For example, a device at 03:05.2 will be re-assigned to 00:00.0. A
-+ second device at 02:1a.1 will be re-assigned to 00:01.0.
++static struct config_field caplist_pm[] = {
++ {
++ .offset = PCI_PM_PMC,
++ .size = 2,
++ .u.w.read = pm_caps_read,
++ },
++ {
++ .offset = PCI_PM_CTRL,
++ .size = 2,
++ .init = pm_ctrl_init,
++ .u.w.read = pciback_read_config_word,
++ .u.w.write = pm_ctrl_write,
++ },
++ {
++ .offset = PCI_PM_PPB_EXTENSIONS,
++ .size = 1,
++ .u.b.read = pciback_read_config_byte,
++ },
++ {
++ .offset = PCI_PM_DATA_REGISTER,
++ .size = 1,
++ .u.b.read = pciback_read_config_byte,
++ },
++ {
++ .size = 0,
++ },
++};
++
++struct pciback_config_capability pciback_config_capability_pm = {
++ .capability = PCI_CAP_ID_PM,
++ .fields = caplist_pm,
++};
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/conf_space_capability_vpd.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/conf_space_capability_vpd.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,42 @@
++/*
++ * PCI Backend - Configuration space overlay for Vital Product Data
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++
++#include <linux/pci.h>
++#include "conf_space.h"
++#include "conf_space_capability.h"
++
++static int vpd_address_write(struct pci_dev *dev, int offset, u16 value,
++ void *data)
++{
++ /* Disallow writes to the vital product data */
++ if (value & PCI_VPD_ADDR_F)
++ return PCIBIOS_SET_FAILED;
++ else
++ return pci_write_config_word(dev, offset, value);
++}
++
++static struct config_field caplist_vpd[] = {
++ {
++ .offset = PCI_VPD_ADDR,
++ .size = 2,
++ .u.w.read = pciback_read_config_word,
++ .u.w.write = vpd_address_write,
++ },
++ {
++ .offset = PCI_VPD_DATA,
++ .size = 4,
++ .u.dw.read = pciback_read_config_dword,
++ .u.dw.write = NULL,
++ },
++ {
++ .size = 0,
++ },
++};
++
++struct pciback_config_capability pciback_config_capability_vpd = {
++ .capability = PCI_CAP_ID_VPD,
++ .fields = caplist_vpd,
++};
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/conf_space_header.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/conf_space_header.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,323 @@
++/*
++ * PCI Backend - Handles the virtual fields in the configuration space headers.
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
+
-+config XEN_PCIDEV_BACKEND_CONTROLLER
-+ bool "Controller"
-+ depends on IA64
-+ ---help---
-+ This PCI backend virtualizes the PCI bus topology by providing a
-+ virtual bus per PCI root device. Devices which are physically under
-+ the same root bus will appear on the same virtual bus. For systems
-+ with complex I/O addressing, this is the only backend which supports
-+ extended I/O port spaces and MMIO translation offsets. This backend
-+ also supports slot virtualization. For example, a device at
-+ 0000:01:02.1 will be re-assigned to 0000:00:00.0. A second device
-+ at 0000:02:05.0 (behind a P2P bridge on bus 0000:01) will be
-+ re-assigned to 0000:00:01.0. A third device at 0000:16:05.0 (under
-+ a different PCI root bus) will be re-assigned to 0000:01:00.0.
++#include <linux/kernel.h>
++#include <linux/pci.h>
++#include "pciback.h"
++#include "conf_space.h"
+
-+endchoice
++struct pci_bar_info {
++ u32 val;
++ u32 len_val;
++ int which;
++};
+
-+config XEN_PCIDEV_BE_DEBUG
-+ bool "PCI Backend Debugging"
-+ depends on XEN_PCIDEV_BACKEND
-+ default n
++#define is_enable_cmd(value) ((value)&(PCI_COMMAND_MEMORY|PCI_COMMAND_IO))
++#define is_master_cmd(value) ((value)&PCI_COMMAND_MASTER)
+
-+config XEN_TPMDEV_BACKEND
-+ tristate "TPM-device backend driver"
-+ depends on XEN_BACKEND
-+ default n
-+ help
-+ The TPM-device backend driver
++static int command_write(struct pci_dev *dev, int offset, u16 value, void *data)
++{
++ int err;
+
-+config XEN_BLKDEV_FRONTEND
-+ tristate "Block-device frontend driver"
-+ depends on XEN
-+ default y
-+ help
-+ The block-device frontend driver allows the kernel to access block
-+ devices mounted within another guest OS. Unless you are building a
-+ dedicated device-driver domain, or your master control domain
-+ (domain 0), then you almost certainly want to say Y here.
++ if (!dev->is_enabled && is_enable_cmd(value)) {
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG "pciback: %s: enable\n",
++ pci_name(dev));
++ err = pci_enable_device(dev);
++ if (err)
++ return err;
++ } else if (dev->is_enabled && !is_enable_cmd(value)) {
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG "pciback: %s: disable\n",
++ pci_name(dev));
++ pci_disable_device(dev);
++ }
+
-+config XEN_NETDEV_FRONTEND
-+ tristate "Network-device frontend driver"
-+ depends on XEN && NET
-+ default y
-+ help
-+ The network-device frontend driver allows the kernel to access
-+ network interfaces within another guest OS. Unless you are building a
-+ dedicated device-driver domain, or your master control domain
-+ (domain 0), then you almost certainly want to say Y here.
++ if (!dev->is_busmaster && is_master_cmd(value)) {
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG "pciback: %s: set bus master\n",
++ pci_name(dev));
++ pci_set_master(dev);
++ }
+
-+config XEN_FRAMEBUFFER
-+ tristate "Framebuffer-device frontend driver"
-+ depends on XEN && FB
-+ select FB_CFB_FILLRECT
-+ select FB_CFB_COPYAREA
-+ select FB_CFB_IMAGEBLIT
-+ default y
-+ help
-+ The framebuffer-device frontend drivers allows the kernel to create a
-+ virtual framebuffer. This framebuffer can be viewed in another
-+ domain. Unless this domain has access to a real video card, you
-+ probably want to say Y here.
++ if (value & PCI_COMMAND_INVALIDATE) {
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG
++ "pciback: %s: enable memory-write-invalidate\n",
++ pci_name(dev));
++ err = pci_set_mwi(dev);
++ if (err) {
++ printk(KERN_WARNING
++ "pciback: %s: cannot enable memory-write-invalidate (%d)\n",
++ pci_name(dev), err);
++ value &= ~PCI_COMMAND_INVALIDATE;
++ }
++ }
+
-+config XEN_KEYBOARD
-+ tristate "Keyboard-device frontend driver"
-+ depends on XEN && XEN_FRAMEBUFFER && INPUT
-+ default y
-+ help
-+ The keyboard-device frontend driver allows the kernel to create a
-+ virtual keyboard. This keyboard can then be driven by another
-+ domain. If you've said Y to CONFIG_XEN_FRAMEBUFFER, you probably
-+ want to say Y here.
++ return pci_write_config_word(dev, offset, value);
++}
+
-+config XEN_SCRUB_PAGES
-+ bool "Scrub memory before freeing it to Xen"
-+ default y
-+ help
-+ Erase memory contents before freeing it back to Xen's global
-+ pool. This ensures that any secrets contained within that
-+ memory (e.g., private keys) cannot be found by other guests that
-+ may be running on the machine. Most people will want to say Y here.
-+ If security is not a concern then you may increase performance by
-+ saying N.
++static int rom_write(struct pci_dev *dev, int offset, u32 value, void *data)
++{
++ struct pci_bar_info *bar = data;
+
-+config XEN_DISABLE_SERIAL
-+ bool "Disable serial port drivers"
-+ default y
-+ help
-+ Disable serial port drivers, allowing the Xen console driver
-+ to provide a serial console at ttyS0.
++ if (unlikely(!bar)) {
++ printk(KERN_WARNING "pciback: driver data not found for %s\n",
++ pci_name(dev));
++ return XEN_PCI_ERR_op_failed;
++ }
+
-+config XEN_SYSFS
-+ tristate "Export Xen attributes in sysfs"
-+ depends on SYSFS
-+ select SYS_HYPERVISOR
-+ default y
-+ help
-+ Xen hypervisor attributes will show up under /sys/hypervisor/.
++ /* A write to obtain the length must happen as a 32-bit write.
++ * This does not (yet) support writing individual bytes
++ */
++ if (value == ~PCI_ROM_ADDRESS_ENABLE)
++ bar->which = 1;
++ else {
++ u32 tmpval;
++ pci_read_config_dword(dev, offset, &tmpval);
++ if (tmpval != bar->val && value == bar->val) {
++ /* Allow restoration of bar value. */
++ pci_write_config_dword(dev, offset, bar->val);
++ }
++ bar->which = 0;
++ }
+
-+choice
-+ prompt "Xen version compatibility"
-+ default XEN_COMPAT_030002_AND_LATER
++ /* Do we need to support enabling/disabling the rom address here? */
+
-+ config XEN_COMPAT_030002_AND_LATER
-+ bool "3.0.2 and later"
++ return 0;
++}
+
-+ config XEN_COMPAT_030004_AND_LATER
-+ bool "3.0.4 and later"
++/* For the BARs, only allow writes which write ~0 or
++ * the correct resource information
++ * (Needed for when the driver probes the resource usage)
++ */
++static int bar_write(struct pci_dev *dev, int offset, u32 value, void *data)
++{
++ struct pci_bar_info *bar = data;
+
-+ config XEN_COMPAT_LATEST_ONLY
-+ bool "no compatibility code"
++ if (unlikely(!bar)) {
++ printk(KERN_WARNING "pciback: driver data not found for %s\n",
++ pci_name(dev));
++ return XEN_PCI_ERR_op_failed;
++ }
+
-+endchoice
++ /* A write to obtain the length must happen as a 32-bit write.
++ * This does not (yet) support writing individual bytes
++ */
++ if (value == ~0)
++ bar->which = 1;
++ else {
++ u32 tmpval;
++ pci_read_config_dword(dev, offset, &tmpval);
++ if (tmpval != bar->val && value == bar->val) {
++ /* Allow restoration of bar value. */
++ pci_write_config_dword(dev, offset, bar->val);
++ }
++ bar->which = 0;
++ }
+
-+config XEN_COMPAT
-+ hex
-+ default 0xffffff if XEN_COMPAT_LATEST_ONLY
-+ default 0x030004 if XEN_COMPAT_030004_AND_LATER
-+ default 0x030002 if XEN_COMPAT_030002_AND_LATER
-+ default 0
++ return 0;
++}
+
-+endmenu
++static int bar_read(struct pci_dev *dev, int offset, u32 * value, void *data)
++{
++ struct pci_bar_info *bar = data;
+
-+config HAVE_IRQ_IGNORE_UNHANDLED
-+ bool
-+ default y
++ if (unlikely(!bar)) {
++ printk(KERN_WARNING "pciback: driver data not found for %s\n",
++ pci_name(dev));
++ return XEN_PCI_ERR_op_failed;
++ }
+
-+config NO_IDLE_HZ
-+ bool
-+ default y
++ *value = bar->which ? bar->len_val : bar->val;
+
-+config XEN_SMPBOOT
-+ bool
-+ default y
-+ depends on SMP
++ return 0;
++}
+
-+endif
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/Makefile tmp-linux-2.6-xen.patch/drivers/xen/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/Makefile 2007-07-30 16:35:11.000000000 +0200
-@@ -0,0 +1,20 @@
-+obj-y += core/
-+obj-y += console/
-+obj-y += evtchn/
-+obj-y += privcmd/
-+obj-y += xenbus/
-+obj-y += gntdev/
-+obj-y += balloon/
-+obj-y += char/
++static inline void read_dev_bar(struct pci_dev *dev,
++ struct pci_bar_info *bar_info, int offset,
++ u32 len_mask)
++{
++ pci_read_config_dword(dev, offset, &bar_info->val);
++ pci_write_config_dword(dev, offset, len_mask);
++ pci_read_config_dword(dev, offset, &bar_info->len_val);
++ pci_write_config_dword(dev, offset, bar_info->val);
++}
+
-+obj-y += util.o
-+obj-$(CONFIG_XEN_BLKDEV_BACKEND) += blkback/
-+obj-$(CONFIG_XEN_BLKDEV_TAP) += blktap/
-+obj-$(CONFIG_XEN_NETDEV_BACKEND) += netback/
-+obj-$(CONFIG_XEN_TPMDEV_BACKEND) += tpmback/
-+obj-$(CONFIG_XEN_BLKDEV_FRONTEND) += blkfront/
-+obj-$(CONFIG_XEN_NETDEV_FRONTEND) += netfront/
-+obj-$(CONFIG_XEN_PCIDEV_BACKEND) += pciback/
-+obj-$(CONFIG_XEN_PCIDEV_FRONTEND) += pcifront/
-+obj-$(CONFIG_XEN_FRAMEBUFFER) += fbfront/
-+obj-$(CONFIG_XEN_KEYBOARD) += fbfront/
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/netback/common.h tmp-linux-2.6-xen.patch/drivers/xen/netback/common.h
---- pristine-linux-2.6.18.2/drivers/xen/netback/common.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/netback/common.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,157 @@
-+/******************************************************************************
-+ * arch/xen/drivers/netif/backend/common.h
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++static void *bar_init(struct pci_dev *dev, int offset)
++{
++ struct pci_bar_info *bar = kmalloc(sizeof(*bar), GFP_KERNEL);
+
-+#ifndef __NETIF__BACKEND__COMMON_H__
-+#define __NETIF__BACKEND__COMMON_H__
++ if (!bar)
++ return ERR_PTR(-ENOMEM);
+
-+#include <linux/version.h>
-+#include <linux/module.h>
-+#include <linux/interrupt.h>
-+#include <linux/slab.h>
-+#include <linux/ip.h>
-+#include <linux/in.h>
-+#include <linux/netdevice.h>
-+#include <linux/etherdevice.h>
-+#include <linux/wait.h>
-+#include <xen/evtchn.h>
-+#include <xen/interface/io/netif.h>
-+#include <asm/io.h>
-+#include <asm/pgalloc.h>
-+#include <xen/interface/grant_table.h>
-+#include <xen/gnttab.h>
-+#include <xen/driver_util.h>
++ read_dev_bar(dev, bar, offset, ~0);
++ bar->which = 0;
+
-+#define DPRINTK(_f, _a...) \
-+ pr_debug("(file=%s, line=%d) " _f, \
-+ __FILE__ , __LINE__ , ## _a )
-+#define IPRINTK(fmt, args...) \
-+ printk(KERN_INFO "xen_net: " fmt, ##args)
-+#define WPRINTK(fmt, args...) \
-+ printk(KERN_WARNING "xen_net: " fmt, ##args)
++ return bar;
++}
+
-+typedef struct netif_st {
-+ /* Unique identifier for this interface. */
-+ domid_t domid;
-+ unsigned int handle;
++static void *rom_init(struct pci_dev *dev, int offset)
++{
++ struct pci_bar_info *bar = kmalloc(sizeof(*bar), GFP_KERNEL);
+
-+ u8 fe_dev_addr[6];
++ if (!bar)
++ return ERR_PTR(-ENOMEM);
+
-+ /* Physical parameters of the comms window. */
-+ grant_handle_t tx_shmem_handle;
-+ grant_ref_t tx_shmem_ref;
-+ grant_handle_t rx_shmem_handle;
-+ grant_ref_t rx_shmem_ref;
-+ unsigned int irq;
++ read_dev_bar(dev, bar, offset, ~PCI_ROM_ADDRESS_ENABLE);
++ bar->which = 0;
+
-+ /* The shared rings and indexes. */
-+ netif_tx_back_ring_t tx;
-+ netif_rx_back_ring_t rx;
-+ struct vm_struct *tx_comms_area;
-+ struct vm_struct *rx_comms_area;
++ return bar;
++}
+
-+ /* Set of features that can be turned on in dev->features. */
-+ int features;
++static void bar_reset(struct pci_dev *dev, int offset, void *data)
++{
++ struct pci_bar_info *bar = data;
+
-+ /* Internal feature information. */
-+ u8 can_queue:1; /* can queue packets for receiver? */
-+ u8 copying_receiver:1; /* copy packets to receiver? */
++ bar->which = 0;
++}
+
-+ /* Allow netif_be_start_xmit() to peek ahead in the rx request ring. */
-+ RING_IDX rx_req_cons_peek;
++static void bar_release(struct pci_dev *dev, int offset, void *data)
++{
++ kfree(data);
++}
+
-+ /* Transmit shaping: allow 'credit_bytes' every 'credit_usec'. */
-+ unsigned long credit_bytes;
-+ unsigned long credit_usec;
-+ unsigned long remaining_credit;
-+ struct timer_list credit_timeout;
++static int interrupt_read(struct pci_dev *dev, int offset, u8 * value,
++ void *data)
++{
++ *value = (u8) dev->irq;
+
-+ /* Enforce draining of the transmit queue. */
-+ struct timer_list tx_queue_timeout;
++ return 0;
++}
+
-+ /* Miscellaneous private stuff. */
-+ struct list_head list; /* scheduling list */
-+ atomic_t refcnt;
-+ struct net_device *dev;
-+ struct net_device_stats stats;
++static int bist_write(struct pci_dev *dev, int offset, u8 value, void *data)
++{
++ u8 cur_value;
++ int err;
+
-+ unsigned int carrier;
++ err = pci_read_config_byte(dev, offset, &cur_value);
++ if (err)
++ goto out;
+
-+ wait_queue_head_t waiting_to_free;
-+} netif_t;
++ if ((cur_value & ~PCI_BIST_START) == (value & ~PCI_BIST_START)
++ || value == PCI_BIST_START)
++ err = pci_write_config_byte(dev, offset, value);
+
-+/*
-+ * Implement our own carrier flag: the network stack's version causes delays
-+ * when the carrier is re-enabled (in particular, dev_activate() may not
-+ * immediately be called, which can cause packet loss; also the etherbridge
-+ * can be rather lazy in activating its port).
-+ */
-+#define netback_carrier_on(netif) ((netif)->carrier = 1)
-+#define netback_carrier_off(netif) ((netif)->carrier = 0)
-+#define netback_carrier_ok(netif) ((netif)->carrier)
++ out:
++ return err;
++}
+
-+#define NET_TX_RING_SIZE __RING_SIZE((netif_tx_sring_t *)0, PAGE_SIZE)
-+#define NET_RX_RING_SIZE __RING_SIZE((netif_rx_sring_t *)0, PAGE_SIZE)
++static struct config_field header_common[] = {
++ {
++ .offset = PCI_COMMAND,
++ .size = 2,
++ .u.w.read = pciback_read_config_word,
++ .u.w.write = command_write,
++ },
++ {
++ .offset = PCI_INTERRUPT_LINE,
++ .size = 1,
++ .u.b.read = interrupt_read,
++ },
++ {
++ .offset = PCI_INTERRUPT_PIN,
++ .size = 1,
++ .u.b.read = pciback_read_config_byte,
++ },
++ {
++ /* Any side effects of letting driver domain control cache line? */
++ .offset = PCI_CACHE_LINE_SIZE,
++ .size = 1,
++ .u.b.read = pciback_read_config_byte,
++ .u.b.write = pciback_write_config_byte,
++ },
++ {
++ .offset = PCI_LATENCY_TIMER,
++ .size = 1,
++ .u.b.read = pciback_read_config_byte,
++ },
++ {
++ .offset = PCI_BIST,
++ .size = 1,
++ .u.b.read = pciback_read_config_byte,
++ .u.b.write = bist_write,
++ },
++ {
++ .size = 0,
++ },
++};
+
-+void netif_disconnect(netif_t *netif);
++#define CFG_FIELD_BAR(reg_offset) \
++ { \
++ .offset = reg_offset, \
++ .size = 4, \
++ .init = bar_init, \
++ .reset = bar_reset, \
++ .release = bar_release, \
++ .u.dw.read = bar_read, \
++ .u.dw.write = bar_write, \
++ }
+
-+netif_t *netif_alloc(domid_t domid, unsigned int handle);
-+int netif_map(netif_t *netif, unsigned long tx_ring_ref,
-+ unsigned long rx_ring_ref, unsigned int evtchn);
++#define CFG_FIELD_ROM(reg_offset) \
++ { \
++ .offset = reg_offset, \
++ .size = 4, \
++ .init = rom_init, \
++ .reset = bar_reset, \
++ .release = bar_release, \
++ .u.dw.read = bar_read, \
++ .u.dw.write = rom_write, \
++ }
+
-+#define netif_get(_b) (atomic_inc(&(_b)->refcnt))
-+#define netif_put(_b) \
-+ do { \
-+ if ( atomic_dec_and_test(&(_b)->refcnt) ) \
-+ wake_up(&(_b)->waiting_to_free); \
-+ } while (0)
++static struct config_field header_0[] = {
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_0),
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_1),
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_2),
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_3),
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_4),
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_5),
++ CFG_FIELD_ROM(PCI_ROM_ADDRESS),
++ {
++ .size = 0,
++ },
++};
+
-+void netif_xenbus_init(void);
++static struct config_field header_1[] = {
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_0),
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_1),
++ CFG_FIELD_ROM(PCI_ROM_ADDRESS1),
++ {
++ .size = 0,
++ },
++};
+
-+#define netif_schedulable(netif) \
-+ (netif_running((netif)->dev) && netback_carrier_ok(netif))
++int pciback_config_header_add_fields(struct pci_dev *dev)
++{
++ int err;
+
-+void netif_schedule_work(netif_t *netif);
-+void netif_deschedule_work(netif_t *netif);
++ err = pciback_config_add_fields(dev, header_common);
++ if (err)
++ goto out;
+
-+int netif_be_start_xmit(struct sk_buff *skb, struct net_device *dev);
-+struct net_device_stats *netif_be_get_stats(struct net_device *dev);
-+irqreturn_t netif_be_int(int irq, void *dev_id, struct pt_regs *regs);
++ switch (dev->hdr_type) {
++ case PCI_HEADER_TYPE_NORMAL:
++ err = pciback_config_add_fields(dev, header_0);
++ break;
++
++ case PCI_HEADER_TYPE_BRIDGE:
++ err = pciback_config_add_fields(dev, header_1);
++ break;
+
-+static inline int netbk_can_queue(struct net_device *dev)
-+{
-+ netif_t *netif = netdev_priv(dev);
-+ return netif->can_queue;
-+}
++ default:
++ err = -EINVAL;
++ printk(KERN_ERR "pciback: %s: Unsupported header type %d!\n",
++ pci_name(dev), dev->hdr_type);
++ break;
++ }
+
-+static inline int netbk_can_sg(struct net_device *dev)
-+{
-+ netif_t *netif = netdev_priv(dev);
-+ return netif->features & NETIF_F_SG;
++ out:
++ return err;
+}
-+
-+#endif /* __NETIF__BACKEND__COMMON_H__ */
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/netback/interface.c tmp-linux-2.6-xen.patch/drivers/xen/netback/interface.c
---- pristine-linux-2.6.18.2/drivers/xen/netback/interface.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/netback/interface.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,336 @@
-+/******************************************************************************
-+ * arch/xen/drivers/netif/backend/interface.c
-+ *
-+ * Network-device interface management.
-+ *
-+ * Copyright (c) 2004-2005, Keir Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/conf_space_quirks.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/conf_space_quirks.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,126 @@
++/*
++ * PCI Backend - Handle special overlays for broken devices.
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ * Author: Chris Bookholt <hap10 at epoch.ncsc.mil>
+ */
+
-+#include "common.h"
-+#include <linux/ethtool.h>
-+#include <linux/rtnetlink.h>
++#include <linux/kernel.h>
++#include <linux/pci.h>
++#include "pciback.h"
++#include "conf_space.h"
++#include "conf_space_quirks.h"
+
-+/*
-+ * Module parameter 'queue_length':
-+ *
-+ * Enables queuing in the network stack when a client has run out of receive
-+ * descriptors. Although this feature can improve receive bandwidth by avoiding
-+ * packet loss, it can also result in packets sitting in the 'tx_queue' for
-+ * unbounded time. This is bad if those packets hold onto foreign resources.
-+ * For example, consider a packet that holds onto resources belonging to the
-+ * guest for which it is queued (e.g., packet received on vif1.0, destined for
-+ * vif1.1 which is not activated in the guest): in this situation the guest
-+ * will never be destroyed, unless vif1.1 is taken down. To avoid this, we
-+ * run a timer (tx_queue_timeout) to drain the queue when the interface is
-+ * blocked.
-+ */
-+static unsigned long netbk_queue_length = 32;
-+module_param_named(queue_length, netbk_queue_length, ulong, 0);
++LIST_HEAD(pciback_quirks);
+
-+static void __netif_up(netif_t *netif)
++struct pciback_config_quirk *pciback_find_quirk(struct pci_dev *dev)
+{
-+ enable_irq(netif->irq);
-+ netif_schedule_work(netif);
++ struct pciback_config_quirk *tmp_quirk;
++
++ list_for_each_entry(tmp_quirk, &pciback_quirks, quirks_list)
++ if (pci_match_id(&tmp_quirk->devid, dev))
++ goto out;
++ tmp_quirk = NULL;
++ printk(KERN_DEBUG
++ "quirk didn't match any device pciback knows about\n");
++ out:
++ return tmp_quirk;
+}
+
-+static void __netif_down(netif_t *netif)
++static inline void register_quirk(struct pciback_config_quirk *quirk)
+{
-+ disable_irq(netif->irq);
-+ netif_deschedule_work(netif);
++ list_add_tail(&quirk->quirks_list, &pciback_quirks);
+}
+
-+static int net_open(struct net_device *dev)
++int pciback_field_is_dup(struct pci_dev *dev, unsigned int reg)
+{
-+ netif_t *netif = netdev_priv(dev);
-+ if (netback_carrier_ok(netif)) {
-+ __netif_up(netif);
-+ netif_start_queue(dev);
++ int ret = 0;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry;
++
++ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
++ if ( OFFSET(cfg_entry) == reg) {
++ ret = 1;
++ break;
++ }
+ }
-+ return 0;
++ return ret;
+}
+
-+static int net_close(struct net_device *dev)
++int pciback_config_quirks_add_field(struct pci_dev *dev, struct config_field
++ *field)
+{
-+ netif_t *netif = netdev_priv(dev);
-+ if (netback_carrier_ok(netif))
-+ __netif_down(netif);
-+ netif_stop_queue(dev);
-+ return 0;
-+}
++ int err = 0;
+
-+static int netbk_change_mtu(struct net_device *dev, int mtu)
-+{
-+ int max = netbk_can_sg(dev) ? 65535 - ETH_HLEN : ETH_DATA_LEN;
++ switch (field->size) {
++ case 1:
++ field->u.b.read = pciback_read_config_byte;
++ field->u.b.write = pciback_write_config_byte;
++ break;
++ case 2:
++ field->u.w.read = pciback_read_config_word;
++ field->u.w.write = pciback_write_config_word;
++ break;
++ case 4:
++ field->u.dw.read = pciback_read_config_dword;
++ field->u.dw.write = pciback_write_config_dword;
++ break;
++ default:
++ err = -EINVAL;
++ goto out;
++ }
+
-+ if (mtu > max)
-+ return -EINVAL;
-+ dev->mtu = mtu;
-+ return 0;
++ pciback_config_add_field(dev, field);
++
++ out:
++ return err;
+}
+
-+static int netbk_set_sg(struct net_device *dev, u32 data)
++int pciback_config_quirks_init(struct pci_dev *dev)
+{
-+ if (data) {
-+ netif_t *netif = netdev_priv(dev);
++ struct pciback_config_quirk *quirk;
++ int ret = 0;
+
-+ if (!(netif->features & NETIF_F_SG))
-+ return -ENOSYS;
++ quirk = kzalloc(sizeof(*quirk), GFP_ATOMIC);
++ if (!quirk) {
++ ret = -ENOMEM;
++ goto out;
+ }
+
-+ return ethtool_op_set_sg(dev, data);
++ quirk->devid.vendor = dev->vendor;
++ quirk->devid.device = dev->device;
++ quirk->devid.subvendor = dev->subsystem_vendor;
++ quirk->devid.subdevice = dev->subsystem_device;
++ quirk->devid.class = 0;
++ quirk->devid.class_mask = 0;
++ quirk->devid.driver_data = 0UL;
++
++ quirk->pdev = dev;
++
++ register_quirk(quirk);
++ out:
++ return ret;
+}
+
-+static int netbk_set_tso(struct net_device *dev, u32 data)
++void pciback_config_field_free(struct config_field *field)
+{
-+ if (data) {
-+ netif_t *netif = netdev_priv(dev);
++ kfree(field);
++}
+
-+ if (!(netif->features & NETIF_F_TSO))
-+ return -ENOSYS;
++int pciback_config_quirk_release(struct pci_dev *dev)
++{
++ struct pciback_config_quirk *quirk;
++ int ret = 0;
++
++ quirk = pciback_find_quirk(dev);
++ if (!quirk) {
++ ret = -ENXIO;
++ goto out;
+ }
+
-+ return ethtool_op_set_tso(dev, data);
++ list_del(&quirk->quirks_list);
++ kfree(quirk);
++
++ out:
++ return ret;
+}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/conf_space_quirks.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/conf_space_quirks.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,35 @@
++/*
++ * PCI Backend - Data structures for special overlays for broken devices.
++ *
++ * Ryan Wilson <hap9 at epoch.ncsc.mil>
++ * Chris Bookholt <hap10 at epoch.ncsc.mil>
++ */
+
-+static struct ethtool_ops network_ethtool_ops =
-+{
-+ .get_tx_csum = ethtool_op_get_tx_csum,
-+ .set_tx_csum = ethtool_op_set_tx_csum,
-+ .get_sg = ethtool_op_get_sg,
-+ .set_sg = netbk_set_sg,
-+ .get_tso = ethtool_op_get_tso,
-+ .set_tso = netbk_set_tso,
-+ .get_link = ethtool_op_get_link,
++#ifndef __XEN_PCIBACK_CONF_SPACE_QUIRKS_H__
++#define __XEN_PCIBACK_CONF_SPACE_QUIRKS_H__
++
++#include <linux/pci.h>
++#include <linux/list.h>
++
++struct pciback_config_quirk {
++ struct list_head quirks_list;
++ struct pci_device_id devid;
++ struct pci_dev *pdev;
+};
+
-+netif_t *netif_alloc(domid_t domid, unsigned int handle)
-+{
-+ int err = 0;
-+ struct net_device *dev;
-+ netif_t *netif;
-+ char name[IFNAMSIZ] = {};
++struct pciback_config_quirk *pciback_find_quirk(struct pci_dev *dev);
+
-+ snprintf(name, IFNAMSIZ - 1, "vif%u.%u", domid, handle);
-+ dev = alloc_netdev(sizeof(netif_t), name, ether_setup);
-+ if (dev == NULL) {
-+ DPRINTK("Could not create netif: out of memory\n");
-+ return ERR_PTR(-ENOMEM);
-+ }
++int pciback_config_quirks_add_field(struct pci_dev *dev, struct config_field
++ *field);
+
-+ netif = netdev_priv(dev);
-+ memset(netif, 0, sizeof(*netif));
-+ netif->domid = domid;
-+ netif->handle = handle;
-+ atomic_set(&netif->refcnt, 1);
-+ init_waitqueue_head(&netif->waiting_to_free);
-+ netif->dev = dev;
++int pciback_config_quirks_remove_field(struct pci_dev *dev, int reg);
+
-+ netback_carrier_off(netif);
++int pciback_config_quirks_init(struct pci_dev *dev);
+
-+ netif->credit_bytes = netif->remaining_credit = ~0UL;
-+ netif->credit_usec = 0UL;
-+ init_timer(&netif->credit_timeout);
-+ /* Initialize 'expires' now: it's used to track the credit window. */
-+ netif->credit_timeout.expires = jiffies;
++void pciback_config_field_free(struct config_field *field);
+
-+ init_timer(&netif->tx_queue_timeout);
++int pciback_config_quirk_release(struct pci_dev *dev);
+
-+ dev->hard_start_xmit = netif_be_start_xmit;
-+ dev->get_stats = netif_be_get_stats;
-+ dev->open = net_open;
-+ dev->stop = net_close;
-+ dev->change_mtu = netbk_change_mtu;
-+ dev->features = NETIF_F_IP_CSUM;
++int pciback_field_is_dup(struct pci_dev *dev, unsigned int reg);
+
-+ SET_ETHTOOL_OPS(dev, &network_ethtool_ops);
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/controller.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/controller.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,408 @@
++/*
++ * Copyright (C) 2007 Hewlett-Packard Development Company, L.P.
++ * Alex Williamson <alex.williamson at hp.com>
++ *
++ * PCI "Controller" Backend - virtualize PCI bus topology based on PCI
++ * controllers. Devices under the same PCI controller are exposed on the
++ * same virtual domain:bus. Within a bus, device slots are virtualized
++ * to compact the bus.
++ *
++ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ */
+
-+ dev->tx_queue_len = netbk_queue_length;
++#include <linux/acpi.h>
++#include <linux/list.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include "pciback.h"
+
-+ /*
-+ * Initialise a dummy MAC address. We choose the numerically
-+ * largest non-broadcast address to prevent the address getting
-+ * stolen by an Ethernet bridge for STP purposes.
-+ * (FE:FF:FF:FF:FF:FF)
-+ */
-+ memset(dev->dev_addr, 0xFF, ETH_ALEN);
-+ dev->dev_addr[0] &= ~0x01;
++#define PCI_MAX_BUSSES 255
++#define PCI_MAX_SLOTS 32
+
-+ rtnl_lock();
-+ err = register_netdevice(dev);
-+ rtnl_unlock();
-+ if (err) {
-+ DPRINTK("Could not register new net device %s: err=%d\n",
-+ dev->name, err);
-+ free_netdev(dev);
-+ return ERR_PTR(err);
++struct controller_dev_entry {
++ struct list_head list;
++ struct pci_dev *dev;
++ unsigned int devfn;
++};
++
++struct controller_list_entry {
++ struct list_head list;
++ struct pci_controller *controller;
++ unsigned int domain;
++ unsigned int bus;
++ unsigned int next_devfn;
++ struct list_head dev_list;
++};
++
++struct controller_dev_data {
++ struct list_head list;
++ unsigned int next_domain;
++ unsigned int next_bus;
++ spinlock_t lock;
++};
++
++struct walk_info {
++ struct pciback_device *pdev;
++ int resource_count;
++ int root_num;
++};
++
++struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn)
++{
++ struct controller_dev_data *dev_data = pdev->pci_dev_data;
++ struct controller_dev_entry *dev_entry;
++ struct controller_list_entry *cntrl_entry;
++ struct pci_dev *dev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&dev_data->lock, flags);
++
++ list_for_each_entry(cntrl_entry, &dev_data->list, list) {
++ if (cntrl_entry->domain != domain ||
++ cntrl_entry->bus != bus)
++ continue;
++
++ list_for_each_entry(dev_entry, &cntrl_entry->dev_list, list) {
++ if (devfn == dev_entry->devfn) {
++ dev = dev_entry->dev;
++ goto found;
++ }
++ }
+ }
++found:
++ spin_unlock_irqrestore(&dev_data->lock, flags);
+
-+ DPRINTK("Successfully created netif\n");
-+ return netif;
++ return dev;
+}
+
-+static int map_frontend_pages(
-+ netif_t *netif, grant_ref_t tx_ring_ref, grant_ref_t rx_ring_ref)
++int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev,
++ int devid, publish_pci_dev_cb publish_cb)
+{
-+ struct gnttab_map_grant_ref op;
++ struct controller_dev_data *dev_data = pdev->pci_dev_data;
++ struct controller_dev_entry *dev_entry;
++ struct controller_list_entry *cntrl_entry;
++ struct pci_controller *dev_controller = PCI_CONTROLLER(dev);
++ unsigned long flags;
++ int ret = 0, found = 0;
+
-+ gnttab_set_map_op(&op, (unsigned long)netif->tx_comms_area->addr,
-+ GNTMAP_host_map, tx_ring_ref, netif->domid);
-+
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
-+ BUG();
++ spin_lock_irqsave(&dev_data->lock, flags);
+
-+ if (op.status) {
-+ DPRINTK(" Gnttab failure mapping tx_ring_ref!\n");
-+ return op.status;
++ /* Look to see if we already have a domain:bus for this controller */
++ list_for_each_entry(cntrl_entry, &dev_data->list, list) {
++ if (cntrl_entry->controller == dev_controller) {
++ found = 1;
++ break;
++ }
+ }
+
-+ netif->tx_shmem_ref = tx_ring_ref;
-+ netif->tx_shmem_handle = op.handle;
++ if (!found) {
++ cntrl_entry = kmalloc(sizeof(*cntrl_entry), GFP_ATOMIC);
++ if (!cntrl_entry) {
++ ret = -ENOMEM;
++ goto out;
++ }
+
-+ gnttab_set_map_op(&op, (unsigned long)netif->rx_comms_area->addr,
-+ GNTMAP_host_map, rx_ring_ref, netif->domid);
++ cntrl_entry->controller = dev_controller;
++ cntrl_entry->next_devfn = PCI_DEVFN(0, 0);
+
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
-+ BUG();
++ cntrl_entry->domain = dev_data->next_domain;
++ cntrl_entry->bus = dev_data->next_bus++;
++ if (dev_data->next_bus > PCI_MAX_BUSSES) {
++ dev_data->next_domain++;
++ dev_data->next_bus = 0;
++ }
+
-+ if (op.status) {
-+ DPRINTK(" Gnttab failure mapping rx_ring_ref!\n");
-+ return op.status;
++ INIT_LIST_HEAD(&cntrl_entry->dev_list);
++
++ list_add_tail(&cntrl_entry->list, &dev_data->list);
+ }
+
-+ netif->rx_shmem_ref = rx_ring_ref;
-+ netif->rx_shmem_handle = op.handle;
++ if (PCI_SLOT(cntrl_entry->next_devfn) > PCI_MAX_SLOTS) {
++ /*
++ * While it seems unlikely, this can actually happen if
++ * a controller has P2P bridges under it.
++ */
++ xenbus_dev_fatal(pdev->xdev, -ENOSPC, "Virtual bus %04x:%02x "
++ "is full, no room to export %04x:%02x:%02x.%x",
++ cntrl_entry->domain, cntrl_entry->bus,
++ pci_domain_nr(dev->bus), dev->bus->number,
++ PCI_SLOT(dev->devfn), PCI_FUNC(dev->devfn));
++ ret = -ENOSPC;
++ goto out;
++ }
+
-+ return 0;
-+}
++ dev_entry = kmalloc(sizeof(*dev_entry), GFP_ATOMIC);
++ if (!dev_entry) {
++ if (list_empty(&cntrl_entry->dev_list)) {
++ list_del(&cntrl_entry->list);
++ kfree(cntrl_entry);
++ }
++ ret = -ENOMEM;
++ goto out;
++ }
+
-+static void unmap_frontend_pages(netif_t *netif)
-+{
-+ struct gnttab_unmap_grant_ref op;
++ dev_entry->dev = dev;
++ dev_entry->devfn = cntrl_entry->next_devfn;
+
-+ gnttab_set_unmap_op(&op, (unsigned long)netif->tx_comms_area->addr,
-+ GNTMAP_host_map, netif->tx_shmem_handle);
++ list_add_tail(&dev_entry->list, &cntrl_entry->dev_list);
+
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
-+ BUG();
++ cntrl_entry->next_devfn += PCI_DEVFN(1, 0);
+
-+ gnttab_set_unmap_op(&op, (unsigned long)netif->rx_comms_area->addr,
-+ GNTMAP_host_map, netif->rx_shmem_handle);
++out:
++ spin_unlock_irqrestore(&dev_data->lock, flags);
+
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
-+ BUG();
++ /* TODO: Publish virtual domain:bus:slot.func here. */
++
++ return ret;
+}
+
-+int netif_map(netif_t *netif, unsigned long tx_ring_ref,
-+ unsigned long rx_ring_ref, unsigned int evtchn)
++void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
+{
-+ int err = -ENOMEM;
-+ netif_tx_sring_t *txs;
-+ netif_rx_sring_t *rxs;
++ struct controller_dev_data *dev_data = pdev->pci_dev_data;
++ struct controller_list_entry *cntrl_entry;
++ struct controller_dev_entry *dev_entry = NULL;
++ struct pci_dev *found_dev = NULL;
++ unsigned long flags;
+
-+ /* Already connected through? */
-+ if (netif->irq)
-+ return 0;
++ spin_lock_irqsave(&dev_data->lock, flags);
+
-+ netif->tx_comms_area = alloc_vm_area(PAGE_SIZE);
-+ if (netif->tx_comms_area == NULL)
-+ return -ENOMEM;
-+ netif->rx_comms_area = alloc_vm_area(PAGE_SIZE);
-+ if (netif->rx_comms_area == NULL)
-+ goto err_rx;
++ list_for_each_entry(cntrl_entry, &dev_data->list, list) {
++ if (cntrl_entry->controller != PCI_CONTROLLER(dev))
++ continue;
++
++ list_for_each_entry(dev_entry, &cntrl_entry->dev_list, list) {
++ if (dev_entry->dev == dev) {
++ found_dev = dev_entry->dev;
++ break;
++ }
++ }
++ }
++
++ if (!found_dev) {
++ spin_unlock_irqrestore(&dev_data->lock, flags);
++ return;
++ }
++
++ list_del(&dev_entry->list);
++ kfree(dev_entry);
+
-+ err = map_frontend_pages(netif, tx_ring_ref, rx_ring_ref);
-+ if (err)
-+ goto err_map;
++ if (list_empty(&cntrl_entry->dev_list)) {
++ list_del(&cntrl_entry->list);
++ kfree(cntrl_entry);
++ }
+
-+ err = bind_interdomain_evtchn_to_irqhandler(
-+ netif->domid, evtchn, netif_be_int, 0,
-+ netif->dev->name, netif);
-+ if (err < 0)
-+ goto err_hypervisor;
-+ netif->irq = err;
-+ disable_irq(netif->irq);
++ spin_unlock_irqrestore(&dev_data->lock, flags);
++ pcistub_put_pci_dev(found_dev);
++}
+
-+ txs = (netif_tx_sring_t *)netif->tx_comms_area->addr;
-+ BACK_RING_INIT(&netif->tx, txs, PAGE_SIZE);
++int pciback_init_devices(struct pciback_device *pdev)
++{
++ struct controller_dev_data *dev_data;
+
-+ rxs = (netif_rx_sring_t *)
-+ ((char *)netif->rx_comms_area->addr);
-+ BACK_RING_INIT(&netif->rx, rxs, PAGE_SIZE);
++ dev_data = kmalloc(sizeof(*dev_data), GFP_KERNEL);
++ if (!dev_data)
++ return -ENOMEM;
+
-+ netif->rx_req_cons_peek = 0;
++ spin_lock_init(&dev_data->lock);
+
-+ netif_get(netif);
++ INIT_LIST_HEAD(&dev_data->list);
+
-+ rtnl_lock();
-+ netback_carrier_on(netif);
-+ if (netif_running(netif->dev))
-+ __netif_up(netif);
-+ rtnl_unlock();
++ /* Starting domain:bus numbers */
++ dev_data->next_domain = 0;
++ dev_data->next_bus = 0;
++
++ pdev->pci_dev_data = dev_data;
+
+ return 0;
-+err_hypervisor:
-+ unmap_frontend_pages(netif);
-+err_map:
-+ free_vm_area(netif->rx_comms_area);
-+err_rx:
-+ free_vm_area(netif->tx_comms_area);
-+ return err;
+}
+
-+void netif_disconnect(netif_t *netif)
++static acpi_status write_xenbus_resource(struct acpi_resource *res, void *data)
+{
-+ if (netback_carrier_ok(netif)) {
-+ rtnl_lock();
-+ netback_carrier_off(netif);
-+ netif_carrier_off(netif->dev); /* discard queued packets */
-+ if (netif_running(netif->dev))
-+ __netif_down(netif);
-+ rtnl_unlock();
-+ netif_put(netif);
-+ }
++ struct walk_info *info = data;
++ struct acpi_resource_address64 addr;
++ acpi_status status;
++ int i, len, err;
++ char str[32], tmp[3];
++ unsigned char *ptr, *buf;
+
-+ atomic_dec(&netif->refcnt);
-+ wait_event(netif->waiting_to_free, atomic_read(&netif->refcnt) == 0);
++ status = acpi_resource_to_address64(res, &addr);
+
-+ del_timer_sync(&netif->credit_timeout);
-+ del_timer_sync(&netif->tx_queue_timeout);
++ /* Do we care about this range? Let's check. */
++ if (!ACPI_SUCCESS(status) ||
++ !(addr.resource_type == ACPI_MEMORY_RANGE ||
++ addr.resource_type == ACPI_IO_RANGE) ||
++ !addr.address_length || addr.producer_consumer != ACPI_PRODUCER)
++ return AE_OK;
+
-+ if (netif->irq)
-+ unbind_from_irqhandler(netif->irq, netif);
-+
-+ unregister_netdev(netif->dev);
++ /*
++ * Furthermore, we really only care to tell the guest about
++ * address ranges that require address translation of some sort.
++ */
++ if (!(addr.resource_type == ACPI_MEMORY_RANGE &&
++ addr.info.mem.translation) &&
++ !(addr.resource_type == ACPI_IO_RANGE &&
++ addr.info.io.translation))
++ return AE_OK;
++
++ /* Store the resource in xenbus for the guest */
++ len = snprintf(str, sizeof(str), "root-%d-resource-%d",
++ info->root_num, info->resource_count);
++ if (unlikely(len >= (sizeof(str) - 1)))
++ return AE_OK;
+
-+ if (netif->tx.sring) {
-+ unmap_frontend_pages(netif);
-+ free_vm_area(netif->tx_comms_area);
-+ free_vm_area(netif->rx_comms_area);
-+ }
++ buf = kzalloc((sizeof(*res) * 2) + 1, GFP_KERNEL);
++ if (!buf)
++ return AE_OK;
+
-+ free_netdev(netif->dev);
-+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/netback/loopback.c tmp-linux-2.6-xen.patch/drivers/xen/netback/loopback.c
---- pristine-linux-2.6.18.2/drivers/xen/netback/loopback.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/netback/loopback.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,324 @@
-+/******************************************************************************
-+ * netback/loopback.c
-+ *
-+ * A two-interface loopback device to emulate a local netfront-netback
-+ * connection. This ensures that local packet delivery looks identical
-+ * to inter-domain delivery. Most importantly, packets delivered locally
-+ * originating from other domains will get *copied* when they traverse this
-+ * driver. This prevents unbounded delays in socket-buffer queues from
-+ * causing the netback driver to "seize up".
-+ *
-+ * This driver creates a symmetric pair of loopback interfaces with names
-+ * vif0.0 and veth0. The intention is that 'vif0.0' is bound to an Ethernet
-+ * bridge, just like a proper netback interface, while a local IP interface
-+ * is configured on 'veth0'.
-+ *
-+ * As with a real netback interface, vif0.0 is configured with a suitable
-+ * dummy MAC address. No default is provided for veth0: a reasonable strategy
-+ * is to transfer eth0's MAC address to veth0, and give eth0 a dummy address
-+ * (to avoid confusing the Etherbridge).
-+ *
-+ * Copyright (c) 2005 K A Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ /* Clean out resource_source */
++ res->data.address64.resource_source.index = 0xFF;
++ res->data.address64.resource_source.string_length = 0;
++ res->data.address64.resource_source.string_ptr = NULL;
+
-+#include <linux/module.h>
-+#include <linux/netdevice.h>
-+#include <linux/inetdevice.h>
-+#include <linux/etherdevice.h>
-+#include <linux/skbuff.h>
-+#include <linux/ethtool.h>
-+#include <net/dst.h>
-+#include <net/xfrm.h> /* secpath_reset() */
-+#include <asm/hypervisor.h> /* is_initial_xendomain() */
++ ptr = (unsigned char *)res;
+
-+static int nloopbacks = -1;
-+module_param(nloopbacks, int, 0);
-+MODULE_PARM_DESC(nloopbacks, "Number of netback-loopback devices to create");
++ /* Turn the acpi_resource into an ASCII byte stream */
++ for (i = 0; i < sizeof(*res); i++) {
++ snprintf(tmp, sizeof(tmp), "%02x", ptr[i]);
++ strncat(buf, tmp, 2);
++ }
+
-+struct net_private {
-+ struct net_device *loopback_dev;
-+ struct net_device_stats stats;
-+};
++ err = xenbus_printf(XBT_NIL, info->pdev->xdev->nodename,
++ str, "%s", buf);
+
-+static int loopback_open(struct net_device *dev)
-+{
-+ struct net_private *np = netdev_priv(dev);
-+ memset(&np->stats, 0, sizeof(np->stats));
-+ netif_start_queue(dev);
-+ return 0;
-+}
++ if (!err)
++ info->resource_count++;
+
-+static int loopback_close(struct net_device *dev)
-+{
-+ netif_stop_queue(dev);
-+ return 0;
-+}
++ kfree(buf);
+
-+#ifdef CONFIG_X86
-+static int is_foreign(unsigned long pfn)
-+{
-+ /* NB. Play it safe for auto-translation mode. */
-+ return (xen_feature(XENFEAT_auto_translated_physmap) ||
-+ (phys_to_machine_mapping[pfn] & FOREIGN_FRAME_BIT));
++ return AE_OK;
+}
-+#else
-+/* How to detect a foreign mapping? Play it safe. */
-+#define is_foreign(pfn) (1)
-+#endif
+
-+static int skb_remove_foreign_references(struct sk_buff *skb)
++int pciback_publish_pci_roots(struct pciback_device *pdev,
++ publish_pci_root_cb publish_root_cb)
+{
-+ struct page *page;
-+ unsigned long pfn;
-+ int i, off;
-+ char *vaddr;
-+
-+ BUG_ON(skb_shinfo(skb)->frag_list);
++ struct controller_dev_data *dev_data = pdev->pci_dev_data;
++ struct controller_list_entry *cntrl_entry;
++ int i, root_num, len, err = 0;
++ unsigned int domain, bus;
++ char str[64];
++ struct walk_info info;
+
-+ if (skb_cloned(skb) &&
-+ unlikely(pskb_expand_head(skb, 0, 0, GFP_ATOMIC)))
-+ return 0;
++ spin_lock(&dev_data->lock);
+
-+ for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
-+ pfn = page_to_pfn(skb_shinfo(skb)->frags[i].page);
-+ if (!is_foreign(pfn))
-+ continue;
-+
-+ page = alloc_page(GFP_ATOMIC | __GFP_NOWARN);
-+ if (unlikely(!page))
-+ return 0;
++ list_for_each_entry(cntrl_entry, &dev_data->list, list) {
++ /* First publish all the domain:bus info */
++ err = publish_root_cb(pdev, cntrl_entry->domain,
++ cntrl_entry->bus);
++ if (err)
++ goto out;
+
-+ vaddr = kmap_skb_frag(&skb_shinfo(skb)->frags[i]);
-+ off = skb_shinfo(skb)->frags[i].page_offset;
-+ memcpy(page_address(page) + off,
-+ vaddr + off,
-+ skb_shinfo(skb)->frags[i].size);
-+ kunmap_skb_frag(vaddr);
++ /*
++ * Now figure out which root-%d this belongs to
++ * so we can associate resources with it.
++ */
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
++ "root_num", "%d", &root_num);
+
-+ put_page(skb_shinfo(skb)->frags[i].page);
-+ skb_shinfo(skb)->frags[i].page = page;
-+ }
++ if (err != 1)
++ goto out;
+
-+ return 1;
-+}
++ for (i = 0; i < root_num; i++) {
++ len = snprintf(str, sizeof(str), "root-%d", i);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
+
-+static int loopback_start_xmit(struct sk_buff *skb, struct net_device *dev)
-+{
-+ struct net_private *np = netdev_priv(dev);
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
++ str, "%x:%x", &domain, &bus);
++ if (err != 2)
++ goto out;
+
-+ if (!skb_remove_foreign_references(skb)) {
-+ np->stats.tx_dropped++;
-+ dev_kfree_skb(skb);
-+ return 0;
-+ }
++ /* Is this the one we just published? */
++ if (domain == cntrl_entry->domain &&
++ bus == cntrl_entry->bus)
++ break;
++ }
+
-+ dst_release(skb->dst);
-+ skb->dst = NULL;
++ if (i == root_num)
++ goto out;
+
-+ skb_orphan(skb);
++ info.pdev = pdev;
++ info.resource_count = 0;
++ info.root_num = i;
+
-+ np->stats.tx_bytes += skb->len;
-+ np->stats.tx_packets++;
++ /* Let ACPI do the heavy lifting on decoding resources */
++ acpi_walk_resources(cntrl_entry->controller->acpi_handle,
++ METHOD_NAME__CRS, write_xenbus_resource,
++ &info);
+
-+ /* Switch to loopback context. */
-+ dev = np->loopback_dev;
-+ np = netdev_priv(dev);
++ /* No resouces. OK. On to the next one */
++ if (!info.resource_count)
++ continue;
+
-+ np->stats.rx_bytes += skb->len;
-+ np->stats.rx_packets++;
++ /* Store the number of resources we wrote for this root-%d */
++ len = snprintf(str, sizeof(str), "root-%d-resources", i);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
+
-+ if (skb->ip_summed == CHECKSUM_HW) {
-+ /* Defer checksum calculation. */
-+ skb->proto_csum_blank = 1;
-+ /* Must be a local packet: assert its integrity. */
-+ skb->proto_data_valid = 1;
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, str,
++ "%d", info.resource_count);
++ if (err)
++ goto out;
+ }
+
-+ skb->ip_summed = skb->proto_data_valid ?
-+ CHECKSUM_UNNECESSARY : CHECKSUM_NONE;
-+
-+ skb->pkt_type = PACKET_HOST; /* overridden by eth_type_trans() */
-+ skb->protocol = eth_type_trans(skb, dev);
-+ skb->dev = dev;
-+ dev->last_rx = jiffies;
++ /* Finally, write some magic to synchronize with the guest. */
++ len = snprintf(str, sizeof(str), "root-resource-magic");
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
+
-+ /* Flush netfilter context: rx'ed skbuffs not expected to have any. */
-+ nf_reset(skb);
-+ secpath_reset(skb);
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, str,
++ "%lx", (sizeof(struct acpi_resource) * 2) + 1);
+
-+ netif_rx(skb);
++out:
++ spin_unlock(&dev_data->lock);
+
-+ return 0;
++ return err;
+}
+
-+static struct net_device_stats *loopback_get_stats(struct net_device *dev)
++void pciback_release_devices(struct pciback_device *pdev)
+{
-+ struct net_private *np = netdev_priv(dev);
-+ return &np->stats;
-+}
++ struct controller_dev_data *dev_data = pdev->pci_dev_data;
++ struct controller_list_entry *cntrl_entry, *c;
++ struct controller_dev_entry *dev_entry, *d;
+
-+static struct ethtool_ops network_ethtool_ops =
-+{
-+ .get_tx_csum = ethtool_op_get_tx_csum,
-+ .set_tx_csum = ethtool_op_set_tx_csum,
-+ .get_sg = ethtool_op_get_sg,
-+ .set_sg = ethtool_op_set_sg,
-+ .get_tso = ethtool_op_get_tso,
-+ .set_tso = ethtool_op_set_tso,
-+ .get_link = ethtool_op_get_link,
-+};
++ list_for_each_entry_safe(cntrl_entry, c, &dev_data->list, list) {
++ list_for_each_entry_safe(dev_entry, d,
++ &cntrl_entry->dev_list, list) {
++ list_del(&dev_entry->list);
++ pcistub_put_pci_dev(dev_entry->dev);
++ kfree(dev_entry);
++ }
++ list_del(&cntrl_entry->list);
++ kfree(cntrl_entry);
++ }
+
++ kfree(dev_data);
++ pdev->pci_dev_data = NULL;
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/passthrough.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/passthrough.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,166 @@
+/*
-+ * Nothing to do here. Virtual interface is point-to-point and the
-+ * physical interface is probably promiscuous anyway.
++ * PCI Backend - Provides restricted access to the real PCI bus topology
++ * to the frontend
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
+ */
-+static void loopback_set_multicast_list(struct net_device *dev)
-+{
-+}
+
-+static void loopback_construct(struct net_device *dev, struct net_device *lo)
-+{
-+ struct net_private *np = netdev_priv(dev);
++#include <linux/list.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include "pciback.h"
+
-+ np->loopback_dev = lo;
++struct passthrough_dev_data {
++ /* Access to dev_list must be protected by lock */
++ struct list_head dev_list;
++ spinlock_t lock;
++};
+
-+ dev->open = loopback_open;
-+ dev->stop = loopback_close;
-+ dev->hard_start_xmit = loopback_start_xmit;
-+ dev->get_stats = loopback_get_stats;
-+ dev->set_multicast_list = loopback_set_multicast_list;
-+ dev->change_mtu = NULL; /* allow arbitrary mtu */
++struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn)
++{
++ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
++ struct pci_dev_entry *dev_entry;
++ struct pci_dev *dev = NULL;
++ unsigned long flags;
+
-+ dev->tx_queue_len = 0;
++ spin_lock_irqsave(&dev_data->lock, flags);
+
-+ dev->features = (NETIF_F_HIGHDMA |
-+ NETIF_F_LLTX |
-+ NETIF_F_TSO |
-+ NETIF_F_SG |
-+ NETIF_F_IP_CSUM);
++ list_for_each_entry(dev_entry, &dev_data->dev_list, list) {
++ if (domain == (unsigned int)pci_domain_nr(dev_entry->dev->bus)
++ && bus == (unsigned int)dev_entry->dev->bus->number
++ && devfn == dev_entry->dev->devfn) {
++ dev = dev_entry->dev;
++ break;
++ }
++ }
+
-+ SET_ETHTOOL_OPS(dev, &network_ethtool_ops);
++ spin_unlock_irqrestore(&dev_data->lock, flags);
+
-+ /*
-+ * We do not set a jumbo MTU on the interface. Otherwise the network
-+ * stack will try to send large packets that will get dropped by the
-+ * Ethernet bridge (unless the physical Ethernet interface is
-+ * configured to transfer jumbo packets). If a larger MTU is desired
-+ * then the system administrator can specify it using the 'ifconfig'
-+ * command.
-+ */
-+ /*dev->mtu = 16*1024;*/
++ return dev;
+}
+
-+static int __init make_loopback(int i)
++int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev,
++ int devid, publish_pci_dev_cb publish_cb)
+{
-+ struct net_device *dev1, *dev2;
-+ char dev_name[IFNAMSIZ];
-+ int err = -ENOMEM;
-+
-+ sprintf(dev_name, "vif0.%d", i);
-+ dev1 = alloc_netdev(sizeof(struct net_private), dev_name, ether_setup);
-+ if (!dev1)
-+ return err;
-+
-+ sprintf(dev_name, "veth%d", i);
-+ dev2 = alloc_netdev(sizeof(struct net_private), dev_name, ether_setup);
-+ if (!dev2)
-+ goto fail_netdev2;
-+
-+ loopback_construct(dev1, dev2);
-+ loopback_construct(dev2, dev1);
-+
-+ /*
-+ * Initialise a dummy MAC address for the 'dummy backend' interface. We
-+ * choose the numerically largest non-broadcast address to prevent the
-+ * address getting stolen by an Ethernet bridge for STP purposes.
-+ */
-+ memset(dev1->dev_addr, 0xFF, ETH_ALEN);
-+ dev1->dev_addr[0] &= ~0x01;
++ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
++ struct pci_dev_entry *dev_entry;
++ unsigned long flags;
++ unsigned int domain, bus, devfn;
++ int err;
+
-+ if ((err = register_netdev(dev1)) != 0)
-+ goto fail;
++ dev_entry = kmalloc(sizeof(*dev_entry), GFP_KERNEL);
++ if (!dev_entry)
++ return -ENOMEM;
++ dev_entry->dev = dev;
+
-+ if ((err = register_netdev(dev2)) != 0) {
-+ unregister_netdev(dev1);
-+ goto fail;
-+ }
++ spin_lock_irqsave(&dev_data->lock, flags);
++ list_add_tail(&dev_entry->list, &dev_data->dev_list);
++ spin_unlock_irqrestore(&dev_data->lock, flags);
+
-+ return 0;
++ /* Publish this device. */
++ domain = (unsigned int)pci_domain_nr(dev->bus);
++ bus = (unsigned int)dev->bus->number;
++ devfn = dev->devfn;
++ err = publish_cb(pdev, domain, bus, devfn, devid);
+
-+ fail:
-+ free_netdev(dev2);
-+ fail_netdev2:
-+ free_netdev(dev1);
+ return err;
+}
+
-+static void __exit clean_loopback(int i)
++void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
+{
-+ struct net_device *dev1, *dev2;
-+ char dev_name[IFNAMSIZ];
++ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
++ struct pci_dev_entry *dev_entry, *t;
++ struct pci_dev *found_dev = NULL;
++ unsigned long flags;
+
-+ sprintf(dev_name, "vif0.%d", i);
-+ dev1 = dev_get_by_name(dev_name);
-+ sprintf(dev_name, "veth%d", i);
-+ dev2 = dev_get_by_name(dev_name);
-+ if (dev1 && dev2) {
-+ unregister_netdev(dev2);
-+ unregister_netdev(dev1);
-+ free_netdev(dev2);
-+ free_netdev(dev1);
++ spin_lock_irqsave(&dev_data->lock, flags);
++
++ list_for_each_entry_safe(dev_entry, t, &dev_data->dev_list, list) {
++ if (dev_entry->dev == dev) {
++ list_del(&dev_entry->list);
++ found_dev = dev_entry->dev;
++ kfree(dev_entry);
++ }
+ }
++
++ spin_unlock_irqrestore(&dev_data->lock, flags);
++
++ if (found_dev)
++ pcistub_put_pci_dev(found_dev);
+}
+
-+static int __init loopback_init(void)
++int pciback_init_devices(struct pciback_device *pdev)
+{
-+ int i, err = 0;
-+
-+ if (nloopbacks == -1)
-+ nloopbacks = is_initial_xendomain() ? 4 : 0;
++ struct passthrough_dev_data *dev_data;
+
-+ for (i = 0; i < nloopbacks; i++)
-+ if ((err = make_loopback(i)) != 0)
-+ break;
++ dev_data = kmalloc(sizeof(*dev_data), GFP_KERNEL);
++ if (!dev_data)
++ return -ENOMEM;
+
-+ return err;
-+}
++ spin_lock_init(&dev_data->lock);
+
-+module_init(loopback_init);
++ INIT_LIST_HEAD(&dev_data->dev_list);
+
-+static void __exit loopback_exit(void)
-+{
-+ int i;
++ pdev->pci_dev_data = dev_data;
+
-+ for (i = nloopbacks; i-- > 0; )
-+ clean_loopback(i);
++ return 0;
+}
+
-+module_exit(loopback_exit);
++int pciback_publish_pci_roots(struct pciback_device *pdev,
++ publish_pci_root_cb publish_root_cb)
++{
++ int err = 0;
++ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
++ struct pci_dev_entry *dev_entry, *e;
++ struct pci_dev *dev;
++ int found;
++ unsigned int domain, bus;
+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/netback/Makefile tmp-linux-2.6-xen.patch/drivers/xen/netback/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/netback/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/netback/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,5 @@
-+obj-$(CONFIG_XEN_NETDEV_BACKEND) := netbk.o
-+obj-$(CONFIG_XEN_NETDEV_LOOPBACK) += netloop.o
++ spin_lock(&dev_data->lock);
+
-+netbk-y := netback.o xenbus.o interface.o
-+netloop-y := loopback.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/netback/netback.c tmp-linux-2.6-xen.patch/drivers/xen/netback/netback.c
---- pristine-linux-2.6.18.2/drivers/xen/netback/netback.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/netback/netback.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,1496 @@
-+/******************************************************************************
-+ * drivers/xen/netback/netback.c
-+ *
-+ * Back-end of the driver for virtual network devices. This portion of the
-+ * driver exports a 'unified' network-device interface that can be accessed
-+ * by any operating system that implements a compatible front end. A
-+ * reference front-end implementation can be found in:
-+ * drivers/xen/netfront/netfront.c
-+ *
-+ * Copyright (c) 2002-2005, K A Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ list_for_each_entry(dev_entry, &dev_data->dev_list, list) {
++ /* Only publish this device as a root if none of its
++ * parent bridges are exported
++ */
++ found = 0;
++ dev = dev_entry->dev->bus->self;
++ for (; !found && dev != NULL; dev = dev->bus->self) {
++ list_for_each_entry(e, &dev_data->dev_list, list) {
++ if (dev == e->dev) {
++ found = 1;
++ break;
++ }
++ }
++ }
+
-+#include "common.h"
-+#include <xen/balloon.h>
-+#include <xen/interface/memory.h>
++ domain = (unsigned int)pci_domain_nr(dev_entry->dev->bus);
++ bus = (unsigned int)dev_entry->dev->bus->number;
+
-+/*define NETBE_DEBUG_INTERRUPT*/
++ if (!found) {
++ err = publish_root_cb(pdev, domain, bus);
++ if (err)
++ break;
++ }
++ }
+
-+/* extra field used in struct page */
-+#define netif_page_index(pg) (*(long *)&(pg)->mapping)
++ spin_unlock(&dev_data->lock);
+
-+struct netbk_rx_meta {
-+ skb_frag_t frag;
-+ int id;
-+ u8 copy:1;
-+};
++ return err;
++}
+
-+static void netif_idx_release(u16 pending_idx);
-+static void netif_page_release(struct page *page);
-+static void make_tx_response(netif_t *netif,
-+ netif_tx_request_t *txp,
-+ s8 st);
-+static netif_rx_response_t *make_rx_response(netif_t *netif,
-+ u16 id,
-+ s8 st,
-+ u16 offset,
-+ u16 size,
-+ u16 flags);
++void pciback_release_devices(struct pciback_device *pdev)
++{
++ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
++ struct pci_dev_entry *dev_entry, *t;
+
-+static void net_tx_action(unsigned long unused);
-+static DECLARE_TASKLET(net_tx_tasklet, net_tx_action, 0);
++ list_for_each_entry_safe(dev_entry, t, &dev_data->dev_list, list) {
++ list_del(&dev_entry->list);
++ pcistub_put_pci_dev(dev_entry->dev);
++ kfree(dev_entry);
++ }
+
-+static void net_rx_action(unsigned long unused);
-+static DECLARE_TASKLET(net_rx_tasklet, net_rx_action, 0);
++ kfree(dev_data);
++ pdev->pci_dev_data = NULL;
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/pci_stub.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/pci_stub.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,948 @@
++/*
++ * PCI Stub Driver - Grabs devices in backend to be exported later
++ *
++ * Ryan Wilson <hap9 at epoch.ncsc.mil>
++ * Chris Bookholt <hap10 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/list.h>
++#include <linux/spinlock.h>
++#include <linux/kref.h>
++#include <asm/atomic.h>
++#include "pciback.h"
++#include "conf_space.h"
++#include "conf_space_quirks.h"
+
-+static struct timer_list net_timer;
++static char *pci_devs_to_hide = NULL;
++module_param_named(hide, pci_devs_to_hide, charp, 0444);
+
-+#define MAX_PENDING_REQS 256
++struct pcistub_device_id {
++ struct list_head slot_list;
++ int domain;
++ unsigned char bus;
++ unsigned int devfn;
++};
++static LIST_HEAD(pcistub_device_ids);
++static DEFINE_SPINLOCK(device_ids_lock);
+
-+static struct sk_buff_head rx_queue;
++struct pcistub_device {
++ struct kref kref;
++ struct list_head dev_list;
++ spinlock_t lock;
+
-+static struct page **mmap_pages;
-+static inline unsigned long idx_to_kaddr(unsigned int idx)
-+{
-+ return (unsigned long)pfn_to_kaddr(page_to_pfn(mmap_pages[idx]));
-+}
++ struct pci_dev *dev;
++ struct pciback_device *pdev; /* non-NULL if struct pci_dev is in use */
++};
+
-+#define PKT_PROT_LEN 64
++/* Access to pcistub_devices & seized_devices lists and the initialize_devices
++ * flag must be locked with pcistub_devices_lock
++ */
++static DEFINE_SPINLOCK(pcistub_devices_lock);
++static LIST_HEAD(pcistub_devices);
+
-+static struct pending_tx_info {
-+ netif_tx_request_t req;
-+ netif_t *netif;
-+} pending_tx_info[MAX_PENDING_REQS];
-+static u16 pending_ring[MAX_PENDING_REQS];
-+typedef unsigned int PEND_RING_IDX;
-+#define MASK_PEND_IDX(_i) ((_i)&(MAX_PENDING_REQS-1))
-+static PEND_RING_IDX pending_prod, pending_cons;
-+#define NR_PENDING_REQS (MAX_PENDING_REQS - pending_prod + pending_cons)
++/* wait for device_initcall before initializing our devices
++ * (see pcistub_init_devices_late)
++ */
++static int initialize_devices = 0;
++static LIST_HEAD(seized_devices);
+
-+/* Freed TX SKBs get batched on this ring before return to pending_ring. */
-+static u16 dealloc_ring[MAX_PENDING_REQS];
-+static PEND_RING_IDX dealloc_prod, dealloc_cons;
++static struct pcistub_device *pcistub_device_alloc(struct pci_dev *dev)
++{
++ struct pcistub_device *psdev;
+
-+static struct sk_buff_head tx_queue;
++ dev_dbg(&dev->dev, "pcistub_device_alloc\n");
+
-+static grant_handle_t grant_tx_handle[MAX_PENDING_REQS];
-+static gnttab_unmap_grant_ref_t tx_unmap_ops[MAX_PENDING_REQS];
-+static gnttab_map_grant_ref_t tx_map_ops[MAX_PENDING_REQS];
++ psdev = kzalloc(sizeof(*psdev), GFP_ATOMIC);
++ if (!psdev)
++ return NULL;
+
-+static struct list_head net_schedule_list;
-+static spinlock_t net_schedule_list_lock;
++ psdev->dev = pci_dev_get(dev);
++ if (!psdev->dev) {
++ kfree(psdev);
++ return NULL;
++ }
+
-+#define MAX_MFN_ALLOC 64
-+static unsigned long mfn_list[MAX_MFN_ALLOC];
-+static unsigned int alloc_index = 0;
++ kref_init(&psdev->kref);
++ spin_lock_init(&psdev->lock);
+
-+static inline unsigned long alloc_mfn(void)
-+{
-+ BUG_ON(alloc_index == 0);
-+ return mfn_list[--alloc_index];
++ return psdev;
+}
+
-+static int check_mfn(int nr)
++/* Don't call this directly as it's called by pcistub_device_put */
++static void pcistub_device_release(struct kref *kref)
+{
-+ struct xen_memory_reservation reservation = {
-+ .extent_order = 0,
-+ .domid = DOMID_SELF
-+ };
++ struct pcistub_device *psdev;
+
-+ if (likely(alloc_index >= nr))
-+ return 0;
++ psdev = container_of(kref, struct pcistub_device, kref);
+
-+ set_xen_guest_handle(reservation.extent_start, mfn_list + alloc_index);
-+ reservation.nr_extents = MAX_MFN_ALLOC - alloc_index;
-+ alloc_index += HYPERVISOR_memory_op(XENMEM_increase_reservation,
-+ &reservation);
++ dev_dbg(&psdev->dev->dev, "pcistub_device_release\n");
+
-+ return alloc_index >= nr ? 0 : -ENOMEM;
++ /* Clean-up the device */
++ pciback_reset_device(psdev->dev);
++ pciback_config_free_dyn_fields(psdev->dev);
++ pciback_config_free_dev(psdev->dev);
++ kfree(pci_get_drvdata(psdev->dev));
++ pci_set_drvdata(psdev->dev, NULL);
++
++ pci_dev_put(psdev->dev);
++
++ kfree(psdev);
+}
+
-+static inline void maybe_schedule_tx_action(void)
++static inline void pcistub_device_get(struct pcistub_device *psdev)
+{
-+ smp_mb();
-+ if ((NR_PENDING_REQS < (MAX_PENDING_REQS/2)) &&
-+ !list_empty(&net_schedule_list))
-+ tasklet_schedule(&net_tx_tasklet);
++ kref_get(&psdev->kref);
+}
+
-+static struct sk_buff *netbk_copy_skb(struct sk_buff *skb)
++static inline void pcistub_device_put(struct pcistub_device *psdev)
+{
-+ struct skb_shared_info *ninfo;
-+ struct sk_buff *nskb;
-+ unsigned long offset;
-+ int ret;
-+ int len;
-+ int headlen;
-+
-+ BUG_ON(skb_shinfo(skb)->frag_list != NULL);
-+
-+ nskb = alloc_skb(SKB_MAX_HEAD(0), GFP_ATOMIC | __GFP_NOWARN);
-+ if (unlikely(!nskb))
-+ goto err;
-+
-+ skb_reserve(nskb, 16 + NET_IP_ALIGN);
-+ headlen = nskb->end - nskb->data;
-+ if (headlen > skb_headlen(skb))
-+ headlen = skb_headlen(skb);
-+ ret = skb_copy_bits(skb, 0, __skb_put(nskb, headlen), headlen);
-+ BUG_ON(ret);
-+
-+ ninfo = skb_shinfo(nskb);
-+ ninfo->gso_size = skb_shinfo(skb)->gso_size;
-+ ninfo->gso_type = skb_shinfo(skb)->gso_type;
-+
-+ offset = headlen;
-+ len = skb->len - headlen;
++ kref_put(&psdev->kref, pcistub_device_release);
++}
+
-+ nskb->len = skb->len;
-+ nskb->data_len = len;
-+ nskb->truesize += len;
++static struct pcistub_device *pcistub_device_find(int domain, int bus,
++ int slot, int func)
++{
++ struct pcistub_device *psdev = NULL;
++ unsigned long flags;
+
-+ while (len) {
-+ struct page *page;
-+ int copy;
-+ int zero;
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
+
-+ if (unlikely(ninfo->nr_frags >= MAX_SKB_FRAGS)) {
-+ dump_stack();
-+ goto err_free;
++ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
++ if (psdev->dev != NULL
++ && domain == pci_domain_nr(psdev->dev->bus)
++ && bus == psdev->dev->bus->number
++ && PCI_DEVFN(slot, func) == psdev->dev->devfn) {
++ pcistub_device_get(psdev);
++ goto out;
+ }
++ }
+
-+ copy = len >= PAGE_SIZE ? PAGE_SIZE : len;
-+ zero = len >= PAGE_SIZE ? 0 : __GFP_ZERO;
++ /* didn't find it */
++ psdev = NULL;
+
-+ page = alloc_page(GFP_ATOMIC | __GFP_NOWARN | zero);
-+ if (unlikely(!page))
-+ goto err_free;
++ out:
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++ return psdev;
++}
+
-+ ret = skb_copy_bits(skb, offset, page_address(page), copy);
-+ BUG_ON(ret);
++static struct pci_dev *pcistub_device_get_pci_dev(struct pciback_device *pdev,
++ struct pcistub_device *psdev)
++{
++ struct pci_dev *pci_dev = NULL;
++ unsigned long flags;
+
-+ ninfo->frags[ninfo->nr_frags].page = page;
-+ ninfo->frags[ninfo->nr_frags].page_offset = 0;
-+ ninfo->frags[ninfo->nr_frags].size = copy;
-+ ninfo->nr_frags++;
++ pcistub_device_get(psdev);
+
-+ offset += copy;
-+ len -= copy;
++ spin_lock_irqsave(&psdev->lock, flags);
++ if (!psdev->pdev) {
++ psdev->pdev = pdev;
++ pci_dev = psdev->dev;
+ }
++ spin_unlock_irqrestore(&psdev->lock, flags);
+
-+ offset = nskb->data - skb->data;
-+
-+ nskb->h.raw = skb->h.raw + offset;
-+ nskb->nh.raw = skb->nh.raw + offset;
-+ nskb->mac.raw = skb->mac.raw + offset;
-+
-+ return nskb;
++ if (!pci_dev)
++ pcistub_device_put(psdev);
+
-+ err_free:
-+ kfree_skb(nskb);
-+ err:
-+ return NULL;
++ return pci_dev;
+}
+
-+static inline int netbk_max_required_rx_slots(netif_t *netif)
++struct pci_dev *pcistub_get_pci_dev_by_slot(struct pciback_device *pdev,
++ int domain, int bus,
++ int slot, int func)
+{
-+ if (netif->features & (NETIF_F_SG|NETIF_F_TSO))
-+ return MAX_SKB_FRAGS + 2; /* header + extra_info + frags */
-+ return 1; /* all in one */
-+}
++ struct pcistub_device *psdev;
++ struct pci_dev *found_dev = NULL;
++ unsigned long flags;
+
-+static inline int netbk_queue_full(netif_t *netif)
-+{
-+ RING_IDX peek = netif->rx_req_cons_peek;
-+ RING_IDX needed = netbk_max_required_rx_slots(netif);
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
+
-+ return ((netif->rx.sring->req_prod - peek) < needed) ||
-+ ((netif->rx.rsp_prod_pvt + NET_RX_RING_SIZE - peek) < needed);
-+}
++ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
++ if (psdev->dev != NULL
++ && domain == pci_domain_nr(psdev->dev->bus)
++ && bus == psdev->dev->bus->number
++ && PCI_DEVFN(slot, func) == psdev->dev->devfn) {
++ found_dev = pcistub_device_get_pci_dev(pdev, psdev);
++ break;
++ }
++ }
+
-+static void tx_queue_callback(unsigned long data)
-+{
-+ netif_t *netif = (netif_t *)data;
-+ if (netif_schedulable(netif))
-+ netif_wake_queue(netif->dev);
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++ return found_dev;
+}
+
-+int netif_be_start_xmit(struct sk_buff *skb, struct net_device *dev)
++struct pci_dev *pcistub_get_pci_dev(struct pciback_device *pdev,
++ struct pci_dev *dev)
+{
-+ netif_t *netif = netdev_priv(dev);
-+
-+ BUG_ON(skb->dev != dev);
++ struct pcistub_device *psdev;
++ struct pci_dev *found_dev = NULL;
++ unsigned long flags;
+
-+ /* Drop the packet if the target domain has no receive buffers. */
-+ if (unlikely(!netif_schedulable(netif) || netbk_queue_full(netif)))
-+ goto drop;
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
+
-+ /*
-+ * Copy the packet here if it's destined for a flipping interface
-+ * but isn't flippable (e.g. extra references to data).
-+ * XXX For now we also copy skbuffs whose head crosses a page
-+ * boundary, because netbk_gop_skb can't handle them.
-+ */
-+ if (!netif->copying_receiver ||
-+ ((skb_headlen(skb) + offset_in_page(skb->data)) >= PAGE_SIZE)) {
-+ struct sk_buff *nskb = netbk_copy_skb(skb);
-+ if ( unlikely(nskb == NULL) )
-+ goto drop;
-+ /* Copy only the header fields we use in this driver. */
-+ nskb->dev = skb->dev;
-+ nskb->ip_summed = skb->ip_summed;
-+ nskb->proto_data_valid = skb->proto_data_valid;
-+ dev_kfree_skb(skb);
-+ skb = nskb;
++ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
++ if (psdev->dev == dev) {
++ found_dev = pcistub_device_get_pci_dev(pdev, psdev);
++ break;
++ }
+ }
+
-+ netif->rx_req_cons_peek += skb_shinfo(skb)->nr_frags + 1 +
-+ !!skb_shinfo(skb)->gso_size;
-+ netif_get(netif);
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++ return found_dev;
++}
+
-+ if (netbk_can_queue(dev) && netbk_queue_full(netif)) {
-+ netif->rx.sring->req_event = netif->rx_req_cons_peek +
-+ netbk_max_required_rx_slots(netif);
-+ mb(); /* request notification /then/ check & stop the queue */
-+ if (netbk_queue_full(netif)) {
-+ netif_stop_queue(dev);
-+ /*
-+ * Schedule 500ms timeout to restart the queue, thus
-+ * ensuring that an inactive queue will be drained.
-+ * Packets will be immediately be dropped until more
-+ * receive buffers become available (see
-+ * netbk_queue_full() check above).
-+ */
-+ netif->tx_queue_timeout.data = (unsigned long)netif;
-+ netif->tx_queue_timeout.function = tx_queue_callback;
-+ __mod_timer(&netif->tx_queue_timeout, jiffies + HZ/2);
++void pcistub_put_pci_dev(struct pci_dev *dev)
++{
++ struct pcistub_device *psdev, *found_psdev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
++ if (psdev->dev == dev) {
++ found_psdev = psdev;
++ break;
+ }
+ }
+
-+ skb_queue_tail(&rx_queue, skb);
-+ tasklet_schedule(&net_rx_tasklet);
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
+
-+ return 0;
++ /* Cleanup our device
++ * (so it's ready for the next domain)
++ */
++ pciback_reset_device(found_psdev->dev);
++ pciback_config_free_dyn_fields(found_psdev->dev);
++ pciback_config_reset_dev(found_psdev->dev);
+
-+ drop:
-+ netif->stats.tx_dropped++;
-+ dev_kfree_skb(skb);
-+ return 0;
-+}
++ spin_lock_irqsave(&found_psdev->lock, flags);
++ found_psdev->pdev = NULL;
++ spin_unlock_irqrestore(&found_psdev->lock, flags);
+
-+#if 0
-+static void xen_network_done_notify(void)
-+{
-+ static struct net_device *eth0_dev = NULL;
-+ if (unlikely(eth0_dev == NULL))
-+ eth0_dev = __dev_get_by_name("eth0");
-+ netif_rx_schedule(eth0_dev);
-+}
-+/*
-+ * Add following to poll() function in NAPI driver (Tigon3 is example):
-+ * if ( xen_network_done() )
-+ * tg3_enable_ints(tp);
-+ */
-+int xen_network_done(void)
-+{
-+ return skb_queue_empty(&rx_queue);
++ pcistub_device_put(found_psdev);
+}
-+#endif
-+
-+struct netrx_pending_operations {
-+ unsigned trans_prod, trans_cons;
-+ unsigned mmu_prod, mmu_cons;
-+ unsigned mcl_prod, mcl_cons;
-+ unsigned copy_prod, copy_cons;
-+ unsigned meta_prod, meta_cons;
-+ mmu_update_t *mmu;
-+ gnttab_transfer_t *trans;
-+ gnttab_copy_t *copy;
-+ multicall_entry_t *mcl;
-+ struct netbk_rx_meta *meta;
-+};
+
-+/* Set up the grant operations for this fragment. If it's a flipping
-+ interface, we also set up the unmap request from here. */
-+static u16 netbk_gop_frag(netif_t *netif, struct netbk_rx_meta *meta,
-+ int i, struct netrx_pending_operations *npo,
-+ struct page *page, unsigned long size,
-+ unsigned long offset)
++static int __devinit pcistub_match_one(struct pci_dev *dev,
++ struct pcistub_device_id *pdev_id)
+{
-+ mmu_update_t *mmu;
-+ gnttab_transfer_t *gop;
-+ gnttab_copy_t *copy_gop;
-+ multicall_entry_t *mcl;
-+ netif_rx_request_t *req;
-+ unsigned long old_mfn, new_mfn;
-+
-+ old_mfn = virt_to_mfn(page_address(page));
++ /* Match the specified device by domain, bus, slot, func and also if
++ * any of the device's parent bridges match.
++ */
++ for (; dev != NULL; dev = dev->bus->self) {
++ if (pci_domain_nr(dev->bus) == pdev_id->domain
++ && dev->bus->number == pdev_id->bus
++ && dev->devfn == pdev_id->devfn)
++ return 1;
+
-+ req = RING_GET_REQUEST(&netif->rx, netif->rx.req_cons + i);
-+ if (netif->copying_receiver) {
-+ /* The fragment needs to be copied rather than
-+ flipped. */
-+ meta->copy = 1;
-+ copy_gop = npo->copy + npo->copy_prod++;
-+ copy_gop->flags = GNTCOPY_dest_gref;
-+ if (PageForeign(page)) {
-+ struct pending_tx_info *src_pend =
-+ &pending_tx_info[netif_page_index(page)];
-+ copy_gop->source.domid = src_pend->netif->domid;
-+ copy_gop->source.u.ref = src_pend->req.gref;
-+ copy_gop->flags |= GNTCOPY_source_gref;
-+ } else {
-+ copy_gop->source.domid = DOMID_SELF;
-+ copy_gop->source.u.gmfn = old_mfn;
-+ }
-+ copy_gop->source.offset = offset;
-+ copy_gop->dest.domid = netif->domid;
-+ copy_gop->dest.offset = 0;
-+ copy_gop->dest.u.ref = req->gref;
-+ copy_gop->len = size;
-+ } else {
-+ meta->copy = 0;
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ new_mfn = alloc_mfn();
++ /* Sometimes topmost bridge links to itself. */
++ if (dev == dev->bus->self)
++ break;
++ }
+
-+ /*
-+ * Set the new P2M table entry before
-+ * reassigning the old data page. Heed the
-+ * comment in pgtable-2level.h:pte_page(). :-)
-+ */
-+ set_phys_to_machine(page_to_pfn(page), new_mfn);
++ return 0;
++}
+
-+ mcl = npo->mcl + npo->mcl_prod++;
-+ MULTI_update_va_mapping(mcl,
-+ (unsigned long)page_address(page),
-+ pfn_pte_ma(new_mfn, PAGE_KERNEL),
-+ 0);
++static int __devinit pcistub_match(struct pci_dev *dev)
++{
++ struct pcistub_device_id *pdev_id;
++ unsigned long flags;
++ int found = 0;
+
-+ mmu = npo->mmu + npo->mmu_prod++;
-+ mmu->ptr = ((maddr_t)new_mfn << PAGE_SHIFT) |
-+ MMU_MACHPHYS_UPDATE;
-+ mmu->val = page_to_pfn(page);
++ spin_lock_irqsave(&device_ids_lock, flags);
++ list_for_each_entry(pdev_id, &pcistub_device_ids, slot_list) {
++ if (pcistub_match_one(dev, pdev_id)) {
++ found = 1;
++ break;
+ }
-+
-+ gop = npo->trans + npo->trans_prod++;
-+ gop->mfn = old_mfn;
-+ gop->domid = netif->domid;
-+ gop->ref = req->gref;
+ }
-+ return req->id;
++ spin_unlock_irqrestore(&device_ids_lock, flags);
++
++ return found;
+}
+
-+static void netbk_gop_skb(struct sk_buff *skb,
-+ struct netrx_pending_operations *npo)
++static int __devinit pcistub_init_device(struct pci_dev *dev)
+{
-+ netif_t *netif = netdev_priv(skb->dev);
-+ int nr_frags = skb_shinfo(skb)->nr_frags;
-+ int i;
-+ int extra;
-+ struct netbk_rx_meta *head_meta, *meta;
++ struct pciback_dev_data *dev_data;
++ int err = 0;
+
-+ head_meta = npo->meta + npo->meta_prod++;
-+ head_meta->frag.page_offset = skb_shinfo(skb)->gso_type;
-+ head_meta->frag.size = skb_shinfo(skb)->gso_size;
-+ extra = !!head_meta->frag.size + 1;
++ dev_dbg(&dev->dev, "initializing...\n");
+
-+ for (i = 0; i < nr_frags; i++) {
-+ meta = npo->meta + npo->meta_prod++;
-+ meta->frag = skb_shinfo(skb)->frags[i];
-+ meta->id = netbk_gop_frag(netif, meta, i + extra, npo,
-+ meta->frag.page,
-+ meta->frag.size,
-+ meta->frag.page_offset);
++ /* The PCI backend is not intended to be a module (or to work with
++ * removable PCI devices (yet). If it were, pciback_config_free()
++ * would need to be called somewhere to free the memory allocated
++ * here and then to call kfree(pci_get_drvdata(psdev->dev)).
++ */
++ dev_data = kzalloc(sizeof(*dev_data), GFP_ATOMIC);
++ if (!dev_data) {
++ err = -ENOMEM;
++ goto out;
+ }
++ pci_set_drvdata(dev, dev_data);
+
-+ /*
-+ * This must occur at the end to ensure that we don't trash skb_shinfo
-+ * until we're done. We know that the head doesn't cross a page
-+ * boundary because such packets get copied in netif_be_start_xmit.
++ dev_dbg(&dev->dev, "initializing config\n");
++ err = pciback_config_init_dev(dev);
++ if (err)
++ goto out;
++
++ /* HACK: Force device (& ACPI) to determine what IRQ it's on - we
++ * must do this here because pcibios_enable_device may specify
++ * the pci device's true irq (and possibly its other resources)
++ * if they differ from what's in the configuration space.
++ * This makes the assumption that the device's resources won't
++ * change after this point (otherwise this code may break!)
+ */
-+ head_meta->id = netbk_gop_frag(netif, head_meta, 0, npo,
-+ virt_to_page(skb->data),
-+ skb_headlen(skb),
-+ offset_in_page(skb->data));
++ dev_dbg(&dev->dev, "enabling device\n");
++ err = pci_enable_device(dev);
++ if (err)
++ goto config_release;
+
-+ netif->rx.req_cons += nr_frags + extra;
-+}
++ /* Now disable the device (this also ensures some private device
++ * data is setup before we export)
++ */
++ dev_dbg(&dev->dev, "reset device\n");
++ pciback_reset_device(dev);
+
-+static inline void netbk_free_pages(int nr_frags, struct netbk_rx_meta *meta)
-+{
-+ int i;
++ return 0;
+
-+ for (i = 0; i < nr_frags; i++)
-+ put_page(meta[i].frag.page);
++ config_release:
++ pciback_config_free_dev(dev);
++
++ out:
++ pci_set_drvdata(dev, NULL);
++ kfree(dev_data);
++ return err;
+}
+
-+/* This is a twin to netbk_gop_skb. Assume that netbk_gop_skb was
-+ used to set up the operations on the top of
-+ netrx_pending_operations, which have since been done. Check that
-+ they didn't give any errors and advance over them. */
-+static int netbk_check_gop(int nr_frags, domid_t domid,
-+ struct netrx_pending_operations *npo)
++/*
++ * Because some initialization still happens on
++ * devices during fs_initcall, we need to defer
++ * full initialization of our devices until
++ * device_initcall.
++ */
++static int __init pcistub_init_devices_late(void)
+{
-+ multicall_entry_t *mcl;
-+ gnttab_transfer_t *gop;
-+ gnttab_copy_t *copy_op;
-+ int status = NETIF_RSP_OKAY;
-+ int i;
++ struct pcistub_device *psdev;
++ unsigned long flags;
++ int err = 0;
+
-+ for (i = 0; i <= nr_frags; i++) {
-+ if (npo->meta[npo->meta_cons + i].copy) {
-+ copy_op = npo->copy + npo->copy_cons++;
-+ if (copy_op->status != GNTST_okay) {
-+ DPRINTK("Bad status %d from copy to DOM%d.\n",
-+ copy_op->status, domid);
-+ status = NETIF_RSP_ERROR;
-+ }
-+ } else {
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ mcl = npo->mcl + npo->mcl_cons++;
-+ /* The update_va_mapping() must not fail. */
-+ BUG_ON(mcl->result != 0);
-+ }
++ pr_debug("pciback: pcistub_init_devices_late\n");
+
-+ gop = npo->trans + npo->trans_cons++;
-+ /* Check the reassignment error code. */
-+ if (gop->status != 0) {
-+ DPRINTK("Bad status %d from grant transfer to DOM%u\n",
-+ gop->status, domid);
-+ /*
-+ * Page no longer belongs to us unless
-+ * GNTST_bad_page, but that should be
-+ * a fatal error anyway.
-+ */
-+ BUG_ON(gop->status == GNTST_bad_page);
-+ status = NETIF_RSP_ERROR;
-+ }
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ while (!list_empty(&seized_devices)) {
++ psdev = container_of(seized_devices.next,
++ struct pcistub_device, dev_list);
++ list_del(&psdev->dev_list);
++
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++
++ err = pcistub_init_device(psdev->dev);
++ if (err) {
++ dev_err(&psdev->dev->dev,
++ "error %d initializing device\n", err);
++ kfree(psdev);
++ psdev = NULL;
+ }
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ if (psdev)
++ list_add_tail(&psdev->dev_list, &pcistub_devices);
+ }
+
-+ return status;
++ initialize_devices = 1;
++
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++
++ return 0;
+}
+
-+static void netbk_add_frag_responses(netif_t *netif, int status,
-+ struct netbk_rx_meta *meta, int nr_frags)
++static int __devinit pcistub_seize(struct pci_dev *dev)
+{
-+ int i;
-+ unsigned long offset;
++ struct pcistub_device *psdev;
++ unsigned long flags;
++ int err = 0;
+
-+ for (i = 0; i < nr_frags; i++) {
-+ int id = meta[i].id;
-+ int flags = (i == nr_frags - 1) ? 0 : NETRXF_more_data;
++ psdev = pcistub_device_alloc(dev);
++ if (!psdev)
++ return -ENOMEM;
+
-+ if (meta[i].copy)
-+ offset = 0;
-+ else
-+ offset = meta[i].frag.page_offset;
-+ make_rx_response(netif, id, status, offset,
-+ meta[i].frag.size, flags);
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ if (initialize_devices) {
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++
++ /* don't want irqs disabled when calling pcistub_init_device */
++ err = pcistub_init_device(psdev->dev);
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ if (!err)
++ list_add(&psdev->dev_list, &pcistub_devices);
++ } else {
++ dev_dbg(&dev->dev, "deferring initialization\n");
++ list_add(&psdev->dev_list, &seized_devices);
+ }
++
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++
++ if (err)
++ pcistub_device_put(psdev);
++
++ return err;
+}
+
-+static void net_rx_action(unsigned long unused)
++static int __devinit pcistub_probe(struct pci_dev *dev,
++ const struct pci_device_id *id)
+{
-+ netif_t *netif = NULL;
-+ s8 status;
-+ u16 id, irq, flags;
-+ netif_rx_response_t *resp;
-+ multicall_entry_t *mcl;
-+ struct sk_buff_head rxq;
-+ struct sk_buff *skb;
-+ int notify_nr = 0;
-+ int ret;
-+ int nr_frags;
-+ int count;
-+ unsigned long offset;
++ int err = 0;
+
-+ /*
-+ * Putting hundreds of bytes on the stack is considered rude.
-+ * Static works because a tasklet can only be on one CPU at any time.
-+ */
-+ static multicall_entry_t rx_mcl[NET_RX_RING_SIZE+3];
-+ static mmu_update_t rx_mmu[NET_RX_RING_SIZE];
-+ static gnttab_transfer_t grant_trans_op[NET_RX_RING_SIZE];
-+ static gnttab_copy_t grant_copy_op[NET_RX_RING_SIZE];
-+ static unsigned char rx_notify[NR_IRQS];
-+ static u16 notify_list[NET_RX_RING_SIZE];
-+ static struct netbk_rx_meta meta[NET_RX_RING_SIZE];
++ dev_dbg(&dev->dev, "probing...\n");
+
-+ struct netrx_pending_operations npo = {
-+ mmu: rx_mmu,
-+ trans: grant_trans_op,
-+ copy: grant_copy_op,
-+ mcl: rx_mcl,
-+ meta: meta};
++ if (pcistub_match(dev)) {
+
-+ skb_queue_head_init(&rxq);
++ if (dev->hdr_type != PCI_HEADER_TYPE_NORMAL
++ && dev->hdr_type != PCI_HEADER_TYPE_BRIDGE) {
++ dev_err(&dev->dev, "can't export pci devices that "
++ "don't have a normal (0) or bridge (1) "
++ "header type!\n");
++ err = -ENODEV;
++ goto out;
++ }
+
-+ count = 0;
++ dev_info(&dev->dev, "seizing device\n");
++ err = pcistub_seize(dev);
++ } else
++ /* Didn't find the device */
++ err = -ENODEV;
+
-+ while ((skb = skb_dequeue(&rx_queue)) != NULL) {
-+ nr_frags = skb_shinfo(skb)->nr_frags;
-+ *(int *)skb->cb = nr_frags;
++ out:
++ return err;
++}
+
-+ if (!xen_feature(XENFEAT_auto_translated_physmap) &&
-+ !((netif_t *)netdev_priv(skb->dev))->copying_receiver &&
-+ check_mfn(nr_frags + 1)) {
-+ /* Memory squeeze? Back off for an arbitrary while. */
-+ if ( net_ratelimit() )
-+ WPRINTK("Memory squeeze in netback "
-+ "driver.\n");
-+ mod_timer(&net_timer, jiffies + HZ);
-+ skb_queue_head(&rx_queue, skb);
-+ break;
-+ }
++static void pcistub_remove(struct pci_dev *dev)
++{
++ struct pcistub_device *psdev, *found_psdev = NULL;
++ unsigned long flags;
+
-+ netbk_gop_skb(skb, &npo);
++ dev_dbg(&dev->dev, "removing\n");
+
-+ count += nr_frags + 1;
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
+
-+ __skb_queue_tail(&rxq, skb);
++ pciback_config_quirk_release(dev);
+
-+ /* Filled the batch queue? */
-+ if (count + MAX_SKB_FRAGS >= NET_RX_RING_SIZE)
++ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
++ if (psdev->dev == dev) {
++ found_psdev = psdev;
+ break;
++ }
+ }
+
-+ if (npo.mcl_prod &&
-+ !xen_feature(XENFEAT_auto_translated_physmap)) {
-+ mcl = npo.mcl + npo.mcl_prod++;
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
+
-+ BUG_ON(mcl[-1].op != __HYPERVISOR_update_va_mapping);
-+ mcl[-1].args[MULTI_UVMFLAGS_INDEX] = UVMF_TLB_FLUSH|UVMF_ALL;
++ if (found_psdev) {
++ dev_dbg(&dev->dev, "found device to remove - in use? %p\n",
++ found_psdev->pdev);
+
-+ mcl->op = __HYPERVISOR_mmu_update;
-+ mcl->args[0] = (unsigned long)rx_mmu;
-+ mcl->args[1] = npo.mmu_prod;
-+ mcl->args[2] = 0;
-+ mcl->args[3] = DOMID_SELF;
-+ }
++ if (found_psdev->pdev) {
++ printk(KERN_WARNING "pciback: ****** removing device "
++ "%s while still in-use! ******\n",
++ pci_name(found_psdev->dev));
++ printk(KERN_WARNING "pciback: ****** driver domain may "
++ "still access this device's i/o resources!\n");
++ printk(KERN_WARNING "pciback: ****** shutdown driver "
++ "domain before binding device\n");
++ printk(KERN_WARNING "pciback: ****** to other drivers "
++ "or domains\n");
+
-+ if (npo.trans_prod) {
-+ mcl = npo.mcl + npo.mcl_prod++;
-+ mcl->op = __HYPERVISOR_grant_table_op;
-+ mcl->args[0] = GNTTABOP_transfer;
-+ mcl->args[1] = (unsigned long)grant_trans_op;
-+ mcl->args[2] = npo.trans_prod;
-+ }
++ pciback_release_pci_dev(found_psdev->pdev,
++ found_psdev->dev);
++ }
+
-+ if (npo.copy_prod) {
-+ mcl = npo.mcl + npo.mcl_prod++;
-+ mcl->op = __HYPERVISOR_grant_table_op;
-+ mcl->args[0] = GNTTABOP_copy;
-+ mcl->args[1] = (unsigned long)grant_copy_op;
-+ mcl->args[2] = npo.copy_prod;
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++ list_del(&found_psdev->dev_list);
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++
++ /* the final put for releasing from the list */
++ pcistub_device_put(found_psdev);
+ }
++}
+
-+ /* Nothing to do? */
-+ if (!npo.mcl_prod)
-+ return;
++static struct pci_device_id pcistub_ids[] = {
++ {
++ .vendor = PCI_ANY_ID,
++ .device = PCI_ANY_ID,
++ .subvendor = PCI_ANY_ID,
++ .subdevice = PCI_ANY_ID,
++ },
++ {0,},
++};
+
-+ BUG_ON(npo.copy_prod > NET_RX_RING_SIZE);
-+ BUG_ON(npo.mmu_prod > NET_RX_RING_SIZE);
-+ BUG_ON(npo.trans_prod > NET_RX_RING_SIZE);
-+ BUG_ON(npo.mcl_prod > NET_RX_RING_SIZE+3);
-+ BUG_ON(npo.meta_prod > NET_RX_RING_SIZE);
++/*
++ * Note: There is no MODULE_DEVICE_TABLE entry here because this isn't
++ * for a normal device. I don't want it to be loaded automatically.
++ */
+
-+ ret = HYPERVISOR_multicall(npo.mcl, npo.mcl_prod);
-+ BUG_ON(ret != 0);
++static struct pci_driver pciback_pci_driver = {
++ .name = "pciback",
++ .id_table = pcistub_ids,
++ .probe = pcistub_probe,
++ .remove = pcistub_remove,
++};
+
-+ while ((skb = __skb_dequeue(&rxq)) != NULL) {
-+ nr_frags = *(int *)skb->cb;
++static inline int str_to_slot(const char *buf, int *domain, int *bus,
++ int *slot, int *func)
++{
++ int err;
+
-+ netif = netdev_priv(skb->dev);
-+ /* We can't rely on skb_release_data to release the
-+ pages used by fragments for us, since it tries to
-+ touch the pages in the fraglist. If we're in
-+ flipping mode, that doesn't work. In copying mode,
-+ we still have access to all of the pages, and so
-+ it's safe to let release_data deal with it. */
-+ /* (Freeing the fragments is safe since we copy
-+ non-linear skbs destined for flipping interfaces) */
-+ if (!netif->copying_receiver) {
-+ atomic_set(&(skb_shinfo(skb)->dataref), 1);
-+ skb_shinfo(skb)->frag_list = NULL;
-+ skb_shinfo(skb)->nr_frags = 0;
-+ netbk_free_pages(nr_frags, meta + npo.meta_cons + 1);
-+ }
++ err = sscanf(buf, " %x:%x:%x.%x", domain, bus, slot, func);
++ if (err == 4)
++ return 0;
++ else if (err < 0)
++ return -EINVAL;
+
-+ netif->stats.tx_bytes += skb->len;
-+ netif->stats.tx_packets++;
++ /* try again without domain */
++ *domain = 0;
++ err = sscanf(buf, " %x:%x.%x", bus, slot, func);
++ if (err == 3)
++ return 0;
+
-+ status = netbk_check_gop(nr_frags, netif->domid, &npo);
++ return -EINVAL;
++}
+
-+ id = meta[npo.meta_cons].id;
-+ flags = nr_frags ? NETRXF_more_data : 0;
++static inline int str_to_quirk(const char *buf, int *domain, int *bus, int
++ *slot, int *func, int *reg, int *size, int *mask)
++{
++ int err;
+
-+ if (skb->ip_summed == CHECKSUM_HW) /* local packet? */
-+ flags |= NETRXF_csum_blank | NETRXF_data_validated;
-+ else if (skb->proto_data_valid) /* remote but checksummed? */
-+ flags |= NETRXF_data_validated;
++ err =
++ sscanf(buf, " %04x:%02x:%02x.%1x-%08x:%1x:%08x", domain, bus, slot,
++ func, reg, size, mask);
++ if (err == 7)
++ return 0;
++ return -EINVAL;
++}
+
-+ if (meta[npo.meta_cons].copy)
-+ offset = 0;
-+ else
-+ offset = offset_in_page(skb->data);
-+ resp = make_rx_response(netif, id, status, offset,
-+ skb_headlen(skb), flags);
++static int pcistub_device_id_add(int domain, int bus, int slot, int func)
++{
++ struct pcistub_device_id *pci_dev_id;
++ unsigned long flags;
++
++ pci_dev_id = kmalloc(sizeof(*pci_dev_id), GFP_KERNEL);
++ if (!pci_dev_id)
++ return -ENOMEM;
++
++ pci_dev_id->domain = domain;
++ pci_dev_id->bus = bus;
++ pci_dev_id->devfn = PCI_DEVFN(slot, func);
++
++ pr_debug("pciback: wants to seize %04x:%02x:%02x.%01x\n",
++ domain, bus, slot, func);
++
++ spin_lock_irqsave(&device_ids_lock, flags);
++ list_add_tail(&pci_dev_id->slot_list, &pcistub_device_ids);
++ spin_unlock_irqrestore(&device_ids_lock, flags);
+
-+ if (meta[npo.meta_cons].frag.size) {
-+ struct netif_extra_info *gso =
-+ (struct netif_extra_info *)
-+ RING_GET_RESPONSE(&netif->rx,
-+ netif->rx.rsp_prod_pvt++);
++ return 0;
++}
+
-+ resp->flags |= NETRXF_extra_info;
++static int pcistub_device_id_remove(int domain, int bus, int slot, int func)
++{
++ struct pcistub_device_id *pci_dev_id, *t;
++ int devfn = PCI_DEVFN(slot, func);
++ int err = -ENOENT;
++ unsigned long flags;
+
-+ gso->u.gso.size = meta[npo.meta_cons].frag.size;
-+ gso->u.gso.type = XEN_NETIF_GSO_TYPE_TCPV4;
-+ gso->u.gso.pad = 0;
-+ gso->u.gso.features = 0;
++ spin_lock_irqsave(&device_ids_lock, flags);
++ list_for_each_entry_safe(pci_dev_id, t, &pcistub_device_ids, slot_list) {
+
-+ gso->type = XEN_NETIF_EXTRA_TYPE_GSO;
-+ gso->flags = 0;
-+ }
++ if (pci_dev_id->domain == domain
++ && pci_dev_id->bus == bus && pci_dev_id->devfn == devfn) {
++ /* Don't break; here because it's possible the same
++ * slot could be in the list more than once
++ */
++ list_del(&pci_dev_id->slot_list);
++ kfree(pci_dev_id);
+
-+ netbk_add_frag_responses(netif, status,
-+ meta + npo.meta_cons + 1,
-+ nr_frags);
++ err = 0;
+
-+ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&netif->rx, ret);
-+ irq = netif->irq;
-+ if (ret && !rx_notify[irq]) {
-+ rx_notify[irq] = 1;
-+ notify_list[notify_nr++] = irq;
++ pr_debug("pciback: removed %04x:%02x:%02x.%01x from "
++ "seize list\n", domain, bus, slot, func);
+ }
++ }
++ spin_unlock_irqrestore(&device_ids_lock, flags);
+
-+ if (netif_queue_stopped(netif->dev) &&
-+ netif_schedulable(netif) &&
-+ !netbk_queue_full(netif))
-+ netif_wake_queue(netif->dev);
++ return err;
++}
+
-+ netif_put(netif);
-+ dev_kfree_skb(skb);
-+ npo.meta_cons += nr_frags + 1;
++static int pcistub_reg_add(int domain, int bus, int slot, int func, int reg,
++ int size, int mask)
++{
++ int err = 0;
++ struct pcistub_device *psdev;
++ struct pci_dev *dev;
++ struct config_field *field;
++
++ psdev = pcistub_device_find(domain, bus, slot, func);
++ if (!psdev || !psdev->dev) {
++ err = -ENODEV;
++ goto out;
+ }
++ dev = psdev->dev;
+
-+ while (notify_nr != 0) {
-+ irq = notify_list[--notify_nr];
-+ rx_notify[irq] = 0;
-+ notify_remote_via_irq(irq);
++ field = kzalloc(sizeof(*field), GFP_ATOMIC);
++ if (!field) {
++ err = -ENOMEM;
++ goto out;
+ }
+
-+ /* More work to do? */
-+ if (!skb_queue_empty(&rx_queue) && !timer_pending(&net_timer))
-+ tasklet_schedule(&net_rx_tasklet);
-+#if 0
-+ else
-+ xen_network_done_notify();
-+#endif
-+}
++ field->offset = reg;
++ field->size = size;
++ field->mask = mask;
++ field->init = NULL;
++ field->reset = NULL;
++ field->release = NULL;
++ field->clean = pciback_config_field_free;
+
-+static void net_alarm(unsigned long unused)
-+{
-+ tasklet_schedule(&net_rx_tasklet);
++ err = pciback_config_quirks_add_field(dev, field);
++ if (err)
++ kfree(field);
++ out:
++ return err;
+}
+
-+struct net_device_stats *netif_be_get_stats(struct net_device *dev)
++static ssize_t pcistub_slot_add(struct device_driver *drv, const char *buf,
++ size_t count)
+{
-+ netif_t *netif = netdev_priv(dev);
-+ return &netif->stats;
-+}
++ int domain, bus, slot, func;
++ int err;
+
-+static int __on_net_schedule_list(netif_t *netif)
-+{
-+ return netif->list.next != NULL;
-+}
++ err = str_to_slot(buf, &domain, &bus, &slot, &func);
++ if (err)
++ goto out;
+
-+static void remove_from_net_schedule_list(netif_t *netif)
-+{
-+ spin_lock_irq(&net_schedule_list_lock);
-+ if (likely(__on_net_schedule_list(netif))) {
-+ list_del(&netif->list);
-+ netif->list.next = NULL;
-+ netif_put(netif);
-+ }
-+ spin_unlock_irq(&net_schedule_list_lock);
++ err = pcistub_device_id_add(domain, bus, slot, func);
++
++ out:
++ if (!err)
++ err = count;
++ return err;
+}
+
-+static void add_to_net_schedule_list_tail(netif_t *netif)
++DRIVER_ATTR(new_slot, S_IWUSR, NULL, pcistub_slot_add);
++
++static ssize_t pcistub_slot_remove(struct device_driver *drv, const char *buf,
++ size_t count)
+{
-+ if (__on_net_schedule_list(netif))
-+ return;
++ int domain, bus, slot, func;
++ int err;
+
-+ spin_lock_irq(&net_schedule_list_lock);
-+ if (!__on_net_schedule_list(netif) &&
-+ likely(netif_schedulable(netif))) {
-+ list_add_tail(&netif->list, &net_schedule_list);
-+ netif_get(netif);
-+ }
-+ spin_unlock_irq(&net_schedule_list_lock);
++ err = str_to_slot(buf, &domain, &bus, &slot, &func);
++ if (err)
++ goto out;
++
++ err = pcistub_device_id_remove(domain, bus, slot, func);
++
++ out:
++ if (!err)
++ err = count;
++ return err;
+}
+
-+/*
-+ * Note on CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER:
-+ * If this driver is pipelining transmit requests then we can be very
-+ * aggressive in avoiding new-packet notifications -- frontend only needs to
-+ * send a notification if there are no outstanding unreceived responses.
-+ * If we may be buffer transmit buffers for any reason then we must be rather
-+ * more conservative and treat this as the final check for pending work.
-+ */
-+void netif_schedule_work(netif_t *netif)
++DRIVER_ATTR(remove_slot, S_IWUSR, NULL, pcistub_slot_remove);
++
++static ssize_t pcistub_slot_show(struct device_driver *drv, char *buf)
+{
-+ int more_to_do;
++ struct pcistub_device_id *pci_dev_id;
++ size_t count = 0;
++ unsigned long flags;
+
-+#ifdef CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER
-+ more_to_do = RING_HAS_UNCONSUMED_REQUESTS(&netif->tx);
-+#else
-+ RING_FINAL_CHECK_FOR_REQUESTS(&netif->tx, more_to_do);
-+#endif
++ spin_lock_irqsave(&device_ids_lock, flags);
++ list_for_each_entry(pci_dev_id, &pcistub_device_ids, slot_list) {
++ if (count >= PAGE_SIZE)
++ break;
+
-+ if (more_to_do) {
-+ add_to_net_schedule_list_tail(netif);
-+ maybe_schedule_tx_action();
++ count += scnprintf(buf + count, PAGE_SIZE - count,
++ "%04x:%02x:%02x.%01x\n",
++ pci_dev_id->domain, pci_dev_id->bus,
++ PCI_SLOT(pci_dev_id->devfn),
++ PCI_FUNC(pci_dev_id->devfn));
+ }
-+}
++ spin_unlock_irqrestore(&device_ids_lock, flags);
+
-+void netif_deschedule_work(netif_t *netif)
-+{
-+ remove_from_net_schedule_list(netif);
++ return count;
+}
+
++DRIVER_ATTR(slots, S_IRUSR, pcistub_slot_show, NULL);
+
-+static void tx_add_credit(netif_t *netif)
++static ssize_t pcistub_quirk_add(struct device_driver *drv, const char *buf,
++ size_t count)
+{
-+ unsigned long max_burst, max_credit;
++ int domain, bus, slot, func, reg, size, mask;
++ int err;
+
-+ /*
-+ * Allow a burst big enough to transmit a jumbo packet of up to 128kB.
-+ * Otherwise the interface can seize up due to insufficient credit.
-+ */
-+ max_burst = RING_GET_REQUEST(&netif->tx, netif->tx.req_cons)->size;
-+ max_burst = min(max_burst, 131072UL);
-+ max_burst = max(max_burst, netif->credit_bytes);
++ err = str_to_quirk(buf, &domain, &bus, &slot, &func, ®, &size,
++ &mask);
++ if (err)
++ goto out;
+
-+ /* Take care that adding a new chunk of credit doesn't wrap to zero. */
-+ max_credit = netif->remaining_credit + netif->credit_bytes;
-+ if (max_credit < netif->remaining_credit)
-+ max_credit = ULONG_MAX; /* wrapped: clamp to ULONG_MAX */
++ err = pcistub_reg_add(domain, bus, slot, func, reg, size, mask);
+
-+ netif->remaining_credit = min(max_credit, max_burst);
++ out:
++ if (!err)
++ err = count;
++ return err;
+}
+
-+static void tx_credit_callback(unsigned long data)
++static ssize_t pcistub_quirk_show(struct device_driver *drv, char *buf)
+{
-+ netif_t *netif = (netif_t *)data;
-+ tx_add_credit(netif);
-+ netif_schedule_work(netif);
-+}
++ int count = 0;
++ unsigned long flags;
++ extern struct list_head pciback_quirks;
++ struct pciback_config_quirk *quirk;
++ struct pciback_dev_data *dev_data;
++ struct config_field *field;
++ struct config_field_entry *cfg_entry;
+
-+inline static void net_tx_action_dealloc(void)
-+{
-+ gnttab_unmap_grant_ref_t *gop;
-+ u16 pending_idx;
-+ PEND_RING_IDX dc, dp;
-+ netif_t *netif;
-+ int ret;
++ spin_lock_irqsave(&device_ids_lock, flags);
++ list_for_each_entry(quirk, &pciback_quirks, quirks_list) {
++ if (count >= PAGE_SIZE)
++ goto out;
+
-+ dc = dealloc_cons;
-+ dp = dealloc_prod;
++ count += scnprintf(buf + count, PAGE_SIZE - count,
++ "%02x:%02x.%01x\n\t%04x:%04x:%04x:%04x\n",
++ quirk->pdev->bus->number,
++ PCI_SLOT(quirk->pdev->devfn),
++ PCI_FUNC(quirk->pdev->devfn),
++ quirk->devid.vendor, quirk->devid.device,
++ quirk->devid.subvendor,
++ quirk->devid.subdevice);
+
-+ /* Ensure we see all indexes enqueued by netif_idx_release(). */
-+ smp_rmb();
++ dev_data = pci_get_drvdata(quirk->pdev);
+
-+ /*
-+ * Free up any grants we have finished using
-+ */
-+ gop = tx_unmap_ops;
-+ while (dc != dp) {
-+ pending_idx = dealloc_ring[MASK_PEND_IDX(dc++)];
-+ gnttab_set_unmap_op(gop, idx_to_kaddr(pending_idx),
-+ GNTMAP_host_map,
-+ grant_tx_handle[pending_idx]);
-+ gop++;
-+ }
-+ ret = HYPERVISOR_grant_table_op(
-+ GNTTABOP_unmap_grant_ref, tx_unmap_ops, gop - tx_unmap_ops);
-+ BUG_ON(ret);
++ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
++ field = cfg_entry->field;
++ if (count >= PAGE_SIZE)
++ goto out;
+
-+ while (dealloc_cons != dp) {
-+ pending_idx = dealloc_ring[MASK_PEND_IDX(dealloc_cons++)];
++ count += scnprintf(buf + count, PAGE_SIZE - count,
++ "\t\t%08x:%01x:%08x\n",
++ cfg_entry->base_offset + field->offset,
++ field->size, field->mask);
++ }
++ }
+
-+ netif = pending_tx_info[pending_idx].netif;
++ out:
++ spin_unlock_irqrestore(&device_ids_lock, flags);
+
-+ make_tx_response(netif, &pending_tx_info[pending_idx].req,
-+ NETIF_RSP_OKAY);
++ return count;
++}
+
-+ pending_ring[MASK_PEND_IDX(pending_prod++)] = pending_idx;
++DRIVER_ATTR(quirks, S_IRUSR | S_IWUSR, pcistub_quirk_show, pcistub_quirk_add);
+
-+ netif_put(netif);
++static ssize_t permissive_add(struct device_driver *drv, const char *buf,
++ size_t count)
++{
++ int domain, bus, slot, func;
++ int err;
++ struct pcistub_device *psdev;
++ struct pciback_dev_data *dev_data;
++ err = str_to_slot(buf, &domain, &bus, &slot, &func);
++ if (err)
++ goto out;
++ psdev = pcistub_device_find(domain, bus, slot, func);
++ if (!psdev) {
++ err = -ENODEV;
++ goto out;
++ }
++ if (!psdev->dev) {
++ err = -ENODEV;
++ goto release;
++ }
++ dev_data = pci_get_drvdata(psdev->dev);
++ /* the driver data for a device should never be null at this point */
++ if (!dev_data) {
++ err = -ENXIO;
++ goto release;
++ }
++ if (!dev_data->permissive) {
++ dev_data->permissive = 1;
++ /* Let user know that what they're doing could be unsafe */
++ dev_warn(&psdev->dev->dev,
++ "enabling permissive mode configuration space accesses!\n");
++ dev_warn(&psdev->dev->dev,
++ "permissive mode is potentially unsafe!\n");
+ }
++ release:
++ pcistub_device_put(psdev);
++ out:
++ if (!err)
++ err = count;
++ return err;
+}
+
-+static void netbk_tx_err(netif_t *netif, netif_tx_request_t *txp, RING_IDX end)
++static ssize_t permissive_show(struct device_driver *drv, char *buf)
+{
-+ RING_IDX cons = netif->tx.req_cons;
-+
-+ do {
-+ make_tx_response(netif, txp, NETIF_RSP_ERROR);
-+ if (cons >= end)
++ struct pcistub_device *psdev;
++ struct pciback_dev_data *dev_data;
++ size_t count = 0;
++ unsigned long flags;
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
++ if (count >= PAGE_SIZE)
+ break;
-+ txp = RING_GET_REQUEST(&netif->tx, cons++);
-+ } while (1);
-+ netif->tx.req_cons = cons;
-+ netif_schedule_work(netif);
-+ netif_put(netif);
++ if (!psdev->dev)
++ continue;
++ dev_data = pci_get_drvdata(psdev->dev);
++ if (!dev_data || !dev_data->permissive)
++ continue;
++ count +=
++ scnprintf(buf + count, PAGE_SIZE - count, "%s\n",
++ pci_name(psdev->dev));
++ }
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++ return count;
+}
+
-+static int netbk_count_requests(netif_t *netif, netif_tx_request_t *first,
-+ netif_tx_request_t *txp, int work_to_do)
-+{
-+ RING_IDX cons = netif->tx.req_cons;
-+ int frags = 0;
-+
-+ if (!(first->flags & NETTXF_more_data))
-+ return 0;
-+
-+ do {
-+ if (frags >= work_to_do) {
-+ DPRINTK("Need more frags\n");
-+ return -frags;
-+ }
++DRIVER_ATTR(permissive, S_IRUSR | S_IWUSR, permissive_show, permissive_add);
+
-+ if (unlikely(frags >= MAX_SKB_FRAGS)) {
-+ DPRINTK("Too many frags\n");
-+ return -frags;
-+ }
++#ifdef CONFIG_PCI_MSI
+
-+ memcpy(txp, RING_GET_REQUEST(&netif->tx, cons + frags),
-+ sizeof(*txp));
-+ if (txp->size > first->size) {
-+ DPRINTK("Frags galore\n");
-+ return -frags;
-+ }
++int pciback_get_owner(struct pci_dev *dev)
++{
++ struct pcistub_device *psdev;
+
-+ first->size -= txp->size;
-+ frags++;
++ psdev = pcistub_device_find(pci_domain_nr(dev->bus), dev->bus->number,
++ PCI_SLOT(dev->devfn), PCI_FUNC(dev->devfn));
+
-+ if (unlikely((txp->offset + txp->size) > PAGE_SIZE)) {
-+ DPRINTK("txp->offset: %x, size: %u\n",
-+ txp->offset, txp->size);
-+ return -frags;
-+ }
-+ } while ((txp++)->flags & NETTXF_more_data);
++ if (!psdev || !psdev->pdev)
++ return -1;
+
-+ return frags;
++ return psdev->pdev->xdev->otherend_id;
+}
++#endif
+
-+static gnttab_map_grant_ref_t *netbk_get_requests(netif_t *netif,
-+ struct sk_buff *skb,
-+ netif_tx_request_t *txp,
-+ gnttab_map_grant_ref_t *mop)
++static void pcistub_exit(void)
+{
-+ struct skb_shared_info *shinfo = skb_shinfo(skb);
-+ skb_frag_t *frags = shinfo->frags;
-+ unsigned long pending_idx = *((u16 *)skb->data);
-+ int i, start;
-+
-+ /* Skip first skb fragment if it is on same page as header fragment. */
-+ start = ((unsigned long)shinfo->frags[0].page == pending_idx);
++ driver_remove_file(&pciback_pci_driver.driver, &driver_attr_new_slot);
++ driver_remove_file(&pciback_pci_driver.driver,
++ &driver_attr_remove_slot);
++ driver_remove_file(&pciback_pci_driver.driver, &driver_attr_slots);
++ driver_remove_file(&pciback_pci_driver.driver, &driver_attr_quirks);
++ driver_remove_file(&pciback_pci_driver.driver, &driver_attr_permissive);
+
-+ for (i = start; i < shinfo->nr_frags; i++, txp++) {
-+ pending_idx = pending_ring[MASK_PEND_IDX(pending_cons++)];
++ pci_unregister_driver(&pciback_pci_driver);
++ WARN_ON(unregister_msi_get_owner(pciback_get_owner));
++}
+
-+ gnttab_set_map_op(mop++, idx_to_kaddr(pending_idx),
-+ GNTMAP_host_map | GNTMAP_readonly,
-+ txp->gref, netif->domid);
++static int __init pcistub_init(void)
++{
++ int pos = 0;
++ int err = 0;
++ int domain, bus, slot, func;
++ int parsed;
+
-+ memcpy(&pending_tx_info[pending_idx].req, txp, sizeof(*txp));
-+ netif_get(netif);
-+ pending_tx_info[pending_idx].netif = netif;
-+ frags[i].page = (void *)pending_idx;
-+ }
++ if (pci_devs_to_hide && *pci_devs_to_hide) {
++ do {
++ parsed = 0;
+
-+ return mop;
-+}
++ err = sscanf(pci_devs_to_hide + pos,
++ " (%x:%x:%x.%x) %n",
++ &domain, &bus, &slot, &func, &parsed);
++ if (err != 4) {
++ domain = 0;
++ err = sscanf(pci_devs_to_hide + pos,
++ " (%x:%x.%x) %n",
++ &bus, &slot, &func, &parsed);
++ if (err != 3)
++ goto parse_error;
++ }
+
-+static int netbk_tx_check_mop(struct sk_buff *skb,
-+ gnttab_map_grant_ref_t **mopp)
-+{
-+ gnttab_map_grant_ref_t *mop = *mopp;
-+ int pending_idx = *((u16 *)skb->data);
-+ netif_t *netif = pending_tx_info[pending_idx].netif;
-+ netif_tx_request_t *txp;
-+ struct skb_shared_info *shinfo = skb_shinfo(skb);
-+ int nr_frags = shinfo->nr_frags;
-+ int i, err, start;
++ err = pcistub_device_id_add(domain, bus, slot, func);
++ if (err)
++ goto out;
+
-+ /* Check status of header. */
-+ err = mop->status;
-+ if (unlikely(err)) {
-+ txp = &pending_tx_info[pending_idx].req;
-+ make_tx_response(netif, txp, NETIF_RSP_ERROR);
-+ pending_ring[MASK_PEND_IDX(pending_prod++)] = pending_idx;
-+ netif_put(netif);
-+ } else {
-+ set_phys_to_machine(
-+ __pa(idx_to_kaddr(pending_idx)) >> PAGE_SHIFT,
-+ FOREIGN_FRAME(mop->dev_bus_addr >> PAGE_SHIFT));
-+ grant_tx_handle[pending_idx] = mop->handle;
++ /* if parsed<=0, we've reached the end of the string */
++ pos += parsed;
++ } while (parsed > 0 && pci_devs_to_hide[pos]);
+ }
+
-+ /* Skip first skb fragment if it is on same page as header fragment. */
-+ start = ((unsigned long)shinfo->frags[0].page == pending_idx);
++ /* If we're the first PCI Device Driver to register, we're the
++ * first one to get offered PCI devices as they become
++ * available (and thus we can be the first to grab them)
++ */
++ err = pci_register_driver(&pciback_pci_driver);
++ if (err < 0)
++ goto out;
+
-+ for (i = start; i < nr_frags; i++) {
-+ int j, newerr;
++ err = driver_create_file(&pciback_pci_driver.driver,
++ &driver_attr_new_slot);
++ if (!err)
++ err = driver_create_file(&pciback_pci_driver.driver,
++ &driver_attr_remove_slot);
++ if (!err)
++ err = driver_create_file(&pciback_pci_driver.driver,
++ &driver_attr_slots);
++ if (!err)
++ err = driver_create_file(&pciback_pci_driver.driver,
++ &driver_attr_quirks);
++ if (!err)
++ err = driver_create_file(&pciback_pci_driver.driver,
++ &driver_attr_permissive);
+
-+ pending_idx = (unsigned long)shinfo->frags[i].page;
++ if (!err)
++ err = register_msi_get_owner(pciback_get_owner);
++ if (err)
++ pcistub_exit();
+
-+ /* Check error status: if okay then remember grant handle. */
-+ newerr = (++mop)->status;
-+ if (likely(!newerr)) {
-+ set_phys_to_machine(
-+ __pa(idx_to_kaddr(pending_idx))>>PAGE_SHIFT,
-+ FOREIGN_FRAME(mop->dev_bus_addr>>PAGE_SHIFT));
-+ grant_tx_handle[pending_idx] = mop->handle;
-+ /* Had a previous error? Invalidate this fragment. */
-+ if (unlikely(err))
-+ netif_idx_release(pending_idx);
-+ continue;
-+ }
++ out:
++ return err;
+
-+ /* Error on this fragment: respond to client with an error. */
-+ txp = &pending_tx_info[pending_idx].req;
-+ make_tx_response(netif, txp, NETIF_RSP_ERROR);
-+ pending_ring[MASK_PEND_IDX(pending_prod++)] = pending_idx;
-+ netif_put(netif);
++ parse_error:
++ printk(KERN_ERR "pciback: Error parsing pci_devs_to_hide at \"%s\"\n",
++ pci_devs_to_hide + pos);
++ return -EINVAL;
++}
+
-+ /* Not the first error? Preceding frags already invalidated. */
-+ if (err)
-+ continue;
++#ifndef MODULE
++/*
++ * fs_initcall happens before device_initcall
++ * so pciback *should* get called first (b/c we
++ * want to suck up any device before other drivers
++ * get a chance by being the first pci device
++ * driver to register)
++ */
++fs_initcall(pcistub_init);
++#endif
+
-+ /* First error: invalidate header and preceding fragments. */
-+ pending_idx = *((u16 *)skb->data);
-+ netif_idx_release(pending_idx);
-+ for (j = start; j < i; j++) {
-+ pending_idx = (unsigned long)shinfo->frags[i].page;
-+ netif_idx_release(pending_idx);
-+ }
++static int __init pciback_init(void)
++{
++ int err;
+
-+ /* Remember the error: invalidate all subsequent fragments. */
-+ err = newerr;
-+ }
++ err = pciback_config_init();
++ if (err)
++ return err;
++
++#ifdef MODULE
++ err = pcistub_init();
++ if (err < 0)
++ return err;
++#endif
++
++ pcistub_init_devices_late();
++ err = pciback_xenbus_register();
++ if (err)
++ pcistub_exit();
+
-+ *mopp = mop + 1;
+ return err;
+}
+
-+static void netbk_fill_frags(struct sk_buff *skb)
++static void __exit pciback_cleanup(void)
+{
-+ struct skb_shared_info *shinfo = skb_shinfo(skb);
-+ int nr_frags = shinfo->nr_frags;
-+ int i;
++ pciback_xenbus_unregister();
++ pcistub_exit();
++}
+
-+ for (i = 0; i < nr_frags; i++) {
-+ skb_frag_t *frag = shinfo->frags + i;
-+ netif_tx_request_t *txp;
-+ unsigned long pending_idx;
++module_init(pciback_init);
++module_exit(pciback_cleanup);
+
-+ pending_idx = (unsigned long)frag->page;
-+ txp = &pending_tx_info[pending_idx].req;
-+ frag->page = virt_to_page(idx_to_kaddr(pending_idx));
-+ frag->size = txp->size;
-+ frag->page_offset = txp->offset;
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/pciback.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/pciback.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,111 @@
++/*
++ * PCI Backend Common Data Structures & Function Declarations
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#ifndef __XEN_PCIBACK_H__
++#define __XEN_PCIBACK_H__
+
-+ skb->len += txp->size;
-+ skb->data_len += txp->size;
-+ skb->truesize += txp->size;
-+ }
-+}
++#include <linux/pci.h>
++#include <linux/interrupt.h>
++#include <xen/xenbus.h>
++#include <linux/list.h>
++#include <linux/spinlock.h>
++#include <linux/workqueue.h>
++#include <asm/atomic.h>
++#include <xen/interface/io/pciif.h>
+
-+int netbk_get_extras(netif_t *netif, struct netif_extra_info *extras,
-+ int work_to_do)
-+{
-+ struct netif_extra_info extra;
-+ RING_IDX cons = netif->tx.req_cons;
++struct pci_dev_entry {
++ struct list_head list;
++ struct pci_dev *dev;
++};
+
-+ do {
-+ if (unlikely(work_to_do-- <= 0)) {
-+ DPRINTK("Missing extra info\n");
-+ return -EBADR;
-+ }
++#define _PDEVF_op_active (0)
++#define PDEVF_op_active (1<<(_PDEVF_op_active))
+
-+ memcpy(&extra, RING_GET_REQUEST(&netif->tx, cons),
-+ sizeof(extra));
-+ if (unlikely(!extra.type ||
-+ extra.type >= XEN_NETIF_EXTRA_TYPE_MAX)) {
-+ netif->tx.req_cons = ++cons;
-+ DPRINTK("Invalid extra type: %d\n", extra.type);
-+ return -EINVAL;
-+ }
++struct pciback_device {
++ void *pci_dev_data;
++ spinlock_t dev_lock;
+
-+ memcpy(&extras[extra.type - 1], &extra, sizeof(extra));
-+ netif->tx.req_cons = ++cons;
-+ } while (extra.flags & XEN_NETIF_EXTRA_FLAG_MORE);
++ struct xenbus_device *xdev;
+
-+ return work_to_do;
-+}
++ struct xenbus_watch be_watch;
++ u8 be_watching;
+
-+static int netbk_set_skb_gso(struct sk_buff *skb, struct netif_extra_info *gso)
-+{
-+ if (!gso->u.gso.size) {
-+ DPRINTK("GSO size must not be zero.\n");
-+ return -EINVAL;
-+ }
++ int evtchn_irq;
+
-+ /* Currently only TCPv4 S.O. is supported. */
-+ if (gso->u.gso.type != XEN_NETIF_GSO_TYPE_TCPV4) {
-+ DPRINTK("Bad GSO type %d.\n", gso->u.gso.type);
-+ return -EINVAL;
-+ }
++ struct vm_struct *sh_area;
++ struct xen_pci_sharedinfo *sh_info;
+
-+ skb_shinfo(skb)->gso_size = gso->u.gso.size;
-+ skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4;
++ unsigned long flags;
+
-+ /* Header must be checked, and gso_segs computed. */
-+ skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY;
-+ skb_shinfo(skb)->gso_segs = 0;
++ struct work_struct op_work;
++};
+
-+ return 0;
-+}
++struct pciback_dev_data {
++ struct list_head config_fields;
++ int permissive;
++ int warned_on_write;
++};
+
-+/* Called after netfront has transmitted */
-+static void net_tx_action(unsigned long unused)
-+{
-+ struct list_head *ent;
-+ struct sk_buff *skb;
-+ netif_t *netif;
-+ netif_tx_request_t txreq;
-+ netif_tx_request_t txfrags[MAX_SKB_FRAGS];
-+ struct netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX - 1];
-+ u16 pending_idx;
-+ RING_IDX i;
-+ gnttab_map_grant_ref_t *mop;
-+ unsigned int data_len;
-+ int ret, work_to_do;
++/* Get/Put PCI Devices that are hidden from the PCI Backend Domain */
++struct pci_dev *pcistub_get_pci_dev_by_slot(struct pciback_device *pdev,
++ int domain, int bus,
++ int slot, int func);
++struct pci_dev *pcistub_get_pci_dev(struct pciback_device *pdev,
++ struct pci_dev *dev);
++void pcistub_put_pci_dev(struct pci_dev *dev);
+
-+ if (dealloc_cons != dealloc_prod)
-+ net_tx_action_dealloc();
++/* Ensure a device is turned off or reset */
++void pciback_reset_device(struct pci_dev *pdev);
+
-+ mop = tx_map_ops;
-+ while (((NR_PENDING_REQS + MAX_SKB_FRAGS) < MAX_PENDING_REQS) &&
-+ !list_empty(&net_schedule_list)) {
-+ /* Get a netif from the list with work to do. */
-+ ent = net_schedule_list.next;
-+ netif = list_entry(ent, netif_t, list);
-+ netif_get(netif);
-+ remove_from_net_schedule_list(netif);
++/* Access a virtual configuration space for a PCI device */
++int pciback_config_init(void);
++int pciback_config_init_dev(struct pci_dev *dev);
++void pciback_config_free_dyn_fields(struct pci_dev *dev);
++void pciback_config_reset_dev(struct pci_dev *dev);
++void pciback_config_free_dev(struct pci_dev *dev);
++int pciback_config_read(struct pci_dev *dev, int offset, int size,
++ u32 * ret_val);
++int pciback_config_write(struct pci_dev *dev, int offset, int size, u32 value);
+
-+ RING_FINAL_CHECK_FOR_REQUESTS(&netif->tx, work_to_do);
-+ if (!work_to_do) {
-+ netif_put(netif);
-+ continue;
-+ }
++/* Handle requests for specific devices from the frontend */
++typedef int (*publish_pci_dev_cb) (struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn, unsigned int devid);
++typedef int (*publish_pci_root_cb) (struct pciback_device * pdev,
++ unsigned int domain, unsigned int bus);
++int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev,
++ int devid, publish_pci_dev_cb publish_cb);
++void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev);
++struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn);
++int pciback_init_devices(struct pciback_device *pdev);
++int pciback_publish_pci_roots(struct pciback_device *pdev,
++ publish_pci_root_cb cb);
++void pciback_release_devices(struct pciback_device *pdev);
+
-+ i = netif->tx.req_cons;
-+ rmb(); /* Ensure that we see the request before we copy it. */
-+ memcpy(&txreq, RING_GET_REQUEST(&netif->tx, i), sizeof(txreq));
++/* Handles events from front-end */
++irqreturn_t pciback_handle_event(int irq, void *dev_id, struct pt_regs *regs);
++void pciback_do_op(void *data);
+
-+ /* Credit-based scheduling. */
-+ if (txreq.size > netif->remaining_credit) {
-+ unsigned long now = jiffies;
-+ unsigned long next_credit =
-+ netif->credit_timeout.expires +
-+ msecs_to_jiffies(netif->credit_usec / 1000);
++int pciback_xenbus_register(void);
++void pciback_xenbus_unregister(void);
+
-+ /* Timer could already be pending in rare cases. */
-+ if (timer_pending(&netif->credit_timeout)) {
-+ netif_put(netif);
-+ continue;
-+ }
++#ifdef CONFIG_PCI_MSI
++int pciback_enable_msi(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op);
+
-+ /* Passed the point where we can replenish credit? */
-+ if (time_after_eq(now, next_credit)) {
-+ netif->credit_timeout.expires = now;
-+ tx_add_credit(netif);
-+ }
++int pciback_disable_msi(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op);
+
-+ /* Still too big to send right now? Set a callback. */
-+ if (txreq.size > netif->remaining_credit) {
-+ netif->credit_timeout.data =
-+ (unsigned long)netif;
-+ netif->credit_timeout.function =
-+ tx_credit_callback;
-+ __mod_timer(&netif->credit_timeout,
-+ next_credit);
-+ netif_put(netif);
-+ continue;
-+ }
-+ }
-+ netif->remaining_credit -= txreq.size;
+
-+ work_to_do--;
-+ netif->tx.req_cons = ++i;
++int pciback_enable_msix(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op);
+
-+ memset(extras, 0, sizeof(extras));
-+ if (txreq.flags & NETTXF_extra_info) {
-+ work_to_do = netbk_get_extras(netif, extras,
-+ work_to_do);
-+ i = netif->tx.req_cons;
-+ if (unlikely(work_to_do < 0)) {
-+ netbk_tx_err(netif, &txreq, i);
-+ continue;
-+ }
-+ }
++int pciback_disable_msix(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op);
++#endif
++extern int verbose_request;
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/pciback_ops.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/pciback_ops.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,117 @@
++/*
++ * PCI Backend Operations - respond to PCI requests from Frontend
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <asm/bitops.h>
++#include <xen/evtchn.h>
++#include "pciback.h"
+
-+ ret = netbk_count_requests(netif, &txreq, txfrags, work_to_do);
-+ if (unlikely(ret < 0)) {
-+ netbk_tx_err(netif, &txreq, i - ret);
-+ continue;
-+ }
-+ i += ret;
++int verbose_request = 0;
++module_param(verbose_request, int, 0644);
+
-+ if (unlikely(txreq.size < ETH_HLEN)) {
-+ DPRINTK("Bad packet size: %d\n", txreq.size);
-+ netbk_tx_err(netif, &txreq, i);
-+ continue;
-+ }
++/* Ensure a device is "turned off" and ready to be exported.
++ * (Also see pciback_config_reset to ensure virtual configuration space is
++ * ready to be re-exported)
++ */
++void pciback_reset_device(struct pci_dev *dev)
++{
++ u16 cmd;
+
-+ /* No crossing a page as the payload mustn't fragment. */
-+ if (unlikely((txreq.offset + txreq.size) > PAGE_SIZE)) {
-+ DPRINTK("txreq.offset: %x, size: %u, end: %lu\n",
-+ txreq.offset, txreq.size,
-+ (txreq.offset &~PAGE_MASK) + txreq.size);
-+ netbk_tx_err(netif, &txreq, i);
-+ continue;
-+ }
++ /* Disable devices (but not bridges) */
++ if (dev->hdr_type == PCI_HEADER_TYPE_NORMAL) {
++ pci_disable_device(dev);
+
-+ pending_idx = pending_ring[MASK_PEND_IDX(pending_cons)];
++ pci_write_config_word(dev, PCI_COMMAND, 0);
+
-+ data_len = (txreq.size > PKT_PROT_LEN &&
-+ ret < MAX_SKB_FRAGS) ?
-+ PKT_PROT_LEN : txreq.size;
++ dev->is_enabled = 0;
++ dev->is_busmaster = 0;
++ } else {
++ pci_read_config_word(dev, PCI_COMMAND, &cmd);
++ if (cmd & (PCI_COMMAND_INVALIDATE)) {
++ cmd &= ~(PCI_COMMAND_INVALIDATE);
++ pci_write_config_word(dev, PCI_COMMAND, cmd);
+
-+ skb = alloc_skb(data_len + 16 + NET_IP_ALIGN,
-+ GFP_ATOMIC | __GFP_NOWARN);
-+ if (unlikely(skb == NULL)) {
-+ DPRINTK("Can't allocate a skb in start_xmit.\n");
-+ netbk_tx_err(netif, &txreq, i);
-+ break;
++ dev->is_busmaster = 0;
+ }
++ }
++}
+
-+ /* Packets passed to netif_rx() must have some headroom. */
-+ skb_reserve(skb, 16 + NET_IP_ALIGN);
++static inline void test_and_schedule_op(struct pciback_device *pdev)
++{
++ /* Check that frontend is requesting an operation and that we are not
++ * already processing a request */
++ if (test_bit(_XEN_PCIF_active, (unsigned long *)&pdev->sh_info->flags)
++ && !test_and_set_bit(_PDEVF_op_active, &pdev->flags))
++ schedule_work(&pdev->op_work);
++}
+
-+ if (extras[XEN_NETIF_EXTRA_TYPE_GSO - 1].type) {
-+ struct netif_extra_info *gso;
-+ gso = &extras[XEN_NETIF_EXTRA_TYPE_GSO - 1];
++/* Performing the configuration space reads/writes must not be done in atomic
++ * context because some of the pci_* functions can sleep (mostly due to ACPI
++ * use of semaphores). This function is intended to be called from a work
++ * queue in process context taking a struct pciback_device as a parameter */
++void pciback_do_op(void *data)
++{
++ struct pciback_device *pdev = data;
++ struct pci_dev *dev;
++ struct xen_pci_op *op = &pdev->sh_info->op;
+
-+ if (netbk_set_skb_gso(skb, gso)) {
-+ kfree_skb(skb);
-+ netbk_tx_err(netif, &txreq, i);
-+ continue;
-+ }
++ dev = pciback_get_pci_dev(pdev, op->domain, op->bus, op->devfn);
++
++ if (dev == NULL)
++ op->err = XEN_PCI_ERR_dev_not_found;
++ else
++ {
++ switch (op->cmd)
++ {
++ case XEN_PCI_OP_conf_read:
++ op->err = pciback_config_read(dev,
++ op->offset, op->size, &op->value);
++ break;
++ case XEN_PCI_OP_conf_write:
++ op->err = pciback_config_write(dev,
++ op->offset, op->size, op->value);
++ break;
++#ifdef CONFIG_PCI_MSI
++ case XEN_PCI_OP_enable_msi:
++ op->err = pciback_enable_msi(pdev, dev, op);
++ break;
++ case XEN_PCI_OP_disable_msi:
++ op->err = pciback_disable_msi(pdev, dev, op);
++ break;
++ case XEN_PCI_OP_enable_msix:
++ op->err = pciback_enable_msix(pdev, dev, op);
++ break;
++ case XEN_PCI_OP_disable_msix:
++ op->err = pciback_disable_msix(pdev, dev, op);
++ break;
++#endif
++ default:
++ op->err = XEN_PCI_ERR_not_implemented;
++ break;
+ }
++ }
++ /* Tell the driver domain that we're done. */
++ wmb();
++ clear_bit(_XEN_PCIF_active, (unsigned long *)&pdev->sh_info->flags);
++ notify_remote_via_irq(pdev->evtchn_irq);
+
-+ gnttab_set_map_op(mop, idx_to_kaddr(pending_idx),
-+ GNTMAP_host_map | GNTMAP_readonly,
-+ txreq.gref, netif->domid);
-+ mop++;
++ /* Mark that we're done. */
++ smp_mb__before_clear_bit(); /* /after/ clearing PCIF_active */
++ clear_bit(_PDEVF_op_active, &pdev->flags);
++ smp_mb__after_clear_bit(); /* /before/ final check for work */
+
-+ memcpy(&pending_tx_info[pending_idx].req,
-+ &txreq, sizeof(txreq));
-+ pending_tx_info[pending_idx].netif = netif;
-+ *((u16 *)skb->data) = pending_idx;
++ /* Check to see if the driver domain tried to start another request in
++ * between clearing _XEN_PCIF_active and clearing _PDEVF_op_active. */
++ test_and_schedule_op(pdev);
++}
+
-+ __skb_put(skb, data_len);
++irqreturn_t pciback_handle_event(int irq, void *dev_id, struct pt_regs *regs)
++{
++ struct pciback_device *pdev = dev_id;
+
-+ skb_shinfo(skb)->nr_frags = ret;
-+ if (data_len < txreq.size) {
-+ skb_shinfo(skb)->nr_frags++;
-+ skb_shinfo(skb)->frags[0].page =
-+ (void *)(unsigned long)pending_idx;
-+ } else {
-+ /* Discriminate from any valid pending_idx value. */
-+ skb_shinfo(skb)->frags[0].page = (void *)~0UL;
-+ }
++ test_and_schedule_op(pdev);
+
-+ __skb_queue_tail(&tx_queue, skb);
++ return IRQ_HANDLED;
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/slot.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/slot.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,157 @@
++/*
++ * PCI Backend - Provides a Virtual PCI bus (with real devices)
++ * to the frontend
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil> (vpci.c)
++ * Author: Tristan Gingold <tristan.gingold at bull.net>, from vpci.c
++ */
+
-+ pending_cons++;
++#include <linux/list.h>
++#include <linux/slab.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include "pciback.h"
+
-+ mop = netbk_get_requests(netif, skb, txfrags, mop);
++/* There are at most 32 slots in a pci bus. */
++#define PCI_SLOT_MAX 32
+
-+ netif->tx.req_cons = i;
-+ netif_schedule_work(netif);
++#define PCI_BUS_NBR 2
+
-+ if ((mop - tx_map_ops) >= ARRAY_SIZE(tx_map_ops))
-+ break;
-+ }
++struct slot_dev_data {
++ /* Access to dev_list must be protected by lock */
++ struct pci_dev *slots[PCI_BUS_NBR][PCI_SLOT_MAX];
++ spinlock_t lock;
++};
+
-+ if (mop == tx_map_ops)
-+ return;
++struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn)
++{
++ struct pci_dev *dev = NULL;
++ struct slot_dev_data *slot_dev = pdev->pci_dev_data;
++ unsigned long flags;
+
-+ ret = HYPERVISOR_grant_table_op(
-+ GNTTABOP_map_grant_ref, tx_map_ops, mop - tx_map_ops);
-+ BUG_ON(ret);
++ if (domain != 0 || PCI_FUNC(devfn) != 0)
++ return NULL;
+
-+ mop = tx_map_ops;
-+ while ((skb = __skb_dequeue(&tx_queue)) != NULL) {
-+ netif_tx_request_t *txp;
++ if (PCI_SLOT(devfn) >= PCI_SLOT_MAX || bus >= PCI_BUS_NBR)
++ return NULL;
+
-+ pending_idx = *((u16 *)skb->data);
-+ netif = pending_tx_info[pending_idx].netif;
-+ txp = &pending_tx_info[pending_idx].req;
++ spin_lock_irqsave(&slot_dev->lock, flags);
++ dev = slot_dev->slots[bus][PCI_SLOT(devfn)];
++ spin_unlock_irqrestore(&slot_dev->lock, flags);
+
-+ /* Check the remap error code. */
-+ if (unlikely(netbk_tx_check_mop(skb, &mop))) {
-+ DPRINTK("netback grant failed.\n");
-+ skb_shinfo(skb)->nr_frags = 0;
-+ kfree_skb(skb);
-+ continue;
-+ }
++ return dev;
++}
+
-+ data_len = skb->len;
-+ memcpy(skb->data,
-+ (void *)(idx_to_kaddr(pending_idx)|txp->offset),
-+ data_len);
-+ if (data_len < txp->size) {
-+ /* Append the packet payload as a fragment. */
-+ txp->offset += data_len;
-+ txp->size -= data_len;
-+ } else {
-+ /* Schedule a response immediately. */
-+ netif_idx_release(pending_idx);
-+ }
++int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev,
++ int devid, publish_pci_dev_cb publish_cb)
++{
++ int err = 0, slot, bus;
++ struct slot_dev_data *slot_dev = pdev->pci_dev_data;
++ unsigned long flags;
+
-+ /*
-+ * Old frontends do not assert data_validated but we
-+ * can infer it from csum_blank so test both flags.
-+ */
-+ if (txp->flags & (NETTXF_data_validated|NETTXF_csum_blank)) {
-+ skb->ip_summed = CHECKSUM_UNNECESSARY;
-+ skb->proto_data_valid = 1;
-+ } else {
-+ skb->ip_summed = CHECKSUM_NONE;
-+ skb->proto_data_valid = 0;
++ if ((dev->class >> 24) == PCI_BASE_CLASS_BRIDGE) {
++ err = -EFAULT;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Can't export bridges on the virtual PCI bus");
++ goto out;
++ }
++
++ spin_lock_irqsave(&slot_dev->lock, flags);
++
++ /* Assign to a new slot on the virtual PCI bus */
++ for (bus = 0; bus < PCI_BUS_NBR; bus++)
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ if (slot_dev->slots[bus][slot] == NULL) {
++ printk(KERN_INFO
++ "pciback: slot: %s: assign to virtual slot %d, bus %d\n",
++ pci_name(dev), slot, bus);
++ slot_dev->slots[bus][slot] = dev;
++ goto unlock;
++ }
+ }
-+ skb->proto_csum_blank = !!(txp->flags & NETTXF_csum_blank);
+
-+ netbk_fill_frags(skb);
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "No more space on root virtual PCI bus");
+
-+ skb->dev = netif->dev;
-+ skb->protocol = eth_type_trans(skb, skb->dev);
++ unlock:
++ spin_unlock_irqrestore(&slot_dev->lock, flags);
+
-+ netif->stats.rx_bytes += skb->len;
-+ netif->stats.rx_packets++;
++ /* Publish this device. */
++ if(!err)
++ err = publish_cb(pdev, 0, 0, PCI_DEVFN(slot, 0), devid);
+
-+ netif_rx(skb);
-+ netif->dev->last_rx = jiffies;
-+ }
++ out:
++ return err;
+}
+
-+static void netif_idx_release(u16 pending_idx)
++void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
+{
-+ static DEFINE_SPINLOCK(_lock);
++ int slot, bus;
++ struct slot_dev_data *slot_dev = pdev->pci_dev_data;
++ struct pci_dev *found_dev = NULL;
+ unsigned long flags;
+
-+ spin_lock_irqsave(&_lock, flags);
-+ dealloc_ring[MASK_PEND_IDX(dealloc_prod)] = pending_idx;
-+ /* Sync with net_tx_action_dealloc: insert idx /then/ incr producer. */
-+ smp_wmb();
-+ dealloc_prod++;
-+ spin_unlock_irqrestore(&_lock, flags);
++ spin_lock_irqsave(&slot_dev->lock, flags);
+
-+ tasklet_schedule(&net_tx_tasklet);
-+}
++ for (bus = 0; bus < PCI_BUS_NBR; bus++)
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ if (slot_dev->slots[bus][slot] == dev) {
++ slot_dev->slots[bus][slot] = NULL;
++ found_dev = dev;
++ goto out;
++ }
++ }
+
-+static void netif_page_release(struct page *page)
-+{
-+ /* Ready for next use. */
-+ init_page_count(page);
++ out:
++ spin_unlock_irqrestore(&slot_dev->lock, flags);
+
-+ netif_idx_release(netif_page_index(page));
++ if (found_dev)
++ pcistub_put_pci_dev(found_dev);
+}
+
-+irqreturn_t netif_be_int(int irq, void *dev_id, struct pt_regs *regs)
++int pciback_init_devices(struct pciback_device *pdev)
+{
-+ netif_t *netif = dev_id;
++ int slot, bus;
++ struct slot_dev_data *slot_dev;
+
-+ add_to_net_schedule_list_tail(netif);
-+ maybe_schedule_tx_action();
++ slot_dev = kmalloc(sizeof(*slot_dev), GFP_KERNEL);
++ if (!slot_dev)
++ return -ENOMEM;
+
-+ if (netif_schedulable(netif) && !netbk_queue_full(netif))
-+ netif_wake_queue(netif->dev);
++ spin_lock_init(&slot_dev->lock);
+
-+ return IRQ_HANDLED;
++ for (bus = 0; bus < PCI_BUS_NBR; bus++)
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++)
++ slot_dev->slots[bus][slot] = NULL;
++
++ pdev->pci_dev_data = slot_dev;
++
++ return 0;
+}
+
-+static void make_tx_response(netif_t *netif,
-+ netif_tx_request_t *txp,
-+ s8 st)
++int pciback_publish_pci_roots(struct pciback_device *pdev,
++ publish_pci_root_cb publish_cb)
+{
-+ RING_IDX i = netif->tx.rsp_prod_pvt;
-+ netif_tx_response_t *resp;
-+ int notify;
-+
-+ resp = RING_GET_RESPONSE(&netif->tx, i);
-+ resp->id = txp->id;
-+ resp->status = st;
++ /* The Virtual PCI bus has only one root */
++ return publish_cb(pdev, 0, 0);
++}
+
-+ if (txp->flags & NETTXF_extra_info)
-+ RING_GET_RESPONSE(&netif->tx, ++i)->status = NETIF_RSP_NULL;
++void pciback_release_devices(struct pciback_device *pdev)
++{
++ int slot, bus;
++ struct slot_dev_data *slot_dev = pdev->pci_dev_data;
++ struct pci_dev *dev;
+
-+ netif->tx.rsp_prod_pvt = ++i;
-+ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&netif->tx, notify);
-+ if (notify)
-+ notify_remote_via_irq(netif->irq);
++ for (bus = 0; bus < PCI_BUS_NBR; bus++)
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ dev = slot_dev->slots[bus][slot];
++ if (dev != NULL)
++ pcistub_put_pci_dev(dev);
++ }
+
-+#ifdef CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER
-+ if (i == netif->tx.req_cons) {
-+ int more_to_do;
-+ RING_FINAL_CHECK_FOR_REQUESTS(&netif->tx, more_to_do);
-+ if (more_to_do)
-+ add_to_net_schedule_list_tail(netif);
-+ }
-+#endif
++ kfree(slot_dev);
++ pdev->pci_dev_data = NULL;
+}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/vpci.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/vpci.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,212 @@
++/*
++ * PCI Backend - Provides a Virtual PCI bus (with real devices)
++ * to the frontend
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
+
-+static netif_rx_response_t *make_rx_response(netif_t *netif,
-+ u16 id,
-+ s8 st,
-+ u16 offset,
-+ u16 size,
-+ u16 flags)
-+{
-+ RING_IDX i = netif->rx.rsp_prod_pvt;
-+ netif_rx_response_t *resp;
++#include <linux/list.h>
++#include <linux/slab.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include "pciback.h"
+
-+ resp = RING_GET_RESPONSE(&netif->rx, i);
-+ resp->offset = offset;
-+ resp->flags = flags;
-+ resp->id = id;
-+ resp->status = (s16)size;
-+ if (st < 0)
-+ resp->status = (s16)st;
++#define PCI_SLOT_MAX 32
+
-+ netif->rx.rsp_prod_pvt = ++i;
++struct vpci_dev_data {
++ /* Access to dev_list must be protected by lock */
++ struct list_head dev_list[PCI_SLOT_MAX];
++ spinlock_t lock;
++};
+
-+ return resp;
++static inline struct list_head *list_first(struct list_head *head)
++{
++ return head->next;
+}
+
-+#ifdef NETBE_DEBUG_INTERRUPT
-+static irqreturn_t netif_be_dbg(int irq, void *dev_id, struct pt_regs *regs)
++struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn)
+{
-+ struct list_head *ent;
-+ netif_t *netif;
-+ int i = 0;
++ struct pci_dev_entry *entry;
++ struct pci_dev *dev = NULL;
++ struct vpci_dev_data *vpci_dev = pdev->pci_dev_data;
++ unsigned long flags;
+
-+ printk(KERN_ALERT "netif_schedule_list:\n");
-+ spin_lock_irq(&net_schedule_list_lock);
++ if (domain != 0 || bus != 0)
++ return NULL;
+
-+ list_for_each (ent, &net_schedule_list) {
-+ netif = list_entry(ent, netif_t, list);
-+ printk(KERN_ALERT " %d: private(rx_req_cons=%08x "
-+ "rx_resp_prod=%08x\n",
-+ i, netif->rx.req_cons, netif->rx.rsp_prod_pvt);
-+ printk(KERN_ALERT " tx_req_cons=%08x tx_resp_prod=%08x)\n",
-+ netif->tx.req_cons, netif->tx.rsp_prod_pvt);
-+ printk(KERN_ALERT " shared(rx_req_prod=%08x "
-+ "rx_resp_prod=%08x\n",
-+ netif->rx.sring->req_prod, netif->rx.sring->rsp_prod);
-+ printk(KERN_ALERT " rx_event=%08x tx_req_prod=%08x\n",
-+ netif->rx.sring->rsp_event, netif->tx.sring->req_prod);
-+ printk(KERN_ALERT " tx_resp_prod=%08x, tx_event=%08x)\n",
-+ netif->tx.sring->rsp_prod, netif->tx.sring->rsp_event);
-+ i++;
++ if (PCI_SLOT(devfn) < PCI_SLOT_MAX) {
++ spin_lock_irqsave(&vpci_dev->lock, flags);
++
++ list_for_each_entry(entry,
++ &vpci_dev->dev_list[PCI_SLOT(devfn)],
++ list) {
++ if (PCI_FUNC(entry->dev->devfn) == PCI_FUNC(devfn)) {
++ dev = entry->dev;
++ break;
++ }
++ }
++
++ spin_unlock_irqrestore(&vpci_dev->lock, flags);
+ }
++ return dev;
++}
+
-+ spin_unlock_irq(&net_schedule_list_lock);
-+ printk(KERN_ALERT " ** End of netif_schedule_list **\n");
++static inline int match_slot(struct pci_dev *l, struct pci_dev *r)
++{
++ if (pci_domain_nr(l->bus) == pci_domain_nr(r->bus)
++ && l->bus == r->bus && PCI_SLOT(l->devfn) == PCI_SLOT(r->devfn))
++ return 1;
+
-+ return IRQ_HANDLED;
++ return 0;
+}
-+#endif
+
-+static int __init netback_init(void)
++int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev,
++ int devid, publish_pci_dev_cb publish_cb)
+{
-+ int i;
-+ struct page *page;
++ int err = 0, slot, func;
++ struct pci_dev_entry *t, *dev_entry;
++ struct vpci_dev_data *vpci_dev = pdev->pci_dev_data;
++ unsigned long flags;
+
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++ if ((dev->class >> 24) == PCI_BASE_CLASS_BRIDGE) {
++ err = -EFAULT;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Can't export bridges on the virtual PCI bus");
++ goto out;
++ }
+
-+ /* We can increase reservation by this much in net_rx_action(). */
-+ balloon_update_driver_allowance(NET_RX_RING_SIZE);
++ dev_entry = kmalloc(sizeof(*dev_entry), GFP_KERNEL);
++ if (!dev_entry) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error adding entry to virtual PCI bus");
++ goto out;
++ }
+
-+ skb_queue_head_init(&rx_queue);
-+ skb_queue_head_init(&tx_queue);
++ dev_entry->dev = dev;
+
-+ init_timer(&net_timer);
-+ net_timer.data = 0;
-+ net_timer.function = net_alarm;
++ spin_lock_irqsave(&vpci_dev->lock, flags);
+
-+ mmap_pages = alloc_empty_pages_and_pagevec(MAX_PENDING_REQS);
-+ if (mmap_pages == NULL) {
-+ printk("%s: out of memory\n", __FUNCTION__);
-+ return -ENOMEM;
-+ }
++ /* Keep multi-function devices together on the virtual PCI bus */
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ if (!list_empty(&vpci_dev->dev_list[slot])) {
++ t = list_entry(list_first(&vpci_dev->dev_list[slot]),
++ struct pci_dev_entry, list);
+
-+ for (i = 0; i < MAX_PENDING_REQS; i++) {
-+ page = mmap_pages[i];
-+ SetPageForeign(page, netif_page_release);
-+ netif_page_index(page) = i;
++ if (match_slot(dev, t->dev)) {
++ pr_info("pciback: vpci: %s: "
++ "assign to virtual slot %d func %d\n",
++ pci_name(dev), slot,
++ PCI_FUNC(dev->devfn));
++ list_add_tail(&dev_entry->list,
++ &vpci_dev->dev_list[slot]);
++ func = PCI_FUNC(dev->devfn);
++ goto unlock;
++ }
++ }
+ }
+
-+ pending_cons = 0;
-+ pending_prod = MAX_PENDING_REQS;
-+ for (i = 0; i < MAX_PENDING_REQS; i++)
-+ pending_ring[i] = i;
++ /* Assign to a new slot on the virtual PCI bus */
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ if (list_empty(&vpci_dev->dev_list[slot])) {
++ printk(KERN_INFO
++ "pciback: vpci: %s: assign to virtual slot %d\n",
++ pci_name(dev), slot);
++ list_add_tail(&dev_entry->list,
++ &vpci_dev->dev_list[slot]);
++ func = PCI_FUNC(dev->devfn);
++ goto unlock;
++ }
++ }
+
-+ spin_lock_init(&net_schedule_list_lock);
-+ INIT_LIST_HEAD(&net_schedule_list);
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "No more space on root virtual PCI bus");
+
-+ netif_xenbus_init();
++ unlock:
++ spin_unlock_irqrestore(&vpci_dev->lock, flags);
+
-+#ifdef NETBE_DEBUG_INTERRUPT
-+ (void)bind_virq_to_irqhandler(VIRQ_DEBUG,
-+ 0,
-+ netif_be_dbg,
-+ SA_SHIRQ,
-+ "net-be-dbg",
-+ &netif_be_dbg);
-+#endif
++ /* Publish this device. */
++ if(!err)
++ err = publish_cb(pdev, 0, 0, PCI_DEVFN(slot, func), devid);
+
-+ return 0;
++ out:
++ return err;
+}
+
-+module_init(netback_init);
++void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
++{
++ int slot;
++ struct vpci_dev_data *vpci_dev = pdev->pci_dev_data;
++ struct pci_dev *found_dev = NULL;
++ unsigned long flags;
+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/netback/xenbus.c tmp-linux-2.6-xen.patch/drivers/xen/netback/xenbus.c
---- pristine-linux-2.6.18.2/drivers/xen/netback/xenbus.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/netback/xenbus.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,448 @@
-+/* Xenbus code for netif backend
-+ Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
-+ Copyright (C) 2005 XenSource Ltd
++ spin_lock_irqsave(&vpci_dev->lock, flags);
+
-+ This program is free software; you can redistribute it and/or modify
-+ it under the terms of the GNU General Public License as published by
-+ the Free Software Foundation; either version 2 of the License, or
-+ (at your option) any later version.
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ struct pci_dev_entry *e, *tmp;
++ list_for_each_entry_safe(e, tmp, &vpci_dev->dev_list[slot],
++ list) {
++ if (e->dev == dev) {
++ list_del(&e->list);
++ found_dev = e->dev;
++ kfree(e);
++ goto out;
++ }
++ }
++ }
+
-+ This program is distributed in the hope that it will be useful,
-+ but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ GNU General Public License for more details.
++ out:
++ spin_unlock_irqrestore(&vpci_dev->lock, flags);
+
-+ You should have received a copy of the GNU General Public License
-+ along with this program; if not, write to the Free Software
-+ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
-+*/
++ if (found_dev)
++ pcistub_put_pci_dev(found_dev);
++}
+
-+#include <stdarg.h>
-+#include <linux/module.h>
-+#include <xen/xenbus.h>
-+#include "common.h"
++int pciback_init_devices(struct pciback_device *pdev)
++{
++ int slot;
++ struct vpci_dev_data *vpci_dev;
+
-+#if 0
-+#undef DPRINTK
-+#define DPRINTK(fmt, args...) \
-+ printk("netback/xenbus (%s:%d) " fmt ".\n", __FUNCTION__, __LINE__, ##args)
-+#endif
++ vpci_dev = kmalloc(sizeof(*vpci_dev), GFP_KERNEL);
++ if (!vpci_dev)
++ return -ENOMEM;
+
-+struct backend_info {
-+ struct xenbus_device *dev;
-+ netif_t *netif;
-+ enum xenbus_state frontend_state;
-+};
++ spin_lock_init(&vpci_dev->lock);
+
-+static int connect_rings(struct backend_info *);
-+static void connect(struct backend_info *);
-+static void backend_create_netif(struct backend_info *be);
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ INIT_LIST_HEAD(&vpci_dev->dev_list[slot]);
++ }
+
-+static int netback_remove(struct xenbus_device *dev)
-+{
-+ struct backend_info *be = dev->dev.driver_data;
++ pdev->pci_dev_data = vpci_dev;
+
-+ if (be->netif) {
-+ netif_disconnect(be->netif);
-+ be->netif = NULL;
-+ }
-+ kfree(be);
-+ dev->dev.driver_data = NULL;
+ return 0;
+}
+
++int pciback_publish_pci_roots(struct pciback_device *pdev,
++ publish_pci_root_cb publish_cb)
++{
++ /* The Virtual PCI bus has only one root */
++ return publish_cb(pdev, 0, 0);
++}
+
-+/**
-+ * Entry point to this code when a new device is created. Allocate the basic
-+ * structures and switch to InitWait.
-+ */
-+static int netback_probe(struct xenbus_device *dev,
-+ const struct xenbus_device_id *id)
++void pciback_release_devices(struct pciback_device *pdev)
+{
-+ const char *message;
-+ struct xenbus_transaction xbt;
-+ int err;
-+ struct backend_info *be = kzalloc(sizeof(struct backend_info),
-+ GFP_KERNEL);
-+ if (!be) {
-+ xenbus_dev_fatal(dev, -ENOMEM,
-+ "allocating backend structure");
-+ return -ENOMEM;
++ int slot;
++ struct vpci_dev_data *vpci_dev = pdev->pci_dev_data;
++
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ struct pci_dev_entry *e, *tmp;
++ list_for_each_entry_safe(e, tmp, &vpci_dev->dev_list[slot],
++ list) {
++ list_del(&e->list);
++ pcistub_put_pci_dev(e->dev);
++ kfree(e);
++ }
+ }
+
-+ be->dev = dev;
-+ dev->dev.driver_data = be;
++ kfree(vpci_dev);
++ pdev->pci_dev_data = NULL;
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pciback/xenbus.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pciback/xenbus.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,704 @@
++/*
++ * PCI Backend Xenbus Setup - handles setup with frontend and xend
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/list.h>
++#include <linux/vmalloc.h>
++#include <xen/xenbus.h>
++#include <xen/evtchn.h>
++#include "pciback.h"
+
-+ do {
-+ err = xenbus_transaction_start(&xbt);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "starting transaction");
-+ goto fail;
-+ }
++#define INVALID_EVTCHN_IRQ (-1)
+
-+ err = xenbus_printf(xbt, dev->nodename, "feature-sg", "%d", 1);
-+ if (err) {
-+ message = "writing feature-sg";
-+ goto abort_transaction;
-+ }
++static struct pciback_device *alloc_pdev(struct xenbus_device *xdev)
++{
++ struct pciback_device *pdev;
+
-+ err = xenbus_printf(xbt, dev->nodename, "feature-gso-tcpv4",
-+ "%d", 1);
-+ if (err) {
-+ message = "writing feature-gso-tcpv4";
-+ goto abort_transaction;
-+ }
++ pdev = kzalloc(sizeof(struct pciback_device), GFP_KERNEL);
++ if (pdev == NULL)
++ goto out;
++ dev_dbg(&xdev->dev, "allocated pdev @ 0x%p\n", pdev);
+
-+ /* We support rx-copy path. */
-+ err = xenbus_printf(xbt, dev->nodename,
-+ "feature-rx-copy", "%d", 1);
-+ if (err) {
-+ message = "writing feature-rx-copy";
-+ goto abort_transaction;
-+ }
++ pdev->xdev = xdev;
++ xdev->dev.driver_data = pdev;
+
-+ /*
-+ * We don't support rx-flip path (except old guests who don't
-+ * grok this feature flag).
-+ */
-+ err = xenbus_printf(xbt, dev->nodename,
-+ "feature-rx-flip", "%d", 0);
-+ if (err) {
-+ message = "writing feature-rx-flip";
-+ goto abort_transaction;
-+ }
++ spin_lock_init(&pdev->dev_lock);
+
-+ err = xenbus_transaction_end(xbt, 0);
-+ } while (err == -EAGAIN);
++ pdev->sh_area = NULL;
++ pdev->sh_info = NULL;
++ pdev->evtchn_irq = INVALID_EVTCHN_IRQ;
++ pdev->be_watching = 0;
+
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "completing transaction");
-+ goto fail;
++ INIT_WORK(&pdev->op_work, pciback_do_op, pdev);
++
++ if (pciback_init_devices(pdev)) {
++ kfree(pdev);
++ pdev = NULL;
+ }
++ out:
++ return pdev;
++}
+
-+ err = xenbus_switch_state(dev, XenbusStateInitWait);
-+ if (err)
-+ goto fail;
++static void pciback_disconnect(struct pciback_device *pdev)
++{
++ spin_lock(&pdev->dev_lock);
+
-+ /* This kicks hotplug scripts, so do it immediately. */
-+ backend_create_netif(be);
++ /* Ensure the guest can't trigger our handler before removing devices */
++ if (pdev->evtchn_irq != INVALID_EVTCHN_IRQ) {
++ unbind_from_irqhandler(pdev->evtchn_irq, pdev);
++ pdev->evtchn_irq = INVALID_EVTCHN_IRQ;
++ }
+
-+ return 0;
++ /* If the driver domain started an op, make sure we complete it or
++ * delete it before releasing the shared memory */
++ cancel_delayed_work(&pdev->op_work);
++ flush_scheduled_work();
+
-+abort_transaction:
-+ xenbus_transaction_end(xbt, 1);
-+ xenbus_dev_fatal(dev, err, "%s", message);
-+fail:
-+ DPRINTK("failed");
-+ netback_remove(dev);
-+ return err;
-+}
++ if (pdev->sh_info != NULL) {
++ xenbus_unmap_ring_vfree(pdev->xdev, pdev->sh_area);
++ pdev->sh_info = NULL;
++ }
+
++ spin_unlock(&pdev->dev_lock);
++}
+
-+/**
-+ * Handle the creation of the hotplug script environment. We add the script
-+ * and vif variables to the environment, for the benefit of the vif-* hotplug
-+ * scripts.
-+ */
-+static int netback_uevent(struct xenbus_device *xdev, char **envp,
-+ int num_envp, char *buffer, int buffer_size)
++static void free_pdev(struct pciback_device *pdev)
+{
-+ struct backend_info *be = xdev->dev.driver_data;
-+ netif_t *netif = be->netif;
-+ int i = 0, length = 0;
-+ char *val;
-+
-+ DPRINTK("netback_uevent");
++ if (pdev->be_watching)
++ unregister_xenbus_watch(&pdev->be_watch);
+
-+ val = xenbus_read(XBT_NIL, xdev->nodename, "script", NULL);
-+ if (IS_ERR(val)) {
-+ int err = PTR_ERR(val);
-+ xenbus_dev_fatal(xdev, err, "reading script");
-+ return err;
-+ }
-+ else {
-+ add_uevent_var(envp, num_envp, &i, buffer, buffer_size,
-+ &length, "script=%s", val);
-+ kfree(val);
-+ }
++ pciback_disconnect(pdev);
+
-+ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
-+ "vif=%s", netif->dev->name);
++ pciback_release_devices(pdev);
+
-+ envp[i] = NULL;
++ pdev->xdev->dev.driver_data = NULL;
++ pdev->xdev = NULL;
+
-+ return 0;
++ kfree(pdev);
+}
+
-+
-+static void backend_create_netif(struct backend_info *be)
++static int pciback_do_attach(struct pciback_device *pdev, int gnt_ref,
++ int remote_evtchn)
+{
-+ int err;
-+ long handle;
-+ struct xenbus_device *dev = be->dev;
++ int err = 0;
++ struct vm_struct *area;
+
-+ if (be->netif != NULL)
-+ return;
++ dev_dbg(&pdev->xdev->dev,
++ "Attaching to frontend resources - gnt_ref=%d evtchn=%d\n",
++ gnt_ref, remote_evtchn);
+
-+ err = xenbus_scanf(XBT_NIL, dev->nodename, "handle", "%li", &handle);
-+ if (err != 1) {
-+ xenbus_dev_fatal(dev, err, "reading handle");
-+ return;
++ area = xenbus_map_ring_valloc(pdev->xdev, gnt_ref);
++ if (IS_ERR(area)) {
++ err = PTR_ERR(area);
++ goto out;
+ }
++ pdev->sh_area = area;
++ pdev->sh_info = area->addr;
+
-+ be->netif = netif_alloc(dev->otherend_id, handle);
-+ if (IS_ERR(be->netif)) {
-+ err = PTR_ERR(be->netif);
-+ be->netif = NULL;
-+ xenbus_dev_fatal(dev, err, "creating interface");
-+ return;
++ err = bind_interdomain_evtchn_to_irqhandler(
++ pdev->xdev->otherend_id, remote_evtchn, pciback_handle_event,
++ SA_SAMPLE_RANDOM, "pciback", pdev);
++ if (err < 0) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error binding event channel to IRQ");
++ goto out;
+ }
++ pdev->evtchn_irq = err;
++ err = 0;
+
-+ kobject_uevent(&dev->dev.kobj, KOBJ_ONLINE);
++ dev_dbg(&pdev->xdev->dev, "Attached!\n");
++ out:
++ return err;
+}
+
-+
-+/**
-+ * Callback received when the frontend's state changes.
-+ */
-+static void frontend_changed(struct xenbus_device *dev,
-+ enum xenbus_state frontend_state)
++static int pciback_attach(struct pciback_device *pdev)
+{
-+ struct backend_info *be = dev->dev.driver_data;
-+
-+ DPRINTK("%s", xenbus_strstate(frontend_state));
-+
-+ be->frontend_state = frontend_state;
++ int err = 0;
++ int gnt_ref, remote_evtchn;
++ char *magic = NULL;
+
-+ switch (frontend_state) {
-+ case XenbusStateInitialising:
-+ if (dev->state == XenbusStateClosed) {
-+ printk(KERN_INFO "%s: %s: prepare for reconnect\n",
-+ __FUNCTION__, dev->nodename);
-+ if (be->netif) {
-+ netif_disconnect(be->netif);
-+ be->netif = NULL;
-+ }
-+ xenbus_switch_state(dev, XenbusStateInitWait);
-+ }
-+ break;
++ spin_lock(&pdev->dev_lock);
+
-+ case XenbusStateInitialised:
-+ break;
++ /* Make sure we only do this setup once */
++ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
++ XenbusStateInitialised)
++ goto out;
+
-+ case XenbusStateConnected:
-+ backend_create_netif(be);
-+ if (be->netif)
-+ connect(be);
-+ break;
++ /* Wait for frontend to state that it has published the configuration */
++ if (xenbus_read_driver_state(pdev->xdev->otherend) !=
++ XenbusStateInitialised)
++ goto out;
+
-+ case XenbusStateClosing:
-+ xenbus_switch_state(dev, XenbusStateClosing);
-+ break;
++ dev_dbg(&pdev->xdev->dev, "Reading frontend config\n");
+
-+ case XenbusStateClosed:
-+ xenbus_switch_state(dev, XenbusStateClosed);
-+ if (xenbus_dev_is_online(dev))
-+ break;
-+ /* fall through if not online */
-+ case XenbusStateUnknown:
-+ if (be->netif != NULL)
-+ kobject_uevent(&dev->dev.kobj, KOBJ_OFFLINE);
-+ device_unregister(&dev->dev);
-+ break;
++ err = xenbus_gather(XBT_NIL, pdev->xdev->otherend,
++ "pci-op-ref", "%u", &gnt_ref,
++ "event-channel", "%u", &remote_evtchn,
++ "magic", NULL, &magic, NULL);
++ if (err) {
++ /* If configuration didn't get read correctly, wait longer */
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading configuration from frontend");
++ goto out;
++ }
+
-+ default:
-+ xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
-+ frontend_state);
-+ break;
++ if (magic == NULL || strcmp(magic, XEN_PCI_MAGIC) != 0) {
++ xenbus_dev_fatal(pdev->xdev, -EFAULT,
++ "version mismatch (%s/%s) with pcifront - "
++ "halting pciback",
++ magic, XEN_PCI_MAGIC);
++ goto out;
+ }
-+}
+
++ err = pciback_do_attach(pdev, gnt_ref, remote_evtchn);
++ if (err)
++ goto out;
+
-+static void xen_net_read_rate(struct xenbus_device *dev,
-+ unsigned long *bytes, unsigned long *usec)
-+{
-+ char *s, *e;
-+ unsigned long b, u;
-+ char *ratestr;
++ dev_dbg(&pdev->xdev->dev, "Connecting...\n");
+
-+ /* Default to unlimited bandwidth. */
-+ *bytes = ~0UL;
-+ *usec = 0;
++ err = xenbus_switch_state(pdev->xdev, XenbusStateConnected);
++ if (err)
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error switching to connected state!");
+
-+ ratestr = xenbus_read(XBT_NIL, dev->nodename, "rate", NULL);
-+ if (IS_ERR(ratestr))
-+ return;
++ dev_dbg(&pdev->xdev->dev, "Connected? %d\n", err);
++ out:
++ spin_unlock(&pdev->dev_lock);
+
-+ s = ratestr;
-+ b = simple_strtoul(s, &e, 10);
-+ if ((s == e) || (*e != ','))
-+ goto fail;
++ if (magic)
++ kfree(magic);
+
-+ s = e + 1;
-+ u = simple_strtoul(s, &e, 10);
-+ if ((s == e) || (*e != '\0'))
-+ goto fail;
++ return err;
++}
+
-+ *bytes = b;
-+ *usec = u;
++static int pciback_publish_pci_dev(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn, unsigned int devid)
++{
++ int err;
++ int len;
++ char str[64];
+
-+ kfree(ratestr);
-+ return;
++ len = snprintf(str, sizeof(str), "vdev-%d", devid);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
+
-+ fail:
-+ WPRINTK("Failed to parse network rate limit. Traffic unlimited.\n");
-+ kfree(ratestr);
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, str,
++ "%04x:%02x:%02x.%02x", domain, bus,
++ PCI_SLOT(devfn), PCI_FUNC(devfn));
++
++ out:
++ return err;
+}
+
-+static int xen_net_read_mac(struct xenbus_device *dev, u8 mac[])
++static int pciback_export_device(struct pciback_device *pdev,
++ int domain, int bus, int slot, int func,
++ int devid)
+{
-+ char *s, *e, *macstr;
-+ int i;
++ struct pci_dev *dev;
++ int err = 0;
+
-+ macstr = s = xenbus_read(XBT_NIL, dev->nodename, "mac", NULL);
-+ if (IS_ERR(macstr))
-+ return PTR_ERR(macstr);
++ dev_dbg(&pdev->xdev->dev, "exporting dom %x bus %x slot %x func %x\n",
++ domain, bus, slot, func);
+
-+ for (i = 0; i < ETH_ALEN; i++) {
-+ mac[i] = simple_strtoul(s, &e, 16);
-+ if ((s == e) || (*e != ((i == ETH_ALEN-1) ? '\0' : ':'))) {
-+ kfree(macstr);
-+ return -ENOENT;
-+ }
-+ s = e+1;
++ dev = pcistub_get_pci_dev_by_slot(pdev, domain, bus, slot, func);
++ if (!dev) {
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Couldn't locate PCI device "
++ "(%04x:%02x:%02x.%01x)! "
++ "perhaps already in-use?",
++ domain, bus, slot, func);
++ goto out;
+ }
+
-+ kfree(macstr);
-+ return 0;
++ err = pciback_add_pci_dev(pdev, dev, devid, pciback_publish_pci_dev);
++ if (err)
++ goto out;
++
++ /* TODO: It'd be nice to export a bridge and have all of its children
++ * get exported with it. This may be best done in xend (which will
++ * have to calculate resource usage anyway) but we probably want to
++ * put something in here to ensure that if a bridge gets given to a
++ * driver domain, that all devices under that bridge are not given
++ * to other driver domains (as he who controls the bridge can disable
++ * it and stop the other devices from working).
++ */
++ out:
++ return err;
+}
+
-+static void connect(struct backend_info *be)
++static int pciback_remove_device(struct pciback_device *pdev,
++ int domain, int bus, int slot, int func)
+{
-+ int err;
-+ struct xenbus_device *dev = be->dev;
++ int err = 0;
++ struct pci_dev *dev;
+
-+ err = connect_rings(be);
-+ if (err)
-+ return;
++ dev_dbg(&pdev->xdev->dev, "removing dom %x bus %x slot %x func %x\n",
++ domain, bus, slot, func);
+
-+ err = xen_net_read_mac(dev, be->netif->fe_dev_addr);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "parsing %s/mac", dev->nodename);
-+ return;
++ dev = pciback_get_pci_dev(pdev, domain, bus, PCI_DEVFN(slot, func));
++ if (!dev) {
++ err = -EINVAL;
++ dev_dbg(&pdev->xdev->dev, "Couldn't locate PCI device "
++ "(%04x:%02x:%02x.%01x)! not owned by this domain\n",
++ domain, bus, slot, func);
++ goto out;
+ }
+
-+ xen_net_read_rate(dev, &be->netif->credit_bytes,
-+ &be->netif->credit_usec);
-+ be->netif->remaining_credit = be->netif->credit_bytes;
-+
-+ xenbus_switch_state(dev, XenbusStateConnected);
-+
-+ netif_wake_queue(be->netif->dev);
++ pciback_release_pci_dev(pdev, dev);
++
++ out:
++ return err;
+}
+
-+
-+static int connect_rings(struct backend_info *be)
++static int pciback_publish_pci_root(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus)
+{
-+ struct xenbus_device *dev = be->dev;
-+ unsigned long tx_ring_ref, rx_ring_ref;
-+ unsigned int evtchn, rx_copy;
-+ int err;
-+ int val;
-+
-+ DPRINTK("");
++ unsigned int d, b;
++ int i, root_num, len, err;
++ char str[64];
+
-+ err = xenbus_gather(XBT_NIL, dev->otherend,
-+ "tx-ring-ref", "%lu", &tx_ring_ref,
-+ "rx-ring-ref", "%lu", &rx_ring_ref,
-+ "event-channel", "%u", &evtchn, NULL);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err,
-+ "reading %s/ring-ref and event-channel",
-+ dev->otherend);
-+ return err;
-+ }
++ dev_dbg(&pdev->xdev->dev, "Publishing pci roots\n");
+
-+ err = xenbus_scanf(XBT_NIL, dev->otherend, "request-rx-copy", "%u",
-+ &rx_copy);
-+ if (err == -ENOENT) {
-+ err = 0;
-+ rx_copy = 0;
-+ }
-+ if (err < 0) {
-+ xenbus_dev_fatal(dev, err, "reading %s/request-rx-copy",
-+ dev->otherend);
-+ return err;
-+ }
-+ be->netif->copying_receiver = !!rx_copy;
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
++ "root_num", "%d", &root_num);
++ if (err == 0 || err == -ENOENT)
++ root_num = 0;
++ else if (err < 0)
++ goto out;
+
-+ if (be->netif->dev->tx_queue_len != 0) {
-+ if (xenbus_scanf(XBT_NIL, dev->otherend,
-+ "feature-rx-notify", "%d", &val) < 0)
-+ val = 0;
-+ if (val)
-+ be->netif->can_queue = 1;
-+ else
-+ /* Must be non-zero for pfifo_fast to work. */
-+ be->netif->dev->tx_queue_len = 1;
-+ }
++ /* Verify that we haven't already published this pci root */
++ for (i = 0; i < root_num; i++) {
++ len = snprintf(str, sizeof(str), "root-%d", i);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
+
-+ if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-sg", "%d", &val) < 0)
-+ val = 0;
-+ if (val) {
-+ be->netif->features |= NETIF_F_SG;
-+ be->netif->dev->features |= NETIF_F_SG;
-+ }
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
++ str, "%x:%x", &d, &b);
++ if (err < 0)
++ goto out;
++ if (err != 2) {
++ err = -EINVAL;
++ goto out;
++ }
+
-+ if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-gso-tcpv4", "%d",
-+ &val) < 0)
-+ val = 0;
-+ if (val) {
-+ be->netif->features |= NETIF_F_TSO;
-+ be->netif->dev->features |= NETIF_F_TSO;
++ if (d == domain && b == bus) {
++ err = 0;
++ goto out;
++ }
+ }
+
-+ if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-no-csum-offload",
-+ "%d", &val) < 0)
-+ val = 0;
-+ if (val) {
-+ be->netif->features &= ~NETIF_F_IP_CSUM;
-+ be->netif->dev->features &= ~NETIF_F_IP_CSUM;
++ len = snprintf(str, sizeof(str), "root-%d", root_num);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
+ }
+
-+ /* Map the shared frame, irq etc. */
-+ err = netif_map(be->netif, tx_ring_ref, rx_ring_ref, evtchn);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err,
-+ "mapping shared-frames %lu/%lu port %u",
-+ tx_ring_ref, rx_ring_ref, evtchn);
-+ return err;
-+ }
-+ return 0;
-+}
++ dev_dbg(&pdev->xdev->dev, "writing root %d at %04x:%02x\n",
++ root_num, domain, bus);
+
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, str,
++ "%04x:%02x", domain, bus);
++ if (err)
++ goto out;
+
-+/* ** Driver Registration ** */
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename,
++ "root_num", "%d", (root_num + 1));
+
++ out:
++ return err;
++}
+
-+static struct xenbus_device_id netback_ids[] = {
-+ { "vif" },
-+ { "" }
-+};
++static int pciback_reconfigure(struct pciback_device *pdev)
++{
++ int err = 0;
++ int num_devs;
++ int domain, bus, slot, func;
++ int substate;
++ int i, len;
++ char state_str[64];
++ char dev_str[64];
+
++ spin_lock(&pdev->dev_lock);
+
-+static struct xenbus_driver netback = {
-+ .name = "vif",
-+ .owner = THIS_MODULE,
-+ .ids = netback_ids,
-+ .probe = netback_probe,
-+ .remove = netback_remove,
-+ .uevent = netback_uevent,
-+ .otherend_changed = frontend_changed,
-+};
++ dev_dbg(&pdev->xdev->dev, "Reconfiguring device ...\n");
+
++ /* Make sure we only reconfigure once */
++ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
++ XenbusStateReconfiguring)
++ goto out;
+
-+void netif_xenbus_init(void)
-+{
-+ xenbus_register_backend(&netback);
-+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/netfront/Makefile tmp-linux-2.6-xen.patch/drivers/xen/netfront/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/netfront/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/netfront/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,4 @@
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename, "num_devs", "%d",
++ &num_devs);
++ if (err != 1) {
++ if (err >= 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading number of devices");
++ goto out;
++ }
+
-+obj-$(CONFIG_XEN_NETDEV_FRONTEND) := xennet.o
++ for (i = 0; i < num_devs; i++) {
++ len = snprintf(state_str, sizeof(state_str), "state-%d", i);
++ if (unlikely(len >= (sizeof(state_str) - 1))) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "String overflow while reading "
++ "configuration");
++ goto out;
++ }
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename, state_str,
++ "%d", &substate);
++ if (err != 1)
++ substate = XenbusStateUnknown;
++
++ switch (substate) {
++ case XenbusStateInitialising:
++ dev_dbg(&pdev->xdev->dev, "Attaching dev-%d ...\n", i);
+
-+xennet-objs := netfront.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/netfront/netfront.c tmp-linux-2.6-xen.patch/drivers/xen/netfront/netfront.c
---- pristine-linux-2.6.18.2/drivers/xen/netfront/netfront.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/netfront/netfront.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,2172 @@
-+/******************************************************************************
-+ * Virtual network driver for conversing with remote driver backends.
-+ *
-+ * Copyright (c) 2002-2005, K A Fraser
-+ * Copyright (c) 2005, XenSource Ltd
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ len = snprintf(dev_str, sizeof(dev_str), "dev-%d", i);
++ if (unlikely(len >= (sizeof(dev_str) - 1))) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "String overflow while "
++ "reading configuration");
++ goto out;
++ }
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
++ dev_str, "%x:%x:%x.%x",
++ &domain, &bus, &slot, &func);
++ if (err < 0) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading device "
++ "configuration");
++ goto out;
++ }
++ if (err != 4) {
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error parsing pci device "
++ "configuration");
++ goto out;
++ }
++
++ err = pciback_export_device(pdev, domain, bus, slot,
++ func, i);
++ if (err)
++ goto out;
+
-+#include <linux/module.h>
-+#include <linux/version.h>
-+#include <linux/kernel.h>
-+#include <linux/sched.h>
-+#include <linux/slab.h>
-+#include <linux/string.h>
-+#include <linux/errno.h>
-+#include <linux/netdevice.h>
-+#include <linux/inetdevice.h>
-+#include <linux/etherdevice.h>
-+#include <linux/skbuff.h>
-+#include <linux/init.h>
-+#include <linux/bitops.h>
-+#include <linux/ethtool.h>
-+#include <linux/in.h>
-+#include <linux/if_ether.h>
-+#include <linux/io.h>
-+#include <linux/moduleparam.h>
-+#include <net/sock.h>
-+#include <net/pkt_sched.h>
-+#include <net/arp.h>
-+#include <net/route.h>
-+#include <asm/uaccess.h>
-+#include <xen/evtchn.h>
-+#include <xen/xenbus.h>
-+#include <xen/interface/io/netif.h>
-+#include <xen/interface/memory.h>
-+#include <xen/balloon.h>
-+#include <asm/page.h>
-+#include <asm/maddr.h>
-+#include <asm/uaccess.h>
-+#include <xen/interface/grant_table.h>
-+#include <xen/gnttab.h>
++ /* Publish pci roots. */
++ err = pciback_publish_pci_roots(pdev, pciback_publish_pci_root);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error while publish PCI root"
++ "buses for frontend");
++ goto out;
++ }
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
-+#endif
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename,
++ state_str, "%d",
++ XenbusStateInitialised);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error switching substate of "
++ "dev-%d\n", i);
++ goto out;
++ }
++ break;
+
-+struct netfront_cb {
-+ struct page *page;
-+ unsigned offset;
-+};
++ case XenbusStateClosing:
++ dev_dbg(&pdev->xdev->dev, "Detaching dev-%d ...\n", i);
+
-+#define NETFRONT_SKB_CB(skb) ((struct netfront_cb *)((skb)->cb))
++ len = snprintf(dev_str, sizeof(dev_str), "vdev-%d", i);
++ if (unlikely(len >= (sizeof(dev_str) - 1))) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "String overflow while "
++ "reading configuration");
++ goto out;
++ }
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
++ dev_str, "%x:%x:%x.%x",
++ &domain, &bus, &slot, &func);
++ if (err < 0) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading device "
++ "configuration");
++ goto out;
++ }
++ if (err != 4) {
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error parsing pci device "
++ "configuration");
++ goto out;
++ }
+
-+/*
-+ * Mutually-exclusive module options to select receive data path:
-+ * rx_copy : Packets are copied by network backend into local memory
-+ * rx_flip : Page containing packet data is transferred to our ownership
-+ * For fully-virtualised guests there is no option - copying must be used.
-+ * For paravirtualised guests, flipping is the default.
-+ */
-+#ifdef CONFIG_XEN
-+static int MODPARM_rx_copy = 0;
-+module_param_named(rx_copy, MODPARM_rx_copy, bool, 0);
-+MODULE_PARM_DESC(rx_copy, "Copy packets from network card (rather than flip)");
-+static int MODPARM_rx_flip = 0;
-+module_param_named(rx_flip, MODPARM_rx_flip, bool, 0);
-+MODULE_PARM_DESC(rx_flip, "Flip packets from network card (rather than copy)");
-+#else
-+static const int MODPARM_rx_copy = 1;
-+static const int MODPARM_rx_flip = 0;
-+#endif
++ err = pciback_remove_device(pdev, domain, bus, slot,
++ func);
++ if(err)
++ goto out;
+
-+#define RX_COPY_THRESHOLD 256
++ /* TODO: If at some point we implement support for pci
++ * root hot-remove on pcifront side, we'll need to
++ * remove unnecessary xenstore nodes of pci roots here.
++ */
+
-+/* If we don't have GSO, fake things up so that we never try to use it. */
-+#if defined(NETIF_F_GSO)
-+#define HAVE_GSO 1
-+#define HAVE_TSO 1 /* TSO is a subset of GSO */
-+static inline void dev_disable_gso_features(struct net_device *dev)
-+{
-+ /* Turn off all GSO bits except ROBUST. */
-+ dev->features &= (1 << NETIF_F_GSO_SHIFT) - 1;
-+ dev->features |= NETIF_F_GSO_ROBUST;
-+}
-+#elif defined(NETIF_F_TSO)
-+#define HAVE_TSO 1
++ break;
+
-+/* Some older kernels cannot cope with incorrect checksums,
-+ * particularly in netfilter. I'm not sure there is 100% correlation
-+ * with the presence of NETIF_F_TSO but it appears to be a good first
-+ * approximiation.
-+ */
-+#define HAVE_NO_CSUM_OFFLOAD 1
++ default:
++ break;
++ }
++ }
+
-+#define gso_size tso_size
-+#define gso_segs tso_segs
-+static inline void dev_disable_gso_features(struct net_device *dev)
-+{
-+ /* Turn off all TSO bits. */
-+ dev->features &= ~NETIF_F_TSO;
-+}
-+static inline int skb_is_gso(const struct sk_buff *skb)
-+{
-+ return skb_shinfo(skb)->tso_size;
-+}
-+static inline int skb_gso_ok(struct sk_buff *skb, int features)
-+{
-+ return (features & NETIF_F_TSO);
++ err = xenbus_switch_state(pdev->xdev, XenbusStateReconfigured);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error switching to reconfigured state!");
++ goto out;
++ }
++
++ out:
++ spin_unlock(&pdev->dev_lock);
++
++ return 0;
+}
+
-+static inline int netif_needs_gso(struct net_device *dev, struct sk_buff *skb)
++static void pciback_frontend_changed(struct xenbus_device *xdev,
++ enum xenbus_state fe_state)
+{
-+ return skb_is_gso(skb) &&
-+ (!skb_gso_ok(skb, dev->features) ||
-+ unlikely(skb->ip_summed != CHECKSUM_HW));
-+}
-+#else
-+#define netif_needs_gso(dev, skb) 0
-+#define dev_disable_gso_features(dev) ((void)0)
-+#endif
++ struct pciback_device *pdev = xdev->dev.driver_data;
+
-+#define GRANT_INVALID_REF 0
++ dev_dbg(&xdev->dev, "fe state changed %d\n", fe_state);
+
-+#define NET_TX_RING_SIZE __RING_SIZE((struct netif_tx_sring *)0, PAGE_SIZE)
-+#define NET_RX_RING_SIZE __RING_SIZE((struct netif_rx_sring *)0, PAGE_SIZE)
++ switch (fe_state) {
++ case XenbusStateInitialised:
++ pciback_attach(pdev);
++ break;
+
-+struct netfront_info {
-+ struct list_head list;
-+ struct net_device *netdev;
++ case XenbusStateReconfiguring:
++ pciback_reconfigure(pdev);
++ break;
+
-+ struct net_device_stats stats;
++ case XenbusStateConnected:
++ /* pcifront switched its state from reconfiguring to connected.
++ * Then switch to connected state.
++ */
++ xenbus_switch_state(xdev, XenbusStateConnected);
++ break;
+
-+ struct netif_tx_front_ring tx;
-+ struct netif_rx_front_ring rx;
++ case XenbusStateClosing:
++ pciback_disconnect(pdev);
++ xenbus_switch_state(xdev, XenbusStateClosing);
++ break;
+
-+ spinlock_t tx_lock;
-+ spinlock_t rx_lock;
++ case XenbusStateClosed:
++ pciback_disconnect(pdev);
++ xenbus_switch_state(xdev, XenbusStateClosed);
++ if (xenbus_dev_is_online(xdev))
++ break;
++ /* fall through if not online */
++ case XenbusStateUnknown:
++ dev_dbg(&xdev->dev, "frontend is gone! unregister device\n");
++ device_unregister(&xdev->dev);
++ break;
+
-+ unsigned int irq;
-+ unsigned int copying_receiver;
-+ unsigned int carrier;
++ default:
++ break;
++ }
++}
+
-+ /* Receive-ring batched refills. */
-+#define RX_MIN_TARGET 8
-+#define RX_DFL_MIN_TARGET 64
-+#define RX_MAX_TARGET min_t(int, NET_RX_RING_SIZE, 256)
-+ unsigned rx_min_target, rx_max_target, rx_target;
-+ struct sk_buff_head rx_batch;
++static int pciback_setup_backend(struct pciback_device *pdev)
++{
++ /* Get configuration from xend (if available now) */
++ int domain, bus, slot, func;
++ int err = 0;
++ int i, num_devs;
++ char dev_str[64];
++ char state_str[64];
+
-+ struct timer_list rx_refill_timer;
++ spin_lock(&pdev->dev_lock);
+
-+ /*
-+ * {tx,rx}_skbs store outstanding skbuffs. The first entry in tx_skbs
-+ * is an index into a chain of free entries.
++ /* It's possible we could get the call to setup twice, so make sure
++ * we're not already connected.
+ */
-+ struct sk_buff *tx_skbs[NET_TX_RING_SIZE+1];
-+ struct sk_buff *rx_skbs[NET_RX_RING_SIZE];
++ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
++ XenbusStateInitWait)
++ goto out;
+
-+#define TX_MAX_TARGET min_t(int, NET_RX_RING_SIZE, 256)
-+ grant_ref_t gref_tx_head;
-+ grant_ref_t grant_tx_ref[NET_TX_RING_SIZE + 1];
-+ grant_ref_t gref_rx_head;
-+ grant_ref_t grant_rx_ref[NET_RX_RING_SIZE];
++ dev_dbg(&pdev->xdev->dev, "getting be setup\n");
+
-+ struct xenbus_device *xbdev;
-+ int tx_ring_ref;
-+ int rx_ring_ref;
-+ u8 mac[ETH_ALEN];
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename, "num_devs", "%d",
++ &num_devs);
++ if (err != 1) {
++ if (err >= 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading number of devices");
++ goto out;
++ }
+
-+ unsigned long rx_pfn_array[NET_RX_RING_SIZE];
-+ struct multicall_entry rx_mcl[NET_RX_RING_SIZE+1];
-+ struct mmu_update rx_mmu[NET_RX_RING_SIZE];
-+};
++ for (i = 0; i < num_devs; i++) {
++ int l = snprintf(dev_str, sizeof(dev_str), "dev-%d", i);
++ if (unlikely(l >= (sizeof(dev_str) - 1))) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "String overflow while reading "
++ "configuration");
++ goto out;
++ }
+
-+struct netfront_rx_info {
-+ struct netif_rx_response rx;
-+ struct netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX - 1];
-+};
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename, dev_str,
++ "%x:%x:%x.%x", &domain, &bus, &slot, &func);
++ if (err < 0) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading device configuration");
++ goto out;
++ }
++ if (err != 4) {
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error parsing pci device "
++ "configuration");
++ goto out;
++ }
+
-+/*
-+ * Implement our own carrier flag: the network stack's version causes delays
-+ * when the carrier is re-enabled (in particular, dev_activate() may not
-+ * immediately be called, which can cause packet loss).
-+ */
-+#define netfront_carrier_on(netif) ((netif)->carrier = 1)
-+#define netfront_carrier_off(netif) ((netif)->carrier = 0)
-+#define netfront_carrier_ok(netif) ((netif)->carrier)
++ err = pciback_export_device(pdev, domain, bus, slot, func, i);
++ if (err)
++ goto out;
+
-+/*
-+ * Access macros for acquiring freeing slots in tx_skbs[].
-+ */
++ /* Switch substate of this device. */
++ l = snprintf(state_str, sizeof(state_str), "state-%d", i);
++ if (unlikely(l >= (sizeof(state_str) - 1))) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "String overflow while reading "
++ "configuration");
++ goto out;
++ }
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, state_str,
++ "%d", XenbusStateInitialised);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err, "Error switching "
++ "substate of dev-%d\n", i);
++ goto out;
++ }
++ }
+
-+static inline void add_id_to_freelist(struct sk_buff **list, unsigned short id)
++ err = pciback_publish_pci_roots(pdev, pciback_publish_pci_root);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error while publish PCI root buses "
++ "for frontend");
++ goto out;
++ }
++
++ err = xenbus_switch_state(pdev->xdev, XenbusStateInitialised);
++ if (err)
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error switching to initialised state!");
++
++ out:
++ spin_unlock(&pdev->dev_lock);
++
++ if (!err)
++ /* see if pcifront is already configured (if not, we'll wait) */
++ pciback_attach(pdev);
++
++ return err;
++}
++
++static void pciback_be_watch(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
+{
-+ list[id] = list[0];
-+ list[0] = (void *)(unsigned long)id;
++ struct pciback_device *pdev =
++ container_of(watch, struct pciback_device, be_watch);
++
++ switch (xenbus_read_driver_state(pdev->xdev->nodename)) {
++ case XenbusStateInitWait:
++ pciback_setup_backend(pdev);
++ break;
++
++ default:
++ break;
++ }
+}
+
-+static inline unsigned short get_id_from_freelist(struct sk_buff **list)
++static int pciback_xenbus_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
+{
-+ unsigned int id = (unsigned int)(unsigned long)list[0];
-+ list[0] = list[id];
-+ return id;
++ int err = 0;
++ struct pciback_device *pdev = alloc_pdev(dev);
++
++ if (pdev == NULL) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(dev, err,
++ "Error allocating pciback_device struct");
++ goto out;
++ }
++
++ /* wait for xend to configure us */
++ err = xenbus_switch_state(dev, XenbusStateInitWait);
++ if (err)
++ goto out;
++
++ /* watch the backend node for backend configuration information */
++ err = xenbus_watch_path(dev, dev->nodename, &pdev->be_watch,
++ pciback_be_watch);
++ if (err)
++ goto out;
++ pdev->be_watching = 1;
++
++ /* We need to force a call to our callback here in case
++ * xend already configured us!
++ */
++ pciback_be_watch(&pdev->be_watch, NULL, 0);
++
++ out:
++ return err;
+}
+
-+static inline int xennet_rxidx(RING_IDX idx)
++static int pciback_xenbus_remove(struct xenbus_device *dev)
+{
-+ return idx & (NET_RX_RING_SIZE - 1);
++ struct pciback_device *pdev = dev->dev.driver_data;
++
++ if (pdev != NULL)
++ free_pdev(pdev);
++
++ return 0;
+}
+
-+static inline struct sk_buff *xennet_get_rx_skb(struct netfront_info *np,
-+ RING_IDX ri)
++static const struct xenbus_device_id xenpci_ids[] = {
++ {"pci"},
++ {{0}},
++};
++
++static struct xenbus_driver xenbus_pciback_driver = {
++ .name = "pciback",
++ .owner = THIS_MODULE,
++ .ids = xenpci_ids,
++ .probe = pciback_xenbus_probe,
++ .remove = pciback_xenbus_remove,
++ .otherend_changed = pciback_frontend_changed,
++};
++
++int __init pciback_xenbus_register(void)
+{
-+ int i = xennet_rxidx(ri);
-+ struct sk_buff *skb = np->rx_skbs[i];
-+ np->rx_skbs[i] = NULL;
-+ return skb;
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ return xenbus_register_backend(&xenbus_pciback_driver);
+}
+
-+static inline grant_ref_t xennet_get_rx_ref(struct netfront_info *np,
-+ RING_IDX ri)
++void __exit pciback_xenbus_unregister(void)
+{
-+ int i = xennet_rxidx(ri);
-+ grant_ref_t ref = np->grant_rx_ref[i];
-+ np->grant_rx_ref[i] = GRANT_INVALID_REF;
-+ return ref;
++ xenbus_unregister_driver(&xenbus_pciback_driver);
+}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pcifront/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pcifront/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,7 @@
++obj-y += pcifront.o
+
-+#define DPRINTK(fmt, args...) \
-+ pr_debug("netfront (%s:%d) " fmt, \
-+ __FUNCTION__, __LINE__, ##args)
-+#define IPRINTK(fmt, args...) \
-+ printk(KERN_INFO "netfront: " fmt, ##args)
-+#define WPRINTK(fmt, args...) \
-+ printk(KERN_WARNING "netfront: " fmt, ##args)
++pcifront-y := pci_op.o xenbus.o pci.o
+
-+static int setup_device(struct xenbus_device *, struct netfront_info *);
-+static struct net_device *create_netdev(struct xenbus_device *);
++ifeq ($(CONFIG_XEN_PCIDEV_FE_DEBUG),y)
++EXTRA_CFLAGS += -DDEBUG
++endif
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pcifront/pci.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pcifront/pci.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,46 @@
++/*
++ * PCI Frontend Operations - ensure only one PCI frontend runs at a time
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include "pcifront.h"
+
-+static void end_access(int, void *);
-+static void netif_disconnect_backend(struct netfront_info *);
++DEFINE_SPINLOCK(pcifront_dev_lock);
++static struct pcifront_device *pcifront_dev = NULL;
+
-+static int network_connect(struct net_device *);
-+static void network_tx_buf_gc(struct net_device *);
-+static void network_alloc_rx_buffers(struct net_device *);
-+static int send_fake_arp(struct net_device *);
++int pcifront_connect(struct pcifront_device *pdev)
++{
++ int err = 0;
+
-+static irqreturn_t netif_int(int irq, void *dev_id, struct pt_regs *ptregs);
++ spin_lock(&pcifront_dev_lock);
+
-+#ifdef CONFIG_SYSFS
-+static int xennet_sysfs_addif(struct net_device *netdev);
-+static void xennet_sysfs_delif(struct net_device *netdev);
-+#else /* !CONFIG_SYSFS */
-+#define xennet_sysfs_addif(dev) (0)
-+#define xennet_sysfs_delif(dev) do { } while(0)
-+#endif
++ if (!pcifront_dev) {
++ dev_info(&pdev->xdev->dev, "Installing PCI frontend\n");
++ pcifront_dev = pdev;
++ }
++ else {
++ dev_err(&pdev->xdev->dev, "PCI frontend already installed!\n");
++ err = -EEXIST;
++ }
+
-+static inline int xennet_can_sg(struct net_device *dev)
-+{
-+ return dev->features & NETIF_F_SG;
++ spin_unlock(&pcifront_dev_lock);
++
++ return err;
+}
+
-+/**
-+ * Entry point to this code when a new device is created. Allocate the basic
-+ * structures and the ring buffers for communication with the backend, and
-+ * inform the backend of the appropriate details for those.
-+ */
-+static int __devinit netfront_probe(struct xenbus_device *dev,
-+ const struct xenbus_device_id *id)
++void pcifront_disconnect(struct pcifront_device *pdev)
+{
-+ int err;
-+ struct net_device *netdev;
-+ struct netfront_info *info;
++ spin_lock(&pcifront_dev_lock);
+
-+ netdev = create_netdev(dev);
-+ if (IS_ERR(netdev)) {
-+ err = PTR_ERR(netdev);
-+ xenbus_dev_fatal(dev, err, "creating netdev");
-+ return err;
++ if (pdev == pcifront_dev) {
++ dev_info(&pdev->xdev->dev,
++ "Disconnecting PCI Frontend Buses\n");
++ pcifront_dev = NULL;
+ }
+
-+ info = netdev_priv(netdev);
-+ dev->dev.driver_data = info;
++ spin_unlock(&pcifront_dev_lock);
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pcifront/pci_op.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pcifront/pci_op.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,551 @@
++/*
++ * PCI Frontend Operations - Communicates with frontend
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <linux/version.h>
++#include <linux/init.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include <linux/time.h>
++#include <xen/evtchn.h>
++#include "pcifront.h"
+
-+ err = register_netdev(info->netdev);
-+ if (err) {
-+ printk(KERN_WARNING "%s: register_netdev err=%d\n",
-+ __FUNCTION__, err);
-+ goto fail;
-+ }
++static int verbose_request = 0;
++module_param(verbose_request, int, 0644);
+
-+ err = xennet_sysfs_addif(info->netdev);
-+ if (err) {
-+ unregister_netdev(info->netdev);
-+ printk(KERN_WARNING "%s: add sysfs failed err=%d\n",
-+ __FUNCTION__, err);
-+ goto fail;
++#ifdef __ia64__
++static void pcifront_init_sd(struct pcifront_sd *sd,
++ unsigned int domain, unsigned int bus,
++ struct pcifront_device *pdev)
++{
++ int err, i, j, k, len, root_num, res_count;
++ struct acpi_resource res;
++ unsigned int d, b, byte;
++ unsigned long magic;
++ char str[64], tmp[3];
++ unsigned char *buf, *bufp;
++ u8 *ptr;
++
++ memset(sd, 0, sizeof(*sd));
++
++ sd->segment = domain;
++ sd->node = -1; /* Revisit for NUMA */
++ sd->platform_data = pdev;
++
++ /* Look for resources for this controller in xenbus. */
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, "root_num",
++ "%d", &root_num);
++ if (err != 1)
++ return;
++
++ for (i = 0; i < root_num; i++) {
++ len = snprintf(str, sizeof(str), "root-%d", i);
++ if (unlikely(len >= (sizeof(str) - 1)))
++ return;
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
++ str, "%x:%x", &d, &b);
++ if (err != 2)
++ return;
++
++ if (d == domain && b == bus)
++ break;
+ }
+
-+ return 0;
++ if (i == root_num)
++ return;
+
-+ fail:
-+ free_netdev(netdev);
-+ dev->dev.driver_data = NULL;
-+ return err;
-+}
++ len = snprintf(str, sizeof(str), "root-resource-magic");
+
-+static int __devexit netfront_remove(struct xenbus_device *dev)
-+{
-+ struct netfront_info *info = dev->dev.driver_data;
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
++ str, "%lx", &magic);
+
-+ DPRINTK("%s\n", dev->nodename);
++ if (err != 1)
++ return; /* No resources, nothing to do */
+
-+ netif_disconnect_backend(info);
++ if (magic != (sizeof(res) * 2) + 1) {
++ printk(KERN_WARNING "pcifront: resource magic mismatch\n");
++ return;
++ }
+
-+ del_timer_sync(&info->rx_refill_timer);
++ len = snprintf(str, sizeof(str), "root-%d-resources", i);
++ if (unlikely(len >= (sizeof(str) - 1)))
++ return;
+
-+ xennet_sysfs_delif(info->netdev);
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
++ str, "%d", &res_count);
+
-+ unregister_netdev(info->netdev);
++ if (err != 1)
++ return; /* No resources, nothing to do */
+
-+ free_netdev(info->netdev);
++ sd->window = kzalloc(sizeof(*sd->window) * res_count, GFP_KERNEL);
++ if (!sd->window)
++ return;
+
-+ return 0;
-+}
++ /* magic is also the size of the byte stream in xenbus */
++ buf = kmalloc(magic, GFP_KERNEL);
++ if (!buf) {
++ kfree(sd->window);
++ sd->window = NULL;
++ return;
++ }
+
-+/**
-+ * We are reconnecting to the backend, due to a suspend/resume, or a backend
-+ * driver restart. We tear down our netif structure and recreate it, but
-+ * leave the device-layer structures intact so that this is transparent to the
-+ * rest of the kernel.
-+ */
-+static int netfront_resume(struct xenbus_device *dev)
-+{
-+ struct netfront_info *info = dev->dev.driver_data;
++ /* Read the resources out of xenbus */
++ for (j = 0; j < res_count; j++) {
++ memset(&res, 0, sizeof(res));
++ memset(buf, 0, magic);
+
-+ DPRINTK("%s\n", dev->nodename);
++ len = snprintf(str, sizeof(str), "root-%d-resource-%d", i, j);
++ if (unlikely(len >= (sizeof(str) - 1)))
++ return;
+
-+ netif_disconnect_backend(info);
-+ return 0;
-+}
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
++ "%s", buf);
++ if (err != 1) {
++ printk(KERN_WARNING "pcifront: error reading "
++ "resource %d on bus %04x:%02x\n",
++ j, domain, bus);
++ continue;
++ }
+
-+static int xen_net_read_mac(struct xenbus_device *dev, u8 mac[])
-+{
-+ char *s, *e, *macstr;
-+ int i;
++ bufp = buf;
++ ptr = (u8 *)&res;
++ memset(tmp, 0, sizeof(tmp));
+
-+ macstr = s = xenbus_read(XBT_NIL, dev->nodename, "mac", NULL);
-+ if (IS_ERR(macstr))
-+ return PTR_ERR(macstr);
++ /* Copy ASCII byte stream into structure */
++ for (k = 0; k < magic - 1; k += 2) {
++ memcpy(tmp, bufp, 2);
++ bufp += 2;
+
-+ for (i = 0; i < ETH_ALEN; i++) {
-+ mac[i] = simple_strtoul(s, &e, 16);
-+ if ((s == e) || (*e != ((i == ETH_ALEN-1) ? '\0' : ':'))) {
-+ kfree(macstr);
-+ return -ENOENT;
++ sscanf(tmp, "%02x", &byte);
++ *ptr = byte;
++ ptr++;
+ }
-+ s = e+1;
-+ }
+
-+ kfree(macstr);
-+ return 0;
++ xen_add_resource(sd, domain, bus, &res);
++ sd->windows++;
++ }
++ kfree(buf);
+}
++#endif
+
-+/* Common code used when first setting up, and when resuming. */
-+static int talk_to_backend(struct xenbus_device *dev,
-+ struct netfront_info *info)
++static int errno_to_pcibios_err(int errno)
+{
-+ const char *message;
-+ struct xenbus_transaction xbt;
-+ int err;
++ switch (errno) {
++ case XEN_PCI_ERR_success:
++ return PCIBIOS_SUCCESSFUL;
+
-+ err = xen_net_read_mac(dev, info->mac);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "parsing %s/mac", dev->nodename);
-+ goto out;
-+ }
++ case XEN_PCI_ERR_dev_not_found:
++ return PCIBIOS_DEVICE_NOT_FOUND;
+
-+ /* Create shared ring, alloc event channel. */
-+ err = setup_device(dev, info);
-+ if (err)
-+ goto out;
++ case XEN_PCI_ERR_invalid_offset:
++ case XEN_PCI_ERR_op_failed:
++ return PCIBIOS_BAD_REGISTER_NUMBER;
+
-+again:
-+ err = xenbus_transaction_start(&xbt);
-+ if (err) {
-+ xenbus_dev_fatal(dev, err, "starting transaction");
-+ goto destroy_ring;
-+ }
++ case XEN_PCI_ERR_not_implemented:
++ return PCIBIOS_FUNC_NOT_SUPPORTED;
+
-+ err = xenbus_printf(xbt, dev->nodename, "tx-ring-ref","%u",
-+ info->tx_ring_ref);
-+ if (err) {
-+ message = "writing tx ring-ref";
-+ goto abort_transaction;
-+ }
-+ err = xenbus_printf(xbt, dev->nodename, "rx-ring-ref","%u",
-+ info->rx_ring_ref);
-+ if (err) {
-+ message = "writing rx ring-ref";
-+ goto abort_transaction;
-+ }
-+ err = xenbus_printf(xbt, dev->nodename,
-+ "event-channel", "%u",
-+ irq_to_evtchn_port(info->irq));
-+ if (err) {
-+ message = "writing event-channel";
-+ goto abort_transaction;
++ case XEN_PCI_ERR_access_denied:
++ return PCIBIOS_SET_FAILED;
+ }
++ return errno;
++}
+
-+ err = xenbus_printf(xbt, dev->nodename, "request-rx-copy", "%u",
-+ info->copying_receiver);
-+ if (err) {
-+ message = "writing request-rx-copy";
-+ goto abort_transaction;
-+ }
++static int do_pci_op(struct pcifront_device *pdev, struct xen_pci_op *op)
++{
++ int err = 0;
++ struct xen_pci_op *active_op = &pdev->sh_info->op;
++ unsigned long irq_flags;
++ evtchn_port_t port = pdev->evtchn;
++ s64 ns, ns_timeout;
++ struct timeval tv;
+
-+ err = xenbus_printf(xbt, dev->nodename, "feature-rx-notify", "%d", 1);
-+ if (err) {
-+ message = "writing feature-rx-notify";
-+ goto abort_transaction;
-+ }
++ spin_lock_irqsave(&pdev->sh_info_lock, irq_flags);
+
-+#ifdef HAVE_NO_CSUM_OFFLOAD
-+ err = xenbus_printf(xbt, dev->nodename, "feature-no-csum-offload", "%d", 1);
-+ if (err) {
-+ message = "writing feature-no-csum-offload";
-+ goto abort_transaction;
-+ }
-+#endif
++ memcpy(active_op, op, sizeof(struct xen_pci_op));
+
-+ err = xenbus_printf(xbt, dev->nodename, "feature-sg", "%d", 1);
-+ if (err) {
-+ message = "writing feature-sg";
-+ goto abort_transaction;
-+ }
++ /* Go */
++ wmb();
++ set_bit(_XEN_PCIF_active, (unsigned long *)&pdev->sh_info->flags);
++ notify_remote_via_evtchn(port);
+
-+#ifdef HAVE_TSO
-+ err = xenbus_printf(xbt, dev->nodename, "feature-gso-tcpv4", "%d", 1);
-+ if (err) {
-+ message = "writing feature-gso-tcpv4";
-+ goto abort_transaction;
-+ }
-+#endif
++ /*
++ * We set a poll timeout of 3 seconds but give up on return after
++ * 2 seconds. It is better to time out too late rather than too early
++ * (in the latter case we end up continually re-executing poll() with a
++ * timeout in the past). 1s difference gives plenty of slack for error.
++ */
++ do_gettimeofday(&tv);
++ ns_timeout = timeval_to_ns(&tv) + 2 * (s64)NSEC_PER_SEC;
+
-+ err = xenbus_transaction_end(xbt, 0);
-+ if (err) {
-+ if (err == -EAGAIN)
-+ goto again;
-+ xenbus_dev_fatal(dev, err, "completing transaction");
-+ goto destroy_ring;
++ clear_evtchn(port);
++
++ while (test_bit(_XEN_PCIF_active,
++ (unsigned long *)&pdev->sh_info->flags)) {
++ if (HYPERVISOR_poll(&port, 1, jiffies + 3*HZ))
++ BUG();
++ clear_evtchn(port);
++ do_gettimeofday(&tv);
++ ns = timeval_to_ns(&tv);
++ if (ns > ns_timeout) {
++ dev_err(&pdev->xdev->dev,
++ "pciback not responding!!!\n");
++ clear_bit(_XEN_PCIF_active,
++ (unsigned long *)&pdev->sh_info->flags);
++ err = XEN_PCI_ERR_dev_not_found;
++ goto out;
++ }
+ }
+
-+ return 0;
++ memcpy(op, active_op, sizeof(struct xen_pci_op));
+
-+ abort_transaction:
-+ xenbus_transaction_end(xbt, 1);
-+ xenbus_dev_fatal(dev, err, "%s", message);
-+ destroy_ring:
-+ netif_disconnect_backend(info);
-+ out:
++ err = op->err;
++ out:
++ spin_unlock_irqrestore(&pdev->sh_info_lock, irq_flags);
+ return err;
+}
+
-+static int setup_device(struct xenbus_device *dev, struct netfront_info *info)
++/* Access to this function is spinlocked in drivers/pci/access.c */
++static int pcifront_bus_read(struct pci_bus *bus, unsigned int devfn,
++ int where, int size, u32 * val)
+{
-+ struct netif_tx_sring *txs;
-+ struct netif_rx_sring *rxs;
-+ int err;
-+ struct net_device *netdev = info->netdev;
-+
-+ info->tx_ring_ref = GRANT_INVALID_REF;
-+ info->rx_ring_ref = GRANT_INVALID_REF;
-+ info->rx.sring = NULL;
-+ info->tx.sring = NULL;
-+ info->irq = 0;
++ int err = 0;
++ struct xen_pci_op op = {
++ .cmd = XEN_PCI_OP_conf_read,
++ .domain = pci_domain_nr(bus),
++ .bus = bus->number,
++ .devfn = devfn,
++ .offset = where,
++ .size = size,
++ };
++ struct pcifront_sd *sd = bus->sysdata;
++ struct pcifront_device *pdev = pcifront_get_pdev(sd);
+
-+ txs = (struct netif_tx_sring *)get_zeroed_page(GFP_KERNEL);
-+ if (!txs) {
-+ err = -ENOMEM;
-+ xenbus_dev_fatal(dev, err, "allocating tx ring page");
-+ goto fail;
-+ }
-+ SHARED_RING_INIT(txs);
-+ FRONT_RING_INIT(&info->tx, txs, PAGE_SIZE);
++ if (verbose_request)
++ dev_info(&pdev->xdev->dev,
++ "read dev=%04x:%02x:%02x.%01x - offset %x size %d\n",
++ pci_domain_nr(bus), bus->number, PCI_SLOT(devfn),
++ PCI_FUNC(devfn), where, size);
+
-+ err = xenbus_grant_ring(dev, virt_to_mfn(txs));
-+ if (err < 0) {
-+ free_page((unsigned long)txs);
-+ goto fail;
-+ }
-+ info->tx_ring_ref = err;
++ err = do_pci_op(pdev, &op);
+
-+ rxs = (struct netif_rx_sring *)get_zeroed_page(GFP_KERNEL);
-+ if (!rxs) {
-+ err = -ENOMEM;
-+ xenbus_dev_fatal(dev, err, "allocating rx ring page");
-+ goto fail;
-+ }
-+ SHARED_RING_INIT(rxs);
-+ FRONT_RING_INIT(&info->rx, rxs, PAGE_SIZE);
++ if (likely(!err)) {
++ if (verbose_request)
++ dev_info(&pdev->xdev->dev, "read got back value %x\n",
++ op.value);
+
-+ err = xenbus_grant_ring(dev, virt_to_mfn(rxs));
-+ if (err < 0) {
-+ free_page((unsigned long)rxs);
-+ goto fail;
++ *val = op.value;
++ } else if (err == -ENODEV) {
++ /* No device here, pretend that it just returned 0 */
++ err = 0;
++ *val = 0;
+ }
-+ info->rx_ring_ref = err;
+
-+ memcpy(netdev->dev_addr, info->mac, ETH_ALEN);
++ return errno_to_pcibios_err(err);
++}
+
-+ err = bind_listening_port_to_irqhandler(
-+ dev->otherend_id, netif_int, SA_SAMPLE_RANDOM, netdev->name,
-+ netdev);
-+ if (err < 0)
-+ goto fail;
-+ info->irq = err;
++/* Access to this function is spinlocked in drivers/pci/access.c */
++static int pcifront_bus_write(struct pci_bus *bus, unsigned int devfn,
++ int where, int size, u32 val)
++{
++ struct xen_pci_op op = {
++ .cmd = XEN_PCI_OP_conf_write,
++ .domain = pci_domain_nr(bus),
++ .bus = bus->number,
++ .devfn = devfn,
++ .offset = where,
++ .size = size,
++ .value = val,
++ };
++ struct pcifront_sd *sd = bus->sysdata;
++ struct pcifront_device *pdev = pcifront_get_pdev(sd);
+
-+ return 0;
++ if (verbose_request)
++ dev_info(&pdev->xdev->dev,
++ "write dev=%04x:%02x:%02x.%01x - "
++ "offset %x size %d val %x\n",
++ pci_domain_nr(bus), bus->number,
++ PCI_SLOT(devfn), PCI_FUNC(devfn), where, size, val);
+
-+ fail:
-+ return err;
++ return errno_to_pcibios_err(do_pci_op(pdev, &op));
+}
+
-+/**
-+ * Callback received when the backend's state changes.
-+ */
-+static void backend_changed(struct xenbus_device *dev,
-+ enum xenbus_state backend_state)
++struct pci_ops pcifront_bus_ops = {
++ .read = pcifront_bus_read,
++ .write = pcifront_bus_write,
++};
++
++#ifdef CONFIG_PCI_MSI
++int pci_frontend_enable_msix(struct pci_dev *dev,
++ struct msix_entry *entries,
++ int nvec)
+{
-+ struct netfront_info *np = dev->dev.driver_data;
-+ struct net_device *netdev = np->netdev;
++ int err;
++ int i;
++ struct xen_pci_op op = {
++ .cmd = XEN_PCI_OP_enable_msix,
++ .domain = pci_domain_nr(dev->bus),
++ .bus = dev->bus->number,
++ .devfn = dev->devfn,
++ .value = nvec,
++ };
++ struct pcifront_sd *sd = dev->bus->sysdata;
++ struct pcifront_device *pdev = pcifront_get_pdev(sd);
+
-+ DPRINTK("%s\n", xenbus_strstate(backend_state));
++ if (nvec > SH_INFO_MAX_VEC) {
++ printk("too much vector for pci frontend%x\n", nvec);
++ return -EINVAL;
++ }
+
-+ switch (backend_state) {
-+ case XenbusStateInitialising:
-+ case XenbusStateInitialised:
-+ case XenbusStateConnected:
-+ case XenbusStateUnknown:
-+ case XenbusStateClosed:
-+ break;
++ for (i = 0; i < nvec; i++) {
++ op.msix_entries[i].entry = entries[i].entry;
++ op.msix_entries[i].vector = entries[i].vector;
++ }
+
-+ case XenbusStateInitWait:
-+ if (dev->state != XenbusStateInitialising)
-+ break;
-+ if (network_connect(netdev) != 0)
-+ break;
-+ xenbus_switch_state(dev, XenbusStateConnected);
-+ (void)send_fake_arp(netdev);
-+ break;
++ err = do_pci_op(pdev, &op);
+
-+ case XenbusStateClosing:
-+ xenbus_frontend_closed(dev);
-+ break;
++ if (!err) {
++ if (!op.value) {
++ /* we get the result */
++ for ( i = 0; i < nvec; i++)
++ entries[i].vector = op.msix_entries[i].vector;
++ return 0;
++ }
++ else {
++ printk("enable msix get value %x\n", op.value);
++ return op.value;
++ }
++ }
++ else {
++ printk("enable msix get err %x\n", err);
++ return err;
+ }
+}
+
-+/** Send a packet on a net device to encourage switches to learn the
-+ * MAC. We send a fake ARP request.
-+ *
-+ * @param dev device
-+ * @return 0 on success, error code otherwise
-+ */
-+static int send_fake_arp(struct net_device *dev)
++void pci_frontend_disable_msix(struct pci_dev* dev)
+{
-+ struct sk_buff *skb;
-+ u32 src_ip, dst_ip;
-+
-+ dst_ip = INADDR_BROADCAST;
-+ src_ip = inet_select_addr(dev, dst_ip, RT_SCOPE_LINK);
-+
-+ /* No IP? Then nothing to do. */
-+ if (src_ip == 0)
-+ return 0;
++ int err;
++ struct xen_pci_op op = {
++ .cmd = XEN_PCI_OP_disable_msix,
++ .domain = pci_domain_nr(dev->bus),
++ .bus = dev->bus->number,
++ .devfn = dev->devfn,
++ };
++ struct pcifront_sd *sd = dev->bus->sysdata;
++ struct pcifront_device *pdev = pcifront_get_pdev(sd);
+
-+ skb = arp_create(ARPOP_REPLY, ETH_P_ARP,
-+ dst_ip, dev, src_ip,
-+ /*dst_hw*/ NULL, /*src_hw*/ NULL,
-+ /*target_hw*/ dev->dev_addr);
-+ if (skb == NULL)
-+ return -ENOMEM;
++ err = do_pci_op(pdev, &op);
+
-+ return dev_queue_xmit(skb);
++ /* What should do for error ? */
++ if (err)
++ printk("pci_disable_msix get err %x\n", err);
+}
+
-+static inline int netfront_tx_slot_available(struct netfront_info *np)
++int pci_frontend_enable_msi(struct pci_dev *dev)
+{
-+ return ((np->tx.req_prod_pvt - np->tx.rsp_cons) <
-+ (TX_MAX_TARGET - MAX_SKB_FRAGS - 2));
++ int err;
++ struct xen_pci_op op = {
++ .cmd = XEN_PCI_OP_enable_msi,
++ .domain = pci_domain_nr(dev->bus),
++ .bus = dev->bus->number,
++ .devfn = dev->devfn,
++ };
++ struct pcifront_sd *sd = dev->bus->sysdata;
++ struct pcifront_device *pdev = pcifront_get_pdev(sd);
++
++ err = do_pci_op(pdev, &op);
++ if (likely(!err)) {
++ dev->irq = op.value;
++ }
++ else {
++ printk("pci frontend enable msi failed for dev %x:%x \n",
++ op.bus, op.devfn);
++ err = -EINVAL;
++ }
++ return err;
+}
+
-+static inline void network_maybe_wake_tx(struct net_device *dev)
++void pci_frontend_disable_msi(struct pci_dev* dev)
+{
-+ struct netfront_info *np = netdev_priv(dev);
++ int err;
++ struct xen_pci_op op = {
++ .cmd = XEN_PCI_OP_disable_msi,
++ .domain = pci_domain_nr(dev->bus),
++ .bus = dev->bus->number,
++ .devfn = dev->devfn,
++ };
++ struct pcifront_sd *sd = dev->bus->sysdata;
++ struct pcifront_device *pdev = pcifront_get_pdev(sd);
+
-+ if (unlikely(netif_queue_stopped(dev)) &&
-+ netfront_tx_slot_available(np) &&
-+ likely(netif_running(dev)))
-+ netif_wake_queue(dev);
++ err = do_pci_op(pdev, &op);
++ if (err == XEN_PCI_ERR_dev_not_found) {
++ /* XXX No response from backend, what shall we do? */
++ printk("get no response from backend for disable MSI\n");
++ return;
++ }
++ if (likely(!err))
++ dev->irq = op.value;
++ else
++ /* how can pciback notify us fail? */
++ printk("get fake response frombackend \n");
+}
++#endif /* CONFIG_PCI_MSI */
+
-+static int network_open(struct net_device *dev)
++/* Claim resources for the PCI frontend as-is, backend won't allow changes */
++static void pcifront_claim_resource(struct pci_dev *dev, void *data)
+{
-+ struct netfront_info *np = netdev_priv(dev);
++ struct pcifront_device *pdev = data;
++ int i;
++ struct resource *r;
+
-+ memset(&np->stats, 0, sizeof(np->stats));
++ for (i = 0; i < PCI_NUM_RESOURCES; i++) {
++ r = &dev->resource[i];
+
-+ spin_lock_bh(&np->rx_lock);
-+ if (netfront_carrier_ok(np)) {
-+ network_alloc_rx_buffers(dev);
-+ np->rx.sring->rsp_event = np->rx.rsp_cons + 1;
-+ if (RING_HAS_UNCONSUMED_RESPONSES(&np->rx))
-+ netif_rx_schedule(dev);
++ if (!r->parent && r->start && r->flags) {
++ dev_dbg(&pdev->xdev->dev, "claiming resource %s/%d\n",
++ pci_name(dev), i);
++ pci_claim_resource(dev, i);
++ }
+ }
-+ spin_unlock_bh(&np->rx_lock);
-+
-+ network_maybe_wake_tx(dev);
-+
-+ return 0;
+}
+
-+static void network_tx_buf_gc(struct net_device *dev)
++int __devinit pcifront_scan_root(struct pcifront_device *pdev,
++ unsigned int domain, unsigned int bus)
+{
-+ RING_IDX cons, prod;
-+ unsigned short id;
-+ struct netfront_info *np = netdev_priv(dev);
-+ struct sk_buff *skb;
++ struct pci_bus *b;
++ struct pcifront_sd *sd = NULL;
++ struct pci_bus_entry *bus_entry = NULL;
++ int err = 0;
+
-+ BUG_ON(!netfront_carrier_ok(np));
++#ifndef CONFIG_PCI_DOMAINS
++ if (domain != 0) {
++ dev_err(&pdev->xdev->dev,
++ "PCI Root in non-zero PCI Domain! domain=%d\n", domain);
++ dev_err(&pdev->xdev->dev,
++ "Please compile with CONFIG_PCI_DOMAINS\n");
++ err = -EINVAL;
++ goto err_out;
++ }
++#endif
+
-+ do {
-+ prod = np->tx.sring->rsp_prod;
-+ rmb(); /* Ensure we see responses up to 'rp'. */
++ dev_info(&pdev->xdev->dev, "Creating PCI Frontend Bus %04x:%02x\n",
++ domain, bus);
+
-+ for (cons = np->tx.rsp_cons; cons != prod; cons++) {
-+ struct netif_tx_response *txrsp;
++ bus_entry = kmalloc(sizeof(*bus_entry), GFP_KERNEL);
++ sd = kmalloc(sizeof(*sd), GFP_KERNEL);
++ if (!bus_entry || !sd) {
++ err = -ENOMEM;
++ goto err_out;
++ }
++ pcifront_init_sd(sd, domain, bus, pdev);
+
-+ txrsp = RING_GET_RESPONSE(&np->tx, cons);
-+ if (txrsp->status == NETIF_RSP_NULL)
-+ continue;
++ b = pci_scan_bus_parented(&pdev->xdev->dev, bus,
++ &pcifront_bus_ops, sd);
++ if (!b) {
++ dev_err(&pdev->xdev->dev,
++ "Error creating PCI Frontend Bus!\n");
++ err = -ENOMEM;
++ goto err_out;
++ }
+
-+ id = txrsp->id;
-+ skb = np->tx_skbs[id];
-+ if (unlikely(gnttab_query_foreign_access(
-+ np->grant_tx_ref[id]) != 0)) {
-+ printk(KERN_ALERT "network_tx_buf_gc: warning "
-+ "-- grant still in use by backend "
-+ "domain.\n");
-+ BUG();
-+ }
-+ gnttab_end_foreign_access_ref(
-+ np->grant_tx_ref[id], GNTMAP_readonly);
-+ gnttab_release_grant_reference(
-+ &np->gref_tx_head, np->grant_tx_ref[id]);
-+ np->grant_tx_ref[id] = GRANT_INVALID_REF;
-+ add_id_to_freelist(np->tx_skbs, id);
-+ dev_kfree_skb_irq(skb);
-+ }
++ pcifront_setup_root_resources(b, sd);
++ bus_entry->bus = b;
+
-+ np->tx.rsp_cons = prod;
++ list_add(&bus_entry->list, &pdev->root_buses);
+
-+ /*
-+ * Set a new event, then check for race with update of tx_cons.
-+ * Note that it is essential to schedule a callback, no matter
-+ * how few buffers are pending. Even if there is space in the
-+ * transmit ring, higher layers may be blocked because too much
-+ * data is outstanding: in such cases notification from Xen is
-+ * likely to be the only kick that we'll get.
-+ */
-+ np->tx.sring->rsp_event =
-+ prod + ((np->tx.sring->req_prod - prod) >> 1) + 1;
-+ mb();
-+ } while ((cons == prod) && (prod != np->tx.sring->rsp_prod));
++ /* Claim resources before going "live" with our devices */
++ pci_walk_bus(b, pcifront_claim_resource, pdev);
+
-+ network_maybe_wake_tx(dev);
-+}
++ pci_bus_add_devices(b);
+
-+static void rx_refill_timeout(unsigned long data)
-+{
-+ struct net_device *dev = (struct net_device *)data;
-+ netif_rx_schedule(dev);
++ return 0;
++
++ err_out:
++ kfree(bus_entry);
++ kfree(sd);
++
++ return err;
+}
+
-+static void network_alloc_rx_buffers(struct net_device *dev)
++int __devinit pcifront_rescan_root(struct pcifront_device *pdev,
++ unsigned int domain, unsigned int bus)
+{
-+ unsigned short id;
-+ struct netfront_info *np = netdev_priv(dev);
-+ struct sk_buff *skb;
-+ struct page *page;
-+ int i, batch_target, notify;
-+ RING_IDX req_prod = np->rx.req_prod_pvt;
-+ struct xen_memory_reservation reservation;
-+ grant_ref_t ref;
-+ unsigned long pfn;
-+ void *vaddr;
-+ int nr_flips;
-+ netif_rx_request_t *req;
++ struct pci_bus *b;
++ struct pci_dev *d;
++ unsigned int devfn;
+
-+ if (unlikely(!netfront_carrier_ok(np)))
-+ return;
++#ifndef CONFIG_PCI_DOMAINS
++ if (domain != 0) {
++ dev_err(&pdev->xdev->dev,
++ "PCI Root in non-zero PCI Domain! domain=%d\n", domain);
++ dev_err(&pdev->xdev->dev,
++ "Please compile with CONFIG_PCI_DOMAINS\n");
++ return -EINVAL;
++ }
++#endif
+
-+ /*
-+ * Allocate skbuffs greedily, even though we batch updates to the
-+ * receive ring. This creates a less bursty demand on the memory
-+ * allocator, so should reduce the chance of failed allocation requests
-+ * both for ourself and for other kernel subsystems.
-+ */
-+ batch_target = np->rx_target - (req_prod - np->rx.rsp_cons);
-+ for (i = skb_queue_len(&np->rx_batch); i < batch_target; i++) {
-+ /*
-+ * Allocate an skb and a page. Do not use __dev_alloc_skb as
-+ * that will allocate page-sized buffers which is not
-+ * necessary here.
-+ * 16 bytes added as necessary headroom for netif_receive_skb.
-+ */
-+ skb = alloc_skb(RX_COPY_THRESHOLD + 16 + NET_IP_ALIGN,
-+ GFP_ATOMIC | __GFP_NOWARN);
-+ if (unlikely(!skb))
-+ goto no_skb;
++ dev_info(&pdev->xdev->dev, "Rescanning PCI Frontend Bus %04x:%02x\n",
++ domain, bus);
+
-+ page = alloc_page(GFP_ATOMIC | __GFP_NOWARN);
-+ if (!page) {
-+ kfree_skb(skb);
-+no_skb:
-+ /* Any skbuffs queued for refill? Force them out. */
-+ if (i != 0)
-+ goto refill;
-+ /* Could not allocate any skbuffs. Try again later. */
-+ mod_timer(&np->rx_refill_timer,
-+ jiffies + (HZ/10));
-+ break;
++ b = pci_find_bus(domain, bus);
++ if(!b)
++ /* If the bus is unknown, create it. */
++ return pcifront_scan_root(pdev, domain, bus);
++
++ /* Rescan the bus for newly attached functions and add.
++ * We omit handling of PCI bridge attachment because pciback prevents
++ * bridges from being exported.
++ */
++ for (devfn = 0; devfn < 0x100; devfn++) {
++ d = pci_get_slot(b, devfn);
++ if(d) {
++ /* Device is already known. */
++ pci_dev_put(d);
++ continue;
+ }
+
-+ skb_reserve(skb, 16 + NET_IP_ALIGN); /* mimic dev_alloc_skb() */
-+ skb_shinfo(skb)->frags[0].page = page;
-+ skb_shinfo(skb)->nr_frags = 1;
-+ __skb_queue_tail(&np->rx_batch, skb);
++ d = pci_scan_single_device(b, devfn);
++ if (d) {
++ dev_info(&pdev->xdev->dev, "New device on "
++ "%04x:%02x:%02x.%02x found.\n", domain, bus,
++ PCI_SLOT(devfn), PCI_FUNC(devfn));
++ pci_bus_add_device(d);
++ }
+ }
+
-+ /* Is the batch large enough to be worthwhile? */
-+ if (i < (np->rx_target/2)) {
-+ if (req_prod > np->rx.sring->req_prod)
-+ goto push;
-+ return;
-+ }
++ return 0;
++}
+
-+ /* Adjust our fill target if we risked running out of buffers. */
-+ if (((req_prod - np->rx.sring->rsp_prod) < (np->rx_target / 4)) &&
-+ ((np->rx_target *= 2) > np->rx_max_target))
-+ np->rx_target = np->rx_max_target;
++static void free_root_bus_devs(struct pci_bus *bus)
++{
++ struct pci_dev *dev;
+
-+ refill:
-+ for (nr_flips = i = 0; ; i++) {
-+ if ((skb = __skb_dequeue(&np->rx_batch)) == NULL)
-+ break;
++ while (!list_empty(&bus->devices)) {
++ dev = container_of(bus->devices.next, struct pci_dev,
++ bus_list);
++ dev_dbg(&dev->dev, "removing device\n");
++ pci_remove_bus_device(dev);
++ }
++}
+
-+ skb->dev = dev;
++void pcifront_free_roots(struct pcifront_device *pdev)
++{
++ struct pci_bus_entry *bus_entry, *t;
+
-+ id = xennet_rxidx(req_prod + i);
++ dev_dbg(&pdev->xdev->dev, "cleaning up root buses\n");
+
-+ BUG_ON(np->rx_skbs[id]);
-+ np->rx_skbs[id] = skb;
++ list_for_each_entry_safe(bus_entry, t, &pdev->root_buses, list) {
++ list_del(&bus_entry->list);
+
-+ ref = gnttab_claim_grant_reference(&np->gref_rx_head);
-+ BUG_ON((signed short)ref < 0);
-+ np->grant_rx_ref[id] = ref;
++ free_root_bus_devs(bus_entry->bus);
+
-+ pfn = page_to_pfn(skb_shinfo(skb)->frags[0].page);
-+ vaddr = page_address(skb_shinfo(skb)->frags[0].page);
++ kfree(bus_entry->bus->sysdata);
+
-+ req = RING_GET_REQUEST(&np->rx, req_prod + i);
-+ if (!np->copying_receiver) {
-+ gnttab_grant_foreign_transfer_ref(ref,
-+ np->xbdev->otherend_id,
-+ pfn);
-+ np->rx_pfn_array[nr_flips] = pfn_to_mfn(pfn);
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ /* Remove this page before passing
-+ * back to Xen. */
-+ set_phys_to_machine(pfn, INVALID_P2M_ENTRY);
-+ MULTI_update_va_mapping(np->rx_mcl+i,
-+ (unsigned long)vaddr,
-+ __pte(0), 0);
-+ }
-+ nr_flips++;
-+ } else {
-+ gnttab_grant_foreign_access_ref(ref,
-+ np->xbdev->otherend_id,
-+ pfn_to_mfn(pfn),
-+ 0);
-+ }
++ device_unregister(bus_entry->bus->bridge);
++ pci_remove_bus(bus_entry->bus);
+
-+ req->id = id;
-+ req->gref = ref;
++ kfree(bus_entry);
+ }
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pcifront/pcifront.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pcifront/pcifront.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,42 @@
++/*
++ * PCI Frontend - Common data structures & function declarations
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#ifndef __XEN_PCIFRONT_H__
++#define __XEN_PCIFRONT_H__
+
-+ if ( nr_flips != 0 ) {
-+ /* Tell the ballon driver what is going on. */
-+ balloon_update_driver_allowance(i);
++#include <linux/spinlock.h>
++#include <linux/pci.h>
++#include <xen/xenbus.h>
++#include <xen/interface/io/pciif.h>
++#include <xen/pcifront.h>
+
-+ set_xen_guest_handle(reservation.extent_start,
-+ np->rx_pfn_array);
-+ reservation.nr_extents = nr_flips;
-+ reservation.extent_order = 0;
-+ reservation.address_bits = 0;
-+ reservation.domid = DOMID_SELF;
++struct pci_bus_entry {
++ struct list_head list;
++ struct pci_bus *bus;
++};
+
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ /* After all PTEs have been zapped, flush the TLB. */
-+ np->rx_mcl[i-1].args[MULTI_UVMFLAGS_INDEX] =
-+ UVMF_TLB_FLUSH|UVMF_ALL;
++struct pcifront_device {
++ struct xenbus_device *xdev;
++ struct list_head root_buses;
++ spinlock_t dev_lock;
+
-+ /* Give away a batch of pages. */
-+ np->rx_mcl[i].op = __HYPERVISOR_memory_op;
-+ np->rx_mcl[i].args[0] = XENMEM_decrease_reservation;
-+ np->rx_mcl[i].args[1] = (unsigned long)&reservation;
++ int evtchn;
++ int gnt_ref;
+
-+ /* Zap PTEs and give away pages in one big
-+ * multicall. */
-+ (void)HYPERVISOR_multicall(np->rx_mcl, i+1);
++ /* Lock this when doing any operations in sh_info */
++ spinlock_t sh_info_lock;
++ struct xen_pci_sharedinfo *sh_info;
++};
+
-+ /* Check return status of HYPERVISOR_memory_op(). */
-+ if (unlikely(np->rx_mcl[i].result != i))
-+ panic("Unable to reduce memory reservation\n");
-+ } else {
-+ if (HYPERVISOR_memory_op(XENMEM_decrease_reservation,
-+ &reservation) != i)
-+ panic("Unable to reduce memory reservation\n");
-+ }
-+ } else {
-+ wmb();
-+ }
++int pcifront_connect(struct pcifront_device *pdev);
++void pcifront_disconnect(struct pcifront_device *pdev);
+
-+ /* Above is a suitable barrier to ensure backend will see requests. */
-+ np->rx.req_prod_pvt = req_prod + i;
-+ push:
-+ RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&np->rx, notify);
-+ if (notify)
-+ notify_remote_via_irq(np->irq);
-+}
++int pcifront_scan_root(struct pcifront_device *pdev,
++ unsigned int domain, unsigned int bus);
++int pcifront_rescan_root(struct pcifront_device *pdev,
++ unsigned int domain, unsigned int bus);
++void pcifront_free_roots(struct pcifront_device *pdev);
+
-+static void xennet_make_frags(struct sk_buff *skb, struct net_device *dev,
-+ struct netif_tx_request *tx)
-+{
-+ struct netfront_info *np = netdev_priv(dev);
-+ char *data = skb->data;
-+ unsigned long mfn;
-+ RING_IDX prod = np->tx.req_prod_pvt;
-+ int frags = skb_shinfo(skb)->nr_frags;
-+ unsigned int offset = offset_in_page(data);
-+ unsigned int len = skb_headlen(skb);
-+ unsigned int id;
-+ grant_ref_t ref;
-+ int i;
++#endif /* __XEN_PCIFRONT_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/pcifront/xenbus.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/pcifront/xenbus.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,455 @@
++/*
++ * PCI Frontend Xenbus Setup - handles setup with backend (imports page/evtchn)
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/mm.h>
++#include <xen/xenbus.h>
++#include <xen/gnttab.h>
++#include "pcifront.h"
+
-+ while (len > PAGE_SIZE - offset) {
-+ tx->size = PAGE_SIZE - offset;
-+ tx->flags |= NETTXF_more_data;
-+ len -= tx->size;
-+ data += tx->size;
-+ offset = 0;
++#ifndef __init_refok
++#define __init_refok
++#endif
+
-+ id = get_id_from_freelist(np->tx_skbs);
-+ np->tx_skbs[id] = skb_get(skb);
-+ tx = RING_GET_REQUEST(&np->tx, prod++);
-+ tx->id = id;
-+ ref = gnttab_claim_grant_reference(&np->gref_tx_head);
-+ BUG_ON((signed short)ref < 0);
++#define INVALID_GRANT_REF (0)
++#define INVALID_EVTCHN (-1)
+
-+ mfn = virt_to_mfn(data);
-+ gnttab_grant_foreign_access_ref(ref, np->xbdev->otherend_id,
-+ mfn, GNTMAP_readonly);
++static struct pcifront_device *alloc_pdev(struct xenbus_device *xdev)
++{
++ struct pcifront_device *pdev;
+
-+ tx->gref = np->grant_tx_ref[id] = ref;
-+ tx->offset = offset;
-+ tx->size = len;
-+ tx->flags = 0;
-+ }
++ pdev = kzalloc(sizeof(struct pcifront_device), GFP_KERNEL);
++ if (pdev == NULL)
++ goto out;
+
-+ for (i = 0; i < frags; i++) {
-+ skb_frag_t *frag = skb_shinfo(skb)->frags + i;
++ pdev->sh_info =
++ (struct xen_pci_sharedinfo *)__get_free_page(GFP_KERNEL);
++ if (pdev->sh_info == NULL) {
++ kfree(pdev);
++ pdev = NULL;
++ goto out;
++ }
++ pdev->sh_info->flags = 0;
+
-+ tx->flags |= NETTXF_more_data;
++ xdev->dev.driver_data = pdev;
++ pdev->xdev = xdev;
+
-+ id = get_id_from_freelist(np->tx_skbs);
-+ np->tx_skbs[id] = skb_get(skb);
-+ tx = RING_GET_REQUEST(&np->tx, prod++);
-+ tx->id = id;
-+ ref = gnttab_claim_grant_reference(&np->gref_tx_head);
-+ BUG_ON((signed short)ref < 0);
++ INIT_LIST_HEAD(&pdev->root_buses);
+
-+ mfn = pfn_to_mfn(page_to_pfn(frag->page));
-+ gnttab_grant_foreign_access_ref(ref, np->xbdev->otherend_id,
-+ mfn, GNTMAP_readonly);
++ spin_lock_init(&pdev->dev_lock);
++ spin_lock_init(&pdev->sh_info_lock);
+
-+ tx->gref = np->grant_tx_ref[id] = ref;
-+ tx->offset = frag->page_offset;
-+ tx->size = frag->size;
-+ tx->flags = 0;
-+ }
++ pdev->evtchn = INVALID_EVTCHN;
++ pdev->gnt_ref = INVALID_GRANT_REF;
+
-+ np->tx.req_prod_pvt = prod;
++ dev_dbg(&xdev->dev, "Allocated pdev @ 0x%p pdev->sh_info @ 0x%p\n",
++ pdev, pdev->sh_info);
++ out:
++ return pdev;
+}
+
-+static int network_start_xmit(struct sk_buff *skb, struct net_device *dev)
++static void free_pdev(struct pcifront_device *pdev)
+{
-+ unsigned short id;
-+ struct netfront_info *np = netdev_priv(dev);
-+ struct netif_tx_request *tx;
-+ struct netif_extra_info *extra;
-+ char *data = skb->data;
-+ RING_IDX i;
-+ grant_ref_t ref;
-+ unsigned long mfn;
-+ int notify;
-+ int frags = skb_shinfo(skb)->nr_frags;
-+ unsigned int offset = offset_in_page(data);
-+ unsigned int len = skb_headlen(skb);
-+
-+ frags += (offset + len + PAGE_SIZE - 1) / PAGE_SIZE;
-+ if (unlikely(frags > MAX_SKB_FRAGS + 1)) {
-+ printk(KERN_ALERT "xennet: skb rides the rocket: %d frags\n",
-+ frags);
-+ dump_stack();
-+ goto drop;
-+ }
++ dev_dbg(&pdev->xdev->dev, "freeing pdev @ 0x%p\n", pdev);
+
-+ spin_lock_irq(&np->tx_lock);
++ pcifront_free_roots(pdev);
+
-+ if (unlikely(!netfront_carrier_ok(np) ||
-+ (frags > 1 && !xennet_can_sg(dev)) ||
-+ netif_needs_gso(dev, skb))) {
-+ spin_unlock_irq(&np->tx_lock);
-+ goto drop;
-+ }
++ if (pdev->evtchn != INVALID_EVTCHN)
++ xenbus_free_evtchn(pdev->xdev, pdev->evtchn);
+
-+ i = np->tx.req_prod_pvt;
++ if (pdev->gnt_ref != INVALID_GRANT_REF)
++ gnttab_end_foreign_access(pdev->gnt_ref,
++ (unsigned long)pdev->sh_info);
+
-+ id = get_id_from_freelist(np->tx_skbs);
-+ np->tx_skbs[id] = skb;
++ pdev->xdev->dev.driver_data = NULL;
+
-+ tx = RING_GET_REQUEST(&np->tx, i);
++ kfree(pdev);
++}
+
-+ tx->id = id;
-+ ref = gnttab_claim_grant_reference(&np->gref_tx_head);
-+ BUG_ON((signed short)ref < 0);
-+ mfn = virt_to_mfn(data);
-+ gnttab_grant_foreign_access_ref(
-+ ref, np->xbdev->otherend_id, mfn, GNTMAP_readonly);
-+ tx->gref = np->grant_tx_ref[id] = ref;
-+ tx->offset = offset;
-+ tx->size = len;
++static int pcifront_publish_info(struct pcifront_device *pdev)
++{
++ int err = 0;
++ struct xenbus_transaction trans;
+
-+ tx->flags = 0;
-+ extra = NULL;
++ err = xenbus_grant_ring(pdev->xdev, virt_to_mfn(pdev->sh_info));
++ if (err < 0)
++ goto out;
+
-+ if (skb->ip_summed == CHECKSUM_HW) /* local packet? */
-+ tx->flags |= NETTXF_csum_blank | NETTXF_data_validated;
-+#ifdef CONFIG_XEN
-+ if (skb->proto_data_valid) /* remote but checksummed? */
-+ tx->flags |= NETTXF_data_validated;
-+#endif
++ pdev->gnt_ref = err;
+
-+#ifdef HAVE_TSO
-+ if (skb_shinfo(skb)->gso_size) {
-+ struct netif_extra_info *gso = (struct netif_extra_info *)
-+ RING_GET_REQUEST(&np->tx, ++i);
++ err = xenbus_alloc_evtchn(pdev->xdev, &pdev->evtchn);
++ if (err)
++ goto out;
+
-+ if (extra)
-+ extra->flags |= XEN_NETIF_EXTRA_FLAG_MORE;
-+ else
-+ tx->flags |= NETTXF_extra_info;
++ do_publish:
++ err = xenbus_transaction_start(&trans);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error writing configuration for backend "
++ "(start transaction)");
++ goto out;
++ }
+
-+ gso->u.gso.size = skb_shinfo(skb)->gso_size;
-+ gso->u.gso.type = XEN_NETIF_GSO_TYPE_TCPV4;
-+ gso->u.gso.pad = 0;
-+ gso->u.gso.features = 0;
++ err = xenbus_printf(trans, pdev->xdev->nodename,
++ "pci-op-ref", "%u", pdev->gnt_ref);
++ if (!err)
++ err = xenbus_printf(trans, pdev->xdev->nodename,
++ "event-channel", "%u", pdev->evtchn);
++ if (!err)
++ err = xenbus_printf(trans, pdev->xdev->nodename,
++ "magic", XEN_PCI_MAGIC);
+
-+ gso->type = XEN_NETIF_EXTRA_TYPE_GSO;
-+ gso->flags = 0;
-+ extra = gso;
++ if (err) {
++ xenbus_transaction_end(trans, 1);
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error writing configuration for backend");
++ goto out;
++ } else {
++ err = xenbus_transaction_end(trans, 0);
++ if (err == -EAGAIN)
++ goto do_publish;
++ else if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error completing transaction "
++ "for backend");
++ goto out;
++ }
+ }
-+#endif
+
-+ np->tx.req_prod_pvt = i + 1;
++ xenbus_switch_state(pdev->xdev, XenbusStateInitialised);
+
-+ xennet_make_frags(skb, dev, tx);
-+ tx->size = skb->len;
++ dev_dbg(&pdev->xdev->dev, "publishing successful!\n");
+
-+ RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&np->tx, notify);
-+ if (notify)
-+ notify_remote_via_irq(np->irq);
++ out:
++ return err;
++}
+
-+ np->stats.tx_bytes += skb->len;
-+ np->stats.tx_packets++;
++static int __devinit pcifront_try_connect(struct pcifront_device *pdev)
++{
++ int err = -EFAULT;
++ int i, num_roots, len;
++ char str[64];
++ unsigned int domain, bus;
+
-+ /* Note: It is not safe to access skb after network_tx_buf_gc()! */
-+ network_tx_buf_gc(dev);
++ spin_lock(&pdev->dev_lock);
+
-+ if (!netfront_tx_slot_available(np))
-+ netif_stop_queue(dev);
++ /* Only connect once */
++ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
++ XenbusStateInitialised)
++ goto out;
+
-+ spin_unlock_irq(&np->tx_lock);
++ err = pcifront_connect(pdev);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error connecting PCI Frontend");
++ goto out;
++ }
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
++ "root_num", "%d", &num_roots);
++ if (err == -ENOENT) {
++ xenbus_dev_error(pdev->xdev, err,
++ "No PCI Roots found, trying 0000:00");
++ err = pcifront_scan_root(pdev, 0, 0);
++ num_roots = 0;
++ } else if (err != 1) {
++ if (err == 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading number of PCI roots");
++ goto out;
++ }
++
++ for (i = 0; i < num_roots; i++) {
++ len = snprintf(str, sizeof(str), "root-%d", i);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
++ "%x:%x", &domain, &bus);
++ if (err != 2) {
++ if (err >= 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading PCI root %d", i);
++ goto out;
++ }
++
++ err = pcifront_scan_root(pdev, domain, bus);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error scanning PCI root %04x:%02x",
++ domain, bus);
++ goto out;
++ }
++ }
+
-+ return 0;
++ err = xenbus_switch_state(pdev->xdev, XenbusStateConnected);
++ if (err)
++ goto out;
+
-+ drop:
-+ np->stats.tx_dropped++;
-+ dev_kfree_skb(skb);
-+ return 0;
++ out:
++ spin_unlock(&pdev->dev_lock);
++ return err;
+}
+
-+static irqreturn_t netif_int(int irq, void *dev_id, struct pt_regs *ptregs)
++static int pcifront_try_disconnect(struct pcifront_device *pdev)
+{
-+ struct net_device *dev = dev_id;
-+ struct netfront_info *np = netdev_priv(dev);
-+ unsigned long flags;
++ int err = 0;
++ enum xenbus_state prev_state;
+
-+ spin_lock_irqsave(&np->tx_lock, flags);
++ spin_lock(&pdev->dev_lock);
+
-+ if (likely(netfront_carrier_ok(np))) {
-+ network_tx_buf_gc(dev);
-+ /* Under tx_lock: protects access to rx shared-ring indexes. */
-+ if (RING_HAS_UNCONSUMED_RESPONSES(&np->rx))
-+ netif_rx_schedule(dev);
++ prev_state = xenbus_read_driver_state(pdev->xdev->nodename);
++
++ if (prev_state >= XenbusStateClosing)
++ goto out;
++
++ if(prev_state == XenbusStateConnected) {
++ pcifront_free_roots(pdev);
++ pcifront_disconnect(pdev);
+ }
+
-+ spin_unlock_irqrestore(&np->tx_lock, flags);
++ err = xenbus_switch_state(pdev->xdev, XenbusStateClosed);
+
-+ return IRQ_HANDLED;
++ out:
++ spin_unlock(&pdev->dev_lock);
++
++ return err;
+}
+
-+static void xennet_move_rx_slot(struct netfront_info *np, struct sk_buff *skb,
-+ grant_ref_t ref)
++static int __devinit pcifront_attach_devices(struct pcifront_device *pdev)
+{
-+ int new = xennet_rxidx(np->rx.req_prod_pvt);
-+
-+ BUG_ON(np->rx_skbs[new]);
-+ np->rx_skbs[new] = skb;
-+ np->grant_rx_ref[new] = ref;
-+ RING_GET_REQUEST(&np->rx, np->rx.req_prod_pvt)->id = new;
-+ RING_GET_REQUEST(&np->rx, np->rx.req_prod_pvt)->gref = ref;
-+ np->rx.req_prod_pvt++;
-+}
++ int err = -EFAULT;
++ int i, num_roots, len;
++ unsigned int domain, bus;
++ char str[64];
+
-+int xennet_get_extras(struct netfront_info *np,
-+ struct netif_extra_info *extras, RING_IDX rp)
++ spin_lock(&pdev->dev_lock);
+
-+{
-+ struct netif_extra_info *extra;
-+ RING_IDX cons = np->rx.rsp_cons;
-+ int err = 0;
++ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
++ XenbusStateReconfiguring)
++ goto out;
+
-+ do {
-+ struct sk_buff *skb;
-+ grant_ref_t ref;
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
++ "root_num", "%d", &num_roots);
++ if (err == -ENOENT) {
++ xenbus_dev_error(pdev->xdev, err,
++ "No PCI Roots found, trying 0000:00");
++ err = pcifront_rescan_root(pdev, 0, 0);
++ num_roots = 0;
++ } else if (err != 1) {
++ if (err == 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading number of PCI roots");
++ goto out;
++ }
+
-+ if (unlikely(cons + 1 == rp)) {
-+ if (net_ratelimit())
-+ WPRINTK("Missing extra info\n");
-+ err = -EBADR;
-+ break;
++ for (i = 0; i < num_roots; i++) {
++ len = snprintf(str, sizeof(str), "root-%d", i);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
+ }
+
-+ extra = (struct netif_extra_info *)
-+ RING_GET_RESPONSE(&np->rx, ++cons);
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
++ "%x:%x", &domain, &bus);
++ if (err != 2) {
++ if (err >= 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading PCI root %d", i);
++ goto out;
++ }
+
-+ if (unlikely(!extra->type ||
-+ extra->type >= XEN_NETIF_EXTRA_TYPE_MAX)) {
-+ if (net_ratelimit())
-+ WPRINTK("Invalid extra type: %d\n",
-+ extra->type);
-+ err = -EINVAL;
-+ } else {
-+ memcpy(&extras[extra->type - 1], extra,
-+ sizeof(*extra));
++ err = pcifront_rescan_root(pdev, domain, bus);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error scanning PCI root %04x:%02x",
++ domain, bus);
++ goto out;
+ }
++ }
+
-+ skb = xennet_get_rx_skb(np, cons);
-+ ref = xennet_get_rx_ref(np, cons);
-+ xennet_move_rx_slot(np, skb, ref);
-+ } while (extra->flags & XEN_NETIF_EXTRA_FLAG_MORE);
++ xenbus_switch_state(pdev->xdev, XenbusStateConnected);
+
-+ np->rx.rsp_cons = cons;
++ out:
++ spin_unlock(&pdev->dev_lock);
+ return err;
+}
+
-+static int xennet_get_responses(struct netfront_info *np,
-+ struct netfront_rx_info *rinfo, RING_IDX rp,
-+ struct sk_buff_head *list,
-+ int *pages_flipped_p)
++static int pcifront_detach_devices(struct pcifront_device *pdev)
+{
-+ int pages_flipped = *pages_flipped_p;
-+ struct mmu_update *mmu;
-+ struct multicall_entry *mcl;
-+ struct netif_rx_response *rx = &rinfo->rx;
-+ struct netif_extra_info *extras = rinfo->extras;
-+ RING_IDX cons = np->rx.rsp_cons;
-+ struct sk_buff *skb = xennet_get_rx_skb(np, cons);
-+ grant_ref_t ref = xennet_get_rx_ref(np, cons);
-+ int max = MAX_SKB_FRAGS + (rx->status <= RX_COPY_THRESHOLD);
-+ int frags = 1;
+ int err = 0;
-+ unsigned long ret;
++ int i, num_devs;
++ unsigned int domain, bus, slot, func;
++ struct pci_bus *pci_bus;
++ struct pci_dev *pci_dev;
++ char str[64];
+
-+ if (rx->flags & NETRXF_extra_info) {
-+ err = xennet_get_extras(np, extras, rp);
-+ cons = np->rx.rsp_cons;
-+ }
++ spin_lock(&pdev->dev_lock);
+
-+ for (;;) {
-+ unsigned long mfn;
++ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
++ XenbusStateConnected)
++ goto out;
+
-+ if (unlikely(rx->status < 0 ||
-+ rx->offset + rx->status > PAGE_SIZE)) {
-+ if (net_ratelimit())
-+ WPRINTK("rx->offset: %x, size: %u\n",
-+ rx->offset, rx->status);
-+ xennet_move_rx_slot(np, skb, ref);
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, "num_devs", "%d",
++ &num_devs);
++ if (err != 1) {
++ if (err >= 0)
+ err = -EINVAL;
-+ goto next;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading number of PCI devices");
++ goto out;
++ }
++
++ /* Find devices being detached and remove them. */
++ for (i = 0; i < num_devs; i++) {
++ int l, state;
++ l = snprintf(str, sizeof(str), "state-%d", i);
++ if (unlikely(l >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
+ }
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str, "%d",
++ &state);
++ if (err != 1)
++ state = XenbusStateUnknown;
+
-+ /*
-+ * This definitely indicates a bug, either in this driver or in
-+ * the backend driver. In future this should flag the bad
-+ * situation to the system controller to reboot the backed.
-+ */
-+ if (ref == GRANT_INVALID_REF) {
-+ if (net_ratelimit())
-+ WPRINTK("Bad rx response id %d.\n", rx->id);
-+ err = -EINVAL;
-+ goto next;
++ if (state != XenbusStateClosing)
++ continue;
++
++ /* Remove device. */
++ l = snprintf(str, sizeof(str), "vdev-%d", i);
++ if (unlikely(l >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
++ "%x:%x:%x.%x", &domain, &bus, &slot, &func);
++ if (err != 4) {
++ if (err >= 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading PCI device %d", i);
++ goto out;
+ }
+
-+ if (!np->copying_receiver) {
-+ /* Memory pressure, insufficient buffer
-+ * headroom, ... */
-+ if (!(mfn = gnttab_end_foreign_transfer_ref(ref))) {
-+ if (net_ratelimit())
-+ WPRINTK("Unfulfilled rx req "
-+ "(id=%d, st=%d).\n",
-+ rx->id, rx->status);
-+ xennet_move_rx_slot(np, skb, ref);
-+ err = -ENOMEM;
-+ goto next;
-+ }
++ pci_bus = pci_find_bus(domain, bus);
++ if(!pci_bus) {
++ dev_dbg(&pdev->xdev->dev, "Cannot get bus %04x:%02x\n",
++ domain, bus);
++ continue;
++ }
++ pci_dev = pci_get_slot(pci_bus, PCI_DEVFN(slot, func));
++ if(!pci_dev) {
++ dev_dbg(&pdev->xdev->dev,
++ "Cannot get PCI device %04x:%02x:%02x.%02x\n",
++ domain, bus, slot, func);
++ continue;
++ }
++ pci_remove_bus_device(pci_dev);
++ pci_dev_put(pci_dev);
+
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ /* Remap the page. */
-+ struct page *page =
-+ skb_shinfo(skb)->frags[0].page;
-+ unsigned long pfn = page_to_pfn(page);
-+ void *vaddr = page_address(page);
++ dev_dbg(&pdev->xdev->dev,
++ "PCI device %04x:%02x:%02x.%02x removed.\n",
++ domain, bus, slot, func);
++ }
+
-+ mcl = np->rx_mcl + pages_flipped;
-+ mmu = np->rx_mmu + pages_flipped;
++ err = xenbus_switch_state(pdev->xdev, XenbusStateReconfiguring);
+
-+ MULTI_update_va_mapping(mcl,
-+ (unsigned long)vaddr,
-+ pfn_pte_ma(mfn,
-+ PAGE_KERNEL),
-+ 0);
-+ mmu->ptr = ((maddr_t)mfn << PAGE_SHIFT)
-+ | MMU_MACHPHYS_UPDATE;
-+ mmu->val = pfn;
++ out:
++ spin_unlock(&pdev->dev_lock);
++ return err;
++}
+
-+ set_phys_to_machine(pfn, mfn);
-+ }
-+ pages_flipped++;
-+ } else {
-+ ret = gnttab_end_foreign_access_ref(ref, 0);
-+ BUG_ON(!ret);
-+ }
++static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
++ enum xenbus_state be_state)
++{
++ struct pcifront_device *pdev = xdev->dev.driver_data;
+
-+ gnttab_release_grant_reference(&np->gref_rx_head, ref);
++ switch (be_state) {
++ case XenbusStateUnknown:
++ case XenbusStateInitialising:
++ case XenbusStateInitWait:
++ case XenbusStateInitialised:
++ case XenbusStateClosed:
++ break;
+
-+ __skb_queue_tail(list, skb);
++ case XenbusStateConnected:
++ pcifront_try_connect(pdev);
++ break;
+
-+next:
-+ if (!(rx->flags & NETRXF_more_data))
-+ break;
++ case XenbusStateClosing:
++ dev_warn(&xdev->dev, "backend going away!\n");
++ pcifront_try_disconnect(pdev);
++ break;
+
-+ if (cons + frags == rp) {
-+ if (net_ratelimit())
-+ WPRINTK("Need more frags\n");
-+ err = -ENOENT;
-+ break;
-+ }
++ case XenbusStateReconfiguring:
++ pcifront_detach_devices(pdev);
++ break;
+
-+ rx = RING_GET_RESPONSE(&np->rx, cons + frags);
-+ skb = xennet_get_rx_skb(np, cons + frags);
-+ ref = xennet_get_rx_ref(np, cons + frags);
-+ frags++;
++ case XenbusStateReconfigured:
++ pcifront_attach_devices(pdev);
++ break;
+ }
++}
+
-+ if (unlikely(frags > max)) {
-+ if (net_ratelimit())
-+ WPRINTK("Too many frags\n");
-+ err = -E2BIG;
-+ }
++static int pcifront_xenbus_probe(struct xenbus_device *xdev,
++ const struct xenbus_device_id *id)
++{
++ int err = 0;
++ struct pcifront_device *pdev = alloc_pdev(xdev);
+
-+ if (unlikely(err))
-+ np->rx.rsp_cons = cons + frags;
++ if (pdev == NULL) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(xdev, err,
++ "Error allocating pcifront_device struct");
++ goto out;
++ }
+
-+ *pages_flipped_p = pages_flipped;
++ err = pcifront_publish_info(pdev);
+
++ out:
+ return err;
+}
+
-+static RING_IDX xennet_fill_frags(struct netfront_info *np,
-+ struct sk_buff *skb,
-+ struct sk_buff_head *list)
++static int pcifront_xenbus_remove(struct xenbus_device *xdev)
+{
-+ struct skb_shared_info *shinfo = skb_shinfo(skb);
-+ int nr_frags = shinfo->nr_frags;
-+ RING_IDX cons = np->rx.rsp_cons;
-+ skb_frag_t *frag = shinfo->frags + nr_frags;
-+ struct sk_buff *nskb;
-+
-+ while ((nskb = __skb_dequeue(list))) {
-+ struct netif_rx_response *rx =
-+ RING_GET_RESPONSE(&np->rx, ++cons);
++ if (xdev->dev.driver_data)
++ free_pdev(xdev->dev.driver_data);
+
-+ frag->page = skb_shinfo(nskb)->frags[0].page;
-+ frag->page_offset = rx->offset;
-+ frag->size = rx->status;
++ return 0;
++}
+
-+ skb->data_len += rx->status;
++static const struct xenbus_device_id xenpci_ids[] = {
++ {"pci"},
++ {{0}},
++};
++MODULE_ALIAS("xen:pci");
+
-+ skb_shinfo(nskb)->nr_frags = 0;
-+ kfree_skb(nskb);
++static struct xenbus_driver xenbus_pcifront_driver = {
++ .name = "pcifront",
++ .owner = THIS_MODULE,
++ .ids = xenpci_ids,
++ .probe = pcifront_xenbus_probe,
++ .remove = pcifront_xenbus_remove,
++ .otherend_changed = pcifront_backend_changed,
++};
+
-+ frag++;
-+ nr_frags++;
-+ }
++static int __init pcifront_init(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+ shinfo->nr_frags = nr_frags;
-+ return cons;
++ return xenbus_register_frontend(&xenbus_pcifront_driver);
+}
+
-+static int xennet_set_skb_gso(struct sk_buff *skb,
-+ struct netif_extra_info *gso)
++/* Initialize after the Xen PCI Frontend Stub is initialized */
++subsys_initcall(pcifront_init);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/privcmd/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/privcmd/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,3 @@
++
++obj-y += privcmd.o
++obj-$(CONFIG_COMPAT) += compat_privcmd.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/privcmd/compat_privcmd.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/privcmd/compat_privcmd.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,73 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
++
++#include <linux/config.h>
++#include <linux/compat.h>
++#include <linux/ioctl.h>
++#include <linux/syscalls.h>
++#include <asm/hypervisor.h>
++#include <asm/uaccess.h>
++#include <xen/public/privcmd.h>
++#include <xen/compat_ioctl.h>
++
++int privcmd_ioctl_32(int fd, unsigned int cmd, unsigned long arg)
+{
-+ if (!gso->u.gso.size) {
-+ if (net_ratelimit())
-+ WPRINTK("GSO size must not be zero.\n");
-+ return -EINVAL;
-+ }
++ int ret;
+
-+ /* Currently only TCPv4 S.O. is supported. */
-+ if (gso->u.gso.type != XEN_NETIF_GSO_TYPE_TCPV4) {
-+ if (net_ratelimit())
-+ WPRINTK("Bad GSO type %d.\n", gso->u.gso.type);
-+ return -EINVAL;
++ switch (cmd) {
++ case IOCTL_PRIVCMD_MMAP_32: {
++ struct privcmd_mmap *p;
++ struct privcmd_mmap_32 *p32;
++ struct privcmd_mmap_32 n32;
++
++ p32 = compat_ptr(arg);
++ p = compat_alloc_user_space(sizeof(*p));
++ if (copy_from_user(&n32, p32, sizeof(n32)) ||
++ put_user(n32.num, &p->num) ||
++ put_user(n32.dom, &p->dom) ||
++ put_user(compat_ptr(n32.entry), &p->entry))
++ return -EFAULT;
++
++ ret = sys_ioctl(fd, IOCTL_PRIVCMD_MMAP, (unsigned long)p);
++ }
++ break;
++ case IOCTL_PRIVCMD_MMAPBATCH_32: {
++ struct privcmd_mmapbatch *p;
++ struct privcmd_mmapbatch_32 *p32;
++ struct privcmd_mmapbatch_32 n32;
++
++ p32 = compat_ptr(arg);
++ p = compat_alloc_user_space(sizeof(*p));
++ if (copy_from_user(&n32, p32, sizeof(n32)) ||
++ put_user(n32.num, &p->num) ||
++ put_user(n32.dom, &p->dom) ||
++ put_user(n32.addr, &p->addr) ||
++ put_user(compat_ptr(n32.arr), &p->arr))
++ return -EFAULT;
++
++ ret = sys_ioctl(fd, IOCTL_PRIVCMD_MMAPBATCH, (unsigned long)p);
++ }
++ break;
++ default:
++ ret = -EINVAL;
++ break;
+ }
++ return ret;
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/privcmd/privcmd.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/privcmd/privcmd.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,356 @@
++/******************************************************************************
++ * privcmd.c
++ *
++ * Interface to privileged domain-0 commands.
++ *
++ * Copyright (c) 2002-2004, K A Fraser, B Dragovic
++ */
+
-+#ifdef HAVE_TSO
-+ skb_shinfo(skb)->gso_size = gso->u.gso.size;
-+#ifdef HAVE_GSO
-+ skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4;
++#include <linux/kernel.h>
++#include <linux/sched.h>
++#include <linux/slab.h>
++#include <linux/string.h>
++#include <linux/errno.h>
++#include <linux/mm.h>
++#include <linux/mman.h>
++#include <linux/swap.h>
++#include <linux/smp_lock.h>
++#include <linux/highmem.h>
++#include <linux/pagemap.h>
++#include <linux/seq_file.h>
++#include <asm/hypervisor.h>
+
-+ /* Header must be checked, and gso_segs computed. */
-+ skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY;
-+#endif
-+ skb_shinfo(skb)->gso_segs = 0;
++#include <asm/pgalloc.h>
++#include <asm/pgtable.h>
++#include <asm/uaccess.h>
++#include <asm/tlb.h>
++#include <asm/hypervisor.h>
++#include <xen/public/privcmd.h>
++#include <xen/interface/xen.h>
++#include <xen/xen_proc.h>
++#include <xen/features.h>
+
-+ return 0;
-+#else
-+ if (net_ratelimit())
-+ WPRINTK("GSO unsupported by this kernel.\n");
-+ return -EINVAL;
++static struct proc_dir_entry *privcmd_intf;
++static struct proc_dir_entry *capabilities_intf;
++
++#ifndef HAVE_ARCH_PRIVCMD_MMAP
++static int privcmd_enforce_singleshot_mapping(struct vm_area_struct *vma);
+#endif
-+}
+
-+static int netif_poll(struct net_device *dev, int *pbudget)
++static long privcmd_ioctl(struct file *file,
++ unsigned int cmd, unsigned long data)
+{
-+ struct netfront_info *np = netdev_priv(dev);
-+ struct sk_buff *skb;
-+ struct netfront_rx_info rinfo;
-+ struct netif_rx_response *rx = &rinfo.rx;
-+ struct netif_extra_info *extras = rinfo.extras;
-+ RING_IDX i, rp;
-+ struct multicall_entry *mcl;
-+ int work_done, budget, more_to_do = 1;
-+ struct sk_buff_head rxq;
-+ struct sk_buff_head errq;
-+ struct sk_buff_head tmpq;
-+ unsigned long flags;
-+ unsigned int len;
-+ int pages_flipped = 0;
-+ int err;
++ int ret = -ENOSYS;
++ void __user *udata = (void __user *) data;
+
-+ spin_lock(&np->rx_lock); /* no need for spin_lock_bh() in ->poll() */
++ switch (cmd) {
++ case IOCTL_PRIVCMD_HYPERCALL: {
++ privcmd_hypercall_t hypercall;
++
++ if (copy_from_user(&hypercall, udata, sizeof(hypercall)))
++ return -EFAULT;
+
-+ if (unlikely(!netfront_carrier_ok(np))) {
-+ spin_unlock(&np->rx_lock);
-+ return 0;
++#if defined(__i386__)
++ if (hypercall.op >= (PAGE_SIZE >> 5))
++ break;
++ __asm__ __volatile__ (
++ "pushl %%ebx; pushl %%ecx; pushl %%edx; "
++ "pushl %%esi; pushl %%edi; "
++ "movl 8(%%eax),%%ebx ;"
++ "movl 16(%%eax),%%ecx ;"
++ "movl 24(%%eax),%%edx ;"
++ "movl 32(%%eax),%%esi ;"
++ "movl 40(%%eax),%%edi ;"
++ "movl (%%eax),%%eax ;"
++ "shll $5,%%eax ;"
++ "addl $hypercall_page,%%eax ;"
++ "call *%%eax ;"
++ "popl %%edi; popl %%esi; popl %%edx; "
++ "popl %%ecx; popl %%ebx"
++ : "=a" (ret) : "0" (&hypercall) : "memory" );
++#elif defined (__x86_64__)
++ if (hypercall.op < (PAGE_SIZE >> 5)) {
++ long ign1, ign2, ign3;
++ __asm__ __volatile__ (
++ "movq %8,%%r10; movq %9,%%r8;"
++ "shll $5,%%eax ;"
++ "addq $hypercall_page,%%rax ;"
++ "call *%%rax"
++ : "=a" (ret), "=D" (ign1),
++ "=S" (ign2), "=d" (ign3)
++ : "0" ((unsigned int)hypercall.op),
++ "1" (hypercall.arg[0]),
++ "2" (hypercall.arg[1]),
++ "3" (hypercall.arg[2]),
++ "g" (hypercall.arg[3]),
++ "g" (hypercall.arg[4])
++ : "r8", "r10", "memory" );
++ }
++#else
++ ret = privcmd_hypercall(&hypercall);
++#endif
+ }
++ break;
+
-+ skb_queue_head_init(&rxq);
-+ skb_queue_head_init(&errq);
-+ skb_queue_head_init(&tmpq);
-+
-+ if ((budget = *pbudget) > dev->quota)
-+ budget = dev->quota;
-+ rp = np->rx.sring->rsp_prod;
-+ rmb(); /* Ensure we see queued responses up to 'rp'. */
++ case IOCTL_PRIVCMD_MMAP: {
++#define MMAP_NR_PER_PAGE (int)((PAGE_SIZE-sizeof(struct list_head))/sizeof(privcmd_mmap_entry_t))
++ privcmd_mmap_t mmapcmd;
++ privcmd_mmap_entry_t *msg;
++ privcmd_mmap_entry_t __user *p;
++ struct mm_struct *mm = current->mm;
++ struct vm_area_struct *vma;
++ unsigned long va;
++ int i, rc;
++ LIST_HEAD(pagelist);
++ struct list_head *l,*l2;
+
-+ i = np->rx.rsp_cons;
-+ work_done = 0;
-+ while ((i != rp) && (work_done < budget)) {
-+ memcpy(rx, RING_GET_RESPONSE(&np->rx, i), sizeof(*rx));
-+ memset(extras, 0, sizeof(rinfo.extras));
++ if (!is_initial_xendomain())
++ return -EPERM;
+
-+ err = xennet_get_responses(np, &rinfo, rp, &tmpq,
-+ &pages_flipped);
++ if (copy_from_user(&mmapcmd, udata, sizeof(mmapcmd)))
++ return -EFAULT;
+
-+ if (unlikely(err)) {
-+err:
-+ while ((skb = __skb_dequeue(&tmpq)))
-+ __skb_queue_tail(&errq, skb);
-+ np->stats.rx_errors++;
-+ i = np->rx.rsp_cons;
-+ continue;
-+ }
++ p = mmapcmd.entry;
++ for (i = 0; i < mmapcmd.num;) {
++ int nr = min(mmapcmd.num - i, MMAP_NR_PER_PAGE);
+
-+ skb = __skb_dequeue(&tmpq);
++ rc = -ENOMEM;
++ l = (struct list_head *) __get_free_page(GFP_KERNEL);
++ if (l == NULL)
++ goto mmap_out;
+
-+ if (extras[XEN_NETIF_EXTRA_TYPE_GSO - 1].type) {
-+ struct netif_extra_info *gso;
-+ gso = &extras[XEN_NETIF_EXTRA_TYPE_GSO - 1];
++ INIT_LIST_HEAD(l);
++ list_add_tail(l, &pagelist);
++ msg = (privcmd_mmap_entry_t*)(l + 1);
+
-+ if (unlikely(xennet_set_skb_gso(skb, gso))) {
-+ __skb_queue_head(&tmpq, skb);
-+ np->rx.rsp_cons += skb_queue_len(&tmpq);
-+ goto err;
-+ }
++ rc = -EFAULT;
++ if (copy_from_user(msg, p, nr*sizeof(*msg)))
++ goto mmap_out;
++ i += nr;
++ p += nr;
+ }
+
-+ NETFRONT_SKB_CB(skb)->page = skb_shinfo(skb)->frags[0].page;
-+ NETFRONT_SKB_CB(skb)->offset = rx->offset;
++ l = pagelist.next;
++ msg = (privcmd_mmap_entry_t*)(l + 1);
+
-+ len = rx->status;
-+ if (len > RX_COPY_THRESHOLD)
-+ len = RX_COPY_THRESHOLD;
-+ skb_put(skb, len);
++ down_write(&mm->mmap_sem);
+
-+ if (rx->status > len) {
-+ skb_shinfo(skb)->frags[0].page_offset =
-+ rx->offset + len;
-+ skb_shinfo(skb)->frags[0].size = rx->status - len;
-+ skb->data_len = rx->status - len;
-+ } else {
-+ skb_shinfo(skb)->frags[0].page = NULL;
-+ skb_shinfo(skb)->nr_frags = 0;
-+ }
++ vma = find_vma(mm, msg->va);
++ rc = -EINVAL;
++ if (!vma || (msg->va != vma->vm_start) ||
++ !privcmd_enforce_singleshot_mapping(vma))
++ goto mmap_out;
+
-+ i = xennet_fill_frags(np, skb, &tmpq);
++ va = vma->vm_start;
+
-+ /*
-+ * Truesize must approximates the size of true data plus
-+ * any supervisor overheads. Adding hypervisor overheads
-+ * has been shown to significantly reduce achievable
-+ * bandwidth with the default receive buffer size. It is
-+ * therefore not wise to account for it here.
-+ *
-+ * After alloc_skb(RX_COPY_THRESHOLD), truesize is set to
-+ * RX_COPY_THRESHOLD + the supervisor overheads. Here, we
-+ * add the size of the data pulled in xennet_fill_frags().
-+ *
-+ * We also adjust for any unused space in the main data
-+ * area by subtracting (RX_COPY_THRESHOLD - len). This is
-+ * especially important with drivers which split incoming
-+ * packets into header and data, using only 66 bytes of
-+ * the main data area (see the e1000 driver for example.)
-+ * On such systems, without this last adjustement, our
-+ * achievable receive throughout using the standard receive
-+ * buffer size was cut by 25%(!!!).
-+ */
-+ skb->truesize += skb->data_len - (RX_COPY_THRESHOLD - len);
-+ skb->len += skb->data_len;
++ i = 0;
++ list_for_each(l, &pagelist) {
++ int nr = i + min(mmapcmd.num - i, MMAP_NR_PER_PAGE);
+
-+ /*
-+ * Old backends do not assert data_validated but we
-+ * can infer it from csum_blank so test both flags.
-+ */
-+ if (rx->flags & (NETRXF_data_validated|NETRXF_csum_blank))
-+ skb->ip_summed = CHECKSUM_UNNECESSARY;
-+ else
-+ skb->ip_summed = CHECKSUM_NONE;
-+#ifdef CONFIG_XEN
-+ skb->proto_data_valid = (skb->ip_summed != CHECKSUM_NONE);
-+ skb->proto_csum_blank = !!(rx->flags & NETRXF_csum_blank);
-+#endif
-+ np->stats.rx_packets++;
-+ np->stats.rx_bytes += skb->len;
++ msg = (privcmd_mmap_entry_t*)(l + 1);
++ while (i<nr) {
+
-+ __skb_queue_tail(&rxq, skb);
++ /* Do not allow range to wrap the address space. */
++ rc = -EINVAL;
++ if ((msg->npages > (LONG_MAX >> PAGE_SHIFT)) ||
++ ((unsigned long)(msg->npages << PAGE_SHIFT) >= -va))
++ goto mmap_out;
++
++ /* Range chunks must be contiguous in va space. */
++ if ((msg->va != va) ||
++ ((msg->va+(msg->npages<<PAGE_SHIFT)) > vma->vm_end))
++ goto mmap_out;
++
++ if ((rc = direct_remap_pfn_range(
++ vma,
++ msg->va & PAGE_MASK,
++ msg->mfn,
++ msg->npages << PAGE_SHIFT,
++ vma->vm_page_prot,
++ mmapcmd.dom)) < 0)
++ goto mmap_out;
+
-+ np->rx.rsp_cons = ++i;
-+ work_done++;
-+ }
++ va += msg->npages << PAGE_SHIFT;
++ msg++;
++ i++;
++ }
++ }
+
-+ if (pages_flipped) {
-+ /* Some pages are no longer absent... */
-+ balloon_update_driver_allowance(-pages_flipped);
++ rc = 0;
+
-+ /* Do all the remapping work and M2P updates. */
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ mcl = np->rx_mcl + pages_flipped;
-+ mcl->op = __HYPERVISOR_mmu_update;
-+ mcl->args[0] = (unsigned long)np->rx_mmu;
-+ mcl->args[1] = pages_flipped;
-+ mcl->args[2] = 0;
-+ mcl->args[3] = DOMID_SELF;
-+ (void)HYPERVISOR_multicall(np->rx_mcl,
-+ pages_flipped + 1);
-+ }
++ mmap_out:
++ up_write(&mm->mmap_sem);
++ list_for_each_safe(l,l2,&pagelist)
++ free_page((unsigned long)l);
++ ret = rc;
+ }
++#undef MMAP_NR_PER_PAGE
++ break;
+
-+ while ((skb = __skb_dequeue(&errq)))
-+ kfree_skb(skb);
++ case IOCTL_PRIVCMD_MMAPBATCH: {
++#define MMAPBATCH_NR_PER_PAGE (unsigned long)((PAGE_SIZE-sizeof(struct list_head))/sizeof(unsigned long))
++ privcmd_mmapbatch_t m;
++ struct mm_struct *mm = current->mm;
++ struct vm_area_struct *vma;
++ xen_pfn_t __user *p;
++ unsigned long addr, *mfn, nr_pages;
++ int i;
++ LIST_HEAD(pagelist);
++ struct list_head *l, *l2;
+
-+ while ((skb = __skb_dequeue(&rxq)) != NULL) {
-+ struct page *page = NETFRONT_SKB_CB(skb)->page;
-+ void *vaddr = page_address(page);
-+ unsigned offset = NETFRONT_SKB_CB(skb)->offset;
++ if (!is_initial_xendomain())
++ return -EPERM;
+
-+ memcpy(skb->data, vaddr + offset, skb_headlen(skb));
++ if (copy_from_user(&m, udata, sizeof(m)))
++ return -EFAULT;
+
-+ if (page != skb_shinfo(skb)->frags[0].page)
-+ __free_page(page);
++ nr_pages = m.num;
++ if ((m.num <= 0) || (nr_pages > (LONG_MAX >> PAGE_SHIFT)))
++ return -EINVAL;
+
-+ /* Ethernet work: Delayed to here as it peeks the header. */
-+ skb->protocol = eth_type_trans(skb, dev);
++ p = m.arr;
++ for (i=0; i<nr_pages; ) {
++ int nr = min(nr_pages - i, MMAPBATCH_NR_PER_PAGE);
+
-+ /* Pass it up. */
-+ netif_receive_skb(skb);
-+ dev->last_rx = jiffies;
-+ }
++ ret = -ENOMEM;
++ l = (struct list_head *)__get_free_page(GFP_KERNEL);
++ if (l == NULL)
++ goto mmapbatch_out;
+
-+ /* If we get a callback with very few responses, reduce fill target. */
-+ /* NB. Note exponential increase, linear decrease. */
-+ if (((np->rx.req_prod_pvt - np->rx.sring->rsp_prod) >
-+ ((3*np->rx_target) / 4)) &&
-+ (--np->rx_target < np->rx_min_target))
-+ np->rx_target = np->rx_min_target;
++ INIT_LIST_HEAD(l);
++ list_add_tail(l, &pagelist);
+
-+ network_alloc_rx_buffers(dev);
++ mfn = (unsigned long*)(l + 1);
++ ret = -EFAULT;
++ if (copy_from_user(mfn, p, nr*sizeof(*mfn)))
++ goto mmapbatch_out;
+
-+ *pbudget -= work_done;
-+ dev->quota -= work_done;
++ i += nr; p+= nr;
++ }
+
-+ if (work_done < budget) {
-+ local_irq_save(flags);
++ down_write(&mm->mmap_sem);
+
-+ RING_FINAL_CHECK_FOR_RESPONSES(&np->rx, more_to_do);
-+ if (!more_to_do)
-+ __netif_rx_complete(dev);
++ vma = find_vma(mm, m.addr);
++ ret = -EINVAL;
++ if (!vma ||
++ (m.addr != vma->vm_start) ||
++ ((m.addr + (nr_pages << PAGE_SHIFT)) != vma->vm_end) ||
++ !privcmd_enforce_singleshot_mapping(vma)) {
++ up_write(&mm->mmap_sem);
++ goto mmapbatch_out;
++ }
+
-+ local_irq_restore(flags);
++ p = m.arr;
++ addr = m.addr;
++ i = 0;
++ ret = 0;
++ list_for_each(l, &pagelist) {
++ int nr = i + min(nr_pages - i, MMAPBATCH_NR_PER_PAGE);
++ mfn = (unsigned long *)(l + 1);
++
++ while (i<nr) {
++ if(direct_remap_pfn_range(vma, addr & PAGE_MASK,
++ *mfn, PAGE_SIZE,
++ vma->vm_page_prot, m.dom) < 0) {
++ *mfn |= 0xf0000000U;
++ ret++;
++ }
++ mfn++; i++; addr += PAGE_SIZE;
++ }
++ }
++
++ up_write(&mm->mmap_sem);
++ if (ret > 0) {
++ p = m.arr;
++ i = 0;
++ ret = 0;
++ list_for_each(l, &pagelist) {
++ int nr = min(nr_pages - i, MMAPBATCH_NR_PER_PAGE);
++ mfn = (unsigned long *)(l + 1);
++ if (copy_to_user(p, mfn, nr*sizeof(*mfn)))
++ ret = -EFAULT;
++ i += nr; p += nr;
++ }
++ }
++ mmapbatch_out:
++ list_for_each_safe(l,l2,&pagelist)
++ free_page((unsigned long)l);
++#undef MMAPBATCH_NR_PER_PAGE
+ }
++ break;
+
-+ spin_unlock(&np->rx_lock);
++ default:
++ ret = -EINVAL;
++ break;
++ }
+
-+ return more_to_do;
++ return ret;
+}
+
-+static void netif_release_tx_bufs(struct netfront_info *np)
++#ifndef HAVE_ARCH_PRIVCMD_MMAP
++static struct page *privcmd_nopage(struct vm_area_struct *vma,
++ unsigned long address,
++ int *type)
+{
-+ struct sk_buff *skb;
-+ int i;
++ return NOPAGE_SIGBUS;
++}
+
-+ for (i = 1; i <= NET_TX_RING_SIZE; i++) {
-+ if ((unsigned long)np->tx_skbs[i] < PAGE_OFFSET)
-+ continue;
++static struct vm_operations_struct privcmd_vm_ops = {
++ .nopage = privcmd_nopage
++};
+
-+ skb = np->tx_skbs[i];
-+ gnttab_end_foreign_access_ref(
-+ np->grant_tx_ref[i], GNTMAP_readonly);
-+ gnttab_release_grant_reference(
-+ &np->gref_tx_head, np->grant_tx_ref[i]);
-+ np->grant_tx_ref[i] = GRANT_INVALID_REF;
-+ add_id_to_freelist(np->tx_skbs, i);
-+ dev_kfree_skb_irq(skb);
-+ }
++static int privcmd_mmap(struct file * file, struct vm_area_struct * vma)
++{
++ /* Unsupported for auto-translate guests. */
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return -ENOSYS;
++
++ /* DONTCOPY is essential for Xen as copy_page_range is broken. */
++ vma->vm_flags |= VM_RESERVED | VM_IO | VM_DONTCOPY;
++ vma->vm_ops = &privcmd_vm_ops;
++ vma->vm_private_data = NULL;
++
++ return 0;
+}
+
-+static void netif_release_rx_bufs_flip(struct netfront_info *np)
++static int privcmd_enforce_singleshot_mapping(struct vm_area_struct *vma)
+{
-+ struct mmu_update *mmu = np->rx_mmu;
-+ struct multicall_entry *mcl = np->rx_mcl;
-+ struct sk_buff_head free_list;
-+ struct sk_buff *skb;
-+ unsigned long mfn;
-+ int xfer = 0, noxfer = 0, unused = 0;
-+ int id, ref, rc;
++ return (xchg(&vma->vm_private_data, (void *)1) == NULL);
++}
++#endif
+
-+ skb_queue_head_init(&free_list);
++static const struct file_operations privcmd_file_ops = {
++ .unlocked_ioctl = privcmd_ioctl,
++ .mmap = privcmd_mmap,
++};
+
-+ spin_lock_bh(&np->rx_lock);
++static int capabilities_read(char *page, char **start, off_t off,
++ int count, int *eof, void *data)
++{
++ int len = 0;
++ *page = 0;
+
-+ for (id = 0; id < NET_RX_RING_SIZE; id++) {
-+ if ((ref = np->grant_rx_ref[id]) == GRANT_INVALID_REF) {
-+ unused++;
-+ continue;
-+ }
++ if (is_initial_xendomain())
++ len = sprintf( page, "control_d\n" );
+
-+ skb = np->rx_skbs[id];
-+ mfn = gnttab_end_foreign_transfer_ref(ref);
-+ gnttab_release_grant_reference(&np->gref_rx_head, ref);
-+ np->grant_rx_ref[id] = GRANT_INVALID_REF;
-+ add_id_to_freelist(np->rx_skbs, id);
++ *eof = 1;
++ return len;
++}
+
-+ if (0 == mfn) {
-+ struct page *page = skb_shinfo(skb)->frags[0].page;
-+ balloon_release_driver_page(page);
-+ skb_shinfo(skb)->nr_frags = 0;
-+ dev_kfree_skb(skb);
-+ noxfer++;
-+ continue;
-+ }
++static int __init privcmd_init(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ /* Remap the page. */
-+ struct page *page = skb_shinfo(skb)->frags[0].page;
-+ unsigned long pfn = page_to_pfn(page);
-+ void *vaddr = page_address(page);
++ privcmd_intf = create_xen_proc_entry("privcmd", 0400);
++ if (privcmd_intf != NULL)
++ privcmd_intf->proc_fops = &privcmd_file_ops;
+
-+ MULTI_update_va_mapping(mcl, (unsigned long)vaddr,
-+ pfn_pte_ma(mfn, PAGE_KERNEL),
-+ 0);
-+ mcl++;
-+ mmu->ptr = ((maddr_t)mfn << PAGE_SHIFT)
-+ | MMU_MACHPHYS_UPDATE;
-+ mmu->val = pfn;
-+ mmu++;
++ capabilities_intf = create_xen_proc_entry("capabilities", 0400 );
++ if (capabilities_intf != NULL)
++ capabilities_intf->read_proc = capabilities_read;
+
-+ set_phys_to_machine(pfn, mfn);
-+ }
-+ __skb_queue_tail(&free_list, skb);
-+ xfer++;
-+ }
++ return 0;
++}
+
-+ DPRINTK("%s: %d xfer, %d noxfer, %d unused\n",
-+ __FUNCTION__, xfer, noxfer, unused);
++__initcall(privcmd_init);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/scsiback/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/scsiback/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,4 @@
++obj-$(CONFIG_XEN_SCSI_BACKEND) := xen-scsibk.o
+
-+ if (xfer) {
-+ /* Some pages are no longer absent... */
-+ balloon_update_driver_allowance(-xfer);
++xen-scsibk-y := interface.o scsiback.o xenbus.o translate.o emulate.o
+
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ /* Do all the remapping work and M2P updates. */
-+ mcl->op = __HYPERVISOR_mmu_update;
-+ mcl->args[0] = (unsigned long)np->rx_mmu;
-+ mcl->args[1] = mmu - np->rx_mmu;
-+ mcl->args[2] = 0;
-+ mcl->args[3] = DOMID_SELF;
-+ mcl++;
-+ rc = HYPERVISOR_multicall_check(
-+ np->rx_mcl, mcl - np->rx_mcl, NULL);
-+ BUG_ON(rc);
-+ }
-+ }
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/scsiback/common.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/scsiback/common.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,181 @@
++/*
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * Based on the blkback driver code.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ while ((skb = __skb_dequeue(&free_list)) != NULL)
-+ dev_kfree_skb(skb);
++#ifndef __SCSIIF__BACKEND__COMMON_H__
++#define __SCSIIF__BACKEND__COMMON_H__
+
-+ spin_unlock_bh(&np->rx_lock);
-+}
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/interrupt.h>
++#include <linux/slab.h>
++#include <linux/vmalloc.h>
++#include <linux/wait.h>
++#include <linux/sched.h>
++#include <linux/kthread.h>
++#include <linux/blkdev.h>
++#include <linux/list.h>
++#include <linux/kthread.h>
++#include <scsi/scsi.h>
++#include <scsi/scsi_cmnd.h>
++#include <scsi/scsi_host.h>
++#include <scsi/scsi_device.h>
++#include <scsi/scsi_dbg.h>
++#include <scsi/scsi_eh.h>
++#include <asm/io.h>
++#include <asm/setup.h>
++#include <asm/pgalloc.h>
++#include <asm/delay.h>
++#include <xen/evtchn.h>
++#include <asm/hypervisor.h>
++#include <xen/gnttab.h>
++#include <xen/driver_util.h>
++#include <xen/xenbus.h>
++#include <xen/interface/io/ring.h>
++#include <xen/interface/grant_table.h>
++#include <xen/interface/io/vscsiif.h>
+
-+static void netif_release_rx_bufs_copy(struct netfront_info *np)
-+{
-+ struct sk_buff *skb;
-+ int i, ref;
-+ int busy = 0, inuse = 0;
+
-+ spin_lock_bh(&np->rx_lock);
++#define DPRINTK(_f, _a...) \
++ pr_debug("(file=%s, line=%d) " _f, \
++ __FILE__ , __LINE__ , ## _a )
+
-+ for (i = 0; i < NET_RX_RING_SIZE; i++) {
-+ ref = np->grant_rx_ref[i];
++struct ids_tuple {
++ unsigned int hst; /* host */
++ unsigned int chn; /* channel */
++ unsigned int tgt; /* target */
++ unsigned int lun; /* LUN */
++};
+
-+ if (ref == GRANT_INVALID_REF)
-+ continue;
++struct v2p_entry {
++ struct ids_tuple v; /* translate from */
++ struct scsi_device *sdev; /* translate to */
++ struct list_head l;
++};
+
-+ inuse++;
++struct vscsibk_info {
++ struct xenbus_device *dev;
+
-+ skb = np->rx_skbs[i];
++ domid_t domid;
++ unsigned int evtchn;
++ unsigned int irq;
+
-+ if (!gnttab_end_foreign_access_ref(ref, 0))
-+ {
-+ busy++;
-+ continue;
-+ }
++ struct vscsiif_back_ring ring;
++ struct vm_struct *ring_area;
++ grant_handle_t shmem_handle;
++ grant_ref_t shmem_ref;
+
-+ gnttab_release_grant_reference(&np->gref_rx_head, ref);
-+ np->grant_rx_ref[i] = GRANT_INVALID_REF;
-+ add_id_to_freelist(np->rx_skbs, i);
++ spinlock_t ring_lock;
++ atomic_t nr_unreplied_reqs;
+
-+ skb_shinfo(skb)->nr_frags = 0;
-+ dev_kfree_skb(skb);
-+ }
++ spinlock_t v2p_lock;
++ struct list_head v2p_entry_lists;
+
-+ if (busy)
-+ DPRINTK("%s: Unable to release %d of %d inuse grant references out of %ld total.\n",
-+ __FUNCTION__, busy, inuse, NET_RX_RING_SIZE);
++ struct task_struct *kthread;
++ wait_queue_head_t waiting_to_free;
++ wait_queue_head_t wq;
++ unsigned int waiting_reqs;
++ struct page **mmap_pages;
+
-+ spin_unlock_bh(&np->rx_lock);
-+}
++};
+
-+static int network_close(struct net_device *dev)
-+{
-+ struct netfront_info *np = netdev_priv(dev);
-+ netif_stop_queue(np->netdev);
-+ return 0;
-+}
++typedef struct {
++ unsigned char act;
++ struct vscsibk_info *info;
++ struct scsi_device *sdev;
+
++ uint16_t rqid;
++
++ uint8_t nr_segments;
++ uint8_t cmnd[VSCSIIF_MAX_COMMAND_SIZE];
++ uint8_t cmd_len;
+
-+static struct net_device_stats *network_get_stats(struct net_device *dev)
-+{
-+ struct netfront_info *np = netdev_priv(dev);
-+ return &np->stats;
-+}
++ uint8_t sc_data_direction;
++ uint16_t timeout_per_command;
++
++ uint32_t request_bufflen;
++ struct scatterlist *sgl;
++ grant_ref_t gref[VSCSIIF_SG_TABLESIZE];
++
++ int32_t rslt;
++ uint32_t resid;
++ uint8_t sense_buffer[VSCSIIF_SENSE_BUFFERSIZE];
+
-+static int xennet_change_mtu(struct net_device *dev, int mtu)
-+{
-+ int max = xennet_can_sg(dev) ? 65535 - ETH_HLEN : ETH_DATA_LEN;
++ struct list_head free_list;
++} pending_req_t;
+
-+ if (mtu > max)
-+ return -EINVAL;
-+ dev->mtu = mtu;
-+ return 0;
-+}
+
-+static int xennet_set_sg(struct net_device *dev, u32 data)
-+{
-+ if (data) {
-+ struct netfront_info *np = netdev_priv(dev);
-+ int val;
+
-+ if (xenbus_scanf(XBT_NIL, np->xbdev->otherend, "feature-sg",
-+ "%d", &val) < 0)
-+ val = 0;
-+ if (!val)
-+ return -ENOSYS;
-+ } else if (dev->mtu > ETH_DATA_LEN)
-+ dev->mtu = ETH_DATA_LEN;
++#define scsiback_get(_b) (atomic_inc(&(_b)->nr_unreplied_reqs))
++#define scsiback_put(_b) \
++ do { \
++ if (atomic_dec_and_test(&(_b)->nr_unreplied_reqs)) \
++ wake_up(&(_b)->waiting_to_free);\
++ } while (0)
+
-+ return ethtool_op_set_sg(dev, data);
-+}
++#define VSCSIIF_TIMEOUT (900*HZ)
+
-+static int xennet_set_tso(struct net_device *dev, u32 data)
-+{
-+#ifdef HAVE_TSO
-+ if (data) {
-+ struct netfront_info *np = netdev_priv(dev);
-+ int val;
+
-+ if (xenbus_scanf(XBT_NIL, np->xbdev->otherend,
-+ "feature-gso-tcpv4", "%d", &val) < 0)
-+ val = 0;
-+ if (!val)
-+ return -ENOSYS;
-+ }
++irqreturn_t scsiback_intr(int, void *, struct pt_regs *);
++int scsiback_init_sring(struct vscsibk_info *info,
++ unsigned long ring_ref, unsigned int evtchn);
++int scsiback_schedule(void *data);
+
-+ return ethtool_op_set_tso(dev, data);
-+#else
-+ return -ENOSYS;
-+#endif
-+}
+
-+static void xennet_set_features(struct net_device *dev)
-+{
-+ dev_disable_gso_features(dev);
-+ xennet_set_sg(dev, 0);
++struct vscsibk_info *vscsibk_info_alloc(domid_t domid);
++void scsiback_free(struct vscsibk_info *info);
++void scsiback_disconnect(struct vscsibk_info *info);
++int __init scsiback_interface_init(void);
++void scsiback_interface_exit(void);
++int scsiback_xenbus_init(void);
++void scsiback_xenbus_unregister(void);
+
-+ /* We need checksum offload to enable scatter/gather and TSO. */
-+ if (!(dev->features & NETIF_F_IP_CSUM))
-+ return;
++void scsiback_init_translation_table(struct vscsibk_info *info);
+
-+ if (xennet_set_sg(dev, 1))
-+ return;
++int scsiback_add_translation_entry(struct vscsibk_info *info,
++ struct scsi_device *sdev, struct ids_tuple *v);
+
-+ /* Before 2.6.9 TSO seems to be unreliable so do not enable it
-+ * on older kernels.
-+ */
-+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,9)
-+ xennet_set_tso(dev, 1);
-+#endif
++int scsiback_del_translation_entry(struct vscsibk_info *info,
++ struct ids_tuple *v);
++struct scsi_device *scsiback_do_translation(struct vscsibk_info *info,
++ struct ids_tuple *v);
++void scsiback_release_translation_entry(struct vscsibk_info *info);
+
-+}
+
-+static int network_connect(struct net_device *dev)
-+{
-+ struct netfront_info *np = netdev_priv(dev);
-+ int i, requeue_idx, err;
-+ struct sk_buff *skb;
-+ grant_ref_t ref;
-+ netif_rx_request_t *req;
-+ unsigned int feature_rx_copy, feature_rx_flip;
++void scsiback_cmd_exec(pending_req_t *pending_req);
++void scsiback_do_resp_with_sense(char *sense_buffer, int32_t result,
++ uint32_t resid, pending_req_t *pending_req);
++void scsiback_fast_flush_area(pending_req_t *req);
+
-+ err = xenbus_scanf(XBT_NIL, np->xbdev->otherend,
-+ "feature-rx-copy", "%u", &feature_rx_copy);
-+ if (err != 1)
-+ feature_rx_copy = 0;
-+ err = xenbus_scanf(XBT_NIL, np->xbdev->otherend,
-+ "feature-rx-flip", "%u", &feature_rx_flip);
-+ if (err != 1)
-+ feature_rx_flip = 1;
++void scsiback_rsp_emulation(pending_req_t *pending_req);
++void scsiback_req_emulation_or_cmdexec(pending_req_t *pending_req);
++void scsiback_emulation_init(void);
+
-+ /*
-+ * Copy packets on receive path if:
-+ * (a) This was requested by user, and the backend supports it; or
-+ * (b) Flipping was requested, but this is unsupported by the backend.
-+ */
-+ np->copying_receiver = ((MODPARM_rx_copy && feature_rx_copy) ||
-+ (MODPARM_rx_flip && !feature_rx_flip));
+
-+ err = talk_to_backend(np->xbdev, np);
-+ if (err)
-+ return err;
++#endif /* __SCSIIF__BACKEND__COMMON_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/scsiback/emulate.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/scsiback/emulate.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,454 @@
++/*
++ * Xen SCSI backend driver
++ *
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ xennet_set_features(dev);
++#include <scsi/scsi.h>
++#include <scsi/scsi_cmnd.h>
++#include <scsi/scsi_device.h>
++#include "common.h"
+
-+ DPRINTK("device %s has %sing receive path.\n",
-+ dev->name, np->copying_receiver ? "copy" : "flipp");
++/* Following SCSI commands are not defined in scsi/scsi.h */
++#define EXTENDED_COPY 0x83 /* EXTENDED COPY command */
++#define REPORT_ALIASES 0xa3 /* REPORT ALIASES command */
++#define CHANGE_ALIASES 0xa4 /* CHANGE ALIASES command */
++#define SET_PRIORITY 0xa4 /* SET PRIORITY command */
+
-+ spin_lock_bh(&np->rx_lock);
-+ spin_lock_irq(&np->tx_lock);
+
-+ /*
-+ * Recovery procedure:
-+ * NB. Freelist index entries are always going to be less than
-+ * PAGE_OFFSET, whereas pointers to skbs will always be equal or
-+ * greater than PAGE_OFFSET: we use this property to distinguish
-+ * them.
-+ */
++/*
++ The bitmap in order to control emulation.
++ (Bit 3 to 7 are reserved for future use.)
++*/
++#define VSCSIIF_NEED_CMD_EXEC 0x01 /* If this bit is set, cmd exec */
++ /* is required. */
++#define VSCSIIF_NEED_EMULATE_REQBUF 0x02 /* If this bit is set, need */
++ /* emulation reqest buff before */
++ /* cmd exec. */
++#define VSCSIIF_NEED_EMULATE_RSPBUF 0x04 /* If this bit is set, need */
++ /* emulation resp buff after */
++ /* cmd exec. */
++
++/* Additional Sense Code (ASC) used */
++#define NO_ADDITIONAL_SENSE 0x0
++#define LOGICAL_UNIT_NOT_READY 0x4
++#define UNRECOVERED_READ_ERR 0x11
++#define PARAMETER_LIST_LENGTH_ERR 0x1a
++#define INVALID_OPCODE 0x20
++#define ADDR_OUT_OF_RANGE 0x21
++#define INVALID_FIELD_IN_CDB 0x24
++#define INVALID_FIELD_IN_PARAM_LIST 0x26
++#define POWERON_RESET 0x29
++#define SAVING_PARAMS_UNSUP 0x39
++#define THRESHOLD_EXCEEDED 0x5d
++#define LOW_POWER_COND_ON 0x5e
+
-+ /* Step 1: Discard all pending TX packet fragments. */
-+ netif_release_tx_bufs(np);
+
-+ /* Step 2: Rebuild the RX buffer freelist and the RX ring itself. */
-+ for (requeue_idx = 0, i = 0; i < NET_RX_RING_SIZE; i++) {
-+ if (!np->rx_skbs[i])
-+ continue;
+
-+ skb = np->rx_skbs[requeue_idx] = xennet_get_rx_skb(np, i);
-+ ref = np->grant_rx_ref[requeue_idx] = xennet_get_rx_ref(np, i);
-+ req = RING_GET_REQUEST(&np->rx, requeue_idx);
++/* Number os SCSI op_code */
++#define VSCSI_MAX_SCSI_OP_CODE 256
++static unsigned char bitmap[VSCSI_MAX_SCSI_OP_CODE];
+
-+ if (!np->copying_receiver) {
-+ gnttab_grant_foreign_transfer_ref(
-+ ref, np->xbdev->otherend_id,
-+ page_to_pfn(skb_shinfo(skb)->frags->page));
-+ } else {
-+ gnttab_grant_foreign_access_ref(
-+ ref, np->xbdev->otherend_id,
-+ pfn_to_mfn(page_to_pfn(skb_shinfo(skb)->
-+ frags->page)),
-+ 0);
-+ }
-+ req->gref = ref;
-+ req->id = requeue_idx;
+
-+ requeue_idx++;
-+ }
+
-+ np->rx.req_prod_pvt = requeue_idx;
++/*
++ Emulation routines for each SCSI op_code.
++*/
++static void (*pre_function[VSCSI_MAX_SCSI_OP_CODE])(pending_req_t *, void *);
++static void (*post_function[VSCSI_MAX_SCSI_OP_CODE])(pending_req_t *, void *);
+
-+ /*
-+ * Step 3: All public and private state should now be sane. Get
-+ * ready to start sending and receiving packets and give the driver
-+ * domain a kick because we've probably just requeued some
-+ * packets.
-+ */
-+ netfront_carrier_on(np);
-+ notify_remote_via_irq(np->irq);
-+ network_tx_buf_gc(dev);
-+ network_alloc_rx_buffers(dev);
+
-+ spin_unlock_irq(&np->tx_lock);
-+ spin_unlock_bh(&np->rx_lock);
++static const int check_condition_result =
++ (DRIVER_SENSE << 24) | SAM_STAT_CHECK_CONDITION;
+
-+ return 0;
++static void scsiback_mk_sense_buffer(uint8_t *data, uint8_t key,
++ uint8_t asc, uint8_t asq)
++{
++ data[0] = 0x70; /* fixed, current */
++ data[2] = key;
++ data[7] = 0xa; /* implies 18 byte sense buffer */
++ data[12] = asc;
++ data[13] = asq;
+}
+
-+static void netif_uninit(struct net_device *dev)
++static void resp_not_supported_cmd(pending_req_t *pending_req, void *data)
+{
-+ struct netfront_info *np = netdev_priv(dev);
-+ netif_release_tx_bufs(np);
-+ if (np->copying_receiver)
-+ netif_release_rx_bufs_copy(np);
-+ else
-+ netif_release_rx_bufs_flip(np);
-+ gnttab_free_grant_references(np->gref_tx_head);
-+ gnttab_free_grant_references(np->gref_rx_head);
++ scsiback_mk_sense_buffer(pending_req->sense_buffer, ILLEGAL_REQUEST,
++ INVALID_OPCODE, 0);
++ pending_req->resid = 0;
++ pending_req->rslt = check_condition_result;
+}
+
-+static struct ethtool_ops network_ethtool_ops =
-+{
-+ .get_tx_csum = ethtool_op_get_tx_csum,
-+ .set_tx_csum = ethtool_op_set_tx_csum,
-+ .get_sg = ethtool_op_get_sg,
-+ .set_sg = xennet_set_sg,
-+ .get_tso = ethtool_op_get_tso,
-+ .set_tso = xennet_set_tso,
-+ .get_link = ethtool_op_get_link,
-+};
+
-+#ifdef CONFIG_SYSFS
-+static ssize_t show_rxbuf_min(struct class_device *cd, char *buf)
++static int __copy_to_sg(struct scatterlist *sg, unsigned int nr_sg,
++ void *buf, unsigned int buflen)
+{
-+ struct net_device *netdev = container_of(cd, struct net_device,
-+ class_dev);
-+ struct netfront_info *info = netdev_priv(netdev);
++ void *from = buf;
++ void *to;
++ unsigned int from_rest = buflen;
++ unsigned int to_capa;
++ unsigned int copy_size = 0;
++ unsigned int i;
++ unsigned long pfn;
+
-+ return sprintf(buf, "%u\n", info->rx_min_target);
++ for (i = 0; i < nr_sg; i++) {
++ if (sg->page == NULL) {
++ printk(KERN_WARNING "%s: inconsistent length field in "
++ "scatterlist\n", __FUNCTION__);
++ return -ENOMEM;
++ }
++
++ to_capa = sg->length;
++ copy_size = min_t(unsigned int, to_capa, from_rest);
++
++ pfn = page_to_pfn(sg->page);
++ to = pfn_to_kaddr(pfn) + (sg->offset);
++ memcpy(to, from, copy_size);
++
++ from_rest -= copy_size;
++ if (from_rest == 0) {
++ return 0;
++ }
++
++ sg++;
++ from += copy_size;
++ }
++
++ printk(KERN_WARNING "%s: no space in scatterlist\n",
++ __FUNCTION__);
++ return -ENOMEM;
+}
+
-+static ssize_t store_rxbuf_min(struct class_device *cd,
-+ const char *buf, size_t len)
++static int __copy_from_sg(struct scatterlist *sg, unsigned int nr_sg,
++ void *buf, unsigned int buflen)
+{
-+ struct net_device *netdev = container_of(cd, struct net_device,
-+ class_dev);
-+ struct netfront_info *np = netdev_priv(netdev);
-+ char *endp;
-+ unsigned long target;
-+
-+ if (!capable(CAP_NET_ADMIN))
-+ return -EPERM;
++ void *from;
++ void *to = buf;
++ unsigned int from_rest;
++ unsigned int to_capa = buflen;
++ unsigned int copy_size;
++ unsigned int i;
++ unsigned long pfn;
+
-+ target = simple_strtoul(buf, &endp, 0);
-+ if (endp == buf)
-+ return -EBADMSG;
++ for (i = 0; i < nr_sg; i++) {
++ if (sg->page == NULL) {
++ printk(KERN_WARNING "%s: inconsistent length field in "
++ "scatterlist\n", __FUNCTION__);
++ return -ENOMEM;
++ }
+
-+ if (target < RX_MIN_TARGET)
-+ target = RX_MIN_TARGET;
-+ if (target > RX_MAX_TARGET)
-+ target = RX_MAX_TARGET;
++ from_rest = sg->length;
++ if ((from_rest > 0) && (to_capa < from_rest)) {
++ printk(KERN_WARNING
++ "%s: no space in destination buffer\n",
++ __FUNCTION__);
++ return -ENOMEM;
++ }
++ copy_size = from_rest;
+
-+ spin_lock_bh(&np->rx_lock);
-+ if (target > np->rx_max_target)
-+ np->rx_max_target = target;
-+ np->rx_min_target = target;
-+ if (target > np->rx_target)
-+ np->rx_target = target;
++ pfn = page_to_pfn(sg->page);
++ from = pfn_to_kaddr(pfn) + (sg->offset);
++ memcpy(to, from, copy_size);
+
-+ network_alloc_rx_buffers(netdev);
++ to_capa -= copy_size;
++
++ sg++;
++ to += copy_size;
++ }
+
-+ spin_unlock_bh(&np->rx_lock);
-+ return len;
++ return 0;
+}
+
-+static ssize_t show_rxbuf_max(struct class_device *cd, char *buf)
++static int __nr_luns_under_host(struct vscsibk_info *info)
+{
-+ struct net_device *netdev = container_of(cd, struct net_device,
-+ class_dev);
-+ struct netfront_info *info = netdev_priv(netdev);
++ struct v2p_entry *entry;
++ struct list_head *head = &(info->v2p_entry_lists);
++ unsigned long flags;
++ int lun_cnt = 0;
+
-+ return sprintf(buf, "%u\n", info->rx_max_target);
++ spin_lock_irqsave(&info->v2p_lock, flags);
++ list_for_each_entry(entry, head, l) {
++ lun_cnt++;
++ }
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++
++ return (lun_cnt);
+}
+
-+static ssize_t store_rxbuf_max(struct class_device *cd,
-+ const char *buf, size_t len)
++
++/* REPORT LUNS Define*/
++#define VSCSI_REPORT_LUNS_HEADER 8
++#define VSCSI_REPORT_LUNS_RETRY 3
++
++/* quoted scsi_debug.c/resp_report_luns() */
++static void __report_luns(pending_req_t *pending_req, void *data)
+{
-+ struct net_device *netdev = container_of(cd, struct net_device,
-+ class_dev);
-+ struct netfront_info *np = netdev_priv(netdev);
-+ char *endp;
-+ unsigned long target;
++ struct vscsibk_info *info = pending_req->info;
++ unsigned int channel = pending_req->sdev->channel;
++ unsigned int target = pending_req->sdev->id;
++ unsigned int nr_seg = pending_req->nr_segments;
++ unsigned char *cmd = (unsigned char *)pending_req->cmnd;
++
++ unsigned char *buff = NULL;
++ unsigned char alloc_len;
++ unsigned int alloc_luns = 0;
++ unsigned int req_bufflen = 0;
++ unsigned int actual_len = 0;
++ unsigned int retry_cnt = 0;
++ int select_report = (int)cmd[2];
++ int i, lun_cnt = 0, lun, upper, err = 0;
++
++ struct v2p_entry *entry;
++ struct list_head *head = &(info->v2p_entry_lists);
++ unsigned long flags;
++
++ struct scsi_lun *one_lun;
+
-+ if (!capable(CAP_NET_ADMIN))
-+ return -EPERM;
++ req_bufflen = cmd[9] + (cmd[8] << 8) + (cmd[7] << 16) + (cmd[6] << 24);
++ if ((req_bufflen < 4) || (select_report != 0))
++ goto fail;
+
-+ target = simple_strtoul(buf, &endp, 0);
-+ if (endp == buf)
-+ return -EBADMSG;
++ alloc_luns = __nr_luns_under_host(info);
++ alloc_len = sizeof(struct scsi_lun) * alloc_luns
++ + VSCSI_REPORT_LUNS_HEADER;
++retry:
++ if ((buff = kmalloc(alloc_len, GFP_KERNEL)) == NULL) {
++ printk(KERN_ERR "scsiback:%s kmalloc err\n", __FUNCTION__);
++ goto fail;
++ }
+
-+ if (target < RX_MIN_TARGET)
-+ target = RX_MIN_TARGET;
-+ if (target > RX_MAX_TARGET)
-+ target = RX_MAX_TARGET;
++ memset(buff, 0, alloc_len);
+
-+ spin_lock_bh(&np->rx_lock);
-+ if (target < np->rx_min_target)
-+ np->rx_min_target = target;
-+ np->rx_max_target = target;
-+ if (target < np->rx_target)
-+ np->rx_target = target;
++ one_lun = (struct scsi_lun *) &buff[8];
++ spin_lock_irqsave(&info->v2p_lock, flags);
++ list_for_each_entry(entry, head, l) {
++ if ((entry->v.chn == channel) &&
++ (entry->v.tgt == target)) {
++
++ /* check overflow */
++ if (lun_cnt >= alloc_luns) {
++ spin_unlock_irqrestore(&info->v2p_lock,
++ flags);
++
++ if (retry_cnt < VSCSI_REPORT_LUNS_RETRY) {
++ retry_cnt++;
++ if (buff)
++ kfree(buff);
++ goto retry;
++ }
+
-+ network_alloc_rx_buffers(netdev);
++ goto fail;
++ }
+
-+ spin_unlock_bh(&np->rx_lock);
-+ return len;
-+}
++ lun = entry->v.lun;
++ upper = (lun >> 8) & 0x3f;
++ if (upper)
++ one_lun[lun_cnt].scsi_lun[0] = upper;
++ one_lun[lun_cnt].scsi_lun[1] = lun & 0xff;
++ lun_cnt++;
++ }
++ }
+
-+static ssize_t show_rxbuf_cur(struct class_device *cd, char *buf)
-+{
-+ struct net_device *netdev = container_of(cd, struct net_device,
-+ class_dev);
-+ struct netfront_info *info = netdev_priv(netdev);
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
+
-+ return sprintf(buf, "%u\n", info->rx_target);
++ buff[2] = ((sizeof(struct scsi_lun) * lun_cnt) >> 8) & 0xff;
++ buff[3] = (sizeof(struct scsi_lun) * lun_cnt) & 0xff;
++
++ actual_len = lun_cnt * sizeof(struct scsi_lun)
++ + VSCSI_REPORT_LUNS_HEADER;
++ req_bufflen = 0;
++ for (i = 0; i < nr_seg; i++)
++ req_bufflen += pending_req->sgl[i].length;
++
++ err = __copy_to_sg(pending_req->sgl, nr_seg, buff,
++ min(req_bufflen, actual_len));
++ if (err)
++ goto fail;
++
++ memset(pending_req->sense_buffer, 0, VSCSIIF_SENSE_BUFFERSIZE);
++ pending_req->rslt = 0x00;
++ pending_req->resid = req_bufflen - min(req_bufflen, actual_len);
++
++ kfree(buff);
++ return;
++
++fail:
++ scsiback_mk_sense_buffer(pending_req->sense_buffer, ILLEGAL_REQUEST,
++ INVALID_FIELD_IN_CDB, 0);
++ pending_req->rslt = check_condition_result;
++ pending_req->resid = 0;
++ if (buff)
++ kfree(buff);
++ return;
+}
+
-+static const struct class_device_attribute xennet_attrs[] = {
-+ __ATTR(rxbuf_min, S_IRUGO|S_IWUSR, show_rxbuf_min, store_rxbuf_min),
-+ __ATTR(rxbuf_max, S_IRUGO|S_IWUSR, show_rxbuf_max, store_rxbuf_max),
-+ __ATTR(rxbuf_cur, S_IRUGO, show_rxbuf_cur, NULL),
-+};
+
-+static int xennet_sysfs_addif(struct net_device *netdev)
++
++int __pre_do_emulation(pending_req_t *pending_req, void *data)
+{
-+ int i;
-+ int error = 0;
++ uint8_t op_code = pending_req->cmnd[0];
+
-+ for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++) {
-+ error = class_device_create_file(&netdev->class_dev,
-+ &xennet_attrs[i]);
-+ if (error)
-+ goto fail;
++ if ((bitmap[op_code] & VSCSIIF_NEED_EMULATE_REQBUF) &&
++ pre_function[op_code] != NULL) {
++ pre_function[op_code](pending_req, data);
+ }
-+ return 0;
+
-+ fail:
-+ while (--i >= 0)
-+ class_device_remove_file(&netdev->class_dev,
-+ &xennet_attrs[i]);
-+ return error;
++ /*
++ 0: no need for native driver call, so should return immediately.
++ 1: non emulation or should call native driver
++ after modifing the request buffer.
++ */
++ return !!(bitmap[op_code] & VSCSIIF_NEED_CMD_EXEC);
+}
+
-+static void xennet_sysfs_delif(struct net_device *netdev)
++void scsiback_rsp_emulation(pending_req_t *pending_req)
+{
-+ int i;
++ uint8_t op_code = pending_req->cmnd[0];
+
-+ for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++) {
-+ class_device_remove_file(&netdev->class_dev,
-+ &xennet_attrs[i]);
++ if ((bitmap[op_code] & VSCSIIF_NEED_EMULATE_RSPBUF) &&
++ post_function[op_code] != NULL) {
++ post_function[op_code](pending_req, NULL);
+ }
-+}
+
-+#endif /* CONFIG_SYSFS */
++ return;
++}
+
+
-+/*
-+ * Nothing to do here. Virtual interface is point-to-point and the
-+ * physical interface is probably promiscuous anyway.
-+ */
-+static void network_set_multicast_list(struct net_device *dev)
++void scsiback_req_emulation_or_cmdexec(pending_req_t *pending_req)
+{
++ if (__pre_do_emulation(pending_req, NULL)) {
++ scsiback_cmd_exec(pending_req);
++ }
++ else {
++ scsiback_fast_flush_area(pending_req);
++ scsiback_do_resp_with_sense(pending_req->sense_buffer,
++ pending_req->rslt, pending_req->resid, pending_req);
++ }
+}
+
-+static struct net_device * __devinit create_netdev(struct xenbus_device *dev)
++
++/*
++ Following are not customizable functions.
++*/
++void scsiback_emulation_init(void)
+{
-+ int i, err = 0;
-+ struct net_device *netdev = NULL;
-+ struct netfront_info *np = NULL;
++ int i;
+
-+ netdev = alloc_etherdev(sizeof(struct netfront_info));
-+ if (!netdev) {
-+ printk(KERN_WARNING "%s> alloc_etherdev failed.\n",
-+ __FUNCTION__);
-+ return ERR_PTR(-ENOMEM);
++ /* Initialize to default state */
++ for (i = 0; i < VSCSI_MAX_SCSI_OP_CODE; i++) {
++ bitmap[i] = (VSCSIIF_NEED_EMULATE_REQBUF |
++ VSCSIIF_NEED_EMULATE_RSPBUF);
++ pre_function[i] = resp_not_supported_cmd;
++ post_function[i] = NULL;
++ /* means,
++ - no need for pre-emulation
++ - no need for post-emulation
++ - call native driver
++ */
+ }
+
-+ np = netdev_priv(netdev);
-+ np->xbdev = dev;
++ /*
++ Register appropriate functions below as you need.
++ (See scsi/scsi.h for definition of SCSI op_code.)
++ */
+
-+ spin_lock_init(&np->tx_lock);
-+ spin_lock_init(&np->rx_lock);
++ /*
++ This command is Non emulation.
++ */
++ bitmap[TEST_UNIT_READY] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[TEST_UNIT_READY] = NULL;
++ post_function[TEST_UNIT_READY] = NULL;
+
-+ skb_queue_head_init(&np->rx_batch);
-+ np->rx_target = RX_DFL_MIN_TARGET;
-+ np->rx_min_target = RX_DFL_MIN_TARGET;
-+ np->rx_max_target = RX_MAX_TARGET;
++ bitmap[REZERO_UNIT] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[REZERO_UNIT] = NULL;
++ post_function[REZERO_UNIT] = NULL;
+
-+ init_timer(&np->rx_refill_timer);
-+ np->rx_refill_timer.data = (unsigned long)netdev;
-+ np->rx_refill_timer.function = rx_refill_timeout;
++ bitmap[REQUEST_SENSE] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[REQUEST_SENSE] = NULL;
++ post_function[REQUEST_SENSE] = NULL;
+
-+ /* Initialise {tx,rx}_skbs as a free chain containing every entry. */
-+ for (i = 0; i <= NET_TX_RING_SIZE; i++) {
-+ np->tx_skbs[i] = (void *)((unsigned long) i+1);
-+ np->grant_tx_ref[i] = GRANT_INVALID_REF;
-+ }
++ bitmap[FORMAT_UNIT] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[FORMAT_UNIT] = NULL;
++ post_function[FORMAT_UNIT] = NULL;
+
-+ for (i = 0; i < NET_RX_RING_SIZE; i++) {
-+ np->rx_skbs[i] = NULL;
-+ np->grant_rx_ref[i] = GRANT_INVALID_REF;
-+ }
++ bitmap[READ_BLOCK_LIMITS] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[READ_BLOCK_LIMITS] = NULL;
++ post_function[READ_BLOCK_LIMITS] = NULL;
+
-+ /* A grant for every tx ring slot */
-+ if (gnttab_alloc_grant_references(TX_MAX_TARGET,
-+ &np->gref_tx_head) < 0) {
-+ printk(KERN_ALERT "#### netfront can't alloc tx grant refs\n");
-+ err = -ENOMEM;
-+ goto exit;
-+ }
-+ /* A grant for every rx ring slot */
-+ if (gnttab_alloc_grant_references(RX_MAX_TARGET,
-+ &np->gref_rx_head) < 0) {
-+ printk(KERN_ALERT "#### netfront can't alloc rx grant refs\n");
-+ err = -ENOMEM;
-+ goto exit_free_tx;
-+ }
++ bitmap[READ_6] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[READ_6] = NULL;
++ post_function[READ_6] = NULL;
+
-+ netdev->open = network_open;
-+ netdev->hard_start_xmit = network_start_xmit;
-+ netdev->stop = network_close;
-+ netdev->get_stats = network_get_stats;
-+ netdev->poll = netif_poll;
-+ netdev->set_multicast_list = network_set_multicast_list;
-+ netdev->uninit = netif_uninit;
-+ netdev->change_mtu = xennet_change_mtu;
-+ netdev->weight = 64;
-+ netdev->features = NETIF_F_IP_CSUM;
++ bitmap[WRITE_6] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[WRITE_6] = NULL;
++ post_function[WRITE_6] = NULL;
+
-+ SET_ETHTOOL_OPS(netdev, &network_ethtool_ops);
-+ SET_MODULE_OWNER(netdev);
-+ SET_NETDEV_DEV(netdev, &dev->dev);
++ bitmap[WRITE_FILEMARKS] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[WRITE_FILEMARKS] = NULL;
++ post_function[WRITE_FILEMARKS] = NULL;
+
-+ np->netdev = netdev;
++ bitmap[SPACE] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[SPACE] = NULL;
++ post_function[SPACE] = NULL;
+
-+ netfront_carrier_off(np);
++ bitmap[INQUIRY] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[INQUIRY] = NULL;
++ post_function[INQUIRY] = NULL;
+
-+ return netdev;
++ bitmap[ERASE] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[ERASE] = NULL;
++ post_function[ERASE] = NULL;
+
-+ exit_free_tx:
-+ gnttab_free_grant_references(np->gref_tx_head);
-+ exit:
-+ free_netdev(netdev);
-+ return ERR_PTR(err);
-+}
++ bitmap[MODE_SENSE] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[MODE_SENSE] = NULL;
++ post_function[MODE_SENSE] = NULL;
+
++ bitmap[SEND_DIAGNOSTIC] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[SEND_DIAGNOSTIC] = NULL;
++ post_function[SEND_DIAGNOSTIC] = NULL;
++
++ bitmap[READ_CAPACITY] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[READ_CAPACITY] = NULL;
++ post_function[READ_CAPACITY] = NULL;
++
++ bitmap[READ_10] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[READ_10] = NULL;
++ post_function[READ_10] = NULL;
++
++ bitmap[WRITE_10] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[WRITE_10] = NULL;
++ post_function[WRITE_10] = NULL;
++
++ /*
++ This command is Full emulation.
++ */
++ pre_function[REPORT_LUNS] = __report_luns;
++ bitmap[REPORT_LUNS] = (VSCSIIF_NEED_EMULATE_REQBUF |
++ VSCSIIF_NEED_EMULATE_RSPBUF);
++
++ return;
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/scsiback/interface.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/scsiback/interface.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,182 @@
+/*
-+ * We use this notifier to send out a fake ARP reply to reset switches and
-+ * router ARP caches when an IP interface is brought up on a VIF.
++ * interface management.
++ *
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * Based on the blkback driver code.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
-+static int
-+inetdev_notify(struct notifier_block *this, unsigned long event, void *ptr)
-+{
-+ struct in_ifaddr *ifa = (struct in_ifaddr *)ptr;
-+ struct net_device *dev = ifa->ifa_dev->dev;
+
-+ /* UP event and is it one of our devices? */
-+ if (event == NETDEV_UP && dev->open == network_open)
-+ (void)send_fake_arp(dev);
++#include <scsi/scsi.h>
++#include <scsi/scsi_host.h>
++#include <scsi/scsi_device.h>
++#include "common.h"
+
-+ return NOTIFY_DONE;
-+}
++#include <xen/evtchn.h>
++#include <linux/kthread.h>
+
+
-+static void netif_disconnect_backend(struct netfront_info *info)
++static kmem_cache_t *scsiback_cachep;
++
++struct vscsibk_info *vscsibk_info_alloc(domid_t domid)
+{
-+ /* Stop old i/f to prevent errors whilst we rebuild the state. */
-+ spin_lock_bh(&info->rx_lock);
-+ spin_lock_irq(&info->tx_lock);
-+ netfront_carrier_off(info);
-+ spin_unlock_irq(&info->tx_lock);
-+ spin_unlock_bh(&info->rx_lock);
++ struct vscsibk_info *info;
+
-+ if (info->irq)
-+ unbind_from_irqhandler(info->irq, info->netdev);
-+ info->irq = 0;
++ info = kmem_cache_alloc(scsiback_cachep, GFP_KERNEL);
++ if (!info)
++ return ERR_PTR(-ENOMEM);
+
-+ end_access(info->tx_ring_ref, info->tx.sring);
-+ end_access(info->rx_ring_ref, info->rx.sring);
-+ info->tx_ring_ref = GRANT_INVALID_REF;
-+ info->rx_ring_ref = GRANT_INVALID_REF;
-+ info->tx.sring = NULL;
-+ info->rx.sring = NULL;
++ memset(info, 0, sizeof(*info));
++ info->domid = domid;
++ spin_lock_init(&info->ring_lock);
++ atomic_set(&info->nr_unreplied_reqs, 0);
++ init_waitqueue_head(&info->wq);
++ init_waitqueue_head(&info->waiting_to_free);
++
++ return info;
+}
+
++static int map_frontend_page( struct vscsibk_info *info,
++ unsigned long ring_ref)
++{
++ struct gnttab_map_grant_ref op;
++ int err;
++
++ gnttab_set_map_op(&op, (unsigned long)info->ring_area->addr,
++ GNTMAP_host_map, ring_ref,
++ info->domid);
+
-+static void end_access(int ref, void *page)
++ err = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1);
++ BUG_ON(err);
++
++ if (op.status) {
++ printk(KERN_ERR "scsiback: Grant table operation failure !\n");
++ return op.status;
++ }
++
++ info->shmem_ref = ring_ref;
++ info->shmem_handle = op.handle;
++
++ return (GNTST_okay);
++}
++
++static void unmap_frontend_page(struct vscsibk_info *info)
+{
-+ if (ref != GRANT_INVALID_REF)
-+ gnttab_end_foreign_access(ref, 0, (unsigned long)page);
++ struct gnttab_unmap_grant_ref op;
++ int err;
++
++ gnttab_set_unmap_op(&op, (unsigned long)info->ring_area->addr,
++ GNTMAP_host_map, info->shmem_handle);
++
++ err = HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1);
++ BUG_ON(err);
++
+}
+
++int scsiback_init_sring(struct vscsibk_info *info,
++ unsigned long ring_ref, unsigned int evtchn)
++{
++ struct vscsiif_sring *sring;
++ int err;
+
-+/* ** Driver registration ** */
++ if (info->irq) {
++ printk(KERN_ERR "scsiback: Already connected through?\n");
++ return -1;
++ }
+
++ info->ring_area = alloc_vm_area(PAGE_SIZE);
++ if (!info)
++ return -ENOMEM;
+
-+static struct xenbus_device_id netfront_ids[] = {
-+ { "vif" },
-+ { "" }
-+};
-+MODULE_ALIAS("xen:vif");
++ err = map_frontend_page(info, ring_ref);
++ if (err)
++ goto free_vm;
+
++ sring = (struct vscsiif_sring *) info->ring_area->addr;
++ BACK_RING_INIT(&info->ring, sring, PAGE_SIZE);
+
-+static struct xenbus_driver netfront = {
-+ .name = "vif",
-+ .owner = THIS_MODULE,
-+ .ids = netfront_ids,
-+ .probe = netfront_probe,
-+ .remove = __devexit_p(netfront_remove),
-+ .resume = netfront_resume,
-+ .otherend_changed = backend_changed,
-+};
++ err = bind_interdomain_evtchn_to_irqhandler(
++ info->domid, evtchn,
++ scsiback_intr, 0, "vscsiif-backend", info);
+
++ if (err < 0)
++ goto unmap_page;
++
++ info->irq = err;
+
-+static struct notifier_block notifier_inetdev = {
-+ .notifier_call = inetdev_notify,
-+ .next = NULL,
-+ .priority = 0
-+};
++ return 0;
+
-+static int __init netif_init(void)
++unmap_page:
++ unmap_frontend_page(info);
++free_vm:
++ free_vm_area(info->ring_area);
++
++ return err;
++}
++
++void scsiback_disconnect(struct vscsibk_info *info)
+{
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++ if (info->kthread) {
++ kthread_stop(info->kthread);
++ info->kthread = NULL;
++ }
+
-+#ifdef CONFIG_XEN
-+ if (MODPARM_rx_flip && MODPARM_rx_copy) {
-+ WPRINTK("Cannot specify both rx_copy and rx_flip.\n");
-+ return -EINVAL;
++ wait_event(info->waiting_to_free,
++ atomic_read(&info->nr_unreplied_reqs) == 0);
++
++ if (info->irq) {
++ unbind_from_irqhandler(info->irq, info);
++ info->irq = 0;
+ }
+
-+ if (!MODPARM_rx_flip && !MODPARM_rx_copy)
-+ MODPARM_rx_flip = 1; /* Default is to flip. */
-+#endif
++ if (info->ring.sring) {
++ unmap_frontend_page(info);
++ free_vm_area(info->ring_area);
++ info->ring.sring = NULL;
++ }
++}
++
++void scsiback_free(struct vscsibk_info *info)
++{
++ kmem_cache_free(scsiback_cachep, info);
++}
++
++int __init scsiback_interface_init(void)
++{
++ scsiback_cachep = kmem_cache_create("vscsiif_cache",
++ sizeof(struct vscsibk_info), 0, 0, NULL, NULL);
++ if (!scsiback_cachep) {
++ printk(KERN_ERR "scsiback: can't init scsi cache\n");
++ return -ENOMEM;
++ }
++
++ return 0;
++}
++
++void scsiback_interface_exit(void)
++{
++ kmem_cache_destroy(scsiback_cachep);
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/scsiback/scsiback.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/scsiback/scsiback.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,717 @@
++/*
++ * Xen SCSI backend driver
++ *
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * Based on the blkback driver code.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ if (is_initial_xendomain())
-+ return 0;
++#include <linux/spinlock.h>
++#include <linux/kthread.h>
++#include <linux/list.h>
++#include <linux/delay.h>
++#include <xen/balloon.h>
++#include <asm/hypervisor.h>
++#include <scsi/scsi.h>
++#include <scsi/scsi_cmnd.h>
++#include <scsi/scsi_host.h>
++#include <scsi/scsi_device.h>
++#include <scsi/scsi_dbg.h>
++#include <scsi/scsi_eh.h>
+
-+ IPRINTK("Initialising virtual ethernet driver.\n");
++#include "common.h"
+
-+ (void)register_inetaddr_notifier(¬ifier_inetdev);
+
-+ return xenbus_register_frontend(&netfront);
-+}
-+module_init(netif_init);
++struct list_head pending_free;
++DEFINE_SPINLOCK(pending_free_lock);
++DECLARE_WAIT_QUEUE_HEAD(pending_free_wq);
++
++int vscsiif_reqs = VSCSIIF_BACK_MAX_PENDING_REQS;
++module_param_named(reqs, vscsiif_reqs, int, 0);
++MODULE_PARM_DESC(reqs, "Number of scsiback requests to allocate");
+
++static unsigned int log_print_stat = 0;
++module_param(log_print_stat, int, 0644);
+
-+static void __exit netif_exit(void)
-+{
-+ if (is_initial_xendomain())
-+ return;
++#define SCSIBACK_INVALID_HANDLE (~0)
+
-+ unregister_inetaddr_notifier(¬ifier_inetdev);
++static pending_req_t *pending_reqs;
++static struct page **pending_pages;
++static grant_handle_t *pending_grant_handles;
+
-+ return xenbus_unregister_driver(&netfront);
++static int vaddr_pagenr(pending_req_t *req, int seg)
++{
++ return (req - pending_reqs) * VSCSIIF_SG_TABLESIZE + seg;
+}
-+module_exit(netif_exit);
-+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,426 @@
-+/*
-+ * PCI Backend - Functions for creating a virtual configuration space for
-+ * exported PCI Devices.
-+ * It's dangerous to allow PCI Driver Domains to change their
-+ * device's resources (memory, i/o ports, interrupts). We need to
-+ * restrict changes to certain PCI Configuration registers:
-+ * BARs, INTERRUPT_PIN, most registers in the header...
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ */
-+
-+#include <linux/kernel.h>
-+#include <linux/pci.h>
-+#include "pciback.h"
-+#include "conf_space.h"
-+#include "conf_space_quirks.h"
+
-+#define DEFINE_PCI_CONFIG(op,size,type) \
-+int pciback_##op##_config_##size \
-+(struct pci_dev *dev, int offset, type value, void *data) \
-+{ \
-+ return pci_##op##_config_##size (dev, offset, value); \
++static unsigned long vaddr(pending_req_t *req, int seg)
++{
++ unsigned long pfn = page_to_pfn(pending_pages[vaddr_pagenr(req, seg)]);
++ return (unsigned long)pfn_to_kaddr(pfn);
+}
+
-+DEFINE_PCI_CONFIG(read, byte, u8 *)
-+DEFINE_PCI_CONFIG(read, word, u16 *)
-+DEFINE_PCI_CONFIG(read, dword, u32 *)
++#define pending_handle(_req, _seg) \
++ (pending_grant_handles[vaddr_pagenr(_req, _seg)])
+
-+DEFINE_PCI_CONFIG(write, byte, u8)
-+DEFINE_PCI_CONFIG(write, word, u16)
-+DEFINE_PCI_CONFIG(write, dword, u32)
+
-+static int conf_space_read(struct pci_dev *dev,
-+ struct config_field_entry *entry, int offset,
-+ u32 * value)
++void scsiback_fast_flush_area(pending_req_t *req)
+{
-+ int ret = 0;
-+ struct config_field *field = entry->field;
++ struct gnttab_unmap_grant_ref unmap[VSCSIIF_SG_TABLESIZE];
++ unsigned int i, invcount = 0;
++ grant_handle_t handle;
++ int err;
+
-+ *value = 0;
++ if (req->nr_segments) {
++ for (i = 0; i < req->nr_segments; i++) {
++ handle = pending_handle(req, i);
++ if (handle == SCSIBACK_INVALID_HANDLE)
++ continue;
++ gnttab_set_unmap_op(&unmap[i], vaddr(req, i),
++ GNTMAP_host_map, handle);
++ pending_handle(req, i) = SCSIBACK_INVALID_HANDLE;
++ invcount++;
++ }
+
-+ switch (field->size) {
-+ case 1:
-+ if (field->u.b.read)
-+ ret = field->u.b.read(dev, offset, (u8 *) value,
-+ entry->data);
-+ break;
-+ case 2:
-+ if (field->u.w.read)
-+ ret = field->u.w.read(dev, offset, (u16 *) value,
-+ entry->data);
-+ break;
-+ case 4:
-+ if (field->u.dw.read)
-+ ret = field->u.dw.read(dev, offset, value, entry->data);
-+ break;
++ err = HYPERVISOR_grant_table_op(
++ GNTTABOP_unmap_grant_ref, unmap, invcount);
++ BUG_ON(err);
++ kfree(req->sgl);
+ }
-+ return ret;
++
++ return;
+}
+
-+static int conf_space_write(struct pci_dev *dev,
-+ struct config_field_entry *entry, int offset,
-+ u32 value)
++
++static pending_req_t * alloc_req(struct vscsibk_info *info)
+{
-+ int ret = 0;
-+ struct config_field *field = entry->field;
++ pending_req_t *req = NULL;
++ unsigned long flags;
+
-+ switch (field->size) {
-+ case 1:
-+ if (field->u.b.write)
-+ ret = field->u.b.write(dev, offset, (u8) value,
-+ entry->data);
-+ break;
-+ case 2:
-+ if (field->u.w.write)
-+ ret = field->u.w.write(dev, offset, (u16) value,
-+ entry->data);
-+ break;
-+ case 4:
-+ if (field->u.dw.write)
-+ ret = field->u.dw.write(dev, offset, value,
-+ entry->data);
-+ break;
++ spin_lock_irqsave(&pending_free_lock, flags);
++ if (!list_empty(&pending_free)) {
++ req = list_entry(pending_free.next, pending_req_t, free_list);
++ list_del(&req->free_list);
+ }
-+ return ret;
-+}
-+
-+static inline u32 get_mask(int size)
-+{
-+ if (size == 1)
-+ return 0xff;
-+ else if (size == 2)
-+ return 0xffff;
-+ else
-+ return 0xffffffff;
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++ return req;
+}
+
-+static inline int valid_request(int offset, int size)
-+{
-+ /* Validate request (no un-aligned requests) */
-+ if ((size == 1 || size == 2 || size == 4) && (offset % size) == 0)
-+ return 1;
-+ return 0;
-+}
+
-+static inline u32 merge_value(u32 val, u32 new_val, u32 new_val_mask,
-+ int offset)
++static void free_req(pending_req_t *req)
+{
-+ if (offset >= 0) {
-+ new_val_mask <<= (offset * 8);
-+ new_val <<= (offset * 8);
-+ } else {
-+ new_val_mask >>= (offset * -8);
-+ new_val >>= (offset * -8);
-+ }
-+ val = (val & ~new_val_mask) | (new_val & new_val_mask);
++ unsigned long flags;
++ int was_empty;
+
-+ return val;
++ spin_lock_irqsave(&pending_free_lock, flags);
++ was_empty = list_empty(&pending_free);
++ list_add(&req->free_list, &pending_free);
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++ if (was_empty)
++ wake_up(&pending_free_wq);
+}
+
-+static int pcibios_err_to_errno(int err)
++
++static void scsiback_notify_work(struct vscsibk_info *info)
+{
-+ switch (err) {
-+ case PCIBIOS_SUCCESSFUL:
-+ return XEN_PCI_ERR_success;
-+ case PCIBIOS_DEVICE_NOT_FOUND:
-+ return XEN_PCI_ERR_dev_not_found;
-+ case PCIBIOS_BAD_REGISTER_NUMBER:
-+ return XEN_PCI_ERR_invalid_offset;
-+ case PCIBIOS_FUNC_NOT_SUPPORTED:
-+ return XEN_PCI_ERR_not_implemented;
-+ case PCIBIOS_SET_FAILED:
-+ return XEN_PCI_ERR_access_denied;
-+ }
-+ return err;
++ info->waiting_reqs = 1;
++ wake_up(&info->wq);
+}
+
-+int pciback_config_read(struct pci_dev *dev, int offset, int size,
-+ u32 * ret_val)
++void scsiback_do_resp_with_sense(char *sense_buffer, int32_t result,
++ uint32_t resid, pending_req_t *pending_req)
+{
-+ int err = 0;
-+ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
-+ struct config_field_entry *cfg_entry;
-+ struct config_field *field;
-+ int req_start, req_end, field_start, field_end;
-+ /* if read fails for any reason, return 0 (as if device didn't respond) */
-+ u32 value = 0, tmp_val;
++ vscsiif_response_t *ring_res;
++ struct vscsibk_info *info = pending_req->info;
++ int notify;
++ int more_to_do = 1;
++ unsigned long flags;
+
-+ if (unlikely(verbose_request))
-+ printk(KERN_DEBUG "pciback: %s: read %d bytes at 0x%x\n",
-+ pci_name(dev), size, offset);
++ DPRINTK("%s\n",__FUNCTION__);
+
-+ if (!valid_request(offset, size)) {
-+ err = XEN_PCI_ERR_invalid_offset;
-+ goto out;
-+ }
++ spin_lock_irqsave(&info->ring_lock, flags);
+
-+ /* Get the real value first, then modify as appropriate */
-+ switch (size) {
-+ case 1:
-+ err = pci_read_config_byte(dev, offset, (u8 *) & value);
-+ break;
-+ case 2:
-+ err = pci_read_config_word(dev, offset, (u16 *) & value);
-+ break;
-+ case 4:
-+ err = pci_read_config_dword(dev, offset, &value);
-+ break;
-+ }
++ ring_res = RING_GET_RESPONSE(&info->ring, info->ring.rsp_prod_pvt);
++ info->ring.rsp_prod_pvt++;
+
-+ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
-+ field = cfg_entry->field;
++ ring_res->rslt = result;
++ ring_res->rqid = pending_req->rqid;
+
-+ req_start = offset;
-+ req_end = offset + size;
-+ field_start = OFFSET(cfg_entry);
-+ field_end = OFFSET(cfg_entry) + field->size;
++ if (sense_buffer != NULL) {
++ memcpy(ring_res->sense_buffer, sense_buffer,
++ VSCSIIF_SENSE_BUFFERSIZE);
++ ring_res->sense_len = VSCSIIF_SENSE_BUFFERSIZE;
++ } else {
++ ring_res->sense_len = 0;
++ }
+
-+ if ((req_start >= field_start && req_start < field_end)
-+ || (req_end > field_start && req_end <= field_end)) {
-+ err = conf_space_read(dev, cfg_entry, field_start,
-+ &tmp_val);
-+ if (err)
-+ goto out;
++ ring_res->residual_len = resid;
+
-+ value = merge_value(value, tmp_val,
-+ get_mask(field->size),
-+ field_start - req_start);
-+ }
++ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&info->ring, notify);
++ if (info->ring.rsp_prod_pvt == info->ring.req_cons) {
++ RING_FINAL_CHECK_FOR_REQUESTS(&info->ring, more_to_do);
++ } else if (RING_HAS_UNCONSUMED_REQUESTS(&info->ring)) {
++ more_to_do = 1;
+ }
++
++ spin_unlock_irqrestore(&info->ring_lock, flags);
+
-+ out:
-+ if (unlikely(verbose_request))
-+ printk(KERN_DEBUG "pciback: %s: read %d bytes at 0x%x = %x\n",
-+ pci_name(dev), size, offset, value);
++ if (more_to_do)
++ scsiback_notify_work(info);
+
-+ *ret_val = value;
-+ return pcibios_err_to_errno(err);
++ if (notify)
++ notify_remote_via_irq(info->irq);
++
++ free_req(pending_req);
+}
+
-+int pciback_config_write(struct pci_dev *dev, int offset, int size, u32 value)
++static void scsiback_print_status(char *sense_buffer, int errors,
++ pending_req_t *pending_req)
+{
-+ int err = 0, handled = 0;
-+ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
-+ struct config_field_entry *cfg_entry;
-+ struct config_field *field;
-+ u32 tmp_val;
-+ int req_start, req_end, field_start, field_end;
++ struct scsi_device *sdev = pending_req->sdev;
++
++ printk(KERN_ERR "scsiback: %d:%d:%d:%d ",sdev->host->host_no,
++ sdev->channel, sdev->id, sdev->lun);
++ printk(KERN_ERR "status = 0x%02x, message = 0x%02x, host = 0x%02x, driver = 0x%02x\n",
++ status_byte(errors), msg_byte(errors),
++ host_byte(errors), driver_byte(errors));
+
-+ if (unlikely(verbose_request))
-+ printk(KERN_DEBUG
-+ "pciback: %s: write request %d bytes at 0x%x = %x\n",
-+ pci_name(dev), size, offset, value);
++ printk(KERN_ERR "scsiback: cmnd[0]=0x%02X\n",
++ pending_req->cmnd[0]);
+
-+ if (!valid_request(offset, size))
-+ return XEN_PCI_ERR_invalid_offset;
++ if (CHECK_CONDITION & status_byte(errors))
++ __scsi_print_sense("scsiback", sense_buffer, SCSI_SENSE_BUFFERSIZE);
++}
+
-+ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
-+ field = cfg_entry->field;
+
-+ req_start = offset;
-+ req_end = offset + size;
-+ field_start = OFFSET(cfg_entry);
-+ field_end = OFFSET(cfg_entry) + field->size;
++static void scsiback_cmd_done(struct request *req, int errors)
++{
++ pending_req_t *pending_req = req->end_io_data;
++ unsigned char *sense_buffer;
++ unsigned int resid;
+
-+ if ((req_start >= field_start && req_start < field_end)
-+ || (req_end > field_start && req_end <= field_end)) {
-+ tmp_val = 0;
++ sense_buffer = req->sense;
++ resid = req->data_len;
+
-+ err = pciback_config_read(dev, field_start,
-+ field->size, &tmp_val);
-+ if (err)
-+ break;
++ if (errors != 0) {
++ if (log_print_stat)
++ scsiback_print_status(sense_buffer, errors, pending_req);
++ }
+
-+ tmp_val = merge_value(tmp_val, value, get_mask(size),
-+ req_start - field_start);
++ scsiback_rsp_emulation(pending_req);
+
-+ err = conf_space_write(dev, cfg_entry, field_start,
-+ tmp_val);
++ scsiback_fast_flush_area(pending_req);
++ scsiback_do_resp_with_sense(sense_buffer, errors, resid, pending_req);
++ scsiback_put(pending_req->info);
+
-+ /* handled is set true here, but not every byte
-+ * may have been written! Properly detecting if
-+ * every byte is handled is unnecessary as the
-+ * flag is used to detect devices that need
-+ * special helpers to work correctly.
-+ */
-+ handled = 1;
++ __blk_put_request(req->q, req);
++}
++
++
++static int scsiback_gnttab_data_map(vscsiif_request_t *ring_req,
++ pending_req_t *pending_req)
++{
++ u32 flags;
++ int write;
++ int i, err = 0;
++ unsigned int data_len = 0;
++ struct gnttab_map_grant_ref map[VSCSIIF_SG_TABLESIZE];
++ struct vscsibk_info *info = pending_req->info;
++
++ int data_dir = (int)pending_req->sc_data_direction;
++ unsigned int nr_segments = (unsigned int)pending_req->nr_segments;
++
++ write = (data_dir == DMA_TO_DEVICE);
++
++ if (nr_segments) {
++ /* free of (sgl) in fast_flush_area()*/
++ pending_req->sgl = kmalloc(sizeof(struct scatterlist) * nr_segments,
++ GFP_KERNEL);
++ if (!pending_req->sgl) {
++ printk(KERN_ERR "scsiback: %s: kmalloc() error.\n", __FUNCTION__);
++ return -ENOMEM;
+ }
-+ }
+
-+ if (!handled && !err) {
-+ /* By default, anything not specificially handled above is
-+ * read-only. The permissive flag changes this behavior so
-+ * that anything not specifically handled above is writable.
-+ * This means that some fields may still be read-only because
-+ * they have entries in the config_field list that intercept
-+ * the write and do nothing. */
-+ if (dev_data->permissive) {
-+ switch (size) {
-+ case 1:
-+ err = pci_write_config_byte(dev, offset,
-+ (u8) value);
-+ break;
-+ case 2:
-+ err = pci_write_config_word(dev, offset,
-+ (u16) value);
-+ break;
-+ case 4:
-+ err = pci_write_config_dword(dev, offset,
-+ (u32) value);
-+ break;
-+ }
-+ } else if (!dev_data->warned_on_write) {
-+ dev_data->warned_on_write = 1;
-+ dev_warn(&dev->dev, "Driver tried to write to a "
-+ "read-only configuration space field at offset "
-+ "0x%x, size %d. This may be harmless, but if "
-+ "you have problems with your device:\n"
-+ "1) see permissive attribute in sysfs\n"
-+ "2) report problems to the xen-devel "
-+ "mailing list along with details of your "
-+ "device obtained from lspci.\n", offset, size);
++ for (i = 0; i < nr_segments; i++) {
++ flags = GNTMAP_host_map;
++ if (write)
++ flags |= GNTMAP_readonly;
++ gnttab_set_map_op(&map[i], vaddr(pending_req, i), flags,
++ ring_req->seg[i].gref,
++ info->domid);
+ }
-+ }
+
-+ return pcibios_err_to_errno(err);
-+}
++ err = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, map, nr_segments);
++ BUG_ON(err);
+
-+void pciback_config_free_dyn_fields(struct pci_dev *dev)
-+{
-+ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
-+ struct config_field_entry *cfg_entry, *t;
-+ struct config_field *field;
++ for (i = 0; i < nr_segments; i++) {
++ if (unlikely(map[i].status != 0)) {
++ printk(KERN_ERR "scsiback: invalid buffer -- could not remap it\n");
++ map[i].handle = SCSIBACK_INVALID_HANDLE;
++ err |= 1;
++ }
+
-+ dev_dbg(&dev->dev,
-+ "free-ing dynamically allocated virtual configuration space fields\n");
++ pending_handle(pending_req, i) = map[i].handle;
+
-+ list_for_each_entry_safe(cfg_entry, t, &dev_data->config_fields, list) {
-+ field = cfg_entry->field;
++ if (err)
++ continue;
+
-+ if (field->clean) {
-+ field->clean(field);
++ set_phys_to_machine(__pa(vaddr(
++ pending_req, i)) >> PAGE_SHIFT,
++ FOREIGN_FRAME(map[i].dev_bus_addr >> PAGE_SHIFT));
++
++ pending_req->sgl[i].page = virt_to_page(vaddr(pending_req, i));
++ pending_req->sgl[i].offset = ring_req->seg[i].offset;
++ pending_req->sgl[i].length = ring_req->seg[i].length;
++ data_len += pending_req->sgl[i].length;
+
-+ if (cfg_entry->data)
-+ kfree(cfg_entry->data);
++ barrier();
++ if (pending_req->sgl[i].offset >= PAGE_SIZE ||
++ pending_req->sgl[i].length > PAGE_SIZE ||
++ pending_req->sgl[i].offset + pending_req->sgl[i].length > PAGE_SIZE)
++ err |= 1;
+
-+ list_del(&cfg_entry->list);
-+ kfree(cfg_entry);
+ }
+
++ if (err)
++ goto fail_flush;
+ }
++
++ pending_req->request_bufflen = data_len;
++
++ return 0;
++
++fail_flush:
++ scsiback_fast_flush_area(pending_req);
++ return -ENOMEM;
+}
+
-+void pciback_config_reset_dev(struct pci_dev *dev)
++/* quoted scsi_lib.c/scsi_merge_bio */
++static int scsiback_merge_bio(struct request *rq, struct bio *bio)
+{
-+ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
-+ struct config_field_entry *cfg_entry;
-+ struct config_field *field;
++ struct request_queue *q = rq->q;
+
-+ dev_dbg(&dev->dev, "resetting virtual configuration space\n");
++ bio->bi_flags &= ~(1 << BIO_SEG_VALID);
++ if (rq_data_dir(rq) == WRITE)
++ bio->bi_rw |= (1 << BIO_RW);
+
-+ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
-+ field = cfg_entry->field;
++ blk_queue_bounce(q, &bio);
+
-+ if (field->reset)
-+ field->reset(dev, OFFSET(cfg_entry), cfg_entry->data);
++ if (!rq->bio)
++ blk_rq_bio_prep(q, rq, bio);
++ else if (!q->back_merge_fn(q, rq, bio))
++ return -EINVAL;
++ else {
++ rq->biotail->bi_next = bio;
++ rq->biotail = bio;
++ rq->hard_nr_sectors += bio_sectors(bio);
++ rq->nr_sectors = rq->hard_nr_sectors;
+ }
-+}
+
-+void pciback_config_free_dev(struct pci_dev *dev)
-+{
-+ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
-+ struct config_field_entry *cfg_entry, *t;
-+ struct config_field *field;
++ return 0;
++}
+
-+ dev_dbg(&dev->dev, "free-ing virtual configuration space fields\n");
+
-+ list_for_each_entry_safe(cfg_entry, t, &dev_data->config_fields, list) {
-+ list_del(&cfg_entry->list);
++/* quoted scsi_lib.c/scsi_bi_endio */
++static int scsiback_bi_endio(struct bio *bio, unsigned int bytes_done, int error)
++{
++ if (bio->bi_size)
++ return 1;
+
-+ field = cfg_entry->field;
++ bio_put(bio);
++ return 0;
++}
+
-+ if (field->release)
-+ field->release(dev, OFFSET(cfg_entry), cfg_entry->data);
+
-+ kfree(cfg_entry);
-+ }
-+}
+
-+int pciback_config_add_field_offset(struct pci_dev *dev,
-+ struct config_field *field,
-+ unsigned int base_offset)
++/* quoted scsi_lib.c/scsi_req_map_sg . */
++static int request_map_sg(struct request *rq, pending_req_t *pending_req, unsigned int count)
+{
-+ int err = 0;
-+ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
-+ struct config_field_entry *cfg_entry;
-+ void *tmp;
++ struct request_queue *q = rq->q;
++ int nr_pages;
++ unsigned int nsegs = count;
+
-+ cfg_entry = kmalloc(sizeof(*cfg_entry), GFP_KERNEL);
-+ if (!cfg_entry) {
-+ err = -ENOMEM;
-+ goto out;
-+ }
++ unsigned int data_len = 0, len, bytes, off;
++ struct page *page;
++ struct bio *bio = NULL;
++ int i, err, nr_vecs = 0;
+
-+ cfg_entry->data = NULL;
-+ cfg_entry->field = field;
-+ cfg_entry->base_offset = base_offset;
++ for (i = 0; i < nsegs; i++) {
++ page = pending_req->sgl[i].page;
++ off = (unsigned int)pending_req->sgl[i].offset;
++ len = (unsigned int)pending_req->sgl[i].length;
++ data_len += len;
++
++ nr_pages = (len + off + PAGE_SIZE - 1) >> PAGE_SHIFT;
++ while (len > 0) {
++ bytes = min_t(unsigned int, len, PAGE_SIZE - off);
++
++ if (!bio) {
++ nr_vecs = min_t(int, BIO_MAX_PAGES, nr_pages);
++ nr_pages -= nr_vecs;
++ bio = bio_alloc(GFP_KERNEL, nr_vecs);
++ if (!bio) {
++ err = -ENOMEM;
++ goto free_bios;
++ }
++ bio->bi_end_io = scsiback_bi_endio;
++ }
+
-+ /* silently ignore duplicate fields */
-+ err = pciback_field_is_dup(dev,OFFSET(cfg_entry));
-+ if (err)
-+ goto out;
++ if (bio_add_pc_page(q, bio, page, bytes, off) !=
++ bytes) {
++ bio_put(bio);
++ err = -EINVAL;
++ goto free_bios;
++ }
+
-+ if (field->init) {
-+ tmp = field->init(dev, OFFSET(cfg_entry));
++ if (bio->bi_vcnt >= nr_vecs) {
++ err = scsiback_merge_bio(rq, bio);
++ if (err) {
++ bio_endio(bio, bio->bi_size, 0);
++ goto free_bios;
++ }
++ bio = NULL;
++ }
+
-+ if (IS_ERR(tmp)) {
-+ err = PTR_ERR(tmp);
-+ goto out;
++ page++;
++ len -= bytes;
++ off = 0;
+ }
-+
-+ cfg_entry->data = tmp;
+ }
+
-+ dev_dbg(&dev->dev, "added config field at offset 0x%02x\n",
-+ OFFSET(cfg_entry));
-+ list_add_tail(&cfg_entry->list, &dev_data->config_fields);
++ rq->buffer = rq->data = NULL;
++ rq->data_len = data_len;
+
-+ out:
-+ if (err)
-+ kfree(cfg_entry);
++ return 0;
++
++free_bios:
++ while ((bio = rq->bio) != NULL) {
++ rq->bio = bio->bi_next;
++ /*
++ * call endio instead of bio_put incase it was bounced
++ */
++ bio_endio(bio, bio->bi_size, 0);
++ }
+
+ return err;
+}
+
-+/* This sets up the device's virtual configuration space to keep track of
-+ * certain registers (like the base address registers (BARs) so that we can
-+ * keep the client from manipulating them directly.
-+ */
-+int pciback_config_init_dev(struct pci_dev *dev)
++
++void scsiback_cmd_exec(pending_req_t *pending_req)
+{
-+ int err = 0;
-+ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ int cmd_len = (int)pending_req->cmd_len;
++ int data_dir = (int)pending_req->sc_data_direction;
++ unsigned int nr_segments = (unsigned int)pending_req->nr_segments;
++ unsigned int timeout;
++ struct request *rq;
++ int write;
+
-+ dev_dbg(&dev->dev, "initializing virtual configuration space\n");
++ DPRINTK("%s\n",__FUNCTION__);
+
-+ INIT_LIST_HEAD(&dev_data->config_fields);
++ /* because it doesn't timeout backend earlier than frontend.*/
++ if (pending_req->timeout_per_command)
++ timeout = pending_req->timeout_per_command * HZ;
++ else
++ timeout = VSCSIIF_TIMEOUT;
+
-+ err = pciback_config_header_add_fields(dev);
-+ if (err)
-+ goto out;
++ write = (data_dir == DMA_TO_DEVICE);
++ rq = blk_get_request(pending_req->sdev->request_queue, write, GFP_KERNEL);
+
-+ err = pciback_config_capability_add_fields(dev);
-+ if (err)
-+ goto out;
++ rq->flags |= REQ_BLOCK_PC;
++ rq->cmd_len = cmd_len;
++ memcpy(rq->cmd, pending_req->cmnd, cmd_len);
+
-+ err = pciback_config_quirks_init(dev);
++ memset(pending_req->sense_buffer, 0, VSCSIIF_SENSE_BUFFERSIZE);
++ rq->sense = pending_req->sense_buffer;
++ rq->sense_len = 0;
+
-+ out:
-+ return err;
++ /* not allowed to retry in backend. */
++ rq->retries = 0;
++ rq->timeout = timeout;
++ rq->end_io_data = pending_req;
++
++ if (nr_segments) {
++
++ if (request_map_sg(rq, pending_req, nr_segments)) {
++ printk(KERN_ERR "scsiback: SG Request Map Error\n");
++ return;
++ }
++ }
++
++ scsiback_get(pending_req->info);
++ blk_execute_rq_nowait(rq->q, NULL, rq, 1, scsiback_cmd_done);
++
++ return ;
+}
+
-+int pciback_config_init(void)
++
++static void scsiback_device_reset_exec(pending_req_t *pending_req)
+{
-+ return pciback_config_capability_init();
-+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_capability.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_capability.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_capability.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_capability.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,71 @@
-+/*
-+ * PCI Backend - Handles the virtual fields found on the capability lists
-+ * in the configuration space.
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ */
++ struct vscsibk_info *info = pending_req->info;
++ int err;
++ struct scsi_device *sdev = pending_req->sdev;
+
-+#include <linux/kernel.h>
-+#include <linux/pci.h>
-+#include "pciback.h"
-+#include "conf_space.h"
-+#include "conf_space_capability.h"
++ scsiback_get(info);
++ err = scsi_reset_provider(sdev, SCSI_TRY_RESET_DEVICE);
+
-+static LIST_HEAD(capabilities);
++ scsiback_do_resp_with_sense(NULL, err, 0, pending_req);
++ scsiback_put(info);
+
-+static struct config_field caplist_header[] = {
-+ {
-+ .offset = PCI_CAP_LIST_ID,
-+ .size = 2, /* encompass PCI_CAP_LIST_ID & PCI_CAP_LIST_NEXT */
-+ .u.w.read = pciback_read_config_word,
-+ .u.w.write = NULL,
-+ },
-+ {
-+ .size = 0,
-+ },
-+};
++ return;
++}
+
-+static inline void register_capability(struct pciback_config_capability *cap)
++
++irqreturn_t scsiback_intr(int irq, void *dev_id, struct pt_regs *regs)
+{
-+ list_add_tail(&cap->cap_list, &capabilities);
++ scsiback_notify_work((struct vscsibk_info *)dev_id);
++ return IRQ_HANDLED;
+}
+
-+int pciback_config_capability_add_fields(struct pci_dev *dev)
++static int prepare_pending_reqs(struct vscsibk_info *info,
++ vscsiif_request_t *ring_req, pending_req_t *pending_req)
+{
-+ int err = 0;
-+ struct pciback_config_capability *cap;
-+ int cap_offset;
++ struct scsi_device *sdev;
++ struct ids_tuple vir;
++ int err = -EINVAL;
+
-+ list_for_each_entry(cap, &capabilities, cap_list) {
-+ cap_offset = pci_find_capability(dev, cap->capability);
-+ if (cap_offset) {
-+ dev_dbg(&dev->dev, "Found capability 0x%x at 0x%x\n",
-+ cap->capability, cap_offset);
++ DPRINTK("%s\n",__FUNCTION__);
+
-+ err = pciback_config_add_fields_offset(dev,
-+ caplist_header,
-+ cap_offset);
-+ if (err)
-+ goto out;
-+ err = pciback_config_add_fields_offset(dev,
-+ cap->fields,
-+ cap_offset);
-+ if (err)
-+ goto out;
-+ }
++ pending_req->rqid = ring_req->rqid;
++ pending_req->act = ring_req->act;
++
++ pending_req->info = info;
++
++ vir.chn = ring_req->channel;
++ vir.tgt = ring_req->id;
++ vir.lun = ring_req->lun;
++
++ rmb();
++ sdev = scsiback_do_translation(info, &vir);
++ if (!sdev) {
++ pending_req->sdev = NULL;
++ DPRINTK("scsiback: doesn't exist.\n");
++ err = -ENODEV;
++ goto invalid_value;
+ }
++ pending_req->sdev = sdev;
+
-+ out:
-+ return err;
-+}
++ /* request range check from frontend */
++ pending_req->sc_data_direction = ring_req->sc_data_direction;
++ barrier();
++ if ((pending_req->sc_data_direction != DMA_BIDIRECTIONAL) &&
++ (pending_req->sc_data_direction != DMA_TO_DEVICE) &&
++ (pending_req->sc_data_direction != DMA_FROM_DEVICE) &&
++ (pending_req->sc_data_direction != DMA_NONE)) {
++ DPRINTK("scsiback: invalid parameter data_dir = %d\n",
++ pending_req->sc_data_direction);
++ err = -EINVAL;
++ goto invalid_value;
++ }
+
-+extern struct pciback_config_capability pciback_config_capability_vpd;
-+extern struct pciback_config_capability pciback_config_capability_pm;
++ pending_req->nr_segments = ring_req->nr_segments;
++ barrier();
++ if (pending_req->nr_segments > VSCSIIF_SG_TABLESIZE) {
++ DPRINTK("scsiback: invalid parameter nr_seg = %d\n",
++ pending_req->nr_segments);
++ err = -EINVAL;
++ goto invalid_value;
++ }
+
-+int pciback_config_capability_init(void)
-+{
-+ register_capability(&pciback_config_capability_vpd);
-+ register_capability(&pciback_config_capability_pm);
++ pending_req->cmd_len = ring_req->cmd_len;
++ barrier();
++ if (pending_req->cmd_len > VSCSIIF_MAX_COMMAND_SIZE) {
++ DPRINTK("scsiback: invalid parameter cmd_len = %d\n",
++ pending_req->cmd_len);
++ err = -EINVAL;
++ goto invalid_value;
++ }
++ memcpy(pending_req->cmnd, ring_req->cmnd, pending_req->cmd_len);
++
++ pending_req->timeout_per_command = ring_req->timeout_per_command;
++
++ if(scsiback_gnttab_data_map(ring_req, pending_req)) {
++ DPRINTK("scsiback: invalid buffer\n");
++ err = -EINVAL;
++ goto invalid_value;
++ }
+
+ return 0;
++
++invalid_value:
++ return err;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_capability.h tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_capability.h
---- pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_capability.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_capability.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,23 @@
-+/*
-+ * PCI Backend - Data structures for special overlays for structures on
-+ * the capability list.
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ */
+
-+#ifndef __PCIBACK_CONFIG_CAPABILITY_H__
-+#define __PCIBACK_CONFIG_CAPABILITY_H__
+
-+#include <linux/pci.h>
-+#include <linux/list.h>
++static int scsiback_do_cmd_fn(struct vscsibk_info *info)
++{
++ struct vscsiif_back_ring *ring = &info->ring;
++ vscsiif_request_t *ring_req;
+
-+struct pciback_config_capability {
-+ struct list_head cap_list;
++ pending_req_t *pending_req;
++ RING_IDX rc, rp;
++ int err, more_to_do = 0;
+
-+ int capability;
++ DPRINTK("%s\n",__FUNCTION__);
+
-+ /* If the device has the capability found above, add these fields */
-+ struct config_field *fields;
-+};
++ rc = ring->req_cons;
++ rp = ring->sring->req_prod;
++ rmb();
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_capability_pm.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_capability_pm.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_capability_pm.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_capability_pm.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,128 @@
-+/*
-+ * PCI Backend - Configuration space overlay for power management
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ */
++ while ((rc != rp)) {
++ if (RING_REQUEST_CONS_OVERFLOW(ring, rc))
++ break;
++ pending_req = alloc_req(info);
++ if (NULL == pending_req) {
++ more_to_do = 1;
++ break;
++ }
+
-+#include <linux/pci.h>
-+#include "conf_space.h"
-+#include "conf_space_capability.h"
++ ring_req = RING_GET_REQUEST(ring, rc);
++ ring->req_cons = ++rc;
+
-+static int pm_caps_read(struct pci_dev *dev, int offset, u16 *value,
-+ void *data)
-+{
-+ int err;
-+ u16 real_value;
++ err = prepare_pending_reqs(info, ring_req,
++ pending_req);
++ if (err == -EINVAL) {
++ scsiback_do_resp_with_sense(NULL, (DRIVER_ERROR << 24),
++ 0, pending_req);
++ continue;
++ } else if (err == -ENODEV) {
++ scsiback_do_resp_with_sense(NULL, (DID_NO_CONNECT << 16),
++ 0, pending_req);
++ continue;
++ }
+
-+ err = pci_read_config_word(dev, offset, &real_value);
-+ if (err)
-+ goto out;
++ if (pending_req->act == VSCSIIF_ACT_SCSI_CDB) {
++ scsiback_req_emulation_or_cmdexec(pending_req);
++ } else if (pending_req->act == VSCSIIF_ACT_SCSI_RESET) {
++ scsiback_device_reset_exec(pending_req);
++ } else {
++ printk(KERN_ERR "scsiback: invalid parameter for request\n");
++ scsiback_do_resp_with_sense(NULL, (DRIVER_ERROR << 24),
++ 0, pending_req);
++ continue;
++ }
++ }
+
-+ *value = real_value & ~PCI_PM_CAP_PME_MASK;
++ if (RING_HAS_UNCONSUMED_REQUESTS(ring))
++ more_to_do = 1;
+
-+ out:
-+ return err;
++ /* Yield point for this unbounded loop. */
++ cond_resched();
++
++ return more_to_do;
+}
+
-+/* PM_OK_BITS specifies the bits that the driver domain is allowed to change.
-+ * Can't allow driver domain to enable PMEs - they're shared */
-+#define PM_OK_BITS (PCI_PM_CTRL_PME_STATUS|PCI_PM_CTRL_DATA_SEL_MASK)
+
-+static int pm_ctrl_write(struct pci_dev *dev, int offset, u16 new_value,
-+ void *data)
++int scsiback_schedule(void *data)
+{
-+ int err;
-+ u16 old_value;
-+ pci_power_t new_state, old_state;
++ struct vscsibk_info *info = (struct vscsibk_info *)data;
+
-+ err = pci_read_config_word(dev, offset, &old_value);
-+ if (err)
-+ goto out;
++ DPRINTK("%s\n",__FUNCTION__);
+
-+ old_state = (pci_power_t)(old_value & PCI_PM_CTRL_STATE_MASK);
-+ new_state = (pci_power_t)(new_value & PCI_PM_CTRL_STATE_MASK);
++ while (!kthread_should_stop()) {
++ wait_event_interruptible(
++ info->wq,
++ info->waiting_reqs || kthread_should_stop());
++ wait_event_interruptible(
++ pending_free_wq,
++ !list_empty(&pending_free) || kthread_should_stop());
+
-+ new_value &= PM_OK_BITS;
-+ if ((old_value & PM_OK_BITS) != new_value) {
-+ new_value = (old_value & ~PM_OK_BITS) | new_value;
-+ err = pci_write_config_word(dev, offset, new_value);
-+ if (err)
-+ goto out;
-+ }
++ info->waiting_reqs = 0;
++ smp_mb();
+
-+ /* Let pci core handle the power management change */
-+ dev_dbg(&dev->dev, "set power state to %x\n", new_state);
-+ err = pci_set_power_state(dev, new_state);
-+ if (err) {
-+ err = PCIBIOS_SET_FAILED;
-+ goto out;
++ if (scsiback_do_cmd_fn(info))
++ info->waiting_reqs = 1;
+ }
+
-+ /*
-+ * Device may lose PCI config info on D3->D0 transition. This
-+ * is a problem for some guests which will not reset BARs. Even
-+ * those that have a go will be foiled by our BAR-write handler
-+ * which will discard the write! Since Linux won't re-init
-+ * the config space automatically in all cases, we do it here.
-+ * Future: Should we re-initialise all first 64 bytes of config space?
-+ */
-+ if (new_state == PCI_D0 &&
-+ (old_state == PCI_D3hot || old_state == PCI_D3cold) &&
-+ !(old_value & PCI_PM_CTRL_NO_SOFT_RESET))
-+ pci_restore_bars(dev);
-+
-+ out:
-+ return err;
++ return 0;
+}
+
-+/* Ensure PMEs are disabled */
-+static void *pm_ctrl_init(struct pci_dev *dev, int offset)
++
++static int __init scsiback_init(void)
+{
-+ int err;
-+ u16 value;
++ int i, mmap_pages;
+
-+ err = pci_read_config_word(dev, offset, &value);
-+ if (err)
-+ goto out;
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+ if (value & PCI_PM_CTRL_PME_ENABLE) {
-+ value &= ~PCI_PM_CTRL_PME_ENABLE;
-+ err = pci_write_config_word(dev, offset, value);
-+ }
++ mmap_pages = vscsiif_reqs * VSCSIIF_SG_TABLESIZE;
+
-+ out:
-+ return ERR_PTR(err);
-+}
++ pending_reqs = kmalloc(sizeof(pending_reqs[0]) *
++ vscsiif_reqs, GFP_KERNEL);
++ pending_grant_handles = kmalloc(sizeof(pending_grant_handles[0]) *
++ mmap_pages, GFP_KERNEL);
++ pending_pages = alloc_empty_pages_and_pagevec(mmap_pages);
+
-+static struct config_field caplist_pm[] = {
-+ {
-+ .offset = PCI_PM_PMC,
-+ .size = 2,
-+ .u.w.read = pm_caps_read,
-+ },
-+ {
-+ .offset = PCI_PM_CTRL,
-+ .size = 2,
-+ .init = pm_ctrl_init,
-+ .u.w.read = pciback_read_config_word,
-+ .u.w.write = pm_ctrl_write,
-+ },
-+ {
-+ .offset = PCI_PM_PPB_EXTENSIONS,
-+ .size = 1,
-+ .u.b.read = pciback_read_config_byte,
-+ },
-+ {
-+ .offset = PCI_PM_DATA_REGISTER,
-+ .size = 1,
-+ .u.b.read = pciback_read_config_byte,
-+ },
-+ {
-+ .size = 0,
-+ },
-+};
++ if (!pending_reqs || !pending_grant_handles || !pending_pages)
++ goto out_of_memory;
+
-+struct pciback_config_capability pciback_config_capability_pm = {
-+ .capability = PCI_CAP_ID_PM,
-+ .fields = caplist_pm,
-+};
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_capability_vpd.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_capability_vpd.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_capability_vpd.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_capability_vpd.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,42 @@
-+/*
-+ * PCI Backend - Configuration space overlay for Vital Product Data
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ */
++ for (i = 0; i < mmap_pages; i++)
++ pending_grant_handles[i] = SCSIBACK_INVALID_HANDLE;
+
-+#include <linux/pci.h>
-+#include "conf_space.h"
-+#include "conf_space_capability.h"
++ if (scsiback_interface_init() < 0)
++ goto out_of_kmem;
+
-+static int vpd_address_write(struct pci_dev *dev, int offset, u16 value,
-+ void *data)
++ memset(pending_reqs, 0, sizeof(pending_reqs));
++ INIT_LIST_HEAD(&pending_free);
++
++ for (i = 0; i < vscsiif_reqs; i++)
++ list_add_tail(&pending_reqs[i].free_list, &pending_free);
++
++ if (scsiback_xenbus_init())
++ goto out_of_xenbus;
++
++ scsiback_emulation_init();
++
++ return 0;
++
++out_of_xenbus:
++ scsiback_xenbus_unregister();
++out_of_kmem:
++ scsiback_interface_exit();
++out_of_memory:
++ kfree(pending_reqs);
++ kfree(pending_grant_handles);
++ free_empty_pages_and_pagevec(pending_pages, mmap_pages);
++ printk(KERN_ERR "scsiback: %s: out of memory\n", __FUNCTION__);
++ return -ENOMEM;
++}
++
++static void __exit scsiback_exit(void)
+{
-+ /* Disallow writes to the vital product data */
-+ if (value & PCI_VPD_ADDR_F)
-+ return PCIBIOS_SET_FAILED;
-+ else
-+ return pci_write_config_word(dev, offset, value);
++ scsiback_xenbus_unregister();
++ scsiback_interface_exit();
++ kfree(pending_reqs);
++ kfree(pending_grant_handles);
++ free_empty_pages_and_pagevec(pending_pages, (vscsiif_reqs * VSCSIIF_SG_TABLESIZE));
++
+}
+
-+static struct config_field caplist_vpd[] = {
-+ {
-+ .offset = PCI_VPD_ADDR,
-+ .size = 2,
-+ .u.w.read = pciback_read_config_word,
-+ .u.w.write = vpd_address_write,
-+ },
-+ {
-+ .offset = PCI_VPD_DATA,
-+ .size = 4,
-+ .u.dw.read = pciback_read_config_dword,
-+ .u.dw.write = NULL,
-+ },
-+ {
-+ .size = 0,
-+ },
-+};
-+
-+struct pciback_config_capability pciback_config_capability_vpd = {
-+ .capability = PCI_CAP_ID_VPD,
-+ .fields = caplist_vpd,
-+};
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space.h tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space.h
---- pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,126 @@
++module_init(scsiback_init);
++module_exit(scsiback_exit);
++
++MODULE_DESCRIPTION("Xen SCSI backend driver");
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/scsiback/translate.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/scsiback/translate.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,168 @@
+/*
-+ * PCI Backend - Common data structures for overriding the configuration space
++ * Xen SCSI backend driver
+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
+
-+#ifndef __XEN_PCIBACK_CONF_SPACE_H__
-+#define __XEN_PCIBACK_CONF_SPACE_H__
-+
+#include <linux/list.h>
-+#include <linux/err.h>
++#include <linux/gfp.h>
+
-+/* conf_field_init can return an errno in a ptr with ERR_PTR() */
-+typedef void *(*conf_field_init) (struct pci_dev * dev, int offset);
-+typedef void (*conf_field_reset) (struct pci_dev * dev, int offset, void *data);
-+typedef void (*conf_field_free) (struct pci_dev * dev, int offset, void *data);
++#include "common.h"
+
-+typedef int (*conf_dword_write) (struct pci_dev * dev, int offset, u32 value,
-+ void *data);
-+typedef int (*conf_word_write) (struct pci_dev * dev, int offset, u16 value,
-+ void *data);
-+typedef int (*conf_byte_write) (struct pci_dev * dev, int offset, u8 value,
-+ void *data);
-+typedef int (*conf_dword_read) (struct pci_dev * dev, int offset, u32 * value,
-+ void *data);
-+typedef int (*conf_word_read) (struct pci_dev * dev, int offset, u16 * value,
-+ void *data);
-+typedef int (*conf_byte_read) (struct pci_dev * dev, int offset, u8 * value,
-+ void *data);
++/*
++ Initialize the translation entry list
++*/
++void scsiback_init_translation_table(struct vscsibk_info *info)
++{
++ INIT_LIST_HEAD(&info->v2p_entry_lists);
++ spin_lock_init(&info->v2p_lock);
++}
+
-+/* These are the fields within the configuration space which we
-+ * are interested in intercepting reads/writes to and changing their
-+ * values.
-+ */
-+struct config_field {
-+ unsigned int offset;
-+ unsigned int size;
-+ unsigned int mask;
-+ conf_field_init init;
-+ conf_field_reset reset;
-+ conf_field_free release;
-+ void (*clean) (struct config_field * field);
-+ union {
-+ struct {
-+ conf_dword_write write;
-+ conf_dword_read read;
-+ } dw;
-+ struct {
-+ conf_word_write write;
-+ conf_word_read read;
-+ } w;
-+ struct {
-+ conf_byte_write write;
-+ conf_byte_read read;
-+ } b;
-+ } u;
-+ struct list_head list;
-+};
+
-+struct config_field_entry {
-+ struct list_head list;
-+ struct config_field *field;
-+ unsigned int base_offset;
-+ void *data;
-+};
++/*
++ Add a new translation entry
++*/
++int scsiback_add_translation_entry(struct vscsibk_info *info,
++ struct scsi_device *sdev, struct ids_tuple *v)
++{
++ int err = 0;
++ struct v2p_entry *entry;
++ struct v2p_entry *new;
++ struct list_head *head = &(info->v2p_entry_lists);
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->v2p_lock, flags);
+
-+#define OFFSET(cfg_entry) ((cfg_entry)->base_offset+(cfg_entry)->field->offset)
++ /* Check double assignment to identical virtual ID */
++ list_for_each_entry(entry, head, l) {
++ if ((entry->v.chn == v->chn) &&
++ (entry->v.tgt == v->tgt) &&
++ (entry->v.lun == v->lun)) {
++ printk(KERN_WARNING "scsiback: Virtual ID is already used. "
++ "Assignment was not performed.\n");
++ err = -EEXIST;
++ goto out;
++ }
+
-+/* Add fields to a device - the add_fields macro expects to get a pointer to
-+ * the first entry in an array (of which the ending is marked by size==0)
-+ */
-+int pciback_config_add_field_offset(struct pci_dev *dev,
-+ struct config_field *field,
-+ unsigned int offset);
++ }
+
-+static inline int pciback_config_add_field(struct pci_dev *dev,
-+ struct config_field *field)
-+{
-+ return pciback_config_add_field_offset(dev, field, 0);
++ /* Create a new translation entry and add to the list */
++ if ((new = kmalloc(sizeof(struct v2p_entry), GFP_ATOMIC)) == NULL) {
++ printk(KERN_ERR "scsiback: %s: kmalloc() error.\n", __FUNCTION__);
++ err = -ENOMEM;
++ goto out;
++ }
++ new->v = *v;
++ new->sdev = sdev;
++ list_add_tail(&new->l, head);
++
++out:
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++ return err;
+}
+
-+static inline int pciback_config_add_fields(struct pci_dev *dev,
-+ struct config_field *field)
++
++/*
++ Delete the translation entry specfied
++*/
++int scsiback_del_translation_entry(struct vscsibk_info *info,
++ struct ids_tuple *v)
+{
-+ int i, err = 0;
-+ for (i = 0; field[i].size != 0; i++) {
-+ err = pciback_config_add_field(dev, &field[i]);
-+ if (err)
-+ break;
++ struct v2p_entry *entry;
++ struct list_head *head = &(info->v2p_entry_lists);
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->v2p_lock, flags);
++ /* Find out the translation entry specified */
++ list_for_each_entry(entry, head, l) {
++ if ((entry->v.chn == v->chn) &&
++ (entry->v.tgt == v->tgt) &&
++ (entry->v.lun == v->lun)) {
++ goto found;
++ }
+ }
-+ return err;
++
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++ return 1;
++
++found:
++ /* Delete the translation entry specfied */
++ scsi_device_put(entry->sdev);
++ list_del(&entry->l);
++ kfree(entry);
++
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++ return 0;
+}
+
-+static inline int pciback_config_add_fields_offset(struct pci_dev *dev,
-+ struct config_field *field,
-+ unsigned int offset)
++
++/*
++ Perform virtual to physical translation
++*/
++struct scsi_device *scsiback_do_translation(struct vscsibk_info *info,
++ struct ids_tuple *v)
+{
-+ int i, err = 0;
-+ for (i = 0; field[i].size != 0; i++) {
-+ err = pciback_config_add_field_offset(dev, &field[i], offset);
-+ if (err)
-+ break;
++ struct v2p_entry *entry;
++ struct list_head *head = &(info->v2p_entry_lists);
++ struct scsi_device *sdev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->v2p_lock, flags);
++ list_for_each_entry(entry, head, l) {
++ if ((entry->v.chn == v->chn) &&
++ (entry->v.tgt == v->tgt) &&
++ (entry->v.lun == v->lun)) {
++ sdev = entry->sdev;
++ goto out;
++ }
+ }
-+ return err;
++out:
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++ return sdev;
+}
+
-+/* Read/Write the real configuration space */
-+int pciback_read_config_byte(struct pci_dev *dev, int offset, u8 * value,
-+ void *data);
-+int pciback_read_config_word(struct pci_dev *dev, int offset, u16 * value,
-+ void *data);
-+int pciback_read_config_dword(struct pci_dev *dev, int offset, u32 * value,
-+ void *data);
-+int pciback_write_config_byte(struct pci_dev *dev, int offset, u8 value,
-+ void *data);
-+int pciback_write_config_word(struct pci_dev *dev, int offset, u16 value,
-+ void *data);
-+int pciback_write_config_dword(struct pci_dev *dev, int offset, u32 value,
-+ void *data);
+
-+int pciback_config_capability_init(void);
++/*
++ Release the translation entry specfied
++*/
++void scsiback_release_translation_entry(struct vscsibk_info *info)
++{
++ struct v2p_entry *entry, *tmp;
++ struct list_head *head = &(info->v2p_entry_lists);
++ unsigned long flags;
+
-+int pciback_config_header_add_fields(struct pci_dev *dev);
-+int pciback_config_capability_add_fields(struct pci_dev *dev);
++ spin_lock_irqsave(&info->v2p_lock, flags);
++ list_for_each_entry_safe(entry, tmp, head, l) {
++ scsi_device_put(entry->sdev);
++ list_del(&entry->l);
++ kfree(entry);
++ }
+
-+#endif /* __XEN_PCIBACK_CONF_SPACE_H__ */
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_header.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_header.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_header.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_header.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,323 @@
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++ return;
++
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/scsiback/xenbus.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/scsiback/xenbus.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,368 @@
+/*
-+ * PCI Backend - Handles the virtual fields in the configuration space headers.
++ * Xen SCSI backend driver
+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * Based on the blkback driver code.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
+
-+#include <linux/kernel.h>
-+#include <linux/pci.h>
-+#include "pciback.h"
-+#include "conf_space.h"
-+
-+struct pci_bar_info {
-+ u32 val;
-+ u32 len_val;
-+ int which;
-+};
++#include <stdarg.h>
++#include <linux/module.h>
++#include <linux/kthread.h>
++#include <scsi/scsi.h>
++#include <scsi/scsi_host.h>
++#include <scsi/scsi_device.h>
+
-+#define is_enable_cmd(value) ((value)&(PCI_COMMAND_MEMORY|PCI_COMMAND_IO))
-+#define is_master_cmd(value) ((value)&PCI_COMMAND_MASTER)
++#include "common.h"
+
-+static int command_write(struct pci_dev *dev, int offset, u16 value, void *data)
++struct backend_info
+{
-+ int err;
++ struct xenbus_device *dev;
++ struct vscsibk_info *info;
++};
+
-+ if (!dev->is_enabled && is_enable_cmd(value)) {
-+ if (unlikely(verbose_request))
-+ printk(KERN_DEBUG "pciback: %s: enable\n",
-+ pci_name(dev));
-+ err = pci_enable_device(dev);
-+ if (err)
-+ return err;
-+ } else if (dev->is_enabled && !is_enable_cmd(value)) {
-+ if (unlikely(verbose_request))
-+ printk(KERN_DEBUG "pciback: %s: disable\n",
-+ pci_name(dev));
-+ pci_disable_device(dev);
-+ }
+
-+ if (!dev->is_busmaster && is_master_cmd(value)) {
-+ if (unlikely(verbose_request))
-+ printk(KERN_DEBUG "pciback: %s: set bus master\n",
-+ pci_name(dev));
-+ pci_set_master(dev);
-+ }
++static int __vscsiif_name(struct backend_info *be, char *buf)
++{
++ struct xenbus_device *dev = be->dev;
++ unsigned int domid, id;
+
-+ if (value & PCI_COMMAND_INVALIDATE) {
-+ if (unlikely(verbose_request))
-+ printk(KERN_DEBUG
-+ "pciback: %s: enable memory-write-invalidate\n",
-+ pci_name(dev));
-+ err = pci_set_mwi(dev);
-+ if (err) {
-+ printk(KERN_WARNING
-+ "pciback: %s: cannot enable memory-write-invalidate (%d)\n",
-+ pci_name(dev), err);
-+ value &= ~PCI_COMMAND_INVALIDATE;
-+ }
-+ }
++ sscanf(dev->nodename, "backend/vscsi/%u/%u", &domid, &id);
++ snprintf(buf, TASK_COMM_LEN, "vscsi.%u.%u", be->info->domid, id);
+
-+ return pci_write_config_word(dev, offset, value);
++ return 0;
+}
+
-+static int rom_write(struct pci_dev *dev, int offset, u32 value, void *data)
++static int scsiback_map(struct backend_info *be)
+{
-+ struct pci_bar_info *bar = data;
++ struct xenbus_device *dev = be->dev;
++ unsigned long ring_ref;
++ unsigned int evtchn;
++ int err;
++ char name[TASK_COMM_LEN];
+
-+ if (unlikely(!bar)) {
-+ printk(KERN_WARNING "pciback: driver data not found for %s\n",
-+ pci_name(dev));
-+ return XEN_PCI_ERR_op_failed;
++ err = xenbus_gather(XBT_NIL, dev->otherend,
++ "ring-ref", "%lu", &ring_ref,
++ "event-channel", "%u", &evtchn, NULL);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "reading %s ring", dev->otherend);
++ return err;
+ }
+
-+ /* A write to obtain the length must happen as a 32-bit write.
-+ * This does not (yet) support writing individual bytes
-+ */
-+ if (value == ~PCI_ROM_ADDRESS_ENABLE)
-+ bar->which = 1;
-+ else {
-+ u32 tmpval;
-+ pci_read_config_dword(dev, offset, &tmpval);
-+ if (tmpval != bar->val && value == bar->val) {
-+ /* Allow restoration of bar value. */
-+ pci_write_config_dword(dev, offset, bar->val);
-+ }
-+ bar->which = 0;
++ err = scsiback_init_sring(be->info, ring_ref, evtchn);
++ if (err)
++ return err;
++
++ err = __vscsiif_name(be, name);
++ if (err) {
++ xenbus_dev_error(dev, err, "get scsiback dev name");
++ return err;
+ }
+
-+ /* Do we need to support enabling/disabling the rom address here? */
++ be->info->kthread = kthread_run(scsiback_schedule, be->info, name);
++ if (IS_ERR(be->info->kthread)) {
++ err = PTR_ERR(be->info->kthread);
++ be->info->kthread = NULL;
++ xenbus_dev_error(be->dev, err, "start vscsiif");
++ return err;
++ }
+
+ return 0;
+}
+
-+/* For the BARs, only allow writes which write ~0 or
-+ * the correct resource information
-+ * (Needed for when the driver probes the resource usage)
-+ */
-+static int bar_write(struct pci_dev *dev, int offset, u32 value, void *data)
++
++struct scsi_device *scsiback_get_scsi_device(struct ids_tuple *phy)
+{
-+ struct pci_bar_info *bar = data;
++ struct Scsi_Host *shost;
++ struct scsi_device *sdev = NULL;
+
-+ if (unlikely(!bar)) {
-+ printk(KERN_WARNING "pciback: driver data not found for %s\n",
-+ pci_name(dev));
-+ return XEN_PCI_ERR_op_failed;
++ shost = scsi_host_lookup(phy->hst);
++ if (IS_ERR(shost)) {
++ printk(KERN_ERR "scsiback: host%d doesn't exist.\n",
++ phy->hst);
++ return NULL;
+ }
-+
-+ /* A write to obtain the length must happen as a 32-bit write.
-+ * This does not (yet) support writing individual bytes
-+ */
-+ if (value == ~0)
-+ bar->which = 1;
-+ else {
-+ u32 tmpval;
-+ pci_read_config_dword(dev, offset, &tmpval);
-+ if (tmpval != bar->val && value == bar->val) {
-+ /* Allow restoration of bar value. */
-+ pci_write_config_dword(dev, offset, bar->val);
-+ }
-+ bar->which = 0;
++ sdev = scsi_device_lookup(shost, phy->chn, phy->tgt, phy->lun);
++ if (!sdev) {
++ printk(KERN_ERR "scsiback: %d:%d:%d:%d doesn't exist.\n",
++ phy->hst, phy->chn, phy->tgt, phy->lun);
++ scsi_host_put(shost);
++ return NULL;
+ }
+
-+ return 0;
++ scsi_host_put(shost);
++ return (sdev);
+}
+
-+static int bar_read(struct pci_dev *dev, int offset, u32 * value, void *data)
++#define VSCSIBACK_OP_ADD_OR_DEL_LUN 1
++#define VSCSIBACK_OP_UPDATEDEV_STATE 2
++
++
++static void scsiback_do_lun_hotplug(struct backend_info *be, int op)
+{
-+ struct pci_bar_info *bar = data;
++ int i, err = 0;
++ struct ids_tuple phy, vir;
++ int device_state;
++ char str[64], state_str[64];
++ char **dir;
++ unsigned int dir_n = 0;
++ struct xenbus_device *dev = be->dev;
++ struct scsi_device *sdev;
+
-+ if (unlikely(!bar)) {
-+ printk(KERN_WARNING "pciback: driver data not found for %s\n",
-+ pci_name(dev));
-+ return XEN_PCI_ERR_op_failed;
-+ }
++ dir = xenbus_directory(XBT_NIL, dev->nodename, "vscsi-devs", &dir_n);
++ if (IS_ERR(dir))
++ return;
+
-+ *value = bar->which ? bar->len_val : bar->val;
++ for (i = 0; i < dir_n; i++) {
++
++ /* read status */
++ snprintf(state_str, sizeof(state_str), "vscsi-devs/%s/state", dir[i]);
++ err = xenbus_scanf(XBT_NIL, dev->nodename, state_str, "%u",
++ &device_state);
++ if (XENBUS_EXIST_ERR(err))
++ continue;
+
-+ return 0;
-+}
++ /* physical SCSI device */
++ snprintf(str, sizeof(str), "vscsi-devs/%s/p-dev", dir[i]);
++ err = xenbus_scanf(XBT_NIL, dev->nodename, str,
++ "%u:%u:%u:%u", &phy.hst, &phy.chn, &phy.tgt, &phy.lun);
++ if (XENBUS_EXIST_ERR(err)) {
++ xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateClosed);
++ continue;
++ }
+
-+static inline void read_dev_bar(struct pci_dev *dev,
-+ struct pci_bar_info *bar_info, int offset,
-+ u32 len_mask)
-+{
-+ pci_read_config_dword(dev, offset, &bar_info->val);
-+ pci_write_config_dword(dev, offset, len_mask);
-+ pci_read_config_dword(dev, offset, &bar_info->len_val);
-+ pci_write_config_dword(dev, offset, bar_info->val);
-+}
++ /* virtual SCSI device */
++ snprintf(str, sizeof(str), "vscsi-devs/%s/v-dev", dir[i]);
++ err = xenbus_scanf(XBT_NIL, dev->nodename, str,
++ "%u:%u:%u:%u", &vir.hst, &vir.chn, &vir.tgt, &vir.lun);
++ if (XENBUS_EXIST_ERR(err)) {
++ xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateClosed);
++ continue;
++ }
+
-+static void *bar_init(struct pci_dev *dev, int offset)
-+{
-+ struct pci_bar_info *bar = kmalloc(sizeof(*bar), GFP_KERNEL);
++ switch (op) {
++ case VSCSIBACK_OP_ADD_OR_DEL_LUN:
++ if (device_state == XenbusStateInitialising) {
++ sdev = scsiback_get_scsi_device(&phy);
++ if (!sdev)
++ xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateClosed);
++ else {
++ err = scsiback_add_translation_entry(be->info, sdev, &vir);
++ if (!err) {
++ if (xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateInitialised)) {
++ printk(KERN_ERR "scsiback: xenbus_printf error %s\n", state_str);
++ scsiback_del_translation_entry(be->info, &vir);
++ }
++ } else {
++ scsi_device_put(sdev);
++ xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateClosed);
++ }
++ }
++ }
+
-+ if (!bar)
-+ return ERR_PTR(-ENOMEM);
++ if (device_state == XenbusStateClosing) {
++ if (!scsiback_del_translation_entry(be->info, &vir)) {
++ if (xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateClosed))
++ printk(KERN_ERR "scsiback: xenbus_printf error %s\n", state_str);
++ }
++ }
++ break;
+
-+ read_dev_bar(dev, bar, offset, ~0);
-+ bar->which = 0;
++ case VSCSIBACK_OP_UPDATEDEV_STATE:
++ if (device_state == XenbusStateInitialised) {
++ /* modify vscsi-devs/dev-x/state */
++ if (xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateConnected)) {
++ printk(KERN_ERR "scsiback: xenbus_printf error %s\n", state_str);
++ scsiback_del_translation_entry(be->info, &vir);
++ xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateClosed);
++ }
++ }
++ break;
++ /*When it is necessary, processing is added here.*/
++ default:
++ break;
++ }
++ }
+
-+ return bar;
++ kfree(dir);
++ return ;
+}
+
-+static void *rom_init(struct pci_dev *dev, int offset)
++
++static void scsiback_frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state)
+{
-+ struct pci_bar_info *bar = kmalloc(sizeof(*bar), GFP_KERNEL);
++ struct backend_info *be = dev->dev.driver_data;
++ int err;
+
-+ if (!bar)
-+ return ERR_PTR(-ENOMEM);
++ switch (frontend_state) {
++ case XenbusStateInitialising:
++ break;
++ case XenbusStateInitialised:
++ err = scsiback_map(be);
++ if (err)
++ break;
+
-+ read_dev_bar(dev, bar, offset, ~PCI_ROM_ADDRESS_ENABLE);
-+ bar->which = 0;
++ scsiback_do_lun_hotplug(be, VSCSIBACK_OP_ADD_OR_DEL_LUN);
++ xenbus_switch_state(dev, XenbusStateConnected);
+
-+ return bar;
-+}
++ break;
++ case XenbusStateConnected:
+
-+static void bar_reset(struct pci_dev *dev, int offset, void *data)
-+{
-+ struct pci_bar_info *bar = data;
++ scsiback_do_lun_hotplug(be, VSCSIBACK_OP_UPDATEDEV_STATE);
+
-+ bar->which = 0;
-+}
++ if (dev->state == XenbusStateConnected)
++ break;
+
-+static void bar_release(struct pci_dev *dev, int offset, void *data)
-+{
-+ kfree(data);
++ xenbus_switch_state(dev, XenbusStateConnected);
++
++ break;
++
++ case XenbusStateClosing:
++ scsiback_disconnect(be->info);
++ xenbus_switch_state(dev, XenbusStateClosing);
++ break;
++
++ case XenbusStateClosed:
++ xenbus_switch_state(dev, XenbusStateClosed);
++ if (xenbus_dev_is_online(dev))
++ break;
++ /* fall through if not online */
++ case XenbusStateUnknown:
++ device_unregister(&dev->dev);
++ break;
++
++ case XenbusStateReconfiguring:
++ scsiback_do_lun_hotplug(be, VSCSIBACK_OP_ADD_OR_DEL_LUN);
++
++ xenbus_switch_state(dev, XenbusStateReconfigured);
++
++ break;
++
++ default:
++ xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
++ frontend_state);
++ break;
++ }
+}
+
-+static int interrupt_read(struct pci_dev *dev, int offset, u8 * value,
-+ void *data)
++
++static int scsiback_remove(struct xenbus_device *dev)
+{
-+ *value = (u8) dev->irq;
++ struct backend_info *be = dev->dev.driver_data;
++
++ if (be->info) {
++ scsiback_disconnect(be->info);
++ scsiback_release_translation_entry(be->info);
++ scsiback_free(be->info);
++ be->info = NULL;
++ }
++
++ kfree(be);
++ dev->dev.driver_data = NULL;
+
+ return 0;
+}
+
-+static int bist_write(struct pci_dev *dev, int offset, u8 value, void *data)
++
++static int scsiback_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
+{
-+ u8 cur_value;
+ int err;
+
-+ err = pci_read_config_byte(dev, offset, &cur_value);
++ struct backend_info *be = kzalloc(sizeof(struct backend_info),
++ GFP_KERNEL);
++
++ DPRINTK("%p %d\n", dev, dev->otherend_id);
++
++ if (!be) {
++ xenbus_dev_fatal(dev, -ENOMEM,
++ "allocating backend structure");
++ return -ENOMEM;
++ }
++ be->dev = dev;
++ dev->dev.driver_data = be;
++
++ be->info = vscsibk_info_alloc(dev->otherend_id);
++ if (IS_ERR(be->info)) {
++ err = PTR_ERR(be->info);
++ be->info = NULL;
++ xenbus_dev_fatal(dev, err, "creating scsihost interface");
++ goto fail;
++ }
++
++ be->info->dev = dev;
++ be->info->irq = 0;
++
++ scsiback_init_translation_table(be->info);
++
++ err = xenbus_switch_state(dev, XenbusStateInitWait);
+ if (err)
-+ goto out;
++ goto fail;
+
-+ if ((cur_value & ~PCI_BIST_START) == (value & ~PCI_BIST_START)
-+ || value == PCI_BIST_START)
-+ err = pci_write_config_byte(dev, offset, value);
++ return 0;
++
++
++fail:
++ printk(KERN_WARNING "scsiback: %s failed\n",__FUNCTION__);
++ scsiback_remove(dev);
+
-+ out:
+ return err;
+}
+
-+static struct config_field header_common[] = {
-+ {
-+ .offset = PCI_COMMAND,
-+ .size = 2,
-+ .u.w.read = pciback_read_config_word,
-+ .u.w.write = command_write,
-+ },
-+ {
-+ .offset = PCI_INTERRUPT_LINE,
-+ .size = 1,
-+ .u.b.read = interrupt_read,
-+ },
-+ {
-+ .offset = PCI_INTERRUPT_PIN,
-+ .size = 1,
-+ .u.b.read = pciback_read_config_byte,
-+ },
-+ {
-+ /* Any side effects of letting driver domain control cache line? */
-+ .offset = PCI_CACHE_LINE_SIZE,
-+ .size = 1,
-+ .u.b.read = pciback_read_config_byte,
-+ .u.b.write = pciback_write_config_byte,
-+ },
-+ {
-+ .offset = PCI_LATENCY_TIMER,
-+ .size = 1,
-+ .u.b.read = pciback_read_config_byte,
-+ },
-+ {
-+ .offset = PCI_BIST,
-+ .size = 1,
-+ .u.b.read = pciback_read_config_byte,
-+ .u.b.write = bist_write,
-+ },
-+ {
-+ .size = 0,
-+ },
++
++static struct xenbus_device_id scsiback_ids[] = {
++ { "vscsi" },
++ { "" }
+};
+
-+#define CFG_FIELD_BAR(reg_offset) \
-+ { \
-+ .offset = reg_offset, \
-+ .size = 4, \
-+ .init = bar_init, \
-+ .reset = bar_reset, \
-+ .release = bar_release, \
-+ .u.dw.read = bar_read, \
-+ .u.dw.write = bar_write, \
-+ }
++static struct xenbus_driver scsiback = {
++ .name = "vscsi",
++ .owner = THIS_MODULE,
++ .ids = scsiback_ids,
++ .probe = scsiback_probe,
++ .remove = scsiback_remove,
++ .otherend_changed = scsiback_frontend_changed
++};
+
-+#define CFG_FIELD_ROM(reg_offset) \
-+ { \
-+ .offset = reg_offset, \
-+ .size = 4, \
-+ .init = rom_init, \
-+ .reset = bar_reset, \
-+ .release = bar_release, \
-+ .u.dw.read = bar_read, \
-+ .u.dw.write = rom_write, \
-+ }
++int scsiback_xenbus_init(void)
++{
++ return xenbus_register_backend(&scsiback);
++}
+
-+static struct config_field header_0[] = {
-+ CFG_FIELD_BAR(PCI_BASE_ADDRESS_0),
-+ CFG_FIELD_BAR(PCI_BASE_ADDRESS_1),
-+ CFG_FIELD_BAR(PCI_BASE_ADDRESS_2),
-+ CFG_FIELD_BAR(PCI_BASE_ADDRESS_3),
-+ CFG_FIELD_BAR(PCI_BASE_ADDRESS_4),
-+ CFG_FIELD_BAR(PCI_BASE_ADDRESS_5),
-+ CFG_FIELD_ROM(PCI_ROM_ADDRESS),
-+ {
-+ .size = 0,
-+ },
-+};
++void scsiback_xenbus_unregister(void)
++{
++ xenbus_unregister_driver(&scsiback);
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/scsifront/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/scsifront/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,3 @@
+
-+static struct config_field header_1[] = {
-+ CFG_FIELD_BAR(PCI_BASE_ADDRESS_0),
-+ CFG_FIELD_BAR(PCI_BASE_ADDRESS_1),
-+ CFG_FIELD_ROM(PCI_ROM_ADDRESS1),
-+ {
-+ .size = 0,
-+ },
++obj-$(CONFIG_XEN_SCSI_FRONTEND) := xenscsi.o
++xenscsi-objs := scsifront.o xenbus.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/scsifront/common.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/scsifront/common.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,129 @@
++/*
++ * Xen SCSI frontend driver
++ *
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_DRIVERS_SCSIFRONT_H__
++#define __XEN_DRIVERS_SCSIFRONT_H__
++
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/kernel.h>
++#include <linux/device.h>
++#include <linux/kthread.h>
++#include <linux/wait.h>
++#include <linux/interrupt.h>
++#include <linux/spinlock.h>
++#include <linux/sched.h>
++#include <linux/blkdev.h>
++#include <scsi/scsi_cmnd.h>
++#include <scsi/scsi_device.h>
++#include <scsi/scsi.h>
++#include <scsi/scsi_host.h>
++#include <xen/xenbus.h>
++#include <xen/gnttab.h>
++#include <xen/evtchn.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/io/ring.h>
++#include <xen/interface/io/vscsiif.h>
++#include <asm/delay.h>
++
++
++#define GRANT_INVALID_REF 0
++#define VSCSI_IN_ABORT 1
++#define VSCSI_IN_RESET 2
++
++/* tuning point*/
++#define VSCSIIF_DEFAULT_CMD_PER_LUN 10
++#define VSCSIIF_MAX_TARGET 64
++#define VSCSIIF_MAX_LUN 255
++
++#define VSCSIIF_RING_SIZE \
++ __RING_SIZE((struct vscsiif_sring *)0, PAGE_SIZE)
++#define VSCSIIF_MAX_REQS VSCSIIF_RING_SIZE
++
++struct vscsifrnt_shadow {
++ uint16_t next_free;
++
++ /* command between backend and frontend
++ * VSCSIIF_ACT_SCSI_CDB or VSCSIIF_ACT_SCSI_RESET */
++ unsigned char act;
++
++ /* do reset function */
++ wait_queue_head_t wq_reset; /* reset work queue */
++ int wait_reset; /* reset work queue condition */
++ int32_t rslt_reset; /* reset response status */
++ /* (SUCESS or FAILED) */
++
++ /* for DMA_TO_DEVICE(1), DMA_FROM_DEVICE(2), DMA_NONE(3)
++ requests */
++ unsigned int sc_data_direction;
++
++ /* Number of pieces of scatter-gather */
++ unsigned int nr_segments;
++
++ /* requested struct scsi_cmnd is stored from kernel */
++ unsigned long req_scsi_cmnd;
++ int gref[VSCSIIF_SG_TABLESIZE];
+};
+
-+int pciback_config_header_add_fields(struct pci_dev *dev)
-+{
-+ int err;
++struct vscsifrnt_info {
++ struct xenbus_device *dev;
+
-+ err = pciback_config_add_fields(dev, header_common);
-+ if (err)
-+ goto out;
++ struct Scsi_Host *host;
+
-+ switch (dev->hdr_type) {
-+ case PCI_HEADER_TYPE_NORMAL:
-+ err = pciback_config_add_fields(dev, header_0);
-+ break;
++ spinlock_t io_lock;
++ spinlock_t shadow_lock;
++ unsigned int evtchn;
++ unsigned int irq;
+
-+ case PCI_HEADER_TYPE_BRIDGE:
-+ err = pciback_config_add_fields(dev, header_1);
-+ break;
++ grant_ref_t ring_ref;
++ struct vscsiif_front_ring ring;
++ struct vscsiif_response ring_res;
+
-+ default:
-+ err = -EINVAL;
-+ printk(KERN_ERR "pciback: %s: Unsupported header type %d!\n",
-+ pci_name(dev), dev->hdr_type);
-+ break;
-+ }
++ struct vscsifrnt_shadow shadow[VSCSIIF_MAX_REQS];
++ uint32_t shadow_free;
+
-+ out:
-+ return err;
-+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_quirks.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_quirks.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_quirks.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_quirks.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,126 @@
++ struct task_struct *kthread;
++ wait_queue_head_t wq;
++ unsigned int waiting_resp;
++
++};
++
++#define DPRINTK(_f, _a...) \
++ pr_debug("(file=%s, line=%d) " _f, \
++ __FILE__ , __LINE__ , ## _a )
++
++int scsifront_xenbus_init(void);
++void scsifront_xenbus_unregister(void);
++int scsifront_schedule(void *data);
++irqreturn_t scsifront_intr(int irq, void *dev_id, struct pt_regs *ptregs);
++int scsifront_cmd_done(struct vscsifrnt_info *info);
++
++
++#endif /* __XEN_DRIVERS_SCSIFRONT_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/scsifront/scsifront.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/scsifront/scsifront.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,511 @@
+/*
-+ * PCI Backend - Handle special overlays for broken devices.
++ * Xen SCSI frontend driver
+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ * Author: Chris Bookholt <hap10 at epoch.ncsc.mil>
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
++
+
-+#include <linux/kernel.h>
-+#include <linux/pci.h>
-+#include "pciback.h"
-+#include "conf_space.h"
-+#include "conf_space_quirks.h"
-+
-+LIST_HEAD(pciback_quirks);
++#include <linux/version.h>
++#include "common.h"
+
-+struct pciback_config_quirk *pciback_find_quirk(struct pci_dev *dev)
++static int get_id_from_freelist(struct vscsifrnt_info *info)
+{
-+ struct pciback_config_quirk *tmp_quirk;
++ unsigned long flags;
++ uint32_t free;
+
-+ list_for_each_entry(tmp_quirk, &pciback_quirks, quirks_list)
-+ if (pci_match_id(&tmp_quirk->devid, dev))
-+ goto out;
-+ tmp_quirk = NULL;
-+ printk(KERN_DEBUG
-+ "quirk didn't match any device pciback knows about\n");
-+ out:
-+ return tmp_quirk;
-+}
++ spin_lock_irqsave(&info->shadow_lock, flags);
+
-+static inline void register_quirk(struct pciback_config_quirk *quirk)
-+{
-+ list_add_tail(&quirk->quirks_list, &pciback_quirks);
++ free = info->shadow_free;
++ BUG_ON(free > VSCSIIF_MAX_REQS);
++ info->shadow_free = info->shadow[free].next_free;
++ info->shadow[free].next_free = 0x0fff;
++
++ info->shadow[free].wait_reset = 0;
++
++ spin_unlock_irqrestore(&info->shadow_lock, flags);
++
++ return free;
+}
+
-+int pciback_field_is_dup(struct pci_dev *dev, unsigned int reg)
++static void add_id_to_freelist(struct vscsifrnt_info *info, uint32_t id)
+{
-+ int ret = 0;
-+ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
-+ struct config_field_entry *cfg_entry;
++ unsigned long flags;
+
-+ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
-+ if ( OFFSET(cfg_entry) == reg) {
-+ ret = 1;
-+ break;
-+ }
-+ }
-+ return ret;
++ spin_lock_irqsave(&info->shadow_lock, flags);
++
++ info->shadow[id].next_free = info->shadow_free;
++ info->shadow[id].req_scsi_cmnd = 0;
++ info->shadow_free = id;
++
++ spin_unlock_irqrestore(&info->shadow_lock, flags);
+}
+
-+int pciback_config_quirks_add_field(struct pci_dev *dev, struct config_field
-+ *field)
++
++struct vscsiif_request * scsifront_pre_request(struct vscsifrnt_info *info)
+{
-+ int err = 0;
++ struct vscsiif_front_ring *ring = &(info->ring);
++ vscsiif_request_t *ring_req;
++ uint32_t id;
+
-+ switch (field->size) {
-+ case 1:
-+ field->u.b.read = pciback_read_config_byte;
-+ field->u.b.write = pciback_write_config_byte;
-+ break;
-+ case 2:
-+ field->u.w.read = pciback_read_config_word;
-+ field->u.w.write = pciback_write_config_word;
-+ break;
-+ case 4:
-+ field->u.dw.read = pciback_read_config_dword;
-+ field->u.dw.write = pciback_write_config_dword;
-+ break;
-+ default:
-+ err = -EINVAL;
-+ goto out;
-+ }
++ ring_req = RING_GET_REQUEST(&(info->ring), ring->req_prod_pvt);
+
-+ pciback_config_add_field(dev, field);
++ ring->req_prod_pvt++;
++
++ id = get_id_from_freelist(info); /* use id by response */
++ ring_req->rqid = (uint16_t)id;
+
-+ out:
-+ return err;
++ return ring_req;
+}
+
-+int pciback_config_quirks_init(struct pci_dev *dev)
-+{
-+ struct pciback_config_quirk *quirk;
-+ int ret = 0;
+
-+ quirk = kzalloc(sizeof(*quirk), GFP_ATOMIC);
-+ if (!quirk) {
-+ ret = -ENOMEM;
-+ goto out;
-+ }
++static void scsifront_notify_work(struct vscsifrnt_info *info)
++{
++ info->waiting_resp = 1;
++ wake_up(&info->wq);
++}
+
-+ quirk->devid.vendor = dev->vendor;
-+ quirk->devid.device = dev->device;
-+ quirk->devid.subvendor = dev->subsystem_vendor;
-+ quirk->devid.subdevice = dev->subsystem_device;
-+ quirk->devid.class = 0;
-+ quirk->devid.class_mask = 0;
-+ quirk->devid.driver_data = 0UL;
+
-+ quirk->pdev = dev;
++static void scsifront_do_request(struct vscsifrnt_info *info)
++{
++ struct vscsiif_front_ring *ring = &(info->ring);
++ unsigned int irq = info->irq;
++ int notify;
+
-+ register_quirk(quirk);
-+ out:
-+ return ret;
++ RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(ring, notify);
++ if (notify)
++ notify_remote_via_irq(irq);
+}
+
-+void pciback_config_field_free(struct config_field *field)
++irqreturn_t scsifront_intr(int irq, void *dev_id, struct pt_regs *ptregs)
+{
-+ kfree(field);
++ scsifront_notify_work((struct vscsifrnt_info *)dev_id);
++ return IRQ_HANDLED;
+}
+
-+int pciback_config_quirk_release(struct pci_dev *dev)
++
++static void scsifront_gnttab_done(struct vscsifrnt_shadow *s, uint32_t id)
+{
-+ struct pciback_config_quirk *quirk;
-+ int ret = 0;
++ int i;
+
-+ quirk = pciback_find_quirk(dev);
-+ if (!quirk) {
-+ ret = -ENXIO;
-+ goto out;
-+ }
++ if (s->sc_data_direction == DMA_NONE)
++ return;
+
-+ list_del(&quirk->quirks_list);
-+ kfree(quirk);
++ if (s->nr_segments) {
++ for (i = 0; i < s->nr_segments; i++) {
++ if (unlikely(gnttab_query_foreign_access(
++ s->gref[i]) != 0)) {
++ printk(KERN_ALERT "scsifront: "
++ "grant still in use by backend.\n");
++ BUG();
++ }
++ gnttab_end_foreign_access(s->gref[i], 0UL);
++ }
++ }
+
-+ out:
-+ return ret;
++ return;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_quirks.h tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_quirks.h
---- pristine-linux-2.6.18.2/drivers/xen/pciback/conf_space_quirks.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/conf_space_quirks.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,35 @@
-+/*
-+ * PCI Backend - Data structures for special overlays for broken devices.
-+ *
-+ * Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ * Chris Bookholt <hap10 at epoch.ncsc.mil>
-+ */
+
-+#ifndef __XEN_PCIBACK_CONF_SPACE_QUIRKS_H__
-+#define __XEN_PCIBACK_CONF_SPACE_QUIRKS_H__
+
-+#include <linux/pci.h>
-+#include <linux/list.h>
++static void scsifront_cdb_cmd_done(struct vscsifrnt_info *info,
++ vscsiif_response_t *ring_res)
++{
++ struct scsi_cmnd *sc;
++ uint32_t id;
++ uint8_t sense_len;
+
-+struct pciback_config_quirk {
-+ struct list_head quirks_list;
-+ struct pci_device_id devid;
-+ struct pci_dev *pdev;
-+};
++ id = ring_res->rqid;
++ sc = (struct scsi_cmnd *)info->shadow[id].req_scsi_cmnd;
+
-+struct pciback_config_quirk *pciback_find_quirk(struct pci_dev *dev);
++ if (sc == NULL)
++ BUG();
+
-+int pciback_config_quirks_add_field(struct pci_dev *dev, struct config_field
-+ *field);
++ scsifront_gnttab_done(&info->shadow[id], id);
++ add_id_to_freelist(info, id);
+
-+int pciback_config_quirks_remove_field(struct pci_dev *dev, int reg);
++ sc->result = ring_res->rslt;
++ sc->resid = ring_res->residual_len;
+
-+int pciback_config_quirks_init(struct pci_dev *dev);
++ if (ring_res->sense_len > VSCSIIF_SENSE_BUFFERSIZE)
++ sense_len = VSCSIIF_SENSE_BUFFERSIZE;
++ else
++ sense_len = ring_res->sense_len;
+
-+void pciback_config_field_free(struct config_field *field);
++ if (sense_len)
++ memcpy(sc->sense_buffer, ring_res->sense_buffer, sense_len);
+
-+int pciback_config_quirk_release(struct pci_dev *dev);
++ sc->scsi_done(sc);
+
-+int pciback_field_is_dup(struct pci_dev *dev, unsigned int reg);
++ return;
++}
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/controller.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/controller.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/controller.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/controller.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,404 @@
-+/*
-+ * Copyright (C) 2007 Hewlett-Packard Development Company, L.P.
-+ * Alex Williamson <alex.williamson at hp.com>
-+ *
-+ * PCI "Controller" Backend - virtualize PCI bus topology based on PCI
-+ * controllers. Devices under the same PCI controller are exposed on the
-+ * same virtual domain:bus. Within a bus, device slots are virtualized
-+ * to compact the bus.
-+ *
-+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-+ * This program is free software; you can redistribute it and/or modify
-+ * it under the terms of the GNU General Public License as published by
-+ * the Free Software Foundation; either version 2 of the License, or
-+ * (at your option) any later version.
-+ *
-+ * This program is distributed in the hope that it will be useful,
-+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ * GNU General Public License for more details.
-+ *
-+ * You should have received a copy of the GNU General Public License
-+ * along with this program; if not, write to the Free Software
-+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
-+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-+ */
+
-+#include <linux/acpi.h>
-+#include <linux/list.h>
-+#include <linux/pci.h>
-+#include <linux/spinlock.h>
-+#include "pciback.h"
++static void scsifront_sync_cmd_done(struct vscsifrnt_info *info,
++ vscsiif_response_t *ring_res)
++{
++ uint16_t id = ring_res->rqid;
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->shadow_lock, flags);
++ info->shadow[id].wait_reset = 1;
++ info->shadow[id].rslt_reset = ring_res->rslt;
++ spin_unlock_irqrestore(&info->shadow_lock, flags);
+
-+#define PCI_MAX_BUSSES 255
-+#define PCI_MAX_SLOTS 32
++ wake_up(&(info->shadow[id].wq_reset));
++}
+
-+struct controller_dev_entry {
-+ struct list_head list;
-+ struct pci_dev *dev;
-+ unsigned int devfn;
-+};
+
-+struct controller_list_entry {
-+ struct list_head list;
-+ struct pci_controller *controller;
-+ unsigned int domain;
-+ unsigned int bus;
-+ unsigned int next_devfn;
-+ struct list_head dev_list;
-+};
++int scsifront_cmd_done(struct vscsifrnt_info *info)
++{
++ vscsiif_response_t *ring_res;
+
-+struct controller_dev_data {
-+ struct list_head list;
-+ unsigned int next_domain;
-+ unsigned int next_bus;
-+ spinlock_t lock;
-+};
++ RING_IDX i, rp;
++ int more_to_do = 0;
++ unsigned long flags;
+
-+struct walk_info {
-+ struct pciback_device *pdev;
-+ int resource_count;
-+ int root_num;
-+};
++ spin_lock_irqsave(&info->io_lock, flags);
+
-+struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
-+ unsigned int domain, unsigned int bus,
-+ unsigned int devfn)
-+{
-+ struct controller_dev_data *dev_data = pdev->pci_dev_data;
-+ struct controller_dev_entry *dev_entry;
-+ struct controller_list_entry *cntrl_entry;
-+ struct pci_dev *dev = NULL;
-+ unsigned long flags;
++ rp = info->ring.sring->rsp_prod;
++ rmb();
++ for (i = info->ring.rsp_cons; i != rp; i++) {
++
++ ring_res = RING_GET_RESPONSE(&info->ring, i);
+
-+ spin_lock_irqsave(&dev_data->lock, flags);
++ if (info->shadow[ring_res->rqid].act == VSCSIIF_ACT_SCSI_CDB)
++ scsifront_cdb_cmd_done(info, ring_res);
++ else
++ scsifront_sync_cmd_done(info, ring_res);
++ }
+
-+ list_for_each_entry(cntrl_entry, &dev_data->list, list) {
-+ if (cntrl_entry->domain != domain ||
-+ cntrl_entry->bus != bus)
-+ continue;
++ info->ring.rsp_cons = i;
+
-+ list_for_each_entry(dev_entry, &cntrl_entry->dev_list, list) {
-+ if (devfn == dev_entry->devfn) {
-+ dev = dev_entry->dev;
-+ goto found;
-+ }
-+ }
++ if (i != info->ring.req_prod_pvt) {
++ RING_FINAL_CHECK_FOR_RESPONSES(&info->ring, more_to_do);
++ } else {
++ info->ring.sring->rsp_event = i + 1;
+ }
-+found:
-+ spin_unlock_irqrestore(&dev_data->lock, flags);
+
-+ return dev;
-+}
++ spin_unlock_irqrestore(&info->io_lock, flags);
+
-+int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
-+{
-+ struct controller_dev_data *dev_data = pdev->pci_dev_data;
-+ struct controller_dev_entry *dev_entry;
-+ struct controller_list_entry *cntrl_entry;
-+ struct pci_controller *dev_controller = PCI_CONTROLLER(dev);
-+ unsigned long flags;
-+ int ret = 0, found = 0;
+
-+ spin_lock_irqsave(&dev_data->lock, flags);
++ /* Yield point for this unbounded loop. */
++ cond_resched();
+
-+ /* Look to see if we already have a domain:bus for this controller */
-+ list_for_each_entry(cntrl_entry, &dev_data->list, list) {
-+ if (cntrl_entry->controller == dev_controller) {
-+ found = 1;
-+ break;
-+ }
-+ }
++ return more_to_do;
++}
+
-+ if (!found) {
-+ cntrl_entry = kmalloc(sizeof(*cntrl_entry), GFP_ATOMIC);
-+ if (!cntrl_entry) {
-+ ret = -ENOMEM;
-+ goto out;
-+ }
+
-+ cntrl_entry->controller = dev_controller;
-+ cntrl_entry->next_devfn = PCI_DEVFN(0, 0);
+
-+ cntrl_entry->domain = dev_data->next_domain;
-+ cntrl_entry->bus = dev_data->next_bus++;
-+ if (dev_data->next_bus > PCI_MAX_BUSSES) {
-+ dev_data->next_domain++;
-+ dev_data->next_bus = 0;
-+ }
+
-+ INIT_LIST_HEAD(&cntrl_entry->dev_list);
++int scsifront_schedule(void *data)
++{
++ struct vscsifrnt_info *info = (struct vscsifrnt_info *)data;
+
-+ list_add_tail(&cntrl_entry->list, &dev_data->list);
-+ }
++ while (!kthread_should_stop()) {
++ wait_event_interruptible(
++ info->wq,
++ info->waiting_resp || kthread_should_stop());
+
-+ if (PCI_SLOT(cntrl_entry->next_devfn) > PCI_MAX_SLOTS) {
-+ /*
-+ * While it seems unlikely, this can actually happen if
-+ * a controller has P2P bridges under it.
-+ */
-+ xenbus_dev_fatal(pdev->xdev, -ENOSPC, "Virtual bus %04x:%02x "
-+ "is full, no room to export %04x:%02x:%02x.%x",
-+ cntrl_entry->domain, cntrl_entry->bus,
-+ pci_domain_nr(dev->bus), dev->bus->number,
-+ PCI_SLOT(dev->devfn), PCI_FUNC(dev->devfn));
-+ ret = -ENOSPC;
-+ goto out;
-+ }
++ info->waiting_resp = 0;
++ smp_mb();
+
-+ dev_entry = kmalloc(sizeof(*dev_entry), GFP_ATOMIC);
-+ if (!dev_entry) {
-+ if (list_empty(&cntrl_entry->dev_list)) {
-+ list_del(&cntrl_entry->list);
-+ kfree(cntrl_entry);
-+ }
-+ ret = -ENOMEM;
-+ goto out;
++ if (scsifront_cmd_done(info))
++ info->waiting_resp = 1;
+ }
+
-+ dev_entry->dev = dev;
-+ dev_entry->devfn = cntrl_entry->next_devfn;
-+
-+ list_add_tail(&dev_entry->list, &cntrl_entry->dev_list);
++ return 0;
++}
+
-+ cntrl_entry->next_devfn += PCI_DEVFN(1, 0);
+
-+out:
-+ spin_unlock_irqrestore(&dev_data->lock, flags);
-+ return ret;
-+}
+
-+void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
++static int map_data_for_request(struct vscsifrnt_info *info,
++ struct scsi_cmnd *sc, vscsiif_request_t *ring_req, uint32_t id)
+{
-+ struct controller_dev_data *dev_data = pdev->pci_dev_data;
-+ struct controller_list_entry *cntrl_entry;
-+ struct controller_dev_entry *dev_entry = NULL;
-+ struct pci_dev *found_dev = NULL;
-+ unsigned long flags;
++ grant_ref_t gref_head;
++ struct page *page;
++ int err, i, ref, ref_cnt = 0;
++ int write = (sc->sc_data_direction == DMA_TO_DEVICE);
++ int nr_pages, off, len, bytes;
++ unsigned long buffer_pfn;
++ unsigned int data_len = 0;
+
-+ spin_lock_irqsave(&dev_data->lock, flags);
++ if (sc->sc_data_direction == DMA_NONE)
++ return 0;
+
-+ list_for_each_entry(cntrl_entry, &dev_data->list, list) {
-+ if (cntrl_entry->controller != PCI_CONTROLLER(dev))
-+ continue;
++ err = gnttab_alloc_grant_references(VSCSIIF_SG_TABLESIZE, &gref_head);
++ if (err) {
++ printk(KERN_ERR "scsifront: gnttab_alloc_grant_references() error\n");
++ return -ENOMEM;
++ }
+
-+ list_for_each_entry(dev_entry, &cntrl_entry->dev_list, list) {
-+ if (dev_entry->dev == dev) {
-+ found_dev = dev_entry->dev;
-+ break;
++ if (sc->use_sg) {
++ /* quoted scsi_lib.c/scsi_req_map_sg . */
++ struct scatterlist *sg = (struct scatterlist *)sc->request_buffer;
++ nr_pages = (sc->request_bufflen + sg[0].offset + PAGE_SIZE - 1) >> PAGE_SHIFT;
++
++ if (nr_pages > VSCSIIF_SG_TABLESIZE) {
++ printk(KERN_ERR "scsifront: Unable to map request_buffer for command!\n");
++ ref_cnt = (-E2BIG);
++ goto big_to_sg;
++ }
++
++ for (i = 0; i < sc->use_sg; i++) {
++ page = sg[i].page;
++ off = sg[i].offset;
++ len = sg[i].length;
++ data_len += len;
++
++ buffer_pfn = page_to_phys(page) >> PAGE_SHIFT;
++
++ while (len > 0) {
++ bytes = min_t(unsigned int, len, PAGE_SIZE - off);
++
++ ref = gnttab_claim_grant_reference(&gref_head);
++ BUG_ON(ref == -ENOSPC);
++
++ gnttab_grant_foreign_access_ref(ref, info->dev->otherend_id,
++ buffer_pfn, write);
++
++ info->shadow[id].gref[ref_cnt] = ref;
++ ring_req->seg[ref_cnt].gref = ref;
++ ring_req->seg[ref_cnt].offset = (uint16_t)off;
++ ring_req->seg[ref_cnt].length = (uint16_t)bytes;
++
++ buffer_pfn++;
++ len -= bytes;
++ off = 0;
++ ref_cnt++;
+ }
+ }
-+ }
++ } else if (sc->request_bufflen) {
++ unsigned long end = ((unsigned long)sc->request_buffer
++ + sc->request_bufflen + PAGE_SIZE - 1) >> PAGE_SHIFT;
++ unsigned long start = (unsigned long)sc->request_buffer >> PAGE_SHIFT;
+
-+ if (!found_dev) {
-+ spin_unlock_irqrestore(&dev_data->lock, flags);
-+ return;
-+ }
++ page = virt_to_page(sc->request_buffer);
++ nr_pages = end - start;
++ len = sc->request_bufflen;
+
-+ list_del(&dev_entry->list);
-+ kfree(dev_entry);
++ if (nr_pages > VSCSIIF_SG_TABLESIZE) {
++ ref_cnt = (-E2BIG);
++ goto big_to_sg;
++ }
+
-+ if (list_empty(&cntrl_entry->dev_list)) {
-+ list_del(&cntrl_entry->list);
-+ kfree(cntrl_entry);
-+ }
++ buffer_pfn = page_to_phys(page) >> PAGE_SHIFT;
+
-+ spin_unlock_irqrestore(&dev_data->lock, flags);
-+ pcistub_put_pci_dev(found_dev);
-+}
++ off = offset_in_page((unsigned long)sc->request_buffer);
++ for (i = 0; i < nr_pages; i++) {
++ bytes = PAGE_SIZE - off;
+
-+int pciback_init_devices(struct pciback_device *pdev)
-+{
-+ struct controller_dev_data *dev_data;
++ if (bytes > len)
++ bytes = len;
+
-+ dev_data = kmalloc(sizeof(*dev_data), GFP_KERNEL);
-+ if (!dev_data)
-+ return -ENOMEM;
++ ref = gnttab_claim_grant_reference(&gref_head);
++ BUG_ON(ref == -ENOSPC);
+
-+ spin_lock_init(&dev_data->lock);
++ gnttab_grant_foreign_access_ref(ref, info->dev->otherend_id,
++ buffer_pfn, write);
+
-+ INIT_LIST_HEAD(&dev_data->list);
++ info->shadow[id].gref[i] = ref;
++ ring_req->seg[i].gref = ref;
++ ring_req->seg[i].offset = (uint16_t)off;
++ ring_req->seg[i].length = (uint16_t)bytes;
+
-+ /* Starting domain:bus numbers */
-+ dev_data->next_domain = 0;
-+ dev_data->next_bus = 0;
++ buffer_pfn++;
++ len -= bytes;
++ off = 0;
++ ref_cnt++;
++ }
++ }
+
-+ pdev->pci_dev_data = dev_data;
++big_to_sg:
+
-+ return 0;
++ gnttab_free_grant_references(gref_head);
++
++ return ref_cnt;
+}
+
-+static acpi_status write_xenbus_resource(struct acpi_resource *res, void *data)
++static int scsifront_queuecommand(struct scsi_cmnd *sc,
++ void (*done)(struct scsi_cmnd *))
+{
-+ struct walk_info *info = data;
-+ struct acpi_resource_address64 addr;
-+ acpi_status status;
-+ int i, len, err;
-+ char str[32], tmp[3];
-+ unsigned char *ptr, *buf;
++ struct vscsifrnt_info *info =
++ (struct vscsifrnt_info *) sc->device->host->hostdata;
++ vscsiif_request_t *ring_req;
++ int ref_cnt;
++ uint16_t rqid;
+
-+ status = acpi_resource_to_address64(res, &addr);
++ if (RING_FULL(&info->ring)) {
++ goto out_host_busy;
++ }
+
-+ /* Do we care about this range? Let's check. */
-+ if (!ACPI_SUCCESS(status) ||
-+ !(addr.resource_type == ACPI_MEMORY_RANGE ||
-+ addr.resource_type == ACPI_IO_RANGE) ||
-+ !addr.address_length || addr.producer_consumer != ACPI_PRODUCER)
-+ return AE_OK;
++ sc->scsi_done = done;
++ sc->result = 0;
+
-+ /*
-+ * Furthermore, we really only care to tell the guest about
-+ * address ranges that require address translation of some sort.
-+ */
-+ if (!(addr.resource_type == ACPI_MEMORY_RANGE &&
-+ addr.info.mem.translation) &&
-+ !(addr.resource_type == ACPI_IO_RANGE &&
-+ addr.info.io.translation))
-+ return AE_OK;
-+
-+ /* Store the resource in xenbus for the guest */
-+ len = snprintf(str, sizeof(str), "root-%d-resource-%d",
-+ info->root_num, info->resource_count);
-+ if (unlikely(len >= (sizeof(str) - 1)))
-+ return AE_OK;
++ ring_req = scsifront_pre_request(info);
++ rqid = ring_req->rqid;
++ ring_req->act = VSCSIIF_ACT_SCSI_CDB;
+
-+ buf = kzalloc((sizeof(*res) * 2) + 1, GFP_KERNEL);
-+ if (!buf)
-+ return AE_OK;
++ ring_req->id = sc->device->id;
++ ring_req->lun = sc->device->lun;
++ ring_req->channel = sc->device->channel;
++ ring_req->cmd_len = sc->cmd_len;
+
-+ /* Clean out resource_source */
-+ res->data.address64.resource_source.index = 0xFF;
-+ res->data.address64.resource_source.string_length = 0;
-+ res->data.address64.resource_source.string_ptr = NULL;
++ BUG_ON(sc->cmd_len > VSCSIIF_MAX_COMMAND_SIZE);
+
-+ ptr = (unsigned char *)res;
++ if ( sc->cmd_len )
++ memcpy(ring_req->cmnd, sc->cmnd, sc->cmd_len);
++ else
++ memset(ring_req->cmnd, 0, VSCSIIF_MAX_COMMAND_SIZE);
+
-+ /* Turn the acpi_resource into an ASCII byte stream */
-+ for (i = 0; i < sizeof(*res); i++) {
-+ snprintf(tmp, sizeof(tmp), "%02x", ptr[i]);
-+ strncat(buf, tmp, 2);
++ ring_req->sc_data_direction = (uint8_t)sc->sc_data_direction;
++ ring_req->timeout_per_command = (sc->timeout_per_command / HZ);
++
++ info->shadow[rqid].req_scsi_cmnd = (unsigned long)sc;
++ info->shadow[rqid].sc_data_direction = sc->sc_data_direction;
++ info->shadow[rqid].act = ring_req->act;
++
++ ref_cnt = map_data_for_request(info, sc, ring_req, rqid);
++ if (ref_cnt < 0) {
++ add_id_to_freelist(info, rqid);
++ if (ref_cnt == (-ENOMEM))
++ goto out_host_busy;
++ else {
++ sc->result = (DID_ERROR << 16);
++ goto out_fail_command;
++ }
+ }
+
-+ err = xenbus_printf(XBT_NIL, info->pdev->xdev->nodename,
-+ str, "%s", buf);
++ ring_req->nr_segments = (uint8_t)ref_cnt;
++ info->shadow[rqid].nr_segments = ref_cnt;
+
-+ if (!err)
-+ info->resource_count++;
++ scsifront_do_request(info);
+
-+ kfree(buf);
++ return 0;
+
-+ return AE_OK;
++out_host_busy:
++ return SCSI_MLQUEUE_HOST_BUSY;
++
++out_fail_command:
++ done(sc);
++ return 0;
+}
+
-+int pciback_publish_pci_roots(struct pciback_device *pdev,
-+ publish_pci_root_cb publish_root_cb)
++
++static int scsifront_eh_abort_handler(struct scsi_cmnd *sc)
+{
-+ struct controller_dev_data *dev_data = pdev->pci_dev_data;
-+ struct controller_list_entry *cntrl_entry;
-+ int i, root_num, len, err = 0;
-+ unsigned int domain, bus;
-+ char str[64];
-+ struct walk_info info;
++ return (FAILED);
++}
+
-+ spin_lock(&dev_data->lock);
++/* vscsi supports only device_reset, because it is each of LUNs */
++static int scsifront_dev_reset_handler(struct scsi_cmnd *sc)
++{
++ struct Scsi_Host *host = sc->device->host;
++ struct vscsifrnt_info *info =
++ (struct vscsifrnt_info *) sc->device->host->hostdata;
+
-+ list_for_each_entry(cntrl_entry, &dev_data->list, list) {
-+ /* First publish all the domain:bus info */
-+ err = publish_root_cb(pdev, cntrl_entry->domain,
-+ cntrl_entry->bus);
-+ if (err)
-+ goto out;
++ vscsiif_request_t *ring_req;
++ uint16_t rqid;
++ int err;
+
-+ /*
-+ * Now figure out which root-%d this belongs to
-+ * so we can associate resources with it.
-+ */
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
-+ "root_num", "%d", &root_num);
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,12)
++ spin_lock_irq(host->host_lock);
++#endif
+
-+ if (err != 1)
-+ goto out;
++ ring_req = scsifront_pre_request(info);
++ ring_req->act = VSCSIIF_ACT_SCSI_RESET;
+
-+ for (i = 0; i < root_num; i++) {
-+ len = snprintf(str, sizeof(str), "root-%d", i);
-+ if (unlikely(len >= (sizeof(str) - 1))) {
-+ err = -ENOMEM;
-+ goto out;
-+ }
++ rqid = ring_req->rqid;
++ info->shadow[rqid].act = VSCSIIF_ACT_SCSI_RESET;
+
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
-+ str, "%x:%x", &domain, &bus);
-+ if (err != 2)
-+ goto out;
++ ring_req->channel = sc->device->channel;
++ ring_req->id = sc->device->id;
++ ring_req->lun = sc->device->lun;
++ ring_req->cmd_len = sc->cmd_len;
+
-+ /* Is this the one we just published? */
-+ if (domain == cntrl_entry->domain &&
-+ bus == cntrl_entry->bus)
-+ break;
-+ }
++ if ( sc->cmd_len )
++ memcpy(ring_req->cmnd, sc->cmnd, sc->cmd_len);
++ else
++ memset(ring_req->cmnd, 0, VSCSIIF_MAX_COMMAND_SIZE);
+
-+ if (i == root_num)
-+ goto out;
++ ring_req->sc_data_direction = (uint8_t)sc->sc_data_direction;
++ ring_req->timeout_per_command = (sc->timeout_per_command / HZ);
++ ring_req->nr_segments = 0;
+
-+ info.pdev = pdev;
-+ info.resource_count = 0;
-+ info.root_num = i;
++ scsifront_do_request(info);
+
-+ /* Let ACPI do the heavy lifting on decoding resources */
-+ acpi_walk_resources(cntrl_entry->controller->acpi_handle,
-+ METHOD_NAME__CRS, write_xenbus_resource,
-+ &info);
++ spin_unlock_irq(host->host_lock);
++ wait_event_interruptible(info->shadow[rqid].wq_reset,
++ info->shadow[rqid].wait_reset);
++ spin_lock_irq(host->host_lock);
+
-+ /* No resouces. OK. On to the next one */
-+ if (!info.resource_count)
-+ continue;
++ err = info->shadow[rqid].rslt_reset;
+
-+ /* Store the number of resources we wrote for this root-%d */
-+ len = snprintf(str, sizeof(str), "root-%d-resources", i);
-+ if (unlikely(len >= (sizeof(str) - 1))) {
-+ err = -ENOMEM;
-+ goto out;
-+ }
++ add_id_to_freelist(info, rqid);
+
-+ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, str,
-+ "%d", info.resource_count);
-+ if (err)
-+ goto out;
-+ }
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,12)
++ spin_unlock_irq(host->host_lock);
++#endif
++ return (err);
++}
+
-+ /* Finally, write some magic to synchronize with the guest. */
-+ len = snprintf(str, sizeof(str), "root-resource-magic");
-+ if (unlikely(len >= (sizeof(str) - 1))) {
-+ err = -ENOMEM;
-+ goto out;
-+ }
+
-+ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, str,
-+ "%lx", (sizeof(struct acpi_resource) * 2) + 1);
++struct scsi_host_template scsifront_sht = {
++ .module = THIS_MODULE,
++ .name = "Xen SCSI frontend driver",
++ .queuecommand = scsifront_queuecommand,
++ .eh_abort_handler = scsifront_eh_abort_handler,
++ .eh_device_reset_handler= scsifront_dev_reset_handler,
++ .cmd_per_lun = VSCSIIF_DEFAULT_CMD_PER_LUN,
++ .can_queue = VSCSIIF_MAX_REQS,
++ .this_id = -1,
++ .sg_tablesize = VSCSIIF_SG_TABLESIZE,
++ .use_clustering = DISABLE_CLUSTERING,
++ .proc_name = "scsifront",
++};
+
-+out:
-+ spin_unlock(&dev_data->lock);
++
++static int __init scsifront_init(void)
++{
++ int err;
++
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ err = scsifront_xenbus_init();
+
+ return err;
+}
+
-+void pciback_release_devices(struct pciback_device *pdev)
++static void __exit scsifront_exit(void)
+{
-+ struct controller_dev_data *dev_data = pdev->pci_dev_data;
-+ struct controller_list_entry *cntrl_entry, *c;
-+ struct controller_dev_entry *dev_entry, *d;
-+
-+ list_for_each_entry_safe(cntrl_entry, c, &dev_data->list, list) {
-+ list_for_each_entry_safe(dev_entry, d,
-+ &cntrl_entry->dev_list, list) {
-+ list_del(&dev_entry->list);
-+ pcistub_put_pci_dev(dev_entry->dev);
-+ kfree(dev_entry);
-+ }
-+ list_del(&cntrl_entry->list);
-+ kfree(cntrl_entry);
-+ }
-+
-+ kfree(dev_data);
-+ pdev->pci_dev_data = NULL;
++ scsifront_xenbus_unregister();
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/Makefile tmp-linux-2.6-xen.patch/drivers/xen/pciback/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/pciback/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/Makefile 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,16 @@
-+obj-$(CONFIG_XEN_PCIDEV_BACKEND) += pciback.o
+
-+pciback-y := pci_stub.o pciback_ops.o xenbus.o
-+pciback-y += conf_space.o conf_space_header.o \
-+ conf_space_capability.o \
-+ conf_space_capability_vpd.o \
-+ conf_space_capability_pm.o \
-+ conf_space_quirks.o
-+pciback-$(CONFIG_XEN_PCIDEV_BACKEND_VPCI) += vpci.o
-+pciback-$(CONFIG_XEN_PCIDEV_BACKEND_SLOT) += slot.o
-+pciback-$(CONFIG_XEN_PCIDEV_BACKEND_PASS) += passthrough.o
-+pciback-$(CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER) += controller.o
++module_init(scsifront_init);
++module_exit(scsifront_exit);
+
-+ifeq ($(CONFIG_XEN_PCIDEV_BE_DEBUG),y)
-+EXTRA_CFLAGS += -DDEBUG
-+endif
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/passthrough.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/passthrough.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/passthrough.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/passthrough.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,157 @@
++MODULE_DESCRIPTION("Xen SCSI frontend driver");
++MODULE_LICENSE("GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/scsifront/xenbus.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/scsifront/xenbus.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,421 @@
+/*
-+ * PCI Backend - Provides restricted access to the real PCI bus topology
-+ * to the frontend
++ * Xen SCSI frontend driver
+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
++
+
-+#include <linux/list.h>
-+#include <linux/pci.h>
-+#include <linux/spinlock.h>
-+#include "pciback.h"
++#include <linux/version.h>
++#include "common.h"
+
-+struct passthrough_dev_data {
-+ /* Access to dev_list must be protected by lock */
-+ struct list_head dev_list;
-+ spinlock_t lock;
-+};
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,11)
++ #define DEFAULT_TASK_COMM_LEN 16
++#else
++ #define DEFAULT_TASK_COMM_LEN TASK_COMM_LEN
++#endif
+
-+struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
-+ unsigned int domain, unsigned int bus,
-+ unsigned int devfn)
++extern struct scsi_host_template scsifront_sht;
++
++static void scsifront_free(struct vscsifrnt_info *info)
+{
-+ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
-+ struct pci_dev_entry *dev_entry;
-+ struct pci_dev *dev = NULL;
-+ unsigned long flags;
++ struct Scsi_Host *host = info->host;
+
-+ spin_lock_irqsave(&dev_data->lock, flags);
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,14)
++ if (host->shost_state != SHOST_DEL) {
++#else
++ if (!test_bit(SHOST_DEL, &host->shost_state)) {
++#endif
++ scsi_remove_host(info->host);
++ }
+
-+ list_for_each_entry(dev_entry, &dev_data->dev_list, list) {
-+ if (domain == (unsigned int)pci_domain_nr(dev_entry->dev->bus)
-+ && bus == (unsigned int)dev_entry->dev->bus->number
-+ && devfn == dev_entry->dev->devfn) {
-+ dev = dev_entry->dev;
-+ break;
-+ }
++ if (info->ring_ref != GRANT_INVALID_REF) {
++ gnttab_end_foreign_access(info->ring_ref,
++ (unsigned long)info->ring.sring);
++ info->ring_ref = GRANT_INVALID_REF;
++ info->ring.sring = NULL;
+ }
+
-+ spin_unlock_irqrestore(&dev_data->lock, flags);
++ if (info->irq)
++ unbind_from_irqhandler(info->irq, info);
++ info->irq = 0;
+
-+ return dev;
++ scsi_host_put(info->host);
+}
+
-+int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
++
++static int scsifront_alloc_ring(struct vscsifrnt_info *info)
+{
-+ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
-+ struct pci_dev_entry *dev_entry;
-+ unsigned long flags;
++ struct xenbus_device *dev = info->dev;
++ struct vscsiif_sring *sring;
++ int err = -ENOMEM;
+
-+ dev_entry = kmalloc(sizeof(*dev_entry), GFP_KERNEL);
-+ if (!dev_entry)
-+ return -ENOMEM;
-+ dev_entry->dev = dev;
+
-+ spin_lock_irqsave(&dev_data->lock, flags);
-+ list_add_tail(&dev_entry->list, &dev_data->dev_list);
-+ spin_unlock_irqrestore(&dev_data->lock, flags);
++ info->ring_ref = GRANT_INVALID_REF;
+
-+ return 0;
-+}
++ /***** Frontend to Backend ring start *****/
++ sring = (struct vscsiif_sring *) __get_free_page(GFP_KERNEL);
++ if (!sring) {
++ xenbus_dev_fatal(dev, err, "fail to allocate shared ring (Front to Back)");
++ return err;
++ }
++ SHARED_RING_INIT(sring);
++ FRONT_RING_INIT(&info->ring, sring, PAGE_SIZE);
+
-+void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
-+{
-+ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
-+ struct pci_dev_entry *dev_entry, *t;
-+ struct pci_dev *found_dev = NULL;
-+ unsigned long flags;
++ err = xenbus_grant_ring(dev, virt_to_mfn(sring));
++ if (err < 0) {
++ free_page((unsigned long) sring);
++ info->ring.sring = NULL;
++ xenbus_dev_fatal(dev, err, "fail to grant shared ring (Front to Back)");
++ goto free_sring;
++ }
++ info->ring_ref = err;
+
-+ spin_lock_irqsave(&dev_data->lock, flags);
++ err = bind_listening_port_to_irqhandler(
++ dev->otherend_id, scsifront_intr,
++ SA_SAMPLE_RANDOM, "scsifront", info);
+
-+ list_for_each_entry_safe(dev_entry, t, &dev_data->dev_list, list) {
-+ if (dev_entry->dev == dev) {
-+ list_del(&dev_entry->list);
-+ found_dev = dev_entry->dev;
-+ kfree(dev_entry);
-+ }
++ if (err <= 0) {
++ xenbus_dev_fatal(dev, err, "bind_listening_port_to_irqhandler");
++ goto free_sring;
+ }
++ info->irq = err;
+
-+ spin_unlock_irqrestore(&dev_data->lock, flags);
++ return 0;
+
-+ if (found_dev)
-+ pcistub_put_pci_dev(found_dev);
++/* free resource */
++free_sring:
++ scsifront_free(info);
++
++ return err;
+}
+
-+int pciback_init_devices(struct pciback_device *pdev)
++
++static int scsifront_init_ring(struct vscsifrnt_info *info)
+{
-+ struct passthrough_dev_data *dev_data;
++ struct xenbus_device *dev = info->dev;
++ struct xenbus_transaction xbt;
++ int err;
+
-+ dev_data = kmalloc(sizeof(*dev_data), GFP_KERNEL);
-+ if (!dev_data)
-+ return -ENOMEM;
++ DPRINTK("%s\n",__FUNCTION__);
+
-+ spin_lock_init(&dev_data->lock);
++ err = scsifront_alloc_ring(info);
++ if (err)
++ return err;
++ DPRINTK("%u %u\n", info->ring_ref, info->evtchn);
+
-+ INIT_LIST_HEAD(&dev_data->dev_list);
++again:
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "starting transaction");
++ }
+
-+ pdev->pci_dev_data = dev_data;
++ err = xenbus_printf(xbt, dev->nodename, "ring-ref", "%u",
++ info->ring_ref);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "%s", "writing ring-ref");
++ goto fail;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
++ irq_to_evtchn_port(info->irq));
++
++ if (err) {
++ xenbus_dev_fatal(dev, err, "%s", "writing event-channel");
++ goto fail;
++ }
++
++ err = xenbus_transaction_end(xbt, 0);
++ if (err) {
++ if (err == -EAGAIN)
++ goto again;
++ xenbus_dev_fatal(dev, err, "completing transaction");
++ goto free_sring;
++ }
+
+ return 0;
++
++fail:
++ xenbus_transaction_end(xbt, 1);
++free_sring:
++ /* free resource */
++ scsifront_free(info);
++
++ return err;
+}
+
-+int pciback_publish_pci_roots(struct pciback_device *pdev,
-+ publish_pci_root_cb publish_root_cb)
++
++static int scsifront_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
+{
-+ int err = 0;
-+ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
-+ struct pci_dev_entry *dev_entry, *e;
-+ struct pci_dev *dev;
-+ int found;
-+ unsigned int domain, bus;
++ struct vscsifrnt_info *info;
++ struct Scsi_Host *host;
++ int i, err = -ENOMEM;
++ char name[DEFAULT_TASK_COMM_LEN];
++
++ host = scsi_host_alloc(&scsifront_sht, sizeof(*info));
++ if (!host) {
++ xenbus_dev_fatal(dev, err, "fail to allocate scsi host");
++ return err;
++ }
++ info = (struct vscsifrnt_info *) host->hostdata;
++ info->host = host;
+
-+ spin_lock(&dev_data->lock);
+
-+ list_for_each_entry(dev_entry, &dev_data->dev_list, list) {
-+ /* Only publish this device as a root if none of its
-+ * parent bridges are exported
-+ */
-+ found = 0;
-+ dev = dev_entry->dev->bus->self;
-+ for (; !found && dev != NULL; dev = dev->bus->self) {
-+ list_for_each_entry(e, &dev_data->dev_list, list) {
-+ if (dev == e->dev) {
-+ found = 1;
-+ break;
-+ }
-+ }
-+ }
++ dev->dev.driver_data = info;
++ info->dev = dev;
+
-+ domain = (unsigned int)pci_domain_nr(dev_entry->dev->bus);
-+ bus = (unsigned int)dev_entry->dev->bus->number;
++ for (i = 0; i < VSCSIIF_MAX_REQS; i++) {
++ info->shadow[i].next_free = i + 1;
++ init_waitqueue_head(&(info->shadow[i].wq_reset));
++ info->shadow[i].wait_reset = 0;
++ }
++ info->shadow[VSCSIIF_MAX_REQS - 1].next_free = 0x0fff;
+
-+ if (!found) {
-+ err = publish_root_cb(pdev, domain, bus);
-+ if (err)
-+ break;
-+ }
++ err = scsifront_init_ring(info);
++ if (err) {
++ scsi_host_put(host);
++ return err;
+ }
+
-+ spin_unlock(&dev_data->lock);
++ init_waitqueue_head(&info->wq);
++ spin_lock_init(&info->io_lock);
++ spin_lock_init(&info->shadow_lock);
+
-+ return err;
-+}
++ snprintf(name, DEFAULT_TASK_COMM_LEN, "vscsiif.%d", info->host->host_no);
+
-+void pciback_release_devices(struct pciback_device *pdev)
-+{
-+ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
-+ struct pci_dev_entry *dev_entry, *t;
++ info->kthread = kthread_run(scsifront_schedule, info, name);
++ if (IS_ERR(info->kthread)) {
++ err = PTR_ERR(info->kthread);
++ info->kthread = NULL;
++ printk(KERN_ERR "scsifront: kthread start err %d\n", err);
++ goto free_sring;
++ }
+
-+ list_for_each_entry_safe(dev_entry, t, &dev_data->dev_list, list) {
-+ list_del(&dev_entry->list);
-+ pcistub_put_pci_dev(dev_entry->dev);
-+ kfree(dev_entry);
++ host->max_id = VSCSIIF_MAX_TARGET;
++ host->max_channel = 0;
++ host->max_lun = VSCSIIF_MAX_LUN;
++ host->max_sectors = (VSCSIIF_SG_TABLESIZE - 1) * PAGE_SIZE / 512;
++
++ err = scsi_add_host(host, &dev->dev);
++ if (err) {
++ printk(KERN_ERR "scsifront: fail to add scsi host %d\n", err);
++ goto free_sring;
+ }
+
-+ kfree(dev_data);
-+ pdev->pci_dev_data = NULL;
++ xenbus_switch_state(dev, XenbusStateInitialised);
++
++ return 0;
++
++free_sring:
++ /* free resource */
++ scsifront_free(info);
++ return err;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/pciback.h tmp-linux-2.6-xen.patch/drivers/xen/pciback/pciback.h
---- pristine-linux-2.6.18.2/drivers/xen/pciback/pciback.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/pciback.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,93 @@
-+/*
-+ * PCI Backend Common Data Structures & Function Declarations
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ */
-+#ifndef __XEN_PCIBACK_H__
-+#define __XEN_PCIBACK_H__
+
-+#include <linux/pci.h>
-+#include <linux/interrupt.h>
-+#include <xen/xenbus.h>
-+#include <linux/list.h>
-+#include <linux/spinlock.h>
-+#include <linux/workqueue.h>
-+#include <asm/atomic.h>
-+#include <xen/interface/io/pciif.h>
++static int scsifront_remove(struct xenbus_device *dev)
++{
++ struct vscsifrnt_info *info = dev->dev.driver_data;
+
-+struct pci_dev_entry {
-+ struct list_head list;
-+ struct pci_dev *dev;
-+};
++ DPRINTK("%s: %s removed\n",__FUNCTION__ ,dev->nodename);
+
-+#define _PDEVF_op_active (0)
-+#define PDEVF_op_active (1<<(_PDEVF_op_active))
++ if (info->kthread) {
++ kthread_stop(info->kthread);
++ info->kthread = NULL;
++ }
+
-+struct pciback_device {
-+ void *pci_dev_data;
-+ spinlock_t dev_lock;
++ scsifront_free(info);
++
++ return 0;
++}
+
-+ struct xenbus_device *xdev;
+
-+ struct xenbus_watch be_watch;
-+ u8 be_watching;
++static int scsifront_disconnect(struct vscsifrnt_info *info)
++{
++ struct xenbus_device *dev = info->dev;
++ struct Scsi_Host *host = info->host;
+
-+ int evtchn_irq;
++ DPRINTK("%s: %s disconnect\n",__FUNCTION__ ,dev->nodename);
+
-+ struct vm_struct *sh_area;
-+ struct xen_pci_sharedinfo *sh_info;
++ /*
++ When this function is executed, all devices of
++ Frontend have been deleted.
++ Therefore, it need not block I/O before remove_host.
++ */
+
-+ unsigned long flags;
++ scsi_remove_host(host);
++ xenbus_frontend_closed(dev);
+
-+ struct work_struct op_work;
-+};
++ return 0;
++}
+
-+struct pciback_dev_data {
-+ struct list_head config_fields;
-+ int permissive;
-+ int warned_on_write;
-+};
++#define VSCSIFRONT_OP_ADD_LUN 1
++#define VSCSIFRONT_OP_DEL_LUN 2
+
-+/* Get/Put PCI Devices that are hidden from the PCI Backend Domain */
-+struct pci_dev *pcistub_get_pci_dev_by_slot(struct pciback_device *pdev,
-+ int domain, int bus,
-+ int slot, int func);
-+struct pci_dev *pcistub_get_pci_dev(struct pciback_device *pdev,
-+ struct pci_dev *dev);
-+void pcistub_put_pci_dev(struct pci_dev *dev);
++static void scsifront_do_lun_hotplug(struct vscsifrnt_info *info, int op)
++{
++ struct xenbus_device *dev = info->dev;
++ int i, err = 0;
++ char str[64], state_str[64];
++ char **dir;
++ unsigned int dir_n = 0;
++ unsigned int device_state;
++ unsigned int hst, chn, tgt, lun;
++ struct scsi_device *sdev;
+
-+/* Ensure a device is turned off or reset */
-+void pciback_reset_device(struct pci_dev *pdev);
++ dir = xenbus_directory(XBT_NIL, dev->otherend, "vscsi-devs", &dir_n);
++ if (IS_ERR(dir))
++ return;
+
-+/* Access a virtual configuration space for a PCI device */
-+int pciback_config_init(void);
-+int pciback_config_init_dev(struct pci_dev *dev);
-+void pciback_config_free_dyn_fields(struct pci_dev *dev);
-+void pciback_config_reset_dev(struct pci_dev *dev);
-+void pciback_config_free_dev(struct pci_dev *dev);
-+int pciback_config_read(struct pci_dev *dev, int offset, int size,
-+ u32 * ret_val);
-+int pciback_config_write(struct pci_dev *dev, int offset, int size, u32 value);
++ for (i = 0; i < dir_n; i++) {
++ /* read status */
++ snprintf(str, sizeof(str), "vscsi-devs/%s/state", dir[i]);
++ err = xenbus_scanf(XBT_NIL, dev->otherend, str, "%u",
++ &device_state);
++ if (XENBUS_EXIST_ERR(err))
++ continue;
++
++ /* virtual SCSI device */
++ snprintf(str, sizeof(str), "vscsi-devs/%s/v-dev", dir[i]);
++ err = xenbus_scanf(XBT_NIL, dev->otherend, str,
++ "%u:%u:%u:%u", &hst, &chn, &tgt, &lun);
++ if (XENBUS_EXIST_ERR(err))
++ continue;
+
-+/* Handle requests for specific devices from the frontend */
-+typedef int (*publish_pci_root_cb) (struct pciback_device * pdev,
-+ unsigned int domain, unsigned int bus);
-+int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev);
-+void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev);
-+struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
-+ unsigned int domain, unsigned int bus,
-+ unsigned int devfn);
-+int pciback_init_devices(struct pciback_device *pdev);
-+int pciback_publish_pci_roots(struct pciback_device *pdev,
-+ publish_pci_root_cb cb);
-+void pciback_release_devices(struct pciback_device *pdev);
++ /* front device state path */
++ snprintf(state_str, sizeof(state_str), "vscsi-devs/%s/state", dir[i]);
+
-+/* Handles events from front-end */
-+irqreturn_t pciback_handle_event(int irq, void *dev_id, struct pt_regs *regs);
-+void pciback_do_op(void *data);
++ switch (op) {
++ case VSCSIFRONT_OP_ADD_LUN:
++ if (device_state == XenbusStateInitialised) {
++ sdev = scsi_device_lookup(info->host, chn, tgt, lun);
++ if (sdev) {
++ printk(KERN_ERR "scsifront: Device already in use.\n");
++ scsi_device_put(sdev);
++ xenbus_printf(XBT_NIL, dev->nodename,
++ state_str, "%d", XenbusStateClosed);
++ } else {
++ scsi_add_device(info->host, chn, tgt, lun);
++ xenbus_printf(XBT_NIL, dev->nodename,
++ state_str, "%d", XenbusStateConnected);
++ }
++ }
++ break;
++ case VSCSIFRONT_OP_DEL_LUN:
++ if (device_state == XenbusStateClosing) {
++ sdev = scsi_device_lookup(info->host, chn, tgt, lun);
++ if (sdev) {
++ scsi_remove_device(sdev);
++ scsi_device_put(sdev);
++ xenbus_printf(XBT_NIL, dev->nodename,
++ state_str, "%d", XenbusStateClosed);
++ }
++ }
++ break;
++ default:
++ break;
++ }
++ }
++
++ kfree(dir);
++ return;
++}
+
-+int pciback_xenbus_register(void);
-+void pciback_xenbus_unregister(void);
+
-+extern int verbose_request;
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/pciback_ops.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/pciback_ops.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/pciback_ops.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/pciback_ops.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,95 @@
-+/*
-+ * PCI Backend Operations - respond to PCI requests from Frontend
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ */
-+#include <linux/module.h>
-+#include <asm/bitops.h>
-+#include <xen/evtchn.h>
-+#include "pciback.h"
+
-+int verbose_request = 0;
-+module_param(verbose_request, int, 0644);
+
-+/* Ensure a device is "turned off" and ready to be exported.
-+ * (Also see pciback_config_reset to ensure virtual configuration space is
-+ * ready to be re-exported)
-+ */
-+void pciback_reset_device(struct pci_dev *dev)
++static void scsifront_backend_changed(struct xenbus_device *dev,
++ enum xenbus_state backend_state)
+{
-+ u16 cmd;
++ struct vscsifrnt_info *info = dev->dev.driver_data;
+
-+ /* Disable devices (but not bridges) */
-+ if (dev->hdr_type == PCI_HEADER_TYPE_NORMAL) {
-+ pci_disable_device(dev);
++ DPRINTK("%p %u %u\n", dev, dev->state, backend_state);
+
-+ pci_write_config_word(dev, PCI_COMMAND, 0);
++ switch (backend_state) {
++ case XenbusStateUnknown:
++ case XenbusStateInitialising:
++ case XenbusStateInitWait:
++ case XenbusStateClosed:
++ break;
+
-+ dev->is_enabled = 0;
-+ dev->is_busmaster = 0;
-+ } else {
-+ pci_read_config_word(dev, PCI_COMMAND, &cmd);
-+ if (cmd & (PCI_COMMAND_INVALIDATE)) {
-+ cmd &= ~(PCI_COMMAND_INVALIDATE);
-+ pci_write_config_word(dev, PCI_COMMAND, cmd);
++ case XenbusStateInitialised:
++ break;
+
-+ dev->is_busmaster = 0;
++ case XenbusStateConnected:
++ if (xenbus_read_driver_state(dev->nodename) ==
++ XenbusStateInitialised) {
++ scsifront_do_lun_hotplug(info, VSCSIFRONT_OP_ADD_LUN);
+ }
++
++ if (dev->state == XenbusStateConnected)
++ break;
++
++ xenbus_switch_state(dev, XenbusStateConnected);
++ break;
++
++ case XenbusStateClosing:
++ scsifront_disconnect(info);
++ break;
++
++ case XenbusStateReconfiguring:
++ scsifront_do_lun_hotplug(info, VSCSIFRONT_OP_DEL_LUN);
++ xenbus_switch_state(dev, XenbusStateReconfiguring);
++ break;
++
++ case XenbusStateReconfigured:
++ scsifront_do_lun_hotplug(info, VSCSIFRONT_OP_ADD_LUN);
++ xenbus_switch_state(dev, XenbusStateConnected);
++ break;
+ }
+}
+
-+static inline void test_and_schedule_op(struct pciback_device *pdev)
++
++static struct xenbus_device_id scsifront_ids[] = {
++ { "vscsi" },
++ { "" }
++};
++
++
++static struct xenbus_driver scsifront_driver = {
++ .name = "vscsi",
++ .owner = THIS_MODULE,
++ .ids = scsifront_ids,
++ .probe = scsifront_probe,
++ .remove = scsifront_remove,
++/* .resume = scsifront_resume, */
++ .otherend_changed = scsifront_backend_changed,
++};
++
++int scsifront_xenbus_init(void)
+{
-+ /* Check that frontend is requesting an operation and that we are not
-+ * already processing a request */
-+ if (test_bit(_XEN_PCIF_active, (unsigned long *)&pdev->sh_info->flags)
-+ && !test_and_set_bit(_PDEVF_op_active, &pdev->flags))
-+ schedule_work(&pdev->op_work);
++ return xenbus_register_frontend(&scsifront_driver);
+}
+
-+/* Performing the configuration space reads/writes must not be done in atomic
-+ * context because some of the pci_* functions can sleep (mostly due to ACPI
-+ * use of semaphores). This function is intended to be called from a work
-+ * queue in process context taking a struct pciback_device as a parameter */
-+void pciback_do_op(void *data)
++void scsifront_xenbus_unregister(void)
+{
-+ struct pciback_device *pdev = data;
-+ struct pci_dev *dev;
-+ struct xen_pci_op *op = &pdev->sh_info->op;
++ xenbus_unregister_driver(&scsifront_driver);
++}
+
-+ dev = pciback_get_pci_dev(pdev, op->domain, op->bus, op->devfn);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/tpmback/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/tpmback/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,4 @@
+
-+ if (dev == NULL)
-+ op->err = XEN_PCI_ERR_dev_not_found;
-+ else if (op->cmd == XEN_PCI_OP_conf_read)
-+ op->err = pciback_config_read(dev, op->offset, op->size,
-+ &op->value);
-+ else if (op->cmd == XEN_PCI_OP_conf_write)
-+ op->err = pciback_config_write(dev, op->offset, op->size,
-+ op->value);
-+ else
-+ op->err = XEN_PCI_ERR_not_implemented;
++obj-$(CONFIG_XEN_TPMDEV_BACKEND) += tpmbk.o
+
-+ /* Tell the driver domain that we're done. */
-+ wmb();
-+ clear_bit(_XEN_PCIF_active, (unsigned long *)&pdev->sh_info->flags);
-+ notify_remote_via_irq(pdev->evtchn_irq);
++tpmbk-y += tpmback.o interface.o xenbus.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/tpmback/common.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/tpmback/common.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,85 @@
++/******************************************************************************
++ * drivers/xen/tpmback/common.h
++ */
+
-+ /* Mark that we're done. */
-+ smp_mb__before_clear_bit(); /* /after/ clearing PCIF_active */
-+ clear_bit(_PDEVF_op_active, &pdev->flags);
-+ smp_mb__after_clear_bit(); /* /before/ final check for work */
++#ifndef __TPM__BACKEND__COMMON_H__
++#define __TPM__BACKEND__COMMON_H__
+
-+ /* Check to see if the driver domain tried to start another request in
-+ * between clearing _XEN_PCIF_active and clearing _PDEVF_op_active. */
-+ test_and_schedule_op(pdev);
-+}
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/interrupt.h>
++#include <linux/slab.h>
++#include <xen/evtchn.h>
++#include <xen/driver_util.h>
++#include <xen/interface/grant_table.h>
++#include <xen/interface/io/tpmif.h>
++#include <asm/io.h>
++#include <asm/pgalloc.h>
+
-+irqreturn_t pciback_handle_event(int irq, void *dev_id, struct pt_regs *regs)
-+{
-+ struct pciback_device *pdev = dev_id;
++#define DPRINTK(_f, _a...) \
++ pr_debug("(file=%s, line=%d) " _f, \
++ __FILE__ , __LINE__ , ## _a )
+
-+ test_and_schedule_op(pdev);
++struct backend_info;
+
-+ return IRQ_HANDLED;
-+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/pci_stub.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/pci_stub.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/pci_stub.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/pci_stub.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,929 @@
-+/*
-+ * PCI Stub Driver - Grabs devices in backend to be exported later
-+ *
-+ * Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ * Chris Bookholt <hap10 at epoch.ncsc.mil>
-+ */
-+#include <linux/module.h>
-+#include <linux/init.h>
-+#include <linux/list.h>
-+#include <linux/spinlock.h>
-+#include <linux/kref.h>
-+#include <asm/atomic.h>
-+#include "pciback.h"
-+#include "conf_space.h"
-+#include "conf_space_quirks.h"
++typedef struct tpmif_st {
++ struct list_head tpmif_list;
++ /* Unique identifier for this interface. */
++ domid_t domid;
++ unsigned int handle;
+
-+static char *pci_devs_to_hide = NULL;
-+module_param_named(hide, pci_devs_to_hide, charp, 0444);
++ /* Physical parameters of the comms window. */
++ unsigned int irq;
+
-+struct pcistub_device_id {
-+ struct list_head slot_list;
-+ int domain;
-+ unsigned char bus;
-+ unsigned int devfn;
-+};
-+static LIST_HEAD(pcistub_device_ids);
-+static DEFINE_SPINLOCK(device_ids_lock);
++ /* The shared rings and indexes. */
++ tpmif_tx_interface_t *tx;
++ struct vm_struct *tx_area;
+
-+struct pcistub_device {
-+ struct kref kref;
-+ struct list_head dev_list;
-+ spinlock_t lock;
++ /* Miscellaneous private stuff. */
++ enum { DISCONNECTED, DISCONNECTING, CONNECTED } status;
++ int active;
+
-+ struct pci_dev *dev;
-+ struct pciback_device *pdev; /* non-NULL if struct pci_dev is in use */
-+};
++ struct tpmif_st *hash_next;
++ struct list_head list; /* scheduling list */
++ atomic_t refcnt;
+
-+/* Access to pcistub_devices & seized_devices lists and the initialize_devices
-+ * flag must be locked with pcistub_devices_lock
-+ */
-+static DEFINE_SPINLOCK(pcistub_devices_lock);
-+static LIST_HEAD(pcistub_devices);
++ struct backend_info *bi;
+
-+/* wait for device_initcall before initializing our devices
-+ * (see pcistub_init_devices_late)
-+ */
-+static int initialize_devices = 0;
-+static LIST_HEAD(seized_devices);
++ grant_handle_t shmem_handle;
++ grant_ref_t shmem_ref;
++ struct page **mmap_pages;
+
-+static struct pcistub_device *pcistub_device_alloc(struct pci_dev *dev)
-+{
-+ struct pcistub_device *psdev;
++ char devname[20];
++} tpmif_t;
+
-+ dev_dbg(&dev->dev, "pcistub_device_alloc\n");
++void tpmif_disconnect_complete(tpmif_t * tpmif);
++tpmif_t *tpmif_find(domid_t domid, struct backend_info *bi);
++void tpmif_interface_init(void);
++void tpmif_interface_exit(void);
++void tpmif_schedule_work(tpmif_t * tpmif);
++void tpmif_deschedule_work(tpmif_t * tpmif);
++void tpmif_xenbus_init(void);
++void tpmif_xenbus_exit(void);
++int tpmif_map(tpmif_t *tpmif, unsigned long shared_page, unsigned int evtchn);
++irqreturn_t tpmif_be_int(int irq, void *dev_id, struct pt_regs *regs);
+
-+ psdev = kzalloc(sizeof(*psdev), GFP_ATOMIC);
-+ if (!psdev)
-+ return NULL;
++long int tpmback_get_instance(struct backend_info *bi);
+
-+ psdev->dev = pci_dev_get(dev);
-+ if (!psdev->dev) {
-+ kfree(psdev);
-+ return NULL;
-+ }
++int vtpm_release_packets(tpmif_t * tpmif, int send_msgs);
++
++
++#define tpmif_get(_b) (atomic_inc(&(_b)->refcnt))
++#define tpmif_put(_b) \
++ do { \
++ if (atomic_dec_and_test(&(_b)->refcnt)) \
++ tpmif_disconnect_complete(_b); \
++ } while (0)
++
++extern int num_frontends;
++
++static inline unsigned long idx_to_kaddr(tpmif_t *t, unsigned int idx)
++{
++ return (unsigned long)pfn_to_kaddr(page_to_pfn(t->mmap_pages[idx]));
++}
++
++#endif /* __TPMIF__BACKEND__COMMON_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/tpmback/interface.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/tpmback/interface.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,168 @@
++ /*****************************************************************************
++ * drivers/xen/tpmback/interface.c
++ *
++ * Vritual TPM interface management.
++ *
++ * Copyright (c) 2005, IBM Corporation
++ *
++ * Author: Stefan Berger, stefanb at us.ibm.com
++ *
++ * This code has been derived from drivers/xen/netback/interface.c
++ * Copyright (c) 2004, Keir Fraser
++ */
+
-+ kref_init(&psdev->kref);
-+ spin_lock_init(&psdev->lock);
++#include "common.h"
++#include <xen/balloon.h>
++#include <xen/gnttab.h>
+
-+ return psdev;
-+}
++static kmem_cache_t *tpmif_cachep;
++int num_frontends = 0;
+
-+/* Don't call this directly as it's called by pcistub_device_put */
-+static void pcistub_device_release(struct kref *kref)
++LIST_HEAD(tpmif_list);
++
++static tpmif_t *alloc_tpmif(domid_t domid, struct backend_info *bi)
+{
-+ struct pcistub_device *psdev;
++ tpmif_t *tpmif;
+
-+ psdev = container_of(kref, struct pcistub_device, kref);
++ tpmif = kmem_cache_alloc(tpmif_cachep, GFP_KERNEL);
++ if (tpmif == NULL)
++ goto out_of_memory;
+
-+ dev_dbg(&psdev->dev->dev, "pcistub_device_release\n");
++ memset(tpmif, 0, sizeof (*tpmif));
++ tpmif->domid = domid;
++ tpmif->status = DISCONNECTED;
++ tpmif->bi = bi;
++ snprintf(tpmif->devname, sizeof(tpmif->devname), "tpmif%d", domid);
++ atomic_set(&tpmif->refcnt, 1);
+
-+ /* Clean-up the device */
-+ pciback_reset_device(psdev->dev);
-+ pciback_config_free_dyn_fields(psdev->dev);
-+ pciback_config_free_dev(psdev->dev);
-+ kfree(pci_get_drvdata(psdev->dev));
-+ pci_set_drvdata(psdev->dev, NULL);
++ tpmif->mmap_pages = alloc_empty_pages_and_pagevec(TPMIF_TX_RING_SIZE);
++ if (tpmif->mmap_pages == NULL)
++ goto out_of_memory;
+
-+ pci_dev_put(psdev->dev);
++ list_add(&tpmif->tpmif_list, &tpmif_list);
++ num_frontends++;
+
-+ kfree(psdev);
-+}
++ return tpmif;
+
-+static inline void pcistub_device_get(struct pcistub_device *psdev)
-+{
-+ kref_get(&psdev->kref);
++ out_of_memory:
++ if (tpmif != NULL)
++ kmem_cache_free(tpmif_cachep, tpmif);
++ printk("%s: out of memory\n", __FUNCTION__);
++ return ERR_PTR(-ENOMEM);
+}
+
-+static inline void pcistub_device_put(struct pcistub_device *psdev)
++static void free_tpmif(tpmif_t * tpmif)
+{
-+ kref_put(&psdev->kref, pcistub_device_release);
++ num_frontends--;
++ list_del(&tpmif->tpmif_list);
++ free_empty_pages_and_pagevec(tpmif->mmap_pages, TPMIF_TX_RING_SIZE);
++ kmem_cache_free(tpmif_cachep, tpmif);
+}
+
-+static struct pcistub_device *pcistub_device_find(int domain, int bus,
-+ int slot, int func)
++tpmif_t *tpmif_find(domid_t domid, struct backend_info *bi)
+{
-+ struct pcistub_device *psdev = NULL;
-+ unsigned long flags;
-+
-+ spin_lock_irqsave(&pcistub_devices_lock, flags);
++ tpmif_t *tpmif;
+
-+ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
-+ if (psdev->dev != NULL
-+ && domain == pci_domain_nr(psdev->dev->bus)
-+ && bus == psdev->dev->bus->number
-+ && PCI_DEVFN(slot, func) == psdev->dev->devfn) {
-+ pcistub_device_get(psdev);
-+ goto out;
++ list_for_each_entry(tpmif, &tpmif_list, tpmif_list) {
++ if (tpmif->bi == bi) {
++ if (tpmif->domid == domid) {
++ tpmif_get(tpmif);
++ return tpmif;
++ } else {
++ return ERR_PTR(-EEXIST);
++ }
+ }
+ }
+
-+ /* didn't find it */
-+ psdev = NULL;
-+
-+ out:
-+ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
-+ return psdev;
++ return alloc_tpmif(domid, bi);
+}
+
-+static struct pci_dev *pcistub_device_get_pci_dev(struct pciback_device *pdev,
-+ struct pcistub_device *psdev)
++static int map_frontend_page(tpmif_t *tpmif, unsigned long shared_page)
+{
-+ struct pci_dev *pci_dev = NULL;
-+ unsigned long flags;
++ struct gnttab_map_grant_ref op;
+
-+ pcistub_device_get(psdev);
++ gnttab_set_map_op(&op, (unsigned long)tpmif->tx_area->addr,
++ GNTMAP_host_map, shared_page, tpmif->domid);
+
-+ spin_lock_irqsave(&psdev->lock, flags);
-+ if (!psdev->pdev) {
-+ psdev->pdev = pdev;
-+ pci_dev = psdev->dev;
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
++
++ if (op.status) {
++ DPRINTK(" Grant table operation failure !\n");
++ return op.status;
+ }
-+ spin_unlock_irqrestore(&psdev->lock, flags);
+
-+ if (!pci_dev)
-+ pcistub_device_put(psdev);
++ tpmif->shmem_ref = shared_page;
++ tpmif->shmem_handle = op.handle;
+
-+ return pci_dev;
++ return 0;
+}
+
-+struct pci_dev *pcistub_get_pci_dev_by_slot(struct pciback_device *pdev,
-+ int domain, int bus,
-+ int slot, int func)
++static void unmap_frontend_page(tpmif_t *tpmif)
+{
-+ struct pcistub_device *psdev;
-+ struct pci_dev *found_dev = NULL;
-+ unsigned long flags;
-+
-+ spin_lock_irqsave(&pcistub_devices_lock, flags);
++ struct gnttab_unmap_grant_ref op;
+
-+ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
-+ if (psdev->dev != NULL
-+ && domain == pci_domain_nr(psdev->dev->bus)
-+ && bus == psdev->dev->bus->number
-+ && PCI_DEVFN(slot, func) == psdev->dev->devfn) {
-+ found_dev = pcistub_device_get_pci_dev(pdev, psdev);
-+ break;
-+ }
-+ }
++ gnttab_set_unmap_op(&op, (unsigned long)tpmif->tx_area->addr,
++ GNTMAP_host_map, tpmif->shmem_handle);
+
-+ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
-+ return found_dev;
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
+}
+
-+struct pci_dev *pcistub_get_pci_dev(struct pciback_device *pdev,
-+ struct pci_dev *dev)
++int tpmif_map(tpmif_t *tpmif, unsigned long shared_page, unsigned int evtchn)
+{
-+ struct pcistub_device *psdev;
-+ struct pci_dev *found_dev = NULL;
-+ unsigned long flags;
-+
-+ spin_lock_irqsave(&pcistub_devices_lock, flags);
-+
-+ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
-+ if (psdev->dev == dev) {
-+ found_dev = pcistub_device_get_pci_dev(pdev, psdev);
-+ break;
-+ }
-+ }
-+
-+ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
-+ return found_dev;
-+}
++ int err;
+
-+void pcistub_put_pci_dev(struct pci_dev *dev)
-+{
-+ struct pcistub_device *psdev, *found_psdev = NULL;
-+ unsigned long flags;
++ if (tpmif->irq)
++ return 0;
+
-+ spin_lock_irqsave(&pcistub_devices_lock, flags);
++ if ((tpmif->tx_area = alloc_vm_area(PAGE_SIZE)) == NULL)
++ return -ENOMEM;
+
-+ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
-+ if (psdev->dev == dev) {
-+ found_psdev = psdev;
-+ break;
-+ }
++ err = map_frontend_page(tpmif, shared_page);
++ if (err) {
++ free_vm_area(tpmif->tx_area);
++ return err;
+ }
+
-+ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++ tpmif->tx = (tpmif_tx_interface_t *)tpmif->tx_area->addr;
++ memset(tpmif->tx, 0, PAGE_SIZE);
+
-+ /* Cleanup our device
-+ * (so it's ready for the next domain)
-+ */
-+ pciback_reset_device(found_psdev->dev);
-+ pciback_config_free_dyn_fields(found_psdev->dev);
-+ pciback_config_reset_dev(found_psdev->dev);
++ err = bind_interdomain_evtchn_to_irqhandler(
++ tpmif->domid, evtchn, tpmif_be_int, 0, tpmif->devname, tpmif);
++ if (err < 0) {
++ unmap_frontend_page(tpmif);
++ free_vm_area(tpmif->tx_area);
++ return err;
++ }
++ tpmif->irq = err;
+
-+ spin_lock_irqsave(&found_psdev->lock, flags);
-+ found_psdev->pdev = NULL;
-+ spin_unlock_irqrestore(&found_psdev->lock, flags);
++ tpmif->shmem_ref = shared_page;
++ tpmif->active = 1;
+
-+ pcistub_device_put(found_psdev);
++ return 0;
+}
+
-+static int __devinit pcistub_match_one(struct pci_dev *dev,
-+ struct pcistub_device_id *pdev_id)
++void tpmif_disconnect_complete(tpmif_t *tpmif)
+{
-+ /* Match the specified device by domain, bus, slot, func and also if
-+ * any of the device's parent bridges match.
-+ */
-+ for (; dev != NULL; dev = dev->bus->self) {
-+ if (pci_domain_nr(dev->bus) == pdev_id->domain
-+ && dev->bus->number == pdev_id->bus
-+ && dev->devfn == pdev_id->devfn)
-+ return 1;
++ if (tpmif->irq)
++ unbind_from_irqhandler(tpmif->irq, tpmif);
+
-+ /* Sometimes topmost bridge links to itself. */
-+ if (dev == dev->bus->self)
-+ break;
++ if (tpmif->tx) {
++ unmap_frontend_page(tpmif);
++ free_vm_area(tpmif->tx_area);
+ }
+
-+ return 0;
++ free_tpmif(tpmif);
+}
+
-+static int __devinit pcistub_match(struct pci_dev *dev)
++void __init tpmif_interface_init(void)
+{
-+ struct pcistub_device_id *pdev_id;
-+ unsigned long flags;
-+ int found = 0;
-+
-+ spin_lock_irqsave(&device_ids_lock, flags);
-+ list_for_each_entry(pdev_id, &pcistub_device_ids, slot_list) {
-+ if (pcistub_match_one(dev, pdev_id)) {
-+ found = 1;
-+ break;
-+ }
-+ }
-+ spin_unlock_irqrestore(&device_ids_lock, flags);
-+
-+ return found;
++ tpmif_cachep = kmem_cache_create("tpmif_cache", sizeof (tpmif_t),
++ 0, 0, NULL, NULL);
+}
+
-+static int __devinit pcistub_init_device(struct pci_dev *dev)
++void __exit tpmif_interface_exit(void)
+{
-+ struct pciback_dev_data *dev_data;
-+ int err = 0;
++ kmem_cache_destroy(tpmif_cachep);
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/tpmback/tpmback.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/tpmback/tpmback.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,944 @@
++/******************************************************************************
++ * drivers/xen/tpmback/tpmback.c
++ *
++ * Copyright (c) 2005, IBM Corporation
++ *
++ * Author: Stefan Berger, stefanb at us.ibm.com
++ * Grant table support: Mahadevan Gomathisankaran
++ *
++ * This code has been derived from drivers/xen/netback/netback.c
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ */
+
-+ dev_dbg(&dev->dev, "initializing...\n");
++#include "common.h"
++#include <xen/evtchn.h>
+
-+ /* The PCI backend is not intended to be a module (or to work with
-+ * removable PCI devices (yet). If it were, pciback_config_free()
-+ * would need to be called somewhere to free the memory allocated
-+ * here and then to call kfree(pci_get_drvdata(psdev->dev)).
-+ */
-+ dev_data = kzalloc(sizeof(*dev_data), GFP_ATOMIC);
-+ if (!dev_data) {
-+ err = -ENOMEM;
-+ goto out;
-+ }
-+ pci_set_drvdata(dev, dev_data);
++#include <linux/types.h>
++#include <linux/list.h>
++#include <linux/miscdevice.h>
++#include <linux/poll.h>
++#include <asm/uaccess.h>
++#include <xen/xenbus.h>
++#include <xen/interface/grant_table.h>
++#include <xen/gnttab.h>
+
-+ dev_dbg(&dev->dev, "initializing config\n");
-+ err = pciback_config_init_dev(dev);
-+ if (err)
-+ goto out;
++/* local data structures */
++struct data_exchange {
++ struct list_head pending_pak;
++ struct list_head current_pak;
++ unsigned int copied_so_far;
++ u8 has_opener:1;
++ u8 aborted:1;
++ rwlock_t pak_lock; // protects all of the previous fields
++ wait_queue_head_t wait_queue;
++};
+
-+ /* HACK: Force device (& ACPI) to determine what IRQ it's on - we
-+ * must do this here because pcibios_enable_device may specify
-+ * the pci device's true irq (and possibly its other resources)
-+ * if they differ from what's in the configuration space.
-+ * This makes the assumption that the device's resources won't
-+ * change after this point (otherwise this code may break!)
-+ */
-+ dev_dbg(&dev->dev, "enabling device\n");
-+ err = pci_enable_device(dev);
-+ if (err)
-+ goto config_release;
++struct vtpm_resp_hdr {
++ uint32_t instance_no;
++ uint16_t tag_no;
++ uint32_t len_no;
++ uint32_t ordinal_no;
++} __attribute__ ((packed));
+
-+ /* Now disable the device (this also ensures some private device
-+ * data is setup before we export)
-+ */
-+ dev_dbg(&dev->dev, "reset device\n");
-+ pciback_reset_device(dev);
++struct packet {
++ struct list_head next;
++ unsigned int data_len;
++ u8 *data_buffer;
++ tpmif_t *tpmif;
++ u32 tpm_instance;
++ u8 req_tag;
++ u32 last_read;
++ u8 flags;
++ struct timer_list processing_timer;
++};
+
-+ return 0;
++enum {
++ PACKET_FLAG_DISCARD_RESPONSE = 1,
++};
+
-+ config_release:
-+ pciback_config_free_dev(dev);
++/* local variables */
++static struct data_exchange dataex;
+
-+ out:
-+ pci_set_drvdata(dev, NULL);
-+ kfree(dev_data);
-+ return err;
-+}
++/* local function prototypes */
++static int _packet_write(struct packet *pak,
++ const char *data, size_t size, int userbuffer);
++static void processing_timeout(unsigned long ptr);
++static int packet_read_shmem(struct packet *pak,
++ tpmif_t * tpmif,
++ u32 offset,
++ char *buffer, int isuserbuffer, u32 left);
++static int vtpm_queue_packet(struct packet *pak);
+
-+/*
-+ * Because some initialization still happens on
-+ * devices during fs_initcall, we need to defer
-+ * full initialization of our devices until
-+ * device_initcall.
-+ */
-+static int __init pcistub_init_devices_late(void)
++/***************************************************************
++ Buffer copying fo user and kernel space buffes.
++***************************************************************/
++static inline int copy_from_buffer(void *to,
++ const void *from, unsigned long size,
++ int isuserbuffer)
+{
-+ struct pcistub_device *psdev;
-+ unsigned long flags;
-+ int err = 0;
++ if (isuserbuffer) {
++ if (copy_from_user(to, (void __user *)from, size))
++ return -EFAULT;
++ } else {
++ memcpy(to, from, size);
++ }
++ return 0;
++}
+
-+ pr_debug("pciback: pcistub_init_devices_late\n");
++static inline int copy_to_buffer(void *to,
++ const void *from, unsigned long size,
++ int isuserbuffer)
++{
++ if (isuserbuffer) {
++ if (copy_to_user((void __user *)to, from, size))
++ return -EFAULT;
++ } else {
++ memcpy(to, from, size);
++ }
++ return 0;
++}
+
-+ spin_lock_irqsave(&pcistub_devices_lock, flags);
+
-+ while (!list_empty(&seized_devices)) {
-+ psdev = container_of(seized_devices.next,
-+ struct pcistub_device, dev_list);
-+ list_del(&psdev->dev_list);
++static void dataex_init(struct data_exchange *dataex)
++{
++ INIT_LIST_HEAD(&dataex->pending_pak);
++ INIT_LIST_HEAD(&dataex->current_pak);
++ dataex->has_opener = 0;
++ rwlock_init(&dataex->pak_lock);
++ init_waitqueue_head(&dataex->wait_queue);
++}
+
-+ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++/***************************************************************
++ Packet-related functions
++***************************************************************/
+
-+ err = pcistub_init_device(psdev->dev);
-+ if (err) {
-+ dev_err(&psdev->dev->dev,
-+ "error %d initializing device\n", err);
-+ kfree(psdev);
-+ psdev = NULL;
-+ }
++static struct packet *packet_find_instance(struct list_head *head,
++ u32 tpm_instance)
++{
++ struct packet *pak;
++ struct list_head *p;
+
-+ spin_lock_irqsave(&pcistub_devices_lock, flags);
++ /*
++ * traverse the list of packets and return the first
++ * one with the given instance number
++ */
++ list_for_each(p, head) {
++ pak = list_entry(p, struct packet, next);
+
-+ if (psdev)
-+ list_add_tail(&psdev->dev_list, &pcistub_devices);
++ if (pak->tpm_instance == tpm_instance) {
++ return pak;
++ }
+ }
++ return NULL;
++}
+
-+ initialize_devices = 1;
++static struct packet *packet_find_packet(struct list_head *head, void *packet)
++{
++ struct packet *pak;
++ struct list_head *p;
+
-+ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++ /*
++ * traverse the list of packets and return the first
++ * one with the given instance number
++ */
++ list_for_each(p, head) {
++ pak = list_entry(p, struct packet, next);
+
-+ return 0;
++ if (pak == packet) {
++ return pak;
++ }
++ }
++ return NULL;
+}
+
-+static int __devinit pcistub_seize(struct pci_dev *dev)
++static struct packet *packet_alloc(tpmif_t * tpmif,
++ u32 size, u8 req_tag, u8 flags)
+{
-+ struct pcistub_device *psdev;
-+ unsigned long flags;
-+ int err = 0;
++ struct packet *pak = NULL;
++ pak = kzalloc(sizeof (struct packet), GFP_ATOMIC);
++ if (NULL != pak) {
++ if (tpmif) {
++ pak->tpmif = tpmif;
++ pak->tpm_instance = tpmback_get_instance(tpmif->bi);
++ tpmif_get(tpmif);
++ }
++ pak->data_len = size;
++ pak->req_tag = req_tag;
++ pak->last_read = 0;
++ pak->flags = flags;
+
-+ psdev = pcistub_device_alloc(dev);
-+ if (!psdev)
-+ return -ENOMEM;
++ /*
++ * cannot do tpmif_get(tpmif); bad things happen
++ * on the last tpmif_put()
++ */
++ init_timer(&pak->processing_timer);
++ pak->processing_timer.function = processing_timeout;
++ pak->processing_timer.data = (unsigned long)pak;
++ }
++ return pak;
++}
+
-+ spin_lock_irqsave(&pcistub_devices_lock, flags);
++static void inline packet_reset(struct packet *pak)
++{
++ pak->last_read = 0;
++}
+
-+ if (initialize_devices) {
-+ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++static void packet_free(struct packet *pak)
++{
++ if (timer_pending(&pak->processing_timer)) {
++ BUG();
++ }
+
-+ /* don't want irqs disabled when calling pcistub_init_device */
-+ err = pcistub_init_device(psdev->dev);
++ if (pak->tpmif)
++ tpmif_put(pak->tpmif);
++ kfree(pak->data_buffer);
++ /*
++ * cannot do tpmif_put(pak->tpmif); bad things happen
++ * on the last tpmif_put()
++ */
++ kfree(pak);
++}
+
-+ spin_lock_irqsave(&pcistub_devices_lock, flags);
+
-+ if (!err)
-+ list_add(&psdev->dev_list, &pcistub_devices);
++/*
++ * Write data to the shared memory and send it to the FE.
++ */
++static int packet_write(struct packet *pak,
++ const char *data, size_t size, int isuserbuffer)
++{
++ int rc = 0;
++
++ if (0 != (pak->flags & PACKET_FLAG_DISCARD_RESPONSE)) {
++ /* Don't send a respone to this packet. Just acknowledge it. */
++ rc = size;
+ } else {
-+ dev_dbg(&dev->dev, "deferring initialization\n");
-+ list_add(&psdev->dev_list, &seized_devices);
++ rc = _packet_write(pak, data, size, isuserbuffer);
+ }
+
-+ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
-+
-+ if (err)
-+ pcistub_device_put(psdev);
-+
-+ return err;
++ return rc;
+}
+
-+static int __devinit pcistub_probe(struct pci_dev *dev,
-+ const struct pci_device_id *id)
++int _packet_write(struct packet *pak,
++ const char *data, size_t size, int isuserbuffer)
+{
-+ int err = 0;
-+
-+ dev_dbg(&dev->dev, "probing...\n");
++ /*
++ * Write into the shared memory pages directly
++ * and send it to the front end.
++ */
++ tpmif_t *tpmif = pak->tpmif;
++ grant_handle_t handle;
++ int rc = 0;
++ unsigned int i = 0;
++ unsigned int offset = 0;
+
-+ if (pcistub_match(dev)) {
++ if (tpmif == NULL) {
++ return -EFAULT;
++ }
+
-+ if (dev->hdr_type != PCI_HEADER_TYPE_NORMAL
-+ && dev->hdr_type != PCI_HEADER_TYPE_BRIDGE) {
-+ dev_err(&dev->dev, "can't export pci devices that "
-+ "don't have a normal (0) or bridge (1) "
-+ "header type!\n");
-+ err = -ENODEV;
-+ goto out;
-+ }
++ if (tpmif->status == DISCONNECTED) {
++ return size;
++ }
+
-+ dev_info(&dev->dev, "seizing device\n");
-+ err = pcistub_seize(dev);
-+ } else
-+ /* Didn't find the device */
-+ err = -ENODEV;
++ while (offset < size && i < TPMIF_TX_RING_SIZE) {
++ unsigned int tocopy;
++ struct gnttab_map_grant_ref map_op;
++ struct gnttab_unmap_grant_ref unmap_op;
++ tpmif_tx_request_t *tx;
+
-+ out:
-+ return err;
-+}
++ tx = &tpmif->tx->ring[i].req;
+
-+static void pcistub_remove(struct pci_dev *dev)
-+{
-+ struct pcistub_device *psdev, *found_psdev = NULL;
-+ unsigned long flags;
++ if (0 == tx->addr) {
++ DPRINTK("ERROR: Buffer for outgoing packet NULL?! i=%d\n", i);
++ return 0;
++ }
+
-+ dev_dbg(&dev->dev, "removing\n");
++ gnttab_set_map_op(&map_op, idx_to_kaddr(tpmif, i),
++ GNTMAP_host_map, tx->ref, tpmif->domid);
+
-+ spin_lock_irqsave(&pcistub_devices_lock, flags);
++ if (unlikely(HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref,
++ &map_op, 1))) {
++ BUG();
++ }
+
-+ pciback_config_quirk_release(dev);
++ handle = map_op.handle;
+
-+ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
-+ if (psdev->dev == dev) {
-+ found_psdev = psdev;
-+ break;
++ if (map_op.status) {
++ DPRINTK(" Grant table operation failure !\n");
++ return 0;
+ }
-+ }
+
-+ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++ tocopy = min_t(size_t, size - offset, PAGE_SIZE);
+
-+ if (found_psdev) {
-+ dev_dbg(&dev->dev, "found device to remove - in use? %p\n",
-+ found_psdev->pdev);
++ if (copy_from_buffer((void *)(idx_to_kaddr(tpmif, i) |
++ (tx->addr & ~PAGE_MASK)),
++ &data[offset], tocopy, isuserbuffer)) {
++ tpmif_put(tpmif);
++ return -EFAULT;
++ }
++ tx->size = tocopy;
+
-+ if (found_psdev->pdev) {
-+ printk(KERN_WARNING "pciback: ****** removing device "
-+ "%s while still in-use! ******\n",
-+ pci_name(found_psdev->dev));
-+ printk(KERN_WARNING "pciback: ****** driver domain may "
-+ "still access this device's i/o resources!\n");
-+ printk(KERN_WARNING "pciback: ****** shutdown driver "
-+ "domain before binding device\n");
-+ printk(KERN_WARNING "pciback: ****** to other drivers "
-+ "or domains\n");
++ gnttab_set_unmap_op(&unmap_op, idx_to_kaddr(tpmif, i),
++ GNTMAP_host_map, handle);
+
-+ pciback_release_pci_dev(found_psdev->pdev,
-+ found_psdev->dev);
++ if (unlikely
++ (HYPERVISOR_grant_table_op
++ (GNTTABOP_unmap_grant_ref, &unmap_op, 1))) {
++ BUG();
+ }
+
-+ spin_lock_irqsave(&pcistub_devices_lock, flags);
-+ list_del(&found_psdev->dev_list);
-+ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
-+
-+ /* the final put for releasing from the list */
-+ pcistub_device_put(found_psdev);
++ offset += tocopy;
++ i++;
+ }
-+}
+
-+static struct pci_device_id pcistub_ids[] = {
-+ {
-+ .vendor = PCI_ANY_ID,
-+ .device = PCI_ANY_ID,
-+ .subvendor = PCI_ANY_ID,
-+ .subdevice = PCI_ANY_ID,
-+ },
-+ {0,},
-+};
++ rc = offset;
++ DPRINTK("Notifying frontend via irq %d\n", tpmif->irq);
++ notify_remote_via_irq(tpmif->irq);
++
++ return rc;
++}
+
+/*
-+ * Note: There is no MODULE_DEVICE_TABLE entry here because this isn't
-+ * for a normal device. I don't want it to be loaded automatically.
++ * Read data from the shared memory and copy it directly into the
++ * provided buffer. Advance the read_last indicator which tells
++ * how many bytes have already been read.
+ */
++static int packet_read(struct packet *pak, size_t numbytes,
++ char *buffer, size_t buffersize, int isuserbuffer)
++{
++ tpmif_t *tpmif = pak->tpmif;
+
-+static struct pci_driver pciback_pci_driver = {
-+ .name = "pciback",
-+ .id_table = pcistub_ids,
-+ .probe = pcistub_probe,
-+ .remove = pcistub_remove,
-+};
++ /*
++ * Read 'numbytes' of data from the buffer. The first 4
++ * bytes are the instance number in network byte order,
++ * after that come the data from the shared memory buffer.
++ */
++ u32 to_copy;
++ u32 offset = 0;
++ u32 room_left = buffersize;
+
-+static inline int str_to_slot(const char *buf, int *domain, int *bus,
-+ int *slot, int *func)
-+{
-+ int err;
++ if (pak->last_read < 4) {
++ /*
++ * copy the instance number into the buffer
++ */
++ u32 instance_no = htonl(pak->tpm_instance);
++ u32 last_read = pak->last_read;
+
-+ err = sscanf(buf, " %x:%x:%x.%x", domain, bus, slot, func);
-+ if (err == 4)
-+ return 0;
-+ else if (err < 0)
-+ return -EINVAL;
++ to_copy = min_t(size_t, 4 - last_read, numbytes);
+
-+ /* try again without domain */
-+ *domain = 0;
-+ err = sscanf(buf, " %x:%x.%x", bus, slot, func);
-+ if (err == 3)
-+ return 0;
++ if (copy_to_buffer(&buffer[0],
++ &(((u8 *) & instance_no)[last_read]),
++ to_copy, isuserbuffer)) {
++ return -EFAULT;
++ }
+
-+ return -EINVAL;
-+}
++ pak->last_read += to_copy;
++ offset += to_copy;
++ room_left -= to_copy;
++ }
+
-+static inline int str_to_quirk(const char *buf, int *domain, int *bus, int
-+ *slot, int *func, int *reg, int *size, int *mask)
-+{
-+ int err;
++ /*
++ * If the packet has a data buffer appended, read from it...
++ */
+
-+ err =
-+ sscanf(buf, " %04x:%02x:%02x.%1x-%08x:%1x:%08x", domain, bus, slot,
-+ func, reg, size, mask);
-+ if (err == 7)
-+ return 0;
-+ return -EINVAL;
++ if (room_left > 0) {
++ if (pak->data_buffer) {
++ u32 to_copy = min_t(u32, pak->data_len - offset, room_left);
++ u32 last_read = pak->last_read - 4;
++
++ if (copy_to_buffer(&buffer[offset],
++ &pak->data_buffer[last_read],
++ to_copy, isuserbuffer)) {
++ return -EFAULT;
++ }
++ pak->last_read += to_copy;
++ offset += to_copy;
++ } else {
++ offset = packet_read_shmem(pak,
++ tpmif,
++ offset,
++ buffer,
++ isuserbuffer, room_left);
++ }
++ }
++ return offset;
+}
+
-+static int pcistub_device_id_add(int domain, int bus, int slot, int func)
++static int packet_read_shmem(struct packet *pak,
++ tpmif_t * tpmif,
++ u32 offset, char *buffer, int isuserbuffer,
++ u32 room_left)
+{
-+ struct pcistub_device_id *pci_dev_id;
-+ unsigned long flags;
++ u32 last_read = pak->last_read - 4;
++ u32 i = (last_read / PAGE_SIZE);
++ u32 pg_offset = last_read & (PAGE_SIZE - 1);
++ u32 to_copy;
++ grant_handle_t handle;
+
-+ pci_dev_id = kmalloc(sizeof(*pci_dev_id), GFP_KERNEL);
-+ if (!pci_dev_id)
-+ return -ENOMEM;
++ tpmif_tx_request_t *tx;
+
-+ pci_dev_id->domain = domain;
-+ pci_dev_id->bus = bus;
-+ pci_dev_id->devfn = PCI_DEVFN(slot, func);
++ tx = &tpmif->tx->ring[0].req;
++ /*
++ * Start copying data at the page with index 'index'
++ * and within that page at offset 'offset'.
++ * Copy a maximum of 'room_left' bytes.
++ */
++ to_copy = min_t(u32, PAGE_SIZE - pg_offset, room_left);
++ while (to_copy > 0) {
++ void *src;
++ struct gnttab_map_grant_ref map_op;
++ struct gnttab_unmap_grant_ref unmap_op;
+
-+ pr_debug("pciback: wants to seize %04x:%02x:%02x.%01x\n",
-+ domain, bus, slot, func);
++ tx = &tpmif->tx->ring[i].req;
+
-+ spin_lock_irqsave(&device_ids_lock, flags);
-+ list_add_tail(&pci_dev_id->slot_list, &pcistub_device_ids);
-+ spin_unlock_irqrestore(&device_ids_lock, flags);
++ gnttab_set_map_op(&map_op, idx_to_kaddr(tpmif, i),
++ GNTMAP_host_map, tx->ref, tpmif->domid);
+
-+ return 0;
-+}
++ if (unlikely(HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref,
++ &map_op, 1))) {
++ BUG();
++ }
+
-+static int pcistub_device_id_remove(int domain, int bus, int slot, int func)
-+{
-+ struct pcistub_device_id *pci_dev_id, *t;
-+ int devfn = PCI_DEVFN(slot, func);
-+ int err = -ENOENT;
-+ unsigned long flags;
++ if (map_op.status) {
++ DPRINTK(" Grant table operation failure !\n");
++ return -EFAULT;
++ }
+
-+ spin_lock_irqsave(&device_ids_lock, flags);
-+ list_for_each_entry_safe(pci_dev_id, t, &pcistub_device_ids, slot_list) {
++ handle = map_op.handle;
+
-+ if (pci_dev_id->domain == domain
-+ && pci_dev_id->bus == bus && pci_dev_id->devfn == devfn) {
-+ /* Don't break; here because it's possible the same
-+ * slot could be in the list more than once
++ if (to_copy > tx->size) {
++ /*
++ * User requests more than what's available
+ */
-+ list_del(&pci_dev_id->slot_list);
-+ kfree(pci_dev_id);
++ to_copy = min_t(u32, tx->size, to_copy);
++ }
+
-+ err = 0;
++ DPRINTK("Copying from mapped memory at %08lx\n",
++ (unsigned long)(idx_to_kaddr(tpmif, i) |
++ (tx->addr & ~PAGE_MASK)));
+
-+ pr_debug("pciback: removed %04x:%02x:%02x.%01x from "
-+ "seize list\n", domain, bus, slot, func);
++ src = (void *)(idx_to_kaddr(tpmif, i) |
++ ((tx->addr & ~PAGE_MASK) + pg_offset));
++ if (copy_to_buffer(&buffer[offset],
++ src, to_copy, isuserbuffer)) {
++ return -EFAULT;
+ }
-+ }
-+ spin_unlock_irqrestore(&device_ids_lock, flags);
+
-+ return err;
-+}
-+
-+static int pcistub_reg_add(int domain, int bus, int slot, int func, int reg,
-+ int size, int mask)
-+{
-+ int err = 0;
-+ struct pcistub_device *psdev;
-+ struct pci_dev *dev;
-+ struct config_field *field;
++ DPRINTK("Data from TPM-FE of domain %d are %d %d %d %d\n",
++ tpmif->domid, buffer[offset], buffer[offset + 1],
++ buffer[offset + 2], buffer[offset + 3]);
+
-+ psdev = pcistub_device_find(domain, bus, slot, func);
-+ if (!psdev || !psdev->dev) {
-+ err = -ENODEV;
-+ goto out;
-+ }
-+ dev = psdev->dev;
++ gnttab_set_unmap_op(&unmap_op, idx_to_kaddr(tpmif, i),
++ GNTMAP_host_map, handle);
+
-+ field = kzalloc(sizeof(*field), GFP_ATOMIC);
-+ if (!field) {
-+ err = -ENOMEM;
-+ goto out;
-+ }
++ if (unlikely
++ (HYPERVISOR_grant_table_op
++ (GNTTABOP_unmap_grant_ref, &unmap_op, 1))) {
++ BUG();
++ }
+
-+ field->offset = reg;
-+ field->size = size;
-+ field->mask = mask;
-+ field->init = NULL;
-+ field->reset = NULL;
-+ field->release = NULL;
-+ field->clean = pciback_config_field_free;
++ offset += to_copy;
++ pg_offset = 0;
++ last_read += to_copy;
++ room_left -= to_copy;
+
-+ err = pciback_config_quirks_add_field(dev, field);
-+ if (err)
-+ kfree(field);
-+ out:
-+ return err;
++ to_copy = min_t(u32, PAGE_SIZE, room_left);
++ i++;
++ } /* while (to_copy > 0) */
++ /*
++ * Adjust the last_read pointer
++ */
++ pak->last_read = last_read + 4;
++ return offset;
+}
+
-+static ssize_t pcistub_slot_add(struct device_driver *drv, const char *buf,
-+ size_t count)
++/* ============================================================
++ * The file layer for reading data from this device
++ * ============================================================
++ */
++static int vtpm_op_open(struct inode *inode, struct file *f)
+{
-+ int domain, bus, slot, func;
-+ int err;
-+
-+ err = str_to_slot(buf, &domain, &bus, &slot, &func);
-+ if (err)
-+ goto out;
-+
-+ err = pcistub_device_id_add(domain, bus, slot, func);
++ int rc = 0;
++ unsigned long flags;
+
-+ out:
-+ if (!err)
-+ err = count;
-+ return err;
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ if (dataex.has_opener == 0) {
++ dataex.has_opener = 1;
++ } else {
++ rc = -EPERM;
++ }
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ return rc;
+}
+
-+DRIVER_ATTR(new_slot, S_IWUSR, NULL, pcistub_slot_add);
-+
-+static ssize_t pcistub_slot_remove(struct device_driver *drv, const char *buf,
-+ size_t count)
++static ssize_t vtpm_op_read(struct file *file,
++ char __user * data, size_t size, loff_t * offset)
+{
-+ int domain, bus, slot, func;
-+ int err;
++ int ret_size = -ENODATA;
++ struct packet *pak = NULL;
++ unsigned long flags;
+
-+ err = str_to_slot(buf, &domain, &bus, &slot, &func);
-+ if (err)
-+ goto out;
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ if (dataex.aborted) {
++ dataex.aborted = 0;
++ dataex.copied_so_far = 0;
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ return -EIO;
++ }
+
-+ err = pcistub_device_id_remove(domain, bus, slot, func);
++ if (list_empty(&dataex.pending_pak)) {
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ wait_event_interruptible(dataex.wait_queue,
++ !list_empty(&dataex.pending_pak));
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ dataex.copied_so_far = 0;
++ }
+
-+ out:
-+ if (!err)
-+ err = count;
-+ return err;
-+}
++ if (!list_empty(&dataex.pending_pak)) {
++ unsigned int left;
+
-+DRIVER_ATTR(remove_slot, S_IWUSR, NULL, pcistub_slot_remove);
++ pak = list_entry(dataex.pending_pak.next, struct packet, next);
++ left = pak->data_len - dataex.copied_so_far;
++ list_del(&pak->next);
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
+
-+static ssize_t pcistub_slot_show(struct device_driver *drv, char *buf)
-+{
-+ struct pcistub_device_id *pci_dev_id;
-+ size_t count = 0;
-+ unsigned long flags;
++ DPRINTK("size given by app: %d, available: %d\n", size, left);
+
-+ spin_lock_irqsave(&device_ids_lock, flags);
-+ list_for_each_entry(pci_dev_id, &pcistub_device_ids, slot_list) {
-+ if (count >= PAGE_SIZE)
-+ break;
++ ret_size = min_t(size_t, size, left);
+
-+ count += scnprintf(buf + count, PAGE_SIZE - count,
-+ "%04x:%02x:%02x.%01x\n",
-+ pci_dev_id->domain, pci_dev_id->bus,
-+ PCI_SLOT(pci_dev_id->devfn),
-+ PCI_FUNC(pci_dev_id->devfn));
-+ }
-+ spin_unlock_irqrestore(&device_ids_lock, flags);
++ ret_size = packet_read(pak, ret_size, data, size, 1);
+
-+ return count;
-+}
++ write_lock_irqsave(&dataex.pak_lock, flags);
+
-+DRIVER_ATTR(slots, S_IRUSR, pcistub_slot_show, NULL);
++ if (ret_size < 0) {
++ del_singleshot_timer_sync(&pak->processing_timer);
++ packet_free(pak);
++ dataex.copied_so_far = 0;
++ } else {
++ DPRINTK("Copied %d bytes to user buffer\n", ret_size);
+
-+static ssize_t pcistub_quirk_add(struct device_driver *drv, const char *buf,
-+ size_t count)
-+{
-+ int domain, bus, slot, func, reg, size, mask;
-+ int err;
++ dataex.copied_so_far += ret_size;
++ if (dataex.copied_so_far >= pak->data_len + 4) {
++ DPRINTK("All data from this packet given to app.\n");
++ /* All data given to app */
+
-+ err = str_to_quirk(buf, &domain, &bus, &slot, &func, ®, &size,
-+ &mask);
-+ if (err)
-+ goto out;
++ del_singleshot_timer_sync(&pak->
++ processing_timer);
++ list_add_tail(&pak->next, &dataex.current_pak);
++ /*
++ * The more fontends that are handled at the same time,
++ * the more time we give the TPM to process the request.
++ */
++ mod_timer(&pak->processing_timer,
++ jiffies + (num_frontends * 60 * HZ));
++ dataex.copied_so_far = 0;
++ } else {
++ list_add(&pak->next, &dataex.pending_pak);
++ }
++ }
++ }
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
+
-+ err = pcistub_reg_add(domain, bus, slot, func, reg, size, mask);
++ DPRINTK("Returning result from read to app: %d\n", ret_size);
+
-+ out:
-+ if (!err)
-+ err = count;
-+ return err;
++ return ret_size;
+}
+
-+static ssize_t pcistub_quirk_show(struct device_driver *drv, char *buf)
++/*
++ * Write operation - only works after a previous read operation!
++ */
++static ssize_t vtpm_op_write(struct file *file,
++ const char __user * data, size_t size,
++ loff_t * offset)
+{
-+ int count = 0;
-+ unsigned long flags;
-+ extern struct list_head pciback_quirks;
-+ struct pciback_config_quirk *quirk;
-+ struct pciback_dev_data *dev_data;
-+ struct config_field *field;
-+ struct config_field_entry *cfg_entry;
++ struct packet *pak;
++ int rc = 0;
++ unsigned int off = 4;
++ unsigned long flags;
++ struct vtpm_resp_hdr vrh;
+
-+ spin_lock_irqsave(&device_ids_lock, flags);
-+ list_for_each_entry(quirk, &pciback_quirks, quirks_list) {
-+ if (count >= PAGE_SIZE)
-+ goto out;
++ /*
++ * Minimum required packet size is:
++ * 4 bytes for instance number
++ * 2 bytes for tag
++ * 4 bytes for paramSize
++ * 4 bytes for the ordinal
++ * sum: 14 bytes
++ */
++ if (size < sizeof (vrh))
++ return -EFAULT;
+
-+ count += scnprintf(buf + count, PAGE_SIZE - count,
-+ "%02x:%02x.%01x\n\t%04x:%04x:%04x:%04x\n",
-+ quirk->pdev->bus->number,
-+ PCI_SLOT(quirk->pdev->devfn),
-+ PCI_FUNC(quirk->pdev->devfn),
-+ quirk->devid.vendor, quirk->devid.device,
-+ quirk->devid.subvendor,
-+ quirk->devid.subdevice);
++ if (copy_from_user(&vrh, data, sizeof (vrh)))
++ return -EFAULT;
+
-+ dev_data = pci_get_drvdata(quirk->pdev);
++ /* malformed packet? */
++ if ((off + ntohl(vrh.len_no)) != size)
++ return -EFAULT;
+
-+ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
-+ field = cfg_entry->field;
-+ if (count >= PAGE_SIZE)
-+ goto out;
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ pak = packet_find_instance(&dataex.current_pak,
++ ntohl(vrh.instance_no));
+
-+ count += scnprintf(buf + count, PAGE_SIZE - count,
-+ "\t\t%08x:%01x:%08x\n",
-+ cfg_entry->base_offset + field->offset,
-+ field->size, field->mask);
-+ }
++ if (pak == NULL) {
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ DPRINTK(KERN_ALERT "No associated packet! (inst=%d)\n",
++ ntohl(vrh.instance_no));
++ return -EFAULT;
+ }
+
-+ out:
-+ spin_unlock_irqrestore(&device_ids_lock, flags);
++ del_singleshot_timer_sync(&pak->processing_timer);
++ list_del(&pak->next);
+
-+ return count;
-+}
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
+
-+DRIVER_ATTR(quirks, S_IRUSR | S_IWUSR, pcistub_quirk_show, pcistub_quirk_add);
++ /*
++ * The first 'offset' bytes must be the instance number - skip them.
++ */
++ size -= off;
+
-+static ssize_t permissive_add(struct device_driver *drv, const char *buf,
-+ size_t count)
-+{
-+ int domain, bus, slot, func;
-+ int err;
-+ struct pcistub_device *psdev;
-+ struct pciback_dev_data *dev_data;
-+ err = str_to_slot(buf, &domain, &bus, &slot, &func);
-+ if (err)
-+ goto out;
-+ psdev = pcistub_device_find(domain, bus, slot, func);
-+ if (!psdev) {
-+ err = -ENODEV;
-+ goto out;
-+ }
-+ if (!psdev->dev) {
-+ err = -ENODEV;
-+ goto release;
-+ }
-+ dev_data = pci_get_drvdata(psdev->dev);
-+ /* the driver data for a device should never be null at this point */
-+ if (!dev_data) {
-+ err = -ENXIO;
-+ goto release;
-+ }
-+ if (!dev_data->permissive) {
-+ dev_data->permissive = 1;
-+ /* Let user know that what they're doing could be unsafe */
-+ dev_warn(&psdev->dev->dev,
-+ "enabling permissive mode configuration space accesses!\n");
-+ dev_warn(&psdev->dev->dev,
-+ "permissive mode is potentially unsafe!\n");
++ rc = packet_write(pak, &data[off], size, 1);
++
++ if (rc > 0) {
++ /* I neglected the first 4 bytes */
++ rc += off;
+ }
-+ release:
-+ pcistub_device_put(psdev);
-+ out:
-+ if (!err)
-+ err = count;
-+ return err;
++ packet_free(pak);
++ return rc;
+}
+
-+static ssize_t permissive_show(struct device_driver *drv, char *buf)
++static int vtpm_op_release(struct inode *inode, struct file *file)
+{
-+ struct pcistub_device *psdev;
-+ struct pciback_dev_data *dev_data;
-+ size_t count = 0;
+ unsigned long flags;
-+ spin_lock_irqsave(&pcistub_devices_lock, flags);
-+ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
-+ if (count >= PAGE_SIZE)
-+ break;
-+ if (!psdev->dev)
-+ continue;
-+ dev_data = pci_get_drvdata(psdev->dev);
-+ if (!dev_data || !dev_data->permissive)
-+ continue;
-+ count +=
-+ scnprintf(buf + count, PAGE_SIZE - count, "%s\n",
-+ pci_name(psdev->dev));
-+ }
-+ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
-+ return count;
-+}
+
-+DRIVER_ATTR(permissive, S_IRUSR | S_IWUSR, permissive_show, permissive_add);
++ vtpm_release_packets(NULL, 1);
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ dataex.has_opener = 0;
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ return 0;
++}
+
-+static void pcistub_exit(void)
++static unsigned int vtpm_op_poll(struct file *file,
++ struct poll_table_struct *pts)
+{
-+ driver_remove_file(&pciback_pci_driver.driver, &driver_attr_new_slot);
-+ driver_remove_file(&pciback_pci_driver.driver,
-+ &driver_attr_remove_slot);
-+ driver_remove_file(&pciback_pci_driver.driver, &driver_attr_slots);
-+ driver_remove_file(&pciback_pci_driver.driver, &driver_attr_quirks);
-+ driver_remove_file(&pciback_pci_driver.driver, &driver_attr_permissive);
++ unsigned int flags = POLLOUT | POLLWRNORM;
+
-+ pci_unregister_driver(&pciback_pci_driver);
++ poll_wait(file, &dataex.wait_queue, pts);
++ if (!list_empty(&dataex.pending_pak)) {
++ flags |= POLLIN | POLLRDNORM;
++ }
++ return flags;
+}
+
-+static int __init pcistub_init(void)
-+{
-+ int pos = 0;
-+ int err = 0;
-+ int domain, bus, slot, func;
-+ int parsed;
-+
-+ if (pci_devs_to_hide && *pci_devs_to_hide) {
-+ do {
-+ parsed = 0;
++static const struct file_operations vtpm_ops = {
++ .owner = THIS_MODULE,
++ .llseek = no_llseek,
++ .open = vtpm_op_open,
++ .read = vtpm_op_read,
++ .write = vtpm_op_write,
++ .release = vtpm_op_release,
++ .poll = vtpm_op_poll,
++};
+
-+ err = sscanf(pci_devs_to_hide + pos,
-+ " (%x:%x:%x.%x) %n",
-+ &domain, &bus, &slot, &func, &parsed);
-+ if (err != 4) {
-+ domain = 0;
-+ err = sscanf(pci_devs_to_hide + pos,
-+ " (%x:%x.%x) %n",
-+ &bus, &slot, &func, &parsed);
-+ if (err != 3)
-+ goto parse_error;
-+ }
++static struct miscdevice vtpms_miscdevice = {
++ .minor = 225,
++ .name = "vtpm",
++ .fops = &vtpm_ops,
++};
+
-+ err = pcistub_device_id_add(domain, bus, slot, func);
-+ if (err)
-+ goto out;
++/***************************************************************
++ Utility functions
++***************************************************************/
+
-+ /* if parsed<=0, we've reached the end of the string */
-+ pos += parsed;
-+ } while (parsed > 0 && pci_devs_to_hide[pos]);
-+ }
++static int tpm_send_fail_message(struct packet *pak, u8 req_tag)
++{
++ int rc;
++ static const unsigned char tpm_error_message_fail[] = {
++ 0x00, 0x00,
++ 0x00, 0x00, 0x00, 0x0a,
++ 0x00, 0x00, 0x00, 0x09 /* TPM_FAIL */
++ };
++ unsigned char buffer[sizeof (tpm_error_message_fail)];
+
-+ /* If we're the first PCI Device Driver to register, we're the
-+ * first one to get offered PCI devices as they become
-+ * available (and thus we can be the first to grab them)
++ memcpy(buffer, tpm_error_message_fail,
++ sizeof (tpm_error_message_fail));
++ /*
++ * Insert the right response tag depending on the given tag
++ * All response tags are '+3' to the request tag.
+ */
-+ err = pci_register_driver(&pciback_pci_driver);
-+ if (err < 0)
-+ goto out;
-+
-+ err = driver_create_file(&pciback_pci_driver.driver,
-+ &driver_attr_new_slot);
-+ if (!err)
-+ err = driver_create_file(&pciback_pci_driver.driver,
-+ &driver_attr_remove_slot);
-+ if (!err)
-+ err = driver_create_file(&pciback_pci_driver.driver,
-+ &driver_attr_slots);
-+ if (!err)
-+ err = driver_create_file(&pciback_pci_driver.driver,
-+ &driver_attr_quirks);
-+ if (!err)
-+ err = driver_create_file(&pciback_pci_driver.driver,
-+ &driver_attr_permissive);
-+
-+ if (err)
-+ pcistub_exit();
++ buffer[1] = req_tag + 3;
+
-+ out:
-+ return err;
++ /*
++ * Write the data to shared memory and notify the front-end
++ */
++ rc = packet_write(pak, buffer, sizeof (buffer), 0);
+
-+ parse_error:
-+ printk(KERN_ERR "pciback: Error parsing pci_devs_to_hide at \"%s\"\n",
-+ pci_devs_to_hide + pos);
-+ return -EINVAL;
++ return rc;
+}
+
-+#ifndef MODULE
-+/*
-+ * fs_initcall happens before device_initcall
-+ * so pciback *should* get called first (b/c we
-+ * want to suck up any device before other drivers
-+ * get a chance by being the first pci device
-+ * driver to register)
-+ */
-+fs_initcall(pcistub_init);
-+#endif
-+
-+static int __init pciback_init(void)
++static int _vtpm_release_packets(struct list_head *head,
++ tpmif_t * tpmif, int send_msgs)
+{
-+ int err;
++ int aborted = 0;
++ int c = 0;
++ struct packet *pak;
++ struct list_head *pos, *tmp;
+
-+ err = pciback_config_init();
-+ if (err)
-+ return err;
++ list_for_each_safe(pos, tmp, head) {
++ pak = list_entry(pos, struct packet, next);
++ c += 1;
+
-+#ifdef MODULE
-+ err = pcistub_init();
-+ if (err < 0)
-+ return err;
-+#endif
++ if (tpmif == NULL || pak->tpmif == tpmif) {
++ int can_send = 0;
+
-+ pcistub_init_devices_late();
-+ err = pciback_xenbus_register();
-+ if (err)
-+ pcistub_exit();
++ del_singleshot_timer_sync(&pak->processing_timer);
++ list_del(&pak->next);
+
-+ return err;
++ if (pak->tpmif && pak->tpmif->status == CONNECTED) {
++ can_send = 1;
++ }
++
++ if (send_msgs && can_send) {
++ tpm_send_fail_message(pak, pak->req_tag);
++ }
++ packet_free(pak);
++ if (c == 1)
++ aborted = 1;
++ }
++ }
++ return aborted;
+}
+
-+static void __exit pciback_cleanup(void)
++int vtpm_release_packets(tpmif_t * tpmif, int send_msgs)
+{
-+ pciback_xenbus_unregister();
-+ pcistub_exit();
-+}
++ unsigned long flags;
+
-+module_init(pciback_init);
-+module_exit(pciback_cleanup);
++ write_lock_irqsave(&dataex.pak_lock, flags);
+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/slot.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/slot.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/slot.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/slot.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,151 @@
-+/*
-+ * PCI Backend - Provides a Virtual PCI bus (with real devices)
-+ * to the frontend
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil> (vpci.c)
-+ * Author: Tristan Gingold <tristan.gingold at bull.net>, from vpci.c
-+ */
++ dataex.aborted = _vtpm_release_packets(&dataex.pending_pak,
++ tpmif,
++ send_msgs);
++ _vtpm_release_packets(&dataex.current_pak, tpmif, send_msgs);
+
-+#include <linux/list.h>
-+#include <linux/slab.h>
-+#include <linux/pci.h>
-+#include <linux/spinlock.h>
-+#include "pciback.h"
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ return 0;
++}
+
-+/* There are at most 32 slots in a pci bus. */
-+#define PCI_SLOT_MAX 32
++static int vtpm_queue_packet(struct packet *pak)
++{
++ int rc = 0;
+
-+#define PCI_BUS_NBR 2
++ if (dataex.has_opener) {
++ unsigned long flags;
+
-+struct slot_dev_data {
-+ /* Access to dev_list must be protected by lock */
-+ struct pci_dev *slots[PCI_BUS_NBR][PCI_SLOT_MAX];
-+ spinlock_t lock;
-+};
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ list_add_tail(&pak->next, &dataex.pending_pak);
++ /* give the TPM some time to pick up the request */
++ mod_timer(&pak->processing_timer, jiffies + (30 * HZ));
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
+
-+struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
-+ unsigned int domain, unsigned int bus,
-+ unsigned int devfn)
-+{
-+ struct pci_dev *dev = NULL;
-+ struct slot_dev_data *slot_dev = pdev->pci_dev_data;
-+ unsigned long flags;
++ wake_up_interruptible(&dataex.wait_queue);
++ } else {
++ rc = -EFAULT;
++ }
++ return rc;
++}
+
-+ if (domain != 0 || PCI_FUNC(devfn) != 0)
-+ return NULL;
++static int vtpm_receive(tpmif_t * tpmif, u32 size)
++{
++ int rc = 0;
++ unsigned char buffer[10];
++ __be32 *native_size;
++ struct packet *pak = packet_alloc(tpmif, size, 0, 0);
+
-+ if (PCI_SLOT(devfn) >= PCI_SLOT_MAX || bus >= PCI_BUS_NBR)
-+ return NULL;
++ if (!pak)
++ return -ENOMEM;
++ /*
++ * Read 10 bytes from the received buffer to test its
++ * content for validity.
++ */
++ if (sizeof (buffer) != packet_read(pak,
++ sizeof (buffer), buffer,
++ sizeof (buffer), 0)) {
++ goto failexit;
++ }
++ /*
++ * Reset the packet read pointer so we can read all its
++ * contents again.
++ */
++ packet_reset(pak);
+
-+ spin_lock_irqsave(&slot_dev->lock, flags);
-+ dev = slot_dev->slots[bus][PCI_SLOT(devfn)];
-+ spin_unlock_irqrestore(&slot_dev->lock, flags);
++ native_size = (__force __be32 *) (&buffer[4 + 2]);
++ /*
++ * Verify that the size of the packet is correct
++ * as indicated and that there's actually someone reading packets.
++ * The minimum size of the packet is '10' for tag, size indicator
++ * and ordinal.
++ */
++ if (size < 10 ||
++ be32_to_cpu(*native_size) != size ||
++ 0 == dataex.has_opener || tpmif->status != CONNECTED) {
++ rc = -EINVAL;
++ goto failexit;
++ } else {
++ rc = vtpm_queue_packet(pak);
++ if (rc < 0)
++ goto failexit;
++ }
++ return 0;
+
-+ return dev;
++ failexit:
++ if (pak) {
++ tpm_send_fail_message(pak, buffer[4 + 1]);
++ packet_free(pak);
++ }
++ return rc;
+}
+
-+int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
++/*
++ * Timeout function that gets invoked when a packet has not been processed
++ * during the timeout period.
++ * The packet must be on a list when this function is invoked. This
++ * also means that once its taken off a list, the timer must be
++ * destroyed as well.
++ */
++static void processing_timeout(unsigned long ptr)
+{
-+ int err = 0, slot, bus;
-+ struct slot_dev_data *slot_dev = pdev->pci_dev_data;
++ struct packet *pak = (struct packet *)ptr;
+ unsigned long flags;
+
-+ if ((dev->class >> 24) == PCI_BASE_CLASS_BRIDGE) {
-+ err = -EFAULT;
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Can't export bridges on the virtual PCI bus");
-+ goto out;
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ /*
++ * The packet needs to be searched whether it
++ * is still on the list.
++ */
++ if (pak == packet_find_packet(&dataex.pending_pak, pak) ||
++ pak == packet_find_packet(&dataex.current_pak, pak)) {
++ if ((pak->flags & PACKET_FLAG_DISCARD_RESPONSE) == 0) {
++ tpm_send_fail_message(pak, pak->req_tag);
++ }
++ /* discard future responses */
++ pak->flags |= PACKET_FLAG_DISCARD_RESPONSE;
+ }
+
-+ spin_lock_irqsave(&slot_dev->lock, flags);
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++}
+
-+ /* Assign to a new slot on the virtual PCI bus */
-+ for (bus = 0; bus < PCI_BUS_NBR; bus++)
-+ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
-+ if (slot_dev->slots[bus][slot] == NULL) {
-+ printk(KERN_INFO
-+ "pciback: slot: %s: assign to virtual slot %d, bus %d\n",
-+ pci_name(dev), slot, bus);
-+ slot_dev->slots[bus][slot] = dev;
-+ goto unlock;
-+ }
-+ }
++static void tpm_tx_action(unsigned long unused);
++static DECLARE_TASKLET(tpm_tx_tasklet, tpm_tx_action, 0);
+
-+ err = -ENOMEM;
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "No more space on root virtual PCI bus");
++static struct list_head tpm_schedule_list;
++static spinlock_t tpm_schedule_list_lock;
+
-+ unlock:
-+ spin_unlock_irqrestore(&slot_dev->lock, flags);
-+ out:
-+ return err;
++static inline void maybe_schedule_tx_action(void)
++{
++ smp_mb();
++ tasklet_schedule(&tpm_tx_tasklet);
+}
+
-+void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
++static inline int __on_tpm_schedule_list(tpmif_t * tpmif)
+{
-+ int slot, bus;
-+ struct slot_dev_data *slot_dev = pdev->pci_dev_data;
-+ struct pci_dev *found_dev = NULL;
-+ unsigned long flags;
++ return tpmif->list.next != NULL;
++}
+
-+ spin_lock_irqsave(&slot_dev->lock, flags);
++static void remove_from_tpm_schedule_list(tpmif_t * tpmif)
++{
++ spin_lock_irq(&tpm_schedule_list_lock);
++ if (likely(__on_tpm_schedule_list(tpmif))) {
++ list_del(&tpmif->list);
++ tpmif->list.next = NULL;
++ tpmif_put(tpmif);
++ }
++ spin_unlock_irq(&tpm_schedule_list_lock);
++}
+
-+ for (bus = 0; bus < PCI_BUS_NBR; bus++)
-+ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
-+ if (slot_dev->slots[bus][slot] == dev) {
-+ slot_dev->slots[bus][slot] = NULL;
-+ found_dev = dev;
-+ goto out;
-+ }
-+ }
++static void add_to_tpm_schedule_list_tail(tpmif_t * tpmif)
++{
++ if (__on_tpm_schedule_list(tpmif))
++ return;
+
-+ out:
-+ spin_unlock_irqrestore(&slot_dev->lock, flags);
++ spin_lock_irq(&tpm_schedule_list_lock);
++ if (!__on_tpm_schedule_list(tpmif) && tpmif->active) {
++ list_add_tail(&tpmif->list, &tpm_schedule_list);
++ tpmif_get(tpmif);
++ }
++ spin_unlock_irq(&tpm_schedule_list_lock);
++}
++
++void tpmif_schedule_work(tpmif_t * tpmif)
++{
++ add_to_tpm_schedule_list_tail(tpmif);
++ maybe_schedule_tx_action();
++}
+
-+ if (found_dev)
-+ pcistub_put_pci_dev(found_dev);
++void tpmif_deschedule_work(tpmif_t * tpmif)
++{
++ remove_from_tpm_schedule_list(tpmif);
+}
+
-+int pciback_init_devices(struct pciback_device *pdev)
++static void tpm_tx_action(unsigned long unused)
+{
-+ int slot, bus;
-+ struct slot_dev_data *slot_dev;
++ struct list_head *ent;
++ tpmif_t *tpmif;
++ tpmif_tx_request_t *tx;
+
-+ slot_dev = kmalloc(sizeof(*slot_dev), GFP_KERNEL);
-+ if (!slot_dev)
-+ return -ENOMEM;
++ DPRINTK("%s: Getting data from front-end(s)!\n", __FUNCTION__);
+
-+ spin_lock_init(&slot_dev->lock);
++ while (!list_empty(&tpm_schedule_list)) {
++ /* Get a tpmif from the list with work to do. */
++ ent = tpm_schedule_list.next;
++ tpmif = list_entry(ent, tpmif_t, list);
++ tpmif_get(tpmif);
++ remove_from_tpm_schedule_list(tpmif);
+
-+ for (bus = 0; bus < PCI_BUS_NBR; bus++)
-+ for (slot = 0; slot < PCI_SLOT_MAX; slot++)
-+ slot_dev->slots[bus][slot] = NULL;
++ tx = &tpmif->tx->ring[0].req;
+
-+ pdev->pci_dev_data = slot_dev;
++ /* pass it up */
++ vtpm_receive(tpmif, tx->size);
+
-+ return 0;
++ tpmif_put(tpmif);
++ }
+}
+
-+int pciback_publish_pci_roots(struct pciback_device *pdev,
-+ publish_pci_root_cb publish_cb)
++irqreturn_t tpmif_be_int(int irq, void *dev_id, struct pt_regs *regs)
+{
-+ /* The Virtual PCI bus has only one root */
-+ return publish_cb(pdev, 0, 0);
++ tpmif_t *tpmif = (tpmif_t *) dev_id;
++
++ add_to_tpm_schedule_list_tail(tpmif);
++ maybe_schedule_tx_action();
++ return IRQ_HANDLED;
+}
+
-+void pciback_release_devices(struct pciback_device *pdev)
++static int __init tpmback_init(void)
+{
-+ int slot, bus;
-+ struct slot_dev_data *slot_dev = pdev->pci_dev_data;
-+ struct pci_dev *dev;
++ int rc;
+
-+ for (bus = 0; bus < PCI_BUS_NBR; bus++)
-+ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
-+ dev = slot_dev->slots[bus][slot];
-+ if (dev != NULL)
-+ pcistub_put_pci_dev(dev);
-+ }
++ if ((rc = misc_register(&vtpms_miscdevice)) != 0) {
++ printk(KERN_ALERT
++ "Could not register misc device for TPM BE.\n");
++ return rc;
++ }
+
-+ kfree(slot_dev);
-+ pdev->pci_dev_data = NULL;
-+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/vpci.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/vpci.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/vpci.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/vpci.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,204 @@
-+/*
-+ * PCI Backend - Provides a Virtual PCI bus (with real devices)
-+ * to the frontend
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ */
++ dataex_init(&dataex);
+
-+#include <linux/list.h>
-+#include <linux/slab.h>
-+#include <linux/pci.h>
-+#include <linux/spinlock.h>
-+#include "pciback.h"
++ spin_lock_init(&tpm_schedule_list_lock);
++ INIT_LIST_HEAD(&tpm_schedule_list);
+
-+#define PCI_SLOT_MAX 32
++ tpmif_interface_init();
++ tpmif_xenbus_init();
+
-+struct vpci_dev_data {
-+ /* Access to dev_list must be protected by lock */
-+ struct list_head dev_list[PCI_SLOT_MAX];
-+ spinlock_t lock;
-+};
++ printk(KERN_ALERT "Successfully initialized TPM backend driver.\n");
+
-+static inline struct list_head *list_first(struct list_head *head)
++ return 0;
++}
++
++module_init(tpmback_init);
++
++void __exit tpmback_exit(void)
+{
-+ return head->next;
++ vtpm_release_packets(NULL, 0);
++ tpmif_xenbus_exit();
++ tpmif_interface_exit();
++ misc_deregister(&vtpms_miscdevice);
+}
+
-+struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
-+ unsigned int domain, unsigned int bus,
-+ unsigned int devfn)
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/tpmback/xenbus.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/tpmback/xenbus.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,289 @@
++/* Xenbus code for tpmif backend
++ Copyright (C) 2005 IBM Corporation
++ Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
++
++ This program is free software; you can redistribute it and/or modify
++ it under the terms of the GNU General Public License as published by
++ the Free Software Foundation; either version 2 of the License, or
++ (at your option) any later version.
++
++ This program is distributed in the hope that it will be useful,
++ but WITHOUT ANY WARRANTY; without even the implied warranty of
++ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ GNU General Public License for more details.
++
++ You should have received a copy of the GNU General Public License
++ along with this program; if not, write to the Free Software
++ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++*/
++#include <stdarg.h>
++#include <linux/module.h>
++#include <xen/xenbus.h>
++#include "common.h"
++
++struct backend_info
+{
-+ struct pci_dev_entry *entry;
-+ struct pci_dev *dev = NULL;
-+ struct vpci_dev_data *vpci_dev = pdev->pci_dev_data;
-+ unsigned long flags;
++ struct xenbus_device *dev;
+
-+ if (domain != 0 || bus != 0)
-+ return NULL;
++ /* our communications channel */
++ tpmif_t *tpmif;
+
-+ if (PCI_SLOT(devfn) < PCI_SLOT_MAX) {
-+ spin_lock_irqsave(&vpci_dev->lock, flags);
++ long int frontend_id;
++ long int instance; // instance of TPM
++ u8 is_instance_set;// whether instance number has been set
+
-+ list_for_each_entry(entry,
-+ &vpci_dev->dev_list[PCI_SLOT(devfn)],
-+ list) {
-+ if (PCI_FUNC(entry->dev->devfn) == PCI_FUNC(devfn)) {
-+ dev = entry->dev;
-+ break;
-+ }
-+ }
++ /* watch front end for changes */
++ struct xenbus_watch backend_watch;
++};
+
-+ spin_unlock_irqrestore(&vpci_dev->lock, flags);
-+ }
-+ return dev;
++static void maybe_connect(struct backend_info *be);
++static void connect(struct backend_info *be);
++static int connect_ring(struct backend_info *be);
++static void backend_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len);
++static void frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state);
++
++long int tpmback_get_instance(struct backend_info *bi)
++{
++ long int res = -1;
++ if (bi && bi->is_instance_set)
++ res = bi->instance;
++ return res;
+}
+
-+static inline int match_slot(struct pci_dev *l, struct pci_dev *r)
++static int tpmback_remove(struct xenbus_device *dev)
+{
-+ if (pci_domain_nr(l->bus) == pci_domain_nr(r->bus)
-+ && l->bus == r->bus && PCI_SLOT(l->devfn) == PCI_SLOT(r->devfn))
-+ return 1;
++ struct backend_info *be = dev->dev.driver_data;
+
++ if (!be) return 0;
++
++ if (be->backend_watch.node) {
++ unregister_xenbus_watch(&be->backend_watch);
++ kfree(be->backend_watch.node);
++ be->backend_watch.node = NULL;
++ }
++ if (be->tpmif) {
++ be->tpmif->bi = NULL;
++ vtpm_release_packets(be->tpmif, 0);
++ tpmif_put(be->tpmif);
++ be->tpmif = NULL;
++ }
++ kfree(be);
++ dev->dev.driver_data = NULL;
+ return 0;
+}
+
-+int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
++static int tpmback_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
+{
-+ int err = 0, slot;
-+ struct pci_dev_entry *t, *dev_entry;
-+ struct vpci_dev_data *vpci_dev = pdev->pci_dev_data;
-+ unsigned long flags;
++ int err;
++ struct backend_info *be = kzalloc(sizeof(struct backend_info),
++ GFP_KERNEL);
+
-+ if ((dev->class >> 24) == PCI_BASE_CLASS_BRIDGE) {
-+ err = -EFAULT;
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Can't export bridges on the virtual PCI bus");
-+ goto out;
++ if (!be) {
++ xenbus_dev_fatal(dev, -ENOMEM,
++ "allocating backend structure");
++ return -ENOMEM;
+ }
+
-+ dev_entry = kmalloc(sizeof(*dev_entry), GFP_KERNEL);
-+ if (!dev_entry) {
-+ err = -ENOMEM;
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error adding entry to virtual PCI bus");
-+ goto out;
++ be->is_instance_set = 0;
++ be->dev = dev;
++ dev->dev.driver_data = be;
++
++ err = xenbus_watch_path2(dev, dev->nodename,
++ "instance", &be->backend_watch,
++ backend_changed);
++ if (err) {
++ goto fail;
+ }
+
-+ dev_entry->dev = dev;
++ err = xenbus_switch_state(dev, XenbusStateInitWait);
++ if (err) {
++ goto fail;
++ }
++ return 0;
++fail:
++ tpmback_remove(dev);
++ return err;
++}
+
-+ spin_lock_irqsave(&vpci_dev->lock, flags);
+
-+ /* Keep multi-function devices together on the virtual PCI bus */
-+ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
-+ if (!list_empty(&vpci_dev->dev_list[slot])) {
-+ t = list_entry(list_first(&vpci_dev->dev_list[slot]),
-+ struct pci_dev_entry, list);
++static void backend_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ int err;
++ long instance;
++ struct backend_info *be
++ = container_of(watch, struct backend_info, backend_watch);
++ struct xenbus_device *dev = be->dev;
+
-+ if (match_slot(dev, t->dev)) {
-+ pr_info("pciback: vpci: %s: "
-+ "assign to virtual slot %d func %d\n",
-+ pci_name(dev), slot,
-+ PCI_FUNC(dev->devfn));
-+ list_add_tail(&dev_entry->list,
-+ &vpci_dev->dev_list[slot]);
-+ goto unlock;
-+ }
-+ }
++ err = xenbus_scanf(XBT_NIL, dev->nodename,
++ "instance","%li", &instance);
++ if (XENBUS_EXIST_ERR(err)) {
++ return;
+ }
+
-+ /* Assign to a new slot on the virtual PCI bus */
-+ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
-+ if (list_empty(&vpci_dev->dev_list[slot])) {
-+ printk(KERN_INFO
-+ "pciback: vpci: %s: assign to virtual slot %d\n",
-+ pci_name(dev), slot);
-+ list_add_tail(&dev_entry->list,
-+ &vpci_dev->dev_list[slot]);
-+ goto unlock;
-+ }
++ if (err != 1) {
++ xenbus_dev_fatal(dev, err, "reading instance");
++ return;
+ }
+
-+ err = -ENOMEM;
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "No more space on root virtual PCI bus");
-+
-+ unlock:
-+ spin_unlock_irqrestore(&vpci_dev->lock, flags);
-+ out:
-+ return err;
++ if (be->is_instance_set == 0) {
++ be->instance = instance;
++ be->is_instance_set = 1;
++ }
+}
+
-+void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
++
++static void frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state)
+{
-+ int slot;
-+ struct vpci_dev_data *vpci_dev = pdev->pci_dev_data;
-+ struct pci_dev *found_dev = NULL;
-+ unsigned long flags;
++ struct backend_info *be = dev->dev.driver_data;
++ int err;
+
-+ spin_lock_irqsave(&vpci_dev->lock, flags);
++ switch (frontend_state) {
++ case XenbusStateInitialising:
++ case XenbusStateInitialised:
++ break;
+
-+ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
-+ struct pci_dev_entry *e, *tmp;
-+ list_for_each_entry_safe(e, tmp, &vpci_dev->dev_list[slot],
-+ list) {
-+ if (e->dev == dev) {
-+ list_del(&e->list);
-+ found_dev = e->dev;
-+ kfree(e);
-+ goto out;
-+ }
++ case XenbusStateConnected:
++ err = connect_ring(be);
++ if (err) {
++ return;
+ }
-+ }
++ maybe_connect(be);
++ break;
+
-+ out:
-+ spin_unlock_irqrestore(&vpci_dev->lock, flags);
++ case XenbusStateClosing:
++ be->instance = -1;
++ xenbus_switch_state(dev, XenbusStateClosing);
++ break;
+
-+ if (found_dev)
-+ pcistub_put_pci_dev(found_dev);
++ case XenbusStateUnknown: /* keep it here */
++ case XenbusStateClosed:
++ xenbus_switch_state(dev, XenbusStateClosed);
++ device_unregister(&be->dev->dev);
++ tpmback_remove(dev);
++ break;
++
++ default:
++ xenbus_dev_fatal(dev, -EINVAL,
++ "saw state %d at frontend",
++ frontend_state);
++ break;
++ }
+}
+
-+int pciback_init_devices(struct pciback_device *pdev)
++
++
++static void maybe_connect(struct backend_info *be)
+{
-+ int slot;
-+ struct vpci_dev_data *vpci_dev;
++ if (be->tpmif == NULL || be->tpmif->status == CONNECTED)
++ return;
+
-+ vpci_dev = kmalloc(sizeof(*vpci_dev), GFP_KERNEL);
-+ if (!vpci_dev)
-+ return -ENOMEM;
++ connect(be);
++}
+
-+ spin_lock_init(&vpci_dev->lock);
+
-+ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
-+ INIT_LIST_HEAD(&vpci_dev->dev_list[slot]);
++static void connect(struct backend_info *be)
++{
++ struct xenbus_transaction xbt;
++ int err;
++ struct xenbus_device *dev = be->dev;
++ unsigned long ready = 1;
++
++again:
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(be->dev, err, "starting transaction");
++ return;
+ }
+
-+ pdev->pci_dev_data = vpci_dev;
++ err = xenbus_printf(xbt, be->dev->nodename,
++ "ready", "%lu", ready);
++ if (err) {
++ xenbus_dev_fatal(be->dev, err, "writing 'ready'");
++ goto abort;
++ }
+
-+ return 0;
-+}
++ err = xenbus_transaction_end(xbt, 0);
++ if (err == -EAGAIN)
++ goto again;
++ if (err)
++ xenbus_dev_fatal(be->dev, err, "end of transaction");
+
-+int pciback_publish_pci_roots(struct pciback_device *pdev,
-+ publish_pci_root_cb publish_cb)
-+{
-+ /* The Virtual PCI bus has only one root */
-+ return publish_cb(pdev, 0, 0);
++ err = xenbus_switch_state(dev, XenbusStateConnected);
++ if (!err)
++ be->tpmif->status = CONNECTED;
++ return;
++abort:
++ xenbus_transaction_end(xbt, 1);
+}
+
-+void pciback_release_devices(struct pciback_device *pdev)
++
++static int connect_ring(struct backend_info *be)
+{
-+ int slot;
-+ struct vpci_dev_data *vpci_dev = pdev->pci_dev_data;
++ struct xenbus_device *dev = be->dev;
++ unsigned long ring_ref;
++ unsigned int evtchn;
++ int err;
+
-+ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
-+ struct pci_dev_entry *e, *tmp;
-+ list_for_each_entry_safe(e, tmp, &vpci_dev->dev_list[slot],
-+ list) {
-+ list_del(&e->list);
-+ pcistub_put_pci_dev(e->dev);
-+ kfree(e);
++ err = xenbus_gather(XBT_NIL, dev->otherend,
++ "ring-ref", "%lu", &ring_ref,
++ "event-channel", "%u", &evtchn, NULL);
++ if (err) {
++ xenbus_dev_error(dev, err,
++ "reading %s/ring-ref and event-channel",
++ dev->otherend);
++ return err;
++ }
++
++ if (!be->tpmif) {
++ be->tpmif = tpmif_find(dev->otherend_id, be);
++ if (IS_ERR(be->tpmif)) {
++ err = PTR_ERR(be->tpmif);
++ be->tpmif = NULL;
++ xenbus_dev_fatal(dev,err,"creating vtpm interface");
++ return err;
+ }
+ }
+
-+ kfree(vpci_dev);
-+ pdev->pci_dev_data = NULL;
++ if (be->tpmif != NULL) {
++ err = tpmif_map(be->tpmif, ring_ref, evtchn);
++ if (err) {
++ xenbus_dev_error(dev, err,
++ "mapping shared-frame %lu port %u",
++ ring_ref, evtchn);
++ return err;
++ }
++ }
++ return 0;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pciback/xenbus.c tmp-linux-2.6-xen.patch/drivers/xen/pciback/xenbus.c
---- pristine-linux-2.6.18.2/drivers/xen/pciback/xenbus.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pciback/xenbus.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,454 @@
-+/*
-+ * PCI Backend Xenbus Setup - handles setup with frontend and xend
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ */
-+#include <linux/module.h>
-+#include <linux/init.h>
-+#include <linux/list.h>
-+#include <linux/vmalloc.h>
-+#include <xen/xenbus.h>
-+#include <xen/evtchn.h>
-+#include "pciback.h"
-+
-+#define INVALID_EVTCHN_IRQ (-1)
+
-+static struct pciback_device *alloc_pdev(struct xenbus_device *xdev)
-+{
-+ struct pciback_device *pdev;
-+
-+ pdev = kzalloc(sizeof(struct pciback_device), GFP_KERNEL);
-+ if (pdev == NULL)
-+ goto out;
-+ dev_dbg(&xdev->dev, "allocated pdev @ 0x%p\n", pdev);
+
-+ pdev->xdev = xdev;
-+ xdev->dev.driver_data = pdev;
++static const struct xenbus_device_id tpmback_ids[] = {
++ { "vtpm" },
++ { "" }
++};
+
-+ spin_lock_init(&pdev->dev_lock);
+
-+ pdev->sh_area = NULL;
-+ pdev->sh_info = NULL;
-+ pdev->evtchn_irq = INVALID_EVTCHN_IRQ;
-+ pdev->be_watching = 0;
++static struct xenbus_driver tpmback = {
++ .name = "vtpm",
++ .owner = THIS_MODULE,
++ .ids = tpmback_ids,
++ .probe = tpmback_probe,
++ .remove = tpmback_remove,
++ .otherend_changed = frontend_changed,
++};
+
-+ INIT_WORK(&pdev->op_work, pciback_do_op, pdev);
+
-+ if (pciback_init_devices(pdev)) {
-+ kfree(pdev);
-+ pdev = NULL;
-+ }
-+ out:
-+ return pdev;
++void tpmif_xenbus_init(void)
++{
++ xenbus_register_backend(&tpmback);
+}
+
-+static void free_pdev(struct pciback_device *pdev)
++void tpmif_xenbus_exit(void)
+{
-+ if (pdev->be_watching)
-+ unregister_xenbus_watch(&pdev->be_watch);
-+
-+ /* Ensure the guest can't trigger our handler before removing devices */
-+ if (pdev->evtchn_irq != INVALID_EVTCHN_IRQ)
-+ unbind_from_irqhandler(pdev->evtchn_irq, pdev);
++ xenbus_unregister_driver(&tpmback);
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/util.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/util.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,65 @@
++#include <linux/mm.h>
++#include <linux/module.h>
++#include <linux/slab.h>
++#include <linux/vmalloc.h>
++#include <asm/uaccess.h>
++#include <xen/driver_util.h>
+
-+ /* If the driver domain started an op, make sure we complete it or
-+ * delete it before releasing the shared memory */
-+ cancel_delayed_work(&pdev->op_work);
-+ flush_scheduled_work();
++struct class *get_xen_class(void)
++{
++ static struct class *xen_class;
+
-+ if (pdev->sh_info)
-+ xenbus_unmap_ring_vfree(pdev->xdev, pdev->sh_area);
++ if (xen_class)
++ return xen_class;
+
-+ pciback_release_devices(pdev);
++ xen_class = class_create(THIS_MODULE, "xen");
++ if (IS_ERR(xen_class)) {
++ printk("Failed to create xen sysfs class.\n");
++ xen_class = NULL;
++ }
+
-+ pdev->xdev->dev.driver_data = NULL;
-+ pdev->xdev = NULL;
++ return xen_class;
++}
++EXPORT_SYMBOL_GPL(get_xen_class);
+
-+ kfree(pdev);
++#ifdef CONFIG_X86
++static int f(pte_t *pte, struct page *pmd_page, unsigned long addr, void *data)
++{
++ /* apply_to_page_range() does all the hard work. */
++ return 0;
+}
+
-+static int pciback_do_attach(struct pciback_device *pdev, int gnt_ref,
-+ int remote_evtchn)
++struct vm_struct *alloc_vm_area(unsigned long size)
+{
-+ int err = 0;
+ struct vm_struct *area;
+
-+ dev_dbg(&pdev->xdev->dev,
-+ "Attaching to frontend resources - gnt_ref=%d evtchn=%d\n",
-+ gnt_ref, remote_evtchn);
++ area = get_vm_area(size, VM_IOREMAP);
++ if (area == NULL)
++ return NULL;
+
-+ area = xenbus_map_ring_valloc(pdev->xdev, gnt_ref);
-+ if (IS_ERR(area)) {
-+ err = PTR_ERR(area);
-+ goto out;
++ /*
++ * This ensures that page tables are constructed for this region
++ * of kernel virtual address space and mapped into init_mm.
++ */
++ if (apply_to_page_range(&init_mm, (unsigned long)area->addr,
++ area->size, f, NULL)) {
++ free_vm_area(area);
++ return NULL;
+ }
-+ pdev->sh_area = area;
-+ pdev->sh_info = area->addr;
+
-+ err = bind_interdomain_evtchn_to_irqhandler(
-+ pdev->xdev->otherend_id, remote_evtchn, pciback_handle_event,
-+ SA_SAMPLE_RANDOM, "pciback", pdev);
-+ if (err < 0) {
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error binding event channel to IRQ");
-+ goto out;
-+ }
-+ pdev->evtchn_irq = err;
-+ err = 0;
++ /* Map page directories into every address space. */
++ vmalloc_sync_all();
+
-+ dev_dbg(&pdev->xdev->dev, "Attached!\n");
-+ out:
-+ return err;
++ return area;
+}
++EXPORT_SYMBOL_GPL(alloc_vm_area);
+
-+static int pciback_attach(struct pciback_device *pdev)
++void free_vm_area(struct vm_struct *area)
+{
-+ int err = 0;
-+ int gnt_ref, remote_evtchn;
-+ char *magic = NULL;
++ struct vm_struct *ret;
++ ret = remove_vm_area(area->addr);
++ BUG_ON(ret != area);
++ kfree(area);
++}
++EXPORT_SYMBOL_GPL(free_vm_area);
++#endif /* CONFIG_X86 */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/xenbus/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/xenbus/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,9 @@
++obj-y += xenbus_client.o xenbus_comms.o xenbus_xs.o xenbus_probe.o
++obj-$(CONFIG_XEN_BACKEND) += xenbus_be.o
+
-+ spin_lock(&pdev->dev_lock);
++xenbus_be-objs =
++xenbus_be-objs += xenbus_backend_client.o
+
-+ /* Make sure we only do this setup once */
-+ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
-+ XenbusStateInitialised)
-+ goto out;
++xenbus-$(CONFIG_XEN_BACKEND) += xenbus_probe_backend.o
++obj-y += $(xenbus-y) $(xenbus-m)
++obj-$(CONFIG_XEN_XENBUS_DEV) += xenbus_dev.o
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/xenbus/xenbus_backend_client.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/xenbus/xenbus_backend_client.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,147 @@
++/******************************************************************************
++ * Backend-client-facing interface for the Xenbus driver. In other words, the
++ * interface between the Xenbus and the device-specific code in the backend
++ * driver.
++ *
++ * Copyright (C) 2005-2006 XenSource Ltd
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ /* Wait for frontend to state that it has published the configuration */
-+ if (xenbus_read_driver_state(pdev->xdev->otherend) !=
-+ XenbusStateInitialised)
-+ goto out;
++#include <linux/err.h>
++#include <xen/gnttab.h>
++#include <xen/xenbus.h>
++#include <xen/driver_util.h>
+
-+ dev_dbg(&pdev->xdev->dev, "Reading frontend config\n");
++/* Based on Rusty Russell's skeleton driver's map_page */
++struct vm_struct *xenbus_map_ring_valloc(struct xenbus_device *dev, int gnt_ref)
++{
++ struct gnttab_map_grant_ref op;
++ struct vm_struct *area;
+
-+ err = xenbus_gather(XBT_NIL, pdev->xdev->otherend,
-+ "pci-op-ref", "%u", &gnt_ref,
-+ "event-channel", "%u", &remote_evtchn,
-+ "magic", NULL, &magic, NULL);
-+ if (err) {
-+ /* If configuration didn't get read correctly, wait longer */
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error reading configuration from frontend");
-+ goto out;
-+ }
++ area = alloc_vm_area(PAGE_SIZE);
++ if (!area)
++ return ERR_PTR(-ENOMEM);
+
-+ if (magic == NULL || strcmp(magic, XEN_PCI_MAGIC) != 0) {
-+ xenbus_dev_fatal(pdev->xdev, -EFAULT,
-+ "version mismatch (%s/%s) with pcifront - "
-+ "halting pciback",
-+ magic, XEN_PCI_MAGIC);
-+ goto out;
++ gnttab_set_map_op(&op, (unsigned long)area->addr, GNTMAP_host_map,
++ gnt_ref, dev->otherend_id);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
++
++ if (op.status != GNTST_okay) {
++ free_vm_area(area);
++ xenbus_dev_fatal(dev, op.status,
++ "mapping in shared page %d from domain %d",
++ gnt_ref, dev->otherend_id);
++ BUG_ON(!IS_ERR(ERR_PTR(op.status)));
++ return ERR_PTR(op.status);
+ }
+
-+ err = pciback_do_attach(pdev, gnt_ref, remote_evtchn);
-+ if (err)
-+ goto out;
++ /* Stuff the handle in an unused field */
++ area->phys_addr = (unsigned long)op.handle;
+
-+ dev_dbg(&pdev->xdev->dev, "Connecting...\n");
++ return area;
++}
++EXPORT_SYMBOL_GPL(xenbus_map_ring_valloc);
+
-+ err = xenbus_switch_state(pdev->xdev, XenbusStateConnected);
-+ if (err)
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error switching to connected state!");
+
-+ dev_dbg(&pdev->xdev->dev, "Connected? %d\n", err);
-+ out:
-+ spin_unlock(&pdev->dev_lock);
++int xenbus_map_ring(struct xenbus_device *dev, int gnt_ref,
++ grant_handle_t *handle, void *vaddr)
++{
++ struct gnttab_map_grant_ref op;
++
++ gnttab_set_map_op(&op, (unsigned long)vaddr, GNTMAP_host_map,
++ gnt_ref, dev->otherend_id);
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
+
-+ if (magic)
-+ kfree(magic);
++ if (op.status != GNTST_okay) {
++ xenbus_dev_fatal(dev, op.status,
++ "mapping in shared page %d from domain %d",
++ gnt_ref, dev->otherend_id);
++ } else
++ *handle = op.handle;
+
-+ return err;
++ return op.status;
+}
++EXPORT_SYMBOL_GPL(xenbus_map_ring);
+
-+static void pciback_frontend_changed(struct xenbus_device *xdev,
-+ enum xenbus_state fe_state)
-+{
-+ struct pciback_device *pdev = xdev->dev.driver_data;
+
-+ dev_dbg(&xdev->dev, "fe state changed %d\n", fe_state);
++/* Based on Rusty Russell's skeleton driver's unmap_page */
++int xenbus_unmap_ring_vfree(struct xenbus_device *dev, struct vm_struct *area)
++{
++ struct gnttab_unmap_grant_ref op;
+
-+ switch (fe_state) {
-+ case XenbusStateInitialised:
-+ pciback_attach(pdev);
-+ break;
++ gnttab_set_unmap_op(&op, (unsigned long)area->addr, GNTMAP_host_map,
++ (grant_handle_t)area->phys_addr);
+
-+ case XenbusStateClosing:
-+ xenbus_switch_state(xdev, XenbusStateClosing);
-+ break;
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
+
-+ case XenbusStateUnknown:
-+ case XenbusStateClosed:
-+ dev_dbg(&xdev->dev, "frontend is gone! unregister device\n");
-+ device_unregister(&xdev->dev);
-+ break;
++ if (op.status == GNTST_okay)
++ free_vm_area(area);
++ else
++ xenbus_dev_error(dev, op.status,
++ "unmapping page at handle %d error %d",
++ (int16_t)area->phys_addr, op.status);
+
-+ default:
-+ break;
-+ }
++ return op.status;
+}
++EXPORT_SYMBOL_GPL(xenbus_unmap_ring_vfree);
+
-+static int pciback_publish_pci_root(struct pciback_device *pdev,
-+ unsigned int domain, unsigned int bus)
++
++int xenbus_unmap_ring(struct xenbus_device *dev,
++ grant_handle_t handle, void *vaddr)
+{
-+ unsigned int d, b;
-+ int i, root_num, len, err;
-+ char str[64];
++ struct gnttab_unmap_grant_ref op;
+
-+ dev_dbg(&pdev->xdev->dev, "Publishing pci roots\n");
++ gnttab_set_unmap_op(&op, (unsigned long)vaddr, GNTMAP_host_map,
++ handle);
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
+
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
-+ "root_num", "%d", &root_num);
-+ if (err == 0 || err == -ENOENT)
-+ root_num = 0;
-+ else if (err < 0)
-+ goto out;
++ if (op.status != GNTST_okay)
++ xenbus_dev_error(dev, op.status,
++ "unmapping page at handle %d error %d",
++ handle, op.status);
+
-+ /* Verify that we haven't already published this pci root */
-+ for (i = 0; i < root_num; i++) {
-+ len = snprintf(str, sizeof(str), "root-%d", i);
-+ if (unlikely(len >= (sizeof(str) - 1))) {
-+ err = -ENOMEM;
-+ goto out;
-+ }
++ return op.status;
++}
++EXPORT_SYMBOL_GPL(xenbus_unmap_ring);
+
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
-+ str, "%x:%x", &d, &b);
-+ if (err < 0)
-+ goto out;
-+ if (err != 2) {
-+ err = -EINVAL;
-+ goto out;
-+ }
++int xenbus_dev_is_online(struct xenbus_device *dev)
++{
++ int rc, val;
+
-+ if (d == domain && b == bus) {
-+ err = 0;
-+ goto out;
-+ }
-+ }
++ rc = xenbus_scanf(XBT_NIL, dev->nodename, "online", "%d", &val);
++ if (rc != 1)
++ val = 0; /* no online node present */
+
-+ len = snprintf(str, sizeof(str), "root-%d", root_num);
-+ if (unlikely(len >= (sizeof(str) - 1))) {
-+ err = -ENOMEM;
-+ goto out;
-+ }
++ return val;
++}
++EXPORT_SYMBOL_GPL(xenbus_dev_is_online);
+
-+ dev_dbg(&pdev->xdev->dev, "writing root %d at %04x:%02x\n",
-+ root_num, domain, bus);
++MODULE_LICENSE("Dual BSD/GPL");
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/xenbus/xenbus_client.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/xenbus/xenbus_client.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,284 @@
++/******************************************************************************
++ * Client-facing interface for the Xenbus driver. In other words, the
++ * interface between the Xenbus and the device-specific code, be it the
++ * frontend or the backend of that driver.
++ *
++ * Copyright (C) 2005 XenSource Ltd
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, str,
-+ "%04x:%02x", domain, bus);
-+ if (err)
-+ goto out;
++#include <linux/slab.h>
++#include <xen/evtchn.h>
++#include <xen/gnttab.h>
++#include <xen/xenbus.h>
++#include <xen/driver_util.h>
+
-+ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename,
-+ "root_num", "%d", (root_num + 1));
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
+
-+ out:
-+ return err;
++#define DPRINTK(fmt, args...) \
++ pr_debug("xenbus_client (%s:%d) " fmt ".\n", __FUNCTION__, __LINE__, ##args)
++
++const char *xenbus_strstate(enum xenbus_state state)
++{
++ static const char *const name[] = {
++ [ XenbusStateUnknown ] = "Unknown",
++ [ XenbusStateInitialising ] = "Initialising",
++ [ XenbusStateInitWait ] = "InitWait",
++ [ XenbusStateInitialised ] = "Initialised",
++ [ XenbusStateConnected ] = "Connected",
++ [ XenbusStateClosing ] = "Closing",
++ [ XenbusStateClosed ] = "Closed",
++ };
++ return (state < ARRAY_SIZE(name)) ? name[state] : "INVALID";
+}
++EXPORT_SYMBOL_GPL(xenbus_strstate);
+
-+static int pciback_export_device(struct pciback_device *pdev,
-+ int domain, int bus, int slot, int func)
++int xenbus_watch_path(struct xenbus_device *dev, const char *path,
++ struct xenbus_watch *watch,
++ void (*callback)(struct xenbus_watch *,
++ const char **, unsigned int))
+{
-+ struct pci_dev *dev;
-+ int err = 0;
++ int err;
+
-+ dev_dbg(&pdev->xdev->dev, "exporting dom %x bus %x slot %x func %x\n",
-+ domain, bus, slot, func);
++ watch->node = path;
++ watch->callback = callback;
+
-+ dev = pcistub_get_pci_dev_by_slot(pdev, domain, bus, slot, func);
-+ if (!dev) {
-+ err = -EINVAL;
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Couldn't locate PCI device "
-+ "(%04x:%02x:%02x.%01x)! "
-+ "perhaps already in-use?",
-+ domain, bus, slot, func);
-+ goto out;
++ err = register_xenbus_watch(watch);
++
++ if (err) {
++ watch->node = NULL;
++ watch->callback = NULL;
++ xenbus_dev_fatal(dev, err, "adding watch on %s", path);
+ }
+
-+ err = pciback_add_pci_dev(pdev, dev);
-+ if (err)
-+ goto out;
++ return err;
++}
++EXPORT_SYMBOL_GPL(xenbus_watch_path);
+
-+ /* TODO: It'd be nice to export a bridge and have all of its children
-+ * get exported with it. This may be best done in xend (which will
-+ * have to calculate resource usage anyway) but we probably want to
-+ * put something in here to ensure that if a bridge gets given to a
-+ * driver domain, that all devices under that bridge are not given
-+ * to other driver domains (as he who controls the bridge can disable
-+ * it and stop the other devices from working).
-+ */
-+ out:
++
++int xenbus_watch_path2(struct xenbus_device *dev, const char *path,
++ const char *path2, struct xenbus_watch *watch,
++ void (*callback)(struct xenbus_watch *,
++ const char **, unsigned int))
++{
++ int err;
++ char *state = kasprintf(GFP_NOIO | __GFP_HIGH, "%s/%s", path, path2);
++ if (!state) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating path for watch");
++ return -ENOMEM;
++ }
++ err = xenbus_watch_path(dev, state, watch, callback);
++
++ if (err)
++ kfree(state);
+ return err;
+}
++EXPORT_SYMBOL_GPL(xenbus_watch_path2);
+
-+static int pciback_setup_backend(struct pciback_device *pdev)
++
++int xenbus_switch_state(struct xenbus_device *dev, enum xenbus_state state)
+{
-+ /* Get configuration from xend (if available now) */
-+ int domain, bus, slot, func;
-+ int err = 0;
-+ int i, num_devs;
-+ char dev_str[64];
++ /* We check whether the state is currently set to the given value, and
++ if not, then the state is set. We don't want to unconditionally
++ write the given state, because we don't want to fire watches
++ unnecessarily. Furthermore, if the node has gone, we don't write
++ to it, as the device will be tearing down, and we don't want to
++ resurrect that directory.
+
-+ spin_lock(&pdev->dev_lock);
++ Note that, because of this cached value of our state, this function
++ will not work inside a Xenstore transaction (something it was
++ trying to in the past) because dev->state would not get reset if
++ the transaction was aborted.
+
-+ /* It's possible we could get the call to setup twice, so make sure
-+ * we're not already connected.
+ */
-+ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
-+ XenbusStateInitWait)
-+ goto out;
+
-+ dev_dbg(&pdev->xdev->dev, "getting be setup\n");
++ int current_state;
++ int err;
+
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename, "num_devs", "%d",
-+ &num_devs);
-+ if (err != 1) {
-+ if (err >= 0)
-+ err = -EINVAL;
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error reading number of devices");
-+ goto out;
++ if (state == dev->state)
++ return 0;
++
++ err = xenbus_scanf(XBT_NIL, dev->nodename, "state", "%d",
++ ¤t_state);
++ if (err != 1)
++ return 0;
++
++ err = xenbus_printf(XBT_NIL, dev->nodename, "state", "%d", state);
++ if (err) {
++ if (state != XenbusStateClosing) /* Avoid looping */
++ xenbus_dev_fatal(dev, err, "writing new state");
++ return err;
+ }
+
-+ for (i = 0; i < num_devs; i++) {
-+ int l = snprintf(dev_str, sizeof(dev_str), "dev-%d", i);
-+ if (unlikely(l >= (sizeof(dev_str) - 1))) {
-+ err = -ENOMEM;
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "String overflow while reading "
-+ "configuration");
-+ goto out;
-+ }
++ dev->state = state;
+
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename, dev_str,
-+ "%x:%x:%x.%x", &domain, &bus, &slot, &func);
-+ if (err < 0) {
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error reading device configuration");
-+ goto out;
-+ }
-+ if (err != 4) {
-+ err = -EINVAL;
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error parsing pci device "
-+ "configuration");
-+ goto out;
-+ }
++ return 0;
++}
++EXPORT_SYMBOL_GPL(xenbus_switch_state);
+
-+ err = pciback_export_device(pdev, domain, bus, slot, func);
-+ if (err)
-+ goto out;
-+ }
++int xenbus_frontend_closed(struct xenbus_device *dev)
++{
++ xenbus_switch_state(dev, XenbusStateClosed);
++ complete(&dev->down);
++ return 0;
++}
++EXPORT_SYMBOL_GPL(xenbus_frontend_closed);
++
++/**
++ * Return the path to the error node for the given device, or NULL on failure.
++ * If the value returned is non-NULL, then it is the caller's to kfree.
++ */
++static char *error_path(struct xenbus_device *dev)
++{
++ return kasprintf(GFP_KERNEL, "error/%s", dev->nodename);
++}
+
-+ err = pciback_publish_pci_roots(pdev, pciback_publish_pci_root);
-+ if (err) {
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error while publish PCI root buses "
-+ "for frontend");
-+ goto out;
-+ }
+
-+ err = xenbus_switch_state(pdev->xdev, XenbusStateInitialised);
-+ if (err)
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error switching to initialised state!");
++void _dev_error(struct xenbus_device *dev, int err, const char *fmt,
++ va_list ap)
++{
++ int ret;
++ unsigned int len;
++ char *printf_buffer = NULL, *path_buffer = NULL;
+
-+ out:
-+ spin_unlock(&pdev->dev_lock);
++#define PRINTF_BUFFER_SIZE 4096
++ printf_buffer = kmalloc(PRINTF_BUFFER_SIZE, GFP_KERNEL);
++ if (printf_buffer == NULL)
++ goto fail;
+
-+ if (!err)
-+ /* see if pcifront is already configured (if not, we'll wait) */
-+ pciback_attach(pdev);
++ len = sprintf(printf_buffer, "%i ", -err);
++ ret = vsnprintf(printf_buffer+len, PRINTF_BUFFER_SIZE-len, fmt, ap);
+
-+ return err;
-+}
++ BUG_ON(len + ret > PRINTF_BUFFER_SIZE-1);
+
-+static void pciback_be_watch(struct xenbus_watch *watch,
-+ const char **vec, unsigned int len)
-+{
-+ struct pciback_device *pdev =
-+ container_of(watch, struct pciback_device, be_watch);
++ dev_err(&dev->dev, "%s\n", printf_buffer);
+
-+ switch (xenbus_read_driver_state(pdev->xdev->nodename)) {
-+ case XenbusStateInitWait:
-+ pciback_setup_backend(pdev);
-+ break;
++ path_buffer = error_path(dev);
+
-+ default:
-+ break;
++ if (path_buffer == NULL) {
++ printk("xenbus: failed to write error node for %s (%s)\n",
++ dev->nodename, printf_buffer);
++ goto fail;
+ }
-+}
-+
-+static int pciback_xenbus_probe(struct xenbus_device *dev,
-+ const struct xenbus_device_id *id)
-+{
-+ int err = 0;
-+ struct pciback_device *pdev = alloc_pdev(dev);
+
-+ if (pdev == NULL) {
-+ err = -ENOMEM;
-+ xenbus_dev_fatal(dev, err,
-+ "Error allocating pciback_device struct");
-+ goto out;
++ if (xenbus_write(XBT_NIL, path_buffer, "error", printf_buffer) != 0) {
++ printk("xenbus: failed to write error node for %s (%s)\n",
++ dev->nodename, printf_buffer);
++ goto fail;
+ }
+
-+ /* wait for xend to configure us */
-+ err = xenbus_switch_state(dev, XenbusStateInitWait);
-+ if (err)
-+ goto out;
++fail:
++ if (printf_buffer)
++ kfree(printf_buffer);
++ if (path_buffer)
++ kfree(path_buffer);
++}
+
-+ /* watch the backend node for backend configuration information */
-+ err = xenbus_watch_path(dev, dev->nodename, &pdev->be_watch,
-+ pciback_be_watch);
-+ if (err)
-+ goto out;
-+ pdev->be_watching = 1;
+
-+ /* We need to force a call to our callback here in case
-+ * xend already configured us!
-+ */
-+ pciback_be_watch(&pdev->be_watch, NULL, 0);
++void xenbus_dev_error(struct xenbus_device *dev, int err, const char *fmt,
++ ...)
++{
++ va_list ap;
+
-+ out:
-+ return err;
++ va_start(ap, fmt);
++ _dev_error(dev, err, fmt, ap);
++ va_end(ap);
+}
++EXPORT_SYMBOL_GPL(xenbus_dev_error);
+
-+static int pciback_xenbus_remove(struct xenbus_device *dev)
++
++void xenbus_dev_fatal(struct xenbus_device *dev, int err, const char *fmt,
++ ...)
+{
-+ struct pciback_device *pdev = dev->dev.driver_data;
++ va_list ap;
+
-+ if (pdev != NULL)
-+ free_pdev(pdev);
++ va_start(ap, fmt);
++ _dev_error(dev, err, fmt, ap);
++ va_end(ap);
+
-+ return 0;
++ xenbus_switch_state(dev, XenbusStateClosing);
+}
++EXPORT_SYMBOL_GPL(xenbus_dev_fatal);
+
-+static struct xenbus_device_id xenpci_ids[] = {
-+ {"pci"},
-+ {{0}},
-+};
-+
-+static struct xenbus_driver xenbus_pciback_driver = {
-+ .name = "pciback",
-+ .owner = THIS_MODULE,
-+ .ids = xenpci_ids,
-+ .probe = pciback_xenbus_probe,
-+ .remove = pciback_xenbus_remove,
-+ .otherend_changed = pciback_frontend_changed,
-+};
+
-+int __init pciback_xenbus_register(void)
++int xenbus_grant_ring(struct xenbus_device *dev, unsigned long ring_mfn)
+{
-+ if (!is_running_on_xen())
-+ return -ENODEV;
-+
-+ return xenbus_register_backend(&xenbus_pciback_driver);
++ int err = gnttab_grant_foreign_access(dev->otherend_id, ring_mfn, 0);
++ if (err < 0)
++ xenbus_dev_fatal(dev, err, "granting access to ring page");
++ return err;
+}
++EXPORT_SYMBOL_GPL(xenbus_grant_ring);
+
-+void __exit pciback_xenbus_unregister(void)
++
++int xenbus_alloc_evtchn(struct xenbus_device *dev, int *port)
+{
-+ xenbus_unregister_driver(&xenbus_pciback_driver);
-+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pcifront/Makefile tmp-linux-2.6-xen.patch/drivers/xen/pcifront/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/pcifront/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pcifront/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,7 @@
-+obj-y += pcifront.o
++ struct evtchn_alloc_unbound alloc_unbound;
++ int err;
+
-+pcifront-y := pci_op.o xenbus.o pci.o
++ alloc_unbound.dom = DOMID_SELF;
++ alloc_unbound.remote_dom = dev->otherend_id;
+
-+ifeq ($(CONFIG_XEN_PCIDEV_FE_DEBUG),y)
-+EXTRA_CFLAGS += -DDEBUG
-+endif
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pcifront/pci.c tmp-linux-2.6-xen.patch/drivers/xen/pcifront/pci.c
---- pristine-linux-2.6.18.2/drivers/xen/pcifront/pci.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pcifront/pci.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,46 @@
-+/*
-+ * PCI Frontend Operations - ensure only one PCI frontend runs at a time
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ */
-+#include <linux/module.h>
-+#include <linux/init.h>
-+#include <linux/pci.h>
-+#include <linux/spinlock.h>
-+#include "pcifront.h"
++ err = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
++ &alloc_unbound);
++ if (err)
++ xenbus_dev_fatal(dev, err, "allocating event channel");
++ else
++ *port = alloc_unbound.port;
+
-+DEFINE_SPINLOCK(pcifront_dev_lock);
-+static struct pcifront_device *pcifront_dev = NULL;
++ return err;
++}
++EXPORT_SYMBOL_GPL(xenbus_alloc_evtchn);
+
-+int pcifront_connect(struct pcifront_device *pdev)
-+{
-+ int err = 0;
+
-+ spin_lock(&pcifront_dev_lock);
++int xenbus_free_evtchn(struct xenbus_device *dev, int port)
++{
++ struct evtchn_close close;
++ int err;
+
-+ if (!pcifront_dev) {
-+ dev_info(&pdev->xdev->dev, "Installing PCI frontend\n");
-+ pcifront_dev = pdev;
-+ }
-+ else {
-+ dev_err(&pdev->xdev->dev, "PCI frontend already installed!\n");
-+ err = -EEXIST;
-+ }
++ close.port = port;
+
-+ spin_unlock(&pcifront_dev_lock);
++ err = HYPERVISOR_event_channel_op(EVTCHNOP_close, &close);
++ if (err)
++ xenbus_dev_error(dev, err, "freeing event channel %d", port);
+
+ return err;
+}
++EXPORT_SYMBOL_GPL(xenbus_free_evtchn);
+
-+void pcifront_disconnect(struct pcifront_device *pdev)
-+{
-+ spin_lock(&pcifront_dev_lock);
+
-+ if (pdev == pcifront_dev) {
-+ dev_info(&pdev->xdev->dev,
-+ "Disconnecting PCI Frontend Buses\n");
-+ pcifront_dev = NULL;
-+ }
++enum xenbus_state xenbus_read_driver_state(const char *path)
++{
++ enum xenbus_state result;
++ int err = xenbus_gather(XBT_NIL, path, "state", "%d", &result, NULL);
++ if (err)
++ result = XenbusStateUnknown;
+
-+ spin_unlock(&pcifront_dev_lock);
++ return result;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pcifront/pcifront.h tmp-linux-2.6-xen.patch/drivers/xen/pcifront/pcifront.h
---- pristine-linux-2.6.18.2/drivers/xen/pcifront/pcifront.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pcifront/pcifront.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,40 @@
-+/*
-+ * PCI Frontend - Common data structures & function declarations
++EXPORT_SYMBOL_GPL(xenbus_read_driver_state);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/xenbus/xenbus_comms.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/xenbus/xenbus_comms.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,233 @@
++/******************************************************************************
++ * xenbus_comms.c
+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ * Low level code to talks to Xen Store: ringbuffer and event channel.
++ *
++ * Copyright (C) 2005 Rusty Russell, IBM Corporation
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
-+#ifndef __XEN_PCIFRONT_H__
-+#define __XEN_PCIFRONT_H__
+
-+#include <linux/spinlock.h>
-+#include <linux/pci.h>
++#include <linux/wait.h>
++#include <linux/interrupt.h>
++#include <linux/sched.h>
++#include <linux/err.h>
++#include <linux/ptrace.h>
++#include <linux/workqueue.h>
++#include <xen/evtchn.h>
+#include <xen/xenbus.h>
-+#include <xen/interface/io/pciif.h>
-+#include <xen/pcifront.h>
+
-+struct pci_bus_entry {
-+ struct list_head list;
-+ struct pci_bus *bus;
-+};
++#include <asm/hypervisor.h>
+
-+struct pcifront_device {
-+ struct xenbus_device *xdev;
-+ struct list_head root_buses;
-+ spinlock_t dev_lock;
++#include "xenbus_comms.h"
+
-+ int evtchn;
-+ int gnt_ref;
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
+
-+ /* Lock this when doing any operations in sh_info */
-+ spinlock_t sh_info_lock;
-+ struct xen_pci_sharedinfo *sh_info;
-+};
++static int xenbus_irq;
+
-+int pcifront_connect(struct pcifront_device *pdev);
-+void pcifront_disconnect(struct pcifront_device *pdev);
++extern void xenbus_probe(void *);
++extern int xenstored_ready;
++static DECLARE_WORK(probe_work, xenbus_probe, NULL);
+
-+int pcifront_scan_root(struct pcifront_device *pdev,
-+ unsigned int domain, unsigned int bus);
-+void pcifront_free_roots(struct pcifront_device *pdev);
++static DECLARE_WAIT_QUEUE_HEAD(xb_waitq);
+
-+#endif /* __XEN_PCIFRONT_H__ */
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pcifront/pci_op.c tmp-linux-2.6-xen.patch/drivers/xen/pcifront/pci_op.c
---- pristine-linux-2.6.18.2/drivers/xen/pcifront/pci_op.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pcifront/pci_op.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,386 @@
-+/*
-+ * PCI Frontend Operations - Communicates with frontend
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ */
-+#include <linux/module.h>
-+#include <linux/version.h>
-+#include <linux/init.h>
-+#include <linux/pci.h>
-+#include <linux/spinlock.h>
-+#include <linux/time.h>
-+#include <xen/evtchn.h>
-+#include "pcifront.h"
++static irqreturn_t wake_waiting(int irq, void *unused, struct pt_regs *regs)
++{
++ if (unlikely(xenstored_ready == 0)) {
++ xenstored_ready = 1;
++ schedule_work(&probe_work);
++ }
+
-+static int verbose_request = 0;
-+module_param(verbose_request, int, 0644);
++ wake_up(&xb_waitq);
++ return IRQ_HANDLED;
++}
+
-+#ifdef __ia64__
-+static void pcifront_init_sd(struct pcifront_sd *sd,
-+ unsigned int domain, unsigned int bus,
-+ struct pcifront_device *pdev)
++static int check_indexes(XENSTORE_RING_IDX cons, XENSTORE_RING_IDX prod)
+{
-+ int err, i, j, k, len, root_num, res_count;
-+ struct acpi_resource res;
-+ unsigned int d, b, byte;
-+ unsigned long magic;
-+ char str[64], tmp[3];
-+ unsigned char *buf, *bufp;
-+ u8 *ptr;
++ return ((prod - cons) <= XENSTORE_RING_SIZE);
++}
+
-+ memset(sd, 0, sizeof(*sd));
++static void *get_output_chunk(XENSTORE_RING_IDX cons,
++ XENSTORE_RING_IDX prod,
++ char *buf, uint32_t *len)
++{
++ *len = XENSTORE_RING_SIZE - MASK_XENSTORE_IDX(prod);
++ if ((XENSTORE_RING_SIZE - (prod - cons)) < *len)
++ *len = XENSTORE_RING_SIZE - (prod - cons);
++ return buf + MASK_XENSTORE_IDX(prod);
++}
+
-+ sd->segment = domain;
-+ sd->node = -1; /* Revisit for NUMA */
-+ sd->platform_data = pdev;
++static const void *get_input_chunk(XENSTORE_RING_IDX cons,
++ XENSTORE_RING_IDX prod,
++ const char *buf, uint32_t *len)
++{
++ *len = XENSTORE_RING_SIZE - MASK_XENSTORE_IDX(cons);
++ if ((prod - cons) < *len)
++ *len = prod - cons;
++ return buf + MASK_XENSTORE_IDX(cons);
++}
+
-+ /* Look for resources for this controller in xenbus. */
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, "root_num",
-+ "%d", &root_num);
-+ if (err != 1)
-+ return;
++int xb_write(const void *data, unsigned len)
++{
++ struct xenstore_domain_interface *intf = xen_store_interface;
++ XENSTORE_RING_IDX cons, prod;
++ int rc;
+
-+ for (i = 0; i < root_num; i++) {
-+ len = snprintf(str, sizeof(str), "root-%d", i);
-+ if (unlikely(len >= (sizeof(str) - 1)))
-+ return;
++ while (len != 0) {
++ void *dst;
++ unsigned int avail;
+
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
-+ str, "%x:%x", &d, &b);
-+ if (err != 2)
-+ return;
++ rc = wait_event_interruptible(
++ xb_waitq,
++ (intf->req_prod - intf->req_cons) !=
++ XENSTORE_RING_SIZE);
++ if (rc < 0)
++ return rc;
+
-+ if (d == domain && b == bus)
-+ break;
-+ }
++ /* Read indexes, then verify. */
++ cons = intf->req_cons;
++ prod = intf->req_prod;
++ if (!check_indexes(cons, prod)) {
++ intf->req_cons = intf->req_prod = 0;
++ return -EIO;
++ }
+
-+ if (i == root_num)
-+ return;
++ dst = get_output_chunk(cons, prod, intf->req, &avail);
++ if (avail == 0)
++ continue;
++ if (avail > len)
++ avail = len;
+
-+ len = snprintf(str, sizeof(str), "root-resource-magic");
++ /* Must write data /after/ reading the consumer index. */
++ mb();
+
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
-+ str, "%lx", &magic);
++ memcpy(dst, data, avail);
++ data += avail;
++ len -= avail;
+
-+ if (err != 1)
-+ return; /* No resources, nothing to do */
++ /* Other side must not see new producer until data is there. */
++ wmb();
++ intf->req_prod += avail;
+
-+ if (magic != (sizeof(res) * 2) + 1) {
-+ printk(KERN_WARNING "pcifront: resource magic mismatch\n");
-+ return;
++ /* Implies mb(): other side will see the updated producer. */
++ notify_remote_via_evtchn(xen_store_evtchn);
+ }
+
-+ len = snprintf(str, sizeof(str), "root-%d-resources", i);
-+ if (unlikely(len >= (sizeof(str) - 1)))
-+ return;
-+
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
-+ str, "%d", &res_count);
-+
-+ if (err != 1)
-+ return; /* No resources, nothing to do */
-+
-+ sd->window = kzalloc(sizeof(*sd->window) * res_count, GFP_KERNEL);
-+ if (!sd->window)
-+ return;
-+
-+ /* magic is also the size of the byte stream in xenbus */
-+ buf = kmalloc(magic, GFP_KERNEL);
-+ if (!buf) {
-+ kfree(sd->window);
-+ sd->window = NULL;
-+ return;
-+ }
++ return 0;
++}
+
-+ /* Read the resources out of xenbus */
-+ for (j = 0; j < res_count; j++) {
-+ memset(&res, 0, sizeof(res));
-+ memset(buf, 0, magic);
++int xb_data_to_read(void)
++{
++ struct xenstore_domain_interface *intf = xen_store_interface;
++ return (intf->rsp_cons != intf->rsp_prod);
++}
+
-+ len = snprintf(str, sizeof(str), "root-%d-resource-%d", i, j);
-+ if (unlikely(len >= (sizeof(str) - 1)))
-+ return;
++int xb_wait_for_data_to_read(void)
++{
++ return wait_event_interruptible(xb_waitq, xb_data_to_read());
++}
+
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
-+ "%s", buf);
-+ if (err != 1) {
-+ printk(KERN_WARNING "pcifront: error reading "
-+ "resource %d on bus %04x:%02x\n",
-+ j, domain, bus);
-+ continue;
-+ }
++int xb_read(void *data, unsigned len)
++{
++ struct xenstore_domain_interface *intf = xen_store_interface;
++ XENSTORE_RING_IDX cons, prod;
++ int rc;
+
-+ bufp = buf;
-+ ptr = (u8 *)&res;
-+ memset(tmp, 0, sizeof(tmp));
++ while (len != 0) {
++ unsigned int avail;
++ const char *src;
+
-+ /* Copy ASCII byte stream into structure */
-+ for (k = 0; k < magic - 1; k += 2) {
-+ memcpy(tmp, bufp, 2);
-+ bufp += 2;
++ rc = xb_wait_for_data_to_read();
++ if (rc < 0)
++ return rc;
+
-+ sscanf(tmp, "%02x", &byte);
-+ *ptr = byte;
-+ ptr++;
++ /* Read indexes, then verify. */
++ cons = intf->rsp_cons;
++ prod = intf->rsp_prod;
++ if (!check_indexes(cons, prod)) {
++ intf->rsp_cons = intf->rsp_prod = 0;
++ return -EIO;
+ }
+
-+ xen_add_resource(sd, domain, bus, &res);
-+ sd->windows++;
-+ }
-+ kfree(buf);
-+}
-+#endif
++ src = get_input_chunk(cons, prod, intf->rsp, &avail);
++ if (avail == 0)
++ continue;
++ if (avail > len)
++ avail = len;
+
-+static int errno_to_pcibios_err(int errno)
-+{
-+ switch (errno) {
-+ case XEN_PCI_ERR_success:
-+ return PCIBIOS_SUCCESSFUL;
++ /* Must read data /after/ reading the producer index. */
++ rmb();
+
-+ case XEN_PCI_ERR_dev_not_found:
-+ return PCIBIOS_DEVICE_NOT_FOUND;
++ memcpy(data, src, avail);
++ data += avail;
++ len -= avail;
+
-+ case XEN_PCI_ERR_invalid_offset:
-+ case XEN_PCI_ERR_op_failed:
-+ return PCIBIOS_BAD_REGISTER_NUMBER;
++ /* Other side must not see free space until we've copied out */
++ mb();
++ intf->rsp_cons += avail;
+
-+ case XEN_PCI_ERR_not_implemented:
-+ return PCIBIOS_FUNC_NOT_SUPPORTED;
++ pr_debug("Finished read of %i bytes (%i to go)\n", avail, len);
+
-+ case XEN_PCI_ERR_access_denied:
-+ return PCIBIOS_SET_FAILED;
++ /* Implies mb(): other side will see the updated consumer. */
++ notify_remote_via_evtchn(xen_store_evtchn);
+ }
-+ return errno;
++
++ return 0;
+}
+
-+static int do_pci_op(struct pcifront_device *pdev, struct xen_pci_op *op)
++/* Set up interrupt handler off store event channel. */
++int xb_init_comms(void)
+{
-+ int err = 0;
-+ struct xen_pci_op *active_op = &pdev->sh_info->op;
-+ unsigned long irq_flags;
-+ evtchn_port_t port = pdev->evtchn;
-+ s64 ns, ns_timeout;
-+ struct timeval tv;
-+
-+ spin_lock_irqsave(&pdev->sh_info_lock, irq_flags);
-+
-+ memcpy(active_op, op, sizeof(struct xen_pci_op));
++ struct xenstore_domain_interface *intf = xen_store_interface;
++ int err;
+
-+ /* Go */
-+ wmb();
-+ set_bit(_XEN_PCIF_active, (unsigned long *)&pdev->sh_info->flags);
-+ notify_remote_via_evtchn(port);
++ if (intf->req_prod != intf->req_cons)
++ printk(KERN_ERR "XENBUS request ring is not quiescent "
++ "(%08x:%08x)!\n", intf->req_cons, intf->req_prod);
+
-+ /*
-+ * We set a poll timeout of 3 seconds but give up on return after
-+ * 2 seconds. It is better to time out too late rather than too early
-+ * (in the latter case we end up continually re-executing poll() with a
-+ * timeout in the past). 1s difference gives plenty of slack for error.
-+ */
-+ do_gettimeofday(&tv);
-+ ns_timeout = timeval_to_ns(&tv) + 2 * (s64)NSEC_PER_SEC;
++ if (intf->rsp_prod != intf->rsp_cons) {
++ printk(KERN_WARNING "XENBUS response ring is not quiescent "
++ "(%08x:%08x): fixing up\n",
++ intf->rsp_cons, intf->rsp_prod);
++ intf->rsp_cons = intf->rsp_prod;
++ }
+
-+ clear_evtchn(port);
++ if (xenbus_irq)
++ unbind_from_irqhandler(xenbus_irq, &xb_waitq);
+
-+ while (test_bit(_XEN_PCIF_active,
-+ (unsigned long *)&pdev->sh_info->flags)) {
-+ if (HYPERVISOR_poll(&port, 1, jiffies + 3*HZ))
-+ BUG();
-+ clear_evtchn(port);
-+ do_gettimeofday(&tv);
-+ ns = timeval_to_ns(&tv);
-+ if (ns > ns_timeout) {
-+ dev_err(&pdev->xdev->dev,
-+ "pciback not responding!!!\n");
-+ clear_bit(_XEN_PCIF_active,
-+ (unsigned long *)&pdev->sh_info->flags);
-+ err = XEN_PCI_ERR_dev_not_found;
-+ goto out;
-+ }
++ err = bind_caller_port_to_irqhandler(
++ xen_store_evtchn, wake_waiting,
++ 0, "xenbus", &xb_waitq);
++ if (err <= 0) {
++ printk(KERN_ERR "XENBUS request irq failed %i\n", err);
++ return err;
+ }
+
-+ memcpy(op, active_op, sizeof(struct xen_pci_op));
++ xenbus_irq = err;
+
-+ err = op->err;
-+ out:
-+ spin_unlock_irqrestore(&pdev->sh_info_lock, irq_flags);
-+ return err;
++ return 0;
+}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/xenbus/xenbus_comms.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/xenbus/xenbus_comms.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,46 @@
++/*
++ * Private include for xenbus communications.
++ *
++ * Copyright (C) 2005 Rusty Russell, IBM Corporation
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+/* Access to this function is spinlocked in drivers/pci/access.c */
-+static int pcifront_bus_read(struct pci_bus *bus, unsigned int devfn,
-+ int where, int size, u32 * val)
-+{
-+ int err = 0;
-+ struct xen_pci_op op = {
-+ .cmd = XEN_PCI_OP_conf_read,
-+ .domain = pci_domain_nr(bus),
-+ .bus = bus->number,
-+ .devfn = devfn,
-+ .offset = where,
-+ .size = size,
-+ };
-+ struct pcifront_sd *sd = bus->sysdata;
-+ struct pcifront_device *pdev = pcifront_get_pdev(sd);
++#ifndef _XENBUS_COMMS_H
++#define _XENBUS_COMMS_H
+
-+ if (verbose_request)
-+ dev_info(&pdev->xdev->dev,
-+ "read dev=%04x:%02x:%02x.%01x - offset %x size %d\n",
-+ pci_domain_nr(bus), bus->number, PCI_SLOT(devfn),
-+ PCI_FUNC(devfn), where, size);
++int xs_init(void);
++int xb_init_comms(void);
+
-+ err = do_pci_op(pdev, &op);
++/* Low level routines. */
++int xb_write(const void *data, unsigned len);
++int xb_read(void *data, unsigned len);
++int xb_data_to_read(void);
++int xb_wait_for_data_to_read(void);
++int xs_input_avail(void);
++extern struct xenstore_domain_interface *xen_store_interface;
++extern int xen_store_evtchn;
+
-+ if (likely(!err)) {
-+ if (verbose_request)
-+ dev_info(&pdev->xdev->dev, "read got back value %x\n",
-+ op.value);
++#endif /* _XENBUS_COMMS_H */
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/xenbus/xenbus_dev.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/xenbus/xenbus_dev.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,408 @@
++/*
++ * xenbus_dev.c
++ *
++ * Driver giving user-space access to the kernel's xenbus connection
++ * to xenstore.
++ *
++ * Copyright (c) 2005, Christian Limpach
++ * Copyright (c) 2005, Rusty Russell, IBM Corporation
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ *val = op.value;
-+ } else if (err == -ENODEV) {
-+ /* No device here, pretend that it just returned 0 */
-+ err = 0;
-+ *val = 0;
-+ }
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/uio.h>
++#include <linux/notifier.h>
++#include <linux/wait.h>
++#include <linux/fs.h>
++#include <linux/poll.h>
++#include <linux/mutex.h>
+
-+ return errno_to_pcibios_err(err);
-+}
++#include "xenbus_comms.h"
+
-+/* Access to this function is spinlocked in drivers/pci/access.c */
-+static int pcifront_bus_write(struct pci_bus *bus, unsigned int devfn,
-+ int where, int size, u32 val)
-+{
-+ struct xen_pci_op op = {
-+ .cmd = XEN_PCI_OP_conf_write,
-+ .domain = pci_domain_nr(bus),
-+ .bus = bus->number,
-+ .devfn = devfn,
-+ .offset = where,
-+ .size = size,
-+ .value = val,
-+ };
-+ struct pcifront_sd *sd = bus->sysdata;
-+ struct pcifront_device *pdev = pcifront_get_pdev(sd);
++#include <asm/uaccess.h>
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
++#include <xen/xen_proc.h>
++#include <asm/hypervisor.h>
+
-+ if (verbose_request)
-+ dev_info(&pdev->xdev->dev,
-+ "write dev=%04x:%02x:%02x.%01x - "
-+ "offset %x size %d val %x\n",
-+ pci_domain_nr(bus), bus->number,
-+ PCI_SLOT(devfn), PCI_FUNC(devfn), where, size, val);
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
+
-+ return errno_to_pcibios_err(do_pci_op(pdev, &op));
-+}
++struct xenbus_dev_transaction {
++ struct list_head list;
++ struct xenbus_transaction handle;
++};
+
-+struct pci_ops pcifront_bus_ops = {
-+ .read = pcifront_bus_read,
-+ .write = pcifront_bus_write,
++struct read_buffer {
++ struct list_head list;
++ unsigned int cons;
++ unsigned int len;
++ char msg[];
+};
+
-+/* Claim resources for the PCI frontend as-is, backend won't allow changes */
-+static void pcifront_claim_resource(struct pci_dev *dev, void *data)
-+{
-+ struct pcifront_device *pdev = data;
-+ int i;
-+ struct resource *r;
++struct xenbus_dev_data {
++ /* In-progress transaction. */
++ struct list_head transactions;
+
-+ for (i = 0; i < PCI_NUM_RESOURCES; i++) {
-+ r = &dev->resource[i];
++ /* Active watches. */
++ struct list_head watches;
+
-+ if (!r->parent && r->start && r->flags) {
-+ dev_dbg(&pdev->xdev->dev, "claiming resource %s/%d\n",
-+ pci_name(dev), i);
-+ pci_claim_resource(dev, i);
-+ }
-+ }
-+}
++ /* Partial request. */
++ unsigned int len;
++ union {
++ struct xsd_sockmsg msg;
++ char buffer[PAGE_SIZE];
++ } u;
+
-+int pcifront_scan_root(struct pcifront_device *pdev,
-+ unsigned int domain, unsigned int bus)
-+{
-+ struct pci_bus *b;
-+ struct pcifront_sd *sd = NULL;
-+ struct pci_bus_entry *bus_entry = NULL;
-+ int err = 0;
++ /* Response queue. */
++ struct list_head read_buffers;
++ wait_queue_head_t read_waitq;
+
-+#ifndef CONFIG_PCI_DOMAINS
-+ if (domain != 0) {
-+ dev_err(&pdev->xdev->dev,
-+ "PCI Root in non-zero PCI Domain! domain=%d\n", domain);
-+ dev_err(&pdev->xdev->dev,
-+ "Please compile with CONFIG_PCI_DOMAINS\n");
-+ err = -EINVAL;
-+ goto err_out;
-+ }
-+#endif
++ struct mutex reply_mutex;
++};
+
-+ dev_info(&pdev->xdev->dev, "Creating PCI Frontend Bus %04x:%02x\n",
-+ domain, bus);
++static struct proc_dir_entry *xenbus_dev_intf;
+
-+ bus_entry = kmalloc(sizeof(*bus_entry), GFP_KERNEL);
-+ sd = kmalloc(sizeof(*sd), GFP_KERNEL);
-+ if (!bus_entry || !sd) {
-+ err = -ENOMEM;
-+ goto err_out;
++static ssize_t xenbus_dev_read(struct file *filp,
++ char __user *ubuf,
++ size_t len, loff_t *ppos)
++{
++ struct xenbus_dev_data *u = filp->private_data;
++ struct read_buffer *rb;
++ int i, ret;
++
++ mutex_lock(&u->reply_mutex);
++ while (list_empty(&u->read_buffers)) {
++ mutex_unlock(&u->reply_mutex);
++ ret = wait_event_interruptible(u->read_waitq,
++ !list_empty(&u->read_buffers));
++ if (ret)
++ return ret;
++ mutex_lock(&u->reply_mutex);
+ }
-+ pcifront_init_sd(sd, domain, bus, pdev);
+
-+ b = pci_scan_bus_parented(&pdev->xdev->dev, bus,
-+ &pcifront_bus_ops, sd);
-+ if (!b) {
-+ dev_err(&pdev->xdev->dev,
-+ "Error creating PCI Frontend Bus!\n");
-+ err = -ENOMEM;
-+ goto err_out;
++ rb = list_entry(u->read_buffers.next, struct read_buffer, list);
++ for (i = 0; i < len;) {
++ put_user(rb->msg[rb->cons], ubuf + i);
++ i++;
++ rb->cons++;
++ if (rb->cons == rb->len) {
++ list_del(&rb->list);
++ kfree(rb);
++ if (list_empty(&u->read_buffers))
++ break;
++ rb = list_entry(u->read_buffers.next,
++ struct read_buffer, list);
++ }
+ }
++ mutex_unlock(&u->reply_mutex);
+
-+ pcifront_setup_root_resources(b, sd);
-+ bus_entry->bus = b;
++ return i;
++}
+
-+ list_add(&bus_entry->list, &pdev->root_buses);
++static void queue_reply(struct xenbus_dev_data *u,
++ char *data, unsigned int len)
++{
++ struct read_buffer *rb;
+
-+ /* Claim resources before going "live" with our devices */
-+ pci_walk_bus(b, pcifront_claim_resource, pdev);
++ if (len == 0)
++ return;
+
-+ pci_bus_add_devices(b);
++ rb = kmalloc(sizeof(*rb) + len, GFP_KERNEL);
++ BUG_ON(rb == NULL);
+
-+ return 0;
++ rb->cons = 0;
++ rb->len = len;
+
-+ err_out:
-+ kfree(bus_entry);
-+ kfree(sd);
++ memcpy(rb->msg, data, len);
+
-+ return err;
++ list_add_tail(&rb->list, &u->read_buffers);
++
++ wake_up(&u->read_waitq);
+}
+
-+static void free_root_bus_devs(struct pci_bus *bus)
++struct watch_adapter
+{
-+ struct pci_dev *dev;
++ struct list_head list;
++ struct xenbus_watch watch;
++ struct xenbus_dev_data *dev_data;
++ char *token;
++};
+
-+ while (!list_empty(&bus->devices)) {
-+ dev = container_of(bus->devices.next, struct pci_dev,
-+ bus_list);
-+ dev_dbg(&dev->dev, "removing device\n");
-+ pci_remove_bus_device(dev);
-+ }
++static void free_watch_adapter (struct watch_adapter *watch)
++{
++ kfree(watch->watch.node);
++ kfree(watch->token);
++ kfree(watch);
+}
+
-+void pcifront_free_roots(struct pcifront_device *pdev)
++static void watch_fired(struct xenbus_watch *watch,
++ const char **vec,
++ unsigned int len)
+{
-+ struct pci_bus_entry *bus_entry, *t;
-+
-+ dev_dbg(&pdev->xdev->dev, "cleaning up root buses\n");
-+
-+ list_for_each_entry_safe(bus_entry, t, &pdev->root_buses, list) {
-+ list_del(&bus_entry->list);
-+
-+ free_root_bus_devs(bus_entry->bus);
++ struct watch_adapter *adap =
++ container_of(watch, struct watch_adapter, watch);
++ struct xsd_sockmsg hdr;
++ const char *path, *token;
++ int path_len, tok_len, body_len, data_len = 0;
++
++ path = vec[XS_WATCH_PATH];
++ token = adap->token;
+
-+ kfree(bus_entry->bus->sysdata);
++ path_len = strlen(path) + 1;
++ tok_len = strlen(token) + 1;
++ if (len > 2)
++ data_len = vec[len] - vec[2] + 1;
++ body_len = path_len + tok_len + data_len;
+
-+ device_unregister(bus_entry->bus->bridge);
-+ pci_remove_bus(bus_entry->bus);
++ hdr.type = XS_WATCH_EVENT;
++ hdr.len = body_len;
+
-+ kfree(bus_entry);
-+ }
++ mutex_lock(&adap->dev_data->reply_mutex);
++ queue_reply(adap->dev_data, (char *)&hdr, sizeof(hdr));
++ queue_reply(adap->dev_data, (char *)path, path_len);
++ queue_reply(adap->dev_data, (char *)token, tok_len);
++ if (len > 2)
++ queue_reply(adap->dev_data, (char *)vec[2], data_len);
++ mutex_unlock(&adap->dev_data->reply_mutex);
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/pcifront/xenbus.c tmp-linux-2.6-xen.patch/drivers/xen/pcifront/xenbus.c
---- pristine-linux-2.6.18.2/drivers/xen/pcifront/xenbus.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/pcifront/xenbus.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,296 @@
-+/*
-+ * PCI Frontend Xenbus Setup - handles setup with backend (imports page/evtchn)
-+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
-+ */
-+#include <linux/module.h>
-+#include <linux/init.h>
-+#include <linux/mm.h>
-+#include <xen/xenbus.h>
-+#include <xen/gnttab.h>
-+#include "pcifront.h"
+
-+#define INVALID_GRANT_REF (0)
-+#define INVALID_EVTCHN (-1)
++static LIST_HEAD(watch_list);
+
-+static struct pcifront_device *alloc_pdev(struct xenbus_device *xdev)
++static ssize_t xenbus_dev_write(struct file *filp,
++ const char __user *ubuf,
++ size_t len, loff_t *ppos)
+{
-+ struct pcifront_device *pdev;
++ struct xenbus_dev_data *u = filp->private_data;
++ struct xenbus_dev_transaction *trans = NULL;
++ uint32_t msg_type;
++ void *reply;
++ char *path, *token;
++ struct watch_adapter *watch, *tmp_watch;
++ int err, rc = len;
+
-+ pdev = kmalloc(sizeof(struct pcifront_device), GFP_KERNEL);
-+ if (pdev == NULL)
++ if ((len + u->len) > sizeof(u->u.buffer)) {
++ rc = -EINVAL;
+ goto out;
++ }
+
-+ pdev->sh_info =
-+ (struct xen_pci_sharedinfo *)__get_free_page(GFP_KERNEL);
-+ if (pdev->sh_info == NULL) {
-+ kfree(pdev);
-+ pdev = NULL;
++ if (copy_from_user(u->u.buffer + u->len, ubuf, len) != 0) {
++ rc = -EFAULT;
+ goto out;
+ }
-+ pdev->sh_info->flags = 0;
+
-+ xdev->dev.driver_data = pdev;
-+ pdev->xdev = xdev;
++ u->len += len;
++ if ((u->len < sizeof(u->u.msg)) ||
++ (u->len < (sizeof(u->u.msg) + u->u.msg.len)))
++ return rc;
++
++ msg_type = u->u.msg.type;
++
++ switch (msg_type) {
++ case XS_TRANSACTION_START:
++ case XS_TRANSACTION_END:
++ case XS_DIRECTORY:
++ case XS_READ:
++ case XS_GET_PERMS:
++ case XS_RELEASE:
++ case XS_GET_DOMAIN_PATH:
++ case XS_WRITE:
++ case XS_MKDIR:
++ case XS_RM:
++ case XS_SET_PERMS:
++ if (msg_type == XS_TRANSACTION_START) {
++ trans = kmalloc(sizeof(*trans), GFP_KERNEL);
++ if (!trans) {
++ rc = -ENOMEM;
++ goto out;
++ }
++ }
++
++ reply = xenbus_dev_request_and_reply(&u->u.msg);
++ if (IS_ERR(reply)) {
++ kfree(trans);
++ rc = PTR_ERR(reply);
++ goto out;
++ }
++
++ if (msg_type == XS_TRANSACTION_START) {
++ trans->handle.id = simple_strtoul(reply, NULL, 0);
++ list_add(&trans->list, &u->transactions);
++ } else if (msg_type == XS_TRANSACTION_END) {
++ list_for_each_entry(trans, &u->transactions, list)
++ if (trans->handle.id == u->u.msg.tx_id)
++ break;
++ BUG_ON(&trans->list == &u->transactions);
++ list_del(&trans->list);
++ kfree(trans);
++ }
++ mutex_lock(&u->reply_mutex);
++ queue_reply(u, (char *)&u->u.msg, sizeof(u->u.msg));
++ queue_reply(u, (char *)reply, u->u.msg.len);
++ mutex_unlock(&u->reply_mutex);
++ kfree(reply);
++ break;
++
++ case XS_WATCH:
++ case XS_UNWATCH: {
++ static const char *XS_RESP = "OK";
++ struct xsd_sockmsg hdr;
++
++ path = u->u.buffer + sizeof(u->u.msg);
++ token = memchr(path, 0, u->u.msg.len);
++ if (token == NULL) {
++ rc = -EILSEQ;
++ goto out;
++ }
++ token++;
++
++ if (msg_type == XS_WATCH) {
++ watch = kzalloc(sizeof(*watch), GFP_KERNEL);
++ watch->watch.node = kmalloc(strlen(path)+1,
++ GFP_KERNEL);
++ strcpy((char *)watch->watch.node, path);
++ watch->watch.callback = watch_fired;
++ watch->token = kmalloc(strlen(token)+1, GFP_KERNEL);
++ strcpy(watch->token, token);
++ watch->dev_data = u;
+
-+ INIT_LIST_HEAD(&pdev->root_buses);
++ err = register_xenbus_watch(&watch->watch);
++ if (err) {
++ free_watch_adapter(watch);
++ rc = err;
++ goto out;
++ }
++
++ list_add(&watch->list, &u->watches);
++ } else {
++ list_for_each_entry_safe(watch, tmp_watch,
++ &u->watches, list) {
++ if (!strcmp(watch->token, token) &&
++ !strcmp(watch->watch.node, path))
++ {
++ unregister_xenbus_watch(&watch->watch);
++ list_del(&watch->list);
++ free_watch_adapter(watch);
++ break;
++ }
++ }
++ }
+
-+ spin_lock_init(&pdev->dev_lock);
-+ spin_lock_init(&pdev->sh_info_lock);
++ hdr.type = msg_type;
++ hdr.len = strlen(XS_RESP) + 1;
++ mutex_lock(&u->reply_mutex);
++ queue_reply(u, (char *)&hdr, sizeof(hdr));
++ queue_reply(u, (char *)XS_RESP, hdr.len);
++ mutex_unlock(&u->reply_mutex);
++ break;
++ }
+
-+ pdev->evtchn = INVALID_EVTCHN;
-+ pdev->gnt_ref = INVALID_GRANT_REF;
++ default:
++ rc = -EINVAL;
++ break;
++ }
+
-+ dev_dbg(&xdev->dev, "Allocated pdev @ 0x%p pdev->sh_info @ 0x%p\n",
-+ pdev, pdev->sh_info);
-+ out:
-+ return pdev;
++ out:
++ u->len = 0;
++ return rc;
+}
+
-+static void free_pdev(struct pcifront_device *pdev)
++static int xenbus_dev_open(struct inode *inode, struct file *filp)
+{
-+ dev_dbg(&pdev->xdev->dev, "freeing pdev @ 0x%p\n", pdev);
-+
-+ pcifront_free_roots(pdev);
++ struct xenbus_dev_data *u;
+
-+ if (pdev->evtchn != INVALID_EVTCHN)
-+ xenbus_free_evtchn(pdev->xdev, pdev->evtchn);
++ if (xen_store_evtchn == 0)
++ return -ENOENT;
+
-+ if (pdev->gnt_ref != INVALID_GRANT_REF)
-+ gnttab_end_foreign_access(pdev->gnt_ref, 0,
-+ (unsigned long)pdev->sh_info);
++ nonseekable_open(inode, filp);
+
-+ pdev->xdev->dev.driver_data = NULL;
++ u = kzalloc(sizeof(*u), GFP_KERNEL);
++ if (u == NULL)
++ return -ENOMEM;
+
-+ kfree(pdev);
-+}
++ INIT_LIST_HEAD(&u->transactions);
++ INIT_LIST_HEAD(&u->watches);
++ INIT_LIST_HEAD(&u->read_buffers);
++ init_waitqueue_head(&u->read_waitq);
+
-+static int pcifront_publish_info(struct pcifront_device *pdev)
-+{
-+ int err = 0;
-+ struct xenbus_transaction trans;
++ mutex_init(&u->reply_mutex);
+
-+ err = xenbus_grant_ring(pdev->xdev, virt_to_mfn(pdev->sh_info));
-+ if (err < 0)
-+ goto out;
++ filp->private_data = u;
+
-+ pdev->gnt_ref = err;
++ return 0;
++}
+
-+ err = xenbus_alloc_evtchn(pdev->xdev, &pdev->evtchn);
-+ if (err)
-+ goto out;
++static int xenbus_dev_release(struct inode *inode, struct file *filp)
++{
++ struct xenbus_dev_data *u = filp->private_data;
++ struct xenbus_dev_transaction *trans, *tmp;
++ struct watch_adapter *watch, *tmp_watch;
+
-+ do_publish:
-+ err = xenbus_transaction_start(&trans);
-+ if (err) {
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error writing configuration for backend "
-+ "(start transaction)");
-+ goto out;
++ list_for_each_entry_safe(trans, tmp, &u->transactions, list) {
++ xenbus_transaction_end(trans->handle, 1);
++ list_del(&trans->list);
++ kfree(trans);
+ }
+
-+ err = xenbus_printf(trans, pdev->xdev->nodename,
-+ "pci-op-ref", "%u", pdev->gnt_ref);
-+ if (!err)
-+ err = xenbus_printf(trans, pdev->xdev->nodename,
-+ "event-channel", "%u", pdev->evtchn);
-+ if (!err)
-+ err = xenbus_printf(trans, pdev->xdev->nodename,
-+ "magic", XEN_PCI_MAGIC);
-+
-+ if (err) {
-+ xenbus_transaction_end(trans, 1);
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error writing configuration for backend");
-+ goto out;
-+ } else {
-+ err = xenbus_transaction_end(trans, 0);
-+ if (err == -EAGAIN)
-+ goto do_publish;
-+ else if (err) {
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error completing transaction "
-+ "for backend");
-+ goto out;
-+ }
++ list_for_each_entry_safe(watch, tmp_watch, &u->watches, list) {
++ unregister_xenbus_watch(&watch->watch);
++ list_del(&watch->list);
++ free_watch_adapter(watch);
+ }
+
-+ xenbus_switch_state(pdev->xdev, XenbusStateInitialised);
-+
-+ dev_dbg(&pdev->xdev->dev, "publishing successful!\n");
++ kfree(u);
+
-+ out:
-+ return err;
++ return 0;
+}
+
-+static int pcifront_try_connect(struct pcifront_device *pdev)
++static unsigned int xenbus_dev_poll(struct file *file, poll_table *wait)
+{
-+ int err = -EFAULT;
-+ int i, num_roots, len;
-+ char str[64];
-+ unsigned int domain, bus;
++ struct xenbus_dev_data *u = file->private_data;
+
-+ spin_lock(&pdev->dev_lock);
++ poll_wait(file, &u->read_waitq, wait);
++ if (!list_empty(&u->read_buffers))
++ return POLLIN | POLLRDNORM;
++ return 0;
++}
+
-+ /* Only connect once */
-+ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
-+ XenbusStateInitialised)
-+ goto out;
++static const struct file_operations xenbus_dev_file_ops = {
++ .read = xenbus_dev_read,
++ .write = xenbus_dev_write,
++ .open = xenbus_dev_open,
++ .release = xenbus_dev_release,
++ .poll = xenbus_dev_poll,
++};
+
-+ err = pcifront_connect(pdev);
-+ if (err) {
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error connecting PCI Frontend");
-+ goto out;
-+ }
++int xenbus_dev_init(void)
++{
++ xenbus_dev_intf = create_xen_proc_entry("xenbus", 0400);
++ if (xenbus_dev_intf)
++ xenbus_dev_intf->proc_fops = &xenbus_dev_file_ops;
+
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
-+ "root_num", "%d", &num_roots);
-+ if (err == -ENOENT) {
-+ xenbus_dev_error(pdev->xdev, err,
-+ "No PCI Roots found, trying 0000:00");
-+ err = pcifront_scan_root(pdev, 0, 0);
-+ num_roots = 0;
-+ } else if (err != 1) {
-+ if (err == 0)
-+ err = -EINVAL;
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error reading number of PCI roots");
-+ goto out;
-+ }
++ return 0;
++}
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/xenbus/xenbus_probe.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/xenbus/xenbus_probe.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1153 @@
++/******************************************************************************
++ * Talks to Xen Store to figure out what devices we have.
++ *
++ * Copyright (C) 2005 Rusty Russell, IBM Corporation
++ * Copyright (C) 2005 Mike Wray, Hewlett-Packard
++ * Copyright (C) 2005, 2006 XenSource Ltd
++ * Copyright (C) 2007 Solarflare Communications, Inc.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ for (i = 0; i < num_roots; i++) {
-+ len = snprintf(str, sizeof(str), "root-%d", i);
-+ if (unlikely(len >= (sizeof(str) - 1))) {
-+ err = -ENOMEM;
-+ goto out;
-+ }
++#define DPRINTK(fmt, args...) \
++ pr_debug("xenbus_probe (%s:%d) " fmt ".\n", \
++ __FUNCTION__, __LINE__, ##args)
+
-+ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
-+ "%x:%x", &domain, &bus);
-+ if (err != 2) {
-+ if (err >= 0)
-+ err = -EINVAL;
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error reading PCI root %d", i);
-+ goto out;
-+ }
++#include <linux/kernel.h>
++#include <linux/err.h>
++#include <linux/string.h>
++#include <linux/ctype.h>
++#include <linux/fcntl.h>
++#include <linux/mm.h>
++#include <linux/notifier.h>
++#include <linux/mutex.h>
++#include <linux/module.h>
+
-+ err = pcifront_scan_root(pdev, domain, bus);
-+ if (err) {
-+ xenbus_dev_fatal(pdev->xdev, err,
-+ "Error scanning PCI root %04x:%02x",
-+ domain, bus);
-+ goto out;
-+ }
-+ }
++#include <asm/io.h>
++#include <asm/page.h>
++#include <asm/maddr.h>
++#include <asm/pgtable.h>
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
++#include <xen/xen_proc.h>
++#include <xen/evtchn.h>
++#include <xen/features.h>
++#ifdef MODULE
++#include <xen/hvm.h>
++#endif
+
-+ err = xenbus_switch_state(pdev->xdev, XenbusStateConnected);
-+ if (err)
-+ goto out;
++#include "xenbus_comms.h"
++#include "xenbus_probe.h"
+
-+ out:
-+ spin_unlock(&pdev->dev_lock);
-+ return err;
-+}
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
+
-+static int pcifront_try_disconnect(struct pcifront_device *pdev)
-+{
-+ int err = 0;
-+ enum xenbus_state prev_state;
++int xen_store_evtchn;
++struct xenstore_domain_interface *xen_store_interface;
++static unsigned long xen_store_mfn;
+
-+ spin_lock(&pdev->dev_lock);
++extern struct mutex xenwatch_mutex;
+
-+ prev_state = xenbus_read_driver_state(pdev->xdev->nodename);
++static BLOCKING_NOTIFIER_HEAD(xenstore_chain);
+
-+ if (prev_state < XenbusStateClosing)
-+ err = xenbus_switch_state(pdev->xdev, XenbusStateClosing);
++static void wait_for_devices(struct xenbus_driver *xendrv);
+
-+ if (!err && prev_state == XenbusStateConnected)
-+ pcifront_disconnect(pdev);
++static int xenbus_probe_frontend(const char *type, const char *name);
+
-+ spin_unlock(&pdev->dev_lock);
++static void xenbus_dev_shutdown(struct device *_dev);
+
-+ return err;
++/* If something in array of ids matches this device, return it. */
++static const struct xenbus_device_id *
++match_device(const struct xenbus_device_id *arr, struct xenbus_device *dev)
++{
++ for (; *arr->devicetype != '\0'; arr++) {
++ if (!strcmp(arr->devicetype, dev->devicetype))
++ return arr;
++ }
++ return NULL;
+}
+
-+static void pcifront_backend_changed(struct xenbus_device *xdev,
-+ enum xenbus_state be_state)
++int xenbus_match(struct device *_dev, struct device_driver *_drv)
+{
-+ struct pcifront_device *pdev = xdev->dev.driver_data;
-+
-+ switch (be_state) {
-+ case XenbusStateClosing:
-+ dev_warn(&xdev->dev, "backend going away!\n");
-+ pcifront_try_disconnect(pdev);
-+ break;
++ struct xenbus_driver *drv = to_xenbus_driver(_drv);
+
-+ case XenbusStateUnknown:
-+ case XenbusStateClosed:
-+ dev_warn(&xdev->dev, "backend went away!\n");
-+ pcifront_try_disconnect(pdev);
++ if (!drv->ids)
++ return 0;
+
-+ device_unregister(&pdev->xdev->dev);
-+ break;
++ return match_device(drv->ids, to_xenbus_device(_dev)) != NULL;
++}
+
-+ case XenbusStateConnected:
-+ pcifront_try_connect(pdev);
-+ break;
++/* device/<type>/<id> => <type>-<id> */
++static int frontend_bus_id(char bus_id[BUS_ID_SIZE], const char *nodename)
++{
++ nodename = strchr(nodename, '/');
++ if (!nodename || strlen(nodename + 1) >= BUS_ID_SIZE) {
++ printk(KERN_WARNING "XENBUS: bad frontend %s\n", nodename);
++ return -EINVAL;
++ }
+
-+ default:
-+ break;
++ strlcpy(bus_id, nodename + 1, BUS_ID_SIZE);
++ if (!strchr(bus_id, '/')) {
++ printk(KERN_WARNING "XENBUS: bus_id %s no slash\n", bus_id);
++ return -EINVAL;
+ }
++ *strchr(bus_id, '/') = '-';
++ return 0;
+}
+
-+static int pcifront_xenbus_probe(struct xenbus_device *xdev,
-+ const struct xenbus_device_id *id)
-+{
-+ int err = 0;
-+ struct pcifront_device *pdev = alloc_pdev(xdev);
+
-+ if (pdev == NULL) {
-+ err = -ENOMEM;
-+ xenbus_dev_fatal(xdev, err,
-+ "Error allocating pcifront_device struct");
-+ goto out;
-+ }
++static void free_otherend_details(struct xenbus_device *dev)
++{
++ kfree(dev->otherend);
++ dev->otherend = NULL;
++}
+
-+ err = pcifront_publish_info(pdev);
+
-+ out:
-+ return err;
++static void free_otherend_watch(struct xenbus_device *dev)
++{
++ if (dev->otherend_watch.node) {
++ unregister_xenbus_watch(&dev->otherend_watch);
++ kfree(dev->otherend_watch.node);
++ dev->otherend_watch.node = NULL;
++ }
+}
+
-+static int pcifront_xenbus_remove(struct xenbus_device *xdev)
++
++int read_otherend_details(struct xenbus_device *xendev,
++ char *id_node, char *path_node)
+{
-+ if (xdev->dev.driver_data)
-+ free_pdev(xdev->dev.driver_data);
++ int err = xenbus_gather(XBT_NIL, xendev->nodename,
++ id_node, "%i", &xendev->otherend_id,
++ path_node, NULL, &xendev->otherend,
++ NULL);
++ if (err) {
++ xenbus_dev_fatal(xendev, err,
++ "reading other end details from %s",
++ xendev->nodename);
++ return err;
++ }
++ if (strlen(xendev->otherend) == 0 ||
++ !xenbus_exists(XBT_NIL, xendev->otherend, "")) {
++ xenbus_dev_fatal(xendev, -ENOENT,
++ "unable to read other end from %s. "
++ "missing or inaccessible.",
++ xendev->nodename);
++ free_otherend_details(xendev);
++ return -ENOENT;
++ }
+
+ return 0;
+}
+
-+static struct xenbus_device_id xenpci_ids[] = {
-+ {"pci"},
-+ {{0}},
-+};
-+MODULE_ALIAS("xen:pci");
-+
-+static struct xenbus_driver xenbus_pcifront_driver = {
-+ .name = "pcifront",
-+ .owner = THIS_MODULE,
-+ .ids = xenpci_ids,
-+ .probe = pcifront_xenbus_probe,
-+ .remove = pcifront_xenbus_remove,
-+ .otherend_changed = pcifront_backend_changed,
-+};
+
-+static int __init pcifront_init(void)
++static int read_backend_details(struct xenbus_device *xendev)
+{
-+ if (!is_running_on_xen())
-+ return -ENODEV;
-+
-+ return xenbus_register_frontend(&xenbus_pcifront_driver);
++ return read_otherend_details(xendev, "backend-id", "backend");
+}
+
-+/* Initialize after the Xen PCI Frontend Stub is initialized */
-+subsys_initcall(pcifront_init);
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/privcmd/Makefile tmp-linux-2.6-xen.patch/drivers/xen/privcmd/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/privcmd/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/privcmd/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,2 @@
-+
-+obj-$(CONFIG_XEN_PRIVCMD) := privcmd.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/privcmd/privcmd.c tmp-linux-2.6-xen.patch/drivers/xen/privcmd/privcmd.c
---- pristine-linux-2.6.18.2/drivers/xen/privcmd/privcmd.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/privcmd/privcmd.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,286 @@
-+/******************************************************************************
-+ * privcmd.c
-+ *
-+ * Interface to privileged domain-0 commands.
-+ *
-+ * Copyright (c) 2002-2004, K A Fraser, B Dragovic
-+ */
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
++static int xenbus_uevent_frontend(struct device *dev, char **envp,
++ int num_envp, char *buffer, int buffer_size)
++{
++ struct xenbus_device *xdev;
++ int length = 0, i = 0;
+
-+#include <linux/kernel.h>
-+#include <linux/sched.h>
-+#include <linux/slab.h>
-+#include <linux/string.h>
-+#include <linux/errno.h>
-+#include <linux/mm.h>
-+#include <linux/mman.h>
-+#include <linux/swap.h>
-+#include <linux/smp_lock.h>
-+#include <linux/highmem.h>
-+#include <linux/pagemap.h>
-+#include <linux/seq_file.h>
-+#include <linux/kthread.h>
-+#include <asm/hypervisor.h>
++ if (dev == NULL)
++ return -ENODEV;
++ xdev = to_xenbus_device(dev);
++ if (xdev == NULL)
++ return -ENODEV;
+
-+#include <asm/pgalloc.h>
-+#include <asm/pgtable.h>
-+#include <asm/uaccess.h>
-+#include <asm/tlb.h>
-+#include <asm/hypervisor.h>
-+#include <xen/public/privcmd.h>
-+#include <xen/interface/xen.h>
-+#include <xen/xen_proc.h>
++ /* stuff we want to pass to /sbin/hotplug */
++ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
++ "XENBUS_TYPE=%s", xdev->devicetype);
++ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
++ "XENBUS_PATH=%s", xdev->nodename);
++ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
++ "MODALIAS=xen:%s", xdev->devicetype);
+
-+static struct proc_dir_entry *privcmd_intf;
-+static struct proc_dir_entry *capabilities_intf;
++ return 0;
++}
++#endif
+
-+#ifndef HAVE_ARCH_PRIVCMD_MMAP
-+static int privcmd_enforce_singleshot_mapping(struct vm_area_struct *vma);
++/* Bus type for frontend drivers. */
++static struct xen_bus_type xenbus_frontend = {
++ .root = "device",
++ .levels = 2, /* device/type/<id> */
++ .get_bus_id = frontend_bus_id,
++ .probe = xenbus_probe_frontend,
++ .error = -ENODEV,
++ .bus = {
++ .name = "xen",
++ .match = xenbus_match,
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
++ .probe = xenbus_dev_probe,
++ .remove = xenbus_dev_remove,
++ .shutdown = xenbus_dev_shutdown,
++ .uevent = xenbus_uevent_frontend,
+#endif
++ },
++ .dev = {
++ .bus_id = "xen",
++ },
++};
+
-+static long privcmd_ioctl(struct file *file,
-+ unsigned int cmd, unsigned long data)
++static void otherend_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
+{
-+ int ret = -ENOSYS;
-+ void __user *udata = (void __user *) data;
++ struct xenbus_device *dev =
++ container_of(watch, struct xenbus_device, otherend_watch);
++ struct xenbus_driver *drv = to_xenbus_driver(dev->dev.driver);
++ enum xenbus_state state;
++
++ /* Protect us against watches firing on old details when the otherend
++ details change, say immediately after a resume. */
++ if (!dev->otherend ||
++ strncmp(dev->otherend, vec[XS_WATCH_PATH],
++ strlen(dev->otherend))) {
++ DPRINTK("Ignoring watch at %s", vec[XS_WATCH_PATH]);
++ return;
++ }
+
-+ switch (cmd) {
-+ case IOCTL_PRIVCMD_HYPERCALL: {
-+ privcmd_hypercall_t hypercall;
-+
-+ if (copy_from_user(&hypercall, udata, sizeof(hypercall)))
-+ return -EFAULT;
++ state = xenbus_read_driver_state(dev->otherend);
+
-+#if defined(__i386__)
-+ if (hypercall.op >= (PAGE_SIZE >> 5))
-+ break;
-+ __asm__ __volatile__ (
-+ "pushl %%ebx; pushl %%ecx; pushl %%edx; "
-+ "pushl %%esi; pushl %%edi; "
-+ "movl 8(%%eax),%%ebx ;"
-+ "movl 16(%%eax),%%ecx ;"
-+ "movl 24(%%eax),%%edx ;"
-+ "movl 32(%%eax),%%esi ;"
-+ "movl 40(%%eax),%%edi ;"
-+ "movl (%%eax),%%eax ;"
-+ "shll $5,%%eax ;"
-+ "addl $hypercall_page,%%eax ;"
-+ "call *%%eax ;"
-+ "popl %%edi; popl %%esi; popl %%edx; "
-+ "popl %%ecx; popl %%ebx"
-+ : "=a" (ret) : "0" (&hypercall) : "memory" );
-+#elif defined (__x86_64__)
-+ if (hypercall.op < (PAGE_SIZE >> 5)) {
-+ long ign1, ign2, ign3;
-+ __asm__ __volatile__ (
-+ "movq %8,%%r10; movq %9,%%r8;"
-+ "shll $5,%%eax ;"
-+ "addq $hypercall_page,%%rax ;"
-+ "call *%%rax"
-+ : "=a" (ret), "=D" (ign1),
-+ "=S" (ign2), "=d" (ign3)
-+ : "0" ((unsigned int)hypercall.op),
-+ "1" (hypercall.arg[0]),
-+ "2" (hypercall.arg[1]),
-+ "3" (hypercall.arg[2]),
-+ "g" (hypercall.arg[3]),
-+ "g" (hypercall.arg[4])
-+ : "r8", "r10", "memory" );
-+ }
-+#elif defined (__ia64__)
-+ ret = privcmd_hypercall(&hypercall);
-+#endif
++ DPRINTK("state is %d (%s), %s, %s", state, xenbus_strstate(state),
++ dev->otherend_watch.node, vec[XS_WATCH_PATH]);
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
++ /*
++ * Ignore xenbus transitions during shutdown. This prevents us doing
++ * work that can fail e.g., when the rootfs is gone.
++ */
++ if (system_state > SYSTEM_RUNNING) {
++ struct xen_bus_type *bus = bus;
++ bus = container_of(dev->dev.bus, struct xen_bus_type, bus);
++ /* If we're frontend, drive the state machine to Closed. */
++ /* This should cause the backend to release our resources. */
++ if ((bus == &xenbus_frontend) && (state == XenbusStateClosing))
++ xenbus_frontend_closed(dev);
++ return;
+ }
-+ break;
++#endif
+
-+ case IOCTL_PRIVCMD_MMAP: {
-+ privcmd_mmap_t mmapcmd;
-+ privcmd_mmap_entry_t msg;
-+ privcmd_mmap_entry_t __user *p;
-+ struct mm_struct *mm = current->mm;
-+ struct vm_area_struct *vma;
-+ unsigned long va;
-+ int i, rc;
++ if (drv->otherend_changed)
++ drv->otherend_changed(dev, state);
++}
+
-+ if (!is_initial_xendomain())
-+ return -EPERM;
+
-+ if (copy_from_user(&mmapcmd, udata, sizeof(mmapcmd)))
-+ return -EFAULT;
++static int talk_to_otherend(struct xenbus_device *dev)
++{
++ struct xenbus_driver *drv = to_xenbus_driver(dev->dev.driver);
+
-+ p = mmapcmd.entry;
-+ if (copy_from_user(&msg, p, sizeof(msg)))
-+ return -EFAULT;
++ free_otherend_watch(dev);
++ free_otherend_details(dev);
+
-+ down_read(&mm->mmap_sem);
++ return drv->read_otherend_details(dev);
++}
+
-+ vma = find_vma(mm, msg.va);
-+ rc = -EINVAL;
-+ if (!vma || (msg.va != vma->vm_start) ||
-+ !privcmd_enforce_singleshot_mapping(vma))
-+ goto mmap_out;
+
-+ va = vma->vm_start;
++static int watch_otherend(struct xenbus_device *dev)
++{
++ return xenbus_watch_path2(dev, dev->otherend, "state",
++ &dev->otherend_watch, otherend_changed);
++}
+
-+ for (i = 0; i < mmapcmd.num; i++) {
-+ rc = -EFAULT;
-+ if (copy_from_user(&msg, p, sizeof(msg)))
-+ goto mmap_out;
+
-+ /* Do not allow range to wrap the address space. */
-+ rc = -EINVAL;
-+ if ((msg.npages > (LONG_MAX >> PAGE_SHIFT)) ||
-+ ((unsigned long)(msg.npages << PAGE_SHIFT) >= -va))
-+ goto mmap_out;
++int xenbus_dev_probe(struct device *_dev)
++{
++ struct xenbus_device *dev = to_xenbus_device(_dev);
++ struct xenbus_driver *drv = to_xenbus_driver(_dev->driver);
++ const struct xenbus_device_id *id;
++ int err;
+
-+ /* Range chunks must be contiguous in va space. */
-+ if ((msg.va != va) ||
-+ ((msg.va+(msg.npages<<PAGE_SHIFT)) > vma->vm_end))
-+ goto mmap_out;
++ DPRINTK("%s", dev->nodename);
+
-+ if ((rc = direct_remap_pfn_range(
-+ vma,
-+ msg.va & PAGE_MASK,
-+ msg.mfn,
-+ msg.npages << PAGE_SHIFT,
-+ vma->vm_page_prot,
-+ mmapcmd.dom)) < 0)
-+ goto mmap_out;
++ if (!drv->probe) {
++ err = -ENODEV;
++ goto fail;
++ }
+
-+ p++;
-+ va += msg.npages << PAGE_SHIFT;
-+ }
++ id = match_device(drv->ids, dev);
++ if (!id) {
++ err = -ENODEV;
++ goto fail;
++ }
+
-+ rc = 0;
++ err = talk_to_otherend(dev);
++ if (err) {
++ printk(KERN_WARNING
++ "xenbus_probe: talk_to_otherend on %s failed.\n",
++ dev->nodename);
++ return err;
++ }
+
-+ mmap_out:
-+ up_read(&mm->mmap_sem);
-+ ret = rc;
++ err = drv->probe(dev, id);
++ if (err)
++ goto fail;
++
++ err = watch_otherend(dev);
++ if (err) {
++ printk(KERN_WARNING
++ "xenbus_probe: watch_otherend on %s failed.\n",
++ dev->nodename);
++ return err;
+ }
-+ break;
+
-+ case IOCTL_PRIVCMD_MMAPBATCH: {
-+ privcmd_mmapbatch_t m;
-+ struct mm_struct *mm = current->mm;
-+ struct vm_area_struct *vma;
-+ xen_pfn_t __user *p;
-+ unsigned long addr, mfn, nr_pages;
-+ int i;
++ return 0;
++fail:
++ xenbus_dev_error(dev, err, "xenbus_dev_probe on %s", dev->nodename);
++ xenbus_switch_state(dev, XenbusStateClosed);
++ return -ENODEV;
++}
+
-+ if (!is_initial_xendomain())
-+ return -EPERM;
++int xenbus_dev_remove(struct device *_dev)
++{
++ struct xenbus_device *dev = to_xenbus_device(_dev);
++ struct xenbus_driver *drv = to_xenbus_driver(_dev->driver);
+
-+ if (copy_from_user(&m, udata, sizeof(m)))
-+ return -EFAULT;
++ DPRINTK("%s", dev->nodename);
+
-+ nr_pages = m.num;
-+ if ((m.num <= 0) || (nr_pages > (LONG_MAX >> PAGE_SHIFT)))
-+ return -EINVAL;
++ free_otherend_watch(dev);
++ free_otherend_details(dev);
+
-+ down_read(&mm->mmap_sem);
++ if (drv->remove)
++ drv->remove(dev);
+
-+ vma = find_vma(mm, m.addr);
-+ if (!vma ||
-+ (m.addr != vma->vm_start) ||
-+ ((m.addr + (nr_pages << PAGE_SHIFT)) != vma->vm_end) ||
-+ !privcmd_enforce_singleshot_mapping(vma)) {
-+ up_read(&mm->mmap_sem);
-+ return -EINVAL;
-+ }
++ xenbus_switch_state(dev, XenbusStateClosed);
++ return 0;
++}
+
-+ p = m.arr;
-+ addr = m.addr;
-+ for (i = 0; i < nr_pages; i++, addr += PAGE_SIZE, p++) {
-+ if (get_user(mfn, p)) {
-+ up_read(&mm->mmap_sem);
-+ return -EFAULT;
-+ }
++static void xenbus_dev_shutdown(struct device *_dev)
++{
++ struct xenbus_device *dev = to_xenbus_device(_dev);
++ unsigned long timeout = 5*HZ;
+
-+ ret = direct_remap_pfn_range(vma, addr & PAGE_MASK,
-+ mfn, PAGE_SIZE,
-+ vma->vm_page_prot, m.dom);
-+ if (ret < 0)
-+ put_user(0xF0000000 | mfn, p);
-+ }
++ DPRINTK("%s", dev->nodename);
+
-+ up_read(&mm->mmap_sem);
-+ ret = 0;
-+ }
-+ break;
++ if (is_initial_xendomain())
++ return;
+
-+ default:
-+ ret = -EINVAL;
-+ break;
++ get_device(&dev->dev);
++ if (dev->state != XenbusStateConnected) {
++ printk("%s: %s: %s != Connected, skipping\n", __FUNCTION__,
++ dev->nodename, xenbus_strstate(dev->state));
++ goto out;
+ }
++ xenbus_switch_state(dev, XenbusStateClosing);
++ timeout = wait_for_completion_timeout(&dev->down, timeout);
++ if (!timeout)
++ printk("%s: %s timeout closing device\n", __FUNCTION__, dev->nodename);
++ out:
++ put_device(&dev->dev);
++}
++
++int xenbus_register_driver_common(struct xenbus_driver *drv,
++ struct xen_bus_type *bus)
++{
++ int ret;
++
++ if (bus->error)
++ return bus->error;
++
++ drv->driver.name = drv->name;
++ drv->driver.bus = &bus->bus;
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,10)
++ drv->driver.owner = drv->owner;
++#endif
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,16)
++ drv->driver.probe = xenbus_dev_probe;
++ drv->driver.remove = xenbus_dev_remove;
++ drv->driver.shutdown = xenbus_dev_shutdown;
++#endif
+
++ mutex_lock(&xenwatch_mutex);
++ ret = driver_register(&drv->driver);
++ mutex_unlock(&xenwatch_mutex);
+ return ret;
+}
+
-+#ifndef HAVE_ARCH_PRIVCMD_MMAP
-+static struct page *privcmd_nopage(struct vm_area_struct *vma,
-+ unsigned long address,
-+ int *type)
++int xenbus_register_frontend(struct xenbus_driver *drv)
+{
-+ return NOPAGE_SIGBUS;
-+}
++ int ret;
+
-+static struct vm_operations_struct privcmd_vm_ops = {
-+ .nopage = privcmd_nopage
-+};
++ drv->read_otherend_details = read_backend_details;
+
-+static int privcmd_mmap(struct file * file, struct vm_area_struct * vma)
-+{
-+#ifndef __powerpc__ /* PowerPC has a trick to safely do this. */
-+ /* Unsupported for auto-translate guests. */
-+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ return -ENOSYS;
-+#endif
++ ret = xenbus_register_driver_common(drv, &xenbus_frontend);
++ if (ret)
++ return ret;
+
-+ /* DONTCOPY is essential for Xen as copy_page_range is broken. */
-+ vma->vm_flags |= VM_RESERVED | VM_IO | VM_DONTCOPY;
-+ vma->vm_ops = &privcmd_vm_ops;
-+ vma->vm_private_data = NULL;
++ /* If this driver is loaded as a module wait for devices to attach. */
++ wait_for_devices(drv);
+
+ return 0;
+}
++EXPORT_SYMBOL_GPL(xenbus_register_frontend);
+
-+static int privcmd_enforce_singleshot_mapping(struct vm_area_struct *vma)
++void xenbus_unregister_driver(struct xenbus_driver *drv)
+{
-+ return (xchg(&vma->vm_private_data, (void *)1) == NULL);
++ driver_unregister(&drv->driver);
+}
-+#endif
++EXPORT_SYMBOL_GPL(xenbus_unregister_driver);
+
-+static const struct file_operations privcmd_file_ops = {
-+ .unlocked_ioctl = privcmd_ioctl,
-+ .mmap = privcmd_mmap,
++struct xb_find_info
++{
++ struct xenbus_device *dev;
++ const char *nodename;
+};
+
-+static int capabilities_read(char *page, char **start, off_t off,
-+ int count, int *eof, void *data)
++static int cmp_dev(struct device *dev, void *data)
+{
-+ int len = 0;
-+ *page = 0;
++ struct xenbus_device *xendev = to_xenbus_device(dev);
++ struct xb_find_info *info = data;
+
-+ if (is_initial_xendomain())
-+ len = sprintf( page, "control_d\n" );
++ if (!strcmp(xendev->nodename, info->nodename)) {
++ info->dev = xendev;
++ get_device(dev);
++ return 1;
++ }
++ return 0;
++}
+
-+ *eof = 1;
-+ return len;
++struct xenbus_device *xenbus_device_find(const char *nodename,
++ struct bus_type *bus)
++{
++ struct xb_find_info info = { .dev = NULL, .nodename = nodename };
++
++ bus_for_each_dev(bus, NULL, &info, cmp_dev);
++ return info.dev;
+}
+
-+static int __init privcmd_init(void)
++static int cleanup_dev(struct device *dev, void *data)
+{
-+ if (!is_running_on_xen())
-+ return -ENODEV;
++ struct xenbus_device *xendev = to_xenbus_device(dev);
++ struct xb_find_info *info = data;
++ int len = strlen(info->nodename);
+
-+ privcmd_intf = create_xen_proc_entry("privcmd", 0400);
-+ if (privcmd_intf != NULL)
-+ privcmd_intf->proc_fops = &privcmd_file_ops;
++ DPRINTK("%s", info->nodename);
+
-+ capabilities_intf = create_xen_proc_entry("capabilities", 0400 );
-+ if (capabilities_intf != NULL)
-+ capabilities_intf->read_proc = capabilities_read;
++ /* Match the info->nodename path, or any subdirectory of that path. */
++ if (strncmp(xendev->nodename, info->nodename, len))
++ return 0;
+
-+ return 0;
++ /* If the node name is longer, ensure it really is a subdirectory. */
++ if ((strlen(xendev->nodename) > len) && (xendev->nodename[len] != '/'))
++ return 0;
++
++ info->dev = xendev;
++ get_device(dev);
++ return 1;
+}
+
-+__initcall(privcmd_init);
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/tpmback/common.h tmp-linux-2.6-xen.patch/drivers/xen/tpmback/common.h
---- pristine-linux-2.6.18.2/drivers/xen/tpmback/common.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/tpmback/common.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,85 @@
-+/******************************************************************************
-+ * drivers/xen/tpmback/common.h
-+ */
++static void xenbus_cleanup_devices(const char *path, struct bus_type *bus)
++{
++ struct xb_find_info info = { .nodename = path };
+
-+#ifndef __TPM__BACKEND__COMMON_H__
-+#define __TPM__BACKEND__COMMON_H__
++ do {
++ info.dev = NULL;
++ bus_for_each_dev(bus, NULL, &info, cleanup_dev);
++ if (info.dev) {
++ device_unregister(&info.dev->dev);
++ put_device(&info.dev->dev);
++ }
++ } while (info.dev);
++}
+
-+#include <linux/version.h>
-+#include <linux/module.h>
-+#include <linux/interrupt.h>
-+#include <linux/slab.h>
-+#include <xen/evtchn.h>
-+#include <xen/driver_util.h>
-+#include <xen/interface/grant_table.h>
-+#include <xen/interface/io/tpmif.h>
-+#include <asm/io.h>
-+#include <asm/pgalloc.h>
++static void xenbus_dev_release(struct device *dev)
++{
++ if (dev)
++ kfree(to_xenbus_device(dev));
++}
+
-+#define DPRINTK(_f, _a...) \
-+ pr_debug("(file=%s, line=%d) " _f, \
-+ __FILE__ , __LINE__ , ## _a )
++static ssize_t xendev_show_nodename(struct device *dev,
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,13)
++ struct device_attribute *attr,
++#endif
++ char *buf)
++{
++ return sprintf(buf, "%s\n", to_xenbus_device(dev)->nodename);
++}
++DEVICE_ATTR(nodename, S_IRUSR | S_IRGRP | S_IROTH, xendev_show_nodename, NULL);
+
-+struct backend_info;
++static ssize_t xendev_show_devtype(struct device *dev,
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,13)
++ struct device_attribute *attr,
++#endif
++ char *buf)
++{
++ return sprintf(buf, "%s\n", to_xenbus_device(dev)->devicetype);
++}
++DEVICE_ATTR(devtype, S_IRUSR | S_IRGRP | S_IROTH, xendev_show_devtype, NULL);
+
-+typedef struct tpmif_st {
-+ struct list_head tpmif_list;
-+ /* Unique identifier for this interface. */
-+ domid_t domid;
-+ unsigned int handle;
+
-+ /* Physical parameters of the comms window. */
-+ unsigned int irq;
++int xenbus_probe_node(struct xen_bus_type *bus,
++ const char *type,
++ const char *nodename)
++{
++ int err;
++ struct xenbus_device *xendev;
++ size_t stringlen;
++ char *tmpstring;
+
-+ /* The shared rings and indexes. */
-+ tpmif_tx_interface_t *tx;
-+ struct vm_struct *tx_area;
++ enum xenbus_state state = xenbus_read_driver_state(nodename);
+
-+ /* Miscellaneous private stuff. */
-+ enum { DISCONNECTED, DISCONNECTING, CONNECTED } status;
-+ int active;
++ if (bus->error)
++ return bus->error;
+
-+ struct tpmif_st *hash_next;
-+ struct list_head list; /* scheduling list */
-+ atomic_t refcnt;
++ if (state != XenbusStateInitialising) {
++ /* Device is not new, so ignore it. This can happen if a
++ device is going away after switching to Closed. */
++ return 0;
++ }
+
-+ struct backend_info *bi;
++ stringlen = strlen(nodename) + 1 + strlen(type) + 1;
++ xendev = kzalloc(sizeof(*xendev) + stringlen, GFP_KERNEL);
++ if (!xendev)
++ return -ENOMEM;
+
-+ grant_handle_t shmem_handle;
-+ grant_ref_t shmem_ref;
-+ struct page **mmap_pages;
++ xendev->state = XenbusStateInitialising;
+
-+ char devname[20];
-+} tpmif_t;
++ /* Copy the strings into the extra space. */
+
-+void tpmif_disconnect_complete(tpmif_t * tpmif);
-+tpmif_t *tpmif_find(domid_t domid, struct backend_info *bi);
-+void tpmif_interface_init(void);
-+void tpmif_interface_exit(void);
-+void tpmif_schedule_work(tpmif_t * tpmif);
-+void tpmif_deschedule_work(tpmif_t * tpmif);
-+void tpmif_xenbus_init(void);
-+void tpmif_xenbus_exit(void);
-+int tpmif_map(tpmif_t *tpmif, unsigned long shared_page, unsigned int evtchn);
-+irqreturn_t tpmif_be_int(int irq, void *dev_id, struct pt_regs *regs);
++ tmpstring = (char *)(xendev + 1);
++ strcpy(tmpstring, nodename);
++ xendev->nodename = tmpstring;
+
-+long int tpmback_get_instance(struct backend_info *bi);
++ tmpstring += strlen(tmpstring) + 1;
++ strcpy(tmpstring, type);
++ xendev->devicetype = tmpstring;
++ init_completion(&xendev->down);
+
-+int vtpm_release_packets(tpmif_t * tpmif, int send_msgs);
++ xendev->dev.parent = &bus->dev;
++ xendev->dev.bus = &bus->bus;
++ xendev->dev.release = xenbus_dev_release;
+
++ err = bus->get_bus_id(xendev->dev.bus_id, xendev->nodename);
++ if (err)
++ goto fail;
+
-+#define tpmif_get(_b) (atomic_inc(&(_b)->refcnt))
-+#define tpmif_put(_b) \
-+ do { \
-+ if (atomic_dec_and_test(&(_b)->refcnt)) \
-+ tpmif_disconnect_complete(_b); \
-+ } while (0)
++ /* Register with generic device framework. */
++ err = device_register(&xendev->dev);
++ if (err)
++ goto fail;
+
-+extern int num_frontends;
++ err = device_create_file(&xendev->dev, &dev_attr_nodename);
++ if (err)
++ goto unregister;
++ err = device_create_file(&xendev->dev, &dev_attr_devtype);
++ if (err)
++ goto unregister;
+
-+static inline unsigned long idx_to_kaddr(tpmif_t *t, unsigned int idx)
-+{
-+ return (unsigned long)pfn_to_kaddr(page_to_pfn(t->mmap_pages[idx]));
++ return 0;
++unregister:
++ device_remove_file(&xendev->dev, &dev_attr_nodename);
++ device_remove_file(&xendev->dev, &dev_attr_devtype);
++ device_unregister(&xendev->dev);
++fail:
++ kfree(xendev);
++ return err;
+}
+
-+#endif /* __TPMIF__BACKEND__COMMON_H__ */
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/tpmback/interface.c tmp-linux-2.6-xen.patch/drivers/xen/tpmback/interface.c
---- pristine-linux-2.6.18.2/drivers/xen/tpmback/interface.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/tpmback/interface.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,167 @@
-+ /*****************************************************************************
-+ * drivers/xen/tpmback/interface.c
-+ *
-+ * Vritual TPM interface management.
-+ *
-+ * Copyright (c) 2005, IBM Corporation
-+ *
-+ * Author: Stefan Berger, stefanb at us.ibm.com
-+ *
-+ * This code has been derived from drivers/xen/netback/interface.c
-+ * Copyright (c) 2004, Keir Fraser
-+ */
++/* device/<typename>/<name> */
++static int xenbus_probe_frontend(const char *type, const char *name)
++{
++ char *nodename;
++ int err;
+
-+#include "common.h"
-+#include <xen/balloon.h>
-+#include <xen/gnttab.h>
++ nodename = kasprintf(GFP_KERNEL, "%s/%s/%s", xenbus_frontend.root, type, name);
++ if (!nodename)
++ return -ENOMEM;
+
-+static kmem_cache_t *tpmif_cachep;
-+int num_frontends = 0;
++ DPRINTK("%s", nodename);
+
-+LIST_HEAD(tpmif_list);
++ err = xenbus_probe_node(&xenbus_frontend, type, nodename);
++ kfree(nodename);
++ return err;
++}
+
-+static tpmif_t *alloc_tpmif(domid_t domid, struct backend_info *bi)
++static int xenbus_probe_device_type(struct xen_bus_type *bus, const char *type)
+{
-+ tpmif_t *tpmif;
++ int err = 0;
++ char **dir;
++ unsigned int dir_n = 0;
++ int i;
+
-+ tpmif = kmem_cache_alloc(tpmif_cachep, GFP_KERNEL);
-+ if (tpmif == NULL)
-+ goto out_of_memory;
++ dir = xenbus_directory(XBT_NIL, bus->root, type, &dir_n);
++ if (IS_ERR(dir))
++ return PTR_ERR(dir);
+
-+ memset(tpmif, 0, sizeof (*tpmif));
-+ tpmif->domid = domid;
-+ tpmif->status = DISCONNECTED;
-+ tpmif->bi = bi;
-+ snprintf(tpmif->devname, sizeof(tpmif->devname), "tpmif%d", domid);
-+ atomic_set(&tpmif->refcnt, 1);
++ for (i = 0; i < dir_n; i++) {
++ err = bus->probe(type, dir[i]);
++ if (err)
++ break;
++ }
++ kfree(dir);
++ return err;
++}
+
-+ tpmif->mmap_pages = alloc_empty_pages_and_pagevec(TPMIF_TX_RING_SIZE);
-+ if (tpmif->mmap_pages == NULL)
-+ goto out_of_memory;
++int xenbus_probe_devices(struct xen_bus_type *bus)
++{
++ int err = 0;
++ char **dir;
++ unsigned int i, dir_n;
+
-+ list_add(&tpmif->tpmif_list, &tpmif_list);
-+ num_frontends++;
++ if (bus->error)
++ return bus->error;
+
-+ return tpmif;
++ dir = xenbus_directory(XBT_NIL, bus->root, "", &dir_n);
++ if (IS_ERR(dir))
++ return PTR_ERR(dir);
+
-+ out_of_memory:
-+ if (tpmif != NULL)
-+ kmem_cache_free(tpmif_cachep, tpmif);
-+ printk("%s: out of memory\n", __FUNCTION__);
-+ return ERR_PTR(-ENOMEM);
++ for (i = 0; i < dir_n; i++) {
++ err = xenbus_probe_device_type(bus, dir[i]);
++ if (err)
++ break;
++ }
++ kfree(dir);
++ return err;
+}
+
-+static void free_tpmif(tpmif_t * tpmif)
++static unsigned int char_count(const char *str, char c)
+{
-+ num_frontends--;
-+ list_del(&tpmif->tpmif_list);
-+ free_empty_pages_and_pagevec(tpmif->mmap_pages, TPMIF_TX_RING_SIZE);
-+ kmem_cache_free(tpmif_cachep, tpmif);
++ unsigned int i, ret = 0;
++
++ for (i = 0; str[i]; i++)
++ if (str[i] == c)
++ ret++;
++ return ret;
+}
+
-+tpmif_t *tpmif_find(domid_t domid, struct backend_info *bi)
++static int strsep_len(const char *str, char c, unsigned int len)
+{
-+ tpmif_t *tpmif;
++ unsigned int i;
+
-+ list_for_each_entry(tpmif, &tpmif_list, tpmif_list) {
-+ if (tpmif->bi == bi) {
-+ if (tpmif->domid == domid) {
-+ tpmif_get(tpmif);
-+ return tpmif;
-+ } else {
-+ return ERR_PTR(-EEXIST);
-+ }
++ for (i = 0; str[i]; i++)
++ if (str[i] == c) {
++ if (len == 0)
++ return i;
++ len--;
+ }
++ return (len == 0) ? i : -ERANGE;
++}
++
++void dev_changed(const char *node, struct xen_bus_type *bus)
++{
++ int exists, rootlen;
++ struct xenbus_device *dev;
++ char type[BUS_ID_SIZE];
++ const char *p, *root;
++
++ if (bus->error || char_count(node, '/') < 2)
++ return;
++
++ exists = xenbus_exists(XBT_NIL, node, "");
++ if (!exists) {
++ xenbus_cleanup_devices(node, &bus->bus);
++ return;
+ }
+
-+ return alloc_tpmif(domid, bi);
++ /* backend/<type>/... or device/<type>/... */
++ p = strchr(node, '/') + 1;
++ snprintf(type, BUS_ID_SIZE, "%.*s", (int)strcspn(p, "/"), p);
++ type[BUS_ID_SIZE-1] = '\0';
++
++ rootlen = strsep_len(node, '/', bus->levels);
++ if (rootlen < 0)
++ return;
++ root = kasprintf(GFP_KERNEL, "%.*s", rootlen, node);
++ if (!root)
++ return;
++
++ dev = xenbus_device_find(root, &bus->bus);
++ if (!dev)
++ xenbus_probe_node(bus, type, root);
++ else
++ put_device(&dev->dev);
++
++ kfree(root);
+}
+
-+static int map_frontend_page(tpmif_t *tpmif, unsigned long shared_page)
++static void frontend_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
+{
-+ struct gnttab_map_grant_ref op;
++ DPRINTK("");
+
-+ gnttab_set_map_op(&op, (unsigned long)tpmif->tx_area->addr,
-+ GNTMAP_host_map, shared_page, tpmif->domid);
++ dev_changed(vec[XS_WATCH_PATH], &xenbus_frontend);
++}
+
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
-+ BUG();
++/* We watch for devices appearing and vanishing. */
++static struct xenbus_watch fe_watch = {
++ .node = "device",
++ .callback = frontend_changed,
++};
+
-+ if (op.status) {
-+ DPRINTK(" Grant table operation failure !\n");
-+ return op.status;
-+ }
++static int suspend_dev(struct device *dev, void *data)
++{
++ int err = 0;
++ struct xenbus_driver *drv;
++ struct xenbus_device *xdev;
+
-+ tpmif->shmem_ref = shared_page;
-+ tpmif->shmem_handle = op.handle;
++ DPRINTK("");
+
++ if (dev->driver == NULL)
++ return 0;
++ drv = to_xenbus_driver(dev->driver);
++ xdev = container_of(dev, struct xenbus_device, dev);
++ if (drv->suspend)
++ err = drv->suspend(xdev);
++ if (err)
++ printk(KERN_WARNING
++ "xenbus: suspend %s failed: %i\n", dev->bus_id, err);
+ return 0;
+}
+
-+static void unmap_frontend_page(tpmif_t *tpmif)
++static int suspend_cancel_dev(struct device *dev, void *data)
+{
-+ struct gnttab_unmap_grant_ref op;
++ int err = 0;
++ struct xenbus_driver *drv;
++ struct xenbus_device *xdev;
+
-+ gnttab_set_unmap_op(&op, (unsigned long)tpmif->tx_area->addr,
-+ GNTMAP_host_map, tpmif->shmem_handle);
++ DPRINTK("");
+
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
-+ BUG();
++ if (dev->driver == NULL)
++ return 0;
++ drv = to_xenbus_driver(dev->driver);
++ xdev = container_of(dev, struct xenbus_device, dev);
++ if (drv->suspend_cancel)
++ err = drv->suspend_cancel(xdev);
++ if (err)
++ printk(KERN_WARNING
++ "xenbus: suspend_cancel %s failed: %i\n",
++ dev->bus_id, err);
++ return 0;
+}
+
-+int tpmif_map(tpmif_t *tpmif, unsigned long shared_page, unsigned int evtchn)
++static int resume_dev(struct device *dev, void *data)
+{
+ int err;
++ struct xenbus_driver *drv;
++ struct xenbus_device *xdev;
+
-+ if (tpmif->irq)
++ DPRINTK("");
++
++ if (dev->driver == NULL)
+ return 0;
+
-+ if ((tpmif->tx_area = alloc_vm_area(PAGE_SIZE)) == NULL)
-+ return -ENOMEM;
++ drv = to_xenbus_driver(dev->driver);
++ xdev = container_of(dev, struct xenbus_device, dev);
+
-+ err = map_frontend_page(tpmif, shared_page);
++ err = talk_to_otherend(xdev);
+ if (err) {
-+ free_vm_area(tpmif->tx_area);
++ printk(KERN_WARNING
++ "xenbus: resume (talk_to_otherend) %s failed: %i\n",
++ dev->bus_id, err);
+ return err;
+ }
+
-+ tpmif->tx = (tpmif_tx_interface_t *)tpmif->tx_area->addr;
++ xdev->state = XenbusStateInitialising;
+
-+ err = bind_interdomain_evtchn_to_irqhandler(
-+ tpmif->domid, evtchn, tpmif_be_int, 0, tpmif->devname, tpmif);
-+ if (err < 0) {
-+ unmap_frontend_page(tpmif);
-+ free_vm_area(tpmif->tx_area);
-+ return err;
++ if (drv->resume) {
++ err = drv->resume(xdev);
++ if (err) {
++ printk(KERN_WARNING
++ "xenbus: resume %s failed: %i\n",
++ dev->bus_id, err);
++ return err;
++ }
+ }
-+ tpmif->irq = err;
+
-+ tpmif->shmem_ref = shared_page;
-+ tpmif->active = 1;
++ err = watch_otherend(xdev);
++ if (err) {
++ printk(KERN_WARNING
++ "xenbus_probe: resume (watch_otherend) %s failed: "
++ "%d.\n", dev->bus_id, err);
++ return err;
++ }
+
+ return 0;
+}
+
-+void tpmif_disconnect_complete(tpmif_t *tpmif)
++void xenbus_suspend(void)
+{
-+ if (tpmif->irq)
-+ unbind_from_irqhandler(tpmif->irq, tpmif);
-+
-+ if (tpmif->tx) {
-+ unmap_frontend_page(tpmif);
-+ free_vm_area(tpmif->tx_area);
-+ }
++ DPRINTK("");
+
-+ free_tpmif(tpmif);
++ if (!xenbus_frontend.error)
++ bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_dev);
++ xenbus_backend_suspend(suspend_dev);
++ xs_suspend();
+}
++EXPORT_SYMBOL_GPL(xenbus_suspend);
+
-+void __init tpmif_interface_init(void)
++void xenbus_resume(void)
+{
-+ tpmif_cachep = kmem_cache_create("tpmif_cache", sizeof (tpmif_t),
-+ 0, 0, NULL, NULL);
++ xb_init_comms();
++ xs_resume();
++ if (!xenbus_frontend.error)
++ bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, resume_dev);
++ xenbus_backend_resume(resume_dev);
+}
++EXPORT_SYMBOL_GPL(xenbus_resume);
+
-+void __exit tpmif_interface_exit(void)
++void xenbus_suspend_cancel(void)
+{
-+ kmem_cache_destroy(tpmif_cachep);
++ xs_suspend_cancel();
++ if (!xenbus_frontend.error)
++ bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_cancel_dev);
++ xenbus_backend_resume(suspend_cancel_dev);
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/tpmback/Makefile tmp-linux-2.6-xen.patch/drivers/xen/tpmback/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/tpmback/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/tpmback/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,4 @@
-+
-+obj-$(CONFIG_XEN_TPMDEV_BACKEND) += tpmbk.o
-+
-+tpmbk-y += tpmback.o interface.o xenbus.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/tpmback/tpmback.c tmp-linux-2.6-xen.patch/drivers/xen/tpmback/tpmback.c
---- pristine-linux-2.6.18.2/drivers/xen/tpmback/tpmback.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/tpmback/tpmback.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,944 @@
-+/******************************************************************************
-+ * drivers/xen/tpmback/tpmback.c
-+ *
-+ * Copyright (c) 2005, IBM Corporation
-+ *
-+ * Author: Stefan Berger, stefanb at us.ibm.com
-+ * Grant table support: Mahadevan Gomathisankaran
-+ *
-+ * This code has been derived from drivers/xen/netback/netback.c
-+ * Copyright (c) 2002-2004, K A Fraser
-+ *
-+ */
-+
-+#include "common.h"
-+#include <xen/evtchn.h>
-+
-+#include <linux/types.h>
-+#include <linux/list.h>
-+#include <linux/miscdevice.h>
-+#include <linux/poll.h>
-+#include <asm/uaccess.h>
-+#include <xen/xenbus.h>
-+#include <xen/interface/grant_table.h>
-+#include <xen/gnttab.h>
-+
-+/* local data structures */
-+struct data_exchange {
-+ struct list_head pending_pak;
-+ struct list_head current_pak;
-+ unsigned int copied_so_far;
-+ u8 has_opener:1;
-+ u8 aborted:1;
-+ rwlock_t pak_lock; // protects all of the previous fields
-+ wait_queue_head_t wait_queue;
-+};
-+
-+struct vtpm_resp_hdr {
-+ uint32_t instance_no;
-+ uint16_t tag_no;
-+ uint32_t len_no;
-+ uint32_t ordinal_no;
-+} __attribute__ ((packed));
-+
-+struct packet {
-+ struct list_head next;
-+ unsigned int data_len;
-+ u8 *data_buffer;
-+ tpmif_t *tpmif;
-+ u32 tpm_instance;
-+ u8 req_tag;
-+ u32 last_read;
-+ u8 flags;
-+ struct timer_list processing_timer;
-+};
-+
-+enum {
-+ PACKET_FLAG_DISCARD_RESPONSE = 1,
-+};
++EXPORT_SYMBOL_GPL(xenbus_suspend_cancel);
+
-+/* local variables */
-+static struct data_exchange dataex;
++/* A flag to determine if xenstored is 'ready' (i.e. has started) */
++int xenstored_ready = 0;
+
-+/* local function prototypes */
-+static int _packet_write(struct packet *pak,
-+ const char *data, size_t size, int userbuffer);
-+static void processing_timeout(unsigned long ptr);
-+static int packet_read_shmem(struct packet *pak,
-+ tpmif_t * tpmif,
-+ u32 offset,
-+ char *buffer, int isuserbuffer, u32 left);
-+static int vtpm_queue_packet(struct packet *pak);
+
-+/***************************************************************
-+ Buffer copying fo user and kernel space buffes.
-+***************************************************************/
-+static inline int copy_from_buffer(void *to,
-+ const void *from, unsigned long size,
-+ int isuserbuffer)
++int register_xenstore_notifier(struct notifier_block *nb)
+{
-+ if (isuserbuffer) {
-+ if (copy_from_user(to, (void __user *)from, size))
-+ return -EFAULT;
-+ } else {
-+ memcpy(to, from, size);
-+ }
-+ return 0;
-+}
++ int ret = 0;
+
-+static inline int copy_to_buffer(void *to,
-+ const void *from, unsigned long size,
-+ int isuserbuffer)
-+{
-+ if (isuserbuffer) {
-+ if (copy_to_user((void __user *)to, from, size))
-+ return -EFAULT;
-+ } else {
-+ memcpy(to, from, size);
-+ }
-+ return 0;
-+}
++ if (xenstored_ready > 0)
++ ret = nb->notifier_call(nb, 0, NULL);
++ else
++ blocking_notifier_chain_register(&xenstore_chain, nb);
+
++ return ret;
++}
++EXPORT_SYMBOL_GPL(register_xenstore_notifier);
+
-+static void dataex_init(struct data_exchange *dataex)
++void unregister_xenstore_notifier(struct notifier_block *nb)
+{
-+ INIT_LIST_HEAD(&dataex->pending_pak);
-+ INIT_LIST_HEAD(&dataex->current_pak);
-+ dataex->has_opener = 0;
-+ rwlock_init(&dataex->pak_lock);
-+ init_waitqueue_head(&dataex->wait_queue);
++ blocking_notifier_chain_unregister(&xenstore_chain, nb);
+}
++EXPORT_SYMBOL_GPL(unregister_xenstore_notifier);
+
-+/***************************************************************
-+ Packet-related functions
-+***************************************************************/
+
-+static struct packet *packet_find_instance(struct list_head *head,
-+ u32 tpm_instance)
++void xenbus_probe(void *unused)
+{
-+ struct packet *pak;
-+ struct list_head *p;
++ BUG_ON((xenstored_ready <= 0));
+
-+ /*
-+ * traverse the list of packets and return the first
-+ * one with the given instance number
-+ */
-+ list_for_each(p, head) {
-+ pak = list_entry(p, struct packet, next);
++ /* Enumerate devices in xenstore and watch for changes. */
++ xenbus_probe_devices(&xenbus_frontend);
++ register_xenbus_watch(&fe_watch);
++ xenbus_backend_probe_and_watch();
+
-+ if (pak->tpm_instance == tpm_instance) {
-+ return pak;
-+ }
-+ }
-+ return NULL;
++ /* Notify others that xenstore is up */
++ blocking_notifier_call_chain(&xenstore_chain, 0, NULL);
+}
+
-+static struct packet *packet_find_packet(struct list_head *head, void *packet)
-+{
-+ struct packet *pak;
-+ struct list_head *p;
-+
-+ /*
-+ * traverse the list of packets and return the first
-+ * one with the given instance number
-+ */
-+ list_for_each(p, head) {
-+ pak = list_entry(p, struct packet, next);
+
-+ if (pak == packet) {
-+ return pak;
-+ }
-+ }
-+ return NULL;
-+}
++#if defined(CONFIG_PROC_FS) && defined(CONFIG_XEN_PRIVILEGED_GUEST)
++static struct file_operations xsd_kva_fops;
++static struct proc_dir_entry *xsd_kva_intf;
++static struct proc_dir_entry *xsd_port_intf;
+
-+static struct packet *packet_alloc(tpmif_t * tpmif,
-+ u32 size, u8 req_tag, u8 flags)
++static int xsd_kva_mmap(struct file *file, struct vm_area_struct *vma)
+{
-+ struct packet *pak = NULL;
-+ pak = kzalloc(sizeof (struct packet), GFP_ATOMIC);
-+ if (NULL != pak) {
-+ if (tpmif) {
-+ pak->tpmif = tpmif;
-+ pak->tpm_instance = tpmback_get_instance(tpmif->bi);
-+ tpmif_get(tpmif);
-+ }
-+ pak->data_len = size;
-+ pak->req_tag = req_tag;
-+ pak->last_read = 0;
-+ pak->flags = flags;
-+
-+ /*
-+ * cannot do tpmif_get(tpmif); bad things happen
-+ * on the last tpmif_put()
-+ */
-+ init_timer(&pak->processing_timer);
-+ pak->processing_timer.function = processing_timeout;
-+ pak->processing_timer.data = (unsigned long)pak;
-+ }
-+ return pak;
-+}
++ size_t size = vma->vm_end - vma->vm_start;
+
-+static void inline packet_reset(struct packet *pak)
-+{
-+ pak->last_read = 0;
-+}
++ if ((size > PAGE_SIZE) || (vma->vm_pgoff != 0))
++ return -EINVAL;
+
-+static void packet_free(struct packet *pak)
-+{
-+ if (timer_pending(&pak->processing_timer)) {
-+ BUG();
-+ }
++ if (remap_pfn_range(vma, vma->vm_start, mfn_to_pfn(xen_store_mfn),
++ size, vma->vm_page_prot))
++ return -EAGAIN;
+
-+ if (pak->tpmif)
-+ tpmif_put(pak->tpmif);
-+ kfree(pak->data_buffer);
-+ /*
-+ * cannot do tpmif_put(pak->tpmif); bad things happen
-+ * on the last tpmif_put()
-+ */
-+ kfree(pak);
++ return 0;
+}
+
-+
-+/*
-+ * Write data to the shared memory and send it to the FE.
-+ */
-+static int packet_write(struct packet *pak,
-+ const char *data, size_t size, int isuserbuffer)
++static int xsd_kva_read(char *page, char **start, off_t off,
++ int count, int *eof, void *data)
+{
-+ int rc = 0;
-+
-+ if (0 != (pak->flags & PACKET_FLAG_DISCARD_RESPONSE)) {
-+ /* Don't send a respone to this packet. Just acknowledge it. */
-+ rc = size;
-+ } else {
-+ rc = _packet_write(pak, data, size, isuserbuffer);
-+ }
++ int len;
+
-+ return rc;
++ len = sprintf(page, "0x%p", xen_store_interface);
++ *eof = 1;
++ return len;
+}
+
-+int _packet_write(struct packet *pak,
-+ const char *data, size_t size, int isuserbuffer)
++static int xsd_port_read(char *page, char **start, off_t off,
++ int count, int *eof, void *data)
+{
-+ /*
-+ * Write into the shared memory pages directly
-+ * and send it to the front end.
-+ */
-+ tpmif_t *tpmif = pak->tpmif;
-+ grant_handle_t handle;
-+ int rc = 0;
-+ unsigned int i = 0;
-+ unsigned int offset = 0;
++ int len;
++
++ len = sprintf(page, "%d", xen_store_evtchn);
++ *eof = 1;
++ return len;
++}
++#endif
+
-+ if (tpmif == NULL) {
-+ return -EFAULT;
-+ }
++static int xenbus_probe_init(void)
++{
++ int err = 0;
++ unsigned long page = 0;
+
-+ if (tpmif->status == DISCONNECTED) {
-+ return size;
-+ }
++ DPRINTK("");
+
-+ while (offset < size && i < TPMIF_TX_RING_SIZE) {
-+ unsigned int tocopy;
-+ struct gnttab_map_grant_ref map_op;
-+ struct gnttab_unmap_grant_ref unmap_op;
-+ tpmif_tx_request_t *tx;
++ if (!is_running_on_xen())
++ return -ENODEV;
+
-+ tx = &tpmif->tx->ring[i].req;
++ /* Register ourselves with the kernel bus subsystem */
++ xenbus_frontend.error = bus_register(&xenbus_frontend.bus);
++ if (xenbus_frontend.error)
++ printk(KERN_WARNING
++ "XENBUS: Error registering frontend bus: %i\n",
++ xenbus_frontend.error);
++ xenbus_backend_bus_register();
+
-+ if (0 == tx->addr) {
-+ DPRINTK("ERROR: Buffer for outgoing packet NULL?! i=%d\n", i);
-+ return 0;
-+ }
++ /*
++ * Domain0 doesn't have a store_evtchn or store_mfn yet.
++ */
++ if (is_initial_xendomain()) {
++ struct evtchn_alloc_unbound alloc_unbound;
+
-+ gnttab_set_map_op(&map_op, idx_to_kaddr(tpmif, i),
-+ GNTMAP_host_map, tx->ref, tpmif->domid);
++ /* Allocate page. */
++ page = get_zeroed_page(GFP_KERNEL);
++ if (!page)
++ return -ENOMEM;
+
-+ if (unlikely(HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref,
-+ &map_op, 1))) {
-+ BUG();
-+ }
++ xen_store_mfn = xen_start_info->store_mfn =
++ pfn_to_mfn(virt_to_phys((void *)page) >>
++ PAGE_SHIFT);
+
-+ handle = map_op.handle;
++ /* Next allocate a local port which xenstored can bind to */
++ alloc_unbound.dom = DOMID_SELF;
++ alloc_unbound.remote_dom = 0;
+
-+ if (map_op.status) {
-+ DPRINTK(" Grant table operation failure !\n");
-+ return 0;
++ err = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
++ &alloc_unbound);
++ if (err == -ENOSYS)
++ goto err;
++ BUG_ON(err);
++ xen_store_evtchn = xen_start_info->store_evtchn =
++ alloc_unbound.port;
++
++#if defined(CONFIG_PROC_FS) && defined(CONFIG_XEN_PRIVILEGED_GUEST)
++ /* And finally publish the above info in /proc/xen */
++ xsd_kva_intf = create_xen_proc_entry("xsd_kva", 0600);
++ if (xsd_kva_intf) {
++ memcpy(&xsd_kva_fops, xsd_kva_intf->proc_fops,
++ sizeof(xsd_kva_fops));
++ xsd_kva_fops.mmap = xsd_kva_mmap;
++ xsd_kva_intf->proc_fops = &xsd_kva_fops;
++ xsd_kva_intf->read_proc = xsd_kva_read;
+ }
++ xsd_port_intf = create_xen_proc_entry("xsd_port", 0400);
++ if (xsd_port_intf)
++ xsd_port_intf->read_proc = xsd_port_read;
++#endif
++ xen_store_interface = mfn_to_virt(xen_store_mfn);
++ } else {
++ xenstored_ready = 1;
++#ifdef CONFIG_XEN
++ xen_store_evtchn = xen_start_info->store_evtchn;
++ xen_store_mfn = xen_start_info->store_mfn;
++ xen_store_interface = mfn_to_virt(xen_store_mfn);
++#else
++ xen_store_evtchn = hvm_get_parameter(HVM_PARAM_STORE_EVTCHN);
++ xen_store_mfn = hvm_get_parameter(HVM_PARAM_STORE_PFN);
++ xen_store_interface = ioremap(xen_store_mfn << PAGE_SHIFT,
++ PAGE_SIZE);
++#endif
++ }
+
-+ tocopy = min_t(size_t, size - offset, PAGE_SIZE);
+
-+ if (copy_from_buffer((void *)(idx_to_kaddr(tpmif, i) |
-+ (tx->addr & ~PAGE_MASK)),
-+ &data[offset], tocopy, isuserbuffer)) {
-+ tpmif_put(tpmif);
-+ return -EFAULT;
-+ }
-+ tx->size = tocopy;
++ xenbus_dev_init();
+
-+ gnttab_set_unmap_op(&unmap_op, idx_to_kaddr(tpmif, i),
-+ GNTMAP_host_map, handle);
++ /* Initialize the interface to xenstore. */
++ err = xs_init();
++ if (err) {
++ printk(KERN_WARNING
++ "XENBUS: Error initializing xenstore comms: %i\n", err);
++ goto err;
++ }
+
-+ if (unlikely
-+ (HYPERVISOR_grant_table_op
-+ (GNTTABOP_unmap_grant_ref, &unmap_op, 1))) {
-+ BUG();
++ /* Register ourselves with the kernel device subsystem */
++ if (!xenbus_frontend.error) {
++ xenbus_frontend.error = device_register(&xenbus_frontend.dev);
++ if (xenbus_frontend.error) {
++ bus_unregister(&xenbus_frontend.bus);
++ printk(KERN_WARNING
++ "XENBUS: Error registering frontend device: %i\n",
++ xenbus_frontend.error);
+ }
-+
-+ offset += tocopy;
-+ i++;
+ }
++ xenbus_backend_device_register();
+
-+ rc = offset;
-+ DPRINTK("Notifying frontend via irq %d\n", tpmif->irq);
-+ notify_remote_via_irq(tpmif->irq);
++ if (!is_initial_xendomain())
++ xenbus_probe(NULL);
+
-+ return rc;
-+}
++ return 0;
+
-+/*
-+ * Read data from the shared memory and copy it directly into the
-+ * provided buffer. Advance the read_last indicator which tells
-+ * how many bytes have already been read.
-+ */
-+static int packet_read(struct packet *pak, size_t numbytes,
-+ char *buffer, size_t buffersize, int isuserbuffer)
-+{
-+ tpmif_t *tpmif = pak->tpmif;
++ err:
++ if (page)
++ free_page(page);
+
+ /*
-+ * Read 'numbytes' of data from the buffer. The first 4
-+ * bytes are the instance number in network byte order,
-+ * after that come the data from the shared memory buffer.
++ * Do not unregister the xenbus front/backend buses here. The buses
++ * must exist because front/backend drivers will use them when they are
++ * registered.
+ */
-+ u32 to_copy;
-+ u32 offset = 0;
-+ u32 room_left = buffersize;
-+
-+ if (pak->last_read < 4) {
-+ /*
-+ * copy the instance number into the buffer
-+ */
-+ u32 instance_no = htonl(pak->tpm_instance);
-+ u32 last_read = pak->last_read;
+
-+ to_copy = min_t(size_t, 4 - last_read, numbytes);
++ return err;
++}
+
-+ if (copy_to_buffer(&buffer[0],
-+ &(((u8 *) & instance_no)[last_read]),
-+ to_copy, isuserbuffer)) {
-+ return -EFAULT;
-+ }
++#ifdef CONFIG_XEN
++postcore_initcall(xenbus_probe_init);
++MODULE_LICENSE("Dual BSD/GPL");
++#else
++int xenbus_init(void)
++{
++ return xenbus_probe_init();
++}
++#endif
+
-+ pak->last_read += to_copy;
-+ offset += to_copy;
-+ room_left -= to_copy;
-+ }
++static int is_disconnected_device(struct device *dev, void *data)
++{
++ struct xenbus_device *xendev = to_xenbus_device(dev);
++ struct device_driver *drv = data;
++ struct xenbus_driver *xendrv;
+
+ /*
-+ * If the packet has a data buffer appended, read from it...
++ * A device with no driver will never connect. We care only about
++ * devices which should currently be in the process of connecting.
+ */
++ if (!dev->driver)
++ return 0;
+
-+ if (room_left > 0) {
-+ if (pak->data_buffer) {
-+ u32 to_copy = min_t(u32, pak->data_len - offset, room_left);
-+ u32 last_read = pak->last_read - 4;
++ /* Is this search limited to a particular driver? */
++ if (drv && (dev->driver != drv))
++ return 0;
+
-+ if (copy_to_buffer(&buffer[offset],
-+ &pak->data_buffer[last_read],
-+ to_copy, isuserbuffer)) {
-+ return -EFAULT;
-+ }
-+ pak->last_read += to_copy;
-+ offset += to_copy;
-+ } else {
-+ offset = packet_read_shmem(pak,
-+ tpmif,
-+ offset,
-+ buffer,
-+ isuserbuffer, room_left);
-+ }
-+ }
-+ return offset;
++ xendrv = to_xenbus_driver(dev->driver);
++ return (xendev->state < XenbusStateConnected ||
++ (xendrv->is_ready && !xendrv->is_ready(xendev)));
+}
+
-+static int packet_read_shmem(struct packet *pak,
-+ tpmif_t * tpmif,
-+ u32 offset, char *buffer, int isuserbuffer,
-+ u32 room_left)
++static int exists_disconnected_device(struct device_driver *drv)
+{
-+ u32 last_read = pak->last_read - 4;
-+ u32 i = (last_read / PAGE_SIZE);
-+ u32 pg_offset = last_read & (PAGE_SIZE - 1);
-+ u32 to_copy;
-+ grant_handle_t handle;
++ if (xenbus_frontend.error)
++ return xenbus_frontend.error;
++ return bus_for_each_dev(&xenbus_frontend.bus, NULL, drv,
++ is_disconnected_device);
++}
+
-+ tpmif_tx_request_t *tx;
++static int print_device_status(struct device *dev, void *data)
++{
++ struct xenbus_device *xendev = to_xenbus_device(dev);
++ struct device_driver *drv = data;
++ struct xenbus_driver *xendrv;
+
-+ tx = &tpmif->tx->ring[0].req;
-+ /*
-+ * Start copying data at the page with index 'index'
-+ * and within that page at offset 'offset'.
-+ * Copy a maximum of 'room_left' bytes.
-+ */
-+ to_copy = min_t(u32, PAGE_SIZE - pg_offset, room_left);
-+ while (to_copy > 0) {
-+ void *src;
-+ struct gnttab_map_grant_ref map_op;
-+ struct gnttab_unmap_grant_ref unmap_op;
++ /* Is this operation limited to a particular driver? */
++ if (drv && (dev->driver != drv))
++ return 0;
+
-+ tx = &tpmif->tx->ring[i].req;
++ if (!dev->driver) {
++ /* Information only: is this too noisy? */
++ printk(KERN_INFO "XENBUS: Device with no driver: %s\n",
++ xendev->nodename);
++ return 0;
++ }
+
-+ gnttab_set_map_op(&map_op, idx_to_kaddr(tpmif, i),
-+ GNTMAP_host_map, tx->ref, tpmif->domid);
++ if (xendev->state < XenbusStateConnected) {
++ enum xenbus_state rstate = XenbusStateUnknown;
++ if (xendev->otherend)
++ rstate = xenbus_read_driver_state(xendev->otherend);
++ printk(KERN_WARNING "XENBUS: Timeout connecting "
++ "to device: %s (local state %d, remote state %d)\n",
++ xendev->nodename, xendev->state, rstate);
++ }
+
-+ if (unlikely(HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref,
-+ &map_op, 1))) {
-+ BUG();
-+ }
++ xendrv = to_xenbus_driver(dev->driver);
++ if (xendrv->is_ready && !xendrv->is_ready(xendev))
++ printk(KERN_WARNING "XENBUS: Device not ready: %s\n",
++ xendev->nodename);
+
-+ if (map_op.status) {
-+ DPRINTK(" Grant table operation failure !\n");
-+ return -EFAULT;
-+ }
++ return 0;
++}
+
-+ handle = map_op.handle;
++/* We only wait for device setup after most initcalls have run. */
++static int ready_to_wait_for_devices;
+
-+ if (to_copy > tx->size) {
-+ /*
-+ * User requests more than what's available
-+ */
-+ to_copy = min_t(u32, tx->size, to_copy);
-+ }
++/*
++ * On a 5-minute timeout, wait for all devices currently configured. We need
++ * to do this to guarantee that the filesystems and / or network devices
++ * needed for boot are available, before we can allow the boot to proceed.
++ *
++ * This needs to be on a late_initcall, to happen after the frontend device
++ * drivers have been initialised, but before the root fs is mounted.
++ *
++ * A possible improvement here would be to have the tools add a per-device
++ * flag to the store entry, indicating whether it is needed at boot time.
++ * This would allow people who knew what they were doing to accelerate their
++ * boot slightly, but of course needs tools or manual intervention to set up
++ * those flags correctly.
++ */
++static void wait_for_devices(struct xenbus_driver *xendrv)
++{
++ unsigned long start = jiffies;
++ struct device_driver *drv = xendrv ? &xendrv->driver : NULL;
++ unsigned int seconds_waited = 0;
+
-+ DPRINTK("Copying from mapped memory at %08lx\n",
-+ (unsigned long)(idx_to_kaddr(tpmif, i) |
-+ (tx->addr & ~PAGE_MASK)));
++ if (!ready_to_wait_for_devices || !is_running_on_xen())
++ return;
+
-+ src = (void *)(idx_to_kaddr(tpmif, i) |
-+ ((tx->addr & ~PAGE_MASK) + pg_offset));
-+ if (copy_to_buffer(&buffer[offset],
-+ src, to_copy, isuserbuffer)) {
-+ return -EFAULT;
++ while (exists_disconnected_device(drv)) {
++ if (time_after(jiffies, start + (seconds_waited+5)*HZ)) {
++ if (!seconds_waited)
++ printk(KERN_WARNING "XENBUS: Waiting for "
++ "devices to initialise: ");
++ seconds_waited += 5;
++ printk("%us...", 300 - seconds_waited);
++ if (seconds_waited == 300)
++ break;
+ }
++
++ schedule_timeout_interruptible(HZ/10);
++ }
+
-+ DPRINTK("Data from TPM-FE of domain %d are %d %d %d %d\n",
-+ tpmif->domid, buffer[offset], buffer[offset + 1],
-+ buffer[offset + 2], buffer[offset + 3]);
++ if (seconds_waited)
++ printk("\n");
+
-+ gnttab_set_unmap_op(&unmap_op, idx_to_kaddr(tpmif, i),
-+ GNTMAP_host_map, handle);
++ bus_for_each_dev(&xenbus_frontend.bus, NULL, drv,
++ print_device_status);
++}
+
-+ if (unlikely
-+ (HYPERVISOR_grant_table_op
-+ (GNTTABOP_unmap_grant_ref, &unmap_op, 1))) {
-+ BUG();
-+ }
++#ifndef MODULE
++static int __init boot_wait_for_devices(void)
++{
++ if (!xenbus_frontend.error) {
++ ready_to_wait_for_devices = 1;
++ wait_for_devices(NULL);
++ }
++ return 0;
++}
+
-+ offset += to_copy;
-+ pg_offset = 0;
-+ last_read += to_copy;
-+ room_left -= to_copy;
++late_initcall(boot_wait_for_devices);
++#endif
+
-+ to_copy = min_t(u32, PAGE_SIZE, room_left);
-+ i++;
-+ } /* while (to_copy > 0) */
-+ /*
-+ * Adjust the last_read pointer
-+ */
-+ pak->last_read = last_read + 4;
-+ return offset;
++int xenbus_for_each_frontend(void *arg, int (*fn)(struct device *, void *))
++{
++ return bus_for_each_dev(&xenbus_frontend.bus, NULL, arg, fn);
+}
-+
-+/* ============================================================
-+ * The file layer for reading data from this device
-+ * ============================================================
++EXPORT_SYMBOL_GPL(xenbus_for_each_frontend);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/xenbus/xenbus_probe.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/xenbus/xenbus_probe.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,75 @@
++/******************************************************************************
++ * xenbus_probe.h
++ *
++ * Talks to Xen Store to figure out what devices we have.
++ *
++ * Copyright (C) 2005 Rusty Russell, IBM Corporation
++ * Copyright (C) 2005 XenSource Ltd.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
-+static int vtpm_op_open(struct inode *inode, struct file *f)
-+{
-+ int rc = 0;
-+ unsigned long flags;
+
-+ write_lock_irqsave(&dataex.pak_lock, flags);
-+ if (dataex.has_opener == 0) {
-+ dataex.has_opener = 1;
-+ } else {
-+ rc = -EPERM;
-+ }
-+ write_unlock_irqrestore(&dataex.pak_lock, flags);
-+ return rc;
-+}
++#ifndef _XENBUS_PROBE_H
++#define _XENBUS_PROBE_H
+
-+static ssize_t vtpm_op_read(struct file *file,
-+ char __user * data, size_t size, loff_t * offset)
++#if defined(CONFIG_XEN_BACKEND) || defined(CONFIG_XEN_BACKEND_MODULE)
++extern void xenbus_backend_suspend(int (*fn)(struct device *, void *));
++extern void xenbus_backend_resume(int (*fn)(struct device *, void *));
++extern void xenbus_backend_probe_and_watch(void);
++extern void xenbus_backend_bus_register(void);
++extern void xenbus_backend_device_register(void);
++#else
++static inline void xenbus_backend_suspend(int (*fn)(struct device *, void *)) {}
++static inline void xenbus_backend_resume(int (*fn)(struct device *, void *)) {}
++static inline void xenbus_backend_probe_and_watch(void) {}
++static inline void xenbus_backend_bus_register(void) {}
++static inline void xenbus_backend_device_register(void) {}
++#endif
++
++struct xen_bus_type
+{
-+ int ret_size = -ENODATA;
-+ struct packet *pak = NULL;
-+ unsigned long flags;
++ char *root;
++ int error;
++ unsigned int levels;
++ int (*get_bus_id)(char bus_id[BUS_ID_SIZE], const char *nodename);
++ int (*probe)(const char *type, const char *dir);
++ struct bus_type bus;
++ struct device dev;
++};
+
-+ write_lock_irqsave(&dataex.pak_lock, flags);
-+ if (dataex.aborted) {
-+ dataex.aborted = 0;
-+ dataex.copied_so_far = 0;
-+ write_unlock_irqrestore(&dataex.pak_lock, flags);
-+ return -EIO;
-+ }
++extern int xenbus_match(struct device *_dev, struct device_driver *_drv);
++extern int xenbus_dev_probe(struct device *_dev);
++extern int xenbus_dev_remove(struct device *_dev);
++extern int xenbus_register_driver_common(struct xenbus_driver *drv,
++ struct xen_bus_type *bus);
++extern int xenbus_probe_node(struct xen_bus_type *bus,
++ const char *type,
++ const char *nodename);
++extern int xenbus_probe_devices(struct xen_bus_type *bus);
++
++extern void dev_changed(const char *node, struct xen_bus_type *bus);
++
++#endif
++
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/xenbus/xenbus_probe_backend.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/xenbus/xenbus_probe_backend.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,292 @@
++/******************************************************************************
++ * Talks to Xen Store to figure out what devices we have (backend half).
++ *
++ * Copyright (C) 2005 Rusty Russell, IBM Corporation
++ * Copyright (C) 2005 Mike Wray, Hewlett-Packard
++ * Copyright (C) 2005, 2006 XenSource Ltd
++ * Copyright (C) 2007 Solarflare Communications, Inc.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#define DPRINTK(fmt, args...) \
++ pr_debug("xenbus_probe (%s:%d) " fmt ".\n", \
++ __FUNCTION__, __LINE__, ##args)
++
++#include <linux/kernel.h>
++#include <linux/err.h>
++#include <linux/string.h>
++#include <linux/ctype.h>
++#include <linux/fcntl.h>
++#include <linux/mm.h>
++#include <linux/notifier.h>
+
-+ if (list_empty(&dataex.pending_pak)) {
-+ write_unlock_irqrestore(&dataex.pak_lock, flags);
-+ wait_event_interruptible(dataex.wait_queue,
-+ !list_empty(&dataex.pending_pak));
-+ write_lock_irqsave(&dataex.pak_lock, flags);
-+ dataex.copied_so_far = 0;
-+ }
++#include <asm/io.h>
++#include <asm/page.h>
++#include <asm/maddr.h>
++#include <asm/pgtable.h>
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
++#include <xen/xen_proc.h>
++#include <xen/evtchn.h>
++#include <xen/features.h>
+
-+ if (!list_empty(&dataex.pending_pak)) {
-+ unsigned int left;
++#include "xenbus_comms.h"
++#include "xenbus_probe.h"
+
-+ pak = list_entry(dataex.pending_pak.next, struct packet, next);
-+ left = pak->data_len - dataex.copied_so_far;
-+ list_del(&pak->next);
-+ write_unlock_irqrestore(&dataex.pak_lock, flags);
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
+
-+ DPRINTK("size given by app: %d, available: %d\n", size, left);
++static int xenbus_uevent_backend(struct device *dev, char **envp,
++ int num_envp, char *buffer, int buffer_size);
++static int xenbus_probe_backend(const char *type, const char *domid);
+
-+ ret_size = min_t(size_t, size, left);
++extern int read_otherend_details(struct xenbus_device *xendev,
++ char *id_node, char *path_node);
+
-+ ret_size = packet_read(pak, ret_size, data, size, 1);
++static int read_frontend_details(struct xenbus_device *xendev)
++{
++ return read_otherend_details(xendev, "frontend-id", "frontend");
++}
+
-+ write_lock_irqsave(&dataex.pak_lock, flags);
++/* backend/<type>/<fe-uuid>/<id> => <type>-<fe-domid>-<id> */
++static int backend_bus_id(char bus_id[BUS_ID_SIZE], const char *nodename)
++{
++ int domid, err;
++ const char *devid, *type, *frontend;
++ unsigned int typelen;
+
-+ if (ret_size < 0) {
-+ del_singleshot_timer_sync(&pak->processing_timer);
-+ packet_free(pak);
-+ dataex.copied_so_far = 0;
-+ } else {
-+ DPRINTK("Copied %d bytes to user buffer\n", ret_size);
++ type = strchr(nodename, '/');
++ if (!type)
++ return -EINVAL;
++ type++;
++ typelen = strcspn(type, "/");
++ if (!typelen || type[typelen] != '/')
++ return -EINVAL;
+
-+ dataex.copied_so_far += ret_size;
-+ if (dataex.copied_so_far >= pak->data_len + 4) {
-+ DPRINTK("All data from this packet given to app.\n");
-+ /* All data given to app */
++ devid = strrchr(nodename, '/') + 1;
+
-+ del_singleshot_timer_sync(&pak->
-+ processing_timer);
-+ list_add_tail(&pak->next, &dataex.current_pak);
-+ /*
-+ * The more fontends that are handled at the same time,
-+ * the more time we give the TPM to process the request.
-+ */
-+ mod_timer(&pak->processing_timer,
-+ jiffies + (num_frontends * 60 * HZ));
-+ dataex.copied_so_far = 0;
-+ } else {
-+ list_add(&pak->next, &dataex.pending_pak);
-+ }
-+ }
-+ }
-+ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ err = xenbus_gather(XBT_NIL, nodename, "frontend-id", "%i", &domid,
++ "frontend", NULL, &frontend,
++ NULL);
++ if (err)
++ return err;
++ if (strlen(frontend) == 0)
++ err = -ERANGE;
++ if (!err && !xenbus_exists(XBT_NIL, frontend, ""))
++ err = -ENOENT;
++ kfree(frontend);
+
-+ DPRINTK("Returning result from read to app: %d\n", ret_size);
++ if (err)
++ return err;
+
-+ return ret_size;
++ if (snprintf(bus_id, BUS_ID_SIZE,
++ "%.*s-%i-%s", typelen, type, domid, devid) >= BUS_ID_SIZE)
++ return -ENOSPC;
++ return 0;
+}
+
-+/*
-+ * Write operation - only works after a previous read operation!
-+ */
-+static ssize_t vtpm_op_write(struct file *file,
-+ const char __user * data, size_t size,
-+ loff_t * offset)
-+{
-+ struct packet *pak;
-+ int rc = 0;
-+ unsigned int off = 4;
-+ unsigned long flags;
-+ struct vtpm_resp_hdr vrh;
-+
-+ /*
-+ * Minimum required packet size is:
-+ * 4 bytes for instance number
-+ * 2 bytes for tag
-+ * 4 bytes for paramSize
-+ * 4 bytes for the ordinal
-+ * sum: 14 bytes
-+ */
-+ if (size < sizeof (vrh))
-+ return -EFAULT;
++static struct xen_bus_type xenbus_backend = {
++ .root = "backend",
++ .levels = 3, /* backend/type/<frontend>/<id> */
++ .get_bus_id = backend_bus_id,
++ .probe = xenbus_probe_backend,
++ .error = -ENODEV,
++ .bus = {
++ .name = "xen-backend",
++ .match = xenbus_match,
++ .probe = xenbus_dev_probe,
++ .remove = xenbus_dev_remove,
++// .shutdown = xenbus_dev_shutdown,
++ .uevent = xenbus_uevent_backend,
++ },
++ .dev = {
++ .bus_id = "xen-backend",
++ },
++};
+
-+ if (copy_from_user(&vrh, data, sizeof (vrh)))
-+ return -EFAULT;
++static int xenbus_uevent_backend(struct device *dev, char **envp,
++ int num_envp, char *buffer, int buffer_size)
++{
++ struct xenbus_device *xdev;
++ struct xenbus_driver *drv;
++ int i = 0;
++ int length = 0;
+
-+ /* malformed packet? */
-+ if ((off + ntohl(vrh.len_no)) != size)
-+ return -EFAULT;
++ DPRINTK("");
+
-+ write_lock_irqsave(&dataex.pak_lock, flags);
-+ pak = packet_find_instance(&dataex.current_pak,
-+ ntohl(vrh.instance_no));
++ if (dev == NULL)
++ return -ENODEV;
+
-+ if (pak == NULL) {
-+ write_unlock_irqrestore(&dataex.pak_lock, flags);
-+ DPRINTK(KERN_ALERT "No associated packet! (inst=%d)\n",
-+ ntohl(vrh.instance_no));
-+ return -EFAULT;
-+ }
++ xdev = to_xenbus_device(dev);
++ if (xdev == NULL)
++ return -ENODEV;
+
-+ del_singleshot_timer_sync(&pak->processing_timer);
-+ list_del(&pak->next);
++ /* stuff we want to pass to /sbin/hotplug */
++ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
++ "XENBUS_TYPE=%s", xdev->devicetype);
+
-+ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
++ "XENBUS_PATH=%s", xdev->nodename);
+
-+ /*
-+ * The first 'offset' bytes must be the instance number - skip them.
-+ */
-+ size -= off;
++ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
++ "XENBUS_BASE_PATH=%s", xenbus_backend.root);
+
-+ rc = packet_write(pak, &data[off], size, 1);
++ /* terminate, set to next free slot, shrink available space */
++ envp[i] = NULL;
++ envp = &envp[i];
++ num_envp -= i;
++ buffer = &buffer[length];
++ buffer_size -= length;
+
-+ if (rc > 0) {
-+ /* I neglected the first 4 bytes */
-+ rc += off;
++ if (dev->driver) {
++ drv = to_xenbus_driver(dev->driver);
++ if (drv && drv->uevent)
++ return drv->uevent(xdev, envp, num_envp, buffer,
++ buffer_size);
+ }
-+ packet_free(pak);
-+ return rc;
-+}
+
-+static int vtpm_op_release(struct inode *inode, struct file *file)
-+{
-+ unsigned long flags;
-+
-+ vtpm_release_packets(NULL, 1);
-+ write_lock_irqsave(&dataex.pak_lock, flags);
-+ dataex.has_opener = 0;
-+ write_unlock_irqrestore(&dataex.pak_lock, flags);
+ return 0;
+}
+
-+static unsigned int vtpm_op_poll(struct file *file,
-+ struct poll_table_struct *pts)
++int xenbus_register_backend(struct xenbus_driver *drv)
+{
-+ unsigned int flags = POLLOUT | POLLWRNORM;
++ drv->read_otherend_details = read_frontend_details;
+
-+ poll_wait(file, &dataex.wait_queue, pts);
-+ if (!list_empty(&dataex.pending_pak)) {
-+ flags |= POLLIN | POLLRDNORM;
-+ }
-+ return flags;
++ return xenbus_register_driver_common(drv, &xenbus_backend);
+}
++EXPORT_SYMBOL_GPL(xenbus_register_backend);
+
-+static const struct file_operations vtpm_ops = {
-+ .owner = THIS_MODULE,
-+ .llseek = no_llseek,
-+ .open = vtpm_op_open,
-+ .read = vtpm_op_read,
-+ .write = vtpm_op_write,
-+ .release = vtpm_op_release,
-+ .poll = vtpm_op_poll,
-+};
-+
-+static struct miscdevice vtpms_miscdevice = {
-+ .minor = 225,
-+ .name = "vtpm",
-+ .fops = &vtpm_ops,
-+};
-+
-+/***************************************************************
-+ Utility functions
-+***************************************************************/
-+
-+static int tpm_send_fail_message(struct packet *pak, u8 req_tag)
++/* backend/<typename>/<frontend-uuid>/<name> */
++static int xenbus_probe_backend_unit(const char *dir,
++ const char *type,
++ const char *name)
+{
-+ int rc;
-+ static const unsigned char tpm_error_message_fail[] = {
-+ 0x00, 0x00,
-+ 0x00, 0x00, 0x00, 0x0a,
-+ 0x00, 0x00, 0x00, 0x09 /* TPM_FAIL */
-+ };
-+ unsigned char buffer[sizeof (tpm_error_message_fail)];
++ char *nodename;
++ int err;
+
-+ memcpy(buffer, tpm_error_message_fail,
-+ sizeof (tpm_error_message_fail));
-+ /*
-+ * Insert the right response tag depending on the given tag
-+ * All response tags are '+3' to the request tag.
-+ */
-+ buffer[1] = req_tag + 3;
++ nodename = kasprintf(GFP_KERNEL, "%s/%s", dir, name);
++ if (!nodename)
++ return -ENOMEM;
+
-+ /*
-+ * Write the data to shared memory and notify the front-end
-+ */
-+ rc = packet_write(pak, buffer, sizeof (buffer), 0);
++ DPRINTK("%s\n", nodename);
+
-+ return rc;
++ err = xenbus_probe_node(&xenbus_backend, type, nodename);
++ kfree(nodename);
++ return err;
+}
+
-+static int _vtpm_release_packets(struct list_head *head,
-+ tpmif_t * tpmif, int send_msgs)
++/* backend/<typename>/<frontend-domid> */
++static int xenbus_probe_backend(const char *type, const char *domid)
+{
-+ int aborted = 0;
-+ int c = 0;
-+ struct packet *pak;
-+ struct list_head *pos, *tmp;
-+
-+ list_for_each_safe(pos, tmp, head) {
-+ pak = list_entry(pos, struct packet, next);
-+ c += 1;
++ char *nodename;
++ int err = 0;
++ char **dir;
++ unsigned int i, dir_n = 0;
+
-+ if (tpmif == NULL || pak->tpmif == tpmif) {
-+ int can_send = 0;
++ DPRINTK("");
+
-+ del_singleshot_timer_sync(&pak->processing_timer);
-+ list_del(&pak->next);
++ nodename = kasprintf(GFP_KERNEL, "%s/%s/%s", xenbus_backend.root, type, domid);
++ if (!nodename)
++ return -ENOMEM;
+
-+ if (pak->tpmif && pak->tpmif->status == CONNECTED) {
-+ can_send = 1;
-+ }
++ dir = xenbus_directory(XBT_NIL, nodename, "", &dir_n);
++ if (IS_ERR(dir)) {
++ kfree(nodename);
++ return PTR_ERR(dir);
++ }
+
-+ if (send_msgs && can_send) {
-+ tpm_send_fail_message(pak, pak->req_tag);
-+ }
-+ packet_free(pak);
-+ if (c == 1)
-+ aborted = 1;
-+ }
++ for (i = 0; i < dir_n; i++) {
++ err = xenbus_probe_backend_unit(nodename, type, dir[i]);
++ if (err)
++ break;
+ }
-+ return aborted;
++ kfree(dir);
++ kfree(nodename);
++ return err;
+}
+
-+int vtpm_release_packets(tpmif_t * tpmif, int send_msgs)
++static void backend_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
+{
-+ unsigned long flags;
++ DPRINTK("");
+
-+ write_lock_irqsave(&dataex.pak_lock, flags);
++ dev_changed(vec[XS_WATCH_PATH], &xenbus_backend);
++}
+
-+ dataex.aborted = _vtpm_release_packets(&dataex.pending_pak,
-+ tpmif,
-+ send_msgs);
-+ _vtpm_release_packets(&dataex.current_pak, tpmif, send_msgs);
++static struct xenbus_watch be_watch = {
++ .node = "backend",
++ .callback = backend_changed,
++};
+
-+ write_unlock_irqrestore(&dataex.pak_lock, flags);
-+ return 0;
++void xenbus_backend_suspend(int (*fn)(struct device *, void *))
++{
++ DPRINTK("");
++ if (!xenbus_backend.error)
++ bus_for_each_dev(&xenbus_backend.bus, NULL, NULL, fn);
+}
+
-+static int vtpm_queue_packet(struct packet *pak)
++void xenbus_backend_resume(int (*fn)(struct device *, void *))
+{
-+ int rc = 0;
++ DPRINTK("");
++ if (!xenbus_backend.error)
++ bus_for_each_dev(&xenbus_backend.bus, NULL, NULL, fn);
++}
+
-+ if (dataex.has_opener) {
-+ unsigned long flags;
++void xenbus_backend_probe_and_watch(void)
++{
++ xenbus_probe_devices(&xenbus_backend);
++ register_xenbus_watch(&be_watch);
++}
+
-+ write_lock_irqsave(&dataex.pak_lock, flags);
-+ list_add_tail(&pak->next, &dataex.pending_pak);
-+ /* give the TPM some time to pick up the request */
-+ mod_timer(&pak->processing_timer, jiffies + (30 * HZ));
-+ write_unlock_irqrestore(&dataex.pak_lock, flags);
++void xenbus_backend_bus_register(void)
++{
++ xenbus_backend.error = bus_register(&xenbus_backend.bus);
++ if (xenbus_backend.error)
++ printk(KERN_WARNING
++ "XENBUS: Error registering backend bus: %i\n",
++ xenbus_backend.error);
++}
+
-+ wake_up_interruptible(&dataex.wait_queue);
-+ } else {
-+ rc = -EFAULT;
++void xenbus_backend_device_register(void)
++{
++ if (xenbus_backend.error)
++ return;
++
++ xenbus_backend.error = device_register(&xenbus_backend.dev);
++ if (xenbus_backend.error) {
++ bus_unregister(&xenbus_backend.bus);
++ printk(KERN_WARNING
++ "XENBUS: Error registering backend device: %i\n",
++ xenbus_backend.error);
+ }
-+ return rc;
+}
+
-+static int vtpm_receive(tpmif_t * tpmif, u32 size)
++int xenbus_for_each_backend(void *arg, int (*fn)(struct device *, void *))
+{
-+ int rc = 0;
-+ unsigned char buffer[10];
-+ __be32 *native_size;
-+ struct packet *pak = packet_alloc(tpmif, size, 0, 0);
++ return bus_for_each_dev(&xenbus_backend.bus, NULL, arg, fn);
++}
++EXPORT_SYMBOL_GPL(xenbus_for_each_backend);
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/xenbus/xenbus_xs.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/xenbus/xenbus_xs.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,902 @@
++/******************************************************************************
++ * xenbus_xs.c
++ *
++ * This is the kernel equivalent of the "xs" library. We don't need everything
++ * and we use xenbus_comms for communication.
++ *
++ * Copyright (C) 2005 Rusty Russell, IBM Corporation
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ if (!pak)
-+ return -ENOMEM;
-+ /*
-+ * Read 10 bytes from the received buffer to test its
-+ * content for validity.
-+ */
-+ if (sizeof (buffer) != packet_read(pak,
-+ sizeof (buffer), buffer,
-+ sizeof (buffer), 0)) {
-+ goto failexit;
-+ }
-+ /*
-+ * Reset the packet read pointer so we can read all its
-+ * contents again.
-+ */
-+ packet_reset(pak);
++#include <linux/unistd.h>
++#include <linux/errno.h>
++#include <linux/types.h>
++#include <linux/uio.h>
++#include <linux/kernel.h>
++#include <linux/string.h>
++#include <linux/err.h>
++#include <linux/slab.h>
++#include <linux/fcntl.h>
++#include <linux/kthread.h>
++#include <linux/rwsem.h>
++#include <linux/module.h>
++#include <linux/mutex.h>
++#include <xen/xenbus.h>
++#include "xenbus_comms.h"
+
-+ native_size = (__force __be32 *) (&buffer[4 + 2]);
-+ /*
-+ * Verify that the size of the packet is correct
-+ * as indicated and that there's actually someone reading packets.
-+ * The minimum size of the packet is '10' for tag, size indicator
-+ * and ordinal.
-+ */
-+ if (size < 10 ||
-+ be32_to_cpu(*native_size) != size ||
-+ 0 == dataex.has_opener || tpmif->status != CONNECTED) {
-+ rc = -EINVAL;
-+ goto failexit;
-+ } else {
-+ rc = vtpm_queue_packet(pak);
-+ if (rc < 0)
-+ goto failexit;
-+ }
-+ return 0;
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
+
-+ failexit:
-+ if (pak) {
-+ tpm_send_fail_message(pak, buffer[4 + 1]);
-+ packet_free(pak);
-+ }
-+ return rc;
-+}
++#ifndef PF_NOFREEZE /* Old kernel (pre-2.6.6). */
++#define PF_NOFREEZE 0
++#endif
+
-+/*
-+ * Timeout function that gets invoked when a packet has not been processed
-+ * during the timeout period.
-+ * The packet must be on a list when this function is invoked. This
-+ * also means that once its taken off a list, the timer must be
-+ * destroyed as well.
-+ */
-+static void processing_timeout(unsigned long ptr)
-+{
-+ struct packet *pak = (struct packet *)ptr;
-+ unsigned long flags;
++struct xs_stored_msg {
++ struct list_head list;
++
++ struct xsd_sockmsg hdr;
++
++ union {
++ /* Queued replies. */
++ struct {
++ char *body;
++ } reply;
++
++ /* Queued watch events. */
++ struct {
++ struct xenbus_watch *handle;
++ char **vec;
++ unsigned int vec_size;
++ } watch;
++ } u;
++};
++
++struct xs_handle {
++ /* A list of replies. Currently only one will ever be outstanding. */
++ struct list_head reply_list;
++ spinlock_t reply_lock;
++ wait_queue_head_t reply_waitq;
+
-+ write_lock_irqsave(&dataex.pak_lock, flags);
+ /*
-+ * The packet needs to be searched whether it
-+ * is still on the list.
++ * Mutex ordering: transaction_mutex -> watch_mutex -> request_mutex.
++ * response_mutex is never taken simultaneously with the other three.
+ */
-+ if (pak == packet_find_packet(&dataex.pending_pak, pak) ||
-+ pak == packet_find_packet(&dataex.current_pak, pak)) {
-+ if ((pak->flags & PACKET_FLAG_DISCARD_RESPONSE) == 0) {
-+ tpm_send_fail_message(pak, pak->req_tag);
-+ }
-+ /* discard future responses */
-+ pak->flags |= PACKET_FLAG_DISCARD_RESPONSE;
-+ }
+
-+ write_unlock_irqrestore(&dataex.pak_lock, flags);
-+}
++ /* One request at a time. */
++ struct mutex request_mutex;
+
-+static void tpm_tx_action(unsigned long unused);
-+static DECLARE_TASKLET(tpm_tx_tasklet, tpm_tx_action, 0);
++ /* Protect xenbus reader thread against save/restore. */
++ struct mutex response_mutex;
+
-+static struct list_head tpm_schedule_list;
-+static spinlock_t tpm_schedule_list_lock;
++ /* Protect transactions against save/restore. */
++ struct rw_semaphore transaction_mutex;
+
-+static inline void maybe_schedule_tx_action(void)
-+{
-+ smp_mb();
-+ tasklet_schedule(&tpm_tx_tasklet);
-+}
++ /* Protect watch (de)register against save/restore. */
++ struct rw_semaphore watch_mutex;
++};
+
-+static inline int __on_tpm_schedule_list(tpmif_t * tpmif)
-+{
-+ return tpmif->list.next != NULL;
-+}
++static struct xs_handle xs_state;
+
-+static void remove_from_tpm_schedule_list(tpmif_t * tpmif)
-+{
-+ spin_lock_irq(&tpm_schedule_list_lock);
-+ if (likely(__on_tpm_schedule_list(tpmif))) {
-+ list_del(&tpmif->list);
-+ tpmif->list.next = NULL;
-+ tpmif_put(tpmif);
-+ }
-+ spin_unlock_irq(&tpm_schedule_list_lock);
-+}
++/* List of registered watches, and a lock to protect it. */
++static LIST_HEAD(watches);
++static DEFINE_SPINLOCK(watches_lock);
+
-+static void add_to_tpm_schedule_list_tail(tpmif_t * tpmif)
-+{
-+ if (__on_tpm_schedule_list(tpmif))
-+ return;
++/* List of pending watch callback events, and a lock to protect it. */
++static LIST_HEAD(watch_events);
++static DEFINE_SPINLOCK(watch_events_lock);
+
-+ spin_lock_irq(&tpm_schedule_list_lock);
-+ if (!__on_tpm_schedule_list(tpmif) && tpmif->active) {
-+ list_add_tail(&tpmif->list, &tpm_schedule_list);
-+ tpmif_get(tpmif);
-+ }
-+ spin_unlock_irq(&tpm_schedule_list_lock);
-+}
++/*
++ * Details of the xenwatch callback kernel thread. The thread waits on the
++ * watch_events_waitq for work to do (queued on watch_events list). When it
++ * wakes up it acquires the xenwatch_mutex before reading the list and
++ * carrying out work.
++ */
++static pid_t xenwatch_pid;
++/* static */ DEFINE_MUTEX(xenwatch_mutex);
++static DECLARE_WAIT_QUEUE_HEAD(watch_events_waitq);
+
-+void tpmif_schedule_work(tpmif_t * tpmif)
++static int get_error(const char *errorstring)
+{
-+ add_to_tpm_schedule_list_tail(tpmif);
-+ maybe_schedule_tx_action();
-+}
++ unsigned int i;
+
-+void tpmif_deschedule_work(tpmif_t * tpmif)
-+{
-+ remove_from_tpm_schedule_list(tpmif);
++ for (i = 0; strcmp(errorstring, xsd_errors[i].errstring) != 0; i++) {
++ if (i == ARRAY_SIZE(xsd_errors) - 1) {
++ printk(KERN_WARNING
++ "XENBUS xen store gave: unknown error %s",
++ errorstring);
++ return EINVAL;
++ }
++ }
++ return xsd_errors[i].errnum;
+}
+
-+static void tpm_tx_action(unsigned long unused)
++static void *read_reply(enum xsd_sockmsg_type *type, unsigned int *len)
+{
-+ struct list_head *ent;
-+ tpmif_t *tpmif;
-+ tpmif_tx_request_t *tx;
++ struct xs_stored_msg *msg;
++ char *body;
+
-+ DPRINTK("%s: Getting data from front-end(s)!\n", __FUNCTION__);
++ spin_lock(&xs_state.reply_lock);
+
-+ while (!list_empty(&tpm_schedule_list)) {
-+ /* Get a tpmif from the list with work to do. */
-+ ent = tpm_schedule_list.next;
-+ tpmif = list_entry(ent, tpmif_t, list);
-+ tpmif_get(tpmif);
-+ remove_from_tpm_schedule_list(tpmif);
++ while (list_empty(&xs_state.reply_list)) {
++ spin_unlock(&xs_state.reply_lock);
++ /* XXX FIXME: Avoid synchronous wait for response here. */
++ wait_event(xs_state.reply_waitq,
++ !list_empty(&xs_state.reply_list));
++ spin_lock(&xs_state.reply_lock);
++ }
+
-+ tx = &tpmif->tx->ring[0].req;
++ msg = list_entry(xs_state.reply_list.next,
++ struct xs_stored_msg, list);
++ list_del(&msg->list);
+
-+ /* pass it up */
-+ vtpm_receive(tpmif, tx->size);
++ spin_unlock(&xs_state.reply_lock);
+
-+ tpmif_put(tpmif);
-+ }
-+}
++ *type = msg->hdr.type;
++ if (len)
++ *len = msg->hdr.len;
++ body = msg->u.reply.body;
+
-+irqreturn_t tpmif_be_int(int irq, void *dev_id, struct pt_regs *regs)
-+{
-+ tpmif_t *tpmif = (tpmif_t *) dev_id;
++ kfree(msg);
+
-+ add_to_tpm_schedule_list_tail(tpmif);
-+ maybe_schedule_tx_action();
-+ return IRQ_HANDLED;
++ return body;
+}
+
-+static int __init tpmback_init(void)
++void *xenbus_dev_request_and_reply(struct xsd_sockmsg *msg)
+{
-+ int rc;
++ void *ret;
++ struct xsd_sockmsg req_msg = *msg;
++ int err;
+
-+ if ((rc = misc_register(&vtpms_miscdevice)) != 0) {
-+ printk(KERN_ALERT
-+ "Could not register misc device for TPM BE.\n");
-+ return rc;
-+ }
++ if (req_msg.type == XS_TRANSACTION_START)
++ down_read(&xs_state.transaction_mutex);
+
-+ dataex_init(&dataex);
++ mutex_lock(&xs_state.request_mutex);
+
-+ spin_lock_init(&tpm_schedule_list_lock);
-+ INIT_LIST_HEAD(&tpm_schedule_list);
++ err = xb_write(msg, sizeof(*msg) + msg->len);
++ if (err) {
++ msg->type = XS_ERROR;
++ ret = ERR_PTR(err);
++ } else
++ ret = read_reply(&msg->type, &msg->len);
+
-+ tpmif_interface_init();
-+ tpmif_xenbus_init();
++ mutex_unlock(&xs_state.request_mutex);
+
-+ printk(KERN_ALERT "Successfully initialized TPM backend driver.\n");
++ if ((req_msg.type == XS_TRANSACTION_END) ||
++ ((req_msg.type == XS_TRANSACTION_START) &&
++ (msg->type == XS_ERROR)))
++ up_read(&xs_state.transaction_mutex);
+
-+ return 0;
++ return ret;
+}
+
-+module_init(tpmback_init);
-+
-+void __exit tpmback_exit(void)
++/* Send message to xs, get kmalloc'ed reply. ERR_PTR() on error. */
++static void *xs_talkv(struct xenbus_transaction t,
++ enum xsd_sockmsg_type type,
++ const struct kvec *iovec,
++ unsigned int num_vecs,
++ unsigned int *len)
+{
-+ vtpm_release_packets(NULL, 0);
-+ tpmif_xenbus_exit();
-+ tpmif_interface_exit();
-+ misc_deregister(&vtpms_miscdevice);
-+}
-+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/tpmback/xenbus.c tmp-linux-2.6-xen.patch/drivers/xen/tpmback/xenbus.c
---- pristine-linux-2.6.18.2/drivers/xen/tpmback/xenbus.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/tpmback/xenbus.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,289 @@
-+/* Xenbus code for tpmif backend
-+ Copyright (C) 2005 IBM Corporation
-+ Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
++ struct xsd_sockmsg msg;
++ void *ret = NULL;
++ unsigned int i;
++ int err;
+
-+ This program is free software; you can redistribute it and/or modify
-+ it under the terms of the GNU General Public License as published by
-+ the Free Software Foundation; either version 2 of the License, or
-+ (at your option) any later version.
++ msg.tx_id = t.id;
++ msg.req_id = 0;
++ msg.type = type;
++ msg.len = 0;
++ for (i = 0; i < num_vecs; i++)
++ msg.len += iovec[i].iov_len;
+
-+ This program is distributed in the hope that it will be useful,
-+ but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ GNU General Public License for more details.
++ mutex_lock(&xs_state.request_mutex);
+
-+ You should have received a copy of the GNU General Public License
-+ along with this program; if not, write to the Free Software
-+ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
-+*/
-+#include <stdarg.h>
-+#include <linux/module.h>
-+#include <xen/xenbus.h>
-+#include "common.h"
++ err = xb_write(&msg, sizeof(msg));
++ if (err) {
++ mutex_unlock(&xs_state.request_mutex);
++ return ERR_PTR(err);
++ }
+
-+struct backend_info
-+{
-+ struct xenbus_device *dev;
++ for (i = 0; i < num_vecs; i++) {
++ err = xb_write(iovec[i].iov_base, iovec[i].iov_len);;
++ if (err) {
++ mutex_unlock(&xs_state.request_mutex);
++ return ERR_PTR(err);
++ }
++ }
+
-+ /* our communications channel */
-+ tpmif_t *tpmif;
++ ret = read_reply(&msg.type, len);
+
-+ long int frontend_id;
-+ long int instance; // instance of TPM
-+ u8 is_instance_set;// whether instance number has been set
++ mutex_unlock(&xs_state.request_mutex);
+
-+ /* watch front end for changes */
-+ struct xenbus_watch backend_watch;
-+};
++ if (IS_ERR(ret))
++ return ret;
+
-+static void maybe_connect(struct backend_info *be);
-+static void connect(struct backend_info *be);
-+static int connect_ring(struct backend_info *be);
-+static void backend_changed(struct xenbus_watch *watch,
-+ const char **vec, unsigned int len);
-+static void frontend_changed(struct xenbus_device *dev,
-+ enum xenbus_state frontend_state);
++ if (msg.type == XS_ERROR) {
++ err = get_error(ret);
++ kfree(ret);
++ return ERR_PTR(-err);
++ }
+
-+long int tpmback_get_instance(struct backend_info *bi)
-+{
-+ long int res = -1;
-+ if (bi && bi->is_instance_set)
-+ res = bi->instance;
-+ return res;
++ if (msg.type != type) {
++ if (printk_ratelimit())
++ printk(KERN_WARNING
++ "XENBUS unexpected type [%d], expected [%d]\n",
++ msg.type, type);
++ kfree(ret);
++ return ERR_PTR(-EINVAL);
++ }
++ return ret;
+}
+
-+static int tpmback_remove(struct xenbus_device *dev)
++/* Simplified version of xs_talkv: single message. */
++static void *xs_single(struct xenbus_transaction t,
++ enum xsd_sockmsg_type type,
++ const char *string,
++ unsigned int *len)
+{
-+ struct backend_info *be = dev->dev.driver_data;
++ struct kvec iovec;
+
-+ if (!be) return 0;
++ iovec.iov_base = (void *)string;
++ iovec.iov_len = strlen(string) + 1;
++ return xs_talkv(t, type, &iovec, 1, len);
++}
+
-+ if (be->backend_watch.node) {
-+ unregister_xenbus_watch(&be->backend_watch);
-+ kfree(be->backend_watch.node);
-+ be->backend_watch.node = NULL;
-+ }
-+ if (be->tpmif) {
-+ be->tpmif->bi = NULL;
-+ vtpm_release_packets(be->tpmif, 0);
-+ tpmif_put(be->tpmif);
-+ be->tpmif = NULL;
-+ }
-+ kfree(be);
-+ dev->dev.driver_data = NULL;
++/* Many commands only need an ack, don't care what it says. */
++static int xs_error(char *reply)
++{
++ if (IS_ERR(reply))
++ return PTR_ERR(reply);
++ kfree(reply);
+ return 0;
+}
+
-+static int tpmback_probe(struct xenbus_device *dev,
-+ const struct xenbus_device_id *id)
++static unsigned int count_strings(const char *strings, unsigned int len)
+{
-+ int err;
-+ struct backend_info *be = kzalloc(sizeof(struct backend_info),
-+ GFP_KERNEL);
++ unsigned int num;
++ const char *p;
+
-+ if (!be) {
-+ xenbus_dev_fatal(dev, -ENOMEM,
-+ "allocating backend structure");
-+ return -ENOMEM;
-+ }
++ for (p = strings, num = 0; p < strings + len; p += strlen(p) + 1)
++ num++;
+
-+ be->is_instance_set = 0;
-+ be->dev = dev;
-+ dev->dev.driver_data = be;
++ return num;
++}
+
-+ err = xenbus_watch_path2(dev, dev->nodename,
-+ "instance", &be->backend_watch,
-+ backend_changed);
-+ if (err) {
-+ goto fail;
-+ }
++/* Return the path to dir with /name appended. Buffer must be kfree()'ed. */
++static char *join(const char *dir, const char *name)
++{
++ char *buffer;
+
-+ err = xenbus_switch_state(dev, XenbusStateInitWait);
-+ if (err) {
-+ goto fail;
-+ }
-+ return 0;
-+fail:
-+ tpmback_remove(dev);
-+ return err;
++ if (strlen(name) == 0)
++ buffer = kasprintf(GFP_NOIO | __GFP_HIGH, "%s", dir);
++ else
++ buffer = kasprintf(GFP_NOIO | __GFP_HIGH, "%s/%s", dir, name);
++ return (!buffer) ? ERR_PTR(-ENOMEM) : buffer;
+}
+
-+
-+static void backend_changed(struct xenbus_watch *watch,
-+ const char **vec, unsigned int len)
++static char **split(char *strings, unsigned int len, unsigned int *num)
+{
-+ int err;
-+ long instance;
-+ struct backend_info *be
-+ = container_of(watch, struct backend_info, backend_watch);
-+ struct xenbus_device *dev = be->dev;
++ char *p, **ret;
+
-+ err = xenbus_scanf(XBT_NIL, dev->nodename,
-+ "instance","%li", &instance);
-+ if (XENBUS_EXIST_ERR(err)) {
-+ return;
-+ }
++ /* Count the strings. */
++ *num = count_strings(strings, len) + 1;
+
-+ if (err != 1) {
-+ xenbus_dev_fatal(dev, err, "reading instance");
-+ return;
++ /* Transfer to one big alloc for easy freeing. */
++ ret = kmalloc(*num * sizeof(char *) + len, GFP_NOIO | __GFP_HIGH);
++ if (!ret) {
++ kfree(strings);
++ return ERR_PTR(-ENOMEM);
+ }
++ memcpy(&ret[*num], strings, len);
++ kfree(strings);
+
-+ if (be->is_instance_set == 0) {
-+ be->instance = instance;
-+ be->is_instance_set = 1;
-+ }
-+}
++ strings = (char *)&ret[*num];
++ for (p = strings, *num = 0; p < strings + len; p += strlen(p) + 1)
++ ret[(*num)++] = p;
++ ret[*num] = strings + len;
+
++ return ret;
++}
+
-+static void frontend_changed(struct xenbus_device *dev,
-+ enum xenbus_state frontend_state)
++char **xenbus_directory(struct xenbus_transaction t,
++ const char *dir, const char *node, unsigned int *num)
+{
-+ struct backend_info *be = dev->dev.driver_data;
-+ int err;
++ char *strings, *path;
++ unsigned int len;
+
-+ switch (frontend_state) {
-+ case XenbusStateInitialising:
-+ case XenbusStateInitialised:
-+ break;
++ path = join(dir, node);
++ if (IS_ERR(path))
++ return (char **)path;
+
-+ case XenbusStateConnected:
-+ err = connect_ring(be);
-+ if (err) {
-+ return;
-+ }
-+ maybe_connect(be);
-+ break;
++ strings = xs_single(t, XS_DIRECTORY, path, &len);
++ kfree(path);
++ if (IS_ERR(strings))
++ return (char **)strings;
+
-+ case XenbusStateClosing:
-+ be->instance = -1;
-+ xenbus_switch_state(dev, XenbusStateClosing);
-+ break;
++ return split(strings, len, num);
++}
++EXPORT_SYMBOL_GPL(xenbus_directory);
+
-+ case XenbusStateUnknown: /* keep it here */
-+ case XenbusStateClosed:
-+ xenbus_switch_state(dev, XenbusStateClosed);
-+ device_unregister(&be->dev->dev);
-+ tpmback_remove(dev);
-+ break;
++/* Check if a path exists. Return 1 if it does. */
++int xenbus_exists(struct xenbus_transaction t,
++ const char *dir, const char *node)
++{
++ char **d;
++ int dir_n;
+
-+ default:
-+ xenbus_dev_fatal(dev, -EINVAL,
-+ "saw state %d at frontend",
-+ frontend_state);
-+ break;
-+ }
++ d = xenbus_directory(t, dir, node, &dir_n);
++ if (IS_ERR(d))
++ return 0;
++ kfree(d);
++ return 1;
+}
++EXPORT_SYMBOL_GPL(xenbus_exists);
++
++/* Get the value of a single file.
++ * Returns a kmalloced value: call free() on it after use.
++ * len indicates length in bytes.
++ */
++void *xenbus_read(struct xenbus_transaction t,
++ const char *dir, const char *node, unsigned int *len)
++{
++ char *path;
++ void *ret;
+
++ path = join(dir, node);
++ if (IS_ERR(path))
++ return (void *)path;
+
++ ret = xs_single(t, XS_READ, path, len);
++ kfree(path);
++ return ret;
++}
++EXPORT_SYMBOL_GPL(xenbus_read);
+
-+static void maybe_connect(struct backend_info *be)
++/* Write the value of a single file.
++ * Returns -err on failure.
++ */
++int xenbus_write(struct xenbus_transaction t,
++ const char *dir, const char *node, const char *string)
+{
-+ if (be->tpmif == NULL || be->tpmif->status == CONNECTED)
-+ return;
++ const char *path;
++ struct kvec iovec[2];
++ int ret;
+
-+ connect(be);
-+}
++ path = join(dir, node);
++ if (IS_ERR(path))
++ return PTR_ERR(path);
++
++ iovec[0].iov_base = (void *)path;
++ iovec[0].iov_len = strlen(path) + 1;
++ iovec[1].iov_base = (void *)string;
++ iovec[1].iov_len = strlen(string);
+
++ ret = xs_error(xs_talkv(t, XS_WRITE, iovec, ARRAY_SIZE(iovec), NULL));
++ kfree(path);
++ return ret;
++}
++EXPORT_SYMBOL_GPL(xenbus_write);
+
-+static void connect(struct backend_info *be)
++/* Create a new directory. */
++int xenbus_mkdir(struct xenbus_transaction t,
++ const char *dir, const char *node)
+{
-+ struct xenbus_transaction xbt;
-+ int err;
-+ struct xenbus_device *dev = be->dev;
-+ unsigned long ready = 1;
++ char *path;
++ int ret;
+
-+again:
-+ err = xenbus_transaction_start(&xbt);
-+ if (err) {
-+ xenbus_dev_fatal(be->dev, err, "starting transaction");
-+ return;
-+ }
++ path = join(dir, node);
++ if (IS_ERR(path))
++ return PTR_ERR(path);
++
++ ret = xs_error(xs_single(t, XS_MKDIR, path, NULL));
++ kfree(path);
++ return ret;
++}
++EXPORT_SYMBOL_GPL(xenbus_mkdir);
+
-+ err = xenbus_printf(xbt, be->dev->nodename,
-+ "ready", "%lu", ready);
-+ if (err) {
-+ xenbus_dev_fatal(be->dev, err, "writing 'ready'");
-+ goto abort;
-+ }
++/* Destroy a file or directory (directories must be empty). */
++int xenbus_rm(struct xenbus_transaction t, const char *dir, const char *node)
++{
++ char *path;
++ int ret;
+
-+ err = xenbus_transaction_end(xbt, 0);
-+ if (err == -EAGAIN)
-+ goto again;
-+ if (err)
-+ xenbus_dev_fatal(be->dev, err, "end of transaction");
++ path = join(dir, node);
++ if (IS_ERR(path))
++ return PTR_ERR(path);
+
-+ err = xenbus_switch_state(dev, XenbusStateConnected);
-+ if (!err)
-+ be->tpmif->status = CONNECTED;
-+ return;
-+abort:
-+ xenbus_transaction_end(xbt, 1);
++ ret = xs_error(xs_single(t, XS_RM, path, NULL));
++ kfree(path);
++ return ret;
+}
++EXPORT_SYMBOL_GPL(xenbus_rm);
+
-+
-+static int connect_ring(struct backend_info *be)
++/* Start a transaction: changes by others will not be seen during this
++ * transaction, and changes will not be visible to others until end.
++ */
++int xenbus_transaction_start(struct xenbus_transaction *t)
+{
-+ struct xenbus_device *dev = be->dev;
-+ unsigned long ring_ref;
-+ unsigned int evtchn;
-+ int err;
++ char *id_str;
+
-+ err = xenbus_gather(XBT_NIL, dev->otherend,
-+ "ring-ref", "%lu", &ring_ref,
-+ "event-channel", "%u", &evtchn, NULL);
-+ if (err) {
-+ xenbus_dev_error(dev, err,
-+ "reading %s/ring-ref and event-channel",
-+ dev->otherend);
-+ return err;
-+ }
++ down_read(&xs_state.transaction_mutex);
+
-+ if (!be->tpmif) {
-+ be->tpmif = tpmif_find(dev->otherend_id, be);
-+ if (IS_ERR(be->tpmif)) {
-+ err = PTR_ERR(be->tpmif);
-+ be->tpmif = NULL;
-+ xenbus_dev_fatal(dev,err,"creating vtpm interface");
-+ return err;
-+ }
++ id_str = xs_single(XBT_NIL, XS_TRANSACTION_START, "", NULL);
++ if (IS_ERR(id_str)) {
++ up_read(&xs_state.transaction_mutex);
++ return PTR_ERR(id_str);
+ }
+
-+ if (be->tpmif != NULL) {
-+ err = tpmif_map(be->tpmif, ring_ref, evtchn);
-+ if (err) {
-+ xenbus_dev_error(dev, err,
-+ "mapping shared-frame %lu port %u",
-+ ring_ref, evtchn);
-+ return err;
-+ }
-+ }
++ t->id = simple_strtoul(id_str, NULL, 0);
++ kfree(id_str);
+ return 0;
+}
++EXPORT_SYMBOL_GPL(xenbus_transaction_start);
+
++/* End a transaction.
++ * If abandon is true, transaction is discarded instead of committed.
++ */
++int xenbus_transaction_end(struct xenbus_transaction t, int abort)
++{
++ char abortstr[2];
++ int err;
+
-+static struct xenbus_device_id tpmback_ids[] = {
-+ { "vtpm" },
-+ { "" }
-+};
-+
++ if (abort)
++ strcpy(abortstr, "F");
++ else
++ strcpy(abortstr, "T");
+
-+static struct xenbus_driver tpmback = {
-+ .name = "vtpm",
-+ .owner = THIS_MODULE,
-+ .ids = tpmback_ids,
-+ .probe = tpmback_probe,
-+ .remove = tpmback_remove,
-+ .otherend_changed = frontend_changed,
-+};
++ err = xs_error(xs_single(t, XS_TRANSACTION_END, abortstr, NULL));
+
++ up_read(&xs_state.transaction_mutex);
+
-+void tpmif_xenbus_init(void)
-+{
-+ xenbus_register_backend(&tpmback);
++ return err;
+}
++EXPORT_SYMBOL_GPL(xenbus_transaction_end);
+
-+void tpmif_xenbus_exit(void)
++/* Single read and scanf: returns -errno or num scanned. */
++int xenbus_scanf(struct xenbus_transaction t,
++ const char *dir, const char *node, const char *fmt, ...)
+{
-+ xenbus_unregister_driver(&tpmback);
++ va_list ap;
++ int ret;
++ char *val;
++
++ val = xenbus_read(t, dir, node, NULL);
++ if (IS_ERR(val))
++ return PTR_ERR(val);
++
++ va_start(ap, fmt);
++ ret = vsscanf(val, fmt, ap);
++ va_end(ap);
++ kfree(val);
++ /* Distinctive errno. */
++ if (ret == 0)
++ return -ERANGE;
++ return ret;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/util.c tmp-linux-2.6-xen.patch/drivers/xen/util.c
---- pristine-linux-2.6.18.2/drivers/xen/util.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/util.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,70 @@
-+#include <linux/mm.h>
-+#include <linux/module.h>
-+#include <linux/slab.h>
-+#include <linux/vmalloc.h>
-+#include <asm/uaccess.h>
-+#include <xen/driver_util.h>
++EXPORT_SYMBOL_GPL(xenbus_scanf);
+
-+struct class *get_xen_class(void)
++/* Single printf and write: returns -errno or 0. */
++int xenbus_printf(struct xenbus_transaction t,
++ const char *dir, const char *node, const char *fmt, ...)
+{
-+ static struct class *xen_class;
++ va_list ap;
++ int ret;
++#define PRINTF_BUFFER_SIZE 4096
++ char *printf_buffer;
+
-+ if (xen_class)
-+ return xen_class;
++ printf_buffer = kmalloc(PRINTF_BUFFER_SIZE, GFP_NOIO | __GFP_HIGH);
++ if (printf_buffer == NULL)
++ return -ENOMEM;
+
-+ xen_class = class_create(THIS_MODULE, "xen");
-+ if (IS_ERR(xen_class)) {
-+ printk("Failed to create xen sysfs class.\n");
-+ xen_class = NULL;
-+ }
++ va_start(ap, fmt);
++ ret = vsnprintf(printf_buffer, PRINTF_BUFFER_SIZE, fmt, ap);
++ va_end(ap);
+
-+ return xen_class;
-+}
-+EXPORT_SYMBOL_GPL(get_xen_class);
++ BUG_ON(ret > PRINTF_BUFFER_SIZE-1);
++ ret = xenbus_write(t, dir, node, printf_buffer);
+
-+/* Todo: merge ia64 ('auto-translate physmap') versions of these functions. */
-+#ifndef __ia64__
++ kfree(printf_buffer);
+
-+static int f(pte_t *pte, struct page *pmd_page, unsigned long addr, void *data)
-+{
-+ /* apply_to_page_range() does all the hard work. */
-+ return 0;
++ return ret;
+}
++EXPORT_SYMBOL_GPL(xenbus_printf);
+
-+struct vm_struct *alloc_vm_area(unsigned long size)
++/* Takes tuples of names, scanf-style args, and void **, NULL terminated. */
++int xenbus_gather(struct xenbus_transaction t, const char *dir, ...)
+{
-+ struct vm_struct *area;
++ va_list ap;
++ const char *name;
++ int ret = 0;
+
-+ area = get_vm_area(size, VM_IOREMAP);
-+ if (area == NULL)
-+ return NULL;
++ va_start(ap, dir);
++ while (ret == 0 && (name = va_arg(ap, char *)) != NULL) {
++ const char *fmt = va_arg(ap, char *);
++ void *result = va_arg(ap, void *);
++ char *p;
+
-+ /*
-+ * This ensures that page tables are constructed for this region
-+ * of kernel virtual address space and mapped into init_mm.
-+ */
-+ if (apply_to_page_range(&init_mm, (unsigned long)area->addr,
-+ area->size, f, NULL)) {
-+ free_vm_area(area);
-+ return NULL;
++ p = xenbus_read(t, dir, name, NULL);
++ if (IS_ERR(p)) {
++ ret = PTR_ERR(p);
++ break;
++ }
++ if (fmt) {
++ if (sscanf(p, fmt, result) == 0)
++ ret = -EINVAL;
++ kfree(p);
++ } else
++ *(char **)result = p;
+ }
++ va_end(ap);
++ return ret;
++}
++EXPORT_SYMBOL_GPL(xenbus_gather);
+
-+ /* Map page directories into every address space. */
-+#ifdef CONFIG_X86
-+ vmalloc_sync_all();
-+#endif
++static int xs_watch(const char *path, const char *token)
++{
++ struct kvec iov[2];
+
-+ return area;
++ iov[0].iov_base = (void *)path;
++ iov[0].iov_len = strlen(path) + 1;
++ iov[1].iov_base = (void *)token;
++ iov[1].iov_len = strlen(token) + 1;
++
++ return xs_error(xs_talkv(XBT_NIL, XS_WATCH, iov,
++ ARRAY_SIZE(iov), NULL));
+}
-+EXPORT_SYMBOL_GPL(alloc_vm_area);
+
-+void free_vm_area(struct vm_struct *area)
++static int xs_unwatch(const char *path, const char *token)
+{
-+ struct vm_struct *ret;
-+ ret = remove_vm_area(area->addr);
-+ BUG_ON(ret != area);
-+ kfree(area);
++ struct kvec iov[2];
++
++ iov[0].iov_base = (char *)path;
++ iov[0].iov_len = strlen(path) + 1;
++ iov[1].iov_base = (char *)token;
++ iov[1].iov_len = strlen(token) + 1;
++
++ return xs_error(xs_talkv(XBT_NIL, XS_UNWATCH, iov,
++ ARRAY_SIZE(iov), NULL));
+}
-+EXPORT_SYMBOL_GPL(free_vm_area);
+
-+#endif /* !__ia64__ */
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/xenbus/Makefile tmp-linux-2.6-xen.patch/drivers/xen/xenbus/Makefile
---- pristine-linux-2.6.18.2/drivers/xen/xenbus/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/xenbus/Makefile 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,9 @@
-+obj-y += xenbus_client.o xenbus_comms.o xenbus_xs.o xenbus_probe.o
-+obj-$(CONFIG_XEN_BACKEND) += xenbus_be.o
++static struct xenbus_watch *find_watch(const char *token)
++{
++ struct xenbus_watch *i, *cmp;
+
-+xenbus_be-objs =
-+xenbus_be-objs += xenbus_backend_client.o
++ cmp = (void *)simple_strtoul(token, NULL, 16);
+
-+xenbus-$(CONFIG_XEN_BACKEND) += xenbus_probe_backend.o
-+obj-y += $(xenbus-y) $(xenbus-m)
-+obj-$(CONFIG_XEN_XENBUS_DEV) += xenbus_dev.o
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_backend_client.c tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_backend_client.c
---- pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_backend_client.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_backend_client.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,147 @@
-+/******************************************************************************
-+ * Backend-client-facing interface for the Xenbus driver. In other words, the
-+ * interface between the Xenbus and the device-specific code in the backend
-+ * driver.
-+ *
-+ * Copyright (C) 2005-2006 XenSource Ltd
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ list_for_each_entry(i, &watches, list)
++ if (i == cmp)
++ return i;
+
-+#include <linux/err.h>
-+#include <xen/gnttab.h>
-+#include <xen/xenbus.h>
-+#include <xen/driver_util.h>
++ return NULL;
++}
+
-+/* Based on Rusty Russell's skeleton driver's map_page */
-+struct vm_struct *xenbus_map_ring_valloc(struct xenbus_device *dev, int gnt_ref)
++/* Register callback to watch this node. */
++int register_xenbus_watch(struct xenbus_watch *watch)
+{
-+ struct gnttab_map_grant_ref op;
-+ struct vm_struct *area;
++ /* Pointer in ascii is the token. */
++ char token[sizeof(watch) * 2 + 1];
++ int err;
+
-+ area = alloc_vm_area(PAGE_SIZE);
-+ if (!area)
-+ return ERR_PTR(-ENOMEM);
++ sprintf(token, "%lX", (long)watch);
+
-+ gnttab_set_map_op(&op, (unsigned long)area->addr, GNTMAP_host_map,
-+ gnt_ref, dev->otherend_id);
-+
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
-+ BUG();
++ down_read(&xs_state.watch_mutex);
+
-+ if (op.status != GNTST_okay) {
-+ free_vm_area(area);
-+ xenbus_dev_fatal(dev, op.status,
-+ "mapping in shared page %d from domain %d",
-+ gnt_ref, dev->otherend_id);
-+ BUG_ON(!IS_ERR(ERR_PTR(op.status)));
-+ return ERR_PTR(op.status);
++ spin_lock(&watches_lock);
++ BUG_ON(find_watch(token));
++ list_add(&watch->list, &watches);
++ spin_unlock(&watches_lock);
++
++ err = xs_watch(watch->node, token);
++
++ /* Ignore errors due to multiple registration. */
++ if ((err != 0) && (err != -EEXIST)) {
++ spin_lock(&watches_lock);
++ list_del(&watch->list);
++ spin_unlock(&watches_lock);
+ }
+
-+ /* Stuff the handle in an unused field */
-+ area->phys_addr = (unsigned long)op.handle;
++ up_read(&xs_state.watch_mutex);
+
-+ return area;
++ return err;
+}
-+EXPORT_SYMBOL_GPL(xenbus_map_ring_valloc);
-+
++EXPORT_SYMBOL_GPL(register_xenbus_watch);
+
-+int xenbus_map_ring(struct xenbus_device *dev, int gnt_ref,
-+ grant_handle_t *handle, void *vaddr)
++void unregister_xenbus_watch(struct xenbus_watch *watch)
+{
-+ struct gnttab_map_grant_ref op;
-+
-+ gnttab_set_map_op(&op, (unsigned long)vaddr, GNTMAP_host_map,
-+ gnt_ref, dev->otherend_id);
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
-+ BUG();
++ struct xs_stored_msg *msg, *tmp;
++ char token[sizeof(watch) * 2 + 1];
++ int err;
+
-+ if (op.status != GNTST_okay) {
-+ xenbus_dev_fatal(dev, op.status,
-+ "mapping in shared page %d from domain %d",
-+ gnt_ref, dev->otherend_id);
-+ } else
-+ *handle = op.handle;
++ BUG_ON(watch->flags & XBWF_new_thread);
+
-+ return op.status;
-+}
-+EXPORT_SYMBOL_GPL(xenbus_map_ring);
++ sprintf(token, "%lX", (long)watch);
+
++ down_read(&xs_state.watch_mutex);
+
-+/* Based on Rusty Russell's skeleton driver's unmap_page */
-+int xenbus_unmap_ring_vfree(struct xenbus_device *dev, struct vm_struct *area)
-+{
-+ struct gnttab_unmap_grant_ref op;
++ spin_lock(&watches_lock);
++ BUG_ON(!find_watch(token));
++ list_del(&watch->list);
++ spin_unlock(&watches_lock);
+
-+ gnttab_set_unmap_op(&op, (unsigned long)area->addr, GNTMAP_host_map,
-+ (grant_handle_t)area->phys_addr);
++ err = xs_unwatch(watch->node, token);
++ if (err)
++ printk(KERN_WARNING
++ "XENBUS Failed to release watch %s: %i\n",
++ watch->node, err);
+
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
-+ BUG();
++ up_read(&xs_state.watch_mutex);
+
-+ if (op.status == GNTST_okay)
-+ free_vm_area(area);
-+ else
-+ xenbus_dev_error(dev, op.status,
-+ "unmapping page at handle %d error %d",
-+ (int16_t)area->phys_addr, op.status);
++ /* Cancel pending watch events. */
++ spin_lock(&watch_events_lock);
++ list_for_each_entry_safe(msg, tmp, &watch_events, list) {
++ if (msg->u.watch.handle != watch)
++ continue;
++ list_del(&msg->list);
++ kfree(msg->u.watch.vec);
++ kfree(msg);
++ }
++ spin_unlock(&watch_events_lock);
+
-+ return op.status;
++ /* Flush any currently-executing callback, unless we are it. :-) */
++ if (current->pid != xenwatch_pid) {
++ mutex_lock(&xenwatch_mutex);
++ mutex_unlock(&xenwatch_mutex);
++ }
+}
-+EXPORT_SYMBOL_GPL(xenbus_unmap_ring_vfree);
++EXPORT_SYMBOL_GPL(unregister_xenbus_watch);
+
++void xs_suspend(void)
++{
++ down_write(&xs_state.transaction_mutex);
++ down_write(&xs_state.watch_mutex);
++ mutex_lock(&xs_state.request_mutex);
++ mutex_lock(&xs_state.response_mutex);
++}
+
-+int xenbus_unmap_ring(struct xenbus_device *dev,
-+ grant_handle_t handle, void *vaddr)
++void xs_resume(void)
+{
-+ struct gnttab_unmap_grant_ref op;
++ struct xenbus_watch *watch;
++ char token[sizeof(watch) * 2 + 1];
+
-+ gnttab_set_unmap_op(&op, (unsigned long)vaddr, GNTMAP_host_map,
-+ handle);
-+ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
-+ BUG();
++ mutex_unlock(&xs_state.response_mutex);
++ mutex_unlock(&xs_state.request_mutex);
++ up_write(&xs_state.transaction_mutex);
+
-+ if (op.status != GNTST_okay)
-+ xenbus_dev_error(dev, op.status,
-+ "unmapping page at handle %d error %d",
-+ handle, op.status);
++ /* No need for watches_lock: the watch_mutex is sufficient. */
++ list_for_each_entry(watch, &watches, list) {
++ sprintf(token, "%lX", (long)watch);
++ xs_watch(watch->node, token);
++ }
+
-+ return op.status;
++ up_write(&xs_state.watch_mutex);
+}
-+EXPORT_SYMBOL_GPL(xenbus_unmap_ring);
+
-+int xenbus_dev_is_online(struct xenbus_device *dev)
++void xs_suspend_cancel(void)
+{
-+ int rc, val;
++ mutex_unlock(&xs_state.response_mutex);
++ mutex_unlock(&xs_state.request_mutex);
++ up_write(&xs_state.watch_mutex);
++ up_write(&xs_state.transaction_mutex);
++}
+
-+ rc = xenbus_scanf(XBT_NIL, dev->nodename, "online", "%d", &val);
-+ if (rc != 1)
-+ val = 0; /* no online node present */
++static int xenwatch_handle_callback(void *data)
++{
++ struct xs_stored_msg *msg = data;
+
-+ return val;
++ msg->u.watch.handle->callback(msg->u.watch.handle,
++ (const char **)msg->u.watch.vec,
++ msg->u.watch.vec_size);
++
++ kfree(msg->u.watch.vec);
++ kfree(msg);
++
++ /* Kill this kthread if we were spawned just for this callback. */
++ if (current->pid != xenwatch_pid)
++ do_exit(0);
++
++ return 0;
+}
-+EXPORT_SYMBOL_GPL(xenbus_dev_is_online);
+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_client.c tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_client.c
---- pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_client.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_client.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,283 @@
-+/******************************************************************************
-+ * Client-facing interface for the Xenbus driver. In other words, the
-+ * interface between the Xenbus and the device-specific code, be it the
-+ * frontend or the backend of that driver.
-+ *
-+ * Copyright (C) 2005 XenSource Ltd
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++static int xenwatch_thread(void *unused)
++{
++ struct list_head *ent;
++ struct xs_stored_msg *msg;
++
++ current->flags |= PF_NOFREEZE;
++ for (;;) {
++ wait_event_interruptible(watch_events_waitq,
++ !list_empty(&watch_events));
++
++ if (kthread_should_stop())
++ break;
++
++ mutex_lock(&xenwatch_mutex);
++
++ spin_lock(&watch_events_lock);
++ ent = watch_events.next;
++ if (ent != &watch_events)
++ list_del(ent);
++ spin_unlock(&watch_events_lock);
+
-+#include <xen/evtchn.h>
-+#include <xen/gnttab.h>
-+#include <xen/xenbus.h>
-+#include <xen/driver_util.h>
++ if (ent == &watch_events) {
++ mutex_unlock(&xenwatch_mutex);
++ continue;
++ }
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
-+#endif
++ msg = list_entry(ent, struct xs_stored_msg, list);
+
-+#define DPRINTK(fmt, args...) \
-+ pr_debug("xenbus_client (%s:%d) " fmt ".\n", __FUNCTION__, __LINE__, ##args)
++ /*
++ * Unlock the mutex before running an XBWF_new_thread
++ * handler. kthread_run can block which can deadlock
++ * against unregister_xenbus_watch() if we need to
++ * unregister other watches in order to make
++ * progress. This can occur on resume before the swap
++ * device is attached.
++ */
++ if (msg->u.watch.handle->flags & XBWF_new_thread) {
++ mutex_unlock(&xenwatch_mutex);
++ kthread_run(xenwatch_handle_callback,
++ msg, "xenwatch_cb");
++ } else {
++ xenwatch_handle_callback(msg);
++ mutex_unlock(&xenwatch_mutex);
++ }
++ }
+
-+const char *xenbus_strstate(enum xenbus_state state)
-+{
-+ static const char *const name[] = {
-+ [ XenbusStateUnknown ] = "Unknown",
-+ [ XenbusStateInitialising ] = "Initialising",
-+ [ XenbusStateInitWait ] = "InitWait",
-+ [ XenbusStateInitialised ] = "Initialised",
-+ [ XenbusStateConnected ] = "Connected",
-+ [ XenbusStateClosing ] = "Closing",
-+ [ XenbusStateClosed ] = "Closed",
-+ };
-+ return (state < ARRAY_SIZE(name)) ? name[state] : "INVALID";
++ return 0;
+}
-+EXPORT_SYMBOL_GPL(xenbus_strstate);
+
-+int xenbus_watch_path(struct xenbus_device *dev, const char *path,
-+ struct xenbus_watch *watch,
-+ void (*callback)(struct xenbus_watch *,
-+ const char **, unsigned int))
++static int process_msg(void)
+{
++ struct xs_stored_msg *msg;
++ char *body;
+ int err;
+
-+ watch->node = path;
-+ watch->callback = callback;
++ /*
++ * We must disallow save/restore while reading a xenstore message.
++ * A partial read across s/r leaves us out of sync with xenstored.
++ */
++ for (;;) {
++ err = xb_wait_for_data_to_read();
++ if (err)
++ return err;
++ mutex_lock(&xs_state.response_mutex);
++ if (xb_data_to_read())
++ break;
++ /* We raced with save/restore: pending data 'disappeared'. */
++ mutex_unlock(&xs_state.response_mutex);
++ }
+
-+ err = register_xenbus_watch(watch);
+
++ msg = kmalloc(sizeof(*msg), GFP_NOIO | __GFP_HIGH);
++ if (msg == NULL) {
++ err = -ENOMEM;
++ goto out;
++ }
++
++ err = xb_read(&msg->hdr, sizeof(msg->hdr));
+ if (err) {
-+ watch->node = NULL;
-+ watch->callback = NULL;
-+ xenbus_dev_fatal(dev, err, "adding watch on %s", path);
++ kfree(msg);
++ goto out;
+ }
+
-+ return err;
-+}
-+EXPORT_SYMBOL_GPL(xenbus_watch_path);
++ body = kmalloc(msg->hdr.len + 1, GFP_NOIO | __GFP_HIGH);
++ if (body == NULL) {
++ kfree(msg);
++ err = -ENOMEM;
++ goto out;
++ }
+
++ err = xb_read(body, msg->hdr.len);
++ if (err) {
++ kfree(body);
++ kfree(msg);
++ goto out;
++ }
++ body[msg->hdr.len] = '\0';
+
-+int xenbus_watch_path2(struct xenbus_device *dev, const char *path,
-+ const char *path2, struct xenbus_watch *watch,
-+ void (*callback)(struct xenbus_watch *,
-+ const char **, unsigned int))
-+{
-+ int err;
-+ char *state = kasprintf(GFP_KERNEL, "%s/%s", path, path2);
-+ if (!state) {
-+ xenbus_dev_fatal(dev, -ENOMEM, "allocating path for watch");
-+ return -ENOMEM;
++ if (msg->hdr.type == XS_WATCH_EVENT) {
++ msg->u.watch.vec = split(body, msg->hdr.len,
++ &msg->u.watch.vec_size);
++ if (IS_ERR(msg->u.watch.vec)) {
++ err = PTR_ERR(msg->u.watch.vec);
++ kfree(msg);
++ goto out;
++ }
++
++ spin_lock(&watches_lock);
++ msg->u.watch.handle = find_watch(
++ msg->u.watch.vec[XS_WATCH_TOKEN]);
++ if (msg->u.watch.handle != NULL) {
++ spin_lock(&watch_events_lock);
++ list_add_tail(&msg->list, &watch_events);
++ wake_up(&watch_events_waitq);
++ spin_unlock(&watch_events_lock);
++ } else {
++ kfree(msg->u.watch.vec);
++ kfree(msg);
++ }
++ spin_unlock(&watches_lock);
++ } else {
++ msg->u.reply.body = body;
++ spin_lock(&xs_state.reply_lock);
++ list_add_tail(&msg->list, &xs_state.reply_list);
++ spin_unlock(&xs_state.reply_lock);
++ wake_up(&xs_state.reply_waitq);
+ }
-+ err = xenbus_watch_path(dev, state, watch, callback);
+
-+ if (err)
-+ kfree(state);
++ out:
++ mutex_unlock(&xs_state.response_mutex);
+ return err;
+}
-+EXPORT_SYMBOL_GPL(xenbus_watch_path2);
-+
+
-+int xenbus_switch_state(struct xenbus_device *dev, enum xenbus_state state)
++static int xenbus_thread(void *unused)
+{
-+ /* We check whether the state is currently set to the given value, and
-+ if not, then the state is set. We don't want to unconditionally
-+ write the given state, because we don't want to fire watches
-+ unnecessarily. Furthermore, if the node has gone, we don't write
-+ to it, as the device will be tearing down, and we don't want to
-+ resurrect that directory.
++ int err;
+
-+ Note that, because of this cached value of our state, this function
-+ will not work inside a Xenstore transaction (something it was
-+ trying to in the past) because dev->state would not get reset if
-+ the transaction was aborted.
++ current->flags |= PF_NOFREEZE;
++ for (;;) {
++ err = process_msg();
++ if (err)
++ printk(KERN_WARNING "XENBUS error %d while reading "
++ "message\n", err);
++ if (kthread_should_stop())
++ break;
++ }
+
-+ */
++ return 0;
++}
+
-+ int current_state;
++int xs_init(void)
++{
+ int err;
++ struct task_struct *task;
+
-+ if (state == dev->state)
-+ return 0;
++ INIT_LIST_HEAD(&xs_state.reply_list);
++ spin_lock_init(&xs_state.reply_lock);
++ init_waitqueue_head(&xs_state.reply_waitq);
+
-+ err = xenbus_scanf(XBT_NIL, dev->nodename, "state", "%d",
-+ ¤t_state);
-+ if (err != 1)
-+ return 0;
++ mutex_init(&xs_state.request_mutex);
++ mutex_init(&xs_state.response_mutex);
++ init_rwsem(&xs_state.transaction_mutex);
++ init_rwsem(&xs_state.watch_mutex);
+
-+ err = xenbus_printf(XBT_NIL, dev->nodename, "state", "%d", state);
-+ if (err) {
-+ if (state != XenbusStateClosing) /* Avoid looping */
-+ xenbus_dev_fatal(dev, err, "writing new state");
++ /* Initialize the shared memory rings to talk to xenstored */
++ err = xb_init_comms();
++ if (err)
+ return err;
-+ }
+
-+ dev->state = state;
++ task = kthread_run(xenwatch_thread, NULL, "xenwatch");
++ if (IS_ERR(task))
++ return PTR_ERR(task);
++ xenwatch_pid = task->pid;
+
-+ return 0;
-+}
-+EXPORT_SYMBOL_GPL(xenbus_switch_state);
++ task = kthread_run(xenbus_thread, NULL, "xenbus");
++ if (IS_ERR(task))
++ return PTR_ERR(task);
+
-+int xenbus_frontend_closed(struct xenbus_device *dev)
-+{
-+ xenbus_switch_state(dev, XenbusStateClosed);
-+ complete(&dev->down);
+ return 0;
+}
-+EXPORT_SYMBOL_GPL(xenbus_frontend_closed);
-+
+diff -r d894e36cfc30 -r 0aa021803deb drivers/xen/xenoprof/xenoprofile.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/drivers/xen/xenoprof/xenoprofile.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,545 @@
+/**
-+ * Return the path to the error node for the given device, or NULL on failure.
-+ * If the value returned is non-NULL, then it is the caller's to kfree.
++ * @file xenoprofile.c
++ *
++ * @remark Copyright 2002 OProfile authors
++ * @remark Read the file COPYING
++ *
++ * @author John Levon <levon at movementarian.org>
++ *
++ * Modified by Aravind Menon and Jose Renato Santos for Xen
++ * These modifications are:
++ * Copyright (C) 2005 Hewlett-Packard Co.
++ *
++ * Separated out arch-generic part
++ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
+ */
-+static char *error_path(struct xenbus_device *dev)
-+{
-+ return kasprintf(GFP_KERNEL, "error/%s", dev->nodename);
-+}
+
++#include <linux/init.h>
++#include <linux/notifier.h>
++#include <linux/smp.h>
++#include <linux/oprofile.h>
++#include <linux/sysdev.h>
++#include <linux/slab.h>
++#include <linux/interrupt.h>
++#include <linux/vmalloc.h>
++#include <asm/pgtable.h>
++#include <xen/evtchn.h>
++#include <xen/xenoprof.h>
++#include <xen/driver_util.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/xenoprof.h>
++#include "../../../drivers/oprofile/cpu_buffer.h"
++#include "../../../drivers/oprofile/event_buffer.h"
+
-+void _dev_error(struct xenbus_device *dev, int err, const char *fmt,
-+ va_list ap)
-+{
-+ int ret;
-+ unsigned int len;
-+ char *printf_buffer = NULL, *path_buffer = NULL;
++#define MAX_XENOPROF_SAMPLES 16
+
-+#define PRINTF_BUFFER_SIZE 4096
-+ printf_buffer = kmalloc(PRINTF_BUFFER_SIZE, GFP_KERNEL);
-+ if (printf_buffer == NULL)
-+ goto fail;
++/* sample buffers shared with Xen */
++static xenoprof_buf_t *xenoprof_buf[MAX_VIRT_CPUS];
++/* Shared buffer area */
++static struct xenoprof_shared_buffer shared_buffer;
+
-+ len = sprintf(printf_buffer, "%i ", -err);
-+ ret = vsnprintf(printf_buffer+len, PRINTF_BUFFER_SIZE-len, fmt, ap);
++/* Passive sample buffers shared with Xen */
++static xenoprof_buf_t *p_xenoprof_buf[MAX_OPROF_DOMAINS][MAX_VIRT_CPUS];
++/* Passive shared buffer area */
++static struct xenoprof_shared_buffer p_shared_buffer[MAX_OPROF_DOMAINS];
+
-+ BUG_ON(len + ret > PRINTF_BUFFER_SIZE-1);
++static int xenoprof_start(void);
++static void xenoprof_stop(void);
+
-+ dev_err(&dev->dev, "%s\n", printf_buffer);
++static int xenoprof_enabled = 0;
++static int xenoprof_is_primary = 0;
++static int active_defined;
+
-+ path_buffer = error_path(dev);
++extern unsigned long backtrace_depth;
+
-+ if (path_buffer == NULL) {
-+ printk("xenbus: failed to write error node for %s (%s)\n",
-+ dev->nodename, printf_buffer);
-+ goto fail;
-+ }
++/* Number of buffers in shared area (one per VCPU) */
++static int nbuf;
++/* Mappings of VIRQ_XENOPROF to irq number (per cpu) */
++static int ovf_irq[NR_CPUS];
++/* cpu model type string - copied from Xen on XENOPROF_init command */
++static char cpu_type[XENOPROF_CPU_TYPE_SIZE];
+
-+ if (xenbus_write(XBT_NIL, path_buffer, "error", printf_buffer) != 0) {
-+ printk("xenbus: failed to write error node for %s (%s)\n",
-+ dev->nodename, printf_buffer);
-+ goto fail;
-+ }
++#ifdef CONFIG_PM
+
-+fail:
-+ if (printf_buffer)
-+ kfree(printf_buffer);
-+ if (path_buffer)
-+ kfree(path_buffer);
++static int xenoprof_suspend(struct sys_device * dev, pm_message_t state)
++{
++ if (xenoprof_enabled == 1)
++ xenoprof_stop();
++ return 0;
+}
+
+
-+void xenbus_dev_error(struct xenbus_device *dev, int err, const char *fmt,
-+ ...)
++static int xenoprof_resume(struct sys_device * dev)
+{
-+ va_list ap;
-+
-+ va_start(ap, fmt);
-+ _dev_error(dev, err, fmt, ap);
-+ va_end(ap);
++ if (xenoprof_enabled == 1)
++ xenoprof_start();
++ return 0;
+}
-+EXPORT_SYMBOL_GPL(xenbus_dev_error);
+
+
-+void xenbus_dev_fatal(struct xenbus_device *dev, int err, const char *fmt,
-+ ...)
-+{
-+ va_list ap;
++static struct sysdev_class oprofile_sysclass = {
++ set_kset_name("oprofile"),
++ .resume = xenoprof_resume,
++ .suspend = xenoprof_suspend
++};
+
-+ va_start(ap, fmt);
-+ _dev_error(dev, err, fmt, ap);
-+ va_end(ap);
+
-+ xenbus_switch_state(dev, XenbusStateClosing);
++static struct sys_device device_oprofile = {
++ .id = 0,
++ .cls = &oprofile_sysclass,
++};
++
++
++static int __init init_driverfs(void)
++{
++ int error;
++ if (!(error = sysdev_class_register(&oprofile_sysclass)))
++ error = sysdev_register(&device_oprofile);
++ return error;
+}
-+EXPORT_SYMBOL_GPL(xenbus_dev_fatal);
+
+
-+int xenbus_grant_ring(struct xenbus_device *dev, unsigned long ring_mfn)
++static void exit_driverfs(void)
+{
-+ int err = gnttab_grant_foreign_access(dev->otherend_id, ring_mfn, 0);
-+ if (err < 0)
-+ xenbus_dev_fatal(dev, err, "granting access to ring page");
-+ return err;
++ sysdev_unregister(&device_oprofile);
++ sysdev_class_unregister(&oprofile_sysclass);
+}
-+EXPORT_SYMBOL_GPL(xenbus_grant_ring);
+
++#else
++#define init_driverfs() do { } while (0)
++#define exit_driverfs() do { } while (0)
++#endif /* CONFIG_PM */
++
++static unsigned long long oprofile_samples;
++static unsigned long long p_oprofile_samples;
+
-+int xenbus_alloc_evtchn(struct xenbus_device *dev, int *port)
++static unsigned int pdomains;
++static struct xenoprof_passive passive_domains[MAX_OPROF_DOMAINS];
++
++/* Check whether the given entry is an escape code */
++static int xenoprof_is_escape(xenoprof_buf_t * buf, int tail)
+{
-+ struct evtchn_alloc_unbound alloc_unbound;
-+ int err;
++ return (buf->event_log[tail].eip == XENOPROF_ESCAPE_CODE);
++}
+
-+ alloc_unbound.dom = DOMID_SELF;
-+ alloc_unbound.remote_dom = dev->otherend_id;
++/* Get the event at the given entry */
++static uint8_t xenoprof_get_event(xenoprof_buf_t * buf, int tail)
++{
++ return (buf->event_log[tail].event);
++}
+
-+ err = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
-+ &alloc_unbound);
-+ if (err)
-+ xenbus_dev_fatal(dev, err, "allocating event channel");
-+ else
-+ *port = alloc_unbound.port;
++static void xenoprof_add_pc(xenoprof_buf_t *buf, int is_passive)
++{
++ int head, tail, size;
++ int tracing = 0;
+
-+ return err;
++ head = buf->event_head;
++ tail = buf->event_tail;
++ size = buf->event_size;
++
++ while (tail != head) {
++ if (xenoprof_is_escape(buf, tail) &&
++ xenoprof_get_event(buf, tail) == XENOPROF_TRACE_BEGIN) {
++ tracing=1;
++ oprofile_add_pc(ESCAPE_CODE, buf->event_log[tail].mode,
++ CPU_TRACE_BEGIN);
++ if (!is_passive)
++ oprofile_samples++;
++ else
++ p_oprofile_samples++;
++
++ } else {
++ oprofile_add_pc(buf->event_log[tail].eip,
++ buf->event_log[tail].mode,
++ buf->event_log[tail].event);
++ if (!tracing) {
++ if (!is_passive)
++ oprofile_samples++;
++ else
++ p_oprofile_samples++;
++ }
++
++ }
++ tail++;
++ if(tail==size)
++ tail=0;
++ }
++ buf->event_tail = tail;
+}
-+EXPORT_SYMBOL_GPL(xenbus_alloc_evtchn);
+
++static void xenoprof_handle_passive(void)
++{
++ int i, j;
++ int flag_domain, flag_switch = 0;
++
++ for (i = 0; i < pdomains; i++) {
++ flag_domain = 0;
++ for (j = 0; j < passive_domains[i].nbuf; j++) {
++ xenoprof_buf_t *buf = p_xenoprof_buf[i][j];
++ if (buf->event_head == buf->event_tail)
++ continue;
++ if (!flag_domain) {
++ if (!oprofile_add_domain_switch(
++ passive_domains[i].domain_id))
++ goto done;
++ flag_domain = 1;
++ }
++ xenoprof_add_pc(buf, 1);
++ flag_switch = 1;
++ }
++ }
++done:
++ if (flag_switch)
++ oprofile_add_domain_switch(COORDINATOR_DOMAIN);
++}
+
-+int xenbus_free_evtchn(struct xenbus_device *dev, int port)
++static irqreturn_t
++xenoprof_ovf_interrupt(int irq, void * dev_id, struct pt_regs * regs)
+{
-+ struct evtchn_close close;
-+ int err;
++ struct xenoprof_buf * buf;
++ static unsigned long flag;
+
-+ close.port = port;
++ buf = xenoprof_buf[smp_processor_id()];
+
-+ err = HYPERVISOR_event_channel_op(EVTCHNOP_close, &close);
-+ if (err)
-+ xenbus_dev_error(dev, err, "freeing event channel %d", port);
++ xenoprof_add_pc(buf, 0);
+
-+ return err;
++ if (xenoprof_is_primary && !test_and_set_bit(0, &flag)) {
++ xenoprof_handle_passive();
++ smp_mb__before_clear_bit();
++ clear_bit(0, &flag);
++ }
++
++ return IRQ_HANDLED;
+}
-+EXPORT_SYMBOL_GPL(xenbus_free_evtchn);
+
+
-+enum xenbus_state xenbus_read_driver_state(const char *path)
++static void unbind_virq(void)
+{
-+ enum xenbus_state result;
-+ int err = xenbus_gather(XBT_NIL, path, "state", "%d", &result, NULL);
-+ if (err)
-+ result = XenbusStateUnknown;
++ unsigned int i;
+
-+ return result;
++ for_each_online_cpu(i) {
++ if (ovf_irq[i] >= 0) {
++ unbind_from_irqhandler(ovf_irq[i], NULL);
++ ovf_irq[i] = -1;
++ }
++ }
+}
-+EXPORT_SYMBOL_GPL(xenbus_read_driver_state);
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_comms.c tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_comms.c
---- pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_comms.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_comms.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,232 @@
-+/******************************************************************************
-+ * xenbus_comms.c
-+ *
-+ * Low level code to talks to Xen Store: ringbuffer and event channel.
-+ *
-+ * Copyright (C) 2005 Rusty Russell, IBM Corporation
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
+
-+#include <linux/wait.h>
-+#include <linux/interrupt.h>
-+#include <linux/sched.h>
-+#include <linux/err.h>
-+#include <linux/ptrace.h>
-+#include <xen/evtchn.h>
-+#include <xen/xenbus.h>
+
-+#include <asm/hypervisor.h>
++static int bind_virq(void)
++{
++ unsigned int i;
++ int result;
+
-+#include "xenbus_comms.h"
++ for_each_online_cpu(i) {
++ result = bind_virq_to_irqhandler(VIRQ_XENOPROF,
++ i,
++ xenoprof_ovf_interrupt,
++ SA_INTERRUPT,
++ "xenoprof",
++ NULL);
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
-+#endif
++ if (result < 0) {
++ unbind_virq();
++ return result;
++ }
+
-+static int xenbus_irq;
++ ovf_irq[i] = result;
++ }
++
++ return 0;
++}
+
-+extern void xenbus_probe(void *);
-+extern int xenstored_ready;
-+static DECLARE_WORK(probe_work, xenbus_probe, NULL);
+
-+static DECLARE_WAIT_QUEUE_HEAD(xb_waitq);
++static void unmap_passive_list(void)
++{
++ int i;
++ for (i = 0; i < pdomains; i++)
++ xenoprof_arch_unmap_shared_buffer(&p_shared_buffer[i]);
++ pdomains = 0;
++}
++
++
++static int map_xenoprof_buffer(int max_samples)
++{
++ struct xenoprof_get_buffer get_buffer;
++ struct xenoprof_buf *buf;
++ int ret, i;
++
++ if ( shared_buffer.buffer )
++ return 0;
++
++ get_buffer.max_samples = max_samples;
++ ret = xenoprof_arch_map_shared_buffer(&get_buffer, &shared_buffer);
++ if (ret)
++ return ret;
++ nbuf = get_buffer.nbuf;
+
-+static irqreturn_t wake_waiting(int irq, void *unused, struct pt_regs *regs)
-+{
-+ if (unlikely(xenstored_ready == 0)) {
-+ xenstored_ready = 1;
-+ schedule_work(&probe_work);
++ for (i=0; i< nbuf; i++) {
++ buf = (struct xenoprof_buf*)
++ &shared_buffer.buffer[i * get_buffer.bufsize];
++ BUG_ON(buf->vcpu_id >= MAX_VIRT_CPUS);
++ xenoprof_buf[buf->vcpu_id] = buf;
+ }
+
-+ wake_up(&xb_waitq);
-+ return IRQ_HANDLED;
-+}
-+
-+static int check_indexes(XENSTORE_RING_IDX cons, XENSTORE_RING_IDX prod)
-+{
-+ return ((prod - cons) <= XENSTORE_RING_SIZE);
++ return 0;
+}
+
-+static void *get_output_chunk(XENSTORE_RING_IDX cons,
-+ XENSTORE_RING_IDX prod,
-+ char *buf, uint32_t *len)
-+{
-+ *len = XENSTORE_RING_SIZE - MASK_XENSTORE_IDX(prod);
-+ if ((XENSTORE_RING_SIZE - (prod - cons)) < *len)
-+ *len = XENSTORE_RING_SIZE - (prod - cons);
-+ return buf + MASK_XENSTORE_IDX(prod);
-+}
+
-+static const void *get_input_chunk(XENSTORE_RING_IDX cons,
-+ XENSTORE_RING_IDX prod,
-+ const char *buf, uint32_t *len)
++static int xenoprof_setup(void)
+{
-+ *len = XENSTORE_RING_SIZE - MASK_XENSTORE_IDX(cons);
-+ if ((prod - cons) < *len)
-+ *len = prod - cons;
-+ return buf + MASK_XENSTORE_IDX(cons);
-+}
++ int ret;
+
-+int xb_write(const void *data, unsigned len)
-+{
-+ struct xenstore_domain_interface *intf = xen_store_interface;
-+ XENSTORE_RING_IDX cons, prod;
-+ int rc;
++ if ( (ret = map_xenoprof_buffer(MAX_XENOPROF_SAMPLES)) )
++ return ret;
+
-+ while (len != 0) {
-+ void *dst;
-+ unsigned int avail;
++ if ( (ret = bind_virq()) )
++ return ret;
+
-+ rc = wait_event_interruptible(
-+ xb_waitq,
-+ (intf->req_prod - intf->req_cons) !=
-+ XENSTORE_RING_SIZE);
-+ if (rc < 0)
-+ return rc;
++ if (xenoprof_is_primary) {
++ /* Define dom0 as an active domain if not done yet */
++ if (!active_defined) {
++ domid_t domid;
++ ret = HYPERVISOR_xenoprof_op(
++ XENOPROF_reset_active_list, NULL);
++ if (ret)
++ goto err;
++ domid = 0;
++ ret = HYPERVISOR_xenoprof_op(
++ XENOPROF_set_active, &domid);
++ if (ret)
++ goto err;
++ active_defined = 1;
++ }
+
-+ /* Read indexes, then verify. */
-+ cons = intf->req_cons;
-+ prod = intf->req_prod;
-+ if (!check_indexes(cons, prod)) {
-+ intf->req_cons = intf->req_prod = 0;
-+ return -EIO;
++ if (backtrace_depth > 0) {
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_set_backtrace,
++ &backtrace_depth);
++ if (ret)
++ backtrace_depth = 0;
+ }
+
-+ dst = get_output_chunk(cons, prod, intf->req, &avail);
-+ if (avail == 0)
-+ continue;
-+ if (avail > len)
-+ avail = len;
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_reserve_counters, NULL);
++ if (ret)
++ goto err;
++
++ xenoprof_arch_counter();
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_setup_events, NULL);
++ if (ret)
++ goto err;
++ }
+
-+ /* Must write data /after/ reading the consumer index. */
-+ mb();
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_enable_virq, NULL);
++ if (ret)
++ goto err;
+
-+ memcpy(dst, data, avail);
-+ data += avail;
-+ len -= avail;
++ xenoprof_enabled = 1;
++ return 0;
++ err:
++ unbind_virq();
++ return ret;
++}
+
-+ /* Other side must not see new producer until data is there. */
-+ wmb();
-+ intf->req_prod += avail;
+
-+ /* Implies mb(): other side will see the updated producer. */
-+ notify_remote_via_evtchn(xen_store_evtchn);
++static void xenoprof_shutdown(void)
++{
++ xenoprof_enabled = 0;
++
++ WARN_ON(HYPERVISOR_xenoprof_op(XENOPROF_disable_virq, NULL));
++
++ if (xenoprof_is_primary) {
++ WARN_ON(HYPERVISOR_xenoprof_op(XENOPROF_release_counters,
++ NULL));
++ active_defined = 0;
+ }
+
-+ return 0;
++ unbind_virq();
++
++ xenoprof_arch_unmap_shared_buffer(&shared_buffer);
++ if (xenoprof_is_primary)
++ unmap_passive_list();
+}
+
-+int xb_data_to_read(void)
++
++static int xenoprof_start(void)
+{
-+ struct xenstore_domain_interface *intf = xen_store_interface;
-+ return (intf->rsp_cons != intf->rsp_prod);
++ int ret = 0;
++
++ if (xenoprof_is_primary)
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_start, NULL);
++ if (!ret)
++ xenoprof_arch_start();
++ return ret;
+}
+
-+int xb_wait_for_data_to_read(void)
++
++static void xenoprof_stop(void)
+{
-+ return wait_event_interruptible(xb_waitq, xb_data_to_read());
++ if (xenoprof_is_primary)
++ WARN_ON(HYPERVISOR_xenoprof_op(XENOPROF_stop, NULL));
++ xenoprof_arch_stop();
+}
+
-+int xb_read(void *data, unsigned len)
++
++static int xenoprof_set_active(int * active_domains,
++ unsigned int adomains)
+{
-+ struct xenstore_domain_interface *intf = xen_store_interface;
-+ XENSTORE_RING_IDX cons, prod;
-+ int rc;
++ int ret = 0;
++ int i;
++ int set_dom0 = 0;
++ domid_t domid;
+
-+ while (len != 0) {
-+ unsigned int avail;
-+ const char *src;
++ if (!xenoprof_is_primary)
++ return 0;
+
-+ rc = xb_wait_for_data_to_read();
-+ if (rc < 0)
-+ return rc;
++ if (adomains > MAX_OPROF_DOMAINS)
++ return -E2BIG;
+
-+ /* Read indexes, then verify. */
-+ cons = intf->rsp_cons;
-+ prod = intf->rsp_prod;
-+ if (!check_indexes(cons, prod)) {
-+ intf->rsp_cons = intf->rsp_prod = 0;
-+ return -EIO;
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_reset_active_list, NULL);
++ if (ret)
++ return ret;
++
++ for (i=0; i<adomains; i++) {
++ domid = active_domains[i];
++ if (domid != active_domains[i]) {
++ ret = -EINVAL;
++ goto out;
+ }
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_set_active, &domid);
++ if (ret)
++ goto out;
++ if (active_domains[i] == 0)
++ set_dom0 = 1;
++ }
++ /* dom0 must always be active but may not be in the list */
++ if (!set_dom0) {
++ domid = 0;
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_set_active, &domid);
++ }
+
-+ src = get_input_chunk(cons, prod, intf->rsp, &avail);
-+ if (avail == 0)
-+ continue;
-+ if (avail > len)
-+ avail = len;
++out:
++ if (ret)
++ WARN_ON(HYPERVISOR_xenoprof_op(XENOPROF_reset_active_list,
++ NULL));
++ active_defined = !ret;
++ return ret;
++}
+
-+ /* Must read data /after/ reading the producer index. */
-+ rmb();
++static int xenoprof_set_passive(int * p_domains,
++ unsigned int pdoms)
++{
++ int ret;
++ unsigned int i, j;
++ struct xenoprof_buf *buf;
+
-+ memcpy(data, src, avail);
-+ data += avail;
-+ len -= avail;
++ if (!xenoprof_is_primary)
++ return 0;
+
-+ /* Other side must not see free space until we've copied out */
-+ mb();
-+ intf->rsp_cons += avail;
++ if (pdoms > MAX_OPROF_DOMAINS)
++ return -E2BIG;
+
-+ pr_debug("Finished read of %i bytes (%i to go)\n", avail, len);
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_reset_passive_list, NULL);
++ if (ret)
++ return ret;
++ unmap_passive_list();
+
-+ /* Implies mb(): other side will see the updated consumer. */
-+ notify_remote_via_evtchn(xen_store_evtchn);
++ for (i = 0; i < pdoms; i++) {
++ passive_domains[i].domain_id = p_domains[i];
++ passive_domains[i].max_samples = 2048;
++ ret = xenoprof_arch_set_passive(&passive_domains[i],
++ &p_shared_buffer[i]);
++ if (ret)
++ goto out;
++ for (j = 0; j < passive_domains[i].nbuf; j++) {
++ buf = (struct xenoprof_buf *)
++ &p_shared_buffer[i].buffer[
++ j * passive_domains[i].bufsize];
++ BUG_ON(buf->vcpu_id >= MAX_VIRT_CPUS);
++ p_xenoprof_buf[i][buf->vcpu_id] = buf;
++ }
+ }
+
++ pdomains = pdoms;
+ return 0;
++
++out:
++ for (j = 0; j < i; j++)
++ xenoprof_arch_unmap_shared_buffer(&p_shared_buffer[i]);
++
++ return ret;
+}
+
-+/* Set up interrupt handler off store event channel. */
-+int xb_init_comms(void)
++
++/* The dummy backtrace function to keep oprofile happy
++ * The real backtrace is done in xen
++ */
++static void xenoprof_dummy_backtrace(struct pt_regs * const regs,
++ unsigned int depth)
+{
-+ struct xenstore_domain_interface *intf = xen_store_interface;
-+ int err;
++ /* this should never be called */
++ BUG();
++ return;
++}
+
-+ if (intf->req_prod != intf->req_cons)
-+ printk(KERN_ERR "XENBUS request ring is not quiescent "
-+ "(%08x:%08x)!\n", intf->req_cons, intf->req_prod);
+
-+ if (intf->rsp_prod != intf->rsp_cons) {
-+ printk(KERN_WARNING "XENBUS response ring is not quiescent "
-+ "(%08x:%08x): fixing up\n",
-+ intf->rsp_cons, intf->rsp_prod);
-+ intf->rsp_cons = intf->rsp_prod;
-+ }
++static struct oprofile_operations xenoprof_ops = {
++#ifdef HAVE_XENOPROF_CREATE_FILES
++ .create_files = xenoprof_create_files,
++#endif
++ .set_active = xenoprof_set_active,
++ .set_passive = xenoprof_set_passive,
++ .setup = xenoprof_setup,
++ .shutdown = xenoprof_shutdown,
++ .start = xenoprof_start,
++ .stop = xenoprof_stop,
++ .backtrace = xenoprof_dummy_backtrace
++};
+
-+ if (xenbus_irq)
-+ unbind_from_irqhandler(xenbus_irq, &xb_waitq);
+
-+ err = bind_caller_port_to_irqhandler(
-+ xen_store_evtchn, wake_waiting,
-+ 0, "xenbus", &xb_waitq);
-+ if (err <= 0) {
-+ printk(KERN_ERR "XENBUS request irq failed %i\n", err);
-+ return err;
-+ }
++/* in order to get driverfs right */
++static int using_xenoprof;
+
-+ xenbus_irq = err;
++int __init xenoprofile_init(struct oprofile_operations * ops)
++{
++ struct xenoprof_init init;
++ unsigned int i;
++ int ret;
+
-+ return 0;
-+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_comms.h tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_comms.h
---- pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_comms.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_comms.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,46 @@
-+/*
-+ * Private include for xenbus communications.
-+ *
-+ * Copyright (C) 2005 Rusty Russell, IBM Corporation
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_init, &init);
++ if (!ret) {
++ xenoprof_arch_init_counter(&init);
++ xenoprof_is_primary = init.is_primary;
+
-+#ifndef _XENBUS_COMMS_H
-+#define _XENBUS_COMMS_H
++ /* cpu_type is detected by Xen */
++ cpu_type[XENOPROF_CPU_TYPE_SIZE-1] = 0;
++ strncpy(cpu_type, init.cpu_type, XENOPROF_CPU_TYPE_SIZE - 1);
++ xenoprof_ops.cpu_type = cpu_type;
+
-+int xs_init(void);
-+int xb_init_comms(void);
++ init_driverfs();
++ using_xenoprof = 1;
++ *ops = xenoprof_ops;
+
-+/* Low level routines. */
-+int xb_write(const void *data, unsigned len);
-+int xb_read(void *data, unsigned len);
-+int xb_data_to_read(void);
-+int xb_wait_for_data_to_read(void);
-+int xs_input_avail(void);
-+extern struct xenstore_domain_interface *xen_store_interface;
-+extern int xen_store_evtchn;
++ for (i=0; i<NR_CPUS; i++)
++ ovf_irq[i] = -1;
+
-+#endif /* _XENBUS_COMMS_H */
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_dev.c tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_dev.c
---- pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_dev.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_dev.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,404 @@
-+/*
-+ * xenbus_dev.c
-+ *
-+ * Driver giving user-space access to the kernel's xenbus connection
-+ * to xenstore.
-+ *
-+ * Copyright (c) 2005, Christian Limpach
-+ * Copyright (c) 2005, Rusty Russell, IBM Corporation
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ active_defined = 0;
++ }
+
-+#include <linux/kernel.h>
-+#include <linux/errno.h>
-+#include <linux/uio.h>
-+#include <linux/notifier.h>
-+#include <linux/wait.h>
-+#include <linux/fs.h>
-+#include <linux/poll.h>
-+#include <linux/mutex.h>
++ printk(KERN_INFO "%s: ret %d, events %d, xenoprof_is_primary %d\n",
++ __func__, ret, init.num_events, xenoprof_is_primary);
++ return ret;
++}
+
-+#include "xenbus_comms.h"
+
-+#include <asm/uaccess.h>
-+#include <asm/hypervisor.h>
-+#include <xen/xenbus.h>
-+#include <xen/xen_proc.h>
-+#include <asm/hypervisor.h>
++void xenoprofile_exit(void)
++{
++ if (using_xenoprof)
++ exit_driverfs();
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
++ xenoprof_arch_unmap_shared_buffer(&shared_buffer);
++ if (xenoprof_is_primary) {
++ unmap_passive_list();
++ WARN_ON(HYPERVISOR_xenoprof_op(XENOPROF_shutdown, NULL));
++ }
++}
+diff -r d894e36cfc30 -r 0aa021803deb fs/Kconfig
+--- a/fs/Kconfig Tue Sep 09 11:37:38 2008 +0200
++++ b/fs/Kconfig Wed Sep 10 10:54:08 2008 +0100
+@@ -826,6 +826,11 @@
+ help
+ Exports the dump image of crashed kernel in ELF format.
+
++config PROC_IOMEM_MACHINE
++ bool
++ depends on PROC_FS && EXPERIMENTAL && KEXEC && XEN && IA64
++ default y
++
+ config SYSFS
+ bool "sysfs file system support" if EMBEDDED
+ default y
+@@ -865,6 +870,7 @@
+ config HUGETLBFS
+ bool "HugeTLB file system support"
+ depends X86 || IA64 || PPC64 || SPARC64 || SUPERH || BROKEN
++ depends on !XEN
+ help
+ hugetlbfs is a filesystem backing for HugeTLB pages, based on
+ ramfs. For architectures that support it, say Y here and read
+diff -r d894e36cfc30 -r 0aa021803deb fs/aio.c
+--- a/fs/aio.c Tue Sep 09 11:37:38 2008 +0200
++++ b/fs/aio.c Wed Sep 10 10:54:08 2008 +0100
+@@ -33,6 +33,11 @@
+ #include <asm/kmap_types.h>
+ #include <asm/uaccess.h>
+ #include <asm/mmu_context.h>
++
++#ifdef CONFIG_EPOLL
++#include <linux/poll.h>
++#include <linux/eventpoll.h>
++#endif
+
+ #if DEBUG > 1
+ #define dprintk printk
+@@ -1015,6 +1020,10 @@
+ if (waitqueue_active(&ctx->wait))
+ wake_up(&ctx->wait);
+
++#ifdef CONFIG_EPOLL
++ if (ctx->file && waitqueue_active(&ctx->poll_wait))
++ wake_up(&ctx->poll_wait);
++#endif
+ if (ret)
+ put_ioctx(ctx);
+
+@@ -1024,6 +1033,8 @@
+ /* aio_read_evt
+ * Pull an event off of the ioctx's event ring. Returns the number of
+ * events fetched (0 or 1 ;-)
++ * If ent parameter is 0, just returns the number of events that would
++ * be fetched.
+ * FIXME: make this use cmpxchg.
+ * TODO: make the ringbuffer user mmap()able (requires FIXME).
+ */
+@@ -1046,13 +1057,18 @@
+
+ head = ring->head % info->nr;
+ if (head != ring->tail) {
+- struct io_event *evp = aio_ring_event(info, head, KM_USER1);
+- *ent = *evp;
+- head = (head + 1) % info->nr;
+- smp_mb(); /* finish reading the event before updatng the head */
+- ring->head = head;
+- ret = 1;
+- put_aio_ring_event(evp, KM_USER1);
++ if (ent) { /* event requested */
++ struct io_event *evp =
++ aio_ring_event(info, head, KM_USER1);
++ *ent = *evp;
++ head = (head + 1) % info->nr;
++ /* finish reading the event before updatng the head */
++ smp_mb();
++ ring->head = head;
++ ret = 1;
++ put_aio_ring_event(evp, KM_USER1);
++ } else /* only need to know availability */
++ ret = 1;
+ }
+ spin_unlock(&info->ring_lock);
+
+@@ -1235,8 +1251,77 @@
+
+ aio_cancel_all(ioctx);
+ wait_for_all_aios(ioctx);
++#ifdef CONFIG_EPOLL
++ /* forget the poll file, but it's up to the user to close it */
++ if (ioctx->file) {
++ ioctx->file->private_data = 0;
++ ioctx->file = 0;
++ }
+#endif
+ put_ioctx(ioctx); /* once for the lookup */
+ }
+
-+struct xenbus_dev_transaction {
-+ struct list_head list;
-+ struct xenbus_transaction handle;
-+};
++#ifdef CONFIG_EPOLL
+
-+struct read_buffer {
-+ struct list_head list;
-+ unsigned int cons;
-+ unsigned int len;
-+ char msg[];
-+};
++static int aio_queue_fd_close(struct inode *inode, struct file *file)
++{
++ struct kioctx *ioctx = file->private_data;
++ if (ioctx) {
++ file->private_data = 0;
++ spin_lock_irq(&ioctx->ctx_lock);
++ ioctx->file = 0;
++ spin_unlock_irq(&ioctx->ctx_lock);
++ }
++ return 0;
++}
+
-+struct xenbus_dev_data {
-+ /* In-progress transaction. */
-+ struct list_head transactions;
++static unsigned int aio_queue_fd_poll(struct file *file, poll_table *wait)
++{ unsigned int pollflags = 0;
++ struct kioctx *ioctx = file->private_data;
+
-+ /* Active watches. */
-+ struct list_head watches;
++ if (ioctx) {
+
-+ /* Partial request. */
-+ unsigned int len;
-+ union {
-+ struct xsd_sockmsg msg;
-+ char buffer[PAGE_SIZE];
-+ } u;
++ spin_lock_irq(&ioctx->ctx_lock);
++ /* Insert inside our poll wait queue */
++ poll_wait(file, &ioctx->poll_wait, wait);
+
-+ /* Response queue. */
-+ struct list_head read_buffers;
-+ wait_queue_head_t read_waitq;
++ /* Check our condition */
++ if (aio_read_evt(ioctx, 0))
++ pollflags = POLLIN | POLLRDNORM;
++ spin_unlock_irq(&ioctx->ctx_lock);
++ }
+
-+ struct mutex reply_mutex;
++ return pollflags;
++}
++
++static const struct file_operations aioq_fops = {
++ .release = aio_queue_fd_close,
++ .poll = aio_queue_fd_poll
+};
+
-+static struct proc_dir_entry *xenbus_dev_intf;
++/* make_aio_fd:
++ * Create a file descriptor that can be used to poll the event queue.
++ * Based and piggybacked on the excellent epoll code.
++ */
+
-+static ssize_t xenbus_dev_read(struct file *filp,
-+ char __user *ubuf,
-+ size_t len, loff_t *ppos)
++static int make_aio_fd(struct kioctx *ioctx)
+{
-+ struct xenbus_dev_data *u = filp->private_data;
-+ struct read_buffer *rb;
-+ int i, ret;
-+
-+ mutex_lock(&u->reply_mutex);
-+ while (list_empty(&u->read_buffers)) {
-+ mutex_unlock(&u->reply_mutex);
-+ ret = wait_event_interruptible(u->read_waitq,
-+ !list_empty(&u->read_buffers));
-+ if (ret)
-+ return ret;
-+ mutex_lock(&u->reply_mutex);
-+ }
++ int error, fd;
++ struct inode *inode;
++ struct file *file;
+
-+ rb = list_entry(u->read_buffers.next, struct read_buffer, list);
-+ for (i = 0; i < len;) {
-+ put_user(rb->msg[rb->cons], ubuf + i);
-+ i++;
-+ rb->cons++;
-+ if (rb->cons == rb->len) {
-+ list_del(&rb->list);
-+ kfree(rb);
-+ if (list_empty(&u->read_buffers))
-+ break;
-+ rb = list_entry(u->read_buffers.next,
-+ struct read_buffer, list);
-+ }
-+ }
-+ mutex_unlock(&u->reply_mutex);
++ error = ep_getfd(&fd, &inode, &file, NULL, &aioq_fops);
++ if (error)
++ return error;
+
-+ return i;
++ /* associate the file with the IO context */
++ file->private_data = ioctx;
++ ioctx->file = file;
++ init_waitqueue_head(&ioctx->poll_wait);
++ return fd;
+}
++#endif
+
-+static void queue_reply(struct xenbus_dev_data *u,
-+ char *data, unsigned int len)
-+{
-+ struct read_buffer *rb;
+
+ /* sys_io_setup:
+ * Create an aio_context capable of receiving at least nr_events.
+@@ -1250,18 +1335,30 @@
+ * resources are available. May fail with -EFAULT if an invalid
+ * pointer is passed for ctxp. Will fail with -ENOSYS if not
+ * implemented.
++ *
++ * To request a selectable fd, the user context has to be initialized
++ * to 1, instead of 0, and the return value is the fd.
++ * This keeps the system call compatible, since a non-zero value
++ * was not allowed so far.
+ */
+ asmlinkage long sys_io_setup(unsigned nr_events, aio_context_t __user *ctxp)
+ {
+ struct kioctx *ioctx = NULL;
+ unsigned long ctx;
+ long ret;
++ int make_fd = 0;
+
+ ret = get_user(ctx, ctxp);
+ if (unlikely(ret))
+ goto out;
+
+ ret = -EINVAL;
++#ifdef CONFIG_EPOLL
++ if (ctx == 1) {
++ make_fd = 1;
++ ctx = 0;
++ }
++#endif
+ if (unlikely(ctx || nr_events == 0)) {
+ pr_debug("EINVAL: io_setup: ctx %lu nr_events %u\n",
+ ctx, nr_events);
+@@ -1272,8 +1369,12 @@
+ ret = PTR_ERR(ioctx);
+ if (!IS_ERR(ioctx)) {
+ ret = put_user(ioctx->user_id, ctxp);
+- if (!ret)
+- return 0;
++#ifdef CONFIG_EPOLL
++ if (make_fd && ret >= 0)
++ ret = make_aio_fd(ioctx);
++#endif
++ if (ret >= 0)
++ return ret;
+
+ get_ioctx(ioctx); /* io_destroy() expects us to hold a ref */
+ io_destroy(ioctx);
+diff -r d894e36cfc30 -r 0aa021803deb fs/binfmt_elf.c
+--- a/fs/binfmt_elf.c Tue Sep 09 11:37:38 2008 +0200
++++ b/fs/binfmt_elf.c Wed Sep 10 10:54:08 2008 +0100
+@@ -1179,6 +1179,10 @@
+ */
+ static int maydump(struct vm_area_struct *vma)
+ {
++ /* The vma can be set up to tell us the answer directly. */
++ if (vma->vm_flags & VM_ALWAYSDUMP)
++ return 1;
+
-+ if (len == 0)
-+ return;
+ /* Do not dump I/O mapped devices or special mappings */
+ if (vma->vm_flags & (VM_IO | VM_RESERVED))
+ return 0;
+diff -r d894e36cfc30 -r 0aa021803deb fs/bio.c
+--- a/fs/bio.c Tue Sep 09 11:37:38 2008 +0200
++++ b/fs/bio.c Wed Sep 10 10:54:08 2008 +0100
+@@ -112,7 +112,8 @@
+
+ BIO_BUG_ON(pool_idx >= BIOVEC_NR_POOLS);
+
+- mempool_free(bio->bi_io_vec, bio_set->bvec_pools[pool_idx]);
++ if (bio->bi_io_vec)
++ mempool_free(bio->bi_io_vec, bio_set->bvec_pools[pool_idx]);
+ mempool_free(bio, bio_set->bio_pool);
+ }
+
+diff -r d894e36cfc30 -r 0aa021803deb fs/compat_ioctl.c
+--- a/fs/compat_ioctl.c Tue Sep 09 11:37:38 2008 +0200
++++ b/fs/compat_ioctl.c Wed Sep 10 10:54:08 2008 +0100
+@@ -122,6 +122,13 @@
+ #include <linux/dvb/frontend.h>
+ #include <linux/dvb/video.h>
+ #include <linux/lp.h>
+
-+ rb = kmalloc(sizeof(*rb) + len, GFP_KERNEL);
-+ BUG_ON(rb == NULL);
++#ifdef CONFIG_XEN
++#include <xen/interface/xen.h>
++#include <xen/public/evtchn.h>
++#include <xen/public/privcmd.h>
++#include <xen/compat_ioctl.h>
++#endif
+
+ /* Aiee. Someone does not find a difference between int and long */
+ #define EXT2_IOC32_GETFLAGS _IOR('f', 1, int)
+@@ -2899,6 +2906,18 @@
+ /*LPGETSTATS not implemented, but no kernels seem to compile it in anyways*/
+ COMPATIBLE_IOCTL(LPGETFLAGS)
+ HANDLE_IOCTL(LPSETTIMEOUT, lp_timeout_trans)
++
++#ifdef CONFIG_XEN
++HANDLE_IOCTL(IOCTL_PRIVCMD_MMAP_32, privcmd_ioctl_32)
++HANDLE_IOCTL(IOCTL_PRIVCMD_MMAPBATCH_32, privcmd_ioctl_32)
++COMPATIBLE_IOCTL(IOCTL_PRIVCMD_HYPERCALL)
++COMPATIBLE_IOCTL(IOCTL_EVTCHN_BIND_VIRQ)
++COMPATIBLE_IOCTL(IOCTL_EVTCHN_BIND_INTERDOMAIN)
++COMPATIBLE_IOCTL(IOCTL_EVTCHN_BIND_UNBOUND_PORT)
++COMPATIBLE_IOCTL(IOCTL_EVTCHN_UNBIND)
++COMPATIBLE_IOCTL(IOCTL_EVTCHN_NOTIFY)
++COMPATIBLE_IOCTL(IOCTL_EVTCHN_RESET)
++#endif
+ };
+
+ int ioctl_table_size = ARRAY_SIZE(ioctl_start);
+diff -r d894e36cfc30 -r 0aa021803deb fs/eventpoll.c
+--- a/fs/eventpoll.c Tue Sep 09 11:37:38 2008 +0200
++++ b/fs/eventpoll.c Wed Sep 10 10:54:08 2008 +0100
+@@ -236,8 +236,6 @@
+
+ static void ep_poll_safewake_init(struct poll_safewake *psw);
+ static void ep_poll_safewake(struct poll_safewake *psw, wait_queue_head_t *wq);
+-static int ep_getfd(int *efd, struct inode **einode, struct file **efile,
+- struct eventpoll *ep);
+ static int ep_alloc(struct eventpoll **pep);
+ static void ep_free(struct eventpoll *ep);
+ static struct epitem *ep_find(struct eventpoll *ep, struct file *file, int fd);
+@@ -267,7 +265,7 @@
+ static int ep_poll(struct eventpoll *ep, struct epoll_event __user *events,
+ int maxevents, long timeout);
+ static int eventpollfs_delete_dentry(struct dentry *dentry);
+-static struct inode *ep_eventpoll_inode(void);
++static struct inode *ep_eventpoll_inode(const struct file_operations *fops);
+ static int eventpollfs_get_sb(struct file_system_type *fs_type,
+ int flags, const char *dev_name,
+ void *data, struct vfsmount *mnt);
+@@ -517,7 +515,7 @@
+ * Creates all the items needed to setup an eventpoll file. That is,
+ * a file structure, and inode and a free file descriptor.
+ */
+- error = ep_getfd(&fd, &inode, &file, ep);
++ error = ep_getfd(&fd, &inode, &file, ep, &eventpoll_fops);
+ if (error)
+ goto eexit_2;
+
+@@ -702,8 +700,8 @@
+ /*
+ * Creates the file descriptor to be used by the epoll interface.
+ */
+-static int ep_getfd(int *efd, struct inode **einode, struct file **efile,
+- struct eventpoll *ep)
++int ep_getfd(int *efd, struct inode **einode, struct file **efile,
++ struct eventpoll *ep, const struct file_operations *fops)
+ {
+ struct qstr this;
+ char name[32];
+@@ -719,7 +717,7 @@
+ goto eexit_1;
+
+ /* Allocates an inode from the eventpoll file system */
+- inode = ep_eventpoll_inode();
++ inode = ep_eventpoll_inode(fops);
+ error = PTR_ERR(inode);
+ if (IS_ERR(inode))
+ goto eexit_2;
+@@ -750,7 +748,7 @@
+
+ file->f_pos = 0;
+ file->f_flags = O_RDONLY;
+- file->f_op = &eventpoll_fops;
++ file->f_op = fops;
+ file->f_mode = FMODE_READ;
+ file->f_version = 0;
+ file->private_data = ep;
+@@ -1569,7 +1567,7 @@
+ }
+
+
+-static struct inode *ep_eventpoll_inode(void)
++static struct inode *ep_eventpoll_inode(const struct file_operations *fops)
+ {
+ int error = -ENOMEM;
+ struct inode *inode = new_inode(eventpoll_mnt->mnt_sb);
+@@ -1577,7 +1575,7 @@
+ if (!inode)
+ goto eexit_1;
+
+- inode->i_fop = &eventpoll_fops;
++ inode->i_fop = fops;
+
+ /*
+ * Mark the inode dirty from the very beginning,
+diff -r d894e36cfc30 -r 0aa021803deb fs/lockd/svc.c
+--- a/fs/lockd/svc.c Tue Sep 09 11:37:38 2008 +0200
++++ b/fs/lockd/svc.c Wed Sep 10 10:54:08 2008 +0100
+@@ -323,9 +323,6 @@
+ * Sysctl parameters (same as module parameters, different interface).
+ */
+
+-/* Something that isn't CTL_ANY, CTL_NONE or a value that may clash. */
+-#define CTL_UNNUMBERED -2
+-
+ static ctl_table nlm_sysctls[] = {
+ {
+ .ctl_name = CTL_UNNUMBERED,
+diff -r d894e36cfc30 -r 0aa021803deb fs/nfs/sysctl.c
+--- a/fs/nfs/sysctl.c Tue Sep 09 11:37:38 2008 +0200
++++ b/fs/nfs/sysctl.c Wed Sep 10 10:54:08 2008 +0100
+@@ -18,11 +18,6 @@
+ static const int nfs_set_port_min = 0;
+ static const int nfs_set_port_max = 65535;
+ static struct ctl_table_header *nfs_callback_sysctl_table;
+-/*
+- * Something that isn't CTL_ANY, CTL_NONE or a value that may clash.
+- * Use the same values as fs/lockd/svc.c
+- */
+-#define CTL_UNNUMBERED -2
+
+ static ctl_table nfs_cb_sysctls[] = {
+ #ifdef CONFIG_NFS_V4
+diff -r d894e36cfc30 -r 0aa021803deb fs/proc/proc_misc.c
+--- a/fs/proc/proc_misc.c Tue Sep 09 11:37:38 2008 +0200
++++ b/fs/proc/proc_misc.c Wed Sep 10 10:54:08 2008 +0100
+@@ -471,7 +471,7 @@
+ (unsigned long long)cputime64_to_clock_t(irq),
+ (unsigned long long)cputime64_to_clock_t(softirq),
+ (unsigned long long)cputime64_to_clock_t(steal));
+- for_each_online_cpu(i) {
++ for_each_possible_cpu(i) {
+
+ /* Copy values here to work around gcc-2.95.3, gcc-2.96 */
+ user = kstat_cpu(i).cpustat.user;
+diff -r d894e36cfc30 -r 0aa021803deb fs/proc/vmcore.c
+--- a/fs/proc/vmcore.c Tue Sep 09 11:37:38 2008 +0200
++++ b/fs/proc/vmcore.c Wed Sep 10 10:54:08 2008 +0100
+@@ -514,7 +514,7 @@
+ /* Do some basic Verification. */
+ if (memcmp(ehdr.e_ident, ELFMAG, SELFMAG) != 0 ||
+ (ehdr.e_type != ET_CORE) ||
+- !elf_check_arch(&ehdr) ||
++ !vmcore_elf_check_arch(&ehdr) ||
+ ehdr.e_ident[EI_CLASS] != ELFCLASS64 ||
+ ehdr.e_ident[EI_VERSION] != EV_CURRENT ||
+ ehdr.e_version != EV_CURRENT ||
+diff -r d894e36cfc30 -r 0aa021803deb fs/xfs/linux-2.6/xfs_buf.c
+--- a/fs/xfs/linux-2.6/xfs_buf.c Tue Sep 09 11:37:38 2008 +0200
++++ b/fs/xfs/linux-2.6/xfs_buf.c Wed Sep 10 10:54:08 2008 +0100
+@@ -181,6 +181,19 @@
+ void *addr)
+ {
+ a_list_t *aentry;
+
-+ rb->cons = 0;
-+ rb->len = len;
++#ifdef CONFIG_XEN
++ /*
++ * Xen needs to be able to make sure it can get an exclusive
++ * RO mapping of pages it wants to turn into a pagetable. If
++ * a newly allocated page is also still being vmap()ed by xfs,
++ * it will cause pagetable construction to fail. This is a
++ * quick workaround to always eagerly unmap pages so that Xen
++ * is happy.
++ */
++ vunmap(addr);
++ return;
++#endif
+
+ aentry = kmalloc(sizeof(a_list_t), GFP_NOWAIT);
+ if (likely(aentry)) {
+diff -r d894e36cfc30 -r 0aa021803deb include/acpi/aclocal.h
+--- a/include/acpi/aclocal.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/acpi/aclocal.h Wed Sep 10 10:54:08 2008 +0100
+@@ -708,7 +708,7 @@
+ * must be preserved.
+ */
+ #define ACPI_PM1_STATUS_PRESERVED_BITS 0x0800 /* Bit 11 */
+-#define ACPI_PM1_CONTROL_PRESERVED_BITS 0x0201 /* Bit 9, Bit 0 (SCI_EN) */
++#define ACPI_PM1_CONTROL_PRESERVED_BITS 0x0200 /* Bit 9 (whatever) */
+
+ /*
+ * Register IDs
+diff -r d894e36cfc30 -r 0aa021803deb include/acpi/pdc_intel.h
+--- a/include/acpi/pdc_intel.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/acpi/pdc_intel.h Wed Sep 10 10:54:08 2008 +0100
+@@ -13,6 +13,7 @@
+ #define ACPI_PDC_SMP_C_SWCOORD (0x0040)
+ #define ACPI_PDC_SMP_T_SWCOORD (0x0080)
+ #define ACPI_PDC_C_C1_FFH (0x0100)
++#define ACPI_PDC_C_C2C3_FFH (0x0200)
+
+ #define ACPI_PDC_EST_CAPABILITY_SMP (ACPI_PDC_SMP_C1PT | \
+ ACPI_PDC_C_C1_HALT | \
+@@ -25,6 +26,8 @@
+
+ #define ACPI_PDC_C_CAPABILITY_SMP (ACPI_PDC_SMP_C2C3 | \
+ ACPI_PDC_SMP_C1PT | \
+- ACPI_PDC_C_C1_HALT)
++ ACPI_PDC_C_C1_HALT | \
++ ACPI_PDC_C_C1_FFH | \
++ ACPI_PDC_C_C2C3_FFH)
+
+ #endif /* __PDC_INTEL_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/acpi/processor.h
+--- a/include/acpi/processor.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/acpi/processor.h Wed Sep 10 10:54:08 2008 +0100
+@@ -21,6 +21,12 @@
+ #define ACPI_PSD_REV0_REVISION 0 /* Support for _PSD as in ACPI 3.0 */
+ #define ACPI_PSD_REV0_ENTRIES 5
+
++#ifdef CONFIG_XEN
++#define NR_ACPI_CPUS (NR_CPUS < 256 ? 256 : NR_CPUS)
++#else
++#define NR_ACPI_CPUS NR_CPUS
++#endif /* CONFIG_XEN */
+
-+ memcpy(rb->msg, data, len);
+ /*
+ * Types of coordination defined in ACPI 3.0. Same macros can be used across
+ * P, C and T states
+@@ -32,6 +38,17 @@
+ /* Power Management */
+
+ struct acpi_processor_cx;
++
++#ifdef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++struct acpi_csd_package {
++ acpi_integer num_entries;
++ acpi_integer revision;
++ acpi_integer domain;
++ acpi_integer coord_type;
++ acpi_integer num_processors;
++ acpi_integer index;
++} __attribute__ ((packed));
++#endif
+
+ struct acpi_power_register {
+ u8 descriptor;
+@@ -63,6 +80,12 @@
+ u32 power;
+ u32 usage;
+ u64 time;
++#ifdef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++ /* Require raw information for external control logic */
++ struct acpi_power_register reg;
++ u32 csd_count;
++ struct acpi_csd_package *domain_info;
++#endif
+ struct acpi_processor_cx_policy promotion;
+ struct acpi_processor_cx_policy demotion;
+ };
+@@ -231,6 +254,9 @@
+ {
+ return;
+ }
++#ifdef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++int acpi_processor_ppc_has_changed(struct acpi_processor *pr);
++#else
+ static inline int acpi_processor_ppc_has_changed(struct acpi_processor *pr)
+ {
+ static unsigned int printout = 1;
+@@ -243,6 +269,7 @@
+ }
+ return 0;
+ }
++#endif /* CONFIG_PROCESSOR_EXTERNAL_CONTROL */
+ #endif /* CONFIG_CPU_FREQ */
+
+ /* in processor_throttling.c */
+@@ -275,4 +302,79 @@
+ }
+ #endif
+
++/*
++ * Following are interfaces geared to external processor PM control
++ * logic like a VMM
++ */
++/* Events notified to external control logic */
++#define PROCESSOR_PM_INIT 1
++#define PROCESSOR_PM_CHANGE 2
++#define PROCESSOR_HOTPLUG 3
+
-+ list_add_tail(&rb->list, &u->read_buffers);
++/* Objects for the PM events */
++#define PM_TYPE_IDLE 0
++#define PM_TYPE_PERF 1
++#define PM_TYPE_THR 2
++#define PM_TYPE_MAX 3
+
-+ wake_up(&u->read_waitq);
-+}
++/* Processor hotplug events */
++#define HOTPLUG_TYPE_ADD 0
++#define HOTPLUG_TYPE_REMOVE 1
+
-+struct watch_adapter
-+{
-+ struct list_head list;
-+ struct xenbus_watch watch;
-+ struct xenbus_dev_data *dev_data;
-+ char *token;
++#ifdef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++struct processor_extcntl_ops {
++ /* Transfer processor PM events to external control logic */
++ int (*pm_ops[PM_TYPE_MAX])(struct acpi_processor *pr, int event);
++ /* Notify physical processor status to external control logic */
++ int (*hotplug)(struct acpi_processor *pr, int event);
+};
++extern const struct processor_extcntl_ops *processor_extcntl_ops;
+
-+static void free_watch_adapter (struct watch_adapter *watch)
++static inline int processor_cntl_external(void)
+{
-+ kfree(watch->watch.node);
-+ kfree(watch->token);
-+ kfree(watch);
++ return (processor_extcntl_ops != NULL);
+}
+
-+static void watch_fired(struct xenbus_watch *watch,
-+ const char **vec,
-+ unsigned int len)
++static inline int processor_pm_external(void)
+{
-+ struct watch_adapter *adap =
-+ container_of(watch, struct watch_adapter, watch);
-+ struct xsd_sockmsg hdr;
-+ const char *path, *token;
-+ int path_len, tok_len, body_len;
-+
-+ path = vec[XS_WATCH_PATH];
-+ token = adap->token;
-+
-+ path_len = strlen(path) + 1;
-+ tok_len = strlen(token) + 1;
-+ body_len = path_len + tok_len;
-+
-+ hdr.type = XS_WATCH_EVENT;
-+ hdr.len = body_len;
-+
-+ mutex_lock(&adap->dev_data->reply_mutex);
-+ queue_reply(adap->dev_data, (char *)&hdr, sizeof(hdr));
-+ queue_reply(adap->dev_data, (char *)path, path_len);
-+ queue_reply(adap->dev_data, (char *)token, tok_len);
-+ mutex_unlock(&adap->dev_data->reply_mutex);
++ return processor_cntl_external() &&
++ (processor_extcntl_ops->pm_ops[PM_TYPE_IDLE] != NULL);
+}
+
-+static LIST_HEAD(watch_list);
++static inline int processor_pmperf_external(void)
++{
++ return processor_cntl_external() &&
++ (processor_extcntl_ops->pm_ops[PM_TYPE_PERF] != NULL);
++}
+
-+static ssize_t xenbus_dev_write(struct file *filp,
-+ const char __user *ubuf,
-+ size_t len, loff_t *ppos)
++static inline int processor_pmthr_external(void)
+{
-+ struct xenbus_dev_data *u = filp->private_data;
-+ struct xenbus_dev_transaction *trans = NULL;
-+ uint32_t msg_type;
-+ void *reply;
-+ char *path, *token;
-+ struct watch_adapter *watch, *tmp_watch;
-+ int err, rc = len;
++ return processor_cntl_external() &&
++ (processor_extcntl_ops->pm_ops[PM_TYPE_THR] != NULL);
++}
+
-+ if ((len + u->len) > sizeof(u->u.buffer)) {
-+ rc = -EINVAL;
-+ goto out;
-+ }
++extern int processor_notify_external(struct acpi_processor *pr,
++ int event, int type);
++extern void processor_extcntl_init(void);
++extern int processor_extcntl_prepare(struct acpi_processor *pr);
++extern int acpi_processor_get_performance_info(struct acpi_processor *pr);
++extern int acpi_processor_get_psd(struct acpi_processor *pr);
++void arch_acpi_processor_init_extcntl(const struct processor_extcntl_ops **);
++#else
++static inline int processor_cntl_external(void) {return 0;}
++static inline int processor_pm_external(void) {return 0;}
++static inline int processor_pmperf_external(void) {return 0;}
++static inline int processor_pmthr_external(void) {return 0;}
++static inline int processor_notify_external(struct acpi_processor *pr,
++ int event, int type)
++{
++ return 0;
++}
++static inline void processor_extcntl_init(void) {}
++static inline int processor_extcntl_prepare(struct acpi_processor *pr)
++{
++ return 0;
++}
++#endif /* CONFIG_PROCESSOR_EXTERNAL_CONTROL */
+
-+ if (copy_from_user(u->u.buffer + u->len, ubuf, len) != 0) {
-+ rc = -EFAULT;
-+ goto out;
-+ }
+ #endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-generic/pci.h
+--- a/include/asm-generic/pci.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-generic/pci.h Wed Sep 10 10:54:08 2008 +0100
+@@ -43,7 +43,9 @@
+ return root;
+ }
+
++#ifndef pcibios_scan_all_fns
+ #define pcibios_scan_all_fns(a, b) 0
++#endif
+
+ #ifndef HAVE_ARCH_PCI_GET_LEGACY_IDE_IRQ
+ static inline int pci_get_legacy_ide_irq(struct pci_dev *dev, int channel)
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-generic/pgtable.h
+--- a/include/asm-generic/pgtable.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-generic/pgtable.h Wed Sep 10 10:54:08 2008 +0100
+@@ -188,6 +188,10 @@
+ })
+ #endif
+
++#ifndef arch_change_pte_range
++#define arch_change_pte_range(mm, pmd, addr, end, newprot) 0
++#endif
+
-+ u->len += len;
-+ if ((u->len < sizeof(u->u.msg)) ||
-+ (u->len < (sizeof(u->u.msg) + u->u.msg.len)))
-+ return rc;
+ #ifndef __ASSEMBLY__
+ /*
+ * When walking page tables, we usually want to skip any p?d_none entries;
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-generic/vmlinux.lds.h
+--- a/include/asm-generic/vmlinux.lds.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-generic/vmlinux.lds.h Wed Sep 10 10:54:08 2008 +0100
+@@ -194,3 +194,6 @@
+ .stab.index 0 : { *(.stab.index) } \
+ .stab.indexstr 0 : { *(.stab.indexstr) } \
+ .comment 0 : { *(.comment) }
+
-+ msg_type = u->u.msg.type;
++#define NOTES \
++ .notes : { *(.note.*) } :note
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/acpi.h
+--- a/include/asm-i386/acpi.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-i386/acpi.h Wed Sep 10 10:54:08 2008 +0100
+@@ -31,6 +31,9 @@
+ #include <acpi/pdc_intel.h>
+
+ #include <asm/system.h> /* defines cmpxchg */
++#ifdef CONFIG_XEN
++#include <xen/interface/platform.h>
++#endif
+
+ #define COMPILER_DEPENDENT_INT64 long long
+ #define COMPILER_DEPENDENT_UINT64 unsigned long long
+@@ -156,6 +159,27 @@
+ }
+ extern int acpi_irq_balance_set(char *str);
+
++#ifdef CONFIG_XEN
++static inline int acpi_notify_hypervisor_state(u8 sleep_state,
++ u32 pm1a_cnt_val,
++ u32 pm1b_cnt_val)
++{
++ struct xen_platform_op op = {
++ .cmd = XENPF_enter_acpi_sleep,
++ .interface_version = XENPF_INTERFACE_VERSION,
++ .u = {
++ .enter_acpi_sleep = {
++ .pm1a_cnt_val = pm1a_cnt_val,
++ .pm1b_cnt_val = pm1b_cnt_val,
++ .sleep_state = sleep_state,
++ },
++ },
++ };
+
-+ switch (msg_type) {
-+ case XS_TRANSACTION_START:
-+ case XS_TRANSACTION_END:
-+ case XS_DIRECTORY:
-+ case XS_READ:
-+ case XS_GET_PERMS:
-+ case XS_RELEASE:
-+ case XS_GET_DOMAIN_PATH:
-+ case XS_WRITE:
-+ case XS_MKDIR:
-+ case XS_RM:
-+ case XS_SET_PERMS:
-+ if (msg_type == XS_TRANSACTION_START) {
-+ trans = kmalloc(sizeof(*trans), GFP_KERNEL);
-+ if (!trans) {
-+ rc = -ENOMEM;
-+ goto out;
-+ }
-+ }
++ return HYPERVISOR_platform_op(&op);
++}
++#endif /* CONFIG_XEN */
+
-+ reply = xenbus_dev_request_and_reply(&u->u.msg);
-+ if (IS_ERR(reply)) {
-+ kfree(trans);
-+ rc = PTR_ERR(reply);
-+ goto out;
-+ }
+ #else /* !CONFIG_ACPI */
+
+ #define acpi_lapic 0
+@@ -181,7 +205,9 @@
+
+ extern u8 x86_acpiid_to_apicid[];
+
++#ifndef CONFIG_XEN
+ #define ARCH_HAS_POWER_INIT 1
++#endif
+
+ #endif /*__KERNEL__*/
+
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/agp.h
+--- a/include/asm-i386/agp.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-i386/agp.h Wed Sep 10 10:54:08 2008 +0100
+@@ -12,8 +12,10 @@
+ * data corruption on some CPUs.
+ */
+
+-int map_page_into_agp(struct page *page);
+-int unmap_page_from_agp(struct page *page);
++/* Caller's responsibility to call global_flush_tlb() for
++ * performance reasons */
++#define map_page_into_agp(page) change_page_attr(page, 1, PAGE_KERNEL_NOCACHE)
++#define unmap_page_from_agp(page) change_page_attr(page, 1, PAGE_KERNEL)
+ #define flush_agp_mappings() global_flush_tlb()
+
+ /* Could use CLFLUSH here if the cpu supports it. But then it would
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/apic.h
+--- a/include/asm-i386/apic.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-i386/apic.h Wed Sep 10 10:54:08 2008 +0100
+@@ -119,10 +119,12 @@
+
+ extern int disable_timer_pin_1;
+
++#ifndef CONFIG_XEN
+ void smp_send_timer_broadcast_ipi(struct pt_regs *regs);
+ void switch_APIC_timer_to_ipi(void *cpumask);
+ void switch_ipi_to_APIC_timer(void *cpumask);
+ #define ARCH_APICTIMER_STOPS_ON_C3 1
++#endif
+
+ extern int timer_over_8254;
+
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/e820.h
+--- a/include/asm-i386/e820.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-i386/e820.h Wed Sep 10 10:54:08 2008 +0100
+@@ -38,6 +38,7 @@
+
+ extern int e820_all_mapped(unsigned long start, unsigned long end,
+ unsigned type);
++extern int e820_any_mapped(u64 start, u64 end, unsigned type);
+
+ #endif/*!__ASSEMBLY__*/
+
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/elf.h
+--- a/include/asm-i386/elf.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-i386/elf.h Wed Sep 10 10:54:08 2008 +0100
+@@ -169,50 +169,6 @@
+ NEW_AUX_ENT(AT_SYSINFO_EHDR, VDSO_COMPAT_BASE); \
+ } while (0)
+
+-/*
+- * These macros parameterize elf_core_dump in fs/binfmt_elf.c to write out
+- * extra segments containing the vsyscall DSO contents. Dumping its
+- * contents makes post-mortem fully interpretable later without matching up
+- * the same kernel and hardware config to see what PC values meant.
+- * Dumping its extra ELF program headers includes all the other information
+- * a debugger needs to easily find how the vsyscall DSO was being used.
+- */
+-#define ELF_CORE_EXTRA_PHDRS (VDSO_HIGH_EHDR->e_phnum)
+-#define ELF_CORE_WRITE_EXTRA_PHDRS \
+-do { \
+- const struct elf_phdr *const vsyscall_phdrs = \
+- (const struct elf_phdr *) (VDSO_HIGH_BASE \
+- + VDSO_HIGH_EHDR->e_phoff); \
+- int i; \
+- Elf32_Off ofs = 0; \
+- for (i = 0; i < VDSO_HIGH_EHDR->e_phnum; ++i) { \
+- struct elf_phdr phdr = vsyscall_phdrs[i]; \
+- if (phdr.p_type == PT_LOAD) { \
+- BUG_ON(ofs != 0); \
+- ofs = phdr.p_offset = offset; \
+- phdr.p_memsz = PAGE_ALIGN(phdr.p_memsz); \
+- phdr.p_filesz = phdr.p_memsz; \
+- offset += phdr.p_filesz; \
+- } \
+- else \
+- phdr.p_offset += ofs; \
+- phdr.p_paddr = 0; /* match other core phdrs */ \
+- DUMP_WRITE(&phdr, sizeof(phdr)); \
+- } \
+-} while (0)
+-#define ELF_CORE_WRITE_EXTRA_DATA \
+-do { \
+- const struct elf_phdr *const vsyscall_phdrs = \
+- (const struct elf_phdr *) (VDSO_HIGH_BASE \
+- + VDSO_HIGH_EHDR->e_phoff); \
+- int i; \
+- for (i = 0; i < VDSO_HIGH_EHDR->e_phnum; ++i) { \
+- if (vsyscall_phdrs[i].p_type == PT_LOAD) \
+- DUMP_WRITE((void *) vsyscall_phdrs[i].p_vaddr, \
+- PAGE_ALIGN(vsyscall_phdrs[i].p_memsz)); \
+- } \
+-} while (0)
+-
+ #endif
+
+ #endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/fixmap.h
+--- a/include/asm-i386/fixmap.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-i386/fixmap.h Wed Sep 10 10:54:08 2008 +0100
+@@ -19,7 +19,7 @@
+ * Leave one empty page between vmalloc'ed areas and
+ * the start of the fixmap.
+ */
+-#define __FIXADDR_TOP 0xfffff000
++extern unsigned long __FIXADDR_TOP;
+
+ #ifndef __ASSEMBLY__
+ #include <linux/kernel.h>
+@@ -94,6 +94,8 @@
+ extern void __set_fixmap (enum fixed_addresses idx,
+ unsigned long phys, pgprot_t flags);
+
++extern void set_fixaddr_top(unsigned long top);
+
-+ if (msg_type == XS_TRANSACTION_START) {
-+ trans->handle.id = simple_strtoul(reply, NULL, 0);
-+ list_add(&trans->list, &u->transactions);
-+ } else if (msg_type == XS_TRANSACTION_END) {
-+ list_for_each_entry(trans, &u->transactions, list)
-+ if (trans->handle.id == u->u.msg.tx_id)
-+ break;
-+ BUG_ON(&trans->list == &u->transactions);
-+ list_del(&trans->list);
-+ kfree(trans);
-+ }
-+ mutex_lock(&u->reply_mutex);
-+ queue_reply(u, (char *)&u->u.msg, sizeof(u->u.msg));
-+ queue_reply(u, (char *)reply, u->u.msg.len);
-+ mutex_unlock(&u->reply_mutex);
-+ kfree(reply);
-+ break;
+ #define set_fixmap(idx, phys) \
+ __set_fixmap(idx, phys, PAGE_KERNEL)
+ /*
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/io_apic.h
+--- a/include/asm-i386/io_apic.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-i386/io_apic.h Wed Sep 10 10:54:08 2008 +0100
+@@ -12,7 +12,7 @@
+
+ #ifdef CONFIG_X86_IO_APIC
+
+-#ifdef CONFIG_PCI_MSI
++#if defined(CONFIG_PCI_MSI) && !defined(CONFIG_XEN)
+ static inline int use_pci_vector(void) {return 1;}
+ static inline void disable_edge_ioapic_vector(unsigned int vector) { }
+ static inline void mask_and_ack_level_ioapic_vector(unsigned int vector) { }
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/kexec.h
+--- a/include/asm-i386/kexec.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-i386/kexec.h Wed Sep 10 10:54:08 2008 +0100
+@@ -1,5 +1,25 @@
+ #ifndef _I386_KEXEC_H
+ #define _I386_KEXEC_H
+
-+ case XS_WATCH:
-+ case XS_UNWATCH: {
-+ static const char *XS_RESP = "OK";
-+ struct xsd_sockmsg hdr;
++#define PA_CONTROL_PAGE 0
++#define VA_CONTROL_PAGE 1
++#define PA_PGD 2
++#define VA_PGD 3
++#define PA_PTE_0 4
++#define VA_PTE_0 5
++#define PA_PTE_1 6
++#define VA_PTE_1 7
++#ifdef CONFIG_X86_PAE
++#define PA_PMD_0 8
++#define VA_PMD_0 9
++#define PA_PMD_1 10
++#define VA_PMD_1 11
++#define PAGES_NR 12
++#else
++#define PAGES_NR 8
++#endif
+
-+ path = u->u.buffer + sizeof(u->u.msg);
-+ token = memchr(path, 0, u->u.msg.len);
-+ if (token == NULL) {
-+ rc = -EILSEQ;
-+ goto out;
-+ }
-+ token++;
++#ifndef __ASSEMBLY__
+
+ #include <asm/fixmap.h>
+ #include <asm/ptrace.h>
+@@ -26,6 +46,9 @@
+
+ /* The native architecture */
+ #define KEXEC_ARCH KEXEC_ARCH_386
+
-+ if (msg_type == XS_WATCH) {
-+ watch = kmalloc(sizeof(*watch), GFP_KERNEL);
-+ watch->watch.node = kmalloc(strlen(path)+1,
-+ GFP_KERNEL);
-+ strcpy((char *)watch->watch.node, path);
-+ watch->watch.callback = watch_fired;
-+ watch->token = kmalloc(strlen(token)+1, GFP_KERNEL);
-+ strcpy(watch->token, token);
-+ watch->dev_data = u;
++/* We can also handle crash dumps from 64 bit kernel. */
++#define vmcore_elf_check_arch_cross(x) ((x)->e_machine == EM_X86_64)
+
+ #define MAX_NOTE_BYTES 1024
+
+@@ -72,5 +95,26 @@
+ newregs->eip = (unsigned long)current_text_addr();
+ }
+ }
++asmlinkage NORET_TYPE void
++relocate_kernel(unsigned long indirection_page,
++ unsigned long control_page,
++ unsigned long start_address,
++ unsigned int has_pae) ATTRIB_NORET;
+
-+ err = register_xenbus_watch(&watch->watch);
-+ if (err) {
-+ free_watch_adapter(watch);
-+ rc = err;
-+ goto out;
-+ }
-+
-+ list_add(&watch->list, &u->watches);
-+ } else {
-+ list_for_each_entry_safe(watch, tmp_watch,
-+ &u->watches, list) {
-+ if (!strcmp(watch->token, token) &&
-+ !strcmp(watch->watch.node, path))
-+ {
-+ unregister_xenbus_watch(&watch->watch);
-+ list_del(&watch->list);
-+ free_watch_adapter(watch);
-+ break;
-+ }
-+ }
-+ }
+
-+ hdr.type = msg_type;
-+ hdr.len = strlen(XS_RESP) + 1;
-+ mutex_lock(&u->reply_mutex);
-+ queue_reply(u, (char *)&hdr, sizeof(hdr));
-+ queue_reply(u, (char *)XS_RESP, hdr.len);
-+ mutex_unlock(&u->reply_mutex);
-+ break;
-+ }
++/* Under Xen we need to work with machine addresses. These macros give the
++ * machine address of a certain page to the generic kexec code instead of
++ * the pseudo physical address which would be given by the default macros.
++ */
+
-+ default:
-+ rc = -EINVAL;
-+ break;
-+ }
++#ifdef CONFIG_XEN
++#define KEXEC_ARCH_HAS_PAGE_MACROS
++#define kexec_page_to_pfn(page) pfn_to_mfn(page_to_pfn(page))
++#define kexec_pfn_to_page(pfn) pfn_to_page(mfn_to_pfn(pfn))
++#define kexec_virt_to_phys(addr) virt_to_machine(addr)
++#define kexec_phys_to_virt(addr) phys_to_virt(machine_to_phys(addr))
++#endif
+
-+ out:
-+ u->len = 0;
-+ return rc;
++#endif /* __ASSEMBLY__ */
+
+ #endif /* _I386_KEXEC_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-default/mach_traps.h
+--- a/include/asm-i386/mach-default/mach_traps.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-i386/mach-default/mach_traps.h Wed Sep 10 10:54:08 2008 +0100
+@@ -12,6 +12,18 @@
+ static inline void clear_mem_error(unsigned char reason)
+ {
+ reason = (reason & 0xf) | 4;
++ outb(reason, 0x61);
+}
+
-+static int xenbus_dev_open(struct inode *inode, struct file *filp)
++static inline void clear_io_check_error(unsigned char reason)
+{
-+ struct xenbus_dev_data *u;
++ unsigned long i;
+
-+ if (xen_store_evtchn == 0)
-+ return -ENOENT;
++ reason = (reason & 0xf) | 8;
++ outb(reason, 0x61);
++ i = 2000;
++ while (--i) udelay(1000);
++ reason &= ~8;
+ outb(reason, 0x61);
+ }
+
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/agp.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/agp.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,44 @@
++#ifndef AGP_H
++#define AGP_H 1
+
-+ nonseekable_open(inode, filp);
++#include <asm/pgtable.h>
++#include <asm/cacheflush.h>
++#include <asm/system.h>
+
-+ u = kzalloc(sizeof(*u), GFP_KERNEL);
-+ if (u == NULL)
-+ return -ENOMEM;
++/*
++ * Functions to keep the agpgart mappings coherent with the MMU.
++ * The GART gives the CPU a physical alias of pages in memory. The alias region is
++ * mapped uncacheable. Make sure there are no conflicting mappings
++ * with different cachability attributes for the same page. This avoids
++ * data corruption on some CPUs.
++ */
+
-+ INIT_LIST_HEAD(&u->transactions);
-+ INIT_LIST_HEAD(&u->watches);
-+ INIT_LIST_HEAD(&u->read_buffers);
-+ init_waitqueue_head(&u->read_waitq);
++/* Caller's responsibility to call global_flush_tlb() for
++ * performance reasons */
++#define map_page_into_agp(page) ( \
++ xen_create_contiguous_region((unsigned long)page_address(page), 0, 32) \
++ ?: change_page_attr(page, 1, PAGE_KERNEL_NOCACHE))
++#define unmap_page_from_agp(page) ( \
++ xen_destroy_contiguous_region((unsigned long)page_address(page), 0), \
++ /* only a fallback: xen_destroy_contiguous_region uses PAGE_KERNEL */ \
++ change_page_attr(page, 1, PAGE_KERNEL))
++#define flush_agp_mappings() global_flush_tlb()
+
-+ mutex_init(&u->reply_mutex);
++/* Could use CLFLUSH here if the cpu supports it. But then it would
++ need to be called for each cacheline of the whole page so it may not be
++ worth it. Would need a page for it. */
++#define flush_agp_cache() wbinvd()
+
-+ filp->private_data = u;
++/* Convert a physical address to an address suitable for the GART. */
++#define phys_to_gart(x) phys_to_machine(x)
++#define gart_to_phys(x) machine_to_phys(x)
+
-+ return 0;
-+}
++/* GATT allocation. Returns/accepts GATT kernel virtual address. */
++#define alloc_gatt_pages(order) ({ \
++ char *_t; dma_addr_t _d; \
++ _t = dma_alloc_coherent(NULL,PAGE_SIZE<<(order),&_d,GFP_KERNEL); \
++ _t; })
++#define free_gatt_pages(table, order) \
++ dma_free_coherent(NULL,PAGE_SIZE<<(order),(table),virt_to_bus(table))
+
-+static int xenbus_dev_release(struct inode *inode, struct file *filp)
-+{
-+ struct xenbus_dev_data *u = filp->private_data;
-+ struct xenbus_dev_transaction *trans, *tmp;
-+ struct watch_adapter *watch, *tmp_watch;
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/desc.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/desc.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,166 @@
++#ifndef __ARCH_DESC_H
++#define __ARCH_DESC_H
+
-+ list_for_each_entry_safe(trans, tmp, &u->transactions, list) {
-+ xenbus_transaction_end(trans->handle, 1);
-+ list_del(&trans->list);
-+ kfree(trans);
-+ }
++#include <asm/ldt.h>
++#include <asm/segment.h>
+
-+ list_for_each_entry_safe(watch, tmp_watch, &u->watches, list) {
-+ unregister_xenbus_watch(&watch->watch);
-+ list_del(&watch->list);
-+ free_watch_adapter(watch);
-+ }
++#define CPU_16BIT_STACK_SIZE 1024
+
-+ kfree(u);
++#ifndef __ASSEMBLY__
+
-+ return 0;
-+}
++#include <linux/preempt.h>
++#include <linux/smp.h>
+
-+static unsigned int xenbus_dev_poll(struct file *file, poll_table *wait)
-+{
-+ struct xenbus_dev_data *u = file->private_data;
++#include <asm/mmu.h>
+
-+ poll_wait(file, &u->read_waitq, wait);
-+ if (!list_empty(&u->read_buffers))
-+ return POLLIN | POLLRDNORM;
-+ return 0;
-+}
++extern struct desc_struct cpu_gdt_table[GDT_ENTRIES];
+
-+static const struct file_operations xenbus_dev_file_ops = {
-+ .read = xenbus_dev_read,
-+ .write = xenbus_dev_write,
-+ .open = xenbus_dev_open,
-+ .release = xenbus_dev_release,
-+ .poll = xenbus_dev_poll,
-+};
++DECLARE_PER_CPU(unsigned char, cpu_16bit_stack[CPU_16BIT_STACK_SIZE]);
+
-+int xenbus_dev_init(void)
-+{
-+ xenbus_dev_intf = create_xen_proc_entry("xenbus", 0400);
-+ if (xenbus_dev_intf)
-+ xenbus_dev_intf->proc_fops = &xenbus_dev_file_ops;
++struct Xgt_desc_struct {
++ unsigned short size;
++ unsigned long address __attribute__((packed));
++ unsigned short pad;
++} __attribute__ ((packed));
+
-+ return 0;
-+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_probe_backend.c tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_probe_backend.c
---- pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_probe_backend.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_probe_backend.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,287 @@
-+/******************************************************************************
-+ * Talks to Xen Store to figure out what devices we have (backend half).
-+ *
-+ * Copyright (C) 2005 Rusty Russell, IBM Corporation
-+ * Copyright (C) 2005 Mike Wray, Hewlett-Packard
-+ * Copyright (C) 2005, 2006 XenSource Ltd
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++extern struct Xgt_desc_struct idt_descr;
++DECLARE_PER_CPU(struct Xgt_desc_struct, cpu_gdt_descr);
+
-+#define DPRINTK(fmt, args...) \
-+ pr_debug("xenbus_probe (%s:%d) " fmt ".\n", \
-+ __FUNCTION__, __LINE__, ##args)
+
-+#include <linux/kernel.h>
-+#include <linux/err.h>
-+#include <linux/string.h>
-+#include <linux/ctype.h>
-+#include <linux/fcntl.h>
-+#include <linux/mm.h>
-+#include <linux/notifier.h>
-+#include <linux/kthread.h>
++static inline struct desc_struct *get_cpu_gdt_table(unsigned int cpu)
++{
++ return (struct desc_struct *)per_cpu(cpu_gdt_descr, cpu).address;
++}
+
-+#include <asm/io.h>
-+#include <asm/page.h>
-+#include <asm/maddr.h>
-+#include <asm/pgtable.h>
-+#include <asm/hypervisor.h>
-+#include <xen/xenbus.h>
-+#include <xen/xen_proc.h>
-+#include <xen/evtchn.h>
-+#include <xen/features.h>
-+#include <xen/hvm.h>
++#define load_TR_desc() __asm__ __volatile__("ltr %w0"::"q" (GDT_ENTRY_TSS*8))
++#define load_LDT_desc() __asm__ __volatile__("lldt %w0"::"q" (GDT_ENTRY_LDT*8))
+
-+#include "xenbus_comms.h"
-+#include "xenbus_probe.h"
++#define load_gdt(dtr) __asm__ __volatile("lgdt %0"::"m" (*dtr))
++#define load_idt(dtr) __asm__ __volatile("lidt %0"::"m" (*dtr))
++#define load_tr(tr) __asm__ __volatile("ltr %0"::"mr" (tr))
++#define load_ldt(ldt) __asm__ __volatile("lldt %0"::"mr" (ldt))
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
-+#endif
++#define store_gdt(dtr) __asm__ ("sgdt %0":"=m" (*dtr))
++#define store_idt(dtr) __asm__ ("sidt %0":"=m" (*dtr))
++#define store_tr(tr) __asm__ ("str %0":"=mr" (tr))
++#define store_ldt(ldt) __asm__ ("sldt %0":"=mr" (ldt))
+
-+static int xenbus_uevent_backend(struct device *dev, char **envp,
-+ int num_envp, char *buffer, int buffer_size);
-+static int xenbus_probe_backend(const char *type, const char *domid);
++/*
++ * This is the ldt that every process will get unless we need
++ * something other than this.
++ */
++extern struct desc_struct default_ldt[];
++extern void set_intr_gate(unsigned int irq, void * addr);
+
-+extern int read_otherend_details(struct xenbus_device *xendev,
-+ char *id_node, char *path_node);
++#define _set_tssldt_desc(n,addr,limit,type) \
++__asm__ __volatile__ ("movw %w3,0(%2)\n\t" \
++ "movw %w1,2(%2)\n\t" \
++ "rorl $16,%1\n\t" \
++ "movb %b1,4(%2)\n\t" \
++ "movb %4,5(%2)\n\t" \
++ "movb $0,6(%2)\n\t" \
++ "movb %h1,7(%2)\n\t" \
++ "rorl $16,%1" \
++ : "=m"(*(n)) : "q" (addr), "r"(n), "ir"(limit), "i"(type))
+
-+static int read_frontend_details(struct xenbus_device *xendev)
++#ifndef CONFIG_X86_NO_TSS
++static inline void __set_tss_desc(unsigned int cpu, unsigned int entry, void *addr)
+{
-+ return read_otherend_details(xendev, "frontend-id", "frontend");
++ _set_tssldt_desc(&get_cpu_gdt_table(cpu)[entry], (int)addr,
++ offsetof(struct tss_struct, __cacheline_filler) - 1, 0x89);
+}
+
-+/* backend/<type>/<fe-uuid>/<id> => <type>-<fe-domid>-<id> */
-+static int backend_bus_id(char bus_id[BUS_ID_SIZE], const char *nodename)
++#define set_tss_desc(cpu,addr) __set_tss_desc(cpu, GDT_ENTRY_TSS, addr)
++#endif
++
++static inline void set_ldt_desc(unsigned int cpu, void *addr, unsigned int size)
+{
-+ int domid, err;
-+ const char *devid, *type, *frontend;
-+ unsigned int typelen;
++ _set_tssldt_desc(&get_cpu_gdt_table(cpu)[GDT_ENTRY_LDT], (int)addr, ((size << 3)-1), 0x82);
++}
+
-+ type = strchr(nodename, '/');
-+ if (!type)
-+ return -EINVAL;
-+ type++;
-+ typelen = strcspn(type, "/");
-+ if (!typelen || type[typelen] != '/')
-+ return -EINVAL;
++#define LDT_entry_a(info) \
++ ((((info)->base_addr & 0x0000ffff) << 16) | ((info)->limit & 0x0ffff))
+
-+ devid = strrchr(nodename, '/') + 1;
++#define LDT_entry_b(info) \
++ (((info)->base_addr & 0xff000000) | \
++ (((info)->base_addr & 0x00ff0000) >> 16) | \
++ ((info)->limit & 0xf0000) | \
++ (((info)->read_exec_only ^ 1) << 9) | \
++ ((info)->contents << 10) | \
++ (((info)->seg_not_present ^ 1) << 15) | \
++ ((info)->seg_32bit << 22) | \
++ ((info)->limit_in_pages << 23) | \
++ ((info)->useable << 20) | \
++ 0x7000)
+
-+ err = xenbus_gather(XBT_NIL, nodename, "frontend-id", "%i", &domid,
-+ "frontend", NULL, &frontend,
-+ NULL);
-+ if (err)
-+ return err;
-+ if (strlen(frontend) == 0)
-+ err = -ERANGE;
-+ if (!err && !xenbus_exists(XBT_NIL, frontend, ""))
-+ err = -ENOENT;
-+ kfree(frontend);
++#define LDT_empty(info) (\
++ (info)->base_addr == 0 && \
++ (info)->limit == 0 && \
++ (info)->contents == 0 && \
++ (info)->read_exec_only == 1 && \
++ (info)->seg_32bit == 0 && \
++ (info)->limit_in_pages == 0 && \
++ (info)->seg_not_present == 1 && \
++ (info)->useable == 0 )
+
-+ if (err)
-+ return err;
++extern int write_ldt_entry(void *ldt, int entry, __u32 entry_a, __u32 entry_b);
+
-+ if (snprintf(bus_id, BUS_ID_SIZE,
-+ "%.*s-%i-%s", typelen, type, domid, devid) >= BUS_ID_SIZE)
-+ return -ENOSPC;
-+ return 0;
++#if TLS_SIZE != 24
++# error update this code.
++#endif
++
++static inline void load_TLS(struct thread_struct *t, unsigned int cpu)
++{
++#define C(i) if (HYPERVISOR_update_descriptor(virt_to_machine(&get_cpu_gdt_table(cpu)[GDT_ENTRY_TLS_MIN + i]), \
++ *(u64 *)&t->tls_array[i])) \
++ BUG();
++ C(0); C(1); C(2);
++#undef C
+}
+
-+static struct xen_bus_type xenbus_backend = {
-+ .root = "backend",
-+ .levels = 3, /* backend/type/<frontend>/<id> */
-+ .get_bus_id = backend_bus_id,
-+ .probe = xenbus_probe_backend,
-+ .error = -ENODEV,
-+ .bus = {
-+ .name = "xen-backend",
-+ .match = xenbus_match,
-+ .probe = xenbus_dev_probe,
-+ .remove = xenbus_dev_remove,
-+// .shutdown = xenbus_dev_shutdown,
-+ .uevent = xenbus_uevent_backend,
-+ },
-+ .dev = {
-+ .bus_id = "xen-backend",
-+ },
-+};
++static inline void clear_LDT(void)
++{
++ int cpu = get_cpu();
+
-+static int xenbus_uevent_backend(struct device *dev, char **envp,
-+ int num_envp, char *buffer, int buffer_size)
++ /*
++ * NB. We load the default_ldt for lcall7/27 handling on demand, as
++ * it slows down context switching. Noone uses it anyway.
++ */
++ cpu = cpu; /* XXX avoid compiler warning */
++ xen_set_ldt(NULL, 0);
++ put_cpu();
++}
++
++/*
++ * load one particular LDT into the current CPU
++ */
++static inline void load_LDT_nolock(mm_context_t *pc, int cpu)
+{
-+ struct xenbus_device *xdev;
-+ struct xenbus_driver *drv;
-+ int i = 0;
-+ int length = 0;
++ void *segments = pc->ldt;
++ int count = pc->size;
+
-+ DPRINTK("");
++ if (likely(!count))
++ segments = NULL;
+
-+ if (dev == NULL)
-+ return -ENODEV;
++ xen_set_ldt(segments, count);
++}
+
-+ xdev = to_xenbus_device(dev);
-+ if (xdev == NULL)
-+ return -ENODEV;
++static inline void load_LDT(mm_context_t *pc)
++{
++ int cpu = get_cpu();
++ load_LDT_nolock(pc, cpu);
++ put_cpu();
++}
+
-+ /* stuff we want to pass to /sbin/hotplug */
-+ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
-+ "XENBUS_TYPE=%s", xdev->devicetype);
++static inline unsigned long get_desc_base(unsigned long *desc)
++{
++ unsigned long base;
++ base = ((desc[0] >> 16) & 0x0000ffff) |
++ ((desc[1] << 16) & 0x00ff0000) |
++ (desc[1] & 0xff000000);
++ return base;
++}
+
-+ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
-+ "XENBUS_PATH=%s", xdev->nodename);
++#endif /* !__ASSEMBLY__ */
+
-+ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
-+ "XENBUS_BASE_PATH=%s", xenbus_backend.root);
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/dma-mapping.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/dma-mapping.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,151 @@
++#ifndef _ASM_I386_DMA_MAPPING_H
++#define _ASM_I386_DMA_MAPPING_H
+
-+ /* terminate, set to next free slot, shrink available space */
-+ envp[i] = NULL;
-+ envp = &envp[i];
-+ num_envp -= i;
-+ buffer = &buffer[length];
-+ buffer_size -= length;
++/*
++ * IOMMU interface. See Documentation/DMA-mapping.txt and DMA-API.txt for
++ * documentation.
++ */
+
-+ if (dev->driver) {
-+ drv = to_xenbus_driver(dev->driver);
-+ if (drv && drv->uevent)
-+ return drv->uevent(xdev, envp, num_envp, buffer,
-+ buffer_size);
-+ }
++#include <linux/mm.h>
++#include <asm/cache.h>
++#include <asm/io.h>
++#include <asm/scatterlist.h>
++#include <asm/swiotlb.h>
+
-+ return 0;
++static inline int
++address_needs_mapping(struct device *hwdev, dma_addr_t addr)
++{
++ dma_addr_t mask = 0xffffffff;
++ /* If the device has a mask, use it, otherwise default to 32 bits */
++ if (hwdev && hwdev->dma_mask)
++ mask = *hwdev->dma_mask;
++ return (addr & ~mask) != 0;
+}
+
-+int xenbus_register_backend(struct xenbus_driver *drv)
-+{
-+ drv->read_otherend_details = read_frontend_details;
++extern int range_straddles_page_boundary(paddr_t p, size_t size);
+
-+ return xenbus_register_driver_common(drv, &xenbus_backend);
-+}
-+EXPORT_SYMBOL_GPL(xenbus_register_backend);
++#define dma_alloc_noncoherent(d, s, h, f) dma_alloc_coherent(d, s, h, f)
++#define dma_free_noncoherent(d, s, v, h) dma_free_coherent(d, s, v, h)
+
-+/* backend/<typename>/<frontend-uuid>/<name> */
-+static int xenbus_probe_backend_unit(const char *dir,
-+ const char *type,
-+ const char *name)
-+{
-+ char *nodename;
-+ int err;
++void *dma_alloc_coherent(struct device *dev, size_t size,
++ dma_addr_t *dma_handle, gfp_t flag);
+
-+ nodename = kasprintf(GFP_KERNEL, "%s/%s", dir, name);
-+ if (!nodename)
-+ return -ENOMEM;
++void dma_free_coherent(struct device *dev, size_t size,
++ void *vaddr, dma_addr_t dma_handle);
+
-+ DPRINTK("%s\n", nodename);
++extern dma_addr_t
++dma_map_single(struct device *dev, void *ptr, size_t size,
++ enum dma_data_direction direction);
+
-+ err = xenbus_probe_node(&xenbus_backend, type, nodename);
-+ kfree(nodename);
-+ return err;
-+}
++extern void
++dma_unmap_single(struct device *dev, dma_addr_t dma_addr, size_t size,
++ enum dma_data_direction direction);
+
-+/* backend/<typename>/<frontend-domid> */
-+static int xenbus_probe_backend(const char *type, const char *domid)
-+{
-+ char *nodename;
-+ int err = 0;
-+ char **dir;
-+ unsigned int i, dir_n = 0;
++extern int dma_map_sg(struct device *hwdev, struct scatterlist *sg,
++ int nents, enum dma_data_direction direction);
++extern void dma_unmap_sg(struct device *hwdev, struct scatterlist *sg,
++ int nents, enum dma_data_direction direction);
+
-+ DPRINTK("");
++#ifdef CONFIG_HIGHMEM
++extern dma_addr_t
++dma_map_page(struct device *dev, struct page *page, unsigned long offset,
++ size_t size, enum dma_data_direction direction);
+
-+ nodename = kasprintf(GFP_KERNEL, "%s/%s/%s", xenbus_backend.root, type, domid);
-+ if (!nodename)
-+ return -ENOMEM;
++extern void
++dma_unmap_page(struct device *dev, dma_addr_t dma_address, size_t size,
++ enum dma_data_direction direction);
++#else
++#define dma_map_page(dev, page, offset, size, dir) \
++ dma_map_single(dev, page_address(page) + (offset), (size), (dir))
++#define dma_unmap_page dma_unmap_single
++#endif
+
-+ dir = xenbus_directory(XBT_NIL, nodename, "", &dir_n);
-+ if (IS_ERR(dir)) {
-+ kfree(nodename);
-+ return PTR_ERR(dir);
-+ }
++extern void
++dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle, size_t size,
++ enum dma_data_direction direction);
++
++extern void
++dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle, size_t size,
++ enum dma_data_direction direction);
+
-+ for (i = 0; i < dir_n; i++) {
-+ err = xenbus_probe_backend_unit(nodename, type, dir[i]);
-+ if (err)
-+ break;
-+ }
-+ kfree(dir);
-+ kfree(nodename);
-+ return err;
++static inline void
++dma_sync_single_range_for_cpu(struct device *dev, dma_addr_t dma_handle,
++ unsigned long offset, size_t size,
++ enum dma_data_direction direction)
++{
++ dma_sync_single_for_cpu(dev, dma_handle+offset, size, direction);
+}
+
-+static void backend_changed(struct xenbus_watch *watch,
-+ const char **vec, unsigned int len)
++static inline void
++dma_sync_single_range_for_device(struct device *dev, dma_addr_t dma_handle,
++ unsigned long offset, size_t size,
++ enum dma_data_direction direction)
+{
-+ DPRINTK("");
-+
-+ dev_changed(vec[XS_WATCH_PATH], &xenbus_backend);
++ dma_sync_single_for_device(dev, dma_handle+offset, size, direction);
+}
+
-+static struct xenbus_watch be_watch = {
-+ .node = "backend",
-+ .callback = backend_changed,
-+};
-+
-+void xenbus_backend_suspend(int (*fn)(struct device *, void *))
++static inline void
++dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg, int nelems,
++ enum dma_data_direction direction)
+{
-+ DPRINTK("");
-+ if (!xenbus_backend.error)
-+ bus_for_each_dev(&xenbus_backend.bus, NULL, NULL, fn);
++ if (swiotlb)
++ swiotlb_sync_sg_for_cpu(dev,sg,nelems,direction);
++ flush_write_buffers();
+}
+
-+void xenbus_backend_resume(int (*fn)(struct device *, void *))
++static inline void
++dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg, int nelems,
++ enum dma_data_direction direction)
+{
-+ DPRINTK("");
-+ if (!xenbus_backend.error)
-+ bus_for_each_dev(&xenbus_backend.bus, NULL, NULL, fn);
++ if (swiotlb)
++ swiotlb_sync_sg_for_device(dev,sg,nelems,direction);
++ flush_write_buffers();
+}
+
-+void xenbus_backend_probe_and_watch(void)
++extern int
++dma_mapping_error(dma_addr_t dma_addr);
++
++extern int
++dma_supported(struct device *dev, u64 mask);
++
++static inline int
++dma_set_mask(struct device *dev, u64 mask)
+{
-+ xenbus_probe_devices(&xenbus_backend);
-+ register_xenbus_watch(&be_watch);
++ if(!dev->dma_mask || !dma_supported(dev, mask))
++ return -EIO;
++
++ *dev->dma_mask = mask;
++
++ return 0;
+}
+
-+void xenbus_backend_bus_register(void)
++static inline int
++dma_get_cache_alignment(void)
+{
-+ xenbus_backend.error = bus_register(&xenbus_backend.bus);
-+ if (xenbus_backend.error)
-+ printk(KERN_WARNING
-+ "XENBUS: Error registering backend bus: %i\n",
-+ xenbus_backend.error);
++ /* no easy way to get cache size on all x86, so return the
++ * maximum possible, to be safe */
++ return (1 << INTERNODE_CACHE_SHIFT);
+}
+
-+void xenbus_backend_device_register(void)
-+{
-+ if (xenbus_backend.error)
-+ return;
++#define dma_is_consistent(d) (1)
+
-+ xenbus_backend.error = device_register(&xenbus_backend.dev);
-+ if (xenbus_backend.error) {
-+ bus_unregister(&xenbus_backend.bus);
-+ printk(KERN_WARNING
-+ "XENBUS: Error registering backend device: %i\n",
-+ xenbus_backend.error);
-+ }
++static inline void
++dma_cache_sync(void *vaddr, size_t size,
++ enum dma_data_direction direction)
++{
++ flush_write_buffers();
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_probe.c tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_probe.c
---- pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_probe.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_probe.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,1132 @@
-+/******************************************************************************
-+ * Talks to Xen Store to figure out what devices we have.
-+ *
-+ * Copyright (C) 2005 Rusty Russell, IBM Corporation
-+ * Copyright (C) 2005 Mike Wray, Hewlett-Packard
-+ * Copyright (C) 2005, 2006 XenSource Ltd
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
-+
-+#define DPRINTK(fmt, args...) \
-+ pr_debug("xenbus_probe (%s:%d) " fmt ".\n", \
-+ __FUNCTION__, __LINE__, ##args)
+
-+#include <linux/kernel.h>
-+#include <linux/err.h>
-+#include <linux/string.h>
-+#include <linux/ctype.h>
-+#include <linux/fcntl.h>
-+#include <linux/mm.h>
-+#include <linux/notifier.h>
-+#include <linux/kthread.h>
-+#include <linux/mutex.h>
++#define ARCH_HAS_DMA_DECLARE_COHERENT_MEMORY
++extern int
++dma_declare_coherent_memory(struct device *dev, dma_addr_t bus_addr,
++ dma_addr_t device_addr, size_t size, int flags);
+
-+#include <asm/io.h>
-+#include <asm/page.h>
-+#include <asm/maddr.h>
-+#include <asm/pgtable.h>
-+#include <asm/hypervisor.h>
-+#include <xen/xenbus.h>
-+#include <xen/xen_proc.h>
-+#include <xen/evtchn.h>
-+#include <xen/features.h>
-+#include <xen/hvm.h>
++extern void
++dma_release_declared_memory(struct device *dev);
+
-+#include "xenbus_comms.h"
-+#include "xenbus_probe.h"
++extern void *
++dma_mark_declared_memory_occupied(struct device *dev,
++ dma_addr_t device_addr, size_t size);
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
+#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/fixmap.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/fixmap.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,155 @@
++/*
++ * fixmap.h: compile-time virtual memory allocation
++ *
++ * This file is subject to the terms and conditions of the GNU General Public
++ * License. See the file "COPYING" in the main directory of this archive
++ * for more details.
++ *
++ * Copyright (C) 1998 Ingo Molnar
++ *
++ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
++ */
+
-+int xen_store_evtchn;
-+struct xenstore_domain_interface *xen_store_interface;
-+static unsigned long xen_store_mfn;
++#ifndef _ASM_FIXMAP_H
++#define _ASM_FIXMAP_H
+
-+extern struct mutex xenwatch_mutex;
+
-+static ATOMIC_NOTIFIER_HEAD(xenstore_chain);
++/* used by vmalloc.c, vsyscall.lds.S.
++ *
++ * Leave one empty page between vmalloc'ed areas and
++ * the start of the fixmap.
++ */
++extern unsigned long __FIXADDR_TOP;
+
-+static void wait_for_devices(struct xenbus_driver *xendrv);
++#ifndef __ASSEMBLY__
++#include <linux/kernel.h>
++#include <asm/acpi.h>
++#include <asm/apicdef.h>
++#include <asm/page.h>
++#ifdef CONFIG_HIGHMEM
++#include <linux/threads.h>
++#include <asm/kmap_types.h>
++#endif
+
-+static int xenbus_probe_frontend(const char *type, const char *name);
++/*
++ * Here we define all the compile-time 'special' virtual
++ * addresses. The point is to have a constant address at
++ * compile time, but to set the physical address only
++ * in the boot process. We allocate these special addresses
++ * from the end of virtual memory (0xfffff000) backwards.
++ * Also this lets us do fail-safe vmalloc(), we
++ * can guarantee that these special addresses and
++ * vmalloc()-ed addresses never overlap.
++ *
++ * these 'compile-time allocated' memory buffers are
++ * fixed-size 4k pages. (or larger if used with an increment
++ * highger than 1) use fixmap_set(idx,phys) to associate
++ * physical memory with fixmap indices.
++ *
++ * TLB entries of such buffers will not be flushed across
++ * task switches.
++ */
++enum fixed_addresses {
++ FIX_HOLE,
++ FIX_VDSO,
++#ifdef CONFIG_X86_LOCAL_APIC
++ FIX_APIC_BASE, /* local (CPU) APIC) -- required for SMP or not */
++#endif
++#ifdef CONFIG_X86_IO_APIC
++ FIX_IO_APIC_BASE_0,
++ FIX_IO_APIC_BASE_END = FIX_IO_APIC_BASE_0 + MAX_IO_APICS-1,
++#endif
++#ifdef CONFIG_X86_VISWS_APIC
++ FIX_CO_CPU, /* Cobalt timer */
++ FIX_CO_APIC, /* Cobalt APIC Redirection Table */
++ FIX_LI_PCIA, /* Lithium PCI Bridge A */
++ FIX_LI_PCIB, /* Lithium PCI Bridge B */
++#endif
++#ifdef CONFIG_X86_F00F_BUG
++ FIX_F00F_IDT, /* Virtual mapping for IDT */
++#endif
++#ifdef CONFIG_X86_CYCLONE_TIMER
++ FIX_CYCLONE_TIMER, /*cyclone timer register*/
++#endif
++#ifdef CONFIG_HIGHMEM
++ FIX_KMAP_BEGIN, /* reserved pte's for temporary kernel mappings */
++ FIX_KMAP_END = FIX_KMAP_BEGIN+(KM_TYPE_NR*NR_CPUS)-1,
++#endif
++#ifdef CONFIG_ACPI
++ FIX_ACPI_BEGIN,
++ FIX_ACPI_END = FIX_ACPI_BEGIN + FIX_ACPI_PAGES - 1,
++#endif
++#ifdef CONFIG_PCI_MMCONFIG
++ FIX_PCIE_MCFG,
++#endif
++ FIX_SHARED_INFO,
++#define NR_FIX_ISAMAPS 256
++ FIX_ISAMAP_END,
++ FIX_ISAMAP_BEGIN = FIX_ISAMAP_END + NR_FIX_ISAMAPS - 1,
++ __end_of_permanent_fixed_addresses,
++ /* temporary boot-time mappings, used before ioremap() is functional */
++#define NR_FIX_BTMAPS 16
++ FIX_BTMAP_END = __end_of_permanent_fixed_addresses,
++ FIX_BTMAP_BEGIN = FIX_BTMAP_END + NR_FIX_BTMAPS - 1,
++ FIX_WP_TEST,
++ __end_of_fixed_addresses
++};
+
-+static void xenbus_dev_shutdown(struct device *_dev);
++extern void set_fixaddr_top(unsigned long top);
+
-+/* If something in array of ids matches this device, return it. */
-+static const struct xenbus_device_id *
-+match_device(const struct xenbus_device_id *arr, struct xenbus_device *dev)
-+{
-+ for (; *arr->devicetype != '\0'; arr++) {
-+ if (!strcmp(arr->devicetype, dev->devicetype))
-+ return arr;
-+ }
-+ return NULL;
-+}
++extern void __set_fixmap(enum fixed_addresses idx,
++ maddr_t phys, pgprot_t flags);
+
-+int xenbus_match(struct device *_dev, struct device_driver *_drv)
-+{
-+ struct xenbus_driver *drv = to_xenbus_driver(_drv);
++#define set_fixmap(idx, phys) \
++ __set_fixmap(idx, phys, PAGE_KERNEL)
++/*
++ * Some hardware wants to get fixmapped without caching.
++ */
++#define set_fixmap_nocache(idx, phys) \
++ __set_fixmap(idx, phys, PAGE_KERNEL_NOCACHE)
+
-+ if (!drv->ids)
-+ return 0;
++#define clear_fixmap(idx) \
++ __set_fixmap(idx, 0, __pgprot(0))
+
-+ return match_device(drv->ids, to_xenbus_device(_dev)) != NULL;
-+}
++#define FIXADDR_TOP ((unsigned long)__FIXADDR_TOP)
+
-+/* device/<type>/<id> => <type>-<id> */
-+static int frontend_bus_id(char bus_id[BUS_ID_SIZE], const char *nodename)
-+{
-+ nodename = strchr(nodename, '/');
-+ if (!nodename || strlen(nodename + 1) >= BUS_ID_SIZE) {
-+ printk(KERN_WARNING "XENBUS: bad frontend %s\n", nodename);
-+ return -EINVAL;
-+ }
++#define __FIXADDR_SIZE (__end_of_permanent_fixed_addresses << PAGE_SHIFT)
++#define __FIXADDR_BOOT_SIZE (__end_of_fixed_addresses << PAGE_SHIFT)
++#define FIXADDR_START (FIXADDR_TOP - __FIXADDR_SIZE)
++#define FIXADDR_BOOT_START (FIXADDR_TOP - __FIXADDR_BOOT_SIZE)
+
-+ strlcpy(bus_id, nodename + 1, BUS_ID_SIZE);
-+ if (!strchr(bus_id, '/')) {
-+ printk(KERN_WARNING "XENBUS: bus_id %s no slash\n", bus_id);
-+ return -EINVAL;
-+ }
-+ *strchr(bus_id, '/') = '-';
-+ return 0;
-+}
++#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT))
++#define __virt_to_fix(x) ((FIXADDR_TOP - ((x)&PAGE_MASK)) >> PAGE_SHIFT)
+
++extern void __this_fixmap_does_not_exist(void);
+
-+static void free_otherend_details(struct xenbus_device *dev)
++/*
++ * 'index to address' translation. If anyone tries to use the idx
++ * directly without tranlation, we catch the bug with a NULL-deference
++ * kernel oops. Illegal ranges of incoming indices are caught too.
++ */
++static __always_inline unsigned long fix_to_virt(const unsigned int idx)
+{
-+ kfree(dev->otherend);
-+ dev->otherend = NULL;
-+}
++ /*
++ * this branch gets completely eliminated after inlining,
++ * except when someone tries to use fixaddr indices in an
++ * illegal way. (such as mixing up address types or using
++ * out-of-range indices).
++ *
++ * If it doesn't get removed, the linker will complain
++ * loudly with a reasonably clear error message..
++ */
++ if (idx >= __end_of_fixed_addresses)
++ __this_fixmap_does_not_exist();
+
++ return __fix_to_virt(idx);
++}
+
-+static void free_otherend_watch(struct xenbus_device *dev)
++static inline unsigned long virt_to_fix(const unsigned long vaddr)
+{
-+ if (dev->otherend_watch.node) {
-+ unregister_xenbus_watch(&dev->otherend_watch);
-+ kfree(dev->otherend_watch.node);
-+ dev->otherend_watch.node = NULL;
-+ }
++ BUG_ON(vaddr >= FIXADDR_TOP || vaddr < FIXADDR_START);
++ return __virt_to_fix(vaddr);
+}
+
++#endif /* !__ASSEMBLY__ */
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/gnttab_dma.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/gnttab_dma.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,41 @@
++/*
++ * Copyright (c) 2007 Herbert Xu <herbert at gondor.apana.org.au>
++ * Copyright (c) 2007 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ */
+
-+int read_otherend_details(struct xenbus_device *xendev,
-+ char *id_node, char *path_node)
-+{
-+ int err = xenbus_gather(XBT_NIL, xendev->nodename,
-+ id_node, "%i", &xendev->otherend_id,
-+ path_node, NULL, &xendev->otherend,
-+ NULL);
-+ if (err) {
-+ xenbus_dev_fatal(xendev, err,
-+ "reading other end details from %s",
-+ xendev->nodename);
-+ return err;
-+ }
-+ if (strlen(xendev->otherend) == 0 ||
-+ !xenbus_exists(XBT_NIL, xendev->otherend, "")) {
-+ xenbus_dev_fatal(xendev, -ENOENT,
-+ "unable to read other end from %s. "
-+ "missing or inaccessible.",
-+ xendev->nodename);
-+ free_otherend_details(xendev);
-+ return -ENOENT;
-+ }
++#ifndef _ASM_I386_GNTTAB_DMA_H
++#define _ASM_I386_GNTTAB_DMA_H
+
-+ return 0;
++static inline int gnttab_dma_local_pfn(struct page *page)
++{
++ /* Has it become a local MFN? */
++ return pfn_valid(mfn_to_local_pfn(pfn_to_mfn(page_to_pfn(page))));
+}
+
-+
-+static int read_backend_details(struct xenbus_device *xendev)
++static inline maddr_t gnttab_dma_map_page(struct page *page)
+{
-+ return read_otherend_details(xendev, "backend-id", "backend");
++ __gnttab_dma_map_page(page);
++ return ((maddr_t)pfn_to_mfn(page_to_pfn(page)) << PAGE_SHIFT);
+}
+
-+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
-+static int xenbus_uevent_frontend(struct device *dev, char **envp,
-+ int num_envp, char *buffer, int buffer_size)
++static inline void gnttab_dma_unmap_page(maddr_t maddr)
+{
-+ struct xenbus_device *xdev;
-+ int length = 0, i = 0;
-+
-+ if (dev == NULL)
-+ return -ENODEV;
-+ xdev = to_xenbus_device(dev);
-+ if (xdev == NULL)
-+ return -ENODEV;
++ __gnttab_dma_unmap_page(virt_to_page(bus_to_virt(maddr)));
++}
+
-+ /* stuff we want to pass to /sbin/hotplug */
-+ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
-+ "XENBUS_TYPE=%s", xdev->devicetype);
-+ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
-+ "XENBUS_PATH=%s", xdev->nodename);
-+ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
-+ "MODALIAS=xen:%s", xdev->devicetype);
++#endif /* _ASM_I386_GNTTAB_DMA_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/highmem.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/highmem.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,80 @@
++/*
++ * highmem.h: virtual kernel memory mappings for high memory
++ *
++ * Used in CONFIG_HIGHMEM systems for memory pages which
++ * are not addressable by direct kernel virtual addresses.
++ *
++ * Copyright (C) 1999 Gerhard Wichert, Siemens AG
++ * Gerhard.Wichert at pdb.siemens.de
++ *
++ *
++ * Redesigned the x86 32-bit VM architecture to deal with
++ * up to 16 Terabyte physical memory. With current x86 CPUs
++ * we now support up to 64 Gigabytes physical RAM.
++ *
++ * Copyright (C) 1999 Ingo Molnar <mingo at redhat.com>
++ */
+
-+ return 0;
-+}
-+#endif
++#ifndef _ASM_HIGHMEM_H
++#define _ASM_HIGHMEM_H
+
-+/* Bus type for frontend drivers. */
-+static struct xen_bus_type xenbus_frontend = {
-+ .root = "device",
-+ .levels = 2, /* device/type/<id> */
-+ .get_bus_id = frontend_bus_id,
-+ .probe = xenbus_probe_frontend,
-+ .error = -ENODEV,
-+ .bus = {
-+ .name = "xen",
-+ .match = xenbus_match,
-+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
-+ .probe = xenbus_dev_probe,
-+ .remove = xenbus_dev_remove,
-+ .shutdown = xenbus_dev_shutdown,
-+ .uevent = xenbus_uevent_frontend,
-+#endif
-+ },
-+ .dev = {
-+ .bus_id = "xen",
-+ },
-+};
++#ifdef __KERNEL__
+
-+static void otherend_changed(struct xenbus_watch *watch,
-+ const char **vec, unsigned int len)
-+{
-+ struct xenbus_device *dev =
-+ container_of(watch, struct xenbus_device, otherend_watch);
-+ struct xenbus_driver *drv = to_xenbus_driver(dev->dev.driver);
-+ enum xenbus_state state;
++#include <linux/interrupt.h>
++#include <linux/threads.h>
++#include <asm/kmap_types.h>
++#include <asm/tlbflush.h>
+
-+ /* Protect us against watches firing on old details when the otherend
-+ details change, say immediately after a resume. */
-+ if (!dev->otherend ||
-+ strncmp(dev->otherend, vec[XS_WATCH_PATH],
-+ strlen(dev->otherend))) {
-+ DPRINTK("Ignoring watch at %s", vec[XS_WATCH_PATH]);
-+ return;
-+ }
++/* declarations for highmem.c */
++extern unsigned long highstart_pfn, highend_pfn;
+
-+ state = xenbus_read_driver_state(dev->otherend);
++extern pte_t *kmap_pte;
++extern pgprot_t kmap_prot;
++extern pte_t *pkmap_page_table;
+
-+ DPRINTK("state is %d (%s), %s, %s", state, xenbus_strstate(state),
-+ dev->otherend_watch.node, vec[XS_WATCH_PATH]);
++/*
++ * Right now we initialize only a single pte table. It can be extended
++ * easily, subsequent pte tables have to be allocated in one physical
++ * chunk of RAM.
++ */
++#ifdef CONFIG_X86_PAE
++#define LAST_PKMAP 512
++#else
++#define LAST_PKMAP 1024
++#endif
++/*
++ * Ordering is:
++ *
++ * FIXADDR_TOP
++ * fixed_addresses
++ * FIXADDR_START
++ * temp fixed addresses
++ * FIXADDR_BOOT_START
++ * Persistent kmap area
++ * PKMAP_BASE
++ * VMALLOC_END
++ * Vmalloc area
++ * VMALLOC_START
++ * high_memory
++ */
++#define PKMAP_BASE ( (FIXADDR_BOOT_START - PAGE_SIZE*(LAST_PKMAP + 1)) & PMD_MASK )
++#define LAST_PKMAP_MASK (LAST_PKMAP-1)
++#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
++#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
+
-+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
-+ /*
-+ * Ignore xenbus transitions during shutdown. This prevents us doing
-+ * work that can fail e.g., when the rootfs is gone.
-+ */
-+ if (system_state > SYSTEM_RUNNING) {
-+ struct xen_bus_type *bus = bus;
-+ bus = container_of(dev->dev.bus, struct xen_bus_type, bus);
-+ /* If we're frontend, drive the state machine to Closed. */
-+ /* This should cause the backend to release our resources. */
-+ if ((bus == &xenbus_frontend) && (state == XenbusStateClosing))
-+ xenbus_frontend_closed(dev);
-+ return;
-+ }
-+#endif
++extern void * FASTCALL(kmap_high(struct page *page));
++extern void FASTCALL(kunmap_high(struct page *page));
+
-+ if (drv->otherend_changed)
-+ drv->otherend_changed(dev, state);
-+}
++void *kmap(struct page *page);
++void kunmap(struct page *page);
++void *kmap_atomic(struct page *page, enum km_type type);
++void *kmap_atomic_pte(struct page *page, enum km_type type);
++void kunmap_atomic(void *kvaddr, enum km_type type);
++void *kmap_atomic_pfn(unsigned long pfn, enum km_type type);
++struct page *kmap_atomic_to_page(void *ptr);
+
++#define flush_cache_kmaps() do { } while (0)
+
-+static int talk_to_otherend(struct xenbus_device *dev)
-+{
-+ struct xenbus_driver *drv = to_xenbus_driver(dev->dev.driver);
++#endif /* __KERNEL__ */
+
-+ free_otherend_watch(dev);
-+ free_otherend_details(dev);
++#endif /* _ASM_HIGHMEM_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/hw_irq.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/hw_irq.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,72 @@
++#ifndef _ASM_HW_IRQ_H
++#define _ASM_HW_IRQ_H
+
-+ return drv->read_otherend_details(dev);
-+}
++/*
++ * linux/include/asm/hw_irq.h
++ *
++ * (C) 1992, 1993 Linus Torvalds, (C) 1997 Ingo Molnar
++ *
++ * moved some of the old arch/i386/kernel/irq.h to here. VY
++ *
++ * IRQ/IPI changes taken from work by Thomas Radke
++ * <tomsoft at informatik.tu-chemnitz.de>
++ */
+
++#include <linux/profile.h>
++#include <asm/atomic.h>
++#include <asm/irq.h>
++#include <asm/sections.h>
+
-+static int watch_otherend(struct xenbus_device *dev)
-+{
-+ return xenbus_watch_path2(dev, dev->otherend, "state",
-+ &dev->otherend_watch, otherend_changed);
-+}
++struct hw_interrupt_type;
+
++#define NMI_VECTOR 0x02
+
-+int xenbus_dev_probe(struct device *_dev)
-+{
-+ struct xenbus_device *dev = to_xenbus_device(_dev);
-+ struct xenbus_driver *drv = to_xenbus_driver(_dev->driver);
-+ const struct xenbus_device_id *id;
-+ int err;
++/*
++ * Various low-level irq details needed by irq.c, process.c,
++ * time.c, io_apic.c and smp.c
++ *
++ * Interrupt entry/exit code at both C and assembly level
++ */
+
-+ DPRINTK("%s", dev->nodename);
++extern u8 irq_vector[NR_IRQ_VECTORS];
++#define IO_APIC_VECTOR(irq) (irq_vector[irq])
++#define AUTO_ASSIGN -1
+
-+ if (!drv->probe) {
-+ err = -ENODEV;
-+ goto fail;
-+ }
++extern void (*interrupt[NR_IRQS])(void);
+
-+ id = match_device(drv->ids, dev);
-+ if (!id) {
-+ err = -ENODEV;
-+ goto fail;
-+ }
++#ifdef CONFIG_SMP
++fastcall void reschedule_interrupt(void);
++fastcall void invalidate_interrupt(void);
++fastcall void call_function_interrupt(void);
++#endif
+
-+ err = talk_to_otherend(dev);
-+ if (err) {
-+ printk(KERN_WARNING
-+ "xenbus_probe: talk_to_otherend on %s failed.\n",
-+ dev->nodename);
-+ return err;
-+ }
++#ifdef CONFIG_X86_LOCAL_APIC
++fastcall void apic_timer_interrupt(void);
++fastcall void error_interrupt(void);
++fastcall void spurious_interrupt(void);
++fastcall void thermal_interrupt(struct pt_regs *);
++#define platform_legacy_irq(irq) ((irq) < 16)
++#endif
+
-+ err = drv->probe(dev, id);
-+ if (err)
-+ goto fail;
++void disable_8259A_irq(unsigned int irq);
++void enable_8259A_irq(unsigned int irq);
++int i8259A_irq_pending(unsigned int irq);
++void make_8259A_irq(unsigned int irq);
++void init_8259A(int aeoi);
++void FASTCALL(send_IPI_self(int vector));
++void init_VISWS_APIC_irqs(void);
++void setup_IO_APIC(void);
++void disable_IO_APIC(void);
++#define print_IO_APIC()
++int IO_APIC_get_PCI_irq_vector(int bus, int slot, int fn);
++void send_IPI(int dest, int vector);
++void setup_ioapic_dest(void);
+
-+ err = watch_otherend(dev);
-+ if (err) {
-+ printk(KERN_WARNING
-+ "xenbus_probe: watch_otherend on %s failed.\n",
-+ dev->nodename);
-+ return err;
-+ }
++extern unsigned long io_apic_irqs;
+
-+ return 0;
-+fail:
-+ xenbus_dev_error(dev, err, "xenbus_dev_probe on %s", dev->nodename);
-+ xenbus_switch_state(dev, XenbusStateClosed);
-+ return -ENODEV;
-+}
++extern atomic_t irq_err_count;
++extern atomic_t irq_mis_count;
+
-+int xenbus_dev_remove(struct device *_dev)
-+{
-+ struct xenbus_device *dev = to_xenbus_device(_dev);
-+ struct xenbus_driver *drv = to_xenbus_driver(_dev->driver);
++#define IO_APIC_IRQ(x) (((x) >= 16) || ((1<<(x)) & io_apic_irqs))
+
-+ DPRINTK("%s", dev->nodename);
++#endif /* _ASM_HW_IRQ_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/hypercall.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/hypercall.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,416 @@
++/******************************************************************************
++ * hypercall.h
++ *
++ * Linux-specific hypervisor handling.
++ *
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ free_otherend_watch(dev);
-+ free_otherend_details(dev);
++#ifndef __HYPERCALL_H__
++#define __HYPERCALL_H__
+
-+ if (drv->remove)
-+ drv->remove(dev);
++#include <linux/string.h> /* memcpy() */
++#include <linux/stringify.h>
+
-+ xenbus_switch_state(dev, XenbusStateClosed);
-+ return 0;
-+}
++#ifndef __HYPERVISOR_H__
++# error "please don't include this file directly"
++#endif
+
-+static void xenbus_dev_shutdown(struct device *_dev)
-+{
-+ struct xenbus_device *dev = to_xenbus_device(_dev);
-+ unsigned long timeout = 5*HZ;
++#ifdef CONFIG_XEN
++#define HYPERCALL_STR(name) \
++ "call hypercall_page + ("__stringify(__HYPERVISOR_##name)" * 32)"
++#else
++#define HYPERCALL_STR(name) \
++ "mov hypercall_stubs,%%eax; " \
++ "add $("__stringify(__HYPERVISOR_##name)" * 32),%%eax; "\
++ "call *%%eax"
++#endif
+
-+ DPRINTK("%s", dev->nodename);
++#define _hypercall0(type, name) \
++({ \
++ type __res; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res) \
++ : \
++ : "memory" ); \
++ __res; \
++})
+
-+ get_device(&dev->dev);
-+ if (dev->state != XenbusStateConnected) {
-+ printk("%s: %s: %s != Connected, skipping\n", __FUNCTION__,
-+ dev->nodename, xenbus_strstate(dev->state));
-+ goto out;
-+ }
-+ xenbus_switch_state(dev, XenbusStateClosing);
-+ timeout = wait_for_completion_timeout(&dev->down, timeout);
-+ if (!timeout)
-+ printk("%s: %s timeout closing device\n", __FUNCTION__, dev->nodename);
-+ out:
-+ put_device(&dev->dev);
-+}
++#define _hypercall1(type, name, a1) \
++({ \
++ type __res; \
++ long __ign1; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=b" (__ign1) \
++ : "1" ((long)(a1)) \
++ : "memory" ); \
++ __res; \
++})
+
-+int xenbus_register_driver_common(struct xenbus_driver *drv,
-+ struct xen_bus_type *bus)
-+{
-+ int ret;
++#define _hypercall2(type, name, a1, a2) \
++({ \
++ type __res; \
++ long __ign1, __ign2; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=b" (__ign1), "=c" (__ign2) \
++ : "1" ((long)(a1)), "2" ((long)(a2)) \
++ : "memory" ); \
++ __res; \
++})
+
-+ if (bus->error)
-+ return bus->error;
++#define _hypercall3(type, name, a1, a2, a3) \
++({ \
++ type __res; \
++ long __ign1, __ign2, __ign3; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=b" (__ign1), "=c" (__ign2), \
++ "=d" (__ign3) \
++ : "1" ((long)(a1)), "2" ((long)(a2)), \
++ "3" ((long)(a3)) \
++ : "memory" ); \
++ __res; \
++})
+
-+ drv->driver.name = drv->name;
-+ drv->driver.bus = &bus->bus;
-+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,10)
-+ drv->driver.owner = drv->owner;
-+#endif
-+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,16)
-+ drv->driver.probe = xenbus_dev_probe;
-+ drv->driver.remove = xenbus_dev_remove;
-+ drv->driver.shutdown = xenbus_dev_shutdown;
-+#endif
++#define _hypercall4(type, name, a1, a2, a3, a4) \
++({ \
++ type __res; \
++ long __ign1, __ign2, __ign3, __ign4; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=b" (__ign1), "=c" (__ign2), \
++ "=d" (__ign3), "=S" (__ign4) \
++ : "1" ((long)(a1)), "2" ((long)(a2)), \
++ "3" ((long)(a3)), "4" ((long)(a4)) \
++ : "memory" ); \
++ __res; \
++})
+
-+ mutex_lock(&xenwatch_mutex);
-+ ret = driver_register(&drv->driver);
-+ mutex_unlock(&xenwatch_mutex);
-+ return ret;
-+}
++#define _hypercall5(type, name, a1, a2, a3, a4, a5) \
++({ \
++ type __res; \
++ long __ign1, __ign2, __ign3, __ign4, __ign5; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=b" (__ign1), "=c" (__ign2), \
++ "=d" (__ign3), "=S" (__ign4), "=D" (__ign5) \
++ : "1" ((long)(a1)), "2" ((long)(a2)), \
++ "3" ((long)(a3)), "4" ((long)(a4)), \
++ "5" ((long)(a5)) \
++ : "memory" ); \
++ __res; \
++})
+
-+int xenbus_register_frontend(struct xenbus_driver *drv)
++static inline int __must_check
++HYPERVISOR_set_trap_table(
++ const trap_info_t *table)
+{
-+ int ret;
-+
-+ drv->read_otherend_details = read_backend_details;
-+
-+ ret = xenbus_register_driver_common(drv, &xenbus_frontend);
-+ if (ret)
-+ return ret;
-+
-+ /* If this driver is loaded as a module wait for devices to attach. */
-+ wait_for_devices(drv);
-+
-+ return 0;
++ return _hypercall1(int, set_trap_table, table);
+}
-+EXPORT_SYMBOL_GPL(xenbus_register_frontend);
+
-+void xenbus_unregister_driver(struct xenbus_driver *drv)
++static inline int __must_check
++HYPERVISOR_mmu_update(
++ mmu_update_t *req, unsigned int count, unsigned int *success_count,
++ domid_t domid)
+{
-+ driver_unregister(&drv->driver);
++ return _hypercall4(int, mmu_update, req, count, success_count, domid);
+}
-+EXPORT_SYMBOL_GPL(xenbus_unregister_driver);
+
-+struct xb_find_info
++static inline int __must_check
++HYPERVISOR_mmuext_op(
++ struct mmuext_op *op, unsigned int count, unsigned int *success_count,
++ domid_t domid)
+{
-+ struct xenbus_device *dev;
-+ const char *nodename;
-+};
++ return _hypercall4(int, mmuext_op, op, count, success_count, domid);
++}
+
-+static int cmp_dev(struct device *dev, void *data)
++static inline int __must_check
++HYPERVISOR_set_gdt(
++ unsigned long *frame_list, unsigned int entries)
+{
-+ struct xenbus_device *xendev = to_xenbus_device(dev);
-+ struct xb_find_info *info = data;
-+
-+ if (!strcmp(xendev->nodename, info->nodename)) {
-+ info->dev = xendev;
-+ get_device(dev);
-+ return 1;
-+ }
-+ return 0;
++ return _hypercall2(int, set_gdt, frame_list, entries);
+}
+
-+struct xenbus_device *xenbus_device_find(const char *nodename,
-+ struct bus_type *bus)
++static inline int __must_check
++HYPERVISOR_stack_switch(
++ unsigned long ss, unsigned long esp)
+{
-+ struct xb_find_info info = { .dev = NULL, .nodename = nodename };
-+
-+ bus_for_each_dev(bus, NULL, &info, cmp_dev);
-+ return info.dev;
++ return _hypercall2(int, stack_switch, ss, esp);
+}
+
-+static int cleanup_dev(struct device *dev, void *data)
++static inline int __must_check
++HYPERVISOR_set_callbacks(
++ unsigned long event_selector, unsigned long event_address,
++ unsigned long failsafe_selector, unsigned long failsafe_address)
+{
-+ struct xenbus_device *xendev = to_xenbus_device(dev);
-+ struct xb_find_info *info = data;
-+ int len = strlen(info->nodename);
-+
-+ DPRINTK("%s", info->nodename);
++ return _hypercall4(int, set_callbacks,
++ event_selector, event_address,
++ failsafe_selector, failsafe_address);
++}
+
-+ /* Match the info->nodename path, or any subdirectory of that path. */
-+ if (strncmp(xendev->nodename, info->nodename, len))
-+ return 0;
++static inline int
++HYPERVISOR_fpu_taskswitch(
++ int set)
++{
++ return _hypercall1(int, fpu_taskswitch, set);
++}
+
-+ /* If the node name is longer, ensure it really is a subdirectory. */
-+ if ((strlen(xendev->nodename) > len) && (xendev->nodename[len] != '/'))
-+ return 0;
++static inline int __must_check
++HYPERVISOR_sched_op_compat(
++ int cmd, unsigned long arg)
++{
++ return _hypercall2(int, sched_op_compat, cmd, arg);
++}
+
-+ info->dev = xendev;
-+ get_device(dev);
-+ return 1;
++static inline int __must_check
++HYPERVISOR_sched_op(
++ int cmd, void *arg)
++{
++ return _hypercall2(int, sched_op, cmd, arg);
+}
+
-+static void xenbus_cleanup_devices(const char *path, struct bus_type *bus)
++static inline long __must_check
++HYPERVISOR_set_timer_op(
++ u64 timeout)
+{
-+ struct xb_find_info info = { .nodename = path };
++ unsigned long timeout_hi = (unsigned long)(timeout>>32);
++ unsigned long timeout_lo = (unsigned long)timeout;
++ return _hypercall2(long, set_timer_op, timeout_lo, timeout_hi);
++}
+
-+ do {
-+ info.dev = NULL;
-+ bus_for_each_dev(bus, NULL, &info, cleanup_dev);
-+ if (info.dev) {
-+ device_unregister(&info.dev->dev);
-+ put_device(&info.dev->dev);
-+ }
-+ } while (info.dev);
++static inline int __must_check
++HYPERVISOR_platform_op(
++ struct xen_platform_op *platform_op)
++{
++ platform_op->interface_version = XENPF_INTERFACE_VERSION;
++ return _hypercall1(int, platform_op, platform_op);
+}
+
-+static void xenbus_dev_release(struct device *dev)
++static inline int __must_check
++HYPERVISOR_set_debugreg(
++ unsigned int reg, unsigned long value)
+{
-+ if (dev)
-+ kfree(to_xenbus_device(dev));
++ return _hypercall2(int, set_debugreg, reg, value);
+}
+
-+static ssize_t xendev_show_nodename(struct device *dev,
-+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,13)
-+ struct device_attribute *attr,
-+#endif
-+ char *buf)
++static inline unsigned long __must_check
++HYPERVISOR_get_debugreg(
++ unsigned int reg)
+{
-+ return sprintf(buf, "%s\n", to_xenbus_device(dev)->nodename);
++ return _hypercall1(unsigned long, get_debugreg, reg);
+}
-+DEVICE_ATTR(nodename, S_IRUSR | S_IRGRP | S_IROTH, xendev_show_nodename, NULL);
+
-+static ssize_t xendev_show_devtype(struct device *dev,
-+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,13)
-+ struct device_attribute *attr,
-+#endif
-+ char *buf)
++static inline int __must_check
++HYPERVISOR_update_descriptor(
++ u64 ma, u64 desc)
+{
-+ return sprintf(buf, "%s\n", to_xenbus_device(dev)->devicetype);
++ return _hypercall4(int, update_descriptor, ma, ma>>32, desc, desc>>32);
+}
-+DEVICE_ATTR(devtype, S_IRUSR | S_IRGRP | S_IROTH, xendev_show_devtype, NULL);
+
++static inline int __must_check
++HYPERVISOR_memory_op(
++ unsigned int cmd, void *arg)
++{
++ return _hypercall2(int, memory_op, cmd, arg);
++}
+
-+int xenbus_probe_node(struct xen_bus_type *bus,
-+ const char *type,
-+ const char *nodename)
++static inline int __must_check
++HYPERVISOR_multicall(
++ multicall_entry_t *call_list, unsigned int nr_calls)
+{
-+ int err;
-+ struct xenbus_device *xendev;
-+ size_t stringlen;
-+ char *tmpstring;
++ return _hypercall2(int, multicall, call_list, nr_calls);
++}
+
-+ enum xenbus_state state = xenbus_read_driver_state(nodename);
++static inline int __must_check
++HYPERVISOR_update_va_mapping(
++ unsigned long va, pte_t new_val, unsigned long flags)
++{
++ unsigned long pte_hi = 0;
++#ifdef CONFIG_X86_PAE
++ pte_hi = new_val.pte_high;
++#endif
++ return _hypercall4(int, update_va_mapping, va,
++ new_val.pte_low, pte_hi, flags);
++}
+
-+ if (bus->error)
-+ return bus->error;
++static inline int __must_check
++HYPERVISOR_event_channel_op(
++ int cmd, void *arg)
++{
++ int rc = _hypercall2(int, event_channel_op, cmd, arg);
+
-+ if (state != XenbusStateInitialising) {
-+ /* Device is not new, so ignore it. This can happen if a
-+ device is going away after switching to Closed. */
-+ return 0;
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ struct evtchn_op op;
++ op.cmd = cmd;
++ memcpy(&op.u, arg, sizeof(op.u));
++ rc = _hypercall1(int, event_channel_op_compat, &op);
++ memcpy(arg, &op.u, sizeof(op.u));
+ }
++#endif
+
-+ stringlen = strlen(nodename) + 1 + strlen(type) + 1;
-+ xendev = kzalloc(sizeof(*xendev) + stringlen, GFP_KERNEL);
-+ if (!xendev)
-+ return -ENOMEM;
-+
-+ xendev->state = XenbusStateInitialising;
-+
-+ /* Copy the strings into the extra space. */
-+
-+ tmpstring = (char *)(xendev + 1);
-+ strcpy(tmpstring, nodename);
-+ xendev->nodename = tmpstring;
-+
-+ tmpstring += strlen(tmpstring) + 1;
-+ strcpy(tmpstring, type);
-+ xendev->devicetype = tmpstring;
-+ init_completion(&xendev->down);
-+
-+ xendev->dev.parent = &bus->dev;
-+ xendev->dev.bus = &bus->bus;
-+ xendev->dev.release = xenbus_dev_release;
-+
-+ err = bus->get_bus_id(xendev->dev.bus_id, xendev->nodename);
-+ if (err)
-+ goto fail;
-+
-+ /* Register with generic device framework. */
-+ err = device_register(&xendev->dev);
-+ if (err)
-+ goto fail;
++ return rc;
++}
+
-+ err = device_create_file(&xendev->dev, &dev_attr_nodename);
-+ if (err)
-+ goto unregister;
-+ err = device_create_file(&xendev->dev, &dev_attr_devtype);
-+ if (err)
-+ goto unregister;
++static inline int __must_check
++HYPERVISOR_acm_op(
++ int cmd, void *arg)
++{
++ return _hypercall2(int, acm_op, cmd, arg);
++}
+
-+ return 0;
-+unregister:
-+ device_remove_file(&xendev->dev, &dev_attr_nodename);
-+ device_remove_file(&xendev->dev, &dev_attr_devtype);
-+ device_unregister(&xendev->dev);
-+fail:
-+ kfree(xendev);
-+ return err;
++static inline int __must_check
++HYPERVISOR_xen_version(
++ int cmd, void *arg)
++{
++ return _hypercall2(int, xen_version, cmd, arg);
+}
+
-+/* device/<typename>/<name> */
-+static int xenbus_probe_frontend(const char *type, const char *name)
++static inline int __must_check
++HYPERVISOR_console_io(
++ int cmd, unsigned int count, char *str)
+{
-+ char *nodename;
-+ int err;
++ return _hypercall3(int, console_io, cmd, count, str);
++}
+
-+ nodename = kasprintf(GFP_KERNEL, "%s/%s/%s", xenbus_frontend.root, type, name);
-+ if (!nodename)
-+ return -ENOMEM;
++static inline int __must_check
++HYPERVISOR_physdev_op(
++ int cmd, void *arg)
++{
++ int rc = _hypercall2(int, physdev_op, cmd, arg);
+
-+ DPRINTK("%s", nodename);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ struct physdev_op op;
++ op.cmd = cmd;
++ memcpy(&op.u, arg, sizeof(op.u));
++ rc = _hypercall1(int, physdev_op_compat, &op);
++ memcpy(arg, &op.u, sizeof(op.u));
++ }
++#endif
+
-+ err = xenbus_probe_node(&xenbus_frontend, type, nodename);
-+ kfree(nodename);
-+ return err;
++ return rc;
+}
+
-+static int xenbus_probe_device_type(struct xen_bus_type *bus, const char *type)
++static inline int __must_check
++HYPERVISOR_grant_table_op(
++ unsigned int cmd, void *uop, unsigned int count)
+{
-+ int err = 0;
-+ char **dir;
-+ unsigned int dir_n = 0;
-+ int i;
++ return _hypercall3(int, grant_table_op, cmd, uop, count);
++}
+
-+ dir = xenbus_directory(XBT_NIL, bus->root, type, &dir_n);
-+ if (IS_ERR(dir))
-+ return PTR_ERR(dir);
++static inline int __must_check
++HYPERVISOR_update_va_mapping_otherdomain(
++ unsigned long va, pte_t new_val, unsigned long flags, domid_t domid)
++{
++ unsigned long pte_hi = 0;
++#ifdef CONFIG_X86_PAE
++ pte_hi = new_val.pte_high;
++#endif
++ return _hypercall5(int, update_va_mapping_otherdomain, va,
++ new_val.pte_low, pte_hi, flags, domid);
++}
+
-+ for (i = 0; i < dir_n; i++) {
-+ err = bus->probe(type, dir[i]);
-+ if (err)
-+ break;
-+ }
-+ kfree(dir);
-+ return err;
++static inline int __must_check
++HYPERVISOR_vm_assist(
++ unsigned int cmd, unsigned int type)
++{
++ return _hypercall2(int, vm_assist, cmd, type);
+}
+
-+int xenbus_probe_devices(struct xen_bus_type *bus)
++static inline int __must_check
++HYPERVISOR_vcpu_op(
++ int cmd, unsigned int vcpuid, void *extra_args)
+{
-+ int err = 0;
-+ char **dir;
-+ unsigned int i, dir_n;
++ return _hypercall3(int, vcpu_op, cmd, vcpuid, extra_args);
++}
+
-+ if (bus->error)
-+ return bus->error;
++static inline int __must_check
++HYPERVISOR_suspend(
++ unsigned long srec)
++{
++ struct sched_shutdown sched_shutdown = {
++ .reason = SHUTDOWN_suspend
++ };
+
-+ dir = xenbus_directory(XBT_NIL, bus->root, "", &dir_n);
-+ if (IS_ERR(dir))
-+ return PTR_ERR(dir);
++ int rc = _hypercall3(int, sched_op, SCHEDOP_shutdown,
++ &sched_shutdown, srec);
+
-+ for (i = 0; i < dir_n; i++) {
-+ err = xenbus_probe_device_type(bus, dir[i]);
-+ if (err)
-+ break;
-+ }
-+ kfree(dir);
-+ return err;
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (rc == -ENOSYS)
++ rc = _hypercall3(int, sched_op_compat, SCHEDOP_shutdown,
++ SHUTDOWN_suspend, srec);
++#endif
++
++ return rc;
+}
+
-+static unsigned int char_count(const char *str, char c)
++#if CONFIG_XEN_COMPAT <= 0x030002
++static inline int
++HYPERVISOR_nmi_op(
++ unsigned long op, void *arg)
+{
-+ unsigned int i, ret = 0;
++ return _hypercall2(int, nmi_op, op, arg);
++}
++#endif
+
-+ for (i = 0; str[i]; i++)
-+ if (str[i] == c)
-+ ret++;
-+ return ret;
++#ifndef CONFIG_XEN
++static inline unsigned long __must_check
++HYPERVISOR_hvm_op(
++ int op, void *arg)
++{
++ return _hypercall2(unsigned long, hvm_op, op, arg);
+}
++#endif
+
-+static int strsep_len(const char *str, char c, unsigned int len)
++static inline int __must_check
++HYPERVISOR_callback_op(
++ int cmd, const void *arg)
+{
-+ unsigned int i;
++ return _hypercall2(int, callback_op, cmd, arg);
++}
+
-+ for (i = 0; str[i]; i++)
-+ if (str[i] == c) {
-+ if (len == 0)
-+ return i;
-+ len--;
-+ }
-+ return (len == 0) ? i : -ERANGE;
++static inline int __must_check
++HYPERVISOR_xenoprof_op(
++ int op, void *arg)
++{
++ return _hypercall2(int, xenoprof_op, op, arg);
+}
+
-+void dev_changed(const char *node, struct xen_bus_type *bus)
++static inline int __must_check
++HYPERVISOR_kexec_op(
++ unsigned long op, void *args)
+{
-+ int exists, rootlen;
-+ struct xenbus_device *dev;
-+ char type[BUS_ID_SIZE];
-+ const char *p, *root;
++ return _hypercall2(int, kexec_op, op, args);
++}
+
-+ if (bus->error || char_count(node, '/') < 2)
-+ return;
+
-+ exists = xenbus_exists(XBT_NIL, node, "");
-+ if (!exists) {
-+ xenbus_cleanup_devices(node, &bus->bus);
-+ return;
-+ }
+
-+ /* backend/<type>/... or device/<type>/... */
-+ p = strchr(node, '/') + 1;
-+ snprintf(type, BUS_ID_SIZE, "%.*s", (int)strcspn(p, "/"), p);
-+ type[BUS_ID_SIZE-1] = '\0';
++#endif /* __HYPERCALL_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/hypervisor.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/hypervisor.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,259 @@
++/******************************************************************************
++ * hypervisor.h
++ *
++ * Linux-specific hypervisor handling.
++ *
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+ rootlen = strsep_len(node, '/', bus->levels);
-+ if (rootlen < 0)
-+ return;
-+ root = kasprintf(GFP_KERNEL, "%.*s", rootlen, node);
-+ if (!root)
-+ return;
++#ifndef __HYPERVISOR_H__
++#define __HYPERVISOR_H__
+
-+ dev = xenbus_device_find(root, &bus->bus);
-+ if (!dev)
-+ xenbus_probe_node(bus, type, root);
-+ else
-+ put_device(&dev->dev);
++#include <linux/types.h>
++#include <linux/kernel.h>
++#include <linux/version.h>
++#include <linux/errno.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/platform.h>
++#include <xen/interface/event_channel.h>
++#include <xen/interface/physdev.h>
++#include <xen/interface/sched.h>
++#include <xen/interface/nmi.h>
++#include <asm/ptrace.h>
++#include <asm/page.h>
++#if defined(__i386__)
++# ifdef CONFIG_X86_PAE
++# include <asm-generic/pgtable-nopud.h>
++# else
++# include <asm-generic/pgtable-nopmd.h>
++# endif
++#elif defined(__x86_64__) && LINUX_VERSION_CODE < KERNEL_VERSION(2,6,11)
++# include <asm-generic/pgtable-nopud.h>
++#endif
+
-+ kfree(root);
-+}
++extern shared_info_t *HYPERVISOR_shared_info;
+
-+static void frontend_changed(struct xenbus_watch *watch,
-+ const char **vec, unsigned int len)
-+{
-+ DPRINTK("");
++#define vcpu_info(cpu) (HYPERVISOR_shared_info->vcpu_info + (cpu))
++#ifdef CONFIG_SMP
++#define current_vcpu_info() vcpu_info(smp_processor_id())
++#else
++#define current_vcpu_info() vcpu_info(0)
++#endif
+
-+ dev_changed(vec[XS_WATCH_PATH], &xenbus_frontend);
-+}
++#ifdef CONFIG_X86_32
++extern unsigned long hypervisor_virt_start;
++#endif
+
-+/* We watch for devices appearing and vanishing. */
-+static struct xenbus_watch fe_watch = {
-+ .node = "device",
-+ .callback = frontend_changed,
-+};
++/* arch/xen/i386/kernel/setup.c */
++extern start_info_t *xen_start_info;
++#ifdef CONFIG_XEN_PRIVILEGED_GUEST
++#define is_initial_xendomain() (xen_start_info->flags & SIF_INITDOMAIN)
++#else
++#define is_initial_xendomain() 0
++#endif
+
-+static int suspend_dev(struct device *dev, void *data)
-+{
-+ int err = 0;
-+ struct xenbus_driver *drv;
-+ struct xenbus_device *xdev;
++/* arch/xen/kernel/evtchn.c */
++/* Force a proper event-channel callback from Xen. */
++void force_evtchn_callback(void);
+
-+ DPRINTK("");
++/* arch/xen/kernel/process.c */
++void xen_cpu_idle (void);
+
-+ if (dev->driver == NULL)
-+ return 0;
-+ drv = to_xenbus_driver(dev->driver);
-+ xdev = container_of(dev, struct xenbus_device, dev);
-+ if (drv->suspend)
-+ err = drv->suspend(xdev);
-+ if (err)
-+ printk(KERN_WARNING
-+ "xenbus: suspend %s failed: %i\n", dev->bus_id, err);
-+ return 0;
-+}
++/* arch/xen/i386/kernel/hypervisor.c */
++void do_hypervisor_callback(struct pt_regs *regs);
+
-+static int suspend_cancel_dev(struct device *dev, void *data)
-+{
-+ int err = 0;
-+ struct xenbus_driver *drv;
-+ struct xenbus_device *xdev;
++/* arch/xen/i386/mm/hypervisor.c */
++/*
++ * NB. ptr values should be PHYSICAL, not MACHINE. 'vals' should be already
++ * be MACHINE addresses.
++ */
+
-+ DPRINTK("");
++void xen_pt_switch(unsigned long ptr);
++void xen_new_user_pt(unsigned long ptr); /* x86_64 only */
++void xen_load_gs(unsigned int selector); /* x86_64 only */
++void xen_tlb_flush(void);
++void xen_invlpg(unsigned long ptr);
+
-+ if (dev->driver == NULL)
-+ return 0;
-+ drv = to_xenbus_driver(dev->driver);
-+ xdev = container_of(dev, struct xenbus_device, dev);
-+ if (drv->suspend_cancel)
-+ err = drv->suspend_cancel(xdev);
-+ if (err)
-+ printk(KERN_WARNING
-+ "xenbus: suspend_cancel %s failed: %i\n",
-+ dev->bus_id, err);
-+ return 0;
-+}
++void xen_l1_entry_update(pte_t *ptr, pte_t val);
++void xen_l2_entry_update(pmd_t *ptr, pmd_t val);
++void xen_l3_entry_update(pud_t *ptr, pud_t val); /* x86_64/PAE */
++void xen_l4_entry_update(pgd_t *ptr, pgd_t val); /* x86_64 only */
++void xen_pgd_pin(unsigned long ptr);
++void xen_pgd_unpin(unsigned long ptr);
+
-+static int resume_dev(struct device *dev, void *data)
-+{
-+ int err;
-+ struct xenbus_driver *drv;
-+ struct xenbus_device *xdev;
++void xen_set_ldt(const void *ptr, unsigned int ents);
+
-+ DPRINTK("");
++#ifdef CONFIG_SMP
++#include <linux/cpumask.h>
++void xen_tlb_flush_all(void);
++void xen_invlpg_all(unsigned long ptr);
++void xen_tlb_flush_mask(cpumask_t *mask);
++void xen_invlpg_mask(cpumask_t *mask, unsigned long ptr);
++#endif
+
-+ if (dev->driver == NULL)
-+ return 0;
++/* Returns zero on success else negative errno. */
++int xen_create_contiguous_region(
++ unsigned long vstart, unsigned int order, unsigned int address_bits);
++void xen_destroy_contiguous_region(
++ unsigned long vstart, unsigned int order);
+
-+ drv = to_xenbus_driver(dev->driver);
-+ xdev = container_of(dev, struct xenbus_device, dev);
++struct page;
+
-+ err = talk_to_otherend(xdev);
-+ if (err) {
-+ printk(KERN_WARNING
-+ "xenbus: resume (talk_to_otherend) %s failed: %i\n",
-+ dev->bus_id, err);
-+ return err;
-+ }
++int xen_limit_pages_to_max_mfn(
++ struct page *pages, unsigned int order, unsigned int address_bits);
+
-+ xdev->state = XenbusStateInitialising;
++/* Turn jiffies into Xen system time. */
++u64 jiffies_to_st(unsigned long jiffies);
+
-+ if (drv->resume) {
-+ err = drv->resume(xdev);
-+ if (err) {
-+ printk(KERN_WARNING
-+ "xenbus: resume %s failed: %i\n",
-+ dev->bus_id, err);
-+ return err;
-+ }
-+ }
++#ifdef CONFIG_XEN_SCRUB_PAGES
++void scrub_pages(void *, unsigned int);
++#else
++#define scrub_pages(_p,_n) ((void)0)
++#endif
+
-+ err = watch_otherend(xdev);
-+ if (err) {
-+ printk(KERN_WARNING
-+ "xenbus_probe: resume (watch_otherend) %s failed: "
-+ "%d.\n", dev->bus_id, err);
-+ return err;
-+ }
++#include <xen/hypercall.h>
+
-+ return 0;
-+}
++#if defined(CONFIG_X86_64)
++#define MULTI_UVMFLAGS_INDEX 2
++#define MULTI_UVMDOMID_INDEX 3
++#else
++#define MULTI_UVMFLAGS_INDEX 3
++#define MULTI_UVMDOMID_INDEX 4
++#endif
+
-+void xenbus_suspend(void)
++#ifdef CONFIG_XEN
++#define is_running_on_xen() 1
++#else
++extern char *hypercall_stubs;
++#define is_running_on_xen() (!!hypercall_stubs)
++#endif
++
++static inline int
++HYPERVISOR_yield(
++ void)
+{
-+ DPRINTK("");
++ int rc = HYPERVISOR_sched_op(SCHEDOP_yield, NULL);
+
-+ if (!xenbus_frontend.error)
-+ bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_dev);
-+ xenbus_backend_suspend(suspend_dev);
-+ xs_suspend();
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (rc == -ENOSYS)
++ rc = HYPERVISOR_sched_op_compat(SCHEDOP_yield, 0);
++#endif
++
++ return rc;
+}
-+EXPORT_SYMBOL_GPL(xenbus_suspend);
+
-+void xenbus_resume(void)
++static inline int
++HYPERVISOR_block(
++ void)
+{
-+ xb_init_comms();
-+ xs_resume();
-+ if (!xenbus_frontend.error)
-+ bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, resume_dev);
-+ xenbus_backend_resume(resume_dev);
++ int rc = HYPERVISOR_sched_op(SCHEDOP_block, NULL);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (rc == -ENOSYS)
++ rc = HYPERVISOR_sched_op_compat(SCHEDOP_block, 0);
++#endif
++
++ return rc;
+}
-+EXPORT_SYMBOL_GPL(xenbus_resume);
+
-+void xenbus_suspend_cancel(void)
++static inline void /*__noreturn*/
++HYPERVISOR_shutdown(
++ unsigned int reason)
+{
-+ xs_suspend_cancel();
-+ if (!xenbus_frontend.error)
-+ bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_cancel_dev);
-+ xenbus_backend_resume(suspend_cancel_dev);
++ struct sched_shutdown sched_shutdown = {
++ .reason = reason
++ };
++
++ VOID(HYPERVISOR_sched_op(SCHEDOP_shutdown, &sched_shutdown));
++#if CONFIG_XEN_COMPAT <= 0x030002
++ VOID(HYPERVISOR_sched_op_compat(SCHEDOP_shutdown, reason));
++#endif
++ /* Don't recurse needlessly. */
++ BUG_ON(reason != SHUTDOWN_crash);
++ for(;;);
+}
-+EXPORT_SYMBOL_GPL(xenbus_suspend_cancel);
+
-+/* A flag to determine if xenstored is 'ready' (i.e. has started) */
-+int xenstored_ready = 0;
++static inline int __must_check
++HYPERVISOR_poll(
++ evtchn_port_t *ports, unsigned int nr_ports, u64 timeout)
++{
++ int rc;
++ struct sched_poll sched_poll = {
++ .nr_ports = nr_ports,
++ .timeout = jiffies_to_st(timeout)
++ };
++ set_xen_guest_handle(sched_poll.ports, ports);
+
++ rc = HYPERVISOR_sched_op(SCHEDOP_poll, &sched_poll);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (rc == -ENOSYS)
++ rc = HYPERVISOR_sched_op_compat(SCHEDOP_yield, 0);
++#endif
+
-+int register_xenstore_notifier(struct notifier_block *nb)
-+{
-+ int ret = 0;
++ return rc;
++}
+
-+ if (xenstored_ready > 0)
-+ ret = nb->notifier_call(nb, 0, NULL);
-+ else
-+ atomic_notifier_chain_register(&xenstore_chain, nb);
++#ifdef CONFIG_XEN
+
-+ return ret;
++static inline void
++MULTI_update_va_mapping(
++ multicall_entry_t *mcl, unsigned long va,
++ pte_t new_val, unsigned long flags)
++{
++ mcl->op = __HYPERVISOR_update_va_mapping;
++ mcl->args[0] = va;
++#if defined(CONFIG_X86_64)
++ mcl->args[1] = new_val.pte;
++#elif defined(CONFIG_X86_PAE)
++ mcl->args[1] = new_val.pte_low;
++ mcl->args[2] = new_val.pte_high;
++#else
++ mcl->args[1] = new_val.pte_low;
++ mcl->args[2] = 0;
++#endif
++ mcl->args[MULTI_UVMFLAGS_INDEX] = flags;
+}
-+EXPORT_SYMBOL_GPL(register_xenstore_notifier);
+
-+void unregister_xenstore_notifier(struct notifier_block *nb)
++static inline void
++MULTI_grant_table_op(multicall_entry_t *mcl, unsigned int cmd,
++ void *uop, unsigned int count)
+{
-+ atomic_notifier_chain_unregister(&xenstore_chain, nb);
++ mcl->op = __HYPERVISOR_grant_table_op;
++ mcl->args[0] = cmd;
++ mcl->args[1] = (unsigned long)uop;
++ mcl->args[2] = count;
+}
-+EXPORT_SYMBOL_GPL(unregister_xenstore_notifier);
+
++#else /* !defined(CONFIG_XEN) */
++
++/* Multicalls not supported for HVM guests. */
++#define MULTI_update_va_mapping(a,b,c,d) ((void)0)
++#define MULTI_grant_table_op(a,b,c,d) ((void)0)
++
++#endif
++
++#endif /* __HYPERVISOR_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/io.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/io.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,389 @@
++#ifndef _ASM_IO_H
++#define _ASM_IO_H
++
++#include <linux/string.h>
++#include <linux/compiler.h>
++
++/*
++ * This file contains the definitions for the x86 IO instructions
++ * inb/inw/inl/outb/outw/outl and the "string versions" of the same
++ * (insb/insw/insl/outsb/outsw/outsl). You can also use "pausing"
++ * versions of the single-IO instructions (inb_p/inw_p/..).
++ *
++ * This file is not meant to be obfuscating: it's just complicated
++ * to (a) handle it all in a way that makes gcc able to optimize it
++ * as well as possible and (b) trying to avoid writing the same thing
++ * over and over again with slight variations and possibly making a
++ * mistake somewhere.
++ */
+
-+void xenbus_probe(void *unused)
-+{
-+ BUG_ON((xenstored_ready <= 0));
++/*
++ * Thanks to James van Artsdalen for a better timing-fix than
++ * the two short jumps: using outb's to a nonexistent port seems
++ * to guarantee better timings even on fast machines.
++ *
++ * On the other hand, I'd like to be sure of a non-existent port:
++ * I feel a bit unsafe about using 0x80 (should be safe, though)
++ *
++ * Linus
++ */
+
-+ /* Enumerate devices in xenstore and watch for changes. */
-+ xenbus_probe_devices(&xenbus_frontend);
-+ register_xenbus_watch(&fe_watch);
-+ xenbus_backend_probe_and_watch();
++ /*
++ * Bit simplified and optimized by Jan Hubicka
++ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999.
++ *
++ * isa_memset_io, isa_memcpy_fromio, isa_memcpy_toio added,
++ * isa_read[wl] and isa_write[wl] fixed
++ * - Arnaldo Carvalho de Melo <acme at conectiva.com.br>
++ */
+
-+ /* Notify others that xenstore is up */
-+ atomic_notifier_call_chain(&xenstore_chain, 0, NULL);
-+}
++#define IO_SPACE_LIMIT 0xffff
+
++#define XQUAD_PORTIO_BASE 0xfe400000
++#define XQUAD_PORTIO_QUAD 0x40000 /* 256k per quad. */
+
-+#if defined(CONFIG_PROC_FS) && defined(CONFIG_XEN_PRIVILEGED_GUEST)
-+static struct file_operations xsd_kva_fops;
-+static struct proc_dir_entry *xsd_kva_intf;
-+static struct proc_dir_entry *xsd_port_intf;
++#ifdef __KERNEL__
+
-+static int xsd_kva_mmap(struct file *file, struct vm_area_struct *vma)
-+{
-+ size_t size = vma->vm_end - vma->vm_start;
++#include <asm-generic/iomap.h>
+
-+ if ((size > PAGE_SIZE) || (vma->vm_pgoff != 0))
-+ return -EINVAL;
++#include <linux/vmalloc.h>
++#include <asm/fixmap.h>
+
-+ if (remap_pfn_range(vma, vma->vm_start, mfn_to_pfn(xen_store_mfn),
-+ size, vma->vm_page_prot))
-+ return -EAGAIN;
++/*
++ * Convert a physical pointer to a virtual kernel pointer for /dev/mem
++ * access
++ */
++#define xlate_dev_mem_ptr(p) __va(p)
+
-+ return 0;
-+}
++/*
++ * Convert a virtual cached pointer to an uncached pointer
++ */
++#define xlate_dev_kmem_ptr(p) p
+
-+static int xsd_kva_read(char *page, char **start, off_t off,
-+ int count, int *eof, void *data)
++/**
++ * virt_to_phys - map virtual addresses to physical
++ * @address: address to remap
++ *
++ * The returned physical address is the physical (CPU) mapping for
++ * the memory address given. It is only valid to use this function on
++ * addresses directly mapped or allocated via kmalloc.
++ *
++ * This function does not give bus mappings for DMA transfers. In
++ * almost all conceivable cases a device driver should not be using
++ * this function
++ */
++
++static inline unsigned long virt_to_phys(volatile void * address)
+{
-+ int len;
-+
-+ len = sprintf(page, "0x%p", xen_store_interface);
-+ *eof = 1;
-+ return len;
++ return __pa(address);
+}
+
-+static int xsd_port_read(char *page, char **start, off_t off,
-+ int count, int *eof, void *data)
-+{
-+ int len;
-+
-+ len = sprintf(page, "%d", xen_store_evtchn);
-+ *eof = 1;
-+ return len;
-+}
-+#endif
++/**
++ * phys_to_virt - map physical address to virtual
++ * @address: address to remap
++ *
++ * The returned virtual address is a current CPU mapping for
++ * the memory address given. It is only valid to use this function on
++ * addresses that have a kernel mapping
++ *
++ * This function does not handle bus mappings for DMA transfers. In
++ * almost all conceivable cases a device driver should not be using
++ * this function
++ */
+
-+static int xenbus_probe_init(void)
++static inline void * phys_to_virt(unsigned long address)
+{
-+ int err = 0;
-+ unsigned long page = 0;
-+
-+ DPRINTK("");
-+
-+ if (!is_running_on_xen())
-+ return -ENODEV;
-+
-+ /* Register ourselves with the kernel bus subsystem */
-+ xenbus_frontend.error = bus_register(&xenbus_frontend.bus);
-+ if (xenbus_frontend.error)
-+ printk(KERN_WARNING
-+ "XENBUS: Error registering frontend bus: %i\n",
-+ xenbus_frontend.error);
-+ xenbus_backend_bus_register();
-+
-+ /*
-+ * Domain0 doesn't have a store_evtchn or store_mfn yet.
-+ */
-+ if (is_initial_xendomain()) {
-+ struct evtchn_alloc_unbound alloc_unbound;
-+
-+ /* Allocate page. */
-+ page = get_zeroed_page(GFP_KERNEL);
-+ if (!page)
-+ return -ENOMEM;
++ return __va(address);
++}
+
-+ xen_store_mfn = xen_start_info->store_mfn =
-+ pfn_to_mfn(virt_to_phys((void *)page) >>
-+ PAGE_SHIFT);
++/*
++ * Change "struct page" to physical address.
++ */
++#define page_to_pseudophys(page) ((dma_addr_t)page_to_pfn(page) << PAGE_SHIFT)
++#define page_to_phys(page) (phys_to_machine(page_to_pseudophys(page)))
++#define page_to_bus(page) (phys_to_machine(page_to_pseudophys(page)))
+
-+ /* Next allocate a local port which xenstored can bind to */
-+ alloc_unbound.dom = DOMID_SELF;
-+ alloc_unbound.remote_dom = 0;
++#define bio_to_pseudophys(bio) (page_to_pseudophys(bio_page((bio))) + \
++ (unsigned long) bio_offset((bio)))
++#define bvec_to_pseudophys(bv) (page_to_pseudophys((bv)->bv_page) + \
++ (unsigned long) (bv)->bv_offset)
+
-+ err = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
-+ &alloc_unbound);
-+ if (err == -ENOSYS)
-+ goto err;
-+ BUG_ON(err);
-+ xen_store_evtchn = xen_start_info->store_evtchn =
-+ alloc_unbound.port;
++#define BIOVEC_PHYS_MERGEABLE(vec1, vec2) \
++ (((bvec_to_phys((vec1)) + (vec1)->bv_len) == bvec_to_phys((vec2))) && \
++ ((bvec_to_pseudophys((vec1)) + (vec1)->bv_len) == \
++ bvec_to_pseudophys((vec2))))
+
-+#if defined(CONFIG_PROC_FS) && defined(CONFIG_XEN_PRIVILEGED_GUEST)
-+ /* And finally publish the above info in /proc/xen */
-+ xsd_kva_intf = create_xen_proc_entry("xsd_kva", 0600);
-+ if (xsd_kva_intf) {
-+ memcpy(&xsd_kva_fops, xsd_kva_intf->proc_fops,
-+ sizeof(xsd_kva_fops));
-+ xsd_kva_fops.mmap = xsd_kva_mmap;
-+ xsd_kva_intf->proc_fops = &xsd_kva_fops;
-+ xsd_kva_intf->read_proc = xsd_kva_read;
-+ }
-+ xsd_port_intf = create_xen_proc_entry("xsd_port", 0400);
-+ if (xsd_port_intf)
-+ xsd_port_intf->read_proc = xsd_port_read;
-+#endif
-+ xen_store_interface = mfn_to_virt(xen_store_mfn);
-+ } else {
-+ xenstored_ready = 1;
-+#ifdef CONFIG_XEN
-+ xen_store_evtchn = xen_start_info->store_evtchn;
-+ xen_store_mfn = xen_start_info->store_mfn;
-+ xen_store_interface = mfn_to_virt(xen_store_mfn);
-+#else
-+ xen_store_evtchn = hvm_get_parameter(HVM_PARAM_STORE_EVTCHN);
-+ xen_store_mfn = hvm_get_parameter(HVM_PARAM_STORE_PFN);
-+ xen_store_interface = ioremap(xen_store_mfn << PAGE_SHIFT,
-+ PAGE_SIZE);
-+#endif
-+ }
++extern void __iomem * __ioremap(unsigned long offset, unsigned long size, unsigned long flags);
+
++/**
++ * ioremap - map bus memory into CPU space
++ * @offset: bus address of the memory
++ * @size: size of the resource to map
++ *
++ * ioremap performs a platform specific sequence of operations to
++ * make bus memory CPU accessible via the readb/readw/readl/writeb/
++ * writew/writel functions and the other mmio helpers. The returned
++ * address is not guaranteed to be usable directly as a virtual
++ * address.
++ */
+
-+ xenbus_dev_init();
++static inline void __iomem * ioremap(unsigned long offset, unsigned long size)
++{
++ return __ioremap(offset, size, 0);
++}
+
-+ /* Initialize the interface to xenstore. */
-+ err = xs_init();
-+ if (err) {
-+ printk(KERN_WARNING
-+ "XENBUS: Error initializing xenstore comms: %i\n", err);
-+ goto err;
-+ }
++extern void __iomem * ioremap_nocache(unsigned long offset, unsigned long size);
++extern void iounmap(volatile void __iomem *addr);
+
-+ /* Register ourselves with the kernel device subsystem */
-+ if (!xenbus_frontend.error) {
-+ xenbus_frontend.error = device_register(&xenbus_frontend.dev);
-+ if (xenbus_frontend.error) {
-+ bus_unregister(&xenbus_frontend.bus);
-+ printk(KERN_WARNING
-+ "XENBUS: Error registering frontend device: %i\n",
-+ xenbus_frontend.error);
-+ }
-+ }
-+ xenbus_backend_device_register();
++/*
++ * bt_ioremap() and bt_iounmap() are for temporary early boot-time
++ * mappings, before the real ioremap() is functional.
++ * A boot-time mapping is currently limited to at most 16 pages.
++ */
++extern void *bt_ioremap(unsigned long offset, unsigned long size);
++extern void bt_iounmap(void *addr, unsigned long size);
+
-+ if (!is_initial_xendomain())
-+ xenbus_probe(NULL);
++/* Use early IO mappings for DMI because it's initialized early */
++#define dmi_ioremap bt_ioremap
++#define dmi_iounmap bt_iounmap
++#define dmi_alloc alloc_bootmem
+
-+ return 0;
++/*
++ * ISA I/O bus memory addresses are 1:1 with the physical address.
++ */
++#define isa_virt_to_bus(_x) ({ BUG(); virt_to_bus(_x); })
++#define isa_page_to_bus(_x) isa_page_to_bus_is_UNSUPPORTED->x
++#define isa_bus_to_virt(_x) (void *)(__fix_to_virt(FIX_ISAMAP_BEGIN) + (_x))
+
-+ err:
-+ if (page)
-+ free_page(page);
++/*
++ * However PCI ones are not necessarily 1:1 and therefore these interfaces
++ * are forbidden in portable PCI drivers.
++ *
++ * Allow them on x86 for legacy drivers, though.
++ */
++#define virt_to_bus(_x) phys_to_machine(__pa(_x))
++#define bus_to_virt(_x) __va(machine_to_phys(_x))
+
-+ /*
-+ * Do not unregister the xenbus front/backend buses here. The buses
-+ * must exist because front/backend drivers will use them when they are
-+ * registered.
-+ */
++/*
++ * readX/writeX() are used to access memory mapped devices. On some
++ * architectures the memory mapped IO stuff needs to be accessed
++ * differently. On the x86 architecture, we just read/write the
++ * memory location directly.
++ */
+
-+ return err;
++static inline unsigned char readb(const volatile void __iomem *addr)
++{
++ return *(volatile unsigned char __force *) addr;
+}
-+
-+#ifdef CONFIG_XEN
-+postcore_initcall(xenbus_probe_init);
-+MODULE_LICENSE("Dual BSD/GPL");
-+#else
-+int xenbus_init(void)
++static inline unsigned short readw(const volatile void __iomem *addr)
+{
-+ return xenbus_probe_init();
++ return *(volatile unsigned short __force *) addr;
+}
-+#endif
-+
-+static int is_disconnected_device(struct device *dev, void *data)
++static inline unsigned int readl(const volatile void __iomem *addr)
+{
-+ struct xenbus_device *xendev = to_xenbus_device(dev);
-+ struct device_driver *drv = data;
-+ struct xenbus_driver *xendrv;
-+
-+ /*
-+ * A device with no driver will never connect. We care only about
-+ * devices which should currently be in the process of connecting.
-+ */
-+ if (!dev->driver)
-+ return 0;
-+
-+ /* Is this search limited to a particular driver? */
-+ if (drv && (dev->driver != drv))
-+ return 0;
-+
-+ xendrv = to_xenbus_driver(dev->driver);
-+ return (xendev->state < XenbusStateConnected ||
-+ (xendrv->is_ready && !xendrv->is_ready(xendev)));
++ return *(volatile unsigned int __force *) addr;
+}
++#define readb_relaxed(addr) readb(addr)
++#define readw_relaxed(addr) readw(addr)
++#define readl_relaxed(addr) readl(addr)
++#define __raw_readb readb
++#define __raw_readw readw
++#define __raw_readl readl
+
-+static int exists_disconnected_device(struct device_driver *drv)
++static inline void writeb(unsigned char b, volatile void __iomem *addr)
+{
-+ if (xenbus_frontend.error)
-+ return xenbus_frontend.error;
-+ return bus_for_each_dev(&xenbus_frontend.bus, NULL, drv,
-+ is_disconnected_device);
++ *(volatile unsigned char __force *) addr = b;
+}
-+
-+static int print_device_status(struct device *dev, void *data)
++static inline void writew(unsigned short b, volatile void __iomem *addr)
+{
-+ struct xenbus_device *xendev = to_xenbus_device(dev);
-+ struct device_driver *drv = data;
-+
-+ /* Is this operation limited to a particular driver? */
-+ if (drv && (dev->driver != drv))
-+ return 0;
++ *(volatile unsigned short __force *) addr = b;
++}
++static inline void writel(unsigned int b, volatile void __iomem *addr)
++{
++ *(volatile unsigned int __force *) addr = b;
++}
++#define __raw_writeb writeb
++#define __raw_writew writew
++#define __raw_writel writel
+
-+ if (!dev->driver) {
-+ /* Information only: is this too noisy? */
-+ printk(KERN_INFO "XENBUS: Device with no driver: %s\n",
-+ xendev->nodename);
-+ } else if (xendev->state < XenbusStateConnected) {
-+ enum xenbus_state rstate = XenbusStateUnknown;
-+ if (xendev->otherend)
-+ rstate = xenbus_read_driver_state(xendev->otherend);
-+ printk(KERN_WARNING "XENBUS: Timeout connecting "
-+ "to device: %s (local state %d, remote state %d)\n",
-+ xendev->nodename, xendev->state, rstate);
-+ }
++#define mmiowb()
+
-+ return 0;
++static inline void memset_io(volatile void __iomem *addr, unsigned char val, int count)
++{
++ memset((void __force *) addr, val, count);
++}
++static inline void memcpy_fromio(void *dst, const volatile void __iomem *src, int count)
++{
++ __memcpy(dst, (void __force *) src, count);
++}
++static inline void memcpy_toio(volatile void __iomem *dst, const void *src, int count)
++{
++ __memcpy((void __force *) dst, src, count);
+}
-+
-+/* We only wait for device setup after most initcalls have run. */
-+static int ready_to_wait_for_devices;
+
+/*
-+ * On a 5-minute timeout, wait for all devices currently configured. We need
-+ * to do this to guarantee that the filesystems and / or network devices
-+ * needed for boot are available, before we can allow the boot to proceed.
-+ *
-+ * This needs to be on a late_initcall, to happen after the frontend device
-+ * drivers have been initialised, but before the root fs is mounted.
-+ *
-+ * A possible improvement here would be to have the tools add a per-device
-+ * flag to the store entry, indicating whether it is needed at boot time.
-+ * This would allow people who knew what they were doing to accelerate their
-+ * boot slightly, but of course needs tools or manual intervention to set up
-+ * those flags correctly.
++ * ISA space is 'always mapped' on a typical x86 system, no need to
++ * explicitly ioremap() it. The fact that the ISA IO space is mapped
++ * to PAGE_OFFSET is pure coincidence - it does not mean ISA values
++ * are physical addresses. The following constant pointer can be
++ * used as the IO-area pointer (it can be iounmapped as well, so the
++ * analogy with PCI is quite large):
+ */
-+static void wait_for_devices(struct xenbus_driver *xendrv)
-+{
-+ unsigned long start = jiffies;
-+ struct device_driver *drv = xendrv ? &xendrv->driver : NULL;
-+ unsigned int seconds_waited = 0;
-+
-+ if (!ready_to_wait_for_devices || !is_running_on_xen())
-+ return;
-+
-+ while (exists_disconnected_device(drv)) {
-+ if (time_after(jiffies, start + (seconds_waited+5)*HZ)) {
-+ if (!seconds_waited)
-+ printk(KERN_WARNING "XENBUS: Waiting for "
-+ "devices to initialise: ");
-+ seconds_waited += 5;
-+ printk("%us...", 300 - seconds_waited);
-+ if (seconds_waited == 300)
-+ break;
-+ }
-+
-+ schedule_timeout_interruptible(HZ/10);
-+ }
++#define __ISA_IO_base ((char __iomem *)(fix_to_virt(FIX_ISAMAP_BEGIN)))
+
-+ if (seconds_waited)
-+ printk("\n");
++/*
++ * Again, i386 does not require mem IO specific function.
++ */
+
-+ bus_for_each_dev(&xenbus_frontend.bus, NULL, drv,
-+ print_device_status);
-+}
++#define eth_io_copy_and_sum(a,b,c,d) eth_copy_and_sum((a),(void __force *)(b),(c),(d))
+
-+#ifndef MODULE
-+static int __init boot_wait_for_devices(void)
++/**
++ * check_signature - find BIOS signatures
++ * @io_addr: mmio address to check
++ * @signature: signature block
++ * @length: length of signature
++ *
++ * Perform a signature comparison with the mmio address io_addr. This
++ * address should have been obtained by ioremap.
++ * Returns 1 on a match.
++ */
++
++static inline int check_signature(volatile void __iomem * io_addr,
++ const unsigned char *signature, int length)
+{
-+ if (!xenbus_frontend.error) {
-+ ready_to_wait_for_devices = 1;
-+ wait_for_devices(NULL);
-+ }
-+ return 0;
++ int retval = 0;
++ do {
++ if (readb(io_addr) != *signature)
++ goto out;
++ io_addr++;
++ signature++;
++ length--;
++ } while (length);
++ retval = 1;
++out:
++ return retval;
+}
+
-+late_initcall(boot_wait_for_devices);
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_probe.h tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_probe.h
---- pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_probe.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_probe.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,75 @@
-+/******************************************************************************
-+ * xenbus_probe.h
-+ *
-+ * Talks to Xen Store to figure out what devices we have.
++/*
++ * Cache management
+ *
-+ * Copyright (C) 2005 Rusty Russell, IBM Corporation
-+ * Copyright (C) 2005 XenSource Ltd.
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
++ * This needed for two cases
++ * 1. Out of order aware processors
++ * 2. Accidentally out of order processors (PPro errata #51)
+ */
++
++#if defined(CONFIG_X86_OOSTORE) || defined(CONFIG_X86_PPRO_FENCE)
+
-+#ifndef _XENBUS_PROBE_H
-+#define _XENBUS_PROBE_H
-+
-+#if defined(CONFIG_XEN_BACKEND) || defined(CONFIG_XEN_BACKEND_MODULE)
-+extern void xenbus_backend_suspend(int (*fn)(struct device *, void *));
-+extern void xenbus_backend_resume(int (*fn)(struct device *, void *));
-+extern void xenbus_backend_probe_and_watch(void);
-+extern void xenbus_backend_bus_register(void);
-+extern void xenbus_backend_device_register(void);
-+#else
-+static inline void xenbus_backend_suspend(int (*fn)(struct device *, void *)) {}
-+static inline void xenbus_backend_resume(int (*fn)(struct device *, void *)) {}
-+static inline void xenbus_backend_probe_and_watch(void) {}
-+static inline void xenbus_backend_bus_register(void) {}
-+static inline void xenbus_backend_device_register(void) {}
-+#endif
-+
-+struct xen_bus_type
++static inline void flush_write_buffers(void)
+{
-+ char *root;
-+ int error;
-+ unsigned int levels;
-+ int (*get_bus_id)(char bus_id[BUS_ID_SIZE], const char *nodename);
-+ int (*probe)(const char *type, const char *dir);
-+ struct bus_type bus;
-+ struct device dev;
-+};
++ __asm__ __volatile__ ("lock; addl $0,0(%%esp)": : :"memory");
++}
++
++#define dma_cache_inv(_start,_size) flush_write_buffers()
++#define dma_cache_wback(_start,_size) flush_write_buffers()
++#define dma_cache_wback_inv(_start,_size) flush_write_buffers()
+
-+extern int xenbus_match(struct device *_dev, struct device_driver *_drv);
-+extern int xenbus_dev_probe(struct device *_dev);
-+extern int xenbus_dev_remove(struct device *_dev);
-+extern int xenbus_register_driver_common(struct xenbus_driver *drv,
-+ struct xen_bus_type *bus);
-+extern int xenbus_probe_node(struct xen_bus_type *bus,
-+ const char *type,
-+ const char *nodename);
-+extern int xenbus_probe_devices(struct xen_bus_type *bus);
++#else
+
-+extern void dev_changed(const char *node, struct xen_bus_type *bus);
++/* Nothing to do */
++
++#define dma_cache_inv(_start,_size) do { } while (0)
++#define dma_cache_wback(_start,_size) do { } while (0)
++#define dma_cache_wback_inv(_start,_size) do { } while (0)
++#define flush_write_buffers()
+
+#endif
+
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_xs.c tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_xs.c
---- pristine-linux-2.6.18.2/drivers/xen/xenbus/xenbus_xs.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/xenbus/xenbus_xs.c 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,886 @@
-+/******************************************************************************
-+ * xenbus_xs.c
-+ *
-+ * This is the kernel equivalent of the "xs" library. We don't need everything
-+ * and we use xenbus_comms for communication.
-+ *
-+ * Copyright (C) 2005 Rusty Russell, IBM Corporation
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++#endif /* __KERNEL__ */
+
-+#include <linux/unistd.h>
-+#include <linux/errno.h>
-+#include <linux/types.h>
-+#include <linux/uio.h>
-+#include <linux/kernel.h>
-+#include <linux/string.h>
-+#include <linux/err.h>
-+#include <linux/slab.h>
-+#include <linux/fcntl.h>
-+#include <linux/kthread.h>
-+#include <linux/rwsem.h>
-+#include <linux/module.h>
-+#include <linux/mutex.h>
-+#include <xen/xenbus.h>
-+#include "xenbus_comms.h"
++#ifdef SLOW_IO_BY_JUMPING
++#define __SLOW_DOWN_IO "jmp 1f; 1: jmp 1f; 1:"
++#else
++#define __SLOW_DOWN_IO "outb %%al,$0x80;"
++#endif
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
++static inline void slow_down_io(void) {
++ __asm__ __volatile__(
++ __SLOW_DOWN_IO
++#ifdef REALLY_SLOW_IO
++ __SLOW_DOWN_IO __SLOW_DOWN_IO __SLOW_DOWN_IO
+#endif
++ : : );
++}
+
-+#ifndef PF_NOFREEZE /* Old kernel (pre-2.6.6). */
-+#define PF_NOFREEZE 0
++#ifdef CONFIG_X86_NUMAQ
++extern void *xquad_portio; /* Where the IO area was mapped */
++#define XQUAD_PORT_ADDR(port, quad) (xquad_portio + (XQUAD_PORTIO_QUAD*quad) + port)
++#define __BUILDIO(bwl,bw,type) \
++static inline void out##bwl##_quad(unsigned type value, int port, int quad) { \
++ if (xquad_portio) \
++ write##bwl(value, XQUAD_PORT_ADDR(port, quad)); \
++ else \
++ out##bwl##_local(value, port); \
++} \
++static inline void out##bwl(unsigned type value, int port) { \
++ out##bwl##_quad(value, port, 0); \
++} \
++static inline unsigned type in##bwl##_quad(int port, int quad) { \
++ if (xquad_portio) \
++ return read##bwl(XQUAD_PORT_ADDR(port, quad)); \
++ else \
++ return in##bwl##_local(port); \
++} \
++static inline unsigned type in##bwl(int port) { \
++ return in##bwl##_quad(port, 0); \
++}
++#else
++#define __BUILDIO(bwl,bw,type) \
++static inline void out##bwl(unsigned type value, int port) { \
++ out##bwl##_local(value, port); \
++} \
++static inline unsigned type in##bwl(int port) { \
++ return in##bwl##_local(port); \
++}
+#endif
+
-+struct xs_stored_msg {
-+ struct list_head list;
+
-+ struct xsd_sockmsg hdr;
++#define BUILDIO(bwl,bw,type) \
++static inline void out##bwl##_local(unsigned type value, int port) { \
++ __asm__ __volatile__("out" #bwl " %" #bw "0, %w1" : : "a"(value), "Nd"(port)); \
++} \
++static inline unsigned type in##bwl##_local(int port) { \
++ unsigned type value; \
++ __asm__ __volatile__("in" #bwl " %w1, %" #bw "0" : "=a"(value) : "Nd"(port)); \
++ return value; \
++} \
++static inline void out##bwl##_local_p(unsigned type value, int port) { \
++ out##bwl##_local(value, port); \
++ slow_down_io(); \
++} \
++static inline unsigned type in##bwl##_local_p(int port) { \
++ unsigned type value = in##bwl##_local(port); \
++ slow_down_io(); \
++ return value; \
++} \
++__BUILDIO(bwl,bw,type) \
++static inline void out##bwl##_p(unsigned type value, int port) { \
++ out##bwl(value, port); \
++ slow_down_io(); \
++} \
++static inline unsigned type in##bwl##_p(int port) { \
++ unsigned type value = in##bwl(port); \
++ slow_down_io(); \
++ return value; \
++} \
++static inline void outs##bwl(int port, const void *addr, unsigned long count) { \
++ __asm__ __volatile__("rep; outs" #bwl : "+S"(addr), "+c"(count) : "d"(port)); \
++} \
++static inline void ins##bwl(int port, void *addr, unsigned long count) { \
++ __asm__ __volatile__("rep; ins" #bwl : "+D"(addr), "+c"(count) : "d"(port)); \
++}
+
-+ union {
-+ /* Queued replies. */
-+ struct {
-+ char *body;
-+ } reply;
++BUILDIO(b,b,char)
++BUILDIO(w,w,short)
++BUILDIO(l,,int)
+
-+ /* Queued watch events. */
-+ struct {
-+ struct xenbus_watch *handle;
-+ char **vec;
-+ unsigned int vec_size;
-+ } watch;
-+ } u;
-+};
++/* We will be supplying our own /dev/mem implementation */
++#define ARCH_HAS_DEV_MEM
+
-+struct xs_handle {
-+ /* A list of replies. Currently only one will ever be outstanding. */
-+ struct list_head reply_list;
-+ spinlock_t reply_lock;
-+ wait_queue_head_t reply_waitq;
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/irqflags.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/irqflags.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,127 @@
++/*
++ * include/asm-i386/irqflags.h
++ *
++ * IRQ flags handling
++ *
++ * This file gets included from lowlevel asm headers too, to provide
++ * wrapped versions of the local_irq_*() APIs, based on the
++ * raw_local_irq_*() functions from the lowlevel headers.
++ */
++#ifndef _ASM_IRQFLAGS_H
++#define _ASM_IRQFLAGS_H
+
-+ /*
-+ * Mutex ordering: transaction_mutex -> watch_mutex -> request_mutex.
-+ * response_mutex is never taken simultaneously with the other three.
-+ */
++#ifndef __ASSEMBLY__
+
-+ /* One request at a time. */
-+ struct mutex request_mutex;
++/*
++ * The use of 'barrier' in the following reflects their use as local-lock
++ * operations. Reentrancy must be prevented (e.g., __cli()) /before/ following
++ * critical operations are executed. All critical operations must complete
++ * /before/ reentrancy is permitted (e.g., __sti()). Alpha architecture also
++ * includes these barriers, for example.
++ */
+
-+ /* Protect xenbus reader thread against save/restore. */
-+ struct mutex response_mutex;
++#define __raw_local_save_flags() (current_vcpu_info()->evtchn_upcall_mask)
+
-+ /* Protect transactions against save/restore. */
-+ struct rw_semaphore transaction_mutex;
++#define raw_local_save_flags(flags) \
++ do { (flags) = __raw_local_save_flags(); } while (0)
+
-+ /* Protect watch (de)register against save/restore. */
-+ struct rw_semaphore watch_mutex;
-+};
++#define raw_local_irq_restore(x) \
++do { \
++ vcpu_info_t *_vcpu; \
++ barrier(); \
++ _vcpu = current_vcpu_info(); \
++ if ((_vcpu->evtchn_upcall_mask = (x)) == 0) { \
++ barrier(); /* unmask then check (avoid races) */ \
++ if (unlikely(_vcpu->evtchn_upcall_pending)) \
++ force_evtchn_callback(); \
++ } \
++} while (0)
+
-+static struct xs_handle xs_state;
++#define raw_local_irq_disable() \
++do { \
++ current_vcpu_info()->evtchn_upcall_mask = 1; \
++ barrier(); \
++} while (0)
+
-+/* List of registered watches, and a lock to protect it. */
-+static LIST_HEAD(watches);
-+static DEFINE_SPINLOCK(watches_lock);
++#define raw_local_irq_enable() \
++do { \
++ vcpu_info_t *_vcpu; \
++ barrier(); \
++ _vcpu = current_vcpu_info(); \
++ _vcpu->evtchn_upcall_mask = 0; \
++ barrier(); /* unmask then check (avoid races) */ \
++ if (unlikely(_vcpu->evtchn_upcall_pending)) \
++ force_evtchn_callback(); \
++} while (0)
+
-+/* List of pending watch callback events, and a lock to protect it. */
-+static LIST_HEAD(watch_events);
-+static DEFINE_SPINLOCK(watch_events_lock);
++/*
++ * Used in the idle loop; sti takes one instruction cycle
++ * to complete:
++ */
++void raw_safe_halt(void);
+
+/*
-+ * Details of the xenwatch callback kernel thread. The thread waits on the
-+ * watch_events_waitq for work to do (queued on watch_events list). When it
-+ * wakes up it acquires the xenwatch_mutex before reading the list and
-+ * carrying out work.
++ * Used when interrupts are already enabled or to
++ * shutdown the processor:
+ */
-+static pid_t xenwatch_pid;
-+/* static */ DEFINE_MUTEX(xenwatch_mutex);
-+static DECLARE_WAIT_QUEUE_HEAD(watch_events_waitq);
++void halt(void);
+
-+static int get_error(const char *errorstring)
++static inline int raw_irqs_disabled_flags(unsigned long flags)
+{
-+ unsigned int i;
-+
-+ for (i = 0; strcmp(errorstring, xsd_errors[i].errstring) != 0; i++) {
-+ if (i == ARRAY_SIZE(xsd_errors) - 1) {
-+ printk(KERN_WARNING
-+ "XENBUS xen store gave: unknown error %s",
-+ errorstring);
-+ return EINVAL;
-+ }
-+ }
-+ return xsd_errors[i].errnum;
++ return (flags != 0);
+}
+
-+static void *read_reply(enum xsd_sockmsg_type *type, unsigned int *len)
-+{
-+ struct xs_stored_msg *msg;
-+ char *body;
++#define raw_irqs_disabled() \
++({ \
++ unsigned long flags = __raw_local_save_flags(); \
++ \
++ raw_irqs_disabled_flags(flags); \
++})
+
-+ spin_lock(&xs_state.reply_lock);
++/*
++ * For spinlocks, etc:
++ */
++#define __raw_local_irq_save() \
++({ \
++ unsigned long flags = __raw_local_save_flags(); \
++ \
++ raw_local_irq_disable(); \
++ \
++ flags; \
++})
+
-+ while (list_empty(&xs_state.reply_list)) {
-+ spin_unlock(&xs_state.reply_lock);
-+ /* XXX FIXME: Avoid synchronous wait for response here. */
-+ wait_event(xs_state.reply_waitq,
-+ !list_empty(&xs_state.reply_list));
-+ spin_lock(&xs_state.reply_lock);
-+ }
++#define raw_local_irq_save(flags) \
++ do { (flags) = __raw_local_irq_save(); } while (0)
+
-+ msg = list_entry(xs_state.reply_list.next,
-+ struct xs_stored_msg, list);
-+ list_del(&msg->list);
++#endif /* __ASSEMBLY__ */
+
-+ spin_unlock(&xs_state.reply_lock);
++/*
++ * Do the CPU's IRQ-state tracing from assembly code. We call a
++ * C function, so save all the C-clobbered registers:
++ */
++#ifdef CONFIG_TRACE_IRQFLAGS
+
-+ *type = msg->hdr.type;
-+ if (len)
-+ *len = msg->hdr.len;
-+ body = msg->u.reply.body;
++# define TRACE_IRQS_ON \
++ pushl %eax; \
++ pushl %ecx; \
++ pushl %edx; \
++ call trace_hardirqs_on; \
++ popl %edx; \
++ popl %ecx; \
++ popl %eax;
+
-+ kfree(msg);
++# define TRACE_IRQS_OFF \
++ pushl %eax; \
++ pushl %ecx; \
++ pushl %edx; \
++ call trace_hardirqs_off; \
++ popl %edx; \
++ popl %ecx; \
++ popl %eax;
+
-+ return body;
-+}
++#else
++# define TRACE_IRQS_ON
++# define TRACE_IRQS_OFF
++#endif
+
-+void *xenbus_dev_request_and_reply(struct xsd_sockmsg *msg)
-+{
-+ void *ret;
-+ struct xsd_sockmsg req_msg = *msg;
-+ int err;
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/maddr.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/maddr.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,193 @@
++#ifndef _I386_MADDR_H
++#define _I386_MADDR_H
+
-+ if (req_msg.type == XS_TRANSACTION_START)
-+ down_read(&xs_state.transaction_mutex);
++#include <xen/features.h>
++#include <xen/interface/xen.h>
+
-+ mutex_lock(&xs_state.request_mutex);
++/**** MACHINE <-> PHYSICAL CONVERSION MACROS ****/
++#define INVALID_P2M_ENTRY (~0UL)
++#define FOREIGN_FRAME_BIT (1UL<<31)
++#define FOREIGN_FRAME(m) ((m) | FOREIGN_FRAME_BIT)
+
-+ err = xb_write(msg, sizeof(*msg) + msg->len);
-+ if (err) {
-+ msg->type = XS_ERROR;
-+ ret = ERR_PTR(err);
-+ } else
-+ ret = read_reply(&msg->type, &msg->len);
++/* Definitions for machine and pseudophysical addresses. */
++#ifdef CONFIG_X86_PAE
++typedef unsigned long long paddr_t;
++typedef unsigned long long maddr_t;
++#else
++typedef unsigned long paddr_t;
++typedef unsigned long maddr_t;
++#endif
+
-+ mutex_unlock(&xs_state.request_mutex);
++#ifdef CONFIG_XEN
+
-+ if ((req_msg.type == XS_TRANSACTION_END) ||
-+ ((req_msg.type == XS_TRANSACTION_START) &&
-+ (msg->type == XS_ERROR)))
-+ up_read(&xs_state.transaction_mutex);
++extern unsigned long *phys_to_machine_mapping;
++extern unsigned long max_mapnr;
+
-+ return ret;
-+}
++#undef machine_to_phys_mapping
++extern unsigned long *machine_to_phys_mapping;
++extern unsigned int machine_to_phys_order;
+
-+/* Send message to xs, get kmalloc'ed reply. ERR_PTR() on error. */
-+static void *xs_talkv(struct xenbus_transaction t,
-+ enum xsd_sockmsg_type type,
-+ const struct kvec *iovec,
-+ unsigned int num_vecs,
-+ unsigned int *len)
++static inline unsigned long pfn_to_mfn(unsigned long pfn)
+{
-+ struct xsd_sockmsg msg;
-+ void *ret = NULL;
-+ unsigned int i;
-+ int err;
-+
-+ msg.tx_id = t.id;
-+ msg.req_id = 0;
-+ msg.type = type;
-+ msg.len = 0;
-+ for (i = 0; i < num_vecs; i++)
-+ msg.len += iovec[i].iov_len;
-+
-+ mutex_lock(&xs_state.request_mutex);
-+
-+ err = xb_write(&msg, sizeof(msg));
-+ if (err) {
-+ mutex_unlock(&xs_state.request_mutex);
-+ return ERR_PTR(err);
-+ }
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return pfn;
++ BUG_ON(max_mapnr && pfn >= max_mapnr);
++ return phys_to_machine_mapping[pfn] & ~FOREIGN_FRAME_BIT;
++}
+
-+ for (i = 0; i < num_vecs; i++) {
-+ err = xb_write(iovec[i].iov_base, iovec[i].iov_len);;
-+ if (err) {
-+ mutex_unlock(&xs_state.request_mutex);
-+ return ERR_PTR(err);
-+ }
-+ }
++static inline int phys_to_machine_mapping_valid(unsigned long pfn)
++{
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return 1;
++ BUG_ON(max_mapnr && pfn >= max_mapnr);
++ return (phys_to_machine_mapping[pfn] != INVALID_P2M_ENTRY);
++}
+
-+ ret = read_reply(&msg.type, len);
++static inline unsigned long mfn_to_pfn(unsigned long mfn)
++{
++ unsigned long pfn;
+
-+ mutex_unlock(&xs_state.request_mutex);
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return mfn;
+
-+ if (IS_ERR(ret))
-+ return ret;
++ if (unlikely((mfn >> machine_to_phys_order) != 0))
++ return max_mapnr;
+
-+ if (msg.type == XS_ERROR) {
-+ err = get_error(ret);
-+ kfree(ret);
-+ return ERR_PTR(-err);
-+ }
++ /* The array access can fail (e.g., device space beyond end of RAM). */
++ asm (
++ "1: movl %1,%0\n"
++ "2:\n"
++ ".section .fixup,\"ax\"\n"
++ "3: movl %2,%0\n"
++ " jmp 2b\n"
++ ".previous\n"
++ ".section __ex_table,\"a\"\n"
++ " .align 4\n"
++ " .long 1b,3b\n"
++ ".previous"
++ : "=r" (pfn)
++ : "m" (machine_to_phys_mapping[mfn]), "m" (max_mapnr) );
+
-+ if (msg.type != type) {
-+ if (printk_ratelimit())
-+ printk(KERN_WARNING
-+ "XENBUS unexpected type [%d], expected [%d]\n",
-+ msg.type, type);
-+ kfree(ret);
-+ return ERR_PTR(-EINVAL);
-+ }
-+ return ret;
++ return pfn;
+}
+
-+/* Simplified version of xs_talkv: single message. */
-+static void *xs_single(struct xenbus_transaction t,
-+ enum xsd_sockmsg_type type,
-+ const char *string,
-+ unsigned int *len)
++/*
++ * We detect special mappings in one of two ways:
++ * 1. If the MFN is an I/O page then Xen will set the m2p entry
++ * to be outside our maximum possible pseudophys range.
++ * 2. If the MFN belongs to a different domain then we will certainly
++ * not have MFN in our p2m table. Conversely, if the page is ours,
++ * then we'll have p2m(m2p(MFN))==MFN.
++ * If we detect a special mapping then it doesn't have a 'struct page'.
++ * We force !pfn_valid() by returning an out-of-range pointer.
++ *
++ * NB. These checks require that, for any MFN that is not in our reservation,
++ * there is no PFN such that p2m(PFN) == MFN. Otherwise we can get confused if
++ * we are foreign-mapping the MFN, and the other domain as m2p(MFN) == PFN.
++ * Yikes! Various places must poke in INVALID_P2M_ENTRY for safety.
++ *
++ * NB2. When deliberately mapping foreign pages into the p2m table, you *must*
++ * use FOREIGN_FRAME(). This will cause pte_pfn() to choke on it, as we
++ * require. In all the cases we care about, the FOREIGN_FRAME bit is
++ * masked (e.g., pfn_to_mfn()) so behaviour there is correct.
++ */
++static inline unsigned long mfn_to_local_pfn(unsigned long mfn)
+{
-+ struct kvec iovec;
++ unsigned long pfn = mfn_to_pfn(mfn);
++ if ((pfn < max_mapnr)
++ && !xen_feature(XENFEAT_auto_translated_physmap)
++ && (phys_to_machine_mapping[pfn] != mfn))
++ return max_mapnr; /* force !pfn_valid() */
++ return pfn;
++}
+
-+ iovec.iov_base = (void *)string;
-+ iovec.iov_len = strlen(string) + 1;
-+ return xs_talkv(t, type, &iovec, 1, len);
++static inline void set_phys_to_machine(unsigned long pfn, unsigned long mfn)
++{
++ BUG_ON(max_mapnr && pfn >= max_mapnr);
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY);
++ return;
++ }
++ phys_to_machine_mapping[pfn] = mfn;
+}
+
-+/* Many commands only need an ack, don't care what it says. */
-+static int xs_error(char *reply)
++static inline maddr_t phys_to_machine(paddr_t phys)
+{
-+ if (IS_ERR(reply))
-+ return PTR_ERR(reply);
-+ kfree(reply);
-+ return 0;
++ maddr_t machine = pfn_to_mfn(phys >> PAGE_SHIFT);
++ machine = (machine << PAGE_SHIFT) | (phys & ~PAGE_MASK);
++ return machine;
+}
+
-+static unsigned int count_strings(const char *strings, unsigned int len)
++static inline paddr_t machine_to_phys(maddr_t machine)
+{
-+ unsigned int num;
-+ const char *p;
-+
-+ for (p = strings, num = 0; p < strings + len; p += strlen(p) + 1)
-+ num++;
-+
-+ return num;
++ paddr_t phys = mfn_to_pfn(machine >> PAGE_SHIFT);
++ phys = (phys << PAGE_SHIFT) | (machine & ~PAGE_MASK);
++ return phys;
+}
+
-+/* Return the path to dir with /name appended. Buffer must be kfree()'ed. */
-+static char *join(const char *dir, const char *name)
++#ifdef CONFIG_X86_PAE
++static inline paddr_t pte_phys_to_machine(paddr_t phys)
+{
-+ char *buffer;
-+
-+ if (strlen(name) == 0)
-+ buffer = kasprintf(GFP_KERNEL, "%s", dir);
-+ else
-+ buffer = kasprintf(GFP_KERNEL, "%s/%s", dir, name);
-+ return (!buffer) ? ERR_PTR(-ENOMEM) : buffer;
++ /*
++ * In PAE mode, the NX bit needs to be dealt with in the value
++ * passed to pfn_to_mfn(). On x86_64, we need to mask it off,
++ * but for i386 the conversion to ulong for the argument will
++ * clip it off.
++ */
++ maddr_t machine = pfn_to_mfn(phys >> PAGE_SHIFT);
++ machine = (machine << PAGE_SHIFT) | (phys & ~PHYSICAL_PAGE_MASK);
++ return machine;
+}
+
-+static char **split(char *strings, unsigned int len, unsigned int *num)
++static inline paddr_t pte_machine_to_phys(maddr_t machine)
+{
-+ char *p, **ret;
-+
-+ /* Count the strings. */
-+ *num = count_strings(strings, len);
-+
-+ /* Transfer to one big alloc for easy freeing. */
-+ ret = kmalloc(*num * sizeof(char *) + len, GFP_KERNEL);
-+ if (!ret) {
-+ kfree(strings);
-+ return ERR_PTR(-ENOMEM);
-+ }
-+ memcpy(&ret[*num], strings, len);
-+ kfree(strings);
-+
-+ strings = (char *)&ret[*num];
-+ for (p = strings, *num = 0; p < strings + len; p += strlen(p) + 1)
-+ ret[(*num)++] = p;
-+
-+ return ret;
++ /*
++ * In PAE mode, the NX bit needs to be dealt with in the value
++ * passed to mfn_to_pfn(). On x86_64, we need to mask it off,
++ * but for i386 the conversion to ulong for the argument will
++ * clip it off.
++ */
++ paddr_t phys = mfn_to_pfn(machine >> PAGE_SHIFT);
++ phys = (phys << PAGE_SHIFT) | (machine & ~PHYSICAL_PAGE_MASK);
++ return phys;
+}
++#endif
+
-+char **xenbus_directory(struct xenbus_transaction t,
-+ const char *dir, const char *node, unsigned int *num)
++#ifdef CONFIG_X86_PAE
++#define __pte_ma(x) ((pte_t) { (x), (maddr_t)(x) >> 32 } )
++static inline pte_t pfn_pte_ma(unsigned long page_nr, pgprot_t pgprot)
+{
-+ char *strings, *path;
-+ unsigned int len;
++ pte_t pte;
+
-+ path = join(dir, node);
-+ if (IS_ERR(path))
-+ return (char **)path;
++ pte.pte_high = (page_nr >> (32 - PAGE_SHIFT)) | \
++ (pgprot_val(pgprot) >> 32);
++ pte.pte_high &= (__supported_pte_mask >> 32);
++ pte.pte_low = ((page_nr << PAGE_SHIFT) | pgprot_val(pgprot)) & \
++ __supported_pte_mask;
++ return pte;
++}
++#else
++#define __pte_ma(x) ((pte_t) { (x) } )
++#define pfn_pte_ma(pfn, prot) __pte_ma(((pfn) << PAGE_SHIFT) | pgprot_val(prot))
++#endif
+
-+ strings = xs_single(t, XS_DIRECTORY, path, &len);
-+ kfree(path);
-+ if (IS_ERR(strings))
-+ return (char **)strings;
++#else /* !CONFIG_XEN */
+
-+ return split(strings, len, num);
-+}
-+EXPORT_SYMBOL_GPL(xenbus_directory);
++#define pfn_to_mfn(pfn) (pfn)
++#define mfn_to_pfn(mfn) (mfn)
++#define mfn_to_local_pfn(mfn) (mfn)
++#define set_phys_to_machine(pfn, mfn) ((void)0)
++#define phys_to_machine_mapping_valid(pfn) (1)
++#define phys_to_machine(phys) ((maddr_t)(phys))
++#define machine_to_phys(mach) ((paddr_t)(mach))
++#define pfn_pte_ma(pfn, prot) pfn_pte(pfn, prot)
++#define __pte_ma(x) __pte(x)
+
-+/* Check if a path exists. Return 1 if it does. */
-+int xenbus_exists(struct xenbus_transaction t,
-+ const char *dir, const char *node)
-+{
-+ char **d;
-+ int dir_n;
++#endif /* !CONFIG_XEN */
+
-+ d = xenbus_directory(t, dir, node, &dir_n);
-+ if (IS_ERR(d))
-+ return 0;
-+ kfree(d);
-+ return 1;
-+}
-+EXPORT_SYMBOL_GPL(xenbus_exists);
++/* VIRT <-> MACHINE conversion */
++#define virt_to_machine(v) (phys_to_machine(__pa(v)))
++#define virt_to_mfn(v) (pfn_to_mfn(__pa(v) >> PAGE_SHIFT))
++#define mfn_to_virt(m) (__va(mfn_to_pfn(m) << PAGE_SHIFT))
+
-+/* Get the value of a single file.
-+ * Returns a kmalloced value: call free() on it after use.
-+ * len indicates length in bytes.
++#endif /* _I386_MADDR_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/mmu.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/mmu.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,29 @@
++#ifndef __i386_MMU_H
++#define __i386_MMU_H
++
++#include <asm/semaphore.h>
++/*
++ * The i386 doesn't have a mmu context, but
++ * we put the segment information here.
++ *
++ * cpu_vm_mask is used to optimize ldt flushing.
+ */
-+void *xenbus_read(struct xenbus_transaction t,
-+ const char *dir, const char *node, unsigned int *len)
-+{
-+ char *path;
-+ void *ret;
++typedef struct {
++ int size;
++ struct semaphore sem;
++ void *ldt;
++ void *vdso;
++#ifdef CONFIG_XEN
++ int has_foreign_mappings;
++#endif
++} mm_context_t;
+
-+ path = join(dir, node);
-+ if (IS_ERR(path))
-+ return (void *)path;
++/* mm/memory.c:exit_mmap hook */
++extern void _arch_exit_mmap(struct mm_struct *mm);
++#define arch_exit_mmap(_mm) _arch_exit_mmap(_mm)
+
-+ ret = xs_single(t, XS_READ, path, len);
-+ kfree(path);
-+ return ret;
-+}
-+EXPORT_SYMBOL_GPL(xenbus_read);
++/* kernel/fork.c:dup_mmap hook */
++extern void _arch_dup_mmap(struct mm_struct *mm);
++#define arch_dup_mmap(mm, oldmm) ((void)(oldmm), _arch_dup_mmap(mm))
+
-+/* Write the value of a single file.
-+ * Returns -err on failure.
-+ */
-+int xenbus_write(struct xenbus_transaction t,
-+ const char *dir, const char *node, const char *string)
-+{
-+ const char *path;
-+ struct kvec iovec[2];
-+ int ret;
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/mmu_context.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/mmu_context.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,108 @@
++#ifndef __I386_SCHED_H
++#define __I386_SCHED_H
+
-+ path = join(dir, node);
-+ if (IS_ERR(path))
-+ return PTR_ERR(path);
++#include <asm/desc.h>
++#include <asm/atomic.h>
++#include <asm/pgalloc.h>
++#include <asm/tlbflush.h>
+
-+ iovec[0].iov_base = (void *)path;
-+ iovec[0].iov_len = strlen(path) + 1;
-+ iovec[1].iov_base = (void *)string;
-+ iovec[1].iov_len = strlen(string);
++/*
++ * Used for LDT copy/destruction.
++ */
++int init_new_context(struct task_struct *tsk, struct mm_struct *mm);
++void destroy_context(struct mm_struct *mm);
+
-+ ret = xs_error(xs_talkv(t, XS_WRITE, iovec, ARRAY_SIZE(iovec), NULL));
-+ kfree(path);
-+ return ret;
-+}
-+EXPORT_SYMBOL_GPL(xenbus_write);
+
-+/* Create a new directory. */
-+int xenbus_mkdir(struct xenbus_transaction t,
-+ const char *dir, const char *node)
++static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk)
+{
-+ char *path;
-+ int ret;
++#if 0 /* XEN: no lazy tlb */
++ unsigned cpu = smp_processor_id();
++ if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_OK)
++ per_cpu(cpu_tlbstate, cpu).state = TLBSTATE_LAZY;
++#endif
++}
+
-+ path = join(dir, node);
-+ if (IS_ERR(path))
-+ return PTR_ERR(path);
++#define prepare_arch_switch(next) __prepare_arch_switch()
+
-+ ret = xs_error(xs_single(t, XS_MKDIR, path, NULL));
-+ kfree(path);
-+ return ret;
++static inline void __prepare_arch_switch(void)
++{
++ /*
++ * Save away %fs and %gs. No need to save %es and %ds, as those
++ * are always kernel segments while inside the kernel. Must
++ * happen before reload of cr3/ldt (i.e., not in __switch_to).
++ */
++ asm volatile ( "mov %%fs,%0 ; mov %%gs,%1"
++ : "=m" (current->thread.fs),
++ "=m" (current->thread.gs));
++ asm volatile ( "movl %0,%%fs ; movl %0,%%gs"
++ : : "r" (0) );
+}
-+EXPORT_SYMBOL_GPL(xenbus_mkdir);
+
-+/* Destroy a file or directory (directories must be empty). */
-+int xenbus_rm(struct xenbus_transaction t, const char *dir, const char *node)
++extern void mm_pin(struct mm_struct *mm);
++extern void mm_unpin(struct mm_struct *mm);
++void mm_pin_all(void);
++
++static inline void switch_mm(struct mm_struct *prev,
++ struct mm_struct *next,
++ struct task_struct *tsk)
+{
-+ char *path;
-+ int ret;
++ int cpu = smp_processor_id();
++ struct mmuext_op _op[2], *op = _op;
+
-+ path = join(dir, node);
-+ if (IS_ERR(path))
-+ return PTR_ERR(path);
++ if (likely(prev != next)) {
++ BUG_ON(!xen_feature(XENFEAT_writable_page_tables) &&
++ !test_bit(PG_pinned, &virt_to_page(next->pgd)->flags));
+
-+ ret = xs_error(xs_single(t, XS_RM, path, NULL));
-+ kfree(path);
-+ return ret;
-+}
-+EXPORT_SYMBOL_GPL(xenbus_rm);
++ /* stop flush ipis for the previous mm */
++ cpu_clear(cpu, prev->cpu_vm_mask);
++#if 0 /* XEN: no lazy tlb */
++ per_cpu(cpu_tlbstate, cpu).state = TLBSTATE_OK;
++ per_cpu(cpu_tlbstate, cpu).active_mm = next;
++#endif
++ cpu_set(cpu, next->cpu_vm_mask);
+
-+/* Start a transaction: changes by others will not be seen during this
-+ * transaction, and changes will not be visible to others until end.
-+ */
-+int xenbus_transaction_start(struct xenbus_transaction *t)
-+{
-+ char *id_str;
++ /* Re-load page tables: load_cr3(next->pgd) */
++ op->cmd = MMUEXT_NEW_BASEPTR;
++ op->arg1.mfn = pfn_to_mfn(__pa(next->pgd) >> PAGE_SHIFT);
++ op++;
+
-+ down_read(&xs_state.transaction_mutex);
++ /*
++ * load the LDT, if the LDT is different:
++ */
++ if (unlikely(prev->context.ldt != next->context.ldt)) {
++ /* load_LDT_nolock(&next->context, cpu) */
++ op->cmd = MMUEXT_SET_LDT;
++ op->arg1.linear_addr = (unsigned long)next->context.ldt;
++ op->arg2.nr_ents = next->context.size;
++ op++;
++ }
+
-+ id_str = xs_single(XBT_NIL, XS_TRANSACTION_START, "", NULL);
-+ if (IS_ERR(id_str)) {
-+ up_read(&xs_state.transaction_mutex);
-+ return PTR_ERR(id_str);
++ BUG_ON(HYPERVISOR_mmuext_op(_op, op-_op, NULL, DOMID_SELF));
+ }
++#if 0 /* XEN: no lazy tlb */
++ else {
++ per_cpu(cpu_tlbstate, cpu).state = TLBSTATE_OK;
++ BUG_ON(per_cpu(cpu_tlbstate, cpu).active_mm != next);
+
-+ t->id = simple_strtoul(id_str, NULL, 0);
-+ kfree(id_str);
-+ return 0;
++ if (!cpu_test_and_set(cpu, next->cpu_vm_mask)) {
++ /* We were in lazy tlb mode and leave_mm disabled
++ * tlb flush IPI delivery. We must reload %cr3.
++ */
++ load_cr3(next->pgd);
++ load_LDT_nolock(&next->context, cpu);
++ }
++ }
++#endif
+}
-+EXPORT_SYMBOL_GPL(xenbus_transaction_start);
+
-+/* End a transaction.
-+ * If abandon is true, transaction is discarded instead of committed.
-+ */
-+int xenbus_transaction_end(struct xenbus_transaction t, int abort)
++#define deactivate_mm(tsk, mm) \
++ asm("movl %0,%%fs ; movl %0,%%gs": :"r" (0))
++
++static inline void activate_mm(struct mm_struct *prev, struct mm_struct *next)
+{
-+ char abortstr[2];
-+ int err;
++ if (!test_bit(PG_pinned, &virt_to_page(next->pgd)->flags))
++ mm_pin(next);
++ switch_mm(prev, next, NULL);
++}
+
-+ if (abort)
-+ strcpy(abortstr, "F");
-+ else
-+ strcpy(abortstr, "T");
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/page.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/page.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,233 @@
++#ifndef _I386_PAGE_H
++#define _I386_PAGE_H
+
-+ err = xs_error(xs_single(t, XS_TRANSACTION_END, abortstr, NULL));
++/* PAGE_SHIFT determines the page size */
++#define PAGE_SHIFT 12
++#define PAGE_SIZE (1UL << PAGE_SHIFT)
++#define PAGE_MASK (~(PAGE_SIZE-1))
+
-+ up_read(&xs_state.transaction_mutex);
++#ifdef CONFIG_X86_PAE
++#define __PHYSICAL_MASK_SHIFT 40
++#define __PHYSICAL_MASK ((1ULL << __PHYSICAL_MASK_SHIFT) - 1)
++#define PHYSICAL_PAGE_MASK (~((1ULL << PAGE_SHIFT) - 1) & __PHYSICAL_MASK)
++#else
++#define __PHYSICAL_MASK_SHIFT 32
++#define __PHYSICAL_MASK (~0UL)
++#define PHYSICAL_PAGE_MASK (PAGE_MASK & __PHYSICAL_MASK)
++#endif
+
-+ return err;
-+}
-+EXPORT_SYMBOL_GPL(xenbus_transaction_end);
++#define LARGE_PAGE_MASK (~(LARGE_PAGE_SIZE-1))
++#define LARGE_PAGE_SIZE (1UL << PMD_SHIFT)
+
-+/* Single read and scanf: returns -errno or num scanned. */
-+int xenbus_scanf(struct xenbus_transaction t,
-+ const char *dir, const char *node, const char *fmt, ...)
-+{
-+ va_list ap;
-+ int ret;
-+ char *val;
++#ifdef __KERNEL__
+
-+ val = xenbus_read(t, dir, node, NULL);
-+ if (IS_ERR(val))
-+ return PTR_ERR(val);
++/*
++ * Need to repeat this here in order to not include pgtable.h (which in turn
++ * depends on definitions made here), but to be able to use the symbolic
++ * below. The preprocessor will warn if the two definitions aren't identical.
++ */
++#define _PAGE_PRESENT 0x001
++#define _PAGE_IO 0x200
+
-+ va_start(ap, fmt);
-+ ret = vsscanf(val, fmt, ap);
-+ va_end(ap);
-+ kfree(val);
-+ /* Distinctive errno. */
-+ if (ret == 0)
-+ return -ERANGE;
-+ return ret;
-+}
-+EXPORT_SYMBOL_GPL(xenbus_scanf);
++#ifndef __ASSEMBLY__
+
-+/* Single printf and write: returns -errno or 0. */
-+int xenbus_printf(struct xenbus_transaction t,
-+ const char *dir, const char *node, const char *fmt, ...)
-+{
-+ va_list ap;
-+ int ret;
-+#define PRINTF_BUFFER_SIZE 4096
-+ char *printf_buffer;
++#include <linux/string.h>
++#include <linux/types.h>
++#include <linux/kernel.h>
++#include <asm/bug.h>
++#include <xen/interface/xen.h>
++#include <xen/features.h>
+
-+ printf_buffer = kmalloc(PRINTF_BUFFER_SIZE, GFP_KERNEL);
-+ if (printf_buffer == NULL)
-+ return -ENOMEM;
++#ifdef CONFIG_X86_USE_3DNOW
+
-+ va_start(ap, fmt);
-+ ret = vsnprintf(printf_buffer, PRINTF_BUFFER_SIZE, fmt, ap);
-+ va_end(ap);
++#include <asm/mmx.h>
+
-+ BUG_ON(ret > PRINTF_BUFFER_SIZE-1);
-+ ret = xenbus_write(t, dir, node, printf_buffer);
++#define clear_page(page) mmx_clear_page((void *)(page))
++#define copy_page(to,from) mmx_copy_page(to,from)
+
-+ kfree(printf_buffer);
++#else
+
-+ return ret;
-+}
-+EXPORT_SYMBOL_GPL(xenbus_printf);
++#define alloc_zeroed_user_highpage(vma, vaddr) alloc_page_vma(GFP_HIGHUSER | __GFP_ZERO, vma, vaddr)
++#define __HAVE_ARCH_ALLOC_ZEROED_USER_HIGHPAGE
+
-+/* Takes tuples of names, scanf-style args, and void **, NULL terminated. */
-+int xenbus_gather(struct xenbus_transaction t, const char *dir, ...)
-+{
-+ va_list ap;
-+ const char *name;
-+ int ret = 0;
++/*
++ * On older X86 processors it's not a win to use MMX here it seems.
++ * Maybe the K6-III ?
++ */
++
++#define clear_page(page) memset((void *)(page), 0, PAGE_SIZE)
++#define copy_page(to,from) memcpy((void *)(to), (void *)(from), PAGE_SIZE)
+
-+ va_start(ap, dir);
-+ while (ret == 0 && (name = va_arg(ap, char *)) != NULL) {
-+ const char *fmt = va_arg(ap, char *);
-+ void *result = va_arg(ap, void *);
-+ char *p;
++#endif
+
-+ p = xenbus_read(t, dir, name, NULL);
-+ if (IS_ERR(p)) {
-+ ret = PTR_ERR(p);
-+ break;
-+ }
-+ if (fmt) {
-+ if (sscanf(p, fmt, result) == 0)
-+ ret = -EINVAL;
-+ kfree(p);
-+ } else
-+ *(char **)result = p;
-+ }
-+ va_end(ap);
++#define clear_user_page(page, vaddr, pg) clear_page(page)
++#define copy_user_page(to, from, vaddr, pg) copy_page(to, from)
++
++/*
++ * These are used to make use of C type-checking..
++ */
++extern int nx_enabled;
++#ifdef CONFIG_X86_PAE
++extern unsigned long long __supported_pte_mask;
++typedef struct { unsigned long pte_low, pte_high; } pte_t;
++typedef struct { unsigned long long pmd; } pmd_t;
++typedef struct { unsigned long long pgd; } pgd_t;
++typedef struct { unsigned long long pgprot; } pgprot_t;
++#define pgprot_val(x) ((x).pgprot)
++#include <asm/maddr.h>
++#define __pte(x) ({ unsigned long long _x = (x); \
++ if ((_x & (_PAGE_PRESENT|_PAGE_IO)) == _PAGE_PRESENT) \
++ _x = pte_phys_to_machine(_x); \
++ ((pte_t) {(unsigned long)(_x), (unsigned long)(_x>>32)}); })
++#define __pgd(x) ({ unsigned long long _x = (x); \
++ (pgd_t) {((_x) & _PAGE_PRESENT) ? pte_phys_to_machine(_x) : (_x)}; })
++#define __pmd(x) ({ unsigned long long _x = (x); \
++ (pmd_t) {((_x) & _PAGE_PRESENT) ? pte_phys_to_machine(_x) : (_x)}; })
++static inline unsigned long long __pte_val(pte_t x)
++{
++ return ((unsigned long long)x.pte_high << 32) | x.pte_low;
++}
++static inline unsigned long long pte_val(pte_t x)
++{
++ unsigned long long ret = __pte_val(x);
++ if ((x.pte_low & (_PAGE_PRESENT|_PAGE_IO)) == _PAGE_PRESENT)
++ ret = pte_machine_to_phys(ret);
++ return ret;
++}
++#define __pmd_val(x) ((x).pmd)
++static inline unsigned long long pmd_val(pmd_t x)
++{
++ unsigned long long ret = __pmd_val(x);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (ret) ret = pte_machine_to_phys(ret) | _PAGE_PRESENT;
++#else
++ if (ret & _PAGE_PRESENT) ret = pte_machine_to_phys(ret);
++#endif
++ return ret;
++}
++#define __pud_val(x) __pgd_val((x).pgd)
++#define __pgd_val(x) ((x).pgd)
++static inline unsigned long long pgd_val(pgd_t x)
++{
++ unsigned long long ret = __pgd_val(x);
++ if (ret & _PAGE_PRESENT) ret = pte_machine_to_phys(ret);
++ return ret;
++}
++#define HPAGE_SHIFT 21
++#else
++typedef struct { unsigned long pte_low; } pte_t;
++typedef struct { unsigned long pgd; } pgd_t;
++typedef struct { unsigned long pgprot; } pgprot_t;
++#define pgprot_val(x) ((x).pgprot)
++#include <asm/maddr.h>
++#define boot_pte_t pte_t /* or would you rather have a typedef */
++#define __pte_val(x) ((x).pte_low)
++#define pte_val(x) ((__pte_val(x) & (_PAGE_PRESENT|_PAGE_IO)) \
++ == _PAGE_PRESENT ? \
++ machine_to_phys(__pte_val(x)) : \
++ __pte_val(x))
++#define __pte(x) ({ unsigned long _x = (x); \
++ if ((_x & (_PAGE_PRESENT|_PAGE_IO)) == _PAGE_PRESENT) \
++ _x = phys_to_machine(_x); \
++ ((pte_t) { _x }); })
++#define __pmd_val(x) __pud_val((x).pud)
++#define __pud_val(x) __pgd_val((x).pgd)
++#define __pgd(x) ({ unsigned long _x = (x); \
++ (pgd_t) {((_x) & _PAGE_PRESENT) ? phys_to_machine(_x) : (_x)}; })
++#define __pgd_val(x) ((x).pgd)
++static inline unsigned long pgd_val(pgd_t x)
++{
++ unsigned long ret = __pgd_val(x);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (ret) ret = machine_to_phys(ret) | _PAGE_PRESENT;
++#else
++ if (ret & _PAGE_PRESENT) ret = machine_to_phys(ret);
++#endif
+ return ret;
+}
-+EXPORT_SYMBOL_GPL(xenbus_gather);
-+
-+static int xs_watch(const char *path, const char *token)
-+{
-+ struct kvec iov[2];
-+
-+ iov[0].iov_base = (void *)path;
-+ iov[0].iov_len = strlen(path) + 1;
-+ iov[1].iov_base = (void *)token;
-+ iov[1].iov_len = strlen(token) + 1;
-+
-+ return xs_error(xs_talkv(XBT_NIL, XS_WATCH, iov,
-+ ARRAY_SIZE(iov), NULL));
-+}
-+
-+static int xs_unwatch(const char *path, const char *token)
-+{
-+ struct kvec iov[2];
++#define HPAGE_SHIFT 22
++#endif
++#define PTE_MASK PHYSICAL_PAGE_MASK
+
-+ iov[0].iov_base = (char *)path;
-+ iov[0].iov_len = strlen(path) + 1;
-+ iov[1].iov_base = (char *)token;
-+ iov[1].iov_len = strlen(token) + 1;
++#ifdef CONFIG_HUGETLB_PAGE
++#define HPAGE_SIZE ((1UL) << HPAGE_SHIFT)
++#define HPAGE_MASK (~(HPAGE_SIZE - 1))
++#define HUGETLB_PAGE_ORDER (HPAGE_SHIFT - PAGE_SHIFT)
++#define HAVE_ARCH_HUGETLB_UNMAPPED_AREA
++#endif
+
-+ return xs_error(xs_talkv(XBT_NIL, XS_UNWATCH, iov,
-+ ARRAY_SIZE(iov), NULL));
-+}
++#define __pgprot(x) ((pgprot_t) { (x) } )
+
-+static struct xenbus_watch *find_watch(const char *token)
-+{
-+ struct xenbus_watch *i, *cmp;
++#endif /* !__ASSEMBLY__ */
+
-+ cmp = (void *)simple_strtoul(token, NULL, 16);
++/* to align the pointer to the (next) page boundary */
++#define PAGE_ALIGN(addr) (((addr)+PAGE_SIZE-1)&PAGE_MASK)
+
-+ list_for_each_entry(i, &watches, list)
-+ if (i == cmp)
-+ return i;
++/*
++ * This handles the memory map.. We could make this a config
++ * option, but too many people screw it up, and too few need
++ * it.
++ *
++ * A __PAGE_OFFSET of 0xC0000000 means that the kernel has
++ * a virtual address space of one gigabyte, which limits the
++ * amount of physical memory you can use to about 950MB.
++ *
++ * If you want more physical memory than this then see the CONFIG_HIGHMEM4G
++ * and CONFIG_HIGHMEM64G options in the kernel configuration.
++ */
+
-+ return NULL;
-+}
++#ifndef __ASSEMBLY__
+
-+/* Register callback to watch this node. */
-+int register_xenbus_watch(struct xenbus_watch *watch)
-+{
-+ /* Pointer in ascii is the token. */
-+ char token[sizeof(watch) * 2 + 1];
-+ int err;
++struct vm_area_struct;
+
-+ sprintf(token, "%lX", (long)watch);
++/*
++ * This much address space is reserved for vmalloc() and iomap()
++ * as well as fixmap mappings.
++ */
++extern unsigned int __VMALLOC_RESERVE;
+
-+ down_read(&xs_state.watch_mutex);
++extern int sysctl_legacy_va_layout;
+
-+ spin_lock(&watches_lock);
-+ BUG_ON(find_watch(token));
-+ list_add(&watch->list, &watches);
-+ spin_unlock(&watches_lock);
++extern int page_is_ram(unsigned long pagenr);
+
-+ err = xs_watch(watch->node, token);
++#endif /* __ASSEMBLY__ */
+
-+ /* Ignore errors due to multiple registration. */
-+ if ((err != 0) && (err != -EEXIST)) {
-+ spin_lock(&watches_lock);
-+ list_del(&watch->list);
-+ spin_unlock(&watches_lock);
-+ }
++#ifdef __ASSEMBLY__
++#define __PAGE_OFFSET CONFIG_PAGE_OFFSET
++#define __PHYSICAL_START CONFIG_PHYSICAL_START
++#else
++#define __PAGE_OFFSET ((unsigned long)CONFIG_PAGE_OFFSET)
++#define __PHYSICAL_START ((unsigned long)CONFIG_PHYSICAL_START)
++#endif
++#define __KERNEL_START (__PAGE_OFFSET + __PHYSICAL_START)
+
-+ up_read(&xs_state.watch_mutex);
++#if CONFIG_XEN_COMPAT <= 0x030002
++#undef LOAD_OFFSET
++#define LOAD_OFFSET 0
++#endif
+
-+ return err;
-+}
-+EXPORT_SYMBOL_GPL(register_xenbus_watch);
++#define PAGE_OFFSET ((unsigned long)__PAGE_OFFSET)
++#define VMALLOC_RESERVE ((unsigned long)__VMALLOC_RESERVE)
++#define MAXMEM (__FIXADDR_TOP-__PAGE_OFFSET-__VMALLOC_RESERVE)
++#define __pa(x) ((unsigned long)(x)-PAGE_OFFSET)
++#define __va(x) ((void *)((unsigned long)(x)+PAGE_OFFSET))
++#define pfn_to_kaddr(pfn) __va((pfn) << PAGE_SHIFT)
++#ifdef CONFIG_FLATMEM
++#define pfn_valid(pfn) ((pfn) < max_mapnr)
++#endif /* CONFIG_FLATMEM */
++#define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT)
+
-+void unregister_xenbus_watch(struct xenbus_watch *watch)
-+{
-+ struct xs_stored_msg *msg, *tmp;
-+ char token[sizeof(watch) * 2 + 1];
-+ int err;
++#define virt_addr_valid(kaddr) pfn_valid(__pa(kaddr) >> PAGE_SHIFT)
+
-+ sprintf(token, "%lX", (long)watch);
++#define VM_DATA_DEFAULT_FLAGS \
++ (VM_READ | VM_WRITE | \
++ ((current->personality & READ_IMPLIES_EXEC) ? VM_EXEC : 0 ) | \
++ VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC)
+
-+ down_read(&xs_state.watch_mutex);
++#include <asm-generic/memory_model.h>
++#include <asm-generic/page.h>
+
-+ spin_lock(&watches_lock);
-+ BUG_ON(!find_watch(token));
-+ list_del(&watch->list);
-+ spin_unlock(&watches_lock);
++#define __HAVE_ARCH_GATE_AREA 1
++#endif /* __KERNEL__ */
+
-+ err = xs_unwatch(watch->node, token);
-+ if (err)
-+ printk(KERN_WARNING
-+ "XENBUS Failed to release watch %s: %i\n",
-+ watch->node, err);
++#endif /* _I386_PAGE_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/pci.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/pci.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,148 @@
++#ifndef __i386_PCI_H
++#define __i386_PCI_H
+
-+ up_read(&xs_state.watch_mutex);
+
-+ /* Cancel pending watch events. */
-+ spin_lock(&watch_events_lock);
-+ list_for_each_entry_safe(msg, tmp, &watch_events, list) {
-+ if (msg->u.watch.handle != watch)
-+ continue;
-+ list_del(&msg->list);
-+ kfree(msg->u.watch.vec);
-+ kfree(msg);
-+ }
-+ spin_unlock(&watch_events_lock);
++#ifdef __KERNEL__
++#include <linux/mm.h> /* for struct page */
+
-+ /* Flush any currently-executing callback, unless we are it. :-) */
-+ if (current->pid != xenwatch_pid) {
-+ mutex_lock(&xenwatch_mutex);
-+ mutex_unlock(&xenwatch_mutex);
-+ }
-+}
-+EXPORT_SYMBOL_GPL(unregister_xenbus_watch);
++/* Can be used to override the logic in pci_scan_bus for skipping
++ already-configured bus numbers - to be used for buggy BIOSes
++ or architectures with incomplete PCI setup by the loader */
+
-+void xs_suspend(void)
-+{
-+ down_write(&xs_state.transaction_mutex);
-+ down_write(&xs_state.watch_mutex);
-+ mutex_lock(&xs_state.request_mutex);
-+ mutex_lock(&xs_state.response_mutex);
-+}
++#ifdef CONFIG_PCI
++extern unsigned int pcibios_assign_all_busses(void);
++#else
++#define pcibios_assign_all_busses() 0
++#endif
+
-+void xs_resume(void)
-+{
-+ struct xenbus_watch *watch;
-+ char token[sizeof(watch) * 2 + 1];
++#include <asm/hypervisor.h>
++#define pcibios_scan_all_fns(a, b) (!is_initial_xendomain())
+
-+ mutex_unlock(&xs_state.response_mutex);
-+ mutex_unlock(&xs_state.request_mutex);
-+ up_write(&xs_state.transaction_mutex);
++extern unsigned long pci_mem_start;
++#define PCIBIOS_MIN_IO 0x1000
++#define PCIBIOS_MIN_MEM (pci_mem_start)
+
-+ /* No need for watches_lock: the watch_mutex is sufficient. */
-+ list_for_each_entry(watch, &watches, list) {
-+ sprintf(token, "%lX", (long)watch);
-+ xs_watch(watch->node, token);
-+ }
++#define PCIBIOS_MIN_CARDBUS_IO 0x4000
+
-+ up_write(&xs_state.watch_mutex);
-+}
++void pcibios_config_init(void);
++struct pci_bus * pcibios_scan_root(int bus);
+
-+void xs_suspend_cancel(void)
-+{
-+ mutex_unlock(&xs_state.response_mutex);
-+ mutex_unlock(&xs_state.request_mutex);
-+ up_write(&xs_state.watch_mutex);
-+ up_write(&xs_state.transaction_mutex);
-+}
++void pcibios_set_master(struct pci_dev *dev);
++void pcibios_penalize_isa_irq(int irq, int active);
++struct irq_routing_table *pcibios_get_irq_routing_table(void);
++int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq);
+
-+static int xenwatch_handle_callback(void *data)
-+{
-+ struct xs_stored_msg *msg = data;
++/* Dynamic DMA mapping stuff.
++ * i386 has everything mapped statically.
++ */
+
-+ msg->u.watch.handle->callback(msg->u.watch.handle,
-+ (const char **)msg->u.watch.vec,
-+ msg->u.watch.vec_size);
++#include <linux/types.h>
++#include <linux/slab.h>
++#include <asm/scatterlist.h>
++#include <linux/string.h>
++#include <asm/io.h>
+
-+ kfree(msg->u.watch.vec);
-+ kfree(msg);
++struct pci_dev;
+
-+ /* Kill this kthread if we were spawned just for this callback. */
-+ if (current->pid != xenwatch_pid)
-+ do_exit(0);
++#ifdef CONFIG_SWIOTLB
+
-+ return 0;
-+}
+
-+static int xenwatch_thread(void *unused)
-+{
-+ struct list_head *ent;
-+ struct xs_stored_msg *msg;
++/* On Xen we use SWIOTLB instead of blk-specific bounce buffers. */
++#define PCI_DMA_BUS_IS_PHYS (0)
+
-+ current->flags |= PF_NOFREEZE;
-+ for (;;) {
-+ wait_event_interruptible(watch_events_waitq,
-+ !list_empty(&watch_events));
++#define DECLARE_PCI_UNMAP_ADDR(ADDR_NAME) \
++ dma_addr_t ADDR_NAME;
++#define DECLARE_PCI_UNMAP_LEN(LEN_NAME) \
++ __u32 LEN_NAME;
++#define pci_unmap_addr(PTR, ADDR_NAME) \
++ ((PTR)->ADDR_NAME)
++#define pci_unmap_addr_set(PTR, ADDR_NAME, VAL) \
++ (((PTR)->ADDR_NAME) = (VAL))
++#define pci_unmap_len(PTR, LEN_NAME) \
++ ((PTR)->LEN_NAME)
++#define pci_unmap_len_set(PTR, LEN_NAME, VAL) \
++ (((PTR)->LEN_NAME) = (VAL))
+
-+ if (kthread_should_stop())
-+ break;
++#else
+
-+ mutex_lock(&xenwatch_mutex);
++/* The PCI address space does equal the physical memory
++ * address space. The networking and block device layers use
++ * this boolean for bounce buffer decisions.
++ */
++#define PCI_DMA_BUS_IS_PHYS (1)
+
-+ spin_lock(&watch_events_lock);
-+ ent = watch_events.next;
-+ if (ent != &watch_events)
-+ list_del(ent);
-+ spin_unlock(&watch_events_lock);
++/* pci_unmap_{page,single} is a nop so... */
++#define DECLARE_PCI_UNMAP_ADDR(ADDR_NAME)
++#define DECLARE_PCI_UNMAP_LEN(LEN_NAME)
++#define pci_unmap_addr(PTR, ADDR_NAME) (0)
++#define pci_unmap_addr_set(PTR, ADDR_NAME, VAL) do { } while (0)
++#define pci_unmap_len(PTR, LEN_NAME) (0)
++#define pci_unmap_len_set(PTR, LEN_NAME, VAL) do { } while (0)
+
-+ if (ent != &watch_events) {
-+ msg = list_entry(ent, struct xs_stored_msg, list);
-+ if (msg->u.watch.handle->flags & XBWF_new_thread)
-+ kthread_run(xenwatch_handle_callback,
-+ msg, "xenwatch_cb");
-+ else
-+ xenwatch_handle_callback(msg);
-+ }
++#endif
+
-+ mutex_unlock(&xenwatch_mutex);
-+ }
++/* This is always fine. */
++#define pci_dac_dma_supported(pci_dev, mask) (1)
+
-+ return 0;
++static inline dma64_addr_t
++pci_dac_page_to_dma(struct pci_dev *pdev, struct page *page, unsigned long offset, int direction)
++{
++ return ((dma64_addr_t) page_to_phys(page) +
++ (dma64_addr_t) offset);
+}
+
-+static int process_msg(void)
++static inline struct page *
++pci_dac_dma_to_page(struct pci_dev *pdev, dma64_addr_t dma_addr)
+{
-+ struct xs_stored_msg *msg;
-+ char *body;
-+ int err;
-+
-+ /*
-+ * We must disallow save/restore while reading a xenstore message.
-+ * A partial read across s/r leaves us out of sync with xenstored.
-+ */
-+ for (;;) {
-+ err = xb_wait_for_data_to_read();
-+ if (err)
-+ return err;
-+ mutex_lock(&xs_state.response_mutex);
-+ if (xb_data_to_read())
-+ break;
-+ /* We raced with save/restore: pending data 'disappeared'. */
-+ mutex_unlock(&xs_state.response_mutex);
-+ }
-+
-+
-+ msg = kmalloc(sizeof(*msg), GFP_KERNEL);
-+ if (msg == NULL) {
-+ err = -ENOMEM;
-+ goto out;
-+ }
-+
-+ err = xb_read(&msg->hdr, sizeof(msg->hdr));
-+ if (err) {
-+ kfree(msg);
-+ goto out;
-+ }
-+
-+ body = kmalloc(msg->hdr.len + 1, GFP_KERNEL);
-+ if (body == NULL) {
-+ kfree(msg);
-+ err = -ENOMEM;
-+ goto out;
-+ }
-+
-+ err = xb_read(body, msg->hdr.len);
-+ if (err) {
-+ kfree(body);
-+ kfree(msg);
-+ goto out;
-+ }
-+ body[msg->hdr.len] = '\0';
-+
-+ if (msg->hdr.type == XS_WATCH_EVENT) {
-+ msg->u.watch.vec = split(body, msg->hdr.len,
-+ &msg->u.watch.vec_size);
-+ if (IS_ERR(msg->u.watch.vec)) {
-+ err = PTR_ERR(msg->u.watch.vec);
-+ kfree(msg);
-+ goto out;
-+ }
-+
-+ spin_lock(&watches_lock);
-+ msg->u.watch.handle = find_watch(
-+ msg->u.watch.vec[XS_WATCH_TOKEN]);
-+ if (msg->u.watch.handle != NULL) {
-+ spin_lock(&watch_events_lock);
-+ list_add_tail(&msg->list, &watch_events);
-+ wake_up(&watch_events_waitq);
-+ spin_unlock(&watch_events_lock);
-+ } else {
-+ kfree(msg->u.watch.vec);
-+ kfree(msg);
-+ }
-+ spin_unlock(&watches_lock);
-+ } else {
-+ msg->u.reply.body = body;
-+ spin_lock(&xs_state.reply_lock);
-+ list_add_tail(&msg->list, &xs_state.reply_list);
-+ spin_unlock(&xs_state.reply_lock);
-+ wake_up(&xs_state.reply_waitq);
-+ }
-+
-+ out:
-+ mutex_unlock(&xs_state.response_mutex);
-+ return err;
++ return pfn_to_page(dma_addr >> PAGE_SHIFT);
+}
+
-+static int xenbus_thread(void *unused)
++static inline unsigned long
++pci_dac_dma_to_offset(struct pci_dev *pdev, dma64_addr_t dma_addr)
+{
-+ int err;
-+
-+ current->flags |= PF_NOFREEZE;
-+ for (;;) {
-+ err = process_msg();
-+ if (err)
-+ printk(KERN_WARNING "XENBUS error %d while reading "
-+ "message\n", err);
-+ if (kthread_should_stop())
-+ break;
-+ }
-+
-+ return 0;
++ return (dma_addr & ~PAGE_MASK);
+}
+
-+int xs_init(void)
++static inline void
++pci_dac_dma_sync_single_for_cpu(struct pci_dev *pdev, dma64_addr_t dma_addr, size_t len, int direction)
+{
-+ int err;
-+ struct task_struct *task;
-+
-+ INIT_LIST_HEAD(&xs_state.reply_list);
-+ spin_lock_init(&xs_state.reply_lock);
-+ init_waitqueue_head(&xs_state.reply_waitq);
++}
+
-+ mutex_init(&xs_state.request_mutex);
-+ mutex_init(&xs_state.response_mutex);
-+ init_rwsem(&xs_state.transaction_mutex);
-+ init_rwsem(&xs_state.watch_mutex);
++static inline void
++pci_dac_dma_sync_single_for_device(struct pci_dev *pdev, dma64_addr_t dma_addr, size_t len, int direction)
++{
++ flush_write_buffers();
++}
+
-+ /* Initialize the shared memory rings to talk to xenstored */
-+ err = xb_init_comms();
-+ if (err)
-+ return err;
++#define HAVE_PCI_MMAP
++extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma,
++ enum pci_mmap_state mmap_state, int write_combine);
+
-+ task = kthread_run(xenwatch_thread, NULL, "xenwatch");
-+ if (IS_ERR(task))
-+ return PTR_ERR(task);
-+ xenwatch_pid = task->pid;
+
-+ task = kthread_run(xenbus_thread, NULL, "xenbus");
-+ if (IS_ERR(task))
-+ return PTR_ERR(task);
++static inline void pcibios_add_platform_entries(struct pci_dev *dev)
++{
++}
+
-+ return 0;
++#ifdef CONFIG_PCI
++static inline void pci_dma_burst_advice(struct pci_dev *pdev,
++ enum pci_dma_burst_strategy *strat,
++ unsigned long *strategy_parameter)
++{
++ *strat = PCI_DMA_BURST_INFINITY;
++ *strategy_parameter = ~0UL;
+}
-diff -Nurp pristine-linux-2.6.18.2/drivers/xen/xenoprof/xenoprofile.c tmp-linux-2.6-xen.patch/drivers/xen/xenoprof/xenoprofile.c
---- pristine-linux-2.6.18.2/drivers/xen/xenoprof/xenoprofile.c 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/drivers/xen/xenoprof/xenoprofile.c 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,500 @@
-+/**
-+ * @file xenoprofile.c
-+ *
-+ * @remark Copyright 2002 OProfile authors
-+ * @remark Read the file COPYING
-+ *
-+ * @author John Levon <levon at movementarian.org>
-+ *
-+ * Modified by Aravind Menon and Jose Renato Santos for Xen
-+ * These modifications are:
-+ * Copyright (C) 2005 Hewlett-Packard Co.
-+ *
-+ * Separated out arch-generic part
-+ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
-+ * VA Linux Systems Japan K.K.
-+ */
++#endif
+
-+#include <linux/init.h>
-+#include <linux/notifier.h>
-+#include <linux/smp.h>
-+#include <linux/oprofile.h>
-+#include <linux/sysdev.h>
-+#include <linux/slab.h>
-+#include <linux/interrupt.h>
-+#include <linux/vmalloc.h>
-+#include <asm/pgtable.h>
-+#include <xen/evtchn.h>
-+#include <xen/xenoprof.h>
-+#include <xen/driver_util.h>
-+#include <xen/interface/xen.h>
-+#include <xen/interface/xenoprof.h>
-+#include "../../../drivers/oprofile/cpu_buffer.h"
-+#include "../../../drivers/oprofile/event_buffer.h"
++#endif /* __KERNEL__ */
+
-+#define MAX_XENOPROF_SAMPLES 16
++#ifdef CONFIG_XEN_PCIDEV_FRONTEND
++#include <xen/pcifront.h>
++#endif /* CONFIG_XEN_PCIDEV_FRONTEND */
+
-+/* sample buffers shared with Xen */
-+xenoprof_buf_t * xenoprof_buf[MAX_VIRT_CPUS];
-+/* Shared buffer area */
-+struct xenoprof_shared_buffer shared_buffer;
++/* implement the pci_ DMA API in terms of the generic device dma_ one */
++#include <asm-generic/pci-dma-compat.h>
+
-+/* Passive sample buffers shared with Xen */
-+xenoprof_buf_t *p_xenoprof_buf[MAX_OPROF_DOMAINS][MAX_VIRT_CPUS];
-+/* Passive shared buffer area */
-+struct xenoprof_shared_buffer p_shared_buffer[MAX_OPROF_DOMAINS];
++/* generic pci stuff */
++#include <asm-generic/pci.h>
+
-+static int xenoprof_start(void);
-+static void xenoprof_stop(void);
++#endif /* __i386_PCI_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/pgalloc.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/pgalloc.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,59 @@
++#ifndef _I386_PGALLOC_H
++#define _I386_PGALLOC_H
+
-+static int xenoprof_enabled = 0;
-+static int xenoprof_is_primary = 0;
-+static int active_defined;
++#include <asm/fixmap.h>
++#include <linux/threads.h>
++#include <linux/mm.h> /* for struct page */
++#include <asm/io.h> /* for phys_to_virt and page_to_pseudophys */
+
-+/* Number of buffers in shared area (one per VCPU) */
-+int nbuf;
-+/* Mappings of VIRQ_XENOPROF to irq number (per cpu) */
-+int ovf_irq[NR_CPUS];
-+/* cpu model type string - copied from Xen memory space on XENOPROF_init command */
-+char cpu_type[XENOPROF_CPU_TYPE_SIZE];
++#define pmd_populate_kernel(mm, pmd, pte) \
++ set_pmd(pmd, __pmd(_PAGE_TABLE + __pa(pte)))
+
-+#ifdef CONFIG_PM
++#define pmd_populate(mm, pmd, pte) \
++do { \
++ unsigned long pfn = page_to_pfn(pte); \
++ if (test_bit(PG_pinned, &virt_to_page((mm)->pgd)->flags)) { \
++ if (!PageHighMem(pte)) \
++ BUG_ON(HYPERVISOR_update_va_mapping( \
++ (unsigned long)__va(pfn << PAGE_SHIFT), \
++ pfn_pte(pfn, PAGE_KERNEL_RO), 0)); \
++ else if (!test_and_set_bit(PG_pinned, &pte->flags)) \
++ kmap_flush_unused(); \
++ set_pmd(pmd, \
++ __pmd(_PAGE_TABLE + ((paddr_t)pfn << PAGE_SHIFT))); \
++ } else \
++ *(pmd) = __pmd(_PAGE_TABLE + ((paddr_t)pfn << PAGE_SHIFT)); \
++} while (0)
+
-+static int xenoprof_suspend(struct sys_device * dev, pm_message_t state)
-+{
-+ if (xenoprof_enabled == 1)
-+ xenoprof_stop();
-+ return 0;
-+}
++/*
++ * Allocate and free page tables.
++ */
++extern pgd_t *pgd_alloc(struct mm_struct *);
++extern void pgd_free(pgd_t *pgd);
+
++extern pte_t *pte_alloc_one_kernel(struct mm_struct *, unsigned long);
++extern struct page *pte_alloc_one(struct mm_struct *, unsigned long);
+
-+static int xenoprof_resume(struct sys_device * dev)
++static inline void pte_free_kernel(pte_t *pte)
+{
-+ if (xenoprof_enabled == 1)
-+ xenoprof_start();
-+ return 0;
++ make_lowmem_page_writable(pte, XENFEAT_writable_page_tables);
++ free_page((unsigned long)pte);
+}
+
++extern void pte_free(struct page *pte);
+
-+static struct sysdev_class oprofile_sysclass = {
-+ set_kset_name("oprofile"),
-+ .resume = xenoprof_resume,
-+ .suspend = xenoprof_suspend
-+};
-+
-+
-+static struct sys_device device_oprofile = {
-+ .id = 0,
-+ .cls = &oprofile_sysclass,
-+};
++#define __pte_free_tlb(tlb,pte) tlb_remove_page((tlb),(pte))
+
++#ifdef CONFIG_X86_PAE
++/*
++ * In the PAE case we free the pmds as part of the pgd.
++ */
++#define pmd_alloc_one(mm, addr) ({ BUG(); ((pmd_t *)2); })
++#define pmd_free(x) do { } while (0)
++#define __pmd_free_tlb(tlb,x) do { } while (0)
++#define pud_populate(mm, pmd, pte) BUG()
++#endif
+
-+static int __init init_driverfs(void)
-+{
-+ int error;
-+ if (!(error = sysdev_class_register(&oprofile_sysclass)))
-+ error = sysdev_register(&device_oprofile);
-+ return error;
-+}
++#define check_pgt_cache() do { } while (0)
+
++#endif /* _I386_PGALLOC_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/pgtable-2level-defs.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/pgtable-2level-defs.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,20 @@
++#ifndef _I386_PGTABLE_2LEVEL_DEFS_H
++#define _I386_PGTABLE_2LEVEL_DEFS_H
+
-+static void exit_driverfs(void)
-+{
-+ sysdev_unregister(&device_oprofile);
-+ sysdev_class_unregister(&oprofile_sysclass);
-+}
++#define HAVE_SHARED_KERNEL_PMD 0
+
-+#else
-+#define init_driverfs() do { } while (0)
-+#define exit_driverfs() do { } while (0)
-+#endif /* CONFIG_PM */
++/*
++ * traditional i386 two-level paging structure:
++ */
+
-+unsigned long long oprofile_samples = 0;
-+unsigned long long p_oprofile_samples = 0;
++#define PGDIR_SHIFT 22
++#define PTRS_PER_PGD 1024
+
-+unsigned int pdomains;
-+struct xenoprof_passive passive_domains[MAX_OPROF_DOMAINS];
++/*
++ * the i386 is two-level, so we don't really have any
++ * PMD directory physically.
++ */
+
-+static void xenoprof_add_pc(xenoprof_buf_t *buf, int is_passive)
-+{
-+ int head, tail, size;
++#define PTRS_PER_PTE 1024
+
-+ head = buf->event_head;
-+ tail = buf->event_tail;
-+ size = buf->event_size;
++#endif /* _I386_PGTABLE_2LEVEL_DEFS_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/pgtable-2level.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/pgtable-2level.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,121 @@
++#ifndef _I386_PGTABLE_2LEVEL_H
++#define _I386_PGTABLE_2LEVEL_H
+
-+ if (tail > head) {
-+ while (tail < size) {
-+ oprofile_add_pc(buf->event_log[tail].eip,
-+ buf->event_log[tail].mode,
-+ buf->event_log[tail].event);
-+ if (!is_passive)
-+ oprofile_samples++;
-+ else
-+ p_oprofile_samples++;
-+ tail++;
-+ }
-+ tail = 0;
-+ }
-+ while (tail < head) {
-+ oprofile_add_pc(buf->event_log[tail].eip,
-+ buf->event_log[tail].mode,
-+ buf->event_log[tail].event);
-+ if (!is_passive)
-+ oprofile_samples++;
-+ else
-+ p_oprofile_samples++;
-+ tail++;
-+ }
++#include <asm-generic/pgtable-nopmd.h>
+
-+ buf->event_tail = tail;
-+}
++#define pte_ERROR(e) \
++ printk("%s:%d: bad pte %08lx (pfn %05lx).\n", __FILE__, __LINE__, \
++ __pte_val(e), pte_pfn(e))
++#define pgd_ERROR(e) \
++ printk("%s:%d: bad pgd %08lx (pfn %05lx).\n", __FILE__, __LINE__, \
++ __pgd_val(e), pgd_val(e) >> PAGE_SHIFT)
+
-+static void xenoprof_handle_passive(void)
-+{
-+ int i, j;
-+ int flag_domain, flag_switch = 0;
-+
-+ for (i = 0; i < pdomains; i++) {
-+ flag_domain = 0;
-+ for (j = 0; j < passive_domains[i].nbuf; j++) {
-+ xenoprof_buf_t *buf = p_xenoprof_buf[i][j];
-+ if (buf->event_head == buf->event_tail)
-+ continue;
-+ if (!flag_domain) {
-+ if (!oprofile_add_domain_switch(passive_domains[i].
-+ domain_id))
-+ goto done;
-+ flag_domain = 1;
-+ }
-+ xenoprof_add_pc(buf, 1);
-+ flag_switch = 1;
-+ }
-+ }
-+done:
-+ if (flag_switch)
-+ oprofile_add_domain_switch(COORDINATOR_DOMAIN);
-+}
++/*
++ * Certain architectures need to do special things when PTEs
++ * within a page table are directly modified. Thus, the following
++ * hook is made available.
++ */
++#define set_pte(pteptr, pteval) (*(pteptr) = pteval)
+
-+static irqreturn_t
-+xenoprof_ovf_interrupt(int irq, void * dev_id, struct pt_regs * regs)
-+{
-+ struct xenoprof_buf * buf;
-+ int cpu;
-+ static unsigned long flag;
++#define set_pte_at(_mm,addr,ptep,pteval) do { \
++ if (((_mm) != current->mm && (_mm) != &init_mm) || \
++ HYPERVISOR_update_va_mapping((addr), (pteval), 0)) \
++ set_pte((ptep), (pteval)); \
++} while (0)
+
-+ cpu = smp_processor_id();
-+ buf = xenoprof_buf[cpu];
++#define set_pte_at_sync(_mm,addr,ptep,pteval) do { \
++ if (((_mm) != current->mm && (_mm) != &init_mm) || \
++ HYPERVISOR_update_va_mapping((addr), (pteval), UVMF_INVLPG)) { \
++ set_pte((ptep), (pteval)); \
++ xen_invlpg((addr)); \
++ } \
++} while (0)
+
-+ xenoprof_add_pc(buf, 0);
++#define set_pte_atomic(pteptr, pteval) set_pte(pteptr,pteval)
+
-+ if (xenoprof_is_primary && !test_and_set_bit(0, &flag)) {
-+ xenoprof_handle_passive();
-+ smp_mb__before_clear_bit();
-+ clear_bit(0, &flag);
-+ }
++#define set_pmd(pmdptr, pmdval) xen_l2_entry_update((pmdptr), (pmdval))
+
-+ return IRQ_HANDLED;
-+}
++#define pte_clear(mm,addr,xp) do { set_pte_at(mm, addr, xp, __pte(0)); } while (0)
++#define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0)
+
++#define pte_none(x) (!(x).pte_low)
+
-+static void unbind_virq(void)
++static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
+{
-+ int i;
-+
-+ for_each_online_cpu(i) {
-+ if (ovf_irq[i] >= 0) {
-+ unbind_from_irqhandler(ovf_irq[i], NULL);
-+ ovf_irq[i] = -1;
-+ }
++ pte_t pte = *ptep;
++ if (!pte_none(pte)) {
++ if ((mm != &init_mm) ||
++ HYPERVISOR_update_va_mapping(addr, __pte(0), 0))
++ pte = __pte_ma(xchg(&ptep->pte_low, 0));
+ }
++ return pte;
+}
+
++#define ptep_clear_flush(vma, addr, ptep) \
++({ \
++ pte_t *__ptep = (ptep); \
++ pte_t __res = *__ptep; \
++ if (!pte_none(__res) && \
++ ((vma)->vm_mm != current->mm || \
++ HYPERVISOR_update_va_mapping(addr, __pte(0), \
++ (unsigned long)(vma)->vm_mm->cpu_vm_mask.bits| \
++ UVMF_INVLPG|UVMF_MULTI))) { \
++ __ptep->pte_low = 0; \
++ flush_tlb_page(vma, addr); \
++ } \
++ __res; \
++})
+
-+static int bind_virq(void)
-+{
-+ int i, result;
-+
-+ for_each_online_cpu(i) {
-+ result = bind_virq_to_irqhandler(VIRQ_XENOPROF,
-+ i,
-+ xenoprof_ovf_interrupt,
-+ SA_INTERRUPT,
-+ "xenoprof",
-+ NULL);
++#define pte_same(a, b) ((a).pte_low == (b).pte_low)
+
-+ if (result < 0) {
-+ unbind_virq();
-+ return result;
-+ }
++#define __pte_mfn(_pte) ((_pte).pte_low >> PAGE_SHIFT)
++#define pte_mfn(_pte) ((_pte).pte_low & _PAGE_PRESENT ? \
++ __pte_mfn(_pte) : pfn_to_mfn(__pte_mfn(_pte)))
++#define pte_pfn(_pte) ((_pte).pte_low & _PAGE_IO ? max_mapnr : \
++ (_pte).pte_low & _PAGE_PRESENT ? \
++ mfn_to_local_pfn(__pte_mfn(_pte)) : \
++ __pte_mfn(_pte))
+
-+ ovf_irq[i] = result;
-+ }
-+
-+ return 0;
-+}
++#define pte_page(_pte) pfn_to_page(pte_pfn(_pte))
+
++#define pfn_pte(pfn, prot) __pte(((pfn) << PAGE_SHIFT) | pgprot_val(prot))
++#define pfn_pmd(pfn, prot) __pmd(((pfn) << PAGE_SHIFT) | pgprot_val(prot))
+
-+static void unmap_passive_list(void)
++/*
++ * All present user pages are user-executable:
++ */
++static inline int pte_exec(pte_t pte)
+{
-+ int i;
-+ for (i = 0; i < pdomains; i++)
-+ xenoprof_arch_unmap_shared_buffer(&p_shared_buffer[i]);
-+ pdomains = 0;
++ return pte_user(pte);
+}
+
-+
-+static int map_xenoprof_buffer(int max_samples)
++/*
++ * All present pages are kernel-executable:
++ */
++static inline int pte_exec_kernel(pte_t pte)
+{
-+ struct xenoprof_get_buffer get_buffer;
-+ struct xenoprof_buf *buf;
-+ int ret, i;
-+
-+ if ( shared_buffer.buffer )
-+ return 0;
-+
-+ get_buffer.max_samples = max_samples;
-+ ret = xenoprof_arch_map_shared_buffer(&get_buffer, &shared_buffer);
-+ if (ret)
-+ return ret;
-+ nbuf = get_buffer.nbuf;
-+
-+ for (i=0; i< nbuf; i++) {
-+ buf = (struct xenoprof_buf*)
-+ &shared_buffer.buffer[i * get_buffer.bufsize];
-+ BUG_ON(buf->vcpu_id >= MAX_VIRT_CPUS);
-+ xenoprof_buf[buf->vcpu_id] = buf;
-+ }
-+
-+ return 0;
++ return 1;
+}
+
++/*
++ * Bits 0, 6 and 7 are taken, split up the 29 bits of offset
++ * into this range:
++ */
++#define PTE_FILE_MAX_BITS 29
+
-+static int xenoprof_setup(void)
-+{
-+ int ret;
-+
-+ if ( (ret = map_xenoprof_buffer(MAX_XENOPROF_SAMPLES)) )
-+ return ret;
++#define pte_to_pgoff(pte) \
++ ((((pte).pte_low >> 1) & 0x1f ) + (((pte).pte_low >> 8) << 5 ))
+
-+ if ( (ret = bind_virq()) )
-+ return ret;
++#define pgoff_to_pte(off) \
++ ((pte_t) { (((off) & 0x1f) << 1) + (((off) >> 5) << 8) + _PAGE_FILE })
+
-+ if (xenoprof_is_primary) {
-+ /* Define dom0 as an active domain if not done yet */
-+ if (!active_defined) {
-+ domid_t domid;
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_reset_active_list, NULL);
-+ if (ret)
-+ goto err;
-+ domid = 0;
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_set_active, &domid);
-+ if (ret)
-+ goto err;
-+ active_defined = 1;
-+ }
++/* Encode and de-code a swap entry */
++#define __swp_type(x) (((x).val >> 1) & 0x1f)
++#define __swp_offset(x) ((x).val >> 8)
++#define __swp_entry(type, offset) ((swp_entry_t) { ((type) << 1) | ((offset) << 8) })
++#define __pte_to_swp_entry(pte) ((swp_entry_t) { (pte).pte_low })
++#define __swp_entry_to_pte(x) ((pte_t) { (x).val })
+
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_reserve_counters, NULL);
-+ if (ret)
-+ goto err;
-+ xenoprof_arch_counter();
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_setup_events, NULL);
++void vmalloc_sync_all(void);
+
-+ if (ret)
-+ goto err;
-+ }
++#endif /* _I386_PGTABLE_2LEVEL_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/pgtable-3level-defs.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/pgtable-3level-defs.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,24 @@
++#ifndef _I386_PGTABLE_3LEVEL_DEFS_H
++#define _I386_PGTABLE_3LEVEL_DEFS_H
+
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_enable_virq, NULL);
-+ if (ret)
-+ goto err;
++#define HAVE_SHARED_KERNEL_PMD 0
+
-+ xenoprof_enabled = 1;
-+ return 0;
-+ err:
-+ unbind_virq();
-+ return ret;
-+}
++/*
++ * PGDIR_SHIFT determines what a top-level page table entry can map
++ */
++#define PGDIR_SHIFT 30
++#define PTRS_PER_PGD 4
+
++/*
++ * PMD_SHIFT determines the size of the area a middle-level
++ * page table can map
++ */
++#define PMD_SHIFT 21
++#define PTRS_PER_PMD 512
+
-+static void xenoprof_shutdown(void)
-+{
-+ xenoprof_enabled = 0;
++/*
++ * entries per page directory level
++ */
++#define PTRS_PER_PTE 512
+
-+ HYPERVISOR_xenoprof_op(XENOPROF_disable_virq, NULL);
++#endif /* _I386_PGTABLE_3LEVEL_DEFS_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/pgtable-3level.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/pgtable-3level.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,211 @@
++#ifndef _I386_PGTABLE_3LEVEL_H
++#define _I386_PGTABLE_3LEVEL_H
+
-+ if (xenoprof_is_primary) {
-+ HYPERVISOR_xenoprof_op(XENOPROF_release_counters, NULL);
-+ active_defined = 0;
-+ }
++#include <asm-generic/pgtable-nopud.h>
+
-+ unbind_virq();
++/*
++ * Intel Physical Address Extension (PAE) Mode - three-level page
++ * tables on PPro+ CPUs.
++ *
++ * Copyright (C) 1999 Ingo Molnar <mingo at redhat.com>
++ */
+
-+ xenoprof_arch_unmap_shared_buffer(&shared_buffer);
-+ if (xenoprof_is_primary)
-+ unmap_passive_list();
-+}
++#define pte_ERROR(e) \
++ printk("%s:%d: bad pte %p(%016Lx pfn %08lx).\n", __FILE__, __LINE__, \
++ &(e), __pte_val(e), pte_pfn(e))
++#define pmd_ERROR(e) \
++ printk("%s:%d: bad pmd %p(%016Lx pfn %08Lx).\n", __FILE__, __LINE__, \
++ &(e), __pmd_val(e), (pmd_val(e) & PTE_MASK) >> PAGE_SHIFT)
++#define pgd_ERROR(e) \
++ printk("%s:%d: bad pgd %p(%016Lx pfn %08Lx).\n", __FILE__, __LINE__, \
++ &(e), __pgd_val(e), (pgd_val(e) & PTE_MASK) >> PAGE_SHIFT)
+
++#define pud_none(pud) 0
++#define pud_bad(pud) 0
++#define pud_present(pud) 1
+
-+static int xenoprof_start(void)
++/*
++ * Is the pte executable?
++ */
++static inline int pte_x(pte_t pte)
+{
-+ int ret = 0;
-+
-+ if (xenoprof_is_primary)
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_start, NULL);
-+ if (!ret)
-+ xenoprof_arch_start();
-+ return ret;
++ return !(__pte_val(pte) & _PAGE_NX);
+}
+
-+
-+static void xenoprof_stop(void)
++/*
++ * All present user-pages with !NX bit are user-executable:
++ */
++static inline int pte_exec(pte_t pte)
+{
-+ if (xenoprof_is_primary)
-+ HYPERVISOR_xenoprof_op(XENOPROF_stop, NULL);
-+ xenoprof_arch_stop();
++ return pte_user(pte) && pte_x(pte);
++}
++/*
++ * All present pages with !NX bit are kernel-executable:
++ */
++static inline int pte_exec_kernel(pte_t pte)
++{
++ return pte_x(pte);
+}
+
++/* Rules for using set_pte: the pte being assigned *must* be
++ * either not present or in a state where the hardware will
++ * not attempt to update the pte. In places where this is
++ * not possible, use pte_get_and_clear to obtain the old pte
++ * value and then use set_pte to update it. -ben
++ */
++#define __HAVE_ARCH_SET_PTE_ATOMIC
+
-+static int xenoprof_set_active(int * active_domains,
-+ unsigned int adomains)
++static inline void set_pte(pte_t *ptep, pte_t pte)
+{
-+ int ret = 0;
-+ int i;
-+ int set_dom0 = 0;
-+ domid_t domid;
++ ptep->pte_high = pte.pte_high;
++ smp_wmb();
++ ptep->pte_low = pte.pte_low;
++}
++#define set_pte_atomic(pteptr,pteval) \
++ set_64bit((unsigned long long *)(pteptr),__pte_val(pteval))
+
-+ if (!xenoprof_is_primary)
-+ return 0;
++#define set_pte_at(_mm,addr,ptep,pteval) do { \
++ if (((_mm) != current->mm && (_mm) != &init_mm) || \
++ HYPERVISOR_update_va_mapping((addr), (pteval), 0)) \
++ set_pte((ptep), (pteval)); \
++} while (0)
+
-+ if (adomains > MAX_OPROF_DOMAINS)
-+ return -E2BIG;
++#define set_pte_at_sync(_mm,addr,ptep,pteval) do { \
++ if (((_mm) != current->mm && (_mm) != &init_mm) || \
++ HYPERVISOR_update_va_mapping((addr), (pteval), UVMF_INVLPG)) { \
++ set_pte((ptep), (pteval)); \
++ xen_invlpg((addr)); \
++ } \
++} while (0)
+
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_reset_active_list, NULL);
-+ if (ret)
-+ return ret;
++#define set_pmd(pmdptr,pmdval) \
++ xen_l2_entry_update((pmdptr), (pmdval))
++#define set_pud(pudptr,pudval) \
++ xen_l3_entry_update((pudptr), (pudval))
+
-+ for (i=0; i<adomains; i++) {
-+ domid = active_domains[i];
-+ if (domid != active_domains[i]) {
-+ ret = -EINVAL;
-+ goto out;
-+ }
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_set_active, &domid);
-+ if (ret)
-+ goto out;
-+ if (active_domains[i] == 0)
-+ set_dom0 = 1;
-+ }
-+ /* dom0 must always be active but may not be in the list */
-+ if (!set_dom0) {
-+ domid = 0;
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_set_active, &domid);
-+ }
++/*
++ * Pentium-II erratum A13: in PAE mode we explicitly have to flush
++ * the TLB via cr3 if the top-level pgd is changed...
++ * We do not let the generic code free and clear pgd entries due to
++ * this erratum.
++ */
++static inline void pud_clear (pud_t * pud) { }
+
-+out:
-+ if (ret)
-+ HYPERVISOR_xenoprof_op(XENOPROF_reset_active_list, NULL);
-+ active_defined = !ret;
-+ return ret;
-+}
++#define pud_page(pud) \
++((struct page *) __va(pud_val(pud) & PAGE_MASK))
+
-+static int xenoprof_set_passive(int * p_domains,
-+ unsigned int pdoms)
-+{
-+ int ret;
-+ int i, j;
-+ struct xenoprof_buf *buf;
++#define pud_page_kernel(pud) \
++((unsigned long) __va(pud_val(pud) & PAGE_MASK))
+
-+ if (!xenoprof_is_primary)
-+ return 0;
+
-+ if (pdoms > MAX_OPROF_DOMAINS)
-+ return -E2BIG;
++/* Find an entry in the second-level page table.. */
++#define pmd_offset(pud, address) ((pmd_t *) pud_page(*(pud)) + \
++ pmd_index(address))
+
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_reset_passive_list, NULL);
-+ if (ret)
-+ return ret;
-+ unmap_passive_list();
++static inline int pte_none(pte_t pte)
++{
++ return !(pte.pte_low | pte.pte_high);
++}
+
-+ for (i = 0; i < pdoms; i++) {
-+ passive_domains[i].domain_id = p_domains[i];
-+ passive_domains[i].max_samples = 2048;
-+ ret = xenoprof_arch_set_passive(&passive_domains[i],
-+ &p_shared_buffer[i]);
-+ if (ret)
-+ goto out;
-+ for (j = 0; j < passive_domains[i].nbuf; j++) {
-+ buf = (struct xenoprof_buf *)
-+ &p_shared_buffer[i].buffer[j * passive_domains[i].bufsize];
-+ BUG_ON(buf->vcpu_id >= MAX_VIRT_CPUS);
-+ p_xenoprof_buf[i][buf->vcpu_id] = buf;
-+ }
++/*
++ * For PTEs and PDEs, we must clear the P-bit first when clearing a page table
++ * entry, so clear the bottom half first and enforce ordering with a compiler
++ * barrier.
++ */
++static inline void pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
++{
++ if ((mm != current->mm && mm != &init_mm)
++ || HYPERVISOR_update_va_mapping(addr, __pte(0), 0)) {
++ ptep->pte_low = 0;
++ smp_wmb();
++ ptep->pte_high = 0;
+ }
++}
+
-+ pdomains = pdoms;
-+ return 0;
++#define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0)
+
-+out:
-+ for (j = 0; j < i; j++)
-+ xenoprof_arch_unmap_shared_buffer(&p_shared_buffer[i]);
++static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
++{
++ pte_t pte = *ptep;
++ if (!pte_none(pte)) {
++ if ((mm != &init_mm) ||
++ HYPERVISOR_update_va_mapping(addr, __pte(0), 0)) {
++ uint64_t val = __pte_val(pte);
++ if (__cmpxchg64(ptep, val, 0) != val) {
++ /* xchg acts as a barrier before the setting of the high bits */
++ pte.pte_low = xchg(&ptep->pte_low, 0);
++ pte.pte_high = ptep->pte_high;
++ ptep->pte_high = 0;
++ }
++ }
++ }
++ return pte;
++}
++
++#define ptep_clear_flush(vma, addr, ptep) \
++({ \
++ pte_t *__ptep = (ptep); \
++ pte_t __res = *__ptep; \
++ if (!pte_none(__res) && \
++ ((vma)->vm_mm != current->mm || \
++ HYPERVISOR_update_va_mapping(addr, __pte(0), \
++ (unsigned long)(vma)->vm_mm->cpu_vm_mask.bits| \
++ UVMF_INVLPG|UVMF_MULTI))) { \
++ __ptep->pte_low = 0; \
++ smp_wmb(); \
++ __ptep->pte_high = 0; \
++ flush_tlb_page(vma, addr); \
++ } \
++ __res; \
++})
+
-+ return ret;
++static inline int pte_same(pte_t a, pte_t b)
++{
++ return a.pte_low == b.pte_low && a.pte_high == b.pte_high;
+}
+
-+struct oprofile_operations xenoprof_ops = {
-+#ifdef HAVE_XENOPROF_CREATE_FILES
-+ .create_files = xenoprof_create_files,
-+#endif
-+ .set_active = xenoprof_set_active,
-+ .set_passive = xenoprof_set_passive,
-+ .setup = xenoprof_setup,
-+ .shutdown = xenoprof_shutdown,
-+ .start = xenoprof_start,
-+ .stop = xenoprof_stop
-+};
++#define pte_page(x) pfn_to_page(pte_pfn(x))
+
++#define __pte_mfn(_pte) (((_pte).pte_low >> PAGE_SHIFT) | \
++ ((_pte).pte_high << (32-PAGE_SHIFT)))
++#define pte_mfn(_pte) ((_pte).pte_low & _PAGE_PRESENT ? \
++ __pte_mfn(_pte) : pfn_to_mfn(__pte_mfn(_pte)))
++#define pte_pfn(_pte) ((_pte).pte_low & _PAGE_IO ? max_mapnr : \
++ (_pte).pte_low & _PAGE_PRESENT ? \
++ mfn_to_local_pfn(__pte_mfn(_pte)) : \
++ __pte_mfn(_pte))
+
-+/* in order to get driverfs right */
-+static int using_xenoprof;
++extern unsigned long long __supported_pte_mask;
+
-+int __init xenoprofile_init(struct oprofile_operations * ops)
++static inline pte_t pfn_pte(unsigned long page_nr, pgprot_t pgprot)
+{
-+ struct xenoprof_init init;
-+ int ret, i;
++ return __pte((((unsigned long long)page_nr << PAGE_SHIFT) |
++ pgprot_val(pgprot)) & __supported_pte_mask);
++}
+
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_init, &init);
-+ if (!ret) {
-+ xenoprof_arch_init_counter(&init);
-+ xenoprof_is_primary = init.is_primary;
++static inline pmd_t pfn_pmd(unsigned long page_nr, pgprot_t pgprot)
++{
++ return __pmd((((unsigned long long)page_nr << PAGE_SHIFT) |
++ pgprot_val(pgprot)) & __supported_pte_mask);
++}
+
-+ /* cpu_type is detected by Xen */
-+ cpu_type[XENOPROF_CPU_TYPE_SIZE-1] = 0;
-+ strncpy(cpu_type, init.cpu_type, XENOPROF_CPU_TYPE_SIZE - 1);
-+ xenoprof_ops.cpu_type = cpu_type;
++/*
++ * Bits 0, 6 and 7 are taken in the low part of the pte,
++ * put the 32 bits of offset into the high part.
++ */
++#define pte_to_pgoff(pte) ((pte).pte_high)
++#define pgoff_to_pte(off) ((pte_t) { _PAGE_FILE, (off) })
++#define PTE_FILE_MAX_BITS 32
+
-+ init_driverfs();
-+ using_xenoprof = 1;
-+ *ops = xenoprof_ops;
++/* Encode and de-code a swap entry */
++#define __swp_type(x) (((x).val) & 0x1f)
++#define __swp_offset(x) ((x).val >> 5)
++#define __swp_entry(type, offset) ((swp_entry_t){(type) | (offset) << 5})
++#define __pte_to_swp_entry(pte) ((swp_entry_t){ (pte).pte_high })
++#define __swp_entry_to_pte(x) ((pte_t){ 0, (x).val })
+
-+ for (i=0; i<NR_CPUS; i++)
-+ ovf_irq[i] = -1;
++#define __pmd_free_tlb(tlb, x) do { } while (0)
+
-+ active_defined = 0;
-+ }
-+ printk(KERN_INFO "%s: ret %d, events %d, xenoprof_is_primary %d\n",
-+ __func__, ret, init.num_events, xenoprof_is_primary);
-+ return ret;
-+}
++void vmalloc_sync_all(void);
+
++#endif /* _I386_PGTABLE_3LEVEL_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/pgtable.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/pgtable.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,537 @@
++#ifndef _I386_PGTABLE_H
++#define _I386_PGTABLE_H
+
-+void xenoprofile_exit(void)
-+{
-+ if (using_xenoprof)
-+ exit_driverfs();
++#include <asm/hypervisor.h>
+
-+ xenoprof_arch_unmap_shared_buffer(&shared_buffer);
-+ if (xenoprof_is_primary) {
-+ unmap_passive_list();
-+ HYPERVISOR_xenoprof_op(XENOPROF_shutdown, NULL);
-+ }
-+}
-diff -Nurp pristine-linux-2.6.18.2/fs/aio.c tmp-linux-2.6-xen.patch/fs/aio.c
---- pristine-linux-2.6.18.2/fs/aio.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/fs/aio.c 2007-10-14 01:51:15.000000000 +0200
-@@ -34,6 +34,11 @@
- #include <asm/uaccess.h>
- #include <asm/mmu_context.h>
-
-+#ifdef CONFIG_EPOLL
-+#include <linux/poll.h>
-+#include <linux/eventpoll.h>
-+#endif
++/*
++ * The Linux memory management assumes a three-level page table setup. On
++ * the i386, we use that, but "fold" the mid level into the top-level page
++ * table, so that we physically have the same two-level page table as the
++ * i386 mmu expects.
++ *
++ * This file contains the functions and defines necessary to modify and use
++ * the i386 page table tree.
++ */
++#ifndef __ASSEMBLY__
++#include <asm/processor.h>
++#include <asm/fixmap.h>
++#include <linux/threads.h>
+
- #if DEBUG > 1
- #define dprintk printk
- #else
-@@ -1015,6 +1020,10 @@ put_rq:
- if (waitqueue_active(&ctx->wait))
- wake_up(&ctx->wait);
-
-+#ifdef CONFIG_EPOLL
-+ if (ctx->file && waitqueue_active(&ctx->poll_wait))
-+ wake_up(&ctx->poll_wait);
-+#endif
- if (ret)
- put_ioctx(ctx);
-
-@@ -1024,6 +1033,8 @@ put_rq:
- /* aio_read_evt
- * Pull an event off of the ioctx's event ring. Returns the number of
- * events fetched (0 or 1 ;-)
-+ * If ent parameter is 0, just returns the number of events that would
-+ * be fetched.
- * FIXME: make this use cmpxchg.
- * TODO: make the ringbuffer user mmap()able (requires FIXME).
- */
-@@ -1046,13 +1057,18 @@ static int aio_read_evt(struct kioctx *i
-
- head = ring->head % info->nr;
- if (head != ring->tail) {
-- struct io_event *evp = aio_ring_event(info, head, KM_USER1);
-- *ent = *evp;
-- head = (head + 1) % info->nr;
-- smp_mb(); /* finish reading the event before updatng the head */
-- ring->head = head;
-- ret = 1;
-- put_aio_ring_event(evp, KM_USER1);
-+ if (ent) { /* event requested */
-+ struct io_event *evp =
-+ aio_ring_event(info, head, KM_USER1);
-+ *ent = *evp;
-+ head = (head + 1) % info->nr;
-+ /* finish reading the event before updatng the head */
-+ smp_mb();
-+ ring->head = head;
-+ ret = 1;
-+ put_aio_ring_event(evp, KM_USER1);
-+ } else /* only need to know availability */
-+ ret = 1;
- }
- spin_unlock(&info->ring_lock);
-
-@@ -1235,9 +1251,78 @@ static void io_destroy(struct kioctx *io
-
- aio_cancel_all(ioctx);
- wait_for_all_aios(ioctx);
-+#ifdef CONFIG_EPOLL
-+ /* forget the poll file, but it's up to the user to close it */
-+ if (ioctx->file) {
-+ ioctx->file->private_data = 0;
-+ ioctx->file = 0;
-+ }
++#ifndef _I386_BITOPS_H
++#include <asm/bitops.h>
+#endif
- put_ioctx(ioctx); /* once for the lookup */
- }
-
-+#ifdef CONFIG_EPOLL
-+
-+static int aio_queue_fd_close(struct inode *inode, struct file *file)
-+{
-+ struct kioctx *ioctx = file->private_data;
-+ if (ioctx) {
-+ file->private_data = 0;
-+ spin_lock_irq(&ioctx->ctx_lock);
-+ ioctx->file = 0;
-+ spin_unlock_irq(&ioctx->ctx_lock);
-+ }
-+ return 0;
-+}
-+
-+static unsigned int aio_queue_fd_poll(struct file *file, poll_table *wait)
-+{ unsigned int pollflags = 0;
-+ struct kioctx *ioctx = file->private_data;
+
-+ if (ioctx) {
++#include <linux/slab.h>
++#include <linux/list.h>
++#include <linux/spinlock.h>
+
-+ spin_lock_irq(&ioctx->ctx_lock);
-+ /* Insert inside our poll wait queue */
-+ poll_wait(file, &ioctx->poll_wait, wait);
++/* Is this pagetable pinned? */
++#define PG_pinned PG_arch_1
+
-+ /* Check our condition */
-+ if (aio_read_evt(ioctx, 0))
-+ pollflags = POLLIN | POLLRDNORM;
-+ spin_unlock_irq(&ioctx->ctx_lock);
-+ }
++struct mm_struct;
++struct vm_area_struct;
+
-+ return pollflags;
-+}
++/*
++ * ZERO_PAGE is a global shared page that is always zero: used
++ * for zero-mapped memory areas etc..
++ */
++#define ZERO_PAGE(vaddr) (virt_to_page(empty_zero_page))
++extern unsigned long empty_zero_page[1024];
++extern pgd_t *swapper_pg_dir;
++extern kmem_cache_t *pgd_cache;
++extern kmem_cache_t *pmd_cache;
++extern spinlock_t pgd_lock;
++extern struct page *pgd_list;
+
-+static const struct file_operations aioq_fops = {
-+ .release = aio_queue_fd_close,
-+ .poll = aio_queue_fd_poll
-+};
++void pmd_ctor(void *, kmem_cache_t *, unsigned long);
++void pgd_ctor(void *, kmem_cache_t *, unsigned long);
++void pgd_dtor(void *, kmem_cache_t *, unsigned long);
++void pgtable_cache_init(void);
++void paging_init(void);
+
-+/* make_aio_fd:
-+ * Create a file descriptor that can be used to poll the event queue.
-+ * Based and piggybacked on the excellent epoll code.
++/*
++ * The Linux x86 paging architecture is 'compile-time dual-mode', it
++ * implements both the traditional 2-level x86 page tables and the
++ * newer 3-level PAE-mode page tables.
+ */
++#ifdef CONFIG_X86_PAE
++# include <asm/pgtable-3level-defs.h>
++# define PMD_SIZE (1UL << PMD_SHIFT)
++# define PMD_MASK (~(PMD_SIZE-1))
++#else
++# include <asm/pgtable-2level-defs.h>
++#endif
+
-+static int make_aio_fd(struct kioctx *ioctx)
-+{
-+ int error, fd;
-+ struct inode *inode;
-+ struct file *file;
++#define PGDIR_SIZE (1UL << PGDIR_SHIFT)
++#define PGDIR_MASK (~(PGDIR_SIZE-1))
+
-+ error = ep_getfd(&fd, &inode, &file, NULL, &aioq_fops);
-+ if (error)
-+ return error;
++#define USER_PTRS_PER_PGD (TASK_SIZE/PGDIR_SIZE)
++#define FIRST_USER_ADDRESS 0
+
-+ /* associate the file with the IO context */
-+ file->private_data = ioctx;
-+ ioctx->file = file;
-+ init_waitqueue_head(&ioctx->poll_wait);
-+ return fd;
-+}
-+#endif
++#define USER_PGD_PTRS (PAGE_OFFSET >> PGDIR_SHIFT)
++#define KERNEL_PGD_PTRS (PTRS_PER_PGD-USER_PGD_PTRS)
+
++#define TWOLEVEL_PGDIR_SHIFT 22
++#define BOOT_USER_PGD_PTRS (__PAGE_OFFSET >> TWOLEVEL_PGDIR_SHIFT)
++#define BOOT_KERNEL_PGD_PTRS (1024-BOOT_USER_PGD_PTRS)
+
- /* sys_io_setup:
- * Create an aio_context capable of receiving at least nr_events.
- * ctxp must not point to an aio_context that already exists, and
-@@ -1250,18 +1335,30 @@ static void io_destroy(struct kioctx *io
- * resources are available. May fail with -EFAULT if an invalid
- * pointer is passed for ctxp. Will fail with -ENOSYS if not
- * implemented.
-+ *
-+ * To request a selectable fd, the user context has to be initialized
-+ * to 1, instead of 0, and the return value is the fd.
-+ * This keeps the system call compatible, since a non-zero value
-+ * was not allowed so far.
- */
- asmlinkage long sys_io_setup(unsigned nr_events, aio_context_t __user *ctxp)
- {
- struct kioctx *ioctx = NULL;
- unsigned long ctx;
- long ret;
-+ int make_fd = 0;
-
- ret = get_user(ctx, ctxp);
- if (unlikely(ret))
- goto out;
-
- ret = -EINVAL;
-+#ifdef CONFIG_EPOLL
-+ if (ctx == 1) {
-+ make_fd = 1;
-+ ctx = 0;
-+ }
-+#endif
- if (unlikely(ctx || nr_events == 0)) {
- pr_debug("EINVAL: io_setup: ctx %lu nr_events %u\n",
- ctx, nr_events);
-@@ -1272,8 +1369,12 @@ asmlinkage long sys_io_setup(unsigned nr
- ret = PTR_ERR(ioctx);
- if (!IS_ERR(ioctx)) {
- ret = put_user(ioctx->user_id, ctxp);
-- if (!ret)
-- return 0;
-+#ifdef CONFIG_EPOLL
-+ if (make_fd && ret >= 0)
-+ ret = make_aio_fd(ioctx);
-+#endif
-+ if (ret >= 0)
-+ return ret;
-
- get_ioctx(ioctx); /* io_destroy() expects us to hold a ref */
- io_destroy(ioctx);
-diff -Nurp pristine-linux-2.6.18.2/fs/eventpoll.c tmp-linux-2.6-xen.patch/fs/eventpoll.c
---- pristine-linux-2.6.18.2/fs/eventpoll.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/fs/eventpoll.c 2007-10-14 01:51:15.000000000 +0200
-@@ -236,8 +236,6 @@ struct ep_pqueue {
-
- static void ep_poll_safewake_init(struct poll_safewake *psw);
- static void ep_poll_safewake(struct poll_safewake *psw, wait_queue_head_t *wq);
--static int ep_getfd(int *efd, struct inode **einode, struct file **efile,
-- struct eventpoll *ep);
- static int ep_alloc(struct eventpoll **pep);
- static void ep_free(struct eventpoll *ep);
- static struct epitem *ep_find(struct eventpoll *ep, struct file *file, int fd);
-@@ -267,7 +265,7 @@ static int ep_events_transfer(struct eve
- static int ep_poll(struct eventpoll *ep, struct epoll_event __user *events,
- int maxevents, long timeout);
- static int eventpollfs_delete_dentry(struct dentry *dentry);
--static struct inode *ep_eventpoll_inode(void);
-+static struct inode *ep_eventpoll_inode(const struct file_operations *fops);
- static int eventpollfs_get_sb(struct file_system_type *fs_type,
- int flags, const char *dev_name,
- void *data, struct vfsmount *mnt);
-@@ -517,7 +515,7 @@ asmlinkage long sys_epoll_create(int siz
- * Creates all the items needed to setup an eventpoll file. That is,
- * a file structure, and inode and a free file descriptor.
- */
-- error = ep_getfd(&fd, &inode, &file, ep);
-+ error = ep_getfd(&fd, &inode, &file, ep, &eventpoll_fops);
- if (error)
- goto eexit_2;
-
-@@ -702,8 +700,8 @@ eexit_1:
- /*
- * Creates the file descriptor to be used by the epoll interface.
- */
--static int ep_getfd(int *efd, struct inode **einode, struct file **efile,
-- struct eventpoll *ep)
-+int ep_getfd(int *efd, struct inode **einode, struct file **efile,
-+ struct eventpoll *ep, const struct file_operations *fops)
- {
- struct qstr this;
- char name[32];
-@@ -719,7 +717,7 @@ static int ep_getfd(int *efd, struct ino
- goto eexit_1;
-
- /* Allocates an inode from the eventpoll file system */
-- inode = ep_eventpoll_inode();
-+ inode = ep_eventpoll_inode(fops);
- error = PTR_ERR(inode);
- if (IS_ERR(inode))
- goto eexit_2;
-@@ -750,7 +748,7 @@ static int ep_getfd(int *efd, struct ino
-
- file->f_pos = 0;
- file->f_flags = O_RDONLY;
-- file->f_op = &eventpoll_fops;
-+ file->f_op = fops;
- file->f_mode = FMODE_READ;
- file->f_version = 0;
- file->private_data = ep;
-@@ -1569,7 +1567,7 @@ static int eventpollfs_delete_dentry(str
- }
-
-
--static struct inode *ep_eventpoll_inode(void)
-+static struct inode *ep_eventpoll_inode(const struct file_operations *fops)
- {
- int error = -ENOMEM;
- struct inode *inode = new_inode(eventpoll_mnt->mnt_sb);
-@@ -1577,7 +1575,7 @@ static struct inode *ep_eventpoll_inode(
- if (!inode)
- goto eexit_1;
-
-- inode->i_fop = &eventpoll_fops;
-+ inode->i_fop = fops;
-
- /*
- * Mark the inode dirty from the very beginning,
-diff -Nurp pristine-linux-2.6.18.2/fs/Kconfig tmp-linux-2.6-xen.patch/fs/Kconfig
---- pristine-linux-2.6.18.2/fs/Kconfig 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/fs/Kconfig 2007-07-30 16:35:12.000000000 +0200
-@@ -865,6 +865,7 @@ config TMPFS
- config HUGETLBFS
- bool "HugeTLB file system support"
- depends X86 || IA64 || PPC64 || SPARC64 || SUPERH || BROKEN
-+ depends !XEN
- help
- hugetlbfs is a filesystem backing for HugeTLB pages, based on
- ramfs. For architectures that support it, say Y here and read
-diff -Nurp pristine-linux-2.6.18.2/fs/proc/proc_misc.c tmp-linux-2.6-xen.patch/fs/proc/proc_misc.c
---- pristine-linux-2.6.18.2/fs/proc/proc_misc.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/fs/proc/proc_misc.c 2007-10-14 01:51:15.000000000 +0200
-@@ -471,7 +471,7 @@ static int show_stat(struct seq_file *p,
- (unsigned long long)cputime64_to_clock_t(irq),
- (unsigned long long)cputime64_to_clock_t(softirq),
- (unsigned long long)cputime64_to_clock_t(steal));
-- for_each_online_cpu(i) {
-+ for_each_possible_cpu(i) {
-
- /* Copy values here to work around gcc-2.95.3, gcc-2.96 */
- user = kstat_cpu(i).cpustat.user;
-diff -Nurp pristine-linux-2.6.18.2/fs/proc/vmcore.c tmp-linux-2.6-xen.patch/fs/proc/vmcore.c
---- pristine-linux-2.6.18.2/fs/proc/vmcore.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/fs/proc/vmcore.c 2007-10-14 01:51:15.000000000 +0200
-@@ -514,7 +514,7 @@ static int __init parse_crash_elf64_head
- /* Do some basic Verification. */
- if (memcmp(ehdr.e_ident, ELFMAG, SELFMAG) != 0 ||
- (ehdr.e_type != ET_CORE) ||
-- !elf_check_arch(&ehdr) ||
-+ !vmcore_elf_check_arch(&ehdr) ||
- ehdr.e_ident[EI_CLASS] != ELFCLASS64 ||
- ehdr.e_ident[EI_VERSION] != EV_CURRENT ||
- ehdr.e_version != EV_CURRENT ||
-diff -Nurp pristine-linux-2.6.18.2/include/asm-generic/pci.h tmp-linux-2.6-xen.patch/include/asm-generic/pci.h
---- pristine-linux-2.6.18.2/include/asm-generic/pci.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-generic/pci.h 2007-09-30 18:06:18.000000000 +0200
-@@ -43,7 +43,9 @@ pcibios_select_root(struct pci_dev *pdev
- return root;
- }
-
-+#ifndef pcibios_scan_all_fns
- #define pcibios_scan_all_fns(a, b) 0
++/* Just any arbitrary offset to the start of the vmalloc VM area: the
++ * current 8MB value just means that there will be a 8MB "hole" after the
++ * physical memory until the kernel virtual memory starts. That means that
++ * any out-of-bounds memory accesses will hopefully be caught.
++ * The vmalloc() routines leaves a hole of 4kB between each vmalloced
++ * area for the same reason. ;)
++ */
++#define VMALLOC_OFFSET (8*1024*1024)
++#define VMALLOC_START (((unsigned long) high_memory + vmalloc_earlyreserve + \
++ 2*VMALLOC_OFFSET-1) & ~(VMALLOC_OFFSET-1))
++#ifdef CONFIG_HIGHMEM
++# define VMALLOC_END (PKMAP_BASE-2*PAGE_SIZE)
++#else
++# define VMALLOC_END (FIXADDR_START-2*PAGE_SIZE)
+#endif
-
- #ifndef HAVE_ARCH_PCI_GET_LEGACY_IDE_IRQ
- static inline int pci_get_legacy_ide_irq(struct pci_dev *dev, int channel)
-diff -Nurp pristine-linux-2.6.18.2/include/asm-generic/vmlinux.lds.h tmp-linux-2.6-xen.patch/include/asm-generic/vmlinux.lds.h
---- pristine-linux-2.6.18.2/include/asm-generic/vmlinux.lds.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-generic/vmlinux.lds.h 2007-10-14 01:51:15.000000000 +0200
-@@ -194,3 +194,6 @@
- .stab.index 0 : { *(.stab.index) } \
- .stab.indexstr 0 : { *(.stab.indexstr) } \
- .comment 0 : { *(.comment) }
+
-+#define NOTES \
-+ .notes : { *(.note.*) } :note
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/apic.h tmp-linux-2.6-xen.patch/include/asm-i386/apic.h
---- pristine-linux-2.6.18.2/include/asm-i386/apic.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/apic.h 2007-07-30 16:35:12.000000000 +0200
-@@ -119,10 +119,12 @@ extern void enable_NMI_through_LVT0 (voi
-
- extern int disable_timer_pin_1;
-
-+#ifndef CONFIG_XEN
- void smp_send_timer_broadcast_ipi(struct pt_regs *regs);
- void switch_APIC_timer_to_ipi(void *cpumask);
- void switch_ipi_to_APIC_timer(void *cpumask);
- #define ARCH_APICTIMER_STOPS_ON_C3 1
-+#endif
-
- extern int timer_over_8254;
-
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/fixmap.h tmp-linux-2.6-xen.patch/include/asm-i386/fixmap.h
---- pristine-linux-2.6.18.2/include/asm-i386/fixmap.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/fixmap.h 2007-10-14 01:51:15.000000000 +0200
-@@ -19,7 +19,7 @@
- * Leave one empty page between vmalloc'ed areas and
- * the start of the fixmap.
- */
--#define __FIXADDR_TOP 0xfffff000
-+extern unsigned long __FIXADDR_TOP;
-
- #ifndef __ASSEMBLY__
- #include <linux/kernel.h>
-@@ -94,6 +94,8 @@ enum fixed_addresses {
- extern void __set_fixmap (enum fixed_addresses idx,
- unsigned long phys, pgprot_t flags);
-
-+extern void set_fixaddr_top(unsigned long top);
++/*
++ * _PAGE_PSE set in the page directory entry just means that
++ * the page directory entry points directly to a 4MB-aligned block of
++ * memory.
++ */
++#define _PAGE_BIT_PRESENT 0
++#define _PAGE_BIT_RW 1
++#define _PAGE_BIT_USER 2
++#define _PAGE_BIT_PWT 3
++#define _PAGE_BIT_PCD 4
++#define _PAGE_BIT_ACCESSED 5
++#define _PAGE_BIT_DIRTY 6
++#define _PAGE_BIT_PSE 7 /* 4 MB (or 2MB) page, Pentium+, if present.. */
++#define _PAGE_BIT_GLOBAL 8 /* Global TLB entry PPro+ */
++/*#define _PAGE_BIT_UNUSED1 9*/ /* available for programmer */
++#define _PAGE_BIT_UNUSED2 10
++#define _PAGE_BIT_UNUSED3 11
++#define _PAGE_BIT_NX 63
+
- #define set_fixmap(idx, phys) \
- __set_fixmap(idx, phys, PAGE_KERNEL)
- /*
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/kexec.h tmp-linux-2.6-xen.patch/include/asm-i386/kexec.h
---- pristine-linux-2.6.18.2/include/asm-i386/kexec.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/kexec.h 2007-07-30 16:35:12.000000000 +0200
-@@ -1,6 +1,26 @@
- #ifndef _I386_KEXEC_H
- #define _I386_KEXEC_H
-
-+#define PA_CONTROL_PAGE 0
-+#define VA_CONTROL_PAGE 1
-+#define PA_PGD 2
-+#define VA_PGD 3
-+#define PA_PTE_0 4
-+#define VA_PTE_0 5
-+#define PA_PTE_1 6
-+#define VA_PTE_1 7
++#define _PAGE_PRESENT 0x001
++#define _PAGE_RW 0x002
++#define _PAGE_USER 0x004
++#define _PAGE_PWT 0x008
++#define _PAGE_PCD 0x010
++#define _PAGE_ACCESSED 0x020
++#define _PAGE_DIRTY 0x040
++#define _PAGE_PSE 0x080 /* 4 MB (or 2MB) page, Pentium+, if present.. */
++#define _PAGE_GLOBAL 0x100 /* Global TLB entry PPro+ */
++/*#define _PAGE_UNUSED1 0x200*/ /* available for programmer */
++#define _PAGE_UNUSED2 0x400
++#define _PAGE_UNUSED3 0x800
++
++/* If _PAGE_PRESENT is clear, we use these: */
++#define _PAGE_FILE 0x040 /* nonlinear file mapping, saved PTE; unset:swap */
++#define _PAGE_PROTNONE 0x080 /* if the user mapped it with PROT_NONE;
++ pte_present gives true */
+#ifdef CONFIG_X86_PAE
-+#define PA_PMD_0 8
-+#define VA_PMD_0 9
-+#define PA_PMD_1 10
-+#define VA_PMD_1 11
-+#define PAGES_NR 12
++#define _PAGE_NX (1ULL<<_PAGE_BIT_NX)
+#else
-+#define PAGES_NR 8
++#define _PAGE_NX 0
+#endif
+
-+#ifndef __ASSEMBLY__
-+
- #include <asm/fixmap.h>
- #include <asm/ptrace.h>
- #include <asm/string.h>
-@@ -27,6 +47,9 @@
- /* The native architecture */
- #define KEXEC_ARCH KEXEC_ARCH_386
-
-+/* We can also handle crash dumps from 64 bit kernel. */
-+#define vmcore_elf_check_arch_cross(x) ((x)->e_machine == EM_X86_64)
++/* Mapped page is I/O or foreign and has no associated page struct. */
++#define _PAGE_IO 0x200
+
- #define MAX_NOTE_BYTES 1024
-
- /* CPU does not save ss and esp on stack if execution is already
-@@ -72,5 +95,26 @@ static inline void crash_setup_regs(stru
- newregs->eip = (unsigned long)current_text_addr();
- }
- }
-+asmlinkage NORET_TYPE void
-+relocate_kernel(unsigned long indirection_page,
-+ unsigned long control_page,
-+ unsigned long start_address,
-+ unsigned int has_pae) ATTRIB_NORET;
++#define _PAGE_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED | _PAGE_DIRTY)
++#define _KERNPG_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY)
++#define _PAGE_CHG_MASK (PTE_MASK | _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_IO)
+
++#define PAGE_NONE \
++ __pgprot(_PAGE_PROTNONE | _PAGE_ACCESSED)
++#define PAGE_SHARED \
++ __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED)
+
-+/* Under Xen we need to work with machine addresses. These macros give the
-+ * machine address of a certain page to the generic kexec code instead of
-+ * the pseudo physical address which would be given by the default macros.
-+ */
++#define PAGE_SHARED_EXEC \
++ __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED)
++#define PAGE_COPY_NOEXEC \
++ __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX)
++#define PAGE_COPY_EXEC \
++ __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED)
++#define PAGE_COPY \
++ PAGE_COPY_NOEXEC
++#define PAGE_READONLY \
++ __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX)
++#define PAGE_READONLY_EXEC \
++ __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED)
+
-+#ifdef CONFIG_XEN
-+#define KEXEC_ARCH_HAS_PAGE_MACROS
-+#define kexec_page_to_pfn(page) pfn_to_mfn(page_to_pfn(page))
-+#define kexec_pfn_to_page(pfn) pfn_to_page(mfn_to_pfn(pfn))
-+#define kexec_virt_to_phys(addr) virt_to_machine(addr)
-+#define kexec_phys_to_virt(addr) phys_to_virt(machine_to_phys(addr))
-+#endif
++#define _PAGE_KERNEL \
++ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_NX)
++#define _PAGE_KERNEL_EXEC \
++ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED)
+
-+#endif /* __ASSEMBLY__ */
-
- #endif /* _I386_KEXEC_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-default/mach_traps.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-default/mach_traps.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-default/mach_traps.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-default/mach_traps.h 2007-07-30 16:35:12.000000000 +0200
-@@ -15,6 +15,18 @@ static inline void clear_mem_error(unsig
- outb(reason, 0x61);
- }
-
-+static inline void clear_io_check_error(unsigned char reason)
-+{
-+ unsigned long i;
++extern unsigned long long __PAGE_KERNEL, __PAGE_KERNEL_EXEC;
++#define __PAGE_KERNEL_RO (__PAGE_KERNEL & ~_PAGE_RW)
++#define __PAGE_KERNEL_NOCACHE (__PAGE_KERNEL | _PAGE_PCD)
++#define __PAGE_KERNEL_LARGE (__PAGE_KERNEL | _PAGE_PSE)
++#define __PAGE_KERNEL_LARGE_EXEC (__PAGE_KERNEL_EXEC | _PAGE_PSE)
+
-+ reason = (reason & 0xf) | 8;
-+ outb(reason, 0x61);
-+ i = 2000;
-+ while (--i) udelay(1000);
-+ reason &= ~8;
-+ outb(reason, 0x61);
-+}
++#define PAGE_KERNEL __pgprot(__PAGE_KERNEL)
++#define PAGE_KERNEL_RO __pgprot(__PAGE_KERNEL_RO)
++#define PAGE_KERNEL_EXEC __pgprot(__PAGE_KERNEL_EXEC)
++#define PAGE_KERNEL_NOCACHE __pgprot(__PAGE_KERNEL_NOCACHE)
++#define PAGE_KERNEL_LARGE __pgprot(__PAGE_KERNEL_LARGE)
++#define PAGE_KERNEL_LARGE_EXEC __pgprot(__PAGE_KERNEL_LARGE_EXEC)
+
- static inline unsigned char get_nmi_reason(void)
- {
- return inb(0x61);
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/agp.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/agp.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/agp.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/agp.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,37 @@
-+#ifndef AGP_H
-+#define AGP_H 1
++/*
++ * The i386 can't do page protection for execute, and considers that
++ * the same are read. Also, write permissions imply read permissions.
++ * This is the closest we can get..
++ */
++#define __P000 PAGE_NONE
++#define __P001 PAGE_READONLY
++#define __P010 PAGE_COPY
++#define __P011 PAGE_COPY
++#define __P100 PAGE_READONLY_EXEC
++#define __P101 PAGE_READONLY_EXEC
++#define __P110 PAGE_COPY_EXEC
++#define __P111 PAGE_COPY_EXEC
+
-+#include <asm/pgtable.h>
-+#include <asm/cacheflush.h>
-+#include <asm/system.h>
++#define __S000 PAGE_NONE
++#define __S001 PAGE_READONLY
++#define __S010 PAGE_SHARED
++#define __S011 PAGE_SHARED
++#define __S100 PAGE_READONLY_EXEC
++#define __S101 PAGE_READONLY_EXEC
++#define __S110 PAGE_SHARED_EXEC
++#define __S111 PAGE_SHARED_EXEC
+
-+/*
-+ * Functions to keep the agpgart mappings coherent with the MMU.
-+ * The GART gives the CPU a physical alias of pages in memory. The alias region is
-+ * mapped uncacheable. Make sure there are no conflicting mappings
-+ * with different cachability attributes for the same page. This avoids
-+ * data corruption on some CPUs.
++/*
++ * Define this if things work differently on an i386 and an i486:
++ * it will (on an i486) warn about kernel memory accesses that are
++ * done without a 'access_ok(VERIFY_WRITE,..)'
+ */
++#undef TEST_ACCESS_OK
+
-+int map_page_into_agp(struct page *page);
-+int unmap_page_from_agp(struct page *page);
-+#define flush_agp_mappings() global_flush_tlb()
++/* The boot page tables (all created as a single array) */
++extern unsigned long pg0[];
+
-+/* Could use CLFLUSH here if the cpu supports it. But then it would
-+ need to be called for each cacheline of the whole page so it may not be
-+ worth it. Would need a page for it. */
-+#define flush_agp_cache() wbinvd()
++#define pte_present(x) ((x).pte_low & (_PAGE_PRESENT | _PAGE_PROTNONE))
+
-+/* Convert a physical address to an address suitable for the GART. */
-+#define phys_to_gart(x) phys_to_machine(x)
-+#define gart_to_phys(x) machine_to_phys(x)
++/* To avoid harmful races, pmd_none(x) should check only the lower when PAE */
++#define pmd_none(x) (!(unsigned long)__pmd_val(x))
++#if CONFIG_XEN_COMPAT <= 0x030002
++/* pmd_present doesn't just test the _PAGE_PRESENT bit since wr.p.t.
++ can temporarily clear it. */
++#define pmd_present(x) (__pmd_val(x))
++#define pmd_bad(x) ((__pmd_val(x) & (~PAGE_MASK & ~_PAGE_USER & ~_PAGE_PRESENT)) != (_KERNPG_TABLE & ~_PAGE_PRESENT))
++#else
++#define pmd_present(x) (__pmd_val(x) & _PAGE_PRESENT)
++#define pmd_bad(x) ((__pmd_val(x) & (~PAGE_MASK & ~_PAGE_USER)) != _KERNPG_TABLE)
++#endif
+
-+/* GATT allocation. Returns/accepts GATT kernel virtual address. */
-+#define alloc_gatt_pages(order) ({ \
-+ char *_t; dma_addr_t _d; \
-+ _t = dma_alloc_coherent(NULL,PAGE_SIZE<<(order),&_d,GFP_KERNEL); \
-+ _t; })
-+#define free_gatt_pages(table, order) \
-+ dma_free_coherent(NULL,PAGE_SIZE<<(order),(table),virt_to_bus(table))
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/desc.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/desc.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/desc.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/desc.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,164 @@
-+#ifndef __ARCH_DESC_H
-+#define __ARCH_DESC_H
++#define pages_to_mb(x) ((x) >> (20-PAGE_SHIFT))
+
-+#include <asm/ldt.h>
-+#include <asm/segment.h>
++/*
++ * The following only work if pte_present() is true.
++ * Undefined behaviour if not..
++ */
++static inline int pte_user(pte_t pte) { return (pte).pte_low & _PAGE_USER; }
++static inline int pte_read(pte_t pte) { return (pte).pte_low & _PAGE_USER; }
++static inline int pte_dirty(pte_t pte) { return (pte).pte_low & _PAGE_DIRTY; }
++static inline int pte_young(pte_t pte) { return (pte).pte_low & _PAGE_ACCESSED; }
++static inline int pte_write(pte_t pte) { return (pte).pte_low & _PAGE_RW; }
++static inline int pte_huge(pte_t pte) { return (pte).pte_low & _PAGE_PSE; }
+
-+#define CPU_16BIT_STACK_SIZE 1024
++/*
++ * The following only works if pte_present() is not true.
++ */
++static inline int pte_file(pte_t pte) { return (pte).pte_low & _PAGE_FILE; }
+
-+#ifndef __ASSEMBLY__
++static inline pte_t pte_rdprotect(pte_t pte) { (pte).pte_low &= ~_PAGE_USER; return pte; }
++static inline pte_t pte_exprotect(pte_t pte) { (pte).pte_low &= ~_PAGE_USER; return pte; }
++static inline pte_t pte_mkclean(pte_t pte) { (pte).pte_low &= ~_PAGE_DIRTY; return pte; }
++static inline pte_t pte_mkold(pte_t pte) { (pte).pte_low &= ~_PAGE_ACCESSED; return pte; }
++static inline pte_t pte_wrprotect(pte_t pte) { (pte).pte_low &= ~_PAGE_RW; return pte; }
++static inline pte_t pte_mkread(pte_t pte) { (pte).pte_low |= _PAGE_USER; return pte; }
++static inline pte_t pte_mkexec(pte_t pte) { (pte).pte_low |= _PAGE_USER; return pte; }
++static inline pte_t pte_mkdirty(pte_t pte) { (pte).pte_low |= _PAGE_DIRTY; return pte; }
++static inline pte_t pte_mkyoung(pte_t pte) { (pte).pte_low |= _PAGE_ACCESSED; return pte; }
++static inline pte_t pte_mkwrite(pte_t pte) { (pte).pte_low |= _PAGE_RW; return pte; }
++static inline pte_t pte_mkhuge(pte_t pte) { (pte).pte_low |= _PAGE_PSE; return pte; }
+
-+#include <linux/preempt.h>
-+#include <linux/smp.h>
++#ifdef CONFIG_X86_PAE
++# include <asm/pgtable-3level.h>
++#else
++# include <asm/pgtable-2level.h>
++#endif
+
-+#include <asm/mmu.h>
++#define ptep_test_and_clear_dirty(vma, addr, ptep) \
++({ \
++ pte_t __pte = *(ptep); \
++ int __ret = pte_dirty(__pte); \
++ if (__ret) { \
++ __pte = pte_mkclean(__pte); \
++ if ((vma)->vm_mm != current->mm || \
++ HYPERVISOR_update_va_mapping(addr, __pte, 0)) \
++ (ptep)->pte_low = __pte.pte_low; \
++ } \
++ __ret; \
++})
+
-+extern struct desc_struct cpu_gdt_table[GDT_ENTRIES];
++#define ptep_test_and_clear_young(vma, addr, ptep) \
++({ \
++ pte_t __pte = *(ptep); \
++ int __ret = pte_young(__pte); \
++ if (__ret) \
++ __pte = pte_mkold(__pte); \
++ if ((vma)->vm_mm != current->mm || \
++ HYPERVISOR_update_va_mapping(addr, __pte, 0)) \
++ (ptep)->pte_low = __pte.pte_low; \
++ __ret; \
++})
+
-+DECLARE_PER_CPU(unsigned char, cpu_16bit_stack[CPU_16BIT_STACK_SIZE]);
++#define ptep_get_and_clear_full(mm, addr, ptep, full) \
++ ((full) ? ({ \
++ pte_t __res = *(ptep); \
++ if (test_bit(PG_pinned, &virt_to_page((mm)->pgd)->flags)) \
++ xen_l1_entry_update(ptep, __pte(0)); \
++ else \
++ *(ptep) = __pte(0); \
++ __res; \
++ }) : \
++ ptep_get_and_clear(mm, addr, ptep))
+
-+struct Xgt_desc_struct {
-+ unsigned short size;
-+ unsigned long address __attribute__((packed));
-+ unsigned short pad;
-+} __attribute__ ((packed));
++static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
++{
++ pte_t pte = *ptep;
++ if (pte_write(pte))
++ set_pte_at(mm, addr, ptep, pte_wrprotect(pte));
++}
+
-+extern struct Xgt_desc_struct idt_descr;
-+DECLARE_PER_CPU(struct Xgt_desc_struct, cpu_gdt_descr);
++/*
++ * clone_pgd_range(pgd_t *dst, pgd_t *src, int count);
++ *
++ * dst - pointer to pgd range anwhere on a pgd page
++ * src - ""
++ * count - the number of pgds to copy.
++ *
++ * dst and src can be on the same page, but the range must not overlap,
++ * and must not cross a page boundary.
++ */
++static inline void clone_pgd_range(pgd_t *dst, pgd_t *src, int count)
++{
++ memcpy(dst, src, count * sizeof(pgd_t));
++}
++
++/*
++ * Macro to mark a page protection value as "uncacheable". On processors which do not support
++ * it, this is a no-op.
++ */
++#define pgprot_noncached(prot) ((boot_cpu_data.x86 > 3) \
++ ? (__pgprot(pgprot_val(prot) | _PAGE_PCD | _PAGE_PWT)) : (prot))
+
++/*
++ * Conversion functions: convert a page and protection to a page entry,
++ * and a page entry and page directory to the page they refer to.
++ */
+
-+static inline struct desc_struct *get_cpu_gdt_table(unsigned int cpu)
++#define mk_pte(page, pgprot) pfn_pte(page_to_pfn(page), (pgprot))
++
++static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
+{
-+ return (struct desc_struct *)per_cpu(cpu_gdt_descr, cpu).address;
++ /*
++ * Since this might change the present bit (which controls whether
++ * a pte_t object has undergone p2m translation), we must use
++ * pte_val() on the input pte and __pte() for the return value.
++ */
++ paddr_t pteval = pte_val(pte);
++
++ pteval &= _PAGE_CHG_MASK;
++ pteval |= pgprot_val(newprot);
++#ifdef CONFIG_X86_PAE
++ pteval &= __supported_pte_mask;
++#endif
++ return __pte(pteval);
+}
+
-+#define load_TR_desc() __asm__ __volatile__("ltr %w0"::"q" (GDT_ENTRY_TSS*8))
-+#define load_LDT_desc() __asm__ __volatile__("lldt %w0"::"q" (GDT_ENTRY_LDT*8))
++#define pmd_large(pmd) \
++((__pmd_val(pmd) & (_PAGE_PSE|_PAGE_PRESENT)) == (_PAGE_PSE|_PAGE_PRESENT))
+
-+#define load_gdt(dtr) __asm__ __volatile("lgdt %0"::"m" (*dtr))
-+#define load_idt(dtr) __asm__ __volatile("lidt %0"::"m" (*dtr))
-+#define load_tr(tr) __asm__ __volatile("ltr %0"::"mr" (tr))
-+#define load_ldt(ldt) __asm__ __volatile("lldt %0"::"mr" (ldt))
++/*
++ * the pgd page can be thought of an array like this: pgd_t[PTRS_PER_PGD]
++ *
++ * this macro returns the index of the entry in the pgd page which would
++ * control the given virtual address
++ */
++#define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD-1))
++#define pgd_index_k(addr) pgd_index(addr)
+
-+#define store_gdt(dtr) __asm__ ("sgdt %0":"=m" (*dtr))
-+#define store_idt(dtr) __asm__ ("sidt %0":"=m" (*dtr))
-+#define store_tr(tr) __asm__ ("str %0":"=mr" (tr))
-+#define store_ldt(ldt) __asm__ ("sldt %0":"=mr" (ldt))
++/*
++ * pgd_offset() returns a (pgd_t *)
++ * pgd_index() is used get the offset into the pgd page's array of pgd_t's;
++ */
++#define pgd_offset(mm, address) ((mm)->pgd+pgd_index(address))
+
+/*
-+ * This is the ldt that every process will get unless we need
-+ * something other than this.
++ * a shortcut which implies the use of the kernel's pgd, instead
++ * of a process's
+ */
-+extern struct desc_struct default_ldt[];
-+extern void set_intr_gate(unsigned int irq, void * addr);
++#define pgd_offset_k(address) pgd_offset(&init_mm, address)
+
-+#define _set_tssldt_desc(n,addr,limit,type) \
-+__asm__ __volatile__ ("movw %w3,0(%2)\n\t" \
-+ "movw %w1,2(%2)\n\t" \
-+ "rorl $16,%1\n\t" \
-+ "movb %b1,4(%2)\n\t" \
-+ "movb %4,5(%2)\n\t" \
-+ "movb $0,6(%2)\n\t" \
-+ "movb %h1,7(%2)\n\t" \
-+ "rorl $16,%1" \
-+ : "=m"(*(n)) : "q" (addr), "r"(n), "ir"(limit), "i"(type))
++/*
++ * the pmd page can be thought of an array like this: pmd_t[PTRS_PER_PMD]
++ *
++ * this macro returns the index of the entry in the pmd page which would
++ * control the given virtual address
++ */
++#define pmd_index(address) \
++ (((address) >> PMD_SHIFT) & (PTRS_PER_PMD-1))
+
-+#ifndef CONFIG_X86_NO_TSS
-+static inline void __set_tss_desc(unsigned int cpu, unsigned int entry, void *addr)
-+{
-+ _set_tssldt_desc(&get_cpu_gdt_table(cpu)[entry], (int)addr,
-+ offsetof(struct tss_struct, __cacheline_filler) - 1, 0x89);
-+}
++/*
++ * the pte page can be thought of an array like this: pte_t[PTRS_PER_PTE]
++ *
++ * this macro returns the index of the entry in the pte page which would
++ * control the given virtual address
++ */
++#define pte_index(address) \
++ (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1))
++#define pte_offset_kernel(dir, address) \
++ ((pte_t *) pmd_page_kernel(*(dir)) + pte_index(address))
+
-+#define set_tss_desc(cpu,addr) __set_tss_desc(cpu, GDT_ENTRY_TSS, addr)
++#define pmd_page(pmd) (pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT))
++
++#define pmd_page_kernel(pmd) \
++ ((unsigned long) __va(pmd_val(pmd) & PAGE_MASK))
++
++/*
++ * Helper function that returns the kernel pagetable entry controlling
++ * the virtual address 'address'. NULL means no pagetable entry present.
++ * NOTE: the return type is pte_t but if the pmd is PSE then we return it
++ * as a pte too.
++ */
++extern pte_t *lookup_address(unsigned long address);
++
++/*
++ * Make a given kernel text page executable/non-executable.
++ * Returns the previous executability setting of that page (which
++ * is used to restore the previous state). Used by the SMP bootup code.
++ * NOTE: this is an __init function for security reasons.
++ */
++#ifdef CONFIG_X86_PAE
++ extern int set_kernel_exec(unsigned long vaddr, int enable);
++#else
++ static inline int set_kernel_exec(unsigned long vaddr, int enable) { return 0;}
+#endif
+
-+static inline void set_ldt_desc(unsigned int cpu, void *addr, unsigned int size)
-+{
-+ _set_tssldt_desc(&get_cpu_gdt_table(cpu)[GDT_ENTRY_LDT], (int)addr, ((size << 3)-1), 0x82);
-+}
++extern void noexec_setup(const char *str);
+
-+#define LDT_entry_a(info) \
-+ ((((info)->base_addr & 0x0000ffff) << 16) | ((info)->limit & 0x0ffff))
++#if defined(CONFIG_HIGHPTE)
++#define pte_offset_map(dir, address) \
++ ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)),KM_PTE0) + \
++ pte_index(address))
++#define pte_offset_map_nested(dir, address) \
++ ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)),KM_PTE1) + \
++ pte_index(address))
++#define pte_unmap(pte) kunmap_atomic(pte, KM_PTE0)
++#define pte_unmap_nested(pte) kunmap_atomic(pte, KM_PTE1)
++#else
++#define pte_offset_map(dir, address) \
++ ((pte_t *)page_address(pmd_page(*(dir))) + pte_index(address))
++#define pte_offset_map_nested(dir, address) pte_offset_map(dir, address)
++#define pte_unmap(pte) do { } while (0)
++#define pte_unmap_nested(pte) do { } while (0)
++#endif
+
-+#define LDT_entry_b(info) \
-+ (((info)->base_addr & 0xff000000) | \
-+ (((info)->base_addr & 0x00ff0000) >> 16) | \
-+ ((info)->limit & 0xf0000) | \
-+ (((info)->read_exec_only ^ 1) << 9) | \
-+ ((info)->contents << 10) | \
-+ (((info)->seg_not_present ^ 1) << 15) | \
-+ ((info)->seg_32bit << 22) | \
-+ ((info)->limit_in_pages << 23) | \
-+ ((info)->useable << 20) | \
-+ 0x7000)
++#define __HAVE_ARCH_PTEP_ESTABLISH
++#define ptep_establish(vma, address, ptep, pteval) \
++ do { \
++ if ( likely((vma)->vm_mm == current->mm) ) { \
++ BUG_ON(HYPERVISOR_update_va_mapping(address, \
++ pteval, \
++ (unsigned long)(vma)->vm_mm->cpu_vm_mask.bits| \
++ UVMF_INVLPG|UVMF_MULTI)); \
++ } else { \
++ xen_l1_entry_update(ptep, pteval); \
++ flush_tlb_page(vma, address); \
++ } \
++ } while (0)
+
-+#define LDT_empty(info) (\
-+ (info)->base_addr == 0 && \
-+ (info)->limit == 0 && \
-+ (info)->contents == 0 && \
-+ (info)->read_exec_only == 1 && \
-+ (info)->seg_32bit == 0 && \
-+ (info)->limit_in_pages == 0 && \
-+ (info)->seg_not_present == 1 && \
-+ (info)->useable == 0 )
++/*
++ * The i386 doesn't have any external MMU info: the kernel page
++ * tables contain all the necessary information.
++ *
++ * Also, we only update the dirty/accessed state if we set
++ * the dirty bit by hand in the kernel, since the hardware
++ * will do the accessed bit for us, and we don't want to
++ * race with other CPU's that might be updating the dirty
++ * bit at the same time.
++ */
++#define update_mmu_cache(vma,address,pte) do { } while (0)
++#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
++#define ptep_set_access_flags(vma, address, ptep, entry, dirty) \
++ do { \
++ if (dirty) \
++ ptep_establish(vma, address, ptep, entry); \
++ } while (0)
+
-+extern int write_ldt_entry(void *ldt, int entry, __u32 entry_a, __u32 entry_b);
++#include <xen/features.h>
++void make_lowmem_page_readonly(void *va, unsigned int feature);
++void make_lowmem_page_writable(void *va, unsigned int feature);
++void make_page_readonly(void *va, unsigned int feature);
++void make_page_writable(void *va, unsigned int feature);
++void make_pages_readonly(void *va, unsigned int nr, unsigned int feature);
++void make_pages_writable(void *va, unsigned int nr, unsigned int feature);
+
-+#if TLS_SIZE != 24
-+# error update this code.
-+#endif
++#define virt_to_ptep(va) \
++({ \
++ pte_t *__ptep = lookup_address((unsigned long)(va)); \
++ BUG_ON(!__ptep || !pte_present(*__ptep)); \
++ __ptep; \
++})
+
-+static inline void load_TLS(struct thread_struct *t, unsigned int cpu)
-+{
-+#define C(i) HYPERVISOR_update_descriptor(virt_to_machine(&get_cpu_gdt_table(cpu)[GDT_ENTRY_TLS_MIN + i]), *(u64 *)&t->tls_array[i])
-+ C(0); C(1); C(2);
-+#undef C
-+}
++#define arbitrary_virt_to_machine(va) \
++ (((maddr_t)pte_mfn(*virt_to_ptep(va)) << PAGE_SHIFT) \
++ | ((unsigned long)(va) & (PAGE_SIZE - 1)))
++
++#endif /* !__ASSEMBLY__ */
++
++#ifdef CONFIG_FLATMEM
++#define kern_addr_valid(addr) (1)
++#endif /* CONFIG_FLATMEM */
++
++int direct_remap_pfn_range(struct vm_area_struct *vma,
++ unsigned long address,
++ unsigned long mfn,
++ unsigned long size,
++ pgprot_t prot,
++ domid_t domid);
++int direct_kernel_remap_pfn_range(unsigned long address,
++ unsigned long mfn,
++ unsigned long size,
++ pgprot_t prot,
++ domid_t domid);
++int create_lookup_pte_addr(struct mm_struct *mm,
++ unsigned long address,
++ uint64_t *ptep);
++int touch_pte_range(struct mm_struct *mm,
++ unsigned long address,
++ unsigned long size);
++
++int xen_change_pte_range(struct mm_struct *mm, pmd_t *pmd,
++ unsigned long addr, unsigned long end, pgprot_t newprot);
++
++#define arch_change_pte_range(mm, pmd, addr, end, newprot) \
++ xen_change_pte_range(mm, pmd, addr, end, newprot)
++
++#define io_remap_pfn_range(vma,from,pfn,size,prot) \
++direct_remap_pfn_range(vma,from,pfn,size,prot,DOMID_IO)
+
-+static inline void clear_LDT(void)
-+{
-+ int cpu = get_cpu();
++#define MK_IOSPACE_PFN(space, pfn) (pfn)
++#define GET_IOSPACE(pfn) 0
++#define GET_PFN(pfn) (pfn)
+
-+ /*
-+ * NB. We load the default_ldt for lcall7/27 handling on demand, as
-+ * it slows down context switching. Noone uses it anyway.
-+ */
-+ cpu = cpu; /* XXX avoid compiler warning */
-+ xen_set_ldt(0UL, 0);
-+ put_cpu();
-+}
++#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
++#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_DIRTY
++#define __HAVE_ARCH_PTEP_GET_AND_CLEAR
++#define __HAVE_ARCH_PTEP_GET_AND_CLEAR_FULL
++#define __HAVE_ARCH_PTEP_CLEAR_FLUSH
++#define __HAVE_ARCH_PTEP_SET_WRPROTECT
++#define __HAVE_ARCH_PTE_SAME
++#include <asm-generic/pgtable.h>
+
++#endif /* _I386_PGTABLE_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/processor.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/processor.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,743 @@
+/*
-+ * load one particular LDT into the current CPU
++ * include/asm-i386/processor.h
++ *
++ * Copyright (C) 1994 Linus Torvalds
+ */
-+static inline void load_LDT_nolock(mm_context_t *pc, int cpu)
-+{
-+ void *segments = pc->ldt;
-+ int count = pc->size;
-+
-+ if (likely(!count))
-+ segments = NULL;
+
-+ xen_set_ldt((unsigned long)segments, count);
-+}
++#ifndef __ASM_I386_PROCESSOR_H
++#define __ASM_I386_PROCESSOR_H
+
-+static inline void load_LDT(mm_context_t *pc)
-+{
-+ int cpu = get_cpu();
-+ load_LDT_nolock(pc, cpu);
-+ put_cpu();
-+}
++#include <asm/vm86.h>
++#include <asm/math_emu.h>
++#include <asm/segment.h>
++#include <asm/page.h>
++#include <asm/types.h>
++#include <asm/sigcontext.h>
++#include <asm/cpufeature.h>
++#include <asm/msr.h>
++#include <asm/system.h>
++#include <linux/cache.h>
++#include <linux/threads.h>
++#include <asm/percpu.h>
++#include <linux/cpumask.h>
++#include <xen/interface/physdev.h>
+
-+static inline unsigned long get_desc_base(unsigned long *desc)
-+{
-+ unsigned long base;
-+ base = ((desc[0] >> 16) & 0x0000ffff) |
-+ ((desc[1] << 16) & 0x00ff0000) |
-+ (desc[1] & 0xff000000);
-+ return base;
-+}
++/* flag for disabling the tsc */
++extern int tsc_disable;
+
-+#endif /* !__ASSEMBLY__ */
++struct desc_struct {
++ unsigned long a,b;
++};
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/dma-mapping.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/dma-mapping.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/dma-mapping.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/dma-mapping.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,157 @@
-+#ifndef _ASM_I386_DMA_MAPPING_H
-+#define _ASM_I386_DMA_MAPPING_H
++#define desc_empty(desc) \
++ (!((desc)->a | (desc)->b))
+
++#define desc_equal(desc1, desc2) \
++ (((desc1)->a == (desc2)->a) && ((desc1)->b == (desc2)->b))
+/*
-+ * IOMMU interface. See Documentation/DMA-mapping.txt and DMA-API.txt for
-+ * documentation.
++ * Default implementation of macro that returns current
++ * instruction pointer ("program counter").
+ */
++#define current_text_addr() ({ void *pc; __asm__("movl $1f,%0\n1:":"=g" (pc)); pc; })
+
-+#include <linux/mm.h>
-+#include <asm/cache.h>
-+#include <asm/io.h>
-+#include <asm/scatterlist.h>
-+#include <asm/swiotlb.h>
-+
-+static inline int
-+address_needs_mapping(struct device *hwdev, dma_addr_t addr)
-+{
-+ dma_addr_t mask = 0xffffffff;
-+ /* If the device has a mask, use it, otherwise default to 32 bits */
-+ if (hwdev && hwdev->dma_mask)
-+ mask = *hwdev->dma_mask;
-+ return (addr & ~mask) != 0;
-+}
-+
-+static inline int
-+range_straddles_page_boundary(paddr_t p, size_t size)
-+{
-+ extern unsigned long *contiguous_bitmap;
-+ return ((((p & ~PAGE_MASK) + size) > PAGE_SIZE) &&
-+ !test_bit(p >> PAGE_SHIFT, contiguous_bitmap));
-+}
++/*
++ * CPU type and hardware bug flags. Kept separately for each CPU.
++ * Members of this structure are referenced in head.S, so think twice
++ * before touching them. [mj]
++ */
+
-+#define dma_alloc_noncoherent(d, s, h, f) dma_alloc_coherent(d, s, h, f)
-+#define dma_free_noncoherent(d, s, v, h) dma_free_coherent(d, s, v, h)
++struct cpuinfo_x86 {
++ __u8 x86; /* CPU family */
++ __u8 x86_vendor; /* CPU vendor */
++ __u8 x86_model;
++ __u8 x86_mask;
++ char wp_works_ok; /* It doesn't on 386's */
++ char hlt_works_ok; /* Problems on some 486Dx4's and old 386's */
++ char hard_math;
++ char rfu;
++ int cpuid_level; /* Maximum supported CPUID level, -1=no CPUID */
++ unsigned long x86_capability[NCAPINTS];
++ char x86_vendor_id[16];
++ char x86_model_id[64];
++ int x86_cache_size; /* in KB - valid for CPUS which support this
++ call */
++ int x86_cache_alignment; /* In bytes */
++ char fdiv_bug;
++ char f00f_bug;
++ char coma_bug;
++ char pad0;
++ int x86_power;
++ unsigned long loops_per_jiffy;
++#ifdef CONFIG_SMP
++ cpumask_t llc_shared_map; /* cpus sharing the last level cache */
++#endif
++ unsigned char x86_max_cores; /* cpuid returned max cores value */
++ unsigned char apicid;
++#ifdef CONFIG_SMP
++ unsigned char booted_cores; /* number of cores as seen by OS */
++ __u8 phys_proc_id; /* Physical processor id. */
++ __u8 cpu_core_id; /* Core id */
++#endif
++} __attribute__((__aligned__(SMP_CACHE_BYTES)));
+
-+void *dma_alloc_coherent(struct device *dev, size_t size,
-+ dma_addr_t *dma_handle, gfp_t flag);
++#define X86_VENDOR_INTEL 0
++#define X86_VENDOR_CYRIX 1
++#define X86_VENDOR_AMD 2
++#define X86_VENDOR_UMC 3
++#define X86_VENDOR_NEXGEN 4
++#define X86_VENDOR_CENTAUR 5
++#define X86_VENDOR_RISE 6
++#define X86_VENDOR_TRANSMETA 7
++#define X86_VENDOR_NSC 8
++#define X86_VENDOR_NUM 9
++#define X86_VENDOR_UNKNOWN 0xff
+
-+void dma_free_coherent(struct device *dev, size_t size,
-+ void *vaddr, dma_addr_t dma_handle);
++/*
++ * capabilities of CPUs
++ */
+
-+extern dma_addr_t
-+dma_map_single(struct device *dev, void *ptr, size_t size,
-+ enum dma_data_direction direction);
++extern struct cpuinfo_x86 boot_cpu_data;
++extern struct cpuinfo_x86 new_cpu_data;
++#ifndef CONFIG_X86_NO_TSS
++extern struct tss_struct doublefault_tss;
++DECLARE_PER_CPU(struct tss_struct, init_tss);
++#endif
+
-+extern void
-+dma_unmap_single(struct device *dev, dma_addr_t dma_addr, size_t size,
-+ enum dma_data_direction direction);
++#ifdef CONFIG_SMP
++extern struct cpuinfo_x86 cpu_data[];
++#define current_cpu_data cpu_data[smp_processor_id()]
++#else
++#define cpu_data (&boot_cpu_data)
++#define current_cpu_data boot_cpu_data
++#endif
+
-+extern int dma_map_sg(struct device *hwdev, struct scatterlist *sg,
-+ int nents, enum dma_data_direction direction);
-+extern void dma_unmap_sg(struct device *hwdev, struct scatterlist *sg,
-+ int nents, enum dma_data_direction direction);
++extern int cpu_llc_id[NR_CPUS];
++extern char ignore_fpu_irq;
+
-+#ifdef CONFIG_HIGHMEM
-+extern dma_addr_t
-+dma_map_page(struct device *dev, struct page *page, unsigned long offset,
-+ size_t size, enum dma_data_direction direction);
++extern void identify_cpu(struct cpuinfo_x86 *);
++extern void print_cpu_info(struct cpuinfo_x86 *);
++extern unsigned int init_intel_cacheinfo(struct cpuinfo_x86 *c);
++extern unsigned short num_cache_leaves;
+
-+extern void
-+dma_unmap_page(struct device *dev, dma_addr_t dma_address, size_t size,
-+ enum dma_data_direction direction);
++#ifdef CONFIG_X86_HT
++extern void detect_ht(struct cpuinfo_x86 *c);
+#else
-+#define dma_map_page(dev, page, offset, size, dir) \
-+ dma_map_single(dev, page_address(page) + (offset), (size), (dir))
-+#define dma_unmap_page dma_unmap_single
++static inline void detect_ht(struct cpuinfo_x86 *c) {}
+#endif
+
-+extern void
-+dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle, size_t size,
-+ enum dma_data_direction direction);
-+
-+extern void
-+dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle, size_t size,
-+ enum dma_data_direction direction);
++/*
++ * EFLAGS bits
++ */
++#define X86_EFLAGS_CF 0x00000001 /* Carry Flag */
++#define X86_EFLAGS_PF 0x00000004 /* Parity Flag */
++#define X86_EFLAGS_AF 0x00000010 /* Auxillary carry Flag */
++#define X86_EFLAGS_ZF 0x00000040 /* Zero Flag */
++#define X86_EFLAGS_SF 0x00000080 /* Sign Flag */
++#define X86_EFLAGS_TF 0x00000100 /* Trap Flag */
++#define X86_EFLAGS_IF 0x00000200 /* Interrupt Flag */
++#define X86_EFLAGS_DF 0x00000400 /* Direction Flag */
++#define X86_EFLAGS_OF 0x00000800 /* Overflow Flag */
++#define X86_EFLAGS_IOPL 0x00003000 /* IOPL mask */
++#define X86_EFLAGS_NT 0x00004000 /* Nested Task */
++#define X86_EFLAGS_RF 0x00010000 /* Resume Flag */
++#define X86_EFLAGS_VM 0x00020000 /* Virtual Mode */
++#define X86_EFLAGS_AC 0x00040000 /* Alignment Check */
++#define X86_EFLAGS_VIF 0x00080000 /* Virtual Interrupt Flag */
++#define X86_EFLAGS_VIP 0x00100000 /* Virtual Interrupt Pending */
++#define X86_EFLAGS_ID 0x00200000 /* CPUID detection flag */
+
-+static inline void
-+dma_sync_single_range_for_cpu(struct device *dev, dma_addr_t dma_handle,
-+ unsigned long offset, size_t size,
-+ enum dma_data_direction direction)
++/*
++ * Generic CPUID function
++ * clear %ecx since some cpus (Cyrix MII) do not set or clear %ecx
++ * resulting in stale register contents being returned.
++ */
++static inline void cpuid(unsigned int op, unsigned int *eax, unsigned int *ebx, unsigned int *ecx, unsigned int *edx)
+{
-+ dma_sync_single_for_cpu(dev, dma_handle+offset, size, direction);
++ __asm__(XEN_CPUID
++ : "=a" (*eax),
++ "=b" (*ebx),
++ "=c" (*ecx),
++ "=d" (*edx)
++ : "0" (op), "c"(0));
+}
+
-+static inline void
-+dma_sync_single_range_for_device(struct device *dev, dma_addr_t dma_handle,
-+ unsigned long offset, size_t size,
-+ enum dma_data_direction direction)
++/* Some CPUID calls want 'count' to be placed in ecx */
++static inline void cpuid_count(int op, int count, int *eax, int *ebx, int *ecx,
++ int *edx)
+{
-+ dma_sync_single_for_device(dev, dma_handle+offset, size, direction);
++ __asm__(XEN_CPUID
++ : "=a" (*eax),
++ "=b" (*ebx),
++ "=c" (*ecx),
++ "=d" (*edx)
++ : "0" (op), "c" (count));
+}
+
-+static inline void
-+dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg, int nelems,
-+ enum dma_data_direction direction)
++/*
++ * CPUID functions returning a single datum
++ */
++static inline unsigned int cpuid_eax(unsigned int op)
+{
-+ if (swiotlb)
-+ swiotlb_sync_sg_for_cpu(dev,sg,nelems,direction);
-+ flush_write_buffers();
-+}
++ unsigned int eax;
+
-+static inline void
-+dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg, int nelems,
-+ enum dma_data_direction direction)
-+{
-+ if (swiotlb)
-+ swiotlb_sync_sg_for_device(dev,sg,nelems,direction);
-+ flush_write_buffers();
++ __asm__(XEN_CPUID
++ : "=a" (eax)
++ : "0" (op)
++ : "bx", "cx", "dx");
++ return eax;
+}
-+
-+extern int
-+dma_mapping_error(dma_addr_t dma_addr);
-+
-+extern int
-+dma_supported(struct device *dev, u64 mask);
-+
-+static inline int
-+dma_set_mask(struct device *dev, u64 mask)
++static inline unsigned int cpuid_ebx(unsigned int op)
+{
-+ if(!dev->dma_mask || !dma_supported(dev, mask))
-+ return -EIO;
-+
-+ *dev->dma_mask = mask;
++ unsigned int eax, ebx;
+
-+ return 0;
++ __asm__(XEN_CPUID
++ : "=a" (eax), "=b" (ebx)
++ : "0" (op)
++ : "cx", "dx" );
++ return ebx;
+}
-+
-+static inline int
-+dma_get_cache_alignment(void)
++static inline unsigned int cpuid_ecx(unsigned int op)
+{
-+ /* no easy way to get cache size on all x86, so return the
-+ * maximum possible, to be safe */
-+ return (1 << INTERNODE_CACHE_SHIFT);
-+}
-+
-+#define dma_is_consistent(d) (1)
++ unsigned int eax, ecx;
+
-+static inline void
-+dma_cache_sync(void *vaddr, size_t size,
-+ enum dma_data_direction direction)
-+{
-+ flush_write_buffers();
++ __asm__(XEN_CPUID
++ : "=a" (eax), "=c" (ecx)
++ : "0" (op)
++ : "bx", "dx" );
++ return ecx;
+}
++static inline unsigned int cpuid_edx(unsigned int op)
++{
++ unsigned int eax, edx;
+
-+#define ARCH_HAS_DMA_DECLARE_COHERENT_MEMORY
-+extern int
-+dma_declare_coherent_memory(struct device *dev, dma_addr_t bus_addr,
-+ dma_addr_t device_addr, size_t size, int flags);
-+
-+extern void
-+dma_release_declared_memory(struct device *dev);
-+
-+extern void *
-+dma_mark_declared_memory_occupied(struct device *dev,
-+ dma_addr_t device_addr, size_t size);
-+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/fixmap.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/fixmap.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/fixmap.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/fixmap.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,155 @@
-+/*
-+ * fixmap.h: compile-time virtual memory allocation
-+ *
-+ * This file is subject to the terms and conditions of the GNU General Public
-+ * License. See the file "COPYING" in the main directory of this archive
-+ * for more details.
-+ *
-+ * Copyright (C) 1998 Ingo Molnar
-+ *
-+ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
-+ */
-+
-+#ifndef _ASM_FIXMAP_H
-+#define _ASM_FIXMAP_H
-+
-+
-+/* used by vmalloc.c, vsyscall.lds.S.
-+ *
-+ * Leave one empty page between vmalloc'ed areas and
-+ * the start of the fixmap.
-+ */
-+extern unsigned long __FIXADDR_TOP;
++ __asm__(XEN_CPUID
++ : "=a" (eax), "=d" (edx)
++ : "0" (op)
++ : "bx", "cx");
++ return edx;
++}
+
-+#ifndef __ASSEMBLY__
-+#include <linux/kernel.h>
-+#include <asm/acpi.h>
-+#include <asm/apicdef.h>
-+#include <asm/page.h>
-+#ifdef CONFIG_HIGHMEM
-+#include <linux/threads.h>
-+#include <asm/kmap_types.h>
-+#endif
++#define load_cr3(pgdir) write_cr3(__pa(pgdir))
+
+/*
-+ * Here we define all the compile-time 'special' virtual
-+ * addresses. The point is to have a constant address at
-+ * compile time, but to set the physical address only
-+ * in the boot process. We allocate these special addresses
-+ * from the end of virtual memory (0xfffff000) backwards.
-+ * Also this lets us do fail-safe vmalloc(), we
-+ * can guarantee that these special addresses and
-+ * vmalloc()-ed addresses never overlap.
-+ *
-+ * these 'compile-time allocated' memory buffers are
-+ * fixed-size 4k pages. (or larger if used with an increment
-+ * highger than 1) use fixmap_set(idx,phys) to associate
-+ * physical memory with fixmap indices.
-+ *
-+ * TLB entries of such buffers will not be flushed across
-+ * task switches.
++ * Intel CPU features in CR4
+ */
-+enum fixed_addresses {
-+ FIX_HOLE,
-+ FIX_VDSO,
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ FIX_APIC_BASE, /* local (CPU) APIC) -- required for SMP or not */
-+#endif
-+#ifdef CONFIG_X86_IO_APIC
-+ FIX_IO_APIC_BASE_0,
-+ FIX_IO_APIC_BASE_END = FIX_IO_APIC_BASE_0 + MAX_IO_APICS-1,
-+#endif
-+#ifdef CONFIG_X86_VISWS_APIC
-+ FIX_CO_CPU, /* Cobalt timer */
-+ FIX_CO_APIC, /* Cobalt APIC Redirection Table */
-+ FIX_LI_PCIA, /* Lithium PCI Bridge A */
-+ FIX_LI_PCIB, /* Lithium PCI Bridge B */
-+#endif
-+#ifdef CONFIG_X86_F00F_BUG
-+ FIX_F00F_IDT, /* Virtual mapping for IDT */
-+#endif
-+#ifdef CONFIG_X86_CYCLONE_TIMER
-+ FIX_CYCLONE_TIMER, /*cyclone timer register*/
-+#endif
-+#ifdef CONFIG_HIGHMEM
-+ FIX_KMAP_BEGIN, /* reserved pte's for temporary kernel mappings */
-+ FIX_KMAP_END = FIX_KMAP_BEGIN+(KM_TYPE_NR*NR_CPUS)-1,
-+#endif
-+#ifdef CONFIG_ACPI
-+ FIX_ACPI_BEGIN,
-+ FIX_ACPI_END = FIX_ACPI_BEGIN + FIX_ACPI_PAGES - 1,
-+#endif
-+#ifdef CONFIG_PCI_MMCONFIG
-+ FIX_PCIE_MCFG,
-+#endif
-+ FIX_SHARED_INFO,
-+#define NR_FIX_ISAMAPS 256
-+ FIX_ISAMAP_END,
-+ FIX_ISAMAP_BEGIN = FIX_ISAMAP_END + NR_FIX_ISAMAPS - 1,
-+ __end_of_permanent_fixed_addresses,
-+ /* temporary boot-time mappings, used before ioremap() is functional */
-+#define NR_FIX_BTMAPS 16
-+ FIX_BTMAP_END = __end_of_permanent_fixed_addresses,
-+ FIX_BTMAP_BEGIN = FIX_BTMAP_END + NR_FIX_BTMAPS - 1,
-+ FIX_WP_TEST,
-+ __end_of_fixed_addresses
-+};
-+
-+extern void set_fixaddr_top(unsigned long top);
-+
-+extern void __set_fixmap(enum fixed_addresses idx,
-+ maddr_t phys, pgprot_t flags);
++#define X86_CR4_VME 0x0001 /* enable vm86 extensions */
++#define X86_CR4_PVI 0x0002 /* virtual interrupts flag enable */
++#define X86_CR4_TSD 0x0004 /* disable time stamp at ipl 3 */
++#define X86_CR4_DE 0x0008 /* enable debugging extensions */
++#define X86_CR4_PSE 0x0010 /* enable page size extensions */
++#define X86_CR4_PAE 0x0020 /* enable physical address extensions */
++#define X86_CR4_MCE 0x0040 /* Machine check enable */
++#define X86_CR4_PGE 0x0080 /* enable global pages */
++#define X86_CR4_PCE 0x0100 /* enable performance counters at ipl 3 */
++#define X86_CR4_OSFXSR 0x0200 /* enable fast FPU save and restore */
++#define X86_CR4_OSXMMEXCPT 0x0400 /* enable unmasked SSE exceptions */
+
-+#define set_fixmap(idx, phys) \
-+ __set_fixmap(idx, phys, PAGE_KERNEL)
+/*
-+ * Some hardware wants to get fixmapped without caching.
++ * Save the cr4 feature set we're using (ie
++ * Pentium 4MB enable and PPro Global page
++ * enable), so that any CPU's that boot up
++ * after us can get the correct flags.
+ */
-+#define set_fixmap_nocache(idx, phys) \
-+ __set_fixmap(idx, phys, PAGE_KERNEL_NOCACHE)
-+
-+#define clear_fixmap(idx) \
-+ __set_fixmap(idx, 0, __pgprot(0))
-+
-+#define FIXADDR_TOP ((unsigned long)__FIXADDR_TOP)
-+
-+#define __FIXADDR_SIZE (__end_of_permanent_fixed_addresses << PAGE_SHIFT)
-+#define __FIXADDR_BOOT_SIZE (__end_of_fixed_addresses << PAGE_SHIFT)
-+#define FIXADDR_START (FIXADDR_TOP - __FIXADDR_SIZE)
-+#define FIXADDR_BOOT_START (FIXADDR_TOP - __FIXADDR_BOOT_SIZE)
-+
-+#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT))
-+#define __virt_to_fix(x) ((FIXADDR_TOP - ((x)&PAGE_MASK)) >> PAGE_SHIFT)
-+
-+extern void __this_fixmap_does_not_exist(void);
++extern unsigned long mmu_cr4_features;
+
-+/*
-+ * 'index to address' translation. If anyone tries to use the idx
-+ * directly without tranlation, we catch the bug with a NULL-deference
-+ * kernel oops. Illegal ranges of incoming indices are caught too.
-+ */
-+static __always_inline unsigned long fix_to_virt(const unsigned int idx)
++static inline void set_in_cr4 (unsigned long mask)
+{
-+ /*
-+ * this branch gets completely eliminated after inlining,
-+ * except when someone tries to use fixaddr indices in an
-+ * illegal way. (such as mixing up address types or using
-+ * out-of-range indices).
-+ *
-+ * If it doesn't get removed, the linker will complain
-+ * loudly with a reasonably clear error message..
-+ */
-+ if (idx >= __end_of_fixed_addresses)
-+ __this_fixmap_does_not_exist();
-+
-+ return __fix_to_virt(idx);
++ unsigned cr4;
++ mmu_cr4_features |= mask;
++ cr4 = read_cr4();
++ cr4 |= mask;
++ write_cr4(cr4);
+}
+
-+static inline unsigned long virt_to_fix(const unsigned long vaddr)
++static inline void clear_in_cr4 (unsigned long mask)
+{
-+ BUG_ON(vaddr >= FIXADDR_TOP || vaddr < FIXADDR_START);
-+ return __virt_to_fix(vaddr);
++ unsigned cr4;
++ mmu_cr4_features &= ~mask;
++ cr4 = read_cr4();
++ cr4 &= ~mask;
++ write_cr4(cr4);
+}
+
-+#endif /* !__ASSEMBLY__ */
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/floppy.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/floppy.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/floppy.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/floppy.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,147 @@
+/*
-+ * Architecture specific parts of the Floppy driver
-+ *
-+ * This file is subject to the terms and conditions of the GNU General Public
-+ * License. See the file "COPYING" in the main directory of this archive
-+ * for more details.
-+ *
-+ * Copyright (C) 1995
-+ *
-+ * Modifications for Xen are Copyright (c) 2004, Keir Fraser.
++ * NSC/Cyrix CPU configuration register indexes
+ */
-+#ifndef __ASM_XEN_I386_FLOPPY_H
-+#define __ASM_XEN_I386_FLOPPY_H
+
-+#include <linux/vmalloc.h>
++#define CX86_PCR0 0x20
++#define CX86_GCR 0xb8
++#define CX86_CCR0 0xc0
++#define CX86_CCR1 0xc1
++#define CX86_CCR2 0xc2
++#define CX86_CCR3 0xc3
++#define CX86_CCR4 0xe8
++#define CX86_CCR5 0xe9
++#define CX86_CCR6 0xea
++#define CX86_CCR7 0xeb
++#define CX86_PCR1 0xf0
++#define CX86_DIR0 0xfe
++#define CX86_DIR1 0xff
++#define CX86_ARR_BASE 0xc4
++#define CX86_RCR_BASE 0xdc
+
-+/* XEN: Hit DMA paths on the head. This trick from asm-m68k/floppy.h. */
-+#include <asm/dma.h>
-+#undef MAX_DMA_ADDRESS
-+#define MAX_DMA_ADDRESS 0
-+#define CROSS_64KB(a,s) (0)
-+
-+#define fd_inb(port) inb_p(port)
-+#define fd_outb(value,port) outb_p(value,port)
-+
-+#define fd_request_dma() (0)
-+#define fd_free_dma() ((void)0)
-+#define fd_enable_irq() enable_irq(FLOPPY_IRQ)
-+#define fd_disable_irq() disable_irq(FLOPPY_IRQ)
-+#define fd_free_irq() free_irq(FLOPPY_IRQ, NULL)
-+#define fd_get_dma_residue() (virtual_dma_count + virtual_dma_residue)
-+#define fd_dma_setup(addr, size, mode, io) vdma_dma_setup(addr, size, mode, io)
-+/*
-+ * Do not use vmalloc/vfree: floppy_release_irq_and_dma() gets called from
-+ * softirq context via motor_off_callback. A generic bug we happen to trigger.
-+ */
-+#define fd_dma_mem_alloc(size) __get_free_pages(GFP_KERNEL, get_order(size))
-+#define fd_dma_mem_free(addr, size) free_pages(addr, get_order(size))
-+
-+static int virtual_dma_count;
-+static int virtual_dma_residue;
-+static char *virtual_dma_addr;
-+static int virtual_dma_mode;
-+static int doing_pdma;
-+
-+static irqreturn_t floppy_hardint(int irq, void *dev_id, struct pt_regs * regs)
-+{
-+ register unsigned char st;
-+ register int lcount;
-+ register char *lptr;
-+
-+ if (!doing_pdma)
-+ return floppy_interrupt(irq, dev_id, regs);
-+
-+ st = 1;
-+ for(lcount=virtual_dma_count, lptr=virtual_dma_addr;
-+ lcount; lcount--, lptr++) {
-+ st=inb(virtual_dma_port+4) & 0xa0 ;
-+ if(st != 0xa0)
-+ break;
-+ if(virtual_dma_mode)
-+ outb_p(*lptr, virtual_dma_port+5);
-+ else
-+ *lptr = inb_p(virtual_dma_port+5);
-+ }
-+ virtual_dma_count = lcount;
-+ virtual_dma_addr = lptr;
-+ st = inb(virtual_dma_port+4);
++/*
++ * NSC/Cyrix CPU indexed register access macros
++ */
+
-+ if(st == 0x20)
-+ return IRQ_HANDLED;
-+ if(!(st & 0x20)) {
-+ virtual_dma_residue += virtual_dma_count;
-+ virtual_dma_count=0;
-+ doing_pdma = 0;
-+ floppy_interrupt(irq, dev_id, regs);
-+ return IRQ_HANDLED;
-+ }
-+ return IRQ_HANDLED;
-+}
++#define getCx86(reg) ({ outb((reg), 0x22); inb(0x23); })
++
++#define setCx86(reg, data) do { \
++ outb((reg), 0x22); \
++ outb((data), 0x23); \
++} while (0)
+
-+static void fd_disable_dma(void)
++/* Stop speculative execution */
++static inline void sync_core(void)
+{
-+ doing_pdma = 0;
-+ virtual_dma_residue += virtual_dma_count;
-+ virtual_dma_count=0;
++ int tmp;
++ asm volatile("cpuid" : "=a" (tmp) : "0" (1) : "ebx","ecx","edx","memory");
+}
+
-+static int fd_request_irq(void)
++static inline void __monitor(const void *eax, unsigned long ecx,
++ unsigned long edx)
+{
-+ return request_irq(FLOPPY_IRQ, floppy_hardint,
-+ IRQF_DISABLED, "floppy", NULL);
++ /* "monitor %eax,%ecx,%edx;" */
++ asm volatile(
++ ".byte 0x0f,0x01,0xc8;"
++ : :"a" (eax), "c" (ecx), "d"(edx));
+}
+
-+static int vdma_dma_setup(char *addr, unsigned long size, int mode, int io)
++static inline void __mwait(unsigned long eax, unsigned long ecx)
+{
-+ doing_pdma = 1;
-+ virtual_dma_port = io;
-+ virtual_dma_mode = (mode == DMA_MODE_WRITE);
-+ virtual_dma_addr = addr;
-+ virtual_dma_count = size;
-+ virtual_dma_residue = 0;
-+ return 0;
++ /* "mwait %eax,%ecx;" */
++ asm volatile(
++ ".byte 0x0f,0x01,0xc9;"
++ : :"a" (eax), "c" (ecx));
+}
+
-+/* XEN: This trick to force 'virtual DMA' is from include/asm-m68k/floppy.h. */
-+#define FDC1 xen_floppy_init()
-+static int FDC2 = -1;
++/* from system description table in BIOS. Mostly for MCA use, but
++others may find it useful. */
++extern unsigned int machine_id;
++extern unsigned int machine_submodel_id;
++extern unsigned int BIOS_revision;
++extern unsigned int mca_pentium_flag;
+
-+static int xen_floppy_init(void)
-+{
-+ use_virtual_dma = 1;
-+ can_use_virtual_dma = 1;
-+ return 0x3f0;
-+}
++/* Boot loader type from the setup header */
++extern int bootloader_type;
+
+/*
-+ * Floppy types are stored in the rtc's CMOS RAM and so rtc_lock
-+ * is needed to prevent corrupted CMOS RAM in case "insmod floppy"
-+ * coincides with another rtc CMOS user. Paul G.
++ * User space process size: 3GB (default).
+ */
-+#define FLOPPY0_TYPE ({ \
-+ unsigned long flags; \
-+ unsigned char val; \
-+ spin_lock_irqsave(&rtc_lock, flags); \
-+ val = (CMOS_READ(0x10) >> 4) & 15; \
-+ spin_unlock_irqrestore(&rtc_lock, flags); \
-+ val; \
-+})
-+
-+#define FLOPPY1_TYPE ({ \
-+ unsigned long flags; \
-+ unsigned char val; \
-+ spin_lock_irqsave(&rtc_lock, flags); \
-+ val = CMOS_READ(0x10) & 15; \
-+ spin_unlock_irqrestore(&rtc_lock, flags); \
-+ val; \
-+})
-+
-+#define N_FDC 2
-+#define N_DRIVE 8
++#define TASK_SIZE (PAGE_OFFSET)
+
-+#define FLOPPY_MOTOR_MASK 0xf0
++/* This decides where the kernel will search for a free chunk of vm
++ * space during mmap's.
++ */
++#define TASK_UNMAPPED_BASE (PAGE_ALIGN(TASK_SIZE / 3))
+
-+#define EXTRA_FLOPPY_PARAMS
++#define HAVE_ARCH_PICK_MMAP_LAYOUT
+
-+#endif /* __ASM_XEN_I386_FLOPPY_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/highmem.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/highmem.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/highmem.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/highmem.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,80 @@
+/*
-+ * highmem.h: virtual kernel memory mappings for high memory
-+ *
-+ * Used in CONFIG_HIGHMEM systems for memory pages which
-+ * are not addressable by direct kernel virtual addresses.
-+ *
-+ * Copyright (C) 1999 Gerhard Wichert, Siemens AG
-+ * Gerhard.Wichert at pdb.siemens.de
-+ *
-+ *
-+ * Redesigned the x86 32-bit VM architecture to deal with
-+ * up to 16 Terabyte physical memory. With current x86 CPUs
-+ * we now support up to 64 Gigabytes physical RAM.
-+ *
-+ * Copyright (C) 1999 Ingo Molnar <mingo at redhat.com>
++ * Size of io_bitmap.
+ */
++#define IO_BITMAP_BITS 65536
++#define IO_BITMAP_BYTES (IO_BITMAP_BITS/8)
++#define IO_BITMAP_LONGS (IO_BITMAP_BYTES/sizeof(long))
++#ifndef CONFIG_X86_NO_TSS
++#define IO_BITMAP_OFFSET offsetof(struct tss_struct,io_bitmap)
++#endif
++#define INVALID_IO_BITMAP_OFFSET 0x8000
++#define INVALID_IO_BITMAP_OFFSET_LAZY 0x9000
+
-+#ifndef _ASM_HIGHMEM_H
-+#define _ASM_HIGHMEM_H
-+
-+#ifdef __KERNEL__
++struct i387_fsave_struct {
++ long cwd;
++ long swd;
++ long twd;
++ long fip;
++ long fcs;
++ long foo;
++ long fos;
++ long st_space[20]; /* 8*10 bytes for each FP-reg = 80 bytes */
++ long status; /* software status information */
++};
+
-+#include <linux/interrupt.h>
-+#include <linux/threads.h>
-+#include <asm/kmap_types.h>
-+#include <asm/tlbflush.h>
++struct i387_fxsave_struct {
++ unsigned short cwd;
++ unsigned short swd;
++ unsigned short twd;
++ unsigned short fop;
++ long fip;
++ long fcs;
++ long foo;
++ long fos;
++ long mxcsr;
++ long mxcsr_mask;
++ long st_space[32]; /* 8*16 bytes for each FP-reg = 128 bytes */
++ long xmm_space[32]; /* 8*16 bytes for each XMM-reg = 128 bytes */
++ long padding[56];
++} __attribute__ ((aligned (16)));
+
-+/* declarations for highmem.c */
-+extern unsigned long highstart_pfn, highend_pfn;
++struct i387_soft_struct {
++ long cwd;
++ long swd;
++ long twd;
++ long fip;
++ long fcs;
++ long foo;
++ long fos;
++ long st_space[20]; /* 8*10 bytes for each FP-reg = 80 bytes */
++ unsigned char ftop, changed, lookahead, no_update, rm, alimit;
++ struct info *info;
++ unsigned long entry_eip;
++};
+
-+extern pte_t *kmap_pte;
-+extern pgprot_t kmap_prot;
-+extern pte_t *pkmap_page_table;
++union i387_union {
++ struct i387_fsave_struct fsave;
++ struct i387_fxsave_struct fxsave;
++ struct i387_soft_struct soft;
++};
+
-+/*
-+ * Right now we initialize only a single pte table. It can be extended
-+ * easily, subsequent pte tables have to be allocated in one physical
-+ * chunk of RAM.
-+ */
-+#ifdef CONFIG_X86_PAE
-+#define LAST_PKMAP 512
-+#else
-+#define LAST_PKMAP 1024
-+#endif
-+/*
-+ * Ordering is:
-+ *
-+ * FIXADDR_TOP
-+ * fixed_addresses
-+ * FIXADDR_START
-+ * temp fixed addresses
-+ * FIXADDR_BOOT_START
-+ * Persistent kmap area
-+ * PKMAP_BASE
-+ * VMALLOC_END
-+ * Vmalloc area
-+ * VMALLOC_START
-+ * high_memory
-+ */
-+#define PKMAP_BASE ( (FIXADDR_BOOT_START - PAGE_SIZE*(LAST_PKMAP + 1)) & PMD_MASK )
-+#define LAST_PKMAP_MASK (LAST_PKMAP-1)
-+#define PKMAP_NR(virt) ((virt-PKMAP_BASE) >> PAGE_SHIFT)
-+#define PKMAP_ADDR(nr) (PKMAP_BASE + ((nr) << PAGE_SHIFT))
++typedef struct {
++ unsigned long seg;
++} mm_segment_t;
+
-+extern void * FASTCALL(kmap_high(struct page *page));
-+extern void FASTCALL(kunmap_high(struct page *page));
++struct thread_struct;
+
-+void *kmap(struct page *page);
-+void kunmap(struct page *page);
-+void *kmap_atomic(struct page *page, enum km_type type);
-+void *kmap_atomic_pte(struct page *page, enum km_type type);
-+void kunmap_atomic(void *kvaddr, enum km_type type);
-+void *kmap_atomic_pfn(unsigned long pfn, enum km_type type);
-+struct page *kmap_atomic_to_page(void *ptr);
++#ifndef CONFIG_X86_NO_TSS
++struct tss_struct {
++ unsigned short back_link,__blh;
++ unsigned long esp0;
++ unsigned short ss0,__ss0h;
++ unsigned long esp1;
++ unsigned short ss1,__ss1h; /* ss1 is used to cache MSR_IA32_SYSENTER_CS */
++ unsigned long esp2;
++ unsigned short ss2,__ss2h;
++ unsigned long __cr3;
++ unsigned long eip;
++ unsigned long eflags;
++ unsigned long eax,ecx,edx,ebx;
++ unsigned long esp;
++ unsigned long ebp;
++ unsigned long esi;
++ unsigned long edi;
++ unsigned short es, __esh;
++ unsigned short cs, __csh;
++ unsigned short ss, __ssh;
++ unsigned short ds, __dsh;
++ unsigned short fs, __fsh;
++ unsigned short gs, __gsh;
++ unsigned short ldt, __ldth;
++ unsigned short trace, io_bitmap_base;
++ /*
++ * The extra 1 is there because the CPU will access an
++ * additional byte beyond the end of the IO permission
++ * bitmap. The extra byte must be all 1 bits, and must
++ * be within the limit.
++ */
++ unsigned long io_bitmap[IO_BITMAP_LONGS + 1];
++ /*
++ * Cache the current maximum and the last task that used the bitmap:
++ */
++ unsigned long io_bitmap_max;
++ struct thread_struct *io_bitmap_owner;
++ /*
++ * pads the TSS to be cacheline-aligned (size is 0x100)
++ */
++ unsigned long __cacheline_filler[35];
++ /*
++ * .. and then another 0x100 bytes for emergency kernel stack
++ */
++ unsigned long stack[64];
++} __attribute__((packed));
++#endif
+
-+#define flush_cache_kmaps() do { } while (0)
++#define ARCH_MIN_TASKALIGN 16
+
-+#endif /* __KERNEL__ */
++struct thread_struct {
++/* cached TLS descriptors. */
++ struct desc_struct tls_array[GDT_ENTRY_TLS_ENTRIES];
++ unsigned long esp0;
++ unsigned long sysenter_cs;
++ unsigned long eip;
++ unsigned long esp;
++ unsigned long fs;
++ unsigned long gs;
++/* Hardware debugging registers */
++ unsigned long debugreg[8]; /* %%db0-7 debug registers */
++/* fault info */
++ unsigned long cr2, trap_no, error_code;
++/* floating point info */
++ union i387_union i387;
++/* virtual 86 mode info */
++ struct vm86_struct __user * vm86_info;
++ unsigned long screen_bitmap;
++ unsigned long v86flags, v86mask, saved_esp0;
++ unsigned int saved_fs, saved_gs;
++/* IO permissions */
++ unsigned long *io_bitmap_ptr;
++ unsigned long iopl;
++/* max allowed port in the bitmap, in bytes: */
++ unsigned long io_bitmap_max;
++};
+
-+#endif /* _ASM_HIGHMEM_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/hw_irq.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/hw_irq.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/hw_irq.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/hw_irq.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,72 @@
-+#ifndef _ASM_HW_IRQ_H
-+#define _ASM_HW_IRQ_H
++#define INIT_THREAD { \
++ .vm86_info = NULL, \
++ .sysenter_cs = __KERNEL_CS, \
++ .io_bitmap_ptr = NULL, \
++}
+
++#ifndef CONFIG_X86_NO_TSS
+/*
-+ * linux/include/asm/hw_irq.h
-+ *
-+ * (C) 1992, 1993 Linus Torvalds, (C) 1997 Ingo Molnar
-+ *
-+ * moved some of the old arch/i386/kernel/irq.h to here. VY
-+ *
-+ * IRQ/IPI changes taken from work by Thomas Radke
-+ * <tomsoft at informatik.tu-chemnitz.de>
++ * Note that the .io_bitmap member must be extra-big. This is because
++ * the CPU will access an additional byte beyond the end of the IO
++ * permission bitmap. The extra byte must be all 1 bits, and must
++ * be within the limit.
+ */
++#define INIT_TSS { \
++ .esp0 = sizeof(init_stack) + (long)&init_stack, \
++ .ss0 = __KERNEL_DS, \
++ .ss1 = __KERNEL_CS, \
++ .io_bitmap_base = INVALID_IO_BITMAP_OFFSET, \
++ .io_bitmap = { [ 0 ... IO_BITMAP_LONGS] = ~0 }, \
++}
+
-+#include <linux/profile.h>
-+#include <asm/atomic.h>
-+#include <asm/irq.h>
-+#include <asm/sections.h>
-+
-+struct hw_interrupt_type;
++static inline void __load_esp0(struct tss_struct *tss, struct thread_struct *thread)
++{
++ tss->esp0 = thread->esp0;
++ /* This can only happen when SEP is enabled, no need to test "SEP"arately */
++ if (unlikely(tss->ss1 != thread->sysenter_cs)) {
++ tss->ss1 = thread->sysenter_cs;
++ wrmsr(MSR_IA32_SYSENTER_CS, thread->sysenter_cs, 0);
++ }
++}
++#define load_esp0(tss, thread) \
++ __load_esp0(tss, thread)
++#else
++#define load_esp0(tss, thread) do { \
++ if (HYPERVISOR_stack_switch(__KERNEL_DS, (thread)->esp0)) \
++ BUG(); \
++} while (0)
++#endif
+
-+#define NMI_VECTOR 0x02
++#define start_thread(regs, new_eip, new_esp) do { \
++ __asm__("movl %0,%%fs ; movl %0,%%gs": :"r" (0)); \
++ set_fs(USER_DS); \
++ regs->xds = __USER_DS; \
++ regs->xes = __USER_DS; \
++ regs->xss = __USER_DS; \
++ regs->xcs = __USER_CS; \
++ regs->eip = new_eip; \
++ regs->esp = new_esp; \
++} while (0)
+
+/*
-+ * Various low-level irq details needed by irq.c, process.c,
-+ * time.c, io_apic.c and smp.c
-+ *
-+ * Interrupt entry/exit code at both C and assembly level
++ * These special macros can be used to get or set a debugging register
+ */
++#define get_debugreg(var, register) \
++ (var) = HYPERVISOR_get_debugreg((register))
++#define set_debugreg(value, register) \
++ WARN_ON(HYPERVISOR_set_debugreg((register), (value)))
+
-+extern u8 irq_vector[NR_IRQ_VECTORS];
-+#define IO_APIC_VECTOR(irq) (irq_vector[irq])
-+#define AUTO_ASSIGN -1
-+
-+extern void (*interrupt[NR_IRQS])(void);
-+
-+#ifdef CONFIG_SMP
-+fastcall void reschedule_interrupt(void);
-+fastcall void invalidate_interrupt(void);
-+fastcall void call_function_interrupt(void);
-+#endif
-+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+fastcall void apic_timer_interrupt(void);
-+fastcall void error_interrupt(void);
-+fastcall void spurious_interrupt(void);
-+fastcall void thermal_interrupt(struct pt_regs *);
-+#define platform_legacy_irq(irq) ((irq) < 16)
-+#endif
++/*
++ * Set IOPL bits in EFLAGS from given mask
++ */
++static inline void set_iopl_mask(unsigned mask)
++{
++ struct physdev_set_iopl set_iopl;
+
-+void disable_8259A_irq(unsigned int irq);
-+void enable_8259A_irq(unsigned int irq);
-+int i8259A_irq_pending(unsigned int irq);
-+void make_8259A_irq(unsigned int irq);
-+void init_8259A(int aeoi);
-+void FASTCALL(send_IPI_self(int vector));
-+void init_VISWS_APIC_irqs(void);
-+void setup_IO_APIC(void);
-+void disable_IO_APIC(void);
-+void print_IO_APIC(void);
-+int IO_APIC_get_PCI_irq_vector(int bus, int slot, int fn);
-+void send_IPI(int dest, int vector);
-+void setup_ioapic_dest(void);
++ /* Force the change at ring 0. */
++ set_iopl.iopl = (mask == 0) ? 1 : (mask >> 12) & 3;
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl));
++}
+
-+extern unsigned long io_apic_irqs;
++/* Forward declaration, a strange C thing */
++struct task_struct;
++struct mm_struct;
+
-+extern atomic_t irq_err_count;
-+extern atomic_t irq_mis_count;
++/* Free all resources held by a thread. */
++extern void release_thread(struct task_struct *);
+
-+#define IO_APIC_IRQ(x) (((x) >= 16) || ((1<<(x)) & io_apic_irqs))
++/* Prepare to copy thread state - unlazy all lazy status */
++extern void prepare_to_copy(struct task_struct *tsk);
+
-+#endif /* _ASM_HW_IRQ_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/hypercall.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/hypercall.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/hypercall.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/hypercall.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,407 @@
-+/******************************************************************************
-+ * hypercall.h
-+ *
-+ * Linux-specific hypervisor handling.
-+ *
-+ * Copyright (c) 2002-2004, K A Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
++/*
++ * create a kernel thread without removing it from tasklists
+ */
++extern int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags);
+
-+#ifndef __HYPERCALL_H__
-+#define __HYPERCALL_H__
++extern unsigned long thread_saved_pc(struct task_struct *tsk);
++void show_trace(struct task_struct *task, struct pt_regs *regs, unsigned long *stack);
+
-+#include <linux/string.h> /* memcpy() */
++unsigned long get_wchan(struct task_struct *p);
+
-+#ifndef __HYPERVISOR_H__
-+# error "please don't include this file directly"
-+#endif
++#define THREAD_SIZE_LONGS (THREAD_SIZE/sizeof(unsigned long))
++#define KSTK_TOP(info) \
++({ \
++ unsigned long *__ptr = (unsigned long *)(info); \
++ (unsigned long)(&__ptr[THREAD_SIZE_LONGS]); \
++})
+
-+#define __STR(x) #x
-+#define STR(x) __STR(x)
++/*
++ * The below -8 is to reserve 8 bytes on top of the ring0 stack.
++ * This is necessary to guarantee that the entire "struct pt_regs"
++ * is accessable even if the CPU haven't stored the SS/ESP registers
++ * on the stack (interrupt gate does not save these registers
++ * when switching to the same priv ring).
++ * Therefore beware: accessing the xss/esp fields of the
++ * "struct pt_regs" is possible, but they may contain the
++ * completely wrong values.
++ */
++#define task_pt_regs(task) \
++({ \
++ struct pt_regs *__regs__; \
++ __regs__ = (struct pt_regs *)(KSTK_TOP(task_stack_page(task))-8); \
++ __regs__ - 1; \
++})
+
-+#ifdef CONFIG_XEN
-+#define HYPERCALL_STR(name) \
-+ "call hypercall_page + ("STR(__HYPERVISOR_##name)" * 32)"
-+#else
-+#define HYPERCALL_STR(name) \
-+ "mov hypercall_stubs,%%eax; " \
-+ "add $("STR(__HYPERVISOR_##name)" * 32),%%eax; " \
-+ "call *%%eax"
-+#endif
++#define KSTK_EIP(task) (task_pt_regs(task)->eip)
++#define KSTK_ESP(task) (task_pt_regs(task)->esp)
+
-+#define _hypercall0(type, name) \
-+({ \
-+ long __res; \
-+ asm volatile ( \
-+ HYPERCALL_STR(name) \
-+ : "=a" (__res) \
-+ : \
-+ : "memory" ); \
-+ (type)__res; \
-+})
+
-+#define _hypercall1(type, name, a1) \
-+({ \
-+ long __res, __ign1; \
-+ asm volatile ( \
-+ HYPERCALL_STR(name) \
-+ : "=a" (__res), "=b" (__ign1) \
-+ : "1" ((long)(a1)) \
-+ : "memory" ); \
-+ (type)__res; \
-+})
++struct microcode_header {
++ unsigned int hdrver;
++ unsigned int rev;
++ unsigned int date;
++ unsigned int sig;
++ unsigned int cksum;
++ unsigned int ldrver;
++ unsigned int pf;
++ unsigned int datasize;
++ unsigned int totalsize;
++ unsigned int reserved[3];
++};
+
-+#define _hypercall2(type, name, a1, a2) \
-+({ \
-+ long __res, __ign1, __ign2; \
-+ asm volatile ( \
-+ HYPERCALL_STR(name) \
-+ : "=a" (__res), "=b" (__ign1), "=c" (__ign2) \
-+ : "1" ((long)(a1)), "2" ((long)(a2)) \
-+ : "memory" ); \
-+ (type)__res; \
-+})
++struct microcode {
++ struct microcode_header hdr;
++ unsigned int bits[0];
++};
+
-+#define _hypercall3(type, name, a1, a2, a3) \
-+({ \
-+ long __res, __ign1, __ign2, __ign3; \
-+ asm volatile ( \
-+ HYPERCALL_STR(name) \
-+ : "=a" (__res), "=b" (__ign1), "=c" (__ign2), \
-+ "=d" (__ign3) \
-+ : "1" ((long)(a1)), "2" ((long)(a2)), \
-+ "3" ((long)(a3)) \
-+ : "memory" ); \
-+ (type)__res; \
-+})
++typedef struct microcode microcode_t;
++typedef struct microcode_header microcode_header_t;
+
-+#define _hypercall4(type, name, a1, a2, a3, a4) \
-+({ \
-+ long __res, __ign1, __ign2, __ign3, __ign4; \
-+ asm volatile ( \
-+ HYPERCALL_STR(name) \
-+ : "=a" (__res), "=b" (__ign1), "=c" (__ign2), \
-+ "=d" (__ign3), "=S" (__ign4) \
-+ : "1" ((long)(a1)), "2" ((long)(a2)), \
-+ "3" ((long)(a3)), "4" ((long)(a4)) \
-+ : "memory" ); \
-+ (type)__res; \
-+})
++/* microcode format is extended from prescott processors */
++struct extended_signature {
++ unsigned int sig;
++ unsigned int pf;
++ unsigned int cksum;
++};
+
-+#define _hypercall5(type, name, a1, a2, a3, a4, a5) \
-+({ \
-+ long __res, __ign1, __ign2, __ign3, __ign4, __ign5; \
-+ asm volatile ( \
-+ HYPERCALL_STR(name) \
-+ : "=a" (__res), "=b" (__ign1), "=c" (__ign2), \
-+ "=d" (__ign3), "=S" (__ign4), "=D" (__ign5) \
-+ : "1" ((long)(a1)), "2" ((long)(a2)), \
-+ "3" ((long)(a3)), "4" ((long)(a4)), \
-+ "5" ((long)(a5)) \
-+ : "memory" ); \
-+ (type)__res; \
-+})
++struct extended_sigtable {
++ unsigned int count;
++ unsigned int cksum;
++ unsigned int reserved[3];
++ struct extended_signature sigs[0];
++};
+
-+static inline int
-+HYPERVISOR_set_trap_table(
-+ trap_info_t *table)
++/* REP NOP (PAUSE) is a good thing to insert into busy-wait loops. */
++static inline void rep_nop(void)
+{
-+ return _hypercall1(int, set_trap_table, table);
++ __asm__ __volatile__("rep;nop": : :"memory");
+}
+
-+static inline int
-+HYPERVISOR_mmu_update(
-+ mmu_update_t *req, int count, int *success_count, domid_t domid)
-+{
-+ return _hypercall4(int, mmu_update, req, count, success_count, domid);
-+}
++#define cpu_relax() rep_nop()
+
-+static inline int
-+HYPERVISOR_mmuext_op(
-+ struct mmuext_op *op, int count, int *success_count, domid_t domid)
-+{
-+ return _hypercall4(int, mmuext_op, op, count, success_count, domid);
-+}
++/* generic versions from gas */
++#define GENERIC_NOP1 ".byte 0x90\n"
++#define GENERIC_NOP2 ".byte 0x89,0xf6\n"
++#define GENERIC_NOP3 ".byte 0x8d,0x76,0x00\n"
++#define GENERIC_NOP4 ".byte 0x8d,0x74,0x26,0x00\n"
++#define GENERIC_NOP5 GENERIC_NOP1 GENERIC_NOP4
++#define GENERIC_NOP6 ".byte 0x8d,0xb6,0x00,0x00,0x00,0x00\n"
++#define GENERIC_NOP7 ".byte 0x8d,0xb4,0x26,0x00,0x00,0x00,0x00\n"
++#define GENERIC_NOP8 GENERIC_NOP1 GENERIC_NOP7
+
-+static inline int
-+HYPERVISOR_set_gdt(
-+ unsigned long *frame_list, int entries)
-+{
-+ return _hypercall2(int, set_gdt, frame_list, entries);
-+}
++/* Opteron nops */
++#define K8_NOP1 GENERIC_NOP1
++#define K8_NOP2 ".byte 0x66,0x90\n"
++#define K8_NOP3 ".byte 0x66,0x66,0x90\n"
++#define K8_NOP4 ".byte 0x66,0x66,0x66,0x90\n"
++#define K8_NOP5 K8_NOP3 K8_NOP2
++#define K8_NOP6 K8_NOP3 K8_NOP3
++#define K8_NOP7 K8_NOP4 K8_NOP3
++#define K8_NOP8 K8_NOP4 K8_NOP4
+
-+static inline int
-+HYPERVISOR_stack_switch(
-+ unsigned long ss, unsigned long esp)
-+{
-+ return _hypercall2(int, stack_switch, ss, esp);
-+}
++/* K7 nops */
++/* uses eax dependencies (arbitary choice) */
++#define K7_NOP1 GENERIC_NOP1
++#define K7_NOP2 ".byte 0x8b,0xc0\n"
++#define K7_NOP3 ".byte 0x8d,0x04,0x20\n"
++#define K7_NOP4 ".byte 0x8d,0x44,0x20,0x00\n"
++#define K7_NOP5 K7_NOP4 ASM_NOP1
++#define K7_NOP6 ".byte 0x8d,0x80,0,0,0,0\n"
++#define K7_NOP7 ".byte 0x8D,0x04,0x05,0,0,0,0\n"
++#define K7_NOP8 K7_NOP7 ASM_NOP1
+
-+static inline int
-+HYPERVISOR_set_callbacks(
-+ unsigned long event_selector, unsigned long event_address,
-+ unsigned long failsafe_selector, unsigned long failsafe_address)
-+{
-+ return _hypercall4(int, set_callbacks,
-+ event_selector, event_address,
-+ failsafe_selector, failsafe_address);
-+}
++#ifdef CONFIG_MK8
++#define ASM_NOP1 K8_NOP1
++#define ASM_NOP2 K8_NOP2
++#define ASM_NOP3 K8_NOP3
++#define ASM_NOP4 K8_NOP4
++#define ASM_NOP5 K8_NOP5
++#define ASM_NOP6 K8_NOP6
++#define ASM_NOP7 K8_NOP7
++#define ASM_NOP8 K8_NOP8
++#elif defined(CONFIG_MK7)
++#define ASM_NOP1 K7_NOP1
++#define ASM_NOP2 K7_NOP2
++#define ASM_NOP3 K7_NOP3
++#define ASM_NOP4 K7_NOP4
++#define ASM_NOP5 K7_NOP5
++#define ASM_NOP6 K7_NOP6
++#define ASM_NOP7 K7_NOP7
++#define ASM_NOP8 K7_NOP8
++#else
++#define ASM_NOP1 GENERIC_NOP1
++#define ASM_NOP2 GENERIC_NOP2
++#define ASM_NOP3 GENERIC_NOP3
++#define ASM_NOP4 GENERIC_NOP4
++#define ASM_NOP5 GENERIC_NOP5
++#define ASM_NOP6 GENERIC_NOP6
++#define ASM_NOP7 GENERIC_NOP7
++#define ASM_NOP8 GENERIC_NOP8
++#endif
+
-+static inline int
-+HYPERVISOR_fpu_taskswitch(
-+ int set)
-+{
-+ return _hypercall1(int, fpu_taskswitch, set);
-+}
++#define ASM_NOP_MAX 8
+
-+static inline int
-+HYPERVISOR_sched_op_compat(
-+ int cmd, unsigned long arg)
++/* Prefetch instructions for Pentium III and AMD Athlon */
++/* It's not worth to care about 3dnow! prefetches for the K6
++ because they are microcoded there and very slow.
++ However we don't do prefetches for pre XP Athlons currently
++ That should be fixed. */
++#define ARCH_HAS_PREFETCH
++static inline void prefetch(const void *x)
+{
-+ return _hypercall2(int, sched_op_compat, cmd, arg);
++ alternative_input(ASM_NOP4,
++ "prefetchnta (%1)",
++ X86_FEATURE_XMM,
++ "r" (x));
+}
+
-+static inline int
-+HYPERVISOR_sched_op(
-+ int cmd, void *arg)
-+{
-+ return _hypercall2(int, sched_op, cmd, arg);
-+}
++#define ARCH_HAS_PREFETCH
++#define ARCH_HAS_PREFETCHW
++#define ARCH_HAS_SPINLOCK_PREFETCH
+
-+static inline long
-+HYPERVISOR_set_timer_op(
-+ u64 timeout)
++/* 3dnow! prefetch to get an exclusive cache line. Useful for
++ spinlocks to avoid one state transition in the cache coherency protocol. */
++static inline void prefetchw(const void *x)
+{
-+ unsigned long timeout_hi = (unsigned long)(timeout>>32);
-+ unsigned long timeout_lo = (unsigned long)timeout;
-+ return _hypercall2(long, set_timer_op, timeout_lo, timeout_hi);
++ alternative_input(ASM_NOP4,
++ "prefetchw (%1)",
++ X86_FEATURE_3DNOW,
++ "r" (x));
+}
++#define spin_lock_prefetch(x) prefetchw(x)
+
-+static inline int
-+HYPERVISOR_platform_op(
-+ struct xen_platform_op *platform_op)
-+{
-+ platform_op->interface_version = XENPF_INTERFACE_VERSION;
-+ return _hypercall1(int, platform_op, platform_op);
-+}
++extern void select_idle_routine(const struct cpuinfo_x86 *c);
+
-+static inline int
-+HYPERVISOR_set_debugreg(
-+ int reg, unsigned long value)
-+{
-+ return _hypercall2(int, set_debugreg, reg, value);
-+}
++#define cache_line_size() (boot_cpu_data.x86_cache_alignment)
+
-+static inline unsigned long
-+HYPERVISOR_get_debugreg(
-+ int reg)
-+{
-+ return _hypercall1(unsigned long, get_debugreg, reg);
-+}
++extern unsigned long boot_option_idle_override;
++extern void enable_sep_cpu(void);
++extern int sysenter_setup(void);
+
-+static inline int
-+HYPERVISOR_update_descriptor(
-+ u64 ma, u64 desc)
-+{
-+ return _hypercall4(int, update_descriptor, ma, ma>>32, desc, desc>>32);
-+}
++#endif /* __ASM_I386_PROCESSOR_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/ptrace.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/ptrace.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,90 @@
++#ifndef _I386_PTRACE_H
++#define _I386_PTRACE_H
+
-+static inline int
-+HYPERVISOR_memory_op(
-+ unsigned int cmd, void *arg)
-+{
-+ return _hypercall2(int, memory_op, cmd, arg);
-+}
++#define EBX 0
++#define ECX 1
++#define EDX 2
++#define ESI 3
++#define EDI 4
++#define EBP 5
++#define EAX 6
++#define DS 7
++#define ES 8
++#define FS 9
++#define GS 10
++#define ORIG_EAX 11
++#define EIP 12
++#define CS 13
++#define EFL 14
++#define UESP 15
++#define SS 16
++#define FRAME_SIZE 17
+
-+static inline int
-+HYPERVISOR_multicall(
-+ multicall_entry_t *call_list, int nr_calls)
-+{
-+ return _hypercall2(int, multicall, call_list, nr_calls);
-+}
++/* this struct defines the way the registers are stored on the
++ stack during a system call. */
+
-+static inline int
-+HYPERVISOR_update_va_mapping(
-+ unsigned long va, pte_t new_val, unsigned long flags)
-+{
-+ unsigned long pte_hi = 0;
-+#ifdef CONFIG_X86_PAE
-+ pte_hi = new_val.pte_high;
-+#endif
-+ return _hypercall4(int, update_va_mapping, va,
-+ new_val.pte_low, pte_hi, flags);
-+}
++struct pt_regs {
++ long ebx;
++ long ecx;
++ long edx;
++ long esi;
++ long edi;
++ long ebp;
++ long eax;
++ int xds;
++ int xes;
++ long orig_eax;
++ long eip;
++ int xcs;
++ long eflags;
++ long esp;
++ int xss;
++};
+
-+static inline int
-+HYPERVISOR_event_channel_op(
-+ int cmd, void *arg)
-+{
-+ int rc = _hypercall2(int, event_channel_op, cmd, arg);
++/* Arbitrarily choose the same ptrace numbers as used by the Sparc code. */
++#define PTRACE_GETREGS 12
++#define PTRACE_SETREGS 13
++#define PTRACE_GETFPREGS 14
++#define PTRACE_SETFPREGS 15
++#define PTRACE_GETFPXREGS 18
++#define PTRACE_SETFPXREGS 19
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (unlikely(rc == -ENOSYS)) {
-+ struct evtchn_op op;
-+ op.cmd = cmd;
-+ memcpy(&op.u, arg, sizeof(op.u));
-+ rc = _hypercall1(int, event_channel_op_compat, &op);
-+ memcpy(arg, &op.u, sizeof(op.u));
-+ }
-+#endif
++#define PTRACE_OLDSETOPTIONS 21
+
-+ return rc;
-+}
++#define PTRACE_GET_THREAD_AREA 25
++#define PTRACE_SET_THREAD_AREA 26
+
-+static inline int
-+HYPERVISOR_acm_op(
-+ int cmd, void *arg)
-+{
-+ return _hypercall2(int, acm_op, cmd, arg);
-+}
++#define PTRACE_SYSEMU 31
++#define PTRACE_SYSEMU_SINGLESTEP 32
+
-+static inline int
-+HYPERVISOR_xen_version(
-+ int cmd, void *arg)
-+{
-+ return _hypercall2(int, xen_version, cmd, arg);
-+}
++#ifdef __KERNEL__
+
-+static inline int
-+HYPERVISOR_console_io(
-+ int cmd, int count, char *str)
++#include <asm/vm86.h>
++
++struct task_struct;
++extern void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs, int error_code);
++
++/*
++ * user_mode_vm(regs) determines whether a register set came from user mode.
++ * This is true if V8086 mode was enabled OR if the register set was from
++ * protected mode with RPL-3 CS value. This tricky test checks that with
++ * one comparison. Many places in the kernel can bypass this full check
++ * if they have already ruled out V8086 mode, so user_mode(regs) can be used.
++ */
++static inline int user_mode(struct pt_regs *regs)
+{
-+ return _hypercall3(int, console_io, cmd, count, str);
++ return (regs->xcs & 2) != 0;
+}
-+
-+static inline int
-+HYPERVISOR_physdev_op(
-+ int cmd, void *arg)
++static inline int user_mode_vm(struct pt_regs *regs)
+{
-+ int rc = _hypercall2(int, physdev_op, cmd, arg);
++ return ((regs->xcs & 2) | (regs->eflags & VM_MASK)) != 0;
++}
++#define instruction_pointer(regs) ((regs)->eip)
++#if defined(CONFIG_SMP) && defined(CONFIG_FRAME_POINTER)
++extern unsigned long profile_pc(struct pt_regs *regs);
++#else
++#define profile_pc(regs) instruction_pointer(regs)
++#endif
++#endif /* __KERNEL__ */
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (unlikely(rc == -ENOSYS)) {
-+ struct physdev_op op;
-+ op.cmd = cmd;
-+ memcpy(&op.u, arg, sizeof(op.u));
-+ rc = _hypercall1(int, physdev_op_compat, &op);
-+ memcpy(arg, &op.u, sizeof(op.u));
-+ }
+#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/scatterlist.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/scatterlist.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,22 @@
++#ifndef _I386_SCATTERLIST_H
++#define _I386_SCATTERLIST_H
+
-+ return rc;
-+}
++struct scatterlist {
++ struct page *page;
++ unsigned int offset;
++ unsigned int length;
++ dma_addr_t dma_address;
++ unsigned int dma_length;
++};
+
-+static inline int
-+HYPERVISOR_grant_table_op(
-+ unsigned int cmd, void *uop, unsigned int count)
-+{
-+ return _hypercall3(int, grant_table_op, cmd, uop, count);
-+}
++/* These macros should be used after a pci_map_sg call has been done
++ * to get bus addresses of each of the SG entries and their lengths.
++ * You should only work with the number of sg entries pci_map_sg
++ * returns.
++ */
++#define sg_dma_address(sg) ((sg)->dma_address)
++#define sg_dma_len(sg) ((sg)->dma_length)
+
-+static inline int
-+HYPERVISOR_update_va_mapping_otherdomain(
-+ unsigned long va, pte_t new_val, unsigned long flags, domid_t domid)
-+{
-+ unsigned long pte_hi = 0;
-+#ifdef CONFIG_X86_PAE
-+ pte_hi = new_val.pte_high;
-+#endif
-+ return _hypercall5(int, update_va_mapping_otherdomain, va,
-+ new_val.pte_low, pte_hi, flags, domid);
-+}
++#define ISA_DMA_THRESHOLD (0x00ffffff)
+
-+static inline int
-+HYPERVISOR_vm_assist(
-+ unsigned int cmd, unsigned int type)
-+{
-+ return _hypercall2(int, vm_assist, cmd, type);
-+}
++#endif /* !(_I386_SCATTERLIST_H) */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/segment.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/segment.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,117 @@
++#ifndef _ASM_SEGMENT_H
++#define _ASM_SEGMENT_H
+
-+static inline int
-+HYPERVISOR_vcpu_op(
-+ int cmd, int vcpuid, void *extra_args)
-+{
-+ return _hypercall3(int, vcpu_op, cmd, vcpuid, extra_args);
-+}
++/*
++ * The layout of the per-CPU GDT under Linux:
++ *
++ * 0 - null
++ * 1 - reserved
++ * 2 - reserved
++ * 3 - reserved
++ *
++ * 4 - unused <==== new cacheline
++ * 5 - unused
++ *
++ * ------- start of TLS (Thread-Local Storage) segments:
++ *
++ * 6 - TLS segment #1 [ glibc's TLS segment ]
++ * 7 - TLS segment #2 [ Wine's %fs Win32 segment ]
++ * 8 - TLS segment #3
++ * 9 - reserved
++ * 10 - reserved
++ * 11 - reserved
++ *
++ * ------- start of kernel segments:
++ *
++ * 12 - kernel code segment <==== new cacheline
++ * 13 - kernel data segment
++ * 14 - default user CS
++ * 15 - default user DS
++ * 16 - TSS
++ * 17 - LDT
++ * 18 - PNPBIOS support (16->32 gate)
++ * 19 - PNPBIOS support
++ * 20 - PNPBIOS support
++ * 21 - PNPBIOS support
++ * 22 - PNPBIOS support
++ * 23 - APM BIOS support
++ * 24 - APM BIOS support
++ * 25 - APM BIOS support
++ *
++ * 26 - ESPFIX small SS
++ * 27 - unused
++ * 28 - unused
++ * 29 - unused
++ * 30 - unused
++ * 31 - TSS for double fault handler
++ */
++#define GDT_ENTRY_TLS_ENTRIES 3
++#define GDT_ENTRY_TLS_MIN 6
++#define GDT_ENTRY_TLS_MAX (GDT_ENTRY_TLS_MIN + GDT_ENTRY_TLS_ENTRIES - 1)
+
-+static inline int
-+HYPERVISOR_suspend(
-+ unsigned long srec)
-+{
-+ struct sched_shutdown sched_shutdown = {
-+ .reason = SHUTDOWN_suspend
-+ };
++#define TLS_SIZE (GDT_ENTRY_TLS_ENTRIES * 8)
+
-+ int rc = _hypercall3(int, sched_op, SCHEDOP_shutdown,
-+ &sched_shutdown, srec);
++#define GDT_ENTRY_DEFAULT_USER_CS 14
++#define __USER_CS (GDT_ENTRY_DEFAULT_USER_CS * 8 + 3)
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (rc == -ENOSYS)
-+ rc = _hypercall3(int, sched_op_compat, SCHEDOP_shutdown,
-+ SHUTDOWN_suspend, srec);
-+#endif
++#define GDT_ENTRY_DEFAULT_USER_DS 15
++#define __USER_DS (GDT_ENTRY_DEFAULT_USER_DS * 8 + 3)
++
++#define GDT_ENTRY_KERNEL_BASE 12
++
++#define GDT_ENTRY_KERNEL_CS (GDT_ENTRY_KERNEL_BASE + 0)
++#define __KERNEL_CS (GDT_ENTRY_KERNEL_CS * 8)
++#define GET_KERNEL_CS() (__KERNEL_CS | (xen_feature(XENFEAT_supervisor_mode_kernel)?0:1) )
++
++#define GDT_ENTRY_KERNEL_DS (GDT_ENTRY_KERNEL_BASE + 1)
++#define __KERNEL_DS (GDT_ENTRY_KERNEL_DS * 8)
++#define GET_KERNEL_DS() (__KERNEL_DS | (xen_feature(XENFEAT_supervisor_mode_kernel)?0:1) )
++
++#define GDT_ENTRY_TSS (GDT_ENTRY_KERNEL_BASE + 4)
++#define GDT_ENTRY_LDT (GDT_ENTRY_KERNEL_BASE + 5)
++
++#define GDT_ENTRY_PNPBIOS_BASE (GDT_ENTRY_KERNEL_BASE + 6)
++#define GDT_ENTRY_APMBIOS_BASE (GDT_ENTRY_KERNEL_BASE + 11)
++
++#define GDT_ENTRY_ESPFIX_SS (GDT_ENTRY_KERNEL_BASE + 14)
++#define __ESPFIX_SS (GDT_ENTRY_ESPFIX_SS * 8)
+
-+ return rc;
-+}
++#define GDT_ENTRY_DOUBLEFAULT_TSS 31
+
-+static inline int
-+HYPERVISOR_nmi_op(
-+ unsigned long op, void *arg)
-+{
-+ return _hypercall2(int, nmi_op, op, arg);
-+}
++/*
++ * The GDT has 32 entries
++ */
++#define GDT_ENTRIES 32
+
-+static inline unsigned long
-+HYPERVISOR_hvm_op(
-+ int op, void *arg)
-+{
-+ return _hypercall2(unsigned long, hvm_op, op, arg);
-+}
++#define GDT_SIZE (GDT_ENTRIES * 8)
+
-+static inline int
-+HYPERVISOR_callback_op(
-+ int cmd, void *arg)
-+{
-+ return _hypercall2(int, callback_op, cmd, arg);
-+}
++/* Simple and small GDT entries for booting only */
+
-+static inline int
-+HYPERVISOR_xenoprof_op(
-+ int op, void *arg)
-+{
-+ return _hypercall2(int, xenoprof_op, op, arg);
-+}
++#define GDT_ENTRY_BOOT_CS 2
++#define __BOOT_CS (GDT_ENTRY_BOOT_CS * 8)
+
-+static inline int
-+HYPERVISOR_kexec_op(
-+ unsigned long op, void *args)
-+{
-+ return _hypercall2(int, kexec_op, op, args);
-+}
++#define GDT_ENTRY_BOOT_DS (GDT_ENTRY_BOOT_CS + 1)
++#define __BOOT_DS (GDT_ENTRY_BOOT_DS * 8)
+
++/* The PnP BIOS entries in the GDT */
++#define GDT_ENTRY_PNPBIOS_CS32 (GDT_ENTRY_PNPBIOS_BASE + 0)
++#define GDT_ENTRY_PNPBIOS_CS16 (GDT_ENTRY_PNPBIOS_BASE + 1)
++#define GDT_ENTRY_PNPBIOS_DS (GDT_ENTRY_PNPBIOS_BASE + 2)
++#define GDT_ENTRY_PNPBIOS_TS1 (GDT_ENTRY_PNPBIOS_BASE + 3)
++#define GDT_ENTRY_PNPBIOS_TS2 (GDT_ENTRY_PNPBIOS_BASE + 4)
+
++/* The PnP BIOS selectors */
++#define PNP_CS32 (GDT_ENTRY_PNPBIOS_CS32 * 8) /* segment for calling fn */
++#define PNP_CS16 (GDT_ENTRY_PNPBIOS_CS16 * 8) /* code segment for BIOS */
++#define PNP_DS (GDT_ENTRY_PNPBIOS_DS * 8) /* data segment for BIOS */
++#define PNP_TS1 (GDT_ENTRY_PNPBIOS_TS1 * 8) /* transfer data segment */
++#define PNP_TS2 (GDT_ENTRY_PNPBIOS_TS2 * 8) /* another data segment */
+
-+#endif /* __HYPERCALL_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/hypervisor.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/hypervisor.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/hypervisor.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/hypervisor.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,265 @@
-+/******************************************************************************
-+ * hypervisor.h
-+ *
-+ * Linux-specific hypervisor handling.
-+ *
-+ * Copyright (c) 2002-2004, K A Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
++/*
++ * The interrupt descriptor table has room for 256 idt's,
++ * the global descriptor table is dependent on the number
++ * of tasks we can have..
+ */
++#define IDT_ENTRIES 256
+
-+#ifndef __HYPERVISOR_H__
-+#define __HYPERVISOR_H__
-+
-+#include <linux/types.h>
-+#include <linux/kernel.h>
-+#include <linux/version.h>
-+#include <linux/errno.h>
-+#include <xen/interface/xen.h>
-+#include <xen/interface/platform.h>
-+#include <xen/interface/event_channel.h>
-+#include <xen/interface/physdev.h>
-+#include <xen/interface/sched.h>
-+#include <xen/interface/nmi.h>
-+#include <asm/ptrace.h>
-+#include <asm/page.h>
-+#if defined(__i386__)
-+# ifdef CONFIG_X86_PAE
-+# include <asm-generic/pgtable-nopud.h>
-+# else
-+# include <asm-generic/pgtable-nopmd.h>
-+# endif
-+#elif defined(__x86_64__) && LINUX_VERSION_CODE < KERNEL_VERSION(2,6,11)
-+# include <asm-generic/pgtable-nopud.h>
+#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/setup.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/setup.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,81 @@
++/*
++ * Just a place holder. We don't want to have to test x86 before
++ * we include stuff
++ */
+
-+extern shared_info_t *HYPERVISOR_shared_info;
++#ifndef _i386_SETUP_H
++#define _i386_SETUP_H
+
-+#define vcpu_info(cpu) (HYPERVISOR_shared_info->vcpu_info + (cpu))
-+#ifdef CONFIG_SMP
-+#define current_vcpu_info() vcpu_info(smp_processor_id())
-+#else
-+#define current_vcpu_info() vcpu_info(0)
-+#endif
++#ifdef __KERNEL__
++#include <linux/pfn.h>
+
-+#ifdef CONFIG_X86_32
-+extern unsigned long hypervisor_virt_start;
++/*
++ * Reserved space for vmalloc and iomap - defined in asm/page.h
++ */
++#define MAXMEM_PFN PFN_DOWN(MAXMEM)
++#define MAX_NONPAE_PFN (1 << 20)
+#endif
+
-+/* arch/xen/i386/kernel/setup.c */
-+extern start_info_t *xen_start_info;
-+#ifdef CONFIG_XEN_PRIVILEGED_GUEST
-+#define is_initial_xendomain() (xen_start_info->flags & SIF_INITDOMAIN)
-+#else
-+#define is_initial_xendomain() 0
-+#endif
++#define PARAM_SIZE 4096
++#define COMMAND_LINE_SIZE 256
+
-+/* arch/xen/kernel/evtchn.c */
-+/* Force a proper event-channel callback from Xen. */
-+void force_evtchn_callback(void);
++#define OLD_CL_MAGIC_ADDR 0x90020
++#define OLD_CL_MAGIC 0xA33F
++#define OLD_CL_BASE_ADDR 0x90000
++#define OLD_CL_OFFSET 0x90022
++#define NEW_CL_POINTER 0x228 /* Relative to real mode data */
+
-+/* arch/xen/kernel/process.c */
-+void xen_cpu_idle (void);
++#ifndef __ASSEMBLY__
++/*
++ * This is set up by the setup-routine at boot-time
++ */
++extern unsigned char boot_params[PARAM_SIZE];
+
-+/* arch/xen/i386/kernel/hypervisor.c */
-+void do_hypervisor_callback(struct pt_regs *regs);
++#define PARAM (boot_params)
++#define SCREEN_INFO (*(struct screen_info *) (PARAM+0))
++#define EXT_MEM_K (*(unsigned short *) (PARAM+2))
++#define ALT_MEM_K (*(unsigned long *) (PARAM+0x1e0))
++#define E820_MAP_NR (*(char*) (PARAM+E820NR))
++#define E820_MAP ((struct e820entry *) (PARAM+E820MAP))
++#define APM_BIOS_INFO (*(struct apm_bios_info *) (PARAM+0x40))
++#define IST_INFO (*(struct ist_info *) (PARAM+0x60))
++#define DRIVE_INFO (*(struct drive_info_struct *) (PARAM+0x80))
++#define SYS_DESC_TABLE (*(struct sys_desc_table_struct*)(PARAM+0xa0))
++#define EFI_SYSTAB ((efi_system_table_t *) *((unsigned long *)(PARAM+0x1c4)))
++#define EFI_MEMDESC_SIZE (*((unsigned long *) (PARAM+0x1c8)))
++#define EFI_MEMDESC_VERSION (*((unsigned long *) (PARAM+0x1cc)))
++#define EFI_MEMMAP ((void *) *((unsigned long *)(PARAM+0x1d0)))
++#define EFI_MEMMAP_SIZE (*((unsigned long *) (PARAM+0x1d4)))
++#define MOUNT_ROOT_RDONLY (*(unsigned short *) (PARAM+0x1F2))
++#define RAMDISK_FLAGS (*(unsigned short *) (PARAM+0x1F8))
++#define VIDEO_MODE (*(unsigned short *) (PARAM+0x1FA))
++#define ORIG_ROOT_DEV (*(unsigned short *) (PARAM+0x1FC))
++#define AUX_DEVICE_INFO (*(unsigned char *) (PARAM+0x1FF))
++#define LOADER_TYPE (*(unsigned char *) (PARAM+0x210))
++#define KERNEL_START (*(unsigned long *) (PARAM+0x214))
++#define INITRD_START (__pa(xen_start_info->mod_start))
++#define INITRD_SIZE (xen_start_info->mod_len)
++#define EDID_INFO (*(struct edid_info *) (PARAM+0x440))
++#define EDD_NR (*(unsigned char *) (PARAM+EDDNR))
++#define EDD_MBR_SIG_NR (*(unsigned char *) (PARAM+EDD_MBR_SIG_NR_BUF))
++#define EDD_MBR_SIGNATURE ((unsigned int *) (PARAM+EDD_MBR_SIG_BUF))
++#define EDD_BUF ((struct edd_info *) (PARAM+EDDBUF))
+
-+/* arch/xen/i386/mm/hypervisor.c */
+/*
-+ * NB. ptr values should be PHYSICAL, not MACHINE. 'vals' should be already
-+ * be MACHINE addresses.
++ * Do NOT EVER look at the BIOS memory size location.
++ * It does not work on many machines.
+ */
++#define LOWMEMSIZE() (0x9f000)
+
-+void xen_pt_switch(unsigned long ptr);
-+void xen_new_user_pt(unsigned long ptr); /* x86_64 only */
-+void xen_load_gs(unsigned int selector); /* x86_64 only */
-+void xen_tlb_flush(void);
-+void xen_invlpg(unsigned long ptr);
++struct e820entry;
+
-+void xen_l1_entry_update(pte_t *ptr, pte_t val);
-+void xen_l2_entry_update(pmd_t *ptr, pmd_t val);
-+void xen_l3_entry_update(pud_t *ptr, pud_t val); /* x86_64/PAE */
-+void xen_l4_entry_update(pgd_t *ptr, pgd_t val); /* x86_64 only */
-+void xen_pgd_pin(unsigned long ptr);
-+void xen_pgd_unpin(unsigned long ptr);
++char * __init machine_specific_memory_setup(void);
+
-+void xen_set_ldt(unsigned long ptr, unsigned long bytes);
++int __init copy_e820_map(struct e820entry * biosmap, int nr_map);
++int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map);
++void __init add_memory_region(unsigned long long start,
++ unsigned long long size, int type);
+
-+#ifdef CONFIG_SMP
++#endif /* __ASSEMBLY__ */
++
++#endif /* _i386_SETUP_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/smp.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/smp.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,103 @@
++#ifndef __ASM_SMP_H
++#define __ASM_SMP_H
++
++/*
++ * We need the APIC definitions automatically as part of 'smp.h'
++ */
++#ifndef __ASSEMBLY__
++#include <linux/kernel.h>
++#include <linux/threads.h>
+#include <linux/cpumask.h>
-+void xen_tlb_flush_all(void);
-+void xen_invlpg_all(unsigned long ptr);
-+void xen_tlb_flush_mask(cpumask_t *mask);
-+void xen_invlpg_mask(cpumask_t *mask, unsigned long ptr);
+#endif
+
-+/* Returns zero on success else negative errno. */
-+int xen_create_contiguous_region(
-+ unsigned long vstart, unsigned int order, unsigned int address_bits);
-+void xen_destroy_contiguous_region(
-+ unsigned long vstart, unsigned int order);
++#ifdef CONFIG_X86_LOCAL_APIC
++#ifndef __ASSEMBLY__
++#include <asm/fixmap.h>
++#include <asm/bitops.h>
++#include <asm/mpspec.h>
++#ifdef CONFIG_X86_IO_APIC
++#include <asm/io_apic.h>
++#endif
++#include <asm/apic.h>
++#endif
++#endif
+
-+struct page;
++#define BAD_APICID 0xFFu
++#ifdef CONFIG_SMP
++#ifndef __ASSEMBLY__
+
-+int xen_limit_pages_to_max_mfn(
-+ struct page *pages, unsigned int order, unsigned int address_bits);
++/*
++ * Private routines/data
++ */
++
++extern void smp_alloc_memory(void);
++extern int pic_mode;
++extern int smp_num_siblings;
++extern cpumask_t cpu_sibling_map[];
++extern cpumask_t cpu_core_map[];
+
-+/* Turn jiffies into Xen system time. */
-+u64 jiffies_to_st(unsigned long jiffies);
++extern void (*mtrr_hook) (void);
++extern void zap_low_mappings (void);
++extern void lock_ipi_call_lock(void);
++extern void unlock_ipi_call_lock(void);
+
-+#ifdef CONFIG_XEN_SCRUB_PAGES
-+#define scrub_pages(_p,_n) memset((void *)(_p), 0, (_n) << PAGE_SHIFT)
-+#else
-+#define scrub_pages(_p,_n) ((void)0)
-+#endif
++#define MAX_APICID 256
++extern u8 x86_cpu_to_apicid[];
+
-+#include <xen/hypercall.h>
++#define cpu_physical_id(cpu) x86_cpu_to_apicid[cpu]
+
-+#if defined(CONFIG_X86_64)
-+#define MULTI_UVMFLAGS_INDEX 2
-+#define MULTI_UVMDOMID_INDEX 3
-+#else
-+#define MULTI_UVMFLAGS_INDEX 3
-+#define MULTI_UVMDOMID_INDEX 4
++#ifdef CONFIG_HOTPLUG_CPU
++extern void cpu_exit_clear(void);
++extern void cpu_uninit(void);
+#endif
+
-+#define is_running_on_xen() 1
-+
-+static inline int
-+HYPERVISOR_yield(
-+ void)
-+{
-+ int rc = HYPERVISOR_sched_op(SCHEDOP_yield, NULL);
-+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (rc == -ENOSYS)
-+ rc = HYPERVISOR_sched_op_compat(SCHEDOP_yield, 0);
-+#endif
++/*
++ * This function is needed by all SMP systems. It must _always_ be valid
++ * from the initial startup. We map APIC_BASE very early in page_setup(),
++ * so this is correct in the x86 case.
++ */
++#define raw_smp_processor_id() (current_thread_info()->cpu)
+
-+ return rc;
-+}
++extern cpumask_t cpu_possible_map;
++#define cpu_callin_map cpu_possible_map
+
-+static inline int
-+HYPERVISOR_block(
-+ void)
++/* We don't mark CPUs online until __cpu_up(), so we need another measure */
++static inline int num_booting_cpus(void)
+{
-+ int rc = HYPERVISOR_sched_op(SCHEDOP_block, NULL);
++ return cpus_weight(cpu_possible_map);
++}
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (rc == -ENOSYS)
-+ rc = HYPERVISOR_sched_op_compat(SCHEDOP_block, 0);
-+#endif
++#ifdef CONFIG_X86_LOCAL_APIC
+
-+ return rc;
++#ifdef APIC_DEFINITION
++extern int hard_smp_processor_id(void);
++#else
++#include <mach_apicdef.h>
++static inline int hard_smp_processor_id(void)
++{
++ /* we don't want to mark this access volatile - bad code generation */
++ return GET_APIC_ID(*(unsigned long *)(APIC_BASE+APIC_ID));
+}
++#endif
+
-+static inline int
-+HYPERVISOR_shutdown(
-+ unsigned int reason)
++static __inline int logical_smp_processor_id(void)
+{
-+ struct sched_shutdown sched_shutdown = {
-+ .reason = reason
-+ };
-+
-+ int rc = HYPERVISOR_sched_op(SCHEDOP_shutdown, &sched_shutdown);
++ /* we don't want to mark this access volatile - bad code generation */
++ return GET_APIC_LOGICAL_ID(*(unsigned long *)(APIC_BASE+APIC_LDR));
++}
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (rc == -ENOSYS)
-+ rc = HYPERVISOR_sched_op_compat(SCHEDOP_shutdown, reason);
+#endif
+
-+ return rc;
-+}
++extern int __cpu_disable(void);
++extern void __cpu_die(unsigned int cpu);
++extern void prefill_possible_map(void);
++#endif /* !__ASSEMBLY__ */
+
-+static inline int
-+HYPERVISOR_poll(
-+ evtchn_port_t *ports, unsigned int nr_ports, u64 timeout)
-+{
-+ int rc;
-+ struct sched_poll sched_poll = {
-+ .nr_ports = nr_ports,
-+ .timeout = jiffies_to_st(timeout)
-+ };
-+ set_xen_guest_handle(sched_poll.ports, ports);
++#else /* CONFIG_SMP */
+
-+ rc = HYPERVISOR_sched_op(SCHEDOP_poll, &sched_poll);
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (rc == -ENOSYS)
-+ rc = HYPERVISOR_sched_op_compat(SCHEDOP_yield, 0);
-+#endif
++#define cpu_physical_id(cpu) boot_cpu_physical_apicid
+
-+ return rc;
-+}
++#define NO_PROC_ID 0xFF /* No processor magic marker */
+
-+static inline void
-+MULTI_update_va_mapping(
-+ multicall_entry_t *mcl, unsigned long va,
-+ pte_t new_val, unsigned long flags)
-+{
-+ mcl->op = __HYPERVISOR_update_va_mapping;
-+ mcl->args[0] = va;
-+#if defined(CONFIG_X86_64)
-+ mcl->args[1] = new_val.pte;
-+#elif defined(CONFIG_X86_PAE)
-+ mcl->args[1] = new_val.pte_low;
-+ mcl->args[2] = new_val.pte_high;
-+#else
-+ mcl->args[1] = new_val.pte_low;
-+ mcl->args[2] = 0;
+#endif
-+ mcl->args[MULTI_UVMFLAGS_INDEX] = flags;
-+}
-+
-+static inline void
-+MULTI_grant_table_op(multicall_entry_t *mcl, unsigned int cmd,
-+ void *uop, unsigned int count)
-+{
-+ mcl->op = __HYPERVISOR_grant_table_op;
-+ mcl->args[0] = cmd;
-+ mcl->args[1] = (unsigned long)uop;
-+ mcl->args[2] = count;
-+}
-+
-+static inline void
-+MULTI_update_va_mapping_otherdomain(
-+ multicall_entry_t *mcl, unsigned long va,
-+ pte_t new_val, unsigned long flags, domid_t domid)
-+{
-+ mcl->op = __HYPERVISOR_update_va_mapping_otherdomain;
-+ mcl->args[0] = va;
-+#if defined(CONFIG_X86_64)
-+ mcl->args[1] = new_val.pte;
-+#elif defined(CONFIG_X86_PAE)
-+ mcl->args[1] = new_val.pte_low;
-+ mcl->args[2] = new_val.pte_high;
-+#else
-+ mcl->args[1] = new_val.pte_low;
-+ mcl->args[2] = 0;
+#endif
-+ mcl->args[MULTI_UVMFLAGS_INDEX] = flags;
-+ mcl->args[MULTI_UVMDOMID_INDEX] = domid;
-+}
-+
-+#endif /* __HYPERVISOR_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/io.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/io.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/io.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/io.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,390 @@
-+#ifndef _ASM_IO_H
-+#define _ASM_IO_H
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/spinlock.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/spinlock.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,202 @@
++#ifndef __ASM_SPINLOCK_H
++#define __ASM_SPINLOCK_H
+
-+#include <linux/string.h>
++#include <asm/atomic.h>
++#include <asm/rwlock.h>
++#include <asm/page.h>
+#include <linux/compiler.h>
+
+/*
-+ * This file contains the definitions for the x86 IO instructions
-+ * inb/inw/inl/outb/outw/outl and the "string versions" of the same
-+ * (insb/insw/insl/outsb/outsw/outsl). You can also use "pausing"
-+ * versions of the single-IO instructions (inb_p/inw_p/..).
++ * Your basic SMP spinlocks, allowing only a single CPU anywhere
+ *
-+ * This file is not meant to be obfuscating: it's just complicated
-+ * to (a) handle it all in a way that makes gcc able to optimize it
-+ * as well as possible and (b) trying to avoid writing the same thing
-+ * over and over again with slight variations and possibly making a
-+ * mistake somewhere.
-+ */
-+
-+/*
-+ * Thanks to James van Artsdalen for a better timing-fix than
-+ * the two short jumps: using outb's to a nonexistent port seems
-+ * to guarantee better timings even on fast machines.
++ * Simple spin lock operations. There are two variants, one clears IRQ's
++ * on the local processor, one does not.
+ *
-+ * On the other hand, I'd like to be sure of a non-existent port:
-+ * I feel a bit unsafe about using 0x80 (should be safe, though)
++ * We make no fairness assumptions. They have a cost.
+ *
-+ * Linus
++ * (the type definitions are in asm/spinlock_types.h)
+ */
+
-+ /*
-+ * Bit simplified and optimized by Jan Hubicka
-+ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999.
-+ *
-+ * isa_memset_io, isa_memcpy_fromio, isa_memcpy_toio added,
-+ * isa_read[wl] and isa_write[wl] fixed
-+ * - Arnaldo Carvalho de Melo <acme at conectiva.com.br>
-+ */
-+
-+#define IO_SPACE_LIMIT 0xffff
-+
-+#define XQUAD_PORTIO_BASE 0xfe400000
-+#define XQUAD_PORTIO_QUAD 0x40000 /* 256k per quad. */
-+
-+#ifdef __KERNEL__
-+
-+#include <asm-generic/iomap.h>
-+
-+#include <linux/vmalloc.h>
-+#include <asm/fixmap.h>
++#define __raw_spin_is_locked(x) \
++ (*(volatile signed char *)(&(x)->slock) <= 0)
+
-+/*
-+ * Convert a physical pointer to a virtual kernel pointer for /dev/mem
-+ * access
-+ */
-+#define xlate_dev_mem_ptr(p, sz) ioremap(p, sz)
-+#define xlate_dev_mem_ptr_unmap(p) iounmap(p)
++#define __raw_spin_lock_string \
++ "\n1:\t" \
++ LOCK_PREFIX " ; decb %0\n\t" \
++ "jns 3f\n" \
++ "2:\t" \
++ "rep;nop\n\t" \
++ "cmpb $0,%0\n\t" \
++ "jle 2b\n\t" \
++ "jmp 1b\n" \
++ "3:\n\t"
+
+/*
-+ * Convert a virtual cached pointer to an uncached pointer
++ * NOTE: there's an irqs-on section here, which normally would have to be
++ * irq-traced, but on CONFIG_TRACE_IRQFLAGS we never use
++ * __raw_spin_lock_string_flags().
+ */
-+#define xlate_dev_kmem_ptr(p) p
++#define __raw_spin_lock_string_flags \
++ "\n1:\t" \
++ LOCK_PREFIX " ; decb %0\n\t" \
++ "jns 5f\n" \
++ "2:\t" \
++ "testl $0x200, %1\n\t" \
++ "jz 4f\n\t" \
++ "#sti\n" \
++ "3:\t" \
++ "rep;nop\n\t" \
++ "cmpb $0, %0\n\t" \
++ "jle 3b\n\t" \
++ "#cli\n\t" \
++ "jmp 1b\n" \
++ "4:\t" \
++ "rep;nop\n\t" \
++ "cmpb $0, %0\n\t" \
++ "jg 1b\n\t" \
++ "jmp 4b\n" \
++ "5:\n\t"
+
-+/**
-+ * virt_to_phys - map virtual addresses to physical
-+ * @address: address to remap
-+ *
-+ * The returned physical address is the physical (CPU) mapping for
-+ * the memory address given. It is only valid to use this function on
-+ * addresses directly mapped or allocated via kmalloc.
-+ *
-+ * This function does not give bus mappings for DMA transfers. In
-+ * almost all conceivable cases a device driver should not be using
-+ * this function
-+ */
-+
-+static inline unsigned long virt_to_phys(volatile void * address)
++static inline void __raw_spin_lock(raw_spinlock_t *lock)
+{
-+ return __pa(address);
++ asm(__raw_spin_lock_string : "+m" (lock->slock) : : "memory");
+}
+
-+/**
-+ * phys_to_virt - map physical address to virtual
-+ * @address: address to remap
-+ *
-+ * The returned virtual address is a current CPU mapping for
-+ * the memory address given. It is only valid to use this function on
-+ * addresses that have a kernel mapping
-+ *
-+ * This function does not handle bus mappings for DMA transfers. In
-+ * almost all conceivable cases a device driver should not be using
-+ * this function
++/*
++ * It is easier for the lock validator if interrupts are not re-enabled
++ * in the middle of a lock-acquire. This is a performance feature anyway
++ * so we turn it off:
+ */
++#ifndef CONFIG_PROVE_LOCKING
++static inline void __raw_spin_lock_flags(raw_spinlock_t *lock, unsigned long flags)
++{
++ asm(__raw_spin_lock_string_flags : "+m" (lock->slock) : "r" (flags) : "memory");
++}
++#endif
+
-+static inline void * phys_to_virt(unsigned long address)
++static inline int __raw_spin_trylock(raw_spinlock_t *lock)
+{
-+ return __va(address);
++ char oldval;
++ __asm__ __volatile__(
++ "xchgb %b0,%1"
++ :"=q" (oldval), "+m" (lock->slock)
++ :"0" (0) : "memory");
++ return oldval > 0;
+}
+
+/*
-+ * Change "struct page" to physical address.
++ * __raw_spin_unlock based on writing $1 to the low byte.
++ * This method works. Despite all the confusion.
++ * (except on PPro SMP or if we are using OOSTORE, so we use xchgb there)
++ * (PPro errata 66, 92)
+ */
-+#define page_to_pseudophys(page) ((dma_addr_t)page_to_pfn(page) << PAGE_SHIFT)
-+#define page_to_phys(page) (phys_to_machine(page_to_pseudophys(page)))
-+#define page_to_bus(page) (phys_to_machine(page_to_pseudophys(page)))
-+
-+#define bio_to_pseudophys(bio) (page_to_pseudophys(bio_page((bio))) + \
-+ (unsigned long) bio_offset((bio)))
-+#define bvec_to_pseudophys(bv) (page_to_pseudophys((bv)->bv_page) + \
-+ (unsigned long) (bv)->bv_offset)
+
-+#define BIOVEC_PHYS_MERGEABLE(vec1, vec2) \
-+ (((bvec_to_phys((vec1)) + (vec1)->bv_len) == bvec_to_phys((vec2))) && \
-+ ((bvec_to_pseudophys((vec1)) + (vec1)->bv_len) == \
-+ bvec_to_pseudophys((vec2))))
++#if !defined(CONFIG_X86_OOSTORE) && !defined(CONFIG_X86_PPRO_FENCE)
+
-+extern void __iomem * __ioremap(unsigned long offset, unsigned long size, unsigned long flags);
++#define __raw_spin_unlock_string \
++ "movb $1,%0" \
++ :"+m" (lock->slock) : : "memory"
+
-+/**
-+ * ioremap - map bus memory into CPU space
-+ * @offset: bus address of the memory
-+ * @size: size of the resource to map
-+ *
-+ * ioremap performs a platform specific sequence of operations to
-+ * make bus memory CPU accessible via the readb/readw/readl/writeb/
-+ * writew/writel functions and the other mmio helpers. The returned
-+ * address is not guaranteed to be usable directly as a virtual
-+ * address.
-+ */
+
-+static inline void __iomem * ioremap(unsigned long offset, unsigned long size)
++static inline void __raw_spin_unlock(raw_spinlock_t *lock)
+{
-+ return __ioremap(offset, size, 0);
++ __asm__ __volatile__(
++ __raw_spin_unlock_string
++ );
+}
+
-+extern void __iomem * ioremap_nocache(unsigned long offset, unsigned long size);
-+extern void iounmap(volatile void __iomem *addr);
++#else
+
-+/*
-+ * bt_ioremap() and bt_iounmap() are for temporary early boot-time
-+ * mappings, before the real ioremap() is functional.
-+ * A boot-time mapping is currently limited to at most 16 pages.
-+ */
-+extern void *bt_ioremap(unsigned long offset, unsigned long size);
-+extern void bt_iounmap(void *addr, unsigned long size);
++#define __raw_spin_unlock_string \
++ "xchgb %b0, %1" \
++ :"=q" (oldval), "+m" (lock->slock) \
++ :"0" (oldval) : "memory"
+
-+/* Use early IO mappings for DMI because it's initialized early */
-+#define dmi_ioremap bt_ioremap
-+#define dmi_iounmap bt_iounmap
-+#define dmi_alloc alloc_bootmem
++static inline void __raw_spin_unlock(raw_spinlock_t *lock)
++{
++ char oldval = 1;
+
-+/*
-+ * ISA I/O bus memory addresses are 1:1 with the physical address.
-+ */
-+#define isa_virt_to_bus(_x) isa_virt_to_bus_is_UNSUPPORTED->x
-+#define isa_page_to_bus(_x) isa_page_to_bus_is_UNSUPPORTED->x
-+#define isa_bus_to_virt(_x) (void *)(__fix_to_virt(FIX_ISAMAP_BEGIN) + (_x))
++ __asm__ __volatile__(
++ __raw_spin_unlock_string
++ );
++}
++
++#endif
++
++#define __raw_spin_unlock_wait(lock) \
++ do { while (__raw_spin_is_locked(lock)) cpu_relax(); } while (0)
+
+/*
-+ * However PCI ones are not necessarily 1:1 and therefore these interfaces
-+ * are forbidden in portable PCI drivers.
++ * Read-write spinlocks, allowing multiple readers
++ * but only one writer.
+ *
-+ * Allow them on x86 for legacy drivers, though.
++ * NOTE! it is quite common to have readers in interrupts
++ * but no interrupt writers. For those circumstances we
++ * can "mix" irq-safe locks - any writer needs to get a
++ * irq-safe write-lock, but readers can get non-irqsafe
++ * read-locks.
++ *
++ * On x86, we implement read-write locks as a 32-bit counter
++ * with the high bit (sign) being the "contended" bit.
++ *
++ * The inline assembly is non-obvious. Think about it.
++ *
++ * Changed to use the same technique as rw semaphores. See
++ * semaphore.h for details. -ben
++ *
++ * the helpers are in arch/i386/kernel/semaphore.c
+ */
-+#define virt_to_bus(_x) phys_to_machine(__pa(_x))
-+#define bus_to_virt(_x) __va(machine_to_phys(_x))
+
-+/*
-+ * readX/writeX() are used to access memory mapped devices. On some
-+ * architectures the memory mapped IO stuff needs to be accessed
-+ * differently. On the x86 architecture, we just read/write the
-+ * memory location directly.
++/**
++ * read_can_lock - would read_trylock() succeed?
++ * @lock: the rwlock in question.
+ */
++#define __raw_read_can_lock(x) ((int)(x)->lock > 0)
+
-+static inline unsigned char readb(const volatile void __iomem *addr)
-+{
-+ return *(volatile unsigned char __force *) addr;
-+}
-+static inline unsigned short readw(const volatile void __iomem *addr)
-+{
-+ return *(volatile unsigned short __force *) addr;
-+}
-+static inline unsigned int readl(const volatile void __iomem *addr)
-+{
-+ return *(volatile unsigned int __force *) addr;
-+}
-+#define readb_relaxed(addr) readb(addr)
-+#define readw_relaxed(addr) readw(addr)
-+#define readl_relaxed(addr) readl(addr)
-+#define __raw_readb readb
-+#define __raw_readw readw
-+#define __raw_readl readl
++/**
++ * write_can_lock - would write_trylock() succeed?
++ * @lock: the rwlock in question.
++ */
++#define __raw_write_can_lock(x) ((x)->lock == RW_LOCK_BIAS)
+
-+static inline void writeb(unsigned char b, volatile void __iomem *addr)
-+{
-+ *(volatile unsigned char __force *) addr = b;
-+}
-+static inline void writew(unsigned short b, volatile void __iomem *addr)
++static inline void __raw_read_lock(raw_rwlock_t *rw)
+{
-+ *(volatile unsigned short __force *) addr = b;
++ __build_read_lock(rw, "__read_lock_failed");
+}
-+static inline void writel(unsigned int b, volatile void __iomem *addr)
++
++static inline void __raw_write_lock(raw_rwlock_t *rw)
+{
-+ *(volatile unsigned int __force *) addr = b;
++ __build_write_lock(rw, "__write_lock_failed");
+}
-+#define __raw_writeb writeb
-+#define __raw_writew writew
-+#define __raw_writel writel
-+
-+#define mmiowb()
+
-+static inline void memset_io(volatile void __iomem *addr, unsigned char val, int count)
++static inline int __raw_read_trylock(raw_rwlock_t *lock)
+{
-+ memset((void __force *) addr, val, count);
++ atomic_t *count = (atomic_t *)lock;
++ atomic_dec(count);
++ if (atomic_read(count) >= 0)
++ return 1;
++ atomic_inc(count);
++ return 0;
+}
-+static inline void memcpy_fromio(void *dst, const volatile void __iomem *src, int count)
++
++static inline int __raw_write_trylock(raw_rwlock_t *lock)
+{
-+ __memcpy(dst, (void __force *) src, count);
++ atomic_t *count = (atomic_t *)lock;
++ if (atomic_sub_and_test(RW_LOCK_BIAS, count))
++ return 1;
++ atomic_add(RW_LOCK_BIAS, count);
++ return 0;
+}
-+static inline void memcpy_toio(volatile void __iomem *dst, const void *src, int count)
++
++static inline void __raw_read_unlock(raw_rwlock_t *rw)
+{
-+ __memcpy((void __force *) dst, src, count);
++ asm volatile(LOCK_PREFIX "incl %0" :"+m" (rw->lock) : : "memory");
+}
+
-+/*
-+ * ISA space is 'always mapped' on a typical x86 system, no need to
-+ * explicitly ioremap() it. The fact that the ISA IO space is mapped
-+ * to PAGE_OFFSET is pure coincidence - it does not mean ISA values
-+ * are physical addresses. The following constant pointer can be
-+ * used as the IO-area pointer (it can be iounmapped as well, so the
-+ * analogy with PCI is quite large):
-+ */
-+#define __ISA_IO_base ((char __iomem *)(fix_to_virt(FIX_ISAMAP_BEGIN)))
-+
-+/*
-+ * Again, i386 does not require mem IO specific function.
-+ */
-+
-+#define eth_io_copy_and_sum(a,b,c,d) eth_copy_and_sum((a),(void __force *)(b),(c),(d))
-+
-+/**
-+ * check_signature - find BIOS signatures
-+ * @io_addr: mmio address to check
-+ * @signature: signature block
-+ * @length: length of signature
-+ *
-+ * Perform a signature comparison with the mmio address io_addr. This
-+ * address should have been obtained by ioremap.
-+ * Returns 1 on a match.
-+ */
-+
-+static inline int check_signature(volatile void __iomem * io_addr,
-+ const unsigned char *signature, int length)
++static inline void __raw_write_unlock(raw_rwlock_t *rw)
+{
-+ int retval = 0;
-+ do {
-+ if (readb(io_addr) != *signature)
-+ goto out;
-+ io_addr++;
-+ signature++;
-+ length--;
-+ } while (length);
-+ retval = 1;
-+out:
-+ return retval;
++ asm volatile(LOCK_PREFIX "addl $" RW_LOCK_BIAS_STR ", %0"
++ : "+m" (rw->lock) : : "memory");
+}
+
-+/*
-+ * Cache management
-+ *
-+ * This needed for two cases
-+ * 1. Out of order aware processors
-+ * 2. Accidentally out of order processors (PPro errata #51)
-+ */
-+
-+#if defined(CONFIG_X86_OOSTORE) || defined(CONFIG_X86_PPRO_FENCE)
++#endif /* __ASM_SPINLOCK_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/swiotlb.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/swiotlb.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,43 @@
++#ifndef _ASM_SWIOTLB_H
++#define _ASM_SWIOTLB_H 1
+
-+static inline void flush_write_buffers(void)
-+{
-+ __asm__ __volatile__ ("lock; addl $0,0(%%esp)": : :"memory");
-+}
++/* SWIOTLB interface */
+
-+#define dma_cache_inv(_start,_size) flush_write_buffers()
-+#define dma_cache_wback(_start,_size) flush_write_buffers()
-+#define dma_cache_wback_inv(_start,_size) flush_write_buffers()
++extern dma_addr_t swiotlb_map_single(struct device *hwdev, void *ptr, size_t size,
++ int dir);
++extern void swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr,
++ size_t size, int dir);
++extern void swiotlb_sync_single_for_cpu(struct device *hwdev,
++ dma_addr_t dev_addr,
++ size_t size, int dir);
++extern void swiotlb_sync_single_for_device(struct device *hwdev,
++ dma_addr_t dev_addr,
++ size_t size, int dir);
++extern void swiotlb_sync_sg_for_cpu(struct device *hwdev,
++ struct scatterlist *sg, int nelems,
++ int dir);
++extern void swiotlb_sync_sg_for_device(struct device *hwdev,
++ struct scatterlist *sg, int nelems,
++ int dir);
++extern int swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg,
++ int nents, int direction);
++extern void swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg,
++ int nents, int direction);
++extern int swiotlb_dma_mapping_error(dma_addr_t dma_addr);
++#ifdef CONFIG_HIGHMEM
++extern dma_addr_t swiotlb_map_page(struct device *hwdev, struct page *page,
++ unsigned long offset, size_t size,
++ enum dma_data_direction direction);
++extern void swiotlb_unmap_page(struct device *hwdev, dma_addr_t dma_address,
++ size_t size, enum dma_data_direction direction);
++#endif
++extern int swiotlb_dma_supported(struct device *hwdev, u64 mask);
++extern void swiotlb_init(void);
+
++#ifdef CONFIG_SWIOTLB
++extern int swiotlb;
+#else
++#define swiotlb 0
++#endif
+
-+/* Nothing to do */
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/synch_bitops.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/synch_bitops.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,126 @@
++#ifndef __XEN_SYNCH_BITOPS_H__
++#define __XEN_SYNCH_BITOPS_H__
+
-+#define dma_cache_inv(_start,_size) do { } while (0)
-+#define dma_cache_wback(_start,_size) do { } while (0)
-+#define dma_cache_wback_inv(_start,_size) do { } while (0)
-+#define flush_write_buffers()
++/*
++ * Copyright 1992, Linus Torvalds.
++ * Heavily modified to provide guaranteed strong synchronisation
++ * when communicating with Xen or other guest OSes running on other CPUs.
++ */
+
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
+#endif
+
-+#endif /* __KERNEL__ */
++#define ADDR (*(volatile long *) addr)
+
-+#ifdef SLOW_IO_BY_JUMPING
-+#define __SLOW_DOWN_IO "jmp 1f; 1: jmp 1f; 1:"
-+#else
-+#define __SLOW_DOWN_IO "outb %%al,$0x80;"
-+#endif
++static __inline__ void synch_set_bit(int nr, volatile void * addr)
++{
++ __asm__ __volatile__ (
++ "lock btsl %1,%0"
++ : "+m" (ADDR) : "Ir" (nr) : "memory" );
++}
+
-+static inline void slow_down_io(void) {
-+ __asm__ __volatile__(
-+ __SLOW_DOWN_IO
-+#ifdef REALLY_SLOW_IO
-+ __SLOW_DOWN_IO __SLOW_DOWN_IO __SLOW_DOWN_IO
-+#endif
-+ : : );
++static __inline__ void synch_clear_bit(int nr, volatile void * addr)
++{
++ __asm__ __volatile__ (
++ "lock btrl %1,%0"
++ : "+m" (ADDR) : "Ir" (nr) : "memory" );
+}
+
-+#ifdef CONFIG_X86_NUMAQ
-+extern void *xquad_portio; /* Where the IO area was mapped */
-+#define XQUAD_PORT_ADDR(port, quad) (xquad_portio + (XQUAD_PORTIO_QUAD*quad) + port)
-+#define __BUILDIO(bwl,bw,type) \
-+static inline void out##bwl##_quad(unsigned type value, int port, int quad) { \
-+ if (xquad_portio) \
-+ write##bwl(value, XQUAD_PORT_ADDR(port, quad)); \
-+ else \
-+ out##bwl##_local(value, port); \
-+} \
-+static inline void out##bwl(unsigned type value, int port) { \
-+ out##bwl##_quad(value, port, 0); \
-+} \
-+static inline unsigned type in##bwl##_quad(int port, int quad) { \
-+ if (xquad_portio) \
-+ return read##bwl(XQUAD_PORT_ADDR(port, quad)); \
-+ else \
-+ return in##bwl##_local(port); \
-+} \
-+static inline unsigned type in##bwl(int port) { \
-+ return in##bwl##_quad(port, 0); \
++static __inline__ void synch_change_bit(int nr, volatile void * addr)
++{
++ __asm__ __volatile__ (
++ "lock btcl %1,%0"
++ : "+m" (ADDR) : "Ir" (nr) : "memory" );
+}
-+#else
-+#define __BUILDIO(bwl,bw,type) \
-+static inline void out##bwl(unsigned type value, int port) { \
-+ out##bwl##_local(value, port); \
-+} \
-+static inline unsigned type in##bwl(int port) { \
-+ return in##bwl##_local(port); \
++
++static __inline__ int synch_test_and_set_bit(int nr, volatile void * addr)
++{
++ int oldbit;
++ __asm__ __volatile__ (
++ "lock btsl %2,%1\n\tsbbl %0,%0"
++ : "=r" (oldbit), "+m" (ADDR) : "Ir" (nr) : "memory");
++ return oldbit;
+}
-+#endif
+
++static __inline__ int synch_test_and_clear_bit(int nr, volatile void * addr)
++{
++ int oldbit;
++ __asm__ __volatile__ (
++ "lock btrl %2,%1\n\tsbbl %0,%0"
++ : "=r" (oldbit), "+m" (ADDR) : "Ir" (nr) : "memory");
++ return oldbit;
++}
+
-+#define BUILDIO(bwl,bw,type) \
-+static inline void out##bwl##_local(unsigned type value, int port) { \
-+ __asm__ __volatile__("out" #bwl " %" #bw "0, %w1" : : "a"(value), "Nd"(port)); \
-+} \
-+static inline unsigned type in##bwl##_local(int port) { \
-+ unsigned type value; \
-+ __asm__ __volatile__("in" #bwl " %w1, %" #bw "0" : "=a"(value) : "Nd"(port)); \
-+ return value; \
-+} \
-+static inline void out##bwl##_local_p(unsigned type value, int port) { \
-+ out##bwl##_local(value, port); \
-+ slow_down_io(); \
-+} \
-+static inline unsigned type in##bwl##_local_p(int port) { \
-+ unsigned type value = in##bwl##_local(port); \
-+ slow_down_io(); \
-+ return value; \
-+} \
-+__BUILDIO(bwl,bw,type) \
-+static inline void out##bwl##_p(unsigned type value, int port) { \
-+ out##bwl(value, port); \
-+ slow_down_io(); \
-+} \
-+static inline unsigned type in##bwl##_p(int port) { \
-+ unsigned type value = in##bwl(port); \
-+ slow_down_io(); \
-+ return value; \
-+} \
-+static inline void outs##bwl(int port, const void *addr, unsigned long count) { \
-+ __asm__ __volatile__("rep; outs" #bwl : "+S"(addr), "+c"(count) : "d"(port)); \
-+} \
-+static inline void ins##bwl(int port, void *addr, unsigned long count) { \
-+ __asm__ __volatile__("rep; ins" #bwl : "+D"(addr), "+c"(count) : "d"(port)); \
++static __inline__ int synch_test_and_change_bit(int nr, volatile void * addr)
++{
++ int oldbit;
++
++ __asm__ __volatile__ (
++ "lock btcl %2,%1\n\tsbbl %0,%0"
++ : "=r" (oldbit), "+m" (ADDR) : "Ir" (nr) : "memory");
++ return oldbit;
+}
+
-+BUILDIO(b,b,char)
-+BUILDIO(w,w,short)
-+BUILDIO(l,,int)
++struct __synch_xchg_dummy { unsigned long a[100]; };
++#define __synch_xg(x) ((struct __synch_xchg_dummy *)(x))
+
-+/* We will be supplying our own /dev/mem implementation */
-+#define ARCH_HAS_DEV_MEM
++#define synch_cmpxchg(ptr, old, new) \
++((__typeof__(*(ptr)))__synch_cmpxchg((ptr),\
++ (unsigned long)(old), \
++ (unsigned long)(new), \
++ sizeof(*(ptr))))
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/irqflags.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/irqflags.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/irqflags.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/irqflags.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,127 @@
-+/*
-+ * include/asm-i386/irqflags.h
-+ *
-+ * IRQ flags handling
-+ *
-+ * This file gets included from lowlevel asm headers too, to provide
-+ * wrapped versions of the local_irq_*() APIs, based on the
-+ * raw_local_irq_*() functions from the lowlevel headers.
-+ */
-+#ifndef _ASM_IRQFLAGS_H
-+#define _ASM_IRQFLAGS_H
++static inline unsigned long __synch_cmpxchg(volatile void *ptr,
++ unsigned long old,
++ unsigned long new, int size)
++{
++ unsigned long prev;
++ switch (size) {
++ case 1:
++ __asm__ __volatile__("lock; cmpxchgb %b1,%2"
++ : "=a"(prev)
++ : "q"(new), "m"(*__synch_xg(ptr)),
++ "0"(old)
++ : "memory");
++ return prev;
++ case 2:
++ __asm__ __volatile__("lock; cmpxchgw %w1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__synch_xg(ptr)),
++ "0"(old)
++ : "memory");
++ return prev;
++#ifdef CONFIG_X86_64
++ case 4:
++ __asm__ __volatile__("lock; cmpxchgl %k1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__synch_xg(ptr)),
++ "0"(old)
++ : "memory");
++ return prev;
++ case 8:
++ __asm__ __volatile__("lock; cmpxchgq %1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__synch_xg(ptr)),
++ "0"(old)
++ : "memory");
++ return prev;
++#else
++ case 4:
++ __asm__ __volatile__("lock; cmpxchgl %1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__synch_xg(ptr)),
++ "0"(old)
++ : "memory");
++ return prev;
++#endif
++ }
++ return old;
++}
+
-+#ifndef __ASSEMBLY__
++#define synch_test_bit test_bit
+
-+/*
-+ * The use of 'barrier' in the following reflects their use as local-lock
-+ * operations. Reentrancy must be prevented (e.g., __cli()) /before/ following
-+ * critical operations are executed. All critical operations must complete
-+ * /before/ reentrancy is permitted (e.g., __sti()). Alpha architecture also
-+ * includes these barriers, for example.
-+ */
++#define synch_cmpxchg_subword synch_cmpxchg
+
-+#define __raw_local_save_flags() (current_vcpu_info()->evtchn_upcall_mask)
++#endif /* __XEN_SYNCH_BITOPS_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/system.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/system.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,488 @@
++#ifndef __ASM_SYSTEM_H
++#define __ASM_SYSTEM_H
+
-+#define raw_local_save_flags(flags) \
-+ do { (flags) = __raw_local_save_flags(); } while (0)
++#include <linux/kernel.h>
++#include <asm/segment.h>
++#include <asm/cpufeature.h>
++#include <linux/bitops.h> /* for LOCK_PREFIX */
++#include <asm/synch_bitops.h>
++#include <asm/hypervisor.h>
+
-+#define raw_local_irq_restore(x) \
-+do { \
-+ vcpu_info_t *_vcpu; \
-+ barrier(); \
-+ _vcpu = current_vcpu_info(); \
-+ if ((_vcpu->evtchn_upcall_mask = (x)) == 0) { \
-+ barrier(); /* unmask then check (avoid races) */ \
-+ if (unlikely(_vcpu->evtchn_upcall_pending)) \
-+ force_evtchn_callback(); \
-+ } \
-+} while (0)
++#ifdef __KERNEL__
+
-+#define raw_local_irq_disable() \
-+do { \
-+ current_vcpu_info()->evtchn_upcall_mask = 1; \
-+ barrier(); \
-+} while (0)
++struct task_struct; /* one of the stranger aspects of C forward declarations.. */
++extern struct task_struct * FASTCALL(__switch_to(struct task_struct *prev, struct task_struct *next));
+
-+#define raw_local_irq_enable() \
-+do { \
-+ vcpu_info_t *_vcpu; \
-+ barrier(); \
-+ _vcpu = current_vcpu_info(); \
-+ _vcpu->evtchn_upcall_mask = 0; \
-+ barrier(); /* unmask then check (avoid races) */ \
-+ if (unlikely(_vcpu->evtchn_upcall_pending)) \
-+ force_evtchn_callback(); \
++/*
++ * Saving eflags is important. It switches not only IOPL between tasks,
++ * it also protects other tasks from NT leaking through sysenter etc.
++ */
++#define switch_to(prev,next,last) do { \
++ unsigned long esi,edi; \
++ asm volatile("pushfl\n\t" /* Save flags */ \
++ "pushl %%ebp\n\t" \
++ "movl %%esp,%0\n\t" /* save ESP */ \
++ "movl %5,%%esp\n\t" /* restore ESP */ \
++ "movl $1f,%1\n\t" /* save EIP */ \
++ "pushl %6\n\t" /* restore EIP */ \
++ "jmp __switch_to\n" \
++ "1:\t" \
++ "popl %%ebp\n\t" \
++ "popfl" \
++ :"=m" (prev->thread.esp),"=m" (prev->thread.eip), \
++ "=a" (last),"=S" (esi),"=D" (edi) \
++ :"m" (next->thread.esp),"m" (next->thread.eip), \
++ "2" (prev), "d" (next)); \
+} while (0)
+
++#define _set_base(addr,base) do { unsigned long __pr; \
++__asm__ __volatile__ ("movw %%dx,%1\n\t" \
++ "rorl $16,%%edx\n\t" \
++ "movb %%dl,%2\n\t" \
++ "movb %%dh,%3" \
++ :"=&d" (__pr) \
++ :"m" (*((addr)+2)), \
++ "m" (*((addr)+4)), \
++ "m" (*((addr)+7)), \
++ "0" (base) \
++ ); } while(0)
++
++#define _set_limit(addr,limit) do { unsigned long __lr; \
++__asm__ __volatile__ ("movw %%dx,%1\n\t" \
++ "rorl $16,%%edx\n\t" \
++ "movb %2,%%dh\n\t" \
++ "andb $0xf0,%%dh\n\t" \
++ "orb %%dh,%%dl\n\t" \
++ "movb %%dl,%2" \
++ :"=&d" (__lr) \
++ :"m" (*(addr)), \
++ "m" (*((addr)+6)), \
++ "0" (limit) \
++ ); } while(0)
++
++#define set_base(ldt,base) _set_base( ((char *)&(ldt)) , (base) )
++#define set_limit(ldt,limit) _set_limit( ((char *)&(ldt)) , ((limit)-1) )
++
+/*
-+ * Used in the idle loop; sti takes one instruction cycle
-+ * to complete:
++ * Load a segment. Fall back on loading the zero
++ * segment if something goes wrong..
+ */
-+void raw_safe_halt(void);
++#define loadsegment(seg,value) \
++ asm volatile("\n" \
++ "1:\t" \
++ "mov %0,%%" #seg "\n" \
++ "2:\n" \
++ ".section .fixup,\"ax\"\n" \
++ "3:\t" \
++ "pushl $0\n\t" \
++ "popl %%" #seg "\n\t" \
++ "jmp 2b\n" \
++ ".previous\n" \
++ ".section __ex_table,\"a\"\n\t" \
++ ".align 4\n\t" \
++ ".long 1b,3b\n" \
++ ".previous" \
++ : :"rm" (value))
+
+/*
-+ * Used when interrupts are already enabled or to
-+ * shutdown the processor:
++ * Save a segment register away
+ */
-+void halt(void);
-+
-+static inline int raw_irqs_disabled_flags(unsigned long flags)
-+{
-+ return (flags != 0);
-+}
++#define savesegment(seg, value) \
++ asm volatile("mov %%" #seg ",%0":"=rm" (value))
+
-+#define raw_irqs_disabled() \
-+({ \
-+ unsigned long flags = __raw_local_save_flags(); \
-+ \
-+ raw_irqs_disabled_flags(flags); \
++#define read_cr0() ({ \
++ unsigned int __dummy; \
++ __asm__ __volatile__( \
++ "movl %%cr0,%0\n\t" \
++ :"=r" (__dummy)); \
++ __dummy; \
+})
++#define write_cr0(x) \
++ __asm__ __volatile__("movl %0,%%cr0": :"r" (x))
+
-+/*
-+ * For spinlocks, etc:
-+ */
-+#define __raw_local_irq_save() \
-+({ \
-+ unsigned long flags = __raw_local_save_flags(); \
-+ \
-+ raw_local_irq_disable(); \
-+ \
-+ flags; \
-+})
++#define read_cr2() (current_vcpu_info()->arch.cr2)
++#define write_cr2(x) \
++ __asm__ __volatile__("movl %0,%%cr2": :"r" (x))
+
-+#define raw_local_irq_save(flags) \
-+ do { (flags) = __raw_local_irq_save(); } while (0)
++#define read_cr3() ({ \
++ unsigned int __dummy; \
++ __asm__ ( \
++ "movl %%cr3,%0\n\t" \
++ :"=r" (__dummy)); \
++ __dummy = xen_cr3_to_pfn(__dummy); \
++ mfn_to_pfn(__dummy) << PAGE_SHIFT; \
++})
++#define write_cr3(x) ({ \
++ unsigned int __dummy = pfn_to_mfn((x) >> PAGE_SHIFT); \
++ __dummy = xen_pfn_to_cr3(__dummy); \
++ __asm__ __volatile__("movl %0,%%cr3": :"r" (__dummy)); \
++})
++#define read_cr4() ({ \
++ unsigned int __dummy; \
++ __asm__( \
++ "movl %%cr4,%0\n\t" \
++ :"=r" (__dummy)); \
++ __dummy; \
++})
++#define read_cr4_safe() ({ \
++ unsigned int __dummy; \
++ /* This could fault if %cr4 does not exist */ \
++ __asm__("1: movl %%cr4, %0 \n" \
++ "2: \n" \
++ ".section __ex_table,\"a\" \n" \
++ ".long 1b,2b \n" \
++ ".previous \n" \
++ : "=r" (__dummy): "0" (0)); \
++ __dummy; \
++})
+
-+#endif /* __ASSEMBLY__ */
++#define write_cr4(x) \
++ __asm__ __volatile__("movl %0,%%cr4": :"r" (x))
+
+/*
-+ * Do the CPU's IRQ-state tracing from assembly code. We call a
-+ * C function, so save all the C-clobbered registers:
++ * Clear and set 'TS' bit respectively
+ */
-+#ifdef CONFIG_TRACE_IRQFLAGS
-+
-+# define TRACE_IRQS_ON \
-+ pushl %eax; \
-+ pushl %ecx; \
-+ pushl %edx; \
-+ call trace_hardirqs_on; \
-+ popl %edx; \
-+ popl %ecx; \
-+ popl %eax;
++#define clts() (HYPERVISOR_fpu_taskswitch(0))
++#define stts() (HYPERVISOR_fpu_taskswitch(1))
+
-+# define TRACE_IRQS_OFF \
-+ pushl %eax; \
-+ pushl %ecx; \
-+ pushl %edx; \
-+ call trace_hardirqs_off; \
-+ popl %edx; \
-+ popl %ecx; \
-+ popl %eax;
++#endif /* __KERNEL__ */
+
-+#else
-+# define TRACE_IRQS_ON
-+# define TRACE_IRQS_OFF
-+#endif
++#define wbinvd() \
++ __asm__ __volatile__ ("wbinvd": : :"memory")
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/maddr.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/maddr.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/maddr.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/maddr.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,193 @@
-+#ifndef _I386_MADDR_H
-+#define _I386_MADDR_H
++static inline unsigned long get_limit(unsigned long segment)
++{
++ unsigned long __limit;
++ __asm__("lsll %1,%0"
++ :"=r" (__limit):"r" (segment));
++ return __limit+1;
++}
+
-+#include <xen/features.h>
-+#include <xen/interface/xen.h>
++#define nop() __asm__ __volatile__ ("nop")
+
-+/**** MACHINE <-> PHYSICAL CONVERSION MACROS ****/
-+#define INVALID_P2M_ENTRY (~0UL)
-+#define FOREIGN_FRAME_BIT (1UL<<31)
-+#define FOREIGN_FRAME(m) ((m) | FOREIGN_FRAME_BIT)
++#define xchg(ptr,v) ((__typeof__(*(ptr)))__xchg((unsigned long)(v),(ptr),sizeof(*(ptr))))
+
-+/* Definitions for machine and pseudophysical addresses. */
-+#ifdef CONFIG_X86_PAE
-+typedef unsigned long long paddr_t;
-+typedef unsigned long long maddr_t;
-+#else
-+typedef unsigned long paddr_t;
-+typedef unsigned long maddr_t;
-+#endif
++#define tas(ptr) (xchg((ptr),1))
+
-+#ifdef CONFIG_XEN
++struct __xchg_dummy { unsigned long a[100]; };
++#define __xg(x) ((struct __xchg_dummy *)(x))
+
-+extern unsigned long *phys_to_machine_mapping;
-+extern unsigned long max_mapnr;
+
-+#undef machine_to_phys_mapping
-+extern unsigned long *machine_to_phys_mapping;
-+extern unsigned int machine_to_phys_order;
++#ifdef CONFIG_X86_CMPXCHG64
+
-+static inline unsigned long pfn_to_mfn(unsigned long pfn)
++/*
++ * The semantics of XCHGCMP8B are a bit strange, this is why
++ * there is a loop and the loading of %%eax and %%edx has to
++ * be inside. This inlines well in most cases, the cached
++ * cost is around ~38 cycles. (in the future we might want
++ * to do an SIMD/3DNOW!/MMX/FPU 64-bit store here, but that
++ * might have an implicit FPU-save as a cost, so it's not
++ * clear which path to go.)
++ *
++ * cmpxchg8b must be used with the lock prefix here to allow
++ * the instruction to be executed atomically, see page 3-102
++ * of the instruction set reference 24319102.pdf. We need
++ * the reader side to see the coherent 64bit value.
++ */
++static inline void __set_64bit (unsigned long long * ptr,
++ unsigned int low, unsigned int high)
+{
-+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ return pfn;
-+ BUG_ON(max_mapnr && pfn >= max_mapnr);
-+ return phys_to_machine_mapping[pfn] & ~FOREIGN_FRAME_BIT;
++ __asm__ __volatile__ (
++ "\n1:\t"
++ "movl (%0), %%eax\n\t"
++ "movl 4(%0), %%edx\n\t"
++ "lock cmpxchg8b (%0)\n\t"
++ "jnz 1b"
++ : /* no outputs */
++ : "D"(ptr),
++ "b"(low),
++ "c"(high)
++ : "ax","dx","memory");
+}
+
-+static inline int phys_to_machine_mapping_valid(unsigned long pfn)
++static inline void __set_64bit_constant (unsigned long long *ptr,
++ unsigned long long value)
+{
-+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ return 1;
-+ BUG_ON(max_mapnr && pfn >= max_mapnr);
-+ return (phys_to_machine_mapping[pfn] != INVALID_P2M_ENTRY);
++ __set_64bit(ptr,(unsigned int)(value), (unsigned int)((value)>>32ULL));
+}
++#define ll_low(x) *(((unsigned int*)&(x))+0)
++#define ll_high(x) *(((unsigned int*)&(x))+1)
+
-+static inline unsigned long mfn_to_pfn(unsigned long mfn)
++static inline void __set_64bit_var (unsigned long long *ptr,
++ unsigned long long value)
+{
-+ unsigned long pfn;
-+
-+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ return mfn;
++ __set_64bit(ptr,ll_low(value), ll_high(value));
++}
+
-+ if (unlikely((mfn >> machine_to_phys_order) != 0))
-+ return max_mapnr;
++#define set_64bit(ptr,value) \
++(__builtin_constant_p(value) ? \
++ __set_64bit_constant(ptr, value) : \
++ __set_64bit_var(ptr, value) )
+
-+ /* The array access can fail (e.g., device space beyond end of RAM). */
-+ asm (
-+ "1: movl %1,%0\n"
-+ "2:\n"
-+ ".section .fixup,\"ax\"\n"
-+ "3: movl %2,%0\n"
-+ " jmp 2b\n"
-+ ".previous\n"
-+ ".section __ex_table,\"a\"\n"
-+ " .align 4\n"
-+ " .long 1b,3b\n"
-+ ".previous"
-+ : "=r" (pfn)
-+ : "m" (machine_to_phys_mapping[mfn]), "m" (max_mapnr) );
++#define _set_64bit(ptr,value) \
++(__builtin_constant_p(value) ? \
++ __set_64bit(ptr, (unsigned int)(value), (unsigned int)((value)>>32ULL) ) : \
++ __set_64bit(ptr, ll_low(value), ll_high(value)) )
+
-+ return pfn;
-+}
++#endif
+
+/*
-+ * We detect special mappings in one of two ways:
-+ * 1. If the MFN is an I/O page then Xen will set the m2p entry
-+ * to be outside our maximum possible pseudophys range.
-+ * 2. If the MFN belongs to a different domain then we will certainly
-+ * not have MFN in our p2m table. Conversely, if the page is ours,
-+ * then we'll have p2m(m2p(MFN))==MFN.
-+ * If we detect a special mapping then it doesn't have a 'struct page'.
-+ * We force !pfn_valid() by returning an out-of-range pointer.
-+ *
-+ * NB. These checks require that, for any MFN that is not in our reservation,
-+ * there is no PFN such that p2m(PFN) == MFN. Otherwise we can get confused if
-+ * we are foreign-mapping the MFN, and the other domain as m2p(MFN) == PFN.
-+ * Yikes! Various places must poke in INVALID_P2M_ENTRY for safety.
-+ *
-+ * NB2. When deliberately mapping foreign pages into the p2m table, you *must*
-+ * use FOREIGN_FRAME(). This will cause pte_pfn() to choke on it, as we
-+ * require. In all the cases we care about, the FOREIGN_FRAME bit is
-+ * masked (e.g., pfn_to_mfn()) so behaviour there is correct.
++ * Note: no "lock" prefix even on SMP: xchg always implies lock anyway
++ * Note 2: xchg has side effect, so that attribute volatile is necessary,
++ * but generally the primitive is invalid, *ptr is output argument. --ANK
+ */
-+static inline unsigned long mfn_to_local_pfn(unsigned long mfn)
++static inline unsigned long __xchg(unsigned long x, volatile void * ptr, int size)
+{
-+ unsigned long pfn = mfn_to_pfn(mfn);
-+ if ((pfn < max_mapnr)
-+ && !xen_feature(XENFEAT_auto_translated_physmap)
-+ && (phys_to_machine_mapping[pfn] != mfn))
-+ return max_mapnr; /* force !pfn_valid() */
-+ return pfn;
++ switch (size) {
++ case 1:
++ __asm__ __volatile__("xchgb %b0,%1"
++ :"=q" (x)
++ :"m" (*__xg(ptr)), "0" (x)
++ :"memory");
++ break;
++ case 2:
++ __asm__ __volatile__("xchgw %w0,%1"
++ :"=r" (x)
++ :"m" (*__xg(ptr)), "0" (x)
++ :"memory");
++ break;
++ case 4:
++ __asm__ __volatile__("xchgl %0,%1"
++ :"=r" (x)
++ :"m" (*__xg(ptr)), "0" (x)
++ :"memory");
++ break;
++ }
++ return x;
+}
+
-+static inline void set_phys_to_machine(unsigned long pfn, unsigned long mfn)
++/*
++ * Atomic compare and exchange. Compare OLD with MEM, if identical,
++ * store NEW in MEM. Return the initial value in MEM. Success is
++ * indicated by comparing RETURN with OLD.
++ */
++
++#ifdef CONFIG_X86_CMPXCHG
++#define __HAVE_ARCH_CMPXCHG 1
++#define cmpxchg(ptr,o,n)\
++ ((__typeof__(*(ptr)))__cmpxchg((ptr),(unsigned long)(o),\
++ (unsigned long)(n),sizeof(*(ptr))))
++#endif
++
++static inline unsigned long __cmpxchg(volatile void *ptr, unsigned long old,
++ unsigned long new, int size)
+{
-+ BUG_ON(max_mapnr && pfn >= max_mapnr);
-+ if (xen_feature(XENFEAT_auto_translated_physmap)) {
-+ BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY);
-+ return;
++ unsigned long prev;
++ switch (size) {
++ case 1:
++ __asm__ __volatile__(LOCK_PREFIX "cmpxchgb %b1,%2"
++ : "=a"(prev)
++ : "q"(new), "m"(*__xg(ptr)), "0"(old)
++ : "memory");
++ return prev;
++ case 2:
++ __asm__ __volatile__(LOCK_PREFIX "cmpxchgw %w1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__xg(ptr)), "0"(old)
++ : "memory");
++ return prev;
++ case 4:
++ __asm__ __volatile__(LOCK_PREFIX "cmpxchgl %1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__xg(ptr)), "0"(old)
++ : "memory");
++ return prev;
+ }
-+ phys_to_machine_mapping[pfn] = mfn;
++ return old;
+}
+
-+static inline maddr_t phys_to_machine(paddr_t phys)
-+{
-+ maddr_t machine = pfn_to_mfn(phys >> PAGE_SHIFT);
-+ machine = (machine << PAGE_SHIFT) | (phys & ~PAGE_MASK);
-+ return machine;
-+}
++#ifndef CONFIG_X86_CMPXCHG
++/*
++ * Building a kernel capable running on 80386. It may be necessary to
++ * simulate the cmpxchg on the 80386 CPU. For that purpose we define
++ * a function for each of the sizes we support.
++ */
+
-+static inline paddr_t machine_to_phys(maddr_t machine)
-+{
-+ paddr_t phys = mfn_to_pfn(machine >> PAGE_SHIFT);
-+ phys = (phys << PAGE_SHIFT) | (machine & ~PAGE_MASK);
-+ return phys;
-+}
++extern unsigned long cmpxchg_386_u8(volatile void *, u8, u8);
++extern unsigned long cmpxchg_386_u16(volatile void *, u16, u16);
++extern unsigned long cmpxchg_386_u32(volatile void *, u32, u32);
+
-+#ifdef CONFIG_X86_PAE
-+static inline paddr_t pte_phys_to_machine(paddr_t phys)
++static inline unsigned long cmpxchg_386(volatile void *ptr, unsigned long old,
++ unsigned long new, int size)
+{
-+ /*
-+ * In PAE mode, the NX bit needs to be dealt with in the value
-+ * passed to pfn_to_mfn(). On x86_64, we need to mask it off,
-+ * but for i386 the conversion to ulong for the argument will
-+ * clip it off.
-+ */
-+ maddr_t machine = pfn_to_mfn(phys >> PAGE_SHIFT);
-+ machine = (machine << PAGE_SHIFT) | (phys & ~PHYSICAL_PAGE_MASK);
-+ return machine;
++ switch (size) {
++ case 1:
++ return cmpxchg_386_u8(ptr, old, new);
++ case 2:
++ return cmpxchg_386_u16(ptr, old, new);
++ case 4:
++ return cmpxchg_386_u32(ptr, old, new);
++ }
++ return old;
+}
+
-+static inline paddr_t pte_machine_to_phys(maddr_t machine)
++#define cmpxchg(ptr,o,n) \
++({ \
++ __typeof__(*(ptr)) __ret; \
++ if (likely(boot_cpu_data.x86 > 3)) \
++ __ret = __cmpxchg((ptr), (unsigned long)(o), \
++ (unsigned long)(n), sizeof(*(ptr))); \
++ else \
++ __ret = cmpxchg_386((ptr), (unsigned long)(o), \
++ (unsigned long)(n), sizeof(*(ptr))); \
++ __ret; \
++})
++#endif
++
++#ifdef CONFIG_X86_CMPXCHG64
++
++static inline unsigned long long __cmpxchg64(volatile void *ptr, unsigned long long old,
++ unsigned long long new)
+{
-+ /*
-+ * In PAE mode, the NX bit needs to be dealt with in the value
-+ * passed to mfn_to_pfn(). On x86_64, we need to mask it off,
-+ * but for i386 the conversion to ulong for the argument will
-+ * clip it off.
-+ */
-+ paddr_t phys = mfn_to_pfn(machine >> PAGE_SHIFT);
-+ phys = (phys << PAGE_SHIFT) | (machine & ~PHYSICAL_PAGE_MASK);
-+ return phys;
++ unsigned long long prev;
++ __asm__ __volatile__(LOCK_PREFIX "cmpxchg8b %3"
++ : "=A"(prev)
++ : "b"((unsigned long)new),
++ "c"((unsigned long)(new >> 32)),
++ "m"(*__xg(ptr)),
++ "0"(old)
++ : "memory");
++ return prev;
+}
-+#endif
+
-+#ifdef CONFIG_X86_PAE
-+static inline pte_t pfn_pte_ma(unsigned long page_nr, pgprot_t pgprot)
-+{
-+ pte_t pte;
++#define cmpxchg64(ptr,o,n)\
++ ((__typeof__(*(ptr)))__cmpxchg64((ptr),(unsigned long long)(o),\
++ (unsigned long long)(n)))
+
-+ pte.pte_high = (page_nr >> (32 - PAGE_SHIFT)) | \
-+ (pgprot_val(pgprot) >> 32);
-+ pte.pte_high &= (__supported_pte_mask >> 32);
-+ pte.pte_low = ((page_nr << PAGE_SHIFT) | pgprot_val(pgprot)) & \
-+ __supported_pte_mask;
-+ return pte;
-+}
-+#else
-+#define pfn_pte_ma(pfn, prot) __pte_ma(((pfn) << PAGE_SHIFT) | pgprot_val(prot))
+#endif
++
++/*
++ * Force strict CPU ordering.
++ * And yes, this is required on UP too when we're talking
++ * to devices.
++ *
++ * For now, "wmb()" doesn't actually do anything, as all
++ * Intel CPU's follow what Intel calls a *Processor Order*,
++ * in which all writes are seen in the program order even
++ * outside the CPU.
++ *
++ * I expect future Intel CPU's to have a weaker ordering,
++ * but I'd also expect them to finally get their act together
++ * and add some real memory barriers if so.
++ *
++ * Some non intel clones support out of order store. wmb() ceases to be a
++ * nop for these.
++ */
++
+
-+#define __pte_ma(x) ((pte_t) { (x) } )
++/*
++ * Actually only lfence would be needed for mb() because all stores done
++ * by the kernel should be already ordered. But keep a full barrier for now.
++ */
+
-+#else /* !CONFIG_XEN */
++#define mb() alternative("lock; addl $0,0(%%esp)", "mfence", X86_FEATURE_XMM2)
++#define rmb() alternative("lock; addl $0,0(%%esp)", "lfence", X86_FEATURE_XMM2)
+
-+#define pfn_to_mfn(pfn) (pfn)
-+#define mfn_to_pfn(mfn) (mfn)
-+#define mfn_to_local_pfn(mfn) (mfn)
-+#define set_phys_to_machine(pfn, mfn) ((void)0)
-+#define phys_to_machine_mapping_valid(pfn) (1)
-+#define phys_to_machine(phys) ((maddr_t)(phys))
-+#define machine_to_phys(mach) ((paddr_t)(mach))
-+#define pfn_pte_ma(pfn, prot) pfn_pte(pfn, prot)
-+#define __pte_ma(x) __pte(x)
++/**
++ * read_barrier_depends - Flush all pending reads that subsequents reads
++ * depend on.
++ *
++ * No data-dependent reads from memory-like regions are ever reordered
++ * over this barrier. All reads preceding this primitive are guaranteed
++ * to access memory (but not necessarily other CPUs' caches) before any
++ * reads following this primitive that depend on the data return by
++ * any of the preceding reads. This primitive is much lighter weight than
++ * rmb() on most CPUs, and is never heavier weight than is
++ * rmb().
++ *
++ * These ordering constraints are respected by both the local CPU
++ * and the compiler.
++ *
++ * Ordering is not guaranteed by anything other than these primitives,
++ * not even by data dependencies. See the documentation for
++ * memory_barrier() for examples and URLs to more information.
++ *
++ * For example, the following code would force ordering (the initial
++ * value of "a" is zero, "b" is one, and "p" is "&a"):
++ *
++ * <programlisting>
++ * CPU 0 CPU 1
++ *
++ * b = 2;
++ * memory_barrier();
++ * p = &b; q = p;
++ * read_barrier_depends();
++ * d = *q;
++ * </programlisting>
++ *
++ * because the read of "*q" depends on the read of "p" and these
++ * two reads are separated by a read_barrier_depends(). However,
++ * the following code, with the same initial values for "a" and "b":
++ *
++ * <programlisting>
++ * CPU 0 CPU 1
++ *
++ * a = 2;
++ * memory_barrier();
++ * b = 3; y = b;
++ * read_barrier_depends();
++ * x = a;
++ * </programlisting>
++ *
++ * does not enforce ordering, since there is no data dependency between
++ * the read of "a" and the read of "b". Therefore, on some CPUs, such
++ * as Alpha, "y" could be set to 3 and "x" to 0. Use rmb()
++ * in cases like this where there are no data dependencies.
++ **/
+
-+#endif /* !CONFIG_XEN */
++#define read_barrier_depends() do { } while(0)
+
-+/* VIRT <-> MACHINE conversion */
-+#define virt_to_machine(v) (phys_to_machine(__pa(v)))
-+#define virt_to_mfn(v) (pfn_to_mfn(__pa(v) >> PAGE_SHIFT))
-+#define mfn_to_virt(m) (__va(mfn_to_pfn(m) << PAGE_SHIFT))
++#ifdef CONFIG_X86_OOSTORE
++/* Actually there are no OOO store capable CPUs for now that do SSE,
++ but make it already an possibility. */
++#define wmb() alternative("lock; addl $0,0(%%esp)", "sfence", X86_FEATURE_XMM)
++#else
++#define wmb() __asm__ __volatile__ ("": : :"memory")
++#endif
+
-+#endif /* _I386_MADDR_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/mmu_context.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/mmu_context.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/mmu_context.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/mmu_context.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,108 @@
-+#ifndef __I386_SCHED_H
-+#define __I386_SCHED_H
++#ifdef CONFIG_SMP
++#define smp_mb() mb()
++#define smp_rmb() rmb()
++#define smp_wmb() wmb()
++#define smp_read_barrier_depends() read_barrier_depends()
++#define set_mb(var, value) do { (void) xchg(&var, value); } while (0)
++#else
++#define smp_mb() barrier()
++#define smp_rmb() barrier()
++#define smp_wmb() barrier()
++#define smp_read_barrier_depends() do { } while(0)
++#define set_mb(var, value) do { var = value; barrier(); } while (0)
++#endif
+
-+#include <asm/desc.h>
-+#include <asm/atomic.h>
-+#include <asm/pgalloc.h>
-+#include <asm/tlbflush.h>
++#include <linux/irqflags.h>
+
+/*
-+ * Used for LDT copy/destruction.
++ * disable hlt during certain critical i/o operations
+ */
-+int init_new_context(struct task_struct *tsk, struct mm_struct *mm);
-+void destroy_context(struct mm_struct *mm);
-+
-+
-+static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk)
-+{
-+#if 0 /* XEN: no lazy tlb */
-+ unsigned cpu = smp_processor_id();
-+ if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_OK)
-+ per_cpu(cpu_tlbstate, cpu).state = TLBSTATE_LAZY;
-+#endif
-+}
++#define HAVE_DISABLE_HLT
++void disable_hlt(void);
++void enable_hlt(void);
+
-+#define prepare_arch_switch(next) __prepare_arch_switch()
++extern int es7000_plat;
++void cpu_idle_wait(void);
+
-+static inline void __prepare_arch_switch(void)
++/*
++ * On SMP systems, when the scheduler does migration-cost autodetection,
++ * it needs a way to flush as much of the CPU's caches as possible:
++ */
++static inline void sched_cacheflush(void)
+{
-+ /*
-+ * Save away %fs and %gs. No need to save %es and %ds, as those
-+ * are always kernel segments while inside the kernel. Must
-+ * happen before reload of cr3/ldt (i.e., not in __switch_to).
-+ */
-+ asm volatile ( "mov %%fs,%0 ; mov %%gs,%1"
-+ : "=m" (current->thread.fs),
-+ "=m" (current->thread.gs));
-+ asm volatile ( "movl %0,%%fs ; movl %0,%%gs"
-+ : : "r" (0) );
++ wbinvd();
+}
+
-+extern void mm_pin(struct mm_struct *mm);
-+extern void mm_unpin(struct mm_struct *mm);
-+void mm_pin_all(void);
-+
-+static inline void switch_mm(struct mm_struct *prev,
-+ struct mm_struct *next,
-+ struct task_struct *tsk)
-+{
-+ int cpu = smp_processor_id();
-+ struct mmuext_op _op[2], *op = _op;
++extern unsigned long arch_align_stack(unsigned long sp);
++extern void free_init_pages(char *what, unsigned long begin, unsigned long end);
+
-+ if (likely(prev != next)) {
-+ BUG_ON(!xen_feature(XENFEAT_writable_page_tables) &&
-+ !test_bit(PG_pinned, &virt_to_page(next->pgd)->flags));
++void default_idle(void);
+
-+ /* stop flush ipis for the previous mm */
-+ cpu_clear(cpu, prev->cpu_vm_mask);
-+#if 0 /* XEN: no lazy tlb */
-+ per_cpu(cpu_tlbstate, cpu).state = TLBSTATE_OK;
-+ per_cpu(cpu_tlbstate, cpu).active_mm = next;
+#endif
-+ cpu_set(cpu, next->cpu_vm_mask);
-+
-+ /* Re-load page tables: load_cr3(next->pgd) */
-+ op->cmd = MMUEXT_NEW_BASEPTR;
-+ op->arg1.mfn = pfn_to_mfn(__pa(next->pgd) >> PAGE_SHIFT);
-+ op++;
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/tlbflush.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/tlbflush.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,101 @@
++#ifndef _I386_TLBFLUSH_H
++#define _I386_TLBFLUSH_H
+
-+ /*
-+ * load the LDT, if the LDT is different:
-+ */
-+ if (unlikely(prev->context.ldt != next->context.ldt)) {
-+ /* load_LDT_nolock(&next->context, cpu) */
-+ op->cmd = MMUEXT_SET_LDT;
-+ op->arg1.linear_addr = (unsigned long)next->context.ldt;
-+ op->arg2.nr_ents = next->context.size;
-+ op++;
-+ }
++#include <linux/mm.h>
++#include <asm/processor.h>
+
-+ BUG_ON(HYPERVISOR_mmuext_op(_op, op-_op, NULL, DOMID_SELF));
-+ }
-+#if 0 /* XEN: no lazy tlb */
-+ else {
-+ per_cpu(cpu_tlbstate, cpu).state = TLBSTATE_OK;
-+ BUG_ON(per_cpu(cpu_tlbstate, cpu).active_mm != next);
++#define __flush_tlb() xen_tlb_flush()
++#define __flush_tlb_global() xen_tlb_flush()
++#define __flush_tlb_all() xen_tlb_flush()
+
-+ if (!cpu_test_and_set(cpu, next->cpu_vm_mask)) {
-+ /* We were in lazy tlb mode and leave_mm disabled
-+ * tlb flush IPI delivery. We must reload %cr3.
-+ */
-+ load_cr3(next->pgd);
-+ load_LDT_nolock(&next->context, cpu);
-+ }
-+ }
-+#endif
-+}
++extern unsigned long pgkern_mask;
+
-+#define deactivate_mm(tsk, mm) \
-+ asm("movl %0,%%fs ; movl %0,%%gs": :"r" (0))
++#define cpu_has_invlpg (boot_cpu_data.x86 > 3)
+
-+static inline void activate_mm(struct mm_struct *prev, struct mm_struct *next)
-+{
-+ if (!test_bit(PG_pinned, &virt_to_page(next->pgd)->flags))
-+ mm_pin(next);
-+ switch_mm(prev, next, NULL);
-+}
++#define __flush_tlb_single(addr) xen_invlpg(addr)
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/mmu.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/mmu.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/mmu.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/mmu.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,29 @@
-+#ifndef __i386_MMU_H
-+#define __i386_MMU_H
++#define __flush_tlb_one(addr) __flush_tlb_single(addr)
+
-+#include <asm/semaphore.h>
+/*
-+ * The i386 doesn't have a mmu context, but
-+ * we put the segment information here.
++ * TLB flushing:
+ *
-+ * cpu_vm_mask is used to optimize ldt flushing.
++ * - flush_tlb() flushes the current mm struct TLBs
++ * - flush_tlb_all() flushes all processes TLBs
++ * - flush_tlb_mm(mm) flushes the specified mm context TLB's
++ * - flush_tlb_page(vma, vmaddr) flushes one page
++ * - flush_tlb_range(vma, start, end) flushes a range of pages
++ * - flush_tlb_kernel_range(start, end) flushes a range of kernel pages
++ * - flush_tlb_pgtables(mm, start, end) flushes a range of page tables
++ *
++ * ..but the i386 has somewhat limited tlb flushing capabilities,
++ * and page-granular flushes are available only on i486 and up.
+ */
-+typedef struct {
-+ int size;
-+ struct semaphore sem;
-+ void *ldt;
-+ void *vdso;
-+#ifdef CONFIG_XEN
-+ int has_foreign_mappings;
-+#endif
-+} mm_context_t;
+
-+/* mm/memory.c:exit_mmap hook */
-+extern void _arch_exit_mmap(struct mm_struct *mm);
-+#define arch_exit_mmap(_mm) _arch_exit_mmap(_mm)
++#ifndef CONFIG_SMP
+
-+/* kernel/fork.c:dup_mmap hook */
-+extern void _arch_dup_mmap(struct mm_struct *mm);
-+#define arch_dup_mmap(mm, oldmm) ((void)(oldmm), _arch_dup_mmap(mm))
++#define flush_tlb() __flush_tlb()
++#define flush_tlb_all() __flush_tlb_all()
++#define local_flush_tlb() __flush_tlb()
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/page.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/page.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/page.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/page.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,235 @@
-+#ifndef _I386_PAGE_H
-+#define _I386_PAGE_H
++static inline void flush_tlb_mm(struct mm_struct *mm)
++{
++ if (mm == current->active_mm)
++ __flush_tlb();
++}
+
-+/* PAGE_SHIFT determines the page size */
-+#define PAGE_SHIFT 12
-+#define PAGE_SIZE (1UL << PAGE_SHIFT)
-+#define PAGE_MASK (~(PAGE_SIZE-1))
++static inline void flush_tlb_page(struct vm_area_struct *vma,
++ unsigned long addr)
++{
++ if (vma->vm_mm == current->active_mm)
++ __flush_tlb_one(addr);
++}
++
++static inline void flush_tlb_range(struct vm_area_struct *vma,
++ unsigned long start, unsigned long end)
++{
++ if (vma->vm_mm == current->active_mm)
++ __flush_tlb();
++}
+
-+#ifdef CONFIG_X86_PAE
-+#define __PHYSICAL_MASK_SHIFT 40
-+#define __PHYSICAL_MASK ((1ULL << __PHYSICAL_MASK_SHIFT) - 1)
-+#define PHYSICAL_PAGE_MASK (~((1ULL << PAGE_SHIFT) - 1) & __PHYSICAL_MASK)
+#else
-+#define __PHYSICAL_MASK_SHIFT 32
-+#define __PHYSICAL_MASK (~0UL)
-+#define PHYSICAL_PAGE_MASK (PAGE_MASK & __PHYSICAL_MASK)
-+#endif
+
-+#define LARGE_PAGE_MASK (~(LARGE_PAGE_SIZE-1))
-+#define LARGE_PAGE_SIZE (1UL << PMD_SHIFT)
++#include <asm/smp.h>
+
-+#ifdef __KERNEL__
++#define local_flush_tlb() \
++ __flush_tlb()
+
-+/*
-+ * Need to repeat this here in order to not include pgtable.h (which in turn
-+ * depends on definitions made here), but to be able to use the symbolic
-+ * below. The preprocessor will warn if the two definitions aren't identical.
-+ */
-+#define _PAGE_PRESENT 0x001
++#define flush_tlb_all xen_tlb_flush_all
++#define flush_tlb_current_task() xen_tlb_flush_mask(¤t->mm->cpu_vm_mask)
++#define flush_tlb_mm(mm) xen_tlb_flush_mask(&(mm)->cpu_vm_mask)
++#define flush_tlb_page(vma, va) xen_invlpg_mask(&(vma)->vm_mm->cpu_vm_mask, va)
+
-+#ifndef __ASSEMBLY__
++#define flush_tlb() flush_tlb_current_task()
+
-+#include <linux/string.h>
-+#include <linux/types.h>
-+#include <linux/kernel.h>
-+#include <asm/bug.h>
-+#include <xen/interface/xen.h>
-+#include <xen/features.h>
++static inline void flush_tlb_range(struct vm_area_struct * vma, unsigned long start, unsigned long end)
++{
++ flush_tlb_mm(vma->vm_mm);
++}
+
-+#define arch_free_page(_page,_order) \
-+({ int foreign = PageForeign(_page); \
-+ if (foreign) \
-+ PageForeignDestructor(_page); \
-+ foreign; \
-+})
-+#define HAVE_ARCH_FREE_PAGE
++#define TLBSTATE_OK 1
++#define TLBSTATE_LAZY 2
+
-+#ifdef CONFIG_X86_USE_3DNOW
++struct tlb_state
++{
++ struct mm_struct *active_mm;
++ int state;
++ char __cacheline_padding[L1_CACHE_BYTES-8];
++};
++DECLARE_PER_CPU(struct tlb_state, cpu_tlbstate);
+
-+#include <asm/mmx.h>
+
-+#define clear_page(page) mmx_clear_page((void *)(page))
-+#define copy_page(to,from) mmx_copy_page(to,from)
++#endif
+
-+#else
++#define flush_tlb_kernel_range(start, end) flush_tlb_all()
+
-+#define alloc_zeroed_user_highpage(vma, vaddr) alloc_page_vma(GFP_HIGHUSER | __GFP_ZERO, vma, vaddr)
-+#define __HAVE_ARCH_ALLOC_ZEROED_USER_HIGHPAGE
++static inline void flush_tlb_pgtables(struct mm_struct *mm,
++ unsigned long start, unsigned long end)
++{
++ /* i386 does not keep any page table caches in TLB */
++}
+
++#endif /* _I386_TLBFLUSH_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/vga.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/vga.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,20 @@
+/*
-+ * On older X86 processors it's not a win to use MMX here it seems.
-+ * Maybe the K6-III ?
++ * Access to VGA videoram
++ *
++ * (c) 1998 Martin Mares <mj at ucw.cz>
+ */
-+
-+#define clear_page(page) memset((void *)(page), 0, PAGE_SIZE)
-+#define copy_page(to,from) memcpy((void *)(to), (void *)(from), PAGE_SIZE)
-+
-+#endif
+
-+#define clear_user_page(page, vaddr, pg) clear_page(page)
-+#define copy_user_page(to, from, vaddr, pg) copy_page(to, from)
++#ifndef _LINUX_ASM_VGA_H_
++#define _LINUX_ASM_VGA_H_
+
+/*
-+ * These are used to make use of C type-checking..
++ * On the PC, we can just recalculate addresses and then
++ * access the videoram directly without any black magic.
+ */
-+extern int nx_enabled;
-+#ifdef CONFIG_X86_PAE
-+extern unsigned long long __supported_pte_mask;
-+typedef struct { unsigned long pte_low, pte_high; } pte_t;
-+typedef struct { unsigned long long pmd; } pmd_t;
-+typedef struct { unsigned long long pgd; } pgd_t;
-+typedef struct { unsigned long long pgprot; } pgprot_t;
-+#define pgprot_val(x) ((x).pgprot)
-+#include <asm/maddr.h>
-+#define __pte(x) ({ unsigned long long _x = (x); \
-+ if (_x & _PAGE_PRESENT) _x = pte_phys_to_machine(_x); \
-+ ((pte_t) {(unsigned long)(_x), (unsigned long)(_x>>32)}); })
-+#define __pgd(x) ({ unsigned long long _x = (x); \
-+ (pgd_t) {((_x) & _PAGE_PRESENT) ? pte_phys_to_machine(_x) : (_x)}; })
-+#define __pmd(x) ({ unsigned long long _x = (x); \
-+ (pmd_t) {((_x) & _PAGE_PRESENT) ? pte_phys_to_machine(_x) : (_x)}; })
-+static inline unsigned long long __pte_val(pte_t x)
-+{
-+ return ((unsigned long long)x.pte_high << 32) | x.pte_low;
-+}
-+static inline unsigned long long pte_val(pte_t x)
-+{
-+ unsigned long long ret = __pte_val(x);
-+ if (x.pte_low & _PAGE_PRESENT) ret = pte_machine_to_phys(ret);
-+ return ret;
-+}
-+#define __pmd_val(x) ((x).pmd)
-+static inline unsigned long long pmd_val(pmd_t x)
-+{
-+ unsigned long long ret = __pmd_val(x);
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (ret) ret = pte_machine_to_phys(ret) | _PAGE_PRESENT;
-+#else
-+ if (ret & _PAGE_PRESENT) ret = pte_machine_to_phys(ret);
-+#endif
-+ return ret;
-+}
-+#define __pud_val(x) __pgd_val((x).pgd)
-+#define __pgd_val(x) ((x).pgd)
-+static inline unsigned long long pgd_val(pgd_t x)
-+{
-+ unsigned long long ret = __pgd_val(x);
-+ if (ret & _PAGE_PRESENT) ret = pte_machine_to_phys(ret);
-+ return ret;
-+}
-+#define HPAGE_SHIFT 21
-+#else
-+typedef struct { unsigned long pte_low; } pte_t;
-+typedef struct { unsigned long pgd; } pgd_t;
-+typedef struct { unsigned long pgprot; } pgprot_t;
-+#define pgprot_val(x) ((x).pgprot)
-+#include <asm/maddr.h>
-+#define boot_pte_t pte_t /* or would you rather have a typedef */
-+#define __pte_val(x) ((x).pte_low)
-+#define pte_val(x) (__pte_val(x) & _PAGE_PRESENT ? \
-+ machine_to_phys(__pte_val(x)) : \
-+ __pte_val(x))
-+#define __pte(x) ({ unsigned long _x = (x); \
-+ (pte_t) {((_x) & _PAGE_PRESENT) ? phys_to_machine(_x) : (_x)}; })
-+#define __pmd_val(x) __pud_val((x).pud)
-+#define __pud_val(x) __pgd_val((x).pgd)
-+#define __pgd(x) ({ unsigned long _x = (x); \
-+ (pgd_t) {((_x) & _PAGE_PRESENT) ? phys_to_machine(_x) : (_x)}; })
-+#define __pgd_val(x) ((x).pgd)
-+static inline unsigned long pgd_val(pgd_t x)
-+{
-+ unsigned long ret = __pgd_val(x);
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (ret) ret = machine_to_phys(ret) | _PAGE_PRESENT;
-+#else
-+ if (ret & _PAGE_PRESENT) ret = machine_to_phys(ret);
-+#endif
-+ return ret;
-+}
-+#define HPAGE_SHIFT 22
-+#endif
-+#define PTE_MASK PHYSICAL_PAGE_MASK
+
-+#ifdef CONFIG_HUGETLB_PAGE
-+#define HPAGE_SIZE ((1UL) << HPAGE_SHIFT)
-+#define HPAGE_MASK (~(HPAGE_SIZE - 1))
-+#define HUGETLB_PAGE_ORDER (HPAGE_SHIFT - PAGE_SHIFT)
-+#define HAVE_ARCH_HUGETLB_UNMAPPED_AREA
-+#endif
++#define VGA_MAP_MEM(x,s) (unsigned long)isa_bus_to_virt(x)
+
-+#define __pgprot(x) ((pgprot_t) { (x) } )
++#define vga_readb(x) (*(x))
++#define vga_writeb(x,y) (*(y) = (x))
+
-+#endif /* !__ASSEMBLY__ */
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/asm/xenoprof.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/asm/xenoprof.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,48 @@
++/******************************************************************************
++ * asm-i386/mach-xen/asm/xenoprof.h
++ *
++ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ */
++#ifndef __ASM_XENOPROF_H__
++#define __ASM_XENOPROF_H__
++#ifdef CONFIG_XEN
+
-+/* to align the pointer to the (next) page boundary */
-+#define PAGE_ALIGN(addr) (((addr)+PAGE_SIZE-1)&PAGE_MASK)
++struct super_block;
++struct dentry;
++int xenoprof_create_files(struct super_block * sb, struct dentry * root);
++#define HAVE_XENOPROF_CREATE_FILES
++
++struct xenoprof_init;
++void xenoprof_arch_init_counter(struct xenoprof_init *init);
++void xenoprof_arch_counter(void);
++void xenoprof_arch_start(void);
++void xenoprof_arch_stop(void);
+
++struct xenoprof_arch_shared_buffer {
++ /* nothing */
++};
++struct xenoprof_shared_buffer;
++void xenoprof_arch_unmap_shared_buffer(struct xenoprof_shared_buffer* sbuf);
++struct xenoprof_get_buffer;
++int xenoprof_arch_map_shared_buffer(struct xenoprof_get_buffer* get_buffer, struct xenoprof_shared_buffer* sbuf);
++struct xenoprof_passive;
++int xenoprof_arch_set_passive(struct xenoprof_passive* pdomain, struct xenoprof_shared_buffer* sbuf);
++
++#endif /* CONFIG_XEN */
++#endif /* __ASM_XENOPROF_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/irq_vectors.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/irq_vectors.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,131 @@
+/*
-+ * This handles the memory map.. We could make this a config
-+ * option, but too many people screw it up, and too few need
-+ * it.
++ * This file should contain #defines for all of the interrupt vector
++ * numbers used by this architecture.
+ *
-+ * A __PAGE_OFFSET of 0xC0000000 means that the kernel has
-+ * a virtual address space of one gigabyte, which limits the
-+ * amount of physical memory you can use to about 950MB.
++ * In addition, there are some standard defines:
+ *
-+ * If you want more physical memory than this then see the CONFIG_HIGHMEM4G
-+ * and CONFIG_HIGHMEM64G options in the kernel configuration.
-+ */
-+
-+#ifndef __ASSEMBLY__
-+
-+struct vm_area_struct;
++ * FIRST_EXTERNAL_VECTOR:
++ * The first free place for external interrupts
++ *
++ * SYSCALL_VECTOR:
++ * The IRQ vector a syscall makes the user to kernel transition
++ * under.
++ *
++ * TIMER_IRQ:
++ * The IRQ number the timer interrupt comes in at.
++ *
++ * NR_IRQS:
++ * The total number of interrupt vectors (including all the
++ * architecture specific interrupts) needed.
++ *
++ */
++#ifndef _ASM_IRQ_VECTORS_H
++#define _ASM_IRQ_VECTORS_H
+
+/*
-+ * This much address space is reserved for vmalloc() and iomap()
-+ * as well as fixmap mappings.
++ * IDT vectors usable for external interrupt sources start
++ * at 0x20:
+ */
-+extern unsigned int __VMALLOC_RESERVE;
-+
-+extern int sysctl_legacy_va_layout;
++#define FIRST_EXTERNAL_VECTOR 0x20
+
-+extern int page_is_ram(unsigned long pagenr);
++#define SYSCALL_VECTOR 0x80
+
-+#endif /* __ASSEMBLY__ */
++/*
++ * Vectors 0x20-0x2f are used for ISA interrupts.
++ */
+
-+#ifdef __ASSEMBLY__
-+#define __PAGE_OFFSET CONFIG_PAGE_OFFSET
-+#define __PHYSICAL_START CONFIG_PHYSICAL_START
-+#else
-+#define __PAGE_OFFSET ((unsigned long)CONFIG_PAGE_OFFSET)
-+#define __PHYSICAL_START ((unsigned long)CONFIG_PHYSICAL_START)
-+#endif
-+#define __KERNEL_START (__PAGE_OFFSET + __PHYSICAL_START)
++#if 0
++/*
++ * Special IRQ vectors used by the SMP architecture, 0xf0-0xff
++ *
++ * some of the following vectors are 'rare', they are merged
++ * into a single vector (CALL_FUNCTION_VECTOR) to save vector space.
++ * TLB, reschedule and local APIC vectors are performance-critical.
++ *
++ * Vectors 0xf0-0xfa are free (reserved for future Linux use).
++ */
++#define SPURIOUS_APIC_VECTOR 0xff
++#define ERROR_APIC_VECTOR 0xfe
++#define INVALIDATE_TLB_VECTOR 0xfd
++#define RESCHEDULE_VECTOR 0xfc
++#define CALL_FUNCTION_VECTOR 0xfb
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+#undef LOAD_OFFSET
-+#define LOAD_OFFSET 0
++#define THERMAL_APIC_VECTOR 0xf0
++/*
++ * Local APIC timer IRQ vector is on a different priority level,
++ * to work around the 'lost local interrupt if more than 2 IRQ
++ * sources per level' errata.
++ */
++#define LOCAL_TIMER_VECTOR 0xef
+#endif
+
-+#define PAGE_OFFSET ((unsigned long)__PAGE_OFFSET)
-+#define VMALLOC_RESERVE ((unsigned long)__VMALLOC_RESERVE)
-+#define MAXMEM (__FIXADDR_TOP-__PAGE_OFFSET-__VMALLOC_RESERVE)
-+#define __pa(x) ((unsigned long)(x)-PAGE_OFFSET)
-+#define __va(x) ((void *)((unsigned long)(x)+PAGE_OFFSET))
-+#define pfn_to_kaddr(pfn) __va((pfn) << PAGE_SHIFT)
-+#ifdef CONFIG_FLATMEM
-+#define pfn_valid(pfn) ((pfn) < max_mapnr)
-+#endif /* CONFIG_FLATMEM */
-+#define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT)
-+
-+#define virt_addr_valid(kaddr) pfn_valid(__pa(kaddr) >> PAGE_SHIFT)
-+
-+#define VM_DATA_DEFAULT_FLAGS \
-+ (VM_READ | VM_WRITE | \
-+ ((current->personality & READ_IMPLIES_EXEC) ? VM_EXEC : 0 ) | \
-+ VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC)
-+
-+#include <asm-generic/memory_model.h>
-+#include <asm-generic/page.h>
++#define SPURIOUS_APIC_VECTOR 0xff
++#define ERROR_APIC_VECTOR 0xfe
+
-+#define __HAVE_ARCH_GATE_AREA 1
-+#endif /* __KERNEL__ */
++/*
++ * First APIC vector available to drivers: (vectors 0x30-0xee)
++ * we start at 0x31 to spread out vectors evenly between priority
++ * levels. (0x80 is the syscall vector)
++ */
++#define FIRST_DEVICE_VECTOR 0x31
++#define FIRST_SYSTEM_VECTOR 0xef
+
-+#endif /* _I386_PAGE_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/param.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/param.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/param.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/param.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,23 @@
-+#ifndef _ASMi386_PARAM_H
-+#define _ASMi386_PARAM_H
++/*
++ * 16 8259A IRQ's, 208 potential APIC interrupt sources.
++ * Right now the APIC is mostly only used for SMP.
++ * 256 vectors is an architectural limit. (we can have
++ * more than 256 devices theoretically, but they will
++ * have to use shared interrupts)
++ * Since vectors 0x00-0x1f are used/reserved for the CPU,
++ * the usable vector space is 0x20-0xff (224 vectors)
++ */
+
-+#ifdef __KERNEL__
-+# define HZ CONFIG_HZ /* Internal kernel timer frequency */
-+# define USER_HZ 100 /* .. some user interfaces are in "ticks" */
-+# define CLOCKS_PER_SEC (USER_HZ) /* like times() */
-+#endif
++#define RESCHEDULE_VECTOR 0
++#define CALL_FUNCTION_VECTOR 1
++#define NR_IPIS 2
+
-+#ifndef HZ
-+#define HZ 100
-+#endif
++/*
++ * The maximum number of vectors supported by i386 processors
++ * is limited to 256. For processors other than i386, NR_VECTORS
++ * should be changed accordingly.
++ */
++#define NR_VECTORS 256
+
-+#define EXEC_PAGESIZE 4096
++#define FPU_IRQ 13
+
-+#ifndef NOGROUP
-+#define NOGROUP (-1)
-+#endif
++#define FIRST_VM86_IRQ 3
++#define LAST_VM86_IRQ 15
++#define invalid_vm86_irq(irq) ((irq) < 3 || (irq) > 15)
+
-+#define MAXHOSTNAMELEN 64 /* max length of hostname */
-+#define COMMAND_LINE_SIZE 256
++/*
++ * The flat IRQ space is divided into two regions:
++ * 1. A one-to-one mapping of real physical IRQs. This space is only used
++ * if we have physical device-access privilege. This region is at the
++ * start of the IRQ space so that existing device drivers do not need
++ * to be modified to translate physical IRQ numbers into our IRQ space.
++ * 3. A dynamic mapping of inter-domain and Xen-sourced virtual IRQs. These
++ * are bound using the provided bind/unbind functions.
++ */
+
++#define PIRQ_BASE 0
++#if !defined(MAX_IO_APICS)
++# define NR_PIRQS (NR_VECTORS + 32 * NR_CPUS)
++#elif NR_CPUS < MAX_IO_APICS
++# define NR_PIRQS (NR_VECTORS + 32 * NR_CPUS)
++#else
++# define NR_PIRQS (NR_VECTORS + 32 * MAX_IO_APICS)
+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pci.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pci.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pci.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pci.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,148 @@
-+#ifndef __i386_PCI_H
-+#define __i386_PCI_H
+
++#define DYNIRQ_BASE (PIRQ_BASE + NR_PIRQS)
++#define NR_DYNIRQS 256
+
-+#ifdef __KERNEL__
-+#include <linux/mm.h> /* for struct page */
-+
-+/* Can be used to override the logic in pci_scan_bus for skipping
-+ already-configured bus numbers - to be used for buggy BIOSes
-+ or architectures with incomplete PCI setup by the loader */
++#define NR_IRQS (NR_PIRQS + NR_DYNIRQS)
++#define NR_IRQ_VECTORS NR_IRQS
+
-+#ifdef CONFIG_PCI
-+extern unsigned int pcibios_assign_all_busses(void);
-+#else
-+#define pcibios_assign_all_busses() 0
-+#endif
++#define pirq_to_irq(_x) ((_x) + PIRQ_BASE)
++#define irq_to_pirq(_x) ((_x) - PIRQ_BASE)
+
-+#include <asm/hypervisor.h>
-+#define pcibios_scan_all_fns(a, b) (!is_initial_xendomain())
++#define dynirq_to_irq(_x) ((_x) + DYNIRQ_BASE)
++#define irq_to_dynirq(_x) ((_x) - DYNIRQ_BASE)
+
-+extern unsigned long pci_mem_start;
-+#define PCIBIOS_MIN_IO 0x1000
-+#define PCIBIOS_MIN_MEM (pci_mem_start)
++#endif /* _ASM_IRQ_VECTORS_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/mach_traps.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/mach_traps.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,33 @@
++/*
++ * include/asm-xen/asm-i386/mach-xen/mach_traps.h
++ *
++ * Machine specific NMI handling for Xen
++ */
++#ifndef _MACH_TRAPS_H
++#define _MACH_TRAPS_H
+
-+#define PCIBIOS_MIN_CARDBUS_IO 0x4000
++#include <linux/bitops.h>
++#include <xen/interface/nmi.h>
+
-+void pcibios_config_init(void);
-+struct pci_bus * pcibios_scan_root(int bus);
++static inline void clear_mem_error(unsigned char reason) {}
++static inline void clear_io_check_error(unsigned char reason) {}
+
-+void pcibios_set_master(struct pci_dev *dev);
-+void pcibios_penalize_isa_irq(int irq, int active);
-+struct irq_routing_table *pcibios_get_irq_routing_table(void);
-+int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq);
++static inline unsigned char get_nmi_reason(void)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ unsigned char reason = 0;
+
-+/* Dynamic DMA mapping stuff.
-+ * i386 has everything mapped statically.
-+ */
++ /* construct a value which looks like it came from
++ * port 0x61.
++ */
++ if (test_bit(_XEN_NMIREASON_io_error, &s->arch.nmi_reason))
++ reason |= 0x40;
++ if (test_bit(_XEN_NMIREASON_parity_error, &s->arch.nmi_reason))
++ reason |= 0x80;
+
-+#include <linux/types.h>
-+#include <linux/slab.h>
-+#include <asm/scatterlist.h>
-+#include <linux/string.h>
-+#include <asm/io.h>
++ return reason;
++}
+
-+struct pci_dev;
++static inline void reassert_nmi(void) {}
+
-+#ifdef CONFIG_SWIOTLB
++#endif /* !_MACH_TRAPS_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/mach-xen/setup_arch.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-i386/mach-xen/setup_arch.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,5 @@
++/* Hook to call BIOS initialisation function */
+
++#define ARCH_SETUP machine_specific_arch_setup();
+
-+/* On Xen we use SWIOTLB instead of blk-specific bounce buffers. */
-+#define PCI_DMA_BUS_IS_PHYS (0)
++void __init machine_specific_arch_setup(void);
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/page.h
+--- a/include/asm-i386/page.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-i386/page.h Wed Sep 10 10:54:08 2008 +0100
+@@ -122,7 +122,7 @@
+
+ #define PAGE_OFFSET ((unsigned long)__PAGE_OFFSET)
+ #define VMALLOC_RESERVE ((unsigned long)__VMALLOC_RESERVE)
+-#define MAXMEM (-__PAGE_OFFSET-__VMALLOC_RESERVE)
++#define MAXMEM (__FIXADDR_TOP-__PAGE_OFFSET-__VMALLOC_RESERVE)
+ #define __pa(x) ((unsigned long)(x)-PAGE_OFFSET)
+ #define __va(x) ((void *)((unsigned long)(x)+PAGE_OFFSET))
+ #define pfn_to_kaddr(pfn) __va((pfn) << PAGE_SHIFT)
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/pgtable-2level-defs.h
+--- a/include/asm-i386/pgtable-2level-defs.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-i386/pgtable-2level-defs.h Wed Sep 10 10:54:08 2008 +0100
+@@ -1,5 +1,7 @@
+ #ifndef _I386_PGTABLE_2LEVEL_DEFS_H
+ #define _I386_PGTABLE_2LEVEL_DEFS_H
+
-+#define DECLARE_PCI_UNMAP_ADDR(ADDR_NAME) \
-+ dma_addr_t ADDR_NAME;
-+#define DECLARE_PCI_UNMAP_LEN(LEN_NAME) \
-+ __u32 LEN_NAME;
-+#define pci_unmap_addr(PTR, ADDR_NAME) \
-+ ((PTR)->ADDR_NAME)
-+#define pci_unmap_addr_set(PTR, ADDR_NAME, VAL) \
-+ (((PTR)->ADDR_NAME) = (VAL))
-+#define pci_unmap_len(PTR, LEN_NAME) \
-+ ((PTR)->LEN_NAME)
-+#define pci_unmap_len_set(PTR, LEN_NAME, VAL) \
-+ (((PTR)->LEN_NAME) = (VAL))
++#define HAVE_SHARED_KERNEL_PMD 0
+
+ /*
+ * traditional i386 two-level paging structure:
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-i386/pgtable-3level-defs.h
+--- a/include/asm-i386/pgtable-3level-defs.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-i386/pgtable-3level-defs.h Wed Sep 10 10:54:08 2008 +0100
+@@ -1,5 +1,7 @@
+ #ifndef _I386_PGTABLE_3LEVEL_DEFS_H
+ #define _I386_PGTABLE_3LEVEL_DEFS_H
+
++#define HAVE_SHARED_KERNEL_PMD 1
+
+ /*
+ * PGDIR_SHIFT determines what a top-level page table entry can map
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/agp.h
+--- a/include/asm-ia64/agp.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/agp.h Wed Sep 10 10:54:08 2008 +0100
+@@ -19,13 +19,44 @@
+ #define flush_agp_cache() mb()
+
+ /* Convert a physical address to an address suitable for the GART. */
++#ifndef CONFIG_XEN
+ #define phys_to_gart(x) (x)
+ #define gart_to_phys(x) (x)
++#else
++#define phys_to_gart(x) phys_to_machine_for_dma(x)
++#define gart_to_phys(x) machine_to_phys_for_dma(x)
++#endif
+
+ /* GATT allocation. Returns/accepts GATT kernel virtual address. */
++#ifndef CONFIG_XEN
+ #define alloc_gatt_pages(order) \
+ ((char *)__get_free_pages(GFP_KERNEL, (order)))
+ #define free_gatt_pages(table, order) \
+ free_pages((unsigned long)(table), (order))
+#else
++#include <asm/hypervisor.h>
++static inline char*
++alloc_gatt_pages(unsigned int order)
++{
++ unsigned long error;
++ unsigned long ret = __get_free_pages(GFP_KERNEL, (order));
++ if (ret == 0) {
++ goto out;
++ }
++ error = xen_create_contiguous_region(ret, order, 0);
++ if (error) {
++ free_pages(ret, order);
++ ret = 0;
++ }
++out:
++ return (char*)ret;
++}
++static inline void
++free_gatt_pages(void* table, unsigned int order)
++{
++ xen_destroy_contiguous_region((unsigned long)table, order);
++ free_pages((unsigned long)table, order);
++}
++#endif /* CONFIG_XEN */
+
+ #endif /* _ASM_IA64_AGP_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/gcc_intrin.h
+--- a/include/asm-ia64/gcc_intrin.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/gcc_intrin.h Wed Sep 10 10:54:08 2008 +0100
+@@ -26,7 +26,7 @@
+
+ register unsigned long ia64_r13 asm ("r13") __attribute_used__;
+
+-#define ia64_setreg(regnum, val) \
++#define __ia64_setreg(regnum, val) \
+ ({ \
+ switch (regnum) { \
+ case _IA64_REG_PSR_L: \
+@@ -55,7 +55,7 @@
+ } \
+ })
+
+-#define ia64_getreg(regnum) \
++#define __ia64_getreg(regnum) \
+ ({ \
+ __u64 ia64_intri_res; \
+ \
+@@ -92,7 +92,7 @@
+
+ #define ia64_hint_pause 0
+
+-#define ia64_hint(mode) \
++#define __ia64_hint(mode) \
+ ({ \
+ switch (mode) { \
+ case ia64_hint_pause: \
+@@ -374,7 +374,7 @@
+
+ #define ia64_invala() asm volatile ("invala" ::: "memory")
+
+-#define ia64_thash(addr) \
++#define __ia64_thash(addr) \
+ ({ \
+ __u64 ia64_intri_res; \
+ asm volatile ("thash %0=%1" : "=r"(ia64_intri_res) : "r" (addr)); \
+@@ -394,18 +394,18 @@
+
+ #define ia64_nop(x) asm volatile ("nop %0"::"i"(x));
+
+-#define ia64_itci(addr) asm volatile ("itc.i %0;;" :: "r"(addr) : "memory")
++#define __ia64_itci(addr) asm volatile ("itc.i %0;;" :: "r"(addr) : "memory")
+
+-#define ia64_itcd(addr) asm volatile ("itc.d %0;;" :: "r"(addr) : "memory")
++#define __ia64_itcd(addr) asm volatile ("itc.d %0;;" :: "r"(addr) : "memory")
+
+
+-#define ia64_itri(trnum, addr) asm volatile ("itr.i itr[%0]=%1" \
++#define __ia64_itri(trnum, addr) asm volatile ("itr.i itr[%0]=%1" \
+ :: "r"(trnum), "r"(addr) : "memory")
+
+-#define ia64_itrd(trnum, addr) asm volatile ("itr.d dtr[%0]=%1" \
++#define __ia64_itrd(trnum, addr) asm volatile ("itr.d dtr[%0]=%1" \
+ :: "r"(trnum), "r"(addr) : "memory")
+
+-#define ia64_tpa(addr) \
++#define __ia64_tpa(addr) \
+ ({ \
+ __u64 ia64_pa; \
+ asm volatile ("tpa %0 = %1" : "=r"(ia64_pa) : "r"(addr) : "memory"); \
+@@ -415,22 +415,22 @@
+ #define __ia64_set_dbr(index, val) \
+ asm volatile ("mov dbr[%0]=%1" :: "r"(index), "r"(val) : "memory")
+
+-#define ia64_set_ibr(index, val) \
++#define __ia64_set_ibr(index, val) \
+ asm volatile ("mov ibr[%0]=%1" :: "r"(index), "r"(val) : "memory")
+
+-#define ia64_set_pkr(index, val) \
++#define __ia64_set_pkr(index, val) \
+ asm volatile ("mov pkr[%0]=%1" :: "r"(index), "r"(val) : "memory")
+
+-#define ia64_set_pmc(index, val) \
++#define __ia64_set_pmc(index, val) \
+ asm volatile ("mov pmc[%0]=%1" :: "r"(index), "r"(val) : "memory")
+
+-#define ia64_set_pmd(index, val) \
++#define __ia64_set_pmd(index, val) \
+ asm volatile ("mov pmd[%0]=%1" :: "r"(index), "r"(val) : "memory")
+
+-#define ia64_set_rr(index, val) \
++#define __ia64_set_rr(index, val) \
+ asm volatile ("mov rr[%0]=%1" :: "r"(index), "r"(val) : "memory");
+
+-#define ia64_get_cpuid(index) \
++#define __ia64_get_cpuid(index) \
+ ({ \
+ __u64 ia64_intri_res; \
+ asm volatile ("mov %0=cpuid[%r1]" : "=r"(ia64_intri_res) : "rO"(index)); \
+@@ -444,21 +444,21 @@
+ ia64_intri_res; \
+ })
+
+-#define ia64_get_ibr(index) \
++#define __ia64_get_ibr(index) \
+ ({ \
+ __u64 ia64_intri_res; \
+ asm volatile ("mov %0=ibr[%1]" : "=r"(ia64_intri_res) : "r"(index)); \
+ ia64_intri_res; \
+ })
+
+-#define ia64_get_pkr(index) \
++#define __ia64_get_pkr(index) \
+ ({ \
+ __u64 ia64_intri_res; \
+ asm volatile ("mov %0=pkr[%1]" : "=r"(ia64_intri_res) : "r"(index)); \
+ ia64_intri_res; \
+ })
+
+-#define ia64_get_pmc(index) \
++#define __ia64_get_pmc(index) \
+ ({ \
+ __u64 ia64_intri_res; \
+ asm volatile ("mov %0=pmc[%1]" : "=r"(ia64_intri_res) : "r"(index)); \
+@@ -466,48 +466,48 @@
+ })
+
+
+-#define ia64_get_pmd(index) \
++#define __ia64_get_pmd(index) \
+ ({ \
+ __u64 ia64_intri_res; \
+ asm volatile ("mov %0=pmd[%1]" : "=r"(ia64_intri_res) : "r"(index)); \
+ ia64_intri_res; \
+ })
+
+-#define ia64_get_rr(index) \
++#define __ia64_get_rr(index) \
+ ({ \
+ __u64 ia64_intri_res; \
+ asm volatile ("mov %0=rr[%1]" : "=r"(ia64_intri_res) : "r" (index)); \
+ ia64_intri_res; \
+ })
+
+-#define ia64_fc(addr) asm volatile ("fc %0" :: "r"(addr) : "memory")
++#define __ia64_fc(addr) asm volatile ("fc %0" :: "r"(addr) : "memory")
+
+
+ #define ia64_sync_i() asm volatile (";; sync.i" ::: "memory")
+
+-#define ia64_ssm(mask) asm volatile ("ssm %0":: "i"((mask)) : "memory")
+-#define ia64_rsm(mask) asm volatile ("rsm %0":: "i"((mask)) : "memory")
++#define __ia64_ssm(mask) asm volatile ("ssm %0":: "i"((mask)) : "memory")
++#define __ia64_rsm(mask) asm volatile ("rsm %0":: "i"((mask)) : "memory")
+ #define ia64_sum(mask) asm volatile ("sum %0":: "i"((mask)) : "memory")
+ #define ia64_rum(mask) asm volatile ("rum %0":: "i"((mask)) : "memory")
+
+-#define ia64_ptce(addr) asm volatile ("ptc.e %0" :: "r"(addr))
++#define __ia64_ptce(addr) asm volatile ("ptc.e %0" :: "r"(addr))
+
+-#define ia64_ptcga(addr, size) \
++#define __ia64_ptcga(addr, size) \
+ do { \
+ asm volatile ("ptc.ga %0,%1" :: "r"(addr), "r"(size) : "memory"); \
+ ia64_dv_serialize_data(); \
+ } while (0)
+
+-#define ia64_ptcl(addr, size) \
++#define __ia64_ptcl(addr, size) \
+ do { \
+ asm volatile ("ptc.l %0,%1" :: "r"(addr), "r"(size) : "memory"); \
+ ia64_dv_serialize_data(); \
+ } while (0)
+
+-#define ia64_ptri(addr, size) \
++#define __ia64_ptri(addr, size) \
+ asm volatile ("ptr.i %0,%1" :: "r"(addr), "r"(size) : "memory")
+
+-#define ia64_ptrd(addr, size) \
++#define __ia64_ptrd(addr, size) \
+ asm volatile ("ptr.d %0,%1" :: "r"(addr), "r"(size) : "memory")
+
+ /* Values for lfhint in ia64_lfetch and ia64_lfetch_fault */
+@@ -589,7 +589,7 @@
+ } \
+ })
+
+-#define ia64_intrin_local_irq_restore(x) \
++#define __ia64_intrin_local_irq_restore(x) \
+ do { \
+ asm volatile (";; cmp.ne p6,p7=%0,r0;;" \
+ "(p6) ssm psr.i;" \
+@@ -598,4 +598,6 @@
+ :: "r"((x)) : "p6", "p7", "memory"); \
+ } while (0)
+
++#define __ia64_get_psr_i() (__ia64_getreg(_IA64_REG_PSR) & 0x4000UL)
+
-+/* The PCI address space does equal the physical memory
-+ * address space. The networking and block device layers use
-+ * this boolean for bounce buffer decisions.
+ #endif /* _ASM_IA64_GCC_INTRIN_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/gnttab_dma.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/gnttab_dma.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,51 @@
++/*
++ * Copyright (c) 2007 Herbert Xu <herbert at gondor.apana.org.au>
++ * Copyright (c) 2007 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ */
-+#define PCI_DMA_BUS_IS_PHYS (1)
-+
-+/* pci_unmap_{page,single} is a nop so... */
-+#define DECLARE_PCI_UNMAP_ADDR(ADDR_NAME)
-+#define DECLARE_PCI_UNMAP_LEN(LEN_NAME)
-+#define pci_unmap_addr(PTR, ADDR_NAME) (0)
-+#define pci_unmap_addr_set(PTR, ADDR_NAME, VAL) do { } while (0)
-+#define pci_unmap_len(PTR, LEN_NAME) (0)
-+#define pci_unmap_len_set(PTR, LEN_NAME, VAL) do { } while (0)
-+
-+#endif
-+
-+/* This is always fine. */
-+#define pci_dac_dma_supported(pci_dev, mask) (1)
+
-+static inline dma64_addr_t
-+pci_dac_page_to_dma(struct pci_dev *pdev, struct page *page, unsigned long offset, int direction)
-+{
-+ return ((dma64_addr_t) page_to_phys(page) +
-+ (dma64_addr_t) offset);
-+}
++#ifndef _ASM_IA64_GNTTAB_DMA_H
++#define _ASM_IA64_GNTTAB_DMA_H
+
-+static inline struct page *
-+pci_dac_dma_to_page(struct pci_dev *pdev, dma64_addr_t dma_addr)
++static inline int gnttab_dma_local_pfn(struct page *page)
+{
-+ return pfn_to_page(dma_addr >> PAGE_SHIFT);
++ return 0;
+}
+
-+static inline unsigned long
-+pci_dac_dma_to_offset(struct pci_dev *pdev, dma64_addr_t dma_addr)
++/* caller must get dma address after calling this function */
++static inline void gnttab_dma_use_page(struct page *page)
+{
-+ return (dma_addr & ~PAGE_MASK);
++ __gnttab_dma_map_page(page);
+}
+
-+static inline void
-+pci_dac_dma_sync_single_for_cpu(struct pci_dev *pdev, dma64_addr_t dma_addr, size_t len, int direction)
++static inline dma_addr_t gnttab_dma_map_page(struct page *page)
+{
++ gnttab_dma_use_page(page);
++ return page_to_bus(page);
+}
+
-+static inline void
-+pci_dac_dma_sync_single_for_device(struct pci_dev *pdev, dma64_addr_t dma_addr, size_t len, int direction)
++static inline dma_addr_t gnttab_dma_map_virt(void *ptr)
+{
-+ flush_write_buffers();
++ return gnttab_dma_map_page(virt_to_page(ptr)) + offset_in_page(ptr);
+}
+
-+#define HAVE_PCI_MMAP
-+extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma,
-+ enum pci_mmap_state mmap_state, int write_combine);
-+
-+
-+static inline void pcibios_add_platform_entries(struct pci_dev *dev)
++static inline void gnttab_dma_unmap_page(dma_addr_t dma_address)
+{
++ __gnttab_dma_unmap_page(virt_to_page(bus_to_virt(dma_address)));
+}
+
-+#ifdef CONFIG_PCI
-+static inline void pci_dma_burst_advice(struct pci_dev *pdev,
-+ enum pci_dma_burst_strategy *strat,
-+ unsigned long *strategy_parameter)
-+{
-+ *strat = PCI_DMA_BURST_INFINITY;
-+ *strategy_parameter = ~0UL;
-+}
++#endif /* _ASM_IA64_GNTTAB_DMA_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/hw_irq.h
+--- a/include/asm-ia64/hw_irq.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/hw_irq.h Wed Sep 10 10:54:08 2008 +0100
+@@ -15,7 +15,11 @@
+ #include <asm/ptrace.h>
+ #include <asm/smp.h>
+
++#ifndef CONFIG_XEN
+ typedef u8 ia64_vector;
++#else
++typedef u16 ia64_vector;
+#endif
-+
-+#endif /* __KERNEL__ */
-+
-+#ifdef CONFIG_XEN_PCIDEV_FRONTEND
-+#include <xen/pcifront.h>
-+#endif /* CONFIG_XEN_PCIDEV_FRONTEND */
-+
-+/* implement the pci_ DMA API in terms of the generic device dma_ one */
-+#include <asm-generic/pci-dma-compat.h>
-+
-+/* generic pci stuff */
-+#include <asm-generic/pci.h>
-+
-+#endif /* __i386_PCI_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pgalloc.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pgalloc.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pgalloc.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pgalloc.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,59 @@
-+#ifndef _I386_PGALLOC_H
-+#define _I386_PGALLOC_H
-+
-+#include <asm/fixmap.h>
-+#include <linux/threads.h>
-+#include <linux/mm.h> /* for struct page */
-+#include <asm/io.h> /* for phys_to_virt and page_to_pseudophys */
-+
-+#define pmd_populate_kernel(mm, pmd, pte) \
-+ set_pmd(pmd, __pmd(_PAGE_TABLE + __pa(pte)))
-+
-+#define pmd_populate(mm, pmd, pte) \
-+do { \
-+ unsigned long pfn = page_to_pfn(pte); \
-+ if (test_bit(PG_pinned, &virt_to_page((mm)->pgd)->flags)) { \
-+ if (!PageHighMem(pte)) \
-+ BUG_ON(HYPERVISOR_update_va_mapping( \
-+ (unsigned long)__va(pfn << PAGE_SHIFT), \
-+ pfn_pte(pfn, PAGE_KERNEL_RO), 0)); \
-+ else if (!test_and_set_bit(PG_pinned, &pte->flags)) \
-+ kmap_flush_unused(); \
-+ set_pmd(pmd, \
-+ __pmd(_PAGE_TABLE + ((paddr_t)pfn << PAGE_SHIFT))); \
-+ } else \
-+ *(pmd) = __pmd(_PAGE_TABLE + ((paddr_t)pfn << PAGE_SHIFT)); \
-+} while (0)
-+
-+/*
-+ * Allocate and free page tables.
+
+ /*
+ * 0 special
+@@ -99,6 +103,12 @@
+
+ static inline void ia64_resend_irq(unsigned int vector)
+ {
++#ifdef CONFIG_XEN
++ extern int resend_irq_on_evtchn(unsigned int i);
++ if (is_running_on_xen())
++ resend_irq_on_evtchn(vector);
++ else
++#endif /* CONFIG_XEN */
+ platform_send_ipi(smp_processor_id(), vector, IA64_IPI_DM_INT, 0);
+ }
+
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/hypercall.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/hypercall.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,458 @@
++/******************************************************************************
++ * hypercall.h
++ *
++ * Linux-specific hypervisor handling.
++ *
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
-+extern pgd_t *pgd_alloc(struct mm_struct *);
-+extern void pgd_free(pgd_t *pgd);
-+
-+extern pte_t *pte_alloc_one_kernel(struct mm_struct *, unsigned long);
-+extern struct page *pte_alloc_one(struct mm_struct *, unsigned long);
-+
-+static inline void pte_free_kernel(pte_t *pte)
-+{
-+ free_page((unsigned long)pte);
-+ make_lowmem_page_writable(pte, XENFEAT_writable_page_tables);
-+}
-+
-+extern void pte_free(struct page *pte);
+
-+#define __pte_free_tlb(tlb,pte) tlb_remove_page((tlb),(pte))
++#ifndef __HYPERCALL_H__
++#define __HYPERCALL_H__
+
-+#ifdef CONFIG_X86_PAE
-+/*
-+ * In the PAE case we free the pmds as part of the pgd.
-+ */
-+#define pmd_alloc_one(mm, addr) ({ BUG(); ((pmd_t *)2); })
-+#define pmd_free(x) do { } while (0)
-+#define __pmd_free_tlb(tlb,x) do { } while (0)
-+#define pud_populate(mm, pmd, pte) BUG()
++#ifndef __HYPERVISOR_H__
++# error "please don't include this file directly"
+#endif
+
-+#define check_pgt_cache() do { } while (0)
-+
-+#endif /* _I386_PGALLOC_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pgtable-2level-defs.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pgtable-2level-defs.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pgtable-2level-defs.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pgtable-2level-defs.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,20 @@
-+#ifndef _I386_PGTABLE_2LEVEL_DEFS_H
-+#define _I386_PGTABLE_2LEVEL_DEFS_H
-+
-+#define HAVE_SHARED_KERNEL_PMD 0
-+
-+/*
-+ * traditional i386 two-level paging structure:
-+ */
-+
-+#define PGDIR_SHIFT 22
-+#define PTRS_PER_PGD 1024
-+
-+/*
-+ * the i386 is two-level, so we don't really have any
-+ * PMD directory physically.
-+ */
-+
-+#define PTRS_PER_PTE 1024
-+
-+#endif /* _I386_PGTABLE_2LEVEL_DEFS_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pgtable-2level.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pgtable-2level.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pgtable-2level.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pgtable-2level.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,120 @@
-+#ifndef _I386_PGTABLE_2LEVEL_H
-+#define _I386_PGTABLE_2LEVEL_H
-+
-+#include <asm-generic/pgtable-nopmd.h>
-+
-+#define pte_ERROR(e) \
-+ printk("%s:%d: bad pte %08lx (pfn %05lx).\n", __FILE__, __LINE__, \
-+ __pte_val(e), pte_pfn(e))
-+#define pgd_ERROR(e) \
-+ printk("%s:%d: bad pgd %08lx (pfn %05lx).\n", __FILE__, __LINE__, \
-+ __pgd_val(e), pgd_val(e) >> PAGE_SHIFT)
++#include <asm/xen/xcom_hcall.h>
++struct xencomm_handle;
++extern unsigned long __hypercall(unsigned long a1, unsigned long a2,
++ unsigned long a3, unsigned long a4,
++ unsigned long a5, unsigned long cmd);
+
+/*
-+ * Certain architectures need to do special things when PTEs
-+ * within a page table are directly modified. Thus, the following
-+ * hook is made available.
++ * Assembler stubs for hyper-calls.
+ */
-+#define set_pte(pteptr, pteval) (*(pteptr) = pteval)
-+
-+#define set_pte_at(_mm,addr,ptep,pteval) do { \
-+ if (((_mm) != current->mm && (_mm) != &init_mm) || \
-+ HYPERVISOR_update_va_mapping((addr), (pteval), 0)) \
-+ set_pte((ptep), (pteval)); \
-+} while (0)
+
-+#define set_pte_at_sync(_mm,addr,ptep,pteval) do { \
-+ if (((_mm) != current->mm && (_mm) != &init_mm) || \
-+ HYPERVISOR_update_va_mapping((addr), (pteval), UVMF_INVLPG)) { \
-+ set_pte((ptep), (pteval)); \
-+ xen_invlpg((addr)); \
-+ } \
-+} while (0)
-+
-+#define set_pte_atomic(pteptr, pteval) set_pte(pteptr,pteval)
-+
-+#define set_pmd(pmdptr, pmdval) xen_l2_entry_update((pmdptr), (pmdval))
-+
-+#define pte_clear(mm,addr,xp) do { set_pte_at(mm, addr, xp, __pte(0)); } while (0)
-+#define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0)
-+
-+#define pte_none(x) (!(x).pte_low)
++#define _hypercall0(type, name) \
++({ \
++ long __res; \
++ __res=__hypercall(0, 0, 0, 0, 0, __HYPERVISOR_##name); \
++ (type)__res; \
++})
+
-+static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
-+{
-+ pte_t pte = *ptep;
-+ if (!pte_none(pte)) {
-+ if (mm != &init_mm)
-+ pte = __pte_ma(xchg(&ptep->pte_low, 0));
-+ else
-+ HYPERVISOR_update_va_mapping(addr, __pte(0), 0);
-+ }
-+ return pte;
-+}
++#define _hypercall1(type, name, a1) \
++({ \
++ long __res; \
++ __res = __hypercall((unsigned long)a1, \
++ 0, 0, 0, 0, __HYPERVISOR_##name); \
++ (type)__res; \
++})
+
-+#define ptep_clear_flush(vma, addr, ptep) \
++#define _hypercall2(type, name, a1, a2) \
+({ \
-+ pte_t *__ptep = (ptep); \
-+ pte_t __res = *__ptep; \
-+ if (!pte_none(__res) && \
-+ ((vma)->vm_mm != current->mm || \
-+ HYPERVISOR_update_va_mapping(addr, __pte(0), \
-+ (unsigned long)(vma)->vm_mm->cpu_vm_mask.bits| \
-+ UVMF_INVLPG|UVMF_MULTI))) { \
-+ __ptep->pte_low = 0; \
-+ flush_tlb_page(vma, addr); \
-+ } \
-+ __res; \
++ long __res; \
++ __res = __hypercall((unsigned long)a1, \
++ (unsigned long)a2, \
++ 0, 0, 0, __HYPERVISOR_##name); \
++ (type)__res; \
+})
+
-+#define pte_same(a, b) ((a).pte_low == (b).pte_low)
++#define _hypercall3(type, name, a1, a2, a3) \
++({ \
++ long __res; \
++ __res = __hypercall((unsigned long)a1, \
++ (unsigned long)a2, \
++ (unsigned long)a3, \
++ 0, 0, __HYPERVISOR_##name); \
++ (type)__res; \
++})
+
-+#define __pte_mfn(_pte) ((_pte).pte_low >> PAGE_SHIFT)
-+#define pte_mfn(_pte) ((_pte).pte_low & _PAGE_PRESENT ? \
-+ __pte_mfn(_pte) : pfn_to_mfn(__pte_mfn(_pte)))
-+#define pte_pfn(_pte) ((_pte).pte_low & _PAGE_PRESENT ? \
-+ mfn_to_local_pfn(__pte_mfn(_pte)) : __pte_mfn(_pte))
++#define _hypercall4(type, name, a1, a2, a3, a4) \
++({ \
++ long __res; \
++ __res = __hypercall((unsigned long)a1, \
++ (unsigned long)a2, \
++ (unsigned long)a3, \
++ (unsigned long)a4, \
++ 0, __HYPERVISOR_##name); \
++ (type)__res; \
++})
+
-+#define pte_page(_pte) pfn_to_page(pte_pfn(_pte))
++#define _hypercall5(type, name, a1, a2, a3, a4, a5) \
++({ \
++ long __res; \
++ __res = __hypercall((unsigned long)a1, \
++ (unsigned long)a2, \
++ (unsigned long)a3, \
++ (unsigned long)a4, \
++ (unsigned long)a5, \
++ __HYPERVISOR_##name); \
++ (type)__res; \
++})
+
-+#define pfn_pte(pfn, prot) __pte(((pfn) << PAGE_SHIFT) | pgprot_val(prot))
-+#define pfn_pmd(pfn, prot) __pmd(((pfn) << PAGE_SHIFT) | pgprot_val(prot))
+
-+/*
-+ * All present user pages are user-executable:
-+ */
-+static inline int pte_exec(pte_t pte)
++static inline int
++xencomm_arch_hypercall_sched_op(int cmd, struct xencomm_handle *arg)
+{
-+ return pte_user(pte);
++ return _hypercall2(int, sched_op, cmd, arg);
+}
+
-+/*
-+ * All present pages are kernel-executable:
-+ */
-+static inline int pte_exec_kernel(pte_t pte)
++static inline long
++HYPERVISOR_set_timer_op(u64 timeout)
+{
-+ return 1;
++ unsigned long timeout_hi = (unsigned long)(timeout >> 32);
++ unsigned long timeout_lo = (unsigned long)timeout;
++ return _hypercall2(long, set_timer_op, timeout_lo, timeout_hi);
+}
+
-+/*
-+ * Bits 0, 6 and 7 are taken, split up the 29 bits of offset
-+ * into this range:
-+ */
-+#define PTE_FILE_MAX_BITS 29
-+
-+#define pte_to_pgoff(pte) \
-+ ((((pte).pte_low >> 1) & 0x1f ) + (((pte).pte_low >> 8) << 5 ))
-+
-+#define pgoff_to_pte(off) \
-+ ((pte_t) { (((off) & 0x1f) << 1) + (((off) >> 5) << 8) + _PAGE_FILE })
-+
-+/* Encode and de-code a swap entry */
-+#define __swp_type(x) (((x).val >> 1) & 0x1f)
-+#define __swp_offset(x) ((x).val >> 8)
-+#define __swp_entry(type, offset) ((swp_entry_t) { ((type) << 1) | ((offset) << 8) })
-+#define __pte_to_swp_entry(pte) ((swp_entry_t) { (pte).pte_low })
-+#define __swp_entry_to_pte(x) ((pte_t) { (x).val })
-+
-+void vmalloc_sync_all(void);
-+
-+#endif /* _I386_PGTABLE_2LEVEL_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pgtable-3level-defs.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pgtable-3level-defs.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pgtable-3level-defs.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pgtable-3level-defs.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,24 @@
-+#ifndef _I386_PGTABLE_3LEVEL_DEFS_H
-+#define _I386_PGTABLE_3LEVEL_DEFS_H
-+
-+#define HAVE_SHARED_KERNEL_PMD 0
-+
-+/*
-+ * PGDIR_SHIFT determines what a top-level page table entry can map
-+ */
-+#define PGDIR_SHIFT 30
-+#define PTRS_PER_PGD 4
++static inline int
++xencomm_arch_hypercall_platform_op(struct xencomm_handle *op)
++{
++ return _hypercall1(int, platform_op, op);
++}
+
-+/*
-+ * PMD_SHIFT determines the size of the area a middle-level
-+ * page table can map
-+ */
-+#define PMD_SHIFT 21
-+#define PTRS_PER_PMD 512
++static inline int
++xencomm_arch_hypercall_sysctl(struct xencomm_handle *op)
++{
++ return _hypercall1(int, sysctl, op);
++}
+
-+/*
-+ * entries per page directory level
-+ */
-+#define PTRS_PER_PTE 512
++static inline int
++xencomm_arch_hypercall_domctl(struct xencomm_handle *op)
++{
++ return _hypercall1(int, domctl, op);
++}
+
-+#endif /* _I386_PGTABLE_3LEVEL_DEFS_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pgtable-3level.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pgtable-3level.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pgtable-3level.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pgtable-3level.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,209 @@
-+#ifndef _I386_PGTABLE_3LEVEL_H
-+#define _I386_PGTABLE_3LEVEL_H
++static inline int
++xencomm_arch_hypercall_multicall(struct xencomm_handle *call_list,
++ int nr_calls)
++{
++ return _hypercall2(int, multicall, call_list, nr_calls);
++}
+
-+#include <asm-generic/pgtable-nopud.h>
++static inline int
++xencomm_arch_hypercall_memory_op(unsigned int cmd, struct xencomm_handle *arg)
++{
++ return _hypercall2(int, memory_op, cmd, arg);
++}
+
-+/*
-+ * Intel Physical Address Extension (PAE) Mode - three-level page
-+ * tables on PPro+ CPUs.
-+ *
-+ * Copyright (C) 1999 Ingo Molnar <mingo at redhat.com>
-+ */
++static inline int
++xencomm_arch_hypercall_event_channel_op(int cmd, struct xencomm_handle *arg)
++{
++ return _hypercall2(int, event_channel_op, cmd, arg);
++}
+
-+#define pte_ERROR(e) \
-+ printk("%s:%d: bad pte %p(%016Lx pfn %08lx).\n", __FILE__, __LINE__, \
-+ &(e), __pte_val(e), pte_pfn(e))
-+#define pmd_ERROR(e) \
-+ printk("%s:%d: bad pmd %p(%016Lx pfn %08Lx).\n", __FILE__, __LINE__, \
-+ &(e), __pmd_val(e), (pmd_val(e) & PTE_MASK) >> PAGE_SHIFT)
-+#define pgd_ERROR(e) \
-+ printk("%s:%d: bad pgd %p(%016Lx pfn %08Lx).\n", __FILE__, __LINE__, \
-+ &(e), __pgd_val(e), (pgd_val(e) & PTE_MASK) >> PAGE_SHIFT)
++static inline int
++xencomm_arch_hypercall_xsm_op(struct xencomm_handle *arg)
++{
++ return _hypercall1(int, xsm_op, arg);
++}
+
-+#define pud_none(pud) 0
-+#define pud_bad(pud) 0
-+#define pud_present(pud) 1
++static inline int
++xencomm_arch_hypercall_xen_version(int cmd, struct xencomm_handle *arg)
++{
++ return _hypercall2(int, xen_version, cmd, arg);
++}
+
-+/*
-+ * Is the pte executable?
-+ */
-+static inline int pte_x(pte_t pte)
++static inline int
++xencomm_arch_hypercall_console_io(int cmd, int count,
++ struct xencomm_handle *str)
+{
-+ return !(__pte_val(pte) & _PAGE_NX);
++ return _hypercall3(int, console_io, cmd, count, str);
+}
+
-+/*
-+ * All present user-pages with !NX bit are user-executable:
-+ */
-+static inline int pte_exec(pte_t pte)
++static inline int
++xencomm_arch_hypercall_physdev_op(int cmd, struct xencomm_handle *arg)
+{
-+ return pte_user(pte) && pte_x(pte);
++ return _hypercall2(int, physdev_op, cmd, arg);
+}
-+/*
-+ * All present pages with !NX bit are kernel-executable:
-+ */
-+static inline int pte_exec_kernel(pte_t pte)
++
++static inline int
++xencomm_arch_hypercall_grant_table_op(unsigned int cmd,
++ struct xencomm_handle *uop,
++ unsigned int count)
+{
-+ return pte_x(pte);
++ return _hypercall3(int, grant_table_op, cmd, uop, count);
+}
+
-+/* Rules for using set_pte: the pte being assigned *must* be
-+ * either not present or in a state where the hardware will
-+ * not attempt to update the pte. In places where this is
-+ * not possible, use pte_get_and_clear to obtain the old pte
-+ * value and then use set_pte to update it. -ben
-+ */
-+#define __HAVE_ARCH_SET_PTE_ATOMIC
++int HYPERVISOR_grant_table_op(unsigned int cmd, void *uop, unsigned int count);
+
-+static inline void set_pte(pte_t *ptep, pte_t pte)
++extern int xencomm_arch_hypercall_suspend(struct xencomm_handle *arg);
++
++static inline int
++xencomm_arch_hypercall_callback_op(int cmd, struct xencomm_handle *arg)
+{
-+ ptep->pte_high = pte.pte_high;
-+ smp_wmb();
-+ ptep->pte_low = pte.pte_low;
++ return _hypercall2(int, callback_op, cmd, arg);
+}
-+#define set_pte_atomic(pteptr,pteval) \
-+ set_64bit((unsigned long long *)(pteptr),__pte_val(pteval))
-+
-+#define set_pte_at(_mm,addr,ptep,pteval) do { \
-+ if (((_mm) != current->mm && (_mm) != &init_mm) || \
-+ HYPERVISOR_update_va_mapping((addr), (pteval), 0)) \
-+ set_pte((ptep), (pteval)); \
-+} while (0)
+
-+#define set_pte_at_sync(_mm,addr,ptep,pteval) do { \
-+ if (((_mm) != current->mm && (_mm) != &init_mm) || \
-+ HYPERVISOR_update_va_mapping((addr), (pteval), UVMF_INVLPG)) { \
-+ set_pte((ptep), (pteval)); \
-+ xen_invlpg((addr)); \
-+ } \
-+} while (0)
++static inline unsigned long
++xencomm_arch_hypercall_hvm_op(int cmd, void *arg)
++{
++ return _hypercall2(unsigned long, hvm_op, cmd, arg);
++}
+
-+#define set_pmd(pmdptr,pmdval) \
-+ xen_l2_entry_update((pmdptr), (pmdval))
-+#define set_pud(pudptr,pudval) \
-+ xen_l3_entry_update((pudptr), (pudval))
++static inline long
++xencomm_arch_hypercall_vcpu_op(int cmd, int cpu, void *arg)
++{
++ return _hypercall3(long, vcpu_op, cmd, cpu, arg);
++}
+
-+/*
-+ * Pentium-II erratum A13: in PAE mode we explicitly have to flush
-+ * the TLB via cr3 if the top-level pgd is changed...
-+ * We do not let the generic code free and clear pgd entries due to
-+ * this erratum.
-+ */
-+static inline void pud_clear (pud_t * pud) { }
++static inline int
++HYPERVISOR_physdev_op(int cmd, void *arg)
++{
++ switch (cmd) {
++ case PHYSDEVOP_eoi:
++ return _hypercall1(int, ia64_fast_eoi,
++ ((struct physdev_eoi *)arg)->irq);
++ default:
++ return xencomm_hypercall_physdev_op(cmd, arg);
++ }
++}
+
-+#define pud_page(pud) \
-+((struct page *) __va(pud_val(pud) & PAGE_MASK))
++static inline int
++xencomm_arch_hypercall_xenoprof_op(int op, struct xencomm_handle *arg)
++{
++ return _hypercall2(int, xenoprof_op, op, arg);
++}
+
-+#define pud_page_kernel(pud) \
-+((unsigned long) __va(pud_val(pud) & PAGE_MASK))
++static inline long
++xencomm_arch_hypercall_opt_feature(struct xencomm_handle *arg)
++{
++ return _hypercall1(long, opt_feature, arg);
++}
+
++extern fastcall unsigned int __do_IRQ(unsigned int irq, struct pt_regs *regs);
++static inline void exit_idle(void) {}
++#define do_IRQ(irq, regs) ({ \
++ irq_enter(); \
++ __do_IRQ((irq), (regs)); \
++ irq_exit(); \
++})
+
-+/* Find an entry in the second-level page table.. */
-+#define pmd_offset(pud, address) ((pmd_t *) pud_page(*(pud)) + \
-+ pmd_index(address))
++#include <linux/err.h>
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
+
-+static inline int pte_none(pte_t pte)
++static inline unsigned long
++__HYPERVISOR_ioremap(unsigned long ioaddr, unsigned long size)
+{
-+ return !(pte.pte_low | pte.pte_high);
++ return _hypercall3(unsigned long, ia64_dom0vp_op,
++ IA64_DOM0VP_ioremap, ioaddr, size);
+}
+
-+/*
-+ * For PTEs and PDEs, we must clear the P-bit first when clearing a page table
-+ * entry, so clear the bottom half first and enforce ordering with a compiler
-+ * barrier.
-+ */
-+static inline void pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
++static inline unsigned long
++HYPERVISOR_ioremap(unsigned long ioaddr, unsigned long size)
+{
-+ if ((mm != current->mm && mm != &init_mm)
-+ || HYPERVISOR_update_va_mapping(addr, __pte(0), 0)) {
-+ ptep->pte_low = 0;
-+ smp_wmb();
-+ ptep->pte_high = 0;
++ unsigned long ret = ioaddr;
++ if (is_running_on_xen()) {
++ ret = __HYPERVISOR_ioremap(ioaddr, size);
++ if (unlikely(ret == -ENOSYS))
++ panic("hypercall %s failed with %ld. "
++ "Please check Xen and Linux config mismatch\n",
++ __func__, -ret);
++ else if (unlikely(IS_ERR_VALUE(ret)))
++ ret = ioaddr;
+ }
++ return ret;
+}
+
-+#define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0)
++static inline unsigned long
++__HYPERVISOR_phystomach(unsigned long gpfn)
++{
++ return _hypercall2(unsigned long, ia64_dom0vp_op,
++ IA64_DOM0VP_phystomach, gpfn);
++}
+
-+static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
++static inline unsigned long
++HYPERVISOR_phystomach(unsigned long gpfn)
+{
-+ pte_t pte = *ptep;
-+ if (!pte_none(pte)) {
-+ if (mm != &init_mm) {
-+ uint64_t val = __pte_val(pte);
-+ if (__cmpxchg64(ptep, val, 0) != val) {
-+ /* xchg acts as a barrier before the setting of the high bits */
-+ pte.pte_low = xchg(&ptep->pte_low, 0);
-+ pte.pte_high = ptep->pte_high;
-+ ptep->pte_high = 0;
-+ }
-+ } else
-+ HYPERVISOR_update_va_mapping(addr, __pte(0), 0);
++ unsigned long ret = gpfn;
++ if (is_running_on_xen()) {
++ ret = __HYPERVISOR_phystomach(gpfn);
+ }
-+ return pte;
++ return ret;
+}
+
-+#define ptep_clear_flush(vma, addr, ptep) \
-+({ \
-+ pte_t *__ptep = (ptep); \
-+ pte_t __res = *__ptep; \
-+ if (!pte_none(__res) && \
-+ ((vma)->vm_mm != current->mm || \
-+ HYPERVISOR_update_va_mapping(addr, __pte(0), \
-+ (unsigned long)(vma)->vm_mm->cpu_vm_mask.bits| \
-+ UVMF_INVLPG|UVMF_MULTI))) { \
-+ __ptep->pte_low = 0; \
-+ smp_wmb(); \
-+ __ptep->pte_high = 0; \
-+ flush_tlb_page(vma, addr); \
-+ } \
-+ __res; \
-+})
-+
-+static inline int pte_same(pte_t a, pte_t b)
++static inline unsigned long
++__HYPERVISOR_machtophys(unsigned long mfn)
+{
-+ return a.pte_low == b.pte_low && a.pte_high == b.pte_high;
++ return _hypercall2(unsigned long, ia64_dom0vp_op,
++ IA64_DOM0VP_machtophys, mfn);
+}
+
-+#define pte_page(x) pfn_to_page(pte_pfn(x))
++static inline unsigned long
++HYPERVISOR_machtophys(unsigned long mfn)
++{
++ unsigned long ret = mfn;
++ if (is_running_on_xen()) {
++ ret = __HYPERVISOR_machtophys(mfn);
++ }
++ return ret;
++}
+
-+#define __pte_mfn(_pte) (((_pte).pte_low >> PAGE_SHIFT) | \
-+ ((_pte).pte_high << (32-PAGE_SHIFT)))
-+#define pte_mfn(_pte) ((_pte).pte_low & _PAGE_PRESENT ? \
-+ __pte_mfn(_pte) : pfn_to_mfn(__pte_mfn(_pte)))
-+#define pte_pfn(_pte) ((_pte).pte_low & _PAGE_PRESENT ? \
-+ mfn_to_local_pfn(__pte_mfn(_pte)) : __pte_mfn(_pte))
++static inline unsigned long
++__HYPERVISOR_zap_physmap(unsigned long gpfn, unsigned int extent_order)
++{
++ return _hypercall3(unsigned long, ia64_dom0vp_op,
++ IA64_DOM0VP_zap_physmap, gpfn, extent_order);
++}
+
-+extern unsigned long long __supported_pte_mask;
++static inline unsigned long
++HYPERVISOR_zap_physmap(unsigned long gpfn, unsigned int extent_order)
++{
++ unsigned long ret = 0;
++ if (is_running_on_xen()) {
++ ret = __HYPERVISOR_zap_physmap(gpfn, extent_order);
++ }
++ return ret;
++}
+
-+static inline pte_t pfn_pte(unsigned long page_nr, pgprot_t pgprot)
++static inline unsigned long
++__HYPERVISOR_add_physmap(unsigned long gpfn, unsigned long mfn,
++ unsigned long flags, domid_t domid)
+{
-+ return __pte((((unsigned long long)page_nr << PAGE_SHIFT) |
-+ pgprot_val(pgprot)) & __supported_pte_mask);
++ return _hypercall5(unsigned long, ia64_dom0vp_op,
++ IA64_DOM0VP_add_physmap, gpfn, mfn, flags, domid);
+}
+
-+static inline pmd_t pfn_pmd(unsigned long page_nr, pgprot_t pgprot)
++static inline unsigned long
++HYPERVISOR_add_physmap(unsigned long gpfn, unsigned long mfn,
++ unsigned long flags, domid_t domid)
+{
-+ return __pmd((((unsigned long long)page_nr << PAGE_SHIFT) |
-+ pgprot_val(pgprot)) & __supported_pte_mask);
++ unsigned long ret = 0;
++ BUG_ON(!is_running_on_xen());//XXX
++ if (is_running_on_xen()) {
++ ret = __HYPERVISOR_add_physmap(gpfn, mfn, flags, domid);
++ }
++ return ret;
+}
+
-+/*
-+ * Bits 0, 6 and 7 are taken in the low part of the pte,
-+ * put the 32 bits of offset into the high part.
-+ */
-+#define pte_to_pgoff(pte) ((pte).pte_high)
-+#define pgoff_to_pte(off) ((pte_t) { _PAGE_FILE, (off) })
-+#define PTE_FILE_MAX_BITS 32
++static inline unsigned long
++__HYPERVISOR_add_physmap_with_gmfn(unsigned long gpfn, unsigned long gmfn,
++ unsigned long flags, domid_t domid)
++{
++ return _hypercall5(unsigned long, ia64_dom0vp_op,
++ IA64_DOM0VP_add_physmap_with_gmfn,
++ gpfn, gmfn, flags, domid);
++}
+
-+/* Encode and de-code a swap entry */
-+#define __swp_type(x) (((x).val) & 0x1f)
-+#define __swp_offset(x) ((x).val >> 5)
-+#define __swp_entry(type, offset) ((swp_entry_t){(type) | (offset) << 5})
-+#define __pte_to_swp_entry(pte) ((swp_entry_t){ (pte).pte_high })
-+#define __swp_entry_to_pte(x) ((pte_t){ 0, (x).val })
++static inline unsigned long
++HYPERVISOR_add_physmap_with_gmfn(unsigned long gpfn, unsigned long gmfn,
++ unsigned long flags, domid_t domid)
++{
++ unsigned long ret = 0;
++ BUG_ON(!is_running_on_xen());//XXX
++ if (is_running_on_xen()) {
++ ret = __HYPERVISOR_add_physmap_with_gmfn(gpfn, gmfn,
++ flags, domid);
++ }
++ return ret;
++}
+
-+#define __pmd_free_tlb(tlb, x) do { } while (0)
++#ifdef CONFIG_XEN_IA64_EXPOSE_P2M
++static inline unsigned long
++HYPERVISOR_expose_p2m(unsigned long conv_start_gpfn,
++ unsigned long assign_start_gpfn,
++ unsigned long expose_size, unsigned long granule_pfn)
++{
++ return _hypercall5(unsigned long, ia64_dom0vp_op,
++ IA64_DOM0VP_expose_p2m, conv_start_gpfn,
++ assign_start_gpfn, expose_size, granule_pfn);
++}
+
-+void vmalloc_sync_all(void);
++static inline int
++xencomm_arch_expose_foreign_p2m(unsigned long gpfn,
++ domid_t domid, struct xencomm_handle *arg,
++ unsigned long flags)
++{
++ return _hypercall5(int, ia64_dom0vp_op,
++ IA64_DOM0VP_expose_foreign_p2m,
++ gpfn, domid, arg, flags);
++}
+
-+#endif /* _I386_PGTABLE_3LEVEL_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pgtable.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pgtable.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/pgtable.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/pgtable.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,531 @@
-+#ifndef _I386_PGTABLE_H
-+#define _I386_PGTABLE_H
++static inline int
++HYPERVISOR_unexpose_foreign_p2m(unsigned long gpfn, domid_t domid)
++{
++ return _hypercall3(int, ia64_dom0vp_op,
++ IA64_DOM0VP_unexpose_foreign_p2m, gpfn, domid);
++}
++#endif
+
-+#include <asm/hypervisor.h>
++static inline int
++xencomm_arch_hypercall_perfmon_op(unsigned long cmd,
++ struct xencomm_handle *arg,
++ unsigned long count)
++{
++ return _hypercall4(int, ia64_dom0vp_op,
++ IA64_DOM0VP_perfmon, cmd, arg, count);
++}
+
-+/*
-+ * The Linux memory management assumes a three-level page table setup. On
-+ * the i386, we use that, but "fold" the mid level into the top-level page
-+ * table, so that we physically have the same two-level page table as the
-+ * i386 mmu expects.
-+ *
-+ * This file contains the functions and defines necessary to modify and use
-+ * the i386 page table tree.
-+ */
-+#ifndef __ASSEMBLY__
-+#include <asm/processor.h>
-+#include <asm/fixmap.h>
-+#include <linux/threads.h>
++static inline int
++xencomm_arch_hypercall_fpswa_revision(struct xencomm_handle *arg)
++{
++ return _hypercall2(int, ia64_dom0vp_op,
++ IA64_DOM0VP_fpswa_revision, arg);
++}
+
-+#ifndef _I386_BITOPS_H
-+#include <asm/bitops.h>
-+#endif
++static inline int
++xencomm_arch_hypercall_ia64_debug_op(unsigned long cmd,
++ unsigned long domain,
++ struct xencomm_handle *arg)
++{
++ return _hypercall3(int, ia64_debug_op, cmd, domain, arg);
++}
+
-+#include <linux/slab.h>
-+#include <linux/list.h>
-+#include <linux/spinlock.h>
++static inline int
++HYPERVISOR_add_io_space(unsigned long phys_base,
++ unsigned long sparse,
++ unsigned long space_number)
++{
++ return _hypercall4(int, ia64_dom0vp_op, IA64_DOM0VP_add_io_space,
++ phys_base, sparse, space_number);
++}
+
-+/* Is this pagetable pinned? */
-+#define PG_pinned PG_arch_1
++static inline int
++xencomm_arch_hypercall_kexec_op(int cmd, struct xencomm_handle *arg)
++{
++ return _hypercall2(int, kexec_op, cmd, arg);
++}
+
-+struct mm_struct;
-+struct vm_area_struct;
++// for balloon driver
++#define HYPERVISOR_update_va_mapping(va, new_val, flags) (0)
+
-+/*
-+ * ZERO_PAGE is a global shared page that is always zero: used
-+ * for zero-mapped memory areas etc..
-+ */
-+#define ZERO_PAGE(vaddr) (virt_to_page(empty_zero_page))
-+extern unsigned long empty_zero_page[1024];
-+extern pgd_t *swapper_pg_dir;
-+extern kmem_cache_t *pgd_cache;
-+extern kmem_cache_t *pmd_cache;
-+extern spinlock_t pgd_lock;
-+extern struct page *pgd_list;
++/* Use xencomm to do hypercalls. */
++#define HYPERVISOR_sched_op xencomm_hypercall_sched_op
++#define HYPERVISOR_event_channel_op xencomm_hypercall_event_channel_op
++#define HYPERVISOR_callback_op xencomm_hypercall_callback_op
++#define HYPERVISOR_multicall xencomm_hypercall_multicall
++#define HYPERVISOR_xen_version xencomm_hypercall_xen_version
++#define HYPERVISOR_console_io xencomm_hypercall_console_io
++#define HYPERVISOR_hvm_op xencomm_hypercall_hvm_op
++#define HYPERVISOR_memory_op xencomm_hypercall_memory_op
++#define HYPERVISOR_xenoprof_op xencomm_hypercall_xenoprof_op
++#define HYPERVISOR_perfmon_op xencomm_hypercall_perfmon_op
++#define HYPERVISOR_fpswa_revision xencomm_hypercall_fpswa_revision
++#define HYPERVISOR_suspend xencomm_hypercall_suspend
++#define HYPERVISOR_vcpu_op xencomm_hypercall_vcpu_op
++#define HYPERVISOR_opt_feature xencomm_hypercall_opt_feature
++#define HYPERVISOR_kexec_op xencomm_hypercall_kexec_op
+
-+void pmd_ctor(void *, kmem_cache_t *, unsigned long);
-+void pgd_ctor(void *, kmem_cache_t *, unsigned long);
-+void pgd_dtor(void *, kmem_cache_t *, unsigned long);
-+void pgtable_cache_init(void);
-+void paging_init(void);
++/* to compile gnttab_copy_grant_page() in drivers/xen/core/gnttab.c */
++#define HYPERVISOR_mmu_update(req, count, success_count, domid) ({BUG();0;})
+
-+/*
-+ * The Linux x86 paging architecture is 'compile-time dual-mode', it
-+ * implements both the traditional 2-level x86 page tables and the
-+ * newer 3-level PAE-mode page tables.
++#endif /* __HYPERCALL_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/hypervisor.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/hypervisor.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,233 @@
++/******************************************************************************
++ * hypervisor.h
++ *
++ * Linux-specific hypervisor handling.
++ *
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
-+#ifdef CONFIG_X86_PAE
-+# include <asm/pgtable-3level-defs.h>
-+# define PMD_SIZE (1UL << PMD_SHIFT)
-+# define PMD_MASK (~(PMD_SIZE-1))
-+#else
-+# include <asm/pgtable-2level-defs.h>
-+#endif
-+
-+#define PGDIR_SIZE (1UL << PGDIR_SHIFT)
-+#define PGDIR_MASK (~(PGDIR_SIZE-1))
+
-+#define USER_PTRS_PER_PGD (TASK_SIZE/PGDIR_SIZE)
-+#define FIRST_USER_ADDRESS 0
++#ifndef __HYPERVISOR_H__
++#define __HYPERVISOR_H__
+
-+#define USER_PGD_PTRS (PAGE_OFFSET >> PGDIR_SHIFT)
-+#define KERNEL_PGD_PTRS (PTRS_PER_PGD-USER_PGD_PTRS)
++#ifdef CONFIG_XEN
++/* running_on_xen is set before executing any C code by early_xen_setup */
++extern const int running_on_xen;
++#define is_running_on_xen() (running_on_xen)
++#else /* CONFIG_XEN */
++# ifdef CONFIG_VMX_GUEST
++# define is_running_on_xen() (1)
++# else /* CONFIG_VMX_GUEST */
++# define is_running_on_xen() (0)
++# define HYPERVISOR_ioremap(offset, size) (offset)
++# endif /* CONFIG_VMX_GUEST */
++#endif /* CONFIG_XEN */
+
-+#define TWOLEVEL_PGDIR_SHIFT 22
-+#define BOOT_USER_PGD_PTRS (__PAGE_OFFSET >> TWOLEVEL_PGDIR_SHIFT)
-+#define BOOT_KERNEL_PGD_PTRS (1024-BOOT_USER_PGD_PTRS)
++#if defined(CONFIG_XEN) || defined(CONFIG_VMX_GUEST)
++#include <linux/types.h>
++#include <linux/kernel.h>
++#include <linux/version.h>
++#include <linux/errno.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/platform.h>
++#include <xen/interface/event_channel.h>
++#include <xen/interface/physdev.h>
++#include <xen/interface/sched.h>
++#include <xen/hypercall.h>
++#include <asm/ptrace.h>
++#include <asm/page.h>
+
-+/* Just any arbitrary offset to the start of the vmalloc VM area: the
-+ * current 8MB value just means that there will be a 8MB "hole" after the
-+ * physical memory until the kernel virtual memory starts. That means that
-+ * any out-of-bounds memory accesses will hopefully be caught.
-+ * The vmalloc() routines leaves a hole of 4kB between each vmalloced
-+ * area for the same reason. ;)
-+ */
-+#define VMALLOC_OFFSET (8*1024*1024)
-+#define VMALLOC_START (((unsigned long) high_memory + vmalloc_earlyreserve + \
-+ 2*VMALLOC_OFFSET-1) & ~(VMALLOC_OFFSET-1))
-+#ifdef CONFIG_HIGHMEM
-+# define VMALLOC_END (PKMAP_BASE-2*PAGE_SIZE)
-+#else
-+# define VMALLOC_END (FIXADDR_START-2*PAGE_SIZE)
-+#endif
++extern shared_info_t *HYPERVISOR_shared_info;
++extern start_info_t *xen_start_info;
+
-+/*
-+ * _PAGE_PSE set in the page directory entry just means that
-+ * the page directory entry points directly to a 4MB-aligned block of
-+ * memory.
-+ */
-+#define _PAGE_BIT_PRESENT 0
-+#define _PAGE_BIT_RW 1
-+#define _PAGE_BIT_USER 2
-+#define _PAGE_BIT_PWT 3
-+#define _PAGE_BIT_PCD 4
-+#define _PAGE_BIT_ACCESSED 5
-+#define _PAGE_BIT_DIRTY 6
-+#define _PAGE_BIT_PSE 7 /* 4 MB (or 2MB) page, Pentium+, if present.. */
-+#define _PAGE_BIT_GLOBAL 8 /* Global TLB entry PPro+ */
-+#define _PAGE_BIT_UNUSED1 9 /* available for programmer */
-+#define _PAGE_BIT_UNUSED2 10
-+#define _PAGE_BIT_UNUSED3 11
-+#define _PAGE_BIT_NX 63
++void force_evtchn_callback(void);
+
-+#define _PAGE_PRESENT 0x001
-+#define _PAGE_RW 0x002
-+#define _PAGE_USER 0x004
-+#define _PAGE_PWT 0x008
-+#define _PAGE_PCD 0x010
-+#define _PAGE_ACCESSED 0x020
-+#define _PAGE_DIRTY 0x040
-+#define _PAGE_PSE 0x080 /* 4 MB (or 2MB) page, Pentium+, if present.. */
-+#define _PAGE_GLOBAL 0x100 /* Global TLB entry PPro+ */
-+#define _PAGE_UNUSED1 0x200 /* available for programmer */
-+#define _PAGE_UNUSED2 0x400
-+#define _PAGE_UNUSED3 0x800
++/* Turn jiffies into Xen system time. XXX Implement me. */
++#define jiffies_to_st(j) 0
+
-+/* If _PAGE_PRESENT is clear, we use these: */
-+#define _PAGE_FILE 0x040 /* nonlinear file mapping, saved PTE; unset:swap */
-+#define _PAGE_PROTNONE 0x080 /* if the user mapped it with PROT_NONE;
-+ pte_present gives true */
-+#ifdef CONFIG_X86_PAE
-+#define _PAGE_NX (1ULL<<_PAGE_BIT_NX)
-+#else
-+#define _PAGE_NX 0
-+#endif
++static inline int
++HYPERVISOR_yield(
++ void)
++{
++ int rc = HYPERVISOR_sched_op(SCHEDOP_yield, NULL);
+
-+#define _PAGE_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED | _PAGE_DIRTY)
-+#define _KERNPG_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY)
-+#define _PAGE_CHG_MASK (PTE_MASK | _PAGE_ACCESSED | _PAGE_DIRTY)
++ return rc;
++}
+
-+#define PAGE_NONE \
-+ __pgprot(_PAGE_PROTNONE | _PAGE_ACCESSED)
-+#define PAGE_SHARED \
-+ __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED)
++static inline int
++HYPERVISOR_block(
++ void)
++{
++ int rc = HYPERVISOR_sched_op(SCHEDOP_block, NULL);
+
-+#define PAGE_SHARED_EXEC \
-+ __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED)
-+#define PAGE_COPY_NOEXEC \
-+ __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX)
-+#define PAGE_COPY_EXEC \
-+ __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED)
-+#define PAGE_COPY \
-+ PAGE_COPY_NOEXEC
-+#define PAGE_READONLY \
-+ __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX)
-+#define PAGE_READONLY_EXEC \
-+ __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED)
++ return rc;
++}
+
-+#define _PAGE_KERNEL \
-+ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_NX)
-+#define _PAGE_KERNEL_EXEC \
-+ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED)
++static inline int
++HYPERVISOR_shutdown(
++ unsigned int reason)
++{
++ struct sched_shutdown sched_shutdown = {
++ .reason = reason
++ };
+
-+extern unsigned long long __PAGE_KERNEL, __PAGE_KERNEL_EXEC;
-+#define __PAGE_KERNEL_RO (__PAGE_KERNEL & ~_PAGE_RW)
-+#define __PAGE_KERNEL_NOCACHE (__PAGE_KERNEL | _PAGE_PCD)
-+#define __PAGE_KERNEL_LARGE (__PAGE_KERNEL | _PAGE_PSE)
-+#define __PAGE_KERNEL_LARGE_EXEC (__PAGE_KERNEL_EXEC | _PAGE_PSE)
++ int rc = HYPERVISOR_sched_op(SCHEDOP_shutdown, &sched_shutdown);
+
-+#define PAGE_KERNEL __pgprot(__PAGE_KERNEL)
-+#define PAGE_KERNEL_RO __pgprot(__PAGE_KERNEL_RO)
-+#define PAGE_KERNEL_EXEC __pgprot(__PAGE_KERNEL_EXEC)
-+#define PAGE_KERNEL_NOCACHE __pgprot(__PAGE_KERNEL_NOCACHE)
-+#define PAGE_KERNEL_LARGE __pgprot(__PAGE_KERNEL_LARGE)
-+#define PAGE_KERNEL_LARGE_EXEC __pgprot(__PAGE_KERNEL_LARGE_EXEC)
++ return rc;
++}
+
-+/*
-+ * The i386 can't do page protection for execute, and considers that
-+ * the same are read. Also, write permissions imply read permissions.
-+ * This is the closest we can get..
-+ */
-+#define __P000 PAGE_NONE
-+#define __P001 PAGE_READONLY
-+#define __P010 PAGE_COPY
-+#define __P011 PAGE_COPY
-+#define __P100 PAGE_READONLY_EXEC
-+#define __P101 PAGE_READONLY_EXEC
-+#define __P110 PAGE_COPY_EXEC
-+#define __P111 PAGE_COPY_EXEC
++static inline int
++HYPERVISOR_poll(
++ evtchn_port_t *ports, unsigned int nr_ports, u64 timeout)
++{
++ struct sched_poll sched_poll = {
++ .nr_ports = nr_ports,
++ .timeout = jiffies_to_st(timeout)
++ };
+
-+#define __S000 PAGE_NONE
-+#define __S001 PAGE_READONLY
-+#define __S010 PAGE_SHARED
-+#define __S011 PAGE_SHARED
-+#define __S100 PAGE_READONLY_EXEC
-+#define __S101 PAGE_READONLY_EXEC
-+#define __S110 PAGE_SHARED_EXEC
-+#define __S111 PAGE_SHARED_EXEC
++ int rc;
+
-+/*
-+ * Define this if things work differently on an i386 and an i486:
-+ * it will (on an i486) warn about kernel memory accesses that are
-+ * done without a 'access_ok(VERIFY_WRITE,..)'
-+ */
-+#undef TEST_ACCESS_OK
++ set_xen_guest_handle(sched_poll.ports, ports);
++ rc = HYPERVISOR_sched_op(SCHEDOP_poll, &sched_poll);
+
-+/* The boot page tables (all created as a single array) */
-+extern unsigned long pg0[];
++ return rc;
++}
+
-+#define pte_present(x) ((x).pte_low & (_PAGE_PRESENT | _PAGE_PROTNONE))
++#ifndef CONFIG_VMX_GUEST
++/* for drivers/xen/privcmd/privcmd.c */
++#define machine_to_phys_mapping 0
++struct vm_area_struct;
++int direct_remap_pfn_range(struct vm_area_struct *vma,
++ unsigned long address,
++ unsigned long mfn,
++ unsigned long size,
++ pgprot_t prot,
++ domid_t domid);
++struct file;
++int privcmd_enforce_singleshot_mapping(struct vm_area_struct *vma);
++int privcmd_mmap(struct file * file, struct vm_area_struct * vma);
++#define HAVE_ARCH_PRIVCMD_MMAP
+
-+/* To avoid harmful races, pmd_none(x) should check only the lower when PAE */
-+#define pmd_none(x) (!(unsigned long)__pmd_val(x))
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+/* pmd_present doesn't just test the _PAGE_PRESENT bit since wr.p.t.
-+ can temporarily clear it. */
-+#define pmd_present(x) (__pmd_val(x))
-+#define pmd_bad(x) ((__pmd_val(x) & (~PAGE_MASK & ~_PAGE_USER & ~_PAGE_PRESENT)) != (_KERNPG_TABLE & ~_PAGE_PRESENT))
++/* for drivers/xen/balloon/balloon.c */
++#ifdef CONFIG_XEN_SCRUB_PAGES
++#define scrub_pages(_p,_n) memset((void *)(_p), 0, (_n) << PAGE_SHIFT)
+#else
-+#define pmd_present(x) (__pmd_val(x) & _PAGE_PRESENT)
-+#define pmd_bad(x) ((__pmd_val(x) & (~PAGE_MASK & ~_PAGE_USER)) != _KERNPG_TABLE)
++#define scrub_pages(_p,_n) ((void)0)
+#endif
++#define pte_mfn(_x) pte_pfn(_x)
++#define phys_to_machine_mapping_valid(_x) (1)
+
++void xen_contiguous_bitmap_init(unsigned long end_pfn);
++int __xen_create_contiguous_region(unsigned long vstart, unsigned int order, unsigned int address_bits);
++static inline int
++xen_create_contiguous_region(unsigned long vstart,
++ unsigned int order, unsigned int address_bits)
++{
++ int ret = 0;
++ if (is_running_on_xen()) {
++ ret = __xen_create_contiguous_region(vstart, order,
++ address_bits);
++ }
++ return ret;
++}
+
-+#define pages_to_mb(x) ((x) >> (20-PAGE_SHIFT))
-+
-+/*
-+ * The following only work if pte_present() is true.
-+ * Undefined behaviour if not..
-+ */
-+static inline int pte_user(pte_t pte) { return (pte).pte_low & _PAGE_USER; }
-+static inline int pte_read(pte_t pte) { return (pte).pte_low & _PAGE_USER; }
-+static inline int pte_dirty(pte_t pte) { return (pte).pte_low & _PAGE_DIRTY; }
-+static inline int pte_young(pte_t pte) { return (pte).pte_low & _PAGE_ACCESSED; }
-+static inline int pte_write(pte_t pte) { return (pte).pte_low & _PAGE_RW; }
-+static inline int pte_huge(pte_t pte) { return (pte).pte_low & _PAGE_PSE; }
++void __xen_destroy_contiguous_region(unsigned long vstart, unsigned int order);
++static inline void
++xen_destroy_contiguous_region(unsigned long vstart, unsigned int order)
++{
++ if (is_running_on_xen())
++ __xen_destroy_contiguous_region(vstart, order);
++}
+
-+/*
-+ * The following only works if pte_present() is not true.
-+ */
-+static inline int pte_file(pte_t pte) { return (pte).pte_low & _PAGE_FILE; }
++struct page;
+
-+static inline pte_t pte_rdprotect(pte_t pte) { (pte).pte_low &= ~_PAGE_USER; return pte; }
-+static inline pte_t pte_exprotect(pte_t pte) { (pte).pte_low &= ~_PAGE_USER; return pte; }
-+static inline pte_t pte_mkclean(pte_t pte) { (pte).pte_low &= ~_PAGE_DIRTY; return pte; }
-+static inline pte_t pte_mkold(pte_t pte) { (pte).pte_low &= ~_PAGE_ACCESSED; return pte; }
-+static inline pte_t pte_wrprotect(pte_t pte) { (pte).pte_low &= ~_PAGE_RW; return pte; }
-+static inline pte_t pte_mkread(pte_t pte) { (pte).pte_low |= _PAGE_USER; return pte; }
-+static inline pte_t pte_mkexec(pte_t pte) { (pte).pte_low |= _PAGE_USER; return pte; }
-+static inline pte_t pte_mkdirty(pte_t pte) { (pte).pte_low |= _PAGE_DIRTY; return pte; }
-+static inline pte_t pte_mkyoung(pte_t pte) { (pte).pte_low |= _PAGE_ACCESSED; return pte; }
-+static inline pte_t pte_mkwrite(pte_t pte) { (pte).pte_low |= _PAGE_RW; return pte; }
-+static inline pte_t pte_mkhuge(pte_t pte) { (pte).pte_low |= _PAGE_PSE; return pte; }
++int xen_limit_pages_to_max_mfn(struct page *pages, unsigned int order,
++ unsigned int address_bits);
+
-+#ifdef CONFIG_X86_PAE
-+# include <asm/pgtable-3level.h>
-+#else
-+# include <asm/pgtable-2level.h>
-+#endif
++/* For drivers/xen/core/machine_reboot.c */
++#define HAVE_XEN_POST_SUSPEND
++void xen_post_suspend(int suspend_cancelled);
+
-+#define ptep_test_and_clear_dirty(vma, addr, ptep) \
-+({ \
-+ pte_t __pte = *(ptep); \
-+ int __ret = pte_dirty(__pte); \
-+ if (__ret) { \
-+ __pte = pte_mkclean(__pte); \
-+ if ((vma)->vm_mm != current->mm || \
-+ HYPERVISOR_update_va_mapping(addr, __pte, 0)) \
-+ (ptep)->pte_low = __pte.pte_low; \
-+ } \
-+ __ret; \
-+})
++/* For setup_arch() in arch/ia64/kernel/setup.c */
++void xen_ia64_enable_opt_feature(void);
++#endif /* !CONFIG_VMX_GUEST */
+
-+#define ptep_test_and_clear_young(vma, addr, ptep) \
-+({ \
-+ pte_t __pte = *(ptep); \
-+ int __ret = pte_young(__pte); \
-+ if (__ret) \
-+ __pte = pte_mkold(__pte); \
-+ if ((vma)->vm_mm != current->mm || \
-+ HYPERVISOR_update_va_mapping(addr, __pte, 0)) \
-+ (ptep)->pte_low = __pte.pte_low; \
-+ __ret; \
-+})
++#define __pte_ma(_x) ((pte_t) {(_x)}) /* unmodified use */
++#define pfn_pte_ma(_x,_y) __pte_ma(0) /* unmodified use */
+
-+#define ptep_get_and_clear_full(mm, addr, ptep, full) \
-+ ((full) ? ({ \
-+ pte_t __res = *(ptep); \
-+ if (test_bit(PG_pinned, &virt_to_page((mm)->pgd)->flags)) \
-+ xen_l1_entry_update(ptep, __pte(0)); \
-+ else \
-+ *(ptep) = __pte(0); \
-+ __res; \
-+ }) : \
-+ ptep_get_and_clear(mm, addr, ptep))
++/* for netfront.c, netback.c */
++#define MULTI_UVMFLAGS_INDEX 0 /* XXX any value */
+
-+static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
++static inline void
++MULTI_update_va_mapping(
++ multicall_entry_t *mcl, unsigned long va,
++ pte_t new_val, unsigned long flags)
+{
-+ pte_t pte = *ptep;
-+ if (pte_write(pte))
-+ set_pte_at(mm, addr, ptep, pte_wrprotect(pte));
++ mcl->op = __HYPERVISOR_update_va_mapping;
++ mcl->result = 0;
+}
+
-+/*
-+ * clone_pgd_range(pgd_t *dst, pgd_t *src, int count);
-+ *
-+ * dst - pointer to pgd range anwhere on a pgd page
-+ * src - ""
-+ * count - the number of pgds to copy.
-+ *
-+ * dst and src can be on the same page, but the range must not overlap,
-+ * and must not cross a page boundary.
-+ */
-+static inline void clone_pgd_range(pgd_t *dst, pgd_t *src, int count)
++static inline void
++MULTI_grant_table_op(multicall_entry_t *mcl, unsigned int cmd,
++ void *uop, unsigned int count)
+{
-+ memcpy(dst, src, count * sizeof(pgd_t));
++ mcl->op = __HYPERVISOR_grant_table_op;
++ mcl->args[0] = cmd;
++ mcl->args[1] = (unsigned long)uop;
++ mcl->args[2] = count;
+}
+
+/*
-+ * Macro to mark a page protection value as "uncacheable". On processors which do not support
-+ * it, this is a no-op.
-+ */
-+#define pgprot_noncached(prot) ((boot_cpu_data.x86 > 3) \
-+ ? (__pgprot(pgprot_val(prot) | _PAGE_PCD | _PAGE_PWT)) : (prot))
-+
-+/*
-+ * Conversion functions: convert a page and protection to a page entry,
-+ * and a page entry and page directory to the page they refer to.
++ * for blktap.c
++ * int create_lookup_pte_addr(struct mm_struct *mm,
++ * unsigned long address,
++ * uint64_t *ptep);
+ */
++#define create_lookup_pte_addr(mm, address, ptep) \
++ ({ \
++ printk(KERN_EMERG \
++ "%s:%d " \
++ "create_lookup_pte_addr() isn't supported.\n", \
++ __func__, __LINE__); \
++ BUG(); \
++ (-ENOSYS); \
++ })
+
-+#define mk_pte(page, pgprot) pfn_pte(page_to_pfn(page), (pgprot))
++/* for debug */
++asmlinkage int xprintk(const char *fmt, ...);
++#define xprintd(fmt, ...) xprintk("%s:%d " fmt, __func__, __LINE__, \
++ ##__VA_ARGS__)
+
-+static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
-+{
-+ /*
-+ * Since this might change the present bit (which controls whether
-+ * a pte_t object has undergone p2m translation), we must use
-+ * pte_val() on the input pte and __pte() for the return value.
-+ */
-+ paddr_t pteval = pte_val(pte);
++#endif /* CONFIG_XEN || CONFIG_VMX_GUEST */
+
-+ pteval &= _PAGE_CHG_MASK;
-+ pteval |= pgprot_val(newprot);
-+#ifdef CONFIG_X86_PAE
-+ pteval &= __supported_pte_mask;
++#ifdef CONFIG_XEN_PRIVILEGED_GUEST
++#define is_initial_xendomain() \
++ (is_running_on_xen() ? xen_start_info->flags & SIF_INITDOMAIN : 0)
++#else
++#define is_initial_xendomain() 0
+#endif
-+ return __pte(pteval);
-+}
+
-+#define pmd_large(pmd) \
-+((__pmd_val(pmd) & (_PAGE_PSE|_PAGE_PRESENT)) == (_PAGE_PSE|_PAGE_PRESENT))
++#endif /* __HYPERVISOR_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/intel_intrin.h
+--- a/include/asm-ia64/intel_intrin.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/intel_intrin.h Wed Sep 10 10:54:08 2008 +0100
+@@ -16,8 +16,8 @@
+ * intrinsic
+ */
+
+-#define ia64_getreg __getReg
+-#define ia64_setreg __setReg
++#define __ia64_getreg __getReg
++#define __ia64_setreg __setReg
+
+ #define ia64_hint __hint
+ #define ia64_hint_pause __hint_pause
+@@ -33,16 +33,16 @@
+ #define ia64_getf_exp __getf_exp
+ #define ia64_shrp _m64_shrp
+
+-#define ia64_tpa __tpa
++#define __ia64_tpa __tpa
+ #define ia64_invala __invala
+ #define ia64_invala_gr __invala_gr
+ #define ia64_invala_fr __invala_fr
+ #define ia64_nop __nop
+ #define ia64_sum __sum
+-#define ia64_ssm __ssm
++#define __ia64_ssm __ssm
+ #define ia64_rum __rum
+-#define ia64_rsm __rsm
+-#define ia64_fc __fc
++#define __ia64_rsm __rsm
++#define __ia64_fc __fc
+
+ #define ia64_ldfs __ldfs
+ #define ia64_ldfd __ldfd
+@@ -80,24 +80,24 @@
+
+ #define __ia64_set_dbr(index, val) \
+ __setIndReg(_IA64_REG_INDR_DBR, index, val)
+-#define ia64_set_ibr(index, val) \
++#define __ia64_set_ibr(index, val) \
+ __setIndReg(_IA64_REG_INDR_IBR, index, val)
+-#define ia64_set_pkr(index, val) \
++#define __ia64_set_pkr(index, val) \
+ __setIndReg(_IA64_REG_INDR_PKR, index, val)
+-#define ia64_set_pmc(index, val) \
++#define __ia64_set_pmc(index, val) \
+ __setIndReg(_IA64_REG_INDR_PMC, index, val)
+-#define ia64_set_pmd(index, val) \
++#define __ia64_set_pmd(index, val) \
+ __setIndReg(_IA64_REG_INDR_PMD, index, val)
+-#define ia64_set_rr(index, val) \
++#define __ia64_set_rr(index, val) \
+ __setIndReg(_IA64_REG_INDR_RR, index, val)
+
+-#define ia64_get_cpuid(index) __getIndReg(_IA64_REG_INDR_CPUID, index)
++#define __ia64_get_cpuid(index) __getIndReg(_IA64_REG_INDR_CPUID, index)
+ #define __ia64_get_dbr(index) __getIndReg(_IA64_REG_INDR_DBR, index)
+-#define ia64_get_ibr(index) __getIndReg(_IA64_REG_INDR_IBR, index)
+-#define ia64_get_pkr(index) __getIndReg(_IA64_REG_INDR_PKR, index)
+-#define ia64_get_pmc(index) __getIndReg(_IA64_REG_INDR_PMC, index)
+-#define ia64_get_pmd(index) __getIndReg(_IA64_REG_INDR_PMD, index)
+-#define ia64_get_rr(index) __getIndReg(_IA64_REG_INDR_RR, index)
++#define __ia64_get_ibr(index) __getIndReg(_IA64_REG_INDR_IBR, index)
++#define __ia64_get_pkr(index) __getIndReg(_IA64_REG_INDR_PKR, index)
++#define __ia64_get_pmc(index) __getIndReg(_IA64_REG_INDR_PMC, index)
++#define __ia64_get_pmd(index) __getIndReg(_IA64_REG_INDR_PMD, index)
++#define __ia64_get_rr(index) __getIndReg(_IA64_REG_INDR_RR, index)
+
+ #define ia64_srlz_d __dsrlz
+ #define ia64_srlz_i __isrlz
+@@ -116,18 +116,18 @@
+ #define ia64_ld8_acq __ld8_acq
+
+ #define ia64_sync_i __synci
+-#define ia64_thash __thash
+-#define ia64_ttag __ttag
+-#define ia64_itcd __itcd
+-#define ia64_itci __itci
+-#define ia64_itrd __itrd
+-#define ia64_itri __itri
+-#define ia64_ptce __ptce
+-#define ia64_ptcl __ptcl
+-#define ia64_ptcg __ptcg
+-#define ia64_ptcga __ptcga
+-#define ia64_ptri __ptri
+-#define ia64_ptrd __ptrd
++#define __ia64_thash __thash
++#define __ia64_ttag __ttag
++#define __ia64_itcd __itcd
++#define __ia64_itci __itci
++#define __ia64_itrd __itrd
++#define __ia64_itri __itri
++#define __ia64_ptce __ptce
++#define __ia64_ptcl __ptcl
++#define __ia64_ptcg __ptcg
++#define __ia64_ptcga __ptcga
++#define __ia64_ptri __ptri
++#define __ia64_ptrd __ptrd
+ #define ia64_dep_mi _m64_dep_mi
+
+ /* Values for lfhint in __lfetch and __lfetch_fault */
+@@ -142,15 +142,17 @@
+ #define ia64_lfetch_fault __lfetch_fault
+ #define ia64_lfetch_fault_excl __lfetch_fault_excl
+
+-#define ia64_intrin_local_irq_restore(x) \
++#define __ia64_intrin_local_irq_restore(x) \
+ do { \
+ if ((x) != 0) { \
+- ia64_ssm(IA64_PSR_I); \
++ __ia64_ssm(IA64_PSR_I); \
+ ia64_srlz_d(); \
+ } else { \
+- ia64_rsm(IA64_PSR_I); \
++ __ia64_rsm(IA64_PSR_I); \
+ } \
+ } while (0)
+
-+/*
-+ * the pgd page can be thought of an array like this: pgd_t[PTRS_PER_PGD]
-+ *
-+ * this macro returns the index of the entry in the pgd page which would
-+ * control the given virtual address
-+ */
-+#define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD-1))
-+#define pgd_index_k(addr) pgd_index(addr)
++#define __ia64_get_psr_i() (__ia64_getreg(_IA64_REG_PSR) & 0x4000UL)
+
+ #define __builtin_trap() __break(0);
+
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/intrinsics.h
+--- a/include/asm-ia64/intrinsics.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/intrinsics.h Wed Sep 10 10:54:08 2008 +0100
+@@ -17,6 +17,15 @@
+ #else
+ # include <asm/gcc_intrin.h>
+ #endif
+
-+/*
-+ * pgd_offset() returns a (pgd_t *)
-+ * pgd_index() is used get the offset into the pgd page's array of pgd_t's;
-+ */
-+#define pgd_offset(mm, address) ((mm)->pgd+pgd_index(address))
++#define __ia64_set_rr0_to_rr4(val0, val1, val2, val3, val4) \
++do { \
++ __ia64_set_rr(0x0000000000000000UL, (val0)); \
++ __ia64_set_rr(0x2000000000000000UL, (val1)); \
++ __ia64_set_rr(0x4000000000000000UL, (val2)); \
++ __ia64_set_rr(0x6000000000000000UL, (val3)); \
++ __ia64_set_rr(0x8000000000000000UL, (val4)); \
++} while (0)
+
+ /*
+ * Force an unresolved reference if someone tries to use
+@@ -177,4 +186,5 @@
+ #endif /* !CONFIG_IA64_DEBUG_CMPXCHG */
+
+ #endif
++#include <asm/privop.h>
+ #endif /* _ASM_IA64_INTRINSICS_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/io.h
+--- a/include/asm-ia64/io.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/io.h Wed Sep 10 10:54:08 2008 +0100
+@@ -96,9 +96,39 @@
+ * The following two macros are deprecated and scheduled for removal.
+ * Please use the PCI-DMA interface defined in <asm/pci.h> instead.
+ */
++#ifndef CONFIG_XEN
+ #define bus_to_virt phys_to_virt
+ #define virt_to_bus virt_to_phys
+ #define page_to_bus page_to_phys
++#else
++#define bus_to_virt(bus) \
++ phys_to_virt(machine_to_phys_for_dma(bus))
++#define virt_to_bus(virt) \
++ phys_to_machine_for_dma(virt_to_phys(virt))
++#define page_to_bus(page) \
++ phys_to_machine_for_dma(page_to_pseudophys(page))
+
-+/*
-+ * a shortcut which implies the use of the kernel's pgd, instead
-+ * of a process's
-+ */
-+#define pgd_offset_k(address) pgd_offset(&init_mm, address)
++#define page_to_pseudophys(page) \
++ ((dma_addr_t)page_to_pfn(page) << PAGE_SHIFT)
+
+/*
-+ * the pmd page can be thought of an array like this: pmd_t[PTRS_PER_PMD]
-+ *
-+ * this macro returns the index of the entry in the pmd page which would
-+ * control the given virtual address
++ * Drivers that use page_to_phys() for bus addresses are broken.
++ * This includes:
++ * drivers/ide/cris/ide-cris.c
++ * drivers/scsi/dec_esp.c
+ */
-+#define pmd_index(address) \
-+ (((address) >> PMD_SHIFT) & (PTRS_PER_PMD-1))
-+
++#define page_to_phys(page) (page_to_pseudophys(page))
++#define bvec_to_bus(bv) (page_to_bus((bv)->bv_page) + \
++ (unsigned long) (bv)->bv_offset)
++#define bio_to_pseudophys(bio) (page_to_pseudophys(bio_page((bio))) + \
++ (unsigned long) bio_offset((bio)))
++#define bvec_to_pseudophys(bv) (page_to_pseudophys((bv)->bv_page) + \
++ (unsigned long) (bv)->bv_offset)
++#define BIOVEC_PHYS_MERGEABLE(vec1, vec2) \
++ (((bvec_to_bus((vec1)) + (vec1)->bv_len) == bvec_to_bus((vec2))) && \
++ ((bvec_to_pseudophys((vec1)) + (vec1)->bv_len) == \
++ bvec_to_pseudophys((vec2))))
++#endif /* CONFIG_XEN */
+
+ # endif /* KERNEL */
+
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/iosapic.h
+--- a/include/asm-ia64/iosapic.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/iosapic.h Wed Sep 10 10:54:08 2008 +0100
+@@ -53,6 +53,7 @@
+
+ #define NR_IOSAPICS 256
+
++#ifndef CONFIG_XEN
+ static inline unsigned int iosapic_read(char __iomem *iosapic, unsigned int reg)
+ {
+ writel(reg, iosapic + IOSAPIC_REG_SELECT);
+@@ -64,6 +65,7 @@
+ writel(reg, iosapic + IOSAPIC_REG_SELECT);
+ writel(val, iosapic + IOSAPIC_WINDOW);
+ }
++#endif
+
+ static inline void iosapic_eoi(char __iomem *iosapic, u32 vector)
+ {
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/irq.h
+--- a/include/asm-ia64/irq.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/irq.h Wed Sep 10 10:54:08 2008 +0100
+@@ -11,8 +11,41 @@
+ * 02/29/00 D.Mosberger moved most things into hw_irq.h
+ */
+
++#ifndef CONFIG_XEN
+ #define NR_IRQS 256
+ #define NR_IRQ_VECTORS NR_IRQS
++#else
+/*
-+ * the pte page can be thought of an array like this: pte_t[PTRS_PER_PTE]
-+ *
-+ * this macro returns the index of the entry in the pte page which would
-+ * control the given virtual address
++ * The flat IRQ space is divided into two regions:
++ * 1. A one-to-one mapping of real physical IRQs. This space is only used
++ * if we have physical device-access privilege. This region is at the
++ * start of the IRQ space so that existing device drivers do not need
++ * to be modified to translate physical IRQ numbers into our IRQ space.
++ * 3. A dynamic mapping of inter-domain and Xen-sourced virtual IRQs. These
++ * are bound using the provided bind/unbind functions.
+ */
-+#define pte_index(address) \
-+ (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1))
-+#define pte_offset_kernel(dir, address) \
-+ ((pte_t *) pmd_page_kernel(*(dir)) + pte_index(address))
+
-+#define pmd_page(pmd) (pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT))
++#define PIRQ_BASE 0
++#define NR_PIRQS 256
+
-+#define pmd_page_kernel(pmd) \
-+ ((unsigned long) __va(pmd_val(pmd) & PAGE_MASK))
++#define DYNIRQ_BASE (PIRQ_BASE + NR_PIRQS)
++#define NR_DYNIRQS (CONFIG_NR_CPUS * 8)
+
-+/*
-+ * Helper function that returns the kernel pagetable entry controlling
-+ * the virtual address 'address'. NULL means no pagetable entry present.
-+ * NOTE: the return type is pte_t but if the pmd is PSE then we return it
-+ * as a pte too.
-+ */
-+extern pte_t *lookup_address(unsigned long address);
++#define NR_IRQS (NR_PIRQS + NR_DYNIRQS)
++#define NR_IRQ_VECTORS NR_IRQS
+
-+/*
-+ * Make a given kernel text page executable/non-executable.
-+ * Returns the previous executability setting of that page (which
-+ * is used to restore the previous state). Used by the SMP bootup code.
-+ * NOTE: this is an __init function for security reasons.
-+ */
-+#ifdef CONFIG_X86_PAE
-+ extern int set_kernel_exec(unsigned long vaddr, int enable);
-+#else
-+ static inline int set_kernel_exec(unsigned long vaddr, int enable) { return 0;}
-+#endif
++#define pirq_to_irq(_x) ((_x) + PIRQ_BASE)
++#define irq_to_pirq(_x) ((_x) - PIRQ_BASE)
+
-+extern void noexec_setup(const char *str);
++#define dynirq_to_irq(_x) ((_x) + DYNIRQ_BASE)
++#define irq_to_dynirq(_x) ((_x) - DYNIRQ_BASE)
+
-+#if defined(CONFIG_HIGHPTE)
-+#define pte_offset_map(dir, address) \
-+ ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)),KM_PTE0) + \
-+ pte_index(address))
-+#define pte_offset_map_nested(dir, address) \
-+ ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)),KM_PTE1) + \
-+ pte_index(address))
-+#define pte_unmap(pte) kunmap_atomic(pte, KM_PTE0)
-+#define pte_unmap_nested(pte) kunmap_atomic(pte, KM_PTE1)
-+#else
-+#define pte_offset_map(dir, address) \
-+ ((pte_t *)page_address(pmd_page(*(dir))) + pte_index(address))
-+#define pte_offset_map_nested(dir, address) pte_offset_map(dir, address)
-+#define pte_unmap(pte) do { } while (0)
-+#define pte_unmap_nested(pte) do { } while (0)
++#define RESCHEDULE_VECTOR 0
++#define IPI_VECTOR 1
++#define CMCP_VECTOR 2
++#define CPEP_VECTOR 3
++#define NR_IPIS 4
++#endif /* CONFIG_XEN */
+
+ static __inline__ int
+ irq_canonicalize (int irq)
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/kexec.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/kexec.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,58 @@
++#ifndef _ASM_IA64_KEXEC_H
++#define _ASM_IA64_KEXEC_H
++
++
++/* Maximum physical address we can use pages from */
++#define KEXEC_SOURCE_MEMORY_LIMIT (-1UL)
++/* Maximum address we can reach in physical address mode */
++#define KEXEC_DESTINATION_MEMORY_LIMIT (-1UL)
++/* Maximum address we can use for the control code buffer */
++#define KEXEC_CONTROL_MEMORY_LIMIT TASK_SIZE
++
++#define KEXEC_CONTROL_CODE_SIZE (8192 + 8192 + 4096)
++
++/* The native architecture */
++#define KEXEC_ARCH KEXEC_ARCH_IA_64
++
++#define MAX_NOTE_BYTES 1024
++
++#define kexec_flush_icache_page(page) do { \
++ unsigned long page_addr = (unsigned long)page_address(page); \
++ flush_icache_range(page_addr, page_addr + PAGE_SIZE); \
++ } while(0)
++
++extern struct kimage *ia64_kimage;
++DECLARE_PER_CPU(u64, ia64_mca_pal_base);
++const extern unsigned int relocate_new_kernel_size;
++extern void relocate_new_kernel(unsigned long, unsigned long,
++ struct ia64_boot_param *, unsigned long);
++static inline void
++crash_setup_regs(struct pt_regs *newregs, struct pt_regs *oldregs)
++{
++}
++extern struct resource efi_memmap_res;
++extern struct resource boot_param_res;
++extern void kdump_smp_send_stop(void);
++extern void kdump_smp_send_init(void);
++extern void kexec_disable_iosapic(void);
++extern void crash_save_this_cpu(void);
++struct rsvd_region;
++extern unsigned long kdump_find_rsvd_region(unsigned long size,
++ struct rsvd_region *rsvd_regions, int n);
++extern void kdump_cpu_freeze(struct unw_frame_info *info, void *arg);
++extern int kdump_status[];
++extern atomic_t kdump_cpu_freezed;
++extern atomic_t kdump_in_progress;
++
++/* Kexec needs to know about the actual physical addresss.
++ * But in xen, on some architectures, a physical address is a
++ * pseudo-physical addresss. */
++#ifdef CONFIG_XEN
++#define KEXEC_ARCH_HAS_PAGE_MACROS
++#define kexec_page_to_pfn(page) pfn_to_mfn_for_dma(page_to_pfn(page))
++#define kexec_pfn_to_page(pfn) pfn_to_page(mfn_to_pfn_for_dma(pfn))
++#define kexec_virt_to_phys(addr) phys_to_machine_for_dma(__pa(addr))
++#define kexec_phys_to_virt(addr) phys_to_virt(machine_to_phys_for_dma(addr))
++#endif
++
++#endif /* _ASM_IA64_KEXEC_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/machvec.h
+--- a/include/asm-ia64/machvec.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/machvec.h Wed Sep 10 10:54:08 2008 +0100
+@@ -35,6 +35,7 @@
+ typedef int ia64_mv_pci_legacy_write_t (struct pci_bus *, u16 port, u32 val,
+ u8 size);
+ typedef void ia64_mv_migrate_t(struct task_struct * task);
++typedef void ia64_mv_kernel_launch_event_t(void);
+
+ /* DMA-mapping interface: */
+ typedef void ia64_mv_dma_init (void);
+@@ -108,6 +109,8 @@
+ # include <asm/machvec_hpzx1_swiotlb.h>
+ # elif defined (CONFIG_IA64_SGI_SN2)
+ # include <asm/machvec_sn2.h>
++# elif defined (CONFIG_IA64_XEN)
++# include <asm/machvec_xen.h>
+ # elif defined (CONFIG_IA64_GENERIC)
+
+ # ifdef MACHVEC_PLATFORM_HEADER
+@@ -205,6 +208,7 @@
+ ia64_mv_readq_relaxed_t *readq_relaxed;
+ ia64_mv_migrate_t *migrate;
+ ia64_mv_msi_init_t *msi_init;
++ ia64_mv_kernel_launch_event_t *kernel_launch_event;
+ } __attribute__((__aligned__(16))); /* align attrib? see above comment */
+
+ #define MACHVEC_INIT(name) \
+@@ -302,6 +306,9 @@
+ #endif
+ #ifndef platform_tlb_migrate_finish
+ # define platform_tlb_migrate_finish machvec_noop_mm
+#endif
++#ifndef platform_kernel_launch_event
++# define platform_kernel_launch_event machvec_noop
+ #endif
+ #ifndef platform_dma_init
+ # define platform_dma_init swiotlb_init
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/machvec_sn2.h
+--- a/include/asm-ia64/machvec_sn2.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/machvec_sn2.h Wed Sep 10 10:54:08 2008 +0100
+@@ -67,6 +67,7 @@
+ extern ia64_mv_dma_mapping_error sn_dma_mapping_error;
+ extern ia64_mv_dma_supported sn_dma_supported;
+ extern ia64_mv_migrate_t sn_migrate;
++extern ia64_mv_kernel_launch_event_t sn_kernel_launch_event;
+ extern ia64_mv_msi_init_t sn_msi_init;
+
+
+@@ -119,6 +120,7 @@
+ #define platform_dma_mapping_error sn_dma_mapping_error
+ #define platform_dma_supported sn_dma_supported
+ #define platform_migrate sn_migrate
++#define platform_kernel_launch_event sn_kernel_launch_event
+ #ifdef CONFIG_PCI_MSI
+ #define platform_msi_init sn_msi_init
+ #else
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/machvec_xen.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/machvec_xen.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,43 @@
++#ifndef _ASM_IA64_MACHVEC_XEN_h
++#define _ASM_IA64_MACHVEC_XEN_h
+
-+#define __HAVE_ARCH_PTEP_ESTABLISH
-+#define ptep_establish(vma, address, ptep, pteval) \
-+ do { \
-+ if ( likely((vma)->vm_mm == current->mm) ) { \
-+ BUG_ON(HYPERVISOR_update_va_mapping(address, \
-+ pteval, \
-+ (unsigned long)(vma)->vm_mm->cpu_vm_mask.bits| \
-+ UVMF_INVLPG|UVMF_MULTI)); \
-+ } else { \
-+ xen_l1_entry_update(ptep, pteval); \
-+ flush_tlb_page(vma, address); \
-+ } \
-+ } while (0)
-+
-+/*
-+ * The i386 doesn't have any external MMU info: the kernel page
-+ * tables contain all the necessary information.
-+ *
-+ * Also, we only update the dirty/accessed state if we set
-+ * the dirty bit by hand in the kernel, since the hardware
-+ * will do the accessed bit for us, and we don't want to
-+ * race with other CPU's that might be updating the dirty
-+ * bit at the same time.
-+ */
-+#define update_mmu_cache(vma,address,pte) do { } while (0)
-+#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
-+#define ptep_set_access_flags(vma, address, ptep, entry, dirty) \
-+ do { \
-+ if (dirty) \
-+ ptep_establish(vma, address, ptep, entry); \
-+ } while (0)
++extern ia64_mv_setup_t xen_setup;
++extern ia64_mv_cpu_init_t xen_cpu_init;
++extern ia64_mv_irq_init_t xen_irq_init;
++extern ia64_mv_send_ipi_t xen_platform_send_ipi;
++extern ia64_mv_dma_alloc_coherent xen_alloc_coherent;
++extern ia64_mv_dma_free_coherent xen_free_coherent;
++extern ia64_mv_dma_map_single xen_map_single;
++extern ia64_mv_dma_unmap_single xen_unmap_single;
++extern ia64_mv_dma_map_sg xen_map_sg;
++extern ia64_mv_dma_unmap_sg xen_unmap_sg;
++extern ia64_mv_dma_supported xen_dma_supported;
++extern ia64_mv_dma_mapping_error xen_dma_mapping_error;
++
++/*
++ * This stuff has dual use!
++ *
++ * For a generic kernel, the macros are used to initialize the
++ * platform's machvec structure. When compiling a non-generic kernel,
++ * the macros are used directly.
++ */
++#define platform_name "xen"
++#define platform_setup xen_setup
++#define platform_cpu_init xen_cpu_init
++#define platform_irq_init xen_irq_init
++#define platform_send_ipi xen_platform_send_ipi
++#define platform_dma_init machvec_noop
++#define platform_dma_alloc_coherent xen_alloc_coherent
++#define platform_dma_free_coherent xen_free_coherent
++#define platform_dma_map_single xen_map_single
++#define platform_dma_unmap_single xen_unmap_single
++#define platform_dma_map_sg xen_map_sg
++#define platform_dma_unmap_sg xen_unmap_sg
++#define platform_dma_sync_single_for_cpu machvec_dma_sync_single
++#define platform_dma_sync_sg_for_cpu machvec_dma_sync_sg
++#define platform_dma_sync_single_for_device machvec_dma_sync_single
++#define platform_dma_sync_sg_for_device machvec_dma_sync_sg
++#define platform_dma_supported xen_dma_supported
++#define platform_dma_mapping_error xen_dma_mapping_error
++
++#endif /* _ASM_IA64_MACHVEC_XEN_h */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/maddr.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/maddr.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,116 @@
++#ifndef _ASM_IA64_MADDR_H
++#define _ASM_IA64_MADDR_H
+
++#include <linux/kernel.h>
++#include <asm/hypervisor.h>
+#include <xen/features.h>
-+void make_lowmem_page_readonly(void *va, unsigned int feature);
-+void make_lowmem_page_writable(void *va, unsigned int feature);
-+void make_page_readonly(void *va, unsigned int feature);
-+void make_page_writable(void *va, unsigned int feature);
-+void make_pages_readonly(void *va, unsigned int nr, unsigned int feature);
-+void make_pages_writable(void *va, unsigned int nr, unsigned int feature);
-+
-+#define virt_to_ptep(__va) \
-+({ \
-+ pgd_t *__pgd = pgd_offset_k((unsigned long)(__va)); \
-+ pud_t *__pud = pud_offset(__pgd, (unsigned long)(__va)); \
-+ pmd_t *__pmd = pmd_offset(__pud, (unsigned long)(__va)); \
-+ pte_offset_kernel(__pmd, (unsigned long)(__va)); \
-+})
++#include <xen/interface/xen.h>
+
-+#define arbitrary_virt_to_machine(__va) \
-+({ \
-+ maddr_t m = (maddr_t)pte_mfn(*virt_to_ptep(__va)) << PAGE_SHIFT;\
-+ m | ((unsigned long)(__va) & (PAGE_SIZE-1)); \
-+})
++#ifdef CONFIG_XEN
+
-+#endif /* !__ASSEMBLY__ */
++#define INVALID_P2M_ENTRY (~0UL)
+
-+#ifdef CONFIG_FLATMEM
-+#define kern_addr_valid(addr) (1)
-+#endif /* CONFIG_FLATMEM */
++#ifdef CONFIG_XEN_IA64_EXPOSE_P2M
++extern int p2m_initialized;
++extern unsigned long p2m_min_low_pfn;
++extern unsigned long p2m_max_low_pfn;
++extern unsigned long p2m_convert_min_pfn;
++extern unsigned long p2m_convert_max_pfn;
++extern volatile const pte_t* p2m_pte;
++unsigned long p2m_phystomach(unsigned long gpfn);
++#else
++#define p2m_initialized (0)
++#define p2m_phystomach(gpfn) INVALID_MFN
++#endif
+
-+int direct_remap_pfn_range(struct vm_area_struct *vma,
-+ unsigned long address,
-+ unsigned long mfn,
-+ unsigned long size,
-+ pgprot_t prot,
-+ domid_t domid);
-+int direct_kernel_remap_pfn_range(unsigned long address,
-+ unsigned long mfn,
-+ unsigned long size,
-+ pgprot_t prot,
-+ domid_t domid);
-+int create_lookup_pte_addr(struct mm_struct *mm,
-+ unsigned long address,
-+ uint64_t *ptep);
-+int touch_pte_range(struct mm_struct *mm,
-+ unsigned long address,
-+ unsigned long size);
++/* XXX xen page size != page size */
++static inline unsigned long
++pfn_to_mfn_for_dma(unsigned long pfn)
++{
++ unsigned long mfn;
++ if (p2m_initialized)
++ return p2m_phystomach(pfn);
++ mfn = HYPERVISOR_phystomach(pfn);
++ BUG_ON(mfn == 0); /* XXX */
++ BUG_ON(mfn == INVALID_P2M_ENTRY); /* XXX */
++ BUG_ON(mfn == INVALID_MFN);
++ return mfn;
++}
+
-+#define io_remap_pfn_range(vma,from,pfn,size,prot) \
-+direct_remap_pfn_range(vma,from,pfn,size,prot,DOMID_IO)
++static inline unsigned long
++phys_to_machine_for_dma(unsigned long phys)
++{
++ unsigned long machine =
++ pfn_to_mfn_for_dma(phys >> PAGE_SHIFT) << PAGE_SHIFT;
++ machine |= (phys & ~PAGE_MASK);
++ return machine;
++}
+
-+#define MK_IOSPACE_PFN(space, pfn) (pfn)
-+#define GET_IOSPACE(pfn) 0
-+#define GET_PFN(pfn) (pfn)
++static inline unsigned long
++mfn_to_pfn_for_dma(unsigned long mfn)
++{
++ unsigned long pfn;
++ pfn = HYPERVISOR_machtophys(mfn);
++ BUG_ON(pfn == 0);
++ /* BUG_ON(pfn == INVALID_M2P_ENTRY); */
++ return pfn;
++}
+
-+#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
-+#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_DIRTY
-+#define __HAVE_ARCH_PTEP_GET_AND_CLEAR
-+#define __HAVE_ARCH_PTEP_GET_AND_CLEAR_FULL
-+#define __HAVE_ARCH_PTEP_CLEAR_FLUSH
-+#define __HAVE_ARCH_PTEP_SET_WRPROTECT
-+#define __HAVE_ARCH_PTE_SAME
-+#include <asm-generic/pgtable.h>
++static inline unsigned long
++machine_to_phys_for_dma(unsigned long machine)
++{
++ unsigned long phys =
++ mfn_to_pfn_for_dma(machine >> PAGE_SHIFT) << PAGE_SHIFT;
++ phys |= (machine & ~PAGE_MASK);
++ return phys;
++}
+
-+#endif /* _I386_PGTABLE_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/processor.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/processor.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/processor.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/processor.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,741 @@
++#ifdef CONFIG_SPARSEMEM
+/*
-+ * include/asm-i386/processor.h
-+ *
-+ * Copyright (C) 1994 Linus Torvalds
++ * When CONFIG_SPARSEMEM=y, pfn_valid() is defined in
++ * linux/include/linux/mmzone.h. Hoever #include <linux/mmzone.h>
++ * causes the header inclusion hell.
+ */
++static inline int pfn_valid(unsigned long pfn);
++#endif
+
-+#ifndef __ASM_I386_PROCESSOR_H
-+#define __ASM_I386_PROCESSOR_H
++static inline unsigned long
++mfn_to_local_pfn(unsigned long mfn)
++{
++ unsigned long pfn = mfn_to_pfn_for_dma(mfn);
++ if (!pfn_valid(pfn))
++ return INVALID_P2M_ENTRY;
++ return pfn;
++}
+
-+#include <asm/vm86.h>
-+#include <asm/math_emu.h>
-+#include <asm/segment.h>
-+#include <asm/page.h>
-+#include <asm/types.h>
-+#include <asm/sigcontext.h>
-+#include <asm/cpufeature.h>
-+#include <asm/msr.h>
-+#include <asm/system.h>
-+#include <linux/cache.h>
-+#include <linux/threads.h>
-+#include <asm/percpu.h>
-+#include <linux/cpumask.h>
-+#include <xen/interface/physdev.h>
++#else /* !CONFIG_XEN */
+
-+/* flag for disabling the tsc */
-+extern int tsc_disable;
++#define pfn_to_mfn_for_dma(pfn) (pfn)
++#define mfn_to_pfn_for_dma(mfn) (mfn)
++#define phys_to_machine_for_dma(phys) (phys)
++#define machine_to_phys_for_dma(machine) (machine)
++#define mfn_to_local_pfn(mfn) (mfn)
+
-+struct desc_struct {
-+ unsigned long a,b;
-+};
++#endif /* !CONFIG_XEN */
+
-+#define desc_empty(desc) \
-+ (!((desc)->a | (desc)->b))
++#define mfn_to_pfn(mfn) (mfn)
++#define pfn_to_mfn(pfn) (pfn)
+
-+#define desc_equal(desc1, desc2) \
-+ (((desc1)->a == (desc2)->a) && ((desc1)->b == (desc2)->b))
-+/*
-+ * Default implementation of macro that returns current
-+ * instruction pointer ("program counter").
-+ */
-+#define current_text_addr() ({ void *pc; __asm__("movl $1f,%0\n1:":"=g" (pc)); pc; })
++#define mfn_to_virt(mfn) (__va((mfn) << PAGE_SHIFT))
++#define virt_to_mfn(virt) (__pa(virt) >> PAGE_SHIFT)
++#define virt_to_machine(virt) __pa(virt) /* for tpmfront.c */
+
-+/*
-+ * CPU type and hardware bug flags. Kept separately for each CPU.
-+ * Members of this structure are referenced in head.S, so think twice
-+ * before touching them. [mj]
-+ */
++#define set_phys_to_machine(pfn, mfn) do { } while (0)
+
-+struct cpuinfo_x86 {
-+ __u8 x86; /* CPU family */
-+ __u8 x86_vendor; /* CPU vendor */
-+ __u8 x86_model;
-+ __u8 x86_mask;
-+ char wp_works_ok; /* It doesn't on 386's */
-+ char hlt_works_ok; /* Problems on some 486Dx4's and old 386's */
-+ char hard_math;
-+ char rfu;
-+ int cpuid_level; /* Maximum supported CPUID level, -1=no CPUID */
-+ unsigned long x86_capability[NCAPINTS];
-+ char x86_vendor_id[16];
-+ char x86_model_id[64];
-+ int x86_cache_size; /* in KB - valid for CPUS which support this
-+ call */
-+ int x86_cache_alignment; /* In bytes */
-+ char fdiv_bug;
-+ char f00f_bug;
-+ char coma_bug;
-+ char pad0;
-+ int x86_power;
-+ unsigned long loops_per_jiffy;
-+#ifdef CONFIG_SMP
-+ cpumask_t llc_shared_map; /* cpus sharing the last level cache */
-+#endif
-+ unsigned char x86_max_cores; /* cpuid returned max cores value */
-+ unsigned char apicid;
-+#ifdef CONFIG_SMP
-+ unsigned char booted_cores; /* number of cores as seen by OS */
-+ __u8 phys_proc_id; /* Physical processor id. */
-+ __u8 cpu_core_id; /* Core id */
++typedef unsigned long maddr_t; /* to compile netback, netfront */
++#ifndef _ASM_IA64_SN_TYPES_H /* paddr_t is defined in asm-ia64/sn/types.h */
++typedef unsigned long paddr_t;
+#endif
-+} __attribute__((__aligned__(SMP_CACHE_BYTES)));
-+
-+#define X86_VENDOR_INTEL 0
-+#define X86_VENDOR_CYRIX 1
-+#define X86_VENDOR_AMD 2
-+#define X86_VENDOR_UMC 3
-+#define X86_VENDOR_NEXGEN 4
-+#define X86_VENDOR_CENTAUR 5
-+#define X86_VENDOR_RISE 6
-+#define X86_VENDOR_TRANSMETA 7
-+#define X86_VENDOR_NSC 8
-+#define X86_VENDOR_NUM 9
-+#define X86_VENDOR_UNKNOWN 0xff
+
-+/*
-+ * capabilities of CPUs
-+ */
-+
-+extern struct cpuinfo_x86 boot_cpu_data;
-+extern struct cpuinfo_x86 new_cpu_data;
-+#ifndef CONFIG_X86_NO_TSS
-+extern struct tss_struct doublefault_tss;
-+DECLARE_PER_CPU(struct tss_struct, init_tss);
++#ifdef CONFIG_XEN
++int range_straddles_page_boundary(paddr_t p, size_t size);
++#else
++#define range_straddles_page_boundary(addr, size) (0)
+#endif
+
-+#ifdef CONFIG_SMP
-+extern struct cpuinfo_x86 cpu_data[];
-+#define current_cpu_data cpu_data[smp_processor_id()]
++#endif /* _ASM_IA64_MADDR_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/meminit.h
+--- a/include/asm-ia64/meminit.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/meminit.h Wed Sep 10 10:54:08 2008 +0100
+@@ -15,11 +15,17 @@
+ * - initrd (optional)
+ * - command line string
+ * - kernel code & data
++ * - crash dumping code reserved region
+ * - Kernel memory map built from EFI memory map
++ * - xen start info
+ *
+ * More could be added if necessary
+ */
+-#define IA64_MAX_RSVD_REGIONS 6
++#ifndef CONFIG_XEN
++#define IA64_MAX_RSVD_REGIONS 7
+#else
-+#define cpu_data (&boot_cpu_data)
-+#define current_cpu_data boot_cpu_data
++#define IA64_MAX_RSVD_REGIONS 8
++#endif
+
+ struct rsvd_region {
+ unsigned long start; /* virtual address of beginning of element */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/mmu_context.h
+--- a/include/asm-ia64/mmu_context.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/mmu_context.h Wed Sep 10 10:54:08 2008 +0100
+@@ -151,11 +151,7 @@
+ # endif
+ #endif
+
+- ia64_set_rr(0x0000000000000000UL, rr0);
+- ia64_set_rr(0x2000000000000000UL, rr1);
+- ia64_set_rr(0x4000000000000000UL, rr2);
+- ia64_set_rr(0x6000000000000000UL, rr3);
+- ia64_set_rr(0x8000000000000000UL, rr4);
++ ia64_set_rr0_to_rr4(rr0, rr1, rr2, rr3, rr4);
+ ia64_srlz_i(); /* srlz.i implies srlz.d */
+ }
+
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/page.h
+--- a/include/asm-ia64/page.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/page.h Wed Sep 10 10:54:08 2008 +0100
+@@ -119,6 +119,7 @@
+ #endif
+
+ #ifdef CONFIG_FLATMEM
++extern unsigned long max_mapnr;
+ # define pfn_valid(pfn) (((pfn) < max_mapnr) && ia64_pfn_valid(pfn))
+ #elif defined(CONFIG_DISCONTIGMEM)
+ extern unsigned long min_low_pfn;
+@@ -126,7 +127,9 @@
+ # define pfn_valid(pfn) (((pfn) >= min_low_pfn) && ((pfn) < max_low_pfn) && ia64_pfn_valid(pfn))
+ #endif
+
++#ifndef CONFIG_XEN
+ #define page_to_phys(page) (page_to_pfn(page) << PAGE_SHIFT)
+#endif
+ #define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT)
+ #define pfn_to_kaddr(pfn) __va((pfn) << PAGE_SHIFT)
+
+@@ -227,5 +230,17 @@
+ (((current->personality & READ_IMPLIES_EXEC) != 0) \
+ ? VM_EXEC : 0))
+
++#ifndef __ASSEMBLY__
+
-+extern int cpu_llc_id[NR_CPUS];
-+extern char ignore_fpu_irq;
++#include <linux/kernel.h>
++#include <asm/hypervisor.h> /* to compile ioremap.c */
+
-+extern void identify_cpu(struct cpuinfo_x86 *);
-+extern void print_cpu_info(struct cpuinfo_x86 *);
-+extern unsigned int init_intel_cacheinfo(struct cpuinfo_x86 *c);
-+extern unsigned short num_cache_leaves;
++#ifdef CONFIG_XEN
+
-+#ifdef CONFIG_X86_HT
-+extern void detect_ht(struct cpuinfo_x86 *c);
++#include <asm/maddr.h>
++
++#endif /* CONFIG_XEN */
++#endif /* __ASSEMBLY__ */
++
+ # endif /* __KERNEL__ */
+ #endif /* _ASM_IA64_PAGE_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/pgalloc.h
+--- a/include/asm-ia64/pgalloc.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/pgalloc.h Wed Sep 10 10:54:08 2008 +0100
+@@ -125,7 +125,11 @@
+ static inline void
+ pmd_populate(struct mm_struct *mm, pmd_t * pmd_entry, struct page *pte)
+ {
++#ifndef CONFIG_XEN
+ pmd_val(*pmd_entry) = page_to_phys(pte);
+#else
-+static inline void detect_ht(struct cpuinfo_x86 *c) {}
++ pmd_val(*pmd_entry) = page_to_pseudophys(pte);
+#endif
+ }
+
+ static inline void
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/privop.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/privop.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,63 @@
++#ifndef _ASM_IA64_PRIVOP_H
++#define _ASM_IA64_PRIVOP_H
+
++#ifndef _ASM_IA64_INTRINSICS_H
++#error "don't include privop.h directly. instead include intrinsics.h"
++#endif
+/*
-+ * EFLAGS bits
++ * Copyright (C) 2005 Hewlett-Packard Co
++ * Dan Magenheimer <dan.magenheimer at hp.com>
++ *
+ */
-+#define X86_EFLAGS_CF 0x00000001 /* Carry Flag */
-+#define X86_EFLAGS_PF 0x00000004 /* Parity Flag */
-+#define X86_EFLAGS_AF 0x00000010 /* Auxillary carry Flag */
-+#define X86_EFLAGS_ZF 0x00000040 /* Zero Flag */
-+#define X86_EFLAGS_SF 0x00000080 /* Sign Flag */
-+#define X86_EFLAGS_TF 0x00000100 /* Trap Flag */
-+#define X86_EFLAGS_IF 0x00000200 /* Interrupt Flag */
-+#define X86_EFLAGS_DF 0x00000400 /* Direction Flag */
-+#define X86_EFLAGS_OF 0x00000800 /* Overflow Flag */
-+#define X86_EFLAGS_IOPL 0x00003000 /* IOPL mask */
-+#define X86_EFLAGS_NT 0x00004000 /* Nested Task */
-+#define X86_EFLAGS_RF 0x00010000 /* Resume Flag */
-+#define X86_EFLAGS_VM 0x00020000 /* Virtual Mode */
-+#define X86_EFLAGS_AC 0x00040000 /* Alignment Check */
-+#define X86_EFLAGS_VIF 0x00080000 /* Virtual Interrupt Flag */
-+#define X86_EFLAGS_VIP 0x00100000 /* Virtual Interrupt Pending */
-+#define X86_EFLAGS_ID 0x00200000 /* CPUID detection flag */
+
-+/*
-+ * Generic CPUID function
-+ * clear %ecx since some cpus (Cyrix MII) do not set or clear %ecx
-+ * resulting in stale register contents being returned.
-+ */
-+static inline void cpuid(unsigned int op, unsigned int *eax, unsigned int *ebx, unsigned int *ecx, unsigned int *edx)
-+{
-+ __asm__(XEN_CPUID
-+ : "=a" (*eax),
-+ "=b" (*ebx),
-+ "=c" (*ecx),
-+ "=d" (*edx)
-+ : "0" (op), "c"(0));
-+}
++#ifdef CONFIG_XEN
++#include <asm/xen/privop.h>
++#endif
+
-+/* Some CPUID calls want 'count' to be placed in ecx */
-+static inline void cpuid_count(int op, int count, int *eax, int *ebx, int *ecx,
-+ int *edx)
-+{
-+ __asm__(XEN_CPUID
-+ : "=a" (*eax),
-+ "=b" (*ebx),
-+ "=c" (*ecx),
-+ "=d" (*edx)
-+ : "0" (op), "c" (count));
-+}
++#ifndef __ASSEMBLY
+
-+/*
-+ * CPUID functions returning a single datum
-+ */
-+static inline unsigned int cpuid_eax(unsigned int op)
-+{
-+ unsigned int eax;
++#ifndef IA64_PARAVIRTUALIZED
+
-+ __asm__(XEN_CPUID
-+ : "=a" (eax)
-+ : "0" (op)
-+ : "bx", "cx", "dx");
-+ return eax;
-+}
-+static inline unsigned int cpuid_ebx(unsigned int op)
-+{
-+ unsigned int eax, ebx;
++#define ia64_getreg __ia64_getreg
++#define ia64_setreg __ia64_setreg
++#define ia64_hint __ia64_hint
++#define ia64_thash __ia64_thash
++#define ia64_itci __ia64_itci
++#define ia64_itcd __ia64_itcd
++#define ia64_itri __ia64_itri
++#define ia64_itrd __ia64_itrd
++#define ia64_tpa __ia64_tpa
++#define ia64_set_ibr __ia64_set_ibr
++#define ia64_set_pkr __ia64_set_pkr
++#define ia64_set_pmc __ia64_set_pmc
++#define ia64_set_pmd __ia64_set_pmd
++#define ia64_set_rr __ia64_set_rr
++#define ia64_get_cpuid __ia64_get_cpuid
++#define ia64_get_ibr __ia64_get_ibr
++#define ia64_get_pkr __ia64_get_pkr
++#define ia64_get_pmc __ia64_get_pmc
++#define ia64_get_pmd __ia64_get_pmd
++#define ia64_get_rr __ia64_get_rr
++#define ia64_fc __ia64_fc
++#define ia64_ssm __ia64_ssm
++#define ia64_rsm __ia64_rsm
++#define ia64_ptce __ia64_ptce
++#define ia64_ptcga __ia64_ptcga
++#define ia64_ptcl __ia64_ptcl
++#define ia64_ptri __ia64_ptri
++#define ia64_ptrd __ia64_ptrd
++#define ia64_get_psr_i __ia64_get_psr_i
++#define ia64_intrin_local_irq_restore __ia64_intrin_local_irq_restore
++#define ia64_leave_kernel __ia64_leave_kernel
++#define ia64_leave_syscall __ia64_leave_syscall
++#define ia64_trace_syscall __ia64_trace_syscall
++#define ia64_ret_from_clone __ia64_ret_from_clone
++#define ia64_switch_to __ia64_switch_to
++#define ia64_pal_call_static __ia64_pal_call_static
++#define ia64_set_rr0_to_rr4 __ia64_set_rr0_to_rr4
+
-+ __asm__(XEN_CPUID
-+ : "=a" (eax), "=b" (ebx)
-+ : "0" (op)
-+ : "cx", "dx" );
-+ return ebx;
-+}
-+static inline unsigned int cpuid_ecx(unsigned int op)
-+{
-+ unsigned int eax, ecx;
++#endif /* !IA64_PARAVIRTUALIZED */
+
-+ __asm__(XEN_CPUID
-+ : "=a" (eax), "=c" (ecx)
-+ : "0" (op)
-+ : "bx", "dx" );
-+ return ecx;
-+}
-+static inline unsigned int cpuid_edx(unsigned int op)
++#endif /* !__ASSEMBLY */
++
++#endif /* _ASM_IA64_PRIVOP_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/sal.h
+--- a/include/asm-ia64/sal.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/sal.h Wed Sep 10 10:54:08 2008 +0100
+@@ -42,6 +42,9 @@
+ #include <asm/pal.h>
+ #include <asm/system.h>
+ #include <asm/fpu.h>
++#ifdef CONFIG_XEN
++#include <asm/xen/xencomm.h>
++#endif
+
+ extern spinlock_t sal_lock;
+
+@@ -687,10 +690,43 @@
+ /* Get the processor and platform information logged by SAL with respect to the machine
+ * state at the time of the MCAs, INITs, CMCs, or CPEs.
+ */
++#ifdef CONFIG_XEN
++static inline u64 ia64_sal_get_state_info_size (u64 sal_info_type);
++typedef struct ia64_mca_xencomm_t {
++ void *record;
++ struct xencomm_handle *handle;
++ struct list_head list;
++} ia64_mca_xencomm_t;
++extern struct list_head ia64_mca_xencomm_list;
++extern spinlock_t ia64_mca_xencomm_lock;
++#endif
++
+ static inline u64
+ ia64_sal_get_state_info (u64 sal_info_type, u64 *sal_info)
+ {
+ struct ia64_sal_retval isrv;
++#ifdef CONFIG_XEN
++ if (is_running_on_xen()) {
++ ia64_mca_xencomm_t *entry;
++ struct xencomm_handle *desc = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&ia64_mca_xencomm_lock, flags);
++ list_for_each_entry(entry, &ia64_mca_xencomm_list, list) {
++ if (entry->record == sal_info) {
++ desc = entry->handle;
++ break;
++ }
++ }
++ spin_unlock_irqrestore(&ia64_mca_xencomm_lock, flags);
++
++ if (desc == NULL)
++ return 0;
++
++ SAL_CALL_REENTRANT(isrv, SAL_GET_STATE_INFO, sal_info_type, 0,
++ desc, 0, 0, 0, 0);
++ } else
++#endif
+ SAL_CALL_REENTRANT(isrv, SAL_GET_STATE_INFO, sal_info_type, 0,
+ sal_info, 0, 0, 0, 0);
+ if (isrv.status)
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/sn/sn_sal.h
+--- a/include/asm-ia64/sn/sn_sal.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/sn/sn_sal.h Wed Sep 10 10:54:08 2008 +0100
+@@ -86,6 +86,8 @@
+ #define SN_SAL_SET_OS_FEATURE_SET 0x02000066
+ #define SN_SAL_INJECT_ERROR 0x02000067
+ #define SN_SAL_SET_CPU_NUMBER 0x02000068
++
++#define SN_SAL_KERNEL_LAUNCH_EVENT 0x02000069
+
+ /*
+ * Service-specific constants
+@@ -1154,4 +1156,11 @@
+ SAL_CALL_NOLOCK(rv, SN_SAL_SET_CPU_NUMBER, cpu, 0, 0, 0, 0, 0, 0);
+ return rv.status;
+ }
++static inline int
++ia64_sn_kernel_launch_event(void)
+{
-+ unsigned int eax, edx;
++ struct ia64_sal_retval rv;
++ SAL_CALL_NOLOCK(rv, SN_SAL_KERNEL_LAUNCH_EVENT, 0, 0, 0, 0, 0, 0, 0);
++ return rv.status;
++}
+ #endif /* _ASM_IA64_SN_SN_SAL_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/sn/types.h
+--- a/include/asm-ia64/sn/types.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/sn/types.h Wed Sep 10 10:54:08 2008 +0100
+@@ -20,7 +20,9 @@
+ typedef unsigned char slabid_t; /* slab (asic) within slot */
+ typedef u64 nic_t;
+ typedef unsigned long iopaddr_t;
++#ifndef _ASM_IA64_MADDR_H /* paddr_t is defined in asm-ia64/maddr.h */
+ typedef unsigned long paddr_t;
++#endif
+ typedef short cnodeid_t;
+
+ #endif /* _ASM_IA64_SN_TYPES_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/swiotlb.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/swiotlb.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,41 @@
++#ifndef _ASM_SWIOTLB_H
++#define _ASM_SWIOTLB_H 1
++
++/* SWIOTLB interface */
++
++extern dma_addr_t swiotlb_map_single(struct device *hwdev, void *ptr, size_t size,
++ int dir);
++extern void swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr,
++ size_t size, int dir);
++extern void swiotlb_sync_single_for_cpu(struct device *hwdev,
++ dma_addr_t dev_addr,
++ size_t size, int dir);
++extern void swiotlb_sync_single_for_device(struct device *hwdev,
++ dma_addr_t dev_addr,
++ size_t size, int dir);
++extern void swiotlb_sync_sg_for_cpu(struct device *hwdev,
++ struct scatterlist *sg, int nelems,
++ int dir);
++extern void swiotlb_sync_sg_for_device(struct device *hwdev,
++ struct scatterlist *sg, int nelems,
++ int dir);
++extern int swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg,
++ int nents, int direction);
++extern void swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg,
++ int nents, int direction);
++extern int swiotlb_dma_mapping_error(dma_addr_t dma_addr);
++extern dma_addr_t swiotlb_map_page(struct device *hwdev, struct page *page,
++ unsigned long offset, size_t size,
++ enum dma_data_direction direction);
++extern void swiotlb_unmap_page(struct device *hwdev, dma_addr_t dma_address,
++ size_t size, enum dma_data_direction direction);
++extern int swiotlb_dma_supported(struct device *hwdev, u64 mask);
++extern void swiotlb_init(void);
+
-+ __asm__(XEN_CPUID
-+ : "=a" (eax), "=d" (edx)
-+ : "0" (op)
-+ : "bx", "cx");
-+ return edx;
-+}
++#ifdef CONFIG_SWIOTLB
++extern int swiotlb;
++#else
++#define swiotlb 0
++#endif
+
-+#define load_cr3(pgdir) write_cr3(__pa(pgdir))
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/synch_bitops.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/synch_bitops.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,61 @@
++#ifndef __XEN_SYNCH_BITOPS_H__
++#define __XEN_SYNCH_BITOPS_H__
+
+/*
-+ * Intel CPU features in CR4
++ * Copyright 1992, Linus Torvalds.
++ * Heavily modified to provide guaranteed strong synchronisation
++ * when communicating with Xen or other guest OSes running on other CPUs.
+ */
-+#define X86_CR4_VME 0x0001 /* enable vm86 extensions */
-+#define X86_CR4_PVI 0x0002 /* virtual interrupts flag enable */
-+#define X86_CR4_TSD 0x0004 /* disable time stamp at ipl 3 */
-+#define X86_CR4_DE 0x0008 /* enable debugging extensions */
-+#define X86_CR4_PSE 0x0010 /* enable page size extensions */
-+#define X86_CR4_PAE 0x0020 /* enable physical address extensions */
-+#define X86_CR4_MCE 0x0040 /* Machine check enable */
-+#define X86_CR4_PGE 0x0080 /* enable global pages */
-+#define X86_CR4_PCE 0x0100 /* enable performance counters at ipl 3 */
-+#define X86_CR4_OSFXSR 0x0200 /* enable fast FPU save and restore */
-+#define X86_CR4_OSXMMEXCPT 0x0400 /* enable unmasked SSE exceptions */
+
-+/*
-+ * Save the cr4 feature set we're using (ie
-+ * Pentium 4MB enable and PPro Global page
-+ * enable), so that any CPU's that boot up
-+ * after us can get the correct flags.
-+ */
-+extern unsigned long mmu_cr4_features;
++#define ADDR (*(volatile long *) addr)
+
-+static inline void set_in_cr4 (unsigned long mask)
++static __inline__ void synch_set_bit(int nr, volatile void * addr)
+{
-+ unsigned cr4;
-+ mmu_cr4_features |= mask;
-+ cr4 = read_cr4();
-+ cr4 |= mask;
-+ write_cr4(cr4);
++ set_bit(nr, addr);
+}
+
-+static inline void clear_in_cr4 (unsigned long mask)
++static __inline__ void synch_clear_bit(int nr, volatile void * addr)
+{
-+ unsigned cr4;
-+ mmu_cr4_features &= ~mask;
-+ cr4 = read_cr4();
-+ cr4 &= ~mask;
-+ write_cr4(cr4);
++ clear_bit(nr, addr);
+}
+
-+/*
-+ * NSC/Cyrix CPU configuration register indexes
-+ */
-+
-+#define CX86_PCR0 0x20
-+#define CX86_GCR 0xb8
-+#define CX86_CCR0 0xc0
-+#define CX86_CCR1 0xc1
-+#define CX86_CCR2 0xc2
-+#define CX86_CCR3 0xc3
-+#define CX86_CCR4 0xe8
-+#define CX86_CCR5 0xe9
-+#define CX86_CCR6 0xea
-+#define CX86_CCR7 0xeb
-+#define CX86_PCR1 0xf0
-+#define CX86_DIR0 0xfe
-+#define CX86_DIR1 0xff
-+#define CX86_ARR_BASE 0xc4
-+#define CX86_RCR_BASE 0xdc
-+
-+/*
-+ * NSC/Cyrix CPU indexed register access macros
-+ */
-+
-+#define getCx86(reg) ({ outb((reg), 0x22); inb(0x23); })
++static __inline__ void synch_change_bit(int nr, volatile void * addr)
++{
++ change_bit(nr, addr);
++}
+
-+#define setCx86(reg, data) do { \
-+ outb((reg), 0x22); \
-+ outb((data), 0x23); \
-+} while (0)
++static __inline__ int synch_test_and_set_bit(int nr, volatile void * addr)
++{
++ return test_and_set_bit(nr, addr);
++}
+
-+/* Stop speculative execution */
-+static inline void sync_core(void)
++static __inline__ int synch_test_and_clear_bit(int nr, volatile void * addr)
+{
-+ int tmp;
-+ asm volatile("cpuid" : "=a" (tmp) : "0" (1) : "ebx","ecx","edx","memory");
++ return test_and_clear_bit(nr, addr);
+}
+
-+static inline void __monitor(const void *eax, unsigned long ecx,
-+ unsigned long edx)
++static __inline__ int synch_test_and_change_bit(int nr, volatile void * addr)
+{
-+ /* "monitor %eax,%ecx,%edx;" */
-+ asm volatile(
-+ ".byte 0x0f,0x01,0xc8;"
-+ : :"a" (eax), "c" (ecx), "d"(edx));
++ return test_and_change_bit(nr, addr);
+}
+
-+static inline void __mwait(unsigned long eax, unsigned long ecx)
++static __inline__ int synch_const_test_bit(int nr, const volatile void * addr)
+{
-+ /* "mwait %eax,%ecx;" */
-+ asm volatile(
-+ ".byte 0x0f,0x01,0xc9;"
-+ : :"a" (eax), "c" (ecx));
++ return test_bit(nr, addr);
+}
+
-+/* from system description table in BIOS. Mostly for MCA use, but
-+others may find it useful. */
-+extern unsigned int machine_id;
-+extern unsigned int machine_submodel_id;
-+extern unsigned int BIOS_revision;
-+extern unsigned int mca_pentium_flag;
++static __inline__ int synch_var_test_bit(int nr, volatile void * addr)
++{
++ return test_bit(nr, addr);
++}
+
-+/* Boot loader type from the setup header */
-+extern int bootloader_type;
++#define synch_cmpxchg ia64_cmpxchg4_acq
+
-+/*
-+ * User space process size: 3GB (default).
-+ */
-+#define TASK_SIZE (PAGE_OFFSET)
++#define synch_test_bit(nr,addr) \
++(__builtin_constant_p(nr) ? \
++ synch_const_test_bit((nr),(addr)) : \
++ synch_var_test_bit((nr),(addr)))
+
-+/* This decides where the kernel will search for a free chunk of vm
-+ * space during mmap's.
-+ */
-+#define TASK_UNMAPPED_BASE (PAGE_ALIGN(TASK_SIZE / 3))
++#define synch_cmpxchg_subword synch_cmpxchg
+
-+#define HAVE_ARCH_PICK_MMAP_LAYOUT
++#endif /* __XEN_SYNCH_BITOPS_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/system.h
+--- a/include/asm-ia64/system.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-ia64/system.h Wed Sep 10 10:54:08 2008 +0100
+@@ -123,7 +123,7 @@
+ #define __local_irq_save(x) \
+ do { \
+ ia64_stop(); \
+- (x) = ia64_getreg(_IA64_REG_PSR); \
++ (x) = ia64_get_psr_i(); \
+ ia64_stop(); \
+ ia64_rsm(IA64_PSR_I); \
+ } while (0)
+@@ -171,7 +171,7 @@
+ #endif /* !CONFIG_IA64_DEBUG_IRQ */
+
+ #define local_irq_enable() ({ ia64_stop(); ia64_ssm(IA64_PSR_I); ia64_srlz_d(); })
+-#define local_save_flags(flags) ({ ia64_stop(); (flags) = ia64_getreg(_IA64_REG_PSR); })
++#define local_save_flags(flags) ({ ia64_stop(); (flags) = ia64_get_psr_i(); })
+
+ #define irqs_disabled() \
+ ({ \
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/xen/privop.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/xen/privop.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,550 @@
++#ifndef _ASM_IA64_XEN_PRIVOP_H
++#define _ASM_IA64_XEN_PRIVOP_H
+
+/*
-+ * Size of io_bitmap.
++ * Copyright (C) 2005 Hewlett-Packard Co
++ * Dan Magenheimer <dan.magenheimer at hp.com>
++ *
++ * Paravirtualizations of privileged operations for Xen/ia64
++ *
+ */
-+#define IO_BITMAP_BITS 65536
-+#define IO_BITMAP_BYTES (IO_BITMAP_BITS/8)
-+#define IO_BITMAP_LONGS (IO_BITMAP_BYTES/sizeof(long))
-+#ifndef CONFIG_X86_NO_TSS
-+#define IO_BITMAP_OFFSET offsetof(struct tss_struct,io_bitmap)
-+#endif
-+#define INVALID_IO_BITMAP_OFFSET 0x8000
-+#define INVALID_IO_BITMAP_OFFSET_LAZY 0x9000
-+
-+struct i387_fsave_struct {
-+ long cwd;
-+ long swd;
-+ long twd;
-+ long fip;
-+ long fcs;
-+ long foo;
-+ long fos;
-+ long st_space[20]; /* 8*10 bytes for each FP-reg = 80 bytes */
-+ long status; /* software status information */
-+};
+
-+struct i387_fxsave_struct {
-+ unsigned short cwd;
-+ unsigned short swd;
-+ unsigned short twd;
-+ unsigned short fop;
-+ long fip;
-+ long fcs;
-+ long foo;
-+ long fos;
-+ long mxcsr;
-+ long mxcsr_mask;
-+ long st_space[32]; /* 8*16 bytes for each FP-reg = 128 bytes */
-+ long xmm_space[32]; /* 8*16 bytes for each XMM-reg = 128 bytes */
-+ long padding[56];
-+} __attribute__ ((aligned (16)));
++#ifndef __ASSEMBLY__
++#include <linux/types.h> /* arch-ia64.h requires uint64_t */
++#include <linux/stringify.h>
++#endif
++#include <xen/interface/arch-ia64.h>
+
-+struct i387_soft_struct {
-+ long cwd;
-+ long swd;
-+ long twd;
-+ long fip;
-+ long fcs;
-+ long foo;
-+ long fos;
-+ long st_space[20]; /* 8*10 bytes for each FP-reg = 80 bytes */
-+ unsigned char ftop, changed, lookahead, no_update, rm, alimit;
-+ struct info *info;
-+ unsigned long entry_eip;
-+};
++#define IA64_PARAVIRTUALIZED
+
-+union i387_union {
-+ struct i387_fsave_struct fsave;
-+ struct i387_fxsave_struct fxsave;
-+ struct i387_soft_struct soft;
-+};
++/* At 1 MB, before per-cpu space but still addressable using addl instead
++ of movl. */
++#define XSI_BASE 0xfffffffffff00000
+
-+typedef struct {
-+ unsigned long seg;
-+} mm_segment_t;
++/* Address of mapped regs. */
++#define XMAPPEDREGS_BASE (XSI_BASE + XSI_SIZE)
+
-+struct thread_struct;
++#ifdef __ASSEMBLY__
++#define XEN_HYPER_RFI break HYPERPRIVOP_RFI
++#define XEN_HYPER_RSM_PSR_DT break HYPERPRIVOP_RSM_DT
++#define XEN_HYPER_SSM_PSR_DT break HYPERPRIVOP_SSM_DT
++#define XEN_HYPER_COVER break HYPERPRIVOP_COVER
++#define XEN_HYPER_ITC_D break HYPERPRIVOP_ITC_D
++#define XEN_HYPER_ITC_I break HYPERPRIVOP_ITC_I
++#define XEN_HYPER_SSM_I break HYPERPRIVOP_SSM_I
++#define XEN_HYPER_GET_IVR break HYPERPRIVOP_GET_IVR
++#define XEN_HYPER_GET_TPR break HYPERPRIVOP_GET_TPR
++#define XEN_HYPER_SET_TPR break HYPERPRIVOP_SET_TPR
++#define XEN_HYPER_EOI break HYPERPRIVOP_EOI
++#define XEN_HYPER_SET_ITM break HYPERPRIVOP_SET_ITM
++#define XEN_HYPER_THASH break HYPERPRIVOP_THASH
++#define XEN_HYPER_PTC_GA break HYPERPRIVOP_PTC_GA
++#define XEN_HYPER_ITR_D break HYPERPRIVOP_ITR_D
++#define XEN_HYPER_GET_RR break HYPERPRIVOP_GET_RR
++#define XEN_HYPER_SET_RR break HYPERPRIVOP_SET_RR
++#define XEN_HYPER_SET_KR break HYPERPRIVOP_SET_KR
++#define XEN_HYPER_FC break HYPERPRIVOP_FC
++#define XEN_HYPER_GET_CPUID break HYPERPRIVOP_GET_CPUID
++#define XEN_HYPER_GET_PMD break HYPERPRIVOP_GET_PMD
++#define XEN_HYPER_GET_EFLAG break HYPERPRIVOP_GET_EFLAG
++#define XEN_HYPER_SET_EFLAG break HYPERPRIVOP_SET_EFLAG
++#define XEN_HYPER_GET_PSR break HYPERPRIVOP_GET_PSR
+
-+#ifndef CONFIG_X86_NO_TSS
-+struct tss_struct {
-+ unsigned short back_link,__blh;
-+ unsigned long esp0;
-+ unsigned short ss0,__ss0h;
-+ unsigned long esp1;
-+ unsigned short ss1,__ss1h; /* ss1 is used to cache MSR_IA32_SYSENTER_CS */
-+ unsigned long esp2;
-+ unsigned short ss2,__ss2h;
-+ unsigned long __cr3;
-+ unsigned long eip;
-+ unsigned long eflags;
-+ unsigned long eax,ecx,edx,ebx;
-+ unsigned long esp;
-+ unsigned long ebp;
-+ unsigned long esi;
-+ unsigned long edi;
-+ unsigned short es, __esh;
-+ unsigned short cs, __csh;
-+ unsigned short ss, __ssh;
-+ unsigned short ds, __dsh;
-+ unsigned short fs, __fsh;
-+ unsigned short gs, __gsh;
-+ unsigned short ldt, __ldth;
-+ unsigned short trace, io_bitmap_base;
-+ /*
-+ * The extra 1 is there because the CPU will access an
-+ * additional byte beyond the end of the IO permission
-+ * bitmap. The extra byte must be all 1 bits, and must
-+ * be within the limit.
-+ */
-+ unsigned long io_bitmap[IO_BITMAP_LONGS + 1];
-+ /*
-+ * Cache the current maximum and the last task that used the bitmap:
-+ */
-+ unsigned long io_bitmap_max;
-+ struct thread_struct *io_bitmap_owner;
-+ /*
-+ * pads the TSS to be cacheline-aligned (size is 0x100)
-+ */
-+ unsigned long __cacheline_filler[35];
-+ /*
-+ * .. and then another 0x100 bytes for emergency kernel stack
-+ */
-+ unsigned long stack[64];
-+} __attribute__((packed));
++#define XSI_IFS (XSI_BASE + XSI_IFS_OFS)
++#define XSI_PRECOVER_IFS (XSI_BASE + XSI_PRECOVER_IFS_OFS)
++#define XSI_IFA (XSI_BASE + XSI_IFA_OFS)
++#define XSI_ISR (XSI_BASE + XSI_ISR_OFS)
++#define XSI_IIM (XSI_BASE + XSI_IIM_OFS)
++#define XSI_ITIR (XSI_BASE + XSI_ITIR_OFS)
++#define XSI_PSR_I_ADDR (XSI_BASE + XSI_PSR_I_ADDR_OFS)
++#define XSI_PSR_IC (XSI_BASE + XSI_PSR_IC_OFS)
++#define XSI_IPSR (XSI_BASE + XSI_IPSR_OFS)
++#define XSI_IIP (XSI_BASE + XSI_IIP_OFS)
++#define XSI_B1NAT (XSI_BASE + XSI_B1NATS_OFS)
++#define XSI_BANK1_R16 (XSI_BASE + XSI_BANK1_R16_OFS)
++#define XSI_BANKNUM (XSI_BASE + XSI_BANKNUM_OFS)
++#define XSI_IHA (XSI_BASE + XSI_IHA_OFS)
+#endif
+
-+#define ARCH_MIN_TASKALIGN 16
-+
-+struct thread_struct {
-+/* cached TLS descriptors. */
-+ struct desc_struct tls_array[GDT_ENTRY_TLS_ENTRIES];
-+ unsigned long esp0;
-+ unsigned long sysenter_cs;
-+ unsigned long eip;
-+ unsigned long esp;
-+ unsigned long fs;
-+ unsigned long gs;
-+/* Hardware debugging registers */
-+ unsigned long debugreg[8]; /* %%db0-7 debug registers */
-+/* fault info */
-+ unsigned long cr2, trap_no, error_code;
-+/* floating point info */
-+ union i387_union i387;
-+/* virtual 86 mode info */
-+ struct vm86_struct __user * vm86_info;
-+ unsigned long screen_bitmap;
-+ unsigned long v86flags, v86mask, saved_esp0;
-+ unsigned int saved_fs, saved_gs;
-+/* IO permissions */
-+ unsigned long *io_bitmap_ptr;
-+ unsigned long iopl;
-+/* max allowed port in the bitmap, in bytes: */
-+ unsigned long io_bitmap_max;
-+};
++#ifndef __ASSEMBLY__
++#define XEN_HYPER_SSM_I asm("break %0" : : "i" (HYPERPRIVOP_SSM_I): "memory")
+
-+#define INIT_THREAD { \
-+ .vm86_info = NULL, \
-+ .sysenter_cs = __KERNEL_CS, \
-+ .io_bitmap_ptr = NULL, \
-+}
++/************************************************/
++/* Instructions paravirtualized for correctness */
++/************************************************/
+
-+#ifndef CONFIG_X86_NO_TSS
-+/*
-+ * Note that the .io_bitmap member must be extra-big. This is because
-+ * the CPU will access an additional byte beyond the end of the IO
-+ * permission bitmap. The extra byte must be all 1 bits, and must
-+ * be within the limit.
-+ */
-+#define INIT_TSS { \
-+ .esp0 = sizeof(init_stack) + (long)&init_stack, \
-+ .ss0 = __KERNEL_DS, \
-+ .ss1 = __KERNEL_CS, \
-+ .io_bitmap_base = INVALID_IO_BITMAP_OFFSET, \
-+ .io_bitmap = { [ 0 ... IO_BITMAP_LONGS] = ~0 }, \
++/* "fc" and "thash" are privilege-sensitive instructions, meaning they
++ * may have different semantics depending on whether they are executed
++ * at PL0 vs PL!=0. When paravirtualized, these instructions mustn't
++ * be allowed to execute directly, lest incorrect semantics result. */
++#ifdef ASM_SUPPORTED
++static inline void
++xen_fc(unsigned long addr)
++{
++ register __u64 __addr asm ("r8") = addr;
++ asm volatile ("break %0":: "i"(HYPERPRIVOP_FC), "r"(__addr): "memory");
+}
+
-+static inline void __load_esp0(struct tss_struct *tss, struct thread_struct *thread)
++static inline unsigned long
++xen_thash(unsigned long addr)
+{
-+ tss->esp0 = thread->esp0;
-+ /* This can only happen when SEP is enabled, no need to test "SEP"arately */
-+ if (unlikely(tss->ss1 != thread->sysenter_cs)) {
-+ tss->ss1 = thread->sysenter_cs;
-+ wrmsr(MSR_IA32_SYSENTER_CS, thread->sysenter_cs, 0);
-+ }
++ register __u64 ia64_intri_res asm ("r8");
++ register __u64 __addr asm ("r8") = addr;
++ asm volatile ("break %1":
++ "=r"(ia64_intri_res):
++ "i"(HYPERPRIVOP_THASH), "0"(__addr));
++ return ia64_intri_res;
+}
-+#define load_esp0(tss, thread) \
-+ __load_esp0(tss, thread)
+#else
-+#define load_esp0(tss, thread) \
-+ HYPERVISOR_stack_switch(__KERNEL_DS, (thread)->esp0)
++extern void xen_fc(unsigned long addr);
++extern unsigned long xen_thash(unsigned long addr);
+#endif
+
-+#define start_thread(regs, new_eip, new_esp) do { \
-+ __asm__("movl %0,%%fs ; movl %0,%%gs": :"r" (0)); \
-+ set_fs(USER_DS); \
-+ regs->xds = __USER_DS; \
-+ regs->xes = __USER_DS; \
-+ regs->xss = __USER_DS; \
-+ regs->xcs = __USER_CS; \
-+ regs->eip = new_eip; \
-+ regs->esp = new_esp; \
++#define ia64_fc(addr) \
++do { \
++ if (is_running_on_xen()) \
++ xen_fc((unsigned long)(addr)); \
++ else \
++ __ia64_fc(addr); \
+} while (0)
+
-+/*
-+ * These special macros can be used to get or set a debugging register
-+ */
-+#define get_debugreg(var, register) \
-+ (var) = HYPERVISOR_get_debugreg((register))
-+#define set_debugreg(value, register) \
-+ HYPERVISOR_set_debugreg((register), (value))
-+
-+/*
-+ * Set IOPL bits in EFLAGS from given mask
-+ */
-+static inline void set_iopl_mask(unsigned mask)
-+{
-+ struct physdev_set_iopl set_iopl;
-+
-+ /* Force the change at ring 0. */
-+ set_iopl.iopl = (mask == 0) ? 1 : (mask >> 12) & 3;
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
-+}
-+
-+/* Forward declaration, a strange C thing */
-+struct task_struct;
-+struct mm_struct;
-+
-+/* Free all resources held by a thread. */
-+extern void release_thread(struct task_struct *);
-+
-+/* Prepare to copy thread state - unlazy all lazy status */
-+extern void prepare_to_copy(struct task_struct *tsk);
-+
-+/*
-+ * create a kernel thread without removing it from tasklists
-+ */
-+extern int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags);
-+
-+extern unsigned long thread_saved_pc(struct task_struct *tsk);
-+void show_trace(struct task_struct *task, struct pt_regs *regs, unsigned long *stack);
-+
-+unsigned long get_wchan(struct task_struct *p);
-+
-+#define THREAD_SIZE_LONGS (THREAD_SIZE/sizeof(unsigned long))
-+#define KSTK_TOP(info) \
-+({ \
-+ unsigned long *__ptr = (unsigned long *)(info); \
-+ (unsigned long)(&__ptr[THREAD_SIZE_LONGS]); \
-+})
-+
-+/*
-+ * The below -8 is to reserve 8 bytes on top of the ring0 stack.
-+ * This is necessary to guarantee that the entire "struct pt_regs"
-+ * is accessable even if the CPU haven't stored the SS/ESP registers
-+ * on the stack (interrupt gate does not save these registers
-+ * when switching to the same priv ring).
-+ * Therefore beware: accessing the xss/esp fields of the
-+ * "struct pt_regs" is possible, but they may contain the
-+ * completely wrong values.
-+ */
-+#define task_pt_regs(task) \
-+({ \
-+ struct pt_regs *__regs__; \
-+ __regs__ = (struct pt_regs *)(KSTK_TOP(task_stack_page(task))-8); \
-+ __regs__ - 1; \
++#define ia64_thash(addr) \
++({ \
++ unsigned long ia64_intri_res; \
++ if (is_running_on_xen()) \
++ ia64_intri_res = \
++ xen_thash((unsigned long)(addr)); \
++ else \
++ ia64_intri_res = __ia64_thash(addr); \
++ ia64_intri_res; \
+})
+
-+#define KSTK_EIP(task) (task_pt_regs(task)->eip)
-+#define KSTK_ESP(task) (task_pt_regs(task)->esp)
-+
-+
-+struct microcode_header {
-+ unsigned int hdrver;
-+ unsigned int rev;
-+ unsigned int date;
-+ unsigned int sig;
-+ unsigned int cksum;
-+ unsigned int ldrver;
-+ unsigned int pf;
-+ unsigned int datasize;
-+ unsigned int totalsize;
-+ unsigned int reserved[3];
-+};
-+
-+struct microcode {
-+ struct microcode_header hdr;
-+ unsigned int bits[0];
-+};
-+
-+typedef struct microcode microcode_t;
-+typedef struct microcode_header microcode_header_t;
-+
-+/* microcode format is extended from prescott processors */
-+struct extended_signature {
-+ unsigned int sig;
-+ unsigned int pf;
-+ unsigned int cksum;
-+};
-+
-+struct extended_sigtable {
-+ unsigned int count;
-+ unsigned int cksum;
-+ unsigned int reserved[3];
-+ struct extended_signature sigs[0];
-+};
++/* Note that "ttag" and "cover" are also privilege-sensitive; "ttag"
++ * is not currently used (though it may be in a long-format VHPT system!)
++ * and the semantics of cover only change if psr.ic is off which is very
++ * rare (and currently non-existent outside of assembly code */
+
-+/* REP NOP (PAUSE) is a good thing to insert into busy-wait loops. */
-+static inline void rep_nop(void)
++/* There are also privilege-sensitive registers. These registers are
++ * readable at any privilege level but only writable at PL0. */
++#ifdef ASM_SUPPORTED
++static inline unsigned long
++xen_get_cpuid(int index)
+{
-+ __asm__ __volatile__("rep;nop": : :"memory");
++ register __u64 ia64_intri_res asm ("r8");
++ register __u64 __index asm ("r8") = index;
++ asm volatile ("break %1":
++ "=r"(ia64_intri_res):
++ "i"(HYPERPRIVOP_GET_CPUID), "0"(__index));
++ return ia64_intri_res;
+}
+
-+#define cpu_relax() rep_nop()
++static inline unsigned long
++xen_get_pmd(int index)
++{
++ register __u64 ia64_intri_res asm ("r8");
++ register __u64 __index asm ("r8") = index;
++ asm volatile ("break %1":
++ "=r"(ia64_intri_res):
++ "i"(HYPERPRIVOP_GET_PMD), "0O"(__index));
++ return ia64_intri_res;
++}
++#else
++extern unsigned long xen_get_cpuid(int index);
++extern unsigned long xen_get_pmd(int index);
++#endif
+
-+/* generic versions from gas */
-+#define GENERIC_NOP1 ".byte 0x90\n"
-+#define GENERIC_NOP2 ".byte 0x89,0xf6\n"
-+#define GENERIC_NOP3 ".byte 0x8d,0x76,0x00\n"
-+#define GENERIC_NOP4 ".byte 0x8d,0x74,0x26,0x00\n"
-+#define GENERIC_NOP5 GENERIC_NOP1 GENERIC_NOP4
-+#define GENERIC_NOP6 ".byte 0x8d,0xb6,0x00,0x00,0x00,0x00\n"
-+#define GENERIC_NOP7 ".byte 0x8d,0xb4,0x26,0x00,0x00,0x00,0x00\n"
-+#define GENERIC_NOP8 GENERIC_NOP1 GENERIC_NOP7
++#define ia64_get_cpuid(i) \
++({ \
++ unsigned long ia64_intri_res; \
++ if (is_running_on_xen()) \
++ ia64_intri_res = xen_get_cpuid(i); \
++ else \
++ ia64_intri_res = __ia64_get_cpuid(i); \
++ ia64_intri_res; \
++})
+
-+/* Opteron nops */
-+#define K8_NOP1 GENERIC_NOP1
-+#define K8_NOP2 ".byte 0x66,0x90\n"
-+#define K8_NOP3 ".byte 0x66,0x66,0x90\n"
-+#define K8_NOP4 ".byte 0x66,0x66,0x66,0x90\n"
-+#define K8_NOP5 K8_NOP3 K8_NOP2
-+#define K8_NOP6 K8_NOP3 K8_NOP3
-+#define K8_NOP7 K8_NOP4 K8_NOP3
-+#define K8_NOP8 K8_NOP4 K8_NOP4
++#define ia64_get_pmd(i) \
++({ \
++ unsigned long ia64_intri_res; \
++ if (is_running_on_xen()) \
++ ia64_intri_res = xen_get_pmd(i); \
++ else \
++ ia64_intri_res = __ia64_get_pmd(i); \
++ ia64_intri_res; \
++})
+
-+/* K7 nops */
-+/* uses eax dependencies (arbitary choice) */
-+#define K7_NOP1 GENERIC_NOP1
-+#define K7_NOP2 ".byte 0x8b,0xc0\n"
-+#define K7_NOP3 ".byte 0x8d,0x04,0x20\n"
-+#define K7_NOP4 ".byte 0x8d,0x44,0x20,0x00\n"
-+#define K7_NOP5 K7_NOP4 ASM_NOP1
-+#define K7_NOP6 ".byte 0x8d,0x80,0,0,0,0\n"
-+#define K7_NOP7 ".byte 0x8D,0x04,0x05,0,0,0,0\n"
-+#define K7_NOP8 K7_NOP7 ASM_NOP1
++#ifdef ASM_SUPPORTED
++static inline unsigned long
++xen_get_eflag(void)
++{
++ register __u64 ia64_intri_res asm ("r8");
++ asm volatile ("break %1":
++ "=r"(ia64_intri_res): "i"(HYPERPRIVOP_GET_EFLAG));
++ return ia64_intri_res;
++}
+
-+#ifdef CONFIG_MK8
-+#define ASM_NOP1 K8_NOP1
-+#define ASM_NOP2 K8_NOP2
-+#define ASM_NOP3 K8_NOP3
-+#define ASM_NOP4 K8_NOP4
-+#define ASM_NOP5 K8_NOP5
-+#define ASM_NOP6 K8_NOP6
-+#define ASM_NOP7 K8_NOP7
-+#define ASM_NOP8 K8_NOP8
-+#elif defined(CONFIG_MK7)
-+#define ASM_NOP1 K7_NOP1
-+#define ASM_NOP2 K7_NOP2
-+#define ASM_NOP3 K7_NOP3
-+#define ASM_NOP4 K7_NOP4
-+#define ASM_NOP5 K7_NOP5
-+#define ASM_NOP6 K7_NOP6
-+#define ASM_NOP7 K7_NOP7
-+#define ASM_NOP8 K7_NOP8
++static inline void
++xen_set_eflag(unsigned long val)
++{
++ register __u64 __val asm ("r8") = val;
++ asm volatile ("break %0":: "i"(HYPERPRIVOP_SET_EFLAG), "r"(__val): "memory");
++}
+#else
-+#define ASM_NOP1 GENERIC_NOP1
-+#define ASM_NOP2 GENERIC_NOP2
-+#define ASM_NOP3 GENERIC_NOP3
-+#define ASM_NOP4 GENERIC_NOP4
-+#define ASM_NOP5 GENERIC_NOP5
-+#define ASM_NOP6 GENERIC_NOP6
-+#define ASM_NOP7 GENERIC_NOP7
-+#define ASM_NOP8 GENERIC_NOP8
++extern unsigned long xen_get_eflag(void); /* see xen_ia64_getreg */
++extern void xen_set_eflag(unsigned long); /* see xen_ia64_setreg */
+#endif
+
-+#define ASM_NOP_MAX 8
++/************************************************/
++/* Instructions paravirtualized for performance */
++/************************************************/
+
-+/* Prefetch instructions for Pentium III and AMD Athlon */
-+/* It's not worth to care about 3dnow! prefetches for the K6
-+ because they are microcoded there and very slow.
-+ However we don't do prefetches for pre XP Athlons currently
-+ That should be fixed. */
-+#define ARCH_HAS_PREFETCH
-+static inline void prefetch(const void *x)
-+{
-+ alternative_input(ASM_NOP4,
-+ "prefetchnta (%1)",
-+ X86_FEATURE_XMM,
-+ "r" (x));
++/* Xen uses memory-mapped virtual privileged registers for access to many
++ * performance-sensitive privileged registers. Some, like the processor
++ * status register (psr), are broken up into multiple memory locations.
++ * Others, like "pend", are abstractions based on privileged registers.
++ * "Pend" is guaranteed to be set if reading cr.ivr would return a
++ * (non-spurious) interrupt. */
++#define XEN_MAPPEDREGS ((struct mapped_regs *)XMAPPEDREGS_BASE)
++#define XSI_PSR_I \
++ (*XEN_MAPPEDREGS->interrupt_mask_addr)
++#define xen_get_virtual_psr_i() \
++ (!XSI_PSR_I)
++#define xen_set_virtual_psr_i(_val) \
++ ({ XSI_PSR_I = (uint8_t)(_val) ? 0 : 1; })
++#define xen_set_virtual_psr_ic(_val) \
++ ({ XEN_MAPPEDREGS->interrupt_collection_enabled = _val ? 1 : 0; })
++#define xen_get_virtual_pend() \
++ (*(((uint8_t *)XEN_MAPPEDREGS->interrupt_mask_addr) - 1))
++
++/* Hyperprivops are "break" instructions with a well-defined API.
++ * In particular, the virtual psr.ic bit must be off; in this way
++ * it is guaranteed to never conflict with a linux break instruction.
++ * Normally, this is done in a xen stub but this one is frequent enough
++ * that we inline it */
++#define xen_hyper_ssm_i() \
++({ \
++ XEN_HYPER_SSM_I; \
++})
++
++/* turning off interrupts can be paravirtualized simply by writing
++ * to a memory-mapped virtual psr.i bit (implemented as a 16-bit bool) */
++#define xen_rsm_i() \
++{ \
++ xen_set_virtual_psr_i(0); \
++ barrier(); \
+}
+
-+#define ARCH_HAS_PREFETCH
-+#define ARCH_HAS_PREFETCHW
-+#define ARCH_HAS_SPINLOCK_PREFETCH
++/* turning on interrupts is a bit more complicated.. write to the
++ * memory-mapped virtual psr.i bit first (to avoid race condition),
++ * then if any interrupts were pending, we have to execute a hyperprivop
++ * to ensure the pending interrupt gets delivered; else we're done! */
++#define xen_ssm_i() \
++({ \
++ int old = xen_get_virtual_psr_i(); \
++ xen_set_virtual_psr_i(1); \
++ barrier(); \
++ if (!old && xen_get_virtual_pend()) \
++ xen_hyper_ssm_i(); \
++})
+
-+/* 3dnow! prefetch to get an exclusive cache line. Useful for
-+ spinlocks to avoid one state transition in the cache coherency protocol. */
-+static inline void prefetchw(const void *x)
-+{
-+ alternative_input(ASM_NOP4,
-+ "prefetchw (%1)",
-+ X86_FEATURE_3DNOW,
-+ "r" (x));
++#define xen_ia64_intrin_local_irq_restore(x) \
++{ \
++ if (is_running_on_xen()) { \
++ if ((x) & IA64_PSR_I) { xen_ssm_i(); } \
++ else { xen_rsm_i(); } \
++ } \
++ else __ia64_intrin_local_irq_restore((x)); \
+}
-+#define spin_lock_prefetch(x) prefetchw(x)
+
-+extern void select_idle_routine(const struct cpuinfo_x86 *c);
++#define xen_get_psr_i() \
++( \
++ (is_running_on_xen()) ? \
++ (xen_get_virtual_psr_i() ? IA64_PSR_I : 0) \
++ : __ia64_get_psr_i() \
++)
+
-+#define cache_line_size() (boot_cpu_data.x86_cache_alignment)
++#define xen_ia64_ssm(mask) \
++{ \
++ if ((mask)==IA64_PSR_I) { \
++ if (is_running_on_xen()) { xen_ssm_i(); } \
++ else { __ia64_ssm(mask); } \
++ } \
++ else { __ia64_ssm(mask); } \
++}
+
-+extern unsigned long boot_option_idle_override;
-+extern void enable_sep_cpu(void);
-+extern int sysenter_setup(void);
++#define xen_ia64_rsm(mask) \
++{ \
++ if ((mask)==IA64_PSR_I) { \
++ if (is_running_on_xen()) { xen_rsm_i(); } \
++ else { __ia64_rsm(mask); } \
++ } \
++ else { __ia64_rsm(mask); } \
++}
+
-+#endif /* __ASM_I386_PROCESSOR_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/ptrace.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/ptrace.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/ptrace.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/ptrace.h 2007-07-30 16:35:12.000000000 +0200
-@@ -0,0 +1,90 @@
-+#ifndef _I386_PTRACE_H
-+#define _I386_PTRACE_H
+
-+#define EBX 0
-+#define ECX 1
-+#define EDX 2
-+#define ESI 3
-+#define EDI 4
-+#define EBP 5
-+#define EAX 6
-+#define DS 7
-+#define ES 8
-+#define FS 9
-+#define GS 10
-+#define ORIG_EAX 11
-+#define EIP 12
-+#define CS 13
-+#define EFL 14
-+#define UESP 15
-+#define SS 16
-+#define FRAME_SIZE 17
++/* Although all privileged operations can be left to trap and will
++ * be properly handled by Xen, some are frequent enough that we use
++ * hyperprivops for performance. */
+
-+/* this struct defines the way the registers are stored on the
-+ stack during a system call. */
++#ifndef ASM_SUPPORTED
++extern unsigned long xen_get_psr(void);
++extern unsigned long xen_get_ivr(void);
++extern unsigned long xen_get_tpr(void);
++extern void xen_set_itm(unsigned long);
++extern void xen_set_tpr(unsigned long);
++extern void xen_eoi(unsigned long);
++extern void xen_set_rr(unsigned long index, unsigned long val);
++extern unsigned long xen_get_rr(unsigned long index);
++extern void xen_set_kr(unsigned long index, unsigned long val);
++extern void xen_ptcga(unsigned long addr, unsigned long size);
++#else
++static inline unsigned long
++xen_get_psr(void)
++{
++ register __u64 ia64_intri_res asm ("r8");
++ asm volatile ("break %1":
++ "=r"(ia64_intri_res): "i"(HYPERPRIVOP_GET_PSR));
++ return ia64_intri_res;
++}
+
-+struct pt_regs {
-+ long ebx;
-+ long ecx;
-+ long edx;
-+ long esi;
-+ long edi;
-+ long ebp;
-+ long eax;
-+ int xds;
-+ int xes;
-+ long orig_eax;
-+ long eip;
-+ int xcs;
-+ long eflags;
-+ long esp;
-+ int xss;
-+};
++static inline unsigned long
++xen_get_ivr(void)
++{
++ register __u64 ia64_intri_res asm ("r8");
++ asm volatile ("break %1":
++ "=r"(ia64_intri_res): "i"(HYPERPRIVOP_GET_IVR));
++ return ia64_intri_res;
++}
+
-+/* Arbitrarily choose the same ptrace numbers as used by the Sparc code. */
-+#define PTRACE_GETREGS 12
-+#define PTRACE_SETREGS 13
-+#define PTRACE_GETFPREGS 14
-+#define PTRACE_SETFPREGS 15
-+#define PTRACE_GETFPXREGS 18
-+#define PTRACE_SETFPXREGS 19
++static inline unsigned long
++xen_get_tpr(void)
++{
++ register __u64 ia64_intri_res asm ("r8");
++ asm volatile ("break %1":
++ "=r"(ia64_intri_res): "i"(HYPERPRIVOP_GET_TPR));
++ return ia64_intri_res;
++}
+
-+#define PTRACE_OLDSETOPTIONS 21
++static inline void
++xen_set_tpr(unsigned long val)
++{
++ register __u64 __val asm ("r8") = val;
++ asm volatile ("break %0"::
++ "i"(HYPERPRIVOP_GET_TPR), "r"(__val): "memory");
++}
+
-+#define PTRACE_GET_THREAD_AREA 25
-+#define PTRACE_SET_THREAD_AREA 26
++static inline void
++xen_eoi(unsigned long val)
++{
++ register __u64 __val asm ("r8") = val;
++ asm volatile ("break %0"::
++ "i"(HYPERPRIVOP_EOI), "r"(__val): "memory");
++}
+
-+#define PTRACE_SYSEMU 31
-+#define PTRACE_SYSEMU_SINGLESTEP 32
++static inline void
++xen_set_itm(unsigned long val)
++{
++ register __u64 __val asm ("r8") = val;
++ asm volatile ("break %0":: "i"(HYPERPRIVOP_SET_ITM), "r"(__val): "memory");
++}
+
-+#ifdef __KERNEL__
++static inline void
++xen_ptcga(unsigned long addr, unsigned long size)
++{
++ register __u64 __addr asm ("r8") = addr;
++ register __u64 __size asm ("r9") = size;
++ asm volatile ("break %0"::
++ "i"(HYPERPRIVOP_PTC_GA), "r"(__addr), "r"(__size): "memory");
++}
+
-+#include <asm/vm86.h>
++static inline unsigned long
++xen_get_rr(unsigned long index)
++{
++ register __u64 ia64_intri_res asm ("r8");
++ register __u64 __index asm ("r8") = index;
++ asm volatile ("break %1":
++ "=r"(ia64_intri_res):
++ "i"(HYPERPRIVOP_GET_RR), "0"(__index));
++ return ia64_intri_res;
++}
+
-+struct task_struct;
-+extern void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs, int error_code);
++static inline void
++xen_set_rr(unsigned long index, unsigned long val)
++{
++ register __u64 __index asm ("r8") = index;
++ register __u64 __val asm ("r9") = val;
++ asm volatile ("break %0"::
++ "i"(HYPERPRIVOP_SET_RR), "r"(__index), "r"(__val): "memory");
++}
+
-+/*
-+ * user_mode_vm(regs) determines whether a register set came from user mode.
-+ * This is true if V8086 mode was enabled OR if the register set was from
-+ * protected mode with RPL-3 CS value. This tricky test checks that with
-+ * one comparison. Many places in the kernel can bypass this full check
-+ * if they have already ruled out V8086 mode, so user_mode(regs) can be used.
-+ */
-+static inline int user_mode(struct pt_regs *regs)
++static inline void
++xen_set_rr0_to_rr4(unsigned long val0, unsigned long val1,
++ unsigned long val2, unsigned long val3, unsigned long val4)
+{
-+ return (regs->xcs & 2) != 0;
++ register __u64 __val0 asm ("r8") = val0;
++ register __u64 __val1 asm ("r9") = val1;
++ register __u64 __val2 asm ("r10") = val2;
++ register __u64 __val3 asm ("r11") = val3;
++ register __u64 __val4 asm ("r14") = val4;
++ asm volatile ("break %0" ::
++ "i"(HYPERPRIVOP_SET_RR0_TO_RR4),
++ "r"(__val0), "r"(__val1),
++ "r"(__val2), "r"(__val3), "r"(__val4): "memory");
+}
-+static inline int user_mode_vm(struct pt_regs *regs)
++
++static inline void
++xen_set_kr(unsigned long index, unsigned long val)
+{
-+ return ((regs->xcs & 2) | (regs->eflags & VM_MASK)) != 0;
++ register __u64 __index asm ("r8") = index;
++ register __u64 __val asm ("r9") = val;
++ asm volatile ("break %0"::
++ "i"(HYPERPRIVOP_SET_KR), "r"(__index), "r"(__val): "memory");
+}
-+#define instruction_pointer(regs) ((regs)->eip)
-+#if defined(CONFIG_SMP) && defined(CONFIG_FRAME_POINTER)
-+extern unsigned long profile_pc(struct pt_regs *regs);
-+#else
-+#define profile_pc(regs) instruction_pointer(regs)
+#endif
-+#endif /* __KERNEL__ */
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/scatterlist.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/scatterlist.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/scatterlist.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/scatterlist.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,22 @@
-+#ifndef _I386_SCATTERLIST_H
-+#define _I386_SCATTERLIST_H
++/* Note: It may look wrong to test for is_running_on_xen() in each case.
++ * However regnum is always a constant so, as written, the compiler
++ * eliminates the switch statement, whereas is_running_on_xen() must be
++ * tested dynamically. */
++#define xen_ia64_getreg(regnum) \
++({ \
++ __u64 ia64_intri_res; \
++ \
++ switch(regnum) { \
++ case _IA64_REG_PSR: \
++ ia64_intri_res = (is_running_on_xen()) ? \
++ xen_get_psr() : \
++ __ia64_getreg(regnum); \
++ break; \
++ case _IA64_REG_CR_IVR: \
++ ia64_intri_res = (is_running_on_xen()) ? \
++ xen_get_ivr() : \
++ __ia64_getreg(regnum); \
++ break; \
++ case _IA64_REG_CR_TPR: \
++ ia64_intri_res = (is_running_on_xen()) ? \
++ xen_get_tpr() : \
++ __ia64_getreg(regnum); \
++ break; \
++ case _IA64_REG_AR_EFLAG: \
++ ia64_intri_res = (is_running_on_xen()) ? \
++ xen_get_eflag() : \
++ __ia64_getreg(regnum); \
++ break; \
++ default: \
++ ia64_intri_res = __ia64_getreg(regnum); \
++ break; \
++ } \
++ ia64_intri_res; \
++})
+
-+struct scatterlist {
-+ struct page *page;
-+ unsigned int offset;
-+ unsigned int length;
-+ dma_addr_t dma_address;
-+ unsigned int dma_length;
-+};
++#define xen_ia64_setreg(regnum,val) \
++({ \
++ switch(regnum) { \
++ case _IA64_REG_AR_KR0 ... _IA64_REG_AR_KR7: \
++ (is_running_on_xen()) ? \
++ xen_set_kr((regnum-_IA64_REG_AR_KR0), val) : \
++ __ia64_setreg(regnum,val); \
++ break; \
++ case _IA64_REG_CR_ITM: \
++ (is_running_on_xen()) ? \
++ xen_set_itm(val) : \
++ __ia64_setreg(regnum,val); \
++ break; \
++ case _IA64_REG_CR_TPR: \
++ (is_running_on_xen()) ? \
++ xen_set_tpr(val) : \
++ __ia64_setreg(regnum,val); \
++ break; \
++ case _IA64_REG_CR_EOI: \
++ (is_running_on_xen()) ? \
++ xen_eoi(val) : \
++ __ia64_setreg(regnum,val); \
++ break; \
++ case _IA64_REG_AR_EFLAG: \
++ (is_running_on_xen()) ? \
++ xen_set_eflag(val) : \
++ __ia64_setreg(regnum,val); \
++ break; \
++ default: \
++ __ia64_setreg(regnum,val); \
++ break; \
++ } \
++})
+
-+/* These macros should be used after a pci_map_sg call has been done
-+ * to get bus addresses of each of the SG entries and their lengths.
-+ * You should only work with the number of sg entries pci_map_sg
-+ * returns.
-+ */
-+#define sg_dma_address(sg) ((sg)->dma_address)
-+#define sg_dma_len(sg) ((sg)->dma_length)
++#define ia64_ptcga(addr, size) \
++do { \
++ if (is_running_on_xen()) \
++ xen_ptcga((addr), (size)); \
++ else \
++ __ia64_ptcga((addr), (size)); \
++} while (0)
+
-+#define ISA_DMA_THRESHOLD (0x00ffffff)
++#define ia64_set_rr(index, val) \
++do { \
++ if (is_running_on_xen()) \
++ xen_set_rr((index), (val)); \
++ else \
++ __ia64_set_rr((index), (val)); \
++} while (0)
+
-+#endif /* !(_I386_SCATTERLIST_H) */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/segment.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/segment.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/segment.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/segment.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,117 @@
-+#ifndef _ASM_SEGMENT_H
-+#define _ASM_SEGMENT_H
++#define ia64_get_rr(index) \
++({ \
++ __u64 ia64_intri_res; \
++ if (is_running_on_xen()) \
++ ia64_intri_res = xen_get_rr((index)); \
++ else \
++ ia64_intri_res = __ia64_get_rr((index)); \
++ ia64_intri_res; \
++})
+
-+/*
-+ * The layout of the per-CPU GDT under Linux:
-+ *
-+ * 0 - null
-+ * 1 - reserved
-+ * 2 - reserved
-+ * 3 - reserved
-+ *
-+ * 4 - unused <==== new cacheline
-+ * 5 - unused
-+ *
-+ * ------- start of TLS (Thread-Local Storage) segments:
-+ *
-+ * 6 - TLS segment #1 [ glibc's TLS segment ]
-+ * 7 - TLS segment #2 [ Wine's %fs Win32 segment ]
-+ * 8 - TLS segment #3
-+ * 9 - reserved
-+ * 10 - reserved
-+ * 11 - reserved
-+ *
-+ * ------- start of kernel segments:
-+ *
-+ * 12 - kernel code segment <==== new cacheline
-+ * 13 - kernel data segment
-+ * 14 - default user CS
-+ * 15 - default user DS
-+ * 16 - TSS
-+ * 17 - LDT
-+ * 18 - PNPBIOS support (16->32 gate)
-+ * 19 - PNPBIOS support
-+ * 20 - PNPBIOS support
-+ * 21 - PNPBIOS support
-+ * 22 - PNPBIOS support
-+ * 23 - APM BIOS support
-+ * 24 - APM BIOS support
-+ * 25 - APM BIOS support
-+ *
-+ * 26 - ESPFIX small SS
-+ * 27 - unused
-+ * 28 - unused
-+ * 29 - unused
-+ * 30 - unused
-+ * 31 - TSS for double fault handler
-+ */
-+#define GDT_ENTRY_TLS_ENTRIES 3
-+#define GDT_ENTRY_TLS_MIN 6
-+#define GDT_ENTRY_TLS_MAX (GDT_ENTRY_TLS_MIN + GDT_ENTRY_TLS_ENTRIES - 1)
++#define ia64_set_rr0_to_rr4(val0, val1, val2, val3, val4) \
++do { \
++ if (is_running_on_xen()) \
++ xen_set_rr0_to_rr4((val0), (val1), (val2), \
++ (val3), (val4)); \
++ else \
++ __ia64_set_rr0_to_rr4((val0), (val1), (val2), \
++ (val3), (val4)); \
++} while (0)
+
-+#define TLS_SIZE (GDT_ENTRY_TLS_ENTRIES * 8)
++#define ia64_getreg xen_ia64_getreg
++#define ia64_setreg xen_ia64_setreg
++#define ia64_ssm xen_ia64_ssm
++#define ia64_rsm xen_ia64_rsm
++#define ia64_intrin_local_irq_restore xen_ia64_intrin_local_irq_restore
++#define ia64_get_psr_i xen_get_psr_i
+
-+#define GDT_ENTRY_DEFAULT_USER_CS 14
-+#define __USER_CS (GDT_ENTRY_DEFAULT_USER_CS * 8 + 3)
++/* the remainder of these are not performance-sensitive so its
++ * OK to not paravirtualize and just take a privop trap and emulate */
++#define ia64_hint __ia64_hint
++#define ia64_set_pmd __ia64_set_pmd
++#define ia64_itci __ia64_itci
++#define ia64_itcd __ia64_itcd
++#define ia64_itri __ia64_itri
++#define ia64_itrd __ia64_itrd
++#define ia64_tpa __ia64_tpa
++#define ia64_set_ibr __ia64_set_ibr
++#define ia64_set_pkr __ia64_set_pkr
++#define ia64_set_pmc __ia64_set_pmc
++#define ia64_get_ibr __ia64_get_ibr
++#define ia64_get_pkr __ia64_get_pkr
++#define ia64_get_pmc __ia64_get_pmc
++#define ia64_ptce __ia64_ptce
++#define ia64_ptcl __ia64_ptcl
++#define ia64_ptri __ia64_ptri
++#define ia64_ptrd __ia64_ptrd
++
++#endif /* !__ASSEMBLY__ */
++
++/* these routines utilize privilege-sensitive or performance-sensitive
++ * privileged instructions so the code must be replaced with
++ * paravirtualized versions */
++#define ia64_leave_kernel xen_leave_kernel
++#define ia64_leave_syscall xen_leave_syscall
++#define ia64_trace_syscall xen_trace_syscall
++#define ia64_ret_from_clone xen_ret_from_clone
++#define ia64_switch_to xen_switch_to
++#define ia64_pal_call_static xen_pal_call_static
++
++#endif /* _ASM_IA64_XEN_PRIVOP_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/xen/xcom_hcall.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/xen/xcom_hcall.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,67 @@
++/*
++ * Copyright (C) 2006 Tristan Gingold <tristan.gingold at bull.net>, Bull SAS
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ */
+
-+#define GDT_ENTRY_DEFAULT_USER_DS 15
-+#define __USER_DS (GDT_ENTRY_DEFAULT_USER_DS * 8 + 3)
++#ifndef _LINUX_XENCOMM_HCALL_H_
++#define _LINUX_XENCOMM_HCALL_H_
+
-+#define GDT_ENTRY_KERNEL_BASE 12
++/* These function creates inline or mini descriptor for the parameters and
++ calls the corresponding xencomm_arch_hypercall_X.
++ Architectures should defines HYPERVISOR_xxx as xencomm_hypercall_xxx unless
++ they want to use their own wrapper. */
++extern int xencomm_hypercall_console_io(int cmd, int count, char *str);
+
-+#define GDT_ENTRY_KERNEL_CS (GDT_ENTRY_KERNEL_BASE + 0)
-+#define __KERNEL_CS (GDT_ENTRY_KERNEL_CS * 8)
-+#define GET_KERNEL_CS() (__KERNEL_CS | (xen_feature(XENFEAT_supervisor_mode_kernel)?0:1) )
++extern int xencomm_hypercall_event_channel_op(int cmd, void *op);
+
-+#define GDT_ENTRY_KERNEL_DS (GDT_ENTRY_KERNEL_BASE + 1)
-+#define __KERNEL_DS (GDT_ENTRY_KERNEL_DS * 8)
-+#define GET_KERNEL_DS() (__KERNEL_DS | (xen_feature(XENFEAT_supervisor_mode_kernel)?0:1) )
++extern int xencomm_hypercall_xen_version(int cmd, void *arg);
+
-+#define GDT_ENTRY_TSS (GDT_ENTRY_KERNEL_BASE + 4)
-+#define GDT_ENTRY_LDT (GDT_ENTRY_KERNEL_BASE + 5)
++extern int xencomm_hypercall_physdev_op(int cmd, void *op);
+
-+#define GDT_ENTRY_PNPBIOS_BASE (GDT_ENTRY_KERNEL_BASE + 6)
-+#define GDT_ENTRY_APMBIOS_BASE (GDT_ENTRY_KERNEL_BASE + 11)
++extern int xencomm_hypercall_grant_table_op(unsigned int cmd, void *op,
++ unsigned int count);
+
-+#define GDT_ENTRY_ESPFIX_SS (GDT_ENTRY_KERNEL_BASE + 14)
-+#define __ESPFIX_SS (GDT_ENTRY_ESPFIX_SS * 8)
++extern int xencomm_hypercall_sched_op(int cmd, void *arg);
+
-+#define GDT_ENTRY_DOUBLEFAULT_TSS 31
++extern int xencomm_hypercall_multicall(void *call_list, int nr_calls);
+
-+/*
-+ * The GDT has 32 entries
-+ */
-+#define GDT_ENTRIES 32
++extern int xencomm_hypercall_callback_op(int cmd, void *arg);
+
-+#define GDT_SIZE (GDT_ENTRIES * 8)
++extern int xencomm_hypercall_memory_op(unsigned int cmd, void *arg);
+
-+/* Simple and small GDT entries for booting only */
++extern unsigned long xencomm_hypercall_hvm_op(int cmd, void *arg);
+
-+#define GDT_ENTRY_BOOT_CS 2
-+#define __BOOT_CS (GDT_ENTRY_BOOT_CS * 8)
++extern int xencomm_hypercall_suspend(unsigned long srec);
+
-+#define GDT_ENTRY_BOOT_DS (GDT_ENTRY_BOOT_CS + 1)
-+#define __BOOT_DS (GDT_ENTRY_BOOT_DS * 8)
++extern int xencomm_hypercall_xenoprof_op(int op, void *arg);
+
-+/* The PnP BIOS entries in the GDT */
-+#define GDT_ENTRY_PNPBIOS_CS32 (GDT_ENTRY_PNPBIOS_BASE + 0)
-+#define GDT_ENTRY_PNPBIOS_CS16 (GDT_ENTRY_PNPBIOS_BASE + 1)
-+#define GDT_ENTRY_PNPBIOS_DS (GDT_ENTRY_PNPBIOS_BASE + 2)
-+#define GDT_ENTRY_PNPBIOS_TS1 (GDT_ENTRY_PNPBIOS_BASE + 3)
-+#define GDT_ENTRY_PNPBIOS_TS2 (GDT_ENTRY_PNPBIOS_BASE + 4)
++extern int xencomm_hypercall_perfmon_op(unsigned long cmd, void* arg,
++ unsigned long count);
+
-+/* The PnP BIOS selectors */
-+#define PNP_CS32 (GDT_ENTRY_PNPBIOS_CS32 * 8) /* segment for calling fn */
-+#define PNP_CS16 (GDT_ENTRY_PNPBIOS_CS16 * 8) /* code segment for BIOS */
-+#define PNP_DS (GDT_ENTRY_PNPBIOS_DS * 8) /* data segment for BIOS */
-+#define PNP_TS1 (GDT_ENTRY_PNPBIOS_TS1 * 8) /* transfer data segment */
-+#define PNP_TS2 (GDT_ENTRY_PNPBIOS_TS2 * 8) /* another data segment */
++extern long xencomm_hypercall_vcpu_op(int cmd, int cpu, void *arg);
+
-+/*
-+ * The interrupt descriptor table has room for 256 idt's,
-+ * the global descriptor table is dependent on the number
-+ * of tasks we can have..
-+ */
-+#define IDT_ENTRIES 256
++extern long xencomm_hypercall_opt_feature(void *arg);
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/setup.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/setup.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/setup.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/setup.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,81 @@
-+/*
-+ * Just a place holder. We don't want to have to test x86 before
-+ * we include stuff
-+ */
++/* For privcmd. Locally declare argument type to avoid include storm.
++ Type coherency will be checked within privcmd.c */
++struct privcmd_hypercall;
++extern int privcmd_hypercall(struct privcmd_hypercall *hypercall);
+
-+#ifndef _i386_SETUP_H
-+#define _i386_SETUP_H
++extern int xen_foreign_p2m_expose(struct privcmd_hypercall *hypercall);
+
-+#ifdef __KERNEL__
-+#include <linux/pfn.h>
++extern int xencomm_hypercall_kexec_op(int cmd, void *arg);
+
++#endif /* _LINUX_XENCOMM_HCALL_H_ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/xen/xencomm.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/xen/xencomm.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,33 @@
+/*
-+ * Reserved space for vmalloc and iomap - defined in asm/page.h
++ * Copyright (C) 2006 Hollis Blanchard <hollisb at us.ibm.com>, IBM Corporation
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ */
-+#define MAXMEM_PFN PFN_DOWN(MAXMEM)
-+#define MAX_NONPAE_PFN (1 << 20)
-+#endif
+
-+#define PARAM_SIZE 4096
-+#define COMMAND_LINE_SIZE 256
++#ifndef _ASM_IA64_XENCOMM_H_
++#define _ASM_IA64_XENCOMM_H_
+
-+#define OLD_CL_MAGIC_ADDR 0x90020
-+#define OLD_CL_MAGIC 0xA33F
-+#define OLD_CL_BASE_ADDR 0x90000
-+#define OLD_CL_OFFSET 0x90022
-+#define NEW_CL_POINTER 0x228 /* Relative to real mode data */
++#define is_kernel_addr(x) \
++ ((PAGE_OFFSET <= (x) && \
++ (x) < (PAGE_OFFSET + (1UL << IA64_MAX_PHYS_BITS))) || \
++ (KERNEL_START <= (x) && \
++ (x) < KERNEL_START + KERNEL_TR_PAGE_SIZE))
+
-+#ifndef __ASSEMBLY__
-+/*
-+ * This is set up by the setup-routine at boot-time
-+ */
-+extern unsigned char boot_params[PARAM_SIZE];
++/* Must be called before any hypercall. */
++extern void xencomm_initialize (void);
+
-+#define PARAM (boot_params)
-+#define SCREEN_INFO (*(struct screen_info *) (PARAM+0))
-+#define EXT_MEM_K (*(unsigned short *) (PARAM+2))
-+#define ALT_MEM_K (*(unsigned long *) (PARAM+0x1e0))
-+#define E820_MAP_NR (*(char*) (PARAM+E820NR))
-+#define E820_MAP ((struct e820entry *) (PARAM+E820MAP))
-+#define APM_BIOS_INFO (*(struct apm_bios_info *) (PARAM+0x40))
-+#define IST_INFO (*(struct ist_info *) (PARAM+0x60))
-+#define DRIVE_INFO (*(struct drive_info_struct *) (PARAM+0x80))
-+#define SYS_DESC_TABLE (*(struct sys_desc_table_struct*)(PARAM+0xa0))
-+#define EFI_SYSTAB ((efi_system_table_t *) *((unsigned long *)(PARAM+0x1c4)))
-+#define EFI_MEMDESC_SIZE (*((unsigned long *) (PARAM+0x1c8)))
-+#define EFI_MEMDESC_VERSION (*((unsigned long *) (PARAM+0x1cc)))
-+#define EFI_MEMMAP ((void *) *((unsigned long *)(PARAM+0x1d0)))
-+#define EFI_MEMMAP_SIZE (*((unsigned long *) (PARAM+0x1d4)))
-+#define MOUNT_ROOT_RDONLY (*(unsigned short *) (PARAM+0x1F2))
-+#define RAMDISK_FLAGS (*(unsigned short *) (PARAM+0x1F8))
-+#define VIDEO_MODE (*(unsigned short *) (PARAM+0x1FA))
-+#define ORIG_ROOT_DEV (*(unsigned short *) (PARAM+0x1FC))
-+#define AUX_DEVICE_INFO (*(unsigned char *) (PARAM+0x1FF))
-+#define LOADER_TYPE (*(unsigned char *) (PARAM+0x210))
-+#define KERNEL_START (*(unsigned long *) (PARAM+0x214))
-+#define INITRD_START (__pa(xen_start_info->mod_start))
-+#define INITRD_SIZE (xen_start_info->mod_len)
-+#define EDID_INFO (*(struct edid_info *) (PARAM+0x440))
-+#define EDD_NR (*(unsigned char *) (PARAM+EDDNR))
-+#define EDD_MBR_SIG_NR (*(unsigned char *) (PARAM+EDD_MBR_SIG_NR_BUF))
-+#define EDD_MBR_SIGNATURE ((unsigned int *) (PARAM+EDD_MBR_SIG_BUF))
-+#define EDD_BUF ((struct edd_info *) (PARAM+EDDBUF))
++#include <xen/xencomm.h>
+
-+/*
-+ * Do NOT EVER look at the BIOS memory size location.
-+ * It does not work on many machines.
++#endif /* _ASM_IA64_XENCOMM_H_ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-ia64/xenoprof.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-ia64/xenoprof.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,48 @@
++/******************************************************************************
++ * asm-ia64/xenoprof.h
++ *
++ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
+ */
-+#define LOWMEMSIZE() (0x9f000)
-+
-+struct e820entry;
++#ifndef __ASM_XENOPROF_H__
++#define __ASM_XENOPROF_H__
++#ifdef CONFIG_XEN
+
-+char * __init machine_specific_memory_setup(void);
++#undef HAVE_XENOPROF_CREATE_FILES
+
-+int __init copy_e820_map(struct e820entry * biosmap, int nr_map);
-+int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map);
-+void __init add_memory_region(unsigned long long start,
-+ unsigned long long size, int type);
++struct xenoprof_init;
++void xenoprof_arch_init_counter(struct xenoprof_init *init);
++void xenoprof_arch_counter(void);
++void xenoprof_arch_start(void);
++void xenoprof_arch_stop(void);
+
-+#endif /* __ASSEMBLY__ */
++struct xenoprof_arch_shared_buffer {
++ struct resource* res;
++};
+
-+#endif /* _i386_SETUP_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/smp.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/smp.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/smp.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/smp.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,103 @@
-+#ifndef __ASM_SMP_H
-+#define __ASM_SMP_H
++struct xenoprof_shared_buffer;
++void xenoprof_arch_unmap_shared_buffer(struct xenoprof_shared_buffer *sbuf);
++struct xenoprof_get_buffer;
++int xenoprof_arch_map_shared_buffer(struct xenoprof_get_buffer *get_buffer,
++ struct xenoprof_shared_buffer *sbuf);
++struct xenoprof_passive;
++int xenoprof_arch_set_passive(struct xenoprof_passive *pdomain,
++ struct xenoprof_shared_buffer *sbuf);
+
++#endif /* CONFIG_XEN */
++#endif /* __ASM_XENOPROF_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-powerpc/mpic.h
+--- a/include/asm-powerpc/mpic.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-powerpc/mpic.h Wed Sep 10 10:54:08 2008 +0100
+@@ -305,6 +305,8 @@
+ #define MPIC_SPV_EOI 0x00000020
+ /* No passthrough disable */
+ #define MPIC_NO_PTHROU_DIS 0x00000040
++/* Skip reset of IPI vectors during init */
++#define MPIC_SKIP_IPI_INIT 0x00000080
+
+ /* MPIC HW modification ID */
+ #define MPIC_REGSET_MASK 0xf0000000
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-powerpc/udbg.h
+--- a/include/asm-powerpc/udbg.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-powerpc/udbg.h Wed Sep 10 10:54:08 2008 +0100
+@@ -42,6 +42,7 @@
+ extern void __init udbg_init_pmac_realmode(void);
+ extern void __init udbg_init_maple_realmode(void);
+ extern void __init udbg_init_iseries(void);
++extern void __init udbg_init_xen(void);
+ extern void __init udbg_init_rtas_panel(void);
+ extern void __init udbg_init_rtas_console(void);
+
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-powerpc/xen/asm/gnttab_dma.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-powerpc/xen/asm/gnttab_dma.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,29 @@
+/*
-+ * We need the APIC definitions automatically as part of 'smp.h'
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright 2007 IBM Corp.
++ *
++ * Authors: Hollis Blanchard <hollisb at us.ibm.com>
+ */
-+#ifndef __ASSEMBLY__
-+#include <linux/kernel.h>
-+#include <linux/threads.h>
-+#include <linux/cpumask.h>
-+#endif
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+#ifndef __ASSEMBLY__
-+#include <asm/fixmap.h>
-+#include <asm/bitops.h>
-+#include <asm/mpspec.h>
-+#ifdef CONFIG_X86_IO_APIC
-+#include <asm/io_apic.h>
-+#endif
-+#include <asm/apic.h>
-+#endif
-+#endif
++#ifndef _ASM_PPC_GNTTAB_DMA_H
++#define _ASM_PPC_GNTTAB_DMA_H
+
-+#define BAD_APICID 0xFFu
-+#ifdef CONFIG_SMP
-+#ifndef __ASSEMBLY__
++static inline int gnttab_dma_local_pfn(struct page *page)
++{
++ return 0;
++}
+
-+/*
-+ * Private routines/data
++#endif /* _ASM_PPC_GNTTAB_DMA_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-powerpc/xen/asm/hypercall.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-powerpc/xen/asm/hypercall.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,90 @@
++/******************************************************************************
++ * hypercall.h
++ *
++ * Linux-specific hypervisor handling.
++ *
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ * This file may be distributed separately from the Linux kernel, or
++ * incorporated into other software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ *
++ * Copyright 2007 IBM Corp.
++ *
++ * Authors: Hollis Blanchard <hollisb at us.ibm.com>
++ * Jimi Xenidis <jimix at watson.ibm.com>
+ */
-+
-+extern void smp_alloc_memory(void);
-+extern int pic_mode;
-+extern int smp_num_siblings;
-+extern cpumask_t cpu_sibling_map[];
-+extern cpumask_t cpu_core_map[];
+
-+extern void (*mtrr_hook) (void);
-+extern void zap_low_mappings (void);
-+extern void lock_ipi_call_lock(void);
-+extern void unlock_ipi_call_lock(void);
++#ifndef __HYPERCALL_H__
++#define __HYPERCALL_H__
+
-+#define MAX_APICID 256
-+extern u8 x86_cpu_to_apicid[];
++#include <asm/hvcall.h>
++#include <asm/page.h>
++#include <xen/xencomm.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/sched.h>
+
-+#define cpu_physical_id(cpu) x86_cpu_to_apicid[cpu]
++#define XEN_MARK(a)((a) | (~0UL << 16))
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+extern void cpu_exit_clear(void);
-+extern void cpu_uninit(void);
-+#endif
++extern int HYPERVISOR_console_io(int cmd, int count, char *str);
++extern int HYPERVISOR_event_channel_op(int cmd, void *op);
++extern int HYPERVISOR_xen_version(int cmd, void *arg);
++extern int HYPERVISOR_physdev_op(int cmd, void *op);
++extern int HYPERVISOR_grant_table_op(unsigned int cmd, void *uop,
++ unsigned int count);
++extern int HYPERVISOR_vcpu_op(int cmd, int vcpuid, void *extra_args);
++extern int HYPERVISOR_memory_op(unsigned int cmd, void *arg);
++extern int HYPERVISOR_multicall(void *call_list, int nr_calls);
++
++extern int HYPERVISOR_sched_op(int cmd, void *arg);
++extern int HYPERVISOR_poll(
++ evtchn_port_t *ports, unsigned int nr_ports, u64 timeout);
+
-+/*
-+ * This function is needed by all SMP systems. It must _always_ be valid
-+ * from the initial startup. We map APIC_BASE very early in page_setup(),
-+ * so this is correct in the x86 case.
-+ */
-+#define raw_smp_processor_id() (current_thread_info()->cpu)
++static inline int HYPERVISOR_shutdown(unsigned int reason)
++{
++ struct sched_shutdown sched_shutdown = {
++ .reason = reason
++ };
+
-+extern cpumask_t cpu_possible_map;
-+#define cpu_callin_map cpu_possible_map
++ return HYPERVISOR_sched_op(SCHEDOP_shutdown, &sched_shutdown);
++}
+
-+/* We don't mark CPUs online until __cpu_up(), so we need another measure */
-+static inline int num_booting_cpus(void)
++static inline int HYPERVISOR_set_timer_op(unsigned long arg)
+{
-+ return cpus_weight(cpu_possible_map);
++ return plpar_hcall_norets(XEN_MARK(__HYPERVISOR_set_timer_op), arg);
+}
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+
-+#ifdef APIC_DEFINITION
-+extern int hard_smp_processor_id(void);
-+#else
-+#include <mach_apicdef.h>
-+static inline int hard_smp_processor_id(void)
-+{
-+ /* we don't want to mark this access volatile - bad code generation */
-+ return GET_APIC_ID(*(unsigned long *)(APIC_BASE+APIC_ID));
++extern int HYPERVISOR_suspend(unsigned long srec);
++extern int HYPERVISOR_kexec_op(unsigned long op, void *args);
++static inline unsigned long HYPERVISOR_hvm_op(int op, void *arg) {
++ return -ENOSYS;
+}
-+#endif
+
-+static __inline int logical_smp_processor_id(void)
++static inline int
++HYPERVISOR_mmu_update(
++ mmu_update_t *req, int count, int *success_count, domid_t domid)
+{
-+ /* we don't want to mark this access volatile - bad code generation */
-+ return GET_APIC_LOGICAL_ID(*(unsigned long *)(APIC_BASE+APIC_LDR));
++ return -ENOSYS;
+}
+
-+#endif
++struct privcmd_hypercall;
++extern int privcmd_hypercall(struct privcmd_hypercall *hypercall);
+
-+extern int __cpu_disable(void);
-+extern void __cpu_die(unsigned int cpu);
-+extern void prefill_possible_map(void);
-+#endif /* !__ASSEMBLY__ */
++#endif /* __HYPERCALL_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-powerpc/xen/asm/hypervisor.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-powerpc/xen/asm/hypervisor.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,276 @@
++/******************************************************************************
++ * hypervisor.h
++ *
++ * Linux-specific hypervisor handling.
++ *
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ * This file may be distributed separately from the Linux kernel, or
++ * incorporated into other software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+#else /* CONFIG_SMP */
++#ifndef __HYPERVISOR_H__
++#define __HYPERVISOR_H__
+
-+#define cpu_physical_id(cpu) boot_cpu_physical_apicid
++#include <linux/config.h>
++#include <linux/types.h>
++#include <linux/kernel.h>
++#include <linux/version.h>
++#include <xen/interface/xen.h>
++#include <asm/ptrace.h>
++#include <asm/page.h>
++#include <asm/irq.h>
+
-+#define NO_PROC_ID 0xFF /* No processor magic marker */
++extern shared_info_t *HYPERVISOR_shared_info;
+
++/* arch/xen/i386/kernel/setup.c */
++extern start_info_t *xen_start_info;
++
++#ifdef CONFIG_XEN_PRIVILEGED_GUEST
++#define is_initial_xendomain() (xen_start_info && \
++ (xen_start_info->flags & SIF_INITDOMAIN))
++#else
++#define is_initial_xendomain() 0
+#endif
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/spinlock.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/spinlock.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/spinlock.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/spinlock.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,202 @@
-+#ifndef __ASM_SPINLOCK_H
-+#define __ASM_SPINLOCK_H
+
-+#include <asm/atomic.h>
-+#include <asm/rwlock.h>
-+#include <asm/page.h>
-+#include <linux/compiler.h>
++/* arch/xen/kernel/evtchn.c */
++/* Force a proper event-channel callback from Xen. */
++void force_evtchn_callback(void);
+
-+/*
-+ * Your basic SMP spinlocks, allowing only a single CPU anywhere
-+ *
-+ * Simple spin lock operations. There are two variants, one clears IRQ's
-+ * on the local processor, one does not.
-+ *
-+ * We make no fairness assumptions. They have a cost.
-+ *
-+ * (the type definitions are in asm/spinlock_types.h)
-+ */
++/* arch/xen/kernel/process.c */
++void xen_cpu_idle (void);
+
-+#define __raw_spin_is_locked(x) \
-+ (*(volatile signed char *)(&(x)->slock) <= 0)
++/* arch/xen/i386/kernel/hypervisor.c */
++void do_hypervisor_callback(struct pt_regs *regs);
+
-+#define __raw_spin_lock_string \
-+ "\n1:\t" \
-+ LOCK_PREFIX " ; decb %0\n\t" \
-+ "jns 3f\n" \
-+ "2:\t" \
-+ "rep;nop\n\t" \
-+ "cmpb $0,%0\n\t" \
-+ "jle 2b\n\t" \
-+ "jmp 1b\n" \
-+ "3:\n\t"
++/* arch/xen/i386/kernel/head.S */
++void lgdt_finish(void);
+
++/* arch/xen/i386/mm/hypervisor.c */
+/*
-+ * NOTE: there's an irqs-on section here, which normally would have to be
-+ * irq-traced, but on CONFIG_TRACE_IRQFLAGS we never use
-+ * __raw_spin_lock_string_flags().
++ * NB. ptr values should be PHYSICAL, not MACHINE. 'vals' should be already
++ * be MACHINE addresses.
+ */
-+#define __raw_spin_lock_string_flags \
-+ "\n1:\t" \
-+ LOCK_PREFIX " ; decb %0\n\t" \
-+ "jns 5f\n" \
-+ "2:\t" \
-+ "testl $0x200, %1\n\t" \
-+ "jz 4f\n\t" \
-+ "#sti\n" \
-+ "3:\t" \
-+ "rep;nop\n\t" \
-+ "cmpb $0, %0\n\t" \
-+ "jle 3b\n\t" \
-+ "#cli\n\t" \
-+ "jmp 1b\n" \
-+ "4:\t" \
-+ "rep;nop\n\t" \
-+ "cmpb $0, %0\n\t" \
-+ "jg 1b\n\t" \
-+ "jmp 4b\n" \
-+ "5:\n\t"
+
-+static inline void __raw_spin_lock(raw_spinlock_t *lock)
-+{
-+ asm(__raw_spin_lock_string : "+m" (lock->slock) : : "memory");
-+}
++void xen_pt_switch(unsigned long ptr);
++void xen_new_user_pt(unsigned long ptr); /* x86_64 only */
++void xen_load_gs(unsigned int selector); /* x86_64 only */
++void xen_tlb_flush(void);
++void xen_invlpg(unsigned long ptr);
+
-+/*
-+ * It is easier for the lock validator if interrupts are not re-enabled
-+ * in the middle of a lock-acquire. This is a performance feature anyway
-+ * so we turn it off:
-+ */
-+#ifndef CONFIG_PROVE_LOCKING
-+static inline void __raw_spin_lock_flags(raw_spinlock_t *lock, unsigned long flags)
-+{
-+ asm(__raw_spin_lock_string_flags : "+m" (lock->slock) : "r" (flags) : "memory");
-+}
++#ifndef CONFIG_XEN_SHADOW_MODE
++void xen_l1_entry_update(pte_t *ptr, pte_t val);
++void xen_l2_entry_update(pmd_t *ptr, pmd_t val);
++void xen_l3_entry_update(pud_t *ptr, pud_t val); /* x86_64/PAE */
++void xen_l4_entry_update(pgd_t *ptr, pgd_t val); /* x86_64 only */
++void xen_pgd_pin(unsigned long ptr);
++void xen_pgd_unpin(unsigned long ptr);
++void xen_pud_pin(unsigned long ptr); /* x86_64 only */
++void xen_pud_unpin(unsigned long ptr); /* x86_64 only */
++void xen_pmd_pin(unsigned long ptr); /* x86_64 only */
++void xen_pmd_unpin(unsigned long ptr); /* x86_64 only */
++void xen_pte_pin(unsigned long ptr);
++void xen_pte_unpin(unsigned long ptr);
++#else
++#define xen_l1_entry_update(_p, _v) set_pte((_p), (_v))
++#define xen_l2_entry_update(_p, _v) set_pgd((_p), (_v))
++#define xen_pgd_pin(_p) ((void)0)
++#define xen_pgd_unpin(_p) ((void)0)
++#define xen_pte_pin(_p) ((void)0)
++#define xen_pte_unpin(_p) ((void)0)
+#endif
+
-+static inline int __raw_spin_trylock(raw_spinlock_t *lock)
-+{
-+ char oldval;
-+ __asm__ __volatile__(
-+ "xchgb %b0,%1"
-+ :"=q" (oldval), "+m" (lock->slock)
-+ :"0" (0) : "memory");
-+ return oldval > 0;
-+}
++void xen_set_ldt(unsigned long ptr, unsigned long bytes);
++void xen_machphys_update(unsigned long mfn, unsigned long pfn);
+
-+/*
-+ * __raw_spin_unlock based on writing $1 to the low byte.
-+ * This method works. Despite all the confusion.
-+ * (except on PPro SMP or if we are using OOSTORE, so we use xchgb there)
-+ * (PPro errata 66, 92)
-+ */
++#ifdef CONFIG_SMP
++#include <linux/cpumask.h>
++void xen_tlb_flush_all(void);
++void xen_invlpg_all(unsigned long ptr);
++void xen_tlb_flush_mask(cpumask_t *mask);
++void xen_invlpg_mask(cpumask_t *mask, unsigned long ptr);
++#endif
+
-+#if !defined(CONFIG_X86_OOSTORE) && !defined(CONFIG_X86_PPRO_FENCE)
++/* Returns zero on success else negative errno. */
++static inline int xen_create_contiguous_region(
++ unsigned long vstart, unsigned int order, unsigned int address_bits)
++{
++ return 0;
++}
++static inline void xen_destroy_contiguous_region(
++ unsigned long vstart, unsigned int order)
++{
++ return;
++}
+
-+#define __raw_spin_unlock_string \
-+ "movb $1,%0" \
-+ :"+m" (lock->slock) : : "memory"
++#include <asm/hypercall.h>
+
++/* BEGIN: all of these need a new home */
++struct vm_area_struct;
++int direct_remap_pfn_range(struct vm_area_struct *vma, unsigned long address,
++ unsigned long mfn, unsigned long size,
++ pgprot_t prot, domid_t domid);
++#define pfn_to_mfn(x) (x)
++#define mfn_to_pfn(x) (x)
++#define phys_to_machine(phys) ((maddr_t)(phys))
++#define phys_to_machine_mapping_valid(pfn) (1)
+
-+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
-+{
-+ __asm__ __volatile__(
-+ __raw_spin_unlock_string
-+ );
-+}
++/* VIRT <-> MACHINE conversion */
++#define virt_to_machine(v) (phys_to_machine(__pa(v)))
++#define machine_to_virt(m) (__va(m))
++#define virt_to_mfn(v) (pfn_to_mfn(__pa(v) >> PAGE_SHIFT))
++#define mfn_to_virt(m) (__va(mfn_to_pfn(m) << PAGE_SHIFT))
+
-+#else
+
-+#define __raw_spin_unlock_string \
-+ "xchgb %b0, %1" \
-+ :"=q" (oldval), "+m" (lock->slock) \
-+ :"0" (oldval) : "memory"
++#define PIRQ_BASE 0
++#define NR_PIRQS 256
+
-+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
-+{
-+ char oldval = 1;
++#define DYNIRQ_BASE (PIRQ_BASE + NR_PIRQS)
++#define NR_DYNIRQS 256
+
-+ __asm__ __volatile__(
-+ __raw_spin_unlock_string
-+ );
-+}
++#define NR_IPIS 4 /* PPC_MSG_DEBUGGER_BREAK + 1 */
+
++#if NR_IRQS < (NR_PIRQS + NR_DYNIRQS)
++#error to many Xen IRQs
+#endif
+
-+#define __raw_spin_unlock_wait(lock) \
-+ do { while (__raw_spin_is_locked(lock)) cpu_relax(); } while (0)
++#define NR_IRQ_VECTORS NR_IRQS
+
-+/*
-+ * Read-write spinlocks, allowing multiple readers
-+ * but only one writer.
-+ *
-+ * NOTE! it is quite common to have readers in interrupts
-+ * but no interrupt writers. For those circumstances we
-+ * can "mix" irq-safe locks - any writer needs to get a
-+ * irq-safe write-lock, but readers can get non-irqsafe
-+ * read-locks.
-+ *
-+ * On x86, we implement read-write locks as a 32-bit counter
-+ * with the high bit (sign) being the "contended" bit.
-+ *
-+ * The inline assembly is non-obvious. Think about it.
-+ *
-+ * Changed to use the same technique as rw semaphores. See
-+ * semaphore.h for details. -ben
-+ *
-+ * the helpers are in arch/i386/kernel/semaphore.c
-+ */
++#define pirq_to_irq(_x) ((_x) + PIRQ_BASE)
++#define irq_to_pirq(_x) ((_x) - PIRQ_BASE)
+
-+/**
-+ * read_can_lock - would read_trylock() succeed?
-+ * @lock: the rwlock in question.
-+ */
-+#define __raw_read_can_lock(x) ((int)(x)->lock > 0)
++#define dynirq_to_irq(_x) ((_x) + DYNIRQ_BASE)
++#define irq_to_dynirq(_x) ((_x) - DYNIRQ_BASE)
+
-+/**
-+ * write_can_lock - would write_trylock() succeed?
-+ * @lock: the rwlock in question.
-+ */
-+#define __raw_write_can_lock(x) ((x)->lock == RW_LOCK_BIAS)
+
-+static inline void __raw_read_lock(raw_rwlock_t *rw)
-+{
-+ __build_read_lock(rw, "__read_lock_failed");
-+}
++/* END: all of these need a new home */
+
-+static inline void __raw_write_lock(raw_rwlock_t *rw)
-+{
-+ __build_write_lock(rw, "__write_lock_failed");
-+}
++#if defined(CONFIG_X86_64)
++#define MULTI_UVMFLAGS_INDEX 2
++#define MULTI_UVMDOMID_INDEX 3
++#else
++#define MULTI_UVMFLAGS_INDEX 3
++#define MULTI_UVMDOMID_INDEX 4
++#endif
+
-+static inline int __raw_read_trylock(raw_rwlock_t *lock)
-+{
-+ atomic_t *count = (atomic_t *)lock;
-+ atomic_dec(count);
-+ if (atomic_read(count) >= 0)
-+ return 1;
-+ atomic_inc(count);
-+ return 0;
-+}
++extern int is_running_on_xen(void);
+
-+static inline int __raw_write_trylock(raw_rwlock_t *lock)
++static inline void
++MULTI_update_va_mapping(
++ multicall_entry_t *mcl, unsigned long va,
++ pte_t new_val, unsigned long flags)
+{
-+ atomic_t *count = (atomic_t *)lock;
-+ if (atomic_sub_and_test(RW_LOCK_BIAS, count))
-+ return 1;
-+ atomic_add(RW_LOCK_BIAS, count);
-+ return 0;
++ mcl->op = __HYPERVISOR_update_va_mapping;
++ mcl->args[0] = va;
++#if defined(CONFIG_X86_64)
++ mcl->args[1] = new_val.pte;
++ mcl->args[2] = flags;
++#elif defined(CONFIG_X86_PAE)
++ mcl->args[1] = new_val.pte_low;
++ mcl->args[2] = new_val.pte_high;
++ mcl->args[3] = flags;
++#elif defined(CONFIG_PPC64)
++ mcl->args[1] = pte_val(new_val);
++ mcl->args[2] = 0;
++ mcl->args[3] = flags;
++#else
++ mcl->args[1] = new_val.pte_low;
++ mcl->args[2] = 0;
++ mcl->args[3] = flags;
++#endif
+}
+
-+static inline void __raw_read_unlock(raw_rwlock_t *rw)
++static inline void
++MULTI_update_va_mapping_otherdomain(
++ multicall_entry_t *mcl, unsigned long va,
++ pte_t new_val, unsigned long flags, domid_t domid)
+{
-+ asm volatile(LOCK_PREFIX "incl %0" :"+m" (rw->lock) : : "memory");
++ mcl->op = __HYPERVISOR_update_va_mapping_otherdomain;
++ mcl->args[0] = va;
++#if defined(CONFIG_X86_64)
++ mcl->args[1] = new_val.pte;
++ mcl->args[2] = flags;
++ mcl->args[3] = domid;
++#elif defined(CONFIG_X86_PAE)
++ mcl->args[1] = new_val.pte_low;
++ mcl->args[2] = new_val.pte_high;
++ mcl->args[3] = flags;
++ mcl->args[4] = domid;
++#elif defined(CONFIG_PPC64)
++ mcl->args[1] = pte_val(new_val);
++ mcl->args[2] = 0;
++ mcl->args[3] = flags;
++ mcl->args[4] = domid;
++#else
++ mcl->args[1] = new_val.pte_low;
++ mcl->args[2] = 0;
++ mcl->args[3] = flags;
++ mcl->args[4] = domid;
++#endif
+}
+
-+static inline void __raw_write_unlock(raw_rwlock_t *rw)
++#define INVALID_P2M_ENTRY (~0UL)
++#define FOREIGN_FRAME(m) (INVALID_P2M_ENTRY)
++static inline void set_phys_to_machine(unsigned long pfn, unsigned long mfn)
+{
-+ asm volatile(LOCK_PREFIX "addl $" RW_LOCK_BIAS_STR ", %0"
-+ : "+m" (rw->lock) : : "memory");
++ if (pfn != mfn && mfn != INVALID_P2M_ENTRY)
++ printk(KERN_EMERG "%s: pfn: 0x%lx mfn: 0x%lx\n",
++ __func__, pfn, mfn);
++
++ return;
+}
++#define pfn_pte_ma(pfn, prot) __pte(((pfn) << PAGE_SHIFT) | pgprot_val(prot))
+
-+#endif /* __ASM_SPINLOCK_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/swiotlb.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/swiotlb.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/swiotlb.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/swiotlb.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,43 @@
-+#ifndef _ASM_SWIOTLB_H
-+#define _ASM_SWIOTLB_H 1
++typedef unsigned long maddr_t;
++typedef unsigned long paddr_t;
+
-+/* SWIOTLB interface */
++#ifdef CONFIG_XEN_SCRUB_PAGES
+
-+extern dma_addr_t swiotlb_map_single(struct device *hwdev, void *ptr, size_t size,
-+ int dir);
-+extern void swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr,
-+ size_t size, int dir);
-+extern void swiotlb_sync_single_for_cpu(struct device *hwdev,
-+ dma_addr_t dev_addr,
-+ size_t size, int dir);
-+extern void swiotlb_sync_single_for_device(struct device *hwdev,
-+ dma_addr_t dev_addr,
-+ size_t size, int dir);
-+extern void swiotlb_sync_sg_for_cpu(struct device *hwdev,
-+ struct scatterlist *sg, int nelems,
-+ int dir);
-+extern void swiotlb_sync_sg_for_device(struct device *hwdev,
-+ struct scatterlist *sg, int nelems,
-+ int dir);
-+extern int swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg,
-+ int nents, int direction);
-+extern void swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg,
-+ int nents, int direction);
-+extern int swiotlb_dma_mapping_error(dma_addr_t dma_addr);
-+#ifdef CONFIG_HIGHMEM
-+extern dma_addr_t swiotlb_map_page(struct device *hwdev, struct page *page,
-+ unsigned long offset, size_t size,
-+ enum dma_data_direction direction);
-+extern void swiotlb_unmap_page(struct device *hwdev, dma_addr_t dma_address,
-+ size_t size, enum dma_data_direction direction);
-+#endif
-+extern int swiotlb_dma_supported(struct device *hwdev, u64 mask);
-+extern void swiotlb_init(void);
++static inline void scrub_pages(void *p, unsigned n)
++{
++ unsigned i;
+
-+#ifdef CONFIG_SWIOTLB
-+extern int swiotlb;
++ for (i = 0; i < n; i++) {
++ clear_page(p);
++ p += PAGE_SIZE;
++ }
++}
+#else
-+#define swiotlb 0
-+#endif
-+
++#define scrub_pages(_p,_n) ((void)0)
+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/synch_bitops.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/synch_bitops.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/synch_bitops.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/synch_bitops.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,145 @@
-+#ifndef __XEN_SYNCH_BITOPS_H__
-+#define __XEN_SYNCH_BITOPS_H__
+
+/*
-+ * Copyright 1992, Linus Torvalds.
-+ * Heavily modified to provide guaranteed strong synchronisation
-+ * when communicating with Xen or other guest OSes running on other CPUs.
++ * for blktap.c
++ * int create_lookup_pte_addr(struct mm_struct *mm,
++ * unsigned long address,
++ * uint64_t *ptep);
+ */
++#define create_lookup_pte_addr(mm, address, ptep) \
++ ({ \
++ printk(KERN_EMERG \
++ "%s:%d " \
++ "create_lookup_pte_addr() isn't supported.\n", \
++ __func__, __LINE__); \
++ BUG(); \
++ (-ENOSYS); \
++ })
+
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
-+#endif
++#endif /* __HYPERVISOR_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-powerpc/xen/asm/maddr.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-powerpc/xen/asm/maddr.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,7 @@
++#ifndef _POWERPC_MADDR_H
++#define _POWERPC_MADDR_H
+
-+#define ADDR (*(volatile long *) addr)
++#include <xen/features.h>
++#include <xen/interface/xen.h>
+
-+static __inline__ void synch_set_bit(int nr, volatile void * addr)
-+{
-+ __asm__ __volatile__ (
-+ "lock btsl %1,%0"
-+ : "+m" (ADDR) : "Ir" (nr) : "memory" );
-+}
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-powerpc/xen/asm/synch_bitops.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-powerpc/xen/asm/synch_bitops.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,100 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright 2006 IBM Corp.
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
+
-+static __inline__ void synch_clear_bit(int nr, volatile void * addr)
-+{
-+ __asm__ __volatile__ (
-+ "lock btrl %1,%0"
-+ : "+m" (ADDR) : "Ir" (nr) : "memory" );
-+}
++#ifndef __SYNCH_BITOPS_H__
++#define __SYNCH_BITOPS_H__
+
-+static __inline__ void synch_change_bit(int nr, volatile void * addr)
-+{
-+ __asm__ __volatile__ (
-+ "lock btcl %1,%0"
-+ : "+m" (ADDR) : "Ir" (nr) : "memory" );
-+}
++#include <linux/config.h>
++#include <xen/interface/xen.h>
+
-+static __inline__ int synch_test_and_set_bit(int nr, volatile void * addr)
-+{
-+ int oldbit;
-+ __asm__ __volatile__ (
-+ "lock btsl %2,%1\n\tsbbl %0,%0"
-+ : "=r" (oldbit), "+m" (ADDR) : "Ir" (nr) : "memory");
-+ return oldbit;
-+}
++#ifdef CONFIG_SMP
++#include <asm/bitops.h>
+
-+static __inline__ int synch_test_and_clear_bit(int nr, volatile void * addr)
-+{
-+ int oldbit;
-+ __asm__ __volatile__ (
-+ "lock btrl %2,%1\n\tsbbl %0,%0"
-+ : "=r" (oldbit), "+m" (ADDR) : "Ir" (nr) : "memory");
-+ return oldbit;
-+}
++#define synch_change_bit(a,b) change_bit(a,b)
++#define synch_clear_bit(a,b) clear_bit(a,b)
++#define synch_const_test_bit(a,b) const_test_bit(a,b)
++#define synch_set_bit(a,b) set_bit(a,b)
++#define synch_test_and_set_bit(a,b) test_and_set_bit(a,b)
++#define synch_test_and_change_bit(a,b) test_and_change_bit(a,b)
++#define synch_test_and_clear_bit(a,b) test_and_clear_bit(a,b)
++#define synch_test_bit(a,b) test_bit(a,b)
+
-+static __inline__ int synch_test_and_change_bit(int nr, volatile void * addr)
++static __inline__ unsigned long
++__synch_cmpxchg_u16(volatile unsigned short *p, unsigned long old, unsigned long new)
+{
-+ int oldbit;
-+
-+ __asm__ __volatile__ (
-+ "lock btcl %2,%1\n\tsbbl %0,%0"
-+ : "=r" (oldbit), "+m" (ADDR) : "Ir" (nr) : "memory");
-+ return oldbit;
++ int idx;
++ volatile unsigned int *xp = (unsigned int *)((ulong)p & ~(0x3UL));
++ union {
++ unsigned int word;
++ struct {
++ unsigned short s[2];
++ }s;
++ } xold, xnew;
++
++ /* we could start the reservation here and copy the u32
++ * assembler, but I don't think it will gain us a whole
++ * lot. */
++ xold.word = *xp;
++ xnew.word = xold.word;
++ idx = ((ulong)p >> 1) & 0x1;
++ xold.s.s[idx] = old;
++ xnew.s.s[idx] = new;
++
++ return __cmpxchg_u32(xp, xold.word, xnew.word);
+}
+
-+struct __synch_xchg_dummy { unsigned long a[100]; };
-+#define __synch_xg(x) ((struct __synch_xchg_dummy *)(x))
-+
-+#define synch_cmpxchg(ptr, old, new) \
-+((__typeof__(*(ptr)))__synch_cmpxchg((ptr),\
-+ (unsigned long)(old), \
-+ (unsigned long)(new), \
-+ sizeof(*(ptr))))
-+
-+static inline unsigned long __synch_cmpxchg(volatile void *ptr,
-+ unsigned long old,
-+ unsigned long new, int size)
++/*
++ * This function doesn't exist, so you'll get a linker error
++ * if something tries to do an invalid xchg().
++ */
++extern void __synch_cmpxchg_called_with_bad_pointer(void);
++static __inline__ unsigned long
++__synch_cmpxchg(volatile void *ptr, unsigned long old, unsigned long new,
++ unsigned int size)
+{
-+ unsigned long prev;
+ switch (size) {
-+ case 1:
-+ __asm__ __volatile__("lock; cmpxchgb %b1,%2"
-+ : "=a"(prev)
-+ : "q"(new), "m"(*__synch_xg(ptr)),
-+ "0"(old)
-+ : "memory");
-+ return prev;
+ case 2:
-+ __asm__ __volatile__("lock; cmpxchgw %w1,%2"
-+ : "=a"(prev)
-+ : "r"(new), "m"(*__synch_xg(ptr)),
-+ "0"(old)
-+ : "memory");
-+ return prev;
-+#ifdef CONFIG_X86_64
++ return __synch_cmpxchg_u16(ptr, old, new);
+ case 4:
-+ __asm__ __volatile__("lock; cmpxchgl %k1,%2"
-+ : "=a"(prev)
-+ : "r"(new), "m"(*__synch_xg(ptr)),
-+ "0"(old)
-+ : "memory");
-+ return prev;
++ return __cmpxchg_u32(ptr, old, new);
++#ifdef CONFIG_PPC64
+ case 8:
-+ __asm__ __volatile__("lock; cmpxchgq %1,%2"
-+ : "=a"(prev)
-+ : "r"(new), "m"(*__synch_xg(ptr)),
-+ "0"(old)
-+ : "memory");
-+ return prev;
-+#else
-+ case 4:
-+ __asm__ __volatile__("lock; cmpxchgl %1,%2"
-+ : "=a"(prev)
-+ : "r"(new), "m"(*__synch_xg(ptr)),
-+ "0"(old)
-+ : "memory");
-+ return prev;
++ return __cmpxchg_u64(ptr, old, new);
+#endif
+ }
++ __synch_cmpxchg_called_with_bad_pointer();
+ return old;
+}
+
-+static __always_inline int synch_const_test_bit(int nr,
-+ const volatile void * addr)
-+{
-+ return ((1UL << (nr & 31)) &
-+ (((const volatile unsigned int *) addr)[nr >> 5])) != 0;
-+}
++#define synch_cmpxchg(ptr,o,n) \
++ ({ \
++ __typeof__(*(ptr)) _o_ = (o); \
++ __typeof__(*(ptr)) _n_ = (n); \
++ (__typeof__(*(ptr))) __synch_cmpxchg((ptr), (unsigned long)_o_, \
++ (unsigned long)_n_, sizeof(*(ptr))); \
++ })
+
-+static __inline__ int synch_var_test_bit(int nr, volatile void * addr)
-+{
-+ int oldbit;
-+ __asm__ __volatile__ (
-+ "btl %2,%1\n\tsbbl %0,%0"
-+ : "=r" (oldbit) : "m" (ADDR), "Ir" (nr) );
-+ return oldbit;
-+}
++#define synch_cmpxchg_subword(ptr,o,n) __synch_cmpxchg_u16((ptr), (o), (n))
+
-+#define synch_test_bit(nr,addr) \
-+(__builtin_constant_p(nr) ? \
-+ synch_const_test_bit((nr),(addr)) : \
-+ synch_var_test_bit((nr),(addr)))
++#else
++#error "this only works for CONFIG_SMP"
++#endif
+
-+#define synch_cmpxchg_subword synch_cmpxchg
++#endif /* __SYNCH_BITOPS_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/acpi.h
+--- a/include/asm-x86_64/acpi.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-x86_64/acpi.h Wed Sep 10 10:54:08 2008 +0100
+@@ -28,6 +28,9 @@
+
+ #ifdef __KERNEL__
+
++#ifdef CONFIG_XEN
++#include <xen/interface/platform.h>
++#endif
+ #include <acpi/pdc_intel.h>
+
+ #define COMPILER_DEPENDENT_INT64 long long
+@@ -129,6 +132,27 @@
+ }
+ extern int acpi_irq_balance_set(char *str);
+
++#ifdef CONFIG_XEN
++static inline int acpi_notify_hypervisor_state(u8 sleep_state,
++ u32 pm1a_cnt_val,
++ u32 pm1b_cnt_val)
++{
++ struct xen_platform_op op = {
++ .cmd = XENPF_enter_acpi_sleep,
++ .interface_version = XENPF_INTERFACE_VERSION,
++ .u = {
++ .enter_acpi_sleep = {
++ .pm1a_cnt_val = pm1a_cnt_val,
++ .pm1b_cnt_val = pm1b_cnt_val,
++ .sleep_state = sleep_state,
++ },
++ },
++ };
+
-+#endif /* __XEN_SYNCH_BITOPS_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/system.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/system.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/system.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/system.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,488 @@
-+#ifndef __ASM_SYSTEM_H
-+#define __ASM_SYSTEM_H
++ return HYPERVISOR_platform_op(&op);
++}
++#endif /* CONFIG_XEN */
+
-+#include <linux/kernel.h>
-+#include <asm/segment.h>
-+#include <asm/cpufeature.h>
-+#include <linux/bitops.h> /* for LOCK_PREFIX */
-+#include <asm/synch_bitops.h>
-+#include <asm/hypervisor.h>
+ #else /* !CONFIG_ACPI */
+
+ #define acpi_lapic 0
+@@ -152,7 +176,6 @@
+
+ /* early initialization routine */
+ extern void acpi_reserve_bootmem(void);
+-
+ #endif /*CONFIG_ACPI_SLEEP*/
+
+ #define boot_cpu_physical_apicid boot_cpu_id
+@@ -162,7 +185,9 @@
+
+ extern u8 x86_acpiid_to_apicid[];
+
++#ifndef CONFIG_XEN
+ #define ARCH_HAS_POWER_INIT 1
++#endif
+
+ extern int acpi_skip_timer_override;
+
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/agp.h
+--- a/include/asm-x86_64/agp.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-x86_64/agp.h Wed Sep 10 10:54:08 2008 +0100
+@@ -10,8 +10,10 @@
+ * with different cachability attributes for the same page.
+ */
+
+-int map_page_into_agp(struct page *page);
+-int unmap_page_from_agp(struct page *page);
++/* Caller's responsibility to call global_flush_tlb() for
++ * performance reasons */
++#define map_page_into_agp(page) change_page_attr(page, 1, PAGE_KERNEL_NOCACHE)
++#define unmap_page_from_agp(page) change_page_attr(page, 1, PAGE_KERNEL)
+ #define flush_agp_mappings() global_flush_tlb()
+
+ /* Could use CLFLUSH here if the cpu supports it. But then it would
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/apic.h
+--- a/include/asm-x86_64/apic.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-x86_64/apic.h Wed Sep 10 10:54:08 2008 +0100
+@@ -98,11 +98,13 @@
+ extern int disable_timer_pin_1;
+
+
++#ifndef CONFIG_XEN
+ void smp_send_timer_broadcast_ipi(void);
+ void switch_APIC_timer_to_ipi(void *cpumask);
+ void switch_ipi_to_APIC_timer(void *cpumask);
+
+ #define ARCH_APICTIMER_STOPS_ON_C3 1
++#endif
+
+ #endif /* CONFIG_X86_LOCAL_APIC */
+
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/io_apic.h
+--- a/include/asm-x86_64/io_apic.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-x86_64/io_apic.h Wed Sep 10 10:54:08 2008 +0100
+@@ -12,7 +12,7 @@
+
+ #ifdef CONFIG_X86_IO_APIC
+
+-#ifdef CONFIG_PCI_MSI
++#if defined(CONFIG_PCI_MSI) && !defined(CONFIG_XEN)
+ static inline int use_pci_vector(void) {return 1;}
+ static inline void disable_edge_ioapic_vector(unsigned int vector) { }
+ static inline void mask_and_ack_level_ioapic_vector(unsigned int vector) { }
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/kexec.h
+--- a/include/asm-x86_64/kexec.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-x86_64/kexec.h Wed Sep 10 10:54:08 2008 +0100
+@@ -1,5 +1,26 @@
+ #ifndef _X86_64_KEXEC_H
+ #define _X86_64_KEXEC_H
+
-+#ifdef __KERNEL__
++#define PA_CONTROL_PAGE 0
++#define VA_CONTROL_PAGE 1
++#define PA_PGD 2
++#define VA_PGD 3
++#define PA_PUD_0 4
++#define VA_PUD_0 5
++#define PA_PMD_0 6
++#define VA_PMD_0 7
++#define PA_PTE_0 8
++#define VA_PTE_0 9
++#define PA_PUD_1 10
++#define VA_PUD_1 11
++#define PA_PMD_1 12
++#define VA_PMD_1 13
++#define PA_PTE_1 14
++#define VA_PTE_1 15
++#define PA_TABLE_PAGE 16
++#define PAGES_NR 17
+
-+struct task_struct; /* one of the stranger aspects of C forward declarations.. */
-+extern struct task_struct * FASTCALL(__switch_to(struct task_struct *prev, struct task_struct *next));
++#ifndef __ASSEMBLY__
+
+ #include <linux/string.h>
+
+@@ -64,4 +85,25 @@
+ newregs->rip = (unsigned long)current_text_addr();
+ }
+ }
+
-+/*
-+ * Saving eflags is important. It switches not only IOPL between tasks,
-+ * it also protects other tasks from NT leaking through sysenter etc.
++NORET_TYPE void
++relocate_kernel(unsigned long indirection_page,
++ unsigned long page_list,
++ unsigned long start_address) ATTRIB_NORET;
++
++/* Under Xen we need to work with machine addresses. These macros give the
++ * machine address of a certain page to the generic kexec code instead of
++ * the pseudo physical address which would be given by the default macros.
+ */
-+#define switch_to(prev,next,last) do { \
-+ unsigned long esi,edi; \
-+ asm volatile("pushfl\n\t" /* Save flags */ \
-+ "pushl %%ebp\n\t" \
-+ "movl %%esp,%0\n\t" /* save ESP */ \
-+ "movl %5,%%esp\n\t" /* restore ESP */ \
-+ "movl $1f,%1\n\t" /* save EIP */ \
-+ "pushl %6\n\t" /* restore EIP */ \
-+ "jmp __switch_to\n" \
-+ "1:\t" \
-+ "popl %%ebp\n\t" \
-+ "popfl" \
-+ :"=m" (prev->thread.esp),"=m" (prev->thread.eip), \
-+ "=a" (last),"=S" (esi),"=D" (edi) \
-+ :"m" (next->thread.esp),"m" (next->thread.eip), \
-+ "2" (prev), "d" (next)); \
-+} while (0)
+
-+#define _set_base(addr,base) do { unsigned long __pr; \
-+__asm__ __volatile__ ("movw %%dx,%1\n\t" \
-+ "rorl $16,%%edx\n\t" \
-+ "movb %%dl,%2\n\t" \
-+ "movb %%dh,%3" \
-+ :"=&d" (__pr) \
-+ :"m" (*((addr)+2)), \
-+ "m" (*((addr)+4)), \
-+ "m" (*((addr)+7)), \
-+ "0" (base) \
-+ ); } while(0)
++#ifdef CONFIG_XEN
++#define KEXEC_ARCH_HAS_PAGE_MACROS
++#define kexec_page_to_pfn(page) pfn_to_mfn(page_to_pfn(page))
++#define kexec_pfn_to_page(pfn) pfn_to_page(mfn_to_pfn(pfn))
++#define kexec_virt_to_phys(addr) virt_to_machine(addr)
++#define kexec_phys_to_virt(addr) phys_to_virt(machine_to_phys(addr))
++#endif
+
-+#define _set_limit(addr,limit) do { unsigned long __lr; \
-+__asm__ __volatile__ ("movw %%dx,%1\n\t" \
-+ "rorl $16,%%edx\n\t" \
-+ "movb %2,%%dh\n\t" \
-+ "andb $0xf0,%%dh\n\t" \
-+ "orb %%dh,%%dl\n\t" \
-+ "movb %%dl,%2" \
-+ :"=&d" (__lr) \
-+ :"m" (*(addr)), \
-+ "m" (*((addr)+6)), \
-+ "0" (limit) \
-+ ); } while(0)
++#endif /* __ASSEMBLY__ */
+
-+#define set_base(ldt,base) _set_base( ((char *)&(ldt)) , (base) )
-+#define set_limit(ldt,limit) _set_limit( ((char *)&(ldt)) , ((limit)-1) )
+ #endif /* _X86_64_KEXEC_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/agp.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/agp.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,40 @@
++#ifndef AGP_H
++#define AGP_H 1
++
++#include <asm/cacheflush.h>
++#include <asm/system.h>
+
+/*
-+ * Load a segment. Fall back on loading the zero
-+ * segment if something goes wrong..
++ * Functions to keep the agpgart mappings coherent.
++ * The GART gives the CPU a physical alias of memory. The alias is
++ * mapped uncacheable. Make sure there are no conflicting mappings
++ * with different cachability attributes for the same page.
+ */
-+#define loadsegment(seg,value) \
-+ asm volatile("\n" \
-+ "1:\t" \
-+ "mov %0,%%" #seg "\n" \
-+ "2:\n" \
-+ ".section .fixup,\"ax\"\n" \
-+ "3:\t" \
-+ "pushl $0\n\t" \
-+ "popl %%" #seg "\n\t" \
-+ "jmp 2b\n" \
-+ ".previous\n" \
-+ ".section __ex_table,\"a\"\n\t" \
-+ ".align 4\n\t" \
-+ ".long 1b,3b\n" \
-+ ".previous" \
-+ : :"rm" (value))
++
++#define map_page_into_agp(page) ( \
++ xen_create_contiguous_region((unsigned long)page_address(page), 0, 32) \
++ ?: change_page_attr(page, 1, PAGE_KERNEL_NOCACHE))
++#define unmap_page_from_agp(page) ( \
++ xen_destroy_contiguous_region((unsigned long)page_address(page), 0), \
++ /* only a fallback: xen_destroy_contiguous_region uses PAGE_KERNEL */ \
++ change_page_attr(page, 1, PAGE_KERNEL))
++#define flush_agp_mappings() global_flush_tlb()
++
++/* Could use CLFLUSH here if the cpu supports it. But then it would
++ need to be called for each cacheline of the whole page so it may not be
++ worth it. Would need a page for it. */
++#define flush_agp_cache() wbinvd()
++
++/* Convert a physical address to an address suitable for the GART. */
++#define phys_to_gart(x) phys_to_machine(x)
++#define gart_to_phys(x) machine_to_phys(x)
++
++/* GATT allocation. Returns/accepts GATT kernel virtual address. */
++#define alloc_gatt_pages(order) ({ \
++ char *_t; dma_addr_t _d; \
++ _t = dma_alloc_coherent(NULL,PAGE_SIZE<<(order),&_d,GFP_KERNEL); \
++ _t; })
++#define free_gatt_pages(table, order) \
++ dma_free_coherent(NULL,PAGE_SIZE<<(order),(table),virt_to_bus(table))
++
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/arch_hooks.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/arch_hooks.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,27 @@
++#ifndef _ASM_ARCH_HOOKS_H
++#define _ASM_ARCH_HOOKS_H
++
++#include <linux/interrupt.h>
+
+/*
-+ * Save a segment register away
++ * linux/include/asm/arch_hooks.h
++ *
++ * define the architecture specific hooks
+ */
-+#define savesegment(seg, value) \
-+ asm volatile("mov %%" #seg ",%0":"=rm" (value))
+
-+#define read_cr0() ({ \
-+ unsigned int __dummy; \
-+ __asm__ __volatile__( \
-+ "movl %%cr0,%0\n\t" \
-+ :"=r" (__dummy)); \
-+ __dummy; \
-+})
-+#define write_cr0(x) \
-+ __asm__ __volatile__("movl %0,%%cr0": :"r" (x))
++/* these aren't arch hooks, they are generic routines
++ * that can be used by the hooks */
++extern void init_ISA_irqs(void);
++extern void apic_intr_init(void);
++extern void smp_intr_init(void);
++extern irqreturn_t timer_interrupt(int irq, void *dev_id, struct pt_regs *regs);
+
-+#define read_cr2() (current_vcpu_info()->arch.cr2)
-+#define write_cr2(x) \
-+ __asm__ __volatile__("movl %0,%%cr2": :"r" (x))
++/* these are the defined hooks */
++extern void intr_init_hook(void);
++extern void pre_intr_init_hook(void);
++extern void pre_setup_arch_hook(void);
++extern void trap_init_hook(void);
++extern void time_init_hook(void);
++extern void mca_nmi_hook(void);
+
-+#define read_cr3() ({ \
-+ unsigned int __dummy; \
-+ __asm__ ( \
-+ "movl %%cr3,%0\n\t" \
-+ :"=r" (__dummy)); \
-+ __dummy = xen_cr3_to_pfn(__dummy); \
-+ mfn_to_pfn(__dummy) << PAGE_SHIFT; \
-+})
-+#define write_cr3(x) ({ \
-+ unsigned int __dummy = pfn_to_mfn((x) >> PAGE_SHIFT); \
-+ __dummy = xen_pfn_to_cr3(__dummy); \
-+ __asm__ __volatile__("movl %0,%%cr3": :"r" (__dummy)); \
-+})
-+#define read_cr4() ({ \
-+ unsigned int __dummy; \
-+ __asm__( \
-+ "movl %%cr4,%0\n\t" \
-+ :"=r" (__dummy)); \
-+ __dummy; \
-+})
-+#define read_cr4_safe() ({ \
-+ unsigned int __dummy; \
-+ /* This could fault if %cr4 does not exist */ \
-+ __asm__("1: movl %%cr4, %0 \n" \
-+ "2: \n" \
-+ ".section __ex_table,\"a\" \n" \
-+ ".long 1b,2b \n" \
-+ ".previous \n" \
-+ : "=r" (__dummy): "0" (0)); \
-+ __dummy; \
-+})
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/bootsetup.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/bootsetup.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,42 @@
+
-+#define write_cr4(x) \
-+ __asm__ __volatile__("movl %0,%%cr4": :"r" (x))
++#ifndef _X86_64_BOOTSETUP_H
++#define _X86_64_BOOTSETUP_H 1
++
++#define BOOT_PARAM_SIZE 4096
++extern char x86_boot_params[BOOT_PARAM_SIZE];
+
+/*
-+ * Clear and set 'TS' bit respectively
++ * This is set up by the setup-routine at boot-time
+ */
-+#define clts() (HYPERVISOR_fpu_taskswitch(0))
-+#define stts() (HYPERVISOR_fpu_taskswitch(1))
++#define PARAM ((unsigned char *)x86_boot_params)
++#define SCREEN_INFO (*(struct screen_info *) (PARAM+0))
++#define EXT_MEM_K (*(unsigned short *) (PARAM+2))
++#define ALT_MEM_K (*(unsigned int *) (PARAM+0x1e0))
++#define E820_MAP_NR (*(char*) (PARAM+E820NR))
++#define E820_MAP ((struct e820entry *) (PARAM+E820MAP))
++#define APM_BIOS_INFO (*(struct apm_bios_info *) (PARAM+0x40))
++#define DRIVE_INFO (*(struct drive_info_struct *) (PARAM+0x80))
++#define SYS_DESC_TABLE (*(struct sys_desc_table_struct*)(PARAM+0xa0))
++#define MOUNT_ROOT_RDONLY (*(unsigned short *) (PARAM+0x1F2))
++#define RAMDISK_FLAGS (*(unsigned short *) (PARAM+0x1F8))
++#define SAVED_VIDEO_MODE (*(unsigned short *) (PARAM+0x1FA))
++#define ORIG_ROOT_DEV (*(unsigned short *) (PARAM+0x1FC))
++#define AUX_DEVICE_INFO (*(unsigned char *) (PARAM+0x1FF))
++#define LOADER_TYPE (*(unsigned char *) (PARAM+0x210))
++#define KERNEL_START (*(unsigned int *) (PARAM+0x214))
+
-+#endif /* __KERNEL__ */
++#define INITRD_START (__pa(xen_start_info->mod_start))
++#define INITRD_SIZE (xen_start_info->mod_len)
++#define EDID_INFO (*(struct edid_info *) (PARAM+0x440))
+
-+#define wbinvd() \
-+ __asm__ __volatile__ ("wbinvd": : :"memory")
++#define EDD_NR (*(unsigned char *) (PARAM+EDDNR))
++#define EDD_MBR_SIG_NR (*(unsigned char *) (PARAM+EDD_MBR_SIG_NR_BUF))
++#define EDD_MBR_SIGNATURE ((unsigned int *) (PARAM+EDD_MBR_SIG_BUF))
++#define EDD_BUF ((struct edd_info *) (PARAM+EDDBUF))
++#define COMMAND_LINE saved_command_line
+
-+static inline unsigned long get_limit(unsigned long segment)
-+{
-+ unsigned long __limit;
-+ __asm__("lsll %1,%0"
-+ :"=r" (__limit):"r" (segment));
-+ return __limit+1;
-+}
++#define RAMDISK_IMAGE_START_MASK 0x07FF
++#define RAMDISK_PROMPT_FLAG 0x8000
++#define RAMDISK_LOAD_FLAG 0x4000
+
-+#define nop() __asm__ __volatile__ ("nop")
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/desc.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/desc.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,265 @@
++/* Written 2000 by Andi Kleen */
++#ifndef __ARCH_DESC_H
++#define __ARCH_DESC_H
+
-+#define xchg(ptr,v) ((__typeof__(*(ptr)))__xchg((unsigned long)(v),(ptr),sizeof(*(ptr))))
++#include <linux/threads.h>
++#include <asm/ldt.h>
+
-+#define tas(ptr) (xchg((ptr),1))
++#ifndef __ASSEMBLY__
+
-+struct __xchg_dummy { unsigned long a[100]; };
-+#define __xg(x) ((struct __xchg_dummy *)(x))
++#include <linux/string.h>
++#include <linux/smp.h>
+
++#include <asm/segment.h>
++#include <asm/mmu.h>
+
-+#ifdef CONFIG_X86_CMPXCHG64
++// 8 byte segment descriptor
++struct desc_struct {
++ u16 limit0;
++ u16 base0;
++ unsigned base1 : 8, type : 4, s : 1, dpl : 2, p : 1;
++ unsigned limit : 4, avl : 1, l : 1, d : 1, g : 1, base2 : 8;
++} __attribute__((packed));
+
-+/*
-+ * The semantics of XCHGCMP8B are a bit strange, this is why
-+ * there is a loop and the loading of %%eax and %%edx has to
-+ * be inside. This inlines well in most cases, the cached
-+ * cost is around ~38 cycles. (in the future we might want
-+ * to do an SIMD/3DNOW!/MMX/FPU 64-bit store here, but that
-+ * might have an implicit FPU-save as a cost, so it's not
-+ * clear which path to go.)
-+ *
-+ * cmpxchg8b must be used with the lock prefix here to allow
-+ * the instruction to be executed atomically, see page 3-102
-+ * of the instruction set reference 24319102.pdf. We need
-+ * the reader side to see the coherent 64bit value.
-+ */
-+static inline void __set_64bit (unsigned long long * ptr,
-+ unsigned int low, unsigned int high)
-+{
-+ __asm__ __volatile__ (
-+ "\n1:\t"
-+ "movl (%0), %%eax\n\t"
-+ "movl 4(%0), %%edx\n\t"
-+ "lock cmpxchg8b (%0)\n\t"
-+ "jnz 1b"
-+ : /* no outputs */
-+ : "D"(ptr),
-+ "b"(low),
-+ "c"(high)
-+ : "ax","dx","memory");
-+}
++struct n_desc_struct {
++ unsigned int a,b;
++};
+
-+static inline void __set_64bit_constant (unsigned long long *ptr,
-+ unsigned long long value)
-+{
-+ __set_64bit(ptr,(unsigned int)(value), (unsigned int)((value)>>32ULL));
-+}
-+#define ll_low(x) *(((unsigned int*)&(x))+0)
-+#define ll_high(x) *(((unsigned int*)&(x))+1)
++enum {
++ GATE_INTERRUPT = 0xE,
++ GATE_TRAP = 0xF,
++ GATE_CALL = 0xC,
++};
+
-+static inline void __set_64bit_var (unsigned long long *ptr,
-+ unsigned long long value)
-+{
-+ __set_64bit(ptr,ll_low(value), ll_high(value));
-+}
++// 16byte gate
++struct gate_struct {
++ u16 offset_low;
++ u16 segment;
++ unsigned ist : 3, zero0 : 5, type : 5, dpl : 2, p : 1;
++ u16 offset_middle;
++ u32 offset_high;
++ u32 zero1;
++} __attribute__((packed));
+
-+#define set_64bit(ptr,value) \
-+(__builtin_constant_p(value) ? \
-+ __set_64bit_constant(ptr, value) : \
-+ __set_64bit_var(ptr, value) )
++#define PTR_LOW(x) ((unsigned long)(x) & 0xFFFF)
++#define PTR_MIDDLE(x) (((unsigned long)(x) >> 16) & 0xFFFF)
++#define PTR_HIGH(x) ((unsigned long)(x) >> 32)
+
-+#define _set_64bit(ptr,value) \
-+(__builtin_constant_p(value) ? \
-+ __set_64bit(ptr, (unsigned int)(value), (unsigned int)((value)>>32ULL) ) : \
-+ __set_64bit(ptr, ll_low(value), ll_high(value)) )
++enum {
++ DESC_TSS = 0x9,
++ DESC_LDT = 0x2,
++};
+
-+#endif
++// LDT or TSS descriptor in the GDT. 16 bytes.
++struct ldttss_desc {
++ u16 limit0;
++ u16 base0;
++ unsigned base1 : 8, type : 5, dpl : 2, p : 1;
++ unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
++ u32 base3;
++ u32 zero1;
++} __attribute__((packed));
+
-+/*
-+ * Note: no "lock" prefix even on SMP: xchg always implies lock anyway
-+ * Note 2: xchg has side effect, so that attribute volatile is necessary,
-+ * but generally the primitive is invalid, *ptr is output argument. --ANK
-+ */
-+static inline unsigned long __xchg(unsigned long x, volatile void * ptr, int size)
-+{
-+ switch (size) {
-+ case 1:
-+ __asm__ __volatile__("xchgb %b0,%1"
-+ :"=q" (x)
-+ :"m" (*__xg(ptr)), "0" (x)
-+ :"memory");
-+ break;
-+ case 2:
-+ __asm__ __volatile__("xchgw %w0,%1"
-+ :"=r" (x)
-+ :"m" (*__xg(ptr)), "0" (x)
-+ :"memory");
-+ break;
-+ case 4:
-+ __asm__ __volatile__("xchgl %0,%1"
-+ :"=r" (x)
-+ :"m" (*__xg(ptr)), "0" (x)
-+ :"memory");
-+ break;
-+ }
-+ return x;
-+}
++struct desc_ptr {
++ unsigned short size;
++ unsigned long address;
++} __attribute__((packed)) ;
+
-+/*
-+ * Atomic compare and exchange. Compare OLD with MEM, if identical,
-+ * store NEW in MEM. Return the initial value in MEM. Success is
-+ * indicated by comparing RETURN with OLD.
-+ */
++extern struct desc_ptr idt_descr, cpu_gdt_descr[NR_CPUS];
+
-+#ifdef CONFIG_X86_CMPXCHG
-+#define __HAVE_ARCH_CMPXCHG 1
-+#define cmpxchg(ptr,o,n)\
-+ ((__typeof__(*(ptr)))__cmpxchg((ptr),(unsigned long)(o),\
-+ (unsigned long)(n),sizeof(*(ptr))))
-+#endif
++extern struct desc_struct cpu_gdt_table[GDT_ENTRIES];
+
-+static inline unsigned long __cmpxchg(volatile void *ptr, unsigned long old,
-+ unsigned long new, int size)
++#define load_TR_desc() asm volatile("ltr %w0"::"r" (GDT_ENTRY_TSS*8))
++#define load_LDT_desc() asm volatile("lldt %w0"::"r" (GDT_ENTRY_LDT*8))
++
++static inline void clear_LDT(void)
+{
-+ unsigned long prev;
-+ switch (size) {
-+ case 1:
-+ __asm__ __volatile__(LOCK_PREFIX "cmpxchgb %b1,%2"
-+ : "=a"(prev)
-+ : "q"(new), "m"(*__xg(ptr)), "0"(old)
-+ : "memory");
-+ return prev;
-+ case 2:
-+ __asm__ __volatile__(LOCK_PREFIX "cmpxchgw %w1,%2"
-+ : "=a"(prev)
-+ : "r"(new), "m"(*__xg(ptr)), "0"(old)
-+ : "memory");
-+ return prev;
-+ case 4:
-+ __asm__ __volatile__(LOCK_PREFIX "cmpxchgl %1,%2"
-+ : "=a"(prev)
-+ : "r"(new), "m"(*__xg(ptr)), "0"(old)
-+ : "memory");
-+ return prev;
-+ }
-+ return old;
++ int cpu = get_cpu();
++
++ /*
++ * NB. We load the default_ldt for lcall7/27 handling on demand, as
++ * it slows down context switching. Noone uses it anyway.
++ */
++ cpu = cpu; /* XXX avoid compiler warning */
++ xen_set_ldt(NULL, 0);
++ put_cpu();
+}
+
-+#ifndef CONFIG_X86_CMPXCHG
+/*
-+ * Building a kernel capable running on 80386. It may be necessary to
-+ * simulate the cmpxchg on the 80386 CPU. For that purpose we define
-+ * a function for each of the sizes we support.
++ * This is the ldt that every process will get unless we need
++ * something other than this.
+ */
++extern struct desc_struct default_ldt[];
++#ifndef CONFIG_X86_NO_IDT
++extern struct gate_struct idt_table[];
++#endif
++extern struct desc_ptr cpu_gdt_descr[];
+
-+extern unsigned long cmpxchg_386_u8(volatile void *, u8, u8);
-+extern unsigned long cmpxchg_386_u16(volatile void *, u16, u16);
-+extern unsigned long cmpxchg_386_u32(volatile void *, u32, u32);
++/* the cpu gdt accessor */
++#define cpu_gdt(_cpu) ((struct desc_struct *)cpu_gdt_descr[_cpu].address)
+
-+static inline unsigned long cmpxchg_386(volatile void *ptr, unsigned long old,
-+ unsigned long new, int size)
++static inline void _set_gate(void *adr, unsigned type, unsigned long func, unsigned dpl, unsigned ist)
+{
-+ switch (size) {
-+ case 1:
-+ return cmpxchg_386_u8(ptr, old, new);
-+ case 2:
-+ return cmpxchg_386_u16(ptr, old, new);
-+ case 4:
-+ return cmpxchg_386_u32(ptr, old, new);
-+ }
-+ return old;
-+}
++ struct gate_struct s;
++ s.offset_low = PTR_LOW(func);
++ s.segment = __KERNEL_CS;
++ s.ist = ist;
++ s.p = 1;
++ s.dpl = dpl;
++ s.zero0 = 0;
++ s.zero1 = 0;
++ s.type = type;
++ s.offset_middle = PTR_MIDDLE(func);
++ s.offset_high = PTR_HIGH(func);
++ /* does not need to be atomic because it is only done once at setup time */
++ memcpy(adr, &s, 16);
++}
+
-+#define cmpxchg(ptr,o,n) \
-+({ \
-+ __typeof__(*(ptr)) __ret; \
-+ if (likely(boot_cpu_data.x86 > 3)) \
-+ __ret = __cmpxchg((ptr), (unsigned long)(o), \
-+ (unsigned long)(n), sizeof(*(ptr))); \
-+ else \
-+ __ret = cmpxchg_386((ptr), (unsigned long)(o), \
-+ (unsigned long)(n), sizeof(*(ptr))); \
-+ __ret; \
-+})
-+#endif
++#ifndef CONFIG_X86_NO_IDT
++static inline void set_intr_gate(int nr, void *func)
++{
++ BUG_ON((unsigned)nr > 0xFF);
++ _set_gate(&idt_table[nr], GATE_INTERRUPT, (unsigned long) func, 0, 0);
++}
+
-+#ifdef CONFIG_X86_CMPXCHG64
++static inline void set_intr_gate_ist(int nr, void *func, unsigned ist)
++{
++ BUG_ON((unsigned)nr > 0xFF);
++ _set_gate(&idt_table[nr], GATE_INTERRUPT, (unsigned long) func, 0, ist);
++}
+
-+static inline unsigned long long __cmpxchg64(volatile void *ptr, unsigned long long old,
-+ unsigned long long new)
++static inline void set_system_gate(int nr, void *func)
++{
++ BUG_ON((unsigned)nr > 0xFF);
++ _set_gate(&idt_table[nr], GATE_INTERRUPT, (unsigned long) func, 3, 0);
++}
++
++static inline void set_system_gate_ist(int nr, void *func, unsigned ist)
+{
-+ unsigned long long prev;
-+ __asm__ __volatile__(LOCK_PREFIX "cmpxchg8b %3"
-+ : "=A"(prev)
-+ : "b"((unsigned long)new),
-+ "c"((unsigned long)(new >> 32)),
-+ "m"(*__xg(ptr)),
-+ "0"(old)
-+ : "memory");
-+ return prev;
++ _set_gate(&idt_table[nr], GATE_INTERRUPT, (unsigned long) func, 3, ist);
+}
-+
-+#define cmpxchg64(ptr,o,n)\
-+ ((__typeof__(*(ptr)))__cmpxchg64((ptr),(unsigned long long)(o),\
-+ (unsigned long long)(n)))
-+
+#endif
-+
-+/*
-+ * Force strict CPU ordering.
-+ * And yes, this is required on UP too when we're talking
-+ * to devices.
-+ *
-+ * For now, "wmb()" doesn't actually do anything, as all
-+ * Intel CPU's follow what Intel calls a *Processor Order*,
-+ * in which all writes are seen in the program order even
-+ * outside the CPU.
-+ *
-+ * I expect future Intel CPU's to have a weaker ordering,
-+ * but I'd also expect them to finally get their act together
-+ * and add some real memory barriers if so.
-+ *
-+ * Some non intel clones support out of order store. wmb() ceases to be a
-+ * nop for these.
-+ */
-+
+
-+/*
-+ * Actually only lfence would be needed for mb() because all stores done
-+ * by the kernel should be already ordered. But keep a full barrier for now.
-+ */
++static inline void set_tssldt_descriptor(void *ptr, unsigned long tss, unsigned type,
++ unsigned size)
++{
++ struct ldttss_desc d;
++ memset(&d,0,sizeof(d));
++ d.limit0 = size & 0xFFFF;
++ d.base0 = PTR_LOW(tss);
++ d.base1 = PTR_MIDDLE(tss) & 0xFF;
++ d.type = type;
++ d.p = 1;
++ d.limit1 = (size >> 16) & 0xF;
++ d.base2 = (PTR_MIDDLE(tss) >> 8) & 0xFF;
++ d.base3 = PTR_HIGH(tss);
++ memcpy(ptr, &d, 16);
++}
+
-+#define mb() alternative("lock; addl $0,0(%%esp)", "mfence", X86_FEATURE_XMM2)
-+#define rmb() alternative("lock; addl $0,0(%%esp)", "lfence", X86_FEATURE_XMM2)
++#ifndef CONFIG_X86_NO_TSS
++static inline void set_tss_desc(unsigned cpu, void *addr)
++{
++ /*
++ * sizeof(unsigned long) coming from an extra "long" at the end
++ * of the iobitmap. See tss_struct definition in processor.h
++ *
++ * -1? seg base+limit should be pointing to the address of the
++ * last valid byte
++ */
++ set_tssldt_descriptor(&cpu_gdt(cpu)[GDT_ENTRY_TSS],
++ (unsigned long)addr, DESC_TSS,
++ IO_BITMAP_OFFSET + IO_BITMAP_BYTES + sizeof(unsigned long) - 1);
++}
++#endif
+
-+/**
-+ * read_barrier_depends - Flush all pending reads that subsequents reads
-+ * depend on.
-+ *
-+ * No data-dependent reads from memory-like regions are ever reordered
-+ * over this barrier. All reads preceding this primitive are guaranteed
-+ * to access memory (but not necessarily other CPUs' caches) before any
-+ * reads following this primitive that depend on the data return by
-+ * any of the preceding reads. This primitive is much lighter weight than
-+ * rmb() on most CPUs, and is never heavier weight than is
-+ * rmb().
-+ *
-+ * These ordering constraints are respected by both the local CPU
-+ * and the compiler.
-+ *
-+ * Ordering is not guaranteed by anything other than these primitives,
-+ * not even by data dependencies. See the documentation for
-+ * memory_barrier() for examples and URLs to more information.
-+ *
-+ * For example, the following code would force ordering (the initial
-+ * value of "a" is zero, "b" is one, and "p" is "&a"):
-+ *
-+ * <programlisting>
-+ * CPU 0 CPU 1
-+ *
-+ * b = 2;
-+ * memory_barrier();
-+ * p = &b; q = p;
-+ * read_barrier_depends();
-+ * d = *q;
-+ * </programlisting>
-+ *
-+ * because the read of "*q" depends on the read of "p" and these
-+ * two reads are separated by a read_barrier_depends(). However,
-+ * the following code, with the same initial values for "a" and "b":
-+ *
-+ * <programlisting>
-+ * CPU 0 CPU 1
-+ *
-+ * a = 2;
-+ * memory_barrier();
-+ * b = 3; y = b;
-+ * read_barrier_depends();
-+ * x = a;
-+ * </programlisting>
-+ *
-+ * does not enforce ordering, since there is no data dependency between
-+ * the read of "a" and the read of "b". Therefore, on some CPUs, such
-+ * as Alpha, "y" could be set to 3 and "x" to 0. Use rmb()
-+ * in cases like this where there are no data dependencies.
-+ **/
++static inline void set_ldt_desc(unsigned cpu, void *addr, int size)
++{
++ set_tssldt_descriptor(&cpu_gdt(cpu)[GDT_ENTRY_LDT], (unsigned long)addr,
++ DESC_LDT, size * 8 - 1);
++}
+
-+#define read_barrier_depends() do { } while(0)
++static inline void set_seg_base(unsigned cpu, int entry, void *base)
++{
++ struct desc_struct *d = &cpu_gdt(cpu)[entry];
++ u32 addr = (u32)(u64)base;
++ BUG_ON((u64)base >> 32);
++ d->base0 = addr & 0xffff;
++ d->base1 = (addr >> 16) & 0xff;
++ d->base2 = (addr >> 24) & 0xff;
++}
+
-+#ifdef CONFIG_X86_OOSTORE
-+/* Actually there are no OOO store capable CPUs for now that do SSE,
-+ but make it already an possibility. */
-+#define wmb() alternative("lock; addl $0,0(%%esp)", "sfence", X86_FEATURE_XMM)
-+#else
-+#define wmb() __asm__ __volatile__ ("": : :"memory")
++#define LDT_entry_a(info) \
++ ((((info)->base_addr & 0x0000ffff) << 16) | ((info)->limit & 0x0ffff))
++/* Don't allow setting of the lm bit. It is useless anyways because
++ 64bit system calls require __USER_CS. */
++#define LDT_entry_b(info) \
++ (((info)->base_addr & 0xff000000) | \
++ (((info)->base_addr & 0x00ff0000) >> 16) | \
++ ((info)->limit & 0xf0000) | \
++ (((info)->read_exec_only ^ 1) << 9) | \
++ ((info)->contents << 10) | \
++ (((info)->seg_not_present ^ 1) << 15) | \
++ ((info)->seg_32bit << 22) | \
++ ((info)->limit_in_pages << 23) | \
++ ((info)->useable << 20) | \
++ /* ((info)->lm << 21) | */ \
++ 0x7000)
++
++#define LDT_empty(info) (\
++ (info)->base_addr == 0 && \
++ (info)->limit == 0 && \
++ (info)->contents == 0 && \
++ (info)->read_exec_only == 1 && \
++ (info)->seg_32bit == 0 && \
++ (info)->limit_in_pages == 0 && \
++ (info)->seg_not_present == 1 && \
++ (info)->useable == 0 && \
++ (info)->lm == 0)
++
++#if TLS_SIZE != 24
++# error update this code.
+#endif
+
-+#ifdef CONFIG_SMP
-+#define smp_mb() mb()
-+#define smp_rmb() rmb()
-+#define smp_wmb() wmb()
-+#define smp_read_barrier_depends() read_barrier_depends()
-+#define set_mb(var, value) do { (void) xchg(&var, value); } while (0)
-+#else
-+#define smp_mb() barrier()
-+#define smp_rmb() barrier()
-+#define smp_wmb() barrier()
-+#define smp_read_barrier_depends() do { } while(0)
-+#define set_mb(var, value) do { var = value; barrier(); } while (0)
++static inline void load_TLS(struct thread_struct *t, unsigned int cpu)
++{
++#if 0
++ u64 *gdt = (u64 *)(cpu_gdt(cpu) + GDT_ENTRY_TLS_MIN);
++ gdt[0] = t->tls_array[0];
++ gdt[1] = t->tls_array[1];
++ gdt[2] = t->tls_array[2];
+#endif
++#define C(i) \
++ if (HYPERVISOR_update_descriptor(virt_to_machine(&cpu_gdt(cpu)[GDT_ENTRY_TLS_MIN + i]), \
++ t->tls_array[i])) \
++ BUG();
+
-+#include <linux/irqflags.h>
++ C(0); C(1); C(2);
++#undef C
++}
+
+/*
-+ * disable hlt during certain critical i/o operations
++ * load one particular LDT into the current CPU
+ */
-+#define HAVE_DISABLE_HLT
-+void disable_hlt(void);
-+void enable_hlt(void);
++static inline void load_LDT_nolock (mm_context_t *pc, int cpu)
++{
++ void *segments = pc->ldt;
++ int count = pc->size;
+
-+extern int es7000_plat;
-+void cpu_idle_wait(void);
++ if (likely(!count))
++ segments = NULL;
+
-+/*
-+ * On SMP systems, when the scheduler does migration-cost autodetection,
-+ * it needs a way to flush as much of the CPU's caches as possible:
-+ */
-+static inline void sched_cacheflush(void)
++ xen_set_ldt(segments, count);
++}
++
++static inline void load_LDT(mm_context_t *pc)
+{
-+ wbinvd();
++ int cpu = get_cpu();
++ load_LDT_nolock(pc, cpu);
++ put_cpu();
+}
+
-+extern unsigned long arch_align_stack(unsigned long sp);
-+extern void free_init_pages(char *what, unsigned long begin, unsigned long end);
++extern struct desc_ptr idt_descr;
+
-+void default_idle(void);
++#endif /* !__ASSEMBLY__ */
+
+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/tlbflush.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/tlbflush.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/tlbflush.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/tlbflush.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,101 @@
-+#ifndef _I386_TLBFLUSH_H
-+#define _I386_TLBFLUSH_H
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/dma-mapping.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/dma-mapping.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,207 @@
++#ifndef _X8664_DMA_MAPPING_H
++#define _X8664_DMA_MAPPING_H 1
+
-+#include <linux/mm.h>
-+#include <asm/processor.h>
++/*
++ * IOMMU interface. See Documentation/DMA-mapping.txt and DMA-API.txt for
++ * documentation.
++ */
+
-+#define __flush_tlb() xen_tlb_flush()
-+#define __flush_tlb_global() xen_tlb_flush()
-+#define __flush_tlb_all() xen_tlb_flush()
+
-+extern unsigned long pgkern_mask;
++#include <asm/scatterlist.h>
++#include <asm/io.h>
++#include <asm/swiotlb.h>
+
-+#define cpu_has_invlpg (boot_cpu_data.x86 > 3)
++struct dma_mapping_ops {
++ int (*mapping_error)(dma_addr_t dma_addr);
++ void* (*alloc_coherent)(struct device *dev, size_t size,
++ dma_addr_t *dma_handle, gfp_t gfp);
++ void (*free_coherent)(struct device *dev, size_t size,
++ void *vaddr, dma_addr_t dma_handle);
++ dma_addr_t (*map_single)(struct device *hwdev, void *ptr,
++ size_t size, int direction);
++ /* like map_single, but doesn't check the device mask */
++ dma_addr_t (*map_simple)(struct device *hwdev, char *ptr,
++ size_t size, int direction);
++ void (*unmap_single)(struct device *dev, dma_addr_t addr,
++ size_t size, int direction);
++ void (*sync_single_for_cpu)(struct device *hwdev,
++ dma_addr_t dma_handle, size_t size,
++ int direction);
++ void (*sync_single_for_device)(struct device *hwdev,
++ dma_addr_t dma_handle, size_t size,
++ int direction);
++ void (*sync_single_range_for_cpu)(struct device *hwdev,
++ dma_addr_t dma_handle, unsigned long offset,
++ size_t size, int direction);
++ void (*sync_single_range_for_device)(struct device *hwdev,
++ dma_addr_t dma_handle, unsigned long offset,
++ size_t size, int direction);
++ void (*sync_sg_for_cpu)(struct device *hwdev,
++ struct scatterlist *sg, int nelems,
++ int direction);
++ void (*sync_sg_for_device)(struct device *hwdev,
++ struct scatterlist *sg, int nelems,
++ int direction);
++ int (*map_sg)(struct device *hwdev, struct scatterlist *sg,
++ int nents, int direction);
++ void (*unmap_sg)(struct device *hwdev,
++ struct scatterlist *sg, int nents,
++ int direction);
++ int (*dma_supported)(struct device *hwdev, u64 mask);
++ int is_phys;
++};
+
-+#define __flush_tlb_single(addr) xen_invlpg(addr)
++extern dma_addr_t bad_dma_address;
++extern struct dma_mapping_ops* dma_ops;
++extern int iommu_merge;
+
-+#define __flush_tlb_one(addr) __flush_tlb_single(addr)
++static inline int valid_dma_direction(int dma_direction)
++{
++ return ((dma_direction == DMA_BIDIRECTIONAL) ||
++ (dma_direction == DMA_TO_DEVICE) ||
++ (dma_direction == DMA_FROM_DEVICE));
++}
+
-+/*
-+ * TLB flushing:
-+ *
-+ * - flush_tlb() flushes the current mm struct TLBs
-+ * - flush_tlb_all() flushes all processes TLBs
-+ * - flush_tlb_mm(mm) flushes the specified mm context TLB's
-+ * - flush_tlb_page(vma, vmaddr) flushes one page
-+ * - flush_tlb_range(vma, start, end) flushes a range of pages
-+ * - flush_tlb_kernel_range(start, end) flushes a range of kernel pages
-+ * - flush_tlb_pgtables(mm, start, end) flushes a range of page tables
-+ *
-+ * ..but the i386 has somewhat limited tlb flushing capabilities,
-+ * and page-granular flushes are available only on i486 and up.
-+ */
++#if 0
++static inline int dma_mapping_error(dma_addr_t dma_addr)
++{
++ if (dma_ops->mapping_error)
++ return dma_ops->mapping_error(dma_addr);
+
-+#ifndef CONFIG_SMP
++ return (dma_addr == bad_dma_address);
++}
+
-+#define flush_tlb() __flush_tlb()
-+#define flush_tlb_all() __flush_tlb_all()
-+#define local_flush_tlb() __flush_tlb()
++extern void *dma_alloc_coherent(struct device *dev, size_t size,
++ dma_addr_t *dma_handle, gfp_t gfp);
++extern void dma_free_coherent(struct device *dev, size_t size, void *vaddr,
++ dma_addr_t dma_handle);
+
-+static inline void flush_tlb_mm(struct mm_struct *mm)
++static inline dma_addr_t
++dma_map_single(struct device *hwdev, void *ptr, size_t size,
++ int direction)
+{
-+ if (mm == current->active_mm)
-+ __flush_tlb();
++ BUG_ON(!valid_dma_direction(direction));
++ return dma_ops->map_single(hwdev, ptr, size, direction);
+}
+
-+static inline void flush_tlb_page(struct vm_area_struct *vma,
-+ unsigned long addr)
++static inline void
++dma_unmap_single(struct device *dev, dma_addr_t addr,size_t size,
++ int direction)
+{
-+ if (vma->vm_mm == current->active_mm)
-+ __flush_tlb_one(addr);
++ BUG_ON(!valid_dma_direction(direction));
++ dma_ops->unmap_single(dev, addr, size, direction);
+}
+
-+static inline void flush_tlb_range(struct vm_area_struct *vma,
-+ unsigned long start, unsigned long end)
++#define dma_map_page(dev,page,offset,size,dir) \
++ dma_map_single((dev), page_address(page)+(offset), (size), (dir))
++
++#define dma_unmap_page dma_unmap_single
++
++static inline void
++dma_sync_single_for_cpu(struct device *hwdev, dma_addr_t dma_handle,
++ size_t size, int direction)
+{
-+ if (vma->vm_mm == current->active_mm)
-+ __flush_tlb();
++ BUG_ON(!valid_dma_direction(direction));
++ if (dma_ops->sync_single_for_cpu)
++ dma_ops->sync_single_for_cpu(hwdev, dma_handle, size,
++ direction);
++ flush_write_buffers();
+}
+
-+#else
++static inline void
++dma_sync_single_for_device(struct device *hwdev, dma_addr_t dma_handle,
++ size_t size, int direction)
++{
++ BUG_ON(!valid_dma_direction(direction));
++ if (dma_ops->sync_single_for_device)
++ dma_ops->sync_single_for_device(hwdev, dma_handle, size,
++ direction);
++ flush_write_buffers();
++}
+
-+#include <asm/smp.h>
++static inline void
++dma_sync_single_range_for_cpu(struct device *hwdev, dma_addr_t dma_handle,
++ unsigned long offset, size_t size, int direction)
++{
++ BUG_ON(!valid_dma_direction(direction));
++ if (dma_ops->sync_single_range_for_cpu) {
++ dma_ops->sync_single_range_for_cpu(hwdev, dma_handle, offset, size, direction);
++ }
+
-+#define local_flush_tlb() \
-+ __flush_tlb()
++ flush_write_buffers();
++}
+
-+extern void flush_tlb_all(void);
-+extern void flush_tlb_current_task(void);
-+extern void flush_tlb_mm(struct mm_struct *);
-+extern void flush_tlb_page(struct vm_area_struct *, unsigned long);
++static inline void
++dma_sync_single_range_for_device(struct device *hwdev, dma_addr_t dma_handle,
++ unsigned long offset, size_t size, int direction)
++{
++ BUG_ON(!valid_dma_direction(direction));
++ if (dma_ops->sync_single_range_for_device)
++ dma_ops->sync_single_range_for_device(hwdev, dma_handle,
++ offset, size, direction);
+
-+#define flush_tlb() flush_tlb_current_task()
++ flush_write_buffers();
++}
+
-+static inline void flush_tlb_range(struct vm_area_struct * vma, unsigned long start, unsigned long end)
++static inline void
++dma_sync_sg_for_cpu(struct device *hwdev, struct scatterlist *sg,
++ int nelems, int direction)
+{
-+ flush_tlb_mm(vma->vm_mm);
++ BUG_ON(!valid_dma_direction(direction));
++ if (dma_ops->sync_sg_for_cpu)
++ dma_ops->sync_sg_for_cpu(hwdev, sg, nelems, direction);
++ flush_write_buffers();
+}
+
-+#define TLBSTATE_OK 1
-+#define TLBSTATE_LAZY 2
++static inline void
++dma_sync_sg_for_device(struct device *hwdev, struct scatterlist *sg,
++ int nelems, int direction)
++{
++ BUG_ON(!valid_dma_direction(direction));
++ if (dma_ops->sync_sg_for_device) {
++ dma_ops->sync_sg_for_device(hwdev, sg, nelems, direction);
++ }
+
-+struct tlb_state
++ flush_write_buffers();
++}
++
++static inline int
++dma_map_sg(struct device *hwdev, struct scatterlist *sg, int nents, int direction)
+{
-+ struct mm_struct *active_mm;
-+ int state;
-+ char __cacheline_padding[L1_CACHE_BYTES-8];
-+};
-+DECLARE_PER_CPU(struct tlb_state, cpu_tlbstate);
++ BUG_ON(!valid_dma_direction(direction));
++ return dma_ops->map_sg(hwdev, sg, nents, direction);
++}
+
++static inline void
++dma_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nents,
++ int direction)
++{
++ BUG_ON(!valid_dma_direction(direction));
++ dma_ops->unmap_sg(hwdev, sg, nents, direction);
++}
+
-+#endif
++extern int dma_supported(struct device *hwdev, u64 mask);
+
-+#define flush_tlb_kernel_range(start, end) flush_tlb_all()
++/* same for gart, swiotlb, and nommu */
++static inline int dma_get_cache_alignment(void)
++{
++ return boot_cpu_data.x86_clflush_size;
++}
+
-+static inline void flush_tlb_pgtables(struct mm_struct *mm,
-+ unsigned long start, unsigned long end)
++#define dma_is_consistent(h) 1
++
++extern int dma_set_mask(struct device *dev, u64 mask);
++
++static inline void
++dma_cache_sync(void *vaddr, size_t size, enum dma_data_direction dir)
+{
-+ /* i386 does not keep any page table caches in TLB */
++ flush_write_buffers();
+}
+
-+#endif /* _I386_TLBFLUSH_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/vga.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/vga.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/vga.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/vga.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,20 @@
++extern struct device fallback_dev;
++extern int panic_on_overflow;
++#endif
++
++#endif /* _X8664_DMA_MAPPING_H */
++
++#include <asm-i386/mach-xen/asm/dma-mapping.h>
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/e820.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/e820.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,66 @@
+/*
-+ * Access to VGA videoram
++ * structures and definitions for the int 15, ax=e820 memory map
++ * scheme.
+ *
-+ * (c) 1998 Martin Mares <mj at ucw.cz>
++ * In a nutshell, setup.S populates a scratch table in the
++ * empty_zero_block that contains a list of usable address/size
++ * duples. setup.c, this information is transferred into the e820map,
++ * and in init.c/numa.c, that new information is used to mark pages
++ * reserved or not.
+ */
++#ifndef __E820_HEADER
++#define __E820_HEADER
+
-+#ifndef _LINUX_ASM_VGA_H_
-+#define _LINUX_ASM_VGA_H_
++#include <linux/mmzone.h>
+
-+/*
-+ * On the PC, we can just recalculate addresses and then
-+ * access the videoram directly without any black magic.
-+ */
++#define E820MAP 0x2d0 /* our map */
++#define E820MAX 128 /* number of entries in E820MAP */
++#define E820NR 0x1e8 /* # entries in E820MAP */
+
-+#define VGA_MAP_MEM(x,s) (unsigned long)isa_bus_to_virt(x)
++#define E820_RAM 1
++#define E820_RESERVED 2
++#define E820_ACPI 3 /* usable as RAM once ACPI tables have been read */
++#define E820_NVS 4
+
-+#define vga_readb(x) (*(x))
-+#define vga_writeb(x,y) (*(y) = (x))
++#define HIGH_MEMORY (1024*1024)
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/xenoprof.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/xenoprof.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/asm/xenoprof.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/asm/xenoprof.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,48 @@
-+/******************************************************************************
-+ * asm-i386/mach-xen/asm/xenoprof.h
-+ *
-+ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
-+ * VA Linux Systems Japan K.K.
++#define LOWMEMSIZE() (0x9f000)
++
++#ifndef __ASSEMBLY__
++struct e820entry {
++ u64 addr; /* start of memory segment */
++ u64 size; /* size of memory segment */
++ u32 type; /* type of memory segment */
++} __attribute__((packed));
++
++struct e820map {
++ int nr_map;
++ struct e820entry map[E820MAX];
++};
++
++extern unsigned long find_e820_area(unsigned long start, unsigned long end,
++ unsigned size);
++extern void add_memory_region(unsigned long start, unsigned long size,
++ int type);
++extern void setup_memory_region(void);
++extern void contig_e820_setup(void);
++extern unsigned long e820_end_of_ram(void);
++extern void e820_reserve_resources(struct e820entry *e820, int nr_map);
++extern void e820_print_map(char *who);
++extern int e820_any_mapped(unsigned long start, unsigned long end, unsigned type);
++extern int e820_all_mapped(unsigned long start, unsigned long end, unsigned type);
++
++extern void e820_bootmem_free(pg_data_t *pgdat, unsigned long start,unsigned long end);
++extern void e820_setup_gap(struct e820entry *e820, int nr_map);
++extern unsigned long e820_hole_size(unsigned long start_pfn,
++ unsigned long end_pfn);
++
++extern void __init parse_memopt(char *p, char **end);
++extern void __init parse_memmapopt(char *p, char **end);
++
++extern struct e820map e820;
++
++extern unsigned ebda_addr, ebda_size;
++#endif/*!__ASSEMBLY__*/
++
++#endif/*__E820_HEADER*/
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/fixmap.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/fixmap.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,112 @@
++/*
++ * fixmap.h: compile-time virtual memory allocation
+ *
-+ * This program is free software; you can redistribute it and/or modify
-+ * it under the terms of the GNU General Public License as published by
-+ * the Free Software Foundation; either version 2 of the License, or
-+ * (at your option) any later version.
++ * This file is subject to the terms and conditions of the GNU General Public
++ * License. See the file "COPYING" in the main directory of this archive
++ * for more details.
+ *
-+ * This program is distributed in the hope that it will be useful,
-+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ * GNU General Public License for more details.
++ * Copyright (C) 1998 Ingo Molnar
++ */
++
++#ifndef _ASM_FIXMAP_H
++#define _ASM_FIXMAP_H
++
++#include <linux/kernel.h>
++#include <asm/apicdef.h>
++#include <asm/page.h>
++#include <asm/vsyscall.h>
++#include <asm/vsyscall32.h>
++#include <asm/acpi.h>
++
++/*
++ * Here we define all the compile-time 'special' virtual
++ * addresses. The point is to have a constant address at
++ * compile time, but to set the physical address only
++ * in the boot process.
+ *
-+ * You should have received a copy of the GNU General Public License
-+ * along with this program; if not, write to the Free Software
-+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ * these 'compile-time allocated' memory buffers are
++ * fixed-size 4k pages. (or larger if used with an increment
++ * highger than 1) use fixmap_set(idx,phys) to associate
++ * physical memory with fixmap indices.
+ *
++ * TLB entries of such buffers will not be flushed across
++ * task switches.
+ */
-+#ifndef __ASM_XENOPROF_H__
-+#define __ASM_XENOPROF_H__
-+#ifdef CONFIG_XEN
+
-+struct super_block;
-+struct dentry;
-+int xenoprof_create_files(struct super_block * sb, struct dentry * root);
-+#define HAVE_XENOPROF_CREATE_FILES
++enum fixed_addresses {
++ VSYSCALL_LAST_PAGE,
++ VSYSCALL_FIRST_PAGE = VSYSCALL_LAST_PAGE + ((VSYSCALL_END-VSYSCALL_START) >> PAGE_SHIFT) - 1,
++ VSYSCALL_HPET,
++ FIX_HPET_BASE,
++#ifdef CONFIG_X86_LOCAL_APIC
++ FIX_APIC_BASE, /* local (CPU) APIC) -- required for SMP or not */
++#endif
++#ifdef CONFIG_X86_IO_APIC
++ FIX_IO_APIC_BASE_0,
++ FIX_IO_APIC_BASE_END = FIX_IO_APIC_BASE_0 + MAX_IO_APICS-1,
++#endif
++#ifdef CONFIG_ACPI
++ FIX_ACPI_BEGIN,
++ FIX_ACPI_END = FIX_ACPI_BEGIN + FIX_ACPI_PAGES - 1,
++#endif
++ FIX_SHARED_INFO,
++#define NR_FIX_ISAMAPS 256
++ FIX_ISAMAP_END,
++ FIX_ISAMAP_BEGIN = FIX_ISAMAP_END + NR_FIX_ISAMAPS - 1,
++ __end_of_permanent_fixed_addresses,
++ /* temporary boot-time mappings, used before ioremap() is functional */
++#define NR_FIX_BTMAPS 16
++ FIX_BTMAP_END = __end_of_permanent_fixed_addresses,
++ FIX_BTMAP_BEGIN = FIX_BTMAP_END + NR_FIX_BTMAPS - 1,
++ __end_of_fixed_addresses
++};
++
++extern void __set_fixmap (enum fixed_addresses idx,
++ unsigned long phys, pgprot_t flags);
++
++#define set_fixmap(idx, phys) \
++ __set_fixmap(idx, phys, PAGE_KERNEL)
++/*
++ * Some hardware wants to get fixmapped without caching.
++ */
++#define set_fixmap_nocache(idx, phys) \
++ __set_fixmap(idx, phys, PAGE_KERNEL_NOCACHE)
++
++#define clear_fixmap(idx) \
++ __set_fixmap(idx, 0, __pgprot(0))
++
++#define FIXADDR_TOP (VSYSCALL_END-PAGE_SIZE)
++#define FIXADDR_SIZE (__end_of_fixed_addresses << PAGE_SHIFT)
++#define FIXADDR_START (FIXADDR_TOP - FIXADDR_SIZE)
++
++/* Only covers 32bit vsyscalls currently. Need another set for 64bit. */
++#define FIXADDR_USER_START ((unsigned long)VSYSCALL32_VSYSCALL)
++#define FIXADDR_USER_END (FIXADDR_USER_START + PAGE_SIZE)
++
++#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT))
++
++extern void __this_fixmap_does_not_exist(void);
++
++/*
++ * 'index to address' translation. If anyone tries to use the idx
++ * directly without translation, we catch the bug with a NULL-deference
++ * kernel oops. Illegal ranges of incoming indices are caught too.
++ */
++static __always_inline unsigned long fix_to_virt(const unsigned int idx)
++{
++ /*
++ * this branch gets completely eliminated after inlining,
++ * except when someone tries to use fixaddr indices in an
++ * illegal way. (such as mixing up address types or using
++ * out-of-range indices).
++ *
++ * If it doesn't get removed, the linker will complain
++ * loudly with a reasonably clear error message..
++ */
++ if (idx >= __end_of_fixed_addresses)
++ __this_fixmap_does_not_exist();
+
-+struct xenoprof_init;
-+void xenoprof_arch_init_counter(struct xenoprof_init *init);
-+void xenoprof_arch_counter(void);
-+void xenoprof_arch_start(void);
-+void xenoprof_arch_stop(void);
++ return __fix_to_virt(idx);
++}
+
-+struct xenoprof_arch_shared_buffer {
-+ /* nothing */
-+};
-+struct xenoprof_shared_buffer;
-+void xenoprof_arch_unmap_shared_buffer(struct xenoprof_shared_buffer* sbuf);
-+struct xenoprof_get_buffer;
-+int xenoprof_arch_map_shared_buffer(struct xenoprof_get_buffer* get_buffer, struct xenoprof_shared_buffer* sbuf);
-+struct xenoprof_passive;
-+int xenoprof_arch_set_passive(struct xenoprof_passive* pdomain, struct xenoprof_shared_buffer* sbuf);
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/gnttab_dma.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/gnttab_dma.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1 @@
++#include <asm-i386/mach-xen/asm/gnttab_dma.h>
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/hw_irq.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/hw_irq.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,136 @@
++#ifndef _ASM_HW_IRQ_H
++#define _ASM_HW_IRQ_H
+
-+#endif /* CONFIG_XEN */
-+#endif /* __ASM_XENOPROF_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/irq_vectors.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/irq_vectors.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/irq_vectors.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/irq_vectors.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,125 @@
+/*
-+ * This file should contain #defines for all of the interrupt vector
-+ * numbers used by this architecture.
-+ *
-+ * In addition, there are some standard defines:
-+ *
-+ * FIRST_EXTERNAL_VECTOR:
-+ * The first free place for external interrupts
++ * linux/include/asm/hw_irq.h
+ *
-+ * SYSCALL_VECTOR:
-+ * The IRQ vector a syscall makes the user to kernel transition
-+ * under.
++ * (C) 1992, 1993 Linus Torvalds, (C) 1997 Ingo Molnar
+ *
-+ * TIMER_IRQ:
-+ * The IRQ number the timer interrupt comes in at.
++ * moved some of the old arch/i386/kernel/irq.h to here. VY
+ *
-+ * NR_IRQS:
-+ * The total number of interrupt vectors (including all the
-+ * architecture specific interrupts) needed.
++ * IRQ/IPI changes taken from work by Thomas Radke
++ * <tomsoft at informatik.tu-chemnitz.de>
+ *
-+ */
-+#ifndef _ASM_IRQ_VECTORS_H
-+#define _ASM_IRQ_VECTORS_H
++ * hacked by Andi Kleen for x86-64.
++ */
++
++#ifndef __ASSEMBLY__
++#include <asm/atomic.h>
++#include <asm/irq.h>
++#include <linux/profile.h>
++#include <linux/smp.h>
++
++struct hw_interrupt_type;
++#endif
+
++#define NMI_VECTOR 0x02
+/*
+ * IDT vectors usable for external interrupt sources start
+ * at 0x20:
+ */
+#define FIRST_EXTERNAL_VECTOR 0x20
+
-+#define SYSCALL_VECTOR 0x80
++#define IA32_SYSCALL_VECTOR 0x80
++
+
+/*
+ * Vectors 0x20-0x2f are used for ISA interrupts.
+ */
+
-+#if 0
+/*
+ * Special IRQ vectors used by the SMP architecture, 0xf0-0xff
+ *
+ * some of the following vectors are 'rare', they are merged
+ * into a single vector (CALL_FUNCTION_VECTOR) to save vector space.
+ * TLB, reschedule and local APIC vectors are performance-critical.
-+ *
-+ * Vectors 0xf0-0xfa are free (reserved for future Linux use).
+ */
++#ifndef CONFIG_XEN
+#define SPURIOUS_APIC_VECTOR 0xff
+#define ERROR_APIC_VECTOR 0xfe
-+#define INVALIDATE_TLB_VECTOR 0xfd
-+#define RESCHEDULE_VECTOR 0xfc
-+#define CALL_FUNCTION_VECTOR 0xfb
++#define RESCHEDULE_VECTOR 0xfd
++#define CALL_FUNCTION_VECTOR 0xfc
++/* fb free - please don't readd KDB here because it's useless
++ (hint - think what a NMI bit does to a vector) */
++#define THERMAL_APIC_VECTOR 0xfa
++#define THRESHOLD_APIC_VECTOR 0xf9
++/* f8 free */
++#define INVALIDATE_TLB_VECTOR_END 0xf7
++#define INVALIDATE_TLB_VECTOR_START 0xf0 /* f0-f7 used for TLB flush */
++
++#define NUM_INVALIDATE_TLB_VECTORS 8
++#endif
+
-+#define THERMAL_APIC_VECTOR 0xf0
+/*
+ * Local APIC timer IRQ vector is on a different priority level,
+ * to work around the 'lost local interrupt if more than 2 IRQ
+ * sources per level' errata.
+ */
+#define LOCAL_TIMER_VECTOR 0xef
-+#endif
-+
-+#define SPURIOUS_APIC_VECTOR 0xff
-+#define ERROR_APIC_VECTOR 0xfe
+
+/*
+ * First APIC vector available to drivers: (vectors 0x30-0xee)
@@ -86316,524 +125603,71 @@
+ * levels. (0x80 is the syscall vector)
+ */
+#define FIRST_DEVICE_VECTOR 0x31
-+#define FIRST_SYSTEM_VECTOR 0xef
-+
-+/*
-+ * 16 8259A IRQ's, 208 potential APIC interrupt sources.
-+ * Right now the APIC is mostly only used for SMP.
-+ * 256 vectors is an architectural limit. (we can have
-+ * more than 256 devices theoretically, but they will
-+ * have to use shared interrupts)
-+ * Since vectors 0x00-0x1f are used/reserved for the CPU,
-+ * the usable vector space is 0x20-0xff (224 vectors)
-+ */
-+
-+#define RESCHEDULE_VECTOR 0
-+#define CALL_FUNCTION_VECTOR 1
-+#define NR_IPIS 2
-+
-+/*
-+ * The maximum number of vectors supported by i386 processors
-+ * is limited to 256. For processors other than i386, NR_VECTORS
-+ * should be changed accordingly.
-+ */
-+#define NR_VECTORS 256
-+
-+#define FPU_IRQ 13
-+
-+#define FIRST_VM86_IRQ 3
-+#define LAST_VM86_IRQ 15
-+#define invalid_vm86_irq(irq) ((irq) < 3 || (irq) > 15)
-+
-+/*
-+ * The flat IRQ space is divided into two regions:
-+ * 1. A one-to-one mapping of real physical IRQs. This space is only used
-+ * if we have physical device-access privilege. This region is at the
-+ * start of the IRQ space so that existing device drivers do not need
-+ * to be modified to translate physical IRQ numbers into our IRQ space.
-+ * 3. A dynamic mapping of inter-domain and Xen-sourced virtual IRQs. These
-+ * are bound using the provided bind/unbind functions.
-+ */
-+
-+#define PIRQ_BASE 0
-+#define NR_PIRQS 256
-+
-+#define DYNIRQ_BASE (PIRQ_BASE + NR_PIRQS)
-+#define NR_DYNIRQS 256
-+
-+#define NR_IRQS (NR_PIRQS + NR_DYNIRQS)
-+#define NR_IRQ_VECTORS NR_IRQS
++#define FIRST_SYSTEM_VECTOR 0xef /* duplicated in irq.h */
+
-+#define pirq_to_irq(_x) ((_x) + PIRQ_BASE)
-+#define irq_to_pirq(_x) ((_x) - PIRQ_BASE)
+
-+#define dynirq_to_irq(_x) ((_x) + DYNIRQ_BASE)
-+#define irq_to_dynirq(_x) ((_x) - DYNIRQ_BASE)
++#ifndef __ASSEMBLY__
++extern u8 irq_vector[NR_IRQ_VECTORS];
++#define IO_APIC_VECTOR(irq) (irq_vector[irq])
++#define AUTO_ASSIGN -1
+
-+#endif /* _ASM_IRQ_VECTORS_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/mach_traps.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/mach_traps.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/mach_traps.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/mach_traps.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,33 @@
+/*
-+ * include/asm-xen/asm-i386/mach-xen/mach_traps.h
++ * Various low-level irq details needed by irq.c, process.c,
++ * time.c, io_apic.c and smp.c
+ *
-+ * Machine specific NMI handling for Xen
++ * Interrupt entry/exit code at both C and assembly level
+ */
-+#ifndef _MACH_TRAPS_H
-+#define _MACH_TRAPS_H
-+
-+#include <linux/bitops.h>
-+#include <xen/interface/nmi.h>
-+
-+static inline void clear_mem_error(unsigned char reason) {}
-+static inline void clear_io_check_error(unsigned char reason) {}
-+
-+static inline unsigned char get_nmi_reason(void)
-+{
-+ shared_info_t *s = HYPERVISOR_shared_info;
-+ unsigned char reason = 0;
-+
-+ /* construct a value which looks like it came from
-+ * port 0x61.
-+ */
-+ if (test_bit(_XEN_NMIREASON_io_error, &s->arch.nmi_reason))
-+ reason |= 0x40;
-+ if (test_bit(_XEN_NMIREASON_parity_error, &s->arch.nmi_reason))
-+ reason |= 0x80;
-+
-+ return reason;
-+}
-+
-+static inline void reassert_nmi(void) {}
-+
-+#endif /* !_MACH_TRAPS_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/mach-xen/setup_arch.h tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/setup_arch.h
---- pristine-linux-2.6.18.2/include/asm-i386/mach-xen/setup_arch.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/mach-xen/setup_arch.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,5 @@
-+/* Hook to call BIOS initialisation function */
-+
-+#define ARCH_SETUP machine_specific_arch_setup();
-+
-+void __init machine_specific_arch_setup(void);
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/page.h tmp-linux-2.6-xen.patch/include/asm-i386/page.h
---- pristine-linux-2.6.18.2/include/asm-i386/page.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/page.h 2007-10-14 01:51:15.000000000 +0200
-@@ -122,7 +122,7 @@ extern int page_is_ram(unsigned long pag
-
- #define PAGE_OFFSET ((unsigned long)__PAGE_OFFSET)
- #define VMALLOC_RESERVE ((unsigned long)__VMALLOC_RESERVE)
--#define MAXMEM (-__PAGE_OFFSET-__VMALLOC_RESERVE)
-+#define MAXMEM (__FIXADDR_TOP-__PAGE_OFFSET-__VMALLOC_RESERVE)
- #define __pa(x) ((unsigned long)(x)-PAGE_OFFSET)
- #define __va(x) ((void *)((unsigned long)(x)+PAGE_OFFSET))
- #define pfn_to_kaddr(pfn) __va((pfn) << PAGE_SHIFT)
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/pgtable-2level-defs.h tmp-linux-2.6-xen.patch/include/asm-i386/pgtable-2level-defs.h
---- pristine-linux-2.6.18.2/include/asm-i386/pgtable-2level-defs.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/pgtable-2level-defs.h 2007-10-14 01:51:15.000000000 +0200
-@@ -1,6 +1,8 @@
- #ifndef _I386_PGTABLE_2LEVEL_DEFS_H
- #define _I386_PGTABLE_2LEVEL_DEFS_H
-
-+#define HAVE_SHARED_KERNEL_PMD 0
-+
- /*
- * traditional i386 two-level paging structure:
- */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-i386/pgtable-3level-defs.h tmp-linux-2.6-xen.patch/include/asm-i386/pgtable-3level-defs.h
---- pristine-linux-2.6.18.2/include/asm-i386/pgtable-3level-defs.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-i386/pgtable-3level-defs.h 2007-10-14 01:51:15.000000000 +0200
-@@ -1,6 +1,8 @@
- #ifndef _I386_PGTABLE_3LEVEL_DEFS_H
- #define _I386_PGTABLE_3LEVEL_DEFS_H
-
-+#define HAVE_SHARED_KERNEL_PMD 1
-+
- /*
- * PGDIR_SHIFT determines what a top-level page table entry can map
- */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/agp.h tmp-linux-2.6-xen.patch/include/asm-ia64/agp.h
---- pristine-linux-2.6.18.2/include/asm-ia64/agp.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/agp.h 2007-07-30 16:35:13.000000000 +0200
-@@ -19,13 +19,44 @@
- #define flush_agp_cache() mb()
-
- /* Convert a physical address to an address suitable for the GART. */
-+#ifndef CONFIG_XEN
- #define phys_to_gart(x) (x)
- #define gart_to_phys(x) (x)
-+#else
-+#define phys_to_gart(x) phys_to_machine_for_dma(x)
-+#define gart_to_phys(x) machine_to_phys_for_dma(x)
-+#endif
-
- /* GATT allocation. Returns/accepts GATT kernel virtual address. */
-+#ifndef CONFIG_XEN
- #define alloc_gatt_pages(order) \
- ((char *)__get_free_pages(GFP_KERNEL, (order)))
- #define free_gatt_pages(table, order) \
- free_pages((unsigned long)(table), (order))
-+#else
-+#include <asm/hypervisor.h>
-+static inline char*
-+alloc_gatt_pages(unsigned int order)
-+{
-+ unsigned long error;
-+ unsigned long ret = __get_free_pages(GFP_KERNEL, (order));
-+ if (ret == 0) {
-+ goto out;
-+ }
-+ error = xen_create_contiguous_region(ret, order, 0);
-+ if (error) {
-+ free_pages(ret, order);
-+ ret = 0;
-+ }
-+out:
-+ return (char*)ret;
-+}
-+static inline void
-+free_gatt_pages(void* table, unsigned int order)
-+{
-+ xen_destroy_contiguous_region((unsigned long)table, order);
-+ free_pages((unsigned long)table, order);
-+}
-+#endif /* CONFIG_XEN */
-
- #endif /* _ASM_IA64_AGP_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/dma-mapping.h tmp-linux-2.6-xen.patch/include/asm-ia64/dma-mapping.h
---- pristine-linux-2.6.18.2/include/asm-ia64/dma-mapping.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/dma-mapping.h 2007-07-30 16:35:13.000000000 +0200
-@@ -6,7 +6,14 @@
- * David Mosberger-Tang <davidm at hpl.hp.com>
- */
- #include <asm/machvec.h>
-+#ifdef CONFIG_XEN
-+/* Needed for arch/i386/kernel/swiotlb.c and arch/i386/kernel/pci-dma-xen.c */
-+#include <asm/hypervisor.h>
-+/* Needed for arch/i386/kernel/swiotlb.c */
-+#include <asm/swiotlb.h>
-+#endif
-
-+#ifndef CONFIG_XEN
- #define dma_alloc_coherent platform_dma_alloc_coherent
- #define dma_alloc_noncoherent platform_dma_alloc_coherent /* coherent mem. is cheap */
- #define dma_free_coherent platform_dma_free_coherent
-@@ -20,6 +27,46 @@
- #define dma_sync_single_for_device platform_dma_sync_single_for_device
- #define dma_sync_sg_for_device platform_dma_sync_sg_for_device
- #define dma_mapping_error platform_dma_mapping_error
-+#else
-+int dma_map_sg(struct device *hwdev, struct scatterlist *sg, int nents,
-+ enum dma_data_direction direction);
-+void dma_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nents,
-+ enum dma_data_direction direction);
-+int dma_supported(struct device *dev, u64 mask);
-+void *dma_alloc_coherent(struct device *dev, size_t size,
-+ dma_addr_t *dma_handle, gfp_t gfp);
-+void dma_free_coherent(struct device *dev, size_t size, void *vaddr,
-+ dma_addr_t dma_handle);
-+dma_addr_t dma_map_single(struct device *dev, void *ptr, size_t size,
-+ enum dma_data_direction direction);
-+void dma_unmap_single(struct device *dev, dma_addr_t dma_addr, size_t size,
-+ enum dma_data_direction direction);
-+void dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle,
-+ size_t size, enum dma_data_direction direction);
-+void dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle,
-+ size_t size,
-+ enum dma_data_direction direction);
-+int dma_mapping_error(dma_addr_t dma_addr);
+
-+#define flush_write_buffers() do { } while (0)
-+static inline void
-+dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg, int nelems,
-+ enum dma_data_direction direction)
-+{
-+ if (swiotlb)
-+ swiotlb_sync_sg_for_cpu(dev,sg,nelems,direction);
-+ flush_write_buffers();
-+}
++extern void disable_8259A_irq(unsigned int irq);
++extern void enable_8259A_irq(unsigned int irq);
++extern int i8259A_irq_pending(unsigned int irq);
++extern void make_8259A_irq(unsigned int irq);
++extern void init_8259A(int aeoi);
++extern void FASTCALL(send_IPI_self(int vector));
++extern void init_VISWS_APIC_irqs(void);
++extern void setup_IO_APIC(void);
++extern void disable_IO_APIC(void);
++#define print_IO_APIC()
++extern int IO_APIC_get_PCI_irq_vector(int bus, int slot, int fn);
++extern void send_IPI(int dest, int vector);
++extern void setup_ioapic_dest(void);
+
-+static inline void
-+dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg, int nelems,
-+ enum dma_data_direction direction)
-+{
-+ if (swiotlb)
-+ swiotlb_sync_sg_for_device(dev,sg,nelems,direction);
-+ flush_write_buffers();
-+}
-+#endif
-
- #define dma_map_page(dev, pg, off, size, dir) \
- dma_map_single(dev, page_address(pg) + (off), (size), (dir))
-@@ -36,7 +83,9 @@
- #define dma_sync_single_range_for_device(dev, dma_handle, offset, size, dir) \
- dma_sync_single_for_device(dev, dma_handle, size, dir)
-
-+#ifndef CONFIG_XEN
- #define dma_supported platform_dma_supported
-+#endif
-
- static inline int
- dma_set_mask (struct device *dev, u64 mask)
-@@ -61,4 +110,27 @@ dma_cache_sync (void *vaddr, size_t size
-
- #define dma_is_consistent(dma_handle) (1) /* all we do is coherent memory... */
-
-+#ifdef CONFIG_XEN
-+/* arch/i386/kernel/swiotlb.o requires */
-+void contiguous_bitmap_init(unsigned long end_pfn);
++extern unsigned long io_apic_irqs;
+
-+static inline int
-+address_needs_mapping(struct device *hwdev, dma_addr_t addr)
-+{
-+ dma_addr_t mask = DMA_64BIT_MASK;
-+ /* If the device has a mask, use it, otherwise default to 64 bits */
-+ if (hwdev && hwdev->dma_mask)
-+ mask = *hwdev->dma_mask;
-+ return (addr & ~mask) != 0;
-+}
++extern atomic_t irq_err_count;
++extern atomic_t irq_mis_count;
+
-+static inline int
-+range_straddles_page_boundary(void *p, size_t size)
-+{
-+ extern unsigned long *contiguous_bitmap;
-+ return (((((unsigned long)p & ~PAGE_MASK) + size) > PAGE_SIZE) &&
-+ !test_bit(__pa(p) >> PAGE_SHIFT, contiguous_bitmap));
-+}
-+#endif
++#define IO_APIC_IRQ(x) (((x) >= 16) || ((1<<(x)) & io_apic_irqs))
+
- #endif /* _ASM_IA64_DMA_MAPPING_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/fixmap.h tmp-linux-2.6-xen.patch/include/asm-ia64/fixmap.h
---- pristine-linux-2.6.18.2/include/asm-ia64/fixmap.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/fixmap.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,2 @@
-+#define clear_fixmap(x) do {} while (0)
-+#define set_fixmap(x,y) do {} while (0)
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/gcc_intrin.h tmp-linux-2.6-xen.patch/include/asm-ia64/gcc_intrin.h
---- pristine-linux-2.6.18.2/include/asm-ia64/gcc_intrin.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/gcc_intrin.h 2007-07-30 16:35:13.000000000 +0200
-@@ -26,7 +26,7 @@ extern void ia64_bad_param_for_getreg (v
-
- register unsigned long ia64_r13 asm ("r13") __attribute_used__;
-
--#define ia64_setreg(regnum, val) \
-+#define __ia64_setreg(regnum, val) \
- ({ \
- switch (regnum) { \
- case _IA64_REG_PSR_L: \
-@@ -55,7 +55,7 @@ register unsigned long ia64_r13 asm ("r1
- } \
- })
-
--#define ia64_getreg(regnum) \
-+#define __ia64_getreg(regnum) \
- ({ \
- __u64 ia64_intri_res; \
- \
-@@ -92,7 +92,7 @@ register unsigned long ia64_r13 asm ("r1
-
- #define ia64_hint_pause 0
-
--#define ia64_hint(mode) \
-+#define __ia64_hint(mode) \
- ({ \
- switch (mode) { \
- case ia64_hint_pause: \
-@@ -374,7 +374,7 @@ register unsigned long ia64_r13 asm ("r1
-
- #define ia64_invala() asm volatile ("invala" ::: "memory")
-
--#define ia64_thash(addr) \
-+#define __ia64_thash(addr) \
- ({ \
- __u64 ia64_intri_res; \
- asm volatile ("thash %0=%1" : "=r"(ia64_intri_res) : "r" (addr)); \
-@@ -394,18 +394,18 @@ register unsigned long ia64_r13 asm ("r1
-
- #define ia64_nop(x) asm volatile ("nop %0"::"i"(x));
-
--#define ia64_itci(addr) asm volatile ("itc.i %0;;" :: "r"(addr) : "memory")
-+#define __ia64_itci(addr) asm volatile ("itc.i %0;;" :: "r"(addr) : "memory")
-
--#define ia64_itcd(addr) asm volatile ("itc.d %0;;" :: "r"(addr) : "memory")
-+#define __ia64_itcd(addr) asm volatile ("itc.d %0;;" :: "r"(addr) : "memory")
-
-
--#define ia64_itri(trnum, addr) asm volatile ("itr.i itr[%0]=%1" \
-+#define __ia64_itri(trnum, addr) asm volatile ("itr.i itr[%0]=%1" \
- :: "r"(trnum), "r"(addr) : "memory")
-
--#define ia64_itrd(trnum, addr) asm volatile ("itr.d dtr[%0]=%1" \
-+#define __ia64_itrd(trnum, addr) asm volatile ("itr.d dtr[%0]=%1" \
- :: "r"(trnum), "r"(addr) : "memory")
-
--#define ia64_tpa(addr) \
-+#define __ia64_tpa(addr) \
- ({ \
- __u64 ia64_pa; \
- asm volatile ("tpa %0 = %1" : "=r"(ia64_pa) : "r"(addr) : "memory"); \
-@@ -415,22 +415,22 @@ register unsigned long ia64_r13 asm ("r1
- #define __ia64_set_dbr(index, val) \
- asm volatile ("mov dbr[%0]=%1" :: "r"(index), "r"(val) : "memory")
-
--#define ia64_set_ibr(index, val) \
-+#define __ia64_set_ibr(index, val) \
- asm volatile ("mov ibr[%0]=%1" :: "r"(index), "r"(val) : "memory")
-
--#define ia64_set_pkr(index, val) \
-+#define __ia64_set_pkr(index, val) \
- asm volatile ("mov pkr[%0]=%1" :: "r"(index), "r"(val) : "memory")
-
--#define ia64_set_pmc(index, val) \
-+#define __ia64_set_pmc(index, val) \
- asm volatile ("mov pmc[%0]=%1" :: "r"(index), "r"(val) : "memory")
-
--#define ia64_set_pmd(index, val) \
-+#define __ia64_set_pmd(index, val) \
- asm volatile ("mov pmd[%0]=%1" :: "r"(index), "r"(val) : "memory")
-
--#define ia64_set_rr(index, val) \
-+#define __ia64_set_rr(index, val) \
- asm volatile ("mov rr[%0]=%1" :: "r"(index), "r"(val) : "memory");
-
--#define ia64_get_cpuid(index) \
-+#define __ia64_get_cpuid(index) \
- ({ \
- __u64 ia64_intri_res; \
- asm volatile ("mov %0=cpuid[%r1]" : "=r"(ia64_intri_res) : "rO"(index)); \
-@@ -444,21 +444,21 @@ register unsigned long ia64_r13 asm ("r1
- ia64_intri_res; \
- })
-
--#define ia64_get_ibr(index) \
-+#define __ia64_get_ibr(index) \
- ({ \
- __u64 ia64_intri_res; \
- asm volatile ("mov %0=ibr[%1]" : "=r"(ia64_intri_res) : "r"(index)); \
- ia64_intri_res; \
- })
-
--#define ia64_get_pkr(index) \
-+#define __ia64_get_pkr(index) \
- ({ \
- __u64 ia64_intri_res; \
- asm volatile ("mov %0=pkr[%1]" : "=r"(ia64_intri_res) : "r"(index)); \
- ia64_intri_res; \
- })
-
--#define ia64_get_pmc(index) \
-+#define __ia64_get_pmc(index) \
- ({ \
- __u64 ia64_intri_res; \
- asm volatile ("mov %0=pmc[%1]" : "=r"(ia64_intri_res) : "r"(index)); \
-@@ -466,48 +466,48 @@ register unsigned long ia64_r13 asm ("r1
- })
-
-
--#define ia64_get_pmd(index) \
-+#define __ia64_get_pmd(index) \
- ({ \
- __u64 ia64_intri_res; \
- asm volatile ("mov %0=pmd[%1]" : "=r"(ia64_intri_res) : "r"(index)); \
- ia64_intri_res; \
- })
-
--#define ia64_get_rr(index) \
-+#define __ia64_get_rr(index) \
- ({ \
- __u64 ia64_intri_res; \
- asm volatile ("mov %0=rr[%1]" : "=r"(ia64_intri_res) : "r" (index)); \
- ia64_intri_res; \
- })
-
--#define ia64_fc(addr) asm volatile ("fc %0" :: "r"(addr) : "memory")
-+#define __ia64_fc(addr) asm volatile ("fc %0" :: "r"(addr) : "memory")
-
-
- #define ia64_sync_i() asm volatile (";; sync.i" ::: "memory")
-
--#define ia64_ssm(mask) asm volatile ("ssm %0":: "i"((mask)) : "memory")
--#define ia64_rsm(mask) asm volatile ("rsm %0":: "i"((mask)) : "memory")
-+#define __ia64_ssm(mask) asm volatile ("ssm %0":: "i"((mask)) : "memory")
-+#define __ia64_rsm(mask) asm volatile ("rsm %0":: "i"((mask)) : "memory")
- #define ia64_sum(mask) asm volatile ("sum %0":: "i"((mask)) : "memory")
- #define ia64_rum(mask) asm volatile ("rum %0":: "i"((mask)) : "memory")
-
--#define ia64_ptce(addr) asm volatile ("ptc.e %0" :: "r"(addr))
-+#define __ia64_ptce(addr) asm volatile ("ptc.e %0" :: "r"(addr))
-
--#define ia64_ptcga(addr, size) \
-+#define __ia64_ptcga(addr, size) \
- do { \
- asm volatile ("ptc.ga %0,%1" :: "r"(addr), "r"(size) : "memory"); \
- ia64_dv_serialize_data(); \
- } while (0)
-
--#define ia64_ptcl(addr, size) \
-+#define __ia64_ptcl(addr, size) \
- do { \
- asm volatile ("ptc.l %0,%1" :: "r"(addr), "r"(size) : "memory"); \
- ia64_dv_serialize_data(); \
- } while (0)
-
--#define ia64_ptri(addr, size) \
-+#define __ia64_ptri(addr, size) \
- asm volatile ("ptr.i %0,%1" :: "r"(addr), "r"(size) : "memory")
-
--#define ia64_ptrd(addr, size) \
-+#define __ia64_ptrd(addr, size) \
- asm volatile ("ptr.d %0,%1" :: "r"(addr), "r"(size) : "memory")
-
- /* Values for lfhint in ia64_lfetch and ia64_lfetch_fault */
-@@ -589,7 +589,7 @@ do { \
- } \
- })
-
--#define ia64_intrin_local_irq_restore(x) \
-+#define __ia64_intrin_local_irq_restore(x) \
- do { \
- asm volatile (";; cmp.ne p6,p7=%0,r0;;" \
- "(p6) ssm psr.i;" \
-@@ -598,4 +598,6 @@ do { \
- :: "r"((x)) : "p6", "p7", "memory"); \
- } while (0)
-
-+#define __ia64_get_psr_i() (__ia64_getreg(_IA64_REG_PSR) & 0x4000UL)
++#define __STR(x) #x
++#define STR(x) __STR(x)
++
++#include <asm/ptrace.h>
++
++#define IRQ_NAME2(nr) nr##_interrupt(void)
++#define IRQ_NAME(nr) IRQ_NAME2(IRQ##nr)
++
++/*
++ * SMP has a few special interrupts for IPI messages
++ */
++
++#define BUILD_IRQ(nr) \
++asmlinkage void IRQ_NAME(nr); \
++__asm__( \
++"\n.p2align\n" \
++"IRQ" #nr "_interrupt:\n\t" \
++ "push $~(" #nr ") ; " \
++ "jmp common_interrupt");
++
++#define platform_legacy_irq(irq) ((irq) < 16)
+
- #endif /* _ASM_IA64_GCC_INTRIN_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/hw_irq.h tmp-linux-2.6-xen.patch/include/asm-ia64/hw_irq.h
---- pristine-linux-2.6.18.2/include/asm-ia64/hw_irq.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/hw_irq.h 2007-07-30 16:35:13.000000000 +0200
-@@ -15,7 +15,11 @@
- #include <asm/ptrace.h>
- #include <asm/smp.h>
-
-+#ifndef CONFIG_XEN
- typedef u8 ia64_vector;
-+#else
-+typedef u16 ia64_vector;
+#endif
-
- /*
- * 0 special
-@@ -99,6 +103,12 @@ extern void register_percpu_irq (ia64_ve
-
- static inline void ia64_resend_irq(unsigned int vector)
- {
-+#ifdef CONFIG_XEN
-+ extern int resend_irq_on_evtchn(unsigned int i);
-+ if (is_running_on_xen())
-+ resend_irq_on_evtchn(vector);
-+ else
-+#endif /* CONFIG_XEN */
- platform_send_ipi(smp_processor_id(), vector, IA64_IPI_DM_INT, 0);
- }
-
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/hypercall.h tmp-linux-2.6-xen.patch/include/asm-ia64/hypercall.h
---- pristine-linux-2.6.18.2/include/asm-ia64/hypercall.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/hypercall.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,425 @@
++
++#endif /* _ASM_HW_IRQ_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/hypercall.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/hypercall.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,415 @@
+/******************************************************************************
+ * hypercall.h
+ *
@@ -86841,6 +125675,10 @@
+ *
+ * Copyright (c) 2002-2004, K A Fraser
+ *
++ * 64-bit updates:
++ * Benjamin Liu <benjamin.liu at intel.com>
++ * Jun Nakajima <jun.nakajima at intel.com>
++ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation; or, when distributed
@@ -86869,9055 +125707,9471 @@
+#ifndef __HYPERCALL_H__
+#define __HYPERCALL_H__
+
++#include <linux/string.h> /* memcpy() */
++#include <linux/stringify.h>
++
+#ifndef __HYPERVISOR_H__
+# error "please don't include this file directly"
+#endif
+
-+#include <asm/xen/xcom_hcall.h>
-+struct xencomm_handle;
-+extern unsigned long __hypercall(unsigned long a1, unsigned long a2,
-+ unsigned long a3, unsigned long a4,
-+ unsigned long a5, unsigned long cmd);
-+
-+/*
-+ * Assembler stubs for hyper-calls.
-+ */
++#ifdef CONFIG_XEN
++#define HYPERCALL_STR(name) \
++ "call hypercall_page + ("__stringify(__HYPERVISOR_##name)" * 32)"
++#else
++#define HYPERCALL_STR(name) \
++ "mov $("__stringify(__HYPERVISOR_##name)" * 32),%%eax; "\
++ "add hypercall_stubs(%%rip),%%rax; " \
++ "call *%%rax"
++#endif
+
-+#define _hypercall0(type, name) \
-+({ \
-+ long __res; \
-+ __res=__hypercall(0, 0, 0, 0, 0, __HYPERVISOR_##name); \
-+ (type)__res; \
++#define _hypercall0(type, name) \
++({ \
++ type __res; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res) \
++ : \
++ : "memory" ); \
++ __res; \
+})
+
+#define _hypercall1(type, name, a1) \
+({ \
-+ long __res; \
-+ __res = __hypercall((unsigned long)a1, \
-+ 0, 0, 0, 0, __HYPERVISOR_##name); \
-+ (type)__res; \
++ type __res; \
++ long __ign1; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=D" (__ign1) \
++ : "1" ((long)(a1)) \
++ : "memory" ); \
++ __res; \
+})
+
+#define _hypercall2(type, name, a1, a2) \
+({ \
-+ long __res; \
-+ __res = __hypercall((unsigned long)a1, \
-+ (unsigned long)a2, \
-+ 0, 0, 0, __HYPERVISOR_##name); \
-+ (type)__res; \
++ type __res; \
++ long __ign1, __ign2; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=D" (__ign1), "=S" (__ign2) \
++ : "1" ((long)(a1)), "2" ((long)(a2)) \
++ : "memory" ); \
++ __res; \
+})
+
+#define _hypercall3(type, name, a1, a2, a3) \
+({ \
-+ long __res; \
-+ __res = __hypercall((unsigned long)a1, \
-+ (unsigned long)a2, \
-+ (unsigned long)a3, \
-+ 0, 0, __HYPERVISOR_##name); \
-+ (type)__res; \
++ type __res; \
++ long __ign1, __ign2, __ign3; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=D" (__ign1), "=S" (__ign2), \
++ "=d" (__ign3) \
++ : "1" ((long)(a1)), "2" ((long)(a2)), \
++ "3" ((long)(a3)) \
++ : "memory" ); \
++ __res; \
+})
+
+#define _hypercall4(type, name, a1, a2, a3, a4) \
+({ \
-+ long __res; \
-+ __res = __hypercall((unsigned long)a1, \
-+ (unsigned long)a2, \
-+ (unsigned long)a3, \
-+ (unsigned long)a4, \
-+ 0, __HYPERVISOR_##name); \
-+ (type)__res; \
++ type __res; \
++ long __ign1, __ign2, __ign3; \
++ register long __arg4 asm("r10") = (long)(a4); \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=D" (__ign1), "=S" (__ign2), \
++ "=d" (__ign3), "+r" (__arg4) \
++ : "1" ((long)(a1)), "2" ((long)(a2)), \
++ "3" ((long)(a3)) \
++ : "memory" ); \
++ __res; \
+})
+
+#define _hypercall5(type, name, a1, a2, a3, a4, a5) \
+({ \
-+ long __res; \
-+ __res = __hypercall((unsigned long)a1, \
-+ (unsigned long)a2, \
-+ (unsigned long)a3, \
-+ (unsigned long)a4, \
-+ (unsigned long)a5, \
-+ __HYPERVISOR_##name); \
-+ (type)__res; \
++ type __res; \
++ long __ign1, __ign2, __ign3; \
++ register long __arg4 asm("r10") = (long)(a4); \
++ register long __arg5 asm("r8") = (long)(a5); \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=D" (__ign1), "=S" (__ign2), \
++ "=d" (__ign3), "+r" (__arg4), "+r" (__arg5) \
++ : "1" ((long)(a1)), "2" ((long)(a2)), \
++ "3" ((long)(a3)) \
++ : "memory" ); \
++ __res; \
+})
+
-+
-+static inline int
-+xencomm_arch_hypercall_sched_op(int cmd, struct xencomm_handle *arg)
-+{
-+ return _hypercall2(int, sched_op, cmd, arg);
-+}
-+
-+static inline long
-+HYPERVISOR_set_timer_op(u64 timeout)
-+{
-+ unsigned long timeout_hi = (unsigned long)(timeout >> 32);
-+ unsigned long timeout_lo = (unsigned long)timeout;
-+ return _hypercall2(long, set_timer_op, timeout_lo, timeout_hi);
-+}
-+
-+static inline int
-+xencomm_arch_hypercall_platform_op(struct xencomm_handle *op)
-+{
-+ return _hypercall1(int, platform_op, op);
-+}
-+
-+static inline int
-+xencomm_arch_hypercall_sysctl(struct xencomm_handle *op)
-+{
-+ return _hypercall1(int, sysctl, op);
-+}
-+
-+static inline int
-+xencomm_arch_hypercall_domctl(struct xencomm_handle *op)
-+{
-+ return _hypercall1(int, domctl, op);
-+}
-+
-+static inline int
-+xencomm_arch_hypercall_multicall(struct xencomm_handle *call_list,
-+ int nr_calls)
-+{
-+ return _hypercall2(int, multicall, call_list, nr_calls);
-+}
-+
-+static inline int
-+xencomm_arch_hypercall_memory_op(unsigned int cmd, struct xencomm_handle *arg)
-+{
-+ return _hypercall2(int, memory_op, cmd, arg);
-+}
-+
-+static inline int
-+xencomm_arch_hypercall_event_channel_op(int cmd, struct xencomm_handle *arg)
-+{
-+ return _hypercall2(int, event_channel_op, cmd, arg);
-+}
-+
-+static inline int
-+xencomm_arch_hypercall_acm_op(struct xencomm_handle *arg)
-+{
-+ return _hypercall1(int, acm_op, arg);
-+}
-+
-+static inline int
-+xencomm_arch_hypercall_xen_version(int cmd, struct xencomm_handle *arg)
-+{
-+ return _hypercall2(int, xen_version, cmd, arg);
-+}
-+
-+static inline int
-+xencomm_arch_hypercall_console_io(int cmd, int count,
-+ struct xencomm_handle *str)
-+{
-+ return _hypercall3(int, console_io, cmd, count, str);
-+}
-+
-+static inline int
-+xencomm_arch_hypercall_physdev_op(int cmd, struct xencomm_handle *arg)
++static inline int __must_check
++HYPERVISOR_set_trap_table(
++ const trap_info_t *table)
+{
-+ return _hypercall2(int, physdev_op, cmd, arg);
++ return _hypercall1(int, set_trap_table, table);
+}
+
-+static inline int
-+xencomm_arch_hypercall_grant_table_op(unsigned int cmd,
-+ struct xencomm_handle *uop,
-+ unsigned int count)
++static inline int __must_check
++HYPERVISOR_mmu_update(
++ mmu_update_t *req, unsigned int count, unsigned int *success_count,
++ domid_t domid)
+{
-+ return _hypercall3(int, grant_table_op, cmd, uop, count);
++ return _hypercall4(int, mmu_update, req, count, success_count, domid);
+}
+
-+int HYPERVISOR_grant_table_op(unsigned int cmd, void *uop, unsigned int count);
-+
-+extern int xencomm_arch_hypercall_suspend(struct xencomm_handle *arg);
-+
-+static inline int
-+xencomm_arch_hypercall_callback_op(int cmd, struct xencomm_handle *arg)
++static inline int __must_check
++HYPERVISOR_mmuext_op(
++ struct mmuext_op *op, unsigned int count, unsigned int *success_count,
++ domid_t domid)
+{
-+ return _hypercall2(int, callback_op, cmd, arg);
++ return _hypercall4(int, mmuext_op, op, count, success_count, domid);
+}
+
-+static inline unsigned long
-+xencomm_arch_hypercall_hvm_op(int cmd, void *arg)
++static inline int __must_check
++HYPERVISOR_set_gdt(
++ unsigned long *frame_list, unsigned int entries)
+{
-+ return _hypercall2(unsigned long, hvm_op, cmd, arg);
++ return _hypercall2(int, set_gdt, frame_list, entries);
+}
+
-+static inline long
-+xencomm_arch_hypercall_vcpu_op(int cmd, int cpu, void *arg)
++static inline int __must_check
++HYPERVISOR_stack_switch(
++ unsigned long ss, unsigned long esp)
+{
-+ return _hypercall3(long, vcpu_op, cmd, cpu, arg);
++ return _hypercall2(int, stack_switch, ss, esp);
+}
+
-+static inline int
-+HYPERVISOR_physdev_op(int cmd, void *arg)
++static inline int __must_check
++HYPERVISOR_set_callbacks(
++ unsigned long event_address, unsigned long failsafe_address,
++ unsigned long syscall_address)
+{
-+ switch (cmd) {
-+ case PHYSDEVOP_eoi:
-+ return _hypercall1(int, ia64_fast_eoi,
-+ ((struct physdev_eoi *)arg)->irq);
-+ default:
-+ return xencomm_hypercall_physdev_op(cmd, arg);
-+ }
++ return _hypercall3(int, set_callbacks,
++ event_address, failsafe_address, syscall_address);
+}
+
+static inline int
-+xencomm_arch_hypercall_xenoprof_op(int op, struct xencomm_handle *arg)
-+{
-+ return _hypercall2(int, xenoprof_op, op, arg);
-+}
-+
-+extern fastcall unsigned int __do_IRQ(unsigned int irq, struct pt_regs *regs);
-+static inline void exit_idle(void) {}
-+#define do_IRQ(irq, regs) ({ \
-+ irq_enter(); \
-+ __do_IRQ((irq), (regs)); \
-+ irq_exit(); \
-+})
-+
-+#include <linux/err.h>
-+#ifdef CONFIG_XEN
-+#include <asm/xen/privop.h>
-+#endif /* CONFIG_XEN */
-+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-+#include <xen/platform-compat.h>
-+#endif
-+
-+static inline unsigned long
-+__HYPERVISOR_ioremap(unsigned long ioaddr, unsigned long size)
-+{
-+ return _hypercall3(unsigned long, ia64_dom0vp_op,
-+ IA64_DOM0VP_ioremap, ioaddr, size);
-+}
-+
-+static inline unsigned long
-+HYPERVISOR_ioremap(unsigned long ioaddr, unsigned long size)
-+{
-+ unsigned long ret = ioaddr;
-+ if (is_running_on_xen()) {
-+ ret = __HYPERVISOR_ioremap(ioaddr, size);
-+ if (unlikely(ret == -ENOSYS))
-+ panic("hypercall %s failed with %ld. "
-+ "Please check Xen and Linux config mismatch\n",
-+ __func__, -ret);
-+ else if (unlikely(IS_ERR_VALUE(ret)))
-+ ret = ioaddr;
-+ }
-+ return ret;
-+}
-+
-+static inline unsigned long
-+__HYPERVISOR_phystomach(unsigned long gpfn)
-+{
-+ return _hypercall2(unsigned long, ia64_dom0vp_op,
-+ IA64_DOM0VP_phystomach, gpfn);
-+}
-+
-+static inline unsigned long
-+HYPERVISOR_phystomach(unsigned long gpfn)
-+{
-+ unsigned long ret = gpfn;
-+ if (is_running_on_xen()) {
-+ ret = __HYPERVISOR_phystomach(gpfn);
-+ }
-+ return ret;
-+}
-+
-+static inline unsigned long
-+__HYPERVISOR_machtophys(unsigned long mfn)
-+{
-+ return _hypercall2(unsigned long, ia64_dom0vp_op,
-+ IA64_DOM0VP_machtophys, mfn);
-+}
-+
-+static inline unsigned long
-+HYPERVISOR_machtophys(unsigned long mfn)
++HYPERVISOR_fpu_taskswitch(
++ int set)
+{
-+ unsigned long ret = mfn;
-+ if (is_running_on_xen()) {
-+ ret = __HYPERVISOR_machtophys(mfn);
-+ }
-+ return ret;
++ return _hypercall1(int, fpu_taskswitch, set);
+}
+
-+static inline unsigned long
-+__HYPERVISOR_zap_physmap(unsigned long gpfn, unsigned int extent_order)
++static inline int __must_check
++HYPERVISOR_sched_op_compat(
++ int cmd, unsigned long arg)
+{
-+ return _hypercall3(unsigned long, ia64_dom0vp_op,
-+ IA64_DOM0VP_zap_physmap, gpfn, extent_order);
++ return _hypercall2(int, sched_op_compat, cmd, arg);
+}
+
-+static inline unsigned long
-+HYPERVISOR_zap_physmap(unsigned long gpfn, unsigned int extent_order)
++static inline int __must_check
++HYPERVISOR_sched_op(
++ int cmd, void *arg)
+{
-+ unsigned long ret = 0;
-+ if (is_running_on_xen()) {
-+ ret = __HYPERVISOR_zap_physmap(gpfn, extent_order);
-+ }
-+ return ret;
++ return _hypercall2(int, sched_op, cmd, arg);
+}
+
-+static inline unsigned long
-+__HYPERVISOR_add_physmap(unsigned long gpfn, unsigned long mfn,
-+ unsigned long flags, domid_t domid)
++static inline long __must_check
++HYPERVISOR_set_timer_op(
++ u64 timeout)
+{
-+ return _hypercall5(unsigned long, ia64_dom0vp_op,
-+ IA64_DOM0VP_add_physmap, gpfn, mfn, flags, domid);
++ return _hypercall1(long, set_timer_op, timeout);
+}
+
-+static inline unsigned long
-+HYPERVISOR_add_physmap(unsigned long gpfn, unsigned long mfn,
-+ unsigned long flags, domid_t domid)
++static inline int __must_check
++HYPERVISOR_platform_op(
++ struct xen_platform_op *platform_op)
+{
-+ unsigned long ret = 0;
-+ BUG_ON(!is_running_on_xen());//XXX
-+ if (is_running_on_xen()) {
-+ ret = __HYPERVISOR_add_physmap(gpfn, mfn, flags, domid);
-+ }
-+ return ret;
++ platform_op->interface_version = XENPF_INTERFACE_VERSION;
++ return _hypercall1(int, platform_op, platform_op);
+}
+
-+static inline unsigned long
-+__HYPERVISOR_add_physmap_with_gmfn(unsigned long gpfn, unsigned long gmfn,
-+ unsigned long flags, domid_t domid)
++static inline int __must_check
++HYPERVISOR_set_debugreg(
++ unsigned int reg, unsigned long value)
+{
-+ return _hypercall5(unsigned long, ia64_dom0vp_op,
-+ IA64_DOM0VP_add_physmap_with_gmfn,
-+ gpfn, gmfn, flags, domid);
++ return _hypercall2(int, set_debugreg, reg, value);
+}
+
-+static inline unsigned long
-+HYPERVISOR_add_physmap_with_gmfn(unsigned long gpfn, unsigned long gmfn,
-+ unsigned long flags, domid_t domid)
++static inline unsigned long __must_check
++HYPERVISOR_get_debugreg(
++ unsigned int reg)
+{
-+ unsigned long ret = 0;
-+ BUG_ON(!is_running_on_xen());//XXX
-+ if (is_running_on_xen()) {
-+ ret = __HYPERVISOR_add_physmap_with_gmfn(gpfn, gmfn,
-+ flags, domid);
-+ }
-+ return ret;
++ return _hypercall1(unsigned long, get_debugreg, reg);
+}
+
-+#ifdef CONFIG_XEN_IA64_EXPOSE_P2M
-+static inline unsigned long
-+HYPERVISOR_expose_p2m(unsigned long conv_start_gpfn,
-+ unsigned long assign_start_gpfn,
-+ unsigned long expose_size, unsigned long granule_pfn)
++static inline int __must_check
++HYPERVISOR_update_descriptor(
++ unsigned long ma, unsigned long word)
+{
-+ return _hypercall5(unsigned long, ia64_dom0vp_op,
-+ IA64_DOM0VP_expose_p2m, conv_start_gpfn,
-+ assign_start_gpfn, expose_size, granule_pfn);
++ return _hypercall2(int, update_descriptor, ma, word);
+}
-+#endif
+
-+static inline int
-+xencomm_arch_hypercall_perfmon_op(unsigned long cmd,
-+ struct xencomm_handle *arg,
-+ unsigned long count)
++static inline int __must_check
++HYPERVISOR_memory_op(
++ unsigned int cmd, void *arg)
+{
-+ return _hypercall4(int, ia64_dom0vp_op,
-+ IA64_DOM0VP_perfmon, cmd, arg, count);
++ return _hypercall2(int, memory_op, cmd, arg);
+}
+
-+static inline int
-+HYPERVISOR_add_io_space(unsigned long phys_base,
-+ unsigned long sparse,
-+ unsigned long space_number)
++static inline int __must_check
++HYPERVISOR_multicall(
++ multicall_entry_t *call_list, unsigned int nr_calls)
+{
-+ return _hypercall4(int, ia64_dom0vp_op, IA64_DOM0VP_add_io_space,
-+ phys_base, sparse, space_number);
++ return _hypercall2(int, multicall, call_list, nr_calls);
+}
+
-+// for balloon driver
-+#define HYPERVISOR_update_va_mapping(va, new_val, flags) (0)
-+
-+/* Use xencomm to do hypercalls. */
-+#ifdef MODULE
-+#define HYPERVISOR_sched_op xencomm_mini_hypercall_sched_op
-+#define HYPERVISOR_event_channel_op xencomm_mini_hypercall_event_channel_op
-+#define HYPERVISOR_callback_op xencomm_mini_hypercall_callback_op
-+#define HYPERVISOR_multicall xencomm_mini_hypercall_multicall
-+#define HYPERVISOR_xen_version xencomm_mini_hypercall_xen_version
-+#define HYPERVISOR_console_io xencomm_mini_hypercall_console_io
-+#define HYPERVISOR_hvm_op xencomm_mini_hypercall_hvm_op
-+#define HYPERVISOR_memory_op xencomm_mini_hypercall_memory_op
-+#define HYPERVISOR_xenoprof_op xencomm_mini_hypercall_xenoprof_op
-+#define HYPERVISOR_perfmon_op xencomm_mini_hypercall_perfmon_op
-+#else
-+#define HYPERVISOR_sched_op xencomm_hypercall_sched_op
-+#define HYPERVISOR_event_channel_op xencomm_hypercall_event_channel_op
-+#define HYPERVISOR_callback_op xencomm_hypercall_callback_op
-+#define HYPERVISOR_multicall xencomm_hypercall_multicall
-+#define HYPERVISOR_xen_version xencomm_hypercall_xen_version
-+#define HYPERVISOR_console_io xencomm_hypercall_console_io
-+#define HYPERVISOR_hvm_op xencomm_hypercall_hvm_op
-+#define HYPERVISOR_memory_op xencomm_hypercall_memory_op
-+#define HYPERVISOR_xenoprof_op xencomm_hypercall_xenoprof_op
-+#define HYPERVISOR_perfmon_op xencomm_hypercall_perfmon_op
-+#endif
-+
-+#define HYPERVISOR_suspend xencomm_hypercall_suspend
-+#define HYPERVISOR_vcpu_op xencomm_hypercall_vcpu_op
-+
-+#endif /* __HYPERCALL_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/hypervisor.h tmp-linux-2.6-xen.patch/include/asm-ia64/hypervisor.h
---- pristine-linux-2.6.18.2/include/asm-ia64/hypervisor.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/hypervisor.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,226 @@
-+/******************************************************************************
-+ * hypervisor.h
-+ *
-+ * Linux-specific hypervisor handling.
-+ *
-+ * Copyright (c) 2002-2004, K A Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
-+
-+#ifndef __HYPERVISOR_H__
-+#define __HYPERVISOR_H__
-+
-+#ifdef CONFIG_XEN
-+extern int running_on_xen;
-+#define is_running_on_xen() (running_on_xen)
-+#else /* CONFIG_XEN */
-+# ifdef CONFIG_VMX_GUEST
-+# define is_running_on_xen() (1)
-+# else /* CONFIG_VMX_GUEST */
-+# define is_running_on_xen() (0)
-+# define HYPERVISOR_ioremap(offset, size) (offset)
-+# endif /* CONFIG_VMX_GUEST */
-+#endif /* CONFIG_XEN */
-+
-+#if defined(CONFIG_XEN) || defined(CONFIG_VMX_GUEST)
-+#include <linux/types.h>
-+#include <linux/kernel.h>
-+#include <linux/version.h>
-+#include <linux/errno.h>
-+#include <xen/interface/xen.h>
-+#include <xen/interface/platform.h>
-+#include <xen/interface/event_channel.h>
-+#include <xen/interface/physdev.h>
-+#include <xen/interface/sched.h>
-+#include <xen/hypercall.h>
-+#include <asm/ptrace.h>
-+#include <asm/page.h>
-+
-+extern shared_info_t *HYPERVISOR_shared_info;
-+extern start_info_t *xen_start_info;
-+
-+void force_evtchn_callback(void);
-+
-+/* Turn jiffies into Xen system time. XXX Implement me. */
-+#define jiffies_to_st(j) 0
++static inline int __must_check
++HYPERVISOR_update_va_mapping(
++ unsigned long va, pte_t new_val, unsigned long flags)
++{
++ return _hypercall3(int, update_va_mapping, va, new_val.pte, flags);
++}
+
-+static inline int
-+HYPERVISOR_yield(
-+ void)
++static inline int __must_check
++HYPERVISOR_event_channel_op(
++ int cmd, void *arg)
+{
-+ int rc = HYPERVISOR_sched_op(SCHEDOP_yield, NULL);
++ int rc = _hypercall2(int, event_channel_op, cmd, arg);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ struct evtchn_op op;
++ op.cmd = cmd;
++ memcpy(&op.u, arg, sizeof(op.u));
++ rc = _hypercall1(int, event_channel_op_compat, &op);
++ memcpy(arg, &op.u, sizeof(op.u));
++ }
++#endif
+
+ return rc;
+}
+
-+static inline int
-+HYPERVISOR_block(
-+ void)
++static inline int __must_check
++HYPERVISOR_acm_op(
++ int cmd, void *arg)
+{
-+ int rc = HYPERVISOR_sched_op(SCHEDOP_block, NULL);
++ return _hypercall2(int, acm_op, cmd, arg);
++}
+
-+ return rc;
++static inline int __must_check
++HYPERVISOR_xen_version(
++ int cmd, void *arg)
++{
++ return _hypercall2(int, xen_version, cmd, arg);
+}
+
-+static inline int
-+HYPERVISOR_shutdown(
-+ unsigned int reason)
++static inline int __must_check
++HYPERVISOR_console_io(
++ int cmd, unsigned int count, char *str)
+{
-+ struct sched_shutdown sched_shutdown = {
-+ .reason = reason
-+ };
++ return _hypercall3(int, console_io, cmd, count, str);
++}
+
-+ int rc = HYPERVISOR_sched_op(SCHEDOP_shutdown, &sched_shutdown);
++static inline int __must_check
++HYPERVISOR_physdev_op(
++ int cmd, void *arg)
++{
++ int rc = _hypercall2(int, physdev_op, cmd, arg);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ struct physdev_op op;
++ op.cmd = cmd;
++ memcpy(&op.u, arg, sizeof(op.u));
++ rc = _hypercall1(int, physdev_op_compat, &op);
++ memcpy(arg, &op.u, sizeof(op.u));
++ }
++#endif
+
+ return rc;
+}
+
-+static inline int
-+HYPERVISOR_poll(
-+ evtchn_port_t *ports, unsigned int nr_ports, u64 timeout)
++static inline int __must_check
++HYPERVISOR_grant_table_op(
++ unsigned int cmd, void *uop, unsigned int count)
+{
-+ struct sched_poll sched_poll = {
-+ .nr_ports = nr_ports,
-+ .timeout = jiffies_to_st(timeout)
-+ };
++ return _hypercall3(int, grant_table_op, cmd, uop, count);
++}
+
-+ int rc;
++static inline int __must_check
++HYPERVISOR_update_va_mapping_otherdomain(
++ unsigned long va, pte_t new_val, unsigned long flags, domid_t domid)
++{
++ return _hypercall4(int, update_va_mapping_otherdomain, va,
++ new_val.pte, flags, domid);
++}
+
-+ set_xen_guest_handle(sched_poll.ports, ports);
-+ rc = HYPERVISOR_sched_op(SCHEDOP_poll, &sched_poll);
++static inline int __must_check
++HYPERVISOR_vm_assist(
++ unsigned int cmd, unsigned int type)
++{
++ return _hypercall2(int, vm_assist, cmd, type);
++}
+
-+ return rc;
++static inline int __must_check
++HYPERVISOR_vcpu_op(
++ int cmd, unsigned int vcpuid, void *extra_args)
++{
++ return _hypercall3(int, vcpu_op, cmd, vcpuid, extra_args);
+}
+
-+#ifndef CONFIG_VMX_GUEST
-+// for drivers/xen/privcmd/privcmd.c
-+#define machine_to_phys_mapping 0
-+struct vm_area_struct;
-+int direct_remap_pfn_range(struct vm_area_struct *vma,
-+ unsigned long address,
-+ unsigned long mfn,
-+ unsigned long size,
-+ pgprot_t prot,
-+ domid_t domid);
-+struct file;
-+int privcmd_enforce_singleshot_mapping(struct vm_area_struct *vma);
-+int privcmd_mmap(struct file * file, struct vm_area_struct * vma);
-+#define HAVE_ARCH_PRIVCMD_MMAP
++static inline int __must_check
++HYPERVISOR_set_segment_base(
++ int reg, unsigned long value)
++{
++ return _hypercall2(int, set_segment_base, reg, value);
++}
+
-+// for drivers/xen/balloon/balloon.c
-+#ifdef CONFIG_XEN_SCRUB_PAGES
-+#define scrub_pages(_p,_n) memset((void *)(_p), 0, (_n) << PAGE_SHIFT)
-+#else
-+#define scrub_pages(_p,_n) ((void)0)
-+#endif
-+#define pte_mfn(_x) pte_pfn(_x)
-+#define phys_to_machine_mapping_valid(_x) (1)
++static inline int __must_check
++HYPERVISOR_suspend(
++ unsigned long srec)
++{
++ struct sched_shutdown sched_shutdown = {
++ .reason = SHUTDOWN_suspend
++ };
+
-+#endif /* !CONFIG_VMX_GUEST */
++ int rc = _hypercall3(int, sched_op, SCHEDOP_shutdown,
++ &sched_shutdown, srec);
+
-+#define __pte_ma(_x) ((pte_t) {(_x)}) /* unmodified use */
-+#define pfn_pte_ma(_x,_y) __pte_ma(0) /* unmodified use */
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (rc == -ENOSYS)
++ rc = _hypercall3(int, sched_op_compat, SCHEDOP_shutdown,
++ SHUTDOWN_suspend, srec);
++#endif
+
-+#ifndef CONFIG_VMX_GUEST
-+int __xen_create_contiguous_region(unsigned long vstart, unsigned int order, unsigned int address_bits);
++ return rc;
++}
++
++#if CONFIG_XEN_COMPAT <= 0x030002
+static inline int
-+xen_create_contiguous_region(unsigned long vstart,
-+ unsigned int order, unsigned int address_bits)
++HYPERVISOR_nmi_op(
++ unsigned long op, void *arg)
+{
-+ int ret = 0;
-+ if (is_running_on_xen()) {
-+ ret = __xen_create_contiguous_region(vstart, order,
-+ address_bits);
-+ }
-+ return ret;
++ return _hypercall2(int, nmi_op, op, arg);
+}
++#endif
+
-+void __xen_destroy_contiguous_region(unsigned long vstart, unsigned int order);
-+static inline void
-+xen_destroy_contiguous_region(unsigned long vstart, unsigned int order)
++#ifndef CONFIG_XEN
++static inline unsigned long __must_check
++HYPERVISOR_hvm_op(
++ int op, void *arg)
+{
-+ if (is_running_on_xen())
-+ __xen_destroy_contiguous_region(vstart, order);
++ return _hypercall2(unsigned long, hvm_op, op, arg);
+}
++#endif
+
-+/* For drivers/xen/core/machine_reboot.c */
-+#define HAVE_XEN_POST_SUSPEND
-+void xen_post_suspend(int suspend_cancelled);
-+#endif /* !CONFIG_VMX_GUEST */
-+
-+// for netfront.c, netback.c
-+#define MULTI_UVMFLAGS_INDEX 0 //XXX any value
++static inline int __must_check
++HYPERVISOR_callback_op(
++ int cmd, const void *arg)
++{
++ return _hypercall2(int, callback_op, cmd, arg);
++}
+
-+static inline void
-+MULTI_update_va_mapping(
-+ multicall_entry_t *mcl, unsigned long va,
-+ pte_t new_val, unsigned long flags)
++static inline int __must_check
++HYPERVISOR_xenoprof_op(
++ int op, void *arg)
+{
-+ mcl->op = __HYPERVISOR_update_va_mapping;
-+ mcl->result = 0;
++ return _hypercall2(int, xenoprof_op, op, arg);
+}
+
-+static inline void
-+MULTI_grant_table_op(multicall_entry_t *mcl, unsigned int cmd,
-+ void *uop, unsigned int count)
++static inline int __must_check
++HYPERVISOR_kexec_op(
++ unsigned long op, void *args)
+{
-+ mcl->op = __HYPERVISOR_grant_table_op;
-+ mcl->args[0] = cmd;
-+ mcl->args[1] = (unsigned long)uop;
-+ mcl->args[2] = count;
++ return _hypercall2(int, kexec_op, op, args);
+}
+
++#endif /* __HYPERCALL_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/hypervisor.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/hypervisor.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,2 @@
++
++#include <asm-i386/mach-xen/asm/hypervisor.h>
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/io.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/io.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,329 @@
++#ifndef _ASM_IO_H
++#define _ASM_IO_H
++
++#include <asm/fixmap.h>
++
+/*
-+ * for blktap.c
-+ * int create_lookup_pte_addr(struct mm_struct *mm,
-+ * unsigned long address,
-+ * uint64_t *ptep);
++ * This file contains the definitions for the x86 IO instructions
++ * inb/inw/inl/outb/outw/outl and the "string versions" of the same
++ * (insb/insw/insl/outsb/outsw/outsl). You can also use "pausing"
++ * versions of the single-IO instructions (inb_p/inw_p/..).
++ *
++ * This file is not meant to be obfuscating: it's just complicated
++ * to (a) handle it all in a way that makes gcc able to optimize it
++ * as well as possible and (b) trying to avoid writing the same thing
++ * over and over again with slight variations and possibly making a
++ * mistake somewhere.
+ */
-+#define create_lookup_pte_addr(mm, address, ptep) \
-+ ({ \
-+ printk(KERN_EMERG \
-+ "%s:%d " \
-+ "create_lookup_pte_addr() isn't supported.\n", \
-+ __func__, __LINE__); \
-+ BUG(); \
-+ (-ENOSYS); \
-+ })
+
-+// for debug
-+asmlinkage int xprintk(const char *fmt, ...);
-+#define xprintd(fmt, ...) xprintk("%s:%d " fmt, __func__, __LINE__, \
-+ ##__VA_ARGS__)
++/*
++ * Thanks to James van Artsdalen for a better timing-fix than
++ * the two short jumps: using outb's to a nonexistent port seems
++ * to guarantee better timings even on fast machines.
++ *
++ * On the other hand, I'd like to be sure of a non-existent port:
++ * I feel a bit unsafe about using 0x80 (should be safe, though)
++ *
++ * Linus
++ */
+
-+#endif /* CONFIG_XEN || CONFIG_VMX_GUEST */
++ /*
++ * Bit simplified and optimized by Jan Hubicka
++ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999.
++ *
++ * isa_memset_io, isa_memcpy_fromio, isa_memcpy_toio added,
++ * isa_read[wl] and isa_write[wl] fixed
++ * - Arnaldo Carvalho de Melo <acme at conectiva.com.br>
++ */
+
-+#ifdef CONFIG_XEN_PRIVILEGED_GUEST
-+#define is_initial_xendomain() \
-+ (is_running_on_xen() ? xen_start_info->flags & SIF_INITDOMAIN : 0)
++#define __SLOW_DOWN_IO "\noutb %%al,$0x80"
++
++#ifdef REALLY_SLOW_IO
++#define __FULL_SLOW_DOWN_IO __SLOW_DOWN_IO __SLOW_DOWN_IO __SLOW_DOWN_IO __SLOW_DOWN_IO
+#else
-+#define is_initial_xendomain() 0
++#define __FULL_SLOW_DOWN_IO __SLOW_DOWN_IO
+#endif
+
-+#endif /* __HYPERVISOR_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/intel_intrin.h tmp-linux-2.6-xen.patch/include/asm-ia64/intel_intrin.h
---- pristine-linux-2.6.18.2/include/asm-ia64/intel_intrin.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/intel_intrin.h 2007-07-30 16:35:13.000000000 +0200
-@@ -16,8 +16,8 @@
- * intrinsic
- */
-
--#define ia64_getreg __getReg
--#define ia64_setreg __setReg
-+#define __ia64_getreg __getReg
-+#define __ia64_setreg __setReg
-
- #define ia64_hint __hint
- #define ia64_hint_pause __hint_pause
-@@ -33,16 +33,16 @@
- #define ia64_getf_exp __getf_exp
- #define ia64_shrp _m64_shrp
-
--#define ia64_tpa __tpa
-+#define __ia64_tpa __tpa
- #define ia64_invala __invala
- #define ia64_invala_gr __invala_gr
- #define ia64_invala_fr __invala_fr
- #define ia64_nop __nop
- #define ia64_sum __sum
--#define ia64_ssm __ssm
-+#define __ia64_ssm __ssm
- #define ia64_rum __rum
--#define ia64_rsm __rsm
--#define ia64_fc __fc
-+#define __ia64_rsm __rsm
-+#define __ia64_fc __fc
-
- #define ia64_ldfs __ldfs
- #define ia64_ldfd __ldfd
-@@ -80,24 +80,24 @@
-
- #define __ia64_set_dbr(index, val) \
- __setIndReg(_IA64_REG_INDR_DBR, index, val)
--#define ia64_set_ibr(index, val) \
-+#define __ia64_set_ibr(index, val) \
- __setIndReg(_IA64_REG_INDR_IBR, index, val)
--#define ia64_set_pkr(index, val) \
-+#define __ia64_set_pkr(index, val) \
- __setIndReg(_IA64_REG_INDR_PKR, index, val)
--#define ia64_set_pmc(index, val) \
-+#define __ia64_set_pmc(index, val) \
- __setIndReg(_IA64_REG_INDR_PMC, index, val)
--#define ia64_set_pmd(index, val) \
-+#define __ia64_set_pmd(index, val) \
- __setIndReg(_IA64_REG_INDR_PMD, index, val)
--#define ia64_set_rr(index, val) \
-+#define __ia64_set_rr(index, val) \
- __setIndReg(_IA64_REG_INDR_RR, index, val)
-
--#define ia64_get_cpuid(index) __getIndReg(_IA64_REG_INDR_CPUID, index)
-+#define __ia64_get_cpuid(index) __getIndReg(_IA64_REG_INDR_CPUID, index)
- #define __ia64_get_dbr(index) __getIndReg(_IA64_REG_INDR_DBR, index)
--#define ia64_get_ibr(index) __getIndReg(_IA64_REG_INDR_IBR, index)
--#define ia64_get_pkr(index) __getIndReg(_IA64_REG_INDR_PKR, index)
--#define ia64_get_pmc(index) __getIndReg(_IA64_REG_INDR_PMC, index)
--#define ia64_get_pmd(index) __getIndReg(_IA64_REG_INDR_PMD, index)
--#define ia64_get_rr(index) __getIndReg(_IA64_REG_INDR_RR, index)
-+#define __ia64_get_ibr(index) __getIndReg(_IA64_REG_INDR_IBR, index)
-+#define __ia64_get_pkr(index) __getIndReg(_IA64_REG_INDR_PKR, index)
-+#define __ia64_get_pmc(index) __getIndReg(_IA64_REG_INDR_PMC, index)
-+#define __ia64_get_pmd(index) __getIndReg(_IA64_REG_INDR_PMD, index)
-+#define __ia64_get_rr(index) __getIndReg(_IA64_REG_INDR_RR, index)
-
- #define ia64_srlz_d __dsrlz
- #define ia64_srlz_i __isrlz
-@@ -116,18 +116,18 @@
- #define ia64_ld8_acq __ld8_acq
-
- #define ia64_sync_i __synci
--#define ia64_thash __thash
--#define ia64_ttag __ttag
--#define ia64_itcd __itcd
--#define ia64_itci __itci
--#define ia64_itrd __itrd
--#define ia64_itri __itri
--#define ia64_ptce __ptce
--#define ia64_ptcl __ptcl
--#define ia64_ptcg __ptcg
--#define ia64_ptcga __ptcga
--#define ia64_ptri __ptri
--#define ia64_ptrd __ptrd
-+#define __ia64_thash __thash
-+#define __ia64_ttag __ttag
-+#define __ia64_itcd __itcd
-+#define __ia64_itci __itci
-+#define __ia64_itrd __itrd
-+#define __ia64_itri __itri
-+#define __ia64_ptce __ptce
-+#define __ia64_ptcl __ptcl
-+#define __ia64_ptcg __ptcg
-+#define __ia64_ptcga __ptcga
-+#define __ia64_ptri __ptri
-+#define __ia64_ptrd __ptrd
- #define ia64_dep_mi _m64_dep_mi
-
- /* Values for lfhint in __lfetch and __lfetch_fault */
-@@ -142,16 +142,18 @@
- #define ia64_lfetch_fault __lfetch_fault
- #define ia64_lfetch_fault_excl __lfetch_fault_excl
-
--#define ia64_intrin_local_irq_restore(x) \
-+#define __ia64_intrin_local_irq_restore(x) \
- do { \
- if ((x) != 0) { \
-- ia64_ssm(IA64_PSR_I); \
-+ __ia64_ssm(IA64_PSR_I); \
- ia64_srlz_d(); \
- } else { \
-- ia64_rsm(IA64_PSR_I); \
-+ __ia64_rsm(IA64_PSR_I); \
- } \
- } while (0)
-
-+#define __ia64_get_psr_i() (__ia64_getreg(_IA64_REG_PSR) & 0x4000UL)
++/*
++ * Talk about misusing macros..
++ */
++#define __OUT1(s,x) \
++static inline void out##s(unsigned x value, unsigned short port) {
+
- #define __builtin_trap() __break(0);
-
- #endif /* _ASM_IA64_INTEL_INTRIN_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/io.h tmp-linux-2.6-xen.patch/include/asm-ia64/io.h
---- pristine-linux-2.6.18.2/include/asm-ia64/io.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/io.h 2007-07-30 16:35:13.000000000 +0200
-@@ -66,9 +66,11 @@ extern unsigned int num_io_spaces;
- #define PIO_RESERVED __IA64_UNCACHED_OFFSET
- #define HAVE_ARCH_PIO_SIZE
-
-+#include <asm/hypervisor.h>
- #include <asm/intrinsics.h>
- #include <asm/machvec.h>
- #include <asm/page.h>
-+#include <asm/privop.h>
- #include <asm/system.h>
- #include <asm-generic/iomap.h>
-
-@@ -96,9 +98,44 @@ extern int valid_mmap_phys_addr_range (u
- * The following two macros are deprecated and scheduled for removal.
- * Please use the PCI-DMA interface defined in <asm/pci.h> instead.
- */
-+#ifndef CONFIG_XEN
- #define bus_to_virt phys_to_virt
- #define virt_to_bus virt_to_phys
- #define page_to_bus page_to_phys
-+#else
-+#define bus_to_virt(bus) \
-+ phys_to_virt(machine_to_phys_for_dma(bus))
-+#define virt_to_bus(virt) \
-+ phys_to_machine_for_dma(virt_to_phys(virt))
-+#define page_to_bus(page) \
-+ phys_to_machine_for_dma(page_to_pseudophys(page))
++#define __OUT2(s,s1,s2) \
++__asm__ __volatile__ ("out" #s " %" s1 "0,%" s2 "1"
+
-+#define page_to_pseudophys(page) \
-+ ((dma_addr_t)page_to_pfn(page) << PAGE_SHIFT)
++#define __OUT(s,s1,x) \
++__OUT1(s,x) __OUT2(s,s1,"w") : : "a" (value), "Nd" (port)); } \
++__OUT1(s##_p,x) __OUT2(s,s1,"w") __FULL_SLOW_DOWN_IO : : "a" (value), "Nd" (port));} \
++
++#define __IN1(s) \
++static inline RETURN_TYPE in##s(unsigned short port) { RETURN_TYPE _v;
++
++#define __IN2(s,s1,s2) \
++__asm__ __volatile__ ("in" #s " %" s2 "1,%" s1 "0"
++
++#define __IN(s,s1,i...) \
++__IN1(s) __IN2(s,s1,"w") : "=a" (_v) : "Nd" (port) ,##i ); return _v; } \
++__IN1(s##_p) __IN2(s,s1,"w") __FULL_SLOW_DOWN_IO : "=a" (_v) : "Nd" (port) ,##i ); return _v; } \
++
++#define __INS(s) \
++static inline void ins##s(unsigned short port, void * addr, unsigned long count) \
++{ __asm__ __volatile__ ("rep ; ins" #s \
++: "=D" (addr), "=c" (count) : "d" (port),"0" (addr),"1" (count)); }
++
++#define __OUTS(s) \
++static inline void outs##s(unsigned short port, const void * addr, unsigned long count) \
++{ __asm__ __volatile__ ("rep ; outs" #s \
++: "=S" (addr), "=c" (count) : "d" (port),"0" (addr),"1" (count)); }
++
++#define RETURN_TYPE unsigned char
++__IN(b,"")
++#undef RETURN_TYPE
++#define RETURN_TYPE unsigned short
++__IN(w,"")
++#undef RETURN_TYPE
++#define RETURN_TYPE unsigned int
++__IN(l,"")
++#undef RETURN_TYPE
++
++__OUT(b,"b",char)
++__OUT(w,"w",short)
++__OUT(l,,int)
++
++__INS(b)
++__INS(w)
++__INS(l)
++
++__OUTS(b)
++__OUTS(w)
++__OUTS(l)
++
++#define IO_SPACE_LIMIT 0xffff
++
++#if defined(__KERNEL__) && __x86_64__
++
++#include <linux/vmalloc.h>
+
++#ifndef __i386__
+/*
-+ * Drivers that use page_to_phys() for bus addresses are broken.
-+ * This includes:
-+ * drivers/ide/cris/ide-cris.c
-+ * drivers/scsi/dec_esp.c
++ * Change virtual addresses to physical addresses and vv.
++ * These are pretty trivial
+ */
-+#define page_to_phys(page) (page_to_pseudophys(page))
-+#define bvec_to_bus(bv) (page_to_bus((bv)->bv_page) + \
-+ (unsigned long) (bv)->bv_offset)
-+#define bio_to_pseudophys(bio) (page_to_pseudophys(bio_page((bio))) + \
-+ (unsigned long) bio_offset((bio)))
-+#define bvec_to_pseudophys(bv) (page_to_pseudophys((bv)->bv_page) + \
-+ (unsigned long) (bv)->bv_offset)
-+#define BIOVEC_PHYS_MERGEABLE(vec1, vec2) \
-+ (((bvec_to_bus((vec1)) + (vec1)->bv_len) == bvec_to_bus((vec2))) && \
-+ ((bvec_to_pseudophys((vec1)) + (vec1)->bv_len) == \
-+ bvec_to_pseudophys((vec2))))
++static inline unsigned long virt_to_phys(volatile void * address)
++{
++ return __pa(address);
++}
+
-+/* We will be supplying our own /dev/mem implementation */
-+#define ARCH_HAS_DEV_MEM
-+#define ARCH_HAS_DEV_MEM_MMAP_MEM
-+int xen_mmap_mem(struct file * file, struct vm_area_struct * vma);
-+#endif /* CONFIG_XEN */
-
- # endif /* KERNEL */
-
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/iosapic.h tmp-linux-2.6-xen.patch/include/asm-ia64/iosapic.h
---- pristine-linux-2.6.18.2/include/asm-ia64/iosapic.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/iosapic.h 2007-07-30 16:35:13.000000000 +0200
-@@ -53,6 +53,7 @@
-
- #define NR_IOSAPICS 256
-
-+#ifndef CONFIG_XEN
- static inline unsigned int iosapic_read(char __iomem *iosapic, unsigned int reg)
- {
- writel(reg, iosapic + IOSAPIC_REG_SELECT);
-@@ -64,6 +65,7 @@ static inline void iosapic_write(char __
- writel(reg, iosapic + IOSAPIC_REG_SELECT);
- writel(val, iosapic + IOSAPIC_WINDOW);
- }
-+#endif
-
- static inline void iosapic_eoi(char __iomem *iosapic, u32 vector)
- {
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/irq.h tmp-linux-2.6-xen.patch/include/asm-ia64/irq.h
---- pristine-linux-2.6.18.2/include/asm-ia64/irq.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/irq.h 2007-07-30 16:35:13.000000000 +0200
-@@ -11,8 +11,41 @@
- * 02/29/00 D.Mosberger moved most things into hw_irq.h
- */
-
-+#ifndef CONFIG_XEN
- #define NR_IRQS 256
- #define NR_IRQ_VECTORS NR_IRQS
-+#else
-+/*
-+ * The flat IRQ space is divided into two regions:
-+ * 1. A one-to-one mapping of real physical IRQs. This space is only used
-+ * if we have physical device-access privilege. This region is at the
-+ * start of the IRQ space so that existing device drivers do not need
-+ * to be modified to translate physical IRQ numbers into our IRQ space.
-+ * 3. A dynamic mapping of inter-domain and Xen-sourced virtual IRQs. These
-+ * are bound using the provided bind/unbind functions.
++static inline void * phys_to_virt(unsigned long address)
++{
++ return __va(address);
++}
++
++#define virt_to_bus(_x) phys_to_machine(__pa(_x))
++#define bus_to_virt(_x) __va(machine_to_phys(_x))
++#endif
++
++/*
++ * Change "struct page" to physical address.
+ */
++#define page_to_pseudophys(page) ((dma_addr_t)page_to_pfn(page) << PAGE_SHIFT)
++#define page_to_phys(page) (phys_to_machine(page_to_pseudophys(page)))
++#define page_to_bus(page) (phys_to_machine(page_to_pseudophys(page)))
+
-+#define PIRQ_BASE 0
-+#define NR_PIRQS 256
++#define bio_to_pseudophys(bio) (page_to_pseudophys(bio_page((bio))) + \
++ (unsigned long) bio_offset((bio)))
++#define bvec_to_pseudophys(bv) (page_to_pseudophys((bv)->bv_page) + \
++ (unsigned long) (bv)->bv_offset)
+
-+#define DYNIRQ_BASE (PIRQ_BASE + NR_PIRQS)
-+#define NR_DYNIRQS 256
++#define BIOVEC_PHYS_MERGEABLE(vec1, vec2) \
++ (((bvec_to_phys((vec1)) + (vec1)->bv_len) == bvec_to_phys((vec2))) && \
++ ((bvec_to_pseudophys((vec1)) + (vec1)->bv_len) == \
++ bvec_to_pseudophys((vec2))))
+
-+#define NR_IRQS (NR_PIRQS + NR_DYNIRQS)
-+#define NR_IRQ_VECTORS NR_IRQS
++#include <asm-generic/iomap.h>
+
-+#define pirq_to_irq(_x) ((_x) + PIRQ_BASE)
-+#define irq_to_pirq(_x) ((_x) - PIRQ_BASE)
++extern void __iomem *__ioremap(unsigned long offset, unsigned long size, unsigned long flags);
+
-+#define dynirq_to_irq(_x) ((_x) + DYNIRQ_BASE)
-+#define irq_to_dynirq(_x) ((_x) - DYNIRQ_BASE)
++static inline void __iomem * ioremap (unsigned long offset, unsigned long size)
++{
++ return __ioremap(offset, size, 0);
++}
+
-+#define RESCHEDULE_VECTOR 0
-+#define IPI_VECTOR 1
-+#define CMCP_VECTOR 2
-+#define CPEP_VECTOR 3
-+#define NR_IPIS 4
-+#endif /* CONFIG_XEN */
-
- static __inline__ int
- irq_canonicalize (int irq)
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/machvec_dig.h tmp-linux-2.6-xen.patch/include/asm-ia64/machvec_dig.h
---- pristine-linux-2.6.18.2/include/asm-ia64/machvec_dig.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/machvec_dig.h 2007-07-30 16:35:13.000000000 +0200
-@@ -13,4 +13,19 @@ extern ia64_mv_setup_t dig_setup;
- #define platform_name "dig"
- #define platform_setup dig_setup
-
-+#ifdef CONFIG_XEN
-+# define platform_dma_map_sg dma_map_sg
-+# define platform_dma_unmap_sg dma_unmap_sg
-+# define platform_dma_mapping_error dma_mapping_error
-+# define platform_dma_supported dma_supported
-+# define platform_dma_alloc_coherent dma_alloc_coherent
-+# define platform_dma_free_coherent dma_free_coherent
-+# define platform_dma_map_single dma_map_single
-+# define platform_dma_unmap_single dma_unmap_single
-+# define platform_dma_sync_single_for_cpu \
-+ dma_sync_single_for_cpu
-+# define platform_dma_sync_single_for_device \
-+ dma_sync_single_for_device
-+#endif
-+
- #endif /* _ASM_IA64_MACHVEC_DIG_h */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/maddr.h tmp-linux-2.6-xen.patch/include/asm-ia64/maddr.h
---- pristine-linux-2.6.18.2/include/asm-ia64/maddr.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/maddr.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,98 @@
-+#ifndef _ASM_IA64_MADDR_H
-+#define _ASM_IA64_MADDR_H
++extern void *bt_ioremap(unsigned long addr, unsigned long size);
++extern void bt_iounmap(void *addr, unsigned long size);
++#define early_ioremap bt_ioremap
++#define early_iounmap bt_iounmap
+
-+#include <linux/kernel.h>
-+#include <asm/hypervisor.h>
-+#include <xen/features.h>
-+#include <xen/interface/xen.h>
++/*
++ * This one maps high address device memory and turns off caching for that area.
++ * it's useful if some control registers are in such an area and write combining
++ * or read caching is not desirable:
++ */
++extern void __iomem * ioremap_nocache (unsigned long offset, unsigned long size);
++extern void iounmap(volatile void __iomem *addr);
+
-+#ifdef CONFIG_XEN
++/*
++ * ISA I/O bus memory addresses are 1:1 with the physical address.
++ */
+
-+#define INVALID_P2M_ENTRY (~0UL)
++#define isa_virt_to_bus(_x) ({ BUG(); virt_to_bus(_x); })
++#define isa_page_to_bus(_x) isa_page_to_bus_is_UNSUPPORTED->x
++#define isa_bus_to_virt(_x) (void *)(__fix_to_virt(FIX_ISAMAP_BEGIN) + (_x))
+
-+#ifdef CONFIG_XEN_IA64_EXPOSE_P2M
-+extern int p2m_initialized;
-+extern unsigned long p2m_min_low_pfn;
-+extern unsigned long p2m_max_low_pfn;
-+extern unsigned long p2m_convert_min_pfn;
-+extern unsigned long p2m_convert_max_pfn;
-+extern volatile const pte_t* p2m_pte;
-+unsigned long p2m_phystomach(unsigned long gpfn);
-+#else
-+#define p2m_initialized (0)
-+#define p2m_phystomach(gpfn) INVALID_MFN
-+#endif
++/*
++ * However PCI ones are not necessarily 1:1 and therefore these interfaces
++ * are forbidden in portable PCI drivers.
++ *
++ * Allow them on x86 for legacy drivers, though.
++ */
++#define virt_to_bus(_x) phys_to_machine(__pa(_x))
++#define bus_to_virt(_x) __va(machine_to_phys(_x))
+
-+/* XXX xen page size != page size */
-+static inline unsigned long
-+pfn_to_mfn_for_dma(unsigned long pfn)
++/*
++ * readX/writeX() are used to access memory mapped devices. On some
++ * architectures the memory mapped IO stuff needs to be accessed
++ * differently. On the x86 architecture, we just read/write the
++ * memory location directly.
++ */
++
++static inline __u8 __readb(const volatile void __iomem *addr)
+{
-+ unsigned long mfn;
-+ if (p2m_initialized)
-+ return p2m_phystomach(pfn);
-+ mfn = HYPERVISOR_phystomach(pfn);
-+ BUG_ON(mfn == 0); // XXX
-+ BUG_ON(mfn == INVALID_P2M_ENTRY); // XXX
-+ BUG_ON(mfn == INVALID_MFN);
-+ return mfn;
++ return *(__force volatile __u8 *)addr;
+}
-+
-+static inline unsigned long
-+phys_to_machine_for_dma(unsigned long phys)
++static inline __u16 __readw(const volatile void __iomem *addr)
+{
-+ unsigned long machine =
-+ pfn_to_mfn_for_dma(phys >> PAGE_SHIFT) << PAGE_SHIFT;
-+ machine |= (phys & ~PAGE_MASK);
-+ return machine;
++ return *(__force volatile __u16 *)addr;
+}
-+
-+static inline unsigned long
-+mfn_to_pfn_for_dma(unsigned long mfn)
++static __always_inline __u32 __readl(const volatile void __iomem *addr)
+{
-+ unsigned long pfn;
-+ pfn = HYPERVISOR_machtophys(mfn);
-+ BUG_ON(pfn == 0);
-+ //BUG_ON(pfn == INVALID_M2P_ENTRY);
-+ return pfn;
++ return *(__force volatile __u32 *)addr;
+}
-+
-+static inline unsigned long
-+machine_to_phys_for_dma(unsigned long machine)
++static inline __u64 __readq(const volatile void __iomem *addr)
+{
-+ unsigned long phys =
-+ mfn_to_pfn_for_dma(machine >> PAGE_SHIFT) << PAGE_SHIFT;
-+ phys |= (machine & ~PAGE_MASK);
-+ return phys;
++ return *(__force volatile __u64 *)addr;
+}
++#define readb(x) __readb(x)
++#define readw(x) __readw(x)
++#define readl(x) __readl(x)
++#define readq(x) __readq(x)
++#define readb_relaxed(a) readb(a)
++#define readw_relaxed(a) readw(a)
++#define readl_relaxed(a) readl(a)
++#define readq_relaxed(a) readq(a)
++#define __raw_readb readb
++#define __raw_readw readw
++#define __raw_readl readl
++#define __raw_readq readq
+
-+static inline unsigned long
-+mfn_to_local_pfn(unsigned long mfn)
++#define mmiowb()
++
++static inline void __writel(__u32 b, volatile void __iomem *addr)
+{
-+ unsigned long pfn = mfn_to_pfn_for_dma(mfn);
-+ if (!pfn_valid(pfn))
-+ return INVALID_P2M_ENTRY;
-+ return pfn;
++ *(__force volatile __u32 *)addr = b;
++}
++static inline void __writeq(__u64 b, volatile void __iomem *addr)
++{
++ *(__force volatile __u64 *)addr = b;
++}
++static inline void __writeb(__u8 b, volatile void __iomem *addr)
++{
++ *(__force volatile __u8 *)addr = b;
++}
++static inline void __writew(__u16 b, volatile void __iomem *addr)
++{
++ *(__force volatile __u16 *)addr = b;
+}
++#define writeq(val,addr) __writeq((val),(addr))
++#define writel(val,addr) __writel((val),(addr))
++#define writew(val,addr) __writew((val),(addr))
++#define writeb(val,addr) __writeb((val),(addr))
++#define __raw_writeb writeb
++#define __raw_writew writew
++#define __raw_writel writel
++#define __raw_writeq writeq
+
-+#else /* !CONFIG_XEN */
++void __memcpy_fromio(void*,unsigned long,unsigned);
++void __memcpy_toio(unsigned long,const void*,unsigned);
+
-+#define pfn_to_mfn_for_dma(pfn) (pfn)
-+#define mfn_to_pfn_for_dma(mfn) (mfn)
-+#define phys_to_machine_for_dma(phys) (phys)
-+#define machine_to_phys_for_dma(machine) (machine)
-+#define mfn_to_local_pfn(mfn) (mfn)
++static inline void memcpy_fromio(void *to, const volatile void __iomem *from, unsigned len)
++{
++ __memcpy_fromio(to,(unsigned long)from,len);
++}
++static inline void memcpy_toio(volatile void __iomem *to, const void *from, unsigned len)
++{
++ __memcpy_toio((unsigned long)to,from,len);
++}
+
-+#endif /* !CONFIG_XEN */
++void memset_io(volatile void __iomem *a, int b, size_t c);
+
-+#define mfn_to_pfn(mfn) (mfn)
-+#define pfn_to_mfn(pfn) (pfn)
++/*
++ * ISA space is 'always mapped' on a typical x86 system, no need to
++ * explicitly ioremap() it. The fact that the ISA IO space is mapped
++ * to PAGE_OFFSET is pure coincidence - it does not mean ISA values
++ * are physical addresses. The following constant pointer can be
++ * used as the IO-area pointer (it can be iounmapped as well, so the
++ * analogy with PCI is quite large):
++ */
++#define __ISA_IO_base ((char __iomem *)(fix_to_virt(FIX_ISAMAP_BEGIN)))
+
-+#define mfn_to_virt(mfn) (__va((mfn) << PAGE_SHIFT))
-+#define virt_to_mfn(virt) (__pa(virt) >> PAGE_SHIFT)
-+#define virt_to_machine(virt) __pa(virt) // for tpmfront.c
++/*
++ * Again, x86-64 does not require mem IO specific function.
++ */
+
-+#define set_phys_to_machine(pfn, mfn) do { } while (0)
++#define eth_io_copy_and_sum(a,b,c,d) eth_copy_and_sum((a),(void *)(b),(c),(d))
+
-+typedef unsigned long maddr_t; // to compile netback, netfront
++/**
++ * check_signature - find BIOS signatures
++ * @io_addr: mmio address to check
++ * @signature: signature block
++ * @length: length of signature
++ *
++ * Perform a signature comparison with the mmio address io_addr. This
++ * address should have been obtained by ioremap.
++ * Returns 1 on a match.
++ */
++
++static inline int check_signature(void __iomem *io_addr,
++ const unsigned char *signature, int length)
++{
++ int retval = 0;
++ do {
++ if (readb(io_addr) != *signature)
++ goto out;
++ io_addr++;
++ signature++;
++ length--;
++ } while (length);
++ retval = 1;
++out:
++ return retval;
++}
+
-+#endif /* _ASM_IA64_MADDR_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/meminit.h tmp-linux-2.6-xen.patch/include/asm-ia64/meminit.h
---- pristine-linux-2.6.18.2/include/asm-ia64/meminit.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/meminit.h 2007-07-30 16:35:13.000000000 +0200
-@@ -16,10 +16,15 @@
- * - command line string
- * - kernel code & data
- * - Kernel memory map built from EFI memory map
-+ * - xen start info
- *
- * More could be added if necessary
- */
-+#ifndef CONFIG_XEN
- #define IA64_MAX_RSVD_REGIONS 6
-+#else
-+#define IA64_MAX_RSVD_REGIONS 7
-+#endif
-
- struct rsvd_region {
- unsigned long start; /* virtual address of beginning of element */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/page.h tmp-linux-2.6-xen.patch/include/asm-ia64/page.h
---- pristine-linux-2.6.18.2/include/asm-ia64/page.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/page.h 2007-07-30 16:35:13.000000000 +0200
-@@ -119,6 +119,7 @@ extern struct page *vmem_map;
- #endif
-
- #ifdef CONFIG_FLATMEM
-+extern unsigned long max_mapnr;
- # define pfn_valid(pfn) (((pfn) < max_mapnr) && ia64_pfn_valid(pfn))
- #elif defined(CONFIG_DISCONTIGMEM)
- extern unsigned long min_low_pfn;
-@@ -126,7 +127,9 @@ extern unsigned long max_low_pfn;
- # define pfn_valid(pfn) (((pfn) >= min_low_pfn) && ((pfn) < max_low_pfn) && ia64_pfn_valid(pfn))
- #endif
-
-+#ifndef CONFIG_XEN
- #define page_to_phys(page) (page_to_pfn(page) << PAGE_SHIFT)
-+#endif
- #define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT)
- #define pfn_to_kaddr(pfn) __va((pfn) << PAGE_SHIFT)
-
-@@ -227,5 +230,25 @@ get_order (unsigned long size)
- (((current->personality & READ_IMPLIES_EXEC) != 0) \
- ? VM_EXEC : 0))
-
-+#ifndef __ASSEMBLY__
-+#ifdef CONFIG_XEN
++/* Nothing to do */
+
-+#include <linux/kernel.h>
-+#include <asm/hypervisor.h>
-+#include <xen/features.h> // to compile netback, netfront
-+#include <asm/maddr.h>
++#define dma_cache_inv(_start,_size) do { } while (0)
++#define dma_cache_wback(_start,_size) do { } while (0)
++#define dma_cache_wback_inv(_start,_size) do { } while (0)
+
-+#define arch_free_page(_page, _order) \
-+({ \
-+ int foreign = PageForeign(_page); \
-+ if (foreign) \
-+ PageForeignDestructor(_page); \
-+ foreign; \
-+})
-+#define HAVE_ARCH_FREE_PAGE
++#define flush_write_buffers()
+
-+#endif /* CONFIG_XEN */
-+#endif /* __ASSEMBLY__ */
++extern int iommu_bio_merge;
++#define BIO_VMERGE_BOUNDARY iommu_bio_merge
+
- # endif /* __KERNEL__ */
- #endif /* _ASM_IA64_PAGE_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/pal.h tmp-linux-2.6-xen.patch/include/asm-ia64/pal.h
---- pristine-linux-2.6.18.2/include/asm-ia64/pal.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/pal.h 2007-07-30 16:35:13.000000000 +0200
-@@ -82,6 +82,7 @@
- #ifndef __ASSEMBLY__
-
- #include <linux/types.h>
-+#include <asm/processor.h>
- #include <asm/fpu.h>
-
- /*
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/pgalloc.h tmp-linux-2.6-xen.patch/include/asm-ia64/pgalloc.h
---- pristine-linux-2.6.18.2/include/asm-ia64/pgalloc.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/pgalloc.h 2007-07-30 16:35:13.000000000 +0200
-@@ -125,7 +125,11 @@ static inline void pmd_free(pmd_t * pmd)
- static inline void
- pmd_populate(struct mm_struct *mm, pmd_t * pmd_entry, struct page *pte)
- {
-+#ifndef CONFIG_XEN
- pmd_val(*pmd_entry) = page_to_phys(pte);
-+#else
-+ pmd_val(*pmd_entry) = page_to_pseudophys(pte);
-+#endif
- }
-
- static inline void
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/privop.h tmp-linux-2.6-xen.patch/include/asm-ia64/privop.h
---- pristine-linux-2.6.18.2/include/asm-ia64/privop.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/privop.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,60 @@
-+#ifndef _ASM_IA64_PRIVOP_H
-+#define _ASM_IA64_PRIVOP_H
++/*
++ * Convert a physical pointer to a virtual kernel pointer for /dev/mem
++ * access
++ */
++#define xlate_dev_mem_ptr(p) __va(p)
+
+/*
-+ * Copyright (C) 2005 Hewlett-Packard Co
-+ * Dan Magenheimer <dan.magenheimer at hp.com>
-+ *
++ * Convert a virtual cached pointer to an uncached pointer
+ */
++#define xlate_dev_kmem_ptr(p) p
+
-+#ifdef CONFIG_XEN
-+#include <asm/xen/privop.h>
-+#endif
++#endif /* __KERNEL__ */
+
-+#ifndef __ASSEMBLY
++#define ARCH_HAS_DEV_MEM
+
-+#ifndef IA64_PARAVIRTUALIZED
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/irq.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/irq.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,38 @@
++#ifndef _ASM_IRQ_H
++#define _ASM_IRQ_H
+
-+#define ia64_getreg __ia64_getreg
-+#define ia64_setreg __ia64_setreg
-+#define ia64_hint __ia64_hint
-+#define ia64_thash __ia64_thash
-+#define ia64_itci __ia64_itci
-+#define ia64_itcd __ia64_itcd
-+#define ia64_itri __ia64_itri
-+#define ia64_itrd __ia64_itrd
-+#define ia64_tpa __ia64_tpa
-+#define ia64_set_ibr __ia64_set_ibr
-+#define ia64_set_pkr __ia64_set_pkr
-+#define ia64_set_pmc __ia64_set_pmc
-+#define ia64_set_pmd __ia64_set_pmd
-+#define ia64_set_rr __ia64_set_rr
-+#define ia64_get_cpuid __ia64_get_cpuid
-+#define ia64_get_ibr __ia64_get_ibr
-+#define ia64_get_pkr __ia64_get_pkr
-+#define ia64_get_pmc __ia64_get_pmc
-+#define ia64_get_pmd __ia64_get_pmd
-+#define ia64_get_rr __ia64_get_rr
-+#define ia64_fc __ia64_fc
-+#define ia64_ssm __ia64_ssm
-+#define ia64_rsm __ia64_rsm
-+#define ia64_ptce __ia64_ptce
-+#define ia64_ptcga __ia64_ptcga
-+#define ia64_ptcl __ia64_ptcl
-+#define ia64_ptri __ia64_ptri
-+#define ia64_ptrd __ia64_ptrd
-+#define ia64_get_psr_i __ia64_get_psr_i
-+#define ia64_intrin_local_irq_restore __ia64_intrin_local_irq_restore
-+#define ia64_pal_halt_light __ia64_pal_halt_light
-+#define ia64_leave_kernel __ia64_leave_kernel
-+#define ia64_leave_syscall __ia64_leave_syscall
-+#define ia64_trace_syscall __ia64_trace_syscall
-+#define ia64_ret_from_clone __ia64_ret_from_clone
-+#define ia64_switch_to __ia64_switch_to
-+#define ia64_pal_call_static __ia64_pal_call_static
++/*
++ * linux/include/asm/irq.h
++ *
++ * (C) 1992, 1993 Linus Torvalds, (C) 1997 Ingo Molnar
++ *
++ * IRQ/IPI changes taken from work by Thomas Radke
++ * <tomsoft at informatik.tu-chemnitz.de>
++ */
+
-+#endif /* !IA64_PARAVIRTUALIZED */
++#include <linux/sched.h>
++/* include comes from machine specific directory */
++#include "irq_vectors.h"
++#include <asm/thread_info.h>
+
-+#endif /* !__ASSEMBLY */
++static __inline__ int irq_canonicalize(int irq)
++{
++ return ((irq == 2) ? 9 : irq);
++}
+
-+#endif /* _ASM_IA64_PRIVOP_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/processor.h tmp-linux-2.6-xen.patch/include/asm-ia64/processor.h
---- pristine-linux-2.6.18.2/include/asm-ia64/processor.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/processor.h 2007-07-30 16:35:13.000000000 +0200
-@@ -18,6 +18,7 @@
- #include <asm/kregs.h>
- #include <asm/ptrace.h>
- #include <asm/ustack.h>
-+#include <asm/privop.h>
-
- #define IA64_NUM_DBG_REGS 8
- /*
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/sal.h tmp-linux-2.6-xen.patch/include/asm-ia64/sal.h
---- pristine-linux-2.6.18.2/include/asm-ia64/sal.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/sal.h 2007-09-30 18:06:18.000000000 +0200
-@@ -42,6 +42,9 @@
- #include <asm/pal.h>
- #include <asm/system.h>
- #include <asm/fpu.h>
-+#ifdef CONFIG_XEN
-+#include <asm/xen/xencomm.h>
-+#endif
-
- extern spinlock_t sal_lock;
-
-@@ -686,10 +689,28 @@ ia64_sal_clear_state_info (u64 sal_info_
- /* Get the processor and platform information logged by SAL with respect to the machine
- * state at the time of the MCAs, INITs, CMCs, or CPEs.
- */
-+#ifdef CONFIG_XEN
-+static inline u64 ia64_sal_get_state_info_size (u64 sal_info_type);
++#ifdef CONFIG_X86_LOCAL_APIC
++#define ARCH_HAS_NMI_WATCHDOG /* See include/linux/nmi.h */
+#endif
+
- static inline u64
- ia64_sal_get_state_info (u64 sal_info_type, u64 *sal_info)
- {
- struct ia64_sal_retval isrv;
-+#ifdef CONFIG_XEN
-+ if (is_running_on_xen()) {
-+ struct xencomm_handle *desc;
++#define KDB_VECTOR 0xf9
+
-+ if (xencomm_create(sal_info,
-+ ia64_sal_get_state_info_size(sal_info_type),
-+ &desc, GFP_ATOMIC))
-+ return 0;
++# define irq_ctx_init(cpu) do { } while (0)
+
-+ SAL_CALL_REENTRANT(isrv, SAL_GET_STATE_INFO, sal_info_type, 0,
-+ desc, 0, 0, 0, 0);
-+ xencomm_free(desc);
-+ } else
++#ifdef CONFIG_HOTPLUG_CPU
++#include <linux/cpumask.h>
++extern void fixup_irqs(cpumask_t map);
+#endif
- SAL_CALL_REENTRANT(isrv, SAL_GET_STATE_INFO, sal_info_type, 0,
- sal_info, 0, 0, 0, 0);
- if (isrv.status)
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/swiotlb.h tmp-linux-2.6-xen.patch/include/asm-ia64/swiotlb.h
---- pristine-linux-2.6.18.2/include/asm-ia64/swiotlb.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/swiotlb.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,41 @@
-+#ifndef _ASM_SWIOTLB_H
-+#define _ASM_SWIOTLB_H 1
-+
-+/* SWIOTLB interface */
+
-+extern dma_addr_t swiotlb_map_single(struct device *hwdev, void *ptr, size_t size,
-+ int dir);
-+extern void swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr,
-+ size_t size, int dir);
-+extern void swiotlb_sync_single_for_cpu(struct device *hwdev,
-+ dma_addr_t dev_addr,
-+ size_t size, int dir);
-+extern void swiotlb_sync_single_for_device(struct device *hwdev,
-+ dma_addr_t dev_addr,
-+ size_t size, int dir);
-+extern void swiotlb_sync_sg_for_cpu(struct device *hwdev,
-+ struct scatterlist *sg, int nelems,
-+ int dir);
-+extern void swiotlb_sync_sg_for_device(struct device *hwdev,
-+ struct scatterlist *sg, int nelems,
-+ int dir);
-+extern int swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg,
-+ int nents, int direction);
-+extern void swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg,
-+ int nents, int direction);
-+extern int swiotlb_dma_mapping_error(dma_addr_t dma_addr);
-+extern dma_addr_t swiotlb_map_page(struct device *hwdev, struct page *page,
-+ unsigned long offset, size_t size,
-+ enum dma_data_direction direction);
-+extern void swiotlb_unmap_page(struct device *hwdev, dma_addr_t dma_address,
-+ size_t size, enum dma_data_direction direction);
-+extern int swiotlb_dma_supported(struct device *hwdev, u64 mask);
-+extern void swiotlb_init(void);
++#define __ARCH_HAS_DO_SOFTIRQ 1
+
-+#ifdef CONFIG_SWIOTLB
-+extern int swiotlb;
-+#else
-+#define swiotlb 0
-+#endif
++#endif /* _ASM_IRQ_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/irqflags.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/irqflags.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,139 @@
++/*
++ * include/asm-x86_64/irqflags.h
++ *
++ * IRQ flags handling
++ *
++ * This file gets included from lowlevel asm headers too, to provide
++ * wrapped versions of the local_irq_*() APIs, based on the
++ * raw_local_irq_*() functions from the lowlevel headers.
++ */
++#ifndef _ASM_IRQFLAGS_H
++#define _ASM_IRQFLAGS_H
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/synch_bitops.h tmp-linux-2.6-xen.patch/include/asm-ia64/synch_bitops.h
---- pristine-linux-2.6.18.2/include/asm-ia64/synch_bitops.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/synch_bitops.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,61 @@
-+#ifndef __XEN_SYNCH_BITOPS_H__
-+#define __XEN_SYNCH_BITOPS_H__
++#ifndef __ASSEMBLY__
++/*
++ * Interrupt control:
++ */
+
+/*
-+ * Copyright 1992, Linus Torvalds.
-+ * Heavily modified to provide guaranteed strong synchronisation
-+ * when communicating with Xen or other guest OSes running on other CPUs.
++ * The use of 'barrier' in the following reflects their use as local-lock
++ * operations. Reentrancy must be prevented (e.g., __cli()) /before/ following
++ * critical operations are executed. All critical operations must complete
++ * /before/ reentrancy is permitted (e.g., __sti()). Alpha architecture also
++ * includes these barriers, for example.
+ */
+
-+#define ADDR (*(volatile long *) addr)
++#define __raw_local_save_flags() (current_vcpu_info()->evtchn_upcall_mask)
+
-+static __inline__ void synch_set_bit(int nr, volatile void * addr)
-+{
-+ set_bit(nr, addr);
-+}
++#define raw_local_save_flags(flags) \
++ do { (flags) = __raw_local_save_flags(); } while (0)
++
++#define raw_local_irq_restore(x) \
++do { \
++ vcpu_info_t *_vcpu; \
++ barrier(); \
++ _vcpu = current_vcpu_info(); \
++ if ((_vcpu->evtchn_upcall_mask = (x)) == 0) { \
++ barrier(); /* unmask then check (avoid races) */ \
++ if ( unlikely(_vcpu->evtchn_upcall_pending) ) \
++ force_evtchn_callback(); \
++ } \
++} while (0)
+
-+static __inline__ void synch_clear_bit(int nr, volatile void * addr)
-+{
-+ clear_bit(nr, addr);
-+}
++#ifdef CONFIG_X86_VSMP
+
-+static __inline__ void synch_change_bit(int nr, volatile void * addr)
-+{
-+ change_bit(nr, addr);
-+}
++/*
++ * Interrupt control for the VSMP architecture:
++ */
+
-+static __inline__ int synch_test_and_set_bit(int nr, volatile void * addr)
++static inline void raw_local_irq_disable(void)
+{
-+ return test_and_set_bit(nr, addr);
-+}
++ unsigned long flags = __raw_local_save_flags();
+
-+static __inline__ int synch_test_and_clear_bit(int nr, volatile void * addr)
-+{
-+ return test_and_clear_bit(nr, addr);
++ raw_local_irq_restore((flags & ~(1 << 9)) | (1 << 18));
+}
+
-+static __inline__ int synch_test_and_change_bit(int nr, volatile void * addr)
++static inline void raw_local_irq_enable(void)
+{
-+ return test_and_change_bit(nr, addr);
-+}
++ unsigned long flags = __raw_local_save_flags();
+
-+static __inline__ int synch_const_test_bit(int nr, const volatile void * addr)
-+{
-+ return test_bit(nr, addr);
++ raw_local_irq_restore((flags | (1 << 9)) & ~(1 << 18));
+}
+
-+static __inline__ int synch_var_test_bit(int nr, volatile void * addr)
++static inline int raw_irqs_disabled_flags(unsigned long flags)
+{
-+ return test_bit(nr, addr);
++ return !(flags & (1<<9)) || (flags & (1 << 18));
+}
+
-+#define synch_cmpxchg ia64_cmpxchg4_acq
++#else /* CONFIG_X86_VSMP */
+
-+#define synch_test_bit(nr,addr) \
-+(__builtin_constant_p(nr) ? \
-+ synch_const_test_bit((nr),(addr)) : \
-+ synch_var_test_bit((nr),(addr)))
++#define raw_local_irq_disable() \
++do { \
++ current_vcpu_info()->evtchn_upcall_mask = 1; \
++ barrier(); \
++} while (0)
+
-+#define synch_cmpxchg_subword synch_cmpxchg
++#define raw_local_irq_enable() \
++do { \
++ vcpu_info_t *_vcpu; \
++ barrier(); \
++ _vcpu = current_vcpu_info(); \
++ _vcpu->evtchn_upcall_mask = 0; \
++ barrier(); /* unmask then check (avoid races) */ \
++ if ( unlikely(_vcpu->evtchn_upcall_pending) ) \
++ force_evtchn_callback(); \
++} while (0)
+
-+#endif /* __XEN_SYNCH_BITOPS_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/system.h tmp-linux-2.6-xen.patch/include/asm-ia64/system.h
---- pristine-linux-2.6.18.2/include/asm-ia64/system.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/system.h 2007-07-30 16:35:13.000000000 +0200
-@@ -123,7 +123,7 @@ extern struct ia64_boot_param {
- #define __local_irq_save(x) \
- do { \
- ia64_stop(); \
-- (x) = ia64_getreg(_IA64_REG_PSR); \
-+ (x) = ia64_get_psr_i(); \
- ia64_stop(); \
- ia64_rsm(IA64_PSR_I); \
- } while (0)
-@@ -171,7 +171,7 @@ do { \
- #endif /* !CONFIG_IA64_DEBUG_IRQ */
-
- #define local_irq_enable() ({ ia64_stop(); ia64_ssm(IA64_PSR_I); ia64_srlz_d(); })
--#define local_save_flags(flags) ({ ia64_stop(); (flags) = ia64_getreg(_IA64_REG_PSR); })
-+#define local_save_flags(flags) ({ ia64_stop(); (flags) = ia64_get_psr_i(); })
-
- #define irqs_disabled() \
- ({ \
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/uaccess.h tmp-linux-2.6-xen.patch/include/asm-ia64/uaccess.h
---- pristine-linux-2.6.18.2/include/asm-ia64/uaccess.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/uaccess.h 2007-07-30 16:35:13.000000000 +0200
-@@ -365,6 +365,7 @@ ia64_done_with_exception (struct pt_regs
- }
-
- #define ARCH_HAS_TRANSLATE_MEM_PTR 1
-+#ifndef CONFIG_XEN
- static __inline__ char *
- xlate_dev_mem_ptr (unsigned long p)
- {
-@@ -379,6 +380,25 @@ xlate_dev_mem_ptr (unsigned long p)
-
- return ptr;
- }
-+#else
-+static __inline__ char *
-+xlate_dev_mem_ptr (unsigned long p, ssize_t sz)
++static inline int raw_irqs_disabled_flags(unsigned long flags)
+{
-+ unsigned long pfn = p >> PAGE_SHIFT;
-+
-+ if (pfn_valid(pfn) && !PageUncached(pfn_to_page(pfn)))
-+ return __va(p);
-+
-+ return ioremap(p, sz);
++ return (flags != 0);
+}
+
-+static __inline__ void
-+xlate_dev_mem_ptr_unmap (char* v)
-+{
-+ if (REGION_NUMBER(v) == RGN_UNCACHED)
-+ iounmap(v);
-+}
+#endif
-
- /*
- * Convert a virtual cached kernel memory pointer to an uncached pointer
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/xen/privop.h tmp-linux-2.6-xen.patch/include/asm-ia64/xen/privop.h
---- pristine-linux-2.6.18.2/include/asm-ia64/xen/privop.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/xen/privop.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,310 @@
-+#ifndef _ASM_IA64_XEN_PRIVOP_H
-+#define _ASM_IA64_XEN_PRIVOP_H
+
+/*
-+ * Copyright (C) 2005 Hewlett-Packard Co
-+ * Dan Magenheimer <dan.magenheimer at hp.com>
-+ *
-+ * Paravirtualizations of privileged operations for Xen/ia64
-+ *
++ * For spinlocks, etc.:
+ */
+
++#define __raw_local_irq_save() \
++({ \
++ unsigned long flags = __raw_local_save_flags(); \
++ \
++ raw_local_irq_disable(); \
++ \
++ flags; \
++})
+
-+#include <xen/interface/arch-ia64.h>
-+
-+#define IA64_PARAVIRTUALIZED
++#define raw_local_irq_save(flags) \
++ do { (flags) = __raw_local_irq_save(); } while (0)
+
-+/* At 1 MB, before per-cpu space but still addressable using addl instead
-+ of movl. */
-+#define XSI_BASE 0xfffffffffff00000
++#define raw_irqs_disabled() \
++({ \
++ unsigned long flags = __raw_local_save_flags(); \
++ \
++ raw_irqs_disabled_flags(flags); \
++})
+
-+/* Address of mapped regs. */
-+#define XMAPPEDREGS_BASE (XSI_BASE + XSI_SIZE)
++/*
++ * Used in the idle loop; sti takes one instruction cycle
++ * to complete:
++ */
++void raw_safe_halt(void);
+
-+#ifdef __ASSEMBLY__
-+#define XEN_HYPER_RFI break HYPERPRIVOP_RFI
-+#define XEN_HYPER_RSM_PSR_DT break HYPERPRIVOP_RSM_DT
-+#define XEN_HYPER_SSM_PSR_DT break HYPERPRIVOP_SSM_DT
-+#define XEN_HYPER_COVER break HYPERPRIVOP_COVER
-+#define XEN_HYPER_ITC_D break HYPERPRIVOP_ITC_D
-+#define XEN_HYPER_ITC_I break HYPERPRIVOP_ITC_I
-+#define XEN_HYPER_SSM_I break HYPERPRIVOP_SSM_I
-+#define XEN_HYPER_GET_IVR break HYPERPRIVOP_GET_IVR
-+#define XEN_HYPER_GET_TPR break HYPERPRIVOP_GET_TPR
-+#define XEN_HYPER_SET_TPR break HYPERPRIVOP_SET_TPR
-+#define XEN_HYPER_EOI break HYPERPRIVOP_EOI
-+#define XEN_HYPER_SET_ITM break HYPERPRIVOP_SET_ITM
-+#define XEN_HYPER_THASH break HYPERPRIVOP_THASH
-+#define XEN_HYPER_PTC_GA break HYPERPRIVOP_PTC_GA
-+#define XEN_HYPER_ITR_D break HYPERPRIVOP_ITR_D
-+#define XEN_HYPER_GET_RR break HYPERPRIVOP_GET_RR
-+#define XEN_HYPER_SET_RR break HYPERPRIVOP_SET_RR
-+#define XEN_HYPER_SET_KR break HYPERPRIVOP_SET_KR
-+#define XEN_HYPER_FC break HYPERPRIVOP_FC
-+#define XEN_HYPER_GET_CPUID break HYPERPRIVOP_GET_CPUID
-+#define XEN_HYPER_GET_PMD break HYPERPRIVOP_GET_PMD
-+#define XEN_HYPER_GET_EFLAG break HYPERPRIVOP_GET_EFLAG
-+#define XEN_HYPER_SET_EFLAG break HYPERPRIVOP_SET_EFLAG
-+#define XEN_HYPER_GET_PSR break HYPERPRIVOP_GET_PSR
++/*
++ * Used when interrupts are already enabled or to
++ * shutdown the processor:
++ */
++void halt(void);
+
-+#define XSI_IFS (XSI_BASE + XSI_IFS_OFS)
-+#define XSI_PRECOVER_IFS (XSI_BASE + XSI_PRECOVER_IFS_OFS)
-+#define XSI_IFA (XSI_BASE + XSI_IFA_OFS)
-+#define XSI_ISR (XSI_BASE + XSI_ISR_OFS)
-+#define XSI_IIM (XSI_BASE + XSI_IIM_OFS)
-+#define XSI_ITIR (XSI_BASE + XSI_ITIR_OFS)
-+#define XSI_PSR_I_ADDR (XSI_BASE + XSI_PSR_I_ADDR_OFS)
-+#define XSI_PSR_IC (XSI_BASE + XSI_PSR_IC_OFS)
-+#define XSI_IPSR (XSI_BASE + XSI_IPSR_OFS)
-+#define XSI_IIP (XSI_BASE + XSI_IIP_OFS)
-+#define XSI_B1NAT (XSI_BASE + XSI_B1NATS_OFS)
-+#define XSI_BANK1_R16 (XSI_BASE + XSI_BANK1_R16_OFS)
-+#define XSI_BANKNUM (XSI_BASE + XSI_BANKNUM_OFS)
-+#define XSI_IHA (XSI_BASE + XSI_IHA_OFS)
++#else /* __ASSEMBLY__: */
++# ifdef CONFIG_TRACE_IRQFLAGS
++# define TRACE_IRQS_ON call trace_hardirqs_on_thunk
++# define TRACE_IRQS_OFF call trace_hardirqs_off_thunk
++# else
++# define TRACE_IRQS_ON
++# define TRACE_IRQS_OFF
++# endif
+#endif
+
-+#ifndef __ASSEMBLY__
-+#define XEN_HYPER_SSM_I asm("break %0" : : "i" (HYPERPRIVOP_SSM_I))
-+#define XEN_HYPER_GET_IVR asm("break %0" : : "i" (HYPERPRIVOP_GET_IVR))
-+
-+/************************************************/
-+/* Instructions paravirtualized for correctness */
-+/************************************************/
-+
-+/* "fc" and "thash" are privilege-sensitive instructions, meaning they
-+ * may have different semantics depending on whether they are executed
-+ * at PL0 vs PL!=0. When paravirtualized, these instructions mustn't
-+ * be allowed to execute directly, lest incorrect semantics result. */
-+extern unsigned long xen_fc(unsigned long addr);
-+#define ia64_fc(addr) xen_fc((unsigned long)(addr))
-+extern unsigned long xen_thash(unsigned long addr);
-+#define ia64_thash(addr) xen_thash((unsigned long)(addr))
-+/* Note that "ttag" and "cover" are also privilege-sensitive; "ttag"
-+ * is not currently used (though it may be in a long-format VHPT system!)
-+ * and the semantics of cover only change if psr.ic is off which is very
-+ * rare (and currently non-existent outside of assembly code */
-+
-+/* There are also privilege-sensitive registers. These registers are
-+ * readable at any privilege level but only writable at PL0. */
-+extern unsigned long xen_get_cpuid(int index);
-+#define ia64_get_cpuid(i) xen_get_cpuid(i)
-+extern unsigned long xen_get_pmd(int index);
-+#define ia64_get_pmd(i) xen_get_pmd(i)
-+extern unsigned long xen_get_eflag(void); /* see xen_ia64_getreg */
-+extern void xen_set_eflag(unsigned long); /* see xen_ia64_setreg */
-+
-+/************************************************/
-+/* Instructions paravirtualized for performance */
-+/************************************************/
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/maddr.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/maddr.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,161 @@
++#ifndef _X86_64_MADDR_H
++#define _X86_64_MADDR_H
+
-+/* Xen uses memory-mapped virtual privileged registers for access to many
-+ * performance-sensitive privileged registers. Some, like the processor
-+ * status register (psr), are broken up into multiple memory locations.
-+ * Others, like "pend", are abstractions based on privileged registers.
-+ * "Pend" is guaranteed to be set if reading cr.ivr would return a
-+ * (non-spurious) interrupt. */
-+#define XEN_MAPPEDREGS ((struct mapped_regs *)XMAPPEDREGS_BASE)
-+#define XSI_PSR_I \
-+ (*XEN_MAPPEDREGS->interrupt_mask_addr)
-+#define xen_get_virtual_psr_i() \
-+ (!XSI_PSR_I)
-+#define xen_set_virtual_psr_i(_val) \
-+ ({ XSI_PSR_I = (uint8_t)(_val) ? 0 : 1; })
-+#define xen_set_virtual_psr_ic(_val) \
-+ ({ XEN_MAPPEDREGS->interrupt_collection_enabled = _val ? 1 : 0; })
-+#define xen_get_virtual_pend() \
-+ (*(((uint8_t *)XEN_MAPPEDREGS->interrupt_mask_addr) - 1))
++#include <xen/features.h>
++#include <xen/interface/xen.h>
+
-+/* Hyperprivops are "break" instructions with a well-defined API.
-+ * In particular, the virtual psr.ic bit must be off; in this way
-+ * it is guaranteed to never conflict with a linux break instruction.
-+ * Normally, this is done in a xen stub but this one is frequent enough
-+ * that we inline it */
-+#define xen_hyper_ssm_i() \
-+({ \
-+ XEN_HYPER_SSM_I; \
-+})
++/**** MACHINE <-> PHYSICAL CONVERSION MACROS ****/
++#define INVALID_P2M_ENTRY (~0UL)
++#define FOREIGN_FRAME_BIT (1UL<<63)
++#define FOREIGN_FRAME(m) ((m) | FOREIGN_FRAME_BIT)
+
-+/* turning off interrupts can be paravirtualized simply by writing
-+ * to a memory-mapped virtual psr.i bit (implemented as a 16-bit bool) */
-+#define xen_rsm_i() xen_set_virtual_psr_i(0)
++/* Definitions for machine and pseudophysical addresses. */
++typedef unsigned long paddr_t;
++typedef unsigned long maddr_t;
+
-+/* turning on interrupts is a bit more complicated.. write to the
-+ * memory-mapped virtual psr.i bit first (to avoid race condition),
-+ * then if any interrupts were pending, we have to execute a hyperprivop
-+ * to ensure the pending interrupt gets delivered; else we're done! */
-+#define xen_ssm_i() \
-+({ \
-+ int old = xen_get_virtual_psr_i(); \
-+ if (!old) { \
-+ if (xen_get_virtual_pend()) \
-+ xen_hyper_ssm_i(); \
-+ else \
-+ xen_set_virtual_psr_i(1); \
-+ } \
-+})
++#ifdef CONFIG_XEN
+
-+#define xen_ia64_intrin_local_irq_restore(x) \
-+{ \
-+ if (is_running_on_xen()) { \
-+ if ((x) & IA64_PSR_I) { xen_ssm_i(); } \
-+ else { xen_rsm_i(); } \
-+ } \
-+ else __ia64_intrin_local_irq_restore((x)); \
-+}
++extern unsigned long *phys_to_machine_mapping;
+
-+#define xen_get_psr_i() \
-+( \
-+ (is_running_on_xen()) ? \
-+ (xen_get_virtual_psr_i() ? IA64_PSR_I : 0) \
-+ : __ia64_get_psr_i() \
-+)
++#undef machine_to_phys_mapping
++extern unsigned long *machine_to_phys_mapping;
++extern unsigned int machine_to_phys_order;
+
-+#define xen_ia64_ssm(mask) \
-+{ \
-+ if ((mask)==IA64_PSR_I) { \
-+ if (is_running_on_xen()) { xen_ssm_i(); } \
-+ else { __ia64_ssm(mask); } \
-+ } \
-+ else { __ia64_ssm(mask); } \
++static inline unsigned long pfn_to_mfn(unsigned long pfn)
++{
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return pfn;
++ BUG_ON(end_pfn && pfn >= end_pfn);
++ return phys_to_machine_mapping[pfn] & ~FOREIGN_FRAME_BIT;
+}
+
-+#define xen_ia64_rsm(mask) \
-+{ \
-+ if ((mask)==IA64_PSR_I) { \
-+ if (is_running_on_xen()) { xen_rsm_i(); } \
-+ else { __ia64_rsm(mask); } \
-+ } \
-+ else { __ia64_rsm(mask); } \
++static inline int phys_to_machine_mapping_valid(unsigned long pfn)
++{
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return 1;
++ BUG_ON(end_pfn && pfn >= end_pfn);
++ return (phys_to_machine_mapping[pfn] != INVALID_P2M_ENTRY);
+}
+
++static inline unsigned long mfn_to_pfn(unsigned long mfn)
++{
++ unsigned long pfn;
+
-+/* Although all privileged operations can be left to trap and will
-+ * be properly handled by Xen, some are frequent enough that we use
-+ * hyperprivops for performance. */
-+
-+extern unsigned long xen_get_psr(void);
-+extern unsigned long xen_get_ivr(void);
-+extern unsigned long xen_get_tpr(void);
-+extern void xen_set_itm(unsigned long);
-+extern void xen_set_tpr(unsigned long);
-+extern void xen_eoi(void);
-+extern void xen_set_rr(unsigned long index, unsigned long val);
-+extern unsigned long xen_get_rr(unsigned long index);
-+extern void xen_set_kr(unsigned long index, unsigned long val);
-+extern void xen_ptcga(unsigned long addr, unsigned long size);
-+
-+/* Note: It may look wrong to test for is_running_on_xen() in each case.
-+ * However regnum is always a constant so, as written, the compiler
-+ * eliminates the switch statement, whereas is_running_on_xen() must be
-+ * tested dynamically. */
-+#define xen_ia64_getreg(regnum) \
-+({ \
-+ __u64 ia64_intri_res; \
-+ \
-+ switch(regnum) { \
-+ case _IA64_REG_PSR: \
-+ ia64_intri_res = (is_running_on_xen()) ? \
-+ xen_get_psr() : \
-+ __ia64_getreg(regnum); \
-+ break; \
-+ case _IA64_REG_CR_IVR: \
-+ ia64_intri_res = (is_running_on_xen()) ? \
-+ xen_get_ivr() : \
-+ __ia64_getreg(regnum); \
-+ break; \
-+ case _IA64_REG_CR_TPR: \
-+ ia64_intri_res = (is_running_on_xen()) ? \
-+ xen_get_tpr() : \
-+ __ia64_getreg(regnum); \
-+ break; \
-+ case _IA64_REG_AR_EFLAG: \
-+ ia64_intri_res = (is_running_on_xen()) ? \
-+ xen_get_eflag() : \
-+ __ia64_getreg(regnum); \
-+ break; \
-+ default: \
-+ ia64_intri_res = __ia64_getreg(regnum); \
-+ break; \
-+ } \
-+ ia64_intri_res; \
-+})
-+
-+#define xen_ia64_setreg(regnum,val) \
-+({ \
-+ switch(regnum) { \
-+ case _IA64_REG_AR_KR0 ... _IA64_REG_AR_KR7: \
-+ (is_running_on_xen()) ? \
-+ xen_set_kr((regnum-_IA64_REG_AR_KR0), val) : \
-+ __ia64_setreg(regnum,val); \
-+ break; \
-+ case _IA64_REG_CR_ITM: \
-+ (is_running_on_xen()) ? \
-+ xen_set_itm(val) : \
-+ __ia64_setreg(regnum,val); \
-+ break; \
-+ case _IA64_REG_CR_TPR: \
-+ (is_running_on_xen()) ? \
-+ xen_set_tpr(val) : \
-+ __ia64_setreg(regnum,val); \
-+ break; \
-+ case _IA64_REG_CR_EOI: \
-+ (is_running_on_xen()) ? \
-+ xen_eoi() : \
-+ __ia64_setreg(regnum,val); \
-+ break; \
-+ case _IA64_REG_AR_EFLAG: \
-+ (is_running_on_xen()) ? \
-+ xen_set_eflag(val) : \
-+ __ia64_setreg(regnum,val); \
-+ break; \
-+ default: \
-+ __ia64_setreg(regnum,val); \
-+ break; \
-+ } \
-+})
-+
-+#define ia64_ssm xen_ia64_ssm
-+#define ia64_rsm xen_ia64_rsm
-+#define ia64_intrin_local_irq_restore xen_ia64_intrin_local_irq_restore
-+#define ia64_ptcga xen_ptcga
-+#define ia64_set_rr(index,val) xen_set_rr(index,val)
-+#define ia64_get_rr(index) xen_get_rr(index)
-+#define ia64_getreg xen_ia64_getreg
-+#define ia64_setreg xen_ia64_setreg
-+#define ia64_get_psr_i xen_get_psr_i
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return mfn;
+
-+/* the remainder of these are not performance-sensitive so its
-+ * OK to not paravirtualize and just take a privop trap and emulate */
-+#define ia64_hint __ia64_hint
-+#define ia64_set_pmd __ia64_set_pmd
-+#define ia64_itci __ia64_itci
-+#define ia64_itcd __ia64_itcd
-+#define ia64_itri __ia64_itri
-+#define ia64_itrd __ia64_itrd
-+#define ia64_tpa __ia64_tpa
-+#define ia64_set_ibr __ia64_set_ibr
-+#define ia64_set_pkr __ia64_set_pkr
-+#define ia64_set_pmc __ia64_set_pmc
-+#define ia64_get_ibr __ia64_get_ibr
-+#define ia64_get_pkr __ia64_get_pkr
-+#define ia64_get_pmc __ia64_get_pmc
-+#define ia64_ptce __ia64_ptce
-+#define ia64_ptcl __ia64_ptcl
-+#define ia64_ptri __ia64_ptri
-+#define ia64_ptrd __ia64_ptrd
++ if (unlikely((mfn >> machine_to_phys_order) != 0))
++ return end_pfn;
+
-+#endif /* !__ASSEMBLY__ */
++ /* The array access can fail (e.g., device space beyond end of RAM). */
++ asm (
++ "1: movq %1,%0\n"
++ "2:\n"
++ ".section .fixup,\"ax\"\n"
++ "3: movq %2,%0\n"
++ " jmp 2b\n"
++ ".previous\n"
++ ".section __ex_table,\"a\"\n"
++ " .align 8\n"
++ " .quad 1b,3b\n"
++ ".previous"
++ : "=r" (pfn)
++ : "m" (machine_to_phys_mapping[mfn]), "m" (end_pfn) );
+
-+/* these routines utilize privilege-sensitive or performance-sensitive
-+ * privileged instructions so the code must be replaced with
-+ * paravirtualized versions */
-+#define ia64_pal_halt_light xen_pal_halt_light
-+#define ia64_leave_kernel xen_leave_kernel
-+#define ia64_leave_syscall xen_leave_syscall
-+#define ia64_trace_syscall xen_trace_syscall
-+#define ia64_ret_from_clone xen_ret_from_clone
-+#define ia64_switch_to xen_switch_to
-+#define ia64_pal_call_static xen_pal_call_static
++ return pfn;
++}
+
-+#endif /* _ASM_IA64_XEN_PRIVOP_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/xen/xcom_hcall.h tmp-linux-2.6-xen.patch/include/asm-ia64/xen/xcom_hcall.h
---- pristine-linux-2.6.18.2/include/asm-ia64/xen/xcom_hcall.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/xen/xcom_hcall.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,88 @@
+/*
-+ * Copyright (C) 2006 Tristan Gingold <tristan.gingold at bull.net>, Bull SAS
++ * We detect special mappings in one of two ways:
++ * 1. If the MFN is an I/O page then Xen will set the m2p entry
++ * to be outside our maximum possible pseudophys range.
++ * 2. If the MFN belongs to a different domain then we will certainly
++ * not have MFN in our p2m table. Conversely, if the page is ours,
++ * then we'll have p2m(m2p(MFN))==MFN.
++ * If we detect a special mapping then it doesn't have a 'struct page'.
++ * We force !pfn_valid() by returning an out-of-range pointer.
+ *
-+ * This program is free software; you can redistribute it and/or modify
-+ * it under the terms of the GNU General Public License as published by
-+ * the Free Software Foundation; either version 2 of the License, or
-+ * (at your option) any later version.
-+ *
-+ * This program is distributed in the hope that it will be useful,
-+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ * GNU General Public License for more details.
-+ *
-+ * You should have received a copy of the GNU General Public License
-+ * along with this program; if not, write to the Free Software
-+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ * NB. These checks require that, for any MFN that is not in our reservation,
++ * there is no PFN such that p2m(PFN) == MFN. Otherwise we can get confused if
++ * we are foreign-mapping the MFN, and the other domain as m2p(MFN) == PFN.
++ * Yikes! Various places must poke in INVALID_P2M_ENTRY for safety.
++ *
++ * NB2. When deliberately mapping foreign pages into the p2m table, you *must*
++ * use FOREIGN_FRAME(). This will cause pte_pfn() to choke on it, as we
++ * require. In all the cases we care about, the FOREIGN_FRAME bit is
++ * masked (e.g., pfn_to_mfn()) so behaviour there is correct.
+ */
++static inline unsigned long mfn_to_local_pfn(unsigned long mfn)
++{
++ unsigned long pfn = mfn_to_pfn(mfn);
++ if ((pfn < end_pfn)
++ && !xen_feature(XENFEAT_auto_translated_physmap)
++ && (phys_to_machine_mapping[pfn] != mfn))
++ return end_pfn; /* force !pfn_valid() */
++ return pfn;
++}
+
-+#ifndef _LINUX_XENCOMM_HCALL_H_
-+#define _LINUX_XENCOMM_HCALL_H_
-+
-+/* These function creates inline descriptor for the parameters and
-+ calls the corresponding xencomm_arch_hypercall_X.
-+ Architectures should defines HYPERVISOR_xxx as xencomm_hypercall_xxx unless
-+ they want to use their own wrapper. */
-+extern int xencomm_hypercall_console_io(int cmd, int count, char *str);
++static inline void set_phys_to_machine(unsigned long pfn, unsigned long mfn)
++{
++ BUG_ON(end_pfn && pfn >= end_pfn);
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY);
++ return;
++ }
++ phys_to_machine_mapping[pfn] = mfn;
++}
+
-+extern int xencomm_hypercall_event_channel_op(int cmd, void *op);
++static inline maddr_t phys_to_machine(paddr_t phys)
++{
++ maddr_t machine = pfn_to_mfn(phys >> PAGE_SHIFT);
++ machine = (machine << PAGE_SHIFT) | (phys & ~PAGE_MASK);
++ return machine;
++}
+
-+extern int xencomm_hypercall_xen_version(int cmd, void *arg);
++static inline paddr_t machine_to_phys(maddr_t machine)
++{
++ paddr_t phys = mfn_to_pfn(machine >> PAGE_SHIFT);
++ phys = (phys << PAGE_SHIFT) | (machine & ~PAGE_MASK);
++ return phys;
++}
+
-+extern int xencomm_hypercall_physdev_op(int cmd, void *op);
++static inline paddr_t pte_phys_to_machine(paddr_t phys)
++{
++ maddr_t machine;
++ machine = pfn_to_mfn((phys & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT);
++ machine = (machine << PAGE_SHIFT) | (phys & ~PHYSICAL_PAGE_MASK);
++ return machine;
++}
+
-+extern int xencomm_hypercall_grant_table_op(unsigned int cmd, void *op,
-+ unsigned int count);
++static inline paddr_t pte_machine_to_phys(maddr_t machine)
++{
++ paddr_t phys;
++ phys = mfn_to_pfn((machine & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT);
++ phys = (phys << PAGE_SHIFT) | (machine & ~PHYSICAL_PAGE_MASK);
++ return phys;
++}
+
-+extern int xencomm_hypercall_sched_op(int cmd, void *arg);
++#define __pte_ma(x) ((pte_t) { (x) } )
++#define pfn_pte_ma(pfn, prot) __pte_ma((((pfn) << PAGE_SHIFT) | pgprot_val(prot)) & __supported_pte_mask)
+
-+extern int xencomm_hypercall_multicall(void *call_list, int nr_calls);
++#else /* !CONFIG_XEN */
+
-+extern int xencomm_hypercall_callback_op(int cmd, void *arg);
++#define pfn_to_mfn(pfn) (pfn)
++#define mfn_to_pfn(mfn) (mfn)
++#define mfn_to_local_pfn(mfn) (mfn)
++#define set_phys_to_machine(pfn, mfn) ((void)0)
++#define phys_to_machine_mapping_valid(pfn) (1)
++#define phys_to_machine(phys) ((maddr_t)(phys))
++#define machine_to_phys(mach) ((paddr_t)(mach))
++#define pfn_pte_ma(pfn, prot) pfn_pte(pfn, prot)
++#define __pte_ma(x) __pte(x)
+
-+extern int xencomm_hypercall_memory_op(unsigned int cmd, void *arg);
++#endif /* !CONFIG_XEN */
+
-+extern unsigned long xencomm_hypercall_hvm_op(int cmd, void *arg);
++/* VIRT <-> MACHINE conversion */
++#define virt_to_machine(v) (phys_to_machine(__pa(v)))
++#define virt_to_mfn(v) (pfn_to_mfn(__pa(v) >> PAGE_SHIFT))
++#define mfn_to_virt(m) (__va(mfn_to_pfn(m) << PAGE_SHIFT))
+
-+extern int xencomm_hypercall_suspend(unsigned long srec);
++#endif /* _X86_64_MADDR_H */
+
-+extern int xencomm_hypercall_xenoprof_op(int op, void *arg);
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/mmu.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/mmu.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,38 @@
++#ifndef __x86_64_MMU_H
++#define __x86_64_MMU_H
+
-+extern int xencomm_hypercall_perfmon_op(unsigned long cmd, void* arg,
-+ unsigned long count);
++#include <linux/spinlock.h>
++#include <asm/semaphore.h>
+
-+extern long xencomm_hypercall_vcpu_op(int cmd, int cpu, void *arg);
++/*
++ * The x86_64 doesn't have a mmu context, but
++ * we put the segment information here.
++ *
++ * cpu_vm_mask is used to optimize ldt flushing.
++ */
++typedef struct {
++ void *ldt;
++ rwlock_t ldtlock;
++ int size;
++ struct semaphore sem;
++#ifdef CONFIG_XEN
++ unsigned pinned:1;
++ unsigned has_foreign_mappings:1;
++ struct list_head unpinned;
++#endif
++} mm_context_t;
+
-+/* Using mini xencomm. */
-+extern int xencomm_mini_hypercall_console_io(int cmd, int count, char *str);
++#ifdef CONFIG_XEN
++extern struct list_head mm_unpinned;
++extern spinlock_t mm_unpinned_lock;
+
-+extern int xencomm_mini_hypercall_event_channel_op(int cmd, void *op);
++/* mm/memory.c:exit_mmap hook */
++extern void _arch_exit_mmap(struct mm_struct *mm);
++#define arch_exit_mmap(_mm) _arch_exit_mmap(_mm)
+
-+extern int xencomm_mini_hypercall_xen_version(int cmd, void *arg);
++/* kernel/fork.c:dup_mmap hook */
++extern void _arch_dup_mmap(struct mm_struct *mm);
++#define arch_dup_mmap(mm, oldmm) ((void)(oldmm), _arch_dup_mmap(mm))
++#endif
+
-+extern int xencomm_mini_hypercall_physdev_op(int cmd, void *op);
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/mmu_context.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/mmu_context.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,136 @@
++#ifndef __X86_64_MMU_CONTEXT_H
++#define __X86_64_MMU_CONTEXT_H
+
-+extern int xencomm_mini_hypercall_grant_table_op(unsigned int cmd, void *op,
-+ unsigned int count);
++#include <asm/desc.h>
++#include <asm/atomic.h>
++#include <asm/pgalloc.h>
++#include <asm/page.h>
++#include <asm/pda.h>
++#include <asm/pgtable.h>
++#include <asm/tlbflush.h>
+
-+extern int xencomm_mini_hypercall_sched_op(int cmd, void *arg);
++/*
++ * possibly do the LDT unload here?
++ */
++int init_new_context(struct task_struct *tsk, struct mm_struct *mm);
++void destroy_context(struct mm_struct *mm);
+
-+extern int xencomm_mini_hypercall_multicall(void *call_list, int nr_calls);
++static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk)
++{
++#if defined(CONFIG_SMP) && !defined(CONFIG_XEN)
++ if (read_pda(mmu_state) == TLBSTATE_OK)
++ write_pda(mmu_state, TLBSTATE_LAZY);
++#endif
++}
+
-+extern int xencomm_mini_hypercall_callback_op(int cmd, void *arg);
++#define prepare_arch_switch(next) __prepare_arch_switch()
+
-+extern int xencomm_mini_hypercall_memory_op(unsigned int cmd, void *arg);
++static inline void __prepare_arch_switch(void)
++{
++ /*
++ * Save away %es, %ds, %fs and %gs. Must happen before reload
++ * of cr3/ldt (i.e., not in __switch_to).
++ */
++ __asm__ __volatile__ (
++ "mov %%es,%0 ; mov %%ds,%1 ; mov %%fs,%2 ; mov %%gs,%3"
++ : "=m" (current->thread.es),
++ "=m" (current->thread.ds),
++ "=m" (current->thread.fsindex),
++ "=m" (current->thread.gsindex) );
+
-+extern unsigned long xencomm_mini_hypercall_hvm_op(int cmd, void *arg);
++ if (current->thread.ds)
++ __asm__ __volatile__ ( "movl %0,%%ds" : : "r" (0) );
+
-+extern int xencomm_mini_hypercall_xenoprof_op(int op, void *arg);
++ if (current->thread.es)
++ __asm__ __volatile__ ( "movl %0,%%es" : : "r" (0) );
+
-+extern int xencomm_mini_hypercall_perfmon_op(unsigned long cmd, void* arg,
-+ unsigned long count);
++ if (current->thread.fsindex) {
++ __asm__ __volatile__ ( "movl %0,%%fs" : : "r" (0) );
++ current->thread.fs = 0;
++ }
+
-+/* For privcmd. Locally declare argument type to avoid include storm.
-+ Type coherency will be checked within privcmd.c */
-+struct privcmd_hypercall;
-+extern int privcmd_hypercall(struct privcmd_hypercall *hypercall);
++ if (current->thread.gsindex) {
++ load_gs_index(0);
++ current->thread.gs = 0;
++ }
++}
+
-+#endif /* _LINUX_XENCOMM_HCALL_H_ */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/xen/xencomm.h tmp-linux-2.6-xen.patch/include/asm-ia64/xen/xencomm.h
---- pristine-linux-2.6.18.2/include/asm-ia64/xen/xencomm.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/xen/xencomm.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,60 @@
-+/*
-+ * Copyright (C) 2006 Hollis Blanchard <hollisb at us.ibm.com>, IBM Corporation
-+ *
-+ * This program is free software; you can redistribute it and/or modify
-+ * it under the terms of the GNU General Public License as published by
-+ * the Free Software Foundation; either version 2 of the License, or
-+ * (at your option) any later version.
-+ *
-+ * This program is distributed in the hope that it will be useful,
-+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ * GNU General Public License for more details.
-+ *
-+ * You should have received a copy of the GNU General Public License
-+ * along with this program; if not, write to the Free Software
-+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
-+ */
++extern void mm_pin(struct mm_struct *mm);
++extern void mm_unpin(struct mm_struct *mm);
++void mm_pin_all(void);
+
-+#ifndef _LINUX_XENCOMM_H_
-+#define _LINUX_XENCOMM_H_
++static inline void load_cr3(pgd_t *pgd)
++{
++ asm volatile("movq %0,%%cr3" :: "r" (phys_to_machine(__pa(pgd))) :
++ "memory");
++}
+
-+#include <xen/interface/xencomm.h>
++static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
++ struct task_struct *tsk)
++{
++ unsigned cpu = smp_processor_id();
++ struct mmuext_op _op[3], *op = _op;
+
-+#define XENCOMM_MINI_ADDRS 3
-+struct xencomm_mini {
-+ struct xencomm_desc _desc;
-+ uint64_t address[XENCOMM_MINI_ADDRS];
-+};
++ if (likely(prev != next)) {
++ BUG_ON(!xen_feature(XENFEAT_writable_page_tables) &&
++ !next->context.pinned);
+
-+/* Must be called before any hypercall. */
-+extern void xencomm_init (void);
++ /* stop flush ipis for the previous mm */
++ cpu_clear(cpu, prev->cpu_vm_mask);
++#if defined(CONFIG_SMP) && !defined(CONFIG_XEN)
++ write_pda(mmu_state, TLBSTATE_OK);
++ write_pda(active_mm, next);
++#endif
++ cpu_set(cpu, next->cpu_vm_mask);
+
-+/* To avoid additionnal virt to phys conversion, an opaque structure is
-+ presented. */
-+struct xencomm_handle;
++ /* load_cr3(next->pgd) */
++ op->cmd = MMUEXT_NEW_BASEPTR;
++ op->arg1.mfn = pfn_to_mfn(__pa(next->pgd) >> PAGE_SHIFT);
++ op++;
+
-+extern int xencomm_create(void *buffer, unsigned long bytes,
-+ struct xencomm_handle **desc, gfp_t type);
-+extern void xencomm_free(struct xencomm_handle *desc);
++ /* xen_new_user_pt(__pa(__user_pgd(next->pgd))) */
++ op->cmd = MMUEXT_NEW_USER_BASEPTR;
++ op->arg1.mfn = pfn_to_mfn(__pa(__user_pgd(next->pgd)) >> PAGE_SHIFT);
++ op++;
++
++ if (unlikely(next->context.ldt != prev->context.ldt)) {
++ /* load_LDT_nolock(&next->context, cpu) */
++ op->cmd = MMUEXT_SET_LDT;
++ op->arg1.linear_addr = (unsigned long)next->context.ldt;
++ op->arg2.nr_ents = next->context.size;
++ op++;
++ }
+
-+extern int xencomm_create_mini(struct xencomm_mini *area, int *nbr_area,
-+ void *buffer, unsigned long bytes,
-+ struct xencomm_handle **ret);
++ BUG_ON(HYPERVISOR_mmuext_op(_op, op-_op, NULL, DOMID_SELF));
++ }
++#if defined(CONFIG_SMP) && !defined(CONFIG_XEN)
++ else {
++ write_pda(mmu_state, TLBSTATE_OK);
++ if (read_pda(active_mm) != next)
++ out_of_line_bug();
++ if (!cpu_test_and_set(cpu, next->cpu_vm_mask)) {
++ /* We were in lazy tlb mode and leave_mm disabled
++ * tlb flush IPI delivery. We must reload CR3
++ * to make sure to use no freed page tables.
++ */
++ load_cr3(next->pgd);
++ xen_new_user_pt(__pa(__user_pgd(next->pgd)));
++ load_LDT_nolock(&next->context, cpu);
++ }
++ }
++#endif
++}
+
-+/* Translate virtual address to physical address. */
-+extern unsigned long xencomm_vaddr_to_paddr(unsigned long vaddr);
++#define deactivate_mm(tsk,mm) do { \
++ load_gs_index(0); \
++ asm volatile("movl %0,%%fs"::"r"(0)); \
++} while(0)
+
-+/* Inline version. To be used only on linear space (kernel space). */
-+static inline struct xencomm_handle *
-+xencomm_create_inline(void *buffer)
++static inline void activate_mm(struct mm_struct *prev, struct mm_struct *next)
+{
-+ unsigned long paddr;
-+
-+ paddr = xencomm_vaddr_to_paddr((unsigned long)buffer);
-+ return (struct xencomm_handle *)(paddr | XENCOMM_INLINE_FLAG);
++ if (!next->context.pinned)
++ mm_pin(next);
++ switch_mm(prev, next, NULL);
+}
+
-+#define xen_guest_handle(hnd) ((hnd).p)
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/msr.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/msr.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,399 @@
++#ifndef X86_64_MSR_H
++#define X86_64_MSR_H 1
+
-+#endif /* _LINUX_XENCOMM_H_ */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-ia64/xenoprof.h tmp-linux-2.6-xen.patch/include/asm-ia64/xenoprof.h
---- pristine-linux-2.6.18.2/include/asm-ia64/xenoprof.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-ia64/xenoprof.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,48 @@
-+/******************************************************************************
-+ * asm-ia64/xenoprof.h
-+ *
-+ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
-+ * VA Linux Systems Japan K.K.
-+ *
-+ * This program is free software; you can redistribute it and/or modify
-+ * it under the terms of the GNU General Public License as published by
-+ * the Free Software Foundation; either version 2 of the License, or
-+ * (at your option) any later version.
-+ *
-+ * This program is distributed in the hope that it will be useful,
-+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ * GNU General Public License for more details.
-+ *
-+ * You should have received a copy of the GNU General Public License
-+ * along with this program; if not, write to the Free Software
-+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
-+ *
++#ifndef __ASSEMBLY__
++/*
++ * Access to machine-specific registers (available on 586 and better only)
++ * Note: the rd* operations modify the parameters directly (without using
++ * pointer indirection), this allows gcc to optimize better
+ */
-+#ifndef __ASM_XENOPROF_H__
-+#define __ASM_XENOPROF_H__
-+#ifdef CONFIG_XEN
+
-+#undef HAVE_XENOPROF_CREATE_FILES
++#define rdmsr(msr,val1,val2) \
++ __asm__ __volatile__("rdmsr" \
++ : "=a" (val1), "=d" (val2) \
++ : "c" (msr))
+
-+struct xenoprof_init;
-+void xenoprof_arch_init_counter(struct xenoprof_init *init);
-+void xenoprof_arch_counter(void);
-+void xenoprof_arch_start(void);
-+void xenoprof_arch_stop(void);
+
-+struct xenoprof_arch_shared_buffer {
-+ struct resource* res;
-+};
++#define rdmsrl(msr,val) do { unsigned long a__,b__; \
++ __asm__ __volatile__("rdmsr" \
++ : "=a" (a__), "=d" (b__) \
++ : "c" (msr)); \
++ val = a__ | (b__<<32); \
++} while(0)
+
-+struct xenoprof_shared_buffer;
-+void xenoprof_arch_unmap_shared_buffer(struct xenoprof_shared_buffer* sbuf);
-+struct xenoprof_get_buffer;
-+int xenoprof_arch_map_shared_buffer(struct xenoprof_get_buffer* get_buffer,
-+ struct xenoprof_shared_buffer* sbuf);
-+struct xenoprof_passive;
-+int xenoprof_arch_set_passive(struct xenoprof_passive* pdomain,
-+ struct xenoprof_shared_buffer* sbuf);
++#define wrmsr(msr,val1,val2) \
++ __asm__ __volatile__("wrmsr" \
++ : /* no outputs */ \
++ : "c" (msr), "a" (val1), "d" (val2))
+
-+#endif /* CONFIG_XEN */
-+#endif /* __ASM_XENOPROF_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-um/page.h tmp-linux-2.6-xen.patch/include/asm-um/page.h
---- pristine-linux-2.6.18.2/include/asm-um/page.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-um/page.h 2007-07-30 16:35:13.000000000 +0200
-@@ -114,7 +114,7 @@ extern unsigned long uml_physmem;
- extern struct page *arch_validate(struct page *page, gfp_t mask, int order);
- #define HAVE_ARCH_VALIDATE
-
--extern void arch_free_page(struct page *page, int order);
-+extern int arch_free_page(struct page *page, int order);
- #define HAVE_ARCH_FREE_PAGE
-
- #include <asm-generic/memory_model.h>
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/apic.h tmp-linux-2.6-xen.patch/include/asm-x86_64/apic.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/apic.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/apic.h 2007-07-30 16:35:13.000000000 +0200
-@@ -98,11 +98,13 @@ extern void setup_APIC_extened_lvt(unsig
- extern int disable_timer_pin_1;
-
-
-+#ifndef CONFIG_XEN
- void smp_send_timer_broadcast_ipi(void);
- void switch_APIC_timer_to_ipi(void *cpumask);
- void switch_ipi_to_APIC_timer(void *cpumask);
-
- #define ARCH_APICTIMER_STOPS_ON_C3 1
-+#endif
-
- #endif /* CONFIG_X86_LOCAL_APIC */
-
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/kexec.h tmp-linux-2.6-xen.patch/include/asm-x86_64/kexec.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/kexec.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/kexec.h 2007-07-30 16:35:13.000000000 +0200
-@@ -1,6 +1,27 @@
- #ifndef _X86_64_KEXEC_H
- #define _X86_64_KEXEC_H
-
-+#define PA_CONTROL_PAGE 0
-+#define VA_CONTROL_PAGE 1
-+#define PA_PGD 2
-+#define VA_PGD 3
-+#define PA_PUD_0 4
-+#define VA_PUD_0 5
-+#define PA_PMD_0 6
-+#define VA_PMD_0 7
-+#define PA_PTE_0 8
-+#define VA_PTE_0 9
-+#define PA_PUD_1 10
-+#define VA_PUD_1 11
-+#define PA_PMD_1 12
-+#define VA_PMD_1 13
-+#define PA_PTE_1 14
-+#define VA_PTE_1 15
-+#define PA_TABLE_PAGE 16
-+#define PAGES_NR 17
++#define wrmsrl(msr,val) wrmsr(msr,(__u32)((__u64)(val)),((__u64)(val))>>32)
+
-+#ifndef __ASSEMBLY__
++/* wrmsr with exception handling */
++#define wrmsr_safe(msr,a,b) ({ int ret__; \
++ asm volatile("2: wrmsr ; xorl %0,%0\n" \
++ "1:\n\t" \
++ ".section .fixup,\"ax\"\n\t" \
++ "3: movl %4,%0 ; jmp 1b\n\t" \
++ ".previous\n\t" \
++ ".section __ex_table,\"a\"\n" \
++ " .align 8\n\t" \
++ " .quad 2b,3b\n\t" \
++ ".previous" \
++ : "=a" (ret__) \
++ : "c" (msr), "0" (a), "d" (b), "i" (-EFAULT)); \
++ ret__; })
+
- #include <linux/string.h>
-
- #include <asm/page.h>
-@@ -64,4 +85,25 @@ static inline void crash_setup_regs(stru
- newregs->rip = (unsigned long)current_text_addr();
- }
- }
++#define checking_wrmsrl(msr,val) wrmsr_safe(msr,(u32)(val),(u32)((val)>>32))
+
-+NORET_TYPE void
-+relocate_kernel(unsigned long indirection_page,
-+ unsigned long page_list,
-+ unsigned long start_address) ATTRIB_NORET;
++#define rdmsr_safe(msr,a,b) \
++ ({ int ret__; \
++ asm volatile ("1: rdmsr\n" \
++ "2:\n" \
++ ".section .fixup,\"ax\"\n" \
++ "3: movl %4,%0\n" \
++ " jmp 2b\n" \
++ ".previous\n" \
++ ".section __ex_table,\"a\"\n" \
++ " .align 8\n" \
++ " .quad 1b,3b\n" \
++ ".previous":"=&bDS" (ret__), "=a"(*(a)), "=d"(*(b))\
++ :"c"(msr), "i"(-EIO), "0"(0)); \
++ ret__; })
+
-+/* Under Xen we need to work with machine addresses. These macros give the
-+ * machine address of a certain page to the generic kexec code instead of
-+ * the pseudo physical address which would be given by the default macros.
-+ */
++#define rdtsc(low,high) \
++ __asm__ __volatile__("rdtsc" : "=a" (low), "=d" (high))
+
-+#ifdef CONFIG_XEN
-+#define KEXEC_ARCH_HAS_PAGE_MACROS
-+#define kexec_page_to_pfn(page) pfn_to_mfn(page_to_pfn(page))
-+#define kexec_pfn_to_page(pfn) pfn_to_page(mfn_to_pfn(pfn))
-+#define kexec_virt_to_phys(addr) virt_to_machine(addr)
-+#define kexec_phys_to_virt(addr) phys_to_virt(machine_to_phys(addr))
-+#endif
++#define rdtscl(low) \
++ __asm__ __volatile__ ("rdtsc" : "=a" (low) : : "edx")
+
-+#endif /* __ASSEMBLY__ */
++#define rdtscll(val) do { \
++ unsigned int __a,__d; \
++ asm volatile("rdtsc" : "=a" (__a), "=d" (__d)); \
++ (val) = ((unsigned long)__a) | (((unsigned long)__d)<<32); \
++} while(0)
+
- #endif /* _X86_64_KEXEC_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/agp.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/agp.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/agp.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/agp.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,35 @@
-+#ifndef AGP_H
-+#define AGP_H 1
++#define write_tsc(val1,val2) wrmsr(0x10, val1, val2)
+
-+#include <asm/cacheflush.h>
-+#include <asm/system.h>
++#define rdpmc(counter,low,high) \
++ __asm__ __volatile__("rdpmc" \
++ : "=a" (low), "=d" (high) \
++ : "c" (counter))
++
++static inline void cpuid(int op, unsigned int *eax, unsigned int *ebx,
++ unsigned int *ecx, unsigned int *edx)
++{
++ __asm__(XEN_CPUID
++ : "=a" (*eax),
++ "=b" (*ebx),
++ "=c" (*ecx),
++ "=d" (*edx)
++ : "0" (op));
++}
++
++/* Some CPUID calls want 'count' to be placed in ecx */
++static inline void cpuid_count(int op, int count, int *eax, int *ebx, int *ecx,
++ int *edx)
++{
++ __asm__(XEN_CPUID
++ : "=a" (*eax),
++ "=b" (*ebx),
++ "=c" (*ecx),
++ "=d" (*edx)
++ : "0" (op), "c" (count));
++}
+
+/*
-+ * Functions to keep the agpgart mappings coherent.
-+ * The GART gives the CPU a physical alias of memory. The alias is
-+ * mapped uncacheable. Make sure there are no conflicting mappings
-+ * with different cachability attributes for the same page.
++ * CPUID functions returning a single datum
+ */
++static inline unsigned int cpuid_eax(unsigned int op)
++{
++ unsigned int eax;
+
-+int map_page_into_agp(struct page *page);
-+int unmap_page_from_agp(struct page *page);
-+#define flush_agp_mappings() global_flush_tlb()
++ __asm__(XEN_CPUID
++ : "=a" (eax)
++ : "0" (op)
++ : "bx", "cx", "dx");
++ return eax;
++}
++static inline unsigned int cpuid_ebx(unsigned int op)
++{
++ unsigned int eax, ebx;
+
-+/* Could use CLFLUSH here if the cpu supports it. But then it would
-+ need to be called for each cacheline of the whole page so it may not be
-+ worth it. Would need a page for it. */
-+#define flush_agp_cache() wbinvd()
++ __asm__(XEN_CPUID
++ : "=a" (eax), "=b" (ebx)
++ : "0" (op)
++ : "cx", "dx" );
++ return ebx;
++}
++static inline unsigned int cpuid_ecx(unsigned int op)
++{
++ unsigned int eax, ecx;
+
-+/* Convert a physical address to an address suitable for the GART. */
-+#define phys_to_gart(x) phys_to_machine(x)
-+#define gart_to_phys(x) machine_to_phys(x)
++ __asm__(XEN_CPUID
++ : "=a" (eax), "=c" (ecx)
++ : "0" (op)
++ : "bx", "dx" );
++ return ecx;
++}
++static inline unsigned int cpuid_edx(unsigned int op)
++{
++ unsigned int eax, edx;
++
++ __asm__(XEN_CPUID
++ : "=a" (eax), "=d" (edx)
++ : "0" (op)
++ : "bx", "cx");
++ return edx;
++}
++
++#define MSR_IA32_UCODE_WRITE 0x79
++#define MSR_IA32_UCODE_REV 0x8b
+
-+/* GATT allocation. Returns/accepts GATT kernel virtual address. */
-+#define alloc_gatt_pages(order) ({ \
-+ char *_t; dma_addr_t _d; \
-+ _t = dma_alloc_coherent(NULL,PAGE_SIZE<<(order),&_d,GFP_KERNEL); \
-+ _t; })
-+#define free_gatt_pages(table, order) \
-+ dma_free_coherent(NULL,PAGE_SIZE<<(order),(table),virt_to_bus(table))
+
+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/arch_hooks.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/arch_hooks.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/arch_hooks.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/arch_hooks.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,27 @@
-+#ifndef _ASM_ARCH_HOOKS_H
-+#define _ASM_ARCH_HOOKS_H
+
-+#include <linux/interrupt.h>
++/* AMD/K8 specific MSRs */
++#define MSR_EFER 0xc0000080 /* extended feature register */
++#define MSR_STAR 0xc0000081 /* legacy mode SYSCALL target */
++#define MSR_LSTAR 0xc0000082 /* long mode SYSCALL target */
++#define MSR_CSTAR 0xc0000083 /* compatibility mode SYSCALL target */
++#define MSR_SYSCALL_MASK 0xc0000084 /* EFLAGS mask for syscall */
++#define MSR_FS_BASE 0xc0000100 /* 64bit GS base */
++#define MSR_GS_BASE 0xc0000101 /* 64bit FS base */
++#define MSR_KERNEL_GS_BASE 0xc0000102 /* SwapGS GS shadow (or USER_GS from kernel) */
++/* EFER bits: */
++#define _EFER_SCE 0 /* SYSCALL/SYSRET */
++#define _EFER_LME 8 /* Long mode enable */
++#define _EFER_LMA 10 /* Long mode active (read-only) */
++#define _EFER_NX 11 /* No execute enable */
+
-+/*
-+ * linux/include/asm/arch_hooks.h
-+ *
-+ * define the architecture specific hooks
-+ */
++#define EFER_SCE (1<<_EFER_SCE)
++#define EFER_LME (1<<_EFER_LME)
++#define EFER_LMA (1<<_EFER_LMA)
++#define EFER_NX (1<<_EFER_NX)
++
++/* Intel MSRs. Some also available on other CPUs */
++#define MSR_IA32_TSC 0x10
++#define MSR_IA32_PLATFORM_ID 0x17
++
++#define MSR_IA32_PERFCTR0 0xc1
++#define MSR_IA32_PERFCTR1 0xc2
+
-+/* these aren't arch hooks, they are generic routines
-+ * that can be used by the hooks */
-+extern void init_ISA_irqs(void);
-+extern void apic_intr_init(void);
-+extern void smp_intr_init(void);
-+extern irqreturn_t timer_interrupt(int irq, void *dev_id, struct pt_regs *regs);
++#define MSR_MTRRcap 0x0fe
++#define MSR_IA32_BBL_CR_CTL 0x119
+
-+/* these are the defined hooks */
-+extern void intr_init_hook(void);
-+extern void pre_intr_init_hook(void);
-+extern void pre_setup_arch_hook(void);
-+extern void trap_init_hook(void);
-+extern void time_init_hook(void);
-+extern void mca_nmi_hook(void);
++#define MSR_IA32_SYSENTER_CS 0x174
++#define MSR_IA32_SYSENTER_ESP 0x175
++#define MSR_IA32_SYSENTER_EIP 0x176
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/bootsetup.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/bootsetup.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/bootsetup.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/bootsetup.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,42 @@
++#define MSR_IA32_MCG_CAP 0x179
++#define MSR_IA32_MCG_STATUS 0x17a
++#define MSR_IA32_MCG_CTL 0x17b
+
-+#ifndef _X86_64_BOOTSETUP_H
-+#define _X86_64_BOOTSETUP_H 1
++#define MSR_IA32_EVNTSEL0 0x186
++#define MSR_IA32_EVNTSEL1 0x187
+
-+#define BOOT_PARAM_SIZE 4096
-+extern char x86_boot_params[BOOT_PARAM_SIZE];
++#define MSR_IA32_DEBUGCTLMSR 0x1d9
++#define MSR_IA32_LASTBRANCHFROMIP 0x1db
++#define MSR_IA32_LASTBRANCHTOIP 0x1dc
++#define MSR_IA32_LASTINTFROMIP 0x1dd
++#define MSR_IA32_LASTINTTOIP 0x1de
+
-+/*
-+ * This is set up by the setup-routine at boot-time
-+ */
-+#define PARAM ((unsigned char *)x86_boot_params)
-+#define SCREEN_INFO (*(struct screen_info *) (PARAM+0))
-+#define EXT_MEM_K (*(unsigned short *) (PARAM+2))
-+#define ALT_MEM_K (*(unsigned int *) (PARAM+0x1e0))
-+#define E820_MAP_NR (*(char*) (PARAM+E820NR))
-+#define E820_MAP ((struct e820entry *) (PARAM+E820MAP))
-+#define APM_BIOS_INFO (*(struct apm_bios_info *) (PARAM+0x40))
-+#define DRIVE_INFO (*(struct drive_info_struct *) (PARAM+0x80))
-+#define SYS_DESC_TABLE (*(struct sys_desc_table_struct*)(PARAM+0xa0))
-+#define MOUNT_ROOT_RDONLY (*(unsigned short *) (PARAM+0x1F2))
-+#define RAMDISK_FLAGS (*(unsigned short *) (PARAM+0x1F8))
-+#define SAVED_VIDEO_MODE (*(unsigned short *) (PARAM+0x1FA))
-+#define ORIG_ROOT_DEV (*(unsigned short *) (PARAM+0x1FC))
-+#define AUX_DEVICE_INFO (*(unsigned char *) (PARAM+0x1FF))
-+#define LOADER_TYPE (*(unsigned char *) (PARAM+0x210))
-+#define KERNEL_START (*(unsigned int *) (PARAM+0x214))
++#define MSR_MTRRfix64K_00000 0x250
++#define MSR_MTRRfix16K_80000 0x258
++#define MSR_MTRRfix16K_A0000 0x259
++#define MSR_MTRRfix4K_C0000 0x268
++#define MSR_MTRRfix4K_C8000 0x269
++#define MSR_MTRRfix4K_D0000 0x26a
++#define MSR_MTRRfix4K_D8000 0x26b
++#define MSR_MTRRfix4K_E0000 0x26c
++#define MSR_MTRRfix4K_E8000 0x26d
++#define MSR_MTRRfix4K_F0000 0x26e
++#define MSR_MTRRfix4K_F8000 0x26f
++#define MSR_MTRRdefType 0x2ff
+
-+#define INITRD_START (__pa(xen_start_info->mod_start))
-+#define INITRD_SIZE (xen_start_info->mod_len)
-+#define EDID_INFO (*(struct edid_info *) (PARAM+0x440))
++#define MSR_IA32_MC0_CTL 0x400
++#define MSR_IA32_MC0_STATUS 0x401
++#define MSR_IA32_MC0_ADDR 0x402
++#define MSR_IA32_MC0_MISC 0x403
+
-+#define EDD_NR (*(unsigned char *) (PARAM+EDDNR))
-+#define EDD_MBR_SIG_NR (*(unsigned char *) (PARAM+EDD_MBR_SIG_NR_BUF))
-+#define EDD_MBR_SIGNATURE ((unsigned int *) (PARAM+EDD_MBR_SIG_BUF))
-+#define EDD_BUF ((struct edd_info *) (PARAM+EDDBUF))
-+#define COMMAND_LINE saved_command_line
++#define MSR_P6_PERFCTR0 0xc1
++#define MSR_P6_PERFCTR1 0xc2
++#define MSR_P6_EVNTSEL0 0x186
++#define MSR_P6_EVNTSEL1 0x187
+
-+#define RAMDISK_IMAGE_START_MASK 0x07FF
-+#define RAMDISK_PROMPT_FLAG 0x8000
-+#define RAMDISK_LOAD_FLAG 0x4000
++/* K7/K8 MSRs. Not complete. See the architecture manual for a more complete list. */
++#define MSR_K7_EVNTSEL0 0xC0010000
++#define MSR_K7_PERFCTR0 0xC0010004
++#define MSR_K7_EVNTSEL1 0xC0010001
++#define MSR_K7_PERFCTR1 0xC0010005
++#define MSR_K7_EVNTSEL2 0xC0010002
++#define MSR_K7_PERFCTR2 0xC0010006
++#define MSR_K7_EVNTSEL3 0xC0010003
++#define MSR_K7_PERFCTR3 0xC0010007
++#define MSR_K8_TOP_MEM1 0xC001001A
++#define MSR_K8_TOP_MEM2 0xC001001D
++#define MSR_K8_SYSCFG 0xC0010010
++#define MSR_K8_HWCR 0xC0010015
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/desc.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/desc.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/desc.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/desc.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,263 @@
-+/* Written 2000 by Andi Kleen */
-+#ifndef __ARCH_DESC_H
-+#define __ARCH_DESC_H
++/* K6 MSRs */
++#define MSR_K6_EFER 0xC0000080
++#define MSR_K6_STAR 0xC0000081
++#define MSR_K6_WHCR 0xC0000082
++#define MSR_K6_UWCCR 0xC0000085
++#define MSR_K6_PSOR 0xC0000087
++#define MSR_K6_PFIR 0xC0000088
+
-+#include <linux/threads.h>
-+#include <asm/ldt.h>
++/* Centaur-Hauls/IDT defined MSRs. */
++#define MSR_IDT_FCR1 0x107
++#define MSR_IDT_FCR2 0x108
++#define MSR_IDT_FCR3 0x109
++#define MSR_IDT_FCR4 0x10a
+
-+#ifndef __ASSEMBLY__
++#define MSR_IDT_MCR0 0x110
++#define MSR_IDT_MCR1 0x111
++#define MSR_IDT_MCR2 0x112
++#define MSR_IDT_MCR3 0x113
++#define MSR_IDT_MCR4 0x114
++#define MSR_IDT_MCR5 0x115
++#define MSR_IDT_MCR6 0x116
++#define MSR_IDT_MCR7 0x117
++#define MSR_IDT_MCR_CTRL 0x120
+
-+#include <linux/string.h>
-+#include <linux/smp.h>
++/* VIA Cyrix defined MSRs*/
++#define MSR_VIA_FCR 0x1107
++#define MSR_VIA_LONGHAUL 0x110a
++#define MSR_VIA_RNG 0x110b
++#define MSR_VIA_BCR2 0x1147
+
-+#include <asm/segment.h>
-+#include <asm/mmu.h>
++/* Intel defined MSRs. */
++#define MSR_IA32_P5_MC_ADDR 0
++#define MSR_IA32_P5_MC_TYPE 1
++#define MSR_IA32_PLATFORM_ID 0x17
++#define MSR_IA32_EBL_CR_POWERON 0x2a
+
-+// 8 byte segment descriptor
-+struct desc_struct {
-+ u16 limit0;
-+ u16 base0;
-+ unsigned base1 : 8, type : 4, s : 1, dpl : 2, p : 1;
-+ unsigned limit : 4, avl : 1, l : 1, d : 1, g : 1, base2 : 8;
-+} __attribute__((packed));
++#define MSR_IA32_APICBASE 0x1b
++#define MSR_IA32_APICBASE_BSP (1<<8)
++#define MSR_IA32_APICBASE_ENABLE (1<<11)
++#define MSR_IA32_APICBASE_BASE (0xfffff<<12)
+
-+struct n_desc_struct {
-+ unsigned int a,b;
-+};
++/* P4/Xeon+ specific */
++#define MSR_IA32_MCG_EAX 0x180
++#define MSR_IA32_MCG_EBX 0x181
++#define MSR_IA32_MCG_ECX 0x182
++#define MSR_IA32_MCG_EDX 0x183
++#define MSR_IA32_MCG_ESI 0x184
++#define MSR_IA32_MCG_EDI 0x185
++#define MSR_IA32_MCG_EBP 0x186
++#define MSR_IA32_MCG_ESP 0x187
++#define MSR_IA32_MCG_EFLAGS 0x188
++#define MSR_IA32_MCG_EIP 0x189
++#define MSR_IA32_MCG_RESERVED 0x18A
+
-+enum {
-+ GATE_INTERRUPT = 0xE,
-+ GATE_TRAP = 0xF,
-+ GATE_CALL = 0xC,
-+};
++#define MSR_P6_EVNTSEL0 0x186
++#define MSR_P6_EVNTSEL1 0x187
+
-+// 16byte gate
-+struct gate_struct {
-+ u16 offset_low;
-+ u16 segment;
-+ unsigned ist : 3, zero0 : 5, type : 5, dpl : 2, p : 1;
-+ u16 offset_middle;
-+ u32 offset_high;
-+ u32 zero1;
-+} __attribute__((packed));
++#define MSR_IA32_PERF_STATUS 0x198
++#define MSR_IA32_PERF_CTL 0x199
+
-+#define PTR_LOW(x) ((unsigned long)(x) & 0xFFFF)
-+#define PTR_MIDDLE(x) (((unsigned long)(x) >> 16) & 0xFFFF)
-+#define PTR_HIGH(x) ((unsigned long)(x) >> 32)
++#define MSR_IA32_THERM_CONTROL 0x19a
++#define MSR_IA32_THERM_INTERRUPT 0x19b
++#define MSR_IA32_THERM_STATUS 0x19c
++#define MSR_IA32_MISC_ENABLE 0x1a0
+
-+enum {
-+ DESC_TSS = 0x9,
-+ DESC_LDT = 0x2,
-+};
++#define MSR_IA32_DEBUGCTLMSR 0x1d9
++#define MSR_IA32_LASTBRANCHFROMIP 0x1db
++#define MSR_IA32_LASTBRANCHTOIP 0x1dc
++#define MSR_IA32_LASTINTFROMIP 0x1dd
++#define MSR_IA32_LASTINTTOIP 0x1de
+
-+// LDT or TSS descriptor in the GDT. 16 bytes.
-+struct ldttss_desc {
-+ u16 limit0;
-+ u16 base0;
-+ unsigned base1 : 8, type : 5, dpl : 2, p : 1;
-+ unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
-+ u32 base3;
-+ u32 zero1;
-+} __attribute__((packed));
++#define MSR_IA32_MC0_CTL 0x400
++#define MSR_IA32_MC0_STATUS 0x401
++#define MSR_IA32_MC0_ADDR 0x402
++#define MSR_IA32_MC0_MISC 0x403
+
-+struct desc_ptr {
-+ unsigned short size;
-+ unsigned long address;
-+} __attribute__((packed)) ;
++/* Pentium IV performance counter MSRs */
++#define MSR_P4_BPU_PERFCTR0 0x300
++#define MSR_P4_BPU_PERFCTR1 0x301
++#define MSR_P4_BPU_PERFCTR2 0x302
++#define MSR_P4_BPU_PERFCTR3 0x303
++#define MSR_P4_MS_PERFCTR0 0x304
++#define MSR_P4_MS_PERFCTR1 0x305
++#define MSR_P4_MS_PERFCTR2 0x306
++#define MSR_P4_MS_PERFCTR3 0x307
++#define MSR_P4_FLAME_PERFCTR0 0x308
++#define MSR_P4_FLAME_PERFCTR1 0x309
++#define MSR_P4_FLAME_PERFCTR2 0x30a
++#define MSR_P4_FLAME_PERFCTR3 0x30b
++#define MSR_P4_IQ_PERFCTR0 0x30c
++#define MSR_P4_IQ_PERFCTR1 0x30d
++#define MSR_P4_IQ_PERFCTR2 0x30e
++#define MSR_P4_IQ_PERFCTR3 0x30f
++#define MSR_P4_IQ_PERFCTR4 0x310
++#define MSR_P4_IQ_PERFCTR5 0x311
++#define MSR_P4_BPU_CCCR0 0x360
++#define MSR_P4_BPU_CCCR1 0x361
++#define MSR_P4_BPU_CCCR2 0x362
++#define MSR_P4_BPU_CCCR3 0x363
++#define MSR_P4_MS_CCCR0 0x364
++#define MSR_P4_MS_CCCR1 0x365
++#define MSR_P4_MS_CCCR2 0x366
++#define MSR_P4_MS_CCCR3 0x367
++#define MSR_P4_FLAME_CCCR0 0x368
++#define MSR_P4_FLAME_CCCR1 0x369
++#define MSR_P4_FLAME_CCCR2 0x36a
++#define MSR_P4_FLAME_CCCR3 0x36b
++#define MSR_P4_IQ_CCCR0 0x36c
++#define MSR_P4_IQ_CCCR1 0x36d
++#define MSR_P4_IQ_CCCR2 0x36e
++#define MSR_P4_IQ_CCCR3 0x36f
++#define MSR_P4_IQ_CCCR4 0x370
++#define MSR_P4_IQ_CCCR5 0x371
++#define MSR_P4_ALF_ESCR0 0x3ca
++#define MSR_P4_ALF_ESCR1 0x3cb
++#define MSR_P4_BPU_ESCR0 0x3b2
++#define MSR_P4_BPU_ESCR1 0x3b3
++#define MSR_P4_BSU_ESCR0 0x3a0
++#define MSR_P4_BSU_ESCR1 0x3a1
++#define MSR_P4_CRU_ESCR0 0x3b8
++#define MSR_P4_CRU_ESCR1 0x3b9
++#define MSR_P4_CRU_ESCR2 0x3cc
++#define MSR_P4_CRU_ESCR3 0x3cd
++#define MSR_P4_CRU_ESCR4 0x3e0
++#define MSR_P4_CRU_ESCR5 0x3e1
++#define MSR_P4_DAC_ESCR0 0x3a8
++#define MSR_P4_DAC_ESCR1 0x3a9
++#define MSR_P4_FIRM_ESCR0 0x3a4
++#define MSR_P4_FIRM_ESCR1 0x3a5
++#define MSR_P4_FLAME_ESCR0 0x3a6
++#define MSR_P4_FLAME_ESCR1 0x3a7
++#define MSR_P4_FSB_ESCR0 0x3a2
++#define MSR_P4_FSB_ESCR1 0x3a3
++#define MSR_P4_IQ_ESCR0 0x3ba
++#define MSR_P4_IQ_ESCR1 0x3bb
++#define MSR_P4_IS_ESCR0 0x3b4
++#define MSR_P4_IS_ESCR1 0x3b5
++#define MSR_P4_ITLB_ESCR0 0x3b6
++#define MSR_P4_ITLB_ESCR1 0x3b7
++#define MSR_P4_IX_ESCR0 0x3c8
++#define MSR_P4_IX_ESCR1 0x3c9
++#define MSR_P4_MOB_ESCR0 0x3aa
++#define MSR_P4_MOB_ESCR1 0x3ab
++#define MSR_P4_MS_ESCR0 0x3c0
++#define MSR_P4_MS_ESCR1 0x3c1
++#define MSR_P4_PMH_ESCR0 0x3ac
++#define MSR_P4_PMH_ESCR1 0x3ad
++#define MSR_P4_RAT_ESCR0 0x3bc
++#define MSR_P4_RAT_ESCR1 0x3bd
++#define MSR_P4_SAAT_ESCR0 0x3ae
++#define MSR_P4_SAAT_ESCR1 0x3af
++#define MSR_P4_SSU_ESCR0 0x3be
++#define MSR_P4_SSU_ESCR1 0x3bf /* guess: not defined in manual */
++#define MSR_P4_TBPU_ESCR0 0x3c2
++#define MSR_P4_TBPU_ESCR1 0x3c3
++#define MSR_P4_TC_ESCR0 0x3c4
++#define MSR_P4_TC_ESCR1 0x3c5
++#define MSR_P4_U2L_ESCR0 0x3b0
++#define MSR_P4_U2L_ESCR1 0x3b1
+
-+extern struct desc_ptr idt_descr, cpu_gdt_descr[NR_CPUS];
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/nmi.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/nmi.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,93 @@
++/*
++ * linux/include/asm-i386/nmi.h
++ */
++#ifndef ASM_NMI_H
++#define ASM_NMI_H
+
-+extern struct desc_struct cpu_gdt_table[GDT_ENTRIES];
++#include <linux/pm.h>
++#include <asm/io.h>
+
-+#define load_TR_desc() asm volatile("ltr %w0"::"r" (GDT_ENTRY_TSS*8))
-+#define load_LDT_desc() asm volatile("lldt %w0"::"r" (GDT_ENTRY_LDT*8))
++#include <xen/interface/nmi.h>
+
-+static inline void clear_LDT(void)
-+{
-+ int cpu = get_cpu();
++struct pt_regs;
+
-+ /*
-+ * NB. We load the default_ldt for lcall7/27 handling on demand, as
-+ * it slows down context switching. Noone uses it anyway.
-+ */
-+ cpu = cpu; /* XXX avoid compiler warning */
-+ xen_set_ldt(0UL, 0);
-+ put_cpu();
-+}
++typedef int (*nmi_callback_t)(struct pt_regs * regs, int cpu);
+
-+/*
-+ * This is the ldt that every process will get unless we need
-+ * something other than this.
++/**
++ * set_nmi_callback
++ *
++ * Set a handler for an NMI. Only one handler may be
++ * set. Return 1 if the NMI was handled.
+ */
-+extern struct desc_struct default_ldt[];
-+#ifndef CONFIG_X86_NO_IDT
-+extern struct gate_struct idt_table[];
-+#endif
-+extern struct desc_ptr cpu_gdt_descr[];
++void set_nmi_callback(nmi_callback_t callback);
+
-+/* the cpu gdt accessor */
-+#define cpu_gdt(_cpu) ((struct desc_struct *)cpu_gdt_descr[_cpu].address)
++/**
++ * unset_nmi_callback
++ *
++ * Remove the handler previously set.
++ */
++void unset_nmi_callback(void);
+
-+static inline void _set_gate(void *adr, unsigned type, unsigned long func, unsigned dpl, unsigned ist)
-+{
-+ struct gate_struct s;
-+ s.offset_low = PTR_LOW(func);
-+ s.segment = __KERNEL_CS;
-+ s.ist = ist;
-+ s.p = 1;
-+ s.dpl = dpl;
-+ s.zero0 = 0;
-+ s.zero1 = 0;
-+ s.type = type;
-+ s.offset_middle = PTR_MIDDLE(func);
-+ s.offset_high = PTR_HIGH(func);
-+ /* does not need to be atomic because it is only done once at setup time */
-+ memcpy(adr, &s, 16);
-+}
++#ifdef CONFIG_PM
++
++/** Replace the PM callback routine for NMI. */
++struct pm_dev * set_nmi_pm_callback(pm_callback callback);
+
-+#ifndef CONFIG_X86_NO_IDT
-+static inline void set_intr_gate(int nr, void *func)
-+{
-+ BUG_ON((unsigned)nr > 0xFF);
-+ _set_gate(&idt_table[nr], GATE_INTERRUPT, (unsigned long) func, 0, 0);
-+}
++/** Unset the PM callback routine back to the default. */
++void unset_nmi_pm_callback(struct pm_dev * dev);
+
-+static inline void set_intr_gate_ist(int nr, void *func, unsigned ist)
-+{
-+ BUG_ON((unsigned)nr > 0xFF);
-+ _set_gate(&idt_table[nr], GATE_INTERRUPT, (unsigned long) func, 0, ist);
-+}
++#else
+
-+static inline void set_system_gate(int nr, void *func)
-+{
-+ BUG_ON((unsigned)nr > 0xFF);
-+ _set_gate(&idt_table[nr], GATE_INTERRUPT, (unsigned long) func, 3, 0);
++static inline struct pm_dev * set_nmi_pm_callback(pm_callback callback)
++{
++ return 0;
+}
-+
-+static inline void set_system_gate_ist(int nr, void *func, unsigned ist)
++
++static inline void unset_nmi_pm_callback(struct pm_dev * dev)
+{
-+ _set_gate(&idt_table[nr], GATE_INTERRUPT, (unsigned long) func, 3, ist);
+}
-+#endif
+
-+static inline void set_tssldt_descriptor(void *ptr, unsigned long tss, unsigned type,
-+ unsigned size)
-+{
-+ struct ldttss_desc d;
-+ memset(&d,0,sizeof(d));
-+ d.limit0 = size & 0xFFFF;
-+ d.base0 = PTR_LOW(tss);
-+ d.base1 = PTR_MIDDLE(tss) & 0xFF;
-+ d.type = type;
-+ d.p = 1;
-+ d.limit1 = (size >> 16) & 0xF;
-+ d.base2 = (PTR_MIDDLE(tss) >> 8) & 0xFF;
-+ d.base3 = PTR_HIGH(tss);
-+ memcpy(ptr, &d, 16);
-+}
++#endif /* CONFIG_PM */
++
++extern void default_do_nmi(struct pt_regs *);
++extern void die_nmi(char *str, struct pt_regs *regs);
+
-+#ifndef CONFIG_X86_NO_TSS
-+static inline void set_tss_desc(unsigned cpu, void *addr)
-+{
-+ /*
-+ * sizeof(unsigned long) coming from an extra "long" at the end
-+ * of the iobitmap. See tss_struct definition in processor.h
-+ *
-+ * -1? seg base+limit should be pointing to the address of the
-+ * last valid byte
-+ */
-+ set_tssldt_descriptor(&cpu_gdt(cpu)[GDT_ENTRY_TSS],
-+ (unsigned long)addr, DESC_TSS,
-+ IO_BITMAP_OFFSET + IO_BITMAP_BYTES + sizeof(unsigned long) - 1);
-+}
-+#endif
++static inline unsigned char get_nmi_reason(void)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ unsigned char reason = 0;
+
-+static inline void set_ldt_desc(unsigned cpu, void *addr, int size)
-+{
-+ set_tssldt_descriptor(&cpu_gdt(cpu)[GDT_ENTRY_LDT], (unsigned long)addr,
-+ DESC_LDT, size * 8 - 1);
-+}
++ /* construct a value which looks like it came from
++ * port 0x61.
++ */
++ if (test_bit(_XEN_NMIREASON_io_error, &s->arch.nmi_reason))
++ reason |= 0x40;
++ if (test_bit(_XEN_NMIREASON_parity_error, &s->arch.nmi_reason))
++ reason |= 0x80;
+
-+static inline void set_seg_base(unsigned cpu, int entry, void *base)
-+{
-+ struct desc_struct *d = &cpu_gdt(cpu)[entry];
-+ u32 addr = (u32)(u64)base;
-+ BUG_ON((u64)base >> 32);
-+ d->base0 = addr & 0xffff;
-+ d->base1 = (addr >> 16) & 0xff;
-+ d->base2 = (addr >> 24) & 0xff;
-+}
++ return reason;
++}
+
-+#define LDT_entry_a(info) \
-+ ((((info)->base_addr & 0x0000ffff) << 16) | ((info)->limit & 0x0ffff))
-+/* Don't allow setting of the lm bit. It is useless anyways because
-+ 64bit system calls require __USER_CS. */
-+#define LDT_entry_b(info) \
-+ (((info)->base_addr & 0xff000000) | \
-+ (((info)->base_addr & 0x00ff0000) >> 16) | \
-+ ((info)->limit & 0xf0000) | \
-+ (((info)->read_exec_only ^ 1) << 9) | \
-+ ((info)->contents << 10) | \
-+ (((info)->seg_not_present ^ 1) << 15) | \
-+ ((info)->seg_32bit << 22) | \
-+ ((info)->limit_in_pages << 23) | \
-+ ((info)->useable << 20) | \
-+ /* ((info)->lm << 21) | */ \
-+ 0x7000)
++extern int panic_on_timeout;
++extern int unknown_nmi_panic;
+
-+#define LDT_empty(info) (\
-+ (info)->base_addr == 0 && \
-+ (info)->limit == 0 && \
-+ (info)->contents == 0 && \
-+ (info)->read_exec_only == 1 && \
-+ (info)->seg_32bit == 0 && \
-+ (info)->limit_in_pages == 0 && \
-+ (info)->seg_not_present == 1 && \
-+ (info)->useable == 0 && \
-+ (info)->lm == 0)
++extern int check_nmi_watchdog(void);
++
++extern void setup_apic_nmi_watchdog (void);
++extern int reserve_lapic_nmi(void);
++extern void release_lapic_nmi(void);
++extern void disable_timer_nmi_watchdog(void);
++extern void enable_timer_nmi_watchdog(void);
++extern void nmi_watchdog_tick (struct pt_regs * regs, unsigned reason);
+
-+#if TLS_SIZE != 24
-+# error update this code.
-+#endif
++extern void nmi_watchdog_default(void);
++extern int setup_nmi_watchdog(char *);
+
-+static inline void load_TLS(struct thread_struct *t, unsigned int cpu)
-+{
-+#if 0
-+ u64 *gdt = (u64 *)(cpu_gdt(cpu) + GDT_ENTRY_TLS_MIN);
-+ gdt[0] = t->tls_array[0];
-+ gdt[1] = t->tls_array[1];
-+ gdt[2] = t->tls_array[2];
-+#endif
-+#define C(i) \
-+ HYPERVISOR_update_descriptor(virt_to_machine(&cpu_gdt(cpu)[GDT_ENTRY_TLS_MIN + i]), t->tls_array[i])
++extern unsigned int nmi_watchdog;
++#define NMI_DEFAULT -1
++#define NMI_NONE 0
++#define NMI_IO_APIC 1
++#define NMI_LOCAL_APIC 2
++#define NMI_INVALID 3
+
-+ C(0); C(1); C(2);
-+#undef C
-+}
++#endif /* ASM_NMI_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/page.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/page.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,212 @@
++#ifndef _X86_64_PAGE_H
++#define _X86_64_PAGE_H
++
++/* #include <linux/string.h> */
++#ifndef __ASSEMBLY__
++#include <linux/kernel.h>
++#include <linux/types.h>
++#include <asm/bug.h>
++#endif
++#include <xen/interface/xen.h>
+
+/*
-+ * load one particular LDT into the current CPU
++ * Need to repeat this here in order to not include pgtable.h (which in turn
++ * depends on definitions made here), but to be able to use the symbolic
++ * below. The preprocessor will warn if the two definitions aren't identical.
+ */
-+static inline void load_LDT_nolock (mm_context_t *pc, int cpu)
-+{
-+ void *segments = pc->ldt;
-+ int count = pc->size;
-+
-+ if (likely(!count))
-+ segments = NULL;
++#define _PAGE_PRESENT 0x001
++#define _PAGE_IO 0x200
+
-+ xen_set_ldt((unsigned long)segments, count);
-+}
++/* PAGE_SHIFT determines the page size */
++#define PAGE_SHIFT 12
++#ifdef __ASSEMBLY__
++#define PAGE_SIZE (0x1 << PAGE_SHIFT)
++#else
++#define PAGE_SIZE (1UL << PAGE_SHIFT)
++#endif
++#define PAGE_MASK (~(PAGE_SIZE-1))
+
-+static inline void load_LDT(mm_context_t *pc)
-+{
-+ int cpu = get_cpu();
-+ load_LDT_nolock(pc, cpu);
-+ put_cpu();
-+}
++/* See Documentation/x86_64/mm.txt for a description of the memory map. */
++#define __PHYSICAL_MASK_SHIFT 46
++#define __PHYSICAL_MASK ((1UL << __PHYSICAL_MASK_SHIFT) - 1)
++#define __VIRTUAL_MASK_SHIFT 48
++#define __VIRTUAL_MASK ((1UL << __VIRTUAL_MASK_SHIFT) - 1)
+
-+extern struct desc_ptr idt_descr;
++#define PHYSICAL_PAGE_MASK (~(PAGE_SIZE-1) & __PHYSICAL_MASK)
+
-+#endif /* !__ASSEMBLY__ */
++#define THREAD_ORDER 1
++#define THREAD_SIZE (PAGE_SIZE << THREAD_ORDER)
++#define CURRENT_MASK (~(THREAD_SIZE-1))
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/dma-mapping.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/dma-mapping.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/dma-mapping.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/dma-mapping.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,207 @@
-+#ifndef _X8664_DMA_MAPPING_H
-+#define _X8664_DMA_MAPPING_H 1
++#define EXCEPTION_STACK_ORDER 0
++#define EXCEPTION_STKSZ (PAGE_SIZE << EXCEPTION_STACK_ORDER)
+
-+/*
-+ * IOMMU interface. See Documentation/DMA-mapping.txt and DMA-API.txt for
-+ * documentation.
-+ */
++#define DEBUG_STACK_ORDER (EXCEPTION_STACK_ORDER + 1)
++#define DEBUG_STKSZ (PAGE_SIZE << DEBUG_STACK_ORDER)
+
++#define IRQSTACK_ORDER 2
++#define IRQSTACKSIZE (PAGE_SIZE << IRQSTACK_ORDER)
+
-+#include <asm/scatterlist.h>
-+#include <asm/io.h>
-+#include <asm/swiotlb.h>
++#define STACKFAULT_STACK 1
++#define DOUBLEFAULT_STACK 2
++#define NMI_STACK 3
++#define DEBUG_STACK 4
++#define MCE_STACK 5
++#define N_EXCEPTION_STACKS 5 /* hw limit: 7 */
+
-+struct dma_mapping_ops {
-+ int (*mapping_error)(dma_addr_t dma_addr);
-+ void* (*alloc_coherent)(struct device *dev, size_t size,
-+ dma_addr_t *dma_handle, gfp_t gfp);
-+ void (*free_coherent)(struct device *dev, size_t size,
-+ void *vaddr, dma_addr_t dma_handle);
-+ dma_addr_t (*map_single)(struct device *hwdev, void *ptr,
-+ size_t size, int direction);
-+ /* like map_single, but doesn't check the device mask */
-+ dma_addr_t (*map_simple)(struct device *hwdev, char *ptr,
-+ size_t size, int direction);
-+ void (*unmap_single)(struct device *dev, dma_addr_t addr,
-+ size_t size, int direction);
-+ void (*sync_single_for_cpu)(struct device *hwdev,
-+ dma_addr_t dma_handle, size_t size,
-+ int direction);
-+ void (*sync_single_for_device)(struct device *hwdev,
-+ dma_addr_t dma_handle, size_t size,
-+ int direction);
-+ void (*sync_single_range_for_cpu)(struct device *hwdev,
-+ dma_addr_t dma_handle, unsigned long offset,
-+ size_t size, int direction);
-+ void (*sync_single_range_for_device)(struct device *hwdev,
-+ dma_addr_t dma_handle, unsigned long offset,
-+ size_t size, int direction);
-+ void (*sync_sg_for_cpu)(struct device *hwdev,
-+ struct scatterlist *sg, int nelems,
-+ int direction);
-+ void (*sync_sg_for_device)(struct device *hwdev,
-+ struct scatterlist *sg, int nelems,
-+ int direction);
-+ int (*map_sg)(struct device *hwdev, struct scatterlist *sg,
-+ int nents, int direction);
-+ void (*unmap_sg)(struct device *hwdev,
-+ struct scatterlist *sg, int nents,
-+ int direction);
-+ int (*dma_supported)(struct device *hwdev, u64 mask);
-+ int is_phys;
-+};
++#define LARGE_PAGE_MASK (~(LARGE_PAGE_SIZE-1))
++#define LARGE_PAGE_SIZE (1UL << PMD_SHIFT)
+
-+extern dma_addr_t bad_dma_address;
-+extern struct dma_mapping_ops* dma_ops;
-+extern int iommu_merge;
++#define HPAGE_SHIFT PMD_SHIFT
++#define HPAGE_SIZE ((1UL) << HPAGE_SHIFT)
++#define HPAGE_MASK (~(HPAGE_SIZE - 1))
++#define HUGETLB_PAGE_ORDER (HPAGE_SHIFT - PAGE_SHIFT)
+
-+static inline int valid_dma_direction(int dma_direction)
-+{
-+ return ((dma_direction == DMA_BIDIRECTIONAL) ||
-+ (dma_direction == DMA_TO_DEVICE) ||
-+ (dma_direction == DMA_FROM_DEVICE));
-+}
++#ifdef __KERNEL__
++#ifndef __ASSEMBLY__
+
-+#if 0
-+static inline int dma_mapping_error(dma_addr_t dma_addr)
-+{
-+ if (dma_ops->mapping_error)
-+ return dma_ops->mapping_error(dma_addr);
++extern unsigned long end_pfn;
+
-+ return (dma_addr == bad_dma_address);
-+}
++#include <asm/maddr.h>
+
-+extern void *dma_alloc_coherent(struct device *dev, size_t size,
-+ dma_addr_t *dma_handle, gfp_t gfp);
-+extern void dma_free_coherent(struct device *dev, size_t size, void *vaddr,
-+ dma_addr_t dma_handle);
++void clear_page(void *);
++void copy_page(void *, void *);
+
-+static inline dma_addr_t
-+dma_map_single(struct device *hwdev, void *ptr, size_t size,
-+ int direction)
-+{
-+ BUG_ON(!valid_dma_direction(direction));
-+ return dma_ops->map_single(hwdev, ptr, size, direction);
-+}
++#define clear_user_page(page, vaddr, pg) clear_page(page)
++#define copy_user_page(to, from, vaddr, pg) copy_page(to, from)
+
-+static inline void
-+dma_unmap_single(struct device *dev, dma_addr_t addr,size_t size,
-+ int direction)
-+{
-+ BUG_ON(!valid_dma_direction(direction));
-+ dma_ops->unmap_single(dev, addr, size, direction);
-+}
++#define alloc_zeroed_user_highpage(vma, vaddr) alloc_page_vma(GFP_HIGHUSER | __GFP_ZERO, vma, vaddr)
++#define __HAVE_ARCH_ALLOC_ZEROED_USER_HIGHPAGE
+
-+#define dma_map_page(dev,page,offset,size,dir) \
-+ dma_map_single((dev), page_address(page)+(offset), (size), (dir))
++/*
++ * These are used to make use of C type-checking..
++ */
++typedef struct { unsigned long pte; } pte_t;
++typedef struct { unsigned long pmd; } pmd_t;
++typedef struct { unsigned long pud; } pud_t;
++typedef struct { unsigned long pgd; } pgd_t;
++#define PTE_MASK PHYSICAL_PAGE_MASK
+
-+#define dma_unmap_page dma_unmap_single
++typedef struct { unsigned long pgprot; } pgprot_t;
+
-+static inline void
-+dma_sync_single_for_cpu(struct device *hwdev, dma_addr_t dma_handle,
-+ size_t size, int direction)
-+{
-+ BUG_ON(!valid_dma_direction(direction));
-+ if (dma_ops->sync_single_for_cpu)
-+ dma_ops->sync_single_for_cpu(hwdev, dma_handle, size,
-+ direction);
-+ flush_write_buffers();
-+}
++#define __pte_val(x) ((x).pte)
++#define pte_val(x) ((__pte_val(x) & (_PAGE_PRESENT|_PAGE_IO)) \
++ == _PAGE_PRESENT ? \
++ pte_machine_to_phys(__pte_val(x)) : \
++ __pte_val(x))
+
-+static inline void
-+dma_sync_single_for_device(struct device *hwdev, dma_addr_t dma_handle,
-+ size_t size, int direction)
++#define __pmd_val(x) ((x).pmd)
++static inline unsigned long pmd_val(pmd_t x)
+{
-+ BUG_ON(!valid_dma_direction(direction));
-+ if (dma_ops->sync_single_for_device)
-+ dma_ops->sync_single_for_device(hwdev, dma_handle, size,
-+ direction);
-+ flush_write_buffers();
++ unsigned long ret = __pmd_val(x);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (ret) ret = pte_machine_to_phys(ret) | _PAGE_PRESENT;
++#else
++ if (ret & _PAGE_PRESENT) ret = pte_machine_to_phys(ret);
++#endif
++ return ret;
+}
+
-+static inline void
-+dma_sync_single_range_for_cpu(struct device *hwdev, dma_addr_t dma_handle,
-+ unsigned long offset, size_t size, int direction)
++#define __pud_val(x) ((x).pud)
++static inline unsigned long pud_val(pud_t x)
+{
-+ BUG_ON(!valid_dma_direction(direction));
-+ if (dma_ops->sync_single_range_for_cpu) {
-+ dma_ops->sync_single_range_for_cpu(hwdev, dma_handle, offset, size, direction);
-+ }
-+
-+ flush_write_buffers();
++ unsigned long ret = __pud_val(x);
++ if (ret & _PAGE_PRESENT) ret = pte_machine_to_phys(ret);
++ return ret;
+}
+
-+static inline void
-+dma_sync_single_range_for_device(struct device *hwdev, dma_addr_t dma_handle,
-+ unsigned long offset, size_t size, int direction)
++#define __pgd_val(x) ((x).pgd)
++static inline unsigned long pgd_val(pgd_t x)
+{
-+ BUG_ON(!valid_dma_direction(direction));
-+ if (dma_ops->sync_single_range_for_device)
-+ dma_ops->sync_single_range_for_device(hwdev, dma_handle,
-+ offset, size, direction);
-+
-+ flush_write_buffers();
++ unsigned long ret = __pgd_val(x);
++ if (ret & _PAGE_PRESENT) ret = pte_machine_to_phys(ret);
++ return ret;
+}
+
-+static inline void
-+dma_sync_sg_for_cpu(struct device *hwdev, struct scatterlist *sg,
-+ int nelems, int direction)
-+{
-+ BUG_ON(!valid_dma_direction(direction));
-+ if (dma_ops->sync_sg_for_cpu)
-+ dma_ops->sync_sg_for_cpu(hwdev, sg, nelems, direction);
-+ flush_write_buffers();
-+}
++#define pgprot_val(x) ((x).pgprot)
+
-+static inline void
-+dma_sync_sg_for_device(struct device *hwdev, struct scatterlist *sg,
-+ int nelems, int direction)
++static inline pte_t __pte(unsigned long x)
+{
-+ BUG_ON(!valid_dma_direction(direction));
-+ if (dma_ops->sync_sg_for_device) {
-+ dma_ops->sync_sg_for_device(hwdev, sg, nelems, direction);
-+ }
-+
-+ flush_write_buffers();
++ if ((x & (_PAGE_PRESENT|_PAGE_IO)) == _PAGE_PRESENT)
++ x = pte_phys_to_machine(x);
++ return ((pte_t) { (x) });
+}
+
-+static inline int
-+dma_map_sg(struct device *hwdev, struct scatterlist *sg, int nents, int direction)
++static inline pmd_t __pmd(unsigned long x)
+{
-+ BUG_ON(!valid_dma_direction(direction));
-+ return dma_ops->map_sg(hwdev, sg, nents, direction);
++ if (x & _PAGE_PRESENT) x = pte_phys_to_machine(x);
++ return ((pmd_t) { (x) });
+}
+
-+static inline void
-+dma_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nents,
-+ int direction)
++static inline pud_t __pud(unsigned long x)
+{
-+ BUG_ON(!valid_dma_direction(direction));
-+ dma_ops->unmap_sg(hwdev, sg, nents, direction);
++ if (x & _PAGE_PRESENT) x = pte_phys_to_machine(x);
++ return ((pud_t) { (x) });
+}
+
-+extern int dma_supported(struct device *hwdev, u64 mask);
-+
-+/* same for gart, swiotlb, and nommu */
-+static inline int dma_get_cache_alignment(void)
++static inline pgd_t __pgd(unsigned long x)
+{
-+ return boot_cpu_data.x86_clflush_size;
++ if (x & _PAGE_PRESENT) x = pte_phys_to_machine(x);
++ return ((pgd_t) { (x) });
+}
+
-+#define dma_is_consistent(h) 1
++#define __pgprot(x) ((pgprot_t) { (x) } )
+
-+extern int dma_set_mask(struct device *dev, u64 mask);
++#define __PHYSICAL_START ((unsigned long)CONFIG_PHYSICAL_START)
++#define __START_KERNEL (__START_KERNEL_map + __PHYSICAL_START)
++#define __START_KERNEL_map 0xffffffff80000000UL
++#define __PAGE_OFFSET 0xffff880000000000UL
+
-+static inline void
-+dma_cache_sync(void *vaddr, size_t size, enum dma_data_direction dir)
-+{
-+ flush_write_buffers();
-+}
++#else
++#define __PHYSICAL_START CONFIG_PHYSICAL_START
++#define __START_KERNEL (__START_KERNEL_map + __PHYSICAL_START)
++#define __START_KERNEL_map 0xffffffff80000000
++#define __PAGE_OFFSET 0xffff880000000000
++#endif /* !__ASSEMBLY__ */
+
-+extern struct device fallback_dev;
-+extern int panic_on_overflow;
++#if CONFIG_XEN_COMPAT <= 0x030002
++#undef LOAD_OFFSET
++#define LOAD_OFFSET 0
+#endif
+
-+#endif /* _X8664_DMA_MAPPING_H */
-+
-+#include <asm-i386/mach-xen/asm/dma-mapping.h>
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/e820.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/e820.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/e820.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/e820.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,66 @@
-+/*
-+ * structures and definitions for the int 15, ax=e820 memory map
-+ * scheme.
-+ *
-+ * In a nutshell, setup.S populates a scratch table in the
-+ * empty_zero_block that contains a list of usable address/size
-+ * duples. setup.c, this information is transferred into the e820map,
-+ * and in init.c/numa.c, that new information is used to mark pages
-+ * reserved or not.
-+ */
-+#ifndef __E820_HEADER
-+#define __E820_HEADER
-+
-+#include <linux/mmzone.h>
-+
-+#define E820MAP 0x2d0 /* our map */
-+#define E820MAX 128 /* number of entries in E820MAP */
-+#define E820NR 0x1e8 /* # entries in E820MAP */
++/* to align the pointer to the (next) page boundary */
++#define PAGE_ALIGN(addr) (((addr)+PAGE_SIZE-1)&PAGE_MASK)
+
-+#define E820_RAM 1
-+#define E820_RESERVED 2
-+#define E820_ACPI 3 /* usable as RAM once ACPI tables have been read */
-+#define E820_NVS 4
++#define KERNEL_TEXT_SIZE (40UL*1024*1024)
++#define KERNEL_TEXT_START 0xffffffff80000000UL
+
-+#define HIGH_MEMORY (1024*1024)
++#define PAGE_OFFSET ((unsigned long)__PAGE_OFFSET)
+
-+#define LOWMEMSIZE() (0x9f000)
++/* Note: __pa(&symbol_visible_to_c) should be always replaced with __pa_symbol.
++ Otherwise you risk miscompilation. */
++#define __pa(x) (((unsigned long)(x)>=__START_KERNEL_map)?(unsigned long)(x) - (unsigned long)__START_KERNEL_map:(unsigned long)(x) - PAGE_OFFSET)
++/* __pa_symbol should be used for C visible symbols.
++ This seems to be the official gcc blessed way to do such arithmetic. */
++#define __pa_symbol(x) \
++ ({unsigned long v; \
++ asm("" : "=r" (v) : "0" (x)); \
++ __pa(v); })
+
-+#ifndef __ASSEMBLY__
-+struct e820entry {
-+ u64 addr; /* start of memory segment */
-+ u64 size; /* size of memory segment */
-+ u32 type; /* type of memory segment */
-+} __attribute__((packed));
++#define __va(x) ((void *)((unsigned long)(x)+PAGE_OFFSET))
++#define __boot_va(x) __va(x)
++#define __boot_pa(x) __pa(x)
++#ifdef CONFIG_FLATMEM
++#define pfn_valid(pfn) ((pfn) < end_pfn)
++#endif
+
-+struct e820map {
-+ int nr_map;
-+ struct e820entry map[E820MAX];
-+};
++#define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT)
++#define virt_addr_valid(kaddr) pfn_valid(__pa(kaddr) >> PAGE_SHIFT)
++#define pfn_to_kaddr(pfn) __va((pfn) << PAGE_SHIFT)
+
-+extern unsigned long find_e820_area(unsigned long start, unsigned long end,
-+ unsigned size);
-+extern void add_memory_region(unsigned long start, unsigned long size,
-+ int type);
-+extern void setup_memory_region(void);
-+extern void contig_e820_setup(void);
-+extern unsigned long e820_end_of_ram(void);
-+extern void e820_reserve_resources(struct e820entry *e820, int nr_map);
-+extern void e820_print_map(char *who);
-+extern int e820_any_mapped(unsigned long start, unsigned long end, unsigned type);
-+extern int e820_all_mapped(unsigned long start, unsigned long end, unsigned type);
++#define VM_DATA_DEFAULT_FLAGS \
++ (((current->personality & READ_IMPLIES_EXEC) ? VM_EXEC : 0 ) | \
++ VM_READ | VM_WRITE | VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC)
+
-+extern void e820_bootmem_free(pg_data_t *pgdat, unsigned long start,unsigned long end);
-+extern void e820_setup_gap(struct e820entry *e820, int nr_map);
-+extern unsigned long e820_hole_size(unsigned long start_pfn,
-+ unsigned long end_pfn);
++#define __HAVE_ARCH_GATE_AREA 1
+
-+extern void __init parse_memopt(char *p, char **end);
-+extern void __init parse_memmapopt(char *p, char **end);
++#include <asm-generic/memory_model.h>
++#include <asm-generic/page.h>
+
-+extern struct e820map e820;
++#endif /* __KERNEL__ */
+
-+extern unsigned ebda_addr, ebda_size;
-+#endif/*!__ASSEMBLY__*/
++#endif /* _X86_64_PAGE_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/pci.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/pci.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,168 @@
++#ifndef __x8664_PCI_H
++#define __x8664_PCI_H
+
-+#endif/*__E820_HEADER*/
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/fixmap.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/fixmap.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/fixmap.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/fixmap.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,112 @@
-+/*
-+ * fixmap.h: compile-time virtual memory allocation
-+ *
-+ * This file is subject to the terms and conditions of the GNU General Public
-+ * License. See the file "COPYING" in the main directory of this archive
-+ * for more details.
-+ *
-+ * Copyright (C) 1998 Ingo Molnar
-+ */
++#include <asm/io.h>
+
-+#ifndef _ASM_FIXMAP_H
-+#define _ASM_FIXMAP_H
++#ifdef __KERNEL__
+
-+#include <linux/kernel.h>
-+#include <asm/apicdef.h>
-+#include <asm/page.h>
-+#include <asm/vsyscall.h>
-+#include <asm/vsyscall32.h>
-+#include <asm/acpi.h>
++#include <linux/mm.h> /* for struct page */
+
-+/*
-+ * Here we define all the compile-time 'special' virtual
-+ * addresses. The point is to have a constant address at
-+ * compile time, but to set the physical address only
-+ * in the boot process.
-+ *
-+ * these 'compile-time allocated' memory buffers are
-+ * fixed-size 4k pages. (or larger if used with an increment
-+ * highger than 1) use fixmap_set(idx,phys) to associate
-+ * physical memory with fixmap indices.
-+ *
-+ * TLB entries of such buffers will not be flushed across
-+ * task switches.
-+ */
++/* Can be used to override the logic in pci_scan_bus for skipping
++ already-configured bus numbers - to be used for buggy BIOSes
++ or architectures with incomplete PCI setup by the loader */
+
-+enum fixed_addresses {
-+ VSYSCALL_LAST_PAGE,
-+ VSYSCALL_FIRST_PAGE = VSYSCALL_LAST_PAGE + ((VSYSCALL_END-VSYSCALL_START) >> PAGE_SHIFT) - 1,
-+ VSYSCALL_HPET,
-+ FIX_HPET_BASE,
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ FIX_APIC_BASE, /* local (CPU) APIC) -- required for SMP or not */
-+#endif
-+#ifdef CONFIG_X86_IO_APIC
-+ FIX_IO_APIC_BASE_0,
-+ FIX_IO_APIC_BASE_END = FIX_IO_APIC_BASE_0 + MAX_IO_APICS-1,
-+#endif
-+#ifdef CONFIG_ACPI
-+ FIX_ACPI_BEGIN,
-+ FIX_ACPI_END = FIX_ACPI_BEGIN + FIX_ACPI_PAGES - 1,
++#ifdef CONFIG_PCI
++extern unsigned int pcibios_assign_all_busses(void);
++#else
++#define pcibios_assign_all_busses() 0
+#endif
-+ FIX_SHARED_INFO,
-+#define NR_FIX_ISAMAPS 256
-+ FIX_ISAMAP_END,
-+ FIX_ISAMAP_BEGIN = FIX_ISAMAP_END + NR_FIX_ISAMAPS - 1,
-+ __end_of_permanent_fixed_addresses,
-+ /* temporary boot-time mappings, used before ioremap() is functional */
-+#define NR_FIX_BTMAPS 16
-+ FIX_BTMAP_END = __end_of_permanent_fixed_addresses,
-+ FIX_BTMAP_BEGIN = FIX_BTMAP_END + NR_FIX_BTMAPS - 1,
-+ __end_of_fixed_addresses
-+};
-+
-+extern void __set_fixmap (enum fixed_addresses idx,
-+ unsigned long phys, pgprot_t flags);
+
-+#define set_fixmap(idx, phys) \
-+ __set_fixmap(idx, phys, PAGE_KERNEL)
-+/*
-+ * Some hardware wants to get fixmapped without caching.
-+ */
-+#define set_fixmap_nocache(idx, phys) \
-+ __set_fixmap(idx, phys, PAGE_KERNEL_NOCACHE)
-+
-+#define clear_fixmap(idx) \
-+ __set_fixmap(idx, 0, __pgprot(0))
-+
-+#define FIXADDR_TOP (VSYSCALL_END-PAGE_SIZE)
-+#define FIXADDR_SIZE (__end_of_fixed_addresses << PAGE_SHIFT)
-+#define FIXADDR_START (FIXADDR_TOP - FIXADDR_SIZE)
++#include <asm/hypervisor.h>
++#define pcibios_scan_all_fns(a, b) (!is_initial_xendomain())
+
-+/* Only covers 32bit vsyscalls currently. Need another set for 64bit. */
-+#define FIXADDR_USER_START ((unsigned long)VSYSCALL32_VSYSCALL)
-+#define FIXADDR_USER_END (FIXADDR_USER_START + PAGE_SIZE)
++extern unsigned long pci_mem_start;
++#define PCIBIOS_MIN_IO 0x1000
++#define PCIBIOS_MIN_MEM (pci_mem_start)
+
-+#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT))
++#define PCIBIOS_MIN_CARDBUS_IO 0x4000
+
-+extern void __this_fixmap_does_not_exist(void);
++void pcibios_config_init(void);
++struct pci_bus * pcibios_scan_root(int bus);
++extern int (*pci_config_read)(int seg, int bus, int dev, int fn, int reg, int len, u32 *value);
++extern int (*pci_config_write)(int seg, int bus, int dev, int fn, int reg, int len, u32 value);
+
-+/*
-+ * 'index to address' translation. If anyone tries to use the idx
-+ * directly without translation, we catch the bug with a NULL-deference
-+ * kernel oops. Illegal ranges of incoming indices are caught too.
-+ */
-+static __always_inline unsigned long fix_to_virt(const unsigned int idx)
-+{
-+ /*
-+ * this branch gets completely eliminated after inlining,
-+ * except when someone tries to use fixaddr indices in an
-+ * illegal way. (such as mixing up address types or using
-+ * out-of-range indices).
-+ *
-+ * If it doesn't get removed, the linker will complain
-+ * loudly with a reasonably clear error message..
-+ */
-+ if (idx >= __end_of_fixed_addresses)
-+ __this_fixmap_does_not_exist();
++void pcibios_set_master(struct pci_dev *dev);
++void pcibios_penalize_isa_irq(int irq, int active);
++struct irq_routing_table *pcibios_get_irq_routing_table(void);
++int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq);
+
-+ return __fix_to_virt(idx);
-+}
++#include <linux/types.h>
++#include <linux/slab.h>
++#include <asm/scatterlist.h>
++#include <linux/string.h>
++#include <asm/page.h>
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/floppy.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/floppy.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/floppy.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/floppy.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,206 @@
-+/*
-+ * Architecture specific parts of the Floppy driver
-+ *
-+ * This file is subject to the terms and conditions of the GNU General Public
-+ * License. See the file "COPYING" in the main directory of this archive
-+ * for more details.
-+ *
-+ * Copyright (C) 1995
++extern void pci_iommu_alloc(void);
++extern int iommu_setup(char *opt);
++
++/* The PCI address space does equal the physical memory
++ * address space. The networking and block device layers use
++ * this boolean for bounce buffer decisions
+ *
-+ * Modifications for Xen are Copyright (c) 2004, Keir Fraser.
++ * On AMD64 it mostly equals, but we set it to zero if a hardware
++ * IOMMU (gart) of sotware IOMMU (swiotlb) is available.
+ */
-+#ifndef __ASM_XEN_X86_64_FLOPPY_H
-+#define __ASM_XEN_X86_64_FLOPPY_H
++#define PCI_DMA_BUS_IS_PHYS (dma_ops->is_phys)
+
-+#include <linux/vmalloc.h>
++#if defined(CONFIG_IOMMU) || defined(CONFIG_CALGARY_IOMMU)
+
+/*
-+ * The DMA channel used by the floppy controller cannot access data at
-+ * addresses >= 16MB
-+ *
-+ * Went back to the 1MB limit, as some people had problems with the floppy
-+ * driver otherwise. It doesn't matter much for performance anyway, as most
-+ * floppy accesses go through the track buffer.
++ * x86-64 always supports DAC, but sometimes it is useful to force
++ * devices through the IOMMU to get automatic sg list merging.
++ * Optional right now.
+ */
-+#define _CROSS_64KB(a,s,vdma) \
-+(!(vdma) && ((unsigned long)(a)/K_64 != ((unsigned long)(a) + (s) - 1) / K_64))
++extern int iommu_sac_force;
++#define pci_dac_dma_supported(pci_dev, mask) (!iommu_sac_force)
+
-+/* XEN: Hit DMA paths on the head. This trick from asm-m68k/floppy.h. */
-+#include <asm/dma.h>
-+#undef MAX_DMA_ADDRESS
-+#define MAX_DMA_ADDRESS 0
-+#define CROSS_64KB(a,s) (0)
-+
-+#define fd_inb(port) inb_p(port)
-+#define fd_outb(value,port) outb_p(value,port)
-+
-+#define fd_request_dma() (0)
-+#define fd_free_dma() ((void)0)
-+#define fd_enable_irq() enable_irq(FLOPPY_IRQ)
-+#define fd_disable_irq() disable_irq(FLOPPY_IRQ)
-+#define fd_free_irq() free_irq(FLOPPY_IRQ, NULL)
-+#define fd_get_dma_residue() vdma_get_dma_residue(FLOPPY_DMA)
-+/*
-+ * Do not use vmalloc/vfree: floppy_release_irq_and_dma() gets called from
-+ * softirq context via motor_off_callback. A generic bug we happen to trigger.
-+ */
-+#define fd_dma_mem_alloc(size) __get_free_pages(GFP_KERNEL|__GFP_NORETRY, get_order(size))
-+#define fd_dma_mem_free(addr, size) free_pages(addr, get_order(size))
-+#define fd_dma_setup(addr, size, mode, io) vdma_dma_setup(addr, size, mode, io)
-+
-+static int virtual_dma_count;
-+static int virtual_dma_residue;
-+static char *virtual_dma_addr;
-+static int virtual_dma_mode;
-+static int doing_pdma;
-+
-+static irqreturn_t floppy_hardint(int irq, void *dev_id, struct pt_regs * regs)
-+{
-+ register unsigned char st;
-+
-+#undef TRACE_FLPY_INT
-+
-+#ifdef TRACE_FLPY_INT
-+ static int calls=0;
-+ static int bytes=0;
-+ static int dma_wait=0;
-+#endif
-+ if (!doing_pdma)
-+ return floppy_interrupt(irq, dev_id, regs);
-+
-+#ifdef TRACE_FLPY_INT
-+ if(!calls)
-+ bytes = virtual_dma_count;
-+#endif
++#define DECLARE_PCI_UNMAP_ADDR(ADDR_NAME) \
++ dma_addr_t ADDR_NAME;
++#define DECLARE_PCI_UNMAP_LEN(LEN_NAME) \
++ __u32 LEN_NAME;
++#define pci_unmap_addr(PTR, ADDR_NAME) \
++ ((PTR)->ADDR_NAME)
++#define pci_unmap_addr_set(PTR, ADDR_NAME, VAL) \
++ (((PTR)->ADDR_NAME) = (VAL))
++#define pci_unmap_len(PTR, LEN_NAME) \
++ ((PTR)->LEN_NAME)
++#define pci_unmap_len_set(PTR, LEN_NAME, VAL) \
++ (((PTR)->LEN_NAME) = (VAL))
+
-+ {
-+ register int lcount;
-+ register char *lptr;
++#elif defined(CONFIG_SWIOTLB)
+
-+ st = 1;
-+ for(lcount=virtual_dma_count, lptr=virtual_dma_addr;
-+ lcount; lcount--, lptr++) {
-+ st=inb(virtual_dma_port+4) & 0xa0 ;
-+ if(st != 0xa0)
-+ break;
-+ if(virtual_dma_mode)
-+ outb_p(*lptr, virtual_dma_port+5);
-+ else
-+ *lptr = inb_p(virtual_dma_port+5);
-+ }
-+ virtual_dma_count = lcount;
-+ virtual_dma_addr = lptr;
-+ st = inb(virtual_dma_port+4);
-+ }
++#define pci_dac_dma_supported(pci_dev, mask) 1
++
++#define DECLARE_PCI_UNMAP_ADDR(ADDR_NAME) \
++ dma_addr_t ADDR_NAME;
++#define DECLARE_PCI_UNMAP_LEN(LEN_NAME) \
++ __u32 LEN_NAME;
++#define pci_unmap_addr(PTR, ADDR_NAME) \
++ ((PTR)->ADDR_NAME)
++#define pci_unmap_addr_set(PTR, ADDR_NAME, VAL) \
++ (((PTR)->ADDR_NAME) = (VAL))
++#define pci_unmap_len(PTR, LEN_NAME) \
++ ((PTR)->LEN_NAME)
++#define pci_unmap_len_set(PTR, LEN_NAME, VAL) \
++ (((PTR)->LEN_NAME) = (VAL))
++
++#else
++/* No IOMMU */
++
++#define pci_dac_dma_supported(pci_dev, mask) 1
++
++#define DECLARE_PCI_UNMAP_ADDR(ADDR_NAME)
++#define DECLARE_PCI_UNMAP_LEN(LEN_NAME)
++#define pci_unmap_addr(PTR, ADDR_NAME) (0)
++#define pci_unmap_addr_set(PTR, ADDR_NAME, VAL) do { } while (0)
++#define pci_unmap_len(PTR, LEN_NAME) (0)
++#define pci_unmap_len_set(PTR, LEN_NAME, VAL) do { } while (0)
+
-+#ifdef TRACE_FLPY_INT
-+ calls++;
-+#endif
-+ if(st == 0x20)
-+ return IRQ_HANDLED;
-+ if(!(st & 0x20)) {
-+ virtual_dma_residue += virtual_dma_count;
-+ virtual_dma_count=0;
-+#ifdef TRACE_FLPY_INT
-+ printk("count=%x, residue=%x calls=%d bytes=%d dma_wait=%d\n",
-+ virtual_dma_count, virtual_dma_residue, calls, bytes,
-+ dma_wait);
-+ calls = 0;
-+ dma_wait=0;
-+#endif
-+ doing_pdma = 0;
-+ floppy_interrupt(irq, dev_id, regs);
-+ return IRQ_HANDLED;
-+ }
-+#ifdef TRACE_FLPY_INT
-+ if(!virtual_dma_count)
-+ dma_wait++;
+#endif
-+ return IRQ_HANDLED;
-+}
+
-+static void fd_disable_dma(void)
++#include <asm-generic/pci-dma-compat.h>
++
++static inline dma64_addr_t
++pci_dac_page_to_dma(struct pci_dev *pdev, struct page *page, unsigned long offset, int direction)
+{
-+ doing_pdma = 0;
-+ virtual_dma_residue += virtual_dma_count;
-+ virtual_dma_count=0;
++ return ((dma64_addr_t) page_to_phys(page) +
++ (dma64_addr_t) offset);
+}
+
-+static int vdma_get_dma_residue(unsigned int dummy)
++static inline struct page *
++pci_dac_dma_to_page(struct pci_dev *pdev, dma64_addr_t dma_addr)
+{
-+ return virtual_dma_count + virtual_dma_residue;
++ return virt_to_page(__va(dma_addr));
+}
+
-+
-+static int fd_request_irq(void)
++static inline unsigned long
++pci_dac_dma_to_offset(struct pci_dev *pdev, dma64_addr_t dma_addr)
+{
-+ return request_irq(FLOPPY_IRQ, floppy_hardint,
-+ IRQF_DISABLED, "floppy", NULL);
++ return (dma_addr & ~PAGE_MASK);
+}
+
-+#if 0
-+static unsigned long vdma_mem_alloc(unsigned long size)
++static inline void
++pci_dac_dma_sync_single_for_cpu(struct pci_dev *pdev, dma64_addr_t dma_addr, size_t len, int direction)
+{
-+ return (unsigned long) vmalloc(size);
-+
+}
+
-+static void vdma_mem_free(unsigned long addr, unsigned long size)
++static inline void
++pci_dac_dma_sync_single_for_device(struct pci_dev *pdev, dma64_addr_t dma_addr, size_t len, int direction)
+{
-+ vfree((void *)addr);
++ flush_write_buffers();
+}
-+#endif
+
-+static int vdma_dma_setup(char *addr, unsigned long size, int mode, int io)
++#ifdef CONFIG_PCI
++static inline void pci_dma_burst_advice(struct pci_dev *pdev,
++ enum pci_dma_burst_strategy *strat,
++ unsigned long *strategy_parameter)
+{
-+ doing_pdma = 1;
-+ virtual_dma_port = io;
-+ virtual_dma_mode = (mode == DMA_MODE_WRITE);
-+ virtual_dma_addr = addr;
-+ virtual_dma_count = size;
-+ virtual_dma_residue = 0;
-+ return 0;
++ *strat = PCI_DMA_BURST_INFINITY;
++ *strategy_parameter = ~0UL;
+}
++#endif
+
-+/* XEN: This trick to force 'virtual DMA' is from include/asm-m68k/floppy.h. */
-+#define FDC1 xen_floppy_init()
-+static int FDC2 = -1;
++#define HAVE_PCI_MMAP
++extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma,
++ enum pci_mmap_state mmap_state, int write_combine);
+
-+static int xen_floppy_init(void)
++static inline void pcibios_add_platform_entries(struct pci_dev *dev)
+{
-+ use_virtual_dma = 1;
-+ can_use_virtual_dma = 1;
-+ return 0x3f0;
+}
+
-+/*
-+ * Floppy types are stored in the rtc's CMOS RAM and so rtc_lock
-+ * is needed to prevent corrupted CMOS RAM in case "insmod floppy"
-+ * coincides with another rtc CMOS user. Paul G.
-+ */
-+#define FLOPPY0_TYPE ({ \
-+ unsigned long flags; \
-+ unsigned char val; \
-+ spin_lock_irqsave(&rtc_lock, flags); \
-+ val = (CMOS_READ(0x10) >> 4) & 15; \
-+ spin_unlock_irqrestore(&rtc_lock, flags); \
-+ val; \
-+})
-+
-+#define FLOPPY1_TYPE ({ \
-+ unsigned long flags; \
-+ unsigned char val; \
-+ spin_lock_irqsave(&rtc_lock, flags); \
-+ val = CMOS_READ(0x10) & 15; \
-+ spin_unlock_irqrestore(&rtc_lock, flags); \
-+ val; \
-+})
-+
-+#define N_FDC 2
-+#define N_DRIVE 8
-+
-+#define FLOPPY_MOTOR_MASK 0xf0
-+
-+#define EXTRA_FLOPPY_PARAMS
-+
-+#endif /* __ASM_XEN_X86_64_FLOPPY_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/hw_irq.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/hw_irq.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/hw_irq.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/hw_irq.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,136 @@
-+#ifndef _ASM_HW_IRQ_H
-+#define _ASM_HW_IRQ_H
-+
-+/*
-+ * linux/include/asm/hw_irq.h
-+ *
-+ * (C) 1992, 1993 Linus Torvalds, (C) 1997 Ingo Molnar
-+ *
-+ * moved some of the old arch/i386/kernel/irq.h to here. VY
-+ *
-+ * IRQ/IPI changes taken from work by Thomas Radke
-+ * <tomsoft at informatik.tu-chemnitz.de>
-+ *
-+ * hacked by Andi Kleen for x86-64.
-+ */
-+
-+#ifndef __ASSEMBLY__
-+#include <asm/atomic.h>
-+#include <asm/irq.h>
-+#include <linux/profile.h>
-+#include <linux/smp.h>
++#endif /* __KERNEL__ */
+
-+struct hw_interrupt_type;
++/* generic pci stuff */
++#ifdef CONFIG_PCI
++#include <asm-generic/pci.h>
+#endif
+
-+#define NMI_VECTOR 0x02
-+/*
-+ * IDT vectors usable for external interrupt sources start
-+ * at 0x20:
-+ */
-+#define FIRST_EXTERNAL_VECTOR 0x20
++#endif /* __x8664_PCI_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/pgalloc.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/pgalloc.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,204 @@
++#ifndef _X86_64_PGALLOC_H
++#define _X86_64_PGALLOC_H
+
-+#define IA32_SYSCALL_VECTOR 0x80
++#include <asm/fixmap.h>
++#include <asm/pda.h>
++#include <linux/threads.h>
++#include <linux/mm.h>
++#include <asm/io.h> /* for phys_to_virt and page_to_pseudophys */
+
++#include <xen/features.h>
++void make_page_readonly(void *va, unsigned int feature);
++void make_page_writable(void *va, unsigned int feature);
++void make_pages_readonly(void *va, unsigned int nr, unsigned int feature);
++void make_pages_writable(void *va, unsigned int nr, unsigned int feature);
+
-+/*
-+ * Vectors 0x20-0x2f are used for ISA interrupts.
-+ */
++#define __user_pgd(pgd) ((pgd) + PTRS_PER_PGD)
+
-+/*
-+ * Special IRQ vectors used by the SMP architecture, 0xf0-0xff
-+ *
-+ * some of the following vectors are 'rare', they are merged
-+ * into a single vector (CALL_FUNCTION_VECTOR) to save vector space.
-+ * TLB, reschedule and local APIC vectors are performance-critical.
-+ */
-+#ifndef CONFIG_XEN
-+#define SPURIOUS_APIC_VECTOR 0xff
-+#define ERROR_APIC_VECTOR 0xfe
-+#define RESCHEDULE_VECTOR 0xfd
-+#define CALL_FUNCTION_VECTOR 0xfc
-+/* fb free - please don't readd KDB here because it's useless
-+ (hint - think what a NMI bit does to a vector) */
-+#define THERMAL_APIC_VECTOR 0xfa
-+#define THRESHOLD_APIC_VECTOR 0xf9
-+/* f8 free */
-+#define INVALIDATE_TLB_VECTOR_END 0xf7
-+#define INVALIDATE_TLB_VECTOR_START 0xf0 /* f0-f7 used for TLB flush */
++static inline void pmd_populate_kernel(struct mm_struct *mm, pmd_t *pmd, pte_t *pte)
++{
++ set_pmd(pmd, __pmd(_PAGE_TABLE | __pa(pte)));
++}
+
-+#define NUM_INVALIDATE_TLB_VECTORS 8
-+#endif
++static inline void pmd_populate(struct mm_struct *mm, pmd_t *pmd, struct page *pte)
++{
++ if (unlikely((mm)->context.pinned)) {
++ BUG_ON(HYPERVISOR_update_va_mapping(
++ (unsigned long)__va(page_to_pfn(pte) << PAGE_SHIFT),
++ pfn_pte(page_to_pfn(pte), PAGE_KERNEL_RO), 0));
++ set_pmd(pmd, __pmd(_PAGE_TABLE | (page_to_pfn(pte) << PAGE_SHIFT)));
++ } else {
++ *(pmd) = __pmd(_PAGE_TABLE | (page_to_pfn(pte) << PAGE_SHIFT));
++ }
++}
+
-+/*
-+ * Local APIC timer IRQ vector is on a different priority level,
-+ * to work around the 'lost local interrupt if more than 2 IRQ
-+ * sources per level' errata.
-+ */
-+#define LOCAL_TIMER_VECTOR 0xef
++static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd)
++{
++ if (unlikely((mm)->context.pinned)) {
++ BUG_ON(HYPERVISOR_update_va_mapping(
++ (unsigned long)pmd,
++ pfn_pte(virt_to_phys(pmd)>>PAGE_SHIFT,
++ PAGE_KERNEL_RO), 0));
++ set_pud(pud, __pud(_PAGE_TABLE | __pa(pmd)));
++ } else {
++ *(pud) = __pud(_PAGE_TABLE | __pa(pmd));
++ }
++}
+
+/*
-+ * First APIC vector available to drivers: (vectors 0x30-0xee)
-+ * we start at 0x31 to spread out vectors evenly between priority
-+ * levels. (0x80 is the syscall vector)
++ * We need to use the batch mode here, but pgd_pupulate() won't be
++ * be called frequently.
+ */
-+#define FIRST_DEVICE_VECTOR 0x31
-+#define FIRST_SYSTEM_VECTOR 0xef /* duplicated in irq.h */
++static inline void pgd_populate(struct mm_struct *mm, pgd_t *pgd, pud_t *pud)
++{
++ if (unlikely((mm)->context.pinned)) {
++ BUG_ON(HYPERVISOR_update_va_mapping(
++ (unsigned long)pud,
++ pfn_pte(virt_to_phys(pud)>>PAGE_SHIFT,
++ PAGE_KERNEL_RO), 0));
++ set_pgd(pgd, __pgd(_PAGE_TABLE | __pa(pud)));
++ set_pgd(__user_pgd(pgd), __pgd(_PAGE_TABLE | __pa(pud)));
++ } else {
++ *(pgd) = __pgd(_PAGE_TABLE | __pa(pud));
++ *(__user_pgd(pgd)) = *(pgd);
++ }
++}
+
++extern struct page *pte_alloc_one(struct mm_struct *mm, unsigned long addr);
++extern void pte_free(struct page *pte);
+
-+#ifndef __ASSEMBLY__
-+extern u8 irq_vector[NR_IRQ_VECTORS];
-+#define IO_APIC_VECTOR(irq) (irq_vector[irq])
-+#define AUTO_ASSIGN -1
++static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long addr)
++{
++ struct page *pg;
+
-+/*
-+ * Various low-level irq details needed by irq.c, process.c,
-+ * time.c, io_apic.c and smp.c
-+ *
-+ * Interrupt entry/exit code at both C and assembly level
-+ */
++ pg = pte_alloc_one(mm, addr);
++ return pg ? page_address(pg) : NULL;
++}
+
-+extern void disable_8259A_irq(unsigned int irq);
-+extern void enable_8259A_irq(unsigned int irq);
-+extern int i8259A_irq_pending(unsigned int irq);
-+extern void make_8259A_irq(unsigned int irq);
-+extern void init_8259A(int aeoi);
-+extern void FASTCALL(send_IPI_self(int vector));
-+extern void init_VISWS_APIC_irqs(void);
-+extern void setup_IO_APIC(void);
-+extern void disable_IO_APIC(void);
-+extern void print_IO_APIC(void);
-+extern int IO_APIC_get_PCI_irq_vector(int bus, int slot, int fn);
-+extern void send_IPI(int dest, int vector);
-+extern void setup_ioapic_dest(void);
++static inline void pmd_free(pmd_t *pmd)
++{
++ BUG_ON((unsigned long)pmd & (PAGE_SIZE-1));
++ pte_free(virt_to_page(pmd));
++}
+
-+extern unsigned long io_apic_irqs;
++static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long addr)
++{
++ struct page *pg;
+
-+extern atomic_t irq_err_count;
-+extern atomic_t irq_mis_count;
++ pg = pte_alloc_one(mm, addr);
++ return pg ? page_address(pg) : NULL;
++}
+
-+#define IO_APIC_IRQ(x) (((x) >= 16) || ((1<<(x)) & io_apic_irqs))
++static inline void pud_free(pud_t *pud)
++{
++ BUG_ON((unsigned long)pud & (PAGE_SIZE-1));
++ pte_free(virt_to_page(pud));
++}
+
-+#define __STR(x) #x
-+#define STR(x) __STR(x)
++static inline void pgd_list_add(pgd_t *pgd)
++{
++ struct page *page = virt_to_page(pgd);
+
-+#include <asm/ptrace.h>
++ spin_lock(&pgd_lock);
++ page->index = (pgoff_t)pgd_list;
++ if (pgd_list)
++ pgd_list->private = (unsigned long)&page->index;
++ pgd_list = page;
++ page->private = (unsigned long)&pgd_list;
++ spin_unlock(&pgd_lock);
++}
+
-+#define IRQ_NAME2(nr) nr##_interrupt(void)
-+#define IRQ_NAME(nr) IRQ_NAME2(IRQ##nr)
++static inline void pgd_list_del(pgd_t *pgd)
++{
++ struct page *next, **pprev, *page = virt_to_page(pgd);
+
-+/*
-+ * SMP has a few special interrupts for IPI messages
-+ */
++ spin_lock(&pgd_lock);
++ next = (struct page *)page->index;
++ pprev = (struct page **)page->private;
++ *pprev = next;
++ if (next)
++ next->private = (unsigned long)pprev;
++ spin_unlock(&pgd_lock);
++}
+
-+#define BUILD_IRQ(nr) \
-+asmlinkage void IRQ_NAME(nr); \
-+__asm__( \
-+"\n.p2align\n" \
-+"IRQ" #nr "_interrupt:\n\t" \
-+ "push $~(" #nr ") ; " \
-+ "jmp common_interrupt");
++static inline pgd_t *pgd_alloc(struct mm_struct *mm)
++{
++ /*
++ * We allocate two contiguous pages for kernel and user.
++ */
++ unsigned boundary;
++ pgd_t *pgd = (pgd_t *)__get_free_pages(GFP_KERNEL|__GFP_REPEAT, 1);
++ if (!pgd)
++ return NULL;
++ pgd_list_add(pgd);
++ /*
++ * Copy kernel pointers in from init.
++ * Could keep a freelist or slab cache of those because the kernel
++ * part never changes.
++ */
++ boundary = pgd_index(__PAGE_OFFSET);
++ memset(pgd, 0, boundary * sizeof(pgd_t));
++ memcpy(pgd + boundary,
++ init_level4_pgt + boundary,
++ (PTRS_PER_PGD - boundary) * sizeof(pgd_t));
+
-+#define platform_legacy_irq(irq) ((irq) < 16)
++ memset(__user_pgd(pgd), 0, PAGE_SIZE); /* clean up user pgd */
++ /*
++ * Set level3_user_pgt for vsyscall area
++ */
++ __user_pgd(pgd)[pgd_index(VSYSCALL_START)] =
++ __pgd(__pa_symbol(level3_user_pgt) | _PAGE_TABLE);
++ return pgd;
++}
+
-+#endif
++static inline void pgd_free(pgd_t *pgd)
++{
++ pte_t *ptep = virt_to_ptep(pgd);
+
-+#endif /* _ASM_HW_IRQ_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/hypercall.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/hypercall.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/hypercall.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/hypercall.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,406 @@
-+/******************************************************************************
-+ * hypercall.h
-+ *
-+ * Linux-specific hypervisor handling.
-+ *
-+ * Copyright (c) 2002-2004, K A Fraser
-+ *
-+ * 64-bit updates:
-+ * Benjamin Liu <benjamin.liu at intel.com>
-+ * Jun Nakajima <jun.nakajima at intel.com>
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++ if (!pte_write(*ptep)) {
++ xen_pgd_unpin(__pa(pgd));
++ BUG_ON(HYPERVISOR_update_va_mapping(
++ (unsigned long)pgd,
++ pfn_pte(virt_to_phys(pgd)>>PAGE_SHIFT, PAGE_KERNEL),
++ 0));
++ }
+
-+#ifndef __HYPERCALL_H__
-+#define __HYPERCALL_H__
++ ptep = virt_to_ptep(__user_pgd(pgd));
+
-+#include <linux/string.h> /* memcpy() */
++ if (!pte_write(*ptep)) {
++ xen_pgd_unpin(__pa(__user_pgd(pgd)));
++ BUG_ON(HYPERVISOR_update_va_mapping(
++ (unsigned long)__user_pgd(pgd),
++ pfn_pte(virt_to_phys(__user_pgd(pgd))>>PAGE_SHIFT,
++ PAGE_KERNEL),
++ 0));
++ }
+
-+#ifndef __HYPERVISOR_H__
-+# error "please don't include this file directly"
-+#endif
++ pgd_list_del(pgd);
++ free_pages((unsigned long)pgd, 1);
++}
+
-+#define __STR(x) #x
-+#define STR(x) __STR(x)
++static inline pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
++{
++ pte_t *pte = (pte_t *)get_zeroed_page(GFP_KERNEL|__GFP_REPEAT);
++ if (pte)
++ make_page_readonly(pte, XENFEAT_writable_page_tables);
+
-+#ifdef CONFIG_XEN
-+#define HYPERCALL_STR(name) \
-+ "call hypercall_page + ("STR(__HYPERVISOR_##name)" * 32)"
-+#else
-+#define HYPERCALL_STR(name) \
-+ "mov hypercall_stubs,%%rax; " \
-+ "add $("STR(__HYPERVISOR_##name)" * 32),%%rax; " \
-+ "call *%%rax"
-+#endif
++ return pte;
++}
+
-+#define _hypercall0(type, name) \
-+({ \
-+ long __res; \
-+ asm volatile ( \
-+ HYPERCALL_STR(name) \
-+ : "=a" (__res) \
-+ : \
-+ : "memory" ); \
-+ (type)__res; \
-+})
++/* Should really implement gc for free page table pages. This could be
++ done with a reference count in struct page. */
+
-+#define _hypercall1(type, name, a1) \
-+({ \
-+ long __res, __ign1; \
-+ asm volatile ( \
-+ HYPERCALL_STR(name) \
-+ : "=a" (__res), "=D" (__ign1) \
-+ : "1" ((long)(a1)) \
-+ : "memory" ); \
-+ (type)__res; \
-+})
++static inline void pte_free_kernel(pte_t *pte)
++{
++ BUG_ON((unsigned long)pte & (PAGE_SIZE-1));
++ make_page_writable(pte, XENFEAT_writable_page_tables);
++ free_page((unsigned long)pte);
++}
+
-+#define _hypercall2(type, name, a1, a2) \
-+({ \
-+ long __res, __ign1, __ign2; \
-+ asm volatile ( \
-+ HYPERCALL_STR(name) \
-+ : "=a" (__res), "=D" (__ign1), "=S" (__ign2) \
-+ : "1" ((long)(a1)), "2" ((long)(a2)) \
-+ : "memory" ); \
-+ (type)__res; \
-+})
++#define __pte_free_tlb(tlb,pte) tlb_remove_page((tlb),(pte))
++#define __pmd_free_tlb(tlb,x) tlb_remove_page((tlb),virt_to_page(x))
++#define __pud_free_tlb(tlb,x) tlb_remove_page((tlb),virt_to_page(x))
+
-+#define _hypercall3(type, name, a1, a2, a3) \
-+({ \
-+ long __res, __ign1, __ign2, __ign3; \
-+ asm volatile ( \
-+ HYPERCALL_STR(name) \
-+ : "=a" (__res), "=D" (__ign1), "=S" (__ign2), \
-+ "=d" (__ign3) \
-+ : "1" ((long)(a1)), "2" ((long)(a2)), \
-+ "3" ((long)(a3)) \
-+ : "memory" ); \
-+ (type)__res; \
-+})
++#endif /* _X86_64_PGALLOC_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/pgtable.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/pgtable.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,583 @@
++#ifndef _X86_64_PGTABLE_H
++#define _X86_64_PGTABLE_H
++
++/*
++ * This file contains the functions and defines necessary to modify and use
++ * the x86-64 page table tree.
++ */
++#include <asm/processor.h>
++#include <asm/fixmap.h>
++#include <asm/bitops.h>
++#include <linux/threads.h>
++#include <linux/sched.h>
++#include <asm/pda.h>
++#ifdef CONFIG_XEN
++#include <asm/hypervisor.h>
+
-+#define _hypercall4(type, name, a1, a2, a3, a4) \
-+({ \
-+ long __res, __ign1, __ign2, __ign3; \
-+ asm volatile ( \
-+ "movq %7,%%r10; " \
-+ HYPERCALL_STR(name) \
-+ : "=a" (__res), "=D" (__ign1), "=S" (__ign2), \
-+ "=d" (__ign3) \
-+ : "1" ((long)(a1)), "2" ((long)(a2)), \
-+ "3" ((long)(a3)), "g" ((long)(a4)) \
-+ : "memory", "r10" ); \
-+ (type)__res; \
-+})
++extern pud_t level3_user_pgt[512];
+
-+#define _hypercall5(type, name, a1, a2, a3, a4, a5) \
-+({ \
-+ long __res, __ign1, __ign2, __ign3; \
-+ asm volatile ( \
-+ "movq %7,%%r10; movq %8,%%r8; " \
-+ HYPERCALL_STR(name) \
-+ : "=a" (__res), "=D" (__ign1), "=S" (__ign2), \
-+ "=d" (__ign3) \
-+ : "1" ((long)(a1)), "2" ((long)(a2)), \
-+ "3" ((long)(a3)), "g" ((long)(a4)), \
-+ "g" ((long)(a5)) \
-+ : "memory", "r10", "r8" ); \
-+ (type)__res; \
++extern void xen_init_pt(void);
++
++extern pte_t *lookup_address(unsigned long address);
++
++#define virt_to_ptep(va) \
++({ \
++ pte_t *__ptep = lookup_address((unsigned long)(va)); \
++ BUG_ON(!__ptep || !pte_present(*__ptep)); \
++ __ptep; \
+})
+
-+static inline int
-+HYPERVISOR_set_trap_table(
-+ trap_info_t *table)
-+{
-+ return _hypercall1(int, set_trap_table, table);
-+}
++#define arbitrary_virt_to_machine(va) \
++ (((maddr_t)pte_mfn(*virt_to_ptep(va)) << PAGE_SHIFT) \
++ | ((unsigned long)(va) & (PAGE_SIZE - 1)))
++#endif
+
-+static inline int
-+HYPERVISOR_mmu_update(
-+ mmu_update_t *req, int count, int *success_count, domid_t domid)
-+{
-+ return _hypercall4(int, mmu_update, req, count, success_count, domid);
-+}
++extern pud_t level3_kernel_pgt[512];
++extern pud_t level3_physmem_pgt[512];
++extern pud_t level3_ident_pgt[512];
++extern pmd_t level2_kernel_pgt[512];
++extern pgd_t init_level4_pgt[];
++extern pgd_t boot_level4_pgt[];
++extern unsigned long __supported_pte_mask;
+
-+static inline int
-+HYPERVISOR_mmuext_op(
-+ struct mmuext_op *op, int count, int *success_count, domid_t domid)
-+{
-+ return _hypercall4(int, mmuext_op, op, count, success_count, domid);
-+}
++#define swapper_pg_dir init_level4_pgt
+
-+static inline int
-+HYPERVISOR_set_gdt(
-+ unsigned long *frame_list, int entries)
-+{
-+ return _hypercall2(int, set_gdt, frame_list, entries);
-+}
++extern int nonx_setup(char *str);
++extern void paging_init(void);
++extern void clear_kernel_mapping(unsigned long addr, unsigned long size);
+
-+static inline int
-+HYPERVISOR_stack_switch(
-+ unsigned long ss, unsigned long esp)
-+{
-+ return _hypercall2(int, stack_switch, ss, esp);
-+}
++extern unsigned long pgkern_mask;
+
-+static inline int
-+HYPERVISOR_set_callbacks(
-+ unsigned long event_address, unsigned long failsafe_address,
-+ unsigned long syscall_address)
-+{
-+ return _hypercall3(int, set_callbacks,
-+ event_address, failsafe_address, syscall_address);
-+}
++/*
++ * ZERO_PAGE is a global shared page that is always zero: used
++ * for zero-mapped memory areas etc..
++ */
++extern unsigned long empty_zero_page[PAGE_SIZE/sizeof(unsigned long)];
++#define ZERO_PAGE(vaddr) (virt_to_page(empty_zero_page))
+
-+static inline int
-+HYPERVISOR_fpu_taskswitch(
-+ int set)
-+{
-+ return _hypercall1(int, fpu_taskswitch, set);
-+}
++/*
++ * PGDIR_SHIFT determines what a top-level page table entry can map
++ */
++#define PGDIR_SHIFT 39
++#define PTRS_PER_PGD 512
+
-+static inline int
-+HYPERVISOR_sched_op_compat(
-+ int cmd, unsigned long arg)
-+{
-+ return _hypercall2(int, sched_op_compat, cmd, arg);
-+}
++/*
++ * 3rd level page
++ */
++#define PUD_SHIFT 30
++#define PTRS_PER_PUD 512
+
-+static inline int
-+HYPERVISOR_sched_op(
-+ int cmd, void *arg)
-+{
-+ return _hypercall2(int, sched_op, cmd, arg);
-+}
++/*
++ * PMD_SHIFT determines the size of the area a middle-level
++ * page table can map
++ */
++#define PMD_SHIFT 21
++#define PTRS_PER_PMD 512
+
-+static inline long
-+HYPERVISOR_set_timer_op(
-+ u64 timeout)
-+{
-+ return _hypercall1(long, set_timer_op, timeout);
-+}
++/*
++ * entries per page directory level
++ */
++#define PTRS_PER_PTE 512
+
-+static inline int
-+HYPERVISOR_platform_op(
-+ struct xen_platform_op *platform_op)
-+{
-+ platform_op->interface_version = XENPF_INTERFACE_VERSION;
-+ return _hypercall1(int, platform_op, platform_op);
-+}
++#define pte_ERROR(e) \
++ printk("%s:%d: bad pte %p(%016lx pfn %010lx).\n", __FILE__, __LINE__, \
++ &(e), __pte_val(e), pte_pfn(e))
++#define pmd_ERROR(e) \
++ printk("%s:%d: bad pmd %p(%016lx pfn %010lx).\n", __FILE__, __LINE__, \
++ &(e), __pmd_val(e), pmd_pfn(e))
++#define pud_ERROR(e) \
++ printk("%s:%d: bad pud %p(%016lx pfn %010lx).\n", __FILE__, __LINE__, \
++ &(e), __pud_val(e), (pud_val(e) & __PHYSICAL_MASK) >> PAGE_SHIFT)
++#define pgd_ERROR(e) \
++ printk("%s:%d: bad pgd %p(%016lx pfn %010lx).\n", __FILE__, __LINE__, \
++ &(e), __pgd_val(e), (pgd_val(e) & __PHYSICAL_MASK) >> PAGE_SHIFT)
+
-+static inline int
-+HYPERVISOR_set_debugreg(
-+ int reg, unsigned long value)
-+{
-+ return _hypercall2(int, set_debugreg, reg, value);
-+}
++#define pgd_none(x) (!__pgd_val(x))
++#define pud_none(x) (!__pud_val(x))
+
-+static inline unsigned long
-+HYPERVISOR_get_debugreg(
-+ int reg)
++static inline void set_pte(pte_t *dst, pte_t val)
+{
-+ return _hypercall1(unsigned long, get_debugreg, reg);
++ *dst = val;
+}
+
-+static inline int
-+HYPERVISOR_update_descriptor(
-+ unsigned long ma, unsigned long word)
-+{
-+ return _hypercall2(int, update_descriptor, ma, word);
-+}
++#define set_pmd(pmdptr, pmdval) xen_l2_entry_update(pmdptr, (pmdval))
++#define set_pud(pudptr, pudval) xen_l3_entry_update(pudptr, (pudval))
++#define set_pgd(pgdptr, pgdval) xen_l4_entry_update(pgdptr, (pgdval))
+
-+static inline int
-+HYPERVISOR_memory_op(
-+ unsigned int cmd, void *arg)
++static inline void pud_clear (pud_t * pud)
+{
-+ return _hypercall2(int, memory_op, cmd, arg);
++ set_pud(pud, __pud(0));
+}
+
-+static inline int
-+HYPERVISOR_multicall(
-+ multicall_entry_t *call_list, int nr_calls)
-+{
-+ return _hypercall2(int, multicall, call_list, nr_calls);
-+}
++#define __user_pgd(pgd) ((pgd) + PTRS_PER_PGD)
+
-+static inline int
-+HYPERVISOR_update_va_mapping(
-+ unsigned long va, pte_t new_val, unsigned long flags)
++static inline void pgd_clear (pgd_t * pgd)
+{
-+ return _hypercall3(int, update_va_mapping, va, new_val.pte, flags);
++ set_pgd(pgd, __pgd(0));
++ set_pgd(__user_pgd(pgd), __pgd(0));
+}
+
-+static inline int
-+HYPERVISOR_event_channel_op(
-+ int cmd, void *arg)
-+{
-+ int rc = _hypercall2(int, event_channel_op, cmd, arg);
++#define pud_page(pud) \
++ ((unsigned long) __va(pud_val(pud) & PHYSICAL_PAGE_MASK))
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (unlikely(rc == -ENOSYS)) {
-+ struct evtchn_op op;
-+ op.cmd = cmd;
-+ memcpy(&op.u, arg, sizeof(op.u));
-+ rc = _hypercall1(int, event_channel_op_compat, &op);
-+ memcpy(arg, &op.u, sizeof(op.u));
-+ }
-+#endif
++#define pte_same(a, b) ((a).pte == (b).pte)
+
-+ return rc;
-+}
++#define pte_pgprot(a) (__pgprot((a).pte & ~PHYSICAL_PAGE_MASK))
+
-+static inline int
-+HYPERVISOR_acm_op(
-+ int cmd, void *arg)
-+{
-+ return _hypercall2(int, acm_op, cmd, arg);
-+}
++#define PMD_SIZE (1UL << PMD_SHIFT)
++#define PMD_MASK (~(PMD_SIZE-1))
++#define PUD_SIZE (1UL << PUD_SHIFT)
++#define PUD_MASK (~(PUD_SIZE-1))
++#define PGDIR_SIZE (1UL << PGDIR_SHIFT)
++#define PGDIR_MASK (~(PGDIR_SIZE-1))
+
-+static inline int
-+HYPERVISOR_xen_version(
-+ int cmd, void *arg)
-+{
-+ return _hypercall2(int, xen_version, cmd, arg);
-+}
++#define USER_PTRS_PER_PGD ((TASK_SIZE-1)/PGDIR_SIZE+1)
++#define FIRST_USER_ADDRESS 0
+
-+static inline int
-+HYPERVISOR_console_io(
-+ int cmd, int count, char *str)
-+{
-+ return _hypercall3(int, console_io, cmd, count, str);
-+}
++#ifndef __ASSEMBLY__
++#define MAXMEM 0x3fffffffffffUL
++#define VMALLOC_START 0xffffc20000000000UL
++#define VMALLOC_END 0xffffe1ffffffffffUL
++#define MODULES_VADDR 0xffffffff88000000UL
++#define MODULES_END 0xfffffffffff00000UL
++#define MODULES_LEN (MODULES_END - MODULES_VADDR)
+
-+static inline int
-+HYPERVISOR_physdev_op(
-+ int cmd, void *arg)
-+{
-+ int rc = _hypercall2(int, physdev_op, cmd, arg);
++#define _PAGE_BIT_PRESENT 0
++#define _PAGE_BIT_RW 1
++#define _PAGE_BIT_USER 2
++#define _PAGE_BIT_PWT 3
++#define _PAGE_BIT_PCD 4
++#define _PAGE_BIT_ACCESSED 5
++#define _PAGE_BIT_DIRTY 6
++#define _PAGE_BIT_PSE 7 /* 4 MB (or 2MB) page */
++#define _PAGE_BIT_GLOBAL 8 /* Global TLB entry PPro+ */
++#define _PAGE_BIT_NX 63 /* No execute: only valid after cpuid check */
++
++#define _PAGE_PRESENT 0x001
++#define _PAGE_RW 0x002
++#define _PAGE_USER 0x004
++#define _PAGE_PWT 0x008
++#define _PAGE_PCD 0x010
++#define _PAGE_ACCESSED 0x020
++#define _PAGE_DIRTY 0x040
++#define _PAGE_PSE 0x080 /* 2MB page */
++#define _PAGE_FILE 0x040 /* nonlinear file mapping, saved PTE; unset:swap */
++#define _PAGE_GLOBAL 0x100 /* Global TLB entry */
++
++#define _PAGE_PROTNONE 0x080 /* If not present */
++#define _PAGE_NX (1UL<<_PAGE_BIT_NX)
++
++/* Mapped page is I/O or foreign and has no associated page struct. */
++#define _PAGE_IO 0x200
+
+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (unlikely(rc == -ENOSYS)) {
-+ struct physdev_op op;
-+ op.cmd = cmd;
-+ memcpy(&op.u, arg, sizeof(op.u));
-+ rc = _hypercall1(int, physdev_op_compat, &op);
-+ memcpy(arg, &op.u, sizeof(op.u));
-+ }
++extern unsigned int __kernel_page_user;
++#else
++#define __kernel_page_user 0
+#endif
+
-+ return rc;
-+}
++#define _PAGE_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED | _PAGE_DIRTY)
++#define _KERNPG_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY | __kernel_page_user)
+
-+static inline int
-+HYPERVISOR_grant_table_op(
-+ unsigned int cmd, void *uop, unsigned int count)
-+{
-+ return _hypercall3(int, grant_table_op, cmd, uop, count);
-+}
++#define _PAGE_CHG_MASK (PTE_MASK | _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_IO)
+
-+static inline int
-+HYPERVISOR_update_va_mapping_otherdomain(
-+ unsigned long va, pte_t new_val, unsigned long flags, domid_t domid)
-+{
-+ return _hypercall4(int, update_va_mapping_otherdomain, va,
-+ new_val.pte, flags, domid);
-+}
++#define PAGE_NONE __pgprot(_PAGE_PROTNONE | _PAGE_ACCESSED)
++#define PAGE_SHARED __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX)
++#define PAGE_SHARED_EXEC __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED)
++#define PAGE_COPY_NOEXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX)
++#define PAGE_COPY PAGE_COPY_NOEXEC
++#define PAGE_COPY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED)
++#define PAGE_READONLY __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX)
++#define PAGE_READONLY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED)
++#define __PAGE_KERNEL \
++ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_NX | __kernel_page_user)
++#define __PAGE_KERNEL_EXEC \
++ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED | __kernel_page_user)
++#define __PAGE_KERNEL_NOCACHE \
++ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_PCD | _PAGE_ACCESSED | _PAGE_NX | __kernel_page_user)
++#define __PAGE_KERNEL_RO \
++ (_PAGE_PRESENT | _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_NX | __kernel_page_user)
++#define __PAGE_KERNEL_VSYSCALL \
++ (_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED)
++#define __PAGE_KERNEL_VSYSCALL_NOCACHE \
++ (_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED | _PAGE_PCD)
++#define __PAGE_KERNEL_LARGE \
++ (__PAGE_KERNEL | _PAGE_PSE)
++#define __PAGE_KERNEL_LARGE_EXEC \
++ (__PAGE_KERNEL_EXEC | _PAGE_PSE)
+
-+static inline int
-+HYPERVISOR_vm_assist(
-+ unsigned int cmd, unsigned int type)
-+{
-+ return _hypercall2(int, vm_assist, cmd, type);
-+}
++/*
++ * We don't support GLOBAL page in xenolinux64
++ */
++#define MAKE_GLOBAL(x) __pgprot((x))
+
-+static inline int
-+HYPERVISOR_vcpu_op(
-+ int cmd, int vcpuid, void *extra_args)
-+{
-+ return _hypercall3(int, vcpu_op, cmd, vcpuid, extra_args);
-+}
++#define PAGE_KERNEL MAKE_GLOBAL(__PAGE_KERNEL)
++#define PAGE_KERNEL_EXEC MAKE_GLOBAL(__PAGE_KERNEL_EXEC)
++#define PAGE_KERNEL_RO MAKE_GLOBAL(__PAGE_KERNEL_RO)
++#define PAGE_KERNEL_NOCACHE MAKE_GLOBAL(__PAGE_KERNEL_NOCACHE)
++#define PAGE_KERNEL_VSYSCALL32 __pgprot(__PAGE_KERNEL_VSYSCALL)
++#define PAGE_KERNEL_VSYSCALL MAKE_GLOBAL(__PAGE_KERNEL_VSYSCALL)
++#define PAGE_KERNEL_LARGE MAKE_GLOBAL(__PAGE_KERNEL_LARGE)
++#define PAGE_KERNEL_VSYSCALL_NOCACHE MAKE_GLOBAL(__PAGE_KERNEL_VSYSCALL_NOCACHE)
+
-+static inline int
-+HYPERVISOR_set_segment_base(
-+ int reg, unsigned long value)
++/* xwr */
++#define __P000 PAGE_NONE
++#define __P001 PAGE_READONLY
++#define __P010 PAGE_COPY
++#define __P011 PAGE_COPY
++#define __P100 PAGE_READONLY_EXEC
++#define __P101 PAGE_READONLY_EXEC
++#define __P110 PAGE_COPY_EXEC
++#define __P111 PAGE_COPY_EXEC
++
++#define __S000 PAGE_NONE
++#define __S001 PAGE_READONLY
++#define __S010 PAGE_SHARED
++#define __S011 PAGE_SHARED
++#define __S100 PAGE_READONLY_EXEC
++#define __S101 PAGE_READONLY_EXEC
++#define __S110 PAGE_SHARED_EXEC
++#define __S111 PAGE_SHARED_EXEC
++
++static inline unsigned long pgd_bad(pgd_t pgd)
+{
-+ return _hypercall2(int, set_segment_base, reg, value);
++ unsigned long val = __pgd_val(pgd);
++ val &= ~PTE_MASK;
++ val &= ~(_PAGE_USER | _PAGE_DIRTY);
++ return val & ~(_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED);
+}
+
-+static inline int
-+HYPERVISOR_suspend(
-+ unsigned long srec)
-+{
-+ struct sched_shutdown sched_shutdown = {
-+ .reason = SHUTDOWN_suspend
-+ };
++static inline unsigned long pud_bad(pud_t pud)
++{
++ unsigned long val = __pud_val(pud);
++ val &= ~PTE_MASK;
++ val &= ~(_PAGE_USER | _PAGE_DIRTY);
++ return val & ~(_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED);
++}
+
-+ int rc = _hypercall3(int, sched_op, SCHEDOP_shutdown,
-+ &sched_shutdown, srec);
++#define set_pte_at(_mm,addr,ptep,pteval) do { \
++ if (((_mm) != current->mm && (_mm) != &init_mm) || \
++ HYPERVISOR_update_va_mapping((addr), (pteval), 0)) \
++ set_pte((ptep), (pteval)); \
++} while (0)
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (rc == -ENOSYS)
-+ rc = _hypercall3(int, sched_op_compat, SCHEDOP_shutdown,
-+ SHUTDOWN_suspend, srec);
-+#endif
++#define pte_none(x) (!(x).pte)
++#define pte_present(x) ((x).pte & (_PAGE_PRESENT | _PAGE_PROTNONE))
++#define pte_clear(mm,addr,xp) do { set_pte_at(mm, addr, xp, __pte(0)); } while (0)
+
-+ return rc;
-+}
++#define pages_to_mb(x) ((x) >> (20-PAGE_SHIFT))
+
-+static inline int
-+HYPERVISOR_nmi_op(
-+ unsigned long op, void *arg)
-+{
-+ return _hypercall2(int, nmi_op, op, arg);
-+}
++#define __pte_mfn(_pte) (((_pte).pte & PTE_MASK) >> PAGE_SHIFT)
++#define pte_mfn(_pte) ((_pte).pte & _PAGE_PRESENT ? \
++ __pte_mfn(_pte) : pfn_to_mfn(__pte_mfn(_pte)))
++#define pte_pfn(_pte) ((_pte).pte & _PAGE_IO ? end_pfn : \
++ (_pte).pte & _PAGE_PRESENT ? \
++ mfn_to_local_pfn(__pte_mfn(_pte)) : \
++ __pte_mfn(_pte))
+
-+static inline unsigned long
-+HYPERVISOR_hvm_op(
-+ int op, void *arg)
-+{
-+ return _hypercall2(unsigned long, hvm_op, op, arg);
-+}
++#define pte_page(x) pfn_to_page(pte_pfn(x))
+
-+static inline int
-+HYPERVISOR_callback_op(
-+ int cmd, void *arg)
++static inline pte_t pfn_pte(unsigned long page_nr, pgprot_t pgprot)
+{
-+ return _hypercall2(int, callback_op, cmd, arg);
++ unsigned long pte = page_nr << PAGE_SHIFT;
++ pte |= pgprot_val(pgprot);
++ pte &= __supported_pte_mask;
++ return __pte(pte);
+}
+
-+static inline int
-+HYPERVISOR_xenoprof_op(
-+ int op, void *arg)
++static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
+{
-+ return _hypercall2(int, xenoprof_op, op, arg);
++ pte_t pte = *ptep;
++ if (!pte_none(pte)) {
++ if ((mm != &init_mm) ||
++ HYPERVISOR_update_va_mapping(addr, __pte(0), 0))
++ pte = __pte_ma(xchg(&ptep->pte, 0));
++ }
++ return pte;
+}
+
-+static inline int
-+HYPERVISOR_kexec_op(
-+ unsigned long op, void *args)
++static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm, unsigned long addr, pte_t *ptep, int full)
+{
-+ return _hypercall2(int, kexec_op, op, args);
++ if (full) {
++ pte_t pte = *ptep;
++ if (mm->context.pinned)
++ xen_l1_entry_update(ptep, __pte(0));
++ else
++ *ptep = __pte(0);
++ return pte;
++ }
++ return ptep_get_and_clear(mm, addr, ptep);
+}
+
-+#endif /* __HYPERCALL_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/hypervisor.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/hypervisor.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/hypervisor.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/hypervisor.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,2 @@
-+
-+#include <asm-i386/mach-xen/asm/hypervisor.h>
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/io.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/io.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/io.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/io.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,330 @@
-+#ifndef _ASM_IO_H
-+#define _ASM_IO_H
-+
-+#include <asm/fixmap.h>
++#define ptep_clear_flush(vma, addr, ptep) \
++({ \
++ pte_t *__ptep = (ptep); \
++ pte_t __res = *__ptep; \
++ if (!pte_none(__res) && \
++ ((vma)->vm_mm != current->mm || \
++ HYPERVISOR_update_va_mapping(addr, __pte(0), \
++ (unsigned long)(vma)->vm_mm->cpu_vm_mask.bits| \
++ UVMF_INVLPG|UVMF_MULTI))) { \
++ __ptep->pte = 0; \
++ flush_tlb_page(vma, addr); \
++ } \
++ __res; \
++})
+
+/*
-+ * This file contains the definitions for the x86 IO instructions
-+ * inb/inw/inl/outb/outw/outl and the "string versions" of the same
-+ * (insb/insw/insl/outsb/outsw/outsl). You can also use "pausing"
-+ * versions of the single-IO instructions (inb_p/inw_p/..).
-+ *
-+ * This file is not meant to be obfuscating: it's just complicated
-+ * to (a) handle it all in a way that makes gcc able to optimize it
-+ * as well as possible and (b) trying to avoid writing the same thing
-+ * over and over again with slight variations and possibly making a
-+ * mistake somewhere.
++ * The following only work if pte_present() is true.
++ * Undefined behaviour if not..
+ */
++#define __LARGE_PTE (_PAGE_PSE|_PAGE_PRESENT)
++static inline int pte_user(pte_t pte) { return __pte_val(pte) & _PAGE_USER; }
++static inline int pte_read(pte_t pte) { return __pte_val(pte) & _PAGE_USER; }
++static inline int pte_exec(pte_t pte) { return __pte_val(pte) & _PAGE_USER; }
++static inline int pte_dirty(pte_t pte) { return __pte_val(pte) & _PAGE_DIRTY; }
++static inline int pte_young(pte_t pte) { return __pte_val(pte) & _PAGE_ACCESSED; }
++static inline int pte_write(pte_t pte) { return __pte_val(pte) & _PAGE_RW; }
++static inline int pte_file(pte_t pte) { return __pte_val(pte) & _PAGE_FILE; }
++static inline int pte_huge(pte_t pte) { return __pte_val(pte) & _PAGE_PSE; }
+
-+/*
-+ * Thanks to James van Artsdalen for a better timing-fix than
-+ * the two short jumps: using outb's to a nonexistent port seems
-+ * to guarantee better timings even on fast machines.
-+ *
-+ * On the other hand, I'd like to be sure of a non-existent port:
-+ * I feel a bit unsafe about using 0x80 (should be safe, though)
-+ *
-+ * Linus
-+ */
++static inline pte_t pte_rdprotect(pte_t pte) { __pte_val(pte) &= ~_PAGE_USER; return pte; }
++static inline pte_t pte_exprotect(pte_t pte) { __pte_val(pte) &= ~_PAGE_USER; return pte; }
++static inline pte_t pte_mkclean(pte_t pte) { __pte_val(pte) &= ~_PAGE_DIRTY; return pte; }
++static inline pte_t pte_mkold(pte_t pte) { __pte_val(pte) &= ~_PAGE_ACCESSED; return pte; }
++static inline pte_t pte_wrprotect(pte_t pte) { __pte_val(pte) &= ~_PAGE_RW; return pte; }
++static inline pte_t pte_mkread(pte_t pte) { __pte_val(pte) |= _PAGE_USER; return pte; }
++static inline pte_t pte_mkexec(pte_t pte) { __pte_val(pte) |= _PAGE_USER; return pte; }
++static inline pte_t pte_mkdirty(pte_t pte) { __pte_val(pte) |= _PAGE_DIRTY; return pte; }
++static inline pte_t pte_mkyoung(pte_t pte) { __pte_val(pte) |= _PAGE_ACCESSED; return pte; }
++static inline pte_t pte_mkwrite(pte_t pte) { __pte_val(pte) |= _PAGE_RW; return pte; }
++static inline pte_t pte_mkhuge(pte_t pte) { __pte_val(pte) |= _PAGE_PSE; return pte; }
+
-+ /*
-+ * Bit simplified and optimized by Jan Hubicka
-+ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999.
-+ *
-+ * isa_memset_io, isa_memcpy_fromio, isa_memcpy_toio added,
-+ * isa_read[wl] and isa_write[wl] fixed
-+ * - Arnaldo Carvalho de Melo <acme at conectiva.com.br>
-+ */
++#define ptep_test_and_clear_dirty(vma, addr, ptep) \
++({ \
++ pte_t __pte = *(ptep); \
++ int __ret = pte_dirty(__pte); \
++ if (__ret) \
++ set_pte_at((vma)->vm_mm, addr, ptep, pte_mkclean(__pte)); \
++ __ret; \
++})
+
-+#define __SLOW_DOWN_IO "\noutb %%al,$0x80"
++#define ptep_test_and_clear_young(vma, addr, ptep) \
++({ \
++ pte_t __pte = *(ptep); \
++ int __ret = pte_young(__pte); \
++ if (__ret) \
++ set_pte_at((vma)->vm_mm, addr, ptep, pte_mkold(__pte)); \
++ __ret; \
++})
+
-+#ifdef REALLY_SLOW_IO
-+#define __FULL_SLOW_DOWN_IO __SLOW_DOWN_IO __SLOW_DOWN_IO __SLOW_DOWN_IO __SLOW_DOWN_IO
-+#else
-+#define __FULL_SLOW_DOWN_IO __SLOW_DOWN_IO
-+#endif
++static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
++{
++ pte_t pte = *ptep;
++ if (pte_write(pte))
++ set_pte_at(mm, addr, ptep, pte_wrprotect(pte));
++}
+
+/*
-+ * Talk about misusing macros..
++ * Macro to mark a page protection value as "uncacheable".
+ */
-+#define __OUT1(s,x) \
-+static inline void out##s(unsigned x value, unsigned short port) {
++#define pgprot_noncached(prot) (__pgprot(pgprot_val(prot) | _PAGE_PCD | _PAGE_PWT))
+
-+#define __OUT2(s,s1,s2) \
-+__asm__ __volatile__ ("out" #s " %" s1 "0,%" s2 "1"
++static inline int pmd_large(pmd_t pte) {
++ return (__pmd_val(pte) & __LARGE_PTE) == __LARGE_PTE;
++}
+
-+#define __OUT(s,s1,x) \
-+__OUT1(s,x) __OUT2(s,s1,"w") : : "a" (value), "Nd" (port)); } \
-+__OUT1(s##_p,x) __OUT2(s,s1,"w") __FULL_SLOW_DOWN_IO : : "a" (value), "Nd" (port));} \
+
-+#define __IN1(s) \
-+static inline RETURN_TYPE in##s(unsigned short port) { RETURN_TYPE _v;
++/*
++ * Conversion functions: convert a page and protection to a page entry,
++ * and a page entry and page directory to the page they refer to.
++ */
+
-+#define __IN2(s,s1,s2) \
-+__asm__ __volatile__ ("in" #s " %" s2 "1,%" s1 "0"
++/*
++ * Level 4 access.
++ * Never use these in the common code.
++ */
++#define pgd_page(pgd) ((unsigned long) __va(pgd_val(pgd) & PTE_MASK))
++#define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD-1))
++#define pgd_offset(mm, addr) ((mm)->pgd + pgd_index(addr))
++#define pgd_offset_k(address) (init_level4_pgt + pgd_index(address))
++#define pgd_present(pgd) (__pgd_val(pgd) & _PAGE_PRESENT)
++#define mk_kernel_pgd(address) __pgd((address) | _KERNPG_TABLE)
+
-+#define __IN(s,s1,i...) \
-+__IN1(s) __IN2(s,s1,"w") : "=a" (_v) : "Nd" (port) ,##i ); return _v; } \
-+__IN1(s##_p) __IN2(s,s1,"w") __FULL_SLOW_DOWN_IO : "=a" (_v) : "Nd" (port) ,##i ); return _v; } \
++/* PUD - Level3 access */
++/* to find an entry in a page-table-directory. */
++#define pud_index(address) (((address) >> PUD_SHIFT) & (PTRS_PER_PUD-1))
++#define pud_offset(pgd, address) ((pud_t *) pgd_page(*(pgd)) + pud_index(address))
++#define pud_present(pud) (__pud_val(pud) & _PAGE_PRESENT)
+
-+#define __INS(s) \
-+static inline void ins##s(unsigned short port, void * addr, unsigned long count) \
-+{ __asm__ __volatile__ ("rep ; ins" #s \
-+: "=D" (addr), "=c" (count) : "d" (port),"0" (addr),"1" (count)); }
++/* PMD - Level 2 access */
++#define pmd_page_kernel(pmd) ((unsigned long) __va(pmd_val(pmd) & PTE_MASK))
++#define pmd_page(pmd) (pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT))
+
-+#define __OUTS(s) \
-+static inline void outs##s(unsigned short port, const void * addr, unsigned long count) \
-+{ __asm__ __volatile__ ("rep ; outs" #s \
-+: "=S" (addr), "=c" (count) : "d" (port),"0" (addr),"1" (count)); }
++#define pmd_index(address) (((address) >> PMD_SHIFT) & (PTRS_PER_PMD-1))
++#define pmd_offset(dir, address) ((pmd_t *) pud_page(*(dir)) + \
++ pmd_index(address))
++#define pmd_none(x) (!__pmd_val(x))
++#if CONFIG_XEN_COMPAT <= 0x030002
++/* pmd_present doesn't just test the _PAGE_PRESENT bit since wr.p.t.
++ can temporarily clear it. */
++#define pmd_present(x) (__pmd_val(x))
++#else
++#define pmd_present(x) (__pmd_val(x) & _PAGE_PRESENT)
++#endif
++#define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0)
++#define pmd_bad(x) ((__pmd_val(x) & ~(PTE_MASK | _PAGE_USER | _PAGE_PRESENT)) \
++ != (_KERNPG_TABLE & ~(_PAGE_USER | _PAGE_PRESENT)))
++#define pfn_pmd(nr,prot) (__pmd(((nr) << PAGE_SHIFT) | pgprot_val(prot)))
++#define pmd_pfn(x) ((pmd_val(x) & __PHYSICAL_MASK) >> PAGE_SHIFT)
+
-+#define RETURN_TYPE unsigned char
-+__IN(b,"")
-+#undef RETURN_TYPE
-+#define RETURN_TYPE unsigned short
-+__IN(w,"")
-+#undef RETURN_TYPE
-+#define RETURN_TYPE unsigned int
-+__IN(l,"")
-+#undef RETURN_TYPE
++#define pte_to_pgoff(pte) ((__pte_val(pte) & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT)
++#define pgoff_to_pte(off) ((pte_t) { ((off) << PAGE_SHIFT) | _PAGE_FILE })
++#define PTE_FILE_MAX_BITS __PHYSICAL_MASK_SHIFT
+
-+__OUT(b,"b",char)
-+__OUT(w,"w",short)
-+__OUT(l,,int)
++/* PTE - Level 1 access. */
+
-+__INS(b)
-+__INS(w)
-+__INS(l)
++/* page, protection -> pte */
++#define mk_pte(page, pgprot) pfn_pte(page_to_pfn(page), (pgprot))
++#define mk_pte_huge(entry) (__pte_val(entry) |= _PAGE_PRESENT | _PAGE_PSE)
++
++/* physical address -> PTE */
++static inline pte_t mk_pte_phys(unsigned long physpage, pgprot_t pgprot)
++{
++ unsigned long pteval;
++ pteval = physpage | pgprot_val(pgprot);
++ return __pte(pteval);
++}
++
++/* Change flags of a PTE */
++static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
++{
++ /*
++ * Since this might change the present bit (which controls whether
++ * a pte_t object has undergone p2m translation), we must use
++ * pte_val() on the input pte and __pte() for the return value.
++ */
++ unsigned long pteval = pte_val(pte);
+
-+__OUTS(b)
-+__OUTS(w)
-+__OUTS(l)
++ pteval &= _PAGE_CHG_MASK;
++ pteval |= pgprot_val(newprot);
++ pteval &= __supported_pte_mask;
++ return __pte(pteval);
++}
+
-+#define IO_SPACE_LIMIT 0xffff
++#define pte_index(address) \
++ (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1))
++#define pte_offset_kernel(dir, address) ((pte_t *) pmd_page_kernel(*(dir)) + \
++ pte_index(address))
+
-+#if defined(__KERNEL__) && __x86_64__
++/* x86-64 always has all page tables mapped. */
++#define pte_offset_map(dir,address) pte_offset_kernel(dir,address)
++#define pte_offset_map_nested(dir,address) pte_offset_kernel(dir,address)
++#define pte_unmap(pte) /* NOP */
++#define pte_unmap_nested(pte) /* NOP */
+
-+#include <linux/vmalloc.h>
++#define update_mmu_cache(vma,address,pte) do { } while (0)
+
-+#ifndef __i386__
+/*
-+ * Change virtual addresses to physical addresses and vv.
-+ * These are pretty trivial
++ * Rules for using ptep_establish: the pte MUST be a user pte, and
++ * must be a present->present transition.
+ */
-+static inline unsigned long virt_to_phys(volatile void * address)
-+{
-+ return __pa(address);
-+}
-+
-+static inline void * phys_to_virt(unsigned long address)
-+{
-+ return __va(address);
-+}
++#define __HAVE_ARCH_PTEP_ESTABLISH
++#define ptep_establish(vma, address, ptep, pteval) \
++ do { \
++ if ( likely((vma)->vm_mm == current->mm) ) { \
++ BUG_ON(HYPERVISOR_update_va_mapping(address, \
++ pteval, \
++ (unsigned long)(vma)->vm_mm->cpu_vm_mask.bits| \
++ UVMF_INVLPG|UVMF_MULTI)); \
++ } else { \
++ xen_l1_entry_update(ptep, pteval); \
++ flush_tlb_page(vma, address); \
++ } \
++ } while (0)
+
-+#define virt_to_bus(_x) phys_to_machine(__pa(_x))
-+#define bus_to_virt(_x) __va(machine_to_phys(_x))
-+#endif
++/* We only update the dirty/accessed state if we set
++ * the dirty bit by hand in the kernel, since the hardware
++ * will do the accessed bit for us, and we don't want to
++ * race with other CPU's that might be updating the dirty
++ * bit at the same time. */
++#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
++#define ptep_set_access_flags(vma, address, ptep, entry, dirty) \
++ do { \
++ if (dirty) \
++ ptep_establish(vma, address, ptep, entry); \
++ } while (0)
+
-+/*
-+ * Change "struct page" to physical address.
-+ */
-+#define page_to_pseudophys(page) ((dma_addr_t)page_to_pfn(page) << PAGE_SHIFT)
-+#define page_to_phys(page) (phys_to_machine(page_to_pseudophys(page)))
-+#define page_to_bus(page) (phys_to_machine(page_to_pseudophys(page)))
++/* Encode and de-code a swap entry */
++#define __swp_type(x) (((x).val >> 1) & 0x3f)
++#define __swp_offset(x) ((x).val >> 8)
++#define __swp_entry(type, offset) ((swp_entry_t) { ((type) << 1) | ((offset) << 8) })
++#define __pte_to_swp_entry(pte) ((swp_entry_t) { __pte_val(pte) })
++#define __swp_entry_to_pte(x) ((pte_t) { (x).val })
+
-+#define bio_to_pseudophys(bio) (page_to_pseudophys(bio_page((bio))) + \
-+ (unsigned long) bio_offset((bio)))
-+#define bvec_to_pseudophys(bv) (page_to_pseudophys((bv)->bv_page) + \
-+ (unsigned long) (bv)->bv_offset)
++extern spinlock_t pgd_lock;
++extern struct page *pgd_list;
++void vmalloc_sync_all(void);
+
-+#define BIOVEC_PHYS_MERGEABLE(vec1, vec2) \
-+ (((bvec_to_phys((vec1)) + (vec1)->bv_len) == bvec_to_phys((vec2))) && \
-+ ((bvec_to_pseudophys((vec1)) + (vec1)->bv_len) == \
-+ bvec_to_pseudophys((vec2))))
++#endif /* !__ASSEMBLY__ */
+
-+#include <asm-generic/iomap.h>
++extern int kern_addr_valid(unsigned long addr);
+
-+extern void __iomem *__ioremap(unsigned long offset, unsigned long size, unsigned long flags);
++#define DOMID_LOCAL (0xFFFFU)
+
-+static inline void __iomem * ioremap (unsigned long offset, unsigned long size)
-+{
-+ return __ioremap(offset, size, 0);
-+}
++struct vm_area_struct;
+
-+extern void *bt_ioremap(unsigned long addr, unsigned long size);
-+extern void bt_iounmap(void *addr, unsigned long size);
-+#define early_ioremap bt_ioremap
-+#define early_iounmap bt_iounmap
++int direct_remap_pfn_range(struct vm_area_struct *vma,
++ unsigned long address,
++ unsigned long mfn,
++ unsigned long size,
++ pgprot_t prot,
++ domid_t domid);
+
-+/*
-+ * This one maps high address device memory and turns off caching for that area.
-+ * it's useful if some control registers are in such an area and write combining
-+ * or read caching is not desirable:
-+ */
-+extern void __iomem * ioremap_nocache (unsigned long offset, unsigned long size);
-+extern void iounmap(volatile void __iomem *addr);
++int direct_kernel_remap_pfn_range(unsigned long address,
++ unsigned long mfn,
++ unsigned long size,
++ pgprot_t prot,
++ domid_t domid);
+
-+/*
-+ * ISA I/O bus memory addresses are 1:1 with the physical address.
-+ */
++int create_lookup_pte_addr(struct mm_struct *mm,
++ unsigned long address,
++ uint64_t *ptep);
+
-+#define isa_virt_to_bus(_x) isa_virt_to_bus_is_UNSUPPORTED->x
-+#define isa_page_to_bus(_x) isa_page_to_bus_is_UNSUPPORTED->x
-+#define isa_bus_to_virt(_x) (void *)(__fix_to_virt(FIX_ISAMAP_BEGIN) + (_x))
++int touch_pte_range(struct mm_struct *mm,
++ unsigned long address,
++ unsigned long size);
+
-+/*
-+ * However PCI ones are not necessarily 1:1 and therefore these interfaces
-+ * are forbidden in portable PCI drivers.
-+ *
-+ * Allow them on x86 for legacy drivers, though.
-+ */
-+#define virt_to_bus(_x) phys_to_machine(__pa(_x))
-+#define bus_to_virt(_x) __va(machine_to_phys(_x))
++int xen_change_pte_range(struct mm_struct *mm, pmd_t *pmd,
++ unsigned long addr, unsigned long end, pgprot_t newprot);
+
-+/*
-+ * readX/writeX() are used to access memory mapped devices. On some
-+ * architectures the memory mapped IO stuff needs to be accessed
-+ * differently. On the x86 architecture, we just read/write the
-+ * memory location directly.
-+ */
++#define arch_change_pte_range(mm, pmd, addr, end, newprot) \
++ xen_change_pte_range(mm, pmd, addr, end, newprot)
+
-+static inline __u8 __readb(const volatile void __iomem *addr)
-+{
-+ return *(__force volatile __u8 *)addr;
-+}
-+static inline __u16 __readw(const volatile void __iomem *addr)
-+{
-+ return *(__force volatile __u16 *)addr;
-+}
-+static __always_inline __u32 __readl(const volatile void __iomem *addr)
-+{
-+ return *(__force volatile __u32 *)addr;
-+}
-+static inline __u64 __readq(const volatile void __iomem *addr)
-+{
-+ return *(__force volatile __u64 *)addr;
-+}
-+#define readb(x) __readb(x)
-+#define readw(x) __readw(x)
-+#define readl(x) __readl(x)
-+#define readq(x) __readq(x)
-+#define readb_relaxed(a) readb(a)
-+#define readw_relaxed(a) readw(a)
-+#define readl_relaxed(a) readl(a)
-+#define readq_relaxed(a) readq(a)
-+#define __raw_readb readb
-+#define __raw_readw readw
-+#define __raw_readl readl
-+#define __raw_readq readq
++#define io_remap_pfn_range(vma, vaddr, pfn, size, prot) \
++ direct_remap_pfn_range(vma,vaddr,pfn,size,prot,DOMID_IO)
+
-+#define mmiowb()
++#define MK_IOSPACE_PFN(space, pfn) (pfn)
++#define GET_IOSPACE(pfn) 0
++#define GET_PFN(pfn) (pfn)
+
-+static inline void __writel(__u32 b, volatile void __iomem *addr)
-+{
-+ *(__force volatile __u32 *)addr = b;
-+}
-+static inline void __writeq(__u64 b, volatile void __iomem *addr)
-+{
-+ *(__force volatile __u64 *)addr = b;
-+}
-+static inline void __writeb(__u8 b, volatile void __iomem *addr)
-+{
-+ *(__force volatile __u8 *)addr = b;
-+}
-+static inline void __writew(__u16 b, volatile void __iomem *addr)
-+{
-+ *(__force volatile __u16 *)addr = b;
-+}
-+#define writeq(val,addr) __writeq((val),(addr))
-+#define writel(val,addr) __writel((val),(addr))
-+#define writew(val,addr) __writew((val),(addr))
-+#define writeb(val,addr) __writeb((val),(addr))
-+#define __raw_writeb writeb
-+#define __raw_writew writew
-+#define __raw_writel writel
-+#define __raw_writeq writeq
++#define HAVE_ARCH_UNMAPPED_AREA
+
-+void __memcpy_fromio(void*,unsigned long,unsigned);
-+void __memcpy_toio(unsigned long,const void*,unsigned);
++#define pgtable_cache_init() do { } while (0)
++#define check_pgt_cache() do { } while (0)
+
-+static inline void memcpy_fromio(void *to, const volatile void __iomem *from, unsigned len)
-+{
-+ __memcpy_fromio(to,(unsigned long)from,len);
-+}
-+static inline void memcpy_toio(volatile void __iomem *to, const void *from, unsigned len)
-+{
-+ __memcpy_toio((unsigned long)to,from,len);
-+}
++#define PAGE_AGP PAGE_KERNEL_NOCACHE
++#define HAVE_PAGE_AGP 1
+
-+void memset_io(volatile void __iomem *a, int b, size_t c);
++/* fs/proc/kcore.c */
++#define kc_vaddr_to_offset(v) ((v) & __VIRTUAL_MASK)
++#define kc_offset_to_vaddr(o) \
++ (((o) & (1UL << (__VIRTUAL_MASK_SHIFT-1))) ? ((o) | (~__VIRTUAL_MASK)) : (o))
+
-+/*
-+ * ISA space is 'always mapped' on a typical x86 system, no need to
-+ * explicitly ioremap() it. The fact that the ISA IO space is mapped
-+ * to PAGE_OFFSET is pure coincidence - it does not mean ISA values
-+ * are physical addresses. The following constant pointer can be
-+ * used as the IO-area pointer (it can be iounmapped as well, so the
-+ * analogy with PCI is quite large):
-+ */
-+#define __ISA_IO_base ((char __iomem *)(fix_to_virt(FIX_ISAMAP_BEGIN)))
++#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
++#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_DIRTY
++#define __HAVE_ARCH_PTEP_GET_AND_CLEAR
++#define __HAVE_ARCH_PTEP_GET_AND_CLEAR_FULL
++#define __HAVE_ARCH_PTEP_CLEAR_FLUSH
++#define __HAVE_ARCH_PTEP_SET_WRPROTECT
++#define __HAVE_ARCH_PTE_SAME
++#include <asm-generic/pgtable.h>
+
++#endif /* _X86_64_PGTABLE_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/processor.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/processor.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,502 @@
+/*
-+ * Again, x86-64 does not require mem IO specific function.
-+ */
-+
-+#define eth_io_copy_and_sum(a,b,c,d) eth_copy_and_sum((a),(void *)(b),(c),(d))
-+
-+/**
-+ * check_signature - find BIOS signatures
-+ * @io_addr: mmio address to check
-+ * @signature: signature block
-+ * @length: length of signature
++ * include/asm-x86_64/processor.h
+ *
-+ * Perform a signature comparison with the mmio address io_addr. This
-+ * address should have been obtained by ioremap.
-+ * Returns 1 on a match.
++ * Copyright (C) 1994 Linus Torvalds
+ */
-+
-+static inline int check_signature(void __iomem *io_addr,
-+ const unsigned char *signature, int length)
-+{
-+ int retval = 0;
-+ do {
-+ if (readb(io_addr) != *signature)
-+ goto out;
-+ io_addr++;
-+ signature++;
-+ length--;
-+ } while (length);
-+ retval = 1;
-+out:
-+ return retval;
-+}
+
-+/* Nothing to do */
++#ifndef __ASM_X86_64_PROCESSOR_H
++#define __ASM_X86_64_PROCESSOR_H
+
-+#define dma_cache_inv(_start,_size) do { } while (0)
-+#define dma_cache_wback(_start,_size) do { } while (0)
-+#define dma_cache_wback_inv(_start,_size) do { } while (0)
++#include <asm/segment.h>
++#include <asm/page.h>
++#include <asm/types.h>
++#include <asm/sigcontext.h>
++#include <asm/cpufeature.h>
++#include <linux/threads.h>
++#include <asm/msr.h>
++#include <asm/current.h>
++#include <asm/system.h>
++#include <asm/mmsegment.h>
++#include <asm/percpu.h>
++#include <linux/personality.h>
++#include <linux/cpumask.h>
+
-+#define flush_write_buffers()
++#define TF_MASK 0x00000100
++#define IF_MASK 0x00000200
++#define IOPL_MASK 0x00003000
++#define NT_MASK 0x00004000
++#define VM_MASK 0x00020000
++#define AC_MASK 0x00040000
++#define VIF_MASK 0x00080000 /* virtual interrupt flag */
++#define VIP_MASK 0x00100000 /* virtual interrupt pending */
++#define ID_MASK 0x00200000
+
-+extern int iommu_bio_merge;
-+#define BIO_VMERGE_BOUNDARY iommu_bio_merge
++#define desc_empty(desc) \
++ (!((desc)->a | (desc)->b))
++
++#define desc_equal(desc1, desc2) \
++ (((desc1)->a == (desc2)->a) && ((desc1)->b == (desc2)->b))
+
+/*
-+ * Convert a physical pointer to a virtual kernel pointer for /dev/mem
-+ * access
++ * Default implementation of macro that returns current
++ * instruction pointer ("program counter").
+ */
-+#define xlate_dev_mem_ptr(p, sz) ioremap(p, sz)
-+#define xlate_dev_mem_ptr_unmap(p) iounmap(p)
++#define current_text_addr() ({ void *pc; asm volatile("leaq 1f(%%rip),%0\n1:":"=r"(pc)); pc; })
+
+/*
-+ * Convert a virtual cached pointer to an uncached pointer
++ * CPU type and hardware bug flags. Kept separately for each CPU.
+ */
-+#define xlate_dev_kmem_ptr(p) p
+
-+#endif /* __KERNEL__ */
++struct cpuinfo_x86 {
++ __u8 x86; /* CPU family */
++ __u8 x86_vendor; /* CPU vendor */
++ __u8 x86_model;
++ __u8 x86_mask;
++ int cpuid_level; /* Maximum supported CPUID level, -1=no CPUID */
++ __u32 x86_capability[NCAPINTS];
++ char x86_vendor_id[16];
++ char x86_model_id[64];
++ int x86_cache_size; /* in KB */
++ int x86_clflush_size;
++ int x86_cache_alignment;
++ int x86_tlbsize; /* number of 4K pages in DTLB/ITLB combined(in pages)*/
++ __u8 x86_virt_bits, x86_phys_bits;
++ __u8 x86_max_cores; /* cpuid returned max cores value */
++ __u32 x86_power;
++ __u32 extended_cpuid_level; /* Max extended CPUID function supported */
++ unsigned long loops_per_jiffy;
++#ifdef CONFIG_SMP
++ cpumask_t llc_shared_map; /* cpus sharing the last level cache */
++#endif
++ __u8 apicid;
++#ifdef CONFIG_SMP
++ __u8 booted_cores; /* number of cores as seen by OS */
++ __u8 phys_proc_id; /* Physical Processor id. */
++ __u8 cpu_core_id; /* Core id. */
++#endif
++} ____cacheline_aligned;
+
-+#define ARCH_HAS_DEV_MEM
++#define X86_VENDOR_INTEL 0
++#define X86_VENDOR_CYRIX 1
++#define X86_VENDOR_AMD 2
++#define X86_VENDOR_UMC 3
++#define X86_VENDOR_NEXGEN 4
++#define X86_VENDOR_CENTAUR 5
++#define X86_VENDOR_RISE 6
++#define X86_VENDOR_TRANSMETA 7
++#define X86_VENDOR_NUM 8
++#define X86_VENDOR_UNKNOWN 0xff
+
++#ifdef CONFIG_SMP
++extern struct cpuinfo_x86 cpu_data[];
++#define current_cpu_data cpu_data[smp_processor_id()]
++#else
++#define cpu_data (&boot_cpu_data)
++#define current_cpu_data boot_cpu_data
+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/irqflags.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/irqflags.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/irqflags.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/irqflags.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,139 @@
-+/*
-+ * include/asm-x86_64/irqflags.h
-+ *
-+ * IRQ flags handling
-+ *
-+ * This file gets included from lowlevel asm headers too, to provide
-+ * wrapped versions of the local_irq_*() APIs, based on the
-+ * raw_local_irq_*() functions from the lowlevel headers.
-+ */
-+#ifndef _ASM_IRQFLAGS_H
-+#define _ASM_IRQFLAGS_H
+
-+#ifndef __ASSEMBLY__
++extern char ignore_irq13;
++
++extern void identify_cpu(struct cpuinfo_x86 *);
++extern void print_cpu_info(struct cpuinfo_x86 *);
++extern unsigned int init_intel_cacheinfo(struct cpuinfo_x86 *c);
++extern unsigned short num_cache_leaves;
++
+/*
-+ * Interrupt control:
++ * EFLAGS bits
+ */
++#define X86_EFLAGS_CF 0x00000001 /* Carry Flag */
++#define X86_EFLAGS_PF 0x00000004 /* Parity Flag */
++#define X86_EFLAGS_AF 0x00000010 /* Auxillary carry Flag */
++#define X86_EFLAGS_ZF 0x00000040 /* Zero Flag */
++#define X86_EFLAGS_SF 0x00000080 /* Sign Flag */
++#define X86_EFLAGS_TF 0x00000100 /* Trap Flag */
++#define X86_EFLAGS_IF 0x00000200 /* Interrupt Flag */
++#define X86_EFLAGS_DF 0x00000400 /* Direction Flag */
++#define X86_EFLAGS_OF 0x00000800 /* Overflow Flag */
++#define X86_EFLAGS_IOPL 0x00003000 /* IOPL mask */
++#define X86_EFLAGS_NT 0x00004000 /* Nested Task */
++#define X86_EFLAGS_RF 0x00010000 /* Resume Flag */
++#define X86_EFLAGS_VM 0x00020000 /* Virtual Mode */
++#define X86_EFLAGS_AC 0x00040000 /* Alignment Check */
++#define X86_EFLAGS_VIF 0x00080000 /* Virtual Interrupt Flag */
++#define X86_EFLAGS_VIP 0x00100000 /* Virtual Interrupt Pending */
++#define X86_EFLAGS_ID 0x00200000 /* CPUID detection flag */
+
+/*
-+ * The use of 'barrier' in the following reflects their use as local-lock
-+ * operations. Reentrancy must be prevented (e.g., __cli()) /before/ following
-+ * critical operations are executed. All critical operations must complete
-+ * /before/ reentrancy is permitted (e.g., __sti()). Alpha architecture also
-+ * includes these barriers, for example.
++ * Intel CPU features in CR4
+ */
-+
-+#define __raw_local_save_flags() (current_vcpu_info()->evtchn_upcall_mask)
-+
-+#define raw_local_save_flags(flags) \
-+ do { (flags) = __raw_local_save_flags(); } while (0)
-+
-+#define raw_local_irq_restore(x) \
-+do { \
-+ vcpu_info_t *_vcpu; \
-+ barrier(); \
-+ _vcpu = current_vcpu_info(); \
-+ if ((_vcpu->evtchn_upcall_mask = (x)) == 0) { \
-+ barrier(); /* unmask then check (avoid races) */ \
-+ if ( unlikely(_vcpu->evtchn_upcall_pending) ) \
-+ force_evtchn_callback(); \
-+ } \
-+} while (0)
-+
-+#ifdef CONFIG_X86_VSMP
++#define X86_CR4_VME 0x0001 /* enable vm86 extensions */
++#define X86_CR4_PVI 0x0002 /* virtual interrupts flag enable */
++#define X86_CR4_TSD 0x0004 /* disable time stamp at ipl 3 */
++#define X86_CR4_DE 0x0008 /* enable debugging extensions */
++#define X86_CR4_PSE 0x0010 /* enable page size extensions */
++#define X86_CR4_PAE 0x0020 /* enable physical address extensions */
++#define X86_CR4_MCE 0x0040 /* Machine check enable */
++#define X86_CR4_PGE 0x0080 /* enable global pages */
++#define X86_CR4_PCE 0x0100 /* enable performance counters at ipl 3 */
++#define X86_CR4_OSFXSR 0x0200 /* enable fast FPU save and restore */
++#define X86_CR4_OSXMMEXCPT 0x0400 /* enable unmasked SSE exceptions */
+
+/*
-+ * Interrupt control for the VSMP architecture:
++ * Save the cr4 feature set we're using (ie
++ * Pentium 4MB enable and PPro Global page
++ * enable), so that any CPU's that boot up
++ * after us can get the correct flags.
+ */
++extern unsigned long mmu_cr4_features;
+
-+static inline void raw_local_irq_disable(void)
-+{
-+ unsigned long flags = __raw_local_save_flags();
-+
-+ raw_local_irq_restore((flags & ~(1 << 9)) | (1 << 18));
-+}
-+
-+static inline void raw_local_irq_enable(void)
-+{
-+ unsigned long flags = __raw_local_save_flags();
-+
-+ raw_local_irq_restore((flags | (1 << 9)) & ~(1 << 18));
-+}
-+
-+static inline int raw_irqs_disabled_flags(unsigned long flags)
++static inline void set_in_cr4 (unsigned long mask)
+{
-+ return !(flags & (1<<9)) || (flags & (1 << 18));
++ mmu_cr4_features |= mask;
++ __asm__("movq %%cr4,%%rax\n\t"
++ "orq %0,%%rax\n\t"
++ "movq %%rax,%%cr4\n"
++ : : "irg" (mask)
++ :"ax");
+}
+
-+#else /* CONFIG_X86_VSMP */
-+
-+#define raw_local_irq_disable() \
-+do { \
-+ current_vcpu_info()->evtchn_upcall_mask = 1; \
-+ barrier(); \
-+} while (0)
-+
-+#define raw_local_irq_enable() \
-+do { \
-+ vcpu_info_t *_vcpu; \
-+ barrier(); \
-+ _vcpu = current_vcpu_info(); \
-+ _vcpu->evtchn_upcall_mask = 0; \
-+ barrier(); /* unmask then check (avoid races) */ \
-+ if ( unlikely(_vcpu->evtchn_upcall_pending) ) \
-+ force_evtchn_callback(); \
-+} while (0)
-+
-+static inline int raw_irqs_disabled_flags(unsigned long flags)
++static inline void clear_in_cr4 (unsigned long mask)
+{
-+ return (flags != 0);
++ mmu_cr4_features &= ~mask;
++ __asm__("movq %%cr4,%%rax\n\t"
++ "andq %0,%%rax\n\t"
++ "movq %%rax,%%cr4\n"
++ : : "irg" (~mask)
++ :"ax");
+}
+
-+#endif
+
+/*
-+ * For spinlocks, etc.:
++ * User space process size. 47bits minus one guard page.
+ */
++#define TASK_SIZE64 (0x800000000000UL - 4096)
+
-+#define __raw_local_irq_save() \
-+({ \
-+ unsigned long flags = __raw_local_save_flags(); \
-+ \
-+ raw_local_irq_disable(); \
-+ \
-+ flags; \
-+})
++/* This decides where the kernel will search for a free chunk of vm
++ * space during mmap's.
++ */
++#define IA32_PAGE_OFFSET ((current->personality & ADDR_LIMIT_3GB) ? 0xc0000000 : 0xFFFFe000)
+
-+#define raw_local_irq_save(flags) \
-+ do { (flags) = __raw_local_irq_save(); } while (0)
++#define TASK_SIZE (test_thread_flag(TIF_IA32) ? IA32_PAGE_OFFSET : TASK_SIZE64)
++#define TASK_SIZE_OF(child) ((test_tsk_thread_flag(child, TIF_IA32)) ? IA32_PAGE_OFFSET : TASK_SIZE64)
+
-+#define raw_irqs_disabled() \
-+({ \
-+ unsigned long flags = __raw_local_save_flags(); \
-+ \
-+ raw_irqs_disabled_flags(flags); \
-+})
++#define TASK_UNMAPPED_BASE PAGE_ALIGN(TASK_SIZE/3)
+
+/*
-+ * Used in the idle loop; sti takes one instruction cycle
-+ * to complete:
++ * Size of io_bitmap.
+ */
-+void raw_safe_halt(void);
++#define IO_BITMAP_BITS 65536
++#define IO_BITMAP_BYTES (IO_BITMAP_BITS/8)
++#define IO_BITMAP_LONGS (IO_BITMAP_BYTES/sizeof(long))
++#ifndef CONFIG_X86_NO_TSS
++#define IO_BITMAP_OFFSET offsetof(struct tss_struct,io_bitmap)
++#endif
++#define INVALID_IO_BITMAP_OFFSET 0x8000
+
-+/*
-+ * Used when interrupts are already enabled or to
-+ * shutdown the processor:
-+ */
-+void halt(void);
++struct i387_fxsave_struct {
++ u16 cwd;
++ u16 swd;
++ u16 twd;
++ u16 fop;
++ u64 rip;
++ u64 rdp;
++ u32 mxcsr;
++ u32 mxcsr_mask;
++ u32 st_space[32]; /* 8*16 bytes for each FP-reg = 128 bytes */
++ u32 xmm_space[64]; /* 16*16 bytes for each XMM-reg = 128 bytes */
++ u32 padding[24];
++} __attribute__ ((aligned (16)));
+
-+#else /* __ASSEMBLY__: */
-+# ifdef CONFIG_TRACE_IRQFLAGS
-+# define TRACE_IRQS_ON call trace_hardirqs_on_thunk
-+# define TRACE_IRQS_OFF call trace_hardirqs_off_thunk
-+# else
-+# define TRACE_IRQS_ON
-+# define TRACE_IRQS_OFF
-+# endif
-+#endif
++union i387_union {
++ struct i387_fxsave_struct fxsave;
++};
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/irq.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/irq.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/irq.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/irq.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,38 @@
-+#ifndef _ASM_IRQ_H
-+#define _ASM_IRQ_H
++#ifndef CONFIG_X86_NO_TSS
++struct tss_struct {
++ u32 reserved1;
++ u64 rsp0;
++ u64 rsp1;
++ u64 rsp2;
++ u64 reserved2;
++ u64 ist[7];
++ u32 reserved3;
++ u32 reserved4;
++ u16 reserved5;
++ u16 io_bitmap_base;
++ /*
++ * The extra 1 is there because the CPU will access an
++ * additional byte beyond the end of the IO permission
++ * bitmap. The extra byte must be all 1 bits, and must
++ * be within the limit. Thus we have:
++ *
++ * 128 bytes, the bitmap itself, for ports 0..0x3ff
++ * 8 bytes, for an extra "long" of ~0UL
++ */
++ unsigned long io_bitmap[IO_BITMAP_LONGS + 1];
++} __attribute__((packed)) ____cacheline_aligned;
+
-+/*
-+ * linux/include/asm/irq.h
-+ *
-+ * (C) 1992, 1993 Linus Torvalds, (C) 1997 Ingo Molnar
-+ *
-+ * IRQ/IPI changes taken from work by Thomas Radke
-+ * <tomsoft at informatik.tu-chemnitz.de>
-+ */
++DECLARE_PER_CPU(struct tss_struct,init_tss);
++#endif
+
-+#include <linux/sched.h>
-+/* include comes from machine specific directory */
-+#include "irq_vectors.h"
-+#include <asm/thread_info.h>
+
-+static __inline__ int irq_canonicalize(int irq)
-+{
-+ return ((irq == 2) ? 9 : irq);
-+}
++extern struct cpuinfo_x86 boot_cpu_data;
++#ifndef CONFIG_X86_NO_TSS
++/* Save the original ist values for checking stack pointers during debugging */
++struct orig_ist {
++ unsigned long ist[7];
++};
++DECLARE_PER_CPU(struct orig_ist, orig_ist);
++#endif
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+#define ARCH_HAS_NMI_WATCHDOG /* See include/linux/nmi.h */
++#ifdef CONFIG_X86_VSMP
++#define ARCH_MIN_TASKALIGN (1 << INTERNODE_CACHE_SHIFT)
++#define ARCH_MIN_MMSTRUCT_ALIGN (1 << INTERNODE_CACHE_SHIFT)
++#else
++#define ARCH_MIN_TASKALIGN 16
++#define ARCH_MIN_MMSTRUCT_ALIGN 0
+#endif
+
-+#define KDB_VECTOR 0xf9
++struct thread_struct {
++ unsigned long rsp0;
++ unsigned long rsp;
++ unsigned long userrsp; /* Copy from PDA */
++ unsigned long fs;
++ unsigned long gs;
++ unsigned short es, ds, fsindex, gsindex;
++/* Hardware debugging registers */
++ unsigned long debugreg0;
++ unsigned long debugreg1;
++ unsigned long debugreg2;
++ unsigned long debugreg3;
++ unsigned long debugreg6;
++ unsigned long debugreg7;
++/* fault info */
++ unsigned long cr2, trap_no, error_code;
++/* floating point info */
++ union i387_union i387 __attribute__((aligned(16)));
++/* IO permissions. the bitmap could be moved into the GDT, that would make
++ switch faster for a limited number of ioperm using tasks. -AK */
++ int ioperm;
++ unsigned long *io_bitmap_ptr;
++ unsigned io_bitmap_max;
++/* cached TLS descriptors. */
++ u64 tls_array[GDT_ENTRY_TLS_ENTRIES];
++ unsigned int iopl;
++} __attribute__((aligned(16)));
+
-+# define irq_ctx_init(cpu) do { } while (0)
++#define INIT_THREAD { \
++ .rsp0 = (unsigned long)&init_stack + sizeof(init_stack) \
++}
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+#include <linux/cpumask.h>
-+extern void fixup_irqs(cpumask_t map);
++#ifndef CONFIG_X86_NO_TSS
++#define INIT_TSS { \
++ .rsp0 = (unsigned long)&init_stack + sizeof(init_stack) \
++}
+#endif
+
-+#define __ARCH_HAS_DO_SOFTIRQ 1
-+
-+#endif /* _ASM_IRQ_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/maddr.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/maddr.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/maddr.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/maddr.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,161 @@
-+#ifndef _X86_64_MADDR_H
-+#define _X86_64_MADDR_H
++#define INIT_MMAP \
++{ &init_mm, 0, 0, NULL, PAGE_SHARED, VM_READ | VM_WRITE | VM_EXEC, 1, NULL, NULL }
+
-+#include <xen/features.h>
-+#include <xen/interface/xen.h>
++#define start_thread(regs,new_rip,new_rsp) do { \
++ asm volatile("movl %0,%%fs; movl %0,%%es; movl %0,%%ds": :"r" (0)); \
++ load_gs_index(0); \
++ (regs)->rip = (new_rip); \
++ (regs)->rsp = (new_rsp); \
++ write_pda(oldrsp, (new_rsp)); \
++ (regs)->cs = __USER_CS; \
++ (regs)->ss = __USER_DS; \
++ (regs)->eflags = 0x200; \
++ set_fs(USER_DS); \
++} while(0)
+
-+/**** MACHINE <-> PHYSICAL CONVERSION MACROS ****/
-+#define INVALID_P2M_ENTRY (~0UL)
-+#define FOREIGN_FRAME_BIT (1UL<<63)
-+#define FOREIGN_FRAME(m) ((m) | FOREIGN_FRAME_BIT)
++#define get_debugreg(var, register) \
++ var = HYPERVISOR_get_debugreg(register)
++#define set_debugreg(value, register) do { \
++ if (HYPERVISOR_set_debugreg(register, value)) \
++ BUG(); \
++} while (0)
+
-+/* Definitions for machine and pseudophysical addresses. */
-+typedef unsigned long paddr_t;
-+typedef unsigned long maddr_t;
++struct task_struct;
++struct mm_struct;
+
-+#ifdef CONFIG_XEN
++/* Free all resources held by a thread. */
++extern void release_thread(struct task_struct *);
+
-+extern unsigned long *phys_to_machine_mapping;
++/* Prepare to copy thread state - unlazy all lazy status */
++extern void prepare_to_copy(struct task_struct *tsk);
+
-+#undef machine_to_phys_mapping
-+extern unsigned long *machine_to_phys_mapping;
-+extern unsigned int machine_to_phys_order;
++/*
++ * create a kernel thread without removing it from tasklists
++ */
++extern long kernel_thread(int (*fn)(void *), void * arg, unsigned long flags);
+
-+static inline unsigned long pfn_to_mfn(unsigned long pfn)
-+{
-+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ return pfn;
-+ BUG_ON(end_pfn && pfn >= end_pfn);
-+ return phys_to_machine_mapping[pfn] & ~FOREIGN_FRAME_BIT;
-+}
++/*
++ * Return saved PC of a blocked thread.
++ * What is this good for? it will be always the scheduler or ret_from_fork.
++ */
++#define thread_saved_pc(t) (*(unsigned long *)((t)->thread.rsp - 8))
+
-+static inline int phys_to_machine_mapping_valid(unsigned long pfn)
-+{
-+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ return 1;
-+ BUG_ON(end_pfn && pfn >= end_pfn);
-+ return (phys_to_machine_mapping[pfn] != INVALID_P2M_ENTRY);
-+}
++extern unsigned long get_wchan(struct task_struct *p);
++#define task_pt_regs(tsk) ((struct pt_regs *)(tsk)->thread.rsp0 - 1)
++#define KSTK_EIP(tsk) (task_pt_regs(tsk)->rip)
++#define KSTK_ESP(tsk) -1 /* sorry. doesn't work for syscall. */
+
-+static inline unsigned long mfn_to_pfn(unsigned long mfn)
-+{
-+ unsigned long pfn;
+
-+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ return mfn;
++struct microcode_header {
++ unsigned int hdrver;
++ unsigned int rev;
++ unsigned int date;
++ unsigned int sig;
++ unsigned int cksum;
++ unsigned int ldrver;
++ unsigned int pf;
++ unsigned int datasize;
++ unsigned int totalsize;
++ unsigned int reserved[3];
++};
+
-+ if (unlikely((mfn >> machine_to_phys_order) != 0))
-+ return end_pfn;
++struct microcode {
++ struct microcode_header hdr;
++ unsigned int bits[0];
++};
+
-+ /* The array access can fail (e.g., device space beyond end of RAM). */
-+ asm (
-+ "1: movq %1,%0\n"
-+ "2:\n"
-+ ".section .fixup,\"ax\"\n"
-+ "3: movq %2,%0\n"
-+ " jmp 2b\n"
-+ ".previous\n"
-+ ".section __ex_table,\"a\"\n"
-+ " .align 8\n"
-+ " .quad 1b,3b\n"
-+ ".previous"
-+ : "=r" (pfn)
-+ : "m" (machine_to_phys_mapping[mfn]), "m" (end_pfn) );
++typedef struct microcode microcode_t;
++typedef struct microcode_header microcode_header_t;
+
-+ return pfn;
-+}
++/* microcode format is extended from prescott processors */
++struct extended_signature {
++ unsigned int sig;
++ unsigned int pf;
++ unsigned int cksum;
++};
+
-+/*
-+ * We detect special mappings in one of two ways:
-+ * 1. If the MFN is an I/O page then Xen will set the m2p entry
-+ * to be outside our maximum possible pseudophys range.
-+ * 2. If the MFN belongs to a different domain then we will certainly
-+ * not have MFN in our p2m table. Conversely, if the page is ours,
-+ * then we'll have p2m(m2p(MFN))==MFN.
-+ * If we detect a special mapping then it doesn't have a 'struct page'.
-+ * We force !pfn_valid() by returning an out-of-range pointer.
-+ *
-+ * NB. These checks require that, for any MFN that is not in our reservation,
-+ * there is no PFN such that p2m(PFN) == MFN. Otherwise we can get confused if
-+ * we are foreign-mapping the MFN, and the other domain as m2p(MFN) == PFN.
-+ * Yikes! Various places must poke in INVALID_P2M_ENTRY for safety.
-+ *
-+ * NB2. When deliberately mapping foreign pages into the p2m table, you *must*
-+ * use FOREIGN_FRAME(). This will cause pte_pfn() to choke on it, as we
-+ * require. In all the cases we care about, the FOREIGN_FRAME bit is
-+ * masked (e.g., pfn_to_mfn()) so behaviour there is correct.
-+ */
-+static inline unsigned long mfn_to_local_pfn(unsigned long mfn)
-+{
-+ unsigned long pfn = mfn_to_pfn(mfn);
-+ if ((pfn < end_pfn)
-+ && !xen_feature(XENFEAT_auto_translated_physmap)
-+ && (phys_to_machine_mapping[pfn] != mfn))
-+ return end_pfn; /* force !pfn_valid() */
-+ return pfn;
-+}
++struct extended_sigtable {
++ unsigned int count;
++ unsigned int cksum;
++ unsigned int reserved[3];
++ struct extended_signature sigs[0];
++};
+
-+static inline void set_phys_to_machine(unsigned long pfn, unsigned long mfn)
-+{
-+ BUG_ON(end_pfn && pfn >= end_pfn);
-+ if (xen_feature(XENFEAT_auto_translated_physmap)) {
-+ BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY);
-+ return;
-+ }
-+ phys_to_machine_mapping[pfn] = mfn;
-+}
+
-+static inline maddr_t phys_to_machine(paddr_t phys)
-+{
-+ maddr_t machine = pfn_to_mfn(phys >> PAGE_SHIFT);
-+ machine = (machine << PAGE_SHIFT) | (phys & ~PAGE_MASK);
-+ return machine;
-+}
++#define ASM_NOP1 K8_NOP1
++#define ASM_NOP2 K8_NOP2
++#define ASM_NOP3 K8_NOP3
++#define ASM_NOP4 K8_NOP4
++#define ASM_NOP5 K8_NOP5
++#define ASM_NOP6 K8_NOP6
++#define ASM_NOP7 K8_NOP7
++#define ASM_NOP8 K8_NOP8
+
-+static inline paddr_t machine_to_phys(maddr_t machine)
-+{
-+ paddr_t phys = mfn_to_pfn(machine >> PAGE_SHIFT);
-+ phys = (phys << PAGE_SHIFT) | (machine & ~PAGE_MASK);
-+ return phys;
-+}
++/* Opteron nops */
++#define K8_NOP1 ".byte 0x90\n"
++#define K8_NOP2 ".byte 0x66,0x90\n"
++#define K8_NOP3 ".byte 0x66,0x66,0x90\n"
++#define K8_NOP4 ".byte 0x66,0x66,0x66,0x90\n"
++#define K8_NOP5 K8_NOP3 K8_NOP2
++#define K8_NOP6 K8_NOP3 K8_NOP3
++#define K8_NOP7 K8_NOP4 K8_NOP3
++#define K8_NOP8 K8_NOP4 K8_NOP4
+
-+static inline paddr_t pte_phys_to_machine(paddr_t phys)
-+{
-+ maddr_t machine;
-+ machine = pfn_to_mfn((phys & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT);
-+ machine = (machine << PAGE_SHIFT) | (phys & ~PHYSICAL_PAGE_MASK);
-+ return machine;
-+}
++#define ASM_NOP_MAX 8
+
-+static inline paddr_t pte_machine_to_phys(maddr_t machine)
++/* REP NOP (PAUSE) is a good thing to insert into busy-wait loops. */
++static inline void rep_nop(void)
+{
-+ paddr_t phys;
-+ phys = mfn_to_pfn((machine & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT);
-+ phys = (phys << PAGE_SHIFT) | (machine & ~PHYSICAL_PAGE_MASK);
-+ return phys;
++ __asm__ __volatile__("rep;nop": : :"memory");
+}
+
-+#define __pte_ma(x) ((pte_t) { (x) } )
-+#define pfn_pte_ma(pfn, prot) __pte_ma((((pfn) << PAGE_SHIFT) | pgprot_val(prot)) & __supported_pte_mask)
++/* Stop speculative execution */
++static inline void sync_core(void)
++{
++ int tmp;
++ asm volatile("cpuid" : "=a" (tmp) : "0" (1) : "ebx","ecx","edx","memory");
++}
+
-+#else /* !CONFIG_XEN */
++#define cpu_has_fpu 1
+
-+#define pfn_to_mfn(pfn) (pfn)
-+#define mfn_to_pfn(mfn) (mfn)
-+#define mfn_to_local_pfn(mfn) (mfn)
-+#define set_phys_to_machine(pfn, mfn) ((void)0)
-+#define phys_to_machine_mapping_valid(pfn) (1)
-+#define phys_to_machine(phys) ((maddr_t)(phys))
-+#define machine_to_phys(mach) ((paddr_t)(mach))
-+#define pfn_pte_ma(pfn, prot) pfn_pte(pfn, prot)
-+#define __pte_ma(x) __pte(x)
++#define ARCH_HAS_PREFETCH
++static inline void prefetch(void *x)
++{
++ asm volatile("prefetcht0 %0" :: "m" (*(unsigned long *)x));
++}
+
-+#endif /* !CONFIG_XEN */
++#define ARCH_HAS_PREFETCHW 1
++static inline void prefetchw(void *x)
++{
++ alternative_input("prefetcht0 (%1)",
++ "prefetchw (%1)",
++ X86_FEATURE_3DNOW,
++ "r" (x));
++}
+
-+/* VIRT <-> MACHINE conversion */
-+#define virt_to_machine(v) (phys_to_machine(__pa(v)))
-+#define virt_to_mfn(v) (pfn_to_mfn(__pa(v) >> PAGE_SHIFT))
-+#define mfn_to_virt(m) (__va(mfn_to_pfn(m) << PAGE_SHIFT))
++#define ARCH_HAS_SPINLOCK_PREFETCH 1
+
-+#endif /* _X86_64_MADDR_H */
++#define spin_lock_prefetch(x) prefetchw(x)
+
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/mmu_context.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/mmu_context.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/mmu_context.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/mmu_context.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,136 @@
-+#ifndef __X86_64_MMU_CONTEXT_H
-+#define __X86_64_MMU_CONTEXT_H
++#define cpu_relax() rep_nop()
+
-+#include <asm/desc.h>
-+#include <asm/atomic.h>
-+#include <asm/pgalloc.h>
-+#include <asm/page.h>
-+#include <asm/pda.h>
-+#include <asm/pgtable.h>
-+#include <asm/tlbflush.h>
++/*
++ * NSC/Cyrix CPU configuration register indexes
++ */
++#define CX86_CCR0 0xc0
++#define CX86_CCR1 0xc1
++#define CX86_CCR2 0xc2
++#define CX86_CCR3 0xc3
++#define CX86_CCR4 0xe8
++#define CX86_CCR5 0xe9
++#define CX86_CCR6 0xea
++#define CX86_CCR7 0xeb
++#define CX86_DIR0 0xfe
++#define CX86_DIR1 0xff
++#define CX86_ARR_BASE 0xc4
++#define CX86_RCR_BASE 0xdc
+
+/*
-+ * possibly do the LDT unload here?
++ * NSC/Cyrix CPU indexed register access macros
+ */
-+int init_new_context(struct task_struct *tsk, struct mm_struct *mm);
-+void destroy_context(struct mm_struct *mm);
+
-+static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk)
-+{
-+#if defined(CONFIG_SMP) && !defined(CONFIG_XEN)
-+ if (read_pda(mmu_state) == TLBSTATE_OK)
-+ write_pda(mmu_state, TLBSTATE_LAZY);
-+#endif
-+}
++#define getCx86(reg) ({ outb((reg), 0x22); inb(0x23); })
+
-+#define prepare_arch_switch(next) __prepare_arch_switch()
++#define setCx86(reg, data) do { \
++ outb((reg), 0x22); \
++ outb((data), 0x23); \
++} while (0)
+
-+static inline void __prepare_arch_switch(void)
++static inline void serialize_cpu(void)
+{
-+ /*
-+ * Save away %es, %ds, %fs and %gs. Must happen before reload
-+ * of cr3/ldt (i.e., not in __switch_to).
-+ */
-+ __asm__ __volatile__ (
-+ "mov %%es,%0 ; mov %%ds,%1 ; mov %%fs,%2 ; mov %%gs,%3"
-+ : "=m" (current->thread.es),
-+ "=m" (current->thread.ds),
-+ "=m" (current->thread.fsindex),
-+ "=m" (current->thread.gsindex) );
-+
-+ if (current->thread.ds)
-+ __asm__ __volatile__ ( "movl %0,%%ds" : : "r" (0) );
-+
-+ if (current->thread.es)
-+ __asm__ __volatile__ ( "movl %0,%%es" : : "r" (0) );
-+
-+ if (current->thread.fsindex) {
-+ __asm__ __volatile__ ( "movl %0,%%fs" : : "r" (0) );
-+ current->thread.fs = 0;
-+ }
-+
-+ if (current->thread.gsindex) {
-+ load_gs_index(0);
-+ current->thread.gs = 0;
-+ }
++ __asm__ __volatile__ ("cpuid" : : : "ax", "bx", "cx", "dx");
+}
+
-+extern void mm_pin(struct mm_struct *mm);
-+extern void mm_unpin(struct mm_struct *mm);
-+void mm_pin_all(void);
-+
-+static inline void load_cr3(pgd_t *pgd)
++static inline void __monitor(const void *eax, unsigned long ecx,
++ unsigned long edx)
+{
-+ asm volatile("movq %0,%%cr3" :: "r" (phys_to_machine(__pa(pgd))) :
-+ "memory");
++ /* "monitor %eax,%ecx,%edx;" */
++ asm volatile(
++ ".byte 0x0f,0x01,0xc8;"
++ : :"a" (eax), "c" (ecx), "d"(edx));
+}
+
-+static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
-+ struct task_struct *tsk)
++static inline void __mwait(unsigned long eax, unsigned long ecx)
+{
-+ unsigned cpu = smp_processor_id();
-+ struct mmuext_op _op[3], *op = _op;
++ /* "mwait %eax,%ecx;" */
++ asm volatile(
++ ".byte 0x0f,0x01,0xc9;"
++ : :"a" (eax), "c" (ecx));
++}
+
-+ if (likely(prev != next)) {
-+ BUG_ON(!xen_feature(XENFEAT_writable_page_tables) &&
-+ !next->context.pinned);
++#define stack_current() \
++({ \
++ struct thread_info *ti; \
++ asm("andq %%rsp,%0; ":"=r" (ti) : "0" (CURRENT_MASK)); \
++ ti->task; \
++})
+
-+ /* stop flush ipis for the previous mm */
-+ cpu_clear(cpu, prev->cpu_vm_mask);
-+#if defined(CONFIG_SMP) && !defined(CONFIG_XEN)
-+ write_pda(mmu_state, TLBSTATE_OK);
-+ write_pda(active_mm, next);
-+#endif
-+ cpu_set(cpu, next->cpu_vm_mask);
++#define cache_line_size() (boot_cpu_data.x86_cache_alignment)
+
-+ /* load_cr3(next->pgd) */
-+ op->cmd = MMUEXT_NEW_BASEPTR;
-+ op->arg1.mfn = pfn_to_mfn(__pa(next->pgd) >> PAGE_SHIFT);
-+ op++;
++extern unsigned long boot_option_idle_override;
++/* Boot loader type from the setup header */
++extern int bootloader_type;
+
-+ /* xen_new_user_pt(__pa(__user_pgd(next->pgd))) */
-+ op->cmd = MMUEXT_NEW_USER_BASEPTR;
-+ op->arg1.mfn = pfn_to_mfn(__pa(__user_pgd(next->pgd)) >> PAGE_SHIFT);
-+ op++;
-+
-+ if (unlikely(next->context.ldt != prev->context.ldt)) {
-+ /* load_LDT_nolock(&next->context, cpu) */
-+ op->cmd = MMUEXT_SET_LDT;
-+ op->arg1.linear_addr = (unsigned long)next->context.ldt;
-+ op->arg2.nr_ents = next->context.size;
-+ op++;
-+ }
++#define HAVE_ARCH_PICK_MMAP_LAYOUT 1
+
-+ BUG_ON(HYPERVISOR_mmuext_op(_op, op-_op, NULL, DOMID_SELF));
-+ }
-+#if defined(CONFIG_SMP) && !defined(CONFIG_XEN)
-+ else {
-+ write_pda(mmu_state, TLBSTATE_OK);
-+ if (read_pda(active_mm) != next)
-+ out_of_line_bug();
-+ if (!cpu_test_and_set(cpu, next->cpu_vm_mask)) {
-+ /* We were in lazy tlb mode and leave_mm disabled
-+ * tlb flush IPI delivery. We must reload CR3
-+ * to make sure to use no freed page tables.
-+ */
-+ load_cr3(next->pgd);
-+ xen_new_user_pt(__pa(__user_pgd(next->pgd)));
-+ load_LDT_nolock(&next->context, cpu);
-+ }
-+ }
-+#endif
-+}
++#endif /* __ASM_X86_64_PROCESSOR_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/ptrace.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/ptrace.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,127 @@
++#ifndef _X86_64_PTRACE_H
++#define _X86_64_PTRACE_H
+
-+#define deactivate_mm(tsk,mm) do { \
-+ load_gs_index(0); \
-+ asm volatile("movl %0,%%fs"::"r"(0)); \
-+} while(0)
++#if defined(__ASSEMBLY__) || defined(__FRAME_OFFSETS)
++#define R15 0
++#define R14 8
++#define R13 16
++#define R12 24
++#define RBP 32
++#define RBX 40
++/* arguments: interrupts/non tracing syscalls only save upto here*/
++#define R11 48
++#define R10 56
++#define R9 64
++#define R8 72
++#define RAX 80
++#define RCX 88
++#define RDX 96
++#define RSI 104
++#define RDI 112
++#define ORIG_RAX 120 /* = ERROR */
++/* end of arguments */
++/* cpu exception frame or undefined in case of fast syscall. */
++#define RIP 128
++#define CS 136
++#define EFLAGS 144
++#define RSP 152
++#define SS 160
++#define ARGOFFSET R11
++#endif /* __ASSEMBLY__ */
+
-+static inline void activate_mm(struct mm_struct *prev, struct mm_struct *next)
-+{
-+ if (!next->context.pinned)
-+ mm_pin(next);
-+ switch_mm(prev, next, NULL);
-+}
++/* top of stack page */
++#define FRAME_SIZE 168
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/mmu.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/mmu.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/mmu.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/mmu.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,38 @@
-+#ifndef __x86_64_MMU_H
-+#define __x86_64_MMU_H
++#define PTRACE_OLDSETOPTIONS 21
+
-+#include <linux/spinlock.h>
-+#include <asm/semaphore.h>
++#ifndef __ASSEMBLY__
++
++struct pt_regs {
++ unsigned long r15;
++ unsigned long r14;
++ unsigned long r13;
++ unsigned long r12;
++ unsigned long rbp;
++ unsigned long rbx;
++/* arguments: non interrupts/non tracing syscalls only save upto here*/
++ unsigned long r11;
++ unsigned long r10;
++ unsigned long r9;
++ unsigned long r8;
++ unsigned long rax;
++ unsigned long rcx;
++ unsigned long rdx;
++ unsigned long rsi;
++ unsigned long rdi;
++ unsigned long orig_rax;
++/* end of arguments */
++/* cpu exception frame or undefined */
++ unsigned long rip;
++ unsigned long cs;
++ unsigned long eflags;
++ unsigned long rsp;
++ unsigned long ss;
++/* top of stack page */
++};
+
-+/*
-+ * The x86_64 doesn't have a mmu context, but
-+ * we put the segment information here.
-+ *
-+ * cpu_vm_mask is used to optimize ldt flushing.
-+ */
-+typedef struct {
-+ void *ldt;
-+ rwlock_t ldtlock;
-+ int size;
-+ struct semaphore sem;
-+#ifdef CONFIG_XEN
-+ unsigned pinned:1;
-+ unsigned has_foreign_mappings:1;
-+ struct list_head unpinned;
+#endif
-+} mm_context_t;
+
-+#ifdef CONFIG_XEN
-+extern struct list_head mm_unpinned;
-+extern spinlock_t mm_unpinned_lock;
++/* Arbitrarily choose the same ptrace numbers as used by the Sparc code. */
++#define PTRACE_GETREGS 12
++#define PTRACE_SETREGS 13
++#define PTRACE_GETFPREGS 14
++#define PTRACE_SETFPREGS 15
++#define PTRACE_GETFPXREGS 18
++#define PTRACE_SETFPXREGS 19
+
-+/* mm/memory.c:exit_mmap hook */
-+extern void _arch_exit_mmap(struct mm_struct *mm);
-+#define arch_exit_mmap(_mm) _arch_exit_mmap(_mm)
++/* only useful for access 32bit programs */
++#define PTRACE_GET_THREAD_AREA 25
++#define PTRACE_SET_THREAD_AREA 26
+
-+/* kernel/fork.c:dup_mmap hook */
-+extern void _arch_dup_mmap(struct mm_struct *mm);
-+#define arch_dup_mmap(mm, oldmm) ((void)(oldmm), _arch_dup_mmap(mm))
-+#endif
++#define PTRACE_ARCH_PRCTL 30 /* arch_prctl for child */
+
++#if defined(__KERNEL__) && !defined(__ASSEMBLY__)
++#define user_mode(regs) (!!((regs)->cs & 3))
++#define user_mode_vm(regs) user_mode(regs)
++#define instruction_pointer(regs) ((regs)->rip)
++#if defined(CONFIG_SMP) && defined(CONFIG_FRAME_POINTER)
++extern unsigned long profile_pc(struct pt_regs *regs);
++#else
++#define profile_pc(regs) instruction_pointer(regs)
+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/msr.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/msr.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/msr.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/msr.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,399 @@
-+#ifndef X86_64_MSR_H
-+#define X86_64_MSR_H 1
+
-+#ifndef __ASSEMBLY__
-+/*
-+ * Access to machine-specific registers (available on 586 and better only)
-+ * Note: the rd* operations modify the parameters directly (without using
-+ * pointer indirection), this allows gcc to optimize better
-+ */
++#include <linux/compiler.h>
+
-+#define rdmsr(msr,val1,val2) \
-+ __asm__ __volatile__("rdmsr" \
-+ : "=a" (val1), "=d" (val2) \
-+ : "c" (msr))
++void signal_fault(struct pt_regs *regs, void __user *frame, char *where);
+
++struct task_struct;
+
-+#define rdmsrl(msr,val) do { unsigned long a__,b__; \
-+ __asm__ __volatile__("rdmsr" \
-+ : "=a" (a__), "=d" (b__) \
-+ : "c" (msr)); \
-+ val = a__ | (b__<<32); \
-+} while(0)
++extern unsigned long
++convert_rip_to_linear(struct task_struct *child, struct pt_regs *regs);
+
-+#define wrmsr(msr,val1,val2) \
-+ __asm__ __volatile__("wrmsr" \
-+ : /* no outputs */ \
-+ : "c" (msr), "a" (val1), "d" (val2))
++enum {
++ EF_CF = 0x00000001,
++ EF_PF = 0x00000004,
++ EF_AF = 0x00000010,
++ EF_ZF = 0x00000040,
++ EF_SF = 0x00000080,
++ EF_TF = 0x00000100,
++ EF_IE = 0x00000200,
++ EF_DF = 0x00000400,
++ EF_OF = 0x00000800,
++ EF_IOPL = 0x00003000,
++ EF_IOPL_RING0 = 0x00000000,
++ EF_IOPL_RING1 = 0x00001000,
++ EF_IOPL_RING2 = 0x00002000,
++ EF_NT = 0x00004000, /* nested task */
++ EF_RF = 0x00010000, /* resume */
++ EF_VM = 0x00020000, /* virtual mode */
++ EF_AC = 0x00040000, /* alignment */
++ EF_VIF = 0x00080000, /* virtual interrupt */
++ EF_VIP = 0x00100000, /* virtual interrupt pending */
++ EF_ID = 0x00200000, /* id */
++};
+
-+#define wrmsrl(msr,val) wrmsr(msr,(__u32)((__u64)(val)),((__u64)(val))>>32)
++#endif
+
-+/* wrmsr with exception handling */
-+#define wrmsr_safe(msr,a,b) ({ int ret__; \
-+ asm volatile("2: wrmsr ; xorl %0,%0\n" \
-+ "1:\n\t" \
-+ ".section .fixup,\"ax\"\n\t" \
-+ "3: movl %4,%0 ; jmp 1b\n\t" \
-+ ".previous\n\t" \
-+ ".section __ex_table,\"a\"\n" \
-+ " .align 8\n\t" \
-+ " .quad 2b,3b\n\t" \
-+ ".previous" \
-+ : "=a" (ret__) \
-+ : "c" (msr), "0" (a), "d" (b), "i" (-EFAULT)); \
-+ ret__; })
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/smp.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/smp.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,150 @@
++#ifndef __ASM_SMP_H
++#define __ASM_SMP_H
+
-+#define checking_wrmsrl(msr,val) wrmsr_safe(msr,(u32)(val),(u32)((val)>>32))
++/*
++ * We need the APIC definitions automatically as part of 'smp.h'
++ */
++#ifndef __ASSEMBLY__
++#include <linux/threads.h>
++#include <linux/cpumask.h>
++#include <linux/bitops.h>
++extern int disable_apic;
++#endif
+
-+#define rdmsr_safe(msr,a,b) \
-+ ({ int ret__; \
-+ asm volatile ("1: rdmsr\n" \
-+ "2:\n" \
-+ ".section .fixup,\"ax\"\n" \
-+ "3: movl %4,%0\n" \
-+ " jmp 2b\n" \
-+ ".previous\n" \
-+ ".section __ex_table,\"a\"\n" \
-+ " .align 8\n" \
-+ " .quad 1b,3b\n" \
-+ ".previous":"=&bDS" (ret__), "=a"(*(a)), "=d"(*(b))\
-+ :"c"(msr), "i"(-EIO), "0"(0)); \
-+ ret__; })
++#ifdef CONFIG_X86_LOCAL_APIC
++#ifndef __ASSEMBLY__
++#include <asm/fixmap.h>
++#include <asm/mpspec.h>
++#ifdef CONFIG_X86_IO_APIC
++#include <asm/io_apic.h>
++#endif
++#include <asm/apic.h>
++#include <asm/thread_info.h>
++#endif
++#endif
+
-+#define rdtsc(low,high) \
-+ __asm__ __volatile__("rdtsc" : "=a" (low), "=d" (high))
++#ifdef CONFIG_SMP
++#ifndef ASSEMBLY
+
-+#define rdtscl(low) \
-+ __asm__ __volatile__ ("rdtsc" : "=a" (low) : : "edx")
++#include <asm/pda.h>
+
-+#define rdtscll(val) do { \
-+ unsigned int __a,__d; \
-+ asm volatile("rdtsc" : "=a" (__a), "=d" (__d)); \
-+ (val) = ((unsigned long)__a) | (((unsigned long)__d)<<32); \
-+} while(0)
++struct pt_regs;
+
-+#define write_tsc(val1,val2) wrmsr(0x10, val1, val2)
++extern cpumask_t cpu_present_mask;
++extern cpumask_t cpu_possible_map;
++extern cpumask_t cpu_online_map;
++extern cpumask_t cpu_initialized;
+
-+#define rdpmc(counter,low,high) \
-+ __asm__ __volatile__("rdpmc" \
-+ : "=a" (low), "=d" (high) \
-+ : "c" (counter))
++/*
++ * Private routines/data
++ */
++
++extern void smp_alloc_memory(void);
++extern volatile unsigned long smp_invalidate_needed;
++extern int pic_mode;
++extern void lock_ipi_call_lock(void);
++extern void unlock_ipi_call_lock(void);
++extern int smp_num_siblings;
++extern void smp_send_reschedule(int cpu);
++void smp_stop_cpu(void);
++extern int smp_call_function_single(int cpuid, void (*func) (void *info),
++ void *info, int retry, int wait);
+
-+static inline void cpuid(int op, unsigned int *eax, unsigned int *ebx,
-+ unsigned int *ecx, unsigned int *edx)
-+{
-+ __asm__(XEN_CPUID
-+ : "=a" (*eax),
-+ "=b" (*ebx),
-+ "=c" (*ecx),
-+ "=d" (*edx)
-+ : "0" (op));
-+}
++extern cpumask_t cpu_sibling_map[NR_CPUS];
++extern cpumask_t cpu_core_map[NR_CPUS];
++extern u8 cpu_llc_id[NR_CPUS];
+
-+/* Some CPUID calls want 'count' to be placed in ecx */
-+static inline void cpuid_count(int op, int count, int *eax, int *ebx, int *ecx,
-+ int *edx)
-+{
-+ __asm__(XEN_CPUID
-+ : "=a" (*eax),
-+ "=b" (*ebx),
-+ "=c" (*ecx),
-+ "=d" (*edx)
-+ : "0" (op), "c" (count));
-+}
++#define SMP_TRAMPOLINE_BASE 0x6000
+
+/*
-+ * CPUID functions returning a single datum
++ * On x86 all CPUs are mapped 1:1 to the APIC space.
++ * This simplifies scheduling and IPI sending and
++ * compresses data structures.
+ */
-+static inline unsigned int cpuid_eax(unsigned int op)
-+{
-+ unsigned int eax;
+
-+ __asm__(XEN_CPUID
-+ : "=a" (eax)
-+ : "0" (op)
-+ : "bx", "cx", "dx");
-+ return eax;
-+}
-+static inline unsigned int cpuid_ebx(unsigned int op)
++static inline int num_booting_cpus(void)
+{
-+ unsigned int eax, ebx;
-+
-+ __asm__(XEN_CPUID
-+ : "=a" (eax), "=b" (ebx)
-+ : "0" (op)
-+ : "cx", "dx" );
-+ return ebx;
++ return cpus_weight(cpu_possible_map);
+}
-+static inline unsigned int cpuid_ecx(unsigned int op)
-+{
-+ unsigned int eax, ecx;
+
-+ __asm__(XEN_CPUID
-+ : "=a" (eax), "=c" (ecx)
-+ : "0" (op)
-+ : "bx", "dx" );
-+ return ecx;
-+}
-+static inline unsigned int cpuid_edx(unsigned int op)
-+{
-+ unsigned int eax, edx;
++#define raw_smp_processor_id() read_pda(cpunumber)
+
-+ __asm__(XEN_CPUID
-+ : "=a" (eax), "=d" (edx)
-+ : "0" (op)
-+ : "bx", "cx");
-+ return edx;
++#ifdef CONFIG_X86_LOCAL_APIC
++static inline int hard_smp_processor_id(void)
++{
++ /* we don't want to mark this access volatile - bad code generation */
++ return GET_APIC_ID(*(unsigned int *)(APIC_BASE+APIC_ID));
+}
++#endif
++
++extern int safe_smp_processor_id(void);
++extern int __cpu_disable(void);
++extern void __cpu_die(unsigned int cpu);
++extern void prefill_possible_map(void);
++extern unsigned num_processors;
++extern unsigned disabled_cpus;
+
-+#define MSR_IA32_UCODE_WRITE 0x79
-+#define MSR_IA32_UCODE_REV 0x8b
++#endif /* !ASSEMBLY */
+
++#define NO_PROC_ID 0xFF /* No processor magic marker */
+
+#endif
+
-+/* AMD/K8 specific MSRs */
-+#define MSR_EFER 0xc0000080 /* extended feature register */
-+#define MSR_STAR 0xc0000081 /* legacy mode SYSCALL target */
-+#define MSR_LSTAR 0xc0000082 /* long mode SYSCALL target */
-+#define MSR_CSTAR 0xc0000083 /* compatibility mode SYSCALL target */
-+#define MSR_SYSCALL_MASK 0xc0000084 /* EFLAGS mask for syscall */
-+#define MSR_FS_BASE 0xc0000100 /* 64bit GS base */
-+#define MSR_GS_BASE 0xc0000101 /* 64bit FS base */
-+#define MSR_KERNEL_GS_BASE 0xc0000102 /* SwapGS GS shadow (or USER_GS from kernel) */
-+/* EFER bits: */
-+#define _EFER_SCE 0 /* SYSCALL/SYSRET */
-+#define _EFER_LME 8 /* Long mode enable */
-+#define _EFER_LMA 10 /* Long mode active (read-only) */
-+#define _EFER_NX 11 /* No execute enable */
++#ifndef ASSEMBLY
++/*
++ * Some lowlevel functions might want to know about
++ * the real APIC ID <-> CPU # mapping.
++ */
++extern u8 x86_cpu_to_apicid[NR_CPUS]; /* physical ID */
++extern u8 x86_cpu_to_log_apicid[NR_CPUS];
++extern u8 bios_cpu_apicid[];
+
-+#define EFER_SCE (1<<_EFER_SCE)
-+#define EFER_LME (1<<_EFER_LME)
-+#define EFER_LMA (1<<_EFER_LMA)
-+#define EFER_NX (1<<_EFER_NX)
++#ifdef CONFIG_X86_LOCAL_APIC
++static inline unsigned int cpu_mask_to_apicid(cpumask_t cpumask)
++{
++ return cpus_addr(cpumask)[0];
++}
+
-+/* Intel MSRs. Some also available on other CPUs */
-+#define MSR_IA32_TSC 0x10
-+#define MSR_IA32_PLATFORM_ID 0x17
++static inline int cpu_present_to_apicid(int mps_cpu)
++{
++ if (mps_cpu < NR_CPUS)
++ return (int)bios_cpu_apicid[mps_cpu];
++ else
++ return BAD_APICID;
++}
++#endif
+
-+#define MSR_IA32_PERFCTR0 0xc1
-+#define MSR_IA32_PERFCTR1 0xc2
++#endif /* !ASSEMBLY */
+
-+#define MSR_MTRRcap 0x0fe
-+#define MSR_IA32_BBL_CR_CTL 0x119
++#ifndef CONFIG_SMP
++#define stack_smp_processor_id() 0
++#define safe_smp_processor_id() 0
++#define cpu_logical_map(x) (x)
++#else
++#include <asm/thread_info.h>
++#define stack_smp_processor_id() \
++({ \
++ struct thread_info *ti; \
++ __asm__("andq %%rsp,%0; ":"=r" (ti) : "0" (CURRENT_MASK)); \
++ ti->cpu; \
++})
++#endif
+
-+#define MSR_IA32_SYSENTER_CS 0x174
-+#define MSR_IA32_SYSENTER_ESP 0x175
-+#define MSR_IA32_SYSENTER_EIP 0x176
++#ifndef __ASSEMBLY__
++#ifdef CONFIG_X86_LOCAL_APIC
++static __inline int logical_smp_processor_id(void)
++{
++ /* we don't want to mark this access volatile - bad code generation */
++ return GET_APIC_LOGICAL_ID(*(unsigned long *)(APIC_BASE+APIC_LDR));
++}
++#endif
++#endif
+
-+#define MSR_IA32_MCG_CAP 0x179
-+#define MSR_IA32_MCG_STATUS 0x17a
-+#define MSR_IA32_MCG_CTL 0x17b
++#ifdef CONFIG_SMP
++#define cpu_physical_id(cpu) x86_cpu_to_apicid[cpu]
++#else
++#define cpu_physical_id(cpu) boot_cpu_id
++#endif
+
-+#define MSR_IA32_EVNTSEL0 0x186
-+#define MSR_IA32_EVNTSEL1 0x187
++#endif
+
-+#define MSR_IA32_DEBUGCTLMSR 0x1d9
-+#define MSR_IA32_LASTBRANCHFROMIP 0x1db
-+#define MSR_IA32_LASTBRANCHTOIP 0x1dc
-+#define MSR_IA32_LASTINTFROMIP 0x1dd
-+#define MSR_IA32_LASTINTTOIP 0x1de
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/synch_bitops.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/synch_bitops.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,2 @@
+
-+#define MSR_MTRRfix64K_00000 0x250
-+#define MSR_MTRRfix16K_80000 0x258
-+#define MSR_MTRRfix16K_A0000 0x259
-+#define MSR_MTRRfix4K_C0000 0x268
-+#define MSR_MTRRfix4K_C8000 0x269
-+#define MSR_MTRRfix4K_D0000 0x26a
-+#define MSR_MTRRfix4K_D8000 0x26b
-+#define MSR_MTRRfix4K_E0000 0x26c
-+#define MSR_MTRRfix4K_E8000 0x26d
-+#define MSR_MTRRfix4K_F0000 0x26e
-+#define MSR_MTRRfix4K_F8000 0x26f
-+#define MSR_MTRRdefType 0x2ff
++#include <asm-i386/mach-xen/asm/synch_bitops.h>
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/system.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/system.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,256 @@
++#ifndef __ASM_SYSTEM_H
++#define __ASM_SYSTEM_H
+
-+#define MSR_IA32_MC0_CTL 0x400
-+#define MSR_IA32_MC0_STATUS 0x401
-+#define MSR_IA32_MC0_ADDR 0x402
-+#define MSR_IA32_MC0_MISC 0x403
++#include <linux/kernel.h>
++#include <asm/segment.h>
++#include <asm/alternative.h>
+
-+#define MSR_P6_PERFCTR0 0xc1
-+#define MSR_P6_PERFCTR1 0xc2
-+#define MSR_P6_EVNTSEL0 0x186
-+#define MSR_P6_EVNTSEL1 0x187
++#include <asm/synch_bitops.h>
++#include <asm/hypervisor.h>
++#include <xen/interface/arch-x86_64.h>
+
-+/* K7/K8 MSRs. Not complete. See the architecture manual for a more complete list. */
-+#define MSR_K7_EVNTSEL0 0xC0010000
-+#define MSR_K7_PERFCTR0 0xC0010004
-+#define MSR_K7_EVNTSEL1 0xC0010001
-+#define MSR_K7_PERFCTR1 0xC0010005
-+#define MSR_K7_EVNTSEL2 0xC0010002
-+#define MSR_K7_PERFCTR2 0xC0010006
-+#define MSR_K7_EVNTSEL3 0xC0010003
-+#define MSR_K7_PERFCTR3 0xC0010007
-+#define MSR_K8_TOP_MEM1 0xC001001A
-+#define MSR_K8_TOP_MEM2 0xC001001D
-+#define MSR_K8_SYSCFG 0xC0010010
-+#define MSR_K8_HWCR 0xC0010015
++#ifdef __KERNEL__
+
-+/* K6 MSRs */
-+#define MSR_K6_EFER 0xC0000080
-+#define MSR_K6_STAR 0xC0000081
-+#define MSR_K6_WHCR 0xC0000082
-+#define MSR_K6_UWCCR 0xC0000085
-+#define MSR_K6_PSOR 0xC0000087
-+#define MSR_K6_PFIR 0xC0000088
++#define __STR(x) #x
++#define STR(x) __STR(x)
+
-+/* Centaur-Hauls/IDT defined MSRs. */
-+#define MSR_IDT_FCR1 0x107
-+#define MSR_IDT_FCR2 0x108
-+#define MSR_IDT_FCR3 0x109
-+#define MSR_IDT_FCR4 0x10a
++#define __SAVE(reg,offset) "movq %%" #reg ",(14-" #offset ")*8(%%rsp)\n\t"
++#define __RESTORE(reg,offset) "movq (14-" #offset ")*8(%%rsp),%%" #reg "\n\t"
+
-+#define MSR_IDT_MCR0 0x110
-+#define MSR_IDT_MCR1 0x111
-+#define MSR_IDT_MCR2 0x112
-+#define MSR_IDT_MCR3 0x113
-+#define MSR_IDT_MCR4 0x114
-+#define MSR_IDT_MCR5 0x115
-+#define MSR_IDT_MCR6 0x116
-+#define MSR_IDT_MCR7 0x117
-+#define MSR_IDT_MCR_CTRL 0x120
++/* frame pointer must be last for get_wchan */
++#define SAVE_CONTEXT "pushf ; pushq %%rbp ; movq %%rsi,%%rbp\n\t"
++#define RESTORE_CONTEXT "movq %%rbp,%%rsi ; popq %%rbp ; popf\n\t"
+
-+/* VIA Cyrix defined MSRs*/
-+#define MSR_VIA_FCR 0x1107
-+#define MSR_VIA_LONGHAUL 0x110a
-+#define MSR_VIA_RNG 0x110b
-+#define MSR_VIA_BCR2 0x1147
++#define __EXTRA_CLOBBER \
++ ,"rcx","rbx","rdx","r8","r9","r10","r11","r12","r13","r14","r15"
+
-+/* Intel defined MSRs. */
-+#define MSR_IA32_P5_MC_ADDR 0
-+#define MSR_IA32_P5_MC_TYPE 1
-+#define MSR_IA32_PLATFORM_ID 0x17
-+#define MSR_IA32_EBL_CR_POWERON 0x2a
++#define switch_to(prev,next,last) \
++ asm volatile(SAVE_CONTEXT \
++ "movq %%rsp,%P[threadrsp](%[prev])\n\t" /* save RSP */ \
++ "movq %P[threadrsp](%[next]),%%rsp\n\t" /* restore RSP */ \
++ "call __switch_to\n\t" \
++ ".globl thread_return\n" \
++ "thread_return:\n\t" \
++ "movq %%gs:%P[pda_pcurrent],%%rsi\n\t" \
++ "movq %P[thread_info](%%rsi),%%r8\n\t" \
++ LOCK_PREFIX "btr %[tif_fork],%P[ti_flags](%%r8)\n\t" \
++ "movq %%rax,%%rdi\n\t" \
++ "jc ret_from_fork\n\t" \
++ RESTORE_CONTEXT \
++ : "=a" (last) \
++ : [next] "S" (next), [prev] "D" (prev), \
++ [threadrsp] "i" (offsetof(struct task_struct, thread.rsp)), \
++ [ti_flags] "i" (offsetof(struct thread_info, flags)),\
++ [tif_fork] "i" (TIF_FORK), \
++ [thread_info] "i" (offsetof(struct task_struct, thread_info)), \
++ [pda_pcurrent] "i" (offsetof(struct x8664_pda, pcurrent)) \
++ : "memory", "cc" __EXTRA_CLOBBER)
++
++extern void load_gs_index(unsigned);
+
-+#define MSR_IA32_APICBASE 0x1b
-+#define MSR_IA32_APICBASE_BSP (1<<8)
-+#define MSR_IA32_APICBASE_ENABLE (1<<11)
-+#define MSR_IA32_APICBASE_BASE (0xfffff<<12)
++/*
++ * Load a segment. Fall back on loading the zero
++ * segment if something goes wrong..
++ */
++#define loadsegment(seg,value) \
++ asm volatile("\n" \
++ "1:\t" \
++ "movl %k0,%%" #seg "\n" \
++ "2:\n" \
++ ".section .fixup,\"ax\"\n" \
++ "3:\t" \
++ "movl %1,%%" #seg "\n\t" \
++ "jmp 2b\n" \
++ ".previous\n" \
++ ".section __ex_table,\"a\"\n\t" \
++ ".align 8\n\t" \
++ ".quad 1b,3b\n" \
++ ".previous" \
++ : :"r" (value), "r" (0))
+
-+/* P4/Xeon+ specific */
-+#define MSR_IA32_MCG_EAX 0x180
-+#define MSR_IA32_MCG_EBX 0x181
-+#define MSR_IA32_MCG_ECX 0x182
-+#define MSR_IA32_MCG_EDX 0x183
-+#define MSR_IA32_MCG_ESI 0x184
-+#define MSR_IA32_MCG_EDI 0x185
-+#define MSR_IA32_MCG_EBP 0x186
-+#define MSR_IA32_MCG_ESP 0x187
-+#define MSR_IA32_MCG_EFLAGS 0x188
-+#define MSR_IA32_MCG_EIP 0x189
-+#define MSR_IA32_MCG_RESERVED 0x18A
++/*
++ * Clear and set 'TS' bit respectively
++ */
++#define clts() (HYPERVISOR_fpu_taskswitch(0))
+
-+#define MSR_P6_EVNTSEL0 0x186
-+#define MSR_P6_EVNTSEL1 0x187
++static inline unsigned long read_cr0(void)
++{
++ unsigned long cr0;
++ asm volatile("movq %%cr0,%0" : "=r" (cr0));
++ return cr0;
++}
+
-+#define MSR_IA32_PERF_STATUS 0x198
-+#define MSR_IA32_PERF_CTL 0x199
++static inline void write_cr0(unsigned long val)
++{
++ asm volatile("movq %0,%%cr0" :: "r" (val));
++}
+
-+#define MSR_IA32_THERM_CONTROL 0x19a
-+#define MSR_IA32_THERM_INTERRUPT 0x19b
-+#define MSR_IA32_THERM_STATUS 0x19c
-+#define MSR_IA32_MISC_ENABLE 0x1a0
++#define read_cr3() ({ \
++ unsigned long __dummy; \
++ asm("movq %%cr3,%0" : "=r" (__dummy)); \
++ machine_to_phys(__dummy); \
++})
+
-+#define MSR_IA32_DEBUGCTLMSR 0x1d9
-+#define MSR_IA32_LASTBRANCHFROMIP 0x1db
-+#define MSR_IA32_LASTBRANCHTOIP 0x1dc
-+#define MSR_IA32_LASTINTFROMIP 0x1dd
-+#define MSR_IA32_LASTINTTOIP 0x1de
++static inline unsigned long read_cr4(void)
++{
++ unsigned long cr4;
++ asm("movq %%cr4,%0" : "=r" (cr4));
++ return cr4;
++}
+
-+#define MSR_IA32_MC0_CTL 0x400
-+#define MSR_IA32_MC0_STATUS 0x401
-+#define MSR_IA32_MC0_ADDR 0x402
-+#define MSR_IA32_MC0_MISC 0x403
++static inline void write_cr4(unsigned long val)
++{
++ asm volatile("movq %0,%%cr4" :: "r" (val));
++}
+
-+/* Pentium IV performance counter MSRs */
-+#define MSR_P4_BPU_PERFCTR0 0x300
-+#define MSR_P4_BPU_PERFCTR1 0x301
-+#define MSR_P4_BPU_PERFCTR2 0x302
-+#define MSR_P4_BPU_PERFCTR3 0x303
-+#define MSR_P4_MS_PERFCTR0 0x304
-+#define MSR_P4_MS_PERFCTR1 0x305
-+#define MSR_P4_MS_PERFCTR2 0x306
-+#define MSR_P4_MS_PERFCTR3 0x307
-+#define MSR_P4_FLAME_PERFCTR0 0x308
-+#define MSR_P4_FLAME_PERFCTR1 0x309
-+#define MSR_P4_FLAME_PERFCTR2 0x30a
-+#define MSR_P4_FLAME_PERFCTR3 0x30b
-+#define MSR_P4_IQ_PERFCTR0 0x30c
-+#define MSR_P4_IQ_PERFCTR1 0x30d
-+#define MSR_P4_IQ_PERFCTR2 0x30e
-+#define MSR_P4_IQ_PERFCTR3 0x30f
-+#define MSR_P4_IQ_PERFCTR4 0x310
-+#define MSR_P4_IQ_PERFCTR5 0x311
-+#define MSR_P4_BPU_CCCR0 0x360
-+#define MSR_P4_BPU_CCCR1 0x361
-+#define MSR_P4_BPU_CCCR2 0x362
-+#define MSR_P4_BPU_CCCR3 0x363
-+#define MSR_P4_MS_CCCR0 0x364
-+#define MSR_P4_MS_CCCR1 0x365
-+#define MSR_P4_MS_CCCR2 0x366
-+#define MSR_P4_MS_CCCR3 0x367
-+#define MSR_P4_FLAME_CCCR0 0x368
-+#define MSR_P4_FLAME_CCCR1 0x369
-+#define MSR_P4_FLAME_CCCR2 0x36a
-+#define MSR_P4_FLAME_CCCR3 0x36b
-+#define MSR_P4_IQ_CCCR0 0x36c
-+#define MSR_P4_IQ_CCCR1 0x36d
-+#define MSR_P4_IQ_CCCR2 0x36e
-+#define MSR_P4_IQ_CCCR3 0x36f
-+#define MSR_P4_IQ_CCCR4 0x370
-+#define MSR_P4_IQ_CCCR5 0x371
-+#define MSR_P4_ALF_ESCR0 0x3ca
-+#define MSR_P4_ALF_ESCR1 0x3cb
-+#define MSR_P4_BPU_ESCR0 0x3b2
-+#define MSR_P4_BPU_ESCR1 0x3b3
-+#define MSR_P4_BSU_ESCR0 0x3a0
-+#define MSR_P4_BSU_ESCR1 0x3a1
-+#define MSR_P4_CRU_ESCR0 0x3b8
-+#define MSR_P4_CRU_ESCR1 0x3b9
-+#define MSR_P4_CRU_ESCR2 0x3cc
-+#define MSR_P4_CRU_ESCR3 0x3cd
-+#define MSR_P4_CRU_ESCR4 0x3e0
-+#define MSR_P4_CRU_ESCR5 0x3e1
-+#define MSR_P4_DAC_ESCR0 0x3a8
-+#define MSR_P4_DAC_ESCR1 0x3a9
-+#define MSR_P4_FIRM_ESCR0 0x3a4
-+#define MSR_P4_FIRM_ESCR1 0x3a5
-+#define MSR_P4_FLAME_ESCR0 0x3a6
-+#define MSR_P4_FLAME_ESCR1 0x3a7
-+#define MSR_P4_FSB_ESCR0 0x3a2
-+#define MSR_P4_FSB_ESCR1 0x3a3
-+#define MSR_P4_IQ_ESCR0 0x3ba
-+#define MSR_P4_IQ_ESCR1 0x3bb
-+#define MSR_P4_IS_ESCR0 0x3b4
-+#define MSR_P4_IS_ESCR1 0x3b5
-+#define MSR_P4_ITLB_ESCR0 0x3b6
-+#define MSR_P4_ITLB_ESCR1 0x3b7
-+#define MSR_P4_IX_ESCR0 0x3c8
-+#define MSR_P4_IX_ESCR1 0x3c9
-+#define MSR_P4_MOB_ESCR0 0x3aa
-+#define MSR_P4_MOB_ESCR1 0x3ab
-+#define MSR_P4_MS_ESCR0 0x3c0
-+#define MSR_P4_MS_ESCR1 0x3c1
-+#define MSR_P4_PMH_ESCR0 0x3ac
-+#define MSR_P4_PMH_ESCR1 0x3ad
-+#define MSR_P4_RAT_ESCR0 0x3bc
-+#define MSR_P4_RAT_ESCR1 0x3bd
-+#define MSR_P4_SAAT_ESCR0 0x3ae
-+#define MSR_P4_SAAT_ESCR1 0x3af
-+#define MSR_P4_SSU_ESCR0 0x3be
-+#define MSR_P4_SSU_ESCR1 0x3bf /* guess: not defined in manual */
-+#define MSR_P4_TBPU_ESCR0 0x3c2
-+#define MSR_P4_TBPU_ESCR1 0x3c3
-+#define MSR_P4_TC_ESCR0 0x3c4
-+#define MSR_P4_TC_ESCR1 0x3c5
-+#define MSR_P4_U2L_ESCR0 0x3b0
-+#define MSR_P4_U2L_ESCR1 0x3b1
++#define stts() (HYPERVISOR_fpu_taskswitch(1))
++
++#define wbinvd() \
++ __asm__ __volatile__ ("wbinvd": : :"memory");
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/nmi.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/nmi.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/nmi.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/nmi.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,93 @@
+/*
-+ * linux/include/asm-i386/nmi.h
++ * On SMP systems, when the scheduler does migration-cost autodetection,
++ * it needs a way to flush as much of the CPU's caches as possible.
+ */
-+#ifndef ASM_NMI_H
-+#define ASM_NMI_H
-+
-+#include <linux/pm.h>
-+#include <asm/io.h>
-+
-+#include <xen/interface/nmi.h>
++static inline void sched_cacheflush(void)
++{
++ wbinvd();
++}
+
-+struct pt_regs;
++#endif /* __KERNEL__ */
+
-+typedef int (*nmi_callback_t)(struct pt_regs * regs, int cpu);
++#define nop() __asm__ __volatile__ ("nop")
+
-+/**
-+ * set_nmi_callback
-+ *
-+ * Set a handler for an NMI. Only one handler may be
-+ * set. Return 1 if the NMI was handled.
-+ */
-+void set_nmi_callback(nmi_callback_t callback);
++#define xchg(ptr,v) ((__typeof__(*(ptr)))__xchg((unsigned long)(v),(ptr),sizeof(*(ptr))))
+
-+/**
-+ * unset_nmi_callback
-+ *
-+ * Remove the handler previously set.
-+ */
-+void unset_nmi_callback(void);
++#define tas(ptr) (xchg((ptr),1))
+
-+#ifdef CONFIG_PM
-+
-+/** Replace the PM callback routine for NMI. */
-+struct pm_dev * set_nmi_pm_callback(pm_callback callback);
++#define __xg(x) ((volatile long *)(x))
+
-+/** Unset the PM callback routine back to the default. */
-+void unset_nmi_pm_callback(struct pm_dev * dev);
++static inline void set_64bit(volatile unsigned long *ptr, unsigned long val)
++{
++ *ptr = val;
++}
+
-+#else
++#define _set_64bit set_64bit
+
-+static inline struct pm_dev * set_nmi_pm_callback(pm_callback callback)
-+{
-+ return 0;
-+}
-+
-+static inline void unset_nmi_pm_callback(struct pm_dev * dev)
++/*
++ * Note: no "lock" prefix even on SMP: xchg always implies lock anyway
++ * Note 2: xchg has side effect, so that attribute volatile is necessary,
++ * but generally the primitive is invalid, *ptr is output argument. --ANK
++ */
++static inline unsigned long __xchg(unsigned long x, volatile void * ptr, int size)
+{
++ switch (size) {
++ case 1:
++ __asm__ __volatile__("xchgb %b0,%1"
++ :"=q" (x)
++ :"m" (*__xg(ptr)), "0" (x)
++ :"memory");
++ break;
++ case 2:
++ __asm__ __volatile__("xchgw %w0,%1"
++ :"=r" (x)
++ :"m" (*__xg(ptr)), "0" (x)
++ :"memory");
++ break;
++ case 4:
++ __asm__ __volatile__("xchgl %k0,%1"
++ :"=r" (x)
++ :"m" (*__xg(ptr)), "0" (x)
++ :"memory");
++ break;
++ case 8:
++ __asm__ __volatile__("xchgq %0,%1"
++ :"=r" (x)
++ :"m" (*__xg(ptr)), "0" (x)
++ :"memory");
++ break;
++ }
++ return x;
+}
+
-+#endif /* CONFIG_PM */
-+
-+extern void default_do_nmi(struct pt_regs *);
-+extern void die_nmi(char *str, struct pt_regs *regs);
++/*
++ * Atomic compare and exchange. Compare OLD with MEM, if identical,
++ * store NEW in MEM. Return the initial value in MEM. Success is
++ * indicated by comparing RETURN with OLD.
++ */
+
-+static inline unsigned char get_nmi_reason(void)
++#define __HAVE_ARCH_CMPXCHG 1
++
++static inline unsigned long __cmpxchg(volatile void *ptr, unsigned long old,
++ unsigned long new, int size)
+{
-+ shared_info_t *s = HYPERVISOR_shared_info;
-+ unsigned char reason = 0;
++ unsigned long prev;
++ switch (size) {
++ case 1:
++ __asm__ __volatile__(LOCK_PREFIX "cmpxchgb %b1,%2"
++ : "=a"(prev)
++ : "q"(new), "m"(*__xg(ptr)), "0"(old)
++ : "memory");
++ return prev;
++ case 2:
++ __asm__ __volatile__(LOCK_PREFIX "cmpxchgw %w1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__xg(ptr)), "0"(old)
++ : "memory");
++ return prev;
++ case 4:
++ __asm__ __volatile__(LOCK_PREFIX "cmpxchgl %k1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__xg(ptr)), "0"(old)
++ : "memory");
++ return prev;
++ case 8:
++ __asm__ __volatile__(LOCK_PREFIX "cmpxchgq %1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__xg(ptr)), "0"(old)
++ : "memory");
++ return prev;
++ }
++ return old;
++}
+
-+ /* construct a value which looks like it came from
-+ * port 0x61.
-+ */
-+ if (test_bit(_XEN_NMIREASON_io_error, &s->arch.nmi_reason))
-+ reason |= 0x40;
-+ if (test_bit(_XEN_NMIREASON_parity_error, &s->arch.nmi_reason))
-+ reason |= 0x80;
++#define cmpxchg(ptr,o,n)\
++ ((__typeof__(*(ptr)))__cmpxchg((ptr),(unsigned long)(o),\
++ (unsigned long)(n),sizeof(*(ptr))))
++
++#ifdef CONFIG_SMP
++#define smp_mb() mb()
++#define smp_rmb() rmb()
++#define smp_wmb() wmb()
++#define smp_read_barrier_depends() do {} while(0)
++#else
++#define smp_mb() barrier()
++#define smp_rmb() barrier()
++#define smp_wmb() barrier()
++#define smp_read_barrier_depends() do {} while(0)
++#endif
+
-+ return reason;
-+}
++
++/*
++ * Force strict CPU ordering.
++ * And yes, this is required on UP too when we're talking
++ * to devices.
++ */
++#define mb() asm volatile("mfence":::"memory")
++#define rmb() asm volatile("lfence":::"memory")
+
-+extern int panic_on_timeout;
-+extern int unknown_nmi_panic;
++#ifdef CONFIG_UNORDERED_IO
++#define wmb() asm volatile("sfence" ::: "memory")
++#else
++#define wmb() asm volatile("" ::: "memory")
++#endif
++#define read_barrier_depends() do {} while(0)
++#define set_mb(var, value) do { (void) xchg(&var, value); } while (0)
+
-+extern int check_nmi_watchdog(void);
-+
-+extern void setup_apic_nmi_watchdog (void);
-+extern int reserve_lapic_nmi(void);
-+extern void release_lapic_nmi(void);
-+extern void disable_timer_nmi_watchdog(void);
-+extern void enable_timer_nmi_watchdog(void);
-+extern void nmi_watchdog_tick (struct pt_regs * regs, unsigned reason);
++#define warn_if_not_ulong(x) do { unsigned long foo; (void) (&(x) == &foo); } while (0)
+
-+extern void nmi_watchdog_default(void);
-+extern int setup_nmi_watchdog(char *);
++#include <linux/irqflags.h>
+
-+extern unsigned int nmi_watchdog;
-+#define NMI_DEFAULT -1
-+#define NMI_NONE 0
-+#define NMI_IO_APIC 1
-+#define NMI_LOCAL_APIC 2
-+#define NMI_INVALID 3
++void cpu_idle_wait(void);
+
-+#endif /* ASM_NMI_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/page.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/page.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/page.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/page.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,217 @@
-+#ifndef _X86_64_PAGE_H
-+#define _X86_64_PAGE_H
++extern unsigned long arch_align_stack(unsigned long sp);
++extern void free_init_pages(char *what, unsigned long begin, unsigned long end);
+
-+/* #include <linux/string.h> */
-+#ifndef __ASSEMBLY__
-+#include <linux/kernel.h>
-+#include <linux/types.h>
-+#include <asm/bug.h>
+#endif
-+#include <xen/interface/xen.h>
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/timer.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/timer.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,67 @@
++#ifndef _ASMi386_TIMER_H
++#define _ASMi386_TIMER_H
++#include <linux/init.h>
+
-+/*
-+ * Need to repeat this here in order to not include pgtable.h (which in turn
-+ * depends on definitions made here), but to be able to use the symbolic
-+ * below. The preprocessor will warn if the two definitions aren't identical.
++/**
++ * struct timer_ops - used to define a timer source
++ *
++ * @name: name of the timer.
++ * @init: Probes and initializes the timer. Takes clock= override
++ * string as an argument. Returns 0 on success, anything else
++ * on failure.
++ * @mark_offset: called by the timer interrupt.
++ * @get_offset: called by gettimeofday(). Returns the number of microseconds
++ * since the last timer interupt.
++ * @monotonic_clock: returns the number of nanoseconds since the init of the
++ * timer.
++ * @delay: delays this many clock cycles.
+ */
-+#define _PAGE_PRESENT 0x001
++struct timer_opts {
++ char* name;
++ void (*mark_offset)(void);
++ unsigned long (*get_offset)(void);
++ unsigned long long (*monotonic_clock)(void);
++ void (*delay)(unsigned long);
++ unsigned long (*read_timer)(void);
++ int (*suspend)(pm_message_t state);
++ int (*resume)(void);
++};
+
-+#define arch_free_page(_page,_order) \
-+({ int foreign = PageForeign(_page); \
-+ if (foreign) \
-+ PageForeignDestructor(_page); \
-+ foreign; \
-+})
-+#define HAVE_ARCH_FREE_PAGE
++struct init_timer_opts {
++ int (*init)(char *override);
++ struct timer_opts *opts;
++};
+
-+/* PAGE_SHIFT determines the page size */
-+#define PAGE_SHIFT 12
-+#ifdef __ASSEMBLY__
-+#define PAGE_SIZE (0x1 << PAGE_SHIFT)
-+#else
-+#define PAGE_SIZE (1UL << PAGE_SHIFT)
-+#endif
-+#define PAGE_MASK (~(PAGE_SIZE-1))
++#define TICK_SIZE (tick_nsec / 1000)
+
-+/* See Documentation/x86_64/mm.txt for a description of the memory map. */
-+#define __PHYSICAL_MASK_SHIFT 46
-+#define __PHYSICAL_MASK ((1UL << __PHYSICAL_MASK_SHIFT) - 1)
-+#define __VIRTUAL_MASK_SHIFT 48
-+#define __VIRTUAL_MASK ((1UL << __VIRTUAL_MASK_SHIFT) - 1)
++extern struct timer_opts* __init select_timer(void);
++extern void clock_fallback(void);
++void setup_pit_timer(void);
+
-+#define PHYSICAL_PAGE_MASK (~(PAGE_SIZE-1) & __PHYSICAL_MASK)
++/* Modifiers for buggy PIT handling */
+
-+#define THREAD_ORDER 1
-+#define THREAD_SIZE (PAGE_SIZE << THREAD_ORDER)
-+#define CURRENT_MASK (~(THREAD_SIZE-1))
++extern int pit_latch_buggy;
+
-+#define EXCEPTION_STACK_ORDER 0
-+#define EXCEPTION_STKSZ (PAGE_SIZE << EXCEPTION_STACK_ORDER)
++extern struct timer_opts *cur_timer;
++extern int timer_ack;
+
-+#define DEBUG_STACK_ORDER (EXCEPTION_STACK_ORDER + 1)
-+#define DEBUG_STKSZ (PAGE_SIZE << DEBUG_STACK_ORDER)
++/* list of externed timers */
++extern struct timer_opts timer_none;
++extern struct timer_opts timer_pit;
++extern struct init_timer_opts timer_pit_init;
++extern struct init_timer_opts timer_tsc_init;
++#ifdef CONFIG_X86_CYCLONE_TIMER
++extern struct init_timer_opts timer_cyclone_init;
++#endif
+
-+#define IRQSTACK_ORDER 2
-+#define IRQSTACKSIZE (PAGE_SIZE << IRQSTACK_ORDER)
++extern unsigned long calibrate_tsc(void);
++extern void init_cpu_khz(void);
++#ifdef CONFIG_HPET_TIMER
++extern struct init_timer_opts timer_hpet_init;
++extern unsigned long calibrate_tsc_hpet(unsigned long *tsc_hpet_quotient_ptr);
++#endif
+
-+#define STACKFAULT_STACK 1
-+#define DOUBLEFAULT_STACK 2
-+#define NMI_STACK 3
-+#define DEBUG_STACK 4
-+#define MCE_STACK 5
-+#define N_EXCEPTION_STACKS 5 /* hw limit: 7 */
++#ifdef CONFIG_X86_PM_TIMER
++extern struct init_timer_opts timer_pmtmr_init;
++#endif
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/tlbflush.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/tlbflush.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,103 @@
++#ifndef _X8664_TLBFLUSH_H
++#define _X8664_TLBFLUSH_H
+
-+#define LARGE_PAGE_MASK (~(LARGE_PAGE_SIZE-1))
-+#define LARGE_PAGE_SIZE (1UL << PMD_SHIFT)
++#include <linux/mm.h>
++#include <asm/processor.h>
+
-+#define HPAGE_SHIFT PMD_SHIFT
-+#define HPAGE_SIZE ((1UL) << HPAGE_SHIFT)
-+#define HPAGE_MASK (~(HPAGE_SIZE - 1))
-+#define HUGETLB_PAGE_ORDER (HPAGE_SHIFT - PAGE_SHIFT)
++#define __flush_tlb() xen_tlb_flush()
+
-+#ifdef __KERNEL__
-+#ifndef __ASSEMBLY__
++/*
++ * Global pages have to be flushed a bit differently. Not a real
++ * performance problem because this does not happen often.
++ */
++#define __flush_tlb_global() xen_tlb_flush()
+
-+extern unsigned long end_pfn;
+
-+#include <asm/maddr.h>
++extern unsigned long pgkern_mask;
+
-+void clear_page(void *);
-+void copy_page(void *, void *);
++#define __flush_tlb_all() __flush_tlb_global()
+
-+#define clear_user_page(page, vaddr, pg) clear_page(page)
-+#define copy_user_page(to, from, vaddr, pg) copy_page(to, from)
++#define __flush_tlb_one(addr) xen_invlpg((unsigned long)addr)
+
-+#define alloc_zeroed_user_highpage(vma, vaddr) alloc_page_vma(GFP_HIGHUSER | __GFP_ZERO, vma, vaddr)
-+#define __HAVE_ARCH_ALLOC_ZEROED_USER_HIGHPAGE
+
+/*
-+ * These are used to make use of C type-checking..
++ * TLB flushing:
++ *
++ * - flush_tlb() flushes the current mm struct TLBs
++ * - flush_tlb_all() flushes all processes TLBs
++ * - flush_tlb_mm(mm) flushes the specified mm context TLB's
++ * - flush_tlb_page(vma, vmaddr) flushes one page
++ * - flush_tlb_range(vma, start, end) flushes a range of pages
++ * - flush_tlb_kernel_range(start, end) flushes a range of kernel pages
++ * - flush_tlb_pgtables(mm, start, end) flushes a range of page tables
++ *
++ * x86-64 can only flush individual pages or full VMs. For a range flush
++ * we always do the full VM. Might be worth trying if for a small
++ * range a few INVLPGs in a row are a win.
+ */
-+typedef struct { unsigned long pte; } pte_t;
-+typedef struct { unsigned long pmd; } pmd_t;
-+typedef struct { unsigned long pud; } pud_t;
-+typedef struct { unsigned long pgd; } pgd_t;
-+#define PTE_MASK PHYSICAL_PAGE_MASK
+
-+typedef struct { unsigned long pgprot; } pgprot_t;
++#ifndef CONFIG_SMP
+
-+#define __pte_val(x) ((x).pte)
-+#define pte_val(x) ((__pte_val(x) & _PAGE_PRESENT) ? \
-+ pte_machine_to_phys(__pte_val(x)) : \
-+ __pte_val(x))
++#define flush_tlb() __flush_tlb()
++#define flush_tlb_all() __flush_tlb_all()
++#define local_flush_tlb() __flush_tlb()
+
-+#define __pmd_val(x) ((x).pmd)
-+static inline unsigned long pmd_val(pmd_t x)
++static inline void flush_tlb_mm(struct mm_struct *mm)
+{
-+ unsigned long ret = __pmd_val(x);
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (ret) ret = pte_machine_to_phys(ret) | _PAGE_PRESENT;
-+#else
-+ if (ret & _PAGE_PRESENT) ret = pte_machine_to_phys(ret);
-+#endif
-+ return ret;
++ if (mm == current->active_mm)
++ __flush_tlb();
+}
+
-+#define __pud_val(x) ((x).pud)
-+static inline unsigned long pud_val(pud_t x)
++static inline void flush_tlb_page(struct vm_area_struct *vma,
++ unsigned long addr)
+{
-+ unsigned long ret = __pud_val(x);
-+ if (ret & _PAGE_PRESENT) ret = pte_machine_to_phys(ret);
-+ return ret;
++ if (vma->vm_mm == current->active_mm)
++ __flush_tlb_one(addr);
+}
+
-+#define __pgd_val(x) ((x).pgd)
-+static inline unsigned long pgd_val(pgd_t x)
++static inline void flush_tlb_range(struct vm_area_struct *vma,
++ unsigned long start, unsigned long end)
+{
-+ unsigned long ret = __pgd_val(x);
-+ if (ret & _PAGE_PRESENT) ret = pte_machine_to_phys(ret);
-+ return ret;
++ if (vma->vm_mm == current->active_mm)
++ __flush_tlb();
+}
+
-+#define pgprot_val(x) ((x).pgprot)
++#else
+
-+static inline pte_t __pte(unsigned long x)
-+{
-+ if (x & _PAGE_PRESENT) x = pte_phys_to_machine(x);
-+ return ((pte_t) { (x) });
-+}
++#include <asm/smp.h>
+
-+static inline pmd_t __pmd(unsigned long x)
-+{
-+ if (x & _PAGE_PRESENT) x = pte_phys_to_machine(x);
-+ return ((pmd_t) { (x) });
-+}
++#define local_flush_tlb() \
++ __flush_tlb()
+
-+static inline pud_t __pud(unsigned long x)
-+{
-+ if (x & _PAGE_PRESENT) x = pte_phys_to_machine(x);
-+ return ((pud_t) { (x) });
-+}
++#define flush_tlb_all xen_tlb_flush_all
++#define flush_tlb_current_task() xen_tlb_flush_mask(¤t->mm->cpu_vm_mask)
++#define flush_tlb_mm(mm) xen_tlb_flush_mask(&(mm)->cpu_vm_mask)
++#define flush_tlb_page(vma, va) xen_invlpg_mask(&(vma)->vm_mm->cpu_vm_mask, va)
+
-+static inline pgd_t __pgd(unsigned long x)
++#define flush_tlb() flush_tlb_current_task()
++
++static inline void flush_tlb_range(struct vm_area_struct * vma, unsigned long start, unsigned long end)
+{
-+ if (x & _PAGE_PRESENT) x = pte_phys_to_machine(x);
-+ return ((pgd_t) { (x) });
++ flush_tlb_mm(vma->vm_mm);
+}
+
-+#define __pgprot(x) ((pgprot_t) { (x) } )
-+
-+#define __PHYSICAL_START ((unsigned long)CONFIG_PHYSICAL_START)
-+#define __START_KERNEL (__START_KERNEL_map + __PHYSICAL_START)
-+#define __START_KERNEL_map 0xffffffff80000000UL
-+#define __PAGE_OFFSET 0xffff880000000000UL
++#define TLBSTATE_OK 1
++#define TLBSTATE_LAZY 2
+
-+#else
-+#define __PHYSICAL_START CONFIG_PHYSICAL_START
-+#define __START_KERNEL (__START_KERNEL_map + __PHYSICAL_START)
-+#define __START_KERNEL_map 0xffffffff80000000
-+#define __PAGE_OFFSET 0xffff880000000000
-+#endif /* !__ASSEMBLY__ */
++/* Roughly an IPI every 20MB with 4k pages for freeing page table
++ ranges. Cost is about 42k of memory for each CPU. */
++#define ARCH_FREE_PTE_NR 5350
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+#undef LOAD_OFFSET
-+#define LOAD_OFFSET 0
+#endif
+
-+/* to align the pointer to the (next) page boundary */
-+#define PAGE_ALIGN(addr) (((addr)+PAGE_SIZE-1)&PAGE_MASK)
-+
-+#define KERNEL_TEXT_SIZE (40UL*1024*1024)
-+#define KERNEL_TEXT_START 0xffffffff80000000UL
++#define flush_tlb_kernel_range(start, end) flush_tlb_all()
+
-+#define PAGE_OFFSET ((unsigned long)__PAGE_OFFSET)
++static inline void flush_tlb_pgtables(struct mm_struct *mm,
++ unsigned long start, unsigned long end)
++{
++ /* x86_64 does not keep any page table caches in a software TLB.
++ The CPUs do in their hardware TLBs, but they are handled
++ by the normal TLB flushing algorithms. */
++}
+
-+/* Note: __pa(&symbol_visible_to_c) should be always replaced with __pa_symbol.
-+ Otherwise you risk miscompilation. */
-+#define __pa(x) (((unsigned long)(x)>=__START_KERNEL_map)?(unsigned long)(x) - (unsigned long)__START_KERNEL_map:(unsigned long)(x) - PAGE_OFFSET)
-+/* __pa_symbol should be used for C visible symbols.
-+ This seems to be the official gcc blessed way to do such arithmetic. */
-+#define __pa_symbol(x) \
-+ ({unsigned long v; \
-+ asm("" : "=r" (v) : "0" (x)); \
-+ __pa(v); })
++#endif /* _X8664_TLBFLUSH_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/vga.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/vga.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,20 @@
++/*
++ * Access to VGA videoram
++ *
++ * (c) 1998 Martin Mares <mj at ucw.cz>
++ */
+
-+#define __va(x) ((void *)((unsigned long)(x)+PAGE_OFFSET))
-+#define __boot_va(x) __va(x)
-+#define __boot_pa(x) __pa(x)
-+#ifdef CONFIG_FLATMEM
-+#define pfn_valid(pfn) ((pfn) < end_pfn)
-+#endif
++#ifndef _LINUX_ASM_VGA_H_
++#define _LINUX_ASM_VGA_H_
+
-+#define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT)
-+#define virt_addr_valid(kaddr) pfn_valid(__pa(kaddr) >> PAGE_SHIFT)
-+#define pfn_to_kaddr(pfn) __va((pfn) << PAGE_SHIFT)
++/*
++ * On the PC, we can just recalculate addresses and then
++ * access the videoram directly without any black magic.
++ */
+
-+#define VM_DATA_DEFAULT_FLAGS \
-+ (((current->personality & READ_IMPLIES_EXEC) ? VM_EXEC : 0 ) | \
-+ VM_READ | VM_WRITE | VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC)
++#define VGA_MAP_MEM(x,s) (unsigned long)isa_bus_to_virt(x)
+
-+#define __HAVE_ARCH_GATE_AREA 1
++#define vga_readb(x) (*(x))
++#define vga_writeb(x,y) (*(y) = (x))
+
-+#include <asm-generic/memory_model.h>
-+#include <asm-generic/page.h>
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/xenoprof.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/xenoprof.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,1 @@
++#include <asm-i386/mach-xen/asm/xenoprof.h>
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/asm/xor.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/asm/xor.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,328 @@
++/*
++ * x86-64 changes / gcc fixes from Andi Kleen.
++ * Copyright 2002 Andi Kleen, SuSE Labs.
++ *
++ * This hasn't been optimized for the hammer yet, but there are likely
++ * no advantages to be gotten from x86-64 here anyways.
++ */
+
-+#endif /* __KERNEL__ */
++typedef struct { unsigned long a,b; } __attribute__((aligned(16))) xmm_store_t;
+
-+#endif /* _X86_64_PAGE_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/pci.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/pci.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/pci.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/pci.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,168 @@
-+#ifndef __x8664_PCI_H
-+#define __x8664_PCI_H
++/* Doesn't use gcc to save the XMM registers, because there is no easy way to
++ tell it to do a clts before the register saving. */
++#define XMMS_SAVE do { \
++ preempt_disable(); \
++ if (!(current_thread_info()->status & TS_USEDFPU)) \
++ clts(); \
++ __asm__ __volatile__ ( \
++ "movups %%xmm0,(%1) ;\n\t" \
++ "movups %%xmm1,0x10(%1) ;\n\t" \
++ "movups %%xmm2,0x20(%1) ;\n\t" \
++ "movups %%xmm3,0x30(%1) ;\n\t" \
++ : "=&r" (cr0) \
++ : "r" (xmm_save) \
++ : "memory"); \
++} while(0)
+
-+#include <asm/io.h>
++#define XMMS_RESTORE do { \
++ asm volatile ( \
++ "sfence ;\n\t" \
++ "movups (%1),%%xmm0 ;\n\t" \
++ "movups 0x10(%1),%%xmm1 ;\n\t" \
++ "movups 0x20(%1),%%xmm2 ;\n\t" \
++ "movups 0x30(%1),%%xmm3 ;\n\t" \
++ : \
++ : "r" (cr0), "r" (xmm_save) \
++ : "memory"); \
++ if (!(current_thread_info()->status & TS_USEDFPU)) \
++ stts(); \
++ preempt_enable(); \
++} while(0)
+
-+#ifdef __KERNEL__
++#define OFFS(x) "16*("#x")"
++#define PF_OFFS(x) "256+16*("#x")"
++#define PF0(x) " prefetchnta "PF_OFFS(x)"(%[p1]) ;\n"
++#define LD(x,y) " movaps "OFFS(x)"(%[p1]), %%xmm"#y" ;\n"
++#define ST(x,y) " movaps %%xmm"#y", "OFFS(x)"(%[p1]) ;\n"
++#define PF1(x) " prefetchnta "PF_OFFS(x)"(%[p2]) ;\n"
++#define PF2(x) " prefetchnta "PF_OFFS(x)"(%[p3]) ;\n"
++#define PF3(x) " prefetchnta "PF_OFFS(x)"(%[p4]) ;\n"
++#define PF4(x) " prefetchnta "PF_OFFS(x)"(%[p5]) ;\n"
++#define PF5(x) " prefetchnta "PF_OFFS(x)"(%[p6]) ;\n"
++#define XO1(x,y) " xorps "OFFS(x)"(%[p2]), %%xmm"#y" ;\n"
++#define XO2(x,y) " xorps "OFFS(x)"(%[p3]), %%xmm"#y" ;\n"
++#define XO3(x,y) " xorps "OFFS(x)"(%[p4]), %%xmm"#y" ;\n"
++#define XO4(x,y) " xorps "OFFS(x)"(%[p5]), %%xmm"#y" ;\n"
++#define XO5(x,y) " xorps "OFFS(x)"(%[p6]), %%xmm"#y" ;\n"
+
-+#include <linux/mm.h> /* for struct page */
+
-+/* Can be used to override the logic in pci_scan_bus for skipping
-+ already-configured bus numbers - to be used for buggy BIOSes
-+ or architectures with incomplete PCI setup by the loader */
++static void
++xor_sse_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
++{
++ unsigned int lines = bytes >> 8;
++ unsigned long cr0;
++ xmm_store_t xmm_save[4];
+
-+#ifdef CONFIG_PCI
-+extern unsigned int pcibios_assign_all_busses(void);
-+#else
-+#define pcibios_assign_all_busses() 0
-+#endif
++ XMMS_SAVE;
+
-+#include <asm/hypervisor.h>
-+#define pcibios_scan_all_fns(a, b) (!is_initial_xendomain())
++ asm volatile (
++#undef BLOCK
++#define BLOCK(i) \
++ LD(i,0) \
++ LD(i+1,1) \
++ PF1(i) \
++ PF1(i+2) \
++ LD(i+2,2) \
++ LD(i+3,3) \
++ PF0(i+4) \
++ PF0(i+6) \
++ XO1(i,0) \
++ XO1(i+1,1) \
++ XO1(i+2,2) \
++ XO1(i+3,3) \
++ ST(i,0) \
++ ST(i+1,1) \
++ ST(i+2,2) \
++ ST(i+3,3) \
+
-+extern unsigned long pci_mem_start;
-+#define PCIBIOS_MIN_IO 0x1000
-+#define PCIBIOS_MIN_MEM (pci_mem_start)
+
-+#define PCIBIOS_MIN_CARDBUS_IO 0x4000
++ PF0(0)
++ PF0(2)
+
-+void pcibios_config_init(void);
-+struct pci_bus * pcibios_scan_root(int bus);
-+extern int (*pci_config_read)(int seg, int bus, int dev, int fn, int reg, int len, u32 *value);
-+extern int (*pci_config_write)(int seg, int bus, int dev, int fn, int reg, int len, u32 value);
++ " .align 32 ;\n"
++ " 1: ;\n"
+
-+void pcibios_set_master(struct pci_dev *dev);
-+void pcibios_penalize_isa_irq(int irq, int active);
-+struct irq_routing_table *pcibios_get_irq_routing_table(void);
-+int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq);
++ BLOCK(0)
++ BLOCK(4)
++ BLOCK(8)
++ BLOCK(12)
+
-+#include <linux/types.h>
-+#include <linux/slab.h>
-+#include <asm/scatterlist.h>
-+#include <linux/string.h>
-+#include <asm/page.h>
++ " addq %[inc], %[p1] ;\n"
++ " addq %[inc], %[p2] ;\n"
++ " decl %[cnt] ; jnz 1b"
++ : [p1] "+r" (p1), [p2] "+r" (p2), [cnt] "+r" (lines)
++ : [inc] "r" (256UL)
++ : "memory");
+
-+extern void pci_iommu_alloc(void);
-+extern int iommu_setup(char *opt);
++ XMMS_RESTORE;
++}
+
-+/* The PCI address space does equal the physical memory
-+ * address space. The networking and block device layers use
-+ * this boolean for bounce buffer decisions
-+ *
-+ * On AMD64 it mostly equals, but we set it to zero if a hardware
-+ * IOMMU (gart) of sotware IOMMU (swiotlb) is available.
-+ */
-+#define PCI_DMA_BUS_IS_PHYS (dma_ops->is_phys)
++static void
++xor_sse_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
++ unsigned long *p3)
++{
++ unsigned int lines = bytes >> 8;
++ xmm_store_t xmm_save[4];
++ unsigned long cr0;
+
-+#if defined(CONFIG_IOMMU) || defined(CONFIG_CALGARY_IOMMU)
++ XMMS_SAVE;
+
-+/*
-+ * x86-64 always supports DAC, but sometimes it is useful to force
-+ * devices through the IOMMU to get automatic sg list merging.
-+ * Optional right now.
-+ */
-+extern int iommu_sac_force;
-+#define pci_dac_dma_supported(pci_dev, mask) (!iommu_sac_force)
++ __asm__ __volatile__ (
++#undef BLOCK
++#define BLOCK(i) \
++ PF1(i) \
++ PF1(i+2) \
++ LD(i,0) \
++ LD(i+1,1) \
++ LD(i+2,2) \
++ LD(i+3,3) \
++ PF2(i) \
++ PF2(i+2) \
++ PF0(i+4) \
++ PF0(i+6) \
++ XO1(i,0) \
++ XO1(i+1,1) \
++ XO1(i+2,2) \
++ XO1(i+3,3) \
++ XO2(i,0) \
++ XO2(i+1,1) \
++ XO2(i+2,2) \
++ XO2(i+3,3) \
++ ST(i,0) \
++ ST(i+1,1) \
++ ST(i+2,2) \
++ ST(i+3,3) \
+
-+#define DECLARE_PCI_UNMAP_ADDR(ADDR_NAME) \
-+ dma_addr_t ADDR_NAME;
-+#define DECLARE_PCI_UNMAP_LEN(LEN_NAME) \
-+ __u32 LEN_NAME;
-+#define pci_unmap_addr(PTR, ADDR_NAME) \
-+ ((PTR)->ADDR_NAME)
-+#define pci_unmap_addr_set(PTR, ADDR_NAME, VAL) \
-+ (((PTR)->ADDR_NAME) = (VAL))
-+#define pci_unmap_len(PTR, LEN_NAME) \
-+ ((PTR)->LEN_NAME)
-+#define pci_unmap_len_set(PTR, LEN_NAME, VAL) \
-+ (((PTR)->LEN_NAME) = (VAL))
+
-+#elif defined(CONFIG_SWIOTLB)
++ PF0(0)
++ PF0(2)
+
-+#define pci_dac_dma_supported(pci_dev, mask) 1
++ " .align 32 ;\n"
++ " 1: ;\n"
+
-+#define DECLARE_PCI_UNMAP_ADDR(ADDR_NAME) \
-+ dma_addr_t ADDR_NAME;
-+#define DECLARE_PCI_UNMAP_LEN(LEN_NAME) \
-+ __u32 LEN_NAME;
-+#define pci_unmap_addr(PTR, ADDR_NAME) \
-+ ((PTR)->ADDR_NAME)
-+#define pci_unmap_addr_set(PTR, ADDR_NAME, VAL) \
-+ (((PTR)->ADDR_NAME) = (VAL))
-+#define pci_unmap_len(PTR, LEN_NAME) \
-+ ((PTR)->LEN_NAME)
-+#define pci_unmap_len_set(PTR, LEN_NAME, VAL) \
-+ (((PTR)->LEN_NAME) = (VAL))
++ BLOCK(0)
++ BLOCK(4)
++ BLOCK(8)
++ BLOCK(12)
+
-+#else
-+/* No IOMMU */
++ " addq %[inc], %[p1] ;\n"
++ " addq %[inc], %[p2] ;\n"
++ " addq %[inc], %[p3] ;\n"
++ " decl %[cnt] ; jnz 1b"
++ : [cnt] "+r" (lines),
++ [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3)
++ : [inc] "r" (256UL)
++ : "memory");
++ XMMS_RESTORE;
++}
+
-+#define pci_dac_dma_supported(pci_dev, mask) 1
++static void
++xor_sse_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
++ unsigned long *p3, unsigned long *p4)
++{
++ unsigned int lines = bytes >> 8;
++ xmm_store_t xmm_save[4];
++ unsigned long cr0;
+
-+#define DECLARE_PCI_UNMAP_ADDR(ADDR_NAME)
-+#define DECLARE_PCI_UNMAP_LEN(LEN_NAME)
-+#define pci_unmap_addr(PTR, ADDR_NAME) (0)
-+#define pci_unmap_addr_set(PTR, ADDR_NAME, VAL) do { } while (0)
-+#define pci_unmap_len(PTR, LEN_NAME) (0)
-+#define pci_unmap_len_set(PTR, LEN_NAME, VAL) do { } while (0)
++ XMMS_SAVE;
+
-+#endif
++ __asm__ __volatile__ (
++#undef BLOCK
++#define BLOCK(i) \
++ PF1(i) \
++ PF1(i+2) \
++ LD(i,0) \
++ LD(i+1,1) \
++ LD(i+2,2) \
++ LD(i+3,3) \
++ PF2(i) \
++ PF2(i+2) \
++ XO1(i,0) \
++ XO1(i+1,1) \
++ XO1(i+2,2) \
++ XO1(i+3,3) \
++ PF3(i) \
++ PF3(i+2) \
++ PF0(i+4) \
++ PF0(i+6) \
++ XO2(i,0) \
++ XO2(i+1,1) \
++ XO2(i+2,2) \
++ XO2(i+3,3) \
++ XO3(i,0) \
++ XO3(i+1,1) \
++ XO3(i+2,2) \
++ XO3(i+3,3) \
++ ST(i,0) \
++ ST(i+1,1) \
++ ST(i+2,2) \
++ ST(i+3,3) \
+
-+#include <asm-generic/pci-dma-compat.h>
+
-+static inline dma64_addr_t
-+pci_dac_page_to_dma(struct pci_dev *pdev, struct page *page, unsigned long offset, int direction)
-+{
-+ return ((dma64_addr_t) page_to_phys(page) +
-+ (dma64_addr_t) offset);
-+}
++ PF0(0)
++ PF0(2)
+
-+static inline struct page *
-+pci_dac_dma_to_page(struct pci_dev *pdev, dma64_addr_t dma_addr)
-+{
-+ return virt_to_page(__va(dma_addr));
-+}
++ " .align 32 ;\n"
++ " 1: ;\n"
+
-+static inline unsigned long
-+pci_dac_dma_to_offset(struct pci_dev *pdev, dma64_addr_t dma_addr)
-+{
-+ return (dma_addr & ~PAGE_MASK);
-+}
++ BLOCK(0)
++ BLOCK(4)
++ BLOCK(8)
++ BLOCK(12)
+
-+static inline void
-+pci_dac_dma_sync_single_for_cpu(struct pci_dev *pdev, dma64_addr_t dma_addr, size_t len, int direction)
-+{
-+}
++ " addq %[inc], %[p1] ;\n"
++ " addq %[inc], %[p2] ;\n"
++ " addq %[inc], %[p3] ;\n"
++ " addq %[inc], %[p4] ;\n"
++ " decl %[cnt] ; jnz 1b"
++ : [cnt] "+c" (lines),
++ [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3), [p4] "+r" (p4)
++ : [inc] "r" (256UL)
++ : "memory" );
+
-+static inline void
-+pci_dac_dma_sync_single_for_device(struct pci_dev *pdev, dma64_addr_t dma_addr, size_t len, int direction)
-+{
-+ flush_write_buffers();
++ XMMS_RESTORE;
+}
+
-+#ifdef CONFIG_PCI
-+static inline void pci_dma_burst_advice(struct pci_dev *pdev,
-+ enum pci_dma_burst_strategy *strat,
-+ unsigned long *strategy_parameter)
++static void
++xor_sse_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
++ unsigned long *p3, unsigned long *p4, unsigned long *p5)
+{
-+ *strat = PCI_DMA_BURST_INFINITY;
-+ *strategy_parameter = ~0UL;
-+}
-+#endif
-+
-+#define HAVE_PCI_MMAP
-+extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma,
-+ enum pci_mmap_state mmap_state, int write_combine);
++ unsigned int lines = bytes >> 8;
++ xmm_store_t xmm_save[4];
++ unsigned long cr0;
+
-+static inline void pcibios_add_platform_entries(struct pci_dev *dev)
-+{
-+}
++ XMMS_SAVE;
+
-+#endif /* __KERNEL__ */
++ __asm__ __volatile__ (
++#undef BLOCK
++#define BLOCK(i) \
++ PF1(i) \
++ PF1(i+2) \
++ LD(i,0) \
++ LD(i+1,1) \
++ LD(i+2,2) \
++ LD(i+3,3) \
++ PF2(i) \
++ PF2(i+2) \
++ XO1(i,0) \
++ XO1(i+1,1) \
++ XO1(i+2,2) \
++ XO1(i+3,3) \
++ PF3(i) \
++ PF3(i+2) \
++ XO2(i,0) \
++ XO2(i+1,1) \
++ XO2(i+2,2) \
++ XO2(i+3,3) \
++ PF4(i) \
++ PF4(i+2) \
++ PF0(i+4) \
++ PF0(i+6) \
++ XO3(i,0) \
++ XO3(i+1,1) \
++ XO3(i+2,2) \
++ XO3(i+3,3) \
++ XO4(i,0) \
++ XO4(i+1,1) \
++ XO4(i+2,2) \
++ XO4(i+3,3) \
++ ST(i,0) \
++ ST(i+1,1) \
++ ST(i+2,2) \
++ ST(i+3,3) \
+
-+/* generic pci stuff */
-+#ifdef CONFIG_PCI
-+#include <asm-generic/pci.h>
-+#endif
+
-+#endif /* __x8664_PCI_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/pgalloc.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/pgalloc.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/pgalloc.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/pgalloc.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,204 @@
-+#ifndef _X86_64_PGALLOC_H
-+#define _X86_64_PGALLOC_H
++ PF0(0)
++ PF0(2)
+
-+#include <asm/fixmap.h>
-+#include <asm/pda.h>
-+#include <linux/threads.h>
-+#include <linux/mm.h>
-+#include <asm/io.h> /* for phys_to_virt and page_to_pseudophys */
++ " .align 32 ;\n"
++ " 1: ;\n"
+
-+#include <xen/features.h>
-+void make_page_readonly(void *va, unsigned int feature);
-+void make_page_writable(void *va, unsigned int feature);
-+void make_pages_readonly(void *va, unsigned int nr, unsigned int feature);
-+void make_pages_writable(void *va, unsigned int nr, unsigned int feature);
++ BLOCK(0)
++ BLOCK(4)
++ BLOCK(8)
++ BLOCK(12)
+
-+#define __user_pgd(pgd) ((pgd) + PTRS_PER_PGD)
++ " addq %[inc], %[p1] ;\n"
++ " addq %[inc], %[p2] ;\n"
++ " addq %[inc], %[p3] ;\n"
++ " addq %[inc], %[p4] ;\n"
++ " addq %[inc], %[p5] ;\n"
++ " decl %[cnt] ; jnz 1b"
++ : [cnt] "+c" (lines),
++ [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3), [p4] "+r" (p4),
++ [p5] "+r" (p5)
++ : [inc] "r" (256UL)
++ : "memory");
+
-+static inline void pmd_populate_kernel(struct mm_struct *mm, pmd_t *pmd, pte_t *pte)
-+{
-+ set_pmd(pmd, __pmd(_PAGE_TABLE | __pa(pte)));
++ XMMS_RESTORE;
+}
+
-+static inline void pmd_populate(struct mm_struct *mm, pmd_t *pmd, struct page *pte)
-+{
-+ if (unlikely((mm)->context.pinned)) {
-+ BUG_ON(HYPERVISOR_update_va_mapping(
-+ (unsigned long)__va(page_to_pfn(pte) << PAGE_SHIFT),
-+ pfn_pte(page_to_pfn(pte), PAGE_KERNEL_RO), 0));
-+ set_pmd(pmd, __pmd(_PAGE_TABLE | (page_to_pfn(pte) << PAGE_SHIFT)));
-+ } else {
-+ *(pmd) = __pmd(_PAGE_TABLE | (page_to_pfn(pte) << PAGE_SHIFT));
-+ }
-+}
++static struct xor_block_template xor_block_sse = {
++ .name = "generic_sse",
++ .do_2 = xor_sse_2,
++ .do_3 = xor_sse_3,
++ .do_4 = xor_sse_4,
++ .do_5 = xor_sse_5,
++};
+
-+static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd)
-+{
-+ if (unlikely((mm)->context.pinned)) {
-+ BUG_ON(HYPERVISOR_update_va_mapping(
-+ (unsigned long)pmd,
-+ pfn_pte(virt_to_phys(pmd)>>PAGE_SHIFT,
-+ PAGE_KERNEL_RO), 0));
-+ set_pud(pud, __pud(_PAGE_TABLE | __pa(pmd)));
-+ } else {
-+ *(pud) = __pud(_PAGE_TABLE | __pa(pmd));
-+ }
-+}
++#undef XOR_TRY_TEMPLATES
++#define XOR_TRY_TEMPLATES \
++ do { \
++ xor_speed(&xor_block_sse); \
++ } while (0)
+
++/* We force the use of the SSE xor block because it can write around L2.
++ We may also be able to load into the L1 only depending on how the cpu
++ deals with a load to a line that is being prefetched. */
++#define XOR_SELECT_TEMPLATE(FASTEST) (&xor_block_sse)
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/irq_vectors.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/irq_vectors.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,123 @@
+/*
-+ * We need to use the batch mode here, but pgd_pupulate() won't be
-+ * be called frequently.
-+ */
-+static inline void pgd_populate(struct mm_struct *mm, pgd_t *pgd, pud_t *pud)
-+{
-+ if (unlikely((mm)->context.pinned)) {
-+ BUG_ON(HYPERVISOR_update_va_mapping(
-+ (unsigned long)pud,
-+ pfn_pte(virt_to_phys(pud)>>PAGE_SHIFT,
-+ PAGE_KERNEL_RO), 0));
-+ set_pgd(pgd, __pgd(_PAGE_TABLE | __pa(pud)));
-+ set_pgd(__user_pgd(pgd), __pgd(_PAGE_TABLE | __pa(pud)));
-+ } else {
-+ *(pgd) = __pgd(_PAGE_TABLE | __pa(pud));
-+ *(__user_pgd(pgd)) = *(pgd);
-+ }
-+}
++ * This file should contain #defines for all of the interrupt vector
++ * numbers used by this architecture.
++ *
++ * In addition, there are some standard defines:
++ *
++ * FIRST_EXTERNAL_VECTOR:
++ * The first free place for external interrupts
++ *
++ * SYSCALL_VECTOR:
++ * The IRQ vector a syscall makes the user to kernel transition
++ * under.
++ *
++ * TIMER_IRQ:
++ * The IRQ number the timer interrupt comes in at.
++ *
++ * NR_IRQS:
++ * The total number of interrupt vectors (including all the
++ * architecture specific interrupts) needed.
++ *
++ */
++#ifndef _ASM_IRQ_VECTORS_H
++#define _ASM_IRQ_VECTORS_H
+
-+extern struct page *pte_alloc_one(struct mm_struct *mm, unsigned long addr);
-+extern void pte_free(struct page *pte);
++/*
++ * IDT vectors usable for external interrupt sources start
++ * at 0x20:
++ */
++#define FIRST_EXTERNAL_VECTOR 0x20
+
-+static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long addr)
-+{
-+ struct page *pg;
++#define SYSCALL_VECTOR 0x80
+
-+ pg = pte_alloc_one(mm, addr);
-+ return pg ? page_address(pg) : NULL;
-+}
++/*
++ * Vectors 0x20-0x2f are used for ISA interrupts.
++ */
+
-+static inline void pmd_free(pmd_t *pmd)
-+{
-+ BUG_ON((unsigned long)pmd & (PAGE_SIZE-1));
-+ pte_free(virt_to_page(pmd));
-+}
++#if 0
++/*
++ * Special IRQ vectors used by the SMP architecture, 0xf0-0xff
++ *
++ * some of the following vectors are 'rare', they are merged
++ * into a single vector (CALL_FUNCTION_VECTOR) to save vector space.
++ * TLB, reschedule and local APIC vectors are performance-critical.
++ *
++ * Vectors 0xf0-0xfa are free (reserved for future Linux use).
++ */
++#define INVALIDATE_TLB_VECTOR 0xfd
++#define RESCHEDULE_VECTOR 0xfc
++#define CALL_FUNCTION_VECTOR 0xfb
+
-+static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long addr)
-+{
-+ struct page *pg;
++#define THERMAL_APIC_VECTOR 0xf0
++/*
++ * Local APIC timer IRQ vector is on a different priority level,
++ * to work around the 'lost local interrupt if more than 2 IRQ
++ * sources per level' errata.
++ */
++#define LOCAL_TIMER_VECTOR 0xef
++#endif
+
-+ pg = pte_alloc_one(mm, addr);
-+ return pg ? page_address(pg) : NULL;
-+}
++#define SPURIOUS_APIC_VECTOR 0xff
++#define ERROR_APIC_VECTOR 0xfe
+
-+static inline void pud_free(pud_t *pud)
-+{
-+ BUG_ON((unsigned long)pud & (PAGE_SIZE-1));
-+ pte_free(virt_to_page(pud));
-+}
++/*
++ * First APIC vector available to drivers: (vectors 0x30-0xee)
++ * we start at 0x31 to spread out vectors evenly between priority
++ * levels. (0x80 is the syscall vector)
++ */
++#define FIRST_DEVICE_VECTOR 0x31
++#define FIRST_SYSTEM_VECTOR 0xef
+
-+static inline void pgd_list_add(pgd_t *pgd)
-+{
-+ struct page *page = virt_to_page(pgd);
++/*
++ * 16 8259A IRQ's, 208 potential APIC interrupt sources.
++ * Right now the APIC is mostly only used for SMP.
++ * 256 vectors is an architectural limit. (we can have
++ * more than 256 devices theoretically, but they will
++ * have to use shared interrupts)
++ * Since vectors 0x00-0x1f are used/reserved for the CPU,
++ * the usable vector space is 0x20-0xff (224 vectors)
++ */
+
-+ spin_lock(&pgd_lock);
-+ page->index = (pgoff_t)pgd_list;
-+ if (pgd_list)
-+ pgd_list->private = (unsigned long)&page->index;
-+ pgd_list = page;
-+ page->private = (unsigned long)&pgd_list;
-+ spin_unlock(&pgd_lock);
-+}
++#define RESCHEDULE_VECTOR 0
++#define CALL_FUNCTION_VECTOR 1
++#define NR_IPIS 2
+
-+static inline void pgd_list_del(pgd_t *pgd)
-+{
-+ struct page *next, **pprev, *page = virt_to_page(pgd);
++/*
++ * The maximum number of vectors supported by i386 processors
++ * is limited to 256. For processors other than i386, NR_VECTORS
++ * should be changed accordingly.
++ */
++#define NR_VECTORS 256
+
-+ spin_lock(&pgd_lock);
-+ next = (struct page *)page->index;
-+ pprev = (struct page **)page->private;
-+ *pprev = next;
-+ if (next)
-+ next->private = (unsigned long)pprev;
-+ spin_unlock(&pgd_lock);
-+}
++#define FPU_IRQ 13
+
-+static inline pgd_t *pgd_alloc(struct mm_struct *mm)
-+{
-+ /*
-+ * We allocate two contiguous pages for kernel and user.
-+ */
-+ unsigned boundary;
-+ pgd_t *pgd = (pgd_t *)__get_free_pages(GFP_KERNEL|__GFP_REPEAT, 1);
-+ if (!pgd)
-+ return NULL;
-+ pgd_list_add(pgd);
-+ /*
-+ * Copy kernel pointers in from init.
-+ * Could keep a freelist or slab cache of those because the kernel
-+ * part never changes.
-+ */
-+ boundary = pgd_index(__PAGE_OFFSET);
-+ memset(pgd, 0, boundary * sizeof(pgd_t));
-+ memcpy(pgd + boundary,
-+ init_level4_pgt + boundary,
-+ (PTRS_PER_PGD - boundary) * sizeof(pgd_t));
++#define FIRST_VM86_IRQ 3
++#define LAST_VM86_IRQ 15
++#define invalid_vm86_irq(irq) ((irq) < 3 || (irq) > 15)
+
-+ memset(__user_pgd(pgd), 0, PAGE_SIZE); /* clean up user pgd */
-+ /*
-+ * Set level3_user_pgt for vsyscall area
-+ */
-+ set_pgd(__user_pgd(pgd) + pgd_index(VSYSCALL_START),
-+ __pgd(__pa_symbol(level3_user_pgt) | _PAGE_TABLE));
-+ return pgd;
-+}
++/*
++ * The flat IRQ space is divided into two regions:
++ * 1. A one-to-one mapping of real physical IRQs. This space is only used
++ * if we have physical device-access privilege. This region is at the
++ * start of the IRQ space so that existing device drivers do not need
++ * to be modified to translate physical IRQ numbers into our IRQ space.
++ * 3. A dynamic mapping of inter-domain and Xen-sourced virtual IRQs. These
++ * are bound using the provided bind/unbind functions.
++ */
+
-+static inline void pgd_free(pgd_t *pgd)
-+{
-+ pte_t *ptep = virt_to_ptep(pgd);
++#define PIRQ_BASE 0
++#define NR_PIRQS 256
+
-+ if (!pte_write(*ptep)) {
-+ xen_pgd_unpin(__pa(pgd));
-+ BUG_ON(HYPERVISOR_update_va_mapping(
-+ (unsigned long)pgd,
-+ pfn_pte(virt_to_phys(pgd)>>PAGE_SHIFT, PAGE_KERNEL),
-+ 0));
-+ }
++#define DYNIRQ_BASE (PIRQ_BASE + NR_PIRQS)
++#define NR_DYNIRQS 256
+
-+ ptep = virt_to_ptep(__user_pgd(pgd));
++#define NR_IRQS (NR_PIRQS + NR_DYNIRQS)
++#define NR_IRQ_VECTORS NR_IRQS
+
-+ if (!pte_write(*ptep)) {
-+ xen_pgd_unpin(__pa(__user_pgd(pgd)));
-+ BUG_ON(HYPERVISOR_update_va_mapping(
-+ (unsigned long)__user_pgd(pgd),
-+ pfn_pte(virt_to_phys(__user_pgd(pgd))>>PAGE_SHIFT,
-+ PAGE_KERNEL),
-+ 0));
-+ }
++#define pirq_to_irq(_x) ((_x) + PIRQ_BASE)
++#define irq_to_pirq(_x) ((_x) - PIRQ_BASE)
+
-+ pgd_list_del(pgd);
-+ free_pages((unsigned long)pgd, 1);
-+}
++#define dynirq_to_irq(_x) ((_x) + DYNIRQ_BASE)
++#define irq_to_dynirq(_x) ((_x) - DYNIRQ_BASE)
+
-+static inline pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
-+{
-+ pte_t *pte = (pte_t *)get_zeroed_page(GFP_KERNEL|__GFP_REPEAT);
-+ if (pte)
-+ make_page_readonly(pte, XENFEAT_writable_page_tables);
++#endif /* _ASM_IRQ_VECTORS_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/mach_time.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/mach_time.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,111 @@
++/*
++ * include/asm-i386/mach-default/mach_time.h
++ *
++ * Machine specific set RTC function for generic.
++ * Split out from time.c by Osamu Tomita <tomita at cinet.co.jp>
++ */
++#ifndef _MACH_TIME_H
++#define _MACH_TIME_H
+
-+ return pte;
-+}
++#include <asm-i386/mc146818rtc.h>
+
-+/* Should really implement gc for free page table pages. This could be
-+ done with a reference count in struct page. */
++/* for check timing call set_rtc_mmss() 500ms */
++/* used in arch/i386/time.c::do_timer_interrupt() */
++#define USEC_AFTER 500000
++#define USEC_BEFORE 500000
+
-+static inline void pte_free_kernel(pte_t *pte)
++/*
++ * In order to set the CMOS clock precisely, set_rtc_mmss has to be
++ * called 500 ms after the second nowtime has started, because when
++ * nowtime is written into the registers of the CMOS clock, it will
++ * jump to the next second precisely 500 ms later. Check the Motorola
++ * MC146818A or Dallas DS12887 data sheet for details.
++ *
++ * BUG: This routine does not handle hour overflow properly; it just
++ * sets the minutes. Usually you'll only notice that after reboot!
++ */
++static inline int mach_set_rtc_mmss(unsigned long nowtime)
+{
-+ BUG_ON((unsigned long)pte & (PAGE_SIZE-1));
-+ make_page_writable(pte, XENFEAT_writable_page_tables);
-+ free_page((unsigned long)pte);
-+}
-+
-+#define __pte_free_tlb(tlb,pte) tlb_remove_page((tlb),(pte))
-+#define __pmd_free_tlb(tlb,x) tlb_remove_page((tlb),virt_to_page(x))
-+#define __pud_free_tlb(tlb,x) tlb_remove_page((tlb),virt_to_page(x))
++ int retval = 0;
++ int real_seconds, real_minutes, cmos_minutes;
++ unsigned char save_control, save_freq_select;
+
-+#endif /* _X86_64_PGALLOC_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/pgtable.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/pgtable.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/pgtable.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/pgtable.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,575 @@
-+#ifndef _X86_64_PGTABLE_H
-+#define _X86_64_PGTABLE_H
++ save_control = CMOS_READ(RTC_CONTROL); /* tell the clock it's being set */
++ CMOS_WRITE((save_control|RTC_SET), RTC_CONTROL);
+
-+/*
-+ * This file contains the functions and defines necessary to modify and use
-+ * the x86-64 page table tree.
-+ */
-+#include <asm/processor.h>
-+#include <asm/fixmap.h>
-+#include <asm/bitops.h>
-+#include <linux/threads.h>
-+#include <linux/sched.h>
-+#include <asm/pda.h>
-+#ifdef CONFIG_XEN
-+#include <asm/hypervisor.h>
++ save_freq_select = CMOS_READ(RTC_FREQ_SELECT); /* stop and reset prescaler */
++ CMOS_WRITE((save_freq_select|RTC_DIV_RESET2), RTC_FREQ_SELECT);
+
-+extern pud_t level3_user_pgt[512];
-+extern pud_t init_level4_user_pgt[];
++ cmos_minutes = CMOS_READ(RTC_MINUTES);
++ if (!(save_control & RTC_DM_BINARY) || RTC_ALWAYS_BCD)
++ BCD_TO_BIN(cmos_minutes);
+
-+extern void xen_init_pt(void);
++ /*
++ * since we're only adjusting minutes and seconds,
++ * don't interfere with hour overflow. This avoids
++ * messing with unknown time zones but requires your
++ * RTC not to be off by more than 15 minutes
++ */
++ real_seconds = nowtime % 60;
++ real_minutes = nowtime / 60;
++ if (((abs(real_minutes - cmos_minutes) + 15)/30) & 1)
++ real_minutes += 30; /* correct for half hour time zone */
++ real_minutes %= 60;
+
-+#define virt_to_ptep(__va) \
-+({ \
-+ pgd_t *__pgd = pgd_offset_k((unsigned long)(__va)); \
-+ pud_t *__pud = pud_offset(__pgd, (unsigned long)(__va)); \
-+ pmd_t *__pmd = pmd_offset(__pud, (unsigned long)(__va)); \
-+ pte_offset_kernel(__pmd, (unsigned long)(__va)); \
-+})
++ if (abs(real_minutes - cmos_minutes) < 30) {
++ if (!(save_control & RTC_DM_BINARY) || RTC_ALWAYS_BCD) {
++ BIN_TO_BCD(real_seconds);
++ BIN_TO_BCD(real_minutes);
++ }
++ CMOS_WRITE(real_seconds,RTC_SECONDS);
++ CMOS_WRITE(real_minutes,RTC_MINUTES);
++ } else {
++ printk(KERN_WARNING
++ "set_rtc_mmss: can't update from %d to %d\n",
++ cmos_minutes, real_minutes);
++ retval = -1;
++ }
+
-+#define arbitrary_virt_to_machine(__va) \
-+({ \
-+ maddr_t m = (maddr_t)pte_mfn(*virt_to_ptep(__va)) << PAGE_SHIFT;\
-+ m | ((unsigned long)(__va) & (PAGE_SIZE-1)); \
-+})
-+#endif
++ /* The following flags have to be released exactly in this order,
++ * otherwise the DS12887 (popular MC146818A clone with integrated
++ * battery and quartz) will not reset the oscillator and will not
++ * update precisely 500 ms later. You won't find this mentioned in
++ * the Dallas Semiconductor data sheets, but who believes data
++ * sheets anyway ... -- Markus Kuhn
++ */
++ CMOS_WRITE(save_control, RTC_CONTROL);
++ CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT);
+
-+extern pud_t level3_kernel_pgt[512];
-+extern pud_t level3_physmem_pgt[512];
-+extern pud_t level3_ident_pgt[512];
-+extern pmd_t level2_kernel_pgt[512];
-+extern pgd_t init_level4_pgt[];
-+extern pgd_t boot_level4_pgt[];
-+extern unsigned long __supported_pte_mask;
++ return retval;
++}
+
-+#define swapper_pg_dir init_level4_pgt
++static inline unsigned long mach_get_cmos_time(void)
++{
++ unsigned int year, mon, day, hour, min, sec;
+
-+extern int nonx_setup(char *str);
-+extern void paging_init(void);
-+extern void clear_kernel_mapping(unsigned long addr, unsigned long size);
++ do {
++ sec = CMOS_READ(RTC_SECONDS);
++ min = CMOS_READ(RTC_MINUTES);
++ hour = CMOS_READ(RTC_HOURS);
++ day = CMOS_READ(RTC_DAY_OF_MONTH);
++ mon = CMOS_READ(RTC_MONTH);
++ year = CMOS_READ(RTC_YEAR);
++ } while (sec != CMOS_READ(RTC_SECONDS));
+
-+extern unsigned long pgkern_mask;
++ if (!(CMOS_READ(RTC_CONTROL) & RTC_DM_BINARY) || RTC_ALWAYS_BCD) {
++ BCD_TO_BIN(sec);
++ BCD_TO_BIN(min);
++ BCD_TO_BIN(hour);
++ BCD_TO_BIN(day);
++ BCD_TO_BIN(mon);
++ BCD_TO_BIN(year);
++ }
+
-+/*
-+ * ZERO_PAGE is a global shared page that is always zero: used
-+ * for zero-mapped memory areas etc..
-+ */
-+extern unsigned long empty_zero_page[PAGE_SIZE/sizeof(unsigned long)];
-+#define ZERO_PAGE(vaddr) (virt_to_page(empty_zero_page))
++ year += 1900;
++ if (year < 1970)
++ year += 100;
+
-+/*
-+ * PGDIR_SHIFT determines what a top-level page table entry can map
-+ */
-+#define PGDIR_SHIFT 39
-+#define PTRS_PER_PGD 512
++ return mktime(year, mon, day, hour, min, sec);
++}
+
++#endif /* !_MACH_TIME_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/mach_timer.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/mach_timer.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,50 @@
+/*
-+ * 3rd level page
++ * include/asm-i386/mach-default/mach_timer.h
++ *
++ * Machine specific calibrate_tsc() for generic.
++ * Split out from timer_tsc.c by Osamu Tomita <tomita at cinet.co.jp>
+ */
-+#define PUD_SHIFT 30
-+#define PTRS_PER_PUD 512
-+
-+/*
-+ * PMD_SHIFT determines the size of the area a middle-level
-+ * page table can map
++/* ------ Calibrate the TSC -------
++ * Return 2^32 * (1 / (TSC clocks per usec)) for do_fast_gettimeoffset().
++ * Too much 64-bit arithmetic here to do this cleanly in C, and for
++ * accuracy's sake we want to keep the overhead on the CTC speaker (channel 2)
++ * output busy loop as low as possible. We avoid reading the CTC registers
++ * directly because of the awkward 8-bit access mechanism of the 82C54
++ * device.
+ */
-+#define PMD_SHIFT 21
-+#define PTRS_PER_PMD 512
++#ifndef _MACH_TIMER_H
++#define _MACH_TIMER_H
+
-+/*
-+ * entries per page directory level
-+ */
-+#define PTRS_PER_PTE 512
++#define CALIBRATE_TIME_MSEC 30 /* 30 msecs */
++#define CALIBRATE_LATCH \
++ ((CLOCK_TICK_RATE * CALIBRATE_TIME_MSEC + 1000/2)/1000)
+
-+#define pte_ERROR(e) \
-+ printk("%s:%d: bad pte %p(%016lx pfn %010lx).\n", __FILE__, __LINE__, \
-+ &(e), __pte_val(e), pte_pfn(e))
-+#define pmd_ERROR(e) \
-+ printk("%s:%d: bad pmd %p(%016lx pfn %010lx).\n", __FILE__, __LINE__, \
-+ &(e), __pmd_val(e), pmd_pfn(e))
-+#define pud_ERROR(e) \
-+ printk("%s:%d: bad pud %p(%016lx pfn %010lx).\n", __FILE__, __LINE__, \
-+ &(e), __pud_val(e), (pud_val(e) & __PHYSICAL_MASK) >> PAGE_SHIFT)
-+#define pgd_ERROR(e) \
-+ printk("%s:%d: bad pgd %p(%016lx pfn %010lx).\n", __FILE__, __LINE__, \
-+ &(e), __pgd_val(e), (pgd_val(e) & __PHYSICAL_MASK) >> PAGE_SHIFT)
++static inline void mach_prepare_counter(void)
++{
++ /* Set the Gate high, disable speaker */
++ outb((inb(0x61) & ~0x02) | 0x01, 0x61);
+
-+#define pgd_none(x) (!__pgd_val(x))
-+#define pud_none(x) (!__pud_val(x))
++ /*
++ * Now let's take care of CTC channel 2
++ *
++ * Set the Gate high, program CTC channel 2 for mode 0,
++ * (interrupt on terminal count mode), binary count,
++ * load 5 * LATCH count, (LSB and MSB) to begin countdown.
++ *
++ * Some devices need a delay here.
++ */
++ outb(0xb0, 0x43); /* binary, mode 0, LSB/MSB, Ch 2 */
++ outb_p(CALIBRATE_LATCH & 0xff, 0x42); /* LSB of count */
++ outb_p(CALIBRATE_LATCH >> 8, 0x42); /* MSB of count */
++}
+
-+static inline void set_pte(pte_t *dst, pte_t val)
++static inline void mach_countup(unsigned long *count_p)
+{
-+ *dst = val;
++ unsigned long count = 0;
++ do {
++ count++;
++ } while ((inb_p(0x61) & 0x20) == 0);
++ *count_p = count;
+}
+
-+#define set_pmd(pmdptr, pmdval) xen_l2_entry_update(pmdptr, (pmdval))
-+#define set_pud(pudptr, pudval) xen_l3_entry_update(pudptr, (pudval))
-+#define set_pgd(pgdptr, pgdval) xen_l4_entry_update(pgdptr, (pgdval))
++#endif /* !_MACH_TIMER_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/setup_arch_post.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/setup_arch_post.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,63 @@
++/**
++ * machine_specific_* - Hooks for machine specific setup.
++ *
++ * Description:
++ * This is included late in kernel/setup.c so that it can make
++ * use of all of the static functions.
++ **/
+
-+static inline void pud_clear (pud_t * pud)
-+{
-+ set_pud(pud, __pud(0));
-+}
++#include <xen/interface/callback.h>
+
-+#define __user_pgd(pgd) ((pgd) + PTRS_PER_PGD)
++extern void hypervisor_callback(void);
++extern void failsafe_callback(void);
++extern void nmi(void);
+
-+static inline void pgd_clear (pgd_t * pgd)
++static void __init machine_specific_arch_setup(void)
+{
-+ set_pgd(pgd, __pgd(0));
-+ set_pgd(__user_pgd(pgd), __pgd(0));
-+}
++ int ret;
++ static struct callback_register __initdata event = {
++ .type = CALLBACKTYPE_event,
++ .address = (unsigned long) hypervisor_callback,
++ };
++ static struct callback_register __initdata failsafe = {
++ .type = CALLBACKTYPE_failsafe,
++ .address = (unsigned long)failsafe_callback,
++ };
++ static struct callback_register __initdata syscall = {
++ .type = CALLBACKTYPE_syscall,
++ .address = (unsigned long)system_call,
++ };
++#ifdef CONFIG_X86_LOCAL_APIC
++ static struct callback_register __initdata nmi_cb = {
++ .type = CALLBACKTYPE_nmi,
++ .address = (unsigned long)nmi,
++ };
++#endif
+
-+#define pud_page(pud) \
-+ ((unsigned long) __va(pud_val(pud) & PHYSICAL_PAGE_MASK))
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &event);
++ if (ret == 0)
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &failsafe);
++ if (ret == 0)
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &syscall);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (ret == -ENOSYS)
++ ret = HYPERVISOR_set_callbacks(
++ event.address,
++ failsafe.address,
++ syscall.address);
++#endif
++ BUG_ON(ret);
+
-+#define pte_same(a, b) ((a).pte == (b).pte)
++#ifdef CONFIG_X86_LOCAL_APIC
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &nmi_cb);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (ret == -ENOSYS) {
++ static struct xennmi_callback __initdata cb = {
++ .handler_address = (unsigned long)nmi
++ };
+
-+#define pte_pgprot(a) (__pgprot((a).pte & ~PHYSICAL_PAGE_MASK))
++ HYPERVISOR_nmi_op(XENNMI_register_callback, &cb);
++ }
++#endif
++#endif
++}
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/mach-xen/setup_arch_pre.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/asm-x86_64/mach-xen/setup_arch_pre.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,5 @@
++/* Hook to call BIOS initialisation function */
+
-+#define PMD_SIZE (1UL << PMD_SHIFT)
-+#define PMD_MASK (~(PMD_SIZE-1))
-+#define PUD_SIZE (1UL << PUD_SHIFT)
-+#define PUD_MASK (~(PUD_SIZE-1))
-+#define PGDIR_SIZE (1UL << PGDIR_SHIFT)
-+#define PGDIR_MASK (~(PGDIR_SIZE-1))
++#define ARCH_SETUP machine_specific_arch_setup();
+
-+#define USER_PTRS_PER_PGD ((TASK_SIZE-1)/PGDIR_SIZE+1)
-+#define FIRST_USER_ADDRESS 0
++static void __init machine_specific_arch_setup(void);
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/msi.h
+--- a/include/asm-x86_64/msi.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-x86_64/msi.h Wed Sep 10 10:54:08 2008 +0100
+@@ -7,14 +7,21 @@
+ #define ASM_MSI_H
+
+ #include <asm/desc.h>
++#ifndef CONFIG_XEN
+ #include <asm/mach_apic.h>
++#endif
+ #include <asm/smp.h>
+
++#ifndef CONFIG_XEN
+ #define LAST_DEVICE_VECTOR (FIRST_SYSTEM_VECTOR - 1)
++#else
++#define LAST_DYNAMIC_VECTOR 0xdf
++#define LAST_DEVICE_VECTOR (LAST_DYNAMIC_VECTOR)
++#endif
+
-+#ifndef __ASSEMBLY__
-+#define MAXMEM 0x3fffffffffffUL
-+#define VMALLOC_START 0xffffc20000000000UL
-+#define VMALLOC_END 0xffffe1ffffffffffUL
-+#define MODULES_VADDR 0xffffffff88000000UL
-+#define MODULES_END 0xfffffffffff00000UL
-+#define MODULES_LEN (MODULES_END - MODULES_VADDR)
+ #define MSI_TARGET_CPU_SHIFT 12
+
+ extern struct msi_ops msi_apic_ops;
+-
+ static inline int msi_arch_init(void)
+ {
+ msi_register(&msi_apic_ops);
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/signal.h
+--- a/include/asm-x86_64/signal.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-x86_64/signal.h Wed Sep 10 10:54:08 2008 +0100
+@@ -22,10 +22,6 @@
+ typedef struct {
+ unsigned long sig[_NSIG_WORDS];
+ } sigset_t;
+-
+-
+-struct pt_regs;
+-asmlinkage int do_signal(struct pt_regs *regs, sigset_t *oldset);
+
+
+ #else
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/thread_info.h
+--- a/include/asm-x86_64/thread_info.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-x86_64/thread_info.h Wed Sep 10 10:54:08 2008 +0100
+@@ -114,6 +114,7 @@
+ #define TIF_IRET 5 /* force IRET */
+ #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */
+ #define TIF_SECCOMP 8 /* secure computing */
++#define TIF_RESTORE_SIGMASK 9 /* restore signal mask in do_signal */
+ /* 16 free */
+ #define TIF_IA32 17 /* 32bit process */
+ #define TIF_FORK 18 /* ret_from_fork */
+@@ -128,6 +129,7 @@
+ #define _TIF_IRET (1<<TIF_IRET)
+ #define _TIF_SYSCALL_AUDIT (1<<TIF_SYSCALL_AUDIT)
+ #define _TIF_SECCOMP (1<<TIF_SECCOMP)
++#define _TIF_RESTORE_SIGMASK (1<<TIF_RESTORE_SIGMASK)
+ #define _TIF_IA32 (1<<TIF_IA32)
+ #define _TIF_FORK (1<<TIF_FORK)
+ #define _TIF_ABI_PENDING (1<<TIF_ABI_PENDING)
+diff -r d894e36cfc30 -r 0aa021803deb include/asm-x86_64/unistd.h
+--- a/include/asm-x86_64/unistd.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/asm-x86_64/unistd.h Wed Sep 10 10:54:08 2008 +0100
+@@ -600,9 +600,9 @@
+ #define __NR_faccessat 269
+ __SYSCALL(__NR_faccessat, sys_faccessat)
+ #define __NR_pselect6 270
+-__SYSCALL(__NR_pselect6, sys_ni_syscall) /* for now */
++__SYSCALL(__NR_pselect6, sys_pselect6)
+ #define __NR_ppoll 271
+-__SYSCALL(__NR_ppoll, sys_ni_syscall) /* for now */
++__SYSCALL(__NR_ppoll, sys_ppoll)
+ #define __NR_unshare 272
+ __SYSCALL(__NR_unshare, sys_unshare)
+ #define __NR_set_robust_list 273
+@@ -658,6 +658,7 @@
+ #define __ARCH_WANT_SYS_SIGPENDING
+ #define __ARCH_WANT_SYS_SIGPROCMASK
+ #define __ARCH_WANT_SYS_RT_SIGACTION
++#define __ARCH_WANT_SYS_RT_SIGSUSPEND
+ #define __ARCH_WANT_SYS_TIME
+ #define __ARCH_WANT_COMPAT_SYS_TIME
+
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/aio.h
+--- a/include/linux/aio.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/aio.h Wed Sep 10 10:54:08 2008 +0100
+@@ -191,6 +191,11 @@
+ struct aio_ring_info ring_info;
+
+ struct work_struct wq;
++#ifdef CONFIG_EPOLL
++ // poll integration
++ wait_queue_head_t poll_wait;
++ struct file *file;
++#endif
+ };
+
+ /* prototypes */
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/bio.h
+--- a/include/linux/bio.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/bio.h Wed Sep 10 10:54:08 2008 +0100
+@@ -172,12 +172,27 @@
+ #define bio_offset(bio) bio_iovec((bio))->bv_offset
+ #define bio_segments(bio) ((bio)->bi_vcnt - (bio)->bi_idx)
+ #define bio_sectors(bio) ((bio)->bi_size >> 9)
+-#define bio_cur_sectors(bio) (bio_iovec(bio)->bv_len >> 9)
+-#define bio_data(bio) (page_address(bio_page((bio))) + bio_offset((bio)))
+ #define bio_barrier(bio) ((bio)->bi_rw & (1 << BIO_RW_BARRIER))
+ #define bio_sync(bio) ((bio)->bi_rw & (1 << BIO_RW_SYNC))
+ #define bio_failfast(bio) ((bio)->bi_rw & (1 << BIO_RW_FAILFAST))
+ #define bio_rw_ahead(bio) ((bio)->bi_rw & (1 << BIO_RW_AHEAD))
++#define bio_empty_barrier(bio) (bio_barrier(bio) && !(bio)->bi_size)
+
-+#define _PAGE_BIT_PRESENT 0
-+#define _PAGE_BIT_RW 1
-+#define _PAGE_BIT_USER 2
-+#define _PAGE_BIT_PWT 3
-+#define _PAGE_BIT_PCD 4
-+#define _PAGE_BIT_ACCESSED 5
-+#define _PAGE_BIT_DIRTY 6
-+#define _PAGE_BIT_PSE 7 /* 4 MB (or 2MB) page */
-+#define _PAGE_BIT_GLOBAL 8 /* Global TLB entry PPro+ */
-+#define _PAGE_BIT_NX 63 /* No execute: only valid after cpuid check */
++static inline unsigned int bio_cur_sectors(struct bio *bio)
++{
++ if (bio->bi_vcnt)
++ return bio_iovec(bio)->bv_len >> 9;
+
-+#define _PAGE_PRESENT 0x001
-+#define _PAGE_RW 0x002
-+#define _PAGE_USER 0x004
-+#define _PAGE_PWT 0x008
-+#define _PAGE_PCD 0x010
-+#define _PAGE_ACCESSED 0x020
-+#define _PAGE_DIRTY 0x040
-+#define _PAGE_PSE 0x080 /* 2MB page */
-+#define _PAGE_FILE 0x040 /* nonlinear file mapping, saved PTE; unset:swap */
-+#define _PAGE_GLOBAL 0x100 /* Global TLB entry */
++ return 0;
++}
+
-+#define _PAGE_PROTNONE 0x080 /* If not present */
-+#define _PAGE_NX (1UL<<_PAGE_BIT_NX)
++static inline void *bio_data(struct bio *bio)
++{
++ if (bio->bi_vcnt)
++ return page_address(bio_page(bio)) + bio_offset(bio);
+
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+extern unsigned int __kernel_page_user;
-+#else
-+#define __kernel_page_user 0
++ return NULL;
++}
+
+ /*
+ * will die
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/blkdev.h
+--- a/include/linux/blkdev.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/blkdev.h Wed Sep 10 10:54:08 2008 +0100
+@@ -506,6 +506,8 @@
+ #define blk_barrier_rq(rq) ((rq)->flags & REQ_HARDBARRIER)
+ #define blk_fua_rq(rq) ((rq)->flags & REQ_FUA)
+
++#define blk_empty_barrier(rq) (blk_barrier_rq(rq) && blk_fs_request(rq) && !(rq)->hard_nr_sectors)
++
+ #define list_entry_rq(ptr) list_entry((ptr), struct request, queuelist)
+
+ #define rq_data_dir(rq) ((rq)->flags & 1)
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/crash_dump.h
+--- a/include/linux/crash_dump.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/crash_dump.h Wed Sep 10 10:54:08 2008 +0100
+@@ -14,5 +14,13 @@
+ extern const struct file_operations proc_vmcore_operations;
+ extern struct proc_dir_entry *proc_vmcore;
+
++/* Architecture code defines this if there are other possible ELF
++ * machine types, e.g. on bi-arch capable hardware. */
++#ifndef vmcore_elf_check_arch_cross
++#define vmcore_elf_check_arch_cross(x) 0
+#endif
+
-+#define _PAGE_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED | _PAGE_DIRTY)
-+#define _KERNPG_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY | __kernel_page_user)
++#define vmcore_elf_check_arch(x) (elf_check_arch(x) || vmcore_elf_check_arch_cross(x))
++
+ #endif /* CONFIG_CRASH_DUMP */
+ #endif /* LINUX_CRASHDUMP_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/elfnote.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/linux/elfnote.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,104 @@
++#ifndef _LINUX_ELFNOTE_H
++#define _LINUX_ELFNOTE_H
++/*
++ * Helper macros to generate ELF Note structures, which are put into a
++ * PT_NOTE segment of the final vmlinux image. These are useful for
++ * including name-value pairs of metadata into the kernel binary (or
++ * modules?) for use by external programs.
++ *
++ * Each note has three parts: a name, a type and a desc. The name is
++ * intended to distinguish the note's originator, so it would be a
++ * company, project, subsystem, etc; it must be in a suitable form for
++ * use in a section name. The type is an integer which is used to tag
++ * the data, and is considered to be within the "name" namespace (so
++ * "FooCo"'s type 42 is distinct from "BarProj"'s type 42). The
++ * "desc" field is the actual data. There are no constraints on the
++ * desc field's contents, though typically they're fairly small.
++ *
++ * All notes from a given NAME are put into a section named
++ * .note.NAME. When the kernel image is finally linked, all the notes
++ * are packed into a single .notes section, which is mapped into the
++ * PT_NOTE segment. Because notes for a given name are grouped into
++ * the same section, they'll all be adjacent the output file.
++ *
++ * This file defines macros for both C and assembler use. Their
++ * syntax is slightly different, but they're semantically similar.
++ *
++ * See the ELF specification for more detail about ELF notes.
++ */
+
-+#define _PAGE_CHG_MASK (PTE_MASK | _PAGE_ACCESSED | _PAGE_DIRTY)
++#ifdef __ASSEMBLER__
++/*
++ * Generate a structure with the same shape as Elf{32,64}_Nhdr (which
++ * turn out to be the same size and shape), followed by the name and
++ * desc data with appropriate padding. The 'desctype' argument is the
++ * assembler pseudo op defining the type of the data e.g. .asciz while
++ * 'descdata' is the data itself e.g. "hello, world".
++ *
++ * e.g. ELFNOTE(XYZCo, 42, .asciz, "forty-two")
++ * ELFNOTE(XYZCo, 12, .long, 0xdeadbeef)
++ */
++#ifdef __STDC__
++#define ELFNOTE(name, type, desctype, descdata...) \
++.pushsection .note.name ; \
++ .align 4 ; \
++ .long 2f - 1f /* namesz */ ; \
++ .long 4f - 3f /* descsz */ ; \
++ .long type ; \
++1:.asciz #name ; \
++2:.align 4 ; \
++3:desctype descdata ; \
++4:.align 4 ; \
++.popsection
++#else /* !__STDC__, i.e. -traditional */
++#define ELFNOTE(name, type, desctype, descdata) \
++.pushsection .note.name ; \
++ .align 4 ; \
++ .long 2f - 1f /* namesz */ ; \
++ .long 4f - 3f /* descsz */ ; \
++ .long type ; \
++1:.asciz "name" ; \
++2:.align 4 ; \
++3:desctype descdata ; \
++4:.align 4 ; \
++.popsection
++#endif /* __STDC__ */
++#else /* !__ASSEMBLER__ */
++#include <linux/elf.h>
++/*
++ * Use an anonymous structure which matches the shape of
++ * Elf{32,64}_Nhdr, but includes the name and desc data. The size and
++ * type of name and desc depend on the macro arguments. "name" must
++ * be a literal string, and "desc" must be passed by value. You may
++ * only define one note per line, since __LINE__ is used to generate
++ * unique symbols.
++ */
++#define _ELFNOTE_PASTE(a,b) a##b
++#define _ELFNOTE(size, name, unique, type, desc) \
++ static const struct { \
++ struct elf##size##_note _nhdr; \
++ unsigned char _name[sizeof(name)] \
++ __attribute__((aligned(sizeof(Elf##size##_Word)))); \
++ typeof(desc) _desc \
++ __attribute__((aligned(sizeof(Elf##size##_Word)))); \
++ } _ELFNOTE_PASTE(_note_, unique) \
++ __attribute_used__ \
++ __attribute__((section(".note." name), \
++ aligned(sizeof(Elf##size##_Word)), \
++ unused)) = { \
++ { \
++ sizeof(name), \
++ sizeof(desc), \
++ type, \
++ }, \
++ name, \
++ desc \
++ }
++#define ELFNOTE(size, name, type, desc) \
++ _ELFNOTE(size, name, __LINE__, type, desc)
+
-+#define PAGE_NONE __pgprot(_PAGE_PROTNONE | _PAGE_ACCESSED)
-+#define PAGE_SHARED __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX)
-+#define PAGE_SHARED_EXEC __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED)
-+#define PAGE_COPY_NOEXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX)
-+#define PAGE_COPY PAGE_COPY_NOEXEC
-+#define PAGE_COPY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED)
-+#define PAGE_READONLY __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX)
-+#define PAGE_READONLY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED)
-+#define __PAGE_KERNEL \
-+ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_NX | __kernel_page_user)
-+#define __PAGE_KERNEL_EXEC \
-+ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED | __kernel_page_user)
-+#define __PAGE_KERNEL_NOCACHE \
-+ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_PCD | _PAGE_ACCESSED | _PAGE_NX | __kernel_page_user)
-+#define __PAGE_KERNEL_RO \
-+ (_PAGE_PRESENT | _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_NX | __kernel_page_user)
-+#define __PAGE_KERNEL_VSYSCALL \
-+ (_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED)
-+#define __PAGE_KERNEL_VSYSCALL_NOCACHE \
-+ (_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED | _PAGE_PCD)
-+#define __PAGE_KERNEL_LARGE \
-+ (__PAGE_KERNEL | _PAGE_PSE)
-+#define __PAGE_KERNEL_LARGE_EXEC \
-+ (__PAGE_KERNEL_EXEC | _PAGE_PSE)
++#define ELFNOTE32(name, type, desc) ELFNOTE(32, name, type, desc)
++#define ELFNOTE64(name, type, desc) ELFNOTE(64, name, type, desc)
++#endif /* __ASSEMBLER__ */
+
++#endif /* _LINUX_ELFNOTE_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/eventpoll.h
+--- a/include/linux/eventpoll.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/eventpoll.h Wed Sep 10 10:54:08 2008 +0100
+@@ -90,6 +90,12 @@
+ eventpoll_release_file(file);
+ }
+
+/*
-+ * We don't support GLOBAL page in xenolinux64
++ * called by aio code to create fd that can poll the aio event queueQ
+ */
-+#define MAKE_GLOBAL(x) __pgprot((x))
-+
-+#define PAGE_KERNEL MAKE_GLOBAL(__PAGE_KERNEL)
-+#define PAGE_KERNEL_EXEC MAKE_GLOBAL(__PAGE_KERNEL_EXEC)
-+#define PAGE_KERNEL_RO MAKE_GLOBAL(__PAGE_KERNEL_RO)
-+#define PAGE_KERNEL_NOCACHE MAKE_GLOBAL(__PAGE_KERNEL_NOCACHE)
-+#define PAGE_KERNEL_VSYSCALL32 __pgprot(__PAGE_KERNEL_VSYSCALL)
-+#define PAGE_KERNEL_VSYSCALL MAKE_GLOBAL(__PAGE_KERNEL_VSYSCALL)
-+#define PAGE_KERNEL_LARGE MAKE_GLOBAL(__PAGE_KERNEL_LARGE)
-+#define PAGE_KERNEL_VSYSCALL_NOCACHE MAKE_GLOBAL(__PAGE_KERNEL_VSYSCALL_NOCACHE)
++struct eventpoll;
++int ep_getfd(int *efd, struct inode **einode, struct file **efile,
++ struct eventpoll *ep, const struct file_operations *fops);
+ #else
+
+ static inline void eventpoll_init_file(struct file *file) {}
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/highmem.h
+--- a/include/linux/highmem.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/highmem.h Wed Sep 10 10:54:08 2008 +0100
+@@ -24,10 +24,16 @@
+
+ /* declarations for linux/mm/highmem.c */
+ unsigned int nr_free_highpages(void);
++#ifdef CONFIG_XEN
++void kmap_flush_unused(void);
++#endif
+
+ #else /* CONFIG_HIGHMEM */
+
+ static inline unsigned int nr_free_highpages(void) { return 0; }
++#ifdef CONFIG_XEN
++static inline void kmap_flush_unused(void) { }
++#endif
+
+ static inline void *kmap(struct page *page)
+ {
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/interrupt.h
+--- a/include/linux/interrupt.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/interrupt.h Wed Sep 10 10:54:08 2008 +0100
+@@ -166,6 +166,12 @@
+
+ #endif /* CONFIG_GENERIC_HARDIRQS */
+
++#ifdef CONFIG_HAVE_IRQ_IGNORE_UNHANDLED
++int irq_ignore_unhandled(unsigned int irq);
++#else
++#define irq_ignore_unhandled(irq) 0
++#endif
+
-+/* xwr */
-+#define __P000 PAGE_NONE
-+#define __P001 PAGE_READONLY
-+#define __P010 PAGE_COPY
-+#define __P011 PAGE_COPY
-+#define __P100 PAGE_READONLY_EXEC
-+#define __P101 PAGE_READONLY_EXEC
-+#define __P110 PAGE_COPY_EXEC
-+#define __P111 PAGE_COPY_EXEC
+ #ifndef __ARCH_SET_SOFTIRQ_PENDING
+ #define set_softirq_pending(x) (local_softirq_pending() = (x))
+ #define or_softirq_pending(x) (local_softirq_pending() |= (x))
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/ioport.h
+--- a/include/linux/ioport.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/ioport.h Wed Sep 10 10:54:08 2008 +0100
+@@ -93,6 +93,9 @@
+ /* PC/ISA/whatever - the normal PC address spaces: IO and memory */
+ extern struct resource ioport_resource;
+ extern struct resource iomem_resource;
++#ifdef CONFIG_PROC_IOMEM_MACHINE
++extern struct resource iomem_machine_resource;
++#endif
+
+ extern int request_resource(struct resource *root, struct resource *new);
+ extern struct resource * ____request_resource(struct resource *root, struct resource *new);
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/kexec.h
+--- a/include/linux/kexec.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/kexec.h Wed Sep 10 10:54:08 2008 +0100
+@@ -29,6 +29,13 @@
+
+ #ifndef KEXEC_ARCH
+ #error KEXEC_ARCH not defined
++#endif
+
-+#define __S000 PAGE_NONE
-+#define __S001 PAGE_READONLY
-+#define __S010 PAGE_SHARED
-+#define __S011 PAGE_SHARED
-+#define __S100 PAGE_READONLY_EXEC
-+#define __S101 PAGE_READONLY_EXEC
-+#define __S110 PAGE_SHARED_EXEC
-+#define __S111 PAGE_SHARED_EXEC
++#ifndef KEXEC_ARCH_HAS_PAGE_MACROS
++#define kexec_page_to_pfn(page) page_to_pfn(page)
++#define kexec_pfn_to_page(pfn) pfn_to_page(pfn)
++#define kexec_virt_to_phys(addr) virt_to_phys(addr)
++#define kexec_phys_to_virt(addr) phys_to_virt(addr)
+ #endif
+
+ /*
+@@ -91,6 +98,12 @@
+ extern NORET_TYPE void machine_kexec(struct kimage *image) ATTRIB_NORET;
+ extern int machine_kexec_prepare(struct kimage *image);
+ extern void machine_kexec_cleanup(struct kimage *image);
++#ifdef CONFIG_XEN
++extern int xen_machine_kexec_load(struct kimage *image);
++extern void xen_machine_kexec_unload(struct kimage *image);
++extern void xen_machine_kexec_setup_resources(void);
++extern void xen_machine_kexec_register_resources(struct resource *res);
++#endif
+ extern asmlinkage long sys_kexec_load(unsigned long entry,
+ unsigned long nr_segments,
+ struct kexec_segment __user *segments,
+@@ -107,6 +120,10 @@
+ int kexec_should_crash(struct task_struct *);
+ extern struct kimage *kexec_image;
+ extern struct kimage *kexec_crash_image;
++
++#ifndef kexec_flush_icache_page
++#define kexec_flush_icache_page(page)
++#endif
+
+ #define KEXEC_ON_CRASH 0x00000001
+ #define KEXEC_ARCH_MASK 0xffff0000
+@@ -131,6 +148,7 @@
+ typedef u32 note_buf_t[MAX_NOTE_BYTES/4];
+ extern note_buf_t *crash_notes;
+
++
+ #else /* !CONFIG_KEXEC */
+ struct pt_regs;
+ struct task_struct;
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/mm.h
+--- a/include/linux/mm.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/mm.h Wed Sep 10 10:54:08 2008 +0100
+@@ -164,6 +164,10 @@
+ #define VM_NONLINEAR 0x00800000 /* Is non-linear (remap_file_pages) */
+ #define VM_MAPPED_COPY 0x01000000 /* T if mapped copy of data (nommu mmap) */
+ #define VM_INSERTPAGE 0x02000000 /* The vma has had "vm_insert_page()" done on it */
++#ifdef CONFIG_XEN
++#define VM_FOREIGN 0x04000000 /* Has pages belonging to another VM */
++#endif
++#define VM_ALWAYSDUMP 0x08000000 /* Always include in core dumps */
+
+ #ifndef VM_STACK_DEFAULT_FLAGS /* arch can override this */
+ #define VM_STACK_DEFAULT_FLAGS VM_DATA_DEFAULT_FLAGS
+@@ -202,6 +206,10 @@
+ /* notification that a previously read-only page is about to become
+ * writable, if an error is returned it will cause a SIGBUS */
+ int (*page_mkwrite)(struct vm_area_struct *vma, struct page *page);
++ /* Area-specific function for clearing the PTE at @ptep. Returns the
++ * original value of @ptep. */
++ pte_t (*zap_pte)(struct vm_area_struct *vma,
++ unsigned long addr, pte_t *ptep, int is_fullmm);
+ #ifdef CONFIG_NUMA
+ int (*set_policy)(struct vm_area_struct *vma, struct mempolicy *new);
+ struct mempolicy *(*get_policy)(struct vm_area_struct *vma,
+@@ -1027,6 +1035,13 @@
+ #define FOLL_GET 0x04 /* do get_page on page */
+ #define FOLL_ANON 0x08 /* give ZERO_PAGE if no pgtable */
+
++#ifdef CONFIG_XEN
++typedef int (*pte_fn_t)(pte_t *pte, struct page *pmd_page, unsigned long addr,
++ void *data);
++extern int apply_to_page_range(struct mm_struct *mm, unsigned long address,
++ unsigned long size, pte_fn_t fn, void *data);
++#endif
+
-+static inline unsigned long pgd_bad(pgd_t pgd)
-+{
-+ unsigned long val = __pgd_val(pgd);
-+ val &= ~PTE_MASK;
-+ val &= ~(_PAGE_USER | _PAGE_DIRTY);
-+ return val & ~(_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED);
-+}
+ #ifdef CONFIG_PROC_FS
+ void vm_stat_account(struct mm_struct *, unsigned long, struct file *, long);
+ #else
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/moduleparam.h
+--- a/include/linux/moduleparam.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/moduleparam.h Wed Sep 10 10:54:08 2008 +0100
+@@ -58,13 +58,23 @@
+ void *elem;
+ };
+
++/* On alpha, ia64 and ppc64 relocations to global data cannot go into
++ read-only sections (which is part of respective UNIX ABI on these
++ platforms). So 'const' makes no sense and even causes compile failures
++ with some compilers. */
++#if defined(CONFIG_ALPHA) || defined(CONFIG_IA64) || defined(CONFIG_PPC64)
++#define __moduleparam_const
++#else
++#define __moduleparam_const const
++#endif
++
+ /* This is the fundamental function for registering boot/module
+ parameters. perm sets the visibility in driverfs: 000 means it's
+ not there, read bits mean it's readable, write bits mean it's
+ writable. */
+ #define __module_param_call(prefix, name, set, get, arg, perm) \
+ static char __param_str_##name[] = prefix #name; \
+- static struct kernel_param const __param_##name \
++ static struct kernel_param __moduleparam_const __param_##name \
+ __attribute_used__ \
+ __attribute__ ((unused,__section__ ("__param"),aligned(sizeof(void *)))) \
+ = { __param_str_##name, perm, set, get, arg }
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/netfilter_bridge.h
+--- a/include/linux/netfilter_bridge.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/netfilter_bridge.h Wed Sep 10 10:54:08 2008 +0100
+@@ -7,6 +7,7 @@
+ #include <linux/netfilter.h>
+ #if defined(__KERNEL__) && defined(CONFIG_BRIDGE_NETFILTER)
+ #include <linux/if_ether.h>
++#include <linux/if_vlan.h>
+ #endif
+
+ /* Bridge Hooks */
+@@ -54,16 +55,19 @@
+
+ if (skb->nf_bridge) {
+ if (skb->protocol == __constant_htons(ETH_P_8021Q)) {
+- err = skb_cow(skb, 18);
++ int header_size = ETH_HLEN + VLAN_HLEN;
++
++ err = skb_cow(skb, header_size);
+ if (err)
+ return err;
+- memcpy(skb->data - 18, skb->nf_bridge->data, 18);
+- skb_push(skb, 4);
++ memcpy(skb->data - header_size, skb->nf_bridge->data,
++ header_size);
++ skb_push(skb, VLAN_HLEN);
+ } else {
+- err = skb_cow(skb, 16);
++ err = skb_cow(skb, ETH_HLEN);
+ if (err)
+ return err;
+- memcpy(skb->data - 16, skb->nf_bridge->data, 16);
++ memcpy(skb->data - ETH_HLEN, skb->nf_bridge->data, ETH_HLEN);
+ }
+ }
+ return 0;
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/oprofile.h
+--- a/include/linux/oprofile.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/oprofile.h Wed Sep 10 10:54:08 2008 +0100
+@@ -16,6 +16,8 @@
+ #include <linux/types.h>
+ #include <linux/spinlock.h>
+ #include <asm/atomic.h>
+
-+static inline unsigned long pud_bad(pud_t pud)
-+{
-+ unsigned long val = __pud_val(pud);
-+ val &= ~PTE_MASK;
-+ val &= ~(_PAGE_USER | _PAGE_DIRTY);
-+ return val & ~(_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED);
-+}
++#include <xen/interface/xenoprof.h>
+
+ struct super_block;
+ struct dentry;
+@@ -27,6 +29,11 @@
+ /* create any necessary configuration files in the oprofile fs.
+ * Optional. */
+ int (*create_files)(struct super_block * sb, struct dentry * root);
++ /* setup active domains with Xen */
++ int (*set_active)(int *active_domains, unsigned int adomains);
++ /* setup passive domains with Xen */
++ int (*set_passive)(int *passive_domains, unsigned int pdomains);
++
+ /* Do any necessary interrupt setup. Optional. */
+ int (*setup)(void);
+ /* Do any necessary interrupt shutdown. Optional. */
+@@ -78,6 +85,8 @@
+ /* add a backtrace entry, to be called from the ->backtrace callback */
+ void oprofile_add_trace(unsigned long eip);
+
++/* add a domain switch entry */
++int oprofile_add_domain_switch(int32_t domain_id);
+
+ /**
+ * Create a file of the given name as a child of the given root, with
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/page-flags.h
+--- a/include/linux/page-flags.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/page-flags.h Wed Sep 10 10:54:08 2008 +0100
+@@ -97,6 +97,8 @@
+ */
+ #define PG_uncached 31 /* Page has been mapped as uncached */
+ #endif
+
-+#define set_pte_at(_mm,addr,ptep,pteval) do { \
-+ if (((_mm) != current->mm && (_mm) != &init_mm) || \
-+ HYPERVISOR_update_va_mapping((addr), (pteval), 0)) \
-+ set_pte((ptep), (pteval)); \
++#define PG_foreign 20 /* Page is owned by foreign allocator. */
+
+ /*
+ * Manipulation of page state flags
+@@ -247,6 +249,19 @@
+ #define SetPageUncached(page) set_bit(PG_uncached, &(page)->flags)
+ #define ClearPageUncached(page) clear_bit(PG_uncached, &(page)->flags)
+
++#define PageForeign(page) test_bit(PG_foreign, &(page)->flags)
++#define SetPageForeign(_page, dtor) do { \
++ set_bit(PG_foreign, &(_page)->flags); \
++ BUG_ON((dtor) == (void (*)(struct page *))0); \
++ (_page)->index = (long)(dtor); \
+} while (0)
++#define ClearPageForeign(page) do { \
++ clear_bit(PG_foreign, &(page)->flags); \
++ (page)->index = 0; \
++} while (0)
++#define PageForeignDestructor(_page) \
++ ((void (*)(struct page *))(_page)->index)(_page)
+
-+#define pte_none(x) (!(x).pte)
-+#define pte_present(x) ((x).pte & (_PAGE_PRESENT | _PAGE_PROTNONE))
-+#define pte_clear(mm,addr,xp) do { set_pte_at(mm, addr, xp, __pte(0)); } while (0)
-+
-+#define pages_to_mb(x) ((x) >> (20-PAGE_SHIFT))
-+
-+#define __pte_mfn(_pte) (((_pte).pte & PTE_MASK) >> PAGE_SHIFT)
-+#define pte_mfn(_pte) ((_pte).pte & _PAGE_PRESENT ? \
-+ __pte_mfn(_pte) : pfn_to_mfn(__pte_mfn(_pte)))
-+#define pte_pfn(_pte) ((_pte).pte & _PAGE_PRESENT ? \
-+ mfn_to_local_pfn(__pte_mfn(_pte)) : __pte_mfn(_pte))
-+
-+#define pte_page(x) pfn_to_page(pte_pfn(x))
-+
-+static inline pte_t pfn_pte(unsigned long page_nr, pgprot_t pgprot)
-+{
-+ unsigned long pte = page_nr << PAGE_SHIFT;
-+ pte |= pgprot_val(pgprot);
-+ pte &= __supported_pte_mask;
-+ return __pte(pte);
-+}
-+
-+static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
-+{
-+ pte_t pte = *ptep;
-+ if (!pte_none(pte)) {
-+ if (mm != &init_mm)
-+ pte = __pte_ma(xchg(&ptep->pte, 0));
-+ else
-+ HYPERVISOR_update_va_mapping(addr, __pte(0), 0);
-+ }
-+ return pte;
-+}
-+
-+static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm, unsigned long addr, pte_t *ptep, int full)
+ struct page; /* forward declaration */
+
+ int test_clear_page_dirty(struct page *page);
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/pci.h
+--- a/include/linux/pci.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/pci.h Wed Sep 10 10:54:08 2008 +0100
+@@ -152,6 +152,9 @@
+ * directly, use the values stored here. They might be different!
+ */
+ unsigned int irq;
++#ifdef CONFIG_XEN
++ unsigned int irq_old;
++#endif
+ struct resource resource[DEVICE_COUNT_RESOURCE]; /* I/O and memory regions + expansion ROMs */
+
+ /* These fields are used by common fixups */
+@@ -596,6 +599,10 @@
+ struct msix_entry *entries, int nvec) {return -1;}
+ static inline void pci_disable_msix(struct pci_dev *dev) {}
+ static inline void msi_remove_pci_irq_vectors(struct pci_dev *dev) {}
++#ifdef CONFIG_XEN
++#define register_msi_get_owner(func) 0
++#define unregister_msi_get_owner(func) 0
++#endif
+ #else
+ extern void pci_scan_msi_device(struct pci_dev *dev);
+ extern int pci_enable_msi(struct pci_dev *dev);
+@@ -604,6 +611,10 @@
+ struct msix_entry *entries, int nvec);
+ extern void pci_disable_msix(struct pci_dev *dev);
+ extern void msi_remove_pci_irq_vectors(struct pci_dev *dev);
++#ifdef CONFIG_XEN
++extern int register_msi_get_owner(int (*func)(struct pci_dev *dev));
++extern int unregister_msi_get_owner(int (*func)(struct pci_dev *dev));
++#endif
+ #endif
+
+ extern void pci_block_user_cfg_access(struct pci_dev *dev);
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/sched.h
+--- a/include/linux/sched.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/sched.h Wed Sep 10 10:54:08 2008 +0100
+@@ -211,10 +211,15 @@
+ extern void scheduler_tick(void);
+
+ #ifdef CONFIG_DETECT_SOFTLOCKUP
++extern unsigned long softlockup_get_next_event(void);
+ extern void softlockup_tick(void);
+ extern void spawn_softlockup_task(void);
+ extern void touch_softlockup_watchdog(void);
+ #else
++static inline unsigned long softlockup_get_next_event(void)
+{
-+ if (full) {
-+ pte_t pte = *ptep;
-+ if (mm->context.pinned)
-+ xen_l1_entry_update(ptep, __pte(0));
-+ else
-+ *ptep = __pte(0);
-+ return pte;
-+ }
-+ return ptep_get_and_clear(mm, addr, ptep);
++ return MAX_JIFFY_OFFSET;
+}
-+
-+#define ptep_clear_flush(vma, addr, ptep) \
-+({ \
-+ pte_t *__ptep = (ptep); \
-+ pte_t __res = *__ptep; \
-+ if (!pte_none(__res) && \
-+ ((vma)->vm_mm != current->mm || \
-+ HYPERVISOR_update_va_mapping(addr, __pte(0), \
-+ (unsigned long)(vma)->vm_mm->cpu_vm_mask.bits| \
-+ UVMF_INVLPG|UVMF_MULTI))) { \
-+ __ptep->pte = 0; \
-+ flush_tlb_page(vma, addr); \
-+ } \
-+ __res; \
-+})
-+
-+/*
-+ * The following only work if pte_present() is true.
-+ * Undefined behaviour if not..
+ static inline void softlockup_tick(void)
+ {
+ }
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/skbuff.h
+--- a/include/linux/skbuff.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/skbuff.h Wed Sep 10 10:54:08 2008 +0100
+@@ -203,6 +203,8 @@
+ * @local_df: allow local fragmentation
+ * @cloned: Head may be cloned (check refcnt to be sure)
+ * @nohdr: Payload reference only, must not modify header
++ * @proto_data_valid: Protocol data validated since arriving at localhost
++ * @proto_csum_blank: Protocol csum must be added before leaving localhost
+ * @pkt_type: Packet class
+ * @fclone: skbuff clone status
+ * @ip_summed: Driver fed us an IP checksum
+@@ -282,7 +284,13 @@
+ nfctinfo:3;
+ __u8 pkt_type:3,
+ fclone:2,
++#ifndef CONFIG_XEN
+ ipvs_property:1;
++#else
++ ipvs_property:1,
++ proto_data_valid:1,
++ proto_csum_blank:1;
++#endif
+ __be16 protocol;
+
+ void (*destructor)(struct sk_buff *skb);
+diff -r d894e36cfc30 -r 0aa021803deb include/linux/sysctl.h
+--- a/include/linux/sysctl.h Tue Sep 09 11:37:38 2008 +0200
++++ b/include/linux/sysctl.h Wed Sep 10 10:54:08 2008 +0100
+@@ -6,10 +6,17 @@
+ ****************************************************************
+ ****************************************************************
+ **
++ ** WARNING:
+ ** The values in this file are exported to user space via
+- ** the sysctl() binary interface. However this interface
+- ** is unstable and deprecated and will be removed in the future.
+- ** For a stable interface use /proc/sys.
++ ** the sysctl() binary interface. Do *NOT* change the
++ ** numbering of any existing values here, and do not change
++ ** any numbers within any one set of values. If you have to
++ ** have to redefine an existing interface, use a new number for it.
++ ** The kernel will then return -ENOTDIR to any application using
++ ** the old binary interface.
++ **
++ ** For new interfaces unless you really need a binary number
++ ** please use CTL_UNNUMBERED.
+ **
+ ****************************************************************
+ ****************************************************************
+@@ -48,6 +55,7 @@
+ #ifdef __KERNEL__
+ #define CTL_ANY -1 /* Matches any name */
+ #define CTL_NONE 0
++#define CTL_UNNUMBERED CTL_NONE /* sysctl without a binary number */
+ #endif
+
+ enum
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/balloon.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/balloon.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,57 @@
++/******************************************************************************
++ * balloon.h
++ *
++ * Xen balloon driver - enables returning/claiming memory to/from Xen.
++ *
++ * Copyright (c) 2003, B Dragovic
++ * Copyright (c) 2003-2004, M Williamson, K Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
-+#define __LARGE_PTE (_PAGE_PSE|_PAGE_PRESENT)
-+static inline int pte_user(pte_t pte) { return __pte_val(pte) & _PAGE_USER; }
-+static inline int pte_read(pte_t pte) { return __pte_val(pte) & _PAGE_USER; }
-+static inline int pte_exec(pte_t pte) { return __pte_val(pte) & _PAGE_USER; }
-+static inline int pte_dirty(pte_t pte) { return __pte_val(pte) & _PAGE_DIRTY; }
-+static inline int pte_young(pte_t pte) { return __pte_val(pte) & _PAGE_ACCESSED; }
-+static inline int pte_write(pte_t pte) { return __pte_val(pte) & _PAGE_RW; }
-+static inline int pte_file(pte_t pte) { return __pte_val(pte) & _PAGE_FILE; }
-+static inline int pte_huge(pte_t pte) { return __pte_val(pte) & _PAGE_PSE; }
-+
-+static inline pte_t pte_rdprotect(pte_t pte) { __pte_val(pte) &= ~_PAGE_USER; return pte; }
-+static inline pte_t pte_exprotect(pte_t pte) { __pte_val(pte) &= ~_PAGE_USER; return pte; }
-+static inline pte_t pte_mkclean(pte_t pte) { __pte_val(pte) &= ~_PAGE_DIRTY; return pte; }
-+static inline pte_t pte_mkold(pte_t pte) { __pte_val(pte) &= ~_PAGE_ACCESSED; return pte; }
-+static inline pte_t pte_wrprotect(pte_t pte) { __pte_val(pte) &= ~_PAGE_RW; return pte; }
-+static inline pte_t pte_mkread(pte_t pte) { __pte_val(pte) |= _PAGE_USER; return pte; }
-+static inline pte_t pte_mkexec(pte_t pte) { __pte_val(pte) |= _PAGE_USER; return pte; }
-+static inline pte_t pte_mkdirty(pte_t pte) { __pte_val(pte) |= _PAGE_DIRTY; return pte; }
-+static inline pte_t pte_mkyoung(pte_t pte) { __pte_val(pte) |= _PAGE_ACCESSED; return pte; }
-+static inline pte_t pte_mkwrite(pte_t pte) { __pte_val(pte) |= _PAGE_RW; return pte; }
-+static inline pte_t pte_mkhuge(pte_t pte) { __pte_val(pte) |= _PAGE_PSE; return pte; }
-+
-+#define ptep_test_and_clear_dirty(vma, addr, ptep) \
-+({ \
-+ pte_t __pte = *(ptep); \
-+ int __ret = pte_dirty(__pte); \
-+ if (__ret) \
-+ set_pte_at((vma)->vm_mm, addr, ptep, pte_mkclean(__pte)); \
-+ __ret; \
-+})
-+
-+#define ptep_test_and_clear_young(vma, addr, ptep) \
-+({ \
-+ pte_t __pte = *(ptep); \
-+ int __ret = pte_young(__pte); \
-+ if (__ret) \
-+ set_pte_at((vma)->vm_mm, addr, ptep, pte_mkold(__pte)); \
-+ __ret; \
-+})
+
-+static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
-+{
-+ pte_t pte = *ptep;
-+ if (pte_write(pte))
-+ set_pte_at(mm, addr, ptep, pte_wrprotect(pte));
-+}
++#ifndef __ASM_BALLOON_H__
++#define __ASM_BALLOON_H__
+
+/*
-+ * Macro to mark a page protection value as "uncacheable".
++ * Inform the balloon driver that it should allow some slop for device-driver
++ * memory activities.
+ */
-+#define pgprot_noncached(prot) (__pgprot(pgprot_val(prot) | _PAGE_PCD | _PAGE_PWT))
++void balloon_update_driver_allowance(long delta);
+
-+static inline int pmd_large(pmd_t pte) {
-+ return (__pmd_val(pte) & __LARGE_PTE) == __LARGE_PTE;
-+}
++/* Allocate/free a set of empty pages in low memory (i.e., no RAM mapped). */
++struct page **alloc_empty_pages_and_pagevec(int nr_pages);
++void free_empty_pages_and_pagevec(struct page **pagevec, int nr_pages);
+
++void balloon_release_driver_page(struct page *page);
+
+/*
-+ * Conversion functions: convert a page and protection to a page entry,
-+ * and a page entry and page directory to the page they refer to.
++ * Prevent the balloon driver from changing the memory reservation during
++ * a driver critical region.
+ */
++extern spinlock_t balloon_lock;
++#define balloon_lock(__flags) spin_lock_irqsave(&balloon_lock, __flags)
++#define balloon_unlock(__flags) spin_unlock_irqrestore(&balloon_lock, __flags)
+
-+/*
-+ * Level 4 access.
-+ * Never use these in the common code.
++#endif /* __ASM_BALLOON_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/blkif.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/blkif.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,123 @@
++/*
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
+ */
-+#define pgd_page(pgd) ((unsigned long) __va(pgd_val(pgd) & PTE_MASK))
-+#define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD-1))
-+#define pgd_offset(mm, addr) ((mm)->pgd + pgd_index(addr))
-+#define pgd_offset_k(address) (pgd_t *)(init_level4_pgt + pgd_index(address))
-+#define pgd_present(pgd) (__pgd_val(pgd) & _PAGE_PRESENT)
-+#define mk_kernel_pgd(address) __pgd((address) | _KERNPG_TABLE)
-+
-+/* PUD - Level3 access */
-+/* to find an entry in a page-table-directory. */
-+#define pud_index(address) (((address) >> PUD_SHIFT) & (PTRS_PER_PUD-1))
-+#define pud_offset(pgd, address) ((pud_t *) pgd_page(*(pgd)) + pud_index(address))
-+#define pud_present(pud) (__pud_val(pud) & _PAGE_PRESENT)
+
-+/* PMD - Level 2 access */
-+#define pmd_page_kernel(pmd) ((unsigned long) __va(pmd_val(pmd) & PTE_MASK))
-+#define pmd_page(pmd) (pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT))
++#ifndef __XEN_BLKIF_H__
++#define __XEN_BLKIF_H__
+
-+#define pmd_index(address) (((address) >> PMD_SHIFT) & (PTRS_PER_PMD-1))
-+#define pmd_offset(dir, address) ((pmd_t *) pud_page(*(dir)) + \
-+ pmd_index(address))
-+#define pmd_none(x) (!__pmd_val(x))
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+/* pmd_present doesn't just test the _PAGE_PRESENT bit since wr.p.t.
-+ can temporarily clear it. */
-+#define pmd_present(x) (__pmd_val(x))
-+#else
-+#define pmd_present(x) (__pmd_val(x) & _PAGE_PRESENT)
-+#endif
-+#define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0)
-+#define pmd_bad(x) ((__pmd_val(x) & ~(PTE_MASK | _PAGE_USER | _PAGE_PRESENT)) \
-+ != (_KERNPG_TABLE & ~(_PAGE_USER | _PAGE_PRESENT)))
-+#define pfn_pmd(nr,prot) (__pmd(((nr) << PAGE_SHIFT) | pgprot_val(prot)))
-+#define pmd_pfn(x) ((pmd_val(x) & __PHYSICAL_MASK) >> PAGE_SHIFT)
++#include <xen/interface/io/ring.h>
++#include <xen/interface/io/blkif.h>
++#include <xen/interface/io/protocols.h>
+
-+#define pte_to_pgoff(pte) ((__pte_val(pte) & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT)
-+#define pgoff_to_pte(off) ((pte_t) { ((off) << PAGE_SHIFT) | _PAGE_FILE })
-+#define PTE_FILE_MAX_BITS __PHYSICAL_MASK_SHIFT
++/* Not a real protocol. Used to generate ring structs which contain
++ * the elements common to all protocols only. This way we get a
++ * compiler-checkable way to use common struct elements, so we can
++ * avoid using switch(protocol) in a number of places. */
++struct blkif_common_request {
++ char dummy;
++};
++struct blkif_common_response {
++ char dummy;
++};
+
-+/* PTE - Level 1 access. */
++/* i386 protocol version */
++#pragma pack(push, 4)
++struct blkif_x86_32_request {
++ uint8_t operation; /* BLKIF_OP_??? */
++ uint8_t nr_segments; /* number of segments */
++ blkif_vdev_t handle; /* only for read/write requests */
++ uint64_t id; /* private guest value, echoed in resp */
++ blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
++ struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++};
++struct blkif_x86_32_response {
++ uint64_t id; /* copied from request */
++ uint8_t operation; /* copied from request */
++ int16_t status; /* BLKIF_RSP_??? */
++};
++typedef struct blkif_x86_32_request blkif_x86_32_request_t;
++typedef struct blkif_x86_32_response blkif_x86_32_response_t;
++#pragma pack(pop)
+
-+/* page, protection -> pte */
-+#define mk_pte(page, pgprot) pfn_pte(page_to_pfn(page), (pgprot))
-+#define mk_pte_huge(entry) (__pte_val(entry) |= _PAGE_PRESENT | _PAGE_PSE)
-+
-+/* physical address -> PTE */
-+static inline pte_t mk_pte_phys(unsigned long physpage, pgprot_t pgprot)
-+{
-+ unsigned long pteval;
-+ pteval = physpage | pgprot_val(pgprot);
-+ return __pte(pteval);
-+}
-+
-+/* Change flags of a PTE */
-+static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
-+{
-+ /*
-+ * Since this might change the present bit (which controls whether
-+ * a pte_t object has undergone p2m translation), we must use
-+ * pte_val() on the input pte and __pte() for the return value.
-+ */
-+ unsigned long pteval = pte_val(pte);
++/* x86_64 protocol version */
++struct blkif_x86_64_request {
++ uint8_t operation; /* BLKIF_OP_??? */
++ uint8_t nr_segments; /* number of segments */
++ blkif_vdev_t handle; /* only for read/write requests */
++ uint64_t __attribute__((__aligned__(8))) id;
++ blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
++ struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++};
++struct blkif_x86_64_response {
++ uint64_t __attribute__((__aligned__(8))) id;
++ uint8_t operation; /* copied from request */
++ int16_t status; /* BLKIF_RSP_??? */
++};
++typedef struct blkif_x86_64_request blkif_x86_64_request_t;
++typedef struct blkif_x86_64_response blkif_x86_64_response_t;
+
-+ pteval &= _PAGE_CHG_MASK;
-+ pteval |= pgprot_val(newprot);
-+ pteval &= __supported_pte_mask;
-+ return __pte(pteval);
-+}
++DEFINE_RING_TYPES(blkif_common, struct blkif_common_request, struct blkif_common_response);
++DEFINE_RING_TYPES(blkif_x86_32, struct blkif_x86_32_request, struct blkif_x86_32_response);
++DEFINE_RING_TYPES(blkif_x86_64, struct blkif_x86_64_request, struct blkif_x86_64_response);
+
-+#define pte_index(address) \
-+ (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1))
-+#define pte_offset_kernel(dir, address) ((pte_t *) pmd_page_kernel(*(dir)) + \
-+ pte_index(address))
++union blkif_back_rings {
++ blkif_back_ring_t native;
++ blkif_common_back_ring_t common;
++ blkif_x86_32_back_ring_t x86_32;
++ blkif_x86_64_back_ring_t x86_64;
++};
++typedef union blkif_back_rings blkif_back_rings_t;
+
-+/* x86-64 always has all page tables mapped. */
-+#define pte_offset_map(dir,address) pte_offset_kernel(dir,address)
-+#define pte_offset_map_nested(dir,address) pte_offset_kernel(dir,address)
-+#define pte_unmap(pte) /* NOP */
-+#define pte_unmap_nested(pte) /* NOP */
++enum blkif_protocol {
++ BLKIF_PROTOCOL_NATIVE = 1,
++ BLKIF_PROTOCOL_X86_32 = 2,
++ BLKIF_PROTOCOL_X86_64 = 3,
++};
+
-+#define update_mmu_cache(vma,address,pte) do { } while (0)
++static void inline blkif_get_x86_32_req(blkif_request_t *dst, blkif_x86_32_request_t *src)
++{
++ int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST;
++ dst->operation = src->operation;
++ dst->nr_segments = src->nr_segments;
++ dst->handle = src->handle;
++ dst->id = src->id;
++ dst->sector_number = src->sector_number;
++ barrier();
++ if (n > dst->nr_segments)
++ n = dst->nr_segments;
++ for (i = 0; i < n; i++)
++ dst->seg[i] = src->seg[i];
++}
++
++static void inline blkif_get_x86_64_req(blkif_request_t *dst, blkif_x86_64_request_t *src)
++{
++ int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST;
++ dst->operation = src->operation;
++ dst->nr_segments = src->nr_segments;
++ dst->handle = src->handle;
++ dst->id = src->id;
++ dst->sector_number = src->sector_number;
++ barrier();
++ if (n > dst->nr_segments)
++ n = dst->nr_segments;
++ for (i = 0; i < n; i++)
++ dst->seg[i] = src->seg[i];
++}
+
++#endif /* __XEN_BLKIF_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/compat_ioctl.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/compat_ioctl.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,45 @@
+/*
-+ * Rules for using ptep_establish: the pte MUST be a user pte, and
-+ * must be a present->present transition.
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License as
++ * published by the Free Software Foundation; either version 2 of the
++ * License, or (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
++ *
++ * Copyright IBM Corp. 2007
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ * Hollis Blanchard <hollisb at us.ibm.com>
+ */
-+#define __HAVE_ARCH_PTEP_ESTABLISH
-+#define ptep_establish(vma, address, ptep, pteval) \
-+ do { \
-+ if ( likely((vma)->vm_mm == current->mm) ) { \
-+ BUG_ON(HYPERVISOR_update_va_mapping(address, \
-+ pteval, \
-+ (unsigned long)(vma)->vm_mm->cpu_vm_mask.bits| \
-+ UVMF_INVLPG|UVMF_MULTI)); \
-+ } else { \
-+ xen_l1_entry_update(ptep, pteval); \
-+ flush_tlb_page(vma, address); \
-+ } \
-+ } while (0)
-+
-+/* We only update the dirty/accessed state if we set
-+ * the dirty bit by hand in the kernel, since the hardware
-+ * will do the accessed bit for us, and we don't want to
-+ * race with other CPU's that might be updating the dirty
-+ * bit at the same time. */
-+#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
-+#define ptep_set_access_flags(vma, address, ptep, entry, dirty) \
-+ do { \
-+ if (dirty) \
-+ ptep_establish(vma, address, ptep, entry); \
-+ } while (0)
+
-+/* Encode and de-code a swap entry */
-+#define __swp_type(x) (((x).val >> 1) & 0x3f)
-+#define __swp_offset(x) ((x).val >> 8)
-+#define __swp_entry(type, offset) ((swp_entry_t) { ((type) << 1) | ((offset) << 8) })
-+#define __pte_to_swp_entry(pte) ((swp_entry_t) { pte_val(pte) })
-+#define __swp_entry_to_pte(x) ((pte_t) { (x).val })
++#ifndef __LINUX_XEN_COMPAT_H__
++#define __LINUX_XEN_COMPAT_H__
+
-+extern spinlock_t pgd_lock;
-+extern struct page *pgd_list;
-+void vmalloc_sync_all(void);
++#include <linux/compat.h>
+
-+#endif /* !__ASSEMBLY__ */
++extern int privcmd_ioctl_32(int fd, unsigned int cmd, unsigned long arg);
++struct privcmd_mmap_32 {
++ int num;
++ domid_t dom;
++ compat_uptr_t entry;
++};
+
-+extern int kern_addr_valid(unsigned long addr);
++struct privcmd_mmapbatch_32 {
++ int num; /* number of pages to populate */
++ domid_t dom; /* target domain */
++ __u64 addr; /* virtual address */
++ compat_uptr_t arr; /* array of mfns - top nibble set on err */
++};
++#define IOCTL_PRIVCMD_MMAP_32 \
++ _IOC(_IOC_NONE, 'P', 2, sizeof(struct privcmd_mmap_32))
++#define IOCTL_PRIVCMD_MMAPBATCH_32 \
++ _IOC(_IOC_NONE, 'P', 3, sizeof(struct privcmd_mmapbatch_32))
++
++#endif /* __LINUX_XEN_COMPAT_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/cpu_hotplug.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/cpu_hotplug.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,41 @@
++#ifndef __XEN_CPU_HOTPLUG_H__
++#define __XEN_CPU_HOTPLUG_H__
+
-+#define DOMID_LOCAL (0xFFFFU)
++#include <linux/kernel.h>
++#include <linux/cpumask.h>
+
-+struct vm_area_struct;
++#if defined(CONFIG_X86) && defined(CONFIG_SMP)
++extern cpumask_t cpu_initialized_map;
++#endif
+
-+int direct_remap_pfn_range(struct vm_area_struct *vma,
-+ unsigned long address,
-+ unsigned long mfn,
-+ unsigned long size,
-+ pgprot_t prot,
-+ domid_t domid);
++#if defined(CONFIG_HOTPLUG_CPU)
+
-+int direct_kernel_remap_pfn_range(unsigned long address,
-+ unsigned long mfn,
-+ unsigned long size,
-+ pgprot_t prot,
-+ domid_t domid);
++int cpu_up_check(unsigned int cpu);
++void init_xenbus_allowed_cpumask(void);
++int smp_suspend(void);
++void smp_resume(void);
+
-+int create_lookup_pte_addr(struct mm_struct *mm,
-+ unsigned long address,
-+ uint64_t *ptep);
++void cpu_bringup(void);
+
-+int touch_pte_range(struct mm_struct *mm,
-+ unsigned long address,
-+ unsigned long size);
++#else /* !defined(CONFIG_HOTPLUG_CPU) */
+
-+#define io_remap_pfn_range(vma, vaddr, pfn, size, prot) \
-+ direct_remap_pfn_range(vma,vaddr,pfn,size,prot,DOMID_IO)
++#define cpu_up_check(cpu) (0)
++#define init_xenbus_allowed_cpumask() ((void)0)
+
-+#define MK_IOSPACE_PFN(space, pfn) (pfn)
-+#define GET_IOSPACE(pfn) 0
-+#define GET_PFN(pfn) (pfn)
++static inline int smp_suspend(void)
++{
++ if (num_online_cpus() > 1) {
++ printk(KERN_WARNING "Can't suspend SMP guests "
++ "without CONFIG_HOTPLUG_CPU\n");
++ return -EOPNOTSUPP;
++ }
++ return 0;
++}
+
-+#define HAVE_ARCH_UNMAPPED_AREA
++static inline void smp_resume(void)
++{
++}
+
-+#define pgtable_cache_init() do { } while (0)
-+#define check_pgt_cache() do { } while (0)
++#endif /* !defined(CONFIG_HOTPLUG_CPU) */
+
-+#define PAGE_AGP PAGE_KERNEL_NOCACHE
-+#define HAVE_PAGE_AGP 1
++#endif /* __XEN_CPU_HOTPLUG_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/driver_util.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/driver_util.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,14 @@
+
-+/* fs/proc/kcore.c */
-+#define kc_vaddr_to_offset(v) ((v) & __VIRTUAL_MASK)
-+#define kc_offset_to_vaddr(o) \
-+ (((o) & (1UL << (__VIRTUAL_MASK_SHIFT-1))) ? ((o) | (~__VIRTUAL_MASK)) : (o))
++#ifndef __ASM_XEN_DRIVER_UTIL_H__
++#define __ASM_XEN_DRIVER_UTIL_H__
+
-+#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
-+#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_DIRTY
-+#define __HAVE_ARCH_PTEP_GET_AND_CLEAR
-+#define __HAVE_ARCH_PTEP_GET_AND_CLEAR_FULL
-+#define __HAVE_ARCH_PTEP_CLEAR_FLUSH
-+#define __HAVE_ARCH_PTEP_SET_WRPROTECT
-+#define __HAVE_ARCH_PTE_SAME
-+#include <asm-generic/pgtable.h>
++#include <linux/vmalloc.h>
++#include <linux/device.h>
+
-+#endif /* _X86_64_PGTABLE_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/processor.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/processor.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/processor.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/processor.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,506 @@
-+/*
-+ * include/asm-x86_64/processor.h
-+ *
-+ * Copyright (C) 1994 Linus Torvalds
-+ */
++/* Allocate/destroy a 'vmalloc' VM area. */
++extern struct vm_struct *alloc_vm_area(unsigned long size);
++extern void free_vm_area(struct vm_struct *area);
+
-+#ifndef __ASM_X86_64_PROCESSOR_H
-+#define __ASM_X86_64_PROCESSOR_H
++extern struct class *get_xen_class(void);
+
-+#include <asm/segment.h>
-+#include <asm/page.h>
-+#include <asm/types.h>
-+#include <asm/sigcontext.h>
-+#include <asm/cpufeature.h>
-+#include <linux/threads.h>
-+#include <asm/msr.h>
-+#include <asm/current.h>
-+#include <asm/system.h>
-+#include <asm/mmsegment.h>
-+#include <asm/percpu.h>
-+#include <linux/personality.h>
-+#include <linux/cpumask.h>
++#endif /* __ASM_XEN_DRIVER_UTIL_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/evtchn.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/evtchn.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,160 @@
++/******************************************************************************
++ * evtchn.h
++ *
++ * Communication via Xen event channels.
++ * Also definitions for the device that demuxes notifications to userspace.
++ *
++ * Copyright (c) 2004-2005, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+#define TF_MASK 0x00000100
-+#define IF_MASK 0x00000200
-+#define IOPL_MASK 0x00003000
-+#define NT_MASK 0x00004000
-+#define VM_MASK 0x00020000
-+#define AC_MASK 0x00040000
-+#define VIF_MASK 0x00080000 /* virtual interrupt flag */
-+#define VIP_MASK 0x00100000 /* virtual interrupt pending */
-+#define ID_MASK 0x00200000
++#ifndef __ASM_EVTCHN_H__
++#define __ASM_EVTCHN_H__
+
-+#define desc_empty(desc) \
-+ (!((desc)->a | (desc)->b))
++#include <linux/interrupt.h>
++#include <asm/hypervisor.h>
++#include <asm/ptrace.h>
++#include <asm/synch_bitops.h>
++#include <xen/interface/event_channel.h>
++#include <linux/smp.h>
+
-+#define desc_equal(desc1, desc2) \
-+ (((desc1)->a == (desc2)->a) && ((desc1)->b == (desc2)->b))
++/*
++ * LOW-LEVEL DEFINITIONS
++ */
+
+/*
-+ * Default implementation of macro that returns current
-+ * instruction pointer ("program counter").
++ * Dynamically bind an event source to an IRQ-like callback handler.
++ * On some platforms this may not be implemented via the Linux IRQ subsystem.
++ * The IRQ argument passed to the callback handler is the same as returned
++ * from the bind call. It may not correspond to a Linux IRQ number.
++ * Returns IRQ or negative errno.
+ */
-+#define current_text_addr() ({ void *pc; asm volatile("leaq 1f(%%rip),%0\n1:":"=r"(pc)); pc; })
++int bind_caller_port_to_irqhandler(
++ unsigned int caller_port,
++ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id);
++int bind_listening_port_to_irqhandler(
++ unsigned int remote_domain,
++ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id);
++int bind_interdomain_evtchn_to_irqhandler(
++ unsigned int remote_domain,
++ unsigned int remote_port,
++ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id);
++int bind_virq_to_irqhandler(
++ unsigned int virq,
++ unsigned int cpu,
++ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id);
++int bind_ipi_to_irqhandler(
++ unsigned int ipi,
++ unsigned int cpu,
++ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id);
+
+/*
-+ * CPU type and hardware bug flags. Kept separately for each CPU.
++ * Common unbind function for all event sources. Takes IRQ to unbind from.
++ * Automatically closes the underlying event channel (except for bindings
++ * made with bind_caller_port_to_irqhandler()).
+ */
++void unbind_from_irqhandler(unsigned int irq, void *dev_id);
+
-+struct cpuinfo_x86 {
-+ __u8 x86; /* CPU family */
-+ __u8 x86_vendor; /* CPU vendor */
-+ __u8 x86_model;
-+ __u8 x86_mask;
-+ int cpuid_level; /* Maximum supported CPUID level, -1=no CPUID */
-+ __u32 x86_capability[NCAPINTS];
-+ char x86_vendor_id[16];
-+ char x86_model_id[64];
-+ int x86_cache_size; /* in KB */
-+ int x86_clflush_size;
-+ int x86_cache_alignment;
-+ int x86_tlbsize; /* number of 4K pages in DTLB/ITLB combined(in pages)*/
-+ __u8 x86_virt_bits, x86_phys_bits;
-+ __u8 x86_max_cores; /* cpuid returned max cores value */
-+ __u32 x86_power;
-+ __u32 extended_cpuid_level; /* Max extended CPUID function supported */
-+ unsigned long loops_per_jiffy;
-+#ifdef CONFIG_SMP
-+ cpumask_t llc_shared_map; /* cpus sharing the last level cache */
-+#endif
-+ __u8 apicid;
-+#ifdef CONFIG_SMP
-+ __u8 booted_cores; /* number of cores as seen by OS */
-+ __u8 phys_proc_id; /* Physical Processor id. */
-+ __u8 cpu_core_id; /* Core id. */
-+#endif
-+} ____cacheline_aligned;
++void irq_resume(void);
+
-+#define X86_VENDOR_INTEL 0
-+#define X86_VENDOR_CYRIX 1
-+#define X86_VENDOR_AMD 2
-+#define X86_VENDOR_UMC 3
-+#define X86_VENDOR_NEXGEN 4
-+#define X86_VENDOR_CENTAUR 5
-+#define X86_VENDOR_RISE 6
-+#define X86_VENDOR_TRANSMETA 7
-+#define X86_VENDOR_NUM 8
-+#define X86_VENDOR_UNKNOWN 0xff
++/* Entry point for notifications into Linux subsystems. */
++asmlinkage void evtchn_do_upcall(struct pt_regs *regs);
++
++/* Entry point for notifications into the userland character device. */
++void evtchn_device_upcall(int port);
++
++/* Mark a PIRQ as unavailable for dynamic allocation. */
++void evtchn_register_pirq(int irq);
++/* Map a Xen-supplied PIRQ to a dynamically allocated one. */
++int evtchn_map_pirq(int irq, int xen_pirq);
++/* Look up a Xen-supplied PIRQ for a dynamically allocated one. */
++int evtchn_get_xen_pirq(int irq);
++
++void mask_evtchn(int port);
++void disable_all_local_evtchn(void);
++void unmask_evtchn(int port);
+
+#ifdef CONFIG_SMP
-+extern struct cpuinfo_x86 cpu_data[];
-+#define current_cpu_data cpu_data[smp_processor_id()]
++void rebind_evtchn_to_cpu(int port, unsigned int cpu);
+#else
-+#define cpu_data (&boot_cpu_data)
-+#define current_cpu_data boot_cpu_data
++#define rebind_evtchn_to_cpu(port, cpu) ((void)0)
+#endif
+
-+extern char ignore_irq13;
++static inline int test_and_set_evtchn_mask(int port)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ return synch_test_and_set_bit(port, s->evtchn_mask);
++}
+
-+extern void identify_cpu(struct cpuinfo_x86 *);
-+extern void print_cpu_info(struct cpuinfo_x86 *);
-+extern unsigned int init_intel_cacheinfo(struct cpuinfo_x86 *c);
-+extern unsigned short num_cache_leaves;
++static inline void clear_evtchn(int port)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ synch_clear_bit(port, s->evtchn_pending);
++}
+
-+/*
-+ * EFLAGS bits
-+ */
-+#define X86_EFLAGS_CF 0x00000001 /* Carry Flag */
-+#define X86_EFLAGS_PF 0x00000004 /* Parity Flag */
-+#define X86_EFLAGS_AF 0x00000010 /* Auxillary carry Flag */
-+#define X86_EFLAGS_ZF 0x00000040 /* Zero Flag */
-+#define X86_EFLAGS_SF 0x00000080 /* Sign Flag */
-+#define X86_EFLAGS_TF 0x00000100 /* Trap Flag */
-+#define X86_EFLAGS_IF 0x00000200 /* Interrupt Flag */
-+#define X86_EFLAGS_DF 0x00000400 /* Direction Flag */
-+#define X86_EFLAGS_OF 0x00000800 /* Overflow Flag */
-+#define X86_EFLAGS_IOPL 0x00003000 /* IOPL mask */
-+#define X86_EFLAGS_NT 0x00004000 /* Nested Task */
-+#define X86_EFLAGS_RF 0x00010000 /* Resume Flag */
-+#define X86_EFLAGS_VM 0x00020000 /* Virtual Mode */
-+#define X86_EFLAGS_AC 0x00040000 /* Alignment Check */
-+#define X86_EFLAGS_VIF 0x00080000 /* Virtual Interrupt Flag */
-+#define X86_EFLAGS_VIP 0x00100000 /* Virtual Interrupt Pending */
-+#define X86_EFLAGS_ID 0x00200000 /* CPUID detection flag */
++static inline void notify_remote_via_evtchn(int port)
++{
++ struct evtchn_send send = { .port = port };
++ VOID(HYPERVISOR_event_channel_op(EVTCHNOP_send, &send));
++}
+
+/*
-+ * Intel CPU features in CR4
++ * Use these to access the event channel underlying the IRQ handle returned
++ * by bind_*_to_irqhandler().
+ */
-+#define X86_CR4_VME 0x0001 /* enable vm86 extensions */
-+#define X86_CR4_PVI 0x0002 /* virtual interrupts flag enable */
-+#define X86_CR4_TSD 0x0004 /* disable time stamp at ipl 3 */
-+#define X86_CR4_DE 0x0008 /* enable debugging extensions */
-+#define X86_CR4_PSE 0x0010 /* enable page size extensions */
-+#define X86_CR4_PAE 0x0020 /* enable physical address extensions */
-+#define X86_CR4_MCE 0x0040 /* Machine check enable */
-+#define X86_CR4_PGE 0x0080 /* enable global pages */
-+#define X86_CR4_PCE 0x0100 /* enable performance counters at ipl 3 */
-+#define X86_CR4_OSFXSR 0x0200 /* enable fast FPU save and restore */
-+#define X86_CR4_OSXMMEXCPT 0x0400 /* enable unmasked SSE exceptions */
++void notify_remote_via_irq(int irq);
++int irq_to_evtchn_port(int irq);
+
-+/*
-+ * Save the cr4 feature set we're using (ie
-+ * Pentium 4MB enable and PPro Global page
-+ * enable), so that any CPU's that boot up
-+ * after us can get the correct flags.
++#define PIRQ_SET_MAPPING 0x0
++#define PIRQ_CLEAR_MAPPING 0x1
++#define PIRQ_GET_MAPPING 0x3
++int pirq_mapstatus(int pirq, int action);
++int set_pirq_hw_action(int pirq, int (*action)(int pirq, int action));
++int clear_pirq_hw_action(int pirq);
++
++#define PIRQ_STARTUP 1
++#define PIRQ_SHUTDOWN 2
++#define PIRQ_ENABLE 3
++#define PIRQ_DISABLE 4
++#define PIRQ_END 5
++#define PIRQ_ACK 6
++
++#endif /* __ASM_EVTCHN_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/features.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/features.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,20 @@
++/******************************************************************************
++ * features.h
++ *
++ * Query the features reported by Xen.
++ *
++ * Copyright (c) 2006, Ian Campbell
+ */
-+extern unsigned long mmu_cr4_features;
+
-+static inline void set_in_cr4 (unsigned long mask)
-+{
-+ mmu_cr4_features |= mask;
-+ __asm__("movq %%cr4,%%rax\n\t"
-+ "orq %0,%%rax\n\t"
-+ "movq %%rax,%%cr4\n"
-+ : : "irg" (mask)
-+ :"ax");
-+}
++#ifndef __ASM_XEN_FEATURES_H__
++#define __ASM_XEN_FEATURES_H__
+
-+static inline void clear_in_cr4 (unsigned long mask)
-+{
-+ mmu_cr4_features &= ~mask;
-+ __asm__("movq %%cr4,%%rax\n\t"
-+ "andq %0,%%rax\n\t"
-+ "movq %%rax,%%cr4\n"
-+ : : "irg" (~mask)
-+ :"ax");
-+}
++#include <xen/interface/version.h>
++
++extern void setup_xen_features(void);
++
++extern u8 xen_features[XENFEAT_NR_SUBMAPS * 32];
++
++#define xen_feature(flag) (xen_features[flag])
++
++#endif /* __ASM_XEN_FEATURES_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/firmware.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/firmware.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,10 @@
++#ifndef __XEN_FIRMWARE_H__
++#define __XEN_FIRMWARE_H__
++
++#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE)
++void copy_edd(void);
++#endif
++
++void copy_edid(void);
++
++#endif /* __XEN_FIRMWARE_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/gnttab.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/gnttab.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,163 @@
++/******************************************************************************
++ * gnttab.h
++ *
++ * Two sets of functionality:
++ * 1. Granting foreign access to our memory reservation.
++ * 2. Accessing others' memory reservations via grant references.
++ * (i.e., mechanisms for both sender and recipient of grant references)
++ *
++ * Copyright (c) 2004-2005, K A Fraser
++ * Copyright (c) 2005, Christopher Clark
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __ASM_GNTTAB_H__
++#define __ASM_GNTTAB_H__
+
++#include <asm/hypervisor.h>
++#include <asm/maddr.h> /* maddr_t */
++#include <linux/mm.h>
++#include <xen/interface/grant_table.h>
++#include <xen/features.h>
++
++struct gnttab_free_callback {
++ struct gnttab_free_callback *next;
++ void (*fn)(void *);
++ void *arg;
++ u16 count;
++};
++
++int gnttab_grant_foreign_access(domid_t domid, unsigned long frame,
++ int flags);
+
+/*
-+ * Bus types
++ * End access through the given grant reference, iff the grant entry is no
++ * longer in use. Return 1 if the grant entry was freed, 0 if it is still in
++ * use.
+ */
-+#define MCA_bus 0
-+#define MCA_bus__is_a_macro
++int gnttab_end_foreign_access_ref(grant_ref_t ref);
+
+/*
-+ * User space process size. 47bits minus one guard page.
++ * Eventually end access through the given grant reference, and once that
++ * access has been ended, free the given page too. Access will be ended
++ * immediately iff the grant entry is not in use, otherwise it will happen
++ * some time later. page may be 0, in which case no freeing will occur.
+ */
-+#define TASK_SIZE64 (0x800000000000UL - 4096)
++void gnttab_end_foreign_access(grant_ref_t ref, unsigned long page);
+
-+/* This decides where the kernel will search for a free chunk of vm
-+ * space during mmap's.
-+ */
-+#define IA32_PAGE_OFFSET ((current->personality & ADDR_LIMIT_3GB) ? 0xc0000000 : 0xFFFFe000)
++int gnttab_grant_foreign_transfer(domid_t domid, unsigned long pfn);
+
-+#define TASK_SIZE (test_thread_flag(TIF_IA32) ? IA32_PAGE_OFFSET : TASK_SIZE64)
-+#define TASK_SIZE_OF(child) ((test_tsk_thread_flag(child, TIF_IA32)) ? IA32_PAGE_OFFSET : TASK_SIZE64)
++unsigned long gnttab_end_foreign_transfer_ref(grant_ref_t ref);
++unsigned long gnttab_end_foreign_transfer(grant_ref_t ref);
+
-+#define TASK_UNMAPPED_BASE PAGE_ALIGN(TASK_SIZE/3)
++int gnttab_query_foreign_access(grant_ref_t ref);
+
+/*
-+ * Size of io_bitmap.
++ * operations on reserved batches of grant references
+ */
-+#define IO_BITMAP_BITS 65536
-+#define IO_BITMAP_BYTES (IO_BITMAP_BITS/8)
-+#define IO_BITMAP_LONGS (IO_BITMAP_BYTES/sizeof(long))
-+#ifndef CONFIG_X86_NO_TSS
-+#define IO_BITMAP_OFFSET offsetof(struct tss_struct,io_bitmap)
-+#endif
-+#define INVALID_IO_BITMAP_OFFSET 0x8000
++int gnttab_alloc_grant_references(u16 count, grant_ref_t *pprivate_head);
+
-+struct i387_fxsave_struct {
-+ u16 cwd;
-+ u16 swd;
-+ u16 twd;
-+ u16 fop;
-+ u64 rip;
-+ u64 rdp;
-+ u32 mxcsr;
-+ u32 mxcsr_mask;
-+ u32 st_space[32]; /* 8*16 bytes for each FP-reg = 128 bytes */
-+ u32 xmm_space[64]; /* 16*16 bytes for each XMM-reg = 128 bytes */
-+ u32 padding[24];
-+} __attribute__ ((aligned (16)));
++void gnttab_free_grant_reference(grant_ref_t ref);
+
-+union i387_union {
-+ struct i387_fxsave_struct fxsave;
-+};
++void gnttab_free_grant_references(grant_ref_t head);
+
-+#ifndef CONFIG_X86_NO_TSS
-+struct tss_struct {
-+ u32 reserved1;
-+ u64 rsp0;
-+ u64 rsp1;
-+ u64 rsp2;
-+ u64 reserved2;
-+ u64 ist[7];
-+ u32 reserved3;
-+ u32 reserved4;
-+ u16 reserved5;
-+ u16 io_bitmap_base;
-+ /*
-+ * The extra 1 is there because the CPU will access an
-+ * additional byte beyond the end of the IO permission
-+ * bitmap. The extra byte must be all 1 bits, and must
-+ * be within the limit. Thus we have:
-+ *
-+ * 128 bytes, the bitmap itself, for ports 0..0x3ff
-+ * 8 bytes, for an extra "long" of ~0UL
-+ */
-+ unsigned long io_bitmap[IO_BITMAP_LONGS + 1];
-+} __attribute__((packed)) ____cacheline_aligned;
++int gnttab_empty_grant_references(const grant_ref_t *pprivate_head);
+
-+DECLARE_PER_CPU(struct tss_struct,init_tss);
-+#endif
++int gnttab_claim_grant_reference(grant_ref_t *pprivate_head);
+
++void gnttab_release_grant_reference(grant_ref_t *private_head,
++ grant_ref_t release);
+
-+extern struct cpuinfo_x86 boot_cpu_data;
-+#ifndef CONFIG_X86_NO_TSS
-+/* Save the original ist values for checking stack pointers during debugging */
-+struct orig_ist {
-+ unsigned long ist[7];
-+};
-+DECLARE_PER_CPU(struct orig_ist, orig_ist);
-+#endif
++void gnttab_request_free_callback(struct gnttab_free_callback *callback,
++ void (*fn)(void *), void *arg, u16 count);
++void gnttab_cancel_free_callback(struct gnttab_free_callback *callback);
+
-+#ifdef CONFIG_X86_VSMP
-+#define ARCH_MIN_TASKALIGN (1 << INTERNODE_CACHE_SHIFT)
-+#define ARCH_MIN_MMSTRUCT_ALIGN (1 << INTERNODE_CACHE_SHIFT)
-+#else
-+#define ARCH_MIN_TASKALIGN 16
-+#define ARCH_MIN_MMSTRUCT_ALIGN 0
-+#endif
++void gnttab_grant_foreign_access_ref(grant_ref_t ref, domid_t domid,
++ unsigned long frame, int flags);
+
-+struct thread_struct {
-+ unsigned long rsp0;
-+ unsigned long rsp;
-+ unsigned long userrsp; /* Copy from PDA */
-+ unsigned long fs;
-+ unsigned long gs;
-+ unsigned short es, ds, fsindex, gsindex;
-+/* Hardware debugging registers */
-+ unsigned long debugreg0;
-+ unsigned long debugreg1;
-+ unsigned long debugreg2;
-+ unsigned long debugreg3;
-+ unsigned long debugreg6;
-+ unsigned long debugreg7;
-+/* fault info */
-+ unsigned long cr2, trap_no, error_code;
-+/* floating point info */
-+ union i387_union i387 __attribute__((aligned(16)));
-+/* IO permissions. the bitmap could be moved into the GDT, that would make
-+ switch faster for a limited number of ioperm using tasks. -AK */
-+ int ioperm;
-+ unsigned long *io_bitmap_ptr;
-+ unsigned io_bitmap_max;
-+/* cached TLS descriptors. */
-+ u64 tls_array[GDT_ENTRY_TLS_ENTRIES];
-+ unsigned int iopl;
-+} __attribute__((aligned(16)));
++void gnttab_grant_foreign_transfer_ref(grant_ref_t, domid_t domid,
++ unsigned long pfn);
+
-+#define INIT_THREAD { \
-+ .rsp0 = (unsigned long)&init_stack + sizeof(init_stack) \
++int gnttab_copy_grant_page(grant_ref_t ref, struct page **pagep);
++void __gnttab_dma_map_page(struct page *page);
++static inline void __gnttab_dma_unmap_page(struct page *page)
++{
+}
+
-+#ifndef CONFIG_X86_NO_TSS
-+#define INIT_TSS { \
-+ .rsp0 = (unsigned long)&init_stack + sizeof(init_stack) \
++void gnttab_reset_grant_page(struct page *page);
++
++int gnttab_suspend(void);
++int gnttab_resume(void);
++
++void *arch_gnttab_alloc_shared(unsigned long *frames);
++
++static inline void
++gnttab_set_map_op(struct gnttab_map_grant_ref *map, maddr_t addr,
++ uint32_t flags, grant_ref_t ref, domid_t domid)
++{
++ if (flags & GNTMAP_contains_pte)
++ map->host_addr = addr;
++ else if (xen_feature(XENFEAT_auto_translated_physmap))
++ map->host_addr = __pa(addr);
++ else
++ map->host_addr = addr;
++
++ map->flags = flags;
++ map->ref = ref;
++ map->dom = domid;
+}
-+#endif
+
-+#define INIT_MMAP \
-+{ &init_mm, 0, 0, NULL, PAGE_SHARED, VM_READ | VM_WRITE | VM_EXEC, 1, NULL, NULL }
++static inline void
++gnttab_set_unmap_op(struct gnttab_unmap_grant_ref *unmap, maddr_t addr,
++ uint32_t flags, grant_handle_t handle)
++{
++ if (flags & GNTMAP_contains_pte)
++ unmap->host_addr = addr;
++ else if (xen_feature(XENFEAT_auto_translated_physmap))
++ unmap->host_addr = __pa(addr);
++ else
++ unmap->host_addr = addr;
+
-+#define start_thread(regs,new_rip,new_rsp) do { \
-+ asm volatile("movl %0,%%fs; movl %0,%%es; movl %0,%%ds": :"r" (0)); \
-+ load_gs_index(0); \
-+ (regs)->rip = (new_rip); \
-+ (regs)->rsp = (new_rsp); \
-+ write_pda(oldrsp, (new_rsp)); \
-+ (regs)->cs = __USER_CS; \
-+ (regs)->ss = __USER_DS; \
-+ (regs)->eflags = 0x200; \
-+ set_fs(USER_DS); \
-+} while(0)
++ unmap->handle = handle;
++ unmap->dev_bus_addr = 0;
++}
+
-+#define get_debugreg(var, register) \
-+ var = HYPERVISOR_get_debugreg(register)
-+#define set_debugreg(value, register) \
-+ HYPERVISOR_set_debugreg(register, value)
++static inline void
++gnttab_set_replace_op(struct gnttab_unmap_and_replace *unmap, maddr_t addr,
++ maddr_t new_addr, grant_handle_t handle)
++{
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ unmap->host_addr = __pa(addr);
++ unmap->new_addr = __pa(new_addr);
++ } else {
++ unmap->host_addr = addr;
++ unmap->new_addr = new_addr;
++ }
+
-+struct task_struct;
-+struct mm_struct;
++ unmap->handle = handle;
++}
+
-+/* Free all resources held by a thread. */
-+extern void release_thread(struct task_struct *);
++#endif /* __ASM_GNTTAB_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/hvm.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/hvm.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,23 @@
++/* Simple wrappers around HVM functions */
++#ifndef XEN_HVM_H__
++#define XEN_HVM_H__
+
-+/* Prepare to copy thread state - unlazy all lazy status */
-+extern void prepare_to_copy(struct task_struct *tsk);
++#include <xen/interface/hvm/params.h>
+
-+/*
-+ * create a kernel thread without removing it from tasklists
-+ */
-+extern long kernel_thread(int (*fn)(void *), void * arg, unsigned long flags);
++static inline unsigned long hvm_get_parameter(int idx)
++{
++ struct xen_hvm_param xhv;
++ int r;
+
-+/*
-+ * Return saved PC of a blocked thread.
-+ * What is this good for? it will be always the scheduler or ret_from_fork.
-+ */
-+#define thread_saved_pc(t) (*(unsigned long *)((t)->thread.rsp - 8))
++ xhv.domid = DOMID_SELF;
++ xhv.index = idx;
++ r = HYPERVISOR_hvm_op(HVMOP_get_param, &xhv);
++ if (r < 0) {
++ printk(KERN_ERR "cannot get hvm parameter %d: %d.\n",
++ idx, r);
++ return 0;
++ }
++ return xhv.value;
++}
+
-+extern unsigned long get_wchan(struct task_struct *p);
-+#define task_pt_regs(tsk) ((struct pt_regs *)(tsk)->thread.rsp0 - 1)
-+#define KSTK_EIP(tsk) (task_pt_regs(tsk)->rip)
-+#define KSTK_ESP(tsk) -1 /* sorry. doesn't work for syscall. */
++#endif /* XEN_HVM_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/hypercall.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/hypercall.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,30 @@
++#ifndef __XEN_HYPERCALL_H__
++#define __XEN_HYPERCALL_H__
+
++#include <asm/hypercall.h>
+
-+struct microcode_header {
-+ unsigned int hdrver;
-+ unsigned int rev;
-+ unsigned int date;
-+ unsigned int sig;
-+ unsigned int cksum;
-+ unsigned int ldrver;
-+ unsigned int pf;
-+ unsigned int datasize;
-+ unsigned int totalsize;
-+ unsigned int reserved[3];
-+};
++static inline int __must_check
++HYPERVISOR_multicall_check(
++ multicall_entry_t *call_list, unsigned int nr_calls,
++ const unsigned long *rc_list)
++{
++ int rc = HYPERVISOR_multicall(call_list, nr_calls);
+
-+struct microcode {
-+ struct microcode_header hdr;
-+ unsigned int bits[0];
-+};
++ if (unlikely(rc < 0))
++ return rc;
++ BUG_ON(rc);
++ BUG_ON((int)nr_calls < 0);
+
-+typedef struct microcode microcode_t;
-+typedef struct microcode_header microcode_header_t;
++ for ( ; nr_calls > 0; --nr_calls, ++call_list)
++ if (unlikely(call_list->result != (rc_list ? *rc_list++ : 0)))
++ return nr_calls;
+
-+/* microcode format is extended from prescott processors */
-+struct extended_signature {
-+ unsigned int sig;
-+ unsigned int pf;
-+ unsigned int cksum;
-+};
++ return 0;
++}
+
-+struct extended_sigtable {
-+ unsigned int count;
-+ unsigned int cksum;
-+ unsigned int reserved[3];
-+ struct extended_signature sigs[0];
-+};
++/* A construct to ignore the return value of hypercall wrappers in a few
++ * exceptional cases (simply casting the function result to void doesn't
++ * avoid the compiler warning): */
++#define VOID(expr) ((void)((expr)?:0))
+
++#endif /* __XEN_HYPERCALL_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/hypervisor_sysfs.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/hypervisor_sysfs.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,30 @@
++/*
++ * copyright (c) 2006 IBM Corporation
++ * Authored by: Mike D. Day <ncmike at us.ibm.com>
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License version 2 as
++ * published by the Free Software Foundation.
++ */
+
-+#define ASM_NOP1 K8_NOP1
-+#define ASM_NOP2 K8_NOP2
-+#define ASM_NOP3 K8_NOP3
-+#define ASM_NOP4 K8_NOP4
-+#define ASM_NOP5 K8_NOP5
-+#define ASM_NOP6 K8_NOP6
-+#define ASM_NOP7 K8_NOP7
-+#define ASM_NOP8 K8_NOP8
++#ifndef _HYP_SYSFS_H_
++#define _HYP_SYSFS_H_
+
-+/* Opteron nops */
-+#define K8_NOP1 ".byte 0x90\n"
-+#define K8_NOP2 ".byte 0x66,0x90\n"
-+#define K8_NOP3 ".byte 0x66,0x66,0x90\n"
-+#define K8_NOP4 ".byte 0x66,0x66,0x66,0x90\n"
-+#define K8_NOP5 K8_NOP3 K8_NOP2
-+#define K8_NOP6 K8_NOP3 K8_NOP3
-+#define K8_NOP7 K8_NOP4 K8_NOP3
-+#define K8_NOP8 K8_NOP4 K8_NOP4
++#include <linux/kobject.h>
++#include <linux/sysfs.h>
+
-+#define ASM_NOP_MAX 8
++#define HYPERVISOR_ATTR_RO(_name) \
++static struct hyp_sysfs_attr _name##_attr = __ATTR_RO(_name)
+
-+/* REP NOP (PAUSE) is a good thing to insert into busy-wait loops. */
-+static inline void rep_nop(void)
-+{
-+ __asm__ __volatile__("rep;nop": : :"memory");
-+}
++#define HYPERVISOR_ATTR_RW(_name) \
++static struct hyp_sysfs_attr _name##_attr = \
++ __ATTR(_name, 0644, _name##_show, _name##_store)
+
-+/* Stop speculative execution */
-+static inline void sync_core(void)
-+{
-+ int tmp;
-+ asm volatile("cpuid" : "=a" (tmp) : "0" (1) : "ebx","ecx","edx","memory");
-+}
++struct hyp_sysfs_attr {
++ struct attribute attr;
++ ssize_t (*show)(struct hyp_sysfs_attr *, char *);
++ ssize_t (*store)(struct hyp_sysfs_attr *, const char *, size_t);
++ void *hyp_attr_data;
++};
+
-+#define cpu_has_fpu 1
++#endif /* _HYP_SYSFS_H_ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/COPYING
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/COPYING Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,38 @@
++XEN NOTICE
++==========
+
-+#define ARCH_HAS_PREFETCH
-+static inline void prefetch(void *x)
-+{
-+ asm volatile("prefetcht0 %0" :: "m" (*(unsigned long *)x));
-+}
++This copyright applies to all files within this subdirectory and its
++subdirectories:
++ include/public/*.h
++ include/public/hvm/*.h
++ include/public/io/*.h
+
-+#define ARCH_HAS_PREFETCHW 1
-+static inline void prefetchw(void *x)
-+{
-+ alternative_input("prefetcht0 (%1)",
-+ "prefetchw (%1)",
-+ X86_FEATURE_3DNOW,
-+ "r" (x));
-+}
++The intention is that these files can be freely copied into the source
++tree of an operating system when porting that OS to run on Xen. Doing
++so does *not* cause the OS to become subject to the terms of the GPL.
+
-+#define ARCH_HAS_SPINLOCK_PREFETCH 1
++All other files in the Xen source distribution are covered by version
++2 of the GNU General Public License except where explicitly stated
++otherwise within individual source files.
+
-+#define spin_lock_prefetch(x) prefetchw(x)
++ -- Keir Fraser (on behalf of the Xen team)
+
-+#define cpu_relax() rep_nop()
++=====================================================================
++
++Permission is hereby granted, free of charge, to any person obtaining a copy
++of this software and associated documentation files (the "Software"), to
++deal in the Software without restriction, including without limitation the
++rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++sell copies of the Software, and to permit persons to whom the Software is
++furnished to do so, subject to the following conditions:
+
++The above copyright notice and this permission notice shall be included in
++all copies or substantial portions of the Software.
++
++THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++DEALINGS IN THE SOFTWARE.
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/acm.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/acm.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,228 @@
+/*
-+ * NSC/Cyrix CPU configuration register indexes
++ * acm.h: Xen access control module interface defintions
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Reiner Sailer <sailer at watson.ibm.com>
++ * Copyright (c) 2005, International Business Machines Corporation.
+ */
-+#define CX86_CCR0 0xc0
-+#define CX86_CCR1 0xc1
-+#define CX86_CCR2 0xc2
-+#define CX86_CCR3 0xc3
-+#define CX86_CCR4 0xe8
-+#define CX86_CCR5 0xe9
-+#define CX86_CCR6 0xea
-+#define CX86_CCR7 0xeb
-+#define CX86_DIR0 0xfe
-+#define CX86_DIR1 0xff
-+#define CX86_ARR_BASE 0xc4
-+#define CX86_RCR_BASE 0xdc
+
-+/*
-+ * NSC/Cyrix CPU indexed register access macros
++#ifndef _XEN_PUBLIC_ACM_H
++#define _XEN_PUBLIC_ACM_H
++
++#include "xen.h"
++
++/* if ACM_DEBUG defined, all hooks should
++ * print a short trace message (comment it out
++ * when not in testing mode )
+ */
++/* #define ACM_DEBUG */
+
-+#define getCx86(reg) ({ outb((reg), 0x22); inb(0x23); })
++#ifdef ACM_DEBUG
++# define printkd(fmt, args...) printk(fmt,## args)
++#else
++# define printkd(fmt, args...)
++#endif
+
-+#define setCx86(reg, data) do { \
-+ outb((reg), 0x22); \
-+ outb((data), 0x23); \
-+} while (0)
++/* default ssid reference value if not supplied */
++#define ACM_DEFAULT_SSID 0x0
++#define ACM_DEFAULT_LOCAL_SSID 0x0
+
-+static inline void serialize_cpu(void)
-+{
-+ __asm__ __volatile__ ("cpuid" : : : "ax", "bx", "cx", "dx");
-+}
++/* Internal ACM ERROR types */
++#define ACM_OK 0
++#define ACM_UNDEF -1
++#define ACM_INIT_SSID_ERROR -2
++#define ACM_INIT_SOID_ERROR -3
++#define ACM_ERROR -4
+
-+static inline void __monitor(const void *eax, unsigned long ecx,
-+ unsigned long edx)
-+{
-+ /* "monitor %eax,%ecx,%edx;" */
-+ asm volatile(
-+ ".byte 0x0f,0x01,0xc8;"
-+ : :"a" (eax), "c" (ecx), "d"(edx));
-+}
++/* External ACCESS DECISIONS */
++#define ACM_ACCESS_PERMITTED 0
++#define ACM_ACCESS_DENIED -111
++#define ACM_NULL_POINTER_ERROR -200
+
-+static inline void __mwait(unsigned long eax, unsigned long ecx)
-+{
-+ /* "mwait %eax,%ecx;" */
-+ asm volatile(
-+ ".byte 0x0f,0x01,0xc9;"
-+ : :"a" (eax), "c" (ecx));
-+}
++/*
++ Error codes reported in when trying to test for a new policy
++ These error codes are reported in an array of tuples where
++ each error code is followed by a parameter describing the error
++ more closely, such as a domain id.
++*/
++#define ACM_EVTCHN_SHARING_VIOLATION 0x100
++#define ACM_GNTTAB_SHARING_VIOLATION 0x101
++#define ACM_DOMAIN_LOOKUP 0x102
++#define ACM_CHWALL_CONFLICT 0x103
++#define ACM_SSIDREF_IN_USE 0x104
+
-+#define stack_current() \
-+({ \
-+ struct thread_info *ti; \
-+ asm("andq %%rsp,%0; ":"=r" (ti) : "0" (CURRENT_MASK)); \
-+ ti->task; \
-+})
+
-+#define cache_line_size() (boot_cpu_data.x86_cache_alignment)
++/* primary policy in lower 4 bits */
++#define ACM_NULL_POLICY 0
++#define ACM_CHINESE_WALL_POLICY 1
++#define ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY 2
++#define ACM_POLICY_UNDEFINED 15
+
-+extern unsigned long boot_option_idle_override;
-+/* Boot loader type from the setup header */
-+extern int bootloader_type;
++/* combinations have secondary policy component in higher 4bit */
++#define ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY \
++ ((ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY << 4) | ACM_CHINESE_WALL_POLICY)
+
-+#define HAVE_ARCH_PICK_MMAP_LAYOUT 1
++/* policy: */
++#define ACM_POLICY_NAME(X) \
++ ((X) == (ACM_NULL_POLICY)) ? "NULL" : \
++ ((X) == (ACM_CHINESE_WALL_POLICY)) ? "CHINESE WALL" : \
++ ((X) == (ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY)) ? "SIMPLE TYPE ENFORCEMENT" : \
++ ((X) == (ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY)) ? "CHINESE WALL AND SIMPLE TYPE ENFORCEMENT" : \
++ "UNDEFINED"
+
-+#endif /* __ASM_X86_64_PROCESSOR_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/ptrace.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/ptrace.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/ptrace.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/ptrace.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,127 @@
-+#ifndef _X86_64_PTRACE_H
-+#define _X86_64_PTRACE_H
++/* the following policy versions must be increased
++ * whenever the interpretation of the related
++ * policy's data structure changes
++ */
++#define ACM_POLICY_VERSION 3
++#define ACM_CHWALL_VERSION 1
++#define ACM_STE_VERSION 1
+
-+#if defined(__ASSEMBLY__) || defined(__FRAME_OFFSETS)
-+#define R15 0
-+#define R14 8
-+#define R13 16
-+#define R12 24
-+#define RBP 32
-+#define RBX 40
-+/* arguments: interrupts/non tracing syscalls only save upto here*/
-+#define R11 48
-+#define R10 56
-+#define R9 64
-+#define R8 72
-+#define RAX 80
-+#define RCX 88
-+#define RDX 96
-+#define RSI 104
-+#define RDI 112
-+#define ORIG_RAX 120 /* = ERROR */
-+/* end of arguments */
-+/* cpu exception frame or undefined in case of fast syscall. */
-+#define RIP 128
-+#define CS 136
-+#define EFLAGS 144
-+#define RSP 152
-+#define SS 160
-+#define ARGOFFSET R11
-+#endif /* __ASSEMBLY__ */
++/* defines a ssid reference used by xen */
++typedef uint32_t ssidref_t;
+
-+/* top of stack page */
-+#define FRAME_SIZE 168
++/* hooks that are known to domains */
++#define ACMHOOK_none 0
++#define ACMHOOK_sharing 1
+
-+#define PTRACE_OLDSETOPTIONS 21
++/* -------security policy relevant type definitions-------- */
+
-+#ifndef __ASSEMBLY__
++/* type identifier; compares to "equal" or "not equal" */
++typedef uint16_t domaintype_t;
+
-+struct pt_regs {
-+ unsigned long r15;
-+ unsigned long r14;
-+ unsigned long r13;
-+ unsigned long r12;
-+ unsigned long rbp;
-+ unsigned long rbx;
-+/* arguments: non interrupts/non tracing syscalls only save upto here*/
-+ unsigned long r11;
-+ unsigned long r10;
-+ unsigned long r9;
-+ unsigned long r8;
-+ unsigned long rax;
-+ unsigned long rcx;
-+ unsigned long rdx;
-+ unsigned long rsi;
-+ unsigned long rdi;
-+ unsigned long orig_rax;
-+/* end of arguments */
-+/* cpu exception frame or undefined */
-+ unsigned long rip;
-+ unsigned long cs;
-+ unsigned long eflags;
-+ unsigned long rsp;
-+ unsigned long ss;
-+/* top of stack page */
-+};
++/* CHINESE WALL POLICY DATA STRUCTURES
++ *
++ * current accumulated conflict type set:
++ * When a domain is started and has a type that is in
++ * a conflict set, the conflicting types are incremented in
++ * the aggregate set. When a domain is destroyed, the
++ * conflicting types to its type are decremented.
++ * If a domain has multiple types, this procedure works over
++ * all those types.
++ *
++ * conflict_aggregate_set[i] holds the number of
++ * running domains that have a conflict with type i.
++ *
++ * running_types[i] holds the number of running domains
++ * that include type i in their ssidref-referenced type set
++ *
++ * conflict_sets[i][j] is "0" if type j has no conflict
++ * with type i and is "1" otherwise.
++ */
++/* high-16 = version, low-16 = check magic */
++#define ACM_MAGIC 0x0001debc
+
-+#endif
++/* each offset in bytes from start of the struct they
++ * are part of */
+
-+/* Arbitrarily choose the same ptrace numbers as used by the Sparc code. */
-+#define PTRACE_GETREGS 12
-+#define PTRACE_SETREGS 13
-+#define PTRACE_GETFPREGS 14
-+#define PTRACE_SETFPREGS 15
-+#define PTRACE_GETFPXREGS 18
-+#define PTRACE_SETFPXREGS 19
++/* V3 of the policy buffer aded a version structure */
++struct acm_policy_version
++{
++ uint32_t major;
++ uint32_t minor;
++};
+
-+/* only useful for access 32bit programs */
-+#define PTRACE_GET_THREAD_AREA 25
-+#define PTRACE_SET_THREAD_AREA 26
+
-+#define PTRACE_ARCH_PRCTL 30 /* arch_prctl for child */
++/* each buffer consists of all policy information for
++ * the respective policy given in the policy code
++ *
++ * acm_policy_buffer, acm_chwall_policy_buffer,
++ * and acm_ste_policy_buffer need to stay 32-bit aligned
++ * because we create binary policies also with external
++ * tools that assume packed representations (e.g. the java tool)
++ */
++struct acm_policy_buffer {
++ uint32_t policy_version; /* ACM_POLICY_VERSION */
++ uint32_t magic;
++ uint32_t len;
++ uint32_t policy_reference_offset;
++ uint32_t primary_policy_code;
++ uint32_t primary_buffer_offset;
++ uint32_t secondary_policy_code;
++ uint32_t secondary_buffer_offset;
++ struct acm_policy_version xml_pol_version; /* add in V3 */
++};
+
-+#if defined(__KERNEL__) && !defined(__ASSEMBLY__)
-+#define user_mode(regs) (!!((regs)->cs & 3))
-+#define user_mode_vm(regs) user_mode(regs)
-+#define instruction_pointer(regs) ((regs)->rip)
-+#if defined(CONFIG_SMP) && defined(CONFIG_FRAME_POINTER)
-+extern unsigned long profile_pc(struct pt_regs *regs);
-+#else
-+#define profile_pc(regs) instruction_pointer(regs)
-+#endif
+
-+#include <linux/compiler.h>
++struct acm_policy_reference_buffer {
++ uint32_t len;
++};
+
-+void signal_fault(struct pt_regs *regs, void __user *frame, char *where);
++struct acm_chwall_policy_buffer {
++ uint32_t policy_version; /* ACM_CHWALL_VERSION */
++ uint32_t policy_code;
++ uint32_t chwall_max_types;
++ uint32_t chwall_max_ssidrefs;
++ uint32_t chwall_max_conflictsets;
++ uint32_t chwall_ssid_offset;
++ uint32_t chwall_conflict_sets_offset;
++ uint32_t chwall_running_types_offset;
++ uint32_t chwall_conflict_aggregate_offset;
++};
+
-+struct task_struct;
++struct acm_ste_policy_buffer {
++ uint32_t policy_version; /* ACM_STE_VERSION */
++ uint32_t policy_code;
++ uint32_t ste_max_types;
++ uint32_t ste_max_ssidrefs;
++ uint32_t ste_ssid_offset;
++};
+
-+extern unsigned long
-+convert_rip_to_linear(struct task_struct *child, struct pt_regs *regs);
++struct acm_stats_buffer {
++ uint32_t magic;
++ uint32_t len;
++ uint32_t primary_policy_code;
++ uint32_t primary_stats_offset;
++ uint32_t secondary_policy_code;
++ uint32_t secondary_stats_offset;
++};
+
-+enum {
-+ EF_CF = 0x00000001,
-+ EF_PF = 0x00000004,
-+ EF_AF = 0x00000010,
-+ EF_ZF = 0x00000040,
-+ EF_SF = 0x00000080,
-+ EF_TF = 0x00000100,
-+ EF_IE = 0x00000200,
-+ EF_DF = 0x00000400,
-+ EF_OF = 0x00000800,
-+ EF_IOPL = 0x00003000,
-+ EF_IOPL_RING0 = 0x00000000,
-+ EF_IOPL_RING1 = 0x00001000,
-+ EF_IOPL_RING2 = 0x00002000,
-+ EF_NT = 0x00004000, /* nested task */
-+ EF_RF = 0x00010000, /* resume */
-+ EF_VM = 0x00020000, /* virtual mode */
-+ EF_AC = 0x00040000, /* alignment */
-+ EF_VIF = 0x00080000, /* virtual interrupt */
-+ EF_VIP = 0x00100000, /* virtual interrupt pending */
-+ EF_ID = 0x00200000, /* id */
++struct acm_ste_stats_buffer {
++ uint32_t ec_eval_count;
++ uint32_t gt_eval_count;
++ uint32_t ec_denied_count;
++ uint32_t gt_denied_count;
++ uint32_t ec_cachehit_count;
++ uint32_t gt_cachehit_count;
+};
+
-+#endif
++struct acm_ssid_buffer {
++ uint32_t len;
++ ssidref_t ssidref;
++ uint32_t policy_reference_offset;
++ uint32_t primary_policy_code;
++ uint32_t primary_max_types;
++ uint32_t primary_types_offset;
++ uint32_t secondary_policy_code;
++ uint32_t secondary_max_types;
++ uint32_t secondary_types_offset;
++};
+
+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/smp.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/smp.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/smp.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/smp.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,150 @@
-+#ifndef __ASM_SMP_H
-+#define __ASM_SMP_H
+
+/*
-+ * We need the APIC definitions automatically as part of 'smp.h'
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/acm_ops.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/acm_ops.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,159 @@
++/*
++ * acm_ops.h: Xen access control module hypervisor commands
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Reiner Sailer <sailer at watson.ibm.com>
++ * Copyright (c) 2005,2006 International Business Machines Corporation.
+ */
-+#ifndef __ASSEMBLY__
-+#include <linux/threads.h>
-+#include <linux/cpumask.h>
-+#include <linux/bitops.h>
-+extern int disable_apic;
-+#endif
-+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+#ifndef __ASSEMBLY__
-+#include <asm/fixmap.h>
-+#include <asm/mpspec.h>
-+#ifdef CONFIG_X86_IO_APIC
-+#include <asm/io_apic.h>
-+#endif
-+#include <asm/apic.h>
-+#include <asm/thread_info.h>
-+#endif
-+#endif
+
-+#ifdef CONFIG_SMP
-+#ifndef ASSEMBLY
++#ifndef __XEN_PUBLIC_ACM_OPS_H__
++#define __XEN_PUBLIC_ACM_OPS_H__
+
-+#include <asm/pda.h>
++#include "xen.h"
++#include "acm.h"
+
-+struct pt_regs;
++/*
++ * Make sure you increment the interface version whenever you modify this file!
++ * This makes sure that old versions of acm tools will stop working in a
++ * well-defined way (rather than crashing the machine, for instance).
++ */
++#define ACM_INTERFACE_VERSION 0xAAAA000A
+
-+extern cpumask_t cpu_present_mask;
-+extern cpumask_t cpu_possible_map;
-+extern cpumask_t cpu_online_map;
-+extern cpumask_t cpu_initialized;
++/************************************************************************/
+
+/*
-+ * Private routines/data
++ * Prototype for this hypercall is:
++ * int acm_op(int cmd, void *args)
++ * @cmd == ACMOP_??? (access control module operation).
++ * @args == Operation-specific extra arguments (NULL if none).
+ */
-+
-+extern void smp_alloc_memory(void);
-+extern volatile unsigned long smp_invalidate_needed;
-+extern int pic_mode;
-+extern void lock_ipi_call_lock(void);
-+extern void unlock_ipi_call_lock(void);
-+extern int smp_num_siblings;
-+extern void smp_send_reschedule(int cpu);
-+void smp_stop_cpu(void);
-+extern int smp_call_function_single(int cpuid, void (*func) (void *info),
-+ void *info, int retry, int wait);
+
-+extern cpumask_t cpu_sibling_map[NR_CPUS];
-+extern cpumask_t cpu_core_map[NR_CPUS];
-+extern u8 cpu_llc_id[NR_CPUS];
+
-+#define SMP_TRAMPOLINE_BASE 0x6000
++#define ACMOP_setpolicy 1
++struct acm_setpolicy {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) pushcache;
++ uint32_t pushcache_size;
++};
+
-+/*
-+ * On x86 all CPUs are mapped 1:1 to the APIC space.
-+ * This simplifies scheduling and IPI sending and
-+ * compresses data structures.
-+ */
+
-+static inline int num_booting_cpus(void)
-+{
-+ return cpus_weight(cpu_possible_map);
-+}
++#define ACMOP_getpolicy 2
++struct acm_getpolicy {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) pullcache;
++ uint32_t pullcache_size;
++};
+
-+#define raw_smp_processor_id() read_pda(cpunumber)
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+static inline int hard_smp_processor_id(void)
-+{
-+ /* we don't want to mark this access volatile - bad code generation */
-+ return GET_APIC_ID(*(unsigned int *)(APIC_BASE+APIC_ID));
-+}
-+#endif
++#define ACMOP_dumpstats 3
++struct acm_dumpstats {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) pullcache;
++ uint32_t pullcache_size;
++};
+
-+extern int safe_smp_processor_id(void);
-+extern int __cpu_disable(void);
-+extern void __cpu_die(unsigned int cpu);
-+extern void prefill_possible_map(void);
-+extern unsigned num_processors;
-+extern unsigned disabled_cpus;
+
-+#endif /* !ASSEMBLY */
++#define ACMOP_getssid 4
++#define ACM_GETBY_ssidref 1
++#define ACM_GETBY_domainid 2
++struct acm_getssid {
++ /* IN */
++ uint32_t get_ssid_by; /* ACM_GETBY_* */
++ union {
++ domaintype_t domainid;
++ ssidref_t ssidref;
++ } id;
++ XEN_GUEST_HANDLE_64(void) ssidbuf;
++ uint32_t ssidbuf_size;
++};
+
-+#define NO_PROC_ID 0xFF /* No processor magic marker */
++#define ACMOP_getdecision 5
++struct acm_getdecision {
++ /* IN */
++ uint32_t get_decision_by1; /* ACM_GETBY_* */
++ uint32_t get_decision_by2; /* ACM_GETBY_* */
++ union {
++ domaintype_t domainid;
++ ssidref_t ssidref;
++ } id1;
++ union {
++ domaintype_t domainid;
++ ssidref_t ssidref;
++ } id2;
++ uint32_t hook;
++ /* OUT */
++ uint32_t acm_decision;
++};
+
-+#endif
+
-+#ifndef ASSEMBLY
-+/*
-+ * Some lowlevel functions might want to know about
-+ * the real APIC ID <-> CPU # mapping.
-+ */
-+extern u8 x86_cpu_to_apicid[NR_CPUS]; /* physical ID */
-+extern u8 x86_cpu_to_log_apicid[NR_CPUS];
-+extern u8 bios_cpu_apicid[];
++#define ACMOP_chgpolicy 6
++struct acm_change_policy {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) policy_pushcache;
++ uint32_t policy_pushcache_size;
++ XEN_GUEST_HANDLE_64(void) del_array;
++ uint32_t delarray_size;
++ XEN_GUEST_HANDLE_64(void) chg_array;
++ uint32_t chgarray_size;
++ /* OUT */
++ /* array with error code */
++ XEN_GUEST_HANDLE_64(void) err_array;
++ uint32_t errarray_size;
++};
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+static inline unsigned int cpu_mask_to_apicid(cpumask_t cpumask)
-+{
-+ return cpus_addr(cpumask)[0];
-+}
++#define ACMOP_relabeldoms 7
++struct acm_relabel_doms {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) relabel_map;
++ uint32_t relabel_map_size;
++ /* OUT */
++ XEN_GUEST_HANDLE_64(void) err_array;
++ uint32_t errarray_size;
++};
+
-+static inline int cpu_present_to_apicid(int mps_cpu)
-+{
-+ if (mps_cpu < NR_CPUS)
-+ return (int)bios_cpu_apicid[mps_cpu];
-+ else
-+ return BAD_APICID;
-+}
-+#endif
++/* future interface to Xen */
++struct xen_acmctl {
++ uint32_t cmd;
++ uint32_t interface_version;
++ union {
++ struct acm_setpolicy setpolicy;
++ struct acm_getpolicy getpolicy;
++ struct acm_dumpstats dumpstats;
++ struct acm_getssid getssid;
++ struct acm_getdecision getdecision;
++ struct acm_change_policy change_policy;
++ struct acm_relabel_doms relabel_doms;
++ } u;
++};
+
-+#endif /* !ASSEMBLY */
++typedef struct xen_acmctl xen_acmctl_t;
++DEFINE_XEN_GUEST_HANDLE(xen_acmctl_t);
+
-+#ifndef CONFIG_SMP
-+#define stack_smp_processor_id() 0
-+#define safe_smp_processor_id() 0
-+#define cpu_logical_map(x) (x)
-+#else
-+#include <asm/thread_info.h>
-+#define stack_smp_processor_id() \
-+({ \
-+ struct thread_info *ti; \
-+ __asm__("andq %%rsp,%0; ":"=r" (ti) : "0" (CURRENT_MASK)); \
-+ ti->cpu; \
-+})
-+#endif
++#endif /* __XEN_PUBLIC_ACM_OPS_H__ */
+
-+#ifndef __ASSEMBLY__
-+#ifdef CONFIG_X86_LOCAL_APIC
-+static __inline int logical_smp_processor_id(void)
-+{
-+ /* we don't want to mark this access volatile - bad code generation */
-+ return GET_APIC_LOGICAL_ID(*(unsigned long *)(APIC_BASE+APIC_LDR));
-+}
-+#endif
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-ia64.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-ia64.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,621 @@
++/******************************************************************************
++ * arch-ia64/hypervisor-if.h
++ *
++ * Guest OS interface to IA64 Xen.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ */
++
++#include "xen.h"
++
++#ifndef __HYPERVISOR_IF_IA64_H__
++#define __HYPERVISOR_IF_IA64_H__
++
++#if !defined(__GNUC__) || defined(__STRICT_ANSI__)
++#error "Anonymous structs/unions are a GNU extension."
+#endif
+
-+#ifdef CONFIG_SMP
-+#define cpu_physical_id(cpu) x86_cpu_to_apicid[cpu]
++/* Structural guest handles introduced in 0x00030201. */
++#if __XEN_INTERFACE_VERSION__ >= 0x00030201
++#define ___DEFINE_XEN_GUEST_HANDLE(name, type) \
++ typedef struct { type *p; } __guest_handle_ ## name
+#else
-+#define cpu_physical_id(cpu) boot_cpu_id
++#define ___DEFINE_XEN_GUEST_HANDLE(name, type) \
++ typedef type * __guest_handle_ ## name
+#endif
+
++#define __DEFINE_XEN_GUEST_HANDLE(name, type) \
++ ___DEFINE_XEN_GUEST_HANDLE(name, type); \
++ ___DEFINE_XEN_GUEST_HANDLE(const_##name, const type)
++
++#define DEFINE_XEN_GUEST_HANDLE(name) __DEFINE_XEN_GUEST_HANDLE(name, name)
++#define XEN_GUEST_HANDLE(name) __guest_handle_ ## name
++#define XEN_GUEST_HANDLE_64(name) XEN_GUEST_HANDLE(name)
++#define uint64_aligned_t uint64_t
++#define set_xen_guest_handle(hnd, val) do { (hnd).p = val; } while (0)
++#ifdef __XEN_TOOLS__
++#define get_xen_guest_handle(val, hnd) do { val = (hnd).p; } while (0)
+#endif
+
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/synch_bitops.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/synch_bitops.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/synch_bitops.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/synch_bitops.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,2 @@
++#ifndef __ASSEMBLY__
++typedef unsigned long xen_pfn_t;
++#define PRI_xen_pfn "lx"
++#endif
+
-+#include <asm-i386/mach-xen/asm/synch_bitops.h>
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/system.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/system.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/system.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/system.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,262 @@
-+#ifndef __ASM_SYSTEM_H
-+#define __ASM_SYSTEM_H
++/* Arch specific VIRQs definition */
++#define VIRQ_ITC VIRQ_ARCH_0 /* V. Virtual itc timer */
++#define VIRQ_MCA_CMC VIRQ_ARCH_1 /* MCA cmc interrupt */
++#define VIRQ_MCA_CPE VIRQ_ARCH_2 /* MCA cpe interrupt */
+
-+#include <linux/kernel.h>
-+#include <asm/segment.h>
-+#include <asm/alternative.h>
++/* Maximum number of virtual CPUs in multi-processor guests. */
++/* WARNING: before changing this, check that shared_info fits on a page */
++#define MAX_VIRT_CPUS 64
+
-+#include <asm/synch_bitops.h>
-+#include <asm/hypervisor.h>
-+#include <xen/interface/arch-x86_64.h>
++/* IO ports location for PV. */
++#define IO_PORTS_PADDR 0x00000ffffc000000UL
++#define IO_PORTS_SIZE 0x0000000004000000UL
+
-+#ifdef __KERNEL__
++#ifndef __ASSEMBLY__
+
-+#ifdef CONFIG_SMP
-+#define __vcpu_id smp_processor_id()
++typedef unsigned long xen_ulong_t;
++
++#ifdef __XEN_TOOLS__
++#define XEN_PAGE_SIZE XC_PAGE_SIZE
+#else
-+#define __vcpu_id 0
++#define XEN_PAGE_SIZE PAGE_SIZE
+#endif
+
-+#define __STR(x) #x
-+#define STR(x) __STR(x)
++#define INVALID_MFN (~0UL)
+
-+#define __SAVE(reg,offset) "movq %%" #reg ",(14-" #offset ")*8(%%rsp)\n\t"
-+#define __RESTORE(reg,offset) "movq (14-" #offset ")*8(%%rsp),%%" #reg "\n\t"
++struct pt_fpreg {
++ union {
++ unsigned long bits[2];
++ long double __dummy; /* force 16-byte alignment */
++ } u;
++};
+
-+/* frame pointer must be last for get_wchan */
-+#define SAVE_CONTEXT "pushq %%rbp ; movq %%rsi,%%rbp\n\t"
-+#define RESTORE_CONTEXT "movq %%rbp,%%rsi ; popq %%rbp\n\t"
++union vac {
++ unsigned long value;
++ struct {
++ int a_int:1;
++ int a_from_int_cr:1;
++ int a_to_int_cr:1;
++ int a_from_psr:1;
++ int a_from_cpuid:1;
++ int a_cover:1;
++ int a_bsw:1;
++ long reserved:57;
++ };
++};
++typedef union vac vac_t;
+
-+#define __EXTRA_CLOBBER \
-+ ,"rcx","rbx","rdx","r8","r9","r10","r11","r12","r13","r14","r15"
++union vdc {
++ unsigned long value;
++ struct {
++ int d_vmsw:1;
++ int d_extint:1;
++ int d_ibr_dbr:1;
++ int d_pmc:1;
++ int d_to_pmd:1;
++ int d_itm:1;
++ long reserved:58;
++ };
++};
++typedef union vdc vdc_t;
+
-+#define switch_to(prev,next,last) \
-+ asm volatile(SAVE_CONTEXT \
-+ "movq %%rsp,%P[threadrsp](%[prev])\n\t" /* save RSP */ \
-+ "movq %P[threadrsp](%[next]),%%rsp\n\t" /* restore RSP */ \
-+ "call __switch_to\n\t" \
-+ ".globl thread_return\n" \
-+ "thread_return:\n\t" \
-+ "movq %%gs:%P[pda_pcurrent],%%rsi\n\t" \
-+ "movq %P[thread_info](%%rsi),%%r8\n\t" \
-+ LOCK_PREFIX "btr %[tif_fork],%P[ti_flags](%%r8)\n\t" \
-+ "movq %%rax,%%rdi\n\t" \
-+ "jc ret_from_fork\n\t" \
-+ RESTORE_CONTEXT \
-+ : "=a" (last) \
-+ : [next] "S" (next), [prev] "D" (prev), \
-+ [threadrsp] "i" (offsetof(struct task_struct, thread.rsp)), \
-+ [ti_flags] "i" (offsetof(struct thread_info, flags)),\
-+ [tif_fork] "i" (TIF_FORK), \
-+ [thread_info] "i" (offsetof(struct task_struct, thread_info)), \
-+ [pda_pcurrent] "i" (offsetof(struct x8664_pda, pcurrent)) \
-+ : "memory", "cc" __EXTRA_CLOBBER)
-+
-+extern void load_gs_index(unsigned);
++struct mapped_regs {
++ union vac vac;
++ union vdc vdc;
++ unsigned long virt_env_vaddr;
++ unsigned long reserved1[29];
++ unsigned long vhpi;
++ unsigned long reserved2[95];
++ union {
++ unsigned long vgr[16];
++ unsigned long bank1_regs[16]; // bank1 regs (r16-r31) when bank0 active
++ };
++ union {
++ unsigned long vbgr[16];
++ unsigned long bank0_regs[16]; // bank0 regs (r16-r31) when bank1 active
++ };
++ unsigned long vnat;
++ unsigned long vbnat;
++ unsigned long vcpuid[5];
++ unsigned long reserved3[11];
++ unsigned long vpsr;
++ unsigned long vpr;
++ unsigned long reserved4[76];
++ union {
++ unsigned long vcr[128];
++ struct {
++ unsigned long dcr; // CR0
++ unsigned long itm;
++ unsigned long iva;
++ unsigned long rsv1[5];
++ unsigned long pta; // CR8
++ unsigned long rsv2[7];
++ unsigned long ipsr; // CR16
++ unsigned long isr;
++ unsigned long rsv3;
++ unsigned long iip;
++ unsigned long ifa;
++ unsigned long itir;
++ unsigned long iipa;
++ unsigned long ifs;
++ unsigned long iim; // CR24
++ unsigned long iha;
++ unsigned long rsv4[38];
++ unsigned long lid; // CR64
++ unsigned long ivr;
++ unsigned long tpr;
++ unsigned long eoi;
++ unsigned long irr[4];
++ unsigned long itv; // CR72
++ unsigned long pmv;
++ unsigned long cmcv;
++ unsigned long rsv5[5];
++ unsigned long lrr0; // CR80
++ unsigned long lrr1;
++ unsigned long rsv6[46];
++ };
++ };
++ union {
++ unsigned long reserved5[128];
++ struct {
++ unsigned long precover_ifs;
++ unsigned long unat; // not sure if this is needed until NaT arch is done
++ int interrupt_collection_enabled; // virtual psr.ic
++ /* virtual interrupt deliverable flag is evtchn_upcall_mask in
++ * shared info area now. interrupt_mask_addr is the address
++ * of evtchn_upcall_mask for current vcpu
++ */
++ unsigned char *interrupt_mask_addr;
++ int pending_interruption;
++ unsigned char vpsr_pp;
++ unsigned char vpsr_dfh;
++ unsigned char hpsr_dfh;
++ unsigned char hpsr_mfh;
++ unsigned long reserved5_1[4];
++ int metaphysical_mode; // 1 = use metaphys mapping, 0 = use virtual
++ int banknum; // 0 or 1, which virtual register bank is active
++ unsigned long rrs[8]; // region registers
++ unsigned long krs[8]; // kernel registers
++ unsigned long tmp[16]; // temp registers (e.g. for hyperprivops)
++ };
++ };
++};
++typedef struct mapped_regs mapped_regs_t;
+
-+/*
-+ * Load a segment. Fall back on loading the zero
-+ * segment if something goes wrong..
-+ */
-+#define loadsegment(seg,value) \
-+ asm volatile("\n" \
-+ "1:\t" \
-+ "movl %k0,%%" #seg "\n" \
-+ "2:\n" \
-+ ".section .fixup,\"ax\"\n" \
-+ "3:\t" \
-+ "movl %1,%%" #seg "\n\t" \
-+ "jmp 2b\n" \
-+ ".previous\n" \
-+ ".section __ex_table,\"a\"\n\t" \
-+ ".align 8\n\t" \
-+ ".quad 1b,3b\n" \
-+ ".previous" \
-+ : :"r" (value), "r" (0))
++struct vpd {
++ struct mapped_regs vpd_low;
++ unsigned long reserved6[3456];
++ unsigned long vmm_avail[128];
++ unsigned long reserved7[4096];
++};
++typedef struct vpd vpd_t;
++
++struct arch_vcpu_info {
++};
++typedef struct arch_vcpu_info arch_vcpu_info_t;
+
+/*
-+ * Clear and set 'TS' bit respectively
-+ */
-+#define clts() (HYPERVISOR_fpu_taskswitch(0))
++ * This structure is used for magic page in domain pseudo physical address
++ * space and the result of XENMEM_machine_memory_map.
++ * As the XENMEM_machine_memory_map result,
++ * xen_memory_map::nr_entries indicates the size in bytes
++ * including struct xen_ia64_memmap_info. Not the number of entries.
++ */
++struct xen_ia64_memmap_info {
++ uint64_t efi_memmap_size; /* size of EFI memory map */
++ uint64_t efi_memdesc_size; /* size of an EFI memory map descriptor */
++ uint32_t efi_memdesc_version; /* memory descriptor version */
++ void *memdesc[0]; /* array of efi_memory_desc_t */
++};
++typedef struct xen_ia64_memmap_info xen_ia64_memmap_info_t;
+
-+static inline unsigned long read_cr0(void)
-+{
-+ unsigned long cr0;
-+ asm volatile("movq %%cr0,%0" : "=r" (cr0));
-+ return cr0;
-+}
++struct arch_shared_info {
++ /* PFN of the start_info page. */
++ unsigned long start_info_pfn;
+
-+static inline void write_cr0(unsigned long val)
-+{
-+ asm volatile("movq %0,%%cr0" :: "r" (val));
-+}
++ /* Interrupt vector for event channel. */
++ int evtchn_vector;
+
-+#define read_cr3() ({ \
-+ unsigned long __dummy; \
-+ asm("movq %%cr3,%0" : "=r" (__dummy)); \
-+ machine_to_phys(__dummy); \
-+})
++ /* PFN of memmap_info page */
++ unsigned int memmap_info_num_pages;/* currently only = 1 case is
++ supported. */
++ unsigned long memmap_info_pfn;
+
-+static inline unsigned long read_cr4(void)
-+{
-+ unsigned long cr4;
-+ asm("movq %%cr4,%0" : "=r" (cr4));
-+ return cr4;
-+}
++ uint64_t pad[31];
++};
++typedef struct arch_shared_info arch_shared_info_t;
+
-+static inline void write_cr4(unsigned long val)
-+{
-+ asm volatile("movq %0,%%cr4" :: "r" (val));
-+}
++typedef unsigned long xen_callback_t;
+
-+#define stts() (HYPERVISOR_fpu_taskswitch(1))
++struct ia64_tr_entry {
++ unsigned long pte;
++ unsigned long itir;
++ unsigned long vadr;
++ unsigned long rid;
++};
++typedef struct ia64_tr_entry ia64_tr_entry_t;
++DEFINE_XEN_GUEST_HANDLE(ia64_tr_entry_t);
+
-+#define wbinvd() \
-+ __asm__ __volatile__ ("wbinvd": : :"memory");
++struct vcpu_tr_regs {
++ struct ia64_tr_entry itrs[12];
++ struct ia64_tr_entry dtrs[12];
++};
+
-+/*
-+ * On SMP systems, when the scheduler does migration-cost autodetection,
-+ * it needs a way to flush as much of the CPU's caches as possible.
-+ */
-+static inline void sched_cacheflush(void)
-+{
-+ wbinvd();
-+}
++union vcpu_ar_regs {
++ unsigned long ar[128];
++ struct {
++ unsigned long kr[8];
++ unsigned long rsv1[8];
++ unsigned long rsc;
++ unsigned long bsp;
++ unsigned long bspstore;
++ unsigned long rnat;
++ unsigned long rsv2;
++ unsigned long fcr;
++ unsigned long rsv3[2];
++ unsigned long eflag;
++ unsigned long csd;
++ unsigned long ssd;
++ unsigned long cflg;
++ unsigned long fsr;
++ unsigned long fir;
++ unsigned long fdr;
++ unsigned long rsv4;
++ unsigned long ccv; /* 32 */
++ unsigned long rsv5[3];
++ unsigned long unat;
++ unsigned long rsv6[3];
++ unsigned long fpsr;
++ unsigned long rsv7[3];
++ unsigned long itc;
++ unsigned long rsv8[3];
++ unsigned long ign1[16];
++ unsigned long pfs; /* 64 */
++ unsigned long lc;
++ unsigned long ec;
++ unsigned long rsv9[45];
++ unsigned long ign2[16];
++ };
++};
+
-+#endif /* __KERNEL__ */
++union vcpu_cr_regs {
++ unsigned long cr[128];
++ struct {
++ unsigned long dcr; // CR0
++ unsigned long itm;
++ unsigned long iva;
++ unsigned long rsv1[5];
++ unsigned long pta; // CR8
++ unsigned long rsv2[7];
++ unsigned long ipsr; // CR16
++ unsigned long isr;
++ unsigned long rsv3;
++ unsigned long iip;
++ unsigned long ifa;
++ unsigned long itir;
++ unsigned long iipa;
++ unsigned long ifs;
++ unsigned long iim; // CR24
++ unsigned long iha;
++ unsigned long rsv4[38];
++ unsigned long lid; // CR64
++ unsigned long ivr;
++ unsigned long tpr;
++ unsigned long eoi;
++ unsigned long irr[4];
++ unsigned long itv; // CR72
++ unsigned long pmv;
++ unsigned long cmcv;
++ unsigned long rsv5[5];
++ unsigned long lrr0; // CR80
++ unsigned long lrr1;
++ unsigned long rsv6[46];
++ };
++};
+
-+#define nop() __asm__ __volatile__ ("nop")
++struct vcpu_guest_context_regs {
++ unsigned long r[32];
++ unsigned long b[8];
++ unsigned long bank[16];
++ unsigned long ip;
++ unsigned long psr;
++ unsigned long cfm;
++ unsigned long pr;
++ unsigned int nats; /* NaT bits for r1-r31. */
++ unsigned int bnats; /* Nat bits for banked registers. */
++ union vcpu_ar_regs ar;
++ union vcpu_cr_regs cr;
++ struct pt_fpreg f[128];
++ unsigned long dbr[8];
++ unsigned long ibr[8];
++ unsigned long rr[8];
++ unsigned long pkr[16];
++
++ /* FIXME: cpuid,pmd,pmc */
++
++ unsigned long xip;
++ unsigned long xpsr;
++ unsigned long xfs;
++ unsigned long xr[4];
++
++ struct vcpu_tr_regs tr;
++
++ /* Physical registers in case of debug event. */
++ unsigned long excp_iipa;
++ unsigned long excp_ifa;
++ unsigned long excp_isr;
++ unsigned int excp_vector;
+
-+#define xchg(ptr,v) ((__typeof__(*(ptr)))__xchg((unsigned long)(v),(ptr),sizeof(*(ptr))))
++ /*
++ * The rbs is intended to be the image of the stacked registers still
++ * in the cpu (not yet stored in memory). It is laid out as if it
++ * were written in memory at a 512 (64*8) aligned address + offset.
++ * rbs_voff is (offset / 8). rbs_nat contains NaT bits for the
++ * remaining rbs registers. rbs_rnat contains NaT bits for in memory
++ * rbs registers.
++ * Note: loadrs is 2**14 bytes == 2**11 slots.
++ */
++ unsigned int rbs_voff;
++ unsigned long rbs[2048];
++ unsigned long rbs_rnat;
+
-+#define tas(ptr) (xchg((ptr),1))
++ /*
++ * RSE.N_STACKED_PHYS via PAL_RSE_INFO
++ * Strictly this isn't cpu context, but this value is necessary
++ * for domain save/restore. So is here.
++ */
++ unsigned long num_phys_stacked;
++};
+
-+#define __xg(x) ((volatile long *)(x))
++struct vcpu_guest_context {
++#define VGCF_EXTRA_REGS (1UL << 1) /* Set extra regs. */
++#define VGCF_SET_CR_IRR (1UL << 2) /* Set cr_irr[0:3]. */
++#define VGCF_online (1UL << 3) /* make this vcpu online */
++ unsigned long flags; /* VGCF_* flags */
+
-+static inline void set_64bit(volatile unsigned long *ptr, unsigned long val)
-+{
-+ *ptr = val;
-+}
++ struct vcpu_guest_context_regs regs;
+
-+#define _set_64bit set_64bit
++ unsigned long event_callback_ip;
+
-+/*
-+ * Note: no "lock" prefix even on SMP: xchg always implies lock anyway
-+ * Note 2: xchg has side effect, so that attribute volatile is necessary,
-+ * but generally the primitive is invalid, *ptr is output argument. --ANK
-+ */
-+static inline unsigned long __xchg(unsigned long x, volatile void * ptr, int size)
-+{
-+ switch (size) {
-+ case 1:
-+ __asm__ __volatile__("xchgb %b0,%1"
-+ :"=q" (x)
-+ :"m" (*__xg(ptr)), "0" (x)
-+ :"memory");
-+ break;
-+ case 2:
-+ __asm__ __volatile__("xchgw %w0,%1"
-+ :"=r" (x)
-+ :"m" (*__xg(ptr)), "0" (x)
-+ :"memory");
-+ break;
-+ case 4:
-+ __asm__ __volatile__("xchgl %k0,%1"
-+ :"=r" (x)
-+ :"m" (*__xg(ptr)), "0" (x)
-+ :"memory");
-+ break;
-+ case 8:
-+ __asm__ __volatile__("xchgq %0,%1"
-+ :"=r" (x)
-+ :"m" (*__xg(ptr)), "0" (x)
-+ :"memory");
-+ break;
-+ }
-+ return x;
-+}
++ /* xen doesn't share privregs pages with hvm domain so that this member
++ * doesn't make sense for hvm domain.
++ * ~0UL is already used for INVALID_P2M_ENTRY. */
++#define VGC_PRIVREGS_HVM (~(-2UL))
++ unsigned long privregs_pfn;
++};
++typedef struct vcpu_guest_context vcpu_guest_context_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_guest_context_t);
+
-+/*
-+ * Atomic compare and exchange. Compare OLD with MEM, if identical,
-+ * store NEW in MEM. Return the initial value in MEM. Success is
-+ * indicated by comparing RETURN with OLD.
-+ */
++/* dom0 vp op */
++#define __HYPERVISOR_ia64_dom0vp_op __HYPERVISOR_arch_0
++/* Map io space in machine address to dom0 physical address space.
++ Currently physical assigned address equals to machine address. */
++#define IA64_DOM0VP_ioremap 0
+
-+#define __HAVE_ARCH_CMPXCHG 1
++/* Convert a pseudo physical page frame number to the corresponding
++ machine page frame number. If no page is assigned, INVALID_MFN or
++ GPFN_INV_MASK is returned depending on domain's non-vti/vti mode. */
++#define IA64_DOM0VP_phystomach 1
+
-+static inline unsigned long __cmpxchg(volatile void *ptr, unsigned long old,
-+ unsigned long new, int size)
-+{
-+ unsigned long prev;
-+ switch (size) {
-+ case 1:
-+ __asm__ __volatile__(LOCK_PREFIX "cmpxchgb %b1,%2"
-+ : "=a"(prev)
-+ : "q"(new), "m"(*__xg(ptr)), "0"(old)
-+ : "memory");
-+ return prev;
-+ case 2:
-+ __asm__ __volatile__(LOCK_PREFIX "cmpxchgw %w1,%2"
-+ : "=a"(prev)
-+ : "r"(new), "m"(*__xg(ptr)), "0"(old)
-+ : "memory");
-+ return prev;
-+ case 4:
-+ __asm__ __volatile__(LOCK_PREFIX "cmpxchgl %k1,%2"
-+ : "=a"(prev)
-+ : "r"(new), "m"(*__xg(ptr)), "0"(old)
-+ : "memory");
-+ return prev;
-+ case 8:
-+ __asm__ __volatile__(LOCK_PREFIX "cmpxchgq %1,%2"
-+ : "=a"(prev)
-+ : "r"(new), "m"(*__xg(ptr)), "0"(old)
-+ : "memory");
-+ return prev;
-+ }
-+ return old;
-+}
++/* Convert a machine page frame number to the corresponding pseudo physical
++ page frame number of the caller domain. */
++#define IA64_DOM0VP_machtophys 3
+
-+#define cmpxchg(ptr,o,n)\
-+ ((__typeof__(*(ptr)))__cmpxchg((ptr),(unsigned long)(o),\
-+ (unsigned long)(n),sizeof(*(ptr))))
++/* Reserved for future use. */
++#define IA64_DOM0VP_iounmap 4
+
-+#ifdef CONFIG_SMP
-+#define smp_mb() mb()
-+#define smp_rmb() rmb()
-+#define smp_wmb() wmb()
-+#define smp_read_barrier_depends() do {} while(0)
-+#else
-+#define smp_mb() barrier()
-+#define smp_rmb() barrier()
-+#define smp_wmb() barrier()
-+#define smp_read_barrier_depends() do {} while(0)
-+#endif
++/* Unmap and free pages contained in the specified pseudo physical region. */
++#define IA64_DOM0VP_zap_physmap 5
+
-+
-+/*
-+ * Force strict CPU ordering.
-+ * And yes, this is required on UP too when we're talking
-+ * to devices.
-+ */
-+#define mb() asm volatile("mfence":::"memory")
-+#define rmb() asm volatile("lfence":::"memory")
++/* Assign machine page frame to dom0's pseudo physical address space. */
++#define IA64_DOM0VP_add_physmap 6
+
-+#ifdef CONFIG_UNORDERED_IO
-+#define wmb() asm volatile("sfence" ::: "memory")
-+#else
-+#define wmb() asm volatile("" ::: "memory")
-+#endif
-+#define read_barrier_depends() do {} while(0)
-+#define set_mb(var, value) do { (void) xchg(&var, value); } while (0)
++/* expose the p2m table into domain */
++#define IA64_DOM0VP_expose_p2m 7
++
++/* xen perfmon */
++#define IA64_DOM0VP_perfmon 8
++
++/* gmfn version of IA64_DOM0VP_add_physmap */
++#define IA64_DOM0VP_add_physmap_with_gmfn 9
++
++/* get fpswa revision */
++#define IA64_DOM0VP_fpswa_revision 10
++
++/* Add an I/O port space range */
++#define IA64_DOM0VP_add_io_space 11
++
++/* expose the foreign domain's p2m table into privileged domain */
++#define IA64_DOM0VP_expose_foreign_p2m 12
++#define IA64_DOM0VP_EFP_ALLOC_PTE 0x1 /* allocate p2m table */
++
++/* unexpose the foreign domain's p2m table into privileged domain */
++#define IA64_DOM0VP_unexpose_foreign_p2m 13
++
++// flags for page assignement to pseudo physical address space
++#define _ASSIGN_readonly 0
++#define ASSIGN_readonly (1UL << _ASSIGN_readonly)
++#define ASSIGN_writable (0UL << _ASSIGN_readonly) // dummy flag
++/* Internal only: memory attribute must be WC/UC/UCE. */
++#define _ASSIGN_nocache 1
++#define ASSIGN_nocache (1UL << _ASSIGN_nocache)
++// tlb tracking
++#define _ASSIGN_tlb_track 2
++#define ASSIGN_tlb_track (1UL << _ASSIGN_tlb_track)
++/* Internal only: associated with PGC_allocated bit */
++#define _ASSIGN_pgc_allocated 3
++#define ASSIGN_pgc_allocated (1UL << _ASSIGN_pgc_allocated)
++/* Page is an IO page. */
++#define _ASSIGN_io 4
++#define ASSIGN_io (1UL << _ASSIGN_io)
++
++/* This structure has the same layout of struct ia64_boot_param, defined in
++ <asm/system.h>. It is redefined here to ease use. */
++struct xen_ia64_boot_param {
++ unsigned long command_line; /* physical address of cmd line args */
++ unsigned long efi_systab; /* physical address of EFI system table */
++ unsigned long efi_memmap; /* physical address of EFI memory map */
++ unsigned long efi_memmap_size; /* size of EFI memory map */
++ unsigned long efi_memdesc_size; /* size of an EFI memory map descriptor */
++ unsigned int efi_memdesc_version; /* memory descriptor version */
++ struct {
++ unsigned short num_cols; /* number of columns on console. */
++ unsigned short num_rows; /* number of rows on console. */
++ unsigned short orig_x; /* cursor's x position */
++ unsigned short orig_y; /* cursor's y position */
++ } console_info;
++ unsigned long fpswa; /* physical address of the fpswa interface */
++ unsigned long initrd_start;
++ unsigned long initrd_size;
++ unsigned long domain_start; /* va where the boot time domain begins */
++ unsigned long domain_size; /* how big is the boot domain */
++};
++
++#endif /* !__ASSEMBLY__ */
++
++/* Size of the shared_info area (this is not related to page size). */
++#define XSI_SHIFT 14
++#define XSI_SIZE (1 << XSI_SHIFT)
++/* Log size of mapped_regs area (64 KB - only 4KB is used). */
++#define XMAPPEDREGS_SHIFT 12
++#define XMAPPEDREGS_SIZE (1 << XMAPPEDREGS_SHIFT)
++/* Offset of XASI (Xen arch shared info) wrt XSI_BASE. */
++#define XMAPPEDREGS_OFS XSI_SIZE
++
++/* Hyperprivops. */
++#define HYPERPRIVOP_START 0x1
++#define HYPERPRIVOP_RFI (HYPERPRIVOP_START + 0x0)
++#define HYPERPRIVOP_RSM_DT (HYPERPRIVOP_START + 0x1)
++#define HYPERPRIVOP_SSM_DT (HYPERPRIVOP_START + 0x2)
++#define HYPERPRIVOP_COVER (HYPERPRIVOP_START + 0x3)
++#define HYPERPRIVOP_ITC_D (HYPERPRIVOP_START + 0x4)
++#define HYPERPRIVOP_ITC_I (HYPERPRIVOP_START + 0x5)
++#define HYPERPRIVOP_SSM_I (HYPERPRIVOP_START + 0x6)
++#define HYPERPRIVOP_GET_IVR (HYPERPRIVOP_START + 0x7)
++#define HYPERPRIVOP_GET_TPR (HYPERPRIVOP_START + 0x8)
++#define HYPERPRIVOP_SET_TPR (HYPERPRIVOP_START + 0x9)
++#define HYPERPRIVOP_EOI (HYPERPRIVOP_START + 0xa)
++#define HYPERPRIVOP_SET_ITM (HYPERPRIVOP_START + 0xb)
++#define HYPERPRIVOP_THASH (HYPERPRIVOP_START + 0xc)
++#define HYPERPRIVOP_PTC_GA (HYPERPRIVOP_START + 0xd)
++#define HYPERPRIVOP_ITR_D (HYPERPRIVOP_START + 0xe)
++#define HYPERPRIVOP_GET_RR (HYPERPRIVOP_START + 0xf)
++#define HYPERPRIVOP_SET_RR (HYPERPRIVOP_START + 0x10)
++#define HYPERPRIVOP_SET_KR (HYPERPRIVOP_START + 0x11)
++#define HYPERPRIVOP_FC (HYPERPRIVOP_START + 0x12)
++#define HYPERPRIVOP_GET_CPUID (HYPERPRIVOP_START + 0x13)
++#define HYPERPRIVOP_GET_PMD (HYPERPRIVOP_START + 0x14)
++#define HYPERPRIVOP_GET_EFLAG (HYPERPRIVOP_START + 0x15)
++#define HYPERPRIVOP_SET_EFLAG (HYPERPRIVOP_START + 0x16)
++#define HYPERPRIVOP_RSM_BE (HYPERPRIVOP_START + 0x17)
++#define HYPERPRIVOP_GET_PSR (HYPERPRIVOP_START + 0x18)
++#define HYPERPRIVOP_SET_RR0_TO_RR4 (HYPERPRIVOP_START + 0x19)
++#define HYPERPRIVOP_MAX (0x1a)
+
-+#define warn_if_not_ulong(x) do { unsigned long foo; (void) (&(x) == &foo); } while (0)
++/* Fast and light hypercalls. */
++#define __HYPERVISOR_ia64_fast_eoi __HYPERVISOR_arch_1
+
-+#include <linux/irqflags.h>
++/* Extra debug features. */
++#define __HYPERVISOR_ia64_debug_op __HYPERVISOR_arch_2
+
-+void cpu_idle_wait(void);
++/* Xencomm macros. */
++#define XENCOMM_INLINE_MASK 0xf800000000000000UL
++#define XENCOMM_INLINE_FLAG 0x8000000000000000UL
+
-+extern unsigned long arch_align_stack(unsigned long sp);
-+extern void free_init_pages(char *what, unsigned long begin, unsigned long end);
++#ifndef __ASSEMBLY__
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/timer.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/timer.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/timer.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/timer.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,67 @@
-+#ifndef _ASMi386_TIMER_H
-+#define _ASMi386_TIMER_H
-+#include <linux/init.h>
++/*
++ * Optimization features.
++ * The hypervisor may do some special optimizations for guests. This hypercall
++ * can be used to switch on/of these special optimizations.
++ */
++#define __HYPERVISOR_opt_feature 0x700UL
+
-+/**
-+ * struct timer_ops - used to define a timer source
-+ *
-+ * @name: name of the timer.
-+ * @init: Probes and initializes the timer. Takes clock= override
-+ * string as an argument. Returns 0 on success, anything else
-+ * on failure.
-+ * @mark_offset: called by the timer interrupt.
-+ * @get_offset: called by gettimeofday(). Returns the number of microseconds
-+ * since the last timer interupt.
-+ * @monotonic_clock: returns the number of nanoseconds since the init of the
-+ * timer.
-+ * @delay: delays this many clock cycles.
++#define XEN_IA64_OPTF_OFF 0x0
++#define XEN_IA64_OPTF_ON 0x1
++
++/*
++ * If this feature is switched on, the hypervisor inserts the
++ * tlb entries without calling the guests traphandler.
++ * This is useful in guests using region 7 for identity mapping
++ * like the linux kernel does.
+ */
-+struct timer_opts {
-+ char* name;
-+ void (*mark_offset)(void);
-+ unsigned long (*get_offset)(void);
-+ unsigned long long (*monotonic_clock)(void);
-+ void (*delay)(unsigned long);
-+ unsigned long (*read_timer)(void);
-+ int (*suspend)(pm_message_t state);
-+ int (*resume)(void);
-+};
++#define XEN_IA64_OPTF_IDENT_MAP_REG7 1
+
-+struct init_timer_opts {
-+ int (*init)(char *override);
-+ struct timer_opts *opts;
-+};
++/* Identity mapping of region 4 addresses in HVM. */
++#define XEN_IA64_OPTF_IDENT_MAP_REG4 2
+
-+#define TICK_SIZE (tick_nsec / 1000)
++/* Identity mapping of region 5 addresses in HVM. */
++#define XEN_IA64_OPTF_IDENT_MAP_REG5 3
+
-+extern struct timer_opts* __init select_timer(void);
-+extern void clock_fallback(void);
-+void setup_pit_timer(void);
++#define XEN_IA64_OPTF_IDENT_MAP_NOT_SET (0)
+
-+/* Modifiers for buggy PIT handling */
++struct xen_ia64_opt_feature {
++ unsigned long cmd; /* Which feature */
++ unsigned char on; /* Switch feature on/off */
++ union {
++ struct {
++ /* The page protection bit mask of the pte.
++ * This will be or'ed with the pte. */
++ unsigned long pgprot;
++ unsigned long key; /* A protection key for itir. */
++ };
++ };
++};
+
-+extern int pit_latch_buggy;
++#endif /* __ASSEMBLY__ */
+
-+extern struct timer_opts *cur_timer;
-+extern int timer_ack;
++/* xen perfmon */
++#ifdef XEN
++#ifndef __ASSEMBLY__
++#ifndef _ASM_IA64_PERFMON_H
+
-+/* list of externed timers */
-+extern struct timer_opts timer_none;
-+extern struct timer_opts timer_pit;
-+extern struct init_timer_opts timer_pit_init;
-+extern struct init_timer_opts timer_tsc_init;
-+#ifdef CONFIG_X86_CYCLONE_TIMER
-+extern struct init_timer_opts timer_cyclone_init;
-+#endif
++#include <xen/list.h> // asm/perfmon.h requires struct list_head
++#include <asm/perfmon.h>
++// for PFM_xxx and pfarg_features_t, pfarg_context_t, pfarg_reg_t, pfarg_load_t
+
-+extern unsigned long calibrate_tsc(void);
-+extern void init_cpu_khz(void);
-+#ifdef CONFIG_HPET_TIMER
-+extern struct init_timer_opts timer_hpet_init;
-+extern unsigned long calibrate_tsc_hpet(unsigned long *tsc_hpet_quotient_ptr);
-+#endif
++#endif /* _ASM_IA64_PERFMON_H */
+
-+#ifdef CONFIG_X86_PM_TIMER
-+extern struct init_timer_opts timer_pmtmr_init;
-+#endif
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/tlbflush.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/tlbflush.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/tlbflush.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/tlbflush.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,103 @@
-+#ifndef _X8664_TLBFLUSH_H
-+#define _X8664_TLBFLUSH_H
++DEFINE_XEN_GUEST_HANDLE(pfarg_features_t);
++DEFINE_XEN_GUEST_HANDLE(pfarg_context_t);
++DEFINE_XEN_GUEST_HANDLE(pfarg_reg_t);
++DEFINE_XEN_GUEST_HANDLE(pfarg_load_t);
++#endif /* __ASSEMBLY__ */
++#endif /* XEN */
+
-+#include <linux/mm.h>
-+#include <asm/processor.h>
++#ifndef __ASSEMBLY__
++#include "arch-ia64/hvm/memmap.h"
++#endif
+
-+#define __flush_tlb() xen_tlb_flush()
++#endif /* __HYPERVISOR_IF_IA64_H__ */
+
+/*
-+ * Global pages have to be flushed a bit differently. Not a real
-+ * performance problem because this does not happen often.
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-ia64/debug_op.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-ia64/debug_op.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,96 @@
++/******************************************************************************
++ * debug_op.h
++ *
++ * Copyright (c) 2007 Tristan Gingold <tgingold at free.fr>
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
+ */
-+#define __flush_tlb_global() xen_tlb_flush()
+
++#ifndef __XEN_PUBLIC_IA64_DEBUG_OP_H__
++#define __XEN_PUBLIC_IA64_DEBUG_OP_H__
+
-+extern unsigned long pgkern_mask;
++/* Set/Get extra conditions to break. */
++#define XEN_IA64_DEBUG_OP_SET_FLAGS 1
++#define XEN_IA64_DEBUG_OP_GET_FLAGS 2
+
-+#define __flush_tlb_all() __flush_tlb_global()
++/* Break on kernel single step. */
++#define XEN_IA64_DEBUG_ON_KERN_SSTEP (1 << 0)
+
-+#define __flush_tlb_one(addr) xen_invlpg((unsigned long)addr)
++/* Break on kernel debug (breakpoint or watch point). */
++#define XEN_IA64_DEBUG_ON_KERN_DEBUG (1 << 1)
+
++/* Break on kernel taken branch. */
++#define XEN_IA64_DEBUG_ON_KERN_TBRANCH (1 << 2)
+
-+/*
-+ * TLB flushing:
-+ *
-+ * - flush_tlb() flushes the current mm struct TLBs
-+ * - flush_tlb_all() flushes all processes TLBs
-+ * - flush_tlb_mm(mm) flushes the specified mm context TLB's
-+ * - flush_tlb_page(vma, vmaddr) flushes one page
-+ * - flush_tlb_range(vma, start, end) flushes a range of pages
-+ * - flush_tlb_kernel_range(start, end) flushes a range of kernel pages
-+ * - flush_tlb_pgtables(mm, start, end) flushes a range of page tables
-+ *
-+ * x86-64 can only flush individual pages or full VMs. For a range flush
-+ * we always do the full VM. Might be worth trying if for a small
-+ * range a few INVLPGs in a row are a win.
-+ */
++/* Break on interrupt injection. */
++#define XEN_IA64_DEBUG_ON_EXTINT (1 << 3)
+
-+#ifndef CONFIG_SMP
++/* Break on interrupt injection. */
++#define XEN_IA64_DEBUG_ON_EXCEPT (1 << 4)
+
-+#define flush_tlb() __flush_tlb()
-+#define flush_tlb_all() __flush_tlb_all()
-+#define local_flush_tlb() __flush_tlb()
++/* Break on event injection. */
++#define XEN_IA64_DEBUG_ON_EVENT (1 << 5)
+
-+static inline void flush_tlb_mm(struct mm_struct *mm)
-+{
-+ if (mm == current->active_mm)
-+ __flush_tlb();
-+}
++/* Break on privop/virtualized instruction (slow path only). */
++#define XEN_IA64_DEBUG_ON_PRIVOP (1 << 6)
+
-+static inline void flush_tlb_page(struct vm_area_struct *vma,
-+ unsigned long addr)
-+{
-+ if (vma->vm_mm == current->active_mm)
-+ __flush_tlb_one(addr);
-+}
++/* Break on emulated PAL call (at entry). */
++#define XEN_IA64_DEBUG_ON_PAL (1 << 7)
+
-+static inline void flush_tlb_range(struct vm_area_struct *vma,
-+ unsigned long start, unsigned long end)
-+{
-+ if (vma->vm_mm == current->active_mm)
-+ __flush_tlb();
-+}
++/* Break on emulated SAL call (at entry). */
++#define XEN_IA64_DEBUG_ON_SAL (1 << 8)
+
-+#else
++/* Break on emulated EFI call (at entry). */
++#define XEN_IA64_DEBUG_ON_EFI (1 << 9)
+
-+#include <asm/smp.h>
++/* Break on rfi emulation (slow path only, before exec). */
++#define XEN_IA64_DEBUG_ON_RFI (1 << 10)
+
-+#define local_flush_tlb() \
-+ __flush_tlb()
++/* Break on address translation switch. */
++#define XEN_IA64_DEBUG_ON_MMU (1 << 11)
+
-+extern void flush_tlb_all(void);
-+extern void flush_tlb_current_task(void);
-+extern void flush_tlb_mm(struct mm_struct *);
-+extern void flush_tlb_page(struct vm_area_struct *, unsigned long);
++/* Break on bad guest physical address. */
++#define XEN_IA64_DEBUG_ON_BAD_MPA (1 << 12)
+
-+#define flush_tlb() flush_tlb_current_task()
++/* Force psr.ss bit. */
++#define XEN_IA64_DEBUG_FORCE_SS (1 << 13)
+
-+static inline void flush_tlb_range(struct vm_area_struct * vma, unsigned long start, unsigned long end)
-+{
-+ flush_tlb_mm(vma->vm_mm);
-+}
++/* Force psr.db bit. */
++#define XEN_IA64_DEBUG_FORCE_DB (1 << 14)
+
-+#define TLBSTATE_OK 1
-+#define TLBSTATE_LAZY 2
++/* Break on ITR/PTR. */
++#define XEN_IA64_DEBUG_ON_TR (1 << 15)
+
-+/* Roughly an IPI every 20MB with 4k pages for freeing page table
-+ ranges. Cost is about 42k of memory for each CPU. */
-+#define ARCH_FREE_PTE_NR 5350
++/* Break on ITC/PTC.L/PTC.G/PTC.GA. */
++#define XEN_IA64_DEBUG_ON_TC (1 << 16)
+
-+#endif
++/* Get translation cache. */
++#define XEN_IA64_DEBUG_OP_GET_TC 3
+
-+#define flush_tlb_kernel_range(start, end) flush_tlb_all()
++/* Translate virtual address to guest physical address. */
++#define XEN_IA64_DEBUG_OP_TRANSLATE 4
+
-+static inline void flush_tlb_pgtables(struct mm_struct *mm,
-+ unsigned long start, unsigned long end)
-+{
-+ /* x86_64 does not keep any page table caches in a software TLB.
-+ The CPUs do in their hardware TLBs, but they are handled
-+ by the normal TLB flushing algorithms. */
-+}
++union xen_ia64_debug_op {
++ uint64_t flags;
++ struct xen_ia64_debug_vtlb {
++ uint64_t nbr; /* IN/OUT */
++ XEN_GUEST_HANDLE_64(ia64_tr_entry_t) tr; /* IN/OUT */
++ } vtlb;
++};
++typedef union xen_ia64_debug_op xen_ia64_debug_op_t;
++DEFINE_XEN_GUEST_HANDLE(xen_ia64_debug_op_t);
+
-+#endif /* _X8664_TLBFLUSH_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/vga.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/vga.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/vga.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/vga.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,20 @@
-+/*
-+ * Access to VGA videoram
++#endif /* __XEN_PUBLIC_IA64_DEBUG_OP_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-ia64/hvm/memmap.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-ia64/hvm/memmap.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,88 @@
++/******************************************************************************
++ * memmap.h
++ *
++ * Copyright (c) 2008 Tristan Gingold <tgingold AT free fr>
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
-+ * (c) 1998 Martin Mares <mj at ucw.cz>
+ */
+
-+#ifndef _LINUX_ASM_VGA_H_
-+#define _LINUX_ASM_VGA_H_
++#ifndef __XEN_PUBLIC_HVM_MEMMAP_IA64_H__
++#define __XEN_PUBLIC_HVM_MEMMAP_IA64_H__
+
-+/*
-+ * On the PC, we can just recalculate addresses and then
-+ * access the videoram directly without any black magic.
-+ */
++#define MEM_G (1UL << 30)
++#define MEM_M (1UL << 20)
++#define MEM_K (1UL << 10)
++
++/* Guest physical address of IO ports space. */
++#define MMIO_START (3 * MEM_G)
++#define MMIO_SIZE (512 * MEM_M)
+
-+#define VGA_MAP_MEM(x,s) (unsigned long)isa_bus_to_virt(x)
++#define VGA_IO_START 0xA0000UL
++#define VGA_IO_SIZE 0x20000
+
-+#define vga_readb(x) (*(x))
-+#define vga_writeb(x,y) (*(y) = (x))
++#define LEGACY_IO_START (MMIO_START + MMIO_SIZE)
++#define LEGACY_IO_SIZE (64 * MEM_M)
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/xenoprof.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/xenoprof.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/xenoprof.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/xenoprof.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1 @@
-+#include <asm-i386/mach-xen/asm/xenoprof.h>
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/xor.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/xor.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/asm/xor.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/asm/xor.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,328 @@
-+/*
-+ * x86-64 changes / gcc fixes from Andi Kleen.
-+ * Copyright 2002 Andi Kleen, SuSE Labs.
-+ *
-+ * This hasn't been optimized for the hammer yet, but there are likely
-+ * no advantages to be gotten from x86-64 here anyways.
-+ */
++#define IO_PAGE_START (LEGACY_IO_START + LEGACY_IO_SIZE)
++#define IO_PAGE_SIZE XEN_PAGE_SIZE
+
-+typedef struct { unsigned long a,b; } __attribute__((aligned(16))) xmm_store_t;
++#define STORE_PAGE_START (IO_PAGE_START + IO_PAGE_SIZE)
++#define STORE_PAGE_SIZE XEN_PAGE_SIZE
+
-+/* Doesn't use gcc to save the XMM registers, because there is no easy way to
-+ tell it to do a clts before the register saving. */
-+#define XMMS_SAVE do { \
-+ preempt_disable(); \
-+ if (!(current_thread_info()->status & TS_USEDFPU)) \
-+ clts(); \
-+ __asm__ __volatile__ ( \
-+ "movups %%xmm0,(%1) ;\n\t" \
-+ "movups %%xmm1,0x10(%1) ;\n\t" \
-+ "movups %%xmm2,0x20(%1) ;\n\t" \
-+ "movups %%xmm3,0x30(%1) ;\n\t" \
-+ : "=&r" (cr0) \
-+ : "r" (xmm_save) \
-+ : "memory"); \
-+} while(0)
++#define BUFFER_IO_PAGE_START (STORE_PAGE_START + STORE_PAGE_SIZE)
++#define BUFFER_IO_PAGE_SIZE XEN_PAGE_SIZE
+
-+#define XMMS_RESTORE do { \
-+ asm volatile ( \
-+ "sfence ;\n\t" \
-+ "movups (%1),%%xmm0 ;\n\t" \
-+ "movups 0x10(%1),%%xmm1 ;\n\t" \
-+ "movups 0x20(%1),%%xmm2 ;\n\t" \
-+ "movups 0x30(%1),%%xmm3 ;\n\t" \
-+ : \
-+ : "r" (cr0), "r" (xmm_save) \
-+ : "memory"); \
-+ if (!(current_thread_info()->status & TS_USEDFPU)) \
-+ stts(); \
-+ preempt_enable(); \
-+} while(0)
++#define BUFFER_PIO_PAGE_START (BUFFER_IO_PAGE_START + BUFFER_IO_PAGE_SIZE)
++#define BUFFER_PIO_PAGE_SIZE XEN_PAGE_SIZE
+
-+#define OFFS(x) "16*("#x")"
-+#define PF_OFFS(x) "256+16*("#x")"
-+#define PF0(x) " prefetchnta "PF_OFFS(x)"(%[p1]) ;\n"
-+#define LD(x,y) " movaps "OFFS(x)"(%[p1]), %%xmm"#y" ;\n"
-+#define ST(x,y) " movaps %%xmm"#y", "OFFS(x)"(%[p1]) ;\n"
-+#define PF1(x) " prefetchnta "PF_OFFS(x)"(%[p2]) ;\n"
-+#define PF2(x) " prefetchnta "PF_OFFS(x)"(%[p3]) ;\n"
-+#define PF3(x) " prefetchnta "PF_OFFS(x)"(%[p4]) ;\n"
-+#define PF4(x) " prefetchnta "PF_OFFS(x)"(%[p5]) ;\n"
-+#define PF5(x) " prefetchnta "PF_OFFS(x)"(%[p6]) ;\n"
-+#define XO1(x,y) " xorps "OFFS(x)"(%[p2]), %%xmm"#y" ;\n"
-+#define XO2(x,y) " xorps "OFFS(x)"(%[p3]), %%xmm"#y" ;\n"
-+#define XO3(x,y) " xorps "OFFS(x)"(%[p4]), %%xmm"#y" ;\n"
-+#define XO4(x,y) " xorps "OFFS(x)"(%[p5]), %%xmm"#y" ;\n"
-+#define XO5(x,y) " xorps "OFFS(x)"(%[p6]), %%xmm"#y" ;\n"
++#define IO_SAPIC_START 0xfec00000UL
++#define IO_SAPIC_SIZE 0x100000
+
++#define PIB_START 0xfee00000UL
++#define PIB_SIZE 0x200000
+
-+static void
-+xor_sse_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
-+{
-+ unsigned int lines = bytes >> 8;
-+ unsigned long cr0;
-+ xmm_store_t xmm_save[4];
++#define GFW_START (4 * MEM_G - 16 * MEM_M)
++#define GFW_SIZE (16 * MEM_M)
+
-+ XMMS_SAVE;
++/* domVTI */
++#define GPFN_FRAME_BUFFER 0x1 /* VGA framebuffer */
++#define GPFN_LOW_MMIO 0x2 /* Low MMIO range */
++#define GPFN_PIB 0x3 /* PIB base */
++#define GPFN_IOSAPIC 0x4 /* IOSAPIC base */
++#define GPFN_LEGACY_IO 0x5 /* Legacy I/O base */
++#define GPFN_HIGH_MMIO 0x6 /* High MMIO range */
+
-+ asm volatile (
-+#undef BLOCK
-+#define BLOCK(i) \
-+ LD(i,0) \
-+ LD(i+1,1) \
-+ PF1(i) \
-+ PF1(i+2) \
-+ LD(i+2,2) \
-+ LD(i+3,3) \
-+ PF0(i+4) \
-+ PF0(i+6) \
-+ XO1(i,0) \
-+ XO1(i+1,1) \
-+ XO1(i+2,2) \
-+ XO1(i+3,3) \
-+ ST(i,0) \
-+ ST(i+1,1) \
-+ ST(i+2,2) \
-+ ST(i+3,3) \
++/* Nvram belongs to GFW memory space */
++#define NVRAM_SIZE (MEM_K * 64)
++#define NVRAM_START (GFW_START + 10 * MEM_M)
+
++#define NVRAM_VALID_SIG 0x4650494e45584948 /* "HIXENIPF" */
++struct nvram_save_addr {
++ unsigned long addr;
++ unsigned long signature;
++};
+
-+ PF0(0)
-+ PF0(2)
++#endif /* __XEN_PUBLIC_HVM_MEMMAP_IA64_H__ */
+
-+ " .align 32 ;\n"
-+ " 1: ;\n"
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-ia64/hvm/save.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-ia64/hvm/save.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,201 @@
++/******************************************************************************
++ * save_types.h
++ *
++ * Copyright (c) 2007 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ */
+
-+ BLOCK(0)
-+ BLOCK(4)
-+ BLOCK(8)
-+ BLOCK(12)
++#ifndef __XEN_PUBLIC_HVM_SAVE_IA64_H__
++#define __XEN_PUBLIC_HVM_SAVE_IA64_H__
+
-+ " addq %[inc], %[p1] ;\n"
-+ " addq %[inc], %[p2] ;\n"
-+ " decl %[cnt] ; jnz 1b"
-+ : [p1] "+r" (p1), [p2] "+r" (p2), [cnt] "+r" (lines)
-+ : [inc] "r" (256UL)
-+ : "memory");
++#include <public/hvm/save.h>
++#include <public/arch-ia64.h>
+
-+ XMMS_RESTORE;
-+}
++/*
++ * Save/restore header: general info about the save file.
++ */
+
-+static void
-+xor_sse_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-+ unsigned long *p3)
-+{
-+ unsigned int lines = bytes >> 8;
-+ xmm_store_t xmm_save[4];
-+ unsigned long cr0;
++/* x86 uses 0x54381286 */
++#define HVM_FILE_MAGIC 0x343641492f6e6558UL /* "Xen/IA64" */
++#define HVM_FILE_VERSION 0x0000000000000001UL
+
-+ XMMS_SAVE;
++struct hvm_save_header {
++ uint64_t magic; /* Must be HVM_FILE_MAGIC */
++ uint64_t version; /* File format version */
++ uint64_t changeset; /* Version of Xen that saved this file */
++ uint64_t cpuid[5]; /* CPUID[0x01][%eax] on the saving machine */
++};
+
-+ __asm__ __volatile__ (
-+#undef BLOCK
-+#define BLOCK(i) \
-+ PF1(i) \
-+ PF1(i+2) \
-+ LD(i,0) \
-+ LD(i+1,1) \
-+ LD(i+2,2) \
-+ LD(i+3,3) \
-+ PF2(i) \
-+ PF2(i+2) \
-+ PF0(i+4) \
-+ PF0(i+6) \
-+ XO1(i,0) \
-+ XO1(i+1,1) \
-+ XO1(i+2,2) \
-+ XO1(i+3,3) \
-+ XO2(i,0) \
-+ XO2(i+1,1) \
-+ XO2(i+2,2) \
-+ XO2(i+3,3) \
-+ ST(i,0) \
-+ ST(i+1,1) \
-+ ST(i+2,2) \
-+ ST(i+3,3) \
++DECLARE_HVM_SAVE_TYPE(HEADER, 1, struct hvm_save_header);
+
++/*
++ * CPU
++ */
++struct hvm_hw_ia64_cpu {
++ uint64_t ipsr;
++};
++DECLARE_HVM_SAVE_TYPE(CPU, 2, struct hvm_hw_ia64_cpu);
+
-+ PF0(0)
-+ PF0(2)
++/*
++ * CPU
++ */
++struct hvm_hw_ia64_vpd {
++ struct vpd vpd;
++};
++DECLARE_HVM_SAVE_TYPE(VPD, 3, struct hvm_hw_ia64_vpd);
+
-+ " .align 32 ;\n"
-+ " 1: ;\n"
++/*
++ * device dependency
++ * vacpi => viosapic => vlsapic
++ */
++/*
++ * vlsapic
++ */
++struct hvm_hw_ia64_vlsapic {
++ uint64_t insvc[4];
++ uint64_t vhpi; // ??? should this be saved in vpd
++ uint8_t xtp;
++ uint8_t pal_init_pending;
++ uint8_t pad[2];
++};
++DECLARE_HVM_SAVE_TYPE(VLSAPIC, 4, struct hvm_hw_ia64_vlsapic);
++/* set
++ * unconditionaly set v->arch.irq_new_peding = 1
++ * unconditionaly set v->arch.irq_new_condition = 0
++ */
+
-+ BLOCK(0)
-+ BLOCK(4)
-+ BLOCK(8)
-+ BLOCK(12)
++/*
++ * vtime
++ */
++/* itc, itm, itv are saved by arch vcpu context */
++struct hvm_hw_ia64_vtime {
++ uint64_t itc;
++ uint64_t itm;
+
-+ " addq %[inc], %[p1] ;\n"
-+ " addq %[inc], %[p2] ;\n"
-+ " addq %[inc], %[p3] ;\n"
-+ " decl %[cnt] ; jnz 1b"
-+ : [cnt] "+r" (lines),
-+ [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3)
-+ : [inc] "r" (256UL)
-+ : "memory");
-+ XMMS_RESTORE;
-+}
++ uint64_t last_itc;
++ uint64_t pending;
++};
++DECLARE_HVM_SAVE_TYPE(VTIME, 5, struct hvm_hw_ia64_vtime);
++/*
++ * calculate v->vtm.vtm_offset
++ * ??? Or should vtm_offset be set by leave_hypervisor_tail()?
++ * start vtm_timer if necessary by vtm_set_itm().
++ * ??? Or should vtm_timer be set by leave_hypervisor_tail()?
++ *
++ * ??? or should be done by schedule_tail()
++ * => schedule_tail() should do.
++ */
+
-+static void
-+xor_sse_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-+ unsigned long *p3, unsigned long *p4)
++/*
++ * viosapic
++ */
++#define VIOSAPIC_NUM_PINS 48
++
++union viosapic_rte
+{
-+ unsigned int lines = bytes >> 8;
-+ xmm_store_t xmm_save[4];
-+ unsigned long cr0;
++ uint64_t bits;
++ struct {
++ uint8_t vector;
+
-+ XMMS_SAVE;
++ uint8_t delivery_mode : 3;
++ uint8_t reserve1 : 1;
++ uint8_t delivery_status: 1;
++ uint8_t polarity : 1;
++ uint8_t reserve2 : 1;
++ uint8_t trig_mode : 1;
++
++ uint8_t mask : 1;
++ uint8_t reserve3 : 7;
++
++ uint8_t reserved[3];
++ uint16_t dest_id;
++ };
++};
++
++struct hvm_hw_ia64_viosapic {
++ uint64_t irr;
++ uint64_t isr;
++ uint32_t ioregsel;
++ uint32_t pad;
++ uint64_t lowest_vcpu_id;
++ uint64_t base_address;
++ union viosapic_rte redirtbl[VIOSAPIC_NUM_PINS];
++};
++DECLARE_HVM_SAVE_TYPE(VIOSAPIC, 6, struct hvm_hw_ia64_viosapic);
++
++/*
++ * vacpi
++ * PM timer
++ */
++struct vacpi_regs {
++ union {
++ struct {
++ uint32_t pm1a_sts:16;/* PM1a_EVT_BLK.PM1a_STS: status register */
++ uint32_t pm1a_en:16; /* PM1a_EVT_BLK.PM1a_EN: enable register */
++ };
++ uint32_t evt_blk;
++ };
++ uint32_t tmr_val; /* PM_TMR_BLK.TMR_VAL: 32bit free-running counter */
++};
+
-+ __asm__ __volatile__ (
-+#undef BLOCK
-+#define BLOCK(i) \
-+ PF1(i) \
-+ PF1(i+2) \
-+ LD(i,0) \
-+ LD(i+1,1) \
-+ LD(i+2,2) \
-+ LD(i+3,3) \
-+ PF2(i) \
-+ PF2(i+2) \
-+ XO1(i,0) \
-+ XO1(i+1,1) \
-+ XO1(i+2,2) \
-+ XO1(i+3,3) \
-+ PF3(i) \
-+ PF3(i+2) \
-+ PF0(i+4) \
-+ PF0(i+6) \
-+ XO2(i,0) \
-+ XO2(i+1,1) \
-+ XO2(i+2,2) \
-+ XO2(i+3,3) \
-+ XO3(i,0) \
-+ XO3(i+1,1) \
-+ XO3(i+2,2) \
-+ XO3(i+3,3) \
-+ ST(i,0) \
-+ ST(i+1,1) \
-+ ST(i+2,2) \
-+ ST(i+3,3) \
++struct hvm_hw_ia64_vacpi {
++ struct vacpi_regs regs;
++};
++DECLARE_HVM_SAVE_TYPE(VACPI, 7, struct hvm_hw_ia64_vacpi);
++/* update last_gtime and setup timer of struct vacpi */
++
++/*
++ * opt_feature: identity mapping of region 4, 5 and 7.
++ * With the c/s 16396:d2935f9c217f of xen-ia64-devel.hg,
++ * opt_feature hypercall supports only region 4,5,7 identity mappings.
++ * structure hvm_hw_ia64_identity_mappings only supports them.
++ * The new structure, struct hvm_hw_ia64_identity_mappings, is created to
++ * avoid to keep up with change of the xen/ia64 internal structure, struct
++ * opt_feature.
++ *
++ * If it is enhanced in the future, new structure will be created.
++ */
++struct hvm_hw_ia64_identity_mapping {
++ uint64_t on; /* on/off */
++ uint64_t pgprot; /* The page protection bit mask of the pte. */
++ uint64_t key; /* A protection key. */
++};
++
++struct hvm_hw_ia64_identity_mappings {
++ struct hvm_hw_ia64_identity_mapping im_reg4;/* Region 4 identity mapping */
++ struct hvm_hw_ia64_identity_mapping im_reg5;/* Region 5 identity mapping */
++ struct hvm_hw_ia64_identity_mapping im_reg7;/* Region 7 identity mapping */
++};
++DECLARE_HVM_SAVE_TYPE(OPT_FEATURE_IDENTITY_MAPPINGS, 8, struct hvm_hw_ia64_identity_mappings);
++
++/*
++ * Largest type-code in use
++ */
++#define HVM_SAVE_CODE_MAX 8
++
++#endif /* __XEN_PUBLIC_HVM_SAVE_IA64_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-ia64/sioemu.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-ia64/sioemu.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,89 @@
++/******************************************************************************
++ * sioemu.h
++ *
++ * Copyright (c) 2008 Tristan Gingold <tgingold at free.fr>
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ */
+
++#ifndef __XEN_PUBLIC_IA64_SIOEMU_H__
++#define __XEN_PUBLIC_IA64_SIOEMU_H__
+
-+ PF0(0)
-+ PF0(2)
++/* SIOEMU specific hypercalls.
++ The numbers are the minor part of FW_HYPERCALL_SIOEMU. */
+
-+ " .align 32 ;\n"
-+ " 1: ;\n"
++/* Defines the callback entry point. r8=ip, r9=data.
++ Must be called per-vcpu. */
++#define SIOEMU_HYPERCALL_SET_CALLBACK 0x01
+
-+ BLOCK(0)
-+ BLOCK(4)
-+ BLOCK(8)
-+ BLOCK(12)
++/* Finish sioemu fw initialization and start firmware. r8=ip. */
++#define SIOEMU_HYPERCALL_START_FW 0x02
+
-+ " addq %[inc], %[p1] ;\n"
-+ " addq %[inc], %[p2] ;\n"
-+ " addq %[inc], %[p3] ;\n"
-+ " addq %[inc], %[p4] ;\n"
-+ " decl %[cnt] ; jnz 1b"
-+ : [cnt] "+c" (lines),
-+ [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3), [p4] "+r" (p4)
-+ : [inc] "r" (256UL)
-+ : "memory" );
++/* Add IO pages in physmap. */
++#define SIOEMU_HYPERCALL_ADD_IO_PHYSMAP 0x03
+
-+ XMMS_RESTORE;
-+}
++/* Get wallclock time. */
++#define SIOEMU_HYPERCALL_GET_TIME 0x04
+
-+static void
-+xor_sse_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-+ unsigned long *p3, unsigned long *p4, unsigned long *p5)
-+{
-+ unsigned int lines = bytes >> 8;
-+ xmm_store_t xmm_save[4];
-+ unsigned long cr0;
++/* Flush cache. */
++#define SIOEMU_HYPERCALL_FLUSH_CACHE 0x07
+
-+ XMMS_SAVE;
++/* Get freq base. */
++#define SIOEMU_HYPERCALL_FREQ_BASE 0x08
+
-+ __asm__ __volatile__ (
-+#undef BLOCK
-+#define BLOCK(i) \
-+ PF1(i) \
-+ PF1(i+2) \
-+ LD(i,0) \
-+ LD(i+1,1) \
-+ LD(i+2,2) \
-+ LD(i+3,3) \
-+ PF2(i) \
-+ PF2(i+2) \
-+ XO1(i,0) \
-+ XO1(i+1,1) \
-+ XO1(i+2,2) \
-+ XO1(i+3,3) \
-+ PF3(i) \
-+ PF3(i+2) \
-+ XO2(i,0) \
-+ XO2(i+1,1) \
-+ XO2(i+2,2) \
-+ XO2(i+3,3) \
-+ PF4(i) \
-+ PF4(i+2) \
-+ PF0(i+4) \
-+ PF0(i+6) \
-+ XO3(i,0) \
-+ XO3(i+1,1) \
-+ XO3(i+2,2) \
-+ XO3(i+3,3) \
-+ XO4(i,0) \
-+ XO4(i+1,1) \
-+ XO4(i+2,2) \
-+ XO4(i+3,3) \
-+ ST(i,0) \
-+ ST(i+1,1) \
-+ ST(i+2,2) \
-+ ST(i+3,3) \
++/* Return from callback. */
++#define SIOEMU_HYPERCALL_CALLBACK_RETURN 0x09
+
++/* Deliver an interrupt. */
++#define SIOEMU_HYPERCALL_DELIVER_INT 0x0a
+
-+ PF0(0)
-+ PF0(2)
++/* SIOEMU callback reason. */
+
-+ " .align 32 ;\n"
-+ " 1: ;\n"
++/* An event (from event channel) has to be delivered. */
++#define SIOEMU_CB_EVENT 0x00
+
-+ BLOCK(0)
-+ BLOCK(4)
-+ BLOCK(8)
-+ BLOCK(12)
++/* Emulate an IO access. */
++#define SIOEMU_CB_IO_EMULATE 0x01
+
-+ " addq %[inc], %[p1] ;\n"
-+ " addq %[inc], %[p2] ;\n"
-+ " addq %[inc], %[p3] ;\n"
-+ " addq %[inc], %[p4] ;\n"
-+ " addq %[inc], %[p5] ;\n"
-+ " decl %[cnt] ; jnz 1b"
-+ : [cnt] "+c" (lines),
-+ [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3), [p4] "+r" (p4),
-+ [p5] "+r" (p5)
-+ : [inc] "r" (256UL)
-+ : "memory");
++/* An IPI is sent to a dead vcpu. */
++#define SIOEMU_CB_WAKEUP_VCPU 0x02
+
-+ XMMS_RESTORE;
-+}
++/* A SAL hypercall is executed. */
++#define SIOEMU_CB_SAL_ASSIST 0x03
+
-+static struct xor_block_template xor_block_sse = {
-+ .name = "generic_sse",
-+ .do_2 = xor_sse_2,
-+ .do_3 = xor_sse_3,
-+ .do_4 = xor_sse_4,
-+ .do_5 = xor_sse_5,
++#ifndef __ASSEMBLY__
++struct sioemu_callback_info {
++ /* Saved registers. */
++ unsigned long ip;
++ unsigned long psr;
++ unsigned long ifs;
++ unsigned long nats;
++ unsigned long r8;
++ unsigned long r9;
++ unsigned long r10;
++ unsigned long r11;
++
++ /* Callback parameters. */
++ unsigned long cause;
++ unsigned long arg0;
++ unsigned long arg1;
++ unsigned long arg2;
++ unsigned long arg3;
++ unsigned long _pad2[2];
++ unsigned long r2;
+};
-+
-+#undef XOR_TRY_TEMPLATES
-+#define XOR_TRY_TEMPLATES \
-+ do { \
-+ xor_speed(&xor_block_sse); \
-+ } while (0)
-+
-+/* We force the use of the SSE xor block because it can write around L2.
-+ We may also be able to load into the L1 only depending on how the cpu
-+ deals with a load to a line that is being prefetched. */
-+#define XOR_SELECT_TEMPLATE(FASTEST) (&xor_block_sse)
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/irq_vectors.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/irq_vectors.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/irq_vectors.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/irq_vectors.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,123 @@
++#endif /* __ASSEMBLY__ */
++#endif /* __XEN_PUBLIC_IA64_SIOEMU_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-powerpc.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-powerpc.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,120 @@
+/*
-+ * This file should contain #defines for all of the interrupt vector
-+ * numbers used by this architecture.
-+ *
-+ * In addition, there are some standard defines:
-+ *
-+ * FIRST_EXTERNAL_VECTOR:
-+ * The first free place for external interrupts
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
+ *
-+ * SYSCALL_VECTOR:
-+ * The IRQ vector a syscall makes the user to kernel transition
-+ * under.
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
+ *
-+ * TIMER_IRQ:
-+ * The IRQ number the timer interrupt comes in at.
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
+ *
-+ * NR_IRQS:
-+ * The total number of interrupt vectors (including all the
-+ * architecture specific interrupts) needed.
++ * Copyright (C) IBM Corp. 2005, 2006
+ *
-+ */
-+#ifndef _ASM_IRQ_VECTORS_H
-+#define _ASM_IRQ_VECTORS_H
-+
-+/*
-+ * IDT vectors usable for external interrupt sources start
-+ * at 0x20:
++ * Authors: Hollis Blanchard <hollisb at us.ibm.com>
+ */
-+#define FIRST_EXTERNAL_VECTOR 0x20
+
-+#define SYSCALL_VECTOR 0x80
++#include "xen.h"
++
++#ifndef __XEN_PUBLIC_ARCH_PPC_64_H__
++#define __XEN_PUBLIC_ARCH_PPC_64_H__
++
++#define ___DEFINE_XEN_GUEST_HANDLE(name, type) \
++ typedef struct { \
++ int __pad[(sizeof (long long) - sizeof (void *)) / sizeof (int)]; \
++ type *p; \
++ } __attribute__((__aligned__(8))) __guest_handle_ ## name
++
++#define __DEFINE_XEN_GUEST_HANDLE(name, type) \
++ ___DEFINE_XEN_GUEST_HANDLE(name, type); \
++ ___DEFINE_XEN_GUEST_HANDLE(const_##name, const type)
++#define DEFINE_XEN_GUEST_HANDLE(name) __DEFINE_XEN_GUEST_HANDLE(name, name)
++#define XEN_GUEST_HANDLE(name) __guest_handle_ ## name
++#define set_xen_guest_handle(hnd, val) \
++ do { \
++ if (sizeof ((hnd).__pad)) \
++ (hnd).__pad[0] = 0; \
++ (hnd).p = val; \
++ } while (0)
++
++#ifdef __XEN_TOOLS__
++#define get_xen_guest_handle(val, hnd) do { val = (hnd).p; } while (0)
++#endif
++
++#ifndef __ASSEMBLY__
++typedef unsigned long long xen_pfn_t;
++#define PRI_xen_pfn "llx"
++#endif
+
+/*
-+ * Vectors 0x20-0x2f are used for ISA interrupts.
++ * Pointers and other address fields inside interface structures are padded to
++ * 64 bits. This means that field alignments aren't different between 32- and
++ * 64-bit architectures.
+ */
++/* NB. Multi-level macro ensures __LINE__ is expanded before concatenation. */
++#define __MEMORY_PADDING(_X)
++#define _MEMORY_PADDING(_X) __MEMORY_PADDING(_X)
++#define MEMORY_PADDING _MEMORY_PADDING(__LINE__)
+
-+#if 0
-+/*
-+ * Special IRQ vectors used by the SMP architecture, 0xf0-0xff
++/* And the trap vector is... */
++#define TRAP_INSTR "li 0,-1; sc" /* XXX just "sc"? */
++
++#ifndef __ASSEMBLY__
++
++#define XENCOMM_INLINE_FLAG (1UL << 63)
++
++typedef uint64_t xen_ulong_t;
++
++/* User-accessible registers: nost of these need to be saved/restored
++ * for every nested Xen invocation. */
++struct cpu_user_regs
++{
++ uint64_t gprs[32];
++ uint64_t lr;
++ uint64_t ctr;
++ uint64_t srr0;
++ uint64_t srr1;
++ uint64_t pc;
++ uint64_t msr;
++ uint64_t fpscr; /* XXX Is this necessary */
++ uint64_t xer;
++ uint64_t hid4; /* debug only */
++ uint64_t dar; /* debug only */
++ uint32_t dsisr; /* debug only */
++ uint32_t cr;
++ uint32_t __pad; /* good spot for another 32bit reg */
++ uint32_t entry_vector;
++};
++typedef struct cpu_user_regs cpu_user_regs_t;
++
++typedef uint64_t tsc_timestamp_t; /* RDTSC timestamp */ /* XXX timebase */
++
++/* ONLY used to communicate with dom0! See also struct exec_domain. */
++struct vcpu_guest_context {
++ cpu_user_regs_t user_regs; /* User-level CPU registers */
++ uint64_t sdr1; /* Pagetable base */
++ /* XXX etc */
++};
++typedef struct vcpu_guest_context vcpu_guest_context_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_guest_context_t);
++
++struct arch_shared_info {
++ uint64_t boot_timebase;
++};
++
++struct arch_vcpu_info {
++};
++
++/* Support for multi-processor guests. */
++#define MAX_VIRT_CPUS 32
++#endif
++
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-x86/cpuid.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-x86/cpuid.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,68 @@
++/******************************************************************************
++ * arch-x86/cpuid.h
++ *
++ * CPUID interface to Xen.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
+ *
-+ * some of the following vectors are 'rare', they are merged
-+ * into a single vector (CALL_FUNCTION_VECTOR) to save vector space.
-+ * TLB, reschedule and local APIC vectors are performance-critical.
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
+ *
-+ * Vectors 0xf0-0xfa are free (reserved for future Linux use).
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2007 Citrix Systems, Inc.
++ *
++ * Authors:
++ * Keir Fraser <keir.fraser at citrix.com>
+ */
-+#define INVALIDATE_TLB_VECTOR 0xfd
-+#define RESCHEDULE_VECTOR 0xfc
-+#define CALL_FUNCTION_VECTOR 0xfb
+
-+#define THERMAL_APIC_VECTOR 0xf0
++#ifndef __XEN_PUBLIC_ARCH_X86_CPUID_H__
++#define __XEN_PUBLIC_ARCH_X86_CPUID_H__
++
++/* Xen identification leaves start at 0x40000000. */
++#define XEN_CPUID_FIRST_LEAF 0x40000000
++#define XEN_CPUID_LEAF(i) (XEN_CPUID_FIRST_LEAF + (i))
++
+/*
-+ * Local APIC timer IRQ vector is on a different priority level,
-+ * to work around the 'lost local interrupt if more than 2 IRQ
-+ * sources per level' errata.
++ * Leaf 1 (0x40000000)
++ * EAX: Largest Xen-information leaf. All leaves up to an including @EAX
++ * are supported by the Xen host.
++ * EBX-EDX: "XenVMMXenVMM" signature, allowing positive identification
++ * of a Xen host.
+ */
-+#define LOCAL_TIMER_VECTOR 0xef
-+#endif
-+
-+#define SPURIOUS_APIC_VECTOR 0xff
-+#define ERROR_APIC_VECTOR 0xfe
++#define XEN_CPUID_SIGNATURE_EBX 0x566e6558 /* "XenV" */
++#define XEN_CPUID_SIGNATURE_ECX 0x65584d4d /* "MMXe" */
++#define XEN_CPUID_SIGNATURE_EDX 0x4d4d566e /* "nVMM" */
+
+/*
-+ * First APIC vector available to drivers: (vectors 0x30-0xee)
-+ * we start at 0x31 to spread out vectors evenly between priority
-+ * levels. (0x80 is the syscall vector)
++ * Leaf 2 (0x40000001)
++ * EAX[31:16]: Xen major version.
++ * EAX[15: 0]: Xen minor version.
++ * EBX-EDX: Reserved (currently all zeroes).
+ */
-+#define FIRST_DEVICE_VECTOR 0x31
-+#define FIRST_SYSTEM_VECTOR 0xef
+
+/*
-+ * 16 8259A IRQ's, 208 potential APIC interrupt sources.
-+ * Right now the APIC is mostly only used for SMP.
-+ * 256 vectors is an architectural limit. (we can have
-+ * more than 256 devices theoretically, but they will
-+ * have to use shared interrupts)
-+ * Since vectors 0x00-0x1f are used/reserved for the CPU,
-+ * the usable vector space is 0x20-0xff (224 vectors)
++ * Leaf 3 (0x40000002)
++ * EAX: Number of hypercall transfer pages. This register is always guaranteed
++ * to specify one hypercall page.
++ * EBX: Base address of Xen-specific MSRs.
++ * ECX: Features 1. Unused bits are set to zero.
++ * EDX: Features 2. Unused bits are set to zero.
+ */
+
-+#define RESCHEDULE_VECTOR 0
-+#define CALL_FUNCTION_VECTOR 1
-+#define NR_IPIS 2
++/* Does the host support MMU_PT_UPDATE_PRESERVE_AD for this guest? */
++#define _XEN_CPUID_FEAT1_MMU_PT_UPDATE_PRESERVE_AD 0
++#define XEN_CPUID_FEAT1_MMU_PT_UPDATE_PRESERVE_AD (1u<<0)
++
++#endif /* __XEN_PUBLIC_ARCH_X86_CPUID_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-x86/hvm/save.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-x86/hvm/save.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,429 @@
++/*
++ * Structure definitions for HVM state that is held by Xen and must
++ * be saved along with the domain's memory and device-model state.
++ *
++ * Copyright (c) 2007 XenSource Ltd.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_PUBLIC_HVM_SAVE_X86_H__
++#define __XEN_PUBLIC_HVM_SAVE_X86_H__
++
++/*
++ * Save/restore header: general info about the save file.
++ */
++
++#define HVM_FILE_MAGIC 0x54381286
++#define HVM_FILE_VERSION 0x00000001
++
++struct hvm_save_header {
++ uint32_t magic; /* Must be HVM_FILE_MAGIC */
++ uint32_t version; /* File format version */
++ uint64_t changeset; /* Version of Xen that saved this file */
++ uint32_t cpuid; /* CPUID[0x01][%eax] on the saving machine */
++ uint32_t pad0;
++};
++
++DECLARE_HVM_SAVE_TYPE(HEADER, 1, struct hvm_save_header);
++
+
+/*
-+ * The maximum number of vectors supported by i386 processors
-+ * is limited to 256. For processors other than i386, NR_VECTORS
-+ * should be changed accordingly.
++ * Processor
+ */
-+#define NR_VECTORS 256
+
-+#define FPU_IRQ 13
++struct hvm_hw_cpu {
++ uint8_t fpu_regs[512];
++
++ uint64_t rax;
++ uint64_t rbx;
++ uint64_t rcx;
++ uint64_t rdx;
++ uint64_t rbp;
++ uint64_t rsi;
++ uint64_t rdi;
++ uint64_t rsp;
++ uint64_t r8;
++ uint64_t r9;
++ uint64_t r10;
++ uint64_t r11;
++ uint64_t r12;
++ uint64_t r13;
++ uint64_t r14;
++ uint64_t r15;
++
++ uint64_t rip;
++ uint64_t rflags;
++
++ uint64_t cr0;
++ uint64_t cr2;
++ uint64_t cr3;
++ uint64_t cr4;
++
++ uint64_t dr0;
++ uint64_t dr1;
++ uint64_t dr2;
++ uint64_t dr3;
++ uint64_t dr6;
++ uint64_t dr7;
++
++ uint32_t cs_sel;
++ uint32_t ds_sel;
++ uint32_t es_sel;
++ uint32_t fs_sel;
++ uint32_t gs_sel;
++ uint32_t ss_sel;
++ uint32_t tr_sel;
++ uint32_t ldtr_sel;
++
++ uint32_t cs_limit;
++ uint32_t ds_limit;
++ uint32_t es_limit;
++ uint32_t fs_limit;
++ uint32_t gs_limit;
++ uint32_t ss_limit;
++ uint32_t tr_limit;
++ uint32_t ldtr_limit;
++ uint32_t idtr_limit;
++ uint32_t gdtr_limit;
++
++ uint64_t cs_base;
++ uint64_t ds_base;
++ uint64_t es_base;
++ uint64_t fs_base;
++ uint64_t gs_base;
++ uint64_t ss_base;
++ uint64_t tr_base;
++ uint64_t ldtr_base;
++ uint64_t idtr_base;
++ uint64_t gdtr_base;
++
++ uint32_t cs_arbytes;
++ uint32_t ds_arbytes;
++ uint32_t es_arbytes;
++ uint32_t fs_arbytes;
++ uint32_t gs_arbytes;
++ uint32_t ss_arbytes;
++ uint32_t tr_arbytes;
++ uint32_t ldtr_arbytes;
++
++ uint32_t sysenter_cs;
++ uint32_t padding0;
++
++ uint64_t sysenter_esp;
++ uint64_t sysenter_eip;
++
++ /* msr for em64t */
++ uint64_t shadow_gs;
++
++ /* msr content saved/restored. */
++ uint64_t msr_flags;
++ uint64_t msr_lstar;
++ uint64_t msr_star;
++ uint64_t msr_cstar;
++ uint64_t msr_syscall_mask;
++ uint64_t msr_efer;
++
++ /* guest's idea of what rdtsc() would return */
++ uint64_t tsc;
++
++ /* pending event, if any */
++ union {
++ uint32_t pending_event;
++ struct {
++ uint8_t pending_vector:8;
++ uint8_t pending_type:3;
++ uint8_t pending_error_valid:1;
++ uint32_t pending_reserved:19;
++ uint8_t pending_valid:1;
++ };
++ };
++ /* error code for pending event */
++ uint32_t error_code;
++};
++
++DECLARE_HVM_SAVE_TYPE(CPU, 2, struct hvm_hw_cpu);
+
-+#define FIRST_VM86_IRQ 3
-+#define LAST_VM86_IRQ 15
-+#define invalid_vm86_irq(irq) ((irq) < 3 || (irq) > 15)
+
+/*
-+ * The flat IRQ space is divided into two regions:
-+ * 1. A one-to-one mapping of real physical IRQs. This space is only used
-+ * if we have physical device-access privilege. This region is at the
-+ * start of the IRQ space so that existing device drivers do not need
-+ * to be modified to translate physical IRQ numbers into our IRQ space.
-+ * 3. A dynamic mapping of inter-domain and Xen-sourced virtual IRQs. These
-+ * are bound using the provided bind/unbind functions.
++ * PIC
+ */
+
-+#define PIRQ_BASE 0
-+#define NR_PIRQS 256
++struct hvm_hw_vpic {
++ /* IR line bitmasks. */
++ uint8_t irr;
++ uint8_t imr;
++ uint8_t isr;
++
++ /* Line IRx maps to IRQ irq_base+x */
++ uint8_t irq_base;
++
++ /*
++ * Where are we in ICW2-4 initialisation (0 means no init in progress)?
++ * Bits 0-1 (=x): Next write at A=1 sets ICW(x+1).
++ * Bit 2: ICW1.IC4 (1 == ICW4 included in init sequence)
++ * Bit 3: ICW1.SNGL (0 == ICW3 included in init sequence)
++ */
++ uint8_t init_state:4;
++
++ /* IR line with highest priority. */
++ uint8_t priority_add:4;
++
++ /* Reads from A=0 obtain ISR or IRR? */
++ uint8_t readsel_isr:1;
++
++ /* Reads perform a polling read? */
++ uint8_t poll:1;
++
++ /* Automatically clear IRQs from the ISR during INTA? */
++ uint8_t auto_eoi:1;
+
-+#define DYNIRQ_BASE (PIRQ_BASE + NR_PIRQS)
-+#define NR_DYNIRQS 256
++ /* Automatically rotate IRQ priorities during AEOI? */
++ uint8_t rotate_on_auto_eoi:1;
+
-+#define NR_IRQS (NR_PIRQS + NR_DYNIRQS)
-+#define NR_IRQ_VECTORS NR_IRQS
++ /* Exclude slave inputs when considering in-service IRQs? */
++ uint8_t special_fully_nested_mode:1;
+
-+#define pirq_to_irq(_x) ((_x) + PIRQ_BASE)
-+#define irq_to_pirq(_x) ((_x) - PIRQ_BASE)
++ /* Special mask mode excludes masked IRs from AEOI and priority checks. */
++ uint8_t special_mask_mode:1;
+
-+#define dynirq_to_irq(_x) ((_x) + DYNIRQ_BASE)
-+#define irq_to_dynirq(_x) ((_x) - DYNIRQ_BASE)
++ /* Is this a master PIC or slave PIC? (NB. This is not programmable.) */
++ uint8_t is_master:1;
+
-+#endif /* _ASM_IRQ_VECTORS_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/mach_time.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/mach_time.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/mach_time.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/mach_time.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,111 @@
-+/*
-+ * include/asm-i386/mach-default/mach_time.h
-+ *
-+ * Machine specific set RTC function for generic.
-+ * Split out from time.c by Osamu Tomita <tomita at cinet.co.jp>
-+ */
-+#ifndef _MACH_TIME_H
-+#define _MACH_TIME_H
++ /* Edge/trigger selection. */
++ uint8_t elcr;
+
-+#include <asm-i386/mc146818rtc.h>
++ /* Virtual INT output. */
++ uint8_t int_output;
++};
++
++DECLARE_HVM_SAVE_TYPE(PIC, 3, struct hvm_hw_vpic);
+
-+/* for check timing call set_rtc_mmss() 500ms */
-+/* used in arch/i386/time.c::do_timer_interrupt() */
-+#define USEC_AFTER 500000
-+#define USEC_BEFORE 500000
+
+/*
-+ * In order to set the CMOS clock precisely, set_rtc_mmss has to be
-+ * called 500 ms after the second nowtime has started, because when
-+ * nowtime is written into the registers of the CMOS clock, it will
-+ * jump to the next second precisely 500 ms later. Check the Motorola
-+ * MC146818A or Dallas DS12887 data sheet for details.
-+ *
-+ * BUG: This routine does not handle hour overflow properly; it just
-+ * sets the minutes. Usually you'll only notice that after reboot!
++ * IO-APIC
+ */
-+static inline int mach_set_rtc_mmss(unsigned long nowtime)
-+{
-+ int retval = 0;
-+ int real_seconds, real_minutes, cmos_minutes;
-+ unsigned char save_control, save_freq_select;
-+
-+ save_control = CMOS_READ(RTC_CONTROL); /* tell the clock it's being set */
-+ CMOS_WRITE((save_control|RTC_SET), RTC_CONTROL);
-+
-+ save_freq_select = CMOS_READ(RTC_FREQ_SELECT); /* stop and reset prescaler */
-+ CMOS_WRITE((save_freq_select|RTC_DIV_RESET2), RTC_FREQ_SELECT);
+
-+ cmos_minutes = CMOS_READ(RTC_MINUTES);
-+ if (!(save_control & RTC_DM_BINARY) || RTC_ALWAYS_BCD)
-+ BCD_TO_BIN(cmos_minutes);
++#ifdef __ia64__
++#define VIOAPIC_IS_IOSAPIC 1
++#define VIOAPIC_NUM_PINS 24
++#else
++#define VIOAPIC_NUM_PINS 48 /* 16 ISA IRQs, 32 non-legacy PCI IRQS. */
++#endif
+
-+ /*
-+ * since we're only adjusting minutes and seconds,
-+ * don't interfere with hour overflow. This avoids
-+ * messing with unknown time zones but requires your
-+ * RTC not to be off by more than 15 minutes
-+ */
-+ real_seconds = nowtime % 60;
-+ real_minutes = nowtime / 60;
-+ if (((abs(real_minutes - cmos_minutes) + 15)/30) & 1)
-+ real_minutes += 30; /* correct for half hour time zone */
-+ real_minutes %= 60;
++struct hvm_hw_vioapic {
++ uint64_t base_address;
++ uint32_t ioregsel;
++ uint32_t id;
++ union vioapic_redir_entry
++ {
++ uint64_t bits;
++ struct {
++ uint8_t vector;
++ uint8_t delivery_mode:3;
++ uint8_t dest_mode:1;
++ uint8_t delivery_status:1;
++ uint8_t polarity:1;
++ uint8_t remote_irr:1;
++ uint8_t trig_mode:1;
++ uint8_t mask:1;
++ uint8_t reserve:7;
++#if !VIOAPIC_IS_IOSAPIC
++ uint8_t reserved[4];
++ uint8_t dest_id;
++#else
++ uint8_t reserved[3];
++ uint16_t dest_id;
++#endif
++ } fields;
++ } redirtbl[VIOAPIC_NUM_PINS];
++};
+
-+ if (abs(real_minutes - cmos_minutes) < 30) {
-+ if (!(save_control & RTC_DM_BINARY) || RTC_ALWAYS_BCD) {
-+ BIN_TO_BCD(real_seconds);
-+ BIN_TO_BCD(real_minutes);
-+ }
-+ CMOS_WRITE(real_seconds,RTC_SECONDS);
-+ CMOS_WRITE(real_minutes,RTC_MINUTES);
-+ } else {
-+ printk(KERN_WARNING
-+ "set_rtc_mmss: can't update from %d to %d\n",
-+ cmos_minutes, real_minutes);
-+ retval = -1;
-+ }
++DECLARE_HVM_SAVE_TYPE(IOAPIC, 4, struct hvm_hw_vioapic);
+
-+ /* The following flags have to be released exactly in this order,
-+ * otherwise the DS12887 (popular MC146818A clone with integrated
-+ * battery and quartz) will not reset the oscillator and will not
-+ * update precisely 500 ms later. You won't find this mentioned in
-+ * the Dallas Semiconductor data sheets, but who believes data
-+ * sheets anyway ... -- Markus Kuhn
-+ */
-+ CMOS_WRITE(save_control, RTC_CONTROL);
-+ CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT);
+
-+ return retval;
-+}
++/*
++ * LAPIC
++ */
+
-+static inline unsigned long mach_get_cmos_time(void)
-+{
-+ unsigned int year, mon, day, hour, min, sec;
++struct hvm_hw_lapic {
++ uint64_t apic_base_msr;
++ uint32_t disabled; /* VLAPIC_xx_DISABLED */
++ uint32_t timer_divisor;
++};
+
-+ do {
-+ sec = CMOS_READ(RTC_SECONDS);
-+ min = CMOS_READ(RTC_MINUTES);
-+ hour = CMOS_READ(RTC_HOURS);
-+ day = CMOS_READ(RTC_DAY_OF_MONTH);
-+ mon = CMOS_READ(RTC_MONTH);
-+ year = CMOS_READ(RTC_YEAR);
-+ } while (sec != CMOS_READ(RTC_SECONDS));
++DECLARE_HVM_SAVE_TYPE(LAPIC, 5, struct hvm_hw_lapic);
+
-+ if (!(CMOS_READ(RTC_CONTROL) & RTC_DM_BINARY) || RTC_ALWAYS_BCD) {
-+ BCD_TO_BIN(sec);
-+ BCD_TO_BIN(min);
-+ BCD_TO_BIN(hour);
-+ BCD_TO_BIN(day);
-+ BCD_TO_BIN(mon);
-+ BCD_TO_BIN(year);
-+ }
++struct hvm_hw_lapic_regs {
++ uint8_t data[1024];
++};
+
-+ year += 1900;
-+ if (year < 1970)
-+ year += 100;
++DECLARE_HVM_SAVE_TYPE(LAPIC_REGS, 6, struct hvm_hw_lapic_regs);
+
-+ return mktime(year, mon, day, hour, min, sec);
-+}
+
-+#endif /* !_MACH_TIME_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/mach_timer.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/mach_timer.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/mach_timer.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/mach_timer.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,50 @@
+/*
-+ * include/asm-i386/mach-default/mach_timer.h
-+ *
-+ * Machine specific calibrate_tsc() for generic.
-+ * Split out from timer_tsc.c by Osamu Tomita <tomita at cinet.co.jp>
-+ */
-+/* ------ Calibrate the TSC -------
-+ * Return 2^32 * (1 / (TSC clocks per usec)) for do_fast_gettimeoffset().
-+ * Too much 64-bit arithmetic here to do this cleanly in C, and for
-+ * accuracy's sake we want to keep the overhead on the CTC speaker (channel 2)
-+ * output busy loop as low as possible. We avoid reading the CTC registers
-+ * directly because of the awkward 8-bit access mechanism of the 82C54
-+ * device.
++ * IRQs
+ */
-+#ifndef _MACH_TIMER_H
-+#define _MACH_TIMER_H
-+
-+#define CALIBRATE_TIME_MSEC 30 /* 30 msecs */
-+#define CALIBRATE_LATCH \
-+ ((CLOCK_TICK_RATE * CALIBRATE_TIME_MSEC + 1000/2)/1000)
+
-+static inline void mach_prepare_counter(void)
-+{
-+ /* Set the Gate high, disable speaker */
-+ outb((inb(0x61) & ~0x02) | 0x01, 0x61);
++struct hvm_hw_pci_irqs {
++ /*
++ * Virtual interrupt wires for a single PCI bus.
++ * Indexed by: device*4 + INTx#.
++ */
++ union {
++ DECLARE_BITMAP(i, 32*4);
++ uint64_t pad[2];
++ };
++};
+
-+ /*
-+ * Now let's take care of CTC channel 2
-+ *
-+ * Set the Gate high, program CTC channel 2 for mode 0,
-+ * (interrupt on terminal count mode), binary count,
-+ * load 5 * LATCH count, (LSB and MSB) to begin countdown.
-+ *
-+ * Some devices need a delay here.
-+ */
-+ outb(0xb0, 0x43); /* binary, mode 0, LSB/MSB, Ch 2 */
-+ outb_p(CALIBRATE_LATCH & 0xff, 0x42); /* LSB of count */
-+ outb_p(CALIBRATE_LATCH >> 8, 0x42); /* MSB of count */
-+}
++DECLARE_HVM_SAVE_TYPE(PCI_IRQ, 7, struct hvm_hw_pci_irqs);
+
-+static inline void mach_countup(unsigned long *count_p)
-+{
-+ unsigned long count = 0;
-+ do {
-+ count++;
-+ } while ((inb_p(0x61) & 0x20) == 0);
-+ *count_p = count;
-+}
++struct hvm_hw_isa_irqs {
++ /*
++ * Virtual interrupt wires for ISA devices.
++ * Indexed by ISA IRQ (assumes no ISA-device IRQ sharing).
++ */
++ union {
++ DECLARE_BITMAP(i, 16);
++ uint64_t pad[1];
++ };
++};
+
-+#endif /* !_MACH_TIMER_H */
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/setup_arch_post.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/setup_arch_post.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/setup_arch_post.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/setup_arch_post.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,63 @@
-+/**
-+ * machine_specific_* - Hooks for machine specific setup.
-+ *
-+ * Description:
-+ * This is included late in kernel/setup.c so that it can make
-+ * use of all of the static functions.
-+ **/
++DECLARE_HVM_SAVE_TYPE(ISA_IRQ, 8, struct hvm_hw_isa_irqs);
+
-+#include <xen/interface/callback.h>
++struct hvm_hw_pci_link {
++ /*
++ * PCI-ISA interrupt router.
++ * Each PCI <device:INTx#> is 'wire-ORed' into one of four links using
++ * the traditional 'barber's pole' mapping ((device + INTx#) & 3).
++ * The router provides a programmable mapping from each link to a GSI.
++ */
++ uint8_t route[4];
++ uint8_t pad0[4];
++};
+
-+extern void hypervisor_callback(void);
-+extern void failsafe_callback(void);
-+extern void nmi(void);
++DECLARE_HVM_SAVE_TYPE(PCI_LINK, 9, struct hvm_hw_pci_link);
+
-+static void __init machine_specific_arch_setup(void)
-+{
-+ int ret;
-+ static struct callback_register __initdata event = {
-+ .type = CALLBACKTYPE_event,
-+ .address = (unsigned long) hypervisor_callback,
-+ };
-+ static struct callback_register __initdata failsafe = {
-+ .type = CALLBACKTYPE_failsafe,
-+ .address = (unsigned long)failsafe_callback,
-+ };
-+ static struct callback_register __initdata syscall = {
-+ .type = CALLBACKTYPE_syscall,
-+ .address = (unsigned long)system_call,
-+ };
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ static struct callback_register __initdata nmi_cb = {
-+ .type = CALLBACKTYPE_nmi,
-+ .address = (unsigned long)nmi,
-+ };
-+#endif
++/*
++ * PIT
++ */
+
-+ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &event);
-+ if (ret == 0)
-+ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &failsafe);
-+ if (ret == 0)
-+ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &syscall);
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (ret == -ENOSYS)
-+ ret = HYPERVISOR_set_callbacks(
-+ event.address,
-+ failsafe.address,
-+ syscall.address);
-+#endif
-+ BUG_ON(ret);
++struct hvm_hw_pit {
++ struct hvm_hw_pit_channel {
++ uint32_t count; /* can be 65536 */
++ uint16_t latched_count;
++ uint8_t count_latched;
++ uint8_t status_latched;
++ uint8_t status;
++ uint8_t read_state;
++ uint8_t write_state;
++ uint8_t write_latch;
++ uint8_t rw_mode;
++ uint8_t mode;
++ uint8_t bcd; /* not supported */
++ uint8_t gate; /* timer start */
++ } channels[3]; /* 3 x 16 bytes */
++ uint32_t speaker_data_on;
++ uint32_t pad0;
++};
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &nmi_cb);
-+#if CONFIG_XEN_COMPAT <= 0x030002
-+ if (ret == -ENOSYS) {
-+ static struct xennmi_callback __initdata cb = {
-+ .handler_address = (unsigned long)nmi
-+ };
++DECLARE_HVM_SAVE_TYPE(PIT, 10, struct hvm_hw_pit);
+
-+ HYPERVISOR_nmi_op(XENNMI_register_callback, &cb);
-+ }
-+#endif
-+#endif
-+}
-diff -Nurp pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/setup_arch_pre.h tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/setup_arch_pre.h
---- pristine-linux-2.6.18.2/include/asm-x86_64/mach-xen/setup_arch_pre.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/asm-x86_64/mach-xen/setup_arch_pre.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,5 @@
-+/* Hook to call BIOS initialisation function */
+
-+#define ARCH_SETUP machine_specific_arch_setup();
++/*
++ * RTC
++ */
+
-+static void __init machine_specific_arch_setup(void);
-diff -Nurp pristine-linux-2.6.18.2/include/linux/aio.h tmp-linux-2.6-xen.patch/include/linux/aio.h
---- pristine-linux-2.6.18.2/include/linux/aio.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/aio.h 2007-10-14 01:51:15.000000000 +0200
-@@ -191,6 +191,11 @@ struct kioctx {
- struct aio_ring_info ring_info;
-
- struct work_struct wq;
-+#ifdef CONFIG_EPOLL
-+ // poll integration
-+ wait_queue_head_t poll_wait;
-+ struct file *file;
-+#endif
- };
-
- /* prototypes */
-diff -Nurp pristine-linux-2.6.18.2/include/linux/crash_dump.h tmp-linux-2.6-xen.patch/include/linux/crash_dump.h
---- pristine-linux-2.6.18.2/include/linux/crash_dump.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/crash_dump.h 2007-10-14 01:51:15.000000000 +0200
-@@ -14,5 +14,13 @@ extern ssize_t copy_oldmem_page(unsigned
- extern const struct file_operations proc_vmcore_operations;
- extern struct proc_dir_entry *proc_vmcore;
-
-+/* Architecture code defines this if there are other possible ELF
-+ * machine types, e.g. on bi-arch capable hardware. */
-+#ifndef vmcore_elf_check_arch_cross
-+#define vmcore_elf_check_arch_cross(x) 0
-+#endif
++#define RTC_CMOS_SIZE 14
++struct hvm_hw_rtc {
++ /* CMOS bytes */
++ uint8_t cmos_data[RTC_CMOS_SIZE];
++ /* Index register for 2-part operations */
++ uint8_t cmos_index;
++ uint8_t pad0;
++};
+
-+#define vmcore_elf_check_arch(x) (elf_check_arch(x) || vmcore_elf_check_arch_cross(x))
++DECLARE_HVM_SAVE_TYPE(RTC, 11, struct hvm_hw_rtc);
+
- #endif /* CONFIG_CRASH_DUMP */
- #endif /* LINUX_CRASHDUMP_H */
-diff -Nurp pristine-linux-2.6.18.2/include/linux/elfnote.h tmp-linux-2.6-xen.patch/include/linux/elfnote.h
---- pristine-linux-2.6.18.2/include/linux/elfnote.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/elfnote.h 2007-10-14 01:51:15.000000000 +0200
-@@ -0,0 +1,104 @@
-+#ifndef _LINUX_ELFNOTE_H
-+#define _LINUX_ELFNOTE_H
-+/*
-+ * Helper macros to generate ELF Note structures, which are put into a
-+ * PT_NOTE segment of the final vmlinux image. These are useful for
-+ * including name-value pairs of metadata into the kernel binary (or
-+ * modules?) for use by external programs.
-+ *
-+ * Each note has three parts: a name, a type and a desc. The name is
-+ * intended to distinguish the note's originator, so it would be a
-+ * company, project, subsystem, etc; it must be in a suitable form for
-+ * use in a section name. The type is an integer which is used to tag
-+ * the data, and is considered to be within the "name" namespace (so
-+ * "FooCo"'s type 42 is distinct from "BarProj"'s type 42). The
-+ * "desc" field is the actual data. There are no constraints on the
-+ * desc field's contents, though typically they're fairly small.
-+ *
-+ * All notes from a given NAME are put into a section named
-+ * .note.NAME. When the kernel image is finally linked, all the notes
-+ * are packed into a single .notes section, which is mapped into the
-+ * PT_NOTE segment. Because notes for a given name are grouped into
-+ * the same section, they'll all be adjacent the output file.
-+ *
-+ * This file defines macros for both C and assembler use. Their
-+ * syntax is slightly different, but they're semantically similar.
-+ *
-+ * See the ELF specification for more detail about ELF notes.
-+ */
+
-+#ifdef __ASSEMBLER__
-+/*
-+ * Generate a structure with the same shape as Elf{32,64}_Nhdr (which
-+ * turn out to be the same size and shape), followed by the name and
-+ * desc data with appropriate padding. The 'desctype' argument is the
-+ * assembler pseudo op defining the type of the data e.g. .asciz while
-+ * 'descdata' is the data itself e.g. "hello, world".
-+ *
-+ * e.g. ELFNOTE(XYZCo, 42, .asciz, "forty-two")
-+ * ELFNOTE(XYZCo, 12, .long, 0xdeadbeef)
-+ */
-+#ifdef __STDC__
-+#define ELFNOTE(name, type, desctype, descdata...) \
-+.pushsection .note.name ; \
-+ .align 4 ; \
-+ .long 2f - 1f /* namesz */ ; \
-+ .long 4f - 3f /* descsz */ ; \
-+ .long type ; \
-+1:.asciz #name ; \
-+2:.align 4 ; \
-+3:desctype descdata ; \
-+4:.align 4 ; \
-+.popsection
-+#else /* !__STDC__, i.e. -traditional */
-+#define ELFNOTE(name, type, desctype, descdata) \
-+.pushsection .note.name ; \
-+ .align 4 ; \
-+ .long 2f - 1f /* namesz */ ; \
-+ .long 4f - 3f /* descsz */ ; \
-+ .long type ; \
-+1:.asciz "name" ; \
-+2:.align 4 ; \
-+3:desctype descdata ; \
-+4:.align 4 ; \
-+.popsection
-+#endif /* __STDC__ */
-+#else /* !__ASSEMBLER__ */
-+#include <linux/elf.h>
+/*
-+ * Use an anonymous structure which matches the shape of
-+ * Elf{32,64}_Nhdr, but includes the name and desc data. The size and
-+ * type of name and desc depend on the macro arguments. "name" must
-+ * be a literal string, and "desc" must be passed by value. You may
-+ * only define one note per line, since __LINE__ is used to generate
-+ * unique symbols.
++ * HPET
+ */
-+#define _ELFNOTE_PASTE(a,b) a##b
-+#define _ELFNOTE(size, name, unique, type, desc) \
-+ static const struct { \
-+ struct elf##size##_note _nhdr; \
-+ unsigned char _name[sizeof(name)] \
-+ __attribute__((aligned(sizeof(Elf##size##_Word)))); \
-+ typeof(desc) _desc \
-+ __attribute__((aligned(sizeof(Elf##size##_Word)))); \
-+ } _ELFNOTE_PASTE(_note_, unique) \
-+ __attribute_used__ \
-+ __attribute__((section(".note." name), \
-+ aligned(sizeof(Elf##size##_Word)), \
-+ unused)) = { \
-+ { \
-+ sizeof(name), \
-+ sizeof(desc), \
-+ type, \
-+ }, \
-+ name, \
-+ desc \
-+ }
-+#define ELFNOTE(size, name, type, desc) \
-+ _ELFNOTE(size, name, __LINE__, type, desc)
-+
-+#define ELFNOTE32(name, type, desc) ELFNOTE(32, name, type, desc)
-+#define ELFNOTE64(name, type, desc) ELFNOTE(64, name, type, desc)
-+#endif /* __ASSEMBLER__ */
+
-+#endif /* _LINUX_ELFNOTE_H */
-diff -Nurp pristine-linux-2.6.18.2/include/linux/eventpoll.h tmp-linux-2.6-xen.patch/include/linux/eventpoll.h
---- pristine-linux-2.6.18.2/include/linux/eventpoll.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/eventpoll.h 2007-10-14 01:51:15.000000000 +0200
-@@ -90,6 +90,12 @@ static inline void eventpoll_release(str
- eventpoll_release_file(file);
- }
-
-+/*
-+ * called by aio code to create fd that can poll the aio event queueQ
-+ */
-+struct eventpoll;
-+int ep_getfd(int *efd, struct inode **einode, struct file **efile,
-+ struct eventpoll *ep, const struct file_operations *fops);
- #else
-
- static inline void eventpoll_init_file(struct file *file) {}
-diff -Nurp pristine-linux-2.6.18.2/include/linux/gfp.h tmp-linux-2.6-xen.patch/include/linux/gfp.h
---- pristine-linux-2.6.18.2/include/linux/gfp.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/gfp.h 2007-07-30 16:35:13.000000000 +0200
-@@ -99,7 +99,11 @@ static inline int gfp_zone(gfp_t gfp)
- */
-
- #ifndef HAVE_ARCH_FREE_PAGE
--static inline void arch_free_page(struct page *page, int order) { }
-+/*
-+ * If arch_free_page returns non-zero then the generic free_page code can
-+ * immediately bail: the arch-specific function has done all the work.
-+ */
-+static inline int arch_free_page(struct page *page, int order) { return 0; }
- #endif
-
- extern struct page *
-diff -Nurp pristine-linux-2.6.18.2/include/linux/highmem.h tmp-linux-2.6-xen.patch/include/linux/highmem.h
---- pristine-linux-2.6.18.2/include/linux/highmem.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/highmem.h 2007-07-30 16:35:13.000000000 +0200
-@@ -24,10 +24,16 @@ static inline void flush_kernel_dcache_p
-
- /* declarations for linux/mm/highmem.c */
- unsigned int nr_free_highpages(void);
-+#ifdef CONFIG_XEN
-+void kmap_flush_unused(void);
-+#endif
-
- #else /* CONFIG_HIGHMEM */
-
- static inline unsigned int nr_free_highpages(void) { return 0; }
-+#ifdef CONFIG_XEN
-+static inline void kmap_flush_unused(void) { }
-+#endif
-
- static inline void *kmap(struct page *page)
- {
-diff -Nurp pristine-linux-2.6.18.2/include/linux/interrupt.h tmp-linux-2.6-xen.patch/include/linux/interrupt.h
---- pristine-linux-2.6.18.2/include/linux/interrupt.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/interrupt.h 2007-07-30 16:35:13.000000000 +0200
-@@ -166,6 +166,12 @@ static inline int disable_irq_wake(unsig
-
- #endif /* CONFIG_GENERIC_HARDIRQS */
-
-+#ifdef CONFIG_HAVE_IRQ_IGNORE_UNHANDLED
-+int irq_ignore_unhandled(unsigned int irq);
-+#else
-+#define irq_ignore_unhandled(irq) 0
-+#endif
++#define HPET_TIMER_NUM 3 /* 3 timers supported now */
++struct hvm_hw_hpet {
++ /* Memory-mapped, software visible registers */
++ uint64_t capability; /* capabilities */
++ uint64_t res0; /* reserved */
++ uint64_t config; /* configuration */
++ uint64_t res1; /* reserved */
++ uint64_t isr; /* interrupt status reg */
++ uint64_t res2[25]; /* reserved */
++ uint64_t mc64; /* main counter */
++ uint64_t res3; /* reserved */
++ struct { /* timers */
++ uint64_t config; /* configuration/cap */
++ uint64_t cmp; /* comparator */
++ uint64_t fsb; /* FSB route, not supported now */
++ uint64_t res4; /* reserved */
++ } timers[HPET_TIMER_NUM];
++ uint64_t res5[4*(24-HPET_TIMER_NUM)]; /* reserved, up to 0x3ff */
+
- #ifndef __ARCH_SET_SOFTIRQ_PENDING
- #define set_softirq_pending(x) (local_softirq_pending() = (x))
- #define or_softirq_pending(x) (local_softirq_pending() |= (x))
-diff -Nurp pristine-linux-2.6.18.2/include/linux/kexec.h tmp-linux-2.6-xen.patch/include/linux/kexec.h
---- pristine-linux-2.6.18.2/include/linux/kexec.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/kexec.h 2007-07-30 16:35:13.000000000 +0200
-@@ -31,6 +31,13 @@
- #error KEXEC_ARCH not defined
- #endif
-
-+#ifndef KEXEC_ARCH_HAS_PAGE_MACROS
-+#define kexec_page_to_pfn(page) page_to_pfn(page)
-+#define kexec_pfn_to_page(pfn) pfn_to_page(pfn)
-+#define kexec_virt_to_phys(addr) virt_to_phys(addr)
-+#define kexec_phys_to_virt(addr) phys_to_virt(addr)
-+#endif
++ /* Hidden register state */
++ uint64_t period[HPET_TIMER_NUM]; /* Last value written to comparator */
++};
+
- /*
- * This structure is used to hold the arguments that are used when loading
- * kernel binaries.
-@@ -91,6 +98,12 @@ struct kimage {
- extern NORET_TYPE void machine_kexec(struct kimage *image) ATTRIB_NORET;
- extern int machine_kexec_prepare(struct kimage *image);
- extern void machine_kexec_cleanup(struct kimage *image);
-+#ifdef CONFIG_XEN
-+extern int xen_machine_kexec_load(struct kimage *image);
-+extern void xen_machine_kexec_unload(struct kimage *image);
-+extern void xen_machine_kexec_setup_resources(void);
-+extern void xen_machine_kexec_register_resources(struct resource *res);
-+#endif
- extern asmlinkage long sys_kexec_load(unsigned long entry,
- unsigned long nr_segments,
- struct kexec_segment __user *segments,
-diff -Nurp pristine-linux-2.6.18.2/include/linux/mm.h tmp-linux-2.6-xen.patch/include/linux/mm.h
---- pristine-linux-2.6.18.2/include/linux/mm.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/mm.h 2007-07-30 16:35:13.000000000 +0200
-@@ -164,6 +164,9 @@ extern unsigned int kobjsize(const void
- #define VM_NONLINEAR 0x00800000 /* Is non-linear (remap_file_pages) */
- #define VM_MAPPED_COPY 0x01000000 /* T if mapped copy of data (nommu mmap) */
- #define VM_INSERTPAGE 0x02000000 /* The vma has had "vm_insert_page()" done on it */
-+#ifdef CONFIG_XEN
-+#define VM_FOREIGN 0x04000000 /* Has pages belonging to another VM */
-+#endif
-
- #ifndef VM_STACK_DEFAULT_FLAGS /* arch can override this */
- #define VM_STACK_DEFAULT_FLAGS VM_DATA_DEFAULT_FLAGS
-@@ -202,6 +205,10 @@ struct vm_operations_struct {
- /* notification that a previously read-only page is about to become
- * writable, if an error is returned it will cause a SIGBUS */
- int (*page_mkwrite)(struct vm_area_struct *vma, struct page *page);
-+ /* Area-specific function for clearing the PTE at @ptep. Returns the
-+ * original value of @ptep. */
-+ pte_t (*zap_pte)(struct vm_area_struct *vma,
-+ unsigned long addr, pte_t *ptep, int is_fullmm);
- #ifdef CONFIG_NUMA
- int (*set_policy)(struct vm_area_struct *vma, struct mempolicy *new);
- struct mempolicy *(*get_policy)(struct vm_area_struct *vma,
-@@ -1027,6 +1034,13 @@ struct page *follow_page(struct vm_area_
- #define FOLL_GET 0x04 /* do get_page on page */
- #define FOLL_ANON 0x08 /* give ZERO_PAGE if no pgtable */
-
-+#ifdef CONFIG_XEN
-+typedef int (*pte_fn_t)(pte_t *pte, struct page *pmd_page, unsigned long addr,
-+ void *data);
-+extern int apply_to_page_range(struct mm_struct *mm, unsigned long address,
-+ unsigned long size, pte_fn_t fn, void *data);
-+#endif
++DECLARE_HVM_SAVE_TYPE(HPET, 12, struct hvm_hw_hpet);
+
- #ifdef CONFIG_PROC_FS
- void vm_stat_account(struct mm_struct *, unsigned long, struct file *, long);
- #else
-diff -Nurp pristine-linux-2.6.18.2/include/linux/oprofile.h tmp-linux-2.6-xen.patch/include/linux/oprofile.h
---- pristine-linux-2.6.18.2/include/linux/oprofile.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/oprofile.h 2007-10-14 01:51:15.000000000 +0200
-@@ -16,6 +16,8 @@
- #include <linux/types.h>
- #include <linux/spinlock.h>
- #include <asm/atomic.h>
+
-+#include <xen/interface/xenoprof.h>
-
- struct super_block;
- struct dentry;
-@@ -27,6 +29,11 @@ struct oprofile_operations {
- /* create any necessary configuration files in the oprofile fs.
- * Optional. */
- int (*create_files)(struct super_block * sb, struct dentry * root);
-+ /* setup active domains with Xen */
-+ int (*set_active)(int *active_domains, unsigned int adomains);
-+ /* setup passive domains with Xen */
-+ int (*set_passive)(int *passive_domains, unsigned int pdomains);
-+
- /* Do any necessary interrupt setup. Optional. */
- int (*setup)(void);
- /* Do any necessary interrupt shutdown. Optional. */
-@@ -78,6 +85,8 @@ void oprofile_add_pc(unsigned long pc, i
- /* add a backtrace entry, to be called from the ->backtrace callback */
- void oprofile_add_trace(unsigned long eip);
-
-+/* add a domain switch entry */
-+int oprofile_add_domain_switch(int32_t domain_id);
-
- /**
- * Create a file of the given name as a child of the given root, with
-diff -Nurp pristine-linux-2.6.18.2/include/linux/page-flags.h tmp-linux-2.6-xen.patch/include/linux/page-flags.h
---- pristine-linux-2.6.18.2/include/linux/page-flags.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/page-flags.h 2007-07-30 16:35:13.000000000 +0200
-@@ -98,6 +98,8 @@
- #define PG_uncached 31 /* Page has been mapped as uncached */
- #endif
-
-+#define PG_foreign 20 /* Page is owned by foreign allocator. */
++/*
++ * PM timer
++ */
+
- /*
- * Manipulation of page state flags
- */
-@@ -247,6 +249,18 @@
- #define SetPageUncached(page) set_bit(PG_uncached, &(page)->flags)
- #define ClearPageUncached(page) clear_bit(PG_uncached, &(page)->flags)
-
-+#define PageForeign(page) test_bit(PG_foreign, &(page)->flags)
-+#define SetPageForeign(page, dtor) do { \
-+ set_bit(PG_foreign, &(page)->flags); \
-+ (page)->index = (long)(dtor); \
-+} while (0)
-+#define ClearPageForeign(page) do { \
-+ clear_bit(PG_foreign, &(page)->flags); \
-+ (page)->index = 0; \
-+} while (0)
-+#define PageForeignDestructor(page) \
-+ ( (void (*) (struct page *)) (page)->index )(page)
++struct hvm_hw_pmtimer {
++ uint32_t tmr_val; /* PM_TMR_BLK.TMR_VAL: 32bit free-running counter */
++ uint16_t pm1a_sts; /* PM1a_EVT_BLK.PM1a_STS: status register */
++ uint16_t pm1a_en; /* PM1a_EVT_BLK.PM1a_EN: enable register */
++};
+
- struct page; /* forward declaration */
-
- int test_clear_page_dirty(struct page *page);
-diff -Nurp pristine-linux-2.6.18.2/include/linux/sched.h tmp-linux-2.6-xen.patch/include/linux/sched.h
---- pristine-linux-2.6.18.2/include/linux/sched.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/sched.h 2007-10-14 01:51:15.000000000 +0200
-@@ -211,10 +211,15 @@ extern void update_process_times(int use
- extern void scheduler_tick(void);
-
- #ifdef CONFIG_DETECT_SOFTLOCKUP
-+extern unsigned long softlockup_get_next_event(void);
- extern void softlockup_tick(void);
- extern void spawn_softlockup_task(void);
- extern void touch_softlockup_watchdog(void);
- #else
-+static inline unsigned long softlockup_get_next_event(void)
-+{
-+ return MAX_JIFFY_OFFSET;
-+}
- static inline void softlockup_tick(void)
- {
- }
-diff -Nurp pristine-linux-2.6.18.2/include/linux/skbuff.h tmp-linux-2.6-xen.patch/include/linux/skbuff.h
---- pristine-linux-2.6.18.2/include/linux/skbuff.h 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/linux/skbuff.h 2007-07-30 16:35:13.000000000 +0200
-@@ -203,6 +203,8 @@ enum {
- * @local_df: allow local fragmentation
- * @cloned: Head may be cloned (check refcnt to be sure)
- * @nohdr: Payload reference only, must not modify header
-+ * @proto_data_valid: Protocol data validated since arriving at localhost
-+ * @proto_csum_blank: Protocol csum must be added before leaving localhost
- * @pkt_type: Packet class
- * @fclone: skbuff clone status
- * @ip_summed: Driver fed us an IP checksum
-@@ -282,7 +284,13 @@ struct sk_buff {
- nfctinfo:3;
- __u8 pkt_type:3,
- fclone:2,
-+#ifndef CONFIG_XEN
- ipvs_property:1;
-+#else
-+ ipvs_property:1,
-+ proto_data_valid:1,
-+ proto_csum_blank:1;
-+#endif
- __be16 protocol;
-
- void (*destructor)(struct sk_buff *skb);
-diff -Nurp pristine-linux-2.6.18.2/include/xen/balloon.h tmp-linux-2.6-xen.patch/include/xen/balloon.h
---- pristine-linux-2.6.18.2/include/xen/balloon.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/balloon.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,57 @@
++DECLARE_HVM_SAVE_TYPE(PMTIMER, 13, struct hvm_hw_pmtimer);
++
++/*
++ * MTRR MSRs
++ */
++
++struct hvm_hw_mtrr {
++#define MTRR_VCNT 8
++#define NUM_FIXED_MSR 11
++ uint64_t msr_pat_cr;
++ /* mtrr physbase & physmask msr pair*/
++ uint64_t msr_mtrr_var[MTRR_VCNT*2];
++ uint64_t msr_mtrr_fixed[NUM_FIXED_MSR];
++ uint64_t msr_mtrr_cap;
++ uint64_t msr_mtrr_def_type;
++};
++
++DECLARE_HVM_SAVE_TYPE(MTRR, 14, struct hvm_hw_mtrr);
++
++/*
++ * Largest type-code in use
++ */
++#define HVM_SAVE_CODE_MAX 14
++
++#endif /* __XEN_PUBLIC_HVM_SAVE_X86_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-x86/xen-mca.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-x86/xen-mca.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,279 @@
+/******************************************************************************
-+ * balloon.h
-+ *
-+ * Xen balloon driver - enables returning/claiming memory to/from Xen.
-+ *
-+ * Copyright (c) 2003, B Dragovic
-+ * Copyright (c) 2003-2004, M Williamson, K Fraser
++ * arch-x86/mca.h
+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
++ * Contributed by Advanced Micro Devices, Inc.
++ * Author: Christoph Egger <Christoph.Egger at amd.com>
++ *
++ * Guest OS machine check interface to x86 Xen.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
-+ *
++ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
+ */
+
-+#ifndef __ASM_BALLOON_H__
-+#define __ASM_BALLOON_H__
++/* Full MCA functionality has the following Usecases from the guest side:
++ *
++ * Must have's:
++ * 1. Dom0 and DomU register machine check trap callback handlers
++ * (already done via "set_trap_table" hypercall)
++ * 2. Dom0 registers machine check event callback handler
++ * (doable via EVTCHNOP_bind_virq)
++ * 3. Dom0 and DomU fetches machine check data
++ * 4. Dom0 wants Xen to notify a DomU
++ * 5. Dom0 gets DomU ID from physical address
++ * 6. Dom0 wants Xen to kill DomU (already done for "xm destroy")
++ *
++ * Nice to have's:
++ * 7. Dom0 wants Xen to deactivate a physical CPU
++ * This is better done as separate task, physical CPU hotplugging,
++ * and hypercall(s) should be sysctl's
++ * 8. Page migration proposed from Xen NUMA work, where Dom0 can tell Xen to
++ * move a DomU (or Dom0 itself) away from a malicious page
++ * producing correctable errors.
++ * 9. offlining physical page:
++ * Xen free's and never re-uses a certain physical page.
++ * 10. Testfacility: Allow Dom0 to write values into machine check MSR's
++ * and tell Xen to trigger a machine check
++ */
++
++#ifndef __XEN_PUBLIC_ARCH_X86_MCA_H__
++#define __XEN_PUBLIC_ARCH_X86_MCA_H__
++
++/* Hypercall */
++#define __HYPERVISOR_mca __HYPERVISOR_arch_0
++
++#define XEN_MCA_INTERFACE_VERSION 0x03000001
++
++/* IN: Dom0 calls hypercall from MC event handler. */
++#define XEN_MC_CORRECTABLE 0x0
++/* IN: Dom0/DomU calls hypercall from MC trap handler. */
++#define XEN_MC_TRAP 0x1
++/* XEN_MC_CORRECTABLE and XEN_MC_TRAP are mutually exclusive. */
++
++/* OUT: All is ok */
++#define XEN_MC_OK 0x0
++/* OUT: Domain could not fetch data. */
++#define XEN_MC_FETCHFAILED 0x1
++/* OUT: There was no machine check data to fetch. */
++#define XEN_MC_NODATA 0x2
++/* OUT: Between notification time and this hypercall an other
++ * (most likely) correctable error happened. The fetched data,
++ * does not match the original machine check data. */
++#define XEN_MC_NOMATCH 0x4
++
++/* OUT: DomU did not register MC NMI handler. Try something else. */
++#define XEN_MC_CANNOTHANDLE 0x8
++/* OUT: Notifying DomU failed. Retry later or try something else. */
++#define XEN_MC_NOTDELIVERED 0x10
++/* Note, XEN_MC_CANNOTHANDLE and XEN_MC_NOTDELIVERED are mutually exclusive. */
+
-+/*
-+ * Inform the balloon driver that it should allow some slop for device-driver
-+ * memory activities.
-+ */
-+void balloon_update_driver_allowance(long delta);
+
-+/* Allocate/free a set of empty pages in low memory (i.e., no RAM mapped). */
-+struct page **alloc_empty_pages_and_pagevec(int nr_pages);
-+void free_empty_pages_and_pagevec(struct page **pagevec, int nr_pages);
++#ifndef __ASSEMBLY__
+
-+void balloon_release_driver_page(struct page *page);
++#define VIRQ_MCA VIRQ_ARCH_0 /* G. (DOM0) Machine Check Architecture */
+
+/*
-+ * Prevent the balloon driver from changing the memory reservation during
-+ * a driver critical region.
++ * Machine Check Architecure:
++ * structs are read-only and used to report all kinds of
++ * correctable and uncorrectable errors detected by the HW.
++ * Dom0 and DomU: register a handler to get notified.
++ * Dom0 only: Correctable errors are reported via VIRQ_MCA
++ * Dom0 and DomU: Uncorrectable errors are reported via nmi handlers
+ */
-+extern spinlock_t balloon_lock;
-+#define balloon_lock(__flags) spin_lock_irqsave(&balloon_lock, __flags)
-+#define balloon_unlock(__flags) spin_unlock_irqrestore(&balloon_lock, __flags)
++#define MC_TYPE_GLOBAL 0
++#define MC_TYPE_BANK 1
++#define MC_TYPE_EXTENDED 2
+
-+#endif /* __ASM_BALLOON_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/blkif.h tmp-linux-2.6-xen.patch/include/xen/blkif.h
---- pristine-linux-2.6.18.2/include/xen/blkif.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/blkif.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,97 @@
-+#ifndef __XEN_BLKIF_H__
-+#define __XEN_BLKIF_H__
++struct mcinfo_common {
++ uint16_t type; /* structure type */
++ uint16_t size; /* size of this struct in bytes */
++};
+
-+#include <xen/interface/io/ring.h>
-+#include <xen/interface/io/blkif.h>
-+#include <xen/interface/io/protocols.h>
+
-+/* Not a real protocol. Used to generate ring structs which contain
-+ * the elements common to all protocols only. This way we get a
-+ * compiler-checkable way to use common struct elements, so we can
-+ * avoid using switch(protocol) in a number of places. */
-+struct blkif_common_request {
-+ char dummy;
++#define MC_FLAG_CORRECTABLE (1 << 0)
++#define MC_FLAG_UNCORRECTABLE (1 << 1)
++
++/* contains global x86 mc information */
++struct mcinfo_global {
++ struct mcinfo_common common;
++
++ /* running domain at the time in error (most likely the impacted one) */
++ uint16_t mc_domid;
++ uint32_t mc_socketid; /* physical socket of the physical core */
++ uint16_t mc_coreid; /* physical impacted core */
++ uint16_t mc_core_threadid; /* core thread of physical core */
++ uint16_t mc_vcpuid; /* virtual cpu scheduled for mc_domid */
++ uint64_t mc_gstatus; /* global status */
++ uint32_t mc_flags;
+};
-+struct blkif_common_response {
-+ char dummy;
++
++/* contains bank local x86 mc information */
++struct mcinfo_bank {
++ struct mcinfo_common common;
++
++ uint16_t mc_bank; /* bank nr */
++ uint16_t mc_domid; /* Usecase 5: domain referenced by mc_addr on dom0
++ * and if mc_addr is valid. Never valid on DomU. */
++ uint64_t mc_status; /* bank status */
++ uint64_t mc_addr; /* bank address, only valid
++ * if addr bit is set in mc_status */
++ uint64_t mc_misc;
+};
+
-+/* i386 protocol version */
-+#pragma pack(push, 4)
-+struct blkif_x86_32_request {
-+ uint8_t operation; /* BLKIF_OP_??? */
-+ uint8_t nr_segments; /* number of segments */
-+ blkif_vdev_t handle; /* only for read/write requests */
-+ uint64_t id; /* private guest value, echoed in resp */
-+ blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
-+ struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++
++struct mcinfo_msr {
++ uint64_t reg; /* MSR */
++ uint64_t value; /* MSR value */
+};
-+struct blkif_x86_32_response {
-+ uint64_t id; /* copied from request */
-+ uint8_t operation; /* copied from request */
-+ int16_t status; /* BLKIF_RSP_??? */
++
++/* contains mc information from other
++ * or additional mc MSRs */
++struct mcinfo_extended {
++ struct mcinfo_common common;
++
++ /* You can fill up to five registers.
++ * If you need more, then use this structure
++ * multiple times. */
++
++ uint32_t mc_msrs; /* Number of msr with valid values. */
++ struct mcinfo_msr mc_msr[5];
+};
-+typedef struct blkif_x86_32_request blkif_x86_32_request_t;
-+typedef struct blkif_x86_32_response blkif_x86_32_response_t;
-+#pragma pack(pop)
+
-+/* x86_64 protocol version */
-+struct blkif_x86_64_request {
-+ uint8_t operation; /* BLKIF_OP_??? */
-+ uint8_t nr_segments; /* number of segments */
-+ blkif_vdev_t handle; /* only for read/write requests */
-+ uint64_t __attribute__((__aligned__(8))) id;
-+ blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
-+ struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++#define MCINFO_HYPERCALLSIZE 1024
++#define MCINFO_MAXSIZE 768
++
++struct mc_info {
++ /* Number of mcinfo_* entries in mi_data */
++ uint32_t mi_nentries;
++
++ uint8_t mi_data[MCINFO_MAXSIZE - sizeof(uint32_t)];
+};
-+struct blkif_x86_64_response {
-+ uint64_t __attribute__((__aligned__(8))) id;
-+ uint8_t operation; /* copied from request */
-+ int16_t status; /* BLKIF_RSP_??? */
++typedef struct mc_info mc_info_t;
++
++
++
++/*
++ * OS's should use these instead of writing their own lookup function
++ * each with its own bugs and drawbacks.
++ * We use macros instead of static inline functions to allow guests
++ * to include this header in assembly files (*.S).
++ */
++/* Prototype:
++ * uint32_t x86_mcinfo_nentries(struct mc_info *mi);
++ */
++#define x86_mcinfo_nentries(_mi) \
++ (_mi)->mi_nentries
++/* Prototype:
++ * struct mcinfo_common *x86_mcinfo_first(struct mc_info *mi);
++ */
++#define x86_mcinfo_first(_mi) \
++ (struct mcinfo_common *)((_mi)->mi_data)
++/* Prototype:
++ * struct mcinfo_common *x86_mcinfo_next(struct mcinfo_common *mic);
++ */
++#define x86_mcinfo_next(_mic) \
++ (struct mcinfo_common *)((uint8_t *)(_mic) + (_mic)->size)
++
++/* Prototype:
++ * void x86_mcinfo_lookup(void *ret, struct mc_info *mi, uint16_t type);
++ */
++#define x86_mcinfo_lookup(_ret, _mi, _type) \
++ do { \
++ uint32_t found, i; \
++ struct mcinfo_common *_mic; \
++ \
++ found = 0; \
++ (_ret) = NULL; \
++ if (_mi == NULL) break; \
++ _mic = x86_mcinfo_first(_mi); \
++ for (i = 0; i < x86_mcinfo_nentries(_mi); i++) { \
++ if (_mic->type == (_type)) { \
++ found = 1; \
++ break; \
++ } \
++ _mic = x86_mcinfo_next(_mic); \
++ } \
++ (_ret) = found ? _mic : NULL; \
++ } while (0)
++
++
++/* Usecase 1
++ * Register machine check trap callback handler
++ * (already done via "set_trap_table" hypercall)
++ */
++
++/* Usecase 2
++ * Dom0 registers machine check event callback handler
++ * done by EVTCHNOP_bind_virq
++ */
++
++/* Usecase 3
++ * Fetch machine check data from hypervisor.
++ * Note, this hypercall is special, because both Dom0 and DomU must use this.
++ */
++#define XEN_MC_fetch 1
++struct xen_mc_fetch {
++ /* IN/OUT variables. */
++ uint32_t flags;
++
++/* IN: XEN_MC_CORRECTABLE, XEN_MC_TRAP */
++/* OUT: XEN_MC_OK, XEN_MC_FETCHFAILED, XEN_MC_NODATA, XEN_MC_NOMATCH */
++
++ /* OUT variables. */
++ uint32_t fetch_idx; /* only useful for Dom0 for the notify hypercall */
++ struct mc_info mc_info;
+};
-+typedef struct blkif_x86_64_request blkif_x86_64_request_t;
-+typedef struct blkif_x86_64_response blkif_x86_64_response_t;
++typedef struct xen_mc_fetch xen_mc_fetch_t;
++DEFINE_XEN_GUEST_HANDLE(xen_mc_fetch_t);
+
-+DEFINE_RING_TYPES(blkif_common, struct blkif_common_request, struct blkif_common_response);
-+DEFINE_RING_TYPES(blkif_x86_32, struct blkif_x86_32_request, struct blkif_x86_32_response);
-+DEFINE_RING_TYPES(blkif_x86_64, struct blkif_x86_64_request, struct blkif_x86_64_response);
+
-+union blkif_back_rings {
-+ blkif_back_ring_t native;
-+ blkif_common_back_ring_t common;
-+ blkif_x86_32_back_ring_t x86_32;
-+ blkif_x86_64_back_ring_t x86_64;
++/* Usecase 4
++ * This tells the hypervisor to notify a DomU about the machine check error
++ */
++#define XEN_MC_notifydomain 2
++struct xen_mc_notifydomain {
++ /* IN variables. */
++ uint16_t mc_domid; /* The unprivileged domain to notify. */
++ uint16_t mc_vcpuid; /* The vcpu in mc_domid to notify.
++ * Usually echo'd value from the fetch hypercall. */
++ uint32_t fetch_idx; /* echo'd value from the fetch hypercall. */
++
++ /* IN/OUT variables. */
++ uint32_t flags;
++
++/* IN: XEN_MC_CORRECTABLE, XEN_MC_TRAP */
++/* OUT: XEN_MC_OK, XEN_MC_CANNOTHANDLE, XEN_MC_NOTDELIVERED, XEN_MC_NOMATCH */
+};
-+typedef union blkif_back_rings blkif_back_rings_t;
++typedef struct xen_mc_notifydomain xen_mc_notifydomain_t;
++DEFINE_XEN_GUEST_HANDLE(xen_mc_notifydomain_t);
+
-+enum blkif_protocol {
-+ BLKIF_PROTOCOL_NATIVE = 1,
-+ BLKIF_PROTOCOL_X86_32 = 2,
-+ BLKIF_PROTOCOL_X86_64 = 3,
++
++struct xen_mc {
++ uint32_t cmd;
++ uint32_t interface_version; /* XEN_MCA_INTERFACE_VERSION */
++ union {
++ struct xen_mc_fetch mc_fetch;
++ struct xen_mc_notifydomain mc_notifydomain;
++ uint8_t pad[MCINFO_HYPERCALLSIZE];
++ } u;
+};
++typedef struct xen_mc xen_mc_t;
++DEFINE_XEN_GUEST_HANDLE(xen_mc_t);
+
-+static void inline blkif_get_x86_32_req(blkif_request_t *dst, blkif_x86_32_request_t *src)
-+{
-+ int i;
-+ dst->operation = src->operation;
-+ dst->nr_segments = src->nr_segments;
-+ dst->handle = src->handle;
-+ dst->id = src->id;
-+ dst->sector_number = src->sector_number;
-+ for (i = 0; i < src->nr_segments; i++)
-+ dst->seg[i] = src->seg[i];
-+}
++#endif /* __ASSEMBLY__ */
+
-+static void inline blkif_get_x86_64_req(blkif_request_t *dst, blkif_x86_64_request_t *src)
-+{
-+ int i;
-+ dst->operation = src->operation;
-+ dst->nr_segments = src->nr_segments;
-+ dst->handle = src->handle;
-+ dst->id = src->id;
-+ dst->sector_number = src->sector_number;
-+ for (i = 0; i < src->nr_segments; i++)
-+ dst->seg[i] = src->seg[i];
-+}
++#endif /* __XEN_PUBLIC_ARCH_X86_MCA_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-x86/xen-x86_32.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-x86/xen-x86_32.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,180 @@
++/******************************************************************************
++ * xen-x86_32.h
++ *
++ * Guest OS interface to x86 32-bit Xen.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2004-2007, K A Fraser
++ */
+
-+#endif /* __XEN_BLKIF_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/cpu_hotplug.h tmp-linux-2.6-xen.patch/include/xen/cpu_hotplug.h
---- pristine-linux-2.6.18.2/include/xen/cpu_hotplug.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/cpu_hotplug.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,41 @@
-+#ifndef __XEN_CPU_HOTPLUG_H__
-+#define __XEN_CPU_HOTPLUG_H__
++#ifndef __XEN_PUBLIC_ARCH_X86_XEN_X86_32_H__
++#define __XEN_PUBLIC_ARCH_X86_XEN_X86_32_H__
+
-+#include <linux/kernel.h>
-+#include <linux/cpumask.h>
++/*
++ * Hypercall interface:
++ * Input: %ebx, %ecx, %edx, %esi, %edi (arguments 1-5)
++ * Output: %eax
++ * Access is via hypercall page (set up by guest loader or via a Xen MSR):
++ * call hypercall_page + hypercall-number * 32
++ * Clobbered: Argument registers (e.g., 2-arg hypercall clobbers %ebx,%ecx)
++ */
+
-+#if defined(CONFIG_X86) && defined(CONFIG_SMP)
-+extern cpumask_t cpu_initialized_map;
++#if __XEN_INTERFACE_VERSION__ < 0x00030203
++/*
++ * Legacy hypercall interface:
++ * As above, except the entry sequence to the hypervisor is:
++ * mov $hypercall-number*32,%eax ; int $0x82
++ */
++#define TRAP_INSTR "int $0x82"
+#endif
+
-+#if defined(CONFIG_HOTPLUG_CPU)
++/*
++ * These flat segments are in the Xen-private section of every GDT. Since these
++ * are also present in the initial GDT, many OSes will be able to avoid
++ * installing their own GDT.
++ */
++#define FLAT_RING1_CS 0xe019 /* GDT index 259 */
++#define FLAT_RING1_DS 0xe021 /* GDT index 260 */
++#define FLAT_RING1_SS 0xe021 /* GDT index 260 */
++#define FLAT_RING3_CS 0xe02b /* GDT index 261 */
++#define FLAT_RING3_DS 0xe033 /* GDT index 262 */
++#define FLAT_RING3_SS 0xe033 /* GDT index 262 */
+
-+int cpu_up_check(unsigned int cpu);
-+void init_xenbus_allowed_cpumask(void);
-+int smp_suspend(void);
-+void smp_resume(void);
++#define FLAT_KERNEL_CS FLAT_RING1_CS
++#define FLAT_KERNEL_DS FLAT_RING1_DS
++#define FLAT_KERNEL_SS FLAT_RING1_SS
++#define FLAT_USER_CS FLAT_RING3_CS
++#define FLAT_USER_DS FLAT_RING3_DS
++#define FLAT_USER_SS FLAT_RING3_SS
+
-+void cpu_bringup(void);
++#define __HYPERVISOR_VIRT_START_PAE 0xF5800000
++#define __MACH2PHYS_VIRT_START_PAE 0xF5800000
++#define __MACH2PHYS_VIRT_END_PAE 0xF6800000
++#define HYPERVISOR_VIRT_START_PAE \
++ mk_unsigned_long(__HYPERVISOR_VIRT_START_PAE)
++#define MACH2PHYS_VIRT_START_PAE \
++ mk_unsigned_long(__MACH2PHYS_VIRT_START_PAE)
++#define MACH2PHYS_VIRT_END_PAE \
++ mk_unsigned_long(__MACH2PHYS_VIRT_END_PAE)
+
-+#else /* !defined(CONFIG_HOTPLUG_CPU) */
++/* Non-PAE bounds are obsolete. */
++#define __HYPERVISOR_VIRT_START_NONPAE 0xFC000000
++#define __MACH2PHYS_VIRT_START_NONPAE 0xFC000000
++#define __MACH2PHYS_VIRT_END_NONPAE 0xFC400000
++#define HYPERVISOR_VIRT_START_NONPAE \
++ mk_unsigned_long(__HYPERVISOR_VIRT_START_NONPAE)
++#define MACH2PHYS_VIRT_START_NONPAE \
++ mk_unsigned_long(__MACH2PHYS_VIRT_START_NONPAE)
++#define MACH2PHYS_VIRT_END_NONPAE \
++ mk_unsigned_long(__MACH2PHYS_VIRT_END_NONPAE)
+
-+#define cpu_up_check(cpu) (0)
-+#define init_xenbus_allowed_cpumask() ((void)0)
++#define __HYPERVISOR_VIRT_START __HYPERVISOR_VIRT_START_PAE
++#define __MACH2PHYS_VIRT_START __MACH2PHYS_VIRT_START_PAE
++#define __MACH2PHYS_VIRT_END __MACH2PHYS_VIRT_END_PAE
+
-+static inline int smp_suspend(void)
-+{
-+ if (num_online_cpus() > 1) {
-+ printk(KERN_WARNING "Can't suspend SMP guests "
-+ "without CONFIG_HOTPLUG_CPU\n");
-+ return -EOPNOTSUPP;
-+ }
-+ return 0;
-+}
++#ifndef HYPERVISOR_VIRT_START
++#define HYPERVISOR_VIRT_START mk_unsigned_long(__HYPERVISOR_VIRT_START)
++#endif
+
-+static inline void smp_resume(void)
-+{
-+}
++#define MACH2PHYS_VIRT_START mk_unsigned_long(__MACH2PHYS_VIRT_START)
++#define MACH2PHYS_VIRT_END mk_unsigned_long(__MACH2PHYS_VIRT_END)
++#define MACH2PHYS_NR_ENTRIES ((MACH2PHYS_VIRT_END-MACH2PHYS_VIRT_START)>>2)
++#ifndef machine_to_phys_mapping
++#define machine_to_phys_mapping ((unsigned long *)MACH2PHYS_VIRT_START)
++#endif
+
-+#endif /* !defined(CONFIG_HOTPLUG_CPU) */
++/* 32-/64-bit invariability for control interfaces (domctl/sysctl). */
++#if defined(__XEN__) || defined(__XEN_TOOLS__)
++#undef ___DEFINE_XEN_GUEST_HANDLE
++#define ___DEFINE_XEN_GUEST_HANDLE(name, type) \
++ typedef struct { type *p; } \
++ __guest_handle_ ## name; \
++ typedef struct { union { type *p; uint64_aligned_t q; }; } \
++ __guest_handle_64_ ## name
++#undef set_xen_guest_handle
++#define set_xen_guest_handle(hnd, val) \
++ do { if ( sizeof(hnd) == 8 ) *(uint64_t *)&(hnd) = 0; \
++ (hnd).p = val; \
++ } while ( 0 )
++#define uint64_aligned_t uint64_t __attribute__((aligned(8)))
++#define __XEN_GUEST_HANDLE_64(name) __guest_handle_64_ ## name
++#define XEN_GUEST_HANDLE_64(name) __XEN_GUEST_HANDLE_64(name)
++#endif
+
-+#endif /* __XEN_CPU_HOTPLUG_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/driver_util.h tmp-linux-2.6-xen.patch/include/xen/driver_util.h
---- pristine-linux-2.6.18.2/include/xen/driver_util.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/driver_util.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,14 @@
++#ifndef __ASSEMBLY__
+
-+#ifndef __ASM_XEN_DRIVER_UTIL_H__
-+#define __ASM_XEN_DRIVER_UTIL_H__
++struct cpu_user_regs {
++ uint32_t ebx;
++ uint32_t ecx;
++ uint32_t edx;
++ uint32_t esi;
++ uint32_t edi;
++ uint32_t ebp;
++ uint32_t eax;
++ uint16_t error_code; /* private */
++ uint16_t entry_vector; /* private */
++ uint32_t eip;
++ uint16_t cs;
++ uint8_t saved_upcall_mask;
++ uint8_t _pad0;
++ uint32_t eflags; /* eflags.IF == !saved_upcall_mask */
++ uint32_t esp;
++ uint16_t ss, _pad1;
++ uint16_t es, _pad2;
++ uint16_t ds, _pad3;
++ uint16_t fs, _pad4;
++ uint16_t gs, _pad5;
++};
++typedef struct cpu_user_regs cpu_user_regs_t;
++DEFINE_XEN_GUEST_HANDLE(cpu_user_regs_t);
++
++/*
++ * Page-directory addresses above 4GB do not fit into architectural %cr3.
++ * When accessing %cr3, or equivalent field in vcpu_guest_context, guests
++ * must use the following accessor macros to pack/unpack valid MFNs.
++ */
++#define xen_pfn_to_cr3(pfn) (((unsigned)(pfn) << 12) | ((unsigned)(pfn) >> 20))
++#define xen_cr3_to_pfn(cr3) (((unsigned)(cr3) >> 12) | ((unsigned)(cr3) << 20))
++
++struct arch_vcpu_info {
++ unsigned long cr2;
++ unsigned long pad[5]; /* sizeof(vcpu_info_t) == 64 */
++};
++typedef struct arch_vcpu_info arch_vcpu_info_t;
+
-+#include <linux/vmalloc.h>
-+#include <linux/device.h>
++struct xen_callback {
++ unsigned long cs;
++ unsigned long eip;
++};
++typedef struct xen_callback xen_callback_t;
+
-+/* Allocate/destroy a 'vmalloc' VM area. */
-+extern struct vm_struct *alloc_vm_area(unsigned long size);
-+extern void free_vm_area(struct vm_struct *area);
++#endif /* !__ASSEMBLY__ */
+
-+extern struct class *get_xen_class(void);
++#endif /* __XEN_PUBLIC_ARCH_X86_XEN_X86_32_H__ */
+
-+#endif /* __ASM_XEN_DRIVER_UTIL_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/evtchn.h tmp-linux-2.6-xen.patch/include/xen/evtchn.h
---- pristine-linux-2.6.18.2/include/xen/evtchn.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/evtchn.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,126 @@
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-x86/xen-x86_64.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-x86/xen-x86_64.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,212 @@
+/******************************************************************************
-+ * evtchn.h
-+ *
-+ * Communication via Xen event channels.
-+ * Also definitions for the device that demuxes notifications to userspace.
-+ *
-+ * Copyright (c) 2004-2005, K A Fraser
++ * xen-x86_64.h
+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
++ * Guest OS interface to x86 64-bit Xen.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
-+ *
++ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2004-2006, K A Fraser
+ */
+
-+#ifndef __ASM_EVTCHN_H__
-+#define __ASM_EVTCHN_H__
-+
-+#include <linux/interrupt.h>
-+#include <asm/hypervisor.h>
-+#include <asm/ptrace.h>
-+#include <asm/synch_bitops.h>
-+#include <xen/interface/event_channel.h>
-+#include <linux/smp.h>
++#ifndef __XEN_PUBLIC_ARCH_X86_XEN_X86_64_H__
++#define __XEN_PUBLIC_ARCH_X86_XEN_X86_64_H__
+
+/*
-+ * LOW-LEVEL DEFINITIONS
++ * Hypercall interface:
++ * Input: %rdi, %rsi, %rdx, %r10, %r8 (arguments 1-5)
++ * Output: %rax
++ * Access is via hypercall page (set up by guest loader or via a Xen MSR):
++ * call hypercall_page + hypercall-number * 32
++ * Clobbered: argument registers (e.g., 2-arg hypercall clobbers %rdi,%rsi)
+ */
+
++#if __XEN_INTERFACE_VERSION__ < 0x00030203
+/*
-+ * Dynamically bind an event source to an IRQ-like callback handler.
-+ * On some platforms this may not be implemented via the Linux IRQ subsystem.
-+ * The IRQ argument passed to the callback handler is the same as returned
-+ * from the bind call. It may not correspond to a Linux IRQ number.
-+ * Returns IRQ or negative errno.
++ * Legacy hypercall interface:
++ * As above, except the entry sequence to the hypervisor is:
++ * mov $hypercall-number*32,%eax ; syscall
++ * Clobbered: %rcx, %r11, argument registers (as above)
+ */
-+int bind_caller_port_to_irqhandler(
-+ unsigned int caller_port,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
-+ unsigned long irqflags,
-+ const char *devname,
-+ void *dev_id);
-+int bind_listening_port_to_irqhandler(
-+ unsigned int remote_domain,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
-+ unsigned long irqflags,
-+ const char *devname,
-+ void *dev_id);
-+int bind_interdomain_evtchn_to_irqhandler(
-+ unsigned int remote_domain,
-+ unsigned int remote_port,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
-+ unsigned long irqflags,
-+ const char *devname,
-+ void *dev_id);
-+int bind_virq_to_irqhandler(
-+ unsigned int virq,
-+ unsigned int cpu,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
-+ unsigned long irqflags,
-+ const char *devname,
-+ void *dev_id);
-+int bind_ipi_to_irqhandler(
-+ unsigned int ipi,
-+ unsigned int cpu,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
-+ unsigned long irqflags,
-+ const char *devname,
-+ void *dev_id);
++#define TRAP_INSTR "syscall"
++#endif
+
+/*
-+ * Common unbind function for all event sources. Takes IRQ to unbind from.
-+ * Automatically closes the underlying event channel (except for bindings
-+ * made with bind_caller_port_to_irqhandler()).
++ * 64-bit segment selectors
++ * These flat segments are in the Xen-private section of every GDT. Since these
++ * are also present in the initial GDT, many OSes will be able to avoid
++ * installing their own GDT.
+ */
-+void unbind_from_irqhandler(unsigned int irq, void *dev_id);
+
-+void irq_resume(void);
++#define FLAT_RING3_CS32 0xe023 /* GDT index 260 */
++#define FLAT_RING3_CS64 0xe033 /* GDT index 261 */
++#define FLAT_RING3_DS32 0xe02b /* GDT index 262 */
++#define FLAT_RING3_DS64 0x0000 /* NULL selector */
++#define FLAT_RING3_SS32 0xe02b /* GDT index 262 */
++#define FLAT_RING3_SS64 0xe02b /* GDT index 262 */
+
-+/* Entry point for notifications into Linux subsystems. */
-+asmlinkage void evtchn_do_upcall(struct pt_regs *regs);
++#define FLAT_KERNEL_DS64 FLAT_RING3_DS64
++#define FLAT_KERNEL_DS32 FLAT_RING3_DS32
++#define FLAT_KERNEL_DS FLAT_KERNEL_DS64
++#define FLAT_KERNEL_CS64 FLAT_RING3_CS64
++#define FLAT_KERNEL_CS32 FLAT_RING3_CS32
++#define FLAT_KERNEL_CS FLAT_KERNEL_CS64
++#define FLAT_KERNEL_SS64 FLAT_RING3_SS64
++#define FLAT_KERNEL_SS32 FLAT_RING3_SS32
++#define FLAT_KERNEL_SS FLAT_KERNEL_SS64
+
-+/* Entry point for notifications into the userland character device. */
-+void evtchn_device_upcall(int port);
++#define FLAT_USER_DS64 FLAT_RING3_DS64
++#define FLAT_USER_DS32 FLAT_RING3_DS32
++#define FLAT_USER_DS FLAT_USER_DS64
++#define FLAT_USER_CS64 FLAT_RING3_CS64
++#define FLAT_USER_CS32 FLAT_RING3_CS32
++#define FLAT_USER_CS FLAT_USER_CS64
++#define FLAT_USER_SS64 FLAT_RING3_SS64
++#define FLAT_USER_SS32 FLAT_RING3_SS32
++#define FLAT_USER_SS FLAT_USER_SS64
+
-+void mask_evtchn(int port);
-+void unmask_evtchn(int port);
++#define __HYPERVISOR_VIRT_START 0xFFFF800000000000
++#define __HYPERVISOR_VIRT_END 0xFFFF880000000000
++#define __MACH2PHYS_VIRT_START 0xFFFF800000000000
++#define __MACH2PHYS_VIRT_END 0xFFFF804000000000
+
-+static inline void clear_evtchn(int port)
-+{
-+ shared_info_t *s = HYPERVISOR_shared_info;
-+ synch_clear_bit(port, s->evtchn_pending);
-+}
++#ifndef HYPERVISOR_VIRT_START
++#define HYPERVISOR_VIRT_START mk_unsigned_long(__HYPERVISOR_VIRT_START)
++#define HYPERVISOR_VIRT_END mk_unsigned_long(__HYPERVISOR_VIRT_END)
++#endif
+
-+static inline void notify_remote_via_evtchn(int port)
-+{
-+ struct evtchn_send send = { .port = port };
-+ (void)HYPERVISOR_event_channel_op(EVTCHNOP_send, &send);
-+}
++#define MACH2PHYS_VIRT_START mk_unsigned_long(__MACH2PHYS_VIRT_START)
++#define MACH2PHYS_VIRT_END mk_unsigned_long(__MACH2PHYS_VIRT_END)
++#define MACH2PHYS_NR_ENTRIES ((MACH2PHYS_VIRT_END-MACH2PHYS_VIRT_START)>>3)
++#ifndef machine_to_phys_mapping
++#define machine_to_phys_mapping ((unsigned long *)HYPERVISOR_VIRT_START)
++#endif
+
+/*
-+ * Use these to access the event channel underlying the IRQ handle returned
-+ * by bind_*_to_irqhandler().
++ * int HYPERVISOR_set_segment_base(unsigned int which, unsigned long base)
++ * @which == SEGBASE_* ; @base == 64-bit base address
++ * Returns 0 on success.
+ */
-+void notify_remote_via_irq(int irq);
-+int irq_to_evtchn_port(int irq);
++#define SEGBASE_FS 0
++#define SEGBASE_GS_USER 1
++#define SEGBASE_GS_KERNEL 2
++#define SEGBASE_GS_USER_SEL 3 /* Set user %gs specified in base[15:0] */
+
-+#endif /* __ASM_EVTCHN_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/features.h tmp-linux-2.6-xen.patch/include/xen/features.h
---- pristine-linux-2.6.18.2/include/xen/features.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/features.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,20 @@
-+/******************************************************************************
-+ * features.h
-+ *
-+ * Query the features reported by Xen.
-+ *
-+ * Copyright (c) 2006, Ian Campbell
++/*
++ * int HYPERVISOR_iret(void)
++ * All arguments are on the kernel stack, in the following format.
++ * Never returns if successful. Current kernel context is lost.
++ * The saved CS is mapped as follows:
++ * RING0 -> RING3 kernel mode.
++ * RING1 -> RING3 kernel mode.
++ * RING2 -> RING3 kernel mode.
++ * RING3 -> RING3 user mode.
++ * However RING0 indicates that the guest kernel should return to iteself
++ * directly with
++ * orb $3,1*8(%rsp)
++ * iretq
++ * If flags contains VGCF_in_syscall:
++ * Restore RAX, RIP, RFLAGS, RSP.
++ * Discard R11, RCX, CS, SS.
++ * Otherwise:
++ * Restore RAX, R11, RCX, CS:RIP, RFLAGS, SS:RSP.
++ * All other registers are saved on hypercall entry and restored to user.
+ */
++/* Guest exited in SYSCALL context? Return to guest with SYSRET? */
++#define _VGCF_in_syscall 8
++#define VGCF_in_syscall (1<<_VGCF_in_syscall)
++#define VGCF_IN_SYSCALL VGCF_in_syscall
+
-+#ifndef __ASM_XEN_FEATURES_H__
-+#define __ASM_XEN_FEATURES_H__
++#ifndef __ASSEMBLY__
+
-+#include <xen/interface/version.h>
++struct iret_context {
++ /* Top of stack (%rsp at point of hypercall). */
++ uint64_t rax, r11, rcx, flags, rip, cs, rflags, rsp, ss;
++ /* Bottom of iret stack frame. */
++};
+
-+extern void setup_xen_features(void);
++#if defined(__GNUC__) && !defined(__STRICT_ANSI__)
++/* Anonymous union includes both 32- and 64-bit names (e.g., eax/rax). */
++#define __DECL_REG(name) union { \
++ uint64_t r ## name, e ## name; \
++ uint32_t _e ## name; \
++}
++#else
++/* Non-gcc sources must always use the proper 64-bit name (e.g., rax). */
++#define __DECL_REG(name) uint64_t r ## name
++#endif
+
-+extern u8 xen_features[XENFEAT_NR_SUBMAPS * 32];
++struct cpu_user_regs {
++ uint64_t r15;
++ uint64_t r14;
++ uint64_t r13;
++ uint64_t r12;
++ __DECL_REG(bp);
++ __DECL_REG(bx);
++ uint64_t r11;
++ uint64_t r10;
++ uint64_t r9;
++ uint64_t r8;
++ __DECL_REG(ax);
++ __DECL_REG(cx);
++ __DECL_REG(dx);
++ __DECL_REG(si);
++ __DECL_REG(di);
++ uint32_t error_code; /* private */
++ uint32_t entry_vector; /* private */
++ __DECL_REG(ip);
++ uint16_t cs, _pad0[1];
++ uint8_t saved_upcall_mask;
++ uint8_t _pad1[3];
++ __DECL_REG(flags); /* rflags.IF == !saved_upcall_mask */
++ __DECL_REG(sp);
++ uint16_t ss, _pad2[3];
++ uint16_t es, _pad3[3];
++ uint16_t ds, _pad4[3];
++ uint16_t fs, _pad5[3]; /* Non-zero => takes precedence over fs_base. */
++ uint16_t gs, _pad6[3]; /* Non-zero => takes precedence over gs_base_usr. */
++};
++typedef struct cpu_user_regs cpu_user_regs_t;
++DEFINE_XEN_GUEST_HANDLE(cpu_user_regs_t);
+
-+#define xen_feature(flag) (xen_features[flag])
++#undef __DECL_REG
+
-+#endif /* __ASM_XEN_FEATURES_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/gnttab.h tmp-linux-2.6-xen.patch/include/xen/gnttab.h
---- pristine-linux-2.6.18.2/include/xen/gnttab.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/gnttab.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,138 @@
++#define xen_pfn_to_cr3(pfn) ((unsigned long)(pfn) << 12)
++#define xen_cr3_to_pfn(cr3) ((unsigned long)(cr3) >> 12)
++
++struct arch_vcpu_info {
++ unsigned long cr2;
++ unsigned long pad; /* sizeof(vcpu_info_t) == 64 */
++};
++typedef struct arch_vcpu_info arch_vcpu_info_t;
++
++typedef unsigned long xen_callback_t;
++
++#endif /* !__ASSEMBLY__ */
++
++#endif /* __XEN_PUBLIC_ARCH_X86_XEN_X86_64_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-x86/xen.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-x86/xen.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,204 @@
+/******************************************************************************
-+ * gnttab.h
-+ *
-+ * Two sets of functionality:
-+ * 1. Granting foreign access to our memory reservation.
-+ * 2. Accessing others' memory reservations via grant references.
-+ * (i.e., mechanisms for both sender and recipient of grant references)
-+ *
-+ * Copyright (c) 2004-2005, K A Fraser
-+ * Copyright (c) 2005, Christopher Clark
++ * arch-x86/xen.h
+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
++ * Guest OS interface to x86 Xen.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
-+ *
++ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2004-2006, K A Fraser
+ */
+
-+#ifndef __ASM_GNTTAB_H__
-+#define __ASM_GNTTAB_H__
++#include "../xen.h"
+
-+#include <asm/hypervisor.h>
-+#include <asm/maddr.h> /* maddr_t */
-+#include <xen/interface/grant_table.h>
-+#include <xen/features.h>
++#ifndef __XEN_PUBLIC_ARCH_X86_XEN_H__
++#define __XEN_PUBLIC_ARCH_X86_XEN_H__
+
-+struct gnttab_free_callback {
-+ struct gnttab_free_callback *next;
-+ void (*fn)(void *);
-+ void *arg;
-+ u16 count;
-+};
++/* Structural guest handles introduced in 0x00030201. */
++#if __XEN_INTERFACE_VERSION__ >= 0x00030201
++#define ___DEFINE_XEN_GUEST_HANDLE(name, type) \
++ typedef struct { type *p; } __guest_handle_ ## name
++#else
++#define ___DEFINE_XEN_GUEST_HANDLE(name, type) \
++ typedef type * __guest_handle_ ## name
++#endif
+
-+int gnttab_grant_foreign_access(domid_t domid, unsigned long frame,
-+ int readonly);
++#define __DEFINE_XEN_GUEST_HANDLE(name, type) \
++ ___DEFINE_XEN_GUEST_HANDLE(name, type); \
++ ___DEFINE_XEN_GUEST_HANDLE(const_##name, const type)
++#define DEFINE_XEN_GUEST_HANDLE(name) __DEFINE_XEN_GUEST_HANDLE(name, name)
++#define __XEN_GUEST_HANDLE(name) __guest_handle_ ## name
++#define XEN_GUEST_HANDLE(name) __XEN_GUEST_HANDLE(name)
++#define set_xen_guest_handle(hnd, val) do { (hnd).p = val; } while (0)
++#ifdef __XEN_TOOLS__
++#define get_xen_guest_handle(val, hnd) do { val = (hnd).p; } while (0)
++#endif
+
-+/*
-+ * End access through the given grant reference, iff the grant entry is no
-+ * longer in use. Return 1 if the grant entry was freed, 0 if it is still in
-+ * use.
-+ */
-+int gnttab_end_foreign_access_ref(grant_ref_t ref, int readonly);
++#if defined(__i386__)
++#include "xen-x86_32.h"
++#elif defined(__x86_64__)
++#include "xen-x86_64.h"
++#endif
++
++#ifndef __ASSEMBLY__
++typedef unsigned long xen_pfn_t;
++#define PRI_xen_pfn "lx"
++#endif
+
+/*
-+ * Eventually end access through the given grant reference, and once that
-+ * access has been ended, free the given page too. Access will be ended
-+ * immediately iff the grant entry is not in use, otherwise it will happen
-+ * some time later. page may be 0, in which case no freeing will occur.
++ * SEGMENT DESCRIPTOR TABLES
+ */
-+void gnttab_end_foreign_access(grant_ref_t ref, int readonly,
-+ unsigned long page);
-+
-+int gnttab_grant_foreign_transfer(domid_t domid, unsigned long pfn);
-+
-+unsigned long gnttab_end_foreign_transfer_ref(grant_ref_t ref);
-+unsigned long gnttab_end_foreign_transfer(grant_ref_t ref);
-+
-+int gnttab_query_foreign_access(grant_ref_t ref);
-+
+/*
-+ * operations on reserved batches of grant references
++ * A number of GDT entries are reserved by Xen. These are not situated at the
++ * start of the GDT because some stupid OSes export hard-coded selector values
++ * in their ABI. These hard-coded values are always near the start of the GDT,
++ * so Xen places itself out of the way, at the far end of the GDT.
+ */
-+int gnttab_alloc_grant_references(u16 count, grant_ref_t *pprivate_head);
-+
-+void gnttab_free_grant_reference(grant_ref_t ref);
-+
-+void gnttab_free_grant_references(grant_ref_t head);
-+
-+int gnttab_empty_grant_references(const grant_ref_t *pprivate_head);
-+
-+int gnttab_claim_grant_reference(grant_ref_t *pprivate_head);
-+
-+void gnttab_release_grant_reference(grant_ref_t *private_head,
-+ grant_ref_t release);
-+
-+void gnttab_request_free_callback(struct gnttab_free_callback *callback,
-+ void (*fn)(void *), void *arg, u16 count);
-+void gnttab_cancel_free_callback(struct gnttab_free_callback *callback);
-+
-+void gnttab_grant_foreign_access_ref(grant_ref_t ref, domid_t domid,
-+ unsigned long frame, int readonly);
-+
-+void gnttab_grant_foreign_transfer_ref(grant_ref_t, domid_t domid,
-+ unsigned long pfn);
-+
-+int gnttab_suspend(void);
-+int gnttab_resume(void);
-+
-+static inline void
-+gnttab_set_map_op(struct gnttab_map_grant_ref *map, maddr_t addr,
-+ uint32_t flags, grant_ref_t ref, domid_t domid)
-+{
-+ if (flags & GNTMAP_contains_pte)
-+ map->host_addr = addr;
-+ else if (xen_feature(XENFEAT_auto_translated_physmap))
-+ map->host_addr = __pa(addr);
-+ else
-+ map->host_addr = addr;
-+
-+ map->flags = flags;
-+ map->ref = ref;
-+ map->dom = domid;
-+}
-+
-+static inline void
-+gnttab_set_unmap_op(struct gnttab_unmap_grant_ref *unmap, maddr_t addr,
-+ uint32_t flags, grant_handle_t handle)
-+{
-+ if (flags & GNTMAP_contains_pte)
-+ unmap->host_addr = addr;
-+ else if (xen_feature(XENFEAT_auto_translated_physmap))
-+ unmap->host_addr = __pa(addr);
-+ else
-+ unmap->host_addr = addr;
-+
-+ unmap->handle = handle;
-+ unmap->dev_bus_addr = 0;
-+}
++#define FIRST_RESERVED_GDT_PAGE 14
++#define FIRST_RESERVED_GDT_BYTE (FIRST_RESERVED_GDT_PAGE * 4096)
++#define FIRST_RESERVED_GDT_ENTRY (FIRST_RESERVED_GDT_BYTE / 8)
+
-+#endif /* __ASM_GNTTAB_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/hvm.h tmp-linux-2.6-xen.patch/include/xen/hvm.h
---- pristine-linux-2.6.18.2/include/xen/hvm.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/hvm.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,23 @@
-+/* Simple wrappers around HVM functions */
-+#ifndef XEN_HVM_H__
-+#define XEN_HVM_H__
++/* Maximum number of virtual CPUs in multi-processor guests. */
++#define MAX_VIRT_CPUS 32
+
-+#include <xen/interface/hvm/params.h>
+
-+static inline unsigned long hvm_get_parameter(int idx)
-+{
-+ struct xen_hvm_param xhv;
-+ int r;
++/* Machine check support */
++#include "xen-mca.h"
+
-+ xhv.domid = DOMID_SELF;
-+ xhv.index = idx;
-+ r = HYPERVISOR_hvm_op(HVMOP_get_param, &xhv);
-+ if (r < 0) {
-+ printk(KERN_ERR "cannot get hvm parameter %d: %d.\n",
-+ idx, r);
-+ return 0;
-+ }
-+ return xhv.value;
-+}
++#ifndef __ASSEMBLY__
+
-+#endif /* XEN_HVM_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/hypercall.h tmp-linux-2.6-xen.patch/include/xen/hypercall.h
---- pristine-linux-2.6.18.2/include/xen/hypercall.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/hypercall.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,24 @@
-+#ifndef __XEN_HYPERCALL_H__
-+#define __XEN_HYPERCALL_H__
++typedef unsigned long xen_ulong_t;
+
-+#include <asm/hypercall.h>
++/*
++ * Send an array of these to HYPERVISOR_set_trap_table().
++ * The privilege level specifies which modes may enter a trap via a software
++ * interrupt. On x86/64, since rings 1 and 2 are unavailable, we allocate
++ * privilege levels as follows:
++ * Level == 0: Noone may enter
++ * Level == 1: Kernel may enter
++ * Level == 2: Kernel may enter
++ * Level == 3: Everyone may enter
++ */
++#define TI_GET_DPL(_ti) ((_ti)->flags & 3)
++#define TI_GET_IF(_ti) ((_ti)->flags & 4)
++#define TI_SET_DPL(_ti,_dpl) ((_ti)->flags |= (_dpl))
++#define TI_SET_IF(_ti,_if) ((_ti)->flags |= ((!!(_if))<<2))
++struct trap_info {
++ uint8_t vector; /* exception vector */
++ uint8_t flags; /* 0-3: privilege level; 4: clear event enable? */
++ uint16_t cs; /* code selector */
++ unsigned long address; /* code offset */
++};
++typedef struct trap_info trap_info_t;
++DEFINE_XEN_GUEST_HANDLE(trap_info_t);
+
-+static inline int
-+HYPERVISOR_multicall_check(
-+ multicall_entry_t *call_list, int nr_calls,
-+ const unsigned long *rc_list)
-+{
-+ int rc = HYPERVISOR_multicall(call_list, nr_calls);
++typedef uint64_t tsc_timestamp_t; /* RDTSC timestamp */
+
-+ if (unlikely(rc < 0))
-+ return rc;
-+ BUG_ON(rc);
++/*
++ * The following is all CPU context. Note that the fpu_ctxt block is filled
++ * in by FXSAVE if the CPU has feature FXSR; otherwise FSAVE is used.
++ */
++struct vcpu_guest_context {
++ /* FPU registers come first so they can be aligned for FXSAVE/FXRSTOR. */
++ struct { char x[512]; } fpu_ctxt; /* User-level FPU registers */
++#define VGCF_I387_VALID (1<<0)
++#define VGCF_IN_KERNEL (1<<2)
++#define _VGCF_i387_valid 0
++#define VGCF_i387_valid (1<<_VGCF_i387_valid)
++#define _VGCF_in_kernel 2
++#define VGCF_in_kernel (1<<_VGCF_in_kernel)
++#define _VGCF_failsafe_disables_events 3
++#define VGCF_failsafe_disables_events (1<<_VGCF_failsafe_disables_events)
++#define _VGCF_syscall_disables_events 4
++#define VGCF_syscall_disables_events (1<<_VGCF_syscall_disables_events)
++#define _VGCF_online 5
++#define VGCF_online (1<<_VGCF_online)
++ unsigned long flags; /* VGCF_* flags */
++ struct cpu_user_regs user_regs; /* User-level CPU registers */
++ struct trap_info trap_ctxt[256]; /* Virtual IDT */
++ unsigned long ldt_base, ldt_ents; /* LDT (linear address, # ents) */
++ unsigned long gdt_frames[16], gdt_ents; /* GDT (machine frames, # ents) */
++ unsigned long kernel_ss, kernel_sp; /* Virtual TSS (only SS1/SP1) */
++ /* NB. User pagetable on x86/64 is placed in ctrlreg[1]. */
++ unsigned long ctrlreg[8]; /* CR0-CR7 (control registers) */
++ unsigned long debugreg[8]; /* DB0-DB7 (debug registers) */
++#ifdef __i386__
++ unsigned long event_callback_cs; /* CS:EIP of event callback */
++ unsigned long event_callback_eip;
++ unsigned long failsafe_callback_cs; /* CS:EIP of failsafe callback */
++ unsigned long failsafe_callback_eip;
++#else
++ unsigned long event_callback_eip;
++ unsigned long failsafe_callback_eip;
++#ifdef __XEN__
++ union {
++ unsigned long syscall_callback_eip;
++ struct {
++ unsigned int event_callback_cs; /* compat CS of event cb */
++ unsigned int failsafe_callback_cs; /* compat CS of failsafe cb */
++ };
++ };
++#else
++ unsigned long syscall_callback_eip;
++#endif
++#endif
++ unsigned long vm_assist; /* VMASST_TYPE_* bitmap */
++#ifdef __x86_64__
++ /* Segment base addresses. */
++ uint64_t fs_base;
++ uint64_t gs_base_kernel;
++ uint64_t gs_base_user;
++#endif
++};
++typedef struct vcpu_guest_context vcpu_guest_context_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_guest_context_t);
+
-+ for ( ; nr_calls > 0; --nr_calls, ++call_list)
-+ if (unlikely(call_list->result != (rc_list ? *rc_list++ : 0)))
-+ return nr_calls;
++struct arch_shared_info {
++ unsigned long max_pfn; /* max pfn that appears in table */
++ /* Frame containing list of mfns containing list of mfns containing p2m. */
++ xen_pfn_t pfn_to_mfn_frame_list_list;
++ unsigned long nmi_reason;
++ uint64_t pad[32];
++};
++typedef struct arch_shared_info arch_shared_info_t;
+
-+ return 0;
-+}
++#endif /* !__ASSEMBLY__ */
+
-+#endif /* __XEN_HYPERCALL_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/hypervisor_sysfs.h tmp-linux-2.6-xen.patch/include/xen/hypervisor_sysfs.h
---- pristine-linux-2.6.18.2/include/xen/hypervisor_sysfs.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/hypervisor_sysfs.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,30 @@
+/*
-+ * copyright (c) 2006 IBM Corporation
-+ * Authored by: Mike D. Day <ncmike at us.ibm.com>
-+ *
-+ * This program is free software; you can redistribute it and/or modify
-+ * it under the terms of the GNU General Public License version 2 as
-+ * published by the Free Software Foundation.
++ * Prefix forces emulation of some non-trapping instructions.
++ * Currently only CPUID.
+ */
++#ifdef __ASSEMBLY__
++#define XEN_EMULATE_PREFIX .byte 0x0f,0x0b,0x78,0x65,0x6e ;
++#define XEN_CPUID XEN_EMULATE_PREFIX cpuid
++#else
++#define XEN_EMULATE_PREFIX ".byte 0x0f,0x0b,0x78,0x65,0x6e ; "
++#define XEN_CPUID XEN_EMULATE_PREFIX "cpuid"
++#endif
+
-+#ifndef _HYP_SYSFS_H_
-+#define _HYP_SYSFS_H_
-+
-+#include <linux/kobject.h>
-+#include <linux/sysfs.h>
-+
-+#define HYPERVISOR_ATTR_RO(_name) \
-+static struct hyp_sysfs_attr _name##_attr = __ATTR_RO(_name)
++#endif /* __XEN_PUBLIC_ARCH_X86_XEN_H__ */
+
-+#define HYPERVISOR_ATTR_RW(_name) \
-+static struct hyp_sysfs_attr _name##_attr = \
-+ __ATTR(_name, 0644, _name##_show, _name##_store)
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-x86_32.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-x86_32.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,27 @@
++/******************************************************************************
++ * arch-x86_32.h
++ *
++ * Guest OS interface to x86 32-bit Xen.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2004-2006, K A Fraser
++ */
+
-+struct hyp_sysfs_attr {
-+ struct attribute attr;
-+ ssize_t (*show)(struct hyp_sysfs_attr *, char *);
-+ ssize_t (*store)(struct hyp_sysfs_attr *, const char *, size_t);
-+ void *hyp_attr_data;
-+};
++#include "arch-x86/xen.h"
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/arch-x86_64.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/arch-x86_64.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,27 @@
++/******************************************************************************
++ * arch-x86_64.h
++ *
++ * Guest OS interface to x86 64-bit Xen.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2004-2006, K A Fraser
++ */
+
-+#endif /* _HYP_SYSFS_H_ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/acm.h tmp-linux-2.6-xen.patch/include/xen/interface/acm.h
---- pristine-linux-2.6.18.2/include/xen/interface/acm.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/acm.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,228 @@
-+/*
-+ * acm.h: Xen access control module interface defintions
++#include "arch-x86/xen.h"
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/callback.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/callback.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,121 @@
++/******************************************************************************
++ * callback.h
++ *
++ * Register guest OS callbacks with Xen.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -95937,204 +135191,95 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Reiner Sailer <sailer at watson.ibm.com>
-+ * Copyright (c) 2005, International Business Machines Corporation.
++ * Copyright (c) 2006, Ian Campbell
+ */
+
-+#ifndef _XEN_PUBLIC_ACM_H
-+#define _XEN_PUBLIC_ACM_H
++#ifndef __XEN_PUBLIC_CALLBACK_H__
++#define __XEN_PUBLIC_CALLBACK_H__
+
+#include "xen.h"
+
-+/* if ACM_DEBUG defined, all hooks should
-+ * print a short trace message (comment it out
-+ * when not in testing mode )
++/*
++ * Prototype for this hypercall is:
++ * long callback_op(int cmd, void *extra_args)
++ * @cmd == CALLBACKOP_??? (callback operation).
++ * @extra_args == Operation-specific extra arguments (NULL if none).
+ */
-+/* #define ACM_DEBUG */
-+
-+#ifdef ACM_DEBUG
-+# define printkd(fmt, args...) printk(fmt,## args)
-+#else
-+# define printkd(fmt, args...)
-+#endif
+
-+/* default ssid reference value if not supplied */
-+#define ACM_DEFAULT_SSID 0x0
-+#define ACM_DEFAULT_LOCAL_SSID 0x0
++/* ia64, x86: Callback for event delivery. */
++#define CALLBACKTYPE_event 0
+
-+/* Internal ACM ERROR types */
-+#define ACM_OK 0
-+#define ACM_UNDEF -1
-+#define ACM_INIT_SSID_ERROR -2
-+#define ACM_INIT_SOID_ERROR -3
-+#define ACM_ERROR -4
++/* x86: Failsafe callback when guest state cannot be restored by Xen. */
++#define CALLBACKTYPE_failsafe 1
+
-+/* External ACCESS DECISIONS */
-+#define ACM_ACCESS_PERMITTED 0
-+#define ACM_ACCESS_DENIED -111
-+#define ACM_NULL_POINTER_ERROR -200
++/* x86/64 hypervisor: Syscall by 64-bit guest app ('64-on-64-on-64'). */
++#define CALLBACKTYPE_syscall 2
+
+/*
-+ Error codes reported in when trying to test for a new policy
-+ These error codes are reported in an array of tuples where
-+ each error code is followed by a parameter describing the error
-+ more closely, such as a domain id.
-+*/
-+#define ACM_EVTCHN_SHARING_VIOLATION 0x100
-+#define ACM_GNTTAB_SHARING_VIOLATION 0x101
-+#define ACM_DOMAIN_LOOKUP 0x102
-+#define ACM_CHWALL_CONFLICT 0x103
-+#define ACM_SSIDREF_IN_USE 0x104
-+
-+
-+/* primary policy in lower 4 bits */
-+#define ACM_NULL_POLICY 0
-+#define ACM_CHINESE_WALL_POLICY 1
-+#define ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY 2
-+#define ACM_POLICY_UNDEFINED 15
-+
-+/* combinations have secondary policy component in higher 4bit */
-+#define ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY \
-+ ((ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY << 4) | ACM_CHINESE_WALL_POLICY)
-+
-+/* policy: */
-+#define ACM_POLICY_NAME(X) \
-+ ((X) == (ACM_NULL_POLICY)) ? "NULL" : \
-+ ((X) == (ACM_CHINESE_WALL_POLICY)) ? "CHINESE WALL" : \
-+ ((X) == (ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY)) ? "SIMPLE TYPE ENFORCEMENT" : \
-+ ((X) == (ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY)) ? "CHINESE WALL AND SIMPLE TYPE ENFORCEMENT" : \
-+ "UNDEFINED"
-+
-+/* the following policy versions must be increased
-+ * whenever the interpretation of the related
-+ * policy's data structure changes
++ * x86/32 hypervisor: Only available on x86/32 when supervisor_mode_kernel
++ * feature is enabled. Do not use this callback type in new code.
+ */
-+#define ACM_POLICY_VERSION 3
-+#define ACM_CHWALL_VERSION 1
-+#define ACM_STE_VERSION 1
-+
-+/* defines a ssid reference used by xen */
-+typedef uint32_t ssidref_t;
-+
-+/* hooks that are known to domains */
-+#define ACMHOOK_none 0
-+#define ACMHOOK_sharing 1
++#define CALLBACKTYPE_sysenter_deprecated 3
+
-+/* -------security policy relevant type definitions-------- */
++/* x86: Callback for NMI delivery. */
++#define CALLBACKTYPE_nmi 4
+
-+/* type identifier; compares to "equal" or "not equal" */
-+typedef uint16_t domaintype_t;
++/*
++ * x86: sysenter is only available as follows:
++ * - 32-bit hypervisor: with the supervisor_mode_kernel feature enabled
++ * - 64-bit hypervisor: 32-bit guest applications on Intel CPUs
++ * ('32-on-32-on-64', '32-on-64-on-64')
++ * [nb. also 64-bit guest applications on Intel CPUs
++ * ('64-on-64-on-64'), but syscall is preferred]
++ */
++#define CALLBACKTYPE_sysenter 5
+
-+/* CHINESE WALL POLICY DATA STRUCTURES
-+ *
-+ * current accumulated conflict type set:
-+ * When a domain is started and has a type that is in
-+ * a conflict set, the conflicting types are incremented in
-+ * the aggregate set. When a domain is destroyed, the
-+ * conflicting types to its type are decremented.
-+ * If a domain has multiple types, this procedure works over
-+ * all those types.
-+ *
-+ * conflict_aggregate_set[i] holds the number of
-+ * running domains that have a conflict with type i.
-+ *
-+ * running_types[i] holds the number of running domains
-+ * that include type i in their ssidref-referenced type set
-+ *
-+ * conflict_sets[i][j] is "0" if type j has no conflict
-+ * with type i and is "1" otherwise.
++/*
++ * x86/64 hypervisor: Syscall by 32-bit guest app on AMD CPUs
++ * ('32-on-32-on-64', '32-on-64-on-64')
+ */
-+/* high-16 = version, low-16 = check magic */
-+#define ACM_MAGIC 0x0001debc
++#define CALLBACKTYPE_syscall32 7
+
-+/* each offset in bytes from start of the struct they
-+ * are part of */
++/*
++ * Disable event deliver during callback? This flag is ignored for event and
++ * NMI callbacks: event delivery is unconditionally disabled.
++ */
++#define _CALLBACKF_mask_events 0
++#define CALLBACKF_mask_events (1U << _CALLBACKF_mask_events)
+
-+/* V3 of the policy buffer aded a version structure */
-+struct acm_policy_version
-+{
-+ uint32_t major;
-+ uint32_t minor;
++/*
++ * Register a callback.
++ */
++#define CALLBACKOP_register 0
++struct callback_register {
++ uint16_t type;
++ uint16_t flags;
++ xen_callback_t address;
+};
++typedef struct callback_register callback_register_t;
++DEFINE_XEN_GUEST_HANDLE(callback_register_t);
+
-+
-+/* each buffer consists of all policy information for
-+ * the respective policy given in the policy code
++/*
++ * Unregister a callback.
+ *
-+ * acm_policy_buffer, acm_chwall_policy_buffer,
-+ * and acm_ste_policy_buffer need to stay 32-bit aligned
-+ * because we create binary policies also with external
-+ * tools that assume packed representations (e.g. the java tool)
++ * Not all callbacks can be unregistered. -EINVAL will be returned if
++ * you attempt to unregister such a callback.
+ */
-+struct acm_policy_buffer {
-+ uint32_t policy_version; /* ACM_POLICY_VERSION */
-+ uint32_t magic;
-+ uint32_t len;
-+ uint32_t policy_reference_offset;
-+ uint32_t primary_policy_code;
-+ uint32_t primary_buffer_offset;
-+ uint32_t secondary_policy_code;
-+ uint32_t secondary_buffer_offset;
-+ struct acm_policy_version xml_pol_version; /* add in V3 */
-+};
-+
-+
-+struct acm_policy_reference_buffer {
-+ uint32_t len;
-+};
-+
-+struct acm_chwall_policy_buffer {
-+ uint32_t policy_version; /* ACM_CHWALL_VERSION */
-+ uint32_t policy_code;
-+ uint32_t chwall_max_types;
-+ uint32_t chwall_max_ssidrefs;
-+ uint32_t chwall_max_conflictsets;
-+ uint32_t chwall_ssid_offset;
-+ uint32_t chwall_conflict_sets_offset;
-+ uint32_t chwall_running_types_offset;
-+ uint32_t chwall_conflict_aggregate_offset;
-+};
-+
-+struct acm_ste_policy_buffer {
-+ uint32_t policy_version; /* ACM_STE_VERSION */
-+ uint32_t policy_code;
-+ uint32_t ste_max_types;
-+ uint32_t ste_max_ssidrefs;
-+ uint32_t ste_ssid_offset;
-+};
-+
-+struct acm_stats_buffer {
-+ uint32_t magic;
-+ uint32_t len;
-+ uint32_t primary_policy_code;
-+ uint32_t primary_stats_offset;
-+ uint32_t secondary_policy_code;
-+ uint32_t secondary_stats_offset;
-+};
-+
-+struct acm_ste_stats_buffer {
-+ uint32_t ec_eval_count;
-+ uint32_t gt_eval_count;
-+ uint32_t ec_denied_count;
-+ uint32_t gt_denied_count;
-+ uint32_t ec_cachehit_count;
-+ uint32_t gt_cachehit_count;
-+};
-+
-+struct acm_ssid_buffer {
-+ uint32_t len;
-+ ssidref_t ssidref;
-+ uint32_t policy_reference_offset;
-+ uint32_t primary_policy_code;
-+ uint32_t primary_max_types;
-+ uint32_t primary_types_offset;
-+ uint32_t secondary_policy_code;
-+ uint32_t secondary_max_types;
-+ uint32_t secondary_types_offset;
++#define CALLBACKOP_unregister 1
++struct callback_unregister {
++ uint16_t type;
++ uint16_t _unused;
+};
++typedef struct callback_unregister callback_unregister_t;
++DEFINE_XEN_GUEST_HANDLE(callback_unregister_t);
+
++#if __XEN_INTERFACE_VERSION__ < 0x00030207
++#undef CALLBACKTYPE_sysenter
++#define CALLBACKTYPE_sysenter CALLBACKTYPE_sysenter_deprecated
+#endif
+
++#endif /* __XEN_PUBLIC_CALLBACK_H__ */
++
+/*
+ * Local variables:
+ * mode: C
@@ -96144,13 +135289,15 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/acm_ops.h tmp-linux-2.6-xen.patch/include/xen/interface/acm_ops.h
---- pristine-linux-2.6.18.2/include/xen/interface/acm_ops.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/acm_ops.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,159 @@
-+/*
-+ * acm_ops.h: Xen access control module hypervisor commands
-+ *
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/dom0_ops.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/dom0_ops.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,120 @@
++/******************************************************************************
++ * dom0_ops.h
++ *
++ * Process command requests from domain-0 guest OS.
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -96169,134 +135316,93 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Reiner Sailer <sailer at watson.ibm.com>
-+ * Copyright (c) 2005,2006 International Business Machines Corporation.
++ * Copyright (c) 2002-2003, B Dragovic
++ * Copyright (c) 2002-2006, K Fraser
+ */
+
-+#ifndef __XEN_PUBLIC_ACM_OPS_H__
-+#define __XEN_PUBLIC_ACM_OPS_H__
++#ifndef __XEN_PUBLIC_DOM0_OPS_H__
++#define __XEN_PUBLIC_DOM0_OPS_H__
+
+#include "xen.h"
-+#include "acm.h"
-+
-+/*
-+ * Make sure you increment the interface version whenever you modify this file!
-+ * This makes sure that old versions of acm tools will stop working in a
-+ * well-defined way (rather than crashing the machine, for instance).
-+ */
-+#define ACM_INTERFACE_VERSION 0xAAAA000A
-+
-+/************************************************************************/
-+
-+/*
-+ * Prototype for this hypercall is:
-+ * int acm_op(int cmd, void *args)
-+ * @cmd == ACMOP_??? (access control module operation).
-+ * @args == Operation-specific extra arguments (NULL if none).
-+ */
-+
++#include "platform.h"
+
-+#define ACMOP_setpolicy 1
-+struct acm_setpolicy {
-+ /* IN */
-+ XEN_GUEST_HANDLE_64(void) pushcache;
-+ uint32_t pushcache_size;
-+};
++#if __XEN_INTERFACE_VERSION__ >= 0x00030204
++#error "dom0_ops.h is a compatibility interface only"
++#endif
+
++#define DOM0_INTERFACE_VERSION XENPF_INTERFACE_VERSION
+
-+#define ACMOP_getpolicy 2
-+struct acm_getpolicy {
-+ /* IN */
-+ XEN_GUEST_HANDLE_64(void) pullcache;
-+ uint32_t pullcache_size;
-+};
++#define DOM0_SETTIME XENPF_settime
++#define dom0_settime xenpf_settime
++#define dom0_settime_t xenpf_settime_t
+
++#define DOM0_ADD_MEMTYPE XENPF_add_memtype
++#define dom0_add_memtype xenpf_add_memtype
++#define dom0_add_memtype_t xenpf_add_memtype_t
+
-+#define ACMOP_dumpstats 3
-+struct acm_dumpstats {
-+ /* IN */
-+ XEN_GUEST_HANDLE_64(void) pullcache;
-+ uint32_t pullcache_size;
-+};
++#define DOM0_DEL_MEMTYPE XENPF_del_memtype
++#define dom0_del_memtype xenpf_del_memtype
++#define dom0_del_memtype_t xenpf_del_memtype_t
+
++#define DOM0_READ_MEMTYPE XENPF_read_memtype
++#define dom0_read_memtype xenpf_read_memtype
++#define dom0_read_memtype_t xenpf_read_memtype_t
+
-+#define ACMOP_getssid 4
-+#define ACM_GETBY_ssidref 1
-+#define ACM_GETBY_domainid 2
-+struct acm_getssid {
-+ /* IN */
-+ uint32_t get_ssid_by; /* ACM_GETBY_* */
-+ union {
-+ domaintype_t domainid;
-+ ssidref_t ssidref;
-+ } id;
-+ XEN_GUEST_HANDLE_64(void) ssidbuf;
-+ uint32_t ssidbuf_size;
-+};
++#define DOM0_MICROCODE XENPF_microcode_update
++#define dom0_microcode xenpf_microcode_update
++#define dom0_microcode_t xenpf_microcode_update_t
+
-+#define ACMOP_getdecision 5
-+struct acm_getdecision {
-+ /* IN */
-+ uint32_t get_decision_by1; /* ACM_GETBY_* */
-+ uint32_t get_decision_by2; /* ACM_GETBY_* */
-+ union {
-+ domaintype_t domainid;
-+ ssidref_t ssidref;
-+ } id1;
-+ union {
-+ domaintype_t domainid;
-+ ssidref_t ssidref;
-+ } id2;
-+ uint32_t hook;
-+ /* OUT */
-+ uint32_t acm_decision;
-+};
++#define DOM0_PLATFORM_QUIRK XENPF_platform_quirk
++#define dom0_platform_quirk xenpf_platform_quirk
++#define dom0_platform_quirk_t xenpf_platform_quirk_t
+
++typedef uint64_t cpumap_t;
+
-+#define ACMOP_chgpolicy 6
-+struct acm_change_policy {
-+ /* IN */
-+ XEN_GUEST_HANDLE_64(void) policy_pushcache;
-+ uint32_t policy_pushcache_size;
-+ XEN_GUEST_HANDLE_64(void) del_array;
-+ uint32_t delarray_size;
-+ XEN_GUEST_HANDLE_64(void) chg_array;
-+ uint32_t chgarray_size;
-+ /* OUT */
-+ /* array with error code */
-+ XEN_GUEST_HANDLE_64(void) err_array;
-+ uint32_t errarray_size;
++/* Unsupported legacy operation -- defined for API compatibility. */
++#define DOM0_MSR 15
++struct dom0_msr {
++ /* IN variables. */
++ uint32_t write;
++ cpumap_t cpu_mask;
++ uint32_t msr;
++ uint32_t in1;
++ uint32_t in2;
++ /* OUT variables. */
++ uint32_t out1;
++ uint32_t out2;
+};
++typedef struct dom0_msr dom0_msr_t;
++DEFINE_XEN_GUEST_HANDLE(dom0_msr_t);
+
-+#define ACMOP_relabeldoms 7
-+struct acm_relabel_doms {
-+ /* IN */
-+ XEN_GUEST_HANDLE_64(void) relabel_map;
-+ uint32_t relabel_map_size;
-+ /* OUT */
-+ XEN_GUEST_HANDLE_64(void) err_array;
-+ uint32_t errarray_size;
++/* Unsupported legacy operation -- defined for API compatibility. */
++#define DOM0_PHYSICAL_MEMORY_MAP 40
++struct dom0_memory_map_entry {
++ uint64_t start, end;
++ uint32_t flags; /* reserved */
++ uint8_t is_ram;
+};
++typedef struct dom0_memory_map_entry dom0_memory_map_entry_t;
++DEFINE_XEN_GUEST_HANDLE(dom0_memory_map_entry_t);
+
-+/* future interface to Xen */
-+struct xen_acmctl {
++struct dom0_op {
+ uint32_t cmd;
-+ uint32_t interface_version;
++ uint32_t interface_version; /* DOM0_INTERFACE_VERSION */
+ union {
-+ struct acm_setpolicy setpolicy;
-+ struct acm_getpolicy getpolicy;
-+ struct acm_dumpstats dumpstats;
-+ struct acm_getssid getssid;
-+ struct acm_getdecision getdecision;
-+ struct acm_change_policy change_policy;
-+ struct acm_relabel_doms relabel_doms;
++ struct dom0_msr msr;
++ struct dom0_settime settime;
++ struct dom0_add_memtype add_memtype;
++ struct dom0_del_memtype del_memtype;
++ struct dom0_read_memtype read_memtype;
++ struct dom0_microcode microcode;
++ struct dom0_platform_quirk platform_quirk;
++ struct dom0_memory_map_entry physical_memory_map;
++ uint8_t pad[128];
+ } u;
+};
++typedef struct dom0_op dom0_op_t;
++DEFINE_XEN_GUEST_HANDLE(dom0_op_t);
+
-+typedef struct xen_acmctl xen_acmctl_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_acmctl_t);
-+
-+#endif /* __XEN_PUBLIC_ACM_OPS_H__ */
++#endif /* __XEN_PUBLIC_DOM0_OPS_H__ */
+
+/*
+ * Local variables:
@@ -96307,15 +135413,15 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/arch-ia64.h tmp-linux-2.6-xen.patch/include/xen/interface/arch-ia64.h
---- pristine-linux-2.6.18.2/include/xen/interface/arch-ia64.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/arch-ia64.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,522 @@
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/domctl.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/domctl.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,676 @@
+/******************************************************************************
-+ * arch-ia64/hypervisor-if.h
++ * domctl.h
++ *
++ * Domain management operations. For use by node control stack.
+ *
-+ * Guest OS interface to IA64 Xen.
-+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -96334,495 +135440,649 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
++ * Copyright (c) 2002-2003, B Dragovic
++ * Copyright (c) 2002-2006, K Fraser
+ */
+
-+#ifndef __HYPERVISOR_IF_IA64_H__
-+#define __HYPERVISOR_IF_IA64_H__
++#ifndef __XEN_PUBLIC_DOMCTL_H__
++#define __XEN_PUBLIC_DOMCTL_H__
+
-+/* Structural guest handles introduced in 0x00030201. */
-+#if __XEN_INTERFACE_VERSION__ >= 0x00030201
-+#define __DEFINE_XEN_GUEST_HANDLE(name, type) \
-+ typedef struct { type *p; } __guest_handle_ ## name
-+#else
-+#define __DEFINE_XEN_GUEST_HANDLE(name, type) \
-+ typedef type * __guest_handle_ ## name
++#if !defined(__XEN__) && !defined(__XEN_TOOLS__)
++#error "domctl operations are intended for use by node control tools only"
+#endif
+
-+#define DEFINE_XEN_GUEST_HANDLE(name) __DEFINE_XEN_GUEST_HANDLE(name, name)
-+#define XEN_GUEST_HANDLE(name) __guest_handle_ ## name
-+#define XEN_GUEST_HANDLE_64(name) XEN_GUEST_HANDLE(name)
-+#define uint64_aligned_t uint64_t
-+#define set_xen_guest_handle(hnd, val) do { (hnd).p = val; } while (0)
-+#ifdef __XEN_TOOLS__
-+#define get_xen_guest_handle(val, hnd) do { val = (hnd).p; } while (0)
-+#endif
++#include "xen.h"
+
-+#ifndef __ASSEMBLY__
-+/* Guest handles for primitive C types. */
-+__DEFINE_XEN_GUEST_HANDLE(uchar, unsigned char);
-+__DEFINE_XEN_GUEST_HANDLE(uint, unsigned int);
-+__DEFINE_XEN_GUEST_HANDLE(ulong, unsigned long);
-+__DEFINE_XEN_GUEST_HANDLE(u64, unsigned long);
-+DEFINE_XEN_GUEST_HANDLE(char);
-+DEFINE_XEN_GUEST_HANDLE(int);
-+DEFINE_XEN_GUEST_HANDLE(long);
-+DEFINE_XEN_GUEST_HANDLE(void);
++#define XEN_DOMCTL_INTERFACE_VERSION 0x00000005
+
-+typedef unsigned long xen_pfn_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_pfn_t);
-+#define PRI_xen_pfn "lx"
-+#endif
++struct xenctl_cpumap {
++ XEN_GUEST_HANDLE_64(uint8) bitmap;
++ uint32_t nr_cpus;
++};
+
-+/* Arch specific VIRQs definition */
-+#define VIRQ_ITC VIRQ_ARCH_0 /* V. Virtual itc timer */
-+#define VIRQ_MCA_CMC VIRQ_ARCH_1 /* MCA cmc interrupt */
-+#define VIRQ_MCA_CPE VIRQ_ARCH_2 /* MCA cpe interrupt */
++/*
++ * NB. xen_domctl.domain is an IN/OUT parameter for this operation.
++ * If it is specified as zero, an id is auto-allocated and returned.
++ */
++#define XEN_DOMCTL_createdomain 1
++struct xen_domctl_createdomain {
++ /* IN parameters */
++ uint32_t ssidref;
++ xen_domain_handle_t handle;
++ /* Is this an HVM guest (as opposed to a PV guest)? */
++#define _XEN_DOMCTL_CDF_hvm_guest 0
++#define XEN_DOMCTL_CDF_hvm_guest (1U<<_XEN_DOMCTL_CDF_hvm_guest)
++ /* Use hardware-assisted paging if available? */
++#define _XEN_DOMCTL_CDF_hap 1
++#define XEN_DOMCTL_CDF_hap (1U<<_XEN_DOMCTL_CDF_hap)
++ uint32_t flags;
++};
++typedef struct xen_domctl_createdomain xen_domctl_createdomain_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_createdomain_t);
+
-+/* Maximum number of virtual CPUs in multi-processor guests. */
-+/* WARNING: before changing this, check that shared_info fits on a page */
-+#define MAX_VIRT_CPUS 64
++#define XEN_DOMCTL_destroydomain 2
++#define XEN_DOMCTL_pausedomain 3
++#define XEN_DOMCTL_unpausedomain 4
++#define XEN_DOMCTL_resumedomain 27
+
-+#ifndef __ASSEMBLY__
++#define XEN_DOMCTL_getdomaininfo 5
++struct xen_domctl_getdomaininfo {
++ /* OUT variables. */
++ domid_t domain; /* Also echoed in domctl.domain */
++ /* Domain is scheduled to die. */
++#define _XEN_DOMINF_dying 0
++#define XEN_DOMINF_dying (1U<<_XEN_DOMINF_dying)
++ /* Domain is an HVM guest (as opposed to a PV guest). */
++#define _XEN_DOMINF_hvm_guest 1
++#define XEN_DOMINF_hvm_guest (1U<<_XEN_DOMINF_hvm_guest)
++ /* The guest OS has shut down. */
++#define _XEN_DOMINF_shutdown 2
++#define XEN_DOMINF_shutdown (1U<<_XEN_DOMINF_shutdown)
++ /* Currently paused by control software. */
++#define _XEN_DOMINF_paused 3
++#define XEN_DOMINF_paused (1U<<_XEN_DOMINF_paused)
++ /* Currently blocked pending an event. */
++#define _XEN_DOMINF_blocked 4
++#define XEN_DOMINF_blocked (1U<<_XEN_DOMINF_blocked)
++ /* Domain is currently running. */
++#define _XEN_DOMINF_running 5
++#define XEN_DOMINF_running (1U<<_XEN_DOMINF_running)
++ /* Being debugged. */
++#define _XEN_DOMINF_debugged 6
++#define XEN_DOMINF_debugged (1U<<_XEN_DOMINF_debugged)
++ /* CPU to which this domain is bound. */
++#define XEN_DOMINF_cpumask 255
++#define XEN_DOMINF_cpushift 8
++ /* XEN_DOMINF_shutdown guest-supplied code. */
++#define XEN_DOMINF_shutdownmask 255
++#define XEN_DOMINF_shutdownshift 16
++ uint32_t flags; /* XEN_DOMINF_* */
++ uint64_aligned_t tot_pages;
++ uint64_aligned_t max_pages;
++ uint64_aligned_t shared_info_frame; /* GMFN of shared_info struct */
++ uint64_aligned_t cpu_time;
++ uint32_t nr_online_vcpus; /* Number of VCPUs currently online. */
++ uint32_t max_vcpu_id; /* Maximum VCPUID in use by this domain. */
++ uint32_t ssidref;
++ xen_domain_handle_t handle;
++};
++typedef struct xen_domctl_getdomaininfo xen_domctl_getdomaininfo_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_getdomaininfo_t);
+
-+typedef unsigned long xen_ulong_t;
+
-+#define INVALID_MFN (~0UL)
++#define XEN_DOMCTL_getmemlist 6
++struct xen_domctl_getmemlist {
++ /* IN variables. */
++ /* Max entries to write to output buffer. */
++ uint64_aligned_t max_pfns;
++ /* Start index in guest's page list. */
++ uint64_aligned_t start_pfn;
++ XEN_GUEST_HANDLE_64(uint64) buffer;
++ /* OUT variables. */
++ uint64_aligned_t num_pfns;
++};
++typedef struct xen_domctl_getmemlist xen_domctl_getmemlist_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_getmemlist_t);
+
-+#define MEM_G (1UL << 30)
-+#define MEM_M (1UL << 20)
-+#define MEM_K (1UL << 10)
+
-+#define MMIO_START (3 * MEM_G)
-+#define MMIO_SIZE (512 * MEM_M)
++#define XEN_DOMCTL_getpageframeinfo 7
+
-+#define VGA_IO_START 0xA0000UL
-+#define VGA_IO_SIZE 0x20000
++#define XEN_DOMCTL_PFINFO_LTAB_SHIFT 28
++#define XEN_DOMCTL_PFINFO_NOTAB (0x0U<<28)
++#define XEN_DOMCTL_PFINFO_L1TAB (0x1U<<28)
++#define XEN_DOMCTL_PFINFO_L2TAB (0x2U<<28)
++#define XEN_DOMCTL_PFINFO_L3TAB (0x3U<<28)
++#define XEN_DOMCTL_PFINFO_L4TAB (0x4U<<28)
++#define XEN_DOMCTL_PFINFO_LTABTYPE_MASK (0x7U<<28)
++#define XEN_DOMCTL_PFINFO_LPINTAB (0x1U<<31)
++#define XEN_DOMCTL_PFINFO_XTAB (0xfU<<28) /* invalid page */
++#define XEN_DOMCTL_PFINFO_LTAB_MASK (0xfU<<28)
+
-+#define LEGACY_IO_START (MMIO_START + MMIO_SIZE)
-+#define LEGACY_IO_SIZE (64*MEM_M)
++struct xen_domctl_getpageframeinfo {
++ /* IN variables. */
++ uint64_aligned_t gmfn; /* GMFN to query */
++ /* OUT variables. */
++ /* Is the page PINNED to a type? */
++ uint32_t type; /* see above type defs */
++};
++typedef struct xen_domctl_getpageframeinfo xen_domctl_getpageframeinfo_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_getpageframeinfo_t);
+
-+#define IO_PAGE_START (LEGACY_IO_START + LEGACY_IO_SIZE)
-+#define IO_PAGE_SIZE PAGE_SIZE
+
-+#define STORE_PAGE_START (IO_PAGE_START + IO_PAGE_SIZE)
-+#define STORE_PAGE_SIZE PAGE_SIZE
++#define XEN_DOMCTL_getpageframeinfo2 8
++struct xen_domctl_getpageframeinfo2 {
++ /* IN variables. */
++ uint64_aligned_t num;
++ /* IN/OUT variables. */
++ XEN_GUEST_HANDLE_64(uint32) array;
++};
++typedef struct xen_domctl_getpageframeinfo2 xen_domctl_getpageframeinfo2_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_getpageframeinfo2_t);
+
-+#define BUFFER_IO_PAGE_START (STORE_PAGE_START+STORE_PAGE_SIZE)
-+#define BUFFER_IO_PAGE_SIZE PAGE_SIZE
+
-+#define BUFFER_PIO_PAGE_START (BUFFER_IO_PAGE_START+BUFFER_IO_PAGE_SIZE)
-+#define BUFFER_PIO_PAGE_SIZE PAGE_SIZE
++/*
++ * Control shadow pagetables operation
++ */
++#define XEN_DOMCTL_shadow_op 10
+
-+#define IO_SAPIC_START 0xfec00000UL
-+#define IO_SAPIC_SIZE 0x100000
++/* Disable shadow mode. */
++#define XEN_DOMCTL_SHADOW_OP_OFF 0
+
-+#define PIB_START 0xfee00000UL
-+#define PIB_SIZE 0x200000
++/* Enable shadow mode (mode contains ORed XEN_DOMCTL_SHADOW_ENABLE_* flags). */
++#define XEN_DOMCTL_SHADOW_OP_ENABLE 32
+
-+#define GFW_START (4*MEM_G -16*MEM_M)
-+#define GFW_SIZE (16*MEM_M)
++/* Log-dirty bitmap operations. */
++ /* Return the bitmap and clean internal copy for next round. */
++#define XEN_DOMCTL_SHADOW_OP_CLEAN 11
++ /* Return the bitmap but do not modify internal copy. */
++#define XEN_DOMCTL_SHADOW_OP_PEEK 12
+
-+/* Nvram belongs to GFW memory space */
-+#define NVRAM_SIZE (MEM_K * 64)
-+#define NVRAM_START (GFW_START + 10 * MEM_M)
++/* Memory allocation accessors. */
++#define XEN_DOMCTL_SHADOW_OP_GET_ALLOCATION 30
++#define XEN_DOMCTL_SHADOW_OP_SET_ALLOCATION 31
+
-+#define NVRAM_VALID_SIG 0x4650494e45584948 // "HIXENIPF"
-+struct nvram_save_addr {
-+ unsigned long addr;
-+ unsigned long signature;
++/* Legacy enable operations. */
++ /* Equiv. to ENABLE with no mode flags. */
++#define XEN_DOMCTL_SHADOW_OP_ENABLE_TEST 1
++ /* Equiv. to ENABLE with mode flag ENABLE_LOG_DIRTY. */
++#define XEN_DOMCTL_SHADOW_OP_ENABLE_LOGDIRTY 2
++ /* Equiv. to ENABLE with mode flags ENABLE_REFCOUNT and ENABLE_TRANSLATE. */
++#define XEN_DOMCTL_SHADOW_OP_ENABLE_TRANSLATE 3
++
++/* Mode flags for XEN_DOMCTL_SHADOW_OP_ENABLE. */
++ /*
++ * Shadow pagetables are refcounted: guest does not use explicit mmu
++ * operations nor write-protect its pagetables.
++ */
++#define XEN_DOMCTL_SHADOW_ENABLE_REFCOUNT (1 << 1)
++ /*
++ * Log pages in a bitmap as they are dirtied.
++ * Used for live relocation to determine which pages must be re-sent.
++ */
++#define XEN_DOMCTL_SHADOW_ENABLE_LOG_DIRTY (1 << 2)
++ /*
++ * Automatically translate GPFNs into MFNs.
++ */
++#define XEN_DOMCTL_SHADOW_ENABLE_TRANSLATE (1 << 3)
++ /*
++ * Xen does not steal virtual address space from the guest.
++ * Requires HVM support.
++ */
++#define XEN_DOMCTL_SHADOW_ENABLE_EXTERNAL (1 << 4)
++
++struct xen_domctl_shadow_op_stats {
++ uint32_t fault_count;
++ uint32_t dirty_count;
+};
++typedef struct xen_domctl_shadow_op_stats xen_domctl_shadow_op_stats_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_shadow_op_stats_t);
+
-+struct pt_fpreg {
-+ union {
-+ unsigned long bits[2];
-+ long double __dummy; /* force 16-byte alignment */
-+ } u;
++struct xen_domctl_shadow_op {
++ /* IN variables. */
++ uint32_t op; /* XEN_DOMCTL_SHADOW_OP_* */
++
++ /* OP_ENABLE */
++ uint32_t mode; /* XEN_DOMCTL_SHADOW_ENABLE_* */
++
++ /* OP_GET_ALLOCATION / OP_SET_ALLOCATION */
++ uint32_t mb; /* Shadow memory allocation in MB */
++
++ /* OP_PEEK / OP_CLEAN */
++ XEN_GUEST_HANDLE_64(uint8) dirty_bitmap;
++ uint64_aligned_t pages; /* Size of buffer. Updated with actual size. */
++ struct xen_domctl_shadow_op_stats stats;
+};
++typedef struct xen_domctl_shadow_op xen_domctl_shadow_op_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_shadow_op_t);
+
-+struct cpu_user_regs {
-+ /* The following registers are saved by SAVE_MIN: */
-+ unsigned long b6; /* scratch */
-+ unsigned long b7; /* scratch */
-+
-+ unsigned long ar_csd; /* used by cmp8xchg16 (scratch) */
-+ unsigned long ar_ssd; /* reserved for future use (scratch) */
-+
-+ unsigned long r8; /* scratch (return value register 0) */
-+ unsigned long r9; /* scratch (return value register 1) */
-+ unsigned long r10; /* scratch (return value register 2) */
-+ unsigned long r11; /* scratch (return value register 3) */
-+
-+ unsigned long cr_ipsr; /* interrupted task's psr */
-+ unsigned long cr_iip; /* interrupted task's instruction pointer */
-+ unsigned long cr_ifs; /* interrupted task's function state */
-+
-+ unsigned long ar_unat; /* interrupted task's NaT register (preserved) */
-+ unsigned long ar_pfs; /* prev function state */
-+ unsigned long ar_rsc; /* RSE configuration */
-+ /* The following two are valid only if cr_ipsr.cpl > 0: */
-+ unsigned long ar_rnat; /* RSE NaT */
-+ unsigned long ar_bspstore; /* RSE bspstore */
-+
-+ unsigned long pr; /* 64 predicate registers (1 bit each) */
-+ unsigned long b0; /* return pointer (bp) */
-+ unsigned long loadrs; /* size of dirty partition << 16 */
-+
-+ unsigned long r1; /* the gp pointer */
-+ unsigned long r12; /* interrupted task's memory stack pointer */
-+ unsigned long r13; /* thread pointer */
-+
-+ unsigned long ar_fpsr; /* floating point status (preserved) */
-+ unsigned long r15; /* scratch */
-+
-+ /* The remaining registers are NOT saved for system calls. */
-+
-+ unsigned long r14; /* scratch */
-+ unsigned long r2; /* scratch */
-+ unsigned long r3; /* scratch */
-+ unsigned long r16; /* scratch */
-+ unsigned long r17; /* scratch */
-+ unsigned long r18; /* scratch */
-+ unsigned long r19; /* scratch */
-+ unsigned long r20; /* scratch */
-+ unsigned long r21; /* scratch */
-+ unsigned long r22; /* scratch */
-+ unsigned long r23; /* scratch */
-+ unsigned long r24; /* scratch */
-+ unsigned long r25; /* scratch */
-+ unsigned long r26; /* scratch */
-+ unsigned long r27; /* scratch */
-+ unsigned long r28; /* scratch */
-+ unsigned long r29; /* scratch */
-+ unsigned long r30; /* scratch */
-+ unsigned long r31; /* scratch */
-+ unsigned long ar_ccv; /* compare/exchange value (scratch) */
+
-+ /*
-+ * Floating point registers that the kernel considers scratch:
-+ */
-+ struct pt_fpreg f6; /* scratch */
-+ struct pt_fpreg f7; /* scratch */
-+ struct pt_fpreg f8; /* scratch */
-+ struct pt_fpreg f9; /* scratch */
-+ struct pt_fpreg f10; /* scratch */
-+ struct pt_fpreg f11; /* scratch */
-+ unsigned long r4; /* preserved */
-+ unsigned long r5; /* preserved */
-+ unsigned long r6; /* preserved */
-+ unsigned long r7; /* preserved */
-+ unsigned long eml_unat; /* used for emulating instruction */
-+ unsigned long pad0; /* alignment pad */
++#define XEN_DOMCTL_max_mem 11
++struct xen_domctl_max_mem {
++ /* IN variables. */
++ uint64_aligned_t max_memkb;
++};
++typedef struct xen_domctl_max_mem xen_domctl_max_mem_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_max_mem_t);
+
++
++#define XEN_DOMCTL_setvcpucontext 12
++#define XEN_DOMCTL_getvcpucontext 13
++struct xen_domctl_vcpucontext {
++ uint32_t vcpu; /* IN */
++ XEN_GUEST_HANDLE_64(vcpu_guest_context_t) ctxt; /* IN/OUT */
+};
-+typedef struct cpu_user_regs cpu_user_regs_t;
++typedef struct xen_domctl_vcpucontext xen_domctl_vcpucontext_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_vcpucontext_t);
+
-+union vac {
-+ unsigned long value;
-+ struct {
-+ int a_int:1;
-+ int a_from_int_cr:1;
-+ int a_to_int_cr:1;
-+ int a_from_psr:1;
-+ int a_from_cpuid:1;
-+ int a_cover:1;
-+ int a_bsw:1;
-+ long reserved:57;
-+ };
++
++#define XEN_DOMCTL_getvcpuinfo 14
++struct xen_domctl_getvcpuinfo {
++ /* IN variables. */
++ uint32_t vcpu;
++ /* OUT variables. */
++ uint8_t online; /* currently online (not hotplugged)? */
++ uint8_t blocked; /* blocked waiting for an event? */
++ uint8_t running; /* currently scheduled on its CPU? */
++ uint64_aligned_t cpu_time; /* total cpu time consumed (ns) */
++ uint32_t cpu; /* current mapping */
+};
-+typedef union vac vac_t;
++typedef struct xen_domctl_getvcpuinfo xen_domctl_getvcpuinfo_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_getvcpuinfo_t);
+
-+union vdc {
-+ unsigned long value;
-+ struct {
-+ int d_vmsw:1;
-+ int d_extint:1;
-+ int d_ibr_dbr:1;
-+ int d_pmc:1;
-+ int d_to_pmd:1;
-+ int d_itm:1;
-+ long reserved:58;
-+ };
++
++/* Get/set which physical cpus a vcpu can execute on. */
++#define XEN_DOMCTL_setvcpuaffinity 9
++#define XEN_DOMCTL_getvcpuaffinity 25
++struct xen_domctl_vcpuaffinity {
++ uint32_t vcpu; /* IN */
++ struct xenctl_cpumap cpumap; /* IN/OUT */
+};
-+typedef union vdc vdc_t;
++typedef struct xen_domctl_vcpuaffinity xen_domctl_vcpuaffinity_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_vcpuaffinity_t);
+
-+struct mapped_regs {
-+ union vac vac;
-+ union vdc vdc;
-+ unsigned long virt_env_vaddr;
-+ unsigned long reserved1[29];
-+ unsigned long vhpi;
-+ unsigned long reserved2[95];
-+ union {
-+ unsigned long vgr[16];
-+ unsigned long bank1_regs[16]; // bank1 regs (r16-r31) when bank0 active
-+ };
-+ union {
-+ unsigned long vbgr[16];
-+ unsigned long bank0_regs[16]; // bank0 regs (r16-r31) when bank1 active
-+ };
-+ unsigned long vnat;
-+ unsigned long vbnat;
-+ unsigned long vcpuid[5];
-+ unsigned long reserved3[11];
-+ unsigned long vpsr;
-+ unsigned long vpr;
-+ unsigned long reserved4[76];
-+ union {
-+ unsigned long vcr[128];
-+ struct {
-+ unsigned long dcr; // CR0
-+ unsigned long itm;
-+ unsigned long iva;
-+ unsigned long rsv1[5];
-+ unsigned long pta; // CR8
-+ unsigned long rsv2[7];
-+ unsigned long ipsr; // CR16
-+ unsigned long isr;
-+ unsigned long rsv3;
-+ unsigned long iip;
-+ unsigned long ifa;
-+ unsigned long itir;
-+ unsigned long iipa;
-+ unsigned long ifs;
-+ unsigned long iim; // CR24
-+ unsigned long iha;
-+ unsigned long rsv4[38];
-+ unsigned long lid; // CR64
-+ unsigned long ivr;
-+ unsigned long tpr;
-+ unsigned long eoi;
-+ unsigned long irr[4];
-+ unsigned long itv; // CR72
-+ unsigned long pmv;
-+ unsigned long cmcv;
-+ unsigned long rsv5[5];
-+ unsigned long lrr0; // CR80
-+ unsigned long lrr1;
-+ unsigned long rsv6[46];
-+ };
-+ };
++
++#define XEN_DOMCTL_max_vcpus 15
++struct xen_domctl_max_vcpus {
++ uint32_t max; /* maximum number of vcpus */
++};
++typedef struct xen_domctl_max_vcpus xen_domctl_max_vcpus_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_max_vcpus_t);
++
++
++#define XEN_DOMCTL_scheduler_op 16
++/* Scheduler types. */
++#define XEN_SCHEDULER_SEDF 4
++#define XEN_SCHEDULER_CREDIT 5
++/* Set or get info? */
++#define XEN_DOMCTL_SCHEDOP_putinfo 0
++#define XEN_DOMCTL_SCHEDOP_getinfo 1
++struct xen_domctl_scheduler_op {
++ uint32_t sched_id; /* XEN_SCHEDULER_* */
++ uint32_t cmd; /* XEN_DOMCTL_SCHEDOP_* */
+ union {
-+ unsigned long reserved5[128];
-+ struct {
-+ unsigned long precover_ifs;
-+ unsigned long unat; // not sure if this is needed until NaT arch is done
-+ int interrupt_collection_enabled; // virtual psr.ic
-+ /* virtual interrupt deliverable flag is evtchn_upcall_mask in
-+ * shared info area now. interrupt_mask_addr is the address
-+ * of evtchn_upcall_mask for current vcpu
-+ */
-+ unsigned char *interrupt_mask_addr;
-+ int pending_interruption;
-+ unsigned char vpsr_pp;
-+ unsigned char vpsr_dfh;
-+ unsigned char hpsr_dfh;
-+ unsigned char hpsr_mfh;
-+ unsigned long reserved5_1[4];
-+ int metaphysical_mode; // 1 = use metaphys mapping, 0 = use virtual
-+ int banknum; // 0 or 1, which virtual register bank is active
-+ unsigned long rrs[8]; // region registers
-+ unsigned long krs[8]; // kernel registers
-+ unsigned long pkrs[8]; // protection key registers
-+ unsigned long tmp[8]; // temp registers (e.g. for hyperprivops)
-+ };
-+ };
++ struct xen_domctl_sched_sedf {
++ uint64_aligned_t period;
++ uint64_aligned_t slice;
++ uint64_aligned_t latency;
++ uint32_t extratime;
++ uint32_t weight;
++ } sedf;
++ struct xen_domctl_sched_credit {
++ uint16_t weight;
++ uint16_t cap;
++ } credit;
++ } u;
+};
-+typedef struct mapped_regs mapped_regs_t;
++typedef struct xen_domctl_scheduler_op xen_domctl_scheduler_op_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_scheduler_op_t);
+
-+struct vpd {
-+ struct mapped_regs vpd_low;
-+ unsigned long reserved6[3456];
-+ unsigned long vmm_avail[128];
-+ unsigned long reserved7[4096];
++
++#define XEN_DOMCTL_setdomainhandle 17
++struct xen_domctl_setdomainhandle {
++ xen_domain_handle_t handle;
+};
-+typedef struct vpd vpd_t;
++typedef struct xen_domctl_setdomainhandle xen_domctl_setdomainhandle_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_setdomainhandle_t);
+
-+struct arch_vcpu_info {
++
++#define XEN_DOMCTL_setdebugging 18
++struct xen_domctl_setdebugging {
++ uint8_t enable;
+};
-+typedef struct arch_vcpu_info arch_vcpu_info_t;
++typedef struct xen_domctl_setdebugging xen_domctl_setdebugging_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_setdebugging_t);
+
-+struct arch_shared_info {
-+ /* PFN of the start_info page. */
-+ unsigned long start_info_pfn;
+
-+ /* Interrupt vector for event channel. */
-+ int evtchn_vector;
++#define XEN_DOMCTL_irq_permission 19
++struct xen_domctl_irq_permission {
++ uint8_t pirq;
++ uint8_t allow_access; /* flag to specify enable/disable of IRQ access */
++};
++typedef struct xen_domctl_irq_permission xen_domctl_irq_permission_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_irq_permission_t);
++
+
-+ uint64_t pad[32];
++#define XEN_DOMCTL_iomem_permission 20
++struct xen_domctl_iomem_permission {
++ uint64_aligned_t first_mfn;/* first page (physical page number) in range */
++ uint64_aligned_t nr_mfns; /* number of pages in range (>0) */
++ uint8_t allow_access; /* allow (!0) or deny (0) access to range? */
+};
-+typedef struct arch_shared_info arch_shared_info_t;
++typedef struct xen_domctl_iomem_permission xen_domctl_iomem_permission_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_iomem_permission_t);
+
-+typedef unsigned long xen_callback_t;
+
-+struct ia64_tr_entry {
-+ unsigned long pte;
-+ unsigned long itir;
-+ unsigned long vadr;
-+ unsigned long rid;
++#define XEN_DOMCTL_ioport_permission 21
++struct xen_domctl_ioport_permission {
++ uint32_t first_port; /* first port int range */
++ uint32_t nr_ports; /* size of port range */
++ uint8_t allow_access; /* allow or deny access to range? */
+};
++typedef struct xen_domctl_ioport_permission xen_domctl_ioport_permission_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_ioport_permission_t);
+
-+struct vcpu_extra_regs {
-+ struct ia64_tr_entry itrs[8];
-+ struct ia64_tr_entry dtrs[8];
-+ unsigned long iva;
-+ unsigned long dcr;
-+ unsigned long event_callback_ip;
++
++#define XEN_DOMCTL_hypercall_init 22
++struct xen_domctl_hypercall_init {
++ uint64_aligned_t gmfn; /* GMFN to be initialised */
+};
++typedef struct xen_domctl_hypercall_init xen_domctl_hypercall_init_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_hypercall_init_t);
+
-+struct vcpu_guest_context {
-+#define VGCF_EXTRA_REGS (1<<1) /* Get/Set extra regs. */
-+ unsigned long flags; /* VGCF_* flags */
+
-+ struct cpu_user_regs user_regs;
-+ struct vcpu_extra_regs extra_regs;
-+ unsigned long privregs_pfn;
-+};
-+typedef struct vcpu_guest_context vcpu_guest_context_t;
-+DEFINE_XEN_GUEST_HANDLE(vcpu_guest_context_t);
++#define XEN_DOMCTL_arch_setup 23
++#define _XEN_DOMAINSETUP_hvm_guest 0
++#define XEN_DOMAINSETUP_hvm_guest (1UL<<_XEN_DOMAINSETUP_hvm_guest)
++#define _XEN_DOMAINSETUP_query 1 /* Get parameters (for save) */
++#define XEN_DOMAINSETUP_query (1UL<<_XEN_DOMAINSETUP_query)
++#define _XEN_DOMAINSETUP_sioemu_guest 2
++#define XEN_DOMAINSETUP_sioemu_guest (1UL<<_XEN_DOMAINSETUP_sioemu_guest)
++typedef struct xen_domctl_arch_setup {
++ uint64_aligned_t flags; /* XEN_DOMAINSETUP_* */
++#ifdef __ia64__
++ uint64_aligned_t bp; /* mpaddr of boot param area */
++ uint64_aligned_t maxmem; /* Highest memory address for MDT. */
++ uint64_aligned_t xsi_va; /* Xen shared_info area virtual address. */
++ uint32_t hypercall_imm; /* Break imm for Xen hypercalls. */
++ int8_t vhpt_size_log2; /* Log2 of VHPT size. */
++#endif
++} xen_domctl_arch_setup_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_arch_setup_t);
+
-+/* dom0 vp op */
-+#define __HYPERVISOR_ia64_dom0vp_op __HYPERVISOR_arch_0
-+/* Map io space in machine address to dom0 physical address space.
-+ Currently physical assigned address equals to machine address. */
-+#define IA64_DOM0VP_ioremap 0
+
-+/* Convert a pseudo physical page frame number to the corresponding
-+ machine page frame number. If no page is assigned, INVALID_MFN or
-+ GPFN_INV_MASK is returned depending on domain's non-vti/vti mode. */
-+#define IA64_DOM0VP_phystomach 1
++#define XEN_DOMCTL_settimeoffset 24
++struct xen_domctl_settimeoffset {
++ int32_t time_offset_seconds; /* applied to domain wallclock time */
++};
++typedef struct xen_domctl_settimeoffset xen_domctl_settimeoffset_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_settimeoffset_t);
+
-+/* Convert a machine page frame number to the corresponding pseudo physical
-+ page frame number of the caller domain. */
-+#define IA64_DOM0VP_machtophys 3
++
++#define XEN_DOMCTL_gethvmcontext 33
++#define XEN_DOMCTL_sethvmcontext 34
++typedef struct xen_domctl_hvmcontext {
++ uint32_t size; /* IN/OUT: size of buffer / bytes filled */
++ XEN_GUEST_HANDLE_64(uint8) buffer; /* IN/OUT: data, or call
++ * gethvmcontext with NULL
++ * buffer to get size req'd */
++} xen_domctl_hvmcontext_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_hvmcontext_t);
+
-+/* Reserved for future use. */
-+#define IA64_DOM0VP_iounmap 4
+
-+/* Unmap and free pages contained in the specified pseudo physical region. */
-+#define IA64_DOM0VP_zap_physmap 5
++#define XEN_DOMCTL_set_address_size 35
++#define XEN_DOMCTL_get_address_size 36
++typedef struct xen_domctl_address_size {
++ uint32_t size;
++} xen_domctl_address_size_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_address_size_t);
+
-+/* Assign machine page frame to dom0's pseudo physical address space. */
-+#define IA64_DOM0VP_add_physmap 6
+
-+/* expose the p2m table into domain */
-+#define IA64_DOM0VP_expose_p2m 7
++#define XEN_DOMCTL_real_mode_area 26
++struct xen_domctl_real_mode_area {
++ uint32_t log; /* log2 of Real Mode Area size */
++};
++typedef struct xen_domctl_real_mode_area xen_domctl_real_mode_area_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_real_mode_area_t);
+
-+/* xen perfmon */
-+#define IA64_DOM0VP_perfmon 8
+
-+/* gmfn version of IA64_DOM0VP_add_physmap */
-+#define IA64_DOM0VP_add_physmap_with_gmfn 9
++#define XEN_DOMCTL_sendtrigger 28
++#define XEN_DOMCTL_SENDTRIGGER_NMI 0
++#define XEN_DOMCTL_SENDTRIGGER_RESET 1
++#define XEN_DOMCTL_SENDTRIGGER_INIT 2
++struct xen_domctl_sendtrigger {
++ uint32_t trigger; /* IN */
++ uint32_t vcpu; /* IN */
++};
++typedef struct xen_domctl_sendtrigger xen_domctl_sendtrigger_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_sendtrigger_t);
+
-+/* Add an I/O port space range */
-+#define IA64_DOM0VP_add_io_space 11
+
-+// flags for page assignement to pseudo physical address space
-+#define _ASSIGN_readonly 0
-+#define ASSIGN_readonly (1UL << _ASSIGN_readonly)
-+#define ASSIGN_writable (0UL << _ASSIGN_readonly) // dummy flag
-+/* Internal only: memory attribute must be WC/UC/UCE. */
-+#define _ASSIGN_nocache 1
-+#define ASSIGN_nocache (1UL << _ASSIGN_nocache)
-+// tlb tracking
-+#define _ASSIGN_tlb_track 2
-+#define ASSIGN_tlb_track (1UL << _ASSIGN_tlb_track)
-+/* Internal only: associated with PGC_allocated bit */
-+#define _ASSIGN_pgc_allocated 3
-+#define ASSIGN_pgc_allocated (1UL << _ASSIGN_pgc_allocated)
++/* Assign PCI device to HVM guest. Sets up IOMMU structures. */
++#define XEN_DOMCTL_assign_device 37
++#define XEN_DOMCTL_test_assign_device 45
++#define XEN_DOMCTL_deassign_device 47
++struct xen_domctl_assign_device {
++ uint32_t machine_bdf; /* machine PCI ID of assigned device */
++};
++typedef struct xen_domctl_assign_device xen_domctl_assign_device_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_assign_device_t);
++
++/* Retrieve sibling devices infomation of machine_bdf */
++#define XEN_DOMCTL_get_device_group 50
++struct xen_domctl_get_device_group {
++ uint32_t machine_bdf; /* IN */
++ uint32_t max_sdevs; /* IN */
++ uint32_t num_sdevs; /* OUT */
++ XEN_GUEST_HANDLE_64(uint32) sdev_array; /* OUT */
++};
++typedef struct xen_domctl_get_device_group xen_domctl_get_device_group_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_get_device_group_t);
++
++/* Pass-through interrupts: bind real irq -> hvm devfn. */
++#define XEN_DOMCTL_bind_pt_irq 38
++#define XEN_DOMCTL_unbind_pt_irq 48
++typedef enum pt_irq_type_e {
++ PT_IRQ_TYPE_PCI,
++ PT_IRQ_TYPE_ISA,
++ PT_IRQ_TYPE_MSI,
++} pt_irq_type_t;
++struct xen_domctl_bind_pt_irq {
++ uint32_t machine_irq;
++ pt_irq_type_t irq_type;
++ uint32_t hvm_domid;
+
-+/* This structure has the same layout of struct ia64_boot_param, defined in
-+ <asm/system.h>. It is redefined here to ease use. */
-+struct xen_ia64_boot_param {
-+ unsigned long command_line; /* physical address of cmd line args */
-+ unsigned long efi_systab; /* physical address of EFI system table */
-+ unsigned long efi_memmap; /* physical address of EFI memory map */
-+ unsigned long efi_memmap_size; /* size of EFI memory map */
-+ unsigned long efi_memdesc_size; /* size of an EFI memory map descriptor */
-+ unsigned int efi_memdesc_version; /* memory descriptor version */
-+ struct {
-+ unsigned short num_cols; /* number of columns on console. */
-+ unsigned short num_rows; /* number of rows on console. */
-+ unsigned short orig_x; /* cursor's x position */
-+ unsigned short orig_y; /* cursor's y position */
-+ } console_info;
-+ unsigned long fpswa; /* physical address of the fpswa interface */
-+ unsigned long initrd_start;
-+ unsigned long initrd_size;
-+ unsigned long domain_start; /* va where the boot time domain begins */
-+ unsigned long domain_size; /* how big is the boot domain */
++ union {
++ struct {
++ uint8_t isa_irq;
++ } isa;
++ struct {
++ uint8_t bus;
++ uint8_t device;
++ uint8_t intx;
++ } pci;
++ struct {
++ uint8_t gvec;
++ uint32_t gflags;
++ } msi;
++ } u;
+};
++typedef struct xen_domctl_bind_pt_irq xen_domctl_bind_pt_irq_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_bind_pt_irq_t);
+
-+#endif /* !__ASSEMBLY__ */
+
-+/* Size of the shared_info area (this is not related to page size). */
-+#define XSI_SHIFT 14
-+#define XSI_SIZE (1 << XSI_SHIFT)
-+/* Log size of mapped_regs area (64 KB - only 4KB is used). */
-+#define XMAPPEDREGS_SHIFT 12
-+#define XMAPPEDREGS_SIZE (1 << XMAPPEDREGS_SHIFT)
-+/* Offset of XASI (Xen arch shared info) wrt XSI_BASE. */
-+#define XMAPPEDREGS_OFS XSI_SIZE
++/* Bind machine I/O address range -> HVM address range. */
++#define XEN_DOMCTL_memory_mapping 39
++#define DPCI_ADD_MAPPING 1
++#define DPCI_REMOVE_MAPPING 0
++struct xen_domctl_memory_mapping {
++ uint64_aligned_t first_gfn; /* first page (hvm guest phys page) in range */
++ uint64_aligned_t first_mfn; /* first page (machine page) in range */
++ uint64_aligned_t nr_mfns; /* number of pages in range (>0) */
++ uint32_t add_mapping; /* add or remove mapping */
++ uint32_t padding; /* padding for 64-bit aligned structure */
++};
++typedef struct xen_domctl_memory_mapping xen_domctl_memory_mapping_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_memory_mapping_t);
++
++
++/* Bind machine I/O port range -> HVM I/O port range. */
++#define XEN_DOMCTL_ioport_mapping 40
++struct xen_domctl_ioport_mapping {
++ uint32_t first_gport; /* first guest IO port*/
++ uint32_t first_mport; /* first machine IO port */
++ uint32_t nr_ports; /* size of port range */
++ uint32_t add_mapping; /* add or remove mapping */
++};
++typedef struct xen_domctl_ioport_mapping xen_domctl_ioport_mapping_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_ioport_mapping_t);
+
-+/* Hyperprivops. */
-+#define HYPERPRIVOP_START 0x1
-+#define HYPERPRIVOP_RFI (HYPERPRIVOP_START + 0x0)
-+#define HYPERPRIVOP_RSM_DT (HYPERPRIVOP_START + 0x1)
-+#define HYPERPRIVOP_SSM_DT (HYPERPRIVOP_START + 0x2)
-+#define HYPERPRIVOP_COVER (HYPERPRIVOP_START + 0x3)
-+#define HYPERPRIVOP_ITC_D (HYPERPRIVOP_START + 0x4)
-+#define HYPERPRIVOP_ITC_I (HYPERPRIVOP_START + 0x5)
-+#define HYPERPRIVOP_SSM_I (HYPERPRIVOP_START + 0x6)
-+#define HYPERPRIVOP_GET_IVR (HYPERPRIVOP_START + 0x7)
-+#define HYPERPRIVOP_GET_TPR (HYPERPRIVOP_START + 0x8)
-+#define HYPERPRIVOP_SET_TPR (HYPERPRIVOP_START + 0x9)
-+#define HYPERPRIVOP_EOI (HYPERPRIVOP_START + 0xa)
-+#define HYPERPRIVOP_SET_ITM (HYPERPRIVOP_START + 0xb)
-+#define HYPERPRIVOP_THASH (HYPERPRIVOP_START + 0xc)
-+#define HYPERPRIVOP_PTC_GA (HYPERPRIVOP_START + 0xd)
-+#define HYPERPRIVOP_ITR_D (HYPERPRIVOP_START + 0xe)
-+#define HYPERPRIVOP_GET_RR (HYPERPRIVOP_START + 0xf)
-+#define HYPERPRIVOP_SET_RR (HYPERPRIVOP_START + 0x10)
-+#define HYPERPRIVOP_SET_KR (HYPERPRIVOP_START + 0x11)
-+#define HYPERPRIVOP_FC (HYPERPRIVOP_START + 0x12)
-+#define HYPERPRIVOP_GET_CPUID (HYPERPRIVOP_START + 0x13)
-+#define HYPERPRIVOP_GET_PMD (HYPERPRIVOP_START + 0x14)
-+#define HYPERPRIVOP_GET_EFLAG (HYPERPRIVOP_START + 0x15)
-+#define HYPERPRIVOP_SET_EFLAG (HYPERPRIVOP_START + 0x16)
-+#define HYPERPRIVOP_RSM_BE (HYPERPRIVOP_START + 0x17)
-+#define HYPERPRIVOP_GET_PSR (HYPERPRIVOP_START + 0x18)
-+#define HYPERPRIVOP_MAX (0x19)
+
-+/* Fast and light hypercalls. */
-+#define __HYPERVISOR_ia64_fast_eoi __HYPERVISOR_arch_1
++/*
++ * Pin caching type of RAM space for x86 HVM domU.
++ */
++#define XEN_DOMCTL_pin_mem_cacheattr 41
++/* Caching types: these happen to be the same as x86 MTRR/PAT type codes. */
++#define XEN_DOMCTL_MEM_CACHEATTR_UC 0
++#define XEN_DOMCTL_MEM_CACHEATTR_WC 1
++#define XEN_DOMCTL_MEM_CACHEATTR_WT 4
++#define XEN_DOMCTL_MEM_CACHEATTR_WP 5
++#define XEN_DOMCTL_MEM_CACHEATTR_WB 6
++#define XEN_DOMCTL_MEM_CACHEATTR_UCM 7
++struct xen_domctl_pin_mem_cacheattr {
++ uint64_aligned_t start, end;
++ unsigned int type; /* XEN_DOMCTL_MEM_CACHEATTR_* */
++};
++typedef struct xen_domctl_pin_mem_cacheattr xen_domctl_pin_mem_cacheattr_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_pin_mem_cacheattr_t);
++
++
++#define XEN_DOMCTL_set_ext_vcpucontext 42
++#define XEN_DOMCTL_get_ext_vcpucontext 43
++struct xen_domctl_ext_vcpucontext {
++ /* IN: VCPU that this call applies to. */
++ uint32_t vcpu;
++ /*
++ * SET: Size of struct (IN)
++ * GET: Size of struct (OUT)
++ */
++ uint32_t size;
++#if defined(__i386__) || defined(__x86_64__)
++ /* SYSCALL from 32-bit mode and SYSENTER callback information. */
++ /* NB. SYSCALL from 64-bit mode is contained in vcpu_guest_context_t */
++ uint64_aligned_t syscall32_callback_eip;
++ uint64_aligned_t sysenter_callback_eip;
++ uint16_t syscall32_callback_cs;
++ uint16_t sysenter_callback_cs;
++ uint8_t syscall32_disables_events;
++ uint8_t sysenter_disables_events;
++#endif
++};
++typedef struct xen_domctl_ext_vcpucontext xen_domctl_ext_vcpucontext_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_ext_vcpucontext_t);
+
-+/* Xencomm macros. */
-+#define XENCOMM_INLINE_MASK 0xf800000000000000UL
-+#define XENCOMM_INLINE_FLAG 0x8000000000000000UL
++/*
++ * Set optimizaton features for a domain
++ */
++#define XEN_DOMCTL_set_opt_feature 44
++struct xen_domctl_set_opt_feature {
++#if defined(__ia64__)
++ struct xen_ia64_opt_feature optf;
++#else
++ /* Make struct non-empty: do not depend on this field name! */
++ uint64_t dummy;
++#endif
++};
++typedef struct xen_domctl_set_opt_feature xen_domctl_set_opt_feature_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_set_opt_feature_t);
+
-+#define XENCOMM_IS_INLINE(addr) \
-+ (((unsigned long)(addr) & XENCOMM_INLINE_MASK) == XENCOMM_INLINE_FLAG)
-+#define XENCOMM_INLINE_ADDR(addr) \
-+ ((unsigned long)(addr) & ~XENCOMM_INLINE_MASK)
++/*
++ * Set the target domain for a domain
++ */
++#define XEN_DOMCTL_set_target 46
++struct xen_domctl_set_target {
++ domid_t target;
++};
++typedef struct xen_domctl_set_target xen_domctl_set_target_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_set_target_t);
+
-+/* xen perfmon */
-+#ifdef XEN
-+#ifndef __ASSEMBLY__
-+#ifndef _ASM_IA64_PERFMON_H
++#if defined(__i386__) || defined(__x86_64__)
++# define XEN_CPUID_INPUT_UNUSED 0xFFFFFFFF
++# define XEN_DOMCTL_set_cpuid 49
++struct xen_domctl_cpuid {
++ unsigned int input[2];
++ unsigned int eax;
++ unsigned int ebx;
++ unsigned int ecx;
++ unsigned int edx;
++};
++typedef struct xen_domctl_cpuid xen_domctl_cpuid_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_cpuid_t);
++#endif
+
-+#include <xen/list.h> // asm/perfmon.h requires struct list_head
-+#include <asm/perfmon.h>
-+// for PFM_xxx and pfarg_features_t, pfarg_context_t, pfarg_reg_t, pfarg_load_t
++#define XEN_DOMCTL_subscribe 29
++struct xen_domctl_subscribe {
++ uint32_t port; /* IN */
++};
++typedef struct xen_domctl_subscribe xen_domctl_subscribe_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_subscribe_t);
+
-+#endif /* _ASM_IA64_PERFMON_H */
++/*
++ * Define the maximum machine address size which should be allocated
++ * to a guest.
++ */
++#define XEN_DOMCTL_set_machine_address_size 51
++#define XEN_DOMCTL_get_machine_address_size 52
+
-+DEFINE_XEN_GUEST_HANDLE(pfarg_features_t);
-+DEFINE_XEN_GUEST_HANDLE(pfarg_context_t);
-+DEFINE_XEN_GUEST_HANDLE(pfarg_reg_t);
-+DEFINE_XEN_GUEST_HANDLE(pfarg_load_t);
-+#endif /* __ASSEMBLY__ */
-+#endif /* XEN */
+
-+#endif /* __HYPERVISOR_IF_IA64_H__ */
++struct xen_domctl {
++ uint32_t cmd;
++ uint32_t interface_version; /* XEN_DOMCTL_INTERFACE_VERSION */
++ domid_t domain;
++ union {
++ struct xen_domctl_createdomain createdomain;
++ struct xen_domctl_getdomaininfo getdomaininfo;
++ struct xen_domctl_getmemlist getmemlist;
++ struct xen_domctl_getpageframeinfo getpageframeinfo;
++ struct xen_domctl_getpageframeinfo2 getpageframeinfo2;
++ struct xen_domctl_vcpuaffinity vcpuaffinity;
++ struct xen_domctl_shadow_op shadow_op;
++ struct xen_domctl_max_mem max_mem;
++ struct xen_domctl_vcpucontext vcpucontext;
++ struct xen_domctl_getvcpuinfo getvcpuinfo;
++ struct xen_domctl_max_vcpus max_vcpus;
++ struct xen_domctl_scheduler_op scheduler_op;
++ struct xen_domctl_setdomainhandle setdomainhandle;
++ struct xen_domctl_setdebugging setdebugging;
++ struct xen_domctl_irq_permission irq_permission;
++ struct xen_domctl_iomem_permission iomem_permission;
++ struct xen_domctl_ioport_permission ioport_permission;
++ struct xen_domctl_hypercall_init hypercall_init;
++ struct xen_domctl_arch_setup arch_setup;
++ struct xen_domctl_settimeoffset settimeoffset;
++ struct xen_domctl_real_mode_area real_mode_area;
++ struct xen_domctl_hvmcontext hvmcontext;
++ struct xen_domctl_address_size address_size;
++ struct xen_domctl_sendtrigger sendtrigger;
++ struct xen_domctl_get_device_group get_device_group;
++ struct xen_domctl_assign_device assign_device;
++ struct xen_domctl_bind_pt_irq bind_pt_irq;
++ struct xen_domctl_memory_mapping memory_mapping;
++ struct xen_domctl_ioport_mapping ioport_mapping;
++ struct xen_domctl_pin_mem_cacheattr pin_mem_cacheattr;
++ struct xen_domctl_ext_vcpucontext ext_vcpucontext;
++ struct xen_domctl_set_opt_feature set_opt_feature;
++ struct xen_domctl_set_target set_target;
++ struct xen_domctl_subscribe subscribe;
++#if defined(__i386__) || defined(__x86_64__)
++ struct xen_domctl_cpuid cpuid;
++#endif
++ uint8_t pad[128];
++ } u;
++};
++typedef struct xen_domctl xen_domctl_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_t);
++
++#endif /* __XEN_PUBLIC_DOMCTL_H__ */
+
+/*
+ * Local variables:
@@ -96833,11 +136093,15 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/arch-powerpc.h tmp-linux-2.6-xen.patch/include/xen/interface/arch-powerpc.h
---- pristine-linux-2.6.18.2/include/xen/interface/arch-powerpc.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/arch-powerpc.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,125 @@
-+/*
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/elfnote.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/elfnote.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,233 @@
++/******************************************************************************
++ * elfnote.h
++ *
++ * Definitions used for the Xen ELF notes.
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -96856,311 +136120,206 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (C) IBM Corp. 2005, 2006
-+ *
-+ * Authors: Hollis Blanchard <hollisb at us.ibm.com>
++ * Copyright (c) 2006, Ian Campbell, XenSource Ltd.
+ */
+
-+#ifndef __XEN_PUBLIC_ARCH_PPC_64_H__
-+#define __XEN_PUBLIC_ARCH_PPC_64_H__
-+
-+#define __DEFINE_XEN_GUEST_HANDLE(name, type) \
-+ typedef struct { \
-+ int __pad[(sizeof (long long) - sizeof (void *)) / sizeof (int)]; \
-+ type *p; \
-+ } __attribute__((__aligned__(8))) __guest_handle_ ## name
-+
-+#define DEFINE_XEN_GUEST_HANDLE(name) __DEFINE_XEN_GUEST_HANDLE(name, name)
-+#define XEN_GUEST_HANDLE(name) __guest_handle_ ## name
-+#define set_xen_guest_handle(hnd, val) \
-+ do { \
-+ if (sizeof ((hnd).__pad)) \
-+ (hnd).__pad[0] = 0; \
-+ (hnd).p = val; \
-+ } while (0)
-+
-+#ifdef __XEN_TOOLS__
-+#define get_xen_guest_handle(val, hnd) do { val = (hnd).p; } while (0)
-+#endif
-+
-+#ifndef __ASSEMBLY__
-+/* Guest handles for primitive C types. */
-+__DEFINE_XEN_GUEST_HANDLE(uchar, unsigned char);
-+__DEFINE_XEN_GUEST_HANDLE(uint, unsigned int);
-+__DEFINE_XEN_GUEST_HANDLE(ulong, unsigned long);
-+DEFINE_XEN_GUEST_HANDLE(char);
-+DEFINE_XEN_GUEST_HANDLE(int);
-+DEFINE_XEN_GUEST_HANDLE(long);
-+DEFINE_XEN_GUEST_HANDLE(void);
-+
-+typedef unsigned long long xen_pfn_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_pfn_t);
-+#define PRI_xen_pfn "llx"
-+#endif
++#ifndef __XEN_PUBLIC_ELFNOTE_H__
++#define __XEN_PUBLIC_ELFNOTE_H__
+
+/*
-+ * Pointers and other address fields inside interface structures are padded to
-+ * 64 bits. This means that field alignments aren't different between 32- and
-+ * 64-bit architectures.
++ * The notes should live in a PT_NOTE segment and have "Xen" in the
++ * name field.
++ *
++ * Numeric types are either 4 or 8 bytes depending on the content of
++ * the desc field.
++ *
++ * LEGACY indicated the fields in the legacy __xen_guest string which
++ * this a note type replaces.
+ */
-+/* NB. Multi-level macro ensures __LINE__ is expanded before concatenation. */
-+#define __MEMORY_PADDING(_X)
-+#define _MEMORY_PADDING(_X) __MEMORY_PADDING(_X)
-+#define MEMORY_PADDING _MEMORY_PADDING(__LINE__)
-+
-+/* And the trap vector is... */
-+#define TRAP_INSTR "li 0,-1; sc" /* XXX just "sc"? */
-+
-+#ifndef __ASSEMBLY__
-+
-+#define XENCOMM_INLINE_FLAG (1UL << 63)
-+
-+typedef uint64_t xen_ulong_t;
-+
-+/* User-accessible registers: nost of these need to be saved/restored
-+ * for every nested Xen invocation. */
-+struct cpu_user_regs
-+{
-+ uint64_t gprs[32];
-+ uint64_t lr;
-+ uint64_t ctr;
-+ uint64_t srr0;
-+ uint64_t srr1;
-+ uint64_t pc;
-+ uint64_t msr;
-+ uint64_t fpscr; /* XXX Is this necessary */
-+ uint64_t xer;
-+ uint64_t hid4; /* debug only */
-+ uint64_t dar; /* debug only */
-+ uint32_t dsisr; /* debug only */
-+ uint32_t cr;
-+ uint32_t __pad; /* good spot for another 32bit reg */
-+ uint32_t entry_vector;
-+};
-+typedef struct cpu_user_regs cpu_user_regs_t;
+
-+typedef uint64_t tsc_timestamp_t; /* RDTSC timestamp */ /* XXX timebase */
-+
-+/* ONLY used to communicate with dom0! See also struct exec_domain. */
-+struct vcpu_guest_context {
-+ cpu_user_regs_t user_regs; /* User-level CPU registers */
-+ uint64_t sdr1; /* Pagetable base */
-+ /* XXX etc */
-+};
-+typedef struct vcpu_guest_context vcpu_guest_context_t;
-+DEFINE_XEN_GUEST_HANDLE(vcpu_guest_context_t);
++/*
++ * NAME=VALUE pair (string).
++ */
++#define XEN_ELFNOTE_INFO 0
+
-+struct arch_shared_info {
-+ uint64_t boot_timebase;
-+};
++/*
++ * The virtual address of the entry point (numeric).
++ *
++ * LEGACY: VIRT_ENTRY
++ */
++#define XEN_ELFNOTE_ENTRY 1
+
-+struct arch_vcpu_info {
-+};
++/* The virtual address of the hypercall transfer page (numeric).
++ *
++ * LEGACY: HYPERCALL_PAGE. (n.b. legacy value is a physical page
++ * number not a virtual address)
++ */
++#define XEN_ELFNOTE_HYPERCALL_PAGE 2
+
-+/* Support for multi-processor guests. */
-+#define MAX_VIRT_CPUS 32
-+#endif
++/* The virtual address where the kernel image should be mapped (numeric).
++ *
++ * Defaults to 0.
++ *
++ * LEGACY: VIRT_BASE
++ */
++#define XEN_ELFNOTE_VIRT_BASE 3
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/arch-x86/xen.h tmp-linux-2.6-xen.patch/include/xen/interface/arch-x86/xen.h
---- pristine-linux-2.6.18.2/include/xen/interface/arch-x86/xen.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/arch-x86/xen.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,205 @@
-+/******************************************************************************
-+ * arch-x86/xen.h
-+ *
-+ * Guest OS interface to x86 Xen.
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
++/*
++ * The offset of the ELF paddr field from the acutal required
++ * psuedo-physical address (numeric).
+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
++ * This is used to maintain backwards compatibility with older kernels
++ * which wrote __PAGE_OFFSET into that field. This field defaults to 0
++ * if not present.
+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
++ * LEGACY: ELF_PADDR_OFFSET. (n.b. legacy default is VIRT_BASE)
++ */
++#define XEN_ELFNOTE_PADDR_OFFSET 4
++
++/*
++ * The version of Xen that we work with (string).
+ *
-+ * Copyright (c) 2004-2006, K A Fraser
++ * LEGACY: XEN_VER
+ */
++#define XEN_ELFNOTE_XEN_VERSION 5
+
-+#ifndef __XEN_PUBLIC_ARCH_X86_XEN_H__
-+#define __XEN_PUBLIC_ARCH_X86_XEN_H__
++/*
++ * The name of the guest operating system (string).
++ *
++ * LEGACY: GUEST_OS
++ */
++#define XEN_ELFNOTE_GUEST_OS 6
+
-+/* Structural guest handles introduced in 0x00030201. */
-+#if __XEN_INTERFACE_VERSION__ >= 0x00030201
-+#define __DEFINE_XEN_GUEST_HANDLE(name, type) \
-+ typedef struct { type *p; } __guest_handle_ ## name
-+#else
-+#define __DEFINE_XEN_GUEST_HANDLE(name, type) \
-+ typedef type * __guest_handle_ ## name
-+#endif
++/*
++ * The version of the guest operating system (string).
++ *
++ * LEGACY: GUEST_VER
++ */
++#define XEN_ELFNOTE_GUEST_VERSION 7
+
-+#define DEFINE_XEN_GUEST_HANDLE(name) __DEFINE_XEN_GUEST_HANDLE(name, name)
-+#define __XEN_GUEST_HANDLE(name) __guest_handle_ ## name
-+#define XEN_GUEST_HANDLE(name) __XEN_GUEST_HANDLE(name)
-+#define set_xen_guest_handle(hnd, val) do { (hnd).p = val; } while (0)
-+#ifdef __XEN_TOOLS__
-+#define get_xen_guest_handle(val, hnd) do { val = (hnd).p; } while (0)
-+#endif
++/*
++ * The loader type (string).
++ *
++ * LEGACY: LOADER
++ */
++#define XEN_ELFNOTE_LOADER 8
+
-+#if defined(__i386__)
-+#include "xen-x86_32.h"
-+#elif defined(__x86_64__)
-+#include "xen-x86_64.h"
-+#endif
++/*
++ * The kernel supports PAE (x86/32 only, string = "yes", "no" or
++ * "bimodal").
++ *
++ * For compatibility with Xen 3.0.3 and earlier the "bimodal" setting
++ * may be given as "yes,bimodal" which will cause older Xen to treat
++ * this kernel as PAE.
++ *
++ * LEGACY: PAE (n.b. The legacy interface included a provision to
++ * indicate 'extended-cr3' support allowing L3 page tables to be
++ * placed above 4G. It is assumed that any kernel new enough to use
++ * these ELF notes will include this and therefore "yes" here is
++ * equivalent to "yes[entended-cr3]" in the __xen_guest interface.
++ */
++#define XEN_ELFNOTE_PAE_MODE 9
+
-+#ifndef __ASSEMBLY__
-+/* Guest handles for primitive C types. */
-+__DEFINE_XEN_GUEST_HANDLE(uchar, unsigned char);
-+__DEFINE_XEN_GUEST_HANDLE(uint, unsigned int);
-+__DEFINE_XEN_GUEST_HANDLE(ulong, unsigned long);
-+DEFINE_XEN_GUEST_HANDLE(char);
-+DEFINE_XEN_GUEST_HANDLE(int);
-+DEFINE_XEN_GUEST_HANDLE(long);
-+DEFINE_XEN_GUEST_HANDLE(void);
++/*
++ * The features supported/required by this kernel (string).
++ *
++ * The string must consist of a list of feature names (as given in
++ * features.h, without the "XENFEAT_" prefix) separated by '|'
++ * characters. If a feature is required for the kernel to function
++ * then the feature name must be preceded by a '!' character.
++ *
++ * LEGACY: FEATURES
++ */
++#define XEN_ELFNOTE_FEATURES 10
+
-+typedef unsigned long xen_pfn_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_pfn_t);
-+#define PRI_xen_pfn "lx"
-+#endif
++/*
++ * The kernel requires the symbol table to be loaded (string = "yes" or "no")
++ * LEGACY: BSD_SYMTAB (n.b. The legacy treated the presence or absence
++ * of this string as a boolean flag rather than requiring "yes" or
++ * "no".
++ */
++#define XEN_ELFNOTE_BSD_SYMTAB 11
+
+/*
-+ * SEGMENT DESCRIPTOR TABLES
++ * The lowest address the hypervisor hole can begin at (numeric).
++ *
++ * This must not be set higher than HYPERVISOR_VIRT_START. Its presence
++ * also indicates to the hypervisor that the kernel can deal with the
++ * hole starting at a higher address.
+ */
++#define XEN_ELFNOTE_HV_START_LOW 12
++
+/*
-+ * A number of GDT entries are reserved by Xen. These are not situated at the
-+ * start of the GDT because some stupid OSes export hard-coded selector values
-+ * in their ABI. These hard-coded values are always near the start of the GDT,
-+ * so Xen places itself out of the way, at the far end of the GDT.
++ * List of maddr_t-sized mask/value pairs describing how to recognize
++ * (non-present) L1 page table entries carrying valid MFNs (numeric).
+ */
-+#define FIRST_RESERVED_GDT_PAGE 14
-+#define FIRST_RESERVED_GDT_BYTE (FIRST_RESERVED_GDT_PAGE * 4096)
-+#define FIRST_RESERVED_GDT_ENTRY (FIRST_RESERVED_GDT_BYTE / 8)
++#define XEN_ELFNOTE_L1_MFN_VALID 13
+
-+/* Maximum number of virtual CPUs in multi-processor guests. */
-+#define MAX_VIRT_CPUS 32
++/*
++ * Whether or not the guest supports cooperative suspend cancellation.
++ */
++#define XEN_ELFNOTE_SUSPEND_CANCEL 14
+
-+#ifndef __ASSEMBLY__
++/*
++ * The number of the highest elfnote defined.
++ */
++#define XEN_ELFNOTE_MAX XEN_ELFNOTE_SUSPEND_CANCEL
+
-+typedef unsigned long xen_ulong_t;
++/*
++ * System information exported through crash notes.
++ *
++ * The kexec / kdump code will create one XEN_ELFNOTE_CRASH_INFO
++ * note in case of a system crash. This note will contain various
++ * information about the system, see xen/include/xen/elfcore.h.
++ */
++#define XEN_ELFNOTE_CRASH_INFO 0x1000001
+
+/*
-+ * Send an array of these to HYPERVISOR_set_trap_table().
-+ * The privilege level specifies which modes may enter a trap via a software
-+ * interrupt. On x86/64, since rings 1 and 2 are unavailable, we allocate
-+ * privilege levels as follows:
-+ * Level == 0: Noone may enter
-+ * Level == 1: Kernel may enter
-+ * Level == 2: Kernel may enter
-+ * Level == 3: Everyone may enter
++ * System registers exported through crash notes.
++ *
++ * The kexec / kdump code will create one XEN_ELFNOTE_CRASH_REGS
++ * note per cpu in case of a system crash. This note is architecture
++ * specific and will contain registers not saved in the "CORE" note.
++ * See xen/include/xen/elfcore.h for more information.
+ */
-+#define TI_GET_DPL(_ti) ((_ti)->flags & 3)
-+#define TI_GET_IF(_ti) ((_ti)->flags & 4)
-+#define TI_SET_DPL(_ti,_dpl) ((_ti)->flags |= (_dpl))
-+#define TI_SET_IF(_ti,_if) ((_ti)->flags |= ((!!(_if))<<2))
-+struct trap_info {
-+ uint8_t vector; /* exception vector */
-+ uint8_t flags; /* 0-3: privilege level; 4: clear event enable? */
-+ uint16_t cs; /* code selector */
-+ unsigned long address; /* code offset */
-+};
-+typedef struct trap_info trap_info_t;
-+DEFINE_XEN_GUEST_HANDLE(trap_info_t);
++#define XEN_ELFNOTE_CRASH_REGS 0x1000002
+
-+typedef uint64_t tsc_timestamp_t; /* RDTSC timestamp */
+
+/*
-+ * The following is all CPU context. Note that the fpu_ctxt block is filled
-+ * in by FXSAVE if the CPU has feature FXSR; otherwise FSAVE is used.
++ * xen dump-core none note.
++ * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_NONE
++ * in its dump file to indicate that the file is xen dump-core
++ * file. This note doesn't have any other information.
++ * See tools/libxc/xc_core.h for more information.
+ */
-+struct vcpu_guest_context {
-+ /* FPU registers come first so they can be aligned for FXSAVE/FXRSTOR. */
-+ struct { char x[512]; } fpu_ctxt; /* User-level FPU registers */
-+#define VGCF_I387_VALID (1<<0)
-+#define VGCF_IN_KERNEL (1<<2)
-+#define _VGCF_i387_valid 0
-+#define VGCF_i387_valid (1<<_VGCF_i387_valid)
-+#define _VGCF_in_kernel 2
-+#define VGCF_in_kernel (1<<_VGCF_in_kernel)
-+#define _VGCF_failsafe_disables_events 3
-+#define VGCF_failsafe_disables_events (1<<_VGCF_failsafe_disables_events)
-+#define _VGCF_syscall_disables_events 4
-+#define VGCF_syscall_disables_events (1<<_VGCF_syscall_disables_events)
-+#define _VGCF_online 5
-+#define VGCF_online (1<<_VGCF_online)
-+ unsigned long flags; /* VGCF_* flags */
-+ struct cpu_user_regs user_regs; /* User-level CPU registers */
-+ struct trap_info trap_ctxt[256]; /* Virtual IDT */
-+ unsigned long ldt_base, ldt_ents; /* LDT (linear address, # ents) */
-+ unsigned long gdt_frames[16], gdt_ents; /* GDT (machine frames, # ents) */
-+ unsigned long kernel_ss, kernel_sp; /* Virtual TSS (only SS1/SP1) */
-+ /* NB. User pagetable on x86/64 is placed in ctrlreg[1]. */
-+ unsigned long ctrlreg[8]; /* CR0-CR7 (control registers) */
-+ unsigned long debugreg[8]; /* DB0-DB7 (debug registers) */
-+#ifdef __i386__
-+ unsigned long event_callback_cs; /* CS:EIP of event callback */
-+ unsigned long event_callback_eip;
-+ unsigned long failsafe_callback_cs; /* CS:EIP of failsafe callback */
-+ unsigned long failsafe_callback_eip;
-+#else
-+ unsigned long event_callback_eip;
-+ unsigned long failsafe_callback_eip;
-+#ifdef __XEN__
-+ union {
-+ unsigned long syscall_callback_eip;
-+ struct {
-+ unsigned int event_callback_cs; /* compat CS of event cb */
-+ unsigned int failsafe_callback_cs; /* compat CS of failsafe cb */
-+ };
-+ };
-+#else
-+ unsigned long syscall_callback_eip;
-+#endif
-+#endif
-+ unsigned long vm_assist; /* VMASST_TYPE_* bitmap */
-+#ifdef __x86_64__
-+ /* Segment base addresses. */
-+ uint64_t fs_base;
-+ uint64_t gs_base_kernel;
-+ uint64_t gs_base_user;
-+#endif
-+};
-+typedef struct vcpu_guest_context vcpu_guest_context_t;
-+DEFINE_XEN_GUEST_HANDLE(vcpu_guest_context_t);
++#define XEN_ELFNOTE_DUMPCORE_NONE 0x2000000
+
-+struct arch_shared_info {
-+ unsigned long max_pfn; /* max pfn that appears in table */
-+ /* Frame containing list of mfns containing list of mfns containing p2m. */
-+ xen_pfn_t pfn_to_mfn_frame_list_list;
-+ unsigned long nmi_reason;
-+ uint64_t pad[32];
-+};
-+typedef struct arch_shared_info arch_shared_info_t;
++/*
++ * xen dump-core header note.
++ * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_HEADER
++ * in its dump file.
++ * See tools/libxc/xc_core.h for more information.
++ */
++#define XEN_ELFNOTE_DUMPCORE_HEADER 0x2000001
+
-+#endif /* !__ASSEMBLY__ */
++/*
++ * xen dump-core xen version note.
++ * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_XEN_VERSION
++ * in its dump file. It contains the xen version obtained via the
++ * XENVER hypercall.
++ * See tools/libxc/xc_core.h for more information.
++ */
++#define XEN_ELFNOTE_DUMPCORE_XEN_VERSION 0x2000002
+
+/*
-+ * Prefix forces emulation of some non-trapping instructions.
-+ * Currently only CPUID.
++ * xen dump-core format version note.
++ * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_FORMAT_VERSION
++ * in its dump file. It contains a format version identifier.
++ * See tools/libxc/xc_core.h for more information.
+ */
-+#ifdef __ASSEMBLY__
-+#define XEN_EMULATE_PREFIX .byte 0x0f,0x0b,0x78,0x65,0x6e ;
-+#define XEN_CPUID XEN_EMULATE_PREFIX cpuid
-+#else
-+#define XEN_EMULATE_PREFIX ".byte 0x0f,0x0b,0x78,0x65,0x6e ; "
-+#define XEN_CPUID XEN_EMULATE_PREFIX "cpuid"
-+#endif
++#define XEN_ELFNOTE_DUMPCORE_FORMAT_VERSION 0x2000003
+
-+#endif /* __XEN_PUBLIC_ARCH_X86_XEN_H__ */
++#endif /* __XEN_PUBLIC_ELFNOTE_H__ */
+
+/*
+ * Local variables:
@@ -97171,448 +136330,545 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/arch-x86/xen-x86_32.h tmp-linux-2.6-xen.patch/include/xen/interface/arch-x86/xen-x86_32.h
---- pristine-linux-2.6.18.2/include/xen/interface/arch-x86/xen-x86_32.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/arch-x86/xen-x86_32.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,184 @@
-+/******************************************************************************
-+ * xen-x86_32.h
-+ *
-+ * Guest OS interface to x86 32-bit Xen.
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/elfstructs.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/elfstructs.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,527 @@
++#ifndef __XEN_PUBLIC_ELFSTRUCTS_H__
++#define __XEN_PUBLIC_ELFSTRUCTS_H__ 1
++/*
++ * Copyright (c) 1995, 1996 Erik Theisen. All rights reserved.
+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
++ * Redistribution and use in source and binary forms, with or without
++ * modification, are permitted provided that the following conditions
++ * are met:
++ * 1. Redistributions of source code must retain the above copyright
++ * notice, this list of conditions and the following disclaimer.
++ * 2. Redistributions in binary form must reproduce the above copyright
++ * notice, this list of conditions and the following disclaimer in the
++ * documentation and/or other materials provided with the distribution.
++ * 3. The name of the author may not be used to endorse or promote products
++ * derived from this software without specific prior written permission
+ *
-+ * Copyright (c) 2004-2007, K A Fraser
++ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
++ * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
++ * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
++ * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
++ * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
++ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
++ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
++ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
++ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
++ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
-+#ifndef __XEN_PUBLIC_ARCH_X86_XEN_X86_32_H__
-+#define __XEN_PUBLIC_ARCH_X86_XEN_X86_32_H__
++typedef uint8_t Elf_Byte;
+
-+/*
-+ * Hypercall interface:
-+ * Input: %ebx, %ecx, %edx, %esi, %edi (arguments 1-5)
-+ * Output: %eax
-+ * Access is via hypercall page (set up by guest loader or via a Xen MSR):
-+ * call hypercall_page + hypercall-number * 32
-+ * Clobbered: Argument registers (e.g., 2-arg hypercall clobbers %ebx,%ecx)
-+ */
++typedef uint32_t Elf32_Addr; /* Unsigned program address */
++typedef uint32_t Elf32_Off; /* Unsigned file offset */
++typedef int32_t Elf32_Sword; /* Signed large integer */
++typedef uint32_t Elf32_Word; /* Unsigned large integer */
++typedef uint16_t Elf32_Half; /* Unsigned medium integer */
++
++typedef uint64_t Elf64_Addr;
++typedef uint64_t Elf64_Off;
++typedef int32_t Elf64_Shalf;
++
++typedef int32_t Elf64_Sword;
++typedef uint32_t Elf64_Word;
++
++typedef int64_t Elf64_Sxword;
++typedef uint64_t Elf64_Xword;
++
++typedef uint32_t Elf64_Half;
++typedef uint16_t Elf64_Quarter;
+
-+#if __XEN_INTERFACE_VERSION__ < 0x00030203
+/*
-+ * Legacy hypercall interface:
-+ * As above, except the entry sequence to the hypervisor is:
-+ * mov $hypercall-number*32,%eax ; int $0x82
++ * e_ident[] identification indexes
++ * See http://www.caldera.com/developers/gabi/2000-07-17/ch4.eheader.html
+ */
-+#define TRAP_INSTR "int $0x82"
-+#endif
++#define EI_MAG0 0 /* file ID */
++#define EI_MAG1 1 /* file ID */
++#define EI_MAG2 2 /* file ID */
++#define EI_MAG3 3 /* file ID */
++#define EI_CLASS 4 /* file class */
++#define EI_DATA 5 /* data encoding */
++#define EI_VERSION 6 /* ELF header version */
++#define EI_OSABI 7 /* OS/ABI ID */
++#define EI_ABIVERSION 8 /* ABI version */
++#define EI_PAD 9 /* start of pad bytes */
++#define EI_NIDENT 16 /* Size of e_ident[] */
++
++/* e_ident[] magic number */
++#define ELFMAG0 0x7f /* e_ident[EI_MAG0] */
++#define ELFMAG1 'E' /* e_ident[EI_MAG1] */
++#define ELFMAG2 'L' /* e_ident[EI_MAG2] */
++#define ELFMAG3 'F' /* e_ident[EI_MAG3] */
++#define ELFMAG "\177ELF" /* magic */
++#define SELFMAG 4 /* size of magic */
++
++/* e_ident[] file class */
++#define ELFCLASSNONE 0 /* invalid */
++#define ELFCLASS32 1 /* 32-bit objs */
++#define ELFCLASS64 2 /* 64-bit objs */
++#define ELFCLASSNUM 3 /* number of classes */
++
++/* e_ident[] data encoding */
++#define ELFDATANONE 0 /* invalid */
++#define ELFDATA2LSB 1 /* Little-Endian */
++#define ELFDATA2MSB 2 /* Big-Endian */
++#define ELFDATANUM 3 /* number of data encode defines */
++
++/* e_ident[] Operating System/ABI */
++#define ELFOSABI_SYSV 0 /* UNIX System V ABI */
++#define ELFOSABI_HPUX 1 /* HP-UX operating system */
++#define ELFOSABI_NETBSD 2 /* NetBSD */
++#define ELFOSABI_LINUX 3 /* GNU/Linux */
++#define ELFOSABI_HURD 4 /* GNU/Hurd */
++#define ELFOSABI_86OPEN 5 /* 86Open common IA32 ABI */
++#define ELFOSABI_SOLARIS 6 /* Solaris */
++#define ELFOSABI_MONTEREY 7 /* Monterey */
++#define ELFOSABI_IRIX 8 /* IRIX */
++#define ELFOSABI_FREEBSD 9 /* FreeBSD */
++#define ELFOSABI_TRU64 10 /* TRU64 UNIX */
++#define ELFOSABI_MODESTO 11 /* Novell Modesto */
++#define ELFOSABI_OPENBSD 12 /* OpenBSD */
++#define ELFOSABI_ARM 97 /* ARM */
++#define ELFOSABI_STANDALONE 255 /* Standalone (embedded) application */
++
++/* e_ident */
++#define IS_ELF(ehdr) ((ehdr).e_ident[EI_MAG0] == ELFMAG0 && \
++ (ehdr).e_ident[EI_MAG1] == ELFMAG1 && \
++ (ehdr).e_ident[EI_MAG2] == ELFMAG2 && \
++ (ehdr).e_ident[EI_MAG3] == ELFMAG3)
++
++/* ELF Header */
++typedef struct elfhdr {
++ unsigned char e_ident[EI_NIDENT]; /* ELF Identification */
++ Elf32_Half e_type; /* object file type */
++ Elf32_Half e_machine; /* machine */
++ Elf32_Word e_version; /* object file version */
++ Elf32_Addr e_entry; /* virtual entry point */
++ Elf32_Off e_phoff; /* program header table offset */
++ Elf32_Off e_shoff; /* section header table offset */
++ Elf32_Word e_flags; /* processor-specific flags */
++ Elf32_Half e_ehsize; /* ELF header size */
++ Elf32_Half e_phentsize; /* program header entry size */
++ Elf32_Half e_phnum; /* number of program header entries */
++ Elf32_Half e_shentsize; /* section header entry size */
++ Elf32_Half e_shnum; /* number of section header entries */
++ Elf32_Half e_shstrndx; /* section header table's "section
++ header string table" entry offset */
++} Elf32_Ehdr;
++
++typedef struct {
++ unsigned char e_ident[EI_NIDENT]; /* Id bytes */
++ Elf64_Quarter e_type; /* file type */
++ Elf64_Quarter e_machine; /* machine type */
++ Elf64_Half e_version; /* version number */
++ Elf64_Addr e_entry; /* entry point */
++ Elf64_Off e_phoff; /* Program hdr offset */
++ Elf64_Off e_shoff; /* Section hdr offset */
++ Elf64_Half e_flags; /* Processor flags */
++ Elf64_Quarter e_ehsize; /* sizeof ehdr */
++ Elf64_Quarter e_phentsize; /* Program header entry size */
++ Elf64_Quarter e_phnum; /* Number of program headers */
++ Elf64_Quarter e_shentsize; /* Section header entry size */
++ Elf64_Quarter e_shnum; /* Number of section headers */
++ Elf64_Quarter e_shstrndx; /* String table index */
++} Elf64_Ehdr;
++
++/* e_type */
++#define ET_NONE 0 /* No file type */
++#define ET_REL 1 /* relocatable file */
++#define ET_EXEC 2 /* executable file */
++#define ET_DYN 3 /* shared object file */
++#define ET_CORE 4 /* core file */
++#define ET_NUM 5 /* number of types */
++#define ET_LOPROC 0xff00 /* reserved range for processor */
++#define ET_HIPROC 0xffff /* specific e_type */
+
++/* e_machine */
++#define EM_NONE 0 /* No Machine */
++#define EM_M32 1 /* AT&T WE 32100 */
++#define EM_SPARC 2 /* SPARC */
++#define EM_386 3 /* Intel 80386 */
++#define EM_68K 4 /* Motorola 68000 */
++#define EM_88K 5 /* Motorola 88000 */
++#define EM_486 6 /* Intel 80486 - unused? */
++#define EM_860 7 /* Intel 80860 */
++#define EM_MIPS 8 /* MIPS R3000 Big-Endian only */
+/*
-+ * These flat segments are in the Xen-private section of every GDT. Since these
-+ * are also present in the initial GDT, many OSes will be able to avoid
-+ * installing their own GDT.
++ * Don't know if EM_MIPS_RS4_BE,
++ * EM_SPARC64, EM_PARISC,
++ * or EM_PPC are ABI compliant
+ */
-+#define FLAT_RING1_CS 0xe019 /* GDT index 259 */
-+#define FLAT_RING1_DS 0xe021 /* GDT index 260 */
-+#define FLAT_RING1_SS 0xe021 /* GDT index 260 */
-+#define FLAT_RING3_CS 0xe02b /* GDT index 261 */
-+#define FLAT_RING3_DS 0xe033 /* GDT index 262 */
-+#define FLAT_RING3_SS 0xe033 /* GDT index 262 */
++#define EM_MIPS_RS4_BE 10 /* MIPS R4000 Big-Endian */
++#define EM_SPARC64 11 /* SPARC v9 64-bit unoffical */
++#define EM_PARISC 15 /* HPPA */
++#define EM_SPARC32PLUS 18 /* Enhanced instruction set SPARC */
++#define EM_PPC 20 /* PowerPC */
++#define EM_PPC64 21 /* PowerPC 64-bit */
++#define EM_ARM 40 /* Advanced RISC Machines ARM */
++#define EM_ALPHA 41 /* DEC ALPHA */
++#define EM_SPARCV9 43 /* SPARC version 9 */
++#define EM_ALPHA_EXP 0x9026 /* DEC ALPHA */
++#define EM_IA_64 50 /* Intel Merced */
++#define EM_X86_64 62 /* AMD x86-64 architecture */
++#define EM_VAX 75 /* DEC VAX */
+
-+#define FLAT_KERNEL_CS FLAT_RING1_CS
-+#define FLAT_KERNEL_DS FLAT_RING1_DS
-+#define FLAT_KERNEL_SS FLAT_RING1_SS
-+#define FLAT_USER_CS FLAT_RING3_CS
-+#define FLAT_USER_DS FLAT_RING3_DS
-+#define FLAT_USER_SS FLAT_RING3_SS
++/* Version */
++#define EV_NONE 0 /* Invalid */
++#define EV_CURRENT 1 /* Current */
++#define EV_NUM 2 /* number of versions */
+
-+#define __HYPERVISOR_VIRT_START_PAE 0xF5800000
-+#define __MACH2PHYS_VIRT_START_PAE 0xF5800000
-+#define __MACH2PHYS_VIRT_END_PAE 0xF6800000
-+#define HYPERVISOR_VIRT_START_PAE \
-+ mk_unsigned_long(__HYPERVISOR_VIRT_START_PAE)
-+#define MACH2PHYS_VIRT_START_PAE \
-+ mk_unsigned_long(__MACH2PHYS_VIRT_START_PAE)
-+#define MACH2PHYS_VIRT_END_PAE \
-+ mk_unsigned_long(__MACH2PHYS_VIRT_END_PAE)
++/* Section Header */
++typedef struct {
++ Elf32_Word sh_name; /* name - index into section header
++ string table section */
++ Elf32_Word sh_type; /* type */
++ Elf32_Word sh_flags; /* flags */
++ Elf32_Addr sh_addr; /* address */
++ Elf32_Off sh_offset; /* file offset */
++ Elf32_Word sh_size; /* section size */
++ Elf32_Word sh_link; /* section header table index link */
++ Elf32_Word sh_info; /* extra information */
++ Elf32_Word sh_addralign; /* address alignment */
++ Elf32_Word sh_entsize; /* section entry size */
++} Elf32_Shdr;
+
-+#define __HYPERVISOR_VIRT_START_NONPAE 0xFC000000
-+#define __MACH2PHYS_VIRT_START_NONPAE 0xFC000000
-+#define __MACH2PHYS_VIRT_END_NONPAE 0xFC400000
-+#define HYPERVISOR_VIRT_START_NONPAE \
-+ mk_unsigned_long(__HYPERVISOR_VIRT_START_NONPAE)
-+#define MACH2PHYS_VIRT_START_NONPAE \
-+ mk_unsigned_long(__MACH2PHYS_VIRT_START_NONPAE)
-+#define MACH2PHYS_VIRT_END_NONPAE \
-+ mk_unsigned_long(__MACH2PHYS_VIRT_END_NONPAE)
++typedef struct {
++ Elf64_Half sh_name; /* section name */
++ Elf64_Half sh_type; /* section type */
++ Elf64_Xword sh_flags; /* section flags */
++ Elf64_Addr sh_addr; /* virtual address */
++ Elf64_Off sh_offset; /* file offset */
++ Elf64_Xword sh_size; /* section size */
++ Elf64_Half sh_link; /* link to another */
++ Elf64_Half sh_info; /* misc info */
++ Elf64_Xword sh_addralign; /* memory alignment */
++ Elf64_Xword sh_entsize; /* table entry size */
++} Elf64_Shdr;
+
-+#ifdef CONFIG_X86_PAE
-+#define __HYPERVISOR_VIRT_START __HYPERVISOR_VIRT_START_PAE
-+#define __MACH2PHYS_VIRT_START __MACH2PHYS_VIRT_START_PAE
-+#define __MACH2PHYS_VIRT_END __MACH2PHYS_VIRT_END_PAE
-+#else
-+#define __HYPERVISOR_VIRT_START __HYPERVISOR_VIRT_START_NONPAE
-+#define __MACH2PHYS_VIRT_START __MACH2PHYS_VIRT_START_NONPAE
-+#define __MACH2PHYS_VIRT_END __MACH2PHYS_VIRT_END_NONPAE
-+#endif
++/* Special Section Indexes */
++#define SHN_UNDEF 0 /* undefined */
++#define SHN_LORESERVE 0xff00 /* lower bounds of reserved indexes */
++#define SHN_LOPROC 0xff00 /* reserved range for processor */
++#define SHN_HIPROC 0xff1f /* specific section indexes */
++#define SHN_ABS 0xfff1 /* absolute value */
++#define SHN_COMMON 0xfff2 /* common symbol */
++#define SHN_HIRESERVE 0xffff /* upper bounds of reserved indexes */
++
++/* sh_type */
++#define SHT_NULL 0 /* inactive */
++#define SHT_PROGBITS 1 /* program defined information */
++#define SHT_SYMTAB 2 /* symbol table section */
++#define SHT_STRTAB 3 /* string table section */
++#define SHT_RELA 4 /* relocation section with addends*/
++#define SHT_HASH 5 /* symbol hash table section */
++#define SHT_DYNAMIC 6 /* dynamic section */
++#define SHT_NOTE 7 /* note section */
++#define SHT_NOBITS 8 /* no space section */
++#define SHT_REL 9 /* relation section without addends */
++#define SHT_SHLIB 10 /* reserved - purpose unknown */
++#define SHT_DYNSYM 11 /* dynamic symbol table section */
++#define SHT_NUM 12 /* number of section types */
++#define SHT_LOPROC 0x70000000 /* reserved range for processor */
++#define SHT_HIPROC 0x7fffffff /* specific section header types */
++#define SHT_LOUSER 0x80000000 /* reserved range for application */
++#define SHT_HIUSER 0xffffffff /* specific indexes */
++
++/* Section names */
++#define ELF_BSS ".bss" /* uninitialized data */
++#define ELF_DATA ".data" /* initialized data */
++#define ELF_DEBUG ".debug" /* debug */
++#define ELF_DYNAMIC ".dynamic" /* dynamic linking information */
++#define ELF_DYNSTR ".dynstr" /* dynamic string table */
++#define ELF_DYNSYM ".dynsym" /* dynamic symbol table */
++#define ELF_FINI ".fini" /* termination code */
++#define ELF_GOT ".got" /* global offset table */
++#define ELF_HASH ".hash" /* symbol hash table */
++#define ELF_INIT ".init" /* initialization code */
++#define ELF_REL_DATA ".rel.data" /* relocation data */
++#define ELF_REL_FINI ".rel.fini" /* relocation termination code */
++#define ELF_REL_INIT ".rel.init" /* relocation initialization code */
++#define ELF_REL_DYN ".rel.dyn" /* relocaltion dynamic link info */
++#define ELF_REL_RODATA ".rel.rodata" /* relocation read-only data */
++#define ELF_REL_TEXT ".rel.text" /* relocation code */
++#define ELF_RODATA ".rodata" /* read-only data */
++#define ELF_SHSTRTAB ".shstrtab" /* section header string table */
++#define ELF_STRTAB ".strtab" /* string table */
++#define ELF_SYMTAB ".symtab" /* symbol table */
++#define ELF_TEXT ".text" /* code */
+
-+#ifndef HYPERVISOR_VIRT_START
-+#define HYPERVISOR_VIRT_START mk_unsigned_long(__HYPERVISOR_VIRT_START)
-+#endif
+
-+#define MACH2PHYS_VIRT_START mk_unsigned_long(__MACH2PHYS_VIRT_START)
-+#define MACH2PHYS_VIRT_END mk_unsigned_long(__MACH2PHYS_VIRT_END)
-+#define MACH2PHYS_NR_ENTRIES ((MACH2PHYS_VIRT_END-MACH2PHYS_VIRT_START)>>2)
-+#ifndef machine_to_phys_mapping
-+#define machine_to_phys_mapping ((unsigned long *)MACH2PHYS_VIRT_START)
-+#endif
++/* Section Attribute Flags - sh_flags */
++#define SHF_WRITE 0x1 /* Writable */
++#define SHF_ALLOC 0x2 /* occupies memory */
++#define SHF_EXECINSTR 0x4 /* executable */
++#define SHF_MASKPROC 0xf0000000 /* reserved bits for processor */
++ /* specific section attributes */
+
-+/* 32-/64-bit invariability for control interfaces (domctl/sysctl). */
-+#if defined(__XEN__) || defined(__XEN_TOOLS__)
-+#undef __DEFINE_XEN_GUEST_HANDLE
-+#define __DEFINE_XEN_GUEST_HANDLE(name, type) \
-+ typedef struct { type *p; } \
-+ __guest_handle_ ## name; \
-+ typedef struct { union { type *p; uint64_aligned_t q; }; } \
-+ __guest_handle_64_ ## name
-+#undef set_xen_guest_handle
-+#define set_xen_guest_handle(hnd, val) \
-+ do { if ( sizeof(hnd) == 8 ) *(uint64_t *)&(hnd) = 0; \
-+ (hnd).p = val; \
-+ } while ( 0 )
-+#define uint64_aligned_t uint64_t __attribute__((aligned(8)))
-+#define XEN_GUEST_HANDLE_64(name) __guest_handle_64_ ## name
-+#endif
++/* Symbol Table Entry */
++typedef struct elf32_sym {
++ Elf32_Word st_name; /* name - index into string table */
++ Elf32_Addr st_value; /* symbol value */
++ Elf32_Word st_size; /* symbol size */
++ unsigned char st_info; /* type and binding */
++ unsigned char st_other; /* 0 - no defined meaning */
++ Elf32_Half st_shndx; /* section header index */
++} Elf32_Sym;
+
-+#ifndef __ASSEMBLY__
++typedef struct {
++ Elf64_Half st_name; /* Symbol name index in str table */
++ Elf_Byte st_info; /* type / binding attrs */
++ Elf_Byte st_other; /* unused */
++ Elf64_Quarter st_shndx; /* section index of symbol */
++ Elf64_Xword st_value; /* value of symbol */
++ Elf64_Xword st_size; /* size of symbol */
++} Elf64_Sym;
+
-+struct cpu_user_regs {
-+ uint32_t ebx;
-+ uint32_t ecx;
-+ uint32_t edx;
-+ uint32_t esi;
-+ uint32_t edi;
-+ uint32_t ebp;
-+ uint32_t eax;
-+ uint16_t error_code; /* private */
-+ uint16_t entry_vector; /* private */
-+ uint32_t eip;
-+ uint16_t cs;
-+ uint8_t saved_upcall_mask;
-+ uint8_t _pad0;
-+ uint32_t eflags; /* eflags.IF == !saved_upcall_mask */
-+ uint32_t esp;
-+ uint16_t ss, _pad1;
-+ uint16_t es, _pad2;
-+ uint16_t ds, _pad3;
-+ uint16_t fs, _pad4;
-+ uint16_t gs, _pad5;
-+};
-+typedef struct cpu_user_regs cpu_user_regs_t;
-+DEFINE_XEN_GUEST_HANDLE(cpu_user_regs_t);
++/* Symbol table index */
++#define STN_UNDEF 0 /* undefined */
+
-+/*
-+ * Page-directory addresses above 4GB do not fit into architectural %cr3.
-+ * When accessing %cr3, or equivalent field in vcpu_guest_context, guests
-+ * must use the following accessor macros to pack/unpack valid MFNs.
-+ */
-+#define xen_pfn_to_cr3(pfn) (((unsigned)(pfn) << 12) | ((unsigned)(pfn) >> 20))
-+#define xen_cr3_to_pfn(cr3) (((unsigned)(cr3) >> 12) | ((unsigned)(cr3) << 20))
++/* Extract symbol info - st_info */
++#define ELF32_ST_BIND(x) ((x) >> 4)
++#define ELF32_ST_TYPE(x) (((unsigned int) x) & 0xf)
++#define ELF32_ST_INFO(b,t) (((b) << 4) + ((t) & 0xf))
+
-+struct arch_vcpu_info {
-+ unsigned long cr2;
-+ unsigned long pad[5]; /* sizeof(vcpu_info_t) == 64 */
-+};
-+typedef struct arch_vcpu_info arch_vcpu_info_t;
++#define ELF64_ST_BIND(x) ((x) >> 4)
++#define ELF64_ST_TYPE(x) (((unsigned int) x) & 0xf)
++#define ELF64_ST_INFO(b,t) (((b) << 4) + ((t) & 0xf))
+
-+struct xen_callback {
-+ unsigned long cs;
-+ unsigned long eip;
-+};
-+typedef struct xen_callback xen_callback_t;
++/* Symbol Binding - ELF32_ST_BIND - st_info */
++#define STB_LOCAL 0 /* Local symbol */
++#define STB_GLOBAL 1 /* Global symbol */
++#define STB_WEAK 2 /* like global - lower precedence */
++#define STB_NUM 3 /* number of symbol bindings */
++#define STB_LOPROC 13 /* reserved range for processor */
++#define STB_HIPROC 15 /* specific symbol bindings */
+
-+#endif /* !__ASSEMBLY__ */
++/* Symbol type - ELF32_ST_TYPE - st_info */
++#define STT_NOTYPE 0 /* not specified */
++#define STT_OBJECT 1 /* data object */
++#define STT_FUNC 2 /* function */
++#define STT_SECTION 3 /* section */
++#define STT_FILE 4 /* file */
++#define STT_NUM 5 /* number of symbol types */
++#define STT_LOPROC 13 /* reserved range for processor */
++#define STT_HIPROC 15 /* specific symbol types */
+
-+#endif /* __XEN_PUBLIC_ARCH_X86_XEN_X86_32_H__ */
++/* Relocation entry with implicit addend */
++typedef struct {
++ Elf32_Addr r_offset; /* offset of relocation */
++ Elf32_Word r_info; /* symbol table index and type */
++} Elf32_Rel;
+
-+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
-+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/arch-x86/xen-x86_64.h tmp-linux-2.6-xen.patch/include/xen/interface/arch-x86/xen-x86_64.h
---- pristine-linux-2.6.18.2/include/xen/interface/arch-x86/xen-x86_64.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/arch-x86/xen-x86_64.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,211 @@
-+/******************************************************************************
-+ * xen-x86_64.h
-+ *
-+ * Guest OS interface to x86 64-bit Xen.
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Copyright (c) 2004-2006, K A Fraser
-+ */
++/* Relocation entry with explicit addend */
++typedef struct {
++ Elf32_Addr r_offset; /* offset of relocation */
++ Elf32_Word r_info; /* symbol table index and type */
++ Elf32_Sword r_addend;
++} Elf32_Rela;
+
-+#ifndef __XEN_PUBLIC_ARCH_X86_XEN_X86_64_H__
-+#define __XEN_PUBLIC_ARCH_X86_XEN_X86_64_H__
++/* Extract relocation info - r_info */
++#define ELF32_R_SYM(i) ((i) >> 8)
++#define ELF32_R_TYPE(i) ((unsigned char) (i))
++#define ELF32_R_INFO(s,t) (((s) << 8) + (unsigned char)(t))
+
-+/*
-+ * Hypercall interface:
-+ * Input: %rdi, %rsi, %rdx, %r10, %r8 (arguments 1-5)
-+ * Output: %rax
-+ * Access is via hypercall page (set up by guest loader or via a Xen MSR):
-+ * call hypercall_page + hypercall-number * 32
-+ * Clobbered: argument registers (e.g., 2-arg hypercall clobbers %rdi,%rsi)
-+ */
++typedef struct {
++ Elf64_Xword r_offset; /* where to do it */
++ Elf64_Xword r_info; /* index & type of relocation */
++} Elf64_Rel;
+
-+#if __XEN_INTERFACE_VERSION__ < 0x00030203
-+/*
-+ * Legacy hypercall interface:
-+ * As above, except the entry sequence to the hypervisor is:
-+ * mov $hypercall-number*32,%eax ; syscall
-+ * Clobbered: %rcx, %r11, argument registers (as above)
-+ */
-+#define TRAP_INSTR "syscall"
-+#endif
++typedef struct {
++ Elf64_Xword r_offset; /* where to do it */
++ Elf64_Xword r_info; /* index & type of relocation */
++ Elf64_Sxword r_addend; /* adjustment value */
++} Elf64_Rela;
+
-+/*
-+ * 64-bit segment selectors
-+ * These flat segments are in the Xen-private section of every GDT. Since these
-+ * are also present in the initial GDT, many OSes will be able to avoid
-+ * installing their own GDT.
-+ */
++#define ELF64_R_SYM(info) ((info) >> 32)
++#define ELF64_R_TYPE(info) ((info) & 0xFFFFFFFF)
++#define ELF64_R_INFO(s,t) (((s) << 32) + (u_int32_t)(t))
+
-+#define FLAT_RING3_CS32 0xe023 /* GDT index 260 */
-+#define FLAT_RING3_CS64 0xe033 /* GDT index 261 */
-+#define FLAT_RING3_DS32 0xe02b /* GDT index 262 */
-+#define FLAT_RING3_DS64 0x0000 /* NULL selector */
-+#define FLAT_RING3_SS32 0xe02b /* GDT index 262 */
-+#define FLAT_RING3_SS64 0xe02b /* GDT index 262 */
++/* Program Header */
++typedef struct {
++ Elf32_Word p_type; /* segment type */
++ Elf32_Off p_offset; /* segment offset */
++ Elf32_Addr p_vaddr; /* virtual address of segment */
++ Elf32_Addr p_paddr; /* physical address - ignored? */
++ Elf32_Word p_filesz; /* number of bytes in file for seg. */
++ Elf32_Word p_memsz; /* number of bytes in mem. for seg. */
++ Elf32_Word p_flags; /* flags */
++ Elf32_Word p_align; /* memory alignment */
++} Elf32_Phdr;
+
-+#define FLAT_KERNEL_DS64 FLAT_RING3_DS64
-+#define FLAT_KERNEL_DS32 FLAT_RING3_DS32
-+#define FLAT_KERNEL_DS FLAT_KERNEL_DS64
-+#define FLAT_KERNEL_CS64 FLAT_RING3_CS64
-+#define FLAT_KERNEL_CS32 FLAT_RING3_CS32
-+#define FLAT_KERNEL_CS FLAT_KERNEL_CS64
-+#define FLAT_KERNEL_SS64 FLAT_RING3_SS64
-+#define FLAT_KERNEL_SS32 FLAT_RING3_SS32
-+#define FLAT_KERNEL_SS FLAT_KERNEL_SS64
++typedef struct {
++ Elf64_Half p_type; /* entry type */
++ Elf64_Half p_flags; /* flags */
++ Elf64_Off p_offset; /* offset */
++ Elf64_Addr p_vaddr; /* virtual address */
++ Elf64_Addr p_paddr; /* physical address */
++ Elf64_Xword p_filesz; /* file size */
++ Elf64_Xword p_memsz; /* memory size */
++ Elf64_Xword p_align; /* memory & file alignment */
++} Elf64_Phdr;
+
-+#define FLAT_USER_DS64 FLAT_RING3_DS64
-+#define FLAT_USER_DS32 FLAT_RING3_DS32
-+#define FLAT_USER_DS FLAT_USER_DS64
-+#define FLAT_USER_CS64 FLAT_RING3_CS64
-+#define FLAT_USER_CS32 FLAT_RING3_CS32
-+#define FLAT_USER_CS FLAT_USER_CS64
-+#define FLAT_USER_SS64 FLAT_RING3_SS64
-+#define FLAT_USER_SS32 FLAT_RING3_SS32
-+#define FLAT_USER_SS FLAT_USER_SS64
++/* Segment types - p_type */
++#define PT_NULL 0 /* unused */
++#define PT_LOAD 1 /* loadable segment */
++#define PT_DYNAMIC 2 /* dynamic linking section */
++#define PT_INTERP 3 /* the RTLD */
++#define PT_NOTE 4 /* auxiliary information */
++#define PT_SHLIB 5 /* reserved - purpose undefined */
++#define PT_PHDR 6 /* program header */
++#define PT_NUM 7 /* Number of segment types */
++#define PT_LOPROC 0x70000000 /* reserved range for processor */
++#define PT_HIPROC 0x7fffffff /* specific segment types */
+
-+#define __HYPERVISOR_VIRT_START 0xFFFF800000000000
-+#define __HYPERVISOR_VIRT_END 0xFFFF880000000000
-+#define __MACH2PHYS_VIRT_START 0xFFFF800000000000
-+#define __MACH2PHYS_VIRT_END 0xFFFF804000000000
++/* Segment flags - p_flags */
++#define PF_X 0x1 /* Executable */
++#define PF_W 0x2 /* Writable */
++#define PF_R 0x4 /* Readable */
++#define PF_MASKPROC 0xf0000000 /* reserved bits for processor */
++ /* specific segment flags */
+
-+#ifndef HYPERVISOR_VIRT_START
-+#define HYPERVISOR_VIRT_START mk_unsigned_long(__HYPERVISOR_VIRT_START)
-+#define HYPERVISOR_VIRT_END mk_unsigned_long(__HYPERVISOR_VIRT_END)
-+#endif
++/* Dynamic structure */
++typedef struct {
++ Elf32_Sword d_tag; /* controls meaning of d_val */
++ union {
++ Elf32_Word d_val; /* Multiple meanings - see d_tag */
++ Elf32_Addr d_ptr; /* program virtual address */
++ } d_un;
++} Elf32_Dyn;
+
-+#define MACH2PHYS_VIRT_START mk_unsigned_long(__MACH2PHYS_VIRT_START)
-+#define MACH2PHYS_VIRT_END mk_unsigned_long(__MACH2PHYS_VIRT_END)
-+#define MACH2PHYS_NR_ENTRIES ((MACH2PHYS_VIRT_END-MACH2PHYS_VIRT_START)>>3)
-+#ifndef machine_to_phys_mapping
-+#define machine_to_phys_mapping ((unsigned long *)HYPERVISOR_VIRT_START)
-+#endif
++typedef struct {
++ Elf64_Xword d_tag; /* controls meaning of d_val */
++ union {
++ Elf64_Addr d_ptr;
++ Elf64_Xword d_val;
++ } d_un;
++} Elf64_Dyn;
+
-+#ifndef __ASSEMBLY__
++/* Dynamic Array Tags - d_tag */
++#define DT_NULL 0 /* marks end of _DYNAMIC array */
++#define DT_NEEDED 1 /* string table offset of needed lib */
++#define DT_PLTRELSZ 2 /* size of relocation entries in PLT */
++#define DT_PLTGOT 3 /* address PLT/GOT */
++#define DT_HASH 4 /* address of symbol hash table */
++#define DT_STRTAB 5 /* address of string table */
++#define DT_SYMTAB 6 /* address of symbol table */
++#define DT_RELA 7 /* address of relocation table */
++#define DT_RELASZ 8 /* size of relocation table */
++#define DT_RELAENT 9 /* size of relocation entry */
++#define DT_STRSZ 10 /* size of string table */
++#define DT_SYMENT 11 /* size of symbol table entry */
++#define DT_INIT 12 /* address of initialization func. */
++#define DT_FINI 13 /* address of termination function */
++#define DT_SONAME 14 /* string table offset of shared obj */
++#define DT_RPATH 15 /* string table offset of library
++ search path */
++#define DT_SYMBOLIC 16 /* start sym search in shared obj. */
++#define DT_REL 17 /* address of rel. tbl. w addends */
++#define DT_RELSZ 18 /* size of DT_REL relocation table */
++#define DT_RELENT 19 /* size of DT_REL relocation entry */
++#define DT_PLTREL 20 /* PLT referenced relocation entry */
++#define DT_DEBUG 21 /* bugger */
++#define DT_TEXTREL 22 /* Allow rel. mod. to unwritable seg */
++#define DT_JMPREL 23 /* add. of PLT's relocation entries */
++#define DT_BIND_NOW 24 /* Bind now regardless of env setting */
++#define DT_NUM 25 /* Number used. */
++#define DT_LOPROC 0x70000000 /* reserved range for processor */
++#define DT_HIPROC 0x7fffffff /* specific dynamic array tags */
+
-+/*
-+ * int HYPERVISOR_set_segment_base(unsigned int which, unsigned long base)
-+ * @which == SEGBASE_* ; @base == 64-bit base address
-+ * Returns 0 on success.
-+ */
-+#define SEGBASE_FS 0
-+#define SEGBASE_GS_USER 1
-+#define SEGBASE_GS_KERNEL 2
-+#define SEGBASE_GS_USER_SEL 3 /* Set user %gs specified in base[15:0] */
++/* Standard ELF hashing function */
++unsigned int elf_hash(const unsigned char *name);
+
+/*
-+ * int HYPERVISOR_iret(void)
-+ * All arguments are on the kernel stack, in the following format.
-+ * Never returns if successful. Current kernel context is lost.
-+ * The saved CS is mapped as follows:
-+ * RING0 -> RING3 kernel mode.
-+ * RING1 -> RING3 kernel mode.
-+ * RING2 -> RING3 kernel mode.
-+ * RING3 -> RING3 user mode.
-+ * However RING0 indicates that the guest kernel should return to iteself
-+ * directly with
-+ * orb $3,1*8(%rsp)
-+ * iretq
-+ * If flags contains VGCF_in_syscall:
-+ * Restore RAX, RIP, RFLAGS, RSP.
-+ * Discard R11, RCX, CS, SS.
-+ * Otherwise:
-+ * Restore RAX, R11, RCX, CS:RIP, RFLAGS, SS:RSP.
-+ * All other registers are saved on hypercall entry and restored to user.
++ * Note Definitions
+ */
-+/* Guest exited in SYSCALL context? Return to guest with SYSRET? */
-+#define _VGCF_in_syscall 8
-+#define VGCF_in_syscall (1<<_VGCF_in_syscall)
-+#define VGCF_IN_SYSCALL VGCF_in_syscall
-+struct iret_context {
-+ /* Top of stack (%rsp at point of hypercall). */
-+ uint64_t rax, r11, rcx, flags, rip, cs, rflags, rsp, ss;
-+ /* Bottom of iret stack frame. */
-+};
++typedef struct {
++ Elf32_Word namesz;
++ Elf32_Word descsz;
++ Elf32_Word type;
++} Elf32_Note;
+
-+#ifdef __GNUC__
-+/* Anonymous union includes both 32- and 64-bit names (e.g., eax/rax). */
-+#define __DECL_REG(name) union { \
-+ uint64_t r ## name, e ## name; \
-+ uint32_t _e ## name; \
-+}
-+#else
-+/* Non-gcc sources must always use the proper 64-bit name (e.g., rax). */
-+#define __DECL_REG(name) uint64_t r ## name
-+#endif
++typedef struct {
++ Elf64_Half namesz;
++ Elf64_Half descsz;
++ Elf64_Half type;
++} Elf64_Note;
+
-+struct cpu_user_regs {
-+ uint64_t r15;
-+ uint64_t r14;
-+ uint64_t r13;
-+ uint64_t r12;
-+ __DECL_REG(bp);
-+ __DECL_REG(bx);
-+ uint64_t r11;
-+ uint64_t r10;
-+ uint64_t r9;
-+ uint64_t r8;
-+ __DECL_REG(ax);
-+ __DECL_REG(cx);
-+ __DECL_REG(dx);
-+ __DECL_REG(si);
-+ __DECL_REG(di);
-+ uint32_t error_code; /* private */
-+ uint32_t entry_vector; /* private */
-+ __DECL_REG(ip);
-+ uint16_t cs, _pad0[1];
-+ uint8_t saved_upcall_mask;
-+ uint8_t _pad1[3];
-+ __DECL_REG(flags); /* rflags.IF == !saved_upcall_mask */
-+ __DECL_REG(sp);
-+ uint16_t ss, _pad2[3];
-+ uint16_t es, _pad3[3];
-+ uint16_t ds, _pad4[3];
-+ uint16_t fs, _pad5[3]; /* Non-zero => takes precedence over fs_base. */
-+ uint16_t gs, _pad6[3]; /* Non-zero => takes precedence over gs_base_usr. */
-+};
-+typedef struct cpu_user_regs cpu_user_regs_t;
-+DEFINE_XEN_GUEST_HANDLE(cpu_user_regs_t);
+
-+#undef __DECL_REG
++#if defined(ELFSIZE)
++#define CONCAT(x,y) __CONCAT(x,y)
++#define ELFNAME(x) CONCAT(elf,CONCAT(ELFSIZE,CONCAT(_,x)))
++#define ELFNAME2(x,y) CONCAT(x,CONCAT(_elf,CONCAT(ELFSIZE,CONCAT(_,y))))
++#define ELFNAMEEND(x) CONCAT(x,CONCAT(_elf,ELFSIZE))
++#define ELFDEFNNAME(x) CONCAT(ELF,CONCAT(ELFSIZE,CONCAT(_,x)))
++#endif
+
-+#define xen_pfn_to_cr3(pfn) ((unsigned long)(pfn) << 12)
-+#define xen_cr3_to_pfn(cr3) ((unsigned long)(cr3) >> 12)
++#if defined(ELFSIZE) && (ELFSIZE == 32)
++#define Elf_Ehdr Elf32_Ehdr
++#define Elf_Phdr Elf32_Phdr
++#define Elf_Shdr Elf32_Shdr
++#define Elf_Sym Elf32_Sym
++#define Elf_Rel Elf32_Rel
++#define Elf_RelA Elf32_Rela
++#define Elf_Dyn Elf32_Dyn
++#define Elf_Word Elf32_Word
++#define Elf_Sword Elf32_Sword
++#define Elf_Addr Elf32_Addr
++#define Elf_Off Elf32_Off
++#define Elf_Nhdr Elf32_Nhdr
++#define Elf_Note Elf32_Note
+
-+struct arch_vcpu_info {
-+ unsigned long cr2;
-+ unsigned long pad; /* sizeof(vcpu_info_t) == 64 */
-+};
-+typedef struct arch_vcpu_info arch_vcpu_info_t;
++#define ELF_R_SYM ELF32_R_SYM
++#define ELF_R_TYPE ELF32_R_TYPE
++#define ELF_R_INFO ELF32_R_INFO
++#define ELFCLASS ELFCLASS32
+
-+typedef unsigned long xen_callback_t;
++#define ELF_ST_BIND ELF32_ST_BIND
++#define ELF_ST_TYPE ELF32_ST_TYPE
++#define ELF_ST_INFO ELF32_ST_INFO
+
-+#endif /* !__ASSEMBLY__ */
++#define AuxInfo Aux32Info
++#elif defined(ELFSIZE) && (ELFSIZE == 64)
++#define Elf_Ehdr Elf64_Ehdr
++#define Elf_Phdr Elf64_Phdr
++#define Elf_Shdr Elf64_Shdr
++#define Elf_Sym Elf64_Sym
++#define Elf_Rel Elf64_Rel
++#define Elf_RelA Elf64_Rela
++#define Elf_Dyn Elf64_Dyn
++#define Elf_Word Elf64_Word
++#define Elf_Sword Elf64_Sword
++#define Elf_Addr Elf64_Addr
++#define Elf_Off Elf64_Off
++#define Elf_Nhdr Elf64_Nhdr
++#define Elf_Note Elf64_Note
+
-+#endif /* __XEN_PUBLIC_ARCH_X86_XEN_X86_64_H__ */
++#define ELF_R_SYM ELF64_R_SYM
++#define ELF_R_TYPE ELF64_R_TYPE
++#define ELF_R_INFO ELF64_R_INFO
++#define ELFCLASS ELFCLASS64
+
-+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
-+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/arch-x86_32.h tmp-linux-2.6-xen.patch/include/xen/interface/arch-x86_32.h
---- pristine-linux-2.6.18.2/include/xen/interface/arch-x86_32.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/arch-x86_32.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,27 @@
-+/******************************************************************************
-+ * arch-x86_32.h
-+ *
-+ * Guest OS interface to x86 32-bit Xen.
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Copyright (c) 2004-2006, K A Fraser
-+ */
++#define ELF_ST_BIND ELF64_ST_BIND
++#define ELF_ST_TYPE ELF64_ST_TYPE
++#define ELF_ST_INFO ELF64_ST_INFO
++
++#define AuxInfo Aux64Info
++#endif
+
-+#include "arch-x86/xen.h"
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/arch-x86_64.h tmp-linux-2.6-xen.patch/include/xen/interface/arch-x86_64.h
---- pristine-linux-2.6.18.2/include/xen/interface/arch-x86_64.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/arch-x86_64.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,27 @@
++#endif /* __XEN_PUBLIC_ELFSTRUCTS_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/event_channel.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/event_channel.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,264 @@
+/******************************************************************************
-+ * arch-x86_64.h
++ * event_channel.h
+ *
-+ * Guest OS interface to x86 64-bit Xen.
++ * Event channels between domains.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -97632,262 +136888,237 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (c) 2004-2006, K A Fraser
-+ */
-+
-+#include "arch-x86/xen.h"
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/callback.h tmp-linux-2.6-xen.patch/include/xen/interface/callback.h
---- pristine-linux-2.6.18.2/include/xen/interface/callback.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/callback.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,92 @@
-+/******************************************************************************
-+ * callback.h
-+ *
-+ * Register guest OS callbacks with Xen.
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Copyright (c) 2006, Ian Campbell
++ * Copyright (c) 2003-2004, K A Fraser.
+ */
+
-+#ifndef __XEN_PUBLIC_CALLBACK_H__
-+#define __XEN_PUBLIC_CALLBACK_H__
-+
-+#include "xen.h"
++#ifndef __XEN_PUBLIC_EVENT_CHANNEL_H__
++#define __XEN_PUBLIC_EVENT_CHANNEL_H__
+
+/*
+ * Prototype for this hypercall is:
-+ * long callback_op(int cmd, void *extra_args)
-+ * @cmd == CALLBACKOP_??? (callback operation).
-+ * @extra_args == Operation-specific extra arguments (NULL if none).
++ * int event_channel_op(int cmd, void *args)
++ * @cmd == EVTCHNOP_??? (event-channel operation).
++ * @args == Operation-specific extra arguments (NULL if none).
+ */
+
-+#define CALLBACKTYPE_event 0
-+#define CALLBACKTYPE_failsafe 1
-+#define CALLBACKTYPE_syscall 2 /* x86_64 only */
++typedef uint32_t evtchn_port_t;
++DEFINE_XEN_GUEST_HANDLE(evtchn_port_t);
++
+/*
-+ * sysenter is only available on x86_32 with the
-+ * supervisor_mode_kernel option enabled.
++ * EVTCHNOP_alloc_unbound: Allocate a port in domain <dom> and mark as
++ * accepting interdomain bindings from domain <remote_dom>. A fresh port
++ * is allocated in <dom> and returned as <port>.
++ * NOTES:
++ * 1. If the caller is unprivileged then <dom> must be DOMID_SELF.
++ * 2. <rdom> may be DOMID_SELF, allowing loopback connections.
+ */
-+#define CALLBACKTYPE_sysenter 3
-+#define CALLBACKTYPE_nmi 4
++#define EVTCHNOP_alloc_unbound 6
++struct evtchn_alloc_unbound {
++ /* IN parameters */
++ domid_t dom, remote_dom;
++ /* OUT parameters */
++ evtchn_port_t port;
++};
++typedef struct evtchn_alloc_unbound evtchn_alloc_unbound_t;
+
+/*
-+ * Disable event deliver during callback? This flag is ignored for event and
-+ * NMI callbacks: event delivery is unconditionally disabled.
++ * EVTCHNOP_bind_interdomain: Construct an interdomain event channel between
++ * the calling domain and <remote_dom>. <remote_dom,remote_port> must identify
++ * a port that is unbound and marked as accepting bindings from the calling
++ * domain. A fresh port is allocated in the calling domain and returned as
++ * <local_port>.
++ * NOTES:
++ * 2. <remote_dom> may be DOMID_SELF, allowing loopback connections.
+ */
-+#define _CALLBACKF_mask_events 0
-+#define CALLBACKF_mask_events (1U << _CALLBACKF_mask_events)
++#define EVTCHNOP_bind_interdomain 0
++struct evtchn_bind_interdomain {
++ /* IN parameters. */
++ domid_t remote_dom;
++ evtchn_port_t remote_port;
++ /* OUT parameters. */
++ evtchn_port_t local_port;
++};
++typedef struct evtchn_bind_interdomain evtchn_bind_interdomain_t;
+
+/*
-+ * Register a callback.
++ * EVTCHNOP_bind_virq: Bind a local event channel to VIRQ <irq> on specified
++ * vcpu.
++ * NOTES:
++ * 1. Virtual IRQs are classified as per-vcpu or global. See the VIRQ list
++ * in xen.h for the classification of each VIRQ.
++ * 2. Global VIRQs must be allocated on VCPU0 but can subsequently be
++ * re-bound via EVTCHNOP_bind_vcpu.
++ * 3. Per-vcpu VIRQs may be bound to at most one event channel per vcpu.
++ * The allocated event channel is bound to the specified vcpu and the
++ * binding cannot be changed.
+ */
-+#define CALLBACKOP_register 0
-+struct callback_register {
-+ uint16_t type;
-+ uint16_t flags;
-+ xen_callback_t address;
++#define EVTCHNOP_bind_virq 1
++struct evtchn_bind_virq {
++ /* IN parameters. */
++ uint32_t virq;
++ uint32_t vcpu;
++ /* OUT parameters. */
++ evtchn_port_t port;
+};
-+typedef struct callback_register callback_register_t;
-+DEFINE_XEN_GUEST_HANDLE(callback_register_t);
++typedef struct evtchn_bind_virq evtchn_bind_virq_t;
+
+/*
-+ * Unregister a callback.
-+ *
-+ * Not all callbacks can be unregistered. -EINVAL will be returned if
-+ * you attempt to unregister such a callback.
++ * EVTCHNOP_bind_pirq: Bind a local event channel to PIRQ <irq>.
++ * NOTES:
++ * 1. A physical IRQ may be bound to at most one event channel per domain.
++ * 2. Only a sufficiently-privileged domain may bind to a physical IRQ.
+ */
-+#define CALLBACKOP_unregister 1
-+struct callback_unregister {
-+ uint16_t type;
-+ uint16_t _unused;
++#define EVTCHNOP_bind_pirq 2
++struct evtchn_bind_pirq {
++ /* IN parameters. */
++ uint32_t pirq;
++#define BIND_PIRQ__WILL_SHARE 1
++ uint32_t flags; /* BIND_PIRQ__* */
++ /* OUT parameters. */
++ evtchn_port_t port;
+};
-+typedef struct callback_unregister callback_unregister_t;
-+DEFINE_XEN_GUEST_HANDLE(callback_unregister_t);
-+
-+#endif /* __XEN_PUBLIC_CALLBACK_H__ */
++typedef struct evtchn_bind_pirq evtchn_bind_pirq_t;
+
+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
++ * EVTCHNOP_bind_ipi: Bind a local event channel to receive events.
++ * NOTES:
++ * 1. The allocated event channel is bound to the specified vcpu. The binding
++ * may not be changed.
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/COPYING tmp-linux-2.6-xen.patch/include/xen/interface/COPYING
---- pristine-linux-2.6.18.2/include/xen/interface/COPYING 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/COPYING 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,38 @@
-+XEN NOTICE
-+==========
-+
-+This copyright applies to all files within this subdirectory and its
-+subdirectories:
-+ include/public/*.h
-+ include/public/hvm/*.h
-+ include/public/io/*.h
-+
-+The intention is that these files can be freely copied into the source
-+tree of an operating system when porting that OS to run on Xen. Doing
-+so does *not* cause the OS to become subject to the terms of the GPL.
-+
-+All other files in the Xen source distribution are covered by version
-+2 of the GNU General Public License except where explicitly stated
-+otherwise within individual source files.
-+
-+ -- Keir Fraser (on behalf of the Xen team)
-+
-+=====================================================================
-+
-+Permission is hereby granted, free of charge, to any person obtaining a copy
-+of this software and associated documentation files (the "Software"), to
-+deal in the Software without restriction, including without limitation the
-+rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+sell copies of the Software, and to permit persons to whom the Software is
-+furnished to do so, subject to the following conditions:
-+
-+The above copyright notice and this permission notice shall be included in
-+all copies or substantial portions of the Software.
++#define EVTCHNOP_bind_ipi 7
++struct evtchn_bind_ipi {
++ uint32_t vcpu;
++ /* OUT parameters. */
++ evtchn_port_t port;
++};
++typedef struct evtchn_bind_ipi evtchn_bind_ipi_t;
+
-+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+DEALINGS IN THE SOFTWARE.
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/dom0_ops.h tmp-linux-2.6-xen.patch/include/xen/interface/dom0_ops.h
---- pristine-linux-2.6.18.2/include/xen/interface/dom0_ops.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/dom0_ops.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,120 @@
-+/******************************************************************************
-+ * dom0_ops.h
-+ *
-+ * Process command requests from domain-0 guest OS.
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Copyright (c) 2002-2003, B Dragovic
-+ * Copyright (c) 2002-2006, K Fraser
++/*
++ * EVTCHNOP_close: Close a local event channel <port>. If the channel is
++ * interdomain then the remote end is placed in the unbound state
++ * (EVTCHNSTAT_unbound), awaiting a new connection.
+ */
++#define EVTCHNOP_close 3
++struct evtchn_close {
++ /* IN parameters. */
++ evtchn_port_t port;
++};
++typedef struct evtchn_close evtchn_close_t;
+
-+#ifndef __XEN_PUBLIC_DOM0_OPS_H__
-+#define __XEN_PUBLIC_DOM0_OPS_H__
-+
-+#include "xen.h"
-+#include "platform.h"
-+
-+#if __XEN_INTERFACE_VERSION__ >= 0x00030204
-+#error "dom0_ops.h is a compatibility interface only"
-+#endif
-+
-+#define DOM0_INTERFACE_VERSION XENPF_INTERFACE_VERSION
-+
-+#define DOM0_SETTIME XENPF_settime
-+#define dom0_settime xenpf_settime
-+#define dom0_settime_t xenpf_settime_t
-+
-+#define DOM0_ADD_MEMTYPE XENPF_add_memtype
-+#define dom0_add_memtype xenpf_add_memtype
-+#define dom0_add_memtype_t xenpf_add_memtype_t
-+
-+#define DOM0_DEL_MEMTYPE XENPF_del_memtype
-+#define dom0_del_memtype xenpf_del_memtype
-+#define dom0_del_memtype_t xenpf_del_memtype_t
-+
-+#define DOM0_READ_MEMTYPE XENPF_read_memtype
-+#define dom0_read_memtype xenpf_read_memtype
-+#define dom0_read_memtype_t xenpf_read_memtype_t
-+
-+#define DOM0_MICROCODE XENPF_microcode_update
-+#define dom0_microcode xenpf_microcode_update
-+#define dom0_microcode_t xenpf_microcode_update_t
++/*
++ * EVTCHNOP_send: Send an event to the remote end of the channel whose local
++ * endpoint is <port>.
++ */
++#define EVTCHNOP_send 4
++struct evtchn_send {
++ /* IN parameters. */
++ evtchn_port_t port;
++};
++typedef struct evtchn_send evtchn_send_t;
+
-+#define DOM0_PLATFORM_QUIRK XENPF_platform_quirk
-+#define dom0_platform_quirk xenpf_platform_quirk
-+#define dom0_platform_quirk_t xenpf_platform_quirk_t
++/*
++ * EVTCHNOP_status: Get the current status of the communication channel which
++ * has an endpoint at <dom, port>.
++ * NOTES:
++ * 1. <dom> may be specified as DOMID_SELF.
++ * 2. Only a sufficiently-privileged domain may obtain the status of an event
++ * channel for which <dom> is not DOMID_SELF.
++ */
++#define EVTCHNOP_status 5
++struct evtchn_status {
++ /* IN parameters */
++ domid_t dom;
++ evtchn_port_t port;
++ /* OUT parameters */
++#define EVTCHNSTAT_closed 0 /* Channel is not in use. */
++#define EVTCHNSTAT_unbound 1 /* Channel is waiting interdom connection.*/
++#define EVTCHNSTAT_interdomain 2 /* Channel is connected to remote domain. */
++#define EVTCHNSTAT_pirq 3 /* Channel is bound to a phys IRQ line. */
++#define EVTCHNSTAT_virq 4 /* Channel is bound to a virtual IRQ line */
++#define EVTCHNSTAT_ipi 5 /* Channel is bound to a virtual IPI line */
++ uint32_t status;
++ uint32_t vcpu; /* VCPU to which this channel is bound. */
++ union {
++ struct {
++ domid_t dom;
++ } unbound; /* EVTCHNSTAT_unbound */
++ struct {
++ domid_t dom;
++ evtchn_port_t port;
++ } interdomain; /* EVTCHNSTAT_interdomain */
++ uint32_t pirq; /* EVTCHNSTAT_pirq */
++ uint32_t virq; /* EVTCHNSTAT_virq */
++ } u;
++};
++typedef struct evtchn_status evtchn_status_t;
+
-+typedef uint64_t cpumap_t;
++/*
++ * EVTCHNOP_bind_vcpu: Specify which vcpu a channel should notify when an
++ * event is pending.
++ * NOTES:
++ * 1. IPI-bound channels always notify the vcpu specified at bind time.
++ * This binding cannot be changed.
++ * 2. Per-VCPU VIRQ channels always notify the vcpu specified at bind time.
++ * This binding cannot be changed.
++ * 3. All other channels notify vcpu0 by default. This default is set when
++ * the channel is allocated (a port that is freed and subsequently reused
++ * has its binding reset to vcpu0).
++ */
++#define EVTCHNOP_bind_vcpu 8
++struct evtchn_bind_vcpu {
++ /* IN parameters. */
++ evtchn_port_t port;
++ uint32_t vcpu;
++};
++typedef struct evtchn_bind_vcpu evtchn_bind_vcpu_t;
+
-+/* Unsupported legacy operation -- defined for API compatibility. */
-+#define DOM0_MSR 15
-+struct dom0_msr {
-+ /* IN variables. */
-+ uint32_t write;
-+ cpumap_t cpu_mask;
-+ uint32_t msr;
-+ uint32_t in1;
-+ uint32_t in2;
-+ /* OUT variables. */
-+ uint32_t out1;
-+ uint32_t out2;
++/*
++ * EVTCHNOP_unmask: Unmask the specified local event-channel port and deliver
++ * a notification to the appropriate VCPU if an event is pending.
++ */
++#define EVTCHNOP_unmask 9
++struct evtchn_unmask {
++ /* IN parameters. */
++ evtchn_port_t port;
+};
-+typedef struct dom0_msr dom0_msr_t;
-+DEFINE_XEN_GUEST_HANDLE(dom0_msr_t);
++typedef struct evtchn_unmask evtchn_unmask_t;
+
-+/* Unsupported legacy operation -- defined for API compatibility. */
-+#define DOM0_PHYSICAL_MEMORY_MAP 40
-+struct dom0_memory_map_entry {
-+ uint64_t start, end;
-+ uint32_t flags; /* reserved */
-+ uint8_t is_ram;
++/*
++ * EVTCHNOP_reset: Close all event channels associated with specified domain.
++ * NOTES:
++ * 1. <dom> may be specified as DOMID_SELF.
++ * 2. Only a sufficiently-privileged domain may specify other than DOMID_SELF.
++ */
++#define EVTCHNOP_reset 10
++struct evtchn_reset {
++ /* IN parameters. */
++ domid_t dom;
+};
-+typedef struct dom0_memory_map_entry dom0_memory_map_entry_t;
-+DEFINE_XEN_GUEST_HANDLE(dom0_memory_map_entry_t);
++typedef struct evtchn_reset evtchn_reset_t;
+
-+struct dom0_op {
-+ uint32_t cmd;
-+ uint32_t interface_version; /* DOM0_INTERFACE_VERSION */
++/*
++ * Argument to event_channel_op_compat() hypercall. Superceded by new
++ * event_channel_op() hypercall since 0x00030202.
++ */
++struct evtchn_op {
++ uint32_t cmd; /* EVTCHNOP_* */
+ union {
-+ struct dom0_msr msr;
-+ struct dom0_settime settime;
-+ struct dom0_add_memtype add_memtype;
-+ struct dom0_del_memtype del_memtype;
-+ struct dom0_read_memtype read_memtype;
-+ struct dom0_microcode microcode;
-+ struct dom0_platform_quirk platform_quirk;
-+ struct dom0_memory_map_entry physical_memory_map;
-+ uint8_t pad[128];
++ struct evtchn_alloc_unbound alloc_unbound;
++ struct evtchn_bind_interdomain bind_interdomain;
++ struct evtchn_bind_virq bind_virq;
++ struct evtchn_bind_pirq bind_pirq;
++ struct evtchn_bind_ipi bind_ipi;
++ struct evtchn_close close;
++ struct evtchn_send send;
++ struct evtchn_status status;
++ struct evtchn_bind_vcpu bind_vcpu;
++ struct evtchn_unmask unmask;
+ } u;
+};
-+typedef struct dom0_op dom0_op_t;
-+DEFINE_XEN_GUEST_HANDLE(dom0_op_t);
++typedef struct evtchn_op evtchn_op_t;
++DEFINE_XEN_GUEST_HANDLE(evtchn_op_t);
+
-+#endif /* __XEN_PUBLIC_DOM0_OPS_H__ */
++#endif /* __XEN_PUBLIC_EVENT_CHANNEL_H__ */
+
+/*
+ * Local variables:
@@ -97898,14 +137129,14 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/domctl.h tmp-linux-2.6-xen.patch/include/xen/interface/domctl.h
---- pristine-linux-2.6.18.2/include/xen/interface/domctl.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/domctl.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,481 @@
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/features.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/features.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,74 @@
+/******************************************************************************
-+ * domctl.h
++ * features.h
+ *
-+ * Domain management operations. For use by node control stack.
++ * Feature flags, reported by XENVER_get_features.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -97925,473 +137156,425 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (c) 2002-2003, B Dragovic
-+ * Copyright (c) 2002-2006, K Fraser
++ * Copyright (c) 2006, Keir Fraser <keir at xensource.com>
+ */
+
-+#ifndef __XEN_PUBLIC_DOMCTL_H__
-+#define __XEN_PUBLIC_DOMCTL_H__
++#ifndef __XEN_PUBLIC_FEATURES_H__
++#define __XEN_PUBLIC_FEATURES_H__
+
-+#if !defined(__XEN__) && !defined(__XEN_TOOLS__)
-+#error "domctl operations are intended for use by node control tools only"
-+#endif
++/*
++ * If set, the guest does not need to write-protect its pagetables, and can
++ * update them via direct writes.
++ */
++#define XENFEAT_writable_page_tables 0
+
-+#include "xen.h"
++/*
++ * If set, the guest does not need to write-protect its segment descriptor
++ * tables, and can update them via direct writes.
++ */
++#define XENFEAT_writable_descriptor_tables 1
+
-+#define XEN_DOMCTL_INTERFACE_VERSION 0x00000005
++/*
++ * If set, translation between the guest's 'pseudo-physical' address space
++ * and the host's machine address space are handled by the hypervisor. In this
++ * mode the guest does not need to perform phys-to/from-machine translations
++ * when performing page table operations.
++ */
++#define XENFEAT_auto_translated_physmap 2
+
-+struct xenctl_cpumap {
-+ XEN_GUEST_HANDLE_64(uint8_t) bitmap;
-+ uint32_t nr_cpus;
-+};
++/* If set, the guest is running in supervisor mode (e.g., x86 ring 0). */
++#define XENFEAT_supervisor_mode_kernel 3
+
+/*
-+ * NB. xen_domctl.domain is an IN/OUT parameter for this operation.
-+ * If it is specified as zero, an id is auto-allocated and returned.
++ * If set, the guest does not need to allocate x86 PAE page directories
++ * below 4GB. This flag is usually implied by auto_translated_physmap.
+ */
-+#define XEN_DOMCTL_createdomain 1
-+struct xen_domctl_createdomain {
-+ /* IN parameters */
-+ uint32_t ssidref;
-+ xen_domain_handle_t handle;
-+ /* Is this an HVM guest (as opposed to a PV guest)? */
-+#define _XEN_DOMCTL_CDF_hvm_guest 0
-+#define XEN_DOMCTL_CDF_hvm_guest (1U<<_XEN_DOMCTL_CDF_hvm_guest)
-+ uint32_t flags;
-+};
-+typedef struct xen_domctl_createdomain xen_domctl_createdomain_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_createdomain_t);
++#define XENFEAT_pae_pgdir_above_4gb 4
+
-+#define XEN_DOMCTL_destroydomain 2
-+#define XEN_DOMCTL_pausedomain 3
-+#define XEN_DOMCTL_unpausedomain 4
-+#define XEN_DOMCTL_resumedomain 27
++/* x86: Does this Xen host support the MMU_PT_UPDATE_PRESERVE_AD hypercall? */
++#define XENFEAT_mmu_pt_update_preserve_ad 5
+
-+#define XEN_DOMCTL_getdomaininfo 5
-+struct xen_domctl_getdomaininfo {
-+ /* OUT variables. */
-+ domid_t domain; /* Also echoed in domctl.domain */
-+ /* Domain is scheduled to die. */
-+#define _XEN_DOMINF_dying 0
-+#define XEN_DOMINF_dying (1U<<_XEN_DOMINF_dying)
-+ /* Domain is an HVM guest (as opposed to a PV guest). */
-+#define _XEN_DOMINF_hvm_guest 1
-+#define XEN_DOMINF_hvm_guest (1U<<_XEN_DOMINF_hvm_guest)
-+ /* The guest OS has shut down. */
-+#define _XEN_DOMINF_shutdown 2
-+#define XEN_DOMINF_shutdown (1U<<_XEN_DOMINF_shutdown)
-+ /* Currently paused by control software. */
-+#define _XEN_DOMINF_paused 3
-+#define XEN_DOMINF_paused (1U<<_XEN_DOMINF_paused)
-+ /* Currently blocked pending an event. */
-+#define _XEN_DOMINF_blocked 4
-+#define XEN_DOMINF_blocked (1U<<_XEN_DOMINF_blocked)
-+ /* Domain is currently running. */
-+#define _XEN_DOMINF_running 5
-+#define XEN_DOMINF_running (1U<<_XEN_DOMINF_running)
-+ /* Being debugged. */
-+#define _XEN_DOMINF_debugged 6
-+#define XEN_DOMINF_debugged (1U<<_XEN_DOMINF_debugged)
-+ /* CPU to which this domain is bound. */
-+#define XEN_DOMINF_cpumask 255
-+#define XEN_DOMINF_cpushift 8
-+ /* XEN_DOMINF_shutdown guest-supplied code. */
-+#define XEN_DOMINF_shutdownmask 255
-+#define XEN_DOMINF_shutdownshift 16
-+ uint32_t flags; /* XEN_DOMINF_* */
-+ uint64_aligned_t tot_pages;
-+ uint64_aligned_t max_pages;
-+ uint64_aligned_t shared_info_frame; /* GMFN of shared_info struct */
-+ uint64_aligned_t cpu_time;
-+ uint32_t nr_online_vcpus; /* Number of VCPUs currently online. */
-+ uint32_t max_vcpu_id; /* Maximum VCPUID in use by this domain. */
-+ uint32_t ssidref;
-+ xen_domain_handle_t handle;
-+};
-+typedef struct xen_domctl_getdomaininfo xen_domctl_getdomaininfo_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_getdomaininfo_t);
++#define XENFEAT_NR_SUBMAPS 1
+
++#endif /* __XEN_PUBLIC_FEATURES_H__ */
+
-+#define XEN_DOMCTL_getmemlist 6
-+struct xen_domctl_getmemlist {
-+ /* IN variables. */
-+ /* Max entries to write to output buffer. */
-+ uint64_aligned_t max_pfns;
-+ /* Start index in guest's page list. */
-+ uint64_aligned_t start_pfn;
-+ XEN_GUEST_HANDLE_64(uint64_t) buffer;
-+ /* OUT variables. */
-+ uint64_aligned_t num_pfns;
-+};
-+typedef struct xen_domctl_getmemlist xen_domctl_getmemlist_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_getmemlist_t);
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/foreign/Makefile
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/foreign/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,37 @@
++XEN_ROOT=../../../..
++include $(XEN_ROOT)/Config.mk
+
++architectures := x86_32 x86_64 ia64
++headers := $(patsubst %, %.h, $(architectures))
++scripts := $(wildcard *.py)
+
-+#define XEN_DOMCTL_getpageframeinfo 7
++.PHONY: all clean check-headers
++all: $(headers) check-headers
+
-+#define XEN_DOMCTL_PFINFO_LTAB_SHIFT 28
-+#define XEN_DOMCTL_PFINFO_NOTAB (0x0U<<28)
-+#define XEN_DOMCTL_PFINFO_L1TAB (0x1U<<28)
-+#define XEN_DOMCTL_PFINFO_L2TAB (0x2U<<28)
-+#define XEN_DOMCTL_PFINFO_L3TAB (0x3U<<28)
-+#define XEN_DOMCTL_PFINFO_L4TAB (0x4U<<28)
-+#define XEN_DOMCTL_PFINFO_LTABTYPE_MASK (0x7U<<28)
-+#define XEN_DOMCTL_PFINFO_LPINTAB (0x1U<<31)
-+#define XEN_DOMCTL_PFINFO_XTAB (0xfU<<28) /* invalid page */
-+#define XEN_DOMCTL_PFINFO_LTAB_MASK (0xfU<<28)
++clean:
++ rm -f $(headers)
++ rm -f checker checker.c $(XEN_TARGET_ARCH).size
++ rm -f *.pyc *.o *~
+
-+struct xen_domctl_getpageframeinfo {
-+ /* IN variables. */
-+ uint64_aligned_t gmfn; /* GMFN to query */
-+ /* OUT variables. */
-+ /* Is the page PINNED to a type? */
-+ uint32_t type; /* see above type defs */
-+};
-+typedef struct xen_domctl_getpageframeinfo xen_domctl_getpageframeinfo_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_getpageframeinfo_t);
++ifeq ($(CROSS_COMPILE)$(XEN_TARGET_ARCH),$(XEN_COMPILE_ARCH))
++check-headers: checker
++ ./checker > $(XEN_TARGET_ARCH).size
++ diff -u reference.size $(XEN_TARGET_ARCH).size
++checker: checker.c $(headers)
++ $(HOSTCC) $(HOSTCFLAGS) -o $@ $<
++else
++check-headers:
++ @echo "cross build: skipping check"
++endif
+
++x86_32.h: ../arch-x86/xen-x86_32.h ../arch-x86/xen.h ../xen.h $(scripts)
++ python mkheader.py $* $@ $(filter %.h,$^)
+
-+#define XEN_DOMCTL_getpageframeinfo2 8
-+struct xen_domctl_getpageframeinfo2 {
-+ /* IN variables. */
-+ uint64_aligned_t num;
-+ /* IN/OUT variables. */
-+ XEN_GUEST_HANDLE_64(uint32_t) array;
-+};
-+typedef struct xen_domctl_getpageframeinfo2 xen_domctl_getpageframeinfo2_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_getpageframeinfo2_t);
++x86_64.h: ../arch-x86/xen-x86_64.h ../arch-x86/xen.h ../xen.h $(scripts)
++ python mkheader.py $* $@ $(filter %.h,$^)
+
++ia64.h: ../arch-ia64.h ../xen.h $(scripts)
++ python mkheader.py $* $@ $(filter %.h,$^)
+
-+/*
-+ * Control shadow pagetables operation
-+ */
-+#define XEN_DOMCTL_shadow_op 10
++checker.c: $(scripts)
++ python mkchecker.py $(XEN_TARGET_ARCH) $@ $(architectures)
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/foreign/mkchecker.py
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/foreign/mkchecker.py Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,58 @@
++#!/usr/bin/python
+
-+/* Disable shadow mode. */
-+#define XEN_DOMCTL_SHADOW_OP_OFF 0
++import sys;
++from structs import structs;
+
-+/* Enable shadow mode (mode contains ORed XEN_DOMCTL_SHADOW_ENABLE_* flags). */
-+#define XEN_DOMCTL_SHADOW_OP_ENABLE 32
++# command line arguments
++arch = sys.argv[1];
++outfile = sys.argv[2];
++archs = sys.argv[3:];
+
-+/* Log-dirty bitmap operations. */
-+ /* Return the bitmap and clean internal copy for next round. */
-+#define XEN_DOMCTL_SHADOW_OP_CLEAN 11
-+ /* Return the bitmap but do not modify internal copy. */
-+#define XEN_DOMCTL_SHADOW_OP_PEEK 12
++f = open(outfile, "w");
++f.write('''
++/*
++ * sanity checks for generated foreign headers:
++ * - verify struct sizes
++ *
++ * generated by %s -- DO NOT EDIT
++ */
++#include <stdio.h>
++#include <stdlib.h>
++#include <stddef.h>
++#include <inttypes.h>
++#include "../xen.h"
++''');
+
-+/* Memory allocation accessors. */
-+#define XEN_DOMCTL_SHADOW_OP_GET_ALLOCATION 30
-+#define XEN_DOMCTL_SHADOW_OP_SET_ALLOCATION 31
++for a in archs:
++ f.write('#include "%s.h"\n' % a);
+
-+/* Legacy enable operations. */
-+ /* Equiv. to ENABLE with no mode flags. */
-+#define XEN_DOMCTL_SHADOW_OP_ENABLE_TEST 1
-+ /* Equiv. to ENABLE with mode flag ENABLE_LOG_DIRTY. */
-+#define XEN_DOMCTL_SHADOW_OP_ENABLE_LOGDIRTY 2
-+ /* Equiv. to ENABLE with mode flags ENABLE_REFCOUNT and ENABLE_TRANSLATE. */
-+#define XEN_DOMCTL_SHADOW_OP_ENABLE_TRANSLATE 3
++f.write('int main(int argc, char *argv[])\n{\n');
+
-+/* Mode flags for XEN_DOMCTL_SHADOW_OP_ENABLE. */
-+ /*
-+ * Shadow pagetables are refcounted: guest does not use explicit mmu
-+ * operations nor write-protect its pagetables.
-+ */
-+#define XEN_DOMCTL_SHADOW_ENABLE_REFCOUNT (1 << 1)
-+ /*
-+ * Log pages in a bitmap as they are dirtied.
-+ * Used for live relocation to determine which pages must be re-sent.
-+ */
-+#define XEN_DOMCTL_SHADOW_ENABLE_LOG_DIRTY (1 << 2)
-+ /*
-+ * Automatically translate GPFNs into MFNs.
-+ */
-+#define XEN_DOMCTL_SHADOW_ENABLE_TRANSLATE (1 << 3)
-+ /*
-+ * Xen does not steal virtual address space from the guest.
-+ * Requires HVM support.
-+ */
-+#define XEN_DOMCTL_SHADOW_ENABLE_EXTERNAL (1 << 4)
++f.write('\tprintf("\\n");');
++f.write('printf("%-25s |", "structs");\n');
++for a in archs:
++ f.write('\tprintf("%%8s", "%s");\n' % a);
++f.write('\tprintf("\\n");');
+
-+struct xen_domctl_shadow_op_stats {
-+ uint32_t fault_count;
-+ uint32_t dirty_count;
-+};
-+typedef struct xen_domctl_shadow_op_stats xen_domctl_shadow_op_stats_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_shadow_op_stats_t);
++f.write('\tprintf("\\n");');
++for struct in structs:
++ f.write('\tprintf("%%-25s |", "%s");\n' % struct);
++ for a in archs:
++ if a == arch:
++ s = struct; # native
++ else:
++ s = struct + "_" + a;
++ f.write('#ifdef %s_has_no_%s\n' % (a, struct));
++ f.write('\tprintf("%8s", "-");\n');
++ f.write("#else\n");
++ f.write('\tprintf("%%8zd", sizeof(struct %s));\n' % s);
++ f.write("#endif\n");
+
-+struct xen_domctl_shadow_op {
-+ /* IN variables. */
-+ uint32_t op; /* XEN_DOMCTL_SHADOW_OP_* */
++ f.write('\tprintf("\\n");\n\n');
+
-+ /* OP_ENABLE */
-+ uint32_t mode; /* XEN_DOMCTL_SHADOW_ENABLE_* */
++f.write('\tprintf("\\n");\n');
++f.write('\texit(0);\n');
++f.write('}\n');
+
-+ /* OP_GET_ALLOCATION / OP_SET_ALLOCATION */
-+ uint32_t mb; /* Shadow memory allocation in MB */
++f.close();
+
-+ /* OP_PEEK / OP_CLEAN */
-+ XEN_GUEST_HANDLE_64(uint8_t) dirty_bitmap;
-+ uint64_aligned_t pages; /* Size of buffer. Updated with actual size. */
-+ struct xen_domctl_shadow_op_stats stats;
-+};
-+typedef struct xen_domctl_shadow_op xen_domctl_shadow_op_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_shadow_op_t);
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/foreign/mkheader.py
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/foreign/mkheader.py Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,167 @@
++#!/usr/bin/python
+
++import sys, re;
++from structs import unions, structs, defines;
+
-+#define XEN_DOMCTL_max_mem 11
-+struct xen_domctl_max_mem {
-+ /* IN variables. */
-+ uint64_aligned_t max_memkb;
-+};
-+typedef struct xen_domctl_max_mem xen_domctl_max_mem_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_max_mem_t);
++# command line arguments
++arch = sys.argv[1];
++outfile = sys.argv[2];
++infiles = sys.argv[3:];
+
+
-+#define XEN_DOMCTL_setvcpucontext 12
-+#define XEN_DOMCTL_getvcpucontext 13
-+struct xen_domctl_vcpucontext {
-+ uint32_t vcpu; /* IN */
-+ XEN_GUEST_HANDLE_64(vcpu_guest_context_t) ctxt; /* IN/OUT */
-+};
-+typedef struct xen_domctl_vcpucontext xen_domctl_vcpucontext_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_vcpucontext_t);
++###########################################################################
++# configuration #2: architecture information
+
++inttypes = {};
++header = {};
++footer = {};
+
-+#define XEN_DOMCTL_getvcpuinfo 14
-+struct xen_domctl_getvcpuinfo {
-+ /* IN variables. */
-+ uint32_t vcpu;
-+ /* OUT variables. */
-+ uint8_t online; /* currently online (not hotplugged)? */
-+ uint8_t blocked; /* blocked waiting for an event? */
-+ uint8_t running; /* currently scheduled on its CPU? */
-+ uint64_aligned_t cpu_time; /* total cpu time consumed (ns) */
-+ uint32_t cpu; /* current mapping */
++# x86_32
++inttypes["x86_32"] = {
++ "unsigned long" : "uint32_t",
++ "long" : "uint32_t",
++ "xen_pfn_t" : "uint32_t",
+};
-+typedef struct xen_domctl_getvcpuinfo xen_domctl_getvcpuinfo_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_getvcpuinfo_t);
-+
++header["x86_32"] = """
++#define __i386___X86_32 1
++#pragma pack(4)
++""";
++footer["x86_32"] = """
++#pragma pack()
++""";
+
-+/* Get/set which physical cpus a vcpu can execute on. */
-+#define XEN_DOMCTL_setvcpuaffinity 9
-+#define XEN_DOMCTL_getvcpuaffinity 25
-+struct xen_domctl_vcpuaffinity {
-+ uint32_t vcpu; /* IN */
-+ struct xenctl_cpumap cpumap; /* IN/OUT */
++# x86_64
++inttypes["x86_64"] = {
++ "unsigned long" : "__align8__ uint64_t",
++ "long" : "__align8__ uint64_t",
++ "xen_pfn_t" : "__align8__ uint64_t",
+};
-+typedef struct xen_domctl_vcpuaffinity xen_domctl_vcpuaffinity_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_vcpuaffinity_t);
-+
++header["x86_64"] = """
++#ifdef __GNUC__
++# define __DECL_REG(name) union { uint64_t r ## name, e ## name; }
++# define __align8__ __attribute__((aligned (8)))
++#else
++# define __DECL_REG(name) uint64_t r ## name
++# define __align8__ FIXME
++#endif
++#define __x86_64___X86_64 1
++""";
+
-+#define XEN_DOMCTL_max_vcpus 15
-+struct xen_domctl_max_vcpus {
-+ uint32_t max; /* maximum number of vcpus */
++# ia64
++inttypes["ia64"] = {
++ "unsigned long" : "__align8__ uint64_t",
++ "long" : "__align8__ uint64_t",
++ "xen_pfn_t" : "__align8__ uint64_t",
++ "long double" : "__align16__ ldouble_t",
+};
-+typedef struct xen_domctl_max_vcpus xen_domctl_max_vcpus_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_max_vcpus_t);
++header["ia64"] = """
++#define __align8__ __attribute__((aligned (8)))
++#define __align16__ __attribute__((aligned (16)))
++typedef unsigned char ldouble_t[16];
++""";
+
+
-+#define XEN_DOMCTL_scheduler_op 16
-+/* Scheduler types. */
-+#define XEN_SCHEDULER_SEDF 4
-+#define XEN_SCHEDULER_CREDIT 5
-+/* Set or get info? */
-+#define XEN_DOMCTL_SCHEDOP_putinfo 0
-+#define XEN_DOMCTL_SCHEDOP_getinfo 1
-+struct xen_domctl_scheduler_op {
-+ uint32_t sched_id; /* XEN_SCHEDULER_* */
-+ uint32_t cmd; /* XEN_DOMCTL_SCHEDOP_* */
-+ union {
-+ struct xen_domctl_sched_sedf {
-+ uint64_aligned_t period;
-+ uint64_aligned_t slice;
-+ uint64_aligned_t latency;
-+ uint32_t extratime;
-+ uint32_t weight;
-+ } sedf;
-+ struct xen_domctl_sched_credit {
-+ uint16_t weight;
-+ uint16_t cap;
-+ } credit;
-+ } u;
-+};
-+typedef struct xen_domctl_scheduler_op xen_domctl_scheduler_op_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_scheduler_op_t);
++###########################################################################
++# main
+
++input = "";
++output = "";
++fileid = re.sub("[-.]", "_", "__FOREIGN_%s__" % outfile.upper());
+
-+#define XEN_DOMCTL_setdomainhandle 17
-+struct xen_domctl_setdomainhandle {
-+ xen_domain_handle_t handle;
-+};
-+typedef struct xen_domctl_setdomainhandle xen_domctl_setdomainhandle_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_setdomainhandle_t);
++# read input header files
++for name in infiles:
++ f = open(name, "r");
++ input += f.read();
++ f.close();
+
++# add header
++output += """
++/*
++ * public xen defines and struct for %s
++ * generated by %s -- DO NOT EDIT
++ */
+
-+#define XEN_DOMCTL_setdebugging 18
-+struct xen_domctl_setdebugging {
-+ uint8_t enable;
-+};
-+typedef struct xen_domctl_setdebugging xen_domctl_setdebugging_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_setdebugging_t);
++#ifndef %s
++#define %s 1
+
++""" % (arch, sys.argv[0], fileid, fileid)
+
-+#define XEN_DOMCTL_irq_permission 19
-+struct xen_domctl_irq_permission {
-+ uint8_t pirq;
-+ uint8_t allow_access; /* flag to specify enable/disable of IRQ access */
-+};
-+typedef struct xen_domctl_irq_permission xen_domctl_irq_permission_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_irq_permission_t);
++if arch in header:
++ output += header[arch];
++ output += "\n";
+
++# add defines to output
++for line in re.findall("#define[^\n]+", input):
++ for define in defines:
++ regex = "#define\s+%s\\b" % define;
++ match = re.search(regex, line);
++ if None == match:
++ continue;
++ if define.upper()[0] == define[0]:
++ replace = define + "_" + arch.upper();
++ else:
++ replace = define + "_" + arch;
++ regex = "\\b%s\\b" % define;
++ output += re.sub(regex, replace, line) + "\n";
++output += "\n";
+
-+#define XEN_DOMCTL_iomem_permission 20
-+struct xen_domctl_iomem_permission {
-+ uint64_aligned_t first_mfn;/* first page (physical page number) in range */
-+ uint64_aligned_t nr_mfns; /* number of pages in range (>0) */
-+ uint8_t allow_access; /* allow (!0) or deny (0) access to range? */
-+};
-+typedef struct xen_domctl_iomem_permission xen_domctl_iomem_permission_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_iomem_permission_t);
++# delete defines, comments, empty lines
++input = re.sub("#define[^\n]+\n", "", input);
++input = re.compile("/\*(.*?)\*/", re.S).sub("", input)
++input = re.compile("\n\s*\n", re.S).sub("\n", input);
+
++# add unions to output
++for union in unions:
++ regex = "union\s+%s\s*\{(.*?)\n\};" % union;
++ match = re.search(regex, input, re.S)
++ if None == match:
++ output += "#define %s_has_no_%s 1\n" % (arch, union);
++ else:
++ output += "union %s_%s {%s\n};\n" % (union, arch, match.group(1));
++ output += "\n";
+
-+#define XEN_DOMCTL_ioport_permission 21
-+struct xen_domctl_ioport_permission {
-+ uint32_t first_port; /* first port int range */
-+ uint32_t nr_ports; /* size of port range */
-+ uint8_t allow_access; /* allow or deny access to range? */
-+};
-+typedef struct xen_domctl_ioport_permission xen_domctl_ioport_permission_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_ioport_permission_t);
++# add structs to output
++for struct in structs:
++ regex = "struct\s+%s\s*\{(.*?)\n\};" % struct;
++ match = re.search(regex, input, re.S)
++ if None == match:
++ output += "#define %s_has_no_%s 1\n" % (arch, struct);
++ else:
++ output += "struct %s_%s {%s\n};\n" % (struct, arch, match.group(1));
++ output += "typedef struct %s_%s %s_%s_t;\n" % (struct, arch, struct, arch);
++ output += "\n";
+
++# add footer
++if arch in footer:
++ output += footer[arch];
++ output += "\n";
++output += "#endif /* %s */\n" % fileid;
+
-+#define XEN_DOMCTL_hypercall_init 22
-+struct xen_domctl_hypercall_init {
-+ uint64_aligned_t gmfn; /* GMFN to be initialised */
-+};
-+typedef struct xen_domctl_hypercall_init xen_domctl_hypercall_init_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_hypercall_init_t);
++# replace: defines
++for define in defines:
++ if define.upper()[0] == define[0]:
++ replace = define + "_" + arch.upper();
++ else:
++ replace = define + "_" + arch;
++ output = re.sub("\\b%s\\b" % define, replace, output);
+
++# replace: unions
++for union in unions:
++ output = re.sub("\\b(union\s+%s)\\b" % union, "\\1_%s" % arch, output);
+
-+#define XEN_DOMCTL_arch_setup 23
-+#define _XEN_DOMAINSETUP_hvm_guest 0
-+#define XEN_DOMAINSETUP_hvm_guest (1UL<<_XEN_DOMAINSETUP_hvm_guest)
-+#define _XEN_DOMAINSETUP_query 1 /* Get parameters (for save) */
-+#define XEN_DOMAINSETUP_query (1UL<<_XEN_DOMAINSETUP_query)
-+typedef struct xen_domctl_arch_setup {
-+ uint64_aligned_t flags; /* XEN_DOMAINSETUP_* */
-+#ifdef __ia64__
-+ uint64_aligned_t bp; /* mpaddr of boot param area */
-+ uint64_aligned_t maxmem; /* Highest memory address for MDT. */
-+ uint64_aligned_t xsi_va; /* Xen shared_info area virtual address. */
-+ uint32_t hypercall_imm; /* Break imm for Xen hypercalls. */
-+#endif
-+} xen_domctl_arch_setup_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_arch_setup_t);
++# replace: structs + struct typedefs
++for struct in structs:
++ output = re.sub("\\b(struct\s+%s)\\b" % struct, "\\1_%s" % arch, output);
++ output = re.sub("\\b(%s)_t\\b" % struct, "\\1_%s_t" % arch, output);
+
++# replace: integer types
++integers = inttypes[arch].keys();
++integers.sort(lambda a, b: cmp(len(b),len(a)));
++for type in integers:
++ output = re.sub("\\b%s\\b" % type, inttypes[arch][type], output);
+
-+#define XEN_DOMCTL_settimeoffset 24
-+struct xen_domctl_settimeoffset {
-+ int32_t time_offset_seconds; /* applied to domain wallclock time */
-+};
-+typedef struct xen_domctl_settimeoffset xen_domctl_settimeoffset_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_settimeoffset_t);
++# print results
++f = open(outfile, "w");
++f.write(output);
++f.close;
+
-+
-+#define XEN_DOMCTL_gethvmcontext 33
-+#define XEN_DOMCTL_sethvmcontext 34
-+typedef struct xen_domctl_hvmcontext {
-+ uint32_t size; /* IN/OUT: size of buffer / bytes filled */
-+ XEN_GUEST_HANDLE_64(uint8_t) buffer; /* IN/OUT: data, or call
-+ * gethvmcontext with NULL
-+ * buffer to get size
-+ * req'd */
-+} xen_domctl_hvmcontext_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_hvmcontext_t);
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/foreign/reference.size
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/foreign/reference.size Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,18 @@
++
++structs | x86_32 x86_64 ia64
++
++start_info | 1104 1152 1152
++trap_info | 8 16 -
++pt_fpreg | - - 16
++cpu_user_regs | 68 200 -
++xen_ia64_boot_param | - - 96
++ia64_tr_entry | - - 32
++vcpu_tr_regs | - - 768
++vcpu_guest_context_regs | - - 22176
++vcpu_guest_context | 2800 5168 22208
++arch_vcpu_info | 24 16 0
++vcpu_time_info | 32 32 32
++vcpu_info | 64 64 48
++arch_shared_info | 268 280 272
++shared_info | 2584 3368 4384
++
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/foreign/structs.py
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/foreign/structs.py Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,58 @@
++# configuration: what needs translation
+
++unions = [ "vcpu_cr_regs",
++ "vcpu_ar_regs" ];
+
-+#define XEN_DOMCTL_set_address_size 35
-+#define XEN_DOMCTL_get_address_size 36
-+typedef struct xen_domctl_address_size {
-+ uint32_t size;
-+} xen_domctl_address_size_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_address_size_t);
++structs = [ "start_info",
++ "trap_info",
++ "pt_fpreg",
++ "cpu_user_regs",
++ "xen_ia64_boot_param",
++ "ia64_tr_entry",
++ "vcpu_tr_regs",
++ "vcpu_guest_context_regs",
++ "vcpu_guest_context",
++ "arch_vcpu_info",
++ "vcpu_time_info",
++ "vcpu_info",
++ "arch_shared_info",
++ "shared_info" ];
+
++defines = [ "__i386__",
++ "__x86_64__",
+
-+#define XEN_DOMCTL_real_mode_area 26
-+struct xen_domctl_real_mode_area {
-+ uint32_t log; /* log2 of Real Mode Area size */
-+};
-+typedef struct xen_domctl_real_mode_area xen_domctl_real_mode_area_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_real_mode_area_t);
++ "FLAT_RING1_CS",
++ "FLAT_RING1_DS",
++ "FLAT_RING1_SS",
+
++ "FLAT_RING3_CS64",
++ "FLAT_RING3_DS64",
++ "FLAT_RING3_SS64",
++ "FLAT_KERNEL_CS64",
++ "FLAT_KERNEL_DS64",
++ "FLAT_KERNEL_SS64",
+
-+#define XEN_DOMCTL_sendtrigger 28
-+#define XEN_DOMCTL_SENDTRIGGER_NMI 0
-+#define XEN_DOMCTL_SENDTRIGGER_RESET 1
-+#define XEN_DOMCTL_SENDTRIGGER_INIT 2
-+struct xen_domctl_sendtrigger {
-+ uint32_t trigger; /* IN */
-+ uint32_t vcpu; /* IN */
-+};
-+typedef struct xen_domctl_sendtrigger xen_domctl_sendtrigger_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_sendtrigger_t);
++ "FLAT_KERNEL_CS",
++ "FLAT_KERNEL_DS",
++ "FLAT_KERNEL_SS",
++
++ # x86_{32,64}
++ "_VGCF_i387_valid",
++ "VGCF_i387_valid",
++ "_VGCF_in_kernel",
++ "VGCF_in_kernel",
++ "_VGCF_failsafe_disables_events",
++ "VGCF_failsafe_disables_events",
++ "_VGCF_syscall_disables_events",
++ "VGCF_syscall_disables_events",
++ "_VGCF_online",
++ "VGCF_online",
+
-+
-+struct xen_domctl {
-+ uint32_t cmd;
-+ uint32_t interface_version; /* XEN_DOMCTL_INTERFACE_VERSION */
-+ domid_t domain;
-+ union {
-+ struct xen_domctl_createdomain createdomain;
-+ struct xen_domctl_getdomaininfo getdomaininfo;
-+ struct xen_domctl_getmemlist getmemlist;
-+ struct xen_domctl_getpageframeinfo getpageframeinfo;
-+ struct xen_domctl_getpageframeinfo2 getpageframeinfo2;
-+ struct xen_domctl_vcpuaffinity vcpuaffinity;
-+ struct xen_domctl_shadow_op shadow_op;
-+ struct xen_domctl_max_mem max_mem;
-+ struct xen_domctl_vcpucontext vcpucontext;
-+ struct xen_domctl_getvcpuinfo getvcpuinfo;
-+ struct xen_domctl_max_vcpus max_vcpus;
-+ struct xen_domctl_scheduler_op scheduler_op;
-+ struct xen_domctl_setdomainhandle setdomainhandle;
-+ struct xen_domctl_setdebugging setdebugging;
-+ struct xen_domctl_irq_permission irq_permission;
-+ struct xen_domctl_iomem_permission iomem_permission;
-+ struct xen_domctl_ioport_permission ioport_permission;
-+ struct xen_domctl_hypercall_init hypercall_init;
-+ struct xen_domctl_arch_setup arch_setup;
-+ struct xen_domctl_settimeoffset settimeoffset;
-+ struct xen_domctl_real_mode_area real_mode_area;
-+ struct xen_domctl_hvmcontext hvmcontext;
-+ struct xen_domctl_address_size address_size;
-+ struct xen_domctl_sendtrigger sendtrigger;
-+ uint8_t pad[128];
-+ } u;
-+};
-+typedef struct xen_domctl xen_domctl_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_domctl_t);
++ # ia64
++ "VGCF_EXTRA_REGS",
+
-+#endif /* __XEN_PUBLIC_DOMCTL_H__ */
++ # all archs
++ "xen_pfn_to_cr3",
++ "MAX_VIRT_CPUS",
++ "MAX_GUEST_CMDLINE" ];
+
-+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
-+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/elfnote.h tmp-linux-2.6-xen.patch/include/xen/interface/elfnote.h
---- pristine-linux-2.6.18.2/include/xen/interface/elfnote.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/elfnote.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,233 @@
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/grant_table.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/grant_table.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,431 @@
+/******************************************************************************
-+ * elfnote.h
-+ *
-+ * Definitions used for the Xen ELF notes.
-+ *
++ * grant_table.h
++ *
++ * Interface for granting foreign access to page frames, and receiving
++ * page-ownership transfers.
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -98410,755 +137593,638 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (c) 2006, Ian Campbell, XenSource Ltd.
-+ */
-+
-+#ifndef __XEN_PUBLIC_ELFNOTE_H__
-+#define __XEN_PUBLIC_ELFNOTE_H__
-+
-+/*
-+ * The notes should live in a PT_NOTE segment and have "Xen" in the
-+ * name field.
-+ *
-+ * Numeric types are either 4 or 8 bytes depending on the content of
-+ * the desc field.
-+ *
-+ * LEGACY indicated the fields in the legacy __xen_guest string which
-+ * this a note type replaces.
-+ */
-+
-+/*
-+ * NAME=VALUE pair (string).
-+ */
-+#define XEN_ELFNOTE_INFO 0
-+
-+/*
-+ * The virtual address of the entry point (numeric).
-+ *
-+ * LEGACY: VIRT_ENTRY
-+ */
-+#define XEN_ELFNOTE_ENTRY 1
-+
-+/* The virtual address of the hypercall transfer page (numeric).
-+ *
-+ * LEGACY: HYPERCALL_PAGE. (n.b. legacy value is a physical page
-+ * number not a virtual address)
-+ */
-+#define XEN_ELFNOTE_HYPERCALL_PAGE 2
-+
-+/* The virtual address where the kernel image should be mapped (numeric).
-+ *
-+ * Defaults to 0.
-+ *
-+ * LEGACY: VIRT_BASE
-+ */
-+#define XEN_ELFNOTE_VIRT_BASE 3
-+
-+/*
-+ * The offset of the ELF paddr field from the acutal required
-+ * psuedo-physical address (numeric).
-+ *
-+ * This is used to maintain backwards compatibility with older kernels
-+ * which wrote __PAGE_OFFSET into that field. This field defaults to 0
-+ * if not present.
-+ *
-+ * LEGACY: ELF_PADDR_OFFSET. (n.b. legacy default is VIRT_BASE)
-+ */
-+#define XEN_ELFNOTE_PADDR_OFFSET 4
-+
-+/*
-+ * The version of Xen that we work with (string).
-+ *
-+ * LEGACY: XEN_VER
-+ */
-+#define XEN_ELFNOTE_XEN_VERSION 5
-+
-+/*
-+ * The name of the guest operating system (string).
-+ *
-+ * LEGACY: GUEST_OS
-+ */
-+#define XEN_ELFNOTE_GUEST_OS 6
-+
-+/*
-+ * The version of the guest operating system (string).
-+ *
-+ * LEGACY: GUEST_VER
-+ */
-+#define XEN_ELFNOTE_GUEST_VERSION 7
-+
-+/*
-+ * The loader type (string).
-+ *
-+ * LEGACY: LOADER
++ * Copyright (c) 2004, K A Fraser
+ */
-+#define XEN_ELFNOTE_LOADER 8
+
-+/*
-+ * The kernel supports PAE (x86/32 only, string = "yes", "no" or
-+ * "bimodal").
-+ *
-+ * For compatibility with Xen 3.0.3 and earlier the "bimodal" setting
-+ * may be given as "yes,bimodal" which will cause older Xen to treat
-+ * this kernel as PAE.
-+ *
-+ * LEGACY: PAE (n.b. The legacy interface included a provision to
-+ * indicate 'extended-cr3' support allowing L3 page tables to be
-+ * placed above 4G. It is assumed that any kernel new enough to use
-+ * these ELF notes will include this and therefore "yes" here is
-+ * equivalent to "yes[entended-cr3]" in the __xen_guest interface.
-+ */
-+#define XEN_ELFNOTE_PAE_MODE 9
++#ifndef __XEN_PUBLIC_GRANT_TABLE_H__
++#define __XEN_PUBLIC_GRANT_TABLE_H__
+
-+/*
-+ * The features supported/required by this kernel (string).
-+ *
-+ * The string must consist of a list of feature names (as given in
-+ * features.h, without the "XENFEAT_" prefix) separated by '|'
-+ * characters. If a feature is required for the kernel to function
-+ * then the feature name must be preceded by a '!' character.
-+ *
-+ * LEGACY: FEATURES
-+ */
-+#define XEN_ELFNOTE_FEATURES 10
+
-+/*
-+ * The kernel requires the symbol table to be loaded (string = "yes" or "no")
-+ * LEGACY: BSD_SYMTAB (n.b. The legacy treated the presence or absence
-+ * of this string as a boolean flag rather than requiring "yes" or
-+ * "no".
++/***********************************
++ * GRANT TABLE REPRESENTATION
+ */
-+#define XEN_ELFNOTE_BSD_SYMTAB 11
+
-+/*
-+ * The lowest address the hypervisor hole can begin at (numeric).
++/* Some rough guidelines on accessing and updating grant-table entries
++ * in a concurrency-safe manner. For more information, Linux contains a
++ * reference implementation for guest OSes (arch/xen/kernel/grant_table.c).
++ *
++ * NB. WMB is a no-op on current-generation x86 processors. However, a
++ * compiler barrier will still be required.
++ *
++ * Introducing a valid entry into the grant table:
++ * 1. Write ent->domid.
++ * 2. Write ent->frame:
++ * GTF_permit_access: Frame to which access is permitted.
++ * GTF_accept_transfer: Pseudo-phys frame slot being filled by new
++ * frame, or zero if none.
++ * 3. Write memory barrier (WMB).
++ * 4. Write ent->flags, inc. valid type.
++ *
++ * Invalidating an unused GTF_permit_access entry:
++ * 1. flags = ent->flags.
++ * 2. Observe that !(flags & (GTF_reading|GTF_writing)).
++ * 3. Check result of SMP-safe CMPXCHG(&ent->flags, flags, 0).
++ * NB. No need for WMB as reuse of entry is control-dependent on success of
++ * step 3, and all architectures guarantee ordering of ctrl-dep writes.
+ *
-+ * This must not be set higher than HYPERVISOR_VIRT_START. Its presence
-+ * also indicates to the hypervisor that the kernel can deal with the
-+ * hole starting at a higher address.
-+ */
-+#define XEN_ELFNOTE_HV_START_LOW 12
-+
-+/*
-+ * List of maddr_t-sized mask/value pairs describing how to recognize
-+ * (non-present) L1 page table entries carrying valid MFNs (numeric).
-+ */
-+#define XEN_ELFNOTE_L1_MFN_VALID 13
-+
-+/*
-+ * Whether or not the guest supports cooperative suspend cancellation.
-+ */
-+#define XEN_ELFNOTE_SUSPEND_CANCEL 14
-+
-+/*
-+ * The number of the highest elfnote defined.
-+ */
-+#define XEN_ELFNOTE_MAX XEN_ELFNOTE_SUSPEND_CANCEL
-+
-+/*
-+ * System information exported through crash notes.
++ * Invalidating an in-use GTF_permit_access entry:
++ * This cannot be done directly. Request assistance from the domain controller
++ * which can set a timeout on the use of a grant entry and take necessary
++ * action. (NB. This is not yet implemented!).
++ *
++ * Invalidating an unused GTF_accept_transfer entry:
++ * 1. flags = ent->flags.
++ * 2. Observe that !(flags & GTF_transfer_committed). [*]
++ * 3. Check result of SMP-safe CMPXCHG(&ent->flags, flags, 0).
++ * NB. No need for WMB as reuse of entry is control-dependent on success of
++ * step 3, and all architectures guarantee ordering of ctrl-dep writes.
++ * [*] If GTF_transfer_committed is set then the grant entry is 'committed'.
++ * The guest must /not/ modify the grant entry until the address of the
++ * transferred frame is written. It is safe for the guest to spin waiting
++ * for this to occur (detect by observing GTF_transfer_completed in
++ * ent->flags).
+ *
-+ * The kexec / kdump code will create one XEN_ELFNOTE_CRASH_INFO
-+ * note in case of a system crash. This note will contain various
-+ * information about the system, see xen/include/xen/elfcore.h.
-+ */
-+#define XEN_ELFNOTE_CRASH_INFO 0x1000001
-+
-+/*
-+ * System registers exported through crash notes.
++ * Invalidating a committed GTF_accept_transfer entry:
++ * 1. Wait for (ent->flags & GTF_transfer_completed).
+ *
-+ * The kexec / kdump code will create one XEN_ELFNOTE_CRASH_REGS
-+ * note per cpu in case of a system crash. This note is architecture
-+ * specific and will contain registers not saved in the "CORE" note.
-+ * See xen/include/xen/elfcore.h for more information.
-+ */
-+#define XEN_ELFNOTE_CRASH_REGS 0x1000002
-+
-+
-+/*
-+ * xen dump-core none note.
-+ * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_NONE
-+ * in its dump file to indicate that the file is xen dump-core
-+ * file. This note doesn't have any other information.
-+ * See tools/libxc/xc_core.h for more information.
-+ */
-+#define XEN_ELFNOTE_DUMPCORE_NONE 0x2000000
-+
-+/*
-+ * xen dump-core header note.
-+ * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_HEADER
-+ * in its dump file.
-+ * See tools/libxc/xc_core.h for more information.
-+ */
-+#define XEN_ELFNOTE_DUMPCORE_HEADER 0x2000001
-+
-+/*
-+ * xen dump-core xen version note.
-+ * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_XEN_VERSION
-+ * in its dump file. It contains the xen version obtained via the
-+ * XENVER hypercall.
-+ * See tools/libxc/xc_core.h for more information.
++ * Changing a GTF_permit_access from writable to read-only:
++ * Use SMP-safe CMPXCHG to set GTF_readonly, while checking !GTF_writing.
++ *
++ * Changing a GTF_permit_access from read-only to writable:
++ * Use SMP-safe bit-setting instruction.
+ */
-+#define XEN_ELFNOTE_DUMPCORE_XEN_VERSION 0x2000002
+
+/*
-+ * xen dump-core format version note.
-+ * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_FORMAT_VERSION
-+ * in its dump file. It contains a format version identifier.
-+ * See tools/libxc/xc_core.h for more information.
++ * A grant table comprises a packed array of grant entries in one or more
++ * page frames shared between Xen and a guest.
++ * [XEN]: This field is written by Xen and read by the sharing guest.
++ * [GST]: This field is written by the guest and read by Xen.
+ */
-+#define XEN_ELFNOTE_DUMPCORE_FORMAT_VERSION 0x2000003
-+
-+#endif /* __XEN_PUBLIC_ELFNOTE_H__ */
++struct grant_entry {
++ /* GTF_xxx: various type and flag information. [XEN,GST] */
++ uint16_t flags;
++ /* The domain being granted foreign privileges. [GST] */
++ domid_t domid;
++ /*
++ * GTF_permit_access: Frame that @domid is allowed to map and access. [GST]
++ * GTF_accept_transfer: Frame whose ownership transferred by @domid. [XEN]
++ */
++ uint32_t frame;
++};
++typedef struct grant_entry grant_entry_t;
+
+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
-+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/elfstructs.h tmp-linux-2.6-xen.patch/include/xen/interface/elfstructs.h
---- pristine-linux-2.6.18.2/include/xen/interface/elfstructs.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/elfstructs.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,527 @@
-+#ifndef __XEN_PUBLIC_ELFSTRUCTS_H__
-+#define __XEN_PUBLIC_ELFSTRUCTS_H__ 1
-+/*
-+ * Copyright (c) 1995, 1996 Erik Theisen. All rights reserved.
-+ *
-+ * Redistribution and use in source and binary forms, with or without
-+ * modification, are permitted provided that the following conditions
-+ * are met:
-+ * 1. Redistributions of source code must retain the above copyright
-+ * notice, this list of conditions and the following disclaimer.
-+ * 2. Redistributions in binary form must reproduce the above copyright
-+ * notice, this list of conditions and the following disclaimer in the
-+ * documentation and/or other materials provided with the distribution.
-+ * 3. The name of the author may not be used to endorse or promote products
-+ * derived from this software without specific prior written permission
-+ *
-+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
-+ * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
-+ * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
-+ * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
-+ * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
-+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
-+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
-+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
-+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
-+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
++ * Type of grant entry.
++ * GTF_invalid: This grant entry grants no privileges.
++ * GTF_permit_access: Allow @domid to map/access @frame.
++ * GTF_accept_transfer: Allow @domid to transfer ownership of one page frame
++ * to this guest. Xen writes the page number to @frame.
+ */
-+
-+typedef uint8_t Elf_Byte;
-+
-+typedef uint32_t Elf32_Addr; /* Unsigned program address */
-+typedef uint32_t Elf32_Off; /* Unsigned file offset */
-+typedef int32_t Elf32_Sword; /* Signed large integer */
-+typedef uint32_t Elf32_Word; /* Unsigned large integer */
-+typedef uint16_t Elf32_Half; /* Unsigned medium integer */
-+
-+typedef uint64_t Elf64_Addr;
-+typedef uint64_t Elf64_Off;
-+typedef int32_t Elf64_Shalf;
-+
-+typedef int32_t Elf64_Sword;
-+typedef uint32_t Elf64_Word;
-+
-+typedef int64_t Elf64_Sxword;
-+typedef uint64_t Elf64_Xword;
-+
-+typedef uint32_t Elf64_Half;
-+typedef uint16_t Elf64_Quarter;
++#define GTF_invalid (0U<<0)
++#define GTF_permit_access (1U<<0)
++#define GTF_accept_transfer (2U<<0)
++#define GTF_type_mask (3U<<0)
+
+/*
-+ * e_ident[] identification indexes
-+ * See http://www.caldera.com/developers/gabi/2000-07-17/ch4.eheader.html
++ * Subflags for GTF_permit_access.
++ * GTF_readonly: Restrict @domid to read-only mappings and accesses. [GST]
++ * GTF_reading: Grant entry is currently mapped for reading by @domid. [XEN]
++ * GTF_writing: Grant entry is currently mapped for writing by @domid. [XEN]
++ * GTF_PAT, GTF_PWT, GTF_PCD: (x86) cache attribute flags for the grant [GST]
+ */
-+#define EI_MAG0 0 /* file ID */
-+#define EI_MAG1 1 /* file ID */
-+#define EI_MAG2 2 /* file ID */
-+#define EI_MAG3 3 /* file ID */
-+#define EI_CLASS 4 /* file class */
-+#define EI_DATA 5 /* data encoding */
-+#define EI_VERSION 6 /* ELF header version */
-+#define EI_OSABI 7 /* OS/ABI ID */
-+#define EI_ABIVERSION 8 /* ABI version */
-+#define EI_PAD 9 /* start of pad bytes */
-+#define EI_NIDENT 16 /* Size of e_ident[] */
-+
-+/* e_ident[] magic number */
-+#define ELFMAG0 0x7f /* e_ident[EI_MAG0] */
-+#define ELFMAG1 'E' /* e_ident[EI_MAG1] */
-+#define ELFMAG2 'L' /* e_ident[EI_MAG2] */
-+#define ELFMAG3 'F' /* e_ident[EI_MAG3] */
-+#define ELFMAG "\177ELF" /* magic */
-+#define SELFMAG 4 /* size of magic */
-+
-+/* e_ident[] file class */
-+#define ELFCLASSNONE 0 /* invalid */
-+#define ELFCLASS32 1 /* 32-bit objs */
-+#define ELFCLASS64 2 /* 64-bit objs */
-+#define ELFCLASSNUM 3 /* number of classes */
-+
-+/* e_ident[] data encoding */
-+#define ELFDATANONE 0 /* invalid */
-+#define ELFDATA2LSB 1 /* Little-Endian */
-+#define ELFDATA2MSB 2 /* Big-Endian */
-+#define ELFDATANUM 3 /* number of data encode defines */
-+
-+/* e_ident[] Operating System/ABI */
-+#define ELFOSABI_SYSV 0 /* UNIX System V ABI */
-+#define ELFOSABI_HPUX 1 /* HP-UX operating system */
-+#define ELFOSABI_NETBSD 2 /* NetBSD */
-+#define ELFOSABI_LINUX 3 /* GNU/Linux */
-+#define ELFOSABI_HURD 4 /* GNU/Hurd */
-+#define ELFOSABI_86OPEN 5 /* 86Open common IA32 ABI */
-+#define ELFOSABI_SOLARIS 6 /* Solaris */
-+#define ELFOSABI_MONTEREY 7 /* Monterey */
-+#define ELFOSABI_IRIX 8 /* IRIX */
-+#define ELFOSABI_FREEBSD 9 /* FreeBSD */
-+#define ELFOSABI_TRU64 10 /* TRU64 UNIX */
-+#define ELFOSABI_MODESTO 11 /* Novell Modesto */
-+#define ELFOSABI_OPENBSD 12 /* OpenBSD */
-+#define ELFOSABI_ARM 97 /* ARM */
-+#define ELFOSABI_STANDALONE 255 /* Standalone (embedded) application */
-+
-+/* e_ident */
-+#define IS_ELF(ehdr) ((ehdr).e_ident[EI_MAG0] == ELFMAG0 && \
-+ (ehdr).e_ident[EI_MAG1] == ELFMAG1 && \
-+ (ehdr).e_ident[EI_MAG2] == ELFMAG2 && \
-+ (ehdr).e_ident[EI_MAG3] == ELFMAG3)
-+
-+/* ELF Header */
-+typedef struct elfhdr {
-+ unsigned char e_ident[EI_NIDENT]; /* ELF Identification */
-+ Elf32_Half e_type; /* object file type */
-+ Elf32_Half e_machine; /* machine */
-+ Elf32_Word e_version; /* object file version */
-+ Elf32_Addr e_entry; /* virtual entry point */
-+ Elf32_Off e_phoff; /* program header table offset */
-+ Elf32_Off e_shoff; /* section header table offset */
-+ Elf32_Word e_flags; /* processor-specific flags */
-+ Elf32_Half e_ehsize; /* ELF header size */
-+ Elf32_Half e_phentsize; /* program header entry size */
-+ Elf32_Half e_phnum; /* number of program header entries */
-+ Elf32_Half e_shentsize; /* section header entry size */
-+ Elf32_Half e_shnum; /* number of section header entries */
-+ Elf32_Half e_shstrndx; /* section header table's "section
-+ header string table" entry offset */
-+} Elf32_Ehdr;
-+
-+typedef struct {
-+ unsigned char e_ident[EI_NIDENT]; /* Id bytes */
-+ Elf64_Quarter e_type; /* file type */
-+ Elf64_Quarter e_machine; /* machine type */
-+ Elf64_Half e_version; /* version number */
-+ Elf64_Addr e_entry; /* entry point */
-+ Elf64_Off e_phoff; /* Program hdr offset */
-+ Elf64_Off e_shoff; /* Section hdr offset */
-+ Elf64_Half e_flags; /* Processor flags */
-+ Elf64_Quarter e_ehsize; /* sizeof ehdr */
-+ Elf64_Quarter e_phentsize; /* Program header entry size */
-+ Elf64_Quarter e_phnum; /* Number of program headers */
-+ Elf64_Quarter e_shentsize; /* Section header entry size */
-+ Elf64_Quarter e_shnum; /* Number of section headers */
-+ Elf64_Quarter e_shstrndx; /* String table index */
-+} Elf64_Ehdr;
-+
-+/* e_type */
-+#define ET_NONE 0 /* No file type */
-+#define ET_REL 1 /* relocatable file */
-+#define ET_EXEC 2 /* executable file */
-+#define ET_DYN 3 /* shared object file */
-+#define ET_CORE 4 /* core file */
-+#define ET_NUM 5 /* number of types */
-+#define ET_LOPROC 0xff00 /* reserved range for processor */
-+#define ET_HIPROC 0xffff /* specific e_type */
++#define _GTF_readonly (2)
++#define GTF_readonly (1U<<_GTF_readonly)
++#define _GTF_reading (3)
++#define GTF_reading (1U<<_GTF_reading)
++#define _GTF_writing (4)
++#define GTF_writing (1U<<_GTF_writing)
++#define _GTF_PWT (5)
++#define GTF_PWT (1U<<_GTF_PWT)
++#define _GTF_PCD (6)
++#define GTF_PCD (1U<<_GTF_PCD)
++#define _GTF_PAT (7)
++#define GTF_PAT (1U<<_GTF_PAT)
+
-+/* e_machine */
-+#define EM_NONE 0 /* No Machine */
-+#define EM_M32 1 /* AT&T WE 32100 */
-+#define EM_SPARC 2 /* SPARC */
-+#define EM_386 3 /* Intel 80386 */
-+#define EM_68K 4 /* Motorola 68000 */
-+#define EM_88K 5 /* Motorola 88000 */
-+#define EM_486 6 /* Intel 80486 - unused? */
-+#define EM_860 7 /* Intel 80860 */
-+#define EM_MIPS 8 /* MIPS R3000 Big-Endian only */
+/*
-+ * Don't know if EM_MIPS_RS4_BE,
-+ * EM_SPARC64, EM_PARISC,
-+ * or EM_PPC are ABI compliant
++ * Subflags for GTF_accept_transfer:
++ * GTF_transfer_committed: Xen sets this flag to indicate that it is committed
++ * to transferring ownership of a page frame. When a guest sees this flag
++ * it must /not/ modify the grant entry until GTF_transfer_completed is
++ * set by Xen.
++ * GTF_transfer_completed: It is safe for the guest to spin-wait on this flag
++ * after reading GTF_transfer_committed. Xen will always write the frame
++ * address, followed by ORing this flag, in a timely manner.
+ */
-+#define EM_MIPS_RS4_BE 10 /* MIPS R4000 Big-Endian */
-+#define EM_SPARC64 11 /* SPARC v9 64-bit unoffical */
-+#define EM_PARISC 15 /* HPPA */
-+#define EM_SPARC32PLUS 18 /* Enhanced instruction set SPARC */
-+#define EM_PPC 20 /* PowerPC */
-+#define EM_PPC64 21 /* PowerPC 64-bit */
-+#define EM_ARM 40 /* Advanced RISC Machines ARM */
-+#define EM_ALPHA 41 /* DEC ALPHA */
-+#define EM_SPARCV9 43 /* SPARC version 9 */
-+#define EM_ALPHA_EXP 0x9026 /* DEC ALPHA */
-+#define EM_IA_64 50 /* Intel Merced */
-+#define EM_X86_64 62 /* AMD x86-64 architecture */
-+#define EM_VAX 75 /* DEC VAX */
++#define _GTF_transfer_committed (2)
++#define GTF_transfer_committed (1U<<_GTF_transfer_committed)
++#define _GTF_transfer_completed (3)
++#define GTF_transfer_completed (1U<<_GTF_transfer_completed)
+
-+/* Version */
-+#define EV_NONE 0 /* Invalid */
-+#define EV_CURRENT 1 /* Current */
-+#define EV_NUM 2 /* number of versions */
+
-+/* Section Header */
-+typedef struct {
-+ Elf32_Word sh_name; /* name - index into section header
-+ string table section */
-+ Elf32_Word sh_type; /* type */
-+ Elf32_Word sh_flags; /* flags */
-+ Elf32_Addr sh_addr; /* address */
-+ Elf32_Off sh_offset; /* file offset */
-+ Elf32_Word sh_size; /* section size */
-+ Elf32_Word sh_link; /* section header table index link */
-+ Elf32_Word sh_info; /* extra information */
-+ Elf32_Word sh_addralign; /* address alignment */
-+ Elf32_Word sh_entsize; /* section entry size */
-+} Elf32_Shdr;
++/***********************************
++ * GRANT TABLE QUERIES AND USES
++ */
+
-+typedef struct {
-+ Elf64_Half sh_name; /* section name */
-+ Elf64_Half sh_type; /* section type */
-+ Elf64_Xword sh_flags; /* section flags */
-+ Elf64_Addr sh_addr; /* virtual address */
-+ Elf64_Off sh_offset; /* file offset */
-+ Elf64_Xword sh_size; /* section size */
-+ Elf64_Half sh_link; /* link to another */
-+ Elf64_Half sh_info; /* misc info */
-+ Elf64_Xword sh_addralign; /* memory alignment */
-+ Elf64_Xword sh_entsize; /* table entry size */
-+} Elf64_Shdr;
++/*
++ * Reference to a grant entry in a specified domain's grant table.
++ */
++typedef uint32_t grant_ref_t;
+
-+/* Special Section Indexes */
-+#define SHN_UNDEF 0 /* undefined */
-+#define SHN_LORESERVE 0xff00 /* lower bounds of reserved indexes */
-+#define SHN_LOPROC 0xff00 /* reserved range for processor */
-+#define SHN_HIPROC 0xff1f /* specific section indexes */
-+#define SHN_ABS 0xfff1 /* absolute value */
-+#define SHN_COMMON 0xfff2 /* common symbol */
-+#define SHN_HIRESERVE 0xffff /* upper bounds of reserved indexes */
++/*
++ * Handle to track a mapping created via a grant reference.
++ */
++typedef uint32_t grant_handle_t;
+
-+/* sh_type */
-+#define SHT_NULL 0 /* inactive */
-+#define SHT_PROGBITS 1 /* program defined information */
-+#define SHT_SYMTAB 2 /* symbol table section */
-+#define SHT_STRTAB 3 /* string table section */
-+#define SHT_RELA 4 /* relocation section with addends*/
-+#define SHT_HASH 5 /* symbol hash table section */
-+#define SHT_DYNAMIC 6 /* dynamic section */
-+#define SHT_NOTE 7 /* note section */
-+#define SHT_NOBITS 8 /* no space section */
-+#define SHT_REL 9 /* relation section without addends */
-+#define SHT_SHLIB 10 /* reserved - purpose unknown */
-+#define SHT_DYNSYM 11 /* dynamic symbol table section */
-+#define SHT_NUM 12 /* number of section types */
-+#define SHT_LOPROC 0x70000000 /* reserved range for processor */
-+#define SHT_HIPROC 0x7fffffff /* specific section header types */
-+#define SHT_LOUSER 0x80000000 /* reserved range for application */
-+#define SHT_HIUSER 0xffffffff /* specific indexes */
++/*
++ * GNTTABOP_map_grant_ref: Map the grant entry (<dom>,<ref>) for access
++ * by devices and/or host CPUs. If successful, <handle> is a tracking number
++ * that must be presented later to destroy the mapping(s). On error, <handle>
++ * is a negative status code.
++ * NOTES:
++ * 1. If GNTMAP_device_map is specified then <dev_bus_addr> is the address
++ * via which I/O devices may access the granted frame.
++ * 2. If GNTMAP_host_map is specified then a mapping will be added at
++ * either a host virtual address in the current address space, or at
++ * a PTE at the specified machine address. The type of mapping to
++ * perform is selected through the GNTMAP_contains_pte flag, and the
++ * address is specified in <host_addr>.
++ * 3. Mappings should only be destroyed via GNTTABOP_unmap_grant_ref. If a
++ * host mapping is destroyed by other means then it is *NOT* guaranteed
++ * to be accounted to the correct grant reference!
++ */
++#define GNTTABOP_map_grant_ref 0
++struct gnttab_map_grant_ref {
++ /* IN parameters. */
++ uint64_t host_addr;
++ uint32_t flags; /* GNTMAP_* */
++ grant_ref_t ref;
++ domid_t dom;
++ /* OUT parameters. */
++ int16_t status; /* GNTST_* */
++ grant_handle_t handle;
++ uint64_t dev_bus_addr;
++};
++typedef struct gnttab_map_grant_ref gnttab_map_grant_ref_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_map_grant_ref_t);
+
-+/* Section names */
-+#define ELF_BSS ".bss" /* uninitialized data */
-+#define ELF_DATA ".data" /* initialized data */
-+#define ELF_DEBUG ".debug" /* debug */
-+#define ELF_DYNAMIC ".dynamic" /* dynamic linking information */
-+#define ELF_DYNSTR ".dynstr" /* dynamic string table */
-+#define ELF_DYNSYM ".dynsym" /* dynamic symbol table */
-+#define ELF_FINI ".fini" /* termination code */
-+#define ELF_GOT ".got" /* global offset table */
-+#define ELF_HASH ".hash" /* symbol hash table */
-+#define ELF_INIT ".init" /* initialization code */
-+#define ELF_REL_DATA ".rel.data" /* relocation data */
-+#define ELF_REL_FINI ".rel.fini" /* relocation termination code */
-+#define ELF_REL_INIT ".rel.init" /* relocation initialization code */
-+#define ELF_REL_DYN ".rel.dyn" /* relocaltion dynamic link info */
-+#define ELF_REL_RODATA ".rel.rodata" /* relocation read-only data */
-+#define ELF_REL_TEXT ".rel.text" /* relocation code */
-+#define ELF_RODATA ".rodata" /* read-only data */
-+#define ELF_SHSTRTAB ".shstrtab" /* section header string table */
-+#define ELF_STRTAB ".strtab" /* string table */
-+#define ELF_SYMTAB ".symtab" /* symbol table */
-+#define ELF_TEXT ".text" /* code */
++/*
++ * GNTTABOP_unmap_grant_ref: Destroy one or more grant-reference mappings
++ * tracked by <handle>. If <host_addr> or <dev_bus_addr> is zero, that
++ * field is ignored. If non-zero, they must refer to a device/host mapping
++ * that is tracked by <handle>
++ * NOTES:
++ * 1. The call may fail in an undefined manner if either mapping is not
++ * tracked by <handle>.
++ * 3. After executing a batch of unmaps, it is guaranteed that no stale
++ * mappings will remain in the device or host TLBs.
++ */
++#define GNTTABOP_unmap_grant_ref 1
++struct gnttab_unmap_grant_ref {
++ /* IN parameters. */
++ uint64_t host_addr;
++ uint64_t dev_bus_addr;
++ grant_handle_t handle;
++ /* OUT parameters. */
++ int16_t status; /* GNTST_* */
++};
++typedef struct gnttab_unmap_grant_ref gnttab_unmap_grant_ref_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_unmap_grant_ref_t);
+
++/*
++ * GNTTABOP_setup_table: Set up a grant table for <dom> comprising at least
++ * <nr_frames> pages. The frame addresses are written to the <frame_list>.
++ * Only <nr_frames> addresses are written, even if the table is larger.
++ * NOTES:
++ * 1. <dom> may be specified as DOMID_SELF.
++ * 2. Only a sufficiently-privileged domain may specify <dom> != DOMID_SELF.
++ * 3. Xen may not support more than a single grant-table page per domain.
++ */
++#define GNTTABOP_setup_table 2
++struct gnttab_setup_table {
++ /* IN parameters. */
++ domid_t dom;
++ uint32_t nr_frames;
++ /* OUT parameters. */
++ int16_t status; /* GNTST_* */
++ XEN_GUEST_HANDLE(ulong) frame_list;
++};
++typedef struct gnttab_setup_table gnttab_setup_table_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_setup_table_t);
+
-+/* Section Attribute Flags - sh_flags */
-+#define SHF_WRITE 0x1 /* Writable */
-+#define SHF_ALLOC 0x2 /* occupies memory */
-+#define SHF_EXECINSTR 0x4 /* executable */
-+#define SHF_MASKPROC 0xf0000000 /* reserved bits for processor */
-+ /* specific section attributes */
++/*
++ * GNTTABOP_dump_table: Dump the contents of the grant table to the
++ * xen console. Debugging use only.
++ */
++#define GNTTABOP_dump_table 3
++struct gnttab_dump_table {
++ /* IN parameters. */
++ domid_t dom;
++ /* OUT parameters. */
++ int16_t status; /* GNTST_* */
++};
++typedef struct gnttab_dump_table gnttab_dump_table_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_dump_table_t);
+
-+/* Symbol Table Entry */
-+typedef struct elf32_sym {
-+ Elf32_Word st_name; /* name - index into string table */
-+ Elf32_Addr st_value; /* symbol value */
-+ Elf32_Word st_size; /* symbol size */
-+ unsigned char st_info; /* type and binding */
-+ unsigned char st_other; /* 0 - no defined meaning */
-+ Elf32_Half st_shndx; /* section header index */
-+} Elf32_Sym;
++/*
++ * GNTTABOP_transfer_grant_ref: Transfer <frame> to a foreign domain. The
++ * foreign domain has previously registered its interest in the transfer via
++ * <domid, ref>.
++ *
++ * Note that, even if the transfer fails, the specified page no longer belongs
++ * to the calling domain *unless* the error is GNTST_bad_page.
++ */
++#define GNTTABOP_transfer 4
++struct gnttab_transfer {
++ /* IN parameters. */
++ xen_pfn_t mfn;
++ domid_t domid;
++ grant_ref_t ref;
++ /* OUT parameters. */
++ int16_t status;
++};
++typedef struct gnttab_transfer gnttab_transfer_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_transfer_t);
+
-+typedef struct {
-+ Elf64_Half st_name; /* Symbol name index in str table */
-+ Elf_Byte st_info; /* type / binding attrs */
-+ Elf_Byte st_other; /* unused */
-+ Elf64_Quarter st_shndx; /* section index of symbol */
-+ Elf64_Xword st_value; /* value of symbol */
-+ Elf64_Xword st_size; /* size of symbol */
-+} Elf64_Sym;
+
-+/* Symbol table index */
-+#define STN_UNDEF 0 /* undefined */
++/*
++ * GNTTABOP_copy: Hypervisor based copy
++ * source and destinations can be eithers MFNs or, for foreign domains,
++ * grant references. the foreign domain has to grant read/write access
++ * in its grant table.
++ *
++ * The flags specify what type source and destinations are (either MFN
++ * or grant reference).
++ *
++ * Note that this can also be used to copy data between two domains
++ * via a third party if the source and destination domains had previously
++ * grant appropriate access to their pages to the third party.
++ *
++ * source_offset specifies an offset in the source frame, dest_offset
++ * the offset in the target frame and len specifies the number of
++ * bytes to be copied.
++ */
+
-+/* Extract symbol info - st_info */
-+#define ELF32_ST_BIND(x) ((x) >> 4)
-+#define ELF32_ST_TYPE(x) (((unsigned int) x) & 0xf)
-+#define ELF32_ST_INFO(b,t) (((b) << 4) + ((t) & 0xf))
++#define _GNTCOPY_source_gref (0)
++#define GNTCOPY_source_gref (1<<_GNTCOPY_source_gref)
++#define _GNTCOPY_dest_gref (1)
++#define GNTCOPY_dest_gref (1<<_GNTCOPY_dest_gref)
+
-+#define ELF64_ST_BIND(x) ((x) >> 4)
-+#define ELF64_ST_TYPE(x) (((unsigned int) x) & 0xf)
-+#define ELF64_ST_INFO(b,t) (((b) << 4) + ((t) & 0xf))
++#define GNTTABOP_copy 5
++typedef struct gnttab_copy {
++ /* IN parameters. */
++ struct {
++ union {
++ grant_ref_t ref;
++ xen_pfn_t gmfn;
++ } u;
++ domid_t domid;
++ uint16_t offset;
++ } source, dest;
++ uint16_t len;
++ uint16_t flags; /* GNTCOPY_* */
++ /* OUT parameters. */
++ int16_t status;
++} gnttab_copy_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_copy_t);
+
-+/* Symbol Binding - ELF32_ST_BIND - st_info */
-+#define STB_LOCAL 0 /* Local symbol */
-+#define STB_GLOBAL 1 /* Global symbol */
-+#define STB_WEAK 2 /* like global - lower precedence */
-+#define STB_NUM 3 /* number of symbol bindings */
-+#define STB_LOPROC 13 /* reserved range for processor */
-+#define STB_HIPROC 15 /* specific symbol bindings */
++/*
++ * GNTTABOP_query_size: Query the current and maximum sizes of the shared
++ * grant table.
++ * NOTES:
++ * 1. <dom> may be specified as DOMID_SELF.
++ * 2. Only a sufficiently-privileged domain may specify <dom> != DOMID_SELF.
++ */
++#define GNTTABOP_query_size 6
++struct gnttab_query_size {
++ /* IN parameters. */
++ domid_t dom;
++ /* OUT parameters. */
++ uint32_t nr_frames;
++ uint32_t max_nr_frames;
++ int16_t status; /* GNTST_* */
++};
++typedef struct gnttab_query_size gnttab_query_size_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_query_size_t);
+
-+/* Symbol type - ELF32_ST_TYPE - st_info */
-+#define STT_NOTYPE 0 /* not specified */
-+#define STT_OBJECT 1 /* data object */
-+#define STT_FUNC 2 /* function */
-+#define STT_SECTION 3 /* section */
-+#define STT_FILE 4 /* file */
-+#define STT_NUM 5 /* number of symbol types */
-+#define STT_LOPROC 13 /* reserved range for processor */
-+#define STT_HIPROC 15 /* specific symbol types */
++/*
++ * GNTTABOP_unmap_and_replace: Destroy one or more grant-reference mappings
++ * tracked by <handle> but atomically replace the page table entry with one
++ * pointing to the machine address under <new_addr>. <new_addr> will be
++ * redirected to the null entry.
++ * NOTES:
++ * 1. The call may fail in an undefined manner if either mapping is not
++ * tracked by <handle>.
++ * 2. After executing a batch of unmaps, it is guaranteed that no stale
++ * mappings will remain in the device or host TLBs.
++ */
++#define GNTTABOP_unmap_and_replace 7
++struct gnttab_unmap_and_replace {
++ /* IN parameters. */
++ uint64_t host_addr;
++ uint64_t new_addr;
++ grant_handle_t handle;
++ /* OUT parameters. */
++ int16_t status; /* GNTST_* */
++};
++typedef struct gnttab_unmap_and_replace gnttab_unmap_and_replace_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_unmap_and_replace_t);
+
-+/* Relocation entry with implicit addend */
-+typedef struct {
-+ Elf32_Addr r_offset; /* offset of relocation */
-+ Elf32_Word r_info; /* symbol table index and type */
-+} Elf32_Rel;
+
-+/* Relocation entry with explicit addend */
-+typedef struct {
-+ Elf32_Addr r_offset; /* offset of relocation */
-+ Elf32_Word r_info; /* symbol table index and type */
-+ Elf32_Sword r_addend;
-+} Elf32_Rela;
++/*
++ * Bitfield values for update_pin_status.flags.
++ */
++ /* Map the grant entry for access by I/O devices. */
++#define _GNTMAP_device_map (0)
++#define GNTMAP_device_map (1<<_GNTMAP_device_map)
++ /* Map the grant entry for access by host CPUs. */
++#define _GNTMAP_host_map (1)
++#define GNTMAP_host_map (1<<_GNTMAP_host_map)
++ /* Accesses to the granted frame will be restricted to read-only access. */
++#define _GNTMAP_readonly (2)
++#define GNTMAP_readonly (1<<_GNTMAP_readonly)
++ /*
++ * GNTMAP_host_map subflag:
++ * 0 => The host mapping is usable only by the guest OS.
++ * 1 => The host mapping is usable by guest OS + current application.
++ */
++#define _GNTMAP_application_map (3)
++#define GNTMAP_application_map (1<<_GNTMAP_application_map)
+
-+/* Extract relocation info - r_info */
-+#define ELF32_R_SYM(i) ((i) >> 8)
-+#define ELF32_R_TYPE(i) ((unsigned char) (i))
-+#define ELF32_R_INFO(s,t) (((s) << 8) + (unsigned char)(t))
++ /*
++ * GNTMAP_contains_pte subflag:
++ * 0 => This map request contains a host virtual address.
++ * 1 => This map request contains the machine addess of the PTE to update.
++ */
++#define _GNTMAP_contains_pte (4)
++#define GNTMAP_contains_pte (1<<_GNTMAP_contains_pte)
+
-+typedef struct {
-+ Elf64_Xword r_offset; /* where to do it */
-+ Elf64_Xword r_info; /* index & type of relocation */
-+} Elf64_Rel;
++/*
++ * Values for error status returns. All errors are -ve.
++ */
++#define GNTST_okay (0) /* Normal return. */
++#define GNTST_general_error (-1) /* General undefined error. */
++#define GNTST_bad_domain (-2) /* Unrecognsed domain id. */
++#define GNTST_bad_gntref (-3) /* Unrecognised or inappropriate gntref. */
++#define GNTST_bad_handle (-4) /* Unrecognised or inappropriate handle. */
++#define GNTST_bad_virt_addr (-5) /* Inappropriate virtual address to map. */
++#define GNTST_bad_dev_addr (-6) /* Inappropriate device address to unmap.*/
++#define GNTST_no_device_space (-7) /* Out of space in I/O MMU. */
++#define GNTST_permission_denied (-8) /* Not enough privilege for operation. */
++#define GNTST_bad_page (-9) /* Specified page was invalid for op. */
++#define GNTST_bad_copy_arg (-10) /* copy arguments cross page boundary. */
++#define GNTST_address_too_big (-11) /* transfer page address too large. */
+
-+typedef struct {
-+ Elf64_Xword r_offset; /* where to do it */
-+ Elf64_Xword r_info; /* index & type of relocation */
-+ Elf64_Sxword r_addend; /* adjustment value */
-+} Elf64_Rela;
++#define GNTTABOP_error_msgs { \
++ "okay", \
++ "undefined error", \
++ "unrecognised domain id", \
++ "invalid grant reference", \
++ "invalid mapping handle", \
++ "invalid virtual address", \
++ "invalid device address", \
++ "no spare translation slot in the I/O MMU", \
++ "permission denied", \
++ "bad page", \
++ "copy arguments cross page boundary", \
++ "page address size too large" \
++}
+
-+#define ELF64_R_SYM(info) ((info) >> 32)
-+#define ELF64_R_TYPE(info) ((info) & 0xFFFFFFFF)
-+#define ELF64_R_INFO(s,t) (((s) << 32) + (u_int32_t)(t))
++#endif /* __XEN_PUBLIC_GRANT_TABLE_H__ */
+
-+/* Program Header */
-+typedef struct {
-+ Elf32_Word p_type; /* segment type */
-+ Elf32_Off p_offset; /* segment offset */
-+ Elf32_Addr p_vaddr; /* virtual address of segment */
-+ Elf32_Addr p_paddr; /* physical address - ignored? */
-+ Elf32_Word p_filesz; /* number of bytes in file for seg. */
-+ Elf32_Word p_memsz; /* number of bytes in mem. for seg. */
-+ Elf32_Word p_flags; /* flags */
-+ Elf32_Word p_align; /* memory alignment */
-+} Elf32_Phdr;
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/hvm/e820.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/hvm/e820.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,34 @@
+
-+typedef struct {
-+ Elf64_Half p_type; /* entry type */
-+ Elf64_Half p_flags; /* flags */
-+ Elf64_Off p_offset; /* offset */
-+ Elf64_Addr p_vaddr; /* virtual address */
-+ Elf64_Addr p_paddr; /* physical address */
-+ Elf64_Xword p_filesz; /* file size */
-+ Elf64_Xword p_memsz; /* memory size */
-+ Elf64_Xword p_align; /* memory & file alignment */
-+} Elf64_Phdr;
++/*
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
+
-+/* Segment types - p_type */
-+#define PT_NULL 0 /* unused */
-+#define PT_LOAD 1 /* loadable segment */
-+#define PT_DYNAMIC 2 /* dynamic linking section */
-+#define PT_INTERP 3 /* the RTLD */
-+#define PT_NOTE 4 /* auxiliary information */
-+#define PT_SHLIB 5 /* reserved - purpose undefined */
-+#define PT_PHDR 6 /* program header */
-+#define PT_NUM 7 /* Number of segment types */
-+#define PT_LOPROC 0x70000000 /* reserved range for processor */
-+#define PT_HIPROC 0x7fffffff /* specific segment types */
++#ifndef __XEN_PUBLIC_HVM_E820_H__
++#define __XEN_PUBLIC_HVM_E820_H__
+
-+/* Segment flags - p_flags */
-+#define PF_X 0x1 /* Executable */
-+#define PF_W 0x2 /* Writable */
-+#define PF_R 0x4 /* Readable */
-+#define PF_MASKPROC 0xf0000000 /* reserved bits for processor */
-+ /* specific segment flags */
++/* E820 location in HVM virtual address space. */
++#define HVM_E820_PAGE 0x00090000
++#define HVM_E820_NR_OFFSET 0x000001E8
++#define HVM_E820_OFFSET 0x000002D0
+
-+/* Dynamic structure */
-+typedef struct {
-+ Elf32_Sword d_tag; /* controls meaning of d_val */
-+ union {
-+ Elf32_Word d_val; /* Multiple meanings - see d_tag */
-+ Elf32_Addr d_ptr; /* program virtual address */
-+ } d_un;
-+} Elf32_Dyn;
++#define HVM_BELOW_4G_RAM_END 0xF0000000
++#define HVM_BELOW_4G_MMIO_START HVM_BELOW_4G_RAM_END
++#define HVM_BELOW_4G_MMIO_LENGTH ((1ULL << 32) - HVM_BELOW_4G_MMIO_START)
++
++#endif /* __XEN_PUBLIC_HVM_E820_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/hvm/hvm_info_table.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/hvm/hvm_info_table.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,41 @@
++/******************************************************************************
++ * hvm/hvm_info_table.h
++ *
++ * HVM parameter and information table, written into guest memory map.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
+
-+typedef struct {
-+ Elf64_Xword d_tag; /* controls meaning of d_val */
-+ union {
-+ Elf64_Addr d_ptr;
-+ Elf64_Xword d_val;
-+ } d_un;
-+} Elf64_Dyn;
++#ifndef __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__
++#define __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__
+
-+/* Dynamic Array Tags - d_tag */
-+#define DT_NULL 0 /* marks end of _DYNAMIC array */
-+#define DT_NEEDED 1 /* string table offset of needed lib */
-+#define DT_PLTRELSZ 2 /* size of relocation entries in PLT */
-+#define DT_PLTGOT 3 /* address PLT/GOT */
-+#define DT_HASH 4 /* address of symbol hash table */
-+#define DT_STRTAB 5 /* address of string table */
-+#define DT_SYMTAB 6 /* address of symbol table */
-+#define DT_RELA 7 /* address of relocation table */
-+#define DT_RELASZ 8 /* size of relocation table */
-+#define DT_RELAENT 9 /* size of relocation entry */
-+#define DT_STRSZ 10 /* size of string table */
-+#define DT_SYMENT 11 /* size of symbol table entry */
-+#define DT_INIT 12 /* address of initialization func. */
-+#define DT_FINI 13 /* address of termination function */
-+#define DT_SONAME 14 /* string table offset of shared obj */
-+#define DT_RPATH 15 /* string table offset of library
-+ search path */
-+#define DT_SYMBOLIC 16 /* start sym search in shared obj. */
-+#define DT_REL 17 /* address of rel. tbl. w addends */
-+#define DT_RELSZ 18 /* size of DT_REL relocation table */
-+#define DT_RELENT 19 /* size of DT_REL relocation entry */
-+#define DT_PLTREL 20 /* PLT referenced relocation entry */
-+#define DT_DEBUG 21 /* bugger */
-+#define DT_TEXTREL 22 /* Allow rel. mod. to unwritable seg */
-+#define DT_JMPREL 23 /* add. of PLT's relocation entries */
-+#define DT_BIND_NOW 24 /* Bind now regardless of env setting */
-+#define DT_NUM 25 /* Number used. */
-+#define DT_LOPROC 0x70000000 /* reserved range for processor */
-+#define DT_HIPROC 0x7fffffff /* specific dynamic array tags */
++#define HVM_INFO_PFN 0x09F
++#define HVM_INFO_OFFSET 0x800
++#define HVM_INFO_PADDR ((HVM_INFO_PFN << 12) + HVM_INFO_OFFSET)
+
-+/* Standard ELF hashing function */
-+unsigned int elf_hash(const unsigned char *name);
++struct hvm_info_table {
++ char signature[8]; /* "HVM INFO" */
++ uint32_t length;
++ uint8_t checksum;
++ uint8_t acpi_enabled;
++ uint8_t apic_mode;
++ uint32_t nr_vcpus;
++};
+
++#endif /* __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/hvm/hvm_op.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/hvm/hvm_op.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,131 @@
+/*
-+ * Note Definitions
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
+ */
-+typedef struct {
-+ Elf32_Word namesz;
-+ Elf32_Word descsz;
-+ Elf32_Word type;
-+} Elf32_Note;
+
-+typedef struct {
-+ Elf64_Half namesz;
-+ Elf64_Half descsz;
-+ Elf64_Half type;
-+} Elf64_Note;
++#ifndef __XEN_PUBLIC_HVM_HVM_OP_H__
++#define __XEN_PUBLIC_HVM_HVM_OP_H__
+
++/* Get/set subcommands: extra argument == pointer to xen_hvm_param struct. */
++#define HVMOP_set_param 0
++#define HVMOP_get_param 1
++struct xen_hvm_param {
++ domid_t domid; /* IN */
++ uint32_t index; /* IN */
++ uint64_t value; /* IN/OUT */
++};
++typedef struct xen_hvm_param xen_hvm_param_t;
++DEFINE_XEN_GUEST_HANDLE(xen_hvm_param_t);
+
-+#if defined(ELFSIZE)
-+#define CONCAT(x,y) __CONCAT(x,y)
-+#define ELFNAME(x) CONCAT(elf,CONCAT(ELFSIZE,CONCAT(_,x)))
-+#define ELFNAME2(x,y) CONCAT(x,CONCAT(_elf,CONCAT(ELFSIZE,CONCAT(_,y))))
-+#define ELFNAMEEND(x) CONCAT(x,CONCAT(_elf,ELFSIZE))
-+#define ELFDEFNNAME(x) CONCAT(ELF,CONCAT(ELFSIZE,CONCAT(_,x)))
-+#endif
++/* Set the logical level of one of a domain's PCI INTx wires. */
++#define HVMOP_set_pci_intx_level 2
++struct xen_hvm_set_pci_intx_level {
++ /* Domain to be updated. */
++ domid_t domid;
++ /* PCI INTx identification in PCI topology (domain:bus:device:intx). */
++ uint8_t domain, bus, device, intx;
++ /* Assertion level (0 = unasserted, 1 = asserted). */
++ uint8_t level;
++};
++typedef struct xen_hvm_set_pci_intx_level xen_hvm_set_pci_intx_level_t;
++DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_pci_intx_level_t);
+
-+#if defined(ELFSIZE) && (ELFSIZE == 32)
-+#define Elf_Ehdr Elf32_Ehdr
-+#define Elf_Phdr Elf32_Phdr
-+#define Elf_Shdr Elf32_Shdr
-+#define Elf_Sym Elf32_Sym
-+#define Elf_Rel Elf32_Rel
-+#define Elf_RelA Elf32_Rela
-+#define Elf_Dyn Elf32_Dyn
-+#define Elf_Word Elf32_Word
-+#define Elf_Sword Elf32_Sword
-+#define Elf_Addr Elf32_Addr
-+#define Elf_Off Elf32_Off
-+#define Elf_Nhdr Elf32_Nhdr
-+#define Elf_Note Elf32_Note
++/* Set the logical level of one of a domain's ISA IRQ wires. */
++#define HVMOP_set_isa_irq_level 3
++struct xen_hvm_set_isa_irq_level {
++ /* Domain to be updated. */
++ domid_t domid;
++ /* ISA device identification, by ISA IRQ (0-15). */
++ uint8_t isa_irq;
++ /* Assertion level (0 = unasserted, 1 = asserted). */
++ uint8_t level;
++};
++typedef struct xen_hvm_set_isa_irq_level xen_hvm_set_isa_irq_level_t;
++DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_isa_irq_level_t);
+
-+#define ELF_R_SYM ELF32_R_SYM
-+#define ELF_R_TYPE ELF32_R_TYPE
-+#define ELF_R_INFO ELF32_R_INFO
-+#define ELFCLASS ELFCLASS32
++#define HVMOP_set_pci_link_route 4
++struct xen_hvm_set_pci_link_route {
++ /* Domain to be updated. */
++ domid_t domid;
++ /* PCI link identifier (0-3). */
++ uint8_t link;
++ /* ISA IRQ (1-15), or 0 (disable link). */
++ uint8_t isa_irq;
++};
++typedef struct xen_hvm_set_pci_link_route xen_hvm_set_pci_link_route_t;
++DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_pci_link_route_t);
+
-+#define ELF_ST_BIND ELF32_ST_BIND
-+#define ELF_ST_TYPE ELF32_ST_TYPE
-+#define ELF_ST_INFO ELF32_ST_INFO
++/* Flushes all VCPU TLBs: @arg must be NULL. */
++#define HVMOP_flush_tlbs 5
+
-+#define AuxInfo Aux32Info
-+#elif defined(ELFSIZE) && (ELFSIZE == 64)
-+#define Elf_Ehdr Elf64_Ehdr
-+#define Elf_Phdr Elf64_Phdr
-+#define Elf_Shdr Elf64_Shdr
-+#define Elf_Sym Elf64_Sym
-+#define Elf_Rel Elf64_Rel
-+#define Elf_RelA Elf64_Rela
-+#define Elf_Dyn Elf64_Dyn
-+#define Elf_Word Elf64_Word
-+#define Elf_Sword Elf64_Sword
-+#define Elf_Addr Elf64_Addr
-+#define Elf_Off Elf64_Off
-+#define Elf_Nhdr Elf64_Nhdr
-+#define Elf_Note Elf64_Note
++/* Following tools-only interfaces may change in future. */
++#if defined(__XEN__) || defined(__XEN_TOOLS__)
+
-+#define ELF_R_SYM ELF64_R_SYM
-+#define ELF_R_TYPE ELF64_R_TYPE
-+#define ELF_R_INFO ELF64_R_INFO
-+#define ELFCLASS ELFCLASS64
++/* Track dirty VRAM. */
++#define HVMOP_track_dirty_vram 6
++struct xen_hvm_track_dirty_vram {
++ /* Domain to be tracked. */
++ domid_t domid;
++ /* First pfn to track. */
++ uint64_aligned_t first_pfn;
++ /* Number of pages to track. */
++ uint64_aligned_t nr;
++ /* OUT variable. */
++ /* Dirty bitmap buffer. */
++ XEN_GUEST_HANDLE_64(uint8) dirty_bitmap;
++};
++typedef struct xen_hvm_track_dirty_vram xen_hvm_track_dirty_vram_t;
++DEFINE_XEN_GUEST_HANDLE(xen_hvm_track_dirty_vram_t);
++
++/* Notify that some pages got modified by the Device Model. */
++#define HVMOP_modified_memory 7
++struct xen_hvm_modified_memory {
++ /* Domain to be updated. */
++ domid_t domid;
++ /* First pfn. */
++ uint64_aligned_t first_pfn;
++ /* Number of pages. */
++ uint64_aligned_t nr;
++};
++typedef struct xen_hvm_modified_memory xen_hvm_modified_memory_t;
++DEFINE_XEN_GUEST_HANDLE(xen_hvm_modified_memory_t);
++
++#define HVMOP_set_mem_type 8
++typedef enum {
++ HVMMEM_ram_rw, /* Normal read/write guest RAM */
++ HVMMEM_ram_ro, /* Read-only; writes are discarded */
++ HVMMEM_mmio_dm, /* Reads and write go to the device model */
++} hvmmem_type_t;
++/* Notify that a region of memory is to be treated in a specific way. */
++struct xen_hvm_set_mem_type {
++ /* Domain to be updated. */
++ domid_t domid;
++ /* Memory type */
++ hvmmem_type_t hvmmem_type;
++ /* First pfn. */
++ uint64_aligned_t first_pfn;
++ /* Number of pages. */
++ uint64_aligned_t nr;
++};
++typedef struct xen_hvm_set_mem_type xen_hvm_set_mem_type_t;
++DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_mem_type_t);
+
-+#define ELF_ST_BIND ELF64_ST_BIND
-+#define ELF_ST_TYPE ELF64_ST_TYPE
-+#define ELF_ST_INFO ELF64_ST_INFO
+
-+#define AuxInfo Aux64Info
-+#endif
++#endif /* defined(__XEN__) || defined(__XEN_TOOLS__) */
+
-+#endif /* __XEN_PUBLIC_ELFSTRUCTS_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/event_channel.h tmp-linux-2.6-xen.patch/include/xen/interface/event_channel.h
---- pristine-linux-2.6.18.2/include/xen/interface/event_channel.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/event_channel.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,264 @@
-+/******************************************************************************
-+ * event_channel.h
-+ *
-+ * Event channels between domains.
++#endif /* __XEN_PUBLIC_HVM_HVM_OP_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/hvm/ioreq.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/hvm/ioreq.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,127 @@
++/*
++ * ioreq.h: I/O request definitions for device models
++ * Copyright (c) 2004, Intel Corporation.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -99177,257 +138243,226 @@
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Copyright (c) 2003-2004, K A Fraser.
+ */
+
-+#ifndef __XEN_PUBLIC_EVENT_CHANNEL_H__
-+#define __XEN_PUBLIC_EVENT_CHANNEL_H__
++#ifndef _IOREQ_H_
++#define _IOREQ_H_
+
-+/*
-+ * Prototype for this hypercall is:
-+ * int event_channel_op(int cmd, void *args)
-+ * @cmd == EVTCHNOP_??? (event-channel operation).
-+ * @args == Operation-specific extra arguments (NULL if none).
-+ */
++#define IOREQ_READ 1
++#define IOREQ_WRITE 0
+
-+typedef uint32_t evtchn_port_t;
-+DEFINE_XEN_GUEST_HANDLE(evtchn_port_t);
++#define STATE_IOREQ_NONE 0
++#define STATE_IOREQ_READY 1
++#define STATE_IOREQ_INPROCESS 2
++#define STATE_IORESP_READY 3
+
-+/*
-+ * EVTCHNOP_alloc_unbound: Allocate a port in domain <dom> and mark as
-+ * accepting interdomain bindings from domain <remote_dom>. A fresh port
-+ * is allocated in <dom> and returned as <port>.
-+ * NOTES:
-+ * 1. If the caller is unprivileged then <dom> must be DOMID_SELF.
-+ * 2. <rdom> may be DOMID_SELF, allowing loopback connections.
-+ */
-+#define EVTCHNOP_alloc_unbound 6
-+struct evtchn_alloc_unbound {
-+ /* IN parameters */
-+ domid_t dom, remote_dom;
-+ /* OUT parameters */
-+ evtchn_port_t port;
-+};
-+typedef struct evtchn_alloc_unbound evtchn_alloc_unbound_t;
++#define IOREQ_TYPE_PIO 0 /* pio */
++#define IOREQ_TYPE_COPY 1 /* mmio ops */
++#define IOREQ_TYPE_TIMEOFFSET 7
++#define IOREQ_TYPE_INVALIDATE 8 /* mapcache */
+
+/*
-+ * EVTCHNOP_bind_interdomain: Construct an interdomain event channel between
-+ * the calling domain and <remote_dom>. <remote_dom,remote_port> must identify
-+ * a port that is unbound and marked as accepting bindings from the calling
-+ * domain. A fresh port is allocated in the calling domain and returned as
-+ * <local_port>.
-+ * NOTES:
-+ * 2. <remote_dom> may be DOMID_SELF, allowing loopback connections.
++ * VMExit dispatcher should cooperate with instruction decoder to
++ * prepare this structure and notify service OS and DM by sending
++ * virq
+ */
-+#define EVTCHNOP_bind_interdomain 0
-+struct evtchn_bind_interdomain {
-+ /* IN parameters. */
-+ domid_t remote_dom;
-+ evtchn_port_t remote_port;
-+ /* OUT parameters. */
-+ evtchn_port_t local_port;
++struct ioreq {
++ uint64_t addr; /* physical address */
++ uint64_t size; /* size in bytes */
++ uint64_t count; /* for rep prefixes */
++ uint64_t data; /* data (or paddr of data) */
++ uint8_t state:4;
++ uint8_t data_is_ptr:1; /* if 1, data above is the guest paddr
++ * of the real data to use. */
++ uint8_t dir:1; /* 1=read, 0=write */
++ uint8_t df:1;
++ uint8_t pad:1;
++ uint8_t type; /* I/O type */
++ uint8_t _pad0[6];
++ uint64_t io_count; /* How many IO done on a vcpu */
+};
-+typedef struct evtchn_bind_interdomain evtchn_bind_interdomain_t;
++typedef struct ioreq ioreq_t;
+
-+/*
-+ * EVTCHNOP_bind_virq: Bind a local event channel to VIRQ <irq> on specified
-+ * vcpu.
-+ * NOTES:
-+ * 1. Virtual IRQs are classified as per-vcpu or global. See the VIRQ list
-+ * in xen.h for the classification of each VIRQ.
-+ * 2. Global VIRQs must be allocated on VCPU0 but can subsequently be
-+ * re-bound via EVTCHNOP_bind_vcpu.
-+ * 3. Per-vcpu VIRQs may be bound to at most one event channel per vcpu.
-+ * The allocated event channel is bound to the specified vcpu and the
-+ * binding cannot be changed.
-+ */
-+#define EVTCHNOP_bind_virq 1
-+struct evtchn_bind_virq {
-+ /* IN parameters. */
-+ uint32_t virq;
-+ uint32_t vcpu;
-+ /* OUT parameters. */
-+ evtchn_port_t port;
++struct vcpu_iodata {
++ struct ioreq vp_ioreq;
++ /* Event channel port, used for notifications to/from the device model. */
++ uint32_t vp_eport;
++ uint32_t _pad0;
+};
-+typedef struct evtchn_bind_virq evtchn_bind_virq_t;
++typedef struct vcpu_iodata vcpu_iodata_t;
+
-+/*
-+ * EVTCHNOP_bind_pirq: Bind a local event channel to PIRQ <irq>.
-+ * NOTES:
-+ * 1. A physical IRQ may be bound to at most one event channel per domain.
-+ * 2. Only a sufficiently-privileged domain may bind to a physical IRQ.
-+ */
-+#define EVTCHNOP_bind_pirq 2
-+struct evtchn_bind_pirq {
-+ /* IN parameters. */
-+ uint32_t pirq;
-+#define BIND_PIRQ__WILL_SHARE 1
-+ uint32_t flags; /* BIND_PIRQ__* */
-+ /* OUT parameters. */
-+ evtchn_port_t port;
++struct shared_iopage {
++ struct vcpu_iodata vcpu_iodata[1];
+};
-+typedef struct evtchn_bind_pirq evtchn_bind_pirq_t;
++typedef struct shared_iopage shared_iopage_t;
+
-+/*
-+ * EVTCHNOP_bind_ipi: Bind a local event channel to receive events.
-+ * NOTES:
-+ * 1. The allocated event channel is bound to the specified vcpu. The binding
-+ * may not be changed.
-+ */
-+#define EVTCHNOP_bind_ipi 7
-+struct evtchn_bind_ipi {
-+ uint32_t vcpu;
-+ /* OUT parameters. */
-+ evtchn_port_t port;
++struct buf_ioreq {
++ uint8_t type; /* I/O type */
++ uint8_t pad:1;
++ uint8_t dir:1; /* 1=read, 0=write */
++ uint8_t size:2; /* 0=>1, 1=>2, 2=>4, 3=>8. If 8, use two buf_ioreqs */
++ uint32_t addr:20;/* physical address */
++ uint32_t data; /* data */
+};
-+typedef struct evtchn_bind_ipi evtchn_bind_ipi_t;
++typedef struct buf_ioreq buf_ioreq_t;
+
-+/*
-+ * EVTCHNOP_close: Close a local event channel <port>. If the channel is
-+ * interdomain then the remote end is placed in the unbound state
-+ * (EVTCHNSTAT_unbound), awaiting a new connection.
-+ */
-+#define EVTCHNOP_close 3
-+struct evtchn_close {
-+ /* IN parameters. */
-+ evtchn_port_t port;
++#define IOREQ_BUFFER_SLOT_NUM 511 /* 8 bytes each, plus 2 4-byte indexes */
++struct buffered_iopage {
++ unsigned int read_pointer;
++ unsigned int write_pointer;
++ buf_ioreq_t buf_ioreq[IOREQ_BUFFER_SLOT_NUM];
++}; /* NB. Size of this structure must be no greater than one page. */
++typedef struct buffered_iopage buffered_iopage_t;
++
++#if defined(__ia64__)
++struct pio_buffer {
++ uint32_t page_offset;
++ uint32_t pointer;
++ uint32_t data_end;
++ uint32_t buf_size;
++ void *opaque;
+};
-+typedef struct evtchn_close evtchn_close_t;
+
-+/*
-+ * EVTCHNOP_send: Send an event to the remote end of the channel whose local
-+ * endpoint is <port>.
-+ */
-+#define EVTCHNOP_send 4
-+struct evtchn_send {
-+ /* IN parameters. */
-+ evtchn_port_t port;
++#define PIO_BUFFER_IDE_PRIMARY 0 /* I/O port = 0x1F0 */
++#define PIO_BUFFER_IDE_SECONDARY 1 /* I/O port = 0x170 */
++#define PIO_BUFFER_ENTRY_NUM 2
++struct buffered_piopage {
++ struct pio_buffer pio[PIO_BUFFER_ENTRY_NUM];
++ uint8_t buffer[1];
+};
-+typedef struct evtchn_send evtchn_send_t;
++#endif /* defined(__ia64__) */
++
++#define ACPI_PM1A_EVT_BLK_ADDRESS 0x0000000000001f40
++#define ACPI_PM1A_CNT_BLK_ADDRESS (ACPI_PM1A_EVT_BLK_ADDRESS + 0x04)
++#define ACPI_PM_TMR_BLK_ADDRESS (ACPI_PM1A_EVT_BLK_ADDRESS + 0x08)
++#define ACPI_GPE0_BLK_ADDRESS (ACPI_PM_TMR_BLK_ADDRESS + 0x20)
++#define ACPI_GPE0_BLK_LEN 0x08
++
++#endif /* _IOREQ_H_ */
+
+/*
-+ * EVTCHNOP_status: Get the current status of the communication channel which
-+ * has an endpoint at <dom, port>.
-+ * NOTES:
-+ * 1. <dom> may be specified as DOMID_SELF.
-+ * 2. Only a sufficiently-privileged domain may obtain the status of an event
-+ * channel for which <dom> is not DOMID_SELF.
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
+ */
-+#define EVTCHNOP_status 5
-+struct evtchn_status {
-+ /* IN parameters */
-+ domid_t dom;
-+ evtchn_port_t port;
-+ /* OUT parameters */
-+#define EVTCHNSTAT_closed 0 /* Channel is not in use. */
-+#define EVTCHNSTAT_unbound 1 /* Channel is waiting interdom connection.*/
-+#define EVTCHNSTAT_interdomain 2 /* Channel is connected to remote domain. */
-+#define EVTCHNSTAT_pirq 3 /* Channel is bound to a phys IRQ line. */
-+#define EVTCHNSTAT_virq 4 /* Channel is bound to a virtual IRQ line */
-+#define EVTCHNSTAT_ipi 5 /* Channel is bound to a virtual IPI line */
-+ uint32_t status;
-+ uint32_t vcpu; /* VCPU to which this channel is bound. */
-+ union {
-+ struct {
-+ domid_t dom;
-+ } unbound; /* EVTCHNSTAT_unbound */
-+ struct {
-+ domid_t dom;
-+ evtchn_port_t port;
-+ } interdomain; /* EVTCHNSTAT_interdomain */
-+ uint32_t pirq; /* EVTCHNSTAT_pirq */
-+ uint32_t virq; /* EVTCHNSTAT_virq */
-+ } u;
-+};
-+typedef struct evtchn_status evtchn_status_t;
-+
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/hvm/params.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/hvm/params.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,98 @@
+/*
-+ * EVTCHNOP_bind_vcpu: Specify which vcpu a channel should notify when an
-+ * event is pending.
-+ * NOTES:
-+ * 1. IPI-bound channels always notify the vcpu specified at bind time.
-+ * This binding cannot be changed.
-+ * 2. Per-VCPU VIRQ channels always notify the vcpu specified at bind time.
-+ * This binding cannot be changed.
-+ * 3. All other channels notify vcpu0 by default. This default is set when
-+ * the channel is allocated (a port that is freed and subsequently reused
-+ * has its binding reset to vcpu0).
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
+ */
-+#define EVTCHNOP_bind_vcpu 8
-+struct evtchn_bind_vcpu {
-+ /* IN parameters. */
-+ evtchn_port_t port;
-+ uint32_t vcpu;
-+};
-+typedef struct evtchn_bind_vcpu evtchn_bind_vcpu_t;
++
++#ifndef __XEN_PUBLIC_HVM_PARAMS_H__
++#define __XEN_PUBLIC_HVM_PARAMS_H__
++
++#include "hvm_op.h"
+
+/*
-+ * EVTCHNOP_unmask: Unmask the specified local event-channel port and deliver
-+ * a notification to the appropriate VCPU if an event is pending.
++ * Parameter space for HVMOP_{set,get}_param.
+ */
-+#define EVTCHNOP_unmask 9
-+struct evtchn_unmask {
-+ /* IN parameters. */
-+ evtchn_port_t port;
-+};
-+typedef struct evtchn_unmask evtchn_unmask_t;
+
+/*
-+ * EVTCHNOP_reset: Close all event channels associated with specified domain.
-+ * NOTES:
-+ * 1. <dom> may be specified as DOMID_SELF.
-+ * 2. Only a sufficiently-privileged domain may specify other than DOMID_SELF.
++ * How should CPU0 event-channel notifications be delivered?
++ * val[63:56] == 0: val[55:0] is a delivery GSI (Global System Interrupt).
++ * val[63:56] == 1: val[55:0] is a delivery PCI INTx line, as follows:
++ * Domain = val[47:32], Bus = val[31:16],
++ * DevFn = val[15: 8], IntX = val[ 1: 0]
++ * If val == 0 then CPU0 event-channel notifications are not delivered.
+ */
-+#define EVTCHNOP_reset 10
-+struct evtchn_reset {
-+ /* IN parameters. */
-+ domid_t dom;
-+};
-+typedef struct evtchn_reset evtchn_reset_t;
++#define HVM_PARAM_CALLBACK_IRQ 0
+
+/*
-+ * Argument to event_channel_op_compat() hypercall. Superceded by new
-+ * event_channel_op() hypercall since 0x00030202.
++ * These are not used by Xen. They are here for convenience of HVM-guest
++ * xenbus implementations.
+ */
-+struct evtchn_op {
-+ uint32_t cmd; /* EVTCHNOP_* */
-+ union {
-+ struct evtchn_alloc_unbound alloc_unbound;
-+ struct evtchn_bind_interdomain bind_interdomain;
-+ struct evtchn_bind_virq bind_virq;
-+ struct evtchn_bind_pirq bind_pirq;
-+ struct evtchn_bind_ipi bind_ipi;
-+ struct evtchn_close close;
-+ struct evtchn_send send;
-+ struct evtchn_status status;
-+ struct evtchn_bind_vcpu bind_vcpu;
-+ struct evtchn_unmask unmask;
-+ } u;
-+};
-+typedef struct evtchn_op evtchn_op_t;
-+DEFINE_XEN_GUEST_HANDLE(evtchn_op_t);
++#define HVM_PARAM_STORE_PFN 1
++#define HVM_PARAM_STORE_EVTCHN 2
++
++#define HVM_PARAM_PAE_ENABLED 4
++
++#define HVM_PARAM_IOREQ_PFN 5
+
-+#endif /* __XEN_PUBLIC_EVENT_CHANNEL_H__ */
++#define HVM_PARAM_BUFIOREQ_PFN 6
++
++#ifdef __ia64__
++#define HVM_PARAM_NVRAM_FD 7
++#define HVM_PARAM_VHPT_SIZE 8
++#define HVM_PARAM_BUFPIOREQ_PFN 9
++#endif
+
+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
-+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/features.h tmp-linux-2.6-xen.patch/include/xen/interface/features.h
---- pristine-linux-2.6.18.2/include/xen/interface/features.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/features.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,71 @@
-+/******************************************************************************
-+ * features.h
-+ *
-+ * Feature flags, reported by XENVER_get_features.
++ * Set mode for virtual timers (currently x86 only):
++ * delay_for_missed_ticks (default):
++ * Do not advance a vcpu's time beyond the correct delivery time for
++ * interrupts that have been missed due to preemption. Deliver missed
++ * interrupts when the vcpu is rescheduled and advance the vcpu's virtual
++ * time stepwise for each one.
++ * no_delay_for_missed_ticks:
++ * As above, missed interrupts are delivered, but guest time always tracks
++ * wallclock (i.e., real) time while doing so.
++ * no_missed_ticks_pending:
++ * No missed interrupts are held pending. Instead, to ensure ticks are
++ * delivered at some non-zero rate, if we detect missed ticks then the
++ * internal tick alarm is not disabled if the VCPU is preempted during the
++ * next tick period.
++ * one_missed_tick_pending:
++ * Missed interrupts are collapsed together and delivered as one 'late tick'.
++ * Guest time always tracks wallclock (i.e., real) time.
++ */
++#define HVM_PARAM_TIMER_MODE 10
++#define HVMPTM_delay_for_missed_ticks 0
++#define HVMPTM_no_delay_for_missed_ticks 1
++#define HVMPTM_no_missed_ticks_pending 2
++#define HVMPTM_one_missed_tick_pending 3
++
++/* Boolean: Enable virtual HPET (high-precision event timer)? (x86-only) */
++#define HVM_PARAM_HPET_ENABLED 11
++
++/* Identity-map page directory used by Intel EPT when CR0.PG=0. */
++#define HVM_PARAM_IDENT_PT 12
++
++/* Device Model domain, defaults to 0. */
++#define HVM_PARAM_DM_DOMAIN 13
++
++/* ACPI S state: currently support S0 and S3 on x86. */
++#define HVM_PARAM_ACPI_S_STATE 14
++
++#define HVM_NR_PARAMS 15
++
++#endif /* __XEN_PUBLIC_HVM_PARAMS_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/hvm/save.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/hvm/save.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,88 @@
++/*
++ * hvm/save.h
++ *
++ * Structure definitions for HVM state that is held by Xen and must
++ * be saved along with the domain's memory and device-model state.
+ *
++ * Copyright (c) 2007 XenSource Ltd.
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -99445,403 +138480,348 @@
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Copyright (c) 2006, Keir Fraser <keir at xensource.com>
-+ */
-+
-+#ifndef __XEN_PUBLIC_FEATURES_H__
-+#define __XEN_PUBLIC_FEATURES_H__
-+
-+/*
-+ * If set, the guest does not need to write-protect its pagetables, and can
-+ * update them via direct writes.
+ */
-+#define XENFEAT_writable_page_tables 0
+
-+/*
-+ * If set, the guest does not need to write-protect its segment descriptor
-+ * tables, and can update them via direct writes.
-+ */
-+#define XENFEAT_writable_descriptor_tables 1
++#ifndef __XEN_PUBLIC_HVM_SAVE_H__
++#define __XEN_PUBLIC_HVM_SAVE_H__
+
+/*
-+ * If set, translation between the guest's 'pseudo-physical' address space
-+ * and the host's machine address space are handled by the hypervisor. In this
-+ * mode the guest does not need to perform phys-to/from-machine translations
-+ * when performing page table operations.
++ * Structures in this header *must* have the same layout in 32bit
++ * and 64bit environments: this means that all fields must be explicitly
++ * sized types and aligned to their sizes, and the structs must be
++ * a multiple of eight bytes long.
++ *
++ * Only the state necessary for saving and restoring (i.e. fields
++ * that are analogous to actual hardware state) should go in this file.
++ * Internal mechanisms should be kept in Xen-private headers.
+ */
-+#define XENFEAT_auto_translated_physmap 2
+
-+/* If set, the guest is running in supervisor mode (e.g., x86 ring 0). */
-+#define XENFEAT_supervisor_mode_kernel 3
++#if !defined(__GNUC__) || defined(__STRICT_ANSI__)
++#error "Anonymous structs/unions are a GNU extension."
++#endif
+
-+/*
-+ * If set, the guest does not need to allocate x86 PAE page directories
-+ * below 4GB. This flag is usually implied by auto_translated_physmap.
++/*
++ * Each entry is preceded by a descriptor giving its type and length
+ */
-+#define XENFEAT_pae_pgdir_above_4gb 4
-+
-+#define XENFEAT_NR_SUBMAPS 1
++struct hvm_save_descriptor {
++ uint16_t typecode; /* Used to demux the various types below */
++ uint16_t instance; /* Further demux within a type */
++ uint32_t length; /* In bytes, *not* including this descriptor */
++};
+
-+#endif /* __XEN_PUBLIC_FEATURES_H__ */
+
-+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
++/*
++ * Each entry has a datatype associated with it: for example, the CPU state
++ * is saved as a HVM_SAVE_TYPE(CPU), which has HVM_SAVE_LENGTH(CPU),
++ * and is identified by a descriptor with typecode HVM_SAVE_CODE(CPU).
++ * DECLARE_HVM_SAVE_TYPE binds these things together with some type-system
++ * ugliness.
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/foreign/Makefile tmp-linux-2.6-xen.patch/include/xen/interface/foreign/Makefile
---- pristine-linux-2.6.18.2/include/xen/interface/foreign/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/foreign/Makefile 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,37 @@
-+XEN_ROOT=../../../..
-+include $(XEN_ROOT)/Config.mk
-+
-+architectures := x86_32 x86_64 ia64
-+headers := $(patsubst %, %.h, $(architectures))
-+scripts := $(wildcard *.py)
-+
-+.PHONY: all clean check-headers
-+all: $(headers) check-headers
-+
-+clean:
-+ rm -f $(headers)
-+ rm -f checker checker.c $(XEN_TARGET_ARCH).size
-+ rm -f *.pyc *.o *~
-+
-+ifeq ($(CROSS_COMPILE)$(XEN_TARGET_ARCH),$(XEN_COMPILE_ARCH))
-+check-headers: checker
-+ ./checker > $(XEN_TARGET_ARCH).size
-+ diff -u reference.size $(XEN_TARGET_ARCH).size
-+checker: checker.c $(headers)
-+ $(HOSTCC) $(HOSTCFLAGS) -o $@ $<
-+else
-+check-headers:
-+ @echo "cross build: skipping check"
-+endif
+
-+x86_32.h: ../arch-x86/xen-x86_32.h ../arch-x86/xen.h ../xen.h $(scripts)
-+ python mkheader.py $* $@ $(filter %.h,$^)
++#define DECLARE_HVM_SAVE_TYPE(_x, _code, _type) \
++ struct __HVM_SAVE_TYPE_##_x { _type t; char c[_code]; }
+
-+x86_64.h: ../arch-x86/xen-x86_64.h ../arch-x86/xen.h ../xen.h $(scripts)
-+ python mkheader.py $* $@ $(filter %.h,$^)
++#define HVM_SAVE_TYPE(_x) typeof (((struct __HVM_SAVE_TYPE_##_x *)(0))->t)
++#define HVM_SAVE_LENGTH(_x) (sizeof (HVM_SAVE_TYPE(_x)))
++#define HVM_SAVE_CODE(_x) (sizeof (((struct __HVM_SAVE_TYPE_##_x *)(0))->c))
+
-+ia64.h: ../arch-ia64.h ../xen.h $(scripts)
-+ python mkheader.py $* $@ $(filter %.h,$^)
+
-+checker.c: $(scripts)
-+ python mkchecker.py $(XEN_TARGET_ARCH) $@ $(architectures)
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/foreign/mkchecker.py tmp-linux-2.6-xen.patch/include/xen/interface/foreign/mkchecker.py
---- pristine-linux-2.6.18.2/include/xen/interface/foreign/mkchecker.py 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/foreign/mkchecker.py 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,58 @@
-+#!/usr/bin/python
++/*
++ * The series of save records is teminated by a zero-type, zero-length
++ * descriptor.
++ */
+
-+import sys;
-+from structs import structs;
++struct hvm_save_end {};
++DECLARE_HVM_SAVE_TYPE(END, 0, struct hvm_save_end);
+
-+# command line arguments
-+arch = sys.argv[1];
-+outfile = sys.argv[2];
-+archs = sys.argv[3:];
++#if defined(__i386__) || defined(__x86_64__)
++#include "../arch-x86/hvm/save.h"
++#elif defined(__ia64__)
++#include "../arch-ia64/hvm/save.h"
++#else
++#error "unsupported architecture"
++#endif
+
-+f = open(outfile, "w");
-+f.write('''
++#endif /* __XEN_PUBLIC_HVM_SAVE_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/hvm/vmx_assist.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/hvm/vmx_assist.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,122 @@
+/*
-+ * sanity checks for generated foreign headers:
-+ * - verify struct sizes
++ * vmx_assist.h: Context definitions for the VMXASSIST world switch.
+ *
-+ * generated by %s -- DO NOT EDIT
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Leendert van Doorn, leendert at watson.ibm.com
++ * Copyright (c) 2005, International Business Machines Corporation.
+ */
-+#include <stdio.h>
-+#include <stdlib.h>
-+#include <stddef.h>
-+#include <inttypes.h>
-+#include "../xen.h"
-+''');
-+
-+for a in archs:
-+ f.write('#include "%s.h"\n' % a);
-+
-+f.write('int main(int argc, char *argv[])\n{\n');
-+
-+f.write('\tprintf("\\n");');
-+f.write('printf("%-20s |", "structs");\n');
-+for a in archs:
-+ f.write('\tprintf("%%8s", "%s");\n' % a);
-+f.write('\tprintf("\\n");');
-+
-+f.write('\tprintf("\\n");');
-+for struct in structs:
-+ f.write('\tprintf("%%-20s |", "%s");\n' % struct);
-+ for a in archs:
-+ if a == arch:
-+ s = struct; # native
-+ else:
-+ s = struct + "_" + a;
-+ f.write('#ifdef %s_has_no_%s\n' % (a, struct));
-+ f.write('\tprintf("%8s", "-");\n');
-+ f.write("#else\n");
-+ f.write('\tprintf("%%8zd", sizeof(struct %s));\n' % s);
-+ f.write("#endif\n");
-+
-+ f.write('\tprintf("\\n");\n\n');
-+
-+f.write('\tprintf("\\n");\n');
-+f.write('\texit(0);\n');
-+f.write('}\n');
-+
-+f.close();
-+
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/foreign/mkheader.py tmp-linux-2.6-xen.patch/include/xen/interface/foreign/mkheader.py
---- pristine-linux-2.6.18.2/include/xen/interface/foreign/mkheader.py 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/foreign/mkheader.py 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,153 @@
-+#!/usr/bin/python
+
-+import sys, re;
-+from structs import structs, defines;
++#ifndef _VMX_ASSIST_H_
++#define _VMX_ASSIST_H_
+
-+# command line arguments
-+arch = sys.argv[1];
-+outfile = sys.argv[2];
-+infiles = sys.argv[3:];
++#define VMXASSIST_BASE 0xD0000
++#define VMXASSIST_MAGIC 0x17101966
++#define VMXASSIST_MAGIC_OFFSET (VMXASSIST_BASE+8)
+
++#define VMXASSIST_NEW_CONTEXT (VMXASSIST_BASE + 12)
++#define VMXASSIST_OLD_CONTEXT (VMXASSIST_NEW_CONTEXT + 4)
+
-+###########################################################################
-+# configuration #2: architecture information
++#ifndef __ASSEMBLY__
+
-+inttypes = {};
-+header = {};
-+footer = {};
++#define NR_EXCEPTION_HANDLER 32
++#define NR_INTERRUPT_HANDLERS 16
++#define NR_TRAPS (NR_EXCEPTION_HANDLER+NR_INTERRUPT_HANDLERS)
+
-+# x86_32
-+inttypes["x86_32"] = {
-+ "unsigned long" : "uint32_t",
-+ "long" : "uint32_t",
-+ "xen_pfn_t" : "uint32_t",
++union vmcs_arbytes {
++ struct arbyte_fields {
++ unsigned int seg_type : 4,
++ s : 1,
++ dpl : 2,
++ p : 1,
++ reserved0 : 4,
++ avl : 1,
++ reserved1 : 1,
++ default_ops_size: 1,
++ g : 1,
++ null_bit : 1,
++ reserved2 : 15;
++ } fields;
++ unsigned int bytes;
+};
-+header["x86_32"] = """
-+#define __i386___X86_32 1
-+#pragma pack(4)
-+""";
-+footer["x86_32"] = """
-+#pragma pack()
-+""";
+
-+# x86_64
-+inttypes["x86_64"] = {
-+ "unsigned long" : "__align8__ uint64_t",
-+ "long" : "__align8__ uint64_t",
-+ "xen_pfn_t" : "__align8__ uint64_t",
-+};
-+header["x86_64"] = """
-+#ifdef __GNUC__
-+# define __DECL_REG(name) union { uint64_t r ## name, e ## name; }
-+# define __align8__ __attribute__((aligned (8)))
-+#else
-+# define __DECL_REG(name) uint64_t r ## name
-+# define __align8__ FIXME
-+#endif
-+#define __x86_64___X86_64 1
-+""";
++/*
++ * World switch state
++ */
++struct vmx_assist_context {
++ uint32_t eip; /* execution pointer */
++ uint32_t esp; /* stack pointer */
++ uint32_t eflags; /* flags register */
++ uint32_t cr0;
++ uint32_t cr3; /* page table directory */
++ uint32_t cr4;
++ uint32_t idtr_limit; /* idt */
++ uint32_t idtr_base;
++ uint32_t gdtr_limit; /* gdt */
++ uint32_t gdtr_base;
++ uint32_t cs_sel; /* cs selector */
++ uint32_t cs_limit;
++ uint32_t cs_base;
++ union vmcs_arbytes cs_arbytes;
++ uint32_t ds_sel; /* ds selector */
++ uint32_t ds_limit;
++ uint32_t ds_base;
++ union vmcs_arbytes ds_arbytes;
++ uint32_t es_sel; /* es selector */
++ uint32_t es_limit;
++ uint32_t es_base;
++ union vmcs_arbytes es_arbytes;
++ uint32_t ss_sel; /* ss selector */
++ uint32_t ss_limit;
++ uint32_t ss_base;
++ union vmcs_arbytes ss_arbytes;
++ uint32_t fs_sel; /* fs selector */
++ uint32_t fs_limit;
++ uint32_t fs_base;
++ union vmcs_arbytes fs_arbytes;
++ uint32_t gs_sel; /* gs selector */
++ uint32_t gs_limit;
++ uint32_t gs_base;
++ union vmcs_arbytes gs_arbytes;
++ uint32_t tr_sel; /* task selector */
++ uint32_t tr_limit;
++ uint32_t tr_base;
++ union vmcs_arbytes tr_arbytes;
++ uint32_t ldtr_sel; /* ldtr selector */
++ uint32_t ldtr_limit;
++ uint32_t ldtr_base;
++ union vmcs_arbytes ldtr_arbytes;
+
-+# ia64
-+inttypes["ia64"] = {
-+ "unsigned long" : "__align8__ uint64_t",
-+ "long" : "__align8__ uint64_t",
-+ "xen_pfn_t" : "__align8__ uint64_t",
-+ "long double" : "__align16__ ldouble_t",
++ unsigned char rm_irqbase[2];
+};
-+header["ia64"] = """
-+#define __align8__ __attribute__((aligned (8)))
-+#define __align16__ __attribute__((aligned (16)))
-+typedef unsigned char ldouble_t[16];
-+""";
-+
-+
-+###########################################################################
-+# main
++typedef struct vmx_assist_context vmx_assist_context_t;
+
-+input = "";
-+output = "";
-+fileid = re.sub("[-.]", "_", "__FOREIGN_%s__" % outfile.upper());
++#endif /* __ASSEMBLY__ */
+
-+# read input header files
-+for name in infiles:
-+ f = open(name, "r");
-+ input += f.read();
-+ f.close();
++#endif /* _VMX_ASSIST_H_ */
+
-+# add header
-+output += """
+/*
-+ * public xen defines and struct for %s
-+ * generated by %s -- DO NOT EDIT
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/blkif.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/blkif.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,141 @@
++/******************************************************************************
++ * blkif.h
++ *
++ * Unified block-device I/O interface for Xen guest OSes.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2003-2004, Keir Fraser
+ */
+
-+#ifndef %s
-+#define %s 1
-+
-+""" % (arch, sys.argv[0], fileid, fileid)
-+
-+if arch in header:
-+ output += header[arch];
-+ output += "\n";
-+
-+# add defines to output
-+for line in re.findall("#define[^\n]+", input):
-+ for define in defines:
-+ regex = "#define\s+%s\\b" % define;
-+ match = re.search(regex, line);
-+ if None == match:
-+ continue;
-+ if define.upper()[0] == define[0]:
-+ replace = define + "_" + arch.upper();
-+ else:
-+ replace = define + "_" + arch;
-+ regex = "\\b%s\\b" % define;
-+ output += re.sub(regex, replace, line) + "\n";
-+output += "\n";
-+
-+# delete defines, comments, empty lines
-+input = re.sub("#define[^\n]+\n", "", input);
-+input = re.compile("/\*(.*?)\*/", re.S).sub("", input)
-+input = re.compile("\n\s*\n", re.S).sub("\n", input);
-+
-+# add structs to output
-+for struct in structs:
-+ regex = "struct\s+%s\s*\{(.*?)\n\};" % struct;
-+ match = re.search(regex, input, re.S)
-+ if None == match:
-+ output += "#define %s_has_no_%s 1\n" % (arch, struct);
-+ else:
-+ output += "struct %s_%s {%s\n};\n" % (struct, arch, match.group(1));
-+ output += "typedef struct %s_%s %s_%s_t;\n" % (struct, arch, struct, arch);
-+ output += "\n";
-+
-+# add footer
-+if arch in footer:
-+ output += footer[arch];
-+ output += "\n";
-+output += "#endif /* %s */\n" % fileid;
-+
-+# replace: defines
-+for define in defines:
-+ if define.upper()[0] == define[0]:
-+ replace = define + "_" + arch.upper();
-+ else:
-+ replace = define + "_" + arch;
-+ output = re.sub("\\b%s\\b" % define, replace, output);
-+
-+# replace: structs + struct typedefs
-+for struct in structs:
-+ output = re.sub("\\b(struct\s+%s)\\b" % struct, "\\1_%s" % arch, output);
-+ output = re.sub("\\b(%s)_t\\b" % struct, "\\1_%s_t" % arch, output);
-+
-+# replace: integer types
-+integers = inttypes[arch].keys();
-+integers.sort(lambda a, b: cmp(len(b),len(a)));
-+for type in integers:
-+ output = re.sub("\\b%s\\b" % type, inttypes[arch][type], output);
-+
-+# print results
-+f = open(outfile, "w");
-+f.write(output);
-+f.close;
++#ifndef __XEN_PUBLIC_IO_BLKIF_H__
++#define __XEN_PUBLIC_IO_BLKIF_H__
+
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/foreign/reference.size tmp-linux-2.6-xen.patch/include/xen/interface/foreign/reference.size
---- pristine-linux-2.6.18.2/include/xen/interface/foreign/reference.size 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/foreign/reference.size 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,17 @@
++#include "ring.h"
++#include "../grant_table.h"
+
-+structs | x86_32 x86_64 ia64
++/*
++ * Front->back notifications: When enqueuing a new request, sending a
++ * notification can be made conditional on req_event (i.e., the generic
++ * hold-off mechanism provided by the ring macros). Backends must set
++ * req_event appropriately (e.g., using RING_FINAL_CHECK_FOR_REQUESTS()).
++ *
++ * Back->front notifications: When enqueuing a new response, sending a
++ * notification can be made conditional on rsp_event (i.e., the generic
++ * hold-off mechanism provided by the ring macros). Frontends must set
++ * rsp_event appropriately (e.g., using RING_FINAL_CHECK_FOR_RESPONSES()).
++ */
+
-+start_info | 1104 1152 1152
-+trap_info | 8 16 -
-+pt_fpreg | - - 16
-+cpu_user_regs | 68 200 496
-+xen_ia64_boot_param | - - 96
-+ia64_tr_entry | - - 32
-+vcpu_extra_regs | - - 536
-+vcpu_guest_context | 2800 5168 1056
-+arch_vcpu_info | 24 16 0
-+vcpu_time_info | 32 32 32
-+vcpu_info | 64 64 48
-+arch_shared_info | 268 280 272
-+shared_info | 2584 3368 4384
-+
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/foreign/structs.py tmp-linux-2.6-xen.patch/include/xen/interface/foreign/structs.py
---- pristine-linux-2.6.18.2/include/xen/interface/foreign/structs.py 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/foreign/structs.py 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,54 @@
-+# configuration: what needs translation
++#ifndef blkif_vdev_t
++#define blkif_vdev_t uint16_t
++#endif
++#define blkif_sector_t uint64_t
+
-+structs = [ "start_info",
-+ "trap_info",
-+ "pt_fpreg",
-+ "cpu_user_regs",
-+ "xen_ia64_boot_param",
-+ "ia64_tr_entry",
-+ "vcpu_extra_regs",
-+ "vcpu_guest_context",
-+ "arch_vcpu_info",
-+ "vcpu_time_info",
-+ "vcpu_info",
-+ "arch_shared_info",
-+ "shared_info" ];
++/*
++ * REQUEST CODES.
++ */
++#define BLKIF_OP_READ 0
++#define BLKIF_OP_WRITE 1
++/*
++ * Recognised only if "feature-barrier" is present in backend xenbus info.
++ * The "feature-barrier" node contains a boolean indicating whether barrier
++ * requests are likely to succeed or fail. Either way, a barrier request
++ * may fail at any time with BLKIF_RSP_EOPNOTSUPP if it is unsupported by
++ * the underlying block-device hardware. The boolean simply indicates whether
++ * or not it is worthwhile for the frontend to attempt barrier requests.
++ * If a backend does not recognise BLKIF_OP_WRITE_BARRIER, it should *not*
++ * create the "feature-barrier" node!
++ */
++#define BLKIF_OP_WRITE_BARRIER 2
++/*
++ * Recognised if "feature-flush-cache" is present in backend xenbus
++ * info. A flush will ask the underlying storage hardware to flush its
++ * non-volatile caches as appropriate. The "feature-flush-cache" node
++ * contains a boolean indicating whether flush requests are likely to
++ * succeed or fail. Either way, a flush request may fail at any time
++ * with BLKIF_RSP_EOPNOTSUPP if it is unsupported by the underlying
++ * block-device hardware. The boolean simply indicates whether or not it
++ * is worthwhile for the frontend to attempt flushes. If a backend does
++ * not recognise BLKIF_OP_WRITE_FLUSH_CACHE, it should *not* create the
++ * "feature-flush-cache" node!
++ */
++#define BLKIF_OP_FLUSH_DISKCACHE 3
+
-+defines = [ "__i386__",
-+ "__x86_64__",
++/*
++ * Maximum scatter/gather segments per request.
++ * This is carefully chosen so that sizeof(blkif_ring_t) <= PAGE_SIZE.
++ * NB. This could be 12 if the ring indexes weren't stored in the same page.
++ */
++#define BLKIF_MAX_SEGMENTS_PER_REQUEST 11
+
-+ "FLAT_RING1_CS",
-+ "FLAT_RING1_DS",
-+ "FLAT_RING1_SS",
++struct blkif_request_segment {
++ grant_ref_t gref; /* reference to I/O buffer frame */
++ /* @first_sect: first sector in frame to transfer (inclusive). */
++ /* @last_sect: last sector in frame to transfer (inclusive). */
++ uint8_t first_sect, last_sect;
++};
+
-+ "FLAT_RING3_CS64",
-+ "FLAT_RING3_DS64",
-+ "FLAT_RING3_SS64",
-+ "FLAT_KERNEL_CS64",
-+ "FLAT_KERNEL_DS64",
-+ "FLAT_KERNEL_SS64",
++struct blkif_request {
++ uint8_t operation; /* BLKIF_OP_??? */
++ uint8_t nr_segments; /* number of segments */
++ blkif_vdev_t handle; /* only for read/write requests */
++ uint64_t id; /* private guest value, echoed in resp */
++ blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
++ struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++};
++typedef struct blkif_request blkif_request_t;
+
-+ "FLAT_KERNEL_CS",
-+ "FLAT_KERNEL_DS",
-+ "FLAT_KERNEL_SS",
++struct blkif_response {
++ uint64_t id; /* copied from request */
++ uint8_t operation; /* copied from request */
++ int16_t status; /* BLKIF_RSP_??? */
++};
++typedef struct blkif_response blkif_response_t;
+
-+ # x86_{32,64}
-+ "_VGCF_i387_valid",
-+ "VGCF_i387_valid",
-+ "_VGCF_in_kernel",
-+ "VGCF_in_kernel",
-+ "_VGCF_failsafe_disables_events",
-+ "VGCF_failsafe_disables_events",
-+ "_VGCF_syscall_disables_events",
-+ "VGCF_syscall_disables_events",
-+ "_VGCF_online",
-+ "VGCF_online",
++/*
++ * STATUS RETURN CODES.
++ */
++ /* Operation not supported (only happens on barrier writes). */
++#define BLKIF_RSP_EOPNOTSUPP -2
++ /* Operation failed for some unspecified reason (-EIO). */
++#define BLKIF_RSP_ERROR -1
++ /* Operation completed successfully. */
++#define BLKIF_RSP_OKAY 0
+
-+ # ia64
-+ "VGCF_EXTRA_REGS",
++/*
++ * Generate blkif ring structures and types.
++ */
+
-+ # all archs
-+ "xen_pfn_to_cr3",
-+ "MAX_VIRT_CPUS",
-+ "MAX_GUEST_CMDLINE" ];
++DEFINE_RING_TYPES(blkif, struct blkif_request, struct blkif_response);
+
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/grant_table.h tmp-linux-2.6-xen.patch/include/xen/interface/grant_table.h
---- pristine-linux-2.6.18.2/include/xen/interface/grant_table.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/grant_table.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,399 @@
++#define VDISK_CDROM 0x1
++#define VDISK_REMOVABLE 0x2
++#define VDISK_READONLY 0x4
++
++#endif /* __XEN_PUBLIC_IO_BLKIF_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/console.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/console.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,51 @@
+/******************************************************************************
-+ * grant_table.h
++ * console.h
+ *
-+ * Interface for granting foreign access to page frames, and receiving
-+ * page-ownership transfers.
++ * Console I/O interface for Xen guest OSes.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -99861,371 +138841,204 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (c) 2004, K A Fraser
++ * Copyright (c) 2005, Keir Fraser
+ */
+
-+#ifndef __XEN_PUBLIC_GRANT_TABLE_H__
-+#define __XEN_PUBLIC_GRANT_TABLE_H__
-+
++#ifndef __XEN_PUBLIC_IO_CONSOLE_H__
++#define __XEN_PUBLIC_IO_CONSOLE_H__
+
-+/***********************************
-+ * GRANT TABLE REPRESENTATION
-+ */
++typedef uint32_t XENCONS_RING_IDX;
+
-+/* Some rough guidelines on accessing and updating grant-table entries
-+ * in a concurrency-safe manner. For more information, Linux contains a
-+ * reference implementation for guest OSes (arch/xen/kernel/grant_table.c).
-+ *
-+ * NB. WMB is a no-op on current-generation x86 processors. However, a
-+ * compiler barrier will still be required.
-+ *
-+ * Introducing a valid entry into the grant table:
-+ * 1. Write ent->domid.
-+ * 2. Write ent->frame:
-+ * GTF_permit_access: Frame to which access is permitted.
-+ * GTF_accept_transfer: Pseudo-phys frame slot being filled by new
-+ * frame, or zero if none.
-+ * 3. Write memory barrier (WMB).
-+ * 4. Write ent->flags, inc. valid type.
-+ *
-+ * Invalidating an unused GTF_permit_access entry:
-+ * 1. flags = ent->flags.
-+ * 2. Observe that !(flags & (GTF_reading|GTF_writing)).
-+ * 3. Check result of SMP-safe CMPXCHG(&ent->flags, flags, 0).
-+ * NB. No need for WMB as reuse of entry is control-dependent on success of
-+ * step 3, and all architectures guarantee ordering of ctrl-dep writes.
-+ *
-+ * Invalidating an in-use GTF_permit_access entry:
-+ * This cannot be done directly. Request assistance from the domain controller
-+ * which can set a timeout on the use of a grant entry and take necessary
-+ * action. (NB. This is not yet implemented!).
-+ *
-+ * Invalidating an unused GTF_accept_transfer entry:
-+ * 1. flags = ent->flags.
-+ * 2. Observe that !(flags & GTF_transfer_committed). [*]
-+ * 3. Check result of SMP-safe CMPXCHG(&ent->flags, flags, 0).
-+ * NB. No need for WMB as reuse of entry is control-dependent on success of
-+ * step 3, and all architectures guarantee ordering of ctrl-dep writes.
-+ * [*] If GTF_transfer_committed is set then the grant entry is 'committed'.
-+ * The guest must /not/ modify the grant entry until the address of the
-+ * transferred frame is written. It is safe for the guest to spin waiting
-+ * for this to occur (detect by observing GTF_transfer_completed in
-+ * ent->flags).
-+ *
-+ * Invalidating a committed GTF_accept_transfer entry:
-+ * 1. Wait for (ent->flags & GTF_transfer_completed).
-+ *
-+ * Changing a GTF_permit_access from writable to read-only:
-+ * Use SMP-safe CMPXCHG to set GTF_readonly, while checking !GTF_writing.
-+ *
-+ * Changing a GTF_permit_access from read-only to writable:
-+ * Use SMP-safe bit-setting instruction.
-+ */
++#define MASK_XENCONS_IDX(idx, ring) ((idx) & (sizeof(ring)-1))
+
-+/*
-+ * A grant table comprises a packed array of grant entries in one or more
-+ * page frames shared between Xen and a guest.
-+ * [XEN]: This field is written by Xen and read by the sharing guest.
-+ * [GST]: This field is written by the guest and read by Xen.
-+ */
-+struct grant_entry {
-+ /* GTF_xxx: various type and flag information. [XEN,GST] */
-+ uint16_t flags;
-+ /* The domain being granted foreign privileges. [GST] */
-+ domid_t domid;
-+ /*
-+ * GTF_permit_access: Frame that @domid is allowed to map and access. [GST]
-+ * GTF_accept_transfer: Frame whose ownership transferred by @domid. [XEN]
-+ */
-+ uint32_t frame;
++struct xencons_interface {
++ char in[1024];
++ char out[2048];
++ XENCONS_RING_IDX in_cons, in_prod;
++ XENCONS_RING_IDX out_cons, out_prod;
+};
-+typedef struct grant_entry grant_entry_t;
+
-+/*
-+ * Type of grant entry.
-+ * GTF_invalid: This grant entry grants no privileges.
-+ * GTF_permit_access: Allow @domid to map/access @frame.
-+ * GTF_accept_transfer: Allow @domid to transfer ownership of one page frame
-+ * to this guest. Xen writes the page number to @frame.
-+ */
-+#define GTF_invalid (0U<<0)
-+#define GTF_permit_access (1U<<0)
-+#define GTF_accept_transfer (2U<<0)
-+#define GTF_type_mask (3U<<0)
++#endif /* __XEN_PUBLIC_IO_CONSOLE_H__ */
+
+/*
-+ * Subflags for GTF_permit_access.
-+ * GTF_readonly: Restrict @domid to read-only mappings and accesses. [GST]
-+ * GTF_reading: Grant entry is currently mapped for reading by @domid. [XEN]
-+ * GTF_writing: Grant entry is currently mapped for writing by @domid. [XEN]
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
+ */
-+#define _GTF_readonly (2)
-+#define GTF_readonly (1U<<_GTF_readonly)
-+#define _GTF_reading (3)
-+#define GTF_reading (1U<<_GTF_reading)
-+#define _GTF_writing (4)
-+#define GTF_writing (1U<<_GTF_writing)
-+
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/fbif.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/fbif.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,176 @@
+/*
-+ * Subflags for GTF_accept_transfer:
-+ * GTF_transfer_committed: Xen sets this flag to indicate that it is committed
-+ * to transferring ownership of a page frame. When a guest sees this flag
-+ * it must /not/ modify the grant entry until GTF_transfer_completed is
-+ * set by Xen.
-+ * GTF_transfer_completed: It is safe for the guest to spin-wait on this flag
-+ * after reading GTF_transfer_committed. Xen will always write the frame
-+ * address, followed by ORing this flag, in a timely manner.
++ * fbif.h -- Xen virtual frame buffer device
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (C) 2005 Anthony Liguori <aliguori at us.ibm.com>
++ * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster <armbru at redhat.com>
+ */
-+#define _GTF_transfer_committed (2)
-+#define GTF_transfer_committed (1U<<_GTF_transfer_committed)
-+#define _GTF_transfer_completed (3)
-+#define GTF_transfer_completed (1U<<_GTF_transfer_completed)
+
++#ifndef __XEN_PUBLIC_IO_FBIF_H__
++#define __XEN_PUBLIC_IO_FBIF_H__
+
-+/***********************************
-+ * GRANT TABLE QUERIES AND USES
-+ */
++/* Out events (frontend -> backend) */
+
+/*
-+ * Reference to a grant entry in a specified domain's grant table.
++ * Out events may be sent only when requested by backend, and receipt
++ * of an unknown out event is an error.
+ */
-+typedef uint32_t grant_ref_t;
+
++/* Event type 1 currently not used */
+/*
-+ * Handle to track a mapping created via a grant reference.
++ * Framebuffer update notification event
++ * Capable frontend sets feature-update in xenstore.
++ * Backend requests it by setting request-update in xenstore.
+ */
-+typedef uint32_t grant_handle_t;
++#define XENFB_TYPE_UPDATE 2
+
-+/*
-+ * GNTTABOP_map_grant_ref: Map the grant entry (<dom>,<ref>) for access
-+ * by devices and/or host CPUs. If successful, <handle> is a tracking number
-+ * that must be presented later to destroy the mapping(s). On error, <handle>
-+ * is a negative status code.
-+ * NOTES:
-+ * 1. If GNTMAP_device_map is specified then <dev_bus_addr> is the address
-+ * via which I/O devices may access the granted frame.
-+ * 2. If GNTMAP_host_map is specified then a mapping will be added at
-+ * either a host virtual address in the current address space, or at
-+ * a PTE at the specified machine address. The type of mapping to
-+ * perform is selected through the GNTMAP_contains_pte flag, and the
-+ * address is specified in <host_addr>.
-+ * 3. Mappings should only be destroyed via GNTTABOP_unmap_grant_ref. If a
-+ * host mapping is destroyed by other means then it is *NOT* guaranteed
-+ * to be accounted to the correct grant reference!
-+ */
-+#define GNTTABOP_map_grant_ref 0
-+struct gnttab_map_grant_ref {
-+ /* IN parameters. */
-+ uint64_t host_addr;
-+ uint32_t flags; /* GNTMAP_* */
-+ grant_ref_t ref;
-+ domid_t dom;
-+ /* OUT parameters. */
-+ int16_t status; /* GNTST_* */
-+ grant_handle_t handle;
-+ uint64_t dev_bus_addr;
++struct xenfb_update
++{
++ uint8_t type; /* XENFB_TYPE_UPDATE */
++ int32_t x; /* source x */
++ int32_t y; /* source y */
++ int32_t width; /* rect width */
++ int32_t height; /* rect height */
+};
-+typedef struct gnttab_map_grant_ref gnttab_map_grant_ref_t;
-+DEFINE_XEN_GUEST_HANDLE(gnttab_map_grant_ref_t);
+
+/*
-+ * GNTTABOP_unmap_grant_ref: Destroy one or more grant-reference mappings
-+ * tracked by <handle>. If <host_addr> or <dev_bus_addr> is zero, that
-+ * field is ignored. If non-zero, they must refer to a device/host mapping
-+ * that is tracked by <handle>
-+ * NOTES:
-+ * 1. The call may fail in an undefined manner if either mapping is not
-+ * tracked by <handle>.
-+ * 3. After executing a batch of unmaps, it is guaranteed that no stale
-+ * mappings will remain in the device or host TLBs.
++ * Framebuffer resize notification event
++ * Capable backend sets feature-resize in xenstore.
+ */
-+#define GNTTABOP_unmap_grant_ref 1
-+struct gnttab_unmap_grant_ref {
-+ /* IN parameters. */
-+ uint64_t host_addr;
-+ uint64_t dev_bus_addr;
-+ grant_handle_t handle;
-+ /* OUT parameters. */
-+ int16_t status; /* GNTST_* */
-+};
-+typedef struct gnttab_unmap_grant_ref gnttab_unmap_grant_ref_t;
-+DEFINE_XEN_GUEST_HANDLE(gnttab_unmap_grant_ref_t);
++#define XENFB_TYPE_RESIZE 3
+
-+/*
-+ * GNTTABOP_setup_table: Set up a grant table for <dom> comprising at least
-+ * <nr_frames> pages. The frame addresses are written to the <frame_list>.
-+ * Only <nr_frames> addresses are written, even if the table is larger.
-+ * NOTES:
-+ * 1. <dom> may be specified as DOMID_SELF.
-+ * 2. Only a sufficiently-privileged domain may specify <dom> != DOMID_SELF.
-+ * 3. Xen may not support more than a single grant-table page per domain.
-+ */
-+#define GNTTABOP_setup_table 2
-+struct gnttab_setup_table {
-+ /* IN parameters. */
-+ domid_t dom;
-+ uint32_t nr_frames;
-+ /* OUT parameters. */
-+ int16_t status; /* GNTST_* */
-+ XEN_GUEST_HANDLE(ulong) frame_list;
++struct xenfb_resize
++{
++ uint8_t type; /* XENFB_TYPE_RESIZE */
++ int32_t width; /* width in pixels */
++ int32_t height; /* height in pixels */
++ int32_t stride; /* stride in bytes */
++ int32_t depth; /* depth in bits */
++ int32_t offset; /* offset of the framebuffer in bytes */
+};
-+typedef struct gnttab_setup_table gnttab_setup_table_t;
-+DEFINE_XEN_GUEST_HANDLE(gnttab_setup_table_t);
+
-+/*
-+ * GNTTABOP_dump_table: Dump the contents of the grant table to the
-+ * xen console. Debugging use only.
-+ */
-+#define GNTTABOP_dump_table 3
-+struct gnttab_dump_table {
-+ /* IN parameters. */
-+ domid_t dom;
-+ /* OUT parameters. */
-+ int16_t status; /* GNTST_* */
-+};
-+typedef struct gnttab_dump_table gnttab_dump_table_t;
-+DEFINE_XEN_GUEST_HANDLE(gnttab_dump_table_t);
++#define XENFB_OUT_EVENT_SIZE 40
+
-+/*
-+ * GNTTABOP_transfer_grant_ref: Transfer <frame> to a foreign domain. The
-+ * foreign domain has previously registered its interest in the transfer via
-+ * <domid, ref>.
-+ *
-+ * Note that, even if the transfer fails, the specified page no longer belongs
-+ * to the calling domain *unless* the error is GNTST_bad_page.
-+ */
-+#define GNTTABOP_transfer 4
-+struct gnttab_transfer {
-+ /* IN parameters. */
-+ xen_pfn_t mfn;
-+ domid_t domid;
-+ grant_ref_t ref;
-+ /* OUT parameters. */
-+ int16_t status;
++union xenfb_out_event
++{
++ uint8_t type;
++ struct xenfb_update update;
++ struct xenfb_resize resize;
++ char pad[XENFB_OUT_EVENT_SIZE];
+};
-+typedef struct gnttab_transfer gnttab_transfer_t;
-+DEFINE_XEN_GUEST_HANDLE(gnttab_transfer_t);
+
++/* In events (backend -> frontend) */
+
+/*
-+ * GNTTABOP_copy: Hypervisor based copy
-+ * source and destinations can be eithers MFNs or, for foreign domains,
-+ * grant references. the foreign domain has to grant read/write access
-+ * in its grant table.
-+ *
-+ * The flags specify what type source and destinations are (either MFN
-+ * or grant reference).
-+ *
-+ * Note that this can also be used to copy data between two domains
-+ * via a third party if the source and destination domains had previously
-+ * grant appropriate access to their pages to the third party.
-+ *
-+ * source_offset specifies an offset in the source frame, dest_offset
-+ * the offset in the target frame and len specifies the number of
-+ * bytes to be copied.
++ * Frontends should ignore unknown in events.
+ */
+
-+#define _GNTCOPY_source_gref (0)
-+#define GNTCOPY_source_gref (1<<_GNTCOPY_source_gref)
-+#define _GNTCOPY_dest_gref (1)
-+#define GNTCOPY_dest_gref (1<<_GNTCOPY_dest_gref)
++/*
++ * Framebuffer refresh period advice
++ * Backend sends it to advise the frontend their preferred period of
++ * refresh. Frontends that keep the framebuffer constantly up-to-date
++ * just ignore it. Frontends that use the advice should immediately
++ * refresh the framebuffer (and send an update notification event if
++ * those have been requested), then use the update frequency to guide
++ * their periodical refreshs.
++ */
++#define XENFB_TYPE_REFRESH_PERIOD 1
++#define XENFB_NO_REFRESH 0
++
++struct xenfb_refresh_period
++{
++ uint8_t type; /* XENFB_TYPE_UPDATE_PERIOD */
++ uint32_t period; /* period of refresh, in ms,
++ * XENFB_NO_REFRESH if no refresh is needed */
++};
+
-+#define GNTTABOP_copy 5
-+typedef struct gnttab_copy {
-+ /* IN parameters. */
-+ struct {
-+ union {
-+ grant_ref_t ref;
-+ xen_pfn_t gmfn;
-+ } u;
-+ domid_t domid;
-+ uint16_t offset;
-+ } source, dest;
-+ uint16_t len;
-+ uint16_t flags; /* GNTCOPY_* */
-+ /* OUT parameters. */
-+ int16_t status;
-+} gnttab_copy_t;
-+DEFINE_XEN_GUEST_HANDLE(gnttab_copy_t);
++#define XENFB_IN_EVENT_SIZE 40
+
-+/*
-+ * GNTTABOP_query_size: Query the current and maximum sizes of the shared
-+ * grant table.
-+ * NOTES:
-+ * 1. <dom> may be specified as DOMID_SELF.
-+ * 2. Only a sufficiently-privileged domain may specify <dom> != DOMID_SELF.
-+ */
-+#define GNTTABOP_query_size 6
-+struct gnttab_query_size {
-+ /* IN parameters. */
-+ domid_t dom;
-+ /* OUT parameters. */
-+ uint32_t nr_frames;
-+ uint32_t max_nr_frames;
-+ int16_t status; /* GNTST_* */
++union xenfb_in_event
++{
++ uint8_t type;
++ struct xenfb_refresh_period refresh_period;
++ char pad[XENFB_IN_EVENT_SIZE];
+};
-+typedef struct gnttab_query_size gnttab_query_size_t;
-+DEFINE_XEN_GUEST_HANDLE(gnttab_query_size_t);
+
++/* shared page */
+
-+/*
-+ * Bitfield values for update_pin_status.flags.
-+ */
-+ /* Map the grant entry for access by I/O devices. */
-+#define _GNTMAP_device_map (0)
-+#define GNTMAP_device_map (1<<_GNTMAP_device_map)
-+ /* Map the grant entry for access by host CPUs. */
-+#define _GNTMAP_host_map (1)
-+#define GNTMAP_host_map (1<<_GNTMAP_host_map)
-+ /* Accesses to the granted frame will be restricted to read-only access. */
-+#define _GNTMAP_readonly (2)
-+#define GNTMAP_readonly (1<<_GNTMAP_readonly)
-+ /*
-+ * GNTMAP_host_map subflag:
-+ * 0 => The host mapping is usable only by the guest OS.
-+ * 1 => The host mapping is usable by guest OS + current application.
-+ */
-+#define _GNTMAP_application_map (3)
-+#define GNTMAP_application_map (1<<_GNTMAP_application_map)
++#define XENFB_IN_RING_SIZE 1024
++#define XENFB_IN_RING_LEN (XENFB_IN_RING_SIZE / XENFB_IN_EVENT_SIZE)
++#define XENFB_IN_RING_OFFS 1024
++#define XENFB_IN_RING(page) \
++ ((union xenfb_in_event *)((char *)(page) + XENFB_IN_RING_OFFS))
++#define XENFB_IN_RING_REF(page, idx) \
++ (XENFB_IN_RING((page))[(idx) % XENFB_IN_RING_LEN])
+
-+ /*
-+ * GNTMAP_contains_pte subflag:
-+ * 0 => This map request contains a host virtual address.
-+ * 1 => This map request contains the machine addess of the PTE to update.
-+ */
-+#define _GNTMAP_contains_pte (4)
-+#define GNTMAP_contains_pte (1<<_GNTMAP_contains_pte)
++#define XENFB_OUT_RING_SIZE 2048
++#define XENFB_OUT_RING_LEN (XENFB_OUT_RING_SIZE / XENFB_OUT_EVENT_SIZE)
++#define XENFB_OUT_RING_OFFS (XENFB_IN_RING_OFFS + XENFB_IN_RING_SIZE)
++#define XENFB_OUT_RING(page) \
++ ((union xenfb_out_event *)((char *)(page) + XENFB_OUT_RING_OFFS))
++#define XENFB_OUT_RING_REF(page, idx) \
++ (XENFB_OUT_RING((page))[(idx) % XENFB_OUT_RING_LEN])
+
-+/*
-+ * Values for error status returns. All errors are -ve.
-+ */
-+#define GNTST_okay (0) /* Normal return. */
-+#define GNTST_general_error (-1) /* General undefined error. */
-+#define GNTST_bad_domain (-2) /* Unrecognsed domain id. */
-+#define GNTST_bad_gntref (-3) /* Unrecognised or inappropriate gntref. */
-+#define GNTST_bad_handle (-4) /* Unrecognised or inappropriate handle. */
-+#define GNTST_bad_virt_addr (-5) /* Inappropriate virtual address to map. */
-+#define GNTST_bad_dev_addr (-6) /* Inappropriate device address to unmap.*/
-+#define GNTST_no_device_space (-7) /* Out of space in I/O MMU. */
-+#define GNTST_permission_denied (-8) /* Not enough privilege for operation. */
-+#define GNTST_bad_page (-9) /* Specified page was invalid for op. */
-+#define GNTST_bad_copy_arg (-10) /* copy arguments cross page boundary */
++struct xenfb_page
++{
++ uint32_t in_cons, in_prod;
++ uint32_t out_cons, out_prod;
+
-+#define GNTTABOP_error_msgs { \
-+ "okay", \
-+ "undefined error", \
-+ "unrecognised domain id", \
-+ "invalid grant reference", \
-+ "invalid mapping handle", \
-+ "invalid virtual address", \
-+ "invalid device address", \
-+ "no spare translation slot in the I/O MMU", \
-+ "permission denied", \
-+ "bad page", \
-+ "copy arguments cross page boundary" \
-+}
++ int32_t width; /* the width of the framebuffer (in pixels) */
++ int32_t height; /* the height of the framebuffer (in pixels) */
++ uint32_t line_length; /* the length of a row of pixels (in bytes) */
++ uint32_t mem_length; /* the length of the framebuffer (in bytes) */
++ uint8_t depth; /* the depth of a pixel (in bits) */
++
++ /*
++ * Framebuffer page directory
++ *
++ * Each directory page holds PAGE_SIZE / sizeof(*pd)
++ * framebuffer pages, and can thus map up to PAGE_SIZE *
++ * PAGE_SIZE / sizeof(*pd) bytes. With PAGE_SIZE == 4096 and
++ * sizeof(unsigned long) == 4/8, that's 4 Megs 32 bit and 2 Megs
++ * 64 bit. 256 directories give enough room for a 512 Meg
++ * framebuffer with a max resolution of 12,800x10,240. Should
++ * be enough for a while with room leftover for expansion.
++ */
++ unsigned long pd[256];
++};
+
-+#endif /* __XEN_PUBLIC_GRANT_TABLE_H__ */
++/*
++ * Wart: xenkbd needs to know default resolution. Put it here until a
++ * better solution is found, but don't leak it to the backend.
++ */
++#ifdef __KERNEL__
++#define XENFB_WIDTH 800
++#define XENFB_HEIGHT 600
++#define XENFB_DEPTH 32
++#endif
++
++#endif
+
+/*
+ * Local variables:
@@ -100236,12 +139049,15 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/hvm/e820.h tmp-linux-2.6-xen.patch/include/xen/interface/hvm/e820.h
---- pristine-linux-2.6.18.2/include/xen/interface/hvm/e820.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/hvm/e820.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,34 @@
-+
-+/*
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/fsif.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/fsif.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,191 @@
++/******************************************************************************
++ * fsif.h
++ *
++ * Interface to FS level split device drivers.
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -100259,29 +139075,181 @@
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2007, Grzegorz Milos, <gm281 at cam.ac.uk>.
+ */
+
-+#ifndef __XEN_PUBLIC_HVM_E820_H__
-+#define __XEN_PUBLIC_HVM_E820_H__
++#ifndef __XEN_PUBLIC_IO_FSIF_H__
++#define __XEN_PUBLIC_IO_FSIF_H__
+
-+/* E820 location in HVM virtual address space. */
-+#define HVM_E820_PAGE 0x00090000
-+#define HVM_E820_NR_OFFSET 0x000001E8
-+#define HVM_E820_OFFSET 0x000002D0
++#include "ring.h"
++#include "../grant_table.h"
+
-+#define HVM_BELOW_4G_RAM_END 0xF0000000
-+#define HVM_BELOW_4G_MMIO_START HVM_BELOW_4G_RAM_END
-+#define HVM_BELOW_4G_MMIO_LENGTH ((1ULL << 32) - HVM_BELOW_4G_MMIO_START)
++#define REQ_FILE_OPEN 1
++#define REQ_FILE_CLOSE 2
++#define REQ_FILE_READ 3
++#define REQ_FILE_WRITE 4
++#define REQ_STAT 5
++#define REQ_FILE_TRUNCATE 6
++#define REQ_REMOVE 7
++#define REQ_RENAME 8
++#define REQ_CREATE 9
++#define REQ_DIR_LIST 10
++#define REQ_CHMOD 11
++#define REQ_FS_SPACE 12
++#define REQ_FILE_SYNC 13
+
-+#endif /* __XEN_PUBLIC_HVM_E820_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/hvm/hvm_info_table.h tmp-linux-2.6-xen.patch/include/xen/interface/hvm/hvm_info_table.h
---- pristine-linux-2.6.18.2/include/xen/interface/hvm/hvm_info_table.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/hvm/hvm_info_table.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,41 @@
-+/******************************************************************************
-+ * hvm/hvm_info_table.h
-+ *
-+ * HVM parameter and information table, written into guest memory map.
++struct fsif_open_request {
++ grant_ref_t gref;
++};
++
++struct fsif_close_request {
++ uint32_t fd;
++};
++
++struct fsif_read_request {
++ uint32_t fd;
++ int32_t pad;
++ uint64_t len;
++ uint64_t offset;
++ grant_ref_t grefs[1]; /* Variable length */
++};
++
++struct fsif_write_request {
++ uint32_t fd;
++ int32_t pad;
++ uint64_t len;
++ uint64_t offset;
++ grant_ref_t grefs[1]; /* Variable length */
++};
++
++struct fsif_stat_request {
++ uint32_t fd;
++};
++
++/* This structure is a copy of some fields from stat structure, returned
++ * via the ring. */
++struct fsif_stat_response {
++ int32_t stat_mode;
++ uint32_t stat_uid;
++ uint32_t stat_gid;
++ int32_t stat_ret;
++ int64_t stat_size;
++ int64_t stat_atime;
++ int64_t stat_mtime;
++ int64_t stat_ctime;
++};
++
++struct fsif_truncate_request {
++ uint32_t fd;
++ int32_t pad;
++ int64_t length;
++};
++
++struct fsif_remove_request {
++ grant_ref_t gref;
++};
++
++struct fsif_rename_request {
++ uint16_t old_name_offset;
++ uint16_t new_name_offset;
++ grant_ref_t gref;
++};
++
++struct fsif_create_request {
++ int8_t directory;
++ int8_t pad;
++ int16_t pad2;
++ int32_t mode;
++ grant_ref_t gref;
++};
++
++struct fsif_list_request {
++ uint32_t offset;
++ grant_ref_t gref;
++};
++
++#define NR_FILES_SHIFT 0
++#define NR_FILES_SIZE 16 /* 16 bits for the number of files mask */
++#define NR_FILES_MASK (((1ULL << NR_FILES_SIZE) - 1) << NR_FILES_SHIFT)
++#define ERROR_SIZE 32 /* 32 bits for the error mask */
++#define ERROR_SHIFT (NR_FILES_SIZE + NR_FILES_SHIFT)
++#define ERROR_MASK (((1ULL << ERROR_SIZE) - 1) << ERROR_SHIFT)
++#define HAS_MORE_SHIFT (ERROR_SHIFT + ERROR_SIZE)
++#define HAS_MORE_FLAG (1ULL << HAS_MORE_SHIFT)
++
++struct fsif_chmod_request {
++ uint32_t fd;
++ int32_t mode;
++};
++
++struct fsif_space_request {
++ grant_ref_t gref;
++};
++
++struct fsif_sync_request {
++ uint32_t fd;
++};
++
++
++/* FS operation request */
++struct fsif_request {
++ uint8_t type; /* Type of the request */
++ uint8_t pad;
++ uint16_t id; /* Request ID, copied to the response */
++ uint32_t pad2;
++ union {
++ struct fsif_open_request fopen;
++ struct fsif_close_request fclose;
++ struct fsif_read_request fread;
++ struct fsif_write_request fwrite;
++ struct fsif_stat_request fstat;
++ struct fsif_truncate_request ftruncate;
++ struct fsif_remove_request fremove;
++ struct fsif_rename_request frename;
++ struct fsif_create_request fcreate;
++ struct fsif_list_request flist;
++ struct fsif_chmod_request fchmod;
++ struct fsif_space_request fspace;
++ struct fsif_sync_request fsync;
++ } u;
++};
++typedef struct fsif_request fsif_request_t;
++
++/* FS operation response */
++struct fsif_response {
++ uint16_t id;
++ uint16_t pad1;
++ uint32_t pad2;
++ union {
++ uint64_t ret_val;
++ struct fsif_stat_response fstat;
++ };
++};
++
++typedef struct fsif_response fsif_response_t;
++
++#define FSIF_RING_ENTRY_SIZE 64
++
++#define FSIF_NR_READ_GNTS ((FSIF_RING_ENTRY_SIZE - sizeof(struct fsif_read_request)) / \
++ sizeof(grant_ref_t) + 1)
++#define FSIF_NR_WRITE_GNTS ((FSIF_RING_ENTRY_SIZE - sizeof(struct fsif_write_request)) / \
++ sizeof(grant_ref_t) + 1)
++
++DEFINE_RING_TYPES(fsif, struct fsif_request, struct fsif_response);
++
++#define STATE_INITIALISED "init"
++#define STATE_READY "ready"
++
++
++
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/kbdif.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/kbdif.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,132 @@
++/*
++ * kbdif.h -- Xen virtual keyboard/mouse
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -100300,30 +139268,127 @@
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (C) 2005 Anthony Liguori <aliguori at us.ibm.com>
++ * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster <armbru at redhat.com>
+ */
+
-+#ifndef __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__
-+#define __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__
++#ifndef __XEN_PUBLIC_IO_KBDIF_H__
++#define __XEN_PUBLIC_IO_KBDIF_H__
+
-+#define HVM_INFO_PFN 0x09F
-+#define HVM_INFO_OFFSET 0x800
-+#define HVM_INFO_PADDR ((HVM_INFO_PFN << 12) + HVM_INFO_OFFSET)
++/* In events (backend -> frontend) */
+
-+struct hvm_info_table {
-+ char signature[8]; /* "HVM INFO" */
-+ uint32_t length;
-+ uint8_t checksum;
-+ uint8_t acpi_enabled;
-+ uint8_t apic_mode;
-+ uint32_t nr_vcpus;
++/*
++ * Frontends should ignore unknown in events.
++ */
++
++/* Pointer movement event */
++#define XENKBD_TYPE_MOTION 1
++/* Event type 2 currently not used */
++/* Key event (includes pointer buttons) */
++#define XENKBD_TYPE_KEY 3
++/*
++ * Pointer position event
++ * Capable backend sets feature-abs-pointer in xenstore.
++ * Frontend requests ot instead of XENKBD_TYPE_MOTION by setting
++ * request-abs-update in xenstore.
++ */
++#define XENKBD_TYPE_POS 4
++
++struct xenkbd_motion
++{
++ uint8_t type; /* XENKBD_TYPE_MOTION */
++ int32_t rel_x; /* relative X motion */
++ int32_t rel_y; /* relative Y motion */
++ int32_t rel_z; /* relative Z motion (wheel) */
+};
+
-+#endif /* __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/hvm/hvm_op.h tmp-linux-2.6-xen.patch/include/xen/interface/hvm/hvm_op.h
---- pristine-linux-2.6.18.2/include/xen/interface/hvm/hvm_op.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/hvm/hvm_op.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,76 @@
++struct xenkbd_key
++{
++ uint8_t type; /* XENKBD_TYPE_KEY */
++ uint8_t pressed; /* 1 if pressed; 0 otherwise */
++ uint32_t keycode; /* KEY_* from linux/input.h */
++};
++
++struct xenkbd_position
++{
++ uint8_t type; /* XENKBD_TYPE_POS */
++ int32_t abs_x; /* absolute X position (in FB pixels) */
++ int32_t abs_y; /* absolute Y position (in FB pixels) */
++ int32_t rel_z; /* relative Z motion (wheel) */
++};
++
++#define XENKBD_IN_EVENT_SIZE 40
++
++union xenkbd_in_event
++{
++ uint8_t type;
++ struct xenkbd_motion motion;
++ struct xenkbd_key key;
++ struct xenkbd_position pos;
++ char pad[XENKBD_IN_EVENT_SIZE];
++};
++
++/* Out events (frontend -> backend) */
++
++/*
++ * Out events may be sent only when requested by backend, and receipt
++ * of an unknown out event is an error.
++ * No out events currently defined.
++ */
++
++#define XENKBD_OUT_EVENT_SIZE 40
++
++union xenkbd_out_event
++{
++ uint8_t type;
++ char pad[XENKBD_OUT_EVENT_SIZE];
++};
++
++/* shared page */
++
++#define XENKBD_IN_RING_SIZE 2048
++#define XENKBD_IN_RING_LEN (XENKBD_IN_RING_SIZE / XENKBD_IN_EVENT_SIZE)
++#define XENKBD_IN_RING_OFFS 1024
++#define XENKBD_IN_RING(page) \
++ ((union xenkbd_in_event *)((char *)(page) + XENKBD_IN_RING_OFFS))
++#define XENKBD_IN_RING_REF(page, idx) \
++ (XENKBD_IN_RING((page))[(idx) % XENKBD_IN_RING_LEN])
++
++#define XENKBD_OUT_RING_SIZE 1024
++#define XENKBD_OUT_RING_LEN (XENKBD_OUT_RING_SIZE / XENKBD_OUT_EVENT_SIZE)
++#define XENKBD_OUT_RING_OFFS (XENKBD_IN_RING_OFFS + XENKBD_IN_RING_SIZE)
++#define XENKBD_OUT_RING(page) \
++ ((union xenkbd_out_event *)((char *)(page) + XENKBD_OUT_RING_OFFS))
++#define XENKBD_OUT_RING_REF(page, idx) \
++ (XENKBD_OUT_RING((page))[(idx) % XENKBD_OUT_RING_LEN])
++
++struct xenkbd_page
++{
++ uint32_t in_cons, in_prod;
++ uint32_t out_cons, out_prod;
++};
++
++#endif
++
+/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/netif.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/netif.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,205 @@
++/******************************************************************************
++ * netif.h
++ *
++ * Unified network-device I/O interface for Xen guest OSes.
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -100341,72 +139406,196 @@
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2003-2004, Keir Fraser
+ */
+
-+#ifndef __XEN_PUBLIC_HVM_HVM_OP_H__
-+#define __XEN_PUBLIC_HVM_HVM_OP_H__
++#ifndef __XEN_PUBLIC_IO_NETIF_H__
++#define __XEN_PUBLIC_IO_NETIF_H__
+
-+/* Get/set subcommands: extra argument == pointer to xen_hvm_param struct. */
-+#define HVMOP_set_param 0
-+#define HVMOP_get_param 1
-+struct xen_hvm_param {
-+ domid_t domid; /* IN */
-+ uint32_t index; /* IN */
-+ uint64_t value; /* IN/OUT */
++#include "ring.h"
++#include "../grant_table.h"
++
++/*
++ * Notifications after enqueuing any type of message should be conditional on
++ * the appropriate req_event or rsp_event field in the shared ring.
++ * If the client sends notification for rx requests then it should specify
++ * feature 'feature-rx-notify' via xenbus. Otherwise the backend will assume
++ * that it cannot safely queue packets (as it may not be kicked to send them).
++ */
++
++/*
++ * This is the 'wire' format for packets:
++ * Request 1: netif_tx_request -- NETTXF_* (any flags)
++ * [Request 2: netif_tx_extra] (only if request 1 has NETTXF_extra_info)
++ * [Request 3: netif_tx_extra] (only if request 2 has XEN_NETIF_EXTRA_MORE)
++ * Request 4: netif_tx_request -- NETTXF_more_data
++ * Request 5: netif_tx_request -- NETTXF_more_data
++ * ...
++ * Request N: netif_tx_request -- 0
++ */
++
++/* Protocol checksum field is blank in the packet (hardware offload)? */
++#define _NETTXF_csum_blank (0)
++#define NETTXF_csum_blank (1U<<_NETTXF_csum_blank)
++
++/* Packet data has been validated against protocol checksum. */
++#define _NETTXF_data_validated (1)
++#define NETTXF_data_validated (1U<<_NETTXF_data_validated)
++
++/* Packet continues in the next request descriptor. */
++#define _NETTXF_more_data (2)
++#define NETTXF_more_data (1U<<_NETTXF_more_data)
++
++/* Packet to be followed by extra descriptor(s). */
++#define _NETTXF_extra_info (3)
++#define NETTXF_extra_info (1U<<_NETTXF_extra_info)
++
++struct netif_tx_request {
++ grant_ref_t gref; /* Reference to buffer page */
++ uint16_t offset; /* Offset within buffer page */
++ uint16_t flags; /* NETTXF_* */
++ uint16_t id; /* Echoed in response message. */
++ uint16_t size; /* Packet size in bytes. */
+};
-+typedef struct xen_hvm_param xen_hvm_param_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_hvm_param_t);
++typedef struct netif_tx_request netif_tx_request_t;
+
-+/* Set the logical level of one of a domain's PCI INTx wires. */
-+#define HVMOP_set_pci_intx_level 2
-+struct xen_hvm_set_pci_intx_level {
-+ /* Domain to be updated. */
-+ domid_t domid;
-+ /* PCI INTx identification in PCI topology (domain:bus:device:intx). */
-+ uint8_t domain, bus, device, intx;
-+ /* Assertion level (0 = unasserted, 1 = asserted). */
-+ uint8_t level;
++/* Types of netif_extra_info descriptors. */
++#define XEN_NETIF_EXTRA_TYPE_NONE (0) /* Never used - invalid */
++#define XEN_NETIF_EXTRA_TYPE_GSO (1) /* u.gso */
++#define XEN_NETIF_EXTRA_TYPE_MCAST_ADD (2) /* u.mcast */
++#define XEN_NETIF_EXTRA_TYPE_MCAST_DEL (3) /* u.mcast */
++#define XEN_NETIF_EXTRA_TYPE_MAX (4)
++
++/* netif_extra_info flags. */
++#define _XEN_NETIF_EXTRA_FLAG_MORE (0)
++#define XEN_NETIF_EXTRA_FLAG_MORE (1U<<_XEN_NETIF_EXTRA_FLAG_MORE)
++
++/* GSO types - only TCPv4 currently supported. */
++#define XEN_NETIF_GSO_TYPE_TCPV4 (1)
++
++/*
++ * This structure needs to fit within both netif_tx_request and
++ * netif_rx_response for compatibility.
++ */
++struct netif_extra_info {
++ uint8_t type; /* XEN_NETIF_EXTRA_TYPE_* */
++ uint8_t flags; /* XEN_NETIF_EXTRA_FLAG_* */
++
++ union {
++ /*
++ * XEN_NETIF_EXTRA_TYPE_GSO:
++ */
++ struct {
++ /*
++ * Maximum payload size of each segment. For example, for TCP this
++ * is just the path MSS.
++ */
++ uint16_t size;
++
++ /*
++ * GSO type. This determines the protocol of the packet and any
++ * extra features required to segment the packet properly.
++ */
++ uint8_t type; /* XEN_NETIF_GSO_TYPE_* */
++
++ /* Future expansion. */
++ uint8_t pad;
++
++ /*
++ * GSO features. This specifies any extra GSO features required
++ * to process this packet, such as ECN support for TCPv4.
++ */
++ uint16_t features; /* XEN_NETIF_GSO_FEAT_* */
++ } gso;
++
++ /*
++ * XEN_NETIF_EXTRA_TYPE_MCAST_{ADD,DEL}:
++ * Backend advertises availability via 'feature-multicast-control'
++ * xenbus node containing value '1'.
++ * Frontend requests this feature by advertising
++ * 'request-multicast-control' xenbus node containing value '1'.
++ * If multicast control is requested then multicast flooding is
++ * disabled and the frontend must explicitly register its interest
++ * in multicast groups using dummy transmit requests containing
++ * MCAST_{ADD,DEL} extra-info fragments.
++ */
++ struct {
++ uint8_t addr[6]; /* Address to add/remove. */
++ } mcast;
++
++ uint16_t pad[3];
++ } u;
+};
-+typedef struct xen_hvm_set_pci_intx_level xen_hvm_set_pci_intx_level_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_pci_intx_level_t);
++typedef struct netif_extra_info netif_extra_info_t;
+
-+/* Set the logical level of one of a domain's ISA IRQ wires. */
-+#define HVMOP_set_isa_irq_level 3
-+struct xen_hvm_set_isa_irq_level {
-+ /* Domain to be updated. */
-+ domid_t domid;
-+ /* ISA device identification, by ISA IRQ (0-15). */
-+ uint8_t isa_irq;
-+ /* Assertion level (0 = unasserted, 1 = asserted). */
-+ uint8_t level;
++struct netif_tx_response {
++ uint16_t id;
++ int16_t status; /* NETIF_RSP_* */
+};
-+typedef struct xen_hvm_set_isa_irq_level xen_hvm_set_isa_irq_level_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_isa_irq_level_t);
++typedef struct netif_tx_response netif_tx_response_t;
+
-+#define HVMOP_set_pci_link_route 4
-+struct xen_hvm_set_pci_link_route {
-+ /* Domain to be updated. */
-+ domid_t domid;
-+ /* PCI link identifier (0-3). */
-+ uint8_t link;
-+ /* ISA IRQ (1-15), or 0 (disable link). */
-+ uint8_t isa_irq;
++struct netif_rx_request {
++ uint16_t id; /* Echoed in response message. */
++ grant_ref_t gref; /* Reference to incoming granted frame */
+};
-+typedef struct xen_hvm_set_pci_link_route xen_hvm_set_pci_link_route_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_pci_link_route_t);
++typedef struct netif_rx_request netif_rx_request_t;
+
-+/* Flushes all VCPU TLBs: @arg must be NULL. */
-+#define HVMOP_flush_tlbs 5
++/* Packet data has been validated against protocol checksum. */
++#define _NETRXF_data_validated (0)
++#define NETRXF_data_validated (1U<<_NETRXF_data_validated)
++
++/* Protocol checksum field is blank in the packet (hardware offload)? */
++#define _NETRXF_csum_blank (1)
++#define NETRXF_csum_blank (1U<<_NETRXF_csum_blank)
++
++/* Packet continues in the next request descriptor. */
++#define _NETRXF_more_data (2)
++#define NETRXF_more_data (1U<<_NETRXF_more_data)
++
++/* Packet to be followed by extra descriptor(s). */
++#define _NETRXF_extra_info (3)
++#define NETRXF_extra_info (1U<<_NETRXF_extra_info)
++
++struct netif_rx_response {
++ uint16_t id;
++ uint16_t offset; /* Offset in page of start of received packet */
++ uint16_t flags; /* NETRXF_* */
++ int16_t status; /* -ve: BLKIF_RSP_* ; +ve: Rx'ed pkt size. */
++};
++typedef struct netif_rx_response netif_rx_response_t;
+
-+#endif /* __XEN_PUBLIC_HVM_HVM_OP_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/hvm/ioreq.h tmp-linux-2.6-xen.patch/include/xen/interface/hvm/ioreq.h
---- pristine-linux-2.6.18.2/include/xen/interface/hvm/ioreq.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/hvm/ioreq.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,122 @@
+/*
-+ * ioreq.h: I/O request definitions for device models
-+ * Copyright (c) 2004, Intel Corporation.
-+ *
++ * Generate netif ring structures and types.
++ */
++
++DEFINE_RING_TYPES(netif_tx, struct netif_tx_request, struct netif_tx_response);
++DEFINE_RING_TYPES(netif_rx, struct netif_rx_request, struct netif_rx_response);
++
++#define NETIF_RSP_DROPPED -2
++#define NETIF_RSP_ERROR -1
++#define NETIF_RSP_OKAY 0
++/* No response: used for auxiliary requests (e.g., netif_tx_extra). */
++#define NETIF_RSP_NULL 1
++
++#endif
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/pciif.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/pciif.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,101 @@
++/*
++ * PCI Backend/Frontend Common Data Structures & Macros
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -100424,97 +139613,77 @@
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
+ */
++#ifndef __XEN_PCI_COMMON_H__
++#define __XEN_PCI_COMMON_H__
+
-+#ifndef _IOREQ_H_
-+#define _IOREQ_H_
++/* Be sure to bump this number if you change this file */
++#define XEN_PCI_MAGIC "7"
+
-+#define IOREQ_READ 1
-+#define IOREQ_WRITE 0
++/* xen_pci_sharedinfo flags */
++#define _XEN_PCIF_active (0)
++#define XEN_PCIF_active (1<<_XEN_PCI_active)
+
-+#define STATE_IOREQ_NONE 0
-+#define STATE_IOREQ_READY 1
-+#define STATE_IOREQ_INPROCESS 2
-+#define STATE_IORESP_READY 3
++/* xen_pci_op commands */
++#define XEN_PCI_OP_conf_read (0)
++#define XEN_PCI_OP_conf_write (1)
++#define XEN_PCI_OP_enable_msi (2)
++#define XEN_PCI_OP_disable_msi (3)
++#define XEN_PCI_OP_enable_msix (4)
++#define XEN_PCI_OP_disable_msix (5)
+
-+#define IOREQ_TYPE_PIO 0 /* pio */
-+#define IOREQ_TYPE_COPY 1 /* mmio ops */
-+#define IOREQ_TYPE_AND 2
-+#define IOREQ_TYPE_OR 3
-+#define IOREQ_TYPE_XOR 4
-+#define IOREQ_TYPE_XCHG 5
-+#define IOREQ_TYPE_ADD 6
-+#define IOREQ_TYPE_TIMEOFFSET 7
-+#define IOREQ_TYPE_INVALIDATE 8 /* mapcache */
-+#define IOREQ_TYPE_SUB 9
++/* xen_pci_op error numbers */
++#define XEN_PCI_ERR_success (0)
++#define XEN_PCI_ERR_dev_not_found (-1)
++#define XEN_PCI_ERR_invalid_offset (-2)
++#define XEN_PCI_ERR_access_denied (-3)
++#define XEN_PCI_ERR_not_implemented (-4)
++/* XEN_PCI_ERR_op_failed - backend failed to complete the operation */
++#define XEN_PCI_ERR_op_failed (-5)
+
+/*
-+ * VMExit dispatcher should cooperate with instruction decoder to
-+ * prepare this structure and notify service OS and DM by sending
-+ * virq
++ * it should be PAGE_SIZE-sizeof(struct xen_pci_op))/sizeof(struct msix_entry))
++ * Should not exceed 128
+ */
-+struct ioreq {
-+ uint64_t addr; /* physical address */
-+ uint64_t size; /* size in bytes */
-+ uint64_t count; /* for rep prefixes */
-+ uint64_t data; /* data (or paddr of data) */
-+ uint8_t state:4;
-+ uint8_t data_is_ptr:1; /* if 1, data above is the guest paddr
-+ * of the real data to use. */
-+ uint8_t dir:1; /* 1=read, 0=write */
-+ uint8_t df:1;
-+ uint8_t type; /* I/O type */
-+ uint8_t _pad0[6];
-+ uint64_t io_count; /* How many IO done on a vcpu */
-+};
-+typedef struct ioreq ioreq_t;
++#define SH_INFO_MAX_VEC 128
+
-+struct vcpu_iodata {
-+ struct ioreq vp_ioreq;
-+ /* Event channel port, used for notifications to/from the device model. */
-+ uint32_t vp_eport;
-+ uint32_t _pad0;
++struct xen_msix_entry {
++ uint16_t vector;
++ uint16_t entry;
+};
-+typedef struct vcpu_iodata vcpu_iodata_t;
++struct xen_pci_op {
++ /* IN: what action to perform: XEN_PCI_OP_* */
++ uint32_t cmd;
+
-+struct shared_iopage {
-+ struct vcpu_iodata vcpu_iodata[1];
-+};
-+typedef struct shared_iopage shared_iopage_t;
++ /* OUT: will contain an error number (if any) from errno.h */
++ int32_t err;
+
-+#define IOREQ_BUFFER_SLOT_NUM 80
-+struct buffered_iopage {
-+ unsigned int read_pointer;
-+ unsigned int write_pointer;
-+ ioreq_t ioreq[IOREQ_BUFFER_SLOT_NUM];
-+}; /* NB. Size of this structure must be no greater than one page. */
-+typedef struct buffered_iopage buffered_iopage_t;
++ /* IN: which device to touch */
++ uint32_t domain; /* PCI Domain/Segment */
++ uint32_t bus;
++ uint32_t devfn;
+
-+#if defined(__ia64__)
-+struct pio_buffer {
-+ uint32_t page_offset;
-+ uint32_t pointer;
-+ uint32_t data_end;
-+ uint32_t buf_size;
-+ void *opaque;
-+};
++ /* IN: which configuration registers to touch */
++ int32_t offset;
++ int32_t size;
+
-+#define PIO_BUFFER_IDE_PRIMARY 0 /* I/O port = 0x1F0 */
-+#define PIO_BUFFER_IDE_SECONDARY 1 /* I/O port = 0x170 */
-+#define PIO_BUFFER_ENTRY_NUM 2
-+struct buffered_piopage {
-+ struct pio_buffer pio[PIO_BUFFER_ENTRY_NUM];
-+ uint8_t buffer[1];
++ /* IN/OUT: Contains the result after a READ or the value to WRITE */
++ uint32_t value;
++ /* IN: Contains extra infor for this operation */
++ uint32_t info;
++ /*IN: param for msi-x */
++ struct xen_msix_entry msix_entries[SH_INFO_MAX_VEC];
+};
-+#endif /* defined(__ia64__) */
+
-+#if defined(__i386__) || defined(__x86_64__)
-+#define ACPI_PM1A_EVT_BLK_ADDRESS 0x0000000000001f40
-+#define ACPI_PM1A_CNT_BLK_ADDRESS (ACPI_PM1A_EVT_BLK_ADDRESS + 0x04)
-+#define ACPI_PM_TMR_BLK_ADDRESS (ACPI_PM1A_EVT_BLK_ADDRESS + 0x08)
-+#endif /* defined(__i386__) || defined(__x86_64__) */
++struct xen_pci_sharedinfo {
++ /* flags - XEN_PCIF_* */
++ uint32_t flags;
++ struct xen_pci_op op;
++};
+
-+#endif /* _IOREQ_H_ */
++#endif /* __XEN_PCI_COMMON_H__ */
+
+/*
+ * Local variables:
@@ -100525,11 +139694,13 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/hvm/params.h tmp-linux-2.6-xen.patch/include/xen/interface/hvm/params.h
---- pristine-linux-2.6.18.2/include/xen/interface/hvm/params.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/hvm/params.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,60 @@
-+/*
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/protocols.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/protocols.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,40 @@
++/******************************************************************************
++ * protocols.h
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -100549,58 +139720,32 @@
+ * DEALINGS IN THE SOFTWARE.
+ */
+
-+#ifndef __XEN_PUBLIC_HVM_PARAMS_H__
-+#define __XEN_PUBLIC_HVM_PARAMS_H__
-+
-+#include "hvm_op.h"
-+
-+/*
-+ * Parameter space for HVMOP_{set,get}_param.
-+ */
-+
-+/*
-+ * How should CPU0 event-channel notifications be delivered?
-+ * val[63:56] == 0: val[55:0] is a delivery GSI (Global System Interrupt).
-+ * val[63:56] == 1: val[55:0] is a delivery PCI INTx line, as follows:
-+ * Domain = val[47:32], Bus = val[31:16],
-+ * DevFn = val[15: 8], IntX = val[ 1: 0]
-+ * If val == 0 then CPU0 event-channel notifications are not delivered.
-+ */
-+#define HVM_PARAM_CALLBACK_IRQ 0
-+
-+/*
-+ * These are not used by Xen. They are here for convenience of HVM-guest
-+ * xenbus implementations.
-+ */
-+#define HVM_PARAM_STORE_PFN 1
-+#define HVM_PARAM_STORE_EVTCHN 2
-+
-+#define HVM_PARAM_PAE_ENABLED 4
-+
-+#define HVM_PARAM_IOREQ_PFN 5
++#ifndef __XEN_PROTOCOLS_H__
++#define __XEN_PROTOCOLS_H__
+
-+#define HVM_PARAM_BUFIOREQ_PFN 6
++#define XEN_IO_PROTO_ABI_X86_32 "x86_32-abi"
++#define XEN_IO_PROTO_ABI_X86_64 "x86_64-abi"
++#define XEN_IO_PROTO_ABI_IA64 "ia64-abi"
+
-+#ifdef __ia64__
-+#define HVM_PARAM_NVRAM_FD 7
-+#define HVM_NR_PARAMS 8
++#if defined(__i386__)
++# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_X86_32
++#elif defined(__x86_64__)
++# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_X86_64
++#elif defined(__ia64__)
++# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_IA64
+#else
-+#define HVM_NR_PARAMS 7
++# error arch fixup needed here
+#endif
+
-+#endif /* __XEN_PUBLIC_HVM_PARAMS_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/hvm/save.h tmp-linux-2.6-xen.patch/include/xen/interface/hvm/save.h
---- pristine-linux-2.6.18.2/include/xen/interface/hvm/save.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/hvm/save.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,462 @@
-+/*
-+ * hvm/save.h
-+ *
-+ * Structure definitions for HVM state that is held by Xen and must
-+ * be saved along with the domain's memory and device-model state.
-+ *
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/ring.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/ring.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,307 @@
++/******************************************************************************
++ * ring.h
+ *
-+ * Copyright (c) 2007 XenSource Ltd.
++ * Shared producer-consumer ring macros.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -100619,448 +139764,299 @@
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
-+ */
-+
-+#ifndef __XEN_PUBLIC_HVM_SAVE_H__
-+#define __XEN_PUBLIC_HVM_SAVE_H__
-+
-+/*
-+ * Structures in this header *must* have the same layout in 32bit
-+ * and 64bit environments: this means that all fields must be explicitly
-+ * sized types and aligned to their sizes, and the structs must be
-+ * a multiple of eight bytes long.
+ *
-+ * Only the state necessary for saving and restoring (i.e. fields
-+ * that are analogous to actual hardware state) should go in this file.
-+ * Internal mechanisms should be kept in Xen-private headers.
-+ */
-+
-+/*
-+ * Each entry is preceded by a descriptor giving its type and length
-+ */
-+struct hvm_save_descriptor {
-+ uint16_t typecode; /* Used to demux the various types below */
-+ uint16_t instance; /* Further demux within a type */
-+ uint32_t length; /* In bytes, *not* including this descriptor */
-+};
-+
-+
-+/*
-+ * Each entry has a datatype associated with it: for example, the CPU state
-+ * is saved as a HVM_SAVE_TYPE(CPU), which has HVM_SAVE_LENGTH(CPU),
-+ * and is identified by a descriptor with typecode HVM_SAVE_CODE(CPU).
-+ * DECLARE_HVM_SAVE_TYPE binds these things together with some type-system
-+ * ugliness.
-+ */
-+
-+#define DECLARE_HVM_SAVE_TYPE(_x, _code, _type) \
-+ struct __HVM_SAVE_TYPE_##_x { _type t; char c[_code]; }
-+
-+#define HVM_SAVE_TYPE(_x) typeof (((struct __HVM_SAVE_TYPE_##_x *)(0))->t)
-+#define HVM_SAVE_LENGTH(_x) (sizeof (HVM_SAVE_TYPE(_x)))
-+#define HVM_SAVE_CODE(_x) (sizeof (((struct __HVM_SAVE_TYPE_##_x *)(0))->c))
-+
-+
-+/*
-+ * Save/restore header: general info about the save file.
-+ */
-+
-+#define HVM_FILE_MAGIC 0x54381286
-+#define HVM_FILE_VERSION 0x00000001
-+
-+struct hvm_save_header {
-+ uint32_t magic; /* Must be HVM_FILE_MAGIC */
-+ uint32_t version; /* File format version */
-+ uint64_t changeset; /* Version of Xen that saved this file */
-+ uint32_t cpuid; /* CPUID[0x01][%eax] on the saving machine */
-+ uint32_t pad0;
-+};
-+
-+DECLARE_HVM_SAVE_TYPE(HEADER, 1, struct hvm_save_header);
-+
-+
-+/*
-+ * Processor
-+ */
-+
-+struct hvm_hw_cpu {
-+ uint8_t fpu_regs[512];
-+
-+ uint64_t rax;
-+ uint64_t rbx;
-+ uint64_t rcx;
-+ uint64_t rdx;
-+ uint64_t rbp;
-+ uint64_t rsi;
-+ uint64_t rdi;
-+ uint64_t rsp;
-+ uint64_t r8;
-+ uint64_t r9;
-+ uint64_t r10;
-+ uint64_t r11;
-+ uint64_t r12;
-+ uint64_t r13;
-+ uint64_t r14;
-+ uint64_t r15;
-+
-+ uint64_t rip;
-+ uint64_t rflags;
-+
-+ uint64_t cr0;
-+ uint64_t cr2;
-+ uint64_t cr3;
-+ uint64_t cr4;
-+
-+ uint64_t dr0;
-+ uint64_t dr1;
-+ uint64_t dr2;
-+ uint64_t dr3;
-+ uint64_t dr6;
-+ uint64_t dr7;
-+
-+ uint32_t cs_sel;
-+ uint32_t ds_sel;
-+ uint32_t es_sel;
-+ uint32_t fs_sel;
-+ uint32_t gs_sel;
-+ uint32_t ss_sel;
-+ uint32_t tr_sel;
-+ uint32_t ldtr_sel;
-+
-+ uint32_t cs_limit;
-+ uint32_t ds_limit;
-+ uint32_t es_limit;
-+ uint32_t fs_limit;
-+ uint32_t gs_limit;
-+ uint32_t ss_limit;
-+ uint32_t tr_limit;
-+ uint32_t ldtr_limit;
-+ uint32_t idtr_limit;
-+ uint32_t gdtr_limit;
-+
-+ uint64_t cs_base;
-+ uint64_t ds_base;
-+ uint64_t es_base;
-+ uint64_t fs_base;
-+ uint64_t gs_base;
-+ uint64_t ss_base;
-+ uint64_t tr_base;
-+ uint64_t ldtr_base;
-+ uint64_t idtr_base;
-+ uint64_t gdtr_base;
-+
-+ uint32_t cs_arbytes;
-+ uint32_t ds_arbytes;
-+ uint32_t es_arbytes;
-+ uint32_t fs_arbytes;
-+ uint32_t gs_arbytes;
-+ uint32_t ss_arbytes;
-+ uint32_t tr_arbytes;
-+ uint32_t ldtr_arbytes;
-+
-+ uint32_t sysenter_cs;
-+ uint32_t padding0;
-+
-+ uint64_t sysenter_esp;
-+ uint64_t sysenter_eip;
-+
-+ /* msr for em64t */
-+ uint64_t shadow_gs;
-+
-+ /* msr content saved/restored. */
-+ uint64_t msr_flags;
-+ uint64_t msr_lstar;
-+ uint64_t msr_star;
-+ uint64_t msr_cstar;
-+ uint64_t msr_syscall_mask;
-+ uint64_t msr_efer;
-+
-+ /* guest's idea of what rdtsc() would return */
-+ uint64_t tsc;
-+
-+ /* pending event, if any */
-+ union {
-+ uint32_t pending_event;
-+ struct {
-+ uint8_t pending_vector:8;
-+ uint8_t pending_type:3;
-+ uint8_t pending_error_valid:1;
-+ uint32_t pending_reserved:19;
-+ uint8_t pending_valid:1;
-+ };
-+ };
-+ /* error code for pending event */
-+ uint32_t error_code;
-+};
-+
-+DECLARE_HVM_SAVE_TYPE(CPU, 2, struct hvm_hw_cpu);
-+
-+
-+/*
-+ * PIC
++ * Tim Deegan and Andrew Warfield November 2004.
+ */
+
-+struct hvm_hw_vpic {
-+ /* IR line bitmasks. */
-+ uint8_t irr;
-+ uint8_t imr;
-+ uint8_t isr;
-+
-+ /* Line IRx maps to IRQ irq_base+x */
-+ uint8_t irq_base;
-+
-+ /*
-+ * Where are we in ICW2-4 initialisation (0 means no init in progress)?
-+ * Bits 0-1 (=x): Next write at A=1 sets ICW(x+1).
-+ * Bit 2: ICW1.IC4 (1 == ICW4 included in init sequence)
-+ * Bit 3: ICW1.SNGL (0 == ICW3 included in init sequence)
-+ */
-+ uint8_t init_state:4;
-+
-+ /* IR line with highest priority. */
-+ uint8_t priority_add:4;
-+
-+ /* Reads from A=0 obtain ISR or IRR? */
-+ uint8_t readsel_isr:1;
-+
-+ /* Reads perform a polling read? */
-+ uint8_t poll:1;
-+
-+ /* Automatically clear IRQs from the ISR during INTA? */
-+ uint8_t auto_eoi:1;
-+
-+ /* Automatically rotate IRQ priorities during AEOI? */
-+ uint8_t rotate_on_auto_eoi:1;
-+
-+ /* Exclude slave inputs when considering in-service IRQs? */
-+ uint8_t special_fully_nested_mode:1;
-+
-+ /* Special mask mode excludes masked IRs from AEOI and priority checks. */
-+ uint8_t special_mask_mode:1;
-+
-+ /* Is this a master PIC or slave PIC? (NB. This is not programmable.) */
-+ uint8_t is_master:1;
++#ifndef __XEN_PUBLIC_IO_RING_H__
++#define __XEN_PUBLIC_IO_RING_H__
+
-+ /* Edge/trigger selection. */
-+ uint8_t elcr;
++#include "../xen-compat.h"
+
-+ /* Virtual INT output. */
-+ uint8_t int_output;
-+};
++#if __XEN_INTERFACE_VERSION__ < 0x00030208
++#define xen_mb() mb()
++#define xen_rmb() rmb()
++#define xen_wmb() wmb()
++#endif
+
-+DECLARE_HVM_SAVE_TYPE(PIC, 3, struct hvm_hw_vpic);
++typedef unsigned int RING_IDX;
+
++/* Round a 32-bit unsigned constant down to the nearest power of two. */
++#define __RD2(_x) (((_x) & 0x00000002) ? 0x2 : ((_x) & 0x1))
++#define __RD4(_x) (((_x) & 0x0000000c) ? __RD2((_x)>>2)<<2 : __RD2(_x))
++#define __RD8(_x) (((_x) & 0x000000f0) ? __RD4((_x)>>4)<<4 : __RD4(_x))
++#define __RD16(_x) (((_x) & 0x0000ff00) ? __RD8((_x)>>8)<<8 : __RD8(_x))
++#define __RD32(_x) (((_x) & 0xffff0000) ? __RD16((_x)>>16)<<16 : __RD16(_x))
+
+/*
-+ * IO-APIC
++ * Calculate size of a shared ring, given the total available space for the
++ * ring and indexes (_sz), and the name tag of the request/response structure.
++ * A ring contains as many entries as will fit, rounded down to the nearest
++ * power of two (so we can mask with (size-1) to loop around).
+ */
-+
-+#ifdef __ia64__
-+#define VIOAPIC_IS_IOSAPIC 1
-+#define VIOAPIC_NUM_PINS 24
-+#else
-+#define VIOAPIC_NUM_PINS 48 /* 16 ISA IRQs, 32 non-legacy PCI IRQS. */
-+#endif
-+
-+struct hvm_hw_vioapic {
-+ uint64_t base_address;
-+ uint32_t ioregsel;
-+ uint32_t id;
-+ union vioapic_redir_entry
-+ {
-+ uint64_t bits;
-+ struct {
-+ uint8_t vector;
-+ uint8_t delivery_mode:3;
-+ uint8_t dest_mode:1;
-+ uint8_t delivery_status:1;
-+ uint8_t polarity:1;
-+ uint8_t remote_irr:1;
-+ uint8_t trig_mode:1;
-+ uint8_t mask:1;
-+ uint8_t reserve:7;
-+#if !VIOAPIC_IS_IOSAPIC
-+ uint8_t reserved[4];
-+ uint8_t dest_id;
-+#else
-+ uint8_t reserved[3];
-+ uint16_t dest_id;
-+#endif
-+ } fields;
-+ } redirtbl[VIOAPIC_NUM_PINS];
-+};
-+
-+DECLARE_HVM_SAVE_TYPE(IOAPIC, 4, struct hvm_hw_vioapic);
-+
++#define __RING_SIZE(_s, _sz) \
++ (__RD32(((_sz) - (long)(_s)->ring + (long)(_s)) / sizeof((_s)->ring[0])))
+
+/*
-+ * LAPIC
++ * Macros to make the correct C datatypes for a new kind of ring.
++ *
++ * To make a new ring datatype, you need to have two message structures,
++ * let's say request_t, and response_t already defined.
++ *
++ * In a header where you want the ring datatype declared, you then do:
++ *
++ * DEFINE_RING_TYPES(mytag, request_t, response_t);
++ *
++ * These expand out to give you a set of types, as you can see below.
++ * The most important of these are:
++ *
++ * mytag_sring_t - The shared ring.
++ * mytag_front_ring_t - The 'front' half of the ring.
++ * mytag_back_ring_t - The 'back' half of the ring.
++ *
++ * To initialize a ring in your code you need to know the location and size
++ * of the shared memory area (PAGE_SIZE, for instance). To initialise
++ * the front half:
++ *
++ * mytag_front_ring_t front_ring;
++ * SHARED_RING_INIT((mytag_sring_t *)shared_page);
++ * FRONT_RING_INIT(&front_ring, (mytag_sring_t *)shared_page, PAGE_SIZE);
++ *
++ * Initializing the back follows similarly (note that only the front
++ * initializes the shared ring):
++ *
++ * mytag_back_ring_t back_ring;
++ * BACK_RING_INIT(&back_ring, (mytag_sring_t *)shared_page, PAGE_SIZE);
+ */
+
-+struct hvm_hw_lapic {
-+ uint64_t apic_base_msr;
-+ uint32_t disabled; /* VLAPIC_xx_DISABLED */
-+ uint32_t timer_divisor;
-+};
-+
-+DECLARE_HVM_SAVE_TYPE(LAPIC, 5, struct hvm_hw_lapic);
-+
-+struct hvm_hw_lapic_regs {
-+ /* A 4k page of register state */
-+ uint8_t data[0x400];
-+};
-+
-+DECLARE_HVM_SAVE_TYPE(LAPIC_REGS, 6, struct hvm_hw_lapic_regs);
-+
++#define DEFINE_RING_TYPES(__name, __req_t, __rsp_t) \
++ \
++/* Shared ring entry */ \
++union __name##_sring_entry { \
++ __req_t req; \
++ __rsp_t rsp; \
++}; \
++ \
++/* Shared ring page */ \
++struct __name##_sring { \
++ RING_IDX req_prod, req_event; \
++ RING_IDX rsp_prod, rsp_event; \
++ uint8_t pad[48]; \
++ union __name##_sring_entry ring[1]; /* variable-length */ \
++}; \
++ \
++/* "Front" end's private variables */ \
++struct __name##_front_ring { \
++ RING_IDX req_prod_pvt; \
++ RING_IDX rsp_cons; \
++ unsigned int nr_ents; \
++ struct __name##_sring *sring; \
++}; \
++ \
++/* "Back" end's private variables */ \
++struct __name##_back_ring { \
++ RING_IDX rsp_prod_pvt; \
++ RING_IDX req_cons; \
++ unsigned int nr_ents; \
++ struct __name##_sring *sring; \
++}; \
++ \
++/* Syntactic sugar */ \
++typedef struct __name##_sring __name##_sring_t; \
++typedef struct __name##_front_ring __name##_front_ring_t; \
++typedef struct __name##_back_ring __name##_back_ring_t
+
+/*
-+ * IRQs
++ * Macros for manipulating rings.
++ *
++ * FRONT_RING_whatever works on the "front end" of a ring: here
++ * requests are pushed on to the ring and responses taken off it.
++ *
++ * BACK_RING_whatever works on the "back end" of a ring: here
++ * requests are taken off the ring and responses put on.
++ *
++ * N.B. these macros do NO INTERLOCKS OR FLOW CONTROL.
++ * This is OK in 1-for-1 request-response situations where the
++ * requestor (front end) never has more than RING_SIZE()-1
++ * outstanding requests.
+ */
+
-+struct hvm_hw_pci_irqs {
-+ /*
-+ * Virtual interrupt wires for a single PCI bus.
-+ * Indexed by: device*4 + INTx#.
-+ */
-+ union {
-+ DECLARE_BITMAP(i, 32*4);
-+ uint64_t pad[2];
-+ };
-+};
-+
-+DECLARE_HVM_SAVE_TYPE(PCI_IRQ, 7, struct hvm_hw_pci_irqs);
-+
-+struct hvm_hw_isa_irqs {
-+ /*
-+ * Virtual interrupt wires for ISA devices.
-+ * Indexed by ISA IRQ (assumes no ISA-device IRQ sharing).
-+ */
-+ union {
-+ DECLARE_BITMAP(i, 16);
-+ uint64_t pad[1];
-+ };
-+};
-+
-+DECLARE_HVM_SAVE_TYPE(ISA_IRQ, 8, struct hvm_hw_isa_irqs);
-+
-+struct hvm_hw_pci_link {
-+ /*
-+ * PCI-ISA interrupt router.
-+ * Each PCI <device:INTx#> is 'wire-ORed' into one of four links using
-+ * the traditional 'barber's pole' mapping ((device + INTx#) & 3).
-+ * The router provides a programmable mapping from each link to a GSI.
-+ */
-+ uint8_t route[4];
-+ uint8_t pad0[4];
-+};
-+
-+DECLARE_HVM_SAVE_TYPE(PCI_LINK, 9, struct hvm_hw_pci_link);
++/* Initialising empty rings */
++#define SHARED_RING_INIT(_s) do { \
++ (_s)->req_prod = (_s)->rsp_prod = 0; \
++ (_s)->req_event = (_s)->rsp_event = 1; \
++ (void)memset((_s)->pad, 0, sizeof((_s)->pad)); \
++} while(0)
+
-+/*
-+ * PIT
-+ */
++#define FRONT_RING_INIT(_r, _s, __size) do { \
++ (_r)->req_prod_pvt = 0; \
++ (_r)->rsp_cons = 0; \
++ (_r)->nr_ents = __RING_SIZE(_s, __size); \
++ (_r)->sring = (_s); \
++} while (0)
+
-+struct hvm_hw_pit {
-+ struct hvm_hw_pit_channel {
-+ uint32_t count; /* can be 65536 */
-+ uint16_t latched_count;
-+ uint8_t count_latched;
-+ uint8_t status_latched;
-+ uint8_t status;
-+ uint8_t read_state;
-+ uint8_t write_state;
-+ uint8_t write_latch;
-+ uint8_t rw_mode;
-+ uint8_t mode;
-+ uint8_t bcd; /* not supported */
-+ uint8_t gate; /* timer start */
-+ } channels[3]; /* 3 x 16 bytes */
-+ uint32_t speaker_data_on;
-+ uint32_t pad0;
-+};
++#define BACK_RING_INIT(_r, _s, __size) do { \
++ (_r)->rsp_prod_pvt = 0; \
++ (_r)->req_cons = 0; \
++ (_r)->nr_ents = __RING_SIZE(_s, __size); \
++ (_r)->sring = (_s); \
++} while (0)
+
-+DECLARE_HVM_SAVE_TYPE(PIT, 10, struct hvm_hw_pit);
++/* Initialize to existing shared indexes -- for recovery */
++#define FRONT_RING_ATTACH(_r, _s, __size) do { \
++ (_r)->sring = (_s); \
++ (_r)->req_prod_pvt = (_s)->req_prod; \
++ (_r)->rsp_cons = (_s)->rsp_prod; \
++ (_r)->nr_ents = __RING_SIZE(_s, __size); \
++} while (0)
+
++#define BACK_RING_ATTACH(_r, _s, __size) do { \
++ (_r)->sring = (_s); \
++ (_r)->rsp_prod_pvt = (_s)->rsp_prod; \
++ (_r)->req_cons = (_s)->req_prod; \
++ (_r)->nr_ents = __RING_SIZE(_s, __size); \
++} while (0)
+
-+/*
-+ * RTC
-+ */
++/* How big is this ring? */
++#define RING_SIZE(_r) \
++ ((_r)->nr_ents)
+
-+#define RTC_CMOS_SIZE 14
-+struct hvm_hw_rtc {
-+ /* CMOS bytes */
-+ uint8_t cmos_data[RTC_CMOS_SIZE];
-+ /* Index register for 2-part operations */
-+ uint8_t cmos_index;
-+ uint8_t pad0;
-+};
++/* Number of free requests (for use on front side only). */
++#define RING_FREE_REQUESTS(_r) \
++ (RING_SIZE(_r) - ((_r)->req_prod_pvt - (_r)->rsp_cons))
+
-+DECLARE_HVM_SAVE_TYPE(RTC, 11, struct hvm_hw_rtc);
++/* Test if there is an empty slot available on the front ring.
++ * (This is only meaningful from the front. )
++ */
++#define RING_FULL(_r) \
++ (RING_FREE_REQUESTS(_r) == 0)
+
++/* Test if there are outstanding messages to be processed on a ring. */
++#define RING_HAS_UNCONSUMED_RESPONSES(_r) \
++ ((_r)->sring->rsp_prod - (_r)->rsp_cons)
+
-+/*
-+ * HPET
-+ */
++#ifdef __GNUC__
++#define RING_HAS_UNCONSUMED_REQUESTS(_r) ({ \
++ unsigned int req = (_r)->sring->req_prod - (_r)->req_cons; \
++ unsigned int rsp = RING_SIZE(_r) - \
++ ((_r)->req_cons - (_r)->rsp_prod_pvt); \
++ req < rsp ? req : rsp; \
++})
++#else
++/* Same as above, but without the nice GCC ({ ... }) syntax. */
++#define RING_HAS_UNCONSUMED_REQUESTS(_r) \
++ ((((_r)->sring->req_prod - (_r)->req_cons) < \
++ (RING_SIZE(_r) - ((_r)->req_cons - (_r)->rsp_prod_pvt))) ? \
++ ((_r)->sring->req_prod - (_r)->req_cons) : \
++ (RING_SIZE(_r) - ((_r)->req_cons - (_r)->rsp_prod_pvt)))
++#endif
+
-+#define HPET_TIMER_NUM 3 /* 3 timers supported now */
-+struct hvm_hw_hpet {
-+ /* Memory-mapped, software visible registers */
-+ uint64_t capability; /* capabilities */
-+ uint64_t res0; /* reserved */
-+ uint64_t config; /* configuration */
-+ uint64_t res1; /* reserved */
-+ uint64_t isr; /* interrupt status reg */
-+ uint64_t res2[25]; /* reserved */
-+ uint64_t mc64; /* main counter */
-+ uint64_t res3; /* reserved */
-+ struct { /* timers */
-+ uint64_t config; /* configuration/cap */
-+ uint64_t cmp; /* comparator */
-+ uint64_t fsb; /* FSB route, not supported now */
-+ uint64_t res4; /* reserved */
-+ } timers[HPET_TIMER_NUM];
-+ uint64_t res5[4*(24-HPET_TIMER_NUM)]; /* reserved, up to 0x3ff */
++/* Direct access to individual ring elements, by index. */
++#define RING_GET_REQUEST(_r, _idx) \
++ (&((_r)->sring->ring[((_idx) & (RING_SIZE(_r) - 1))].req))
+
-+ /* Hidden register state */
-+ uint64_t period[HPET_TIMER_NUM]; /* Last value written to comparator */
-+};
++#define RING_GET_RESPONSE(_r, _idx) \
++ (&((_r)->sring->ring[((_idx) & (RING_SIZE(_r) - 1))].rsp))
+
-+DECLARE_HVM_SAVE_TYPE(HPET, 12, struct hvm_hw_hpet);
++/* Loop termination condition: Would the specified index overflow the ring? */
++#define RING_REQUEST_CONS_OVERFLOW(_r, _cons) \
++ (((_cons) - (_r)->rsp_prod_pvt) >= RING_SIZE(_r))
++
++#define RING_PUSH_REQUESTS(_r) do { \
++ xen_wmb(); /* back sees requests /before/ updated producer index */ \
++ (_r)->sring->req_prod = (_r)->req_prod_pvt; \
++} while (0)
+
++#define RING_PUSH_RESPONSES(_r) do { \
++ xen_wmb(); /* front sees resps /before/ updated producer index */ \
++ (_r)->sring->rsp_prod = (_r)->rsp_prod_pvt; \
++} while (0)
+
+/*
-+ * PM timer
++ * Notification hold-off (req_event and rsp_event):
++ *
++ * When queueing requests or responses on a shared ring, it may not always be
++ * necessary to notify the remote end. For example, if requests are in flight
++ * in a backend, the front may be able to queue further requests without
++ * notifying the back (if the back checks for new requests when it queues
++ * responses).
++ *
++ * When enqueuing requests or responses:
++ *
++ * Use RING_PUSH_{REQUESTS,RESPONSES}_AND_CHECK_NOTIFY(). The second argument
++ * is a boolean return value. True indicates that the receiver requires an
++ * asynchronous notification.
++ *
++ * After dequeuing requests or responses (before sleeping the connection):
++ *
++ * Use RING_FINAL_CHECK_FOR_REQUESTS() or RING_FINAL_CHECK_FOR_RESPONSES().
++ * The second argument is a boolean return value. True indicates that there
++ * are pending messages on the ring (i.e., the connection should not be put
++ * to sleep).
++ *
++ * These macros will set the req_event/rsp_event field to trigger a
++ * notification on the very next message that is enqueued. If you want to
++ * create batches of work (i.e., only receive a notification after several
++ * messages have been enqueued) then you will need to create a customised
++ * version of the FINAL_CHECK macro in your own code, which sets the event
++ * field appropriately.
+ */
+
-+struct hvm_hw_pmtimer {
-+ uint32_t tmr_val; /* PM_TMR_BLK.TMR_VAL: 32bit free-running counter */
-+ uint16_t pm1a_sts; /* PM1a_EVT_BLK.PM1a_STS: status register */
-+ uint16_t pm1a_en; /* PM1a_EVT_BLK.PM1a_EN: enable register */
-+};
-+
-+DECLARE_HVM_SAVE_TYPE(PMTIMER, 13, struct hvm_hw_pmtimer);
++#define RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(_r, _notify) do { \
++ RING_IDX __old = (_r)->sring->req_prod; \
++ RING_IDX __new = (_r)->req_prod_pvt; \
++ xen_wmb(); /* back sees requests /before/ updated producer index */ \
++ (_r)->sring->req_prod = __new; \
++ xen_mb(); /* back sees new requests /before/ we check req_event */ \
++ (_notify) = ((RING_IDX)(__new - (_r)->sring->req_event) < \
++ (RING_IDX)(__new - __old)); \
++} while (0)
+
-+/*
-+ * Largest type-code in use
-+ */
-+#define HVM_SAVE_CODE_MAX 13
++#define RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(_r, _notify) do { \
++ RING_IDX __old = (_r)->sring->rsp_prod; \
++ RING_IDX __new = (_r)->rsp_prod_pvt; \
++ xen_wmb(); /* front sees resps /before/ updated producer index */ \
++ (_r)->sring->rsp_prod = __new; \
++ xen_mb(); /* front sees new resps /before/ we check rsp_event */ \
++ (_notify) = ((RING_IDX)(__new - (_r)->sring->rsp_event) < \
++ (RING_IDX)(__new - __old)); \
++} while (0)
+
++#define RING_FINAL_CHECK_FOR_REQUESTS(_r, _work_to_do) do { \
++ (_work_to_do) = RING_HAS_UNCONSUMED_REQUESTS(_r); \
++ if (_work_to_do) break; \
++ (_r)->sring->req_event = (_r)->req_cons + 1; \
++ xen_mb(); \
++ (_work_to_do) = RING_HAS_UNCONSUMED_REQUESTS(_r); \
++} while (0)
+
-+/*
-+ * The series of save records is teminated by a zero-type, zero-length
-+ * descriptor.
-+ */
++#define RING_FINAL_CHECK_FOR_RESPONSES(_r, _work_to_do) do { \
++ (_work_to_do) = RING_HAS_UNCONSUMED_RESPONSES(_r); \
++ if (_work_to_do) break; \
++ (_r)->sring->rsp_event = (_r)->rsp_cons + 1; \
++ xen_mb(); \
++ (_work_to_do) = RING_HAS_UNCONSUMED_RESPONSES(_r); \
++} while (0)
+
-+struct hvm_save_end {};
-+DECLARE_HVM_SAVE_TYPE(END, 0, struct hvm_save_end);
++#endif /* __XEN_PUBLIC_IO_RING_H__ */
+
-+#endif /* __XEN_PUBLIC_HVM_SAVE_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/hvm/vmx_assist.h tmp-linux-2.6-xen.patch/include/xen/interface/hvm/vmx_assist.h
---- pristine-linux-2.6.18.2/include/xen/interface/hvm/vmx_assist.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/hvm/vmx_assist.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,122 @@
+/*
-+ * vmx_assist.h: Context definitions for the VMXASSIST world switch.
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/tpmif.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/tpmif.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,77 @@
++/******************************************************************************
++ * tpmif.h
++ *
++ * TPM I/O interface for Xen guest OSes.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -101080,97 +140076,50 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Leendert van Doorn, leendert at watson.ibm.com
-+ * Copyright (c) 2005, International Business Machines Corporation.
++ * Copyright (c) 2005, IBM Corporation
++ *
++ * Author: Stefan Berger, stefanb at us.ibm.com
++ * Grant table support: Mahadevan Gomathisankaran
++ *
++ * This code has been derived from tools/libxc/xen/io/netif.h
++ *
++ * Copyright (c) 2003-2004, Keir Fraser
+ */
+
-+#ifndef _VMX_ASSIST_H_
-+#define _VMX_ASSIST_H_
-+
-+#define VMXASSIST_BASE 0xD0000
-+#define VMXASSIST_MAGIC 0x17101966
-+#define VMXASSIST_MAGIC_OFFSET (VMXASSIST_BASE+8)
-+
-+#define VMXASSIST_NEW_CONTEXT (VMXASSIST_BASE + 12)
-+#define VMXASSIST_OLD_CONTEXT (VMXASSIST_NEW_CONTEXT + 4)
-+
-+#ifndef __ASSEMBLY__
++#ifndef __XEN_PUBLIC_IO_TPMIF_H__
++#define __XEN_PUBLIC_IO_TPMIF_H__
+
-+#define NR_EXCEPTION_HANDLER 32
-+#define NR_INTERRUPT_HANDLERS 16
-+#define NR_TRAPS (NR_EXCEPTION_HANDLER+NR_INTERRUPT_HANDLERS)
++#include "../grant_table.h"
+
-+union vmcs_arbytes {
-+ struct arbyte_fields {
-+ unsigned int seg_type : 4,
-+ s : 1,
-+ dpl : 2,
-+ p : 1,
-+ reserved0 : 4,
-+ avl : 1,
-+ reserved1 : 1,
-+ default_ops_size: 1,
-+ g : 1,
-+ null_bit : 1,
-+ reserved2 : 15;
-+ } fields;
-+ unsigned int bytes;
++struct tpmif_tx_request {
++ unsigned long addr; /* Machine address of packet. */
++ grant_ref_t ref; /* grant table access reference */
++ uint16_t unused;
++ uint16_t size; /* Packet size in bytes. */
+};
++typedef struct tpmif_tx_request tpmif_tx_request_t;
+
+/*
-+ * World switch state
++ * The TPMIF_TX_RING_SIZE defines the number of pages the
++ * front-end and backend can exchange (= size of array).
+ */
-+struct vmx_assist_context {
-+ uint32_t eip; /* execution pointer */
-+ uint32_t esp; /* stack pointer */
-+ uint32_t eflags; /* flags register */
-+ uint32_t cr0;
-+ uint32_t cr3; /* page table directory */
-+ uint32_t cr4;
-+ uint32_t idtr_limit; /* idt */
-+ uint32_t idtr_base;
-+ uint32_t gdtr_limit; /* gdt */
-+ uint32_t gdtr_base;
-+ uint32_t cs_sel; /* cs selector */
-+ uint32_t cs_limit;
-+ uint32_t cs_base;
-+ union vmcs_arbytes cs_arbytes;
-+ uint32_t ds_sel; /* ds selector */
-+ uint32_t ds_limit;
-+ uint32_t ds_base;
-+ union vmcs_arbytes ds_arbytes;
-+ uint32_t es_sel; /* es selector */
-+ uint32_t es_limit;
-+ uint32_t es_base;
-+ union vmcs_arbytes es_arbytes;
-+ uint32_t ss_sel; /* ss selector */
-+ uint32_t ss_limit;
-+ uint32_t ss_base;
-+ union vmcs_arbytes ss_arbytes;
-+ uint32_t fs_sel; /* fs selector */
-+ uint32_t fs_limit;
-+ uint32_t fs_base;
-+ union vmcs_arbytes fs_arbytes;
-+ uint32_t gs_sel; /* gs selector */
-+ uint32_t gs_limit;
-+ uint32_t gs_base;
-+ union vmcs_arbytes gs_arbytes;
-+ uint32_t tr_sel; /* task selector */
-+ uint32_t tr_limit;
-+ uint32_t tr_base;
-+ union vmcs_arbytes tr_arbytes;
-+ uint32_t ldtr_sel; /* ldtr selector */
-+ uint32_t ldtr_limit;
-+ uint32_t ldtr_base;
-+ union vmcs_arbytes ldtr_arbytes;
++typedef uint32_t TPMIF_RING_IDX;
+
-+ unsigned char rm_irqbase[2];
++#define TPMIF_TX_RING_SIZE 1
++
++/* This structure must fit in a memory page. */
++
++struct tpmif_ring {
++ struct tpmif_tx_request req;
+};
-+typedef struct vmx_assist_context vmx_assist_context_t;
++typedef struct tpmif_ring tpmif_ring_t;
+
-+#endif /* __ASSEMBLY__ */
++struct tpmif_tx_interface {
++ struct tpmif_ring ring[TPMIF_TX_RING_SIZE];
++};
++typedef struct tpmif_tx_interface tpmif_tx_interface_t;
+
-+#endif /* _VMX_ASSIST_H_ */
++#endif
+
+/*
+ * Local variables:
@@ -101181,14 +140130,14 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/io/blkif.h tmp-linux-2.6-xen.patch/include/xen/interface/io/blkif.h
---- pristine-linux-2.6.18.2/include/xen/interface/io/blkif.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/io/blkif.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,128 @@
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/vscsiif.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/vscsiif.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,105 @@
+/******************************************************************************
-+ * blkif.h
++ * vscsiif.h
+ *
-+ * Unified block-device I/O interface for Xen guest OSes.
++ * Based on the blkif.h code.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -101208,102 +140157,79 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (c) 2003-2004, Keir Fraser
++ * Copyright(c) FUJITSU Limited 2008.
+ */
+
-+#ifndef __XEN_PUBLIC_IO_BLKIF_H__
-+#define __XEN_PUBLIC_IO_BLKIF_H__
++#ifndef __XEN__PUBLIC_IO_SCSI_H__
++#define __XEN__PUBLIC_IO_SCSI_H__
+
+#include "ring.h"
+#include "../grant_table.h"
+
-+/*
-+ * Front->back notifications: When enqueuing a new request, sending a
-+ * notification can be made conditional on req_event (i.e., the generic
-+ * hold-off mechanism provided by the ring macros). Backends must set
-+ * req_event appropriately (e.g., using RING_FINAL_CHECK_FOR_REQUESTS()).
-+ *
-+ * Back->front notifications: When enqueuing a new response, sending a
-+ * notification can be made conditional on rsp_event (i.e., the generic
-+ * hold-off mechanism provided by the ring macros). Frontends must set
-+ * rsp_event appropriately (e.g., using RING_FINAL_CHECK_FOR_RESPONSES()).
-+ */
++/* command between backend and frontend */
++#define VSCSIIF_ACT_SCSI_CDB 1 /* SCSI CDB command */
++#define VSCSIIF_ACT_SCSI_ABORT 2 /* SCSI Device(Lun) Abort*/
++#define VSCSIIF_ACT_SCSI_RESET 3 /* SCSI Device(Lun) Reset*/
+
-+#ifndef blkif_vdev_t
-+#define blkif_vdev_t uint16_t
-+#endif
-+#define blkif_sector_t uint64_t
+
-+/*
-+ * REQUEST CODES.
-+ */
-+#define BLKIF_OP_READ 0
-+#define BLKIF_OP_WRITE 1
-+/*
-+ * Recognised only if "feature-barrier" is present in backend xenbus info.
-+ * The "feature_barrier" node contains a boolean indicating whether barrier
-+ * requests are likely to succeed or fail. Either way, a barrier request
-+ * may fail at any time with BLKIF_RSP_EOPNOTSUPP if it is unsupported by
-+ * the underlying block-device hardware. The boolean simply indicates whether
-+ * or not it is worthwhile for the frontend to attempt barrier requests.
-+ * If a backend does not recognise BLKIF_OP_WRITE_BARRIER, it should *not*
-+ * create the "feature-barrier" node!
-+ */
-+#define BLKIF_OP_WRITE_BARRIER 2
++#define VSCSIIF_BACK_MAX_PENDING_REQS 128
+
+/*
+ * Maximum scatter/gather segments per request.
-+ * This is carefully chosen so that sizeof(blkif_ring_t) <= PAGE_SIZE.
-+ * NB. This could be 12 if the ring indexes weren't stored in the same page.
++ *
++ * Considering balance between allocating al least 16 "vscsiif_request"
++ * structures on one page (4096bytes) and number of scatter gather
++ * needed, we decided to use 26 as a magic number.
+ */
-+#define BLKIF_MAX_SEGMENTS_PER_REQUEST 11
++#define VSCSIIF_SG_TABLESIZE 26
+
-+struct blkif_request_segment {
-+ grant_ref_t gref; /* reference to I/O buffer frame */
-+ /* @first_sect: first sector in frame to transfer (inclusive). */
-+ /* @last_sect: last sector in frame to transfer (inclusive). */
-+ uint8_t first_sect, last_sect;
-+};
++/*
++ * base on linux kernel 2.6.18
++ */
++#define VSCSIIF_MAX_COMMAND_SIZE 16
++#define VSCSIIF_SENSE_BUFFERSIZE 96
+
-+struct blkif_request {
-+ uint8_t operation; /* BLKIF_OP_??? */
-+ uint8_t nr_segments; /* number of segments */
-+ blkif_vdev_t handle; /* only for read/write requests */
-+ uint64_t id; /* private guest value, echoed in resp */
-+ blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
-+ struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
-+};
-+typedef struct blkif_request blkif_request_t;
+
-+struct blkif_response {
-+ uint64_t id; /* copied from request */
-+ uint8_t operation; /* copied from request */
-+ int16_t status; /* BLKIF_RSP_??? */
-+};
-+typedef struct blkif_response blkif_response_t;
++struct vscsiif_request {
++ uint16_t rqid; /* private guest value, echoed in resp */
++ uint8_t act; /* command between backend and frontend */
++ uint8_t cmd_len;
+
-+/*
-+ * STATUS RETURN CODES.
-+ */
-+ /* Operation not supported (only happens on barrier writes). */
-+#define BLKIF_RSP_EOPNOTSUPP -2
-+ /* Operation failed for some unspecified reason (-EIO). */
-+#define BLKIF_RSP_ERROR -1
-+ /* Operation completed successfully. */
-+#define BLKIF_RSP_OKAY 0
++ uint8_t cmnd[VSCSIIF_MAX_COMMAND_SIZE];
++ uint16_t timeout_per_command; /* The command is issued by twice
++ the value in Backend. */
++ uint16_t channel, id, lun;
++ uint16_t padding;
++ uint8_t sc_data_direction; /* for DMA_TO_DEVICE(1)
++ DMA_FROM_DEVICE(2)
++ DMA_NONE(3) requests */
++ uint8_t nr_segments; /* Number of pieces of scatter-gather */
+
-+/*
-+ * Generate blkif ring structures and types.
-+ */
++ struct scsiif_request_segment {
++ grant_ref_t gref;
++ uint16_t offset;
++ uint16_t length;
++ } seg[VSCSIIF_SG_TABLESIZE];
++ uint32_t reserved[3];
++};
++typedef struct vscsiif_request vscsiif_request_t;
+
-+DEFINE_RING_TYPES(blkif, struct blkif_request, struct blkif_response);
++struct vscsiif_response {
++ uint16_t rqid;
++ uint8_t padding;
++ uint8_t sense_len;
++ uint8_t sense_buffer[VSCSIIF_SENSE_BUFFERSIZE];
++ int32_t rslt;
++ uint32_t residual_len; /* request bufflen -
++ return the value from physical device */
++ uint32_t reserved[36];
++};
++typedef struct vscsiif_response vscsiif_response_t;
+
-+#define VDISK_CDROM 0x1
-+#define VDISK_REMOVABLE 0x2
-+#define VDISK_READONLY 0x4
++DEFINE_RING_TYPES(vscsiif, struct vscsiif_request, struct vscsiif_response);
+
-+#endif /* __XEN_PUBLIC_IO_BLKIF_H__ */
+
++#endif /*__XEN__PUBLIC_IO_SCSI_H__*/
+/*
+ * Local variables:
+ * mode: C
@@ -101313,15 +140239,15 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/io/console.h tmp-linux-2.6-xen.patch/include/xen/interface/io/console.h
---- pristine-linux-2.6.18.2/include/xen/interface/io/console.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/io/console.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,51 @@
-+/******************************************************************************
-+ * console.h
-+ *
-+ * Console I/O interface for Xen guest OSes.
-+ *
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/xenbus.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/xenbus.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,80 @@
++/*****************************************************************************
++ * xenbus.h
++ *
++ * Xenbus protocol details.
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -101340,24 +140266,53 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (c) 2005, Keir Fraser
++ * Copyright (C) 2005 XenSource Ltd.
+ */
+
-+#ifndef __XEN_PUBLIC_IO_CONSOLE_H__
-+#define __XEN_PUBLIC_IO_CONSOLE_H__
++#ifndef _XEN_PUBLIC_IO_XENBUS_H
++#define _XEN_PUBLIC_IO_XENBUS_H
+
-+typedef uint32_t XENCONS_RING_IDX;
++/*
++ * The state of either end of the Xenbus, i.e. the current communication
++ * status of initialisation across the bus. States here imply nothing about
++ * the state of the connection between the driver and the kernel's device
++ * layers.
++ */
++enum xenbus_state {
++ XenbusStateUnknown = 0,
+
-+#define MASK_XENCONS_IDX(idx, ring) ((idx) & (sizeof(ring)-1))
++ XenbusStateInitialising = 1,
+
-+struct xencons_interface {
-+ char in[1024];
-+ char out[2048];
-+ XENCONS_RING_IDX in_cons, in_prod;
-+ XENCONS_RING_IDX out_cons, out_prod;
++ /*
++ * InitWait: Finished early initialisation but waiting for information
++ * from the peer or hotplug scripts.
++ */
++ XenbusStateInitWait = 2,
++
++ /*
++ * Initialised: Waiting for a connection from the peer.
++ */
++ XenbusStateInitialised = 3,
++
++ XenbusStateConnected = 4,
++
++ /*
++ * Closing: The device is being closed due to an error or an unplug event.
++ */
++ XenbusStateClosing = 5,
++
++ XenbusStateClosed = 6,
++
++ /*
++ * Reconfiguring: The device is being reconfigured.
++ */
++ XenbusStateReconfiguring = 7,
++
++ XenbusStateReconfigured = 8
+};
++typedef enum xenbus_state XenbusState;
+
-+#endif /* __XEN_PUBLIC_IO_CONSOLE_H__ */
++#endif /* _XEN_PUBLIC_IO_XENBUS_H */
+
+/*
+ * Local variables:
@@ -101368,12 +140323,13 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/io/fbif.h tmp-linux-2.6-xen.patch/include/xen/interface/io/fbif.h
---- pristine-linux-2.6.18.2/include/xen/interface/io/fbif.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/io/fbif.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,138 @@
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/io/xs_wire.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/io/xs_wire.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,130 @@
+/*
-+ * fbif.h -- Xen virtual frame buffer device
++ * Details of the "wire" protocol between Xen Store Daemon and client
++ * library or guest kernel.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -101393,113 +140349,104 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (C) 2005 Anthony Liguori <aliguori at us.ibm.com>
-+ * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster <armbru at redhat.com>
++ * Copyright (C) 2005 Rusty Russell IBM Corporation
+ */
+
-+#ifndef __XEN_PUBLIC_IO_FBIF_H__
-+#define __XEN_PUBLIC_IO_FBIF_H__
-+
-+/* Out events (frontend -> backend) */
++#ifndef _XS_WIRE_H
++#define _XS_WIRE_H
+
-+/*
-+ * Out events may be sent only when requested by backend, and receipt
-+ * of an unknown out event is an error.
-+ */
++enum xsd_sockmsg_type
++{
++ XS_DEBUG,
++ XS_DIRECTORY,
++ XS_READ,
++ XS_GET_PERMS,
++ XS_WATCH,
++ XS_UNWATCH,
++ XS_TRANSACTION_START,
++ XS_TRANSACTION_END,
++ XS_INTRODUCE,
++ XS_RELEASE,
++ XS_GET_DOMAIN_PATH,
++ XS_WRITE,
++ XS_MKDIR,
++ XS_RM,
++ XS_SET_PERMS,
++ XS_WATCH_EVENT,
++ XS_ERROR,
++ XS_IS_DOMAIN_INTRODUCED,
++ XS_RESUME,
++ XS_SET_TARGET
++};
+
-+/* Event type 1 currently not used */
-+/*
-+ * Framebuffer update notification event
-+ * Capable frontend sets feature-update in xenstore.
-+ * Backend requests it by setting request-update in xenstore.
-+ */
-+#define XENFB_TYPE_UPDATE 2
++#define XS_WRITE_NONE "NONE"
++#define XS_WRITE_CREATE "CREATE"
++#define XS_WRITE_CREATE_EXCL "CREATE|EXCL"
+
-+struct xenfb_update
++/* We hand errors as strings, for portability. */
++struct xsd_errors
+{
-+ uint8_t type; /* XENFB_TYPE_UPDATE */
-+ int32_t x; /* source x */
-+ int32_t y; /* source y */
-+ int32_t width; /* rect width */
-+ int32_t height; /* rect height */
++ int errnum;
++ const char *errstring;
++};
++#define XSD_ERROR(x) { x, #x }
++/* LINTED: static unused */
++static struct xsd_errors xsd_errors[]
++#if defined(__GNUC__)
++__attribute__((unused))
++#endif
++ = {
++ XSD_ERROR(EINVAL),
++ XSD_ERROR(EACCES),
++ XSD_ERROR(EEXIST),
++ XSD_ERROR(EISDIR),
++ XSD_ERROR(ENOENT),
++ XSD_ERROR(ENOMEM),
++ XSD_ERROR(ENOSPC),
++ XSD_ERROR(EIO),
++ XSD_ERROR(ENOTEMPTY),
++ XSD_ERROR(ENOSYS),
++ XSD_ERROR(EROFS),
++ XSD_ERROR(EBUSY),
++ XSD_ERROR(EAGAIN),
++ XSD_ERROR(EISCONN)
+};
+
-+#define XENFB_OUT_EVENT_SIZE 40
-+
-+union xenfb_out_event
++struct xsd_sockmsg
+{
-+ uint8_t type;
-+ struct xenfb_update update;
-+ char pad[XENFB_OUT_EVENT_SIZE];
-+};
-+
-+/* In events (backend -> frontend) */
-+
-+/*
-+ * Frontends should ignore unknown in events.
-+ * No in events currently defined.
-+ */
-+
-+#define XENFB_IN_EVENT_SIZE 40
++ uint32_t type; /* XS_??? */
++ uint32_t req_id;/* Request identifier, echoed in daemon's response. */
++ uint32_t tx_id; /* Transaction id (0 if not related to a transaction). */
++ uint32_t len; /* Length of data following this. */
+
-+union xenfb_in_event
-+{
-+ uint8_t type;
-+ char pad[XENFB_IN_EVENT_SIZE];
++ /* Generally followed by nul-terminated string(s). */
+};
+
-+/* shared page */
-+
-+#define XENFB_IN_RING_SIZE 1024
-+#define XENFB_IN_RING_LEN (XENFB_IN_RING_SIZE / XENFB_IN_EVENT_SIZE)
-+#define XENFB_IN_RING_OFFS 1024
-+#define XENFB_IN_RING(page) \
-+ ((union xenfb_in_event *)((char *)(page) + XENFB_IN_RING_OFFS))
-+#define XENFB_IN_RING_REF(page, idx) \
-+ (XENFB_IN_RING((page))[(idx) % XENFB_IN_RING_LEN])
-+
-+#define XENFB_OUT_RING_SIZE 2048
-+#define XENFB_OUT_RING_LEN (XENFB_OUT_RING_SIZE / XENFB_OUT_EVENT_SIZE)
-+#define XENFB_OUT_RING_OFFS (XENFB_IN_RING_OFFS + XENFB_IN_RING_SIZE)
-+#define XENFB_OUT_RING(page) \
-+ ((union xenfb_out_event *)((char *)(page) + XENFB_OUT_RING_OFFS))
-+#define XENFB_OUT_RING_REF(page, idx) \
-+ (XENFB_OUT_RING((page))[(idx) % XENFB_OUT_RING_LEN])
-+
-+struct xenfb_page
++enum xs_watch_type
+{
-+ uint32_t in_cons, in_prod;
-+ uint32_t out_cons, out_prod;
-+
-+ int32_t width; /* the width of the framebuffer (in pixels) */
-+ int32_t height; /* the height of the framebuffer (in pixels) */
-+ uint32_t line_length; /* the length of a row of pixels (in bytes) */
-+ uint32_t mem_length; /* the length of the framebuffer (in bytes) */
-+ uint8_t depth; /* the depth of a pixel (in bits) */
++ XS_WATCH_PATH = 0,
++ XS_WATCH_TOKEN
++};
+
-+ /*
-+ * Framebuffer page directory
-+ *
-+ * Each directory page holds PAGE_SIZE / sizeof(*pd)
-+ * framebuffer pages, and can thus map up to PAGE_SIZE *
-+ * PAGE_SIZE / sizeof(*pd) bytes. With PAGE_SIZE == 4096 and
-+ * sizeof(unsigned long) == 4, that's 4 Megs. Two directory
-+ * pages should be enough for a while.
-+ */
-+ unsigned long pd[2];
++/* Inter-domain shared memory communications. */
++#define XENSTORE_RING_SIZE 1024
++typedef uint32_t XENSTORE_RING_IDX;
++#define MASK_XENSTORE_IDX(idx) ((idx) & (XENSTORE_RING_SIZE-1))
++struct xenstore_domain_interface {
++ char req[XENSTORE_RING_SIZE]; /* Requests to xenstore daemon. */
++ char rsp[XENSTORE_RING_SIZE]; /* Replies and async watch events. */
++ XENSTORE_RING_IDX req_cons, req_prod;
++ XENSTORE_RING_IDX rsp_cons, rsp_prod;
+};
+
-+/*
-+ * Wart: xenkbd needs to know resolution. Put it here until a better
-+ * solution is found, but don't leak it to the backend.
-+ */
-+#ifdef __KERNEL__
-+#define XENFB_WIDTH 800
-+#define XENFB_HEIGHT 600
-+#define XENFB_DEPTH 32
-+#endif
++/* Violating this is very bad. See docs/misc/xenstore.txt. */
++#define XENSTORE_PAYLOAD_MAX 4096
+
-+#endif
++/* Violating these just gets you an error back */
++#define XENSTORE_ABS_PATH_MAX 3072
++#define XENSTORE_REL_PATH_MAX 2048
++
++#endif /* _XS_WIRE_H */
+
+/*
+ * Local variables:
@@ -101510,13 +140457,13 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/io/kbdif.h tmp-linux-2.6-xen.patch/include/xen/interface/io/kbdif.h
---- pristine-linux-2.6.18.2/include/xen/interface/io/kbdif.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/io/kbdif.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,130 @@
-+/*
-+ * kbdif.h -- Xen virtual keyboard/mouse
-+ *
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/kexec.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/kexec.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,189 @@
++/******************************************************************************
++ * kexec.h - Public portion
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -101534,106 +140481,165 @@
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
++ *
++ * Xen port written by:
++ * - Simon 'Horms' Horman <horms at verge.net.au>
++ * - Magnus Damm <magnus at valinux.co.jp>
++ */
++
++#ifndef _XEN_PUBLIC_KEXEC_H
++#define _XEN_PUBLIC_KEXEC_H
++
++
++/* This file describes the Kexec / Kdump hypercall interface for Xen.
+ *
-+ * Copyright (C) 2005 Anthony Liguori <aliguori at us.ibm.com>
-+ * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster <armbru at redhat.com>
++ * Kexec under vanilla Linux allows a user to reboot the physical machine
++ * into a new user-specified kernel. The Xen port extends this idea
++ * to allow rebooting of the machine from dom0. When kexec for dom0
++ * is used to reboot, both the hypervisor and the domains get replaced
++ * with some other kernel. It is possible to kexec between vanilla
++ * Linux and Xen and back again. Xen to Xen works well too.
++ *
++ * The hypercall interface for kexec can be divided into three main
++ * types of hypercall operations:
++ *
++ * 1) Range information:
++ * This is used by the dom0 kernel to ask the hypervisor about various
++ * address information. This information is needed to allow kexec-tools
++ * to fill in the ELF headers for /proc/vmcore properly.
++ *
++ * 2) Load and unload of images:
++ * There are no big surprises here, the kexec binary from kexec-tools
++ * runs in userspace in dom0. The tool loads/unloads data into the
++ * dom0 kernel such as new kernel, initramfs and hypervisor. When
++ * loaded the dom0 kernel performs a load hypercall operation, and
++ * before releasing all page references the dom0 kernel calls unload.
++ *
++ * 3) Kexec operation:
++ * This is used to start a previously loaded kernel.
+ */
+
-+#ifndef __XEN_PUBLIC_IO_KBDIF_H__
-+#define __XEN_PUBLIC_IO_KBDIF_H__
++#include "xen.h"
+
-+/* In events (backend -> frontend) */
++#if defined(__i386__) || defined(__x86_64__)
++#define KEXEC_XEN_NO_PAGES 17
++#endif
+
+/*
-+ * Frontends should ignore unknown in events.
++ * Prototype for this hypercall is:
++ * int kexec_op(int cmd, void *args)
++ * @cmd == KEXEC_CMD_...
++ * KEXEC operation to perform
++ * @args == Operation-specific extra arguments (NULL if none).
+ */
+
-+/* Pointer movement event */
-+#define XENKBD_TYPE_MOTION 1
-+/* Event type 2 currently not used */
-+/* Key event (includes pointer buttons) */
-+#define XENKBD_TYPE_KEY 3
+/*
-+ * Pointer position event
-+ * Capable backend sets feature-abs-pointer in xenstore.
-+ * Frontend requests ot instead of XENKBD_TYPE_MOTION by setting
-+ * request-abs-update in xenstore.
++ * Kexec supports two types of operation:
++ * - kexec into a regular kernel, very similar to a standard reboot
++ * - KEXEC_TYPE_DEFAULT is used to specify this type
++ * - kexec into a special "crash kernel", aka kexec-on-panic
++ * - KEXEC_TYPE_CRASH is used to specify this type
++ * - parts of our system may be broken at kexec-on-panic time
++ * - the code should be kept as simple and self-contained as possible
+ */
-+#define XENKBD_TYPE_POS 4
-+
-+struct xenkbd_motion
-+{
-+ uint8_t type; /* XENKBD_TYPE_MOTION */
-+ int32_t rel_x; /* relative X motion */
-+ int32_t rel_y; /* relative Y motion */
-+};
-+
-+struct xenkbd_key
-+{
-+ uint8_t type; /* XENKBD_TYPE_KEY */
-+ uint8_t pressed; /* 1 if pressed; 0 otherwise */
-+ uint32_t keycode; /* KEY_* from linux/input.h */
-+};
-+
-+struct xenkbd_position
-+{
-+ uint8_t type; /* XENKBD_TYPE_POS */
-+ int32_t abs_x; /* absolute X position (in FB pixels) */
-+ int32_t abs_y; /* absolute Y position (in FB pixels) */
-+};
+
-+#define XENKBD_IN_EVENT_SIZE 40
++#define KEXEC_TYPE_DEFAULT 0
++#define KEXEC_TYPE_CRASH 1
+
-+union xenkbd_in_event
-+{
-+ uint8_t type;
-+ struct xenkbd_motion motion;
-+ struct xenkbd_key key;
-+ struct xenkbd_position pos;
-+ char pad[XENKBD_IN_EVENT_SIZE];
-+};
+
-+/* Out events (frontend -> backend) */
++/* The kexec implementation for Xen allows the user to load two
++ * types of kernels, KEXEC_TYPE_DEFAULT and KEXEC_TYPE_CRASH.
++ * All data needed for a kexec reboot is kept in one xen_kexec_image_t
++ * per "instance". The data mainly consists of machine address lists to pages
++ * together with destination addresses. The data in xen_kexec_image_t
++ * is passed to the "code page" which is one page of code that performs
++ * the final relocations before jumping to the new kernel.
++ */
++
++typedef struct xen_kexec_image {
++#if defined(__i386__) || defined(__x86_64__)
++ unsigned long page_list[KEXEC_XEN_NO_PAGES];
++#endif
++#if defined(__ia64__)
++ unsigned long reboot_code_buffer;
++#endif
++ unsigned long indirection_page;
++ unsigned long start_address;
++} xen_kexec_image_t;
+
+/*
-+ * Out events may be sent only when requested by backend, and receipt
-+ * of an unknown out event is an error.
-+ * No out events currently defined.
++ * Perform kexec having previously loaded a kexec or kdump kernel
++ * as appropriate.
++ * type == KEXEC_TYPE_DEFAULT or KEXEC_TYPE_CRASH [in]
+ */
++#define KEXEC_CMD_kexec 0
++typedef struct xen_kexec_exec {
++ int type;
++} xen_kexec_exec_t;
+
-+#define XENKBD_OUT_EVENT_SIZE 40
-+
-+union xenkbd_out_event
-+{
-+ uint8_t type;
-+ char pad[XENKBD_OUT_EVENT_SIZE];
-+};
-+
-+/* shared page */
++/*
++ * Load/Unload kernel image for kexec or kdump.
++ * type == KEXEC_TYPE_DEFAULT or KEXEC_TYPE_CRASH [in]
++ * image == relocation information for kexec (ignored for unload) [in]
++ */
++#define KEXEC_CMD_kexec_load 1
++#define KEXEC_CMD_kexec_unload 2
++typedef struct xen_kexec_load {
++ int type;
++ xen_kexec_image_t image;
++} xen_kexec_load_t;
+
-+#define XENKBD_IN_RING_SIZE 2048
-+#define XENKBD_IN_RING_LEN (XENKBD_IN_RING_SIZE / XENKBD_IN_EVENT_SIZE)
-+#define XENKBD_IN_RING_OFFS 1024
-+#define XENKBD_IN_RING(page) \
-+ ((union xenkbd_in_event *)((char *)(page) + XENKBD_IN_RING_OFFS))
-+#define XENKBD_IN_RING_REF(page, idx) \
-+ (XENKBD_IN_RING((page))[(idx) % XENKBD_IN_RING_LEN])
++#define KEXEC_RANGE_MA_CRASH 0 /* machine address and size of crash area */
++#define KEXEC_RANGE_MA_XEN 1 /* machine address and size of Xen itself */
++#define KEXEC_RANGE_MA_CPU 2 /* machine address and size of a CPU note */
++#define KEXEC_RANGE_MA_XENHEAP 3 /* machine address and size of xenheap
++ * Note that although this is adjacent
++ * to Xen it exists in a separate EFI
++ * region on ia64, and thus needs to be
++ * inserted into iomem_machine separately */
++#define KEXEC_RANGE_MA_BOOT_PARAM 4 /* machine address and size of
++ * the ia64_boot_param */
++#define KEXEC_RANGE_MA_EFI_MEMMAP 5 /* machine address and size of
++ * of the EFI Memory Map */
++#define KEXEC_RANGE_MA_VMCOREINFO 6 /* machine address and size of vmcoreinfo */
+
-+#define XENKBD_OUT_RING_SIZE 1024
-+#define XENKBD_OUT_RING_LEN (XENKBD_OUT_RING_SIZE / XENKBD_OUT_EVENT_SIZE)
-+#define XENKBD_OUT_RING_OFFS (XENKBD_IN_RING_OFFS + XENKBD_IN_RING_SIZE)
-+#define XENKBD_OUT_RING(page) \
-+ ((union xenkbd_out_event *)((char *)(page) + XENKBD_OUT_RING_OFFS))
-+#define XENKBD_OUT_RING_REF(page, idx) \
-+ (XENKBD_OUT_RING((page))[(idx) % XENKBD_OUT_RING_LEN])
++/*
++ * Find the address and size of certain memory areas
++ * range == KEXEC_RANGE_... [in]
++ * nr == physical CPU number (starting from 0) if KEXEC_RANGE_MA_CPU [in]
++ * size == number of bytes reserved in window [out]
++ * start == address of the first byte in the window [out]
++ */
++#define KEXEC_CMD_kexec_get_range 3
++typedef struct xen_kexec_range {
++ int range;
++ int nr;
++ unsigned long size;
++ unsigned long start;
++} xen_kexec_range_t;
+
-+struct xenkbd_page
-+{
-+ uint32_t in_cons, in_prod;
-+ uint32_t out_cons, out_prod;
-+};
++/* vmcoreinfo stuff */
++#define VMCOREINFO_BYTES (4096)
++#define VMCOREINFO_NOTE_NAME "VMCOREINFO_XEN"
++void arch_crash_save_vmcoreinfo(void);
++void vmcoreinfo_append_str(const char *fmt, ...)
++ __attribute__ ((format (printf, 1, 2)));
++#define VMCOREINFO_PAGESIZE(value) \
++ vmcoreinfo_append_str("PAGESIZE=%ld\n", value)
++#define VMCOREINFO_SYMBOL(name) \
++ vmcoreinfo_append_str("SYMBOL(%s)=%lx\n", #name, (unsigned long)&name)
++#define VMCOREINFO_SYMBOL_ALIAS(alias, name) \
++ vmcoreinfo_append_str("SYMBOL(%s)=%lx\n", #alias, (unsigned long)&name)
++#define VMCOREINFO_STRUCT_SIZE(name) \
++ vmcoreinfo_append_str("SIZE(%s)=%zu\n", #name, sizeof(struct name))
++#define VMCOREINFO_OFFSET(name, field) \
++ vmcoreinfo_append_str("OFFSET(%s.%s)=%lu\n", #name, #field, \
++ (unsigned long)offsetof(struct name, field))
++#define VMCOREINFO_OFFSET_ALIAS(name, field, alias) \
++ vmcoreinfo_append_str("OFFSET(%s.%s)=%lu\n", #name, #alias, \
++ (unsigned long)offsetof(struct name, field))
+
-+#endif
++#endif /* _XEN_PUBLIC_KEXEC_H */
+
+/*
+ * Local variables:
@@ -101644,14 +140650,12 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/io/netif.h tmp-linux-2.6-xen.patch/include/xen/interface/io/netif.h
---- pristine-linux-2.6.18.2/include/xen/interface/io/netif.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/io/netif.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,184 @@
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/libelf.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/libelf.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,265 @@
+/******************************************************************************
-+ * netif.h
-+ *
-+ * Unified network-device I/O interface for Xen guest OSes.
++ * libelf.h
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -101670,175 +140674,260 @@
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Copyright (c) 2003-2004, Keir Fraser
+ */
+
-+#ifndef __XEN_PUBLIC_IO_NETIF_H__
-+#define __XEN_PUBLIC_IO_NETIF_H__
++#ifndef __XC_LIBELF__
++#define __XC_LIBELF__ 1
+
-+#include "ring.h"
-+#include "../grant_table.h"
++#if defined(__i386__) || defined(__x86_64__) || defined(__ia64__)
++#define XEN_ELF_LITTLE_ENDIAN
++#else
++#error define architectural endianness
++#endif
+
-+/*
-+ * Notifications after enqueuing any type of message should be conditional on
-+ * the appropriate req_event or rsp_event field in the shared ring.
-+ * If the client sends notification for rx requests then it should specify
-+ * feature 'feature-rx-notify' via xenbus. Otherwise the backend will assume
-+ * that it cannot safely queue packets (as it may not be kicked to send them).
-+ */
++#undef ELFSIZE
++#include "elfnote.h"
++#include "elfstructs.h"
++#include "features.h"
+
-+/*
-+ * This is the 'wire' format for packets:
-+ * Request 1: netif_tx_request -- NETTXF_* (any flags)
-+ * [Request 2: netif_tx_extra] (only if request 1 has NETTXF_extra_info)
-+ * [Request 3: netif_tx_extra] (only if request 2 has XEN_NETIF_EXTRA_MORE)
-+ * Request 4: netif_tx_request -- NETTXF_more_data
-+ * Request 5: netif_tx_request -- NETTXF_more_data
-+ * ...
-+ * Request N: netif_tx_request -- 0
-+ */
++/* ------------------------------------------------------------------------ */
+
-+/* Protocol checksum field is blank in the packet (hardware offload)? */
-+#define _NETTXF_csum_blank (0)
-+#define NETTXF_csum_blank (1U<<_NETTXF_csum_blank)
++typedef union {
++ Elf32_Ehdr e32;
++ Elf64_Ehdr e64;
++} elf_ehdr;
+
-+/* Packet data has been validated against protocol checksum. */
-+#define _NETTXF_data_validated (1)
-+#define NETTXF_data_validated (1U<<_NETTXF_data_validated)
++typedef union {
++ Elf32_Phdr e32;
++ Elf64_Phdr e64;
++} elf_phdr;
+
-+/* Packet continues in the next request descriptor. */
-+#define _NETTXF_more_data (2)
-+#define NETTXF_more_data (1U<<_NETTXF_more_data)
++typedef union {
++ Elf32_Shdr e32;
++ Elf64_Shdr e64;
++} elf_shdr;
+
-+/* Packet to be followed by extra descriptor(s). */
-+#define _NETTXF_extra_info (3)
-+#define NETTXF_extra_info (1U<<_NETTXF_extra_info)
++typedef union {
++ Elf32_Sym e32;
++ Elf64_Sym e64;
++} elf_sym;
+
-+struct netif_tx_request {
-+ grant_ref_t gref; /* Reference to buffer page */
-+ uint16_t offset; /* Offset within buffer page */
-+ uint16_t flags; /* NETTXF_* */
-+ uint16_t id; /* Echoed in response message. */
-+ uint16_t size; /* Packet size in bytes. */
++typedef union {
++ Elf32_Rel e32;
++ Elf64_Rel e64;
++} elf_rel;
++
++typedef union {
++ Elf32_Rela e32;
++ Elf64_Rela e64;
++} elf_rela;
++
++typedef union {
++ Elf32_Note e32;
++ Elf64_Note e64;
++} elf_note;
++
++struct elf_binary {
++ /* elf binary */
++ const char *image;
++ size_t size;
++ char class;
++ char data;
++
++ const elf_ehdr *ehdr;
++ const char *sec_strtab;
++ const elf_shdr *sym_tab;
++ const char *sym_strtab;
++
++ /* loaded to */
++ char *dest;
++ uint64_t pstart;
++ uint64_t pend;
++ uint64_t reloc_offset;
++
++ uint64_t bsd_symtab_pstart;
++ uint64_t bsd_symtab_pend;
++
++#ifndef __XEN__
++ /* misc */
++ FILE *log;
++#endif
++ int verbose;
+};
-+typedef struct netif_tx_request netif_tx_request_t;
+
-+/* Types of netif_extra_info descriptors. */
-+#define XEN_NETIF_EXTRA_TYPE_NONE (0) /* Never used - invalid */
-+#define XEN_NETIF_EXTRA_TYPE_GSO (1) /* u.gso */
-+#define XEN_NETIF_EXTRA_TYPE_MAX (2)
++/* ------------------------------------------------------------------------ */
++/* accessing elf header fields */
+
-+/* netif_extra_info flags. */
-+#define _XEN_NETIF_EXTRA_FLAG_MORE (0)
-+#define XEN_NETIF_EXTRA_FLAG_MORE (1U<<_XEN_NETIF_EXTRA_FLAG_MORE)
++#ifdef XEN_ELF_BIG_ENDIAN
++# define NATIVE_ELFDATA ELFDATA2MSB
++#else
++# define NATIVE_ELFDATA ELFDATA2LSB
++#endif
+
-+/* GSO types - only TCPv4 currently supported. */
-+#define XEN_NETIF_GSO_TYPE_TCPV4 (1)
++#define elf_32bit(elf) (ELFCLASS32 == (elf)->class)
++#define elf_64bit(elf) (ELFCLASS64 == (elf)->class)
++#define elf_msb(elf) (ELFDATA2MSB == (elf)->data)
++#define elf_lsb(elf) (ELFDATA2LSB == (elf)->data)
++#define elf_swap(elf) (NATIVE_ELFDATA != (elf)->data)
+
-+/*
-+ * This structure needs to fit within both netif_tx_request and
-+ * netif_rx_response for compatibility.
-+ */
-+struct netif_extra_info {
-+ uint8_t type; /* XEN_NETIF_EXTRA_TYPE_* */
-+ uint8_t flags; /* XEN_NETIF_EXTRA_FLAG_* */
++#define elf_uval(elf, str, elem) \
++ ((ELFCLASS64 == (elf)->class) \
++ ? elf_access_unsigned((elf), (str), \
++ offsetof(typeof(*(str)),e64.elem), \
++ sizeof((str)->e64.elem)) \
++ : elf_access_unsigned((elf), (str), \
++ offsetof(typeof(*(str)),e32.elem), \
++ sizeof((str)->e32.elem)))
++
++#define elf_sval(elf, str, elem) \
++ ((ELFCLASS64 == (elf)->class) \
++ ? elf_access_signed((elf), (str), \
++ offsetof(typeof(*(str)),e64.elem), \
++ sizeof((str)->e64.elem)) \
++ : elf_access_signed((elf), (str), \
++ offsetof(typeof(*(str)),e32.elem), \
++ sizeof((str)->e32.elem)))
++
++#define elf_size(elf, str) \
++ ((ELFCLASS64 == (elf)->class) \
++ ? sizeof((str)->e64) : sizeof((str)->e32))
+
-+ union {
-+ struct {
-+ /*
-+ * Maximum payload size of each segment. For example, for TCP this
-+ * is just the path MSS.
-+ */
-+ uint16_t size;
++uint64_t elf_access_unsigned(struct elf_binary *elf, const void *ptr,
++ uint64_t offset, size_t size);
++int64_t elf_access_signed(struct elf_binary *elf, const void *ptr,
++ uint64_t offset, size_t size);
+
-+ /*
-+ * GSO type. This determines the protocol of the packet and any
-+ * extra features required to segment the packet properly.
-+ */
-+ uint8_t type; /* XEN_NETIF_GSO_TYPE_* */
++uint64_t elf_round_up(struct elf_binary *elf, uint64_t addr);
+
-+ /* Future expansion. */
-+ uint8_t pad;
++/* ------------------------------------------------------------------------ */
++/* xc_libelf_tools.c */
+
-+ /*
-+ * GSO features. This specifies any extra GSO features required
-+ * to process this packet, such as ECN support for TCPv4.
-+ */
-+ uint16_t features; /* XEN_NETIF_GSO_FEAT_* */
-+ } gso;
++int elf_shdr_count(struct elf_binary *elf);
++int elf_phdr_count(struct elf_binary *elf);
+
-+ uint16_t pad[3];
-+ } u;
-+};
++const elf_shdr *elf_shdr_by_name(struct elf_binary *elf, const char *name);
++const elf_shdr *elf_shdr_by_index(struct elf_binary *elf, int index);
++const elf_phdr *elf_phdr_by_index(struct elf_binary *elf, int index);
+
-+struct netif_tx_response {
-+ uint16_t id;
-+ int16_t status; /* NETIF_RSP_* */
-+};
-+typedef struct netif_tx_response netif_tx_response_t;
++const char *elf_section_name(struct elf_binary *elf, const elf_shdr * shdr);
++const void *elf_section_start(struct elf_binary *elf, const elf_shdr * shdr);
++const void *elf_section_end(struct elf_binary *elf, const elf_shdr * shdr);
+
-+struct netif_rx_request {
-+ uint16_t id; /* Echoed in response message. */
-+ grant_ref_t gref; /* Reference to incoming granted frame */
-+};
-+typedef struct netif_rx_request netif_rx_request_t;
++const void *elf_segment_start(struct elf_binary *elf, const elf_phdr * phdr);
++const void *elf_segment_end(struct elf_binary *elf, const elf_phdr * phdr);
+
-+/* Packet data has been validated against protocol checksum. */
-+#define _NETRXF_data_validated (0)
-+#define NETRXF_data_validated (1U<<_NETRXF_data_validated)
++const elf_sym *elf_sym_by_name(struct elf_binary *elf, const char *symbol);
++const elf_sym *elf_sym_by_index(struct elf_binary *elf, int index);
+
-+/* Protocol checksum field is blank in the packet (hardware offload)? */
-+#define _NETRXF_csum_blank (1)
-+#define NETRXF_csum_blank (1U<<_NETRXF_csum_blank)
++const char *elf_note_name(struct elf_binary *elf, const elf_note * note);
++const void *elf_note_desc(struct elf_binary *elf, const elf_note * note);
++uint64_t elf_note_numeric(struct elf_binary *elf, const elf_note * note);
++const elf_note *elf_note_next(struct elf_binary *elf, const elf_note * note);
+
-+/* Packet continues in the next request descriptor. */
-+#define _NETRXF_more_data (2)
-+#define NETRXF_more_data (1U<<_NETRXF_more_data)
++int elf_is_elfbinary(const void *image);
++int elf_phdr_is_loadable(struct elf_binary *elf, const elf_phdr * phdr);
+
-+/* Packet to be followed by extra descriptor(s). */
-+#define _NETRXF_extra_info (3)
-+#define NETRXF_extra_info (1U<<_NETRXF_extra_info)
++/* ------------------------------------------------------------------------ */
++/* xc_libelf_loader.c */
++
++int elf_init(struct elf_binary *elf, const char *image, size_t size);
++#ifdef __XEN__
++void elf_set_verbose(struct elf_binary *elf);
++#else
++void elf_set_logfile(struct elf_binary *elf, FILE * log, int verbose);
++#endif
++
++void elf_parse_binary(struct elf_binary *elf);
++void elf_load_binary(struct elf_binary *elf);
++
++void *elf_get_ptr(struct elf_binary *elf, unsigned long addr);
++uint64_t elf_lookup_addr(struct elf_binary *elf, const char *symbol);
++
++void elf_parse_bsdsyms(struct elf_binary *elf, uint64_t pstart); /* private */
++
++/* ------------------------------------------------------------------------ */
++/* xc_libelf_relocate.c */
++
++int elf_reloc(struct elf_binary *elf);
++
++/* ------------------------------------------------------------------------ */
++/* xc_libelf_dominfo.c */
++
++#define UNSET_ADDR ((uint64_t)-1)
+
-+struct netif_rx_response {
-+ uint16_t id;
-+ uint16_t offset; /* Offset in page of start of received packet */
-+ uint16_t flags; /* NETRXF_* */
-+ int16_t status; /* -ve: BLKIF_RSP_* ; +ve: Rx'ed pkt size. */
++enum xen_elfnote_type {
++ XEN_ENT_NONE = 0,
++ XEN_ENT_LONG = 1,
++ XEN_ENT_STR = 2
+};
-+typedef struct netif_rx_response netif_rx_response_t;
+
-+/*
-+ * Generate netif ring structures and types.
-+ */
++struct xen_elfnote {
++ enum xen_elfnote_type type;
++ const char *name;
++ union {
++ const char *str;
++ uint64_t num;
++ } data;
++};
+
-+DEFINE_RING_TYPES(netif_tx, struct netif_tx_request, struct netif_tx_response);
-+DEFINE_RING_TYPES(netif_rx, struct netif_rx_request, struct netif_rx_response);
++struct elf_dom_parms {
++ /* raw */
++ const char *guest_info;
++ const void *elf_note_start;
++ const void *elf_note_end;
++ struct xen_elfnote elf_notes[XEN_ELFNOTE_MAX + 1];
++
++ /* parsed */
++ char guest_os[16];
++ char guest_ver[16];
++ char xen_ver[16];
++ char loader[16];
++ int pae;
++ int bsd_symtab;
++ uint64_t virt_base;
++ uint64_t virt_entry;
++ uint64_t virt_hypercall;
++ uint64_t virt_hv_start_low;
++ uint64_t elf_paddr_offset;
++ uint32_t f_supported[XENFEAT_NR_SUBMAPS];
++ uint32_t f_required[XENFEAT_NR_SUBMAPS];
+
-+#define NETIF_RSP_DROPPED -2
-+#define NETIF_RSP_ERROR -1
-+#define NETIF_RSP_OKAY 0
-+/* No response: used for auxiliary requests (e.g., netif_tx_extra). */
-+#define NETIF_RSP_NULL 1
++ /* calculated */
++ uint64_t virt_offset;
++ uint64_t virt_kstart;
++ uint64_t virt_kend;
++};
+
-+#endif
++static inline void elf_xen_feature_set(int nr, uint32_t * addr)
++{
++ addr[nr >> 5] |= 1 << (nr & 31);
++}
++static inline int elf_xen_feature_get(int nr, uint32_t * addr)
++{
++ return !!(addr[nr >> 5] & (1 << (nr & 31)));
++}
+
-+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
-+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/io/pciif.h tmp-linux-2.6-xen.patch/include/xen/interface/io/pciif.h
---- pristine-linux-2.6.18.2/include/xen/interface/io/pciif.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/io/pciif.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,83 @@
-+/*
-+ * PCI Backend/Frontend Common Data Structures & Macros
-+ *
++int elf_xen_parse_features(const char *features,
++ uint32_t *supported,
++ uint32_t *required);
++int elf_xen_parse_note(struct elf_binary *elf,
++ struct elf_dom_parms *parms,
++ const elf_note *note);
++int elf_xen_parse_guest_info(struct elf_binary *elf,
++ struct elf_dom_parms *parms);
++int elf_xen_parse(struct elf_binary *elf,
++ struct elf_dom_parms *parms);
++
++#endif /* __XC_LIBELF__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/memory.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/memory.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,312 @@
++/******************************************************************************
++ * memory.h
++ *
++ * Memory reservation and information.
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -101857,386 +140946,367 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ * Copyright (c) 2005, Keir Fraser <keir at xensource.com>
+ */
-+#ifndef __XEN_PCI_COMMON_H__
-+#define __XEN_PCI_COMMON_H__
-+
-+/* Be sure to bump this number if you change this file */
-+#define XEN_PCI_MAGIC "7"
-+
-+/* xen_pci_sharedinfo flags */
-+#define _XEN_PCIF_active (0)
-+#define XEN_PCIF_active (1<<_XEN_PCI_active)
+
-+/* xen_pci_op commands */
-+#define XEN_PCI_OP_conf_read (0)
-+#define XEN_PCI_OP_conf_write (1)
++#ifndef __XEN_PUBLIC_MEMORY_H__
++#define __XEN_PUBLIC_MEMORY_H__
+
-+/* xen_pci_op error numbers */
-+#define XEN_PCI_ERR_success (0)
-+#define XEN_PCI_ERR_dev_not_found (-1)
-+#define XEN_PCI_ERR_invalid_offset (-2)
-+#define XEN_PCI_ERR_access_denied (-3)
-+#define XEN_PCI_ERR_not_implemented (-4)
-+/* XEN_PCI_ERR_op_failed - backend failed to complete the operation */
-+#define XEN_PCI_ERR_op_failed (-5)
++/*
++ * Increase or decrease the specified domain's memory reservation. Returns the
++ * number of extents successfully allocated or freed.
++ * arg == addr of struct xen_memory_reservation.
++ */
++#define XENMEM_increase_reservation 0
++#define XENMEM_decrease_reservation 1
++#define XENMEM_populate_physmap 6
+
-+struct xen_pci_op {
-+ /* IN: what action to perform: XEN_PCI_OP_* */
-+ uint32_t cmd;
++#if __XEN_INTERFACE_VERSION__ >= 0x00030209
++/*
++ * Maximum # bits addressable by the user of the allocated region (e.g., I/O
++ * devices often have a 32-bit limitation even in 64-bit systems). If zero
++ * then the user has no addressing restriction. This field is not used by
++ * XENMEM_decrease_reservation.
++ */
++#define XENMEMF_address_bits(x) (x)
++#define XENMEMF_get_address_bits(x) ((x) & 0xffu)
++/* NUMA node to allocate from. */
++#define XENMEMF_node(x) (((x) + 1) << 8)
++#define XENMEMF_get_node(x) ((((x) >> 8) - 1) & 0xffu)
++#endif
+
-+ /* OUT: will contain an error number (if any) from errno.h */
-+ int32_t err;
++struct xen_memory_reservation {
+
-+ /* IN: which device to touch */
-+ uint32_t domain; /* PCI Domain/Segment */
-+ uint32_t bus;
-+ uint32_t devfn;
++ /*
++ * XENMEM_increase_reservation:
++ * OUT: MFN (*not* GMFN) bases of extents that were allocated
++ * XENMEM_decrease_reservation:
++ * IN: GMFN bases of extents to free
++ * XENMEM_populate_physmap:
++ * IN: GPFN bases of extents to populate with memory
++ * OUT: GMFN bases of extents that were allocated
++ * (NB. This command also updates the mach_to_phys translation table)
++ */
++ XEN_GUEST_HANDLE(xen_pfn_t) extent_start;
+
-+ /* IN: which configuration registers to touch */
-+ int32_t offset;
-+ int32_t size;
++ /* Number of extents, and size/alignment of each (2^extent_order pages). */
++ xen_ulong_t nr_extents;
++ unsigned int extent_order;
+
-+ /* IN/OUT: Contains the result after a READ or the value to WRITE */
-+ uint32_t value;
-+};
++#if __XEN_INTERFACE_VERSION__ >= 0x00030209
++ /* XENMEMF flags. */
++ unsigned int mem_flags;
++#else
++ unsigned int address_bits;
++#endif
+
-+struct xen_pci_sharedinfo {
-+ /* flags - XEN_PCIF_* */
-+ uint32_t flags;
-+ struct xen_pci_op op;
++ /*
++ * Domain whose reservation is being changed.
++ * Unprivileged domains can specify only DOMID_SELF.
++ */
++ domid_t domid;
+};
-+
-+#endif /* __XEN_PCI_COMMON_H__ */
++typedef struct xen_memory_reservation xen_memory_reservation_t;
++DEFINE_XEN_GUEST_HANDLE(xen_memory_reservation_t);
+
+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
++ * An atomic exchange of memory pages. If return code is zero then
++ * @out.extent_list provides GMFNs of the newly-allocated memory.
++ * Returns zero on complete success, otherwise a negative error code.
++ * On complete success then always @nr_exchanged == @in.nr_extents.
++ * On partial success @nr_exchanged indicates how much work was done.
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/io/protocols.h tmp-linux-2.6-xen.patch/include/xen/interface/io/protocols.h
---- pristine-linux-2.6.18.2/include/xen/interface/io/protocols.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/io/protocols.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,21 @@
-+#ifndef __XEN_PROTOCOLS_H__
-+#define __XEN_PROTOCOLS_H__
++#define XENMEM_exchange 11
++struct xen_memory_exchange {
++ /*
++ * [IN] Details of memory extents to be exchanged (GMFN bases).
++ * Note that @in.address_bits is ignored and unused.
++ */
++ struct xen_memory_reservation in;
+
-+#define XEN_IO_PROTO_ABI_X86_32 "x86_32-abi"
-+#define XEN_IO_PROTO_ABI_X86_64 "x86_64-abi"
-+#define XEN_IO_PROTO_ABI_IA64 "ia64-abi"
-+#define XEN_IO_PROTO_ABI_POWERPC64 "powerpc64-abi"
++ /*
++ * [IN/OUT] Details of new memory extents.
++ * We require that:
++ * 1. @in.domid == @out.domid
++ * 2. @in.nr_extents << @in.extent_order ==
++ * @out.nr_extents << @out.extent_order
++ * 3. @in.extent_start and @out.extent_start lists must not overlap
++ * 4. @out.extent_start lists GPFN bases to be populated
++ * 5. @out.extent_start is overwritten with allocated GMFN bases
++ */
++ struct xen_memory_reservation out;
+
-+#if defined(__i386__)
-+# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_X86_32
-+#elif defined(__x86_64__)
-+# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_X86_64
-+#elif defined(__ia64__)
-+# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_IA64
-+#elif defined(__powerpc64__)
-+# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_POWERPC64
-+#else
-+# error arch fixup needed here
-+#endif
++ /*
++ * [OUT] Number of input extents that were successfully exchanged:
++ * 1. The first @nr_exchanged input extents were successfully
++ * deallocated.
++ * 2. The corresponding first entries in the output extent list correctly
++ * indicate the GMFNs that were successfully exchanged.
++ * 3. All other input and output extents are untouched.
++ * 4. If not all input exents are exchanged then the return code of this
++ * command will be non-zero.
++ * 5. THIS FIELD MUST BE INITIALISED TO ZERO BY THE CALLER!
++ */
++ xen_ulong_t nr_exchanged;
++};
++typedef struct xen_memory_exchange xen_memory_exchange_t;
++DEFINE_XEN_GUEST_HANDLE(xen_memory_exchange_t);
+
-+#endif
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/io/ring.h tmp-linux-2.6-xen.patch/include/xen/interface/io/ring.h
---- pristine-linux-2.6.18.2/include/xen/interface/io/ring.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/io/ring.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,299 @@
-+/******************************************************************************
-+ * ring.h
-+ *
-+ * Shared producer-consumer ring macros.
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Tim Deegan and Andrew Warfield November 2004.
++/*
++ * Returns the maximum machine frame number of mapped RAM in this system.
++ * This command always succeeds (it never returns an error code).
++ * arg == NULL.
+ */
++#define XENMEM_maximum_ram_page 2
+
-+#ifndef __XEN_PUBLIC_IO_RING_H__
-+#define __XEN_PUBLIC_IO_RING_H__
-+
-+typedef unsigned int RING_IDX;
-+
-+/* Round a 32-bit unsigned constant down to the nearest power of two. */
-+#define __RD2(_x) (((_x) & 0x00000002) ? 0x2 : ((_x) & 0x1))
-+#define __RD4(_x) (((_x) & 0x0000000c) ? __RD2((_x)>>2)<<2 : __RD2(_x))
-+#define __RD8(_x) (((_x) & 0x000000f0) ? __RD4((_x)>>4)<<4 : __RD4(_x))
-+#define __RD16(_x) (((_x) & 0x0000ff00) ? __RD8((_x)>>8)<<8 : __RD8(_x))
-+#define __RD32(_x) (((_x) & 0xffff0000) ? __RD16((_x)>>16)<<16 : __RD16(_x))
++/*
++ * Returns the current or maximum memory reservation, in pages, of the
++ * specified domain (may be DOMID_SELF). Returns -ve errcode on failure.
++ * arg == addr of domid_t.
++ */
++#define XENMEM_current_reservation 3
++#define XENMEM_maximum_reservation 4
+
+/*
-+ * Calculate size of a shared ring, given the total available space for the
-+ * ring and indexes (_sz), and the name tag of the request/response structure.
-+ * A ring contains as many entries as will fit, rounded down to the nearest
-+ * power of two (so we can mask with (size-1) to loop around).
++ * Returns the maximum GPFN in use by the guest, or -ve errcode on failure.
+ */
-+#define __RING_SIZE(_s, _sz) \
-+ (__RD32(((_sz) - (long)(_s)->ring + (long)(_s)) / sizeof((_s)->ring[0])))
++#define XENMEM_maximum_gpfn 14
+
+/*
-+ * Macros to make the correct C datatypes for a new kind of ring.
-+ *
-+ * To make a new ring datatype, you need to have two message structures,
-+ * let's say request_t, and response_t already defined.
-+ *
-+ * In a header where you want the ring datatype declared, you then do:
-+ *
-+ * DEFINE_RING_TYPES(mytag, request_t, response_t);
-+ *
-+ * These expand out to give you a set of types, as you can see below.
-+ * The most important of these are:
-+ *
-+ * mytag_sring_t - The shared ring.
-+ * mytag_front_ring_t - The 'front' half of the ring.
-+ * mytag_back_ring_t - The 'back' half of the ring.
-+ *
-+ * To initialize a ring in your code you need to know the location and size
-+ * of the shared memory area (PAGE_SIZE, for instance). To initialise
-+ * the front half:
-+ *
-+ * mytag_front_ring_t front_ring;
-+ * SHARED_RING_INIT((mytag_sring_t *)shared_page);
-+ * FRONT_RING_INIT(&front_ring, (mytag_sring_t *)shared_page, PAGE_SIZE);
-+ *
-+ * Initializing the back follows similarly (note that only the front
-+ * initializes the shared ring):
-+ *
-+ * mytag_back_ring_t back_ring;
-+ * BACK_RING_INIT(&back_ring, (mytag_sring_t *)shared_page, PAGE_SIZE);
++ * Returns a list of MFN bases of 2MB extents comprising the machine_to_phys
++ * mapping table. Architectures which do not have a m2p table do not implement
++ * this command.
++ * arg == addr of xen_machphys_mfn_list_t.
+ */
++#define XENMEM_machphys_mfn_list 5
++struct xen_machphys_mfn_list {
++ /*
++ * Size of the 'extent_start' array. Fewer entries will be filled if the
++ * machphys table is smaller than max_extents * 2MB.
++ */
++ unsigned int max_extents;
+
-+#define DEFINE_RING_TYPES(__name, __req_t, __rsp_t) \
-+ \
-+/* Shared ring entry */ \
-+union __name##_sring_entry { \
-+ __req_t req; \
-+ __rsp_t rsp; \
-+}; \
-+ \
-+/* Shared ring page */ \
-+struct __name##_sring { \
-+ RING_IDX req_prod, req_event; \
-+ RING_IDX rsp_prod, rsp_event; \
-+ uint8_t pad[48]; \
-+ union __name##_sring_entry ring[1]; /* variable-length */ \
-+}; \
-+ \
-+/* "Front" end's private variables */ \
-+struct __name##_front_ring { \
-+ RING_IDX req_prod_pvt; \
-+ RING_IDX rsp_cons; \
-+ unsigned int nr_ents; \
-+ struct __name##_sring *sring; \
-+}; \
-+ \
-+/* "Back" end's private variables */ \
-+struct __name##_back_ring { \
-+ RING_IDX rsp_prod_pvt; \
-+ RING_IDX req_cons; \
-+ unsigned int nr_ents; \
-+ struct __name##_sring *sring; \
-+}; \
-+ \
-+/* Syntactic sugar */ \
-+typedef struct __name##_sring __name##_sring_t; \
-+typedef struct __name##_front_ring __name##_front_ring_t; \
-+typedef struct __name##_back_ring __name##_back_ring_t
++ /*
++ * Pointer to buffer to fill with list of extent starts. If there are
++ * any large discontiguities in the machine address space, 2MB gaps in
++ * the machphys table will be represented by an MFN base of zero.
++ */
++ XEN_GUEST_HANDLE(xen_pfn_t) extent_start;
++
++ /*
++ * Number of extents written to the above array. This will be smaller
++ * than 'max_extents' if the machphys table is smaller than max_e * 2MB.
++ */
++ unsigned int nr_extents;
++};
++typedef struct xen_machphys_mfn_list xen_machphys_mfn_list_t;
++DEFINE_XEN_GUEST_HANDLE(xen_machphys_mfn_list_t);
+
+/*
-+ * Macros for manipulating rings.
-+ *
-+ * FRONT_RING_whatever works on the "front end" of a ring: here
-+ * requests are pushed on to the ring and responses taken off it.
-+ *
-+ * BACK_RING_whatever works on the "back end" of a ring: here
-+ * requests are taken off the ring and responses put on.
-+ *
-+ * N.B. these macros do NO INTERLOCKS OR FLOW CONTROL.
-+ * This is OK in 1-for-1 request-response situations where the
-+ * requestor (front end) never has more than RING_SIZE()-1
-+ * outstanding requests.
++ * Returns the location in virtual address space of the machine_to_phys
++ * mapping table. Architectures which do not have a m2p table, or which do not
++ * map it by default into guest address space, do not implement this command.
++ * arg == addr of xen_machphys_mapping_t.
+ */
++#define XENMEM_machphys_mapping 12
++struct xen_machphys_mapping {
++ xen_ulong_t v_start, v_end; /* Start and end virtual addresses. */
++ xen_ulong_t max_mfn; /* Maximum MFN that can be looked up. */
++};
++typedef struct xen_machphys_mapping xen_machphys_mapping_t;
++DEFINE_XEN_GUEST_HANDLE(xen_machphys_mapping_t);
+
-+/* Initialising empty rings */
-+#define SHARED_RING_INIT(_s) do { \
-+ (_s)->req_prod = (_s)->rsp_prod = 0; \
-+ (_s)->req_event = (_s)->rsp_event = 1; \
-+ (void)memset((_s)->pad, 0, sizeof((_s)->pad)); \
-+} while(0)
-+
-+#define FRONT_RING_INIT(_r, _s, __size) do { \
-+ (_r)->req_prod_pvt = 0; \
-+ (_r)->rsp_cons = 0; \
-+ (_r)->nr_ents = __RING_SIZE(_s, __size); \
-+ (_r)->sring = (_s); \
-+} while (0)
++/*
++ * Sets the GPFN at which a particular page appears in the specified guest's
++ * pseudophysical address space.
++ * arg == addr of xen_add_to_physmap_t.
++ */
++#define XENMEM_add_to_physmap 7
++struct xen_add_to_physmap {
++ /* Which domain to change the mapping for. */
++ domid_t domid;
+
-+#define BACK_RING_INIT(_r, _s, __size) do { \
-+ (_r)->rsp_prod_pvt = 0; \
-+ (_r)->req_cons = 0; \
-+ (_r)->nr_ents = __RING_SIZE(_s, __size); \
-+ (_r)->sring = (_s); \
-+} while (0)
++ /* Source mapping space. */
++#define XENMAPSPACE_shared_info 0 /* shared info page */
++#define XENMAPSPACE_grant_table 1 /* grant table page */
++#define XENMAPSPACE_mfn 2 /* usual MFN */
++ unsigned int space;
+
-+/* Initialize to existing shared indexes -- for recovery */
-+#define FRONT_RING_ATTACH(_r, _s, __size) do { \
-+ (_r)->sring = (_s); \
-+ (_r)->req_prod_pvt = (_s)->req_prod; \
-+ (_r)->rsp_cons = (_s)->rsp_prod; \
-+ (_r)->nr_ents = __RING_SIZE(_s, __size); \
-+} while (0)
++ /* Index into source mapping space. */
++ xen_ulong_t idx;
+
-+#define BACK_RING_ATTACH(_r, _s, __size) do { \
-+ (_r)->sring = (_s); \
-+ (_r)->rsp_prod_pvt = (_s)->rsp_prod; \
-+ (_r)->req_cons = (_s)->req_prod; \
-+ (_r)->nr_ents = __RING_SIZE(_s, __size); \
-+} while (0)
++ /* GPFN where the source mapping page should appear. */
++ xen_pfn_t gpfn;
++};
++typedef struct xen_add_to_physmap xen_add_to_physmap_t;
++DEFINE_XEN_GUEST_HANDLE(xen_add_to_physmap_t);
+
-+/* How big is this ring? */
-+#define RING_SIZE(_r) \
-+ ((_r)->nr_ents)
++/*
++ * Unmaps the page appearing at a particular GPFN from the specified guest's
++ * pseudophysical address space.
++ * arg == addr of xen_remove_from_physmap_t.
++ */
++#define XENMEM_remove_from_physmap 15
++struct xen_remove_from_physmap {
++ /* Which domain to change the mapping for. */
++ domid_t domid;
+
-+/* Number of free requests (for use on front side only). */
-+#define RING_FREE_REQUESTS(_r) \
-+ (RING_SIZE(_r) - ((_r)->req_prod_pvt - (_r)->rsp_cons))
++ /* GPFN of the current mapping of the page. */
++ xen_pfn_t gpfn;
++};
++typedef struct xen_remove_from_physmap xen_remove_from_physmap_t;
++DEFINE_XEN_GUEST_HANDLE(xen_remove_from_physmap_t);
+
-+/* Test if there is an empty slot available on the front ring.
-+ * (This is only meaningful from the front. )
++/*
++ * Translates a list of domain-specific GPFNs into MFNs. Returns a -ve error
++ * code on failure. This call only works for auto-translated guests.
+ */
-+#define RING_FULL(_r) \
-+ (RING_FREE_REQUESTS(_r) == 0)
++#define XENMEM_translate_gpfn_list 8
++struct xen_translate_gpfn_list {
++ /* Which domain to translate for? */
++ domid_t domid;
+
-+/* Test if there are outstanding messages to be processed on a ring. */
-+#define RING_HAS_UNCONSUMED_RESPONSES(_r) \
-+ ((_r)->sring->rsp_prod - (_r)->rsp_cons)
++ /* Length of list. */
++ xen_ulong_t nr_gpfns;
+
-+#ifdef __GNUC__
-+#define RING_HAS_UNCONSUMED_REQUESTS(_r) ({ \
-+ unsigned int req = (_r)->sring->req_prod - (_r)->req_cons; \
-+ unsigned int rsp = RING_SIZE(_r) - \
-+ ((_r)->req_cons - (_r)->rsp_prod_pvt); \
-+ req < rsp ? req : rsp; \
-+})
-+#else
-+/* Same as above, but without the nice GCC ({ ... }) syntax. */
-+#define RING_HAS_UNCONSUMED_REQUESTS(_r) \
-+ ((((_r)->sring->req_prod - (_r)->req_cons) < \
-+ (RING_SIZE(_r) - ((_r)->req_cons - (_r)->rsp_prod_pvt))) ? \
-+ ((_r)->sring->req_prod - (_r)->req_cons) : \
-+ (RING_SIZE(_r) - ((_r)->req_cons - (_r)->rsp_prod_pvt)))
-+#endif
++ /* List of GPFNs to translate. */
++ XEN_GUEST_HANDLE(xen_pfn_t) gpfn_list;
+
-+/* Direct access to individual ring elements, by index. */
-+#define RING_GET_REQUEST(_r, _idx) \
-+ (&((_r)->sring->ring[((_idx) & (RING_SIZE(_r) - 1))].req))
++ /*
++ * Output list to contain MFN translations. May be the same as the input
++ * list (in which case each input GPFN is overwritten with the output MFN).
++ */
++ XEN_GUEST_HANDLE(xen_pfn_t) mfn_list;
++};
++typedef struct xen_translate_gpfn_list xen_translate_gpfn_list_t;
++DEFINE_XEN_GUEST_HANDLE(xen_translate_gpfn_list_t);
++
++/*
++ * Returns the pseudo-physical memory map as it was when the domain
++ * was started (specified by XENMEM_set_memory_map).
++ * arg == addr of xen_memory_map_t.
++ */
++#define XENMEM_memory_map 9
++struct xen_memory_map {
++ /*
++ * On call the number of entries which can be stored in buffer. On
++ * return the number of entries which have been stored in
++ * buffer.
++ */
++ unsigned int nr_entries;
+
-+#define RING_GET_RESPONSE(_r, _idx) \
-+ (&((_r)->sring->ring[((_idx) & (RING_SIZE(_r) - 1))].rsp))
++ /*
++ * Entries in the buffer are in the same format as returned by the
++ * BIOS INT 0x15 EAX=0xE820 call.
++ */
++ XEN_GUEST_HANDLE(void) buffer;
++};
++typedef struct xen_memory_map xen_memory_map_t;
++DEFINE_XEN_GUEST_HANDLE(xen_memory_map_t);
+
-+/* Loop termination condition: Would the specified index overflow the ring? */
-+#define RING_REQUEST_CONS_OVERFLOW(_r, _cons) \
-+ (((_cons) - (_r)->rsp_prod_pvt) >= RING_SIZE(_r))
++/*
++ * Returns the real physical memory map. Passes the same structure as
++ * XENMEM_memory_map.
++ * arg == addr of xen_memory_map_t.
++ */
++#define XENMEM_machine_memory_map 10
+
-+#define RING_PUSH_REQUESTS(_r) do { \
-+ wmb(); /* back sees requests /before/ updated producer index */ \
-+ (_r)->sring->req_prod = (_r)->req_prod_pvt; \
-+} while (0)
++/*
++ * Set the pseudo-physical memory map of a domain, as returned by
++ * XENMEM_memory_map.
++ * arg == addr of xen_foreign_memory_map_t.
++ */
++#define XENMEM_set_memory_map 13
++struct xen_foreign_memory_map {
++ domid_t domid;
++ struct xen_memory_map map;
++};
++typedef struct xen_foreign_memory_map xen_foreign_memory_map_t;
++DEFINE_XEN_GUEST_HANDLE(xen_foreign_memory_map_t);
+
-+#define RING_PUSH_RESPONSES(_r) do { \
-+ wmb(); /* front sees responses /before/ updated producer index */ \
-+ (_r)->sring->rsp_prod = (_r)->rsp_prod_pvt; \
-+} while (0)
++#endif /* __XEN_PUBLIC_MEMORY_H__ */
+
+/*
-+ * Notification hold-off (req_event and rsp_event):
-+ *
-+ * When queueing requests or responses on a shared ring, it may not always be
-+ * necessary to notify the remote end. For example, if requests are in flight
-+ * in a backend, the front may be able to queue further requests without
-+ * notifying the back (if the back checks for new requests when it queues
-+ * responses).
-+ *
-+ * When enqueuing requests or responses:
-+ *
-+ * Use RING_PUSH_{REQUESTS,RESPONSES}_AND_CHECK_NOTIFY(). The second argument
-+ * is a boolean return value. True indicates that the receiver requires an
-+ * asynchronous notification.
-+ *
-+ * After dequeuing requests or responses (before sleeping the connection):
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/nmi.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/nmi.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,78 @@
++/******************************************************************************
++ * nmi.h
+ *
-+ * Use RING_FINAL_CHECK_FOR_REQUESTS() or RING_FINAL_CHECK_FOR_RESPONSES().
-+ * The second argument is a boolean return value. True indicates that there
-+ * are pending messages on the ring (i.e., the connection should not be put
-+ * to sleep).
++ * NMI callback registration and reason codes.
+ *
-+ * These macros will set the req_event/rsp_event field to trigger a
-+ * notification on the very next message that is enqueued. If you want to
-+ * create batches of work (i.e., only receive a notification after several
-+ * messages have been enqueued) then you will need to create a customised
-+ * version of the FINAL_CHECK macro in your own code, which sets the event
-+ * field appropriately.
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2005, Keir Fraser <keir at xensource.com>
+ */
+
-+#define RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(_r, _notify) do { \
-+ RING_IDX __old = (_r)->sring->req_prod; \
-+ RING_IDX __new = (_r)->req_prod_pvt; \
-+ wmb(); /* back sees requests /before/ updated producer index */ \
-+ (_r)->sring->req_prod = __new; \
-+ mb(); /* back sees new requests /before/ we check req_event */ \
-+ (_notify) = ((RING_IDX)(__new - (_r)->sring->req_event) < \
-+ (RING_IDX)(__new - __old)); \
-+} while (0)
++#ifndef __XEN_PUBLIC_NMI_H__
++#define __XEN_PUBLIC_NMI_H__
+
-+#define RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(_r, _notify) do { \
-+ RING_IDX __old = (_r)->sring->rsp_prod; \
-+ RING_IDX __new = (_r)->rsp_prod_pvt; \
-+ wmb(); /* front sees responses /before/ updated producer index */ \
-+ (_r)->sring->rsp_prod = __new; \
-+ mb(); /* front sees new responses /before/ we check rsp_event */ \
-+ (_notify) = ((RING_IDX)(__new - (_r)->sring->rsp_event) < \
-+ (RING_IDX)(__new - __old)); \
-+} while (0)
++/*
++ * NMI reason codes:
++ * Currently these are x86-specific, stored in arch_shared_info.nmi_reason.
++ */
++ /* I/O-check error reported via ISA port 0x61, bit 6. */
++#define _XEN_NMIREASON_io_error 0
++#define XEN_NMIREASON_io_error (1UL << _XEN_NMIREASON_io_error)
++ /* Parity error reported via ISA port 0x61, bit 7. */
++#define _XEN_NMIREASON_parity_error 1
++#define XEN_NMIREASON_parity_error (1UL << _XEN_NMIREASON_parity_error)
++ /* Unknown hardware-generated NMI. */
++#define _XEN_NMIREASON_unknown 2
++#define XEN_NMIREASON_unknown (1UL << _XEN_NMIREASON_unknown)
+
-+#define RING_FINAL_CHECK_FOR_REQUESTS(_r, _work_to_do) do { \
-+ (_work_to_do) = RING_HAS_UNCONSUMED_REQUESTS(_r); \
-+ if (_work_to_do) break; \
-+ (_r)->sring->req_event = (_r)->req_cons + 1; \
-+ mb(); \
-+ (_work_to_do) = RING_HAS_UNCONSUMED_REQUESTS(_r); \
-+} while (0)
++/*
++ * long nmi_op(unsigned int cmd, void *arg)
++ * NB. All ops return zero on success, else a negative error code.
++ */
+
-+#define RING_FINAL_CHECK_FOR_RESPONSES(_r, _work_to_do) do { \
-+ (_work_to_do) = RING_HAS_UNCONSUMED_RESPONSES(_r); \
-+ if (_work_to_do) break; \
-+ (_r)->sring->rsp_event = (_r)->rsp_cons + 1; \
-+ mb(); \
-+ (_work_to_do) = RING_HAS_UNCONSUMED_RESPONSES(_r); \
-+} while (0)
++/*
++ * Register NMI callback for this (calling) VCPU. Currently this only makes
++ * sense for domain 0, vcpu 0. All other callers will be returned EINVAL.
++ * arg == pointer to xennmi_callback structure.
++ */
++#define XENNMI_register_callback 0
++struct xennmi_callback {
++ unsigned long handler_address;
++ unsigned long pad;
++};
++typedef struct xennmi_callback xennmi_callback_t;
++DEFINE_XEN_GUEST_HANDLE(xennmi_callback_t);
+
-+#endif /* __XEN_PUBLIC_IO_RING_H__ */
++/*
++ * Deregister NMI callback for this (calling) VCPU.
++ * arg == NULL.
++ */
++#define XENNMI_unregister_callback 1
++
++#endif /* __XEN_PUBLIC_NMI_H__ */
+
+/*
+ * Local variables:
@@ -102247,15 +141317,11 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/io/tpmif.h tmp-linux-2.6-xen.patch/include/xen/interface/io/tpmif.h
---- pristine-linux-2.6.18.2/include/xen/interface/io/tpmif.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/io/tpmif.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,77 @@
-+/******************************************************************************
-+ * tpmif.h
-+ *
-+ * TPM I/O interface for Xen guest OSes.
-+ *
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/physdev.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/physdev.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,219 @@
++/*
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -102273,51 +141339,197 @@
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Copyright (c) 2005, IBM Corporation
-+ *
-+ * Author: Stefan Berger, stefanb at us.ibm.com
-+ * Grant table support: Mahadevan Gomathisankaran
-+ *
-+ * This code has been derived from tools/libxc/xen/io/netif.h
-+ *
-+ * Copyright (c) 2003-2004, Keir Fraser
+ */
+
-+#ifndef __XEN_PUBLIC_IO_TPMIF_H__
-+#define __XEN_PUBLIC_IO_TPMIF_H__
++#ifndef __XEN_PUBLIC_PHYSDEV_H__
++#define __XEN_PUBLIC_PHYSDEV_H__
+
-+#include "../grant_table.h"
++/*
++ * Prototype for this hypercall is:
++ * int physdev_op(int cmd, void *args)
++ * @cmd == PHYSDEVOP_??? (physdev operation).
++ * @args == Operation-specific extra arguments (NULL if none).
++ */
+
-+struct tpmif_tx_request {
-+ unsigned long addr; /* Machine address of packet. */
-+ grant_ref_t ref; /* grant table access reference */
-+ uint16_t unused;
-+ uint16_t size; /* Packet size in bytes. */
++/*
++ * Notify end-of-interrupt (EOI) for the specified IRQ.
++ * @arg == pointer to physdev_eoi structure.
++ */
++#define PHYSDEVOP_eoi 12
++struct physdev_eoi {
++ /* IN */
++ uint32_t irq;
+};
-+typedef struct tpmif_tx_request tpmif_tx_request_t;
++typedef struct physdev_eoi physdev_eoi_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_eoi_t);
+
+/*
-+ * The TPMIF_TX_RING_SIZE defines the number of pages the
-+ * front-end and backend can exchange (= size of array).
++ * Query the status of an IRQ line.
++ * @arg == pointer to physdev_irq_status_query structure.
+ */
-+typedef uint32_t TPMIF_RING_IDX;
++#define PHYSDEVOP_irq_status_query 5
++struct physdev_irq_status_query {
++ /* IN */
++ uint32_t irq;
++ /* OUT */
++ uint32_t flags; /* XENIRQSTAT_* */
++};
++typedef struct physdev_irq_status_query physdev_irq_status_query_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_irq_status_query_t);
+
-+#define TPMIF_TX_RING_SIZE 1
++/* Need to call PHYSDEVOP_eoi when the IRQ has been serviced? */
++#define _XENIRQSTAT_needs_eoi (0)
++#define XENIRQSTAT_needs_eoi (1U<<_XENIRQSTAT_needs_eoi)
+
-+/* This structure must fit in a memory page. */
++/* IRQ shared by multiple guests? */
++#define _XENIRQSTAT_shared (1)
++#define XENIRQSTAT_shared (1U<<_XENIRQSTAT_shared)
+
-+struct tpmif_ring {
-+ struct tpmif_tx_request req;
++/*
++ * Set the current VCPU's I/O privilege level.
++ * @arg == pointer to physdev_set_iopl structure.
++ */
++#define PHYSDEVOP_set_iopl 6
++struct physdev_set_iopl {
++ /* IN */
++ uint32_t iopl;
+};
-+typedef struct tpmif_ring tpmif_ring_t;
++typedef struct physdev_set_iopl physdev_set_iopl_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_set_iopl_t);
+
-+struct tpmif_tx_interface {
-+ struct tpmif_ring ring[TPMIF_TX_RING_SIZE];
++/*
++ * Set the current VCPU's I/O-port permissions bitmap.
++ * @arg == pointer to physdev_set_iobitmap structure.
++ */
++#define PHYSDEVOP_set_iobitmap 7
++struct physdev_set_iobitmap {
++ /* IN */
++#if __XEN_INTERFACE_VERSION__ >= 0x00030205
++ XEN_GUEST_HANDLE(uint8) bitmap;
++#else
++ uint8_t *bitmap;
++#endif
++ uint32_t nr_ports;
+};
-+typedef struct tpmif_tx_interface tpmif_tx_interface_t;
++typedef struct physdev_set_iobitmap physdev_set_iobitmap_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_set_iobitmap_t);
+
-+#endif
++/*
++ * Read or write an IO-APIC register.
++ * @arg == pointer to physdev_apic structure.
++ */
++#define PHYSDEVOP_apic_read 8
++#define PHYSDEVOP_apic_write 9
++struct physdev_apic {
++ /* IN */
++ unsigned long apic_physbase;
++ uint32_t reg;
++ /* IN or OUT */
++ uint32_t value;
++};
++typedef struct physdev_apic physdev_apic_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_apic_t);
++
++/*
++ * Allocate or free a physical upcall vector for the specified IRQ line.
++ * @arg == pointer to physdev_irq structure.
++ */
++#define PHYSDEVOP_alloc_irq_vector 10
++#define PHYSDEVOP_free_irq_vector 11
++struct physdev_irq {
++ /* IN */
++ uint32_t irq;
++ /* IN or OUT */
++ uint32_t vector;
++};
++typedef struct physdev_irq physdev_irq_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_irq_t);
++
++#define MAP_PIRQ_TYPE_MSI 0x0
++#define MAP_PIRQ_TYPE_GSI 0x1
++#define MAP_PIRQ_TYPE_UNKNOWN 0x2
++
++#define PHYSDEVOP_map_pirq 13
++struct physdev_map_pirq {
++ domid_t domid;
++ /* IN */
++ int type;
++ /* IN */
++ int index;
++ /* IN or OUT */
++ int pirq;
++ /* IN */
++ int bus;
++ /* IN */
++ int devfn;
++ /* IN */
++ int entry_nr;
++ /* IN */
++ uint64_t table_base;
++};
++typedef struct physdev_map_pirq physdev_map_pirq_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_map_pirq_t);
++
++#define PHYSDEVOP_unmap_pirq 14
++struct physdev_unmap_pirq {
++ domid_t domid;
++ /* IN */
++ int pirq;
++};
++
++typedef struct physdev_unmap_pirq physdev_unmap_pirq_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_unmap_pirq_t);
++
++#define PHYSDEVOP_manage_pci_add 15
++#define PHYSDEVOP_manage_pci_remove 16
++struct physdev_manage_pci {
++ /* IN */
++ uint8_t bus;
++ uint8_t devfn;
++};
++
++typedef struct physdev_manage_pci physdev_manage_pci_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_manage_pci_t);
++
++/*
++ * Argument to physdev_op_compat() hypercall. Superceded by new physdev_op()
++ * hypercall since 0x00030202.
++ */
++struct physdev_op {
++ uint32_t cmd;
++ union {
++ struct physdev_irq_status_query irq_status_query;
++ struct physdev_set_iopl set_iopl;
++ struct physdev_set_iobitmap set_iobitmap;
++ struct physdev_apic apic_op;
++ struct physdev_irq irq_op;
++ } u;
++};
++typedef struct physdev_op physdev_op_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_op_t);
++
++/*
++ * Notify that some PIRQ-bound event channels have been unmasked.
++ * ** This command is obsolete since interface version 0x00030202 and is **
++ * ** unsupported by newer versions of Xen. **
++ */
++#define PHYSDEVOP_IRQ_UNMASK_NOTIFY 4
++
++/*
++ * These all-capitals physdev operation names are superceded by the new names
++ * (defined above) since interface version 0x00030202.
++ */
++#define PHYSDEVOP_IRQ_STATUS_QUERY PHYSDEVOP_irq_status_query
++#define PHYSDEVOP_SET_IOPL PHYSDEVOP_set_iopl
++#define PHYSDEVOP_SET_IOBITMAP PHYSDEVOP_set_iobitmap
++#define PHYSDEVOP_APIC_READ PHYSDEVOP_apic_read
++#define PHYSDEVOP_APIC_WRITE PHYSDEVOP_apic_write
++#define PHYSDEVOP_ASSIGN_VECTOR PHYSDEVOP_alloc_irq_vector
++#define PHYSDEVOP_FREE_VECTOR PHYSDEVOP_free_irq_vector
++#define PHYSDEVOP_IRQ_NEEDS_UNMASK_NOTIFY XENIRQSTAT_needs_eoi
++#define PHYSDEVOP_IRQ_SHARED XENIRQSTAT_shared
++
++#endif /* __XEN_PUBLIC_PHYSDEV_H__ */
+
+/*
+ * Local variables:
@@ -102328,15 +141540,15 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/io/xenbus.h tmp-linux-2.6-xen.patch/include/xen/interface/io/xenbus.h
---- pristine-linux-2.6.18.2/include/xen/interface/io/xenbus.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/io/xenbus.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,73 @@
-+/*****************************************************************************
-+ * xenbus.h
-+ *
-+ * Xenbus protocol details.
-+ *
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/platform.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/platform.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,346 @@
++/******************************************************************************
++ * platform.h
++ *
++ * Hardware platform operations. Intended for use by domain-0 kernel.
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -102355,46 +141567,319 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (C) 2005 XenSource Ltd.
++ * Copyright (c) 2002-2006, K Fraser
+ */
+
-+#ifndef _XEN_PUBLIC_IO_XENBUS_H
-+#define _XEN_PUBLIC_IO_XENBUS_H
++#ifndef __XEN_PUBLIC_PLATFORM_H__
++#define __XEN_PUBLIC_PLATFORM_H__
++
++#include "xen.h"
++
++#define XENPF_INTERFACE_VERSION 0x03000001
+
+/*
-+ * The state of either end of the Xenbus, i.e. the current communication
-+ * status of initialisation across the bus. States here imply nothing about
-+ * the state of the connection between the driver and the kernel's device
-+ * layers.
++ * Set clock such that it would read <secs,nsecs> after 00:00:00 UTC,
++ * 1 January, 1970 if the current system time was <system_time>.
+ */
-+enum xenbus_state {
-+ XenbusStateUnknown = 0,
++#define XENPF_settime 17
++struct xenpf_settime {
++ /* IN variables. */
++ uint32_t secs;
++ uint32_t nsecs;
++ uint64_t system_time;
++};
++typedef struct xenpf_settime xenpf_settime_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_settime_t);
+
-+ XenbusStateInitialising = 1,
++/*
++ * Request memory range (@mfn, @mfn+ at nr_mfns-1) to have type @type.
++ * On x86, @type is an architecture-defined MTRR memory type.
++ * On success, returns the MTRR that was used (@reg) and a handle that can
++ * be passed to XENPF_DEL_MEMTYPE to accurately tear down the new setting.
++ * (x86-specific).
++ */
++#define XENPF_add_memtype 31
++struct xenpf_add_memtype {
++ /* IN variables. */
++ xen_pfn_t mfn;
++ uint64_t nr_mfns;
++ uint32_t type;
++ /* OUT variables. */
++ uint32_t handle;
++ uint32_t reg;
++};
++typedef struct xenpf_add_memtype xenpf_add_memtype_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_add_memtype_t);
+
-+ /*
-+ * InitWait: Finished early initialisation but waiting for information
-+ * from the peer or hotplug scripts.
-+ */
-+ XenbusStateInitWait = 2,
++/*
++ * Tear down an existing memory-range type. If @handle is remembered then it
++ * should be passed in to accurately tear down the correct setting (in case
++ * of overlapping memory regions with differing types). If it is not known
++ * then @handle should be set to zero. In all cases @reg must be set.
++ * (x86-specific).
++ */
++#define XENPF_del_memtype 32
++struct xenpf_del_memtype {
++ /* IN variables. */
++ uint32_t handle;
++ uint32_t reg;
++};
++typedef struct xenpf_del_memtype xenpf_del_memtype_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_del_memtype_t);
+
-+ /*
-+ * Initialised: Waiting for a connection from the peer.
-+ */
-+ XenbusStateInitialised = 3,
++/* Read current type of an MTRR (x86-specific). */
++#define XENPF_read_memtype 33
++struct xenpf_read_memtype {
++ /* IN variables. */
++ uint32_t reg;
++ /* OUT variables. */
++ xen_pfn_t mfn;
++ uint64_t nr_mfns;
++ uint32_t type;
++};
++typedef struct xenpf_read_memtype xenpf_read_memtype_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_read_memtype_t);
++
++#define XENPF_microcode_update 35
++struct xenpf_microcode_update {
++ /* IN variables. */
++ XEN_GUEST_HANDLE(const_void) data;/* Pointer to microcode data */
++ uint32_t length; /* Length of microcode data. */
++};
++typedef struct xenpf_microcode_update xenpf_microcode_update_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_microcode_update_t);
++
++#define XENPF_platform_quirk 39
++#define QUIRK_NOIRQBALANCING 1 /* Do not restrict IO-APIC RTE targets */
++#define QUIRK_IOAPIC_BAD_REGSEL 2 /* IO-APIC REGSEL forgets its value */
++#define QUIRK_IOAPIC_GOOD_REGSEL 3 /* IO-APIC REGSEL behaves properly */
++struct xenpf_platform_quirk {
++ /* IN variables. */
++ uint32_t quirk_id;
++};
++typedef struct xenpf_platform_quirk xenpf_platform_quirk_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_platform_quirk_t);
++
++#define XENPF_firmware_info 50
++#define XEN_FW_DISK_INFO 1 /* from int 13 AH=08/41/48 */
++#define XEN_FW_DISK_MBR_SIGNATURE 2 /* from MBR offset 0x1b8 */
++#define XEN_FW_VBEDDC_INFO 3 /* from int 10 AX=4f15 */
++struct xenpf_firmware_info {
++ /* IN variables. */
++ uint32_t type;
++ uint32_t index;
++ /* OUT variables. */
++ union {
++ struct {
++ /* Int13, Fn48: Check Extensions Present. */
++ uint8_t device; /* %dl: bios device number */
++ uint8_t version; /* %ah: major version */
++ uint16_t interface_support; /* %cx: support bitmap */
++ /* Int13, Fn08: Legacy Get Device Parameters. */
++ uint16_t legacy_max_cylinder; /* %cl[7:6]:%ch: max cyl # */
++ uint8_t legacy_max_head; /* %dh: max head # */
++ uint8_t legacy_sectors_per_track; /* %cl[5:0]: max sector # */
++ /* Int13, Fn41: Get Device Parameters (as filled into %ds:%esi). */
++ /* NB. First uint16_t of buffer must be set to buffer size. */
++ XEN_GUEST_HANDLE(void) edd_params;
++ } disk_info; /* XEN_FW_DISK_INFO */
++ struct {
++ uint8_t device; /* bios device number */
++ uint32_t mbr_signature; /* offset 0x1b8 in mbr */
++ } disk_mbr_signature; /* XEN_FW_DISK_MBR_SIGNATURE */
++ struct {
++ /* Int10, AX=4F15: Get EDID info. */
++ uint8_t capabilities;
++ uint8_t edid_transfer_time;
++ /* must refer to 128-byte buffer */
++ XEN_GUEST_HANDLE(uint8) edid;
++ } vbeddc_info; /* XEN_FW_VBEDDC_INFO */
++ } u;
++};
++typedef struct xenpf_firmware_info xenpf_firmware_info_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_firmware_info_t);
++
++#define XENPF_enter_acpi_sleep 51
++struct xenpf_enter_acpi_sleep {
++ /* IN variables */
++ uint16_t pm1a_cnt_val; /* PM1a control value. */
++ uint16_t pm1b_cnt_val; /* PM1b control value. */
++ uint32_t sleep_state; /* Which state to enter (Sn). */
++ uint32_t flags; /* Must be zero. */
++};
++typedef struct xenpf_enter_acpi_sleep xenpf_enter_acpi_sleep_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_enter_acpi_sleep_t);
++
++#define XENPF_change_freq 52
++struct xenpf_change_freq {
++ /* IN variables */
++ uint32_t flags; /* Must be zero. */
++ uint32_t cpu; /* Physical cpu. */
++ uint64_t freq; /* New frequency (Hz). */
++};
++typedef struct xenpf_change_freq xenpf_change_freq_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_change_freq_t);
++
++/*
++ * Get idle times (nanoseconds since boot) for physical CPUs specified in the
++ * @cpumap_bitmap with range [0.. at cpumap_nr_cpus-1]. The @idletime array is
++ * indexed by CPU number; only entries with the corresponding @cpumap_bitmap
++ * bit set are written to. On return, @cpumap_bitmap is modified so that any
++ * non-existent CPUs are cleared. Such CPUs have their @idletime array entry
++ * cleared.
++ */
++#define XENPF_getidletime 53
++struct xenpf_getidletime {
++ /* IN/OUT variables */
++ /* IN: CPUs to interrogate; OUT: subset of IN which are present */
++ XEN_GUEST_HANDLE(uint8) cpumap_bitmap;
++ /* IN variables */
++ /* Size of cpumap bitmap. */
++ uint32_t cpumap_nr_cpus;
++ /* Must be indexable for every cpu in cpumap_bitmap. */
++ XEN_GUEST_HANDLE(uint64) idletime;
++ /* OUT variables */
++ /* System time when the idletime snapshots were taken. */
++ uint64_t now;
++};
++typedef struct xenpf_getidletime xenpf_getidletime_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_getidletime_t);
+
-+ XenbusStateConnected = 4,
++#define XENPF_set_processor_pminfo 54
+
-+ /*
-+ * Closing: The device is being closed due to an error or an unplug event.
-+ */
-+ XenbusStateClosing = 5,
++/* ability bits */
++#define XEN_PROCESSOR_PM_CX 1
++#define XEN_PROCESSOR_PM_PX 2
++#define XEN_PROCESSOR_PM_TX 4
++
++/* cmd type */
++#define XEN_PM_CX 0
++#define XEN_PM_PX 1
++#define XEN_PM_TX 2
++
++/* Px sub info type */
++#define XEN_PX_PCT 1
++#define XEN_PX_PSS 2
++#define XEN_PX_PPC 4
++#define XEN_PX_PSD 8
++
++struct xen_power_register {
++ uint32_t space_id;
++ uint32_t bit_width;
++ uint32_t bit_offset;
++ uint32_t access_size;
++ uint64_t address;
++};
++
++struct xen_processor_csd {
++ uint32_t domain; /* domain number of one dependent group */
++ uint32_t coord_type; /* coordination type */
++ uint32_t num; /* number of processors in same domain */
++};
++typedef struct xen_processor_csd xen_processor_csd_t;
++DEFINE_XEN_GUEST_HANDLE(xen_processor_csd_t);
++
++struct xen_processor_cx {
++ struct xen_power_register reg; /* GAS for Cx trigger register */
++ uint8_t type; /* cstate value, c0: 0, c1: 1, ... */
++ uint32_t latency; /* worst latency (ms) to enter/exit this cstate */
++ uint32_t power; /* average power consumption(mW) */
++ uint32_t dpcnt; /* number of dependency entries */
++ XEN_GUEST_HANDLE(xen_processor_csd_t) dp; /* NULL if no dependency */
++};
++typedef struct xen_processor_cx xen_processor_cx_t;
++DEFINE_XEN_GUEST_HANDLE(xen_processor_cx_t);
++
++struct xen_processor_flags {
++ uint32_t bm_control:1;
++ uint32_t bm_check:1;
++ uint32_t has_cst:1;
++ uint32_t power_setup_done:1;
++ uint32_t bm_rld_set:1;
++};
++
++struct xen_processor_power {
++ uint32_t count; /* number of C state entries in array below */
++ struct xen_processor_flags flags; /* global flags of this processor */
++ XEN_GUEST_HANDLE(xen_processor_cx_t) states; /* supported c states */
++};
++
++struct xen_pct_register {
++ uint8_t descriptor;
++ uint16_t length;
++ uint8_t space_id;
++ uint8_t bit_width;
++ uint8_t bit_offset;
++ uint8_t reserved;
++ uint64_t address;
++};
++
++struct xen_processor_px {
++ uint64_t core_frequency; /* megahertz */
++ uint64_t power; /* milliWatts */
++ uint64_t transition_latency; /* microseconds */
++ uint64_t bus_master_latency; /* microseconds */
++ uint64_t control; /* control value */
++ uint64_t status; /* success indicator */
++};
++typedef struct xen_processor_px xen_processor_px_t;
++DEFINE_XEN_GUEST_HANDLE(xen_processor_px_t);
++
++struct xen_psd_package {
++ uint64_t num_entries;
++ uint64_t revision;
++ uint64_t domain;
++ uint64_t coord_type;
++ uint64_t num_processors;
++};
++
++struct xen_processor_performance {
++ uint32_t flags; /* flag for Px sub info type */
++ uint32_t ppc; /* Platform limitation on freq usage */
++ struct xen_pct_register control_register;
++ struct xen_pct_register status_register;
++ uint32_t state_count; /* total available performance states */
++ XEN_GUEST_HANDLE(xen_processor_px_t) states;
++ struct xen_psd_package domain_info;
++ uint32_t shared_type; /* coordination type of this processor */
++};
++typedef struct xen_processor_performance xen_processor_performance_t;
++DEFINE_XEN_GUEST_HANDLE(xen_processor_performance_t);
+
-+ XenbusStateClosed = 6
++struct xenpf_set_processor_pminfo {
++ /* IN variables */
++ uint32_t id; /* ACPI CPU ID */
++ uint32_t type; /* {XEN_PM_CX, XEN_PM_PX} */
++ union {
++ struct xen_processor_power power;/* Cx: _CST/_CSD */
++ struct xen_processor_performance perf; /* Px: _PPC/_PCT/_PSS/_PSD */
++ };
+};
-+typedef enum xenbus_state XenbusState;
++typedef struct xenpf_set_processor_pminfo xenpf_set_processor_pminfo_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_set_processor_pminfo_t);
+
-+#endif /* _XEN_PUBLIC_IO_XENBUS_H */
++struct xen_platform_op {
++ uint32_t cmd;
++ uint32_t interface_version; /* XENPF_INTERFACE_VERSION */
++ union {
++ struct xenpf_settime settime;
++ struct xenpf_add_memtype add_memtype;
++ struct xenpf_del_memtype del_memtype;
++ struct xenpf_read_memtype read_memtype;
++ struct xenpf_microcode_update microcode;
++ struct xenpf_platform_quirk platform_quirk;
++ struct xenpf_firmware_info firmware_info;
++ struct xenpf_enter_acpi_sleep enter_acpi_sleep;
++ struct xenpf_change_freq change_freq;
++ struct xenpf_getidletime getidletime;
++ struct xenpf_set_processor_pminfo set_pminfo;
++ uint8_t pad[128];
++ } u;
++};
++typedef struct xen_platform_op xen_platform_op_t;
++DEFINE_XEN_GUEST_HANDLE(xen_platform_op_t);
++
++#endif /* __XEN_PUBLIC_PLATFORM_H__ */
+
+/*
+ * Local variables:
@@ -102405,14 +141890,15 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/io/xs_wire.h tmp-linux-2.6-xen.patch/include/xen/interface/io/xs_wire.h
---- pristine-linux-2.6.18.2/include/xen/interface/io/xs_wire.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/io/xs_wire.h 2007-09-30 18:06:19.000000000 +0200
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/sched.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/sched.h Wed Sep 10 10:54:08 2008 +0100
@@ -0,0 +1,121 @@
-+/*
-+ * Details of the "wire" protocol between Xen Store Daemon and client
-+ * library or guest kernel.
-+ *
++/******************************************************************************
++ * sched.h
++ *
++ * Scheduler state interactions
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -102431,236 +141917,94 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (C) 2005 Rusty Russell IBM Corporation
++ * Copyright (c) 2005, Keir Fraser <keir at xensource.com>
+ */
+
-+#ifndef _XS_WIRE_H
-+#define _XS_WIRE_H
-+
-+enum xsd_sockmsg_type
-+{
-+ XS_DEBUG,
-+ XS_DIRECTORY,
-+ XS_READ,
-+ XS_GET_PERMS,
-+ XS_WATCH,
-+ XS_UNWATCH,
-+ XS_TRANSACTION_START,
-+ XS_TRANSACTION_END,
-+ XS_INTRODUCE,
-+ XS_RELEASE,
-+ XS_GET_DOMAIN_PATH,
-+ XS_WRITE,
-+ XS_MKDIR,
-+ XS_RM,
-+ XS_SET_PERMS,
-+ XS_WATCH_EVENT,
-+ XS_ERROR,
-+ XS_IS_DOMAIN_INTRODUCED,
-+ XS_RESUME
-+};
-+
-+#define XS_WRITE_NONE "NONE"
-+#define XS_WRITE_CREATE "CREATE"
-+#define XS_WRITE_CREATE_EXCL "CREATE|EXCL"
-+
-+/* We hand errors as strings, for portability. */
-+struct xsd_errors
-+{
-+ int errnum;
-+ const char *errstring;
-+};
-+#define XSD_ERROR(x) { x, #x }
-+static struct xsd_errors xsd_errors[]
-+#if defined(__GNUC__)
-+__attribute__((unused))
-+#endif
-+ = {
-+ XSD_ERROR(EINVAL),
-+ XSD_ERROR(EACCES),
-+ XSD_ERROR(EEXIST),
-+ XSD_ERROR(EISDIR),
-+ XSD_ERROR(ENOENT),
-+ XSD_ERROR(ENOMEM),
-+ XSD_ERROR(ENOSPC),
-+ XSD_ERROR(EIO),
-+ XSD_ERROR(ENOTEMPTY),
-+ XSD_ERROR(ENOSYS),
-+ XSD_ERROR(EROFS),
-+ XSD_ERROR(EBUSY),
-+ XSD_ERROR(EAGAIN),
-+ XSD_ERROR(EISCONN)
-+};
-+
-+struct xsd_sockmsg
-+{
-+ uint32_t type; /* XS_??? */
-+ uint32_t req_id;/* Request identifier, echoed in daemon's response. */
-+ uint32_t tx_id; /* Transaction id (0 if not related to a transaction). */
-+ uint32_t len; /* Length of data following this. */
-+
-+ /* Generally followed by nul-terminated string(s). */
-+};
-+
-+enum xs_watch_type
-+{
-+ XS_WATCH_PATH = 0,
-+ XS_WATCH_TOKEN
-+};
-+
-+/* Inter-domain shared memory communications. */
-+#define XENSTORE_RING_SIZE 1024
-+typedef uint32_t XENSTORE_RING_IDX;
-+#define MASK_XENSTORE_IDX(idx) ((idx) & (XENSTORE_RING_SIZE-1))
-+struct xenstore_domain_interface {
-+ char req[XENSTORE_RING_SIZE]; /* Requests to xenstore daemon. */
-+ char rsp[XENSTORE_RING_SIZE]; /* Replies and async watch events. */
-+ XENSTORE_RING_IDX req_cons, req_prod;
-+ XENSTORE_RING_IDX rsp_cons, rsp_prod;
-+};
++#ifndef __XEN_PUBLIC_SCHED_H__
++#define __XEN_PUBLIC_SCHED_H__
+
-+#endif /* _XS_WIRE_H */
++#include "event_channel.h"
+
+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
-+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/kexec.h tmp-linux-2.6-xen.patch/include/xen/interface/kexec.h
---- pristine-linux-2.6.18.2/include/xen/interface/kexec.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/kexec.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,137 @@
-+/******************************************************************************
-+ * kexec.h - Public portion
++ * The prototype for this hypercall is:
++ * long sched_op(int cmd, void *arg)
++ * @cmd == SCHEDOP_??? (scheduler operation).
++ * @arg == Operation-specific extra argument(s), as described below.
+ *
-+ * Xen port written by:
-+ * - Simon 'Horms' Horman <horms at verge.net.au>
-+ * - Magnus Damm <magnus at valinux.co.jp>
-+ */
-+
-+#ifndef _XEN_PUBLIC_KEXEC_H
-+#define _XEN_PUBLIC_KEXEC_H
-+
-+
-+/* This file describes the Kexec / Kdump hypercall interface for Xen.
-+ *
-+ * Kexec under vanilla Linux allows a user to reboot the physical machine
-+ * into a new user-specified kernel. The Xen port extends this idea
-+ * to allow rebooting of the machine from dom0. When kexec for dom0
-+ * is used to reboot, both the hypervisor and the domains get replaced
-+ * with some other kernel. It is possible to kexec between vanilla
-+ * Linux and Xen and back again. Xen to Xen works well too.
-+ *
-+ * The hypercall interface for kexec can be divided into three main
-+ * types of hypercall operations:
-+ *
-+ * 1) Range information:
-+ * This is used by the dom0 kernel to ask the hypervisor about various
-+ * address information. This information is needed to allow kexec-tools
-+ * to fill in the ELF headers for /proc/vmcore properly.
-+ *
-+ * 2) Load and unload of images:
-+ * There are no big surprises here, the kexec binary from kexec-tools
-+ * runs in userspace in dom0. The tool loads/unloads data into the
-+ * dom0 kernel such as new kernel, initramfs and hypervisor. When
-+ * loaded the dom0 kernel performs a load hypercall operation, and
-+ * before releasing all page references the dom0 kernel calls unload.
-+ *
-+ * 3) Kexec operation:
-+ * This is used to start a previously loaded kernel.
++ * Versions of Xen prior to 3.0.2 provided only the following legacy version
++ * of this hypercall, supporting only the commands yield, block and shutdown:
++ * long sched_op(int cmd, unsigned long arg)
++ * @cmd == SCHEDOP_??? (scheduler operation).
++ * @arg == 0 (SCHEDOP_yield and SCHEDOP_block)
++ * == SHUTDOWN_* code (SCHEDOP_shutdown)
++ * This legacy version is available to new guests as sched_op_compat().
+ */
+
-+#include "xen.h"
-+
-+#if defined(__i386__) || defined(__x86_64__)
-+#define KEXEC_XEN_NO_PAGES 17
-+#endif
-+
+/*
-+ * Prototype for this hypercall is:
-+ * int kexec_op(int cmd, void *args)
-+ * @cmd == KEXEC_CMD_...
-+ * KEXEC operation to perform
-+ * @args == Operation-specific extra arguments (NULL if none).
++ * Voluntarily yield the CPU.
++ * @arg == NULL.
+ */
++#define SCHEDOP_yield 0
+
+/*
-+ * Kexec supports two types of operation:
-+ * - kexec into a regular kernel, very similar to a standard reboot
-+ * - KEXEC_TYPE_DEFAULT is used to specify this type
-+ * - kexec into a special "crash kernel", aka kexec-on-panic
-+ * - KEXEC_TYPE_CRASH is used to specify this type
-+ * - parts of our system may be broken at kexec-on-panic time
-+ * - the code should be kept as simple and self-contained as possible
++ * Block execution of this VCPU until an event is received for processing.
++ * If called with event upcalls masked, this operation will atomically
++ * reenable event delivery and check for pending events before blocking the
++ * VCPU. This avoids a "wakeup waiting" race.
++ * @arg == NULL.
+ */
++#define SCHEDOP_block 1
+
-+#define KEXEC_TYPE_DEFAULT 0
-+#define KEXEC_TYPE_CRASH 1
-+
-+
-+/* The kexec implementation for Xen allows the user to load two
-+ * types of kernels, KEXEC_TYPE_DEFAULT and KEXEC_TYPE_CRASH.
-+ * All data needed for a kexec reboot is kept in one xen_kexec_image_t
-+ * per "instance". The data mainly consists of machine address lists to pages
-+ * together with destination addresses. The data in xen_kexec_image_t
-+ * is passed to the "code page" which is one page of code that performs
-+ * the final relocations before jumping to the new kernel.
++/*
++ * Halt execution of this domain (all VCPUs) and notify the system controller.
++ * @arg == pointer to sched_shutdown structure.
+ */
-+
-+typedef struct xen_kexec_image {
-+#if defined(__i386__) || defined(__x86_64__)
-+ unsigned long page_list[KEXEC_XEN_NO_PAGES];
-+#endif
-+ unsigned long indirection_page;
-+ unsigned long start_address;
-+} xen_kexec_image_t;
++#define SCHEDOP_shutdown 2
++struct sched_shutdown {
++ unsigned int reason; /* SHUTDOWN_* */
++};
++typedef struct sched_shutdown sched_shutdown_t;
++DEFINE_XEN_GUEST_HANDLE(sched_shutdown_t);
+
+/*
-+ * Perform kexec having previously loaded a kexec or kdump kernel
-+ * as appropriate.
-+ * type == KEXEC_TYPE_DEFAULT or KEXEC_TYPE_CRASH [in]
++ * Poll a set of event-channel ports. Return when one or more are pending. An
++ * optional timeout may be specified.
++ * @arg == pointer to sched_poll structure.
+ */
-+#define KEXEC_CMD_kexec 0
-+typedef struct xen_kexec_exec {
-+ int type;
-+} xen_kexec_exec_t;
++#define SCHEDOP_poll 3
++struct sched_poll {
++ XEN_GUEST_HANDLE(evtchn_port_t) ports;
++ unsigned int nr_ports;
++ uint64_t timeout;
++};
++typedef struct sched_poll sched_poll_t;
++DEFINE_XEN_GUEST_HANDLE(sched_poll_t);
+
+/*
-+ * Load/Unload kernel image for kexec or kdump.
-+ * type == KEXEC_TYPE_DEFAULT or KEXEC_TYPE_CRASH [in]
-+ * image == relocation information for kexec (ignored for unload) [in]
++ * Declare a shutdown for another domain. The main use of this function is
++ * in interpreting shutdown requests and reasons for fully-virtualized
++ * domains. A para-virtualized domain may use SCHEDOP_shutdown directly.
++ * @arg == pointer to sched_remote_shutdown structure.
+ */
-+#define KEXEC_CMD_kexec_load 1
-+#define KEXEC_CMD_kexec_unload 2
-+typedef struct xen_kexec_load {
-+ int type;
-+ xen_kexec_image_t image;
-+} xen_kexec_load_t;
-+
-+#define KEXEC_RANGE_MA_CRASH 0 /* machine address and size of crash area */
-+#define KEXEC_RANGE_MA_XEN 1 /* machine address and size of Xen itself */
-+#define KEXEC_RANGE_MA_CPU 2 /* machine address and size of a CPU note */
++#define SCHEDOP_remote_shutdown 4
++struct sched_remote_shutdown {
++ domid_t domain_id; /* Remote domain ID */
++ unsigned int reason; /* SHUTDOWN_xxx reason */
++};
++typedef struct sched_remote_shutdown sched_remote_shutdown_t;
++DEFINE_XEN_GUEST_HANDLE(sched_remote_shutdown_t);
+
+/*
-+ * Find the address and size of certain memory areas
-+ * range == KEXEC_RANGE_... [in]
-+ * nr == physical CPU number (starting from 0) if KEXEC_RANGE_MA_CPU [in]
-+ * size == number of bytes reserved in window [out]
-+ * start == address of the first byte in the window [out]
++ * Reason codes for SCHEDOP_shutdown. These may be interpreted by control
++ * software to determine the appropriate action. For the most part, Xen does
++ * not care about the shutdown code.
+ */
-+#define KEXEC_CMD_kexec_get_range 3
-+typedef struct xen_kexec_range {
-+ int range;
-+ int nr;
-+ unsigned long size;
-+ unsigned long start;
-+} xen_kexec_range_t;
++#define SHUTDOWN_poweroff 0 /* Domain exited normally. Clean up and kill. */
++#define SHUTDOWN_reboot 1 /* Clean up, kill, and then restart. */
++#define SHUTDOWN_suspend 2 /* Clean up, save suspend info, kill. */
++#define SHUTDOWN_crash 3 /* Tell controller we've crashed. */
+
-+#endif /* _XEN_PUBLIC_KEXEC_H */
++#endif /* __XEN_PUBLIC_SCHED_H__ */
+
+/*
+ * Local variables:
@@ -102671,259 +142015,489 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/libelf.h tmp-linux-2.6-xen.patch/include/xen/interface/libelf.h
---- pristine-linux-2.6.18.2/include/xen/interface/libelf.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/libelf.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,241 @@
-+#ifndef __XC_LIBELF__
-+#define __XC_LIBELF__ 1
-+
-+#if defined(__i386__) || defined(__x86_64) || defined(__ia64__)
-+#define XEN_ELF_LITTLE_ENDIAN
-+#elif defined(__powerpc__)
-+#define XEN_ELF_BIG_ENDIAN
-+#else
-+#error define architectural endianness
-+#endif
-+
-+#undef ELFSIZE
-+#include "elfnote.h"
-+#include "elfstructs.h"
-+#include "features.h"
-+
-+/* ------------------------------------------------------------------------ */
-+
-+typedef union {
-+ Elf32_Ehdr e32;
-+ Elf64_Ehdr e64;
-+} elf_ehdr;
-+
-+typedef union {
-+ Elf32_Phdr e32;
-+ Elf64_Phdr e64;
-+} elf_phdr;
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/sysctl.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/sysctl.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,295 @@
++/******************************************************************************
++ * sysctl.h
++ *
++ * System management operations. For use by node control stack.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2002-2006, K Fraser
++ */
+
-+typedef union {
-+ Elf32_Shdr e32;
-+ Elf64_Shdr e64;
-+} elf_shdr;
++#ifndef __XEN_PUBLIC_SYSCTL_H__
++#define __XEN_PUBLIC_SYSCTL_H__
+
-+typedef union {
-+ Elf32_Sym e32;
-+ Elf64_Sym e64;
-+} elf_sym;
++#if !defined(__XEN__) && !defined(__XEN_TOOLS__)
++#error "sysctl operations are intended for use by node control tools only"
++#endif
+
-+typedef union {
-+ Elf32_Rel e32;
-+ Elf64_Rel e64;
-+} elf_rel;
++#include "xen.h"
++#include "domctl.h"
+
-+typedef union {
-+ Elf32_Rela e32;
-+ Elf64_Rela e64;
-+} elf_rela;
++#define XEN_SYSCTL_INTERFACE_VERSION 0x00000006
+
-+typedef union {
-+ Elf32_Note e32;
-+ Elf64_Note e64;
-+} elf_note;
++/*
++ * Read console content from Xen buffer ring.
++ */
++#define XEN_SYSCTL_readconsole 1
++struct xen_sysctl_readconsole {
++ /* IN: Non-zero -> clear after reading. */
++ uint8_t clear;
++ /* IN: Non-zero -> start index specified by @index field. */
++ uint8_t incremental;
++ uint8_t pad0, pad1;
++ /*
++ * IN: Start index for consuming from ring buffer (if @incremental);
++ * OUT: End index after consuming from ring buffer.
++ */
++ uint32_t index;
++ /* IN: Virtual address to write console data. */
++ XEN_GUEST_HANDLE_64(char) buffer;
++ /* IN: Size of buffer; OUT: Bytes written to buffer. */
++ uint32_t count;
++};
++typedef struct xen_sysctl_readconsole xen_sysctl_readconsole_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_readconsole_t);
+
-+struct elf_binary {
-+ /* elf binary */
-+ const char *image;
-+ size_t size;
-+ char class;
-+ char data;
++/* Get trace buffers machine base address */
++#define XEN_SYSCTL_tbuf_op 2
++struct xen_sysctl_tbuf_op {
++ /* IN variables */
++#define XEN_SYSCTL_TBUFOP_get_info 0
++#define XEN_SYSCTL_TBUFOP_set_cpu_mask 1
++#define XEN_SYSCTL_TBUFOP_set_evt_mask 2
++#define XEN_SYSCTL_TBUFOP_set_size 3
++#define XEN_SYSCTL_TBUFOP_enable 4
++#define XEN_SYSCTL_TBUFOP_disable 5
++ uint32_t cmd;
++ /* IN/OUT variables */
++ struct xenctl_cpumap cpu_mask;
++ uint32_t evt_mask;
++ /* OUT variables */
++ uint64_aligned_t buffer_mfn;
++ uint32_t size;
++};
++typedef struct xen_sysctl_tbuf_op xen_sysctl_tbuf_op_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_tbuf_op_t);
+
-+ const elf_ehdr *ehdr;
-+ const char *sec_strtab;
-+ const elf_shdr *sym_tab;
-+ const char *sym_strtab;
++/*
++ * Get physical information about the host machine
++ */
++#define XEN_SYSCTL_physinfo 3
++ /* (x86) The platform supports HVM guests. */
++#define _XEN_SYSCTL_PHYSCAP_hvm 0
++#define XEN_SYSCTL_PHYSCAP_hvm (1u<<_XEN_SYSCTL_PHYSCAP_hvm)
++ /* (x86) The platform supports HVM-guest direct access to I/O devices. */
++#define _XEN_SYSCTL_PHYSCAP_hvm_directio 1
++#define XEN_SYSCTL_PHYSCAP_hvm_directio (1u<<_XEN_SYSCTL_PHYSCAP_hvm_directio)
++struct xen_sysctl_physinfo {
++ uint32_t threads_per_core;
++ uint32_t cores_per_socket;
++ uint32_t nr_cpus;
++ uint32_t nr_nodes;
++ uint32_t cpu_khz;
++ uint64_aligned_t total_pages;
++ uint64_aligned_t free_pages;
++ uint64_aligned_t scrub_pages;
++ uint32_t hw_cap[8];
+
-+ /* loaded to */
-+ char *dest;
-+ uint64_t pstart;
-+ uint64_t pend;
-+ uint64_t reloc_offset;
++ /*
++ * IN: maximum addressable entry in the caller-provided cpu_to_node array.
++ * OUT: largest cpu identifier in the system.
++ * If OUT is greater than IN then the cpu_to_node array is truncated!
++ */
++ uint32_t max_cpu_id;
++ /*
++ * If not NULL, this array is filled with node identifier for each cpu.
++ * If a cpu has no node information (e.g., cpu not present) then the
++ * sentinel value ~0u is written.
++ * The size of this array is specified by the caller in @max_cpu_id.
++ * If the actual @max_cpu_id is smaller than the array then the trailing
++ * elements of the array will not be written by the sysctl.
++ */
++ XEN_GUEST_HANDLE_64(uint32) cpu_to_node;
+
-+#ifndef __XEN__
-+ /* misc */
-+ FILE *log;
-+#endif
-+ int verbose;
++ /* XEN_SYSCTL_PHYSCAP_??? */
++ uint32_t capabilities;
+};
++typedef struct xen_sysctl_physinfo xen_sysctl_physinfo_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_physinfo_t);
+
-+/* ------------------------------------------------------------------------ */
-+/* accessing elf header fields */
-+
-+#ifdef XEN_ELF_BIG_ENDIAN
-+# define NATIVE_ELFDATA ELFDATA2MSB
-+#else
-+# define NATIVE_ELFDATA ELFDATA2LSB
-+#endif
-+
-+#define elf_32bit(elf) (ELFCLASS32 == (elf)->class)
-+#define elf_64bit(elf) (ELFCLASS64 == (elf)->class)
-+#define elf_msb(elf) (ELFDATA2MSB == (elf)->data)
-+#define elf_lsb(elf) (ELFDATA2LSB == (elf)->data)
-+#define elf_swap(elf) (NATIVE_ELFDATA != (elf)->data)
++/*
++ * Get the ID of the current scheduler.
++ */
++#define XEN_SYSCTL_sched_id 4
++struct xen_sysctl_sched_id {
++ /* OUT variable */
++ uint32_t sched_id;
++};
++typedef struct xen_sysctl_sched_id xen_sysctl_sched_id_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_sched_id_t);
+
-+#define elf_uval(elf, str, elem) \
-+ ((ELFCLASS64 == (elf)->class) \
-+ ? elf_access_unsigned((elf), (str), \
-+ offsetof(typeof(*(str)),e64.elem), \
-+ sizeof((str)->e64.elem)) \
-+ : elf_access_unsigned((elf), (str), \
-+ offsetof(typeof(*(str)),e32.elem), \
-+ sizeof((str)->e32.elem)))
-+
-+#define elf_sval(elf, str, elem) \
-+ ((ELFCLASS64 == (elf)->class) \
-+ ? elf_access_signed((elf), (str), \
-+ offsetof(typeof(*(str)),e64.elem), \
-+ sizeof((str)->e64.elem)) \
-+ : elf_access_signed((elf), (str), \
-+ offsetof(typeof(*(str)),e32.elem), \
-+ sizeof((str)->e32.elem)))
-+
-+#define elf_size(elf, str) \
-+ ((ELFCLASS64 == (elf)->class) \
-+ ? sizeof((str)->e64) \
-+ : sizeof((str)->e32))
++/* Interface for controlling Xen software performance counters. */
++#define XEN_SYSCTL_perfc_op 5
++/* Sub-operations: */
++#define XEN_SYSCTL_PERFCOP_reset 1 /* Reset all counters to zero. */
++#define XEN_SYSCTL_PERFCOP_query 2 /* Get perfctr information. */
++struct xen_sysctl_perfc_desc {
++ char name[80]; /* name of perf counter */
++ uint32_t nr_vals; /* number of values for this counter */
++};
++typedef struct xen_sysctl_perfc_desc xen_sysctl_perfc_desc_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_perfc_desc_t);
++typedef uint32_t xen_sysctl_perfc_val_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_perfc_val_t);
+
-+uint64_t elf_access_unsigned(struct elf_binary *elf, const void *ptr,
-+ uint64_t offset, size_t size);
-+int64_t elf_access_signed(struct elf_binary *elf, const void *ptr,
-+ uint64_t offset, size_t size);
++struct xen_sysctl_perfc_op {
++ /* IN variables. */
++ uint32_t cmd; /* XEN_SYSCTL_PERFCOP_??? */
++ /* OUT variables. */
++ uint32_t nr_counters; /* number of counters description */
++ uint32_t nr_vals; /* number of values */
++ /* counter information (or NULL) */
++ XEN_GUEST_HANDLE_64(xen_sysctl_perfc_desc_t) desc;
++ /* counter values (or NULL) */
++ XEN_GUEST_HANDLE_64(xen_sysctl_perfc_val_t) val;
++};
++typedef struct xen_sysctl_perfc_op xen_sysctl_perfc_op_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_perfc_op_t);
+
-+uint64_t elf_round_up(struct elf_binary *elf, uint64_t addr);
++#define XEN_SYSCTL_getdomaininfolist 6
++struct xen_sysctl_getdomaininfolist {
++ /* IN variables. */
++ domid_t first_domain;
++ uint32_t max_domains;
++ XEN_GUEST_HANDLE_64(xen_domctl_getdomaininfo_t) buffer;
++ /* OUT variables. */
++ uint32_t num_domains;
++};
++typedef struct xen_sysctl_getdomaininfolist xen_sysctl_getdomaininfolist_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_getdomaininfolist_t);
+
-+/* ------------------------------------------------------------------------ */
-+/* xc_libelf_tools.c */
++/* Inject debug keys into Xen. */
++#define XEN_SYSCTL_debug_keys 7
++struct xen_sysctl_debug_keys {
++ /* IN variables. */
++ XEN_GUEST_HANDLE_64(char) keys;
++ uint32_t nr_keys;
++};
++typedef struct xen_sysctl_debug_keys xen_sysctl_debug_keys_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_debug_keys_t);
+
-+int elf_shdr_count(struct elf_binary *elf);
-+int elf_phdr_count(struct elf_binary *elf);
++/* Get physical CPU information. */
++#define XEN_SYSCTL_getcpuinfo 8
++struct xen_sysctl_cpuinfo {
++ uint64_aligned_t idletime;
++};
++typedef struct xen_sysctl_cpuinfo xen_sysctl_cpuinfo_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_cpuinfo_t);
++struct xen_sysctl_getcpuinfo {
++ /* IN variables. */
++ uint32_t max_cpus;
++ XEN_GUEST_HANDLE_64(xen_sysctl_cpuinfo_t) info;
++ /* OUT variables. */
++ uint32_t nr_cpus;
++};
++typedef struct xen_sysctl_getcpuinfo xen_sysctl_getcpuinfo_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_getcpuinfo_t);
+
-+const elf_shdr *elf_shdr_by_name(struct elf_binary *elf, const char *name);
-+const elf_shdr *elf_shdr_by_index(struct elf_binary *elf, int index);
-+const elf_phdr *elf_phdr_by_index(struct elf_binary *elf, int index);
++#define XEN_SYSCTL_availheap 9
++struct xen_sysctl_availheap {
++ /* IN variables. */
++ uint32_t min_bitwidth; /* Smallest address width (zero if don't care). */
++ uint32_t max_bitwidth; /* Largest address width (zero if don't care). */
++ int32_t node; /* NUMA node of interest (-1 for all nodes). */
++ /* OUT variables. */
++ uint64_aligned_t avail_bytes;/* Bytes available in the specified region. */
++};
++typedef struct xen_sysctl_availheap xen_sysctl_availheap_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_availheap_t);
+
-+const char *elf_section_name(struct elf_binary *elf, const elf_shdr * shdr);
-+const void *elf_section_start(struct elf_binary *elf, const elf_shdr * shdr);
-+const void *elf_section_end(struct elf_binary *elf, const elf_shdr * shdr);
++#define XEN_SYSCTL_get_pmstat 10
++struct pm_px_val {
++ uint64_aligned_t freq; /* Px core frequency */
++ uint64_aligned_t residency; /* Px residency time */
++ uint64_aligned_t count; /* Px transition count */
++};
++typedef struct pm_px_val pm_px_val_t;
++DEFINE_XEN_GUEST_HANDLE(pm_px_val_t);
++
++struct pm_px_stat {
++ uint8_t total; /* total Px states */
++ uint8_t usable; /* usable Px states */
++ uint8_t last; /* last Px state */
++ uint8_t cur; /* current Px state */
++ XEN_GUEST_HANDLE_64(uint64) trans_pt; /* Px transition table */
++ XEN_GUEST_HANDLE_64(pm_px_val_t) pt;
++};
++typedef struct pm_px_stat pm_px_stat_t;
++DEFINE_XEN_GUEST_HANDLE(pm_px_stat_t);
++
++struct pm_cx_stat {
++ uint32_t nr; /* entry nr in triggers & residencies, including C0 */
++ uint32_t last; /* last Cx state */
++ uint64_aligned_t idle_time; /* idle time from boot */
++ XEN_GUEST_HANDLE_64(uint64) triggers; /* Cx trigger counts */
++ XEN_GUEST_HANDLE_64(uint64) residencies; /* Cx residencies */
++};
++
++struct xen_sysctl_get_pmstat {
++#define PMSTAT_CATEGORY_MASK 0xf0
++#define PMSTAT_PX 0x10
++#define PMSTAT_CX 0x20
++#define PMSTAT_get_max_px (PMSTAT_PX | 0x1)
++#define PMSTAT_get_pxstat (PMSTAT_PX | 0x2)
++#define PMSTAT_reset_pxstat (PMSTAT_PX | 0x3)
++#define PMSTAT_get_max_cx (PMSTAT_CX | 0x1)
++#define PMSTAT_get_cxstat (PMSTAT_CX | 0x2)
++#define PMSTAT_reset_cxstat (PMSTAT_CX | 0x3)
++ uint32_t type;
++ uint32_t cpuid;
++ union {
++ struct pm_px_stat getpx;
++ struct pm_cx_stat getcx;
++ /* other struct for tx, etc */
++ } u;
++};
++typedef struct xen_sysctl_get_pmstat xen_sysctl_get_pmstat_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_get_pmstat_t);
+
-+const void *elf_segment_start(struct elf_binary *elf, const elf_phdr * phdr);
-+const void *elf_segment_end(struct elf_binary *elf, const elf_phdr * phdr);
++struct xen_sysctl {
++ uint32_t cmd;
++ uint32_t interface_version; /* XEN_SYSCTL_INTERFACE_VERSION */
++ union {
++ struct xen_sysctl_readconsole readconsole;
++ struct xen_sysctl_tbuf_op tbuf_op;
++ struct xen_sysctl_physinfo physinfo;
++ struct xen_sysctl_sched_id sched_id;
++ struct xen_sysctl_perfc_op perfc_op;
++ struct xen_sysctl_getdomaininfolist getdomaininfolist;
++ struct xen_sysctl_debug_keys debug_keys;
++ struct xen_sysctl_getcpuinfo getcpuinfo;
++ struct xen_sysctl_availheap availheap;
++ struct xen_sysctl_get_pmstat get_pmstat;
++ uint8_t pad[128];
++ } u;
++};
++typedef struct xen_sysctl xen_sysctl_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_t);
+
-+const elf_sym *elf_sym_by_name(struct elf_binary *elf, const char *symbol);
-+const elf_sym *elf_sym_by_index(struct elf_binary *elf, int index);
++#endif /* __XEN_PUBLIC_SYSCTL_H__ */
+
-+const char *elf_note_name(struct elf_binary *elf, const elf_note * note);
-+const void *elf_note_desc(struct elf_binary *elf, const elf_note * note);
-+uint64_t elf_note_numeric(struct elf_binary *elf, const elf_note * note);
-+const elf_note *elf_note_next(struct elf_binary *elf, const elf_note * note);
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/trace.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/trace.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,172 @@
++/******************************************************************************
++ * include/public/trace.h
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Mark Williamson, (C) 2004 Intel Research Cambridge
++ * Copyright (C) 2005 Bin Ren
++ */
+
-+int elf_is_elfbinary(const void *image);
-+int elf_phdr_is_loadable(struct elf_binary *elf, const elf_phdr * phdr);
++#ifndef __XEN_PUBLIC_TRACE_H__
++#define __XEN_PUBLIC_TRACE_H__
+
-+/* ------------------------------------------------------------------------ */
-+/* xc_libelf_loader.c */
++#define TRACE_EXTRA_MAX 7
++#define TRACE_EXTRA_SHIFT 28
+
-+int elf_init(struct elf_binary *elf, const char *image, size_t size);
-+#ifdef __XEN__
-+void elf_set_verbose(struct elf_binary *elf);
-+#else
-+void elf_set_logfile(struct elf_binary *elf, FILE * log, int verbose);
-+#endif
++/* Trace classes */
++#define TRC_CLS_SHIFT 16
++#define TRC_GEN 0x0001f000 /* General trace */
++#define TRC_SCHED 0x0002f000 /* Xen Scheduler trace */
++#define TRC_DOM0OP 0x0004f000 /* Xen DOM0 operation trace */
++#define TRC_HVM 0x0008f000 /* Xen HVM trace */
++#define TRC_MEM 0x0010f000 /* Xen memory trace */
++#define TRC_PV 0x0020f000 /* Xen PV traces */
++#define TRC_ALL 0x0ffff000
++#define TRC_HD_TO_EVENT(x) ((x)&0x0fffffff)
++#define TRC_HD_CYCLE_FLAG (1UL<<31)
++#define TRC_HD_INCLUDES_CYCLE_COUNT(x) ( !!( (x) & TRC_HD_CYCLE_FLAG ) )
++#define TRC_HD_EXTRA(x) (((x)>>TRACE_EXTRA_SHIFT)&TRACE_EXTRA_MAX)
+
-+void elf_parse_binary(struct elf_binary *elf);
-+void elf_load_binary(struct elf_binary *elf);
++/* Trace subclasses */
++#define TRC_SUBCLS_SHIFT 12
+
-+void *elf_get_ptr(struct elf_binary *elf, unsigned long addr);
-+uint64_t elf_lookup_addr(struct elf_binary *elf, const char *symbol);
++/* trace subclasses for SVM */
++#define TRC_HVM_ENTRYEXIT 0x00081000 /* VMENTRY and #VMEXIT */
++#define TRC_HVM_HANDLER 0x00082000 /* various HVM handlers */
+
-+/* ------------------------------------------------------------------------ */
-+/* xc_libelf_relocate.c */
++/* Trace events per class */
++#define TRC_LOST_RECORDS (TRC_GEN + 1)
++#define TRC_TRACE_WRAP_BUFFER (TRC_GEN + 2)
++#define TRC_TRACE_CPU_CHANGE (TRC_GEN + 3)
+
-+int elf_reloc(struct elf_binary *elf);
++#define TRC_SCHED_DOM_ADD (TRC_SCHED + 1)
++#define TRC_SCHED_DOM_REM (TRC_SCHED + 2)
++#define TRC_SCHED_SLEEP (TRC_SCHED + 3)
++#define TRC_SCHED_WAKE (TRC_SCHED + 4)
++#define TRC_SCHED_YIELD (TRC_SCHED + 5)
++#define TRC_SCHED_BLOCK (TRC_SCHED + 6)
++#define TRC_SCHED_SHUTDOWN (TRC_SCHED + 7)
++#define TRC_SCHED_CTL (TRC_SCHED + 8)
++#define TRC_SCHED_ADJDOM (TRC_SCHED + 9)
++#define TRC_SCHED_SWITCH (TRC_SCHED + 10)
++#define TRC_SCHED_S_TIMER_FN (TRC_SCHED + 11)
++#define TRC_SCHED_T_TIMER_FN (TRC_SCHED + 12)
++#define TRC_SCHED_DOM_TIMER_FN (TRC_SCHED + 13)
++#define TRC_SCHED_SWITCH_INFPREV (TRC_SCHED + 14)
++#define TRC_SCHED_SWITCH_INFNEXT (TRC_SCHED + 15)
+
-+/* ------------------------------------------------------------------------ */
-+/* xc_libelf_dominfo.c */
++#define TRC_MEM_PAGE_GRANT_MAP (TRC_MEM + 1)
++#define TRC_MEM_PAGE_GRANT_UNMAP (TRC_MEM + 2)
++#define TRC_MEM_PAGE_GRANT_TRANSFER (TRC_MEM + 3)
+
-+#define UNSET_ADDR ((uint64_t)-1)
++#define TRC_PV_HYPERCALL (TRC_PV + 1)
++#define TRC_PV_TRAP (TRC_PV + 3)
++#define TRC_PV_PAGE_FAULT (TRC_PV + 4)
++#define TRC_PV_FORCED_INVALID_OP (TRC_PV + 5)
++#define TRC_PV_EMULATE_PRIVOP (TRC_PV + 6)
++#define TRC_PV_EMULATE_4GB (TRC_PV + 7)
++#define TRC_PV_MATH_STATE_RESTORE (TRC_PV + 8)
++#define TRC_PV_PAGING_FIXUP (TRC_PV + 9)
++#define TRC_PV_GDT_LDT_MAPPING_FAULT (TRC_PV + 10)
++#define TRC_PV_PTWR_EMULATION (TRC_PV + 11)
++#define TRC_PV_PTWR_EMULATION_PAE (TRC_PV + 12)
++ /* Indicates that addresses in trace record are 64 bits */
++#define TRC_64_FLAG (0x100)
+
-+enum xen_elfnote_type {
-+ XEN_ENT_NONE = 0,
-+ XEN_ENT_LONG = 1,
-+ XEN_ENT_STR = 2
-+};
++/* trace events per subclass */
++#define TRC_HVM_VMENTRY (TRC_HVM_ENTRYEXIT + 0x01)
++#define TRC_HVM_VMEXIT (TRC_HVM_ENTRYEXIT + 0x02)
++#define TRC_HVM_VMEXIT64 (TRC_HVM_ENTRYEXIT + TRC_64_FLAG + 0x02)
++#define TRC_HVM_PF_XEN (TRC_HVM_HANDLER + 0x01)
++#define TRC_HVM_PF_XEN64 (TRC_HVM_HANDLER + TRC_64_FLAG + 0x01)
++#define TRC_HVM_PF_INJECT (TRC_HVM_HANDLER + 0x02)
++#define TRC_HVM_PF_INJECT64 (TRC_HVM_HANDLER + TRC_64_FLAG + 0x02)
++#define TRC_HVM_INJ_EXC (TRC_HVM_HANDLER + 0x03)
++#define TRC_HVM_INJ_VIRQ (TRC_HVM_HANDLER + 0x04)
++#define TRC_HVM_REINJ_VIRQ (TRC_HVM_HANDLER + 0x05)
++#define TRC_HVM_IO_READ (TRC_HVM_HANDLER + 0x06)
++#define TRC_HVM_IO_WRITE (TRC_HVM_HANDLER + 0x07)
++#define TRC_HVM_CR_READ (TRC_HVM_HANDLER + 0x08)
++#define TRC_HVM_CR_READ64 (TRC_HVM_HANDLER + TRC_64_FLAG + 0x08)
++#define TRC_HVM_CR_WRITE (TRC_HVM_HANDLER + 0x09)
++#define TRC_HVM_CR_WRITE64 (TRC_HVM_HANDLER + TRC_64_FLAG + 0x09)
++#define TRC_HVM_DR_READ (TRC_HVM_HANDLER + 0x0A)
++#define TRC_HVM_DR_WRITE (TRC_HVM_HANDLER + 0x0B)
++#define TRC_HVM_MSR_READ (TRC_HVM_HANDLER + 0x0C)
++#define TRC_HVM_MSR_WRITE (TRC_HVM_HANDLER + 0x0D)
++#define TRC_HVM_CPUID (TRC_HVM_HANDLER + 0x0E)
++#define TRC_HVM_INTR (TRC_HVM_HANDLER + 0x0F)
++#define TRC_HVM_NMI (TRC_HVM_HANDLER + 0x10)
++#define TRC_HVM_SMI (TRC_HVM_HANDLER + 0x11)
++#define TRC_HVM_VMMCALL (TRC_HVM_HANDLER + 0x12)
++#define TRC_HVM_HLT (TRC_HVM_HANDLER + 0x13)
++#define TRC_HVM_INVLPG (TRC_HVM_HANDLER + 0x14)
++#define TRC_HVM_INVLPG64 (TRC_HVM_HANDLER + TRC_64_FLAG + 0x14)
++#define TRC_HVM_MCE (TRC_HVM_HANDLER + 0x15)
++#define TRC_HVM_IO_ASSIST (TRC_HVM_HANDLER + 0x16)
++#define TRC_HVM_MMIO_ASSIST (TRC_HVM_HANDLER + 0x17)
++#define TRC_HVM_CLTS (TRC_HVM_HANDLER + 0x18)
++#define TRC_HVM_LMSW (TRC_HVM_HANDLER + 0x19)
++#define TRC_HVM_LMSW64 (TRC_HVM_HANDLER + TRC_64_FLAG + 0x19)
+
-+struct xen_elfnote {
-+ enum xen_elfnote_type type;
-+ const char *name;
++/* This structure represents a single trace buffer record. */
++struct t_rec {
++ uint32_t event:28;
++ uint32_t extra_u32:3; /* # entries in trailing extra_u32[] array */
++ uint32_t cycles_included:1; /* u.cycles or u.no_cycles? */
+ union {
-+ const char *str;
-+ uint64_t num;
-+ } data;
++ struct {
++ uint32_t cycles_lo, cycles_hi; /* cycle counter timestamp */
++ uint32_t extra_u32[7]; /* event data items */
++ } cycles;
++ struct {
++ uint32_t extra_u32[7]; /* event data items */
++ } nocycles;
++ } u;
+};
+
-+struct elf_dom_parms {
-+ /* raw */
-+ const char *guest_info;
-+ const void *elf_note_start;
-+ const void *elf_note_end;
-+ struct xen_elfnote elf_notes[XEN_ELFNOTE_MAX + 1];
-+
-+ /* parsed */
-+ char guest_os[16];
-+ char guest_ver[16];
-+ char xen_ver[16];
-+ char loader[16];
-+ int pae;
-+ int bsd_symtab;
-+ uint64_t virt_base;
-+ uint64_t virt_entry;
-+ uint64_t virt_hypercall;
-+ uint64_t virt_hv_start_low;
-+ uint64_t elf_paddr_offset;
-+ uint32_t f_supported[XENFEAT_NR_SUBMAPS];
-+ uint32_t f_required[XENFEAT_NR_SUBMAPS];
-+
-+ /* calculated */
-+ uint64_t virt_offset;
-+ uint64_t virt_kstart;
-+ uint64_t virt_kend;
++/*
++ * This structure contains the metadata for a single trace buffer. The head
++ * field, indexes into an array of struct t_rec's.
++ */
++struct t_buf {
++ /* Assume the data buffer size is X. X is generally not a power of 2.
++ * CONS and PROD are incremented modulo (2*X):
++ * 0 <= cons < 2*X
++ * 0 <= prod < 2*X
++ * This is done because addition modulo X breaks at 2^32 when X is not a
++ * power of 2:
++ * (((2^32 - 1) % X) + 1) % X != (2^32) % X
++ */
++ uint32_t cons; /* Offset of next item to be consumed by control tools. */
++ uint32_t prod; /* Offset of next item to be produced by Xen. */
++ /* Records follow immediately after the meta-data header. */
+};
+
-+static inline void elf_xen_feature_set(int nr, uint32_t * addr)
-+{
-+ addr[nr >> 5] |= 1 << (nr & 31);
-+}
-+static inline int elf_xen_feature_get(int nr, uint32_t * addr)
-+{
-+ return !!(addr[nr >> 5] & (1 << (nr & 31)));
-+}
-+
-+int elf_xen_parse_features(const char *features,
-+ uint32_t *supported,
-+ uint32_t *required);
-+int elf_xen_parse_note(struct elf_binary *elf,
-+ struct elf_dom_parms *parms,
-+ const elf_note *note);
-+int elf_xen_parse_guest_info(struct elf_binary *elf,
-+ struct elf_dom_parms *parms);
-+int elf_xen_parse(struct elf_binary *elf,
-+ struct elf_dom_parms *parms);
++#endif /* __XEN_PUBLIC_TRACE_H__ */
+
-+#endif /* __XC_LIBELF__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/memory.h tmp-linux-2.6-xen.patch/include/xen/interface/memory.h
---- pristine-linux-2.6.18.2/include/xen/interface/memory.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/memory.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,281 @@
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/vcpu.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/vcpu.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,213 @@
+/******************************************************************************
-+ * memory.h
++ * vcpu.h
+ *
-+ * Memory reservation and information.
++ * VCPU initialisation, query, and hotplug.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -102946,251 +142520,183 @@
+ * Copyright (c) 2005, Keir Fraser <keir at xensource.com>
+ */
+
-+#ifndef __XEN_PUBLIC_MEMORY_H__
-+#define __XEN_PUBLIC_MEMORY_H__
++#ifndef __XEN_PUBLIC_VCPU_H__
++#define __XEN_PUBLIC_VCPU_H__
+
+/*
-+ * Increase or decrease the specified domain's memory reservation. Returns the
-+ * number of extents successfully allocated or freed.
-+ * arg == addr of struct xen_memory_reservation.
++ * Prototype for this hypercall is:
++ * int vcpu_op(int cmd, int vcpuid, void *extra_args)
++ * @cmd == VCPUOP_??? (VCPU operation).
++ * @vcpuid == VCPU to operate on.
++ * @extra_args == Operation-specific extra arguments (NULL if none).
+ */
-+#define XENMEM_increase_reservation 0
-+#define XENMEM_decrease_reservation 1
-+#define XENMEM_populate_physmap 6
-+struct xen_memory_reservation {
-+
-+ /*
-+ * XENMEM_increase_reservation:
-+ * OUT: MFN (*not* GMFN) bases of extents that were allocated
-+ * XENMEM_decrease_reservation:
-+ * IN: GMFN bases of extents to free
-+ * XENMEM_populate_physmap:
-+ * IN: GPFN bases of extents to populate with memory
-+ * OUT: GMFN bases of extents that were allocated
-+ * (NB. This command also updates the mach_to_phys translation table)
-+ */
-+ XEN_GUEST_HANDLE(xen_pfn_t) extent_start;
-+
-+ /* Number of extents, and size/alignment of each (2^extent_order pages). */
-+ xen_ulong_t nr_extents;
-+ unsigned int extent_order;
-+
-+ /*
-+ * Maximum # bits addressable by the user of the allocated region (e.g.,
-+ * I/O devices often have a 32-bit limitation even in 64-bit systems). If
-+ * zero then the user has no addressing restriction.
-+ * This field is not used by XENMEM_decrease_reservation.
-+ */
-+ unsigned int address_bits;
-+
-+ /*
-+ * Domain whose reservation is being changed.
-+ * Unprivileged domains can specify only DOMID_SELF.
-+ */
-+ domid_t domid;
-+};
-+typedef struct xen_memory_reservation xen_memory_reservation_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_memory_reservation_t);
+
+/*
-+ * An atomic exchange of memory pages. If return code is zero then
-+ * @out.extent_list provides GMFNs of the newly-allocated memory.
-+ * Returns zero on complete success, otherwise a negative error code.
-+ * On complete success then always @nr_exchanged == @in.nr_extents.
-+ * On partial success @nr_exchanged indicates how much work was done.
++ * Initialise a VCPU. Each VCPU can be initialised only once. A
++ * newly-initialised VCPU will not run until it is brought up by VCPUOP_up.
++ *
++ * @extra_arg == pointer to vcpu_guest_context structure containing initial
++ * state for the VCPU.
+ */
-+#define XENMEM_exchange 11
-+struct xen_memory_exchange {
-+ /*
-+ * [IN] Details of memory extents to be exchanged (GMFN bases).
-+ * Note that @in.address_bits is ignored and unused.
-+ */
-+ struct xen_memory_reservation in;
-+
-+ /*
-+ * [IN/OUT] Details of new memory extents.
-+ * We require that:
-+ * 1. @in.domid == @out.domid
-+ * 2. @in.nr_extents << @in.extent_order ==
-+ * @out.nr_extents << @out.extent_order
-+ * 3. @in.extent_start and @out.extent_start lists must not overlap
-+ * 4. @out.extent_start lists GPFN bases to be populated
-+ * 5. @out.extent_start is overwritten with allocated GMFN bases
-+ */
-+ struct xen_memory_reservation out;
-+
-+ /*
-+ * [OUT] Number of input extents that were successfully exchanged:
-+ * 1. The first @nr_exchanged input extents were successfully
-+ * deallocated.
-+ * 2. The corresponding first entries in the output extent list correctly
-+ * indicate the GMFNs that were successfully exchanged.
-+ * 3. All other input and output extents are untouched.
-+ * 4. If not all input exents are exchanged then the return code of this
-+ * command will be non-zero.
-+ * 5. THIS FIELD MUST BE INITIALISED TO ZERO BY THE CALLER!
-+ */
-+ xen_ulong_t nr_exchanged;
-+};
-+typedef struct xen_memory_exchange xen_memory_exchange_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_memory_exchange_t);
++#define VCPUOP_initialise 0
+
+/*
-+ * Returns the maximum machine frame number of mapped RAM in this system.
-+ * This command always succeeds (it never returns an error code).
-+ * arg == NULL.
++ * Bring up a VCPU. This makes the VCPU runnable. This operation will fail
++ * if the VCPU has not been initialised (VCPUOP_initialise).
+ */
-+#define XENMEM_maximum_ram_page 2
++#define VCPUOP_up 1
+
+/*
-+ * Returns the current or maximum memory reservation, in pages, of the
-+ * specified domain (may be DOMID_SELF). Returns -ve errcode on failure.
-+ * arg == addr of domid_t.
++ * Bring down a VCPU (i.e., make it non-runnable).
++ * There are a few caveats that callers should observe:
++ * 1. This operation may return, and VCPU_is_up may return false, before the
++ * VCPU stops running (i.e., the command is asynchronous). It is a good
++ * idea to ensure that the VCPU has entered a non-critical loop before
++ * bringing it down. Alternatively, this operation is guaranteed
++ * synchronous if invoked by the VCPU itself.
++ * 2. After a VCPU is initialised, there is currently no way to drop all its
++ * references to domain memory. Even a VCPU that is down still holds
++ * memory references via its pagetable base pointer and GDT. It is good
++ * practise to move a VCPU onto an 'idle' or default page table, LDT and
++ * GDT before bringing it down.
+ */
-+#define XENMEM_current_reservation 3
-+#define XENMEM_maximum_reservation 4
++#define VCPUOP_down 2
+
-+/*
-+ * Returns the maximum GPFN in use by the guest, or -ve errcode on failure.
-+ */
-+#define XENMEM_maximum_gpfn 14
++/* Returns 1 if the given VCPU is up. */
++#define VCPUOP_is_up 3
+
+/*
-+ * Returns a list of MFN bases of 2MB extents comprising the machine_to_phys
-+ * mapping table. Architectures which do not have a m2p table do not implement
-+ * this command.
-+ * arg == addr of xen_machphys_mfn_list_t.
++ * Return information about the state and running time of a VCPU.
++ * @extra_arg == pointer to vcpu_runstate_info structure.
+ */
-+#define XENMEM_machphys_mfn_list 5
-+struct xen_machphys_mfn_list {
++#define VCPUOP_get_runstate_info 4
++struct vcpu_runstate_info {
++ /* VCPU's current state (RUNSTATE_*). */
++ int state;
++ /* When was current state entered (system time, ns)? */
++ uint64_t state_entry_time;
+ /*
-+ * Size of the 'extent_start' array. Fewer entries will be filled if the
-+ * machphys table is smaller than max_extents * 2MB.
++ * Time spent in each RUNSTATE_* (ns). The sum of these times is
++ * guaranteed not to drift from system time.
+ */
-+ unsigned int max_extents;
++ uint64_t time[4];
++};
++typedef struct vcpu_runstate_info vcpu_runstate_info_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_runstate_info_t);
+
-+ /*
-+ * Pointer to buffer to fill with list of extent starts. If there are
-+ * any large discontiguities in the machine address space, 2MB gaps in
-+ * the machphys table will be represented by an MFN base of zero.
-+ */
-+ XEN_GUEST_HANDLE(xen_pfn_t) extent_start;
++/* VCPU is currently running on a physical CPU. */
++#define RUNSTATE_running 0
+
-+ /*
-+ * Number of extents written to the above array. This will be smaller
-+ * than 'max_extents' if the machphys table is smaller than max_e * 2MB.
-+ */
-+ unsigned int nr_extents;
-+};
-+typedef struct xen_machphys_mfn_list xen_machphys_mfn_list_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_machphys_mfn_list_t);
++/* VCPU is runnable, but not currently scheduled on any physical CPU. */
++#define RUNSTATE_runnable 1
+
-+/*
-+ * Returns the location in virtual address space of the machine_to_phys
-+ * mapping table. Architectures which do not have a m2p table, or which do not
-+ * map it by default into guest address space, do not implement this command.
-+ * arg == addr of xen_machphys_mapping_t.
-+ */
-+#define XENMEM_machphys_mapping 12
-+struct xen_machphys_mapping {
-+ xen_ulong_t v_start, v_end; /* Start and end virtual addresses. */
-+ xen_ulong_t max_mfn; /* Maximum MFN that can be looked up. */
-+};
-+typedef struct xen_machphys_mapping xen_machphys_mapping_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_machphys_mapping_t);
++/* VCPU is blocked (a.k.a. idle). It is therefore not runnable. */
++#define RUNSTATE_blocked 2
+
+/*
-+ * Sets the GPFN at which a particular page appears in the specified guest's
-+ * pseudophysical address space.
-+ * arg == addr of xen_add_to_physmap_t.
++ * VCPU is not runnable, but it is not blocked.
++ * This is a 'catch all' state for things like hotplug and pauses by the
++ * system administrator (or for critical sections in the hypervisor).
++ * RUNSTATE_blocked dominates this state (it is the preferred state).
+ */
-+#define XENMEM_add_to_physmap 7
-+struct xen_add_to_physmap {
-+ /* Which domain to change the mapping for. */
-+ domid_t domid;
-+
-+ /* Source mapping space. */
-+#define XENMAPSPACE_shared_info 0 /* shared info page */
-+#define XENMAPSPACE_grant_table 1 /* grant table page */
-+ unsigned int space;
-+
-+ /* Index into source mapping space. */
-+ xen_ulong_t idx;
-+
-+ /* GPFN where the source mapping page should appear. */
-+ xen_pfn_t gpfn;
-+};
-+typedef struct xen_add_to_physmap xen_add_to_physmap_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_add_to_physmap_t);
++#define RUNSTATE_offline 3
+
+/*
-+ * Translates a list of domain-specific GPFNs into MFNs. Returns a -ve error
-+ * code on failure. This call only works for auto-translated guests.
++ * Register a shared memory area from which the guest may obtain its own
++ * runstate information without needing to execute a hypercall.
++ * Notes:
++ * 1. The registered address may be virtual or physical or guest handle,
++ * depending on the platform. Virtual address or guest handle should be
++ * registered on x86 systems.
++ * 2. Only one shared area may be registered per VCPU. The shared area is
++ * updated by the hypervisor each time the VCPU is scheduled. Thus
++ * runstate.state will always be RUNSTATE_running and
++ * runstate.state_entry_time will indicate the system time at which the
++ * VCPU was last scheduled to run.
++ * @extra_arg == pointer to vcpu_register_runstate_memory_area structure.
+ */
-+#define XENMEM_translate_gpfn_list 8
-+struct xen_translate_gpfn_list {
-+ /* Which domain to translate for? */
-+ domid_t domid;
-+
-+ /* Length of list. */
-+ xen_ulong_t nr_gpfns;
-+
-+ /* List of GPFNs to translate. */
-+ XEN_GUEST_HANDLE(xen_pfn_t) gpfn_list;
-+
-+ /*
-+ * Output list to contain MFN translations. May be the same as the input
-+ * list (in which case each input GPFN is overwritten with the output MFN).
-+ */
-+ XEN_GUEST_HANDLE(xen_pfn_t) mfn_list;
++#define VCPUOP_register_runstate_memory_area 5
++struct vcpu_register_runstate_memory_area {
++ union {
++ XEN_GUEST_HANDLE(vcpu_runstate_info_t) h;
++ struct vcpu_runstate_info *v;
++ uint64_t p;
++ } addr;
+};
-+typedef struct xen_translate_gpfn_list xen_translate_gpfn_list_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_translate_gpfn_list_t);
++typedef struct vcpu_register_runstate_memory_area vcpu_register_runstate_memory_area_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_register_runstate_memory_area_t);
+
+/*
-+ * Returns the pseudo-physical memory map as it was when the domain
-+ * was started (specified by XENMEM_set_memory_map).
-+ * arg == addr of xen_memory_map_t.
-+ */
-+#define XENMEM_memory_map 9
-+struct xen_memory_map {
-+ /*
-+ * On call the number of entries which can be stored in buffer. On
-+ * return the number of entries which have been stored in
-+ * buffer.
-+ */
-+ unsigned int nr_entries;
-+
-+ /*
-+ * Entries in the buffer are in the same format as returned by the
-+ * BIOS INT 0x15 EAX=0xE820 call.
-+ */
-+ XEN_GUEST_HANDLE(void) buffer;
++ * Set or stop a VCPU's periodic timer. Every VCPU has one periodic timer
++ * which can be set via these commands. Periods smaller than one millisecond
++ * may not be supported.
++ */
++#define VCPUOP_set_periodic_timer 6 /* arg == vcpu_set_periodic_timer_t */
++#define VCPUOP_stop_periodic_timer 7 /* arg == NULL */
++struct vcpu_set_periodic_timer {
++ uint64_t period_ns;
+};
-+typedef struct xen_memory_map xen_memory_map_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_memory_map_t);
++typedef struct vcpu_set_periodic_timer vcpu_set_periodic_timer_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_set_periodic_timer_t);
+
+/*
-+ * Returns the real physical memory map. Passes the same structure as
-+ * XENMEM_memory_map.
-+ * arg == addr of xen_memory_map_t.
++ * Set or stop a VCPU's single-shot timer. Every VCPU has one single-shot
++ * timer which can be set via these commands.
+ */
-+#define XENMEM_machine_memory_map 10
++#define VCPUOP_set_singleshot_timer 8 /* arg == vcpu_set_singleshot_timer_t */
++#define VCPUOP_stop_singleshot_timer 9 /* arg == NULL */
++struct vcpu_set_singleshot_timer {
++ uint64_t timeout_abs_ns; /* Absolute system time value in nanoseconds. */
++ uint32_t flags; /* VCPU_SSHOTTMR_??? */
++};
++typedef struct vcpu_set_singleshot_timer vcpu_set_singleshot_timer_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_set_singleshot_timer_t);
+
-+/*
-+ * Set the pseudo-physical memory map of a domain, as returned by
-+ * XENMEM_memory_map.
-+ * arg == addr of xen_foreign_memory_map_t.
++/* Flags to VCPUOP_set_singleshot_timer. */
++ /* Require the timeout to be in the future (return -ETIME if it's passed). */
++#define _VCPU_SSHOTTMR_future (0)
++#define VCPU_SSHOTTMR_future (1U << _VCPU_SSHOTTMR_future)
++
++/*
++ * Register a memory location in the guest address space for the
++ * vcpu_info structure. This allows the guest to place the vcpu_info
++ * structure in a convenient place, such as in a per-cpu data area.
++ * The pointer need not be page aligned, but the structure must not
++ * cross a page boundary.
++ *
++ * This may be called only once per vcpu.
+ */
-+#define XENMEM_set_memory_map 13
-+struct xen_foreign_memory_map {
-+ domid_t domid;
-+ struct xen_memory_map map;
++#define VCPUOP_register_vcpu_info 10 /* arg == vcpu_register_vcpu_info_t */
++struct vcpu_register_vcpu_info {
++ uint64_t mfn; /* mfn of page to place vcpu_info */
++ uint32_t offset; /* offset within page */
++ uint32_t rsvd; /* unused */
+};
-+typedef struct xen_foreign_memory_map xen_foreign_memory_map_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_foreign_memory_map_t);
++typedef struct vcpu_register_vcpu_info vcpu_register_vcpu_info_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_register_vcpu_info_t);
+
-+#endif /* __XEN_PUBLIC_MEMORY_H__ */
++/* Send an NMI to the specified VCPU. @extra_arg == NULL. */
++#define VCPUOP_send_nmi 11
++
++/*
++ * Get the physical ID information for a pinned vcpu's underlying physical
++ * processor. The physical ID informmation is architecture-specific.
++ * On x86: id[31:0]=apic_id, id[63:32]=acpi_id, and all values 0xff and
++ * greater are reserved.
++ * This command returns -EINVAL if it is not a valid operation for this VCPU.
++ */
++#define VCPUOP_get_physid 12 /* arg == vcpu_get_physid_t */
++struct vcpu_get_physid {
++ uint64_t phys_id;
++};
++typedef struct vcpu_get_physid vcpu_get_physid_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_get_physid_t);
++#define xen_vcpu_physid_to_x86_apicid(physid) \
++ ((((uint32_t)(physid)) >= 0xff) ? 0xff : ((uint8_t)(physid)))
++#define xen_vcpu_physid_to_x86_acpiid(physid) \
++ ((((uint32_t)((physid)>>32)) >= 0xff) ? 0xff : ((uint8_t)((physid)>>32)))
++
++#endif /* __XEN_PUBLIC_VCPU_H__ */
+
+/*
+ * Local variables:
@@ -103201,14 +142707,14 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/nmi.h tmp-linux-2.6-xen.patch/include/xen/interface/nmi.h
---- pristine-linux-2.6.18.2/include/xen/interface/nmi.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/nmi.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,78 @@
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/version.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/version.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,91 @@
+/******************************************************************************
-+ * nmi.h
++ * version.h
+ *
-+ * NMI callback registration and reason codes.
++ * Xen version, type, and compile information.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -103228,51 +142734,64 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
++ * Copyright (c) 2005, Nguyen Anh Quynh <aquynh at gmail.com>
+ * Copyright (c) 2005, Keir Fraser <keir at xensource.com>
+ */
+
-+#ifndef __XEN_PUBLIC_NMI_H__
-+#define __XEN_PUBLIC_NMI_H__
++#ifndef __XEN_PUBLIC_VERSION_H__
++#define __XEN_PUBLIC_VERSION_H__
+
-+/*
-+ * NMI reason codes:
-+ * Currently these are x86-specific, stored in arch_shared_info.nmi_reason.
-+ */
-+ /* I/O-check error reported via ISA port 0x61, bit 6. */
-+#define _XEN_NMIREASON_io_error 0
-+#define XEN_NMIREASON_io_error (1UL << _XEN_NMIREASON_io_error)
-+ /* Parity error reported via ISA port 0x61, bit 7. */
-+#define _XEN_NMIREASON_parity_error 1
-+#define XEN_NMIREASON_parity_error (1UL << _XEN_NMIREASON_parity_error)
-+ /* Unknown hardware-generated NMI. */
-+#define _XEN_NMIREASON_unknown 2
-+#define XEN_NMIREASON_unknown (1UL << _XEN_NMIREASON_unknown)
++/* NB. All ops return zero on success, except XENVER_{version,pagesize} */
+
-+/*
-+ * long nmi_op(unsigned int cmd, void *arg)
-+ * NB. All ops return zero on success, else a negative error code.
-+ */
++/* arg == NULL; returns major:minor (16:16). */
++#define XENVER_version 0
+
-+/*
-+ * Register NMI callback for this (calling) VCPU. Currently this only makes
-+ * sense for domain 0, vcpu 0. All other callers will be returned EINVAL.
-+ * arg == pointer to xennmi_callback structure.
-+ */
-+#define XENNMI_register_callback 0
-+struct xennmi_callback {
-+ unsigned long handler_address;
-+ unsigned long pad;
++/* arg == xen_extraversion_t. */
++#define XENVER_extraversion 1
++typedef char xen_extraversion_t[16];
++#define XEN_EXTRAVERSION_LEN (sizeof(xen_extraversion_t))
++
++/* arg == xen_compile_info_t. */
++#define XENVER_compile_info 2
++struct xen_compile_info {
++ char compiler[64];
++ char compile_by[16];
++ char compile_domain[32];
++ char compile_date[32];
+};
-+typedef struct xennmi_callback xennmi_callback_t;
-+DEFINE_XEN_GUEST_HANDLE(xennmi_callback_t);
++typedef struct xen_compile_info xen_compile_info_t;
+
-+/*
-+ * Deregister NMI callback for this (calling) VCPU.
-+ * arg == NULL.
-+ */
-+#define XENNMI_unregister_callback 1
++#define XENVER_capabilities 3
++typedef char xen_capabilities_info_t[1024];
++#define XEN_CAPABILITIES_INFO_LEN (sizeof(xen_capabilities_info_t))
+
-+#endif /* __XEN_PUBLIC_NMI_H__ */
++#define XENVER_changeset 4
++typedef char xen_changeset_info_t[64];
++#define XEN_CHANGESET_INFO_LEN (sizeof(xen_changeset_info_t))
++
++#define XENVER_platform_parameters 5
++struct xen_platform_parameters {
++ unsigned long virt_start;
++};
++typedef struct xen_platform_parameters xen_platform_parameters_t;
++
++#define XENVER_get_features 6
++struct xen_feature_info {
++ unsigned int submap_idx; /* IN: which 32-bit submap to return */
++ uint32_t submap; /* OUT: 32-bit submap */
++};
++typedef struct xen_feature_info xen_feature_info_t;
++
++/* Declares the features reported by XENVER_get_features. */
++#include "features.h"
++
++/* arg == NULL; returns host memory page size. */
++#define XENVER_pagesize 7
++
++/* arg == xen_domain_handle_t. */
++#define XENVER_guest_handle 8
++
++#endif /* __XEN_PUBLIC_VERSION_H__ */
+
+/*
+ * Local variables:
@@ -103283,11 +142802,15 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/physdev.h tmp-linux-2.6-xen.patch/include/xen/interface/physdev.h
---- pristine-linux-2.6.18.2/include/xen/interface/physdev.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/physdev.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,169 @@
-+/*
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/xen-compat.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/xen-compat.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,44 @@
++/******************************************************************************
++ * xen-compat.h
++ *
++ * Guest OS interface to Xen. Compatibility layer.
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -103305,147 +142828,661 @@
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2006, Christian Limpach
+ */
+
-+#ifndef __XEN_PUBLIC_PHYSDEV_H__
-+#define __XEN_PUBLIC_PHYSDEV_H__
++#ifndef __XEN_PUBLIC_XEN_COMPAT_H__
++#define __XEN_PUBLIC_XEN_COMPAT_H__
++
++#define __XEN_LATEST_INTERFACE_VERSION__ 0x00030209
++
++#if defined(__XEN__) || defined(__XEN_TOOLS__)
++/* Xen is built with matching headers and implements the latest interface. */
++#define __XEN_INTERFACE_VERSION__ __XEN_LATEST_INTERFACE_VERSION__
++#elif !defined(__XEN_INTERFACE_VERSION__)
++/* Guests which do not specify a version get the legacy interface. */
++#define __XEN_INTERFACE_VERSION__ 0x00000000
++#endif
++
++#if __XEN_INTERFACE_VERSION__ > __XEN_LATEST_INTERFACE_VERSION__
++#error "These header files do not support the requested interface version."
++#endif
++
++#endif /* __XEN_PUBLIC_XEN_COMPAT_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/xen.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/xen.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,639 @@
++/******************************************************************************
++ * xen.h
++ *
++ * Guest OS interface to Xen.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2004, K A Fraser
++ */
++
++#ifndef __XEN_PUBLIC_XEN_H__
++#define __XEN_PUBLIC_XEN_H__
++
++#include "xen-compat.h"
++
++#if defined(__i386__) || defined(__x86_64__)
++#include "arch-x86/xen.h"
++#elif defined(__ia64__)
++#include "arch-ia64.h"
++#else
++#error "Unsupported architecture"
++#endif
++
++#ifndef __ASSEMBLY__
++/* Guest handles for primitive C types. */
++DEFINE_XEN_GUEST_HANDLE(char);
++__DEFINE_XEN_GUEST_HANDLE(uchar, unsigned char);
++DEFINE_XEN_GUEST_HANDLE(int);
++__DEFINE_XEN_GUEST_HANDLE(uint, unsigned int);
++DEFINE_XEN_GUEST_HANDLE(long);
++__DEFINE_XEN_GUEST_HANDLE(ulong, unsigned long);
++DEFINE_XEN_GUEST_HANDLE(void);
++
++DEFINE_XEN_GUEST_HANDLE(xen_pfn_t);
++#endif
+
+/*
-+ * Prototype for this hypercall is:
-+ * int physdev_op(int cmd, void *args)
-+ * @cmd == PHYSDEVOP_??? (physdev operation).
-+ * @args == Operation-specific extra arguments (NULL if none).
++ * HYPERCALLS
++ */
++
++#define __HYPERVISOR_set_trap_table 0
++#define __HYPERVISOR_mmu_update 1
++#define __HYPERVISOR_set_gdt 2
++#define __HYPERVISOR_stack_switch 3
++#define __HYPERVISOR_set_callbacks 4
++#define __HYPERVISOR_fpu_taskswitch 5
++#define __HYPERVISOR_sched_op_compat 6 /* compat since 0x00030101 */
++#define __HYPERVISOR_platform_op 7
++#define __HYPERVISOR_set_debugreg 8
++#define __HYPERVISOR_get_debugreg 9
++#define __HYPERVISOR_update_descriptor 10
++#define __HYPERVISOR_memory_op 12
++#define __HYPERVISOR_multicall 13
++#define __HYPERVISOR_update_va_mapping 14
++#define __HYPERVISOR_set_timer_op 15
++#define __HYPERVISOR_event_channel_op_compat 16 /* compat since 0x00030202 */
++#define __HYPERVISOR_xen_version 17
++#define __HYPERVISOR_console_io 18
++#define __HYPERVISOR_physdev_op_compat 19 /* compat since 0x00030202 */
++#define __HYPERVISOR_grant_table_op 20
++#define __HYPERVISOR_vm_assist 21
++#define __HYPERVISOR_update_va_mapping_otherdomain 22
++#define __HYPERVISOR_iret 23 /* x86 only */
++#define __HYPERVISOR_vcpu_op 24
++#define __HYPERVISOR_set_segment_base 25 /* x86/64 only */
++#define __HYPERVISOR_mmuext_op 26
++#define __HYPERVISOR_xsm_op 27
++#define __HYPERVISOR_nmi_op 28
++#define __HYPERVISOR_sched_op 29
++#define __HYPERVISOR_callback_op 30
++#define __HYPERVISOR_xenoprof_op 31
++#define __HYPERVISOR_event_channel_op 32
++#define __HYPERVISOR_physdev_op 33
++#define __HYPERVISOR_hvm_op 34
++#define __HYPERVISOR_sysctl 35
++#define __HYPERVISOR_domctl 36
++#define __HYPERVISOR_kexec_op 37
++
++/* Architecture-specific hypercall definitions. */
++#define __HYPERVISOR_arch_0 48
++#define __HYPERVISOR_arch_1 49
++#define __HYPERVISOR_arch_2 50
++#define __HYPERVISOR_arch_3 51
++#define __HYPERVISOR_arch_4 52
++#define __HYPERVISOR_arch_5 53
++#define __HYPERVISOR_arch_6 54
++#define __HYPERVISOR_arch_7 55
++
++/*
++ * HYPERCALL COMPATIBILITY.
++ */
++
++/* New sched_op hypercall introduced in 0x00030101. */
++#if __XEN_INTERFACE_VERSION__ < 0x00030101
++#undef __HYPERVISOR_sched_op
++#define __HYPERVISOR_sched_op __HYPERVISOR_sched_op_compat
++#endif
++
++/* New event-channel and physdev hypercalls introduced in 0x00030202. */
++#if __XEN_INTERFACE_VERSION__ < 0x00030202
++#undef __HYPERVISOR_event_channel_op
++#define __HYPERVISOR_event_channel_op __HYPERVISOR_event_channel_op_compat
++#undef __HYPERVISOR_physdev_op
++#define __HYPERVISOR_physdev_op __HYPERVISOR_physdev_op_compat
++#endif
++
++/* New platform_op hypercall introduced in 0x00030204. */
++#if __XEN_INTERFACE_VERSION__ < 0x00030204
++#define __HYPERVISOR_dom0_op __HYPERVISOR_platform_op
++#endif
++
++/*
++ * VIRTUAL INTERRUPTS
++ *
++ * Virtual interrupts that a guest OS may receive from Xen.
++ *
++ * In the side comments, 'V.' denotes a per-VCPU VIRQ while 'G.' denotes a
++ * global VIRQ. The former can be bound once per VCPU and cannot be re-bound.
++ * The latter can be allocated only once per guest: they must initially be
++ * allocated to VCPU0 but can subsequently be re-bound.
+ */
++#define VIRQ_TIMER 0 /* V. Timebase update, and/or requested timeout. */
++#define VIRQ_DEBUG 1 /* V. Request guest to dump debug info. */
++#define VIRQ_CONSOLE 2 /* G. (DOM0) Bytes received on emergency console. */
++#define VIRQ_DOM_EXC 3 /* G. (DOM0) Exceptional event for some domain. */
++#define VIRQ_TBUF 4 /* G. (DOM0) Trace buffer has records available. */
++#define VIRQ_DEBUGGER 6 /* G. (DOM0) A domain has paused for debugging. */
++#define VIRQ_XENOPROF 7 /* V. XenOprofile interrupt: new sample available */
++#define VIRQ_CON_RING 8 /* G. (DOM0) Bytes received on console */
++
++/* Architecture-specific VIRQ definitions. */
++#define VIRQ_ARCH_0 16
++#define VIRQ_ARCH_1 17
++#define VIRQ_ARCH_2 18
++#define VIRQ_ARCH_3 19
++#define VIRQ_ARCH_4 20
++#define VIRQ_ARCH_5 21
++#define VIRQ_ARCH_6 22
++#define VIRQ_ARCH_7 23
++
++#define NR_VIRQS 24
+
+/*
-+ * Notify end-of-interrupt (EOI) for the specified IRQ.
-+ * @arg == pointer to physdev_eoi structure.
++ * MMU-UPDATE REQUESTS
++ *
++ * HYPERVISOR_mmu_update() accepts a list of (ptr, val) pairs.
++ * A foreigndom (FD) can be specified (or DOMID_SELF for none).
++ * Where the FD has some effect, it is described below.
++ * ptr[1:0] specifies the appropriate MMU_* command.
++ *
++ * ptr[1:0] == MMU_NORMAL_PT_UPDATE:
++ * Updates an entry in a page table. If updating an L1 table, and the new
++ * table entry is valid/present, the mapped frame must belong to the FD, if
++ * an FD has been specified. If attempting to map an I/O page then the
++ * caller assumes the privilege of the FD.
++ * FD == DOMID_IO: Permit /only/ I/O mappings, at the priv level of the caller.
++ * FD == DOMID_XEN: Map restricted areas of Xen's heap space.
++ * ptr[:2] -- Machine address of the page-table entry to modify.
++ * val -- Value to write.
++ *
++ * ptr[1:0] == MMU_MACHPHYS_UPDATE:
++ * Updates an entry in the machine->pseudo-physical mapping table.
++ * ptr[:2] -- Machine address within the frame whose mapping to modify.
++ * The frame must belong to the FD, if one is specified.
++ * val -- Value to write into the mapping entry.
++ *
++ * ptr[1:0] == MMU_PT_UPDATE_PRESERVE_AD:
++ * As MMU_NORMAL_PT_UPDATE above, but A/D bits currently in the PTE are ORed
++ * with those in @val.
++ */
++#define MMU_NORMAL_PT_UPDATE 0 /* checked '*ptr = val'. ptr is MA. */
++#define MMU_MACHPHYS_UPDATE 1 /* ptr = MA of frame to modify entry for */
++#define MMU_PT_UPDATE_PRESERVE_AD 2 /* atomically: *ptr = val | (*ptr&(A|D)) */
++
++/*
++ * MMU EXTENDED OPERATIONS
++ *
++ * HYPERVISOR_mmuext_op() accepts a list of mmuext_op structures.
++ * A foreigndom (FD) can be specified (or DOMID_SELF for none).
++ * Where the FD has some effect, it is described below.
++ *
++ * cmd: MMUEXT_(UN)PIN_*_TABLE
++ * mfn: Machine frame number to be (un)pinned as a p.t. page.
++ * The frame must belong to the FD, if one is specified.
++ *
++ * cmd: MMUEXT_NEW_BASEPTR
++ * mfn: Machine frame number of new page-table base to install in MMU.
++ *
++ * cmd: MMUEXT_NEW_USER_BASEPTR [x86/64 only]
++ * mfn: Machine frame number of new page-table base to install in MMU
++ * when in user space.
++ *
++ * cmd: MMUEXT_TLB_FLUSH_LOCAL
++ * No additional arguments. Flushes local TLB.
++ *
++ * cmd: MMUEXT_INVLPG_LOCAL
++ * linear_addr: Linear address to be flushed from the local TLB.
++ *
++ * cmd: MMUEXT_TLB_FLUSH_MULTI
++ * vcpumask: Pointer to bitmap of VCPUs to be flushed.
++ *
++ * cmd: MMUEXT_INVLPG_MULTI
++ * linear_addr: Linear address to be flushed.
++ * vcpumask: Pointer to bitmap of VCPUs to be flushed.
++ *
++ * cmd: MMUEXT_TLB_FLUSH_ALL
++ * No additional arguments. Flushes all VCPUs' TLBs.
++ *
++ * cmd: MMUEXT_INVLPG_ALL
++ * linear_addr: Linear address to be flushed from all VCPUs' TLBs.
++ *
++ * cmd: MMUEXT_FLUSH_CACHE
++ * No additional arguments. Writes back and flushes cache contents.
++ *
++ * cmd: MMUEXT_SET_LDT
++ * linear_addr: Linear address of LDT base (NB. must be page-aligned).
++ * nr_ents: Number of entries in LDT.
+ */
-+#define PHYSDEVOP_eoi 12
-+struct physdev_eoi {
-+ /* IN */
-+ uint32_t irq;
++#define MMUEXT_PIN_L1_TABLE 0
++#define MMUEXT_PIN_L2_TABLE 1
++#define MMUEXT_PIN_L3_TABLE 2
++#define MMUEXT_PIN_L4_TABLE 3
++#define MMUEXT_UNPIN_TABLE 4
++#define MMUEXT_NEW_BASEPTR 5
++#define MMUEXT_TLB_FLUSH_LOCAL 6
++#define MMUEXT_INVLPG_LOCAL 7
++#define MMUEXT_TLB_FLUSH_MULTI 8
++#define MMUEXT_INVLPG_MULTI 9
++#define MMUEXT_TLB_FLUSH_ALL 10
++#define MMUEXT_INVLPG_ALL 11
++#define MMUEXT_FLUSH_CACHE 12
++#define MMUEXT_SET_LDT 13
++#define MMUEXT_NEW_USER_BASEPTR 15
++
++#ifndef __ASSEMBLY__
++struct mmuext_op {
++ unsigned int cmd;
++ union {
++ /* [UN]PIN_TABLE, NEW_BASEPTR, NEW_USER_BASEPTR */
++ xen_pfn_t mfn;
++ /* INVLPG_LOCAL, INVLPG_ALL, SET_LDT */
++ unsigned long linear_addr;
++ } arg1;
++ union {
++ /* SET_LDT */
++ unsigned int nr_ents;
++ /* TLB_FLUSH_MULTI, INVLPG_MULTI */
++#if __XEN_INTERFACE_VERSION__ >= 0x00030205
++ XEN_GUEST_HANDLE(void) vcpumask;
++#else
++ void *vcpumask;
++#endif
++ } arg2;
+};
-+typedef struct physdev_eoi physdev_eoi_t;
-+DEFINE_XEN_GUEST_HANDLE(physdev_eoi_t);
++typedef struct mmuext_op mmuext_op_t;
++DEFINE_XEN_GUEST_HANDLE(mmuext_op_t);
++#endif
++
++/* These are passed as 'flags' to update_va_mapping. They can be ORed. */
++/* When specifying UVMF_MULTI, also OR in a pointer to a CPU bitmap. */
++/* UVMF_LOCAL is merely UVMF_MULTI with a NULL bitmap pointer. */
++#define UVMF_NONE (0UL<<0) /* No flushing at all. */
++#define UVMF_TLB_FLUSH (1UL<<0) /* Flush entire TLB(s). */
++#define UVMF_INVLPG (2UL<<0) /* Flush only one entry. */
++#define UVMF_FLUSHTYPE_MASK (3UL<<0)
++#define UVMF_MULTI (0UL<<2) /* Flush subset of TLBs. */
++#define UVMF_LOCAL (0UL<<2) /* Flush local TLB. */
++#define UVMF_ALL (1UL<<2) /* Flush all TLBs. */
++
++/*
++ * Commands to HYPERVISOR_console_io().
++ */
++#define CONSOLEIO_write 0
++#define CONSOLEIO_read 1
++
++/*
++ * Commands to HYPERVISOR_vm_assist().
++ */
++#define VMASST_CMD_enable 0
++#define VMASST_CMD_disable 1
++
++/* x86/32 guests: simulate full 4GB segment limits. */
++#define VMASST_TYPE_4gb_segments 0
++
++/* x86/32 guests: trap (vector 15) whenever above vmassist is used. */
++#define VMASST_TYPE_4gb_segments_notify 1
++
++/*
++ * x86 guests: support writes to bottom-level PTEs.
++ * NB1. Page-directory entries cannot be written.
++ * NB2. Guest must continue to remove all writable mappings of PTEs.
++ */
++#define VMASST_TYPE_writable_pagetables 2
++
++/* x86/PAE guests: support PDPTs above 4GB. */
++#define VMASST_TYPE_pae_extended_cr3 3
++
++#define MAX_VMASST_TYPE 3
++
++#ifndef __ASSEMBLY__
++
++typedef uint16_t domid_t;
++
++/* Domain ids >= DOMID_FIRST_RESERVED cannot be used for ordinary domains. */
++#define DOMID_FIRST_RESERVED (0x7FF0U)
++
++/* DOMID_SELF is used in certain contexts to refer to oneself. */
++#define DOMID_SELF (0x7FF0U)
++
++/*
++ * DOMID_IO is used to restrict page-table updates to mapping I/O memory.
++ * Although no Foreign Domain need be specified to map I/O pages, DOMID_IO
++ * is useful to ensure that no mappings to the OS's own heap are accidentally
++ * installed. (e.g., in Linux this could cause havoc as reference counts
++ * aren't adjusted on the I/O-mapping code path).
++ * This only makes sense in MMUEXT_SET_FOREIGNDOM, but in that context can
++ * be specified by any calling domain.
++ */
++#define DOMID_IO (0x7FF1U)
++
++/*
++ * DOMID_XEN is used to allow privileged domains to map restricted parts of
++ * Xen's heap space (e.g., the machine_to_phys table).
++ * This only makes sense in MMUEXT_SET_FOREIGNDOM, and is only permitted if
++ * the caller is privileged.
++ */
++#define DOMID_XEN (0x7FF2U)
+
+/*
-+ * Query the status of an IRQ line.
-+ * @arg == pointer to physdev_irq_status_query structure.
++ * Send an array of these to HYPERVISOR_mmu_update().
++ * NB. The fields are natural pointer/address size for this architecture.
+ */
-+#define PHYSDEVOP_irq_status_query 5
-+struct physdev_irq_status_query {
-+ /* IN */
-+ uint32_t irq;
-+ /* OUT */
-+ uint32_t flags; /* XENIRQSTAT_* */
++struct mmu_update {
++ uint64_t ptr; /* Machine address of PTE. */
++ uint64_t val; /* New contents of PTE. */
+};
-+typedef struct physdev_irq_status_query physdev_irq_status_query_t;
-+DEFINE_XEN_GUEST_HANDLE(physdev_irq_status_query_t);
-+
-+/* Need to call PHYSDEVOP_eoi when the IRQ has been serviced? */
-+#define _XENIRQSTAT_needs_eoi (0)
-+#define XENIRQSTAT_needs_eoi (1U<<_XENIRQSTAT_needs_eoi)
-+
-+/* IRQ shared by multiple guests? */
-+#define _XENIRQSTAT_shared (1)
-+#define XENIRQSTAT_shared (1U<<_XENIRQSTAT_shared)
++typedef struct mmu_update mmu_update_t;
++DEFINE_XEN_GUEST_HANDLE(mmu_update_t);
+
+/*
-+ * Set the current VCPU's I/O privilege level.
-+ * @arg == pointer to physdev_set_iopl structure.
++ * Send an array of these to HYPERVISOR_multicall().
++ * NB. The fields are natural register size for this architecture.
+ */
-+#define PHYSDEVOP_set_iopl 6
-+struct physdev_set_iopl {
-+ /* IN */
-+ uint32_t iopl;
++struct multicall_entry {
++ unsigned long op, result;
++ unsigned long args[6];
+};
-+typedef struct physdev_set_iopl physdev_set_iopl_t;
-+DEFINE_XEN_GUEST_HANDLE(physdev_set_iopl_t);
++typedef struct multicall_entry multicall_entry_t;
++DEFINE_XEN_GUEST_HANDLE(multicall_entry_t);
+
+/*
-+ * Set the current VCPU's I/O-port permissions bitmap.
-+ * @arg == pointer to physdev_set_iobitmap structure.
++ * Event channel endpoints per domain:
++ * 1024 if a long is 32 bits; 4096 if a long is 64 bits.
+ */
-+#define PHYSDEVOP_set_iobitmap 7
-+struct physdev_set_iobitmap {
-+ /* IN */
-+ XEN_GUEST_HANDLE_00030205(uint8_t) bitmap;
-+ uint32_t nr_ports;
-+};
-+typedef struct physdev_set_iobitmap physdev_set_iobitmap_t;
-+DEFINE_XEN_GUEST_HANDLE(physdev_set_iobitmap_t);
++#define NR_EVENT_CHANNELS (sizeof(unsigned long) * sizeof(unsigned long) * 64)
++
++struct vcpu_time_info {
++ /*
++ * Updates to the following values are preceded and followed by an
++ * increment of 'version'. The guest can therefore detect updates by
++ * looking for changes to 'version'. If the least-significant bit of
++ * the version number is set then an update is in progress and the guest
++ * must wait to read a consistent set of values.
++ * The correct way to interact with the version number is similar to
++ * Linux's seqlock: see the implementations of read_seqbegin/read_seqretry.
++ */
++ uint32_t version;
++ uint32_t pad0;
++ uint64_t tsc_timestamp; /* TSC at last update of time vals. */
++ uint64_t system_time; /* Time, in nanosecs, since boot. */
++ /*
++ * Current system time:
++ * system_time +
++ * ((((tsc - tsc_timestamp) << tsc_shift) * tsc_to_system_mul) >> 32)
++ * CPU frequency (Hz):
++ * ((10^9 << 32) / tsc_to_system_mul) >> tsc_shift
++ */
++ uint32_t tsc_to_system_mul;
++ int8_t tsc_shift;
++ int8_t pad1[3];
++}; /* 32 bytes */
++typedef struct vcpu_time_info vcpu_time_info_t;
++
++struct vcpu_info {
++ /*
++ * 'evtchn_upcall_pending' is written non-zero by Xen to indicate
++ * a pending notification for a particular VCPU. It is then cleared
++ * by the guest OS /before/ checking for pending work, thus avoiding
++ * a set-and-check race. Note that the mask is only accessed by Xen
++ * on the CPU that is currently hosting the VCPU. This means that the
++ * pending and mask flags can be updated by the guest without special
++ * synchronisation (i.e., no need for the x86 LOCK prefix).
++ * This may seem suboptimal because if the pending flag is set by
++ * a different CPU then an IPI may be scheduled even when the mask
++ * is set. However, note:
++ * 1. The task of 'interrupt holdoff' is covered by the per-event-
++ * channel mask bits. A 'noisy' event that is continually being
++ * triggered can be masked at source at this very precise
++ * granularity.
++ * 2. The main purpose of the per-VCPU mask is therefore to restrict
++ * reentrant execution: whether for concurrency control, or to
++ * prevent unbounded stack usage. Whatever the purpose, we expect
++ * that the mask will be asserted only for short periods at a time,
++ * and so the likelihood of a 'spurious' IPI is suitably small.
++ * The mask is read before making an event upcall to the guest: a
++ * non-zero mask therefore guarantees that the VCPU will not receive
++ * an upcall activation. The mask is cleared when the VCPU requests
++ * to block: this avoids wakeup-waiting races.
++ */
++ uint8_t evtchn_upcall_pending;
++ uint8_t evtchn_upcall_mask;
++ unsigned long evtchn_pending_sel;
++ struct arch_vcpu_info arch;
++ struct vcpu_time_info time;
++}; /* 64 bytes (x86) */
++#ifndef __XEN__
++typedef struct vcpu_info vcpu_info_t;
++#endif
+
+/*
-+ * Read or write an IO-APIC register.
-+ * @arg == pointer to physdev_apic structure.
++ * Xen/kernel shared data -- pointer provided in start_info.
++ *
++ * This structure is defined to be both smaller than a page, and the
++ * only data on the shared page, but may vary in actual size even within
++ * compatible Xen versions; guests should not rely on the size
++ * of this structure remaining constant.
+ */
-+#define PHYSDEVOP_apic_read 8
-+#define PHYSDEVOP_apic_write 9
-+struct physdev_apic {
-+ /* IN */
-+ unsigned long apic_physbase;
-+ uint32_t reg;
-+ /* IN or OUT */
-+ uint32_t value;
++struct shared_info {
++ struct vcpu_info vcpu_info[MAX_VIRT_CPUS];
++
++ /*
++ * A domain can create "event channels" on which it can send and receive
++ * asynchronous event notifications. There are three classes of event that
++ * are delivered by this mechanism:
++ * 1. Bi-directional inter- and intra-domain connections. Domains must
++ * arrange out-of-band to set up a connection (usually by allocating
++ * an unbound 'listener' port and avertising that via a storage service
++ * such as xenstore).
++ * 2. Physical interrupts. A domain with suitable hardware-access
++ * privileges can bind an event-channel port to a physical interrupt
++ * source.
++ * 3. Virtual interrupts ('events'). A domain can bind an event-channel
++ * port to a virtual interrupt source, such as the virtual-timer
++ * device or the emergency console.
++ *
++ * Event channels are addressed by a "port index". Each channel is
++ * associated with two bits of information:
++ * 1. PENDING -- notifies the domain that there is a pending notification
++ * to be processed. This bit is cleared by the guest.
++ * 2. MASK -- if this bit is clear then a 0->1 transition of PENDING
++ * will cause an asynchronous upcall to be scheduled. This bit is only
++ * updated by the guest. It is read-only within Xen. If a channel
++ * becomes pending while the channel is masked then the 'edge' is lost
++ * (i.e., when the channel is unmasked, the guest must manually handle
++ * pending notifications as no upcall will be scheduled by Xen).
++ *
++ * To expedite scanning of pending notifications, any 0->1 pending
++ * transition on an unmasked channel causes a corresponding bit in a
++ * per-vcpu selector word to be set. Each bit in the selector covers a
++ * 'C long' in the PENDING bitfield array.
++ */
++ unsigned long evtchn_pending[sizeof(unsigned long) * 8];
++ unsigned long evtchn_mask[sizeof(unsigned long) * 8];
++
++ /*
++ * Wallclock time: updated only by control software. Guests should base
++ * their gettimeofday() syscall on this wallclock-base value.
++ */
++ uint32_t wc_version; /* Version counter: see vcpu_time_info_t. */
++ uint32_t wc_sec; /* Secs 00:00:00 UTC, Jan 1, 1970. */
++ uint32_t wc_nsec; /* Nsecs 00:00:00 UTC, Jan 1, 1970. */
++
++ struct arch_shared_info arch;
++
+};
-+typedef struct physdev_apic physdev_apic_t;
-+DEFINE_XEN_GUEST_HANDLE(physdev_apic_t);
++#ifndef __XEN__
++typedef struct shared_info shared_info_t;
++#endif
+
+/*
-+ * Allocate or free a physical upcall vector for the specified IRQ line.
-+ * @arg == pointer to physdev_irq structure.
++ * Start-of-day memory layout:
++ * 1. The domain is started within contiguous virtual-memory region.
++ * 2. The contiguous region ends on an aligned 4MB boundary.
++ * 3. This the order of bootstrap elements in the initial virtual region:
++ * a. relocated kernel image
++ * b. initial ram disk [mod_start, mod_len]
++ * c. list of allocated page frames [mfn_list, nr_pages]
++ * d. start_info_t structure [register ESI (x86)]
++ * e. bootstrap page tables [pt_base, CR3 (x86)]
++ * f. bootstrap stack [register ESP (x86)]
++ * 4. Bootstrap elements are packed together, but each is 4kB-aligned.
++ * 5. The initial ram disk may be omitted.
++ * 6. The list of page frames forms a contiguous 'pseudo-physical' memory
++ * layout for the domain. In particular, the bootstrap virtual-memory
++ * region is a 1:1 mapping to the first section of the pseudo-physical map.
++ * 7. All bootstrap elements are mapped read-writable for the guest OS. The
++ * only exception is the bootstrap page table, which is mapped read-only.
++ * 8. There is guaranteed to be at least 512kB padding after the final
++ * bootstrap element. If necessary, the bootstrap virtual region is
++ * extended by an extra 4MB to ensure this.
+ */
-+#define PHYSDEVOP_alloc_irq_vector 10
-+#define PHYSDEVOP_free_irq_vector 11
-+struct physdev_irq {
-+ /* IN */
-+ uint32_t irq;
-+ /* IN or OUT */
-+ uint32_t vector;
++
++#define MAX_GUEST_CMDLINE 1024
++struct start_info {
++ /* THE FOLLOWING ARE FILLED IN BOTH ON INITIAL BOOT AND ON RESUME. */
++ char magic[32]; /* "xen-<version>-<platform>". */
++ unsigned long nr_pages; /* Total pages allocated to this domain. */
++ unsigned long shared_info; /* MACHINE address of shared info struct. */
++ uint32_t flags; /* SIF_xxx flags. */
++ xen_pfn_t store_mfn; /* MACHINE page number of shared page. */
++ uint32_t store_evtchn; /* Event channel for store communication. */
++ union {
++ struct {
++ xen_pfn_t mfn; /* MACHINE page number of console page. */
++ uint32_t evtchn; /* Event channel for console page. */
++ } domU;
++ struct {
++ uint32_t info_off; /* Offset of console_info struct. */
++ uint32_t info_size; /* Size of console_info struct from start.*/
++ } dom0;
++ } console;
++ /* THE FOLLOWING ARE ONLY FILLED IN ON INITIAL BOOT (NOT RESUME). */
++ unsigned long pt_base; /* VIRTUAL address of page directory. */
++ unsigned long nr_pt_frames; /* Number of bootstrap p.t. frames. */
++ unsigned long mfn_list; /* VIRTUAL address of page-frame list. */
++ unsigned long mod_start; /* VIRTUAL address of pre-loaded module. */
++ unsigned long mod_len; /* Size (bytes) of pre-loaded module. */
++ int8_t cmd_line[MAX_GUEST_CMDLINE];
+};
-+typedef struct physdev_irq physdev_irq_t;
-+DEFINE_XEN_GUEST_HANDLE(physdev_irq_t);
++typedef struct start_info start_info_t;
++
++/* New console union for dom0 introduced in 0x00030203. */
++#if __XEN_INTERFACE_VERSION__ < 0x00030203
++#define console_mfn console.domU.mfn
++#define console_evtchn console.domU.evtchn
++#endif
++
++/* These flags are passed in the 'flags' field of start_info_t. */
++#define SIF_PRIVILEGED (1<<0) /* Is the domain privileged? */
++#define SIF_INITDOMAIN (1<<1) /* Is this the initial control domain? */
++#define SIF_PM_MASK (0xFF<<8) /* reserve 1 byte for xen-pm options */
++
++typedef struct dom0_vga_console_info {
++ uint8_t video_type; /* DOM0_VGA_CONSOLE_??? */
++#define XEN_VGATYPE_TEXT_MODE_3 0x03
++#define XEN_VGATYPE_VESA_LFB 0x23
+
-+/*
-+ * Argument to physdev_op_compat() hypercall. Superceded by new physdev_op()
-+ * hypercall since 0x00030202.
-+ */
-+struct physdev_op {
-+ uint32_t cmd;
+ union {
-+ struct physdev_irq_status_query irq_status_query;
-+ struct physdev_set_iopl set_iopl;
-+ struct physdev_set_iobitmap set_iobitmap;
-+ struct physdev_apic apic_op;
-+ struct physdev_irq irq_op;
++ struct {
++ /* Font height, in pixels. */
++ uint16_t font_height;
++ /* Cursor location (column, row). */
++ uint16_t cursor_x, cursor_y;
++ /* Number of rows and columns (dimensions in characters). */
++ uint16_t rows, columns;
++ } text_mode_3;
++
++ struct {
++ /* Width and height, in pixels. */
++ uint16_t width, height;
++ /* Bytes per scan line. */
++ uint16_t bytes_per_line;
++ /* Bits per pixel. */
++ uint16_t bits_per_pixel;
++ /* LFB physical address, and size (in units of 64kB). */
++ uint32_t lfb_base;
++ uint32_t lfb_size;
++ /* RGB mask offsets and sizes, as defined by VBE 1.2+ */
++ uint8_t red_pos, red_size;
++ uint8_t green_pos, green_size;
++ uint8_t blue_pos, blue_size;
++ uint8_t rsvd_pos, rsvd_size;
++#if __XEN_INTERFACE_VERSION__ >= 0x00030206
++ /* VESA capabilities (offset 0xa, VESA command 0x4f00). */
++ uint32_t gbl_caps;
++ /* Mode attributes (offset 0x0, VESA command 0x4f01). */
++ uint16_t mode_attrs;
++#endif
++ } vesa_lfb;
+ } u;
-+};
-+typedef struct physdev_op physdev_op_t;
-+DEFINE_XEN_GUEST_HANDLE(physdev_op_t);
++} dom0_vga_console_info_t;
++#define xen_vga_console_info dom0_vga_console_info
++#define xen_vga_console_info_t dom0_vga_console_info_t
+
-+/*
-+ * Notify that some PIRQ-bound event channels have been unmasked.
-+ * ** This command is obsolete since interface version 0x00030202 and is **
-+ * ** unsupported by newer versions of Xen. **
-+ */
-+#define PHYSDEVOP_IRQ_UNMASK_NOTIFY 4
++typedef uint8_t xen_domain_handle_t[16];
+
-+/*
-+ * These all-capitals physdev operation names are superceded by the new names
-+ * (defined above) since interface version 0x00030202.
-+ */
-+#define PHYSDEVOP_IRQ_STATUS_QUERY PHYSDEVOP_irq_status_query
-+#define PHYSDEVOP_SET_IOPL PHYSDEVOP_set_iopl
-+#define PHYSDEVOP_SET_IOBITMAP PHYSDEVOP_set_iobitmap
-+#define PHYSDEVOP_APIC_READ PHYSDEVOP_apic_read
-+#define PHYSDEVOP_APIC_WRITE PHYSDEVOP_apic_write
-+#define PHYSDEVOP_ASSIGN_VECTOR PHYSDEVOP_alloc_irq_vector
-+#define PHYSDEVOP_FREE_VECTOR PHYSDEVOP_free_irq_vector
-+#define PHYSDEVOP_IRQ_NEEDS_UNMASK_NOTIFY XENIRQSTAT_needs_eoi
-+#define PHYSDEVOP_IRQ_SHARED XENIRQSTAT_shared
++/* Turn a plain number into a C unsigned long constant. */
++#define __mk_unsigned_long(x) x ## UL
++#define mk_unsigned_long(x) __mk_unsigned_long(x)
+
-+#endif /* __XEN_PUBLIC_PHYSDEV_H__ */
++__DEFINE_XEN_GUEST_HANDLE(uint8, uint8_t);
++__DEFINE_XEN_GUEST_HANDLE(uint16, uint16_t);
++__DEFINE_XEN_GUEST_HANDLE(uint32, uint32_t);
++__DEFINE_XEN_GUEST_HANDLE(uint64, uint64_t);
++
++#else /* __ASSEMBLY__ */
++
++/* In assembly code we cannot use C numeric constant suffixes. */
++#define mk_unsigned_long(x) x
++
++#endif /* !__ASSEMBLY__ */
++
++/* Default definitions for macros used by domctl/sysctl. */
++#if defined(__XEN__) || defined(__XEN_TOOLS__)
++#ifndef uint64_aligned_t
++#define uint64_aligned_t uint64_t
++#endif
++#ifndef XEN_GUEST_HANDLE_64
++#define XEN_GUEST_HANDLE_64(name) XEN_GUEST_HANDLE(name)
++#endif
++#endif
++
++#endif /* __XEN_PUBLIC_XEN_H__ */
+
+/*
+ * Local variables:
@@ -103456,15 +143493,11 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/platform.h tmp-linux-2.6-xen.patch/include/xen/interface/platform.h
---- pristine-linux-2.6.18.2/include/xen/interface/platform.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/platform.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,183 @@
-+/******************************************************************************
-+ * platform.h
-+ *
-+ * Hardware platform operations. Intended for use by domain-0 kernel.
-+ *
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/xencomm.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/xencomm.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,41 @@
++/*
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -103483,174 +143516,37 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (c) 2002-2006, K Fraser
-+ */
-+
-+#ifndef __XEN_PUBLIC_PLATFORM_H__
-+#define __XEN_PUBLIC_PLATFORM_H__
-+
-+#include "xen.h"
-+
-+#define XENPF_INTERFACE_VERSION 0x03000001
-+
-+/*
-+ * Set clock such that it would read <secs,nsecs> after 00:00:00 UTC,
-+ * 1 January, 1970 if the current system time was <system_time>.
-+ */
-+#define XENPF_settime 17
-+struct xenpf_settime {
-+ /* IN variables. */
-+ uint32_t secs;
-+ uint32_t nsecs;
-+ uint64_t system_time;
-+};
-+typedef struct xenpf_settime xenpf_settime_t;
-+DEFINE_XEN_GUEST_HANDLE(xenpf_settime_t);
-+
-+/*
-+ * Request memory range (@mfn, @mfn+ at nr_mfns-1) to have type @type.
-+ * On x86, @type is an architecture-defined MTRR memory type.
-+ * On success, returns the MTRR that was used (@reg) and a handle that can
-+ * be passed to XENPF_DEL_MEMTYPE to accurately tear down the new setting.
-+ * (x86-specific).
-+ */
-+#define XENPF_add_memtype 31
-+struct xenpf_add_memtype {
-+ /* IN variables. */
-+ xen_pfn_t mfn;
-+ uint64_t nr_mfns;
-+ uint32_t type;
-+ /* OUT variables. */
-+ uint32_t handle;
-+ uint32_t reg;
-+};
-+typedef struct xenpf_add_memtype xenpf_add_memtype_t;
-+DEFINE_XEN_GUEST_HANDLE(xenpf_add_memtype_t);
-+
-+/*
-+ * Tear down an existing memory-range type. If @handle is remembered then it
-+ * should be passed in to accurately tear down the correct setting (in case
-+ * of overlapping memory regions with differing types). If it is not known
-+ * then @handle should be set to zero. In all cases @reg must be set.
-+ * (x86-specific).
++ * Copyright (C) IBM Corp. 2006
+ */
-+#define XENPF_del_memtype 32
-+struct xenpf_del_memtype {
-+ /* IN variables. */
-+ uint32_t handle;
-+ uint32_t reg;
-+};
-+typedef struct xenpf_del_memtype xenpf_del_memtype_t;
-+DEFINE_XEN_GUEST_HANDLE(xenpf_del_memtype_t);
-+
-+/* Read current type of an MTRR (x86-specific). */
-+#define XENPF_read_memtype 33
-+struct xenpf_read_memtype {
-+ /* IN variables. */
-+ uint32_t reg;
-+ /* OUT variables. */
-+ xen_pfn_t mfn;
-+ uint64_t nr_mfns;
-+ uint32_t type;
-+};
-+typedef struct xenpf_read_memtype xenpf_read_memtype_t;
-+DEFINE_XEN_GUEST_HANDLE(xenpf_read_memtype_t);
-+
-+#define XENPF_microcode_update 35
-+struct xenpf_microcode_update {
-+ /* IN variables. */
-+ XEN_GUEST_HANDLE(void) data; /* Pointer to microcode data */
-+ uint32_t length; /* Length of microcode data. */
-+};
-+typedef struct xenpf_microcode_update xenpf_microcode_update_t;
-+DEFINE_XEN_GUEST_HANDLE(xenpf_microcode_update_t);
-+
-+#define XENPF_platform_quirk 39
-+#define QUIRK_NOIRQBALANCING 1 /* Do not restrict IO-APIC RTE targets */
-+#define QUIRK_IOAPIC_BAD_REGSEL 2 /* IO-APIC REGSEL forgets its value */
-+#define QUIRK_IOAPIC_GOOD_REGSEL 3 /* IO-APIC REGSEL behaves properly */
-+struct xenpf_platform_quirk {
-+ /* IN variables. */
-+ uint32_t quirk_id;
-+};
-+typedef struct xenpf_platform_quirk xenpf_platform_quirk_t;
-+DEFINE_XEN_GUEST_HANDLE(xenpf_platform_quirk_t);
-+
-+#define XENPF_firmware_info 50
-+#define XEN_FW_DISK_INFO 1 /* from int 13 AH=08/41/48 */
-+#define XEN_FW_DISK_MBR_SIGNATURE 2 /* from MBR offset 0x1b8 */
-+#define XEN_FW_VBEDDC_INFO 3 /* from int 10 AX=4f15 */
-+struct xenpf_firmware_info {
-+ /* IN variables. */
-+ uint32_t type;
-+ uint32_t index;
-+ /* OUT variables. */
-+ union {
-+ struct {
-+ /* Int13, Fn48: Check Extensions Present. */
-+ uint8_t device; /* %dl: bios device number */
-+ uint8_t version; /* %ah: major version */
-+ uint16_t interface_support; /* %cx: support bitmap */
-+ /* Int13, Fn08: Legacy Get Device Parameters. */
-+ uint16_t legacy_max_cylinder; /* %cl[7:6]:%ch: max cyl # */
-+ uint8_t legacy_max_head; /* %dh: max head # */
-+ uint8_t legacy_sectors_per_track; /* %cl[5:0]: max sector # */
-+ /* Int13, Fn41: Get Device Parameters (as filled into %ds:%esi). */
-+ /* NB. First uint16_t of buffer must be set to buffer size. */
-+ XEN_GUEST_HANDLE(void) edd_params;
-+ } disk_info; /* XEN_FW_DISK_INFO */
-+ struct {
-+ uint8_t device; /* bios device number */
-+ uint32_t mbr_signature; /* offset 0x1b8 in mbr */
-+ } disk_mbr_signature; /* XEN_FW_DISK_MBR_SIGNATURE */
-+ struct {
-+ /* Int10, AX=4F15: Get EDID info. */
-+ uint8_t capabilities;
-+ uint8_t edid_transfer_time;
-+ /* must refer to 128-byte buffer */
-+ XEN_GUEST_HANDLE(uint8_t) edid;
-+ } vbeddc_info; /* XEN_FW_VBEDDC_INFO */
-+ } u;
-+};
-+typedef struct xenpf_firmware_info xenpf_firmware_info_t;
-+DEFINE_XEN_GUEST_HANDLE(xenpf_firmware_info_t);
+
-+struct xen_platform_op {
-+ uint32_t cmd;
-+ uint32_t interface_version; /* XENPF_INTERFACE_VERSION */
-+ union {
-+ struct xenpf_settime settime;
-+ struct xenpf_add_memtype add_memtype;
-+ struct xenpf_del_memtype del_memtype;
-+ struct xenpf_read_memtype read_memtype;
-+ struct xenpf_microcode_update microcode;
-+ struct xenpf_platform_quirk platform_quirk;
-+ struct xenpf_firmware_info firmware_info;
-+ uint8_t pad[128];
-+ } u;
-+};
-+typedef struct xen_platform_op xen_platform_op_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_platform_op_t);
-+
-+#endif /* __XEN_PUBLIC_PLATFORM_H__ */
++#ifndef _XEN_XENCOMM_H_
++#define _XEN_XENCOMM_H_
+
-+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
++/* A xencomm descriptor is a scatter/gather list containing physical
++ * addresses corresponding to a virtually contiguous memory area. The
++ * hypervisor translates these physical addresses to machine addresses to copy
++ * to and from the virtually contiguous area.
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/sched.h tmp-linux-2.6-xen.patch/include/xen/interface/sched.h
---- pristine-linux-2.6.18.2/include/xen/interface/sched.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/sched.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,121 @@
++
++#define XENCOMM_MAGIC 0x58434F4D /* 'XCOM' */
++#define XENCOMM_INVALID (~0UL)
++
++struct xencomm_desc {
++ uint32_t magic;
++ uint32_t nr_addrs; /* the number of entries in address[] */
++ uint64_t address[0];
++};
++
++#endif /* _XEN_XENCOMM_H_ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/xenoprof.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/xenoprof.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,138 @@
+/******************************************************************************
-+ * sched.h
++ * xenoprof.h
+ *
-+ * Scheduler state interactions
++ * Interface for enabling system wide profiling based on hardware performance
++ * counters
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
@@ -103670,94 +143566,110 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (c) 2005, Keir Fraser <keir at xensource.com>
++ * Copyright (C) 2005 Hewlett-Packard Co.
++ * Written by Aravind Menon & Jose Renato Santos
+ */
+
-+#ifndef __XEN_PUBLIC_SCHED_H__
-+#define __XEN_PUBLIC_SCHED_H__
++#ifndef __XEN_PUBLIC_XENOPROF_H__
++#define __XEN_PUBLIC_XENOPROF_H__
+
-+#include "event_channel.h"
++#include "xen.h"
+
+/*
-+ * The prototype for this hypercall is:
-+ * long sched_op(int cmd, void *arg)
-+ * @cmd == SCHEDOP_??? (scheduler operation).
-+ * @arg == Operation-specific extra argument(s), as described below.
-+ *
-+ * Versions of Xen prior to 3.0.2 provided only the following legacy version
-+ * of this hypercall, supporting only the commands yield, block and shutdown:
-+ * long sched_op(int cmd, unsigned long arg)
-+ * @cmd == SCHEDOP_??? (scheduler operation).
-+ * @arg == 0 (SCHEDOP_yield and SCHEDOP_block)
-+ * == SHUTDOWN_* code (SCHEDOP_shutdown)
-+ * This legacy version is available to new guests as sched_op_compat().
++ * Commands to HYPERVISOR_xenoprof_op().
+ */
++#define XENOPROF_init 0
++#define XENOPROF_reset_active_list 1
++#define XENOPROF_reset_passive_list 2
++#define XENOPROF_set_active 3
++#define XENOPROF_set_passive 4
++#define XENOPROF_reserve_counters 5
++#define XENOPROF_counter 6
++#define XENOPROF_setup_events 7
++#define XENOPROF_enable_virq 8
++#define XENOPROF_start 9
++#define XENOPROF_stop 10
++#define XENOPROF_disable_virq 11
++#define XENOPROF_release_counters 12
++#define XENOPROF_shutdown 13
++#define XENOPROF_get_buffer 14
++#define XENOPROF_set_backtrace 15
++#define XENOPROF_last_op 15
+
-+/*
-+ * Voluntarily yield the CPU.
-+ * @arg == NULL.
-+ */
-+#define SCHEDOP_yield 0
++#define MAX_OPROF_EVENTS 32
++#define MAX_OPROF_DOMAINS 25
++#define XENOPROF_CPU_TYPE_SIZE 64
+
-+/*
-+ * Block execution of this VCPU until an event is received for processing.
-+ * If called with event upcalls masked, this operation will atomically
-+ * reenable event delivery and check for pending events before blocking the
-+ * VCPU. This avoids a "wakeup waiting" race.
-+ * @arg == NULL.
-+ */
-+#define SCHEDOP_block 1
++/* Xenoprof performance events (not Xen events) */
++struct event_log {
++ uint64_t eip;
++ uint8_t mode;
++ uint8_t event;
++};
+
-+/*
-+ * Halt execution of this domain (all VCPUs) and notify the system controller.
-+ * @arg == pointer to sched_shutdown structure.
-+ */
-+#define SCHEDOP_shutdown 2
-+struct sched_shutdown {
-+ unsigned int reason; /* SHUTDOWN_* */
++/* PC value that indicates a special code */
++#define XENOPROF_ESCAPE_CODE ~0UL
++/* Transient events for the xenoprof->oprofile cpu buf */
++#define XENOPROF_TRACE_BEGIN 1
++
++/* Xenoprof buffer shared between Xen and domain - 1 per VCPU */
++struct xenoprof_buf {
++ uint32_t event_head;
++ uint32_t event_tail;
++ uint32_t event_size;
++ uint32_t vcpu_id;
++ uint64_t xen_samples;
++ uint64_t kernel_samples;
++ uint64_t user_samples;
++ uint64_t lost_samples;
++ struct event_log event_log[1];
+};
-+typedef struct sched_shutdown sched_shutdown_t;
-+DEFINE_XEN_GUEST_HANDLE(sched_shutdown_t);
++#ifndef __XEN__
++typedef struct xenoprof_buf xenoprof_buf_t;
++DEFINE_XEN_GUEST_HANDLE(xenoprof_buf_t);
++#endif
+
-+/*
-+ * Poll a set of event-channel ports. Return when one or more are pending. An
-+ * optional timeout may be specified.
-+ * @arg == pointer to sched_poll structure.
-+ */
-+#define SCHEDOP_poll 3
-+struct sched_poll {
-+ XEN_GUEST_HANDLE(evtchn_port_t) ports;
-+ unsigned int nr_ports;
-+ uint64_t timeout;
++struct xenoprof_init {
++ int32_t num_events;
++ int32_t is_primary;
++ char cpu_type[XENOPROF_CPU_TYPE_SIZE];
+};
-+typedef struct sched_poll sched_poll_t;
-+DEFINE_XEN_GUEST_HANDLE(sched_poll_t);
++typedef struct xenoprof_init xenoprof_init_t;
++DEFINE_XEN_GUEST_HANDLE(xenoprof_init_t);
+
-+/*
-+ * Declare a shutdown for another domain. The main use of this function is
-+ * in interpreting shutdown requests and reasons for fully-virtualized
-+ * domains. A para-virtualized domain may use SCHEDOP_shutdown directly.
-+ * @arg == pointer to sched_remote_shutdown structure.
-+ */
-+#define SCHEDOP_remote_shutdown 4
-+struct sched_remote_shutdown {
-+ domid_t domain_id; /* Remote domain ID */
-+ unsigned int reason; /* SHUTDOWN_xxx reason */
++struct xenoprof_get_buffer {
++ int32_t max_samples;
++ int32_t nbuf;
++ int32_t bufsize;
++ uint64_t buf_gmaddr;
+};
-+typedef struct sched_remote_shutdown sched_remote_shutdown_t;
-+DEFINE_XEN_GUEST_HANDLE(sched_remote_shutdown_t);
++typedef struct xenoprof_get_buffer xenoprof_get_buffer_t;
++DEFINE_XEN_GUEST_HANDLE(xenoprof_get_buffer_t);
+
-+/*
-+ * Reason codes for SCHEDOP_shutdown. These may be interpreted by control
-+ * software to determine the appropriate action. For the most part, Xen does
-+ * not care about the shutdown code.
-+ */
-+#define SHUTDOWN_poweroff 0 /* Domain exited normally. Clean up and kill. */
-+#define SHUTDOWN_reboot 1 /* Clean up, kill, and then restart. */
-+#define SHUTDOWN_suspend 2 /* Clean up, save suspend info, kill. */
-+#define SHUTDOWN_crash 3 /* Tell controller we've crashed. */
++struct xenoprof_counter {
++ uint32_t ind;
++ uint64_t count;
++ uint32_t enabled;
++ uint32_t event;
++ uint32_t hypervisor;
++ uint32_t kernel;
++ uint32_t user;
++ uint64_t unit_mask;
++};
++typedef struct xenoprof_counter xenoprof_counter_t;
++DEFINE_XEN_GUEST_HANDLE(xenoprof_counter_t);
++
++typedef struct xenoprof_passive {
++ uint16_t domain_id;
++ int32_t max_samples;
++ int32_t nbuf;
++ int32_t bufsize;
++ uint64_t buf_gmaddr;
++} xenoprof_passive_t;
++DEFINE_XEN_GUEST_HANDLE(xenoprof_passive_t);
+
-+#endif /* __XEN_PUBLIC_SCHED_H__ */
++
++#endif /* __XEN_PUBLIC_XENOPROF_H__ */
+
+/*
+ * Local variables:
@@ -103768,15 +143680,13 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/sysctl.h tmp-linux-2.6-xen.patch/include/xen/interface/sysctl.h
---- pristine-linux-2.6.18.2/include/xen/interface/sysctl.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/sysctl.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,198 @@
-+/******************************************************************************
-+ * sysctl.h
-+ *
-+ * System management operations. For use by node control stack.
-+ *
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/xsm/acm.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/xsm/acm.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,235 @@
++/*
++ * acm.h: Xen access control module interface defintions
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -103795,295 +143705,210 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (c) 2002-2006, K Fraser
++ * Reiner Sailer <sailer at watson.ibm.com>
++ * Copyright (c) 2005, International Business Machines Corporation.
+ */
+
-+#ifndef __XEN_PUBLIC_SYSCTL_H__
-+#define __XEN_PUBLIC_SYSCTL_H__
++#ifndef _XEN_PUBLIC_ACM_H
++#define _XEN_PUBLIC_ACM_H
+
-+#if !defined(__XEN__) && !defined(__XEN_TOOLS__)
-+#error "sysctl operations are intended for use by node control tools only"
-+#endif
++#include "../xen.h"
+
-+#include "xen.h"
-+#include "domctl.h"
++/* if ACM_DEBUG defined, all hooks should
++ * print a short trace message (comment it out
++ * when not in testing mode )
++ */
++/* #define ACM_DEBUG */
+
-+#define XEN_SYSCTL_INTERFACE_VERSION 0x00000003
++#ifdef ACM_DEBUG
++# define printkd(fmt, args...) printk(fmt,## args)
++#else
++# define printkd(fmt, args...)
++#endif
+
-+/*
-+ * Read console content from Xen buffer ring.
-+ */
-+#define XEN_SYSCTL_readconsole 1
-+struct xen_sysctl_readconsole {
-+ /* IN variables. */
-+ uint32_t clear; /* Non-zero -> clear after reading. */
-+ XEN_GUEST_HANDLE_64(char) buffer; /* Buffer start */
-+ /* IN/OUT variables. */
-+ uint32_t count; /* In: Buffer size; Out: Used buffer size */
-+};
-+typedef struct xen_sysctl_readconsole xen_sysctl_readconsole_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_sysctl_readconsole_t);
++/* default ssid reference value if not supplied */
++#define ACM_DEFAULT_SSID 0x0
++#define ACM_DEFAULT_LOCAL_SSID 0x0
+
-+/* Get trace buffers machine base address */
-+#define XEN_SYSCTL_tbuf_op 2
-+struct xen_sysctl_tbuf_op {
-+ /* IN variables */
-+#define XEN_SYSCTL_TBUFOP_get_info 0
-+#define XEN_SYSCTL_TBUFOP_set_cpu_mask 1
-+#define XEN_SYSCTL_TBUFOP_set_evt_mask 2
-+#define XEN_SYSCTL_TBUFOP_set_size 3
-+#define XEN_SYSCTL_TBUFOP_enable 4
-+#define XEN_SYSCTL_TBUFOP_disable 5
-+ uint32_t cmd;
-+ /* IN/OUT variables */
-+ struct xenctl_cpumap cpu_mask;
-+ uint32_t evt_mask;
-+ /* OUT variables */
-+ uint64_aligned_t buffer_mfn;
-+ uint32_t size;
-+};
-+typedef struct xen_sysctl_tbuf_op xen_sysctl_tbuf_op_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_sysctl_tbuf_op_t);
++/* Internal ACM ERROR types */
++#define ACM_OK 0
++#define ACM_UNDEF -1
++#define ACM_INIT_SSID_ERROR -2
++#define ACM_INIT_SOID_ERROR -3
++#define ACM_ERROR -4
+
-+/*
-+ * Get physical information about the host machine
-+ */
-+#define XEN_SYSCTL_physinfo 3
-+struct xen_sysctl_physinfo {
-+ uint32_t threads_per_core;
-+ uint32_t cores_per_socket;
-+ uint32_t sockets_per_node;
-+ uint32_t nr_nodes;
-+ uint32_t cpu_khz;
-+ uint64_aligned_t total_pages;
-+ uint64_aligned_t free_pages;
-+ uint64_aligned_t scrub_pages;
-+ uint32_t hw_cap[8];
-+};
-+typedef struct xen_sysctl_physinfo xen_sysctl_physinfo_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_sysctl_physinfo_t);
++/* External ACCESS DECISIONS */
++#define ACM_ACCESS_PERMITTED 0
++#define ACM_ACCESS_DENIED -111
++#define ACM_NULL_POINTER_ERROR -200
+
+/*
-+ * Get the ID of the current scheduler.
-+ */
-+#define XEN_SYSCTL_sched_id 4
-+struct xen_sysctl_sched_id {
-+ /* OUT variable */
-+ uint32_t sched_id;
-+};
-+typedef struct xen_sysctl_sched_id xen_sysctl_sched_id_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_sysctl_sched_id_t);
++ Error codes reported in when trying to test for a new policy
++ These error codes are reported in an array of tuples where
++ each error code is followed by a parameter describing the error
++ more closely, such as a domain id.
++*/
++#define ACM_EVTCHN_SHARING_VIOLATION 0x100
++#define ACM_GNTTAB_SHARING_VIOLATION 0x101
++#define ACM_DOMAIN_LOOKUP 0x102
++#define ACM_CHWALL_CONFLICT 0x103
++#define ACM_SSIDREF_IN_USE 0x104
+
-+/* Interface for controlling Xen software performance counters. */
-+#define XEN_SYSCTL_perfc_op 5
-+/* Sub-operations: */
-+#define XEN_SYSCTL_PERFCOP_reset 1 /* Reset all counters to zero. */
-+#define XEN_SYSCTL_PERFCOP_query 2 /* Get perfctr information. */
-+struct xen_sysctl_perfc_desc {
-+ char name[80]; /* name of perf counter */
-+ uint32_t nr_vals; /* number of values for this counter */
-+};
-+typedef struct xen_sysctl_perfc_desc xen_sysctl_perfc_desc_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_sysctl_perfc_desc_t);
-+typedef uint32_t xen_sysctl_perfc_val_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_sysctl_perfc_val_t);
+
-+struct xen_sysctl_perfc_op {
-+ /* IN variables. */
-+ uint32_t cmd; /* XEN_SYSCTL_PERFCOP_??? */
-+ /* OUT variables. */
-+ uint32_t nr_counters; /* number of counters description */
-+ uint32_t nr_vals; /* number of values */
-+ /* counter information (or NULL) */
-+ XEN_GUEST_HANDLE_64(xen_sysctl_perfc_desc_t) desc;
-+ /* counter values (or NULL) */
-+ XEN_GUEST_HANDLE_64(xen_sysctl_perfc_val_t) val;
-+};
-+typedef struct xen_sysctl_perfc_op xen_sysctl_perfc_op_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_sysctl_perfc_op_t);
++/* primary policy in lower 4 bits */
++#define ACM_NULL_POLICY 0
++#define ACM_CHINESE_WALL_POLICY 1
++#define ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY 2
++#define ACM_POLICY_UNDEFINED 15
+
-+#define XEN_SYSCTL_getdomaininfolist 6
-+struct xen_sysctl_getdomaininfolist {
-+ /* IN variables. */
-+ domid_t first_domain;
-+ uint32_t max_domains;
-+ XEN_GUEST_HANDLE_64(xen_domctl_getdomaininfo_t) buffer;
-+ /* OUT variables. */
-+ uint32_t num_domains;
-+};
-+typedef struct xen_sysctl_getdomaininfolist xen_sysctl_getdomaininfolist_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_sysctl_getdomaininfolist_t);
++/* combinations have secondary policy component in higher 4bit */
++#define ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY \
++ ((ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY << 4) | ACM_CHINESE_WALL_POLICY)
+
-+/* Inject debug keys into Xen. */
-+#define XEN_SYSCTL_debug_keys 7
-+struct xen_sysctl_debug_keys {
-+ /* IN variables. */
-+ XEN_GUEST_HANDLE_64(char) keys;
-+ uint32_t nr_keys;
-+};
-+typedef struct xen_sysctl_debug_keys xen_sysctl_debug_keys_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_sysctl_debug_keys_t);
++/* policy: */
++#define ACM_POLICY_NAME(X) \
++ ((X) == (ACM_NULL_POLICY)) ? "NULL" : \
++ ((X) == (ACM_CHINESE_WALL_POLICY)) ? "CHINESE WALL" : \
++ ((X) == (ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY)) ? "SIMPLE TYPE ENFORCEMENT" : \
++ ((X) == (ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY)) ? "CHINESE WALL AND SIMPLE TYPE ENFORCEMENT" : \
++ "UNDEFINED"
+
-+/* Get physical CPU information. */
-+#define XEN_SYSCTL_getcpuinfo 8
-+struct xen_sysctl_cpuinfo {
-+ uint64_t idletime;
-+};
-+typedef struct xen_sysctl_cpuinfo xen_sysctl_cpuinfo_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_sysctl_cpuinfo_t);
-+struct xen_sysctl_getcpuinfo {
-+ /* IN variables. */
-+ uint32_t max_cpus;
-+ XEN_GUEST_HANDLE_64(xen_sysctl_cpuinfo_t) info;
-+ /* OUT variables. */
-+ uint32_t nr_cpus;
-+};
-+typedef struct xen_sysctl_getcpuinfo xen_sysctl_getcpuinfo_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_sysctl_getcpuinfo_t);
++/* the following policy versions must be increased
++ * whenever the interpretation of the related
++ * policy's data structure changes
++ */
++#define ACM_POLICY_VERSION 4
++#define ACM_CHWALL_VERSION 1
++#define ACM_STE_VERSION 1
+
-+struct xen_sysctl {
-+ uint32_t cmd;
-+ uint32_t interface_version; /* XEN_SYSCTL_INTERFACE_VERSION */
-+ union {
-+ struct xen_sysctl_readconsole readconsole;
-+ struct xen_sysctl_tbuf_op tbuf_op;
-+ struct xen_sysctl_physinfo physinfo;
-+ struct xen_sysctl_sched_id sched_id;
-+ struct xen_sysctl_perfc_op perfc_op;
-+ struct xen_sysctl_getdomaininfolist getdomaininfolist;
-+ struct xen_sysctl_debug_keys debug_keys;
-+ struct xen_sysctl_getcpuinfo getcpuinfo;
-+ uint8_t pad[128];
-+ } u;
-+};
-+typedef struct xen_sysctl xen_sysctl_t;
-+DEFINE_XEN_GUEST_HANDLE(xen_sysctl_t);
++/* defines a ssid reference used by xen */
++typedef uint32_t ssidref_t;
+
-+#endif /* __XEN_PUBLIC_SYSCTL_H__ */
++/* hooks that are known to domains */
++#define ACMHOOK_none 0
++#define ACMHOOK_sharing 1
++#define ACMHOOK_authorization 2
++#define ACMHOOK_conflictset 3
+
-+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
-+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/trace.h tmp-linux-2.6-xen.patch/include/xen/interface/trace.h
---- pristine-linux-2.6.18.2/include/xen/interface/trace.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/trace.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,120 @@
-+/******************************************************************************
-+ * include/public/trace.h
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
++/* -------security policy relevant type definitions-------- */
++
++/* type identifier; compares to "equal" or "not equal" */
++typedef uint16_t domaintype_t;
++
++/* CHINESE WALL POLICY DATA STRUCTURES
+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
++ * current accumulated conflict type set:
++ * When a domain is started and has a type that is in
++ * a conflict set, the conflicting types are incremented in
++ * the aggregate set. When a domain is destroyed, the
++ * conflicting types to its type are decremented.
++ * If a domain has multiple types, this procedure works over
++ * all those types.
+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
++ * conflict_aggregate_set[i] holds the number of
++ * running domains that have a conflict with type i.
+ *
-+ * Mark Williamson, (C) 2004 Intel Research Cambridge
-+ * Copyright (C) 2005 Bin Ren
++ * running_types[i] holds the number of running domains
++ * that include type i in their ssidref-referenced type set
++ *
++ * conflict_sets[i][j] is "0" if type j has no conflict
++ * with type i and is "1" otherwise.
+ */
++/* high-16 = version, low-16 = check magic */
++#define ACM_MAGIC 0x0001debc
+
-+#ifndef __XEN_PUBLIC_TRACE_H__
-+#define __XEN_PUBLIC_TRACE_H__
++/* size of the SHA1 hash identifying the XML policy from which the
++ binary policy was created */
++#define ACM_SHA1_HASH_SIZE 20
+
-+/* Trace classes */
-+#define TRC_CLS_SHIFT 16
-+#define TRC_GEN 0x0001f000 /* General trace */
-+#define TRC_SCHED 0x0002f000 /* Xen Scheduler trace */
-+#define TRC_DOM0OP 0x0004f000 /* Xen DOM0 operation trace */
-+#define TRC_HVM 0x0008f000 /* Xen HVM trace */
-+#define TRC_MEM 0x0010f000 /* Xen memory trace */
-+#define TRC_ALL 0xfffff000
++/* each offset in bytes from start of the struct they
++ * are part of */
+
-+/* Trace subclasses */
-+#define TRC_SUBCLS_SHIFT 12
++/* V3 of the policy buffer aded a version structure */
++struct acm_policy_version
++{
++ uint32_t major;
++ uint32_t minor;
++};
+
-+/* trace subclasses for SVM */
-+#define TRC_HVM_ENTRYEXIT 0x00081000 /* VMENTRY and #VMEXIT */
-+#define TRC_HVM_HANDLER 0x00082000 /* various HVM handlers */
+
-+/* Trace events per class */
-+#define TRC_LOST_RECORDS (TRC_GEN + 1)
++/* each buffer consists of all policy information for
++ * the respective policy given in the policy code
++ *
++ * acm_policy_buffer, acm_chwall_policy_buffer,
++ * and acm_ste_policy_buffer need to stay 32-bit aligned
++ * because we create binary policies also with external
++ * tools that assume packed representations (e.g. the java tool)
++ */
++struct acm_policy_buffer {
++ uint32_t magic;
++ uint32_t policy_version; /* ACM_POLICY_VERSION */
++ uint32_t len;
++ uint32_t policy_reference_offset;
++ uint32_t primary_policy_code;
++ uint32_t primary_buffer_offset;
++ uint32_t secondary_policy_code;
++ uint32_t secondary_buffer_offset;
++ struct acm_policy_version xml_pol_version; /* add in V3 */
++ uint8_t xml_policy_hash[ACM_SHA1_HASH_SIZE]; /* added in V4 */
++};
+
-+#define TRC_SCHED_DOM_ADD (TRC_SCHED + 1)
-+#define TRC_SCHED_DOM_REM (TRC_SCHED + 2)
-+#define TRC_SCHED_SLEEP (TRC_SCHED + 3)
-+#define TRC_SCHED_WAKE (TRC_SCHED + 4)
-+#define TRC_SCHED_YIELD (TRC_SCHED + 5)
-+#define TRC_SCHED_BLOCK (TRC_SCHED + 6)
-+#define TRC_SCHED_SHUTDOWN (TRC_SCHED + 7)
-+#define TRC_SCHED_CTL (TRC_SCHED + 8)
-+#define TRC_SCHED_ADJDOM (TRC_SCHED + 9)
-+#define TRC_SCHED_SWITCH (TRC_SCHED + 10)
-+#define TRC_SCHED_S_TIMER_FN (TRC_SCHED + 11)
-+#define TRC_SCHED_T_TIMER_FN (TRC_SCHED + 12)
-+#define TRC_SCHED_DOM_TIMER_FN (TRC_SCHED + 13)
-+#define TRC_SCHED_SWITCH_INFPREV (TRC_SCHED + 14)
-+#define TRC_SCHED_SWITCH_INFNEXT (TRC_SCHED + 15)
+
-+#define TRC_MEM_PAGE_GRANT_MAP (TRC_MEM + 1)
-+#define TRC_MEM_PAGE_GRANT_UNMAP (TRC_MEM + 2)
-+#define TRC_MEM_PAGE_GRANT_TRANSFER (TRC_MEM + 3)
++struct acm_policy_reference_buffer {
++ uint32_t len;
++};
++
++struct acm_chwall_policy_buffer {
++ uint32_t policy_version; /* ACM_CHWALL_VERSION */
++ uint32_t policy_code;
++ uint32_t chwall_max_types;
++ uint32_t chwall_max_ssidrefs;
++ uint32_t chwall_max_conflictsets;
++ uint32_t chwall_ssid_offset;
++ uint32_t chwall_conflict_sets_offset;
++ uint32_t chwall_running_types_offset;
++ uint32_t chwall_conflict_aggregate_offset;
++};
++
++struct acm_ste_policy_buffer {
++ uint32_t policy_version; /* ACM_STE_VERSION */
++ uint32_t policy_code;
++ uint32_t ste_max_types;
++ uint32_t ste_max_ssidrefs;
++ uint32_t ste_ssid_offset;
++};
+
-+/* trace events per subclass */
-+#define TRC_HVM_VMENTRY (TRC_HVM_ENTRYEXIT + 0x01)
-+#define TRC_HVM_VMEXIT (TRC_HVM_ENTRYEXIT + 0x02)
-+#define TRC_HVM_PF_XEN (TRC_HVM_HANDLER + 0x01)
-+#define TRC_HVM_PF_INJECT (TRC_HVM_HANDLER + 0x02)
-+#define TRC_HVM_INJ_EXC (TRC_HVM_HANDLER + 0x03)
-+#define TRC_HVM_INJ_VIRQ (TRC_HVM_HANDLER + 0x04)
-+#define TRC_HVM_REINJ_VIRQ (TRC_HVM_HANDLER + 0x05)
-+#define TRC_HVM_IO_READ (TRC_HVM_HANDLER + 0x06)
-+#define TRC_HVM_IO_WRITE (TRC_HVM_HANDLER + 0x07)
-+#define TRC_HVM_CR_READ (TRC_HVM_HANDLER + 0x08)
-+#define TRC_HVM_CR_WRITE (TRC_HVM_HANDLER + 0x09)
-+#define TRC_HVM_DR_READ (TRC_HVM_HANDLER + 0x0A)
-+#define TRC_HVM_DR_WRITE (TRC_HVM_HANDLER + 0x0B)
-+#define TRC_HVM_MSR_READ (TRC_HVM_HANDLER + 0x0C)
-+#define TRC_HVM_MSR_WRITE (TRC_HVM_HANDLER + 0x0D)
-+#define TRC_HVM_CPUID (TRC_HVM_HANDLER + 0x0E)
-+#define TRC_HVM_INTR (TRC_HVM_HANDLER + 0x0F)
-+#define TRC_HVM_NMI (TRC_HVM_HANDLER + 0x10)
-+#define TRC_HVM_SMI (TRC_HVM_HANDLER + 0x11)
-+#define TRC_HVM_VMMCALL (TRC_HVM_HANDLER + 0x12)
-+#define TRC_HVM_HLT (TRC_HVM_HANDLER + 0x13)
-+#define TRC_HVM_INVLPG (TRC_HVM_HANDLER + 0x14)
-+#define TRC_HVM_MCE (TRC_HVM_HANDLER + 0x15)
++struct acm_stats_buffer {
++ uint32_t magic;
++ uint32_t len;
++ uint32_t primary_policy_code;
++ uint32_t primary_stats_offset;
++ uint32_t secondary_policy_code;
++ uint32_t secondary_stats_offset;
++};
+
-+/* This structure represents a single trace buffer record. */
-+struct t_rec {
-+ uint64_t cycles; /* cycle counter timestamp */
-+ uint32_t event; /* event ID */
-+ unsigned long data[5]; /* event data items */
++struct acm_ste_stats_buffer {
++ uint32_t ec_eval_count;
++ uint32_t gt_eval_count;
++ uint32_t ec_denied_count;
++ uint32_t gt_denied_count;
++ uint32_t ec_cachehit_count;
++ uint32_t gt_cachehit_count;
+};
+
-+/*
-+ * This structure contains the metadata for a single trace buffer. The head
-+ * field, indexes into an array of struct t_rec's.
-+ */
-+struct t_buf {
-+ uint32_t cons; /* Next item to be consumed by control tools. */
-+ uint32_t prod; /* Next item to be produced by Xen. */
-+ /* 'nr_recs' records follow immediately after the meta-data header. */
++struct acm_ssid_buffer {
++ uint32_t len;
++ ssidref_t ssidref;
++ uint32_t policy_reference_offset;
++ uint32_t primary_policy_code;
++ uint32_t primary_max_types;
++ uint32_t primary_types_offset;
++ uint32_t secondary_policy_code;
++ uint32_t secondary_max_types;
++ uint32_t secondary_types_offset;
+};
+
-+#endif /* __XEN_PUBLIC_TRACE_H__ */
++#endif
+
+/*
+ * Local variables:
@@ -104094,15 +143919,13 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/vcpu.h tmp-linux-2.6-xen.patch/include/xen/interface/vcpu.h
---- pristine-linux-2.6.18.2/include/xen/interface/vcpu.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/vcpu.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,192 @@
-+/******************************************************************************
-+ * vcpu.h
-+ *
-+ * VCPU initialisation, query, and hotplug.
-+ *
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/xsm/acm_ops.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/xsm/acm_ops.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,159 @@
++/*
++ * acm_ops.h: Xen access control module hypervisor commands
++ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
@@ -104121,165 +143944,134 @@
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
-+ * Copyright (c) 2005, Keir Fraser <keir at xensource.com>
++ * Reiner Sailer <sailer at watson.ibm.com>
++ * Copyright (c) 2005,2006 International Business Machines Corporation.
+ */
+
-+#ifndef __XEN_PUBLIC_VCPU_H__
-+#define __XEN_PUBLIC_VCPU_H__
++#ifndef __XEN_PUBLIC_ACM_OPS_H__
++#define __XEN_PUBLIC_ACM_OPS_H__
+
-+/*
-+ * Prototype for this hypercall is:
-+ * int vcpu_op(int cmd, int vcpuid, void *extra_args)
-+ * @cmd == VCPUOP_??? (VCPU operation).
-+ * @vcpuid == VCPU to operate on.
-+ * @extra_args == Operation-specific extra arguments (NULL if none).
-+ */
++#include "../xen.h"
++#include "acm.h"
+
+/*
-+ * Initialise a VCPU. Each VCPU can be initialised only once. A
-+ * newly-initialised VCPU will not run until it is brought up by VCPUOP_up.
-+ *
-+ * @extra_arg == pointer to vcpu_guest_context structure containing initial
-+ * state for the VCPU.
++ * Make sure you increment the interface version whenever you modify this file!
++ * This makes sure that old versions of acm tools will stop working in a
++ * well-defined way (rather than crashing the machine, for instance).
+ */
-+#define VCPUOP_initialise 0
++#define ACM_INTERFACE_VERSION 0xAAAA000A
+
-+/*
-+ * Bring up a VCPU. This makes the VCPU runnable. This operation will fail
-+ * if the VCPU has not been initialised (VCPUOP_initialise).
-+ */
-+#define VCPUOP_up 1
++/************************************************************************/
+
+/*
-+ * Bring down a VCPU (i.e., make it non-runnable).
-+ * There are a few caveats that callers should observe:
-+ * 1. This operation may return, and VCPU_is_up may return false, before the
-+ * VCPU stops running (i.e., the command is asynchronous). It is a good
-+ * idea to ensure that the VCPU has entered a non-critical loop before
-+ * bringing it down. Alternatively, this operation is guaranteed
-+ * synchronous if invoked by the VCPU itself.
-+ * 2. After a VCPU is initialised, there is currently no way to drop all its
-+ * references to domain memory. Even a VCPU that is down still holds
-+ * memory references via its pagetable base pointer and GDT. It is good
-+ * practise to move a VCPU onto an 'idle' or default page table, LDT and
-+ * GDT before bringing it down.
++ * Prototype for this hypercall is:
++ * int acm_op(int cmd, void *args)
++ * @cmd == ACMOP_??? (access control module operation).
++ * @args == Operation-specific extra arguments (NULL if none).
+ */
-+#define VCPUOP_down 2
+
-+/* Returns 1 if the given VCPU is up. */
-+#define VCPUOP_is_up 3
+
-+/*
-+ * Return information about the state and running time of a VCPU.
-+ * @extra_arg == pointer to vcpu_runstate_info structure.
-+ */
-+#define VCPUOP_get_runstate_info 4
-+struct vcpu_runstate_info {
-+ /* VCPU's current state (RUNSTATE_*). */
-+ int state;
-+ /* When was current state entered (system time, ns)? */
-+ uint64_t state_entry_time;
-+ /*
-+ * Time spent in each RUNSTATE_* (ns). The sum of these times is
-+ * guaranteed not to drift from system time.
-+ */
-+ uint64_t time[4];
++#define ACMOP_setpolicy 1
++struct acm_setpolicy {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) pushcache;
++ uint32_t pushcache_size;
+};
-+typedef struct vcpu_runstate_info vcpu_runstate_info_t;
-+DEFINE_XEN_GUEST_HANDLE(vcpu_runstate_info_t);
+
-+/* VCPU is currently running on a physical CPU. */
-+#define RUNSTATE_running 0
+
-+/* VCPU is runnable, but not currently scheduled on any physical CPU. */
-+#define RUNSTATE_runnable 1
++#define ACMOP_getpolicy 2
++struct acm_getpolicy {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) pullcache;
++ uint32_t pullcache_size;
++};
+
-+/* VCPU is blocked (a.k.a. idle). It is therefore not runnable. */
-+#define RUNSTATE_blocked 2
+
-+/*
-+ * VCPU is not runnable, but it is not blocked.
-+ * This is a 'catch all' state for things like hotplug and pauses by the
-+ * system administrator (or for critical sections in the hypervisor).
-+ * RUNSTATE_blocked dominates this state (it is the preferred state).
-+ */
-+#define RUNSTATE_offline 3
++#define ACMOP_dumpstats 3
++struct acm_dumpstats {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) pullcache;
++ uint32_t pullcache_size;
++};
+
-+/*
-+ * Register a shared memory area from which the guest may obtain its own
-+ * runstate information without needing to execute a hypercall.
-+ * Notes:
-+ * 1. The registered address may be virtual or physical or guest handle,
-+ * depending on the platform. Virtual address or guest handle should be
-+ * registered on x86 systems.
-+ * 2. Only one shared area may be registered per VCPU. The shared area is
-+ * updated by the hypervisor each time the VCPU is scheduled. Thus
-+ * runstate.state will always be RUNSTATE_running and
-+ * runstate.state_entry_time will indicate the system time at which the
-+ * VCPU was last scheduled to run.
-+ * @extra_arg == pointer to vcpu_register_runstate_memory_area structure.
-+ */
-+#define VCPUOP_register_runstate_memory_area 5
-+struct vcpu_register_runstate_memory_area {
++
++#define ACMOP_getssid 4
++#define ACM_GETBY_ssidref 1
++#define ACM_GETBY_domainid 2
++struct acm_getssid {
++ /* IN */
++ uint32_t get_ssid_by; /* ACM_GETBY_* */
+ union {
-+ XEN_GUEST_HANDLE(vcpu_runstate_info_t) h;
-+ struct vcpu_runstate_info *v;
-+ uint64_t p;
-+ } addr;
++ domaintype_t domainid;
++ ssidref_t ssidref;
++ } id;
++ XEN_GUEST_HANDLE_64(void) ssidbuf;
++ uint32_t ssidbuf_size;
+};
-+typedef struct vcpu_register_runstate_memory_area vcpu_register_runstate_memory_area_t;
-+DEFINE_XEN_GUEST_HANDLE(vcpu_register_runstate_memory_area_t);
+
-+/*
-+ * Set or stop a VCPU's periodic timer. Every VCPU has one periodic timer
-+ * which can be set via these commands. Periods smaller than one millisecond
-+ * may not be supported.
-+ */
-+#define VCPUOP_set_periodic_timer 6 /* arg == vcpu_set_periodic_timer_t */
-+#define VCPUOP_stop_periodic_timer 7 /* arg == NULL */
-+struct vcpu_set_periodic_timer {
-+ uint64_t period_ns;
++#define ACMOP_getdecision 5
++struct acm_getdecision {
++ /* IN */
++ uint32_t get_decision_by1; /* ACM_GETBY_* */
++ uint32_t get_decision_by2; /* ACM_GETBY_* */
++ union {
++ domaintype_t domainid;
++ ssidref_t ssidref;
++ } id1;
++ union {
++ domaintype_t domainid;
++ ssidref_t ssidref;
++ } id2;
++ uint32_t hook;
++ /* OUT */
++ uint32_t acm_decision;
+};
-+typedef struct vcpu_set_periodic_timer vcpu_set_periodic_timer_t;
-+DEFINE_XEN_GUEST_HANDLE(vcpu_set_periodic_timer_t);
+
-+/*
-+ * Set or stop a VCPU's single-shot timer. Every VCPU has one single-shot
-+ * timer which can be set via these commands.
-+ */
-+#define VCPUOP_set_singleshot_timer 8 /* arg == vcpu_set_singleshot_timer_t */
-+#define VCPUOP_stop_singleshot_timer 9 /* arg == NULL */
-+struct vcpu_set_singleshot_timer {
-+ uint64_t timeout_abs_ns; /* Absolute system time value in nanoseconds. */
-+ uint32_t flags; /* VCPU_SSHOTTMR_??? */
++
++#define ACMOP_chgpolicy 6
++struct acm_change_policy {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) policy_pushcache;
++ uint32_t policy_pushcache_size;
++ XEN_GUEST_HANDLE_64(void) del_array;
++ uint32_t delarray_size;
++ XEN_GUEST_HANDLE_64(void) chg_array;
++ uint32_t chgarray_size;
++ /* OUT */
++ /* array with error code */
++ XEN_GUEST_HANDLE_64(void) err_array;
++ uint32_t errarray_size;
+};
-+typedef struct vcpu_set_singleshot_timer vcpu_set_singleshot_timer_t;
-+DEFINE_XEN_GUEST_HANDLE(vcpu_set_singleshot_timer_t);
+
-+/* Flags to VCPUOP_set_singleshot_timer. */
-+ /* Require the timeout to be in the future (return -ETIME if it's passed). */
-+#define _VCPU_SSHOTTMR_future (0)
-+#define VCPU_SSHOTTMR_future (1U << _VCPU_SSHOTTMR_future)
++#define ACMOP_relabeldoms 7
++struct acm_relabel_doms {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) relabel_map;
++ uint32_t relabel_map_size;
++ /* OUT */
++ XEN_GUEST_HANDLE_64(void) err_array;
++ uint32_t errarray_size;
++};
+
-+/*
-+ * Register a memory location in the guest address space for the
-+ * vcpu_info structure. This allows the guest to place the vcpu_info
-+ * structure in a convenient place, such as in a per-cpu data area.
-+ * The pointer need not be page aligned, but the structure must not
-+ * cross a page boundary.
-+ *
-+ * This may be called only once per vcpu.
-+ */
-+#define VCPUOP_register_vcpu_info 10 /* arg == struct vcpu_info */
-+struct vcpu_register_vcpu_info {
-+ uint64_t mfn; /* mfn of page to place vcpu_info */
-+ uint32_t offset; /* offset within page */
-+ uint32_t rsvd; /* unused */
++/* future interface to Xen */
++struct xen_acmctl {
++ uint32_t cmd;
++ uint32_t interface_version;
++ union {
++ struct acm_setpolicy setpolicy;
++ struct acm_getpolicy getpolicy;
++ struct acm_dumpstats dumpstats;
++ struct acm_getssid getssid;
++ struct acm_getdecision getdecision;
++ struct acm_change_policy change_policy;
++ struct acm_relabel_doms relabel_doms;
++ } u;
+};
-+typedef struct vcpu_register_vcpu_info vcpu_register_vcpu_info_t;
-+DEFINE_XEN_GUEST_HANDLE(vcpu_register_vcpu_info_t);
+
-+#endif /* __XEN_PUBLIC_VCPU_H__ */
++typedef struct xen_acmctl xen_acmctl_t;
++DEFINE_XEN_GUEST_HANDLE(xen_acmctl_t);
++
++#endif /* __XEN_PUBLIC_ACM_OPS_H__ */
+
+/*
+ * Local variables:
@@ -104290,2036 +144082,2077 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/version.h tmp-linux-2.6-xen.patch/include/xen/interface/version.h
---- pristine-linux-2.6.18.2/include/xen/interface/version.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/version.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,91 @@
-+/******************************************************************************
-+ * version.h
-+ *
-+ * Xen version, type, and compile information.
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/interface/xsm/flask_op.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/interface/xsm/flask_op.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,45 @@
++/*
++ * This file contains the flask_op hypercall commands and definitions.
+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
++ * Author: George Coker, <gscoker at alpha.ncsc.mil>
+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License version 2,
++ * as published by the Free Software Foundation.
++ */
++
++#ifndef __FLASK_OP_H__
++#define __FLASK_OP_H__
++
++#define FLASK_LOAD 1
++#define FLASK_GETENFORCE 2
++#define FLASK_SETENFORCE 3
++#define FLASK_CONTEXT_TO_SID 4
++#define FLASK_SID_TO_CONTEXT 5
++#define FLASK_ACCESS 6
++#define FLASK_CREATE 7
++#define FLASK_RELABEL 8
++#define FLASK_USER 9
++#define FLASK_POLICYVERS 10
++#define FLASK_GETBOOL 11
++#define FLASK_SETBOOL 12
++#define FLASK_COMMITBOOLS 13
++#define FLASK_MLS 14
++#define FLASK_DISABLE 15
++#define FLASK_GETAVC_THRESHOLD 16
++#define FLASK_SETAVC_THRESHOLD 17
++#define FLASK_AVC_HASHSTATS 18
++#define FLASK_AVC_CACHESTATS 19
++#define FLASK_MEMBER 20
++
++#define FLASK_LAST FLASK_MEMBER
++
++typedef struct flask_op {
++ uint32_t cmd;
++ uint32_t size;
++ char *buf;
++} flask_op_t;
++
++DEFINE_XEN_GUEST_HANDLE(flask_op_t);
++
++#endif
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/pcifront.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/pcifront.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,83 @@
++/*
++ * PCI Frontend - arch-dependendent declarations
+ *
-+ * Copyright (c) 2005, Nguyen Anh Quynh <aquynh at gmail.com>
-+ * Copyright (c) 2005, Keir Fraser <keir at xensource.com>
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
+ */
++#ifndef __XEN_ASM_PCIFRONT_H__
++#define __XEN_ASM_PCIFRONT_H__
+
-+#ifndef __XEN_PUBLIC_VERSION_H__
-+#define __XEN_PUBLIC_VERSION_H__
++#include <linux/spinlock.h>
+
-+/* NB. All ops return zero on success, except XENVER_{version,pagesize} */
++#ifdef __KERNEL__
+
-+/* arg == NULL; returns major:minor (16:16). */
-+#define XENVER_version 0
++#ifndef __ia64__
+
-+/* arg == xen_extraversion_t. */
-+#define XENVER_extraversion 1
-+typedef char xen_extraversion_t[16];
-+#define XEN_EXTRAVERSION_LEN (sizeof(xen_extraversion_t))
++struct pcifront_device;
++struct pci_bus;
+
-+/* arg == xen_compile_info_t. */
-+#define XENVER_compile_info 2
-+struct xen_compile_info {
-+ char compiler[64];
-+ char compile_by[16];
-+ char compile_domain[32];
-+ char compile_date[32];
++struct pcifront_sd {
++ int domain;
++ struct pcifront_device *pdev;
+};
-+typedef struct xen_compile_info xen_compile_info_t;
-+
-+#define XENVER_capabilities 3
-+typedef char xen_capabilities_info_t[1024];
-+#define XEN_CAPABILITIES_INFO_LEN (sizeof(xen_capabilities_info_t))
+
-+#define XENVER_changeset 4
-+typedef char xen_changeset_info_t[64];
-+#define XEN_CHANGESET_INFO_LEN (sizeof(xen_changeset_info_t))
++static inline struct pcifront_device *
++pcifront_get_pdev(struct pcifront_sd *sd)
++{
++ return sd->pdev;
++}
+
-+#define XENVER_platform_parameters 5
-+struct xen_platform_parameters {
-+ unsigned long virt_start;
-+};
-+typedef struct xen_platform_parameters xen_platform_parameters_t;
++static inline void pcifront_init_sd(struct pcifront_sd *sd,
++ unsigned int domain, unsigned int bus,
++ struct pcifront_device *pdev)
++{
++ sd->domain = domain;
++ sd->pdev = pdev;
++}
+
-+#define XENVER_get_features 6
-+struct xen_feature_info {
-+ unsigned int submap_idx; /* IN: which 32-bit submap to return */
-+ uint32_t submap; /* OUT: 32-bit submap */
-+};
-+typedef struct xen_feature_info xen_feature_info_t;
++#if defined(CONFIG_PCI_DOMAINS)
++static inline int pci_domain_nr(struct pci_bus *bus)
++{
++ struct pcifront_sd *sd = bus->sysdata;
++ return sd->domain;
++}
++static inline int pci_proc_domain(struct pci_bus *bus)
++{
++ return pci_domain_nr(bus);
++}
++#endif /* CONFIG_PCI_DOMAINS */
+
-+/* Declares the features reported by XENVER_get_features. */
-+#include "features.h"
++static inline void pcifront_setup_root_resources(struct pci_bus *bus,
++ struct pcifront_sd *sd)
++{
++}
+
-+/* arg == NULL; returns host memory page size. */
-+#define XENVER_pagesize 7
++#else /* __ia64__ */
+
-+/* arg == xen_domain_handle_t. */
-+#define XENVER_guest_handle 8
++#include <linux/acpi.h>
++#include <asm/pci.h>
++#define pcifront_sd pci_controller
+
-+#endif /* __XEN_PUBLIC_VERSION_H__ */
++extern void xen_add_resource(struct pci_controller *, unsigned int,
++ unsigned int, struct acpi_resource *);
++extern void xen_pcibios_setup_root_windows(struct pci_bus *,
++ struct pci_controller *);
+
-+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
-+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/xencomm.h tmp-linux-2.6-xen.patch/include/xen/interface/xencomm.h
---- pristine-linux-2.6.18.2/include/xen/interface/xencomm.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/xencomm.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,41 @@
-+/*
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Copyright (C) IBM Corp. 2006
-+ */
++static inline struct pcifront_device *
++pcifront_get_pdev(struct pcifront_sd *sd)
++{
++ return (struct pcifront_device *)sd->platform_data;
++}
+
-+#ifndef _XEN_XENCOMM_H_
-+#define _XEN_XENCOMM_H_
++static inline void pcifront_setup_root_resources(struct pci_bus *bus,
++ struct pcifront_sd *sd)
++{
++ xen_pcibios_setup_root_windows(bus, sd);
++}
+
-+/* A xencomm descriptor is a scatter/gather list containing physical
-+ * addresses corresponding to a virtually contiguous memory area. The
-+ * hypervisor translates these physical addresses to machine addresses to copy
-+ * to and from the virtually contiguous area.
-+ */
++#endif /* __ia64__ */
+
-+#define XENCOMM_MAGIC 0x58434F4D /* 'XCOM' */
-+#define XENCOMM_INVALID (~0UL)
++extern struct rw_semaphore pci_bus_sem;
+
-+struct xencomm_desc {
-+ uint32_t magic;
-+ uint32_t nr_addrs; /* the number of entries in address[] */
-+ uint64_t address[0];
-+};
++#endif /* __KERNEL__ */
+
-+#endif /* _XEN_XENCOMM_H_ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/xen-compat.h tmp-linux-2.6-xen.patch/include/xen/interface/xen-compat.h
---- pristine-linux-2.6.18.2/include/xen/interface/xen-compat.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/xen-compat.h 2007-07-30 16:35:28.000000000 +0200
-@@ -0,0 +1,51 @@
++#endif /* __XEN_ASM_PCIFRONT_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/public/evtchn.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/public/evtchn.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,88 @@
+/******************************************************************************
-+ * xen-compat.h
++ * evtchn.h
+ *
-+ * Guest OS interface to Xen. Compatibility layer.
++ * Interface to /dev/xen/evtchn.
++ *
++ * Copyright (c) 2003-2005, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
-+ *
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
-+ *
++ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Copyright (c) 2006, Christian Limpach
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
+
-+#ifndef __XEN_PUBLIC_XEN_COMPAT_H__
-+#define __XEN_PUBLIC_XEN_COMPAT_H__
++#ifndef __LINUX_PUBLIC_EVTCHN_H__
++#define __LINUX_PUBLIC_EVTCHN_H__
+
-+#define __XEN_LATEST_INTERFACE_VERSION__ 0x00030205
++/*
++ * Bind a fresh port to VIRQ @virq.
++ * Return allocated port.
++ */
++#define IOCTL_EVTCHN_BIND_VIRQ \
++ _IOC(_IOC_NONE, 'E', 0, sizeof(struct ioctl_evtchn_bind_virq))
++struct ioctl_evtchn_bind_virq {
++ unsigned int virq;
++};
+
-+#if defined(__XEN__) || defined(__XEN_TOOLS__)
-+/* Xen is built with matching headers and implements the latest interface. */
-+#define __XEN_INTERFACE_VERSION__ __XEN_LATEST_INTERFACE_VERSION__
-+#elif !defined(__XEN_INTERFACE_VERSION__)
-+/* Guests which do not specify a version get the legacy interface. */
-+#define __XEN_INTERFACE_VERSION__ 0x00000000
-+#endif
++/*
++ * Bind a fresh port to remote <@remote_domain, @remote_port>.
++ * Return allocated port.
++ */
++#define IOCTL_EVTCHN_BIND_INTERDOMAIN \
++ _IOC(_IOC_NONE, 'E', 1, sizeof(struct ioctl_evtchn_bind_interdomain))
++struct ioctl_evtchn_bind_interdomain {
++ unsigned int remote_domain, remote_port;
++};
+
-+#if __XEN_INTERFACE_VERSION__ > __XEN_LATEST_INTERFACE_VERSION__
-+#error "These header files do not support the requested interface version."
-+#endif
++/*
++ * Allocate a fresh port for binding to @remote_domain.
++ * Return allocated port.
++ */
++#define IOCTL_EVTCHN_BIND_UNBOUND_PORT \
++ _IOC(_IOC_NONE, 'E', 2, sizeof(struct ioctl_evtchn_bind_unbound_port))
++struct ioctl_evtchn_bind_unbound_port {
++ unsigned int remote_domain;
++};
+
-+/* Fields defined as a Xen guest handle since 0x00030205. */
-+#if __XEN_INTERFACE_VERSION__ >= 0x00030205
-+#define XEN_GUEST_HANDLE_00030205(type) XEN_GUEST_HANDLE(type)
-+#else
-+#define XEN_GUEST_HANDLE_00030205(type) type *
-+#endif
++/*
++ * Unbind previously allocated @port.
++ */
++#define IOCTL_EVTCHN_UNBIND \
++ _IOC(_IOC_NONE, 'E', 3, sizeof(struct ioctl_evtchn_unbind))
++struct ioctl_evtchn_unbind {
++ unsigned int port;
++};
+
-+#endif /* __XEN_PUBLIC_XEN_COMPAT_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/xen.h tmp-linux-2.6-xen.patch/include/xen/interface/xen.h
---- pristine-linux-2.6.18.2/include/xen/interface/xen.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/xen.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,612 @@
++/*
++ * Unbind previously allocated @port.
++ */
++#define IOCTL_EVTCHN_NOTIFY \
++ _IOC(_IOC_NONE, 'E', 4, sizeof(struct ioctl_evtchn_notify))
++struct ioctl_evtchn_notify {
++ unsigned int port;
++};
++
++/* Clear and reinitialise the event buffer. Clear error condition. */
++#define IOCTL_EVTCHN_RESET \
++ _IOC(_IOC_NONE, 'E', 5, 0)
++
++#endif /* __LINUX_PUBLIC_EVTCHN_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/public/gntdev.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/public/gntdev.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,119 @@
+/******************************************************************************
-+ * xen.h
++ * gntdev.h
+ *
-+ * Guest OS interface to Xen.
++ * Interface to /dev/xen/gntdev.
++ *
++ * Copyright (c) 2007, D G Murray
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
-+ *
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
-+ *
++ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Copyright (c) 2004, K A Fraser
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
+
-+#ifndef __XEN_PUBLIC_XEN_H__
-+#define __XEN_PUBLIC_XEN_H__
-+
-+#include "xen-compat.h"
++#ifndef __LINUX_PUBLIC_GNTDEV_H__
++#define __LINUX_PUBLIC_GNTDEV_H__
+
-+#if defined(__i386__) || defined(__x86_64__)
-+#include "arch-x86/xen.h"
-+#elif defined(__ia64__)
-+#include "arch-ia64.h"
-+#elif defined(__powerpc__)
-+#include "arch-powerpc.h"
-+#else
-+#error "Unsupported architecture"
-+#endif
++struct ioctl_gntdev_grant_ref {
++ /* The domain ID of the grant to be mapped. */
++ uint32_t domid;
++ /* The grant reference of the grant to be mapped. */
++ uint32_t ref;
++};
+
+/*
-+ * HYPERCALLS
++ * Inserts the grant references into the mapping table of an instance
++ * of gntdev. N.B. This does not perform the mapping, which is deferred
++ * until mmap() is called with @index as the offset.
+ */
-+
-+#define __HYPERVISOR_set_trap_table 0
-+#define __HYPERVISOR_mmu_update 1
-+#define __HYPERVISOR_set_gdt 2
-+#define __HYPERVISOR_stack_switch 3
-+#define __HYPERVISOR_set_callbacks 4
-+#define __HYPERVISOR_fpu_taskswitch 5
-+#define __HYPERVISOR_sched_op_compat 6 /* compat since 0x00030101 */
-+#define __HYPERVISOR_platform_op 7
-+#define __HYPERVISOR_set_debugreg 8
-+#define __HYPERVISOR_get_debugreg 9
-+#define __HYPERVISOR_update_descriptor 10
-+#define __HYPERVISOR_memory_op 12
-+#define __HYPERVISOR_multicall 13
-+#define __HYPERVISOR_update_va_mapping 14
-+#define __HYPERVISOR_set_timer_op 15
-+#define __HYPERVISOR_event_channel_op_compat 16 /* compat since 0x00030202 */
-+#define __HYPERVISOR_xen_version 17
-+#define __HYPERVISOR_console_io 18
-+#define __HYPERVISOR_physdev_op_compat 19 /* compat since 0x00030202 */
-+#define __HYPERVISOR_grant_table_op 20
-+#define __HYPERVISOR_vm_assist 21
-+#define __HYPERVISOR_update_va_mapping_otherdomain 22
-+#define __HYPERVISOR_iret 23 /* x86 only */
-+#define __HYPERVISOR_vcpu_op 24
-+#define __HYPERVISOR_set_segment_base 25 /* x86/64 only */
-+#define __HYPERVISOR_mmuext_op 26
-+#define __HYPERVISOR_acm_op 27
-+#define __HYPERVISOR_nmi_op 28
-+#define __HYPERVISOR_sched_op 29
-+#define __HYPERVISOR_callback_op 30
-+#define __HYPERVISOR_xenoprof_op 31
-+#define __HYPERVISOR_event_channel_op 32
-+#define __HYPERVISOR_physdev_op 33
-+#define __HYPERVISOR_hvm_op 34
-+#define __HYPERVISOR_sysctl 35
-+#define __HYPERVISOR_domctl 36
-+#define __HYPERVISOR_kexec_op 37
-+
-+/* Architecture-specific hypercall definitions. */
-+#define __HYPERVISOR_arch_0 48
-+#define __HYPERVISOR_arch_1 49
-+#define __HYPERVISOR_arch_2 50
-+#define __HYPERVISOR_arch_3 51
-+#define __HYPERVISOR_arch_4 52
-+#define __HYPERVISOR_arch_5 53
-+#define __HYPERVISOR_arch_6 54
-+#define __HYPERVISOR_arch_7 55
++#define IOCTL_GNTDEV_MAP_GRANT_REF \
++_IOC(_IOC_NONE, 'G', 0, sizeof(struct ioctl_gntdev_map_grant_ref))
++struct ioctl_gntdev_map_grant_ref {
++ /* IN parameters */
++ /* The number of grants to be mapped. */
++ uint32_t count;
++ uint32_t pad;
++ /* OUT parameters */
++ /* The offset to be used on a subsequent call to mmap(). */
++ uint64_t index;
++ /* Variable IN parameter. */
++ /* Array of grant references, of size @count. */
++ struct ioctl_gntdev_grant_ref refs[1];
++};
+
+/*
-+ * HYPERCALL COMPATIBILITY.
++ * Removes the grant references from the mapping table of an instance of
++ * of gntdev. N.B. munmap() must be called on the relevant virtual address(es)
++ * before this ioctl is called, or an error will result.
+ */
++#define IOCTL_GNTDEV_UNMAP_GRANT_REF \
++_IOC(_IOC_NONE, 'G', 1, sizeof(struct ioctl_gntdev_unmap_grant_ref))
++struct ioctl_gntdev_unmap_grant_ref {
++ /* IN parameters */
++ /* The offset was returned by the corresponding map operation. */
++ uint64_t index;
++ /* The number of pages to be unmapped. */
++ uint32_t count;
++ uint32_t pad;
++};
+
-+/* New sched_op hypercall introduced in 0x00030101. */
-+#if __XEN_INTERFACE_VERSION__ < 0x00030101
-+#undef __HYPERVISOR_sched_op
-+#define __HYPERVISOR_sched_op __HYPERVISOR_sched_op_compat
-+#endif
-+
-+/* New event-channel and physdev hypercalls introduced in 0x00030202. */
-+#if __XEN_INTERFACE_VERSION__ < 0x00030202
-+#undef __HYPERVISOR_event_channel_op
-+#define __HYPERVISOR_event_channel_op __HYPERVISOR_event_channel_op_compat
-+#undef __HYPERVISOR_physdev_op
-+#define __HYPERVISOR_physdev_op __HYPERVISOR_physdev_op_compat
-+#endif
-+
-+/* New platform_op hypercall introduced in 0x00030204. */
-+#if __XEN_INTERFACE_VERSION__ < 0x00030204
-+#define __HYPERVISOR_dom0_op __HYPERVISOR_platform_op
-+#endif
-+
-+/*
-+ * VIRTUAL INTERRUPTS
-+ *
-+ * Virtual interrupts that a guest OS may receive from Xen.
-+ *
-+ * In the side comments, 'V.' denotes a per-VCPU VIRQ while 'G.' denotes a
-+ * global VIRQ. The former can be bound once per VCPU and cannot be re-bound.
-+ * The latter can be allocated only once per guest: they must initially be
-+ * allocated to VCPU0 but can subsequently be re-bound.
++/*
++ * Returns the offset in the driver's address space that corresponds
++ * to @vaddr. This can be used to perform a munmap(), followed by an
++ * UNMAP_GRANT_REF ioctl, where no state about the offset is retained by
++ * the caller. The number of pages that were allocated at the same time as
++ * @vaddr is returned in @count.
++ *
++ * N.B. Where more than one page has been mapped into a contiguous range, the
++ * supplied @vaddr must correspond to the start of the range; otherwise
++ * an error will result. It is only possible to munmap() the entire
++ * contiguously-allocated range at once, and not any subrange thereof.
+ */
-+#define VIRQ_TIMER 0 /* V. Timebase update, and/or requested timeout. */
-+#define VIRQ_DEBUG 1 /* V. Request guest to dump debug info. */
-+#define VIRQ_CONSOLE 2 /* G. (DOM0) Bytes received on emergency console. */
-+#define VIRQ_DOM_EXC 3 /* G. (DOM0) Exceptional event for some domain. */
-+#define VIRQ_TBUF 4 /* G. (DOM0) Trace buffer has records available. */
-+#define VIRQ_DEBUGGER 6 /* G. (DOM0) A domain has paused for debugging. */
-+#define VIRQ_XENOPROF 7 /* V. XenOprofile interrupt: new sample available */
-+#define VIRQ_CON_RING 8 /* G. (DOM0) Bytes received on console */
-+
-+/* Architecture-specific VIRQ definitions. */
-+#define VIRQ_ARCH_0 16
-+#define VIRQ_ARCH_1 17
-+#define VIRQ_ARCH_2 18
-+#define VIRQ_ARCH_3 19
-+#define VIRQ_ARCH_4 20
-+#define VIRQ_ARCH_5 21
-+#define VIRQ_ARCH_6 22
-+#define VIRQ_ARCH_7 23
-+
-+#define NR_VIRQS 24
++#define IOCTL_GNTDEV_GET_OFFSET_FOR_VADDR \
++_IOC(_IOC_NONE, 'G', 2, sizeof(struct ioctl_gntdev_get_offset_for_vaddr))
++struct ioctl_gntdev_get_offset_for_vaddr {
++ /* IN parameters */
++ /* The virtual address of the first mapped page in a range. */
++ uint64_t vaddr;
++ /* OUT parameters */
++ /* The offset that was used in the initial mmap() operation. */
++ uint64_t offset;
++ /* The number of pages mapped in the VM area that begins at @vaddr. */
++ uint32_t count;
++ uint32_t pad;
++};
+
+/*
-+ * MMU-UPDATE REQUESTS
-+ *
-+ * HYPERVISOR_mmu_update() accepts a list of (ptr, val) pairs.
-+ * A foreigndom (FD) can be specified (or DOMID_SELF for none).
-+ * Where the FD has some effect, it is described below.
-+ * ptr[1:0] specifies the appropriate MMU_* command.
-+ *
-+ * ptr[1:0] == MMU_NORMAL_PT_UPDATE:
-+ * Updates an entry in a page table. If updating an L1 table, and the new
-+ * table entry is valid/present, the mapped frame must belong to the FD, if
-+ * an FD has been specified. If attempting to map an I/O page then the
-+ * caller assumes the privilege of the FD.
-+ * FD == DOMID_IO: Permit /only/ I/O mappings, at the priv level of the caller.
-+ * FD == DOMID_XEN: Map restricted areas of Xen's heap space.
-+ * ptr[:2] -- Machine address of the page-table entry to modify.
-+ * val -- Value to write.
-+ *
-+ * ptr[1:0] == MMU_MACHPHYS_UPDATE:
-+ * Updates an entry in the machine->pseudo-physical mapping table.
-+ * ptr[:2] -- Machine address within the frame whose mapping to modify.
-+ * The frame must belong to the FD, if one is specified.
-+ * val -- Value to write into the mapping entry.
++ * Sets the maximum number of grants that may mapped at once by this gntdev
++ * instance.
++ *
++ * N.B. This must be called before any other ioctl is performed on the device.
+ */
-+#define MMU_NORMAL_PT_UPDATE 0 /* checked '*ptr = val'. ptr is MA. */
-+#define MMU_MACHPHYS_UPDATE 1 /* ptr = MA of frame to modify entry for */
++#define IOCTL_GNTDEV_SET_MAX_GRANTS \
++_IOC(_IOC_NONE, 'G', 3, sizeof(struct ioctl_gntdev_set_max_grants))
++struct ioctl_gntdev_set_max_grants {
++ /* IN parameter */
++ /* The maximum number of grants that may be mapped at once. */
++ uint32_t count;
++};
+
-+/*
-+ * MMU EXTENDED OPERATIONS
-+ *
-+ * HYPERVISOR_mmuext_op() accepts a list of mmuext_op structures.
-+ * A foreigndom (FD) can be specified (or DOMID_SELF for none).
-+ * Where the FD has some effect, it is described below.
-+ *
-+ * cmd: MMUEXT_(UN)PIN_*_TABLE
-+ * mfn: Machine frame number to be (un)pinned as a p.t. page.
-+ * The frame must belong to the FD, if one is specified.
-+ *
-+ * cmd: MMUEXT_NEW_BASEPTR
-+ * mfn: Machine frame number of new page-table base to install in MMU.
-+ *
-+ * cmd: MMUEXT_NEW_USER_BASEPTR [x86/64 only]
-+ * mfn: Machine frame number of new page-table base to install in MMU
-+ * when in user space.
-+ *
-+ * cmd: MMUEXT_TLB_FLUSH_LOCAL
-+ * No additional arguments. Flushes local TLB.
-+ *
-+ * cmd: MMUEXT_INVLPG_LOCAL
-+ * linear_addr: Linear address to be flushed from the local TLB.
++#endif /* __LINUX_PUBLIC_GNTDEV_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/public/privcmd.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/public/privcmd.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,79 @@
++/******************************************************************************
++ * privcmd.h
+ *
-+ * cmd: MMUEXT_TLB_FLUSH_MULTI
-+ * vcpumask: Pointer to bitmap of VCPUs to be flushed.
++ * Interface to /proc/xen/privcmd.
+ *
-+ * cmd: MMUEXT_INVLPG_MULTI
-+ * linear_addr: Linear address to be flushed.
-+ * vcpumask: Pointer to bitmap of VCPUs to be flushed.
++ * Copyright (c) 2003-2005, K A Fraser
+ *
-+ * cmd: MMUEXT_TLB_FLUSH_ALL
-+ * No additional arguments. Flushes all VCPUs' TLBs.
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
+ *
-+ * cmd: MMUEXT_INVLPG_ALL
-+ * linear_addr: Linear address to be flushed from all VCPUs' TLBs.
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
+ *
-+ * cmd: MMUEXT_FLUSH_CACHE
-+ * No additional arguments. Writes back and flushes cache contents.
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
+ *
-+ * cmd: MMUEXT_SET_LDT
-+ * linear_addr: Linear address of LDT base (NB. must be page-aligned).
-+ * nr_ents: Number of entries in LDT.
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
+ */
-+#define MMUEXT_PIN_L1_TABLE 0
-+#define MMUEXT_PIN_L2_TABLE 1
-+#define MMUEXT_PIN_L3_TABLE 2
-+#define MMUEXT_PIN_L4_TABLE 3
-+#define MMUEXT_UNPIN_TABLE 4
-+#define MMUEXT_NEW_BASEPTR 5
-+#define MMUEXT_TLB_FLUSH_LOCAL 6
-+#define MMUEXT_INVLPG_LOCAL 7
-+#define MMUEXT_TLB_FLUSH_MULTI 8
-+#define MMUEXT_INVLPG_MULTI 9
-+#define MMUEXT_TLB_FLUSH_ALL 10
-+#define MMUEXT_INVLPG_ALL 11
-+#define MMUEXT_FLUSH_CACHE 12
-+#define MMUEXT_SET_LDT 13
-+#define MMUEXT_NEW_USER_BASEPTR 15
+
-+#ifndef __ASSEMBLY__
-+struct mmuext_op {
-+ unsigned int cmd;
-+ union {
-+ /* [UN]PIN_TABLE, NEW_BASEPTR, NEW_USER_BASEPTR */
-+ xen_pfn_t mfn;
-+ /* INVLPG_LOCAL, INVLPG_ALL, SET_LDT */
-+ unsigned long linear_addr;
-+ } arg1;
-+ union {
-+ /* SET_LDT */
-+ unsigned int nr_ents;
-+ /* TLB_FLUSH_MULTI, INVLPG_MULTI */
-+ XEN_GUEST_HANDLE_00030205(void) vcpumask;
-+ } arg2;
-+};
-+typedef struct mmuext_op mmuext_op_t;
-+DEFINE_XEN_GUEST_HANDLE(mmuext_op_t);
-+#endif
++#ifndef __LINUX_PUBLIC_PRIVCMD_H__
++#define __LINUX_PUBLIC_PRIVCMD_H__
+
-+/* These are passed as 'flags' to update_va_mapping. They can be ORed. */
-+/* When specifying UVMF_MULTI, also OR in a pointer to a CPU bitmap. */
-+/* UVMF_LOCAL is merely UVMF_MULTI with a NULL bitmap pointer. */
-+#define UVMF_NONE (0UL<<0) /* No flushing at all. */
-+#define UVMF_TLB_FLUSH (1UL<<0) /* Flush entire TLB(s). */
-+#define UVMF_INVLPG (2UL<<0) /* Flush only one entry. */
-+#define UVMF_FLUSHTYPE_MASK (3UL<<0)
-+#define UVMF_MULTI (0UL<<2) /* Flush subset of TLBs. */
-+#define UVMF_LOCAL (0UL<<2) /* Flush local TLB. */
-+#define UVMF_ALL (1UL<<2) /* Flush all TLBs. */
++#include <linux/types.h>
+
-+/*
-+ * Commands to HYPERVISOR_console_io().
-+ */
-+#define CONSOLEIO_write 0
-+#define CONSOLEIO_read 1
++#ifndef __user
++#define __user
++#endif
+
-+/*
-+ * Commands to HYPERVISOR_vm_assist().
-+ */
-+#define VMASST_CMD_enable 0
-+#define VMASST_CMD_disable 1
++typedef struct privcmd_hypercall
++{
++ __u64 op;
++ __u64 arg[5];
++} privcmd_hypercall_t;
+
-+/* x86/32 guests: simulate full 4GB segment limits. */
-+#define VMASST_TYPE_4gb_segments 0
++typedef struct privcmd_mmap_entry {
++ __u64 va;
++ __u64 mfn;
++ __u64 npages;
++} privcmd_mmap_entry_t;
+
-+/* x86/32 guests: trap (vector 15) whenever above vmassist is used. */
-+#define VMASST_TYPE_4gb_segments_notify 1
++typedef struct privcmd_mmap {
++ int num;
++ domid_t dom; /* target domain */
++ privcmd_mmap_entry_t __user *entry;
++} privcmd_mmap_t;
++
++typedef struct privcmd_mmapbatch {
++ int num; /* number of pages to populate */
++ domid_t dom; /* target domain */
++ __u64 addr; /* virtual address */
++ xen_pfn_t __user *arr; /* array of mfns - top nibble set on err */
++} privcmd_mmapbatch_t;
+
+/*
-+ * x86 guests: support writes to bottom-level PTEs.
-+ * NB1. Page-directory entries cannot be written.
-+ * NB2. Guest must continue to remove all writable mappings of PTEs.
++ * @cmd: IOCTL_PRIVCMD_HYPERCALL
++ * @arg: &privcmd_hypercall_t
++ * Return: Value returned from execution of the specified hypercall.
+ */
-+#define VMASST_TYPE_writable_pagetables 2
++#define IOCTL_PRIVCMD_HYPERCALL \
++ _IOC(_IOC_NONE, 'P', 0, sizeof(privcmd_hypercall_t))
++#define IOCTL_PRIVCMD_MMAP \
++ _IOC(_IOC_NONE, 'P', 2, sizeof(privcmd_mmap_t))
++#define IOCTL_PRIVCMD_MMAPBATCH \
++ _IOC(_IOC_NONE, 'P', 3, sizeof(privcmd_mmapbatch_t))
+
-+/* x86/PAE guests: support PDPTs above 4GB. */
-+#define VMASST_TYPE_pae_extended_cr3 3
++#endif /* __LINUX_PUBLIC_PRIVCMD_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/xen_proc.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/xen_proc.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,12 @@
+
-+#define MAX_VMASST_TYPE 3
++#ifndef __ASM_XEN_PROC_H__
++#define __ASM_XEN_PROC_H__
+
-+#ifndef __ASSEMBLY__
++#include <linux/proc_fs.h>
+
-+typedef uint16_t domid_t;
++extern struct proc_dir_entry *create_xen_proc_entry(
++ const char *name, mode_t mode);
++extern void remove_xen_proc_entry(
++ const char *name);
++
++#endif /* __ASM_XEN_PROC_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/xenbus.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/xenbus.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,307 @@
++/******************************************************************************
++ * xenbus.h
++ *
++ * Talks to Xen Store to figure out what devices we have.
++ *
++ * Copyright (C) 2005 Rusty Russell, IBM Corporation
++ * Copyright (C) 2005 XenSource Ltd.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
+
-+/* Domain ids >= DOMID_FIRST_RESERVED cannot be used for ordinary domains. */
-+#define DOMID_FIRST_RESERVED (0x7FF0U)
++#ifndef _XEN_XENBUS_H
++#define _XEN_XENBUS_H
+
-+/* DOMID_SELF is used in certain contexts to refer to oneself. */
-+#define DOMID_SELF (0x7FF0U)
++#include <linux/device.h>
++#include <linux/notifier.h>
++#include <linux/mutex.h>
++#include <linux/completion.h>
++#include <linux/init.h>
++#include <linux/err.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/grant_table.h>
++#include <xen/interface/io/xenbus.h>
++#include <xen/interface/io/xs_wire.h>
+
-+/*
-+ * DOMID_IO is used to restrict page-table updates to mapping I/O memory.
-+ * Although no Foreign Domain need be specified to map I/O pages, DOMID_IO
-+ * is useful to ensure that no mappings to the OS's own heap are accidentally
-+ * installed. (e.g., in Linux this could cause havoc as reference counts
-+ * aren't adjusted on the I/O-mapping code path).
-+ * This only makes sense in MMUEXT_SET_FOREIGNDOM, but in that context can
-+ * be specified by any calling domain.
-+ */
-+#define DOMID_IO (0x7FF1U)
++/* Register callback to watch this node. */
++struct xenbus_watch
++{
++ struct list_head list;
+
-+/*
-+ * DOMID_XEN is used to allow privileged domains to map restricted parts of
-+ * Xen's heap space (e.g., the machine_to_phys table).
-+ * This only makes sense in MMUEXT_SET_FOREIGNDOM, and is only permitted if
-+ * the caller is privileged.
-+ */
-+#define DOMID_XEN (0x7FF2U)
++ /* Path being watched. */
++ const char *node;
+
-+/*
-+ * Send an array of these to HYPERVISOR_mmu_update().
-+ * NB. The fields are natural pointer/address size for this architecture.
-+ */
-+struct mmu_update {
-+ uint64_t ptr; /* Machine address of PTE. */
-+ uint64_t val; /* New contents of PTE. */
-+};
-+typedef struct mmu_update mmu_update_t;
-+DEFINE_XEN_GUEST_HANDLE(mmu_update_t);
++ /* Callback (executed in a process context with no locks held). */
++ void (*callback)(struct xenbus_watch *,
++ const char **vec, unsigned int len);
+
-+/*
-+ * Send an array of these to HYPERVISOR_multicall().
-+ * NB. The fields are natural register size for this architecture.
-+ */
-+struct multicall_entry {
-+ unsigned long op, result;
-+ unsigned long args[6];
++ /* See XBWF_ definitions below. */
++ unsigned long flags;
+};
-+typedef struct multicall_entry multicall_entry_t;
-+DEFINE_XEN_GUEST_HANDLE(multicall_entry_t);
+
+/*
-+ * Event channel endpoints per domain:
-+ * 1024 if a long is 32 bits; 4096 if a long is 64 bits.
++ * Execute callback in its own kthread. Useful if the callback is long
++ * running or heavily serialised, to avoid taking out the main xenwatch thread
++ * for a long period of time (or even unwittingly causing a deadlock).
+ */
-+#define NR_EVENT_CHANNELS (sizeof(unsigned long) * sizeof(unsigned long) * 64)
++#define XBWF_new_thread 1
+
-+struct vcpu_time_info {
-+ /*
-+ * Updates to the following values are preceded and followed by an
-+ * increment of 'version'. The guest can therefore detect updates by
-+ * looking for changes to 'version'. If the least-significant bit of
-+ * the version number is set then an update is in progress and the guest
-+ * must wait to read a consistent set of values.
-+ * The correct way to interact with the version number is similar to
-+ * Linux's seqlock: see the implementations of read_seqbegin/read_seqretry.
-+ */
-+ uint32_t version;
-+ uint32_t pad0;
-+ uint64_t tsc_timestamp; /* TSC at last update of time vals. */
-+ uint64_t system_time; /* Time, in nanosecs, since boot. */
-+ /*
-+ * Current system time:
-+ * system_time +
-+ * ((((tsc - tsc_timestamp) << tsc_shift) * tsc_to_system_mul) >> 32)
-+ * CPU frequency (Hz):
-+ * ((10^9 << 32) / tsc_to_system_mul) >> tsc_shift
-+ */
-+ uint32_t tsc_to_system_mul;
-+ int8_t tsc_shift;
-+ int8_t pad1[3];
-+}; /* 32 bytes */
-+typedef struct vcpu_time_info vcpu_time_info_t;
++/* A xenbus device. */
++struct xenbus_device {
++ const char *devicetype;
++ const char *nodename;
++ const char *otherend;
++ int otherend_id;
++ struct xenbus_watch otherend_watch;
++ struct device dev;
++ enum xenbus_state state;
++ struct completion down;
++};
+
-+struct vcpu_info {
-+ /*
-+ * 'evtchn_upcall_pending' is written non-zero by Xen to indicate
-+ * a pending notification for a particular VCPU. It is then cleared
-+ * by the guest OS /before/ checking for pending work, thus avoiding
-+ * a set-and-check race. Note that the mask is only accessed by Xen
-+ * on the CPU that is currently hosting the VCPU. This means that the
-+ * pending and mask flags can be updated by the guest without special
-+ * synchronisation (i.e., no need for the x86 LOCK prefix).
-+ * This may seem suboptimal because if the pending flag is set by
-+ * a different CPU then an IPI may be scheduled even when the mask
-+ * is set. However, note:
-+ * 1. The task of 'interrupt holdoff' is covered by the per-event-
-+ * channel mask bits. A 'noisy' event that is continually being
-+ * triggered can be masked at source at this very precise
-+ * granularity.
-+ * 2. The main purpose of the per-VCPU mask is therefore to restrict
-+ * reentrant execution: whether for concurrency control, or to
-+ * prevent unbounded stack usage. Whatever the purpose, we expect
-+ * that the mask will be asserted only for short periods at a time,
-+ * and so the likelihood of a 'spurious' IPI is suitably small.
-+ * The mask is read before making an event upcall to the guest: a
-+ * non-zero mask therefore guarantees that the VCPU will not receive
-+ * an upcall activation. The mask is cleared when the VCPU requests
-+ * to block: this avoids wakeup-waiting races.
-+ */
-+ uint8_t evtchn_upcall_pending;
-+ uint8_t evtchn_upcall_mask;
-+ unsigned long evtchn_pending_sel;
-+ struct arch_vcpu_info arch;
-+ struct vcpu_time_info time;
-+}; /* 64 bytes (x86) */
-+#ifndef __XEN__
-+typedef struct vcpu_info vcpu_info_t;
-+#endif
++static inline struct xenbus_device *to_xenbus_device(struct device *dev)
++{
++ return container_of(dev, struct xenbus_device, dev);
++}
+
-+/*
-+ * Xen/kernel shared data -- pointer provided in start_info.
-+ *
-+ * This structure is defined to be both smaller than a page, and the
-+ * only data on the shared page, but may vary in actual size even within
-+ * compatible Xen versions; guests should not rely on the size
-+ * of this structure remaining constant.
-+ */
-+struct shared_info {
-+ struct vcpu_info vcpu_info[MAX_VIRT_CPUS];
++struct xenbus_device_id
++{
++ /* .../device/<device_type>/<identifier> */
++ char devicetype[32]; /* General class of device. */
++};
+
-+ /*
-+ * A domain can create "event channels" on which it can send and receive
-+ * asynchronous event notifications. There are three classes of event that
-+ * are delivered by this mechanism:
-+ * 1. Bi-directional inter- and intra-domain connections. Domains must
-+ * arrange out-of-band to set up a connection (usually by allocating
-+ * an unbound 'listener' port and avertising that via a storage service
-+ * such as xenstore).
-+ * 2. Physical interrupts. A domain with suitable hardware-access
-+ * privileges can bind an event-channel port to a physical interrupt
-+ * source.
-+ * 3. Virtual interrupts ('events'). A domain can bind an event-channel
-+ * port to a virtual interrupt source, such as the virtual-timer
-+ * device or the emergency console.
-+ *
-+ * Event channels are addressed by a "port index". Each channel is
-+ * associated with two bits of information:
-+ * 1. PENDING -- notifies the domain that there is a pending notification
-+ * to be processed. This bit is cleared by the guest.
-+ * 2. MASK -- if this bit is clear then a 0->1 transition of PENDING
-+ * will cause an asynchronous upcall to be scheduled. This bit is only
-+ * updated by the guest. It is read-only within Xen. If a channel
-+ * becomes pending while the channel is masked then the 'edge' is lost
-+ * (i.e., when the channel is unmasked, the guest must manually handle
-+ * pending notifications as no upcall will be scheduled by Xen).
-+ *
-+ * To expedite scanning of pending notifications, any 0->1 pending
-+ * transition on an unmasked channel causes a corresponding bit in a
-+ * per-vcpu selector word to be set. Each bit in the selector covers a
-+ * 'C long' in the PENDING bitfield array.
-+ */
-+ unsigned long evtchn_pending[sizeof(unsigned long) * 8];
-+ unsigned long evtchn_mask[sizeof(unsigned long) * 8];
++/* A xenbus driver. */
++struct xenbus_driver {
++ char *name;
++ struct module *owner;
++ const struct xenbus_device_id *ids;
++ int (*probe)(struct xenbus_device *dev,
++ const struct xenbus_device_id *id);
++ void (*otherend_changed)(struct xenbus_device *dev,
++ enum xenbus_state backend_state);
++ int (*remove)(struct xenbus_device *dev);
++ int (*suspend)(struct xenbus_device *dev);
++ int (*suspend_cancel)(struct xenbus_device *dev);
++ int (*resume)(struct xenbus_device *dev);
++ int (*uevent)(struct xenbus_device *, char **, int, char *, int);
++ struct device_driver driver;
++ int (*read_otherend_details)(struct xenbus_device *dev);
++ int (*is_ready)(struct xenbus_device *dev);
++};
+
-+ /*
-+ * Wallclock time: updated only by control software. Guests should base
-+ * their gettimeofday() syscall on this wallclock-base value.
-+ */
-+ uint32_t wc_version; /* Version counter: see vcpu_time_info_t. */
-+ uint32_t wc_sec; /* Secs 00:00:00 UTC, Jan 1, 1970. */
-+ uint32_t wc_nsec; /* Nsecs 00:00:00 UTC, Jan 1, 1970. */
++static inline struct xenbus_driver *to_xenbus_driver(struct device_driver *drv)
++{
++ return container_of(drv, struct xenbus_driver, driver);
++}
+
-+ struct arch_shared_info arch;
++int xenbus_register_frontend(struct xenbus_driver *drv);
++int xenbus_register_backend(struct xenbus_driver *drv);
++void xenbus_unregister_driver(struct xenbus_driver *drv);
+
++struct xenbus_transaction
++{
++ u32 id;
+};
-+#ifndef __XEN__
-+typedef struct shared_info shared_info_t;
-+#endif
-+
-+/*
-+ * Start-of-day memory layout:
-+ * 1. The domain is started within contiguous virtual-memory region.
-+ * 2. The contiguous region ends on an aligned 4MB boundary.
-+ * 3. This the order of bootstrap elements in the initial virtual region:
-+ * a. relocated kernel image
-+ * b. initial ram disk [mod_start, mod_len]
-+ * c. list of allocated page frames [mfn_list, nr_pages]
-+ * d. start_info_t structure [register ESI (x86)]
-+ * e. bootstrap page tables [pt_base, CR3 (x86)]
-+ * f. bootstrap stack [register ESP (x86)]
-+ * 4. Bootstrap elements are packed together, but each is 4kB-aligned.
-+ * 5. The initial ram disk may be omitted.
-+ * 6. The list of page frames forms a contiguous 'pseudo-physical' memory
-+ * layout for the domain. In particular, the bootstrap virtual-memory
-+ * region is a 1:1 mapping to the first section of the pseudo-physical map.
-+ * 7. All bootstrap elements are mapped read-writable for the guest OS. The
-+ * only exception is the bootstrap page table, which is mapped read-only.
-+ * 8. There is guaranteed to be at least 512kB padding after the final
-+ * bootstrap element. If necessary, the bootstrap virtual region is
-+ * extended by an extra 4MB to ensure this.
-+ */
+
-+#define MAX_GUEST_CMDLINE 1024
-+struct start_info {
-+ /* THE FOLLOWING ARE FILLED IN BOTH ON INITIAL BOOT AND ON RESUME. */
-+ char magic[32]; /* "xen-<version>-<platform>". */
-+ unsigned long nr_pages; /* Total pages allocated to this domain. */
-+ unsigned long shared_info; /* MACHINE address of shared info struct. */
-+ uint32_t flags; /* SIF_xxx flags. */
-+ xen_pfn_t store_mfn; /* MACHINE page number of shared page. */
-+ uint32_t store_evtchn; /* Event channel for store communication. */
-+ union {
-+ struct {
-+ xen_pfn_t mfn; /* MACHINE page number of console page. */
-+ uint32_t evtchn; /* Event channel for console page. */
-+ } domU;
-+ struct {
-+ uint32_t info_off; /* Offset of console_info struct. */
-+ uint32_t info_size; /* Size of console_info struct from start.*/
-+ } dom0;
-+ } console;
-+ /* THE FOLLOWING ARE ONLY FILLED IN ON INITIAL BOOT (NOT RESUME). */
-+ unsigned long pt_base; /* VIRTUAL address of page directory. */
-+ unsigned long nr_pt_frames; /* Number of bootstrap p.t. frames. */
-+ unsigned long mfn_list; /* VIRTUAL address of page-frame list. */
-+ unsigned long mod_start; /* VIRTUAL address of pre-loaded module. */
-+ unsigned long mod_len; /* Size (bytes) of pre-loaded module. */
-+ int8_t cmd_line[MAX_GUEST_CMDLINE];
-+};
-+typedef struct start_info start_info_t;
++/* Nil transaction ID. */
++#define XBT_NIL ((struct xenbus_transaction) { 0 })
+
-+/* New console union for dom0 introduced in 0x00030203. */
-+#if __XEN_INTERFACE_VERSION__ < 0x00030203
-+#define console_mfn console.domU.mfn
-+#define console_evtchn console.domU.evtchn
-+#endif
++char **xenbus_directory(struct xenbus_transaction t,
++ const char *dir, const char *node, unsigned int *num);
++void *xenbus_read(struct xenbus_transaction t,
++ const char *dir, const char *node, unsigned int *len);
++int xenbus_write(struct xenbus_transaction t,
++ const char *dir, const char *node, const char *string);
++int xenbus_mkdir(struct xenbus_transaction t,
++ const char *dir, const char *node);
++int xenbus_exists(struct xenbus_transaction t,
++ const char *dir, const char *node);
++int xenbus_rm(struct xenbus_transaction t, const char *dir, const char *node);
++int xenbus_transaction_start(struct xenbus_transaction *t);
++int xenbus_transaction_end(struct xenbus_transaction t, int abort);
+
-+/* These flags are passed in the 'flags' field of start_info_t. */
-+#define SIF_PRIVILEGED (1<<0) /* Is the domain privileged? */
-+#define SIF_INITDOMAIN (1<<1) /* Is this the initial control domain? */
++/* Single read and scanf: returns -errno or num scanned if > 0. */
++int xenbus_scanf(struct xenbus_transaction t,
++ const char *dir, const char *node, const char *fmt, ...)
++ __attribute__((format(scanf, 4, 5)));
+
-+typedef struct dom0_vga_console_info {
-+ uint8_t video_type; /* DOM0_VGA_CONSOLE_??? */
-+#define XEN_VGATYPE_TEXT_MODE_3 0x03
-+#define XEN_VGATYPE_VESA_LFB 0x23
++/* Single printf and write: returns -errno or 0. */
++int xenbus_printf(struct xenbus_transaction t,
++ const char *dir, const char *node, const char *fmt, ...)
++ __attribute__((format(printf, 4, 5)));
+
-+ union {
-+ struct {
-+ /* Font height, in pixels. */
-+ uint16_t font_height;
-+ /* Cursor location (column, row). */
-+ uint16_t cursor_x, cursor_y;
-+ /* Number of rows and columns (dimensions in characters). */
-+ uint16_t rows, columns;
-+ } text_mode_3;
++/* Generic read function: NULL-terminated triples of name,
++ * sprintf-style type string, and pointer. Returns 0 or errno.*/
++int xenbus_gather(struct xenbus_transaction t, const char *dir, ...);
+
-+ struct {
-+ /* Width and height, in pixels. */
-+ uint16_t width, height;
-+ /* Bytes per scan line. */
-+ uint16_t bytes_per_line;
-+ /* Bits per pixel. */
-+ uint16_t bits_per_pixel;
-+ /* LFB physical address, and size (in units of 64kB). */
-+ uint32_t lfb_base;
-+ uint32_t lfb_size;
-+ /* RGB mask offsets and sizes, as defined by VBE 1.2+ */
-+ uint8_t red_pos, red_size;
-+ uint8_t green_pos, green_size;
-+ uint8_t blue_pos, blue_size;
-+ uint8_t rsvd_pos, rsvd_size;
-+ } vesa_lfb;
-+ } u;
-+} dom0_vga_console_info_t;
-+#define xen_vga_console_info dom0_vga_console_info
-+#define xen_vga_console_info_t dom0_vga_console_info_t
++/* notifer routines for when the xenstore comes up */
++int register_xenstore_notifier(struct notifier_block *nb);
++void unregister_xenstore_notifier(struct notifier_block *nb);
+
-+typedef uint8_t xen_domain_handle_t[16];
++int register_xenbus_watch(struct xenbus_watch *watch);
++void unregister_xenbus_watch(struct xenbus_watch *watch);
++void xs_suspend(void);
++void xs_resume(void);
++void xs_suspend_cancel(void);
+
-+/* Turn a plain number into a C unsigned long constant. */
-+#define __mk_unsigned_long(x) x ## UL
-+#define mk_unsigned_long(x) __mk_unsigned_long(x)
++/* Used by xenbus_dev to borrow kernel's store connection. */
++void *xenbus_dev_request_and_reply(struct xsd_sockmsg *msg);
+
-+DEFINE_XEN_GUEST_HANDLE(uint8_t);
-+DEFINE_XEN_GUEST_HANDLE(uint16_t);
-+DEFINE_XEN_GUEST_HANDLE(uint32_t);
-+DEFINE_XEN_GUEST_HANDLE(uint64_t);
++/* Prepare for domain suspend: then resume or cancel the suspend. */
++void xenbus_suspend(void);
++void xenbus_resume(void);
++void xenbus_suspend_cancel(void);
+
-+#else /* __ASSEMBLY__ */
++#define XENBUS_IS_ERR_READ(str) ({ \
++ if (!IS_ERR(str) && strlen(str) == 0) { \
++ kfree(str); \
++ str = ERR_PTR(-ERANGE); \
++ } \
++ IS_ERR(str); \
++})
+
-+/* In assembly code we cannot use C numeric constant suffixes. */
-+#define mk_unsigned_long(x) x
++#define XENBUS_EXIST_ERR(err) ((err) == -ENOENT || (err) == -ERANGE)
+
-+#endif /* !__ASSEMBLY__ */
+
-+/* Default definitions for macros used by domctl/sysctl. */
-+#if defined(__XEN__) || defined(__XEN_TOOLS__)
-+#ifndef uint64_aligned_t
-+#define uint64_aligned_t uint64_t
-+#endif
-+#ifndef XEN_GUEST_HANDLE_64
-+#define XEN_GUEST_HANDLE_64(name) XEN_GUEST_HANDLE(name)
-+#endif
-+#endif
++/**
++ * Register a watch on the given path, using the given xenbus_watch structure
++ * for storage, and the given callback function as the callback. Return 0 on
++ * success, or -errno on error. On success, the given path will be saved as
++ * watch->node, and remains the caller's to free. On error, watch->node will
++ * be NULL, the device will switch to XenbusStateClosing, and the error will
++ * be saved in the store.
++ */
++int xenbus_watch_path(struct xenbus_device *dev, const char *path,
++ struct xenbus_watch *watch,
++ void (*callback)(struct xenbus_watch *,
++ const char **, unsigned int));
+
-+#endif /* __XEN_PUBLIC_XEN_H__ */
+
-+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
++/**
++ * Register a watch on the given path/path2, using the given xenbus_watch
++ * structure for storage, and the given callback function as the callback.
++ * Return 0 on success, or -errno on error. On success, the watched path
++ * (path/path2) will be saved as watch->node, and becomes the caller's to
++ * kfree(). On error, watch->node will be NULL, so the caller has nothing to
++ * free, the device will switch to XenbusStateClosing, and the error will be
++ * saved in the store.
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/interface/xenoprof.h tmp-linux-2.6-xen.patch/include/xen/interface/xenoprof.h
---- pristine-linux-2.6.18.2/include/xen/interface/xenoprof.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/interface/xenoprof.h 2007-09-30 18:06:19.000000000 +0200
-@@ -0,0 +1,138 @@
-+/******************************************************************************
-+ * xenoprof.h
-+ *
-+ * Interface for enabling system wide profiling based on hardware performance
-+ * counters
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this software and associated documentation files (the "Software"), to
-+ * deal in the Software without restriction, including without limitation the
-+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
-+ * sell copies of the Software, and to permit persons to whom the Software is
-+ * furnished to do so, subject to the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-+ * DEALINGS IN THE SOFTWARE.
-+ *
-+ * Copyright (C) 2005 Hewlett-Packard Co.
-+ * Written by Aravind Menon & Jose Renato Santos
++int xenbus_watch_path2(struct xenbus_device *dev, const char *path,
++ const char *path2, struct xenbus_watch *watch,
++ void (*callback)(struct xenbus_watch *,
++ const char **, unsigned int));
++
++
++/**
++ * Advertise in the store a change of the given driver to the given new_state.
++ * Return 0 on success, or -errno on error. On error, the device will switch
++ * to XenbusStateClosing, and the error will be saved in the store.
+ */
++int xenbus_switch_state(struct xenbus_device *dev, enum xenbus_state new_state);
+
-+#ifndef __XEN_PUBLIC_XENOPROF_H__
-+#define __XEN_PUBLIC_XENOPROF_H__
+
-+#include "xen.h"
++/**
++ * Grant access to the given ring_mfn to the peer of the given device. Return
++ * 0 on success, or -errno on error. On error, the device will switch to
++ * XenbusStateClosing, and the error will be saved in the store.
++ */
++int xenbus_grant_ring(struct xenbus_device *dev, unsigned long ring_mfn);
+
-+/*
-+ * Commands to HYPERVISOR_xenoprof_op().
++
++/**
++ * Map a page of memory into this domain from another domain's grant table.
++ * xenbus_map_ring_valloc allocates a page of virtual address space, maps the
++ * page to that address, and sets *vaddr to that address.
++ * xenbus_map_ring does not allocate the virtual address space (you must do
++ * this yourself!). It only maps in the page to the specified address.
++ * Returns 0 on success, and GNTST_* (see xen/include/interface/grant_table.h)
++ * or -ENOMEM on error. If an error is returned, device will switch to
++ * XenbusStateClosing and the error message will be saved in XenStore.
+ */
-+#define XENOPROF_init 0
-+#define XENOPROF_reset_active_list 1
-+#define XENOPROF_reset_passive_list 2
-+#define XENOPROF_set_active 3
-+#define XENOPROF_set_passive 4
-+#define XENOPROF_reserve_counters 5
-+#define XENOPROF_counter 6
-+#define XENOPROF_setup_events 7
-+#define XENOPROF_enable_virq 8
-+#define XENOPROF_start 9
-+#define XENOPROF_stop 10
-+#define XENOPROF_disable_virq 11
-+#define XENOPROF_release_counters 12
-+#define XENOPROF_shutdown 13
-+#define XENOPROF_get_buffer 14
-+#define XENOPROF_set_backtrace 15
-+#define XENOPROF_last_op 15
++struct vm_struct *xenbus_map_ring_valloc(struct xenbus_device *dev,
++ int gnt_ref);
++int xenbus_map_ring(struct xenbus_device *dev, int gnt_ref,
++ grant_handle_t *handle, void *vaddr);
+
-+#define MAX_OPROF_EVENTS 32
-+#define MAX_OPROF_DOMAINS 25
-+#define XENOPROF_CPU_TYPE_SIZE 64
+
-+/* Xenoprof performance events (not Xen events) */
-+struct event_log {
-+ uint64_t eip;
-+ uint8_t mode;
-+ uint8_t event;
-+};
++/**
++ * Unmap a page of memory in this domain that was imported from another domain.
++ * Use xenbus_unmap_ring_vfree if you mapped in your memory with
++ * xenbus_map_ring_valloc (it will free the virtual address space).
++ * Returns 0 on success and returns GNTST_* on error
++ * (see xen/include/interface/grant_table.h).
++ */
++int xenbus_unmap_ring_vfree(struct xenbus_device *dev, struct vm_struct *);
++int xenbus_unmap_ring(struct xenbus_device *dev,
++ grant_handle_t handle, void *vaddr);
++
+
-+/* PC value that indicates a special code */
-+#define XENOPROF_ESCAPE_CODE ~0UL
-+/* Transient events for the xenoprof->oprofile cpu buf */
-+#define XENOPROF_TRACE_BEGIN 1
++/**
++ * Allocate an event channel for the given xenbus_device, assigning the newly
++ * created local port to *port. Return 0 on success, or -errno on error. On
++ * error, the device will switch to XenbusStateClosing, and the error will be
++ * saved in the store.
++ */
++int xenbus_alloc_evtchn(struct xenbus_device *dev, int *port);
+
-+/* Xenoprof buffer shared between Xen and domain - 1 per VCPU */
-+struct xenoprof_buf {
-+ uint32_t event_head;
-+ uint32_t event_tail;
-+ uint32_t event_size;
-+ uint32_t vcpu_id;
-+ uint64_t xen_samples;
-+ uint64_t kernel_samples;
-+ uint64_t user_samples;
-+ uint64_t lost_samples;
-+ struct event_log event_log[1];
-+};
-+#ifndef __XEN__
-+typedef struct xenoprof_buf xenoprof_buf_t;
-+DEFINE_XEN_GUEST_HANDLE(xenoprof_buf_t);
-+#endif
+
-+struct xenoprof_init {
-+ int32_t num_events;
-+ int32_t is_primary;
-+ char cpu_type[XENOPROF_CPU_TYPE_SIZE];
-+};
-+typedef struct xenoprof_init xenoprof_init_t;
-+DEFINE_XEN_GUEST_HANDLE(xenoprof_init_t);
++/**
++ * Free an existing event channel. Returns 0 on success or -errno on error.
++ */
++int xenbus_free_evtchn(struct xenbus_device *dev, int port);
+
-+struct xenoprof_get_buffer {
-+ int32_t max_samples;
-+ int32_t nbuf;
-+ int32_t bufsize;
-+ uint64_t buf_gmaddr;
-+};
-+typedef struct xenoprof_get_buffer xenoprof_get_buffer_t;
-+DEFINE_XEN_GUEST_HANDLE(xenoprof_get_buffer_t);
+
-+struct xenoprof_counter {
-+ uint32_t ind;
-+ uint64_t count;
-+ uint32_t enabled;
-+ uint32_t event;
-+ uint32_t hypervisor;
-+ uint32_t kernel;
-+ uint32_t user;
-+ uint64_t unit_mask;
-+};
-+typedef struct xenoprof_counter xenoprof_counter_t;
-+DEFINE_XEN_GUEST_HANDLE(xenoprof_counter_t);
++/**
++ * Return the state of the driver rooted at the given store path, or
++ * XenbusStateUnknown if no state can be read.
++ */
++enum xenbus_state xenbus_read_driver_state(const char *path);
+
-+typedef struct xenoprof_passive {
-+ uint16_t domain_id;
-+ int32_t max_samples;
-+ int32_t nbuf;
-+ int32_t bufsize;
-+ uint64_t buf_gmaddr;
-+} xenoprof_passive_t;
-+DEFINE_XEN_GUEST_HANDLE(xenoprof_passive_t);
+
++/***
++ * Report the given negative errno into the store, along with the given
++ * formatted message.
++ */
++void xenbus_dev_error(struct xenbus_device *dev, int err, const char *fmt,
++ ...);
+
-+#endif /* __XEN_PUBLIC_XENOPROF_H__ */
+
-+/*
-+ * Local variables:
-+ * mode: C
-+ * c-set-style: "BSD"
-+ * c-basic-offset: 4
-+ * tab-width: 4
-+ * indent-tabs-mode: nil
-+ * End:
++/***
++ * Equivalent to xenbus_dev_error(dev, err, fmt, args), followed by
++ * xenbus_switch_state(dev, NULL, XenbusStateClosing) to schedule an orderly
++ * closedown of this driver and its peer.
+ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/pcifront.h tmp-linux-2.6-xen.patch/include/xen/pcifront.h
---- pristine-linux-2.6.18.2/include/xen/pcifront.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/pcifront.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,83 @@
++void xenbus_dev_fatal(struct xenbus_device *dev, int err, const char *fmt,
++ ...);
++
++int xenbus_dev_init(void);
++
++const char *xenbus_strstate(enum xenbus_state state);
++int xenbus_dev_is_online(struct xenbus_device *dev);
++int xenbus_frontend_closed(struct xenbus_device *dev);
++
++int xenbus_for_each_backend(void *arg, int (*fn)(struct device *, void *));
++int xenbus_for_each_frontend(void *arg, int (*fn)(struct device *, void *));
++
++#endif /* _XEN_XENBUS_H */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/xencomm.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/xencomm.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,77 @@
+/*
-+ * PCI Frontend - arch-dependendent declarations
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
-+ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Hollis Blanchard <hollisb at us.ibm.com>
++ * Jerone Young <jyoung5 at us.ibm.com>
+ */
-+#ifndef __XEN_ASM_PCIFRONT_H__
-+#define __XEN_ASM_PCIFRONT_H__
+
-+#include <linux/spinlock.h>
++#ifndef _LINUX_XENCOMM_H_
++#define _LINUX_XENCOMM_H_
+
-+#ifdef __KERNEL__
++#include <xen/interface/xencomm.h>
+
-+#ifndef __ia64__
++#define XENCOMM_MINI_ADDRS 3
++struct xencomm_mini {
++ struct xencomm_desc _desc;
++ uint64_t address[XENCOMM_MINI_ADDRS];
++};
+
-+struct pcifront_device;
-+struct pci_bus;
++/* To avoid additionnal virt to phys conversion, an opaque structure is
++ presented. */
++struct xencomm_handle;
+
-+struct pcifront_sd {
-+ int domain;
-+ struct pcifront_device *pdev;
-+};
++extern void xencomm_free(struct xencomm_handle *desc);
++extern struct xencomm_handle *xencomm_map(void *ptr, unsigned long bytes);
++extern struct xencomm_handle *__xencomm_map_no_alloc(void *ptr,
++ unsigned long bytes, struct xencomm_mini *xc_area);
+
-+static inline struct pcifront_device *
-+pcifront_get_pdev(struct pcifront_sd *sd)
-+{
-+ return sd->pdev;
-+}
++#if 0
++#define XENCOMM_MINI_ALIGNED(xc_desc, n) \
++ struct xencomm_mini xc_desc ## _base[(n)] \
++ __attribute__((__aligned__(sizeof(struct xencomm_mini)))); \
++ struct xencomm_mini* xc_desc = &xc_desc ## _base[0];
++#else
++/*
++ * gcc bug workaround:
++ * http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16660
++ * gcc doesn't handle properly stack variable with
++ * __attribute__((__align__(sizeof(struct xencomm_mini))))
++ */
++#define XENCOMM_MINI_ALIGNED(xc_desc, n) \
++ unsigned char xc_desc ## _base[((n) + 1 ) * \
++ sizeof(struct xencomm_mini)]; \
++ struct xencomm_mini *xc_desc = (struct xencomm_mini*) \
++ ((unsigned long)xc_desc ## _base + \
++ (sizeof(struct xencomm_mini) - \
++ ((unsigned long)xc_desc ## _base) % \
++ sizeof(struct xencomm_mini)));
++#endif
++#define xencomm_map_no_alloc(ptr, bytes) \
++ ({XENCOMM_MINI_ALIGNED(xc_desc, 1); \
++ __xencomm_map_no_alloc(ptr, bytes, xc_desc);})
+
-+static inline void pcifront_init_sd(struct pcifront_sd *sd,
-+ unsigned int domain, unsigned int bus,
-+ struct pcifront_device *pdev)
-+{
-+ sd->domain = domain;
-+ sd->pdev = pdev;
-+}
++/* provided by architecture code: */
++extern unsigned long xencomm_vtop(unsigned long vaddr);
+
-+#if defined(CONFIG_PCI_DOMAINS)
-+static inline int pci_domain_nr(struct pci_bus *bus)
-+{
-+ struct pcifront_sd *sd = bus->sysdata;
-+ return sd->domain;
-+}
-+static inline int pci_proc_domain(struct pci_bus *bus)
++static inline void *xencomm_pa(void *ptr)
+{
-+ return pci_domain_nr(bus);
++ return (void *)xencomm_vtop((unsigned long)ptr);
+}
-+#endif /* CONFIG_PCI_DOMAINS */
+
-+static inline void pcifront_setup_root_resources(struct pci_bus *bus,
-+ struct pcifront_sd *sd)
-+{
-+}
++#define xen_guest_handle(hnd) ((hnd).p)
+
-+#else /* __ia64__ */
++#endif /* _LINUX_XENCOMM_H_ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/xencons.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/xencons.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,17 @@
++#ifndef __ASM_XENCONS_H__
++#define __ASM_XENCONS_H__
+
-+#include <linux/acpi.h>
-+#include <asm/pci.h>
-+#define pcifront_sd pci_controller
++struct dom0_vga_console_info;
++void dom0_init_screen_info(const struct dom0_vga_console_info *, size_t);
+
-+extern void xen_add_resource(struct pci_controller *, unsigned int,
-+ unsigned int, struct acpi_resource *);
-+extern void xen_pcibios_setup_root_windows(struct pci_bus *,
-+ struct pci_controller *);
++void xencons_force_flush(void);
++void xencons_resume(void);
+
-+static inline struct pcifront_device *
-+pcifront_get_pdev(struct pcifront_sd *sd)
-+{
-+ return (struct pcifront_device *)sd->platform_data;
-+}
++/* Interrupt work hooks. Receive data, or kick data out. */
++void xencons_rx(char *buf, unsigned len, struct pt_regs *regs);
++void xencons_tx(void);
+
-+static inline void pcifront_setup_root_resources(struct pci_bus *bus,
-+ struct pcifront_sd *sd)
-+{
-+ xen_pcibios_setup_root_windows(bus, sd);
-+}
++int xencons_ring_init(void);
++int xencons_ring_send(const char *data, unsigned len);
+
-+#endif /* __ia64__ */
++#endif /* __ASM_XENCONS_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb include/xen/xenoprof.h
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/include/xen/xenoprof.h Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,42 @@
++/******************************************************************************
++ * xen/xenoprof.h
++ *
++ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ */
+
-+extern struct rw_semaphore pci_bus_sem;
++#ifndef __XEN_XENOPROF_H__
++#define __XEN_XENOPROF_H__
++#ifdef CONFIG_XEN
+
-+#endif /* __KERNEL__ */
++#include <asm/xenoprof.h>
+
-+#endif /* __XEN_ASM_PCIFRONT_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/public/evtchn.h tmp-linux-2.6-xen.patch/include/xen/public/evtchn.h
---- pristine-linux-2.6.18.2/include/xen/public/evtchn.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/public/evtchn.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,88 @@
-+/******************************************************************************
-+ * evtchn.h
-+ *
-+ * Interface to /dev/xen/evtchn.
-+ *
-+ * Copyright (c) 2003-2005, K A Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++struct oprofile_operations;
++int xenoprofile_init(struct oprofile_operations * ops);
++void xenoprofile_exit(void);
+
-+#ifndef __LINUX_PUBLIC_EVTCHN_H__
-+#define __LINUX_PUBLIC_EVTCHN_H__
++struct xenoprof_shared_buffer {
++ char *buffer;
++ struct xenoprof_arch_shared_buffer arch;
++};
++#else
++#define xenoprofile_init(ops) (-ENOSYS)
++#define xenoprofile_exit() do { } while (0)
++
++#endif /* CONFIG_XEN */
++#endif /* __XEN_XENOPROF_H__ */
+diff -r d894e36cfc30 -r 0aa021803deb kernel/Kconfig.preempt
+--- a/kernel/Kconfig.preempt Tue Sep 09 11:37:38 2008 +0200
++++ b/kernel/Kconfig.preempt Wed Sep 10 10:54:08 2008 +0100
+@@ -35,6 +35,7 @@
+
+ config PREEMPT
+ bool "Preemptible Kernel (Low-Latency Desktop)"
++ depends on !XEN
+ help
+ This option reduces the latency of the kernel by making
+ all kernel code (that is not executing in a critical section)
+diff -r d894e36cfc30 -r 0aa021803deb kernel/cpu.c
+--- a/kernel/cpu.c Tue Sep 09 11:37:38 2008 +0200
++++ b/kernel/cpu.c Wed Sep 10 10:54:08 2008 +0100
+@@ -48,7 +48,10 @@
+
+ void unlock_cpu_hotplug(void)
+ {
++/* cpufreq lock-takers fixed in mainline; shut up until dom0 kernel catches up*/
++#ifdef CONFIG_XEN
+ WARN_ON(recursive != current);
++#endif
+ if (recursive_depth) {
+ recursive_depth--;
+ return;
+diff -r d894e36cfc30 -r 0aa021803deb kernel/fork.c
+--- a/kernel/fork.c Tue Sep 09 11:37:38 2008 +0200
++++ b/kernel/fork.c Wed Sep 10 10:54:08 2008 +0100
+@@ -276,6 +276,9 @@
+ if (retval)
+ goto out;
+ }
++#ifdef arch_dup_mmap
++ arch_dup_mmap(mm, oldmm);
++#endif
+ retval = 0;
+ out:
+ up_write(&mm->mmap_sem);
+diff -r d894e36cfc30 -r 0aa021803deb kernel/irq/spurious.c
+--- a/kernel/irq/spurious.c Tue Sep 09 11:37:38 2008 +0200
++++ b/kernel/irq/spurious.c Wed Sep 10 10:54:08 2008 +0100
+@@ -139,7 +139,8 @@
+ irqreturn_t action_ret, struct pt_regs *regs)
+ {
+ if (unlikely(action_ret != IRQ_HANDLED)) {
+- desc->irqs_unhandled++;
++ if (!irq_ignore_unhandled(irq))
++ desc->irqs_unhandled++;
+ if (unlikely(action_ret != IRQ_NONE))
+ report_bad_irq(irq, desc, action_ret);
+ }
+diff -r d894e36cfc30 -r 0aa021803deb kernel/kexec.c
+--- a/kernel/kexec.c Tue Sep 09 11:37:38 2008 +0200
++++ b/kernel/kexec.c Wed Sep 10 10:54:08 2008 +0100
+@@ -330,13 +330,26 @@
+ return 0;
+ }
+
+-static struct page *kimage_alloc_pages(gfp_t gfp_mask, unsigned int order)
++static struct page *kimage_alloc_pages(gfp_t gfp_mask, unsigned int order, unsigned long limit)
+ {
+ struct page *pages;
+
+ pages = alloc_pages(gfp_mask, order);
+ if (pages) {
+ unsigned int count, i;
++#ifdef CONFIG_XEN
++ int address_bits;
++
++ if (limit == ~0UL)
++ address_bits = BITS_PER_LONG;
++ else
++ address_bits = long_log2(limit);
++
++ if (xen_limit_pages_to_max_mfn(pages, order, address_bits) < 0) {
++ __free_pages(pages, order);
++ return NULL;
++ }
++#endif
+ pages->mapping = NULL;
+ set_page_private(pages, order);
+ count = 1 << order;
+@@ -355,6 +368,9 @@
+ count = 1 << order;
+ for (i = 0; i < count; i++)
+ ClearPageReserved(page + i);
++#ifdef CONFIG_XEN
++ xen_destroy_contiguous_region((unsigned long)page_address(page), order);
++#endif
+ __free_pages(page, order);
+ }
+
+@@ -400,10 +416,10 @@
+ do {
+ unsigned long pfn, epfn, addr, eaddr;
+
+- pages = kimage_alloc_pages(GFP_KERNEL, order);
++ pages = kimage_alloc_pages(GFP_KERNEL, order, KEXEC_CONTROL_MEMORY_LIMIT);
+ if (!pages)
+ break;
+- pfn = page_to_pfn(pages);
++ pfn = kexec_page_to_pfn(pages);
+ epfn = pfn + count;
+ addr = pfn << PAGE_SHIFT;
+ eaddr = epfn << PAGE_SHIFT;
+@@ -437,6 +453,7 @@
+ return pages;
+ }
+
++#ifndef CONFIG_XEN
+ static struct page *kimage_alloc_crash_control_pages(struct kimage *image,
+ unsigned int order)
+ {
+@@ -490,7 +507,7 @@
+ }
+ /* If I don't overlap any segments I have found my hole! */
+ if (i == image->nr_segments) {
+- pages = pfn_to_page(hole_start >> PAGE_SHIFT);
++ pages = kexec_pfn_to_page(hole_start >> PAGE_SHIFT);
+ break;
+ }
+ }
+@@ -517,6 +534,13 @@
+
+ return pages;
+ }
++#else /* !CONFIG_XEN */
++struct page *kimage_alloc_control_pages(struct kimage *image,
++ unsigned int order)
++{
++ return kimage_alloc_normal_control_pages(image, order);
++}
++#endif
+
+ static int kimage_add_entry(struct kimage *image, kimage_entry_t entry)
+ {
+@@ -532,7 +556,7 @@
+ return -ENOMEM;
+
+ ind_page = page_address(page);
+- *image->entry = virt_to_phys(ind_page) | IND_INDIRECTION;
++ *image->entry = kexec_virt_to_phys(ind_page) | IND_INDIRECTION;
+ image->entry = ind_page;
+ image->last_entry = ind_page +
+ ((PAGE_SIZE/sizeof(kimage_entry_t)) - 1);
+@@ -593,13 +617,13 @@
+ #define for_each_kimage_entry(image, ptr, entry) \
+ for (ptr = &image->head; (entry = *ptr) && !(entry & IND_DONE); \
+ ptr = (entry & IND_INDIRECTION)? \
+- phys_to_virt((entry & PAGE_MASK)): ptr +1)
++ kexec_phys_to_virt((entry & PAGE_MASK)): ptr +1)
+
+ static void kimage_free_entry(kimage_entry_t entry)
+ {
+ struct page *page;
+
+- page = pfn_to_page(entry >> PAGE_SHIFT);
++ page = kexec_pfn_to_page(entry >> PAGE_SHIFT);
+ kimage_free_pages(page);
+ }
+
+@@ -610,6 +634,10 @@
+
+ if (!image)
+ return;
++
++#ifdef CONFIG_XEN
++ xen_machine_kexec_unload(image);
++#endif
+
+ kimage_free_extra_pages(image);
+ for_each_kimage_entry(image, ptr, entry) {
+@@ -686,7 +714,7 @@
+ * have a match.
+ */
+ list_for_each_entry(page, &image->dest_pages, lru) {
+- addr = page_to_pfn(page) << PAGE_SHIFT;
++ addr = kexec_page_to_pfn(page) << PAGE_SHIFT;
+ if (addr == destination) {
+ list_del(&page->lru);
+ return page;
+@@ -697,16 +725,16 @@
+ kimage_entry_t *old;
+
+ /* Allocate a page, if we run out of memory give up */
+- page = kimage_alloc_pages(gfp_mask, 0);
++ page = kimage_alloc_pages(gfp_mask, 0, KEXEC_SOURCE_MEMORY_LIMIT);
+ if (!page)
+ return NULL;
+ /* If the page cannot be used file it away */
+- if (page_to_pfn(page) >
++ if (kexec_page_to_pfn(page) >
+ (KEXEC_SOURCE_MEMORY_LIMIT >> PAGE_SHIFT)) {
+ list_add(&page->lru, &image->unuseable_pages);
+ continue;
+ }
+- addr = page_to_pfn(page) << PAGE_SHIFT;
++ addr = kexec_page_to_pfn(page) << PAGE_SHIFT;
+
+ /* If it is the destination page we want use it */
+ if (addr == destination)
+@@ -729,7 +757,7 @@
+ struct page *old_page;
+
+ old_addr = *old & PAGE_MASK;
+- old_page = pfn_to_page(old_addr >> PAGE_SHIFT);
++ old_page = kexec_pfn_to_page(old_addr >> PAGE_SHIFT);
+ copy_highpage(page, old_page);
+ *old = addr | (*old & ~PAGE_MASK);
+
+@@ -779,7 +807,7 @@
+ result = -ENOMEM;
+ goto out;
+ }
+- result = kimage_add_page(image, page_to_pfn(page)
++ result = kimage_add_page(image, kexec_page_to_pfn(page)
+ << PAGE_SHIFT);
+ if (result < 0)
+ goto out;
+@@ -811,6 +839,7 @@
+ return result;
+ }
+
++#ifndef CONFIG_XEN
+ static int kimage_load_crash_segment(struct kimage *image,
+ struct kexec_segment *segment)
+ {
+@@ -833,7 +862,7 @@
+ char *ptr;
+ size_t uchunk, mchunk;
+
+- page = pfn_to_page(maddr >> PAGE_SHIFT);
++ page = kexec_pfn_to_page(maddr >> PAGE_SHIFT);
+ if (page == 0) {
+ result = -ENOMEM;
+ goto out;
+@@ -851,6 +880,7 @@
+ memset(ptr + uchunk, 0, mchunk - uchunk);
+ }
+ result = copy_from_user(ptr, buf, uchunk);
++ kexec_flush_icache_page(page);
+ kunmap(page);
+ if (result) {
+ result = (result < 0) ? result : -EIO;
+@@ -881,6 +911,13 @@
+
+ return result;
+ }
++#else /* CONFIG_XEN */
++static int kimage_load_segment(struct kimage *image,
++ struct kexec_segment *segment)
++{
++ return kimage_load_normal_segment(image, segment);
++}
++#endif
+
+ /*
+ * Exec Kernel system call: for obvious reasons only root may call it.
+@@ -991,6 +1028,13 @@
+ if (result)
+ goto out;
+ }
++#ifdef CONFIG_XEN
++ if (image) {
++ result = xen_machine_kexec_load(image);
++ if (result)
++ goto out;
++ }
++#endif
+ /* Install the new kernel, and Uninstall the old */
+ image = xchg(dest_image, image);
+
+@@ -1044,7 +1088,6 @@
+ {
+ int locked;
+
+-
+ /* Take the kexec_lock here to prevent sys_kexec_load
+ * running on one cpu from replacing the crash kernel
+ * we are using after a panic on a different cpu.
+diff -r d894e36cfc30 -r 0aa021803deb kernel/resource.c
+--- a/kernel/resource.c Tue Sep 09 11:37:38 2008 +0200
++++ b/kernel/resource.c Wed Sep 10 10:54:08 2008 +0100
+@@ -35,6 +35,16 @@
+ .flags = IORESOURCE_MEM,
+ };
+ EXPORT_SYMBOL(iomem_resource);
+
-+/*
-+ * Bind a fresh port to VIRQ @virq.
-+ * Return allocated port.
-+ */
-+#define IOCTL_EVTCHN_BIND_VIRQ \
-+ _IOC(_IOC_NONE, 'E', 0, sizeof(struct ioctl_evtchn_bind_virq))
-+struct ioctl_evtchn_bind_virq {
-+ unsigned int virq;
++#ifdef CONFIG_PROC_IOMEM_MACHINE
++struct resource iomem_machine_resource = {
++ .name = "Machine PCI mem",
++ .start = 0,
++ .end = -1,
++ .flags = IORESOURCE_MEM,
+};
++EXPORT_SYMBOL(iomem_machine_resource);
++#endif
+
+ static DEFINE_RWLOCK(resource_lock);
+
+@@ -115,6 +125,18 @@
+ return res;
+ }
+
++#ifdef CONFIG_PROC_IOMEM_MACHINE
++static int iomem_machine_open(struct inode *inode, struct file *file)
++{
++ int res = seq_open(file, &resource_op);
++ if (!res) {
++ struct seq_file *m = file->private_data;
++ m->private = &iomem_machine_resource;
++ }
++ return res;
++}
++#endif
+
-+/*
-+ * Bind a fresh port to remote <@remote_domain, @remote_port>.
-+ * Return allocated port.
-+ */
-+#define IOCTL_EVTCHN_BIND_INTERDOMAIN \
-+ _IOC(_IOC_NONE, 'E', 1, sizeof(struct ioctl_evtchn_bind_interdomain))
-+struct ioctl_evtchn_bind_interdomain {
-+ unsigned int remote_domain, remote_port;
+ static struct file_operations proc_ioports_operations = {
+ .open = ioports_open,
+ .read = seq_read,
+@@ -129,6 +151,15 @@
+ .release = seq_release,
+ };
+
++#ifdef CONFIG_PROC_IOMEM_MACHINE
++static struct file_operations proc_iomem_machine_operations = {
++ .open = iomem_machine_open,
++ .read = seq_read,
++ .llseek = seq_lseek,
++ .release = seq_release,
+};
++#endif
++
+ static int __init ioresources_init(void)
+ {
+ struct proc_dir_entry *entry;
+@@ -139,6 +170,13 @@
+ entry = create_proc_entry("iomem", 0, NULL);
+ if (entry)
+ entry->proc_fops = &proc_iomem_operations;
++#ifdef CONFIG_PROC_IOMEM_MACHINE
++ if (is_initial_xendomain()) {
++ entry = create_proc_entry("iomem_machine", 0, NULL);
++ if (entry)
++ entry->proc_fops = &proc_iomem_machine_operations;
++ }
++#endif
+ return 0;
+ }
+ __initcall(ioresources_init);
+diff -r d894e36cfc30 -r 0aa021803deb kernel/softlockup.c
+--- a/kernel/softlockup.c Tue Sep 09 11:37:38 2008 +0200
++++ b/kernel/softlockup.c Wed Sep 10 10:54:08 2008 +0100
+@@ -39,6 +39,19 @@
+ __raw_get_cpu_var(touch_timestamp) = jiffies;
+ }
+ EXPORT_SYMBOL(touch_softlockup_watchdog);
++
++unsigned long softlockup_get_next_event(void)
++{
++ int this_cpu = smp_processor_id();
++ unsigned long touch_timestamp = per_cpu(touch_timestamp, this_cpu);
++
++ if (per_cpu(print_timestamp, this_cpu) == touch_timestamp ||
++ did_panic ||
++ !per_cpu(watchdog_task, this_cpu))
++ return MAX_JIFFY_OFFSET;
+
++ return max_t(long, 0, touch_timestamp + HZ - jiffies);
++}
+
+ /*
+ * This callback runs from the timer interrupt, and checks
+diff -r d894e36cfc30 -r 0aa021803deb kernel/sysctl.c
+--- a/kernel/sysctl.c Tue Sep 09 11:37:38 2008 +0200
++++ b/kernel/sysctl.c Wed Sep 10 10:54:08 2008 +0100
+@@ -661,7 +661,7 @@
+ .proc_handler = &proc_dointvec,
+ },
+ #endif
+-#ifdef CONFIG_ACPI_SLEEP
++#if defined(CONFIG_ACPI_SLEEP) && !defined(CONFIG_ACPI_PV_SLEEP)
+ {
+ .ctl_name = KERN_ACPI_VIDEO_FLAGS,
+ .procname = "acpi_video_flags",
+diff -r d894e36cfc30 -r 0aa021803deb kernel/timer.c
+--- a/kernel/timer.c Tue Sep 09 11:37:38 2008 +0200
++++ b/kernel/timer.c Wed Sep 10 10:54:08 2008 +0100
+@@ -485,7 +485,9 @@
+ if (hr_expires < 3)
+ return hr_expires + jiffies;
+ }
+- hr_expires += jiffies;
++ hr_expires = min_t(unsigned long,
++ softlockup_get_next_event(),
++ hr_expires) + jiffies;
+
+ base = __get_cpu_var(tvec_bases);
+ spin_lock(&base->lock);
+diff -r d894e36cfc30 -r 0aa021803deb lib/Makefile
+--- a/lib/Makefile Tue Sep 09 11:37:38 2008 +0200
++++ b/lib/Makefile Wed Sep 10 10:54:08 2008 +0100
+@@ -52,6 +52,9 @@
+ obj-$(CONFIG_AUDIT_GENERIC) += audit.o
+
+ obj-$(CONFIG_SWIOTLB) += swiotlb.o
++ifeq ($(CONFIG_IA64),y)
++swiotlb-$(CONFIG_XEN) := ../arch/ia64/xen/swiotlb.o
++endif
+
+ hostprogs-y := gen_crc32table
+ clean-files := crc32table.h
+diff -r d894e36cfc30 -r 0aa021803deb lib/swiotlb-xen.c
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/lib/swiotlb-xen.c Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,739 @@
+/*
-+ * Allocate a fresh port for binding to @remote_domain.
-+ * Return allocated port.
++ * Dynamic DMA mapping support.
++ *
++ * This implementation is a fallback for platforms that do not support
++ * I/O TLBs (aka DMA address translation hardware).
++ * Copyright (C) 2000 Asit Mallick <Asit.K.Mallick at intel.com>
++ * Copyright (C) 2000 Goutham Rao <goutham.rao at intel.com>
++ * Copyright (C) 2000, 2003 Hewlett-Packard Co
++ * David Mosberger-Tang <davidm at hpl.hp.com>
++ * Copyright (C) 2005 Keir Fraser <keir at xensource.com>
+ */
-+#define IOCTL_EVTCHN_BIND_UNBOUND_PORT \
-+ _IOC(_IOC_NONE, 'E', 2, sizeof(struct ioctl_evtchn_bind_unbound_port))
-+struct ioctl_evtchn_bind_unbound_port {
-+ unsigned int remote_domain;
-+};
++
++#include <linux/cache.h>
++#include <linux/mm.h>
++#include <linux/module.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include <linux/string.h>
++#include <linux/types.h>
++#include <linux/ctype.h>
++#include <linux/init.h>
++#include <linux/bootmem.h>
++#include <linux/highmem.h>
++#include <asm/io.h>
++#include <asm/pci.h>
++#include <asm/dma.h>
++#include <asm/uaccess.h>
++#include <xen/gnttab.h>
++#include <xen/interface/memory.h>
++#include <asm-i386/mach-xen/asm/gnttab_dma.h>
++
++int swiotlb;
++EXPORT_SYMBOL(swiotlb);
++
++#define OFFSET(val,align) ((unsigned long)((val) & ( (align) - 1)))
+
+/*
-+ * Unbind previously allocated @port.
++ * Maximum allowable number of contiguous slabs to map,
++ * must be a power of 2. What is the appropriate value ?
++ * The complexity of {map,unmap}_single is linearly dependent on this value.
+ */
-+#define IOCTL_EVTCHN_UNBIND \
-+ _IOC(_IOC_NONE, 'E', 3, sizeof(struct ioctl_evtchn_unbind))
-+struct ioctl_evtchn_unbind {
-+ unsigned int port;
-+};
++#define IO_TLB_SEGSIZE 128
+
+/*
-+ * Unbind previously allocated @port.
++ * log of the size of each IO TLB slab. The number of slabs is command line
++ * controllable.
+ */
-+#define IOCTL_EVTCHN_NOTIFY \
-+ _IOC(_IOC_NONE, 'E', 4, sizeof(struct ioctl_evtchn_notify))
-+struct ioctl_evtchn_notify {
-+ unsigned int port;
-+};
++#define IO_TLB_SHIFT 11
+
-+/* Clear and reinitialise the event buffer. Clear error condition. */
-+#define IOCTL_EVTCHN_RESET \
-+ _IOC(_IOC_NONE, 'E', 5, 0)
++int swiotlb_force;
+
-+#endif /* __LINUX_PUBLIC_EVTCHN_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/public/gntdev.h tmp-linux-2.6-xen.patch/include/xen/public/gntdev.h
---- pristine-linux-2.6.18.2/include/xen/public/gntdev.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/public/gntdev.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,105 @@
-+/******************************************************************************
-+ * gntdev.h
-+ *
-+ * Interface to /dev/xen/gntdev.
-+ *
-+ * Copyright (c) 2007, D G Murray
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
-+ */
++static char *iotlb_virt_start;
++static unsigned long iotlb_nslabs;
+
-+#ifndef __LINUX_PUBLIC_GNTDEV_H__
-+#define __LINUX_PUBLIC_GNTDEV_H__
++/*
++ * Used to do a quick range check in swiotlb_unmap_single and
++ * swiotlb_sync_single_*, to see if the memory was in fact allocated by this
++ * API.
++ */
++static unsigned long iotlb_pfn_start, iotlb_pfn_end;
+
-+struct ioctl_gntdev_grant_ref {
-+ /* The domain ID of the grant to be mapped. */
-+ uint32_t domid;
-+ /* The grant reference of the grant to be mapped. */
-+ uint32_t ref;
-+};
++/* Does the given dma address reside within the swiotlb aperture? */
++static inline int in_swiotlb_aperture(dma_addr_t dev_addr)
++{
++ unsigned long pfn = mfn_to_local_pfn(dev_addr >> PAGE_SHIFT);
++ return (pfn_valid(pfn)
++ && (pfn >= iotlb_pfn_start)
++ && (pfn < iotlb_pfn_end));
++}
+
+/*
-+ * Inserts the grant references into the mapping table of an instance
-+ * of gntdev. N.B. This does not perform the mapping, which is deferred
-+ * until mmap() is called with @index as the offset.
++ * When the IOMMU overflows we return a fallback buffer. This sets the size.
+ */
-+#define IOCTL_GNTDEV_MAP_GRANT_REF \
-+_IOC(_IOC_NONE, 'G', 0, sizeof(struct ioctl_gntdev_map_grant_ref))
-+struct ioctl_gntdev_map_grant_ref {
-+ /* IN parameters */
-+ /* The number of grants to be mapped. */
-+ uint32_t count;
-+ uint32_t pad;
-+ /* OUT parameters */
-+ /* The offset to be used on a subsequent call to mmap(). */
-+ uint64_t index;
-+ /* Variable IN parameter. */
-+ /* Array of grant references, of size @count. */
-+ struct ioctl_gntdev_grant_ref refs[1];
-+};
++static unsigned long io_tlb_overflow = 32*1024;
++
++void *io_tlb_overflow_buffer;
+
+/*
-+ * Removes the grant references from the mapping table of an instance of
-+ * of gntdev. N.B. munmap() must be called on the relevant virtual address(es)
-+ * before this ioctl is called, or an error will result.
++ * This is a free list describing the number of free entries available from
++ * each index
+ */
-+#define IOCTL_GNTDEV_UNMAP_GRANT_REF \
-+_IOC(_IOC_NONE, 'G', 1, sizeof(struct ioctl_gntdev_unmap_grant_ref))
-+struct ioctl_gntdev_unmap_grant_ref {
-+ /* IN parameters */
-+ /* The offset was returned by the corresponding map operation. */
-+ uint64_t index;
-+ /* The number of pages to be unmapped. */
-+ uint32_t count;
-+ uint32_t pad;
-+};
++static unsigned int *io_tlb_list;
++static unsigned int io_tlb_index;
+
+/*
-+ * Returns the offset in the driver's address space that corresponds
-+ * to @vaddr. This can be used to perform a munmap(), followed by an
-+ * UNMAP_GRANT_REF ioctl, where no state about the offset is retained by
-+ * the caller. The number of pages that were allocated at the same time as
-+ * @vaddr is returned in @count.
-+ *
-+ * N.B. Where more than one page has been mapped into a contiguous range, the
-+ * supplied @vaddr must correspond to the start of the range; otherwise
-+ * an error will result. It is only possible to munmap() the entire
-+ * contiguously-allocated range at once, and not any subrange thereof.
++ * We need to save away the original address corresponding to a mapped entry
++ * for the sync operations.
+ */
-+#define IOCTL_GNTDEV_GET_OFFSET_FOR_VADDR \
-+_IOC(_IOC_NONE, 'G', 2, sizeof(struct ioctl_gntdev_get_offset_for_vaddr))
-+struct ioctl_gntdev_get_offset_for_vaddr {
-+ /* IN parameters */
-+ /* The virtual address of the first mapped page in a range. */
-+ uint64_t vaddr;
-+ /* OUT parameters */
-+ /* The offset that was used in the initial mmap() operation. */
-+ uint64_t offset;
-+ /* The number of pages mapped in the VM area that begins at @vaddr. */
-+ uint32_t count;
-+ uint32_t pad;
-+};
++static struct phys_addr {
++ struct page *page;
++ unsigned int offset;
++} *io_tlb_orig_addr;
+
-+#endif /* __LINUX_PUBLIC_GNTDEV_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/public/privcmd.h tmp-linux-2.6-xen.patch/include/xen/public/privcmd.h
---- pristine-linux-2.6.18.2/include/xen/public/privcmd.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/public/privcmd.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,79 @@
-+/******************************************************************************
-+ * privcmd.h
-+ *
-+ * Interface to /proc/xen/privcmd.
-+ *
-+ * Copyright (c) 2003-2005, K A Fraser
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
++/*
++ * Protect the above data structures in the map and unmap calls
+ */
++static DEFINE_SPINLOCK(io_tlb_lock);
+
-+#ifndef __LINUX_PUBLIC_PRIVCMD_H__
-+#define __LINUX_PUBLIC_PRIVCMD_H__
-+
-+#include <linux/types.h>
-+
-+#ifndef __user
-+#define __user
-+#endif
-+
-+typedef struct privcmd_hypercall
++static unsigned int dma_bits;
++static unsigned int __initdata max_dma_bits = 32;
++static int __init
++setup_dma_bits(char *str)
+{
-+ __u64 op;
-+ __u64 arg[5];
-+} privcmd_hypercall_t;
-+
-+typedef struct privcmd_mmap_entry {
-+ __u64 va;
-+ __u64 mfn;
-+ __u64 npages;
-+} privcmd_mmap_entry_t;
-+
-+typedef struct privcmd_mmap {
-+ int num;
-+ domid_t dom; /* target domain */
-+ privcmd_mmap_entry_t __user *entry;
-+} privcmd_mmap_t;
++ max_dma_bits = simple_strtoul(str, NULL, 0);
++ return 0;
++}
++__setup("dma_bits=", setup_dma_bits);
+
-+typedef struct privcmd_mmapbatch {
-+ int num; /* number of pages to populate */
-+ domid_t dom; /* target domain */
-+ __u64 addr; /* virtual address */
-+ xen_pfn_t __user *arr; /* array of mfns - top nibble set on err */
-+} privcmd_mmapbatch_t;
++static int __init
++setup_io_tlb_npages(char *str)
++{
++ /* Unlike ia64, the size is aperture in megabytes, not 'slabs'! */
++ if (isdigit(*str)) {
++ iotlb_nslabs = simple_strtoul(str, &str, 0) <<
++ (20 - IO_TLB_SHIFT);
++ iotlb_nslabs = ALIGN(iotlb_nslabs, IO_TLB_SEGSIZE);
++ }
++ if (*str == ',')
++ ++str;
++ /*
++ * NB. 'force' enables the swiotlb, but doesn't force its use for
++ * every DMA like it does on native Linux. 'off' forcibly disables
++ * use of the swiotlb.
++ */
++ if (!strcmp(str, "force"))
++ swiotlb_force = 1;
++ else if (!strcmp(str, "off"))
++ swiotlb_force = -1;
++ return 1;
++}
++__setup("swiotlb=", setup_io_tlb_npages);
++/* make io_tlb_overflow tunable too? */
+
+/*
-+ * @cmd: IOCTL_PRIVCMD_HYPERCALL
-+ * @arg: &privcmd_hypercall_t
-+ * Return: Value returned from execution of the specified hypercall.
-+ */
-+#define IOCTL_PRIVCMD_HYPERCALL \
-+ _IOC(_IOC_NONE, 'P', 0, sizeof(privcmd_hypercall_t))
-+#define IOCTL_PRIVCMD_MMAP \
-+ _IOC(_IOC_NONE, 'P', 2, sizeof(privcmd_mmap_t))
-+#define IOCTL_PRIVCMD_MMAPBATCH \
-+ _IOC(_IOC_NONE, 'P', 3, sizeof(privcmd_mmapbatch_t))
-+
-+#endif /* __LINUX_PUBLIC_PRIVCMD_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/xenbus.h tmp-linux-2.6-xen.patch/include/xen/xenbus.h
---- pristine-linux-2.6.18.2/include/xen/xenbus.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/xenbus.h 2007-09-30 18:06:18.000000000 +0200
-@@ -0,0 +1,304 @@
-+/******************************************************************************
-+ * xenbus.h
-+ *
-+ * Talks to Xen Store to figure out what devices we have.
-+ *
-+ * Copyright (C) 2005 Rusty Russell, IBM Corporation
-+ * Copyright (C) 2005 XenSource Ltd.
-+ *
-+ * This program is free software; you can redistribute it and/or
-+ * modify it under the terms of the GNU General Public License version 2
-+ * as published by the Free Software Foundation; or, when distributed
-+ * separately from the Linux kernel or incorporated into other
-+ * software packages, subject to the following license:
-+ *
-+ * Permission is hereby granted, free of charge, to any person obtaining a copy
-+ * of this source file (the "Software"), to deal in the Software without
-+ * restriction, including without limitation the rights to use, copy, modify,
-+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
-+ * and to permit persons to whom the Software is furnished to do so, subject to
-+ * the following conditions:
-+ *
-+ * The above copyright notice and this permission notice shall be included in
-+ * all copies or substantial portions of the Software.
-+ *
-+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
-+ * IN THE SOFTWARE.
++ * Statically reserve bounce buffer space and initialize bounce buffer data
++ * structures for the software IO TLB used to implement the PCI DMA API.
+ */
++void
++swiotlb_init_with_default_size (size_t default_size)
++{
++ unsigned long i, bytes;
++ int rc;
+
-+#ifndef _XEN_XENBUS_H
-+#define _XEN_XENBUS_H
++ if (!iotlb_nslabs) {
++ iotlb_nslabs = (default_size >> IO_TLB_SHIFT);
++ iotlb_nslabs = ALIGN(iotlb_nslabs, IO_TLB_SEGSIZE);
++ }
+
-+#include <linux/device.h>
-+#include <linux/notifier.h>
-+#include <linux/mutex.h>
-+#include <linux/completion.h>
-+#include <linux/init.h>
-+#include <linux/err.h>
-+#include <xen/interface/xen.h>
-+#include <xen/interface/grant_table.h>
-+#include <xen/interface/io/xenbus.h>
-+#include <xen/interface/io/xs_wire.h>
++ bytes = iotlb_nslabs * (1UL << IO_TLB_SHIFT);
+
-+/* Register callback to watch this node. */
-+struct xenbus_watch
-+{
-+ struct list_head list;
++ /*
++ * Get IO TLB memory from the low pages
++ */
++ iotlb_virt_start = alloc_bootmem_low_pages(bytes);
++ if (!iotlb_virt_start)
++ panic("Cannot allocate SWIOTLB buffer!\n");
+
-+ /* Path being watched. */
-+ const char *node;
++ dma_bits = get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT) + PAGE_SHIFT;
++ for (i = 0; i < iotlb_nslabs; i += IO_TLB_SEGSIZE) {
++ do {
++ rc = xen_create_contiguous_region(
++ (unsigned long)iotlb_virt_start + (i << IO_TLB_SHIFT),
++ get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT),
++ dma_bits);
++ } while (rc && dma_bits++ < max_dma_bits);
++ if (rc) {
++ if (i == 0)
++ panic("No suitable physical memory available for SWIOTLB buffer!\n"
++ "Use dom0_mem Xen boot parameter to reserve\n"
++ "some DMA memory (e.g., dom0_mem=-128M).\n");
++ iotlb_nslabs = i;
++ i <<= IO_TLB_SHIFT;
++ free_bootmem(__pa(iotlb_virt_start + i), bytes - i);
++ bytes = i;
++ for (dma_bits = 0; i > 0; i -= IO_TLB_SEGSIZE << IO_TLB_SHIFT) {
++ unsigned int bits = fls64(virt_to_bus(iotlb_virt_start + i - 1));
+
-+ /* Callback (executed in a process context with no locks held). */
-+ void (*callback)(struct xenbus_watch *,
-+ const char **vec, unsigned int len);
++ if (bits > dma_bits)
++ dma_bits = bits;
++ }
++ break;
++ }
++ }
+
-+ /* See XBWF_ definitions below. */
-+ unsigned long flags;
-+};
++ /*
++ * Allocate and initialize the free list array. This array is used
++ * to find contiguous free memory regions of size up to IO_TLB_SEGSIZE.
++ */
++ io_tlb_list = alloc_bootmem(iotlb_nslabs * sizeof(int));
++ for (i = 0; i < iotlb_nslabs; i++)
++ io_tlb_list[i] = IO_TLB_SEGSIZE - OFFSET(i, IO_TLB_SEGSIZE);
++ io_tlb_index = 0;
++ io_tlb_orig_addr = alloc_bootmem(
++ iotlb_nslabs * sizeof(*io_tlb_orig_addr));
+
-+/*
-+ * Execute callback in its own kthread. Useful if the callback is long
-+ * running or heavily serialised, to avoid taking out the main xenwatch thread
-+ * for a long period of time (or even unwittingly causing a deadlock).
-+ */
-+#define XBWF_new_thread 1
++ /*
++ * Get the overflow emergency buffer
++ */
++ io_tlb_overflow_buffer = alloc_bootmem_low(io_tlb_overflow);
++ if (!io_tlb_overflow_buffer)
++ panic("Cannot allocate SWIOTLB overflow buffer!\n");
+
-+/* A xenbus device. */
-+struct xenbus_device {
-+ const char *devicetype;
-+ const char *nodename;
-+ const char *otherend;
-+ int otherend_id;
-+ struct xenbus_watch otherend_watch;
-+ struct device dev;
-+ enum xenbus_state state;
-+ struct completion down;
-+};
++ do {
++ rc = xen_create_contiguous_region(
++ (unsigned long)io_tlb_overflow_buffer,
++ get_order(io_tlb_overflow),
++ dma_bits);
++ } while (rc && dma_bits++ < max_dma_bits);
++ if (rc)
++ panic("No suitable physical memory available for SWIOTLB overflow buffer!\n");
+
-+static inline struct xenbus_device *to_xenbus_device(struct device *dev)
-+{
-+ return container_of(dev, struct xenbus_device, dev);
++ iotlb_pfn_start = __pa(iotlb_virt_start) >> PAGE_SHIFT;
++ iotlb_pfn_end = iotlb_pfn_start + (bytes >> PAGE_SHIFT);
++
++ printk(KERN_INFO "Software IO TLB enabled: \n"
++ " Aperture: %lu megabytes\n"
++ " Kernel range: %p - %p\n"
++ " Address size: %u bits\n",
++ bytes >> 20,
++ iotlb_virt_start, iotlb_virt_start + bytes,
++ dma_bits);
+}
+
-+struct xenbus_device_id
++void
++swiotlb_init(void)
+{
-+ /* .../device/<device_type>/<identifier> */
-+ char devicetype[32]; /* General class of device. */
-+};
++ long ram_end;
++ size_t defsz = 64 * (1 << 20); /* 64MB default size */
+
-+/* A xenbus driver. */
-+struct xenbus_driver {
-+ char *name;
-+ struct module *owner;
-+ const struct xenbus_device_id *ids;
-+ int (*probe)(struct xenbus_device *dev,
-+ const struct xenbus_device_id *id);
-+ void (*otherend_changed)(struct xenbus_device *dev,
-+ enum xenbus_state backend_state);
-+ int (*remove)(struct xenbus_device *dev);
-+ int (*suspend)(struct xenbus_device *dev);
-+ int (*suspend_cancel)(struct xenbus_device *dev);
-+ int (*resume)(struct xenbus_device *dev);
-+ int (*uevent)(struct xenbus_device *, char **, int, char *, int);
-+ struct device_driver driver;
-+ int (*read_otherend_details)(struct xenbus_device *dev);
-+ int (*is_ready)(struct xenbus_device *dev);
-+};
++ if (swiotlb_force == 1) {
++ swiotlb = 1;
++ } else if ((swiotlb_force != -1) &&
++ is_running_on_xen() &&
++ is_initial_xendomain()) {
++ /* Domain 0 always has a swiotlb. */
++ ram_end = HYPERVISOR_memory_op(XENMEM_maximum_ram_page, NULL);
++ if (ram_end <= 0x7ffff)
++ defsz = 2 * (1 << 20); /* 2MB on <2GB on systems. */
++ swiotlb = 1;
++ }
+
-+static inline struct xenbus_driver *to_xenbus_driver(struct device_driver *drv)
-+{
-+ return container_of(drv, struct xenbus_driver, driver);
++ if (swiotlb)
++ swiotlb_init_with_default_size(defsz);
++ else
++ printk(KERN_INFO "Software IO TLB disabled\n");
+}
+
-+int xenbus_register_frontend(struct xenbus_driver *drv);
-+int xenbus_register_backend(struct xenbus_driver *drv);
-+void xenbus_unregister_driver(struct xenbus_driver *drv);
-+
-+struct xenbus_transaction
++/*
++ * We use __copy_to_user_inatomic to transfer to the host buffer because the
++ * buffer may be mapped read-only (e.g, in blkback driver) but lower-level
++ * drivers map the buffer for DMA_BIDIRECTIONAL access. This causes an
++ * unnecessary copy from the aperture to the host buffer, and a page fault.
++ */
++static void
++__sync_single(struct phys_addr buffer, char *dma_addr, size_t size, int dir)
+{
-+ u32 id;
-+};
-+
-+/* Nil transaction ID. */
-+#define XBT_NIL ((struct xenbus_transaction) { 0 })
-+
-+char **xenbus_directory(struct xenbus_transaction t,
-+ const char *dir, const char *node, unsigned int *num);
-+void *xenbus_read(struct xenbus_transaction t,
-+ const char *dir, const char *node, unsigned int *len);
-+int xenbus_write(struct xenbus_transaction t,
-+ const char *dir, const char *node, const char *string);
-+int xenbus_mkdir(struct xenbus_transaction t,
-+ const char *dir, const char *node);
-+int xenbus_exists(struct xenbus_transaction t,
-+ const char *dir, const char *node);
-+int xenbus_rm(struct xenbus_transaction t, const char *dir, const char *node);
-+int xenbus_transaction_start(struct xenbus_transaction *t);
-+int xenbus_transaction_end(struct xenbus_transaction t, int abort);
++ if (PageHighMem(buffer.page)) {
++ size_t len, bytes;
++ char *dev, *host, *kmp;
++ len = size;
++ while (len != 0) {
++ unsigned long flags;
+
-+/* Single read and scanf: returns -errno or num scanned if > 0. */
-+int xenbus_scanf(struct xenbus_transaction t,
-+ const char *dir, const char *node, const char *fmt, ...)
-+ __attribute__((format(scanf, 4, 5)));
++ if (((bytes = len) + buffer.offset) > PAGE_SIZE)
++ bytes = PAGE_SIZE - buffer.offset;
++ local_irq_save(flags); /* protects KM_BOUNCE_READ */
++ kmp = kmap_atomic(buffer.page, KM_BOUNCE_READ);
++ dev = dma_addr + size - len;
++ host = kmp + buffer.offset;
++ if (dir == DMA_FROM_DEVICE) {
++ if (__copy_to_user_inatomic(host, dev, bytes))
++ /* inaccessible */;
++ } else
++ memcpy(dev, host, bytes);
++ kunmap_atomic(kmp, KM_BOUNCE_READ);
++ local_irq_restore(flags);
++ len -= bytes;
++ buffer.page++;
++ buffer.offset = 0;
++ }
++ } else {
++ char *host = (char *)phys_to_virt(
++ page_to_pseudophys(buffer.page)) + buffer.offset;
++ if (dir == DMA_FROM_DEVICE) {
++ if (__copy_to_user_inatomic(host, dma_addr, size))
++ /* inaccessible */;
++ } else if (dir == DMA_TO_DEVICE)
++ memcpy(dma_addr, host, size);
++ }
++}
+
-+/* Single printf and write: returns -errno or 0. */
-+int xenbus_printf(struct xenbus_transaction t,
-+ const char *dir, const char *node, const char *fmt, ...)
-+ __attribute__((format(printf, 4, 5)));
++/*
++ * Allocates bounce buffer and returns its kernel virtual address.
++ */
++static void *
++map_single(struct device *hwdev, struct phys_addr buffer, size_t size, int dir)
++{
++ unsigned long flags;
++ char *dma_addr;
++ unsigned int nslots, stride, index, wrap;
++ struct phys_addr slot_buf;
++ int i;
+
-+/* Generic read function: NULL-terminated triples of name,
-+ * sprintf-style type string, and pointer. Returns 0 or errno.*/
-+int xenbus_gather(struct xenbus_transaction t, const char *dir, ...);
++ /*
++ * For mappings greater than a page, we limit the stride (and
++ * hence alignment) to a page size.
++ */
++ nslots = ALIGN(size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT;
++ if (size > PAGE_SIZE)
++ stride = (1 << (PAGE_SHIFT - IO_TLB_SHIFT));
++ else
++ stride = 1;
+
-+/* notifer routines for when the xenstore comes up */
-+int register_xenstore_notifier(struct notifier_block *nb);
-+void unregister_xenstore_notifier(struct notifier_block *nb);
++ BUG_ON(!nslots);
+
-+int register_xenbus_watch(struct xenbus_watch *watch);
-+void unregister_xenbus_watch(struct xenbus_watch *watch);
-+void xs_suspend(void);
-+void xs_resume(void);
-+void xs_suspend_cancel(void);
++ /*
++ * Find suitable number of IO TLB entries size that will fit this
++ * request and allocate a buffer from that IO TLB pool.
++ */
++ spin_lock_irqsave(&io_tlb_lock, flags);
++ {
++ wrap = index = ALIGN(io_tlb_index, stride);
+
-+/* Used by xenbus_dev to borrow kernel's store connection. */
-+void *xenbus_dev_request_and_reply(struct xsd_sockmsg *msg);
++ if (index >= iotlb_nslabs)
++ wrap = index = 0;
+
-+/* Prepare for domain suspend: then resume or cancel the suspend. */
-+void xenbus_suspend(void);
-+void xenbus_resume(void);
-+void xenbus_suspend_cancel(void);
++ do {
++ /*
++ * If we find a slot that indicates we have 'nslots'
++ * number of contiguous buffers, we allocate the
++ * buffers from that slot and mark the entries as '0'
++ * indicating unavailable.
++ */
++ if (io_tlb_list[index] >= nslots) {
++ int count = 0;
+
-+#define XENBUS_IS_ERR_READ(str) ({ \
-+ if (!IS_ERR(str) && strlen(str) == 0) { \
-+ kfree(str); \
-+ str = ERR_PTR(-ERANGE); \
-+ } \
-+ IS_ERR(str); \
-+})
++ for (i = index; i < (int)(index + nslots); i++)
++ io_tlb_list[i] = 0;
++ for (i = index - 1;
++ (OFFSET(i, IO_TLB_SEGSIZE) !=
++ IO_TLB_SEGSIZE -1) && io_tlb_list[i];
++ i--)
++ io_tlb_list[i] = ++count;
++ dma_addr = iotlb_virt_start +
++ (index << IO_TLB_SHIFT);
+
-+#define XENBUS_EXIST_ERR(err) ((err) == -ENOENT || (err) == -ERANGE)
++ /*
++ * Update the indices to avoid searching in
++ * the next round.
++ */
++ io_tlb_index =
++ ((index + nslots) < iotlb_nslabs
++ ? (index + nslots) : 0);
+
++ goto found;
++ }
++ index += stride;
++ if (index >= iotlb_nslabs)
++ index = 0;
++ } while (index != wrap);
+
-+/**
-+ * Register a watch on the given path, using the given xenbus_watch structure
-+ * for storage, and the given callback function as the callback. Return 0 on
-+ * success, or -errno on error. On success, the given path will be saved as
-+ * watch->node, and remains the caller's to free. On error, watch->node will
-+ * be NULL, the device will switch to XenbusStateClosing, and the error will
-+ * be saved in the store.
-+ */
-+int xenbus_watch_path(struct xenbus_device *dev, const char *path,
-+ struct xenbus_watch *watch,
-+ void (*callback)(struct xenbus_watch *,
-+ const char **, unsigned int));
++ spin_unlock_irqrestore(&io_tlb_lock, flags);
++ return NULL;
++ }
++ found:
++ spin_unlock_irqrestore(&io_tlb_lock, flags);
+
++ /*
++ * Save away the mapping from the original address to the DMA address.
++ * This is needed when we sync the memory. Then we sync the buffer if
++ * needed.
++ */
++ slot_buf = buffer;
++ for (i = 0; i < nslots; i++) {
++ slot_buf.page += slot_buf.offset >> PAGE_SHIFT;
++ slot_buf.offset &= PAGE_SIZE - 1;
++ io_tlb_orig_addr[index+i] = slot_buf;
++ slot_buf.offset += 1 << IO_TLB_SHIFT;
++ }
++ if ((dir == DMA_TO_DEVICE) || (dir == DMA_BIDIRECTIONAL))
++ __sync_single(buffer, dma_addr, size, DMA_TO_DEVICE);
+
-+/**
-+ * Register a watch on the given path/path2, using the given xenbus_watch
-+ * structure for storage, and the given callback function as the callback.
-+ * Return 0 on success, or -errno on error. On success, the watched path
-+ * (path/path2) will be saved as watch->node, and becomes the caller's to
-+ * kfree(). On error, watch->node will be NULL, so the caller has nothing to
-+ * free, the device will switch to XenbusStateClosing, and the error will be
-+ * saved in the store.
-+ */
-+int xenbus_watch_path2(struct xenbus_device *dev, const char *path,
-+ const char *path2, struct xenbus_watch *watch,
-+ void (*callback)(struct xenbus_watch *,
-+ const char **, unsigned int));
++ return dma_addr;
++}
+
++static struct phys_addr dma_addr_to_phys_addr(char *dma_addr)
++{
++ int index = (dma_addr - iotlb_virt_start) >> IO_TLB_SHIFT;
++ struct phys_addr buffer = io_tlb_orig_addr[index];
++ buffer.offset += (long)dma_addr & ((1 << IO_TLB_SHIFT) - 1);
++ buffer.page += buffer.offset >> PAGE_SHIFT;
++ buffer.offset &= PAGE_SIZE - 1;
++ return buffer;
++}
+
-+/**
-+ * Advertise in the store a change of the given driver to the given new_state.
-+ * Return 0 on success, or -errno on error. On error, the device will switch
-+ * to XenbusStateClosing, and the error will be saved in the store.
++/*
++ * dma_addr is the kernel virtual address of the bounce buffer to unmap.
+ */
-+int xenbus_switch_state(struct xenbus_device *dev, enum xenbus_state new_state);
++static void
++unmap_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
++{
++ unsigned long flags;
++ int i, count, nslots = ALIGN(size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT;
++ int index = (dma_addr - iotlb_virt_start) >> IO_TLB_SHIFT;
++ struct phys_addr buffer = dma_addr_to_phys_addr(dma_addr);
+
++ /*
++ * First, sync the memory before unmapping the entry
++ */
++ if ((dir == DMA_FROM_DEVICE) || (dir == DMA_BIDIRECTIONAL))
++ __sync_single(buffer, dma_addr, size, DMA_FROM_DEVICE);
+
-+/**
-+ * Grant access to the given ring_mfn to the peer of the given device. Return
-+ * 0 on success, or -errno on error. On error, the device will switch to
-+ * XenbusStateClosing, and the error will be saved in the store.
-+ */
-+int xenbus_grant_ring(struct xenbus_device *dev, unsigned long ring_mfn);
++ /*
++ * Return the buffer to the free list by setting the corresponding
++ * entries to indicate the number of contigous entries available.
++ * While returning the entries to the free list, we merge the entries
++ * with slots below and above the pool being returned.
++ */
++ spin_lock_irqsave(&io_tlb_lock, flags);
++ {
++ count = ((index + nslots) < ALIGN(index + 1, IO_TLB_SEGSIZE) ?
++ io_tlb_list[index + nslots] : 0);
++ /*
++ * Step 1: return the slots to the free list, merging the
++ * slots with superceeding slots
++ */
++ for (i = index + nslots - 1; i >= index; i--)
++ io_tlb_list[i] = ++count;
++ /*
++ * Step 2: merge the returned slots with the preceding slots,
++ * if available (non zero)
++ */
++ for (i = index - 1;
++ (OFFSET(i, IO_TLB_SEGSIZE) !=
++ IO_TLB_SEGSIZE -1) && io_tlb_list[i];
++ i--)
++ io_tlb_list[i] = ++count;
++ }
++ spin_unlock_irqrestore(&io_tlb_lock, flags);
++}
+
++static void
++sync_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
++{
++ struct phys_addr buffer = dma_addr_to_phys_addr(dma_addr);
++ BUG_ON((dir != DMA_FROM_DEVICE) && (dir != DMA_TO_DEVICE));
++ __sync_single(buffer, dma_addr, size, dir);
++}
+
-+/**
-+ * Map a page of memory into this domain from another domain's grant table.
-+ * xenbus_map_ring_valloc allocates a page of virtual address space, maps the
-+ * page to that address, and sets *vaddr to that address.
-+ * xenbus_map_ring does not allocate the virtual address space (you must do
-+ * this yourself!). It only maps in the page to the specified address.
-+ * Returns 0 on success, and GNTST_* (see xen/include/interface/grant_table.h)
-+ * or -ENOMEM on error. If an error is returned, device will switch to
-+ * XenbusStateClosing and the error message will be saved in XenStore.
-+ */
-+struct vm_struct *xenbus_map_ring_valloc(struct xenbus_device *dev,
-+ int gnt_ref);
-+int xenbus_map_ring(struct xenbus_device *dev, int gnt_ref,
-+ grant_handle_t *handle, void *vaddr);
++static void
++swiotlb_full(struct device *dev, size_t size, int dir, int do_panic)
++{
++ /*
++ * Ran out of IOMMU space for this operation. This is very bad.
++ * Unfortunately the drivers cannot handle this operation properly.
++ * unless they check for pci_dma_mapping_error (most don't)
++ * When the mapping is small enough return a static buffer to limit
++ * the damage, or panic when the transfer is too big.
++ */
++ printk(KERN_ERR "PCI-DMA: Out of SW-IOMMU space for %lu bytes at "
++ "device %s\n", (unsigned long)size, dev ? dev->bus_id : "?");
+
++ if (size > io_tlb_overflow && do_panic) {
++ if (dir == PCI_DMA_FROMDEVICE || dir == PCI_DMA_BIDIRECTIONAL)
++ panic("PCI-DMA: Memory would be corrupted\n");
++ if (dir == PCI_DMA_TODEVICE || dir == PCI_DMA_BIDIRECTIONAL)
++ panic("PCI-DMA: Random memory would be DMAed\n");
++ }
++}
+
-+/**
-+ * Unmap a page of memory in this domain that was imported from another domain.
-+ * Use xenbus_unmap_ring_vfree if you mapped in your memory with
-+ * xenbus_map_ring_valloc (it will free the virtual address space).
-+ * Returns 0 on success and returns GNTST_* on error
-+ * (see xen/include/interface/grant_table.h).
++/*
++ * Map a single buffer of the indicated size for DMA in streaming mode. The
++ * PCI address to use is returned.
++ *
++ * Once the device is given the dma address, the device owns this memory until
++ * either swiotlb_unmap_single or swiotlb_dma_sync_single is performed.
+ */
-+int xenbus_unmap_ring_vfree(struct xenbus_device *dev, struct vm_struct *);
-+int xenbus_unmap_ring(struct xenbus_device *dev,
-+ grant_handle_t handle, void *vaddr);
-+
++dma_addr_t
++swiotlb_map_single(struct device *hwdev, void *ptr, size_t size, int dir)
++{
++ dma_addr_t dev_addr = gnttab_dma_map_page(virt_to_page(ptr)) +
++ offset_in_page(ptr);
++ void *map;
++ struct phys_addr buffer;
+
-+/**
-+ * Allocate an event channel for the given xenbus_device, assigning the newly
-+ * created local port to *port. Return 0 on success, or -errno on error. On
-+ * error, the device will switch to XenbusStateClosing, and the error will be
-+ * saved in the store.
-+ */
-+int xenbus_alloc_evtchn(struct xenbus_device *dev, int *port);
++ BUG_ON(dir == DMA_NONE);
+
++ /*
++ * If the pointer passed in happens to be in the device's DMA window,
++ * we can safely return the device addr and not worry about bounce
++ * buffering it.
++ */
++ if (!range_straddles_page_boundary(__pa(ptr), size) &&
++ !address_needs_mapping(hwdev, dev_addr))
++ return dev_addr;
+
-+/**
-+ * Free an existing event channel. Returns 0 on success or -errno on error.
-+ */
-+int xenbus_free_evtchn(struct xenbus_device *dev, int port);
++ /*
++ * Oh well, have to allocate and map a bounce buffer.
++ */
++ gnttab_dma_unmap_page(dev_addr);
++ buffer.page = virt_to_page(ptr);
++ buffer.offset = (unsigned long)ptr & ~PAGE_MASK;
++ map = map_single(hwdev, buffer, size, dir);
++ if (!map) {
++ swiotlb_full(hwdev, size, dir, 1);
++ map = io_tlb_overflow_buffer;
++ }
+
++ dev_addr = virt_to_bus(map);
++ return dev_addr;
++}
+
-+/**
-+ * Return the state of the driver rooted at the given store path, or
-+ * XenbusStateUnknown if no state can be read.
++/*
++ * Unmap a single streaming mode DMA translation. The dma_addr and size must
++ * match what was provided for in a previous swiotlb_map_single call. All
++ * other usages are undefined.
++ *
++ * After this call, reads by the cpu to the buffer are guaranteed to see
++ * whatever the device wrote there.
+ */
-+enum xenbus_state xenbus_read_driver_state(const char *path);
-+
++void
++swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr, size_t size,
++ int dir)
++{
++ BUG_ON(dir == DMA_NONE);
++ if (in_swiotlb_aperture(dev_addr))
++ unmap_single(hwdev, bus_to_virt(dev_addr), size, dir);
++ else
++ gnttab_dma_unmap_page(dev_addr);
++}
+
-+/***
-+ * Report the given negative errno into the store, along with the given
-+ * formatted message.
++/*
++ * Make physical memory consistent for a single streaming mode DMA translation
++ * after a transfer.
++ *
++ * If you perform a swiotlb_map_single() but wish to interrogate the buffer
++ * using the cpu, yet do not wish to teardown the PCI dma mapping, you must
++ * call this function before doing so. At the next point you give the PCI dma
++ * address back to the card, you must first perform a
++ * swiotlb_dma_sync_for_device, and then the device again owns the buffer
+ */
-+void xenbus_dev_error(struct xenbus_device *dev, int err, const char *fmt,
-+ ...);
++void
++swiotlb_sync_single_for_cpu(struct device *hwdev, dma_addr_t dev_addr,
++ size_t size, int dir)
++{
++ BUG_ON(dir == DMA_NONE);
++ if (in_swiotlb_aperture(dev_addr))
++ sync_single(hwdev, bus_to_virt(dev_addr), size, dir);
++}
+
++void
++swiotlb_sync_single_for_device(struct device *hwdev, dma_addr_t dev_addr,
++ size_t size, int dir)
++{
++ BUG_ON(dir == DMA_NONE);
++ if (in_swiotlb_aperture(dev_addr))
++ sync_single(hwdev, bus_to_virt(dev_addr), size, dir);
++}
+
-+/***
-+ * Equivalent to xenbus_dev_error(dev, err, fmt, args), followed by
-+ * xenbus_switch_state(dev, NULL, XenbusStateClosing) to schedule an orderly
-+ * closedown of this driver and its peer.
++/*
++ * Map a set of buffers described by scatterlist in streaming mode for DMA.
++ * This is the scatter-gather version of the above swiotlb_map_single
++ * interface. Here the scatter gather list elements are each tagged with the
++ * appropriate dma address and length. They are obtained via
++ * sg_dma_{address,length}(SG).
++ *
++ * NOTE: An implementation may be able to use a smaller number of
++ * DMA address/length pairs than there are SG table elements.
++ * (for example via virtual mapping capabilities)
++ * The routine returns the number of addr/length pairs actually
++ * used, at most nents.
++ *
++ * Device ownership issues as mentioned above for swiotlb_map_single are the
++ * same here.
+ */
-+void xenbus_dev_fatal(struct xenbus_device *dev, int err, const char *fmt,
-+ ...);
-+
-+int xenbus_dev_init(void);
-+
-+const char *xenbus_strstate(enum xenbus_state state);
-+int xenbus_dev_is_online(struct xenbus_device *dev);
-+int xenbus_frontend_closed(struct xenbus_device *dev);
++int
++swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg, int nelems,
++ int dir)
++{
++ struct phys_addr buffer;
++ dma_addr_t dev_addr;
++ char *map;
++ int i;
+
-+#endif /* _XEN_XENBUS_H */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/xencons.h tmp-linux-2.6-xen.patch/include/xen/xencons.h
---- pristine-linux-2.6.18.2/include/xen/xencons.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/xencons.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,19 @@
-+#ifndef __ASM_XENCONS_H__
-+#define __ASM_XENCONS_H__
++ BUG_ON(dir == DMA_NONE);
+
-+struct dom0_vga_console_info;
-+void dom0_init_screen_info(const struct dom0_vga_console_info *info);
++ for (i = 0; i < nelems; i++, sg++) {
++ dev_addr = gnttab_dma_map_page(sg->page) + sg->offset;
+
-+void xencons_force_flush(void);
-+void xencons_resume(void);
++ if (range_straddles_page_boundary(page_to_pseudophys(sg->page)
++ + sg->offset, sg->length)
++ || address_needs_mapping(hwdev, dev_addr)) {
++ gnttab_dma_unmap_page(dev_addr);
++ buffer.page = sg->page;
++ buffer.offset = sg->offset;
++ map = map_single(hwdev, buffer, sg->length, dir);
++ if (!map) {
++ /* Don't panic here, we expect map_sg users
++ to do proper error handling. */
++ swiotlb_full(hwdev, sg->length, dir, 0);
++ swiotlb_unmap_sg(hwdev, sg - i, i, dir);
++ sg[0].dma_length = 0;
++ return 0;
++ }
++ sg->dma_address = (dma_addr_t)virt_to_bus(map);
++ } else
++ sg->dma_address = dev_addr;
++ sg->dma_length = sg->length;
++ }
++ return nelems;
++}
+
-+/* Interrupt work hooks. Receive data, or kick data out. */
-+void xencons_rx(char *buf, unsigned len, struct pt_regs *regs);
-+void xencons_tx(void);
++/*
++ * Unmap a set of streaming mode DMA translations. Again, cpu read rules
++ * concerning calls here are the same as for swiotlb_unmap_single() above.
++ */
++void
++swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nelems,
++ int dir)
++{
++ int i;
+
-+int xencons_ring_init(void);
-+int xencons_ring_send(const char *data, unsigned len);
++ BUG_ON(dir == DMA_NONE);
+
-+void xencons_early_setup(void);
++ for (i = 0; i < nelems; i++, sg++)
++ if (in_swiotlb_aperture(sg->dma_address))
++ unmap_single(hwdev,
++ (void *)bus_to_virt(sg->dma_address),
++ sg->dma_length, dir);
++ else
++ gnttab_dma_unmap_page(sg->dma_address);
++}
+
-+#endif /* __ASM_XENCONS_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/xenoprof.h tmp-linux-2.6-xen.patch/include/xen/xenoprof.h
---- pristine-linux-2.6.18.2/include/xen/xenoprof.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/xenoprof.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,42 @@
-+/******************************************************************************
-+ * xen/xenoprof.h
-+ *
-+ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
-+ * VA Linux Systems Japan K.K.
-+ *
-+ * This program is free software; you can redistribute it and/or modify
-+ * it under the terms of the GNU General Public License as published by
-+ * the Free Software Foundation; either version 2 of the License, or
-+ * (at your option) any later version.
-+ *
-+ * This program is distributed in the hope that it will be useful,
-+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
-+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-+ * GNU General Public License for more details.
-+ *
-+ * You should have received a copy of the GNU General Public License
-+ * along with this program; if not, write to the Free Software
-+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++/*
++ * Make physical memory consistent for a set of streaming mode DMA translations
++ * after a transfer.
+ *
++ * The same as swiotlb_sync_single_* but for a scatter-gather list, same rules
++ * and usage.
+ */
++void
++swiotlb_sync_sg_for_cpu(struct device *hwdev, struct scatterlist *sg,
++ int nelems, int dir)
++{
++ int i;
+
-+#ifndef __XEN_XENOPROF_H__
-+#define __XEN_XENOPROF_H__
-+#ifdef CONFIG_XEN
-+
-+#include <asm/xenoprof.h>
++ BUG_ON(dir == DMA_NONE);
+
-+struct oprofile_operations;
-+int xenoprofile_init(struct oprofile_operations * ops);
-+void xenoprofile_exit(void);
++ for (i = 0; i < nelems; i++, sg++)
++ if (in_swiotlb_aperture(sg->dma_address))
++ sync_single(hwdev,
++ (void *)bus_to_virt(sg->dma_address),
++ sg->dma_length, dir);
++}
+
-+struct xenoprof_shared_buffer {
-+ char *buffer;
-+ struct xenoprof_arch_shared_buffer arch;
-+};
-+#else
-+#define xenoprofile_init(ops) (-ENOSYS)
-+#define xenoprofile_exit() do { } while (0)
++void
++swiotlb_sync_sg_for_device(struct device *hwdev, struct scatterlist *sg,
++ int nelems, int dir)
++{
++ int i;
+
-+#endif /* CONFIG_XEN */
-+#endif /* __XEN_XENOPROF_H__ */
-diff -Nurp pristine-linux-2.6.18.2/include/xen/xen_proc.h tmp-linux-2.6-xen.patch/include/xen/xen_proc.h
---- pristine-linux-2.6.18.2/include/xen/xen_proc.h 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/include/xen/xen_proc.h 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,12 @@
++ BUG_ON(dir == DMA_NONE);
+
-+#ifndef __ASM_XEN_PROC_H__
-+#define __ASM_XEN_PROC_H__
++ for (i = 0; i < nelems; i++, sg++)
++ if (in_swiotlb_aperture(sg->dma_address))
++ sync_single(hwdev,
++ (void *)bus_to_virt(sg->dma_address),
++ sg->dma_length, dir);
++}
+
-+#include <linux/proc_fs.h>
++#ifdef CONFIG_HIGHMEM
+
-+extern struct proc_dir_entry *create_xen_proc_entry(
-+ const char *name, mode_t mode);
-+extern void remove_xen_proc_entry(
-+ const char *name);
++dma_addr_t
++swiotlb_map_page(struct device *hwdev, struct page *page,
++ unsigned long offset, size_t size,
++ enum dma_data_direction direction)
++{
++ struct phys_addr buffer;
++ dma_addr_t dev_addr;
++ char *map;
+
-+#endif /* __ASM_XEN_PROC_H__ */
-diff -Nurp pristine-linux-2.6.18.2/kernel/fork.c tmp-linux-2.6-xen.patch/kernel/fork.c
---- pristine-linux-2.6.18.2/kernel/fork.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/kernel/fork.c 2007-07-30 16:35:13.000000000 +0200
-@@ -276,6 +276,9 @@ static inline int dup_mmap(struct mm_str
- if (retval)
- goto out;
- }
-+#ifdef arch_dup_mmap
-+ arch_dup_mmap(mm, oldmm);
-+#endif
- retval = 0;
- out:
- up_write(&mm->mmap_sem);
-diff -Nurp pristine-linux-2.6.18.2/kernel/irq/spurious.c tmp-linux-2.6-xen.patch/kernel/irq/spurious.c
---- pristine-linux-2.6.18.2/kernel/irq/spurious.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/kernel/irq/spurious.c 2007-07-30 16:35:13.000000000 +0200
-@@ -139,7 +139,8 @@ void note_interrupt(unsigned int irq, st
- irqreturn_t action_ret, struct pt_regs *regs)
- {
- if (unlikely(action_ret != IRQ_HANDLED)) {
-- desc->irqs_unhandled++;
-+ if (!irq_ignore_unhandled(irq))
-+ desc->irqs_unhandled++;
- if (unlikely(action_ret != IRQ_NONE))
- report_bad_irq(irq, desc, action_ret);
- }
-diff -Nurp pristine-linux-2.6.18.2/kernel/Kconfig.preempt tmp-linux-2.6-xen.patch/kernel/Kconfig.preempt
---- pristine-linux-2.6.18.2/kernel/Kconfig.preempt 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/kernel/Kconfig.preempt 2007-07-30 16:35:13.000000000 +0200
-@@ -35,6 +35,7 @@ config PREEMPT_VOLUNTARY
-
- config PREEMPT
- bool "Preemptible Kernel (Low-Latency Desktop)"
-+ depends on !XEN
- help
- This option reduces the latency of the kernel by making
- all kernel code (that is not executing in a critical section)
-diff -Nurp pristine-linux-2.6.18.2/kernel/kexec.c tmp-linux-2.6-xen.patch/kernel/kexec.c
---- pristine-linux-2.6.18.2/kernel/kexec.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/kernel/kexec.c 2007-09-30 18:06:18.000000000 +0200
-@@ -330,13 +330,26 @@ static int kimage_is_destination_range(s
- return 0;
- }
-
--static struct page *kimage_alloc_pages(gfp_t gfp_mask, unsigned int order)
-+static struct page *kimage_alloc_pages(gfp_t gfp_mask, unsigned int order, unsigned long limit)
- {
- struct page *pages;
-
- pages = alloc_pages(gfp_mask, order);
- if (pages) {
- unsigned int count, i;
-+#ifdef CONFIG_XEN
-+ int address_bits;
++ dev_addr = gnttab_dma_map_page(page) + offset;
++ if (address_needs_mapping(hwdev, dev_addr)) {
++ gnttab_dma_unmap_page(dev_addr);
++ buffer.page = page;
++ buffer.offset = offset;
++ map = map_single(hwdev, buffer, size, direction);
++ if (!map) {
++ swiotlb_full(hwdev, size, direction, 1);
++ map = io_tlb_overflow_buffer;
++ }
++ dev_addr = (dma_addr_t)virt_to_bus(map);
++ }
+
-+ if (limit == ~0UL)
-+ address_bits = BITS_PER_LONG;
-+ else
-+ address_bits = long_log2(limit);
++ return dev_addr;
++}
+
-+ if (xen_limit_pages_to_max_mfn(pages, order, address_bits) < 0) {
-+ __free_pages(pages, order);
-+ return NULL;
-+ }
-+#endif
- pages->mapping = NULL;
- set_page_private(pages, order);
- count = 1 << order;
-@@ -355,6 +368,9 @@ static void kimage_free_pages(struct pag
- count = 1 << order;
- for (i = 0; i < count; i++)
- ClearPageReserved(page + i);
-+#ifdef CONFIG_XEN
-+ xen_destroy_contiguous_region((unsigned long)page_address(page), order);
-+#endif
- __free_pages(page, order);
- }
-
-@@ -400,10 +416,10 @@ static struct page *kimage_alloc_normal_
- do {
- unsigned long pfn, epfn, addr, eaddr;
-
-- pages = kimage_alloc_pages(GFP_KERNEL, order);
-+ pages = kimage_alloc_pages(GFP_KERNEL, order, KEXEC_CONTROL_MEMORY_LIMIT);
- if (!pages)
- break;
-- pfn = page_to_pfn(pages);
-+ pfn = kexec_page_to_pfn(pages);
- epfn = pfn + count;
- addr = pfn << PAGE_SHIFT;
- eaddr = epfn << PAGE_SHIFT;
-@@ -437,6 +453,7 @@ static struct page *kimage_alloc_normal_
- return pages;
- }
-
-+#ifndef CONFIG_XEN
- static struct page *kimage_alloc_crash_control_pages(struct kimage *image,
- unsigned int order)
- {
-@@ -490,7 +507,7 @@ static struct page *kimage_alloc_crash_c
- }
- /* If I don't overlap any segments I have found my hole! */
- if (i == image->nr_segments) {
-- pages = pfn_to_page(hole_start >> PAGE_SHIFT);
-+ pages = kexec_pfn_to_page(hole_start >> PAGE_SHIFT);
- break;
- }
- }
-@@ -517,6 +534,13 @@ struct page *kimage_alloc_control_pages(
-
- return pages;
- }
-+#else /* !CONFIG_XEN */
-+struct page *kimage_alloc_control_pages(struct kimage *image,
-+ unsigned int order)
++void
++swiotlb_unmap_page(struct device *hwdev, dma_addr_t dma_address,
++ size_t size, enum dma_data_direction direction)
+{
-+ return kimage_alloc_normal_control_pages(image, order);
++ BUG_ON(direction == DMA_NONE);
++ if (in_swiotlb_aperture(dma_address))
++ unmap_single(hwdev, bus_to_virt(dma_address), size, direction);
++ else
++ gnttab_dma_unmap_page(dma_address);
+}
-+#endif
-
- static int kimage_add_entry(struct kimage *image, kimage_entry_t entry)
- {
-@@ -532,7 +556,7 @@ static int kimage_add_entry(struct kimag
- return -ENOMEM;
-
- ind_page = page_address(page);
-- *image->entry = virt_to_phys(ind_page) | IND_INDIRECTION;
-+ *image->entry = kexec_virt_to_phys(ind_page) | IND_INDIRECTION;
- image->entry = ind_page;
- image->last_entry = ind_page +
- ((PAGE_SIZE/sizeof(kimage_entry_t)) - 1);
-@@ -593,13 +617,13 @@ static int kimage_terminate(struct kimag
- #define for_each_kimage_entry(image, ptr, entry) \
- for (ptr = &image->head; (entry = *ptr) && !(entry & IND_DONE); \
- ptr = (entry & IND_INDIRECTION)? \
-- phys_to_virt((entry & PAGE_MASK)): ptr +1)
-+ kexec_phys_to_virt((entry & PAGE_MASK)): ptr +1)
-
- static void kimage_free_entry(kimage_entry_t entry)
- {
- struct page *page;
-
-- page = pfn_to_page(entry >> PAGE_SHIFT);
-+ page = kexec_pfn_to_page(entry >> PAGE_SHIFT);
- kimage_free_pages(page);
- }
-
-@@ -611,6 +635,10 @@ static void kimage_free(struct kimage *i
- if (!image)
- return;
-
-+#ifdef CONFIG_XEN
-+ xen_machine_kexec_unload(image);
++
+#endif
+
- kimage_free_extra_pages(image);
- for_each_kimage_entry(image, ptr, entry) {
- if (entry & IND_INDIRECTION) {
-@@ -686,7 +714,7 @@ static struct page *kimage_alloc_page(st
- * have a match.
- */
- list_for_each_entry(page, &image->dest_pages, lru) {
-- addr = page_to_pfn(page) << PAGE_SHIFT;
-+ addr = kexec_page_to_pfn(page) << PAGE_SHIFT;
- if (addr == destination) {
- list_del(&page->lru);
- return page;
-@@ -697,16 +725,16 @@ static struct page *kimage_alloc_page(st
- kimage_entry_t *old;
-
- /* Allocate a page, if we run out of memory give up */
-- page = kimage_alloc_pages(gfp_mask, 0);
-+ page = kimage_alloc_pages(gfp_mask, 0, KEXEC_SOURCE_MEMORY_LIMIT);
- if (!page)
- return NULL;
- /* If the page cannot be used file it away */
-- if (page_to_pfn(page) >
-+ if (kexec_page_to_pfn(page) >
- (KEXEC_SOURCE_MEMORY_LIMIT >> PAGE_SHIFT)) {
- list_add(&page->lru, &image->unuseable_pages);
- continue;
- }
-- addr = page_to_pfn(page) << PAGE_SHIFT;
-+ addr = kexec_page_to_pfn(page) << PAGE_SHIFT;
-
- /* If it is the destination page we want use it */
- if (addr == destination)
-@@ -729,7 +757,7 @@ static struct page *kimage_alloc_page(st
- struct page *old_page;
-
- old_addr = *old & PAGE_MASK;
-- old_page = pfn_to_page(old_addr >> PAGE_SHIFT);
-+ old_page = kexec_pfn_to_page(old_addr >> PAGE_SHIFT);
- copy_highpage(page, old_page);
- *old = addr | (*old & ~PAGE_MASK);
-
-@@ -779,7 +807,7 @@ static int kimage_load_normal_segment(st
- result = -ENOMEM;
- goto out;
- }
-- result = kimage_add_page(image, page_to_pfn(page)
-+ result = kimage_add_page(image, kexec_page_to_pfn(page)
- << PAGE_SHIFT);
- if (result < 0)
- goto out;
-@@ -811,6 +839,7 @@ out:
- return result;
- }
-
-+#ifndef CONFIG_XEN
- static int kimage_load_crash_segment(struct kimage *image,
- struct kexec_segment *segment)
- {
-@@ -833,7 +862,7 @@ static int kimage_load_crash_segment(str
- char *ptr;
- size_t uchunk, mchunk;
-
-- page = pfn_to_page(maddr >> PAGE_SHIFT);
-+ page = kexec_pfn_to_page(maddr >> PAGE_SHIFT);
- if (page == 0) {
- result = -ENOMEM;
- goto out;
-@@ -881,6 +910,13 @@ static int kimage_load_segment(struct ki
-
- return result;
- }
-+#else /* CONFIG_XEN */
-+static int kimage_load_segment(struct kimage *image,
-+ struct kexec_segment *segment)
++int
++swiotlb_dma_mapping_error(dma_addr_t dma_addr)
+{
-+ return kimage_load_normal_segment(image, segment);
++ return (dma_addr == virt_to_bus(io_tlb_overflow_buffer));
+}
-+#endif
-
- /*
- * Exec Kernel system call: for obvious reasons only root may call it.
-@@ -991,6 +1027,13 @@ asmlinkage long sys_kexec_load(unsigned
- if (result)
- goto out;
- }
-+#ifdef CONFIG_XEN
-+ if (image) {
-+ result = xen_machine_kexec_load(image);
-+ if (result)
-+ goto out;
-+ }
-+#endif
- /* Install the new kernel, and Uninstall the old */
- image = xchg(dest_image, image);
-
-@@ -1044,7 +1087,6 @@ void crash_kexec(struct pt_regs *regs)
- {
- int locked;
-
--
- /* Take the kexec_lock here to prevent sys_kexec_load
- * running on one cpu from replacing the crash kernel
- * we are using after a panic on a different cpu.
-diff -Nurp pristine-linux-2.6.18.2/kernel/softlockup.c tmp-linux-2.6-xen.patch/kernel/softlockup.c
---- pristine-linux-2.6.18.2/kernel/softlockup.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/kernel/softlockup.c 2007-10-14 01:51:15.000000000 +0200
-@@ -40,6 +40,19 @@ void touch_softlockup_watchdog(void)
- }
- EXPORT_SYMBOL(touch_softlockup_watchdog);
-
-+unsigned long softlockup_get_next_event(void)
-+{
-+ int this_cpu = smp_processor_id();
-+ unsigned long touch_timestamp = per_cpu(touch_timestamp, this_cpu);
-+
-+ if (per_cpu(print_timestamp, this_cpu) == touch_timestamp ||
-+ did_panic ||
-+ !per_cpu(watchdog_task, this_cpu))
-+ return MAX_JIFFY_OFFSET;
+
-+ return max_t(long, 0, touch_timestamp + HZ - jiffies);
++/*
++ * Return whether the given PCI device DMA address mask can be supported
++ * properly. For example, if your device can only drive the low 24-bits
++ * during PCI bus mastering, then you would pass 0x00ffffff as the mask to
++ * this function.
++ */
++int
++swiotlb_dma_supported (struct device *hwdev, u64 mask)
++{
++ return (mask >= ((1UL << dma_bits) - 1));
+}
+
- /*
- * This callback runs from the timer interrupt, and checks
- * whether the watchdog thread has hung or not:
-diff -Nurp pristine-linux-2.6.18.2/kernel/timer.c tmp-linux-2.6-xen.patch/kernel/timer.c
---- pristine-linux-2.6.18.2/kernel/timer.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/kernel/timer.c 2007-10-14 01:51:15.000000000 +0200
-@@ -485,7 +485,9 @@ unsigned long next_timer_interrupt(void)
- if (hr_expires < 3)
- return hr_expires + jiffies;
- }
-- hr_expires += jiffies;
-+ hr_expires = min_t(unsigned long,
-+ softlockup_get_next_event(),
-+ hr_expires) + jiffies;
-
- base = __get_cpu_var(tvec_bases);
- spin_lock(&base->lock);
-diff -Nurp pristine-linux-2.6.18.2/lib/Makefile tmp-linux-2.6-xen.patch/lib/Makefile
---- pristine-linux-2.6.18.2/lib/Makefile 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/lib/Makefile 2007-07-30 16:35:13.000000000 +0200
-@@ -52,6 +52,7 @@ obj-$(CONFIG_SMP) += percpu_counter.o
- obj-$(CONFIG_AUDIT_GENERIC) += audit.o
-
- obj-$(CONFIG_SWIOTLB) += swiotlb.o
-+swiotlb-$(CONFIG_XEN) := ../arch/i386/kernel/swiotlb.o
++EXPORT_SYMBOL(swiotlb_init);
++EXPORT_SYMBOL(swiotlb_map_single);
++EXPORT_SYMBOL(swiotlb_unmap_single);
++EXPORT_SYMBOL(swiotlb_map_sg);
++EXPORT_SYMBOL(swiotlb_unmap_sg);
++EXPORT_SYMBOL(swiotlb_sync_single_for_cpu);
++EXPORT_SYMBOL(swiotlb_sync_single_for_device);
++EXPORT_SYMBOL(swiotlb_sync_sg_for_cpu);
++EXPORT_SYMBOL(swiotlb_sync_sg_for_device);
++EXPORT_SYMBOL(swiotlb_dma_mapping_error);
++EXPORT_SYMBOL(swiotlb_dma_supported);
+diff -r d894e36cfc30 -r 0aa021803deb mm/highmem.c
+--- a/mm/highmem.c Tue Sep 09 11:37:38 2008 +0200
++++ b/mm/highmem.c Wed Sep 10 10:54:08 2008 +0100
+@@ -141,6 +141,17 @@
- hostprogs-y := gen_crc32table
- clean-files := crc32table.h
-diff -Nurp pristine-linux-2.6.18.2/mm/highmem.c tmp-linux-2.6-xen.patch/mm/highmem.c
---- pristine-linux-2.6.18.2/mm/highmem.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/mm/highmem.c 2007-07-30 16:35:13.000000000 +0200
-@@ -142,6 +142,17 @@ start:
return vaddr;
}
-
++
+#ifdef CONFIG_XEN
+void kmap_flush_unused(void)
+{
@@ -106330,32 +146163,39 @@
+
+EXPORT_SYMBOL(kmap_flush_unused);
+#endif
-+
+
void fastcall *kmap_high(struct page *page)
{
- unsigned long vaddr;
-diff -Nurp pristine-linux-2.6.18.2/mm/Kconfig tmp-linux-2.6-xen.patch/mm/Kconfig
---- pristine-linux-2.6.18.2/mm/Kconfig 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/mm/Kconfig 2007-07-30 16:35:13.000000000 +0200
-@@ -127,11 +127,14 @@ comment "Memory hotplug is currently inc
- # Default to 4 for wider testing, though 8 might be more appropriate.
- # ARM's adjust_pte (unused if VIPT) depends on mm-wide page_table_lock.
- # PA-RISC 7xxx's spinlock_t would enlarge struct page from 32 to 44 bytes.
-+# XEN on x86 architecture uses the mapping field on pagetable pages to store a
-+# pointer to the destructor. This conflicts with pte_lock_deinit().
- #
- config SPLIT_PTLOCK_CPUS
- int
- default "4096" if ARM && !CPU_CACHE_VIPT
- default "4096" if PARISC && !PA20
-+ default "4096" if X86_XEN || X86_64_XEN
- default "4"
+@@ -457,6 +468,12 @@
+ mempool_t *pool;
- #
-diff -Nurp pristine-linux-2.6.18.2/mm/memory.c tmp-linux-2.6-xen.patch/mm/memory.c
---- pristine-linux-2.6.18.2/mm/memory.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/mm/memory.c 2007-07-30 16:35:13.000000000 +0200
-@@ -403,7 +403,8 @@ struct page *vm_normal_page(struct vm_ar
+ /*
++ * Data-less bio, nothing to bounce
++ */
++ if (bio_empty_barrier(*bio_orig))
++ return;
++
++ /*
+ * for non-isa bounce case, just check if the bounce pfn is equal
+ * to or bigger than the highest pfn in the system -- in that case,
+ * don't waste time iterating over bio segments
+diff -r d894e36cfc30 -r 0aa021803deb mm/memory.c
+--- a/mm/memory.c Tue Sep 09 11:37:38 2008 +0200
++++ b/mm/memory.c Wed Sep 10 10:54:08 2008 +0100
+@@ -396,6 +396,12 @@
+ return NULL;
+ }
+
++#if defined(CONFIG_XEN) && defined(CONFIG_X86)
++ /* XEN: Covers user-space grant mappings (even of local pages). */
++ if (unlikely(vma->vm_flags & VM_FOREIGN))
++ return NULL;
++#endif
++
+ /*
+ * Add some anal sanity checks for now. Eventually,
+ * we should just do "return pfn_to_page(pfn)", but
+@@ -403,7 +409,8 @@
* and that the resulting page looks ok.
*/
if (unlikely(!pfn_valid(pfn))) {
@@ -106365,7 +146205,7 @@
return NULL;
}
-@@ -658,8 +659,12 @@ static unsigned long zap_pte_range(struc
+@@ -658,8 +665,12 @@
page->index > details->last_index))
continue;
}
@@ -106380,7 +146220,7 @@
tlb_remove_tlb_entry(tlb, pte, addr);
if (unlikely(!page))
continue;
-@@ -754,6 +759,7 @@ static unsigned long unmap_page_range(st
+@@ -754,6 +765,7 @@
details = NULL;
BUG_ON(addr >= end);
@@ -106388,7 +146228,7 @@
tlb_start_vma(tlb, vma);
pgd = pgd_offset(vma->vm_mm, addr);
do {
-@@ -891,6 +897,7 @@ unsigned long zap_page_range(struct vm_a
+@@ -891,6 +903,7 @@
tlb_finish_mmu(tlb, address, end);
return end;
}
@@ -106396,7 +146236,7 @@
/*
* Do a quick page-table lookup for a single page.
-@@ -1030,6 +1037,26 @@ int get_user_pages(struct task_struct *t
+@@ -1030,6 +1043,26 @@
continue;
}
@@ -106423,10 +146263,11 @@
if (!vma || (vma->vm_flags & (VM_IO | VM_PFNMAP))
|| !(vm_flags & vma->vm_flags))
return i ? : -EFAULT;
-@@ -1369,6 +1396,102 @@ int remap_pfn_range(struct vm_area_struc
+@@ -1378,6 +1411,102 @@
+ return err;
}
EXPORT_SYMBOL(remap_pfn_range);
-
++
+#ifdef CONFIG_XEN
+static inline int apply_to_pte_range(struct mm_struct *mm, pmd_t *pmd,
+ unsigned long addr, unsigned long end,
@@ -106522,30 +146363,13 @@
+}
+EXPORT_SYMBOL_GPL(apply_to_page_range);
+#endif
-+
+
/*
* handle_pte_fault chooses page fault handler according to an entry
- * which was read non-atomically. Before making any commitment, on
-@@ -1551,14 +1674,7 @@ gotten:
- entry = mk_pte(new_page, vma->vm_page_prot);
- entry = maybe_mkwrite(pte_mkdirty(entry), vma);
- lazy_mmu_prot_update(entry);
-- /*
-- * Clear the pte entry and flush it first, before updating the
-- * pte with the new entry. This will avoid a race condition
-- * seen in the presence of one thread doing SMC and another
-- * thread doing COW.
-- */
-- ptep_clear_flush(vma, address, page_table);
-- set_pte_at(mm, address, page_table, entry);
-+ ptep_establish(vma, address, page_table, entry);
- update_mmu_cache(vma, address, entry);
- lru_cache_add_active(new_page);
- page_add_new_anon_rmap(new_page, vma, address);
-diff -Nurp pristine-linux-2.6.18.2/mm/mmap.c tmp-linux-2.6-xen.patch/mm/mmap.c
---- pristine-linux-2.6.18.2/mm/mmap.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/mm/mmap.c 2007-07-30 16:35:13.000000000 +0200
-@@ -1963,6 +1963,10 @@ void exit_mmap(struct mm_struct *mm)
+diff -r d894e36cfc30 -r 0aa021803deb mm/mmap.c
+--- a/mm/mmap.c Tue Sep 09 11:37:38 2008 +0200
++++ b/mm/mmap.c Wed Sep 10 10:54:08 2008 +0100
+@@ -1970,6 +1970,10 @@
unsigned long nr_accounted = 0;
unsigned long end;
@@ -106556,10 +146380,22 @@
lru_add_drain();
flush_cache_mm(mm);
tlb = tlb_gather_mmu(mm, 1);
-diff -Nurp pristine-linux-2.6.18.2/mm/page_alloc.c tmp-linux-2.6-xen.patch/mm/page_alloc.c
---- pristine-linux-2.6.18.2/mm/page_alloc.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/mm/page_alloc.c 2007-07-30 21:01:00.000000000 +0200
-@@ -154,7 +154,11 @@ static void bad_page(struct page *page)
+diff -r d894e36cfc30 -r 0aa021803deb mm/mprotect.c
+--- a/mm/mprotect.c Tue Sep 09 11:37:38 2008 +0200
++++ b/mm/mprotect.c Wed Sep 10 10:54:08 2008 +0100
+@@ -75,6 +75,8 @@
+ do {
+ next = pmd_addr_end(addr, end);
+ if (pmd_none_or_clear_bad(pmd))
++ continue;
++ if (arch_change_pte_range(mm, pmd, addr, next, newprot))
+ continue;
+ change_pte_range(mm, pmd, addr, next, newprot);
+ } while (pmd++, addr = next, addr != end);
+diff -r d894e36cfc30 -r 0aa021803deb mm/page_alloc.c
+--- a/mm/page_alloc.c Tue Sep 09 11:37:38 2008 +0200
++++ b/mm/page_alloc.c Wed Sep 10 10:54:08 2008 +0100
+@@ -154,7 +154,11 @@
1 << PG_slab |
1 << PG_swapcache |
1 << PG_writeback |
@@ -106572,7 +146408,7 @@
set_page_count(page, 0);
reset_page_mapcount(page);
page->mapping = NULL;
-@@ -389,7 +393,11 @@ static inline int free_pages_check(struc
+@@ -389,7 +393,11 @@
1 << PG_swapcache |
1 << PG_writeback |
1 << PG_reserved |
@@ -106585,17 +146421,20 @@
bad_page(page);
if (PageDirty(page))
__ClearPageDirty(page);
-@@ -443,7 +451,8 @@ static void __free_pages_ok(struct page
+@@ -443,6 +451,12 @@
int i;
int reserved = 0;
-- arch_free_page(page, order);
-+ if (arch_free_page(page, order))
++#ifdef CONFIG_XEN
++ if (PageForeign(page)) {
++ PageForeignDestructor(page);
+ return;
++ }
++#endif
+ arch_free_page(page, order);
if (!PageHighMem(page))
debug_check_no_locks_freed(page_address(page),
- PAGE_SIZE<<order);
-@@ -538,7 +547,11 @@ static int prep_new_page(struct page *pa
+@@ -538,7 +552,11 @@
1 << PG_swapcache |
1 << PG_writeback |
1 << PG_reserved |
@@ -106608,63 +146447,39 @@
bad_page(page);
/*
-@@ -717,7 +730,8 @@ static void fastcall free_hot_cold_page(
+@@ -717,6 +735,12 @@
struct per_cpu_pages *pcp;
unsigned long flags;
-- arch_free_page(page, 0);
-+ if (arch_free_page(page, 0))
++#ifdef CONFIG_XEN
++ if (PageForeign(page)) {
++ PageForeignDestructor(page);
+ return;
++ }
++#endif
+ arch_free_page(page, 0);
if (PageAnon(page))
- page->mapping = NULL;
-@@ -1847,10 +1861,8 @@ static inline void free_zone_pagesets(in
- for_each_zone(zone) {
- struct per_cpu_pageset *pset = zone_pcp(zone, cpu);
-
-- /* Free per_cpu_pageset if it is slab allocated */
-- if (pset != &boot_pageset[cpu])
-- kfree(pset);
- zone_pcp(zone, cpu) = NULL;
-+ kfree(pset);
- }
- }
+diff -r d894e36cfc30 -r 0aa021803deb net/bridge/br_netfilter.c
+--- a/net/bridge/br_netfilter.c Tue Sep 09 11:37:38 2008 +0200
++++ b/net/bridge/br_netfilter.c Wed Sep 10 10:54:08 2008 +0100
+@@ -127,10 +127,10 @@
-@@ -2012,7 +2024,6 @@ static void __meminit free_area_init_cor
- #ifdef CONFIG_NUMA
- zone->min_unmapped_ratio = (realsize*sysctl_min_unmapped_ratio)
- / 100;
-- zone->min_slab_pages = (realsize * sysctl_min_slab_ratio) / 100;
- #endif
- zone->name = zone_names[j];
- spin_lock_init(&zone->lock);
-@@ -2323,22 +2334,6 @@ int sysctl_min_unmapped_ratio_sysctl_han
- sysctl_min_unmapped_ratio) / 100;
- return 0;
- }
--
--int sysctl_min_slab_ratio_sysctl_handler(ctl_table *table, int write,
-- struct file *file, void __user *buffer, size_t *length, loff_t *ppos)
--{
-- struct zone *zone;
-- int rc;
--
-- rc = proc_dointvec_minmax(table, write, file, buffer, length, ppos);
-- if (rc)
-- return rc;
--
-- for_each_zone(zone)
-- zone->min_slab_pages = (zone->present_pages *
-- sysctl_min_slab_ratio) / 100;
-- return 0;
--}
- #endif
+ static inline void nf_bridge_save_header(struct sk_buff *skb)
+ {
+- int header_size = 16;
++ int header_size = ETH_HLEN;
- /*
-diff -Nurp pristine-linux-2.6.18.2/net/core/dev.c tmp-linux-2.6-xen.patch/net/core/dev.c
---- pristine-linux-2.6.18.2/net/core/dev.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/net/core/dev.c 2007-07-30 16:35:13.000000000 +0200
-@@ -113,11 +113,18 @@
+ if (skb->protocol == htons(ETH_P_8021Q))
+- header_size = 18;
++ header_size += VLAN_HLEN;
+
+ memcpy(skb->nf_bridge->data, skb->data - header_size, header_size);
+ }
+diff -r d894e36cfc30 -r 0aa021803deb net/core/dev.c
+--- a/net/core/dev.c Tue Sep 09 11:37:38 2008 +0200
++++ b/net/core/dev.c Wed Sep 10 10:54:08 2008 +0100
+@@ -113,10 +113,17 @@
#include <linux/wireless.h>
#include <net/iw_handler.h>
#include <asm/current.h>
@@ -106673,17 +146488,16 @@
#include <linux/dmaengine.h>
#include <linux/err.h>
#include <linux/ctype.h>
-
++
+#ifdef CONFIG_XEN
+#include <net/ip.h>
+#include <linux/tcp.h>
+#include <linux/udp.h>
+#endif
-+
+
/*
* The list of packet types we will receive (as opposed to discard)
- * and the routines to invoke.
-@@ -1328,7 +1335,7 @@ static int dev_gso_segment(struct sk_buf
+@@ -1328,7 +1335,7 @@
/* Verifying header integrity only. */
if (!segs)
return 0;
@@ -106692,7 +146506,7 @@
if (unlikely(IS_ERR(segs)))
return PTR_ERR(segs);
-@@ -1391,6 +1398,43 @@ out_kfree_skb:
+@@ -1391,6 +1398,43 @@
} \
}
@@ -106736,46 +146550,26 @@
/**
* dev_queue_xmit - transmit a buffer
* @skb: buffer to transmit
-@@ -1423,6 +1467,12 @@ int dev_queue_xmit(struct sk_buff *skb)
+@@ -1422,6 +1466,12 @@
+ struct net_device *dev = skb->dev;
struct Qdisc *q;
int rc = -ENOMEM;
-
++
+ /* If a checksum-deferred packet is forwarded to a device that needs a
+ * checksum, correct the pointers and force checksumming.
+ */
+ if (skb_checksum_setup(skb))
+ goto out_kfree_skb;
-+
+
/* GSO will handle the following emulations directly. */
if (netif_needs_gso(dev, skb))
- goto gso;
-@@ -1478,16 +1528,14 @@ gso:
- if (q->enqueue) {
- /* Grab device queue */
- spin_lock(&dev->queue_lock);
-- q = dev->qdisc;
-- if (q->enqueue) {
-- rc = q->enqueue(skb, q);
-- qdisc_run(dev);
-- spin_unlock(&dev->queue_lock);
-
-- rc = rc == NET_XMIT_BYPASS ? NET_XMIT_SUCCESS : rc;
-- goto out;
-- }
-+ rc = q->enqueue(skb, q);
-+
-+ qdisc_run(dev);
+@@ -1795,6 +1845,19 @@
+ if (skb->tc_verd & TC_NCLS) {
+ skb->tc_verd = CLR_TC_NCLS(skb->tc_verd);
+ goto ncls;
++ }
++#endif
+
- spin_unlock(&dev->queue_lock);
-+ rc = rc == NET_XMIT_BYPASS ? NET_XMIT_SUCCESS : rc;
-+ goto out;
- }
-
- /* The device has no queue. Common case for software devices:
-@@ -1798,6 +1846,19 @@ int netif_receive_skb(struct sk_buff *sk
- }
- #endif
-
+#ifdef CONFIG_XEN
+ switch (skb->ip_summed) {
+ case CHECKSUM_UNNECESSARY:
@@ -106786,13 +146580,10 @@
+ default:
+ skb->proto_data_valid = 0;
+ break;
-+ }
-+#endif
-+
- list_for_each_entry_rcu(ptype, &ptype_all, list) {
- if (!ptype->dev || ptype->dev == skb->dev) {
- if (pt_prev)
-@@ -3584,6 +3645,7 @@ EXPORT_SYMBOL(unregister_netdevice_notif
+ }
+ #endif
+
+@@ -3584,6 +3647,7 @@
EXPORT_SYMBOL(net_enable_timestamp);
EXPORT_SYMBOL(net_disable_timestamp);
EXPORT_SYMBOL(dev_get_flags);
@@ -106800,10 +146591,21 @@
#if defined(CONFIG_BRIDGE) || defined(CONFIG_BRIDGE_MODULE)
EXPORT_SYMBOL(br_handle_frame_hook);
-diff -Nurp pristine-linux-2.6.18.2/net/core/skbuff.c tmp-linux-2.6-xen.patch/net/core/skbuff.c
---- pristine-linux-2.6.18.2/net/core/skbuff.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/net/core/skbuff.c 2007-07-30 16:35:13.000000000 +0200
-@@ -240,6 +240,7 @@ struct sk_buff *alloc_skb_from_cache(kme
+diff -r d894e36cfc30 -r 0aa021803deb net/core/neighbour.c
+--- a/net/core/neighbour.c Tue Sep 09 11:37:38 2008 +0200
++++ b/net/core/neighbour.c Wed Sep 10 10:54:08 2008 +0100
+@@ -2679,7 +2679,6 @@
+ EXPORT_SYMBOL(neigh_table_init);
+ EXPORT_SYMBOL(neigh_table_init_no_netlink);
+ EXPORT_SYMBOL(neigh_update);
+-EXPORT_SYMBOL(neigh_update_hhs);
+ EXPORT_SYMBOL(pneigh_enqueue);
+ EXPORT_SYMBOL(pneigh_lookup);
+ EXPORT_SYMBOL(neightbl_dump_info);
+diff -r d894e36cfc30 -r 0aa021803deb net/core/skbuff.c
+--- a/net/core/skbuff.c Tue Sep 09 11:37:38 2008 +0200
++++ b/net/core/skbuff.c Wed Sep 10 10:54:08 2008 +0100
+@@ -240,6 +240,7 @@
skb_shinfo(skb)->gso_size = 0;
skb_shinfo(skb)->gso_segs = 0;
skb_shinfo(skb)->gso_type = 0;
@@ -106811,7 +146613,7 @@
skb_shinfo(skb)->frag_list = NULL;
out:
return skb;
-@@ -464,6 +465,10 @@ struct sk_buff *skb_clone(struct sk_buff
+@@ -464,6 +465,10 @@
C(local_df);
n->cloned = 1;
n->nohdr = 0;
@@ -106822,10 +146624,10 @@
C(pkt_type);
C(ip_summed);
C(priority);
-diff -Nurp pristine-linux-2.6.18.2/net/ipv4/netfilter/ip_nat_proto_tcp.c tmp-linux-2.6-xen.patch/net/ipv4/netfilter/ip_nat_proto_tcp.c
---- pristine-linux-2.6.18.2/net/ipv4/netfilter/ip_nat_proto_tcp.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/net/ipv4/netfilter/ip_nat_proto_tcp.c 2007-10-14 01:51:15.000000000 +0200
-@@ -129,7 +129,12 @@ tcp_manip_pkt(struct sk_buff **pskb,
+diff -r d894e36cfc30 -r 0aa021803deb net/ipv4/netfilter/ip_nat_proto_tcp.c
+--- a/net/ipv4/netfilter/ip_nat_proto_tcp.c Tue Sep 09 11:37:38 2008 +0200
++++ b/net/ipv4/netfilter/ip_nat_proto_tcp.c Wed Sep 10 10:54:08 2008 +0100
+@@ -129,7 +129,12 @@
if (hdrsize < sizeof(*hdr))
return 1;
@@ -106839,10 +146641,10 @@
ip_nat_cheat_check(oldport ^ 0xFFFF,
newport,
hdr->check));
-diff -Nurp pristine-linux-2.6.18.2/net/ipv4/netfilter/ip_nat_proto_udp.c tmp-linux-2.6-xen.patch/net/ipv4/netfilter/ip_nat_proto_udp.c
---- pristine-linux-2.6.18.2/net/ipv4/netfilter/ip_nat_proto_udp.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/net/ipv4/netfilter/ip_nat_proto_udp.c 2007-10-14 01:51:15.000000000 +0200
-@@ -113,11 +113,17 @@ udp_manip_pkt(struct sk_buff **pskb,
+diff -r d894e36cfc30 -r 0aa021803deb net/ipv4/netfilter/ip_nat_proto_udp.c
+--- a/net/ipv4/netfilter/ip_nat_proto_udp.c Tue Sep 09 11:37:38 2008 +0200
++++ b/net/ipv4/netfilter/ip_nat_proto_udp.c Wed Sep 10 10:54:08 2008 +0100
+@@ -113,11 +113,17 @@
newport = tuple->dst.u.udp.port;
portptr = &hdr->dest;
}
@@ -106862,10 +146664,10 @@
*portptr = newport;
return 1;
}
-diff -Nurp pristine-linux-2.6.18.2/net/ipv4/tcp_input.c tmp-linux-2.6-xen.patch/net/ipv4/tcp_input.c
---- pristine-linux-2.6.18.2/net/ipv4/tcp_input.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/net/ipv4/tcp_input.c 2007-10-14 01:51:15.000000000 +0200
-@@ -127,7 +127,7 @@ static void tcp_measure_rcv_mss(struct s
+diff -r d894e36cfc30 -r 0aa021803deb net/ipv4/tcp_input.c
+--- a/net/ipv4/tcp_input.c Tue Sep 09 11:37:38 2008 +0200
++++ b/net/ipv4/tcp_input.c Wed Sep 10 10:54:08 2008 +0100
+@@ -127,7 +127,7 @@
/* skb->len may jitter because of SACKs, even if peer
* sends good full-sized frames.
*/
@@ -106874,19 +146676,19 @@
if (len >= icsk->icsk_ack.rcv_mss) {
icsk->icsk_ack.rcv_mss = len;
} else {
-diff -Nurp pristine-linux-2.6.18.2/net/ipv4/xfrm4_output.c tmp-linux-2.6-xen.patch/net/ipv4/xfrm4_output.c
---- pristine-linux-2.6.18.2/net/ipv4/xfrm4_output.c 2006-11-04 02:33:58.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/net/ipv4/xfrm4_output.c 2007-10-14 01:51:15.000000000 +0200
-@@ -18,6 +18,8 @@
+diff -r d894e36cfc30 -r 0aa021803deb net/ipv4/xfrm4_output.c
+--- a/net/ipv4/xfrm4_output.c Tue Sep 09 11:37:38 2008 +0200
++++ b/net/ipv4/xfrm4_output.c Wed Sep 10 10:54:08 2008 +0100
+@@ -17,6 +17,8 @@
+ #include <net/ip.h>
#include <net/xfrm.h>
#include <net/icmp.h>
-
-+extern int skb_checksum_setup(struct sk_buff *skb);
+
++extern int skb_checksum_setup(struct sk_buff *skb);
+
static int xfrm4_tunnel_check_size(struct sk_buff *skb)
{
- int mtu, ret = 0;
-@@ -48,6 +50,10 @@ static int xfrm4_output_one(struct sk_bu
+@@ -48,6 +50,10 @@
struct xfrm_state *x = dst->xfrm;
int err;
@@ -106897,21 +146699,79 @@
if (skb->ip_summed == CHECKSUM_HW) {
err = skb_checksum_help(skb, 0);
if (err)
-diff -Nurp pristine-linux-2.6.18.2/scripts/Makefile.xen tmp-linux-2.6-xen.patch/scripts/Makefile.xen
---- pristine-linux-2.6.18.2/scripts/Makefile.xen 1970-01-01 01:00:00.000000000 +0100
-+++ tmp-linux-2.6-xen.patch/scripts/Makefile.xen 2007-07-30 16:35:13.000000000 +0200
-@@ -0,0 +1,14 @@
+diff -r d894e36cfc30 -r 0aa021803deb scripts/Makefile.build
+--- a/scripts/Makefile.build Tue Sep 09 11:37:38 2008 +0200
++++ b/scripts/Makefile.build Wed Sep 10 10:54:08 2008 +0100
+@@ -66,6 +66,18 @@
+
+ ifndef obj
+ $(warning kbuild: Makefile.build is included improperly)
++endif
++
++ifeq ($(CONFIG_XEN),y)
++$(objtree)/scripts/Makefile.xen: $(srctree)/scripts/Makefile.xen.awk $(srctree)/scripts/Makefile.build
++ @echo ' Updating $@'
++ @$(AWK) -f $< $(filter-out $<,$^) >$@
++
++xen-src-single-used-m := $(patsubst $(srctree)/%,%,$(wildcard $(addprefix $(srctree)/,$(single-used-m:.o=-xen.c))))
++xen-single-used-m := $(xen-src-single-used-m:-xen.c=.o)
++single-used-m := $(filter-out $(xen-single-used-m),$(single-used-m))
++
++-include $(objtree)/scripts/Makefile.xen
+ endif
+
+ # ===========================================================================
+diff -r d894e36cfc30 -r 0aa021803deb scripts/Makefile.lib
+--- a/scripts/Makefile.lib Tue Sep 09 11:37:38 2008 +0200
++++ b/scripts/Makefile.lib Wed Sep 10 10:54:08 2008 +0100
+@@ -12,6 +12,12 @@
+ # Filter out objects already built-in
+
+ lib-y := $(filter-out $(obj-y), $(sort $(lib-y) $(lib-m)))
++
++# Remove objects forcibly disabled
++
++obj-y := $(filter-out $(disabled-obj-y),$(obj-y))
++obj-m := $(filter-out $(disabled-obj-y),$(obj-m))
++lib-y := $(filter-out $(disabled-obj-y),$(lib-y))
+
+
+ # Handle objects in subdirs
+diff -r d894e36cfc30 -r 0aa021803deb scripts/Makefile.xen.awk
+--- /dev/null Thu Jan 01 00:00:00 1970 +0000
++++ b/scripts/Makefile.xen.awk Wed Sep 10 10:54:08 2008 +0100
+@@ -0,0 +1,34 @@
++BEGIN {
++ is_rule = 0
++}
++
++/^[[:space:]]*#/ {
++ next
++}
++
++/^[[:space:]]*$/ {
++ if (is_rule)
++ print("")
++ is_rule = 0
++ next
++}
++
++/:[[:space:]]*%\.[cS][[:space:]]/ {
++ line = gensub(/%.([cS])/, "%-xen.\\1", "g", $0)
++ line = gensub(/(single-used-m)/, "xen-\\1", "g", line)
++ print line
++ is_rule = 1
++ next
++}
++
++/^[^\t]$/ {
++ if (is_rule)
++ print("")
++ is_rule = 0
++ next
++}
+
-+# cherrypickxen($1 = allobj)
-+cherrypickxen = $(foreach var, $(1), \
-+ $(shell o=$(var); \
-+ c=$${o%.o}-xen.c; \
-+ s=$${o%.o}-xen.S; \
-+ oxen=$${o%.o}-xen.o; \
-+ [ -f $(srctree)/$(src)/$${c} ] || \
-+ [ -f $(srctree)/$(src)/$${s} ] \
-+ && echo $$oxen \
-+ || echo $(var) ) \
-+ )
-+# filterxen($1 = allobj, $2 = noobjs)
-+filterxen = $(filter-out $(2), $(1))
++is_rule {
++ print $0
++ next
++}
Modified: people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/series/1-extra
==============================================================================
--- people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/series/1-extra (original)
+++ people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/patches-update/series/1-extra Sat Oct 18 14:44:52 2008
@@ -4,7 +4,7 @@
+ ../patches-base/features/all/vserver/vs2.0.2.2-rc9.patch featureset=xen-vserver
+ ../patches-base/features/all/vserver/bindmount-dev.patch featureset=xen-vserver
+ features/all/xen/vserver-clash.patch featureset=xen-vserver
-+ features/all/xen/xen-3.1-15467.patch
++ features/all/xen/xen-748f324a4b2d62d89fe40c4aa52861977e1a2cae.patch
+ features/all/xen/remove-4gb-warning.patch arch=i386
+ features/all/xen/console-hvc-overtake.patch
+ features/all/xen/vserver-update.patch featureset=xen-vserver
Modified: people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/rules.real
==============================================================================
--- people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/rules.real (original)
+++ people/waldi/xen-extra/linux-2.6.18-xen-3.3/debian/rules.real Sat Oct 18 14:44:52 2008
@@ -363,7 +363,7 @@
dh_clean -d -k
dh_installdirs -p$(PACKAGE_NAME) boot 'var/lib/$(PACKAGE_NAME)'
dh_installdirs -p$(MODULES_PACKAGE_NAME) boot
- cp '$(DIR)'/vmlinuz $(INSTALL_DIR)/vmlinuz-$(REAL_VERSION)
+ cp '$(DIR)'/arch/$(KERNEL_ARCH)/boot/vmlinuz $(INSTALL_DIR)/vmlinuz-$(REAL_VERSION)
$(MAKE_CLEAN) -C $(DIR) modules_install INSTALL_MOD_PATH='$(CURDIR)'/$(MODULES_PACKAGE_DIR)
cp $(DIR)/.config $(MODULES_PACKAGE_DIR)/boot/config-$(REAL_VERSION)
cp $(DIR)/System.map $(MODULES_PACKAGE_DIR)/boot/System.map-$(REAL_VERSION)
More information about the Kernel-svn-changes
mailing list