[kernel] r8364 - in dists/trunk/linux-2.6/debian: . patches/features/all/xen patches/series
Bastian Blank
waldi at alioth.debian.org
Fri Mar 16 19:32:49 UTC 2007
Author: waldi
Date: Fri Mar 16 19:32:46 2007
New Revision: 8364
Added:
dists/trunk/linux-2.6/debian/patches/features/all/xen/fedora-2.6.20-48670.patch
- copied, changed from r8229, dists/trunk/linux-2.6/debian/patches/features/all/xen/fedora-2.6.18-36186.patch
dists/trunk/linux-2.6/debian/patches/features/all/xen/update.patch (contents, props changed)
Removed:
dists/trunk/linux-2.6/debian/patches/features/all/xen/fedora-2.6.18-36186.patch
Modified:
dists/trunk/linux-2.6/debian/changelog
dists/trunk/linux-2.6/debian/patches/series/1~experimental.1-extra
Log:
* Update xen patch to changeset 48670 from fedora 2.6.20 branch.
* Fix identified problems with this patch.
* debian/changelog: Update.
* debian/patches/features/all/xen/fedora-2.6.18-36186.patch: Remove.
* debian/patches/features/all/xen/fedora-2.6.20-48670.patch,
debian/patches/features/all/xen/update.patch: Add.
* debian/patches/series/1~experimental.1-extra: Update.
Modified: dists/trunk/linux-2.6/debian/changelog
==============================================================================
--- dists/trunk/linux-2.6/debian/changelog (original)
+++ dists/trunk/linux-2.6/debian/changelog Fri Mar 16 19:32:46 2007
@@ -125,6 +125,7 @@
- rtc-pcf8563: detect polarity of century bit automatically
- x86_64: fix 2.6.18 regression - PTRACE_OLDSETOPTIONS should be accepted
- ocfs2: ocfs2_link() journal credits update
+ * Update xen patch to changeset 48670 from fedora 2.6.20 branch.
[ Rod Whitby ]
* arm/ixp4xx: Enable PATA_ARTOP for the nas100d and dsmg600.
@@ -163,7 +164,7 @@
[ Frederik Schüler ]
* Disable NAPI on forcedeth, it is broken.
- -- Bastian Blank <waldi at debian.org> Thu, 15 Mar 2007 14:28:09 +0100
+ -- Bastian Blank <waldi at debian.org> Fri, 16 Mar 2007 19:24:10 +0100
linux-2.6 (2.6.18.dfsg.1-10) unstable; urgency=low
Copied: dists/trunk/linux-2.6/debian/patches/features/all/xen/fedora-2.6.20-48670.patch (from r8229, dists/trunk/linux-2.6/debian/patches/features/all/xen/fedora-2.6.18-36186.patch)
==============================================================================
--- dists/trunk/linux-2.6/debian/patches/features/all/xen/fedora-2.6.18-36186.patch (original)
+++ dists/trunk/linux-2.6/debian/patches/features/all/xen/fedora-2.6.20-48670.patch Fri Mar 16 19:32:46 2007
@@ -1,6 +1,6 @@
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/boot-xen/Makefile linux-2.6.18-xen/arch/i386/boot-xen/Makefile
---- linux-2.6.18.3/arch/i386/boot-xen/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/boot-xen/Makefile 2006-11-19 14:26:21.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/boot-xen/Makefile b/arch/i386/boot-xen/Makefile
+--- a/arch/i386/boot-xen/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/boot-xen/Makefile 2007-03-14 10:55:14.000000000 +0100
@@ -0,0 +1,21 @@
+
+OBJCOPYFLAGS := -g --strip-unneeded
@@ -23,9 +23,9 @@
+ 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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/Kconfig linux-2.6.18-xen/arch/i386/Kconfig
---- linux-2.6.18.3/arch/i386/Kconfig 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/Kconfig 2006-12-05 18:42:36.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/Kconfig b/arch/i386/Kconfig
+--- a/arch/i386/Kconfig 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/Kconfig 2007-03-14 10:55:14.000000000 +0100
@@ -16,6 +16,7 @@
config GENERIC_TIME
@@ -34,7 +34,7 @@
default y
config LOCKDEP_SUPPORT
-@@ -103,6 +104,15 @@
+@@ -108,6 +109,15 @@
help
Choose this option if your computer is a standard PC or compatible.
@@ -50,7 +50,7 @@
config X86_ELAN
bool "AMD Elan"
help
-@@ -213,6 +223,7 @@
+@@ -229,6 +239,7 @@
config HPET_TIMER
bool "HPET Timer Support"
@@ -58,31 +58,31 @@
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 @@
+@@ -279,7 +290,7 @@
config X86_UP_APIC
bool "Local APIC support on uniprocessors"
-- depends on !SMP && !(X86_VISWS || X86_VOYAGER)
-+ depends on !SMP && !(X86_VISWS || X86_VOYAGER || XEN_UNPRIVILEGED_GUEST)
+- depends on !SMP && !(X86_VISWS || X86_VOYAGER || X86_GENERICARCH)
++ depends on !SMP && !(X86_VISWS || X86_VOYAGER || X86_GENERICARCH || XEN_UNPRIVILEGED_GUEST)
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 @@
+@@ -304,12 +315,12 @@
config X86_LOCAL_APIC
bool
-- depends on X86_UP_APIC || ((X86_VISWS || SMP) && !X86_VOYAGER)
-+ depends on X86_UP_APIC || ((X86_VISWS || SMP) && !(X86_VOYAGER || XEN_UNPRIVILEGED_GUEST))
+- depends on X86_UP_APIC || ((X86_VISWS || SMP) && !X86_VOYAGER) || X86_GENERICARCH
++ depends on X86_UP_APIC || ((X86_VISWS || SMP) && !(X86_VOYAGER || XEN_UNPRIVILEGED_GUEST)) || X86_GENERICARCH
default y
config X86_IO_APIC
bool
-- depends on X86_UP_IOAPIC || (SMP && !(X86_VISWS || X86_VOYAGER))
-+ depends on X86_UP_IOAPIC || (SMP && !(X86_VISWS || X86_VOYAGER || XEN_UNPRIVILEGED_GUEST))
+- depends on X86_UP_IOAPIC || (SMP && !(X86_VISWS || X86_VOYAGER)) || X86_GENERICARCH
++ depends on X86_UP_IOAPIC || (SMP && !(X86_VISWS || X86_VOYAGER || XEN_UNPRIVILEGED_GUEST)) || X86_GENERICARCH
default y
config X86_VISWS_APIC
-@@ -303,7 +314,7 @@
+@@ -319,7 +330,7 @@
config X86_MCE
bool "Machine Check Exception"
@@ -91,15 +91,15 @@
---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 @@
+@@ -418,6 +429,7 @@
config MICROCODE
tristate "/dev/cpu/microcode - Intel IA32 CPU microcode support"
+ depends on !XEN_UNPRIVILEGED_GUEST
+ select FW_LOADER
---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 @@
+@@ -441,6 +453,7 @@
config X86_MSR
tristate "/dev/cpu/*/msr - Model-specific register support"
@@ -107,7 +107,7 @@
help
This device gives privileged processes access to the x86
Model-Specific Registers (MSRs). It is a character device with
-@@ -434,6 +447,10 @@
+@@ -456,6 +469,10 @@
with major 203 and minors 0 to 31 for /dev/cpu/0/cpuid to
/dev/cpu/31/cpuid.
@@ -118,7 +118,7 @@
source "drivers/firmware/Kconfig"
choice
-@@ -607,7 +624,7 @@
+@@ -629,7 +646,7 @@
config HIGHPTE
bool "Allocate 3rd-level pagetables from highmem"
@@ -127,7 +127,7 @@
help
The VM uses one page table entry for each page of physical memory.
For systems with a lot of RAM, this can be wasteful of precious
-@@ -616,6 +633,7 @@
+@@ -638,6 +655,7 @@
config MATH_EMULATION
bool "Math emulation"
@@ -135,7 +135,7 @@
---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 @@
+@@ -663,6 +681,8 @@
config MTRR
bool "MTRR (Memory Type Range Register) support"
@@ -144,7 +144,7 @@
---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 @@
+@@ -697,7 +717,7 @@
config EFI
bool "Boot from EFI support"
@@ -152,8 +152,8 @@
+ depends on ACPI && !X86_XEN
default n
---help---
- This enables the the kernel to boot on EFI platforms using
-@@ -693,7 +713,7 @@
+ This enables the kernel to boot on EFI platforms using
+@@ -715,7 +735,7 @@
config IRQBALANCE
bool "Enable kernel irq balancing"
@@ -162,24 +162,24 @@
default y
help
The default yes will allow the kernel to do irq load balancing.
-@@ -741,7 +761,7 @@
+@@ -749,6 +769,7 @@
config KEXEC
- bool "kexec system call (EXPERIMENTAL)"
-- depends on EXPERIMENTAL
-+ depends on EXPERIMENTAL && !X86_XEN
+ bool "kexec system call"
++ depends on !X86_XEN
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
-@@ -794,6 +814,7 @@
+@@ -865,7 +886,7 @@
config COMPAT_VDSO
bool "Compat VDSO support"
default y
+- depends on !PARAVIRT
+ depends on !X86_XEN
help
Map the VDSO to the predictable old-style address too.
---help---
-@@ -810,18 +831,20 @@
+@@ -882,18 +903,20 @@
depends on HIGHMEM
menu "Power management options (ACPI, APM)"
@@ -203,7 +203,7 @@
---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 @@
+@@ -1078,6 +1101,7 @@
config PCI_GOBIOS
bool "BIOS"
@@ -211,7 +211,7 @@
config PCI_GOMMCONFIG
bool "MMConfig"
-@@ -1013,6 +1037,13 @@
+@@ -1085,6 +1109,13 @@
config PCI_GODIRECT
bool "Direct"
@@ -225,7 +225,7 @@
config PCI_GOANY
bool "Any"
-@@ -1020,7 +1051,7 @@
+@@ -1092,7 +1123,7 @@
config PCI_BIOS
bool
@@ -234,7 +234,7 @@
default y
config PCI_DIRECT
-@@ -1033,6 +1064,18 @@
+@@ -1105,6 +1136,18 @@
depends on PCI && ACPI && (PCI_GOMMCONFIG || PCI_GOANY)
default y
@@ -253,7 +253,7 @@
source "drivers/pci/pcie/Kconfig"
source "drivers/pci/Kconfig"
-@@ -1043,7 +1086,7 @@
+@@ -1115,7 +1158,7 @@
config ISA
bool "ISA support"
@@ -262,7 +262,7 @@
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 @@
+@@ -1142,7 +1185,7 @@
source "drivers/eisa/Kconfig"
config MCA
@@ -271,7 +271,7 @@
default y if X86_VOYAGER
help
MicroChannel Architecture is found in some IBM PS/2 machines and
-@@ -1146,6 +1189,8 @@
+@@ -1218,6 +1261,8 @@
source "crypto/Kconfig"
@@ -280,7 +280,7 @@
source "lib/Kconfig"
#
-@@ -1171,7 +1216,7 @@
+@@ -1243,7 +1288,7 @@
config X86_HT
bool
@@ -289,7 +289,7 @@
default y
config X86_BIOS_REBOOT
-@@ -1184,6 +1229,16 @@
+@@ -1256,6 +1301,16 @@
depends on X86_SMP || (X86_VOYAGER && SMP)
default y
@@ -306,10 +306,10 @@
config KTIME_SCALAR
bool
default y
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/Kconfig.cpu linux-2.6.18-xen/arch/i386/Kconfig.cpu
---- linux-2.6.18.3/arch/i386/Kconfig.cpu 2006-12-06 09:06:08.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/Kconfig.cpu 2006-12-05 18:42:36.000000000 +0100
-@@ -252,7 +252,7 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/Kconfig.cpu b/arch/i386/Kconfig.cpu
+--- a/arch/i386/Kconfig.cpu 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/Kconfig.cpu 2007-03-14 10:55:14.000000000 +0100
+@@ -267,7 +267,7 @@
config X86_F00F_BUG
bool
@@ -318,16 +318,16 @@
default y
config X86_WP_WORKS_OK
-@@ -312,5 +312,5 @@
+@@ -327,5 +327,5 @@
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
+- depends on (MWINCHIP3D || MWINCHIP2 || MCRUSOE || MEFFICEON || MCYRIXIII || MK7 || MK6 || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || M586MMX || M586TSC || MK8 || MVIAC3_2 || MGEODEGX1 || MGEODE_LX || MCORE2) && !X86_NUMAQ
++ depends on (MWINCHIP3D || MWINCHIP2 || MCRUSOE || MEFFICEON || MCYRIXIII || MK7 || MK6 || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || M586MMX || M586TSC || MK8 || MVIAC3_2 || MGEODEGX1 || MGEODE_LX || MCORE2) && !X86_NUMAQ && !X86_XEN
default y
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/Kconfig.debug linux-2.6.18-xen/arch/i386/Kconfig.debug
---- linux-2.6.18.3/arch/i386/Kconfig.debug 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/Kconfig.debug 2006-12-05 18:42:36.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/Kconfig.debug b/arch/i386/Kconfig.debug
+--- a/arch/i386/Kconfig.debug 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/Kconfig.debug 2007-03-14 10:55:14.000000000 +0100
@@ -79,6 +79,7 @@
config DOUBLEFAULT
default y
@@ -336,3092 +336,2436 @@
help
This option allows trapping of rare doublefault exceptions that
would otherwise cause a system to silently reboot. Disabling this
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/acpi/boot-xen.c linux-2.6.18-xen/arch/i386/kernel/acpi/boot-xen.c
---- linux-2.6.18.3/arch/i386/kernel/acpi/boot-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/acpi/boot-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,1168 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/acpi/boot.c b/arch/i386/kernel/acpi/boot.c
+--- a/arch/i386/kernel/acpi/boot.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/acpi/boot.c 2007-03-14 10:55:14.000000000 +0100
+@@ -107,7 +107,7 @@
+ */
+ enum acpi_irq_model_id acpi_irq_model = ACPI_IRQ_MODEL_PIC;
+
+-#ifdef CONFIG_X86_64
++#if defined(CONFIG_X86_64) && !defined(CONFIG_XEN)
+
+ /* rely on all ACPI tables being in the direct mapping */
+ char *__acpi_map_table(unsigned long phys_addr, unsigned long size)
+@@ -140,8 +140,10 @@
+ unsigned long base, offset, mapped_size;
+ int idx;
+
++#ifndef CONFIG_XEN
+ if (phys + size < 8 * 1024 * 1024)
+ return __va(phys);
++#endif
+
+ offset = phys & (PAGE_SIZE - 1);
+ mapped_size = PAGE_SIZE - offset;
+@@ -611,7 +613,11 @@
+ * RSDP signature.
+ */
+ for (offset = 0; offset < length; offset += 16) {
++#ifdef CONFIG_XEN
++ if (strncmp((char *)((unsigned long)isa_bus_to_virt(start) + offset), "RSD PTR ", sig_len))
++#else
+ if (strncmp((char *)(phys_to_virt(start) + offset), "RSD PTR ", sig_len))
++#endif
+ continue;
+ return (start + offset);
+ }
+@@ -724,7 +730,7 @@
+ acpi_fadt.force_apic_physical_destination_mode =
+ fadt->force_apic_physical_destination_mode;
+
+-#ifdef CONFIG_X86_PM_TIMER
++#if defined(CONFIG_X86_PM_TIMER) && !defined(CONFIG_XEN)
+ /* detect the location of the ACPI PM Timer */
+ if (fadt->revision >= FADT2_REVISION_ID) {
+ /* FADT rev. 2 */
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/acpi/Makefile b/arch/i386/kernel/acpi/Makefile
+--- a/arch/i386/kernel/acpi/Makefile 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/acpi/Makefile 2007-03-14 10:55:14.000000000 +0100
+@@ -8,3 +8,7 @@
+ obj-y += cstate.o processor.o
+ endif
+
++ifdef CONFIG_XEN
++include $(srctree)/scripts/Makefile.xen
++obj-y := $(call cherrypickxen, $(obj-y), $(src))
++endif
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/alternative.c b/arch/i386/kernel/alternative.c
+--- a/arch/i386/kernel/alternative.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/alternative.c 2007-03-14 10:55:14.000000000 +0100
+@@ -5,7 +5,11 @@
+ #include <asm/alternative.h>
+ #include <asm/sections.h>
+
++#ifdef CONFIG_X86_64_XEN
++static int no_replacement = 1;
++#else
+ static int no_replacement = 0;
++#endif
+ static int smp_alt_once = 0;
+ static int debug_alternative = 0;
+
+@@ -165,7 +169,11 @@
+ #ifdef CONFIG_X86_64
+ /* vsyscall code is not mapped yet. resolve it manually. */
+ if (instr >= (u8 *)VSYSCALL_START && instr < (u8*)VSYSCALL_END) {
++#ifdef CONFIG_XEN
++ instr = __va(instr - (u8*)VSYSCALL_START + (u8*)phys_to_machine(__pa_symbol(&__vsyscall_0)));
++#else
+ instr = __va(instr - (u8*)VSYSCALL_START + (u8*)__pa_symbol(&__vsyscall_0));
++#endif
+ DPRINTK("%s: vsyscall fixup: %p => %p\n",
+ __FUNCTION__, a->instr, instr);
+ }
+@@ -388,9 +396,12 @@
+ unsigned long flags;
+ if (no_replacement) {
+ printk(KERN_INFO "(SMP-)alternatives turned off\n");
++#ifndef CONFIG_X86_64
++/* ToDo: x86_64 put something strange there, not sure what yet */
+ free_init_pages("SMP alternatives",
+ (unsigned long)__smp_alt_begin,
+ (unsigned long)__smp_alt_end);
++#endif
+ return;
+ }
+
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/apic-xen.c b/arch/i386/kernel/apic-xen.c
+--- a/arch/i386/kernel/apic-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/apic-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,223 @@
+/*
-+ * boot.c - Architecture-Specific Low-Level ACPI Boot Support
-+ *
-+ * Copyright (C) 2001, 2002 Paul Diefenbaugh <paul.s.diefenbaugh at intel.com>
-+ * Copyright (C) 2001 Jun Nakajima <jun.nakajima at intel.com>
-+ *
-+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ * Local APIC handling, local APIC timers
+ *
-+ * 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) 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/acpi.h>
-+#include <linux/efi.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/cpu.h>
+#include <linux/module.h>
-+#include <linux/dmi.h>
-+#include <linux/irq.h>
+
-+#include <asm/pgtable.h>
-+#include <asm/io_apic.h>
-+#include <asm/apic.h>
-+#include <asm/io.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/i8253.h>
++#include <asm/nmi.h>
+
-+#ifdef CONFIG_X86_64
-+
-+extern void __init clustered_apic_check(void);
-+
-+extern int gsi_irq_sharing(int gsi);
-+#include <asm/proto.h>
-+
-+static inline int acpi_madt_oem_check(char *oem_id, char *oem_table_id) { return 0; }
-+
-+
-+#else /* X86 */
-+
-+#ifdef CONFIG_X86_LOCAL_APIC
+#include <mach_apic.h>
-+#include <mach_mpparse.h>
-+#endif /* CONFIG_X86_LOCAL_APIC */
-+
-+static inline int gsi_irq_sharing(int gsi) { return gsi; }
-+
-+#endif /* X86 */
-+
-+#define BAD_MADT_ENTRY(entry, end) ( \
-+ (!entry) || (unsigned long)entry + sizeof(*entry) > end || \
-+ ((acpi_table_entry_header *)entry)->length < sizeof(*entry))
-+
-+#define PREFIX "ACPI: "
-+
-+int acpi_noirq __initdata; /* skip ACPI IRQ initialization */
-+int acpi_pci_disabled __initdata; /* skip ACPI PCI scan and IRQ initialization */
-+int acpi_ht __initdata = 1; /* enable HT */
-+
-+int acpi_lapic;
-+int acpi_ioapic;
-+int acpi_strict;
-+EXPORT_SYMBOL(acpi_strict);
-+
-+acpi_interrupt_flags acpi_sci_flags __initdata;
-+int acpi_sci_override_gsi __initdata;
-+int acpi_skip_timer_override __initdata;
++#include <mach_apicdef.h>
++#include <mach_ipi.h>
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+static u64 acpi_lapic_addr __initdata = APIC_DEFAULT_PHYS_BASE;
-+#endif
++#include "io_ports.h"
+
-+#ifndef __HAVE_ARCH_CMPXCHG
-+#warning ACPI uses CMPXCHG, i486 and later hardware
++#ifndef CONFIG_XEN
++/*
++ * cpu_mask that denotes the CPUs that needs timer interrupt coming in as
++ * IPIs in place of local APIC timers
++ */
++static cpumask_t timer_bcast_ipi;
+#endif
+
-+#define MAX_MADT_ENTRIES 256
-+u8 x86_acpiid_to_apicid[MAX_MADT_ENTRIES] =
-+ {[0 ... MAX_MADT_ENTRIES - 1] = 0xff };
-+EXPORT_SYMBOL(x86_acpiid_to_apicid);
-+
-+/* --------------------------------------------------------------------------
-+ Boot-time Configuration
-+ -------------------------------------------------------------------------- */
-+
+/*
-+ * The default interrupt routing model is PIC (8259). This gets
-+ * overriden if IOAPICs are enumerated (below).
++ * Knob to control our willingness to enable the local APIC.
+ */
-+enum acpi_irq_model_id acpi_irq_model = ACPI_IRQ_MODEL_PIC;
-+
-+#if defined(CONFIG_X86_64) && !defined(CONFIG_XEN)
++static int enable_local_apic __initdata = 0; /* -1=force-disable, +1=force-enable */
+
-+/* rely on all ACPI tables being in the direct mapping */
-+char *__acpi_map_table(unsigned long phys_addr, unsigned long size)
++static inline void lapic_disable(void)
+{
-+ if (!phys_addr || !size)
-+ return NULL;
-+
-+ if (phys_addr+size <= (end_pfn_map << PAGE_SHIFT) + PAGE_SIZE)
-+ return __va(phys_addr);
-+
-+ return NULL;
++ enable_local_apic = -1;
++ clear_bit(X86_FEATURE_APIC, boot_cpu_data.x86_capability);
+}
+
-+#else
++static inline void lapic_enable(void)
++{
++ enable_local_apic = 1;
++}
+
+/*
-+ * Temporarily use the virtual area starting from FIX_IO_APIC_BASE_END,
-+ * to map the target physical address. The problem is that set_fixmap()
-+ * provides a single page, and it is possible that the page is not
-+ * sufficient.
-+ * By using this area, we can map up to MAX_IO_APICS pages temporarily,
-+ * i.e. until the next __va_range() call.
-+ *
-+ * Important Safety Note: The fixed I/O APIC page numbers are *subtracted*
-+ * from the fixed base. That's why we start at FIX_IO_APIC_BASE_END and
-+ * count idx down while incrementing the phys address.
++ * Debug level
+ */
-+char *__acpi_map_table(unsigned long phys, unsigned long size)
-+{
-+ unsigned long base, offset, mapped_size;
-+ int idx;
++int apic_verbosity;
+
++static int modern_apic(void)
++{
+#ifndef CONFIG_XEN
-+ if (phys + size < 8 * 1024 * 1024)
-+ return __va(phys);
++ unsigned int lvr, version;
++ /* AMD systems use old APIC versions, so check the CPU */
++ if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD &&
++ boot_cpu_data.x86 >= 0xf)
++ return 1;
++ lvr = apic_read(APIC_LVR);
++ version = GET_APIC_VERSION(lvr);
++ return version >= 0x14;
++#else
++ return 1;
+#endif
++}
+
-+ offset = phys & (PAGE_SIZE - 1);
-+ mapped_size = PAGE_SIZE - offset;
-+ set_fixmap(FIX_ACPI_END, phys);
-+ base = fix_to_virt(FIX_ACPI_END);
-+
++/*
++ * '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);
+ /*
-+ * Most cases can be covered by the below.
++ * 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 only ack when the APIC is enabled -AK
+ */
-+ idx = FIX_ACPI_END;
-+ while (mapped_size < size) {
-+ if (--idx < FIX_ACPI_BEGIN)
-+ return NULL; /* cannot handle this */
-+ phys += PAGE_SIZE;
-+ set_fixmap(idx, phys);
-+ mapped_size += PAGE_SIZE;
-+ }
-+
-+ return ((unsigned char *)base + offset);
++ if (cpu_has_apic)
++ ack_APIC_irq();
+}
-+#endif
-+
-+#ifdef CONFIG_PCI_MMCONFIG
-+/* The physical address of the MMCONFIG aperture. Set from ACPI tables. */
-+struct acpi_table_mcfg_config *pci_mmcfg_config;
-+int pci_mmcfg_config_num;
+
-+int __init acpi_parse_mcfg(unsigned long phys_addr, unsigned long size)
++#ifndef CONFIG_XEN
++void __init apic_intr_init(void)
+{
-+ struct acpi_table_mcfg *mcfg;
-+ unsigned long i;
-+ int config_size;
++#ifdef CONFIG_SMP
++ smp_intr_init();
++#endif
++ /* self generated IPI for local APIC timer */
++ set_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt);
+
-+ if (!phys_addr || !size)
-+ return -EINVAL;
++ /* IPI vectors for APIC spurious and error interrupts */
++ set_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt);
++ set_intr_gate(ERROR_APIC_VECTOR, error_interrupt);
+
-+ mcfg = (struct acpi_table_mcfg *)__acpi_map_table(phys_addr, size);
-+ if (!mcfg) {
-+ printk(KERN_WARNING PREFIX "Unable to map MCFG\n");
-+ return -ENODEV;
-+ }
++ /* thermal monitor LVT interrupt */
++#ifdef CONFIG_X86_MCE_P4THERMAL
++ set_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt);
++#endif
++}
+
-+ /* how many config structures do we have */
-+ pci_mmcfg_config_num = 0;
-+ i = size - sizeof(struct acpi_table_mcfg);
-+ while (i >= sizeof(struct acpi_table_mcfg_config)) {
-+ ++pci_mmcfg_config_num;
-+ i -= sizeof(struct acpi_table_mcfg_config);
-+ };
-+ if (pci_mmcfg_config_num == 0) {
-+ printk(KERN_ERR PREFIX "MMCONFIG has no entries\n");
-+ return -ENODEV;
-+ }
++/* Using APIC to generate smp_local_timer_interrupt? */
++int using_apic_timer __read_mostly = 0;
+
-+ config_size = pci_mmcfg_config_num * sizeof(*pci_mmcfg_config);
-+ pci_mmcfg_config = kmalloc(config_size, GFP_KERNEL);
-+ if (!pci_mmcfg_config) {
-+ printk(KERN_WARNING PREFIX
-+ "No memory for MCFG config tables\n");
-+ return -ENOMEM;
-+ }
++static int enabled_via_apicbase;
+
-+ memcpy(pci_mmcfg_config, &mcfg->config, config_size);
-+ for (i = 0; i < pci_mmcfg_config_num; ++i) {
-+ if (mcfg->config[i].base_reserved) {
-+ printk(KERN_ERR PREFIX
-+ "MMCONFIG not in low 4GB of memory\n");
-+ kfree(pci_mmcfg_config);
-+ pci_mmcfg_config_num = 0;
-+ return -ENODEV;
-+ }
-+ }
++void enable_NMI_through_LVT0 (void * dummy)
++{
++ unsigned int v, ver;
+
-+ return 0;
++ ver = apic_read(APIC_LVR);
++ ver = GET_APIC_VERSION(ver);
++ v = APIC_DM_NMI; /* unmask and set to NMI */
++ if (!APIC_INTEGRATED(ver)) /* 82489DX */
++ v |= APIC_LVT_LEVEL_TRIGGER;
++ apic_write_around(APIC_LVT0, v);
+}
-+#endif /* CONFIG_PCI_MMCONFIG */
++#endif /* !CONFIG_XEN */
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+static int __init acpi_parse_madt(unsigned long phys_addr, unsigned long size)
++int get_physical_broadcast(void)
+{
-+ struct acpi_table_madt *madt = NULL;
-+
-+ if (!phys_addr || !size || !cpu_has_apic)
-+ return -EINVAL;
-+
-+ madt = (struct acpi_table_madt *)__acpi_map_table(phys_addr, size);
-+ if (!madt) {
-+ printk(KERN_WARNING PREFIX "Unable to map MADT\n");
-+ return -ENODEV;
-+ }
-+
-+ if (madt->lapic_address) {
-+ acpi_lapic_addr = (u64) madt->lapic_address;
-+
-+ printk(KERN_DEBUG PREFIX "Local APIC address 0x%08x\n",
-+ madt->lapic_address);
-+ }
-+
-+ acpi_madt_oem_check(madt->header.oem_id, madt->header.oem_table_id);
-+
-+ return 0;
++ if (modern_apic())
++ return 0xff;
++ else
++ return 0xf;
+}
+
-+static int __init
-+acpi_parse_lapic(acpi_table_entry_header * header, const unsigned long end)
++#ifndef CONFIG_XEN
++#ifndef CONFIG_SMP
++static void up_apic_timer_interrupt_call(void)
+{
-+ struct acpi_table_lapic *processor = NULL;
-+
-+ processor = (struct acpi_table_lapic *)header;
-+
-+ if (BAD_MADT_ENTRY(processor, end))
-+ return -EINVAL;
-+
-+ acpi_table_print_madt_entry(header);
-+
-+ /* Record local apic id only when enabled */
-+ if (processor->flags.enabled)
-+ x86_acpiid_to_apicid[processor->acpi_id] = processor->id;
++ int cpu = smp_processor_id();
+
+ /*
-+ * We need to register disabled CPU as well to permit
-+ * counting disabled CPUs. This allows us to size
-+ * cpus_possible_map more accurately, to permit
-+ * to not preallocating memory for all NR_CPUS
-+ * when we use CPU hotplug.
++ * the NMI deadlock-detector uses this.
+ */
-+ mp_register_lapic(processor->id, /* APIC ID */
-+ processor->flags.enabled); /* Enabled? */
++ per_cpu(irq_stat, cpu).apic_timer_irqs++;
+
-+ return 0;
++ smp_local_timer_interrupt();
+}
++#endif
+
-+static int __init
-+acpi_parse_lapic_addr_ovr(acpi_table_entry_header * header,
-+ const unsigned long end)
++void smp_send_timer_broadcast_ipi(void)
+{
-+ struct acpi_table_lapic_addr_ovr *lapic_addr_ovr = NULL;
-+
-+ lapic_addr_ovr = (struct acpi_table_lapic_addr_ovr *)header;
-+
-+ if (BAD_MADT_ENTRY(lapic_addr_ovr, end))
-+ return -EINVAL;
-+
-+ acpi_lapic_addr = lapic_addr_ovr->address;
++ cpumask_t mask;
+
-+ return 0;
++ cpus_and(mask, cpu_online_map, timer_bcast_ipi);
++ if (!cpus_empty(mask)) {
++#ifdef CONFIG_SMP
++ send_IPI_mask(mask, LOCAL_TIMER_VECTOR);
++#else
++ /*
++ * We can directly call the apic timer interrupt handler
++ * in UP case. Minus all irq related functions
++ */
++ up_apic_timer_interrupt_call();
++#endif
++ }
+}
++#endif
+
-+static int __init
-+acpi_parse_lapic_nmi(acpi_table_entry_header * header, const unsigned long end)
++int setup_profiling_timer(unsigned int multiplier)
+{
-+ struct acpi_table_lapic_nmi *lapic_nmi = NULL;
-+
-+ lapic_nmi = (struct acpi_table_lapic_nmi *)header;
-+
-+ if (BAD_MADT_ENTRY(lapic_nmi, end))
-+ return -EINVAL;
-+
-+ acpi_table_print_madt_entry(header);
++ return -EINVAL;
++}
+
-+ if (lapic_nmi->lint != 1)
-+ printk(KERN_WARNING PREFIX "NMI not connected to LINT 1!\n");
++/*
++ * 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 0;
+}
+
-+#endif /*CONFIG_X86_LOCAL_APIC */
-+
-+#ifdef CONFIG_X86_IO_APIC
-+
-+static int __init
-+acpi_parse_ioapic(acpi_table_entry_header * header, const unsigned long end)
++static int __init parse_lapic(char *arg)
+{
-+ struct acpi_table_ioapic *ioapic = NULL;
-+
-+ ioapic = (struct acpi_table_ioapic *)header;
-+
-+ if (BAD_MADT_ENTRY(ioapic, end))
-+ return -EINVAL;
-+
-+ acpi_table_print_madt_entry(header);
-+
-+ mp_register_ioapic(ioapic->id,
-+ ioapic->address, ioapic->global_irq_base);
-+
++ lapic_enable();
+ return 0;
+}
++early_param("lapic", parse_lapic);
+
-+/*
-+ * Parse Interrupt Source Override for the ACPI SCI
-+ */
-+static void acpi_sci_ioapic_setup(u32 gsi, u16 polarity, u16 trigger)
++static int __init parse_nolapic(char *arg)
+{
-+ if (trigger == 0) /* compatible SCI trigger is level */
-+ trigger = 3;
-+
-+ if (polarity == 0) /* compatible SCI polarity is low */
-+ polarity = 3;
++ lapic_disable();
++ return 0;
++}
++early_param("nolapic", parse_nolapic);
+
-+ /* Command-line over-ride via acpi_sci= */
-+ if (acpi_sci_flags.trigger)
-+ trigger = acpi_sci_flags.trigger;
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/asm-offsets.c b/arch/i386/kernel/asm-offsets.c
+--- a/arch/i386/kernel/asm-offsets.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/asm-offsets.c 2007-03-14 10:55:14.000000000 +0100
+@@ -89,9 +89,14 @@
+ OFFSET(pbe_orig_address, pbe, orig_address);
+ OFFSET(pbe_next, pbe, next);
+
++#ifndef CONFIG_X86_NO_TSS
+ /* Offset from the sysenter stack to tss.esp0 */
+- DEFINE(TSS_sysenter_esp0, offsetof(struct tss_struct, esp0) -
++ DEFINE(SYSENTER_stack_esp0, offsetof(struct tss_struct, esp0) -
+ sizeof(struct tss_struct));
++#else
++ /* sysenter stack points directly to esp0 */
++ DEFINE(SYSENTER_stack_esp0, 0);
++#endif
+
+ DEFINE(PAGE_SIZE_asm, PAGE_SIZE);
+ DEFINE(VDSO_PRELINK, VDSO_PRELINK);
+@@ -111,4 +116,10 @@
+ OFFSET(PARAVIRT_iret, paravirt_ops, iret);
+ OFFSET(PARAVIRT_read_cr0, paravirt_ops, read_cr0);
+ #endif
+
-+ if (acpi_sci_flags.polarity)
-+ polarity = acpi_sci_flags.polarity;
+
-+ /*
-+ * mp_config_acpi_legacy_irqs() already setup IRQs < 16
-+ * If GSI is < 16, this will update its flags,
-+ * else it will create a new mp_irqs[] entry.
-+ */
-+ mp_override_legacy_irq(gsi, polarity, trigger, gsi);
++#ifdef CONFIG_XEN
++ BLANK();
++ OFFSET(XEN_START_mfn_list, start_info, mfn_list);
++#endif
+ }
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/cpu/common-xen.c b/arch/i386/kernel/cpu/common-xen.c
+--- a/arch/i386/kernel/cpu/common-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/cpu/common-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,868 @@
++#include <linux/init.h>
++#include <linux/string.h>
++#include <linux/delay.h>
++#include <linux/smp.h>
++#include <linux/module.h>
++#include <linux/percpu.h>
++#include <linux/bootmem.h>
++#include <asm/semaphore.h>
++#include <asm/processor.h>
++#include <asm/i387.h>
++#include <asm/msr.h>
++#include <asm/io.h>
++#include <asm/mmu_context.h>
++#include <asm/mtrr.h>
++#include <asm/mce.h>
++#ifdef CONFIG_X86_LOCAL_APIC
++#include <asm/mpspec.h>
++#include <asm/apic.h>
++#include <mach_apic.h>
++#endif
++#include <asm/pda.h>
++#include <asm/hypervisor.h>
+
-+ /*
-+ * stash over-ride to indicate we've been here
-+ * and for later update of acpi_fadt
-+ */
-+ acpi_sci_override_gsi = gsi;
-+ return;
-+}
++#include "cpu.h"
+
-+static int __init
-+acpi_parse_int_src_ovr(acpi_table_entry_header * header,
-+ const unsigned long end)
-+{
-+ struct acpi_table_int_src_ovr *intsrc = NULL;
++DEFINE_PER_CPU(struct Xgt_desc_struct, cpu_gdt_descr);
++EXPORT_PER_CPU_SYMBOL(cpu_gdt_descr);
+
-+ intsrc = (struct acpi_table_int_src_ovr *)header;
++struct i386_pda *_cpu_pda[NR_CPUS] __read_mostly;
++EXPORT_SYMBOL(_cpu_pda);
+
-+ if (BAD_MADT_ENTRY(intsrc, end))
-+ return -EINVAL;
++static int cachesize_override __cpuinitdata = -1;
++static int disable_x86_fxsr __cpuinitdata;
++static int disable_x86_serial_nr __cpuinitdata = 1;
++static int disable_x86_sep __cpuinitdata;
+
-+ acpi_table_print_madt_entry(header);
++struct cpu_dev * cpu_devs[X86_VENDOR_NUM] = {};
+
-+ if (intsrc->bus_irq == acpi_fadt.sci_int) {
-+ acpi_sci_ioapic_setup(intsrc->global_irq,
-+ intsrc->flags.polarity,
-+ intsrc->flags.trigger);
-+ return 0;
-+ }
++extern int disable_pse;
+
-+ if (acpi_skip_timer_override &&
-+ intsrc->bus_irq == 0 && intsrc->global_irq == 2) {
-+ printk(PREFIX "BIOS IRQ0 pin2 override ignored.\n");
-+ return 0;
++static void __cpuinit default_init(struct cpuinfo_x86 * c)
++{
++ /* Not much we can do here... */
++ /* Check if at least it has cpuid */
++ if (c->cpuid_level == -1) {
++ /* No cpuid. It must be an ancient CPU */
++ if (c->x86 == 4)
++ strcpy(c->x86_model_id, "486");
++ else if (c->x86 == 3)
++ strcpy(c->x86_model_id, "386");
+ }
++}
+
-+ mp_override_legacy_irq(intsrc->bus_irq,
-+ intsrc->flags.polarity,
-+ intsrc->flags.trigger, intsrc->global_irq);
++static struct cpu_dev __cpuinitdata default_cpu = {
++ .c_init = default_init,
++ .c_vendor = "Unknown",
++};
++static struct cpu_dev * this_cpu __cpuinitdata = &default_cpu;
+
-+ return 0;
++static int __init cachesize_setup(char *str)
++{
++ get_option (&str, &cachesize_override);
++ return 1;
+}
++__setup("cachesize=", cachesize_setup);
+
-+static int __init
-+acpi_parse_nmi_src(acpi_table_entry_header * header, const unsigned long end)
++int __cpuinit get_model_name(struct cpuinfo_x86 *c)
+{
-+ struct acpi_table_nmi_src *nmi_src = NULL;
-+
-+ nmi_src = (struct acpi_table_nmi_src *)header;
++ unsigned int *v;
++ char *p, *q;
+
-+ if (BAD_MADT_ENTRY(nmi_src, end))
-+ return -EINVAL;
++ if (cpuid_eax(0x80000000) < 0x80000004)
++ return 0;
+
-+ acpi_table_print_madt_entry(header);
++ 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;
+
-+ /* TBD: Support nimsrc entries? */
++ /* Intel chips right-justify this string for some dumb reason;
++ undo that brain damage */
++ p = q = &c->x86_model_id[0];
++ while ( *p == ' ' )
++ p++;
++ if ( p != q ) {
++ while ( *p )
++ *q++ = *p++;
++ while ( q <= &c->x86_model_id[48] )
++ *q++ = '\0'; /* Zero-pad the rest */
++ }
+
-+ return 0;
++ return 1;
+}
+
-+#endif /* CONFIG_X86_IO_APIC */
-+
-+/*
-+ * acpi_pic_sci_set_trigger()
-+ *
-+ * use ELCR to set PIC-mode trigger type for SCI
-+ *
-+ * If a PIC-mode SCI is not recognized or gives spurious IRQ7's
-+ * it may require Edge Trigger -- use "acpi_sci=edge"
-+ *
-+ * Port 0x4d0-4d1 are ECLR1 and ECLR2, the Edge/Level Control Registers
-+ * for the 8259 PIC. bit[n] = 1 means irq[n] is Level, otherwise Edge.
-+ * ECLR1 is IRQ's 0-7 (IRQ 0, 1, 2 must be 0)
-+ * ECLR2 is IRQ's 8-15 (IRQ 8, 13 must be 0)
-+ */
+
-+void __init acpi_pic_sci_set_trigger(unsigned int irq, u16 trigger)
++void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c)
+{
-+ unsigned int mask = 1 << irq;
-+ unsigned int old, new;
-+
-+ /* Real old ELCR mask */
-+ old = inb(0x4d0) | (inb(0x4d1) << 8);
++ unsigned int n, dummy, ecx, edx, l2size;
+
-+ /*
-+ * If we use ACPI to set PCI irq's, then we should clear ELCR
-+ * since we will set it correctly as we enable the PCI irq
-+ * routing.
-+ */
-+ new = acpi_noirq ? old : 0;
++ n = cpuid_eax(0x80000000);
+
-+ /*
-+ * Update SCI information in the ELCR, it isn't in the PCI
-+ * routing tables..
-+ */
-+ switch (trigger) {
-+ case 1: /* Edge - clear */
-+ new &= ~mask;
-+ break;
-+ case 3: /* Level - set */
-+ new |= mask;
-+ break;
++ if (n >= 0x80000005) {
++ cpuid(0x80000005, &dummy, &dummy, &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);
+ }
+
-+ if (old == new)
++ if (n < 0x80000006) /* Some chips just has a large L1. */
+ return;
+
-+ printk(PREFIX "setting ELCR to %04x (from %04x)\n", new, old);
-+ outb(new, 0x4d0);
-+ outb(new >> 8, 0x4d1);
-+}
-+
-+int acpi_gsi_to_irq(u32 gsi, unsigned int *irq)
-+{
-+#ifdef CONFIG_X86_IO_APIC
-+ if (use_pci_vector() && !platform_legacy_irq(gsi))
-+ *irq = IO_APIC_VECTOR(gsi);
-+ else
-+#endif
-+ *irq = gsi_irq_sharing(gsi);
-+ return 0;
-+}
++ ecx = cpuid_ecx(0x80000006);
++ l2size = ecx >> 16;
++
++ /* do processor-specific cache resizing */
++ if (this_cpu->c_size_cache)
++ l2size = this_cpu->c_size_cache(c,l2size);
+
-+/*
-+ * success: return IRQ number (>=0)
-+ * failure: return < 0
-+ */
-+int acpi_register_gsi(u32 gsi, int triggering, int polarity)
-+{
-+ unsigned int irq;
-+ unsigned int plat_gsi = gsi;
++ /* Allow user to override all this if necessary. */
++ if (cachesize_override != -1)
++ l2size = cachesize_override;
+
-+#ifdef CONFIG_PCI
-+ /*
-+ * Make sure all (legacy) PCI IRQs are set as level-triggered.
-+ */
-+ if (acpi_irq_model == ACPI_IRQ_MODEL_PIC) {
-+ extern void eisa_set_level_irq(unsigned int irq);
++ if ( l2size == 0 )
++ return; /* Again, no L2 cache is possible */
+
-+ if (triggering == ACPI_LEVEL_SENSITIVE)
-+ eisa_set_level_irq(gsi);
-+ }
-+#endif
++ c->x86_cache_size = l2size;
+
-+#ifdef CONFIG_X86_IO_APIC
-+ if (acpi_irq_model == ACPI_IRQ_MODEL_IOAPIC) {
-+ plat_gsi = mp_register_gsi(gsi, triggering, polarity);
-+ }
-+#endif
-+ acpi_gsi_to_irq(plat_gsi, &irq);
-+ return irq;
++ printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n",
++ l2size, ecx & 0xFF);
+}
+
-+EXPORT_SYMBOL(acpi_register_gsi);
++/* Naming convention should be: <Name> [(<Codename>)] */
++/* This table only is used unless init_<vendor>() below doesn't set it; */
++/* in particular, if CPUID levels 0x80000002..4 are supported, this isn't used */
+
-+/*
-+ * ACPI based hotplug support for CPU
-+ */
-+#ifdef CONFIG_ACPI_HOTPLUG_CPU
-+int acpi_map_lsapic(acpi_handle handle, int *pcpu)
++/* Look up CPU names by table lookup. */
++static char __cpuinit *table_lookup_model(struct cpuinfo_x86 *c)
+{
-+ /* TBD */
-+ return -EINVAL;
-+}
++ struct cpu_model_info *info;
+
-+EXPORT_SYMBOL(acpi_map_lsapic);
++ if ( c->x86_model >= 16 )
++ return NULL; /* Range check */
+
-+int acpi_unmap_lsapic(int cpu)
-+{
-+ /* TBD */
-+ return -EINVAL;
-+}
++ if (!this_cpu)
++ return NULL;
+
-+EXPORT_SYMBOL(acpi_unmap_lsapic);
-+#endif /* CONFIG_ACPI_HOTPLUG_CPU */
++ info = this_cpu->c_models;
+
-+int acpi_register_ioapic(acpi_handle handle, u64 phys_addr, u32 gsi_base)
-+{
-+ /* TBD */
-+ return -EINVAL;
++ while (info && info->family) {
++ if (info->family == c->x86)
++ return info->model_names[c->x86_model];
++ info++;
++ }
++ return NULL; /* Not found */
+}
+
-+EXPORT_SYMBOL(acpi_register_ioapic);
+
-+int acpi_unregister_ioapic(acpi_handle handle, u32 gsi_base)
++static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c, int early)
+{
-+ /* TBD */
-+ return -EINVAL;
++ char *v = c->x86_vendor_id;
++ int i;
++ static int printed;
++
++ for (i = 0; i < X86_VENDOR_NUM; i++) {
++ if (cpu_devs[i]) {
++ if (!strcmp(v,cpu_devs[i]->c_ident[0]) ||
++ (cpu_devs[i]->c_ident[1] &&
++ !strcmp(v,cpu_devs[i]->c_ident[1]))) {
++ c->x86_vendor = i;
++ if (!early)
++ this_cpu = cpu_devs[i];
++ return;
++ }
++ }
++ }
++ if (!printed) {
++ printed++;
++ printk(KERN_ERR "CPU: Vendor unknown, using generic init.\n");
++ printk(KERN_ERR "CPU: Your system may be unstable.\n");
++ }
++ c->x86_vendor = X86_VENDOR_UNKNOWN;
++ this_cpu = &default_cpu;
+}
+
-+EXPORT_SYMBOL(acpi_unregister_ioapic);
+
-+static unsigned long __init
-+acpi_scan_rsdp(unsigned long start, unsigned long length)
++static int __init x86_fxsr_setup(char * s)
+{
-+ unsigned long offset = 0;
-+ unsigned long sig_len = sizeof("RSD PTR ") - 1;
-+ unsigned long vstart = (unsigned long)isa_bus_to_virt(start);
++ /* Tell all the other CPU's to not use it... */
++ disable_x86_fxsr = 1;
+
+ /*
-+ * Scan all 16-byte boundaries of the physical memory region for the
-+ * RSDP signature.
++ * ... and clear the bits early in the boot_cpu_data
++ * so that the bootup process doesn't try to do this
++ * either.
+ */
-+ for (offset = 0; offset < length; offset += 16) {
-+ if (strncmp((char *)(vstart + offset), "RSD PTR ", sig_len))
-+ continue;
-+ return (start + offset);
-+ }
-+
-+ return 0;
++ clear_bit(X86_FEATURE_FXSR, boot_cpu_data.x86_capability);
++ clear_bit(X86_FEATURE_XMM, boot_cpu_data.x86_capability);
++ return 1;
+}
++__setup("nofxsr", x86_fxsr_setup);
+
-+static int __init acpi_parse_sbf(unsigned long phys_addr, unsigned long size)
-+{
-+ struct acpi_table_sbf *sb;
-+
-+ if (!phys_addr || !size)
-+ return -EINVAL;
-+
-+ sb = (struct acpi_table_sbf *)__acpi_map_table(phys_addr, size);
-+ if (!sb) {
-+ printk(KERN_WARNING PREFIX "Unable to map SBF\n");
-+ return -ENODEV;
-+ }
-+
-+ sbf_port = sb->sbf_cmos; /* Save CMOS port */
+
-+ return 0;
++static int __init x86_sep_setup(char * s)
++{
++ disable_x86_sep = 1;
++ return 1;
+}
++__setup("nosep", x86_sep_setup);
+
-+#ifdef CONFIG_HPET_TIMER
+
-+static int __init acpi_parse_hpet(unsigned long phys, unsigned long size)
++/* Standard macro to see if a specific flag is changeable */
++static inline int flag_is_changeable_p(u32 flag)
+{
-+ struct acpi_table_hpet *hpet_tbl;
++ u32 f1, f2;
+
-+ if (!phys || !size)
-+ return -EINVAL;
++ asm("pushfl\n\t"
++ "pushfl\n\t"
++ "popl %0\n\t"
++ "movl %0,%1\n\t"
++ "xorl %2,%0\n\t"
++ "pushl %0\n\t"
++ "popfl\n\t"
++ "pushfl\n\t"
++ "popl %0\n\t"
++ "popfl\n\t"
++ : "=&r" (f1), "=&r" (f2)
++ : "ir" (flag));
+
-+ hpet_tbl = (struct acpi_table_hpet *)__acpi_map_table(phys, size);
-+ if (!hpet_tbl) {
-+ printk(KERN_WARNING PREFIX "Unable to map HPET\n");
-+ return -ENODEV;
-+ }
++ return ((f1^f2) & flag) != 0;
++}
+
-+ if (hpet_tbl->addr.space_id != ACPI_SPACE_MEM) {
-+ printk(KERN_WARNING PREFIX "HPET timers must be located in "
-+ "memory.\n");
-+ return -1;
-+ }
-+#ifdef CONFIG_X86_64
-+ vxtime.hpet_address = hpet_tbl->addr.addrl |
-+ ((long)hpet_tbl->addr.addrh << 32);
-+
-+ printk(KERN_INFO PREFIX "HPET id: %#x base: %#lx\n",
-+ hpet_tbl->id, vxtime.hpet_address);
-+#else /* X86 */
-+ {
-+ extern unsigned long hpet_address;
+
-+ hpet_address = hpet_tbl->addr.addrl;
-+ printk(KERN_INFO PREFIX "HPET id: %#x base: %#lx\n",
-+ hpet_tbl->id, hpet_address);
-+ }
-+#endif /* X86 */
++/* Probe for the CPUID instruction */
++static int __cpuinit have_cpuid_p(void)
++{
++ return flag_is_changeable_p(X86_EFLAGS_ID);
++}
+
-+ return 0;
++void __init cpu_detect(struct cpuinfo_x86 *c)
++{
++ /* Get vendor name */
++ cpuid(0x00000000, &c->cpuid_level,
++ (int *)&c->x86_vendor_id[0],
++ (int *)&c->x86_vendor_id[8],
++ (int *)&c->x86_vendor_id[4]);
++
++ c->x86 = 4;
++ if (c->cpuid_level >= 0x00000001) {
++ u32 junk, tfms, cap0, misc;
++ cpuid(0x00000001, &tfms, &misc, &junk, &cap0);
++ c->x86 = (tfms >> 8) & 15;
++ c->x86_model = (tfms >> 4) & 15;
++ if (c->x86 == 0xf)
++ c->x86 += (tfms >> 20) & 0xff;
++ if (c->x86 >= 0x6)
++ c->x86_model += ((tfms >> 16) & 0xF) << 4;
++ c->x86_mask = tfms & 15;
++ if (cap0 & (1<<19))
++ c->x86_cache_alignment = ((misc >> 8) & 0xff) * 8;
++ }
+}
-+#else
-+#define acpi_parse_hpet NULL
-+#endif
+
-+#ifdef CONFIG_X86_PM_TIMER
-+extern u32 pmtmr_ioport;
-+#endif
++/* Do minimum CPU detection early.
++ Fields really needed: vendor, cpuid_level, family, model, mask, cache alignment.
++ The others are not touched to avoid unwanted side effects.
+
-+static int __init acpi_parse_fadt(unsigned long phys, unsigned long size)
++ WARNING: this function is only called on the BP. Don't add code here
++ that is supposed to run on all CPUs. */
++static void __init early_cpu_detect(void)
+{
-+ struct fadt_descriptor *fadt = NULL;
++ struct cpuinfo_x86 *c = &boot_cpu_data;
+
-+ fadt = (struct fadt_descriptor *)__acpi_map_table(phys, size);
-+ if (!fadt) {
-+ printk(KERN_WARNING PREFIX "Unable to map FADT\n");
-+ return 0;
-+ }
-+ /* initialize sci_int early for INT_SRC_OVR MADT parsing */
-+ acpi_fadt.sci_int = fadt->sci_int;
++ c->x86_cache_alignment = 32;
+
-+ /* initialize rev and apic_phys_dest_mode for x86_64 genapic */
-+ acpi_fadt.revision = fadt->revision;
-+ acpi_fadt.force_apic_physical_destination_mode =
-+ fadt->force_apic_physical_destination_mode;
++ if (!have_cpuid_p())
++ return;
+
-+#if defined(CONFIG_X86_PM_TIMER) && !defined(CONFIG_XEN)
-+ /* detect the location of the ACPI PM Timer */
-+ if (fadt->revision >= FADT2_REVISION_ID) {
-+ /* FADT rev. 2 */
-+ if (fadt->xpm_tmr_blk.address_space_id !=
-+ ACPI_ADR_SPACE_SYSTEM_IO)
-+ return 0;
++ cpu_detect(c);
+
-+ pmtmr_ioport = fadt->xpm_tmr_blk.address;
-+ /*
-+ * "X" fields are optional extensions to the original V1.0
-+ * fields, so we must selectively expand V1.0 fields if the
-+ * corresponding X field is zero.
-+ */
-+ if (!pmtmr_ioport)
-+ pmtmr_ioport = fadt->V1_pm_tmr_blk;
-+ } else {
-+ /* FADT rev. 1 */
-+ pmtmr_ioport = fadt->V1_pm_tmr_blk;
++ get_cpu_vendor(c, 1);
++}
++
++static void __cpuinit generic_identify(struct cpuinfo_x86 * c)
++{
++ u32 tfms, xlvl;
++ int ebx;
++
++ if (have_cpuid_p()) {
++ /* Get vendor name */
++ cpuid(0x00000000, &c->cpuid_level,
++ (int *)&c->x86_vendor_id[0],
++ (int *)&c->x86_vendor_id[8],
++ (int *)&c->x86_vendor_id[4]);
++
++ get_cpu_vendor(c, 0);
++ /* 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 capability, excap;
++ cpuid(0x00000001, &tfms, &ebx, &excap, &capability);
++ c->x86_capability[0] = capability;
++ c->x86_capability[4] = excap;
++ c->x86 = (tfms >> 8) & 15;
++ c->x86_model = (tfms >> 4) & 15;
++ if (c->x86 == 0xf)
++ c->x86 += (tfms >> 20) & 0xff;
++ if (c->x86 >= 0x6)
++ c->x86_model += ((tfms >> 16) & 0xF) << 4;
++ c->x86_mask = tfms & 15;
++#ifdef CONFIG_X86_HT
++ c->apicid = phys_pkg_id((ebx >> 24) & 0xFF, 0);
++#else
++ c->apicid = (ebx >> 24) & 0xFF;
++#endif
++ if (c->x86_capability[0] & (1<<19))
++ c->x86_clflush_size = ((ebx >> 8) & 0xff) * 8;
++ } else {
++ /* Have CPUID level 0 only - unheard of */
++ c->x86 = 4;
++ }
++
++ /* AMD-defined flags: level 0x80000001 */
++ xlvl = cpuid_eax(0x80000000);
++ 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 */
++ }
+ }
-+ if (pmtmr_ioport)
-+ printk(KERN_INFO PREFIX "PM-Timer IO Port: %#x\n",
-+ pmtmr_ioport);
++
++ early_intel_workaround(c);
++
++#ifdef CONFIG_X86_HT
++ c->phys_proc_id = (cpuid_ebx(1) >> 24) & 0xff;
+#endif
-+ return 0;
+}
+
-+unsigned long __init acpi_find_rsdp(void)
++static void __cpuinit squash_the_stupid_serial_number(struct cpuinfo_x86 *c)
+{
-+ unsigned long rsdp_phys = 0;
++ if (cpu_has(c, X86_FEATURE_PN) && disable_x86_serial_nr ) {
++ /* Disable processor serial number */
++ unsigned long lo,hi;
++ rdmsr(MSR_IA32_BBL_CR_CTL,lo,hi);
++ lo |= 0x200000;
++ wrmsr(MSR_IA32_BBL_CR_CTL,lo,hi);
++ printk(KERN_NOTICE "CPU serial number disabled.\n");
++ clear_bit(X86_FEATURE_PN, c->x86_capability);
+
-+ if (efi_enabled) {
-+ if (efi.acpi20 != EFI_INVALID_TABLE_ADDR)
-+ return efi.acpi20;
-+ else if (efi.acpi != EFI_INVALID_TABLE_ADDR)
-+ return efi.acpi;
++ /* Disabling the serial number may affect the cpuid level */
++ c->cpuid_level = cpuid_eax(0);
+ }
-+ /*
-+ * Scan memory looking for the RSDP signature. First search EBDA (low
-+ * memory) paragraphs and then search upper memory (E0000-FFFFF).
-+ */
-+ rsdp_phys = acpi_scan_rsdp(0, 0x400);
-+ if (!rsdp_phys)
-+ rsdp_phys = acpi_scan_rsdp(0xE0000, 0x20000);
++}
+
-+ return rsdp_phys;
++static int __init x86_serial_nr_setup(char *s)
++{
++ disable_x86_serial_nr = 0;
++ return 1;
+}
++__setup("serialnumber", x86_serial_nr_setup);
++
++
+
-+#ifdef CONFIG_X86_LOCAL_APIC
+/*
-+ * Parse LAPIC entries in MADT
-+ * returns 0 on success, < 0 on error
++ * This does the hard work of actually picking apart the CPU stuff...
+ */
-+static int __init acpi_parse_madt_lapic_entries(void)
++void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
+{
-+ int count;
-+
-+ if (!cpu_has_apic)
-+ return -ENODEV;
++ int i;
+
-+ /*
-+ * Note that the LAPIC address is obtained from the MADT (32-bit value)
-+ * and (optionally) overriden by a LAPIC_ADDR_OVR entry (64-bit value).
-+ */
++ c->loops_per_jiffy = loops_per_jiffy;
++ c->x86_cache_size = -1;
++ c->x86_vendor = X86_VENDOR_UNKNOWN;
++ c->cpuid_level = -1; /* CPUID not detected */
++ 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_max_cores = 1;
++ c->x86_clflush_size = 32;
++ memset(&c->x86_capability, 0, sizeof c->x86_capability);
+
-+ count =
-+ acpi_table_parse_madt(ACPI_MADT_LAPIC_ADDR_OVR,
-+ acpi_parse_lapic_addr_ovr, 0);
-+ if (count < 0) {
-+ printk(KERN_ERR PREFIX
-+ "Error parsing LAPIC address override entry\n");
-+ return count;
++ if (!have_cpuid_p()) {
++ /* First of all, decide if this is a 486 or higher */
++ /* It's a 486 if we can modify the AC flag */
++ if ( flag_is_changeable_p(X86_EFLAGS_AC) )
++ c->x86 = 4;
++ else
++ c->x86 = 3;
+ }
+
-+ mp_register_lapic_address(acpi_lapic_addr);
++ generic_identify(c);
+
-+ count = acpi_table_parse_madt(ACPI_MADT_LAPIC, acpi_parse_lapic,
-+ MAX_APICS);
-+ if (!count) {
-+ printk(KERN_ERR PREFIX "No LAPIC entries present\n");
-+ /* TBD: Cleanup to allow fallback to MPS */
-+ return -ENODEV;
-+ } else if (count < 0) {
-+ printk(KERN_ERR PREFIX "Error parsing LAPIC entry\n");
-+ /* TBD: Cleanup to allow fallback to MPS */
-+ return count;
-+ }
++ printk(KERN_DEBUG "CPU: After generic identify, caps:");
++ for (i = 0; i < NCAPINTS; i++)
++ printk(" %08lx", c->x86_capability[i]);
++ printk("\n");
+
-+ count =
-+ acpi_table_parse_madt(ACPI_MADT_LAPIC_NMI, acpi_parse_lapic_nmi, 0);
-+ if (count < 0) {
-+ printk(KERN_ERR PREFIX "Error parsing LAPIC NMI entry\n");
-+ /* TBD: Cleanup to allow fallback to MPS */
-+ return count;
-+ }
-+ return 0;
-+}
-+#endif /* CONFIG_X86_LOCAL_APIC */
++ if (this_cpu->c_identify) {
++ this_cpu->c_identify(c);
+
-+#ifdef CONFIG_X86_IO_APIC
-+/*
-+ * Parse IOAPIC related entries in MADT
-+ * returns 0 on success, < 0 on error
-+ */
-+static int __init acpi_parse_madt_ioapic_entries(void)
-+{
-+ int count;
++ printk(KERN_DEBUG "CPU: After vendor identify, caps:");
++ for (i = 0; i < NCAPINTS; i++)
++ printk(" %08lx", c->x86_capability[i]);
++ printk("\n");
++ }
+
+ /*
-+ * ACPI interpreter is required to complete interrupt setup,
-+ * so if it is off, don't enumerate the io-apics with ACPI.
-+ * If MPS is present, it will handle them,
-+ * otherwise the system will stay in PIC mode
++ * 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!
+ */
-+ if (acpi_disabled || acpi_noirq) {
-+ return -ENODEV;
-+ }
++ if (this_cpu->c_init)
++ this_cpu->c_init(c);
+
-+ if (!cpu_has_apic)
-+ return -ENODEV;
++ /* Disable the PN if appropriate */
++ squash_the_stupid_serial_number(c);
+
+ /*
-+ * if "noapic" boot option, don't look for IO-APICs
++ * The vendor-specific functions might have changed features. Now
++ * we do "generic changes."
+ */
-+ if (skip_ioapic_setup) {
-+ printk(KERN_INFO PREFIX "Skipping IOAPIC probe "
-+ "due to 'noapic' option.\n");
-+ return -ENODEV;
-+ }
+
-+ count =
-+ acpi_table_parse_madt(ACPI_MADT_IOAPIC, acpi_parse_ioapic,
-+ MAX_IO_APICS);
-+ if (!count) {
-+ printk(KERN_ERR PREFIX "No IOAPIC entries present\n");
-+ return -ENODEV;
-+ } else if (count < 0) {
-+ printk(KERN_ERR PREFIX "Error parsing IOAPIC entry\n");
-+ return count;
++ /* TSC disabled? */
++ if ( tsc_disable )
++ clear_bit(X86_FEATURE_TSC, c->x86_capability);
++
++ /* FXSR disabled? */
++ if (disable_x86_fxsr) {
++ clear_bit(X86_FEATURE_FXSR, c->x86_capability);
++ clear_bit(X86_FEATURE_XMM, c->x86_capability);
+ }
+
-+ count =
-+ acpi_table_parse_madt(ACPI_MADT_INT_SRC_OVR, acpi_parse_int_src_ovr,
-+ NR_IRQ_VECTORS);
-+ if (count < 0) {
-+ printk(KERN_ERR PREFIX
-+ "Error parsing interrupt source overrides entry\n");
-+ /* TBD: Cleanup to allow fallback to MPS */
-+ return count;
++ /* SEP disabled? */
++ if (disable_x86_sep)
++ clear_bit(X86_FEATURE_SEP, c->x86_capability);
++
++ if (disable_pse)
++ clear_bit(X86_FEATURE_PSE, c->x86_capability);
++
++ /* If the model name is still unset, do table lookup. */
++ if ( !c->x86_model_id[0] ) {
++ char *p;
++ p = table_lookup_model(c);
++ if ( p )
++ strcpy(c->x86_model_id, p);
++ else
++ /* Last resort... */
++ sprintf(c->x86_model_id, "%02x/%02x",
++ c->x86, c->x86_model);
+ }
+
++ /* Now the feature flags better reflect actual CPU features! */
++
++ printk(KERN_DEBUG "CPU: After all inits, caps:");
++ for (i = 0; i < NCAPINTS; i++)
++ printk(" %08lx", c->x86_capability[i]);
++ printk("\n");
++
+ /*
-+ * If BIOS did not supply an INT_SRC_OVR for the SCI
-+ * pretend we got one so we can set the SCI flags.
++ * 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 (!acpi_sci_override_gsi)
-+ acpi_sci_ioapic_setup(acpi_fadt.sci_int, 0, 0);
++ 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];
++ }
+
-+ /* Fill in identity legacy mapings where no override */
-+ mp_config_acpi_legacy_irqs();
++ /* Init Machine Check Exception if available. */
++ mcheck_init(c);
+
-+ count =
-+ acpi_table_parse_madt(ACPI_MADT_NMI_SRC, acpi_parse_nmi_src,
-+ NR_IRQ_VECTORS);
-+ if (count < 0) {
-+ printk(KERN_ERR PREFIX "Error parsing NMI SRC entry\n");
-+ /* TBD: Cleanup to allow fallback to MPS */
-+ return count;
-+ }
++ if (c == &boot_cpu_data)
++ sysenter_setup();
++ enable_sep_cpu();
+
-+ return 0;
-+}
-+#else
-+static inline int acpi_parse_madt_ioapic_entries(void)
-+{
-+ return -1;
++ if (c == &boot_cpu_data)
++ mtrr_bp_init();
++ else
++ mtrr_ap_init();
+}
-+#endif /* !CONFIG_X86_IO_APIC */
+
-+static void __init acpi_process_madt(void)
++#ifdef CONFIG_X86_HT
++void __cpuinit detect_ht(struct cpuinfo_x86 *c)
+{
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ int count, error;
++ u32 eax, ebx, ecx, edx;
++ int index_msb, core_bits;
+
-+ count = acpi_table_parse(ACPI_APIC, acpi_parse_madt);
-+ if (count >= 1) {
++ cpuid(1, &eax, &ebx, &ecx, &edx);
+
-+ /*
-+ * Parse MADT LAPIC entries
-+ */
-+ error = acpi_parse_madt_lapic_entries();
-+ if (!error) {
-+ acpi_lapic = 1;
++ if (!cpu_has(c, X86_FEATURE_HT) || cpu_has(c, X86_FEATURE_CMP_LEGACY))
++ return;
+
-+#ifdef CONFIG_X86_GENERICARCH
-+ generic_bigsmp_probe();
-+#endif
-+ /*
-+ * Parse MADT IO-APIC entries
-+ */
-+ error = acpi_parse_madt_ioapic_entries();
-+ if (!error) {
-+ acpi_irq_model = ACPI_IRQ_MODEL_IOAPIC;
-+ acpi_irq_balance_set(NULL);
-+ acpi_ioapic = 1;
++ smp_num_siblings = (ebx & 0xff0000) >> 16;
+
-+ smp_found_config = 1;
-+ clustered_apic_check();
-+ }
-+ }
-+ if (error == -EINVAL) {
-+ /*
-+ * Dell Precision Workstation 410, 610 come here.
-+ */
-+ printk(KERN_ERR PREFIX
-+ "Invalid BIOS MADT, disabling ACPI\n");
-+ disable_acpi();
++ 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;
+ }
-+ }
-+#endif
-+ return;
-+}
+
-+extern int acpi_force;
++ index_msb = get_count_order(smp_num_siblings);
++ c->phys_proc_id = phys_pkg_id((ebx >> 24) & 0xFF, index_msb);
+
-+#ifdef __i386__
++ printk(KERN_INFO "CPU: Physical Processor ID: %d\n",
++ c->phys_proc_id);
+
-+static int __init disable_acpi_irq(struct dmi_system_id *d)
-+{
-+ if (!acpi_force) {
-+ printk(KERN_NOTICE "%s detected: force use of acpi=noirq\n",
-+ d->ident);
-+ acpi_noirq_set();
-+ }
-+ return 0;
-+}
++ smp_num_siblings = smp_num_siblings / c->x86_max_cores;
+
-+static int __init disable_acpi_pci(struct dmi_system_id *d)
-+{
-+ if (!acpi_force) {
-+ printk(KERN_NOTICE "%s detected: force use of pci=noacpi\n",
-+ d->ident);
-+ acpi_disable_pci();
-+ }
-+ return 0;
-+}
++ index_msb = get_count_order(smp_num_siblings) ;
+
-+static int __init dmi_disable_acpi(struct dmi_system_id *d)
-+{
-+ if (!acpi_force) {
-+ printk(KERN_NOTICE "%s detected: acpi off\n", d->ident);
-+ disable_acpi();
-+ } else {
-+ printk(KERN_NOTICE
-+ "Warning: DMI blacklist says broken, but acpi forced\n");
++ core_bits = get_count_order(c->x86_max_cores);
++
++ c->cpu_core_id = phys_pkg_id((ebx >> 24) & 0xFF, index_msb) &
++ ((1 << core_bits) - 1);
++
++ if (c->x86_max_cores > 1)
++ printk(KERN_INFO "CPU: Processor Core ID: %d\n",
++ c->cpu_core_id);
+ }
-+ return 0;
+}
++#endif
+
-+/*
-+ * Limit ACPI to CPU enumeration for HT
-+ */
-+static int __init force_acpi_ht(struct dmi_system_id *d)
++void __cpuinit print_cpu_info(struct cpuinfo_x86 *c)
+{
-+ if (!acpi_force) {
-+ printk(KERN_NOTICE "%s detected: force use of acpi=ht\n",
-+ d->ident);
-+ disable_acpi();
-+ acpi_ht = 1;
-+ } else {
-+ printk(KERN_NOTICE
-+ "Warning: acpi=force overrules DMI blacklist: acpi=ht\n");
-+ }
-+ return 0;
-+}
++ char *vendor = NULL;
+
-+/*
-+ * If your system is blacklisted here, but you find that acpi=force
-+ * works for you, please contact acpi-devel at sourceforge.net
-+ */
-+static struct dmi_system_id __initdata acpi_dmi_table[] = {
-+ /*
-+ * Boxes that need ACPI disabled
-+ */
-+ {
-+ .callback = dmi_disable_acpi,
-+ .ident = "IBM Thinkpad",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "IBM"),
-+ DMI_MATCH(DMI_BOARD_NAME, "2629H1G"),
-+ },
-+ },
++ if (c->x86_vendor < X86_VENDOR_NUM)
++ vendor = this_cpu->c_vendor;
++ else if (c->cpuid_level >= 0)
++ vendor = c->x86_vendor_id;
+
-+ /*
-+ * Boxes that need acpi=ht
-+ */
-+ {
-+ .callback = force_acpi_ht,
-+ .ident = "FSC Primergy T850",
-+ .matches = {
-+ DMI_MATCH(DMI_SYS_VENDOR, "FUJITSU SIEMENS"),
-+ DMI_MATCH(DMI_PRODUCT_NAME, "PRIMERGY T850"),
-+ },
-+ },
-+ {
-+ .callback = force_acpi_ht,
-+ .ident = "DELL GX240",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "Dell Computer Corporation"),
-+ DMI_MATCH(DMI_BOARD_NAME, "OptiPlex GX240"),
-+ },
-+ },
-+ {
-+ .callback = force_acpi_ht,
-+ .ident = "HP VISUALIZE NT Workstation",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "Hewlett-Packard"),
-+ DMI_MATCH(DMI_PRODUCT_NAME, "HP VISUALIZE NT Workstation"),
-+ },
-+ },
-+ {
-+ .callback = force_acpi_ht,
-+ .ident = "Compaq Workstation W8000",
-+ .matches = {
-+ DMI_MATCH(DMI_SYS_VENDOR, "Compaq"),
-+ DMI_MATCH(DMI_PRODUCT_NAME, "Workstation W8000"),
-+ },
-+ },
-+ {
-+ .callback = force_acpi_ht,
-+ .ident = "ASUS P4B266",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."),
-+ DMI_MATCH(DMI_BOARD_NAME, "P4B266"),
-+ },
-+ },
-+ {
-+ .callback = force_acpi_ht,
-+ .ident = "ASUS P2B-DS",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."),
-+ DMI_MATCH(DMI_BOARD_NAME, "P2B-DS"),
-+ },
-+ },
-+ {
-+ .callback = force_acpi_ht,
-+ .ident = "ASUS CUR-DLS",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."),
-+ DMI_MATCH(DMI_BOARD_NAME, "CUR-DLS"),
-+ },
-+ },
-+ {
-+ .callback = force_acpi_ht,
-+ .ident = "ABIT i440BX-W83977",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "ABIT <http://www.abit.com>"),
-+ DMI_MATCH(DMI_BOARD_NAME, "i440BX-W83977 (BP6)"),
-+ },
-+ },
-+ {
-+ .callback = force_acpi_ht,
-+ .ident = "IBM Bladecenter",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "IBM"),
-+ DMI_MATCH(DMI_BOARD_NAME, "IBM eServer BladeCenter HS20"),
-+ },
-+ },
-+ {
-+ .callback = force_acpi_ht,
-+ .ident = "IBM eServer xSeries 360",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "IBM"),
-+ DMI_MATCH(DMI_BOARD_NAME, "eServer xSeries 360"),
-+ },
-+ },
-+ {
-+ .callback = force_acpi_ht,
-+ .ident = "IBM eserver xSeries 330",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "IBM"),
-+ DMI_MATCH(DMI_BOARD_NAME, "eserver xSeries 330"),
-+ },
-+ },
-+ {
-+ .callback = force_acpi_ht,
-+ .ident = "IBM eserver xSeries 440",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "IBM"),
-+ DMI_MATCH(DMI_PRODUCT_NAME, "eserver xSeries 440"),
-+ },
-+ },
++ if (vendor && strncmp(c->x86_model_id, vendor, strlen(vendor)))
++ printk("%s ", vendor);
+
-+ /*
-+ * Boxes that need ACPI PCI IRQ routing disabled
-+ */
-+ {
-+ .callback = disable_acpi_irq,
-+ .ident = "ASUS A7V",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC"),
-+ DMI_MATCH(DMI_BOARD_NAME, "<A7V>"),
-+ /* newer BIOS, Revision 1011, does work */
-+ DMI_MATCH(DMI_BIOS_VERSION,
-+ "ASUS A7V ACPI BIOS Revision 1007"),
-+ },
-+ },
++ if (!c->x86_model_id[0])
++ printk("%d86", c->x86);
++ else
++ printk("%s", c->x86_model_id);
+
-+ /*
-+ * Boxes that need ACPI PCI IRQ routing and PCI scan disabled
-+ */
-+ { /* _BBN 0 bug */
-+ .callback = disable_acpi_pci,
-+ .ident = "ASUS PR-DLS",
-+ .matches = {
-+ DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."),
-+ DMI_MATCH(DMI_BOARD_NAME, "PR-DLS"),
-+ DMI_MATCH(DMI_BIOS_VERSION,
-+ "ASUS PR-DLS ACPI BIOS Revision 1010"),
-+ DMI_MATCH(DMI_BIOS_DATE, "03/21/2003")
-+ },
-+ },
-+ {
-+ .callback = disable_acpi_pci,
-+ .ident = "Acer TravelMate 36x Laptop",
-+ .matches = {
-+ DMI_MATCH(DMI_SYS_VENDOR, "Acer"),
-+ DMI_MATCH(DMI_PRODUCT_NAME, "TravelMate 360"),
-+ },
-+ },
-+ {}
-+};
++ if (c->x86_mask || c->cpuid_level >= 0)
++ printk(" stepping %02x\n", c->x86_mask);
++ else
++ printk("\n");
++}
+
-+#endif /* __i386__ */
++cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE;
+
-+/*
-+ * acpi_boot_table_init() and acpi_boot_init()
-+ * called from setup_arch(), always.
-+ * 1. checksums all tables
-+ * 2. enumerates lapics
-+ * 3. enumerates io-apics
-+ *
-+ * acpi_table_init() is separate to allow reading SRAT without
-+ * other side effects.
-+ *
-+ * side effects of acpi_boot_init:
-+ * acpi_lapic = 1 if LAPIC found
-+ * acpi_ioapic = 1 if IOAPIC found
-+ * if (acpi_lapic && acpi_ioapic) smp_found_config = 1;
-+ * if acpi_blacklisted() acpi_disabled = 1;
-+ * acpi_irq_model=...
-+ * ...
-+ *
-+ * return value: (currently ignored)
-+ * 0: success
-+ * !0: failure
++/* This is hacky. :)
++ * We're emulating future behavior.
++ * In the future, the cpu-specific init functions will be called implicitly
++ * via the magic of initcalls.
++ * They will insert themselves into the cpu_devs structure.
++ * Then, when cpu_init() is called, we can just iterate over that array.
+ */
+
-+int __init acpi_boot_table_init(void)
++extern int intel_cpu_init(void);
++extern int cyrix_init_cpu(void);
++extern int nsc_init_cpu(void);
++extern int amd_init_cpu(void);
++extern int centaur_init_cpu(void);
++extern int transmeta_init_cpu(void);
++extern int rise_init_cpu(void);
++extern int nexgen_init_cpu(void);
++extern int umc_init_cpu(void);
++
++void __init early_cpu_init(void)
+{
-+ int error;
++ intel_cpu_init();
++ cyrix_init_cpu();
++ nsc_init_cpu();
++ amd_init_cpu();
++ centaur_init_cpu();
++ transmeta_init_cpu();
++ rise_init_cpu();
++ nexgen_init_cpu();
++ umc_init_cpu();
++ early_cpu_detect();
+
-+#ifdef __i386__
-+ dmi_check_system(acpi_dmi_table);
++#ifdef CONFIG_DEBUG_PAGEALLOC
++ /* pse is not compatible with on-the-fly unmapping,
++ * disable it even if the cpus claim to support it.
++ */
++ clear_bit(X86_FEATURE_PSE, boot_cpu_data.x86_capability);
++ disable_pse = 1;
+#endif
++}
+
-+ /*
-+ * If acpi_disabled, bail out
-+ * One exception: acpi=ht continues far enough to enumerate LAPICs
-+ */
-+ if (acpi_disabled && !acpi_ht)
-+ return 1;
++/* We can't move load_gdt to asm/desc.h because it lacks make_lowmen_page_readonly()
++ definition, and as this is still the only user of load_gdt in xen.
++ ToDo: JQ
++ */
+
-+ /*
-+ * Initialize the ACPI boot-time table parser.
-+ */
-+ error = acpi_table_init();
-+ if (error) {
-+ disable_acpi();
-+ return error;
++#ifdef CONFIG_XEN
++#undef load_gdt
++static void __cpuinit load_gdt(struct Xgt_desc_struct *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_lowmem_page_readonly(
++ (void *)va, XENFEAT_writable_descriptor_tables);
+ }
++ if (HYPERVISOR_set_gdt(frames, gdt_descr->size / 8))
++ BUG();
++}
++#endif /* CONFIG_XEN */
++
++/* Make sure %gs is initialized properly in idle threads */
++struct pt_regs * __devinit idle_regs(struct pt_regs *regs)
++{
++ memset(regs, 0, sizeof(struct pt_regs));
++ regs->xgs = __KERNEL_PDA;
++ return regs;
++}
+
-+ acpi_table_parse(ACPI_BOOT, acpi_parse_sbf);
++static __cpuinit int alloc_gdt(int cpu)
++{
++ struct Xgt_desc_struct *cpu_gdt_descr = &per_cpu(cpu_gdt_descr, cpu);
++ struct desc_struct *gdt;
++ struct i386_pda *pda;
++
++ gdt = (struct desc_struct *)cpu_gdt_descr->address;
++ pda = cpu_pda(cpu);
+
+ /*
-+ * blacklist may disable ACPI entirely
++ * This is a horrible hack to allocate the GDT. The problem
++ * is that cpu_init() is called really early for the boot CPU
++ * (and hence needs bootmem) but much later for the secondary
++ * CPUs, when bootmem will have gone away
+ */
-+ error = acpi_blacklisted();
-+ if (error) {
-+ if (acpi_force) {
-+ printk(KERN_WARNING PREFIX "acpi=force override\n");
-+ } else {
-+ printk(KERN_WARNING PREFIX "Disabling ACPI support\n");
-+ disable_acpi();
-+ return error;
++ if (NODE_DATA(0)->bdata->node_bootmem_map) {
++ BUG_ON(gdt != NULL || pda != NULL);
++
++ gdt = alloc_bootmem_pages(PAGE_SIZE);
++ pda = alloc_bootmem(sizeof(*pda));
++ /* alloc_bootmem(_pages) panics on failure, so no check */
++
++ memset(gdt, 0, PAGE_SIZE);
++ memset(pda, 0, sizeof(*pda));
++#ifdef CONFIG_XEN
++ memcpy(gdt, cpu_gdt_table, GDT_SIZE);
++ cpu_gdt_descr->size = GDT_SIZE;
++#endif
++ } else {
++ /* GDT and PDA might already have been allocated if
++ this is a CPU hotplug re-insertion. */
++ if (gdt == NULL)
++ gdt = (struct desc_struct *)get_zeroed_page(GFP_KERNEL);
++
++ if (pda == NULL)
++ pda = kmalloc_node(sizeof(*pda), GFP_KERNEL, cpu_to_node(cpu));
++
++ if (unlikely(!gdt || !pda)) {
++ free_pages((unsigned long)gdt, 0);
++ kfree(pda);
++ return 0;
+ }
+ }
+
-+ return 0;
++ cpu_gdt_descr->address = (unsigned long)gdt;
++ cpu_pda(cpu) = pda;
++
++ return 1;
+}
+
-+int __init acpi_boot_init(void)
++/* Initial PDA used by boot CPU */
++struct i386_pda boot_pda = {
++ ._pda = &boot_pda,
++ .cpu_number = 0,
++ .pcurrent = &init_task,
++};
++
++static inline void set_kernel_gs(void)
+{
-+ /*
-+ * If acpi_disabled, bail out
-+ * One exception: acpi=ht continues far enough to enumerate LAPICs
-+ */
-+ if (acpi_disabled && !acpi_ht)
-+ return 1;
++ /* Set %gs for this CPU's PDA. Memory clobber is to create a
++ barrier with respect to any PDA operations, so the compiler
++ doesn't move any before here. */
++ asm volatile ("mov %0, %%gs" : : "r" (__KERNEL_PDA) : "memory");
++}
+
-+ acpi_table_parse(ACPI_BOOT, acpi_parse_sbf);
++/* Initialize the CPU's GDT and PDA. The boot CPU does this for
++ itself, but secondaries find this done for them. */
++__cpuinit int init_gdt(int cpu, struct task_struct *idle)
++{
++ struct Xgt_desc_struct *cpu_gdt_descr = &per_cpu(cpu_gdt_descr, cpu);
++ struct desc_struct *gdt;
++ struct i386_pda *pda;
+
-+ /*
-+ * set sci_int and PM timer address
-+ */
-+ acpi_table_parse(ACPI_FADT, acpi_parse_fadt);
++ /* For non-boot CPUs, the GDT and PDA should already have been
++ allocated. */
++ if (!alloc_gdt(cpu)) {
++ printk(KERN_CRIT "CPU%d failed to allocate GDT or PDA\n", cpu);
++ return 0;
++ }
+
-+ /*
-+ * Process the Multiple APIC Description Table (MADT), if present
-+ */
-+ acpi_process_madt();
++ gdt = (struct desc_struct *)cpu_gdt_descr->address;
++ pda = cpu_pda(cpu);
+
-+ acpi_table_parse(ACPI_HPET, acpi_parse_hpet);
++ BUG_ON(gdt == NULL || pda == NULL);
+
-+ return 0;
-+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/acpi/Makefile linux-2.6.18-xen/arch/i386/kernel/acpi/Makefile
---- linux-2.6.18.3/arch/i386/kernel/acpi/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/kernel/acpi/Makefile 2006-11-19 14:26:21.000000000 +0100
-@@ -6,3 +6,7 @@
- obj-y += cstate.o processor.o
- endif
-
-+ifdef CONFIG_XEN
-+include $(srctree)/scripts/Makefile.xen
-+obj-y := $(call cherrypickxen, $(obj-y), $(src))
-+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/alternative.c linux-2.6.18-xen/arch/i386/kernel/alternative.c
---- linux-2.6.18.3/arch/i386/kernel/alternative.c 2006-12-06 09:06:08.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/alternative.c 2006-12-05 18:42:36.000000000 +0100
-@@ -4,7 +4,11 @@
- #include <asm/alternative.h>
- #include <asm/sections.h>
-
-+#ifdef CONFIG_X86_64_XEN
-+static int no_replacement = 1;
-+#else
- static int no_replacement = 0;
-+#endif
- static int smp_alt_once = 0;
- static int debug_alternative = 0;
-
-@@ -151,7 +155,11 @@
- #ifdef CONFIG_X86_64
- /* vsyscall code is not mapped yet. resolve it manually. */
- if (instr >= (u8 *)VSYSCALL_START && instr < (u8*)VSYSCALL_END) {
-+#ifdef CONFIG_XEN
-+ instr = __va(instr - (u8*)VSYSCALL_START + (u8*)phys_to_machine(__pa_symbol(&__vsyscall_0)));
-+#else
- instr = __va(instr - (u8*)VSYSCALL_START + (u8*)__pa_symbol(&__vsyscall_0));
++#ifndef CONFIG_XEN
++ /*
++ * Initialize the per-CPU GDT with the boot GDT,
++ * and set up the GDT descriptor:
++ */
++ memcpy(gdt, cpu_gdt_table, GDT_SIZE);
++ cpu_gdt_descr->size = GDT_SIZE - 1;
+#endif
- DPRINTK("%s: vsyscall fixup: %p => %p\n",
- __FUNCTION__, a->instr, instr);
- }
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/apic-xen.c linux-2.6.18-xen/arch/i386/kernel/apic-xen.c
---- linux-2.6.18.3/arch/i386/kernel/apic-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/apic-xen.c 2006-11-19 14:26:21.000000000 +0100
-@@ -0,0 +1,160 @@
-+/*
-+ * 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/cpu.h>
-+#include <linux/module.h>
++ if (cpu == 0)
++ pack_descriptor((u32 *)&gdt[GDT_ENTRY_PDA].a,
++ (u32 *)&gdt[GDT_ENTRY_PDA].b,
++ (unsigned long)pda, sizeof(*pda) - 1,
++ 0x80 | DESCTYPE_S | 0x2, 0); /* present read-write data segment */
++
++ memset(pda, 0, sizeof(*pda));
++ pda->_pda = pda;
++ pda->cpu_number = cpu;
++ pda->pcurrent = idle;
+
-+#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/i8253.h>
-+#include <asm/nmi.h>
++ return 1;
++}
+
-+#include <mach_apic.h>
-+#include <mach_apicdef.h>
-+#include <mach_ipi.h>
++void __cpuinit cpu_set_gdt(int cpu)
++{
++ struct Xgt_desc_struct *cpu_gdt_descr = &per_cpu(cpu_gdt_descr, cpu);
+
-+#include "io_ports.h"
++ /* Reinit these anyway, even if they've already been done (on
++ the boot CPU, this will transition from the boot gdt+pda to
++ the real ones). */
++ load_gdt(cpu_gdt_descr);
++ set_kernel_gs();
++}
+
-+#ifndef CONFIG_XEN
-+/*
-+ * cpu_mask that denotes the CPUs that needs timer interrupt coming in as
-+ * IPIs in place of local APIC timers
-+ */
-+static cpumask_t timer_bcast_ipi;
++/* Common CPU init for both boot and secondary CPUs */
++static void __cpuinit _cpu_init(int cpu, struct task_struct *curr)
++{
++#ifndef CONFIG_X86_NO_TSS
++ struct tss_struct * t = &per_cpu(init_tss, cpu);
+#endif
++ struct thread_struct *thread = &curr->thread;
+
-+/*
-+ * Knob to control our willingness to enable the local APIC.
-+ */
-+int enable_local_apic __initdata = 0; /* -1=force-disable, +1=force-enable */
++ if (cpu_test_and_set(cpu, cpu_initialized)) {
++ printk(KERN_WARNING "CPU#%d already initialized!\n", cpu);
++ for (;;) local_irq_enable();
++ }
+
-+/*
-+ * Debug level
-+ */
-+int apic_verbosity;
++ printk(KERN_INFO "Initializing CPU#%d\n", cpu);
+
-+static int modern_apic(void)
-+{
-+#ifndef CONFIG_XEN
-+ unsigned int lvr, version;
-+ /* AMD systems use old APIC versions, so check the CPU */
-+ if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD &&
-+ boot_cpu_data.x86 >= 0xf)
-+ return 1;
-+ lvr = apic_read(APIC_LVR);
-+ version = GET_APIC_VERSION(lvr);
-+ return version >= 0x14;
-+#else
-+ return 1;
++ if (cpu_has_vme || cpu_has_tsc || cpu_has_de)
++ clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE);
++ if (tsc_disable && cpu_has_tsc) {
++ printk(KERN_NOTICE "Disabling TSC...\n");
++ /**** FIX-HPA: DOES THIS REALLY BELONG HERE? ****/
++ clear_bit(X86_FEATURE_TSC, boot_cpu_data.x86_capability);
++ set_in_cr4(X86_CR4_TSD);
++ }
++
++#ifndef CONFIG_X86_NO_IDT
++ load_idt(&idt_descr);
+#endif
-+}
+
-+/*
-+ * '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 only ack when the APIC is enabled -AK
++ * Set up and load the per-CPU TSS and LDT
+ */
-+ if (cpu_has_apic)
-+ ack_APIC_irq();
-+}
++ atomic_inc(&init_mm.mm_count);
++ curr->active_mm = &init_mm;
++ if (curr->mm)
++ BUG();
++ enter_lazy_tlb(&init_mm, curr);
+
-+int get_physical_broadcast(void)
-+{
-+ if (modern_apic())
-+ return 0xff;
-+ else
-+ return 0xf;
-+}
++ load_esp0(t, thread);
++#ifndef CONFIG_X86_NO_TSS
++ set_tss_desc(cpu,t);
++ load_TR_desc();
++#endif
++ load_LDT(&init_mm.context);
+
-+#ifndef CONFIG_XEN
-+#ifndef CONFIG_SMP
-+static void up_apic_timer_interrupt_call(struct pt_regs *regs)
-+{
-+ int cpu = smp_processor_id();
++#ifdef CONFIG_DOUBLEFAULT
++ /* Set up doublefault TSS pointer in the GDT */
++ __set_tss_desc(cpu, GDT_ENTRY_DOUBLEFAULT_TSS, &doublefault_tss);
++#endif
++
++ /* Clear %fs. */
++ asm volatile ("mov %0, %%fs" : : "r" (0));
++
++ /* Clear all 6 debug registers: */
++ set_debugreg(0, 0);
++ set_debugreg(0, 1);
++ set_debugreg(0, 2);
++ set_debugreg(0, 3);
++ set_debugreg(0, 6);
++ set_debugreg(0, 7);
+
+ /*
-+ * the NMI deadlock-detector uses this.
++ * Force FPU initialization:
+ */
-+ per_cpu(irq_stat, cpu).apic_timer_irqs++;
-+
-+ smp_local_timer_interrupt(regs);
++ current_thread_info()->status = 0;
++ clear_used_math();
++ mxcsr_feature_mask_init();
+}
-+#endif
+
-+void smp_send_timer_broadcast_ipi(struct pt_regs *regs)
++/* Entrypoint to initialize secondary CPU */
++void __cpuinit secondary_cpu_init(void)
+{
-+ cpumask_t mask;
++ int cpu = smp_processor_id();
++ struct task_struct *curr = current;
+
-+ cpus_and(mask, cpu_online_map, timer_bcast_ipi);
-+ if (!cpus_empty(mask)) {
-+#ifdef CONFIG_SMP
-+ send_IPI_mask(mask, LOCAL_TIMER_VECTOR);
-+#else
-+ /*
-+ * We can directly call the apic timer interrupt handler
-+ * in UP case. Minus all irq related functions
-+ */
-+ up_apic_timer_interrupt_call(regs);
-+#endif
-+ }
++ _cpu_init(cpu, curr);
+}
-+#endif
+
-+int setup_profiling_timer(unsigned int multiplier)
++/*
++ * 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.
++ */
++void __cpuinit cpu_init(void)
+{
-+ return -EINVAL;
++ int cpu = smp_processor_id();
++ struct task_struct *curr = current;
++
++ /* Set up the real GDT and PDA, so we can transition from the
++ boot versions. */
++ if (!init_gdt(cpu, curr)) {
++ /* failed to allocate something; not much we can do... */
++ for (;;)
++ local_irq_enable();
++ }
++
++ cpu_set_gdt(cpu);
++ _cpu_init(cpu, curr);
+}
+
-+/*
-+ * This initializes the IO-APIC and APIC hardware if this is
-+ * a UP kernel.
-+ */
-+int __init APIC_init_uniprocessor (void)
++#ifdef CONFIG_HOTPLUG_CPU
++void __cpuinit cpu_uninit(void)
+{
-+#ifdef CONFIG_X86_IO_APIC
-+ if (smp_found_config)
-+ if (!skip_ioapic_setup && nr_ioapics)
-+ setup_IO_APIC();
-+#endif
++ int cpu = raw_smp_processor_id();
++ cpu_clear(cpu, cpu_initialized);
+
-+ return 0;
++ /* lazy TLB state */
++ per_cpu(cpu_tlbstate, cpu).state = 0;
++ per_cpu(cpu_tlbstate, cpu).active_mm = &init_mm;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/asm-offsets.c linux-2.6.18-xen/arch/i386/kernel/asm-offsets.c
---- linux-2.6.18.3/arch/i386/kernel/asm-offsets.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/kernel/asm-offsets.c 2006-11-19 14:26:21.000000000 +0100
-@@ -66,9 +66,14 @@
- OFFSET(pbe_orig_address, pbe, orig_address);
- OFFSET(pbe_next, pbe, next);
-
-+#ifndef CONFIG_X86_NO_TSS
- /* Offset from the sysenter stack to tss.esp0 */
-- DEFINE(TSS_sysenter_esp0, offsetof(struct tss_struct, esp0) -
-+ DEFINE(SYSENTER_stack_esp0, offsetof(struct tss_struct, esp0) -
- sizeof(struct tss_struct));
-+#else
-+ /* sysenter stack points directly to esp0 */
-+ DEFINE(SYSENTER_stack_esp0, 0);
+#endif
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/cpu/Makefile b/arch/i386/kernel/cpu/Makefile
+--- a/arch/i386/kernel/cpu/Makefile 2007-03-12 21:58:05.000000000 +0100
++++ b/arch/i386/kernel/cpu/Makefile 2007-03-14 10:55:14.000000000 +0100
+@@ -17,3 +17,8 @@
- DEFINE(PAGE_SIZE_asm, PAGE_SIZE);
- DEFINE(VDSO_PRELINK, VDSO_PRELINK);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/cpu/common-xen.c linux-2.6.18-xen/arch/i386/kernel/cpu/common-xen.c
---- linux-2.6.18.3/arch/i386/kernel/cpu/common-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/cpu/common-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,739 @@
+ obj-$(CONFIG_MTRR) += mtrr/
+ obj-$(CONFIG_CPU_FREQ) += cpufreq/
++
++ifdef CONFIG_XEN
++include $(srctree)/scripts/Makefile.xen
++obj-y := $(call cherrypickxen, $(obj-y), $(src))
++endif
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/cpu/mtrr/main-xen.c b/arch/i386/kernel/cpu/mtrr/main-xen.c
+--- a/arch/i386/kernel/cpu/mtrr/main-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/cpu/mtrr/main-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,197 @@
+#include <linux/init.h>
-+#include <linux/string.h>
-+#include <linux/delay.h>
-+#include <linux/smp.h>
++#include <linux/proc_fs.h>
++#include <linux/ctype.h>
+#include <linux/module.h>
-+#include <linux/percpu.h>
-+#include <linux/bootmem.h>
-+#include <asm/semaphore.h>
-+#include <asm/processor.h>
-+#include <asm/i387.h>
-+#include <asm/msr.h>
-+#include <asm/io.h>
-+#include <asm/mmu_context.h>
-+#include <asm/mtrr.h>
-+#include <asm/mce.h>
-+#ifdef CONFIG_X86_LOCAL_APIC
-+#include <asm/mpspec.h>
-+#include <asm/apic.h>
-+#include <mach_apic.h>
-+#endif
-+#include <asm/hypervisor.h>
-+
-+#include "cpu.h"
-+
-+DEFINE_PER_CPU(struct Xgt_desc_struct, cpu_gdt_descr);
-+EXPORT_PER_CPU_SYMBOL(cpu_gdt_descr);
++#include <linux/seq_file.h>
++#include <linux/mutex.h>
++#include <asm/uaccess.h>
+
-+#ifndef CONFIG_XEN
-+DEFINE_PER_CPU(unsigned char, cpu_16bit_stack[CPU_16BIT_STACK_SIZE]);
-+EXPORT_PER_CPU_SYMBOL(cpu_16bit_stack);
-+#endif
++#include <asm/mtrr.h>
++#include "mtrr.h"
+
-+static int cachesize_override __cpuinitdata = -1;
-+static int disable_x86_fxsr __cpuinitdata;
-+static int disable_x86_serial_nr __cpuinitdata = 1;
-+static int disable_x86_sep __cpuinitdata;
++static DEFINE_MUTEX(mtrr_mutex);
+
-+struct cpu_dev * cpu_devs[X86_VENDOR_NUM] = {};
++static void generic_get_mtrr(unsigned int reg, unsigned long *base,
++ unsigned long *size, mtrr_type * type)
++{
++ dom0_op_t op;
+
-+extern int disable_pse;
++ op.cmd = DOM0_READ_MEMTYPE;
++ op.u.read_memtype.reg = reg;
++ (void)HYPERVISOR_dom0_op(&op);
+
-+static void default_init(struct cpuinfo_x86 * c)
-+{
-+ /* Not much we can do here... */
-+ /* Check if at least it has cpuid */
-+ if (c->cpuid_level == -1) {
-+ /* No cpuid. It must be an ancient CPU */
-+ if (c->x86 == 4)
-+ strcpy(c->x86_model_id, "486");
-+ else if (c->x86 == 3)
-+ strcpy(c->x86_model_id, "386");
-+ }
++ *size = op.u.read_memtype.nr_mfns;
++ *base = op.u.read_memtype.mfn;
++ *type = op.u.read_memtype.type;
+}
+
-+static struct cpu_dev default_cpu = {
-+ .c_init = default_init,
-+ .c_vendor = "Unknown",
++struct mtrr_ops generic_mtrr_ops = {
++ .use_intel_if = 1,
++ .get = generic_get_mtrr,
+};
-+static struct cpu_dev * this_cpu = &default_cpu;
+
-+static int __init cachesize_setup(char *str)
-+{
-+ get_option (&str, &cachesize_override);
-+ return 1;
-+}
-+__setup("cachesize=", cachesize_setup);
++struct mtrr_ops *mtrr_if = &generic_mtrr_ops;
++unsigned int num_var_ranges;
++unsigned int *usage_table;
+
-+int __cpuinit get_model_name(struct cpuinfo_x86 *c)
++/* This function returns the number of variable MTRRs */
++static void __init set_num_var_ranges(void)
+{
-+ unsigned int *v;
-+ char *p, *q;
-+
-+ if (cpuid_eax(0x80000000) < 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;
++ dom0_op_t op;
+
-+ /* Intel chips right-justify this string for some dumb reason;
-+ undo that brain damage */
-+ p = q = &c->x86_model_id[0];
-+ while ( *p == ' ' )
-+ p++;
-+ if ( p != q ) {
-+ while ( *p )
-+ *q++ = *p++;
-+ while ( q <= &c->x86_model_id[48] )
-+ *q++ = '\0'; /* Zero-pad the rest */
++ for (num_var_ranges = 0; ; num_var_ranges++) {
++ op.cmd = DOM0_READ_MEMTYPE;
++ op.u.read_memtype.reg = num_var_ranges;
++ if (HYPERVISOR_dom0_op(&op) != 0)
++ break;
+ }
-+
-+ return 1;
+}
+
-+
-+void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c)
++static void __init init_table(void)
+{
-+ unsigned int n, dummy, ecx, edx, l2size;
-+
-+ n = cpuid_eax(0x80000000);
++ int i, max;
+
-+ if (n >= 0x80000005) {
-+ cpuid(0x80000005, &dummy, &dummy, &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);
++ max = num_var_ranges;
++ if ((usage_table = kmalloc(max * sizeof *usage_table, GFP_KERNEL))
++ == NULL) {
++ printk(KERN_ERR "mtrr: could not allocate\n");
++ return;
+ }
++ for (i = 0; i < max; i++)
++ usage_table[i] = 0;
++}
+
-+ if (n < 0x80000006) /* Some chips just has a large L1. */
-+ return;
++int mtrr_add_page(unsigned long base, unsigned long size,
++ unsigned int type, char increment)
++{
++ int error;
++ dom0_op_t op;
+
-+ ecx = cpuid_ecx(0x80000006);
-+ l2size = ecx >> 16;
-+
-+ /* do processor-specific cache resizing */
-+ if (this_cpu->c_size_cache)
-+ l2size = this_cpu->c_size_cache(c,l2size);
++ mutex_lock(&mtrr_mutex);
+
-+ /* Allow user to override all this if necessary. */
-+ if (cachesize_override != -1)
-+ l2size = cachesize_override;
++ op.cmd = DOM0_ADD_MEMTYPE;
++ op.u.add_memtype.mfn = base;
++ op.u.add_memtype.nr_mfns = size;
++ op.u.add_memtype.type = type;
++ error = HYPERVISOR_dom0_op(&op);
++ if (error) {
++ mutex_unlock(&mtrr_mutex);
++ BUG_ON(error > 0);
++ return error;
++ }
+
-+ if ( l2size == 0 )
-+ return; /* Again, no L2 cache is possible */
++ if (increment)
++ ++usage_table[op.u.add_memtype.reg];
+
-+ c->x86_cache_size = l2size;
++ mutex_unlock(&mtrr_mutex);
+
-+ printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n",
-+ l2size, ecx & 0xFF);
++ return op.u.add_memtype.reg;
+}
+
-+/* Naming convention should be: <Name> [(<Codename>)] */
-+/* This table only is used unless init_<vendor>() below doesn't set it; */
-+/* in particular, if CPUID levels 0x80000002..4 are supported, this isn't used */
-+
-+/* Look up CPU names by table lookup. */
-+static char __cpuinit *table_lookup_model(struct cpuinfo_x86 *c)
++static int mtrr_check(unsigned long base, unsigned long size)
+{
-+ struct cpu_model_info *info;
-+
-+ if ( c->x86_model >= 16 )
-+ return NULL; /* Range check */
-+
-+ if (!this_cpu)
-+ return NULL;
-+
-+ info = this_cpu->c_models;
-+
-+ while (info && info->family) {
-+ if (info->family == c->x86)
-+ return info->model_names[c->x86_model];
-+ info++;
++ if ((base & (PAGE_SIZE - 1)) || (size & (PAGE_SIZE - 1))) {
++ printk(KERN_WARNING
++ "mtrr: size and base must be multiples of 4 kiB\n");
++ printk(KERN_DEBUG
++ "mtrr: size: 0x%lx base: 0x%lx\n", size, base);
++ dump_stack();
++ return -1;
+ }
-+ return NULL; /* Not found */
++ return 0;
+}
+
++int
++mtrr_add(unsigned long base, unsigned long size, unsigned int type,
++ char increment)
++{
++ if (mtrr_check(base, size))
++ return -EINVAL;
++ return mtrr_add_page(base >> PAGE_SHIFT, size >> PAGE_SHIFT, type,
++ increment);
++}
+
-+static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c, int early)
++int mtrr_del_page(int reg, unsigned long base, unsigned long size)
+{
-+ char *v = c->x86_vendor_id;
-+ int i;
-+ static int printed;
++ unsigned i;
++ mtrr_type ltype;
++ unsigned long lbase, lsize;
++ int error = -EINVAL;
++ dom0_op_t op;
+
-+ for (i = 0; i < X86_VENDOR_NUM; i++) {
-+ if (cpu_devs[i]) {
-+ if (!strcmp(v,cpu_devs[i]->c_ident[0]) ||
-+ (cpu_devs[i]->c_ident[1] &&
-+ !strcmp(v,cpu_devs[i]->c_ident[1]))) {
-+ c->x86_vendor = i;
-+ if (!early)
-+ this_cpu = cpu_devs[i];
-+ return;
++ mutex_lock(&mtrr_mutex);
++
++ if (reg < 0) {
++ /* Search for existing MTRR */
++ for (i = 0; i < num_var_ranges; ++i) {
++ mtrr_if->get(i, &lbase, &lsize, <ype);
++ if (lbase == base && lsize == size) {
++ reg = i;
++ break;
+ }
+ }
++ if (reg < 0) {
++ printk(KERN_DEBUG "mtrr: no MTRR for %lx000,%lx000 found\n", base,
++ size);
++ goto out;
++ }
+ }
-+ if (!printed) {
-+ printed++;
-+ printk(KERN_ERR "CPU: Vendor unknown, using generic init.\n");
-+ printk(KERN_ERR "CPU: Your system may be unstable.\n");
++ if (usage_table[reg] < 1) {
++ printk(KERN_WARNING "mtrr: reg: %d has count=0\n", reg);
++ goto out;
+ }
-+ c->x86_vendor = X86_VENDOR_UNKNOWN;
-+ this_cpu = &default_cpu;
++ if (--usage_table[reg] < 1) {
++ op.cmd = DOM0_DEL_MEMTYPE;
++ op.u.del_memtype.handle = 0;
++ op.u.del_memtype.reg = reg;
++ error = HYPERVISOR_dom0_op(&op);
++ if (error) {
++ BUG_ON(error > 0);
++ goto out;
++ }
++ }
++ error = reg;
++ out:
++ mutex_unlock(&mtrr_mutex);
++ return error;
+}
+
-+
-+static int __init x86_fxsr_setup(char * s)
++int
++mtrr_del(int reg, unsigned long base, unsigned long size)
+{
-+ disable_x86_fxsr = 1;
-+ return 1;
-+}
-+__setup("nofxsr", x86_fxsr_setup);
-+
-+
-+static int __init x86_sep_setup(char * s)
-+{
-+ disable_x86_sep = 1;
-+ return 1;
++ if (mtrr_check(base, size))
++ return -EINVAL;
++ return mtrr_del_page(reg, base >> PAGE_SHIFT, size >> PAGE_SHIFT);
+}
-+__setup("nosep", x86_sep_setup);
+
++EXPORT_SYMBOL(mtrr_add);
++EXPORT_SYMBOL(mtrr_del);
+
-+/* Standard macro to see if a specific flag is changeable */
-+static inline int flag_is_changeable_p(u32 flag)
++void __init mtrr_bp_init(void)
+{
-+ u32 f1, f2;
-+
-+ asm("pushfl\n\t"
-+ "pushfl\n\t"
-+ "popl %0\n\t"
-+ "movl %0,%1\n\t"
-+ "xorl %2,%0\n\t"
-+ "pushl %0\n\t"
-+ "popfl\n\t"
-+ "pushfl\n\t"
-+ "popl %0\n\t"
-+ "popfl\n\t"
-+ : "=&r" (f1), "=&r" (f2)
-+ : "ir" (flag));
-+
-+ return ((f1^f2) & flag) != 0;
+}
+
-+
-+/* Probe for the CPUID instruction */
-+static int __cpuinit have_cpuid_p(void)
++void mtrr_ap_init(void)
+{
-+ return flag_is_changeable_p(X86_EFLAGS_ID);
+}
+
-+/* Do minimum CPU detection early.
-+ Fields really needed: vendor, cpuid_level, family, model, mask, cache alignment.
-+ The others are not touched to avoid unwanted side effects.
-+
-+ WARNING: this function is only called on the BP. Don't add code here
-+ that is supposed to run on all CPUs. */
-+static void __init early_cpu_detect(void)
++static int __init mtrr_init(void)
+{
+ struct cpuinfo_x86 *c = &boot_cpu_data;
+
-+ c->x86_cache_alignment = 32;
-+
-+ if (!have_cpuid_p())
-+ return;
++ if (!is_initial_xendomain())
++ return -ENODEV;
+
-+ /* Get vendor name */
-+ cpuid(0x00000000, &c->cpuid_level,
-+ (int *)&c->x86_vendor_id[0],
-+ (int *)&c->x86_vendor_id[8],
-+ (int *)&c->x86_vendor_id[4]);
++ if ((!cpu_has(c, X86_FEATURE_MTRR)) &&
++ (!cpu_has(c, X86_FEATURE_K6_MTRR)) &&
++ (!cpu_has(c, X86_FEATURE_CYRIX_ARR)) &&
++ (!cpu_has(c, X86_FEATURE_CENTAUR_MCR)))
++ return -ENODEV;
+
-+ get_cpu_vendor(c, 1);
++ set_num_var_ranges();
++ init_table();
+
-+ c->x86 = 4;
-+ if (c->cpuid_level >= 0x00000001) {
-+ u32 junk, tfms, cap0, misc;
-+ cpuid(0x00000001, &tfms, &misc, &junk, &cap0);
-+ c->x86 = (tfms >> 8) & 15;
-+ c->x86_model = (tfms >> 4) & 15;
-+ if (c->x86 == 0xf)
-+ c->x86 += (tfms >> 20) & 0xff;
-+ if (c->x86 >= 0x6)
-+ c->x86_model += ((tfms >> 16) & 0xF) << 4;
-+ c->x86_mask = tfms & 15;
-+ if (cap0 & (1<<19))
-+ c->x86_cache_alignment = ((misc >> 8) & 0xff) * 8;
-+ }
++ return 0;
+}
+
-+void __cpuinit generic_identify(struct cpuinfo_x86 * c)
-+{
-+ u32 tfms, xlvl;
-+ int ebx;
++subsys_initcall(mtrr_init);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/cpu/mtrr/Makefile b/arch/i386/kernel/cpu/mtrr/Makefile
+--- a/arch/i386/kernel/cpu/mtrr/Makefile 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/cpu/mtrr/Makefile 2007-03-14 10:55:14.000000000 +0100
+@@ -1,3 +1,10 @@
+ obj-y := main.o if.o generic.o state.o
+ obj-$(CONFIG_X86_32) += amd.o cyrix.o centaur.o
+
++ifdef CONFIG_XEN
++include $(srctree)/scripts/Makefile.xen
++n-obj-xen := generic.o state.o amd.o cyrix.o centaur.o
+
-+ if (have_cpuid_p()) {
-+ /* Get vendor name */
-+ cpuid(0x00000000, &c->cpuid_level,
-+ (int *)&c->x86_vendor_id[0],
-+ (int *)&c->x86_vendor_id[8],
-+ (int *)&c->x86_vendor_id[4]);
-+
-+ get_cpu_vendor(c, 0);
-+ /* 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 capability, excap;
-+ cpuid(0x00000001, &tfms, &ebx, &excap, &capability);
-+ c->x86_capability[0] = capability;
-+ c->x86_capability[4] = excap;
-+ c->x86 = (tfms >> 8) & 15;
-+ c->x86_model = (tfms >> 4) & 15;
-+ if (c->x86 == 0xf)
-+ c->x86 += (tfms >> 20) & 0xff;
-+ if (c->x86 >= 0x6)
-+ c->x86_model += ((tfms >> 16) & 0xF) << 4;
-+ c->x86_mask = tfms & 15;
-+#ifdef CONFIG_X86_HT
-+ c->apicid = phys_pkg_id((ebx >> 24) & 0xFF, 0);
-+#else
-+ c->apicid = (ebx >> 24) & 0xFF;
++obj-y := $(call filterxen, $(obj-y), $(n-obj-xen))
++obj-y := $(call cherrypickxen, $(obj-y))
++endif
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/e820.c b/arch/i386/kernel/e820.c
+--- a/arch/i386/kernel/e820.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/e820.c 2007-03-14 10:55:14.000000000 +0100
+@@ -15,11 +15,25 @@
+ #include <asm/page.h>
+ #include <asm/e820.h>
+
++#ifdef CONFIG_XEN
++#include <asm/hypervisor.h>
++#include <xen/interface/physdev.h>
++#include <xen/interface/memory.h>
++#include <xen/features.h>
++#include <xen/xencons.h>
+#endif
-+ } else {
-+ /* Have CPUID level 0 only - unheard of */
-+ c->x86 = 4;
-+ }
-+
-+ /* AMD-defined flags: level 0x80000001 */
-+ xlvl = cpuid_eax(0x80000000);
-+ 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 */
-+ }
-+ }
-+
-+ early_intel_workaround(c);
+
-+#ifdef CONFIG_X86_HT
-+ c->phys_proc_id = (cpuid_ebx(1) >> 24) & 0xff;
+ #ifdef CONFIG_EFI
+ int efi_enabled = 0;
+ EXPORT_SYMBOL(efi_enabled);
+ #endif
+
++#ifdef CONFIG_XEN
++struct e820map machine_e820;
+#endif
-+}
++static void __init
++e820_setup_gap(struct e820entry *e820, int nr_map);
+
-+static void __cpuinit squash_the_stupid_serial_number(struct cpuinfo_x86 *c)
-+{
-+ if (cpu_has(c, X86_FEATURE_PN) && disable_x86_serial_nr ) {
-+ /* Disable processor serial number */
-+ unsigned long lo,hi;
-+ rdmsr(MSR_IA32_BBL_CR_CTL,lo,hi);
-+ lo |= 0x200000;
-+ wrmsr(MSR_IA32_BBL_CR_CTL,lo,hi);
-+ printk(KERN_NOTICE "CPU serial number disabled.\n");
-+ clear_bit(X86_FEATURE_PN, c->x86_capability);
+ struct e820map e820;
+ struct change_member {
+ struct e820entry *pbios; /* pointer to original bios entry */
+@@ -180,6 +194,12 @@
+ unsigned char *rom;
+ int i;
+
++#ifdef CONFIG_XEN
++ /* Nothing to do if not running in dom0. */
++ if (!is_initial_xendomain())
++ return;
++#endif
+
-+ /* Disabling the serial number may affect the cpuid level */
-+ c->cpuid_level = cpuid_eax(0);
-+ }
-+}
+ /* video rom */
+ upper = adapter_rom_resources[0].start;
+ for (start = video_rom_resource.start; start < upper; start += 2048) {
+@@ -247,36 +267,54 @@
+ legacy_init_iomem_resources(struct resource *code_resource, struct resource *data_resource)
+ {
+ int i;
++ struct e820entry *map = e820.map;
++ int nr_map = e820.nr_map;
++#ifdef CONFIG_XEN_PRIVILEGED_GUEST
++ struct xen_memory_map memmap;
+
-+static int __init x86_serial_nr_setup(char *s)
-+{
-+ disable_x86_serial_nr = 0;
-+ return 1;
-+}
-+__setup("serialnumber", x86_serial_nr_setup);
++ map = machine_e820.map;
++ memmap.nr_entries = E820MAX;
+
++ set_xen_guest_handle(memmap.buffer, map);
+
++ if(HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap))
++ BUG();
++ machine_e820.nr_map = memmap.nr_entries;
++ nr_map = memmap.nr_entries;
++ e820_setup_gap(map, memmap.nr_entries);
++#endif
+
+ probe_roms();
+- for (i = 0; i < e820.nr_map; i++) {
++ for (i = 0; i < nr_map; i++) {
+ struct resource *res;
+ #ifndef CONFIG_RESOURCES_64BIT
+- if (e820.map[i].addr + e820.map[i].size > 0x100000000ULL)
++ if (map[i].addr + map[i].size > 0x100000000ULL)
+ continue;
+ #endif
+ res = kzalloc(sizeof(struct resource), GFP_ATOMIC);
+- switch (e820.map[i].type) {
++ switch (map[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.map[i].addr;
+- res->end = res->start + e820.map[i].size - 1;
++ res->start = map[i].addr;
++ res->end = res->start + map[i].size - 1;
+ res->flags = IORESOURCE_MEM | IORESOURCE_BUSY;
+ if (request_resource(&iomem_resource, res)) {
+ kfree(res);
+ continue;
+ }
+- if (e820.map[i].type == E820_RAM) {
++ if (map[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
+ request_resource(res, &crashk_res);
+ #endif
+@@ -295,6 +333,11 @@
+ int i;
+
+ printk("Setting up standard PCI resources\n");
++#ifdef CONFIG_XEN
++ /* Nothing to do if not running in dom0. */
++ if (!is_initial_xendomain())
++ return 0;
++#endif
+ if (efi_enabled)
+ efi_initialize_iomem_resources(&code_resource, &data_resource);
+ else
+@@ -514,10 +557,13 @@
+ */
+ 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 long start = biosmap->addr;
+ unsigned long long size = biosmap->size;
+@@ -529,6 +575,7 @@
+ if (start > end)
+ return -1;
+
++#ifndef CONFIG_XEN
+ /*
+ * Some BIOSes claim RAM in the 640k - 1M region.
+ * Not right. Fix it up.
+@@ -549,6 +596,7 @@
+ size = end - start;
+ }
+ }
++#endif
+ add_memory_region(start, size, type);
+ } while (biosmap++,--nr_map);
+ return 0;
+@@ -653,6 +701,15 @@
+ */
+ last_pfn = PFN_DOWN(e820.map[i].addr + e820.map[i].size);
+
++#ifdef CONFIG_XEN
++ /*
++ * Truncate to the number of actual pages currently
++ * present.
++ */
++ if (last_pfn > xen_start_info->nr_pages)
++ last_pfn = xen_start_info->nr_pages;
++#endif
+
+ if (last_pfn > max_low_pfn)
+ last_pfn = max_low_pfn;
+
+@@ -668,7 +725,12 @@
+ }
+ }
+
+-void __init e820_register_memory(void)
+/*
-+ * This does the hard work of actually picking apart the CPU stuff...
++ * Locate a unused range of the physical address space below 4G which
++ * can be used for PCI mappings.
+ */
-+void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
++static void __init
++e820_setup_gap(struct e820entry *e820, int nr_map)
+ {
+ unsigned long gapstart, gapsize, round;
+ unsigned long long last;
+@@ -681,10 +743,10 @@
+ last = 0x100000000ull;
+ gapstart = 0x10000000;
+ gapsize = 0x400000;
+- i = e820.nr_map;
++ i = nr_map;
+ while (--i >= 0) {
+- unsigned long long start = e820.map[i].addr;
+- unsigned long long end = start + e820.map[i].size;
++ 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
+@@ -716,6 +778,13 @@
+ pci_mem_start, gapstart, gapsize);
+ }
+
++void __init e820_register_memory(void)
+{
-+ int i;
-+
-+ c->loops_per_jiffy = loops_per_jiffy;
-+ c->x86_cache_size = -1;
-+ c->x86_vendor = X86_VENDOR_UNKNOWN;
-+ c->cpuid_level = -1; /* CPUID not detected */
-+ 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_max_cores = 1;
-+ memset(&c->x86_capability, 0, sizeof c->x86_capability);
++#ifndef CONFIG_XEN
++ e820_setup_gap(e820.map, e820.nr_map);
++#endif
++}
+
-+ if (!have_cpuid_p()) {
-+ /* First of all, decide if this is a 486 or higher */
-+ /* It's a 486 if we can modify the AC flag */
-+ if ( flag_is_changeable_p(X86_EFLAGS_AC) )
-+ c->x86 = 4;
+ void __init print_memory_map(char *who)
+ {
+ int i;
+@@ -784,7 +853,7 @@
+
+ void __init limit_regions(unsigned long long size)
+ {
+- unsigned long long current_addr;
++ unsigned long long current_addr = 0;
+ int i;
+
+ print_memory_map("limit_regions start");
+@@ -813,6 +882,19 @@
+ print_memory_map("limit_regions endfor");
+ return;
+ }
++#ifdef CONFIG_XEN
++ if (i==e820.nr_map && current_addr < size) {
++ /*
++ * The e820 map finished before our requested size so
++ * extend the final entry to the requested address.
++ */
++ --i;
++ if (e820.map[i].type == E820_RAM)
++ e820.map[i].size -= current_addr - size;
+ else
-+ c->x86 = 3;
++ add_memory_region(current_addr, size - current_addr, E820_RAM);
+ }
++#endif
+ print_memory_map("limit_regions endfunc");
+ }
+
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/entry.S b/arch/i386/kernel/entry.S
+--- a/arch/i386/kernel/entry.S 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/entry.S 2007-03-14 10:55:14.000000000 +0100
+@@ -284,7 +284,7 @@
+ CFI_SIGNAL_FRAME
+ 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
+@@ -727,7 +727,7 @@
+ * 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.
+ *
+@@ -739,7 +739,7 @@
+ cmpw $__KERNEL_CS,4(%esp); \
+ jne ok; \
+ label: \
+- movl TSS_sysenter_esp0+offset(%esp),%esp; \
++ movl SYSENTER_stack_esp0+offset(%esp),%esp; \
+ CFI_DEF_CFA esp, 0; \
+ CFI_UNDEFINED eip; \
+ pushfl; \
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/entry-xen.S b/arch/i386/kernel/entry-xen.S
+--- a/arch/i386/kernel/entry-xen.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/entry-xen.S 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,1254 @@
++/*
++ * linux/arch/i386/entry.S
++ *
++ * Copyright (C) 1991, 1992 Linus Torvalds
++ */
+
-+ generic_identify(c);
-+
-+ printk(KERN_DEBUG "CPU: After generic identify, caps:");
-+ for (i = 0; i < NCAPINTS; i++)
-+ printk(" %08lx", c->x86_capability[i]);
-+ printk("\n");
++/*
++ * entry.S contains the system-call and fault low-level handling routines.
++ * This also contains the timer-interrupt handler, as well as all interrupts
++ * and faults that can result in a task-switch.
++ *
++ * NOTE: This code handles signal-recognition, which happens every time
++ * after a timer-interrupt and after each system call.
++ *
++ * I changed all the .align's to 4 (16 byte alignment), as that's faster
++ * on a 486.
++ *
++ * Stack layout in 'ret_from_system_call':
++ * ptrace needs to have all regs on the stack.
++ * if the order here is changed, it needs to be
++ * updated in fork.c:copy_process, signal.c:do_signal,
++ * ptrace.c and ptrace.h
++ *
++ * 0(%esp) - %ebx
++ * 4(%esp) - %ecx
++ * 8(%esp) - %edx
++ * C(%esp) - %esi
++ * 10(%esp) - %edi
++ * 14(%esp) - %ebp
++ * 18(%esp) - %eax
++ * 1C(%esp) - %ds
++ * 20(%esp) - %es
++ * 24(%esp) - %gs
++ * 28(%esp) - orig_eax
++ * 2C(%esp) - %eip
++ * 30(%esp) - %cs
++ * 34(%esp) - %eflags
++ * 38(%esp) - %oldesp
++ * 3C(%esp) - %oldss
++ *
++ * "current" is in register %ebx during any slow entries.
++ */
+
-+ if (this_cpu->c_identify) {
-+ this_cpu->c_identify(c);
++#include <linux/linkage.h>
++#include <asm/thread_info.h>
++#include <asm/irqflags.h>
++#include <asm/errno.h>
++#include <asm/segment.h>
++#include <asm/smp.h>
++#include <asm/page.h>
++#include <asm/desc.h>
++#include <asm/percpu.h>
++#include <asm/dwarf2.h>
++#include "irq_vectors.h"
++#include <xen/interface/xen.h>
+
-+ printk(KERN_DEBUG "CPU: After vendor identify, caps:");
-+ for (i = 0; i < NCAPINTS; i++)
-+ printk(" %08lx", c->x86_capability[i]);
-+ printk("\n");
-+ }
++/*
++ * We use macros for low-level operations which need to be overridden
++ * for paravirtualization. The following will never clobber any registers:
++ * INTERRUPT_RETURN (aka. "iret")
++ * GET_CR0_INTO_EAX (aka. "movl %cr0, %eax")
++ * ENABLE_INTERRUPTS_SYSEXIT (aka "sti; sysexit").
++ *
++ * For DISABLE_INTERRUPTS/ENABLE_INTERRUPTS (aka "cli"/"sti"), you must
++ * specify what registers can be overwritten (CLBR_NONE, CLBR_EAX/EDX/ECX/ANY).
++ * Allowing a register to be clobbered can shrink the paravirt replacement
++ * enough to patch inline, increasing performance.
++ */
+
-+ /*
-+ * 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!
-+ */
-+ if (this_cpu->c_init)
-+ this_cpu->c_init(c);
++#define nr_syscalls ((syscall_table_size)/4)
+
-+ /* Disable the PN if appropriate */
-+ squash_the_stupid_serial_number(c);
++CF_MASK = 0x00000001
++TF_MASK = 0x00000100
++IF_MASK = 0x00000200
++DF_MASK = 0x00000400
++NT_MASK = 0x00004000
++VM_MASK = 0x00020000
++/* Pseudo-eflags. */
++NMI_MASK = 0x80000000
+
-+ /*
-+ * The vendor-specific functions might have changed features. Now
-+ * we do "generic changes."
-+ */
++#ifdef CONFIG_XEN
++/* Offsets into shared_info_t. */
++#define evtchn_upcall_pending /* 0 */
++#define evtchn_upcall_mask 1
+
-+ /* TSC disabled? */
-+ if ( tsc_disable )
-+ clear_bit(X86_FEATURE_TSC, c->x86_capability);
++#define sizeof_vcpu_shift 6
+
-+ /* FXSR disabled? */
-+ if (disable_x86_fxsr) {
-+ clear_bit(X86_FEATURE_FXSR, c->x86_capability);
-+ clear_bit(X86_FEATURE_XMM, c->x86_capability);
-+ }
++#ifdef CONFIG_SMP
++#define GET_VCPU_INFO movl %gs:PDA_cpu,%esi ; \
++ shl $sizeof_vcpu_shift,%esi ; \
++ addl HYPERVISOR_shared_info,%esi
++#else
++#define GET_VCPU_INFO movl HYPERVISOR_shared_info,%esi
++#endif
+
-+ /* SEP disabled? */
-+ if (disable_x86_sep)
-+ clear_bit(X86_FEATURE_SEP, c->x86_capability);
++#define __DISABLE_INTERRUPTS movb $1,evtchn_upcall_mask(%esi)
++#define __ENABLE_INTERRUPTS movb $0,evtchn_upcall_mask(%esi)
++#define __TEST_PENDING testb $0xFF,evtchn_upcall_pending(%esi)
++#endif
+
-+ if (disable_pse)
-+ clear_bit(X86_FEATURE_PSE, c->x86_capability);
++#ifdef CONFIG_PREEMPT
++#define preempt_stop(clobbers) DISABLE_INTERRUPTS(clobbers); TRACE_IRQS_OFF
++#else
++#define preempt_stop(clobbers)
++#define resume_kernel restore_nocheck
++#endif
+
-+ /* If the model name is still unset, do table lookup. */
-+ if ( !c->x86_model_id[0] ) {
-+ char *p;
-+ p = table_lookup_model(c);
-+ if ( p )
-+ strcpy(c->x86_model_id, p);
-+ else
-+ /* Last resort... */
-+ sprintf(c->x86_model_id, "%02x/%02x",
-+ c->x86, c->x86_model);
-+ }
++.macro TRACE_IRQS_IRET
++#ifdef CONFIG_TRACE_IRQFLAGS
++ testl $IF_MASK,PT_EFLAGS(%esp) # interrupts off?
++ jz 1f
++ TRACE_IRQS_ON
++1:
++#endif
++.endm
+
-+ /* Now the feature flags better reflect actual CPU features! */
++#ifdef CONFIG_VM86
++#define resume_userspace_sig check_userspace
++#else
++#define resume_userspace_sig resume_userspace
++#endif
+
-+ printk(KERN_DEBUG "CPU: After all inits, caps:");
-+ for (i = 0; i < NCAPINTS; i++)
-+ printk(" %08lx", c->x86_capability[i]);
-+ printk("\n");
++#define SAVE_ALL \
++ cld; \
++ pushl %gs; \
++ CFI_ADJUST_CFA_OFFSET 4;\
++ /*CFI_REL_OFFSET gs, 0;*/\
++ pushl %es; \
++ CFI_ADJUST_CFA_OFFSET 4;\
++ /*CFI_REL_OFFSET es, 0;*/\
++ pushl %ds; \
++ CFI_ADJUST_CFA_OFFSET 4;\
++ /*CFI_REL_OFFSET ds, 0;*/\
++ pushl %eax; \
++ CFI_ADJUST_CFA_OFFSET 4;\
++ CFI_REL_OFFSET eax, 0;\
++ pushl %ebp; \
++ CFI_ADJUST_CFA_OFFSET 4;\
++ CFI_REL_OFFSET ebp, 0;\
++ pushl %edi; \
++ CFI_ADJUST_CFA_OFFSET 4;\
++ CFI_REL_OFFSET edi, 0;\
++ pushl %esi; \
++ CFI_ADJUST_CFA_OFFSET 4;\
++ CFI_REL_OFFSET esi, 0;\
++ pushl %edx; \
++ CFI_ADJUST_CFA_OFFSET 4;\
++ CFI_REL_OFFSET edx, 0;\
++ pushl %ecx; \
++ CFI_ADJUST_CFA_OFFSET 4;\
++ CFI_REL_OFFSET ecx, 0;\
++ pushl %ebx; \
++ CFI_ADJUST_CFA_OFFSET 4;\
++ CFI_REL_OFFSET ebx, 0;\
++ movl $(__USER_DS), %edx; \
++ movl %edx, %ds; \
++ movl %edx, %es; \
++ movl $(__KERNEL_PDA), %edx; \
++ movl %edx, %gs
+
-+ /*
-+ * 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];
-+ }
-+
-+ /* Init Machine Check Exception if available. */
-+ mcheck_init(c);
-+
-+ if (c == &boot_cpu_data)
-+ sysenter_setup();
-+ enable_sep_cpu();
-+
-+ if (c == &boot_cpu_data)
-+ mtrr_bp_init();
-+ else
-+ mtrr_ap_init();
-+}
-+
-+#ifdef CONFIG_X86_HT
-+void __cpuinit detect_ht(struct cpuinfo_x86 *c)
-+{
-+ u32 eax, ebx, ecx, edx;
-+ int index_msb, core_bits;
-+
-+ cpuid(1, &eax, &ebx, &ecx, &edx);
-+
-+ if (!cpu_has(c, X86_FEATURE_HT) || cpu_has(c, X86_FEATURE_CMP_LEGACY))
-+ return;
-+
-+ smp_num_siblings = (ebx & 0xff0000) >> 16;
++#define RESTORE_INT_REGS \
++ popl %ebx; \
++ CFI_ADJUST_CFA_OFFSET -4;\
++ CFI_RESTORE ebx;\
++ popl %ecx; \
++ CFI_ADJUST_CFA_OFFSET -4;\
++ CFI_RESTORE ecx;\
++ popl %edx; \
++ CFI_ADJUST_CFA_OFFSET -4;\
++ CFI_RESTORE edx;\
++ popl %esi; \
++ CFI_ADJUST_CFA_OFFSET -4;\
++ CFI_RESTORE esi;\
++ popl %edi; \
++ CFI_ADJUST_CFA_OFFSET -4;\
++ CFI_RESTORE edi;\
++ popl %ebp; \
++ CFI_ADJUST_CFA_OFFSET -4;\
++ CFI_RESTORE ebp;\
++ popl %eax; \
++ CFI_ADJUST_CFA_OFFSET -4;\
++ CFI_RESTORE eax
+
-+ if (smp_num_siblings == 1) {
-+ printk(KERN_INFO "CPU: Hyper-Threading is disabled\n");
-+ } else if (smp_num_siblings > 1 ) {
++#define RESTORE_REGS \
++ RESTORE_INT_REGS; \
++1: popl %ds; \
++ CFI_ADJUST_CFA_OFFSET -4;\
++ /*CFI_RESTORE ds;*/\
++2: popl %es; \
++ CFI_ADJUST_CFA_OFFSET -4;\
++ /*CFI_RESTORE es;*/\
++3: popl %gs; \
++ CFI_ADJUST_CFA_OFFSET -4;\
++ /*CFI_RESTORE gs;*/\
++.pushsection .fixup,"ax"; \
++4: movl $0,(%esp); \
++ jmp 1b; \
++5: movl $0,(%esp); \
++ jmp 2b; \
++6: movl $0,(%esp); \
++ jmp 3b; \
++.section __ex_table,"a";\
++ .align 4; \
++ .long 1b,4b; \
++ .long 2b,5b; \
++ .long 3b,6b; \
++.popsection
+
-+ if (smp_num_siblings > NR_CPUS) {
-+ printk(KERN_WARNING "CPU: Unsupported number of the "
-+ "siblings %d", smp_num_siblings);
-+ smp_num_siblings = 1;
-+ return;
-+ }
++#define RING0_INT_FRAME \
++ CFI_STARTPROC simple;\
++ CFI_SIGNAL_FRAME;\
++ CFI_DEF_CFA esp, 3*4;\
++ /*CFI_OFFSET cs, -2*4;*/\
++ CFI_OFFSET eip, -3*4
+
-+ index_msb = get_count_order(smp_num_siblings);
-+ c->phys_proc_id = phys_pkg_id((ebx >> 24) & 0xFF, index_msb);
++#define RING0_EC_FRAME \
++ CFI_STARTPROC simple;\
++ CFI_SIGNAL_FRAME;\
++ CFI_DEF_CFA esp, 4*4;\
++ /*CFI_OFFSET cs, -2*4;*/\
++ CFI_OFFSET eip, -3*4
+
-+ printk(KERN_INFO "CPU: Physical Processor ID: %d\n",
-+ c->phys_proc_id);
++#define RING0_PTREGS_FRAME \
++ CFI_STARTPROC simple;\
++ CFI_SIGNAL_FRAME;\
++ CFI_DEF_CFA esp, PT_OLDESP-PT_EBX;\
++ /*CFI_OFFSET cs, PT_CS-PT_OLDESP;*/\
++ CFI_OFFSET eip, PT_EIP-PT_OLDESP;\
++ /*CFI_OFFSET es, PT_ES-PT_OLDESP;*/\
++ /*CFI_OFFSET ds, PT_DS-PT_OLDESP;*/\
++ CFI_OFFSET eax, PT_EAX-PT_OLDESP;\
++ CFI_OFFSET ebp, PT_EBP-PT_OLDESP;\
++ CFI_OFFSET edi, PT_EDI-PT_OLDESP;\
++ CFI_OFFSET esi, PT_ESI-PT_OLDESP;\
++ CFI_OFFSET edx, PT_EDX-PT_OLDESP;\
++ CFI_OFFSET ecx, PT_ECX-PT_OLDESP;\
++ CFI_OFFSET ebx, PT_EBX-PT_OLDESP
+
-+ smp_num_siblings = smp_num_siblings / c->x86_max_cores;
++ENTRY(ret_from_fork)
++ CFI_STARTPROC
++ pushl %eax
++ CFI_ADJUST_CFA_OFFSET 4
++ call schedule_tail
++ GET_THREAD_INFO(%ebp)
++ popl %eax
++ CFI_ADJUST_CFA_OFFSET -4
++ pushl $0x0202 # Reset kernel eflags
++ CFI_ADJUST_CFA_OFFSET 4
++ popfl
++ CFI_ADJUST_CFA_OFFSET -4
++ jmp syscall_exit
++ CFI_ENDPROC
+
-+ index_msb = get_count_order(smp_num_siblings) ;
++/*
++ * Return to user mode is not as complex as all this looks,
++ * but we want the default path for a system call return to
++ * go as quickly as possible which is why some of this is
++ * less clear than it otherwise should be.
++ */
+
-+ core_bits = get_count_order(c->x86_max_cores);
++ # userspace resumption stub bypassing syscall exit tracing
++ ALIGN
++ RING0_PTREGS_FRAME
++ret_from_exception:
++ preempt_stop(CLBR_ANY)
++ret_from_intr:
++ GET_THREAD_INFO(%ebp)
++check_userspace:
++ movl PT_EFLAGS(%esp), %eax # mix EFLAGS and CS
++ movb PT_CS(%esp), %al
++ andl $(VM_MASK | SEGMENT_RPL_MASK), %eax
++ cmpl $USER_RPL, %eax
++ jb resume_kernel # not returning to v8086 or userspace
+
-+ c->cpu_core_id = phys_pkg_id((ebx >> 24) & 0xFF, index_msb) &
-+ ((1 << core_bits) - 1);
++ENTRY(resume_userspace)
++ DISABLE_INTERRUPTS(CLBR_ANY) # make sure we don't miss an interrupt
++ # setting need_resched or sigpending
++ # between sampling and the iret
++ movl TI_flags(%ebp), %ecx
++ andl $_TIF_WORK_MASK, %ecx # is there any work to be done on
++ # int/exception return?
++ jne work_pending
++ jmp restore_all
+
-+ if (c->x86_max_cores > 1)
-+ printk(KERN_INFO "CPU: Processor Core ID: %d\n",
-+ c->cpu_core_id);
-+ }
-+}
++#ifdef CONFIG_PREEMPT
++ENTRY(resume_kernel)
++ DISABLE_INTERRUPTS(CLBR_ANY)
++ cmpl $0,TI_preempt_count(%ebp) # non-zero preempt_count ?
++ jnz restore_nocheck
++need_resched:
++ movl TI_flags(%ebp), %ecx # need_resched set ?
++ testb $_TIF_NEED_RESCHED, %cl
++ jz restore_all
++ testl $IF_MASK,PT_EFLAGS(%esp) # interrupts off (exception path) ?
++ jz restore_all
++ call preempt_schedule_irq
++ jmp need_resched
+#endif
++ CFI_ENDPROC
+
-+void __cpuinit print_cpu_info(struct cpuinfo_x86 *c)
-+{
-+ char *vendor = NULL;
-+
-+ if (c->x86_vendor < X86_VENDOR_NUM)
-+ vendor = this_cpu->c_vendor;
-+ else if (c->cpuid_level >= 0)
-+ vendor = c->x86_vendor_id;
-+
-+ if (vendor && strncmp(c->x86_model_id, vendor, strlen(vendor)))
-+ printk("%s ", vendor);
-+
-+ if (!c->x86_model_id[0])
-+ printk("%d86", c->x86);
-+ else
-+ printk("%s", c->x86_model_id);
-+
-+ if (c->x86_mask || c->cpuid_level >= 0)
-+ printk(" stepping %02x\n", c->x86_mask);
-+ else
-+ printk("\n");
-+}
-+
-+cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE;
-+
-+/* This is hacky. :)
-+ * We're emulating future behavior.
-+ * In the future, the cpu-specific init functions will be called implicitly
-+ * via the magic of initcalls.
-+ * They will insert themselves into the cpu_devs structure.
-+ * Then, when cpu_init() is called, we can just iterate over that array.
-+ */
-+
-+extern int intel_cpu_init(void);
-+extern int cyrix_init_cpu(void);
-+extern int nsc_init_cpu(void);
-+extern int amd_init_cpu(void);
-+extern int centaur_init_cpu(void);
-+extern int transmeta_init_cpu(void);
-+extern int rise_init_cpu(void);
-+extern int nexgen_init_cpu(void);
-+extern int umc_init_cpu(void);
-+
-+void __init early_cpu_init(void)
-+{
-+ intel_cpu_init();
-+ cyrix_init_cpu();
-+ nsc_init_cpu();
-+ amd_init_cpu();
-+ centaur_init_cpu();
-+ transmeta_init_cpu();
-+ rise_init_cpu();
-+ nexgen_init_cpu();
-+ umc_init_cpu();
-+ early_cpu_detect();
++/* SYSENTER_RETURN points to after the "sysenter" instruction in
++ the vsyscall page. See vsyscall-sysentry.S, which defines the symbol. */
+
-+#ifdef CONFIG_DEBUG_PAGEALLOC
-+ /* pse is not compatible with on-the-fly unmapping,
-+ * disable it even if the cpus claim to support it.
++ # sysenter call handler stub
++ENTRY(sysenter_entry)
++ CFI_STARTPROC simple
++ CFI_SIGNAL_FRAME
++ CFI_DEF_CFA esp, 0
++ CFI_REGISTER esp, ebp
++ movl SYSENTER_stack_esp0(%esp),%esp
++sysenter_past_esp:
++ /*
++ * No need to follow this irqs on/off section: the syscall
++ * disabled irqs and here we enable it straight after entry:
+ */
-+ clear_bit(X86_FEATURE_PSE, boot_cpu_data.x86_capability);
-+ disable_pse = 1;
++ ENABLE_INTERRUPTS(CLBR_NONE)
++ pushl $(__USER_DS)
++ CFI_ADJUST_CFA_OFFSET 4
++ /*CFI_REL_OFFSET ss, 0*/
++ pushl %ebp
++ CFI_ADJUST_CFA_OFFSET 4
++ CFI_REL_OFFSET esp, 0
++ pushfl
++ CFI_ADJUST_CFA_OFFSET 4
++ pushl $(__USER_CS)
++ CFI_ADJUST_CFA_OFFSET 4
++ /*CFI_REL_OFFSET cs, 0*/
++#ifndef CONFIG_COMPAT_VDSO
++ /*
++ * Push current_thread_info()->sysenter_return to the stack.
++ * A tiny bit of offset fixup is necessary - 4*4 means the 4 words
++ * pushed above; +8 corresponds to copy_thread's esp0 setting.
++ */
++ pushl (TI_sysenter_return-THREAD_SIZE+8+4*4)(%esp)
++#else
++ pushl $SYSENTER_RETURN
+#endif
-+}
-+
-+void __cpuinit cpu_gdt_init(struct Xgt_desc_struct *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_lowmem_page_readonly(
-+ (void *)va, XENFEAT_writable_descriptor_tables);
-+ }
-+ if (HYPERVISOR_set_gdt(frames, gdt_descr->size / 8))
-+ BUG();
-+}
++ CFI_ADJUST_CFA_OFFSET 4
++ CFI_REL_OFFSET eip, 0
+
+/*
-+ * 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.
++ * Load the potential sixth argument from user stack.
++ * Careful about security.
+ */
-+void __cpuinit cpu_init(void)
-+{
-+ int cpu = smp_processor_id();
-+#ifndef CONFIG_X86_NO_TSS
-+ struct tss_struct * t = &per_cpu(init_tss, cpu);
-+#endif
-+ struct thread_struct *thread = ¤t->thread;
-+ struct desc_struct *gdt;
-+ struct Xgt_desc_struct *cpu_gdt_descr = &per_cpu(cpu_gdt_descr, cpu);
-+
-+ if (cpu_test_and_set(cpu, cpu_initialized)) {
-+ printk(KERN_WARNING "CPU#%d already initialized!\n", cpu);
-+ for (;;) local_irq_enable();
-+ }
-+ printk(KERN_INFO "Initializing CPU#%d\n", cpu);
++ cmpl $__PAGE_OFFSET-3,%ebp
++ jae syscall_fault
++1: movl (%ebp),%ebp
++.section __ex_table,"a"
++ .align 4
++ .long 1b,syscall_fault
++.previous
+
-+ if (cpu_has_vme || cpu_has_tsc || cpu_has_de)
-+ clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE);
-+ if (tsc_disable && cpu_has_tsc) {
-+ printk(KERN_NOTICE "Disabling TSC...\n");
-+ /**** FIX-HPA: DOES THIS REALLY BELONG HERE? ****/
-+ clear_bit(X86_FEATURE_TSC, boot_cpu_data.x86_capability);
-+ set_in_cr4(X86_CR4_TSD);
-+ }
++ pushl %eax
++ CFI_ADJUST_CFA_OFFSET 4
++ SAVE_ALL
++ GET_THREAD_INFO(%ebp)
+
-+#ifndef CONFIG_XEN
-+ /* The CPU hotplug case */
-+ if (cpu_gdt_descr->address) {
-+ gdt = (struct desc_struct *)cpu_gdt_descr->address;
-+ memset(gdt, 0, PAGE_SIZE);
-+ goto old_gdt;
-+ }
-+ /*
-+ * This is a horrible hack to allocate the GDT. The problem
-+ * is that cpu_init() is called really early for the boot CPU
-+ * (and hence needs bootmem) but much later for the secondary
-+ * CPUs, when bootmem will have gone away
-+ */
-+ if (NODE_DATA(0)->bdata->node_bootmem_map) {
-+ gdt = (struct desc_struct *)alloc_bootmem_pages(PAGE_SIZE);
-+ /* alloc_bootmem_pages panics on failure, so no check */
-+ memset(gdt, 0, PAGE_SIZE);
-+ } else {
-+ gdt = (struct desc_struct *)get_zeroed_page(GFP_KERNEL);
-+ if (unlikely(!gdt)) {
-+ printk(KERN_CRIT "CPU%d failed to allocate GDT\n", cpu);
-+ for (;;)
-+ local_irq_enable();
-+ }
-+ }
-+old_gdt:
-+ /*
-+ * Initialize the per-CPU GDT with the boot GDT,
-+ * and set up the GDT descriptor:
-+ */
-+ memcpy(gdt, cpu_gdt_table, GDT_SIZE);
++ /* Note, _TIF_SECCOMP is bit number 8, and so it needs testw and not testb */
++ testw $(_TIF_SYSCALL_EMU|_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT),TI_flags(%ebp)
++ jnz syscall_trace_entry
++ cmpl $(nr_syscalls), %eax
++ jae syscall_badsys
++ call *sys_call_table(,%eax,4)
++ movl %eax,PT_EAX(%esp)
++ DISABLE_INTERRUPTS(CLBR_ECX|CLBR_EDX)
++ TRACE_IRQS_OFF
++ movl TI_flags(%ebp), %ecx
++ testw $_TIF_ALLWORK_MASK, %cx
++ jne syscall_exit_work
++/* if something modifies registers it must also disable sysexit */
++ movl PT_EIP(%esp), %edx
++ movl PT_OLDESP(%esp), %ecx
++ xorl %ebp,%ebp
++ TRACE_IRQS_ON
++1: mov PT_GS(%esp), %gs
++#ifdef CONFIG_XEN
++ __ENABLE_INTERRUPTS
++sysexit_scrit: /**** START OF SYSEXIT CRITICAL REGION ****/
++ __TEST_PENDING
++ jnz 14f # process more events if necessary...
++ movl PT_ESI(%esp), %esi
++ sysexit
++14: __DISABLE_INTERRUPTS
++sysexit_ecrit: /**** END OF SYSEXIT CRITICAL REGION ****/
++ push %esp
++ CFI_ADJUST_CFA_OFFSET 4
++ call evtchn_do_upcall
++ add $4,%esp
++ CFI_ADJUST_CFA_OFFSET -4
++ jmp ret_from_intr
++#else
++ ENABLE_INTERRUPTS_SYSEXIT
++#endif /* !CONFIG_XEN */
++ CFI_ENDPROC
++.pushsection .fixup,"ax"
++2: movl $0,PT_GS(%esp)
++ jmp 1b
++.section __ex_table,"a"
++ .align 4
++ .long 1b,2b
++.popsection
+
-+ /* Set up GDT entry for 16bit stack */
-+ *(__u64 *)(&gdt[GDT_ENTRY_ESPFIX_SS]) |=
-+ ((((__u64)stk16_off) << 16) & 0x000000ffffff0000ULL) |
-+ ((((__u64)stk16_off) << 32) & 0xff00000000000000ULL) |
-+ (CPU_16BIT_STACK_SIZE - 1);
++ # system call handler stub
++ENTRY(system_call)
++ RING0_INT_FRAME # can't unwind into user space anyway
++ pushl %eax # save orig_eax
++ CFI_ADJUST_CFA_OFFSET 4
++ SAVE_ALL
++ GET_THREAD_INFO(%ebp)
++ testl $TF_MASK,PT_EFLAGS(%esp)
++ jz no_singlestep
++ orl $_TIF_SINGLESTEP,TI_flags(%ebp)
++no_singlestep:
++ # system call tracing in operation / emulation
++ /* Note, _TIF_SECCOMP is bit number 8, and so it needs testw and not testb */
++ testw $(_TIF_SYSCALL_EMU|_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT),TI_flags(%ebp)
++ jnz syscall_trace_entry
++ cmpl $(nr_syscalls), %eax
++ jae syscall_badsys
++syscall_call:
++ call *sys_call_table(,%eax,4)
++ movl %eax,PT_EAX(%esp) # store the return value
++syscall_exit:
++ DISABLE_INTERRUPTS(CLBR_ANY) # make sure we don't miss an interrupt
++ # setting need_resched or sigpending
++ # between sampling and the iret
++ TRACE_IRQS_OFF
++ movl TI_flags(%ebp), %ecx
++ testw $_TIF_ALLWORK_MASK, %cx # current->work
++ jne syscall_exit_work
+
-+ cpu_gdt_descr->size = GDT_SIZE - 1;
-+ cpu_gdt_descr->address = (unsigned long)gdt;
++restore_all:
++#ifndef CONFIG_XEN
++ movl PT_EFLAGS(%esp), %eax # mix EFLAGS, SS and CS
++ # Warning: PT_OLDSS(%esp) contains the wrong/random values if we
++ # are returning to the kernel.
++ # See comments in process.c:copy_thread() for details.
++ movb PT_OLDSS(%esp), %ah
++ movb PT_CS(%esp), %al
++ andl $(VM_MASK | (SEGMENT_TI_MASK << 8) | SEGMENT_RPL_MASK), %eax
++ cmpl $((SEGMENT_LDT << 8) | USER_RPL), %eax
++ CFI_REMEMBER_STATE
++ je ldt_ss # returning to user-space with LDT SS
++restore_nocheck:
+#else
-+ if (cpu == 0 && cpu_gdt_descr->address == 0) {
-+ gdt = (struct desc_struct *)alloc_bootmem_pages(PAGE_SIZE);
-+ /* alloc_bootmem_pages panics on failure, so no check */
-+ memset(gdt, 0, PAGE_SIZE);
-+
-+ memcpy(gdt, cpu_gdt_table, GDT_SIZE);
-+
-+ cpu_gdt_descr->size = GDT_SIZE;
-+ cpu_gdt_descr->address = (unsigned long)gdt;
-+ }
++restore_nocheck:
++ movl PT_EFLAGS(%esp), %eax
++ testl $(VM_MASK|NMI_MASK), %eax
++ jnz hypervisor_iret
++ shr $9, %eax # EAX[0] == IRET_EFLAGS.IF
++ GET_VCPU_INFO
++ andb evtchn_upcall_mask(%esi),%al
++ andb $1,%al # EAX[0] == IRET_EFLAGS.IF & event_mask
++ jnz restore_all_enable_events # != 0 => enable event delivery
++ CFI_REMEMBER_STATE
++#endif
++ TRACE_IRQS_IRET
++restore_nocheck_notrace:
++ RESTORE_REGS
++ addl $4, %esp # skip orig_eax/error_code
++ CFI_ADJUST_CFA_OFFSET -4
++1: INTERRUPT_RETURN
++.section .fixup,"ax"
++iret_exc:
++#ifndef CONFIG_XEN
++ TRACE_IRQS_ON
++ ENABLE_INTERRUPTS(CLBR_NONE)
+#endif
++ pushl $0 # no error code
++ pushl $do_iret_error
++ jmp error_code
++.previous
++.section __ex_table,"a"
++ .align 4
++ .long 1b,iret_exc
++.previous
+
-+ cpu_gdt_init(cpu_gdt_descr);
++ CFI_RESTORE_STATE
++#ifndef CONFIG_XEN
++ldt_ss:
++ larl PT_OLDSS(%esp), %eax
++ jnz restore_nocheck
++ testl $0x00400000, %eax # returning to 32bit stack?
++ jnz restore_nocheck # allright, normal return
+
++#ifdef CONFIG_PARAVIRT
+ /*
-+ * Set up and load the per-CPU TSS and LDT
++ * The kernel can't run on a non-flat stack if paravirt mode
++ * is active. Rather than try to fixup the high bits of
++ * ESP, bypass this code entirely. This may break DOSemu
++ * and/or Wine support in a paravirt VM, although the option
++ * is still available to implement the setting of the high
++ * 16-bits in the INTERRUPT_RETURN paravirt-op.
+ */
-+ atomic_inc(&init_mm.mm_count);
-+ current->active_mm = &init_mm;
-+ if (current->mm)
-+ BUG();
-+ enter_lazy_tlb(&init_mm, current);
-+
-+ load_esp0(t, thread);
-+
-+ load_LDT(&init_mm.context);
-+
-+#ifdef CONFIG_DOUBLEFAULT
-+ /* Set up doublefault TSS pointer in the GDT */
-+ __set_tss_desc(cpu, GDT_ENTRY_DOUBLEFAULT_TSS, &doublefault_tss);
++ cmpl $0, paravirt_ops+PARAVIRT_enabled
++ jne restore_nocheck
+#endif
+
-+ /* Clear %fs and %gs. */
-+ asm volatile ("xorl %eax, %eax; movl %eax, %fs; movl %eax, %gs");
-+
-+ /* Clear all 6 debug registers: */
-+ set_debugreg(0, 0);
-+ set_debugreg(0, 1);
-+ set_debugreg(0, 2);
-+ set_debugreg(0, 3);
-+ set_debugreg(0, 6);
-+ set_debugreg(0, 7);
++ /* If returning to userspace with 16bit stack,
++ * try to fix the higher word of ESP, as the CPU
++ * won't restore it.
++ * This is an "official" bug of all the x86-compatible
++ * CPUs, which we can try to work around to make
++ * dosemu and wine happy. */
++ movl PT_OLDESP(%esp), %eax
++ movl %esp, %edx
++ call patch_espfix_desc
++ pushl $__ESPFIX_SS
++ CFI_ADJUST_CFA_OFFSET 4
++ pushl %eax
++ CFI_ADJUST_CFA_OFFSET 4
++ DISABLE_INTERRUPTS(CLBR_EAX)
++ TRACE_IRQS_OFF
++ lss (%esp), %esp
++ CFI_ADJUST_CFA_OFFSET -8
++ jmp restore_nocheck
++#else
++hypervisor_iret:
++ andl $~NMI_MASK, PT_EFLAGS(%esp)
++ RESTORE_REGS
++ addl $4, %esp
++ jmp hypercall_page + (__HYPERVISOR_iret * 32)
++#endif
++ CFI_ENDPROC
+
-+ /*
-+ * Force FPU initialization:
-+ */
-+ current_thread_info()->status = 0;
-+ clear_used_math();
-+ mxcsr_feature_mask_init();
-+}
++ # perform work that needs to be done immediately before resumption
++ ALIGN
++ RING0_PTREGS_FRAME # can't unwind into user space anyway
++work_pending:
++ testb $_TIF_NEED_RESCHED, %cl
++ jz work_notifysig
++work_resched:
++ call schedule
++ DISABLE_INTERRUPTS(CLBR_ANY) # make sure we don't miss an interrupt
++ # setting need_resched or sigpending
++ # between sampling and the iret
++ TRACE_IRQS_OFF
++ movl TI_flags(%ebp), %ecx
++ andl $_TIF_WORK_MASK, %ecx # is there any work to be done other
++ # than syscall tracing?
++ jz restore_all
++ testb $_TIF_NEED_RESCHED, %cl
++ jnz work_resched
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+void __cpuinit cpu_uninit(void)
-+{
-+ int cpu = raw_smp_processor_id();
-+ cpu_clear(cpu, cpu_initialized);
++work_notifysig: # deal with pending signals and
++ # notify-resume requests
++#ifdef CONFIG_VM86
++ testl $VM_MASK, PT_EFLAGS(%esp)
++ movl %esp, %eax
++ jne work_notifysig_v86 # returning to kernel-space or
++ # vm86-space
++ xorl %edx, %edx
++ call do_notify_resume
++ jmp resume_userspace_sig
+
-+ /* lazy TLB state */
-+ per_cpu(cpu_tlbstate, cpu).state = 0;
-+ per_cpu(cpu_tlbstate, cpu).active_mm = &init_mm;
-+}
++ ALIGN
++work_notifysig_v86:
++ pushl %ecx # save ti_flags for do_notify_resume
++ CFI_ADJUST_CFA_OFFSET 4
++ call save_v86_state # %eax contains pt_regs pointer
++ popl %ecx
++ CFI_ADJUST_CFA_OFFSET -4
++ movl %eax, %esp
++#else
++ movl %esp, %eax
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/cpu/Makefile linux-2.6.18-xen/arch/i386/kernel/cpu/Makefile
---- linux-2.6.18.3/arch/i386/kernel/cpu/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/kernel/cpu/Makefile 2006-11-19 14:26:21.000000000 +0100
-@@ -17,3 +17,8 @@
-
- obj-$(CONFIG_MTRR) += mtrr/
- obj-$(CONFIG_CPU_FREQ) += cpufreq/
++ xorl %edx, %edx
++ call do_notify_resume
++ jmp resume_userspace_sig
+
-+ifdef CONFIG_XEN
-+include $(srctree)/scripts/Makefile.xen
-+obj-y := $(call cherrypickxen, $(obj-y), $(src))
-+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/cpu/mtrr/main-xen.c linux-2.6.18-xen/arch/i386/kernel/cpu/mtrr/main-xen.c
---- linux-2.6.18.3/arch/i386/kernel/cpu/mtrr/main-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/cpu/mtrr/main-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,197 @@
-+#include <linux/init.h>
-+#include <linux/proc_fs.h>
-+#include <linux/ctype.h>
-+#include <linux/module.h>
-+#include <linux/seq_file.h>
-+#include <linux/mutex.h>
-+#include <asm/uaccess.h>
-+
-+#include <asm/mtrr.h>
-+#include "mtrr.h"
-+
-+static DEFINE_MUTEX(mtrr_mutex);
-+
-+void generic_get_mtrr(unsigned int reg, unsigned long *base,
-+ unsigned int *size, mtrr_type * type)
-+{
-+ dom0_op_t op;
-+
-+ op.cmd = DOM0_READ_MEMTYPE;
-+ op.u.read_memtype.reg = reg;
-+ (void)HYPERVISOR_dom0_op(&op);
-+
-+ *size = op.u.read_memtype.nr_mfns;
-+ *base = op.u.read_memtype.mfn;
-+ *type = op.u.read_memtype.type;
-+}
-+
-+struct mtrr_ops generic_mtrr_ops = {
-+ .use_intel_if = 1,
-+ .get = generic_get_mtrr,
-+};
-+
-+struct mtrr_ops *mtrr_if = &generic_mtrr_ops;
-+unsigned int num_var_ranges;
-+unsigned int *usage_table;
-+
-+static void __init set_num_var_ranges(void)
-+{
-+ dom0_op_t op;
-+
-+ for (num_var_ranges = 0; ; num_var_ranges++) {
-+ op.cmd = DOM0_READ_MEMTYPE;
-+ op.u.read_memtype.reg = num_var_ranges;
-+ if (HYPERVISOR_dom0_op(&op) != 0)
-+ break;
-+ }
-+}
-+
-+static void __init init_table(void)
-+{
-+ int i, max;
-+
-+ max = num_var_ranges;
-+ if ((usage_table = kmalloc(max * sizeof *usage_table, GFP_KERNEL))
-+ == NULL) {
-+ printk(KERN_ERR "mtrr: could not allocate\n");
-+ return;
-+ }
-+ for (i = 0; i < max; i++)
-+ usage_table[i] = 0;
-+}
-+
-+int mtrr_add_page(unsigned long base, unsigned long size,
-+ unsigned int type, char increment)
-+{
-+ int error;
-+ dom0_op_t op;
-+
-+ mutex_lock(&mtrr_mutex);
-+
-+ op.cmd = DOM0_ADD_MEMTYPE;
-+ op.u.add_memtype.mfn = base;
-+ op.u.add_memtype.nr_mfns = size;
-+ op.u.add_memtype.type = type;
-+ error = HYPERVISOR_dom0_op(&op);
-+ if (error) {
-+ mutex_unlock(&mtrr_mutex);
-+ BUG_ON(error > 0);
-+ return error;
-+ }
-+
-+ if (increment)
-+ ++usage_table[op.u.add_memtype.reg];
-+
-+ mutex_unlock(&mtrr_mutex);
-+
-+ return op.u.add_memtype.reg;
-+}
-+
-+static int mtrr_check(unsigned long base, unsigned long size)
-+{
-+ if ((base & (PAGE_SIZE - 1)) || (size & (PAGE_SIZE - 1))) {
-+ printk(KERN_WARNING
-+ "mtrr: size and base must be multiples of 4 kiB\n");
-+ printk(KERN_DEBUG
-+ "mtrr: size: 0x%lx base: 0x%lx\n", size, base);
-+ dump_stack();
-+ return -1;
-+ }
-+ return 0;
-+}
-+
-+int
-+mtrr_add(unsigned long base, unsigned long size, unsigned int type,
-+ char increment)
-+{
-+ if (mtrr_check(base, size))
-+ return -EINVAL;
-+ return mtrr_add_page(base >> PAGE_SHIFT, size >> PAGE_SHIFT, type,
-+ increment);
-+}
-+
-+int mtrr_del_page(int reg, unsigned long base, unsigned long size)
-+{
-+ unsigned i;
-+ mtrr_type ltype;
-+ unsigned long lbase;
-+ unsigned int lsize;
-+ int error = -EINVAL;
-+ dom0_op_t op;
-+
-+ mutex_lock(&mtrr_mutex);
-+
-+ if (reg < 0) {
-+ /* Search for existing MTRR */
-+ for (i = 0; i < num_var_ranges; ++i) {
-+ mtrr_if->get(i, &lbase, &lsize, <ype);
-+ if (lbase == base && lsize == size) {
-+ reg = i;
-+ break;
-+ }
-+ }
-+ if (reg < 0) {
-+ printk(KERN_DEBUG "mtrr: no MTRR for %lx000,%lx000 found\n", base,
-+ size);
-+ goto out;
-+ }
-+ }
-+ if (usage_table[reg] < 1) {
-+ printk(KERN_WARNING "mtrr: reg: %d has count=0\n", reg);
-+ goto out;
-+ }
-+ if (--usage_table[reg] < 1) {
-+ op.cmd = DOM0_DEL_MEMTYPE;
-+ op.u.del_memtype.handle = 0;
-+ op.u.del_memtype.reg = reg;
-+ error = HYPERVISOR_dom0_op(&op);
-+ if (error) {
-+ BUG_ON(error > 0);
-+ goto out;
-+ }
-+ }
-+ error = reg;
-+ out:
-+ mutex_unlock(&mtrr_mutex);
-+ return error;
-+}
-+
-+int
-+mtrr_del(int reg, unsigned long base, unsigned long size)
-+{
-+ if (mtrr_check(base, size))
-+ return -EINVAL;
-+ return mtrr_del_page(reg, base >> PAGE_SHIFT, size >> PAGE_SHIFT);
-+}
-+
-+EXPORT_SYMBOL(mtrr_add);
-+EXPORT_SYMBOL(mtrr_del);
-+
-+void __init mtrr_bp_init(void)
-+{
-+}
-+
-+void mtrr_ap_init(void)
-+{
-+}
-+
-+static int __init mtrr_init(void)
-+{
-+ struct cpuinfo_x86 *c = &boot_cpu_data;
-+
-+ if (!is_initial_xendomain())
-+ return -ENODEV;
-+
-+ if ((!cpu_has(c, X86_FEATURE_MTRR)) &&
-+ (!cpu_has(c, X86_FEATURE_K6_MTRR)) &&
-+ (!cpu_has(c, X86_FEATURE_CYRIX_ARR)) &&
-+ (!cpu_has(c, X86_FEATURE_CENTAUR_MCR)))
-+ return -ENODEV;
++ # perform syscall exit tracing
++ ALIGN
++syscall_trace_entry:
++ movl $-ENOSYS,PT_EAX(%esp)
++ movl %esp, %eax
++ xorl %edx,%edx
++ call do_syscall_trace
++ cmpl $0, %eax
++ jne resume_userspace # ret != 0 -> running under PTRACE_SYSEMU,
++ # so must skip actual syscall
++ movl PT_ORIG_EAX(%esp), %eax
++ cmpl $(nr_syscalls), %eax
++ jnae syscall_call
++ jmp syscall_exit
+
-+ set_num_var_ranges();
-+ init_table();
++ # perform syscall exit tracing
++ ALIGN
++syscall_exit_work:
++ testb $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP), %cl
++ jz work_pending
++ TRACE_IRQS_ON
++ ENABLE_INTERRUPTS(CLBR_ANY) # could let do_syscall_trace() call
++ # schedule() instead
++ movl %esp, %eax
++ movl $1, %edx
++ call do_syscall_trace
++ jmp resume_userspace
++ CFI_ENDPROC
+
-+ return 0;
-+}
++ RING0_INT_FRAME # can't unwind into user space anyway
++syscall_fault:
++ pushl %eax # save orig_eax
++ CFI_ADJUST_CFA_OFFSET 4
++ SAVE_ALL
++ GET_THREAD_INFO(%ebp)
++ movl $-EFAULT,PT_EAX(%esp)
++ jmp resume_userspace
+
-+subsys_initcall(mtrr_init);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/cpu/mtrr/Makefile linux-2.6.18-xen/arch/i386/kernel/cpu/mtrr/Makefile
---- linux-2.6.18.3/arch/i386/kernel/cpu/mtrr/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/kernel/cpu/mtrr/Makefile 2006-11-19 14:26:22.000000000 +0100
-@@ -3,3 +3,10 @@
- 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
++syscall_badsys:
++ movl $-ENOSYS,PT_EAX(%esp)
++ jmp resume_userspace
++ CFI_ENDPROC
+
-+obj-y := $(call filterxen, $(obj-y), $(n-obj-xen))
-+obj-y := $(call cherrypickxen, $(obj-y))
-+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/early_printk-xen.c linux-2.6.18-xen/arch/i386/kernel/early_printk-xen.c
---- linux-2.6.18.3/arch/i386/kernel/early_printk-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/early_printk-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,2 @@
++#ifndef CONFIG_XEN
++#define FIXUP_ESPFIX_STACK \
++ /* since we are on a wrong stack, we cant make it a C code :( */ \
++ movl %gs:PDA_cpu, %ebx; \
++ PER_CPU(cpu_gdt_descr, %ebx); \
++ movl GDS_address(%ebx), %ebx; \
++ GET_DESC_BASE(GDT_ENTRY_ESPFIX_SS, %ebx, %eax, %ax, %al, %ah); \
++ addl %esp, %eax; \
++ pushl $__KERNEL_DS; \
++ CFI_ADJUST_CFA_OFFSET 4; \
++ pushl %eax; \
++ CFI_ADJUST_CFA_OFFSET 4; \
++ lss (%esp), %esp; \
++ CFI_ADJUST_CFA_OFFSET -8;
++#define UNWIND_ESPFIX_STACK \
++ movl %ss, %eax; \
++ /* see if on espfix stack */ \
++ cmpw $__ESPFIX_SS, %ax; \
++ jne 27f; \
++ movl $__KERNEL_DS, %eax; \
++ movl %eax, %ds; \
++ movl %eax, %es; \
++ /* switch to normal stack */ \
++ FIXUP_ESPFIX_STACK; \
++27:;
+
-+#include "../../x86_64/kernel/early_printk-xen.c"
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/entry.S linux-2.6.18-xen/arch/i386/kernel/entry.S
---- linux-2.6.18.3/arch/i386/kernel/entry.S 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/kernel/entry.S 2006-11-19 14:26:22.000000000 +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
-@@ -689,7 +689,7 @@
- * 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.
- *
-@@ -701,7 +701,7 @@
- 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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/entry-xen.S linux-2.6.18-xen/arch/i386/kernel/entry-xen.S
---- linux-2.6.18.3/arch/i386/kernel/entry-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/entry-xen.S 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,1213 @@
+/*
-+ * linux/arch/i386/entry.S
-+ *
-+ * Copyright (C) 1991, 1992 Linus Torvalds
++ * Build the entry stubs and pointer table with
++ * some assembler magic.
+ */
++.data
++ENTRY(interrupt)
++.text
++
++vector=0
++ENTRY(irq_entries_start)
++ RING0_INT_FRAME
++.rept NR_IRQS
++ ALIGN
++ .if vector
++ CFI_ADJUST_CFA_OFFSET -4
++ .endif
++1: pushl $~(vector)
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp common_interrupt
++.data
++ .long 1b
++.text
++vector=vector+1
++.endr
+
+/*
-+ * entry.S contains the system-call and fault low-level handling routines.
-+ * This also contains the timer-interrupt handler, as well as all interrupts
-+ * and faults that can result in a task-switch.
-+ *
-+ * NOTE: This code handles signal-recognition, which happens every time
-+ * after a timer-interrupt and after each system call.
-+ *
-+ * I changed all the .align's to 4 (16 byte alignment), as that's faster
-+ * on a 486.
-+ *
-+ * Stack layout in 'ret_from_system_call':
-+ * ptrace needs to have all regs on the stack.
-+ * if the order here is changed, it needs to be
-+ * updated in fork.c:copy_process, signal.c:do_signal,
-+ * ptrace.c and ptrace.h
-+ *
-+ * 0(%esp) - %ebx
-+ * 4(%esp) - %ecx
-+ * 8(%esp) - %edx
-+ * C(%esp) - %esi
-+ * 10(%esp) - %edi
-+ * 14(%esp) - %ebp
-+ * 18(%esp) - %eax
-+ * 1C(%esp) - %ds
-+ * 20(%esp) - %es
-+ * 24(%esp) - orig_eax
-+ * 28(%esp) - %eip
-+ * 2C(%esp) - %cs
-+ * 30(%esp) - %eflags
-+ * 34(%esp) - %oldesp
-+ * 38(%esp) - %oldss
-+ *
-+ * "current" is in register %ebx during any slow entries.
++ * the CPU automatically disables interrupts when executing an IRQ vector,
++ * so IRQ-flags tracing has to follow that:
+ */
++ ALIGN
++common_interrupt:
++ SAVE_ALL
++ TRACE_IRQS_OFF
++ movl %esp,%eax
++ call do_IRQ
++ jmp ret_from_intr
++ CFI_ENDPROC
+
-+#include <linux/linkage.h>
-+#include <asm/thread_info.h>
-+#include <asm/irqflags.h>
-+#include <asm/errno.h>
-+#include <asm/segment.h>
-+#include <asm/smp.h>
-+#include <asm/page.h>
-+#include <asm/desc.h>
-+#include <asm/dwarf2.h>
-+#include "irq_vectors.h"
-+#include <xen/interface/xen.h>
-+
-+#define nr_syscalls ((syscall_table_size)/4)
-+
-+EBX = 0x00
-+ECX = 0x04
-+EDX = 0x08
-+ESI = 0x0C
-+EDI = 0x10
-+EBP = 0x14
-+EAX = 0x18
-+DS = 0x1C
-+ES = 0x20
-+ORIG_EAX = 0x24
-+EIP = 0x28
-+CS = 0x2C
-+EFLAGS = 0x30
-+OLDESP = 0x34
-+OLDSS = 0x38
-+
-+CF_MASK = 0x00000001
-+TF_MASK = 0x00000100
-+IF_MASK = 0x00000200
-+DF_MASK = 0x00000400
-+NT_MASK = 0x00004000
-+VM_MASK = 0x00020000
-+/* Pseudo-eflags. */
-+NMI_MASK = 0x80000000
-+
-+#ifndef CONFIG_XEN
-+#define DISABLE_INTERRUPTS cli
-+#define ENABLE_INTERRUPTS sti
-+#else
-+/* Offsets into shared_info_t. */
-+#define evtchn_upcall_pending /* 0 */
-+#define evtchn_upcall_mask 1
-+
-+#define sizeof_vcpu_shift 6
-+
-+#ifdef CONFIG_SMP
-+#define GET_VCPU_INFO movl TI_cpu(%ebp),%esi ; \
-+ shl $sizeof_vcpu_shift,%esi ; \
-+ addl HYPERVISOR_shared_info,%esi
-+#else
-+#define GET_VCPU_INFO movl HYPERVISOR_shared_info,%esi
-+#endif
-+
-+#define __DISABLE_INTERRUPTS movb $1,evtchn_upcall_mask(%esi)
-+#define __ENABLE_INTERRUPTS movb $0,evtchn_upcall_mask(%esi)
-+#define DISABLE_INTERRUPTS GET_VCPU_INFO ; \
-+ __DISABLE_INTERRUPTS
-+#define ENABLE_INTERRUPTS GET_VCPU_INFO ; \
-+ __ENABLE_INTERRUPTS
-+#define __TEST_PENDING testb $0xFF,evtchn_upcall_pending(%esi)
-+#endif
-+
-+#ifdef CONFIG_PREEMPT
-+#define preempt_stop cli; TRACE_IRQS_OFF
-+#else
-+#define preempt_stop
-+#define resume_kernel restore_nocheck
-+#endif
-+
-+.macro TRACE_IRQS_IRET
-+#ifdef CONFIG_TRACE_IRQFLAGS
-+ testl $IF_MASK,EFLAGS(%esp) # interrupts off?
-+ jz 1f
-+ TRACE_IRQS_ON
-+1:
-+#endif
-+.endm
++#define BUILD_INTERRUPT(name, nr) \
++ENTRY(name) \
++ RING0_INT_FRAME; \
++ pushl $~(nr); \
++ CFI_ADJUST_CFA_OFFSET 4; \
++ SAVE_ALL; \
++ TRACE_IRQS_OFF \
++ movl %esp,%eax; \
++ call smp_/**/name; \
++ jmp ret_from_intr; \
++ CFI_ENDPROC
+
-+#ifdef CONFIG_VM86
-+#define resume_userspace_sig check_userspace
++/* The include is where all of the SMP etc. interrupts come from */
++#include "entry_arch.h"
+#else
-+#define resume_userspace_sig resume_userspace
++#define UNWIND_ESPFIX_STACK
+#endif
+
-+#define SAVE_ALL \
-+ cld; \
-+ pushl %es; \
-+ CFI_ADJUST_CFA_OFFSET 4;\
-+ /*CFI_REL_OFFSET es, 0;*/\
-+ pushl %ds; \
-+ CFI_ADJUST_CFA_OFFSET 4;\
-+ /*CFI_REL_OFFSET ds, 0;*/\
-+ pushl %eax; \
-+ CFI_ADJUST_CFA_OFFSET 4;\
-+ CFI_REL_OFFSET eax, 0;\
-+ pushl %ebp; \
-+ CFI_ADJUST_CFA_OFFSET 4;\
-+ CFI_REL_OFFSET ebp, 0;\
-+ pushl %edi; \
-+ CFI_ADJUST_CFA_OFFSET 4;\
-+ CFI_REL_OFFSET edi, 0;\
-+ pushl %esi; \
-+ CFI_ADJUST_CFA_OFFSET 4;\
-+ CFI_REL_OFFSET esi, 0;\
-+ pushl %edx; \
-+ CFI_ADJUST_CFA_OFFSET 4;\
-+ CFI_REL_OFFSET edx, 0;\
-+ pushl %ecx; \
-+ CFI_ADJUST_CFA_OFFSET 4;\
-+ CFI_REL_OFFSET ecx, 0;\
-+ pushl %ebx; \
-+ CFI_ADJUST_CFA_OFFSET 4;\
-+ CFI_REL_OFFSET ebx, 0;\
-+ movl $(__USER_DS), %edx; \
-+ movl %edx, %ds; \
-+ movl %edx, %es;
-+
-+#define RESTORE_INT_REGS \
-+ popl %ebx; \
-+ CFI_ADJUST_CFA_OFFSET -4;\
-+ CFI_RESTORE ebx;\
-+ popl %ecx; \
-+ CFI_ADJUST_CFA_OFFSET -4;\
-+ CFI_RESTORE ecx;\
-+ popl %edx; \
-+ CFI_ADJUST_CFA_OFFSET -4;\
-+ CFI_RESTORE edx;\
-+ popl %esi; \
-+ CFI_ADJUST_CFA_OFFSET -4;\
-+ CFI_RESTORE esi;\
-+ popl %edi; \
-+ CFI_ADJUST_CFA_OFFSET -4;\
-+ CFI_RESTORE edi;\
-+ popl %ebp; \
-+ CFI_ADJUST_CFA_OFFSET -4;\
-+ CFI_RESTORE ebp;\
-+ popl %eax; \
-+ CFI_ADJUST_CFA_OFFSET -4;\
-+ CFI_RESTORE eax
-+
-+#define RESTORE_REGS \
-+ RESTORE_INT_REGS; \
-+1: popl %ds; \
-+ CFI_ADJUST_CFA_OFFSET -4;\
-+ /*CFI_RESTORE ds;*/\
-+2: popl %es; \
-+ CFI_ADJUST_CFA_OFFSET -4;\
-+ /*CFI_RESTORE es;*/\
-+.section .fixup,"ax"; \
-+3: movl $0,(%esp); \
-+ jmp 1b; \
-+4: movl $0,(%esp); \
-+ jmp 2b; \
-+.previous; \
-+.section __ex_table,"a";\
-+ .align 4; \
-+ .long 1b,3b; \
-+ .long 2b,4b; \
-+.previous
-+
-+#define RING0_INT_FRAME \
-+ CFI_STARTPROC simple;\
-+ CFI_DEF_CFA esp, 3*4;\
-+ /*CFI_OFFSET cs, -2*4;*/\
-+ CFI_OFFSET eip, -3*4
-+
-+#define RING0_EC_FRAME \
-+ CFI_STARTPROC simple;\
-+ CFI_DEF_CFA esp, 4*4;\
-+ /*CFI_OFFSET cs, -2*4;*/\
-+ CFI_OFFSET eip, -3*4
-+
-+#define RING0_PTREGS_FRAME \
-+ CFI_STARTPROC simple;\
-+ CFI_DEF_CFA esp, OLDESP-EBX;\
-+ /*CFI_OFFSET cs, CS-OLDESP;*/\
-+ CFI_OFFSET eip, EIP-OLDESP;\
-+ /*CFI_OFFSET es, ES-OLDESP;*/\
-+ /*CFI_OFFSET ds, DS-OLDESP;*/\
-+ CFI_OFFSET eax, EAX-OLDESP;\
-+ CFI_OFFSET ebp, EBP-OLDESP;\
-+ CFI_OFFSET edi, EDI-OLDESP;\
-+ CFI_OFFSET esi, ESI-OLDESP;\
-+ CFI_OFFSET edx, EDX-OLDESP;\
-+ CFI_OFFSET ecx, ECX-OLDESP;\
-+ CFI_OFFSET ebx, EBX-OLDESP
-+
-+ENTRY(ret_from_fork)
-+ CFI_STARTPROC
++KPROBE_ENTRY(page_fault)
++ RING0_EC_FRAME
++ pushl $do_page_fault
++ CFI_ADJUST_CFA_OFFSET 4
++ ALIGN
++error_code:
++ /* the function address is in %gs's slot on the stack */
++ pushl %es
++ CFI_ADJUST_CFA_OFFSET 4
++ /*CFI_REL_OFFSET es, 0*/
++ pushl %ds
++ CFI_ADJUST_CFA_OFFSET 4
++ /*CFI_REL_OFFSET ds, 0*/
+ pushl %eax
+ CFI_ADJUST_CFA_OFFSET 4
-+ call schedule_tail
-+ GET_THREAD_INFO(%ebp)
-+ popl %eax
-+ CFI_ADJUST_CFA_OFFSET -4
-+ pushl $0x0202 # Reset kernel eflags
-+ CFI_ADJUST_CFA_OFFSET 4
-+ popfl
-+ CFI_ADJUST_CFA_OFFSET -4
-+ jmp syscall_exit
-+ CFI_ENDPROC
-+
-+/*
-+ * Return to user mode is not as complex as all this looks,
-+ * but we want the default path for a system call return to
-+ * go as quickly as possible which is why some of this is
-+ * less clear than it otherwise should be.
-+ */
-+
-+ # userspace resumption stub bypassing syscall exit tracing
-+ ALIGN
-+ RING0_PTREGS_FRAME
-+ret_from_exception:
-+ preempt_stop
-+ret_from_intr:
-+ GET_THREAD_INFO(%ebp)
-+check_userspace:
-+ movl EFLAGS(%esp), %eax # mix EFLAGS and CS
-+ movb CS(%esp), %al
-+ testl $(VM_MASK | 2), %eax
-+ jz resume_kernel
-+ENTRY(resume_userspace)
-+ DISABLE_INTERRUPTS # make sure we don't miss an interrupt
-+ # setting need_resched or sigpending
-+ # between sampling and the iret
-+ movl TI_flags(%ebp), %ecx
-+ andl $_TIF_WORK_MASK, %ecx # is there any work to be done on
-+ # int/exception return?
-+ jne work_pending
-+ jmp restore_all
-+
-+#ifdef CONFIG_PREEMPT
-+ENTRY(resume_kernel)
-+ cli
-+ cmpl $0,TI_preempt_count(%ebp) # non-zero preempt_count ?
-+ jnz restore_nocheck
-+need_resched:
-+ movl TI_flags(%ebp), %ecx # need_resched set ?
-+ testb $_TIF_NEED_RESCHED, %cl
-+ jz restore_all
-+ testl $IF_MASK,EFLAGS(%esp) # interrupts off (exception path) ?
-+ jz restore_all
-+ call preempt_schedule_irq
-+ jmp need_resched
-+#endif
-+ CFI_ENDPROC
-+
-+/* SYSENTER_RETURN points to after the "sysenter" instruction in
-+ the vsyscall page. See vsyscall-sysentry.S, which defines the symbol. */
-+
-+ # sysenter call handler stub
-+ENTRY(sysenter_entry)
-+ CFI_STARTPROC simple
-+ CFI_DEF_CFA esp, 0
-+ CFI_REGISTER esp, ebp
-+ movl SYSENTER_stack_esp0(%esp),%esp
-+sysenter_past_esp:
-+ /*
-+ * No need to follow this irqs on/off section: the syscall
-+ * disabled irqs and here we enable it straight after entry:
-+ */
-+ sti
-+ pushl $(__USER_DS)
-+ CFI_ADJUST_CFA_OFFSET 4
-+ /*CFI_REL_OFFSET ss, 0*/
-+ pushl %ebp
-+ CFI_ADJUST_CFA_OFFSET 4
-+ CFI_REL_OFFSET esp, 0
-+ pushfl
-+ CFI_ADJUST_CFA_OFFSET 4
-+ pushl $(__USER_CS)
-+ CFI_ADJUST_CFA_OFFSET 4
-+ /*CFI_REL_OFFSET cs, 0*/
-+ /*
-+ * Push current_thread_info()->sysenter_return to the stack.
-+ * A tiny bit of offset fixup is necessary - 4*4 means the 4 words
-+ * pushed above; +8 corresponds to copy_thread's esp0 setting.
-+ */
-+ pushl (TI_sysenter_return-THREAD_SIZE+8+4*4)(%esp)
-+ CFI_ADJUST_CFA_OFFSET 4
-+ CFI_REL_OFFSET eip, 0
-+
-+/*
-+ * 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
-+
-+ pushl %eax
-+ CFI_ADJUST_CFA_OFFSET 4
-+ SAVE_ALL
-+ GET_THREAD_INFO(%ebp)
-+
-+ /* Note, _TIF_SECCOMP is bit number 8, and so it needs testw and not testb */
-+ testw $(_TIF_SYSCALL_EMU|_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT),TI_flags(%ebp)
-+ jnz syscall_trace_entry
-+ cmpl $(nr_syscalls), %eax
-+ jae syscall_badsys
-+ call *sys_call_table(,%eax,4)
-+ movl %eax,EAX(%esp)
-+ DISABLE_INTERRUPTS
-+ TRACE_IRQS_OFF
-+ movl TI_flags(%ebp), %ecx
-+ testw $_TIF_ALLWORK_MASK, %cx
-+ jne syscall_exit_work
-+/* if something modifies registers it must also disable sysexit */
-+ movl EIP(%esp), %edx
-+ movl OLDESP(%esp), %ecx
-+ xorl %ebp,%ebp
-+ TRACE_IRQS_ON
-+#ifdef CONFIG_XEN
-+ __ENABLE_INTERRUPTS
-+sysexit_scrit: /**** START OF SYSEXIT CRITICAL REGION ****/
-+ __TEST_PENDING
-+ jnz 14f # process more events if necessary...
-+ movl ESI(%esp), %esi
-+ sysexit
-+14: __DISABLE_INTERRUPTS
-+sysexit_ecrit: /**** END OF SYSEXIT CRITICAL REGION ****/
-+ push %esp
-+ CFI_ADJUST_CFA_OFFSET 4
-+ call evtchn_do_upcall
-+ add $4,%esp
-+ CFI_ADJUST_CFA_OFFSET -4
-+ jmp ret_from_intr
-+#else
-+ sti
-+ sysexit
-+#endif /* !CONFIG_XEN */
-+ CFI_ENDPROC
-+
-+
-+ # system call handler stub
-+ENTRY(system_call)
-+ RING0_INT_FRAME # can't unwind into user space anyway
-+ pushl %eax # save orig_eax
-+ CFI_ADJUST_CFA_OFFSET 4
-+ SAVE_ALL
-+ GET_THREAD_INFO(%ebp)
-+ testl $TF_MASK,EFLAGS(%esp)
-+ jz no_singlestep
-+ orl $_TIF_SINGLESTEP,TI_flags(%ebp)
-+no_singlestep:
-+ # system call tracing in operation / emulation
-+ /* Note, _TIF_SECCOMP is bit number 8, and so it needs testw and not testb */
-+ testw $(_TIF_SYSCALL_EMU|_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT),TI_flags(%ebp)
-+ jnz syscall_trace_entry
-+ cmpl $(nr_syscalls), %eax
-+ jae syscall_badsys
-+syscall_call:
-+ call *sys_call_table(,%eax,4)
-+ movl %eax,EAX(%esp) # store the return value
-+syscall_exit:
-+ DISABLE_INTERRUPTS # make sure we don't miss an interrupt
-+ # setting need_resched or sigpending
-+ # between sampling and the iret
-+ TRACE_IRQS_OFF
-+ movl TI_flags(%ebp), %ecx
-+ testw $_TIF_ALLWORK_MASK, %cx # current->work
-+ jne syscall_exit_work
-+
-+restore_all:
-+#ifndef CONFIG_XEN
-+ movl EFLAGS(%esp), %eax # mix EFLAGS, SS and CS
-+ # Warning: OLDSS(%esp) contains the wrong/random values if we
-+ # are returning to the kernel.
-+ # See comments in process.c:copy_thread() for details.
-+ movb OLDSS(%esp), %ah
-+ movb CS(%esp), %al
-+ andl $(VM_MASK | (4 << 8) | 3), %eax
-+ cmpl $((4 << 8) | 3), %eax
-+ je ldt_ss # returning to user-space with LDT SS
-+restore_nocheck:
-+#else
-+restore_nocheck:
-+ movl EFLAGS(%esp), %eax
-+ testl $(VM_MASK|NMI_MASK), %eax
-+ jnz hypervisor_iret
-+ shr $9, %eax # EAX[0] == IRET_EFLAGS.IF
-+ GET_VCPU_INFO
-+ andb evtchn_upcall_mask(%esi),%al
-+ andb $1,%al # EAX[0] == IRET_EFLAGS.IF & event_mask
-+ jnz restore_all_enable_events # != 0 => enable event delivery
-+#endif
-+ TRACE_IRQS_IRET
-+ CFI_REMEMBER_STATE
-+restore_nocheck_notrace:
-+ RESTORE_REGS
-+ addl $4, %esp
-+ CFI_ADJUST_CFA_OFFSET -4
-+1: iret
-+.section .fixup,"ax"
-+iret_exc:
-+#ifndef CONFIG_XEN
-+ TRACE_IRQS_ON
-+ sti
-+#endif
-+ pushl $0 # no error code
-+ pushl $do_iret_error
-+ jmp error_code
-+.previous
-+.section __ex_table,"a"
-+ .align 4
-+ .long 1b,iret_exc
-+.previous
-+
-+ CFI_RESTORE_STATE
-+#ifndef CONFIG_XEN
-+ldt_ss:
-+ larl OLDSS(%esp), %eax
-+ jnz restore_nocheck
-+ testl $0x00400000, %eax # returning to 32bit stack?
-+ jnz restore_nocheck # allright, normal return
-+ /* If returning to userspace with 16bit stack,
-+ * try to fix the higher word of ESP, as the CPU
-+ * won't restore it.
-+ * This is an "official" bug of all the x86-compatible
-+ * CPUs, which we can try to work around to make
-+ * dosemu and wine happy. */
-+ subl $8, %esp # reserve space for switch16 pointer
-+ CFI_ADJUST_CFA_OFFSET 8
-+ cli
-+ TRACE_IRQS_OFF
-+ movl %esp, %eax
-+ /* Set up the 16bit stack frame with switch32 pointer on top,
-+ * and a switch16 pointer on top of the current frame. */
-+ call setup_x86_bogus_stack
-+ CFI_ADJUST_CFA_OFFSET -8 # frame has moved
-+ TRACE_IRQS_IRET
-+ RESTORE_REGS
-+ lss 20+4(%esp), %esp # switch to 16bit stack
-+1: iret
-+.section __ex_table,"a"
-+ .align 4
-+ .long 1b,iret_exc
-+.previous
-+#else
-+hypervisor_iret:
-+ andl $~NMI_MASK, EFLAGS(%esp)
-+ RESTORE_REGS
-+ addl $4, %esp
-+ jmp hypercall_page + (__HYPERVISOR_iret * 32)
-+#endif
-+ CFI_ENDPROC
-+
-+ # perform work that needs to be done immediately before resumption
-+ ALIGN
-+ RING0_PTREGS_FRAME # can't unwind into user space anyway
-+work_pending:
-+ testb $_TIF_NEED_RESCHED, %cl
-+ jz work_notifysig
-+work_resched:
-+ call schedule
-+ DISABLE_INTERRUPTS # make sure we don't miss an interrupt
-+ # setting need_resched or sigpending
-+ # between sampling and the iret
-+ TRACE_IRQS_OFF
-+ movl TI_flags(%ebp), %ecx
-+ andl $_TIF_WORK_MASK, %ecx # is there any work to be done other
-+ # than syscall tracing?
-+ jz restore_all
-+ testb $_TIF_NEED_RESCHED, %cl
-+ jnz work_resched
-+
-+work_notifysig: # deal with pending signals and
-+ # notify-resume requests
-+ testl $VM_MASK, EFLAGS(%esp)
-+ movl %esp, %eax
-+ jne work_notifysig_v86 # returning to kernel-space or
-+ # vm86-space
-+ xorl %edx, %edx
-+ call do_notify_resume
-+ jmp resume_userspace_sig
-+
-+ ALIGN
-+work_notifysig_v86:
-+#ifdef CONFIG_VM86
-+ pushl %ecx # save ti_flags for do_notify_resume
-+ CFI_ADJUST_CFA_OFFSET 4
-+ call save_v86_state # %eax contains pt_regs pointer
-+ popl %ecx
-+ CFI_ADJUST_CFA_OFFSET -4
-+ movl %eax, %esp
-+ xorl %edx, %edx
-+ call do_notify_resume
-+ jmp resume_userspace_sig
-+#endif
-+
-+ # perform syscall exit tracing
-+ ALIGN
-+syscall_trace_entry:
-+ movl $-ENOSYS,EAX(%esp)
-+ movl %esp, %eax
-+ xorl %edx,%edx
-+ call do_syscall_trace
-+ cmpl $0, %eax
-+ jne resume_userspace # ret != 0 -> running under PTRACE_SYSEMU,
-+ # so must skip actual syscall
-+ movl ORIG_EAX(%esp), %eax
-+ cmpl $(nr_syscalls), %eax
-+ jnae syscall_call
-+ jmp syscall_exit
-+
-+ # perform syscall exit tracing
-+ ALIGN
-+syscall_exit_work:
-+ testb $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP), %cl
-+ jz work_pending
-+ TRACE_IRQS_ON
-+ ENABLE_INTERRUPTS # could let do_syscall_trace() call
-+ # schedule() instead
-+ movl %esp, %eax
-+ movl $1, %edx
-+ call do_syscall_trace
-+ jmp resume_userspace
-+ CFI_ENDPROC
-+
-+ RING0_INT_FRAME # can't unwind into user space anyway
-+syscall_fault:
-+ pushl %eax # save orig_eax
-+ CFI_ADJUST_CFA_OFFSET 4
-+ SAVE_ALL
-+ GET_THREAD_INFO(%ebp)
-+ movl $-EFAULT,EAX(%esp)
-+ jmp resume_userspace
-+
-+syscall_badsys:
-+ movl $-ENOSYS,EAX(%esp)
-+ jmp resume_userspace
-+ CFI_ENDPROC
-+
-+#ifndef CONFIG_XEN
-+#define FIXUP_ESPFIX_STACK \
-+ movl %esp, %eax; \
-+ /* switch to 32bit stack using the pointer on top of 16bit stack */ \
-+ lss %ss:CPU_16BIT_STACK_SIZE-8, %esp; \
-+ /* copy data from 16bit stack to 32bit stack */ \
-+ call fixup_x86_bogus_stack; \
-+ /* put ESP to the proper location */ \
-+ movl %eax, %esp;
-+#define UNWIND_ESPFIX_STACK \
-+ pushl %eax; \
-+ CFI_ADJUST_CFA_OFFSET 4; \
-+ movl %ss, %eax; \
-+ /* see if on 16bit stack */ \
-+ cmpw $__ESPFIX_SS, %ax; \
-+ je 28f; \
-+27: popl %eax; \
-+ CFI_ADJUST_CFA_OFFSET -4; \
-+.section .fixup,"ax"; \
-+28: movl $__KERNEL_DS, %eax; \
-+ movl %eax, %ds; \
-+ movl %eax, %es; \
-+ /* switch to 32bit stack */ \
-+ FIXUP_ESPFIX_STACK; \
-+ jmp 27b; \
-+.previous
-+
-+/*
-+ * Build the entry stubs and pointer table with
-+ * some assembler magic.
-+ */
-+.data
-+ENTRY(interrupt)
-+.text
-+
-+vector=0
-+ENTRY(irq_entries_start)
-+ RING0_INT_FRAME
-+.rept NR_IRQS
-+ ALIGN
-+ .if vector
-+ CFI_ADJUST_CFA_OFFSET -4
-+ .endif
-+1: pushl $~(vector)
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp common_interrupt
-+.data
-+ .long 1b
-+.text
-+vector=vector+1
-+.endr
-+
-+/*
-+ * the CPU automatically disables interrupts when executing an IRQ vector,
-+ * so IRQ-flags tracing has to follow that:
-+ */
-+ ALIGN
-+common_interrupt:
-+ SAVE_ALL
-+ TRACE_IRQS_OFF
-+ movl %esp,%eax
-+ call do_IRQ
-+ jmp ret_from_intr
-+ CFI_ENDPROC
-+
-+#define BUILD_INTERRUPT(name, nr) \
-+ENTRY(name) \
-+ SAVE_ALL \
-+ RING0_INT_FRAME; \
-+ pushl $~(nr); \
-+ CFI_ADJUST_CFA_OFFSET 4; \
-+ SAVE_ALL; \
-+ TRACE_IRQS_OFF \
-+ movl %esp,%eax; \
-+ call smp_/**/name; \
-+ jmp ret_from_intr; \
-+ CFI_ENDPROC
-+
-+/* The include is where all of the SMP etc. interrupts come from */
-+#include "entry_arch.h"
-+#else
-+#define UNWIND_ESPFIX_STACK
-+#endif
-+
-+ENTRY(divide_error)
-+ RING0_INT_FRAME
-+ pushl $0 # no error code
-+ CFI_ADJUST_CFA_OFFSET 4
-+ pushl $do_divide_error
-+ CFI_ADJUST_CFA_OFFSET 4
-+ ALIGN
-+error_code:
-+ pushl %ds
-+ CFI_ADJUST_CFA_OFFSET 4
-+ /*CFI_REL_OFFSET ds, 0*/
-+ pushl %eax
-+ CFI_ADJUST_CFA_OFFSET 4
-+ CFI_REL_OFFSET eax, 0
-+ xorl %eax, %eax
-+ pushl %ebp
++ CFI_REL_OFFSET eax, 0
++ pushl %ebp
+ CFI_ADJUST_CFA_OFFSET 4
+ CFI_REL_OFFSET ebp, 0
+ pushl %edi
@@ -3433,7 +2777,6 @@
+ pushl %edx
+ CFI_ADJUST_CFA_OFFSET 4
+ CFI_REL_OFFSET edx, 0
-+ decl %eax # eax = -1
+ pushl %ecx
+ CFI_ADJUST_CFA_OFFSET 4
+ CFI_REL_OFFSET ecx, 0
@@ -3441,18 +2784,20 @@
+ CFI_ADJUST_CFA_OFFSET 4
+ CFI_REL_OFFSET ebx, 0
+ cld
-+ pushl %es
++ pushl %gs
+ CFI_ADJUST_CFA_OFFSET 4
-+ /*CFI_REL_OFFSET es, 0*/
++ /*CFI_REL_OFFSET gs, 0*/
++ movl $(__KERNEL_PDA), %ecx
++ movl %ecx, %gs
+ UNWIND_ESPFIX_STACK
+ popl %ecx
+ CFI_ADJUST_CFA_OFFSET -4
+ /*CFI_REGISTER es, ecx*/
-+ movl ES(%esp), %edi # get the function address
-+ movl ORIG_EAX(%esp), %edx # get the error code
-+ movl %eax, ORIG_EAX(%esp)
-+ movl %ecx, ES(%esp)
-+ /*CFI_REL_OFFSET es, ES*/
++ movl PT_GS(%esp), %edi # get the function address
++ movl PT_ORIG_EAX(%esp), %edx # get the error code
++ movl $-1, PT_ORIG_EAX(%esp) # no syscall to restart
++ mov %ecx, PT_GS(%esp)
++ /*CFI_REL_OFFSET gs, ES*/
+ movl $(__USER_DS), %ecx
+ movl %ecx, %ds
+ movl %ecx, %es
@@ -3460,6 +2805,7 @@
+ call *%edi
+ jmp ret_from_exception
+ CFI_ENDPROC
++KPROBE_END(page_fault)
+
+#ifdef CONFIG_XEN
+# A note on the "critical region" in our callback handler.
@@ -3483,7 +2829,7 @@
+ pushl %eax
+ CFI_ADJUST_CFA_OFFSET 4
+ SAVE_ALL
-+ movl EIP(%esp),%eax
++ movl PT_EIP(%esp),%eax
+ cmpl $scrit,%eax
+ jb 11f
+ cmpl $ecrit,%eax
@@ -3492,14368 +2838,10832 @@
+ jb 11f
+ cmpl $sysexit_ecrit,%eax
+ ja 11f
-+ # interrupted in sysexit critical
-+ addl $0x34,%esp # Remove cs...ebx from stack frame.
-+ # this popped off new frame to reuse the old one, therefore no
-+ # CFI_ADJUST_CFA_OFFSET here
-+11: push %esp
-+ CFI_ADJUST_CFA_OFFSET 4
-+ call evtchn_do_upcall
-+ add $4,%esp
-+ CFI_ADJUST_CFA_OFFSET -4
-+ jmp ret_from_intr
-+
-+ ALIGN
-+restore_all_enable_events:
-+ __ENABLE_INTERRUPTS
-+scrit: /**** START OF CRITICAL REGION ****/
-+ __TEST_PENDING
-+ jnz 14f # process more events if necessary...
-+ RESTORE_REGS
-+ addl $4, %esp
-+ CFI_ADJUST_CFA_OFFSET -4
-+1: iret
-+.section __ex_table,"a"
-+ .align 4
-+ .long 1b,iret_exc
-+.previous
-+14: __DISABLE_INTERRUPTS
-+ jmp 11b
-+ecrit: /**** END OF CRITICAL REGION ****/
-+# [How we do the fixup]. We want to merge the current stack frame with the
-+# just-interrupted frame. How we do this depends on where in the critical
-+# region the interrupted handler was executing, and so how many saved
-+# registers are in each frame. We do this quickly using the lookup table
-+# 'critical_fixup_table'. For each byte offset in the critical region, it
-+# provides the number of bytes which have already been popped from the
-+# interrupted stack frame.
-+critical_region_fixup:
-+ addl $critical_fixup_table-scrit,%eax
-+ movzbl (%eax),%eax # %eax contains num bytes popped
-+ cmpb $0xff,%al # 0xff => vcpu_info critical region
-+ jne 15f
-+ GET_THREAD_INFO(%ebp)
-+ xorl %eax,%eax
-+15: mov %esp,%esi
-+ add %eax,%esi # %esi points at end of src region
-+ mov %esp,%edi
-+ add $0x34,%edi # %edi points at end of dst region
-+ mov %eax,%ecx
-+ shr $2,%ecx # convert words to bytes
-+ je 17f # skip loop if nothing to copy
-+16: subl $4,%esi # pre-decrementing copy loop
-+ subl $4,%edi
-+ movl (%esi),%eax
-+ movl %eax,(%edi)
-+ loop 16b
-+17: movl %edi,%esp # final %edi is top of merged stack
-+ # this popped off new frame to reuse the old one, therefore no
-+ # CFI_DEF_CFA_OFFSET here
-+ jmp 11b
-+ CFI_ENDPROC
-+
-+critical_fixup_table:
-+ .byte 0xff,0xff,0xff # testb $0xff,(%esi) = __TEST_PENDING
-+ .byte 0xff,0xff # jnz 14f
-+ .byte 0x00 # pop %ebx
-+ .byte 0x04 # pop %ecx
-+ .byte 0x08 # pop %edx
-+ .byte 0x0c # pop %esi
-+ .byte 0x10 # pop %edi
-+ .byte 0x14 # pop %ebp
-+ .byte 0x18 # pop %eax
-+ .byte 0x1c # pop %ds
-+ .byte 0x20 # pop %es
-+ .byte 0x24,0x24,0x24 # add $4,%esp
-+ .byte 0x28 # iret
-+ .byte 0xff,0xff,0xff,0xff # movb $1,1(%esi)
-+ .byte 0x00,0x00 # jmp 11b
-+
-+# 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 fix up by reattempting the load, and zeroing the segment
-+# register if the load fails.
-+# Category 2 we fix up by jumping to do_iret_error. 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 maintaining a status value in EAX.
-+ENTRY(failsafe_callback)
-+ RING0_INT_FRAME
-+ pushl %eax
-+ CFI_ADJUST_CFA_OFFSET 4
-+ movl $1,%eax
-+1: mov 4(%esp),%ds
-+2: mov 8(%esp),%es
-+3: mov 12(%esp),%fs
-+4: mov 16(%esp),%gs
-+ testl %eax,%eax
-+ popl %eax
-+ CFI_ADJUST_CFA_OFFSET -4
-+ jz 5f
-+ addl $16,%esp # EAX != 0 => Category 2 (Bad IRET)
-+ CFI_ADJUST_CFA_OFFSET -16
-+ jmp iret_exc
-+ CFI_ADJUST_CFA_OFFSET 16
-+5: addl $16,%esp # EAX == 0 => Category 1 (Bad segment)
-+ CFI_ADJUST_CFA_OFFSET -16
-+ pushl $0
-+ CFI_ADJUST_CFA_OFFSET 4
-+ SAVE_ALL
-+ jmp ret_from_exception
-+.section .fixup,"ax"; \
-+6: xorl %eax,%eax; \
-+ movl %eax,4(%esp); \
-+ jmp 1b; \
-+7: xorl %eax,%eax; \
-+ movl %eax,8(%esp); \
-+ jmp 2b; \
-+8: xorl %eax,%eax; \
-+ movl %eax,12(%esp); \
-+ jmp 3b; \
-+9: xorl %eax,%eax; \
-+ movl %eax,16(%esp); \
-+ jmp 4b; \
-+.previous; \
-+.section __ex_table,"a"; \
-+ .align 4; \
-+ .long 1b,6b; \
-+ .long 2b,7b; \
-+ .long 3b,8b; \
-+ .long 4b,9b; \
-+.previous
-+ CFI_ENDPROC
-+#endif
-+
-+ENTRY(coprocessor_error)
-+ RING0_INT_FRAME
-+ pushl $0
-+ CFI_ADJUST_CFA_OFFSET 4
-+ pushl $do_coprocessor_error
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+
-+ENTRY(simd_coprocessor_error)
-+ RING0_INT_FRAME
-+ pushl $0
-+ CFI_ADJUST_CFA_OFFSET 4
-+ pushl $do_simd_coprocessor_error
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+
-+ENTRY(device_not_available)
-+ RING0_INT_FRAME
-+ pushl $-1 # mark this as an int
-+ CFI_ADJUST_CFA_OFFSET 4
-+ SAVE_ALL
-+#ifndef CONFIG_XEN
-+ movl %cr0, %eax
-+ testl $0x4, %eax # EM (math emulation bit)
-+ je device_available_emulate
-+ pushl $0 # temporary storage for ORIG_EIP
-+ CFI_ADJUST_CFA_OFFSET 4
-+ call math_emulate
-+ addl $4, %esp
-+ CFI_ADJUST_CFA_OFFSET -4
-+ jmp ret_from_exception
-+device_available_emulate:
-+#endif
-+ preempt_stop
-+ call math_state_restore
-+ jmp ret_from_exception
-+ CFI_ENDPROC
-+
-+#ifndef CONFIG_XEN
-+/*
-+ * Debug traps and NMI can happen at the one SYSENTER instruction
-+ * that sets up the real kernel stack. Check here, since we can't
-+ * allow the wrong stack to be used.
-+ *
-+ * "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.
-+ *
-+ * We just load the right stack, and push the three (known) values
-+ * by hand onto the new stack - while updating the return eip past
-+ * the instruction that would have done it for sysenter.
-+ */
-+#define FIX_STACK(offset, ok, label) \
-+ cmpw $__KERNEL_CS,4(%esp); \
-+ jne ok; \
-+label: \
-+ movl SYSENTER_stack_esp0+offset(%esp),%esp; \
-+ pushfl; \
-+ pushl $__KERNEL_CS; \
-+ pushl $sysenter_past_esp
-+#endif /* CONFIG_XEN */
-+
-+KPROBE_ENTRY(debug)
-+ RING0_INT_FRAME
-+#ifndef CONFIG_XEN
-+ cmpl $sysenter_entry,(%esp)
-+ jne debug_stack_correct
-+ FIX_STACK(12, debug_stack_correct, debug_esp_fix_insn)
-+debug_stack_correct:
-+#endif /* !CONFIG_XEN */
-+ pushl $-1 # mark this as an int
-+ CFI_ADJUST_CFA_OFFSET 4
-+ SAVE_ALL
-+ xorl %edx,%edx # error code 0
-+ movl %esp,%eax # pt_regs pointer
-+ call do_debug
-+ jmp ret_from_exception
-+ CFI_ENDPROC
-+ .previous .text
-+
-+#ifndef CONFIG_XEN
-+/*
-+ * NMI is doubly nasty. It can happen _while_ we're handling
-+ * a debug fault, and the debug fault hasn't yet been able to
-+ * clear up the stack. So we first check whether we got an
-+ * NMI on the sysenter entry path, but after that we need to
-+ * check whether we got an NMI on the debug path where the debug
-+ * fault happened on the sysenter path.
-+ */
-+ENTRY(nmi)
-+ RING0_INT_FRAME
-+ pushl %eax
-+ CFI_ADJUST_CFA_OFFSET 4
-+ movl %ss, %eax
-+ cmpw $__ESPFIX_SS, %ax
-+ popl %eax
-+ CFI_ADJUST_CFA_OFFSET -4
-+ je nmi_16bit_stack
-+ cmpl $sysenter_entry,(%esp)
-+ je nmi_stack_fixup
-+ pushl %eax
-+ CFI_ADJUST_CFA_OFFSET 4
-+ movl %esp,%eax
-+ /* Do not access memory above the end of our stack page,
-+ * it might not exist.
-+ */
-+ andl $(THREAD_SIZE-1),%eax
-+ cmpl $(THREAD_SIZE-20),%eax
-+ popl %eax
-+ CFI_ADJUST_CFA_OFFSET -4
-+ jae nmi_stack_correct
-+ cmpl $sysenter_entry,12(%esp)
-+ je nmi_debug_stack_check
-+nmi_stack_correct:
-+ pushl %eax
-+ CFI_ADJUST_CFA_OFFSET 4
-+ SAVE_ALL
-+ xorl %edx,%edx # zero error code
-+ movl %esp,%eax # pt_regs pointer
-+ call do_nmi
-+ jmp restore_nocheck_notrace
-+ CFI_ENDPROC
-+
-+nmi_stack_fixup:
-+ FIX_STACK(12,nmi_stack_correct, 1)
-+ jmp nmi_stack_correct
-+nmi_debug_stack_check:
-+ cmpw $__KERNEL_CS,16(%esp)
-+ jne nmi_stack_correct
-+ cmpl $debug,(%esp)
-+ jb nmi_stack_correct
-+ cmpl $debug_esp_fix_insn,(%esp)
-+ ja nmi_stack_correct
-+ FIX_STACK(24,nmi_stack_correct, 1)
-+ jmp nmi_stack_correct
-+
-+nmi_16bit_stack:
-+ RING0_INT_FRAME
-+ /* create the pointer to lss back */
-+ pushl %ss
-+ CFI_ADJUST_CFA_OFFSET 4
-+ pushl %esp
-+ CFI_ADJUST_CFA_OFFSET 4
-+ movzwl %sp, %esp
-+ addw $4, (%esp)
-+ /* copy the iret frame of 12 bytes */
-+ .rept 3
-+ pushl 16(%esp)
-+ CFI_ADJUST_CFA_OFFSET 4
-+ .endr
-+ pushl %eax
-+ CFI_ADJUST_CFA_OFFSET 4
-+ SAVE_ALL
-+ FIXUP_ESPFIX_STACK # %eax == %esp
-+ CFI_ADJUST_CFA_OFFSET -20 # the frame has now moved
-+ xorl %edx,%edx # zero error code
-+ call do_nmi
-+ RESTORE_REGS
-+ lss 12+4(%esp), %esp # back to 16bit stack
-+1: iret
-+ CFI_ENDPROC
-+.section __ex_table,"a"
-+ .align 4
-+ .long 1b,iret_exc
-+.previous
-+#else
-+ENTRY(nmi)
-+ RING0_INT_FRAME
-+ pushl %eax
-+ CFI_ADJUST_CFA_OFFSET 4
-+ SAVE_ALL
-+ xorl %edx,%edx # zero error code
-+ movl %esp,%eax # pt_regs pointer
-+ call do_nmi
-+ orl $NMI_MASK, EFLAGS(%esp)
-+ jmp restore_all
-+ CFI_ENDPROC
-+#endif
-+
-+KPROBE_ENTRY(int3)
-+ RING0_INT_FRAME
-+ pushl $-1 # mark this as an int
-+ CFI_ADJUST_CFA_OFFSET 4
-+ SAVE_ALL
-+ xorl %edx,%edx # zero error code
-+ movl %esp,%eax # pt_regs pointer
-+ call do_int3
-+ jmp ret_from_exception
-+ CFI_ENDPROC
-+ .previous .text
-+
-+ENTRY(overflow)
-+ RING0_INT_FRAME
-+ pushl $0
-+ CFI_ADJUST_CFA_OFFSET 4
-+ pushl $do_overflow
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+
-+ENTRY(bounds)
-+ RING0_INT_FRAME
-+ pushl $0
-+ CFI_ADJUST_CFA_OFFSET 4
-+ pushl $do_bounds
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+
-+ENTRY(invalid_op)
-+ RING0_INT_FRAME
-+ pushl $0
-+ CFI_ADJUST_CFA_OFFSET 4
-+ pushl $do_invalid_op
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+
-+ENTRY(coprocessor_segment_overrun)
-+ RING0_INT_FRAME
-+ pushl $0
-+ CFI_ADJUST_CFA_OFFSET 4
-+ pushl $do_coprocessor_segment_overrun
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+
-+ENTRY(invalid_TSS)
-+ RING0_EC_FRAME
-+ pushl $do_invalid_TSS
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+
-+ENTRY(segment_not_present)
-+ RING0_EC_FRAME
-+ pushl $do_segment_not_present
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+
-+ENTRY(stack_segment)
-+ RING0_EC_FRAME
-+ pushl $do_stack_segment
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+
-+KPROBE_ENTRY(general_protection)
-+ RING0_EC_FRAME
-+ pushl $do_general_protection
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+ .previous .text
-+
-+ENTRY(alignment_check)
-+ RING0_EC_FRAME
-+ pushl $do_alignment_check
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+
-+KPROBE_ENTRY(page_fault)
-+ RING0_EC_FRAME
-+ pushl $do_page_fault
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+ .previous .text
-+
-+#ifdef CONFIG_X86_MCE
-+ENTRY(machine_check)
-+ RING0_INT_FRAME
-+ pushl $0
-+ CFI_ADJUST_CFA_OFFSET 4
-+ pushl machine_check_vector
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+#endif
-+
-+ENTRY(fixup_4gb_segment)
-+ RING0_INT_FRAME
-+ pushl $do_fixup_4gb_segment
-+ CFI_ADJUST_CFA_OFFSET 4
-+ jmp error_code
-+ CFI_ENDPROC
-+
-+#ifdef CONFIG_STACK_UNWIND
-+ENTRY(arch_unwind_init_running)
-+ CFI_STARTPROC
-+ movl 4(%esp), %edx
-+ movl (%esp), %ecx
-+ leal 4(%esp), %eax
-+ movl %ebx, EBX(%edx)
-+ xorl %ebx, %ebx
-+ movl %ebx, ECX(%edx)
-+ movl %ebx, EDX(%edx)
-+ movl %esi, ESI(%edx)
-+ movl %edi, EDI(%edx)
-+ movl %ebp, EBP(%edx)
-+ movl %ebx, EAX(%edx)
-+ movl $__USER_DS, DS(%edx)
-+ movl $__USER_DS, ES(%edx)
-+ movl %ebx, ORIG_EAX(%edx)
-+ movl %ecx, EIP(%edx)
-+ movl 12(%esp), %ecx
-+ movl $__KERNEL_CS, CS(%edx)
-+ movl %ebx, EFLAGS(%edx)
-+ movl %eax, OLDESP(%edx)
-+ movl 8(%esp), %eax
-+ movl %ecx, 8(%esp)
-+ movl EBX(%edx), %ebx
-+ movl $__KERNEL_DS, OLDSS(%edx)
-+ jmpl *%eax
-+ CFI_ENDPROC
-+ENDPROC(arch_unwind_init_running)
-+#endif
-+
-+.section .rodata,"a"
-+#include "syscall_table.S"
-+
-+syscall_table_size=(.-sys_call_table)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/fixup.c linux-2.6.18-xen/arch/i386/kernel/fixup.c
---- linux-2.6.18.3/arch/i386/kernel/fixup.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/fixup.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,92 @@
-+/******************************************************************************
-+ * fixup.c
-+ *
-+ * Binary-rewriting of certain IA32 instructions, on notification by Xen.
-+ * Used to avoid repeated slow emulation of common instructions used by the
-+ * user-space TLS (Thread-Local Storage) libraries.
-+ *
-+ * **** NOTE ****
-+ * Issues with the binary rewriting have caused it to be removed. Instead
-+ * we rely on Xen's emulator to boot the kernel, and then print a banner
-+ * message recommending that the user disables /lib/tls.
-+ *
-+ * Copyright (c) 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 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/init.h>
-+#include <linux/sched.h>
-+#include <linux/slab.h>
-+#include <linux/kernel.h>
-+#include <linux/delay.h>
-+#include <linux/version.h>
-+
-+#define DP(_f, _args...) printk(KERN_ALERT " " _f "\n" , ## _args )
-+
-+fastcall void do_fixup_4gb_segment(struct pt_regs *regs, long error_code)
-+{
-+#if 0
-+ static unsigned long printed = 0;
-+ char info[100];
-+ int i;
-+
-+ if (test_and_set_bit(0, &printed))
-+ return;
-+
-+ HYPERVISOR_vm_assist(
-+ VMASST_CMD_disable, VMASST_TYPE_4gb_segments_notify);
-+
-+ sprintf(info, "%s (pid=%d)", current->comm, current->tgid);
-+
-+
-+ DP("");
-+ DP("***************************************************************");
-+ DP("***************************************************************");
-+ DP("** WARNING: Currently emulating unsupported memory accesses **");
-+ DP("** in /lib/tls glibc libraries. The emulation is **");
-+ DP("** slow. To ensure full performance you should **");
-+ DP("** install a 'xen-friendly' (nosegneg) version of **");
-+ DP("** the library, or disable tls support by executing **");
-+ DP("** the following as root: **");
-+ DP("** mv /lib/tls /lib/tls.disabled **");
-+ DP("** Offending process: %-38.38s **", info);
-+ DP("***************************************************************");
-+ DP("***************************************************************");
-+ DP("");
-+
-+ for (i = 5; i > 0; i--) {
-+ touch_softlockup_watchdog();
-+ printk("Pausing... %d", i);
-+ mdelay(1000);
-+ printk("\b\b\b\b\b\b\b\b\b\b\b\b");
-+ }
-+
-+ printk("Continuing...\n\n");
-+#else
-+ if (printk_ratelimit())
-+ printk(KERN_WARNING
-+ "4gb seg fixup, process %s (pid %d), cs:ip %02x:%08lx\n",
-+ current->comm, current->tgid, regs->xcs, regs->eip);
-+#endif
-+}
-+
-+static int __init fixup_init(void)
-+{
-+ HYPERVISOR_vm_assist(
-+ VMASST_CMD_enable, VMASST_TYPE_4gb_segments_notify);
-+ return 0;
-+}
-+__initcall(fixup_init);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/head-xen.S linux-2.6.18-xen/arch/i386/kernel/head-xen.S
---- linux-2.6.18.3/arch/i386/kernel/head-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/head-xen.S 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,200 @@
-+
-+
-+.text
-+#include <linux/elfnote.h>
-+#include <linux/threads.h>
-+#include <linux/linkage.h>
-+#include <asm/segment.h>
-+#include <asm/page.h>
-+#include <asm/thread_info.h>
-+#include <asm/asm-offsets.h>
-+#include <xen/interface/arch-x86_32.h>
-+#include <xen/interface/elfnote.h>
-+
-+/*
-+ * References to members of the new_cpu_data structure.
-+ */
-+
-+#define X86 new_cpu_data+CPUINFO_x86
-+#define X86_VENDOR new_cpu_data+CPUINFO_x86_vendor
-+#define X86_MODEL new_cpu_data+CPUINFO_x86_model
-+#define X86_MASK new_cpu_data+CPUINFO_x86_mask
-+#define X86_HARD_MATH new_cpu_data+CPUINFO_hard_math
-+#define X86_CPUID new_cpu_data+CPUINFO_cpuid_level
-+#define X86_CAPABILITY new_cpu_data+CPUINFO_x86_capability
-+#define X86_VENDOR_ID new_cpu_data+CPUINFO_x86_vendor_id
-+
-+#define VIRT_ENTRY_OFFSET 0x0
-+.org VIRT_ENTRY_OFFSET
-+ENTRY(startup_32)
-+ movl %esi,xen_start_info
-+ cld
-+
-+ /* Set up the stack pointer */
-+ movl $(init_thread_union+THREAD_SIZE),%esp
-+
-+ /* get vendor info */
-+ xorl %eax,%eax # call CPUID with 0 -> return vendor ID
-+ XEN_CPUID
-+ movl %eax,X86_CPUID # save CPUID level
-+ movl %ebx,X86_VENDOR_ID # lo 4 chars
-+ movl %edx,X86_VENDOR_ID+4 # next 4 chars
-+ movl %ecx,X86_VENDOR_ID+8 # last 4 chars
-+
-+ movl $1,%eax # Use the CPUID instruction to get CPU type
-+ XEN_CPUID
-+ movb %al,%cl # save reg for future use
-+ andb $0x0f,%ah # mask processor family
-+ movb %ah,X86
-+ andb $0xf0,%al # mask model
-+ shrb $4,%al
-+ movb %al,X86_MODEL
-+ andb $0x0f,%cl # mask mask revision
-+ movb %cl,X86_MASK
-+ movl %edx,X86_CAPABILITY
-+
-+ movb $1,X86_HARD_MATH
-+
-+ xorl %eax,%eax # Clear FS/GS and LDT
-+ movl %eax,%fs
-+ movl %eax,%gs
-+ cld # gcc2 wants the direction flag cleared at all times
-+
-+ call start_kernel
-+L6:
-+ jmp L6 # main should never return here, but
-+ # just in case, we know what happens.
-+
-+#define HYPERCALL_PAGE_OFFSET 0x1000
-+.org HYPERCALL_PAGE_OFFSET
-+ENTRY(hypercall_page)
-+.skip 0x1000
-+
-+/*
-+ * Real beginning of normal "text" segment
-+ */
-+ENTRY(stext)
-+ENTRY(_stext)
-+
-+/*
-+ * BSS section
-+ */
-+.section ".bss.page_aligned","w"
-+ENTRY(empty_zero_page)
-+ .fill 4096,1,0
-+
-+/*
-+ * This starts the data section.
-+ */
-+.data
-+
-+/*
-+ * The Global Descriptor Table contains 28 quadwords, per-CPU.
-+ */
-+ENTRY(cpu_gdt_table)
-+ .quad 0x0000000000000000 /* NULL descriptor */
-+ .quad 0x0000000000000000 /* 0x0b reserved */
-+ .quad 0x0000000000000000 /* 0x13 reserved */
-+ .quad 0x0000000000000000 /* 0x1b reserved */
-+ .quad 0x0000000000000000 /* 0x20 unused */
-+ .quad 0x0000000000000000 /* 0x28 unused */
-+ .quad 0x0000000000000000 /* 0x33 TLS entry 1 */
-+ .quad 0x0000000000000000 /* 0x3b TLS entry 2 */
-+ .quad 0x0000000000000000 /* 0x43 TLS entry 3 */
-+ .quad 0x0000000000000000 /* 0x4b reserved */
-+ .quad 0x0000000000000000 /* 0x53 reserved */
-+ .quad 0x0000000000000000 /* 0x5b reserved */
-+
-+ .quad 0x00cf9a000000ffff /* 0x60 kernel 4GB code at 0x00000000 */
-+ .quad 0x00cf92000000ffff /* 0x68 kernel 4GB data at 0x00000000 */
-+ .quad 0x00cffa000000ffff /* 0x73 user 4GB code at 0x00000000 */
-+ .quad 0x00cff2000000ffff /* 0x7b user 4GB data at 0x00000000 */
-+
-+ .quad 0x0000000000000000 /* 0x80 TSS descriptor */
-+ .quad 0x0000000000000000 /* 0x88 LDT descriptor */
-+
-+ /*
-+ * Segments used for calling PnP BIOS have byte granularity.
-+ * They code segments and data segments have fixed 64k limits,
-+ * the transfer segment sizes are set at run time.
-+ */
-+ .quad 0x0000000000000000 /* 0x90 32-bit code */
-+ .quad 0x0000000000000000 /* 0x98 16-bit code */
-+ .quad 0x0000000000000000 /* 0xa0 16-bit data */
-+ .quad 0x0000000000000000 /* 0xa8 16-bit data */
-+ .quad 0x0000000000000000 /* 0xb0 16-bit data */
-+
-+ /*
-+ * The APM segments have byte granularity and their bases
-+ * are set at run time. All have 64k limits.
-+ */
-+ .quad 0x0000000000000000 /* 0xb8 APM CS code */
-+ .quad 0x0000000000000000 /* 0xc0 APM CS 16 code (16 bit) */
-+ .quad 0x0000000000000000 /* 0xc8 APM DS data */
-+
-+ .quad 0x0000000000000000 /* 0xd0 - ESPFIX 16-bit SS */
-+ .quad 0x0000000000000000 /* 0xd8 - unused */
-+ .quad 0x0000000000000000 /* 0xe0 - unused */
-+ .quad 0x0000000000000000 /* 0xe8 - unused */
-+ .quad 0x0000000000000000 /* 0xf0 - unused */
-+ .quad 0x0000000000000000 /* 0xf8 - GDT entry 31: double-fault TSS */
-+
-+#ifdef CONFIG_XEN_COMPAT_030002
-+/*
-+ * __xen_guest information
-+ */
-+.macro utoa value
-+ .if (\value) < 0 || (\value) >= 0x10
-+ utoa (((\value)>>4)&0x0fffffff)
-+ .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"
-+ utoa __PAGE_OFFSET
-+ .ascii ",ELF_PADDR_OFFSET=0x"
-+ utoa __PAGE_OFFSET
-+ .ascii ",VIRT_ENTRY=0x"
-+ utoa (__PAGE_OFFSET + __PHYSICAL_START + VIRT_ENTRY_OFFSET)
-+ .ascii ",HYPERCALL_PAGE=0x"
-+ utoa ((__PHYSICAL_START+HYPERCALL_PAGE_OFFSET)>>PAGE_SHIFT)
-+ .ascii ",FEATURES=writable_page_tables"
-+ .ascii "|writable_descriptor_tables"
-+ .ascii "|auto_translated_physmap"
-+ .ascii "|pae_pgdir_above_4gb"
-+ .ascii "|supervisor_mode_kernel"
-+#ifdef CONFIG_X86_PAE
-+ .ascii ",PAE=yes[extended-cr3]"
-+#else
-+ .ascii ",PAE=no"
-+#endif
-+ .ascii ",LOADER=generic"
-+ .byte 0
-+#endif /* CONFIG_XEN_COMPAT_030002 */
-+
-+
-+ 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, .long, __PAGE_OFFSET)
-+#ifdef CONFIG_XEN_COMPAT_030002
-+ ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET, .long, __PAGE_OFFSET)
-+#else
-+ ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET, .long, 0)
-+#endif /* !CONFIG_XEN_COMPAT_030002 */
-+ ELFNOTE(Xen, XEN_ELFNOTE_ENTRY, .long, startup_32)
-+ ELFNOTE(Xen, XEN_ELFNOTE_HYPERCALL_PAGE, .long, hypercall_page)
-+ ELFNOTE(Xen, XEN_ELFNOTE_FEATURES, .asciz, "writable_page_tables|writable_descriptor_tables|auto_translated_physmap|pae_pgdir_above_4gb|supervisor_mode_kernel")
-+#ifdef CONFIG_X86_PAE
-+ ELFNOTE(Xen, XEN_ELFNOTE_PAE_MODE, .asciz, "yes")
-+#else
-+ ELFNOTE(Xen, XEN_ELFNOTE_PAE_MODE, .asciz, "no")
-+#endif
-+ ELFNOTE(Xen, XEN_ELFNOTE_LOADER, .asciz, "generic")
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/init_task-xen.c linux-2.6.18-xen/arch/i386/kernel/init_task-xen.c
---- linux-2.6.18.3/arch/i386/kernel/init_task-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/init_task-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,51 @@
-+#include <linux/mm.h>
-+#include <linux/module.h>
-+#include <linux/sched.h>
-+#include <linux/init.h>
-+#include <linux/init_task.h>
-+#include <linux/fs.h>
-+#include <linux/mqueue.h>
-+
-+#include <asm/uaccess.h>
-+#include <asm/pgtable.h>
-+#include <asm/desc.h>
-+
-+static struct fs_struct init_fs = INIT_FS;
-+static struct files_struct init_files = INIT_FILES;
-+static struct signal_struct init_signals = INIT_SIGNALS(init_signals);
-+static struct sighand_struct init_sighand = INIT_SIGHAND(init_sighand);
-+
-+#define swapper_pg_dir ((pgd_t *)NULL)
-+struct mm_struct init_mm = INIT_MM(init_mm);
-+#undef swapper_pg_dir
-+
-+EXPORT_SYMBOL(init_mm);
-+
-+/*
-+ * Initial thread structure.
-+ *
-+ * We need to make sure that this is THREAD_SIZE aligned due to the
-+ * way process stacks are handled. This is done by having a special
-+ * "init_task" linker map entry..
-+ */
-+union thread_union init_thread_union
-+ __attribute__((__section__(".data.init_task"))) =
-+ { INIT_THREAD_INFO(init_task) };
-+
-+/*
-+ * Initial task structure.
-+ *
-+ * All other task structs will be allocated on slabs in fork.c
-+ */
-+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.
-+ */
-+DEFINE_PER_CPU(struct tss_struct, init_tss) ____cacheline_internodealigned_in_smp = INIT_TSS;
-+#endif
-+
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/io_apic-xen.c linux-2.6.18-xen/arch/i386/kernel/io_apic-xen.c
---- linux-2.6.18.3/arch/i386/kernel/io_apic-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/io_apic-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,2771 @@
-+/*
-+ * 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/compiler.h>
-+#include <linux/acpi.h>
-+#include <linux/module.h>
-+#include <linux/sysdev.h>
-+
-+#include <asm/io.h>
-+#include <asm/smp.h>
-+#include <asm/desc.h>
-+#include <asm/timer.h>
-+#include <asm/i8259.h>
-+#include <asm/nmi.h>
-+
-+#include <mach_apic.h>
-+
-+#include "io_ports.h"
-+
-+#ifdef CONFIG_XEN
-+
-+#include <xen/interface/xen.h>
-+#include <xen/interface/physdev.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;
-+ 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)
-+
-+#endif /* CONFIG_XEN */
-+
-+int (*ioapic_renumber_irq)(int ioapic, int irq);
-+atomic_t irq_mis_count;
-+
-+/* Where if anywhere is the i8259 connect in external int mode */
-+static struct { int pin, apic; } ioapic_i8259 = { -1, -1 };
-+
-+static DEFINE_SPINLOCK(ioapic_lock);
-+static DEFINE_SPINLOCK(vector_lock);
-+
-+int timer_over_8254 __initdata = 1;
-+
-+/*
-+ * Is the SiS APIC rmw bug present ?
-+ * -1 = don't know, 0 = no, 1 = yes
-+ */
-+int sis_apic_bug = -1;
-+
-+/*
-+ * # of IRQ routing registers
-+ */
-+int nr_ioapic_registers[MAX_IO_APICS];
-+
-+int disable_timer_pin_1 __initdata;
-+
-+/*
-+ * Rough estimation of how many shared IRQs there are, can
-+ * be changed anytime.
-+ */
-+#define MAX_PLUS_SHARED_IRQS NR_IRQS
-+#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 {
-+ int 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
-+
-+/*
-+ * 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;
-+
-+ 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: whoops");
-+ }
-+ entry->apic = apic;
-+ entry->pin = pin;
-+}
-+
-+#ifdef CONFIG_XEN
-+#define clear_IO_APIC() ((void)0)
-+#else
-+/*
-+ * Reroute an IRQ to a different pin.
-+ */
-+static void __init replace_pin_at_irq(unsigned int irq,
-+ int oldapic, int oldpin,
-+ int newapic, int newpin)
-+{
-+ struct irq_pin_list *entry = irq_2_pin + irq;
-+
-+ while (1) {
-+ if (entry->apic == oldapic && entry->pin == oldpin) {
-+ entry->apic = newapic;
-+ entry->pin = newpin;
-+ }
-+ if (!entry->next)
-+ break;
-+ entry = irq_2_pin + entry->next;
-+ }
-+}
-+
-+static void __modify_IO_APIC_irq (unsigned int irq, unsigned long enable, unsigned long disable)
-+{
-+ struct irq_pin_list *entry = irq_2_pin + irq;
-+ unsigned int pin, reg;
-+
-+ for (;;) {
-+ pin = entry->pin;
-+ if (pin == -1)
-+ break;
-+ reg = io_apic_read(entry->apic, 0x10 + pin*2);
-+ reg &= ~disable;
-+ reg |= enable;
-+ io_apic_modify(entry->apic, 0x10 + pin*2, reg);
-+ if (!entry->next)
-+ break;
-+ entry = irq_2_pin + entry->next;
-+ }
-+}
-+
-+/* mask = 1 */
-+static void __mask_IO_APIC_irq (unsigned int irq)
-+{
-+ __modify_IO_APIC_irq(irq, 0x00010000, 0);
-+}
-+
-+/* mask = 0 */
-+static void __unmask_IO_APIC_irq (unsigned int irq)
-+{
-+ __modify_IO_APIC_irq(irq, 0, 0x00010000);
-+}
-+
-+/* mask = 1, trigger = 0 */
-+static void __mask_and_edge_IO_APIC_irq (unsigned int irq)
-+{
-+ __modify_IO_APIC_irq(irq, 0x00010000, 0x00008000);
-+}
-+
-+/* mask = 0, trigger = 1 */
-+static void __unmask_and_level_IO_APIC_irq (unsigned int irq)
-+{
-+ __modify_IO_APIC_irq(irq, 0x00008000, 0x00010000);
-+}
-+
-+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);
-+}
-+
-+#ifdef CONFIG_SMP
-+static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t cpumask)
-+{
-+ unsigned long flags;
-+ int pin;
-+ struct irq_pin_list *entry = irq_2_pin + irq;
-+ unsigned int apicid_value;
-+ cpumask_t tmp;
-+
-+ cpus_and(tmp, cpumask, cpu_online_map);
-+ if (cpus_empty(tmp))
-+ tmp = TARGET_CPUS;
-+
-+ cpus_and(cpumask, tmp, CPU_MASK_ALL);
-+
-+ apicid_value = cpu_mask_to_apicid(cpumask);
-+ /* Prepare to do the io_apic_write */
-+ apicid_value = apicid_value << 24;
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ for (;;) {
-+ pin = entry->pin;
-+ if (pin == -1)
-+ break;
-+ io_apic_write(entry->apic, 0x10 + 1 + pin*2, apicid_value);
-+ if (!entry->next)
-+ break;
-+ entry = irq_2_pin + entry->next;
-+ }
-+ set_irq_info(irq, cpumask);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+}
-+
-+#if defined(CONFIG_IRQBALANCE)
-+# include <asm/processor.h> /* kernel_thread() */
-+# include <linux/kernel_stat.h> /* kstat */
-+# include <linux/slab.h> /* kmalloc() */
-+# include <linux/timer.h> /* time_after() */
-+
-+#ifdef CONFIG_BALANCED_IRQ_DEBUG
-+# define TDprintk(x...) do { printk("<%ld:%s:%d>: ", jiffies, __FILE__, __LINE__); printk(x); } while (0)
-+# define Dprintk(x...) do { TDprintk(x); } while (0)
-+# else
-+# define TDprintk(x...)
-+# define Dprintk(x...)
-+# endif
-+
-+#define IRQBALANCE_CHECK_ARCH -999
-+#define MAX_BALANCED_IRQ_INTERVAL (5*HZ)
-+#define MIN_BALANCED_IRQ_INTERVAL (HZ/2)
-+#define BALANCED_IRQ_MORE_DELTA (HZ/10)
-+#define BALANCED_IRQ_LESS_DELTA (HZ)
-+
-+static int irqbalance_disabled __read_mostly = IRQBALANCE_CHECK_ARCH;
-+static int physical_balance __read_mostly;
-+static long balanced_irq_interval __read_mostly = MAX_BALANCED_IRQ_INTERVAL;
-+
-+static struct irq_cpu_info {
-+ unsigned long * last_irq;
-+ unsigned long * irq_delta;
-+ unsigned long irq;
-+} irq_cpu_data[NR_CPUS];
-+
-+#define CPU_IRQ(cpu) (irq_cpu_data[cpu].irq)
-+#define LAST_CPU_IRQ(cpu,irq) (irq_cpu_data[cpu].last_irq[irq])
-+#define IRQ_DELTA(cpu,irq) (irq_cpu_data[cpu].irq_delta[irq])
-+
-+#define IDLE_ENOUGH(cpu,now) \
-+ (idle_cpu(cpu) && ((now) - per_cpu(irq_stat, (cpu)).idle_timestamp > 1))
-+
-+#define IRQ_ALLOWED(cpu, allowed_mask) cpu_isset(cpu, allowed_mask)
-+
-+#define CPU_TO_PACKAGEINDEX(i) (first_cpu(cpu_sibling_map[i]))
-+
-+static cpumask_t balance_irq_affinity[NR_IRQS] = {
-+ [0 ... NR_IRQS-1] = CPU_MASK_ALL
-+};
-+
-+void set_balance_irq_affinity(unsigned int irq, cpumask_t mask)
-+{
-+ balance_irq_affinity[irq] = mask;
-+}
-+
-+static unsigned long move(int curr_cpu, cpumask_t allowed_mask,
-+ unsigned long now, int direction)
-+{
-+ int search_idle = 1;
-+ int cpu = curr_cpu;
-+
-+ goto inside;
-+
-+ do {
-+ if (unlikely(cpu == curr_cpu))
-+ search_idle = 0;
-+inside:
-+ if (direction == 1) {
-+ cpu++;
-+ if (cpu >= NR_CPUS)
-+ cpu = 0;
-+ } else {
-+ cpu--;
-+ if (cpu == -1)
-+ cpu = NR_CPUS-1;
-+ }
-+ } while (!cpu_online(cpu) || !IRQ_ALLOWED(cpu,allowed_mask) ||
-+ (search_idle && !IDLE_ENOUGH(cpu,now)));
-+
-+ return cpu;
-+}
-+
-+static inline void balance_irq(int cpu, int irq)
-+{
-+ unsigned long now = jiffies;
-+ cpumask_t allowed_mask;
-+ unsigned int new_cpu;
-+
-+ if (irqbalance_disabled)
-+ return;
-+
-+ cpus_and(allowed_mask, cpu_online_map, balance_irq_affinity[irq]);
-+ new_cpu = move(cpu, allowed_mask, now, 1);
-+ if (cpu != new_cpu) {
-+ set_pending_irq(irq, cpumask_of_cpu(new_cpu));
-+ }
-+}
-+
-+static inline void rotate_irqs_among_cpus(unsigned long useful_load_threshold)
-+{
-+ int i, j;
-+ Dprintk("Rotating IRQs among CPUs.\n");
-+ for_each_online_cpu(i) {
-+ for (j = 0; j < NR_IRQS; j++) {
-+ if (!irq_desc[j].action)
-+ continue;
-+ /* Is it a significant load ? */
-+ if (IRQ_DELTA(CPU_TO_PACKAGEINDEX(i),j) <
-+ useful_load_threshold)
-+ continue;
-+ balance_irq(i, j);
-+ }
-+ }
-+ balanced_irq_interval = max((long)MIN_BALANCED_IRQ_INTERVAL,
-+ balanced_irq_interval - BALANCED_IRQ_LESS_DELTA);
-+ return;
-+}
-+
-+static void do_irq_balance(void)
-+{
-+ int i, j;
-+ unsigned long max_cpu_irq = 0, min_cpu_irq = (~0);
-+ unsigned long move_this_load = 0;
-+ int max_loaded = 0, min_loaded = 0;
-+ int load;
-+ unsigned long useful_load_threshold = balanced_irq_interval + 10;
-+ int selected_irq;
-+ int tmp_loaded, first_attempt = 1;
-+ unsigned long tmp_cpu_irq;
-+ unsigned long imbalance = 0;
-+ cpumask_t allowed_mask, target_cpu_mask, tmp;
-+
-+ for_each_possible_cpu(i) {
-+ int package_index;
-+ CPU_IRQ(i) = 0;
-+ if (!cpu_online(i))
-+ continue;
-+ package_index = CPU_TO_PACKAGEINDEX(i);
-+ for (j = 0; j < NR_IRQS; j++) {
-+ unsigned long value_now, delta;
-+ /* Is this an active IRQ? */
-+ if (!irq_desc[j].action)
-+ continue;
-+ if ( package_index == i )
-+ IRQ_DELTA(package_index,j) = 0;
-+ /* Determine the total count per processor per IRQ */
-+ value_now = (unsigned long) kstat_cpu(i).irqs[j];
-+
-+ /* Determine the activity per processor per IRQ */
-+ delta = value_now - LAST_CPU_IRQ(i,j);
-+
-+ /* Update last_cpu_irq[][] for the next time */
-+ LAST_CPU_IRQ(i,j) = value_now;
-+
-+ /* Ignore IRQs whose rate is less than the clock */
-+ if (delta < useful_load_threshold)
-+ continue;
-+ /* update the load for the processor or package total */
-+ IRQ_DELTA(package_index,j) += delta;
-+
-+ /* Keep track of the higher numbered sibling as well */
-+ if (i != package_index)
-+ CPU_IRQ(i) += delta;
-+ /*
-+ * We have sibling A and sibling B in the package
-+ *
-+ * cpu_irq[A] = load for cpu A + load for cpu B
-+ * cpu_irq[B] = load for cpu B
-+ */
-+ CPU_IRQ(package_index) += delta;
-+ }
-+ }
-+ /* Find the least loaded processor package */
-+ for_each_online_cpu(i) {
-+ if (i != CPU_TO_PACKAGEINDEX(i))
-+ continue;
-+ if (min_cpu_irq > CPU_IRQ(i)) {
-+ min_cpu_irq = CPU_IRQ(i);
-+ min_loaded = i;
-+ }
-+ }
-+ max_cpu_irq = ULONG_MAX;
-+
-+tryanothercpu:
-+ /* Look for heaviest loaded processor.
-+ * We may come back to get the next heaviest loaded processor.
-+ * Skip processors with trivial loads.
-+ */
-+ tmp_cpu_irq = 0;
-+ tmp_loaded = -1;
-+ for_each_online_cpu(i) {
-+ if (i != CPU_TO_PACKAGEINDEX(i))
-+ continue;
-+ if (max_cpu_irq <= CPU_IRQ(i))
-+ continue;
-+ if (tmp_cpu_irq < CPU_IRQ(i)) {
-+ tmp_cpu_irq = CPU_IRQ(i);
-+ tmp_loaded = i;
-+ }
-+ }
-+
-+ if (tmp_loaded == -1) {
-+ /* In the case of small number of heavy interrupt sources,
-+ * loading some of the cpus too much. We use Ingo's original
-+ * approach to rotate them around.
-+ */
-+ if (!first_attempt && imbalance >= useful_load_threshold) {
-+ rotate_irqs_among_cpus(useful_load_threshold);
-+ return;
-+ }
-+ goto not_worth_the_effort;
-+ }
-+
-+ first_attempt = 0; /* heaviest search */
-+ max_cpu_irq = tmp_cpu_irq; /* load */
-+ max_loaded = tmp_loaded; /* processor */
-+ imbalance = (max_cpu_irq - min_cpu_irq) / 2;
-+
-+ Dprintk("max_loaded cpu = %d\n", max_loaded);
-+ Dprintk("min_loaded cpu = %d\n", min_loaded);
-+ Dprintk("max_cpu_irq load = %ld\n", max_cpu_irq);
-+ Dprintk("min_cpu_irq load = %ld\n", min_cpu_irq);
-+ Dprintk("load imbalance = %lu\n", imbalance);
-+
-+ /* if imbalance is less than approx 10% of max load, then
-+ * observe diminishing returns action. - quit
-+ */
-+ if (imbalance < (max_cpu_irq >> 3)) {
-+ Dprintk("Imbalance too trivial\n");
-+ goto not_worth_the_effort;
-+ }
-+
-+tryanotherirq:
-+ /* if we select an IRQ to move that can't go where we want, then
-+ * see if there is another one to try.
-+ */
-+ move_this_load = 0;
-+ selected_irq = -1;
-+ for (j = 0; j < NR_IRQS; j++) {
-+ /* Is this an active IRQ? */
-+ if (!irq_desc[j].action)
-+ continue;
-+ if (imbalance <= IRQ_DELTA(max_loaded,j))
-+ continue;
-+ /* Try to find the IRQ that is closest to the imbalance
-+ * without going over.
-+ */
-+ if (move_this_load < IRQ_DELTA(max_loaded,j)) {
-+ move_this_load = IRQ_DELTA(max_loaded,j);
-+ selected_irq = j;
-+ }
-+ }
-+ if (selected_irq == -1) {
-+ goto tryanothercpu;
-+ }
-+
-+ imbalance = move_this_load;
-+
-+ /* For physical_balance case, we accumlated both load
-+ * values in the one of the siblings cpu_irq[],
-+ * to use the same code for physical and logical processors
-+ * as much as possible.
-+ *
-+ * NOTE: the cpu_irq[] array holds the sum of the load for
-+ * sibling A and sibling B in the slot for the lowest numbered
-+ * sibling (A), _AND_ the load for sibling B in the slot for
-+ * the higher numbered sibling.
-+ *
-+ * We seek the least loaded sibling by making the comparison
-+ * (A+B)/2 vs B
-+ */
-+ load = CPU_IRQ(min_loaded) >> 1;
-+ for_each_cpu_mask(j, cpu_sibling_map[min_loaded]) {
-+ if (load > CPU_IRQ(j)) {
-+ /* This won't change cpu_sibling_map[min_loaded] */
-+ load = CPU_IRQ(j);
-+ min_loaded = j;
-+ }
-+ }
-+
-+ cpus_and(allowed_mask,
-+ cpu_online_map,
-+ balance_irq_affinity[selected_irq]);
-+ target_cpu_mask = cpumask_of_cpu(min_loaded);
-+ cpus_and(tmp, target_cpu_mask, allowed_mask);
-+
-+ if (!cpus_empty(tmp)) {
-+
-+ Dprintk("irq = %d moved to cpu = %d\n",
-+ selected_irq, min_loaded);
-+ /* mark for change destination */
-+ set_pending_irq(selected_irq, cpumask_of_cpu(min_loaded));
-+
-+ /* Since we made a change, come back sooner to
-+ * check for more variation.
-+ */
-+ balanced_irq_interval = max((long)MIN_BALANCED_IRQ_INTERVAL,
-+ balanced_irq_interval - BALANCED_IRQ_LESS_DELTA);
-+ return;
-+ }
-+ goto tryanotherirq;
-+
-+not_worth_the_effort:
-+ /*
-+ * if we did not find an IRQ to move, then adjust the time interval
-+ * upward
-+ */
-+ balanced_irq_interval = min((long)MAX_BALANCED_IRQ_INTERVAL,
-+ balanced_irq_interval + BALANCED_IRQ_MORE_DELTA);
-+ Dprintk("IRQ worth rotating not found\n");
-+ return;
-+}
-+
-+static int balanced_irq(void *unused)
-+{
-+ int i;
-+ unsigned long prev_balance_time = jiffies;
-+ long time_remaining = balanced_irq_interval;
-+
-+ daemonize("kirqd");
-+
-+ /* push everything to CPU 0 to give us a starting point. */
-+ for (i = 0 ; i < NR_IRQS ; i++) {
-+ irq_desc[i].pending_mask[i] = cpumask_of_cpu(0);
-+ set_pending_irq(i, cpumask_of_cpu(0));
-+ }
-+
-+ for ( ; ; ) {
-+ time_remaining = schedule_timeout_interruptible(time_remaining);
-+ try_to_freeze();
-+ if (time_after(jiffies,
-+ prev_balance_time+balanced_irq_interval)) {
-+ preempt_disable();
-+ do_irq_balance();
-+ prev_balance_time = jiffies;
-+ time_remaining = balanced_irq_interval;
-+ preempt_enable();
-+ }
-+ }
-+ return 0;
-+}
-+
-+static int __init balanced_irq_init(void)
-+{
-+ int i;
-+ struct cpuinfo_x86 *c;
-+ cpumask_t tmp;
-+
-+ cpus_shift_right(tmp, cpu_online_map, 2);
-+ c = &boot_cpu_data;
-+ /* When not overwritten by the command line ask subarchitecture. */
-+ if (irqbalance_disabled == IRQBALANCE_CHECK_ARCH)
-+ irqbalance_disabled = NO_BALANCE_IRQ;
-+ if (irqbalance_disabled)
-+ return 0;
-+
-+ /* disable irqbalance completely if there is only one processor online */
-+ if (num_online_cpus() < 2) {
-+ irqbalance_disabled = 1;
-+ return 0;
-+ }
-+ /*
-+ * Enable physical balance only if more than 1 physical processor
-+ * is present
-+ */
-+ if (smp_num_siblings > 1 && !cpus_empty(tmp))
-+ physical_balance = 1;
-+
-+ for_each_online_cpu(i) {
-+ irq_cpu_data[i].irq_delta = kmalloc(sizeof(unsigned long) * NR_IRQS, GFP_KERNEL);
-+ irq_cpu_data[i].last_irq = kmalloc(sizeof(unsigned long) * NR_IRQS, GFP_KERNEL);
-+ if (irq_cpu_data[i].irq_delta == NULL || irq_cpu_data[i].last_irq == NULL) {
-+ printk(KERN_ERR "balanced_irq_init: out of memory");
-+ goto failed;
-+ }
-+ memset(irq_cpu_data[i].irq_delta,0,sizeof(unsigned long) * NR_IRQS);
-+ memset(irq_cpu_data[i].last_irq,0,sizeof(unsigned long) * NR_IRQS);
-+ }
-+
-+ printk(KERN_INFO "Starting balanced_irq\n");
-+ if (kernel_thread(balanced_irq, NULL, CLONE_KERNEL) >= 0)
-+ return 0;
-+ else
-+ printk(KERN_ERR "balanced_irq_init: failed to spawn balanced_irq");
-+failed:
-+ for_each_possible_cpu(i) {
-+ kfree(irq_cpu_data[i].irq_delta);
-+ irq_cpu_data[i].irq_delta = NULL;
-+ kfree(irq_cpu_data[i].last_irq);
-+ irq_cpu_data[i].last_irq = NULL;
-+ }
-+ return 0;
-+}
-+
-+int __init irqbalance_disable(char *str)
-+{
-+ irqbalance_disabled = 1;
-+ return 1;
-+}
-+
-+__setup("noirqbalance", irqbalance_disable);
-+
-+late_initcall(balanced_irq_init);
-+#endif /* CONFIG_IRQBALANCE */
-+#endif /* CONFIG_SMP */
-+#endif
-+
-+#ifndef CONFIG_SMP
-+void fastcall send_IPI_self(int vector)
-+{
-+#ifndef CONFIG_XEN
-+ unsigned int cfg;
-+
-+ /*
-+ * Wait for idle.
-+ */
-+ apic_wait_icr_idle();
-+ cfg = APIC_DM_FIXED | APIC_DEST_SELF | vector | APIC_DEST_LOGICAL;
-+ /*
-+ * Send the IPI. The write to APIC_ICR fires this off.
-+ */
-+ apic_write_around(APIC_ICR, cfg);
-+#endif
-+}
-+#endif /* !CONFIG_SMP */
-+
-+
-+/*
-+ * 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;
-+
-+static int __init ioapic_setup(char *str)
-+{
-+ skip_ioapic_setup = 1;
-+ return 1;
-+}
-+
-+__setup("noapic", ioapic_setup);
-+
-+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, KERN_INFO
-+ "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, KERN_DEBUG
-+ "... 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;
-+}
-+
-+/*
-+ * 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_bus_id_to_type[lbus] == MP_BUS_NEC98
-+ ) &&
-+ (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_bus_id_to_type[lbus] == MP_BUS_NEC98
-+ ) &&
-+ (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;
-+}
-+
-+/*
-+ * 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) {
-+ printk(KERN_WARNING "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;
-+ }
-+ }
-+ return best_guess;
-+}
-+EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector);
-+
-+/*
-+ * 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
-+#ifndef CONFIG_XEN
-+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 /* !CONFIG_XEN */
-+#endif
-+
-+/*
-+ * 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, KERN_INFO
-+ "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)
-+
-+/* NEC98 interrupts are always polarity zero edge triggered,
-+ * when listed as conforming in the MP table. */
-+
-+#define default_NEC98_trigger(idx) (0)
-+#define default_NEC98_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;
-+ }
-+ case MP_BUS_NEC98: /* NEC 98 pin */
-+ {
-+ polarity = default_NEC98_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;
-+ }
-+ case MP_BUS_NEC98: /* NEC 98 pin */
-+ {
-+ trigger = default_NEC98_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 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:
-+ case MP_BUS_NEC98:
-+ {
-+ 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;
-+
-+ /*
-+ * For MPS mode, so far only needed by ES7000 platform
-+ */
-+ if (ioapic_renumber_irq)
-+ irq = ioapic_renumber_irq(apic, irq);
-+
-+ break;
-+ }
-+ default:
-+ {
-+ printk(KERN_ERR "unknown bus type %d.\n",bus);
-+ irq = 0;
-+ break;
-+ }
-+ }
-+
-+ /*
-+ * 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, KERN_DEBUG
-+ "disabling PIRQ%d\n", pin-16);
-+ } else {
-+ irq = pirq_entries[pin-16];
-+ apic_printk(APIC_VERBOSE, KERN_DEBUG
-+ "using PIRQ%d -> IRQ %d\n",
-+ pin-16, irq);
-+ }
-+ }
-+ }
-+ 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; /* = { FIRST_DEVICE_VECTOR , 0 }; */
-+
-+int assign_irq_vector(int irq)
-+{
-+ struct physdev_irq irq_op;
-+ unsigned long flags;
-+
-+ 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))
-+ return -ENOSPC;
-+
-+ vector_irq[irq_op.vector] = irq;
-+ if (irq != AUTO_ASSIGN)
-+ IO_APIC_VECTOR(irq) = irq_op.vector;
-+
-+ spin_unlock_irqrestore(&vector_lock, flags);
-+
-+ return irq_op.vector;
-+}
-+
-+#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) ((void)0)
-+#endif
-+
-+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;
-+ }
-+
-+ irq = pin_2_irq(idx, apic, pin);
-+ /*
-+ * skip adding the timer int on secondary nodes, which causes
-+ * a small but painful rift in the time-space continuum
-+ */
-+ if (multi_timer_check(apic, irq))
-+ continue;
-+ else
-+ 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");
-+}
-+
-+/*
-+ * Set up the 8259A-master output pin:
-+ */
-+#ifndef CONFIG_XEN
-+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_around(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);
-+}
-+
-+static inline void UNEXPECTED_IO_APIC(void)
-+{
-+}
-+
-+void __init 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;
-+ union IO_APIC_reg_03 reg_03;
-+ 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);
-+ if (reg_01.bits.version >= 0x20)
-+ reg_03.raw = io_apic_read(apic, 3);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
-+
-+ 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);
-+ printk(KERN_DEBUG "....... : Delivery Type: %X\n", reg_00.bits.delivery_type);
-+ printk(KERN_DEBUG "....... : LTS : %X\n", reg_00.bits.LTS);
-+ if (reg_00.bits.ID >= get_physical_broadcast())
-+ UNEXPECTED_IO_APIC();
-+ if (reg_00.bits.__reserved_1 || reg_00.bits.__reserved_2)
-+ UNEXPECTED_IO_APIC();
-+
-+ printk(KERN_DEBUG ".... register #01: %08X\n", reg_01.raw);
-+ 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)
-+ )
-+ 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 != 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();
-+
-+ /*
-+ * Some Intel chipsets with IO APIC VERSION of 0x1? don't have reg_02,
-+ * but the value of reg_02 is read as the previous read register
-+ * value, so ignore it if reg_02 == reg_01.
-+ */
-+ if (reg_01.bits.version >= 0x10 && reg_02.raw != reg_01.raw) {
-+ 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();
-+ }
-+
-+ /*
-+ * Some Intel chipsets with IO APIC VERSION of 0x2? don't have reg_02
-+ * or reg_03, but the value of reg_0[23] is read as the previous read
-+ * register value, so ignore it if reg_03 == reg_0[12].
-+ */
-+ if (reg_01.bits.version >= 0x20 && reg_03.raw != reg_02.raw &&
-+ reg_03.raw != reg_01.raw) {
-+ printk(KERN_DEBUG ".... register #03: %08X\n", reg_03.raw);
-+ printk(KERN_DEBUG "....... : Boot DT : %X\n", reg_03.bits.boot_DT);
-+ if (reg_03.bits.__reserved_1)
-+ 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;
-+}
-+
-+#if 0
-+
-+static 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 /*__init*/ 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);
-+
-+ if (APIC_INTEGRATED(ver)) { /* !82489DX */
-+ 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);
-+
-+ if (APIC_INTEGRATED(ver)) { /* !82489DX */
-+ if (maxlvt > 3) /* Due to the Pentium erratum 3AP. */
-+ apic_write(APIC_ESR, 0);
-+ 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 /*__init*/ 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 /* 0 */
-+
-+#else
-+void __init print_IO_APIC(void) { }
-+#endif /* !CONFIG_XEN */
-+
-+static void __init enable_IO_APIC(void)
-+{
-+ union IO_APIC_reg_01 reg_01;
-+ int i8259_apic, i8259_pin;
-+ 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;
-+ }
-+ 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 */
-+ /* If we could not find the appropriate pin by looking at the ioapic
-+ * the i8259 probably is not connected the ioapic but give the
-+ * mptable a chance anyway.
-+ */
-+ 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");
-+ }
-+
-+ /*
-+ * 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
-+ */
-+
-+#if !defined(CONFIG_XEN) && !defined(CONFIG_X86_NUMAQ)
-+static void __init setup_ioapic_ids_from_mpc(void)
-+{
-+ union IO_APIC_reg_00 reg_00;
-+ physid_mask_t phys_id_present_map;
-+ int apic;
-+ int i;
-+ unsigned char old_id;
-+ unsigned long flags;
-+
-+ /*
-+ * Don't check I/O APIC IDs for xAPIC systems. They have
-+ * no meaning without the serial APIC bus.
-+ */
-+ if (!(boot_cpu_data.x86_vendor == X86_VENDOR_INTEL)
-+ || APIC_XAPIC(apic_version[boot_cpu_physical_apicid]))
-+ return;
-+ /*
-+ * This is broken; anything with a real cpu count has to
-+ * circumvent this idiocy regardless.
-+ */
-+ phys_id_present_map = ioapic_phys_id_map(phys_cpu_present_map);
-+
-+ /*
-+ * 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;
-+
-+ if (mp_ioapics[apic].mpc_apicid >= get_physical_broadcast()) {
-+ printk(KERN_ERR "BIOS bug, IO-APIC#%d ID is %d in the MPC table!...\n",
-+ apic, mp_ioapics[apic].mpc_apicid);
-+ printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n",
-+ reg_00.bits.ID);
-+ mp_ioapics[apic].mpc_apicid = reg_00.bits.ID;
-+ }
-+
-+ /*
-+ * Sanity check, is the ID really free? Every APIC in a
-+ * system must have a unique ID or we get lots of nice
-+ * 'stuck on smp_invalidate_needed IPI wait' messages.
-+ */
-+ if (check_apicid_used(phys_id_present_map,
-+ mp_ioapics[apic].mpc_apicid)) {
-+ printk(KERN_ERR "BIOS bug, IO-APIC#%d ID %d is already used!...\n",
-+ apic, mp_ioapics[apic].mpc_apicid);
-+ for (i = 0; i < get_physical_broadcast(); i++)
-+ if (!physid_isset(i, phys_id_present_map))
-+ break;
-+ if (i >= get_physical_broadcast())
-+ panic("Max APIC ID exceeded!\n");
-+ printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n",
-+ i);
-+ physid_set(i, phys_id_present_map);
-+ mp_ioapics[apic].mpc_apicid = i;
-+ } else {
-+ physid_mask_t tmp;
-+ tmp = apicid_to_cpu_present(mp_ioapics[apic].mpc_apicid);
-+ apic_printk(APIC_VERBOSE, "Setting %d in the "
-+ "phys_id_present_map\n",
-+ mp_ioapics[apic].mpc_apicid);
-+ physids_or(phys_id_present_map, phys_id_present_map, tmp);
-+ }
-+
-+
-+ /*
-+ * 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
-+
-+#ifndef CONFIG_XEN
-+/*
-+ * 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
-+ */
-+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.
-+ */
-+ 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)
-+{
-+ unsigned long v;
-+ int i;
-+
-+ move_irq(irq);
-+/*
-+ * It appears there is an erratum which affects at least version 0x11
-+ * of I/O APIC (that's the 82093AA and cores integrated into various
-+ * chipsets). Under certain conditions a level-triggered interrupt is
-+ * erroneously delivered as edge-triggered one but the respective IRR
-+ * bit gets set nevertheless. As a result the I/O unit expects an EOI
-+ * message but it will never arrive and further interrupts are blocked
-+ * from the source. The exact reason is so far unknown, but the
-+ * phenomenon was observed when two consecutive interrupt requests
-+ * from a given source get delivered to the same CPU and the source is
-+ * temporarily disabled in between.
-+ *
-+ * A workaround is to simulate an EOI message manually. We achieve it
-+ * by setting the trigger mode to edge and then to level when the edge
-+ * trigger mode gets detected in the TMR of a local APIC for a
-+ * level-triggered interrupt. We mask the source for the time of the
-+ * operation to prevent an edge-triggered interrupt escaping meanwhile.
-+ * The idea is from Manfred Spraul. --macro
-+ */
-+ i = IO_APIC_VECTOR(irq);
-+
-+ v = apic_read(APIC_TMR + ((i & ~0x1f) >> 1));
-+
-+ ack_APIC_irq();
-+
-+ if (!(v & (1 << (i & 0x1f)))) {
-+ atomic_inc(&irq_mis_count);
-+ spin_lock(&ioapic_lock);
-+ __mask_and_edge_IO_APIC_irq(irq);
-+ __unmask_and_level_IO_APIC_irq(irq);
-+ spin_unlock(&ioapic_lock);
-+ }
-+}
-+
-+#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
-+#endif
-+
-+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
-+ }
-+ }
-+}
-+
-+int timer_uses_ioapic_pin_0;
-+
-+#ifndef CONFIG_XEN
-+static void enable_lapic_irq (unsigned int irq)
-+{
-+ unsigned long v;
-+
-+ v = apic_read(APIC_LVT0);
-+ apic_write_around(APIC_LVT0, v & ~APIC_LVT_MASKED);
-+}
-+
-+static void disable_lapic_irq (unsigned int irq)
-+{
-+ unsigned long v;
-+
-+ v = apic_read(APIC_LVT0);
-+ apic_write_around(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.
-+ */
-+ apic_printk(APIC_VERBOSE, KERN_INFO "activating NMI Watchdog ...");
-+
-+ on_each_cpu(enable_NMI_through_LVT0, NULL, 1, 1);
-+
-+ apic_printk(APIC_VERBOSE, " 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);
-+}
-+
-+/*
-+ * 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.
-+ */
-+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_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
-+ init_8259A(1);
-+ timer_ack = 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;
-+
-+ printk(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 (timer_irq_works()) {
-+ 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);
-+ printk(KERN_ERR "..MP-BIOS bug: 8254 timer not connected to "
-+ "IO-APIC\n");
-+ }
-+
-+ printk(KERN_INFO "...trying to set up timer (IRQ0) through the 8259A ... ");
-+ if (pin2 != -1) {
-+ printk("\n..... (found pin %d) ...", pin2);
-+ /*
-+ * legacy devices should be connected to IO APIC #0
-+ */
-+ setup_ExtINT_IRQ0_pin(apic2, pin2, vector);
-+ if (timer_irq_works()) {
-+ printk("works.\n");
-+ if (pin1 != -1)
-+ replace_pin_at_irq(0, apic1, pin1, apic2, pin2);
-+ else
-+ add_pin_to_irq(0, apic2, pin2);
-+ if (nmi_watchdog == NMI_IO_APIC) {
-+ setup_nmi();
-+ }
-+ return;
-+ }
-+ /*
-+ * Cleanup, just in case ...
-+ */
-+ clear_IO_APIC_pin(apic2, pin2);
-+ }
-+ printk(" 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;
-+ }
-+
-+ printk(KERN_INFO "...trying to set up timer as Virtual Wire IRQ...");
++ # interrupted in sysexit critical
++ addl $0x38,%esp # Remove cs...ebx from stack frame.
++ # this popped off new frame to reuse the old one, therefore no
++ # CFI_ADJUST_CFA_OFFSET here
++11: push %esp
++ CFI_ADJUST_CFA_OFFSET 4
++ call evtchn_do_upcall
++ add $4,%esp
++ CFI_ADJUST_CFA_OFFSET -4
++ jmp ret_from_intr
+
-+ disable_8259A_irq(0);
-+ irq_desc[0].chip = &lapic_irq_type;
-+ apic_write_around(APIC_LVT0, APIC_DM_FIXED | vector); /* Fixed mode */
-+ enable_8259A_irq(0);
++ ALIGN
++restore_all_enable_events:
++ __ENABLE_INTERRUPTS
++scrit: /**** START OF CRITICAL REGION ****/
++ __TEST_PENDING
++ jnz 14f # process more events if necessary...
++ RESTORE_REGS
++ addl $4, %esp
++ CFI_ADJUST_CFA_OFFSET -4
++1: iret
++.section __ex_table,"a"
++ .align 4
++ .long 1b,iret_exc
++.previous
++14: __DISABLE_INTERRUPTS
++ jmp 11b
++ecrit: /**** END OF CRITICAL REGION ****/
++# [How we do the fixup]. We want to merge the current stack frame with the
++# just-interrupted frame. How we do this depends on where in the critical
++# region the interrupted handler was executing, and so how many saved
++# registers are in each frame. We do this quickly using the lookup table
++# 'critical_fixup_table'. For each byte offset in the critical region, it
++# provides the number of bytes which have already been popped from the
++# interrupted stack frame.
++critical_region_fixup:
++ addl $critical_fixup_table-scrit,%eax
++ movzbl (%eax),%eax # %eax contains num bytes popped
++ cmpb $0xff,%al # 0xff => vcpu_info critical region
++ jne 15f
++ GET_THREAD_INFO(%ebp)
++ xorl %eax,%eax
++15: mov %esp,%esi
++ add %eax,%esi # %esi points at end of src region
++ mov %esp,%edi
++ add $0x38,%edi # %edi points at end of dst region
++ mov %eax,%ecx
++ shr $2,%ecx # convert words to bytes
++ je 17f # skip loop if nothing to copy
++16: subl $4,%esi # pre-decrementing copy loop
++ subl $4,%edi
++ movl (%esi),%eax
++ movl %eax,(%edi)
++ loop 16b
++17: movl %edi,%esp # final %edi is top of merged stack
++ # this popped off new frame to reuse the old one, therefore no
++ # CFI_DEF_CFA_OFFSET here
++ jmp 11b
++ CFI_ENDPROC
+
-+ if (timer_irq_works()) {
-+ printk(" works.\n");
-+ return;
-+ }
-+ apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | vector);
-+ printk(" failed.\n");
++critical_fixup_table:
++ .byte 0xff,0xff,0xff # testb $0xff,(%esi) = __TEST_PENDING
++ .byte 0xff,0xff # jnz 14f
++ .byte 0x00 # pop %ebx
++ .byte 0x04 # pop %ecx
++ .byte 0x08 # pop %edx
++ .byte 0x0c # pop %esi
++ .byte 0x10 # pop %edi
++ .byte 0x14 # pop %ebp
++ .byte 0x18 # pop %eax
++ .byte 0x1c # pop %ds
++ .byte 0x20 # pop %es
++ .byte 0x24 # pop %gs
++ .byte 0x28,0x28,0x28 # add $4,%esp
++ .byte 0x2c # iret
++ .byte 0xff,0xff,0xff,0xff # movb $1,1(%esi)
++ .byte 0x00,0x00 # jmp 11b
+
-+ printk(KERN_INFO "...trying to set up timer as ExtINT IRQ...");
++# 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 fix up by reattempting the load, and zeroing the segment
++# register if the load fails.
++# Category 2 we fix up by jumping to do_iret_error. 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 maintaining a status value in EAX.
++ENTRY(failsafe_callback)
++ RING0_INT_FRAME
++ pushl %eax
++ CFI_ADJUST_CFA_OFFSET 4
++ movl $1,%eax
++1: mov 4(%esp),%ds
++2: mov 8(%esp),%es
++3: mov 12(%esp),%fs
++4: mov 16(%esp),%gs
++ testl %eax,%eax
++ popl %eax
++ CFI_ADJUST_CFA_OFFSET -4
++ jz 5f
++ addl $16,%esp # EAX != 0 => Category 2 (Bad IRET)
++ CFI_ADJUST_CFA_OFFSET -16
++ jmp iret_exc
++ CFI_ADJUST_CFA_OFFSET 16
++5: addl $16,%esp # EAX == 0 => Category 1 (Bad segment)
++ CFI_ADJUST_CFA_OFFSET -16
++ pushl $0
++ CFI_ADJUST_CFA_OFFSET 4
++ SAVE_ALL
++ jmp ret_from_exception
++.section .fixup,"ax"; \
++6: xorl %eax,%eax; \
++ movl %eax,4(%esp); \
++ jmp 1b; \
++7: xorl %eax,%eax; \
++ movl %eax,8(%esp); \
++ jmp 2b; \
++8: xorl %eax,%eax; \
++ movl %eax,12(%esp); \
++ jmp 3b; \
++9: xorl %eax,%eax; \
++ movl %eax,16(%esp); \
++ jmp 4b; \
++.previous; \
++.section __ex_table,"a"; \
++ .align 4; \
++ .long 1b,6b; \
++ .long 2b,7b; \
++ .long 3b,8b; \
++ .long 4b,9b; \
++.previous
++ CFI_ENDPROC
++#endif
+
-+ timer_ack = 0;
-+ init_8259A(0);
-+ make_8259A_irq(0);
-+ apic_write_around(APIC_LVT0, APIC_DM_EXTINT);
++ENTRY(coprocessor_error)
++ RING0_INT_FRAME
++ pushl $0
++ CFI_ADJUST_CFA_OFFSET 4
++ pushl $do_coprocessor_error
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
+
-+ unlock_ExtINT_logic();
++ENTRY(simd_coprocessor_error)
++ RING0_INT_FRAME
++ pushl $0
++ CFI_ADJUST_CFA_OFFSET 4
++ pushl $do_simd_coprocessor_error
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
+
-+ if (timer_irq_works()) {
-+ printk(" works.\n");
-+ return;
-+ }
-+ printk(" failed :(.\n");
-+ panic("IO-APIC + timer doesn't work! Boot with apic=debug and send a "
-+ "report. Then try booting with the 'noapic' option");
-+}
-+#else
-+#define check_timer() ((void)0)
++ENTRY(device_not_available)
++ RING0_INT_FRAME
++ pushl $-1 # mark this as an int
++ CFI_ADJUST_CFA_OFFSET 4
++ SAVE_ALL
++#ifndef CONFIG_XEN
++ GET_CR0_INTO_EAX
++ testl $0x4, %eax # EM (math emulation bit)
++ je device_available_emulate
++ pushl $0 # temporary storage for ORIG_EIP
++ CFI_ADJUST_CFA_OFFSET 4
++ call math_emulate
++ addl $4, %esp
++ CFI_ADJUST_CFA_OFFSET -4
++ jmp ret_from_exception
++device_available_emulate:
+#endif
++ preempt_stop(CLBR_ANY)
++ call math_state_restore
++ jmp ret_from_exception
++ CFI_ENDPROC
+
++#ifndef CONFIG_XEN
+/*
++ * Debug traps and NMI can happen at the one SYSENTER instruction
++ * that sets up the real kernel stack. Check here, since we can't
++ * allow the wrong stack to be used.
+ *
-+ * 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.
++ * "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.
++ *
++ * We just load the right stack, and push the three (known) values
++ * by hand onto the new stack - while updating the return eip past
++ * the instruction that would have done it for sysenter.
+ */
-+#define PIC_IRQS (1 << PIC_CASCADE_IR)
-+
-+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;
-+
-+ printk("ENABLING IO-APIC IRQs\n");
++#define FIX_STACK(offset, ok, label) \
++ cmpw $__KERNEL_CS,4(%esp); \
++ jne ok; \
++label: \
++ movl SYSENTER_stack_esp0+offset(%esp),%esp; \
++ CFI_DEF_CFA esp, 0; \
++ CFI_UNDEFINED eip; \
++ pushfl; \
++ CFI_ADJUST_CFA_OFFSET 4; \
++ pushl $__KERNEL_CS; \
++ CFI_ADJUST_CFA_OFFSET 4; \
++ pushl $sysenter_past_esp; \
++ CFI_ADJUST_CFA_OFFSET 4; \
++ CFI_REL_OFFSET eip, 0
++#endif /* CONFIG_XEN */
+
-+ /*
-+ * Set up IO-APIC IRQ routing.
-+ */
-+ if (!acpi_ioapic)
-+ setup_ioapic_ids_from_mpc();
++KPROBE_ENTRY(debug)
++ RING0_INT_FRAME
+#ifndef CONFIG_XEN
-+ sync_Arb_IDs();
-+#endif
-+ setup_IO_APIC_irqs();
-+ init_IO_APIC_traps();
-+ check_timer();
-+ if (!acpi_ioapic)
-+ print_IO_APIC();
-+}
-+
-+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);
++ cmpl $sysenter_entry,(%esp)
++ jne debug_stack_correct
++ FIX_STACK(12, debug_stack_correct, debug_esp_fix_insn)
++debug_stack_correct:
++#endif /* !CONFIG_XEN */
++ pushl $-1 # mark this as an int
++ CFI_ADJUST_CFA_OFFSET 4
++ SAVE_ALL
++ xorl %edx,%edx # error code 0
++ movl %esp,%eax # pt_regs pointer
++ call do_debug
++ jmp ret_from_exception
++ CFI_ENDPROC
++KPROBE_END(debug)
+
++#ifndef CONFIG_XEN
+/*
-+ * Called after all the initialization is done. If we didnt find any
-+ * APIC bugs then we can allow the modify fast path
++ * NMI is doubly nasty. It can happen _while_ we're handling
++ * a debug fault, and the debug fault hasn't yet been able to
++ * clear up the stack. So we first check whether we got an
++ * NMI on the sysenter entry path, but after that we need to
++ * check whether we got an NMI on the debug path where the debug
++ * fault happened on the sysenter path.
+ */
-+
-+static int __init io_apic_bug_finalize(void)
-+{
-+ if(sis_apic_bug == -1)
-+ sis_apic_bug = 0;
-+ if (is_initial_xendomain()) {
-+ dom0_op_t op = { .cmd = DOM0_PLATFORM_QUIRK };
-+ op.u.platform_quirk.quirk_id = sis_apic_bug ?
-+ QUIRK_IOAPIC_BAD_REGSEL : QUIRK_IOAPIC_GOOD_REGSEL;
-+ HYPERVISOR_dom0_op(&op);
-+ }
-+ return 0;
-+}
++KPROBE_ENTRY(nmi)
++ RING0_INT_FRAME
++ pushl %eax
++ CFI_ADJUST_CFA_OFFSET 4
++ movl %ss, %eax
++ cmpw $__ESPFIX_SS, %ax
++ popl %eax
++ CFI_ADJUST_CFA_OFFSET -4
++ je nmi_espfix_stack
++ cmpl $sysenter_entry,(%esp)
++ je nmi_stack_fixup
++ pushl %eax
++ CFI_ADJUST_CFA_OFFSET 4
++ movl %esp,%eax
++ /* Do not access memory above the end of our stack page,
++ * it might not exist.
++ */
++ andl $(THREAD_SIZE-1),%eax
++ cmpl $(THREAD_SIZE-20),%eax
++ popl %eax
++ CFI_ADJUST_CFA_OFFSET -4
++ jae nmi_stack_correct
++ cmpl $sysenter_entry,12(%esp)
++ je nmi_debug_stack_check
++nmi_stack_correct:
++ /* We have a RING0_INT_FRAME here */
++ pushl %eax
++ CFI_ADJUST_CFA_OFFSET 4
++ SAVE_ALL
++ xorl %edx,%edx # zero error code
++ movl %esp,%eax # pt_regs pointer
++ call do_nmi
++ jmp restore_nocheck_notrace
++ CFI_ENDPROC
+
-+late_initcall(io_apic_bug_finalize);
++nmi_stack_fixup:
++ RING0_INT_FRAME
++ FIX_STACK(12,nmi_stack_correct, 1)
++ jmp nmi_stack_correct
+
-+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];
++nmi_debug_stack_check:
++ /* We have a RING0_INT_FRAME here */
++ cmpw $__KERNEL_CS,16(%esp)
++ jne nmi_stack_correct
++ cmpl $debug,(%esp)
++ jb nmi_stack_correct
++ cmpl $debug_esp_fix_insn,(%esp)
++ ja nmi_stack_correct
++ FIX_STACK(24,nmi_stack_correct, 1)
++ jmp nmi_stack_correct
+
-+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);
++nmi_espfix_stack:
++ /* We have a RING0_INT_FRAME here.
++ *
++ * create the pointer to lss back
++ */
++ pushl %ss
++ CFI_ADJUST_CFA_OFFSET 4
++ pushl %esp
++ CFI_ADJUST_CFA_OFFSET 4
++ addw $4, (%esp)
++ /* copy the iret frame of 12 bytes */
++ .rept 3
++ pushl 16(%esp)
++ CFI_ADJUST_CFA_OFFSET 4
++ .endr
++ pushl %eax
++ CFI_ADJUST_CFA_OFFSET 4
++ SAVE_ALL
++ FIXUP_ESPFIX_STACK # %eax == %esp
++ xorl %edx,%edx # zero error code
++ call do_nmi
++ RESTORE_REGS
++ lss 12+4(%esp), %esp # back to espfix stack
++ CFI_ADJUST_CFA_OFFSET -24
++1: INTERRUPT_RETURN
++ CFI_ENDPROC
++.section __ex_table,"a"
++ .align 4
++ .long 1b,iret_exc
++.previous
++KPROBE_END(nmi)
++#else
++KPROBE_ENTRY(nmi)
++ RING0_INT_FRAME
++ pushl %eax
++ CFI_ADJUST_CFA_OFFSET 4
++ SAVE_ALL
++ xorl %edx,%edx # zero error code
++ movl %esp,%eax # pt_regs pointer
++ call do_nmi
++ orl $NMI_MASK, PT_EFLAGS(%esp)
++ jmp restore_all
++ CFI_ENDPROC
++KPROBE_END(nmi)
++#endif
+
-+ return 0;
-+}
++#ifdef CONFIG_PARAVIRT
++ENTRY(native_iret)
++1: iret
++.section __ex_table,"a"
++ .align 4
++ .long 1b,iret_exc
++.previous
+
-+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;
++ENTRY(native_irq_enable_sysexit)
++ sti
++ sysexit
++#endif
+
-+ 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);
++KPROBE_ENTRY(int3)
++ RING0_INT_FRAME
++ pushl $-1 # mark this as an int
++ CFI_ADJUST_CFA_OFFSET 4
++ SAVE_ALL
++ xorl %edx,%edx # zero error code
++ movl %esp,%eax # pt_regs pointer
++ call do_int3
++ jmp ret_from_exception
++ CFI_ENDPROC
++KPROBE_END(int3)
+
-+ return 0;
-+}
++ENTRY(overflow)
++ RING0_INT_FRAME
++ pushl $0
++ CFI_ADJUST_CFA_OFFSET 4
++ pushl $do_overflow
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
+
-+static struct sysdev_class ioapic_sysdev_class = {
-+ set_kset_name("ioapic"),
-+ .suspend = ioapic_suspend,
-+ .resume = ioapic_resume,
-+};
++ENTRY(bounds)
++ RING0_INT_FRAME
++ pushl $0
++ CFI_ADJUST_CFA_OFFSET 4
++ pushl $do_bounds
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
+
-+static int __init ioapic_init_sysfs(void)
-+{
-+ struct sys_device * dev;
-+ int i, size, error = 0;
++ENTRY(invalid_op)
++ RING0_INT_FRAME
++ pushl $0
++ CFI_ADJUST_CFA_OFFSET 4
++ pushl $do_invalid_op
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
+
-+ error = sysdev_class_register(&ioapic_sysdev_class);
-+ if (error)
-+ return error;
++ENTRY(coprocessor_segment_overrun)
++ RING0_INT_FRAME
++ pushl $0
++ CFI_ADJUST_CFA_OFFSET 4
++ pushl $do_coprocessor_segment_overrun
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
+
-+ 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;
-+ }
-+ }
++ENTRY(invalid_TSS)
++ RING0_EC_FRAME
++ pushl $do_invalid_TSS
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
+
-+ return 0;
-+}
++ENTRY(segment_not_present)
++ RING0_EC_FRAME
++ pushl $do_segment_not_present
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
+
-+device_initcall(ioapic_init_sysfs);
++ENTRY(stack_segment)
++ RING0_EC_FRAME
++ pushl $do_stack_segment
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
++
++KPROBE_ENTRY(general_protection)
++ RING0_EC_FRAME
++ pushl $do_general_protection
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
++KPROBE_END(general_protection)
+
-+/* --------------------------------------------------------------------------
-+ ACPI-based IOAPIC Configuration
-+ -------------------------------------------------------------------------- */
++ENTRY(alignment_check)
++ RING0_EC_FRAME
++ pushl $do_alignment_check
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
+
-+#ifdef CONFIG_ACPI
++ENTRY(divide_error)
++ RING0_INT_FRAME
++ pushl $0 # no error code
++ CFI_ADJUST_CFA_OFFSET 4
++ pushl $do_divide_error
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
+
-+int __init io_apic_get_unique_id (int ioapic, int apic_id)
-+{
-+#ifndef CONFIG_XEN
-+ union IO_APIC_reg_00 reg_00;
-+ static physid_mask_t apic_id_map = PHYSID_MASK_NONE;
-+ physid_mask_t tmp;
-+ unsigned long flags;
-+ int i = 0;
++#ifdef CONFIG_X86_MCE
++ENTRY(machine_check)
++ RING0_INT_FRAME
++ pushl $0
++ CFI_ADJUST_CFA_OFFSET 4
++ pushl machine_check_vector
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
++#endif
+
-+ /*
-+ * The P4 platform supports up to 256 APIC IDs on two separate APIC
-+ * buses (one for LAPICs, one for IOAPICs), where predecessors only
-+ * supports up to 16 on one shared APIC bus.
-+ *
-+ * TBD: Expand LAPIC/IOAPIC support on P4-class systems to take full
-+ * advantage of new APIC bus architecture.
-+ */
++ENTRY(fixup_4gb_segment)
++ RING0_INT_FRAME
++ pushl $do_fixup_4gb_segment
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
+
-+ if (physids_empty(apic_id_map))
-+ apic_id_map = ioapic_phys_id_map(phys_cpu_present_map);
++ENTRY(kernel_thread_helper)
++ pushl $0 # fake return address for unwinder
++ CFI_STARTPROC
++ movl %edx,%eax
++ push %edx
++ CFI_ADJUST_CFA_OFFSET 4
++ call *%ebx
++ push %eax
++ CFI_ADJUST_CFA_OFFSET 4
++ call do_exit
++ CFI_ENDPROC
++ENDPROC(kernel_thread_helper)
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ reg_00.raw = io_apic_read(ioapic, 0);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
++.section .rodata,"a"
++#include "syscall_table.S"
+
-+ if (apic_id >= get_physical_broadcast()) {
-+ printk(KERN_WARNING "IOAPIC[%d]: Invalid apic_id %d, trying "
-+ "%d\n", ioapic, apic_id, reg_00.bits.ID);
-+ apic_id = reg_00.bits.ID;
-+ }
++syscall_table_size=(.-sys_call_table)
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/fixup.c b/arch/i386/kernel/fixup.c
+--- a/arch/i386/kernel/fixup.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/fixup.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,92 @@
++/******************************************************************************
++ * fixup.c
++ *
++ * Binary-rewriting of certain IA32 instructions, on notification by Xen.
++ * Used to avoid repeated slow emulation of common instructions used by the
++ * user-space TLS (Thread-Local Storage) libraries.
++ *
++ * **** NOTE ****
++ * Issues with the binary rewriting have caused it to be removed. Instead
++ * we rely on Xen's emulator to boot the kernel, and then print a banner
++ * message recommending that the user disables /lib/tls.
++ *
++ * Copyright (c) 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 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
++ */
+
-+ /*
-+ * Every APIC in a system must have a unique ID or we get lots of nice
-+ * 'stuck on smp_invalidate_needed IPI wait' messages.
-+ */
-+ if (check_apicid_used(apic_id_map, apic_id)) {
++#include <linux/init.h>
++#include <linux/sched.h>
++#include <linux/slab.h>
++#include <linux/kernel.h>
++#include <linux/delay.h>
++#include <linux/version.h>
+
-+ for (i = 0; i < get_physical_broadcast(); i++) {
-+ if (!check_apicid_used(apic_id_map, i))
-+ break;
-+ }
++#define DP(_f, _args...) printk(KERN_ALERT " " _f "\n" , ## _args )
+
-+ if (i == get_physical_broadcast())
-+ panic("Max apic_id exceeded!\n");
++fastcall void do_fixup_4gb_segment(struct pt_regs *regs, long error_code)
++{
++#if 0
++ static unsigned long printed = 0;
++ char info[100];
++ int i;
+
-+ printk(KERN_WARNING "IOAPIC[%d]: apic_id %d already used, "
-+ "trying %d\n", ioapic, apic_id, i);
++ if (test_and_set_bit(0, &printed))
++ return;
+
-+ apic_id = i;
-+ }
++ HYPERVISOR_vm_assist(
++ VMASST_CMD_disable, VMASST_TYPE_4gb_segments_notify);
+
-+ tmp = apicid_to_cpu_present(apic_id);
-+ physids_or(apic_id_map, apic_id_map, tmp);
++ sprintf(info, "%s (pid=%d)", current->comm, current->tgid);
+
-+ if (reg_00.bits.ID != apic_id) {
-+ reg_00.bits.ID = apic_id;
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ io_apic_write(ioapic, 0, reg_00.raw);
-+ reg_00.raw = io_apic_read(ioapic, 0);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
++ DP("");
++ DP("***************************************************************");
++ DP("***************************************************************");
++ DP("** WARNING: Currently emulating unsupported memory accesses **");
++ DP("** in /lib/tls glibc libraries. The emulation is **");
++ DP("** slow. To ensure full performance you should **");
++ DP("** install a 'xen-friendly' (nosegneg) version of **");
++ DP("** the library, or disable tls support by executing **");
++ DP("** the following as root: **");
++ DP("** mv /lib/tls /lib/tls.disabled **");
++ DP("** Offending process: %-38.38s **", info);
++ DP("***************************************************************");
++ DP("***************************************************************");
++ DP("");
+
-+ /* Sanity check */
-+ if (reg_00.bits.ID != apic_id) {
-+ printk("IOAPIC[%d]: Unable to change apic_id!\n", ioapic);
-+ return -1;
-+ }
++ for (i = 5; i > 0; i--) {
++ touch_softlockup_watchdog();
++ printk("Pausing... %d", i);
++ mdelay(1000);
++ printk("\b\b\b\b\b\b\b\b\b\b\b\b");
+ }
+
-+ apic_printk(APIC_VERBOSE, KERN_INFO
-+ "IOAPIC[%d]: Assigned apic_id %d\n", ioapic, apic_id);
-+#endif /* !CONFIG_XEN */
++ printk("Continuing...\n\n");
++#else
++ if (printk_ratelimit())
++ printk(KERN_WARNING
++ "4gb seg fixup, process %s (pid %d), cs:ip %02x:%08lx\n",
++ current->comm, current->tgid, regs->xcs, regs->eip);
++#endif
++}
+
-+ return apic_id;
++static int __init fixup_init(void)
++{
++ HYPERVISOR_vm_assist(
++ VMASST_CMD_enable, VMASST_TYPE_4gb_segments_notify);
++ return 0;
+}
++__initcall(fixup_init);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/head-xen.S b/arch/i386/kernel/head-xen.S
+--- a/arch/i386/kernel/head-xen.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/head-xen.S 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,286 @@
+
+
-+int __init io_apic_get_version (int ioapic)
-+{
-+ union IO_APIC_reg_01 reg_01;
-+ unsigned long flags;
++.text
++#include <linux/elfnote.h>
++#include <linux/threads.h>
++#include <linux/linkage.h>
++#include <asm/segment.h>
++#include <asm/page.h>
++#include <asm/boot.h>
++#include <asm/thread_info.h>
++#include <asm/asm-offsets.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/elfnote.h>
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ reg_01.raw = io_apic_read(ioapic, 1);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
++/*
++ * References to members of the new_cpu_data structure.
++ */
+
-+ return reg_01.bits.version;
-+}
++#define X86 new_cpu_data+CPUINFO_x86
++#define X86_VENDOR new_cpu_data+CPUINFO_x86_vendor
++#define X86_MODEL new_cpu_data+CPUINFO_x86_model
++#define X86_MASK new_cpu_data+CPUINFO_x86_mask
++#define X86_HARD_MATH new_cpu_data+CPUINFO_hard_math
++#define X86_CPUID new_cpu_data+CPUINFO_cpuid_level
++#define X86_CAPABILITY new_cpu_data+CPUINFO_x86_capability
++#define X86_VENDOR_ID new_cpu_data+CPUINFO_x86_vendor_id
+
++#define VIRT_ENTRY_OFFSET 0x0
++.org VIRT_ENTRY_OFFSET
++ENTRY(startup_32)
+
-+int __init io_apic_get_redir_entries (int ioapic)
-+{
-+ union IO_APIC_reg_01 reg_01;
-+ unsigned long flags;
++#ifdef CONFIG_PARAVIRT
++ movl %cs, %eax
++ testl $0x3, %eax
++ jnz startup_paravirt
++#endif
+
-+ spin_lock_irqsave(&ioapic_lock, flags);
-+ reg_01.raw = io_apic_read(ioapic, 1);
-+ spin_unlock_irqrestore(&ioapic_lock, flags);
++ movl %esi,xen_start_info
++ cld
+
-+ return reg_01.bits.entries;
-+}
++ call setup_pda
+
++ /* Set up the stack pointer */
++ movl $(init_thread_union+THREAD_SIZE),%esp
+
-+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;
++ /* get vendor info */
++ xorl %eax,%eax # call CPUID with 0 -> return vendor ID
++ XEN_CPUID
++ movl %eax,X86_CPUID # save CPUID level
++ movl %ebx,X86_VENDOR_ID # lo 4 chars
++ movl %edx,X86_VENDOR_ID+4 # next 4 chars
++ movl %ecx,X86_VENDOR_ID+8 # last 4 chars
+
-+ if (!IO_APIC_IRQ(irq)) {
-+ printk(KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
-+ ioapic);
-+ return -EINVAL;
-+ }
++ movl $1,%eax # Use the CPUID instruction to get CPU type
++ XEN_CPUID
++ movb %al,%cl # save reg for future use
++ andb $0x0f,%ah # mask processor family
++ movb %ah,X86
++ andb $0xf0,%al # mask model
++ shrb $4,%al
++ movb %al,X86_MODEL
++ andb $0x0f,%cl # mask mask revision
++ movb %cl,X86_MASK
++ movl %edx,X86_CAPABILITY
+
-+ /*
-+ * 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.
-+ */
++ movb $1,X86_HARD_MATH
+
-+ memset(&entry,0,sizeof(entry));
++ xorl %eax,%eax # Clear FS and LDT
++ movl %eax,%fs
+
-+ 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;
++ movl $(__KERNEL_PDA),%eax
++ mov %eax,%gs
+
-+ /*
-+ * IRQs < 16 are already in the irq_2_pin[] map
-+ */
-+ if (irq >= 16)
-+ add_pin_to_irq(irq, ioapic, pin);
++ cld # gcc2 wants the direction flag cleared at all times
+
-+ entry.vector = assign_irq_vector(irq);
++ call start_kernel
++L6:
++ jmp L6 # main should never return here, but
++ # just in case, we know what happens.
+
-+ apic_printk(APIC_DEBUG, 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);
++/*
++ * Point the GDT at this CPU's PDA. This will be
++ * cpu_gdt_table and boot_pda.
++ */
++setup_pda:
++ /* get the PDA pointer */
++ movl $boot_pda, %eax
++
++ /* slot the PDA address into the GDT */
++ mov $cpu_gdt_table, %ecx
++ mov %ax, (__KERNEL_PDA+0+2)(%ecx) /* base & 0x0000ffff */
++ shr $16, %eax
++ mov %al, (__KERNEL_PDA+4+0)(%ecx) /* base & 0x00ff0000 */
++ mov %ah, (__KERNEL_PDA+4+3)(%ecx) /* base & 0xff000000 */
++
++ # %esi still points to start_info, and no registers
++ # need to be preserved.
++
++ movl XEN_START_mfn_list(%esi), %ebx
++ movl $(cpu_gdt_table - __PAGE_OFFSET), %eax
++ shrl $PAGE_SHIFT, %eax
++ movl (%ebx,%eax,4), %ecx
++ pushl %ecx # frame number for set_gdt below
+
-+ ioapic_register_intr(irq, entry.vector, edge_level);
++ xorl %esi, %esi
++ xorl %edx, %edx
++ shldl $PAGE_SHIFT, %ecx, %edx
++ shll $PAGE_SHIFT, %ecx
++ orl $0x61, %ecx
++ movl $cpu_gdt_table, %ebx
++ movl $__HYPERVISOR_update_va_mapping, %eax
++ int $0x82
++
++ movl $(PAGE_SIZE_asm / 8), %ecx
++ movl %esp, %ebx
++ movl $__HYPERVISOR_set_gdt, %eax
++ int $0x82
+
-+ if (!ioapic && (irq < 16))
-+ disable_8259A_irq(irq);
++ popl %ecx
++ ret
+
-+ 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);
++#define HYPERCALL_PAGE_OFFSET 0x1000
++.org HYPERCALL_PAGE_OFFSET
++ENTRY(hypercall_page)
++.skip 0x1000
+
-+ return 0;
-+}
++/*
++ * Real beginning of normal "text" segment
++ */
++ENTRY(stext)
++ENTRY(_stext)
+
-+#endif /* CONFIG_ACPI */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/ioport-xen.c linux-2.6.18-xen/arch/i386/kernel/ioport-xen.c
---- linux-2.6.18.3/arch/i386/kernel/ioport-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/ioport-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,121 @@
+/*
-+ * linux/arch/i386/kernel/ioport.c
-+ *
-+ * This contains the io-permission bitmap code - written by obz, with changes
-+ * by Linus.
++ * BSS section
+ */
++.section ".bss.page_aligned","w"
++ENTRY(empty_zero_page)
++ .fill 4096,1,0
+
-+#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/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>
++/*
++ * This starts the data section.
++ */
++.data
++ENTRY(start_pda)
++ .long boot_pda
+
-+/* 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)
-+{
-+ unsigned long mask;
-+ unsigned long *bitmap_base = bitmap + (base / BITS_PER_LONG);
-+ unsigned int low_index = base & (BITS_PER_LONG-1);
-+ int length = low_index + extent;
++#ifdef CONFIG_PARAVIRT
++startup_paravirt:
++ cld
++ movl $(init_thread_union+THREAD_SIZE),%esp
+
-+ if (low_index != 0) {
-+ mask = (~0UL << low_index);
-+ if (length < BITS_PER_LONG)
-+ mask &= ~(~0UL << length);
-+ if (new_value)
-+ *bitmap_base++ |= mask;
-+ else
-+ *bitmap_base++ &= ~mask;
-+ length -= BITS_PER_LONG;
-+ }
++ /* We take pains to preserve all the regs. */
++ pushl %edx
++ pushl %ecx
++ pushl %eax
+
-+ mask = (new_value ? ~0UL : 0UL);
-+ while (length >= BITS_PER_LONG) {
-+ *bitmap_base++ = mask;
-+ length -= BITS_PER_LONG;
-+ }
++ /* paravirt.o is last in link, and that probe fn never returns */
++ pushl $__start_paravirtprobe
++1:
++ movl 0(%esp), %eax
++ pushl (%eax)
++ movl 8(%esp), %eax
++ call *(%esp)
++ popl %eax
+
-+ if (length > 0) {
-+ mask = ~(~0UL << length);
-+ if (new_value)
-+ *bitmap_base++ |= mask;
-+ else
-+ *bitmap_base++ &= ~mask;
-+ }
-+}
++ movl 4(%esp), %eax
++ movl 8(%esp), %ecx
++ movl 12(%esp), %edx
+
++ addl $4, (%esp)
++ jmp 1b
++#endif
+
+/*
-+ * this changes the io permissions bitmap in the current task.
++ * The Global Descriptor Table contains 28 quadwords, per-CPU.
+ */
-+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;
++ .section .data.page_aligned, "aw"
++ .align PAGE_SIZE_asm
++ENTRY(cpu_gdt_table)
++ .quad 0x0000000000000000 /* NULL descriptor */
++ .quad 0x0000000000000000 /* 0x0b reserved */
++ .quad 0x0000000000000000 /* 0x13 reserved */
++ .quad 0x0000000000000000 /* 0x1b reserved */
++ .quad 0x0000000000000000 /* 0x20 unused */
++ .quad 0x0000000000000000 /* 0x28 unused */
++ .quad 0x0000000000000000 /* 0x33 TLS entry 1 */
++ .quad 0x0000000000000000 /* 0x3b TLS entry 2 */
++ .quad 0x0000000000000000 /* 0x43 TLS entry 3 */
++ .quad 0x0000000000000000 /* 0x4b reserved */
++ .quad 0x0000000000000000 /* 0x53 reserved */
++ .quad 0x0000000000000000 /* 0x5b reserved */
+
-+ if ((from + num <= from) || (from + num > IO_BITMAP_BITS))
-+ return -EINVAL;
-+ if (turn_on && !capable(CAP_SYS_RAWIO))
-+ return -EPERM;
++ .quad 0x00cf9a000000ffff /* 0x60 kernel 4GB code at 0x00000000 */
++ .quad 0x00cf92000000ffff /* 0x68 kernel 4GB data at 0x00000000 */
++ .quad 0x00cffa000000ffff /* 0x73 user 4GB code at 0x00000000 */
++ .quad 0x00cff2000000ffff /* 0x7b user 4GB data at 0x00000000 */
++
++ .quad 0x0000000000000000 /* 0x80 TSS descriptor */
++ .quad 0x0000000000000000 /* 0x88 LDT descriptor */
+
+ /*
-+ * 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:
++ * Segments used for calling PnP BIOS have byte granularity.
++ * They code segments and data segments have fixed 64k limits,
++ * the transfer segment sizes are set at run time.
+ */
-+ 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_iobitmap.bitmap = (char *)bitmap;
-+ set_iobitmap.nr_ports = IO_BITMAP_BITS;
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap, &set_iobitmap);
-+ }
++ .quad 0x0000000000000000 /* 0x90 32-bit code */
++ .quad 0x0000000000000000 /* 0x98 16-bit code */
++ .quad 0x0000000000000000 /* 0xa0 16-bit data */
++ .quad 0x0000000000000000 /* 0xa8 16-bit data */
++ .quad 0x0000000000000000 /* 0xb0 16-bit data */
+
-+ set_bitmap(t->io_bitmap_ptr, from, num, !turn_on);
++ /*
++ * The APM segments have byte granularity and their bases
++ * are set at run time. All have 64k limits.
++ */
++ .quad 0x0000000000000000 /* 0xb8 APM CS code */
++ .quad 0x0000000000000000 /* 0xc0 APM CS 16 code (16 bit) */
++ .quad 0x0000000000000000 /* 0xc8 APM DS data */
+
-+ return 0;
-+}
++ .quad 0x0000000000000000 /* 0xd0 - ESPFIX SS */
++ .quad 0x00cf92000000ffff /* 0xd8 - PDA */
++ .quad 0x0000000000000000 /* 0xe0 - unused */
++ .quad 0x0000000000000000 /* 0xe8 - unused */
++ .quad 0x0000000000000000 /* 0xf0 - unused */
++ .quad 0x0000000000000000 /* 0xf8 - GDT entry 31: double-fault TSS */
++ .align PAGE_SIZE_asm
+
++#ifdef CONFIG_XEN_COMPAT_030002
+/*
-+ * 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.
-+ *
-+ * Here we just change the eflags value on the stack: we allow
-+ * only the super-user to do it. This depends on the stack-layout
-+ * on system-call entry - see also fork() and the signal handling
-+ * code.
++ * __xen_guest information
+ */
++.macro utoa value
++ .if (\value) < 0 || (\value) >= 0x10
++ utoa (((\value)>>4)&0x0fffffff)
++ .endif
++ .if ((\value) & 0xf) < 10
++ .byte '0' + ((\value) & 0xf)
++ .else
++ .byte 'A' + ((\value) & 0xf) - 10
++ .endif
++.endm
+
-+asmlinkage long sys_iopl(unsigned long unused)
-+{
-+ volatile struct pt_regs * regs = (struct pt_regs *) &unused;
-+ unsigned int level = regs->ebx;
-+ struct thread_struct *t = ¤t->thread;
-+ unsigned int old = (t->iopl >> 12) & 3;
++.section __xen_guest
++ .ascii "GUEST_OS=linux,GUEST_VER=2.6"
++ .ascii ",XEN_VER=xen-3.0"
++ .ascii ",VIRT_BASE=0x"
++ utoa __PAGE_OFFSET
++ .ascii ",ELF_PADDR_OFFSET=0x"
++ utoa __PAGE_OFFSET
++ .ascii ",VIRT_ENTRY=0x"
++ utoa (__PAGE_OFFSET + LOAD_PHYSICAL_ADDR + VIRT_ENTRY_OFFSET)
++ .ascii ",HYPERCALL_PAGE=0x"
++ utoa ((LOAD_PHYSICAL_ADDR+HYPERCALL_PAGE_OFFSET)>>PAGE_SHIFT)
++ .ascii ",FEATURES=writable_page_tables"
++ .ascii "|writable_descriptor_tables"
++ .ascii "|auto_translated_physmap"
++ .ascii "|pae_pgdir_above_4gb"
++ .ascii "|supervisor_mode_kernel"
++#ifdef CONFIG_X86_PAE
++ .ascii ",PAE=yes[extended-cr3]"
++#else
++ .ascii ",PAE=no"
++#endif
++ .ascii ",LOADER=generic"
++ .byte 0
++#endif /* CONFIG_XEN_COMPAT_030002 */
+
-+ if (level > 3)
-+ return -EINVAL;
-+ /* Trying to gain more privileges? */
-+ if (level > old) {
-+ if (!capable(CAP_SYS_RAWIO))
-+ return -EPERM;
-+ }
-+ t->iopl = level << 12;
-+ set_iopl_mask(t->iopl);
-+ return 0;
-+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/irq-xen.c linux-2.6.18-xen/arch/i386/kernel/irq-xen.c
---- linux-2.6.18.3/arch/i386/kernel/irq-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/irq-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,324 @@
-+/*
-+ * linux/arch/i386/kernel/irq.c
-+ *
-+ * Copyright (C) 1992, 1998 Linus Torvalds, Ingo Molnar
-+ *
-+ * This file contains the lowest level x86-specific interrupt
-+ * entry, irq-stacks and irq statistics code. All the remaining
-+ * irq logic is done by the generic kernel/irq/ code and
-+ * by the x86-specific irq controller code. (e.g. i8259.c and
-+ * io_apic.c.)
-+ */
++
++ 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, .long, __PAGE_OFFSET)
++#ifdef CONFIG_XEN_COMPAT_030002
++ ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET, .long, __PAGE_OFFSET)
++#else
++ ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET, .long, 0)
++#endif /* !CONFIG_XEN_COMPAT_030002 */
++ ELFNOTE(Xen, XEN_ELFNOTE_ENTRY, .long, startup_32)
++ ELFNOTE(Xen, XEN_ELFNOTE_HYPERCALL_PAGE, .long, hypercall_page)
++ ELFNOTE(Xen, XEN_ELFNOTE_FEATURES, .asciz, "writable_page_tables|writable_descriptor_tables|auto_translated_physmap|pae_pgdir_above_4gb|supervisor_mode_kernel")
++#ifdef CONFIG_X86_PAE
++ ELFNOTE(Xen, XEN_ELFNOTE_PAE_MODE, .asciz, "yes")
++#else
++ ELFNOTE(Xen, XEN_ELFNOTE_PAE_MODE, .asciz, "no")
++#endif
++ ELFNOTE(Xen, XEN_ELFNOTE_LOADER, .asciz, "generic")
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/init_task-xen.c b/arch/i386/kernel/init_task-xen.c
+--- a/arch/i386/kernel/init_task-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/init_task-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,51 @@
++#include <linux/mm.h>
++#include <linux/module.h>
++#include <linux/sched.h>
++#include <linux/init.h>
++#include <linux/init_task.h>
++#include <linux/fs.h>
++#include <linux/mqueue.h>
+
+#include <asm/uaccess.h>
-+#include <linux/module.h>
-+#include <linux/seq_file.h>
-+#include <linux/interrupt.h>
-+#include <linux/kernel_stat.h>
-+#include <linux/notifier.h>
-+#include <linux/cpu.h>
-+#include <linux/delay.h>
++#include <asm/pgtable.h>
++#include <asm/desc.h>
++
++static struct fs_struct init_fs = INIT_FS;
++static struct files_struct init_files = INIT_FILES;
++static struct signal_struct init_signals = INIT_SIGNALS(init_signals);
++static struct sighand_struct init_sighand = INIT_SIGHAND(init_sighand);
++
++#define swapper_pg_dir ((pgd_t *)NULL)
++struct mm_struct init_mm = INIT_MM(init_mm);
++#undef swapper_pg_dir
+
-+DEFINE_PER_CPU(irq_cpustat_t, irq_stat) ____cacheline_internodealigned_in_smp;
-+EXPORT_PER_CPU_SYMBOL(irq_stat);
++EXPORT_SYMBOL(init_mm);
+
-+#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.
++ * Initial thread structure.
++ *
++ * We need to make sure that this is THREAD_SIZE aligned due to the
++ * way process stacks are handled. This is done by having a special
++ * "init_task" linker map entry..
+ */
-+void ack_bad_irq(unsigned int irq)
-+{
-+ printk("unexpected IRQ trap at vector %02x\n", irq);
-+}
-+#endif
++union thread_union init_thread_union
++ __attribute__((__section__(".data.init_task"))) =
++ { INIT_THREAD_INFO(init_task) };
+
-+#ifdef CONFIG_4KSTACKS
+/*
-+ * per-CPU IRQ handling contexts (thread information and stack)
++ * Initial task structure.
++ *
++ * All other task structs will be allocated on slabs in fork.c
+ */
-+union irq_ctx {
-+ struct thread_info tinfo;
-+ u32 stack[THREAD_SIZE/sizeof(u32)];
-+};
++struct task_struct init_task = INIT_TASK(init_task);
+
-+static union irq_ctx *hardirq_ctx[NR_CPUS] __read_mostly;
-+static union irq_ctx *softirq_ctx[NR_CPUS] __read_mostly;
++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.
++ */
++DEFINE_PER_CPU(struct tss_struct, init_tss) ____cacheline_internodealigned_in_smp = INIT_TSS;
+#endif
+
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/io_apic-xen.c b/arch/i386/kernel/io_apic-xen.c
+--- a/arch/i386/kernel/io_apic-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/io_apic-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,2973 @@
+/*
-+ * do_IRQ handles all normal device IRQ's (the special
-+ * SMP cross-CPU interrupts have their own specific
-+ * handlers).
++ * 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
+ */
-+fastcall unsigned int do_IRQ(struct pt_regs *regs)
-+{
-+ /* high bit used in ret_from_ code */
-+ int irq = ~regs->orig_eax;
-+#ifdef CONFIG_4KSTACKS
-+ union irq_ctx *curctx, *irqctx;
-+ u32 *isp;
-+#endif
+
-+ if (unlikely((unsigned)irq >= NR_IRQS)) {
-+ printk(KERN_EMERG "%s: cannot handle IRQ %d\n",
-+ __FUNCTION__, irq);
-+ BUG();
-+ }
++#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/compiler.h>
++#include <linux/acpi.h>
++#include <linux/module.h>
++#include <linux/sysdev.h>
++#include <linux/pci.h>
++#include <linux/msi.h>
++#include <linux/htirq.h>
++#include <linux/freezer.h>
+
-+ irq_enter();
-+#ifdef CONFIG_DEBUG_STACKOVERFLOW
-+ /* Debugging check for stack overflow: is there less than 1KB free? */
-+ {
-+ long esp;
++#include <asm/io.h>
++#include <asm/smp.h>
++#include <asm/desc.h>
++#include <asm/timer.h>
++#include <asm/i8259.h>
++#include <asm/nmi.h>
++#include <asm/msidef.h>
++#include <asm/hypertransport.h>
+
-+ __asm__ __volatile__("andl %%esp,%0" :
-+ "=r" (esp) : "0" (THREAD_SIZE - 1));
-+ if (unlikely(esp < (sizeof(struct thread_info) + STACK_WARN))) {
-+ printk("do_IRQ: stack overflow: %ld\n",
-+ esp - sizeof(struct thread_info));
-+ dump_stack();
-+ }
-+ }
-+#endif
++#include <mach_apic.h>
++#include <mach_apicdef.h>
+
-+#ifdef CONFIG_4KSTACKS
++#include "io_ports.h"
+
-+ curctx = (union irq_ctx *) current_thread_info();
-+ irqctx = hardirq_ctx[smp_processor_id()];
++#ifdef CONFIG_XEN
+
-+ /*
-+ * this is where we switch to the IRQ stack. However, if we are
-+ * already using the IRQ stack (because we interrupted a hardirq
-+ * handler) we can't do that and just have to keep using the
-+ * current stack (which is the irq stack already after all)
-+ */
-+ if (curctx != irqctx) {
-+ int arg1, arg2, ebx;
++#include <xen/interface/xen.h>
++#include <xen/interface/physdev.h>
+
-+ /* build the stack frame on the IRQ stack */
-+ isp = (u32*) ((char*)irqctx + sizeof(*irqctx));
-+ irqctx->tinfo.task = curctx->tinfo.task;
-+ irqctx->tinfo.previous_esp = current_stack_pointer;
++/* 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)
+
-+ /*
-+ * Copy the softirq bits in preempt_count so that the
-+ * softirq checks work in the hardirq context.
-+ */
-+ irqctx->tinfo.preempt_count =
-+ (irqctx->tinfo.preempt_count & ~SOFTIRQ_MASK) |
-+ (curctx->tinfo.preempt_count & SOFTIRQ_MASK);
-+
-+ asm volatile(
-+ " xchgl %%ebx,%%esp \n"
-+ " call __do_IRQ \n"
-+ " movl %%ebx,%%esp \n"
-+ : "=a" (arg1), "=d" (arg2), "=b" (ebx)
-+ : "0" (irq), "1" (regs), "2" (isp)
-+ : "memory", "cc", "ecx"
-+ );
-+ } else
-+#endif
-+ __do_IRQ(irq, regs);
++unsigned long io_apic_irqs;
+
-+ irq_exit();
++static inline unsigned int xen_io_apic_read(unsigned int apic, unsigned int reg)
++{
++ struct physdev_apic apic_op;
++ int ret;
+
-+ return 1;
++ 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;
++ HYPERVISOR_physdev_op(PHYSDEVOP_apic_write, &apic_op);
+}
+
-+#ifdef CONFIG_4KSTACKS
++#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)
++
++#endif /* CONFIG_XEN */
++
++int (*ioapic_renumber_irq)(int ioapic, int irq);
++atomic_t irq_mis_count;
++
++/* Where if anywhere is the i8259 connect in external int mode */
++static struct { int pin, apic; } ioapic_i8259 = { -1, -1 };
++
++static DEFINE_SPINLOCK(ioapic_lock);
++static DEFINE_SPINLOCK(vector_lock);
++
++int timer_over_8254 __initdata = 1;
+
+/*
-+ * These should really be __section__(".bss.page_aligned") as well, but
-+ * gcc's 3.0 and earlier don't handle that correctly.
++ * Is the SiS APIC rmw bug present ?
++ * -1 = don't know, 0 = no, 1 = yes
+ */
-+static char softirq_stack[NR_CPUS * THREAD_SIZE]
-+ __attribute__((__aligned__(THREAD_SIZE)));
-+
-+static char hardirq_stack[NR_CPUS * THREAD_SIZE]
-+ __attribute__((__aligned__(THREAD_SIZE)));
++int sis_apic_bug = -1;
+
+/*
-+ * allocate per-cpu stacks for hardirq and for softirq processing
++ * # of IRQ routing registers
+ */
-+void irq_ctx_init(int cpu)
-+{
-+ union irq_ctx *irqctx;
++int nr_ioapic_registers[MAX_IO_APICS];
+
-+ if (hardirq_ctx[cpu])
-+ return;
++static int disable_timer_pin_1 __initdata;
+
-+ irqctx = (union irq_ctx*) &hardirq_stack[cpu*THREAD_SIZE];
-+ irqctx->tinfo.task = NULL;
-+ irqctx->tinfo.exec_domain = NULL;
-+ irqctx->tinfo.cpu = cpu;
-+ irqctx->tinfo.preempt_count = HARDIRQ_OFFSET;
-+ irqctx->tinfo.addr_limit = MAKE_MM_SEG(0);
++/*
++ * Rough estimation of how many shared IRQs there are, can
++ * be changed anytime.
++ */
++#define MAX_PLUS_SHARED_IRQS NR_IRQS
++#define PIN_MAP_SIZE (MAX_PLUS_SHARED_IRQS + NR_IRQS)
+
-+ hardirq_ctx[cpu] = irqctx;
++/*
++ * 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.
++ */
+
-+ irqctx = (union irq_ctx*) &softirq_stack[cpu*THREAD_SIZE];
-+ irqctx->tinfo.task = NULL;
-+ irqctx->tinfo.exec_domain = NULL;
-+ irqctx->tinfo.cpu = cpu;
-+ irqctx->tinfo.preempt_count = 0;
-+ irqctx->tinfo.addr_limit = MAKE_MM_SEG(0);
++static struct irq_pin_list {
++ int apic, pin, next;
++} irq_2_pin[PIN_MAP_SIZE];
+
-+ softirq_ctx[cpu] = irqctx;
++#ifndef CONFIG_XEN
++struct io_apic {
++ unsigned int index;
++ unsigned int unused[3];
++ unsigned int data;
++};
+
-+ printk("CPU %u irqstacks, hard=%p soft=%p\n",
-+ cpu,hardirq_ctx[cpu],softirq_ctx[cpu]);
++static __attribute_const__ struct io_apic __iomem *io_apic_base(int idx)
++{
++ return (void __iomem *) __fix_to_virt(FIX_IO_APIC_BASE_0 + idx)
++ + (mp_ioapics[idx].mpc_apicaddr & ~PAGE_MASK);
+}
+
-+void irq_ctx_exit(int cpu)
++static inline unsigned int io_apic_read(unsigned int apic, unsigned int reg)
+{
-+ hardirq_ctx[cpu] = NULL;
++ struct io_apic __iomem *io_apic = io_apic_base(apic);
++ writel(reg, &io_apic->index);
++ return readl(&io_apic->data);
+}
+
-+extern asmlinkage void __do_softirq(void);
-+
-+asmlinkage void do_softirq(void)
++static inline void io_apic_write(unsigned int apic, unsigned int reg, unsigned int value)
+{
-+ unsigned long flags;
-+ struct thread_info *curctx;
-+ union irq_ctx *irqctx;
-+ u32 *isp;
-+
-+ if (in_interrupt())
-+ return;
++ struct io_apic __iomem *io_apic = io_apic_base(apic);
++ writel(reg, &io_apic->index);
++ writel(value, &io_apic->data);
++}
+
-+ local_irq_save(flags);
++/*
++ * Re-write a value: to be used for read-modify-write
++ * cycles where the read already set up the index register.
++ *
++ * Older SiS APIC requires we rewrite the index register
++ */
++static inline void io_apic_modify(unsigned int apic, unsigned int reg, unsigned int value)
++{
++ volatile struct io_apic *io_apic = io_apic_base(apic);
++ if (sis_apic_bug)
++ writel(reg, &io_apic->index);
++ writel(value, &io_apic->data);
++}
++#endif /* !CONFIG_XEN */
+
-+ if (local_softirq_pending()) {
-+ curctx = current_thread_info();
-+ irqctx = softirq_ctx[smp_processor_id()];
-+ irqctx->tinfo.task = curctx->task;
-+ irqctx->tinfo.previous_esp = current_stack_pointer;
-+
-+ /* build the stack frame on the softirq stack */
-+ isp = (u32*) ((char*)irqctx + sizeof(*irqctx));
-+
-+ asm volatile(
-+ " xchgl %%ebx,%%esp \n"
-+ " call __do_softirq \n"
-+ " movl %%ebx,%%esp \n"
-+ : "=b"(isp)
-+ : "0"(isp)
-+ : "memory", "cc", "edx", "ecx", "eax"
-+ );
-+ /*
-+ * Shouldnt happen, we returned above if in_interrupt():
-+ */
-+ WARN_ON_ONCE(softirq_count());
-+ }
++union entry_union {
++ struct { u32 w1, w2; };
++ struct IO_APIC_route_entry entry;
++};
+
-+ local_irq_restore(flags);
++static struct IO_APIC_route_entry ioapic_read_entry(int apic, int pin)
++{
++ union entry_union eu;
++ unsigned long flags;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ eu.w1 = io_apic_read(apic, 0x10 + 2 * pin);
++ eu.w2 = io_apic_read(apic, 0x11 + 2 * pin);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++ return eu.entry;
+}
+
-+EXPORT_SYMBOL(do_softirq);
-+#endif
-+
+/*
-+ * Interrupt statistics:
++ * When we write a new IO APIC routing entry, we need to write the high
++ * word first! If the mask bit in the low word is clear, we will enable
++ * the interrupt, and we need to make sure the entry is fully populated
++ * before that happens.
+ */
++static void
++__ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e)
++{
++ union entry_union eu;
++ eu.entry = e;
++ io_apic_write(apic, 0x11 + 2*pin, eu.w2);
++ io_apic_write(apic, 0x10 + 2*pin, eu.w1);
++}
+
-+atomic_t irq_err_count;
++static void ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e)
++{
++ unsigned long flags;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ __ioapic_write_entry(apic, pin, e);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
+
+/*
-+ * /proc/interrupts printing:
++ * When we mask an IO APIC routing entry, we need to write the low
++ * word first, in order to set the mask bit before we change the
++ * high bits!
+ */
+
-+int show_interrupts(struct seq_file *p, void *v)
++#ifndef CONFIG_XEN
++static void ioapic_mask_entry(int apic, int pin)
+{
-+ int i = *(loff_t *) v, j;
-+ struct irqaction * action;
+ unsigned long flags;
++ union entry_union eu = { .entry.mask = 1 };
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ io_apic_write(apic, 0x10 + 2*pin, eu.w1);
++ io_apic_write(apic, 0x11 + 2*pin, eu.w2);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
++#endif
++
++/*
++ * 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;
++
++ while (entry->next)
++ entry = irq_2_pin + entry->next;
+
-+ if (i == 0) {
-+ seq_printf(p, " ");
-+ for_each_online_cpu(j)
-+ seq_printf(p, "CPU%-8d ",j);
-+ seq_putc(p, '\n');
++ 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: whoops");
+ }
++ entry->apic = apic;
++ entry->pin = pin;
++}
+
-+ 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));
++#ifdef CONFIG_XEN
++#define clear_IO_APIC() ((void)0)
+#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);
++/*
++ * Reroute an IRQ to a different pin.
++ */
++static void __init replace_pin_at_irq(unsigned int irq,
++ int oldapic, int oldpin,
++ int newapic, int newpin)
++{
++ struct irq_pin_list *entry = irq_2_pin + irq;
+
-+ 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 ", nmi_count(j));
-+ seq_putc(p, '\n');
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ seq_printf(p, "LOC: ");
-+ for_each_online_cpu(j)
-+ seq_printf(p, "%10u ",
-+ per_cpu(irq_stat,j).apic_timer_irqs);
-+ seq_putc(p, '\n');
-+#endif
-+ seq_printf(p, "ERR: %10u\n", atomic_read(&irq_err_count));
-+#if defined(CONFIG_X86_IO_APIC)
-+ seq_printf(p, "MIS: %10u\n", atomic_read(&irq_mis_count));
-+#endif
++ while (1) {
++ if (entry->apic == oldapic && entry->pin == oldpin) {
++ entry->apic = newapic;
++ entry->pin = newpin;
++ }
++ if (!entry->next)
++ break;
++ entry = irq_2_pin + entry->next;
+ }
-+ return 0;
+}
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+
-+void fixup_irqs(cpumask_t map)
++static void __modify_IO_APIC_irq (unsigned int irq, unsigned long enable, unsigned long disable)
+{
-+ unsigned int irq;
-+ static int warned;
-+
-+ for (irq = 0; irq < NR_IRQS; irq++) {
-+ cpumask_t mask;
-+ if (irq == 2)
-+ continue;
++ struct irq_pin_list *entry = irq_2_pin + irq;
++ unsigned int pin, reg;
+
-+ 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);
++ for (;;) {
++ pin = entry->pin;
++ if (pin == -1)
++ break;
++ reg = io_apic_read(entry->apic, 0x10 + pin*2);
++ reg &= ~disable;
++ reg |= enable;
++ io_apic_modify(entry->apic, 0x10 + pin*2, reg);
++ if (!entry->next)
++ break;
++ entry = irq_2_pin + entry->next;
+ }
-+
-+#if 0
-+ barrier();
-+ /* Ingo Molnar says: "after the IO-APIC masks have been redirected
-+ [note the nop - the interrupt-enable boundary on x86 is two
-+ instructions from sti] - to flush out pending hardirqs and
-+ IPIs. After this point nothing is supposed to reach this CPU." */
-+ __asm__ __volatile__("sti; nop; cli");
-+ barrier();
-+#else
-+ /* That doesn't seem sufficient. Give it 1ms. */
-+ local_irq_enable();
-+ mdelay(1);
-+ local_irq_disable();
-+#endif
+}
-+#endif
-+
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/ldt-xen.c linux-2.6.18-xen/arch/i386/kernel/ldt-xen.c
---- linux-2.6.18.3/arch/i386/kernel/ldt-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/ldt-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,270 @@
-+/*
-+ * linux/kernel/ldt.c
-+ *
-+ * Copyright (C) 1992 Krishna Balasubramanian and Linus Torvalds
-+ * Copyright (C) 1999 Ingo Molnar <mingo at redhat.com>
-+ */
+
-+#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>
++/* mask = 1 */
++static void __mask_IO_APIC_irq (unsigned int irq)
++{
++ __modify_IO_APIC_irq(irq, 0x00010000, 0);
++}
+
-+#include <asm/uaccess.h>
-+#include <asm/system.h>
-+#include <asm/ldt.h>
-+#include <asm/desc.h>
-+#include <asm/mmu_context.h>
++/* mask = 0 */
++static void __unmask_IO_APIC_irq (unsigned int irq)
++{
++ __modify_IO_APIC_irq(irq, 0, 0x00010000);
++}
+
-+#ifdef CONFIG_SMP /* avoids "defined but not used" warnig */
-+static void flush_ldt(void *null)
++/* mask = 1, trigger = 0 */
++static void __mask_and_edge_IO_APIC_irq (unsigned int irq)
+{
-+ if (current->active_mm)
-+ load_LDT(¤t->active_mm->context);
++ __modify_IO_APIC_irq(irq, 0x00010000, 0x00008000);
+}
-+#endif
+
-+static int alloc_ldt(mm_context_t *pc, int mincount, int reload)
++/* mask = 0, trigger = 1 */
++static void __unmask_and_level_IO_APIC_irq (unsigned int irq)
+{
-+ void *oldldt;
-+ void *newldt;
-+ int oldsize;
++ __modify_IO_APIC_irq(irq, 0x00008000, 0x00010000);
++}
+
-+ if (mincount <= 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);
++static void mask_IO_APIC_irq (unsigned int irq)
++{
++ unsigned long flags;
+
-+ if (!newldt)
-+ return -ENOMEM;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ __mask_IO_APIC_irq(irq);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
+
-+ 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);
-+ pc->ldt = newldt;
-+ wmb();
-+ pc->size = mincount;
-+ wmb();
++static void unmask_IO_APIC_irq (unsigned int irq)
++{
++ unsigned long flags;
+
-+ 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;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ __unmask_IO_APIC_irq(irq);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
+}
+
-+static inline int copy_ldt(mm_context_t *new, mm_context_t *old)
++static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin)
+{
-+ 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;
++ struct IO_APIC_route_entry entry;
++
++ /* Check delivery_mode to be sure we're not clearing an SMI pin */
++ entry = ioapic_read_entry(apic, pin);
++ if (entry.delivery_mode == dest_SMI)
++ return;
++
++ /*
++ * Disable it in the IO-APIC irq-routing table:
++ */
++ ioapic_mask_entry(apic, pin);
+}
+
-+/*
-+ * 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)
++static void clear_IO_APIC (void)
+{
-+ struct mm_struct * old_mm;
-+ int retval = 0;
++ int apic, pin;
+
-+ init_MUTEX(&mm->context.sem);
-+ mm->context.size = 0;
-+ mm->context.has_foreign_mappings = 0;
-+ 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);
-+ }
-+ return retval;
++ for (apic = 0; apic < nr_ioapics; apic++)
++ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++)
++ clear_IO_APIC_pin(apic, pin);
+}
+
-+/*
-+ * No need to lock the MM as we are the last user
-+ */
-+void destroy_context(struct mm_struct *mm)
++#ifdef CONFIG_SMP
++static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t cpumask)
+{
-+ 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;
++ unsigned long flags;
++ int pin;
++ struct irq_pin_list *entry = irq_2_pin + irq;
++ unsigned int apicid_value;
++ cpumask_t tmp;
++
++ cpus_and(tmp, cpumask, cpu_online_map);
++ if (cpus_empty(tmp))
++ tmp = TARGET_CPUS;
++
++ cpus_and(cpumask, tmp, CPU_MASK_ALL);
++
++ apicid_value = cpu_mask_to_apicid(cpumask);
++ /* Prepare to do the io_apic_write */
++ apicid_value = apicid_value << 24;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ for (;;) {
++ pin = entry->pin;
++ if (pin == -1)
++ break;
++ io_apic_write(entry->apic, 0x10 + 1 + pin*2, apicid_value);
++ if (!entry->next)
++ break;
++ entry = irq_2_pin + entry->next;
+ }
++ set_native_irq_info(irq, cpumask);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
+}
+
-+static int read_ldt(void __user * ptr, unsigned long bytecount)
-+{
-+ int err;
-+ unsigned long size;
-+ struct mm_struct * mm = current->mm;
++#if defined(CONFIG_IRQBALANCE)
++# include <asm/processor.h> /* kernel_thread() */
++# include <linux/kernel_stat.h> /* kstat */
++# include <linux/slab.h> /* kmalloc() */
++# include <linux/timer.h> /* time_after() */
++
++#ifdef CONFIG_BALANCED_IRQ_DEBUG
++# define TDprintk(x...) do { printk("<%ld:%s:%d>: ", jiffies, __FILE__, __LINE__); printk(x); } while (0)
++# define Dprintk(x...) do { TDprintk(x); } while (0)
++# else
++# define TDprintk(x...)
++# define Dprintk(x...)
++# endif
++
++#define IRQBALANCE_CHECK_ARCH -999
++#define MAX_BALANCED_IRQ_INTERVAL (5*HZ)
++#define MIN_BALANCED_IRQ_INTERVAL (HZ/2)
++#define BALANCED_IRQ_MORE_DELTA (HZ/10)
++#define BALANCED_IRQ_LESS_DELTA (HZ)
+
-+ if (!mm->context.size)
-+ return 0;
-+ if (bytecount > LDT_ENTRY_SIZE*LDT_ENTRIES)
-+ bytecount = LDT_ENTRY_SIZE*LDT_ENTRIES;
++static int irqbalance_disabled __read_mostly = IRQBALANCE_CHECK_ARCH;
++static int physical_balance __read_mostly;
++static long balanced_irq_interval __read_mostly = MAX_BALANCED_IRQ_INTERVAL;
+
-+ down(&mm->context.sem);
-+ size = mm->context.size*LDT_ENTRY_SIZE;
-+ if (size > bytecount)
-+ size = bytecount;
++static struct irq_cpu_info {
++ unsigned long * last_irq;
++ unsigned long * irq_delta;
++ unsigned long irq;
++} irq_cpu_data[NR_CPUS];
+
-+ 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;
-+}
++#define CPU_IRQ(cpu) (irq_cpu_data[cpu].irq)
++#define LAST_CPU_IRQ(cpu,irq) (irq_cpu_data[cpu].last_irq[irq])
++#define IRQ_DELTA(cpu,irq) (irq_cpu_data[cpu].irq_delta[irq])
+
-+static int read_default_ldt(void __user * ptr, unsigned long bytecount)
-+{
-+ int err;
-+ unsigned long size;
-+ void *address;
++#define IDLE_ENOUGH(cpu,now) \
++ (idle_cpu(cpu) && ((now) - per_cpu(irq_stat, (cpu)).idle_timestamp > 1))
+
-+ err = 0;
-+ address = &default_ldt[0];
-+ size = 5*sizeof(struct desc_struct);
-+ if (size > bytecount)
-+ size = bytecount;
++#define IRQ_ALLOWED(cpu, allowed_mask) cpu_isset(cpu, allowed_mask)
+
-+ err = size;
-+ if (copy_to_user(ptr, address, size))
-+ err = -EFAULT;
++#define CPU_TO_PACKAGEINDEX(i) (first_cpu(cpu_sibling_map[i]))
+
-+ return err;
-+}
++static cpumask_t balance_irq_affinity[NR_IRQS] = {
++ [0 ... NR_IRQS-1] = CPU_MASK_ALL
++};
+
-+static int write_ldt(void __user * ptr, unsigned long bytecount, int oldmode)
++void set_balance_irq_affinity(unsigned int irq, cpumask_t mask)
+{
-+ struct mm_struct * mm = current->mm;
-+ __u32 entry_1, entry_2;
-+ 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, sizeof(ldt_info)))
-+ goto out;
++ balance_irq_affinity[irq] = mask;
++}
+
-+ 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;
-+ }
++static unsigned long move(int curr_cpu, cpumask_t allowed_mask,
++ unsigned long now, int direction)
++{
++ int search_idle = 1;
++ int cpu = curr_cpu;
+
-+ down(&mm->context.sem);
-+ if (ldt_info.entry_number >= mm->context.size) {
-+ error = alloc_ldt(¤t->mm->context, ldt_info.entry_number+1, 1);
-+ if (error < 0)
-+ goto out_unlock;
-+ }
++ goto inside;
+
-+ /* 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;
++ do {
++ if (unlikely(cpu == curr_cpu))
++ search_idle = 0;
++inside:
++ if (direction == 1) {
++ cpu++;
++ if (cpu >= NR_CPUS)
++ cpu = 0;
++ } else {
++ cpu--;
++ if (cpu == -1)
++ cpu = NR_CPUS-1;
+ }
-+ }
++ } while (!cpu_online(cpu) || !IRQ_ALLOWED(cpu,allowed_mask) ||
++ (search_idle && !IDLE_ENOUGH(cpu,now)));
+
-+ entry_1 = LDT_entry_a(&ldt_info);
-+ entry_2 = LDT_entry_b(&ldt_info);
-+ if (oldmode)
-+ entry_2 &= ~(1 << 20);
++ return cpu;
++}
+
-+ /* Install the new entry ... */
-+install:
-+ error = write_ldt_entry(mm->context.ldt, ldt_info.entry_number,
-+ entry_1, entry_2);
++static inline void balance_irq(int cpu, int irq)
++{
++ unsigned long now = jiffies;
++ cpumask_t allowed_mask;
++ unsigned int new_cpu;
++
++ if (irqbalance_disabled)
++ return;
+
-+out_unlock:
-+ up(&mm->context.sem);
-+out:
-+ return error;
++ cpus_and(allowed_mask, cpu_online_map, balance_irq_affinity[irq]);
++ new_cpu = move(cpu, allowed_mask, now, 1);
++ if (cpu != new_cpu) {
++ set_pending_irq(irq, cpumask_of_cpu(new_cpu));
++ }
+}
+
-+asmlinkage int sys_modify_ldt(int func, void __user *ptr, unsigned long bytecount)
++static inline void rotate_irqs_among_cpus(unsigned long useful_load_threshold)
+{
-+ 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;
++ int i, j;
++ Dprintk("Rotating IRQs among CPUs.\n");
++ for_each_online_cpu(i) {
++ for (j = 0; j < NR_IRQS; j++) {
++ if (!irq_desc[j].action)
++ continue;
++ /* Is it a significant load ? */
++ if (IRQ_DELTA(CPU_TO_PACKAGEINDEX(i),j) <
++ useful_load_threshold)
++ continue;
++ balance_irq(i, j);
++ }
+ }
-+ return ret;
++ balanced_irq_interval = max((long)MIN_BALANCED_IRQ_INTERVAL,
++ balanced_irq_interval - BALANCED_IRQ_LESS_DELTA);
++ return;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/Makefile linux-2.6.18-xen/arch/i386/kernel/Makefile
---- linux-2.6.18.3/arch/i386/kernel/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/kernel/Makefile 2006-11-19 14:26:21.000000000 +0100
-@@ -44,6 +44,12 @@
-
- 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 @@
-
- $(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 +83,19 @@
-
- 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 $(srctree)/scripts/Makefile.xen
++static void do_irq_balance(void)
++{
++ int i, j;
++ unsigned long max_cpu_irq = 0, min_cpu_irq = (~0);
++ unsigned long move_this_load = 0;
++ int max_loaded = 0, min_loaded = 0;
++ int load;
++ unsigned long useful_load_threshold = balanced_irq_interval + 10;
++ int selected_irq;
++ int tmp_loaded, first_attempt = 1;
++ unsigned long tmp_cpu_irq;
++ unsigned long imbalance = 0;
++ cpumask_t allowed_mask, target_cpu_mask, tmp;
+
-+obj-y += fixup.o
-+microcode-$(subst m,y,$(CONFIG_MICROCODE)) := microcode-xen.o
-+n-obj-xen := i8259.o timers/ reboot.o smpboot.o trampoline.o
++ for_each_possible_cpu(i) {
++ int package_index;
++ CPU_IRQ(i) = 0;
++ if (!cpu_online(i))
++ continue;
++ package_index = CPU_TO_PACKAGEINDEX(i);
++ for (j = 0; j < NR_IRQS; j++) {
++ unsigned long value_now, delta;
++ /* Is this an active IRQ? */
++ if (!irq_desc[j].action)
++ continue;
++ if ( package_index == i )
++ IRQ_DELTA(package_index,j) = 0;
++ /* Determine the total count per processor per IRQ */
++ value_now = (unsigned long) kstat_cpu(i).irqs[j];
+
-+obj-y := $(call filterxen, $(obj-y), $(n-obj-xen))
-+obj-y := $(call cherrypickxen, $(obj-y))
-+extra-y := $(call cherrypickxen, $(extra-y))
-+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/microcode-xen.c linux-2.6.18-xen/arch/i386/kernel/microcode-xen.c
---- linux-2.6.18.3/arch/i386/kernel/microcode-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/microcode-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,147 @@
-+/*
-+ * Intel CPU Microcode Update Driver for Linux
-+ *
-+ * Copyright (C) 2000-2004 Tigran Aivazian
-+ *
-+ * This driver allows to upgrade microcode on Intel processors
-+ * belonging to IA-32 family - PentiumPro, Pentium II,
-+ * Pentium III, Xeon, Pentium 4, etc.
-+ *
-+ * Reference: Section 8.10 of Volume III, Intel Pentium 4 Manual,
-+ * Order Number 245472 or free download from:
-+ *
-+ * http://developer.intel.com/design/pentium4/manuals/245472.htm
-+ *
-+ * For more information, go to http://www.urbanmyth.org/microcode
-+ *
-+ * 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.
-+ */
++ /* Determine the activity per processor per IRQ */
++ delta = value_now - LAST_CPU_IRQ(i,j);
+
-+//#define DEBUG /* pr_debug */
-+#include <linux/capability.h>
-+#include <linux/kernel.h>
-+#include <linux/init.h>
-+#include <linux/sched.h>
-+#include <linux/cpumask.h>
-+#include <linux/module.h>
-+#include <linux/slab.h>
-+#include <linux/vmalloc.h>
-+#include <linux/miscdevice.h>
-+#include <linux/spinlock.h>
-+#include <linux/mm.h>
-+#include <linux/mutex.h>
-+#include <linux/syscalls.h>
++ /* Update last_cpu_irq[][] for the next time */
++ LAST_CPU_IRQ(i,j) = value_now;
+
-+#include <asm/msr.h>
-+#include <asm/uaccess.h>
-+#include <asm/processor.h>
++ /* Ignore IRQs whose rate is less than the clock */
++ if (delta < useful_load_threshold)
++ continue;
++ /* update the load for the processor or package total */
++ IRQ_DELTA(package_index,j) += delta;
+
-+MODULE_DESCRIPTION("Intel CPU (IA-32) Microcode Update Driver");
-+MODULE_AUTHOR("Tigran Aivazian <tigran at veritas.com>");
-+MODULE_LICENSE("GPL");
++ /* Keep track of the higher numbered sibling as well */
++ if (i != package_index)
++ CPU_IRQ(i) += delta;
++ /*
++ * We have sibling A and sibling B in the package
++ *
++ * cpu_irq[A] = load for cpu A + load for cpu B
++ * cpu_irq[B] = load for cpu B
++ */
++ CPU_IRQ(package_index) += delta;
++ }
++ }
++ /* Find the least loaded processor package */
++ for_each_online_cpu(i) {
++ if (i != CPU_TO_PACKAGEINDEX(i))
++ continue;
++ if (min_cpu_irq > CPU_IRQ(i)) {
++ min_cpu_irq = CPU_IRQ(i);
++ min_loaded = i;
++ }
++ }
++ max_cpu_irq = ULONG_MAX;
+
-+#define MICROCODE_VERSION "1.14-xen"
++tryanothercpu:
++ /* Look for heaviest loaded processor.
++ * We may come back to get the next heaviest loaded processor.
++ * Skip processors with trivial loads.
++ */
++ tmp_cpu_irq = 0;
++ tmp_loaded = -1;
++ for_each_online_cpu(i) {
++ if (i != CPU_TO_PACKAGEINDEX(i))
++ continue;
++ if (max_cpu_irq <= CPU_IRQ(i))
++ continue;
++ if (tmp_cpu_irq < CPU_IRQ(i)) {
++ tmp_cpu_irq = CPU_IRQ(i);
++ tmp_loaded = i;
++ }
++ }
+
-+#define DEFAULT_UCODE_DATASIZE (2000) /* 2000 bytes */
-+#define MC_HEADER_SIZE (sizeof (microcode_header_t)) /* 48 bytes */
-+#define DEFAULT_UCODE_TOTALSIZE (DEFAULT_UCODE_DATASIZE + MC_HEADER_SIZE) /* 2048 bytes */
++ if (tmp_loaded == -1) {
++ /* In the case of small number of heavy interrupt sources,
++ * loading some of the cpus too much. We use Ingo's original
++ * approach to rotate them around.
++ */
++ if (!first_attempt && imbalance >= useful_load_threshold) {
++ rotate_irqs_among_cpus(useful_load_threshold);
++ return;
++ }
++ goto not_worth_the_effort;
++ }
++
++ first_attempt = 0; /* heaviest search */
++ max_cpu_irq = tmp_cpu_irq; /* load */
++ max_loaded = tmp_loaded; /* processor */
++ imbalance = (max_cpu_irq - min_cpu_irq) / 2;
++
++ Dprintk("max_loaded cpu = %d\n", max_loaded);
++ Dprintk("min_loaded cpu = %d\n", min_loaded);
++ Dprintk("max_cpu_irq load = %ld\n", max_cpu_irq);
++ Dprintk("min_cpu_irq load = %ld\n", min_cpu_irq);
++ Dprintk("load imbalance = %lu\n", imbalance);
+
-+/* no concurrent ->write()s are allowed on /dev/cpu/microcode */
-+static DEFINE_MUTEX(microcode_mutex);
++ /* if imbalance is less than approx 10% of max load, then
++ * observe diminishing returns action. - quit
++ */
++ if (imbalance < (max_cpu_irq >> 3)) {
++ Dprintk("Imbalance too trivial\n");
++ goto not_worth_the_effort;
++ }
+
-+static void __user *user_buffer; /* user area microcode data buffer */
-+static unsigned int user_buffer_size; /* it's size */
-+
-+static int microcode_open (struct inode *unused1, struct file *unused2)
-+{
-+ return capable(CAP_SYS_RAWIO) ? 0 : -EPERM;
-+}
++tryanotherirq:
++ /* if we select an IRQ to move that can't go where we want, then
++ * see if there is another one to try.
++ */
++ move_this_load = 0;
++ selected_irq = -1;
++ for (j = 0; j < NR_IRQS; j++) {
++ /* Is this an active IRQ? */
++ if (!irq_desc[j].action)
++ continue;
++ if (imbalance <= IRQ_DELTA(max_loaded,j))
++ continue;
++ /* Try to find the IRQ that is closest to the imbalance
++ * without going over.
++ */
++ if (move_this_load < IRQ_DELTA(max_loaded,j)) {
++ move_this_load = IRQ_DELTA(max_loaded,j);
++ selected_irq = j;
++ }
++ }
++ if (selected_irq == -1) {
++ goto tryanothercpu;
++ }
+
++ imbalance = move_this_load;
++
++ /* For physical_balance case, we accumlated both load
++ * values in the one of the siblings cpu_irq[],
++ * to use the same code for physical and logical processors
++ * as much as possible.
++ *
++ * NOTE: the cpu_irq[] array holds the sum of the load for
++ * sibling A and sibling B in the slot for the lowest numbered
++ * sibling (A), _AND_ the load for sibling B in the slot for
++ * the higher numbered sibling.
++ *
++ * We seek the least loaded sibling by making the comparison
++ * (A+B)/2 vs B
++ */
++ load = CPU_IRQ(min_loaded) >> 1;
++ for_each_cpu_mask(j, cpu_sibling_map[min_loaded]) {
++ if (load > CPU_IRQ(j)) {
++ /* This won't change cpu_sibling_map[min_loaded] */
++ load = CPU_IRQ(j);
++ min_loaded = j;
++ }
++ }
+
-+static int do_microcode_update (void)
-+{
-+ int err;
-+ dom0_op_t op;
++ cpus_and(allowed_mask,
++ cpu_online_map,
++ balance_irq_affinity[selected_irq]);
++ target_cpu_mask = cpumask_of_cpu(min_loaded);
++ cpus_and(tmp, target_cpu_mask, allowed_mask);
+
-+ err = sys_mlock((unsigned long)user_buffer, user_buffer_size);
-+ if (err != 0)
-+ return err;
++ if (!cpus_empty(tmp)) {
+
-+ op.cmd = DOM0_MICROCODE;
-+ set_xen_guest_handle(op.u.microcode.data, user_buffer);
-+ op.u.microcode.length = user_buffer_size;
-+ err = HYPERVISOR_dom0_op(&op);
++ Dprintk("irq = %d moved to cpu = %d\n",
++ selected_irq, min_loaded);
++ /* mark for change destination */
++ set_pending_irq(selected_irq, cpumask_of_cpu(min_loaded));
+
-+ (void)sys_munlock((unsigned long)user_buffer, user_buffer_size);
++ /* Since we made a change, come back sooner to
++ * check for more variation.
++ */
++ balanced_irq_interval = max((long)MIN_BALANCED_IRQ_INTERVAL,
++ balanced_irq_interval - BALANCED_IRQ_LESS_DELTA);
++ return;
++ }
++ goto tryanotherirq;
+
-+ return err;
++not_worth_the_effort:
++ /*
++ * if we did not find an IRQ to move, then adjust the time interval
++ * upward
++ */
++ balanced_irq_interval = min((long)MAX_BALANCED_IRQ_INTERVAL,
++ balanced_irq_interval + BALANCED_IRQ_MORE_DELTA);
++ Dprintk("IRQ worth rotating not found\n");
++ return;
+}
+
-+static ssize_t microcode_write (struct file *file, const char __user *buf, size_t len, loff_t *ppos)
++static int balanced_irq(void *unused)
+{
-+ ssize_t ret;
++ int i;
++ unsigned long prev_balance_time = jiffies;
++ long time_remaining = balanced_irq_interval;
+
-+ if (len < DEFAULT_UCODE_TOTALSIZE) {
-+ printk(KERN_ERR "microcode: not enough data\n");
-+ return -EINVAL;
++ daemonize("kirqd");
++
++ /* push everything to CPU 0 to give us a starting point. */
++ for (i = 0 ; i < NR_IRQS ; i++) {
++ irq_desc[i].pending_mask = cpumask_of_cpu(0);
++ set_pending_irq(i, cpumask_of_cpu(0));
+ }
+
-+ if ((len >> PAGE_SHIFT) > num_physpages) {
-+ printk(KERN_ERR "microcode: too much data (max %ld pages)\n", num_physpages);
-+ return -EINVAL;
++ for ( ; ; ) {
++ time_remaining = schedule_timeout_interruptible(time_remaining);
++ try_to_freeze();
++ if (time_after(jiffies,
++ prev_balance_time+balanced_irq_interval)) {
++ preempt_disable();
++ do_irq_balance();
++ prev_balance_time = jiffies;
++ time_remaining = balanced_irq_interval;
++ preempt_enable();
++ }
+ }
-+
-+ mutex_lock(µcode_mutex);
-+
-+ user_buffer = (void __user *) buf;
-+ user_buffer_size = (int) len;
-+
-+ ret = do_microcode_update();
-+ if (!ret)
-+ ret = (ssize_t)len;
-+
-+ mutex_unlock(µcode_mutex);
-+
-+ return ret;
++ return 0;
+}
+
-+static struct file_operations microcode_fops = {
-+ .owner = THIS_MODULE,
-+ .write = microcode_write,
-+ .open = microcode_open,
-+};
-+
-+static struct miscdevice microcode_dev = {
-+ .minor = MICROCODE_MINOR,
-+ .name = "microcode",
-+ .fops = µcode_fops,
-+};
-+
-+static int __init microcode_init (void)
++static int __init balanced_irq_init(void)
+{
-+ int error;
++ int i;
++ struct cpuinfo_x86 *c;
++ cpumask_t tmp;
+
-+ error = misc_register(µcode_dev);
-+ if (error) {
-+ printk(KERN_ERR
-+ "microcode: can't misc_register on minor=%d\n",
-+ MICROCODE_MINOR);
-+ return error;
++ cpus_shift_right(tmp, cpu_online_map, 2);
++ c = &boot_cpu_data;
++ /* When not overwritten by the command line ask subarchitecture. */
++ if (irqbalance_disabled == IRQBALANCE_CHECK_ARCH)
++ irqbalance_disabled = NO_BALANCE_IRQ;
++ if (irqbalance_disabled)
++ return 0;
++
++ /* disable irqbalance completely if there is only one processor online */
++ if (num_online_cpus() < 2) {
++ irqbalance_disabled = 1;
++ return 0;
+ }
++ /*
++ * Enable physical balance only if more than 1 physical processor
++ * is present
++ */
++ if (smp_num_siblings > 1 && !cpus_empty(tmp))
++ physical_balance = 1;
+
-+ printk(KERN_INFO
-+ "IA-32 Microcode Update Driver: v" MICROCODE_VERSION " <tigran at veritas.com>\n");
++ for_each_online_cpu(i) {
++ irq_cpu_data[i].irq_delta = kmalloc(sizeof(unsigned long) * NR_IRQS, GFP_KERNEL);
++ irq_cpu_data[i].last_irq = kmalloc(sizeof(unsigned long) * NR_IRQS, GFP_KERNEL);
++ if (irq_cpu_data[i].irq_delta == NULL || irq_cpu_data[i].last_irq == NULL) {
++ printk(KERN_ERR "balanced_irq_init: out of memory");
++ goto failed;
++ }
++ memset(irq_cpu_data[i].irq_delta,0,sizeof(unsigned long) * NR_IRQS);
++ memset(irq_cpu_data[i].last_irq,0,sizeof(unsigned long) * NR_IRQS);
++ }
++
++ printk(KERN_INFO "Starting balanced_irq\n");
++ if (kernel_thread(balanced_irq, NULL, CLONE_KERNEL) >= 0)
++ return 0;
++ else
++ printk(KERN_ERR "balanced_irq_init: failed to spawn balanced_irq");
++failed:
++ for_each_possible_cpu(i) {
++ kfree(irq_cpu_data[i].irq_delta);
++ irq_cpu_data[i].irq_delta = NULL;
++ kfree(irq_cpu_data[i].last_irq);
++ irq_cpu_data[i].last_irq = NULL;
++ }
+ return 0;
+}
+
-+static void __exit microcode_exit (void)
++int __init irqbalance_disable(char *str)
+{
-+ misc_deregister(µcode_dev);
++ irqbalance_disabled = 1;
++ return 1;
+}
+
-+module_init(microcode_init)
-+module_exit(microcode_exit)
-+MODULE_ALIAS_MISCDEV(MICROCODE_MINOR);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/mpparse-xen.c linux-2.6.18-xen/arch/i386/kernel/mpparse-xen.c
---- linux-2.6.18.3/arch/i386/kernel/mpparse-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/mpparse-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,1185 @@
-+/*
-+ * 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
-+ */
++__setup("noirqbalance", irqbalance_disable);
+
-+#include <linux/mm.h>
-+#include <linux/init.h>
-+#include <linux/acpi.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/bitops.h>
++late_initcall(balanced_irq_init);
++#endif /* CONFIG_IRQBALANCE */
++#endif /* CONFIG_SMP */
++#endif /* !CONFIG_XEN */
+
-+#include <asm/smp.h>
-+#include <asm/acpi.h>
-+#include <asm/mtrr.h>
-+#include <asm/mpspec.h>
-+#include <asm/io_apic.h>
++#ifndef CONFIG_SMP
++void fastcall send_IPI_self(int vector)
++{
++#ifndef CONFIG_XEN
++ unsigned int cfg;
+
-+#include <mach_apic.h>
-+#include <mach_mpparse.h>
-+#include <bios_ebda.h>
++ /*
++ * Wait for idle.
++ */
++ apic_wait_icr_idle();
++ cfg = APIC_DM_FIXED | APIC_DEST_SELF | vector | APIC_DEST_LOGICAL;
++ /*
++ * Send the IPI. The write to APIC_ICR fires this off.
++ */
++ apic_write_around(APIC_ICR, cfg);
++#endif
++}
++#endif /* !CONFIG_SMP */
+
-+/* Have we found an MP table */
-+int smp_found_config;
-+unsigned int __initdata maxcpus = NR_CPUS;
+
+/*
-+ * Various Linux-internal data structures created from the
-+ * MP-table.
++ * support for broken MP BIOSs, enables hand-redirection of PIRQ0-7 to
++ * specific CPU-side IRQs.
+ */
-+int apic_version [MAX_APICS];
-+int mp_bus_id_to_type [MAX_MP_BUSSES];
-+int mp_bus_id_to_node [MAX_MP_BUSSES];
-+int mp_bus_id_to_local [MAX_MP_BUSSES];
-+int quad_local_to_mp_bus_id [NR_CPUS/4][4];
-+int mp_bus_id_to_pci_bus [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 };
-+static int mp_current_pci_id;
+
-+/* 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;
++#define MAX_PIRQS 8
++static int pirq_entries [MAX_PIRQS];
++static int pirqs_enabled;
++int skip_ioapic_setup;
+
-+int nr_ioapics;
++static int __init ioapic_setup(char *str)
++{
++ skip_ioapic_setup = 1;
++ return 1;
++}
+
-+int pic_mode;
-+unsigned long mp_lapic_addr;
++__setup("noapic", ioapic_setup);
+
-+unsigned int def_to_bigsmp = 0;
++static int __init ioapic_pirq_setup(char *str)
++{
++ int i, max;
++ int ints[MAX_PIRQS+1];
+
-+/* Processor that is doing the boot up */
-+unsigned int boot_cpu_physical_apicid = -1U;
-+/* Internal processor count */
-+static unsigned int __devinitdata num_processors;
++ get_options(str, ARRAY_SIZE(ints), ints);
+
-+/* Bitmask of physically existing CPUs */
-+physid_mask_t phys_cpu_present_map;
++ for (i = 0; i < MAX_PIRQS; i++)
++ pirq_entries[i] = -1;
+
-+u8 bios_cpu_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = BAD_APICID };
++ pirqs_enabled = 1;
++ apic_printk(APIC_VERBOSE, KERN_INFO
++ "PIRQ redirection, working around broken MP-BIOS.\n");
++ max = MAX_PIRQS;
++ if (ints[0] < MAX_PIRQS)
++ max = ints[0];
+
-+/*
-+ * Intel MP BIOS table parsing routines:
-+ */
++ for (i = 0; i < max; i++) {
++ apic_printk(APIC_VERBOSE, KERN_DEBUG
++ "... 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);
+
+/*
-+ * Checksum an MP configuration block.
++ * Find the IRQ entry number of a certain pin.
+ */
-+
-+static int __init mpf_checksum(unsigned char *mp, int len)
++static int find_irq_entry(int apic, int pin, int type)
+{
-+ int sum = 0;
++ int i;
+
-+ while (len--)
-+ sum += *mp++;
++ 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 sum & 0xFF;
++ return -1;
+}
+
+/*
-+ * Have to match translation table entries to main table entries by counter
-+ * hence the mpc_record variable .... can't see a less disgusting way of
-+ * doing this ....
++ * Find the pin to which IRQ[irq] (ISA) is connected
+ */
++static int __init find_isa_irq_pin(int irq, int type)
++{
++ int i;
+
-+static int mpc_record;
-+static struct mpc_config_translation *translation_table[MAX_MPC_ENTRY] __initdata;
++ for (i = 0; i < mp_irq_entries; i++) {
++ int lbus = mp_irqs[i].mpc_srcbus;
+
-+#ifndef CONFIG_XEN
-+static void __devinit MP_processor_info (struct mpc_config_processor *m)
-+{
-+ int ver, apicid;
-+ physid_mask_t phys_cpu;
-+
-+ if (!(m->mpc_cpuflag & CPU_ENABLED))
-+ return;
-+
-+ apicid = mpc_apic_id(m, translation_table[mpc_record]);
-+
-+ if (m->mpc_featureflag&(1<<0))
-+ Dprintk(" Floating point unit present.\n");
-+ if (m->mpc_featureflag&(1<<7))
-+ Dprintk(" Machine Exception supported.\n");
-+ if (m->mpc_featureflag&(1<<8))
-+ Dprintk(" 64 bit compare & exchange supported.\n");
-+ if (m->mpc_featureflag&(1<<9))
-+ Dprintk(" Internal APIC present.\n");
-+ if (m->mpc_featureflag&(1<<11))
-+ Dprintk(" SEP present.\n");
-+ if (m->mpc_featureflag&(1<<12))
-+ Dprintk(" MTRR present.\n");
-+ if (m->mpc_featureflag&(1<<13))
-+ Dprintk(" PGE present.\n");
-+ if (m->mpc_featureflag&(1<<14))
-+ Dprintk(" MCA present.\n");
-+ if (m->mpc_featureflag&(1<<15))
-+ Dprintk(" CMOV present.\n");
-+ if (m->mpc_featureflag&(1<<16))
-+ Dprintk(" PAT present.\n");
-+ if (m->mpc_featureflag&(1<<17))
-+ Dprintk(" PSE present.\n");
-+ if (m->mpc_featureflag&(1<<18))
-+ Dprintk(" PSN present.\n");
-+ if (m->mpc_featureflag&(1<<19))
-+ Dprintk(" Cache Line Flush Instruction present.\n");
-+ /* 20 Reserved */
-+ if (m->mpc_featureflag&(1<<21))
-+ Dprintk(" Debug Trace and EMON Store present.\n");
-+ if (m->mpc_featureflag&(1<<22))
-+ Dprintk(" ACPI Thermal Throttle Registers present.\n");
-+ if (m->mpc_featureflag&(1<<23))
-+ Dprintk(" MMX present.\n");
-+ if (m->mpc_featureflag&(1<<24))
-+ Dprintk(" FXSR present.\n");
-+ if (m->mpc_featureflag&(1<<25))
-+ Dprintk(" XMM present.\n");
-+ if (m->mpc_featureflag&(1<<26))
-+ Dprintk(" Willamette New Instructions present.\n");
-+ if (m->mpc_featureflag&(1<<27))
-+ Dprintk(" Self Snoop present.\n");
-+ if (m->mpc_featureflag&(1<<28))
-+ Dprintk(" HT present.\n");
-+ if (m->mpc_featureflag&(1<<29))
-+ Dprintk(" Thermal Monitor present.\n");
-+ /* 30, 31 Reserved */
-+
-+
-+ if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) {
-+ Dprintk(" Bootup CPU\n");
-+ boot_cpu_physical_apicid = m->mpc_apicid;
-+ }
-+
-+ ver = m->mpc_apicver;
-+
-+ /*
-+ * Validate version
-+ */
-+ if (ver == 0x0) {
-+ printk(KERN_WARNING "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;
-+
-+ phys_cpu = apicid_to_cpu_present(apicid);
-+ physids_or(phys_cpu_present_map, phys_cpu_present_map, phys_cpu);
-+
-+ if (num_processors >= NR_CPUS) {
-+ printk(KERN_WARNING "WARNING: NR_CPUS limit of %i reached."
-+ " Processor ignored.\n", NR_CPUS);
-+ return;
-+ }
-+
-+ if (num_processors >= maxcpus) {
-+ printk(KERN_WARNING "WARNING: maxcpus limit of %i reached."
-+ " Processor ignored.\n", maxcpus);
-+ return;
-+ }
-+
-+ cpu_set(num_processors, cpu_possible_map);
-+ num_processors++;
-+
-+ /*
-+ * Would be preferable to switch to bigsmp when CONFIG_HOTPLUG_CPU=y
-+ * but we need to work other dependencies like SMP_SUSPEND etc
-+ * before this can be done without some confusion.
-+ * if (CPU_HOTPLUG_ENABLED || num_processors > 8)
-+ * - Ashok Raj <ashok.raj at intel.com>
-+ */
-+ if (num_processors > 8) {
-+ switch (boot_cpu_data.x86_vendor) {
-+ case X86_VENDOR_INTEL:
-+ if (!APIC_XAPIC(ver)) {
-+ def_to_bigsmp = 0;
-+ break;
-+ }
-+ /* If P4 and above fall through */
-+ case X86_VENDOR_AMD:
-+ def_to_bigsmp = 1;
-+ }
++ 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;
+ }
-+ bios_cpu_apicid[num_processors - 1] = m->mpc_apicid;
-+}
-+#else
-+void __init MP_processor_info (struct mpc_config_processor *m)
-+{
-+ num_processors++;
++ return -1;
+}
-+#endif /* CONFIG_XEN */
+
-+static void __init MP_bus_info (struct mpc_config_bus *m)
++static int __init find_isa_irq_apic(int irq, int type)
+{
-+ char str[7];
-+
-+ memcpy(str, m->mpc_bustype, 6);
-+ str[6] = 0;
++ int i;
+
-+ mpc_oem_bus_info(m, str, translation_table[mpc_record]);
++ for (i = 0; i < mp_irq_entries; i++) {
++ int lbus = mp_irqs[i].mpc_srcbus;
+
-+ if (m->mpc_busid >= MAX_MP_BUSSES) {
-+ printk(KERN_WARNING "MP table busid value (%d) for bustype %s "
-+ " is too large, max. supported is %d\n",
-+ m->mpc_busid, str, MAX_MP_BUSSES - 1);
-+ return;
++ 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 (strncmp(str, BUSTYPE_ISA, sizeof(BUSTYPE_ISA)-1) == 0) {
-+ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_ISA;
-+ } else if (strncmp(str, BUSTYPE_EISA, sizeof(BUSTYPE_EISA)-1) == 0) {
-+ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_EISA;
-+ } else if (strncmp(str, BUSTYPE_PCI, sizeof(BUSTYPE_PCI)-1) == 0) {
-+ mpc_oem_pci_bus(m, translation_table[mpc_record]);
-+ 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, BUSTYPE_MCA, sizeof(BUSTYPE_MCA)-1) == 0) {
-+ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_MCA;
-+ } else if (strncmp(str, BUSTYPE_NEC98, sizeof(BUSTYPE_NEC98)-1) == 0) {
-+ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_NEC98;
-+ } else {
-+ printk(KERN_WARNING "Unknown bustype %s - ignoring\n", str);
++ 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;
+}
+
-+static void __init MP_ioapic_info (struct mpc_config_ioapic *m)
++/*
++ * 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)
+{
-+ if (!(m->mpc_flags & MPC_APIC_USABLE))
-+ return;
++ int apic, i, best_guess = -1;
+
-+ printk(KERN_INFO "I/O APIC #%d Version %d at 0x%lX.\n",
-+ m->mpc_apicid, m->mpc_apicver, m->mpc_apicaddr);
-+ if (nr_ioapics >= MAX_IO_APICS) {
-+ printk(KERN_CRIT "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;
++ 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) {
++ printk(KERN_WARNING "PCI BIOS passed nonexistent PCI bus %d!\n", bus);
++ return -1;
+ }
-+ mp_ioapics[nr_ioapics] = *m;
-+ nr_ioapics++;
-+}
++ for (i = 0; i < mp_irq_entries; i++) {
++ int lbus = mp_irqs[i].mpc_srcbus;
+
-+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");
-+}
++ 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;
+
-+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();
-+}
++ 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);
+
-+#ifdef CONFIG_X86_NUMAQ
-+static void __init MP_translation_info (struct mpc_config_translation *m)
-+{
-+ printk(KERN_INFO "Translation: record %d, type %d, quad %d, global %d, local %d\n", mpc_record, m->trans_type, m->trans_quad, m->trans_global, m->trans_local);
++ if (!(apic || IO_APIC_IRQ(irq)))
++ continue;
+
-+ if (mpc_record >= MAX_MPC_ENTRY)
-+ printk(KERN_ERR "MAX_MPC_ENTRY exceeded!\n");
-+ else
-+ translation_table[mpc_record] = m; /* stash this for later */
-+ if (m->trans_quad < MAX_NUMNODES && !node_online(m->trans_quad))
-+ node_set_online(m->trans_quad);
++ 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;
++ }
++ }
++ return best_guess;
+}
++EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector);
+
+/*
-+ * Read/parse the MPC oem tables
++ * 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
+ */
-+
-+static void __init smp_read_mpc_oem(struct mp_config_oemtable *oemtable, \
-+ unsigned short oemsize)
++#ifdef CONFIG_SMP
++#ifndef CONFIG_XEN
++void __init setup_ioapic_dest(void)
+{
-+ int count = sizeof (*oemtable); /* the header size */
-+ unsigned char *oemptr = ((unsigned char *)oemtable)+count;
-+
-+ mpc_record = 0;
-+ printk(KERN_INFO "Found an OEM MPC table at %8p - parsing it ... \n", oemtable);
-+ if (memcmp(oemtable->oem_signature,MPC_OEM_SIGNATURE,4))
-+ {
-+ printk(KERN_WARNING "SMP mpc oemtable: bad signature [%c%c%c%c]!\n",
-+ oemtable->oem_signature[0],
-+ oemtable->oem_signature[1],
-+ oemtable->oem_signature[2],
-+ oemtable->oem_signature[3]);
-+ return;
-+ }
-+ if (mpf_checksum((unsigned char *)oemtable,oemtable->oem_length))
-+ {
-+ printk(KERN_WARNING "SMP oem mptable: checksum error!\n");
++ int pin, ioapic, irq, irq_entry;
++
++ if (skip_ioapic_setup == 1)
+ return;
-+ }
-+ while (count < oemtable->oem_length) {
-+ switch (*oemptr) {
-+ case MP_TRANSLATION:
-+ {
-+ struct mpc_config_translation *m=
-+ (struct mpc_config_translation *)oemptr;
-+ MP_translation_info(m);
-+ oemptr += sizeof(*m);
-+ count += sizeof(*m);
-+ ++mpc_record;
-+ break;
-+ }
-+ default:
-+ {
-+ printk(KERN_WARNING "Unrecognised OEM table entry type! - %d\n", (int) *oemptr);
-+ 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);
+ }
-+ }
-+}
+
-+static inline void mps_oem_check(struct mp_config_table *mpc, char *oem,
-+ char *productid)
-+{
-+ if (strncmp(oem, "IBM NUMA", 8))
-+ printk("Warning! May not be a NUMA-Q system!\n");
-+ if (mpc->mpc_oemptr)
-+ smp_read_mpc_oem((struct mp_config_oemtable *) mpc->mpc_oemptr,
-+ mpc->mpc_oemsize);
++ }
+}
-+#endif /* CONFIG_X86_NUMAQ */
++#endif /* !CONFIG_XEN */
++#endif
+
+/*
-+ * Read/parse the MPC
++ * EISA Edge/Level control register, ELCR
+ */
-+
-+static int __init smp_read_mpc(struct mp_config_table *mpc)
++static int EISA_ELCR(unsigned int irq)
+{
-+ char str[16];
-+ char oem[10];
-+ int count=sizeof(*mpc);
-+ unsigned char *mpt=((unsigned char *)mpc)+count;
-+
-+ if (memcmp(mpc->mpc_signature,MPC_SIGNATURE,4)) {
-+ printk(KERN_ERR "SMP mptable: bad signature [0x%x]!\n",
-+ *(u32 *)mpc->mpc_signature);
-+ return 0;
-+ }
-+ if (mpf_checksum((unsigned char *)mpc,mpc->mpc_length)) {
-+ printk(KERN_ERR "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 (irq < 16) {
++ unsigned int port = 0x4d0 + (irq >> 3);
++ return (inb(port) >> (irq & 7)) & 1;
+ }
-+ memcpy(oem,mpc->mpc_oem,8);
-+ oem[8]=0;
-+ printk(KERN_INFO "OEM ID: %s ",oem);
++ apic_printk(APIC_VERBOSE, KERN_INFO
++ "Broken MPtable reports ISA irq %d\n", irq);
++ return 0;
++}
+
-+ memcpy(str,mpc->mpc_productid,12);
-+ str[12]=0;
-+ printk("Product ID: %s ",str);
++/* 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 */
+
-+ mps_oem_check(mpc, oem, str);
++#define default_EISA_trigger(idx) (EISA_ELCR(mp_irqs[idx].mpc_srcbusirq))
++#define default_EISA_polarity(idx) (0)
+
-+ printk("APIC at: 0x%lX\n",mpc->mpc_lapic);
++/* ISA interrupts are always polarity zero edge triggered,
++ * when listed as conforming in the MP table. */
+
-+ /*
-+ * Save the local APIC address (it might be non-default) -- but only
-+ * if we're not using ACPI.
-+ */
-+ if (!acpi_lapic)
-+ mp_lapic_addr = mpc->mpc_lapic;
++#define default_ISA_trigger(idx) (0)
++#define default_ISA_polarity(idx) (0)
+
-+ /*
-+ * Now process the configuration blocks.
-+ */
-+ mpc_record = 0;
-+ while (count < mpc->mpc_length) {
-+ switch(*mpt) {
-+ case MP_PROCESSOR:
-+ {
-+ struct mpc_config_processor *m=
-+ (struct mpc_config_processor *)mpt;
-+ /* ACPI may have already provided this data */
-+ 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;
++/* PCI interrupts are always polarity one level triggered,
++ * when listed as conforming in the MP table. */
+
-+ 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;
-+ }
-+ default:
-+ {
-+ count = mpc->mpc_length;
-+ break;
-+ }
-+ }
-+ ++mpc_record;
-+ }
-+ clustered_apic_check();
-+ if (!num_processors)
-+ printk(KERN_ERR "SMP mptable: no processors registered!\n");
-+ return num_processors;
-+}
++#define default_PCI_trigger(idx) (1)
++#define default_PCI_polarity(idx) (1)
+
-+static int __init ELCR_trigger(unsigned int irq)
-+{
-+ unsigned int port;
++/* MCA interrupts are always polarity zero level triggered,
++ * when listed as conforming in the MP table. */
+
-+ port = 0x4d0 + (irq >> 3);
-+ return (inb(port) >> (irq & 7)) & 1;
-+}
++#define default_MCA_trigger(idx) (1)
++#define default_MCA_polarity(idx) (0)
+
-+static void __init construct_default_ioirq_mptable(int mpc_default_type)
++static int __init MPBIOS_polarity(int idx)
+{
-+ 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;
++ int bus = mp_irqs[idx].mpc_srcbus;
++ int polarity;
+
+ /*
-+ * 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.
++ * Determine IRQ line polarity (high active or low active):
+ */
-+ 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_WARNING "ELCR contains invalid data... not using ELCR\n");
-+ else {
-+ printk(KERN_INFO "Using ELCR to identify PCI interrupts\n");
-+ ELCR_fallback = 1;
++ 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;
+ }
-+ }
-+
-+ 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 */
++ case 1: /* high active */
++ {
++ polarity = 0;
++ 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;
++ 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;
+ }
-+
-+ 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);
++ return polarity;
+}
+
-+static inline void __init construct_default_ISA_mptable(int mpc_default_type)
++static int MPBIOS_trigger(int idx)
+{
-+ 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;
++ int bus = mp_irqs[idx].mpc_srcbus;
++ int trigger;
+
+ /*
-+ * 2 CPUs, numbered 0 & 1.
++ * Determine IRQ trigger mode (edge or level sensitive):
+ */
-+ 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("???\n");
-+ printk(KERN_ERR "Unknown standard configuration %d\n",
-+ mpc_default_type);
-+ /* fall through */
-+ case 1:
-+ case 5:
-+ memcpy(bus.mpc_bustype, "ISA ", 6);
++ 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 2:
-+ case 6:
-+ case 3:
-+ memcpy(bus.mpc_bustype, "EISA ", 6);
++ }
++ case 1: /* edge */
++ {
++ trigger = 0;
+ 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);
++ }
++ 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 struct intel_mp_floating *mpf_found;
-+
-+/*
-+ * Scan the memory blocks for an SMP configuration block.
-+ */
-+void __init get_smp_config (void)
++static inline int irq_polarity(int idx)
+{
-+ struct intel_mp_floating *mpf = mpf_found;
++ return MPBIOS_polarity(idx);
++}
+
-+ /*
-+ * 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");
++static inline int irq_trigger(int idx)
++{
++ return MPBIOS_trigger(idx);
++}
+
-+ printk(KERN_INFO "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;
-+ }
++static int pin_2_irq(int idx, int apic, int pin)
++{
++ int irq, i;
++ int bus = mp_irqs[idx].mpc_srcbus;
+
+ /*
-+ * Now see if we need to read further.
++ * Debugging check, we are in big trouble if this message pops up!
+ */
-+ 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) {
++ if (mp_irqs[idx].mpc_dstirq != pin)
++ printk(KERN_ERR "broken BIOS or MPTABLE parser, ayiee!!\n");
+
-+ /*
-+ * 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;
++ 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;
+ }
-+ /*
-+ * 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");
++ 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;
+
-+ bus.mpc_type = MP_BUS;
-+ bus.mpc_busid = 0;
-+ memcpy(bus.mpc_bustype, "ISA ", 6);
-+ MP_bus_info(&bus);
++ /*
++ * For MPS mode, so far only needed by ES7000 platform
++ */
++ if (ioapic_renumber_irq)
++ irq = ioapic_renumber_irq(apic, irq);
+
-+ construct_default_ioirq_mptable(0);
++ break;
+ }
++ default:
++ {
++ printk(KERN_ERR "unknown bus type %d.\n",bus);
++ irq = 0;
++ break;
++ }
++ }
+
-+ } else
-+ BUG();
-+
-+ printk(KERN_INFO "Processors: %d\n", num_processors);
+ /*
-+ * Only use the first configuration found.
++ * 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, KERN_DEBUG
++ "disabling PIRQ%d\n", pin-16);
++ } else {
++ irq = pirq_entries[pin-16];
++ apic_printk(APIC_VERBOSE, KERN_DEBUG
++ "using PIRQ%d -> IRQ %d\n",
++ pin-16, irq);
++ }
++ }
++ }
++ return irq;
+}
+
-+static int __init smp_scan_config (unsigned long base, unsigned long length)
++static inline int IO_APIC_irq_trigger(int irq)
+{
-+ unsigned long *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)
-+ printk("Error: MPF size\n");
-+
-+ 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;
-+#ifndef CONFIG_XEN
-+ printk(KERN_INFO "found SMP MP-table at %08lx\n",
-+ virt_to_phys(mpf));
-+ reserve_bootmem(virt_to_phys(mpf), PAGE_SIZE);
-+ if (mpf->mpf_physptr) {
-+ /*
-+ * We cannot access to MPC table to compute
-+ * table size yet, as only few megabytes from
-+ * the bottom is mapped now.
-+ * PC-9800's MPC table places on the very last
-+ * of physical memory; so that simply reserving
-+ * PAGE_SIZE from mpg->mpf_physptr yields BUG()
-+ * in reserve_bootmem.
-+ */
-+ unsigned long size = PAGE_SIZE;
-+ unsigned long end = max_low_pfn * PAGE_SIZE;
-+ if (mpf->mpf_physptr + size > end)
-+ size = end - mpf->mpf_physptr;
-+ reserve_bootmem(mpf->mpf_physptr, size);
-+ }
-+#else
-+ printk(KERN_INFO "found SMP MP-table at %08lx\n",
-+ ((unsigned long)bp - (unsigned long)isa_bus_to_virt(base)) + base);
-+#endif
++ int apic, idx, pin;
+
-+ mpf_found = mpf;
-+ return 1;
++ 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);
+ }
-+ bp += 4;
-+ length -= 16;
+ }
++ /*
++ * nonexistent IRQs are edge default
++ */
+ return 0;
+}
+
-+void __init find_smp_config (void)
-+{
-+#ifndef CONFIG_XEN
-+ unsigned int address;
-+#endif
++/* irq_vectors is indexed by the sum of all RTEs in all I/O APICs. */
++static u8 irq_vector[NR_IRQ_VECTORS] __read_mostly; /* = { FIRST_DEVICE_VECTOR , 0 }; */
+
-+ /*
-+ * 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.
-+ *
-+ * MP1.4 SPEC states to only scan first 1K of 4K EBDA.
-+ */
++static int __assign_irq_vector(int irq)
++{
++ struct physdev_irq irq_op;
++ int vector;
+
-+#ifndef CONFIG_XEN
-+ address = get_bios_ebda();
-+ if (address)
-+ smp_scan_config(address, 0x400);
-+#endif
-+}
++ BUG_ON((unsigned)irq >= NR_IRQ_VECTORS);
+
-+int es7000_plat;
++ if (irq_vector[irq] > 0)
++ return irq_vector[irq];
++ irq_op.irq = irq;
++ if (HYPERVISOR_physdev_op(PHYSDEVOP_alloc_irq_vector, &irq_op))
++ return -ENOSPC;
+
-+/* --------------------------------------------------------------------------
-+ ACPI-based MP Configuration
-+ -------------------------------------------------------------------------- */
++ vector = irq_op.vector;
++ irq_vector[irq] = vector;
+
-+#ifdef CONFIG_ACPI
++ return vector;
++}
+
-+void __init mp_register_lapic_address (
-+ u64 address)
++static int assign_irq_vector(int irq)
+{
-+#ifndef CONFIG_XEN
-+ mp_lapic_addr = (unsigned long) address;
-+
-+ set_fixmap_nocache(FIX_APIC_BASE, mp_lapic_addr);
++ unsigned long flags;
++ int vector;
+
-+ if (boot_cpu_physical_apicid == -1U)
-+ boot_cpu_physical_apicid = GET_APIC_ID(apic_read(APIC_ID));
++ spin_lock_irqsave(&vector_lock, flags);
++ vector = __assign_irq_vector(irq);
++ spin_unlock_irqrestore(&vector_lock, flags);
+
-+ Dprintk("Boot CPU = %d\n", boot_cpu_physical_apicid);
-+#endif
++ return vector;
+}
++#ifndef CONFIG_XEN
++static struct irq_chip ioapic_chip;
+
++#define IOAPIC_AUTO -1
++#define IOAPIC_EDGE 0
++#define IOAPIC_LEVEL 1
+
-+void __devinit mp_register_lapic (
-+ u8 id,
-+ u8 enabled)
++static void ioapic_register_intr(int irq, int vector, unsigned long trigger)
+{
-+ struct mpc_config_processor processor;
-+ int boot_cpu = 0;
-+
-+ if (MAX_APICS - id <= 0) {
-+ printk(KERN_WARNING "Processor #%d invalid (max %d)\n",
-+ id, MAX_APICS);
-+ return;
++ if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
++ trigger == IOAPIC_LEVEL)
++ set_irq_chip_and_handler_name(irq, &ioapic_chip,
++ handle_fasteoi_irq, "fasteoi");
++ else {
++ irq_desc[irq].status |= IRQ_DELAYED_DISABLE;
++ set_irq_chip_and_handler_name(irq, &ioapic_chip,
++ handle_edge_irq, "edge");
+ }
-+
-+ 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;
++ set_intr_gate(vector, interrupt[irq]);
++}
++#else
++#define ioapic_register_intr(_irq,_vector,_trigger) ((void)0)
+#endif
+
-+ MP_processor_info(&processor);
-+}
++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;
+
-+#ifdef CONFIG_X86_IO_APIC
++ apic_printk(APIC_VERBOSE, KERN_DEBUG "init IO_APIC IRQs\n");
+
-+#define MP_ISA_BUS 0
-+#define MP_MAX_IOAPIC_PIN 127
++ for (apic = 0; apic < nr_ioapics; apic++) {
++ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
+
-+static struct mp_ioapic_routing {
-+ int apic_id;
-+ int gsi_base;
-+ int gsi_end;
-+ u32 pin_programmed[4];
-+} mp_ioapic_routing[MAX_IO_APICS];
++ /*
++ * 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);
+
-+static int mp_find_ioapic (
-+ int gsi)
-+{
-+ int i = 0;
++ 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;
++ }
+
-+ /* Find the IOAPIC that manages this GSI. */
-+ for (i = 0; i < nr_ioapics; i++) {
-+ if ((gsi >= mp_ioapic_routing[i].gsi_base)
-+ && (gsi <= mp_ioapic_routing[i].gsi_end))
-+ return i;
-+ }
++ entry.trigger = irq_trigger(idx);
++ entry.polarity = irq_polarity(idx);
+
-+ printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi);
++ if (irq_trigger(idx)) {
++ entry.trigger = 1;
++ entry.mask = 1;
++ }
+
-+ return -1;
-+}
-+
++ irq = pin_2_irq(idx, apic, pin);
++ /*
++ * skip adding the timer int on secondary nodes, which causes
++ * a small but painful rift in the time-space continuum
++ */
++ if (multi_timer_check(apic, irq))
++ continue;
++ else
++ add_pin_to_irq(irq, apic, pin);
+
-+void __init mp_register_ioapic (
-+ u8 id,
-+ u32 address,
-+ u32 gsi_base)
-+{
-+ int idx = 0;
-+ int tmpid;
++ if (/*!apic &&*/ !IO_APIC_IRQ(irq))
++ continue;
+
-+ 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 (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);
++ __ioapic_write_entry(apic, pin, entry);
++ set_native_irq_info(irq, TARGET_CPUS);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
+ }
-+ 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;
++ if (!first_notcon)
++ apic_printk(APIC_VERBOSE, " not connected.\n");
++}
+
++/*
++ * Set up the 8259A-master output pin:
++ */
+#ifndef CONFIG_XEN
-+ set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address);
-+#endif
-+ if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL)
-+ && !APIC_XAPIC(apic_version[boot_cpu_physical_apicid]))
-+ tmpid = io_apic_get_unique_id(idx, id);
-+ else
-+ tmpid = id;
-+ if (tmpid == -1) {
-+ nr_ioapics--;
-+ return;
-+ }
-+ mp_ioapics[idx].mpc_apicid = tmpid;
-+ mp_ioapics[idx].mpc_apicver = io_apic_get_version(idx);
-+
-+ /*
-+ * Build basic GSI lookup table to facilitate gsi->io_apic lookups
-+ * and to prevent reprogramming of IOAPIC pins (PCI GSIs).
-+ */
-+ mp_ioapic_routing[idx].apic_id = mp_ioapics[idx].mpc_apicid;
-+ mp_ioapic_routing[idx].gsi_base = gsi_base;
-+ mp_ioapic_routing[idx].gsi_end = gsi_base +
-+ io_apic_get_redir_entries(idx);
-+
-+ printk("IOAPIC[%d]: apic_id %d, version %d, address 0x%lx, "
-+ "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_base,
-+ mp_ioapic_routing[idx].gsi_end);
++static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, int vector)
++{
++ struct IO_APIC_route_entry entry;
+
-+ return;
-+}
++ memset(&entry,0,sizeof(entry));
+
++ disable_8259A_irq(0);
+
-+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;
++ /* mask LVT0 */
++ apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
+
-+ /*
-+ * Convert 'gsi' to 'ioapic.pin'.
++ /*
++ * We use logical delivery to get the timer IRQ
++ * to the first CPU.
+ */
-+ ioapic = mp_find_ioapic(gsi);
-+ if (ioapic < 0)
-+ return;
-+ pin = gsi - mp_ioapic_routing[ioapic].gsi_base;
++ 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;
+
+ /*
-+ * 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!
++ * The timer IRQ doesn't have to know that behind the
++ * scene we have a 8259A-master in AEOI mode ...
+ */
-+ if ((bus_irq == 0) && (trigger == 3))
-+ trigger = 1;
++ irq_desc[0].chip = &ioapic_chip;
++ set_irq_handler(0, handle_edge_irq);
+
-+ 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");
++ /*
++ * Add it to the IO-APIC irq-routing table:
++ */
++ ioapic_write_entry(apic, pin, entry);
+
-+ return;
++ enable_8259A_irq(0);
+}
+
-+void __init mp_config_acpi_legacy_irqs (void)
++static inline void UNEXPECTED_IO_APIC(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);
++}
+
-+ /*
-+ * Older generations of ES7000 have no legacy identity mappings
-+ */
-+ if (es7000_plat == 1)
-+ return;
++void __init 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;
++ union IO_APIC_reg_03 reg_03;
++ unsigned long flags;
+
-+ /*
-+ * Locate the IOAPIC that manages the ISA IRQs (0-15).
-+ */
-+ ioapic = mp_find_ioapic(0);
-+ if (ioapic < 0)
++ if (apic_verbosity == APIC_QUIET)
+ 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;
++ 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]);
+
-+ /*
-+ * Use the default configuration for the IRQs 0-15. Unless
-+ * overriden by (MADT) interrupt source override entries.
++ /*
++ * We are a bit conservative about what we expect. We have to
++ * know about every hardware change ASAP.
+ */
-+ for (i = 0; i < 16; i++) {
-+ int idx;
-+
-+ for (idx = 0; idx < mp_irq_entries; idx++) {
-+ struct mpc_config_intsrc *irq = mp_irqs + idx;
++ printk(KERN_INFO "testing the IO APIC.......................\n");
+
-+ /* Do we already have a mapping for this ISA IRQ? */
-+ if (irq->mpc_srcbus == MP_ISA_BUS && irq->mpc_srcbusirq == i)
-+ break;
++ for (apic = 0; apic < nr_ioapics; apic++) {
+
-+ /* 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(&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);
++ if (reg_01.bits.version >= 0x20)
++ reg_03.raw = io_apic_read(apic, 3);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
+
-+ if (idx != mp_irq_entries) {
-+ printk(KERN_DEBUG "ACPI: IRQ%d used by override.\n", i);
-+ continue; /* IRQ already used */
-+ }
++ 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);
++ printk(KERN_DEBUG "....... : Delivery Type: %X\n", reg_00.bits.delivery_type);
++ printk(KERN_DEBUG "....... : LTS : %X\n", reg_00.bits.LTS);
++ if (reg_00.bits.ID >= get_physical_broadcast())
++ UNEXPECTED_IO_APIC();
++ if (reg_00.bits.__reserved_1 || reg_00.bits.__reserved_2)
++ UNEXPECTED_IO_APIC();
+
-+ intsrc.mpc_irqtype = mp_INT;
-+ intsrc.mpc_srcbusirq = i; /* Identity mapped */
-+ intsrc.mpc_dstirq = i;
++ printk(KERN_DEBUG ".... register #01: %08X\n", reg_01.raw);
++ 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)
++ )
++ UNEXPECTED_IO_APIC();
+
-+ 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);
++ 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 != 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();
+
-+ mp_irqs[mp_irq_entries] = intsrc;
-+ if (++mp_irq_entries == MAX_IRQ_SOURCES)
-+ panic("Max # of irq sources exceeded!\n");
++ /*
++ * Some Intel chipsets with IO APIC VERSION of 0x1? don't have reg_02,
++ * but the value of reg_02 is read as the previous read register
++ * value, so ignore it if reg_02 == reg_01.
++ */
++ if (reg_01.bits.version >= 0x10 && reg_02.raw != reg_01.raw) {
++ 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();
+ }
-+}
+
-+#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 Interrups, which
-+ * represent all possible interrupts, and IRQs
-+ * assigned to actual devices.
++ * Some Intel chipsets with IO APIC VERSION of 0x2? don't have reg_02
++ * or reg_03, but the value of reg_0[23] is read as the previous read
++ * register value, so ignore it if reg_03 == reg_0[12].
+ */
-+ static int gsi_to_irq[MAX_GSI_NUM];
-+
-+ /* 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;
++ if (reg_01.bits.version >= 0x20 && reg_03.raw != reg_02.raw &&
++ reg_03.raw != reg_01.raw) {
++ printk(KERN_DEBUG ".... register #03: %08X\n", reg_03.raw);
++ printk(KERN_DEBUG "....... : Boot DT : %X\n", reg_03.bits.boot_DT);
++ if (reg_03.bits.__reserved_1)
++ UNEXPECTED_IO_APIC();
+ }
+
-+ ioapic_pin = gsi - mp_ioapic_routing[ioapic].gsi_base;
++ printk(KERN_DEBUG ".... IRQ redirection table:\n");
++
++ printk(KERN_DEBUG " NR Log Phy Mask Trig IRR Pol"
++ " Stat Dest Deli Vect: \n");
+
-+ if (ioapic_renumber_irq)
-+ gsi = ioapic_renumber_irq(ioapic, gsi);
++ for (i = 0; i <= reg_01.bits.entries; i++) {
++ struct IO_APIC_route_entry entry;
+
-+ /*
-+ * 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];
-+ }
++ entry = ioapic_read_entry(apic, i);
+
-+ mp_ioapic_routing[ioapic].pin_programmed[idx] |= (1<<bit);
++ printk(KERN_DEBUG " %02x %03X %02X ",
++ i,
++ entry.dest.logical.logical_dest,
++ entry.dest.physical.physical_dest
++ );
+
-+ 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;
++ 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
++ );
++ }
++ }
++ 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;
++ 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");
+ }
+
-+ 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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/pci-dma-xen.c linux-2.6.18-xen/arch/i386/kernel/pci-dma-xen.c
---- linux-2.6.18.3/arch/i386/kernel/pci-dma-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/pci-dma-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,379 @@
-+/*
-+ * Dynamic DMA mapping support.
-+ *
-+ * On i386 there is no hardware dynamic DMA address translation,
-+ * so consistent alloc/free are merely page allocation/freeing.
-+ * The rest of the dynamic DMA mapping interface is implemented
-+ * in asm/pci.h.
-+ */
-+
-+#include <linux/types.h>
-+#include <linux/mm.h>
-+#include <linux/string.h>
-+#include <linux/pci.h>
-+#include <linux/module.h>
-+#include <linux/version.h>
-+#include <asm/io.h>
-+#include <xen/balloon.h>
-+#include <asm/tlbflush.h>
-+#include <asm-i386/mach-xen/asm/swiotlb.h>
-+#include <asm/bug.h>
-+
-+#ifdef __x86_64__
-+int iommu_merge __read_mostly = 0;
-+EXPORT_SYMBOL(iommu_merge);
-+
-+dma_addr_t bad_dma_address __read_mostly;
-+EXPORT_SYMBOL(bad_dma_address);
-+
-+/* This tells the BIO block layer to assume merging. Default to off
-+ because we cannot guarantee merging later. */
-+int iommu_bio_merge __read_mostly = 0;
-+EXPORT_SYMBOL(iommu_bio_merge);
-+
-+int iommu_sac_force __read_mostly = 0;
-+EXPORT_SYMBOL(iommu_sac_force);
-+
-+int no_iommu __read_mostly;
-+#ifdef CONFIG_IOMMU_DEBUG
-+int panic_on_overflow __read_mostly = 1;
-+int force_iommu __read_mostly = 1;
-+#else
-+int panic_on_overflow __read_mostly = 0;
-+int force_iommu __read_mostly= 0;
-+#endif
-+
-+/* Set this to 1 if there is a HW IOMMU in the system */
-+int iommu_detected __read_mostly = 0;
-+
-+void __init pci_iommu_alloc(void)
-+{
-+ /*
-+ * The order of these functions is important for
-+ * fall-back/fail-over reasons
-+ */
-+#ifdef CONFIG_IOMMU
-+ iommu_hole_init();
-+#endif
-+
-+#ifdef CONFIG_CALGARY_IOMMU
-+#include <asm/calgary.h>
-+ detect_calgary();
-+#endif
-+
-+#ifdef CONFIG_SWIOTLB
-+ pci_swiotlb_init();
-+#endif
-+}
++ printk(KERN_INFO ".................................... done.\n");
+
-+__init int iommu_setup(char *p)
-+{
-+ return 1;
++ return;
+}
-+#endif
-+
-+struct dma_coherent_mem {
-+ void *virt_base;
-+ u32 device_base;
-+ int size;
-+ int flags;
-+ unsigned long *bitmap;
-+};
+
-+#define IOMMU_BUG_ON(test) \
-+do { \
-+ if (unlikely(test)) { \
-+ printk(KERN_ALERT "Fatal DMA error! " \
-+ "Please use 'swiotlb=force'\n"); \
-+ BUG(); \
-+ } \
-+} while (0)
++#if 0
+
-+int
-+dma_map_sg(struct device *hwdev, struct scatterlist *sg, int nents,
-+ enum dma_data_direction direction)
++static void print_APIC_bitfield (int base)
+{
-+ int i, rc;
++ unsigned int v;
++ int i, j;
+
-+ if (direction == DMA_NONE)
-+ BUG();
-+ WARN_ON(nents == 0 || sg[0].length == 0);
++ if (apic_verbosity == APIC_QUIET)
++ return;
+
-+ if (swiotlb) {
-+ rc = swiotlb_map_sg(hwdev, sg, nents, direction);
-+ } else {
-+ for (i = 0; i < nents; i++ ) {
-+ sg[i].dma_address =
-+ page_to_bus(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));
++ 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");
+ }
-+ rc = nents;
++ printk("\n");
+ }
-+
-+ flush_write_buffers();
-+ return rc;
+}
-+EXPORT_SYMBOL(dma_map_sg);
+
-+void
-+dma_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nents,
-+ enum dma_data_direction direction)
++void /*__init*/ print_local_APIC(void * dummy)
+{
-+ BUG_ON(direction == DMA_NONE);
-+ if (swiotlb)
-+ swiotlb_unmap_sg(hwdev, sg, nents, direction);
-+}
-+EXPORT_SYMBOL(dma_unmap_sg);
++ unsigned int v, ver, maxlvt;
+
-+/*
-+ * XXX This file is also used by xenLinux/ia64.
-+ * "defined(__i386__) || defined (__x86_64__)" means "!defined(__ia64__)".
-+ * This #if work around should be removed once this file is merbed back into
-+ * i386' pci-dma or is moved to drivers/xen/core.
-+ */
-+#if defined(__i386__) || defined(__x86_64__)
-+dma_addr_t
-+dma_map_page(struct device *dev, struct page *page, unsigned long offset,
-+ size_t size, enum dma_data_direction direction)
-+{
-+ dma_addr_t dma_addr;
++ 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();
+
-+ BUG_ON(direction == DMA_NONE);
++ v = apic_read(APIC_TASKPRI);
++ printk(KERN_DEBUG "... APIC TASKPRI: %08x (%02x)\n", v, v & APIC_TPRI_MASK);
+
-+ if (swiotlb) {
-+ dma_addr = swiotlb_map_page(
-+ dev, page, offset, size, direction);
-+ } else {
-+ dma_addr = page_to_bus(page) + offset;
-+ IOMMU_BUG_ON(address_needs_mapping(dev, dma_addr));
++ if (APIC_INTEGRATED(ver)) { /* !82489DX */
++ 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);
+ }
+
-+ return dma_addr;
-+}
-+EXPORT_SYMBOL(dma_map_page);
++ 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);
+
-+void
-+dma_unmap_page(struct device *dev, dma_addr_t dma_address, size_t size,
-+ enum dma_data_direction direction)
-+{
-+ BUG_ON(direction == DMA_NONE);
-+ if (swiotlb)
-+ swiotlb_unmap_page(dev, dma_address, size, direction);
-+}
-+EXPORT_SYMBOL(dma_unmap_page);
-+#endif /* defined(__i386__) || defined(__x86_64__) */
++ 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);
+
-+int
-+dma_mapping_error(dma_addr_t dma_addr)
-+{
-+ if (swiotlb)
-+ return swiotlb_dma_mapping_error(dma_addr);
-+ return 0;
-+}
-+EXPORT_SYMBOL(dma_mapping_error);
++ if (APIC_INTEGRATED(ver)) { /* !82489DX */
++ if (maxlvt > 3) /* Due to the Pentium erratum 3AP. */
++ apic_write(APIC_ESR, 0);
++ v = apic_read(APIC_ESR);
++ printk(KERN_DEBUG "... APIC ESR: %08x\n", v);
++ }
+
-+int
-+dma_supported(struct device *dev, u64 mask)
-+{
-+ if (swiotlb)
-+ return swiotlb_dma_supported(dev, mask);
-+ /*
-+ * By default we'll BUG when an infeasible DMA is requested, and
-+ * request swiotlb=force (see IOMMU_BUG_ON).
-+ */
-+ return 1;
-+}
-+EXPORT_SYMBOL(dma_supported);
++ 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);
+
-+void *dma_alloc_coherent(struct device *dev, size_t size,
-+ dma_addr_t *dma_handle, gfp_t gfp)
-+{
-+ void *ret;
-+ struct dma_coherent_mem *mem = dev ? dev->dma_mem : NULL;
-+ unsigned int order = get_order(size);
-+ unsigned long vstart;
-+ /* ignore region specifiers */
-+ gfp &= ~(__GFP_DMA | __GFP_HIGHMEM);
++ v = apic_read(APIC_LVTT);
++ printk(KERN_DEBUG "... APIC LVTT: %08x\n", v);
+
-+ if (mem) {
-+ int page = bitmap_find_free_region(mem->bitmap, mem->size,
-+ order);
-+ if (page >= 0) {
-+ *dma_handle = mem->device_base + (page << PAGE_SHIFT);
-+ ret = mem->virt_base + (page << PAGE_SHIFT);
-+ memset(ret, 0, size);
-+ return ret;
-+ }
-+ if (mem->flags & DMA_MEMORY_EXCLUSIVE)
-+ return NULL;
++ 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 (dev == NULL || (dev->coherent_dma_mask < 0xffffffff))
-+ gfp |= GFP_DMA;
-+
-+ vstart = __get_free_pages(gfp, order);
-+ ret = (void *)vstart;
-+
-+ if (ret != NULL) {
-+ /* NB. Hardcode 31 address bits for now: aacraid limitation. */
-+ if (xen_create_contiguous_region(vstart, order, 31) != 0) {
-+ free_pages(vstart, order);
-+ return NULL;
-+ }
-+ memset(ret, 0, size);
-+ *dma_handle = virt_to_bus(ret);
++ if (maxlvt > 2) { /* ERR is LVT#3. */
++ v = apic_read(APIC_LVTERR);
++ printk(KERN_DEBUG "... APIC LVTERR: %08x\n", v);
+ }
-+ return ret;
++
++ 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");
+}
-+EXPORT_SYMBOL(dma_alloc_coherent);
+
-+void dma_free_coherent(struct device *dev, size_t size,
-+ void *vaddr, dma_addr_t dma_handle)
++void print_all_local_APICs (void)
+{
-+ struct dma_coherent_mem *mem = dev ? dev->dma_mem : NULL;
-+ int order = get_order(size);
-+
-+ if (mem && vaddr >= mem->virt_base && vaddr < (mem->virt_base + (mem->size << PAGE_SHIFT))) {
-+ int page = (vaddr - mem->virt_base) >> PAGE_SHIFT;
-+
-+ bitmap_release_region(mem->bitmap, page, order);
-+ } else {
-+ xen_destroy_contiguous_region((unsigned long)vaddr, order);
-+ free_pages((unsigned long)vaddr, order);
-+ }
++ on_each_cpu(print_local_APIC, NULL, 1, 1);
+}
-+EXPORT_SYMBOL(dma_free_coherent);
+
-+#ifdef ARCH_HAS_DMA_DECLARE_COHERENT_MEMORY
-+int dma_declare_coherent_memory(struct device *dev, dma_addr_t bus_addr,
-+ dma_addr_t device_addr, size_t size, int flags)
++void /*__init*/ print_PIC(void)
+{
-+ void __iomem *mem_base;
-+ int pages = size >> PAGE_SHIFT;
-+ int bitmap_size = (pages + 31)/32;
++ unsigned int v;
++ unsigned long flags;
+
-+ if ((flags & (DMA_MEMORY_MAP | DMA_MEMORY_IO)) == 0)
-+ goto out;
-+ if (!size)
-+ goto out;
-+ if (dev->dma_mem)
-+ goto out;
++ if (apic_verbosity == APIC_QUIET)
++ return;
+
-+ /* FIXME: this routine just ignores DMA_MEMORY_INCLUDES_CHILDREN */
++ printk(KERN_DEBUG "\nprinting PIC contents\n");
+
-+ mem_base = ioremap(bus_addr, size);
-+ if (!mem_base)
-+ goto out;
++ spin_lock_irqsave(&i8259A_lock, flags);
+
-+ dev->dma_mem = kmalloc(sizeof(struct dma_coherent_mem), GFP_KERNEL);
-+ if (!dev->dma_mem)
-+ goto out;
-+ memset(dev->dma_mem, 0, sizeof(struct dma_coherent_mem));
-+ dev->dma_mem->bitmap = kmalloc(bitmap_size, GFP_KERNEL);
-+ if (!dev->dma_mem->bitmap)
-+ goto free1_out;
-+ memset(dev->dma_mem->bitmap, 0, bitmap_size);
++ v = inb(0xa1) << 8 | inb(0x21);
++ printk(KERN_DEBUG "... PIC IMR: %04x\n", v);
+
-+ dev->dma_mem->virt_base = mem_base;
-+ dev->dma_mem->device_base = device_addr;
-+ dev->dma_mem->size = pages;
-+ dev->dma_mem->flags = flags;
++ v = inb(0xa0) << 8 | inb(0x20);
++ printk(KERN_DEBUG "... PIC IRR: %04x\n", v);
+
-+ if (flags & DMA_MEMORY_MAP)
-+ return DMA_MEMORY_MAP;
++ outb(0x0b,0xa0);
++ outb(0x0b,0x20);
++ v = inb(0xa0) << 8 | inb(0x20);
++ outb(0x0a,0xa0);
++ outb(0x0a,0x20);
+
-+ return DMA_MEMORY_IO;
++ spin_unlock_irqrestore(&i8259A_lock, flags);
+
-+ free1_out:
-+ kfree(dev->dma_mem->bitmap);
-+ out:
-+ return 0;
-+}
-+EXPORT_SYMBOL(dma_declare_coherent_memory);
++ printk(KERN_DEBUG "... PIC ISR: %04x\n", v);
+
-+void dma_release_declared_memory(struct device *dev)
-+{
-+ struct dma_coherent_mem *mem = dev->dma_mem;
-+
-+ if(!mem)
-+ return;
-+ dev->dma_mem = NULL;
-+ iounmap(mem->virt_base);
-+ kfree(mem->bitmap);
-+ kfree(mem);
++ v = inb(0x4d1) << 8 | inb(0x4d0);
++ printk(KERN_DEBUG "... PIC ELCR: %04x\n", v);
+}
-+EXPORT_SYMBOL(dma_release_declared_memory);
-+
-+void *dma_mark_declared_memory_occupied(struct device *dev,
-+ dma_addr_t device_addr, size_t size)
-+{
-+ struct dma_coherent_mem *mem = dev->dma_mem;
-+ int pages = (size + (device_addr & ~PAGE_MASK) + PAGE_SIZE - 1) >> PAGE_SHIFT;
-+ int pos, err;
+
-+ if (!mem)
-+ return ERR_PTR(-EINVAL);
++#endif /* 0 */
+
-+ pos = (device_addr - mem->device_base) >> PAGE_SHIFT;
-+ err = bitmap_allocate_region(mem->bitmap, pos, get_order(pages));
-+ if (err != 0)
-+ return ERR_PTR(err);
-+ return mem->virt_base + (pos << PAGE_SHIFT);
-+}
-+EXPORT_SYMBOL(dma_mark_declared_memory_occupied);
-+#endif /* ARCH_HAS_DMA_DECLARE_COHERENT_MEMORY */
++#else
++void __init print_IO_APIC(void) { }
++#endif /* !CONFIG_XEN */
+
-+dma_addr_t
-+dma_map_single(struct device *dev, void *ptr, size_t size,
-+ enum dma_data_direction direction)
++static void __init enable_IO_APIC(void)
+{
-+ dma_addr_t dma;
++ union IO_APIC_reg_01 reg_01;
++ int i8259_apic, i8259_pin;
++ int i, apic;
++ unsigned long flags;
+
-+ if (direction == DMA_NONE)
-+ BUG();
-+ WARN_ON(size == 0);
++ 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 (swiotlb) {
-+ dma = swiotlb_map_single(dev, ptr, size, direction);
-+ } else {
-+ dma = virt_to_bus(ptr);
-+ IOMMU_BUG_ON(range_straddles_page_boundary(ptr, size));
-+ IOMMU_BUG_ON(address_needs_mapping(dev, dma));
++ /*
++ * 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;
+ }
++ 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;
++ entry = ioapic_read_entry(apic, pin);
+
-+ flush_write_buffers();
-+ return dma;
-+}
-+EXPORT_SYMBOL(dma_map_single);
+
-+void
-+dma_unmap_single(struct device *dev, dma_addr_t dma_addr, size_t size,
-+ enum dma_data_direction direction)
-+{
-+ if (direction == DMA_NONE)
-+ BUG();
-+ if (swiotlb)
-+ swiotlb_unmap_single(dev, dma_addr, size, direction);
-+}
-+EXPORT_SYMBOL(dma_unmap_single);
++ /* 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 */
++ /* If we could not find the appropriate pin by looking at the ioapic
++ * the i8259 probably is not connected the ioapic but give the
++ * mptable a chance anyway.
++ */
++ 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");
++ }
+
-+void
-+dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle, size_t size,
-+ enum dma_data_direction direction)
-+{
-+ if (swiotlb)
-+ swiotlb_sync_single_for_cpu(dev, dma_handle, size, direction);
++ /*
++ * Do not trust the IO-APIC being empty at bootup
++ */
++ clear_IO_APIC();
+}
-+EXPORT_SYMBOL(dma_sync_single_for_cpu);
+
-+void
-+dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle, size_t size,
-+ enum dma_data_direction direction)
++/*
++ * Not an __init, needed by the reboot code
++ */
++void disable_IO_APIC(void)
+{
-+ if (swiotlb)
-+ swiotlb_sync_single_for_device(dev, dma_handle, size, direction);
++ /*
++ * 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;
++
++ 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:
++ */
++ ioapic_write_entry(ioapic_i8259.apic, ioapic_i8259.pin, entry);
++ }
++ disconnect_bsp_APIC(ioapic_i8259.pin != -1);
++#endif
+}
-+EXPORT_SYMBOL(dma_sync_single_for_device);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/process-xen.c linux-2.6.18-xen/arch/i386/kernel/process-xen.c
---- linux-2.6.18.3/arch/i386/kernel/process-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/process-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,812 @@
++
+/*
-+ * linux/arch/i386/kernel/process.c
-+ *
-+ * Copyright (C) 1995 Linus Torvalds
++ * function to set the IO-APIC physical IDs based on the
++ * values stored in the MPC table.
+ *
-+ * Pentium III FXSR, SSE support
-+ * Gareth Hughes <gareth at valinux.com>, May 2000
++ * by Matt Domsch <Matt_Domsch at dell.com> Tue Dec 21 12:25:05 CST 1999
+ */
+
-+/*
-+ * This file handles the architecture-dependent parts of process handling..
-+ */
++#if !defined(CONFIG_XEN) && !defined(CONFIG_X86_NUMAQ)
++static void __init setup_ioapic_ids_from_mpc(void)
++{
++ union IO_APIC_reg_00 reg_00;
++ physid_mask_t phys_id_present_map;
++ int apic;
++ int i;
++ unsigned char old_id;
++ unsigned long flags;
+
-+#include <stdarg.h>
++ /*
++ * Don't check I/O APIC IDs for xAPIC systems. They have
++ * no meaning without the serial APIC bus.
++ */
++ if (!(boot_cpu_data.x86_vendor == X86_VENDOR_INTEL)
++ || APIC_XAPIC(apic_version[boot_cpu_physical_apicid]))
++ return;
++ /*
++ * This is broken; anything with a real cpu count has to
++ * circumvent this idiocy regardless.
++ */
++ phys_id_present_map = ioapic_phys_id_map(phys_cpu_present_map);
+
-+#include <linux/cpu.h>
-+#include <linux/errno.h>
-+#include <linux/sched.h>
-+#include <linux/fs.h>
-+#include <linux/kernel.h>
-+#include <linux/mm.h>
-+#include <linux/elfcore.h>
-+#include <linux/smp.h>
-+#include <linux/smp_lock.h>
-+#include <linux/stddef.h>
-+#include <linux/slab.h>
-+#include <linux/vmalloc.h>
-+#include <linux/user.h>
-+#include <linux/a.out.h>
-+#include <linux/interrupt.h>
-+#include <linux/utsname.h>
-+#include <linux/delay.h>
-+#include <linux/reboot.h>
-+#include <linux/init.h>
-+#include <linux/mc146818rtc.h>
-+#include <linux/module.h>
-+#include <linux/kallsyms.h>
-+#include <linux/ptrace.h>
-+#include <linux/random.h>
++ /*
++ * Set the IOAPIC ID to the value stored in the MPC table.
++ */
++ for (apic = 0; apic < nr_ioapics; apic++) {
+
-+#include <asm/uaccess.h>
-+#include <asm/pgtable.h>
-+#include <asm/system.h>
-+#include <asm/io.h>
-+#include <asm/ldt.h>
-+#include <asm/processor.h>
-+#include <asm/i387.h>
-+#include <asm/desc.h>
-+#include <asm/vm86.h>
-+#ifdef CONFIG_MATH_EMULATION
-+#include <asm/math_emu.h>
-+#endif
++ /* 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;
++
++ if (mp_ioapics[apic].mpc_apicid >= get_physical_broadcast()) {
++ printk(KERN_ERR "BIOS bug, IO-APIC#%d ID is %d in the MPC table!...\n",
++ apic, mp_ioapics[apic].mpc_apicid);
++ printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n",
++ reg_00.bits.ID);
++ mp_ioapics[apic].mpc_apicid = reg_00.bits.ID;
++ }
++
++ /*
++ * Sanity check, is the ID really free? Every APIC in a
++ * system must have a unique ID or we get lots of nice
++ * 'stuck on smp_invalidate_needed IPI wait' messages.
++ */
++ if (check_apicid_used(phys_id_present_map,
++ mp_ioapics[apic].mpc_apicid)) {
++ printk(KERN_ERR "BIOS bug, IO-APIC#%d ID %d is already used!...\n",
++ apic, mp_ioapics[apic].mpc_apicid);
++ for (i = 0; i < get_physical_broadcast(); i++)
++ if (!physid_isset(i, phys_id_present_map))
++ break;
++ if (i >= get_physical_broadcast())
++ panic("Max APIC ID exceeded!\n");
++ printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n",
++ i);
++ physid_set(i, phys_id_present_map);
++ mp_ioapics[apic].mpc_apicid = i;
++ } else {
++ physid_mask_t tmp;
++ tmp = apicid_to_cpu_present(mp_ioapics[apic].mpc_apicid);
++ apic_printk(APIC_VERBOSE, "Setting %d in the "
++ "phys_id_present_map\n",
++ mp_ioapics[apic].mpc_apicid);
++ physids_or(phys_id_present_map, phys_id_present_map, tmp);
++ }
+
-+#include <xen/interface/physdev.h>
-+#include <xen/interface/vcpu.h>
-+#include <xen/cpu_hotplug.h>
+
-+#include <linux/err.h>
++ /*
++ * 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;
+
-+#include <asm/tlbflush.h>
-+#include <asm/cpu.h>
++ /*
++ * 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);
+
-+asmlinkage void ret_from_fork(void) __asm__("ret_from_fork");
++ 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);
+
-+static int hlt_counter;
++ /*
++ * 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
+
-+unsigned long boot_option_idle_override = 0;
-+EXPORT_SYMBOL(boot_option_idle_override);
++static int no_timer_check __initdata;
+
-+/*
-+ * Return saved PC of a blocked thread.
-+ */
-+unsigned long thread_saved_pc(struct task_struct *tsk)
++static int __init notimercheck(char *s)
+{
-+ return ((unsigned long *)tsk->thread.esp)[3];
++ no_timer_check = 1;
++ return 1;
+}
++__setup("no_timer_check", notimercheck);
+
++#ifndef CONFIG_XEN
+/*
-+ * Powermanagement idle function, if any..
++ * 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
+ */
-+void (*pm_idle)(void);
-+EXPORT_SYMBOL(pm_idle);
-+static DEFINE_PER_CPU(unsigned int, cpu_idle_state);
-+
-+void disable_hlt(void)
++int __init timer_irq_works(void)
+{
-+ hlt_counter++;
-+}
++ unsigned long t1 = jiffies;
+
-+EXPORT_SYMBOL(disable_hlt);
++ if (no_timer_check)
++ return 1;
+
-+void enable_hlt(void)
-+{
-+ hlt_counter--;
++ 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.
++ */
++ if (jiffies - t1 > 4)
++ return 1;
++
++ return 0;
+}
+
-+EXPORT_SYMBOL(enable_hlt);
++/*
++ * 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.
++ */
+
-+/* XXX XEN doesn't use default_idle(), poll_idle(). Use xen_idle() instead. */
-+void xen_idle(void)
++/*
++ * Startup quirk:
++ *
++ * 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...
++ *
++ * (We do this for level-triggered IRQs too - it cannot hurt.)
++ */
++static unsigned int startup_ioapic_irq(unsigned int irq)
+{
-+ local_irq_disable();
++ int was_pending = 0;
++ unsigned long flags;
+
-+ 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;
++ 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;
+}
-+#ifdef CONFIG_APM_MODULE
-+EXPORT_SYMBOL(default_idle);
-+#endif
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+extern cpumask_t cpu_initialized;
-+static inline void play_dead(void)
++static void ack_ioapic_irq(unsigned int irq)
+{
-+ 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();
++ move_native_irq(irq);
++ ack_APIC_irq();
+}
-+#else
-+static inline void play_dead(void)
++
++static void ack_ioapic_quirk_irq(unsigned int irq)
+{
-+ BUG();
-+}
-+#endif /* CONFIG_HOTPLUG_CPU */
++ unsigned long v;
++ int i;
+
++ move_native_irq(irq);
+/*
-+ * 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)
++ * It appears there is an erratum which affects at least version 0x11
++ * of I/O APIC (that's the 82093AA and cores integrated into various
++ * chipsets). Under certain conditions a level-triggered interrupt is
++ * erroneously delivered as edge-triggered one but the respective IRR
++ * bit gets set nevertheless. As a result the I/O unit expects an EOI
++ * message but it will never arrive and further interrupts are blocked
++ * from the source. The exact reason is so far unknown, but the
++ * phenomenon was observed when two consecutive interrupt requests
++ * from a given source get delivered to the same CPU and the source is
++ * temporarily disabled in between.
++ *
++ * A workaround is to simulate an EOI message manually. We achieve it
++ * by setting the trigger mode to edge and then to level when the edge
++ * trigger mode gets detected in the TMR of a local APIC for a
++ * level-triggered interrupt. We mask the source for the time of the
++ * operation to prevent an edge-triggered interrupt escaping meanwhile.
++ * The idea is from Manfred Spraul. --macro
+ */
-+void cpu_idle(void)
-+{
-+ int cpu = smp_processor_id();
++ i = irq_vector[irq];
+
-+ current_thread_info()->status |= TS_POLLING;
++ v = apic_read(APIC_TMR + ((i & ~0x1f) >> 1));
+
++ ack_APIC_irq();
+
-+ /* endless idle loop with no priority at all */
-+ while (1) {
-+ while (!need_resched()) {
++ if (!(v & (1 << (i & 0x1f)))) {
++ atomic_inc(&irq_mis_count);
++ spin_lock(&ioapic_lock);
++ __mask_and_edge_IO_APIC_irq(irq);
++ __unmask_and_level_IO_APIC_irq(irq);
++ spin_unlock(&ioapic_lock);
++ }
++}
+
-+ if (__get_cpu_var(cpu_idle_state))
-+ __get_cpu_var(cpu_idle_state) = 0;
++static int ioapic_retrigger_irq(unsigned int irq)
++{
++ send_IPI_self(irq_vector[irq]);
+
-+ rmb();
++ return 1;
++}
+
-+ if (cpu_is_offline(cpu))
-+ play_dead();
++static struct irq_chip ioapic_chip __read_mostly = {
++ .name = "IO-APIC",
++ .startup = startup_ioapic_irq,
++ .mask = mask_IO_APIC_irq,
++ .unmask = unmask_IO_APIC_irq,
++ .ack = ack_ioapic_irq,
++ .eoi = ack_ioapic_quirk_irq,
++#ifdef CONFIG_SMP
++ .set_affinity = set_ioapic_affinity_irq,
++#endif
++ .retrigger = ioapic_retrigger_irq,
++};
+
-+ __get_cpu_var(irq_stat).idle_timestamp = jiffies;
-+ xen_idle();
++#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 (IO_APIC_IRQ(tmp) && !irq_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_chip;
++#endif
+ }
-+ preempt_enable_no_resched();
-+ schedule();
-+ preempt_disable();
+ }
+}
+
-+void cpu_idle_wait(void)
++#ifndef CONFIG_XEN
++/*
++ * The local APIC irq-chip implementation:
++ */
++
++static void ack_apic(unsigned int irq)
+{
-+ unsigned int cpu, this_cpu = get_cpu();
-+ cpumask_t map;
++ ack_APIC_irq();
++}
+
-+ set_cpus_allowed(current, cpumask_of_cpu(this_cpu));
-+ put_cpu();
++static void mask_lapic_irq (unsigned int irq)
++{
++ unsigned long v;
+
-+ cpus_clear(map);
-+ for_each_online_cpu(cpu) {
-+ per_cpu(cpu_idle_state, cpu) = 1;
-+ cpu_set(cpu, map);
-+ }
++ v = apic_read(APIC_LVT0);
++ apic_write_around(APIC_LVT0, v | APIC_LVT_MASKED);
++}
+
-+ __get_cpu_var(cpu_idle_state) = 0;
++static void unmask_lapic_irq (unsigned int irq)
++{
++ unsigned long v;
+
-+ 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));
++ v = apic_read(APIC_LVT0);
++ apic_write_around(APIC_LVT0, v & ~APIC_LVT_MASKED);
+}
-+EXPORT_SYMBOL_GPL(cpu_idle_wait);
+
-+/* XXX XEN doesn't use mwait_idle(), select_idle_routine(), idle_setup(). */
-+/* Always use xen_idle() instead. */
-+void __devinit select_idle_routine(const struct cpuinfo_x86 *c) {}
++static struct irq_chip lapic_chip __read_mostly = {
++ .name = "local-APIC-edge",
++ .mask = mask_lapic_irq,
++ .unmask = unmask_lapic_irq,
++ .eoi = ack_apic,
++};
+
-+void show_regs(struct pt_regs * regs)
++static void setup_nmi (void)
+{
-+ unsigned long cr0 = 0L, cr2 = 0L, cr3 = 0L, cr4 = 0L;
-+
-+ printk("\n");
-+ printk("Pid: %d, comm: %20s\n", current->pid, current->comm);
-+ printk("EIP: %04x:[<%08lx>] CPU: %d\n",0xffff & regs->xcs,regs->eip, smp_processor_id());
-+ print_symbol("EIP is at %s\n", regs->eip);
++ /*
++ * 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.
++ */
++ apic_printk(APIC_VERBOSE, KERN_INFO "activating NMI Watchdog ...");
+
-+ if (user_mode_vm(regs))
-+ printk(" ESP: %04x:%08lx",0xffff & regs->xss,regs->esp);
-+ printk(" EFLAGS: %08lx %s (%s %.*s)\n",
-+ regs->eflags, print_tainted(), system_utsname.release,
-+ (int)strcspn(system_utsname.version, " "),
-+ system_utsname.version);
-+ printk("EAX: %08lx EBX: %08lx ECX: %08lx EDX: %08lx\n",
-+ regs->eax,regs->ebx,regs->ecx,regs->edx);
-+ printk("ESI: %08lx EDI: %08lx EBP: %08lx",
-+ regs->esi, regs->edi, regs->ebp);
-+ printk(" DS: %04x ES: %04x\n",
-+ 0xffff & regs->xds,0xffff & regs->xes);
++ on_each_cpu(enable_NMI_through_LVT0, NULL, 1, 1);
+
-+ cr0 = read_cr0();
-+ cr2 = read_cr2();
-+ cr3 = read_cr3();
-+ cr4 = read_cr4_safe();
-+ printk("CR0: %08lx CR2: %08lx CR3: %08lx CR4: %08lx\n", cr0, cr2, cr3, cr4);
-+ show_trace(NULL, regs, ®s->esp);
++ apic_printk(APIC_VERBOSE, " done.\n");
+}
+
+/*
-+ * This gets run with %ebx containing the
-+ * function to call, and %edx containing
-+ * the "args".
-+ */
-+extern void kernel_thread_helper(void);
-+__asm__(".section .text\n"
-+ ".align 4\n"
-+ "kernel_thread_helper:\n\t"
-+ "movl %edx,%eax\n\t"
-+ "pushl %edx\n\t"
-+ "call *%ebx\n\t"
-+ "pushl %eax\n\t"
-+ "call do_exit\n"
-+ ".previous");
-+
-+/*
-+ * Create a kernel thread
++ * 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
+ */
-+int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
++static inline void unlock_ExtINT_logic(void)
+{
-+ struct pt_regs regs;
++ int apic, pin, i;
++ struct IO_APIC_route_entry entry0, entry1;
++ unsigned char save_control, save_freq_select;
+
-+ memset(®s, 0, sizeof(regs));
++ pin = find_isa_irq_pin(8, mp_INT);
++ if (pin == -1) {
++ WARN_ON_ONCE(1);
++ return;
++ }
++ apic = find_isa_irq_apic(8, mp_INT);
++ if (apic == -1) {
++ WARN_ON_ONCE(1);
++ return;
++ }
+
-+ regs.ebx = (unsigned long) fn;
-+ regs.edx = (unsigned long) arg;
++ entry0 = ioapic_read_entry(apic, pin);
++ clear_IO_APIC_pin(apic, pin);
+
-+ regs.xds = __USER_DS;
-+ regs.xes = __USER_DS;
-+ regs.orig_eax = -1;
-+ regs.eip = (unsigned long) kernel_thread_helper;
-+ regs.xcs = GET_KERNEL_CS();
-+ regs.eflags = X86_EFLAGS_IF | X86_EFLAGS_SF | X86_EFLAGS_PF | 0x2;
++ memset(&entry1, 0, sizeof(entry1));
+
-+ /* Ok, create the new process.. */
-+ return do_fork(flags | CLONE_VM | CLONE_UNTRACED, 0, ®s, 0, NULL, NULL);
-+}
-+EXPORT_SYMBOL(kernel_thread);
++ 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;
+
-+/*
-+ * Free current thread data structures etc..
-+ */
-+void exit_thread(void)
-+{
-+ /* The process may have allocated an io port bitmap... nuke it. */
-+ if (unlikely(test_thread_flag(TIF_IO_BITMAP))) {
-+ struct task_struct *tsk = current;
-+ struct thread_struct *t = &tsk->thread;
++ ioapic_write_entry(apic, pin, entry1);
+
-+ struct physdev_set_iobitmap set_iobitmap = { 0 };
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap, &set_iobitmap);
-+ kfree(t->io_bitmap_ptr);
-+ t->io_bitmap_ptr = NULL;
-+ clear_thread_flag(TIF_IO_BITMAP);
++ 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;
+ }
-+}
+
-+void flush_thread(void)
-+{
-+ struct task_struct *tsk = current;
++ CMOS_WRITE(save_control, RTC_CONTROL);
++ CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT);
++ clear_IO_APIC_pin(apic, pin);
+
-+ memset(tsk->thread.debugreg, 0, sizeof(unsigned long)*8);
-+ memset(tsk->thread.tls_array, 0, sizeof(tsk->thread.tls_array));
-+ clear_tsk_thread_flag(tsk, TIF_DEBUG);
-+ /*
-+ * Forget coprocessor state..
-+ */
-+ clear_fpu(tsk);
-+ clear_used_math();
++ ioapic_write_entry(apic, pin, entry0);
+}
++#endif /* !CONFIG_XEN */
+
-+void release_thread(struct task_struct *dead_task)
-+{
-+ BUG_ON(dead_task->mm);
-+ release_vm86_irqs(dead_task);
-+}
++int timer_uses_ioapic_pin_0;
+
++#ifndef CONFIG_XEN
+/*
-+ * This gets called before we allocate a new thread and copy
-+ * the current task into it.
++ * 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.
+ */
-+void prepare_to_copy(struct task_struct *tsk)
++static inline void __init check_timer(void)
+{
-+ unlazy_fpu(tsk);
-+}
++ int apic1, pin1, apic2, pin2;
++ int vector;
+
-+int copy_thread(int nr, unsigned long clone_flags, unsigned long esp,
-+ unsigned long unused,
-+ struct task_struct * p, struct pt_regs * regs)
-+{
-+ struct pt_regs * childregs;
-+ struct task_struct *tsk;
-+ int err;
++ /*
++ * get/set the timer IRQ vector:
++ */
++ disable_8259A_irq(0);
++ vector = assign_irq_vector(0);
++ set_intr_gate(vector, interrupt[0]);
+
-+ childregs = task_pt_regs(p);
-+ *childregs = *regs;
-+ childregs->eax = 0;
-+ childregs->esp = esp;
++ /*
++ * 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_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
++ init_8259A(1);
++ timer_ack = 1;
++ if (timer_over_8254 > 0)
++ enable_8259A_irq(0);
+
-+ p->thread.esp = (unsigned long) childregs;
-+ p->thread.esp0 = (unsigned long) (childregs+1);
++ pin1 = find_isa_irq_pin(0, mp_INT);
++ apic1 = find_isa_irq_apic(0, mp_INT);
++ pin2 = ioapic_i8259.pin;
++ apic2 = ioapic_i8259.apic;
+
-+ p->thread.eip = (unsigned long) ret_from_fork;
++ if (pin1 == 0)
++ timer_uses_ioapic_pin_0 = 1;
+
-+ savesegment(fs,p->thread.fs);
-+ savesegment(gs,p->thread.gs);
++ printk(KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n",
++ vector, apic1, pin1, apic2, pin2);
+
-+ tsk = current;
-+ if (unlikely(test_tsk_thread_flag(tsk, TIF_IO_BITMAP))) {
-+ 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;
++ if (pin1 != -1) {
++ /*
++ * Ok, does IRQ0 through the IOAPIC work?
++ */
++ unmask_IO_APIC_irq(0);
++ if (timer_irq_works()) {
++ 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;
+ }
-+ memcpy(p->thread.io_bitmap_ptr, tsk->thread.io_bitmap_ptr,
-+ IO_BITMAP_BYTES);
-+ set_tsk_thread_flag(p, TIF_IO_BITMAP);
++ clear_IO_APIC_pin(apic1, pin1);
++ printk(KERN_ERR "..MP-BIOS bug: 8254 timer not connected to "
++ "IO-APIC\n");
++ }
++
++ printk(KERN_INFO "...trying to set up timer (IRQ0) through the 8259A ... ");
++ if (pin2 != -1) {
++ printk("\n..... (found pin %d) ...", pin2);
++ /*
++ * legacy devices should be connected to IO APIC #0
++ */
++ setup_ExtINT_IRQ0_pin(apic2, pin2, vector);
++ if (timer_irq_works()) {
++ printk("works.\n");
++ if (pin1 != -1)
++ replace_pin_at_irq(0, apic1, pin1, apic2, pin2);
++ else
++ add_pin_to_irq(0, apic2, pin2);
++ if (nmi_watchdog == NMI_IO_APIC) {
++ setup_nmi();
++ }
++ return;
++ }
++ /*
++ * Cleanup, just in case ...
++ */
++ clear_IO_APIC_pin(apic2, pin2);
++ }
++ printk(" 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;
+ }
+
-+ /*
-+ * Set a new TLS for the child thread?
-+ */
-+ if (clone_flags & CLONE_SETTLS) {
-+ struct desc_struct *desc;
-+ struct user_desc info;
-+ int idx;
-+
-+ err = -EFAULT;
-+ if (copy_from_user(&info, (void __user *)childregs->esi, sizeof(info)))
-+ goto out;
-+ err = -EINVAL;
-+ if (LDT_empty(&info))
-+ goto out;
++ printk(KERN_INFO "...trying to set up timer as Virtual Wire IRQ...");
+
-+ idx = info.entry_number;
-+ if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX)
-+ goto out;
++ disable_8259A_irq(0);
++ set_irq_chip_and_handler_name(0, &lapic_chip, handle_fasteoi_irq,
++ "fasteio");
++ apic_write_around(APIC_LVT0, APIC_DM_FIXED | vector); /* Fixed mode */
++ enable_8259A_irq(0);
+
-+ desc = p->thread.tls_array + idx - GDT_ENTRY_TLS_MIN;
-+ desc->a = LDT_entry_a(&info);
-+ desc->b = LDT_entry_b(&info);
++ if (timer_irq_works()) {
++ printk(" works.\n");
++ return;
+ }
++ apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | vector);
++ printk(" failed.\n");
+
-+ p->thread.iopl = current->thread.iopl;
++ printk(KERN_INFO "...trying to set up timer as ExtINT IRQ...");
+
-+ err = 0;
-+ out:
-+ if (err && p->thread.io_bitmap_ptr) {
-+ kfree(p->thread.io_bitmap_ptr);
-+ p->thread.io_bitmap_max = 0;
++ timer_ack = 0;
++ init_8259A(0);
++ make_8259A_irq(0);
++ apic_write_around(APIC_LVT0, APIC_DM_EXTINT);
++
++ unlock_ExtINT_logic();
++
++ if (timer_irq_works()) {
++ printk(" works.\n");
++ return;
+ }
-+ return err;
++ printk(" failed :(.\n");
++ panic("IO-APIC + timer doesn't work! Boot with apic=debug and send a "
++ "report. Then try booting with the 'noapic' option");
+}
++#else
++#define check_timer() ((void)0)
++#endif /* CONFIG_XEN */
+
+/*
-+ * fill in the user structure for a core dump..
++ *
++ * 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.
+ */
-+void dump_thread(struct pt_regs * regs, struct user * dump)
-+{
-+ int i;
++#define PIC_IRQS (1 << PIC_CASCADE_IR)
+
-+/* changed the size calculations - should hopefully work better. lbt */
-+ dump->magic = CMAGIC;
-+ dump->start_code = 0;
-+ dump->start_stack = regs->esp & ~(PAGE_SIZE - 1);
-+ dump->u_tsize = ((unsigned long) current->mm->end_code) >> PAGE_SHIFT;
-+ dump->u_dsize = ((unsigned long) (current->mm->brk + (PAGE_SIZE-1))) >> PAGE_SHIFT;
-+ dump->u_dsize -= dump->u_tsize;
-+ dump->u_ssize = 0;
-+ for (i = 0; i < 8; i++)
-+ dump->u_debugreg[i] = current->thread.debugreg[i];
++void __init setup_IO_APIC(void)
++{
++ enable_IO_APIC();
+
-+ if (dump->start_stack < TASK_SIZE)
-+ dump->u_ssize = ((unsigned long) (TASK_SIZE - dump->start_stack)) >> PAGE_SHIFT;
++ if (acpi_ioapic)
++ io_apic_irqs = ~0; /* all IRQs go through IOAPIC */
++ else
++ io_apic_irqs = ~PIC_IRQS;
+
-+ dump->regs.ebx = regs->ebx;
-+ dump->regs.ecx = regs->ecx;
-+ dump->regs.edx = regs->edx;
-+ dump->regs.esi = regs->esi;
-+ dump->regs.edi = regs->edi;
-+ dump->regs.ebp = regs->ebp;
-+ dump->regs.eax = regs->eax;
-+ dump->regs.ds = regs->xds;
-+ dump->regs.es = regs->xes;
-+ savesegment(fs,dump->regs.fs);
-+ savesegment(gs,dump->regs.gs);
-+ dump->regs.orig_eax = regs->orig_eax;
-+ dump->regs.eip = regs->eip;
-+ dump->regs.cs = regs->xcs;
-+ dump->regs.eflags = regs->eflags;
-+ dump->regs.esp = regs->esp;
-+ dump->regs.ss = regs->xss;
++ printk("ENABLING IO-APIC IRQs\n");
+
-+ dump->u_fpvalid = dump_fpu (regs, &dump->i387);
++ /*
++ * Set up IO-APIC IRQ routing.
++ */
++ if (!acpi_ioapic)
++ setup_ioapic_ids_from_mpc();
++#ifndef CONFIG_XEN
++ sync_Arb_IDs();
++#endif
++ setup_IO_APIC_irqs();
++ init_IO_APIC_traps();
++ check_timer();
++ if (!acpi_ioapic)
++ print_IO_APIC();
+}
-+EXPORT_SYMBOL(dump_thread);
+
-+/*
-+ * 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)
++static int __init setup_disable_8254_timer(char *s)
+{
-+ struct pt_regs ptregs = *task_pt_regs(tsk);
-+ ptregs.xcs &= 0xffff;
-+ ptregs.xds &= 0xffff;
-+ ptregs.xes &= 0xffff;
-+ ptregs.xss &= 0xffff;
-+
-+ elf_core_copy_regs(regs, &ptregs);
-+
++ 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);
+
+/*
-+ * This function selects if the context switch from prev to next
-+ * has to tweak the TSC disable bit in the cr4.
++ * Called after all the initialization is done. If we didnt find any
++ * APIC bugs then we can allow the modify fast path
+ */
-+static inline void disable_tsc(struct task_struct *prev_p,
-+ struct task_struct *next_p)
++
++static int __init io_apic_bug_finalize(void)
+{
-+ struct thread_info *prev, *next;
-+
-+ /*
-+ * gcc should eliminate the ->thread_info dereference if
-+ * has_secure_computing returns 0 at compile time (SECCOMP=n).
-+ */
-+ prev = task_thread_info(prev_p);
-+ next = task_thread_info(next_p);
-+
-+ if (has_secure_computing(prev) || has_secure_computing(next)) {
-+ /* slow path here */
-+ if (has_secure_computing(prev) &&
-+ !has_secure_computing(next)) {
-+ write_cr4(read_cr4() & ~X86_CR4_TSD);
-+ } else if (!has_secure_computing(prev) &&
-+ has_secure_computing(next))
-+ write_cr4(read_cr4() | X86_CR4_TSD);
++ if(sis_apic_bug == -1)
++ sis_apic_bug = 0;
++ if (is_initial_xendomain()) {
++ dom0_op_t op = { .cmd = DOM0_PLATFORM_QUIRK };
++ op.u.platform_quirk.quirk_id = sis_apic_bug ?
++ QUIRK_IOAPIC_BAD_REGSEL : QUIRK_IOAPIC_GOOD_REGSEL;
++ HYPERVISOR_dom0_op(&op);
+ }
++ return 0;
+}
+
-+/*
-+ * switch_to(x,yn) should switch tasks from x to y.
-+ *
-+ * We fsave/fwait so that an exception goes off at the right time
-+ * (as a call from the fsave or fwait in effect) rather than to
-+ * the wrong process. Lazy FP saving no longer makes any sense
-+ * with modern CPU's, and this simplifies a lot of things (SMP
-+ * and UP become the same).
-+ *
-+ * NOTE! We used to use the x86 hardware context switching. The
-+ * reason for not using it any more becomes apparent when you
-+ * try to recover gracefully from saved state that is no longer
-+ * valid (stale segment register values in particular). With the
-+ * hardware task-switch, there is no way to fix up bad state in
-+ * a reasonable manner.
-+ *
-+ * The fact that Intel documents the hardware task-switching to
-+ * be slow is a fairly red herring - this code is not noticeably
-+ * faster. However, there _is_ some room for improvement here,
-+ * so the performance issues may eventually be a valid point.
-+ * More important, however, is the fact that this allows us much
-+ * more flexibility.
-+ *
-+ * The return value (in %eax) will be the "prev" task after
-+ * the task-switch, and shows up in ret_from_fork in entry.S,
-+ * for example.
-+ */
-+struct task_struct fastcall * __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
-+ struct physdev_set_iopl iopl_op;
-+ struct physdev_set_iobitmap iobmp_op;
-+ multicall_entry_t _mcl[8], *mcl = _mcl;
++late_initcall(io_apic_bug_finalize);
+
-+ /* XEN NOTE: FS/GS saved in switch_mm(), not here. */
++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];
+
-+ /*
-+ * This is basically '__unlazy_fpu', except that we queue a
-+ * multicall to indicate FPU task switch, rather than
-+ * synchronously trapping to Xen.
-+ */
-+ 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++;
-+ }
-+#if 0 /* lazy fpu sanity check */
-+ else BUG_ON(!(read_cr0() & 8));
-+#endif
++static int ioapic_suspend(struct sys_device *dev, pm_message_t state)
++{
++ struct IO_APIC_route_entry *entry;
++ struct sysfs_ioapic_data *data;
++ int i;
++
++ data = container_of(dev, struct sysfs_ioapic_data, dev);
++ entry = data->entry;
++ for (i = 0; i < nr_ioapic_registers[dev->id]; i ++)
++ entry[i] = ioapic_read_entry(dev->id, i);
+
-+ /*
-+ * Reload esp0.
-+ * This is load_esp0(tss, next) with a multicall.
-+ */
-+ mcl->op = __HYPERVISOR_stack_switch;
-+ mcl->args[0] = __KERNEL_DS;
-+ mcl->args[1] = next->esp0;
-+ mcl++;
++ return 0;
++}
+
-+ /*
-+ * 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].a != prev->tls_array[i].a || \
-+ next->tls_array[i].b != prev->tls_array[i].b)) { \
-+ mcl->op = __HYPERVISOR_update_descriptor; \
-+ *(u64 *)&mcl->args[0] = virt_to_machine( \
-+ &get_cpu_gdt_table(cpu)[GDT_ENTRY_TLS_MIN + i]);\
-+ *(u64 *)&mcl->args[2] = *(u64 *)&next->tls_array[i]; \
-+ mcl++; \
-+ } \
-+} while (0)
-+ C(0); C(1); C(2);
-+#undef C
++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;
+
-+ if (unlikely(prev->iopl != next->iopl)) {
-+ iopl_op.iopl = (next->iopl == 0) ? 1 : (next->iopl >> 12) & 3;
-+ mcl->op = __HYPERVISOR_physdev_op;
-+ mcl->args[0] = PHYSDEVOP_set_iopl;
-+ mcl->args[1] = (unsigned long)&iopl_op;
-+ mcl++;
++ 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);
+ }
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++ for (i = 0; i < nr_ioapic_registers[dev->id]; i ++)
++ ioapic_write_entry(dev->id, i, entry[i]);
+
-+ if (unlikely(prev->io_bitmap_ptr || next->io_bitmap_ptr)) {
-+ 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++;
-+ }
++ return 0;
++}
+
-+ (void)HYPERVISOR_multicall(_mcl, mcl - _mcl);
++static struct sysdev_class ioapic_sysdev_class = {
++ set_kset_name("ioapic"),
++ .suspend = ioapic_suspend,
++ .resume = ioapic_resume,
++};
+
-+ /*
-+ * Restore %fs and %gs if needed.
-+ *
-+ * Glibc normally makes %fs be zero, and %gs is one of
-+ * the TLS segments.
-+ */
-+ if (unlikely(next->fs))
-+ loadsegment(fs, next->fs);
++static int __init ioapic_init_sysfs(void)
++{
++ struct sys_device * dev;
++ int i, size, error = 0;
+
-+ if (next->gs)
-+ loadsegment(gs, next->gs);
++ error = sysdev_class_register(&ioapic_sysdev_class);
++ if (error)
++ return error;
+
-+ /*
-+ * Now maybe reload the debug registers
-+ */
-+ if (unlikely(next->debugreg[7])) {
-+ set_debugreg(next->debugreg[0], 0);
-+ set_debugreg(next->debugreg[1], 1);
-+ set_debugreg(next->debugreg[2], 2);
-+ set_debugreg(next->debugreg[3], 3);
-+ /* no 4 and 5 */
-+ set_debugreg(next->debugreg[6], 6);
-+ set_debugreg(next->debugreg[7], 7);
++ 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;
++ }
+ }
+
-+ disable_tsc(prev_p, next_p);
-+
-+ return prev_p;
-+}
-+
-+asmlinkage int sys_fork(struct pt_regs regs)
-+{
-+ return do_fork(SIGCHLD, regs.esp, ®s, 0, NULL, NULL);
++ return 0;
+}
+
-+asmlinkage int sys_clone(struct pt_regs regs)
-+{
-+ unsigned long clone_flags;
-+ unsigned long newsp;
-+ int __user *parent_tidptr, *child_tidptr;
-+
-+ clone_flags = regs.ebx;
-+ newsp = regs.ecx;
-+ parent_tidptr = (int __user *)regs.edx;
-+ child_tidptr = (int __user *)regs.edi;
-+ if (!newsp)
-+ newsp = regs.esp;
-+ return do_fork(clone_flags, newsp, ®s, 0, parent_tidptr, child_tidptr);
-+}
++device_initcall(ioapic_init_sysfs);
+
+/*
-+ * 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.
++ * Dynamic irq allocate and deallocation
+ */
-+asmlinkage int sys_vfork(struct pt_regs regs)
++int create_irq(void)
+{
-+ return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, regs.esp, ®s, 0, NULL, NULL);
-+}
++ /* Allocate an unused irq */
++ int irq, new, vector = 0;
++ unsigned long flags;
+
-+/*
-+ * sys_execve() executes a new program.
-+ */
-+asmlinkage int sys_execve(struct pt_regs regs)
-+{
-+ int error;
-+ char * filename;
++ irq = -ENOSPC;
++ spin_lock_irqsave(&vector_lock, flags);
++ for (new = (NR_IRQS - 1); new >= 0; new--) {
++ if (platform_legacy_irq(new))
++ continue;
++ if (irq_vector[new] != 0)
++ continue;
++ vector = __assign_irq_vector(new);
++ if (likely(vector > 0))
++ irq = new;
++ break;
++ }
++ spin_unlock_irqrestore(&vector_lock, flags);
+
-+ filename = getname((char __user *) regs.ebx);
-+ error = PTR_ERR(filename);
-+ if (IS_ERR(filename))
-+ goto out;
-+ error = do_execve(filename,
-+ (char __user * __user *) regs.ecx,
-+ (char __user * __user *) regs.edx,
-+ ®s);
-+ if (error == 0) {
-+ task_lock(current);
-+ current->ptrace &= ~PT_DTRACE;
-+ task_unlock(current);
-+ /* Make sure we don't return using sysenter.. */
-+ set_thread_flag(TIF_IRET);
++ if (irq >= 0) {
++#ifndef CONFIG_XEN
++ set_intr_gate(vector, interrupt[irq]);
++#endif
++ dynamic_irq_init(irq);
+ }
-+ putname(filename);
-+out:
-+ return error;
++ return irq;
+}
+
-+#define top_esp (THREAD_SIZE - sizeof(unsigned long))
-+#define top_ebp (THREAD_SIZE - 2*sizeof(unsigned long))
-+
-+unsigned long get_wchan(struct task_struct *p)
++void destroy_irq(unsigned int irq)
+{
-+ unsigned long ebp, esp, eip;
-+ unsigned long stack_page;
-+ int count = 0;
-+ if (!p || p == current || p->state == TASK_RUNNING)
-+ return 0;
-+ stack_page = (unsigned long)task_stack_page(p);
-+ esp = p->thread.esp;
-+ if (!stack_page || esp < stack_page || esp > top_esp+stack_page)
-+ return 0;
-+ /* include/asm-i386/system.h:switch_to() pushes ebp last. */
-+ ebp = *(unsigned long *) esp;
-+ do {
-+ if (ebp < stack_page || ebp > top_ebp+stack_page)
-+ return 0;
-+ eip = *(unsigned long *) (ebp+4);
-+ if (!in_sched_functions(eip))
-+ return eip;
-+ ebp = *(unsigned long *) ebp;
-+ } while (count++ < 16);
-+ return 0;
++ unsigned long flags;
++
++ dynamic_irq_cleanup(irq);
++
++ spin_lock_irqsave(&vector_lock, flags);
++ irq_vector[irq] = 0;
++ spin_unlock_irqrestore(&vector_lock, flags);
+}
+
+/*
-+ * sys_alloc_thread_area: get a yet unused TLS descriptor index.
++ * MSI mesage composition
+ */
-+static int get_free_idx(void)
++#ifdef CONFIG_PCI_MSI
++static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_msg *msg)
+{
-+ struct thread_struct *t = ¤t->thread;
-+ int idx;
++ int vector;
++ unsigned dest;
+
-+ for (idx = 0; idx < GDT_ENTRY_TLS_ENTRIES; idx++)
-+ if (desc_empty(t->tls_array + idx))
-+ return idx + GDT_ENTRY_TLS_MIN;
-+ return -ESRCH;
++ vector = assign_irq_vector(irq);
++ if (vector >= 0) {
++ dest = cpu_mask_to_apicid(TARGET_CPUS);
++
++ msg->address_hi = MSI_ADDR_BASE_HI;
++ msg->address_lo =
++ MSI_ADDR_BASE_LO |
++ ((INT_DEST_MODE == 0) ?
++ MSI_ADDR_DEST_MODE_PHYSICAL:
++ MSI_ADDR_DEST_MODE_LOGICAL) |
++ ((INT_DELIVERY_MODE != dest_LowestPrio) ?
++ MSI_ADDR_REDIRECTION_CPU:
++ MSI_ADDR_REDIRECTION_LOWPRI) |
++ MSI_ADDR_DEST_ID(dest);
++
++ msg->data =
++ MSI_DATA_TRIGGER_EDGE |
++ MSI_DATA_LEVEL_ASSERT |
++ ((INT_DELIVERY_MODE != dest_LowestPrio) ?
++ MSI_DATA_DELIVERY_FIXED:
++ MSI_DATA_DELIVERY_LOWPRI) |
++ MSI_DATA_VECTOR(vector);
++ }
++ return vector;
+}
+
-+/*
-+ * Set a given TLS descriptor:
-+ */
-+asmlinkage int sys_set_thread_area(struct user_desc __user *u_info)
++#ifdef CONFIG_SMP
++static void set_msi_irq_affinity(unsigned int irq, cpumask_t mask)
+{
-+ struct thread_struct *t = ¤t->thread;
-+ struct user_desc info;
-+ struct desc_struct *desc;
-+ int cpu, idx;
++ struct msi_msg msg;
++ unsigned int dest;
++ cpumask_t tmp;
++ int vector;
+
-+ if (copy_from_user(&info, u_info, sizeof(info)))
-+ return -EFAULT;
-+ idx = info.entry_number;
++ cpus_and(tmp, mask, cpu_online_map);
++ if (cpus_empty(tmp))
++ tmp = TARGET_CPUS;
+
-+ /*
-+ * index -1 means the kernel should try to find and
-+ * allocate an empty descriptor:
-+ */
-+ if (idx == -1) {
-+ idx = get_free_idx();
-+ if (idx < 0)
-+ return idx;
-+ if (put_user(idx, &u_info->entry_number))
-+ return -EFAULT;
-+ }
++ vector = assign_irq_vector(irq);
++ if (vector < 0)
++ return;
+
-+ if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX)
-+ return -EINVAL;
++ dest = cpu_mask_to_apicid(mask);
+
-+ desc = t->tls_array + idx - GDT_ENTRY_TLS_MIN;
++ read_msi_msg(irq, &msg);
+
-+ /*
-+ * We must not get preempted while modifying the TLS.
-+ */
-+ cpu = get_cpu();
++ msg.data &= ~MSI_DATA_VECTOR_MASK;
++ msg.data |= MSI_DATA_VECTOR(vector);
++ msg.address_lo &= ~MSI_ADDR_DEST_ID_MASK;
++ msg.address_lo |= MSI_ADDR_DEST_ID(dest);
+
-+ if (LDT_empty(&info)) {
-+ desc->a = 0;
-+ desc->b = 0;
-+ } else {
-+ desc->a = LDT_entry_a(&info);
-+ desc->b = LDT_entry_b(&info);
-+ }
-+ load_TLS(t, cpu);
++ write_msi_msg(irq, &msg);
++ set_native_irq_info(irq, mask);
++}
++#endif /* CONFIG_SMP */
+
-+ put_cpu();
++/*
++ * IRQ Chip for MSI PCI/PCI-X/PCI-Express Devices,
++ * which implement the MSI or MSI-X Capability Structure.
++ */
++static struct irq_chip msi_chip = {
++ .name = "PCI-MSI",
++ .unmask = unmask_msi_irq,
++ .mask = mask_msi_irq,
++ .ack = ack_ioapic_irq,
++#ifdef CONFIG_SMP
++ .set_affinity = set_msi_irq_affinity,
++#endif
++ .retrigger = ioapic_retrigger_irq,
++};
++
++int arch_setup_msi_irq(unsigned int irq, struct pci_dev *dev)
++{
++ struct msi_msg msg;
++ int ret;
++ ret = msi_compose_msg(dev, irq, &msg);
++ if (ret < 0)
++ return ret;
++
++ write_msi_msg(irq, &msg);
++
++ set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq,
++ "edge");
+
+ return 0;
+}
+
++void arch_teardown_msi_irq(unsigned int irq)
++{
++ return;
++}
++
++#endif /* CONFIG_PCI_MSI */
++
+/*
-+ * Get the current Thread-Local Storage area:
++ * Hypertransport interrupt support
+ */
++#ifdef CONFIG_HT_IRQ
+
-+#define GET_BASE(desc) ( \
-+ (((desc)->a >> 16) & 0x0000ffff) | \
-+ (((desc)->b << 16) & 0x00ff0000) | \
-+ ( (desc)->b & 0xff000000) )
-+
-+#define GET_LIMIT(desc) ( \
-+ ((desc)->a & 0x0ffff) | \
-+ ((desc)->b & 0xf0000) )
-+
-+#define GET_32BIT(desc) (((desc)->b >> 22) & 1)
-+#define GET_CONTENTS(desc) (((desc)->b >> 10) & 3)
-+#define GET_WRITABLE(desc) (((desc)->b >> 9) & 1)
-+#define GET_LIMIT_PAGES(desc) (((desc)->b >> 23) & 1)
-+#define GET_PRESENT(desc) (((desc)->b >> 15) & 1)
-+#define GET_USEABLE(desc) (((desc)->b >> 20) & 1)
++#ifdef CONFIG_SMP
+
-+asmlinkage int sys_get_thread_area(struct user_desc __user *u_info)
++static void target_ht_irq(unsigned int irq, unsigned int dest)
+{
-+ struct user_desc info;
-+ struct desc_struct *desc;
-+ int idx;
++ struct ht_irq_msg msg;
++ fetch_ht_irq_msg(irq, &msg);
+
-+ if (get_user(idx, &u_info->entry_number))
-+ return -EFAULT;
-+ if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX)
-+ return -EINVAL;
++ msg.address_lo &= ~(HT_IRQ_LOW_DEST_ID_MASK);
++ msg.address_hi &= ~(HT_IRQ_HIGH_DEST_ID_MASK);
+
-+ memset(&info, 0, sizeof(info));
++ msg.address_lo |= HT_IRQ_LOW_DEST_ID(dest);
++ msg.address_hi |= HT_IRQ_HIGH_DEST_ID(dest);
+
-+ desc = current->thread.tls_array + idx - GDT_ENTRY_TLS_MIN;
++ write_ht_irq_msg(irq, &msg);
++}
+
-+ info.entry_number = idx;
-+ info.base_addr = GET_BASE(desc);
-+ info.limit = GET_LIMIT(desc);
-+ info.seg_32bit = GET_32BIT(desc);
-+ info.contents = GET_CONTENTS(desc);
-+ info.read_exec_only = !GET_WRITABLE(desc);
-+ info.limit_in_pages = GET_LIMIT_PAGES(desc);
-+ info.seg_not_present = !GET_PRESENT(desc);
-+ info.useable = GET_USEABLE(desc);
++static void set_ht_irq_affinity(unsigned int irq, cpumask_t mask)
++{
++ unsigned int dest;
++ cpumask_t tmp;
+
-+ if (copy_to_user(u_info, &info, sizeof(info)))
-+ return -EFAULT;
-+ return 0;
++ 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);
++
++ target_ht_irq(irq, dest);
++ set_native_irq_info(irq, mask);
+}
++#endif
+
-+unsigned long arch_align_stack(unsigned long sp)
++static struct irq_chip ht_irq_chip = {
++ .name = "PCI-HT",
++ .mask = mask_ht_irq,
++ .unmask = unmask_ht_irq,
++ .ack = ack_ioapic_irq,
++#ifdef CONFIG_SMP
++ .set_affinity = set_ht_irq_affinity,
++#endif
++ .retrigger = ioapic_retrigger_irq,
++};
++
++int arch_setup_ht_irq(unsigned int irq, struct pci_dev *dev)
+{
-+ if (randomize_va_space)
-+ sp -= get_random_int() % 8192;
-+ return sp & ~0xf;
++ int vector;
++
++ vector = assign_irq_vector(irq);
++ if (vector >= 0) {
++ struct ht_irq_msg msg;
++ unsigned dest;
++ cpumask_t tmp;
++
++ cpus_clear(tmp);
++ cpu_set(vector >> 8, tmp);
++ dest = cpu_mask_to_apicid(tmp);
++
++ msg.address_hi = HT_IRQ_HIGH_DEST_ID(dest);
++
++ msg.address_lo =
++ HT_IRQ_LOW_BASE |
++ HT_IRQ_LOW_DEST_ID(dest) |
++ HT_IRQ_LOW_VECTOR(vector) |
++ ((INT_DEST_MODE == 0) ?
++ HT_IRQ_LOW_DM_PHYSICAL :
++ HT_IRQ_LOW_DM_LOGICAL) |
++ HT_IRQ_LOW_RQEOI_EDGE |
++ ((INT_DELIVERY_MODE != dest_LowestPrio) ?
++ HT_IRQ_LOW_MT_FIXED :
++ HT_IRQ_LOW_MT_ARBITRATED) |
++ HT_IRQ_LOW_IRQ_MASKED;
++
++ write_ht_irq_msg(irq, &msg);
++
++ set_irq_chip_and_handler_name(irq, &ht_irq_chip,
++ handle_edge_irq, "edge");
++ }
++ return vector;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/quirks-xen.c linux-2.6.18-xen/arch/i386/kernel/quirks-xen.c
---- linux-2.6.18.3/arch/i386/kernel/quirks-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/quirks-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,47 @@
-+/*
-+ * This file contains work-arounds for x86 and x86_64 platform bugs.
-+ */
-+#include <linux/pci.h>
-+#include <linux/irq.h>
++#endif /* CONFIG_HT_IRQ */
+
-+#if defined(CONFIG_X86_IO_APIC) && (defined(CONFIG_SMP) || defined(CONFIG_XEN)) && defined(CONFIG_PCI)
++/* --------------------------------------------------------------------------
++ ACPI-based IOAPIC Configuration
++ -------------------------------------------------------------------------- */
++
++#ifdef CONFIG_ACPI
+
-+static void __devinit quirk_intel_irqbalance(struct pci_dev *dev)
++int __init io_apic_get_unique_id (int ioapic, int apic_id)
+{
-+ u8 config, rev;
-+ u32 word;
++#ifndef CONFIG_XEN
++ union IO_APIC_reg_00 reg_00;
++ static physid_mask_t apic_id_map = PHYSID_MASK_NONE;
++ physid_mask_t tmp;
++ unsigned long flags;
++ int i = 0;
+
-+ /* BIOS may enable hardware IRQ balancing for
-+ * E7520/E7320/E7525(revision ID 0x9 and below)
-+ * based platforms.
-+ * Disable SW irqbalance/affinity on those platforms.
++ /*
++ * The P4 platform supports up to 256 APIC IDs on two separate APIC
++ * buses (one for LAPICs, one for IOAPICs), where predecessors only
++ * supports up to 16 on one shared APIC bus.
++ *
++ * TBD: Expand LAPIC/IOAPIC support on P4-class systems to take full
++ * advantage of new APIC bus architecture.
+ */
-+ pci_read_config_byte(dev, PCI_CLASS_REVISION, &rev);
-+ if (rev > 0x9)
-+ return;
-+
-+ printk(KERN_INFO "Intel E7520/7320/7525 detected.");
+
-+ /* enable access to config space*/
-+ pci_read_config_byte(dev, 0xf4, &config);
-+ pci_write_config_byte(dev, 0xf4, config|0x2);
++ if (physids_empty(apic_id_map))
++ apic_id_map = ioapic_phys_id_map(phys_cpu_present_map);
+
-+ /* read xTPR register */
-+ raw_pci_ops->read(0, 0, 0x40, 0x4c, 2, &word);
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_00.raw = io_apic_read(ioapic, 0);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
+
-+ if (!(word & (1 << 13))) {
-+ dom0_op_t op;
-+ printk(KERN_INFO "Disabling irq balancing and affinity\n");
-+ op.cmd = DOM0_PLATFORM_QUIRK;
-+ op.u.platform_quirk.quirk_id = QUIRK_NOIRQBALANCING;
-+ (void)HYPERVISOR_dom0_op(&op);
++ if (apic_id >= get_physical_broadcast()) {
++ printk(KERN_WARNING "IOAPIC[%d]: Invalid apic_id %d, trying "
++ "%d\n", ioapic, apic_id, reg_00.bits.ID);
++ apic_id = reg_00.bits.ID;
+ }
+
-+ /* put back the original value for config space*/
-+ if (!(config & 0x2))
-+ pci_write_config_byte(dev, 0xf4, config);
-+}
-+DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_E7320_MCH, quirk_intel_irqbalance);
-+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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/setup-xen.c linux-2.6.18-xen/arch/i386/kernel/setup-xen.c
---- linux-2.6.18.3/arch/i386/kernel/setup-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/setup-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,1835 @@
-+/*
-+ * linux/arch/i386/kernel/setup.c
-+ *
-+ * Copyright (C) 1995 Linus Torvalds
-+ *
-+ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
-+ *
-+ * Memory region support
-+ * David Parsons <orc at pell.chi.il.us>, July-August 1999
-+ *
-+ * Added E820 sanitization routine (removes overlapping memory regions);
-+ * Brian Moyle <bmoyle at mvista.com>, February 2001
-+ *
-+ * Moved CPU detection code to cpu/${cpu}.c
-+ * Patrick Mochel <mochel at osdl.org>, March 2002
-+ *
-+ * Provisions for empty E820 memory regions (reported by certain BIOSes).
-+ * Alex Achenbach <xela at slit.de>, December 2002.
-+ *
-+ */
-+
-+/*
-+ * This file handles the architecture-dependent parts of initialization
-+ */
++ /*
++ * Every APIC in a system must have a unique ID or we get lots of nice
++ * 'stuck on smp_invalidate_needed IPI wait' messages.
++ */
++ if (check_apicid_used(apic_id_map, apic_id)) {
+
-+#include <linux/sched.h>
-+#include <linux/mm.h>
-+#include <linux/mmzone.h>
-+#include <linux/screen_info.h>
-+#include <linux/ioport.h>
-+#include <linux/acpi.h>
-+#include <linux/apm_bios.h>
-+#include <linux/initrd.h>
-+#include <linux/bootmem.h>
-+#include <linux/seq_file.h>
-+#include <linux/platform_device.h>
-+#include <linux/console.h>
-+#include <linux/mca.h>
-+#include <linux/root_dev.h>
-+#include <linux/highmem.h>
-+#include <linux/module.h>
-+#include <linux/efi.h>
-+#include <linux/init.h>
-+#include <linux/edd.h>
-+#include <linux/nodemask.h>
-+#include <linux/kernel.h>
-+#include <linux/percpu.h>
-+#include <linux/notifier.h>
-+#include <linux/kexec.h>
-+#include <linux/crash_dump.h>
-+#include <linux/dmi.h>
-+#include <linux/pfn.h>
++ for (i = 0; i < get_physical_broadcast(); i++) {
++ if (!check_apicid_used(apic_id_map, i))
++ break;
++ }
+
-+#include <video/edid.h>
++ if (i == get_physical_broadcast())
++ panic("Max apic_id exceeded!\n");
+
-+#include <asm/apic.h>
-+#include <asm/e820.h>
-+#include <asm/mpspec.h>
-+#include <asm/setup.h>
-+#include <asm/arch_hooks.h>
-+#include <asm/sections.h>
-+#include <asm/io_apic.h>
-+#include <asm/ist.h>
-+#include <asm/io.h>
-+#include <asm/hypervisor.h>
-+#include <xen/interface/physdev.h>
-+#include <xen/interface/memory.h>
-+#include <xen/features.h>
-+#include <xen/xencons.h>
-+#include "setup_arch.h"
-+#include <bios_ebda.h>
++ printk(KERN_WARNING "IOAPIC[%d]: apic_id %d already used, "
++ "trying %d\n", ioapic, apic_id, i);
+
-+/* Forward Declaration. */
-+void __init find_max_pfn(void);
++ apic_id = i;
++ }
+
-+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 */
-+};
++ tmp = apicid_to_cpu_present(apic_id);
++ physids_or(apic_id_map, apic_id_map, tmp);
+
-+extern char hypercall_page[PAGE_SIZE];
-+EXPORT_SYMBOL(hypercall_page);
++ if (reg_00.bits.ID != apic_id) {
++ reg_00.bits.ID = apic_id;
+
-+int disable_pse __devinitdata = 0;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ io_apic_write(ioapic, 0, reg_00.raw);
++ reg_00.raw = io_apic_read(ioapic, 0);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
+
-+/*
-+ * Machine setup..
-+ */
++ /* Sanity check */
++ if (reg_00.bits.ID != apic_id) {
++ printk("IOAPIC[%d]: Unable to change apic_id!\n", ioapic);
++ return -1;
++ }
++ }
+
-+#ifdef CONFIG_EFI
-+int efi_enabled = 0;
-+EXPORT_SYMBOL(efi_enabled);
-+#endif
++ apic_printk(APIC_VERBOSE, KERN_INFO
++ "IOAPIC[%d]: Assigned apic_id %d\n", ioapic, apic_id);
++#endif /* !CONFIG_XEN */
+
-+/* cpu data as detected by the assembly code in head.S */
-+struct cpuinfo_x86 new_cpu_data __initdata = { 0, 0, 0, 0, -1, 1, 0, 0, -1 };
-+/* common cpu data for all cpus */
-+struct cpuinfo_x86 boot_cpu_data __read_mostly = { 0, 0, 0, 0, -1, 1, 0, 0, -1 };
-+EXPORT_SYMBOL(boot_cpu_data);
++ return apic_id;
++}
+
-+unsigned long mmu_cr4_features;
+
-+#ifdef CONFIG_ACPI
-+ int acpi_disabled = 0;
-+#else
-+ int acpi_disabled = 1;
-+#endif
-+EXPORT_SYMBOL(acpi_disabled);
++int __init io_apic_get_version (int ioapic)
++{
++ union IO_APIC_reg_01 reg_01;
++ unsigned long flags;
+
-+#ifdef CONFIG_ACPI
-+int __initdata acpi_force = 0;
-+extern acpi_interrupt_flags acpi_sci_flags;
-+#endif
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_01.raw = io_apic_read(ioapic, 1);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
+
-+/* for MCA, but anyone else can use it if they want */
-+unsigned int machine_id;
-+#ifdef CONFIG_MCA
-+EXPORT_SYMBOL(machine_id);
-+#endif
-+unsigned int machine_submodel_id;
-+unsigned int BIOS_revision;
-+unsigned int mca_pentium_flag;
++ return reg_01.bits.version;
++}
+
-+/* For PCI or other memory-mapped resources */
-+unsigned long pci_mem_start = 0x10000000;
-+#ifdef CONFIG_PCI
-+EXPORT_SYMBOL(pci_mem_start);
-+#endif
+
-+/* Boot loader ID as an integer, for the benefit of proc_dointvec */
-+int bootloader_type;
++int __init io_apic_get_redir_entries (int ioapic)
++{
++ union IO_APIC_reg_01 reg_01;
++ unsigned long flags;
+
-+/* user-defined highmem size */
-+static unsigned int highmem_pages = -1;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_01.raw = io_apic_read(ioapic, 1);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
+
-+/*
-+ * Setup options
-+ */
-+struct drive_info_struct { char dummy[32]; } drive_info;
-+#if defined(CONFIG_BLK_DEV_IDE) || defined(CONFIG_BLK_DEV_HD) || \
-+ defined(CONFIG_BLK_DEV_IDE_MODULE) || defined(CONFIG_BLK_DEV_HD_MODULE)
-+EXPORT_SYMBOL(drive_info);
-+#endif
-+struct screen_info screen_info;
-+EXPORT_SYMBOL(screen_info);
-+struct apm_info apm_info;
-+EXPORT_SYMBOL(apm_info);
-+struct sys_desc_table_struct {
-+ unsigned short length;
-+ unsigned char table[0];
-+};
-+struct edid_info edid_info;
-+EXPORT_SYMBOL_GPL(edid_info);
-+struct ist_info ist_info;
-+#if defined(CONFIG_X86_SPEEDSTEP_SMI) || \
-+ defined(CONFIG_X86_SPEEDSTEP_SMI_MODULE)
-+EXPORT_SYMBOL(ist_info);
-+#endif
-+struct e820map e820;
-+static void __init e820_setup_gap(struct e820entry *e820, int nr_map);
-+#ifdef CONFIG_XEN
-+struct e820map machine_e820;
-+#endif
++ return reg_01.bits.entries;
++}
+
-+extern void early_cpu_init(void);
-+extern void generic_apic_probe(char *);
-+extern int root_mountflags;
+
-+unsigned long saved_videomode;
++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;
+
-+#define RAMDISK_IMAGE_START_MASK 0x07FF
-+#define RAMDISK_PROMPT_FLAG 0x8000
-+#define RAMDISK_LOAD_FLAG 0x4000
++ if (!IO_APIC_IRQ(irq)) {
++ printk(KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
++ ioapic);
++ return -EINVAL;
++ }
+
-+static char command_line[COMMAND_LINE_SIZE];
++ /*
++ * 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.
++ */
+
-+unsigned char __initdata boot_params[PARAM_SIZE];
++ memset(&entry,0,sizeof(entry));
+
-+static struct resource data_resource = {
-+ .name = "Kernel data",
-+ .start = 0,
-+ .end = 0,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_MEM
-+};
++ 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;
+
-+static struct resource code_resource = {
-+ .name = "Kernel code",
-+ .start = 0,
-+ .end = 0,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_MEM
-+};
++ /*
++ * IRQs < 16 are already in the irq_2_pin[] map
++ */
++ if (irq >= 16)
++ add_pin_to_irq(irq, ioapic, pin);
+
-+static struct resource system_rom_resource = {
-+ .name = "System ROM",
-+ .start = 0xf0000,
-+ .end = 0xfffff,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM
-+};
++ entry.vector = assign_irq_vector(irq);
+
-+static struct resource extension_rom_resource = {
-+ .name = "Extension ROM",
-+ .start = 0xe0000,
-+ .end = 0xeffff,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM
-+};
-+
-+static struct resource adapter_rom_resources[] = { {
-+ .name = "Adapter ROM",
-+ .start = 0xc8000,
-+ .end = 0,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM
-+}, {
-+ .name = "Adapter ROM",
-+ .start = 0,
-+ .end = 0,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM
-+}, {
-+ .name = "Adapter ROM",
-+ .start = 0,
-+ .end = 0,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM
-+}, {
-+ .name = "Adapter ROM",
-+ .start = 0,
-+ .end = 0,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM
-+}, {
-+ .name = "Adapter ROM",
-+ .start = 0,
-+ .end = 0,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM
-+}, {
-+ .name = "Adapter ROM",
-+ .start = 0,
-+ .end = 0,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM
-+} };
++ apic_printk(APIC_DEBUG, 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);
+
-+#define ADAPTER_ROM_RESOURCES \
-+ (sizeof adapter_rom_resources / sizeof adapter_rom_resources[0])
++ ioapic_register_intr(irq, entry.vector, edge_level);
+
-+static struct resource video_rom_resource = {
-+ .name = "Video ROM",
-+ .start = 0xc0000,
-+ .end = 0xc7fff,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM
-+};
++ if (!ioapic && (irq < 16))
++ disable_8259A_irq(irq);
+
-+static struct resource video_ram_resource = {
-+ .name = "Video RAM area",
-+ .start = 0xa0000,
-+ .end = 0xbffff,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_MEM
-+};
-+
-+static struct resource standard_io_resources[] = { {
-+ .name = "dma1",
-+ .start = 0x0000,
-+ .end = 0x001f,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO
-+}, {
-+ .name = "pic1",
-+ .start = 0x0020,
-+ .end = 0x0021,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO
-+}, {
-+ .name = "timer0",
-+ .start = 0x0040,
-+ .end = 0x0043,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO
-+}, {
-+ .name = "timer1",
-+ .start = 0x0050,
-+ .end = 0x0053,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO
-+}, {
-+ .name = "keyboard",
-+ .start = 0x0060,
-+ .end = 0x006f,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO
-+}, {
-+ .name = "dma page reg",
-+ .start = 0x0080,
-+ .end = 0x008f,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO
-+}, {
-+ .name = "pic2",
-+ .start = 0x00a0,
-+ .end = 0x00a1,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO
-+}, {
-+ .name = "dma2",
-+ .start = 0x00c0,
-+ .end = 0x00df,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO
-+}, {
-+ .name = "fpu",
-+ .start = 0x00f0,
-+ .end = 0x00ff,
-+ .flags = IORESOURCE_BUSY | IORESOURCE_IO
-+} };
++ spin_lock_irqsave(&ioapic_lock, flags);
++ __ioapic_write_entry(ioapic, pin, entry);
++ set_native_irq_info(irq, TARGET_CPUS);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
+
-+#define STANDARD_IO_RESOURCES \
-+ (sizeof standard_io_resources / sizeof standard_io_resources[0])
++ return 0;
++}
+
-+#define romsignature(x) (*(unsigned short *)(x) == 0xaa55)
++#endif /* CONFIG_ACPI */
+
-+static int __init romchecksum(unsigned char *rom, unsigned long length)
++static int __init parse_disable_timer_pin_1(char *arg)
+{
-+ unsigned char *p, sum = 0;
++ disable_timer_pin_1 = 1;
++ return 0;
++}
++early_param("disable_timer_pin_1", parse_disable_timer_pin_1);
+
-+ for (p = rom; p < rom + length; p++)
-+ sum += *p;
-+ return sum == 0;
++static int __init parse_enable_timer_pin_1(char *arg)
++{
++ disable_timer_pin_1 = -1;
++ return 0;
+}
++early_param("enable_timer_pin_1", parse_enable_timer_pin_1);
+
-+static void __init probe_roms(void)
++static int __init parse_noapic(char *arg)
+{
-+ unsigned long start, length, upper;
-+ unsigned char *rom;
-+ int i;
++ /* disable IO-APIC */
++ disable_ioapic_setup();
++ return 0;
++}
++early_param("noapic", parse_noapic);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/ioport-xen.c b/arch/i386/kernel/ioport-xen.c
+--- a/arch/i386/kernel/ioport-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/ioport-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,121 @@
++/*
++ * linux/arch/i386/kernel/ioport.c
++ *
++ * This contains the io-permission bitmap code - written by obz, with changes
++ * by Linus.
++ */
+
-+#ifdef CONFIG_XEN
-+ /* Nothing to do if not running in dom0. */
-+ if (!is_initial_xendomain())
-+ return;
-+#endif
++#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/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>
+
-+ /* 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;
++/* 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)
++{
++ unsigned long mask;
++ unsigned long *bitmap_base = bitmap + (base / BITS_PER_LONG);
++ unsigned int low_index = base & (BITS_PER_LONG-1);
++ int length = low_index + extent;
+
-+ video_rom_resource.start = start;
++ if (low_index != 0) {
++ mask = (~0UL << low_index);
++ if (length < BITS_PER_LONG)
++ mask &= ~(~0UL << length);
++ if (new_value)
++ *bitmap_base++ |= mask;
++ else
++ *bitmap_base++ &= ~mask;
++ length -= BITS_PER_LONG;
++ }
+
-+ /* 0 < length <= 0x7f * 512, historically */
-+ length = rom[2] * 512;
++ mask = (new_value ? ~0UL : 0UL);
++ while (length >= BITS_PER_LONG) {
++ *bitmap_base++ = mask;
++ length -= BITS_PER_LONG;
++ }
+
-+ /* if checksum okay, trust length byte */
-+ if (length && romchecksum(rom, length))
-+ video_rom_resource.end = start + length - 1;
-+ break;
++ if (length > 0) {
++ mask = ~(~0UL << length);
++ if (new_value)
++ *bitmap_base++ |= mask;
++ else
++ *bitmap_base++ &= ~mask;
+ }
++}
+
-+ 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;
++/*
++ * 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;
+
-+ /* 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;
-+ }
++ 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_iobitmap.bitmap = (char *)bitmap;
++ set_iobitmap.nr_ports = IO_BITMAP_BITS;
++ HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap, &set_iobitmap);
+ }
+
-+ /* 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;
++ set_bitmap(t->io_bitmap_ptr, from, num, !turn_on);
+
-+ /* 0 < length <= 0x7f * 512, historically */
-+ length = rom[2] * 512;
++ return 0;
++}
+
-+ /* but accept any length that fits if checksum okay */
-+ if (!length || start + length > upper || !romchecksum(rom, length))
-+ continue;
++/*
++ * 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.
++ *
++ * Here we just change the eflags value on the stack: we allow
++ * only the super-user to do it. This depends on the stack-layout
++ * on system-call entry - see also fork() and the signal handling
++ * code.
++ */
+
-+ adapter_rom_resources[i].start = start;
-+ adapter_rom_resources[i].end = start + length - 1;
-+ request_resource(&iomem_resource, &adapter_rom_resources[i]);
++asmlinkage long sys_iopl(unsigned long unused)
++{
++ volatile struct pt_regs * regs = (struct pt_regs *) &unused;
++ unsigned int level = regs->ebx;
++ struct thread_struct *t = ¤t->thread;
++ unsigned int old = (t->iopl >> 12) & 3;
+
-+ start = adapter_rom_resources[i++].end & ~2047UL;
++ if (level > 3)
++ return -EINVAL;
++ /* Trying to gain more privileges? */
++ if (level > old) {
++ if (!capable(CAP_SYS_RAWIO))
++ return -EPERM;
+ }
++ t->iopl = level << 12;
++ set_iopl_mask(t->iopl);
++ return 0;
+}
-+
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/irq.c b/arch/i386/kernel/irq.c
+--- a/arch/i386/kernel/irq.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/irq.c 2007-03-14 10:55:14.000000000 +0100
+@@ -302,7 +302,9 @@
+
+ cpus_and(mask, irq_desc[irq].affinity, map);
+ if (any_online_cpu(mask) == NR_CPUS) {
++#ifndef CONFIG_XEN
+ printk("Breaking affinity for irq %i\n", irq);
++#endif
+ mask = map;
+ }
+ if (irq_desc[irq].chip->set_affinity)
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/ldt-xen.c b/arch/i386/kernel/ldt-xen.c
+--- a/arch/i386/kernel/ldt-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/ldt-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,268 @@
+/*
-+ * Point at the empty zero page to start with. We map the real shared_info
-+ * page as soon as fixmap is up and running.
++ * linux/arch/i386/kernel/ldt.c
++ *
++ * Copyright (C) 1992 Krishna Balasubramanian and Linus Torvalds
++ * Copyright (C) 1999 Ingo Molnar <mingo at redhat.com>
+ */
-+shared_info_t *HYPERVISOR_shared_info = (shared_info_t *)empty_zero_page;
-+EXPORT_SYMBOL(HYPERVISOR_shared_info);
+
-+unsigned long *phys_to_machine_mapping;
-+unsigned long *pfn_to_mfn_frame_list_list, *pfn_to_mfn_frame_list[16];
-+EXPORT_SYMBOL(phys_to_machine_mapping);
++#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>
+
-+/* Raw start-of-day parameters from the hypervisor. */
-+start_info_t *xen_start_info;
-+EXPORT_SYMBOL(xen_start_info);
++#include <asm/uaccess.h>
++#include <asm/system.h>
++#include <asm/ldt.h>
++#include <asm/desc.h>
++#include <asm/mmu_context.h>
+
-+void __init add_memory_region(unsigned long long start,
-+ unsigned long long size, int type)
++#ifdef CONFIG_SMP /* avoids "defined but not used" warnig */
++static void flush_ldt(void *null)
+{
-+ int x;
-+
-+ if (!efi_enabled) {
-+ 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++;
-+ }
-+} /* add_memory_region */
++ if (current->active_mm)
++ load_LDT(¤t->active_mm->context);
++}
++#endif
+
-+static void __init limit_regions(unsigned long long size)
++static int alloc_ldt(mm_context_t *pc, int mincount, int reload)
+{
-+ unsigned long long current_addr = 0;
-+ int i;
++ void *oldldt;
++ void *newldt;
++ int oldsize;
+
-+ if (efi_enabled) {
-+ efi_memory_desc_t *md;
-+ void *p;
++ if (mincount <= 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);
+
-+ for (p = memmap.map, i = 0; p < memmap.map_end;
-+ p += memmap.desc_size, i++) {
-+ md = p;
-+ current_addr = md->phys_addr + (md->num_pages << 12);
-+ if (md->type == EFI_CONVENTIONAL_MEMORY) {
-+ if (current_addr >= size) {
-+ md->num_pages -=
-+ (((current_addr-size) + PAGE_SIZE-1) >> PAGE_SHIFT);
-+ memmap.nr_map = i + 1;
-+ return;
-+ }
-+ }
-+ }
-+ }
-+ for (i = 0; i < e820.nr_map; i++) {
-+ current_addr = e820.map[i].addr + e820.map[i].size;
-+ if (current_addr < size)
-+ continue;
++ if (!newldt)
++ return -ENOMEM;
+
-+ if (e820.map[i].type != E820_RAM)
-+ continue;
++ 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);
++ pc->ldt = newldt;
++ wmb();
++ pc->size = mincount;
++ wmb();
+
-+ if (e820.map[i].addr >= size) {
-+ /*
-+ * This region starts past the end of the
-+ * requested size, skip it completely.
-+ */
-+ e820.nr_map = i;
-+ } else {
-+ e820.nr_map = i + 1;
-+ e820.map[i].size -= current_addr - size;
-+ }
-+ return;
++ 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
+ }
-+#ifdef CONFIG_XEN
-+ if (i==e820.nr_map && current_addr < size) {
-+ /*
-+ * The e820 map finished before our requested size so
-+ * extend the final entry to the requested address.
-+ */
-+ --i;
-+ if (e820.map[i].type == E820_RAM)
-+ e820.map[i].size -= current_addr - size;
++ 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
-+ add_memory_region(current_addr, size - current_addr, E820_RAM);
++ kfree(oldldt);
+ }
-+#endif
++ return 0;
+}
+
-+#define E820_DEBUG 1
++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;
++}
+
-+static void __init print_memory_map(char *who)
++/*
++ * 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)
+{
-+ int i;
++ struct mm_struct * old_mm;
++ int retval = 0;
+
-+ for (i = 0; i < e820.nr_map; i++) {
-+ printk(" %s: %016Lx - %016Lx ", who,
-+ e820.map[i].addr,
-+ 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 %lu\n", e820.map[i].type);
-+ break;
-+ }
++ init_MUTEX(&mm->context.sem);
++ mm->context.size = 0;
++ mm->context.has_foreign_mappings = 0;
++ 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);
+ }
++ return retval;
+}
+
+/*
-+ * Sanitize the BIOS e820 map.
-+ *
-+ * Some e820 responses include overlapping entries. The following
-+ * replaces the original e820 map with a new one, removing overlaps.
-+ *
++ * No need to lock the MM as we are the last user
+ */
-+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;
++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;
++ }
++}
+
-+int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map)
++static int read_ldt(void __user * ptr, unsigned long bytecount)
+{
-+ 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;
++ int err;
++ unsigned long size;
++ struct mm_struct * mm = current->mm;
+
-+ /*
-+ Visually we're performing the following (1,2,3,4 = memory types)...
++ if (!mm->context.size)
++ return 0;
++ if (bytecount > LDT_ENTRY_SIZE*LDT_ENTRIES)
++ bytecount = LDT_ENTRY_SIZE*LDT_ENTRIES;
+
-+ Sample memory map (w/overlaps):
-+ ____22__________________
-+ ______________________4_
-+ ____1111________________
-+ _44_____________________
-+ 11111111________________
-+ ____________________33__
-+ ___________44___________
-+ __________33333_________
-+ ______________22________
-+ ___________________2222_
-+ _________111111111______
-+ _____________________11_
-+ _________________4______
++ down(&mm->context.sem);
++ size = mm->context.size*LDT_ENTRY_SIZE;
++ if (size > bytecount)
++ size = bytecount;
+
-+ Sanitized equivalent (no overlap):
-+ 1_______________________
-+ _44_____________________
-+ ___1____________________
-+ ____22__________________
-+ ______11________________
-+ _________1______________
-+ __________3_____________
-+ ___________44___________
-+ _____________33_________
-+ _______________2________
-+ ________________1_______
-+ _________________4______
-+ ___________________2____
-+ ____________________33__
-+ ______________________4_
-+ */
++ 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;
++}
+
-+ /* if there's only one memory region, don't bother */
-+ if (*pnr_map < 2)
-+ return -1;
++static int read_default_ldt(void __user * ptr, unsigned long bytecount)
++{
++ int err;
++ unsigned long size;
+
-+ old_nr = *pnr_map;
++ err = 0;
++ size = 5*sizeof(struct desc_struct);
++ if (size > bytecount)
++ size = bytecount;
+
-+ /* 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;
++ err = size;
++ if (clear_user(ptr, size))
++ err = -EFAULT;
+
-+ /* create pointers for initial change-point information (for sorting) */
-+ for (i=0; i < 2*old_nr; i++)
-+ change_point[i] = &change_point_list[i];
++ return err;
++}
+
-+ /* 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];
-+ }
++static int write_ldt(void __user * ptr, unsigned long bytecount, int oldmode)
++{
++ struct mm_struct * mm = current->mm;
++ __u32 entry_1, entry_2;
++ 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, sizeof(ldt_info)))
++ 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;
+ }
-+ chg_nr = chgidx; /* true number of change-points */
+
-+ /* 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;
-+ }
-+ }
++ down(&mm->context.sem);
++ if (ldt_info.entry_number >= mm->context.size) {
++ error = alloc_ldt(¤t->mm->context, ldt_info.entry_number+1, 1);
++ if (error < 0)
++ goto out_unlock;
+ }
+
-+ /* 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;
++ /* 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;
+ }
+ }
-+ 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;
++ entry_1 = LDT_entry_a(&ldt_info);
++ entry_2 = LDT_entry_b(&ldt_info);
++ if (oldmode)
++ entry_2 &= ~(1 << 20);
+
-+ return 0;
++ /* Install the new entry ... */
++install:
++ error = write_ldt_entry(mm->context.ldt, ldt_info.entry_number,
++ entry_1, entry_2);
++
++out_unlock:
++ up(&mm->context.sem);
++out:
++ return error;
+}
+
-+/*
-+ * 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.)
-+ */
-+int __init copy_e820_map(struct e820entry * biosmap, int nr_map)
++asmlinkage int sys_modify_ldt(int func, void __user *ptr, unsigned long bytecount)
+{
-+#ifndef CONFIG_XEN
-+ /* Only one memory region (or negative)? Ignore it */
-+ if (nr_map < 2)
-+ return -1;
-+#else
-+ BUG_ON(nr_map < 1);
-+#endif
++ int ret = -ENOSYS;
+
-+ do {
-+ unsigned long long start = biosmap->addr;
-+ unsigned long long size = biosmap->size;
-+ unsigned long long end = start + size;
-+ unsigned long type = biosmap->type;
++ 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 -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/Makefile b/arch/i386/kernel/Makefile
+--- a/arch/i386/kernel/Makefile 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/Makefile 2007-03-14 10:55:14.000000000 +0100
+@@ -47,6 +47,12 @@
+
+ obj-$(CONFIG_SCx200) += scx200.o
+
++ifdef CONFIG_XEN
++vsyscall_note := vsyscall-note-xen.o
++else
++vsyscall_note := vsyscall-note.o
++endif
+
-+ /* Overflow in 64 bits? Ignore the memory map. */
-+ if (start > end)
-+ return -1;
+ # 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
+@@ -68,7 +74,7 @@
+
+ $(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
+@@ -80,9 +86,20 @@
+
+ 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
+ stacktrace-y += ../../x86_64/kernel/stacktrace.o
+
++ifdef CONFIG_XEN
++include $(srctree)/scripts/Makefile.xen
+
-+#ifndef CONFIG_XEN
-+ /*
-+ * Some BIOSes claim RAM in the 640k - 1M region.
-+ * Not right. Fix it up.
-+ */
-+ if (type == E820_RAM) {
-+ if (start < 0x100000ULL && end > 0xA0000ULL) {
-+ if (start < 0xA0000ULL)
-+ add_memory_region(start, 0xA0000ULL-start, type);
-+ if (end <= 0x100000ULL)
-+ continue;
-+ start = 0x100000ULL;
-+ size = end - start;
-+ }
-+ }
-+#endif
-+ add_memory_region(start, size, type);
-+ } while (biosmap++,--nr_map);
-+ return 0;
-+}
++obj-y += fixup.o
++microcode-$(subst m,y,$(CONFIG_MICROCODE)) := microcode-xen.o
++n-obj-xen := i8259.o timers/ reboot.o smpboot.o trampoline.o
+
-+#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.
++obj-y := $(call filterxen, $(obj-y), $(n-obj-xen))
++obj-y := $(call cherrypickxen, $(obj-y))
++extra-y := $(call cherrypickxen, $(extra-y))
++endif
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/microcode-xen.c b/arch/i386/kernel/microcode-xen.c
+--- a/arch/i386/kernel/microcode-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/microcode-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,147 @@
++/*
++ * Intel CPU Microcode Update Driver for Linux
++ *
++ * Copyright (C) 2000-2004 Tigran Aivazian
++ *
++ * This driver allows to upgrade microcode on Intel processors
++ * belonging to IA-32 family - PentiumPro, Pentium II,
++ * Pentium III, Xeon, Pentium 4, etc.
++ *
++ * Reference: Section 8.10 of Volume III, Intel Pentium 4 Manual,
++ * Order Number 245472 or free download from:
++ *
++ * http://developer.intel.com/design/pentium4/manuals/245472.htm
++ *
++ * For more information, go to http://www.urbanmyth.org/microcode
+ *
++ * 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.
+ */
-+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
+
-+static void __init parse_cmdline_early (char ** cmdline_p)
-+{
-+ char c = ' ', *to = command_line, *from = saved_command_line;
-+ int len = 0, max_cmdline;
-+ int userdef = 0;
-+
-+ 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);
-+ /* Save unparsed command line copy for /proc/cmdline */
-+ saved_command_line[max_cmdline-1] = '\0';
++//#define DEBUG /* pr_debug */
++#include <linux/capability.h>
++#include <linux/kernel.h>
++#include <linux/init.h>
++#include <linux/sched.h>
++#include <linux/cpumask.h>
++#include <linux/module.h>
++#include <linux/slab.h>
++#include <linux/vmalloc.h>
++#include <linux/miscdevice.h>
++#include <linux/spinlock.h>
++#include <linux/mm.h>
++#include <linux/mutex.h>
++#include <linux/syscalls.h>
+
-+ for (;;) {
-+ if (c != ' ')
-+ goto next_char;
-+ /*
-+ * "mem=nopentium" disables the 4MB page tables.
-+ * "mem=XXX[kKmM]" defines a memory region from HIGH_MEM
-+ * to <mem>, overriding the bios size.
-+ * "memmap=XXX[KkmM]@XXX[KkmM]" defines a memory region from
-+ * <start> to <start>+<mem>, overriding the bios size.
-+ *
-+ * HPA tells me bootloaders need to parse mem=, so no new
-+ * option should be mem= [also see Documentation/i386/boot.txt]
-+ */
-+ if (!memcmp(from, "mem=", 4)) {
-+ if (to != command_line)
-+ to--;
-+ if (!memcmp(from+4, "nopentium", 9)) {
-+ from += 9+4;
-+ clear_bit(X86_FEATURE_PSE, boot_cpu_data.x86_capability);
-+ disable_pse = 1;
-+ } else {
-+ /* If the user specifies memory size, we
-+ * limit the BIOS-provided memory map to
-+ * that size. exactmap can be used to specify
-+ * the exact map. mem=number can be used to
-+ * trim the existing memory map.
-+ */
-+ unsigned long long mem_size;
-+
-+ mem_size = memparse(from+4, &from);
-+ limit_regions(mem_size);
-+ userdef=1;
-+ }
-+ }
++#include <asm/msr.h>
++#include <asm/uaccess.h>
++#include <asm/processor.h>
+
-+ else if (!memcmp(from, "memmap=", 7)) {
-+ if (to != command_line)
-+ to--;
-+ 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.
-+ */
-+ find_max_pfn();
-+ saved_max_pfn = max_pfn;
-+#endif
-+ from += 8+7;
-+ e820.nr_map = 0;
-+ userdef = 1;
-+ } else {
-+ /* If the user specifies memory size, we
-+ * limit the BIOS-provided memory map to
-+ * that size. exactmap can be used to specify
-+ * the exact map. mem=number can be used to
-+ * trim the existing memory map.
-+ */
-+ unsigned long long start_at, mem_size;
-+
-+ mem_size = memparse(from+7, &from);
-+ if (*from == '@') {
-+ start_at = memparse(from+1, &from);
-+ add_memory_region(start_at, mem_size, E820_RAM);
-+ } else if (*from == '#') {
-+ start_at = memparse(from+1, &from);
-+ add_memory_region(start_at, mem_size, E820_ACPI);
-+ } else if (*from == '$') {
-+ start_at = memparse(from+1, &from);
-+ add_memory_region(start_at, mem_size, E820_RESERVED);
-+ } else {
-+ limit_regions(mem_size);
-+ userdef=1;
-+ }
-+ }
-+ }
++MODULE_DESCRIPTION("Intel CPU (IA-32) Microcode Update Driver");
++MODULE_AUTHOR("Tigran Aivazian <tigran at veritas.com>");
++MODULE_LICENSE("GPL");
+
-+ else if (!memcmp(from, "noexec=", 7))
-+ noexec_setup(from + 7);
++#define MICROCODE_VERSION "1.14-xen"
+
++#define DEFAULT_UCODE_DATASIZE (2000) /* 2000 bytes */
++#define MC_HEADER_SIZE (sizeof (microcode_header_t)) /* 48 bytes */
++#define DEFAULT_UCODE_TOTALSIZE (DEFAULT_UCODE_DATASIZE + MC_HEADER_SIZE) /* 2048 bytes */
+
-+#ifdef CONFIG_X86_MPPARSE
-+ /*
-+ * 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;
++/* no concurrent ->write()s are allowed on /dev/cpu/microcode */
++static DEFINE_MUTEX(microcode_mutex);
+
-+ maxcpus = simple_strtoul(from + 8, NULL, 0);
-+ }
-+#endif
++static void __user *user_buffer; /* user area microcode data buffer */
++static unsigned int user_buffer_size; /* it's size */
++
++static int microcode_open (struct inode *unused1, struct file *unused2)
++{
++ return capable(CAP_SYS_RAWIO) ? 0 : -EPERM;
++}
+
-+#ifdef CONFIG_ACPI
-+ /* "acpi=off" disables both ACPI table parsing and interpreter */
-+ else if (!memcmp(from, "acpi=off", 8)) {
-+ disable_acpi();
-+ }
+
-+ /* acpi=force to over-ride black-list */
-+ else if (!memcmp(from, "acpi=force", 10)) {
-+ acpi_force = 1;
-+ acpi_ht = 1;
-+ acpi_disabled = 0;
-+ }
++static int do_microcode_update (void)
++{
++ int err;
++ dom0_op_t op;
+
-+ /* acpi=strict disables out-of-spec workarounds */
-+ else if (!memcmp(from, "acpi=strict", 11)) {
-+ acpi_strict = 1;
-+ }
++ err = sys_mlock((unsigned long)user_buffer, user_buffer_size);
++ if (err != 0)
++ return err;
+
-+ /* Limit ACPI just to boot-time to enable HT */
-+ else if (!memcmp(from, "acpi=ht", 7)) {
-+ if (!acpi_force)
-+ disable_acpi();
-+ acpi_ht = 1;
-+ }
-+
-+ /* "pci=noacpi" disable ACPI IRQ routing and PCI scan */
-+ else if (!memcmp(from, "pci=noacpi", 10)) {
-+ acpi_disable_pci();
-+ }
-+ /* "acpi=noirq" disables ACPI interrupt routing */
-+ else if (!memcmp(from, "acpi=noirq", 10)) {
-+ acpi_noirq_set();
-+ }
++ op.cmd = DOM0_MICROCODE;
++ set_xen_guest_handle(op.u.microcode.data, user_buffer);
++ op.u.microcode.length = user_buffer_size;
++ err = HYPERVISOR_dom0_op(&op);
+
-+ else if (!memcmp(from, "acpi_sci=edge", 13))
-+ acpi_sci_flags.trigger = 1;
++ (void)sys_munlock((unsigned long)user_buffer, user_buffer_size);
+
-+ else if (!memcmp(from, "acpi_sci=level", 14))
-+ acpi_sci_flags.trigger = 3;
++ return err;
++}
+
-+ else if (!memcmp(from, "acpi_sci=high", 13))
-+ acpi_sci_flags.polarity = 1;
++static ssize_t microcode_write (struct file *file, const char __user *buf, size_t len, loff_t *ppos)
++{
++ ssize_t ret;
+
-+ else if (!memcmp(from, "acpi_sci=low", 12))
-+ acpi_sci_flags.polarity = 3;
++ if (len < DEFAULT_UCODE_TOTALSIZE) {
++ printk(KERN_ERR "microcode: not enough data\n");
++ return -EINVAL;
++ }
+
-+#ifdef CONFIG_X86_IO_APIC
-+ else if (!memcmp(from, "acpi_skip_timer_override", 24))
-+ acpi_skip_timer_override = 1;
++ if ((len >> PAGE_SHIFT) > num_physpages) {
++ printk(KERN_ERR "microcode: too much data (max %ld pages)\n", num_physpages);
++ return -EINVAL;
++ }
+
-+ if (!memcmp(from, "disable_timer_pin_1", 19))
-+ disable_timer_pin_1 = 1;
-+ if (!memcmp(from, "enable_timer_pin_1", 18))
-+ disable_timer_pin_1 = -1;
-+
-+ /* disable IO-APIC */
-+ else if (!memcmp(from, "noapic", 6))
-+ disable_ioapic_setup();
-+#endif /* CONFIG_X86_IO_APIC */
-+#endif /* CONFIG_ACPI */
++ mutex_lock(µcode_mutex);
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ /* enable local APIC */
-+ else if (!memcmp(from, "lapic", 5))
-+ lapic_enable();
-+
-+ /* disable local APIC */
-+ else if (!memcmp(from, "nolapic", 6))
-+ lapic_disable();
-+#endif /* CONFIG_X86_LOCAL_APIC */
++ user_buffer = (void __user *) buf;
++ user_buffer_size = (int) len;
+
-+#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)) {
-+ 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;
-+ }
-+ }
-+#endif
-+#ifdef CONFIG_PROC_VMCORE
-+ /* elfcorehdr= specifies the location of elf core header
-+ * stored by the crashed kernel.
-+ */
-+ else if (!memcmp(from, "elfcorehdr=", 11))
-+ elfcorehdr_addr = memparse(from+11, &from);
-+#endif
++ ret = do_microcode_update();
++ if (!ret)
++ ret = (ssize_t)len;
+
-+ /*
-+ * highmem=size forces highmem to be exactly 'size' bytes.
-+ * This works even on boxes that have no highmem otherwise.
-+ * This also works to reduce highmem size on bigger boxes.
-+ */
-+ else if (!memcmp(from, "highmem=", 8))
-+ highmem_pages = memparse(from+8, &from) >> PAGE_SHIFT;
-+
-+ /*
-+ * vmalloc=size forces the vmalloc area to be exactly 'size'
-+ * bytes. This can be used to increase (or decrease) the
-+ * vmalloc area - the default is 128m.
-+ */
-+ else if (!memcmp(from, "vmalloc=", 8))
-+ __VMALLOC_RESERVE = memparse(from+8, &from);
++ mutex_unlock(µcode_mutex);
+
-+ next_char:
-+ c = *(from++);
-+ if (!c)
-+ break;
-+ if (COMMAND_LINE_SIZE <= ++len)
-+ break;
-+ *(to++) = c;
-+ }
-+ *to = '\0';
-+ *cmdline_p = command_line;
-+ if (userdef) {
-+ printk(KERN_INFO "user-defined physical RAM map:\n");
-+ print_memory_map("user");
-+ }
++ return ret;
+}
+
-+/*
-+ * Callback for efi_memory_walk.
-+ */
-+static int __init
-+efi_find_max_pfn(unsigned long start, unsigned long end, void *arg)
++static struct file_operations microcode_fops = {
++ .owner = THIS_MODULE,
++ .write = microcode_write,
++ .open = microcode_open,
++};
++
++static struct miscdevice microcode_dev = {
++ .minor = MICROCODE_MINOR,
++ .name = "microcode",
++ .fops = µcode_fops,
++};
++
++static int __init microcode_init (void)
+{
-+ unsigned long *max_pfn = arg, pfn;
++ int error;
+
-+ if (start < end) {
-+ pfn = PFN_UP(end -1);
-+ if (pfn > *max_pfn)
-+ *max_pfn = pfn;
++ error = misc_register(µcode_dev);
++ if (error) {
++ printk(KERN_ERR
++ "microcode: can't misc_register on minor=%d\n",
++ MICROCODE_MINOR);
++ return error;
+ }
-+ return 0;
-+}
+
-+static int __init
-+efi_memory_present_wrapper(unsigned long start, unsigned long end, void *arg)
-+{
-+ memory_present(0, start, end);
++ printk(KERN_INFO
++ "IA-32 Microcode Update Driver: v" MICROCODE_VERSION " <tigran at veritas.com>\n");
+ return 0;
+}
+
-+ /*
-+ * 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 s, unsigned long e, unsigned type)
++static void __exit microcode_exit (void)
+{
-+ u64 start = s;
-+ u64 end = e;
-+ int i;
-+ for (i = 0; i < e820.nr_map; i++) {
-+ struct e820entry *ei = &e820.map[i];
-+ 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;
++ misc_deregister(µcode_dev);
+}
+
++module_init(microcode_init)
++module_exit(microcode_exit)
++MODULE_ALIAS_MISCDEV(MICROCODE_MINOR);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/mpparse.c b/arch/i386/kernel/mpparse.c
+--- a/arch/i386/kernel/mpparse.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/mpparse.c 2007-03-14 10:55:14.000000000 +0100
+@@ -106,6 +106,7 @@
+
+ static void __cpuinit MP_processor_info (struct mpc_config_processor *m)
+ {
++#ifndef CONFIG_XEN
+ int ver, apicid;
+ physid_mask_t phys_cpu;
+
+@@ -196,8 +197,9 @@
+ }
+
+ cpu_set(num_processors, cpu_possible_map);
++#endif /* CONFIG_XEN */
+ num_processors++;
+-
++#ifndef CONFIG_XEN
+ /*
+ * Would be preferable to switch to bigsmp when CONFIG_HOTPLUG_CPU=y
+ * but we need to work other dependencies like SMP_SUSPEND etc
+@@ -218,6 +220,7 @@
+ }
+ }
+ bios_cpu_apicid[num_processors - 1] = m->mpc_apicid;
++#endif /* CONFIG_XEN */
+ }
+
+ static void __init MP_bus_info (struct mpc_config_bus *m)
+@@ -684,7 +687,11 @@
+ * Read the physical hardware table. Anything here will
+ * override the defaults.
+ */
++#ifdef CONFIG_XEN
++ if (!smp_read_mpc(isa_bus_to_virt(mpf->mpf_physptr))) {
++#else
+ if (!smp_read_mpc(phys_to_virt(mpf->mpf_physptr))) {
++#endif
+ 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");
+@@ -719,7 +726,11 @@
+
+ static int __init smp_scan_config (unsigned long base, unsigned long length)
+ {
++#ifdef CONFIG_XEN
++ unsigned long *bp = isa_bus_to_virt(base);
++#else
+ unsigned long *bp = phys_to_virt(base);
++#endif
+ struct intel_mp_floating *mpf;
+
+ Dprintk("Scan SMP from %p for %ld bytes.\n", bp,length);
+@@ -735,6 +746,7 @@
+ || (mpf->mpf_specification == 4)) ) {
+
+ smp_found_config = 1;
++#ifndef CONFIG_XEN
+ printk(KERN_INFO "found SMP MP-table at %08lx\n",
+ virt_to_phys(mpf));
+ reserve_bootmem(virt_to_phys(mpf), PAGE_SIZE);
+@@ -754,6 +766,10 @@
+ size = end - mpf->mpf_physptr;
+ reserve_bootmem(mpf->mpf_physptr, size);
+ }
++#else
++ printk(KERN_INFO "found SMP MP-table at %08lx\n",
++ ((unsigned long)bp - (unsigned long)isa_bus_to_virt(base)) + base);
++#endif
+
+ mpf_found = mpf;
+ return 1;
+@@ -766,7 +782,9 @@
+
+ void __init find_smp_config (void)
+ {
++#ifndef CONFIG_XEN
+ unsigned int address;
++#endif
+
+ /*
+ * FIXME: Linux assumes you have 640K of base ram..
+@@ -797,9 +815,11 @@
+ * MP1.4 SPEC states to only scan first 1K of 4K EBDA.
+ */
+
++#ifndef CONFIG_XEN
+ address = get_bios_ebda();
+ if (address)
+ smp_scan_config(address, 0x400);
++#endif
+ }
+
+ int es7000_plat;
+@@ -812,6 +832,7 @@
+
+ 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);
+@@ -820,6 +841,7 @@
+ boot_cpu_physical_apicid = 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)
+@@ -836,6 +858,7 @@
+ 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));
+@@ -846,6 +869,7 @@
+ processor.mpc_featureflag = boot_cpu_data.x86_capability[0];
+ processor.mpc_reserved[0] = 0;
+ processor.mpc_reserved[1] = 0;
++#endif
+
+ MP_processor_info(&processor);
+ }
+@@ -900,7 +924,9 @@
+ 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
+ if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL)
+ && !APIC_XAPIC(apic_version[boot_cpu_physical_apicid]))
+ tmpid = io_apic_get_unique_id(idx, id);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/pci-dma-xen.c b/arch/i386/kernel/pci-dma-xen.c
+--- a/arch/i386/kernel/pci-dma-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/pci-dma-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,374 @@
+/*
-+ * Find the highest page frame number we have available
++ * Dynamic DMA mapping support.
++ *
++ * On i386 there is no hardware dynamic DMA address translation,
++ * so consistent alloc/free are merely page allocation/freeing.
++ * The rest of the dynamic DMA mapping interface is implemented
++ * in asm/pci.h.
+ */
-+void __init find_max_pfn(void)
-+{
-+ int i;
+
-+ max_pfn = 0;
-+ if (efi_enabled) {
-+ efi_memmap_walk(efi_find_max_pfn, &max_pfn);
-+ efi_memmap_walk(efi_memory_present_wrapper, NULL);
-+ return;
-+ }
++#include <linux/types.h>
++#include <linux/mm.h>
++#include <linux/string.h>
++#include <linux/pci.h>
++#include <linux/module.h>
++#include <linux/version.h>
++#include <asm/io.h>
++#include <xen/balloon.h>
++#include <asm/tlbflush.h>
++#include <asm-i386/mach-xen/asm/swiotlb.h>
++#include <asm/bug.h>
+
-+ for (i = 0; i < e820.nr_map; i++) {
-+ unsigned long start, end;
-+ /* RAM? */
-+ if (e820.map[i].type != E820_RAM)
-+ continue;
-+ start = PFN_UP(e820.map[i].addr);
-+ end = PFN_DOWN(e820.map[i].addr + e820.map[i].size);
-+ if (start >= end)
-+ continue;
-+ if (end > max_pfn)
-+ max_pfn = end;
-+ memory_present(0, start, end);
-+ }
-+}
++#ifdef __x86_64__
++int iommu_merge __read_mostly = 0;
++EXPORT_SYMBOL(iommu_merge);
+
-+/*
-+ * Determine low and high memory ranges:
-+ */
-+unsigned long __init find_max_low_pfn(void)
-+{
-+ unsigned long max_low_pfn;
++dma_addr_t bad_dma_address __read_mostly;
++EXPORT_SYMBOL(bad_dma_address);
+
-+ max_low_pfn = max_pfn;
-+ if (max_low_pfn > MAXMEM_PFN) {
-+ if (highmem_pages == -1)
-+ highmem_pages = max_pfn - MAXMEM_PFN;
-+ if (highmem_pages + MAXMEM_PFN < max_pfn)
-+ max_pfn = MAXMEM_PFN + highmem_pages;
-+ if (highmem_pages + MAXMEM_PFN > max_pfn) {
-+ printk("only %luMB highmem pages available, ignoring highmem size of %uMB.\n", pages_to_mb(max_pfn - MAXMEM_PFN), pages_to_mb(highmem_pages));
-+ highmem_pages = 0;
-+ }
-+ max_low_pfn = MAXMEM_PFN;
-+#ifndef CONFIG_HIGHMEM
-+ /* Maximum memory usable is what is directly addressable */
-+ printk(KERN_WARNING "Warning only %ldMB will be used.\n",
-+ MAXMEM>>20);
-+ if (max_pfn > MAX_NONPAE_PFN)
-+ printk(KERN_WARNING "Use a PAE enabled kernel.\n");
-+ else
-+ printk(KERN_WARNING "Use a HIGHMEM enabled kernel.\n");
-+ max_pfn = MAXMEM_PFN;
-+#else /* !CONFIG_HIGHMEM */
-+#ifndef CONFIG_X86_PAE
-+ if (max_pfn > MAX_NONPAE_PFN) {
-+ max_pfn = MAX_NONPAE_PFN;
-+ printk(KERN_WARNING "Warning only 4GB will be used.\n");
-+ printk(KERN_WARNING "Use a PAE enabled kernel.\n");
-+ }
-+#endif /* !CONFIG_X86_PAE */
-+#endif /* !CONFIG_HIGHMEM */
-+ } else {
-+ if (highmem_pages == -1)
-+ highmem_pages = 0;
-+#ifdef CONFIG_HIGHMEM
-+ if (highmem_pages >= max_pfn) {
-+ printk(KERN_ERR "highmem size specified (%uMB) is bigger than pages available (%luMB)!.\n", pages_to_mb(highmem_pages), pages_to_mb(max_pfn));
-+ highmem_pages = 0;
-+ }
-+ if (highmem_pages) {
-+ if (max_low_pfn-highmem_pages < 64*1024*1024/PAGE_SIZE){
-+ printk(KERN_ERR "highmem size %uMB results in smaller than 64MB lowmem, ignoring it.\n", pages_to_mb(highmem_pages));
-+ highmem_pages = 0;
-+ }
-+ max_low_pfn -= highmem_pages;
-+ }
++/* This tells the BIO block layer to assume merging. Default to off
++ because we cannot guarantee merging later. */
++int iommu_bio_merge __read_mostly = 0;
++EXPORT_SYMBOL(iommu_bio_merge);
++
++int iommu_sac_force __read_mostly = 0;
++EXPORT_SYMBOL(iommu_sac_force);
++
++int no_iommu __read_mostly;
++#ifdef CONFIG_IOMMU_DEBUG
++int panic_on_overflow __read_mostly = 1;
++int force_iommu __read_mostly = 1;
+#else
-+ if (highmem_pages)
-+ printk(KERN_ERR "ignoring highmem size on non-highmem kernel!\n");
++int panic_on_overflow __read_mostly = 0;
++int force_iommu __read_mostly= 0;
+#endif
-+ }
-+ return max_low_pfn;
-+}
+
-+/*
-+ * Free all available memory for boot time allocation. Used
-+ * as a callback function by efi_memory_walk()
-+ */
++/* Set this to 1 if there is a HW IOMMU in the system */
++int iommu_detected __read_mostly = 0;
+
-+static int __init
-+free_available_memory(unsigned long start, unsigned long end, void *arg)
++void __init pci_iommu_alloc(void)
+{
-+ /* check max_low_pfn */
-+ if (start >= (max_low_pfn << PAGE_SHIFT))
-+ return 0;
-+ if (end >= (max_low_pfn << PAGE_SHIFT))
-+ end = max_low_pfn << PAGE_SHIFT;
-+ if (start < end)
-+ free_bootmem(start, end - start);
++ /*
++ * The order of these functions is important for
++ * fall-back/fail-over reasons
++ */
++#ifdef CONFIG_IOMMU
++ iommu_hole_init();
++#endif
+
-+ return 0;
-+}
-+/*
-+ * Register fully available low RAM pages with the bootmem allocator.
-+ */
-+static void __init register_bootmem_low_pages(unsigned long max_low_pfn)
-+{
-+ int i;
++#ifdef CONFIG_CALGARY_IOMMU
++#include <asm/calgary.h>
++ /* shut up compiler */
++ use_calgary = use_calgary;
++ detect_calgary();
++#endif
+
-+ if (efi_enabled) {
-+ efi_memmap_walk(free_available_memory, NULL);
-+ return;
-+ }
-+ for (i = 0; i < e820.nr_map; i++) {
-+ unsigned long curr_pfn, last_pfn, size;
-+ /*
-+ * Reserve usable low memory
-+ */
-+ if (e820.map[i].type != E820_RAM)
-+ continue;
-+ /*
-+ * We are rounding up the start address of usable memory:
-+ */
-+ curr_pfn = PFN_UP(e820.map[i].addr);
-+ if (curr_pfn >= max_low_pfn)
-+ continue;
-+ /*
-+ * ... and at the end of the usable range downwards:
-+ */
-+ last_pfn = PFN_DOWN(e820.map[i].addr + e820.map[i].size);
++#ifdef CONFIG_SWIOTLB
++ pci_swiotlb_init();
++#endif
++}
+
-+#ifdef CONFIG_XEN
-+ /*
-+ * Truncate to the number of actual pages currently
-+ * present.
-+ */
-+ if (last_pfn > xen_start_info->nr_pages)
-+ last_pfn = xen_start_info->nr_pages;
+#endif
+
-+ if (last_pfn > max_low_pfn)
-+ last_pfn = max_low_pfn;
++struct dma_coherent_mem {
++ void *virt_base;
++ u32 device_base;
++ int size;
++ int flags;
++ unsigned long *bitmap;
++};
++
++#define IOMMU_BUG_ON(test) \
++do { \
++ if (unlikely(test)) { \
++ printk(KERN_ALERT "Fatal DMA error! " \
++ "Please use 'swiotlb=force'\n"); \
++ BUG(); \
++ } \
++} while (0)
+
-+ /*
-+ * .. finally, did all the rounding and playing
-+ * around just make the area go away?
-+ */
-+ if (last_pfn <= curr_pfn)
-+ continue;
++int
++dma_map_sg(struct device *hwdev, struct scatterlist *sg, int nents,
++ enum dma_data_direction direction)
++{
++ int i, rc;
++
++ BUG_ON(!valid_dma_direction(direction));
++ WARN_ON(nents == 0 || sg[0].length == 0);
+
-+ size = last_pfn - curr_pfn;
-+ free_bootmem(PFN_PHYS(curr_pfn), PFN_PHYS(size));
++ if (swiotlb) {
++ rc = swiotlb_map_sg(hwdev, sg, nents, direction);
++ } else {
++ for (i = 0; i < nents; i++ ) {
++ sg[i].dma_address =
++ page_to_bus(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));
++ }
++ rc = nents;
+ }
++
++ flush_write_buffers();
++ return rc;
+}
++EXPORT_SYMBOL(dma_map_sg);
+
-+#ifndef CONFIG_XEN
-+/*
-+ * workaround for Dell systems that neglect to reserve EBDA
-+ */
-+static void __init reserve_ebda_region(void)
++void
++dma_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nents,
++ enum dma_data_direction direction)
+{
-+ unsigned int addr;
-+ addr = get_bios_ebda();
-+ if (addr)
-+ reserve_bootmem(addr, PAGE_SIZE);
++ BUG_ON(!valid_dma_direction(direction));
++ if (swiotlb)
++ swiotlb_unmap_sg(hwdev, sg, nents, direction);
+}
-+#endif
++EXPORT_SYMBOL(dma_unmap_sg);
+
-+#ifndef CONFIG_NEED_MULTIPLE_NODES
-+void __init setup_bootmem_allocator(void);
-+static unsigned long __init setup_memory(void)
++/*
++ * XXX This file is also used by xenLinux/ia64.
++ * "defined(__i386__) || defined (__x86_64__)" means "!defined(__ia64__)".
++ * This #if work around should be removed once this file is merbed back into
++ * i386' pci-dma or is moved to drivers/xen/core.
++ */
++#if defined(__i386__) || defined(__x86_64__)
++dma_addr_t
++dma_map_page(struct device *dev, struct page *page, unsigned long offset,
++ size_t size, enum dma_data_direction direction)
+{
-+ /*
-+ * partially used pages are not usable - thus
-+ * we are rounding upwards:
-+ */
-+ min_low_pfn = PFN_UP(__pa(xen_start_info->pt_base)) +
-+ xen_start_info->nr_pt_frames;
-+
-+ find_max_pfn();
-+
-+ max_low_pfn = find_max_low_pfn();
++ dma_addr_t dma_addr;
+
-+#ifdef CONFIG_HIGHMEM
-+ highstart_pfn = highend_pfn = max_pfn;
-+ if (max_pfn > max_low_pfn) {
-+ highstart_pfn = max_low_pfn;
-+ }
-+ printk(KERN_NOTICE "%ldMB HIGHMEM available.\n",
-+ pages_to_mb(highend_pfn - highstart_pfn));
-+#endif
-+ printk(KERN_NOTICE "%ldMB LOWMEM available.\n",
-+ pages_to_mb(max_low_pfn));
++ BUG_ON(!valid_dma_direction(direction));
+
-+ setup_bootmem_allocator();
++ if (swiotlb) {
++ dma_addr = swiotlb_map_page(
++ dev, page, offset, size, direction);
++ } else {
++ dma_addr = page_to_bus(page) + offset;
++ IOMMU_BUG_ON(address_needs_mapping(dev, dma_addr));
++ }
+
-+ return max_low_pfn;
++ return dma_addr;
+}
++EXPORT_SYMBOL(dma_map_page);
+
-+void __init zone_sizes_init(void)
++void
++dma_unmap_page(struct device *dev, dma_addr_t dma_address, size_t size,
++ enum dma_data_direction direction)
+{
-+ 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;
-+ low = max_low_pfn;
-+
-+ if (low < max_dma)
-+ zones_size[ZONE_DMA] = low;
-+ else {
-+ zones_size[ZONE_DMA] = max_dma;
-+ zones_size[ZONE_NORMAL] = low - max_dma;
-+#ifdef CONFIG_HIGHMEM
-+ zones_size[ZONE_HIGHMEM] = highend_pfn - low;
-+#endif
-+ }
-+ free_area_init(zones_size);
++ BUG_ON(!valid_dma_direction(direction));
++ if (swiotlb)
++ swiotlb_unmap_page(dev, dma_address, size, direction);
+}
-+#else
-+extern unsigned long __init setup_memory(void);
-+extern void zone_sizes_init(void);
-+#endif /* !CONFIG_NEED_MULTIPLE_NODES */
++EXPORT_SYMBOL(dma_unmap_page);
++#endif /* defined(__i386__) || defined(__x86_64__) */
+
-+void __init setup_bootmem_allocator(void)
++int
++dma_mapping_error(dma_addr_t dma_addr)
+{
-+ unsigned long bootmap_size;
-+ /*
-+ * Initialize the boot-time allocator (with low memory only):
-+ */
-+ bootmap_size = init_bootmem(min_low_pfn, max_low_pfn);
-+
-+ register_bootmem_low_pages(max_low_pfn);
++ if (swiotlb)
++ return swiotlb_dma_mapping_error(dma_addr);
++ return 0;
++}
++EXPORT_SYMBOL(dma_mapping_error);
+
++int
++dma_supported(struct device *dev, u64 mask)
++{
++ if (swiotlb)
++ return swiotlb_dma_supported(dev, mask);
+ /*
-+ * Reserve the bootmem bitmap itself as well. We do this in two
-+ * steps (first step was init_bootmem()) because this catches
-+ * the (very unlikely) case of us accidentally initializing the
-+ * bootmem allocator with an invalid RAM area.
++ * By default we'll BUG when an infeasible DMA is requested, and
++ * request swiotlb=force (see IOMMU_BUG_ON).
+ */
-+ reserve_bootmem(__PHYSICAL_START, (PFN_PHYS(min_low_pfn) +
-+ bootmap_size + PAGE_SIZE-1) - (__PHYSICAL_START));
++ return 1;
++}
++EXPORT_SYMBOL(dma_supported);
+
-+#ifndef CONFIG_XEN
-+ /*
-+ * reserve physical page 0 - it's a special BIOS page on many boxes,
-+ * enabling clean reboots, SMP operation, laptop functions.
-+ */
-+ reserve_bootmem(0, PAGE_SIZE);
++void *dma_alloc_coherent(struct device *dev, size_t size,
++ dma_addr_t *dma_handle, gfp_t gfp)
++{
++ void *ret;
++ struct dma_coherent_mem *mem = dev ? dev->dma_mem : NULL;
++ unsigned int order = get_order(size);
++ unsigned long vstart;
++ /* ignore region specifiers */
++ gfp &= ~(__GFP_DMA | __GFP_HIGHMEM);
+
-+ /* reserve EBDA region, it's a 4K region */
-+ reserve_ebda_region();
++ if (mem) {
++ int page = bitmap_find_free_region(mem->bitmap, mem->size,
++ order);
++ if (page >= 0) {
++ *dma_handle = mem->device_base + (page << PAGE_SHIFT);
++ ret = mem->virt_base + (page << PAGE_SHIFT);
++ memset(ret, 0, size);
++ return ret;
++ }
++ if (mem->flags & DMA_MEMORY_EXCLUSIVE)
++ return NULL;
++ }
+
-+ /* could be an AMD 768MPX chipset. Reserve a page before VGA to prevent
-+ PCI prefetch into it (errata #56). Usually the page is reserved anyways,
-+ unless you have no PS/2 mouse plugged in. */
-+ if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD &&
-+ boot_cpu_data.x86 == 6)
-+ reserve_bootmem(0xa0000 - 4096, 4096);
++ if (dev == NULL || (dev->coherent_dma_mask < 0xffffffff))
++ gfp |= GFP_DMA;
+
-+#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(PAGE_SIZE, PAGE_SIZE);
-+#endif
-+#ifdef CONFIG_ACPI_SLEEP
-+ /*
-+ * Reserve low memory region for sleep support.
-+ */
-+ acpi_reserve_bootmem();
-+#endif
-+#endif /* !CONFIG_XEN */
++ vstart = __get_free_pages(gfp, order);
++ ret = (void *)vstart;
+
-+#ifdef CONFIG_BLK_DEV_INITRD
-+ if (xen_start_info->mod_start) {
-+ if (INITRD_START + INITRD_SIZE <= (max_low_pfn << PAGE_SHIFT)) {
-+ /*reserve_bootmem(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",
-+ INITRD_START + INITRD_SIZE,
-+ max_low_pfn << PAGE_SHIFT);
-+ initrd_start = 0;
++ if (ret != NULL) {
++ /* NB. Hardcode 31 address bits for now: aacraid limitation. */
++ if (xen_create_contiguous_region(vstart, order, 31) != 0) {
++ free_pages(vstart, order);
++ return NULL;
+ }
++ memset(ret, 0, size);
++ *dma_handle = virt_to_bus(ret);
+ }
-+#endif
-+#ifdef CONFIG_KEXEC
-+ if (crashk_res.start != crashk_res.end)
-+ reserve_bootmem(crashk_res.start,
-+ crashk_res.end - crashk_res.start + 1);
-+#endif
-+
-+ if (!xen_feature(XENFEAT_auto_translated_physmap))
-+ phys_to_machine_mapping =
-+ (unsigned long *)xen_start_info->mfn_list;
++ return ret;
+}
++EXPORT_SYMBOL(dma_alloc_coherent);
+
-+/*
-+ * The node 0 pgdat is initialized before all of these because
-+ * it's needed for bootmem. node>0 pgdats have their virtual
-+ * space allocated before the pagetables are in place to access
-+ * them, so they can't be cleared then.
-+ *
-+ * This should all compile down to nothing when NUMA is off.
-+ */
-+void __init remapped_pgdat_init(void)
++void dma_free_coherent(struct device *dev, size_t size,
++ void *vaddr, dma_addr_t dma_handle)
+{
-+ int nid;
++ struct dma_coherent_mem *mem = dev ? dev->dma_mem : NULL;
++ int order = get_order(size);
++
++ if (mem && vaddr >= mem->virt_base && vaddr < (mem->virt_base + (mem->size << PAGE_SHIFT))) {
++ int page = (vaddr - mem->virt_base) >> PAGE_SHIFT;
+
-+ for_each_online_node(nid) {
-+ if (nid != 0)
-+ memset(NODE_DATA(nid), 0, sizeof(struct pglist_data));
++ bitmap_release_region(mem->bitmap, page, order);
++ } else {
++ xen_destroy_contiguous_region((unsigned long)vaddr, order);
++ free_pages((unsigned long)vaddr, order);
+ }
+}
++EXPORT_SYMBOL(dma_free_coherent);
+
-+/*
-+ * Request address space for all standard RAM and ROM resources
-+ * and also for regions reported as reserved by the e820.
-+ */
-+static void __init
-+legacy_init_iomem_resources(struct resource *code_resource, struct resource *data_resource)
++#ifdef ARCH_HAS_DMA_DECLARE_COHERENT_MEMORY
++int dma_declare_coherent_memory(struct device *dev, dma_addr_t bus_addr,
++ dma_addr_t device_addr, size_t size, int flags)
+{
-+ int i;
-+ struct e820entry *map = e820.map;
-+ int nr_map = e820.nr_map;
-+#ifdef CONFIG_XEN_PRIVILEGED_GUEST
-+ struct xen_memory_map memmap;
-+
-+ map = machine_e820.map;
-+ memmap.nr_entries = E820MAX;
-+
-+ set_xen_guest_handle(memmap.buffer, map);
++ void __iomem *mem_base = NULL;
++ int pages = size >> PAGE_SHIFT;
++ int bitmap_size = (pages + 31)/32;
+
-+ if(HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap))
-+ BUG();
-+ machine_e820.nr_map = memmap.nr_entries;
-+ nr_map = memmap.nr_entries;
-+ e820_setup_gap(map, memmap.nr_entries);
-+#endif
++ if ((flags & (DMA_MEMORY_MAP | DMA_MEMORY_IO)) == 0)
++ goto out;
++ if (!size)
++ goto out;
++ if (dev->dma_mem)
++ goto out;
+
-+ probe_roms();
++ /* FIXME: this routine just ignores DMA_MEMORY_INCLUDES_CHILDREN */
+
-+ for (i = 0; i < nr_map; i++) {
-+ struct resource *res;
-+ if (map[i].addr + map[i].size > 0x100000000ULL)
-+ continue;
-+ res = kzalloc(sizeof(struct resource), GFP_ATOMIC);
-+ switch (map[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 = map[i].addr;
-+ res->end = res->start + map[i].size - 1;
-+ res->flags = IORESOURCE_MEM | IORESOURCE_BUSY;
-+ if (request_resource(&iomem_resource, res)) {
-+ kfree(res);
-+ continue;
-+ }
-+ if (map[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
-+ request_resource(res, &crashk_res);
-+#endif
-+ }
-+ }
-+}
++ mem_base = ioremap(bus_addr, size);
++ if (!mem_base)
++ goto out;
+
-+/*
-+ * Request address space for all standard resources
-+ *
-+ * This is called just before pcibios_init(), which is also a
-+ * subsys_initcall, but is linked in later (in arch/i386/pci/common.c).
-+ */
-+static int __init request_standard_resources(void)
-+{
-+ int i;
++ dev->dma_mem = kzalloc(sizeof(struct dma_coherent_mem), GFP_KERNEL);
++ if (!dev->dma_mem)
++ goto out;
++ dev->dma_mem->bitmap = kzalloc(bitmap_size, GFP_KERNEL);
++ if (!dev->dma_mem->bitmap)
++ goto free1_out;
+
-+ /* Nothing to do if not running in dom0. */
-+ if (!is_initial_xendomain())
-+ return 0;
++ dev->dma_mem->virt_base = mem_base;
++ dev->dma_mem->device_base = device_addr;
++ dev->dma_mem->size = pages;
++ dev->dma_mem->flags = flags;
+
-+ printk("Setting up standard PCI resources\n");
-+ if (efi_enabled)
-+ efi_initialize_iomem_resources(&code_resource, &data_resource);
-+ else
-+ legacy_init_iomem_resources(&code_resource, &data_resource);
++ if (flags & DMA_MEMORY_MAP)
++ return DMA_MEMORY_MAP;
+
-+ /* EFI systems may still have VGA */
-+ request_resource(&iomem_resource, &video_ram_resource);
++ return DMA_MEMORY_IO;
+
-+ /* 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]);
++ free1_out:
++ kfree(dev->dma_mem->bitmap);
++ out:
++ if (mem_base)
++ iounmap(mem_base);
+ return 0;
+}
++EXPORT_SYMBOL(dma_declare_coherent_memory);
+
-+subsys_initcall(request_standard_resources);
++void dma_release_declared_memory(struct device *dev)
++{
++ struct dma_coherent_mem *mem = dev->dma_mem;
++
++ if(!mem)
++ return;
++ dev->dma_mem = NULL;
++ iounmap(mem->virt_base);
++ kfree(mem->bitmap);
++ kfree(mem);
++}
++EXPORT_SYMBOL(dma_release_declared_memory);
+
-+/*
-+ * Locate a unused range of the physical address space below 4G which
-+ * can be used for PCI mappings.
-+ */
-+static void __init
-+e820_setup_gap(struct e820entry *e820, int nr_map)
++void *dma_mark_declared_memory_occupied(struct device *dev,
++ dma_addr_t device_addr, size_t size)
+{
-+ unsigned long gapstart, gapsize, round;
-+ unsigned long long last;
-+ int i;
++ struct dma_coherent_mem *mem = dev->dma_mem;
++ int pages = (size + (device_addr & ~PAGE_MASK) + PAGE_SIZE - 1) >> PAGE_SHIFT;
++ int pos, err;
+
-+ /*
-+ * Search for the bigest gap in the low 32 bits of the e820
-+ * memory space.
-+ */
-+ 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;
++ if (!mem)
++ return ERR_PTR(-EINVAL);
+
-+ /*
-+ * 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;
++ pos = (device_addr - mem->device_base) >> PAGE_SHIFT;
++ err = bitmap_allocate_region(mem->bitmap, pos, get_order(pages));
++ if (err != 0)
++ return ERR_PTR(err);
++ return mem->virt_base + (pos << PAGE_SHIFT);
++}
++EXPORT_SYMBOL(dma_mark_declared_memory_occupied);
++#endif /* ARCH_HAS_DMA_DECLARE_COHERENT_MEMORY */
+
-+ if (gap > gapsize) {
-+ gapsize = gap;
-+ gapstart = end;
-+ }
-+ }
-+ if (start < last)
-+ last = start;
-+ }
++dma_addr_t
++dma_map_single(struct device *dev, void *ptr, size_t size,
++ enum dma_data_direction direction)
++{
++ dma_addr_t dma;
+
-+ /*
-+ * 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;
++ BUG_ON(!valid_dma_direction(direction));
++ WARN_ON(size == 0);
+
-+ printk("Allocating PCI resources starting at %08lx (gap: %08lx:%08lx)\n",
-+ pci_mem_start, gapstart, gapsize);
++ if (swiotlb) {
++ dma = swiotlb_map_single(dev, ptr, size, direction);
++ } else {
++ dma = virt_to_bus(ptr);
++ IOMMU_BUG_ON(range_straddles_page_boundary(ptr, size));
++ IOMMU_BUG_ON(address_needs_mapping(dev, dma));
++ }
++
++ flush_write_buffers();
++ return dma;
+}
++EXPORT_SYMBOL(dma_map_single);
+
-+static void __init register_memory(void)
++void
++dma_unmap_single(struct device *dev, dma_addr_t dma_addr, size_t size,
++ enum dma_data_direction direction)
+{
-+#ifndef CONFIG_XEN
-+ e820_setup_gap(e820.map, e820.nr_map);
-+#endif
++ BUG_ON(!valid_dma_direction(direction));
++ if (swiotlb)
++ swiotlb_unmap_single(dev, dma_addr, size, direction);
+}
++EXPORT_SYMBOL(dma_unmap_single);
+
-+#ifdef CONFIG_MCA
-+static void set_mca_bus(int x)
++void
++dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle, size_t size,
++ enum dma_data_direction direction)
+{
-+ MCA_bus = x;
++ if (swiotlb)
++ swiotlb_sync_single_for_cpu(dev, dma_handle, size, direction);
+}
-+#else
-+static void set_mca_bus(int x) { }
-+#endif
++EXPORT_SYMBOL(dma_sync_single_for_cpu);
+
++void
++dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle, size_t size,
++ enum dma_data_direction direction)
++{
++ if (swiotlb)
++ swiotlb_sync_single_for_device(dev, dma_handle, size, direction);
++}
++EXPORT_SYMBOL(dma_sync_single_for_device);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/process-xen.c b/arch/i386/kernel/process-xen.c
+--- a/arch/i386/kernel/process-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/process-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,851 @@
+/*
-+ * Determine if we were loaded by an EFI loader. If so, then we have also been
-+ * passed the efi memmap, systab, etc., so we should use these data structures
-+ * for initialization. Note, the efi init code path is determined by the
-+ * global efi_enabled. This allows the same kernel image to be used on existing
-+ * systems (with a traditional BIOS) as well as on EFI systems.
++ * linux/arch/i386/kernel/process.c
++ *
++ * Copyright (C) 1995 Linus Torvalds
++ *
++ * Pentium III FXSR, SSE support
++ * Gareth Hughes <gareth at valinux.com>, May 2000
+ */
-+void __init setup_arch(char **cmdline_p)
-+{
-+ int i, j, k, fpp;
-+ struct physdev_set_iopl set_iopl;
-+ unsigned long max_low_pfn;
-+
-+ /* Force a quick death if the kernel panics (not domain 0). */
-+ extern int panic_timeout;
-+ if (!panic_timeout && !is_initial_xendomain())
-+ panic_timeout = 1;
+
-+ /* Register a call for panic conditions. */
-+ atomic_notifier_chain_register(&panic_notifier_list, &xen_panic_block);
++/*
++ * This file handles the architecture-dependent parts of process handling..
++ */
+
-+ HYPERVISOR_vm_assist(VMASST_CMD_enable, VMASST_TYPE_4gb_segments);
-+ HYPERVISOR_vm_assist(VMASST_CMD_enable,
-+ VMASST_TYPE_writable_pagetables);
++#include <stdarg.h>
+
-+ memcpy(&boot_cpu_data, &new_cpu_data, sizeof(new_cpu_data));
-+ pre_setup_arch_hook();
-+ early_cpu_init();
++#include <linux/cpu.h>
++#include <linux/errno.h>
++#include <linux/sched.h>
++#include <linux/fs.h>
++#include <linux/kernel.h>
++#include <linux/mm.h>
++#include <linux/elfcore.h>
++#include <linux/smp.h>
++#include <linux/smp_lock.h>
++#include <linux/stddef.h>
++#include <linux/slab.h>
++#include <linux/vmalloc.h>
++#include <linux/user.h>
++#include <linux/a.out.h>
++#include <linux/interrupt.h>
++#include <linux/utsname.h>
++#include <linux/delay.h>
++#include <linux/reboot.h>
++#include <linux/init.h>
++#include <linux/mc146818rtc.h>
++#include <linux/module.h>
++#include <linux/kallsyms.h>
++#include <linux/ptrace.h>
++#include <linux/random.h>
++#include <linux/personality.h>
+
-+ /*
-+ * FIXME: This isn't an official loader_type right
-+ * now but does currently work with elilo.
-+ * If we were configured as an EFI kernel, check to make
-+ * sure that we were loaded correctly from elilo and that
-+ * the system table is valid. If not, then initialize normally.
-+ */
-+#ifdef CONFIG_EFI
-+ if ((LOADER_TYPE == 0x50) && EFI_SYSTAB)
-+ efi_enabled = 1;
++#include <asm/uaccess.h>
++#include <asm/pgtable.h>
++#include <asm/system.h>
++#include <asm/io.h>
++#include <asm/ldt.h>
++#include <asm/processor.h>
++#include <asm/i387.h>
++#include <asm/desc.h>
++#include <asm/vm86.h>
++#ifdef CONFIG_MATH_EMULATION
++#include <asm/math_emu.h>
+#endif
+
-+ /* This must be initialized to UNNAMED_MAJOR for ipconfig to work
-+ properly. Setting ROOT_DEV to default to /dev/ram0 breaks initrd.
-+ */
-+ ROOT_DEV = MKDEV(UNNAMED_MAJOR,0);
-+ drive_info = DRIVE_INFO;
-+ screen_info = SCREEN_INFO;
-+ edid_info = EDID_INFO;
-+ apm_info.bios = APM_BIOS_INFO;
-+ ist_info = IST_INFO;
-+ saved_videomode = VIDEO_MODE;
-+ if( SYS_DESC_TABLE.length != 0 ) {
-+ set_mca_bus(SYS_DESC_TABLE.table[3] & 0x2);
-+ machine_id = SYS_DESC_TABLE.table[0];
-+ machine_submodel_id = SYS_DESC_TABLE.table[1];
-+ BIOS_revision = SYS_DESC_TABLE.table[2];
-+ }
-+ 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);
-+ }
-+ xen_start_info->console.domU.mfn = 0;
-+ xen_start_info->console.domU.evtchn = 0;
-+ } else
-+ screen_info.orig_video_isVGA = 0;
++#include <xen/interface/physdev.h>
++#include <xen/interface/vcpu.h>
++#include <xen/cpu_hotplug.h>
+
-+#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
++#include <linux/err.h>
+
-+ setup_xen_features();
++#include <asm/tlbflush.h>
++#include <asm/cpu.h>
++#include <asm/pda.h>
+
-+ ARCH_SETUP
-+ if (efi_enabled)
-+ efi_init();
-+ else {
-+ printk(KERN_INFO "BIOS-provided physical RAM map:\n");
-+ print_memory_map(machine_specific_memory_setup());
-+ }
++asmlinkage void ret_from_fork(void) __asm__("ret_from_fork");
+
-+ copy_edd();
++static int hlt_counter;
+
-+ 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 = (PFN_UP(__pa(xen_start_info->pt_base)) +
-+ xen_start_info->nr_pt_frames) << PAGE_SHIFT;
++unsigned long boot_option_idle_override = 0;
++EXPORT_SYMBOL(boot_option_idle_override);
+
-+ 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;
++/*
++ * Return saved PC of a blocked thread.
++ */
++unsigned long thread_saved_pc(struct task_struct *tsk)
++{
++ return ((unsigned long *)tsk->thread.esp)[3];
++}
+
-+ parse_cmdline_early(cmdline_p);
++/*
++ * Powermanagement idle function, if any..
++ */
++void (*pm_idle)(void);
++EXPORT_SYMBOL(pm_idle);
++static DEFINE_PER_CPU(unsigned int, cpu_idle_state);
+
-+#ifdef CONFIG_EARLY_PRINTK
-+ {
-+ char *s = strstr(*cmdline_p, "earlyprintk=");
-+ if (s) {
-+ setup_early_printk(strchr(s, '=') + 1);
-+ printk("early console enabled\n");
-+ }
-+ }
-+#endif
++void disable_hlt(void)
++{
++ hlt_counter++;
++}
+
-+ max_low_pfn = setup_memory();
++EXPORT_SYMBOL(disable_hlt);
+
-+ /*
-+ * NOTE: before this point _nobody_ is allowed to allocate
-+ * any memory using the bootmem allocator. Although the
-+ * alloctor is now initialised only the first 8Mb of the kernel
-+ * virtual address space has been mapped. All allocations before
-+ * paging_init() has completed must use the alloc_bootmem_low_pages()
-+ * variant (which allocates DMA'able memory) and care must be taken
-+ * not to exceed the 8Mb limit.
-+ */
++void enable_hlt(void)
++{
++ hlt_counter--;
++}
+
-+#ifdef CONFIG_SMP
-+ smp_alloc_memory(); /* AP processor realmode stacks in low memory*/
-+#endif
-+ paging_init();
-+ remapped_pgdat_init();
-+ sparse_init();
-+ zone_sizes_init();
++EXPORT_SYMBOL(enable_hlt);
+
-+#ifdef CONFIG_X86_FIND_SMP_CONFIG
++/* XXX XEN doesn't use default_idle(), poll_idle(). Use xen_idle() instead. */
++void xen_idle(void)
++{
++ current_thread_info()->status &= ~TS_POLLING;
+ /*
-+ * Find and reserve possible boot-time SMP configuration:
++ * TS_POLLING-cleared state must be visible before we
++ * test NEED_RESCHED:
+ */
-+ find_smp_config();
++ smp_mb();
++
++ local_irq_disable();
++ if (!need_resched())
++ safe_halt(); /* enables interrupts racelessly */
++ else
++ local_irq_enable();
++ current_thread_info()->status |= TS_POLLING;
++}
++#ifdef CONFIG_APM_MODULE
++EXPORT_SYMBOL(default_idle);
+#endif
+
-+ /* Make sure we have a correctly sized P->M table. */
-+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
-+ phys_to_machine_mapping = alloc_bootmem_low_pages(
-+ max_pfn * sizeof(unsigned long));
-+ memset(phys_to_machine_mapping, ~0,
-+ max_pfn * sizeof(unsigned long));
-+ memcpy(phys_to_machine_mapping,
-+ (unsigned long *)xen_start_info->mfn_list,
-+ xen_start_info->nr_pages * sizeof(unsigned long));
-+ free_bootmem(
-+ __pa(xen_start_info->mfn_list),
-+ PFN_PHYS(PFN_UP(xen_start_info->nr_pages *
-+ sizeof(unsigned long))));
++#ifdef CONFIG_HOTPLUG_CPU
++extern cpumask_t cpu_initialized;
++static inline void play_dead(void)
++{
++ 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();
++}
++#else
++static inline void play_dead(void)
++{
++ BUG();
++}
++#endif /* CONFIG_HOTPLUG_CPU */
+
-+ /*
-+ * 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_low_pages(PAGE_SIZE);
-+ HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list =
-+ virt_to_mfn(pfn_to_mfn_frame_list_list);
++/*
++ * 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)
++{
++ int cpu = smp_processor_id();
+
-+ fpp = PAGE_SIZE/sizeof(unsigned long);
-+ for (i=0, j=0, k=-1; i< max_pfn; i+=fpp, j++) {
-+ if ((j % fpp) == 0) {
-+ k++;
-+ BUG_ON(k>=16);
-+ pfn_to_mfn_frame_list[k] =
-+ alloc_bootmem_low_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 = max_pfn;
-+ }
++ current_thread_info()->status |= TS_POLLING;
+
-+ /*
-+ * NOTE: at this point the bootmem allocator is fully available.
-+ */
++ /* endless idle loop with no priority at all */
++ while (1) {
++ while (!need_resched()) {
+
-+ if (is_initial_xendomain())
-+ dmi_scan_machine();
++ if (__get_cpu_var(cpu_idle_state))
++ __get_cpu_var(cpu_idle_state) = 0;
+
-+#ifdef CONFIG_X86_GENERICARCH
-+ generic_apic_probe(*cmdline_p);
-+#endif
-+ if (efi_enabled)
-+ efi_map_memmap();
++ rmb();
+
-+ set_iopl.iopl = 1;
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
++ if (cpu_is_offline(cpu))
++ play_dead();
+
-+#ifdef CONFIG_ACPI
-+ if (!is_initial_xendomain()) {
-+ printk(KERN_INFO "ACPI in unprivileged domain disabled\n");
-+ acpi_disabled = 1;
-+ acpi_ht = 0;
++ __get_cpu_var(irq_stat).idle_timestamp = jiffies;
++ xen_idle();
++ }
++ preempt_enable_no_resched();
++ schedule();
++ preempt_disable();
+ }
++}
+
-+ /*
-+ * Parse the ACPI tables for possible boot-time SMP configuration.
-+ */
-+ acpi_boot_table_init();
-+#endif
++void cpu_idle_wait(void)
++{
++ unsigned int cpu, this_cpu = get_cpu();
++ cpumask_t map, tmp = current->cpus_allowed;
+
-+#ifdef CONFIG_X86_IO_APIC
-+ check_acpi_pci(); /* Checks more than just ACPI actually */
-+#endif
++ set_cpus_allowed(current, cpumask_of_cpu(this_cpu));
++ put_cpu();
+
-+#ifdef CONFIG_ACPI
-+ acpi_boot_init();
++ cpus_clear(map);
++ for_each_online_cpu(cpu) {
++ per_cpu(cpu_idle_state, cpu) = 1;
++ cpu_set(cpu, map);
++ }
+
-+#if defined(CONFIG_SMP) && defined(CONFIG_X86_PC)
-+ if (def_to_bigsmp)
-+ printk(KERN_WARNING "More than 8 CPUs detected and "
-+ "CONFIG_X86_PC cannot handle it.\nUse "
-+ "CONFIG_X86_GENERICARCH or CONFIG_X86_BIGSMP.\n");
-+#endif
-+#endif
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ if (smp_found_config)
-+ get_smp_config();
-+#endif
-+#if defined(CONFIG_XEN) && defined(CONFIG_SMP)
-+ prefill_possible_map();
-+#endif
++ __get_cpu_var(cpu_idle_state) = 0;
+
-+ register_memory();
++ 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));
+
-+ if (is_initial_xendomain()) {
-+#ifdef CONFIG_VT
-+#if defined(CONFIG_VGA_CONSOLE)
-+ if (!efi_enabled ||
-+ (efi_mem_type(0xa0000) != EFI_CONVENTIONAL_MEMORY))
-+ conswitchp = &vga_con;
-+#elif defined(CONFIG_DUMMY_CONSOLE)
-+ conswitchp = &dummy_con;
-+#endif
-+#endif
-+ } else {
-+ extern int console_use_vt;
-+ console_use_vt = 0;
-+ }
-+#ifdef CONFIG_X86_TSC
-+ tsc_init();
-+#endif
++ set_cpus_allowed(current, tmp);
+}
++EXPORT_SYMBOL_GPL(cpu_idle_wait);
+
-+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;
-+}
++/* XXX XEN doesn't use mwait_idle(), select_idle_routine(), idle_setup(). */
++/* Always use xen_idle() instead. */
++void mwait_idle_with_hints(unsigned long eax, unsigned long ecx) {}
+
-+static __init int add_pcspkr(void)
++void __devinit select_idle_routine(const struct cpuinfo_x86 *c) {}
++
++void show_regs(struct pt_regs * regs)
+{
-+ struct platform_device *pd;
-+ int ret;
++ unsigned long cr0 = 0L, cr2 = 0L, cr3 = 0L, cr4 = 0L;
+
-+ pd = platform_device_alloc("pcspkr", -1);
-+ if (!pd)
-+ return -ENOMEM;
++ printk("\n");
++ printk("Pid: %d, comm: %20s\n", current->pid, current->comm);
++ printk("EIP: %04x:[<%08lx>] CPU: %d\n",0xffff & regs->xcs,regs->eip, smp_processor_id());
++ print_symbol("EIP is at %s\n", regs->eip);
+
-+ ret = platform_device_add(pd);
-+ if (ret)
-+ platform_device_put(pd);
++ if (user_mode_vm(regs))
++ printk(" ESP: %04x:%08lx",0xffff & regs->xss,regs->esp);
++ printk(" EFLAGS: %08lx %s (%s %.*s)\n",
++ regs->eflags, print_tainted(), init_utsname()->release,
++ (int)strcspn(init_utsname()->version, " "),
++ init_utsname()->version);
++ printk("EAX: %08lx EBX: %08lx ECX: %08lx EDX: %08lx\n",
++ regs->eax,regs->ebx,regs->ecx,regs->edx);
++ printk("ESI: %08lx EDI: %08lx EBP: %08lx",
++ regs->esi, regs->edi, regs->ebp);
++ printk(" DS: %04x ES: %04x GS: %04x\n",
++ 0xffff & regs->xds,0xffff & regs->xes, 0xffff & regs->xgs);
+
-+ return ret;
++ cr0 = read_cr0();
++ cr2 = read_cr2();
++ cr3 = read_cr3();
++ cr4 = read_cr4_safe();
++ printk("CR0: %08lx CR2: %08lx CR3: %08lx CR4: %08lx\n", cr0, cr2, cr3, cr4);
++ show_trace(NULL, regs, ®s->esp);
+}
-+device_initcall(add_pcspkr);
+
+/*
-+ * Local Variables:
-+ * mode:c
-+ * c-file-style:"k&r"
-+ * c-basic-offset:8
-+ * End:
++ * This gets run with %ebx containing the
++ * function to call, and %edx containing
++ * the "args".
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/smp-xen.c linux-2.6.18-xen/arch/i386/kernel/smp-xen.c
---- linux-2.6.18.3/arch/i386/kernel/smp-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/smp-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,624 @@
++extern void kernel_thread_helper(void);
++
+/*
-+ * Intel SMP support routines.
-+ *
-+ * (c) 1995 Alan Cox, Building #3 <alan at redhat.com>
-+ * (c) 1998-99, 2000 Ingo Molnar <mingo at redhat.com>
-+ *
-+ * This code is released under the GNU General Public License version 2 or
-+ * later.
++ * Create a kernel thread
+ */
++int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
++{
++ struct pt_regs regs;
+
-+#include <linux/init.h>
-+
-+#include <linux/mm.h>
-+#include <linux/delay.h>
-+#include <linux/spinlock.h>
-+#include <linux/smp_lock.h>
-+#include <linux/kernel_stat.h>
-+#include <linux/mc146818rtc.h>
-+#include <linux/cache.h>
-+#include <linux/interrupt.h>
-+#include <linux/cpu.h>
-+#include <linux/module.h>
++ memset(®s, 0, sizeof(regs));
+
-+#include <asm/mtrr.h>
-+#include <asm/tlbflush.h>
-+#if 0
-+#include <mach_apic.h>
-+#endif
-+#include <xen/evtchn.h>
++ regs.ebx = (unsigned long) fn;
++ regs.edx = (unsigned long) arg;
+
-+/*
-+ * Some notes on x86 processor bugs affecting SMP operation:
-+ *
-+ * Pentium, Pentium Pro, II, III (and all CPUs) have bugs.
-+ * The Linux implications for SMP are handled as follows:
-+ *
-+ * Pentium III / [Xeon]
-+ * None of the E1AP-E3AP errata are visible to the user.
-+ *
-+ * E1AP. see PII A1AP
-+ * E2AP. see PII A2AP
-+ * E3AP. see PII A3AP
-+ *
-+ * Pentium II / [Xeon]
-+ * None of the A1AP-A3AP errata are visible to the user.
-+ *
-+ * A1AP. see PPro 1AP
-+ * A2AP. see PPro 2AP
-+ * A3AP. see PPro 7AP
-+ *
-+ * Pentium Pro
-+ * None of 1AP-9AP errata are visible to the normal user,
-+ * except occasional delivery of 'spurious interrupt' as trap #15.
-+ * This is very rare and a non-problem.
-+ *
-+ * 1AP. Linux maps APIC as non-cacheable
-+ * 2AP. worked around in hardware
-+ * 3AP. fixed in C0 and above steppings microcode update.
-+ * Linux does not use excessive STARTUP_IPIs.
-+ * 4AP. worked around in hardware
-+ * 5AP. symmetric IO mode (normal Linux operation) not affected.
-+ * 'noapic' mode has vector 0xf filled out properly.
-+ * 6AP. 'noapic' mode might be affected - fixed in later steppings
-+ * 7AP. We do not assume writes to the LVT deassering IRQs
-+ * 8AP. We do not enable low power mode (deep sleep) during MP bootup
-+ * 9AP. We do not use mixed mode
-+ *
-+ * Pentium
-+ * There is a marginal case where REP MOVS on 100MHz SMP
-+ * machines with B stepping processors can fail. XXX should provide
-+ * an L1cache=Writethrough or L1cache=off option.
-+ *
-+ * B stepping CPUs may hang. There are hardware work arounds
-+ * for this. We warn about it in case your board doesn't have the work
-+ * arounds. Basically thats so I can tell anyone with a B stepping
-+ * CPU and SMP problems "tough".
-+ *
-+ * Specific items [From Pentium Processor Specification Update]
-+ *
-+ * 1AP. Linux doesn't use remote read
-+ * 2AP. Linux doesn't trust APIC errors
-+ * 3AP. We work around this
-+ * 4AP. Linux never generated 3 interrupts of the same priority
-+ * to cause a lost local interrupt.
-+ * 5AP. Remote read is never used
-+ * 6AP. not affected - worked around in hardware
-+ * 7AP. not affected - worked around in hardware
-+ * 8AP. worked around in hardware - we get explicit CS errors if not
-+ * 9AP. only 'noapic' mode affected. Might generate spurious
-+ * interrupts, we log only the first one and count the
-+ * rest silently.
-+ * 10AP. not affected - worked around in hardware
-+ * 11AP. Linux reads the APIC between writes to avoid this, as per
-+ * the documentation. Make sure you preserve this as it affects
-+ * the C stepping chips too.
-+ * 12AP. not affected - worked around in hardware
-+ * 13AP. not affected - worked around in hardware
-+ * 14AP. we always deassert INIT during bootup
-+ * 15AP. not affected - worked around in hardware
-+ * 16AP. not affected - worked around in hardware
-+ * 17AP. not affected - worked around in hardware
-+ * 18AP. not affected - worked around in hardware
-+ * 19AP. not affected - worked around in BIOS
-+ *
-+ * If this sounds worrying believe me these bugs are either ___RARE___,
-+ * or are signal timing bugs worked around in hardware and there's
-+ * about nothing of note with C stepping upwards.
-+ */
++ regs.xds = __USER_DS;
++ regs.xes = __USER_DS;
++ regs.xgs = __KERNEL_PDA;
++ regs.orig_eax = -1;
++ regs.eip = (unsigned long) kernel_thread_helper;
++ regs.xcs = __KERNEL_CS | get_kernel_rpl();
++ regs.eflags = X86_EFLAGS_IF | X86_EFLAGS_SF | X86_EFLAGS_PF | 0x2;
+
-+DEFINE_PER_CPU(struct tlb_state, cpu_tlbstate) ____cacheline_aligned = { &init_mm, 0, };
++ /* Ok, create the new process.. */
++ return do_fork(flags | CLONE_VM | CLONE_UNTRACED, 0, ®s, 0, NULL, NULL);
++}
++EXPORT_SYMBOL(kernel_thread);
+
+/*
-+ * the following functions deal with sending IPIs between CPUs.
-+ *
-+ * We use 'broadcast', CPU->CPU IPIs and self-IPIs too.
++ * Free current thread data structures etc..
+ */
-+
-+static inline int __prepare_ICR (unsigned int shortcut, int vector)
++void exit_thread(void)
+{
-+ unsigned int icr = shortcut | APIC_DEST_LOGICAL;
++ /* The process may have allocated an io port bitmap... nuke it. */
++ if (unlikely(test_thread_flag(TIF_IO_BITMAP))) {
++ struct task_struct *tsk = current;
++ struct thread_struct *t = &tsk->thread;
+
-+ switch (vector) {
-+ default:
-+ icr |= APIC_DM_FIXED | vector;
-+ break;
-+ case NMI_VECTOR:
-+ icr |= APIC_DM_NMI;
-+ break;
++ struct physdev_set_iobitmap set_iobitmap = { 0 };
++ HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap, &set_iobitmap);
++ kfree(t->io_bitmap_ptr);
++ t->io_bitmap_ptr = NULL;
++ clear_thread_flag(TIF_IO_BITMAP);
+ }
-+ return icr;
+}
+
-+static inline int __prepare_ICR2 (unsigned int mask)
++void flush_thread(void)
+{
-+ return SET_APIC_DEST_FIELD(mask);
++ struct task_struct *tsk = current;
++
++ memset(tsk->thread.debugreg, 0, sizeof(unsigned long)*8);
++ memset(tsk->thread.tls_array, 0, sizeof(tsk->thread.tls_array));
++ clear_tsk_thread_flag(tsk, TIF_DEBUG);
++ /*
++ * Forget coprocessor state..
++ */
++ clear_fpu(tsk);
++ clear_used_math();
+}
+
-+DECLARE_PER_CPU(int, ipi_to_irq[NR_IPIS]);
++void release_thread(struct task_struct *dead_task)
++{
++ BUG_ON(dead_task->mm);
++ release_vm86_irqs(dead_task);
++}
+
-+static inline void __send_IPI_one(unsigned int cpu, int vector)
++/*
++ * This gets called before we allocate a new thread and copy
++ * the current task into it.
++ */
++void prepare_to_copy(struct task_struct *tsk)
+{
-+ int irq = per_cpu(ipi_to_irq, cpu)[vector];
-+ BUG_ON(irq < 0);
-+ notify_remote_via_irq(irq);
++ unlazy_fpu(tsk);
+}
+
-+void __send_IPI_shortcut(unsigned int shortcut, int vector)
++int copy_thread(int nr, unsigned long clone_flags, unsigned long esp,
++ unsigned long unused,
++ struct task_struct * p, struct pt_regs * regs)
+{
-+ int cpu;
++ struct pt_regs * childregs;
++ struct task_struct *tsk;
++ int err;
+
-+ 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);
-+ }
++ childregs = task_pt_regs(p);
++ *childregs = *regs;
++ childregs->eax = 0;
++ childregs->esp = esp;
++
++ p->thread.esp = (unsigned long) childregs;
++ p->thread.esp0 = (unsigned long) (childregs+1);
++
++ p->thread.eip = (unsigned long) ret_from_fork;
++
++ savesegment(fs,p->thread.fs);
++
++ tsk = current;
++ if (unlikely(test_tsk_thread_flag(tsk, TIF_IO_BITMAP))) {
++ p->thread.io_bitmap_ptr = kmemdup(tsk->thread.io_bitmap_ptr,
++ IO_BITMAP_BYTES, GFP_KERNEL);
++ if (!p->thread.io_bitmap_ptr) {
++ p->thread.io_bitmap_max = 0;
++ return -ENOMEM;
+ }
-+ break;
-+ default:
-+ printk("XXXXXX __send_IPI_shortcut %08x vector %d\n", shortcut,
-+ vector);
-+ break;
++ set_tsk_thread_flag(p, TIF_IO_BITMAP);
++ }
++
++ /*
++ * Set a new TLS for the child thread?
++ */
++ if (clone_flags & CLONE_SETTLS) {
++ struct desc_struct *desc;
++ struct user_desc info;
++ int idx;
++
++ err = -EFAULT;
++ if (copy_from_user(&info, (void __user *)childregs->esi, sizeof(info)))
++ goto out;
++ err = -EINVAL;
++ if (LDT_empty(&info))
++ goto out;
++
++ idx = info.entry_number;
++ if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX)
++ goto out;
++
++ desc = p->thread.tls_array + idx - GDT_ENTRY_TLS_MIN;
++ desc->a = LDT_entry_a(&info);
++ desc->b = LDT_entry_b(&info);
+ }
-+}
+
-+void fastcall send_IPI_self(int vector)
-+{
-+ __send_IPI_shortcut(APIC_DEST_SELF, vector);
++ 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;
+}
+
+/*
-+ * This is only used on smaller machines.
++ * fill in the user structure for a core dump..
+ */
-+void send_IPI_mask_bitmask(cpumask_t mask, int vector)
++void dump_thread(struct pt_regs * regs, struct user * dump)
+{
-+ unsigned long flags;
-+ unsigned int cpu;
++ int i;
+
-+ local_irq_save(flags);
-+ WARN_ON(cpus_addr(mask)[0] & ~cpus_addr(cpu_online_map)[0]);
++/* changed the size calculations - should hopefully work better. lbt */
++ dump->magic = CMAGIC;
++ dump->start_code = 0;
++ dump->start_stack = regs->esp & ~(PAGE_SIZE - 1);
++ dump->u_tsize = ((unsigned long) current->mm->end_code) >> PAGE_SHIFT;
++ dump->u_dsize = ((unsigned long) (current->mm->brk + (PAGE_SIZE-1))) >> PAGE_SHIFT;
++ dump->u_dsize -= dump->u_tsize;
++ dump->u_ssize = 0;
++ for (i = 0; i < 8; i++)
++ dump->u_debugreg[i] = current->thread.debugreg[i];
+
-+ for (cpu = 0; cpu < NR_CPUS; ++cpu) {
-+ if (cpu_isset(cpu, mask)) {
-+ __send_IPI_one(cpu, vector);
-+ }
-+ }
++ if (dump->start_stack < TASK_SIZE)
++ dump->u_ssize = ((unsigned long) (TASK_SIZE - dump->start_stack)) >> PAGE_SHIFT;
+
-+ local_irq_restore(flags);
++ dump->regs.ebx = regs->ebx;
++ dump->regs.ecx = regs->ecx;
++ dump->regs.edx = regs->edx;
++ dump->regs.esi = regs->esi;
++ dump->regs.edi = regs->edi;
++ dump->regs.ebp = regs->ebp;
++ dump->regs.eax = regs->eax;
++ dump->regs.ds = regs->xds;
++ dump->regs.es = regs->xes;
++ savesegment(fs,dump->regs.fs);
++ dump->regs.gs = regs->xgs;
++ dump->regs.orig_eax = regs->orig_eax;
++ dump->regs.eip = regs->eip;
++ dump->regs.cs = regs->xcs;
++ dump->regs.eflags = regs->eflags;
++ dump->regs.esp = regs->esp;
++ dump->regs.ss = regs->xss;
++
++ dump->u_fpvalid = dump_fpu (regs, &dump->i387);
+}
++EXPORT_SYMBOL(dump_thread);
+
-+void send_IPI_mask_sequence(cpumask_t mask, int vector)
++/*
++ * 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 ptregs = *task_pt_regs(tsk);
++ ptregs.xcs &= 0xffff;
++ ptregs.xds &= 0xffff;
++ ptregs.xes &= 0xffff;
++ ptregs.xss &= 0xffff;
+
-+ send_IPI_mask_bitmask(mask, vector);
++ elf_core_copy_regs(regs, &ptregs);
++
++ return 1;
+}
+
-+#include <mach_ipi.h> /* must come after the send_IPI functions above for inlining */
++static noinline void __switch_to_xtra(struct task_struct *next_p)
++{
++ struct thread_struct *next;
+
-+#if 0 /* 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>
-+ */
++ next = &next_p->thread;
+
-+static cpumask_t flush_cpumask;
-+static struct mm_struct * flush_mm;
-+static unsigned long flush_va;
-+static DEFINE_SPINLOCK(tlbstate_lock);
-+#define FLUSH_ALL 0xffffffff
++ if (test_tsk_thread_flag(next_p, TIF_DEBUG)) {
++ set_debugreg(next->debugreg[0], 0);
++ set_debugreg(next->debugreg[1], 1);
++ set_debugreg(next->debugreg[2], 2);
++ set_debugreg(next->debugreg[3], 3);
++ /* no 4 and 5 */
++ set_debugreg(next->debugreg[6], 6);
++ set_debugreg(next->debugreg[7], 7);
++ }
++#ifndef CONFIG_XEN
++ if (!test_tsk_thread_flag(next_p, TIF_IO_BITMAP)) {
++ /*
++ * Disable the bitmap via an invalid offset. We still cache
++ * the previous bitmap owner and the IO bitmap contents:
++ */
++ tss->io_bitmap_base = INVALID_IO_BITMAP_OFFSET;
++ return;
++ }
++
++ if (likely(next == tss->io_bitmap_owner)) {
++ /*
++ * Previous owner of the bitmap (hence the bitmap content)
++ * matches the next task, we dont have to do anything but
++ * to set a valid offset in the TSS:
++ */
++ tss->io_bitmap_base = IO_BITMAP_OFFSET;
++ return;
++ }
++ /*
++ * Lazy TSS's I/O bitmap copy. We set an invalid offset here
++ * and we let the task to get a GPF in case an I/O instruction
++ * is performed. The handler of the GPF will verify that the
++ * faulting task has a valid I/O bitmap and, it true, does the
++ * real copy and restart the instruction. This will save us
++ * redundant copies when the currently switched task does not
++ * perform any I/O during its timeslice.
++ */
++ tss->io_bitmap_base = INVALID_IO_BITMAP_OFFSET_LAZY;
++#endif /* !CONFIG_XEN */
++}
+
+/*
-+ * We cannot call mmdrop() because we are in interrupt context,
-+ * instead update mm->cpu_vm_mask.
-+ *
-+ * We need to reload %cr3 since the page tables may be going
-+ * away from under us..
++ * This function selects if the context switch from prev to next
++ * has to tweak the TSC disable bit in the cr4.
+ */
-+static inline void leave_mm (unsigned long cpu)
++static inline void disable_tsc(struct task_struct *prev_p,
++ struct task_struct *next_p)
+{
-+ if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_OK)
-+ BUG();
-+ cpu_clear(cpu, per_cpu(cpu_tlbstate, cpu).active_mm->cpu_vm_mask);
-+ load_cr3(swapper_pg_dir);
++ struct thread_info *prev, *next;
++
++ /*
++ * gcc should eliminate the ->thread_info dereference if
++ * has_secure_computing returns 0 at compile time (SECCOMP=n).
++ */
++ prev = task_thread_info(prev_p);
++ next = task_thread_info(next_p);
++
++ if (has_secure_computing(prev) || has_secure_computing(next)) {
++ /* slow path here */
++ if (has_secure_computing(prev) &&
++ !has_secure_computing(next)) {
++ write_cr4(read_cr4() & ~X86_CR4_TSD);
++ } else if (!has_secure_computing(prev) &&
++ has_secure_computing(next))
++ write_cr4(read_cr4() | X86_CR4_TSD);
++ }
+}
+
+/*
++ * switch_to(x,yn) should switch tasks from x to y.
+ *
-+ * 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 superflous
-+ * tlb flush.
-+ * 1a2) set cpu_tlbstate to TLBSTATE_OK
-+ * Now the smp_invalidate_interrupt won't call leave_mm if cpu0
-+ * was in lazy tlb mode.
-+ * 1a3) update cpu_tlbstate[].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_tlbstate[].active_mm is correct, cpu0 already handles
-+ * flush ipis.
-+ * 1b1) set cpu_tlbstate 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
++ * We fsave/fwait so that an exception goes off at the right time
++ * (as a call from the fsave or fwait in effect) rather than to
++ * the wrong process. Lazy FP saving no longer makes any sense
++ * with modern CPU's, and this simplifies a lot of things (SMP
++ * and UP become the same).
+ *
-+ * 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.
++ * NOTE! We used to use the x86 hardware context switching. The
++ * reason for not using it any more becomes apparent when you
++ * try to recover gracefully from saved state that is no longer
++ * valid (stale segment register values in particular). With the
++ * hardware task-switch, there is no way to fix up bad state in
++ * a reasonable manner.
+ *
-+ * The good news is that cpu_tlbstate is local to each cpu, no
-+ * write/read ordering problems.
-+ */
-+
-+/*
-+ * TLB flush IPI:
++ * The fact that Intel documents the hardware task-switching to
++ * be slow is a fairly red herring - this code is not noticeably
++ * faster. However, there _is_ some room for improvement here,
++ * so the performance issues may eventually be a valid point.
++ * More important, however, is the fact that this allows us much
++ * more flexibility.
+ *
-+ * 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.
++ * The return value (in %eax) will be the "prev" task after
++ * the task-switch, and shows up in ret_from_fork in entry.S,
++ * for example.
+ */
-+
-+irqreturn_t smp_invalidate_interrupt(int irq, void *dev_id,
-+ struct pt_regs *regs)
++struct task_struct fastcall * __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
+{
-+ unsigned long cpu;
++ 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
++ struct physdev_set_iobitmap iobmp_op;
++ multicall_entry_t _mcl[8], *mcl = _mcl;
+
-+ cpu = get_cpu();
++ /* XEN NOTE: FS/GS saved in switch_mm(), not here. */
+
-+ if (!cpu_isset(cpu, 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 (flush_mm == per_cpu(cpu_tlbstate, cpu).active_mm) {
-+ if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_OK) {
-+ if (flush_va == FLUSH_ALL)
-+ local_flush_tlb();
-+ else
-+ __flush_tlb_one(flush_va);
-+ } else
-+ leave_mm(cpu);
++ /*
++ * This is basically '__unlazy_fpu', except that we queue a
++ * multicall to indicate FPU task switch, rather than
++ * synchronously trapping to Xen.
++ */
++ 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++;
+ }
-+ smp_mb__before_clear_bit();
-+ cpu_clear(cpu, flush_cpumask);
-+ smp_mb__after_clear_bit();
-+out:
-+ put_cpu_no_resched();
-+
-+ return IRQ_HANDLED;
-+}
++#if 0 /* lazy fpu sanity check */
++ else BUG_ON(!(read_cr0() & 8));
++#endif
+
-+static void flush_tlb_others(cpumask_t cpumask, struct mm_struct *mm,
-+ unsigned long va)
-+{
+ /*
-+ * A couple of (to be removed) sanity checks:
-+ *
-+ * - current CPU must not be in mask
-+ * - mask must exist :)
++ * Reload esp0.
++ * This is load_esp0(tss, next) with a multicall.
+ */
-+ BUG_ON(cpus_empty(cpumask));
-+ BUG_ON(cpu_isset(smp_processor_id(), cpumask));
-+ BUG_ON(!mm);
-+
-+ /* If a CPU which we ran on has gone down, OK. */
-+ cpus_and(cpumask, cpumask, cpu_online_map);
-+ if (cpus_empty(cpumask))
-+ return;
++ mcl->op = __HYPERVISOR_stack_switch;
++ mcl->args[0] = __KERNEL_DS;
++ mcl->args[1] = next->esp0;
++ mcl++;
+
+ /*
-+ * i'm not happy about this global shared spinlock in the
-+ * MM hot path, but we'll see how contended it is.
-+ * Temporarily this turns IRQs off, so that lockups are
-+ * detected by the NMI watchdog.
++ * Load the per-thread Thread-Local Storage descriptor.
++ * This is load_TLS(next, cpu) with multicalls.
+ */
-+ spin_lock(&tlbstate_lock);
-+
-+ flush_mm = mm;
-+ flush_va = va;
-+#if NR_CPUS <= BITS_PER_LONG
-+ atomic_set_mask(cpumask, &flush_cpumask);
-+#else
-+ {
-+ int k;
-+ unsigned long *flush_mask = (unsigned long *)&flush_cpumask;
-+ unsigned long *cpu_mask = (unsigned long *)&cpumask;
-+ for (k = 0; k < BITS_TO_LONGS(NR_CPUS); ++k)
-+ atomic_set_mask(cpu_mask[k], &flush_mask[k]);
++#define C(i) do { \
++ if (unlikely(next->tls_array[i].a != prev->tls_array[i].a || \
++ next->tls_array[i].b != prev->tls_array[i].b)) { \
++ mcl->op = __HYPERVISOR_update_descriptor; \
++ *(u64 *)&mcl->args[0] = virt_to_machine( \
++ &get_cpu_gdt_table(cpu)[GDT_ENTRY_TLS_MIN + i]);\
++ *(u64 *)&mcl->args[2] = *(u64 *)&next->tls_array[i]; \
++ mcl++; \
++ } \
++} while (0)
++ C(0); C(1); C(2);
++#undef C
++
++ if (unlikely(prev->io_bitmap_ptr || next->io_bitmap_ptr)) {
++ 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++;
+ }
-+#endif
-+ /*
-+ * We have to send the IPI only to
-+ * CPUs affected.
-+ */
-+ send_IPI_mask(cpumask, INVALIDATE_TLB_VECTOR);
+
-+ while (!cpus_empty(flush_cpumask))
-+ /* nothing. lockup detection does not belong here */
-+ mb();
++ (void)HYPERVISOR_multicall(_mcl, mcl - _mcl);
+
-+ flush_mm = NULL;
-+ flush_va = 0;
-+ spin_unlock(&tlbstate_lock);
-+}
-+
-+void flush_tlb_current_task(void)
-+{
-+ struct mm_struct *mm = current->mm;
-+ cpumask_t cpu_mask;
++ /*
++ * Restore %fs if needed.
++ *
++ * Glibc normally makes %fs be zero.
++ */
++ if (unlikely(next->fs))
++ loadsegment(fs, next->fs);
+
-+ preempt_disable();
-+ cpu_mask = mm->cpu_vm_mask;
-+ cpu_clear(smp_processor_id(), cpu_mask);
++ write_pda(pcurrent, next_p);
+
-+ local_flush_tlb();
-+ if (!cpus_empty(cpu_mask))
-+ flush_tlb_others(cpu_mask, mm, FLUSH_ALL);
-+ preempt_enable();
-+}
++ /* we're going to use this soon, after a few expensive things */
++ if (next_p->fpu_counter > 5)
++ prefetch(&next->i387.fxsave);
+
-+void flush_tlb_mm (struct mm_struct * mm)
-+{
-+ cpumask_t cpu_mask;
++ /*
++ * Now maybe handle debug registers and/or IO bitmaps
++ */
++ if (unlikely((task_thread_info(next_p)->flags & _TIF_WORK_CTXSW)
++ || test_tsk_thread_flag(prev_p, TIF_IO_BITMAP)))
++ __switch_to_xtra(next_p);
+
-+ preempt_disable();
-+ cpu_mask = mm->cpu_vm_mask;
-+ cpu_clear(smp_processor_id(), cpu_mask);
++ disable_tsc(prev_p, next_p);
+
-+ 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);
++ /* If the task has used fpu the last 5 timeslices, just do a full
++ * restore of the math state immediately to avoid the trap; the
++ * chances of needing FPU soon are obviously high now
++ */
++ if (next_p->fpu_counter > 5)
++ math_state_restore();
+
-+ preempt_enable();
++ return prev_p;
+}
+
-+void flush_tlb_page(struct vm_area_struct * vma, unsigned long va)
++asmlinkage int sys_fork(struct pt_regs regs)
+{
-+ 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();
++ return do_fork(SIGCHLD, regs.esp, ®s, 0, NULL, NULL);
+}
-+EXPORT_SYMBOL(flush_tlb_page);
+
-+static void do_flush_tlb_all(void* info)
++asmlinkage int sys_clone(struct pt_regs regs)
+{
-+ unsigned long cpu = smp_processor_id();
-+
-+ __flush_tlb_all();
-+ if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_LAZY)
-+ leave_mm(cpu);
-+}
++ unsigned long clone_flags;
++ unsigned long newsp;
++ int __user *parent_tidptr, *child_tidptr;
+
-+void flush_tlb_all(void)
-+{
-+ on_each_cpu(do_flush_tlb_all, NULL, 1, 1);
++ clone_flags = regs.ebx;
++ newsp = regs.ecx;
++ parent_tidptr = (int __user *)regs.edx;
++ child_tidptr = (int __user *)regs.edi;
++ if (!newsp)
++ newsp = regs.esp;
++ return do_fork(clone_flags, newsp, ®s, 0, parent_tidptr, child_tidptr);
+}
+
-+#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 */
-+
+/*
-+ * 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 ...
++ * 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.
+ */
-+void smp_send_reschedule(int cpu)
++asmlinkage int sys_vfork(struct pt_regs regs)
+{
-+ WARN_ON(cpu_is_offline(cpu));
-+ send_IPI_mask(cpumask_of_cpu(cpu), RESCHEDULE_VECTOR);
++ return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, regs.esp, ®s, 0, NULL, NULL);
+}
+
+/*
-+ * Structure and data for smp_call_function(). This is designed to minimise
-+ * static memory requirements. It also looks cleaner.
++ * sys_execve() executes a new program.
+ */
-+static DEFINE_SPINLOCK(call_lock);
++asmlinkage int sys_execve(struct pt_regs regs)
++{
++ int error;
++ char * filename;
+
-+struct call_data_struct {
-+ void (*func) (void *info);
-+ void *info;
-+ atomic_t started;
-+ atomic_t finished;
-+ int wait;
-+};
++ filename = getname((char __user *) regs.ebx);
++ error = PTR_ERR(filename);
++ if (IS_ERR(filename))
++ goto out;
++ error = do_execve(filename,
++ (char __user * __user *) regs.ecx,
++ (char __user * __user *) regs.edx,
++ ®s);
++ if (error == 0) {
++ task_lock(current);
++ current->ptrace &= ~PT_DTRACE;
++ task_unlock(current);
++ /* Make sure we don't return using sysenter.. */
++ set_thread_flag(TIF_IRET);
++ }
++ putname(filename);
++out:
++ return error;
++}
+
-+void lock_ipi_call_lock(void)
++#define top_esp (THREAD_SIZE - sizeof(unsigned long))
++#define top_ebp (THREAD_SIZE - 2*sizeof(unsigned long))
++
++unsigned long get_wchan(struct task_struct *p)
+{
-+ spin_lock_irq(&call_lock);
++ unsigned long ebp, esp, eip;
++ unsigned long stack_page;
++ int count = 0;
++ if (!p || p == current || p->state == TASK_RUNNING)
++ return 0;
++ stack_page = (unsigned long)task_stack_page(p);
++ esp = p->thread.esp;
++ if (!stack_page || esp < stack_page || esp > top_esp+stack_page)
++ return 0;
++ /* include/asm-i386/system.h:switch_to() pushes ebp last. */
++ ebp = *(unsigned long *) esp;
++ do {
++ if (ebp < stack_page || ebp > top_ebp+stack_page)
++ return 0;
++ eip = *(unsigned long *) (ebp+4);
++ if (!in_sched_functions(eip))
++ return eip;
++ ebp = *(unsigned long *) ebp;
++ } while (count++ < 16);
++ return 0;
+}
+
-+void unlock_ipi_call_lock(void)
++/*
++ * sys_alloc_thread_area: get a yet unused TLS descriptor index.
++ */
++static int get_free_idx(void)
+{
-+ spin_unlock_irq(&call_lock);
-+}
++ struct thread_struct *t = ¤t->thread;
++ int idx;
+
-+static struct call_data_struct *call_data;
++ for (idx = 0; idx < GDT_ENTRY_TLS_ENTRIES; idx++)
++ if (desc_empty(t->tls_array + idx))
++ return idx + GDT_ENTRY_TLS_MIN;
++ return -ESRCH;
++}
+
-+/**
-+ * 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.
++/*
++ * Set a given TLS descriptor:
+ */
-+int smp_call_function (void (*func) (void *info), void *info, int nonatomic,
-+ int wait)
++asmlinkage int sys_set_thread_area(struct user_desc __user *u_info)
+{
-+ struct call_data_struct data;
-+ int cpus;
++ struct thread_struct *t = ¤t->thread;
++ struct user_desc info;
++ struct desc_struct *desc;
++ int cpu, idx;
+
-+ /* Holding any lock stops cpus from going down. */
-+ spin_lock(&call_lock);
-+ cpus = num_online_cpus() - 1;
-+ if (!cpus) {
-+ spin_unlock(&call_lock);
-+ return 0;
++ if (copy_from_user(&info, u_info, sizeof(info)))
++ return -EFAULT;
++ idx = info.entry_number;
++
++ /*
++ * index -1 means the kernel should try to find and
++ * allocate an empty descriptor:
++ */
++ if (idx == -1) {
++ idx = get_free_idx();
++ if (idx < 0)
++ return idx;
++ if (put_user(idx, &u_info->entry_number))
++ return -EFAULT;
+ }
+
-+ /* Can deadlock when called with interrupts disabled */
-+ WARN_ON(irqs_disabled());
++ if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX)
++ return -EINVAL;
+
-+ data.func = func;
-+ data.info = info;
-+ atomic_set(&data.started, 0);
-+ data.wait = wait;
-+ if (wait)
-+ atomic_set(&data.finished, 0);
++ desc = t->tls_array + idx - GDT_ENTRY_TLS_MIN;
+
-+ call_data = &data;
-+ mb();
-+
-+ /* Send a message to all other CPUs and wait for them to respond */
-+ send_IPI_allbutself(CALL_FUNCTION_VECTOR);
++ /*
++ * We must not get preempted while modifying the TLS.
++ */
++ cpu = get_cpu();
+
-+ /* Wait for response */
-+ while (atomic_read(&data.started) != cpus)
-+ barrier();
++ if (LDT_empty(&info)) {
++ desc->a = 0;
++ desc->b = 0;
++ } else {
++ desc->a = LDT_entry_a(&info);
++ desc->b = LDT_entry_b(&info);
++ }
++ load_TLS(t, cpu);
+
-+ if (wait)
-+ while (atomic_read(&data.finished) != cpus)
-+ barrier();
-+ spin_unlock(&call_lock);
++ put_cpu();
+
+ return 0;
+}
-+EXPORT_SYMBOL(smp_call_function);
-+
-+static void stop_this_cpu (void * dummy)
-+{
-+ /*
-+ * Remove this CPU:
-+ */
-+ cpu_clear(smp_processor_id(), cpu_online_map);
-+ local_irq_disable();
-+#if 0
-+ disable_local_APIC();
-+#endif
-+ if (cpu_data[smp_processor_id()].hlt_works_ok)
-+ for(;;) halt();
-+ for (;;);
-+}
+
+/*
-+ * this function calls the 'stop' function on all other CPUs in the system.
++ * Get the current Thread-Local Storage area:
+ */
+
-+void smp_send_stop(void)
-+{
-+ smp_call_function(stop_this_cpu, NULL, 1, 0);
++#define GET_BASE(desc) ( \
++ (((desc)->a >> 16) & 0x0000ffff) | \
++ (((desc)->b << 16) & 0x00ff0000) | \
++ ( (desc)->b & 0xff000000) )
+
-+ local_irq_disable();
-+#if 0
-+ disable_local_APIC();
-+#endif
-+ local_irq_enable();
-+}
++#define GET_LIMIT(desc) ( \
++ ((desc)->a & 0x0ffff) | \
++ ((desc)->b & 0xf0000) )
++
++#define GET_32BIT(desc) (((desc)->b >> 22) & 1)
++#define GET_CONTENTS(desc) (((desc)->b >> 10) & 3)
++#define GET_WRITABLE(desc) (((desc)->b >> 9) & 1)
++#define GET_LIMIT_PAGES(desc) (((desc)->b >> 23) & 1)
++#define GET_PRESENT(desc) (((desc)->b >> 15) & 1)
++#define GET_USEABLE(desc) (((desc)->b >> 20) & 1)
+
-+/*
-+ * Reschedule call back. Nothing to do,
-+ * all the work is done automatically when
-+ * we return from the interrupt.
-+ */
-+irqreturn_t smp_reschedule_interrupt(int irq, void *dev_id,
-+ struct pt_regs *regs)
++asmlinkage int sys_get_thread_area(struct user_desc __user *u_info)
+{
++ struct user_desc info;
++ struct desc_struct *desc;
++ int idx;
+
-+ return IRQ_HANDLED;
-+}
++ if (get_user(idx, &u_info->entry_number))
++ return -EFAULT;
++ if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX)
++ return -EINVAL;
+
-+#include <linux/kallsyms.h>
-+irqreturn_t smp_call_function_interrupt(int irq, void *dev_id,
-+ struct pt_regs *regs)
-+{
-+ void (*func) (void *info) = call_data->func;
-+ void *info = call_data->info;
-+ int wait = call_data->wait;
++ memset(&info, 0, sizeof(info));
+
-+ /*
-+ * 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
-+ */
-+ irq_enter();
-+ (*func)(info);
-+ irq_exit();
++ desc = current->thread.tls_array + idx - GDT_ENTRY_TLS_MIN;
+
-+ if (wait) {
-+ mb();
-+ atomic_inc(&call_data->finished);
-+ }
++ info.entry_number = idx;
++ info.base_addr = GET_BASE(desc);
++ info.limit = GET_LIMIT(desc);
++ info.seg_32bit = GET_32BIT(desc);
++ info.contents = GET_CONTENTS(desc);
++ info.read_exec_only = !GET_WRITABLE(desc);
++ info.limit_in_pages = GET_LIMIT_PAGES(desc);
++ info.seg_not_present = !GET_PRESENT(desc);
++ info.useable = GET_USEABLE(desc);
+
-+ return IRQ_HANDLED;
++ if (copy_to_user(u_info, &info, sizeof(info)))
++ return -EFAULT;
++ return 0;
+}
+
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/swiotlb.c linux-2.6.18-xen/arch/i386/kernel/swiotlb.c
---- linux-2.6.18.3/arch/i386/kernel/swiotlb.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/swiotlb.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,672 @@
++unsigned long arch_align_stack(unsigned long sp)
++{
++ if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space)
++ sp -= get_random_int() % 8192;
++ return sp & ~0xf;
++}
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/quirks.c b/arch/i386/kernel/quirks.c
+--- a/arch/i386/kernel/quirks.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/quirks.c 2007-03-14 10:55:14.000000000 +0100
+@@ -7,7 +7,7 @@
+ #include <asm/genapic.h>
+ #include <asm/cpu.h>
+
+-#if defined(CONFIG_X86_IO_APIC) && defined(CONFIG_SMP) && defined(CONFIG_PCI)
++#if defined(CONFIG_X86_IO_APIC) && (defined(CONFIG_SMP) || defined(CONFIG_XEN)) && defined(CONFIG_PCI)
+ static void __devinit verify_quirk_intel_irqbalance(struct pci_dev *dev)
+ {
+ u8 config, rev;
+@@ -68,11 +68,19 @@
+ word = read_pci_config_16(0, 0, 0x40, 0x4c);
+
+ if (!(word & (1 << 13))) {
++#ifdef CONFIG_XEN
++ dom0_op_t op;
++ printk(KERN_INFO "Disabling irq balancing and affinity\n");
++ op.cmd = DOM0_PLATFORM_QUIRK;
++ op.u.platform_quirk.quirk_id = QUIRK_NOIRQBALANCING;
++ (void)HYPERVISOR_dom0_op(&op);
++#else
+ printk(KERN_INFO "Disabling irq balancing and affinity\n");
+ #ifdef CONFIG_IRQBALANCE
+ irqbalance_disable("");
+ #endif
+ noirqdebug_setup("");
++#endif /* CONFIG_XEN */
+ #ifdef CONFIG_PROC_FS
+ no_irq_affinity = 1;
+ #endif
+@@ -80,7 +88,7 @@
+ printk(KERN_INFO "Disabling cpu hotplug control\n");
+ enable_cpu_hotplug = 0;
+ #endif
+-#ifdef CONFIG_X86_64
++#if defined(CONFIG_X86_64) && !defined(CONFIG_XEN)
+ /* force the genapic selection to flat mode so that
+ * interrupts can be redirected to more than one CPU.
+ */
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/setup-xen.c b/arch/i386/kernel/setup-xen.c
+--- a/arch/i386/kernel/setup-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/setup-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,835 @@
+/*
-+ * Dynamic DMA mapping support.
++ * linux/arch/i386/kernel/setup.c
++ *
++ * Copyright (C) 1995 Linus Torvalds
++ *
++ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
++ *
++ * Memory region support
++ * David Parsons <orc at pell.chi.il.us>, July-August 1999
++ *
++ * Added E820 sanitization routine (removes overlapping memory regions);
++ * Brian Moyle <bmoyle at mvista.com>, February 2001
++ *
++ * Moved CPU detection code to cpu/${cpu}.c
++ * Patrick Mochel <mochel at osdl.org>, March 2002
++ *
++ * Provisions for empty E820 memory regions (reported by certain BIOSes).
++ * Alex Achenbach <xela at slit.de>, December 2002.
+ *
-+ * 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>
+ */
+
-+#include <linux/cache.h>
++/*
++ * This file handles the architecture-dependent parts of initialization
++ */
++
++#include <linux/sched.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/mmzone.h>
++#include <linux/screen_info.h>
++#include <linux/ioport.h>
++#include <linux/acpi.h>
++#include <linux/apm_bios.h>
++#include <linux/initrd.h>
+#include <linux/bootmem.h>
++#include <linux/seq_file.h>
++#include <linux/platform_device.h>
++#include <linux/console.h>
++#include <linux/mca.h>
++#include <linux/root_dev.h>
+#include <linux/highmem.h>
++#include <linux/module.h>
++#include <linux/efi.h>
++#include <linux/init.h>
++#include <linux/edd.h>
++#include <linux/nodemask.h>
++#include <linux/notifier.h>
++#include <linux/kexec.h>
++#include <linux/crash_dump.h>
++#include <linux/dmi.h>
++#include <linux/pfn.h>
++
++#include <video/edid.h>
++
++#include <asm/apic.h>
++#include <asm/e820.h>
++#include <asm/mpspec.h>
++#include <asm/mmzone.h>
++#include <asm/setup.h>
++#include <asm/arch_hooks.h>
++#include <asm/sections.h>
++#include <asm/io_apic.h>
++#include <asm/ist.h>
+#include <asm/io.h>
-+#include <asm/pci.h>
-+#include <asm/dma.h>
-+#include <asm/uaccess.h>
++#include <setup_arch.h>
++#include <bios_ebda.h>
++
++#include <asm/hypervisor.h>
++#include <xen/interface/physdev.h>
+#include <xen/interface/memory.h>
++#include <xen/features.h>
++#include <xen/xencons.h>
+
-+int swiotlb;
-+EXPORT_SYMBOL(swiotlb);
++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 */
++};
+
-+#define OFFSET(val,align) ((unsigned long)((val) & ( (align) - 1)))
++extern char hypercall_page[PAGE_SIZE];
++EXPORT_SYMBOL(hypercall_page);
+
-+#define SG_ENT_PHYS_ADDRESS(sg) (page_to_bus((sg)->page) + (sg)->offset)
++int disable_pse __devinitdata = 0;
+
+/*
-+ * 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.
++ * Machine setup..
+ */
-+#define IO_TLB_SEGSIZE 128
++extern struct resource code_resource;
++extern struct resource data_resource;
+
-+/*
-+ * log of the size of each IO TLB slab. The number of slabs is command line
-+ * controllable.
-+ */
-+#define IO_TLB_SHIFT 11
++/* cpu data as detected by the assembly code in head.S */
++struct cpuinfo_x86 new_cpu_data __cpuinitdata = { 0, 0, 0, 0, -1, 1, 0, 0, -1 };
++/* common cpu data for all cpus */
++struct cpuinfo_x86 boot_cpu_data __read_mostly = { 0, 0, 0, 0, -1, 1, 0, 0, -1 };
++EXPORT_SYMBOL(boot_cpu_data);
+
-+/* Width of DMA addresses in the IO TLB. 31 bits is an aacraid limitation. */
-+#define IO_TLB_DMA_BITS 31
++unsigned long mmu_cr4_features;
+
-+int swiotlb_force;
-+static char *iotlb_virt_start;
-+static unsigned long iotlb_nslabs;
++/* for MCA, but anyone else can use it if they want */
++unsigned int machine_id;
++#ifdef CONFIG_MCA
++EXPORT_SYMBOL(machine_id);
++#endif
++unsigned int machine_submodel_id;
++unsigned int BIOS_revision;
++unsigned int mca_pentium_flag;
+
-+/*
-+ * 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;
++/* Boot loader ID as an integer, for the benefit of proc_dointvec */
++int bootloader_type;
+
-+/* 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));
-+}
++/* user-defined highmem size */
++static unsigned int highmem_pages = -1;
+
+/*
-+ * When the IOMMU overflows we return a fallback buffer. This sets the size.
++ * Setup options
+ */
-+static unsigned long io_tlb_overflow = 32*1024;
++struct drive_info_struct { char dummy[32]; } drive_info;
++#if defined(CONFIG_BLK_DEV_IDE) || defined(CONFIG_BLK_DEV_HD) || \
++ defined(CONFIG_BLK_DEV_IDE_MODULE) || defined(CONFIG_BLK_DEV_HD_MODULE)
++EXPORT_SYMBOL(drive_info);
++#endif
++struct screen_info screen_info;
++EXPORT_SYMBOL(screen_info);
++struct apm_info apm_info;
++EXPORT_SYMBOL(apm_info);
++struct sys_desc_table_struct {
++ unsigned short length;
++ unsigned char table[0];
++};
++struct edid_info edid_info;
++EXPORT_SYMBOL_GPL(edid_info);
++struct ist_info ist_info;
++#if defined(CONFIG_X86_SPEEDSTEP_SMI) || \
++ defined(CONFIG_X86_SPEEDSTEP_SMI_MODULE)
++EXPORT_SYMBOL(ist_info);
++#endif
+
-+void *io_tlb_overflow_buffer;
++extern void early_cpu_init(void);
++extern int root_mountflags;
+
-+/*
-+ * 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;
++unsigned long saved_videomode;
+
-+/*
-+ * 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;
++#define RAMDISK_IMAGE_START_MASK 0x07FF
++#define RAMDISK_PROMPT_FLAG 0x8000
++#define RAMDISK_LOAD_FLAG 0x4000
+
-+/*
-+ * Protect the above data structures in the map and unmap calls
-+ */
-+static DEFINE_SPINLOCK(io_tlb_lock);
++static char command_line[COMMAND_LINE_SIZE];
+
-+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);
-+ /* Round up to power of two (xen_create_contiguous_region). */
-+ while (iotlb_nslabs & (iotlb_nslabs-1))
-+ iotlb_nslabs += iotlb_nslabs & ~(iotlb_nslabs-1);
-+ }
-+ 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? */
++unsigned char __initdata boot_params[PARAM_SIZE];
+
+/*
-+ * Statically reserve bounce buffer space and initialize bounce buffer data
-+ * structures for the software IO TLB used to implement the PCI DMA API.
++ * Point at the empty zero page to start with. We map the real shared_info
++ * page as soon as fixmap is up and running.
+ */
-+void
-+swiotlb_init_with_default_size (size_t default_size)
-+{
-+ unsigned long i, bytes;
-+
-+ 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);
-+ }
-+
-+ bytes = iotlb_nslabs * (1UL << IO_TLB_SHIFT);
-+
-+ /*
-+ * 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"
-+ "Use dom0_mem Xen boot parameter to reserve\n"
-+ "some DMA memory (e.g., dom0_mem=-128M).\n");
-+
-+ for (i = 0; i < iotlb_nslabs; i += IO_TLB_SEGSIZE) {
-+ int rc = xen_create_contiguous_region(
-+ (unsigned long)iotlb_virt_start + (i << IO_TLB_SHIFT),
-+ get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT),
-+ IO_TLB_DMA_BITS);
-+ BUG_ON(rc);
-+ }
++shared_info_t *HYPERVISOR_shared_info = (shared_info_t *)empty_zero_page;
++EXPORT_SYMBOL(HYPERVISOR_shared_info);
+
-+ /*
-+ * 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));
++unsigned long *phys_to_machine_mapping;
++unsigned long *pfn_to_mfn_frame_list_list, *pfn_to_mfn_frame_list[16];
++EXPORT_SYMBOL(phys_to_machine_mapping);
+
-+ /*
-+ * Get the overflow emergency buffer
-+ */
-+ io_tlb_overflow_buffer = alloc_bootmem_low(io_tlb_overflow);
++/* Raw start-of-day parameters from the hypervisor. */
++start_info_t *xen_start_info;
++EXPORT_SYMBOL(xen_start_info);
+
-+ 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: 0x%016lx - 0x%016lx\n",
-+ bytes >> 20,
-+ (unsigned long)iotlb_virt_start,
-+ (unsigned long)iotlb_virt_start + bytes);
++#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;
+}
-+
-+void
-+swiotlb_init(void)
++#else
++static inline void copy_edd(void)
+{
-+ long ram_end;
-+ size_t defsz = 64 * (1 << 20); /* 64MB default size */
-+
-+ 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;
-+ }
-+
-+ if (swiotlb)
-+ swiotlb_init_with_default_size(defsz);
-+ else
-+ printk(KERN_INFO "Software IO TLB disabled\n");
+}
++#endif
++
++int __initdata user_defined_memmap = 0;
+
+/*
-+ * 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.
++ * "mem=nopentium" disables the 4MB page tables.
++ * "mem=XXX[kKmM]" defines a memory region from HIGH_MEM
++ * to <mem>, overriding the bios size.
++ * "memmap=XXX[KkmM]@XXX[KkmM]" defines a memory region from
++ * <start> to <start>+<mem>, overriding the bios size.
++ *
++ * HPA tells me bootloaders need to parse mem=, so no new
++ * option should be mem= [also see Documentation/i386/boot.txt]
+ */
-+static void
-+__sync_single(struct phys_addr buffer, char *dma_addr, size_t size, int dir)
++static int __init parse_mem(char *arg)
+{
-+ if (PageHighMem(buffer.page)) {
-+ size_t len, bytes;
-+ char *dev, *host, *kmp;
-+ len = size;
-+ while (len != 0) {
-+ if (((bytes = len) + buffer.offset) > PAGE_SIZE)
-+ bytes = PAGE_SIZE - buffer.offset;
-+ kmp = kmap_atomic(buffer.page, KM_SWIOTLB);
-+ 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_SWIOTLB);
-+ len -= bytes;
-+ buffer.page++;
-+ buffer.offset = 0;
-+ }
++ if (!arg)
++ return -EINVAL;
++
++ if (strcmp(arg, "nopentium") == 0) {
++ clear_bit(X86_FEATURE_PSE, boot_cpu_data.x86_capability);
++ disable_pse = 1;
+ } 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);
++ /* If the user specifies memory size, we
++ * limit the BIOS-provided memory map to
++ * that size. exactmap can be used to specify
++ * the exact map. mem=number can be used to
++ * trim the existing memory map.
++ */
++ unsigned long long mem_size;
++
++ mem_size = memparse(arg, &arg);
++ limit_regions(mem_size);
++ user_defined_memmap = 1;
+ }
++ return 0;
++}
++early_param("mem", parse_mem);
++
++#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 */
+
+/*
-+ * Allocates bounce buffer and returns its kernel virtual address.
++ * highmem=size forces highmem to be exactly 'size' bytes.
++ * This works even on boxes that have no highmem otherwise.
++ * This also works to reduce highmem size on bigger boxes.
+ */
-+static void *
-+map_single(struct device *hwdev, struct phys_addr buffer, size_t size, int dir)
++static int __init parse_highmem(char *arg)
+{
-+ unsigned long flags;
-+ char *dma_addr;
-+ unsigned int nslots, stride, index, wrap;
-+ int i;
++ if (!arg)
++ return -EINVAL;
+
-+ /*
-+ * 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;
++ highmem_pages = memparse(arg, &arg) >> PAGE_SHIFT;
++ return 0;
++}
++early_param("highmem", parse_highmem);
+
-+ BUG_ON(!nslots);
++/*
++ * vmalloc=size forces the vmalloc area to be exactly 'size'
++ * bytes. This can be used to increase (or decrease) the
++ * vmalloc area - the default is 128m.
++ */
++static int __init parse_vmalloc(char *arg)
++{
++ if (!arg)
++ return -EINVAL;
+
-+ /*
-+ * 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);
++ __VMALLOC_RESERVE = memparse(arg, &arg);
++ return 0;
++}
++early_param("vmalloc", parse_vmalloc);
+
-+ if (index >= iotlb_nslabs)
-+ wrap = index = 0;
++/*
++ * reservetop=size reserves a hole at the top of the kernel address space which
++ * a hypervisor can load into later. Needed for dynamically loaded hypervisors,
++ * so relocating the fixmap can be done before paging initialization.
++ */
++static int __init parse_reservetop(char *arg)
++{
++ unsigned long address;
+
-+ 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;
++ if (!arg)
++ return -EINVAL;
+
-+ 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);
++ address = memparse(arg, &arg);
++ reserve_top_address(address);
++ return 0;
++}
++early_param("reservetop", parse_reservetop);
+
-+ /*
-+ * Update the indices to avoid searching in
-+ * the next round.
-+ */
-+ io_tlb_index =
-+ ((index + nslots) < iotlb_nslabs
-+ ? (index + nslots) : 0);
++/*
++ * Determine low and high memory ranges:
++ */
++unsigned long __init find_max_low_pfn(void)
++{
++ unsigned long max_low_pfn;
+
-+ goto found;
++ max_low_pfn = max_pfn;
++ if (max_low_pfn > MAXMEM_PFN) {
++ if (highmem_pages == -1)
++ highmem_pages = max_pfn - MAXMEM_PFN;
++ if (highmem_pages + MAXMEM_PFN < max_pfn)
++ max_pfn = MAXMEM_PFN + highmem_pages;
++ if (highmem_pages + MAXMEM_PFN > max_pfn) {
++ printk("only %luMB highmem pages available, ignoring highmem size of %uMB.\n", pages_to_mb(max_pfn - MAXMEM_PFN), pages_to_mb(highmem_pages));
++ highmem_pages = 0;
++ }
++ max_low_pfn = MAXMEM_PFN;
++#ifndef CONFIG_HIGHMEM
++ /* Maximum memory usable is what is directly addressable */
++ printk(KERN_WARNING "Warning only %ldMB will be used.\n",
++ MAXMEM>>20);
++ if (max_pfn > MAX_NONPAE_PFN)
++ printk(KERN_WARNING "Use a PAE enabled kernel.\n");
++ else
++ printk(KERN_WARNING "Use a HIGHMEM enabled kernel.\n");
++ max_pfn = MAXMEM_PFN;
++#else /* !CONFIG_HIGHMEM */
++#ifndef CONFIG_X86_PAE
++ if (max_pfn > MAX_NONPAE_PFN) {
++ max_pfn = MAX_NONPAE_PFN;
++ printk(KERN_WARNING "Warning only 4GB will be used.\n");
++ printk(KERN_WARNING "Use a PAE enabled kernel.\n");
++ }
++#endif /* !CONFIG_X86_PAE */
++#endif /* !CONFIG_HIGHMEM */
++ } else {
++ if (highmem_pages == -1)
++ highmem_pages = 0;
++#ifdef CONFIG_HIGHMEM
++ if (highmem_pages >= max_pfn) {
++ printk(KERN_ERR "highmem size specified (%uMB) is bigger than pages available (%luMB)!.\n", pages_to_mb(highmem_pages), pages_to_mb(max_pfn));
++ highmem_pages = 0;
++ }
++ if (highmem_pages) {
++ if (max_low_pfn-highmem_pages < 64*1024*1024/PAGE_SIZE){
++ printk(KERN_ERR "highmem size %uMB results in smaller than 64MB lowmem, ignoring it.\n", pages_to_mb(highmem_pages));
++ highmem_pages = 0;
+ }
-+ index += stride;
-+ if (index >= iotlb_nslabs)
-+ index = 0;
-+ } while (index != wrap);
-+
-+ spin_unlock_irqrestore(&io_tlb_lock, flags);
-+ return NULL;
++ max_low_pfn -= highmem_pages;
++ }
++#else
++ if (highmem_pages)
++ printk(KERN_ERR "ignoring highmem size on non-highmem kernel!\n");
++#endif
+ }
-+ 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.
-+ */
-+ io_tlb_orig_addr[index] = buffer;
-+ if ((dir == DMA_TO_DEVICE) || (dir == DMA_BIDIRECTIONAL))
-+ __sync_single(buffer, dma_addr, size, DMA_TO_DEVICE);
-+
-+ return dma_addr;
++ return max_low_pfn;
+}
+
++#ifndef CONFIG_XEN
+/*
-+ * dma_addr is the kernel virtual address of the bounce buffer to unmap.
++ * workaround for Dell systems that neglect to reserve EBDA
+ */
-+static void
-+unmap_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
++static void __init reserve_ebda_region(void)
+{
-+ 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 = io_tlb_orig_addr[index];
++ unsigned int addr;
++ addr = get_bios_ebda();
++ if (addr)
++ reserve_bootmem(addr, PAGE_SIZE);
++}
++#endif
+
++#ifndef CONFIG_NEED_MULTIPLE_NODES
++void __init setup_bootmem_allocator(void);
++static unsigned long __init setup_memory(void)
++{
+ /*
-+ * First, sync the memory before unmapping the entry
++ * partially used pages are not usable - thus
++ * we are rounding upwards:
+ */
-+ if ((dir == DMA_FROM_DEVICE) || (dir == DMA_BIDIRECTIONAL))
-+ __sync_single(buffer, dma_addr, size, DMA_FROM_DEVICE);
++ min_low_pfn = PFN_UP(__pa(xen_start_info->pt_base)) +
++ xen_start_info->nr_pt_frames;
+
-+ /*
-+ * 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;
++ find_max_pfn();
++
++ max_low_pfn = find_max_low_pfn();
++
++#ifdef CONFIG_HIGHMEM
++ highstart_pfn = highend_pfn = max_pfn;
++ if (max_pfn > max_low_pfn) {
++ highstart_pfn = max_low_pfn;
+ }
-+ spin_unlock_irqrestore(&io_tlb_lock, flags);
-+}
++ printk(KERN_NOTICE "%ldMB HIGHMEM available.\n",
++ pages_to_mb(highend_pfn - highstart_pfn));
++ num_physpages = highend_pfn;
++ high_memory = (void *) __va(highstart_pfn * PAGE_SIZE - 1) + 1;
++#else
++ num_physpages = max_low_pfn;
++ high_memory = (void *) __va(max_low_pfn * PAGE_SIZE - 1) + 1;
++#endif
++#ifdef CONFIG_FLATMEM
++ max_mapnr = num_physpages;
++#endif
++ printk(KERN_NOTICE "%ldMB LOWMEM available.\n",
++ pages_to_mb(max_low_pfn));
+
-+static void
-+sync_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
-+{
-+ int index = (dma_addr - iotlb_virt_start) >> IO_TLB_SHIFT;
-+ struct phys_addr buffer = io_tlb_orig_addr[index];
-+ BUG_ON((dir != DMA_FROM_DEVICE) && (dir != DMA_TO_DEVICE));
-+ __sync_single(buffer, dma_addr, size, dir);
++ setup_bootmem_allocator();
++
++ return max_low_pfn;
+}
+
-+static void
-+swiotlb_full(struct device *dev, size_t size, int dir, int do_panic)
++void __init zone_sizes_init(void)
+{
+ /*
-+ * 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.
++ * 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!
+ */
-+ printk(KERN_ERR "PCI-DMA: Out of SW-IOMMU space for %lu bytes at "
-+ "device %s\n", (unsigned long)size, dev ? dev->bus_id : "?");
++ unsigned long max_zone_pfns[MAX_NR_ZONES];
++ memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
++ max_zone_pfns[ZONE_DMA] = max_low_pfn;
++ max_zone_pfns[ZONE_NORMAL] = max_low_pfn;
++#ifdef CONFIG_HIGHMEM
++ max_zone_pfns[ZONE_HIGHMEM] = highend_pfn;
++ add_active_range(0, 0, highend_pfn);
++#else
++ add_active_range(0, 0, max_low_pfn);
++#endif
+
-+ 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");
-+ }
++ free_area_init_nodes(max_zone_pfns);
+}
++#else
++extern unsigned long __init setup_memory(void);
++extern void zone_sizes_init(void);
++#endif /* !CONFIG_NEED_MULTIPLE_NODES */
+
-+/*
-+ * 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)
++void __init setup_bootmem_allocator(void)
+{
-+ dma_addr_t dev_addr = virt_to_bus(ptr);
-+ void *map;
-+ struct phys_addr buffer;
++ unsigned long bootmap_size;
++ /*
++ * Initialize the boot-time allocator (with low memory only):
++ */
++ bootmap_size = init_bootmem(min_low_pfn, max_low_pfn);
+
-+ BUG_ON(dir == DMA_NONE);
++ register_bootmem_low_pages(max_low_pfn);
+
+ /*
-+ * 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.
++ * Reserve the bootmem bitmap itself as well. We do this in two
++ * steps (first step was init_bootmem()) because this catches
++ * the (very unlikely) case of us accidentally initializing the
++ * bootmem allocator with an invalid RAM area.
+ */
-+ if (!range_straddles_page_boundary(ptr, size) &&
-+ !address_needs_mapping(hwdev, dev_addr))
-+ return dev_addr;
++ reserve_bootmem(__pa_symbol(_text), (PFN_PHYS(min_low_pfn) +
++ bootmap_size + PAGE_SIZE-1) - __pa_symbol(_text));
+
++#ifndef CONFIG_XEN
+ /*
-+ * Oh well, have to allocate and map a bounce buffer.
++ * reserve physical page 0 - it's a special BIOS page on many boxes,
++ * enabling clean reboots, SMP operation, laptop functions.
+ */
-+ 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;
-+}
-+
-+/*
-+ * 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);
-+}
-+
-+/*
-+ * 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);
-+}
++ reserve_bootmem(0, PAGE_SIZE);
+
-+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);
-+}
++ /* reserve EBDA region, it's a 4K region */
++ reserve_ebda_region();
+
-+/*
-+ * 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;
++ /* could be an AMD 768MPX chipset. Reserve a page before VGA to prevent
++ PCI prefetch into it (errata #56). Usually the page is reserved anyways,
++ unless you have no PS/2 mouse plugged in. */
++ if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD &&
++ boot_cpu_data.x86 == 6)
++ reserve_bootmem(0xa0000 - 4096, 4096);
+
-+ BUG_ON(dir == DMA_NONE);
++#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(PAGE_SIZE, PAGE_SIZE);
++#endif
++#ifdef CONFIG_ACPI_SLEEP
++ /*
++ * Reserve low memory region for sleep support.
++ */
++ acpi_reserve_bootmem();
++#endif
++#endif /* !CONFIG_XEN */
+
-+ for (i = 0; i < nelems; i++, sg++) {
-+ dev_addr = SG_ENT_PHYS_ADDRESS(sg);
-+ if (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;
++#ifdef CONFIG_BLK_DEV_INITRD
++ if (xen_start_info->mod_start) {
++ if (INITRD_START + INITRD_SIZE <= (max_low_pfn << PAGE_SHIFT)) {
++ /*reserve_bootmem(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",
++ INITRD_START + INITRD_SIZE,
++ max_low_pfn << PAGE_SHIFT);
++ initrd_start = 0;
++ }
+ }
-+ return nelems;
++#endif
++#ifdef CONFIG_KEXEC
++ if (crashk_res.start != crashk_res.end)
++ reserve_bootmem(crashk_res.start,
++ crashk_res.end - crashk_res.start + 1);
++#endif
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap))
++ phys_to_machine_mapping =
++ (unsigned long *)xen_start_info->mfn_list;
+}
+
+/*
-+ * Unmap a set of streaming mode DMA translations. Again, cpu read rules
-+ * concerning calls here are the same as for swiotlb_unmap_single() above.
++ * The node 0 pgdat is initialized before all of these because
++ * it's needed for bootmem. node>0 pgdats have their virtual
++ * space allocated before the pagetables are in place to access
++ * them, so they can't be cleared then.
++ *
++ * This should all compile down to nothing when NUMA is off.
+ */
-+void
-+swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nelems,
-+ int dir)
++void __init remapped_pgdat_init(void)
+{
-+ int i;
++ int nid;
+
-+ BUG_ON(dir == DMA_NONE);
++ for_each_online_node(nid) {
++ if (nid != 0)
++ memset(NODE_DATA(nid), 0, sizeof(struct pglist_data));
++ }
++}
+
-+ 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);
++#ifdef CONFIG_MCA
++static void set_mca_bus(int x)
++{
++ MCA_bus = x;
++}
++#else
++static void set_mca_bus(int x) { }
++#endif
++
++/* Overridden in paravirt.c if CONFIG_PARAVIRT */
++char * __init __attribute__((weak)) memory_setup(void)
++{
++ return machine_specific_memory_setup();
+}
+
+/*
-+ * 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.
++ * Determine if we were loaded by an EFI loader. If so, then we have also been
++ * passed the efi memmap, systab, etc., so we should use these data structures
++ * for initialization. Note, the efi init code path is determined by the
++ * global efi_enabled. This allows the same kernel image to be used on existing
++ * systems (with a traditional BIOS) as well as on EFI systems.
+ */
-+void
-+swiotlb_sync_sg_for_cpu(struct device *hwdev, struct scatterlist *sg,
-+ int nelems, int dir)
++void __init setup_arch(char **cmdline_p)
+{
-+ int i;
++ int i, j, k, fpp;
++ struct physdev_set_iopl set_iopl;
++ unsigned long max_low_pfn;
+
-+ BUG_ON(dir == DMA_NONE);
++ /* Force a quick death if the kernel panics (not domain 0). */
++ extern int panic_timeout;
++ if (!panic_timeout && !is_initial_xendomain())
++ panic_timeout = 1;
+
-+ 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);
-+}
++ /* Register a call for panic conditions. */
++ atomic_notifier_chain_register(&panic_notifier_list, &xen_panic_block);
+
-+void
-+swiotlb_sync_sg_for_device(struct device *hwdev, struct scatterlist *sg,
-+ int nelems, int dir)
-+{
-+ int i;
++ HYPERVISOR_vm_assist(VMASST_CMD_enable, VMASST_TYPE_4gb_segments);
++ HYPERVISOR_vm_assist(VMASST_CMD_enable,
++ VMASST_TYPE_writable_pagetables);
+
-+ BUG_ON(dir == DMA_NONE);
++ memcpy(&boot_cpu_data, &new_cpu_data, sizeof(new_cpu_data));
++ pre_setup_arch_hook();
++ early_cpu_init();
+
-+ 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);
-+}
++ /*
++ * FIXME: This isn't an official loader_type right
++ * now but does currently work with elilo.
++ * If we were configured as an EFI kernel, check to make
++ * sure that we were loaded correctly from elilo and that
++ * the system table is valid. If not, then initialize normally.
++ */
++#ifdef CONFIG_EFI
++ if ((LOADER_TYPE == 0x50) && EFI_SYSTAB)
++ efi_enabled = 1;
++#endif
+
-+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;
++ /* This must be initialized to UNNAMED_MAJOR for ipconfig to work
++ properly. Setting ROOT_DEV to default to /dev/ram0 breaks initrd.
++ */
++ ROOT_DEV = MKDEV(UNNAMED_MAJOR,0);
++ drive_info = DRIVE_INFO;
++ screen_info = SCREEN_INFO;
++ edid_info = EDID_INFO;
++ apm_info.bios = APM_BIOS_INFO;
++ ist_info = IST_INFO;
++ saved_videomode = VIDEO_MODE;
++ if( SYS_DESC_TABLE.length != 0 ) {
++ set_mca_bus(SYS_DESC_TABLE.table[3] & 0x2);
++ machine_id = SYS_DESC_TABLE.table[0];
++ machine_submodel_id = SYS_DESC_TABLE.table[1];
++ BIOS_revision = SYS_DESC_TABLE.table[2];
++ }
++ bootloader_type = LOADER_TYPE;
+
-+ 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;
++ 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);
+ }
-+ dev_addr = (dma_addr_t)virt_to_bus(map);
-+ }
++ xen_start_info->console.domU.mfn = 0;
++ xen_start_info->console.domU.evtchn = 0;
++ } else
++ screen_info.orig_video_isVGA = 0;
+
-+ return dev_addr;
-+}
++#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
+
-+void
-+swiotlb_unmap_page(struct device *hwdev, dma_addr_t dma_address,
-+ size_t size, enum dma_data_direction direction)
-+{
-+ BUG_ON(direction == DMA_NONE);
-+ if (in_swiotlb_aperture(dma_address))
-+ unmap_single(hwdev, bus_to_virt(dma_address), size, direction);
-+}
++ setup_xen_features();
+
-+int
-+swiotlb_dma_mapping_error(dma_addr_t dma_addr)
-+{
-+ return (dma_addr == virt_to_bus(io_tlb_overflow_buffer));
-+}
++ ARCH_SETUP
++ if (efi_enabled)
++ efi_init();
++ else {
++ printk(KERN_INFO "BIOS-provided physical RAM map:\n");
++ print_memory_map(memory_setup());
++ }
+
-+/*
-+ * 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 << IO_TLB_DMA_BITS) - 1));
-+}
++ copy_edd();
+
-+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_map_page);
-+EXPORT_SYMBOL(swiotlb_unmap_page);
-+EXPORT_SYMBOL(swiotlb_dma_mapping_error);
-+EXPORT_SYMBOL(swiotlb_dma_supported);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/sysenter.c linux-2.6.18-xen/arch/i386/kernel/sysenter.c
---- linux-2.6.18.3/arch/i386/kernel/sysenter.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/kernel/sysenter.c 2006-11-19 14:26:22.000000000 +0100
-@@ -23,6 +23,10 @@
- #include <asm/pgtable.h>
- #include <asm/unistd.h>
-
-+#ifdef CONFIG_XEN
-+#include <xen/interface/callback.h>
-+#endif
++ 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 = (PFN_UP(__pa(xen_start_info->pt_base)) +
++ xen_start_info->nr_pt_frames) << PAGE_SHIFT;
+
- /*
- * 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_X86_NO_TSS
- int cpu = get_cpu();
- struct tss_struct *tss = &per_cpu(init_tss, cpu);
-
-@@ -58,6 +63,7 @@
- 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 @@
- {
- syscall_page = (void *)get_zeroed_page(GFP_ATOMIC);
-
-+#ifdef CONFIG_XEN
-+ if (boot_cpu_has(X86_FEATURE_SEP)) {
-+ struct callback_register sysenter = {
-+ .type = CALLBACKTYPE_sysenter,
-+ .address = { __KERNEL_CS, (unsigned long)sysenter_entry },
-+ };
++ 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 (HYPERVISOR_callback_op(CALLBACKOP_register, &sysenter) < 0)
-+ clear_bit(X86_FEATURE_SEP, boot_cpu_data.x86_capability);
++ parse_early_param();
++
++ if (user_defined_memmap) {
++ printk(KERN_INFO "user-defined physical RAM map:\n");
++ print_memory_map("user");
+ }
++
++ strlcpy(command_line, saved_command_line, COMMAND_LINE_SIZE);
++ *cmdline_p = command_line;
++
++ max_low_pfn = setup_memory();
++
++ /*
++ * NOTE: before this point _nobody_ is allowed to allocate
++ * any memory using the bootmem allocator. Although the
++ * alloctor is now initialised only the first 8Mb of the kernel
++ * virtual address space has been mapped. All allocations before
++ * paging_init() has completed must use the alloc_bootmem_low_pages()
++ * variant (which allocates DMA'able memory) and care must be taken
++ * not to exceed the 8Mb limit.
++ */
++
++#ifdef CONFIG_SMP
++ smp_alloc_memory(); /* AP processor realmode stacks in low memory*/
+#endif
++ paging_init();
++ remapped_pgdat_init();
++ sparse_init();
++ zone_sizes_init();
+
- #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));
-@@ -79,8 +97,12 @@
- /*
- * In the non-compat case the ELF coredumping code needs the fixmap:
- */
-+#ifdef CONFIG_XEN
-+ __set_fixmap(FIX_VDSO, virt_to_machine(syscall_page), PAGE_KERNEL_RO);
-+#else
- __set_fixmap(FIX_VDSO, __pa(syscall_page), PAGE_KERNEL_RO);
- #endif
++#ifdef CONFIG_X86_FIND_SMP_CONFIG
++ /*
++ * Find and reserve possible boot-time SMP configuration:
++ */
++ find_smp_config();
+#endif
-
- if (!boot_cpu_has(X86_FEATURE_SEP)) {
- memcpy(syscall_page,
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/time-xen.c linux-2.6.18-xen/arch/i386/kernel/time-xen.c
---- linux-2.6.18.3/arch/i386/kernel/time-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/time-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,1101 @@
-+/*
-+ * 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).
-+ */
++ numa_kva_reserve();
+
-+#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>
++ /* Make sure we have a correctly sized P->M table. */
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ phys_to_machine_mapping = alloc_bootmem_low_pages(
++ max_pfn * sizeof(unsigned long));
++ memset(phys_to_machine_mapping, ~0,
++ max_pfn * sizeof(unsigned long));
++ memcpy(phys_to_machine_mapping,
++ (unsigned long *)xen_start_info->mfn_list,
++ xen_start_info->nr_pages * sizeof(unsigned long));
++ free_bootmem(
++ __pa(xen_start_info->mfn_list),
++ PFN_PHYS(PFN_UP(xen_start_info->nr_pages *
++ sizeof(unsigned long))));
+
-+#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>
++ /*
++ * 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_low_pages(PAGE_SIZE);
++ HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list =
++ virt_to_mfn(pfn_to_mfn_frame_list_list);
+
-+#include "mach_time.h"
++ fpp = PAGE_SIZE/sizeof(unsigned long);
++ for (i=0, j=0, k=-1; i< max_pfn; i+=fpp, j++) {
++ if ((j % fpp) == 0) {
++ k++;
++ BUG_ON(k>=16);
++ pfn_to_mfn_frame_list[k] =
++ alloc_bootmem_low_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 = max_pfn;
++ }
+
-+#include <linux/timex.h>
++ /*
++ * NOTE: at this point the bootmem allocator is fully available.
++ */
+
-+#include <asm/hpet.h>
++ if (is_initial_xendomain())
++ dmi_scan_machine();
+
-+#include <asm/arch_hooks.h>
++#ifdef CONFIG_X86_GENERICARCH
++ generic_apic_probe();
++#endif
++ if (efi_enabled)
++ efi_map_memmap();
+
-+#include <xen/evtchn.h>
-+#include <xen/interface/vcpu.h>
++ set_iopl.iopl = 1;
++ HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
+
-+#if defined (__i386__)
-+#include <asm/i8259.h>
++#ifdef CONFIG_ACPI
++ if (!is_initial_xendomain()) {
++ printk(KERN_INFO "ACPI in unprivileged domain disabled\n");
++ acpi_disabled = 1;
++ acpi_ht = 0;
++ }
++
++ /*
++ * Parse the ACPI tables for possible boot-time SMP configuration.
++ */
++ acpi_boot_table_init();
+#endif
+
-+int pit_latch_buggy; /* extern */
++#ifdef CONFIG_PCI
++#ifdef CONFIG_X86_IO_APIC
++ check_acpi_pci(); /* Checks more than just ACPI actually */
++#endif
++#endif
+
-+#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;
++#ifdef CONFIG_ACPI
++ acpi_boot_init();
++
++#if defined(CONFIG_SMP) && defined(CONFIG_X86_PC)
++ if (def_to_bigsmp)
++ printk(KERN_WARNING "More than 8 CPUs detected and "
++ "CONFIG_X86_PC cannot handle it.\nUse "
++ "CONFIG_X86_GENERICARCH or CONFIG_X86_BIGSMP.\n");
++#endif
++#endif
++#ifdef CONFIG_X86_LOCAL_APIC
++ if (smp_found_config)
++ get_smp_config();
++#endif
++#if defined(CONFIG_XEN) && defined(CONFIG_SMP)
++ prefill_possible_map();
+#endif
+
-+#define USEC_PER_TICK (USEC_PER_SEC / HZ)
-+#define NSEC_PER_TICK (NSEC_PER_SEC / HZ)
-+#define FSEC_PER_TICK (FSEC_PER_SEC / HZ)
++ e820_register_memory();
+
-+#define NS_SCALE 10 /* 2^10, carefully chosen */
-+#define US_SCALE 32 /* 2^32, arbitralrily chosen */
++ if (is_initial_xendomain()) {
++#ifdef CONFIG_VT
++#if defined(CONFIG_VGA_CONSOLE)
++ if (!efi_enabled ||
++ (efi_mem_type(0xa0000) != EFI_CONVENTIONAL_MEMORY))
++ 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
++ }
++#ifdef CONFIG_X86_TSC
++ tsc_init();
++#endif
++}
+
-+unsigned int cpu_khz; /* Detected as we calibrate the TSC */
-+EXPORT_SYMBOL(cpu_khz);
++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;
++}
+
-+extern unsigned long wall_jiffies;
++static __init int add_pcspkr(void)
++{
++ struct platform_device *pd;
++ int ret;
+
-+DEFINE_SPINLOCK(rtc_lock);
-+EXPORT_SYMBOL(rtc_lock);
++ pd = platform_device_alloc("pcspkr", -1);
++ if (!pd)
++ return -ENOMEM;
+
-+extern struct init_timer_opts timer_tsc_init;
-+extern struct timer_opts timer_tsc;
-+#define timer_none timer_tsc
++ ret = platform_device_add(pd);
++ if (ret)
++ platform_device_put(pd);
+
-+/* 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;
++ return ret;
++}
++device_initcall(add_pcspkr);
++
++/*
++ * Local Variables:
++ * mode:c
++ * c-file-style:"k&r"
++ * c-basic-offset:8
++ * End:
++ */
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/smp-xen.c b/arch/i386/kernel/smp-xen.c
+--- a/arch/i386/kernel/smp-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/smp-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,624 @@
++/*
++ * Intel SMP support routines.
++ *
++ * (c) 1995 Alan Cox, Building #3 <alan at redhat.com>
++ * (c) 1998-99, 2000 Ingo Molnar <mingo at redhat.com>
++ *
++ * This code is released under the GNU General Public License version 2 or
++ * later.
++ */
+
-+/* 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);
++#include <linux/init.h>
+
-+/* 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);
++#include <linux/mm.h>
++#include <linux/delay.h>
++#include <linux/spinlock.h>
++#include <linux/smp_lock.h>
++#include <linux/kernel_stat.h>
++#include <linux/mc146818rtc.h>
++#include <linux/cache.h>
++#include <linux/interrupt.h>
++#include <linux/cpu.h>
++#include <linux/module.h>
+
-+/* Current runstate of each CPU (updated automatically by the hypervisor). */
-+static DEFINE_PER_CPU(struct vcpu_runstate_info, runstate);
++#include <asm/mtrr.h>
++#include <asm/tlbflush.h>
++#if 0
++#include <mach_apic.h>
++#endif
++#include <xen/evtchn.h>
+
-+/* Must be signed, as it's compared with s64 quantities which can be -ve. */
-+#define NS_PER_TICK (1000000000LL/HZ)
++/*
++ * Some notes on x86 processor bugs affecting SMP operation:
++ *
++ * Pentium, Pentium Pro, II, III (and all CPUs) have bugs.
++ * The Linux implications for SMP are handled as follows:
++ *
++ * Pentium III / [Xeon]
++ * None of the E1AP-E3AP errata are visible to the user.
++ *
++ * E1AP. see PII A1AP
++ * E2AP. see PII A2AP
++ * E3AP. see PII A3AP
++ *
++ * Pentium II / [Xeon]
++ * None of the A1AP-A3AP errata are visible to the user.
++ *
++ * A1AP. see PPro 1AP
++ * A2AP. see PPro 2AP
++ * A3AP. see PPro 7AP
++ *
++ * Pentium Pro
++ * None of 1AP-9AP errata are visible to the normal user,
++ * except occasional delivery of 'spurious interrupt' as trap #15.
++ * This is very rare and a non-problem.
++ *
++ * 1AP. Linux maps APIC as non-cacheable
++ * 2AP. worked around in hardware
++ * 3AP. fixed in C0 and above steppings microcode update.
++ * Linux does not use excessive STARTUP_IPIs.
++ * 4AP. worked around in hardware
++ * 5AP. symmetric IO mode (normal Linux operation) not affected.
++ * 'noapic' mode has vector 0xf filled out properly.
++ * 6AP. 'noapic' mode might be affected - fixed in later steppings
++ * 7AP. We do not assume writes to the LVT deassering IRQs
++ * 8AP. We do not enable low power mode (deep sleep) during MP bootup
++ * 9AP. We do not use mixed mode
++ *
++ * Pentium
++ * There is a marginal case where REP MOVS on 100MHz SMP
++ * machines with B stepping processors can fail. XXX should provide
++ * an L1cache=Writethrough or L1cache=off option.
++ *
++ * B stepping CPUs may hang. There are hardware work arounds
++ * for this. We warn about it in case your board doesn't have the work
++ * arounds. Basically thats so I can tell anyone with a B stepping
++ * CPU and SMP problems "tough".
++ *
++ * Specific items [From Pentium Processor Specification Update]
++ *
++ * 1AP. Linux doesn't use remote read
++ * 2AP. Linux doesn't trust APIC errors
++ * 3AP. We work around this
++ * 4AP. Linux never generated 3 interrupts of the same priority
++ * to cause a lost local interrupt.
++ * 5AP. Remote read is never used
++ * 6AP. not affected - worked around in hardware
++ * 7AP. not affected - worked around in hardware
++ * 8AP. worked around in hardware - we get explicit CS errors if not
++ * 9AP. only 'noapic' mode affected. Might generate spurious
++ * interrupts, we log only the first one and count the
++ * rest silently.
++ * 10AP. not affected - worked around in hardware
++ * 11AP. Linux reads the APIC between writes to avoid this, as per
++ * the documentation. Make sure you preserve this as it affects
++ * the C stepping chips too.
++ * 12AP. not affected - worked around in hardware
++ * 13AP. not affected - worked around in hardware
++ * 14AP. we always deassert INIT during bootup
++ * 15AP. not affected - worked around in hardware
++ * 16AP. not affected - worked around in hardware
++ * 17AP. not affected - worked around in hardware
++ * 18AP. not affected - worked around in hardware
++ * 19AP. not affected - worked around in BIOS
++ *
++ * If this sounds worrying believe me these bugs are either ___RARE___,
++ * or are signal timing bugs worked around in hardware and there's
++ * about nothing of note with C stepping upwards.
++ */
+
-+static inline void __normalize_time(time_t *sec, s64 *nsec)
++DEFINE_PER_CPU(struct tlb_state, cpu_tlbstate) ____cacheline_aligned = { &init_mm, 0, };
++
++/*
++ * the following functions deal with sending IPIs between CPUs.
++ *
++ * We use 'broadcast', CPU->CPU IPIs and self-IPIs too.
++ */
++
++static inline int __prepare_ICR (unsigned int shortcut, int vector)
+{
-+ while (*nsec >= NSEC_PER_SEC) {
-+ (*nsec) -= NSEC_PER_SEC;
-+ (*sec)++;
-+ }
-+ while (*nsec < 0) {
-+ (*nsec) += NSEC_PER_SEC;
-+ (*sec)--;
++ unsigned int icr = shortcut | APIC_DEST_LOGICAL;
++
++ switch (vector) {
++ default:
++ icr |= APIC_DM_FIXED | vector;
++ break;
++ case NMI_VECTOR:
++ icr |= APIC_DM_NMI;
++ break;
+ }
++ return icr;
+}
+
-+/* 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)
++static inline int __prepare_ICR2 (unsigned int mask)
+{
-+ independent_wallclock = 1;
-+ return 1;
++ return SET_APIC_DEST_FIELD(mask);
+}
-+__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)
++DECLARE_PER_CPU(int, ipi_to_irq[NR_IPIS]);
++
++static inline void __send_IPI_one(unsigned int cpu, int vector)
+{
-+ permitted_clock_jitter = simple_strtoul(str, NULL, 0);
-+ return 1;
++ int irq = per_cpu(ipi_to_irq, cpu)[vector];
++ BUG_ON(irq < 0);
++ notify_remote_via_irq(irq);
+}
-+__setup("permitted_clock_jitter=", __permitted_clock_jitter);
+
-+#ifndef CONFIG_X86
-+int tsc_disable __devinitdata = 0;
-+#endif
-+
-+static void delay_tsc(unsigned long loops)
++void __send_IPI_shortcut(unsigned int shortcut, int vector)
+{
-+ unsigned long bclock, now;
++ int cpu;
+
-+ rdtscl(bclock);
-+ do {
-+ rep_nop();
-+ rdtscl(now);
-+ } while ((now - bclock) < loops);
++ 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;
++ default:
++ printk("XXXXXX __send_IPI_shortcut %08x vector %d\n", shortcut,
++ vector);
++ break;
++ }
+}
+
-+struct timer_opts timer_tsc = {
-+ .name = "tsc",
-+ .delay = delay_tsc,
-+};
++void fastcall send_IPI_self(int vector)
++{
++ __send_IPI_shortcut(APIC_DEST_SELF, vector);
++}
+
+/*
-+ * Scale a 64-bit delta by scaling and multiplying by a 32-bit fraction,
-+ * yielding a 64-bit result.
++ * This is only used on smaller machines.
+ */
-+static inline u64 scale_delta(u64 delta, u32 mul_frac, int shift)
++void send_IPI_mask_bitmask(cpumask_t cpumask, int vector)
+{
-+ u64 product;
-+#ifdef __i386__
-+ u32 tmp1, tmp2;
-+#endif
++ unsigned long mask = cpus_addr(cpumask)[0];
++ unsigned long flags;
++ unsigned int cpu;
+
-+ if (shift < 0)
-+ delta >>= -shift;
-+ else
-+ delta <<= shift;
++ local_irq_save(flags);
++ WARN_ON(mask & ~cpus_addr(cpu_online_map)[0]);
+
-+#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 >> US_SCALE)), "2" (mul_frac) );
-+#else
-+ __asm__ (
-+ "mul %%rdx ; shrd $32,%%rdx,%%rax"
-+ : "=a" (product) : "0" (delta), "d" ((u64)mul_frac) );
-+#endif
++ for (cpu = 0; cpu < NR_CPUS; ++cpu) {
++ if (cpu_isset(cpu, cpumask)) {
++ __send_IPI_one(cpu, vector);
++ }
++ }
+
-+ return product;
++ local_irq_restore(flags);
+}
+
-+#if defined (__i386__)
-+int read_current_timer(unsigned long *timer_val)
++void send_IPI_mask_sequence(cpumask_t mask, int vector)
+{
-+ rdtscl(*timer_val);
-+ return 0;
-+}
-+#endif
+
-+void init_cpu_khz(void)
-+{
-+ u64 __cpu_khz = 1000000ULL << US_SCALE;
-+ struct vcpu_time_info *info;
-+ info = &HYPERVISOR_shared_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;
++ send_IPI_mask_bitmask(mask, vector);
+}
+
-+static u64 get_nsec_offset(struct shadow_time_info *shadow)
++#include <mach_ipi.h> /* must come after the send_IPI functions above for inlining */
++
++#if 0 /* 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>
++ */
++
++static cpumask_t flush_cpumask;
++static struct mm_struct * flush_mm;
++static unsigned long flush_va;
++static DEFINE_SPINLOCK(tlbstate_lock);
++#define FLUSH_ALL 0xffffffff
++
++/*
++ * We cannot call mmdrop() because we are in interrupt context,
++ * instead update mm->cpu_vm_mask.
++ *
++ * We need to reload %cr3 since the page tables may be going
++ * away from under us..
++ */
++static inline void leave_mm (unsigned long cpu)
+{
-+ u64 now, delta;
-+ rdtscll(now);
-+ delta = now - shadow->tsc_timestamp;
-+ return scale_delta(delta, shadow->tsc_to_nsec_mul, shadow->tsc_shift);
++ if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_OK)
++ BUG();
++ cpu_clear(cpu, per_cpu(cpu_tlbstate, cpu).active_mm->cpu_vm_mask);
++ load_cr3(swapper_pg_dir);
+}
+
-+static unsigned long get_usec_offset(struct shadow_time_info *shadow)
++/*
++ *
++ * 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 superflous
++ * tlb flush.
++ * 1a2) set cpu_tlbstate to TLBSTATE_OK
++ * Now the smp_invalidate_interrupt won't call leave_mm if cpu0
++ * was in lazy tlb mode.
++ * 1a3) update cpu_tlbstate[].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_tlbstate[].active_mm is correct, cpu0 already handles
++ * flush ipis.
++ * 1b1) set cpu_tlbstate 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_tlbstate 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.
++ */
++
++irqreturn_t smp_invalidate_interrupt(int irq, void *dev_id,
++ struct pt_regs *regs)
+{
-+ u64 now, delta;
-+ rdtscll(now);
-+ delta = now - shadow->tsc_timestamp;
-+ return scale_delta(delta, shadow->tsc_to_usec_mul, shadow->tsc_shift);
++ unsigned long cpu;
++
++ cpu = get_cpu();
++
++ if (!cpu_isset(cpu, 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 (flush_mm == per_cpu(cpu_tlbstate, cpu).active_mm) {
++ if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_OK) {
++ if (flush_va == FLUSH_ALL)
++ local_flush_tlb();
++ else
++ __flush_tlb_one(flush_va);
++ } else
++ leave_mm(cpu);
++ }
++ smp_mb__before_clear_bit();
++ cpu_clear(cpu, flush_cpumask);
++ smp_mb__after_clear_bit();
++out:
++ put_cpu_no_resched();
++
++ return IRQ_HANDLED;
+}
+
-+static void __update_wallclock(time_t sec, long nsec)
++static void flush_tlb_others(cpumask_t cpumask, struct mm_struct *mm,
++ unsigned long va)
+{
-+ long wtm_nsec, xtime_nsec;
-+ time_t wtm_sec, xtime_sec;
-+ u64 tmp, wc_nsec;
++ /*
++ * A couple of (to be removed) sanity checks:
++ *
++ * - current CPU must not be in mask
++ * - mask must exist :)
++ */
++ BUG_ON(cpus_empty(cpumask));
++ BUG_ON(cpu_isset(smp_processor_id(), cpumask));
++ BUG_ON(!mm);
+
-+ /* 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;
++ /* If a CPU which we ran on has gone down, OK. */
++ cpus_and(cpumask, cpumask, cpu_online_map);
++ if (cpus_empty(cpumask))
++ return;
+
-+ /* Split wallclock base into seconds and nanoseconds. */
-+ tmp = wc_nsec;
-+ xtime_nsec = do_div(tmp, 1000000000);
-+ xtime_sec = (time_t)tmp;
++ /*
++ * i'm not happy about this global shared spinlock in the
++ * MM hot path, but we'll see how contended it is.
++ * Temporarily this turns IRQs off, so that lockups are
++ * detected by the NMI watchdog.
++ */
++ spin_lock(&tlbstate_lock);
++
++ flush_mm = mm;
++ flush_va = va;
++#if NR_CPUS <= BITS_PER_LONG
++ atomic_set_mask(cpumask, &flush_cpumask);
++#else
++ {
++ int k;
++ unsigned long *flush_mask = (unsigned long *)&flush_cpumask;
++ unsigned long *cpu_mask = (unsigned long *)&cpumask;
++ for (k = 0; k < BITS_TO_LONGS(NR_CPUS); ++k)
++ atomic_set_mask(cpu_mask[k], &flush_mask[k]);
++ }
++#endif
++ /*
++ * We have to send the IPI only to
++ * CPUs affected.
++ */
++ send_IPI_mask(cpumask, INVALIDATE_TLB_VECTOR);
+
-+ wtm_sec = wall_to_monotonic.tv_sec + (xtime.tv_sec - xtime_sec);
-+ wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - xtime_nsec);
++ while (!cpus_empty(flush_cpumask))
++ /* nothing. lockup detection does not belong here */
++ mb();
+
-+ set_normalized_timespec(&xtime, xtime_sec, xtime_nsec);
-+ set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec);
++ flush_mm = NULL;
++ flush_va = 0;
++ spin_unlock(&tlbstate_lock);
++}
++
++void flush_tlb_current_task(void)
++{
++ struct mm_struct *mm = current->mm;
++ cpumask_t cpu_mask;
+
-+ ntp_clear();
++ 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();
+}
+
-+static void update_wallclock(void)
++void flush_tlb_mm (struct mm_struct * mm)
+{
-+ shared_info_t *s = HYPERVISOR_shared_info;
++ cpumask_t cpu_mask;
+
-+ 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));
++ preempt_disable();
++ cpu_mask = mm->cpu_vm_mask;
++ cpu_clear(smp_processor_id(), cpu_mask);
+
-+ if (!independent_wallclock)
-+ __update_wallclock(shadow_tv.tv_sec, shadow_tv.tv_nsec);
++ 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();
+}
+
-+/*
-+ * Reads a consistent set of time-base values from Xen, into a shadow data
-+ * area.
-+ */
-+static void get_time_values_from_xen(void)
++void flush_tlb_page(struct vm_area_struct * vma, unsigned long va)
+{
-+ shared_info_t *s = HYPERVISOR_shared_info;
-+ struct vcpu_time_info *src;
-+ struct shadow_time_info *dst;
++ struct mm_struct *mm = vma->vm_mm;
++ cpumask_t cpu_mask;
+
-+ src = &s->vcpu_info[smp_processor_id()].time;
-+ dst = &per_cpu(shadow_time, smp_processor_id());
++ preempt_disable();
++ cpu_mask = mm->cpu_vm_mask;
++ cpu_clear(smp_processor_id(), 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));
++ 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);
+
-+ dst->tsc_to_usec_mul = dst->tsc_to_nsec_mul / 1000;
++ preempt_enable();
+}
++EXPORT_SYMBOL(flush_tlb_page);
+
-+static inline int time_values_up_to_date(int cpu)
++static void do_flush_tlb_all(void* info)
+{
-+ struct vcpu_time_info *src;
-+ struct shadow_time_info *dst;
-+
-+ src = &HYPERVISOR_shared_info->vcpu_info[cpu].time;
-+ dst = &per_cpu(shadow_time, cpu);
++ unsigned long cpu = smp_processor_id();
+
-+ rmb();
-+ return (dst->version == src->version);
++ __flush_tlb_all();
++ if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_LAZY)
++ leave_mm(cpu);
+}
+
-+/*
-+ * 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)
++void flush_tlb_all(void)
+{
-+ 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;
++ on_each_cpu(do_flush_tlb_all, NULL, 1, 1);
+}
-+EXPORT_SYMBOL(rtc_cmos_read);
+
-+void rtc_cmos_write(unsigned char val, unsigned char addr)
++#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 */
++
++/*
++ * 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)
+{
-+ lock_cmos_prefix(addr);
-+ outb_p(addr, RTC_PORT(0));
-+ outb_p(val, RTC_PORT(1));
-+ lock_cmos_suffix(addr);
++ WARN_ON(cpu_is_offline(cpu));
++ send_IPI_mask(cpumask_of_cpu(cpu), RESCHEDULE_VECTOR);
+}
-+EXPORT_SYMBOL(rtc_cmos_write);
+
+/*
-+ * This version of gettimeofday has microsecond resolution
-+ * and better than microsecond precision on fast x86 machines with TSC.
++ * Structure and data for smp_call_function(). This is designed to minimise
++ * static memory requirements. It also looks cleaner.
+ */
-+void do_gettimeofday(struct timeval *tv)
-+{
-+ unsigned long seq;
-+ unsigned long usec, sec;
-+ unsigned long max_ntp_tick;
-+ s64 nsec;
-+ unsigned int cpu;
-+ struct shadow_time_info *shadow;
-+ u32 local_time_version;
++static DEFINE_SPINLOCK(call_lock);
+
-+ cpu = get_cpu();
-+ shadow = &per_cpu(shadow_time, cpu);
++struct call_data_struct {
++ void (*func) (void *info);
++ void *info;
++ atomic_t started;
++ atomic_t finished;
++ int wait;
++};
+
-+ do {
-+ unsigned long lost;
++void lock_ipi_call_lock(void)
++{
++ spin_lock_irq(&call_lock);
++}
+
-+ local_time_version = shadow->version;
-+ seq = read_seqbegin(&xtime_lock);
++void unlock_ipi_call_lock(void)
++{
++ spin_unlock_irq(&call_lock);
++}
+
-+ usec = get_usec_offset(shadow);
-+ lost = jiffies - wall_jiffies;
++static struct call_data_struct *call_data;
+
-+ /*
-+ * If time_adjust is negative then NTP is slowing the clock
-+ * so make sure not to go into next possible interval.
-+ * Better to lose some accuracy than have time go backwards..
-+ */
-+ if (unlikely(time_adjust < 0)) {
-+ max_ntp_tick = (USEC_PER_SEC / HZ) - tickadj;
-+ usec = min(usec, max_ntp_tick);
++/**
++ * 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.
++ */
++int smp_call_function (void (*func) (void *info), void *info, int nonatomic,
++ int wait)
++{
++ struct call_data_struct data;
++ int cpus;
+
-+ if (lost)
-+ usec += lost * max_ntp_tick;
-+ }
-+ else if (unlikely(lost))
-+ usec += lost * (USEC_PER_SEC / HZ);
++ /* Holding any lock stops cpus from going down. */
++ spin_lock(&call_lock);
++ cpus = num_online_cpus() - 1;
++ if (!cpus) {
++ spin_unlock(&call_lock);
++ return 0;
++ }
+
-+ sec = xtime.tv_sec;
-+ usec += (xtime.tv_nsec / NSEC_PER_USEC);
++ /* Can deadlock when called with interrupts disabled */
++ WARN_ON(irqs_disabled());
+
-+ nsec = shadow->system_timestamp - processed_system_time;
-+ __normalize_time(&sec, &nsec);
-+ usec += (long)nsec / NSEC_PER_USEC;
++ data.func = func;
++ data.info = info;
++ atomic_set(&data.started, 0);
++ data.wait = wait;
++ if (wait)
++ atomic_set(&data.finished, 0);
+
-+ 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();
-+ continue;
-+ }
-+ } while (read_seqretry(&xtime_lock, seq) ||
-+ (local_time_version != shadow->version));
++ call_data = &data;
++ mb();
++
++ /* Send a message to all other CPUs and wait for them to respond */
++ send_IPI_allbutself(CALL_FUNCTION_VECTOR);
+
-+ put_cpu();
++ /* Wait for response */
++ while (atomic_read(&data.started) != cpus)
++ barrier();
+
-+ while (usec >= USEC_PER_SEC) {
-+ usec -= USEC_PER_SEC;
-+ sec++;
-+ }
++ if (wait)
++ while (atomic_read(&data.finished) != cpus)
++ barrier();
++ spin_unlock(&call_lock);
+
-+ tv->tv_sec = sec;
-+ tv->tv_usec = usec;
++ return 0;
+}
++EXPORT_SYMBOL(smp_call_function);
+
-+EXPORT_SYMBOL(do_gettimeofday);
-+
-+int do_settimeofday(struct timespec *tv)
++static void stop_this_cpu (void * dummy)
+{
-+ time_t sec;
-+ s64 nsec;
-+ unsigned int cpu;
-+ struct shadow_time_info *shadow;
-+ dom0_op_t op;
-+
-+ if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC)
-+ return -EINVAL;
-+
-+ cpu = get_cpu();
-+ shadow = &per_cpu(shadow_time, cpu);
-+
-+ write_seqlock_irq(&xtime_lock);
-+
+ /*
-+ * 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.
++ * Remove this CPU:
+ */
-+ for (;;) {
-+ nsec = tv->tv_nsec - get_nsec_offset(shadow);
-+ if (time_values_up_to_date(cpu))
-+ break;
-+ get_time_values_from_xen();
-+ }
-+ sec = tv->tv_sec;
-+ __normalize_time(&sec, &nsec);
-+
-+ if (is_initial_xendomain() && !independent_wallclock) {
-+ op.cmd = DOM0_SETTIME;
-+ op.u.settime.secs = sec;
-+ op.u.settime.nsecs = nsec;
-+ op.u.settime.system_time = shadow->system_timestamp;
-+ HYPERVISOR_dom0_op(&op);
-+ update_wallclock();
-+ } else if (independent_wallclock) {
-+ nsec -= shadow->system_timestamp;
-+ __normalize_time(&sec, &nsec);
-+ __update_wallclock(sec, nsec);
-+ }
-+
-+ write_sequnlock_irq(&xtime_lock);
-+
-+ put_cpu();
-+
-+ clock_was_set();
-+ return 0;
++ cpu_clear(smp_processor_id(), cpu_online_map);
++ local_irq_disable();
++#if 0
++ disable_local_APIC();
++#endif
++ if (cpu_data[smp_processor_id()].hlt_works_ok)
++ for(;;) halt();
++ for (;;);
+}
+
-+EXPORT_SYMBOL(do_settimeofday);
++/*
++ * this function calls the 'stop' function on all other CPUs in the system.
++ */
+
-+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)
++void smp_send_stop(void)
+{
-+ time_t sec;
-+ s64 nsec;
-+ dom0_op_t op;
-+
-+ if (!ntp_synced() || independent_wallclock || !is_initial_xendomain())
-+ return;
++ smp_call_function(stop_this_cpu, NULL, 1, 0);
+
-+ write_seqlock_irq(&xtime_lock);
++ local_irq_disable();
++#if 0
++ disable_local_APIC();
++#endif
++ local_irq_enable();
++}
+
-+ sec = xtime.tv_sec;
-+ nsec = xtime.tv_nsec + ((jiffies - wall_jiffies) * (u64)NS_PER_TICK);
-+ __normalize_time(&sec, &nsec);
++/*
++ * Reschedule call back. Nothing to do,
++ * all the work is done automatically when
++ * we return from the interrupt.
++ */
++irqreturn_t smp_reschedule_interrupt(int irq, void *dev_id,
++ struct pt_regs *regs)
++{
++ return IRQ_HANDLED;
++}
+
-+ op.cmd = DOM0_SETTIME;
-+ op.u.settime.secs = sec;
-+ op.u.settime.nsecs = nsec;
-+ op.u.settime.system_time = processed_system_time;
-+ HYPERVISOR_dom0_op(&op);
++#include <linux/kallsyms.h>
++irqreturn_t smp_call_function_interrupt(int irq, void *dev_id,
++ struct pt_regs *regs)
++{
++ void (*func) (void *info) = call_data->func;
++ void *info = call_data->info;
++ int wait = call_data->wait;
+
-+ update_wallclock();
++ /*
++ * 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
++ */
++ irq_enter();
++ (*func)(info);
++ irq_exit();
+
-+ write_sequnlock_irq(&xtime_lock);
++ if (wait) {
++ mb();
++ atomic_inc(&call_data->finished);
++ }
+
-+ /* Once per minute. */
-+ mod_timer(&sync_xen_wallclock_timer, jiffies + 60*HZ);
++ return IRQ_HANDLED;
+}
+
-+static int set_rtc_mmss(unsigned long nowtime)
-+{
-+ int retval;
-+ unsigned long flags;
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/swiotlb.c b/arch/i386/kernel/swiotlb.c
+--- a/arch/i386/kernel/swiotlb.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/swiotlb.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,672 @@
++/*
++ * 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>
++ */
+
-+ if (independent_wallclock || !is_initial_xendomain())
-+ return 0;
++#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/interface/memory.h>
+
-+ /* gets recalled with irq locally disabled */
-+ 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);
++int swiotlb;
++EXPORT_SYMBOL(swiotlb);
+
-+ return retval;
-+}
++#define OFFSET(val,align) ((unsigned long)((val) & ( (align) - 1)))
+
-+/* 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.
++#define SG_ENT_PHYS_ADDRESS(sg) (page_to_bus((sg)->page) + (sg)->offset)
++
++/*
++ * 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.
+ */
-+unsigned long long monotonic_clock(void)
-+{
-+ int cpu = get_cpu();
-+ struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
-+ u64 time;
-+ u32 local_time_version;
++#define IO_TLB_SEGSIZE 128
+
-+ 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();
-+ barrier();
-+ } while (local_time_version != shadow->version);
++/*
++ * log of the size of each IO TLB slab. The number of slabs is command line
++ * controllable.
++ */
++#define IO_TLB_SHIFT 11
+
-+ put_cpu();
++/* Width of DMA addresses in the IO TLB. 31 bits is an aacraid limitation. */
++#define IO_TLB_DMA_BITS 31
+
-+ return time;
-+}
-+EXPORT_SYMBOL(monotonic_clock);
++int swiotlb_force;
++static char *iotlb_virt_start;
++static unsigned long iotlb_nslabs;
+
-+unsigned long long sched_clock(void)
-+{
-+ return monotonic_clock();
-+}
++/*
++ * 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(CONFIG_SMP) && defined(CONFIG_FRAME_POINTER)
-+unsigned long profile_pc(struct pt_regs *regs)
++/* Does the given dma address reside within the swiotlb aperture? */
++static inline int in_swiotlb_aperture(dma_addr_t dev_addr)
+{
-+ unsigned long pc = instruction_pointer(regs);
-+
-+#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;
++ unsigned long pfn = mfn_to_local_pfn(dev_addr >> PAGE_SHIFT);
++ return (pfn_valid(pfn)
++ && (pfn >= iotlb_pfn_start)
++ && (pfn < iotlb_pfn_end));
+}
-+EXPORT_SYMBOL(profile_pc);
-+#endif
+
-+irqreturn_t timer_interrupt(int irq, void *dev_id, struct pt_regs *regs)
-+{
-+ 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);
++/*
++ * When the IOMMU overflows we return a fallback buffer. This sets the size.
++ */
++static unsigned long io_tlb_overflow = 32*1024;
+
-+ write_seqlock(&xtime_lock);
++void *io_tlb_overflow_buffer;
+
-+ do {
-+ get_time_values_from_xen();
++/*
++ * 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;
+
-+ /* 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);
++/*
++ * 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;
+
-+ /*
-+ * 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));
++/*
++ * Protect the above data structures in the map and unmap calls
++ */
++static DEFINE_SPINLOCK(io_tlb_lock);
+
-+ 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));
++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);
++ /* Round up to power of two (xen_create_contiguous_region). */
++ while (iotlb_nslabs & (iotlb_nslabs-1))
++ iotlb_nslabs += iotlb_nslabs & ~(iotlb_nslabs-1);
+ }
++ 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? */
+
-+ /* System-wide jiffy work. */
-+ while (delta >= NS_PER_TICK) {
-+ delta -= NS_PER_TICK;
-+ processed_system_time += NS_PER_TICK;
-+ do_timer(regs);
-+ }
++/*
++ * 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;
+
-+ if (shadow_tv_version != HYPERVISOR_shared_info->wc_version) {
-+ update_wallclock();
-+ clock_was_set();
++ 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);
+ }
+
-+ write_sequnlock(&xtime_lock);
++ bytes = iotlb_nslabs * (1UL << IO_TLB_SHIFT);
+
+ /*
-+ * Account stolen ticks.
-+ * HACK: Passing NULL to account_steal_time()
-+ * ensures that the ticks are accounted as stolen.
++ * Get IO TLB memory from the low pages
+ */
-+ 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);
++ iotlb_virt_start = alloc_bootmem_low_pages(bytes);
++ if (!iotlb_virt_start)
++ panic("Cannot allocate SWIOTLB buffer!\n"
++ "Use dom0_mem Xen boot parameter to reserve\n"
++ "some DMA memory (e.g., dom0_mem=-128M).\n");
++
++ for (i = 0; i < iotlb_nslabs; i += IO_TLB_SEGSIZE) {
++ int rc = xen_create_contiguous_region(
++ (unsigned long)iotlb_virt_start + (i << IO_TLB_SHIFT),
++ get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT),
++ IO_TLB_DMA_BITS);
++ BUG_ON(rc);
+ }
+
+ /*
-+ * Account blocked ticks.
-+ * HACK: Passing idle_task to account_steal_time()
-+ * ensures that the ticks are accounted as idle/wait.
++ * Allocate and initialize the free list array. This array is used
++ * to find contiguous free memory regions of size up to IO_TLB_SEGSIZE.
+ */
-+ 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);
-+ }
++ 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));
+
-+ /* 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(regs))
-+ account_user_time(current, (cputime_t)delta_cpu);
-+ else
-+ account_system_time(current, HARDIRQ_OFFSET,
-+ (cputime_t)delta_cpu);
-+ }
++ /*
++ * Get the overflow emergency buffer
++ */
++ io_tlb_overflow_buffer = alloc_bootmem_low(io_tlb_overflow);
+
-+ /* Local timer processing (see update_process_times()). */
-+ run_local_timers();
-+ if (rcu_pending(cpu))
-+ rcu_check_callbacks(cpu, user_mode(regs));
-+ scheduler_tick();
-+ run_posix_cpu_timers(current);
++ iotlb_pfn_start = __pa(iotlb_virt_start) >> PAGE_SHIFT;
++ iotlb_pfn_end = iotlb_pfn_start + (bytes >> PAGE_SHIFT);
+
-+ return IRQ_HANDLED;
++ printk(KERN_INFO "Software IO TLB enabled: \n"
++ " Aperture: %lu megabytes\n"
++ " Kernel range: 0x%016lx - 0x%016lx\n",
++ bytes >> 20,
++ (unsigned long)iotlb_virt_start,
++ (unsigned long)iotlb_virt_start + bytes);
+}
+
-+static void init_missing_ticks_accounting(int cpu)
++void
++swiotlb_init(void)
+{
-+ struct vcpu_register_runstate_memory_area area;
-+ struct vcpu_runstate_info *runstate = &per_cpu(runstate, cpu);
++ long ram_end;
++ size_t defsz = 64 * (1 << 20); /* 64MB default size */
+
-+ memset(runstate, 0, sizeof(*runstate));
++ 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;
++ }
+
-+ area.addr.v = runstate;
-+ HYPERVISOR_vcpu_op(VCPUOP_register_runstate_memory_area, cpu, &area);
++ if (swiotlb)
++ swiotlb_init_with_default_size(defsz);
++ else
++ printk(KERN_INFO "Software IO TLB disabled\n");
++}
+
-+ per_cpu(processed_blocked_time, cpu) =
-+ runstate->time[RUNSTATE_blocked];
-+ per_cpu(processed_stolen_time, cpu) =
-+ runstate->time[RUNSTATE_runnable] +
-+ runstate->time[RUNSTATE_offline];
++/*
++ * 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)
++{
++ if (PageHighMem(buffer.page)) {
++ size_t len, bytes;
++ char *dev, *host, *kmp;
++ len = size;
++ while (len != 0) {
++ if (((bytes = len) + buffer.offset) > PAGE_SIZE)
++ bytes = PAGE_SIZE - buffer.offset;
++ kmp = kmap_atomic(buffer.page, KM_SWIOTLB);
++ 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_SWIOTLB);
++ 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);
++ }
+}
+
-+/* not static: needed by APM */
-+unsigned long get_cmos_time(void)
++/*
++ * 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 retval;
+ unsigned long flags;
++ char *dma_addr;
++ unsigned int nslots, stride, index, wrap;
++ int i;
+
-+ spin_lock_irqsave(&rtc_lock, flags);
-+
-+ if (efi_enabled)
-+ retval = efi_get_time();
++ /*
++ * 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
-+ retval = mach_get_cmos_time();
-+
-+ spin_unlock_irqrestore(&rtc_lock, flags);
-+
-+ return retval;
-+}
-+EXPORT_SYMBOL(get_cmos_time);
-+
-+static void sync_cmos_clock(unsigned long dummy);
-+
-+static DEFINE_TIMER(sync_cmos_timer, sync_cmos_clock, 0, 0);
++ stride = 1;
+
-+static void sync_cmos_clock(unsigned long dummy)
-+{
-+ struct timeval now, next;
-+ int fail = 1;
++ BUG_ON(!nslots);
+
+ /*
-+ * 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...
++ * Find suitable number of IO TLB entries size that will fit this
++ * request and allocate a buffer from that IO TLB pool.
+ */
-+ if (!ntp_synced())
-+ /*
-+ * Not synced, exit, do not restart a timer (if one is
-+ * running, let it run out).
-+ */
-+ return;
++ spin_lock_irqsave(&io_tlb_lock, flags);
++ {
++ wrap = index = ALIGN(io_tlb_index, stride);
+
-+ 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);
++ if (index >= iotlb_nslabs)
++ wrap = index = 0;
+
-+ next.tv_usec = USEC_AFTER - now.tv_usec;
-+ if (next.tv_usec <= 0)
-+ next.tv_usec += USEC_PER_SEC;
++ 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;
+
-+ if (!fail)
-+ next.tv_sec = 659;
-+ else
-+ next.tv_sec = 0;
++ 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);
+
-+ 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));
-+}
++ /*
++ * Update the indices to avoid searching in
++ * the next round.
++ */
++ io_tlb_index =
++ ((index + nslots) < iotlb_nslabs
++ ? (index + nslots) : 0);
+
-+void notify_arch_cmos_timer(void)
-+{
-+ mod_timer(&sync_cmos_timer, jiffies + 1);
-+ mod_timer(&sync_xen_wallclock_timer, jiffies + 1);
-+}
++ goto found;
++ }
++ index += stride;
++ if (index >= iotlb_nslabs)
++ index = 0;
++ } while (index != wrap);
+
-+static long clock_cmos_diff, sleep_start;
++ spin_unlock_irqrestore(&io_tlb_lock, flags);
++ return NULL;
++ }
++ found:
++ spin_unlock_irqrestore(&io_tlb_lock, flags);
+
-+static int timer_suspend(struct sys_device *dev, pm_message_t state)
-+{
+ /*
-+ * Estimate time zone so that set_time can update the clock
++ * 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.
+ */
-+ clock_cmos_diff = -get_cmos_time();
-+ clock_cmos_diff += get_seconds();
-+ sleep_start = get_cmos_time();
-+ return 0;
++ io_tlb_orig_addr[index] = buffer;
++ if ((dir == DMA_TO_DEVICE) || (dir == DMA_BIDIRECTIONAL))
++ __sync_single(buffer, dma_addr, size, DMA_TO_DEVICE);
++
++ return dma_addr;
+}
+
-+static int timer_resume(struct sys_device *dev)
++/*
++ * dma_addr is the kernel virtual address of the bounce buffer to unmap.
++ */
++static void
++unmap_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
+{
+ 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;
-+}
-+
-+static struct sysdev_class timer_sysclass = {
-+ .resume = timer_resume,
-+ .suspend = timer_suspend,
-+ set_kset_name("timer"),
-+};
++ 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 = io_tlb_orig_addr[index];
+
++ /*
++ * 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);
+
-+/* XXX this driverfs stuff should probably go elsewhere later -john */
-+static struct sys_device device_timer = {
-+ .id = 0,
-+ .cls = &timer_sysclass,
-+};
++ /*
++ * 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 int time_init_device(void)
++static void
++sync_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
+{
-+ int error = sysdev_class_register(&timer_sysclass);
-+ if (!error)
-+ error = sysdev_register(&device_timer);
-+ return error;
++ int index = (dma_addr - iotlb_virt_start) >> IO_TLB_SHIFT;
++ struct phys_addr buffer = io_tlb_orig_addr[index];
++ BUG_ON((dir != DMA_FROM_DEVICE) && (dir != DMA_TO_DEVICE));
++ __sync_single(buffer, dma_addr, size, dir);
+}
+
-+device_initcall(time_init_device);
-+
-+#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)
++static void
++swiotlb_full(struct device *dev, size_t size, int dir, int do_panic)
+{
-+ 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);
++ /*
++ * 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 ((hpet_enable() >= 0) && hpet_use_timer) {
-+ printk("Using HPET for base-timer\n");
++ 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");
+ }
-+ time_init_hook();
+}
-+#endif
-+
-+/* Dynamically-mapped IRQ. */
-+DEFINE_PER_CPU(int, timer_irq);
+
-+extern void (*late_time_init)(void);
-+static void setup_cpu0_timer_irq(void)
++/*
++ * 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)
+{
-+ 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);
-+}
++ dma_addr_t dev_addr = virt_to_bus(ptr);
++ void *map;
++ struct phys_addr buffer;
+
-+void __init time_init(void)
-+{
-+#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
-+ get_time_values_from_xen();
++ BUG_ON(dir == DMA_NONE);
+
-+ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
-+ per_cpu(processed_system_time, 0) = processed_system_time;
-+ init_missing_ticks_accounting(0);
++ /*
++ * 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(ptr, size) &&
++ !address_needs_mapping(hwdev, dev_addr))
++ return dev_addr;
+
-+ update_wallclock();
++ /*
++ * 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;
++ }
+
-+ init_cpu_khz();
-+ printk(KERN_INFO "Xen reported: %u.%03u MHz processor.\n",
-+ cpu_khz / 1000, cpu_khz % 1000);
++ dev_addr = virt_to_bus(map);
++ return dev_addr;
++}
+
-+#if defined(__x86_64__)
-+ vxtime.mode = VXTIME_TSC;
-+ vxtime.quot = (1000000L << US_SCALE) / vxtime_hz;
-+ vxtime.tsc_quot = (1000L << US_SCALE) / cpu_khz;
-+ sync_core();
-+ rdtscll(vxtime.last_tsc);
-+#endif
++/*
++ * 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);
++}
+
-+ /* Cannot request_irq() until kmem is initialised. */
-+ late_time_init = setup_cpu0_timer_irq;
++/*
++ * 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);
+}
+
-+/* Convert jiffies to system time. */
-+u64 jiffies_to_st(unsigned long j)
++void
++swiotlb_sync_single_for_device(struct device *hwdev, dma_addr_t dev_addr,
++ size_t size, int dir)
+{
-+ 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));
-+
-+ return st;
++ BUG_ON(dir == DMA_NONE);
++ if (in_swiotlb_aperture(dev_addr))
++ sync_single(hwdev, bus_to_virt(dev_addr), size, dir);
+}
-+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
++ * 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.
+ */
-+static void stop_hz_timer(void)
++int
++swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg, int nelems,
++ int dir)
+{
-+ unsigned int cpu = smp_processor_id();
-+ unsigned long j;
-+
-+ cpu_set(cpu, nohz_cpu_mask);
-+
-+ /* 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. */
++ struct phys_addr buffer;
++ dma_addr_t dev_addr;
++ char *map;
++ int i;
+
-+ smp_mb();
++ BUG_ON(dir == DMA_NONE);
+
-+ /* 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;
++ for (i = 0; i < nelems; i++, sg++) {
++ dev_addr = SG_ENT_PHYS_ADDRESS(sg);
++ if (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;
+ }
-+
-+ if (HYPERVISOR_set_timer_op(jiffies_to_st(j)) != 0)
-+ BUG();
++ return nelems;
+}
+
-+static void start_hz_timer(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)
+{
-+ cpu_clear(smp_processor_id(), nohz_cpu_mask);
-+}
++ int i;
+
-+void raw_safe_halt(void)
-+{
-+ stop_hz_timer();
-+ /* Blocking includes an implicit local_irq_enable(). */
-+ HYPERVISOR_block();
-+ start_hz_timer();
++ 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);
+}
-+EXPORT_SYMBOL(raw_safe_halt);
+
-+void halt(void)
++/*
++ * 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)
+{
-+ if (irqs_disabled())
-+ HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL);
++ int i;
++
++ BUG_ON(dir == DMA_NONE);
++
++ 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);
+}
-+EXPORT_SYMBOL(halt);
+
-+/* No locking required. We are only CPU running, and interrupts are off. */
-+void time_resume(void)
++void
++swiotlb_sync_sg_for_device(struct device *hwdev, struct scatterlist *sg,
++ int nelems, int dir)
+{
-+ init_cpu_khz();
-+
-+ get_time_values_from_xen();
++ int i;
+
-+ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
-+ per_cpu(processed_system_time, 0) = processed_system_time;
-+ init_missing_ticks_accounting(0);
++ BUG_ON(dir == DMA_NONE);
+
-+ update_wallclock();
++ 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);
+}
+
-+#ifdef CONFIG_SMP
-+static char timer_name[NR_CPUS][15];
-+
-+void local_setup_timer(unsigned int cpu)
++dma_addr_t
++swiotlb_map_page(struct device *hwdev, struct page *page,
++ unsigned long offset, size_t size,
++ enum dma_data_direction direction)
+{
-+ int seq;
++ struct phys_addr buffer;
++ dma_addr_t dev_addr;
++ char *map;
+
-+ BUG_ON(cpu == 0);
++ 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);
++ }
+
-+ 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));
++ return dev_addr;
++}
+
-+ sprintf(timer_name[cpu], "timer%d", cpu);
-+ per_cpu(timer_irq, cpu) =
-+ bind_virq_to_irqhandler(
-+ VIRQ_TIMER,
-+ cpu,
-+ timer_interrupt,
-+ SA_INTERRUPT,
-+ timer_name[cpu],
-+ NULL);
-+ BUG_ON(per_cpu(timer_irq, cpu) < 0);
++void
++swiotlb_unmap_page(struct device *hwdev, dma_addr_t dma_address,
++ size_t size, enum dma_data_direction direction)
++{
++ BUG_ON(!valid_dma_direction(direction));
++ if (in_swiotlb_aperture(dma_address))
++ unmap_single(hwdev, bus_to_virt(dma_address), size, direction);
+}
+
-+void local_teardown_timer(unsigned int cpu)
++int
++swiotlb_dma_mapping_error(dma_addr_t dma_addr)
+{
-+ BUG_ON(cpu == 0);
-+ unbind_from_irqhandler(per_cpu(timer_irq, cpu), NULL);
++ return (dma_addr == virt_to_bus(io_tlb_overflow_buffer));
+}
-+#endif
+
+/*
-+ * /proc/sys/xen: This really belongs in another file. It can stay here for
-+ * now however.
++ * 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.
+ */
-+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)
++int
++swiotlb_dma_supported (struct device *hwdev, u64 mask)
+{
-+ (void)register_sysctl_table(xen_table, 0);
-+ return 0;
++ return (mask >= ((1UL << IO_TLB_DMA_BITS) - 1));
+}
-+__initcall(xen_sysctl_init);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/traps.c linux-2.6.18-xen/arch/i386/kernel/traps.c
---- linux-2.6.18.3/arch/i386/kernel/traps.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/kernel/traps.c 2006-11-19 14:26:22.000000000 +0100
-@@ -642,18 +642,11 @@
++
++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_map_page);
++EXPORT_SYMBOL(swiotlb_unmap_page);
++EXPORT_SYMBOL(swiotlb_dma_mapping_error);
++EXPORT_SYMBOL(swiotlb_dma_supported);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/sysenter.c b/arch/i386/kernel/sysenter.c
+--- a/arch/i386/kernel/sysenter.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/sysenter.c 2007-03-14 10:55:14.000000000 +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()?
+@@ -48,6 +52,7 @@
- static void io_check_error(unsigned char reason, struct pt_regs * regs)
+ void enable_sep_cpu(void)
{
-- unsigned long i;
--
- printk(KERN_EMERG "NMI: IOCK error (debug interrupt?)\n");
- show_registers(regs);
++#ifndef CONFIG_X86_NO_TSS
+ int cpu = get_cpu();
+ struct tss_struct *tss = &per_cpu(init_tss, cpu);
- /* 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);
+@@ -62,6 +67,7 @@
+ wrmsr(MSR_IA32_SYSENTER_ESP, tss->esp1, 0);
+ wrmsr(MSR_IA32_SYSENTER_EIP, (unsigned long) sysenter_entry, 0);
+ put_cpu();
++#endif
}
- static void unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/traps-xen.c linux-2.6.18-xen/arch/i386/kernel/traps-xen.c
---- linux-2.6.18.3/arch/i386/kernel/traps-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/traps-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,1185 @@
+ /*
+@@ -76,6 +82,18 @@
+ {
+ syscall_page = (void *)get_zeroed_page(GFP_ATOMIC);
+
++#ifdef CONFIG_XEN
++ if (boot_cpu_has(X86_FEATURE_SEP)) {
++ struct callback_register sysenter = {
++ .type = CALLBACKTYPE_sysenter,
++ .address = { __KERNEL_CS, (unsigned long)sysenter_entry },
++ };
++
++ if (HYPERVISOR_callback_op(CALLBACKOP_register, &sysenter) < 0)
++ clear_bit(X86_FEATURE_SEP, boot_cpu_data.x86_capability);
++ }
++#endif
++
+ #ifdef CONFIG_COMPAT_VDSO
+ __set_fixmap(FIX_VDSO, __pa(syscall_page), PAGE_READONLY_EXEC);
+ printk("Compat vDSO mapped to %08lx.\n", __fix_to_virt(FIX_VDSO));
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/time-xen.c b/arch/i386/kernel/time-xen.c
+--- a/arch/i386/kernel/time-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/time-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,1065 @@
+/*
-+ * linux/arch/i386/traps.c
++ * linux/arch/i386/kernel/time.c
+ *
-+ * Copyright (C) 1991, 1992 Linus Torvalds
++ * Copyright (C) 1991, 1992, 1995 Linus Torvalds
+ *
-+ * Pentium III FXSR, SSE support
-+ * Gareth Hughes <gareth at valinux.com>, May 2000
++ * 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).
+ */
+
-+/*
-+ * 'Traps.c' handles hardware traps and faults after we have saved some
-+ * state in 'asm.s'.
-+ */
++#include <linux/errno.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>
++#include <linux/param.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/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>
-+#endif
++#include <linux/sysctl.h>
++#include <linux/percpu.h>
++#include <linux/kernel_stat.h>
++#include <linux/posix-timers.h>
+
-+#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/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/time.h>
++
++#include "mach_time.h"
++
++#include <linux/timex.h>
++
++#include <asm/hpet.h>
++
+#include <asm/arch_hooks.h>
-+#include <asm/kdebug.h>
+
-+#include <linux/module.h>
++#include <xen/evtchn.h>
++#include <xen/interface/vcpu.h>
+
-+#include "mach_traps.h"
++#include <asm/i8259.h>
+
-+asmlinkage int system_call(void);
++int pit_latch_buggy; /* extern */
+
-+struct desc_struct default_ldt[] = { { 0, 0 }, { 0, 0 }, { 0, 0 },
-+ { 0, 0 }, { 0, 0 } };
++#define USEC_PER_TICK (USEC_PER_SEC / HZ)
++#define NSEC_PER_TICK (NSEC_PER_SEC / HZ)
++#define FSEC_PER_TICK (FSEC_PER_SEC / HZ)
+
-+/* Do we ignore FPU interrupts ? */
-+char ignore_fpu_irq = 0;
++#define NS_SCALE 10 /* 2^10, carefully chosen */
++#define US_SCALE 32 /* 2^32, arbitralrily chosen */
+
-+#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
++unsigned int cpu_khz; /* Detected as we calibrate the TSC */
++EXPORT_SYMBOL(cpu_khz);
+
-+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);
++DEFINE_SPINLOCK(rtc_lock);
++EXPORT_SYMBOL(rtc_lock);
+
-+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);
++extern struct init_timer_opts timer_tsc_init;
++extern struct timer_opts timer_tsc;
++#define timer_none timer_tsc
+
-+int register_die_notifier(struct notifier_block *nb)
++/* 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;
++
++/* 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 inline void __normalize_time(time_t *sec, s64 *nsec)
++{
++ while (*nsec >= NSEC_PER_SEC) {
++ (*nsec) -= NSEC_PER_SEC;
++ (*sec)++;
++ }
++ 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)
+{
-+ vmalloc_sync_all();
-+ return atomic_notifier_chain_register(&i386die_chain, nb);
++ independent_wallclock = 1;
++ return 1;
+}
-+EXPORT_SYMBOL(register_die_notifier); /* used modular by kdb */
++__setup("independent_wallclock", __independent_wallclock);
+
-+int unregister_die_notifier(struct notifier_block *nb)
++/* 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)
+{
-+ return atomic_notifier_chain_unregister(&i386die_chain, nb);
++ permitted_clock_jitter = simple_strtoul(str, NULL, 0);
++ return 1;
+}
-+EXPORT_SYMBOL(unregister_die_notifier); /* used modular by kdb */
++__setup("permitted_clock_jitter=", __permitted_clock_jitter);
+
-+static inline int valid_stack_ptr(struct thread_info *tinfo, void *p)
++#ifndef CONFIG_X86
++int tsc_disable __devinitdata = 0;
++#endif
++
++static void delay_tsc(unsigned long loops)
+{
-+ return p > (void *)tinfo &&
-+ p < (void *)tinfo + THREAD_SIZE - 3;
++ unsigned long bclock, now;
++
++ rdtscl(bclock);
++ do {
++ rep_nop();
++ rdtscl(now);
++ } while ((now - bclock) < loops);
+}
+
++struct timer_opts timer_tsc = {
++ .name = "tsc",
++ .delay = delay_tsc,
++};
++
+/*
-+ * Print one address/symbol entries per line.
++ * Scale a 64-bit delta by scaling and multiplying by a 32-bit fraction,
++ * yielding a 64-bit result.
+ */
-+static inline void print_addr_and_symbol(unsigned long addr, char *log_lvl)
++static inline u64 scale_delta(u64 delta, u32 mul_frac, int shift)
+{
-+ printk(" [<%08lx>] ", addr);
++ u64 product;
++ u32 tmp1, tmp2;
+
-+ print_symbol("%s\n", addr);
++ if (shift < 0)
++ delta >>= -shift;
++ else
++ delta <<= shift;
++
++ __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 >> US_SCALE)), "2" (mul_frac) );
++
++ return product;
+}
+
-+static inline unsigned long print_context_stack(struct thread_info *tinfo,
-+ unsigned long *stack, unsigned long ebp,
-+ char *log_lvl)
++int read_current_timer(unsigned long *timer_val)
+{
-+ unsigned long addr;
++ rdtscl(*timer_val);
++ return 0;
++}
+
-+#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);
-+ }
-+#endif
-+ return ebp;
++void init_cpu_khz(void)
++{
++ u64 __cpu_khz = 1000000ULL << US_SCALE;
++ struct vcpu_time_info *info;
++ info = &HYPERVISOR_shared_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;
+}
+
-+static asmlinkage int
-+show_trace_unwind(struct unwind_frame_info *info, void *log_lvl)
++static u64 get_nsec_offset(struct shadow_time_info *shadow)
+{
-+ int n = 0;
++ u64 now, delta;
++ rdtscll(now);
++ delta = now - shadow->tsc_timestamp;
++ return scale_delta(delta, shadow->tsc_to_nsec_mul, shadow->tsc_shift);
++}
+
-+ 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;
++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 void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
-+ unsigned long *stack, char *log_lvl)
++static void __update_wallclock(time_t sec, long nsec)
+{
-+ unsigned long ebp;
++ long wtm_nsec, xtime_nsec;
++ time_t wtm_sec, xtime_sec;
++ u64 tmp, wc_nsec;
+
-+ if (!task)
-+ task = current;
++ /* Adjust wall-clock time base based on jiffies ticks. */
++ wc_nsec = processed_system_time;
++ wc_nsec += sec * (u64)NSEC_PER_SEC;
++ wc_nsec += nsec;
+
-+ if (call_trace >= 0) {
-+ int unw_ret = 0;
-+ struct unwind_frame_info info;
-+
-+ 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");
-+ }
++ /* Split wallclock base into seconds and nanoseconds. */
++ tmp = wc_nsec;
++ xtime_nsec = do_div(tmp, 1000000000);
++ xtime_sec = (time_t)tmp;
+
-+ 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;
-+ }
++ wtm_sec = wall_to_monotonic.tv_sec + (xtime.tv_sec - xtime_sec);
++ wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - xtime_nsec);
+
-+ 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);
-+ }
-+}
++ set_normalized_timespec(&xtime, xtime_sec, xtime_nsec);
++ set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec);
+
-+void show_trace(struct task_struct *task, struct pt_regs *regs, unsigned long * stack)
-+{
-+ show_trace_log_lvl(task, regs, stack, "");
++ ntp_clear();
+}
+
-+static void show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
-+ unsigned long *esp, char *log_lvl)
++static void update_wallclock(void)
+{
-+ unsigned long *stack;
-+ int i;
-+
-+ if (esp == NULL) {
-+ if (task)
-+ esp = (unsigned long*)task->thread.esp;
-+ else
-+ esp = (unsigned long *)&esp;
-+ }
++ shared_info_t *s = HYPERVISOR_shared_info;
+
-+ 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);
-+}
++ 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));
+
-+void show_stack(struct task_struct *task, unsigned long *esp)
-+{
-+ printk(" ");
-+ show_stack_log_lvl(task, NULL, esp, "");
++ if (!independent_wallclock)
++ __update_wallclock(shadow_tv.tv_sec, shadow_tv.tv_nsec);
+}
+
+/*
-+ * The architecture-independent dump_stack generator
++ * Reads a consistent set of time-base values from Xen, into a shadow data
++ * area.
+ */
-+void dump_stack(void)
++static void get_time_values_from_xen(void)
+{
-+ unsigned long stack;
-+
-+ show_trace(current, NULL, &stack);
-+}
++ shared_info_t *s = HYPERVISOR_shared_info;
++ struct vcpu_time_info *src;
++ struct shadow_time_info *dst;
+
-+EXPORT_SYMBOL(dump_stack);
++ src = &s->vcpu_info[smp_processor_id()].time;
++ dst = &per_cpu(shadow_time, smp_processor_id());
+
-+void show_registers(struct pt_regs *regs)
-+{
-+ int i;
-+ int in_kernel = 1;
-+ unsigned long esp;
-+ unsigned short ss;
++ 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));
+
-+ 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;
++ dst->tsc_to_usec_mul = dst->tsc_to_nsec_mul / 1000;
++}
+
-+ printk("\n" KERN_EMERG "Stack: ");
-+ show_stack_log_lvl(NULL, regs, (unsigned long *)esp, KERN_EMERG);
++static inline int time_values_up_to_date(int cpu)
++{
++ struct vcpu_time_info *src;
++ struct shadow_time_info *dst;
+
-+ printk(KERN_EMERG "Code: ");
++ src = &HYPERVISOR_shared_info->vcpu_info[cpu].time;
++ dst = &per_cpu(shadow_time, cpu);
+
-+ eip = (u8 __user *)regs->eip - 43;
-+ for (i = 0; i < 64; i++, eip++) {
-+ unsigned char c;
++ rmb();
++ return (dst->version == src->version);
++}
+
-+ 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");
-+}
++/*
++ * 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);
+
-+static void handle_BUG(struct pt_regs *regs)
++/* Routines for accessing the CMOS RAM/RTC. */
++unsigned char rtc_cmos_read(unsigned char addr)
+{
-+ 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;
-+
-+ 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>";
++ 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);
+
-+ 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");
++void rtc_cmos_write(unsigned char val, unsigned char addr)
++{
++ 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);
+
-+/* 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)
++/*
++ * This version of gettimeofday has microsecond resolution
++ * and better than microsecond precision on fast x86 machines with TSC.
++ */
++void do_gettimeofday(struct timeval *tv)
+{
-+ 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 seq;
++ unsigned long usec, sec;
++ unsigned long max_ntp_tick;
++ s64 nsec;
++ unsigned int cpu;
++ struct shadow_time_info *shadow;
++ u32 local_time_version;
+
-+ oops_enter();
++ cpu = get_cpu();
++ shadow = &per_cpu(shadow_time, cpu);
+
-+ 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);
++ do {
++ local_time_version = shadow->version;
++ seq = read_seqbegin(&xtime_lock);
+
-+ if (++die.lock_owner_depth < 3) {
-+ int nl = 0;
-+ unsigned long esp;
-+ unsigned short ss;
++ usec = get_usec_offset(shadow);
+
-+ 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);
++ /*
++ * If time_adjust is negative then NTP is slowing the clock
++ * so make sure not to go into next possible interval.
++ * Better to lose some accuracy than have time go backwards..
++ */
++ if (unlikely(time_adjust < 0)) {
++ max_ntp_tick = (USEC_PER_SEC / HZ) - tickadj;
++ usec = min(usec, max_ntp_tick);
+ }
-+ else
-+ regs = NULL;
-+ } else
-+ printk(KERN_EMERG "Recursive die() failure, output suppressed\n");
+
-+ bust_spinlocks(0);
-+ die.lock_owner = -1;
-+ spin_unlock_irqrestore(&die.lock, flags);
++ sec = xtime.tv_sec;
++ usec += (xtime.tv_nsec / NSEC_PER_USEC);
+
-+ if (!regs)
-+ return;
++ nsec = shadow->system_timestamp - processed_system_time;
++ __normalize_time(&sec, &nsec);
++ usec += (long)nsec / NSEC_PER_USEC;
+
-+ if (kexec_should_crash(current))
-+ crash_kexec(regs);
++ 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();
++ continue;
++ }
++ } while (read_seqretry(&xtime_lock, seq) ||
++ (local_time_version != shadow->version));
+
-+ if (in_interrupt())
-+ panic("Fatal exception in interrupt");
++ put_cpu();
+
-+ if (panic_on_oops)
-+ panic("Fatal exception");
++ while (usec >= USEC_PER_SEC) {
++ usec -= USEC_PER_SEC;
++ sec++;
++ }
+
-+ oops_exit();
-+ do_exit(SIGSEGV);
++ tv->tv_sec = sec;
++ tv->tv_usec = usec;
+}
+
-+static inline void die_if_kernel(const char * str, struct pt_regs * regs, long err)
-+{
-+ if (!user_mode_vm(regs))
-+ die(str, regs, err);
-+}
++EXPORT_SYMBOL(do_gettimeofday);
+
-+static void __kprobes do_trap(int trapnr, int signr, char *str, int vm86,
-+ struct pt_regs * regs, long error_code,
-+ siginfo_t *info)
++int do_settimeofday(struct timespec *tv)
+{
-+ struct task_struct *tsk = current;
-+ tsk->thread.error_code = error_code;
-+ tsk->thread.trap_no = trapnr;
-+
-+ if (regs->eflags & VM_MASK) {
-+ if (vm86)
-+ goto vm86_trap;
-+ goto trap_signal;
-+ }
++ time_t sec;
++ s64 nsec;
++ unsigned int cpu;
++ struct shadow_time_info *shadow;
++ dom0_op_t op;
+
-+ if (!user_mode(regs))
-+ goto kernel_trap;
++ if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC)
++ return -EINVAL;
+
-+ trap_signal: {
-+ if (info)
-+ force_sig_info(signr, info, tsk);
-+ else
-+ force_sig(signr, tsk);
-+ return;
-+ }
++ cpu = get_cpu();
++ shadow = &per_cpu(shadow_time, cpu);
+
-+ kernel_trap: {
-+ if (!fixup_exception(regs))
-+ die(str, regs, error_code);
-+ return;
-+ }
++ write_seqlock_irq(&xtime_lock);
+
-+ vm86_trap: {
-+ int ret = handle_vm86_trap((struct kernel_vm86_regs *) regs, error_code, trapnr);
-+ if (ret) goto trap_signal;
-+ return;
++ /*
++ * 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();
+ }
-+}
-+
-+#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); \
-+}
++ sec = tv->tv_sec;
++ __normalize_time(&sec, &nsec);
+
-+#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); \
-+}
++ if (is_initial_xendomain() && !independent_wallclock) {
++ op.cmd = DOM0_SETTIME;
++ op.u.settime.secs = sec;
++ op.u.settime.nsecs = nsec;
++ op.u.settime.system_time = shadow->system_timestamp;
++ HYPERVISOR_dom0_op(&op);
++ update_wallclock();
++ } else if (independent_wallclock) {
++ nsec -= shadow->system_timestamp;
++ __normalize_time(&sec, &nsec);
++ __update_wallclock(sec, nsec);
++ }
+
-+#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); \
-+}
++ write_sequnlock_irq(&xtime_lock);
+
-+#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); \
++ put_cpu();
++
++ clock_was_set();
++ return 0;
+}
+
-+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)
++EXPORT_SYMBOL(do_settimeofday);
+
-+fastcall void __kprobes do_general_protection(struct pt_regs * regs,
-+ long error_code)
++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)
+{
-+ current->thread.error_code = error_code;
-+ current->thread.trap_no = 13;
++ time_t sec;
++ s64 nsec;
++ dom0_op_t op;
+
-+ if (regs->eflags & VM_MASK)
-+ goto gp_in_vm86;
++ if (!ntp_synced() || independent_wallclock || !is_initial_xendomain())
++ return;
+
-+ if (!user_mode(regs))
-+ goto gp_in_kernel;
++ write_seqlock_irq(&xtime_lock);
+
-+ current->thread.error_code = error_code;
-+ current->thread.trap_no = 13;
-+ force_sig(SIGSEGV, current);
-+ return;
++ sec = xtime.tv_sec;
++ nsec = xtime.tv_nsec;
++ __normalize_time(&sec, &nsec);
+
-+gp_in_vm86:
-+ local_irq_enable();
-+ handle_vm86_fault((struct kernel_vm86_regs *) regs, error_code);
-+ return;
++ op.cmd = DOM0_SETTIME;
++ op.u.settime.secs = sec;
++ op.u.settime.nsecs = nsec;
++ op.u.settime.system_time = processed_system_time;
++ HYPERVISOR_dom0_op(&op);
+
-+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);
-+ }
++ update_wallclock();
++
++ write_sequnlock_irq(&xtime_lock);
++
++ /* Once per minute. */
++ mod_timer(&sync_xen_wallclock_timer, jiffies + 60*HZ);
+}
+
-+static void mem_parity_error(unsigned char reason, struct pt_regs * regs)
++static int set_rtc_mmss(unsigned long nowtime)
+{
-+ 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");
++ int retval;
++ unsigned long flags;
+
-+ /* Clear and disable the memory parity error line. */
-+ clear_mem_error(reason);
++ if (independent_wallclock || !is_initial_xendomain())
++ return 0;
++
++ /* gets recalled with irq locally disabled */
++ spin_lock_irqsave(&rtc_lock, flags);
++ retval = set_wallclock(nowtime);
++ spin_unlock_irqrestore(&rtc_lock, flags);
++
++ return retval;
+}
+
-+static void io_check_error(unsigned char reason, struct pt_regs * regs)
++/* 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)
+{
-+ printk(KERN_EMERG "NMI: IOCK error (debug interrupt?)\n");
-+ show_registers(regs);
++ int cpu = get_cpu();
++ struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
++ u64 time;
++ u32 local_time_version;
+
-+ /* Re-enable the IOCK line, wait for a few seconds */
-+ clear_io_check_error(reason);
++ 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();
++ barrier();
++ } while (local_time_version != shadow->version);
++
++ put_cpu();
++
++ return time;
++}
++EXPORT_SYMBOL(monotonic_clock);
++
++unsigned long long sched_clock(void)
++{
++ return monotonic_clock();
+}
+
-+static void unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
++unsigned long profile_pc(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;
++ unsigned long pc = instruction_pointer(regs);
++
++#ifdef CONFIG_SMP
++ if (!user_mode_vm(regs) && in_lock_functions(pc)) {
++#ifdef CONFIG_FRAME_POINTER
++ return *(unsigned long *)(regs->ebp + 4);
++#else
++ unsigned long *sp;
++ if ((regs->xcs & 3) == 0)
++ sp = (unsigned long *)®s->esp;
++ else
++ sp = (unsigned long *)regs->esp;
++ /* Return address is either directly at stack pointer
++ or above a saved eflags. Eflags has bits 22-31 zero,
++ kernel addresses don't. */
++ if (sp[0] >> 22)
++ return sp[0];
++ if (sp[1] >> 22)
++ return sp[1];
++#endif
+ }
+#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");
++ return pc;
+}
++EXPORT_SYMBOL(profile_pc);
+
-+static DEFINE_SPINLOCK(nmi_print_lock);
-+
-+void die_nmi (struct pt_regs *regs, const char *msg)
++irqreturn_t timer_interrupt(int irq, void *dev_id)
+{
-+ if (notify_die(DIE_NMIWATCHDOG, msg, regs, 0, 2, SIGINT) ==
-+ NOTIFY_STOP)
-+ return;
-+
-+ 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);
++ 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);
+
-+ /* 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);
-+ }
++ write_seqlock(&xtime_lock);
+
-+ do_exit(SIGSEGV);
-+}
++ do {
++ get_time_values_from_xen();
+
-+static void default_do_nmi(struct pt_regs * regs)
-+{
-+ unsigned char reason = 0;
++ /* 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);
+
-+ /* 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.
++ * Obtain a consistent snapshot of stolen/blocked cycles. We
++ * can use state_entry_time to detect if we get preempted here.
+ */
-+ if (nmi_watchdog) {
-+ nmi_watchdog_tick(regs);
-+ return;
-+ }
-+#endif
-+ unknown_nmi_error(reason, regs);
-+ return;
++ 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 ((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 (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();
-+}
+
-+static int dummy_nmi_callback(struct pt_regs * regs, int cpu)
-+{
-+ return 0;
-+}
-+
-+static nmi_callback_t nmi_callback = dummy_nmi_callback;
-+
-+fastcall void do_nmi(struct pt_regs * regs, long error_code)
-+{
-+ int cpu;
++ /* System-wide jiffy work. */
++ while (delta >= NS_PER_TICK) {
++ delta -= NS_PER_TICK;
++ processed_system_time += NS_PER_TICK;
++ do_timer(1);
++ }
+
-+ nmi_enter();
++ if (shadow_tv_version != HYPERVISOR_shared_info->wc_version) {
++ update_wallclock();
++ clock_was_set();
++ }
+
-+ cpu = smp_processor_id();
++ write_sequnlock(&xtime_lock);
+
-+ ++nmi_count(cpu);
++ /*
++ * 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);
++ }
+
-+ if (!rcu_dereference(nmi_callback)(regs, cpu))
-+ default_do_nmi(regs);
++ /*
++ * 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);
++ }
+
-+ nmi_exit();
-+}
++ /* 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(get_irq_regs()))
++ account_user_time(current, (cputime_t)delta_cpu);
++ else
++ account_system_time(current, HARDIRQ_OFFSET,
++ (cputime_t)delta_cpu);
++ }
+
-+void set_nmi_callback(nmi_callback_t callback)
-+{
-+ vmalloc_sync_all();
-+ rcu_assign_pointer(nmi_callback, callback);
-+}
-+EXPORT_SYMBOL_GPL(set_nmi_callback);
++ /* Local timer processing (see update_process_times()). */
++ run_local_timers();
++ if (rcu_pending(cpu))
++ rcu_check_callbacks(cpu, user_mode(get_irq_regs()));
++ scheduler_tick();
++ run_posix_cpu_timers(current);
+
-+void unset_nmi_callback(void)
-+{
-+ nmi_callback = dummy_nmi_callback;
++ return IRQ_HANDLED;
+}
-+EXPORT_SYMBOL_GPL(unset_nmi_callback);
+
-+#ifdef CONFIG_KPROBES
-+fastcall void __kprobes do_int3(struct pt_regs *regs, long error_code)
++static void init_missing_ticks_accounting(int cpu)
+{
-+ 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);
++ struct vcpu_register_runstate_memory_area area;
++ struct vcpu_runstate_info *runstate = &per_cpu(runstate, cpu);
++
++ memset(runstate, 0, sizeof(*runstate));
++
++ area.addr.v = runstate;
++ HYPERVISOR_vcpu_op(VCPUOP_register_runstate_memory_area, cpu, &area);
++
++ per_cpu(processed_blocked_time, cpu) =
++ runstate->time[RUNSTATE_blocked];
++ per_cpu(processed_stolen_time, cpu) =
++ runstate->time[RUNSTATE_runnable] +
++ runstate->time[RUNSTATE_offline];
+}
-+#endif
+
-+/*
-+ * 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)
-+ */
-+fastcall void __kprobes do_debug(struct pt_regs * regs, long error_code)
++/* not static: needed by APM */
++unsigned long get_cmos_time(void)
+{
-+ unsigned int condition;
-+ struct task_struct *tsk = current;
++ unsigned long retval;
++ unsigned long flags;
+
-+ get_debugreg(condition, 6);
++ spin_lock_irqsave(&rtc_lock, flags);
+
-+ 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();
++ retval = get_wallclock();
+
-+ /* 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;
-+ }
++ spin_unlock_irqrestore(&rtc_lock, flags);
+
-+ if (regs->eflags & VM_MASK)
-+ goto debug_vm86;
++ return retval;
++}
++EXPORT_SYMBOL(get_cmos_time);
+
-+ /* Save debug status register where ptrace can see it */
-+ tsk->thread.debugreg[6] = condition;
++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)
++{
++ struct timeval now, next;
++ int fail = 1;
+
+ /*
-+ * Single-stepping through TF: make sure we ignore any events in
-+ * kernel space (but re-enable TF when returning to user mode).
++ * 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 (condition & DR_STEP) {
++ if (!ntp_synced())
+ /*
-+ * We already checked v86 mode above, so we can
-+ * check for kernel mode by just checking the CPL
-+ * of CS.
++ * Not synced, exit, do not restart a timer (if one is
++ * running, let it run out).
+ */
-+ if (!user_mode(regs))
-+ goto clear_TF_reenable;
-+ }
++ return;
+
-+ /* Ok, finally something we can handle */
-+ send_sigtrap(tsk, regs, error_code);
++ 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);
+
-+ /* Disable additional traps. They'll be re-enabled when
-+ * the signal is delivered.
-+ */
-+clear_dr7:
-+ set_debugreg(0, 7);
-+ return;
++ next.tv_usec = USEC_AFTER - now.tv_usec;
++ if (next.tv_usec <= 0)
++ next.tv_usec += USEC_PER_SEC;
+
-+debug_vm86:
-+ handle_vm86_trap((struct kernel_vm86_regs *) regs, error_code, 1);
-+ return;
++ if (!fail)
++ next.tv_sec = 659;
++ else
++ next.tv_sec = 0;
+
-+clear_TF_reenable:
-+ set_tsk_thread_flag(tsk, TIF_SINGLESTEP);
-+ regs->eflags &= ~TF_MASK;
-+ return;
++ 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));
+}
+
-+/*
-+ * 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)
++void notify_arch_cmos_timer(void)
+{
-+ struct task_struct * task;
-+ siginfo_t info;
-+ unsigned short cwd, swd;
++ mod_timer(&sync_cmos_timer, jiffies + 1);
++ mod_timer(&sync_xen_wallclock_timer, jiffies + 1);
++}
+
++static long clock_cmos_diff;
++static unsigned long sleep_start;
++
++static int timer_suspend(struct sys_device *dev, pm_message_t state)
++{
+ /*
-+ * 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
++ * Estimate time zone so that set_time can update the clock
+ */
-+ 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;
++ unsigned long ctime = get_cmos_time();
++
++ clock_cmos_diff = -ctime;
++ clock_cmos_diff += get_seconds();
++ sleep_start = ctime;
++ return 0;
++}
++
++static int timer_resume(struct sys_device *dev)
++{
++ unsigned long flags;
++ unsigned long sec;
++ unsigned long ctime = get_cmos_time();
++ long sleep_length = (ctime - sleep_start) * HZ;
++
++ if (sleep_length < 0) {
++ printk(KERN_WARNING "CMOS clock skew detected in timer resume!\n");
++ /* The time after the resume must not be earlier than the time
++ * before the suspend or some nasty things will happen
++ */
++ sleep_length = 0;
++ ctime = sleep_start;
+ }
-+ force_sig_info(SIGFPE, &info, task);
++
++#ifdef CONFIG_HPET_TIMER
++ if (is_hpet_enabled())
++ hpet_reenable();
++#endif
++ sec = ctime + clock_cmos_diff;
++ write_seqlock_irqsave(&xtime_lock, flags);
++ xtime.tv_sec = sec;
++ xtime.tv_nsec = 0;
++ jiffies_64 += sleep_length;
++ write_sequnlock_irqrestore(&xtime_lock, flags);
++ touch_softlockup_watchdog();
++ return 0;
+}
+
-+fastcall void do_coprocessor_error(struct pt_regs * regs, long error_code)
++static struct sysdev_class timer_sysclass = {
++ .resume = timer_resume,
++ .suspend = timer_suspend,
++ set_kset_name("timer"),
++};
++
++
++/* XXX this driverfs stuff should probably go elsewhere later -john */
++static struct sys_device device_timer = {
++ .id = 0,
++ .cls = &timer_sysclass,
++};
++
++static int time_init_device(void)
+{
-+ ignore_fpu_irq = 1;
-+ math_error((void __user *)regs->eip);
++ int error = sysdev_class_register(&timer_sysclass);
++ if (!error)
++ error = sysdev_register(&device_timer);
++ return error;
+}
+
-+static void simd_math_error(void __user *eip)
++device_initcall(time_init_device);
++
++#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)
+{
-+ struct task_struct * task;
-+ siginfo_t info;
-+ unsigned short mxcsr;
++ 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);
+
-+ /*
-+ * 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;
++ if ((hpet_enable() >= 0) && hpet_use_timer) {
++ printk("Using HPET for base-timer\n");
+ }
-+ force_sig_info(SIGFPE, &info, task);
++
++ do_time_init();
+}
++#endif
+
-+fastcall void do_simd_coprocessor_error(struct pt_regs * regs,
-+ long error_code)
++/* Dynamically-mapped IRQ. */
++DEFINE_PER_CPU(int, timer_irq);
++
++extern void (*late_time_init)(void);
++static void setup_cpu0_timer_irq(void)
+{
-+ 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);
-+ }
++ 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);
+}
+
-+#ifndef CONFIG_XEN
-+fastcall void do_spurious_interrupt_bug(struct pt_regs * regs,
-+ long error_code)
++void __init time_init(void)
+{
-+#if 0
-+ /* No need to warn about this any longer. */
-+ printk("Ignoring P6 Local APIC Spurious Interrupt Bug...\n");
++#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
++ get_time_values_from_xen();
++
++ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
++ per_cpu(processed_system_time, 0) = processed_system_time;
++ init_missing_ticks_accounting(0);
++
++ update_wallclock();
++
++ init_cpu_khz();
++ printk(KERN_INFO "Xen reported: %u.%03u MHz processor.\n",
++ cpu_khz / 1000, cpu_khz % 1000);
++
++ /* Cannot request_irq() until kmem is initialised. */
++ late_time_init = setup_cpu0_timer_irq;
+}
+
-+fastcall void setup_x86_bogus_stack(unsigned char * stk)
++/* Convert jiffies to system time. */
++u64 jiffies_to_st(unsigned long j)
+{
-+ 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;
++ 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));
++
++ return st;
+}
-+#endif
++EXPORT_SYMBOL(jiffies_to_st);
+
+/*
-+ * '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).
++ * 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
+ */
-+asmlinkage void math_state_restore(struct pt_regs regs)
++static void stop_hz_timer(void)
+{
-+ struct thread_info *thread = current_thread_info();
-+ struct task_struct *tsk = thread->task;
++ unsigned int cpu = smp_processor_id();
++ unsigned long j;
+
-+ /* 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() */
-+}
++ cpu_set(cpu, nohz_cpu_mask);
+
-+#ifndef CONFIG_MATH_EMULATION
++ /* 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. */
+
-+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();
-+}
++ smp_mb();
+
-+#endif /* CONFIG_MATH_EMULATION */
++ /* 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;
++ }
+
-+#ifdef CONFIG_X86_F00F_BUG
-+void __init trap_init_f00f_bug(void)
-+{
-+ __set_fixmap(FIX_F00F_IDT, __pa(&idt_table), PAGE_KERNEL_RO);
++ if (HYPERVISOR_set_timer_op(jiffies_to_st(j)) != 0)
++ BUG();
++}
+
-+ /*
-+ * 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);
++static void start_hz_timer(void)
++{
++ cpu_clear(smp_processor_id(), nohz_cpu_mask);
+}
-+#endif
+
++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);
+
-+/*
-+ * 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 halt(void)
++{
++ if (irqs_disabled())
++ HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL);
++}
++EXPORT_SYMBOL(halt);
+
-+void __init trap_init(void)
++/* No locking required. We are only CPU running, and interrupts are off. */
++void time_resume(void)
+{
-+ HYPERVISOR_set_trap_table(trap_table);
++ init_cpu_khz();
+
-+ 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);
-+ };
++ get_time_values_from_xen();
+
-+ 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");
-+ }
++ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
++ per_cpu(processed_system_time, 0) = processed_system_time;
++ init_missing_ticks_accounting(0);
+
-+ /*
-+ * Should be a barrier for any external CPU state.
-+ */
-+ cpu_init();
++ update_wallclock();
+}
+
-+void smp_trap_init(trap_info_t *trap_ctxt)
++#ifdef CONFIG_SMP
++static char timer_name[NR_CPUS][15];
++
++void local_setup_timer(unsigned int cpu)
+{
-+ trap_info_t *t = trap_table;
++ int seq;
+
-+ 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(cpu == 0);
++
++ 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));
++
++ sprintf(timer_name[cpu], "timer%d", cpu);
++ per_cpu(timer_irq, cpu) =
++ bind_virq_to_irqhandler(
++ VIRQ_TIMER,
++ cpu,
++ timer_interrupt,
++ SA_INTERRUPT,
++ timer_name[cpu],
++ NULL);
++ BUG_ON(per_cpu(timer_irq, cpu) < 0);
+}
+
-+static int __init kstack_setup(char *s)
++void local_teardown_timer(unsigned int cpu)
+{
-+ kstack_depth_to_print = simple_strtoul(s, NULL, 0);
-+ return 1;
++ BUG_ON(cpu == 0);
++ unbind_from_irqhandler(per_cpu(timer_irq, cpu), NULL);
+}
-+__setup("kstack=", kstack_setup);
++#endif
+
-+#ifdef CONFIG_STACK_UNWIND
-+static int __init call_trace_setup(char *s)
++/*
++ * /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)
+{
-+ 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;
++ (void)register_sysctl_table(xen_table, 0);
++ return 0;
+}
-+__setup("call_trace=", call_trace_setup);
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/tsc.c linux-2.6.18-xen/arch/i386/kernel/tsc.c
---- linux-2.6.18.3/arch/i386/kernel/tsc.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/kernel/tsc.c 2006-11-19 14:26:22.000000000 +0100
-@@ -101,6 +101,7 @@
- return (cyc * cyc2ns_scale) >> CYC2NS_SCALE_FACTOR;
- }
-
-+#ifndef CONFIG_XEN
- /*
- * Scheduler clock - returns current time in nanosec units.
- */
-@@ -124,6 +125,7 @@
- /* return the value in ns */
- return cycles_2_ns(this_offset);
- }
-+#endif
-
- static unsigned long calculate_cpu_khz(void)
- {
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/vm86.c linux-2.6.18-xen/arch/i386/kernel/vm86.c
---- linux-2.6.18.3/arch/i386/kernel/vm86.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/kernel/vm86.c 2006-11-19 14:26:22.000000000 +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
- 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
- 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)
++__initcall(xen_sysctl_init);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c
+--- a/arch/i386/kernel/traps.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/traps.c 2007-03-14 10:55:14.000000000 +0100
+@@ -646,18 +646,11 @@
+ static __kprobes void
+ io_check_error(unsigned char reason, struct pt_regs * regs)
{
-+#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
+- unsigned long i;
+-
+ printk(KERN_EMERG "NMI: IOCK error (debug interrupt?)\n");
+ show_registers(regs);
- tsk->thread.screen_bitmap = info->screen_bitmap;
- if (info->flags & VM86_SCREEN_BITMAP)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/vmlinux.lds.S linux-2.6.18-xen/arch/i386/kernel/vmlinux.lds.S
---- linux-2.6.18.3/arch/i386/kernel/vmlinux.lds.S 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/kernel/vmlinux.lds.S 2006-11-19 14:26:22.000000000 +0100
-@@ -13,6 +13,12 @@
- 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__ */
-+}
- SECTIONS
- {
- . = __KERNEL_START;
-@@ -26,7 +32,7 @@
- KPROBES_TEXT
- *(.fixup)
- *(.gnu.warning)
-- } = 0x9090
-+ } :text = 0x9090
-
- _etext = .; /* End of text section */
-
-@@ -48,7 +54,7 @@
- .data : AT(ADDR(.data) - LOAD_OFFSET) { /* Data */
- *(.data)
- CONSTRUCTORS
-- }
-+ } :data
-
- . = ALIGN(4096);
- __nosave_begin = .;
-@@ -184,4 +190,6 @@
- STABS_DEBUG
+ /* 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);
+ }
- DWARF_DEBUG
+ static __kprobes void
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/traps-xen.c b/arch/i386/kernel/traps-xen.c
+--- a/arch/i386/kernel/traps-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/traps-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,1132 @@
++/*
++ * 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>
++#include <linux/uaccess.h>
++#include <linux/nmi.h>
++#include <linux/bug.h>
+
-+ NOTES
- }
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/kernel/vsyscall-note-xen.S linux-2.6.18-xen/arch/i386/kernel/vsyscall-note-xen.S
---- linux-2.6.18.3/arch/i386/kernel/vsyscall-note-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/kernel/vsyscall-note-xen.S 2006-11-19 14:26:22.000000000 +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.
-+ */
++#ifdef CONFIG_EISA
++#include <linux/ioport.h>
++#include <linux/eisa.h>
++#endif
+
-+#include "vsyscall-note.S"
++#ifdef CONFIG_MCA
++#include <linux/mca.h>
++#endif
+
-+/*
-+ * 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
++#include <asm/processor.h>
++#include <asm/system.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 <asm/stacktrace.h>
+
-+NOTE_KERNELCAP_BEGIN(1, 1)
-+NOTE_KERNELCAP(1, "nosegneg") /* Change 1 back to 0 when glibc is fixed! */
-+NOTE_KERNELCAP_END
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/lib/delay.c linux-2.6.18-xen/arch/i386/lib/delay.c
---- linux-2.6.18.3/arch/i386/lib/delay.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/lib/delay.c 2006-11-19 14:26:22.000000000 +0100
-@@ -61,6 +61,7 @@
- delay_fn = delay_tsc;
- }
-
-+#ifndef CONFIG_X86_XEN
- int read_current_timer(unsigned long *timer_val)
- {
- if (delay_fn == delay_tsc) {
-@@ -69,7 +70,7 @@
- }
- return -1;
- }
--
-+#endif
- void __delay(unsigned long loops)
- {
- delay_fn(loops);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/mach-xen/irqflags.c linux-2.6.18-xen/arch/i386/mach-xen/irqflags.c
---- linux-2.6.18.3/arch/i386/mach-xen/irqflags.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/mach-xen/irqflags.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,99 @@
+#include <linux/module.h>
-+#include <linux/smp.h>
-+#include <asm/irqflags.h>
-+#include <asm/hypervisor.h>
-+
-+/* interrupt control.. */
+
-+/*
-+ * 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.
-+ */
++#include "mach_traps.h"
+
-+unsigned long __raw_local_save_flags(void)
-+{
-+ struct vcpu_info *_vcpu;
-+ unsigned long flags;
++int panic_on_unrecovered_nmi;
+
-+ preempt_disable();
-+ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
-+ flags = _vcpu->evtchn_upcall_mask;
-+ preempt_enable();
++asmlinkage int system_call(void);
+
-+ return flags;
-+}
-+EXPORT_SYMBOL(__raw_local_save_flags);
++/* Do we ignore FPU interrupts ? */
++char ignore_fpu_irq = 0;
+
-+void raw_local_irq_restore(unsigned long flags)
-+{
-+ struct vcpu_info *_vcpu;
++#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
+
-+ preempt_disable();
-+ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
-+ if ((_vcpu->evtchn_upcall_mask = flags) == 0) {
-+ barrier(); /* unmask then check (avoid races) */
-+ if (unlikely(_vcpu->evtchn_upcall_pending))
-+ force_evtchn_callback();
-+ preempt_enable();
-+ } else
-+ preempt_enable_no_resched();
++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);
+
-+}
-+EXPORT_SYMBOL(raw_local_irq_restore);
++int kstack_depth_to_print = 24;
++ATOMIC_NOTIFIER_HEAD(i386die_chain);
+
-+void raw_local_irq_disable(void)
++int register_die_notifier(struct notifier_block *nb)
+{
-+ struct vcpu_info *_vcpu;
-+
-+ preempt_disable();
-+ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
-+ _vcpu->evtchn_upcall_mask = 1;
-+ preempt_enable_no_resched();
++ vmalloc_sync_all();
++ return atomic_notifier_chain_register(&i386die_chain, nb);
+}
-+EXPORT_SYMBOL(raw_local_irq_disable);
++EXPORT_SYMBOL(register_die_notifier); /* used modular by kdb */
+
-+void raw_local_irq_enable(void)
++int unregister_die_notifier(struct notifier_block *nb)
+{
-+ struct vcpu_info *_vcpu;
-+
-+ preempt_disable();
-+ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
-+ _vcpu->evtchn_upcall_mask = 0;
-+ barrier(); /* unmask then check (avoid races) */
-+ if (unlikely(_vcpu->evtchn_upcall_pending))
-+ force_evtchn_callback();
-+ preempt_enable();
++ return atomic_notifier_chain_unregister(&i386die_chain, nb);
+}
-+EXPORT_SYMBOL(raw_local_irq_enable);
++EXPORT_SYMBOL(unregister_die_notifier); /* used modular by kdb */
+
-+/* Cannot use preempt_enable() here as we would recurse in preempt_sched(). */
-+int raw_irqs_disabled(void)
++static inline int valid_stack_ptr(struct thread_info *tinfo, void *p)
+{
-+ struct vcpu_info *_vcpu;
-+ int disabled;
-+
-+ preempt_disable();
-+ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
-+ disabled = (_vcpu->evtchn_upcall_mask != 0);
-+ preempt_enable_no_resched();
-+ return disabled;
++ return p > (void *)tinfo &&
++ p < (void *)tinfo + THREAD_SIZE - 3;
+}
-+EXPORT_SYMBOL(raw_irqs_disabled);
+
-+unsigned long __raw_local_irq_save(void)
++static inline unsigned long print_context_stack(struct thread_info *tinfo,
++ unsigned long *stack, unsigned long ebp,
++ struct stacktrace_ops *ops, void *data)
+{
-+ struct vcpu_info *_vcpu;
-+ unsigned long flags;
-+
-+ preempt_disable();
-+ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
-+ flags = _vcpu->evtchn_upcall_mask;
-+ _vcpu->evtchn_upcall_mask = 1;
-+ preempt_enable_no_resched();
++ unsigned long addr;
+
-+ return flags;
++#ifdef CONFIG_FRAME_POINTER
++ while (valid_stack_ptr(tinfo, (void *)ebp)) {
++ unsigned long new_ebp;
++ addr = *(unsigned long *)(ebp + 4);
++ ops->address(data, addr);
++ /*
++ * break out of recursive entries (such as
++ * end_of_stack_stop_unwind_function). Also,
++ * we can never allow a frame pointer to
++ * move downwards!
++ */
++ new_ebp = *(unsigned long *)ebp;
++ if (new_ebp <= ebp)
++ break;
++ ebp = new_ebp;
++ }
++#else
++ while (valid_stack_ptr(tinfo, stack)) {
++ addr = *stack++;
++ if (__kernel_text_address(addr))
++ ops->address(data, addr);
++ }
++#endif
++ return ebp;
+}
-+EXPORT_SYMBOL(__raw_local_irq_save);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/mach-xen/Makefile linux-2.6.18-xen/arch/i386/mach-xen/Makefile
---- linux-2.6.18.3/arch/i386/mach-xen/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/mach-xen/Makefile 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,5 @@
-+#
-+# Makefile for the linux kernel.
-+#
+
-+obj-y := setup.o irqflags.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/mach-xen/setup.c linux-2.6.18-xen/arch/i386/mach-xen/setup.c
---- linux-2.6.18.3/arch/i386/mach-xen/setup.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/mach-xen/setup.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,154 @@
-+/*
-+ * Machine specific setup for generic
-+ */
++#define MSG(msg) ops->warning(data, msg)
+
-+#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>
++void dump_trace(struct task_struct *task, struct pt_regs *regs,
++ unsigned long *stack,
++ struct stacktrace_ops *ops, void *data)
++{
++ unsigned long ebp = 0;
+
-+#include <xen/interface/callback.h>
-+#include <xen/interface/memory.h>
++ if (!task)
++ task = current;
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+#define DEFAULT_SEND_IPI (1)
-+#else
-+#define DEFAULT_SEND_IPI (0)
++ if (!stack) {
++ unsigned long dummy;
++ stack = &dummy;
++ if (task && task != current)
++ stack = (unsigned long *)task->thread.esp;
++ }
++
++#ifdef CONFIG_FRAME_POINTER
++ if (!ebp) {
++ 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;
++ }
++ }
+#endif
+
-+int no_broadcast=DEFAULT_SEND_IPI;
++ while (1) {
++ struct thread_info *context;
++ context = (struct thread_info *)
++ ((unsigned long)stack & (~(THREAD_SIZE - 1)));
++ ebp = print_context_stack(context, stack, ebp, ops, data);
++ /* Should be after the line below, but somewhere
++ in early boot context comes out corrupted and we
++ can't reference it -AK */
++ if (ops->stack(data, "IRQ") < 0)
++ break;
++ stack = (unsigned long*)context->previous_esp;
++ if (!stack)
++ break;
++ touch_nmi_watchdog();
++ }
++}
++EXPORT_SYMBOL(dump_trace);
+
-+static __init int no_ipi_broadcast(char *str)
++static void
++print_trace_warning_symbol(void *data, char *msg, unsigned long symbol)
+{
-+ get_option(&str, &no_broadcast);
-+ printk ("Using %s mode\n", no_broadcast ? "No IPI Broadcast" :
-+ "IPI Broadcast");
-+ return 1;
++ printk(data);
++ print_symbol(msg, symbol);
++ printk("\n");
+}
+
-+__setup("no_ipi_broadcast", no_ipi_broadcast);
++static void print_trace_warning(void *data, char *msg)
++{
++ printk("%s%s\n", (char *)data, msg);
++}
+
-+static int __init print_ipi_mode(void)
++static int print_trace_stack(void *data, char *name)
+{
-+ printk ("Using IPI %s mode\n", no_broadcast ? "No-Shortcut" :
-+ "Shortcut");
+ return 0;
+}
+
-+late_initcall(print_ipi_mode);
-+
-+/**
-+ * 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.
-+ **/
-+
-+char * __init machine_specific_memory_setup(void)
++/*
++ * Print one address/symbol entries per line.
++ */
++static void print_trace_address(void *data, unsigned long addr)
+{
-+ 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 = PFN_PHYS(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);
-+
-+ sanitize_e820_map(map, (char *)&memmap.nr_entries);
-+
-+ BUG_ON(copy_e820_map(map, (char)memmap.nr_entries) < 0);
-+
-+ return "Xen";
++ printk("%s [<%08lx>] ", (char *)data, addr);
++ print_symbol("%s\n", addr);
+}
+
-+extern void hypervisor_callback(void);
-+extern void failsafe_callback(void);
-+extern void nmi(void);
-+
-+unsigned long *machine_to_phys_mapping;
-+EXPORT_SYMBOL(machine_to_phys_mapping);
-+unsigned int machine_to_phys_order;
-+EXPORT_SYMBOL(machine_to_phys_order);
++static struct stacktrace_ops print_trace_ops = {
++ .warning = print_trace_warning,
++ .warning_symbol = print_trace_warning_symbol,
++ .stack = print_trace_stack,
++ .address = print_trace_address,
++};
+
-+void __init machine_specific_arch_setup(void)
++static void
++show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
++ unsigned long * stack, char *log_lvl)
+{
-+ int ret;
-+ struct xen_machphys_mapping mapping;
-+ unsigned long machine_to_phys_nr_ents;
-+ struct xen_platform_parameters pp;
-+ struct callback_register event = {
-+ .type = CALLBACKTYPE_event,
-+ .address = { __KERNEL_CS, (unsigned long)hypervisor_callback },
-+ };
-+ struct callback_register failsafe = {
-+ .type = CALLBACKTYPE_failsafe,
-+ .address = { __KERNEL_CS, (unsigned long)failsafe_callback },
-+ };
-+ struct callback_register nmi_cb = {
-+ .type = CALLBACKTYPE_nmi,
-+ .address = { __KERNEL_CS, (unsigned long)nmi },
-+ };
++ dump_trace(task, regs, stack, &print_trace_ops, log_lvl);
++ printk("%s =======================\n", log_lvl);
++}
+
-+ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &event);
-+ if (ret == 0)
-+ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &failsafe);
-+ if (ret == -ENOSYS)
-+ ret = HYPERVISOR_set_callbacks(
-+ event.address.cs, event.address.eip,
-+ failsafe.address.cs, failsafe.address.eip);
-+ BUG_ON(ret);
++void show_trace(struct task_struct *task, struct pt_regs *regs,
++ unsigned long * stack)
++{
++ show_trace_log_lvl(task, regs, stack, "");
++}
+
-+ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &nmi_cb);
-+ if (ret == -ENOSYS) {
-+ struct xennmi_callback cb;
++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;
+
-+ cb.handler_address = nmi_cb.address.eip;
-+ HYPERVISOR_nmi_op(XENNMI_register_callback, &cb);
++ if (esp == NULL) {
++ if (task)
++ esp = (unsigned long*)task->thread.esp;
++ else
++ esp = (unsigned long *)&esp;
+ }
+
-+ if (HYPERVISOR_xen_version(XENVER_platform_parameters,
-+ &pp) == 0)
-+ set_fixaddr_top(pp.virt_start - PAGE_SIZE);
-+
-+ 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;
++ 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++);
+ }
-+ while ((1UL << machine_to_phys_order) < machine_to_phys_nr_ents )
-+ machine_to_phys_order++;
++ printk("\n%sCall Trace:\n", log_lvl);
++ show_trace_log_lvl(task, regs, esp, log_lvl);
+}
+
-+/**
-+ * pre_setup_arch_hook - hook called prior to any setup_arch() execution
-+ *
-+ * Description:
-+ * generally used to activate any machine specific identification
-+ * routines that may be needed before setup_arch() runs. On VISWS
-+ * this is used to get the board revision and type.
-+ **/
-+void __init pre_setup_arch_hook(void)
++void show_stack(struct task_struct *task, unsigned long *esp)
+{
++ printk(" ");
++ show_stack_log_lvl(task, NULL, esp, "");
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/Makefile linux-2.6.18-xen/arch/i386/Makefile
---- linux-2.6.18.3/arch/i386/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/Makefile 2006-11-19 14:26:21.000000000 +0100
-@@ -48,6 +48,11 @@
-
- CFLAGS += $(cflags-y)
-
-+cppflags-$(CONFIG_XEN) += \
-+ -D__XEN_INTERFACE_VERSION__=$(CONFIG_XEN_INTERFACE_VERSION)
+
-+CPPFLAGS += $(cppflags-y)
++/*
++ * The architecture-independent dump_stack generator
++ */
++void dump_stack(void)
++{
++ unsigned long stack;
+
- # Default subarch .c files
- mcore-y := mach-default
-
-@@ -71,6 +76,10 @@
- mflags-$(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-$(CONFIG_X86_XEN) := mach-xen
++ show_trace(current, NULL, &stack);
++}
+
- # generic subarchitecture
- mflags-$(CONFIG_X86_GENERICARCH) := -Iinclude/asm-i386/mach-generic
- mcore-$(CONFIG_X86_GENERICARCH) := mach-default
-@@ -105,6 +114,19 @@
- PHONY += zImage bzImage compressed zlilo bzlilo \
- zdisk bzdisk fdimage fdimage144 fdimage288 isoimage install
-
-+ifdef CONFIG_XEN
-+CPPFLAGS := -Iinclude$(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
++EXPORT_SYMBOL(dump_stack);
+
-+vmlinuz: vmlinux
-+ $(Q)$(MAKE) $(build)=$(boot) $@
++void show_registers(struct pt_regs *regs)
++{
++ int i;
++ int in_kernel = 1;
++ unsigned long esp;
++ unsigned short ss;
+
-+install:
-+ $(Q)$(MAKE) $(build)=$(boot) XENGUEST=$(XENGUEST) $@
-+else
- all: bzImage
-
- # KBUILD_IMAGE specify target image being built
-@@ -127,6 +149,7 @@
-
- install:
- $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(KBUILD_IMAGE) install
-+endif
-
- archclean:
- $(Q)$(MAKE) $(clean)=arch/i386/boot
-@@ -145,3 +168,4 @@
- CLEAN_FILES += arch/$(ARCH)/boot/fdimage \
- arch/$(ARCH)/boot/image.iso \
- arch/$(ARCH)/boot/mtools.conf
-+CLEAN_FILES += vmlinuz vmlinux-stripped
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/mm/fault-xen.c linux-2.6.18-xen/arch/i386/mm/fault-xen.c
---- linux-2.6.18.3/arch/i386/mm/fault-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/mm/fault-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,770 @@
-+/*
-+ * linux/arch/i386/mm/fault.c
-+ *
-+ * Copyright (C) 1995 Linus Torvalds
-+ */
++ 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\n"
++ KERN_EMERG "EIP: %04x:[<%08lx>] %s VLI\n"
++ KERN_EMERG "EFLAGS: %08lx (%s %.*s)\n",
++ smp_processor_id(), 0xffff & regs->xcs, regs->eip,
++ print_tainted(), regs->eflags, init_utsname()->release,
++ (int)strcspn(init_utsname()->version, " "),
++ init_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 *eip;
++ int code_bytes = 64;
++ unsigned char c;
+
-+#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>
++ printk("\n" KERN_EMERG "Stack: ");
++ show_stack_log_lvl(NULL, regs, (unsigned long *)esp, KERN_EMERG);
+
-+#include <asm/system.h>
-+#include <asm/uaccess.h>
-+#include <asm/desc.h>
-+#include <asm/kdebug.h>
++ printk(KERN_EMERG "Code: ");
+
-+extern void die(const char *,struct pt_regs *,long);
++ eip = (u8 *)regs->eip - 43;
++ if (eip < (u8 *)PAGE_OFFSET ||
++ probe_kernel_address(eip, c)) {
++ /* try starting at EIP */
++ eip = (u8 *)regs->eip;
++ code_bytes = 32;
++ }
++ for (i = 0; i < code_bytes; i++, eip++) {
++ if (eip < (u8 *)PAGE_OFFSET ||
++ probe_kernel_address(eip, c)) {
++ printk(" Bad EIP value.");
++ break;
++ }
++ if (eip == (u8 *)regs->eip)
++ printk("<%02x> ", c);
++ else
++ printk("%02x ", c);
++ }
++ }
++ printk("\n");
++}
+
-+#ifdef CONFIG_KPROBES
-+ATOMIC_NOTIFIER_HEAD(notify_page_fault_chain);
-+int register_page_fault_notifier(struct notifier_block *nb)
++int is_valid_bugaddr(unsigned long eip)
+{
-+ vmalloc_sync_all();
-+ return atomic_notifier_chain_register(¬ify_page_fault_chain, nb);
-+}
++ unsigned short ud2;
+
-+int unregister_page_fault_notifier(struct notifier_block *nb)
-+{
-+ return atomic_notifier_chain_unregister(¬ify_page_fault_chain, nb);
-+}
++ if (eip < PAGE_OFFSET)
++ return 0;
++ if (probe_kernel_address((unsigned short *)eip, ud2))
++ return 0;
+
-+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;
++ return ud2 == 0x0b0f;
+}
-+#endif
+
+/*
-+ * Unlock any spinlocks which will prevent us from getting the
-+ * message out
++ * This is gone through when something in the kernel has done something bad and
++ * is about to be terminated.
+ */
-+void bust_spinlocks(int yes)
++void die(const char * str, struct pt_regs * regs, long err)
+{
-+ int loglevel_save = console_loglevel;
++ static struct {
++ spinlock_t lock;
++ u32 lock_owner;
++ int lock_owner_depth;
++ } die = {
++ .lock = __SPIN_LOCK_UNLOCKED(die.lock),
++ .lock_owner = -1,
++ .lock_owner_depth = 0
++ };
++ static int die_counter;
++ unsigned long flags;
+
-+ if (yes) {
-+ oops_in_progress = 1;
-+ return;
++ oops_enter();
++
++ 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);
+ }
-+#ifdef CONFIG_VT
-+ unblank_screen();
++ else
++ local_save_flags(flags);
++
++ if (++die.lock_owner_depth < 3) {
++ int nl = 0;
++ unsigned long esp;
++ unsigned short ss;
++
++ report_bug(regs->eip);
++
++ printk(KERN_EMERG "%s: %04lx [#%d]\n", str, err & 0xffff, ++die_counter);
++#ifdef CONFIG_PREEMPT
++ printk(KERN_EMERG "PREEMPT ");
++ nl = 1;
+#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;
++#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");
++
++ bust_spinlocks(0);
++ die.lock_owner = -1;
++ spin_unlock_irqrestore(&die.lock, flags);
++
++ if (!regs)
++ return;
++
++ if (kexec_should_crash(current))
++ crash_kexec(regs);
++
++ if (in_interrupt())
++ panic("Fatal exception in interrupt");
++
++ if (panic_on_oops)
++ panic("Fatal exception");
++
++ oops_exit();
++ do_exit(SIGSEGV);
+}
+
-+/*
-+ * 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.
-+ */
-+static inline unsigned long get_segment_eip(struct pt_regs *regs,
-+ unsigned long *eip_limit)
++static inline void die_if_kernel(const char * str, struct pt_regs * regs, long err)
+{
-+ unsigned long eip = regs->eip;
-+ unsigned seg = regs->xcs & 0xffff;
-+ u32 seg_ar, seg_limit, base, *desc;
++ if (!user_mode_vm(regs))
++ die(str, regs, err);
++}
+
-+ /* Unlikely, but must come before segment checks. */
-+ if (unlikely(regs->eflags & VM_MASK)) {
-+ base = seg << 4;
-+ *eip_limit = base + 0xffff;
-+ return base + (eip & 0xffff);
++static void __kprobes do_trap(int trapnr, int signr, char *str, int vm86,
++ 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 (regs->eflags & VM_MASK) {
++ if (vm86)
++ goto vm86_trap;
++ goto trap_signal;
+ }
+
-+ /* The standard kernel/user address space limit. */
-+ *eip_limit = (seg & 2) ? USER_DS.seg : KERNEL_DS.seg;
++ if (!user_mode(regs))
++ goto kernel_trap;
+
-+ /* By far the most common cases. */
-+ if (likely(seg == __USER_CS || seg == GET_KERNEL_CS()))
-+ return eip;
++ trap_signal: {
++ if (info)
++ force_sig_info(signr, info, tsk);
++ else
++ force_sig(signr, tsk);
++ return;
++ }
+
-+ /* 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. */
++ kernel_trap: {
++ if (!fixup_exception(regs))
++ die(str, regs, error_code);
++ return;
+ }
+
-+ /* 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);
++ vm86_trap: {
++ int ret = handle_vm86_trap((struct kernel_vm86_regs *) regs, error_code, trapnr);
++ if (ret) goto trap_signal;
++ return;
+ }
++}
+
-+ /* Decode the code segment base from the descriptor */
-+ base = get_desc_base((unsigned long *)desc);
++#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); \
++}
+
-+ if (seg & (1<<2)) {
-+ up(¤t->mm->context.sem);
-+ } else
-+ put_cpu();
++#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); \
++}
+
-+ /* 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;
++#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); \
+}
+
-+/*
-+ * Sometimes AMD Athlon/Opteron CPUs report invalid exceptions on prefetch.
-+ * Check that here and ignore it.
-+ */
-+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;
++#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); \
++}
+
-+ for (i = 0; scan_more && i < 15; i++) {
-+ unsigned char opcode;
-+ unsigned char instr_hi;
-+ unsigned char instr_lo;
++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)
+
-+ if (instr > limit)
-+ break;
-+ if (__get_user(opcode, (unsigned char __user *) instr))
-+ break;
++fastcall void __kprobes do_general_protection(struct pt_regs * regs,
++ long error_code)
++{
++ current->thread.error_code = error_code;
++ current->thread.trap_no = 13;
+
-+ instr_hi = opcode & 0xf0;
-+ instr_lo = opcode & 0x0f;
-+ instr++;
++ if (regs->eflags & VM_MASK)
++ goto gp_in_vm86;
+
-+ 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;
-+ }
++ if (!user_mode(regs))
++ goto gp_in_kernel;
++
++ current->thread.error_code = error_code;
++ current->thread.trap_no = 13;
++ force_sig(SIGSEGV, current);
++ return;
++
++gp_in_vm86:
++ local_irq_enable();
++ handle_vm86_fault((struct kernel_vm86_regs *) regs, error_code);
++ return;
++
++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);
+ }
-+ return prefetch;
+}
+
-+static inline int is_prefetch(struct pt_regs *regs, unsigned long addr,
-+ unsigned long error_code)
++static __kprobes void
++mem_parity_error(unsigned char reason, struct pt_regs * regs)
+{
-+ 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;
-+}
++ printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x on "
++ "CPU %d.\n", reason, smp_processor_id());
++ printk(KERN_EMERG "You have some hardware problem, likely on the PCI bus.\n");
++ if (panic_on_unrecovered_nmi)
++ panic("NMI: Not continuing");
+
-+static noinline void force_sig_info_fault(int si_signo, int si_code,
-+ unsigned long address, struct task_struct *tsk)
-+{
-+ siginfo_t info;
++ printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
+
-+ 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);
++ /* Clear and disable the memory parity error line. */
++ clear_mem_error(reason);
+}
+
-+fastcall void do_invalid_op(struct pt_regs *, unsigned long);
-+
-+#ifdef CONFIG_X86_PAE
-+static void dump_fault_path(unsigned long address)
++static __kprobes void
++io_check_error(unsigned char reason, struct pt_regs * regs)
+{
-+ unsigned long *p, page;
-+ unsigned long mfn;
++ printk(KERN_EMERG "NMI: IOCK error (debug interrupt?)\n");
++ show_registers(regs);
+
-+ 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] & 0x7) << 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]);
-+#ifndef CONFIG_HIGHPTE
-+ if (p[0] & 1) {
-+ mfn = (p[0] >> PAGE_SHIFT) | ((p[1] & 0x7) << 20);
-+ 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]);
-+ }
-+#endif
-+ }
++ /* Re-enable the IOCK line, wait for a few seconds */
++ clear_io_check_error(reason);
+}
-+#else
-+static void dump_fault_path(unsigned long address)
-+{
-+ 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, the page table might be allocated in highmem.
-+ * And lets rather not kmap-atomic the pte, just in case
-+ * it's allocated already.
-+ */
-+#ifndef CONFIG_HIGHPTE
-+ 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));
++static __kprobes 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;
+ }
+#endif
-+}
-+#endif
++ printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x on "
++ "CPU %d.\n", reason, smp_processor_id());
++ printk(KERN_EMERG "Do you have a strange power saving mode enabled?\n");
++ if (panic_on_unrecovered_nmi)
++ panic("NMI: Not continuing");
+
-+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;
++ printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
++}
+
-+#ifdef CONFIG_XEN
-+ /* Faults in hypervisor area are never spurious. */
-+ if (address >= HYPERVISOR_VIRT_START)
-+ return 0;
-+#endif
++static DEFINE_SPINLOCK(nmi_print_lock);
+
-+ /* Reserved-bit violation or user access to kernel space? */
-+ if (error_code & 0x0c)
-+ return 0;
++void __kprobes die_nmi(struct pt_regs *regs, const char *msg)
++{
++ if (notify_die(DIE_NMIWATCHDOG, msg, regs, 0, 2, SIGINT) ==
++ NOTIFY_STOP)
++ return;
+
-+ pgd = init_mm.pgd + pgd_index(address);
-+ if (!pgd_present(*pgd))
-+ return 0;
++ 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);
++ console_silent();
++ spin_unlock(&nmi_print_lock);
++ bust_spinlocks(0);
+
-+ pud = pud_offset(pgd, address);
-+ if (!pud_present(*pud))
-+ return 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);
++ }
+
-+ pmd = pmd_offset(pud, address);
-+ if (!pmd_present(*pmd))
-+ return 0;
++ do_exit(SIGSEGV);
++}
+
-+ 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;
++static __kprobes void default_do_nmi(struct pt_regs * regs)
++{
++ 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_tick(regs, reason))
++ return;
++ if (!do_nmi_callback(regs, smp_processor_id()))
+#endif
++ unknown_nmi_error(reason, regs);
+
-+ return 1;
++ 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();
+}
+
-+static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address)
++fastcall __kprobes void do_nmi(struct pt_regs * regs, long error_code)
+{
-+ unsigned index = pgd_index(address);
-+ pgd_t *pgd_k;
-+ pud_t *pud, *pud_k;
-+ pmd_t *pmd, *pmd_k;
++ int cpu;
+
-+ pgd += index;
-+ pgd_k = init_mm.pgd + index;
++ nmi_enter();
+
-+ if (!pgd_present(*pgd_k))
-+ return NULL;
++ cpu = smp_processor_id();
+
-+ /*
-+ * set_pgd(pgd, *pgd_k); here would be useless on PAE
-+ * and redundant with the set_pmd() on non-PAE. As would
-+ * set_pud.
-+ */
++ ++nmi_count(cpu);
+
-+ pud = pud_offset(pgd, address);
-+ pud_k = pud_offset(pgd_k, address);
-+ if (!pud_present(*pud_k))
-+ return NULL;
++ default_do_nmi(regs);
+
-+ 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);
-+#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)));
-+#endif
-+ else
-+ BUG_ON(pmd_page(*pmd) != pmd_page(*pmd_k));
-+ return pmd_k;
++ nmi_exit();
+}
+
-+/*
-+ * 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)
++#ifdef CONFIG_KPROBES
++fastcall void __kprobes do_int3(struct pt_regs *regs, long error_code)
+{
-+ 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;
++ 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
+
+/*
-+ * This routine handles page faults. It determines the address,
-+ * and the problem, and then passes it off to one of the appropriate
-+ * routines.
++ * 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.
+ *
-+ * 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
++ * 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)
+ */
-+fastcall void __kprobes do_page_fault(struct pt_regs *regs,
-+ unsigned long error_code)
++fastcall void __kprobes do_debug(struct pt_regs * regs, long error_code)
+{
-+ struct task_struct *tsk;
-+ struct mm_struct *mm;
-+ struct vm_area_struct * vma;
-+ unsigned long address;
-+ int write, si_code;
++ unsigned int condition;
++ struct task_struct *tsk = current;
+
-+ /* get the address */
-+ address = read_cr2();
++ get_debugreg(condition, 6);
+
-+ /* 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;
++ 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();
+
-+ tsk = current;
++ /* 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;
++ }
+
-+ si_code = SEGV_MAPERR;
++ if (regs->eflags & VM_MASK)
++ goto debug_vm86;
++
++ /* Save debug status register where ptrace can see it */
++ tsk->thread.debugreg[6] = condition;
+
+ /*
-+ * 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.
++ * Single-stepping through TF: make sure we ignore any events in
++ * kernel space (but re-enable TF when returning to user mode).
+ */
-+ 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;
++ if (condition & DR_STEP) {
+ /*
-+ * Don't take the mm semaphore here. If we fixup a prefetch
-+ * fault we could otherwise deadlock.
++ * We already checked v86 mode above, so we can
++ * check for kernel mode by just checking the CPL
++ * of CS.
+ */
-+ goto bad_area_nosemaphore;
++ if (!user_mode(regs))
++ goto clear_TF_reenable;
+ }
+
-+ if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
-+ SIGSEGV) == NOTIFY_STOP)
-+ return;
-+
-+ /* 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();
-+
-+ mm = tsk->mm;
++ /* Ok, finally something we can handle */
++ send_sigtrap(tsk, regs, error_code);
+
-+ /*
-+ * If we're in an interrupt, have no user context or are running in an
-+ * atomic region then we must not take the fault..
++ /* Disable additional traps. They'll be re-enabled when
++ * the signal is delivered.
+ */
-+ if (in_atomic() || !mm)
-+ goto bad_area_nosemaphore;
++clear_dr7:
++ set_debugreg(0, 7);
++ return;
+
-+ /* 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);
-+ }
++debug_vm86:
++ handle_vm86_trap((struct kernel_vm86_regs *) regs, error_code, 1);
++ return;
++
++clear_TF_reenable:
++ set_tsk_thread_flag(tsk, TIF_SINGLESTEP);
++ regs->eflags &= ~TF_MASK;
++ return;
++}
+
-+ 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..
++ * 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
+ */
-+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 math_error(void __user *eip)
++{
++ struct task_struct * task;
++ siginfo_t info;
++ unsigned short cwd, swd;
+
-+ survive:
+ /*
-+ * If for any reason at all we couldn't handle the fault,
-+ * make sure we exit gracefully rather than endlessly redo
-+ * the fault.
++ * Save the info for the exception handler and clear the error.
+ */
-+ 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();
-+ }
-+
++ 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;
+ /*
-+ * Did it hit the DOS screen memory VA from vm86 mode?
++ * (~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
+ */
-+ if (regs->eflags & VM_MASK) {
-+ unsigned long bit = (address - 0xA0000) >> PAGE_SHIFT;
-+ if (bit < 32)
-+ tsk->thread.screen_bitmap |= 1 << bit;
++ 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;
+ }
-+ 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);
++ force_sig_info(SIGFPE, &info, task);
++}
+
-+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;
++fastcall void do_coprocessor_error(struct pt_regs * regs, long error_code)
++{
++ ignore_fpu_irq = 1;
++ math_error((void __user *)regs->eip);
++}
+
-+ 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;
-+ }
++static void simd_math_error(void __user *eip)
++{
++ struct task_struct * task;
++ siginfo_t info;
++ unsigned short mxcsr;
+
-+#ifdef CONFIG_X86_F00F_BUG
+ /*
-+ * Pentium F0 0F C7 C8 bug workaround.
++ * Save the info for the exception handler and clear the error.
+ */
-+ if (boot_cpu_data.f00f_bug) {
-+ unsigned long nr;
-+
-+ nr = (address - idt_descr.address) >> 3;
-+
-+ if (nr == 6) {
-+ do_invalid_op(regs, 0);
-+ return;
-+ }
-+ }
-+#endif
-+
-+no_context:
-+ /* Are we prepared to handle this kernel fault? */
-+ if (fixup_exception(regs))
-+ return;
-+
-+ /*
-+ * Valid to do another page fault here, because if this fault
-+ * had been triggered by is_prefetch fixup_exception would have
-+ * handled it.
++ 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.
+ */
-+ if (is_prefetch(regs, address, error_code))
-+ return;
-+
-+/*
-+ * Oops. The kernel tried to access some bad page. We'll have to
-+ * terminate things with extreme prejudice.
-+ */
-+
-+ 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);
-+
-+/*
-+ * 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;
++ 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;
+ }
-+ 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 & 4))
-+ goto no_context;
-+
-+ /* User space => ok to do another page fault */
-+ if (is_prefetch(regs, address, error_code))
-+ return;
-+
-+ tsk->thread.cr2 = address;
-+ tsk->thread.error_code = error_code;
-+ tsk->thread.trap_no = 14;
-+ force_sig_info_fault(SIGBUS, BUS_ADRERR, address, tsk);
++ force_sig_info(SIGFPE, &info, task);
+}
+
-+#ifndef CONFIG_X86_PAE
-+void vmalloc_sync_all(void)
++fastcall void do_simd_coprocessor_error(struct pt_regs * regs,
++ long error_code)
+{
-+ /*
-+ * 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 = TASK_SIZE;
-+ unsigned long address;
-+
-+ BUILD_BUG_ON(TASK_SIZE & ~PGDIR_MASK);
-+ for (address = start; address >= TASK_SIZE; address += PGDIR_SIZE) {
-+ if (!test_bit(pgd_index(address), insync)) {
-+ unsigned long flags;
-+ struct page *page;
-+
-+ spin_lock_irqsave(&pgd_lock, flags);
-+ 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(pgd_index(address), insync);
++ 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;
+ }
-+ if (address == start && test_bit(pgd_index(address), insync))
-+ start = address + PGDIR_SIZE;
++ current->thread.trap_no = 19;
++ current->thread.error_code = error_code;
++ die_if_kernel("cache flush denied", regs, error_code);
++ force_sig(SIGSEGV, current);
+ }
+}
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/mm/highmem-xen.c linux-2.6.18-xen/arch/i386/mm/highmem-xen.c
---- linux-2.6.18.3/arch/i386/mm/highmem-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/mm/highmem-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,133 @@
-+#include <linux/highmem.h>
-+#include <linux/module.h>
+
-+void *kmap(struct page *page)
++#ifndef CONFIG_XEN
++fastcall void do_spurious_interrupt_bug(struct pt_regs * regs,
++ long error_code)
+{
-+ might_sleep();
-+ if (!PageHighMem(page))
-+ return page_address(page);
-+ return kmap_high(page);
++#if 0
++ /* No need to warn about this any longer. */
++ printk("Ignoring P6 Local APIC Spurious Interrupt Bug...\n");
++#endif
+}
+
-+void kunmap(struct page *page)
++fastcall unsigned long patch_espfix_desc(unsigned long uesp,
++ unsigned long kesp)
+{
-+ if (in_interrupt())
-+ BUG();
-+ if (!PageHighMem(page))
-+ return;
-+ kunmap_high(page);
++ int cpu = smp_processor_id();
++ struct Xgt_desc_struct *cpu_gdt_descr = &per_cpu(cpu_gdt_descr, cpu);
++ struct desc_struct *gdt = (struct desc_struct *)cpu_gdt_descr->address;
++ unsigned long base = (kesp - uesp) & -THREAD_SIZE;
++ unsigned long new_kesp = kesp - base;
++ unsigned long lim_pages = (new_kesp | (THREAD_SIZE - 1)) >> PAGE_SHIFT;
++ __u64 desc = *(__u64 *)&gdt[GDT_ENTRY_ESPFIX_SS];
++ /* Set up base for espfix segment */
++ desc &= 0x00f0ff0000000000ULL;
++ desc |= ((((__u64)base) << 16) & 0x000000ffffff0000ULL) |
++ ((((__u64)base) << 32) & 0xff00000000000000ULL) |
++ ((((__u64)lim_pages) << 32) & 0x000f000000000000ULL) |
++ (lim_pages & 0xffff);
++ *(__u64 *)&gdt[GDT_ENTRY_ESPFIX_SS] = desc;
++ return new_kesp;
+}
++#endif
+
+/*
-+ * 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.
++ * 'math_state_restore()' saves the current math information in the
++ * old math state array, and gets the new ones from the current task
+ *
-+ * However when holding an atomic kmap is is not legal to sleep, so atomic
-+ * kmaps are appropriate for short, tight code paths only.
++ * 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).
+ */
-+static void *__kmap_atomic(struct page *page, enum km_type type, pgprot_t prot)
++asmlinkage void math_state_restore(void)
+{
-+ enum fixed_addresses idx;
-+ unsigned long vaddr;
-+
-+ /* 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));
++ struct thread_info *thread = current_thread_info();
++ struct task_struct *tsk = thread->task;
+
-+ return (void*) vaddr;
++ /* 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() */
++ tsk->fpu_counter++;
+}
+
-+void *kmap_atomic(struct page *page, enum km_type type)
-+{
-+ return __kmap_atomic(page, type, kmap_prot);
-+}
++#ifndef CONFIG_MATH_EMULATION
+
-+/* Same as kmap_atomic but with PAGE_KERNEL_RO page protection. */
-+void *kmap_atomic_pte(struct page *page, enum km_type type)
++asmlinkage void math_emulate(long arg)
+{
-+ return __kmap_atomic(page, type, PAGE_KERNEL_RO);
++ 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();
+}
+
-+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;
-+ }
-+#endif
++#endif /* CONFIG_MATH_EMULATION */
+
-+#if defined(CONFIG_DEBUG_HIGHMEM)
-+ if (vaddr != __fix_to_virt(FIX_KMAP_BEGIN+idx))
-+ BUG();
++#ifdef CONFIG_X86_F00F_BUG
++void __init trap_init_f00f_bug(void)
++{
++ __set_fixmap(FIX_F00F_IDT, __pa(&idt_table), PAGE_KERNEL_RO);
+
+ /*
-+ * 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.
++ * Update the IDT descriptor and reload the IDT so that
++ * it uses the read-only mapped virtual address.
+ */
-+ pte_clear(&init_mm, vaddr, kmap_pte-idx);
-+#endif
-+
-+ dec_preempt_count();
-+ preempt_check_resched();
-+}
-+
-+/* 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;
-+
-+ inc_preempt_count();
-+
-+ 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);
-+
-+ return (void*) vaddr;
++ idt_descr.address = fix_to_virt(FIX_F00F_IDT);
++ load_idt(&idt_descr);
+}
++#endif
+
-+struct page *kmap_atomic_to_page(void *ptr)
-+{
-+ unsigned long idx, vaddr = (unsigned long)ptr;
-+ pte_t *pte;
-+
-+ if (vaddr < FIXADDR_START)
-+ return virt_to_page(ptr);
-+
-+ idx = virt_to_fix(vaddr);
-+ pte = kmap_pte - (idx - FIX_KMAP_BEGIN);
-+ return pte_page(*pte);
-+}
+
-+EXPORT_SYMBOL(kmap);
-+EXPORT_SYMBOL(kunmap);
-+EXPORT_SYMBOL(kmap_atomic);
-+EXPORT_SYMBOL(kunmap_atomic);
-+EXPORT_SYMBOL(kmap_atomic_to_page);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/mm/hypervisor.c linux-2.6.18-xen/arch/i386/mm/hypervisor.c
---- linux-2.6.18.3/arch/i386/mm/hypervisor.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/mm/hypervisor.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,457 @@
-+/******************************************************************************
-+ * 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.
++/*
++ * NB. All these are "trap gates" (i.e. events_mask isn't set) except
++ * for those that specify <dpl>|4 in the second field.
+ */
-+
-+#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>
-+
-+#ifdef CONFIG_X86_64
-+#define pmd_val_ma(v) (v).pmd
-+#else
-+#ifdef CONFIG_X86_PAE
-+# define pmd_val_ma(v) ((v).pmd)
-+# define pud_val_ma(v) ((v).pgd.pgd)
-+#else
-+# define pmd_val_ma(v) ((v).pud.pgd.pgd)
-+#endif
++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 xen_l1_entry_update(pte_t *ptr, pte_t val)
++void __init trap_init(void)
+{
-+ mmu_update_t u;
-+ u.ptr = virt_to_machine(ptr);
-+ u.val = pte_val_ma(val);
-+ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
-+}
++ HYPERVISOR_set_trap_table(trap_table);
+
-+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_ma(val);
-+ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
-+}
++ 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);
++ };
+
-+#ifdef CONFIG_X86_PAE
-+void xen_l3_entry_update(pud_t *ptr, pud_t val)
-+{
-+ mmu_update_t u;
-+ u.ptr = virt_to_machine(ptr);
-+ u.val = pud_val_ma(val);
-+ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
-+}
-+#endif
++ 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");
++ }
+
-+#ifdef CONFIG_X86_64
-+void xen_l3_entry_update(pud_t *ptr, pud_t val)
-+{
-+ mmu_update_t u;
-+ u.ptr = virt_to_machine(ptr);
-+ u.val = val.pud;
-+ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
++ /*
++ * Should be a barrier for any external CPU state.
++ */
++ cpu_init();
+}
+
-+void xen_l4_entry_update(pgd_t *ptr, pgd_t val)
++void smp_trap_init(trap_info_t *trap_ctxt)
+{
-+ mmu_update_t u;
-+ u.ptr = virt_to_machine(ptr);
-+ u.val = val.pgd;
-+ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
-+}
-+#endif /* CONFIG_X86_64 */
++ trap_info_t *t = trap_table;
+
-+void xen_machphys_update(unsigned long mfn, unsigned long pfn)
-+{
-+ mmu_update_t u;
-+ if (xen_feature(XENFEAT_auto_translated_physmap)) {
-+ BUG_ON(pfn != mfn);
-+ return;
++ 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;
+ }
-+ u.ptr = ((unsigned long long)mfn << PAGE_SHIFT) | MMU_MACHPHYS_UPDATE;
-+ u.val = pfn;
-+ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
+}
+
-+void xen_pt_switch(unsigned long ptr)
++static int __init kstack_setup(char *s)
+{
-+ 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);
++ kstack_depth_to_print = simple_strtoul(s, NULL, 0);
++ return 1;
+}
++__setup("kstack=", kstack_setup);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/tsc.c b/arch/i386/kernel/tsc.c
+--- a/arch/i386/kernel/tsc.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/tsc.c 2007-03-14 10:55:14.000000000 +0100
+@@ -100,6 +100,7 @@
+ return (cyc * cyc2ns_scale) >> CYC2NS_SCALE_FACTOR;
+ }
+
++#ifndef CONFIG_XEN
+ /*
+ * Scheduler clock - returns current time in nanosec units.
+ */
+@@ -123,6 +124,7 @@
+ /* return the value in ns */
+ return cycles_2_ns(this_offset);
+ }
++#endif
+
+ static unsigned long calculate_cpu_khz(void)
+ {
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/vm86.c b/arch/i386/kernel/vm86.c
+--- a/arch/i386/kernel/vm86.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/vm86.c 2007-03-14 10:55:14.000000000 +0100
+@@ -125,7 +125,9 @@
+ 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;
+
+@@ -148,12 +150,16 @@
+ 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
+
+ ret = KVM86->regs32;
+
+@@ -279,7 +285,9 @@
+
+ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk)
+ {
++#ifndef CONFIG_X86_NO_TSS
+ struct tss_struct *tss;
++#endif
+ /*
+ * make sure the vm86() system call doesn't try to do anything silly
+ */
+@@ -324,12 +332,16 @@
+ savesegment(fs, tsk->thread.saved_fs);
+ tsk->thread.saved_gs = info->regs32->xgs;
+
++#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 -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/vmlinux.lds.S b/arch/i386/kernel/vmlinux.lds.S
+--- a/arch/i386/kernel/vmlinux.lds.S 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/kernel/vmlinux.lds.S 2007-03-14 10:55:14.000000000 +0100
+@@ -35,7 +35,13 @@
+ }
+ SECTIONS
+ {
++/* xen i386 redefineds LOAD_OFFSET to zero on page.h
++ quintela at redhat.com */
++#ifdef CONFIG_XEN
++ . = __PAGE_OFFSET + LOAD_PHYSICAL_ADDR;
++#else
+ . = LOAD_OFFSET + LOAD_PHYSICAL_ADDR;
++#endif
+ phys_startup_32 = startup_32 - LOAD_OFFSET;
+ /* read-only */
+ .text : AT(ADDR(.text) - LOAD_OFFSET) {
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/kernel/vsyscall-note-xen.S b/arch/i386/kernel/vsyscall-note-xen.S
+--- a/arch/i386/kernel/vsyscall-note-xen.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/kernel/vsyscall-note-xen.S 2007-03-14 10:55:14.000000000 +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.
++ */
+
-+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);
-+}
++#include "vsyscall-note.S"
+
-+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);
++/*
++ * 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
++
++NOTE_KERNELCAP_BEGIN(1, 1)
++NOTE_KERNELCAP(1, "nosegneg") /* Change 1 back to 0 when glibc is fixed! */
++NOTE_KERNELCAP_END
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/lib/delay.c b/arch/i386/lib/delay.c
+--- a/arch/i386/lib/delay.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/lib/delay.c 2007-03-14 10:55:14.000000000 +0100
+@@ -60,6 +60,7 @@
+ delay_fn = delay_tsc;
+ }
+
++#ifndef CONFIG_X86_XEN
+ int read_current_timer(unsigned long *timer_val)
+ {
+ if (delay_fn == delay_tsc) {
+@@ -68,7 +69,7 @@
+ }
+ return -1;
+ }
+-
++#endif
+ void __delay(unsigned long loops)
+ {
+ delay_fn(loops);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/mach-xen/irqflags.c b/arch/i386/mach-xen/irqflags.c
+--- a/arch/i386/mach-xen/irqflags.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/mach-xen/irqflags.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,103 @@
++#include <linux/module.h>
++#include <linux/smp.h>
++#include <asm/irqflags.h>
++#include <asm/hypervisor.h>
+
-+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);
++/* interrupt control.. */
+
-+#ifdef CONFIG_SMP
++/*
++ * 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.
++ */
+
-+void xen_tlb_flush_all(void)
++unsigned long __raw_local_save_flags(void)
+{
-+ struct mmuext_op op;
-+ op.cmd = MMUEXT_TLB_FLUSH_ALL;
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
-+}
++ struct vcpu_info *_vcpu;
++ unsigned long flags;
+
-+void xen_tlb_flush_mask(cpumask_t *mask)
-+{
-+ struct mmuext_op op;
-+ if ( cpus_empty(*mask) )
-+ return;
-+ op.cmd = MMUEXT_TLB_FLUSH_MULTI;
-+ op.arg2.vcpumask = mask->bits;
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
-+}
++ preempt_disable();
++ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
++ flags = _vcpu->evtchn_upcall_mask;
++ preempt_enable();
+
-+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);
++ return flags;
+}
++EXPORT_SYMBOL(__raw_local_save_flags);
+
-+void xen_invlpg_mask(cpumask_t *mask, unsigned long ptr)
++void raw_local_irq_restore(unsigned long flags)
+{
-+ struct mmuext_op op;
-+ if ( cpus_empty(*mask) )
-+ return;
-+ op.cmd = MMUEXT_INVLPG_MULTI;
-+ op.arg1.linear_addr = ptr & PAGE_MASK;
-+ op.arg2.vcpumask = mask->bits;
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
-+}
-+
-+#endif /* CONFIG_SMP */
++ struct vcpu_info *_vcpu;
+
-+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
-+ op.cmd = MMUEXT_PIN_L2_TABLE;
-+#endif
-+ op.arg1.mfn = pfn_to_mfn(ptr >> PAGE_SHIFT);
-+ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++ preempt_disable();
++ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
++ if ((_vcpu->evtchn_upcall_mask = flags) == 0) {
++ barrier(); /* unmask then check (avoid races) */
++ if (unlikely(_vcpu->evtchn_upcall_pending))
++ force_evtchn_callback();
++ preempt_enable();
++ } else
++ preempt_enable_no_resched();
+}
++EXPORT_SYMBOL(raw_local_irq_restore);
+
-+void xen_pgd_unpin(unsigned long ptr)
++void raw_local_irq_disable(void)
+{
-+ 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);
++ struct vcpu_info *_vcpu;
++
++ preempt_disable();
++ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
++ _vcpu->evtchn_upcall_mask = 1;
++ preempt_enable_no_resched();
+}
++EXPORT_SYMBOL(raw_local_irq_disable);
+
-+void xen_set_ldt(unsigned long ptr, unsigned long len)
++void raw_local_irq_enable(void)
+{
-+ 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);
++ struct vcpu_info *_vcpu;
++
++ preempt_disable();
++ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
++ _vcpu->evtchn_upcall_mask = 0;
++ barrier(); /* unmask then check (avoid races) */
++ if (unlikely(_vcpu->evtchn_upcall_pending))
++ force_evtchn_callback();
++ preempt_enable();
+}
++EXPORT_SYMBOL(raw_local_irq_enable);
+
+/*
-+ * Bitmap is indexed by page number. If bit is set, the page is part of a
-+ * xen_create_contiguous_region() area of memory.
++ * For spinlocks, etc.:
+ */
-+unsigned long *contiguous_bitmap;
+
-+static void contiguous_bitmap_set(
-+ unsigned long first_page, unsigned long nr_pages)
++unsigned long __raw_local_irq_save(void)
+{
-+ unsigned long start_off, end_off, curr_idx, end_idx;
++ struct vcpu_info *_vcpu;
++ unsigned long flags;
+
-+ 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);
++ preempt_disable();
++ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
++ flags = _vcpu->evtchn_upcall_mask;
++ _vcpu->evtchn_upcall_mask = 1;
++ preempt_enable_no_resched();
+
-+ 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;
-+ }
++ return flags;
+}
++EXPORT_SYMBOL(__raw_local_irq_save);
+
-+static void contiguous_bitmap_clear(
-+ unsigned long first_page, unsigned long nr_pages)
++/* Cannot use preempt_enable() here as we would recurse in preempt_sched(). */
++int raw_irqs_disabled(void)
+{
-+ unsigned long start_off, end_off, curr_idx, end_idx;
++ struct vcpu_info *_vcpu;
++ int disabled;
+
-+ 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);
++ preempt_disable();
++ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
++ disabled = (_vcpu->evtchn_upcall_mask != 0);
++ preempt_enable_no_resched();
+
-+ 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);
-+ }
++ return disabled;
+}
++EXPORT_SYMBOL(raw_irqs_disabled);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/mach-xen/Makefile b/arch/i386/mach-xen/Makefile
+--- a/arch/i386/mach-xen/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/mach-xen/Makefile 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,5 @@
++#
++# Makefile for the linux kernel.
++#
+
-+/* Protected by balloon_lock. */
-+#define MAX_CONTIG_ORDER 9 /* 2MB */
-+static unsigned long discontig_frames[1<<MAX_CONTIG_ORDER];
-+static multicall_entry_t cr_mcl[1<<MAX_CONTIG_ORDER];
++obj-y := setup.o irqflags.o
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/mach-xen/setup.c b/arch/i386/mach-xen/setup.c
+--- a/arch/i386/mach-xen/setup.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/mach-xen/setup.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,161 @@
++/*
++ * Machine specific setup for generic
++ */
+
-+/* Ensure multi-page extents are contiguous in machine memory. */
-+int xen_create_contiguous_region(
-+ unsigned long vstart, unsigned int order, unsigned int address_bits)
++#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)
+{
-+ 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
-+ }
-+ };
++ get_option(&str, &no_broadcast);
++ printk ("Using %s mode\n", no_broadcast ? "No IPI Broadcast" :
++ "IPI Broadcast");
++ return 1;
++}
+
-+ /*
-+ * 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;
++__setup("no_ipi_broadcast", no_ipi_broadcast);
+
-+ if (unlikely(order > MAX_CONTIG_ORDER))
-+ return -ENOMEM;
++static int __init print_ipi_mode(void)
++{
++ printk ("Using IPI %s mode\n", no_broadcast ? "No-Shortcut" :
++ "Shortcut");
++ return 0;
++}
+
-+ set_xen_guest_handle(exchange.in.extent_start, in_frames);
-+ set_xen_guest_handle(exchange.out.extent_start, &out_frame);
++late_initcall(print_ipi_mode);
+
-+ scrub_pages(vstart, 1 << order);
++/**
++ * 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.
++ **/
+
-+ balloon_lock(flags);
++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];
+
-+ /* 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(cr_mcl, i))
-+ BUG();
++ memmap.nr_entries = E820MAX;
++ set_xen_guest_handle(memmap.buffer, map);
+
-+ /* 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 (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();
-+ }
++ rc = HYPERVISOR_memory_op(XENMEM_memory_map, &memmap);
++ if ( rc == -ENOSYS ) {
++ memmap.nr_entries = 1;
++ map[0].addr = 0ULL;
++ map[0].size = PFN_PHYS(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);
+
-+ /* 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);
-+ }
++ sanitize_e820_map(map, (char *)&memmap.nr_entries);
+
-+ cr_mcl[i - 1].args[MULTI_UVMFLAGS_INDEX] = order
-+ ? UVMF_TLB_FLUSH|UVMF_ALL
-+ : UVMF_INVLPG|UVMF_ALL;
-+ if (HYPERVISOR_multicall(cr_mcl, i))
-+ BUG();
++ BUG_ON(copy_e820_map(map, (char)memmap.nr_entries) < 0);
+
-+ if (success)
-+ contiguous_bitmap_set(__pa(vstart) >> PAGE_SHIFT,
-+ 1UL << order);
++ return "Xen";
++}
+
-+ balloon_unlock(flags);
++extern void hypervisor_callback(void);
++extern void failsafe_callback(void);
++extern void nmi(void);
+
-+ return success ? 0 : -ENOMEM;
-+}
++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 xen_destroy_contiguous_region(unsigned long vstart, unsigned int order)
++void __init machine_specific_arch_setup(void)
+{
-+ 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
-+ }
++ int ret;
++ struct xen_machphys_mapping mapping;
++ unsigned long machine_to_phys_nr_ents;
++ struct xen_platform_parameters pp;
++ struct callback_register event = {
++ .type = CALLBACKTYPE_event,
++ .address = { __KERNEL_CS, (unsigned long)hypervisor_callback },
++ };
++ struct callback_register failsafe = {
++ .type = CALLBACKTYPE_failsafe,
++ .address = { __KERNEL_CS, (unsigned long)failsafe_callback },
++ };
++ struct callback_register nmi_cb = {
++ .type = CALLBACKTYPE_nmi,
++ .address = { __KERNEL_CS, (unsigned long)nmi },
+ };
+
-+ if (xen_feature(XENFEAT_auto_translated_physmap) ||
-+ !test_bit(__pa(vstart) >> PAGE_SHIFT, contiguous_bitmap))
-+ return;
-+
-+ if (unlikely(order > MAX_CONTIG_ORDER))
-+ return;
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &event);
++ if (ret == 0)
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &failsafe);
++ if (ret == -ENOSYS)
++ ret = HYPERVISOR_set_callbacks(
++ event.address.cs, event.address.eip,
++ failsafe.address.cs, failsafe.address.eip);
++ BUG_ON(ret);
+
-+ set_xen_guest_handle(exchange.in.extent_start, &in_frame);
-+ set_xen_guest_handle(exchange.out.extent_start, out_frames);
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &nmi_cb);
++ if (ret == -ENOSYS) {
++ struct xennmi_callback cb;
+
-+ scrub_pages(vstart, 1 << order);
++ cb.handler_address = nmi_cb.address.eip;
++ HYPERVISOR_nmi_op(XENNMI_register_callback, &cb);
++ }
+
-+ balloon_lock(flags);
++ if (HYPERVISOR_xen_version(XENVER_platform_parameters,
++ &pp) == 0)
++ set_fixaddr_top(pp.virt_start - PAGE_SIZE);
+
-+ contiguous_bitmap_clear(__pa(vstart) >> PAGE_SHIFT, 1UL << order);
++ 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++;
++}
+
-+ /* 1. Find start MFN of contiguous extent. */
-+ in_frame = pfn_to_mfn(__pa(vstart) >> PAGE_SHIFT);
++/**
++ * pre_setup_arch_hook - hook called prior to any setup_arch() execution
++ *
++ * Description:
++ * generally used to activate any machine specific identification
++ * routines that may be needed before setup_arch() runs. On VISWS
++ * this is used to get the board revision and type.
++ **/
++void __init pre_setup_arch_hook(void)
++{
++ int max_cmdline;
+
-+ /* 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;
-+ }
-+ if (HYPERVISOR_multicall(cr_mcl, i))
-+ BUG();
++ 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);
++ /* Save unparsed command line copy for /proc/cmdline */
++ saved_command_line[max_cmdline-1] = '\0';
++}
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/Makefile b/arch/i386/Makefile
+--- a/arch/i386/Makefile 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/Makefile 2007-03-14 10:55:14.000000000 +0100
+@@ -60,6 +60,11 @@
+
+ CFLAGS += $(cflags-y)
+
++cppflags-$(CONFIG_XEN) += \
++ -D__XEN_INTERFACE_VERSION__=$(CONFIG_XEN_INTERFACE_VERSION)
+
-+ /* 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 (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;
-+ }
++CPPFLAGS += $(cppflags-y)
+
-+ /* 4. Map new pages in place of old pages. */
-+ for (i = 0; i < (1UL<<order); i++) {
-+ frame = success ? out_frames[i] : (in_frame + i);
-+ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
-+ pfn_pte_ma(frame, PAGE_KERNEL), 0);
-+ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i, frame);
-+ }
+ # Default subarch .c files
+ mcore-y := mach-default
+
+@@ -83,6 +88,10 @@
+ mflags-$(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-$(CONFIG_X86_XEN) := mach-xen
+
-+ cr_mcl[i - 1].args[MULTI_UVMFLAGS_INDEX] = order
-+ ? UVMF_TLB_FLUSH|UVMF_ALL
-+ : UVMF_INVLPG|UVMF_ALL;
-+ if (HYPERVISOR_multicall(cr_mcl, i))
-+ BUG();
+ # generic subarchitecture
+ mflags-$(CONFIG_X86_GENERICARCH) := -Iinclude/asm-i386/mach-generic
+ mcore-$(CONFIG_X86_GENERICARCH) := mach-default
+@@ -117,6 +126,19 @@
+ PHONY += zImage bzImage compressed zlilo bzlilo \
+ zdisk bzdisk fdimage fdimage144 fdimage288 isoimage install
+
++ifdef CONFIG_XEN
++CPPFLAGS := -Iinclude$(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
+
-+ balloon_unlock(flags);
-+}
++vmlinuz: vmlinux
++ $(Q)$(MAKE) $(build)=$(boot) $@
+
-+#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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/mm/init-xen.c linux-2.6.18-xen/arch/i386/mm/init-xen.c
---- linux-2.6.18.3/arch/i386/mm/init-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/mm/init-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,851 @@
++install:
++ $(Q)$(MAKE) $(build)=$(boot) XENGUEST=$(XENGUEST) $@
++else
+ all: bzImage
+
+ # KBUILD_IMAGE specify target image being built
+@@ -139,6 +161,7 @@
+
+ install:
+ $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(KBUILD_IMAGE) install
++endif
+
+ archclean:
+ $(Q)$(MAKE) $(clean)=arch/i386/boot
+@@ -157,3 +180,4 @@
+ CLEAN_FILES += arch/$(ARCH)/boot/fdimage \
+ arch/$(ARCH)/boot/image.iso \
+ arch/$(ARCH)/boot/mtools.conf
++CLEAN_FILES += vmlinuz vmlinux-stripped
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/mm/fault-xen.c b/arch/i386/mm/fault-xen.c
+--- a/arch/i386/mm/fault-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/mm/fault-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,762 @@
+/*
-+ * linux/arch/i386/mm/init.c
++ * linux/arch/i386/mm/fault.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>
@@ -17863,3832 +13673,4047 @@
+#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/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/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 <linux/module.h>
++#include <linux/kprobes.h>
++#include <linux/uaccess.h>
+
-+#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>
-+
-+extern unsigned long *contiguous_bitmap;
++#include <asm/desc.h>
++#include <asm/kdebug.h>
++#include <asm/segment.h>
+
-+unsigned int __VMALLOC_RESERVE = 128 << 20;
++extern void die(const char *,struct pt_regs *,long);
+
-+DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
-+unsigned long highstart_pfn, highend_pfn;
++static ATOMIC_NOTIFIER_HEAD(notify_page_fault_chain);
+
-+static int noinline do_test_wp_bit(void);
++int register_page_fault_notifier(struct notifier_block *nb)
++{
++ vmalloc_sync_all();
++ return atomic_notifier_chain_register(¬ify_page_fault_chain, nb);
++}
++EXPORT_SYMBOL_GPL(register_page_fault_notifier);
+
-+/*
-+ * 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)
++int unregister_page_fault_notifier(struct notifier_block *nb)
+{
-+ 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 atomic_notifier_chain_unregister(¬ify_page_fault_chain, nb);
++}
++EXPORT_SYMBOL_GPL(unregister_page_fault_notifier);
+
-+ return pmd_table;
++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);
+}
+
+/*
-+ * Create a page table and place a pointer to it in a middle page
-+ * directory entry.
++ * Unlock any spinlocks which will prevent us from getting the
++ * message out
+ */
-+static pte_t * __init one_page_table_init(pmd_t *pmd)
++void bust_spinlocks(int yes)
+{
-+ 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();
++ int loglevel_save = console_loglevel;
+
-+ return page_table;
++ if (yes) {
++ oops_in_progress = 1;
++ return;
+ }
-+
-+ return pte_offset_kernel(pmd, 0);
++#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;
+}
+
+/*
-+ * This function initializes a certain range of kernel virtual memory
-+ * with new bootmem page tables, everywhere page tables are missing in
-+ * the given range.
-+ */
-+
-+/*
-+ * 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.
++ * 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.
+ */
-+static void __init page_table_range_init (unsigned long start, unsigned long end, pgd_t *pgd_base)
++static inline unsigned long get_segment_eip(struct pt_regs *regs,
++ unsigned long *eip_limit)
+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ int pgd_idx, pmd_idx;
-+ unsigned long vaddr;
++ unsigned long eip = regs->eip;
++ unsigned seg = regs->xcs & 0xffff;
++ u32 seg_ar, seg_limit, base, *desc;
+
-+ vaddr = start;
-+ pgd_idx = pgd_index(vaddr);
-+ pmd_idx = pmd_index(vaddr);
-+ pgd = pgd_base + pgd_idx;
++ /* Unlikely, but must come before segment checks. */
++ if (unlikely(regs->eflags & VM_MASK)) {
++ base = seg << 4;
++ *eip_limit = base + 0xffff;
++ return base + (eip & 0xffff);
++ }
+
-+ 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);
++ /* The standard kernel/user address space limit. */
++ *eip_limit = user_mode(regs) ? USER_DS.seg : KERNEL_DS.seg;
++
++ /* By far the most common cases. */
++ if (likely(SEGMENT_IS_FLAT_CODE(seg)))
++ return eip;
+
-+ vaddr += PMD_SIZE;
-+ }
-+ pmd_idx = 0;
++ /* 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. */
+ }
-+}
+
-+static inline int is_kernel_text(unsigned long addr)
-+{
-+ if (addr >= PAGE_OFFSET && addr <= (unsigned long)__init_end)
-+ return 1;
-+ return 0;
-+}
++ /* 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);
++ }
+
-+/*
-+ * 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)
-+{
-+ unsigned long pfn;
-+ pgd_t *pgd;
-+ pmd_t *pmd;
-+ pte_t *pte;
-+ int pgd_idx, pmd_idx, pte_ofs;
++ /* Decode the code segment base from the descriptor */
++ base = get_desc_base((unsigned long *)desc);
+
-+ unsigned long max_ram_pfn = xen_start_info->nr_pages;
-+ if (max_ram_pfn > max_low_pfn)
-+ max_ram_pfn = max_low_pfn;
++ if (seg & (1<<2)) {
++ up(¤t->mm->context.sem);
++ } else
++ put_cpu();
+
-+ 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);
++ /* 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;
++}
+
-+ 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);
-+#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;
++/*
++ * Sometimes AMD Athlon/Opteron CPUs report invalid exceptions on prefetch.
++ * Check that here and ignore it.
++ */
++static int __is_prefetch(struct pt_regs *regs, unsigned long addr)
++{
++ unsigned long limit;
++ unsigned char *instr = (unsigned char *)get_segment_eip (regs, &limit);
++ int scan_more = 1;
++ int prefetch = 0;
++ int i;
+
-+ /* 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;
++ for (i = 0; scan_more && i < 15; i++) {
++ unsigned char opcode;
++ unsigned char instr_hi;
++ unsigned char instr_lo;
+
-+ 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);
++ if (instr > (unsigned char *)limit)
++ break;
++ if (probe_kernel_address(instr, opcode))
++ break;
+
-+ 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;
++ 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 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 > (unsigned char *)limit)
++ break;
++ if (probe_kernel_address(instr, opcode))
++ break;
++ prefetch = (instr_lo == 0xF) &&
++ (opcode == 0x0D || opcode == 0x18);
++ break;
++ default:
++ scan_more = 0;
++ break;
++ }
+ }
++ return prefetch;
+}
+
-+#ifndef CONFIG_XEN
-+
-+static inline int page_kills_ppro(unsigned long pagenr)
++static inline int is_prefetch(struct pt_regs *regs, unsigned long addr,
++ unsigned long error_code)
+{
-+ if (pagenr >= 0x70000 && pagenr <= 0x7003F)
-+ return 1;
++ 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;
-+}
-+
-+#else
++}
+
-+#define page_kills_ppro(p) 0
++static noinline void force_sig_info_fault(int si_signo, int si_code,
++ unsigned long address, struct task_struct *tsk)
++{
++ siginfo_t info;
+
-+#endif
++ 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);
++}
+
-+extern int is_available_memory(efi_memory_desc_t *);
++fastcall void do_invalid_op(struct pt_regs *, unsigned long);
+
-+int page_is_ram(unsigned long pagenr)
++#ifdef CONFIG_X86_PAE
++static void dump_fault_path(unsigned long address)
+{
-+ 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;
++ unsigned long *p, page;
++ unsigned long mfn;
+
-+ if ((pagenr >= addr) && (pagenr < end))
-+ return 1;
++ 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] & 0x7) << 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]);
++#ifndef CONFIG_HIGHPTE
++ if (p[0] & 1) {
++ mfn = (p[0] >> PAGE_SHIFT) | ((p[1] & 0x7) << 20);
++ 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]);
+ }
-+ return 0;
++#endif
+ }
++}
++#else
++static void dump_fault_path(unsigned long address)
++{
++ unsigned long page;
+
-+ for (i = 0; i < e820.nr_map; i++) {
-+
-+ 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;
++ 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, the page table might be allocated in highmem.
++ * And lets rather not kmap-atomic the pte, just in case
++ * it's allocated already.
++ */
++#ifndef CONFIG_HIGHPTE
++ 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));
+ }
-+ return 0;
++#endif
+}
++#endif
+
-+#ifdef CONFIG_HIGHMEM
-+pte_t *kmap_pte;
-+pgprot_t kmap_prot;
++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;
+
-+#define kmap_get_fixmap_pte(vaddr) \
-+ pte_offset_kernel(pmd_offset(pud_offset(pgd_offset_k(vaddr), vaddr), (vaddr)), (vaddr))
++#ifdef CONFIG_XEN
++ /* Faults in hypervisor area are never spurious. */
++ if (address >= HYPERVISOR_VIRT_START)
++ return 0;
++#endif
+
-+static void __init kmap_init(void)
-+{
-+ unsigned long kmap_vstart;
++ /* Reserved-bit violation or user access to kernel space? */
++ if (error_code & 0x0c)
++ return 0;
+
-+ /* cache the first kmap pte */
-+ kmap_vstart = __fix_to_virt(FIX_KMAP_BEGIN);
-+ kmap_pte = kmap_get_fixmap_pte(kmap_vstart);
++ pgd = init_mm.pgd + pgd_index(address);
++ if (!pgd_present(*pgd))
++ return 0;
+
-+ kmap_prot = PAGE_KERNEL;
++ 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 & 0x02) && !pte_write(*pte))
++ return 0;
++#ifdef CONFIG_X86_PAE
++ if ((error_code & 0x10) && (pte_val(*pte) & _PAGE_NX))
++ return 0;
++#endif
++
++ return 1;
+}
+
-+static void __init permanent_kmaps_init(pgd_t *pgd_base)
++static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address)
+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte;
-+ unsigned long vaddr;
++ 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;
+
-+ vaddr = PKMAP_BASE;
-+ page_table_range_init(vaddr, vaddr + PAGE_SIZE*LAST_PKMAP, pgd_base);
++ if (!pgd_present(*pgd_k))
++ return NULL;
+
-+ 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_pgd(pgd, *pgd_k); here would be useless on PAE
++ * and redundant with the set_pmd() on non-PAE. As would
++ * set_pud.
++ */
+
-+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++;
-+}
++ pud = pud_offset(pgd, address);
++ pud_k = pud_offset(pgd_k, address);
++ if (!pud_present(*pud_k))
++ return NULL;
+
-+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);
++ 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);
++#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)));
++#endif
++ else
++ BUG_ON(pmd_page(*pmd) != pmd_page(*pmd_k));
++ return pmd_k;
+}
+
-+static int add_one_highpage_hotplug(struct page *page, unsigned long pfn)
++/*
++ * 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)
+{
-+ free_new_highpage(page, pfn);
-+ totalram_pages++;
-+#ifdef CONFIG_FLATMEM
-+ max_mapnr = max(pfn, max_mapnr);
-+#endif
-+ num_physpages++;
++ 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;
+}
+
+/*
-+ * 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 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
+ */
-+void online_page(struct page *page)
++fastcall void __kprobes do_page_fault(struct pt_regs *regs,
++ unsigned long error_code)
+{
-+ ClearPageReserved(page);
-+ add_one_highpage_hotplug(page, page_to_pfn(page));
-+}
++ 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();
+
-+#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 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;
+
-+#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 */
++ tsk = current;
+
-+unsigned long long __PAGE_KERNEL = _PAGE_KERNEL;
-+EXPORT_SYMBOL(__PAGE_KERNEL);
-+unsigned long long __PAGE_KERNEL_EXEC = _PAGE_KERNEL_EXEC;
++ si_code = SEGV_MAPERR;
+
-+#ifdef CONFIG_NUMA
-+extern void __init remap_numa_kva(void);
-+#else
-+#define remap_numa_kva() do {} while (0)
++ /*
++ * 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;
++ }
+
-+pgd_t *swapper_pg_dir;
++ if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
++ SIGSEGV) == NOTIFY_STOP)
++ return;
+
-+static void __init pagetable_init (void)
-+{
-+ unsigned long vaddr;
-+ pgd_t *pgd_base = (pgd_t *)xen_start_info->pt_base;
++ /* 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();
+
-+ swapper_pg_dir = pgd_base;
-+ init_mm.pgd = pgd_base;
++ mm = tsk->mm;
+
-+ /* Enable PSE if available */
-+ if (cpu_has_pse) {
-+ set_in_cr4(X86_CR4_PSE);
-+ }
++ /*
++ * 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;
+
-+ /* Enable PGE if available */
-+ if (cpu_has_pge) {
-+ set_in_cr4(X86_CR4_PGE);
-+ __PAGE_KERNEL |= _PAGE_GLOBAL;
-+ __PAGE_KERNEL_EXEC |= _PAGE_GLOBAL;
++ /* 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);
+ }
+
-+ kernel_physical_mapping_init(pgd_base);
-+ remap_numa_kva();
++ 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 */
++ /* 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 | VM_WRITE)))
++ goto bad_area;
++ }
+
++ survive:
+ /*
-+ * Fixed mappings, only the page table structure has to be
-+ * created - mappings will be set by set_fixmap():
++ * If for any reason at all we couldn't handle the fault,
++ * make sure we exit gracefully rather than endlessly redo
++ * the fault.
+ */
-+ vaddr = __fix_to_virt(__end_of_fixed_addresses - 1) & PMD_MASK;
-+ page_table_range_init(vaddr, 0, pgd_base);
++ 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();
++ }
+
-+ permanent_kmaps_init(pgd_base);
-+}
++ /*
++ * 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;
+
-+#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.
++ * Something tried to access memory that isn't in our memory map..
++ * Fix it, but check if it's kernel or user first..
+ */
-+char __nosavedata swsusp_pg_dir[PAGE_SIZE]
-+ __attribute__ ((aligned (PAGE_SIZE)));
-+
-+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
++bad_area:
++ up_read(&mm->mmap_sem);
+
-+void zap_low_mappings (void)
-+{
-+ int i;
++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;
+
-+ save_pg_dir();
++ 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;
++ }
+
++#ifdef CONFIG_X86_F00F_BUG
+ /*
-+ * Zap initial low-memory mappings.
-+ *
-+ * Note that "pgd_clear()" doesn't do it for
-+ * us, because pgd_clear() is a no-op on i386.
++ * Pentium F0 0F C7 C8 bug workaround.
+ */
-+ 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));
++ if (boot_cpu_data.f00f_bug) {
++ unsigned long nr;
++
++ nr = (address - idt_descr.address) >> 3;
++
++ if (nr == 6) {
++ do_invalid_op(regs, 0);
++ return;
++ }
++ }
+#endif
-+ flush_tlb_all();
-+}
+
-+static int disable_nx __initdata = 0;
-+u64 __supported_pte_mask __read_mostly = ~_PAGE_NX;
-+EXPORT_SYMBOL(__supported_pte_mask);
++no_context:
++ /* Are we prepared to handle this kernel fault? */
++ if (fixup_exception(regs))
++ 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;
+
+/*
-+ * noexec = on|off
-+ *
-+ * Control non executable mappings.
-+ *
-+ * on Enable
-+ * off Disable
++ * Oops. The kernel tried to access some bad page. We'll have to
++ * terminate things with extreme prejudice.
+ */
-+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;
++
++ 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);
+
-+int nx_enabled = 0;
-+#ifdef CONFIG_X86_PAE
++/*
++ * 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 (is_init(tsk)) {
++ 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;
+
-+static void __init set_nx(void)
++do_sigbus:
++ up_read(&mm->mmap_sem);
++
++ /* Kernel mode? Handle exceptions or die */
++ if (!(error_code & 4))
++ goto no_context;
++
++ /* User space => ok to do another page fault */
++ if (is_prefetch(regs, address, error_code))
++ return;
++
++ tsk->thread.cr2 = address;
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = 14;
++ force_sig_info_fault(SIGBUS, BUS_ADRERR, address, tsk);
++}
++
++#ifndef CONFIG_X86_PAE
++void vmalloc_sync_all(void)
+{
-+ unsigned int v[4], l, h;
++ /*
++ * 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 = TASK_SIZE;
++ unsigned long address;
+
-+ 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;
++ BUILD_BUG_ON(TASK_SIZE & ~PGDIR_MASK);
++ for (address = start; address >= TASK_SIZE; address += PGDIR_SIZE) {
++ if (!test_bit(pgd_index(address), insync)) {
++ unsigned long flags;
++ struct page *page;
++
++ spin_lock_irqsave(&pgd_lock, flags);
++ 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(pgd_index(address), insync);
+ }
++ if (address == start && test_bit(pgd_index(address), insync))
++ start = address + PGDIR_SIZE;
+ }
+}
++#endif
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/mm/highmem-xen.c b/arch/i386/mm/highmem-xen.c
+--- a/arch/i386/mm/highmem-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/mm/highmem-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,117 @@
++#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);
++}
++
++void kunmap(struct page *page)
++{
++ if (in_interrupt())
++ BUG();
++ if (!PageHighMem(page))
++ return;
++ kunmap_high(page);
++}
+
+/*
-+ * Enables/disables executability of a given kernel page and
-+ * returns the previous setting.
++ * 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.
+ */
-+int __init set_kernel_exec(unsigned long vaddr, int enable)
++static void *__kmap_atomic(struct page *page, enum km_type type, pgprot_t prot)
+{
-+ pte_t *pte;
-+ int ret = 1;
++ enum fixed_addresses idx;
++ unsigned long vaddr;
+
-+ if (!nx_enabled)
-+ goto out;
++ /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */
++ pagefault_disable();
++ if (!PageHighMem(page))
++ return page_address(page);
+
-+ pte = lookup_address(vaddr);
-+ BUG_ON(!pte);
++ idx = type + KM_TYPE_NR*smp_processor_id();
++ vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
++ if (!pte_none(*(kmap_pte-idx)))
++ BUG();
++ set_pte_at(&init_mm, vaddr, kmap_pte-idx, mk_pte(page, prot));
+
-+ if (!pte_exec_kernel(*pte))
-+ ret = 0;
++ return (void*) vaddr;
++}
+
-+ 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;
++void *kmap_atomic(struct page *page, enum km_type type)
++{
++ return __kmap_atomic(page, type, kmap_prot);
+}
+
-+#endif
++/* 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, PAGE_KERNEL_RO);
++}
+
-+/*
-+ * 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)
++void kunmap_atomic(void *kvaddr, enum km_type type)
+{
-+ int i;
++ unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
++ enum fixed_addresses idx = type + KM_TYPE_NR*smp_processor_id();
+
-+#ifdef CONFIG_X86_PAE
-+ set_nx();
-+ if (nx_enabled)
-+ printk("NX (Execute Disable) protection: active\n");
++ /*
++ * Force other mappings to Oops if they'll try to access this pte
++ * without first remap it. Keeping stale mappings around is a bad idea
++ * also, in case the page changes cacheability attributes or becomes
++ * a protected page in a hypervisor.
++ */
++ if (vaddr == __fix_to_virt(FIX_KMAP_BEGIN+idx)) {
++ kpte_clear_flush(kmap_pte-idx, vaddr);
++ __flush_tlb_one(vaddr);
++ } else {
++#ifdef CONFIG_DEBUG_HIGHMEM
++ BUG_ON(vaddr < PAGE_OFFSET);
++ BUG_ON(vaddr >= (unsigned long)high_memory);
+#endif
++ }
+
-+ pagetable_init();
++ pagefault_enable();
++}
+
-+#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();
++/* 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;
+
-+ kmap_init();
++ pagefault_disable();
+
-+ /* 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));
++ 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));
+
-+ /* 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);
++ return (void*) vaddr;
+}
+
-+/*
-+ * 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)
++struct page *kmap_atomic_to_page(void *ptr)
+{
-+ printk("Checking if this processor honours the WP bit even in supervisor mode... ");
++ unsigned long idx, vaddr = (unsigned long)ptr;
++ pte_t *pte;
+
-+ /* 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 (vaddr < FIXADDR_START)
++ return virt_to_page(ptr);
+
-+ 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");
-+ }
++ idx = virt_to_fix(vaddr);
++ pte = kmap_pte - (idx - FIX_KMAP_BEGIN);
++ return pte_page(*pte);
+}
+
-+static void __init set_max_mapnr_init(void)
-+{
-+#ifdef CONFIG_HIGHMEM
-+ num_physpages = highend_pfn;
++EXPORT_SYMBOL(kmap);
++EXPORT_SYMBOL(kunmap);
++EXPORT_SYMBOL(kmap_atomic);
++EXPORT_SYMBOL(kunmap_atomic);
++EXPORT_SYMBOL(kmap_atomic_to_page);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/mm/hypervisor.c b/arch/i386/mm/hypervisor.c
+--- a/arch/i386/mm/hypervisor.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/mm/hypervisor.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,457 @@
++/******************************************************************************
++ * 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.
++ */
++
++#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>
++
++#ifdef CONFIG_X86_64
++#define pmd_val_ma(v) (v).pmd
+#else
-+ num_physpages = max_low_pfn;
++#ifdef CONFIG_X86_PAE
++# define pmd_val_ma(v) ((v).pmd)
++# define pud_val_ma(v) ((v).pgd.pgd)
++#else
++# define pmd_val_ma(v) ((v).pud.pgd.pgd)
+#endif
-+#ifdef CONFIG_FLATMEM
-+ max_mapnr = num_physpages;
+#endif
-+}
-+
-+static struct kcore_list kcore_mem, kcore_vmalloc;
+
-+void __init mem_init(void)
++void xen_l1_entry_update(pte_t *ptr, pte_t val)
+{
-+ extern int ppro_with_ram_bug(void);
-+ int codesize, reservedpages, datasize, initsize;
-+ int tmp;
-+ int bad_ppro;
-+ unsigned long pfn;
++ mmu_update_t u;
++ u.ptr = virt_to_machine(ptr);
++ u.val = pte_val_ma(val);
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
++}
+
-+ 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);
++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_ma(val);
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
++}
+
-+#if defined(CONFIG_SWIOTLB)
-+ swiotlb_init();
++#ifdef CONFIG_X86_PAE
++void xen_l3_entry_update(pud_t *ptr, pud_t val)
++{
++ mmu_update_t u;
++ u.ptr = virt_to_machine(ptr);
++ u.val = pud_val_ma(val);
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
++}
+#endif
+
-+#ifdef CONFIG_FLATMEM
-+ if (!mem_map)
-+ BUG();
-+#endif
-+
-+ bad_ppro = ppro_with_ram_bug();
++#ifdef CONFIG_X86_64
++void xen_l3_entry_update(pud_t *ptr, pud_t val)
++{
++ mmu_update_t u;
++ u.ptr = virt_to_machine(ptr);
++ u.val = val.pud;
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 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();
-+ }
-+#endif
-+
-+ set_max_mapnr_init();
++void xen_l4_entry_update(pgd_t *ptr, pgd_t val)
++{
++ mmu_update_t u;
++ u.ptr = virt_to_machine(ptr);
++ u.val = val.pgd;
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
++}
++#endif /* CONFIG_X86_64 */
+
-+#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(&mem_map[pfn]);
-+ init_page_count(&mem_map[pfn]);
-+ totalram_pages++;
++void xen_machphys_update(unsigned long mfn, unsigned long pfn)
++{
++ mmu_update_t u;
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ BUG_ON(pfn != mfn);
++ return;
+ }
++ u.ptr = ((unsigned long long)mfn << PAGE_SHIFT) | MMU_MACHPHYS_UPDATE;
++ u.val = pfn;
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
++}
+
-+ 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;
++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);
++}
+
-+ kclist_add(&kcore_mem, __va(0), max_low_pfn << PAGE_SHIFT);
-+ kclist_add(&kcore_vmalloc, (void *)VMALLOC_START,
-+ VMALLOC_END-VMALLOC_START);
++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);
++}
+
-+ 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))
-+ );
++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);
+
-+#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();
++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);
+
-+ /*
-+ * 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
++#ifdef CONFIG_SMP
+
-+ set_bit(PG_pinned, &virt_to_page(init_mm.pgd)->flags);
++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);
+}
+
-+/*
-+ * 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(u64 start, u64 size)
++void xen_tlb_flush_mask(cpumask_t *mask)
+{
-+ 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;
++ struct mmuext_op op;
++ if ( cpus_empty(*mask) )
++ return;
++ op.cmd = MMUEXT_TLB_FLUSH_MULTI;
++ op.arg2.vcpumask = mask->bits;
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++}
+
-+ return __add_pages(zone, start_pfn, nr_pages);
++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);
+}
+
-+int remove_memory(u64 start, u64 size)
++void xen_invlpg_mask(cpumask_t *mask, unsigned long ptr)
+{
-+ return -EINVAL;
++ struct mmuext_op op;
++ if ( cpus_empty(*mask) )
++ return;
++ op.cmd = MMUEXT_INVLPG_MULTI;
++ op.arg1.linear_addr = ptr & PAGE_MASK;
++ op.arg2.vcpumask = mask->bits;
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
+}
-+#endif
-+#endif
+
-+kmem_cache_t *pgd_cache;
-+kmem_cache_t *pmd_cache;
++#endif /* CONFIG_SMP */
+
-+void __init pgtable_cache_init(void)
++void xen_pgd_pin(unsigned long ptr)
+{
-+ 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),
++ 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
-+ PAGE_SIZE,
-+ PAGE_SIZE,
++ op.cmd = MMUEXT_PIN_L2_TABLE;
+#endif
-+ 0,
-+ pgd_ctor,
-+ PTRS_PER_PMD == 1 ? pgd_dtor : NULL);
-+ if (!pgd_cache)
-+ panic("pgtable_cache_init(): Cannot create pgd cache");
++ op.arg1.mfn = pfn_to_mfn(ptr >> PAGE_SHIFT);
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++}
++
++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);
++}
++
++void xen_set_ldt(unsigned long ptr, unsigned long len)
++{
++ 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);
+}
+
+/*
-+ * This function cannot be __init, since exceptions don't work in that
-+ * section. Put this after the callers, so that it cannot be inlined.
++ * Bitmap is indexed by page number. If bit is set, the page is part of a
++ * xen_create_contiguous_region() area of memory.
+ */
-+static int noinline do_test_wp_bit(void)
++unsigned long *contiguous_bitmap;
++
++static void contiguous_bitmap_set(
++ unsigned long first_page, unsigned long nr_pages)
+{
-+ char tmp_reg;
-+ int flag;
++ unsigned long start_off, end_off, curr_idx, end_idx;
+
-+ __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;
-+}
++ 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);
+
-+#ifdef CONFIG_DEBUG_RODATA
++ 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;
++ }
++}
+
-+void mark_rodata_ro(void)
++static void contiguous_bitmap_clear(
++ unsigned long first_page, unsigned long nr_pages)
+{
-+ unsigned long addr = (unsigned long)__start_rodata;
++ unsigned long start_off, end_off, curr_idx, end_idx;
+
-+ for (; addr < (unsigned long)__end_rodata; addr += PAGE_SIZE)
-+ change_page_attr(virt_to_page(addr), 1, PAGE_KERNEL_RO);
++ 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);
+
-+ printk("Write protecting the kernel read-only data: %uk\n",
-+ (__end_rodata - __start_rodata) >> 10);
++ 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 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, 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
++ }
++ };
+
+ /*
-+ * 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.
++ * 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.
+ */
-+ global_flush_tlb();
-+}
-+#endif
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return 0;
+
-+void free_init_pages(char *what, unsigned long begin, unsigned long end)
-+{
-+ unsigned long addr;
++ if (unlikely(order > MAX_CONTIG_ORDER))
++ return -ENOMEM;
+
-+ 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);
-+}
++ set_xen_guest_handle(exchange.in.extent_start, in_frames);
++ set_xen_guest_handle(exchange.out.extent_start, &out_frame);
+
-+void free_initmem(void)
-+{
-+ free_init_pages("unused kernel memory",
-+ (unsigned long)(&__init_begin),
-+ (unsigned long)(&__init_end));
-+}
++ scrub_pages(vstart, 1 << order);
+
-+#ifdef CONFIG_BLK_DEV_INITRD
-+void free_initrd_mem(unsigned long start, unsigned long end)
-+{
-+ free_init_pages("initrd memory", start, end);
-+}
-+#endif
++ balloon_lock(flags);
+
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/mm/ioremap-xen.c linux-2.6.18-xen/arch/i386/mm/ioremap-xen.c
---- linux-2.6.18.3/arch/i386/mm/ioremap-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/mm/ioremap-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,443 @@
-+/*
-+ * arch/i386/mm/ioremap.c
-+ *
-+ * Re-map IO memory to kernel address space so that we can access it.
-+ * This is needed for high PCI addresses that aren't mapped in the
-+ * 640k-1MB IO memory area on PC's
-+ *
-+ * (C) Copyright 1995 1996 Linus Torvalds
-+ */
++ /* 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(cr_mcl, i))
++ BUG();
+
-+#include <linux/vmalloc.h>
-+#include <linux/init.h>
-+#include <linux/slab.h>
-+#include <linux/module.h>
-+#include <asm/io.h>
-+#include <asm/fixmap.h>
-+#include <asm/cacheflush.h>
-+#include <asm/tlbflush.h>
-+#include <asm/pgtable.h>
-+#include <asm/pgalloc.h>
++ /* 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 (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();
++ }
++ }
+
-+#define ISA_START_ADDRESS 0x0
-+#define ISA_END_ADDRESS 0x100000
++ /* 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);
++ }
+
-+static int direct_remap_area_pte_fn(pte_t *pte,
-+ struct page *pmd_page,
-+ unsigned long address,
-+ void *data)
-+{
-+ mmu_update_t **v = (mmu_update_t **)data;
++ cr_mcl[i - 1].args[MULTI_UVMFLAGS_INDEX] = order
++ ? UVMF_TLB_FLUSH|UVMF_ALL
++ : UVMF_INVLPG|UVMF_ALL;
++ if (HYPERVISOR_multicall(cr_mcl, i))
++ BUG();
+
-+ BUG_ON(!pte_none(*pte));
++ if (success)
++ contiguous_bitmap_set(__pa(vstart) >> PAGE_SHIFT,
++ 1UL << order);
+
-+ (*v)->ptr = ((u64)pfn_to_mfn(page_to_pfn(pmd_page)) <<
-+ PAGE_SHIFT) | ((unsigned long)pte & ~PAGE_MASK);
-+ (*v)++;
++ balloon_unlock(flags);
+
-+ return 0;
++ return success ? 0 : -ENOMEM;
+}
+
-+static int __direct_remap_pfn_range(struct mm_struct *mm,
-+ unsigned long address,
-+ unsigned long mfn,
-+ unsigned long size,
-+ pgprot_t prot,
-+ domid_t domid)
++void xen_destroy_contiguous_region(unsigned long vstart, unsigned int order)
+{
-+ int rc;
-+ unsigned long i, start_address;
-+ mmu_update_t *u, *v, *w;
++ 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
++ }
++ };
+
-+ u = v = w = (mmu_update_t *)__get_free_page(GFP_KERNEL|__GFP_REPEAT);
-+ if (u == NULL)
-+ return -ENOMEM;
++ if (xen_feature(XENFEAT_auto_translated_physmap) ||
++ !test_bit(__pa(vstart) >> PAGE_SHIFT, contiguous_bitmap))
++ return;
+
-+ start_address = address;
++ if (unlikely(order > MAX_CONTIG_ORDER))
++ return;
+
-+ flush_cache_all();
++ set_xen_guest_handle(exchange.in.extent_start, &in_frame);
++ set_xen_guest_handle(exchange.out.extent_start, out_frames);
+
-+ for (i = 0; i < size; i += PAGE_SIZE) {
-+ if ((v - u) == (PAGE_SIZE / sizeof(mmu_update_t))) {
-+ /* Flush a full batch after filling in the PTE ptrs. */
-+ rc = apply_to_page_range(mm, start_address,
-+ address - start_address,
-+ direct_remap_area_pte_fn, &w);
-+ if (rc)
-+ goto out;
-+ rc = -EFAULT;
-+ if (HYPERVISOR_mmu_update(u, v - u, NULL, domid) < 0)
-+ goto out;
-+ v = w = u;
-+ start_address = address;
-+ }
++ scrub_pages(vstart, 1 << order);
+
-+ /*
-+ * Fill in the machine address: PTE ptr is done later by
-+ * __direct_remap_area_pages().
-+ */
-+ v->val = pte_val_ma(pfn_pte_ma(mfn, prot));
++ balloon_lock(flags);
+
-+ mfn++;
-+ address += PAGE_SIZE;
-+ v++;
-+ }
++ contiguous_bitmap_clear(__pa(vstart) >> PAGE_SHIFT, 1UL << order);
+
-+ if (v != u) {
-+ /* Final batch. */
-+ rc = apply_to_page_range(mm, start_address,
-+ address - start_address,
-+ direct_remap_area_pte_fn, &w);
-+ if (rc)
-+ goto out;
-+ rc = -EFAULT;
-+ if (unlikely(HYPERVISOR_mmu_update(u, v - u, NULL, domid) < 0))
-+ goto out;
++ /* 1. Find start MFN of contiguous extent. */
++ in_frame = pfn_to_mfn(__pa(vstart) >> PAGE_SHIFT);
++
++ /* 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;
+ }
++ if (HYPERVISOR_multicall(cr_mcl, i))
++ BUG();
+
-+ rc = 0;
++ /* 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 (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;
++ }
+
-+ out:
-+ flush_tlb_all();
++ /* 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);
++ }
+
-+ free_page((unsigned long)u);
++ cr_mcl[i - 1].args[MULTI_UVMFLAGS_INDEX] = order
++ ? UVMF_TLB_FLUSH|UVMF_ALL
++ : UVMF_INVLPG|UVMF_ALL;
++ if (HYPERVISOR_multicall(cr_mcl, i))
++ BUG();
+
-+ return rc;
++ balloon_unlock(flags);
+}
+
-+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)
++#ifdef __i386__
++int write_ldt_entry(void *ldt, int entry, __u32 entry_a, __u32 entry_b)
+{
-+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ return remap_pfn_range(vma, address, mfn, size, prot);
++ __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 -x '.hg*' -x '.git*' -urN a/arch/i386/mm/init-xen.c b/arch/i386/mm/init-xen.c
+--- a/arch/i386/mm/init-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/mm/init-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,871 @@
++/*
++ * linux/arch/i386/mm/init.c
++ *
++ * Copyright (C) 1995 Linus Torvalds
++ *
++ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
++ */
+
-+ if (domid == DOMID_SELF)
-+ return -EINVAL;
++#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>
+
-+ vma->vm_flags |= VM_IO | VM_RESERVED;
++#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>
+
-+ vma->vm_mm->context.has_foreign_mappings = 1;
++extern unsigned long *contiguous_bitmap;
+
-+ return __direct_remap_pfn_range(
-+ vma->vm_mm, address, mfn, size, prot, domid);
-+}
-+EXPORT_SYMBOL(direct_remap_pfn_range);
++unsigned int __VMALLOC_RESERVE = 128 << 20;
+
-+int direct_kernel_remap_pfn_range(unsigned long address,
-+ unsigned long mfn,
-+ unsigned long size,
-+ pgprot_t prot,
-+ domid_t domid)
-+{
-+ return __direct_remap_pfn_range(
-+ &init_mm, address, mfn, size, prot, domid);
-+}
-+EXPORT_SYMBOL(direct_kernel_remap_pfn_range);
++DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
++unsigned long highstart_pfn, highend_pfn;
+
-+static int lookup_pte_fn(
-+ pte_t *pte, struct page *pmd_page, unsigned long addr, void *data)
-+{
-+ uint64_t *ptep = (uint64_t *)data;
-+ if (ptep)
-+ *ptep = ((uint64_t)pfn_to_mfn(page_to_pfn(pmd_page)) <<
-+ PAGE_SHIFT) | ((unsigned long)pte & ~PAGE_MASK);
-+ return 0;
-+}
++static int noinline do_test_wp_bit(void);
+
-+int create_lookup_pte_addr(struct mm_struct *mm,
-+ unsigned long address,
-+ uint64_t *ptep)
++/*
++ * 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)
+{
-+ return apply_to_page_range(mm, address, PAGE_SIZE,
-+ lookup_pte_fn, ptep);
-+}
-+
-+EXPORT_SYMBOL(create_lookup_pte_addr);
++ 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
+
-+static int noop_fn(
-+ pte_t *pte, struct page *pmd_page, unsigned long addr, void *data)
-+{
-+ return 0;
++ return pmd_table;
+}
+
-+int touch_pte_range(struct mm_struct *mm,
-+ unsigned long address,
-+ unsigned long size)
-+{
-+ return apply_to_page_range(mm, address, size, noop_fn, NULL);
-+}
-+
-+EXPORT_SYMBOL(touch_pte_range);
-+
+/*
-+ * Does @address reside within a non-highmem page that is local to this virtual
-+ * machine (i.e., not an I/O page, nor a memory page belonging to another VM).
-+ * See the comment that accompanies mfn_to_local_pfn() in page.h to understand
-+ * why this works.
++ * Create a page table and place a pointer to it in a middle page
++ * directory entry.
+ */
-+static inline int is_local_lowmem(unsigned long address)
++static pte_t * __init one_page_table_init(pmd_t *pmd)
+{
-+ extern unsigned long max_low_pfn;
-+ return (mfn_to_local_pfn(address >> PAGE_SHIFT) < max_low_pfn);
++ 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);
+}
+
+/*
-+ * Generic mapping function (not visible outside):
++ * This function initializes a certain range of kernel virtual memory
++ * with new bootmem page tables, everywhere page tables are missing in
++ * the given range.
+ */
+
+/*
-+ * Remap an arbitrary physical address space into the kernel virtual
-+ * address space. Needed when the kernel wants to access high addresses
-+ * directly.
-+ *
-+ * NOTE! We need to allow non-page-aligned mappings too: we will obviously
-+ * have to convert them into an offset in a page-aligned mapping, but the
-+ * caller shouldn't need to know that small detail.
++ * 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.
+ */
-+void __iomem * __ioremap(unsigned long phys_addr, unsigned long size, unsigned long flags)
++static void __init page_table_range_init (unsigned long start, unsigned long end, pgd_t *pgd_base)
+{
-+ void __iomem * addr;
-+ struct vm_struct * area;
-+ unsigned long offset, last_addr;
-+ domid_t domid = DOMID_IO;
-+
-+ /* Don't allow wraparound or zero size */
-+ last_addr = phys_addr + size - 1;
-+ if (!size || last_addr < phys_addr)
-+ return NULL;
-+
-+ /*
-+ * Don't remap the low PCI/ISA area, it's always mapped..
-+ */
-+ if (is_initial_xendomain() &&
-+ phys_addr >= ISA_START_ADDRESS && last_addr < ISA_END_ADDRESS)
-+ return (void __iomem *) isa_bus_to_virt(phys_addr);
-+
-+ /*
-+ * Don't allow anybody to remap normal RAM that we're using..
-+ */
-+ if (is_local_lowmem(phys_addr)) {
-+ char *t_addr, *t_end;
-+ struct page *page;
-+
-+ t_addr = bus_to_virt(phys_addr);
-+ t_end = t_addr + (size - 1);
-+
-+ for(page = virt_to_page(t_addr); page <= virt_to_page(t_end); page++)
-+ if(!PageReserved(page))
-+ return NULL;
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ int pgd_idx, pmd_idx;
++ unsigned long vaddr;
+
-+ domid = DOMID_SELF;
-+ }
++ vaddr = start;
++ pgd_idx = pgd_index(vaddr);
++ pmd_idx = pmd_index(vaddr);
++ pgd = pgd_base + pgd_idx;
+
-+ /*
-+ * Mappings have to be page-aligned
-+ */
-+ offset = phys_addr & ~PAGE_MASK;
-+ phys_addr &= PAGE_MASK;
-+ size = PAGE_ALIGN(last_addr+1) - phys_addr;
++ 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);
+
-+ /*
-+ * Ok, go for it..
-+ */
-+ area = get_vm_area(size, VM_IOREMAP | (flags << 20));
-+ if (!area)
-+ return NULL;
-+ area->phys_addr = phys_addr;
-+ addr = (void __iomem *) area->addr;
-+ flags |= _PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED;
-+ if (__direct_remap_pfn_range(&init_mm, (unsigned long)addr,
-+ phys_addr>>PAGE_SHIFT,
-+ size, __pgprot(flags), domid)) {
-+ vunmap((void __force *) addr);
-+ return NULL;
++ vaddr += PMD_SIZE;
++ }
++ pmd_idx = 0;
+ }
-+ return (void __iomem *) (offset + (char __iomem *)addr);
+}
-+EXPORT_SYMBOL(__ioremap);
+
-+/**
-+ * ioremap_nocache - map bus memory into CPU space
-+ * @offset: bus address of the memory
-+ * @size: size of the resource to map
-+ *
-+ * ioremap_nocache 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.
-+ *
-+ * This version of ioremap ensures that the memory is marked uncachable
-+ * on the CPU as well as honouring existing caching rules from things like
-+ * the PCI bus. Note that there are other caches and buffers on many
-+ * busses. In particular driver authors should read up on PCI writes
-+ *
-+ * It's useful if some control registers are in such an area and
-+ * write combining or read caching is not desirable:
-+ *
-+ * Must be freed with iounmap.
-+ */
++static inline int is_kernel_text(unsigned long addr)
++{
++ if (addr >= PAGE_OFFSET && addr <= (unsigned long)__init_end)
++ return 1;
++ return 0;
++}
+
-+void __iomem *ioremap_nocache (unsigned long phys_addr, unsigned long size)
++/*
++ * 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)
+{
-+ unsigned long last_addr;
-+ void __iomem *p = __ioremap(phys_addr, size, _PAGE_PCD);
-+ if (!p)
-+ return p;
++ unsigned long pfn;
++ pgd_t *pgd;
++ pmd_t *pmd;
++ pte_t *pte;
++ int pgd_idx, pmd_idx, pte_ofs;
+
-+ /* Guaranteed to be > phys_addr, as per __ioremap() */
-+ last_addr = phys_addr + size - 1;
++ unsigned long max_ram_pfn = xen_start_info->nr_pages;
++ if (max_ram_pfn > max_low_pfn)
++ max_ram_pfn = max_low_pfn;
+
-+ if (is_local_lowmem(last_addr)) {
-+ struct page *ppage = virt_to_page(bus_to_virt(phys_addr));
-+ unsigned long npages;
++ 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);
+
-+ phys_addr &= PAGE_MASK;
++ 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);
++#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;
+
-+ /* This might overflow and become zero.. */
-+ last_addr = PAGE_ALIGN(last_addr);
++ /* 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;
+
-+ /* .. but that's ok, because modulo-2**n arithmetic will make
-+ * the page-aligned "last - first" come out right.
-+ */
-+ npages = (last_addr - phys_addr) >> PAGE_SHIFT;
++ 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);
+
-+ if (change_page_attr(ppage, npages, PAGE_KERNEL_NOCACHE) < 0) {
-+ iounmap(p);
-+ p = NULL;
++ 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;
++ }
+ }
-+ global_flush_tlb();
++ pmd_idx = 0;
+ }
-+
-+ return p;
+}
-+EXPORT_SYMBOL(ioremap_nocache);
+
-+/**
-+ * iounmap - Free a IO remapping
-+ * @addr: virtual address from ioremap_*
-+ *
-+ * Caller must ensure there is only one unmapping for the same pointer.
-+ */
-+void iounmap(volatile void __iomem *addr)
++#ifndef CONFIG_XEN
++
++static inline int page_kills_ppro(unsigned long pagenr)
+{
-+ struct vm_struct *p, *o;
++ if (pagenr >= 0x70000 && pagenr <= 0x7003F)
++ return 1;
++ return 0;
++}
+
-+ if ((void __force *)addr <= high_memory)
-+ return;
++#else
+
-+ /*
-+ * __ioremap special-cases the PCI/ISA range by not instantiating a
-+ * vm_area and by simply returning an address into the kernel mapping
-+ * of ISA space. So handle that here.
-+ */
-+ if ((unsigned long) addr >= fix_to_virt(FIX_ISAMAP_BEGIN))
-+ return;
++#define page_kills_ppro(p) 0
+
-+ addr = (volatile void __iomem *)(PAGE_MASK & (unsigned long __force)addr);
++#endif
+
-+ /* Use the vm area unlocked, assuming the caller
-+ ensures there isn't another iounmap for the same address
-+ in parallel. Reuse of the virtual address is prevented by
-+ leaving it in the global lists until we're done with it.
-+ cpa takes care of the direct mappings. */
-+ read_lock(&vmlist_lock);
-+ for (p = vmlist; p; p = p->next) {
-+ if (p->addr == addr)
-+ break;
-+ }
-+ read_unlock(&vmlist_lock);
++int page_is_ram(unsigned long pagenr)
++{
++ int i;
++ unsigned long addr, end;
+
-+ if (!p) {
-+ printk("iounmap: bad address %p\n", addr);
-+ dump_stack();
-+ return;
++ 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;
+ }
+
-+ /* Reset the direct mapping. Can block */
-+ if ((p->flags >> 20) && is_local_lowmem(p->phys_addr)) {
-+ /* p->size includes the guard page, but cpa doesn't like that */
-+ change_page_attr(virt_to_page(bus_to_virt(p->phys_addr)),
-+ (p->size - PAGE_SIZE) >> PAGE_SHIFT,
-+ PAGE_KERNEL);
-+ global_flush_tlb();
-+ }
++ for (i = 0; i < e820.nr_map; i++) {
+
-+ /* Finally remove it */
-+ o = remove_vm_area((void *)addr);
-+ BUG_ON(p != o || o == NULL);
-+ kfree(p);
++ 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;
+}
-+EXPORT_SYMBOL(iounmap);
-+
-+void __init *bt_ioremap(unsigned long phys_addr, unsigned long size)
-+{
-+ unsigned long offset, last_addr;
-+ unsigned int nrpages;
-+ enum fixed_addresses idx;
+
-+ /* Don't allow wraparound or zero size */
-+ last_addr = phys_addr + size - 1;
-+ if (!size || last_addr < phys_addr)
-+ return NULL;
++#ifdef CONFIG_HIGHMEM
++pte_t *kmap_pte;
++pgprot_t kmap_prot;
+
-+ /*
-+ * Don't remap the low PCI/ISA area, it's always mapped..
-+ */
-+ if (is_initial_xendomain() &&
-+ phys_addr >= ISA_START_ADDRESS && last_addr < ISA_END_ADDRESS)
-+ return isa_bus_to_virt(phys_addr);
++#define kmap_get_fixmap_pte(vaddr) \
++ pte_offset_kernel(pmd_offset(pud_offset(pgd_offset_k(vaddr), vaddr), (vaddr)), (vaddr))
+
-+ /*
-+ * Mappings have to be page-aligned
-+ */
-+ offset = phys_addr & ~PAGE_MASK;
-+ phys_addr &= PAGE_MASK;
-+ size = PAGE_ALIGN(last_addr) - phys_addr;
++static void __init kmap_init(void)
++{
++ unsigned long kmap_vstart;
+
-+ /*
-+ * Mappings have to fit in the FIX_BTMAP area.
-+ */
-+ nrpages = size >> PAGE_SHIFT;
-+ if (nrpages > NR_FIX_BTMAPS)
-+ return NULL;
++ /* cache the first kmap pte */
++ kmap_vstart = __fix_to_virt(FIX_KMAP_BEGIN);
++ kmap_pte = kmap_get_fixmap_pte(kmap_vstart);
+
-+ /*
-+ * Ok, go for it..
-+ */
-+ idx = FIX_BTMAP_BEGIN;
-+ while (nrpages > 0) {
-+ set_fixmap(idx, phys_addr);
-+ phys_addr += PAGE_SIZE;
-+ --idx;
-+ --nrpages;
-+ }
-+ return (void*) (offset + fix_to_virt(FIX_BTMAP_BEGIN));
++ kmap_prot = PAGE_KERNEL;
+}
+
-+void __init bt_iounmap(void *addr, unsigned long size)
++static void __init permanent_kmaps_init(pgd_t *pgd_base)
+{
-+ unsigned long virt_addr;
-+ unsigned long offset;
-+ unsigned int nrpages;
-+ enum fixed_addresses idx;
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
++ unsigned long vaddr;
+
-+ virt_addr = (unsigned long)addr;
-+ if (virt_addr < fix_to_virt(FIX_BTMAP_BEGIN))
-+ return;
-+ if (virt_addr >= fix_to_virt(FIX_ISAMAP_BEGIN))
-+ return;
-+ offset = virt_addr & ~PAGE_MASK;
-+ nrpages = PAGE_ALIGN(offset + size - 1) >> PAGE_SHIFT;
++ vaddr = PKMAP_BASE;
++ page_table_range_init(vaddr, vaddr + PAGE_SIZE*LAST_PKMAP, pgd_base);
+
-+ idx = FIX_BTMAP_BEGIN;
-+ while (nrpages > 0) {
-+ clear_fixmap(idx);
-+ --idx;
-+ --nrpages;
-+ }
++ 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;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/mm/Makefile linux-2.6.18-xen/arch/i386/mm/Makefile
---- linux-2.6.18.3/arch/i386/mm/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/mm/Makefile 2006-11-19 14:26:22.000000000 +0100
-@@ -8,3 +8,11 @@
- 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
++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++;
++}
+
-+obj-y += hypervisor.o
++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);
++}
++
++static int __meminit 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;
++}
+
-+obj-y := $(call cherrypickxen, $(obj-y))
-+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/mm/pageattr.c linux-2.6.18-xen/arch/i386/mm/pageattr.c
---- linux-2.6.18.3/arch/i386/mm/pageattr.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/mm/pageattr.c 2006-11-19 14:26:22.000000000 +0100
-@@ -84,7 +84,7 @@
- unsigned long flags;
-
- set_pte_atomic(kpte, pte); /* change init_mm */
-- if (PTRS_PER_PMD > 1)
-+ if (HAVE_SHARED_KERNEL_PMD)
- return;
-
- spin_lock_irqsave(&pgd_lock, flags);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/mm/pgtable-xen.c linux-2.6.18-xen/arch/i386/mm/pgtable-xen.c
---- linux-2.6.18.3/arch/i386/mm/pgtable-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/mm/pgtable-xen.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,699 @@
+/*
-+ * linux/arch/i386/mm/pgtable.c
++ * 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 __meminit online_page(struct page *page)
++{
++ ClearPageReserved(page);
++ add_one_highpage_hotplug(page, page_to_pfn(page));
++}
+
-+#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 <xen/foreign_page.h>
-+#include <asm/hypervisor.h>
++#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 */
+
-+static void pgd_test_and_unpin(pgd_t *pgd);
++#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 */
+
-+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;
++unsigned long long __PAGE_KERNEL = _PAGE_KERNEL;
++EXPORT_SYMBOL(__PAGE_KERNEL);
++unsigned long long __PAGE_KERNEL_EXEC = _PAGE_KERNEL_EXEC;
+
-+ 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) {
-+ pgdat_resize_lock(pgdat, &flags);
-+ for (i = 0; i < pgdat->node_spanned_pages; ++i) {
-+ page = pgdat_page_nr(pgdat, i);
-+ total++;
-+ if (PageHighMem(page))
-+ highmem++;
-+ if (PageReserved(page))
-+ reserved++;
-+ else if (PageSwapCache(page))
-+ cached++;
-+ else if (page_count(page))
-+ shared += page_count(page) - 1;
-+ }
-+ pgdat_resize_unlock(pgdat, &flags);
-+ }
-+ printk(KERN_INFO "%d pages of RAM\n", total);
-+ printk(KERN_INFO "%d pages of HIGHMEM\n", highmem);
-+ printk(KERN_INFO "%d reserved pages\n", reserved);
-+ printk(KERN_INFO "%d pages shared\n", shared);
-+ printk(KERN_INFO "%d pages swap cached\n", cached);
++#ifdef CONFIG_NUMA
++extern void __init remap_numa_kva(void);
++#else
++#define remap_numa_kva() do {} while (0)
++#endif
+
-+ printk(KERN_INFO "%lu pages dirty\n", global_page_state(NR_FILE_DIRTY));
-+ printk(KERN_INFO "%lu pages writeback\n",
-+ global_page_state(NR_WRITEBACK));
-+ printk(KERN_INFO "%lu pages mapped\n", global_page_state(NR_FILE_MAPPED));
-+ printk(KERN_INFO "%lu pages slab\n", global_page_state(NR_SLAB));
-+ printk(KERN_INFO "%lu pages pagetables\n",
-+ global_page_state(NR_PAGETABLE));
-+}
++pgd_t *swapper_pg_dir;
+
-+/*
-+ * 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)
++static void __init pagetable_init (void)
+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte;
++ unsigned long vaddr;
++ pgd_t *pgd_base = (pgd_t *)xen_start_info->pt_base;
+
-+ pgd = swapper_pg_dir + pgd_index(vaddr);
-+ if (pgd_none(*pgd)) {
-+ BUG();
-+ return;
-+ }
-+ pud = pud_offset(pgd, vaddr);
-+ if (pud_none(*pud)) {
-+ BUG();
-+ return;
++ swapper_pg_dir = pgd_base;
++ init_mm.pgd = pgd_base;
++
++ /* Enable PSE if available */
++ if (cpu_has_pse) {
++ set_in_cr4(X86_CR4_PSE);
+ }
-+ pmd = pmd_offset(pud, vaddr);
-+ if (pmd_none(*pmd)) {
-+ BUG();
-+ return;
++
++ /* Enable PGE if available */
++ if (cpu_has_pge) {
++ set_in_cr4(X86_CR4_PGE);
++ __PAGE_KERNEL |= _PAGE_GLOBAL;
++ __PAGE_KERNEL_EXEC |= _PAGE_GLOBAL;
+ }
-+ pte = pte_offset_kernel(pmd, vaddr);
-+ /* <pfn,flags> stored as-is, to permit clearing entries */
-+ set_pte(pte, pfn_pte(pfn, flags));
++
++ kernel_physical_mapping_init(pgd_base);
++ remap_numa_kva();
+
+ /*
-+ * It's enough to flush this one mapping.
-+ * (PGE mappings get flushed as well)
++ * Fixed mappings, only the page table structure has to be
++ * created - mappings will be set by set_fixmap():
+ */
-+ __flush_tlb_one(vaddr);
++ vaddr = __fix_to_virt(__end_of_fixed_addresses - 1) & PMD_MASK;
++ page_table_range_init(vaddr, 0, pgd_base);
++
++ permanent_kmaps_init(pgd_base);
+}
+
++#if defined(CONFIG_SOFTWARE_SUSPEND) || defined(CONFIG_ACPI_SLEEP)
+/*
-+ * 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;
++ * 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)));
+
-+ 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);
-+ /* <pfn,flags> stored as-is, to permit clearing entries */
-+ set_pte(pte, pfn_pte_ma(pfn, flags));
++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
++
++void zap_low_mappings (void)
++{
++ int i;
++
++ save_pg_dir();
+
+ /*
-+ * It's enough to flush this one mapping.
-+ * (PGE mappings get flushed as well)
++ * Zap initial low-memory mappings.
++ *
++ * Note that "pgd_clear()" doesn't do it for
++ * us, because pgd_clear() is a no-op on i386.
+ */
-+ __flush_tlb_one(vaddr);
++ 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);
++
+/*
-+ * 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.
-+ * The pmd must already be instantiated. Assumes PAE mode.
-+ */
-+void set_pmd_pfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags)
++ * noexec = on|off
++ *
++ * Control non executable mappings.
++ *
++ * on Enable
++ * off Disable
++ */
++static int __init noexec_setup(char *str)
+{
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
++ if (!str || !strcmp(str, "on")) {
++ if (cpu_has_nx) {
++ __supported_pte_mask |= _PAGE_NX;
++ disable_nx = 0;
++ }
++ } else if (!strcmp(str,"off")) {
++ disable_nx = 1;
++ __supported_pte_mask &= ~_PAGE_NX;
++ } else
++ return -EINVAL;
+
-+ if (vaddr & (PMD_SIZE-1)) { /* vaddr is misaligned */
-+ printk(KERN_WARNING "set_pmd_pfn: vaddr misaligned\n");
-+ return; /* BUG(); */
-+ }
-+ if (pfn & (PTRS_PER_PTE-1)) { /* pfn is misaligned */
-+ printk(KERN_WARNING "set_pmd_pfn: pfn misaligned\n");
-+ return; /* BUG(); */
-+ }
-+ pgd = swapper_pg_dir + pgd_index(vaddr);
-+ if (pgd_none(*pgd)) {
-+ printk(KERN_WARNING "set_pmd_pfn: pgd_none\n");
-+ return; /* BUG(); */
-+ }
-+ pud = pud_offset(pgd, vaddr);
-+ pmd = pmd_offset(pud, vaddr);
-+ set_pmd(pmd, pfn_pmd(pfn, flags));
-+ /*
-+ * It's enough to flush this one mapping.
-+ * (PGE mappings get flushed as well)
-+ */
-+ __flush_tlb_one(vaddr);
++ return 0;
+}
++early_param("noexec", noexec_setup);
+
-+static int nr_fixmaps = 0;
-+unsigned long __FIXADDR_TOP = (HYPERVISOR_VIRT_START - 2 * PAGE_SIZE);
-+EXPORT_SYMBOL(__FIXADDR_TOP);
++int nx_enabled = 0;
++#ifdef CONFIG_X86_PAE
+
-+void __set_fixmap (enum fixed_addresses idx, maddr_t phys, pgprot_t flags)
++static void __init set_nx(void)
+{
-+ unsigned long address = __fix_to_virt(idx);
++ unsigned int v[4], l, h;
+
-+ if (idx >= __end_of_fixed_addresses) {
-+ BUG();
-+ return;
-+ }
-+ switch (idx) {
-+ case FIX_WP_TEST:
-+#ifdef CONFIG_X86_F00F_BUG
-+ case FIX_F00F_IDT:
-+#endif
-+ set_pte_pfn(address, phys >> PAGE_SHIFT, flags);
-+ break;
-+ default:
-+ set_pte_pfn_ma(address, phys >> PAGE_SHIFT, flags);
-+ break;
++ 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;
++ }
+ }
-+ nr_fixmaps++;
+}
+
-+void set_fixaddr_top(unsigned long top)
++/*
++ * Enables/disables executability of a given kernel page and
++ * returns the previous setting.
++ */
++int __init set_kernel_exec(unsigned long vaddr, int enable)
+{
-+ BUG_ON(nr_fixmaps > 0);
-+ __FIXADDR_TOP = top - PAGE_SIZE;
-+}
++ pte_t *pte;
++ int ret = 1;
+
-+pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
-+{
-+ pte_t *pte = (pte_t *)__get_free_page(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO);
-+ if (pte)
-+ make_lowmem_page_readonly(pte, XENFEAT_writable_page_tables);
-+ return pte;
++ 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);
++ pte_update_defer(&init_mm, vaddr, pte);
++ __flush_tlb_all();
++out:
++ return ret;
+}
+
-+struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
++#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)
+{
-+ struct page *pte;
++ int i;
+
-+#ifdef CONFIG_HIGHPTE
-+ pte = alloc_pages(GFP_KERNEL|__GFP_HIGHMEM|__GFP_REPEAT|__GFP_ZERO, 0);
-+#else
-+ pte = alloc_pages(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO, 0);
-+ if (pte) {
-+ SetPageForeign(pte, pte_free);
-+ init_page_count(pte);
-+ }
++#ifdef CONFIG_X86_PAE
++ set_nx();
++ if (nx_enabled)
++ printk("NX (Execute Disable) protection: active\n");
+#endif
-+ return pte;
-+}
+
-+void pte_free(struct page *pte)
-+{
-+ unsigned long va = (unsigned long)__va(page_to_pfn(pte)<<PAGE_SHIFT);
++ pagetable_init();
+
-+ if (!pte_write(*virt_to_ptep(va)))
-+ BUG_ON(HYPERVISOR_update_va_mapping(
-+ va, pfn_pte(page_to_pfn(pte), PAGE_KERNEL), 0));
++#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();
+
-+ ClearPageForeign(pte);
-+ init_page_count(pte);
++ kmap_init();
+
-+ __free_page(pte);
-+}
++ /* 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));
+
-+void pmd_ctor(void *pmd, kmem_cache_t *cache, unsigned long flags)
-+{
-+ memset(pmd, 0, PTRS_PER_PMD*sizeof(pmd_t));
++ /* 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);
+}
+
+/*
-+ * List of all pgd's needed for non-PAE so it can invalidate entries
-+ * in both cached and uncached pgd's; not needed for PAE since the
-+ * kernel pmd is shared. If PAE were not to share the pmd a similar
-+ * tactic would be needed. This is essentially codepath-based locking
-+ * against pageattr.c; it is the unique case in which a valid change
-+ * of kernel pagetables can't be lazily synchronized by vmalloc faults.
-+ * vmalloc faults work because attached pagetables are never freed.
-+ * The locking scheme was chosen on the basis of manfred's
-+ * recommendations and having no core impact whatsoever.
-+ * -- wli
++ * 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.
+ */
-+DEFINE_SPINLOCK(pgd_lock);
-+struct page *pgd_list;
-+
-+static inline void pgd_list_add(pgd_t *pgd)
-+{
-+ struct page *page = virt_to_page(pgd);
-+ page->index = (unsigned long)pgd_list;
-+ if (pgd_list)
-+ set_page_private(pgd_list, (unsigned long)&page->index);
-+ pgd_list = page;
-+ set_page_private(page, (unsigned long)&pgd_list);
-+}
+
-+static inline void pgd_list_del(pgd_t *pgd)
++static void __init test_wp_bit(void)
+{
-+ struct page *next, **pprev, *page = virt_to_page(pgd);
-+ next = (struct page *)page->index;
-+ pprev = (struct page **)page_private(page);
-+ *pprev = next;
-+ if (next)
-+ set_page_private(next, (unsigned long)pprev);
-+}
++ printk("Checking if this processor honours the WP bit even in supervisor mode... ");
+
-+void pgd_ctor(void *pgd, kmem_cache_t *cache, unsigned long unused)
-+{
-+ unsigned long flags;
++ /* 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 (PTRS_PER_PMD > 1) {
-+ if (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 (!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 {
-+ 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);
-+ memset(pgd, 0, USER_PTRS_PER_PGD*sizeof(pgd_t));
-+ pgd_list_add(pgd);
-+ spin_unlock_irqrestore(&pgd_lock, flags);
++ printk("Ok.\n");
+ }
+}
+
-+/* never called when PTRS_PER_PMD > 1 */
-+void pgd_dtor(void *pgd, kmem_cache_t *cache, unsigned long unused)
++static struct kcore_list kcore_mem, kcore_vmalloc;
++
++void __init mem_init(void)
+{
-+ unsigned long flags; /* can be called from interrupt context */
++ extern int ppro_with_ram_bug(void);
++ int codesize, reservedpages, datasize, initsize;
++ int tmp;
++ int bad_ppro;
++ unsigned long pfn;
+
-+ spin_lock_irqsave(&pgd_lock, flags);
-+ pgd_list_del(pgd);
-+ spin_unlock_irqrestore(&pgd_lock, flags);
++ 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);
+
-+ pgd_test_and_unpin(pgd);
-+}
++#if defined(CONFIG_SWIOTLB)
++ swiotlb_init();
++#endif
+
-+pgd_t *pgd_alloc(struct mm_struct *mm)
-+{
-+ int i;
-+ pgd_t *pgd = kmem_cache_alloc(pgd_cache, GFP_KERNEL);
-+ pmd_t **pmd;
-+ unsigned long flags;
++#ifdef CONFIG_FLATMEM
++ BUG_ON(!mem_map);
++#endif
++
++ bad_ppro = ppro_with_ram_bug();
+
-+ pgd_test_and_unpin(pgd);
++#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
++
++ 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(&mem_map[pfn]);
++ init_page_count(&mem_map[pfn]);
++ totalram_pages++;
++ }
+
-+ if (PTRS_PER_PMD == 1 || !pgd)
-+ return pgd;
++ 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++;
+
-+ if (HAVE_SHARED_KERNEL_PMD) {
-+ for (i = 0; i < USER_PTRS_PER_PGD; ++i) {
-+ pmd_t *pmd = kmem_cache_alloc(pmd_cache, GFP_KERNEL);
-+ if (!pmd)
-+ goto out_oom;
-+ set_pgd(&pgd[i], __pgd(1 + __pa(pmd)));
-+ }
-+ return pgd;
-+ }
++ set_highmem_pages_init(bad_ppro);
+
-+ /*
-+ * We can race save/restore (if we sleep during a GFP_KERNEL memory
-+ * allocation). We therefore store virtual addresses of pmds as they
-+ * do not change across save/restore, and poke the machine addresses
-+ * into the pgdir under the pgd_lock.
-+ */
-+ pmd = kmalloc(PTRS_PER_PGD * sizeof(pmd_t *), GFP_KERNEL);
-+ if (!pmd) {
-+ kmem_cache_free(pgd_cache, pgd);
-+ return NULL;
-+ }
++ codesize = (unsigned long) &_etext - (unsigned long) &_text;
++ datasize = (unsigned long) &_edata - (unsigned long) &_etext;
++ initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin;
+
-+ /* Allocate pmds, remember virtual addresses. */
-+ for (i = 0; i < PTRS_PER_PGD; ++i) {
-+ pmd[i] = kmem_cache_alloc(pmd_cache, GFP_KERNEL);
-+ if (!pmd[i])
-+ goto out_oom;
-+ }
++ kclist_add(&kcore_mem, __va(0), max_low_pfn << PAGE_SHIFT);
++ kclist_add(&kcore_vmalloc, (void *)VMALLOC_START,
++ VMALLOC_END-VMALLOC_START);
+
-+ spin_lock_irqsave(&pgd_lock, flags);
++ 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))
++ );
+
-+ /* Protect against save/restore: move below 4GB under pgd_lock. */
-+ if (!xen_feature(XENFEAT_pae_pgdir_above_4gb)) {
-+ int rc = xen_create_contiguous_region(
-+ (unsigned long)pgd, 0, 32);
-+ if (rc) {
-+ spin_unlock_irqrestore(&pgd_lock, flags);
-+ goto out_oom;
-+ }
-+ }
++#if 1 /* double-sanity-check paranoia */
++ printk("virtual kernel memory layout:\n"
++ " fixmap : 0x%08lx - 0x%08lx (%4ld kB)\n"
++#ifdef CONFIG_HIGHMEM
++ " pkmap : 0x%08lx - 0x%08lx (%4ld kB)\n"
++#endif
++ " vmalloc : 0x%08lx - 0x%08lx (%4ld MB)\n"
++ " lowmem : 0x%08lx - 0x%08lx (%4ld MB)\n"
++ " .init : 0x%08lx - 0x%08lx (%4ld kB)\n"
++ " .data : 0x%08lx - 0x%08lx (%4ld kB)\n"
++ " .text : 0x%08lx - 0x%08lx (%4ld kB)\n",
++ FIXADDR_START, FIXADDR_TOP,
++ (FIXADDR_TOP - FIXADDR_START) >> 10,
+
-+ /* Copy kernel pmd contents and write-protect the new pmds. */
-+ 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);
-+ memcpy(pmd[i], kpmd, PAGE_SIZE);
-+ make_lowmem_page_readonly(
-+ pmd[i], XENFEAT_writable_page_tables);
-+ }
++#ifdef CONFIG_HIGHMEM
++ PKMAP_BASE, PKMAP_BASE+LAST_PKMAP*PAGE_SIZE,
++ (LAST_PKMAP*PAGE_SIZE) >> 10,
++#endif
+
-+ /* It is safe to poke machine addresses of pmds under the pmd_lock. */
-+ for (i = 0; i < PTRS_PER_PGD; i++)
-+ set_pgd(&pgd[i], __pgd(1 + __pa(pmd[i])));
++ VMALLOC_START, VMALLOC_END,
++ (VMALLOC_END - VMALLOC_START) >> 20,
+
-+ /* Ensure this pgd gets picked up and pinned on save/restore. */
-+ pgd_list_add(pgd);
++ (unsigned long)__va(0), (unsigned long)high_memory,
++ ((unsigned long)high_memory - (unsigned long)__va(0)) >> 20,
+
-+ spin_unlock_irqrestore(&pgd_lock, flags);
++ (unsigned long)&__init_begin, (unsigned long)&__init_end,
++ ((unsigned long)&__init_end - (unsigned long)&__init_begin) >> 10,
+
-+ kfree(pmd);
++ (unsigned long)&_etext, (unsigned long)&_edata,
++ ((unsigned long)&_edata - (unsigned long)&_etext) >> 10,
+
-+ return pgd;
++ (unsigned long)&_text, (unsigned long)&_etext,
++ ((unsigned long)&_etext - (unsigned long)&_text) >> 10);
+
-+out_oom:
-+ if (HAVE_SHARED_KERNEL_PMD) {
-+ for (i--; i >= 0; i--)
-+ kmem_cache_free(pmd_cache,
-+ (void *)__va(pgd_val(pgd[i])-1));
-+ } else {
-+ for (i--; i >= 0; i--)
-+ kmem_cache_free(pmd_cache, pmd[i]);
-+ kfree(pmd);
++#ifdef CONFIG_HIGHMEM
++ BUG_ON(PKMAP_BASE+LAST_PKMAP*PAGE_SIZE > FIXADDR_START);
++ BUG_ON(VMALLOC_END > PKMAP_BASE);
++#endif
++ BUG_ON(VMALLOC_START > VMALLOC_END);
++ BUG_ON((unsigned long)high_memory > VMALLOC_START);
++#endif /* double-sanity-check paranoia */
++
++#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);
++}
++
++#ifdef CONFIG_MEMORY_HOTPLUG
++int arch_add_memory(int nid, u64 start, u64 size)
++{
++ struct pglist_data *pgdata = NODE_DATA(nid);
++ struct zone *zone = pgdata->node_zones + ZONE_HIGHMEM;
++ 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;
++}
++EXPORT_SYMBOL_GPL(remove_memory);
++#endif
++
++struct kmem_cache *pgd_cache;
++struct kmem_cache *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");
+ }
-+ kmem_cache_free(pgd_cache, pgd);
-+ return NULL;
++ 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");
+}
+
-+void pgd_free(pgd_t *pgd)
++/*
++ * 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)
+{
-+ int i;
++ char tmp_reg;
++ int flag;
+
-+ /*
-+ * After this the pgd should not be pinned for the duration of this
-+ * function's execution. We should never sleep and thus never race:
-+ * 1. User pmds will not become write-protected under our feet due
-+ * to a concurrent mm_pin_all().
-+ * 2. The machine addresses in PGD entries will not become invalid
-+ * due to a concurrent save/restore.
-+ */
-+ pgd_test_and_unpin(pgd);
++ __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;
++}
+
-+ /* 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) {
-+ pmd_t *pmd = (void *)__va(pgd_val(pgd[i])-1);
-+ kmem_cache_free(pmd_cache, pmd);
-+ }
++#ifdef CONFIG_DEBUG_RODATA
+
-+ if (!HAVE_SHARED_KERNEL_PMD) {
-+ unsigned long flags;
-+ spin_lock_irqsave(&pgd_lock, flags);
-+ pgd_list_del(pgd);
-+ spin_unlock_irqrestore(&pgd_lock, flags);
++void mark_rodata_ro(void)
++{
++ unsigned long addr = (unsigned long)__start_rodata;
+
-+ for (i = USER_PTRS_PER_PGD; i < PTRS_PER_PGD; i++) {
-+ pmd_t *pmd = (void *)__va(pgd_val(pgd[i])-1);
-+ make_lowmem_page_writable(
-+ pmd, XENFEAT_writable_page_tables);
-+ memset(pmd, 0, PTRS_PER_PMD*sizeof(pmd_t));
-+ kmem_cache_free(pmd_cache, pmd);
-+ }
++ for (; addr < (unsigned long)__end_rodata; addr += PAGE_SIZE)
++ change_page_attr(virt_to_page(addr), 1, PAGE_KERNEL_RO);
+
-+ if (!xen_feature(XENFEAT_pae_pgdir_above_4gb))
-+ xen_destroy_contiguous_region(
-+ (unsigned long)pgd, 0);
-+ }
-+ }
++ printk("Write protecting the kernel read-only data: %uk\n",
++ (__end_rodata - __start_rodata) >> 10);
+
-+ /* in the non-PAE case, free_pgtables() clears user pgd entries */
-+ kmem_cache_free(pgd_cache, pgd);
++ /*
++ * 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 make_lowmem_page_readonly(void *va, unsigned int feature)
++void free_init_pages(char *what, unsigned long begin, unsigned long end)
+{
-+ pte_t *pte;
-+ int rc;
++ unsigned long addr;
+
-+ if (xen_feature(feature))
-+ return;
++ 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);
++}
+
-+ pte = virt_to_ptep(va);
-+ rc = HYPERVISOR_update_va_mapping(
-+ (unsigned long)va, pte_wrprotect(*pte), 0);
-+ BUG_ON(rc);
++void free_initmem(void)
++{
++ free_init_pages("unused kernel memory",
++ (unsigned long)(&__init_begin),
++ (unsigned long)(&__init_end));
+}
+
-+void make_lowmem_page_writable(void *va, unsigned int feature)
++#ifdef CONFIG_BLK_DEV_INITRD
++void free_initrd_mem(unsigned long start, unsigned long end)
+{
-+ pte_t *pte;
-+ int rc;
++ free_init_pages("initrd memory", start, end);
++}
++#endif
+
-+ if (xen_feature(feature))
-+ return;
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/mm/ioremap-xen.c b/arch/i386/mm/ioremap-xen.c
+--- a/arch/i386/mm/ioremap-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/mm/ioremap-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,443 @@
++/*
++ * arch/i386/mm/ioremap.c
++ *
++ * Re-map IO memory to kernel address space so that we can access it.
++ * This is needed for high PCI addresses that aren't mapped in the
++ * 640k-1MB IO memory area on PC's
++ *
++ * (C) Copyright 1995 1996 Linus Torvalds
++ */
+
-+ pte = virt_to_ptep(va);
-+ rc = HYPERVISOR_update_va_mapping(
-+ (unsigned long)va, pte_mkwrite(*pte), 0);
-+ BUG_ON(rc);
-+}
++#include <linux/vmalloc.h>
++#include <linux/init.h>
++#include <linux/slab.h>
++#include <linux/module.h>
++#include <linux/io.h>
++#include <asm/fixmap.h>
++#include <asm/cacheflush.h>
++#include <asm/tlbflush.h>
++#include <asm/pgtable.h>
++#include <asm/pgalloc.h>
+
-+void make_page_readonly(void *va, unsigned int feature)
++#define ISA_START_ADDRESS 0x0
++#define ISA_END_ADDRESS 0x100000
++
++static int direct_remap_area_pte_fn(pte_t *pte,
++ struct page *pmd_page,
++ unsigned long address,
++ void *data)
+{
-+ pte_t *pte;
-+ int rc;
++ mmu_update_t **v = (mmu_update_t **)data;
+
-+ if (xen_feature(feature))
-+ return;
++ BUG_ON(!pte_none(*pte));
+
-+ pte = virt_to_ptep(va);
-+ rc = HYPERVISOR_update_va_mapping(
-+ (unsigned long)va, pte_wrprotect(*pte), 0);
-+ if (rc) /* fallback? */
-+ xen_l1_entry_update(pte, pte_wrprotect(*pte));
-+ if ((unsigned long)va >= (unsigned long)high_memory) {
-+ unsigned long pfn = pte_pfn(*pte);
-+#ifdef CONFIG_HIGHMEM
-+ if (pfn >= highstart_pfn)
-+ kmap_flush_unused(); /* flush stale writable kmaps */
-+ else
-+#endif
-+ make_lowmem_page_readonly(
-+ phys_to_virt(pfn << PAGE_SHIFT), feature);
-+ }
++ (*v)->ptr = ((u64)pfn_to_mfn(page_to_pfn(pmd_page)) <<
++ PAGE_SHIFT) | ((unsigned long)pte & ~PAGE_MASK);
++ (*v)++;
++
++ return 0;
+}
+
-+void make_page_writable(void *va, unsigned int feature)
++static int __direct_remap_pfn_range(struct mm_struct *mm,
++ unsigned long address,
++ unsigned long mfn,
++ unsigned long size,
++ pgprot_t prot,
++ domid_t domid)
+{
-+ pte_t *pte;
+ int rc;
++ unsigned long i, start_address;
++ mmu_update_t *u, *v, *w;
+
-+ if (xen_feature(feature))
-+ return;
++ u = v = w = (mmu_update_t *)__get_free_page(GFP_KERNEL|__GFP_REPEAT);
++ if (u == NULL)
++ return -ENOMEM;
+
-+ pte = virt_to_ptep(va);
-+ rc = HYPERVISOR_update_va_mapping(
-+ (unsigned long)va, pte_mkwrite(*pte), 0);
-+ if (rc) /* fallback? */
-+ xen_l1_entry_update(pte, pte_mkwrite(*pte));
-+ if ((unsigned long)va >= (unsigned long)high_memory) {
-+ unsigned long pfn = pte_pfn(*pte);
-+#ifdef CONFIG_HIGHMEM
-+ if (pfn < highstart_pfn)
-+#endif
-+ make_lowmem_page_writable(
-+ phys_to_virt(pfn << PAGE_SHIFT), feature);
-+ }
-+}
++ start_address = address;
+
-+void make_pages_readonly(void *va, unsigned int nr, unsigned int feature)
-+{
-+ if (xen_feature(feature))
-+ return;
++ flush_cache_all();
+
-+ while (nr-- != 0) {
-+ make_page_readonly(va, feature);
-+ va = (void *)((unsigned long)va + PAGE_SIZE);
-+ }
-+}
++ for (i = 0; i < size; i += PAGE_SIZE) {
++ if ((v - u) == (PAGE_SIZE / sizeof(mmu_update_t))) {
++ /* Flush a full batch after filling in the PTE ptrs. */
++ rc = apply_to_page_range(mm, start_address,
++ address - start_address,
++ direct_remap_area_pte_fn, &w);
++ if (rc)
++ goto out;
++ rc = -EFAULT;
++ if (HYPERVISOR_mmu_update(u, v - u, NULL, domid) < 0)
++ goto out;
++ v = w = u;
++ start_address = address;
++ }
+
-+void make_pages_writable(void *va, unsigned int nr, unsigned int feature)
-+{
-+ if (xen_feature(feature))
-+ return;
++ /*
++ * Fill in the machine address: PTE ptr is done later by
++ * __direct_remap_area_pages().
++ */
++ v->val = pte_val_ma(pfn_pte_ma(mfn, prot));
+
-+ while (nr-- != 0) {
-+ make_page_writable(va, feature);
-+ va = (void *)((unsigned long)va + PAGE_SIZE);
++ mfn++;
++ address += PAGE_SIZE;
++ v++;
+ }
-+}
+
-+static inline void pgd_walk_set_prot(void *pt, pgprot_t flags)
-+{
-+ struct page *page = virt_to_page(pt);
-+ unsigned long pfn = page_to_pfn(page);
++ if (v != u) {
++ /* Final batch. */
++ rc = apply_to_page_range(mm, start_address,
++ address - start_address,
++ direct_remap_area_pte_fn, &w);
++ if (rc)
++ goto out;
++ rc = -EFAULT;
++ if (unlikely(HYPERVISOR_mmu_update(u, v - u, NULL, domid) < 0))
++ goto out;
++ }
+
-+ if (PageHighMem(page))
-+ return;
-+ BUG_ON(HYPERVISOR_update_va_mapping(
-+ (unsigned long)__va(pfn << PAGE_SHIFT),
-+ pfn_pte(pfn, flags), 0));
++ rc = 0;
++
++ out:
++ flush_tlb_all();
++
++ free_page((unsigned long)u);
++
++ return rc;
+}
+
-+static void pgd_walk(pgd_t *pgd_base, pgprot_t flags)
++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)
+{
-+ pgd_t *pgd = pgd_base;
-+ pud_t *pud;
-+ pmd_t *pmd;
-+ pte_t *pte;
-+ int g, u, m;
-+
+ if (xen_feature(XENFEAT_auto_translated_physmap))
-+ return;
++ return remap_pfn_range(vma, address, mfn, size, prot);
+
-+ for (g = 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(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 */
-+ pgd_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);
-+ pgd_walk_set_prot(pte,flags);
-+ }
-+ }
-+ }
++ if (domid == DOMID_SELF)
++ return -EINVAL;
+
-+ BUG_ON(HYPERVISOR_update_va_mapping(
-+ (unsigned long)pgd_base,
-+ pfn_pte(virt_to_phys(pgd_base)>>PAGE_SHIFT, flags),
-+ UVMF_TLB_FLUSH));
-+}
++ vma->vm_flags |= VM_IO | VM_RESERVED;
+
-+static void __pgd_pin(pgd_t *pgd)
-+{
-+ pgd_walk(pgd, PAGE_KERNEL_RO);
-+ xen_pgd_pin(__pa(pgd));
-+ set_bit(PG_pinned, &virt_to_page(pgd)->flags);
-+}
++ vma->vm_mm->context.has_foreign_mappings = 1;
+
-+static void __pgd_unpin(pgd_t *pgd)
-+{
-+ xen_pgd_unpin(__pa(pgd));
-+ pgd_walk(pgd, PAGE_KERNEL);
-+ clear_bit(PG_pinned, &virt_to_page(pgd)->flags);
++ return __direct_remap_pfn_range(
++ vma->vm_mm, address, mfn, size, prot, domid);
+}
++EXPORT_SYMBOL(direct_remap_pfn_range);
+
-+static void pgd_test_and_unpin(pgd_t *pgd)
++int direct_kernel_remap_pfn_range(unsigned long address,
++ unsigned long mfn,
++ unsigned long size,
++ pgprot_t prot,
++ domid_t domid)
+{
-+ if (test_bit(PG_pinned, &virt_to_page(pgd)->flags))
-+ __pgd_unpin(pgd);
++ return __direct_remap_pfn_range(
++ &init_mm, address, mfn, size, prot, domid);
+}
++EXPORT_SYMBOL(direct_kernel_remap_pfn_range);
+
-+void mm_pin(struct mm_struct *mm)
++static int lookup_pte_fn(
++ pte_t *pte, struct page *pmd_page, unsigned long addr, void *data)
+{
-+ if (xen_feature(XENFEAT_writable_page_tables))
-+ return;
-+ spin_lock(&mm->page_table_lock);
-+ __pgd_pin(mm->pgd);
-+ spin_unlock(&mm->page_table_lock);
++ uint64_t *ptep = (uint64_t *)data;
++ if (ptep)
++ *ptep = ((uint64_t)pfn_to_mfn(page_to_pfn(pmd_page)) <<
++ PAGE_SHIFT) | ((unsigned long)pte & ~PAGE_MASK);
++ return 0;
+}
+
-+void mm_unpin(struct mm_struct *mm)
++int create_lookup_pte_addr(struct mm_struct *mm,
++ unsigned long address,
++ uint64_t *ptep)
+{
-+ if (xen_feature(XENFEAT_writable_page_tables))
-+ return;
-+ spin_lock(&mm->page_table_lock);
-+ __pgd_unpin(mm->pgd);
-+ spin_unlock(&mm->page_table_lock);
++ return apply_to_page_range(mm, address, PAGE_SIZE,
++ lookup_pte_fn, ptep);
+}
+
-+void mm_pin_all(void)
-+{
-+ struct page *page;
-+
-+ /* Only pgds on the pgd_list please: none hidden in the slab cache. */
-+ kmem_cache_shrink(pgd_cache);
-+
-+ if (xen_feature(XENFEAT_writable_page_tables))
-+ return;
-+
-+ for (page = pgd_list; page; page = (struct page *)page->index) {
-+ if (!test_bit(PG_pinned, &page->flags))
-+ __pgd_pin((pgd_t *)page_address(page));
-+ }
-+}
++EXPORT_SYMBOL(create_lookup_pte_addr);
+
-+void _arch_dup_mmap(struct mm_struct *mm)
++static int noop_fn(
++ pte_t *pte, struct page *pmd_page, unsigned long addr, void *data)
+{
-+ if (!test_bit(PG_pinned, &virt_to_page(mm->pgd)->flags))
-+ mm_pin(mm);
++ return 0;
+}
+
-+void _arch_exit_mmap(struct mm_struct *mm)
++int touch_pte_range(struct mm_struct *mm,
++ unsigned long address,
++ unsigned long size)
+{
-+ 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);
++ return apply_to_page_range(mm, address, size, noop_fn, NULL);
++}
+
-+ switch_mm(mm, &init_mm, tsk);
++EXPORT_SYMBOL(touch_pte_range);
+
-+ atomic_dec(&mm->mm_count);
-+ BUG_ON(atomic_read(&mm->mm_count) == 0);
-+ }
++/*
++ * Does @address reside within a non-highmem page that is local to this virtual
++ * machine (i.e., not an I/O page, nor a memory page belonging to another VM).
++ * See the comment that accompanies mfn_to_local_pfn() in page.h to understand
++ * why this works.
++ */
++static inline int is_local_lowmem(unsigned long address)
++{
++ extern unsigned long max_low_pfn;
++ return (mfn_to_local_pfn(address >> PAGE_SHIFT) < max_low_pfn);
++}
+
-+ task_unlock(tsk);
++/*
++ * Generic mapping function (not visible outside):
++ */
+
-+ if (test_bit(PG_pinned, &virt_to_page(mm->pgd)->flags) &&
-+ (atomic_read(&mm->mm_count) == 1) &&
-+ !mm->context.has_foreign_mappings)
-+ mm_unpin(mm);
-+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/oprofile/Makefile linux-2.6.18-xen/arch/i386/oprofile/Makefile
---- linux-2.6.18.3/arch/i386/oprofile/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/oprofile/Makefile 2006-11-19 14:26:22.000000000 +0100
-@@ -6,7 +6,11 @@
- oprofilefs.o oprofile_stats.o \
- timer_int.o )
-
-+ifdef CONFIG_XEN
-+oprofile-y := $(DRIVER_OBJS) xenoprof.o
-+else
- oprofile-y := $(DRIVER_OBJS) init.o backtrace.o
- oprofile-$(CONFIG_X86_LOCAL_APIC) += nmi_int.o op_model_athlon.o \
- op_model_ppro.o op_model_p4.o
- oprofile-$(CONFIG_X86_IO_APIC) += nmi_timer_int.o
-+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/oprofile/xenoprof.c linux-2.6.18-xen/arch/i386/oprofile/xenoprof.c
---- linux-2.6.18.3/arch/i386/oprofile/xenoprof.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/oprofile/xenoprof.c 2006-11-19 14:26:22.000000000 +0100
-@@ -0,0 +1,584 @@
-+/**
-+ * @file xenoprof.c
-+ *
-+ * @remark Copyright 2002 OProfile authors
-+ * @remark Read the file COPYING
-+ *
-+ * @author John Levon <levon at movementarian.org>
++/*
++ * Remap an arbitrary physical address space into the kernel virtual
++ * address space. Needed when the kernel wants to access high addresses
++ * directly.
+ *
-+ * Modified by Aravind Menon and Jose Renato Santos for Xen
-+ * These modifications are:
-+ * Copyright (C) 2005 Hewlett-Packard Co.
++ * NOTE! We need to allow non-page-aligned mappings too: we will obviously
++ * have to convert them into an offset in a page-aligned mapping, but the
++ * caller shouldn't need to know that small detail.
+ */
++void __iomem * __ioremap(unsigned long phys_addr, unsigned long size, unsigned long flags)
++{
++ void __iomem * addr;
++ struct vm_struct * area;
++ unsigned long offset, last_addr;
++ domid_t domid = DOMID_IO;
+
-+#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/nmi.h>
-+#include <asm/msr.h>
-+#include <asm/apic.h>
-+#include <asm/pgtable.h>
-+#include <xen/evtchn.h>
-+#include "op_counter.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>
-+
-+#define MAX_XENOPROF_SAMPLES 16
++ /* Don't allow wraparound or zero size */
++ last_addr = phys_addr + size - 1;
++ if (!size || last_addr < phys_addr)
++ return NULL;
+
-+static int xenoprof_start(void);
-+static void xenoprof_stop(void);
++ /*
++ * Don't remap the low PCI/ISA area, it's always mapped..
++ */
++ if (is_initial_xendomain() &&
++ phys_addr >= ISA_START_ADDRESS && last_addr < ISA_END_ADDRESS)
++ return (void __iomem *) isa_bus_to_virt(phys_addr);
+
-+static int xenoprof_enabled = 0;
-+static unsigned int num_events = 0;
-+static int is_primary = 0;
-+static int active_defined;
++ /*
++ * Don't allow anybody to remap normal RAM that we're using..
++ */
++ if (is_local_lowmem(phys_addr)) {
++ char *t_addr, *t_end;
++ struct page *page;
+
-+/* sample buffers shared with Xen */
-+xenoprof_buf_t * xenoprof_buf[MAX_VIRT_CPUS];
-+/* Shared buffer area */
-+char * shared_buffer = NULL;
-+/* 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];
++ t_addr = bus_to_virt(phys_addr);
++ t_end = t_addr + (size - 1);
++
++ for(page = virt_to_page(t_addr); page <= virt_to_page(t_end); page++)
++ if(!PageReserved(page))
++ return NULL;
+
-+/* Passive sample buffers shared with Xen */
-+xenoprof_buf_t *p_xenoprof_buf[MAX_OPROF_DOMAINS][MAX_VIRT_CPUS];
-+/* Passive shared buffer area */
-+char *p_shared_buffer[MAX_OPROF_DOMAINS];
++ domid = DOMID_SELF;
++ }
+
-+#ifdef CONFIG_PM
++ /*
++ * Mappings have to be page-aligned
++ */
++ offset = phys_addr & ~PAGE_MASK;
++ phys_addr &= PAGE_MASK;
++ size = PAGE_ALIGN(last_addr+1) - phys_addr;
+
-+static int xenoprof_suspend(struct sys_device * dev, pm_message_t state)
-+{
-+ if (xenoprof_enabled == 1)
-+ xenoprof_stop();
-+ return 0;
++ /*
++ * Ok, go for it..
++ */
++ area = get_vm_area(size, VM_IOREMAP | (flags << 20));
++ if (!area)
++ return NULL;
++ area->phys_addr = phys_addr;
++ addr = (void __iomem *) area->addr;
++ flags |= _PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED;
++ if (__direct_remap_pfn_range(&init_mm, (unsigned long)addr,
++ phys_addr>>PAGE_SHIFT,
++ size, __pgprot(flags), domid)) {
++ vunmap((void __force *) addr);
++ return NULL;
++ }
++ return (void __iomem *) (offset + (char __iomem *)addr);
+}
++EXPORT_SYMBOL(__ioremap);
+
++/**
++ * ioremap_nocache - map bus memory into CPU space
++ * @offset: bus address of the memory
++ * @size: size of the resource to map
++ *
++ * ioremap_nocache 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.
++ *
++ * This version of ioremap ensures that the memory is marked uncachable
++ * on the CPU as well as honouring existing caching rules from things like
++ * the PCI bus. Note that there are other caches and buffers on many
++ * busses. In particular driver authors should read up on PCI writes
++ *
++ * It's useful if some control registers are in such an area and
++ * write combining or read caching is not desirable:
++ *
++ * Must be freed with iounmap.
++ */
+
-+static int xenoprof_resume(struct sys_device * dev)
++void __iomem *ioremap_nocache (unsigned long phys_addr, unsigned long size)
+{
-+ if (xenoprof_enabled == 1)
-+ xenoprof_start();
-+ return 0;
-+}
-+
++ unsigned long last_addr;
++ void __iomem *p = __ioremap(phys_addr, size, _PAGE_PCD);
++ if (!p)
++ return p;
+
-+static struct sysdev_class oprofile_sysclass = {
-+ set_kset_name("oprofile"),
-+ .resume = xenoprof_resume,
-+ .suspend = xenoprof_suspend
-+};
++ /* Guaranteed to be > phys_addr, as per __ioremap() */
++ last_addr = phys_addr + size - 1;
+
++ if (is_local_lowmem(last_addr)) {
++ struct page *ppage = virt_to_page(bus_to_virt(phys_addr));
++ unsigned long npages;
+
-+static struct sys_device device_oprofile = {
-+ .id = 0,
-+ .cls = &oprofile_sysclass,
-+};
++ phys_addr &= PAGE_MASK;
+
++ /* This might overflow and become zero.. */
++ last_addr = PAGE_ALIGN(last_addr);
+
-+static int __init init_driverfs(void)
-+{
-+ int error;
-+ if (!(error = sysdev_class_register(&oprofile_sysclass)))
-+ error = sysdev_register(&device_oprofile);
-+ return error;
-+}
++ /* .. but that's ok, because modulo-2**n arithmetic will make
++ * the page-aligned "last - first" come out right.
++ */
++ npages = (last_addr - phys_addr) >> PAGE_SHIFT;
+
++ if (change_page_attr(ppage, npages, PAGE_KERNEL_NOCACHE) < 0) {
++ iounmap(p);
++ p = NULL;
++ }
++ global_flush_tlb();
++ }
+
-+static void __exit exit_driverfs(void)
-+{
-+ sysdev_unregister(&device_oprofile);
-+ sysdev_class_unregister(&oprofile_sysclass);
++ return p;
+}
++EXPORT_SYMBOL(ioremap_nocache);
+
-+#else
-+#define init_driverfs() do { } while (0)
-+#define exit_driverfs() do { } while (0)
-+#endif /* CONFIG_PM */
-+
-+unsigned long long oprofile_samples = 0;
-+unsigned long long p_oprofile_samples = 0;
++/**
++ * iounmap - Free a IO remapping
++ * @addr: virtual address from ioremap_*
++ *
++ * Caller must ensure there is only one unmapping for the same pointer.
++ */
++void iounmap(volatile void __iomem *addr)
++{
++ struct vm_struct *p, *o;
+
-+unsigned int pdomains;
-+struct xenoprof_passive passive_domains[MAX_OPROF_DOMAINS];
++ if ((void __force *)addr <= high_memory)
++ return;
+
-+static void xenoprof_add_pc(xenoprof_buf_t *buf, int is_passive)
-+{
-+ int head, tail, size;
++ /*
++ * __ioremap special-cases the PCI/ISA range by not instantiating a
++ * vm_area and by simply returning an address into the kernel mapping
++ * of ISA space. So handle that here.
++ */
++ if ((unsigned long) addr >= fix_to_virt(FIX_ISAMAP_BEGIN))
++ return;
+
-+ head = buf->event_head;
-+ tail = buf->event_tail;
-+ size = buf->event_size;
++ addr = (volatile void __iomem *)(PAGE_MASK & (unsigned long __force)addr);
+
-+ 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;
++ /* Use the vm area unlocked, assuming the caller
++ ensures there isn't another iounmap for the same address
++ in parallel. Reuse of the virtual address is prevented by
++ leaving it in the global lists until we're done with it.
++ cpa takes care of the direct mappings. */
++ read_lock(&vmlist_lock);
++ for (p = vmlist; p; p = p->next) {
++ if (p->addr == addr)
++ break;
+ }
-+ 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++;
++ read_unlock(&vmlist_lock);
++
++ if (!p) {
++ printk("iounmap: bad address %p\n", addr);
++ dump_stack();
++ return;
+ }
+
-+ buf->event_tail = tail;
-+}
++ /* Reset the direct mapping. Can block */
++ if ((p->flags >> 20) && is_local_lowmem(p->phys_addr)) {
++ /* p->size includes the guard page, but cpa doesn't like that */
++ change_page_attr(virt_to_page(bus_to_virt(p->phys_addr)),
++ (p->size - PAGE_SIZE) >> PAGE_SHIFT,
++ PAGE_KERNEL);
++ global_flush_tlb();
++ }
+
-+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);
++ /* Finally remove it */
++ o = remove_vm_area((void *)addr);
++ BUG_ON(p != o || o == NULL);
++ kfree(p);
+}
++EXPORT_SYMBOL(iounmap);
+
-+static irqreturn_t
-+xenoprof_ovf_interrupt(int irq, void * dev_id, struct pt_regs * regs)
++void __init *bt_ioremap(unsigned long phys_addr, unsigned long size)
+{
-+ struct xenoprof_buf * buf;
-+ int cpu;
-+ static unsigned long flag;
-+
-+ cpu = smp_processor_id();
-+ buf = xenoprof_buf[cpu];
-+
-+ xenoprof_add_pc(buf, 0);
++ unsigned long offset, last_addr;
++ unsigned int nrpages;
++ enum fixed_addresses idx;
+
-+ if (is_primary && !test_and_set_bit(0, &flag)) {
-+ xenoprof_handle_passive();
-+ smp_mb__before_clear_bit();
-+ clear_bit(0, &flag);
-+ }
++ /* Don't allow wraparound or zero size */
++ last_addr = phys_addr + size - 1;
++ if (!size || last_addr < phys_addr)
++ return NULL;
+
-+ return IRQ_HANDLED;
-+}
++ /*
++ * Don't remap the low PCI/ISA area, it's always mapped..
++ */
++ if (is_initial_xendomain() &&
++ phys_addr >= ISA_START_ADDRESS && last_addr < ISA_END_ADDRESS)
++ return isa_bus_to_virt(phys_addr);
+
++ /*
++ * Mappings have to be page-aligned
++ */
++ offset = phys_addr & ~PAGE_MASK;
++ phys_addr &= PAGE_MASK;
++ size = PAGE_ALIGN(last_addr) - phys_addr;
+
-+static void unbind_virq(void)
-+{
-+ int i;
++ /*
++ * Mappings have to fit in the FIX_BTMAP area.
++ */
++ nrpages = size >> PAGE_SHIFT;
++ if (nrpages > NR_FIX_BTMAPS)
++ return NULL;
+
-+ for_each_possible_cpu(i) {
-+ if (ovf_irq[i] >= 0) {
-+ unbind_from_irqhandler(ovf_irq[i], NULL);
-+ ovf_irq[i] = -1;
-+ }
++ /*
++ * Ok, go for it..
++ */
++ idx = FIX_BTMAP_BEGIN;
++ while (nrpages > 0) {
++ set_fixmap(idx, phys_addr);
++ phys_addr += PAGE_SIZE;
++ --idx;
++ --nrpages;
+ }
++ return (void*) (offset + fix_to_virt(FIX_BTMAP_BEGIN));
+}
+
-+
-+static int bind_virq(void)
++void __init bt_iounmap(void *addr, unsigned long size)
+{
-+ int i, result;
-+
-+ for_each_possible_cpu(i) {
-+ result = bind_virq_to_irqhandler(VIRQ_XENOPROF,
-+ i,
-+ xenoprof_ovf_interrupt,
-+ SA_INTERRUPT,
-+ "xenoprof",
-+ NULL);
++ unsigned long virt_addr;
++ unsigned long offset;
++ unsigned int nrpages;
++ enum fixed_addresses idx;
+
-+ if (result < 0) {
-+ unbind_virq();
-+ return result;
-+ }
++ virt_addr = (unsigned long)addr;
++ if (virt_addr < fix_to_virt(FIX_BTMAP_BEGIN))
++ return;
++ if (virt_addr >= fix_to_virt(FIX_ISAMAP_BEGIN))
++ return;
++ offset = virt_addr & ~PAGE_MASK;
++ nrpages = PAGE_ALIGN(offset + size - 1) >> PAGE_SHIFT;
+
-+ ovf_irq[i] = result;
++ idx = FIX_BTMAP_BEGIN;
++ while (nrpages > 0) {
++ clear_fixmap(idx);
++ --idx;
++ --nrpages;
+ }
-+
-+ return 0;
+}
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/mm/Makefile b/arch/i386/mm/Makefile
+--- a/arch/i386/mm/Makefile 2007-03-12 21:58:05.000000000 +0100
++++ b/arch/i386/mm/Makefile 2007-03-14 10:55:14.000000000 +0100
+@@ -8,3 +8,11 @@
+ 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
+
-+static int map_xenoprof_buffer(int max_samples)
-+{
-+ struct xenoprof_get_buffer get_buffer;
-+ struct xenoprof_buf *buf;
-+ int npages, ret, i;
-+ struct vm_struct *area;
++obj-y += hypervisor.o
+
-+ if ( shared_buffer )
-+ return 0;
++obj-y := $(call cherrypickxen, $(obj-y))
++endif
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/mm/pageattr.c b/arch/i386/mm/pageattr.c
+--- a/arch/i386/mm/pageattr.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/mm/pageattr.c 2007-03-14 10:55:14.000000000 +0100
+@@ -90,7 +90,7 @@
+ unsigned long flags;
+
+ set_pte_atomic(kpte, pte); /* change init_mm */
+- if (PTRS_PER_PMD > 1)
++ if (HAVE_SHARED_KERNEL_PMD)
+ return;
+
+ spin_lock_irqsave(&pgd_lock, flags);
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/mm/pgtable-xen.c b/arch/i386/mm/pgtable-xen.c
+--- a/arch/i386/mm/pgtable-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/mm/pgtable-xen.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,724 @@
++/*
++ * linux/arch/i386/mm/pgtable.c
++ */
+
-+ get_buffer.max_samples = max_samples;
++#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>
+
-+ if ( (ret = HYPERVISOR_xenoprof_op(XENOPROF_get_buffer, &get_buffer)) )
-+ return ret;
++#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>
+
-+ nbuf = get_buffer.nbuf;
-+ npages = (get_buffer.bufsize * nbuf - 1) / PAGE_SIZE + 1;
++#include <xen/features.h>
++#include <xen/foreign_page.h>
++#include <asm/hypervisor.h>
+
-+ area = alloc_vm_area(npages * PAGE_SIZE);
-+ if (area == NULL)
-+ return -ENOMEM;
++static void pgd_test_and_unpin(pgd_t *pgd);
+
-+ if ( (ret = direct_kernel_remap_pfn_range(
-+ (unsigned long)area->addr,
-+ get_buffer.buf_maddr >> PAGE_SHIFT,
-+ npages * PAGE_SIZE, __pgprot(_KERNPG_TABLE), DOMID_SELF)) ) {
-+ vunmap(area->addr);
-+ return ret;
-+ }
++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;
+
-+ shared_buffer = area->addr;
-+ for (i=0; i< nbuf; i++) {
-+ buf = (struct xenoprof_buf*)
-+ &shared_buffer[i * get_buffer.bufsize];
-+ BUG_ON(buf->vcpu_id >= MAX_VIRT_CPUS);
-+ xenoprof_buf[buf->vcpu_id] = buf;
++ 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) {
++ pgdat_resize_lock(pgdat, &flags);
++ for (i = 0; i < pgdat->node_spanned_pages; ++i) {
++ page = pgdat_page_nr(pgdat, i);
++ total++;
++ if (PageHighMem(page))
++ highmem++;
++ if (PageReserved(page))
++ reserved++;
++ else if (PageSwapCache(page))
++ cached++;
++ else if (page_count(page))
++ shared += page_count(page) - 1;
++ }
++ pgdat_resize_unlock(pgdat, &flags);
+ }
++ printk(KERN_INFO "%d pages of RAM\n", total);
++ printk(KERN_INFO "%d pages of HIGHMEM\n", highmem);
++ printk(KERN_INFO "%d reserved pages\n", reserved);
++ printk(KERN_INFO "%d pages shared\n", shared);
++ printk(KERN_INFO "%d pages swap cached\n", cached);
+
-+ return 0;
++ printk(KERN_INFO "%lu pages dirty\n", global_page_state(NR_FILE_DIRTY));
++ printk(KERN_INFO "%lu pages writeback\n",
++ global_page_state(NR_WRITEBACK));
++ printk(KERN_INFO "%lu pages mapped\n", global_page_state(NR_FILE_MAPPED));
++ printk(KERN_INFO "%lu pages slab\n",
++ global_page_state(NR_SLAB_RECLAIMABLE) +
++ global_page_state(NR_SLAB_UNRECLAIMABLE));
++ printk(KERN_INFO "%lu pages pagetables\n",
++ global_page_state(NR_PAGETABLE));
+}
+
-+
-+static int xenoprof_setup(void)
++/*
++ * 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)
+{
-+ int ret;
-+ int i;
-+
-+ if ( (ret = map_xenoprof_buffer(MAX_XENOPROF_SAMPLES)) )
-+ return ret;
-+
-+ if ( (ret = bind_virq()) )
-+ return ret;
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
+
-+ if (is_primary) {
-+ struct xenoprof_counter counter;
++ 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);
+
-+ /* 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;
-+ }
++ /*
++ * It's enough to flush this one mapping.
++ * (PGE mappings get flushed as well)
++ */
++ __flush_tlb_one(vaddr);
++}
+
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_reserve_counters, NULL);
-+ if (ret)
-+ goto err;
-+ for (i=0; i<num_events; i++) {
-+ counter.ind = i;
-+ counter.count = (uint64_t)counter_config[i].count;
-+ counter.enabled = (uint32_t)counter_config[i].enabled;
-+ counter.event = (uint32_t)counter_config[i].event;
-+ 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);
-+ }
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_setup_events, NULL);
++/*
++ * 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;
+
-+ if (ret)
-+ goto err;
++ 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);
++ /* <pfn,flags> stored as-is, to permit clearing entries */
++ set_pte(pte, pfn_pte_ma(pfn, flags));
+
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_enable_virq, NULL);
-+ if (ret)
-+ goto err;
-+
-+ xenoprof_enabled = 1;
-+ return 0;
-+ err:
-+ unbind_virq();
-+ return ret;
++ /*
++ * It's enough to flush this one mapping.
++ * (PGE mappings get flushed as well)
++ */
++ __flush_tlb_one(vaddr);
+}
+
-+
-+static void xenoprof_shutdown(void)
++/*
++ * 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.
++ * The pmd must already be instantiated. Assumes PAE mode.
++ */
++void set_pmd_pfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags)
+{
-+ xenoprof_enabled = 0;
-+
-+ HYPERVISOR_xenoprof_op(XENOPROF_disable_virq, NULL);
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
+
-+ if (is_primary) {
-+ HYPERVISOR_xenoprof_op(XENOPROF_release_counters, NULL);
-+ active_defined = 0;
++ if (vaddr & (PMD_SIZE-1)) { /* vaddr is misaligned */
++ printk(KERN_WARNING "set_pmd_pfn: vaddr misaligned\n");
++ return; /* BUG(); */
+ }
-+
-+ unbind_virq();
-+
++ if (pfn & (PTRS_PER_PTE-1)) { /* pfn is misaligned */
++ printk(KERN_WARNING "set_pmd_pfn: pfn misaligned\n");
++ return; /* BUG(); */
++ }
++ pgd = swapper_pg_dir + pgd_index(vaddr);
++ if (pgd_none(*pgd)) {
++ printk(KERN_WARNING "set_pmd_pfn: pgd_none\n");
++ return; /* BUG(); */
++ }
++ pud = pud_offset(pgd, vaddr);
++ pmd = pmd_offset(pud, vaddr);
++ set_pmd(pmd, pfn_pmd(pfn, flags));
++ /*
++ * It's enough to flush this one mapping.
++ * (PGE mappings get flushed as well)
++ */
++ __flush_tlb_one(vaddr);
+}
+
++static int fixmaps;
++#ifndef CONFIG_COMPAT_VDSO
++unsigned long __FIXADDR_TOP = (HYPERVISOR_VIRT_START - 2 * PAGE_SIZE);
++EXPORT_SYMBOL(__FIXADDR_TOP);
++#endif
+
-+static int xenoprof_start(void)
++void __set_fixmap (enum fixed_addresses idx, maddr_t phys, pgprot_t flags)
+{
-+ int ret = 0;
-+
-+ if (is_primary)
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_start, NULL);
++ unsigned long address = __fix_to_virt(idx);
+
-+ return ret;
++ if (idx >= __end_of_fixed_addresses) {
++ BUG();
++ return;
++ }
++ switch (idx) {
++ case FIX_WP_TEST:
++#ifdef CONFIG_X86_F00F_BUG
++ case FIX_F00F_IDT:
++#endif
++ set_pte_pfn(address, phys >> PAGE_SHIFT, flags);
++ break;
++ default:
++ set_pte_pfn_ma(address, phys >> PAGE_SHIFT, flags);
++ break;
++ }
++ fixmaps++;
+}
+
-+
-+static void xenoprof_stop(void)
++/**
++ * reserve_top_address - reserves a hole in the top of kernel address space
++ * @reserve - size of hole to reserve
++ *
++ * Can be used to relocate the fixmap area and poke a hole in the top
++ * of kernel address space to make room for a hypervisor.
++ */
++void reserve_top_address(unsigned long reserve)
+{
-+ if (is_primary)
-+ HYPERVISOR_xenoprof_op(XENOPROF_stop, NULL);
++ BUG_ON(fixmaps > 0);
++#ifdef CONFIG_COMPAT_VDSO
++ BUG_ON(reserve != 0);
++#else
++ __FIXADDR_TOP = -reserve - PAGE_SIZE;
++ __VMALLOC_RESERVE += reserve;
++#endif
+}
+
-+
-+static int xenoprof_set_active(int * active_domains,
-+ unsigned int adomains)
++void set_fixaddr_top(unsigned long top)
+{
-+ int ret = 0;
-+ int i;
-+ int set_dom0 = 0;
-+ domid_t domid;
-+
-+ if (!is_primary)
-+ return 0;
++ BUG_ON(fixmaps > 0);
++ __FIXADDR_TOP = top - PAGE_SIZE;
++}
+
-+ if (adomains > MAX_OPROF_DOMAINS)
-+ return -E2BIG;
++pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
++{
++ pte_t *pte = (pte_t *)__get_free_page(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO);
++ if (pte)
++ make_lowmem_page_readonly(pte, XENFEAT_writable_page_tables);
++ return pte;
++}
+
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_reset_active_list, NULL);
-+ if (ret)
-+ return ret;
++struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
++{
++ struct page *pte;
+
-+ 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);
++#ifdef CONFIG_HIGHPTE
++ pte = alloc_pages(GFP_KERNEL|__GFP_HIGHMEM|__GFP_REPEAT|__GFP_ZERO, 0);
++#else
++ pte = alloc_pages(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO, 0);
++ if (pte) {
++ SetPageForeign(pte, pte_free);
++ init_page_count(pte);
+ }
-+
-+out:
-+ if (ret)
-+ HYPERVISOR_xenoprof_op(XENOPROF_reset_active_list, NULL);
-+ active_defined = !ret;
-+ return ret;
++#endif
++ return pte;
+}
+
-+static int xenoprof_set_passive(int * p_domains,
-+ unsigned int pdoms)
++void pte_free(struct page *pte)
+{
-+ int ret;
-+ int i, j;
-+ int npages;
-+ struct xenoprof_buf *buf;
-+ struct vm_struct *area;
-+ pgprot_t prot = __pgprot(_KERNPG_TABLE);
-+
-+ if (!is_primary)
-+ return 0;
-+
-+ if (pdoms > MAX_OPROF_DOMAINS)
-+ return -E2BIG;
-+
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_reset_passive_list, NULL);
-+ if (ret)
-+ return ret;
-+
-+ for (i = 0; i < pdoms; i++) {
-+ passive_domains[i].domain_id = p_domains[i];
-+ passive_domains[i].max_samples = 2048;
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_set_passive,
-+ &passive_domains[i]);
-+ if (ret)
-+ goto out;
-+
-+ npages = (passive_domains[i].bufsize * passive_domains[i].nbuf - 1) / PAGE_SIZE + 1;
-+
-+ area = alloc_vm_area(npages * PAGE_SIZE);
-+ if (area == NULL) {
-+ ret = -ENOMEM;
-+ goto out;
-+ }
-+
-+ ret = direct_kernel_remap_pfn_range(
-+ (unsigned long)area->addr,
-+ passive_domains[i].buf_maddr >> PAGE_SHIFT,
-+ npages * PAGE_SIZE, prot, DOMID_SELF);
-+ if (ret) {
-+ vunmap(area->addr);
-+ goto out;
-+ }
++ unsigned long va = (unsigned long)__va(page_to_pfn(pte)<<PAGE_SHIFT);
+
-+ p_shared_buffer[i] = area->addr;
++ if (!pte_write(*virt_to_ptep(va)))
++ BUG_ON(HYPERVISOR_update_va_mapping(
++ va, pfn_pte(page_to_pfn(pte), PAGE_KERNEL), 0));
+
-+ for (j = 0; j < passive_domains[i].nbuf; j++) {
-+ buf = (struct xenoprof_buf *)
-+ &p_shared_buffer[i][j * passive_domains[i].bufsize];
-+ BUG_ON(buf->vcpu_id >= MAX_VIRT_CPUS);
-+ p_xenoprof_buf[i][buf->vcpu_id] = buf;
-+ }
++ ClearPageForeign(pte);
++ init_page_count(pte);
+
-+ }
++ __free_page(pte);
++}
+
-+ pdomains = pdoms;
-+ return 0;
++void pmd_ctor(void *pmd, struct kmem_cache *cache, unsigned long flags)
++{
++ memset(pmd, 0, PTRS_PER_PMD*sizeof(pmd_t));
++}
+
-+out:
-+ for (j = 0; j < i; j++) {
-+ vunmap(p_shared_buffer[j]);
-+ p_shared_buffer[j] = NULL;
-+ }
++/*
++ * List of all pgd's needed for non-PAE so it can invalidate entries
++ * in both cached and uncached pgd's; not needed for PAE since the
++ * kernel pmd is shared. If PAE were not to share the pmd a similar
++ * tactic would be needed. This is essentially codepath-based locking
++ * against pageattr.c; it is the unique case in which a valid change
++ * of kernel pagetables can't be lazily synchronized by vmalloc faults.
++ * vmalloc faults work because attached pagetables are never freed.
++ * The locking scheme was chosen on the basis of manfred's
++ * recommendations and having no core impact whatsoever.
++ * -- wli
++ */
++DEFINE_SPINLOCK(pgd_lock);
++struct page *pgd_list;
+
-+ return ret;
++static inline void pgd_list_add(pgd_t *pgd)
++{
++ struct page *page = virt_to_page(pgd);
++ page->index = (unsigned long)pgd_list;
++ if (pgd_list)
++ set_page_private(pgd_list, (unsigned long)&page->index);
++ pgd_list = page;
++ set_page_private(page, (unsigned long)&pgd_list);
+}
+
-+struct op_counter_config counter_config[OP_MAX_COUNTER];
++static inline void pgd_list_del(pgd_t *pgd)
++{
++ struct page *next, **pprev, *page = virt_to_page(pgd);
++ next = (struct page *)page->index;
++ pprev = (struct page **)page_private(page);
++ *pprev = next;
++ if (next)
++ set_page_private(next, (unsigned long)pprev);
++}
+
-+static int xenoprof_create_files(struct super_block * sb, struct dentry * root)
++void pgd_ctor(void *pgd, struct kmem_cache *cache, unsigned long unused)
+{
-+ unsigned int i;
++ unsigned long flags;
+
-+ for (i = 0; i < num_events; ++i) {
-+ struct dentry * dir;
-+ char buf[2];
-+
-+ snprintf(buf, 2, "%d", i);
-+ dir = oprofilefs_mkdir(sb, root, buf);
-+ oprofilefs_create_ulong(sb, dir, "enabled",
-+ &counter_config[i].enabled);
-+ oprofilefs_create_ulong(sb, dir, "event",
-+ &counter_config[i].event);
-+ oprofilefs_create_ulong(sb, dir, "count",
-+ &counter_config[i].count);
-+ oprofilefs_create_ulong(sb, dir, "unit_mask",
-+ &counter_config[i].unit_mask);
-+ oprofilefs_create_ulong(sb, dir, "kernel",
-+ &counter_config[i].kernel);
-+ oprofilefs_create_ulong(sb, dir, "user",
-+ &counter_config[i].user);
++ if (PTRS_PER_PMD > 1) {
++ if (HAVE_SHARED_KERNEL_PMD)
++ clone_pgd_range((pgd_t *)pgd + USER_PTRS_PER_PGD,
++ swapper_pg_dir + USER_PTRS_PER_PGD,
++ KERNEL_PGD_PTRS);
++ } else {
++ 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);
++ memset(pgd, 0, USER_PTRS_PER_PGD*sizeof(pgd_t));
++ pgd_list_add(pgd);
++ spin_unlock_irqrestore(&pgd_lock, flags);
+ }
-+
-+ return 0;
+}
+
-+
-+struct oprofile_operations xenoprof_ops = {
-+ .create_files = xenoprof_create_files,
-+ .set_active = xenoprof_set_active,
-+ .set_passive = xenoprof_set_passive,
-+ .setup = xenoprof_setup,
-+ .shutdown = xenoprof_shutdown,
-+ .start = xenoprof_start,
-+ .stop = xenoprof_stop
-+};
-+
-+
-+/* in order to get driverfs right */
-+static int using_xenoprof;
-+
-+int __init oprofile_arch_init(struct oprofile_operations * ops)
++/* never called when PTRS_PER_PMD > 1 */
++void pgd_dtor(void *pgd, struct kmem_cache *cache, unsigned long unused)
+{
-+ struct xenoprof_init init;
-+ int ret, i;
-+
-+ ret = HYPERVISOR_xenoprof_op(XENOPROF_init, &init);
++ unsigned long flags; /* can be called from interrupt context */
+
-+ if (!ret) {
-+ num_events = init.num_events;
-+ is_primary = init.is_primary;
++ spin_lock_irqsave(&pgd_lock, flags);
++ pgd_list_del(pgd);
++ spin_unlock_irqrestore(&pgd_lock, flags);
+
-+ /* just in case - make sure we do not overflow event list
-+ (i.e. counter_config list) */
-+ if (num_events > OP_MAX_COUNTER)
-+ num_events = OP_MAX_COUNTER;
++ pgd_test_and_unpin(pgd);
++}
+
-+ /* 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;
++pgd_t *pgd_alloc(struct mm_struct *mm)
++{
++ int i;
++ pgd_t *pgd = kmem_cache_alloc(pgd_cache, GFP_KERNEL);
++ pmd_t **pmd;
++ unsigned long flags;
+
-+ init_driverfs();
-+ using_xenoprof = 1;
-+ *ops = xenoprof_ops;
++ pgd_test_and_unpin(pgd);
+
-+ for (i=0; i<NR_CPUS; i++)
-+ ovf_irq[i] = -1;
++ if (PTRS_PER_PMD == 1 || !pgd)
++ return pgd;
+
-+ active_defined = 0;
++ if (HAVE_SHARED_KERNEL_PMD) {
++ for (i = 0; i < USER_PTRS_PER_PGD; ++i) {
++ pmd_t *pmd = kmem_cache_alloc(pmd_cache, GFP_KERNEL);
++ if (!pmd)
++ goto out_oom;
++ set_pgd(&pgd[i], __pgd(1 + __pa(pmd)));
++ }
++ return pgd;
+ }
-+ printk(KERN_INFO "oprofile_arch_init: ret %d, events %d, "
-+ "is_primary %d\n", ret, num_events, is_primary);
-+ return ret;
-+}
-+
-+
-+void __exit oprofile_arch_exit(void)
-+{
-+ int i;
+
-+ if (using_xenoprof)
-+ exit_driverfs();
++ /*
++ * We can race save/restore (if we sleep during a GFP_KERNEL memory
++ * allocation). We therefore store virtual addresses of pmds as they
++ * do not change across save/restore, and poke the machine addresses
++ * into the pgdir under the pgd_lock.
++ */
++ pmd = kmalloc(PTRS_PER_PGD * sizeof(pmd_t *), GFP_KERNEL);
++ if (!pmd) {
++ kmem_cache_free(pgd_cache, pgd);
++ return NULL;
++ }
+
-+ if (shared_buffer) {
-+ vunmap(shared_buffer);
-+ shared_buffer = NULL;
++ /* Allocate pmds, remember virtual addresses. */
++ for (i = 0; i < PTRS_PER_PGD; ++i) {
++ pmd[i] = kmem_cache_alloc(pmd_cache, GFP_KERNEL);
++ if (!pmd[i])
++ goto out_oom;
+ }
-+ if (is_primary) {
-+ for (i = 0; i < pdomains; i++)
-+ if (p_shared_buffer[i]) {
-+ vunmap(p_shared_buffer[i]);
-+ p_shared_buffer[i] = NULL;
-+ }
-+ HYPERVISOR_xenoprof_op(XENOPROF_shutdown, NULL);
-+ }
+
-+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/pci/irq-xen.c linux-2.6.18-xen/arch/i386/pci/irq-xen.c
---- linux-2.6.18.3/arch/i386/pci/irq-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/pci/irq-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,1206 @@
-+/*
-+ * Low-Level PCI Support for PC -- Routing of Interrupts
-+ *
-+ * (c) 1999--2000 Martin Mares <mj at ucw.cz>
-+ */
++ spin_lock_irqsave(&pgd_lock, flags);
+
-+#include <linux/types.h>
-+#include <linux/kernel.h>
-+#include <linux/pci.h>
-+#include <linux/init.h>
-+#include <linux/slab.h>
-+#include <linux/interrupt.h>
-+#include <linux/dmi.h>
-+#include <asm/io.h>
-+#include <asm/smp.h>
-+#include <asm/io_apic.h>
-+#include <linux/irq.h>
-+#include <linux/acpi.h>
++ /* Protect against save/restore: move below 4GB under pgd_lock. */
++ if (!xen_feature(XENFEAT_pae_pgdir_above_4gb)) {
++ int rc = xen_create_contiguous_region(
++ (unsigned long)pgd, 0, 32);
++ if (rc) {
++ spin_unlock_irqrestore(&pgd_lock, flags);
++ goto out_oom;
++ }
++ }
+
-+#include "pci.h"
++ /* Copy kernel pmd contents and write-protect the new pmds. */
++ 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);
++ memcpy(pmd[i], kpmd, PAGE_SIZE);
++ make_lowmem_page_readonly(
++ pmd[i], XENFEAT_writable_page_tables);
++ }
+
-+#define PIRQ_SIGNATURE (('$' << 0) + ('P' << 8) + ('I' << 16) + ('R' << 24))
-+#define PIRQ_VERSION 0x0100
++ /* It is safe to poke machine addresses of pmds under the pmd_lock. */
++ for (i = 0; i < PTRS_PER_PGD; i++)
++ set_pgd(&pgd[i], __pgd(1 + __pa(pmd[i])));
+
-+static int broken_hp_bios_irq9;
-+static int acer_tm360_irqrouting;
++ /* Ensure this pgd gets picked up and pinned on save/restore. */
++ pgd_list_add(pgd);
+
-+static struct irq_routing_table *pirq_table;
++ spin_unlock_irqrestore(&pgd_lock, flags);
+
-+static int pirq_enable_irq(struct pci_dev *dev);
++ kfree(pmd);
+
-+/*
-+ * Never use: 0, 1, 2 (timer, keyboard, and cascade)
-+ * Avoid using: 13, 14 and 15 (FP error and IDE).
-+ * Penalize: 3, 4, 6, 7, 12 (known ISA uses: serial, floppy, parallel and mouse)
-+ */
-+unsigned int pcibios_irq_mask = 0xfff8;
++ return pgd;
+
-+static int pirq_penalty[16] = {
-+ 1000000, 1000000, 1000000, 1000, 1000, 0, 1000, 1000,
-+ 0, 0, 0, 0, 1000, 100000, 100000, 100000
-+};
++out_oom:
++ if (HAVE_SHARED_KERNEL_PMD) {
++ for (i--; i >= 0; i--)
++ kmem_cache_free(pmd_cache,
++ (void *)__va(pgd_val(pgd[i])-1));
++ } else {
++ for (i--; i >= 0; i--)
++ kmem_cache_free(pmd_cache, pmd[i]);
++ kfree(pmd);
++ }
++ kmem_cache_free(pgd_cache, pgd);
++ return NULL;
++}
+
-+struct irq_router {
-+ char *name;
-+ u16 vendor, device;
-+ int (*get)(struct pci_dev *router, struct pci_dev *dev, int pirq);
-+ int (*set)(struct pci_dev *router, struct pci_dev *dev, int pirq, int new);
-+};
++void pgd_free(pgd_t *pgd)
++{
++ int i;
+
-+struct irq_router_handler {
-+ u16 vendor;
-+ int (*probe)(struct irq_router *r, struct pci_dev *router, u16 device);
-+};
++ /*
++ * After this the pgd should not be pinned for the duration of this
++ * function's execution. We should never sleep and thus never race:
++ * 1. User pmds will not become write-protected under our feet due
++ * to a concurrent mm_pin_all().
++ * 2. The machine addresses in PGD entries will not become invalid
++ * due to a concurrent save/restore.
++ */
++ pgd_test_and_unpin(pgd);
+
-+int (*pcibios_enable_irq)(struct pci_dev *dev) = NULL;
-+void (*pcibios_disable_irq)(struct pci_dev *dev) = NULL;
++ /* 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) {
++ pmd_t *pmd = (void *)__va(pgd_val(pgd[i])-1);
++ kmem_cache_free(pmd_cache, pmd);
++ }
+
-+/*
-+ * Check passed address for the PCI IRQ Routing Table signature
-+ * and perform checksum verification.
-+ */
++ if (!HAVE_SHARED_KERNEL_PMD) {
++ unsigned long flags;
++ spin_lock_irqsave(&pgd_lock, flags);
++ pgd_list_del(pgd);
++ spin_unlock_irqrestore(&pgd_lock, flags);
+
-+static inline struct irq_routing_table * pirq_check_routing_table(u8 *addr)
-+{
-+ struct irq_routing_table *rt;
-+ int i;
-+ u8 sum;
++ for (i = USER_PTRS_PER_PGD; i < PTRS_PER_PGD; i++) {
++ pmd_t *pmd = (void *)__va(pgd_val(pgd[i])-1);
++ make_lowmem_page_writable(
++ pmd, XENFEAT_writable_page_tables);
++ memset(pmd, 0, PTRS_PER_PMD*sizeof(pmd_t));
++ kmem_cache_free(pmd_cache, pmd);
++ }
+
-+ rt = (struct irq_routing_table *) addr;
-+ if (rt->signature != PIRQ_SIGNATURE ||
-+ rt->version != PIRQ_VERSION ||
-+ rt->size % 16 ||
-+ rt->size < sizeof(struct irq_routing_table))
-+ return NULL;
-+ sum = 0;
-+ for (i=0; i < rt->size; i++)
-+ sum += addr[i];
-+ if (!sum) {
-+ DBG(KERN_DEBUG "PCI: Interrupt Routing Table found at 0x%p\n", rt);
-+ return rt;
++ if (!xen_feature(XENFEAT_pae_pgdir_above_4gb))
++ xen_destroy_contiguous_region(
++ (unsigned long)pgd, 0);
++ }
+ }
-+ return NULL;
++
++ /* in the non-PAE case, free_pgtables() clears user pgd entries */
++ kmem_cache_free(pgd_cache, pgd);
+}
+
++void make_lowmem_page_readonly(void *va, unsigned int feature)
++{
++ pte_t *pte;
++ int rc;
+
++ if (xen_feature(feature))
++ return;
+
-+/*
-+ * Search 0xf0000 -- 0xfffff for the PCI IRQ Routing Table.
-+ */
++ pte = virt_to_ptep(va);
++ rc = HYPERVISOR_update_va_mapping(
++ (unsigned long)va, pte_wrprotect(*pte), 0);
++ BUG_ON(rc);
++}
+
-+static struct irq_routing_table * __init pirq_find_routing_table(void)
++void make_lowmem_page_writable(void *va, unsigned int feature)
+{
-+ u8 *addr;
-+ struct irq_routing_table *rt;
++ pte_t *pte;
++ int rc;
+
-+#ifdef CONFIG_XEN
-+ if (!is_initial_xendomain())
-+ return NULL;
-+#endif
-+ if (pirq_table_addr) {
-+ rt = pirq_check_routing_table((u8 *) isa_bus_to_virt(pirq_table_addr));
-+ if (rt)
-+ return rt;
-+ printk(KERN_WARNING "PCI: PIRQ table NOT found at pirqaddr\n");
-+ }
-+ for(addr = (u8 *) isa_bus_to_virt(0xf0000); addr < (u8 *) isa_bus_to_virt(0x100000); addr += 16) {
-+ rt = pirq_check_routing_table(addr);
-+ if (rt)
-+ return rt;
-+ }
-+
-+ return NULL;
++ if (xen_feature(feature))
++ return;
++
++ pte = virt_to_ptep(va);
++ rc = HYPERVISOR_update_va_mapping(
++ (unsigned long)va, pte_mkwrite(*pte), 0);
++ BUG_ON(rc);
+}
+
-+/*
-+ * If we have a IRQ routing table, use it to search for peer host
-+ * bridges. It's a gross hack, but since there are no other known
-+ * ways how to get a list of buses, we have to go this way.
-+ */
-+
-+static void __init pirq_peer_trick(void)
++void make_page_readonly(void *va, unsigned int feature)
+{
-+ struct irq_routing_table *rt = pirq_table;
-+ u8 busmap[256];
-+ int i;
-+ struct irq_info *e;
++ pte_t *pte;
++ int rc;
+
-+ memset(busmap, 0, sizeof(busmap));
-+ for(i=0; i < (rt->size - sizeof(struct irq_routing_table)) / sizeof(struct irq_info); i++) {
-+ e = &rt->slots[i];
-+#ifdef DEBUG
-+ {
-+ int j;
-+ DBG(KERN_DEBUG "%02x:%02x slot=%02x", e->bus, e->devfn/8, e->slot);
-+ for(j=0; j<4; j++)
-+ DBG(" %d:%02x/%04x", j, e->irq[j].link, e->irq[j].bitmap);
-+ DBG("\n");
-+ }
++ if (xen_feature(feature))
++ return;
++
++ pte = virt_to_ptep(va);
++ rc = HYPERVISOR_update_va_mapping(
++ (unsigned long)va, pte_wrprotect(*pte), 0);
++ if (rc) /* fallback? */
++ xen_l1_entry_update(pte, pte_wrprotect(*pte));
++ if ((unsigned long)va >= (unsigned long)high_memory) {
++ unsigned long pfn = pte_pfn(*pte);
++#ifdef CONFIG_HIGHMEM
++ if (pfn >= highstart_pfn)
++ kmap_flush_unused(); /* flush stale writable kmaps */
++ else
+#endif
-+ busmap[e->bus] = 1;
-+ }
-+ for(i = 1; i < 256; i++) {
-+ if (!busmap[i] || pci_find_bus(0, i))
-+ continue;
-+ if (pci_scan_bus(i, &pci_root_ops, NULL))
-+ printk(KERN_INFO "PCI: Discovered primary peer bus %02x [IRQ]\n", i);
++ make_lowmem_page_readonly(
++ phys_to_virt(pfn << PAGE_SHIFT), feature);
+ }
-+ pcibios_last_bus = -1;
+}
+
-+/*
-+ * Code for querying and setting of IRQ routes on various interrupt routers.
-+ */
-+
-+void eisa_set_level_irq(unsigned int irq)
++void make_page_writable(void *va, unsigned int feature)
+{
-+ unsigned char mask = 1 << (irq & 7);
-+ unsigned int port = 0x4d0 + (irq >> 3);
-+ unsigned char val;
-+ static u16 eisa_irq_mask;
++ pte_t *pte;
++ int rc;
+
-+ if (irq >= 16 || (1 << irq) & eisa_irq_mask)
++ if (xen_feature(feature))
+ return;
+
-+ eisa_irq_mask |= (1 << irq);
-+ printk(KERN_DEBUG "PCI: setting IRQ %u as level-triggered\n", irq);
-+ val = inb(port);
-+ if (!(val & mask)) {
-+ DBG(KERN_DEBUG " -> edge");
-+ outb(val | mask, port);
++ pte = virt_to_ptep(va);
++ rc = HYPERVISOR_update_va_mapping(
++ (unsigned long)va, pte_mkwrite(*pte), 0);
++ if (rc) /* fallback? */
++ xen_l1_entry_update(pte, pte_mkwrite(*pte));
++ if ((unsigned long)va >= (unsigned long)high_memory) {
++ unsigned long pfn = pte_pfn(*pte);
++#ifdef CONFIG_HIGHMEM
++ if (pfn < highstart_pfn)
++#endif
++ make_lowmem_page_writable(
++ phys_to_virt(pfn << PAGE_SHIFT), feature);
+ }
+}
+
-+/*
-+ * Common IRQ routing practice: nybbles in config space,
-+ * offset by some magic constant.
-+ */
-+static unsigned int read_config_nybble(struct pci_dev *router, unsigned offset, unsigned nr)
++void make_pages_readonly(void *va, unsigned int nr, unsigned int feature)
+{
-+ u8 x;
-+ unsigned reg = offset + (nr >> 1);
++ if (xen_feature(feature))
++ return;
+
-+ pci_read_config_byte(router, reg, &x);
-+ return (nr & 1) ? (x >> 4) : (x & 0xf);
++ while (nr-- != 0) {
++ make_page_readonly(va, feature);
++ va = (void *)((unsigned long)va + PAGE_SIZE);
++ }
+}
+
-+static void write_config_nybble(struct pci_dev *router, unsigned offset, unsigned nr, unsigned int val)
++void make_pages_writable(void *va, unsigned int nr, unsigned int feature)
+{
-+ u8 x;
-+ unsigned reg = offset + (nr >> 1);
++ if (xen_feature(feature))
++ return;
+
-+ pci_read_config_byte(router, reg, &x);
-+ x = (nr & 1) ? ((x & 0x0f) | (val << 4)) : ((x & 0xf0) | val);
-+ pci_write_config_byte(router, reg, x);
++ while (nr-- != 0) {
++ make_page_writable(va, feature);
++ va = (void *)((unsigned long)va + PAGE_SIZE);
++ }
+}
+
-+/*
-+ * ALI pirq entries are damn ugly, and completely undocumented.
-+ * This has been figured out from pirq tables, and it's not a pretty
-+ * picture.
-+ */
-+static int pirq_ali_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
++static inline void pgd_walk_set_prot(void *pt, pgprot_t flags)
+{
-+ static const unsigned char irqmap[16] = { 0, 9, 3, 10, 4, 5, 7, 6, 1, 11, 0, 12, 0, 14, 0, 15 };
++ struct page *page = virt_to_page(pt);
++ unsigned long pfn = page_to_pfn(page);
+
-+ return irqmap[read_config_nybble(router, 0x48, pirq-1)];
++ if (PageHighMem(page))
++ return;
++ BUG_ON(HYPERVISOR_update_va_mapping(
++ (unsigned long)__va(pfn << PAGE_SHIFT),
++ pfn_pte(pfn, flags), 0));
+}
+
-+static int pirq_ali_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++static void pgd_walk(pgd_t *pgd_base, pgprot_t flags)
+{
-+ static const unsigned char irqmap[16] = { 0, 8, 0, 2, 4, 5, 7, 6, 0, 1, 3, 9, 11, 0, 13, 15 };
-+ unsigned int val = irqmap[irq];
-+
-+ if (val) {
-+ write_config_nybble(router, 0x48, pirq-1, val);
-+ return 1;
-+ }
-+ return 0;
-+}
++ pgd_t *pgd = pgd_base;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
++ int g, u, m;
+
-+/*
-+ * The Intel PIIX4 pirq rules are fairly simple: "pirq" is
-+ * just a pointer to the config space.
-+ */
-+static int pirq_piix_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
-+{
-+ u8 x;
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return;
+
-+ pci_read_config_byte(router, pirq, &x);
-+ return (x < 16) ? x : 0;
-+}
++ for (g = 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(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 */
++ pgd_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);
++ pgd_walk_set_prot(pte,flags);
++ }
++ }
++ }
+
-+static int pirq_piix_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
-+{
-+ pci_write_config_byte(router, pirq, irq);
-+ return 1;
++ BUG_ON(HYPERVISOR_update_va_mapping(
++ (unsigned long)pgd_base,
++ pfn_pte(virt_to_phys(pgd_base)>>PAGE_SHIFT, flags),
++ UVMF_TLB_FLUSH));
+}
+
-+/*
-+ * The VIA pirq rules are nibble-based, like ALI,
-+ * but without the ugly irq number munging.
-+ * However, PIRQD is in the upper instead of lower 4 bits.
-+ */
-+static int pirq_via_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
++static void __pgd_pin(pgd_t *pgd)
+{
-+ return read_config_nybble(router, 0x55, pirq == 4 ? 5 : pirq);
++ pgd_walk(pgd, PAGE_KERNEL_RO);
++ xen_pgd_pin(__pa(pgd));
++ set_bit(PG_pinned, &virt_to_page(pgd)->flags);
+}
+
-+static int pirq_via_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++static void __pgd_unpin(pgd_t *pgd)
+{
-+ write_config_nybble(router, 0x55, pirq == 4 ? 5 : pirq, irq);
-+ return 1;
++ xen_pgd_unpin(__pa(pgd));
++ pgd_walk(pgd, PAGE_KERNEL);
++ clear_bit(PG_pinned, &virt_to_page(pgd)->flags);
+}
+
-+/*
-+ * The VIA pirq rules are nibble-based, like ALI,
-+ * but without the ugly irq number munging.
-+ * However, for 82C586, nibble map is different .
-+ */
-+static int pirq_via586_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
++static void pgd_test_and_unpin(pgd_t *pgd)
+{
-+ static const unsigned int pirqmap[5] = { 3, 2, 5, 1, 1 };
-+ return read_config_nybble(router, 0x55, pirqmap[pirq-1]);
++ if (test_bit(PG_pinned, &virt_to_page(pgd)->flags))
++ __pgd_unpin(pgd);
+}
+
-+static int pirq_via586_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++void mm_pin(struct mm_struct *mm)
+{
-+ static const unsigned int pirqmap[5] = { 3, 2, 5, 1, 1 };
-+ write_config_nybble(router, 0x55, pirqmap[pirq-1], irq);
-+ return 1;
++ if (xen_feature(XENFEAT_writable_page_tables))
++ return;
++ spin_lock(&mm->page_table_lock);
++ __pgd_pin(mm->pgd);
++ spin_unlock(&mm->page_table_lock);
+}
+
-+/*
-+ * ITE 8330G pirq rules are nibble-based
-+ * FIXME: pirqmap may be { 1, 0, 3, 2 },
-+ * 2+3 are both mapped to irq 9 on my system
-+ */
-+static int pirq_ite_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
++void mm_unpin(struct mm_struct *mm)
+{
-+ static const unsigned char pirqmap[4] = { 1, 0, 2, 3 };
-+ return read_config_nybble(router,0x43, pirqmap[pirq-1]);
++ if (xen_feature(XENFEAT_writable_page_tables))
++ return;
++ spin_lock(&mm->page_table_lock);
++ __pgd_unpin(mm->pgd);
++ spin_unlock(&mm->page_table_lock);
+}
+
-+static int pirq_ite_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++void mm_pin_all(void)
+{
-+ static const unsigned char pirqmap[4] = { 1, 0, 2, 3 };
-+ write_config_nybble(router, 0x43, pirqmap[pirq-1], irq);
-+ return 1;
-+}
++ struct page *page;
+
-+/*
-+ * OPTI: high four bits are nibble pointer..
-+ * I wonder what the low bits do?
-+ */
-+static int pirq_opti_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
-+{
-+ return read_config_nybble(router, 0xb8, pirq >> 4);
-+}
++ /* Only pgds on the pgd_list please: none hidden in the slab cache. */
++ kmem_cache_shrink(pgd_cache);
+
-+static int pirq_opti_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
-+{
-+ write_config_nybble(router, 0xb8, pirq >> 4, irq);
-+ return 1;
++ if (xen_feature(XENFEAT_writable_page_tables))
++ return;
++
++ for (page = pgd_list; page; page = (struct page *)page->index) {
++ if (!test_bit(PG_pinned, &page->flags))
++ __pgd_pin((pgd_t *)page_address(page));
++ }
+}
+
-+/*
-+ * Cyrix: nibble offset 0x5C
-+ * 0x5C bits 7:4 is INTB bits 3:0 is INTA
-+ * 0x5D bits 7:4 is INTD bits 3:0 is INTC
-+ */
-+static int pirq_cyrix_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
++void _arch_dup_mmap(struct mm_struct *mm)
+{
-+ return read_config_nybble(router, 0x5C, (pirq-1)^1);
++ if (!test_bit(PG_pinned, &virt_to_page(mm->pgd)->flags))
++ mm_pin(mm);
+}
+
-+static int pirq_cyrix_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++void _arch_exit_mmap(struct mm_struct *mm)
+{
-+ write_config_nybble(router, 0x5C, (pirq-1)^1, irq);
-+ return 1;
-+}
++ struct task_struct *tsk = current;
+
-+/*
-+ * PIRQ routing for SiS 85C503 router used in several SiS chipsets.
-+ * We have to deal with the following issues here:
-+ * - vendors have different ideas about the meaning of link values
-+ * - some onboard devices (integrated in the chipset) have special
-+ * links and are thus routed differently (i.e. not via PCI INTA-INTD)
-+ * - different revision of the router have a different layout for
-+ * the routing registers, particularly for the onchip devices
-+ *
-+ * For all routing registers the common thing is we have one byte
-+ * per routeable link which is defined as:
-+ * bit 7 IRQ mapping enabled (0) or disabled (1)
-+ * bits [6:4] reserved (sometimes used for onchip devices)
-+ * bits [3:0] IRQ to map to
-+ * allowed: 3-7, 9-12, 14-15
-+ * reserved: 0, 1, 2, 8, 13
-+ *
-+ * The config-space registers located at 0x41/0x42/0x43/0x44 are
-+ * always used to route the normal PCI INT A/B/C/D respectively.
-+ * Apparently there are systems implementing PCI routing table using
-+ * link values 0x01-0x04 and others using 0x41-0x44 for PCI INTA..D.
-+ * We try our best to handle both link mappings.
-+ *
-+ * Currently (2003-05-21) it appears most SiS chipsets follow the
-+ * definition of routing registers from the SiS-5595 southbridge.
-+ * According to the SiS 5595 datasheets the revision id's of the
-+ * router (ISA-bridge) should be 0x01 or 0xb0.
-+ *
-+ * Furthermore we've also seen lspci dumps with revision 0x00 and 0xb1.
-+ * Looks like these are used in a number of SiS 5xx/6xx/7xx chipsets.
-+ * They seem to work with the current routing code. However there is
-+ * some concern because of the two USB-OHCI HCs (original SiS 5595
-+ * had only one). YMMV.
-+ *
-+ * Onchip routing for router rev-id 0x01/0xb0 and probably 0x00/0xb1:
-+ *
-+ * 0x61: IDEIRQ:
-+ * bits [6:5] must be written 01
-+ * bit 4 channel-select primary (0), secondary (1)
-+ *
-+ * 0x62: USBIRQ:
-+ * bit 6 OHCI function disabled (0), enabled (1)
-+ *
-+ * 0x6a: ACPI/SCI IRQ: bits 4-6 reserved
-+ *
-+ * 0x7e: Data Acq. Module IRQ - bits 4-6 reserved
-+ *
-+ * We support USBIRQ (in addition to INTA-INTD) and keep the
-+ * IDE, ACPI and DAQ routing untouched as set by the BIOS.
++ 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 (test_bit(PG_pinned, &virt_to_page(mm->pgd)->flags) &&
++ (atomic_read(&mm->mm_count) == 1) &&
++ !mm->context.has_foreign_mappings)
++ mm_unpin(mm);
++}
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/oprofile/Makefile b/arch/i386/oprofile/Makefile
+--- a/arch/i386/oprofile/Makefile 2007-03-12 21:58:05.000000000 +0100
++++ b/arch/i386/oprofile/Makefile 2007-03-14 10:55:14.000000000 +0100
+@@ -6,7 +6,11 @@
+ oprofilefs.o oprofile_stats.o \
+ timer_int.o )
+
++ifdef CONFIG_XEN
++oprofile-y := $(DRIVER_OBJS) xenoprof.o
++else
+ oprofile-y := $(DRIVER_OBJS) init.o backtrace.o
+ oprofile-$(CONFIG_X86_LOCAL_APIC) += nmi_int.o op_model_athlon.o \
+ op_model_ppro.o op_model_p4.o
+ oprofile-$(CONFIG_X86_IO_APIC) += nmi_timer_int.o
++endif
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/oprofile/xenoprof.c b/arch/i386/oprofile/xenoprof.c
+--- a/arch/i386/oprofile/xenoprof.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/oprofile/xenoprof.c 2007-03-14 10:55:14.000000000 +0100
+@@ -0,0 +1,584 @@
++/**
++ * @file xenoprof.c
+ *
-+ * Currently the only reported exception is the new SiS 65x chipset
-+ * which includes the SiS 69x southbridge. Here we have the 85C503
-+ * router revision 0x04 and there are changes in the register layout
-+ * mostly related to the different USB HCs with USB 2.0 support.
++ * @remark Copyright 2002 OProfile authors
++ * @remark Read the file COPYING
+ *
-+ * Onchip routing for router rev-id 0x04 (try-and-error observation)
++ * @author John Levon <levon at movementarian.org>
+ *
-+ * 0x60/0x61/0x62/0x63: 1xEHCI and 3xOHCI (companion) USB-HCs
-+ * bit 6-4 are probably unused, not like 5595
++ * Modified by Aravind Menon and Jose Renato Santos for Xen
++ * These modifications are:
++ * Copyright (C) 2005 Hewlett-Packard Co.
+ */
+
-+#define PIRQ_SIS_IRQ_MASK 0x0f
-+#define PIRQ_SIS_IRQ_DISABLE 0x80
-+#define PIRQ_SIS_USB_ENABLE 0x40
++#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/nmi.h>
++#include <asm/msr.h>
++#include <asm/apic.h>
++#include <asm/pgtable.h>
++#include <xen/evtchn.h>
++#include "op_counter.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>
+
-+static int pirq_sis_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
-+{
-+ u8 x;
-+ int reg;
++#define MAX_XENOPROF_SAMPLES 16
+
-+ reg = pirq;
-+ if (reg >= 0x01 && reg <= 0x04)
-+ reg += 0x40;
-+ pci_read_config_byte(router, reg, &x);
-+ return (x & PIRQ_SIS_IRQ_DISABLE) ? 0 : (x & PIRQ_SIS_IRQ_MASK);
-+}
++static int xenoprof_start(void);
++static void xenoprof_stop(void);
++
++static int xenoprof_enabled = 0;
++static unsigned int num_events = 0;
++static int is_primary = 0;
++static int active_defined;
++
++/* sample buffers shared with Xen */
++xenoprof_buf_t * xenoprof_buf[MAX_VIRT_CPUS];
++/* Shared buffer area */
++char * shared_buffer = NULL;
++/* 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];
++
++/* Passive sample buffers shared with Xen */
++xenoprof_buf_t *p_xenoprof_buf[MAX_OPROF_DOMAINS][MAX_VIRT_CPUS];
++/* Passive shared buffer area */
++char *p_shared_buffer[MAX_OPROF_DOMAINS];
++
++#ifdef CONFIG_PM
+
-+static int pirq_sis_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++static int xenoprof_suspend(struct sys_device * dev, pm_message_t state)
+{
-+ u8 x;
-+ int reg;
++ if (xenoprof_enabled == 1)
++ xenoprof_stop();
++ return 0;
++}
+
-+ reg = pirq;
-+ if (reg >= 0x01 && reg <= 0x04)
-+ reg += 0x40;
-+ pci_read_config_byte(router, reg, &x);
-+ x &= ~(PIRQ_SIS_IRQ_MASK | PIRQ_SIS_IRQ_DISABLE);
-+ x |= irq ? irq: PIRQ_SIS_IRQ_DISABLE;
-+ pci_write_config_byte(router, reg, x);
-+ return 1;
++
++static int xenoprof_resume(struct sys_device * dev)
++{
++ if (xenoprof_enabled == 1)
++ xenoprof_start();
++ return 0;
+}
+
+
-+/*
-+ * VLSI: nibble offset 0x74 - educated guess due to routing table and
-+ * config space of VLSI 82C534 PCI-bridge/router (1004:0102)
-+ * Tested on HP OmniBook 800 covering PIRQ 1, 2, 4, 8 for onboard
-+ * devices, PIRQ 3 for non-pci(!) soundchip and (untested) PIRQ 6
-+ * for the busbridge to the docking station.
-+ */
++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,
++};
+
-+static int pirq_vlsi_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
++
++static int __init init_driverfs(void)
+{
-+ if (pirq > 8) {
-+ printk(KERN_INFO "VLSI router pirq escape (%d)\n", pirq);
-+ return 0;
-+ }
-+ return read_config_nybble(router, 0x74, pirq-1);
++ int error;
++ if (!(error = sysdev_class_register(&oprofile_sysclass)))
++ error = sysdev_register(&device_oprofile);
++ return error;
+}
+
-+static int pirq_vlsi_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++
++static void __exit exit_driverfs(void)
+{
-+ if (pirq > 8) {
-+ printk(KERN_INFO "VLSI router pirq escape (%d)\n", pirq);
-+ return 0;
-+ }
-+ write_config_nybble(router, 0x74, pirq-1, irq);
-+ return 1;
++ sysdev_unregister(&device_oprofile);
++ sysdev_class_unregister(&oprofile_sysclass);
+}
+
-+/*
-+ * ServerWorks: PCI interrupts mapped to system IRQ lines through Index
-+ * and Redirect I/O registers (0x0c00 and 0x0c01). The Index register
-+ * format is (PCIIRQ## | 0x10), e.g.: PCIIRQ10=0x1a. The Redirect
-+ * register is a straight binary coding of desired PIC IRQ (low nibble).
-+ *
-+ * The 'link' value in the PIRQ table is already in the correct format
-+ * for the Index register. There are some special index values:
-+ * 0x00 for ACPI (SCI), 0x01 for USB, 0x02 for IDE0, 0x04 for IDE1,
-+ * and 0x03 for SMBus.
-+ */
-+static int pirq_serverworks_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
++#else
++#define init_driverfs() do { } while (0)
++#define exit_driverfs() do { } while (0)
++#endif /* CONFIG_PM */
++
++unsigned long long oprofile_samples = 0;
++unsigned long long p_oprofile_samples = 0;
++
++unsigned int pdomains;
++struct xenoprof_passive passive_domains[MAX_OPROF_DOMAINS];
++
++static void xenoprof_add_pc(xenoprof_buf_t *buf, int is_passive)
+{
-+ outb_p(pirq, 0xc00);
-+ return inb(0xc01) & 0xf;
++ int head, tail, size;
++
++ head = buf->event_head;
++ tail = buf->event_tail;
++ size = buf->event_size;
++
++ 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++;
++ }
++
++ buf->event_tail = tail;
+}
+
-+static int pirq_serverworks_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++static void xenoprof_handle_passive(void)
+{
-+ outb_p(pirq, 0xc00);
-+ outb_p(irq, 0xc01);
-+ return 1;
++ 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);
+}
+
-+/* Support for AMD756 PCI IRQ Routing
-+ * Jhon H. Caicedo <jhcaiced at osso.org.co>
-+ * Jun/21/2001 0.2.0 Release, fixed to use "nybble" functions... (jhcaiced)
-+ * Jun/19/2001 Alpha Release 0.1.0 (jhcaiced)
-+ * The AMD756 pirq rules are nibble-based
-+ * offset 0x56 0-3 PIRQA 4-7 PIRQB
-+ * offset 0x57 0-3 PIRQC 4-7 PIRQD
-+ */
-+static int pirq_amd756_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
++static irqreturn_t
++xenoprof_ovf_interrupt(int irq, void * dev_id)
+{
-+ u8 irq;
-+ irq = 0;
-+ if (pirq <= 4)
-+ {
-+ irq = read_config_nybble(router, 0x56, pirq - 1);
++ struct xenoprof_buf * buf;
++ int cpu;
++ static unsigned long flag;
++
++ cpu = smp_processor_id();
++ buf = xenoprof_buf[cpu];
++
++ xenoprof_add_pc(buf, 0);
++
++ if (is_primary && !test_and_set_bit(0, &flag)) {
++ xenoprof_handle_passive();
++ smp_mb__before_clear_bit();
++ clear_bit(0, &flag);
+ }
-+ printk(KERN_INFO "AMD756: dev %04x:%04x, router pirq : %d get irq : %2d\n",
-+ dev->vendor, dev->device, pirq, irq);
-+ return irq;
++
++ return IRQ_HANDLED;
+}
+
-+static int pirq_amd756_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++
++static void unbind_virq(void)
+{
-+ printk(KERN_INFO "AMD756: dev %04x:%04x, router pirq : %d SET irq : %2d\n",
-+ dev->vendor, dev->device, pirq, irq);
-+ if (pirq <= 4)
-+ {
-+ write_config_nybble(router, 0x56, pirq - 1, irq);
++ int i;
++
++ for_each_possible_cpu(i) {
++ if (ovf_irq[i] >= 0) {
++ unbind_from_irqhandler(ovf_irq[i], NULL);
++ ovf_irq[i] = -1;
++ }
+ }
-+ return 1;
+}
+
-+#ifdef CONFIG_PCI_BIOS
+
-+static int pirq_bios_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++static int bind_virq(void)
+{
-+ struct pci_dev *bridge;
-+ int pin = pci_get_interrupt_pin(dev, &bridge);
-+ return pcibios_set_irq_routing(bridge, pin, irq);
++ int i, result;
++
++ for_each_possible_cpu(i) {
++ result = bind_virq_to_irqhandler(VIRQ_XENOPROF,
++ i,
++ xenoprof_ovf_interrupt,
++ SA_INTERRUPT,
++ "xenoprof",
++ NULL);
++
++ if (result < 0) {
++ unbind_virq();
++ return result;
++ }
++
++ ovf_irq[i] = result;
++ }
++
++ return 0;
+}
+
-+#endif
+
-+static __init int intel_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
++static int map_xenoprof_buffer(int max_samples)
+{
-+ static struct pci_device_id __initdata pirq_440gx[] = {
-+ { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82443GX_0) },
-+ { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82443GX_2) },
-+ { },
-+ };
++ struct xenoprof_get_buffer get_buffer;
++ struct xenoprof_buf *buf;
++ int npages, ret, i;
++ struct vm_struct *area;
+
-+ /* 440GX has a proprietary PIRQ router -- don't use it */
-+ if (pci_dev_present(pirq_440gx))
++ if ( shared_buffer )
+ return 0;
+
-+ switch(device)
-+ {
-+ case PCI_DEVICE_ID_INTEL_82371FB_0:
-+ case PCI_DEVICE_ID_INTEL_82371SB_0:
-+ case PCI_DEVICE_ID_INTEL_82371AB_0:
-+ case PCI_DEVICE_ID_INTEL_82371MX:
-+ case PCI_DEVICE_ID_INTEL_82443MX_0:
-+ case PCI_DEVICE_ID_INTEL_82801AA_0:
-+ case PCI_DEVICE_ID_INTEL_82801AB_0:
-+ case PCI_DEVICE_ID_INTEL_82801BA_0:
-+ case PCI_DEVICE_ID_INTEL_82801BA_10:
-+ case PCI_DEVICE_ID_INTEL_82801CA_0:
-+ case PCI_DEVICE_ID_INTEL_82801CA_12:
-+ case PCI_DEVICE_ID_INTEL_82801DB_0:
-+ case PCI_DEVICE_ID_INTEL_82801E_0:
-+ case PCI_DEVICE_ID_INTEL_82801EB_0:
-+ case PCI_DEVICE_ID_INTEL_ESB_1:
-+ case PCI_DEVICE_ID_INTEL_ICH6_0:
-+ case PCI_DEVICE_ID_INTEL_ICH6_1:
-+ case PCI_DEVICE_ID_INTEL_ICH7_0:
-+ case PCI_DEVICE_ID_INTEL_ICH7_1:
-+ case PCI_DEVICE_ID_INTEL_ICH7_30:
-+ case PCI_DEVICE_ID_INTEL_ICH7_31:
-+ case PCI_DEVICE_ID_INTEL_ESB2_0:
-+ case PCI_DEVICE_ID_INTEL_ICH8_0:
-+ case PCI_DEVICE_ID_INTEL_ICH8_1:
-+ case PCI_DEVICE_ID_INTEL_ICH8_2:
-+ case PCI_DEVICE_ID_INTEL_ICH8_3:
-+ case PCI_DEVICE_ID_INTEL_ICH8_4:
-+ r->name = "PIIX/ICH";
-+ r->get = pirq_piix_get;
-+ r->set = pirq_piix_set;
-+ return 1;
++ get_buffer.max_samples = max_samples;
++
++ if ( (ret = HYPERVISOR_xenoprof_op(XENOPROF_get_buffer, &get_buffer)) )
++ return ret;
++
++ nbuf = get_buffer.nbuf;
++ npages = (get_buffer.bufsize * nbuf - 1) / PAGE_SIZE + 1;
++
++ area = alloc_vm_area(npages * PAGE_SIZE);
++ if (area == NULL)
++ return -ENOMEM;
++
++ if ( (ret = direct_kernel_remap_pfn_range(
++ (unsigned long)area->addr,
++ get_buffer.buf_maddr >> PAGE_SHIFT,
++ npages * PAGE_SIZE, __pgprot(_KERNPG_TABLE), DOMID_SELF)) ) {
++ vunmap(area->addr);
++ return ret;
+ }
++
++ shared_buffer = area->addr;
++ for (i=0; i< nbuf; i++) {
++ buf = (struct xenoprof_buf*)
++ &shared_buffer[i * get_buffer.bufsize];
++ BUG_ON(buf->vcpu_id >= MAX_VIRT_CPUS);
++ xenoprof_buf[buf->vcpu_id] = buf;
++ }
++
+ return 0;
+}
+
-+static __init int via_router_probe(struct irq_router *r,
-+ struct pci_dev *router, u16 device)
++
++static int xenoprof_setup(void)
+{
-+ /* FIXME: We should move some of the quirk fixup stuff here */
++ int ret;
++ int i;
+
-+ /*
-+ * work arounds for some buggy BIOSes
-+ */
-+ if (device == PCI_DEVICE_ID_VIA_82C586_0) {
-+ switch(router->device) {
-+ case PCI_DEVICE_ID_VIA_82C686:
-+ /*
-+ * Asus k7m bios wrongly reports 82C686A
-+ * as 586-compatible
-+ */
-+ device = PCI_DEVICE_ID_VIA_82C686;
-+ break;
-+ case PCI_DEVICE_ID_VIA_8235:
-+ /**
-+ * Asus a7v-x bios wrongly reports 8235
-+ * as 586-compatible
-+ */
-+ device = PCI_DEVICE_ID_VIA_8235;
-+ break;
++ if ( (ret = map_xenoprof_buffer(MAX_XENOPROF_SAMPLES)) )
++ return ret;
++
++ if ( (ret = bind_virq()) )
++ return ret;
++
++ if (is_primary) {
++ struct xenoprof_counter counter;
++
++ /* 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;
++ }
++
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_reserve_counters, NULL);
++ if (ret)
++ goto err;
++ for (i=0; i<num_events; i++) {
++ counter.ind = i;
++ counter.count = (uint64_t)counter_config[i].count;
++ counter.enabled = (uint32_t)counter_config[i].enabled;
++ counter.event = (uint32_t)counter_config[i].event;
++ 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);
+ }
-+ }
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_setup_events, NULL);
+
-+ switch(device) {
-+ case PCI_DEVICE_ID_VIA_82C586_0:
-+ r->name = "VIA";
-+ r->get = pirq_via586_get;
-+ r->set = pirq_via586_set;
-+ return 1;
-+ case PCI_DEVICE_ID_VIA_82C596:
-+ case PCI_DEVICE_ID_VIA_82C686:
-+ case PCI_DEVICE_ID_VIA_8231:
-+ case PCI_DEVICE_ID_VIA_8233A:
-+ case PCI_DEVICE_ID_VIA_8235:
-+ case PCI_DEVICE_ID_VIA_8237:
-+ /* FIXME: add new ones for 8233/5 */
-+ r->name = "VIA";
-+ r->get = pirq_via_get;
-+ r->set = pirq_via_set;
-+ return 1;
++ if (ret)
++ goto err;
+ }
-+ return 0;
-+}
+
-+static __init int vlsi_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
-+{
-+ switch(device)
-+ {
-+ case PCI_DEVICE_ID_VLSI_82C534:
-+ r->name = "VLSI 82C534";
-+ r->get = pirq_vlsi_get;
-+ r->set = pirq_vlsi_set;
-+ return 1;
-+ }
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_enable_virq, NULL);
++ if (ret)
++ goto err;
++
++ xenoprof_enabled = 1;
+ return 0;
++ err:
++ unbind_virq();
++ return ret;
+}
+
+
-+static __init int serverworks_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
++static void xenoprof_shutdown(void)
+{
-+ switch(device)
-+ {
-+ case PCI_DEVICE_ID_SERVERWORKS_OSB4:
-+ case PCI_DEVICE_ID_SERVERWORKS_CSB5:
-+ r->name = "ServerWorks";
-+ r->get = pirq_serverworks_get;
-+ r->set = pirq_serverworks_set;
-+ return 1;
-+ }
-+ return 0;
-+}
++ xenoprof_enabled = 0;
+
-+static __init int sis_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
-+{
-+ if (device != PCI_DEVICE_ID_SI_503)
-+ return 0;
-+
-+ r->name = "SIS";
-+ r->get = pirq_sis_get;
-+ r->set = pirq_sis_set;
-+ return 1;
-+}
++ HYPERVISOR_xenoprof_op(XENOPROF_disable_virq, NULL);
+
-+static __init int cyrix_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
-+{
-+ switch(device)
-+ {
-+ case PCI_DEVICE_ID_CYRIX_5520:
-+ r->name = "NatSemi";
-+ r->get = pirq_cyrix_get;
-+ r->set = pirq_cyrix_set;
-+ return 1;
++ if (is_primary) {
++ HYPERVISOR_xenoprof_op(XENOPROF_release_counters, NULL);
++ active_defined = 0;
+ }
-+ return 0;
-+}
+
-+static __init int opti_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
-+{
-+ switch(device)
-+ {
-+ case PCI_DEVICE_ID_OPTI_82C700:
-+ r->name = "OPTI";
-+ r->get = pirq_opti_get;
-+ r->set = pirq_opti_set;
-+ return 1;
-+ }
-+ return 0;
-+}
++ unbind_virq();
+
-+static __init int ite_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
-+{
-+ switch(device)
-+ {
-+ case PCI_DEVICE_ID_ITE_IT8330G_0:
-+ r->name = "ITE";
-+ r->get = pirq_ite_get;
-+ r->set = pirq_ite_set;
-+ return 1;
-+ }
-+ return 0;
+}
+
-+static __init int ali_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
++
++static int xenoprof_start(void)
+{
-+ switch(device)
-+ {
-+ case PCI_DEVICE_ID_AL_M1533:
-+ case PCI_DEVICE_ID_AL_M1563:
-+ printk(KERN_DEBUG "PCI: Using ALI IRQ Router\n");
-+ r->name = "ALI";
-+ r->get = pirq_ali_get;
-+ r->set = pirq_ali_set;
-+ return 1;
-+ }
-+ return 0;
++ int ret = 0;
++
++ if (is_primary)
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_start, NULL);
++
++ return ret;
+}
+
-+static __init int amd_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
++
++static void xenoprof_stop(void)
+{
-+ switch(device)
-+ {
-+ case PCI_DEVICE_ID_AMD_VIPER_740B:
-+ r->name = "AMD756";
-+ break;
-+ case PCI_DEVICE_ID_AMD_VIPER_7413:
-+ r->name = "AMD766";
-+ break;
-+ case PCI_DEVICE_ID_AMD_VIPER_7443:
-+ r->name = "AMD768";
-+ break;
-+ default:
-+ return 0;
-+ }
-+ r->get = pirq_amd756_get;
-+ r->set = pirq_amd756_set;
-+ return 1;
++ if (is_primary)
++ HYPERVISOR_xenoprof_op(XENOPROF_stop, NULL);
+}
-+
-+static __initdata struct irq_router_handler pirq_routers[] = {
-+ { PCI_VENDOR_ID_INTEL, intel_router_probe },
-+ { PCI_VENDOR_ID_AL, ali_router_probe },
-+ { PCI_VENDOR_ID_ITE, ite_router_probe },
-+ { PCI_VENDOR_ID_VIA, via_router_probe },
-+ { PCI_VENDOR_ID_OPTI, opti_router_probe },
-+ { PCI_VENDOR_ID_SI, sis_router_probe },
-+ { PCI_VENDOR_ID_CYRIX, cyrix_router_probe },
-+ { PCI_VENDOR_ID_VLSI, vlsi_router_probe },
-+ { PCI_VENDOR_ID_SERVERWORKS, serverworks_router_probe },
-+ { PCI_VENDOR_ID_AMD, amd_router_probe },
-+ /* Someone with docs needs to add the ATI Radeon IGP */
-+ { 0, NULL }
-+};
-+static struct irq_router pirq_router;
-+static struct pci_dev *pirq_router_dev;
+
+
-+/*
-+ * FIXME: should we have an option to say "generic for
-+ * chipset" ?
-+ */
-+
-+static void __init pirq_find_router(struct irq_router *r)
++static int xenoprof_set_active(int * active_domains,
++ unsigned int adomains)
+{
-+ struct irq_routing_table *rt = pirq_table;
-+ struct irq_router_handler *h;
++ int ret = 0;
++ int i;
++ int set_dom0 = 0;
++ domid_t domid;
+
-+#ifdef CONFIG_PCI_BIOS
-+ if (!rt->signature) {
-+ printk(KERN_INFO "PCI: Using BIOS for IRQ routing\n");
-+ r->set = pirq_bios_set;
-+ r->name = "BIOS";
-+ return;
-+ }
-+#endif
++ if (!is_primary)
++ return 0;
+
-+ /* Default unless a driver reloads it */
-+ r->name = "default";
-+ r->get = NULL;
-+ r->set = NULL;
-+
-+ DBG(KERN_DEBUG "PCI: Attempting to find IRQ router for %04x:%04x\n",
-+ rt->rtr_vendor, rt->rtr_device);
++ if (adomains > MAX_OPROF_DOMAINS)
++ return -E2BIG;
+
-+ pirq_router_dev = pci_find_slot(rt->rtr_bus, rt->rtr_devfn);
-+ if (!pirq_router_dev) {
-+ DBG(KERN_DEBUG "PCI: Interrupt router not found at "
-+ "%02x:%02x\n", rt->rtr_bus, rt->rtr_devfn);
-+ return;
-+ }
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_reset_active_list, NULL);
++ if (ret)
++ return ret;
+
-+ for( h = pirq_routers; h->vendor; h++) {
-+ /* First look for a router match */
-+ if (rt->rtr_vendor == h->vendor && h->probe(r, pirq_router_dev, rt->rtr_device))
-+ break;
-+ /* Fall back to a device match */
-+ if (pirq_router_dev->vendor == h->vendor && h->probe(r, pirq_router_dev, pirq_router_dev->device))
-+ break;
++ 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);
+ }
-+ printk(KERN_INFO "PCI: Using IRQ router %s [%04x/%04x] at %s\n",
-+ pirq_router.name,
-+ pirq_router_dev->vendor,
-+ pirq_router_dev->device,
-+ pci_name(pirq_router_dev));
-+}
-+
-+static struct irq_info *pirq_get_info(struct pci_dev *dev)
-+{
-+ struct irq_routing_table *rt = pirq_table;
-+ int entries = (rt->size - sizeof(struct irq_routing_table)) / sizeof(struct irq_info);
-+ struct irq_info *info;
+
-+ for (info = rt->slots; entries--; info++)
-+ if (info->bus == dev->bus->number && PCI_SLOT(info->devfn) == PCI_SLOT(dev->devfn))
-+ return info;
-+ return NULL;
++out:
++ if (ret)
++ HYPERVISOR_xenoprof_op(XENOPROF_reset_active_list, NULL);
++ active_defined = !ret;
++ return ret;
+}
+
-+static int pcibios_lookup_irq(struct pci_dev *dev, int assign)
++static int xenoprof_set_passive(int * p_domains,
++ unsigned int pdoms)
+{
-+ u8 pin;
-+ struct irq_info *info;
-+ int i, pirq, newirq;
-+ int irq = 0;
-+ u32 mask;
-+ struct irq_router *r = &pirq_router;
-+ struct pci_dev *dev2 = NULL;
-+ char *msg = NULL;
-+
-+ /* Find IRQ pin */
-+ pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &pin);
-+ if (!pin) {
-+ DBG(KERN_DEBUG " -> no interrupt pin\n");
-+ return 0;
-+ }
-+ pin = pin - 1;
++ int ret;
++ int i, j;
++ int npages;
++ struct xenoprof_buf *buf;
++ struct vm_struct *area;
++ pgprot_t prot = __pgprot(_KERNPG_TABLE);
+
-+ /* Find IRQ routing entry */
++ if (!is_primary)
++ return 0;
+
-+ if (!pirq_table)
-+ return 0;
-+
-+ DBG(KERN_DEBUG "IRQ for %s[%c]", pci_name(dev), 'A' + pin);
-+ info = pirq_get_info(dev);
-+ if (!info) {
-+ DBG(" -> not found in routing table\n" KERN_DEBUG);
-+ return 0;
-+ }
-+ pirq = info->irq[pin].link;
-+ mask = info->irq[pin].bitmap;
-+ if (!pirq) {
-+ DBG(" -> not routed\n" KERN_DEBUG);
-+ return 0;
-+ }
-+ DBG(" -> PIRQ %02x, mask %04x, excl %04x", pirq, mask, pirq_table->exclusive_irqs);
-+ mask &= pcibios_irq_mask;
++ if (pdoms > MAX_OPROF_DOMAINS)
++ return -E2BIG;
+
-+ /* Work around broken HP Pavilion Notebooks which assign USB to
-+ IRQ 9 even though it is actually wired to IRQ 11 */
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_reset_passive_list, NULL);
++ if (ret)
++ return ret;
+
-+ if (broken_hp_bios_irq9 && pirq == 0x59 && dev->irq == 9) {
-+ dev->irq = 11;
-+ pci_write_config_byte(dev, PCI_INTERRUPT_LINE, 11);
-+ r->set(pirq_router_dev, dev, pirq, 11);
-+ }
++ for (i = 0; i < pdoms; i++) {
++ passive_domains[i].domain_id = p_domains[i];
++ passive_domains[i].max_samples = 2048;
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_set_passive,
++ &passive_domains[i]);
++ if (ret)
++ goto out;
+
-+ /* same for Acer Travelmate 360, but with CB and irq 11 -> 10 */
-+ if (acer_tm360_irqrouting && dev->irq == 11 && dev->vendor == PCI_VENDOR_ID_O2) {
-+ pirq = 0x68;
-+ mask = 0x400;
-+ dev->irq = r->get(pirq_router_dev, dev, pirq);
-+ pci_write_config_byte(dev, PCI_INTERRUPT_LINE, dev->irq);
-+ }
++ npages = (passive_domains[i].bufsize * passive_domains[i].nbuf - 1) / PAGE_SIZE + 1;
+
-+ /*
-+ * Find the best IRQ to assign: use the one
-+ * reported by the device if possible.
-+ */
-+ newirq = dev->irq;
-+ if (newirq && !((1 << newirq) & mask)) {
-+ if ( pci_probe & PCI_USE_PIRQ_MASK) newirq = 0;
-+ else printk("\n" KERN_WARNING
-+ "PCI: IRQ %i for device %s doesn't match PIRQ mask "
-+ "- try pci=usepirqmask\n" KERN_DEBUG, newirq,
-+ pci_name(dev));
-+ }
-+ if (!newirq && assign) {
-+ for (i = 0; i < 16; i++) {
-+ if (!(mask & (1 << i)))
-+ continue;
-+ if (pirq_penalty[i] < pirq_penalty[newirq] && can_request_irq(i, IRQF_SHARED))
-+ newirq = i;
++ area = alloc_vm_area(npages * PAGE_SIZE);
++ if (area == NULL) {
++ ret = -ENOMEM;
++ goto out;
+ }
-+ }
-+ DBG(" -> newirq=%d", newirq);
+
-+ /* Check if it is hardcoded */
-+ if ((pirq & 0xf0) == 0xf0) {
-+ irq = pirq & 0xf;
-+ DBG(" -> hardcoded IRQ %d\n", irq);
-+ msg = "Hardcoded";
-+ } else if ( r->get && (irq = r->get(pirq_router_dev, dev, pirq)) && \
-+ ((!(pci_probe & PCI_USE_PIRQ_MASK)) || ((1 << irq) & mask)) ) {
-+ DBG(" -> got IRQ %d\n", irq);
-+ msg = "Found";
-+ eisa_set_level_irq(irq);
-+ } else if (newirq && r->set && (dev->class >> 8) != PCI_CLASS_DISPLAY_VGA) {
-+ DBG(" -> assigning IRQ %d", newirq);
-+ if (r->set(pirq_router_dev, dev, pirq, newirq)) {
-+ eisa_set_level_irq(newirq);
-+ DBG(" ... OK\n");
-+ msg = "Assigned";
-+ irq = newirq;
++ ret = direct_kernel_remap_pfn_range(
++ (unsigned long)area->addr,
++ passive_domains[i].buf_maddr >> PAGE_SHIFT,
++ npages * PAGE_SIZE, prot, DOMID_SELF);
++ if (ret) {
++ vunmap(area->addr);
++ goto out;
+ }
-+ }
+
-+ if (!irq) {
-+ DBG(" ... failed\n");
-+ if (newirq && mask == (1 << newirq)) {
-+ msg = "Guessed";
-+ irq = newirq;
-+ } else
-+ return 0;
-+ }
-+ printk(KERN_INFO "PCI: %s IRQ %d for device %s\n", msg, irq, pci_name(dev));
++ p_shared_buffer[i] = area->addr;
+
-+ /* Update IRQ for all devices with the same pirq value */
-+ while ((dev2 = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev2)) != NULL) {
-+ pci_read_config_byte(dev2, PCI_INTERRUPT_PIN, &pin);
-+ if (!pin)
-+ continue;
-+ pin--;
-+ info = pirq_get_info(dev2);
-+ if (!info)
-+ continue;
-+ if (info->irq[pin].link == pirq) {
-+ /* We refuse to override the dev->irq information. Give a warning! */
-+ if ( dev2->irq && dev2->irq != irq && \
-+ (!(pci_probe & PCI_USE_PIRQ_MASK) || \
-+ ((1 << dev2->irq) & mask)) ) {
-+#ifndef CONFIG_PCI_MSI
-+ printk(KERN_INFO "IRQ routing conflict for %s, have irq %d, want irq %d\n",
-+ pci_name(dev2), dev2->irq, irq);
-+#endif
-+ continue;
-+ }
-+ dev2->irq = irq;
-+ pirq_penalty[irq]++;
-+ if (dev != dev2)
-+ printk(KERN_INFO "PCI: Sharing IRQ %d with %s\n", irq, pci_name(dev2));
++ for (j = 0; j < passive_domains[i].nbuf; j++) {
++ buf = (struct xenoprof_buf *)
++ &p_shared_buffer[i][j * passive_domains[i].bufsize];
++ BUG_ON(buf->vcpu_id >= MAX_VIRT_CPUS);
++ p_xenoprof_buf[i][buf->vcpu_id] = buf;
+ }
-+ }
-+ return 1;
-+}
-+
-+static void __init pcibios_fixup_irqs(void)
-+{
-+ struct pci_dev *dev = NULL;
-+ u8 pin;
-+
-+ DBG(KERN_DEBUG "PCI: IRQ fixup\n");
-+ while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
-+ /*
-+ * If the BIOS has set an out of range IRQ number, just ignore it.
-+ * Also keep track of which IRQ's are already in use.
-+ */
-+ if (dev->irq >= 16) {
-+ DBG(KERN_DEBUG "%s: ignoring bogus IRQ %d\n", pci_name(dev), dev->irq);
-+ dev->irq = 0;
-+ }
-+ /* If the IRQ is already assigned to a PCI device, ignore its ISA use penalty */
-+ if (pirq_penalty[dev->irq] >= 100 && pirq_penalty[dev->irq] < 100000)
-+ pirq_penalty[dev->irq] = 0;
-+ pirq_penalty[dev->irq]++;
-+ }
-+
-+ dev = NULL;
-+ while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
-+ pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &pin);
-+#ifdef CONFIG_X86_IO_APIC
-+ /*
-+ * Recalculate IRQ numbers if we use the I/O APIC.
-+ */
-+ if (io_apic_assign_pci_irqs)
-+ {
-+ int irq;
+
-+ if (pin) {
-+ pin--; /* interrupt pins are numbered starting from 1 */
-+ irq = IO_APIC_get_PCI_irq_vector(dev->bus->number, PCI_SLOT(dev->devfn), pin);
-+ /*
-+ * Busses behind bridges are typically not listed in the MP-table.
-+ * In this case we have to look up the IRQ based on the parent bus,
-+ * parent slot, and pin number. The SMP code detects such bridged
-+ * busses itself so we should get into this branch reliably.
-+ */
-+ if (irq < 0 && dev->bus->parent) { /* go back to the bridge */
-+ struct pci_dev * bridge = dev->bus->self;
-+
-+ pin = (pin + PCI_SLOT(dev->devfn)) % 4;
-+ irq = IO_APIC_get_PCI_irq_vector(bridge->bus->number,
-+ PCI_SLOT(bridge->devfn), pin);
-+ if (irq >= 0)
-+ printk(KERN_WARNING "PCI: using PPB %s[%c] to get irq %d\n",
-+ pci_name(bridge), 'A' + pin, irq);
-+ }
-+ if (irq >= 0) {
-+ if (use_pci_vector() &&
-+ !platform_legacy_irq(irq))
-+ irq = IO_APIC_VECTOR(irq);
-+
-+ printk(KERN_INFO "PCI->APIC IRQ transform: %s[%c] -> IRQ %d\n",
-+ pci_name(dev), 'A' + pin, irq);
-+ dev->irq = irq;
-+ }
-+ }
-+ }
-+#endif
-+ /*
-+ * Still no IRQ? Try to lookup one...
-+ */
-+ if (pin && !dev->irq)
-+ pcibios_lookup_irq(dev, 0);
+ }
-+}
+
-+/*
-+ * Work around broken HP Pavilion Notebooks which assign USB to
-+ * IRQ 9 even though it is actually wired to IRQ 11
-+ */
-+static int __init fix_broken_hp_bios_irq9(struct dmi_system_id *d)
-+{
-+ if (!broken_hp_bios_irq9) {
-+ broken_hp_bios_irq9 = 1;
-+ printk(KERN_INFO "%s detected - fixing broken IRQ routing\n", d->ident);
-+ }
++ pdomains = pdoms;
+ return 0;
-+}
+
-+/*
-+ * Work around broken Acer TravelMate 360 Notebooks which assign
-+ * Cardbus to IRQ 11 even though it is actually wired to IRQ 10
-+ */
-+static int __init fix_acer_tm360_irqrouting(struct dmi_system_id *d)
-+{
-+ if (!acer_tm360_irqrouting) {
-+ acer_tm360_irqrouting = 1;
-+ printk(KERN_INFO "%s detected - fixing broken IRQ routing\n", d->ident);
++out:
++ for (j = 0; j < i; j++) {
++ vunmap(p_shared_buffer[j]);
++ p_shared_buffer[j] = NULL;
+ }
-+ return 0;
++
++ return ret;
+}
+
-+static struct dmi_system_id __initdata pciirq_dmi_table[] = {
-+ {
-+ .callback = fix_broken_hp_bios_irq9,
-+ .ident = "HP Pavilion N5400 Series Laptop",
-+ .matches = {
-+ DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"),
-+ DMI_MATCH(DMI_BIOS_VERSION, "GE.M1.03"),
-+ DMI_MATCH(DMI_PRODUCT_VERSION, "HP Pavilion Notebook Model GE"),
-+ DMI_MATCH(DMI_BOARD_VERSION, "OmniBook N32N-736"),
-+ },
-+ },
-+ {
-+ .callback = fix_acer_tm360_irqrouting,
-+ .ident = "Acer TravelMate 36x Laptop",
-+ .matches = {
-+ DMI_MATCH(DMI_SYS_VENDOR, "Acer"),
-+ DMI_MATCH(DMI_PRODUCT_NAME, "TravelMate 360"),
-+ },
-+ },
-+ { }
-+};
++struct op_counter_config counter_config[OP_MAX_COUNTER];
+
-+static int __init pcibios_irq_init(void)
++static int xenoprof_create_files(struct super_block * sb, struct dentry * root)
+{
-+ DBG(KERN_DEBUG "PCI: IRQ init\n");
-+
-+ if (pcibios_enable_irq || raw_pci_ops == NULL)
-+ return 0;
-+
-+ dmi_check_system(pciirq_dmi_table);
-+
-+ pirq_table = pirq_find_routing_table();
++ unsigned int i;
+
-+#ifdef CONFIG_PCI_BIOS
-+ if (!pirq_table && (pci_probe & PCI_BIOS_IRQ_SCAN))
-+ pirq_table = pcibios_get_irq_routing_table();
-+#endif
-+ if (pirq_table) {
-+ pirq_peer_trick();
-+ pirq_find_router(&pirq_router);
-+ if (pirq_table->exclusive_irqs) {
-+ int i;
-+ for (i=0; i<16; i++)
-+ if (!(pirq_table->exclusive_irqs & (1 << i)))
-+ pirq_penalty[i] += 100;
-+ }
-+ /* If we're using the I/O APIC, avoid using the PCI IRQ routing table */
-+ if (io_apic_assign_pci_irqs)
-+ pirq_table = NULL;
++ for (i = 0; i < num_events; ++i) {
++ struct dentry * dir;
++ char buf[2];
++
++ snprintf(buf, 2, "%d", i);
++ dir = oprofilefs_mkdir(sb, root, buf);
++ oprofilefs_create_ulong(sb, dir, "enabled",
++ &counter_config[i].enabled);
++ oprofilefs_create_ulong(sb, dir, "event",
++ &counter_config[i].event);
++ oprofilefs_create_ulong(sb, dir, "count",
++ &counter_config[i].count);
++ oprofilefs_create_ulong(sb, dir, "unit_mask",
++ &counter_config[i].unit_mask);
++ oprofilefs_create_ulong(sb, dir, "kernel",
++ &counter_config[i].kernel);
++ oprofilefs_create_ulong(sb, dir, "user",
++ &counter_config[i].user);
+ }
+
-+ pcibios_enable_irq = pirq_enable_irq;
-+
-+ pcibios_fixup_irqs();
+ return 0;
+}
+
-+subsys_initcall(pcibios_irq_init);
+
++struct oprofile_operations xenoprof_ops = {
++ .create_files = xenoprof_create_files,
++ .set_active = xenoprof_set_active,
++ .set_passive = xenoprof_set_passive,
++ .setup = xenoprof_setup,
++ .shutdown = xenoprof_shutdown,
++ .start = xenoprof_start,
++ .stop = xenoprof_stop
++};
+
-+static void pirq_penalize_isa_irq(int irq, int active)
-+{
-+ /*
-+ * If any ISAPnP device reports an IRQ in its list of possible
-+ * IRQ's, we try to avoid assigning it to PCI devices.
-+ */
-+ if (irq < 16) {
-+ if (active)
-+ pirq_penalty[irq] += 1000;
-+ else
-+ pirq_penalty[irq] += 100;
-+ }
-+}
+
-+void pcibios_penalize_isa_irq(int irq, int active)
-+{
-+#ifdef CONFIG_ACPI
-+ if (!acpi_noirq)
-+ acpi_penalize_isa_irq(irq, active);
-+ else
-+#endif
-+ pirq_penalize_isa_irq(irq, active);
-+}
++/* in order to get driverfs right */
++static int using_xenoprof;
+
-+static int pirq_enable_irq(struct pci_dev *dev)
++int __init oprofile_arch_init(struct oprofile_operations * ops)
+{
-+ u8 pin;
-+ struct pci_dev *temp_dev;
++ struct xenoprof_init init;
++ int ret, i;
+
-+ pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &pin);
-+ if (pin && !pcibios_lookup_irq(dev, 1) && !dev->irq) {
-+ char *msg = "";
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_init, &init);
+
-+ pin--; /* interrupt pins are numbered starting from 1 */
++ if (!ret) {
++ num_events = init.num_events;
++ is_primary = init.is_primary;
+
-+ if (io_apic_assign_pci_irqs) {
-+ int irq;
++ /* just in case - make sure we do not overflow event list
++ (i.e. counter_config list) */
++ if (num_events > OP_MAX_COUNTER)
++ num_events = OP_MAX_COUNTER;
+
-+ irq = IO_APIC_get_PCI_irq_vector(dev->bus->number, PCI_SLOT(dev->devfn), pin);
-+ /*
-+ * Busses behind bridges are typically not listed in the MP-table.
-+ * In this case we have to look up the IRQ based on the parent bus,
-+ * parent slot, and pin number. The SMP code detects such bridged
-+ * busses itself so we should get into this branch reliably.
-+ */
-+ temp_dev = dev;
-+ while (irq < 0 && dev->bus->parent) { /* go back to the bridge */
-+ struct pci_dev * bridge = dev->bus->self;
-+
-+ pin = (pin + PCI_SLOT(dev->devfn)) % 4;
-+ irq = IO_APIC_get_PCI_irq_vector(bridge->bus->number,
-+ PCI_SLOT(bridge->devfn), pin);
-+ if (irq >= 0)
-+ printk(KERN_WARNING "PCI: using PPB %s[%c] to get irq %d\n",
-+ pci_name(bridge), 'A' + pin, irq);
-+ dev = bridge;
-+ }
-+ dev = temp_dev;
-+ if (irq >= 0) {
-+#ifdef CONFIG_PCI_MSI
-+ if (!platform_legacy_irq(irq))
-+ irq = IO_APIC_VECTOR(irq);
-+#endif
-+ printk(KERN_INFO "PCI->APIC IRQ transform: %s[%c] -> IRQ %d\n",
-+ pci_name(dev), 'A' + pin, irq);
-+ dev->irq = irq;
-+ return 0;
-+ } else
-+ msg = " Probably buggy MP table.";
-+ } else if (pci_probe & PCI_BIOS_IRQ_SCAN)
-+ msg = "";
-+ else
-+ msg = " Please try using pci=biosirq.";
++ /* 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;
+
-+ /* With IDE legacy devices the IRQ lookup failure is not a problem.. */
-+ if (dev->class >> 8 == PCI_CLASS_STORAGE_IDE && !(dev->class & 0x5))
-+ return 0;
++ init_driverfs();
++ using_xenoprof = 1;
++ *ops = xenoprof_ops;
++
++ for (i=0; i<NR_CPUS; i++)
++ ovf_irq[i] = -1;
+
-+ printk(KERN_WARNING "PCI: No IRQ known for interrupt pin %c of device %s.%s\n",
-+ 'A' + pin, pci_name(dev), msg);
++ active_defined = 0;
+ }
-+ return 0;
++ printk(KERN_INFO "oprofile_arch_init: ret %d, events %d, "
++ "is_primary %d\n", ret, num_events, is_primary);
++ return ret;
+}
+
-+int pci_vector_resources(int last, int nr_released)
++
++void __exit oprofile_arch_exit(void)
+{
-+ int count = nr_released;
++ int i;
+
-+ int next = last;
-+ int offset = (last % 8);
++ if (using_xenoprof)
++ exit_driverfs();
+
-+ while (next < FIRST_SYSTEM_VECTOR) {
-+ next += 8;
-+#ifdef CONFIG_X86_64
-+ if (next == IA32_SYSCALL_VECTOR)
-+ continue;
-+#else
-+ if (next == SYSCALL_VECTOR)
-+ continue;
-+#endif
-+ count++;
-+ if (next >= FIRST_SYSTEM_VECTOR) {
-+ if (offset%8) {
-+ next = FIRST_DEVICE_VECTOR + offset;
-+ offset++;
-+ continue;
-+ }
-+ count--;
-+ }
++ if (shared_buffer) {
++ vunmap(shared_buffer);
++ shared_buffer = NULL;
+ }
++ if (is_primary) {
++ for (i = 0; i < pdomains; i++)
++ if (p_shared_buffer[i]) {
++ vunmap(p_shared_buffer[i]);
++ p_shared_buffer[i] = NULL;
++ }
++ HYPERVISOR_xenoprof_op(XENOPROF_shutdown, NULL);
++ }
+
-+ return count;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/pci/Makefile linux-2.6.18-xen/arch/i386/pci/Makefile
---- linux-2.6.18.3/arch/i386/pci/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/pci/Makefile 2006-11-19 14:26:22.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/pci/irq.c b/arch/i386/pci/irq.c
+--- a/arch/i386/pci/irq.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/pci/irq.c 2007-03-14 10:55:14.000000000 +0100
+@@ -94,13 +94,25 @@
+ u8 *addr;
+ struct irq_routing_table *rt;
+
++#ifdef CONFIG_XEN
++ if (!is_initial_xendomain())
++ return NULL;
++#endif
+ if (pirq_table_addr) {
++#ifdef CONFIG_XEN
++ rt = pirq_check_routing_table((u8 *) isa_bus_to_virt(pirq_table_addr));
++#else
+ rt = pirq_check_routing_table((u8 *) __va(pirq_table_addr));
++#endif
+ if (rt)
+ return rt;
+ printk(KERN_WARNING "PCI: PIRQ table NOT found at pirqaddr\n");
+ }
++#ifdef CONFIG_XEN
++ for(addr = (u8 *) isa_bus_to_virt(0xf0000); addr < (u8 *) isa_bus_to_virt(0x100000); addr += 16) {
++#else
+ for(addr = (u8 *) __va(0xf0000); addr < (u8 *) __va(0x100000); addr += 16) {
++#endif
+ rt = pirq_check_routing_table(addr);
+ if (rt)
+ return rt;
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/pci/Makefile b/arch/i386/pci/Makefile
+--- a/arch/i386/pci/Makefile 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/pci/Makefile 2007-03-14 10:55:14.000000000 +0100
@@ -4,6 +4,10 @@
obj-$(CONFIG_PCI_MMCONFIG) += mmconfig.o direct.o
obj-$(CONFIG_PCI_DIRECT) += direct.o
@@ -21703,15 +17728,15 @@
@@ -12,3 +16,8 @@
pci-$(CONFIG_X86_NUMAQ) := numa.o irq.o
- obj-y += $(pci-y) common.o
+ obj-y += $(pci-y) common.o early.o
+
+ifdef CONFIG_XEN
+include $(srctree)/scripts/Makefile.xen
+obj-y := $(call cherrypickxen, $(obj-y))
+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/pci/pcifront.c linux-2.6.18-xen/arch/i386/pci/pcifront.c
---- linux-2.6.18.3/arch/i386/pci/pcifront.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/i386/pci/pcifront.c 2006-11-19 14:26:22.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/pci/pcifront.c b/arch/i386/pci/pcifront.c
+--- a/arch/i386/pci/pcifront.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/i386/pci/pcifront.c 2007-03-14 10:55:14.000000000 +0100
@@ -0,0 +1,55 @@
+/*
+ * PCI Frontend Stub - puts some "dummy" functions in to the Linux x86 PCI core
@@ -21768,18 +17793,18 @@
+}
+
+arch_initcall(pcifront_x86_stub_init);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/i386/power/Makefile linux-2.6.18-xen/arch/i386/power/Makefile
---- linux-2.6.18.3/arch/i386/power/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/i386/power/Makefile 2006-11-19 14:26:22.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/i386/power/Makefile b/arch/i386/power/Makefile
+--- a/arch/i386/power/Makefile 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/i386/power/Makefile 2007-03-14 10:55:14.000000000 +0100
@@ -1,2 +1,4 @@
-obj-$(CONFIG_PM) += cpu.o
+obj-$(CONFIG_PM_LEGACY) += cpu.o
+obj-$(CONFIG_SOFTWARE_SUSPEND) += cpu.o
+obj-$(CONFIG_ACPI_SLEEP) += cpu.o
- obj-$(CONFIG_SOFTWARE_SUSPEND) += swsusp.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/dig/setup.c linux-2.6.18-xen/arch/ia64/dig/setup.c
---- linux-2.6.18.3/arch/ia64/dig/setup.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/dig/setup.c 2006-11-19 14:26:22.000000000 +0100
+ obj-$(CONFIG_SOFTWARE_SUSPEND) += swsusp.o suspend.o
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/dig/setup.c b/arch/ia64/dig/setup.c
+--- a/arch/ia64/dig/setup.c 2007-03-12 21:58:05.000000000 +0100
++++ b/arch/ia64/dig/setup.c 2007-03-14 10:55:14.000000000 +0100
@@ -24,6 +24,8 @@
#include <asm/machvec.h>
#include <asm/system.h>
@@ -21809,19 +17834,19 @@
+ xen_start_info->console.domU.evtchn = 0;
+#endif
}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/hp/sim/Makefile linux-2.6.18-xen/arch/ia64/hp/sim/Makefile
---- linux-2.6.18.3/arch/ia64/hp/sim/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/hp/sim/Makefile 2006-11-19 14:26:22.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/hp/sim/Makefile b/arch/ia64/hp/sim/Makefile
+--- a/arch/ia64/hp/sim/Makefile 2007-03-12 21:58:05.000000000 +0100
++++ b/arch/ia64/hp/sim/Makefile 2007-03-14 10:55:14.000000000 +0100
@@ -14,3 +14,5 @@
obj-$(CONFIG_HP_SIMSERIAL) += simserial.o
obj-$(CONFIG_HP_SIMSERIAL_CONSOLE) += hpsim_console.o
obj-$(CONFIG_HP_SIMSCSI) += simscsi.o
+obj-$(CONFIG_XEN) += simserial.o
+obj-$(CONFIG_XEN) += hpsim_console.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/Kconfig linux-2.6.18-xen/arch/ia64/Kconfig
---- linux-2.6.18.3/arch/ia64/Kconfig 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/Kconfig 2006-11-19 14:26:22.000000000 +0100
-@@ -58,6 +58,20 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/Kconfig b/arch/ia64/Kconfig
+--- a/arch/ia64/Kconfig 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/ia64/Kconfig 2007-03-14 10:55:14.000000000 +0100
+@@ -67,6 +67,20 @@
bool
default y
@@ -21842,7 +17867,7 @@
config SCHED_NO_NO_OMIT_FRAME_POINTER
bool
default y
-@@ -465,6 +479,21 @@
+@@ -500,6 +514,21 @@
bool
default PCI
@@ -21864,7 +17889,7 @@
source "drivers/pci/pcie/Kconfig"
source "drivers/pci/Kconfig"
-@@ -528,3 +557,34 @@
+@@ -572,3 +601,34 @@
source "security/Kconfig"
source "crypto/Kconfig"
@@ -21899,9 +17924,9 @@
+endif
+
+source "drivers/xen/Kconfig"
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/kernel/asm-offsets.c linux-2.6.18-xen/arch/ia64/kernel/asm-offsets.c
---- linux-2.6.18.3/arch/ia64/kernel/asm-offsets.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/kernel/asm-offsets.c 2006-11-19 14:26:22.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/kernel/asm-offsets.c b/arch/ia64/kernel/asm-offsets.c
+--- a/arch/ia64/kernel/asm-offsets.c 2007-03-15 15:56:05.000000000 +0100
++++ b/arch/ia64/kernel/asm-offsets.c 2007-03-14 10:55:14.000000000 +0100
@@ -268,4 +268,29 @@
DEFINE(IA64_TIME_SOURCE_MMIO64, TIME_SOURCE_MMIO64);
DEFINE(IA64_TIME_SOURCE_MMIO32, TIME_SOURCE_MMIO32);
@@ -21932,9 +17957,9 @@
+ DEFINE_MAPPED_REG_OFS(XSI_BANK1_R16_OFS, bank1_regs[0]);
+#endif /* CONFIG_XEN */
}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/kernel/entry.S linux-2.6.18-xen/arch/ia64/kernel/entry.S
---- linux-2.6.18.3/arch/ia64/kernel/entry.S 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/kernel/entry.S 2006-11-19 14:26:22.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/kernel/entry.S b/arch/ia64/kernel/entry.S
+--- a/arch/ia64/kernel/entry.S 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/ia64/kernel/entry.S 2007-03-14 10:55:14.000000000 +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.
@@ -22054,9 +18079,9 @@
;;
mov ar.unat=r9
br.many b7
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/kernel/gate.lds.S linux-2.6.18-xen/arch/ia64/kernel/gate.lds.S
---- linux-2.6.18.3/arch/ia64/kernel/gate.lds.S 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/kernel/gate.lds.S 2006-11-19 14:26:22.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/kernel/gate.lds.S b/arch/ia64/kernel/gate.lds.S
+--- a/arch/ia64/kernel/gate.lds.S 2007-03-12 21:58:05.000000000 +0100
++++ b/arch/ia64/kernel/gate.lds.S 2007-03-14 10:55:14.000000000 +0100
@@ -43,6 +43,28 @@
__start_gate_brl_fsys_bubble_down_patchlist = .;
*(.data.patch.brl_fsys_bubble_down)
@@ -22086,9 +18111,9 @@
} :readable
.IA_64.unwind_info : { *(.IA_64.unwind_info*) }
.IA_64.unwind : { *(.IA_64.unwind*) } :readable :unwind
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/kernel/gate.S linux-2.6.18-xen/arch/ia64/kernel/gate.S
---- linux-2.6.18.3/arch/ia64/kernel/gate.S 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/kernel/gate.S 2006-11-19 14:26:22.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/kernel/gate.S b/arch/ia64/kernel/gate.S
+--- a/arch/ia64/kernel/gate.S 2007-03-12 21:58:05.000000000 +0100
++++ b/arch/ia64/kernel/gate.S 2007-03-14 10:55:14.000000000 +0100
@@ -6,13 +6,15 @@
* David Mosberger-Tang <davidm at hpl.hp.com>
*/
@@ -22258,9 +18283,9 @@
(p9) mov r8=ENOSYS
FSYS_RETURN
END(__kernel_syscall_via_epc)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/kernel/head.S linux-2.6.18-xen/arch/ia64/kernel/head.S
---- linux-2.6.18.3/arch/ia64/kernel/head.S 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/kernel/head.S 2006-11-19 14:26:22.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/kernel/head.S b/arch/ia64/kernel/head.S
+--- a/arch/ia64/kernel/head.S 2007-03-12 21:58:05.000000000 +0100
++++ b/arch/ia64/kernel/head.S 2007-03-14 10:55:14.000000000 +0100
@@ -367,6 +367,12 @@
;;
(isBP) st8 [r2]=r28 // save the address of the boot param area passed by the bootloader
@@ -22274,9 +18299,9 @@
#ifdef CONFIG_SMP
(isAP) br.call.sptk.many rp=start_secondary
.ret0:
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/kernel/iosapic.c linux-2.6.18-xen/arch/ia64/kernel/iosapic.c
---- linux-2.6.18.3/arch/ia64/kernel/iosapic.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/kernel/iosapic.c 2006-11-19 14:26:22.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/kernel/iosapic.c b/arch/ia64/kernel/iosapic.c
+--- a/arch/ia64/kernel/iosapic.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/ia64/kernel/iosapic.c 2007-03-14 10:55:14.000000000 +0100
@@ -159,6 +159,65 @@
static int iosapic_kmalloc_ok;
static LIST_HEAD(free_rte_list);
@@ -22343,7 +18368,7 @@
/*
* Find an IOSAPIC associated with a GSI
*/
-@@ -653,6 +712,9 @@
+@@ -674,6 +733,9 @@
iosapic_intr_info[vector].dmode = delivery;
iosapic_intr_info[vector].trigger = trigger;
@@ -22353,7 +18378,7 @@
if (trigger == IOSAPIC_EDGE)
irq_type = &irq_type_iosapic_edge;
else
-@@ -1015,6 +1077,9 @@
+@@ -1036,6 +1098,9 @@
}
pcat_compat = system_pcat_compat;
@@ -22363,20 +18388,21 @@
if (pcat_compat) {
/*
* Disable the compatibility mode interrupts (8259 style),
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/kernel/irq_ia64.c linux-2.6.18-xen/arch/ia64/kernel/irq_ia64.c
---- linux-2.6.18.3/arch/ia64/kernel/irq_ia64.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/kernel/irq_ia64.c 2006-11-19 14:26:22.000000000 +0100
-@@ -30,6 +30,9 @@
- #include <linux/smp_lock.h>
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/kernel/irq_ia64.c b/arch/ia64/kernel/irq_ia64.c
+--- a/arch/ia64/kernel/irq_ia64.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/ia64/kernel/irq_ia64.c 2007-03-14 10:55:14.000000000 +0100
+@@ -31,6 +31,10 @@
#include <linux/threads.h>
#include <linux/bitops.h>
+ #include <linux/irq.h>
+#ifdef CONFIG_XEN
+#include <linux/cpu.h>
+#endif
++
#include <asm/delay.h>
#include <asm/intrinsics.h>
-@@ -69,6 +72,13 @@
+@@ -70,6 +74,13 @@
assign_irq_vector (int irq)
{
int pos, vector;
@@ -22390,7 +18416,7 @@
again:
pos = find_first_zero_bit(ia64_vector_mask, IA64_NUM_DEVICE_VECTORS);
vector = IA64_FIRST_DEVICE_VECTOR + pos;
-@@ -240,14 +250,215 @@
+@@ -280,14 +291,215 @@
};
#endif
@@ -22606,7 +18632,7 @@
desc = irq_desc + irq;
desc->status |= IRQ_PER_CPU;
desc->chip = &irq_type_ia64_lsapic;
-@@ -259,6 +470,21 @@
+@@ -299,6 +511,21 @@
void __init
init_IRQ (void)
{
@@ -22628,7 +18654,7 @@
register_percpu_irq(IA64_SPURIOUS_INT_VECTOR, NULL);
#ifdef CONFIG_SMP
register_percpu_irq(IA64_IPI_VECTOR, &ipi_irqaction);
-@@ -276,6 +502,39 @@
+@@ -317,6 +544,39 @@
unsigned long ipi_data;
unsigned long phys_cpu_id;
@@ -22668,9 +18694,9 @@
#ifdef CONFIG_SMP
phys_cpu_id = cpu_physical_id(cpu);
#else
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/kernel/pal.S linux-2.6.18-xen/arch/ia64/kernel/pal.S
---- linux-2.6.18.3/arch/ia64/kernel/pal.S 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/kernel/pal.S 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/kernel/pal.S b/arch/ia64/kernel/pal.S
+--- a/arch/ia64/kernel/pal.S 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/ia64/kernel/pal.S 2007-03-14 10:55:14.000000000 +0100
@@ -16,6 +16,7 @@
#include <asm/processor.h>
@@ -22679,16 +18705,7 @@
pal_entry_point:
data8 ia64_pal_default_handler
.text
-@@ -53,7 +54,7 @@
- * in4 1 ==> clear psr.ic, 0 ==> don't clear psr.ic
- *
- */
--GLOBAL_ENTRY(ia64_pal_call_static)
-+GLOBAL_ENTRY(__ia64_pal_call_static)
- .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 @@
+@@ -86,7 +87,7 @@
;;
srlz.d // seralize restoration of psr.l
br.ret.sptk.many b0
@@ -22697,9 +18714,9 @@
/*
* Make a PAL call using the stacked registers calling convention.
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/kernel/patch.c linux-2.6.18-xen/arch/ia64/kernel/patch.c
---- linux-2.6.18.3/arch/ia64/kernel/patch.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/kernel/patch.c 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/kernel/patch.c b/arch/ia64/kernel/patch.c
+--- a/arch/ia64/kernel/patch.c 2007-03-12 21:58:05.000000000 +0100
++++ b/arch/ia64/kernel/patch.c 2007-03-14 10:55:14.000000000 +0100
@@ -184,6 +184,73 @@
ia64_srlz_i();
}
@@ -22785,10 +18802,10 @@
ia64_patch_vtop(START(vtop), END(vtop));
ia64_patch_mckinley_e9(START(mckinley_e9), END(mckinley_e9));
}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/kernel/setup.c linux-2.6.18-xen/arch/ia64/kernel/setup.c
---- linux-2.6.18.3/arch/ia64/kernel/setup.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/kernel/setup.c 2006-11-19 14:26:23.000000000 +0100
-@@ -60,6 +60,10 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/kernel/setup.c b/arch/ia64/kernel/setup.c
+--- a/arch/ia64/kernel/setup.c 2007-03-16 18:49:20.000000000 +0100
++++ b/arch/ia64/kernel/setup.c 2007-03-14 10:55:14.000000000 +0100
+@@ -61,6 +61,10 @@
#include <asm/system.h>
#include <asm/unistd.h>
#include <asm/system.h>
@@ -22799,7 +18816,7 @@
#if defined(CONFIG_SMP) && (IA64_CPU_SIZE > PAGE_SIZE)
# error "struct cpuinfo_ia64 too big!"
-@@ -70,6 +74,22 @@
+@@ -71,6 +75,22 @@
EXPORT_SYMBOL(__per_cpu_offset);
#endif
@@ -22822,7 +18839,7 @@
extern void ia64_setup_printk_clock(void);
DEFINE_PER_CPU(struct cpuinfo_ia64, cpu_info);
-@@ -176,15 +196,33 @@
+@@ -177,15 +197,33 @@
return 0;
}
@@ -22857,7 +18874,7 @@
struct rsvd_region tmp;
tmp = rsvd_region[j];
rsvd_region[j] = rsvd_region[j + 1];
-@@ -192,6 +230,36 @@
+@@ -193,6 +231,36 @@
}
}
}
@@ -22894,7 +18911,7 @@
}
/*
-@@ -242,6 +310,14 @@
+@@ -243,6 +311,14 @@
rsvd_region[n].end = (unsigned long) ia64_imva(_end);
n++;
@@ -22909,7 +18926,7 @@
#ifdef CONFIG_BLK_DEV_INITRD
if (ia64_boot_param->initrd_start) {
rsvd_region[n].start = (unsigned long)__va(ia64_boot_param->initrd_start);
-@@ -333,6 +409,16 @@
+@@ -376,6 +452,16 @@
{
int earlycons = 0;
@@ -22926,7 +18943,7 @@
#ifdef CONFIG_SERIAL_SGI_L1_CONSOLE
{
extern int sn_serial_console_early_setup(void);
-@@ -402,6 +488,14 @@
+@@ -460,6 +546,14 @@
{
unw_init();
@@ -22941,7 +18958,7 @@
ia64_patch_vtop((u64) __start___vtop_patchlist, (u64) __end___vtop_patchlist);
*cmdline_p = __va(ia64_boot_param->command_line);
-@@ -478,6 +572,23 @@
+@@ -538,6 +632,23 @@
conswitchp = &vga_con;
# endif
}
@@ -22956,16 +18973,16 @@
+ xen_start_info->nr_pages, xen_start_info->flags);
+
+ if (!is_initial_xendomain()) {
-+ extern int console_use_vt;
++#if !defined(CONFIG_VT) || !defined(CONFIG_DUMMY_CONSOLE)
+ conswitchp = NULL;
-+ console_use_vt = 0;
++#endif
+ }
+ }
+#endif
#endif
/* enable IA-64 Machine Check Abort Handling unless disabled */
-@@ -486,6 +597,9 @@
+@@ -546,6 +657,9 @@
platform_setup(cmdline_p);
paging_init();
@@ -22975,7 +18992,7 @@
}
/*
-@@ -870,6 +984,15 @@
+@@ -951,6 +1065,15 @@
/* 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();
@@ -22991,9 +19008,9 @@
pm_idle = default_idle;
}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/Makefile linux-2.6.18-xen/arch/ia64/Makefile
---- linux-2.6.18.3/arch/ia64/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/Makefile 2006-11-19 14:26:22.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/Makefile b/arch/ia64/Makefile
+--- a/arch/ia64/Makefile 2007-03-12 21:58:05.000000000 +0100
++++ b/arch/ia64/Makefile 2007-03-14 10:55:14.000000000 +0100
@@ -45,6 +45,12 @@
endif
@@ -23034,9 +19051,9 @@
define archhelp
echo '* compressed - Build compressed kernel image'
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/mm/ioremap.c linux-2.6.18-xen/arch/ia64/mm/ioremap.c
---- linux-2.6.18.3/arch/ia64/mm/ioremap.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/ia64/mm/ioremap.c 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/mm/ioremap.c b/arch/ia64/mm/ioremap.c
+--- a/arch/ia64/mm/ioremap.c 2007-03-12 21:58:05.000000000 +0100
++++ b/arch/ia64/mm/ioremap.c 2007-03-14 10:55:15.000000000 +0100
@@ -16,6 +16,9 @@
static inline void __iomem *
__ioremap (unsigned long offset, unsigned long size)
@@ -23047,15 +19064,15 @@
return (void __iomem *) (__IA64_UNCACHED_OFFSET | offset);
}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/xen/drivers/README linux-2.6.18-xen/arch/ia64/xen/drivers/README
---- linux-2.6.18.3/arch/ia64/xen/drivers/README 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/ia64/xen/drivers/README 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/xen/drivers/README b/arch/ia64/xen/drivers/README
+--- a/arch/ia64/xen/drivers/README 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/ia64/xen/drivers/README 2007-03-14 10:55:15.000000000 +0100
@@ -0,0 +1,2 @@
+This is a temporary location for source/Makefiles that need to be
+patched/reworked in drivers/xen to work with xenlinux/ia64.
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/xen/hypercall.S linux-2.6.18-xen/arch/ia64/xen/hypercall.S
---- linux-2.6.18.3/arch/ia64/xen/hypercall.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/ia64/xen/hypercall.S 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/xen/hypercall.S b/arch/ia64/xen/hypercall.S
+--- a/arch/ia64/xen/hypercall.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/ia64/xen/hypercall.S 2007-03-14 10:55:15.000000000 +0100
@@ -0,0 +1,413 @@
+/*
+ * Support routines for Xen hypercalls
@@ -23470,9 +19487,9 @@
+ ;;
+END(xen_ssm_i_1)
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/xen/hypervisor.c linux-2.6.18-xen/arch/ia64/xen/hypervisor.c
---- linux-2.6.18.3/arch/ia64/xen/hypervisor.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/ia64/xen/hypervisor.c 2006-12-05 18:42:36.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/xen/hypervisor.c b/arch/ia64/xen/hypervisor.c
+--- a/arch/ia64/xen/hypervisor.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/ia64/xen/hypervisor.c 2007-03-14 10:55:15.000000000 +0100
@@ -0,0 +1,847 @@
+/******************************************************************************
+ * include/asm-ia64/shadow.h
@@ -24321,9 +20338,9 @@
+ /* Just trigger a tick. */
+ ia64_cpu_local_tick();
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/xen/Makefile linux-2.6.18-xen/arch/ia64/xen/Makefile
---- linux-2.6.18.3/arch/ia64/xen/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/ia64/xen/Makefile 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/xen/Makefile b/arch/ia64/xen/Makefile
+--- a/arch/ia64/xen/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/ia64/xen/Makefile 2007-03-14 10:55:15.000000000 +0100
@@ -0,0 +1,8 @@
+#
+# Makefile for Xen components
@@ -24333,9 +20350,9 @@
+ hypervisor.o pci-dma-xen.o util.o
+
+pci-dma-xen-y := ../../i386/kernel/pci-dma-xen.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/xen/util.c linux-2.6.18-xen/arch/ia64/xen/util.c
---- linux-2.6.18.3/arch/ia64/xen/util.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/ia64/xen/util.c 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/xen/util.c b/arch/ia64/xen/util.c
+--- a/arch/ia64/xen/util.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/ia64/xen/util.c 2007-03-14 10:55:15.000000000 +0100
@@ -0,0 +1,115 @@
+/******************************************************************************
+ * arch/ia64/xen/util.c
@@ -24452,9 +20469,9 @@
+ * tab-width: 8
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/xen/xenentry.S linux-2.6.18-xen/arch/ia64/xen/xenentry.S
---- linux-2.6.18.3/arch/ia64/xen/xenentry.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/ia64/xen/xenentry.S 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/xen/xenentry.S b/arch/ia64/xen/xenentry.S
+--- a/arch/ia64/xen/xenentry.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/ia64/xen/xenentry.S 2007-03-14 10:55:15.000000000 +0100
@@ -0,0 +1,867 @@
+/*
+ * ia64/xen/entry.S
@@ -25323,9 +21340,9 @@
+#else
+END(ia64_leave_kernel)
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/xen/xenhpski.c linux-2.6.18-xen/arch/ia64/xen/xenhpski.c
---- linux-2.6.18.3/arch/ia64/xen/xenhpski.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/ia64/xen/xenhpski.c 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/xen/xenhpski.c b/arch/ia64/xen/xenhpski.c
+--- a/arch/ia64/xen/xenhpski.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/ia64/xen/xenhpski.c 2007-03-14 10:55:15.000000000 +0100
@@ -0,0 +1,19 @@
+
+extern unsigned long xen_get_cpuid(int);
@@ -25346,9 +21363,9 @@
+ return 1;
+}
+
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/xen/xenivt.S linux-2.6.18-xen/arch/ia64/xen/xenivt.S
---- linux-2.6.18.3/arch/ia64/xen/xenivt.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/ia64/xen/xenivt.S 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/xen/xenivt.S b/arch/ia64/xen/xenivt.S
+--- a/arch/ia64/xen/xenivt.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/ia64/xen/xenivt.S 2007-03-14 10:55:15.000000000 +0100
@@ -0,0 +1,2169 @@
+/*
+ * arch/ia64/xen/ivt.S
@@ -27519,9 +23536,9 @@
+ br.call.sptk.many b6=evtchn_do_upcall
+END(xen_event_callback)
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/xen/xenminstate.h linux-2.6.18-xen/arch/ia64/xen/xenminstate.h
---- linux-2.6.18.3/arch/ia64/xen/xenminstate.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/ia64/xen/xenminstate.h 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/xen/xenminstate.h b/arch/ia64/xen/xenminstate.h
+--- a/arch/ia64/xen/xenminstate.h 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/ia64/xen/xenminstate.h 2007-03-14 10:55:15.000000000 +0100
@@ -0,0 +1,368 @@
+
+#include <asm/cache.h>
@@ -27891,9 +23908,9 @@
+#else
+#define SAVE_MIN DO_SAVE_MIN( , mov r30=r0, )
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/xen/xenpal.S linux-2.6.18-xen/arch/ia64/xen/xenpal.S
---- linux-2.6.18.3/arch/ia64/xen/xenpal.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/ia64/xen/xenpal.S 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/xen/xenpal.S b/arch/ia64/xen/xenpal.S
+--- a/arch/ia64/xen/xenpal.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/ia64/xen/xenpal.S 2007-03-14 10:55:15.000000000 +0100
@@ -0,0 +1,76 @@
+/*
+ * ia64/xen/xenpal.S
@@ -27971,9 +23988,9 @@
+ srlz.d // seralize restoration of psr.l
+ br.ret.sptk.many b0
+END(xen_pal_call_static)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/ia64/xen/xensetup.S linux-2.6.18-xen/arch/ia64/xen/xensetup.S
---- linux-2.6.18.3/arch/ia64/xen/xensetup.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/ia64/xen/xensetup.S 2006-11-19 14:26:23.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/ia64/xen/xensetup.S b/arch/ia64/xen/xensetup.S
+--- a/arch/ia64/xen/xensetup.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/ia64/xen/xensetup.S 2007-03-14 10:55:15.000000000 +0100
@@ -0,0 +1,54 @@
+/*
+ * Support routines for Xen
@@ -28029,9 +24046,9 @@
+ mov ar.pfs=r20
+ br.ret.sptk.many b0
+END(HYPERVISOR_suspend)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/um/kernel/physmem.c linux-2.6.18-xen/arch/um/kernel/physmem.c
---- linux-2.6.18.3/arch/um/kernel/physmem.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/um/kernel/physmem.c 2006-11-19 14:26:32.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/um/kernel/physmem.c b/arch/um/kernel/physmem.c
+--- a/arch/um/kernel/physmem.c 2007-03-15 15:56:05.000000000 +0100
++++ b/arch/um/kernel/physmem.c 2007-03-14 10:55:20.000000000 +0100
@@ -226,7 +226,7 @@
EXPORT_SYMBOL(physmem_remove_mapping);
EXPORT_SYMBOL(physmem_subst_mapping);
@@ -28050,9 +24067,9 @@
}
int is_remapped(void *virt)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/ia32/ia32entry-xen.S linux-2.6.18-xen/arch/x86_64/ia32/ia32entry-xen.S
---- linux-2.6.18.3/arch/x86_64/ia32/ia32entry-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/ia32/ia32entry-xen.S 2006-11-19 14:26:33.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/ia32/ia32entry-xen.S b/arch/x86_64/ia32/ia32entry-xen.S
+--- a/arch/x86_64/ia32/ia32entry-xen.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/ia32/ia32entry-xen.S 2007-03-14 10:55:21.000000000 +0100
@@ -0,0 +1,743 @@
+/*
+ * Compatibility mode system call entry point for x86-64.
@@ -28148,7 +24165,7 @@
+ CFI_REGISTER rsp,rbp
+ __swapgs
+ movq %gs:pda_kernelstack, %rsp
-+ addq $(PDA_STACKOFFSET),%rsp
++ addq $(PDA_STACKOFFSET),%rsp
+ /*
+ * No need to follow this irqs on/off section: the syscall
+ * disabled irqs, here we enable it straight after entry:
@@ -28218,7 +24235,7 @@
+ XEN_UNBLOCK_EVENTS(%r11)
+ __sti /* sti only takes effect after the next instruction */
+ /* sysexit */
-+ .byte 0xf, 0x35 /* TBD */
++ .byte 0xf, 0x35
+
+sysenter_tracesys:
+ CFI_RESTORE_STATE
@@ -28322,7 +24339,7 @@
+ movl RSP-ARGOFFSET(%rsp),%esp
+ CFI_RESTORE rsp
+ __swapgs
-+ sysretl /* TBD */
++ sysretl
+
+cstar_tracesys:
+ CFI_RESTORE_STATE
@@ -28391,7 +24408,6 @@
+ pushq %rax
+ CFI_ADJUST_CFA_OFFSET 8
+ cld
-+/* 1: jmp 1b */
+ /* note the registers are not zero extended to the sf.
+ this could be a problem. */
+ SAVE_ARGS 0,0,1
@@ -28786,8 +24802,8 @@
+ .quad sys_readlinkat /* 305 */
+ .quad sys_fchmodat
+ .quad sys_faccessat
-+ .quad quiet_ni_syscall /* pselect6 for now */
-+ .quad quiet_ni_syscall /* ppoll for now */
++ .quad compat_sys_pselect6
++ .quad compat_sys_ppoll
+ .quad sys_unshare /* 310 */
+ .quad compat_sys_set_robust_list
+ .quad compat_sys_get_robust_list
@@ -28796,11 +24812,32 @@
+ .quad sys_tee
+ .quad compat_sys_vmsplice
+ .quad compat_sys_move_pages
++ .quad sys_getcpu
+ia32_syscall_end:
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/ia32/Makefile linux-2.6.18-xen/arch/x86_64/ia32/Makefile
---- linux-2.6.18.3/arch/x86_64/ia32/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/x86_64/ia32/Makefile 2006-11-19 14:26:33.000000000 +0100
-@@ -27,9 +27,25 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/ia32/Makefile b/arch/x86_64/ia32/Makefile
+--- a/arch/x86_64/ia32/Makefile 2007-03-12 21:58:06.000000000 +0100
++++ b/arch/x86_64/ia32/Makefile 2007-03-14 10:55:21.000000000 +0100
+@@ -14,11 +14,19 @@
+ audit-class-$(CONFIG_AUDIT) := audit.o
+ obj-$(CONFIG_IA32_EMULATION) += $(audit-class-y)
+
++ifdef CONFIG_XEN
++$(obj)/syscall32_syscall.o: \
++ $(foreach F,int80 sysenter syscall,$(obj)/vsyscall-$F.so)
++
++targets := $(foreach F,int80 sysenter syscall,vsyscall-$F.o vsyscall-$F.so)
++
++else
+ $(obj)/syscall32_syscall.o: \
+ $(foreach F,sysenter syscall,$(obj)/vsyscall-$F.so)
+
+ # Teach kbuild about targets
+ targets := $(foreach F,sysenter syscall,vsyscall-$F.o vsyscall-$F.so)
++endif
+
+ # The DSO images are built using a special linker script
+ quiet_cmd_syscall = SYSCALL $@
+@@ -27,9 +35,20 @@
-Wl,-soname=linux-gate.so.1 -o $@ \
-Wl,-T,$(filter-out FORCE,$^)
@@ -28816,141 +24853,38 @@
+
+ifdef CONFIG_XEN
+AFLAGS_vsyscall-int80.o = -m32 -Wa,-32 -Iarch/i386/kernel
-+CFLAGS_syscall32-xen.o += -DUSE_INT80
-+AFLAGS_syscall32_syscall-xen.o += -DUSE_INT80
-+
-+$(obj)/syscall32_syscall-xen.o: \
-+ $(foreach F,int80 sysenter syscall,$(obj)/vsyscall-$F.so)
-+
-+targets := $(foreach F,int80 sysenter syscall,vsyscall-$F.o vsyscall-$F.so)
++CFLAGS_syscall32.o += -DUSE_INT80
++AFLAGS_syscall32_syscall.o += -DUSE_INT80
+
+include $(srctree)/scripts/Makefile.xen
+
+obj-y := $(call cherrypickxen, $(obj-y))
+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/ia32/syscall32_syscall-xen.S linux-2.6.18-xen/arch/x86_64/ia32/syscall32_syscall-xen.S
---- linux-2.6.18.3/arch/x86_64/ia32/syscall32_syscall-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/ia32/syscall32_syscall-xen.S 2006-11-19 14:26:33.000000000 +0100
-@@ -0,0 +1,28 @@
-+/* 32bit VDSOs mapped into user space. */
-+
-+ .section ".init.data","aw"
-+
-+#ifdef USE_INT80
-+
-+ .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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/ia32/syscall32-xen.c linux-2.6.18-xen/arch/x86_64/ia32/syscall32-xen.c
---- linux-2.6.18.3/arch/x86_64/ia32/syscall32-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/ia32/syscall32-xen.c 2006-11-19 14:26:33.000000000 +0100
-@@ -0,0 +1,128 @@
-+/* 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>
-+
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/ia32/syscall32.c b/arch/x86_64/ia32/syscall32.c
+--- a/arch/x86_64/ia32/syscall32.c 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/ia32/syscall32.c 2007-03-14 10:55:21.000000000 +0100
+@@ -14,12 +14,17 @@
+ #include <asm/tlbflush.h>
+ #include <asm/ia32_unistd.h>
+
+#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;
-+
-+char *syscall32_page;
+ extern unsigned char syscall32_syscall[], syscall32_syscall_end[];
+ extern unsigned char syscall32_sysenter[], syscall32_sysenter_end[];
+ extern int sysctl_vsyscall32;
+
+ char *syscall32_page;
+#ifndef USE_INT80
-+static int use_sysenter = -1;
+ static int use_sysenter = -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");
+
+ static struct page *
+ syscall32_nopage(struct vm_area_struct *vma, unsigned long adr, int *type)
+@@ -95,6 +100,14 @@
+ syscall32_page = (void *)get_zeroed_page(GFP_KERNEL);
+ if (!syscall32_page)
+ panic("Cannot allocate syscall32 page");
+
+#ifdef USE_INT80
+ /*
@@ -28959,42 +24893,62 @@
+ 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);
-+ }
+ if (use_sysenter > 0) {
+ memcpy(syscall32_page, syscall32_sysenter,
+ syscall32_sysenter_end - syscall32_sysenter);
+@@ -102,14 +115,20 @@
+ memcpy(syscall32_page, syscall32_syscall,
+ syscall32_syscall_end - syscall32_syscall);
+ }
+#endif
-+ return 0;
-+}
+ return 0;
+ }
+-
+-__initcall(init_syscall32);
+
+/*
+ * 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)
-+{
+
+ /* 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);
+ if (use_sysenter < 0)
+ use_sysenter = (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL);
+
+@@ -120,4 +139,5 @@
+ checking_wrmsrl(MSR_IA32_SYSENTER_EIP, (u64)ia32_sysenter_target);
+
+ wrmsrl(MSR_CSTAR, ia32_cstar_target);
++#endif
+ }
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/ia32/syscall32_syscall.S b/arch/x86_64/ia32/syscall32_syscall.S
+--- a/arch/x86_64/ia32/syscall32_syscall.S 2007-03-12 21:58:06.000000000 +0100
++++ b/arch/x86_64/ia32/syscall32_syscall.S 2007-03-14 10:55:21.000000000 +0100
+@@ -2,6 +2,17 @@
+
+ .section ".init.data","aw"
+
++#ifdef USE_INT80
++
++ .globl syscall32_int80
++ .globl syscall32_int80_end
+
-+ /* 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);
++syscall32_int80:
++ .incbin "arch/x86_64/ia32/vsyscall-int80.so"
++syscall32_int80_end:
+
-+ wrmsrl(MSR_CSTAR, ia32_cstar_target);
+#endif
-+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/ia32/vsyscall-int80.S linux-2.6.18-xen/arch/x86_64/ia32/vsyscall-int80.S
---- linux-2.6.18.3/arch/x86_64/ia32/vsyscall-int80.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/ia32/vsyscall-int80.S 2006-11-19 14:26:33.000000000 +0100
++
+ .globl syscall32_syscall
+ .globl syscall32_syscall_end
+
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/ia32/vsyscall-int80.S b/arch/x86_64/ia32/vsyscall-int80.S
+--- a/arch/x86_64/ia32/vsyscall-int80.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/ia32/vsyscall-int80.S 2007-03-14 10:55:21.000000000 +0100
@@ -0,0 +1,58 @@
+/*
+ * Code for the vsyscall page. This version uses the old int $0x80 method.
@@ -29054,9 +25008,9 @@
+ */
+#define SYSCALL_ENTER_KERNEL int $0x80
+#include "vsyscall-sigreturn.S"
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/ia32/vsyscall-sigreturn.S linux-2.6.18-xen/arch/x86_64/ia32/vsyscall-sigreturn.S
---- linux-2.6.18.3/arch/x86_64/ia32/vsyscall-sigreturn.S 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/x86_64/ia32/vsyscall-sigreturn.S 2006-11-19 14:26:33.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/ia32/vsyscall-sigreturn.S b/arch/x86_64/ia32/vsyscall-sigreturn.S
+--- a/arch/x86_64/ia32/vsyscall-sigreturn.S 2007-03-12 21:58:06.000000000 +0100
++++ b/arch/x86_64/ia32/vsyscall-sigreturn.S 2007-03-14 10:55:21.000000000 +0100
@@ -139,5 +139,5 @@
.align 4
.LENDFDE3:
@@ -29064,10 +25018,10 @@
-#include "../../i386/kernel/vsyscall-note.S"
+#include <vsyscall-note.S>
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/Kconfig linux-2.6.18-xen/arch/x86_64/Kconfig
---- linux-2.6.18.3/arch/x86_64/Kconfig 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/x86_64/Kconfig 2006-12-05 18:42:36.000000000 +0100
-@@ -30,6 +30,7 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/Kconfig b/arch/x86_64/Kconfig
+--- a/arch/x86_64/Kconfig 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/Kconfig 2007-03-14 10:55:21.000000000 +0100
+@@ -34,6 +34,7 @@
config STACKTRACE_SUPPORT
bool
@@ -29075,7 +25029,7 @@
default y
config SEMAPHORE_SLEEPERS
-@@ -135,6 +136,22 @@
+@@ -171,6 +172,22 @@
endchoice
@@ -29098,7 +25052,7 @@
#
# Define implied options from the CPU selection here
#
-@@ -155,6 +172,7 @@
+@@ -191,6 +208,7 @@
config X86_TSC
bool
@@ -29106,7 +25060,7 @@
default y
config X86_GOOD_APIC
-@@ -197,7 +215,7 @@
+@@ -239,7 +257,7 @@
config X86_HT
bool
@@ -29115,7 +25069,7 @@
default y
config MATH_EMULATION
-@@ -211,14 +229,22 @@
+@@ -253,14 +271,22 @@
config X86_IO_APIC
bool
@@ -29138,7 +25092,7 @@
---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 @@
+@@ -301,7 +327,7 @@
config SCHED_SMT
bool "SMT (Hyperthreading) scheduler support"
@@ -29147,7 +25101,7 @@
default n
help
SMT scheduler support improves the CPU scheduler's decision making
-@@ -269,7 +295,7 @@
+@@ -311,7 +337,7 @@
config SCHED_MC
bool "Multi-core scheduler support"
@@ -29156,7 +25110,7 @@
default y
help
Multi-core scheduler support improves the CPU scheduler's decision
-@@ -280,7 +306,7 @@
+@@ -322,7 +348,7 @@
config NUMA
bool "Non Uniform Memory Access (NUMA) Support"
@@ -29165,7 +25119,7 @@
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 @@
+@@ -378,7 +404,7 @@
config ARCH_SPARSEMEM_ENABLE
def_bool y
@@ -29174,7 +25128,7 @@
config ARCH_MEMORY_PROBE
def_bool y
-@@ -365,6 +391,7 @@
+@@ -406,6 +432,7 @@
int "Maximum number of CPUs (2-256)"
range 2 255
depends on SMP
@@ -29182,7 +25136,7 @@
default "8"
help
This allows you to specify the maximum number of CPUs which this
-@@ -387,6 +414,7 @@
+@@ -428,6 +455,7 @@
config HPET_TIMER
bool
@@ -29190,7 +25144,7 @@
default y
help
Use the IA-PC HPET (High Precision Event Timer) to manage
-@@ -407,7 +435,7 @@
+@@ -448,7 +476,7 @@
default y
select SWIOTLB
select AGP
@@ -29199,7 +25153,7 @@
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,
-@@ -444,6 +472,7 @@
+@@ -495,6 +523,7 @@
config X86_MCE
bool "Machine check support" if EMBEDDED
@@ -29207,16 +25161,15 @@
default y
help
Include a machine check error handler to report hardware errors.
-@@ -469,7 +498,7 @@
+@@ -520,6 +549,7 @@
config KEXEC
- bool "kexec system call (EXPERIMENTAL)"
-- depends on EXPERIMENTAL
-+ depends on EXPERIMENTAL && !X86_64_XEN
+ bool "kexec system call"
++ depends on !X86_64_XEN
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 +593,11 @@
+@@ -645,8 +675,11 @@
default y
menu "Power management options"
@@ -29228,7 +25181,7 @@
source "drivers/acpi/Kconfig"
-@@ -588,6 +620,21 @@
+@@ -669,6 +702,21 @@
bool "Support mmconfig PCI config space access"
depends on PCI && ACPI
@@ -29250,25 +25203,25 @@
source "drivers/pci/pcie/Kconfig"
source "drivers/pci/Kconfig"
-@@ -658,4 +705,6 @@
+@@ -739,4 +787,6 @@
source "crypto/Kconfig"
+source "drivers/xen/Kconfig"
+
source "lib/Kconfig"
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/acpi/Makefile linux-2.6.18-xen/arch/x86_64/kernel/acpi/Makefile
---- linux-2.6.18.3/arch/x86_64/kernel/acpi/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/x86_64/kernel/acpi/Makefile 2006-11-19 14:26:33.000000000 +0100
-@@ -7,3 +7,4 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/acpi/Makefile b/arch/x86_64/kernel/acpi/Makefile
+--- a/arch/x86_64/kernel/acpi/Makefile 2007-03-12 21:58:06.000000000 +0100
++++ b/arch/x86_64/kernel/acpi/Makefile 2007-03-14 10:55:21.000000000 +0100
+@@ -6,4 +6,3 @@
+ 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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/apic-xen.c linux-2.6.18-xen/arch/x86_64/kernel/apic-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/apic-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/apic-xen.c 2006-11-19 14:26:33.000000000 +0100
-@@ -0,0 +1,197 @@
+-
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/apic-xen.c b/arch/x86_64/kernel/apic-xen.c
+--- a/arch/x86_64/kernel/apic-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/apic-xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,254 @@
+/*
+ * Local APIC handling, local APIC timers
+ *
@@ -29296,16 +25249,21 @@
+#include <linux/kernel_stat.h>
+#include <linux/sysdev.h>
+#include <linux/module.h>
++#include <linux/ioport.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/pgalloc.h>
++#include <asm/mach_apic.h>
++#include <asm/nmi.h>
+#include <asm/idle.h>
++#include <asm/proto.h>
++#include <asm/timex.h>
++#include <asm/apic.h>
+
++int apic_mapped;
+int apic_verbosity;
+
+/*
@@ -29333,19 +25291,21 @@
+ return -EINVAL;
+}
+
-+void smp_local_timer_interrupt(struct pt_regs *regs)
++void smp_local_timer_interrupt(void)
+{
-+ profile_tick(CPU_PROFILING, regs);
++ profile_tick(CPU_PROFILING);
+#ifndef CONFIG_XEN
+#ifdef CONFIG_SMP
-+ update_process_times(user_mode(regs));
++ update_process_times(user_mode(get_irq_regs()));
+#endif
++ if (apic_runs_main_timer > 1 && smp_processor_id() == boot_cpu_id)
++ main_timer_handler();
+#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',
++ * 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),
@@ -29363,6 +25323,8 @@
+ */
+void smp_apic_timer_interrupt(struct pt_regs *regs)
+{
++ struct pt_regs *old_regs = set_irq_regs(regs);
++
+ /*
+ * the NMI deadlock-detector uses this.
+ */
@@ -29380,8 +25342,9 @@
+ */
+ exit_idle();
+ irq_enter();
-+ smp_local_timer_interrupt(regs);
++ smp_local_timer_interrupt();
+ irq_exit();
++ set_irq_regs(old_regs);
+}
+
+/*
@@ -29450,7 +25413,7 @@
+ irq_exit();
+}
+
-+int disable_apic;
++int disable_apic;
+
+/*
+ * This initializes the IO-APIC and APIC hardware if this is
@@ -29458,17 +25421,64 @@
+ */
+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
++ if (smp_found_config && !skip_ioapic_setup && nr_ioapics)
++ setup_IO_APIC();
++ return 0;
++}
++
++#ifndef CONFIG_XEN
++static __init int setup_disableapic(char *str)
++{
++ disable_apic = 1;
++ clear_bit(X86_FEATURE_APIC, boot_cpu_data.x86_capability);
++ return 0;
++}
++early_param("disableapic", setup_disableapic);
++
++/* same as disableapic, for compatibility */
++static __init int setup_nolapic(char *str)
++{
++ return setup_disableapic(str);
++}
++early_param("nolapic", setup_nolapic);
++
++static __init int setup_noapictimer(char *str)
++{
++ if (str[0] != ' ' && str[0] != 0)
++ return 0;
++ disable_apic_timer = 1;
++ return 1;
++}
++
++static __init int setup_apicmaintimer(char *str)
++{
++ apic_runs_main_timer = 1;
++ nohpet = 1;
++ return 1;
++}
++__setup("apicmaintimer", setup_apicmaintimer);
+
++static __init int setup_noapicmaintimer(char *str)
++{
++ apic_runs_main_timer = -1;
+ return 1;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/asm-offsets.c linux-2.6.18-xen/arch/x86_64/kernel/asm-offsets.c
---- linux-2.6.18.3/arch/x86_64/kernel/asm-offsets.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/x86_64/kernel/asm-offsets.c 2006-11-19 14:26:33.000000000 +0100
++__setup("noapicmaintimer", setup_noapicmaintimer);
++
++static __init int setup_apicpmtimer(char *s)
++{
++ apic_calibrate_pmtmr = 1;
++ notsc_setup(NULL);
++ return setup_apicmaintimer(NULL);
++}
++__setup("apicpmtimer", setup_apicpmtimer);
++
++__setup("noapictimer", setup_noapictimer);
++
++#endif
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/asm-offsets.c b/arch/x86_64/kernel/asm-offsets.c
+--- a/arch/x86_64/kernel/asm-offsets.c 2007-03-12 21:58:06.000000000 +0100
++++ b/arch/x86_64/kernel/asm-offsets.c 2007-03-14 10:55:21.000000000 +0100
@@ -67,8 +67,10 @@
DEFINE(pbe_address, offsetof(struct pbe, address));
DEFINE(pbe_orig_address, offsetof(struct pbe, orig_address));
@@ -29480,10 +25490,10 @@
BLANK();
DEFINE(crypto_tfm_ctx_offset, offsetof(struct crypto_tfm, __crt_ctx));
return 0;
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/e820-xen.c linux-2.6.18-xen/arch/x86_64/kernel/e820-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/e820-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/e820-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,757 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/e820-xen.c b/arch/x86_64/kernel/e820-xen.c
+--- a/arch/x86_64/kernel/e820-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/e820-xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,758 @@
+/*
+ * Handle the memory map.
+ * The functions here do the job until bootmem takes over.
@@ -29502,6 +25512,7 @@
+#include <linux/string.h>
+#include <linux/kexec.h>
+#include <linux/module.h>
++#include <linux/mm.h>
+
+#include <asm/pgtable.h>
+#include <asm/page.h>
@@ -29511,6 +25522,8 @@
+#include <asm/sections.h>
+#include <xen/interface/memory.h>
+
++struct e820map e820 __initdata;
++
+/*
+ * PFN of last memory page.
+ */
@@ -29527,7 +25540,7 @@
+/*
+ * Last pfn which the user wants to use.
+ */
-+unsigned long end_user_pfn = MAXMEM>>PAGE_SHIFT;
++static unsigned long __initdata end_user_pfn = MAXMEM>>PAGE_SHIFT;
+
+extern struct resource code_resource, data_resource;
+
@@ -29539,13 +25552,13 @@
+#ifndef CONFIG_XEN
+ /* various gunk below that needed for SMP startup */
+ if (addr < 0x8000) {
-+ *addrp = 0x8000;
++ *addrp = PAGE_ALIGN(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;
++ *addrp = PAGE_ALIGN(table_end << PAGE_SHIFT);
+ return 1;
+ }
+
@@ -29553,23 +25566,18 @@
+#ifdef CONFIG_BLK_DEV_INITRD
+ if (LOADER_TYPE && INITRD_START && last >= INITRD_START &&
+ addr < INITRD_START+INITRD_SIZE) {
-+ *addrp = INITRD_START + INITRD_SIZE;
++ *addrp = PAGE_ALIGN(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);
++ /* kernel code */
++ if (last >= __pa_symbol(&_text) && addr < __pa_symbol(&_end)) {
++ *addrp = PAGE_ALIGN(__pa_symbol(&_end));
+ return 1;
+ }
+
+ if (last >= ebda_addr && addr < ebda_addr + ebda_size) {
-+ *addrp = ebda_addr + ebda_size;
++ *addrp = PAGE_ALIGN(ebda_addr + ebda_size);
+ return 1;
+ }
+
@@ -29648,7 +25656,7 @@
+ continue;
+ while (bad_addr(&addr, size) && addr+size <= ei->addr+ei->size)
+ ;
-+ last = addr + size;
++ last = PAGE_ALIGN(addr) + size;
+ if (last > ei->addr + ei->size)
+ continue;
+ if (last > end)
@@ -29658,59 +25666,14 @@
+ 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;
++ end_pfn = find_max_pfn_with_active_regions();
+
-+ 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)
@@ -29720,43 +25683,10 @@
+ if (end_pfn > end_pfn_map)
+ end_pfn = end_pfn_map;
+
++ printk("end_pfn_map = %lu\n", 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.
+ */
@@ -29793,6 +25723,96 @@
+ }
+}
+
++/* Mark pages corresponding to given address range as nosave */
++static void __init
++e820_mark_nosave_range(unsigned long start, unsigned long end)
++{
++ unsigned long pfn, max_pfn;
++
++ if (start >= end)
++ return;
++
++ printk("Nosave address range: %016lx - %016lx\n", start, end);
++ max_pfn = end >> PAGE_SHIFT;
++ for (pfn = start >> PAGE_SHIFT; pfn < max_pfn; pfn++)
++ if (pfn_valid(pfn))
++ SetPageNosave(pfn_to_page(pfn));
++}
++
++/*
++ * Find the ranges of physical addresses that do not correspond to
++ * e820 RAM areas and mark the corresponding pages as nosave for software
++ * suspend and suspend to RAM.
++ *
++ * This function requires the e820 map to be sorted and without any
++ * overlapping entries and assumes the first e820 area to be RAM.
++ */
++void __init e820_mark_nosave_regions(void)
++{
++ int i;
++ unsigned long paddr;
++
++ paddr = round_down(e820.map[0].addr + e820.map[0].size, PAGE_SIZE);
++ for (i = 1; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++
++ if (paddr < ei->addr)
++ e820_mark_nosave_range(paddr,
++ round_up(ei->addr, PAGE_SIZE));
++
++ paddr = round_down(ei->addr + ei->size, PAGE_SIZE);
++ if (ei->type != E820_RAM)
++ e820_mark_nosave_range(round_up(ei->addr, PAGE_SIZE),
++ paddr);
++
++ if (paddr >= (end_pfn << PAGE_SHIFT))
++ break;
++ }
++}
++
++/* Walk the e820 map and register active regions within a node */
++void __init
++e820_register_active_regions(int nid, unsigned long start_pfn,
++ unsigned long end_pfn)
++{
++ int i;
++ unsigned long ei_startpfn, ei_endpfn;
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++ ei_startpfn = round_up(ei->addr, PAGE_SIZE) >> PAGE_SHIFT;
++ ei_endpfn = round_down(ei->addr + ei->size, PAGE_SIZE)
++ >> PAGE_SHIFT;
++
++ /* Skip map entries smaller than a page */
++ if (ei_startpfn >= ei_endpfn)
++ continue;
++
++ /* Check if end_pfn_map should be updated */
++ if (ei->type != E820_RAM && ei_endpfn > end_pfn_map)
++ end_pfn_map = ei_endpfn;
++
++ /* Skip if map is outside the node */
++ if (ei->type != E820_RAM ||
++ ei_endpfn <= start_pfn ||
++ ei_startpfn >= end_pfn)
++ continue;
++
++ /* Check for overlaps */
++ if (ei_startpfn < start_pfn)
++ ei_startpfn = start_pfn;
++ if (ei_endpfn > end_pfn)
++ ei_endpfn = end_pfn;
++
++ /* Obey end_user_pfn to save on memmap */
++ if (ei_startpfn >= end_user_pfn)
++ continue;
++ if (ei_endpfn > end_user_pfn)
++ ei_endpfn = end_user_pfn;
++
++ add_active_range(nid, ei_startpfn, ei_endpfn);
++ }
++}
++
+/*
+ * Add a memory region to the kernel e820 map.
+ */
@@ -30013,13 +26033,6 @@
+ * 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)
+{
@@ -30041,37 +26054,20 @@
+ 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);
+ return 0;
+}
+
-+#ifndef CONFIG_XEN
-+void __init setup_memory_region(void)
++void early_panic(char *msg)
+{
-+ char *who = "BIOS-e820";
++ early_printk(msg);
++ panic(msg);
++}
+
++void __init setup_memory_region(void)
++{
++#ifndef CONFIG_XEN
+ /*
+ * Try to copy the BIOS-supplied E820-map.
+ *
@@ -30079,30 +26075,11 @@
+ * 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);
-+ }
++ if (copy_e820_map(E820_MAP, E820_MAP_NR) < 0)
++ early_panic("Cannot find a valid memory map");
+ printk(KERN_INFO "BIOS-provided physical RAM map:\n");
-+ e820_print_map(who);
-+}
-+
++ e820_print_map("BIOS-e820");
+#else /* CONFIG_XEN */
-+
-+void __init setup_memory_region(void)
-+{
+ int rc;
+ struct xen_memory_map memmap;
+ /*
@@ -30127,21 +26104,22 @@
+ BUG_ON(rc);
+
+ sanitize_e820_map(map, (char *)&memmap.nr_entries);
-+
-+ BUG_ON(copy_e820_map(map, (char)memmap.nr_entries) < 0);
-+
++ if (copy_e820_map(map, (char)memmap.nr_entries) < 0)
++ early_panic("Cannot find a valid memory map");
+ printk(KERN_INFO "BIOS-provided physical RAM map:\n");
+ e820_print_map("Xen");
-+}
+#endif
++}
+
-+void __init parse_memopt(char *p, char **from)
++static int __init parse_memopt(char *p)
+{
+ int i;
+ unsigned long current_end;
+ unsigned long end;
+
-+ end_user_pfn = memparse(p, from);
++ if (!p)
++ return -EINVAL;
++ end_user_pfn = memparse(p, &p);
+ end_user_pfn >>= PAGE_SHIFT;
+
+ end = end_user_pfn<<PAGE_SHIFT;
@@ -30158,27 +26136,60 @@
+ else
+ add_memory_region(current_end, end - current_end, E820_RAM);
+ }
++ return 0;
+}
++early_param("mem", parse_memopt);
++
++static int userdef __initdata;
+
-+void __init parse_memmapopt(char *p, char **from)
++static int __init parse_memmap_opt(char *p)
+{
++ char *oldp;
+ unsigned long long start_at, mem_size;
+
-+ mem_size = memparse(p, from);
-+ p = *from;
++ if (!strcmp(p, "exactmap")) {
++#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.
++ */
++ e820_register_active_regions(0, 0, -1UL);
++ saved_max_pfn = e820_end_of_ram();
++ remove_all_active_ranges();
++#endif
++ end_pfn_map = 0;
++ e820.nr_map = 0;
++ userdef = 1;
++ return 0;
++ }
++
++ oldp = p;
++ mem_size = memparse(p, &p);
++ if (p == oldp)
++ return -EINVAL;
+ if (*p == '@') {
-+ start_at = memparse(p+1, from);
++ start_at = memparse(p+1, &p);
+ add_memory_region(start_at, mem_size, E820_RAM);
+ } else if (*p == '#') {
-+ start_at = memparse(p+1, from);
++ start_at = memparse(p+1, &p);
+ add_memory_region(start_at, mem_size, E820_ACPI);
+ } else if (*p == '$') {
-+ start_at = memparse(p+1, from);
++ start_at = memparse(p+1, &p);
+ add_memory_region(start_at, mem_size, E820_RESERVED);
+ } else {
+ end_user_pfn = (mem_size >> PAGE_SHIFT);
+ }
-+ p = *from;
++ return *p == '\0' ? 0 : -EINVAL;
++}
++early_param("memmap", parse_memmap_opt);
++
++void finish_e820_parsing(void)
++{
++ if (userdef) {
++ printk(KERN_INFO "user-defined physical RAM map:\n");
++ e820_print_map("user");
++ }
+}
+
+unsigned long pci_mem_start = 0xaeedbabe;
@@ -30241,163 +26252,21 @@
+ printk(KERN_INFO "Allocating PCI resources starting at %lx (gap: %lx:%lx)\n",
+ pci_mem_start, gapstart, gapsize);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/early_printk-xen.c linux-2.6.18-xen/arch/x86_64/kernel/early_printk-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/early_printk-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/early_printk-xen.c 2006-11-19 14:26:33.000000000 +0100
-@@ -0,0 +1,304 @@
-+
-+#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
-+
-+static int max_ypos = 25, max_xpos = 80;
-+static int current_ypos = 25, current_xpos = 0;
-+
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/early_printk.c b/arch/x86_64/kernel/early_printk.c
+--- a/arch/x86_64/kernel/early_printk.c 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/kernel/early_printk.c 2007-03-14 10:55:21.000000000 +0100
+@@ -20,6 +20,7 @@
+ static int max_ypos = 25, max_xpos = 80;
+ static int current_ypos = 25, current_xpos = 0;
+
+#ifndef CONFIG_XEN
-+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);
-+}
-+
+ static void early_vga_write(struct console *con, const char *str, unsigned n)
+ {
+ char c;
+@@ -149,6 +150,39 @@
+ outb(c & ~DLAB, early_serial_base + LCR);
+ }
+
+#else /* CONFIG_XEN */
+
+#undef SCREEN_INFO
@@ -30431,128 +26300,13 @@
+
+#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)
-+ && 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;
-+ 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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/entry-xen.S linux-2.6.18-xen/arch/x86_64/kernel/entry-xen.S
---- linux-2.6.18.3/arch/x86_64/kernel/entry-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/entry-xen.S 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,1213 @@
+ static struct console early_serial_console = {
+ .name = "earlyser",
+ .write = early_serial_write,
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/entry-xen.S b/arch/x86_64/kernel/entry-xen.S
+--- a/arch/x86_64/kernel/entry-xen.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/entry-xen.S 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,1160 @@
+/*
+ * linux/arch/x86_64/entry.S
+ *
@@ -30560,8 +26314,6 @@
+ * 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
@@ -30581,18 +26333,25 @@
+ * 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.
++ *
++ * Some macro usage:
++ * - CFI macros are used to generate dwarf2 unwind information for better
++ * backtraces. They don't change any code.
++ * - SAVE_ALL/RESTORE_ALL - Save/restore all registers
++ * - SAVE_ARGS/RESTORE_ARGS - Save/restore registers that C functions modify.
++ * There are unfortunately lots of special cases where some registers
++ * not touched. The macro is a big mess that should be cleaned up.
++ * - SAVE_REST/RESTORE_REST - Handle the registers not saved by SAVE_ARGS.
++ * Gives a full stack frame.
++ * - ENTRY/END Define functions in the symbol table.
++ * - FIXUP_TOP_OF_STACK/RESTORE_TOP_OF_STACK - Fix up the hardware stack
++ * frame that is otherwise undefined after a SYSCALL
++ * - TRACE_IRQ_* - Trace hard interrupt state for lock debugging.
++ * - errorentry/paranoidentry/zeroentry - Define exception entry points.
+ */
+
-+#define ASSEMBLY 1
-+#ifdef CONFIG_DEBUG_INFO
-+#undef CONFIG_DEBUG_INFO
-+#endif
+#include <linux/linkage.h>
+#include <asm/segment.h>
-+#include <asm/smp.h>
+#include <asm/cache.h>
+#include <asm/errno.h>
+#include <asm/dwarf2.h>
@@ -30611,7 +26370,7 @@
+#include "irq_vectors.h"
+
+#include "xen_entry.S"
-+
++
+ .code64
+
+#ifndef CONFIG_PREEMPT
@@ -30629,7 +26388,7 @@
+.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
@@ -30677,6 +26436,7 @@
+ .macro CFI_DEFAULT_STACK start=1
+ .if \start
+ CFI_STARTPROC simple
++ CFI_SIGNAL_FRAME
+ CFI_DEF_CFA rsp,SS+8
+ .else
+ CFI_DEF_CFA_OFFSET SS+8
@@ -30703,13 +26463,13 @@
+ /*CFI_REL_OFFSET ss,SS*/
+ .endm
+
-+ /*
-+ * Must be consistent with the definition in arch-x86_64.h:
-+ * struct iret_context {
-+ * u64 rax, r11, rcx, flags, rip, cs, rflags, rsp, ss;
-+ * };
-+ * #define VGCF_IN_SYSCALL (1<<8)
-+ */
++ /*
++ * Must be consistent with the definition in arch-x86_64.h:
++ * struct iret_context {
++ * u64 rax, r11, rcx, flags, rip, cs, rflags, rsp, ss;
++ * };
++ * #define VGCF_IN_SYSCALL (1<<8)
++ */
+ .macro HYPERVISOR_IRET flag
+ testb $3,1*8(%rsp)
+ jnz 2f
@@ -30730,11 +26490,11 @@
+ jmp hypercall_page + (__HYPERVISOR_iret * 32)
+ .endm
+
-+ .macro SWITCH_TO_KERNEL ssoff,adjust=0
++ .macro SWITCH_TO_KERNEL ssoff,adjust=0
+ jc 1f
+ orb $1,\ssoff-\adjust+4(%rsp)
+1:
-+ .endm
++ .endm
+
+/*
+ * A newly forked process directly context switches into this.
@@ -30742,9 +26502,13 @@
+/* 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|_TIF_SECCOMP),threadinfo_flags(%rcx)
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT),threadinfo_flags(%rcx)
+ jnz rff_trace
+rff_action:
+ RESTORE_REST
@@ -30795,6 +26559,7 @@
+
+ENTRY(system_call)
+ CFI_STARTPROC simple
++ CFI_SIGNAL_FRAME
+ CFI_DEF_CFA rsp,PDA_STACKOFFSET
+ CFI_REGISTER rip,rcx
+ /*CFI_REGISTER rflags,r11*/
@@ -30804,10 +26569,9 @@
+ * No need to follow this irqs off/on section - it's straight
+ * and short:
+ */
-+ XEN_UNBLOCK_EVENTS(%r11)
++ 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
@@ -30818,36 +26582,35 @@
+ * 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)
++ XEN_BLOCK_EVENTS(%rsi)
+ TRACE_IRQS_OFF
+ movl threadinfo_flags(%rcx),%edx
+ andl %edi,%edx
-+ CFI_REMEMBER_STATE
+ jnz sysret_careful
++ CFI_REMEMBER_STATE
+ /*
+ * sysretq will re-enable interrupts:
+ */
+ TRACE_IRQS_ON
-+ XEN_UNBLOCK_EVENTS(%rsi)
++ XEN_UNBLOCK_EVENTS(%rsi)
+ CFI_REGISTER rip,rcx
+ RESTORE_ARGS 0,8,0
+ /*CFI_REGISTER rflags,r11*/
-+ HYPERVISOR_IRET VGCF_IN_SYSCALL
++ HYPERVISOR_IRET VGCF_IN_SYSCALL
+
++ CFI_RESTORE_STATE
+ /* 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)
++ XEN_UNBLOCK_EVENTS(%rsi)
+ pushq %rdi
+ CFI_ADJUST_CFA_OFFSET 8
+ call schedule
@@ -30858,8 +26621,7 @@
+ /* Handle a signal */
+sysret_signal:
+ TRACE_IRQS_ON
-+/* sti */
-+ XEN_UNBLOCK_EVENTS(%rsi)
++ XEN_UNBLOCK_EVENTS(%rsi)
+ testl $(_TIF_SIGPENDING|_TIF_NOTIFY_RESUME|_TIF_SINGLESTEP),%edx
+ jz 1f
+
@@ -30882,7 +26644,6 @@
+
+ /* Do syscall tracing */
+tracesys:
-+ CFI_RESTORE_STATE
+ SAVE_REST
+ movq $-ENOSYS,RAX(%rsp)
+ FIXUP_TOP_OF_STACK %rdi
@@ -30891,44 +26652,28 @@
+ LOAD_ARGS ARGOFFSET /* reload args from stack in case ptrace changed it */
+ RESTORE_REST
+ cmpq $__NR_syscall_max,%rax
++ movq $-ENOSYS,%rcx
++ cmova %rcx,%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)
++ */
++ .globl int_ret_from_sys_call
++int_ret_from_sys_call:
++ 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:
++ 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:
@@ -30946,8 +26691,7 @@
+ bt $TIF_NEED_RESCHED,%edx
+ jnc int_very_careful
+ TRACE_IRQS_ON
-+/* sti */
-+ XEN_UNBLOCK_EVENTS(%rsi)
++ XEN_UNBLOCK_EVENTS(%rsi)
+ pushq %rdi
+ CFI_ADJUST_CFA_OFFSET 8
+ call schedule
@@ -30960,8 +26704,7 @@
+ /* handle signals and tracing -- both require a full stack frame */
+int_very_careful:
+ TRACE_IRQS_ON
-+/* sti */
-+ XEN_UNBLOCK_EVENTS(%rsi)
++ XEN_UNBLOCK_EVENTS(%rsi)
+ SAVE_REST
+ /* Check for syscall exit trace */
+ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP),%edx
@@ -30973,8 +26716,6 @@
+ 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:
@@ -30990,7 +26731,7 @@
+ TRACE_IRQS_OFF
+ jmp int_with_check
+ CFI_ENDPROC
-+END(int_ret_from_sys_call)
++END(system_call)
+
+/*
+ * Certain special system calls that need to save a complete full stack frame.
@@ -31072,6 +26813,7 @@
+ */
+ .macro _frame ref
+ CFI_STARTPROC simple
++ CFI_SIGNAL_FRAME
+ CFI_DEF_CFA rsp,SS+8-\ref
+ /*CFI_REL_OFFSET ss,SS-\ref*/
+ CFI_REL_OFFSET rsp,RSP-\ref
@@ -31115,30 +26857,28 @@
+ 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)
++ XEN_BLOCK_EVENTS(%rsi)
+ TRACE_IRQS_OFF
-+/* cli */
+ 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)
++ 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)
++ XEN_BLOCK_EVENTS(%rsi)
+ TRACE_IRQS_OFF
+ movl $_TIF_NEED_RESCHED,%edi
+ GET_THREAD_INFO(%rcx)
@@ -31147,8 +26887,7 @@
+#ifdef CONFIG_PREEMPT
+ /* Returning to kernel space. Check if we need preemption */
+ /* rcx: threadinfo. interrupts off. */
-+ .p2align
-+retint_kernel:
++ENTRY(retint_kernel)
+ cmpl $0,threadinfo_preempt_count(%rcx)
+ jnz retint_restore_args
+ bt $TIF_NEED_RESCHED,threadinfo_flags(%rcx)
@@ -31208,7 +26947,6 @@
+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)
@@ -31220,7 +26958,6 @@
+ENTRY(spurious_interrupt)
+ apicinterrupt SPURIOUS_APIC_VECTOR,smp_spurious_interrupt
+END(spurious_interrupt)
-+#endif
+#endif /* !CONFIG_XEN */
+
+/*
@@ -31228,9 +26965,9 @@
+ */
+ .macro zeroentry sym
+ INTR_FRAME
-+ movq (%rsp),%rcx
-+ movq 8(%rsp),%r11
-+ addq $0x10,%rsp /* skip rcx and r11 */
++ movq (%rsp),%rcx
++ movq 8(%rsp),%r11
++ addq $0x10,%rsp /* skip rcx and r11 */
+ pushq $0 /* push error code/oldrax */
+ CFI_ADJUST_CFA_OFFSET 8
+ pushq %rax /* push real oldrax to the rdi slot */
@@ -31242,9 +26979,9 @@
+
+ .macro errorentry sym
+ XCPT_FRAME
-+ movq (%rsp),%rcx
-+ movq 8(%rsp),%r11
-+ addq $0x10,%rsp /* rsp points to the error code */
++ movq (%rsp),%rcx
++ movq 8(%rsp),%r11
++ addq $0x10,%rsp /* rsp points to the error code */
+ pushq %rax
+ CFI_ADJUST_CFA_OFFSET 8
+ leaq \sym(%rip),%rax
@@ -31256,9 +26993,9 @@
+ /* error code is on the stack already */
+ /* handle NMI like exceptions that can happen everywhere */
+ .macro paranoidentry sym, ist=0
-+ movq (%rsp),%rcx
-+ movq 8(%rsp),%r11
-+ addq $0x10,%rsp /* skip rcx and r11 */
++ movq (%rsp),%rcx
++ movq 8(%rsp),%r11
++ addq $0x10,%rsp /* skip rcx and r11 */
+ SAVE_ALL
+ cld
+#if 0 /* not XEN */
@@ -31284,17 +27021,18 @@
+ .if \ist
+ addq $EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp)
+ .endif
-+/* cli */
-+ TRACE_IRQS_OFF
+ XEN_BLOCK_EVENTS(%rsi)
++ .if \irqtrace
++ TRACE_IRQS_OFF
++ .endif
+ .endm
+#endif
-+
++
+/*
+ * Exception entry point. This expects an error code/orig_rax on the stack
+ * and the exception handler in %rax.
+ */
-+ENTRY(error_entry)
++KPROBE_ENTRY(error_entry)
+ _frame RDI
+ /* rdi slot contains rax, oldrax contains error code */
+ cld
@@ -31329,20 +27067,19 @@
+ CFI_REL_OFFSET r14,R14
+ movq %r15,(%rsp)
+ CFI_REL_OFFSET r15,R15
-+#if 0
++#if 0
+ cmpl $__KERNEL_CS,CS(%rsp)
+ je error_kernelspace
-+#endif
++#endif
+error_call_handler:
-+ movq %rdi, RDI(%rsp)
++ movq %rdi,RDI(%rsp)
+ movq %rsp,%rdi
-+ movq ORIG_RAX(%rsp),%rsi # get error code
++ movq ORIG_RAX(%rsp),%rsi /* get error code */
+ movq $-1,ORIG_RAX(%rsp)
+ call *%rax
+error_exit:
+ RESTORE_REST
-+/* cli */
-+ XEN_BLOCK_EVENTS(%rsi)
++ XEN_BLOCK_EVENTS(%rsi)
+ TRACE_IRQS_OFF
+ GET_THREAD_INFO(%rcx)
+ testb $3,CS-ARGOFFSET(%rsp)
@@ -31354,13 +27091,13 @@
+ jmp retint_restore_args
+
+error_kernelspace:
-+ /*
-+ * 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).
-+ *
-+ */
++ /*
++ * 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).
++ *
++ */
+#if 0
+ incl %ebx
+ /* There are two places in the kernel that can potentially fault with
@@ -31377,9 +27114,9 @@
+ cmpq $gs_change,RIP(%rsp)
+ je error_swapgs
+ jmp error_sti
-+#endif
-+END(error_entry)
-+
++#endif
++KPROBE_END(error_entry)
++
+ENTRY(hypervisor_callback)
+ zeroentry do_hypervisor_callback
+
@@ -31410,19 +27147,17 @@
+ decl %gs:pda_irqcount
+ jmp error_exit
+
-+#ifdef CONFIG_X86_LOCAL_APIC
+KPROBE_ENTRY(nmi)
+ zeroentry do_nmi_callback
-+ENTRY(do_nmi_callback)
-+ addq $8, %rsp
-+ call do_nmi
-+ orl $NMI_MASK,EFLAGS(%rsp)
-+ RESTORE_REST
-+ XEN_BLOCK_EVENTS(%rsi)
-+ GET_THREAD_INFO(%rcx)
-+ jmp retint_restore_args
++ENTRY(xen_do_nmi_callback)
++ addq $8, %rsp
++ call do_nmi
++ orl $NMI_MASK,EFLAGS(%rsp)
++ RESTORE_REST
++ XEN_BLOCK_EVENTS(%rsi)
++ GET_THREAD_INFO(%rcx)
++ jmp retint_restore_args
+ .previous .text
-+#endif
+
+ ALIGN
+restore_all_enable_events:
@@ -31434,11 +27169,11 @@
+ XEN_PUT_VCPU_INFO(%rsi)
+ RESTORE_ARGS 0,8,0
+ HYPERVISOR_IRET 0
-+
++
+14: XEN_LOCKED_BLOCK_EVENTS(%rsi)
+ XEN_PUT_VCPU_INFO(%rsi)
+ SAVE_REST
-+ movq %rsp,%rdi # set the argument again
++ movq %rsp,%rdi # set the argument again
+ jmp 11b
+ecrit: /**** END OF CRITICAL REGION ****/
+# At this point, unlike on x86-32, we don't do the fixup to simplify the
@@ -31571,7 +27306,7 @@
+ * do_sys_execve asm fallback arguments:
+ * rdi: name, rsi: argv, rdx: envp, fake frame on the stack
+ */
-+ENTRY(execve)
++ENTRY(kernel_execve)
+ CFI_STARTPROC
+ FAKE_STACK_FRAME $0
+ SAVE_ALL
@@ -31580,17 +27315,16 @@
+ RESTORE_REST
+ testq %rax,%rax
+ jne 1f
-+ jmp int_ret_from_sys_call
-+1: RESTORE_ARGS
++ jmp int_ret_from_sys_call
++1: RESTORE_ARGS
+ UNFAKE_STACK_FRAME
+ ret
+ CFI_ENDPROC
-+ENDPROC(execve)
++ENDPROC(kernel_execve)
+
+KPROBE_ENTRY(page_fault)
+ errorentry do_page_fault
-+END(page_fault)
-+ .previous .text
++KPROBE_END(page_fault)
+
+ENTRY(coprocessor_error)
+ zeroentry do_coprocessor_error
@@ -31612,7 +27346,7 @@
+ zeroentry do_debug
+/* jmp paranoid_exit */
+ CFI_ENDPROC
-+END(debug)
++KPROBE_END(debug)
+ .previous .text
+
+#if 0
@@ -31628,7 +27362,7 @@
+ jmp paranoid_exit1
+ CFI_ENDPROC
+#endif
-+END(nmi)
++KPROBE_END(nmi)
+ .previous .text
+#endif
+
@@ -31639,8 +27373,7 @@
+ zeroentry do_int3
+/* jmp paranoid_exit */
+ CFI_ENDPROC
-+END(int3)
-+ .previous .text
++KPROBE_END(int3)
+
+ENTRY(overflow)
+ zeroentry do_overflow
@@ -31679,6 +27412,7 @@
+ENTRY(segment_not_present)
+ errorentry do_segment_not_present
+END(segment_not_present)
++
+ /* runs on exception stack */
+ENTRY(stack_segment)
+ XCPT_FRAME
@@ -31688,8 +27422,7 @@
+
+KPROBE_ENTRY(general_protection)
+ errorentry do_general_protection
-+END(general_protection)
-+ .previous .text
++KPROBE_END(general_protection)
+
+ENTRY(alignment_check)
+ errorentry do_alignment_check
@@ -31715,6 +27448,7 @@
+END(machine_check)
+#endif
+
++/* Call softirq on interrupt stack. Interrupts are off. */
+ENTRY(call_softirq)
+ CFI_STARTPROC
+ push %rbp
@@ -31733,43 +27467,10 @@
+ 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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/genapic_xen.c linux-2.6.18-xen/arch/x86_64/kernel/genapic_xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/genapic_xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/genapic_xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,176 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/genapic_xen.c b/arch/x86_64/kernel/genapic_xen.c
+--- a/arch/x86_64/kernel/genapic_xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/genapic_xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,190 @@
+/*
+ * Copyright 2004 James Cleverdon, IBM.
+ * Subject to the GNU Public License, v.2
@@ -31843,6 +27544,20 @@
+ return cpu_online_map;
+}
+
++static cpumask_t xen_vector_allocation_domain(int cpu)
++{
++ /* Careful. Some cpus do not strictly honor the set of cpus
++ * specified in the interrupt destination when using lowest
++ * priority interrupt delivery mode.
++ *
++ * In particular there was a hyperthreading cpu observed to
++ * deliver interrupts to the wrong hyperthread when only one
++ * hyperthread was specified in the interrupt desitination.
++ */
++ cpumask_t domain = { { [0] = APIC_ALL_CPUS, } };
++ return domain;
++}
++
+/*
+ * Set up the logical destination ID.
+ * Do nothing, not called now.
@@ -31934,8 +27649,8 @@
+ .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,
++ .vector_allocation_domain = xen_vector_allocation_domain,
+#ifdef CONFIG_XEN_PRIVILEGED_GUEST
+ .apic_id_registered = xen_apic_id_registered,
+#endif
@@ -31946,9 +27661,9 @@
+ .cpu_mask_to_apicid = xen_cpu_mask_to_apicid,
+ .phys_pkg_id = phys_pkg_id,
+};
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/genapic-xen.c linux-2.6.18-xen/arch/x86_64/kernel/genapic-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/genapic-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/genapic-xen.c 2006-11-19 14:26:33.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/genapic-xen.c b/arch/x86_64/kernel/genapic-xen.c
+--- a/arch/x86_64/kernel/genapic-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/genapic-xen.c 2007-03-14 10:55:21.000000000 +0100
@@ -0,0 +1,143 @@
+/*
+ * Copyright 2004 James Cleverdon, IBM.
@@ -32093,116 +27808,86 @@
+ xen_send_IPI_shortcut(APIC_DEST_SELF, vector, APIC_DEST_PHYSICAL);
+#endif
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/head64-xen.c linux-2.6.18-xen/arch/x86_64/kernel/head64-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/head64-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/head64-xen.c 2006-11-19 14:26:33.000000000 +0100
-@@ -0,0 +1,161 @@
-+/*
-+ * linux/arch/x86_64/kernel/head64.c -- prepare to run common code
-+ *
-+ * Copyright (C) 2000 Andrea Arcangeli <andrea at suse.de> SuSE
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/head64.c b/arch/x86_64/kernel/head64.c
+--- a/arch/x86_64/kernel/head64.c 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/kernel/head64.c 2007-03-14 10:55:21.000000000 +0100
+@@ -2,6 +2,9 @@
+ * 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/init.h>
+@@ -10,6 +13,7 @@
+ #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>
-+
+
+ #include <asm/processor.h>
+ #include <asm/proto.h>
+@@ -20,13 +24,19 @@
+ #include <asm/pgtable.h>
+ #include <asm/sections.h>
+
++#ifdef CONFIG_XEN
+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)
-+{
+ /* Don't add a printk in there. printk relies on the PDA which is not initialized
+ yet. */
+#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);
+ 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
+@@ -38,6 +48,7 @@
+
+ static void __init copy_bootdata(char *real_mode_data)
+ {
++#ifndef CONFIG_XEN
+ int new_data;
+ char * command_line;
+
+@@ -51,26 +62,67 @@
+ }
+ 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;
-+}
-+
+ }
+
++#ifdef CONFIG_XEN
+#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);
++#endif
+
-+void __init x86_64_start_kernel(char * real_mode_data)
-+{
+ void __init x86_64_start_kernel(char * real_mode_data)
+ {
++#ifdef CONFIG_XEN
+ struct xen_machphys_mapping mapping;
+ unsigned long machine_to_phys_nr_ents;
-+ char *s;
-+ int i;
-+
++#endif
+ int i;
+
++#ifdef CONFIG_XEN
+ xen_start_info = (struct start_info *)real_mode_data;
+ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
+ phys_to_machine_mapping =
@@ -32220,48 +27905,31 @@
+ 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;
++#else
+ /* clear bss before set_intr_gate with early_idt_handler */
+ clear_bss();
+
+ for (i = 0; i < IDT_ENTRIES; i++)
+ set_intr_gate(i, early_idt_handler);
+ asm volatile("lidt %0" :: "m" (idt_descr));
+-
++#endif /* CONFIG_XEN */
+ early_printk("Kernel alive\n");
+
++#ifndef CONFIG_XEN
+ /*
+ * switch to init_level4_pgt from boot_level4_pgt
+ */
+ memcpy(init_level4_pgt, boot_level4_pgt, PTRS_PER_PGD*sizeof(pgd_t));
+ asm volatile("movq %0,%%cr3" :: "r" (__pa_symbol(&init_level4_pgt)));
+#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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/head-xen.S linux-2.6.18-xen/arch/x86_64/kernel/head-xen.S
---- linux-2.6.18.3/arch/x86_64/kernel/head-xen.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/head-xen.S 2006-11-19 14:26:33.000000000 +0100
-@@ -0,0 +1,191 @@
+
+ for (i = 0; i < NR_CPUS; i++)
+ cpu_pda(i) = &boot_cpu_pda[i];
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/head-xen.S b/arch/x86_64/kernel/head-xen.S
+--- a/arch/x86_64/kernel/head-xen.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/head-xen.S 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,189 @@
+/*
+ * linux/arch/x86_64/kernel/head.S -- start in 32bit and switch to 64bit
+ *
@@ -32270,8 +27938,6 @@
+ * 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
+ */
@@ -32374,22 +28040,22 @@
+
+/* 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 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 */
++ .quad 0,0,0 /* three TLS descriptors */
++ .quad 0x0000f40000000000 /* node/CPU stored in limit */
++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 */
+
@@ -32453,9 +28119,9 @@
+ ELFNOTE(Xen, XEN_ELFNOTE_HYPERCALL_PAGE, .quad, hypercall_page)
+ 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")
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/init_task.c linux-2.6.18-xen/arch/x86_64/kernel/init_task.c
---- linux-2.6.18.3/arch/x86_64/kernel/init_task.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/x86_64/kernel/init_task.c 2006-11-19 14:26:33.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/init_task.c b/arch/x86_64/kernel/init_task.c
+--- a/arch/x86_64/kernel/init_task.c 2007-03-12 21:58:06.000000000 +0100
++++ b/arch/x86_64/kernel/init_task.c 2007-03-14 10:55:21.000000000 +0100
@@ -37,6 +37,8 @@
struct task_struct init_task = INIT_TASK(init_task);
@@ -32473,10 +28139,10 @@
/* Copies of the original ist values from the tss are only accessed during
* debugging, no special alignment required.
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/io_apic-xen.c linux-2.6.18-xen/arch/x86_64/kernel/io_apic-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/io_apic-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/io_apic-xen.c 2006-11-19 14:26:33.000000000 +0100
-@@ -0,0 +1,2263 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/io_apic-xen.c b/arch/x86_64/kernel/io_apic-xen.c
+--- a/arch/x86_64/kernel/io_apic-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/io_apic-xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,2221 @@
+/*
+ * Intel IO-APIC support for multi-Pentium hosts.
+ *
@@ -32505,9 +28171,12 @@
+#include <linux/delay.h>
+#include <linux/sched.h>
+#include <linux/smp_lock.h>
++#include <linux/pci.h>
+#include <linux/mc146818rtc.h>
+#include <linux/acpi.h>
+#include <linux/sysdev.h>
++#include <linux/msi.h>
++#include <linux/htirq.h>
+#ifdef CONFIG_ACPI
+#include <acpi/acpi_bus.h>
+#endif
@@ -32520,6 +28189,31 @@
+#include <asm/acpi.h>
+#include <asm/dma.h>
+#include <asm/nmi.h>
++#include <asm/msidef.h>
++#include <asm/hypertransport.h>
++
++DEFINE_PER_CPU(vector_irq_t, vector_irq) = {
++ [0 ... FIRST_EXTERNAL_VECTOR - 1] = -1,
++ [FIRST_EXTERNAL_VECTOR + 0] = 0,
++ [FIRST_EXTERNAL_VECTOR + 1] = 1,
++ [FIRST_EXTERNAL_VECTOR + 2] = 2,
++ [FIRST_EXTERNAL_VECTOR + 3] = 3,
++ [FIRST_EXTERNAL_VECTOR + 4] = 4,
++ [FIRST_EXTERNAL_VECTOR + 5] = 5,
++ [FIRST_EXTERNAL_VECTOR + 6] = 6,
++ [FIRST_EXTERNAL_VECTOR + 7] = 7,
++ [FIRST_EXTERNAL_VECTOR + 8] = 8,
++ [FIRST_EXTERNAL_VECTOR + 9] = 9,
++ [FIRST_EXTERNAL_VECTOR + 10] = 10,
++ [FIRST_EXTERNAL_VECTOR + 11] = 11,
++ [FIRST_EXTERNAL_VECTOR + 12] = 12,
++ [FIRST_EXTERNAL_VECTOR + 13] = 13,
++ [FIRST_EXTERNAL_VECTOR + 14] = 14,
++ [FIRST_EXTERNAL_VECTOR + 15] = 15,
++ [FIRST_EXTERNAL_VECTOR + 16 ... NR_VECTORS - 1] = -1
++};
++
++static int assign_irq_vector(int irq, cpumask_t mask, cpumask_t *result);
+
+#define __apicdebuginit __init
+
@@ -32527,17 +28221,17 @@
+
+static int no_timer_check;
+
-+int disable_timer_pin_1 __initdata;
++static int disable_timer_pin_1 __initdata;
+
-+#ifndef CONFIG_XEN
-+int timer_over_8254 __initdata = 0;
++int timer_over_8254 __initdata = 1;
+
++#ifndef CONFIG_XEN
+/* 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);
++DEFINE_SPINLOCK(vector_lock);
+
+/*
+ * # of IRQ routing registers
@@ -32562,13 +28256,51 @@
+ 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
++struct io_apic {
++ unsigned int index;
++ unsigned int unused[3];
++ unsigned int data;
++};
++
++static __attribute_const__ struct io_apic __iomem *io_apic_base(int idx)
++{
++ return (void __iomem *) __fix_to_virt(FIX_IO_APIC_BASE_0 + idx)
++ + (mp_ioapics[idx].mpc_apicaddr & ~PAGE_MASK);
++}
++
++static inline unsigned int io_apic_read(unsigned int apic, unsigned int reg)
++{
++ struct io_apic __iomem *io_apic = io_apic_base(apic);
++ writel(reg, &io_apic->index);
++ return readl(&io_apic->data);
++}
++
++static inline void io_apic_write(unsigned int apic, unsigned int reg, unsigned int value)
++{
++ struct io_apic __iomem *io_apic = io_apic_base(apic);
++ writel(reg, &io_apic->index);
++ writel(value, &io_apic->data);
++}
++
++/*
++ * Re-write a value: to be used for read-modify-write
++ * cycles where the read already set up the index register.
++ */
++static inline void io_apic_modify(unsigned int apic, unsigned int value)
++{
++ struct io_apic __iomem *io_apic = io_apic_base(apic);
++ writel(value, &io_apic->data);
++}
++
++/*
++ * Synchronize the IO-APIC and the CPU by doing
++ * a dummy read from the IO-APIC
++ */
++static inline void io_apic_sync(unsigned int apic)
++{
++ struct io_apic __iomem *io_apic = io_apic_base(apic);
++ readl(&io_apic->data);
++}
+
+#ifdef CONFIG_XEN
+
@@ -32605,19 +28337,122 @@
+ 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 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)
++#endif
++
++#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; \
++}
++#endif /* !CONFIG_XEN */
++
++union entry_union {
++ struct { u32 w1, w2; };
++ struct IO_APIC_route_entry entry;
++};
++
++static struct IO_APIC_route_entry ioapic_read_entry(int apic, int pin)
++{
++ union entry_union eu;
++ unsigned long flags;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ eu.w1 = io_apic_read(apic, 0x10 + 2 * pin);
++ eu.w2 = io_apic_read(apic, 0x11 + 2 * pin);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++ return eu.entry;
++}
++
++/*
++ * When we write a new IO APIC routing entry, we need to write the high
++ * word first! If the mask bit in the low word is clear, we will enable
++ * the interrupt, and we need to make sure the entry is fully populated
++ * before that happens.
++ */
++static void
++__ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e)
++{
++ union entry_union eu;
++ eu.entry = e;
++ io_apic_write(apic, 0x11 + 2*pin, eu.w2);
++ io_apic_write(apic, 0x10 + 2*pin, eu.w1);
++}
++
++static void ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e)
++{
++ unsigned long flags;
++ spin_lock_irqsave(&ioapic_lock, flags);
++ __ioapic_write_entry(apic, pin, e);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
++
++/*
++ * When we mask an IO APIC routing entry, we need to write the low
++ * word first, in order to set the mask bit before we change the
++ * high bits!
++ */
++#ifndef CONFIG_XEN
++static void ioapic_mask_entry(int apic, int pin)
++{
++ unsigned long flags;
++ union entry_union eu = { .entry.mask = 1 };
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ io_apic_write(apic, 0x10 + 2*pin, eu.w1);
++ io_apic_write(apic, 0x11 + 2*pin, eu.w2);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
+
-+#define clear_IO_APIC() ((void)0)
++#ifdef CONFIG_SMP
++static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, u8 vector)
++{
++ int apic, pin;
++ struct irq_pin_list *entry = irq_2_pin + irq;
+
-+#else
++ BUG_ON(irq >= NR_IRQS);
++ for (;;) {
++ unsigned int reg;
++ apic = entry->apic;
++ pin = entry->pin;
++ if (pin == -1)
++ break;
++ io_apic_write(apic, 0x11 + pin*2, dest);
++ reg = io_apic_read(apic, 0x10 + pin*2);
++ reg &= ~0x000000ff;
++ reg |= vector;
++ io_apic_modify(apic, reg);
++ if (!entry->next)
++ break;
++ entry = irq_2_pin + entry->next;
++ }
++}
+
-+#ifdef CONFIG_SMP
+static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t mask)
+{
+ unsigned long flags;
+ unsigned int dest;
+ cpumask_t tmp;
++ int vector;
+
+ cpus_and(tmp, mask, cpu_online_map);
+ if (cpus_empty(tmp))
@@ -32625,7 +28460,11 @@
+
+ cpus_and(mask, tmp, CPU_MASK_ALL);
+
-+ dest = cpu_mask_to_apicid(mask);
++ vector = assign_irq_vector(irq, mask, &tmp);
++ if (vector < 0)
++ return;
++
++ dest = cpu_mask_to_apicid(tmp);
+
+ /*
+ * Only the high 8 bits are valid.
@@ -32633,12 +28472,11 @@
+ dest = SET_APIC_LOGICAL_ID(dest);
+
+ spin_lock_irqsave(&ioapic_lock, flags);
-+ __DO_ACTION(1, = dest, )
-+ set_irq_info(irq, mask);
++ __target_IO_APIC_irq(irq, dest, vector);
++ set_native_irq_info(irq, mask);
+ spin_unlock_irqrestore(&ioapic_lock, flags);
+}
+#endif
-+
+#endif /* !CONFIG_XEN */
+
+/*
@@ -32666,27 +28504,6 @@
+}
+
+#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) \
+ \
@@ -32719,24 +28536,15 @@
+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);
++ entry = ioapic_read_entry(apic, pin);
+ 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);
++ ioapic_mask_entry(apic, pin);
+}
+
+static void clear_IO_APIC (void)
@@ -32749,17 +28557,6 @@
+}
+
+#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;
+
@@ -32768,20 +28565,18 @@
+static int __init disable_ioapic_setup(char *str)
+{
+ skip_ioapic_setup = 1;
-+ return 1;
++ return 0;
+}
++early_param("noapic", disable_ioapic_setup);
+
-+static int __init enable_ioapic_setup(char *str)
++/* Actually the next is obsolete, but keep it for paranoid reasons -AK */
++static int __init disable_timer_pin_setup(char *arg)
+{
-+ ioapic_force = 1;
-+ skip_ioapic_setup = 0;
++ disable_timer_pin_1 = 1;
+ return 1;
+}
++__setup("disable_timer_pin_1", disable_timer_pin_setup);
+
-+__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;
@@ -32795,139 +28590,7 @@
+
+__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.
@@ -32957,9 +28620,7 @@
+ 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) &&
++ if (test_bit(lbus, mp_bus_not_pci) &&
+ (mp_irqs[i].mpc_irqtype == type) &&
+ (mp_irqs[i].mpc_srcbusirq == irq))
+
@@ -32975,9 +28636,7 @@
+ 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) &&
++ if (test_bit(lbus, mp_bus_not_pci) &&
+ (mp_irqs[i].mpc_irqtype == type) &&
+ (mp_irqs[i].mpc_srcbusirq == irq))
+ break;
@@ -33018,7 +28677,7 @@
+ mp_irqs[i].mpc_dstapic == MP_APIC_ALL)
+ break;
+
-+ if ((mp_bus_id_to_type[lbus] == MP_BUS_PCI) &&
++ if (!test_bit(lbus, mp_bus_not_pci) &&
+ !mp_irqs[i].mpc_irqtype &&
+ (bus == lbus) &&
+ (slot == ((mp_irqs[i].mpc_srcbusirq >> 2) & 0x1f))) {
@@ -33041,27 +28700,6 @@
+ 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. */
+
@@ -33074,12 +28712,6 @@
+#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;
@@ -33091,38 +28723,11 @@
+ 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;
-+ }
-+ }
++ if (test_bit(bus, mp_bus_not_pci))
++ polarity = default_ISA_polarity(idx);
++ else
++ polarity = default_PCI_polarity(idx);
+ break;
-+ }
+ case 1: /* high active */
+ {
+ polarity = 0;
@@ -33160,38 +28765,11 @@
+ 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;
-+ }
-+ }
++ if (test_bit(bus, mp_bus_not_pci))
++ trigger = default_ISA_trigger(idx);
++ else
++ trigger = default_PCI_trigger(idx);
+ break;
-+ }
+ case 1: /* edge */
+ {
+ trigger = 0;
@@ -33228,64 +28806,6 @@
+ 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;
@@ -33297,49 +28817,16 @@
+ 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);
-+ }
-+ }
++ if (test_bit(bus, mp_bus_not_pci)) {
++ irq = mp_irqs[idx].mpc_srcbusirq;
++ } else {
++ /*
++ * PCI IRQs are mapped in order
++ */
++ i = irq = 0;
++ while (i < apic)
++ irq += nr_ioapic_registers[i++];
++ irq += pin;
+ }
+ BUG_ON(irq >= NR_IRQS);
+ return irq;
@@ -33363,37 +28850,106 @@
+}
+
+/* irq_vectors is indexed by the sum of all RTEs in all I/O APICs. */
-+u8 irq_vector[NR_IRQ_VECTORS] __read_mostly;
++static u8 irq_vector[NR_IRQ_VECTORS] __read_mostly;
+
-+int assign_irq_vector(int irq)
++static cpumask_t irq_domain[NR_IRQ_VECTORS] __read_mostly = {
++ [0] = CPU_MASK_ALL,
++ [1] = CPU_MASK_ALL,
++ [2] = CPU_MASK_ALL,
++ [3] = CPU_MASK_ALL,
++ [4] = CPU_MASK_ALL,
++ [5] = CPU_MASK_ALL,
++ [6] = CPU_MASK_ALL,
++ [7] = CPU_MASK_ALL,
++ [8] = CPU_MASK_ALL,
++ [9] = CPU_MASK_ALL,
++ [10] = CPU_MASK_ALL,
++ [11] = CPU_MASK_ALL,
++ [12] = CPU_MASK_ALL,
++ [13] = CPU_MASK_ALL,
++ [14] = CPU_MASK_ALL,
++ [15] = CPU_MASK_ALL,
++};
++
++
++static int __assign_irq_vector(int irq, cpumask_t mask, cpumask_t *result)
+{
+ struct physdev_irq irq_op;
-+ unsigned long flags;
-+
-+ BUG_ON(irq != AUTO_ASSIGN && (unsigned)irq >= NR_IRQ_VECTORS);
++ int vector;
+
-+ spin_lock_irqsave(&vector_lock, flags);
++ BUG_ON((unsigned)irq >= NR_IRQ_VECTORS);
+
-+ if (irq != AUTO_ASSIGN && IO_APIC_VECTOR(irq) > 0) {
-+ spin_unlock_irqrestore(&vector_lock, flags);
-+ return IO_APIC_VECTOR(irq);
++ if (irq_vector[irq] > 0) {
++ cpus_and(*result, irq_domain[irq], mask);
++ return irq_vector[irq];
+ }
+ irq_op.irq = irq;
+ if (HYPERVISOR_physdev_op(PHYSDEVOP_alloc_irq_vector, &irq_op))
+ return -ENOSPC;
+
-+ vector_irq[irq_op.vector] = irq;
-+ if (irq != AUTO_ASSIGN)
-+ IO_APIC_VECTOR(irq) = irq_op.vector;
++ vector = irq_op.vector;
++ per_cpu(vector_irq,0)[vector] = irq;
++ irq_vector[irq] = vector;
++ cpus_and(*result, irq_domain[irq], mask);
++
++ return vector;
++}
++
++static int assign_irq_vector(int irq, cpumask_t mask, cpumask_t *result)
++{
++ int vector;
++ unsigned long flags;
+
++ spin_lock_irqsave(&vector_lock, flags);
++ vector = __assign_irq_vector(irq, mask, result);
+ spin_unlock_irqrestore(&vector_lock, flags);
-+ return irq_op.vector;
++ return vector;
++}
++
++static void __clear_irq_vector(int irq)
++{
++ cpumask_t mask;
++ int cpu, vector;
++
++ BUG_ON(!irq_vector[irq]);
++
++ vector = irq_vector[irq];
++ cpus_and(mask, irq_domain[irq], cpu_online_map);
++ for_each_cpu_mask(cpu, mask)
++ per_cpu(vector_irq, cpu)[vector] = -1;
++
++ irq_vector[irq] = 0;
++ irq_domain[irq] = CPU_MASK_NONE;
++}
++
++void __setup_vector_irq(int cpu)
++{
++ /* Initialize vector_irq on a new cpu */
++ /* This function must be called with vector_lock held */
++ int irq, vector;
++
++ /* Mark the inuse vectors */
++ for (irq = 0; irq < NR_IRQ_VECTORS; ++irq) {
++ if (!cpu_isset(cpu, irq_domain[irq]))
++ continue;
++ vector = irq_vector[irq];
++ per_cpu(vector_irq, cpu)[vector] = irq;
++ }
++ /* Mark the free vectors */
++ for (vector = 0; vector < NR_VECTORS; ++vector) {
++ irq = per_cpu(vector_irq, cpu)[vector];
++ if (irq < 0)
++ continue;
++ if (!cpu_isset(cpu, irq_domain[irq]))
++ per_cpu(vector_irq, cpu)[vector] = -1;
++ }
+}
+
++
+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;
++static struct irq_chip ioapic_chip;
+
+#define IOAPIC_AUTO -1
+#define IOAPIC_EDGE 0
@@ -33401,41 +28957,79 @@
+
+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]);
++ set_irq_chip_and_handler_name(irq, &ioapic_chip,
++ handle_fasteoi_irq, "fasteoi");
++ else {
++ irq_desc[irq].status |= IRQ_DELAYED_DISABLE;
++ set_irq_chip_and_handler_name(irq, &ioapic_chip,
++ handle_edge_irq, "edge");
++ }
+}
+#else
+#define ioapic_register_intr(_irq,_vector,_trigger) ((void)0)
+#endif /* !CONFIG_XEN */
+
-+static void __init setup_IO_APIC_irqs(void)
++static void __init setup_IO_APIC_irq(int apic, int pin, int idx, int irq)
+{
+ struct IO_APIC_route_entry entry;
-+ int apic, pin, idx, irq, first_notcon = 1, vector;
++ int 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));
+
-+ /*
-+ * 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);
+
-+ entry.delivery_mode = INT_DELIVERY_MODE;
-+ entry.dest_mode = INT_DEST_MODE;
-+ entry.mask = 0; /* enable IRQ */
++ 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);
++ }
++
++ if (!apic && !IO_APIC_IRQ(irq))
++ return;
++
++ if (IO_APIC_IRQ(irq)) {
++ cpumask_t mask;
++ vector = assign_irq_vector(irq, TARGET_CPUS, &mask);
++ if (vector < 0)
++ return;
++
++ entry.dest.logical.logical_dest = cpu_mask_to_apicid(mask);
++ entry.vector = vector;
++
++ ioapic_register_intr(irq, vector, IOAPIC_AUTO);
++ if (!apic && (irq < 16))
++ disable_8259A_irq(irq);
++ }
++
++ ioapic_write_entry(apic, pin, entry);
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ set_native_irq_info(irq, TARGET_CPUS);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++}
++
++static void __init setup_IO_APIC_irqs(void)
++{
++ int apic, pin, idx, irq, first_notcon = 1;
++
++ 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++) {
+
+ idx = find_irq_entry(apic,pin,mp_INT);
+ if (idx == -1) {
@@ -33447,34 +29041,11 @@
+ 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;
++ setup_IO_APIC_irq(apic, pin, idx, irq);
+
-+ 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);
+ }
+ }
+
@@ -33515,7 +29086,7 @@
+ * 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;
++ set_irq_chip_and_handler_name(0, &ioapic_chip, handle_edge_irq, "edge");
+
+ /*
+ * Add it to the IO-APIC irq-routing table:
@@ -33611,10 +29182,7 @@
+ 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);
++ entry = ioapic_read_entry(apic, i);
+
+ printk(KERN_DEBUG " %02x %03X %02X ",
+ i,
@@ -33634,17 +29202,12 @@
+ );
+ }
+ }
-+ 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);
++ printk(KERN_DEBUG "IRQ%d ", i);
+ for (;;) {
+ printk("-> %d:%d", entry->apic, entry->pin);
+ if (!entry->next)
@@ -33815,9 +29378,6 @@
+ 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):
@@ -33834,11 +29394,7 @@
+ /* 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);
-+
++ entry = ioapic_read_entry(apic, pin);
+
+ /* If the interrupt line is enabled and in ExtInt mode
+ * I have found the pin where the i8259 is connected.
@@ -33892,7 +29448,6 @@
+ */
+ if (ioapic_i8259.pin != -1) {
+ struct IO_APIC_route_entry entry;
-+ unsigned long flags;
+
+ memset(&entry, 0, sizeof(entry));
+ entry.mask = 0; /* Enabled */
@@ -33909,12 +29464,7 @@
+ /*
+ * 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);
++ ioapic_write_entry(ioapic_i8259.apic, ioapic_i8259.pin, entry);
+ }
+
+ disconnect_bsp_APIC(ioapic_i8259.pin != -1);
@@ -33922,76 +29472,6 @@
+}
+
+/*
-+ * 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:
+ *
@@ -34045,7 +29525,7 @@
+ * an edge even if it isn't on the 8259A...
+ */
+
-+static unsigned int startup_edge_ioapic_irq(unsigned int irq)
++static unsigned int startup_ioapic_irq(unsigned int irq)
+{
+ int was_pending = 0;
+ unsigned long flags;
@@ -34062,107 +29542,19 @@
+ 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)
++static int ioapic_retrigger_irq(unsigned int irq)
+{
-+ 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);
++ cpumask_t mask;
++ unsigned vector;
++ unsigned long flags;
+
-+ set_native_irq_info(vector, cpu_mask);
-+ set_ioapic_affinity_irq(irq, cpu_mask);
-+}
-+#endif // CONFIG_SMP
-+#endif // CONFIG_PCI_MSI
++ spin_lock_irqsave(&vector_lock, flags);
++ vector = irq_vector[irq];
++ cpus_clear(mask);
++ cpu_set(first_cpu(irq_domain[irq]), mask);
+
-+static int ioapic_retrigger(unsigned int irq)
-+{
-+ send_IPI_self(IO_APIC_VECTOR(irq));
++ send_IPI_mask(mask, vector);
++ spin_unlock_irqrestore(&vector_lock, flags);
+
+ return 1;
+}
@@ -34176,32 +29568,47 @@
+ * 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,
++static void ack_apic_edge(unsigned int irq)
++{
++ move_native_irq(irq);
++ ack_APIC_irq();
++}
++
++static void ack_apic_level(unsigned int irq)
++{
++ int do_unmask_irq = 0;
++
++#if defined(CONFIG_GENERIC_PENDING_IRQ) || defined(CONFIG_IRQBALANCE)
++ /* If we are moving the irq we need to mask it */
++ if (unlikely(irq_desc[irq].status & IRQ_MOVE_PENDING)) {
++ do_unmask_irq = 1;
++ mask_IO_APIC_irq(irq);
++ }
+#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,
++ /*
++ * We must acknowledge the irq before we move it or the acknowledge will
++ * not propogate properly.
++ */
++ ack_APIC_irq();
++
++ /* Now we can move and renable the irq */
++ move_masked_irq(irq);
++ if (unlikely(do_unmask_irq))
++ unmask_IO_APIC_irq(irq);
++}
++
++static struct irq_chip ioapic_chip __read_mostly = {
++ .name = "IO-APIC",
++ .startup = startup_ioapic_irq,
++ .mask = mask_IO_APIC_irq,
++ .unmask = unmask_IO_APIC_irq,
++ .ack = ack_apic_edge,
++ .eoi = ack_apic_level,
+#ifdef CONFIG_SMP
-+ .set_affinity = set_ioapic_affinity,
++ .set_affinity = set_ioapic_affinity_irq,
+#endif
-+ .retrigger = ioapic_retrigger,
++ .retrigger = ioapic_retrigger_irq,
+};
+#endif /* !CONFIG_XEN */
+
@@ -34222,12 +29629,7 @@
+ */
+ 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)) {
++ if (IO_APIC_IRQ(tmp) && !irq_vector[tmp]) {
+ /*
+ * Hmm.. We don't have an entry for this,
+ * so default to an old-fashioned 8259
@@ -34238,7 +29640,7 @@
+#ifndef CONFIG_XEN
+ else
+ /* Strange. Oh, well.. */
-+ irq_desc[irq].chip = &no_irq_type;
++ irq_desc[irq].chip = &no_irq_chip;
+#endif
+ }
+ }
@@ -34359,8 +29761,6 @@
+ 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
@@ -34373,13 +29773,13 @@
+{
+ int apic1, pin1, apic2, pin2;
+ int vector;
++ cpumask_t mask;
+
+ /*
+ * get/set the timer IRQ vector:
+ */
+ disable_8259A_irq(0);
-+ vector = assign_irq_vector(0);
-+ set_intr_gate(vector, interrupt[0]);
++ vector = assign_irq_vector(0, TARGET_CPUS, &mask);
+
+ /*
+ * Subtle, code in do_timer_interrupt() expects an AEOI
@@ -34398,9 +29798,6 @@
+ 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);
+
@@ -34484,7 +29881,6 @@
+ panic("IO-APIC + timer doesn't work! Try using the 'noapic' kernel parameter\n");
+}
+#else
-+int timer_uses_ioapic_pin_0;
+#define check_timer() ((void)0)
+#endif /* !CONFIG_XEN */
+
@@ -34515,11 +29911,6 @@
+
+ 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 */
@@ -34540,17 +29931,12 @@
+{
+ 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);
++ for (i = 0; i < nr_ioapic_registers[dev->id]; i ++, entry ++ )
++ *entry = ioapic_read_entry(dev->id, i);
+
+ return 0;
+}
@@ -34572,11 +29958,9 @@
+ 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);
++ for (i = 0; i < nr_ioapic_registers[dev->id]; i++)
++ ioapic_write_entry(dev->id, i, entry[i]);
+
+ return 0;
+}
@@ -34622,26 +30006,253 @@
+
+device_initcall(ioapic_init_sysfs);
+
-+/* --------------------------------------------------------------------------
-+ ACPI-based IOAPIC Configuration
-+ -------------------------------------------------------------------------- */
++/*
++ * Dynamic irq allocate and deallocation
++ */
++int create_irq(void)
++{
++ /* Allocate an unused irq */
++ int irq;
++ int new;
++ int vector = 0;
++ unsigned long flags;
++ cpumask_t mask;
+
-+#ifdef CONFIG_ACPI
++ irq = -ENOSPC;
++ spin_lock_irqsave(&vector_lock, flags);
++ for (new = (NR_IRQS - 1); new >= 0; new--) {
++ if (platform_legacy_irq(new))
++ continue;
++ if (irq_vector[new] != 0)
++ continue;
++ vector = __assign_irq_vector(new, TARGET_CPUS, &mask);
++ if (likely(vector > 0))
++ irq = new;
++ break;
++ }
++ spin_unlock_irqrestore(&vector_lock, flags);
+
-+#define IO_APIC_MAX_ID 0xFE
++ if (irq >= 0) {
++ dynamic_irq_init(irq);
++ }
++ return irq;
++}
+
-+int __init io_apic_get_version (int ioapic)
++void destroy_irq(unsigned int irq)
+{
-+ 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);
++ dynamic_irq_cleanup(irq);
+
-+ return reg_01.bits.version;
++ spin_lock_irqsave(&vector_lock, flags);
++ __clear_irq_vector(irq);
++ spin_unlock_irqrestore(&vector_lock, flags);
++}
++
++/*
++ * MSI mesage composition
++ */
++#ifndef CONFIG_XEN
++#ifdef CONFIG_PCI_MSI
++static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_msg *msg)
++{
++ int vector;
++ unsigned dest;
++ cpumask_t tmp;
++
++ vector = assign_irq_vector(irq, TARGET_CPUS, &tmp);
++ if (vector >= 0) {
++ dest = cpu_mask_to_apicid(tmp);
++
++ msg->address_hi = MSI_ADDR_BASE_HI;
++ msg->address_lo =
++ MSI_ADDR_BASE_LO |
++ ((INT_DEST_MODE == 0) ?
++ MSI_ADDR_DEST_MODE_PHYSICAL:
++ MSI_ADDR_DEST_MODE_LOGICAL) |
++ ((INT_DELIVERY_MODE != dest_LowestPrio) ?
++ MSI_ADDR_REDIRECTION_CPU:
++ MSI_ADDR_REDIRECTION_LOWPRI) |
++ MSI_ADDR_DEST_ID(dest);
++
++ msg->data =
++ MSI_DATA_TRIGGER_EDGE |
++ MSI_DATA_LEVEL_ASSERT |
++ ((INT_DELIVERY_MODE != dest_LowestPrio) ?
++ MSI_DATA_DELIVERY_FIXED:
++ MSI_DATA_DELIVERY_LOWPRI) |
++ MSI_DATA_VECTOR(vector);
++ }
++ return vector;
++}
++
++#ifdef CONFIG_SMP
++static void set_msi_irq_affinity(unsigned int irq, cpumask_t mask)
++{
++ struct msi_msg msg;
++ unsigned int dest;
++ cpumask_t tmp;
++ int vector;
++
++ cpus_and(tmp, mask, cpu_online_map);
++ if (cpus_empty(tmp))
++ tmp = TARGET_CPUS;
++
++ cpus_and(mask, tmp, CPU_MASK_ALL);
++
++ vector = assign_irq_vector(irq, mask, &tmp);
++ if (vector < 0)
++ return;
++
++ dest = cpu_mask_to_apicid(tmp);
++
++ read_msi_msg(irq, &msg);
++
++ msg.data &= ~MSI_DATA_VECTOR_MASK;
++ msg.data |= MSI_DATA_VECTOR(vector);
++ msg.address_lo &= ~MSI_ADDR_DEST_ID_MASK;
++ msg.address_lo |= MSI_ADDR_DEST_ID(dest);
++
++ write_msi_msg(irq, &msg);
++ set_native_irq_info(irq, mask);
++}
++#endif /* CONFIG_SMP */
++
++/*
++ * IRQ Chip for MSI PCI/PCI-X/PCI-Express Devices,
++ * which implement the MSI or MSI-X Capability Structure.
++ */
++static struct irq_chip msi_chip = {
++ .name = "PCI-MSI",
++ .unmask = unmask_msi_irq,
++ .mask = mask_msi_irq,
++ .ack = ack_apic_edge,
++#ifdef CONFIG_SMP
++ .set_affinity = set_msi_irq_affinity,
++#endif
++ .retrigger = ioapic_retrigger_irq,
++};
++
++int arch_setup_msi_irq(unsigned int irq, struct pci_dev *dev)
++{
++ struct msi_msg msg;
++ int ret;
++ ret = msi_compose_msg(dev, irq, &msg);
++ if (ret < 0)
++ return ret;
++
++ write_msi_msg(irq, &msg);
++
++ set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq, "edge");
++
++ return 0;
++}
++
++void arch_teardown_msi_irq(unsigned int irq)
++{
++ return;
++}
++
++#endif /* CONFIG_PCI_MSI */
++#endif /* !CONFIG_XEN */
++/*
++ * Hypertransport interrupt support
++ */
++#ifdef CONFIG_HT_IRQ
++
++#ifdef CONFIG_SMP
++
++static void target_ht_irq(unsigned int irq, unsigned int dest, u8 vector)
++{
++ struct ht_irq_msg msg;
++ fetch_ht_irq_msg(irq, &msg);
++
++ msg.address_lo &= ~(HT_IRQ_LOW_VECTOR_MASK | HT_IRQ_LOW_DEST_ID_MASK);
++ msg.address_hi &= ~(HT_IRQ_HIGH_DEST_ID_MASK);
++
++ msg.address_lo |= HT_IRQ_LOW_VECTOR(vector) | HT_IRQ_LOW_DEST_ID(dest);
++ msg.address_hi |= HT_IRQ_HIGH_DEST_ID(dest);
++
++ write_ht_irq_msg(irq, &msg);
++}
++
++static void set_ht_irq_affinity(unsigned int irq, cpumask_t mask)
++{
++ unsigned int dest;
++ cpumask_t tmp;
++ int vector;
++
++ cpus_and(tmp, mask, cpu_online_map);
++ if (cpus_empty(tmp))
++ tmp = TARGET_CPUS;
++
++ cpus_and(mask, tmp, CPU_MASK_ALL);
++
++ vector = assign_irq_vector(irq, mask, &tmp);
++ if (vector < 0)
++ return;
++
++ dest = cpu_mask_to_apicid(tmp);
++
++ target_ht_irq(irq, dest, vector);
++ set_native_irq_info(irq, mask);
++}
++#endif
++
++static struct irq_chip ht_irq_chip = {
++ .name = "PCI-HT",
++ .mask = mask_ht_irq,
++ .unmask = unmask_ht_irq,
++ .ack = ack_apic_edge,
++#ifdef CONFIG_SMP
++ .set_affinity = set_ht_irq_affinity,
++#endif
++ .retrigger = ioapic_retrigger_irq,
++};
++
++int arch_setup_ht_irq(unsigned int irq, struct pci_dev *dev)
++{
++ int vector;
++ cpumask_t tmp;
++
++ vector = assign_irq_vector(irq, TARGET_CPUS, &tmp);
++ if (vector >= 0) {
++ struct ht_irq_msg msg;
++ unsigned dest;
++
++ dest = cpu_mask_to_apicid(tmp);
++
++ msg.address_hi = HT_IRQ_HIGH_DEST_ID(dest);
++
++ msg.address_lo =
++ HT_IRQ_LOW_BASE |
++ HT_IRQ_LOW_DEST_ID(dest) |
++ HT_IRQ_LOW_VECTOR(vector) |
++ ((INT_DEST_MODE == 0) ?
++ HT_IRQ_LOW_DM_PHYSICAL :
++ HT_IRQ_LOW_DM_LOGICAL) |
++ HT_IRQ_LOW_RQEOI_EDGE |
++ ((INT_DELIVERY_MODE != dest_LowestPrio) ?
++ HT_IRQ_LOW_MT_FIXED :
++ HT_IRQ_LOW_MT_ARBITRATED) |
++ HT_IRQ_LOW_IRQ_MASKED;
++
++ write_ht_irq_msg(irq, &msg);
++
++ set_irq_chip_and_handler_name(irq, &ht_irq_chip,
++ handle_edge_irq, "edge");
++ }
++ return vector;
+}
++#endif /* CONFIG_HT_IRQ */
++
++/* --------------------------------------------------------------------------
++ ACPI-based IOAPIC Configuration
++ -------------------------------------------------------------------------- */
+
++#ifdef CONFIG_ACPI
++
++#define IO_APIC_MAX_ID 0xFE
+
+int __init io_apic_get_redir_entries (int ioapic)
+{
@@ -34656,10 +30267,12 @@
+}
+
+
-+int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int active_high_low)
++int io_apic_set_pci_routing (int ioapic, int pin, int irq, int triggering, int polarity)
+{
+ struct IO_APIC_route_entry entry;
+ unsigned long flags;
++ int vector;
++ cpumask_t mask;
+
+ if (!IO_APIC_IRQ(irq)) {
+ apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
@@ -34668,6 +30281,17 @@
+ }
+
+ /*
++ * IRQs < 16 are already in the irq_2_pin[] map
++ */
++ if (irq >= 16)
++ add_pin_to_irq(irq, ioapic, pin);
++
++
++ vector = assign_irq_vector(irq, TARGET_CPUS, &mask);
++ if (vector < 0)
++ return vector;
++
++ /*
+ * 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.
@@ -34677,34 +30301,26 @@
+
+ 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.dest.logical.logical_dest = cpu_mask_to_apicid(mask);
++ entry.trigger = triggering;
++ entry.polarity = polarity;
+ 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);
++ entry.vector = vector & 0xff;
+
+ 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);
++ triggering, polarity);
+
-+ ioapic_register_intr(irq, entry.vector, edge_level);
++ ioapic_register_intr(irq, entry.vector, triggering);
+
+ if (!ioapic && (irq < 16))
+ disable_8259A_irq(irq);
+
++ ioapic_write_entry(ioapic, pin, entry);
++
+ 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);
++ set_native_irq_info(irq, TARGET_CPUS);
+ spin_unlock_irqrestore(&ioapic_lock, flags);
+
+ return 0;
@@ -34733,16 +30349,24 @@
+ if (irq_entry == -1)
+ continue;
+ irq = pin_2_irq(irq_entry, ioapic, pin);
-+ set_ioapic_affinity_irq(irq, TARGET_CPUS);
++
++ /* setup_IO_APIC_irqs could fail to get vector for some device
++ * when you have too many devices, because at that time only boot
++ * cpu is online.
++ */
++ if(!irq_vector[irq])
++ setup_IO_APIC_irq(ioapic, pin, irq_entry, irq);
++ else
++ set_ioapic_affinity_irq(irq, TARGET_CPUS);
+ }
+
+ }
+}
+#endif
+#endif /* !CONFIG_XEN */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/ioport-xen.c linux-2.6.18-xen/arch/x86_64/kernel/ioport-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/ioport-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/ioport-xen.c 2006-11-19 14:26:33.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/ioport-xen.c b/arch/x86_64/kernel/ioport-xen.c
+--- a/arch/x86_64/kernel/ioport-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/ioport-xen.c 2007-03-14 10:55:21.000000000 +0100
@@ -0,0 +1,99 @@
+/*
+ * linux/arch/x86_64/kernel/ioport.c
@@ -34769,13 +30393,12 @@
+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);
++ 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);
+}
+
+/*
@@ -34804,6 +30427,7 @@
+
+ memset(bitmap, 0xff, IO_BITMAP_BYTES);
+ t->io_bitmap_ptr = bitmap;
++ set_thread_flag(TIF_IO_BITMAP);
+
+ set_iobitmap.bitmap = (char *)bitmap;
+ set_iobitmap.nr_ports = IO_BITMAP_BITS;
@@ -34822,337 +30446,46 @@
+ *
+ */
+
-+asmlinkage long sys_iopl(unsigned int new_iopl, struct pt_regs *regs)
++asmlinkage long sys_iopl(unsigned int level, struct pt_regs *regs)
+{
-+ unsigned int old_iopl = current->thread.iopl;
++ unsigned int old = current->thread.iopl;
+ struct physdev_set_iopl set_iopl;
+
-+ if (new_iopl > 3)
++ if (level > 3)
+ return -EINVAL;
-+
-+ /* Need "raw I/O" privileges for direct port access. */
-+ if ((new_iopl > old_iopl) && !capable(CAP_SYS_RAWIO))
-+ return -EPERM;
-+
++ /* Trying to gain more privileges? */
++ if (level > old) {
++ if (!capable(CAP_SYS_RAWIO))
++ return -EPERM;
++ }
+ /* Change our version of the privilege levels. */
-+ current->thread.iopl = new_iopl;
++ current->thread.iopl = level;
+
+ /* Force the change at ring 0. */
-+ set_iopl.iopl = (new_iopl == 0) ? 1 : new_iopl;
++ set_iopl.iopl = (level == 0) ? 1 : level;
+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
+
+ return 0;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/irqflags-xen.c linux-2.6.18-xen/arch/x86_64/kernel/irqflags-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/irqflags-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/irqflags-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,100 @@
-+#include <linux/module.h>
-+#include <linux/smp.h>
-+#include <asm/irqflags.h>
-+#include <asm/hypervisor.h>
-+
-+/*
-+ * 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.
-+ */
-+
-+unsigned long __raw_local_save_flags(void)
-+{
-+ struct vcpu_info *_vcpu;
-+ unsigned long flags;
-+
-+ preempt_disable();
-+ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
-+ flags = _vcpu->evtchn_upcall_mask;
-+ preempt_enable();
-+
-+ return flags;
-+}
-+EXPORT_SYMBOL(__raw_local_save_flags);
-+
-+void raw_local_irq_restore(unsigned long flags)
-+{
-+ struct vcpu_info *_vcpu;
-+ preempt_disable();
-+ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
-+ if ((_vcpu->evtchn_upcall_mask = flags) == 0) {
-+ barrier(); /* unmask then check (avoid races) */
-+ if ( unlikely(_vcpu->evtchn_upcall_pending) )
-+ force_evtchn_callback();
-+ preempt_enable();
-+ } else
-+ preempt_enable_no_resched();
-+}
-+EXPORT_SYMBOL(raw_local_irq_restore);
-+
-+void raw_local_irq_disable(void)
-+{
-+ struct vcpu_info *_vcpu;
-+
-+ preempt_disable();
-+ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
-+ _vcpu->evtchn_upcall_mask = 1;
-+ preempt_enable_no_resched();
-+}
-+EXPORT_SYMBOL(raw_local_irq_disable);
-+
-+void raw_local_irq_enable(void)
-+{
-+ struct vcpu_info *_vcpu;
-+
-+ preempt_disable();
-+ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
-+ _vcpu->evtchn_upcall_mask = 0;
-+ barrier(); /* unmask then check (avoid races) */
-+ if ( unlikely(_vcpu->evtchn_upcall_pending) )
-+ force_evtchn_callback();
-+ preempt_enable();
-+}
-+EXPORT_SYMBOL(raw_local_irq_enable);
-+
-+/*
-+ * For spinlocks, etc.:
-+ */
-+
-+unsigned long __raw_local_irq_save(void)
-+{
-+ struct vcpu_info *_vcpu;
-+ unsigned long flags;
-+
-+ preempt_disable();
-+ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
-+ flags = _vcpu->evtchn_upcall_mask;
-+ _vcpu->evtchn_upcall_mask = 1;
-+ preempt_enable_no_resched();
-+
-+ return flags;
-+}
-+EXPORT_SYMBOL(__raw_local_irq_save);
-+
-+/* Cannot use preempt_enable() here as we would recurse in preempt_sched(). */
-+int raw_irqs_disabled(void)
-+{
-+ struct vcpu_info *_vcpu;
-+ int disabled;
-+
-+ preempt_disable();
-+ _vcpu = &HYPERVISOR_shared_info->vcpu_info[__vcpu_id];
-+ disabled = (_vcpu->evtchn_upcall_mask != 0);
-+ preempt_enable_no_resched();
-+
-+ return disabled;
-+}
-+EXPORT_SYMBOL(raw_irqs_disabled);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/irq-xen.c linux-2.6.18-xen/arch/x86_64/kernel/irq-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/irq-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/irq-xen.c 2006-11-19 14:26:33.000000000 +0100
-@@ -0,0 +1,198 @@
-+/*
-+ * 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));
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/irq.c b/arch/x86_64/kernel/irq.c
+--- a/arch/x86_64/kernel/irq.c 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/kernel/irq.c 2007-03-14 10:55:21.000000000 +0100
+@@ -113,7 +113,11 @@
+
+ exit_idle();
+ irq_enter();
++#ifdef CONFIG_XEN
++ irq = vector;
+#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);
-+}
+ irq = __get_cpu_var(vector_irq)[vector];
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/ldt-xen.c linux-2.6.18-xen/arch/x86_64/kernel/ldt-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/ldt-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/ldt-xen.c 2006-11-19 14:26:33.000000000 +0100
-@@ -0,0 +1,282 @@
+
+ #ifdef CONFIG_DEBUG_STACKOVERFLOW
+ stack_overflow_check(regs);
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/ldt-xen.c b/arch/x86_64/kernel/ldt-xen.c
+--- a/arch/x86_64/kernel/ldt-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/ldt-xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,283 @@
+/*
+ * linux/arch/x86_64/kernel/ldt.c
+ *
@@ -35219,6 +30552,7 @@
+ cpumask_t mask;
+
+ preempt_disable();
++ mask = cpumask_of_cpu(smp_processor_id());
+#endif
+ make_pages_readonly(
+ pc->ldt,
@@ -35226,7 +30560,6 @@
+ 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();
@@ -35269,6 +30602,7 @@
+
+ memset(&mm->context, 0, sizeof(mm->context));
+ init_MUTEX(&mm->context.sem);
++ mm->context.size = 0;
+ old_mm = current->mm;
+ if (old_mm && old_mm->context.size > 0) {
+ down(&old_mm->context.sem);
@@ -35435,15 +30769,15 @@
+ }
+ return ret;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/Makefile linux-2.6.18-xen/arch/x86_64/kernel/Makefile
---- linux-2.6.18.3/arch/x86_64/kernel/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/x86_64/kernel/Makefile 2006-12-05 18:42:36.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/Makefile b/arch/x86_64/kernel/Makefile
+--- a/arch/x86_64/kernel/Makefile 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/kernel/Makefile 2007-03-14 10:55:21.000000000 +0100
@@ -21,11 +21,13 @@
obj-$(CONFIG_X86_CPUID) += cpuid.o
obj-$(CONFIG_SMP) += smp.o smpboot.o trampoline.o
- obj-$(CONFIG_X86_LOCAL_APIC) += apic.o nmi.o
+ obj-y += apic.o nmi.o
+obj-$(CONFIG_X86_XEN_GENAPIC) += genapic.o genapic_xen.o
- obj-$(CONFIG_X86_IO_APIC) += io_apic.o mpparse.o \
+ obj-y += io_apic.o mpparse.o \
genapic.o genapic_cluster.o genapic_flat.o
obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o crash.o
obj-$(CONFIG_CRASH_DUMP) += crash_dump.o
@@ -35453,4503 +30787,4528 @@
obj-$(CONFIG_SOFTWARE_SUSPEND) += suspend_asm.o
obj-$(CONFIG_CPU_FREQ) += cpufreq/
obj-$(CONFIG_EARLY_PRINTK) += early_printk.o
-@@ -55,3 +57,18 @@
+@@ -56,3 +58,17 @@
+ 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
+obj-y += irqflags.o
-+time-y += ../../i386/kernel/time-xen.o
++irqflags-y += ../../i386/mach-xen/irqflags.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))
++endif
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/mpparse.c b/arch/x86_64/kernel/mpparse.c
+--- a/arch/x86_64/kernel/mpparse.c 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/kernel/mpparse.c 2007-03-14 10:55:21.000000000 +0100
+@@ -90,6 +90,7 @@
+
+ static void __cpuinit MP_processor_info (struct mpc_config_processor *m)
+ {
++#ifndef CONFIG_XEN
+ int cpu;
+ cpumask_t tmp_map;
+ char *bootup_cpu = "";
+@@ -110,8 +111,9 @@
+ " Processor ignored.\n", NR_CPUS);
+ return;
+ }
+-
++#endif /* !CONFIG_XEN */
+ num_processors++;
++#ifndef CONFIG_XEN
+ cpus_complement(tmp_map, cpu_present_map);
+ cpu = first_cpu(tmp_map);
+
+@@ -129,6 +131,7 @@
+
+ cpu_set(cpu, cpu_possible_map);
+ cpu_set(cpu, cpu_present_map);
++#endif /* CONFIG_XEN */
+ }
+
+ static void __init MP_bus_info (struct mpc_config_bus *m)
+@@ -488,7 +491,11 @@
+ * Read the physical hardware table. Anything here will
+ * override the defaults.
+ */
++#ifdef CONFIG_XEN
++ if (!smp_read_mpc(isa_bus_to_virt(mpf->mpf_physptr))) {
++#else
+ if (!smp_read_mpc(phys_to_virt(mpf->mpf_physptr))) {
++#endif
+ 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");
+@@ -524,7 +531,12 @@
+ static int __init smp_scan_config (unsigned long base, unsigned long length)
+ {
+ extern void __bad_mpf_size(void);
++#ifdef CONFIG_XEN
++ unsigned int *bp = isa_bus_to_virt(base);
++#else
+ unsigned int *bp = phys_to_virt(base);
++#endif
++
+ struct intel_mp_floating *mpf;
+
+ Dprintk("Scan SMP from %p for %ld bytes.\n", bp,length);
+@@ -540,9 +552,11 @@
+ || (mpf->mpf_specification == 4)) ) {
+
+ smp_found_config = 1;
++#ifndef CONFIG_XEN
+ reserve_bootmem_generic(virt_to_phys(mpf), PAGE_SIZE);
+ if (mpf->mpf_physptr)
+ reserve_bootmem_generic(mpf->mpf_physptr, PAGE_SIZE);
++#endif
+ mpf_found = mpf;
+ return 1;
+ }
+@@ -598,10 +612,12 @@
+
+ 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));
++#endif
+ }
+
+ void __cpuinit mp_register_lapic (u8 id, u8 enabled)
+@@ -612,6 +628,7 @@
+ if (id == boot_cpu_id)
+ boot_cpu = 1;
+
++#ifndef CONFIG_XEN
+ processor.mpc_type = MP_PROCESSOR;
+ processor.mpc_apicid = id;
+ processor.mpc_apicver = 0;
+@@ -621,6 +638,7 @@
+ processor.mpc_featureflag = 0;
+ processor.mpc_reserved[0] = 0;
+ processor.mpc_reserved[1] = 0;
++#endif
+
+ MP_processor_info(&processor);
+ }
+@@ -663,7 +681,9 @@
+ 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 = 0;
+
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/pci-swiotlb.c b/arch/x86_64/kernel/pci-swiotlb.c
+--- a/arch/x86_64/kernel/pci-swiotlb.c 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/kernel/pci-swiotlb.c 2007-03-14 10:55:21.000000000 +0100
+@@ -9,10 +9,13 @@
+ #include <asm/swiotlb.h>
+ #include <asm/dma.h>
+
++#ifndef CONFIG_XEN
+ int swiotlb __read_mostly;
+ EXPORT_SYMBOL(swiotlb);
++#endif
+
+ struct dma_mapping_ops swiotlb_dma_ops = {
++#ifndef CONFIG_XEN
+ .mapping_error = swiotlb_dma_mapping_error,
+ .alloc_coherent = swiotlb_alloc_coherent,
+ .free_coherent = swiotlb_free_coherent,
+@@ -27,10 +30,18 @@
+ .map_sg = swiotlb_map_sg,
+ .unmap_sg = swiotlb_unmap_sg,
+ .dma_supported = NULL,
++#endif
+ };
+
+ void pci_swiotlb_init(void)
+ {
++#ifdef CONFIG_XEN
++ swiotlb_init();
++ if (swiotlb) {
++ printk(KERN_INFO "PCI-DMA: Using software bounce buffering for IO (SWIOTLB)\n");
++ dma_ops = &swiotlb_dma_ops;
++ }
++#else
+ /* don't initialize swiotlb if iommu=off (no_iommu=1) */
+ if (!iommu_detected && !no_iommu && end_pfn > MAX_DMA32_PFN)
+ swiotlb = 1;
+@@ -41,4 +52,5 @@
+ swiotlb_init();
+ dma_ops = &swiotlb_dma_ops;
+ }
++#endif /* CONFIG_XEN */
+ }
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/process-xen.c b/arch/x86_64/kernel/process-xen.c
+--- a/arch/x86_64/kernel/process-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/process-xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,865 @@
++/*
++ * 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/dom0_ops.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>
+
-+n-obj-xen := i8259.o reboot.o i8237.o smpboot.o trampoline.o
++#include <xen/cpu_hotplug.h>
+
-+include $(srctree)/scripts/Makefile.xen
++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);
+
-+obj-y := $(call filterxen, $(obj-y), $(n-obj-xen))
-+obj-y := $(call cherrypickxen, $(obj-y))
-+extra-y := $(call cherrypickxen, $(extra-y))
-+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/mpparse-xen.c linux-2.6.18-xen/arch/x86_64/kernel/mpparse-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/mpparse-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/mpparse-xen.c 2006-11-19 14:26:33.000000000 +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
++ * Powermanagement idle function, if any..
+ */
++void (*pm_idle)(void);
++static DEFINE_PER_CPU(unsigned int, cpu_idle_state);
+
-+#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>
++static ATOMIC_NOTIFIER_HEAD(idle_notifier);
+
-+#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>
++void idle_notifier_register(struct notifier_block *n)
++{
++ atomic_notifier_chain_register(&idle_notifier, n);
++}
++EXPORT_SYMBOL_GPL(idle_notifier_register);
+
-+/* Have we found an MP table */
-+int smp_found_config;
-+unsigned int __initdata maxcpus = NR_CPUS;
++void idle_notifier_unregister(struct notifier_block *n)
++{
++ atomic_notifier_chain_unregister(&idle_notifier, n);
++}
++EXPORT_SYMBOL(idle_notifier_unregister);
+
-+int acpi_found_madt;
++void enter_idle(void)
++{
++ write_pda(isidle, 1);
++ atomic_notifier_call_chain(&idle_notifier, IDLE_START, NULL);
++}
+
-+/*
-+ * 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 void __exit_idle(void)
++{
++ if (test_and_clear_bit_pda(0, isidle) == 0)
++ return;
++ atomic_notifier_call_chain(&idle_notifier, IDLE_END, NULL);
++}
+
-+static int mp_current_pci_id = 0;
-+/* I/O APIC entries */
-+struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS];
++/* Called from interrupts to signify idle end */
++void exit_idle(void)
++{
++ /* idle loop has pid 0 */
++ if (current->pid)
++ return;
++ __exit_idle();
++}
+
-+/* # of MP IRQ source entries */
-+struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES];
++/* XXX XEN doesn't use default_idle(), poll_idle(). Use xen_idle() instead. */
++void xen_idle(void)
++{
++ current_thread_info()->status &= ~TS_POLLING;
++ /*
++ * TS_POLLING-cleared state must be visible before we
++ * test NEED_RESCHED:
++ */
++ smp_mb();
++ local_irq_disable();
++ if (!need_resched()) {
++ /* Enables interrupts one instruction before HLT.
++ x86 special cases this so there is no race. */
++ safe_halt();
++ } else
++ local_irq_enable();
++ current_thread_info()->status |= TS_POLLING;
++}
+
-+/* MP IRQ source entries */
-+int mp_irq_entries;
++#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();
++ HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL);
++ cpu_bringup();
++}
++#else
++static inline void play_dead(void)
++{
++ BUG();
++}
++#endif /* CONFIG_HOTPLUG_CPU */
+
-+int nr_ioapics;
-+int pic_mode;
-+unsigned long mp_lapic_addr = 0;
++/*
++ * 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()) {
++ if (__get_cpu_var(cpu_idle_state))
++ __get_cpu_var(cpu_idle_state) = 0;
+
++ rmb();
++ if (cpu_is_offline(smp_processor_id()))
++ play_dead();
++ /*
++ * Idle routines should keep interrupts disabled
++ * from here on, until they go to idle.
++ * Otherwise, idle callbacks can misfire.
++ */
++ local_irq_disable();
++ enter_idle();
++ xen_idle();
++ /* In many cases the interrupt that ended idle
++ has already called exit_idle. But some idle
++ loops can be woken up without interrupt. */
++ __exit_idle();
++ }
+
++ preempt_enable_no_resched();
++ schedule();
++ preempt_disable();
++ }
++}
+
-+/* Processor that is doing the boot up */
-+unsigned int boot_cpu_id = -1U;
-+/* Internal processor count */
-+unsigned int num_processors __initdata = 0;
++void cpu_idle_wait(void)
++{
++ unsigned int cpu, this_cpu = get_cpu();
++ cpumask_t map, tmp = current->cpus_allowed;
+
-+unsigned disabled_cpus __initdata;
++ set_cpus_allowed(current, cpumask_of_cpu(this_cpu));
++ put_cpu();
+
-+/* Bitmask of physically existing CPUs */
-+physid_mask_t phys_cpu_present_map = PHYSID_MASK_NONE;
++ cpus_clear(map);
++ for_each_online_cpu(cpu) {
++ per_cpu(cpu_idle_state, cpu) = 1;
++ cpu_set(cpu, map);
++ }
+
-+/* 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*/
++ __get_cpu_var(cpu_idle_state) = 0;
+
-+u8 bios_cpu_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = BAD_APICID };
++ 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));
+
++ set_cpus_allowed(current, tmp);
++}
++EXPORT_SYMBOL_GPL(cpu_idle_wait);
+
-+/*
-+ * Intel MP BIOS table parsing routines:
-+ */
++/* XXX XEN doesn't use mwait_idle(), select_idle_routine(), idle_setup(). */
++/* Always use xen_idle() instead. */
++void mwait_idle_with_hints(unsigned long eax, unsigned long ecx) {}
+
-+/*
-+ * Checksum an MP configuration block.
-+ */
++void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c) {}
+
-+static int __init mpf_checksum(unsigned char *mp, int len)
++/* Prints also some state that isn't saved in the pt_regs */
++void __show_regs(struct pt_regs * regs)
+{
-+ int sum = 0;
++ unsigned long cr0 = 0L, cr2 = 0L, cr3 = 0L, cr4 = 0L, fs, gs, shadowgs;
++ unsigned int fsindex,gsindex;
++ unsigned int ds,cs,es;
+
-+ while (len--)
-+ sum += *mp++;
++ printk("\n");
++ print_modules();
++ printk("Pid: %d, comm: %.20s %s %s %.*s\n",
++ current->pid, current->comm, print_tainted(),
++ init_utsname()->release,
++ (int)strcspn(init_utsname()->version, " "),
++ init_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);
+
-+ return sum & 0xFF;
-+}
++ asm("movl %%ds,%0" : "=r" (ds));
++ asm("movl %%cs,%0" : "=r" (cs));
++ asm("movl %%es,%0" : "=r" (es));
++ asm("movl %%fs,%0" : "=r" (fsindex));
++ asm("movl %%gs,%0" : "=r" (gsindex));
+
-+#ifndef CONFIG_XEN
-+static void __cpuinit MP_processor_info (struct mpc_config_processor *m)
-+{
-+ int cpu;
-+ unsigned char ver;
-+ cpumask_t tmp_map;
++ rdmsrl(MSR_FS_BASE, fs);
++ rdmsrl(MSR_GS_BASE, gs);
++ rdmsrl(MSR_KERNEL_GS_BASE, shadowgs);
+
-+ if (!(m->mpc_cpuflag & CPU_ENABLED)) {
-+ disabled_cpus++;
-+ return;
-+ }
++ cr0 = read_cr0();
++ cr2 = 0; /* No real clue how to read it. JQ */
++ cr3 = read_cr3();
++ cr4 = read_cr4();
+
-+ 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);
++ printk("FS: %016lx(%04x) GS:%016lx(%04x) knlGS:%016lx\n",
++ fs,fsindex,gs,gsindex,shadowgs);
++ printk("CS: %04x DS: %04x ES: %04x CR0: %016lx\n", cs, ds, es, cr0);
++ printk("CR2: %016lx CR3: %016lx CR4: %016lx\n", cr2, cr3, cr4);
++}
+
-+ 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;
-+ }
++void show_regs(struct pt_regs *regs)
++{
++ printk("CPU %d:", smp_processor_id());
++ __show_regs(regs);
++ show_trace(NULL, regs, ®s->rsp);
++}
+
-+ num_processors++;
-+ cpus_complement(tmp_map, cpu_present_map);
-+ cpu = first_cpu(tmp_map);
++/*
++ * Free current thread data structures etc..
++ */
++void exit_thread(void)
++{
++ struct task_struct *me = current;
++ struct thread_struct *t = &me->thread;
+
-+#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;
-+ }
++ 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 = { 0 };
+#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;
++ kfree(t->io_bitmap_ptr);
++ t->io_bitmap_ptr = NULL;
++ clear_thread_flag(TIF_IO_BITMAP);
++ /*
++ * 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;
+ }
-+ 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
-+void __init MP_processor_info (struct mpc_config_processor *m)
++
++void load_gs_index(unsigned gs)
+{
-+ num_processors++;
++ HYPERVISOR_set_segment_base(SEGBASE_GS_USER_SEL, gs);
+}
-+#endif /* CONFIG_XEN */
+
-+static void __init MP_bus_info (struct mpc_config_bus *m)
++void flush_thread(void)
+{
-+ char str[7];
++ struct task_struct *tsk = current;
++ struct thread_info *t = current_thread_info();
+
-+ 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);
++ if (t->flags & _TIF_ABI_PENDING) {
++ t->flags ^= (_TIF_ABI_PENDING | _TIF_IA32);
++ if (t->flags & _TIF_IA32)
++ current_thread_info()->status |= TS_COMPAT;
+ }
++ t->flags &= ~_TIF_DEBUG;
++
++ 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();
+}
+
-+static void __init MP_ioapic_info (struct mpc_config_ioapic *m)
++void release_thread(struct task_struct *dead_task)
+{
-+ 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;
++ 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();
++ }
+ }
-+ mp_ioapics[nr_ioapics] = *m;
-+ nr_ioapics++;
+}
+
-+static void __init MP_intsrc_info (struct mpc_config_intsrc *m)
++static inline void set_32bit_tls(struct task_struct *t, int tls, u32 addr)
+{
-+ 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");
++ 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 void __init MP_lintsrc_info (struct mpc_config_lintsrc *m)
++static inline u32 read_32bit_tls(struct task_struct *t, int tls)
+{
-+ 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();
++ struct desc_struct *desc = (void *)t->thread.tls_array;
++ desc += tls;
++ return desc->base0 |
++ (((u32)desc->base1) << 16) |
++ (((u32)desc->base2) << 24);
+}
+
+/*
-+ * Read/parse the MPC
++ * 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);
++}
+
-+static int __init smp_read_mpc(struct mp_config_table *mpc)
++int copy_thread(int nr, unsigned long clone_flags, unsigned long rsp,
++ unsigned long unused,
++ struct task_struct * p, struct pt_regs * regs)
+{
-+ char str[16];
-+ int count=sizeof(*mpc);
-+ unsigned char *mpt=((unsigned char *)mpc)+count;
++ int err;
++ struct pt_regs * childregs;
++ struct task_struct *me = current;
+
-+ 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);
++ 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);
+
-+ memcpy(str,mpc->mpc_productid,12);
-+ str[12]=0;
-+ printk("Product ID: %s ",str);
++ p->thread.fs = me->thread.fs;
++ p->thread.gs = me->thread.gs;
+
-+ printk("APIC at: 0x%X\n",mpc->mpc_lapic);
++ 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));
+
-+ /* save the local APIC address, it might be non-default */
-+ if (!acpi_lapic)
-+ mp_lapic_addr = mpc->mpc_lapic;
++ if (unlikely(test_tsk_thread_flag(me, TIF_IO_BITMAP))) {
++ 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_tsk_thread_flag(p, TIF_IO_BITMAP);
++ }
+
+ /*
-+ * Now process the configuration blocks.
++ * Set a new TLS for the child thread?
+ */
-+ 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;
++ 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;
+
-+ 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;
-+ }
-+ }
++ err = 0;
++out:
++ if (err && p->thread.io_bitmap_ptr) {
++ kfree(p->thread.io_bitmap_ptr);
++ p->thread.io_bitmap_max = 0;
+ }
-+ clustered_apic_check();
-+ if (!num_processors)
-+ printk(KERN_ERR "SMP mptable: no processors registered!\n");
-+ return num_processors;
++ return err;
+}
+
-+static int __init ELCR_trigger(unsigned int irq)
++static inline void __save_init_fpu( struct task_struct *tsk )
+{
-+ unsigned int port;
++ asm volatile( "rex64 ; fxsave %0 ; fnclex"
++ : "=m" (tsk->thread.i387.fxsave));
++ tsk->thread_info->status &= ~TS_USEDFPU;
++}
+
-+ port = 0x4d0 + (irq >> 3);
-+ return (inb(port) >> (irq & 7)) & 1;
++/*
++ * This special macro can be used to load a debugging register
++ */
++#define loaddebug(thread,r) set_debugreg(thread->debugreg ## r, r)
++
++static inline void __switch_to_xtra(struct task_struct *prev_p,
++ struct task_struct *next_p)
++{
++ struct thread_struct *prev, *next;
++
++ prev = &prev_p->thread,
++ next = &next_p->thread;
++
++ if (test_tsk_thread_flag(next_p, TIF_DEBUG)) {
++ loaddebug(next, 0);
++ loaddebug(next, 1);
++ loaddebug(next, 2);
++ loaddebug(next, 3);
++ /* no 4 and 5 */
++ loaddebug(next, 6);
++ loaddebug(next, 7);
++ }
++#ifndef CONFIG_XEN
++ if (test_tsk_thread_flag(next_p, TIF_IO_BITMAP)) {
++ /*
++ * Copy the relevant range of the IO bitmap.
++ * Normally this is 128 bytes or less:
++ */
++ memcpy(tss->io_bitmap, next->io_bitmap_ptr,
++ max(prev->io_bitmap_max, next->io_bitmap_max));
++ } else if (test_tsk_thread_flag(prev_p, TIF_IO_BITMAP)) {
++ /*
++ * Clear any possible leftover bits:
++ */
++ memset(tss->io_bitmap, 0xff, prev->io_bitmap_max);
++ }
++#endif
+}
+
-+static void __init construct_default_ioirq_mptable(int mpc_default_type)
++/*
++ * 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 mpc_config_intsrc intsrc;
-+ int i;
-+ int ELCR_fallback = 0;
++ 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
++ struct physdev_set_iopl iopl_op;
++ struct physdev_set_iobitmap iobmp_op;
++ multicall_entry_t _mcl[8], *mcl = _mcl;
+
-+ intsrc.mpc_type = MP_INTSRC;
-+ intsrc.mpc_irqflag = 0; /* conforming */
-+ intsrc.mpc_srcbus = 0;
-+ intsrc.mpc_dstapic = mp_ioapics[0].mpc_apicid;
++ /*
++ * Reload esp0, LDT and the page table pointer:
++ */
++ mcl->op = __HYPERVISOR_stack_switch;
++ mcl->args[0] = __KERNEL_DS;
++ mcl->args[1] = next->rsp0;
++ mcl++;
+
-+ intsrc.mpc_irqtype = mp_INT;
++ /* we're going to use this soon, after a few expensive things */
++ if (next_p->fpu_counter>5)
++ prefetch(&next->i387.fxsave);
+
+ /*
-+ * 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.
++ * Load the per-thread Thread-Local Storage descriptor.
++ * This is load_TLS(next, cpu) with multicalls.
+ */
-+ if (mpc_default_type == 5) {
-+ printk(KERN_INFO "ISA/PCI bus type with no IRQ information... falling back to ELCR\n");
++#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 (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;
-+ }
++ 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++;
+ }
+
-+ 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 (test_tsk_thread_flag(prev_p, TIF_IO_BITMAP) ||
++ test_tsk_thread_flag(next_p, TIF_IO_BITMAP)) {
++ 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++;
++ }
+
-+ 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;
-+ }
++ (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);
+
-+ intsrc.mpc_srcbusirq = i;
-+ intsrc.mpc_dstirq = i ? i : 2; /* IRQ0 to INTIN2 */
-+ MP_intsrc_info(&intsrc);
-+ }
++ /*
++ * Switch FS and GS.
++ */
++ if (unlikely(next->fsindex))
++ loadsegment(fs, next->fsindex);
+
-+ intsrc.mpc_irqtype = mp_ExtINT;
-+ intsrc.mpc_srcbusirq = 0;
-+ intsrc.mpc_dstirq = 0; /* 8259A to INTIN0 */
-+ MP_intsrc_info(&intsrc);
-+}
++ if (next->fs)
++ HYPERVISOR_set_segment_base(SEGBASE_FS, next->fs);
++
++ if (unlikely(next->gsindex))
++ load_gs_index(next->gsindex);
+
-+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;
++ if (next->gs)
++ HYPERVISOR_set_segment_base(SEGBASE_GS_USER, next->gs);
++
++ /* Must be after DS reload */
++ /*
++ * This is basically '__unlazy_fpu'
++ */
++ if (prev_p->thread_info->status & TS_USEDFPU) {
++ __save_init_fpu(prev_p); /* _not_ save_init_fpu() */
++ HYPERVISOR_fpu_taskswitch(1);
++ }
++
++ /*
++ * Switch the PDA and FPU contexts.
++ */
++ prev->userrsp = read_pda(oldrsp);
++ write_pda(oldrsp, next->userrsp);
++ write_pda(pcurrent, next_p);
+
++ write_pda(kernelstack,
++ (unsigned long)task_stack_page(next_p) + THREAD_SIZE - PDA_STACKOFFSET);
++#ifdef CONFIG_CC_STACKPROTECTOR
++ write_pda(stack_canary, next_p->stack_canary);
+ /*
-+ * local APIC has default address
++ * Build time only check to make sure the stack_canary is at
++ * offset 40 in the pda; this is a gcc ABI requirement
+ */
-+ mp_lapic_addr = APIC_DEFAULT_PHYS_BASE;
++ BUILD_BUG_ON(offsetof(struct x8664_pda, stack_canary) != 40);
++#endif
+
+ /*
-+ * 2 CPUs, numbered 0 & 1.
++ * Now maybe reload the debug registers and handle I/O bitmaps
+ */
-+ 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);
++ if (unlikely((task_thread_info(next_p)->flags & _TIF_WORK_CTXSW))
++ || test_tsk_thread_flag(prev_p, TIF_IO_BITMAP))
++ __switch_to_xtra(prev_p, next_p);
+
-+ /*
-+ * We set up most of the low 16 IO-APIC pins according to MPS rules.
++ /* If the task has used fpu the last 5 timeslices, just do a full
++ * restore of the math state immediately to avoid the trap; the
++ * chances of needing FPU soon are obviously high now
+ */
-+ 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);
-+ }
++ if (next_p->fpu_counter>5)
++ math_state_restore();
++ return prev_p;
+}
+
-+static struct intel_mp_floating *mpf_found;
-+
+/*
-+ * Scan the memory blocks for an SMP configuration block.
++ * sys_execve() executes a new program.
+ */
-+void __init get_smp_config (void)
++asmlinkage
++long sys_execve(char __user *name, char __user * __user *argv,
++ char __user * __user *envp, struct pt_regs regs)
+{
-+ 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");
++ long error;
++ char * filename;
+
-+ 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;
++ 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;
++}
+
-+ /*
-+ * 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;
++void set_personality_64bit(void)
++{
++ /* inherit personality from parent */
+
-+ printk(KERN_ERR "BIOS bug, no explicit IRQ entries, using default mptable. (tell your hw vendor)\n");
++ /* Make sure to be in 64bit mode */
++ clear_thread_flag(TIF_IA32);
+
-+ bus.mpc_type = MP_BUS;
-+ bus.mpc_busid = 0;
-+ memcpy(bus.mpc_bustype, "ISA ", 6);
-+ MP_bus_info(&bus);
++ /* 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;
++}
+
-+ construct_default_ioirq_mptable(0);
-+ }
++asmlinkage long sys_fork(struct pt_regs *regs)
++{
++ return do_fork(SIGCHLD, regs->rsp, regs, 0, NULL, NULL);
++}
+
-+ } else
-+ BUG();
++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);
++}
+
-+ printk(KERN_INFO "Processors: %d\n", num_processors);
-+ /*
-+ * Only use the first configuration found.
-+ */
++/*
++ * 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);
+}
+
-+static int __init smp_scan_config (unsigned long base, unsigned long length)
++unsigned long get_wchan(struct task_struct *p)
+{
-+ extern void __bad_mpf_size(void);
-+ unsigned int *bp = isa_bus_to_virt(base);
-+ struct intel_mp_floating *mpf;
++ 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;
++}
+
-+ 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)) ) {
++long do_arch_prctl(struct task_struct *task, int code, unsigned long addr)
++{
++ int ret = 0;
++ int doit = task == current;
++ int cpu;
+
-+ smp_found_config = 1;
-+ mpf_found = mpf;
-+ return 1;
++ 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);
++ }
+ }
-+ bp += 4;
-+ length -= 16;
++ 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;
+ }
-+ return 0;
-+}
++ 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;
++ }
+
-+void __init find_intel_smp (void)
++ return ret;
++}
++
++long sys_arch_prctl(int code, unsigned long addr)
+{
-+ unsigned int address;
++ return do_arch_prctl(current, code, addr);
++}
+
-+ /*
-+ * 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.
-+ */
++/*
++ * 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;
+
-+ address = *(unsigned short *)phys_to_virt(0x40E);
-+ address <<= 4;
-+ if (smp_scan_config(address, 0x1000))
-+ return;
++ pp = task_pt_regs(tsk);
++
++ ptregs = *pp;
++ ptregs.cs &= 0xffff;
++ ptregs.ss &= 0xffff;
+
-+ /* If we have come this far, we did not find an MP table */
-+ printk(KERN_INFO "No mptable found.\n");
++ elf_core_copy_regs(regs, &ptregs);
++
++ boot_option_idle_override = 1;
++ return 1;
+}
+
-+/*
-+ * - Intel MP Configuration Table
-+ */
-+void __init find_smp_config (void)
++unsigned long arch_align_stack(unsigned long sp)
+{
-+#ifdef CONFIG_X86_LOCAL_APIC
-+ find_intel_smp();
-+#endif
++ if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space)
++ sp -= get_random_int() % 8192;
++ return sp & ~0xf;
+}
+
++#ifndef CONFIG_SMP
++void _restore_vcpu(void)
++{
++}
++#endif
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/setup64-xen.c b/arch/x86_64/kernel/setup64-xen.c
+--- a/arch/x86_64/kernel/setup64-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/setup64-xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,356 @@
++/*
++ * 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
+
-+/* --------------------------------------------------------------------------
-+ ACPI-based MP Configuration
-+ -------------------------------------------------------------------------- */
++char x86_boot_params[BOOT_PARAM_SIZE] __initdata;
+
-+#ifdef CONFIG_ACPI
++cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE;
+
-+void __init mp_register_lapic_address (
-+ u64 address)
-+{
-+#ifndef CONFIG_XEN
-+ mp_lapic_addr = (unsigned long) address;
++struct x8664_pda *_cpu_pda[NR_CPUS] __read_mostly;
++EXPORT_SYMBOL(_cpu_pda);
++struct x8664_pda boot_cpu_pda[NR_CPUS] __cacheline_aligned;
+
-+ set_fixmap_nocache(FIX_APIC_BASE, mp_lapic_addr);
++#ifndef CONFIG_X86_NO_IDT
++struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table };
++#endif
+
-+ if (boot_cpu_id == -1U)
-+ boot_cpu_id = GET_APIC_ID(apic_read(APIC_ID));
++char boot_cpu_stack[IRQSTACKSIZE] __attribute__((section(".bss.page_aligned")));
+
-+ Dprintk("Boot CPU = %d\n", boot_cpu_physical_apicid);
-+#endif
-+}
++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.
+
-+void __cpuinit mp_register_lapic (
-+ u8 id,
-+ u8 enabled)
++on Enable(default)
++off Disable
++*/
++static int __init nonx_setup(char *str)
+{
-+ 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 (!str)
++ return -EINVAL;
++ 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 0;
++}
++early_param("noexec", nonx_setup);
+
-+ if (id == boot_cpu_physical_apicid)
-+ boot_cpu = 1;
++int force_personality32 = 0;
+
-+#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
++/* noexec32=on|off
++Control non executable heap for 32bit processes.
++To control the stack too use noexec=off
+
-+ MP_processor_info(&processor);
++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);
+
-+#ifdef CONFIG_X86_IO_APIC
++/*
++ * 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;
+
-+#define MP_ISA_BUS 0
-+#define MP_MAX_IOAPIC_PIN 127
++#ifdef CONFIG_HOTPLUG_CPU
++ prefill_possible_map();
++#endif
+
-+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;
-+ }
++ /* Copy section for each CPU (we discard the original) */
++ size = PERCPU_ENOUGH_ROOM;
+
-+ printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi);
++ printk(KERN_INFO "PERCPU: Allocating %lu bytes of per cpu data\n", size);
++ for_each_cpu_mask (i, cpu_possible_map) {
++ char *ptr;
+
-+ return -1;
++ 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(init_level4_pgt));
++ xen_new_user_pt(__pa(init_level4_user_pgt));
+}
-+
+
-+void __init mp_register_ioapic (
-+ u8 id,
-+ u32 address,
-+ u32 gsi_base)
++void __cpuinit cpu_gdt_init(struct desc_ptr *gdt_descr)
+{
-+ int idx = 0;
++ unsigned long frames[16];
++ unsigned long va;
++ int f;
+
-+ 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;
++ 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 /
++ sizeof (struct desc_struct)))
++ BUG();
++}
++#else
++static void switch_pt(void)
++{
++ asm volatile("movq %0,%%cr3" :: "r" (__pa_symbol(&init_level4_pgt)));
++}
+
-+ idx = nr_ioapics++;
++void __init cpu_gdt_init(struct desc_ptr *gdt_descr)
++{
++ asm volatile("lgdt %0" :: "m" (*gdt_descr));
++ asm volatile("lidt %0" :: "m" (idt_descr));
++}
++#endif
+
-+ mp_ioapics[idx].mpc_type = MP_IOAPIC;
-+ mp_ioapics[idx].mpc_flags = MPC_APIC_USABLE;
-+ mp_ioapics[idx].mpc_apicaddr = address;
++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
-+ set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address);
++ /* Memory clobbers used to order PDA accessed */
++ mb();
++ wrmsrl(MSR_GS_BASE, pda);
++ mb();
++#else
++ HYPERVISOR_set_segment_base(SEGBASE_GS_KERNEL, (unsigned long)pda);
+#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# */
++ 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;
+
-+ 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);
++ 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);
++ }
+
-+ mp_irqs[mp_irq_entries] = intsrc;
-+ if (++mp_irq_entries == MAX_IRQ_SOURCES)
-+ panic("Max # of irq sources exceeded!\n");
++ switch_pt();
+
-+ return;
-+}
++ 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
+
-+void __init mp_config_acpi_legacy_irqs (void)
++/* May not be marked __init: used by software suspend */
++void syscall_init(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);
-+
++#ifndef CONFIG_XEN
+ /*
-+ * Locate the IOAPIC that manages the ISA IRQs (0-15).
-+ */
-+ ioapic = mp_find_ioapic(0);
-+ if (ioapic < 0)
-+ return;
++ * 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);
+
-+ intsrc.mpc_type = MP_INTSRC;
-+ intsrc.mpc_irqflag = 0; /* Conforming */
-+ intsrc.mpc_srcbus = MP_ISA_BUS;
-+ intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid;
++ /* 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
++}
+
-+ /*
-+ * 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;
++void __cpuinit check_efer(void)
++{
++ unsigned long efer;
+
-+ for (idx = 0; idx < mp_irq_entries; idx++) {
-+ struct mpc_config_intsrc *irq = mp_irqs + idx;
++ rdmsrl(MSR_EFER, efer);
++ if (!(efer & EFER_NX) || do_not_nx) {
++ __supported_pte_mask &= ~_PAGE_NX;
++ }
++}
+
-+ /* Do we already have a mapping for this ISA IRQ? */
-+ if (irq->mpc_srcbus == MP_ISA_BUS && irq->mpc_srcbusirq == i)
-+ break;
++unsigned long kernel_eflags;
+
-+ /* Do we already have a mapping for this IOAPIC pin */
-+ if ((irq->mpc_dstapic == intsrc.mpc_dstapic) &&
-+ (irq->mpc_dstirq == i))
-+ break;
-+ }
++/*
++ * 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;
++ int i;
++#endif
++ struct task_struct *me;
+
-+ 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");
++ /* 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
+
-+ return;
-+}
++ me = current;
+
-+#define MAX_GSI_NUM 4096
++ if (cpu_test_and_set(cpu, cpu_initialized))
++ panic("CPU#%d already initialized!\n", cpu);
+
-+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];
++ printk("Initializing CPU#%d\n", cpu);
+
-+ if (acpi_irq_model != ACPI_IRQ_MODEL_IOAPIC)
-+ return gsi;
++ clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE);
+
-+ /* Don't set up the ACPI SCI because it's already set up */
-+ if (acpi_fadt.sci_int == gsi)
-+ return gsi;
++ /*
++ * 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
+
-+ ioapic = mp_find_ioapic(gsi);
-+ if (ioapic < 0) {
-+ printk(KERN_WARNING "No IOAPIC for GSI %u\n", gsi);
-+ return gsi;
-+ }
++ cpu_gdt_descr[cpu].size = GDT_SIZE;
++ cpu_gdt_init(&cpu_gdt_descr[cpu]);
+
-+ ioapic_pin = gsi - mp_ioapic_routing[ioapic].gsi_start;
++ memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8);
++ syscall_init();
+
-+ /*
-+ * 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];
-+ }
++ wrmsrl(MSR_FS_BASE, 0);
++ wrmsrl(MSR_KERNEL_GS_BASE, 0);
++ barrier();
+
-+ mp_ioapic_routing[ioapic].pin_programmed[idx] |= (1<<bit);
++ check_efer();
+
-+ 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;
++#ifndef CONFIG_X86_NO_TSS
++ /*
++ * set up and load the per-CPU TSS
++ */
++ for (v = 0; v < N_EXCEPTION_STACKS; v++) {
++ static const unsigned int order[N_EXCEPTION_STACKS] = {
++ [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STACK_ORDER,
++ [DEBUG_STACK - 1] = DEBUG_STACK_ORDER
++ };
++ if (cpu) {
++ estacks = (char *)__get_free_pages(GFP_ATOMIC, order[v]);
++ if (!estacks)
++ panic("Cannot allocate exception stack %ld %d\n",
++ v, cpu);
+ }
++ estacks += PAGE_SIZE << order[v];
++ orig_ist->ist[v] = t->ist[v] = (unsigned long)estacks;
+ }
+
-+ 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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/pci-swiotlb-xen.c linux-2.6.18-xen/arch/x86_64/kernel/pci-swiotlb-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/pci-swiotlb-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/pci-swiotlb-xen.c 2006-11-19 14:26:33.000000000 +0100
-@@ -0,0 +1,54 @@
-+/* Glue code to lib/swiotlb.c */
++ 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
+
-+#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>
++ atomic_inc(&init_mm.mm_count);
++ me->active_mm = &init_mm;
++ if (me->mm)
++ BUG();
++ enter_lazy_tlb(&init_mm, me);
+
-+#if 0
-+int swiotlb __read_mostly;
-+EXPORT_SYMBOL(swiotlb);
++#ifndef CONFIG_X86_NO_TSS
++ set_tss_desc(cpu, t);
+#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,
++#ifndef CONFIG_XEN
++ load_TR_desc();
+#endif
-+};
++ load_LDT(&init_mm.context);
+
-+void pci_swiotlb_init(void)
-+{
-+#if 0
-+ /* don't initialize swiotlb if iommu=off (no_iommu=1) */
-+ if (!iommu_aperture && !no_iommu &&
-+ (end_pfn > MAX_DMA32_PFN || force_iommu))
-+ 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
++ /*
++ * 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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/process-xen.c linux-2.6.18-xen/arch/x86_64/kernel/process-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/process-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/process-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,787 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/setup-xen.c b/arch/x86_64/kernel/setup-xen.c
+--- a/arch/x86_64/kernel/setup-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/setup-xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,1478 @@
+/*
-+ * linux/arch/x86-64/kernel/process.c
++ * linux/arch/x86-64/kernel/setup.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
++ * Nov 2001 Dave Jones <davej at suse.de>
++ * Forked from i386 setup code.
+ */
+
+/*
-+ * This file handles the architecture-dependent parts of process handling..
++ * This file handles the architecture-dependent parts of initialization
+ */
+
-+#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/stddef.h>
++#include <linux/unistd.h>
++#include <linux/ptrace.h>
+#include <linux/slab.h>
+#include <linux/user.h>
-+#include <linux/module.h>
+#include <linux/a.out.h>
-+#include <linux/interrupt.h>
++#include <linux/screen_info.h>
++#include <linux/ioport.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 <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/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/dom0_ops.h>
-+#include <xen/interface/physdev.h>
-+#include <xen/interface/vcpu.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/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);
-+static DEFINE_PER_CPU(unsigned int, cpu_idle_state);
-+
-+static ATOMIC_NOTIFIER_HEAD(idle_notifier);
++#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 <linux/pfn.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>
++#include <asm/mach-xen/setup_arch_post.h>
++#include <xen/interface/memory.h>
+
-+void idle_notifier_register(struct notifier_block *n)
-+{
-+ atomic_notifier_chain_register(&idle_notifier, n);
-+}
-+EXPORT_SYMBOL_GPL(idle_notifier_register);
++extern unsigned long start_pfn;
+
-+void idle_notifier_unregister(struct notifier_block *n)
-+{
-+ atomic_notifier_chain_unregister(&idle_notifier, n);
-+}
-+EXPORT_SYMBOL(idle_notifier_unregister);
++shared_info_t *HYPERVISOR_shared_info = (shared_info_t *)empty_zero_page;
++EXPORT_SYMBOL(HYPERVISOR_shared_info);
+
-+enum idle_state { CPU_IDLE, CPU_NOT_IDLE };
-+static DEFINE_PER_CPU(enum idle_state, idle_state) = CPU_NOT_IDLE;
++extern char hypercall_page[PAGE_SIZE];
++EXPORT_SYMBOL(hypercall_page);
+
-+void enter_idle(void)
-+{
-+ __get_cpu_var(idle_state) = CPU_IDLE;
-+ atomic_notifier_call_chain(&idle_notifier, IDLE_START, NULL);
-+}
++/* Allows setting of maximum possible memory size */
++unsigned long xen_override_max_pfn;
+
-+static void __exit_idle(void)
-+{
-+ __get_cpu_var(idle_state) = CPU_NOT_IDLE;
-+ atomic_notifier_call_chain(&idle_notifier, IDLE_END, NULL);
-+}
++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 */
++};
+
-+/* Called from interrupts to signify idle end */
-+void exit_idle(void)
-+{
-+ if (current->pid | read_pda(irqcount))
-+ return;
-+ __exit_idle();
-+}
++unsigned long *phys_to_machine_mapping;
++unsigned long *pfn_to_mfn_frame_list_list, *pfn_to_mfn_frame_list[512];
+
-+/* XXX XEN doesn't use default_idle(), poll_idle(). Use xen_idle() instead. */
-+void xen_idle(void)
-+{
-+ local_irq_disable();
++EXPORT_SYMBOL(phys_to_machine_mapping);
+
-+ 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;
-+ }
-+}
++DEFINE_PER_CPU(multicall_entry_t, multicall_list[8]);
++DEFINE_PER_CPU(int, nr_multicall_ents);
+
-+#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();
-+ HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL);
-+ cpu_bringup();
-+}
-+#else
-+static inline void play_dead(void)
-+{
-+ BUG();
-+}
-+#endif /* CONFIG_HOTPLUG_CPU */
++/* Raw start-of-day parameters from the hypervisor. */
++start_info_t *xen_start_info;
++EXPORT_SYMBOL(xen_start_info);
++#endif
+
+/*
-+ * 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)
++ * Machine setup..
+ */
-+void cpu_idle (void)
-+{
-+ current_thread_info()->status |= TS_POLLING;
-+ /* endless idle loop with no priority at all */
-+ while (1) {
-+ while (!need_resched()) {
-+ if (__get_cpu_var(cpu_idle_state))
-+ __get_cpu_var(cpu_idle_state) = 0;
-+ rmb();
-+
-+ if (cpu_is_offline(smp_processor_id()))
-+ play_dead();
-+ enter_idle();
-+ xen_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);
-+
-+/* XXX XEN doesn't use mwait_idle(), select_idle_routine(), idle_setup(). */
-+/* Always use xen_idle() instead. */
-+void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c) {}
-+
-+/* 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));
++struct cpuinfo_x86 boot_cpu_data __read_mostly;
++EXPORT_SYMBOL(boot_cpu_data);
+
-+ rdmsrl(MSR_FS_BASE, fs);
-+ rdmsrl(MSR_GS_BASE, gs);
-+ rdmsrl(MSR_KERNEL_GS_BASE, shadowgs);
++unsigned long mmu_cr4_features;
+
-+ 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);
++/* Boot loader ID as an integer, for the benefit of proc_dointvec */
++int bootloader_type;
+
-+}
++unsigned long saved_video_mode;
+
-+void show_regs(struct pt_regs *regs)
-+{
-+ printk("CPU %d:", smp_processor_id());
-+ __show_regs(regs);
-+ show_trace(NULL, regs, ®s->rsp);
-+}
++/*
++ * Early DMI memory
++ */
++int dmi_alloc_index;
++char dmi_alloc_data[DMI_MAX_DATA];
+
+/*
-+ * Free current thread data structures etc..
++ * Setup options
+ */
-+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 = { 0 };
-+#endif
++struct screen_info screen_info;
++EXPORT_SYMBOL(screen_info);
++struct sys_desc_table_struct {
++ unsigned short length;
++ unsigned char table[0];
++};
+
-+ 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
++struct edid_info edid_info;
++EXPORT_SYMBOL_GPL(edid_info);
+#ifdef CONFIG_XEN
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap, &iobmp_op);
++struct e820map machine_e820;
+#endif
-+ t->io_bitmap_max = 0;
-+ }
-+}
-+
-+void load_gs_index(unsigned gs)
-+{
-+ 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();
-+ }
-+ }
-+}
++extern int root_mountflags;
+
-+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);
-+}
++char command_line[COMMAND_LINE_SIZE];
+
-+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);
-+}
++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 }
++};
+
-+/*
-+ * 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);
-+}
++#define IORESOURCE_RAM (IORESOURCE_BUSY | IORESOURCE_MEM)
+
-+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;
++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,
++};
+
-+ childregs = ((struct pt_regs *)
-+ (THREAD_SIZE + task_stack_page(p))) - 1;
-+ *childregs = *regs;
++#define IORESOURCE_ROM (IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM)
+
-+ childregs->rax = 0;
-+ childregs->rsp = rsp;
-+ if (rsp == ~0UL)
-+ childregs->rsp = (unsigned long)childregs;
++static struct resource system_rom_resource = {
++ .name = "System ROM",
++ .start = 0xf0000,
++ .end = 0xfffff,
++ .flags = IORESOURCE_ROM,
++};
+
-+ p->thread.rsp = (unsigned long) childregs;
-+ p->thread.rsp0 = (unsigned long) (childregs+1);
-+ p->thread.userrsp = me->thread.userrsp;
++static struct resource extension_rom_resource = {
++ .name = "Extension ROM",
++ .start = 0xe0000,
++ .end = 0xeffff,
++ .flags = IORESOURCE_ROM,
++};
+
-+ set_tsk_thread_flag(p, TIF_FORK);
++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 }
++};
+
-+ p->thread.fs = me->thread.fs;
-+ p->thread.gs = me->thread.gs;
++static struct resource video_rom_resource = {
++ .name = "Video ROM",
++ .start = 0xc0000,
++ .end = 0xc7fff,
++ .flags = IORESOURCE_ROM,
++};
+
-+ 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));
++static struct resource video_ram_resource = {
++ .name = "Video RAM area",
++ .start = 0xa0000,
++ .end = 0xbffff,
++ .flags = IORESOURCE_RAM,
++};
+
-+ 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);
-+ }
++#define romsignature(x) (*(unsigned short *)(x) == 0xaa55)
+
-+ /*
-+ * 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;
++static int __init romchecksum(unsigned char *rom, unsigned long length)
++{
++ unsigned char *p, sum = 0;
+
-+ err = 0;
-+out:
-+ if (err && p->thread.io_bitmap_ptr) {
-+ kfree(p->thread.io_bitmap_ptr);
-+ p->thread.io_bitmap_max = 0;
-+ }
-+ return err;
++ for (p = rom; p < rom + length; p++)
++ sum += *p;
++ return sum == 0;
+}
+
-+static inline void __save_init_fpu( struct task_struct *tsk )
++static void __init probe_roms(void)
+{
-+ asm volatile( "rex64 ; fxsave %0 ; fnclex"
-+ : "=m" (tsk->thread.i387.fxsave));
-+ tsk->thread_info->status &= ~TS_USEDFPU;
-+}
++ unsigned long start, length, upper;
++ unsigned char *rom;
++ int i;
+
-+/*
-+ * 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);
++#ifdef CONFIG_XEN
++ /* Nothing to do if not running in dom0. */
++ if (!is_initial_xendomain())
++ return;
+#endif
-+ struct physdev_set_iopl iopl_op;
-+ struct physdev_set_iobitmap iobmp_op;
-+ multicall_entry_t _mcl[8], *mcl = _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
++ /* 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;
+
-+ 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++;
-+ }
++ video_rom_resource.start = start;
+
-+ if (unlikely(prev->io_bitmap_ptr || next->io_bitmap_ptr)) {
-+ 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++;
-+ }
++ /* 0 < length <= 0x7f * 512, historically */
++ length = rom[2] * 512;
+
-+ (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);
++ /* if checksum okay, trust length byte */
++ if (length && romchecksum(rom, length))
++ video_rom_resource.end = start + length - 1;
+
-+ /*
-+ * Switch FS and GS.
-+ */
-+ if (unlikely(next->fsindex))
-+ loadsegment(fs, next->fsindex);
++ request_resource(&iomem_resource, &video_rom_resource);
++ break;
++ }
+
-+ if (next->fs)
-+ HYPERVISOR_set_segment_base(SEGBASE_FS, next->fs);
-+
-+ if (unlikely(next->gsindex))
-+ load_gs_index(next->gsindex);
++ start = (video_rom_resource.end + 1 + 2047) & ~2047UL;
++ if (start < upper)
++ start = upper;
+
-+ if (next->gs)
-+ HYPERVISOR_set_segment_base(SEGBASE_GS_USER, next->gs);
++ /* system rom */
++ request_resource(&iomem_resource, &system_rom_resource);
++ upper = system_rom_resource.start;
+
-+ /* Must be after DS reload */
-+ if (prev_p->thread_info->status & TS_USEDFPU) {
-+ __save_init_fpu(prev_p); /* _not_ save_init_fpu() */
-+ HYPERVISOR_fpu_taskswitch(1);
++ /* 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;
++ }
+ }
+
-+ /*
-+ * Switch the PDA and FPU contexts.
-+ */
-+ 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);
++ /* check for adapter roms on 2k boundaries */
++ for (i = 0; i < ARRAY_SIZE(adapter_rom_resources) && start < upper;
++ start += 2048) {
++ rom = isa_bus_to_virt(start);
++ if (!romsignature(rom))
++ continue;
+
-+ /*
-+ * 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);
-+ }
++ /* 0 < length <= 0x7f * 512, historically */
++ length = rom[2] * 512;
+
-+ return prev_p;
-+}
++ /* but accept any length that fits if checksum okay */
++ if (!length || start + length > upper || !romchecksum(rom, length))
++ continue;
+
-+/*
-+ * 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;
++ adapter_rom_resources[i].start = start;
++ adapter_rom_resources[i].end = start + length - 1;
++ request_resource(&iomem_resource, &adapter_rom_resources[i]);
+
-+ 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);
++ start = adapter_rom_resources[i++].end & ~2047UL;
+ }
-+ putname(filename);
-+ return error;
+}
+
-+void set_personality_64bit(void)
++#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.
++ */
++static int __init setup_elfcorehdr(char *arg)
+{
-+ /* 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;
++ char *end;
++ if (!arg)
++ return -EINVAL;
++ elfcorehdr_addr = memparse(arg, &end);
++ return end > arg ? 0 : -EINVAL;
+}
++early_param("elfcorehdr", setup_elfcorehdr);
++#endif
+
-+asmlinkage long sys_fork(struct pt_regs *regs)
++#ifndef CONFIG_NUMA
++static void __init
++contig_initmem_init(unsigned long start_pfn, unsigned long end_pfn)
+{
-+ return do_fork(SIGCHLD, regs->rsp, regs, 0, NULL, NULL);
-+}
++ unsigned long bootmap_size, bootmap;
+
-+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);
-+}
++ 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);
++ e820_register_active_regions(0, start_pfn, end_pfn);
++#ifdef CONFIG_XEN
++ free_bootmem_with_active_regions(0, xen_start_info->nr_pages);
++#else
++ free_bootmem_with_active_regions(0, end_pfn);
++#endif
++ reserve_bootmem(bootmap, bootmap_size);
++}
++#endif
+
-+/*
-+ * This is trivial, and on the face of it looks like it
-+ * could equally well be done in user mode.
++#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.
+ *
-+ * 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 inline void copy_edd(void)
+{
-+ return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, regs->rsp, regs, 0,
-+ NULL, NULL);
++ 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;
+}
-+
-+unsigned long get_wchan(struct task_struct *p)
++#else
++static inline void copy_edd(void)
+{
-+ 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;
+}
++#endif
+
-+long do_arch_prctl(struct task_struct *task, int code, unsigned long addr)
-+{
-+ int ret = 0;
-+ int doit = task == current;
-+ int cpu;
++#ifndef CONFIG_XEN
++#define EBDA_ADDR_POINTER 0x40E
+
-+ 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;
-+ }
++unsigned __initdata ebda_addr;
++unsigned __initdata ebda_size;
+
-+ default:
-+ ret = -EINVAL;
-+ break;
-+ }
++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;
+
-+ return ret;
-+}
++ ebda_size = *(unsigned short *)(unsigned long)ebda_addr;
+
-+long sys_arch_prctl(int code, unsigned long addr)
-+{
-+ return do_arch_prctl(current, code, 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;
++}
++#endif
+
-+/*
-+ * 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)
++void __init setup_arch(char **cmdline_p)
+{
-+ struct pt_regs *pp, ptregs;
++ printk(KERN_INFO "Command line: %s\n", saved_command_line);
+
-+ pp = task_pt_regs(tsk);
++#ifdef CONFIG_XEN
++ /* Register a call for panic conditions. */
++ atomic_notifier_chain_register(&panic_notifier_list, &xen_panic_block);
+
-+ ptregs = *pp;
-+ ptregs.cs &= 0xffff;
-+ ptregs.ss &= 0xffff;
++ ROOT_DEV = MKDEV(RAMDISK_MAJOR,0);
++ screen_info = SCREEN_INFO;
+
-+ elf_core_copy_regs(regs, &ptregs);
-+
-+ boot_option_idle_override = 1;
-+ return 1;
-+}
++ 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;
++#else
++ ROOT_DEV = old_decode_dev(ORIG_ROOT_DEV);
++ screen_info = SCREEN_INFO;
++#endif /* !CONFIG_XEN */
+
-+unsigned long arch_align_stack(unsigned long sp)
-+{
-+ if (randomize_va_space)
-+ sp -= get_random_int() % 8192;
-+ return sp & ~0xf;
-+}
++ edid_info = EDID_INFO;
++ saved_video_mode = SAVED_VIDEO_MODE;
++ bootloader_type = LOADER_TYPE;
+
-+#ifndef CONFIG_SMP
-+void _restore_vcpu(void)
-+{
-+}
++#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
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/setup64-xen.c linux-2.6.18-xen/arch/x86_64/kernel/setup64-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/setup64-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/setup64-xen.c 2006-11-19 14:26:34.000000000 +0100
-@@ -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>
++ setup_xen_features();
++
++ HYPERVISOR_vm_assist(VMASST_CMD_enable,
++ VMASST_TYPE_writable_pagetables);
++
++ ARCH_SETUP
+#endif
+
-+char x86_boot_params[BOOT_PARAM_SIZE] __initdata = {0,};
++ setup_memory_region();
++ copy_edd();
+
-+cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE;
++ 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;
+
-+struct x8664_pda *_cpu_pda[NR_CPUS] __read_mostly;
-+EXPORT_SYMBOL(_cpu_pda);
-+struct x8664_pda boot_cpu_pda[NR_CPUS] __cacheline_aligned;
++ 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;
+
-+#ifndef CONFIG_X86_NO_IDT
-+struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table };
-+#endif
++ early_identify_cpu(&boot_cpu_data);
+
-+char boot_cpu_stack[IRQSTACKSIZE] __attribute__((section(".bss.page_aligned")));
++ strlcpy(command_line, saved_command_line, COMMAND_LINE_SIZE);
++ *cmdline_p = command_line;
+
-+unsigned long __supported_pte_mask __read_mostly = ~0UL;
-+EXPORT_SYMBOL(__supported_pte_mask);
-+static int do_not_nx __cpuinitdata = 0;
++ parse_early_param();
+
-+/* noexec=on|off
-+Control non executable mappings for 64bit processes.
++ finish_e820_parsing();
+
-+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 */
++ e820_register_active_regions(0, 0, -1UL);
++ /*
++ * partially used pages are not usable - thus
++ * we are rounding upwards:
++ */
++ end_pfn = e820_end_of_ram();
++ num_physpages = end_pfn;
+
-+int force_personality32 = 0;
++ check_efer();
+
-+/* noexec32=on|off
-+Control non executable heap for 32bit processes.
-+To control the stack too use noexec=off
++#ifndef CONFIG_XEN
++ discover_ebda();
++#endif
+
-+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);
++ init_memory_mapping(0, (end_pfn_map << PAGE_SHIFT));
+
-+/*
-+ * 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;
++ zap_low_mappings(0);
+
-+#ifdef CONFIG_HOTPLUG_CPU
-+ prefill_possible_map();
++ /* How many end-of-memory variables you have, grandma! */
++ max_low_pfn = end_pfn;
++ max_pfn = end_pfn;
++ high_memory = (void *)__va(end_pfn * PAGE_SIZE - 1) + 1;
++
++ /* Remove active ranges so rediscovery with NUMA-awareness happens */
++ remove_all_active_ranges();
++
++#ifdef CONFIG_ACPI_NUMA
++ /*
++ * Parse SRAT to discover nodes.
++ */
++ acpi_numa_init();
+#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;
++#ifdef CONFIG_NUMA
++ numa_initmem_init(0, end_pfn);
++#else
++ contig_initmem_init(0, end_pfn);
+#endif
+
-+ for_each_cpu_mask (i, cpu_possible_map) {
-+ char *ptr;
++ /* Reserve direct mapping */
++ reserve_bootmem_generic(table_start << PAGE_SHIFT,
++ (table_end - table_start) << PAGE_SHIFT);
+
-+ 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);
-+ }
-+}
++ /* reserve kernel */
++ reserve_bootmem_generic(__pa_symbol(&_text),
++ __pa_symbol(&_end) - __pa_symbol(&_text));
+
+#ifdef CONFIG_XEN
-+static void switch_pt(void)
-+{
-+ xen_pt_switch(__pa(init_level4_pgt));
-+ xen_new_user_pt(__pa(init_level4_user_pgt));
-+}
-+
-+void __cpuinit cpu_gdt_init(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 /
-+ sizeof (struct desc_struct)))
-+ BUG();
-+}
++ /* reserve physmap, start info and initial page tables */
++ reserve_bootmem(__pa_symbol(&_end), (table_start<<PAGE_SHIFT)-__pa_symbol(&_end));
+#else
-+static void switch_pt(void)
-+{
-+ asm volatile("movq %0,%%cr3" :: "r" (__pa_symbol(&init_level4_pgt)));
-+}
++ /*
++ * 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);
+
-+void __init cpu_gdt_init(struct desc_ptr *gdt_descr)
-+{
-+ asm volatile("lgdt %0" :: "m" (*gdt_descr));
-+ asm volatile("lidt %0" :: "m" (idt_descr));
-+}
++ /* reserve ebda region */
++ if (ebda_addr)
++ reserve_bootmem_generic(ebda_addr, ebda_size);
+#endif
+
-+void pda_init(int cpu)
-+{
-+ struct x8664_pda *pda = cpu_pda(cpu);
++#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);
+
-+ /* 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);
++ /* Reserve SMP trampoline */
++ reserve_bootmem_generic(SMP_TRAMPOLINE_BASE, PAGE_SIZE);
+#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_ACPI_SLEEP
++ /*
++ * Reserve low memory region for sleep support.
++ */
++ acpi_reserve_bootmem();
++#endif
+#ifdef CONFIG_XEN
-+ xen_init_pt();
++#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
-+ /* 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);
++#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 + PAGE_OFFSET;
++ 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
++ if (crashk_res.start != crashk_res.end) {
++ reserve_bootmem_generic(crashk_res.start,
++ crashk_res.end - crashk_res.start + 1);
+ }
++#endif
+
-+ switch_pt();
++ paging_init();
++ /*
++ * Find and reserve possible boot-time SMP configuration:
++ */
++ find_smp_config();
++#ifdef CONFIG_XEN
++ {
++ int i, j, k, fpp;
+
-+ pda->irqstackptr += IRQSTACKSIZE-64;
-+}
++ 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,
++ xen_start_info->nr_pages * sizeof(unsigned long));
++ free_bootmem(
++ __pa(xen_start_info->mfn_list),
++ PFN_PHYS(PFN_UP(xen_start_info->nr_pages *
++ sizeof(unsigned long))));
+
-+#ifndef CONFIG_X86_NO_TSS
-+char boot_exception_stacks[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + DEBUG_STKSZ]
-+__attribute__((section(".bss.page_aligned")));
++ /*
++ * 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);
++ HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list =
++ virt_to_mfn(pfn_to_mfn_frame_list_list);
++
++ 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;
++ }
++
++ }
++
++ if (is_initial_xendomain())
++ dmi_scan_machine();
++
++#ifdef CONFIG_ACPI
++ if (!is_initial_xendomain()) {
++ acpi_disabled = 1;
++ acpi_ht = 0;
++ }
++#endif
+#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);
++#ifdef CONFIG_PCI
++ early_quirks();
++#endif
++#endif
+
-+ /* Flags to clear on syscall */
-+ wrmsrl(MSR_SYSCALL_MASK, EF_TF|EF_DF|EF_IE|0x3000);
++ /*
++ * 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
-+#ifdef CONFIG_IA32_EMULATION
-+ syscall32_cpu_init ();
++
++ init_cpu_to_node();
++
++ /*
++ * get boot-time SMP configuration:
++ */
++ if (smp_found_config)
++ get_smp_config();
++#ifndef CONFIG_XEN
++ init_apic_mappings();
++#endif
++#if defined(CONFIG_XEN) && defined(CONFIG_SMP) && !defined(CONFIG_HOTPLUG_CPU)
++ prefill_possible_map();
+#endif
-+}
+
-+void __cpuinit check_efer(void)
-+{
-+ unsigned long efer;
++ /*
++ * 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;
+
-+ rdmsrl(MSR_EFER, efer);
-+ if (!(efer & EFER_NX) || do_not_nx) {
-+ __supported_pte_mask &= ~_PAGE_NX;
-+ }
-+}
++ memmap.nr_entries = E820MAX;
++ set_xen_guest_handle(memmap.buffer, machine_e820.map);
+
-+/*
-+ * 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;
++ if (HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap))
++ BUG();
++ machine_e820.nr_map = memmap.nr_entries;
++
++ e820_reserve_resources(machine_e820.map, machine_e820.nr_map);
++ }
++#else
++ e820_reserve_resources(e820.map, e820.nr_map);
++#endif
++ e820_mark_nosave_regions();
++
++ 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 < ARRAY_SIZE(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;
++ 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
-+ 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
++ } else {
++#if defined(CONFIG_VT) && defined(CONFIG_DUMMY_CONSOLE)
++ conswitchp = &dummy_con;
++#endif
++ }
++ }
++#else /* CONFIG_XEN */
+
-+ me = current;
++#ifdef CONFIG_VT
++#if defined(CONFIG_VGA_CONSOLE)
++ conswitchp = &vga_con;
++#elif defined(CONFIG_DUMMY_CONSOLE)
++ conswitchp = &dummy_con;
++#endif
++#endif
+
-+ if (cpu_test_and_set(cpu, cpu_initialized))
-+ panic("CPU#%d already initialized!\n", cpu);
++#endif /* !CONFIG_XEN */
++}
+
-+ printk("Initializing CPU#%d\n", cpu);
++#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 */
+
-+ 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
++static int __cpuinit get_model_name(struct cpuinfo_x86 *c)
++{
++ unsigned int *v;
+
-+ cpu_gdt_descr[cpu].size = GDT_SIZE;
-+ cpu_gdt_init(&cpu_gdt_descr[cpu]);
++ if (c->extended_cpuid_level < 0x80000004)
++ return 0;
+
-+ memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8);
-+ syscall_init();
++ 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;
++}
+
-+ wrmsrl(MSR_FS_BASE, 0);
-+ wrmsrl(MSR_KERNEL_GS_BASE, 0);
-+ barrier();
+
-+ check_efer();
++static void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c)
++{
++ unsigned int n, dummy, eax, ebx, ecx, edx;
+
-+#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
-+ };
++ n = c->extended_cpuid_level;
+
-+ 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;
++ 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;
+ }
+
-+ 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);
++ 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);
+
-+ /*
-+ * Clear all 6 debug registers:
-+ */
++ printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n",
++ c->x86_cache_size, ecx & 0xFF);
++ }
+
-+ set_debugreg(0UL, 0);
-+ set_debugreg(0UL, 1);
-+ set_debugreg(0UL, 2);
-+ set_debugreg(0UL, 3);
-+ set_debugreg(0UL, 6);
-+ set_debugreg(0UL, 7);
++ 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;
++ }
++}
+
-+ fpu_init();
++#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 */
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/setup-xen.c linux-2.6.18-xen/arch/x86_64/kernel/setup-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/setup-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/setup-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,1637 @@
-+/*
-+ * 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.
-+ */
++#endif
+
+/*
-+ * This file handles the architecture-dependent parts of initialization
++ * 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);
+
-+#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 <linux/pfn.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>
-+#include <asm/mach-xen/setup_arch_post.h>
-+#include <xen/interface/memory.h>
-+
-+extern unsigned long start_pfn;
-+extern struct edid_info edid_info;
-+EXPORT_SYMBOL_GPL(edid_info);
++ c->x86_max_cores = (ecx & 0xff) + 1;
+
-+shared_info_t *HYPERVISOR_shared_info = (shared_info_t *)empty_zero_page;
-+EXPORT_SYMBOL(HYPERVISOR_shared_info);
++ /* CPU telling us the core id bits shift? */
++ bits = (ecx >> 12) & 0xF;
+
-+extern char hypercall_page[PAGE_SIZE];
-+EXPORT_SYMBOL(hypercall_page);
++ /* Otherwise recompute */
++ if (bits == 0) {
++ while ((1 << bits) < c->x86_max_cores)
++ bits++;
++ }
+
-+/* Allows setting of maximum possible memory size */
-+unsigned long xen_override_max_pfn;
++ /* 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);
+
-+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 */
-+};
++#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 long *phys_to_machine_mapping;
-+unsigned long *pfn_to_mfn_frame_list_list, *pfn_to_mfn_frame_list[512];
++ printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node);
++#endif
++#endif
++}
+
-+EXPORT_SYMBOL(phys_to_machine_mapping);
++static void __cpuinit init_amd(struct cpuinfo_x86 *c)
++{
++ unsigned level;
+
-+DEFINE_PER_CPU(multicall_entry_t, multicall_list[8]);
-+DEFINE_PER_CPU(int, nr_multicall_ents);
++#ifdef CONFIG_SMP
++ unsigned long value;
+
-+/* Raw start-of-day parameters from the hypervisor. */
-+start_info_t *xen_start_info;
-+EXPORT_SYMBOL(xen_start_info);
++ /*
++ * 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
+
-+/*
-+ * Machine setup..
-+ */
++ /* 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);
+
-+struct cpuinfo_x86 boot_cpu_data __read_mostly;
-+EXPORT_SYMBOL(boot_cpu_data);
++ /* Enable workaround for FXSAVE leak */
++ if (c->x86 >= 6)
++ set_bit(X86_FEATURE_FXSAVE_LEAK, &c->x86_capability);
+
-+unsigned long mmu_cr4_features;
++ 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);
+
-+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
++ /* 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);
+
-+int acpi_numa __initdata;
++ /* Multi core CPU? */
++ if (c->extended_cpuid_level >= 0x80000008)
++ amd_detect_cmp(c);
+
-+/* Boot loader ID as an integer, for the benefit of proc_dointvec */
-+int bootloader_type;
++ /* Fix cpuid4 emulation for more */
++ num_cache_leaves = 3;
+
-+unsigned long saved_video_mode;
++ /* RDTSC can be speculated around */
++ clear_bit(X86_FEATURE_SYNC_RDTSC, &c->x86_capability);
++}
+
-+/*
-+ * Early DMI memory
-+ */
-+int dmi_alloc_index;
-+char dmi_alloc_data[DMI_MAX_DATA];
++static void __cpuinit detect_ht(struct cpuinfo_x86 *c)
++{
++#ifdef CONFIG_SMP
++ u32 eax, ebx, ecx, edx;
++ int index_msb, core_bits;
+
-+/*
-+ * Setup options
-+ */
-+struct screen_info screen_info;
-+EXPORT_SYMBOL(screen_info);
-+struct sys_desc_table_struct {
-+ unsigned short length;
-+ unsigned char table[0];
-+};
++ cpuid(1, &eax, &ebx, &ecx, &edx);
+
-+struct edid_info edid_info;
-+struct e820map e820;
-+#ifdef CONFIG_XEN
-+struct e820map machine_e820;
-+#endif
+
-+extern int root_mountflags;
++ if (!cpu_has(c, X86_FEATURE_HT))
++ return;
++ if (cpu_has(c, X86_FEATURE_CMP_LEGACY))
++ goto out;
+
-+char command_line[COMMAND_LINE_SIZE];
++ smp_num_siblings = (ebx & 0xff0000) >> 16;
+
-+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 }
-+};
++ if (smp_num_siblings == 1) {
++ printk(KERN_INFO "CPU: Hyper-Threading is disabled\n");
++ } else if (smp_num_siblings > 1 ) {
+
-+#define STANDARD_IO_RESOURCES \
-+ (sizeof standard_io_resources / sizeof standard_io_resources[0])
++ if (smp_num_siblings > NR_CPUS) {
++ printk(KERN_WARNING "CPU: Unsupported number of the siblings %d", smp_num_siblings);
++ smp_num_siblings = 1;
++ return;
++ }
+
-+#define IORESOURCE_RAM (IORESOURCE_BUSY | IORESOURCE_MEM)
++ index_msb = get_count_order(smp_num_siblings);
++ c->phys_proc_id = phys_pkg_id(index_msb);
+
-+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,
-+};
++ smp_num_siblings = smp_num_siblings / c->x86_max_cores;
+
-+#define IORESOURCE_ROM (IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM)
++ index_msb = get_count_order(smp_num_siblings) ;
+
-+static struct resource system_rom_resource = {
-+ .name = "System ROM",
-+ .start = 0xf0000,
-+ .end = 0xfffff,
-+ .flags = IORESOURCE_ROM,
-+};
++ core_bits = get_count_order(c->x86_max_cores);
+
-+static struct resource extension_rom_resource = {
-+ .name = "Extension ROM",
-+ .start = 0xe0000,
-+ .end = 0xeffff,
-+ .flags = IORESOURCE_ROM,
-+};
++ 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);
++ }
+
-+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 }
-+};
++#endif
++}
+
-+#define ADAPTER_ROM_RESOURCES \
-+ (sizeof adapter_rom_resources / sizeof adapter_rom_resources[0])
++/*
++ * 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;
+
-+static struct resource video_rom_resource = {
-+ .name = "Video ROM",
-+ .start = 0xc0000,
-+ .end = 0xc7fff,
-+ .flags = IORESOURCE_ROM,
-+};
++ if (c->cpuid_level < 4)
++ return 1;
+
-+static struct resource video_ram_resource = {
-+ .name = "Video RAM area",
-+ .start = 0xa0000,
-+ .end = 0xbffff,
-+ .flags = IORESOURCE_RAM,
-+};
++ cpuid_count(4, 0, &eax, &t, &t, &t);
+
-+#define romsignature(x) (*(unsigned short *)(x) == 0xaa55)
++ if (eax & 0x1f)
++ return ((eax >> 26) + 1);
++ else
++ return 1;
++}
+
-+static int __init romchecksum(unsigned char *rom, unsigned long length)
++static void srat_detect_node(void)
+{
-+ unsigned char *p, sum = 0;
++#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);
+
-+ for (p = rom; p < rom + length; p++)
-+ sum += *p;
-+ return sum == 0;
++ printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node);
++#endif
+}
+
-+static void __init probe_roms(void)
++static void __cpuinit init_intel(struct cpuinfo_x86 *c)
+{
-+ 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
++ /* Cache sizes */
++ unsigned n;
+
-+ /* 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;
++ 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);
++ }
+
-+ video_rom_resource.start = start;
++ if (cpu_has_ds) {
++ unsigned int l1, l2;
++ rdmsr(MSR_IA32_MISC_ENABLE, l1, l2);
++ if (!(l1 & (1<<11)))
++ set_bit(X86_FEATURE_BTS, c->x86_capability);
++ if (!(l1 & (1<<12)))
++ set_bit(X86_FEATURE_PEBS, c->x86_capability);
++ }
+
-+ /* 0 < length <= 0x7f * 512, historically */
-+ length = rom[2] * 512;
++ 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 checksum okay, trust length byte */
-+ if (length && romchecksum(rom, length))
-+ video_rom_resource.end = start + length - 1;
++ 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);
++ if (c->x86 == 6)
++ set_bit(X86_FEATURE_REP_GOOD, &c->x86_capability);
++ if (c->x86 == 15)
++ set_bit(X86_FEATURE_SYNC_RDTSC, &c->x86_capability);
++ else
++ clear_bit(X86_FEATURE_SYNC_RDTSC, &c->x86_capability);
++ c->x86_max_cores = intel_num_cpu_cores(c);
+
-+ request_resource(&iomem_resource, &video_rom_resource);
-+ break;
-+ }
++ srat_detect_node();
++}
+
-+ start = (video_rom_resource.end + 1 + 2047) & ~2047UL;
-+ if (start < upper)
-+ start = upper;
++static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c)
++{
++ char *v = c->x86_vendor_id;
+
-+ /* system rom */
-+ request_resource(&iomem_resource, &system_rom_resource);
-+ upper = system_rom_resource.start;
++ 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;
++}
+
-+ /* 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;
-+ }
-+ }
++struct cpu_model_info {
++ int vendor;
++ int family;
++ char *model_names[16];
++};
+
-+ /* 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;
++/* 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;
+
-+ /* 0 < length <= 0x7f * 512, historically */
-+ length = rom[2] * 512;
++ 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);
+
-+ /* but accept any length that fits if checksum okay */
-+ if (!length || start + length > upper || !romchecksum(rom, length))
-+ continue;
++ /* 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);
+
-+ adapter_rom_resources[i].start = start;
-+ adapter_rom_resources[i].end = start + length - 1;
-+ request_resource(&iomem_resource, &adapter_rom_resources[i]);
++ /* Initialize the standard set of capabilities */
++ /* Note that the vendor-specific code below might override */
+
-+ start = adapter_rom_resources[i++].end & ~2047UL;
++ /* 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;
+ }
-+}
+
-+/* 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]));
++#ifdef CONFIG_SMP
++ c->phys_proc_id = (cpuid_ebx(1) >> 24) & 0xff;
++#endif
+}
+
-+static __init void parse_cmdline_early (char ** cmdline_p)
++/*
++ * This does the hard work of actually picking apart the CPU stuff...
++ */
++void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
+{
-+ char c = ' ', *to = command_line, *from = COMMAND_LINE;
-+ int len = 0;
-+ int userdef = 0;
-+
-+ for (;;) {
-+ if (c != ' ')
-+ goto next_char;
++ int i;
++ u32 xlvl;
+
-+#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;
++ early_identify_cpu(c);
+
-+ 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;
++ /* 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);
+ }
-+#ifdef CONFIG_X86_IO_APIC
-+ else if (fullarg(from, "acpi_skip_timer_override"))
-+ acpi_skip_timer_override = 1;
-+#endif
-+#endif
++ if (xlvl >= 0x80000004)
++ get_model_name(c); /* Default name */
++ }
+
-+#ifndef CONFIG_XEN
-+ if (fullarg(from, "disable_timer_pin_1"))
-+ disable_timer_pin_1 = 1;
-+ if (fullarg(from, "enable_timer_pin_1"))
-+ disable_timer_pin_1 = -1;
-+
-+ 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);
++ /* 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);
++ }
+
-+ 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;
-+ }
-+ }
++ c->apicid = phys_pkg_id(0);
+
-+#ifdef CONFIG_NUMA
-+ if (!memcmp(from, "numa=", 5))
-+ numa_setup(from+5);
-+#endif
++ /*
++ * 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;
+
-+ if (!memcmp(from,"iommu=",6)) {
-+ iommu_setup(from+6);
-+ }
++ case X86_VENDOR_INTEL:
++ init_intel(c);
++ break;
+
-+ if (fullarg(from,"oops=panic"))
-+ panic_on_oops = 1;
++ case X86_VENDOR_UNKNOWN:
++ default:
++ display_cacheinfo(c);
++ break;
++ }
+
-+ if (!memcmp(from, "noexec=", 7))
-+ nonx_setup(from + 7);
++ select_idle_routine(c);
++ detect_ht(c);
+
-+#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)) {
-+ 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;
-+ }
-+ }
-+#endif
++ /*
++ * 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_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);
++#ifdef CONFIG_X86_MCE
++ mcheck_init(c);
+#endif
-+
-+#if defined(CONFIG_HOTPLUG_CPU) && !defined(CONFIG_XEN)
-+ else if (!memcmp(from, "additional_cpus=", 16))
-+ setup_additional_cpus(from+16);
++ if (c == &boot_cpu_data)
++ mtrr_bp_init();
++ else
++ mtrr_ap_init();
++#ifdef CONFIG_NUMA
++ numa_add_cpu(smp_processor_id());
+#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)
++void __cpuinit print_cpu_info(struct cpuinfo_x86 *c)
+{
-+ unsigned long bootmap_size, bootmap;
++ if (c->x86_model_id[0])
++ printk("%s", c->x86_model_id);
+
-+ 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 (c->x86_mask || c->cpuid_level >= 0)
++ printk(" stepping %02x\n", c->x86_mask);
++ else
++ printk("\n");
++}
+
-+#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.
-+ *
++/*
++ * Get CPU information for use by the procfs.
+ */
-+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)
++
++static int show_cpuinfo(struct seq_file *m, void *v)
+{
-+}
-+#endif
++ struct cpuinfo_x86 *c = v;
+
-+#ifndef CONFIG_XEN
-+#define EBDA_ADDR_POINTER 0x40E
++ /*
++ * 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,
+
-+unsigned __initdata ebda_addr;
-+unsigned __initdata ebda_size;
++ /* 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",
+
-+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;
++ /* 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,
+
-+ ebda_size = *(unsigned short *)(unsigned long)ebda_addr;
++ /* 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,
+
-+ /* 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;
-+}
++ /* Intel-defined (#2) */
++ "pni", NULL, NULL, "monitor", "ds_cpl", "vmx", "smx", "est",
++ "tm2", "ssse3", "cid", NULL, NULL, "cx16", "xtpr", NULL,
++ NULL, NULL, "dca", 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
+
-+void __init setup_arch(char **cmdline_p)
-+{
-+ struct xen_memory_map memmap;
++ 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));
++ }
+
-+#ifdef CONFIG_XEN
-+ /* Register a call for panic conditions. */
-+ atomic_notifier_chain_register(&panic_notifier_list, &xen_panic_block);
++ /* 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
+
-+ ROOT_DEV = MKDEV(RAMDISK_MAJOR,0);
-+ screen_info = SCREEN_INFO;
++ 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);
+
-+ 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;
++ {
++ 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);
+
-+ edid_info = EDID_INFO;
-+ saved_video_mode = SAVED_VIDEO_MODE;
-+ bootloader_type = LOADER_TYPE;
++ 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);
+
-+#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);
++ 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");
+
-+#endif
++ return 0;
++}
+
-+ setup_xen_features();
++static void *c_start(struct seq_file *m, loff_t *pos)
++{
++ return *pos < NR_CPUS ? cpu_data + *pos : NULL;
++}
+
-+ HYPERVISOR_vm_assist(VMASST_CMD_enable,
-+ VMASST_TYPE_writable_pagetables);
++static void *c_next(struct seq_file *m, void *v, loff_t *pos)
++{
++ ++*pos;
++ return c_start(m, pos);
++}
+
-+ 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;
++static void c_stop(struct seq_file *m, void *v)
++{
++}
+
-+#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();
++struct seq_operations cpuinfo_op = {
++ .start =c_start,
++ .next = c_next,
++ .stop = c_stop,
++ .show = show_cpuinfo,
++};
+
-+ 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;
++#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;
+
-+ 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;
++ pd = platform_device_alloc("pcspkr", -1);
++ if (!pd)
++ return -ENOMEM;
+
-+ parse_cmdline_early(cmdline_p);
++ ret = platform_device_add(pd);
++ if (ret)
++ platform_device_put(pd);
+
-+ early_identify_cpu(&boot_cpu_data);
++ return ret;
++}
++device_initcall(add_pcspkr);
++#endif
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/smp-xen.c b/arch/x86_64/kernel/smp-xen.c
+--- a/arch/x86_64/kernel/smp-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/smp-xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,576 @@
++/*
++ * 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.
++ */
+
-+ /*
-+ * partially used pages are not usable - thus
-+ * we are rounding upwards:
-+ */
-+ end_pfn = e820_end_of_ram();
-+ num_physpages = end_pfn; /* for pfn_valid */
++#include <linux/init.h>
+
-+ check_efer();
++#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>
+
-+#ifndef CONFIG_XEN
-+ discover_ebda();
++#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
+
-+ init_memory_mapping(0, (end_pfn_map << PAGE_SHIFT));
++#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.
++ */
+
-+#ifdef CONFIG_ACPI_NUMA
-+ /*
-+ * Parse SRAT to discover nodes.
-+ */
-+ acpi_numa_init();
-+#endif
++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;
+
-+#ifdef CONFIG_NUMA
-+ numa_initmem_init(0, end_pfn);
-+#else
-+ contig_initmem_init(0, end_pfn);
++/* 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
+
-+ /* Reserve direct mapping */
-+ reserve_bootmem_generic(table_start << PAGE_SHIFT,
-+ (table_end - table_start) << PAGE_SHIFT);
++/*
++ * We cannot call mmdrop() because we are in interrupt context,
++ * instead update mm->cpu_vm_mask.
++ */
++static inline void leave_mm(int cpu)
++{
++ if (read_pda(mmu_state) == TLBSTATE_OK)
++ BUG();
++ cpu_clear(cpu, read_pda(active_mm)->cpu_vm_mask);
++ load_cr3(swapper_pg_dir);
++}
+
-+ /* reserve kernel */
-+ reserve_bootmem_generic(__pa_symbol(&_text),
-+ __pa_symbol(&_end) - __pa_symbol(&_text));
++#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.
++ */
+
-+#ifdef CONFIG_XEN
-+ /* reserve physmap, start info and initial page tables */
-+ reserve_bootmem(__pa_symbol(&_end), (table_start<<PAGE_SHIFT)-__pa_symbol(&_end));
-+#else
-+ /*
-+ * 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);
++/*
++ * 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.
++ */
+
-+ /* reserve ebda region */
-+ if (ebda_addr)
-+ reserve_bootmem_generic(ebda_addr, ebda_size);
-+#endif
++asmlinkage void smp_invalidate_interrupt(struct pt_regs *regs)
++{
++ int cpu;
++ int sender;
++ union smp_flush_state *f;
+
-+#ifdef CONFIG_SMP
++ cpu = smp_processor_id();
+ /*
-+ * 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)
++ * orig_rax contains the interrupt vector - 256.
++ * Use that to determine where the sender put the data.
+ */
-+ reserve_bootmem_generic(PAGE_SIZE, PAGE_SIZE);
-+
-+ /* Reserve SMP trampoline */
-+ reserve_bootmem_generic(SMP_TRAMPOLINE_BASE, PAGE_SIZE);
-+#endif
++ sender = regs->orig_rax + 256 - INVALIDATE_TLB_VECTOR_START;
++ f = &per_cpu(flush_state, sender);
+
-+#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
-+ if (crashk_res.start != crashk_res.end) {
-+ reserve_bootmem_generic(crashk_res.start,
-+ crashk_res.end - crashk_res.start + 1);
++ 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);
+ }
-+#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;
-+
-+ 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,
-+ xen_start_info->nr_pages * sizeof(unsigned long));
-+ free_bootmem(
-+ __pa(xen_start_info->mfn_list),
-+ PFN_PHYS(PFN_UP(xen_start_info->nr_pages *
-+ sizeof(unsigned long))));
++out:
++ ack_APIC_irq();
++ cpu_clear(cpu, f->flush_cpumask);
++}
+
-+ /*
-+ * 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);
-+ HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list =
-+ virt_to_mfn(pfn_to_mfn_frame_list_list);
++static void flush_tlb_others(cpumask_t cpumask, struct mm_struct *mm,
++ unsigned long va)
++{
++ int sender;
++ union smp_flush_state *f;
+
-+ 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;
-+ }
++ /* 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);
+
-+ if (is_initial_xendomain())
-+ dmi_scan_machine();
++ f->flush_mm = mm;
++ f->flush_va = va;
++ cpus_or(f->flush_cpumask, cpumask, f->flush_cpumask);
+
-+ if (!is_initial_xendomain()) {
-+ acpi_disabled = 1;
-+#ifdef CONFIG_ACPI
-+ acpi_ht = 0;
-+#endif
-+ }
-+#endif
++ /*
++ * We have to send the IPI only to
++ * CPUs affected.
++ */
++ send_IPI_mask(cpumask, INVALIDATE_TLB_VECTOR_START + sender);
+
-+#ifndef CONFIG_XEN
-+ check_ioapic();
-+#endif
++ while (!cpus_empty(f->flush_cpumask))
++ cpu_relax();
+
-+ zap_low_mappings(0);
++ f->flush_mm = NULL;
++ f->flush_va = 0;
++ spin_unlock(&f->tlbstate_lock);
++}
+
-+ /*
-+ * 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();
++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;
++}
+
-+ /*
-+ * Read APIC and some other early information from ACPI tables.
-+ */
-+ acpi_boot_init();
-+#endif
++core_initcall(init_smp_flush);
++
++void flush_tlb_current_task(void)
++{
++ struct mm_struct *mm = current->mm;
++ cpumask_t cpu_mask;
+
-+ init_cpu_to_node();
++ preempt_disable();
++ cpu_mask = mm->cpu_vm_mask;
++ cpu_clear(smp_processor_id(), cpu_mask);
+
-+#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
++ local_flush_tlb();
++ if (!cpus_empty(cpu_mask))
++ flush_tlb_others(cpu_mask, mm, FLUSH_ALL);
++ preempt_enable();
++}
++EXPORT_SYMBOL(flush_tlb_current_task);
+
-+ /*
-+ * 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()) {
-+ memmap.nr_entries = E820MAX;
-+ set_xen_guest_handle(memmap.buffer, machine_e820.map);
++void flush_tlb_mm (struct mm_struct * mm)
++{
++ cpumask_t cpu_mask;
+
-+ if (HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap))
-+ BUG();
-+ machine_e820.nr_map = memmap.nr_entries;
++ preempt_disable();
++ cpu_mask = mm->cpu_vm_mask;
++ cpu_clear(smp_processor_id(), cpu_mask);
+
-+ e820_reserve_resources(machine_e820.map, machine_e820.nr_map);
++ if (current->active_mm == mm) {
++ if (current->mm)
++ local_flush_tlb();
++ else
++ leave_mm(smp_processor_id());
+ }
-+#else
-+ e820_reserve_resources(e820.map, e820.nr_map);
-+#endif
++ if (!cpus_empty(cpu_mask))
++ flush_tlb_others(cpu_mask, mm, FLUSH_ALL);
+
-+ request_resource(&iomem_resource, &video_ram_resource);
++ preempt_enable();
++}
++EXPORT_SYMBOL(flush_tlb_mm);
+
-+ {
-+ 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]);
-+ }
++void flush_tlb_page(struct vm_area_struct * vma, unsigned long va)
++{
++ struct mm_struct *mm = vma->vm_mm;
++ cpumask_t cpu_mask;
+
-+#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
++ preempt_disable();
++ cpu_mask = mm->cpu_vm_mask;
++ cpu_clear(smp_processor_id(), cpu_mask);
+
-+#ifdef CONFIG_XEN
-+ {
-+ struct physdev_set_iopl set_iopl;
++ if (current->active_mm == mm) {
++ if(current->mm)
++ __flush_tlb_one(va);
++ else
++ leave_mm(smp_processor_id());
++ }
+
-+ set_iopl.iopl = 1;
-+ HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
++ if (!cpus_empty(cpu_mask))
++ flush_tlb_others(cpu_mask, mm, va);
+
-+ 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 {
-+ extern int console_use_vt;
-+ console_use_vt = 0;
-+ }
-+ }
-+#else /* CONFIG_XEN */
++ preempt_enable();
++}
++EXPORT_SYMBOL(flush_tlb_page);
+
-+#ifdef CONFIG_VT
-+#if defined(CONFIG_VGA_CONSOLE)
-+ conswitchp = &vga_con;
-+#elif defined(CONFIG_DUMMY_CONSOLE)
-+ conswitchp = &dummy_con;
-+#endif
-+#endif
++static void do_flush_tlb_all(void* info)
++{
++ unsigned long cpu = smp_processor_id();
+
-+#endif /* !CONFIG_XEN */
++ __flush_tlb_all();
++ if (read_pda(mmu_state) == TLBSTATE_LAZY)
++ leave_mm(cpu);
+}
+
-+#ifdef CONFIG_XEN
-+static int
-+xen_panic_event(struct notifier_block *this, unsigned long event, void *ptr)
++void flush_tlb_all(void)
+{
-+ HYPERVISOR_shutdown(SHUTDOWN_crash);
-+ /* we're never actually going to get here... */
-+ return NOTIFY_DONE;
++ on_each_cpu(do_flush_tlb_all, NULL, 1, 1);
+}
-+#endif /* !CONFIG_XEN */
++#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 ...
++ */
+
-+static int __cpuinit get_model_name(struct cpuinfo_x86 *c)
++void smp_send_reschedule(int cpu)
+{
-+ unsigned int *v;
++ send_IPI_mask(cpumask_of_cpu(cpu), RESCHEDULE_VECTOR);
++}
+
-+ if (c->extended_cpuid_level < 0x80000004)
-+ return 0;
++/*
++ * Structure and data for smp_call_function(). This is designed to minimise
++ * static memory requirements. It also looks cleaner.
++ */
++static DEFINE_SPINLOCK(call_lock);
+
-+ 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;
++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);
++}
+
-+static void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c)
++/*
++ * 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)
+{
-+ unsigned int n, dummy, eax, ebx, ecx, edx;
++ struct call_data_struct data;
++ int cpus = 1;
+
-+ n = c->extended_cpuid_level;
++ data.func = func;
++ data.info = info;
++ atomic_set(&data.started, 0);
++ data.wait = wait;
++ if (wait)
++ atomic_set(&data.finished, 0);
+
-+ 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;
-+ }
++ 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);
+
-+ 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);
++ /* Wait for response */
++ while (atomic_read(&data.started) != cpus)
++ cpu_relax();
+
-+ printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n",
-+ c->x86_cache_size, ecx & 0xFF);
-+ }
++ if (!wait)
++ return;
+
-+ 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;
-+ }
++ while (atomic_read(&data.finished) != cpus)
++ cpu_relax();
+}
+
-+#ifdef CONFIG_NUMA
-+static int nearby_node(int apicid)
++/*
++ * 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)
+{
-+ 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;
++ /* prevent preemption and reschedule on another processor */
++ int me = get_cpu();
++ if (cpu == me) {
++ put_cpu();
++ return 0;
+ }
-+ return first_node(node_online_map); /* Shouldn't happen */
++
++ /* Can deadlock when called with interrupts disabled */
++ WARN_ON(irqs_disabled());
++
++ spin_lock_bh(&call_lock);
++ __smp_call_function_single(cpu, func, info, nonatomic, wait);
++ spin_unlock_bh(&call_lock);
++ put_cpu();
++ return 0;
+}
-+#endif
++EXPORT_SYMBOL(smp_call_function_single);
+
+/*
-+ * 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.
++ * this function sends a 'generic call function' IPI to all other CPUs
++ * in the system.
+ */
-+static void __init amd_detect_cmp(struct cpuinfo_x86 *c)
++static void __smp_call_function (void (*func) (void *info), void *info,
++ int nonatomic, int wait)
+{
-+#ifdef CONFIG_SMP
-+ unsigned bits;
-+#ifdef CONFIG_NUMA
-+ int cpu = smp_processor_id();
-+ int node = 0;
-+ unsigned apicid = hard_smp_processor_id();
++ 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)
++#ifndef CONFIG_XEN
++ cpu_relax();
++#else
++ barrier();
+#endif
-+ unsigned ecx = cpuid_ecx(0x80000008);
+
-+ c->x86_max_cores = (ecx & 0xff) + 1;
++ if (!wait)
++ return;
+
-+ /* CPU telling us the core id bits shift? */
-+ bits = (ecx >> 12) & 0xF;
++ while (atomic_read(&data.finished) != cpus)
++#ifndef CONFIG_XEN
++ cpu_relax();
++#else
++ barrier();
++#endif
++}
+
-+ /* Otherwise recompute */
-+ if (bits == 0) {
-+ while ((1 << bits) < c->x86_max_cores)
-+ bits++;
-+ }
++/*
++ * 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);
+
-+ /* 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);
++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);
++}
+
-+#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);
++static void smp_really_stop_cpu(void *dummy)
++{
++ smp_stop_cpu();
++ for (;;)
++ halt();
++}
+
-+ printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node);
++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();
++#ifndef CONFIG_XEN
++ disable_local_APIC();
+#endif
++ local_irq_enable();
+}
+
-+static void __init init_amd(struct cpuinfo_x86 *c)
++/*
++ * 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
+{
-+ unsigned level;
++#ifndef CONFIG_XEN
++ ack_APIC_irq();
++#else
++ return IRQ_HANDLED;
++#endif
++}
+
-+#ifdef CONFIG_SMP
-+ unsigned long value;
++#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
+ /*
-+ * 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)
++ * Notify initiating CPU that I've grabbed the data and am
++ * about to execute the function
+ */
-+ if (c->x86 == 15) {
-+ rdmsrl(MSR_K8_HWCR, value);
-+ value |= 1 << 6;
-+ wrmsrl(MSR_K8_HWCR, value);
++ 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
++}
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/time-xen.c b/arch/x86_64/kernel/time-xen.c
+--- a/arch/x86_64/kernel/time-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/time-xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,1057 @@
++/*
++ * 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).
++ */
+
-+ /* 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);
++#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>
+
-+ /* Enable workaround for FXSAVE leak */
-+ if (c->x86 >= 6)
-+ set_bit(X86_FEATURE_FXSAVE_LEAK, &c->x86_capability);
++#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>
+
-+ 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);
++#include "mach_time.h"
+
-+ /* 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);
++#include <linux/timex.h>
+
-+ /* Multi core CPU? */
-+ if (c->extended_cpuid_level >= 0x80000008)
-+ amd_detect_cmp(c);
++#include <asm/hpet.h>
+
-+ /* Fix cpuid4 emulation for more */
-+ num_cache_leaves = 3;
-+}
++#include <asm/arch_hooks.h>
+
-+static void __cpuinit detect_ht(struct cpuinfo_x86 *c)
-+{
-+#ifdef CONFIG_SMP
-+ u32 eax, ebx, ecx, edx;
-+ int index_msb, core_bits;
-+ int cpu = smp_processor_id();
++#include <xen/evtchn.h>
++#include <xen/interface/vcpu.h>
+
-+ cpuid(1, &eax, &ebx, &ecx, &edx);
++int pit_latch_buggy; /* extern */
++
++unsigned long vxtime_hz = PIT_TICK_RATE;
++struct vxtime_data __vxtime __section_vxtime; /* for vsyscalls */
++volatile unsigned long __jiffies __section_jiffies = INITIAL_JIFFIES;
++struct timespec __xtime __section_xtime;
++struct timezone __sys_tz __section_sys_tz;
++
++#define USEC_PER_TICK (USEC_PER_SEC / HZ)
++#define NSEC_PER_TICK (NSEC_PER_SEC / HZ)
++#define FSEC_PER_TICK (FSEC_PER_SEC / HZ)
+
++#define NS_SCALE 10 /* 2^10, carefully chosen */
++#define US_SCALE 32 /* 2^32, arbitralrily chosen */
+
-+ if (!cpu_has(c, X86_FEATURE_HT))
-+ return;
-+ if (cpu_has(c, X86_FEATURE_CMP_LEGACY))
-+ goto out;
++unsigned int cpu_khz; /* Detected as we calibrate the TSC */
++EXPORT_SYMBOL(cpu_khz);
+
-+ smp_num_siblings = (ebx & 0xff0000) >> 16;
++DEFINE_SPINLOCK(rtc_lock);
++EXPORT_SYMBOL(rtc_lock);
+
-+ if (smp_num_siblings == 1) {
-+ printk(KERN_INFO "CPU: Hyper-Threading is disabled\n");
-+ } else if (smp_num_siblings > 1 ) {
++extern struct init_timer_opts timer_tsc_init;
++extern struct timer_opts timer_tsc;
++#define timer_none timer_tsc
+
-+ if (smp_num_siblings > NR_CPUS) {
-+ printk(KERN_WARNING "CPU: Unsupported number of the siblings %d", smp_num_siblings);
-+ smp_num_siblings = 1;
-+ return;
-+ }
++/* 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;
+
-+ index_msb = get_count_order(smp_num_siblings);
-+ c->phys_proc_id = phys_pkg_id(index_msb);
++/* 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);
+
-+ smp_num_siblings = smp_num_siblings / c->x86_max_cores;
++/* 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);
+
-+ index_msb = get_count_order(smp_num_siblings) ;
++/* Current runstate of each CPU (updated automatically by the hypervisor). */
++static DEFINE_PER_CPU(struct vcpu_runstate_info, runstate);
+
-+ core_bits = get_count_order(c->x86_max_cores);
++/* Must be signed, as it's compared with s64 quantities which can be -ve. */
++#define NS_PER_TICK (1000000000LL/HZ)
+
-+ c->cpu_core_id = phys_pkg_id(index_msb) &
-+ ((1 << core_bits) - 1);
++static inline void __normalize_time(time_t *sec, s64 *nsec)
++{
++ while (*nsec >= NSEC_PER_SEC) {
++ (*nsec) -= NSEC_PER_SEC;
++ (*sec)++;
+ }
-+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);
++ 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("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)
++{
++ permitted_clock_jitter = simple_strtoul(str, NULL, 0);
++ return 1;
++}
++__setup("permitted_clock_jitter=", __permitted_clock_jitter);
++
++#ifndef CONFIG_X86
++int tsc_disable __devinitdata = 0;
+#endif
++
++static void delay_tsc(unsigned long loops)
++{
++ unsigned long bclock, now;
++
++ rdtscl(bclock);
++ do {
++ rep_nop();
++ rdtscl(now);
++ } while ((now - bclock) < loops);
+}
+
++struct timer_opts timer_tsc = {
++ .name = "tsc",
++ .delay = delay_tsc,
++};
++
+/*
-+ * find out the number of processor cores on the die
++ * Scale a 64-bit delta by scaling and multiplying by a 32-bit fraction,
++ * yielding a 64-bit result.
+ */
-+static int __cpuinit intel_num_cpu_cores(struct cpuinfo_x86 *c)
++static inline u64 scale_delta(u64 delta, u32 mul_frac, int shift)
+{
-+ unsigned int eax, t;
++ u64 product;
+
-+ if (c->cpuid_level < 4)
-+ return 1;
++ if (shift < 0)
++ delta >>= -shift;
++ else
++ delta <<= shift;
+
-+ cpuid_count(4, 0, &eax, &t, &t, &t);
++ __asm__ (
++ "mul %%rdx ; shrd $32,%%rdx,%%rax"
++ : "=a" (product) : "0" (delta), "d" ((u64)mul_frac) );
+
-+ if (eax & 0x1f)
-+ return ((eax >> 26) + 1);
-+ else
-+ return 1;
++ return product;
+}
+
-+static void srat_detect_node(void)
++void init_cpu_khz(void)
+{
-+#ifdef CONFIG_NUMA
-+ unsigned node;
-+ int cpu = smp_processor_id();
-+ int apicid = hard_smp_processor_id();
++ u64 __cpu_khz = 1000000ULL << US_SCALE;
++ struct vcpu_time_info *info;
++ info = &HYPERVISOR_shared_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;
++}
+
-+ /* 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);
++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);
++}
+
-+ if (acpi_numa > 0)
-+ printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node);
-+#endif
++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 void __cpuinit init_intel(struct cpuinfo_x86 *c)
++static void __update_wallclock(time_t sec, long nsec)
+{
-+ /* Cache sizes */
-+ unsigned n;
++ long wtm_nsec, xtime_nsec;
++ time_t wtm_sec, xtime_sec;
++ u64 tmp, wc_nsec;
+
-+ 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);
-+ }
++ /* Adjust wall-clock time base based on jiffies ticks. */
++ wc_nsec = processed_system_time;
++ wc_nsec += sec * (u64)NSEC_PER_SEC;
++ wc_nsec += nsec;
+
-+ 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;
-+ }
++ /* Split wallclock base into seconds and nanoseconds. */
++ tmp = wc_nsec;
++ xtime_nsec = do_div(tmp, 1000000000);
++ xtime_sec = (time_t)tmp;
+
-+ 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);
++ wtm_sec = wall_to_monotonic.tv_sec + (xtime.tv_sec - xtime_sec);
++ wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - xtime_nsec);
+
-+ srat_detect_node();
++ set_normalized_timespec(&xtime, xtime_sec, xtime_nsec);
++ set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec);
++
++ ntp_clear();
+}
+
-+static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c)
++static void update_wallclock(void)
+{
-+ char *v = c->x86_vendor_id;
++ shared_info_t *s = HYPERVISOR_shared_info;
+
-+ 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;
++ 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);
+}
+
-+struct cpu_model_info {
-+ int vendor;
-+ int family;
-+ char *model_names[16];
-+};
++/*
++ * Reads a consistent set of time-base values from Xen, into a shadow data
++ * area.
++ */
++static void get_time_values_from_xen(void)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ struct vcpu_time_info *src;
++ struct shadow_time_info *dst;
+
-+/* 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)
++ src = &s->vcpu_info[smp_processor_id()].time;
++ dst = &per_cpu(shadow_time, smp_processor_id());
++
++ 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));
++
++ dst->tsc_to_usec_mul = dst->tsc_to_nsec_mul / 1000;
++}
++
++static inline int time_values_up_to_date(int cpu)
+{
-+ u32 tfms;
++ struct vcpu_time_info *src;
++ struct shadow_time_info *dst;
+
-+ 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);
++ src = &HYPERVISOR_shared_info->vcpu_info[cpu].time;
++ dst = &per_cpu(shadow_time, cpu);
+
-+ /* 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);
++ rmb();
++ return (dst->version == src->version);
++}
+
-+ /* Initialize the standard set of capabilities */
-+ /* Note that the vendor-specific code below might override */
++/*
++ * 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);
+
-+ /* 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;
-+ }
++/* 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);
+
-+#ifdef CONFIG_SMP
-+ c->phys_proc_id = (cpuid_ebx(1) >> 24) & 0xff;
-+#endif
++void rtc_cmos_write(unsigned char val, unsigned char addr)
++{
++ 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);
+
+/*
-+ * This does the hard work of actually picking apart the CPU stuff...
++ * This version of gettimeofday has microsecond resolution
++ * and better than microsecond precision on fast x86 machines with TSC.
+ */
-+void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
++void do_gettimeofday(struct timeval *tv)
+{
-+ int i;
-+ u32 xlvl;
++ unsigned long seq;
++ unsigned long usec, sec;
++ unsigned long max_ntp_tick;
++ s64 nsec;
++ unsigned int cpu;
++ struct shadow_time_info *shadow;
++ u32 local_time_version;
+
-+ early_identify_cpu(c);
++ cpu = get_cpu();
++ shadow = &per_cpu(shadow_time, cpu);
+
-+ /* 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 */
-+ }
++ do {
++ local_time_version = shadow->version;
++ seq = read_seqbegin(&xtime_lock);
+
-+ /* 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);
-+ }
++ usec = get_usec_offset(shadow);
+
-+ c->apicid = phys_pkg_id(0);
++ /*
++ * If time_adjust is negative then NTP is slowing the clock
++ * so make sure not to go into next possible interval.
++ * Better to lose some accuracy than have time go backwards..
++ */
++ if (unlikely(time_adjust < 0)) {
++ max_ntp_tick = (USEC_PER_SEC / HZ) - tickadj;
++ usec = min(usec, max_ntp_tick);
++ }
+
-+ /*
-+ * 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;
++ sec = xtime.tv_sec;
++ usec += (xtime.tv_nsec / NSEC_PER_USEC);
+
-+ case X86_VENDOR_INTEL:
-+ init_intel(c);
-+ break;
++ nsec = shadow->system_timestamp - processed_system_time;
++ __normalize_time(&sec, &nsec);
++ usec += (long)nsec / NSEC_PER_USEC;
+
-+ case X86_VENDOR_UNKNOWN:
-+ default:
-+ display_cacheinfo(c);
-+ break;
-+ }
++ 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();
++ continue;
++ }
++ } while (read_seqretry(&xtime_lock, seq) ||
++ (local_time_version != shadow->version));
+
-+ select_idle_routine(c);
-+ detect_ht(c);
++ put_cpu();
+
-+ /*
-+ * 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];
++ while (usec >= USEC_PER_SEC) {
++ usec -= USEC_PER_SEC;
++ sec++;
+ }
+
-+#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
++ tv->tv_sec = sec;
++ tv->tv_usec = usec;
+}
-+
+
-+void __cpuinit print_cpu_info(struct cpuinfo_x86 *c)
++EXPORT_SYMBOL(do_gettimeofday);
++
++int do_settimeofday(struct timespec *tv)
+{
-+ if (c->x86_model_id[0])
-+ printk("%s", c->x86_model_id);
++ time_t sec;
++ s64 nsec;
++ unsigned int cpu;
++ struct shadow_time_info *shadow;
++ dom0_op_t op;
+
-+ if (c->x86_mask || c->cpuid_level >= 0)
-+ printk(" stepping %02x\n", c->x86_mask);
-+ else
-+ printk("\n");
-+}
++ if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC)
++ return -EINVAL;
+
-+/*
-+ * Get CPU information for use by the procfs.
-+ */
++ cpu = get_cpu();
++ shadow = &per_cpu(shadow_time, cpu);
+
-+static int show_cpuinfo(struct seq_file *m, void *v)
-+{
-+ struct cpuinfo_x86 *c = v;
++ write_seqlock_irq(&xtime_lock);
+
-+ /*
-+ * 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.
++ /*
++ * 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.
+ */
-+ 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,
++ for (;;) {
++ nsec = tv->tv_nsec - get_nsec_offset(shadow);
++ if (time_values_up_to_date(cpu))
++ break;
++ get_time_values_from_xen();
++ }
++ sec = tv->tv_sec;
++ __normalize_time(&sec, &nsec);
+
-+ /* 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",
++ if (is_initial_xendomain() && !independent_wallclock) {
++ op.cmd = DOM0_SETTIME;
++ op.u.settime.secs = sec;
++ op.u.settime.nsecs = nsec;
++ op.u.settime.system_time = shadow->system_timestamp;
++ HYPERVISOR_dom0_op(&op);
++ update_wallclock();
++ } else if (independent_wallclock) {
++ nsec -= shadow->system_timestamp;
++ __normalize_time(&sec, &nsec);
++ __update_wallclock(sec, nsec);
++ }
+
-+ /* 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,
++ write_sequnlock_irq(&xtime_lock);
+
-+ /* Other (Linux-defined) */
-+ "cxmmx", NULL, "cyrix_arr", "centaur_mcr", NULL,
-+ "constant_tsc", NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ "up", NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-+ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
++ put_cpu();
+
-+ /* 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,
++ clock_was_set();
++ return 0;
++}
+
-+ /* 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,
++EXPORT_SYMBOL(do_settimeofday);
+
-+ /* 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 */
-+ };
++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;
++ dom0_op_t op;
+
++ if (!ntp_synced() || independent_wallclock || !is_initial_xendomain())
++ return;
+
-+#ifdef CONFIG_SMP
-+ if (!cpu_online(c-cpu_data))
-+ return 0;
-+#endif
++ write_seqlock_irq(&xtime_lock);
+
-+ 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));
-+ }
++ sec = xtime.tv_sec;
++ nsec = xtime.tv_nsec;
++ __normalize_time(&sec, &nsec);
+
-+ /* 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
++ op.cmd = DOM0_SETTIME;
++ op.u.settime.secs = sec;
++ op.u.settime.nsecs = nsec;
++ op.u.settime.system_time = processed_system_time;
++ HYPERVISOR_dom0_op(&op);
+
-+ 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);
++ update_wallclock();
+
-+ {
-+ 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);
++ write_sequnlock_irq(&xtime_lock);
+
-+ 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);
++ /* Once per minute. */
++ mod_timer(&sync_xen_wallclock_timer, jiffies + 60*HZ);
++}
+
-+ seq_printf(m, "address sizes\t: %u bits physical, %u bits virtual\n",
-+ c->x86_phys_bits, c->x86_virt_bits);
++static int set_rtc_mmss(unsigned long nowtime)
++{
++ int retval;
++ unsigned long flags;
+
-+ 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);
-+ }
-+ }
++ if (independent_wallclock || !is_initial_xendomain())
++ return 0;
+
-+ seq_printf(m, "\n\n");
++ /* gets recalled with irq locally disabled */
++ 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 0;
++ return retval;
+}
+
-+static void *c_start(struct seq_file *m, loff_t *pos)
++/* 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)
+{
-+ return *pos < NR_CPUS ? cpu_data + *pos : NULL;
-+}
++ int cpu = get_cpu();
++ struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
++ u64 time;
++ u32 local_time_version;
+
-+static void *c_next(struct seq_file *m, void *v, loff_t *pos)
-+{
-+ ++*pos;
-+ return c_start(m, pos);
++ 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();
++ barrier();
++ } while (local_time_version != shadow->version);
++
++ put_cpu();
++
++ return time;
+}
++EXPORT_SYMBOL(monotonic_clock);
+
-+static void c_stop(struct seq_file *m, void *v)
++unsigned long long sched_clock(void)
+{
++ return monotonic_clock();
+}
+
-+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)
++unsigned long profile_pc(struct pt_regs *regs)
+{
-+ struct platform_device *pd;
-+ int ret;
-+
-+ pd = platform_device_alloc("pcspkr", -1);
-+ if (!pd)
-+ return -ENOMEM;
-+
-+ ret = platform_device_add(pd);
-+ if (ret)
-+ platform_device_put(pd);
++ unsigned long pc = instruction_pointer(regs);
+
-+ return ret;
++ /* Assume the lock function has either no stack frame or a copy
++ of eflags from PUSHF
++ Eflags always has bits 22 and up cleared unlike kernel addresses. */
++ if (!user_mode_vm(regs) && in_lock_functions(pc)) {
++ unsigned long *sp = (unsigned long *)regs->rsp;
++ if (sp[0] >> 22)
++ return sp[0];
++ if (sp[1] >> 22)
++ return sp[1];
++ }
++ return pc;
+}
-+device_initcall(add_pcspkr);
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/smp-xen.c linux-2.6.18-xen/arch/x86_64/kernel/smp-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/smp-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/smp-xen.c 2006-11-19 14:26:34.000000000 +0100
-@@ -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.
-+ */
++EXPORT_SYMBOL(profile_pc);
+
-+#include <linux/init.h>
++irqreturn_t timer_interrupt(int irq, void *dev_id)
++{
++ 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);
+
-+#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>
++ write_seqlock(&xtime_lock);
+
-+#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
++ do {
++ get_time_values_from_xen();
+
-+#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.
-+ */
++ /* 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);
+
-+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;
++ /*
++ * 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));
+
-+/* 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
++ 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));
++ }
+
-+/*
-+ * 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);
-+}
++ /* System-wide jiffy work. */
++ while (delta >= NS_PER_TICK) {
++ delta -= NS_PER_TICK;
++ processed_system_time += NS_PER_TICK;
++ do_timer(1);
++ }
+
-+#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.
-+ */
++ if (shadow_tv_version != HYPERVISOR_shared_info->wc_version) {
++ update_wallclock();
++ clock_was_set();
++ }
+
-+/*
-+ * 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.
-+ */
++ write_sequnlock(&xtime_lock);
+
-+asmlinkage void smp_invalidate_interrupt(struct pt_regs *regs)
-+{
-+ int cpu;
-+ int sender;
-+ union smp_flush_state *f;
++ /*
++ * 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);
++ }
+
-+ cpu = smp_processor_id();
+ /*
-+ * orig_rax contains the interrupt vector - 256.
-+ * Use that to determine where the sender put the data.
++ * Account blocked ticks.
++ * HACK: Passing idle_task to account_steal_time()
++ * ensures that the ticks are accounted as idle/wait.
+ */
-+ sender = regs->orig_rax + 256 - INVALIDATE_TLB_VECTOR_START;
-+ f = &per_cpu(flush_state, sender);
++ 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);
++ }
+
-+ 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);
++ /* 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(get_irq_regs()))
++ account_user_time(current, (cputime_t)delta_cpu);
++ else
++ account_system_time(current, HARDIRQ_OFFSET,
++ (cputime_t)delta_cpu);
+ }
-+out:
-+ ack_APIC_irq();
-+ cpu_clear(cpu, f->flush_cpumask);
++
++ /* Local timer processing (see update_process_times()). */
++ run_local_timers();
++ if (rcu_pending(cpu))
++ rcu_check_callbacks(cpu, user_mode(get_irq_regs()));
++ scheduler_tick();
++ run_posix_cpu_timers(current);
++
++ return IRQ_HANDLED;
+}
+
-+static void flush_tlb_others(cpumask_t cpumask, struct mm_struct *mm,
-+ unsigned long va)
++static void init_missing_ticks_accounting(int cpu)
+{
-+ int sender;
-+ union smp_flush_state *f;
++ struct vcpu_register_runstate_memory_area area;
++ struct vcpu_runstate_info *runstate = &per_cpu(runstate, cpu);
+
-+ /* Caller has disabled preemption */
-+ sender = smp_processor_id() % NUM_INVALIDATE_TLB_VECTORS;
-+ f = &per_cpu(flush_state, sender);
++ memset(runstate, 0, sizeof(*runstate));
+
-+ /* 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);
++ area.addr.v = runstate;
++ HYPERVISOR_vcpu_op(VCPUOP_register_runstate_memory_area, cpu, &area);
+
-+ f->flush_mm = mm;
-+ f->flush_va = va;
-+ cpus_or(f->flush_cpumask, cpumask, f->flush_cpumask);
++ per_cpu(processed_blocked_time, cpu) =
++ runstate->time[RUNSTATE_blocked];
++ per_cpu(processed_stolen_time, cpu) =
++ runstate->time[RUNSTATE_runnable] +
++ runstate->time[RUNSTATE_offline];
++}
+
-+ /*
-+ * We have to send the IPI only to
-+ * CPUs affected.
-+ */
-+ send_IPI_mask(cpumask, INVALIDATE_TLB_VECTOR_START + sender);
++/* not static: needed by APM */
++unsigned long get_cmos_time(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();
+
-+ while (!cpus_empty(f->flush_cpumask))
-+ cpu_relax();
++ spin_unlock_irqrestore(&rtc_lock, flags);
+
-+ f->flush_mm = NULL;
-+ f->flush_va = 0;
-+ spin_unlock(&f->tlbstate_lock);
++ return retval;
+}
++EXPORT_SYMBOL(get_cmos_time);
+
-+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;
-+}
++static void sync_cmos_clock(unsigned long dummy);
+
-+core_initcall(init_smp_flush);
-+
-+void flush_tlb_current_task(void)
++static DEFINE_TIMER(sync_cmos_timer, sync_cmos_clock, 0, 0);
++
++static void sync_cmos_clock(unsigned long dummy)
+{
-+ struct mm_struct *mm = current->mm;
-+ cpumask_t cpu_mask;
++ struct timeval now, next;
++ int fail = 1;
+
-+ preempt_disable();
-+ cpu_mask = mm->cpu_vm_mask;
-+ cpu_clear(smp_processor_id(), cpu_mask);
++ /*
++ * 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;
+
-+ local_flush_tlb();
-+ if (!cpus_empty(cpu_mask))
-+ flush_tlb_others(cpu_mask, mm, FLUSH_ALL);
-+ preempt_enable();
-+}
-+EXPORT_SYMBOL(flush_tlb_current_task);
++ 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);
+
-+void flush_tlb_mm (struct mm_struct * mm)
-+{
-+ cpumask_t cpu_mask;
++ next.tv_usec = USEC_AFTER - now.tv_usec;
++ if (next.tv_usec <= 0)
++ next.tv_usec += USEC_PER_SEC;
+
-+ preempt_disable();
-+ cpu_mask = mm->cpu_vm_mask;
-+ cpu_clear(smp_processor_id(), cpu_mask);
++ if (!fail)
++ next.tv_sec = 659;
++ else
++ next.tv_sec = 0;
+
-+ if (current->active_mm == mm) {
-+ if (current->mm)
-+ local_flush_tlb();
-+ else
-+ leave_mm(smp_processor_id());
++ if (next.tv_usec >= USEC_PER_SEC) {
++ next.tv_sec++;
++ next.tv_usec -= USEC_PER_SEC;
+ }
-+ if (!cpus_empty(cpu_mask))
-+ flush_tlb_others(cpu_mask, mm, FLUSH_ALL);
-+
-+ preempt_enable();
++ mod_timer(&sync_cmos_timer, jiffies + timeval_to_jiffies(&next));
+}
-+EXPORT_SYMBOL(flush_tlb_mm);
+
-+void flush_tlb_page(struct vm_area_struct * vma, unsigned long va)
++void notify_arch_cmos_timer(void)
+{
-+ 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);
++ mod_timer(&sync_cmos_timer, jiffies + 1);
++ mod_timer(&sync_xen_wallclock_timer, jiffies + 1);
++}
+
-+ if (current->active_mm == mm) {
-+ if(current->mm)
-+ __flush_tlb_one(va);
-+ else
-+ leave_mm(smp_processor_id());
-+ }
++static long clock_cmos_diff;
++static unsigned long sleep_start;
+
-+ if (!cpus_empty(cpu_mask))
-+ flush_tlb_others(cpu_mask, mm, va);
++static int timer_suspend(struct sys_device *dev, pm_message_t state)
++{
++ /*
++ * Estimate time zone so that set_time can update the clock
++ */
++ unsigned long ctime = get_cmos_time();
+
-+ preempt_enable();
++ clock_cmos_diff = -ctime;
++ clock_cmos_diff += get_seconds();
++ sleep_start = ctime;
++ return 0;
+}
-+EXPORT_SYMBOL(flush_tlb_page);
+
-+static void do_flush_tlb_all(void* info)
++static int timer_resume(struct sys_device *dev)
+{
-+ unsigned long cpu = smp_processor_id();
++ unsigned long flags;
++ unsigned long sec;
++ unsigned long ctime = get_cmos_time();
++ long sleep_length = (ctime - sleep_start) * HZ;
+
-+ __flush_tlb_all();
-+ if (read_pda(mmu_state) == TLBSTATE_LAZY)
-+ leave_mm(cpu);
-+}
++ if (sleep_length < 0) {
++ printk(KERN_WARNING "CMOS clock skew detected in timer resume!\n");
++ /* The time after the resume must not be earlier than the time
++ * before the suspend or some nasty things will happen
++ */
++ sleep_length = 0;
++ ctime = sleep_start;
++ }
+
-+void flush_tlb_all(void)
-+{
-+ on_each_cpu(do_flush_tlb_all, NULL, 1, 1);
++#ifdef CONFIG_HPET_TIMER
++ if (is_hpet_enabled())
++ hpet_reenable();
++#endif
++ sec = ctime + clock_cmos_diff;
++ write_seqlock_irqsave(&xtime_lock, flags);
++ xtime.tv_sec = sec;
++ xtime.tv_nsec = 0;
++ jiffies_64 += sleep_length;
++ write_sequnlock_irqrestore(&xtime_lock, flags);
++ touch_softlockup_watchdog();
++ return 0;
+}
-+#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 ...
-+ */
++static struct sysdev_class timer_sysclass = {
++ .resume = timer_resume,
++ .suspend = timer_suspend,
++ set_kset_name("timer"),
++};
+
-+void smp_send_reschedule(int cpu)
++
++/* XXX this driverfs stuff should probably go elsewhere later -john */
++static struct sys_device device_timer = {
++ .id = 0,
++ .cls = &timer_sysclass,
++};
++
++static int time_init_device(void)
+{
-+ send_IPI_mask(cpumask_of_cpu(cpu), RESCHEDULE_VECTOR);
++ int error = sysdev_class_register(&timer_sysclass);
++ if (!error)
++ error = sysdev_register(&device_timer);
++ return error;
+}
+
-+/*
-+ * Structure and data for smp_call_function(). This is designed to minimise
-+ * static memory requirements. It also looks cleaner.
-+ */
-+static DEFINE_SPINLOCK(call_lock);
++device_initcall(time_init_device);
+
-+struct call_data_struct {
-+ void (*func) (void *info);
-+ void *info;
-+ atomic_t started;
-+ atomic_t finished;
-+ int wait;
-+};
++#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);
+
-+static struct call_data_struct * call_data;
++ if ((hpet_enable() >= 0) && hpet_use_timer) {
++ printk("Using HPET for base-timer\n");
++ }
+
-+void lock_ipi_call_lock(void)
-+{
-+ spin_lock_irq(&call_lock);
++ time_init_hook();
+}
++#endif
+
-+void unlock_ipi_call_lock(void)
++/* Dynamically-mapped IRQ. */
++DEFINE_PER_CPU(int, timer_irq);
++
++extern void (*late_time_init)(void);
++static void setup_cpu0_timer_irq(void)
+{
-+ spin_unlock_irq(&call_lock);
++ 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);
+}
+
-+/*
-+ * 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)
++void __init time_init(void)
+{
-+ struct call_data_struct data;
-+ int cpus = 1;
++#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
++ get_time_values_from_xen();
+
-+ data.func = func;
-+ data.info = info;
-+ atomic_set(&data.started, 0);
-+ data.wait = wait;
-+ if (wait)
-+ atomic_set(&data.finished, 0);
++ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
++ per_cpu(processed_system_time, 0) = processed_system_time;
++ init_missing_ticks_accounting(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);
++ update_wallclock();
+
-+ /* Wait for response */
-+ while (atomic_read(&data.started) != cpus)
-+ cpu_relax();
++ init_cpu_khz();
++ printk(KERN_INFO "Xen reported: %u.%03u MHz processor.\n",
++ cpu_khz / 1000, cpu_khz % 1000);
+
-+ if (!wait)
-+ return;
++ vxtime.mode = VXTIME_TSC;
++ vxtime.quot = (1000000L << US_SCALE) / vxtime_hz;
++ vxtime.tsc_quot = (1000L << US_SCALE) / cpu_khz;
++ sync_core();
++ rdtscll(vxtime.last_tsc);
+
-+ while (atomic_read(&data.finished) != cpus)
-+ cpu_relax();
++ /* Cannot request_irq() until kmem is initialised. */
++ late_time_init = setup_cpu0_timer_irq;
+}
+
-+/*
-+ * 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)
++/* Convert jiffies to system time. */
++u64 jiffies_to_st(unsigned long j)
+{
-+ /* 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;
++ 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));
++
++ return st;
+}
++EXPORT_SYMBOL(jiffies_to_st);
+
+/*
-+ * this function sends a 'generic call function' IPI to all other CPUs
-+ * in the system.
++ * 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 __smp_call_function (void (*func) (void *info), void *info,
-+ int nonatomic, int wait)
++static void stop_hz_timer(void)
+{
-+ struct call_data_struct data;
-+ int cpus = num_online_cpus()-1;
-+
-+ if (!cpus)
-+ return;
++ unsigned int cpu = smp_processor_id();
++ unsigned long j;
+
-+ data.func = func;
-+ data.info = info;
-+ atomic_set(&data.started, 0);
-+ data.wait = wait;
-+ if (wait)
-+ atomic_set(&data.finished, 0);
++ cpu_set(cpu, nohz_cpu_mask);
+
-+ call_data = &data;
-+ wmb();
-+ /* Send a message to all other CPUs and wait for them to respond */
-+ send_IPI_allbutself(CALL_FUNCTION_VECTOR);
++ /* 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. */
+
-+ /* Wait for response */
-+ while (atomic_read(&data.started) != cpus)
-+#ifndef CONFIG_XEN
-+ cpu_relax();
-+#else
-+ barrier();
-+#endif
++ smp_mb();
+
-+ if (!wait)
-+ 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;
++ }
+
-+ while (atomic_read(&data.finished) != cpus)
-+#ifndef CONFIG_XEN
-+ cpu_relax();
-+#else
-+ barrier();
-+#endif
++ if (HYPERVISOR_set_timer_op(jiffies_to_st(j)) != 0)
++ BUG();
+}
+
-+/*
-+ * 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)
++static void start_hz_timer(void)
+{
-+ spin_lock(&call_lock);
-+ __smp_call_function(func,info,nonatomic,wait);
-+ spin_unlock(&call_lock);
-+ return 0;
++ cpu_clear(smp_processor_id(), nohz_cpu_mask);
+}
-+EXPORT_SYMBOL(smp_call_function);
+
-+void smp_stop_cpu(void)
++void raw_safe_halt(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);
++ stop_hz_timer();
++ /* Blocking includes an implicit local_irq_enable(). */
++ HYPERVISOR_block();
++ start_hz_timer();
+}
++EXPORT_SYMBOL(raw_safe_halt);
+
-+static void smp_really_stop_cpu(void *dummy)
-+{
-+ smp_stop_cpu();
-+ for (;;)
-+ halt();
-+}
-+
-+void smp_send_stop(void)
++void halt(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();
-+#ifndef CONFIG_XEN
-+ disable_local_APIC();
-+#endif
-+ local_irq_enable();
++ if (irqs_disabled())
++ HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL);
+}
++EXPORT_SYMBOL(halt);
+
-+/*
-+ * 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
++/* No locking required. We are only CPU running, and interrupts are off. */
++void time_resume(void)
+{
-+#ifndef CONFIG_XEN
-+ ack_APIC_irq();
-+#else
-+ return IRQ_HANDLED;
-+#endif
-+}
++ init_cpu_khz();
+
-+#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;
++ get_time_values_from_xen();
+
-+#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
++ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
++ per_cpu(processed_system_time, 0) = processed_system_time;
++ init_missing_ticks_accounting(0);
++
++ update_wallclock();
+}
+
-+int safe_smp_processor_id(void)
-+{
-+#ifdef CONFIG_XEN
-+ return smp_processor_id();
-+#else
-+ unsigned apicid, i;
++#ifdef CONFIG_SMP
++static char timer_name[NR_CPUS][15];
+
-+ if (disable_apic)
-+ return 0;
++void local_setup_timer(unsigned int cpu)
++{
++ int seq;
+
-+ apicid = hard_smp_processor_id();
-+ if (apicid < NR_CPUS && x86_cpu_to_apicid[apicid] == apicid)
-+ return apicid;
++ BUG_ON(cpu == 0);
+
-+ for (i = 0; i < NR_CPUS; ++i) {
-+ if (x86_cpu_to_apicid[i] == apicid)
-+ return i;
-+ }
++ 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));
+
-+ /* 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;
++ sprintf(timer_name[cpu], "timer%d", cpu);
++ per_cpu(timer_irq, cpu) =
++ bind_virq_to_irqhandler(
++ VIRQ_TIMER,
++ cpu,
++ timer_interrupt,
++ SA_INTERRUPT,
++ timer_name[cpu],
++ NULL);
++ BUG_ON(per_cpu(timer_irq, cpu) < 0);
++}
+
-+ return 0; /* Should not happen */
++void local_teardown_timer(unsigned int cpu)
++{
++ BUG_ON(cpu == 0);
++ unbind_from_irqhandler(per_cpu(timer_irq, cpu), NULL);
++}
+#endif
++
++/*
++ * /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;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/traps-xen.c linux-2.6.18-xen/arch/x86_64/kernel/traps-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/traps-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/traps-xen.c 2006-11-19 14:26:34.000000000 +0100
-@@ -0,0 +1,1172 @@
++__initcall(xen_sysctl_init);
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/traps-xen.c b/arch/x86_64/kernel/traps-xen.c
+--- a/arch/x86_64/kernel/traps-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/traps-xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,1156 @@
+/*
+ * linux/arch/x86-64/traps.c
+ *
@@ -39975,15 +35334,17 @@
+#include <linux/delay.h>
+#include <linux/spinlock.h>
+#include <linux/interrupt.h>
++#include <linux/kallsyms.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 <linux/uaccess.h>
++#include <linux/bug.h>
+
+#include <asm/system.h>
-+#include <asm/uaccess.h>
+#include <asm/io.h>
+#include <asm/atomic.h>
+#include <asm/debugreg.h>
@@ -39997,6 +35358,7 @@
+#include <asm/pda.h>
+#include <asm/proto.h>
+#include <asm/nmi.h>
++#include <asm/stacktrace.h>
+
+asmlinkage void divide_error(void);
+asmlinkage void debug(void);
@@ -40021,6 +35383,7 @@
+asmlinkage void spurious_interrupt_bug(void);
+
+ATOMIC_NOTIFIER_HEAD(die_chain);
++EXPORT_SYMBOL(die_chain);
+
+int register_die_notifier(struct notifier_block *nb)
+{
@@ -40057,15 +35420,9 @@
+ preempt_enable_no_resched();
+}
+
-+static int kstack_depth_to_print = 10;
-+#ifdef CONFIG_STACK_UNWIND
-+static int call_trace = 1;
-+#else
-+#define call_trace (-1)
-+#endif
++int kstack_depth_to_print = 12;
+
+#ifdef CONFIG_KALLSYMS
-+# include <linux/kallsyms.h>
+void printk_address(unsigned long address)
+{
+ unsigned long offset = 0, symsize;
@@ -40093,7 +35450,7 @@
+#endif
+
+static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
-+ unsigned *usedp, const char **idp)
++ unsigned *usedp, char **idp)
+{
+#ifndef CONFIG_X86_NO_TSS
+ static char ids[][8] = {
@@ -40113,26 +35470,7 @@
+ * '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;
-+ }
++ unsigned long end = per_cpu(orig_ist, cpu).ist[k];
+ /*
+ * Is 'stack' above this exception frame's end?
+ * If yes then skip to the next frame.
@@ -40187,18 +35525,7 @@
+ return NULL;
+}
+
-+static int show_trace_unwind(struct unwind_frame_info *info, void *context)
-+{
-+ int n = 0;
-+
-+ while (unwind(info) == 0 && UNW_PC(info)) {
-+ if (arch_unw_user_mode(info))
-+ break;
-+ n++;
-+ printk_address(UNW_PC(info));
-+ }
-+ return n;
-+}
++#define MSG(txt) ops->warning(data, txt)
+
+/*
+ * x86-64 can have upto three kernel stacks:
@@ -40207,55 +35534,42 @@
+ * 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)
++static inline int valid_stack_ptr(struct thread_info *tinfo, void *p)
+{
-+ const unsigned cpu = safe_smp_processor_id();
-+ unsigned long *irqstack_end = (unsigned long *)cpu_pda(cpu)->irqstackptr;
-+ unsigned used = 0;
++ void *t = (void *)tinfo;
++ return p > t && p < t + THREAD_SIZE - 3;
++}
+
-+ printk("\nCall Trace:\n");
++void dump_trace(struct task_struct *tsk, struct pt_regs *regs,
++ unsigned long *stack,
++ struct stacktrace_ops *ops, void *data)
++{
++ const unsigned cpu = get_cpu();
++ unsigned long *irqstack_end = (unsigned long*)cpu_pda(cpu)->irqstackptr;
++ unsigned used = 0;
++ struct thread_info *tinfo;
+
+ 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");
++ if (!stack) {
++ unsigned long dummy;
++ stack = &dummy;
++ if (tsk && tsk != current)
++ stack = (unsigned long *)tsk->thread.rsp;
+ }
-+ /*
-+ * Print function call entries within a stack. 'cond' is the
-+ * "end of stackframe" condition, that the 'stack++'
-+ * iteration will eventually trigger.
-+ */
++
++ /*
++ * 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)) { \
++ /* Use unlocked access here because except for NMIs \
++ we should be already protected against module unloads */ \
++ 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 \
@@ -40264,7 +35578,7 @@
+ * down the cause of the crash will be able to figure \
+ * out the call path that was taken. \
+ */ \
-+ printk_address(addr); \
++ ops->address(data, addr); \
+ } \
+ } while (0)
+
@@ -40273,16 +35587,17 @@
+ * current stack address. If the stacks consist of nested
+ * exceptions
+ */
-+ for ( ; ; ) {
-+ const char *id;
++ for (;;) {
++ char *id;
+ unsigned long *estack_end;
+ estack_end = in_exception_stack(cpu, (unsigned long)stack,
+ &used, &id);
+
+ if (estack_end) {
-+ printk(" <%s>", id);
++ if (ops->stack(data, id) < 0)
++ break;
+ HANDLE_STACK (stack < estack_end);
-+ printk(" <EOE>");
++ ops->stack(data, "<EOE>");
+ /*
+ * We link to the next stack via the
+ * second-to-last pointer (index -2 to end) in the
@@ -40297,7 +35612,8 @@
+ (IRQSTACKSIZE - 64) / sizeof(*irqstack);
+
+ if (stack >= irqstack && stack < irqstack_end) {
-+ printk(" <IRQ>");
++ if (ops->stack(data, "IRQ") < 0)
++ break;
+ HANDLE_STACK (stack < irqstack_end);
+ /*
+ * We link to the next stack (which would be
@@ -40306,7 +35622,7 @@
+ */
+ stack = (unsigned long *) (irqstack_end[-1]);
+ irqstack_end = NULL;
-+ printk(" <EOI>");
++ ops->stack(data, "EOI");
+ continue;
+ }
+ }
@@ -40314,19 +35630,59 @@
+ }
+
+ /*
-+ * This prints the process stack:
++ * This handles the process stack:
+ */
-+ HANDLE_STACK (((long) stack & (THREAD_SIZE-1)) != 0);
++ tinfo = task_thread_info(tsk);
++ HANDLE_STACK (valid_stack_ptr(tinfo, stack));
+#undef HANDLE_STACK
++ put_cpu();
++}
++EXPORT_SYMBOL(dump_trace);
++
++static void
++print_trace_warning_symbol(void *data, char *msg, unsigned long symbol)
++{
++ print_symbol(msg, symbol);
++ printk("\n");
++}
++
++static void print_trace_warning(void *data, char *msg)
++{
++ printk("%s\n", msg);
++}
++
++static int print_trace_stack(void *data, char *name)
++{
++ printk(" <%s> ", name);
++ return 0;
++}
++
++static void print_trace_address(void *data, unsigned long addr)
++{
++ printk_address(addr);
++}
++
++static struct stacktrace_ops print_trace_ops = {
++ .warning = print_trace_warning,
++ .warning_symbol = print_trace_warning_symbol,
++ .stack = print_trace_stack,
++ .address = print_trace_address,
++};
+
++void
++show_trace(struct task_struct *tsk, struct pt_regs *regs, unsigned long *stack)
++{
++ printk("\nCall Trace:\n");
++ dump_trace(tsk, regs, stack, &print_trace_ops, NULL);
+ printk("\n");
+}
+
-+static void _show_stack(struct task_struct *tsk, struct pt_regs *regs, unsigned long * rsp)
++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();
++ const int cpu = smp_processor_id();
+ unsigned long *irqstack_end = (unsigned long *) (cpu_pda(cpu)->irqstackptr);
+ unsigned long *irqstack = (unsigned long *) (cpu_pda(cpu)->irqstackptr - IRQSTACKSIZE);
+
@@ -40353,7 +35709,7 @@
+ }
+ if (i && ((i % 4) == 0))
+ printk("\n");
-+ printk(" %016lx ", *stack++);
++ printk(" %016lx", *stack++);
+ touch_nmi_watchdog();
+ }
+ show_trace(tsk, regs, rsp);
@@ -40380,7 +35736,7 @@
+ int i;
+ int in_kernel = !user_mode(regs);
+ unsigned long rsp;
-+ const int cpu = safe_smp_processor_id();
++ const int cpu = smp_processor_id();
+ struct task_struct *cur = cpu_pda(cpu)->pcurrent;
+
+ rsp = regs->rsp;
@@ -40416,36 +35772,22 @@
+ printk("\n");
+}
+
-+void handle_BUG(struct pt_regs *regs)
-+{
-+ struct bug_frame f;
-+ long len;
-+ const char *prefix = "";
++int is_valid_bugaddr(unsigned long rip)
++{
++ unsigned short ud2;
+
-+ 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);
-+}
++ if (__copy_from_user(&ud2, (const void __user *) rip, sizeof(ud2)))
++ return 0;
++
++ return ud2 == 0x0b0f;
++}
+
+#ifdef CONFIG_BUG
+void out_of_line_bug(void)
+{
+ BUG();
+}
++EXPORT_SYMBOL(out_of_line_bug);
+#endif
+
+static DEFINE_SPINLOCK(die_lock);
@@ -40454,9 +35796,11 @@
+
+unsigned __kprobes long oops_begin(void)
+{
-+ int cpu = safe_smp_processor_id();
++ int cpu = smp_processor_id();
+ unsigned long flags;
+
++ oops_enter();
++
+ /* racy, but better than risking deadlock. */
+ local_irq_save(flags);
+ if (!spin_trylock(&die_lock)) {
@@ -40485,6 +35829,7 @@
+ spin_unlock_irqrestore(&die_lock, flags);
+ if (panic_on_oops)
+ panic("Fatal exception");
++ oops_exit();
+}
+
+void __kprobes __die(const char * str, struct pt_regs * regs, long err)
@@ -40515,14 +35860,15 @@
+{
+ unsigned long flags = oops_begin();
+
-+ handle_BUG(regs);
++ if (!user_mode(regs))
++ report_bug(regs->rip);
++
+ __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)
++void __kprobes die_nmi(char *str, struct pt_regs *regs, int do_panic)
+{
+ unsigned long flags = oops_begin();
+
@@ -40530,19 +35876,17 @@
+ * We are in trouble anyway, lets at least try
+ * to get a message out.
+ */
-+ printk(str, safe_smp_processor_id());
++ printk(str, 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");
++ if (do_panic || panic_on_oops)
++ panic("Non maskable interrupt");
+ 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,
@@ -40683,8 +36027,14 @@
+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");
++ printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x.\n",
++ reason);
++ printk(KERN_EMERG "You have some hardware problem, likely on the PCI bus.\n");
++
++ if (panic_on_unrecovered_nmi)
++ panic("NMI: Not continuing");
++
++ printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
+
+#if 0 /* XEN */
+ /* Clear and disable the memory parity error line. */
@@ -40711,9 +36061,15 @@
+
+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");
++{
++ printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x.\n",
++ reason);
++ printk(KERN_EMERG "Do you have a strange power saving mode enabled?\n");
++
++ if (panic_on_unrecovered_nmi)
++ panic("NMI: Not continuing");
++
++ printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
+}
+
+/* Runs on IST stack. This code must keep interrupts off all the time.
@@ -40733,17 +36089,15 @@
+ 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);
++ if (nmi_watchdog_tick(regs,reason))
+ return;
-+ }
-+#endif
-+ unknown_nmi_error(reason, regs);
++ if (!do_nmi_callback(regs,cpu))
++ unknown_nmi_error(reason, regs);
++
+ return;
+ }
+ if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP)
@@ -41030,6 +36384,7 @@
+ init_fpu(me);
+ restore_fpu_checking(&me->thread.i387.fxsave);
+ task_thread_info(me)->status |= TS_USEDFPU;
++ me->fpu_counter++;
+}
+
+
@@ -41092,334 +36447,57 @@
+}
+
+
-+/* Actual parsing is done early in setup.c. */
-+static int __init oops_dummy(char *s)
++static int __init oops_setup(char *s)
+{
-+ panic_on_oops = 1;
-+ return 1;
++ if (!s)
++ return -EINVAL;
++ if (!strcmp(s, "panic"))
++ panic_on_oops = 1;
++ return 0;
+}
-+__setup("oops=", oops_dummy);
++early_param("oops", oops_setup);
+
+static int __init kstack_setup(char *s)
+{
++ if (!s)
++ return -EINVAL;
+ 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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/vmlinux.lds.S linux-2.6.18-xen/arch/x86_64/kernel/vmlinux.lds.S
---- linux-2.6.18.3/arch/x86_64/kernel/vmlinux.lds.S 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/x86_64/kernel/vmlinux.lds.S 2006-11-19 14:26:34.000000000 +0100
-@@ -13,6 +13,12 @@
- 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 */
-+ note PT_NOTE FLAGS(4); /* R__ */
++ return 0;
+}
- SECTIONS
- {
- . = __START_KERNEL;
-@@ -31,7 +37,7 @@
- KPROBES_TEXT
- *(.fixup)
- *(.gnu.warning)
-- } = 0x9090
-+ } :text = 0x9090
- /* out-of-line lock text */
- .text.lock : AT(ADDR(.text.lock) - LOAD_OFFSET) { *(.text.lock) }
-
-@@ -57,17 +63,10 @@
- .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 +88,7 @@
- #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 +131,7 @@
- . = ALIGN(8192); /* init_task */
- .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) {
- *(.data.init_task)
-- }
-+ } :data
-
- . = ALIGN(4096);
- .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) {
-@@ -222,6 +221,14 @@
- . = 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 +242,6 @@
++early_param("kstack", kstack_setup);
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/vmlinux.lds.S b/arch/x86_64/kernel/vmlinux.lds.S
+--- a/arch/x86_64/kernel/vmlinux.lds.S 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/kernel/vmlinux.lds.S 2007-03-14 10:55:21.000000000 +0100
+@@ -227,4 +227,6 @@
STABS_DEBUG
DWARF_DEBUG
+
+ NOTES
}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/vsyscall-xen.c linux-2.6.18-xen/arch/x86_64/kernel/vsyscall-xen.c
---- linux-2.6.18.3/arch/x86_64/kernel/vsyscall-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/vsyscall-xen.c 2006-11-19 14:26:34.000000000 +0100
-@@ -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.
-+ */
-+
-+#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 },
-+};
-+
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/vsyscall.c b/arch/x86_64/kernel/vsyscall.c
+--- a/arch/x86_64/kernel/vsyscall.c 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/kernel/vsyscall.c 2007-03-14 10:55:21.000000000 +0100
+@@ -246,6 +246,7 @@
+
+ #endif
+
++#ifndef CONFIG_XEN
+ /* Assume __initcall executes before all user space. Hopefully kmod
+ doesn't violate that. We'll find out if it does. */
+ static void __cpuinit vsyscall_set_cpu(int cpu)
+@@ -282,6 +283,7 @@
+ smp_call_function_single(cpu, cpu_vsyscall_init, NULL, 0, 1);
+ return NOTIFY_DONE;
+ }
+#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 void __init map_vsyscall(void)
+ {
+@@ -292,6 +294,17 @@
+ __set_fixmap(VSYSCALL_FIRST_PAGE, physaddr_page0, PAGE_KERNEL_VSYSCALL);
+ }
+
+#ifdef CONFIG_XEN
+static void __init map_vsyscall_user(void)
+{
@@ -41431,27 +36509,30 @@
+}
+#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();
+ static int __init vsyscall_init(void)
+ {
+ BUG_ON(((unsigned long) &vgettimeofday !=
+@@ -300,11 +313,17 @@
+ BUG_ON((VSYSCALL_ADDR(0) != __fix_to_virt(VSYSCALL_FIRST_PAGE)));
+ BUG_ON((unsigned long) &vgetcpu != VSYSCALL_ADDR(__NR_vgetcpu));
+ map_vsyscall();
+#ifdef CONFIG_XEN
+ map_vsyscall_user();
+ sysctl_vsyscall = 0; /* disable vgettimeofay() */
+#endif
-+#ifdef CONFIG_SYSCTL
-+ register_sysctl_table(kernel_root_table2, 0);
+ #ifdef CONFIG_SYSCTL
+ register_sysctl_table(kernel_root_table2, 0);
+ #endif
++#ifndef CONFIG_XEN
+ on_each_cpu(cpu_vsyscall_init, NULL, 0, 1);
+ hotcpu_notifier(cpu_vsyscall_notifier, 0);
+#endif
-+ return 0;
-+}
-+
-+__initcall(vsyscall_init);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/kernel/xen_entry.S linux-2.6.18-xen/arch/x86_64/kernel/xen_entry.S
---- linux-2.6.18.3/arch/x86_64/kernel/xen_entry.S 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/kernel/xen_entry.S 2006-11-19 14:26:34.000000000 +0100
+ return 0;
+ }
+
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/kernel/xen_entry.S b/arch/x86_64/kernel/xen_entry.S
+--- a/arch/x86_64/kernel/xen_entry.S 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/kernel/xen_entry.S 2007-03-14 10:55:21.000000000 +0100
@@ -0,0 +1,40 @@
+/*
+ * Copied from arch/xen/i386/kernel/entry.S
@@ -41493,11 +36574,11 @@
+VGCF_IN_SYSCALL = (1<<8)
+
+
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/Makefile linux-2.6.18-xen/arch/x86_64/Makefile
---- linux-2.6.18.3/arch/x86_64/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/x86_64/Makefile 2006-11-19 14:26:33.000000000 +0100
-@@ -32,6 +32,10 @@
- cflags-$(CONFIG_MPSC) += $(call cc-option,-march=nocona)
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/Makefile b/arch/x86_64/Makefile
+--- a/arch/x86_64/Makefile 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/Makefile 2007-03-14 10:55:21.000000000 +0100
+@@ -36,6 +36,10 @@
+ $(call cc-option,-march=core2,$(call cc-option,-mtune=generic))
cflags-$(CONFIG_GENERIC_CPU) += $(call cc-option,-mtune=generic)
+cppflags-$(CONFIG_XEN) += \
@@ -41507,13 +36588,13 @@
cflags-y += -m64
cflags-y += -mno-red-zone
cflags-y += -mcmodel=kernel
-@@ -74,6 +78,21 @@
+@@ -90,6 +94,21 @@
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
++head-y := arch/x86_64/kernel/head-xen.o arch/x86_64/kernel/head64.o arch/x86_64/kernel/init_task.o
+LDFLAGS_vmlinux := -e _start
+boot := arch/i386/boot-xen
+.PHONY: vmlinuz
@@ -41529,7 +36610,7 @@
#Default target when executing "make"
all: bzImage
-@@ -94,6 +113,7 @@
+@@ -110,6 +129,7 @@
install:
$(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(BOOTIMAGE) $@
@@ -41537,10 +36618,10 @@
archclean:
$(Q)$(MAKE) $(clean)=$(boot)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/mm/fault-xen.c linux-2.6.18-xen/arch/x86_64/mm/fault-xen.c
---- linux-2.6.18.3/arch/x86_64/mm/fault-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/mm/fault-xen.c 2006-11-19 14:26:34.000000000 +0100
-@@ -0,0 +1,723 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/mm/fault-xen.c b/arch/x86_64/mm/fault-xen.c
+--- a/arch/x86_64/mm/fault-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/mm/fault-xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,717 @@
+/*
+ * linux/arch/x86-64/mm/fault.c
+ *
@@ -41566,9 +36647,9 @@
+#include <linux/compiler.h>
+#include <linux/module.h>
+#include <linux/kprobes.h>
++#include <linux/uaccess.h>
+
+#include <asm/system.h>
-+#include <asm/uaccess.h>
+#include <asm/pgalloc.h>
+#include <asm/smp.h>
+#include <asm/tlbflush.h>
@@ -41583,8 +36664,7 @@
+#define PF_RSVD (1<<3)
+#define PF_INSTR (1<<4)
+
-+#ifdef CONFIG_KPROBES
-+ATOMIC_NOTIFIER_HEAD(notify_page_fault_chain);
++static ATOMIC_NOTIFIER_HEAD(notify_page_fault_chain);
+
+/* Hook to register for page fault notifications */
+int register_page_fault_notifier(struct notifier_block *nb)
@@ -41592,11 +36672,13 @@
+ vmalloc_sync_all();
+ return atomic_notifier_chain_register(¬ify_page_fault_chain, nb);
+}
++EXPORT_SYMBOL_GPL(register_page_fault_notifier);
+
+int unregister_page_fault_notifier(struct notifier_block *nb)
+{
+ return atomic_notifier_chain_unregister(¬ify_page_fault_chain, nb);
+}
++EXPORT_SYMBOL_GPL(unregister_page_fault_notifier);
+
+static inline int notify_page_fault(enum die_val val, const char *str,
+ struct pt_regs *regs, long err, int trap, int sig)
@@ -41610,13 +36692,6 @@
+ };
+ 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)
+{
@@ -41654,7 +36729,7 @@
+ if (error_code & PF_INSTR)
+ return 0;
+
-+ instr = (unsigned char *)convert_rip_to_linear(current, regs);
++ instr = (unsigned char __user *)convert_rip_to_linear(current, regs);
+ max_instr = instr + 15;
+
+ if (user_mode(regs) && instr >= (unsigned char *)TASK_SIZE)
@@ -41665,7 +36740,7 @@
+ unsigned char instr_hi;
+ unsigned char instr_lo;
+
-+ if (__get_user(opcode, instr))
++ if (probe_kernel_address(instr, opcode))
+ break;
+
+ instr_hi = opcode & 0xf0;
@@ -41703,7 +36778,7 @@
+ case 0x00:
+ /* Prefetch instruction is 0x0F0D or 0x0F18 */
+ scan_more = 0;
-+ if (__get_user(opcode, instr))
++ if (probe_kernel_address(instr, opcode))
+ break;
+ prefetch = (instr_lo == 0xF) &&
+ (opcode == 0x0D || opcode == 0x18);
@@ -41719,7 +36794,7 @@
+static int bad_address(void *p)
+{
+ unsigned long dummy;
-+ return __get_user(dummy, (unsigned long *)p);
++ return probe_kernel_address((unsigned long *)p, dummy);
+}
+
+void dump_pagetable(unsigned long address)
@@ -41791,7 +36866,7 @@
+
+int unhandled_signal(struct task_struct *tsk, int sig)
+{
-+ if (tsk->pid == 1)
++ if (is_init(tsk))
+ return 1;
+ if (tsk->ptrace & PT_PTRACED)
+ return 0;
@@ -41843,7 +36918,7 @@
+ if (pgd_none(*pgd))
+ set_pgd(pgd, *pgd_ref);
+ else
-+ BUG_ON(pgd_page(*pgd) != pgd_page(*pgd_ref));
++ BUG_ON(pgd_page_vaddr(*pgd) != pgd_page_vaddr(*pgd_ref));
+
+ /* Below here mismatches are bugs because these lower tables
+ are shared */
@@ -41852,7 +36927,7 @@
+ 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))
++ if (pud_none(*pud) || pud_page_vaddr(*pud) != pud_page_vaddr(*pud_ref))
+ BUG();
+ pmd = pmd_offset(pud, address);
+ pmd_ref = pmd_offset(pud_ref, address);
@@ -42073,7 +37148,7 @@
+ case PF_PROT: /* read, present */
+ goto bad_area;
+ case 0: /* read, not present */
-+ if (!(vma->vm_flags & (VM_READ | VM_EXEC)))
++ if (!(vma->vm_flags & (VM_READ | VM_EXEC | VM_WRITE)))
+ goto bad_area;
+ }
+
@@ -42189,7 +37264,7 @@
+ */
+out_of_memory:
+ up_read(&mm->mmap_sem);
-+ if (current->pid == 1) {
++ if (is_init(current)) {
+ yield();
+ goto again;
+ }
@@ -42244,7 +37319,7 @@
+ if (pgd_none(*pgd))
+ set_pgd(pgd, *pgd_ref);
+ else
-+ BUG_ON(pgd_page(*pgd) != pgd_page(*pgd_ref));
++ BUG_ON(pgd_page_vaddr(*pgd) != pgd_page_vaddr(*pgd_ref));
+ }
+ spin_unlock(&pgd_lock);
+ set_bit(pgd_index(address), insync);
@@ -42264,10 +37339,10 @@
+ return 1;
+}
+__setup("pagefaulttrace", enable_pagefaulttrace);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/mm/init-xen.c linux-2.6.18-xen/arch/x86_64/mm/init-xen.c
---- linux-2.6.18.3/arch/x86_64/mm/init-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/mm/init-xen.c 2006-11-19 14:26:34.000000000 +0100
-@@ -0,0 +1,1200 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/mm/init-xen.c b/arch/x86_64/mm/init-xen.c
+--- a/arch/x86_64/mm/init-xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/arch/x86_64/mm/init-xen.c 2007-03-14 10:55:21.000000000 +0100
+@@ -0,0 +1,1148 @@
+/*
+ * linux/arch/x86_64/mm/init.c
+ *
@@ -42668,7 +37743,6 @@
+
+ if (after_bootmem) {
+ void *adr = (void *)get_zeroed_page(GFP_ATOMIC);
-+
+ *phys = __pa(adr);
+ return adr;
+ }
@@ -42720,7 +37794,6 @@
+
+ /* 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));
@@ -42742,13 +37815,15 @@
+#endif /* !CONFIG_XEN */
+
+static void __meminit
-+phys_pmd_init(pmd_t *pmd, unsigned long address, unsigned long end)
++phys_pmd_init(pmd_t *pmd_page, unsigned long address, unsigned long end)
+{
-+ int i, k;
++ int i = pmd_index(address);
++ int k;
+
-+ for (i = 0; i < PTRS_PER_PMD; pmd++, i++) {
++ for (; i < PTRS_PER_PMD; i++) {
+ unsigned long pte_phys;
+ pte_t *pte, *pte_save;
++ pmd_t *pmd = pmd_page + pmd_index(address);
+
+ if (address >= end) {
+ if (!after_bootmem)
@@ -42756,6 +37831,10 @@
+ set_pmd(pmd, __pmd(0));
+ break;
+ }
++
++ if (pmd_val(*pmd))
++ continue;
++
+ pte = alloc_static_page(&pte_phys);
+ pte_save = pte;
+ for (k = 0; k < PTRS_PER_PTE; pte++, k++, address += PTE_SIZE) {
@@ -42781,40 +37860,36 @@
+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();
-+ }
++ pmd_t *pmd = pmd_offset(pud,0);
++ 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)
++static void __meminit phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end)
+{
-+ long i = pud_index(address);
-+
-+ pud = pud + i;
++ int i = pud_index(addr);
+
-+ 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;
++ for (; i < PTRS_PER_PUD; i++, addr = (addr & PUD_MASK) + PUD_SIZE ) {
++ unsigned long pmd_phys;
++ pud_t *pud = pud_page + pud_index(addr);
+ pmd_t *pmd;
+
-+ paddr = (address & PGDIR_MASK) + i*PUD_SIZE;
-+ if (paddr >= end)
++ if (addr >= end)
+ break;
+
++ if (pud_val(*pud)) {
++ phys_pmd_update(pud, addr, end);
++ continue;
++ }
++
+ 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);
++ phys_pmd_init(pmd, addr, end);
+ spin_unlock(&init_mm.page_table_lock);
+ }
+ __flush_tlb();
@@ -43018,78 +38093,19 @@
+#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;
-+#ifndef CONFIG_XEN
-+ unsigned long w;
-+#endif
-+
-+ for (i = 0; i < MAX_NR_ZONES; i++)
-+ z[i] = 0;
-+
-+#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;
-+
-+ /* 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];
-+ }
-+ }
-+#else
-+ z[ZONE_DMA] = end_pfn;
-+ for (i = 0; i < MAX_NR_ZONES; i++)
-+ h[i] = 0;
-+#endif
-+}
-+
+#ifndef CONFIG_NUMA
+void __init paging_init(void)
+{
-+ unsigned long zones[MAX_NR_ZONES], holes[MAX_NR_ZONES];
+ int i;
++ unsigned long max_zone_pfns[MAX_NR_ZONES];
++ memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
++ max_zone_pfns[ZONE_DMA] = end_pfn;
++ max_zone_pfns[ZONE_DMA32] = end_pfn;
++ max_zone_pfns[ZONE_NORMAL] = end_pfn;
+
+ 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);
++ free_area_init_nodes(max_zone_pfns);
+
+ /* Switch to the real shared_info page, and clear the
+ * dummy page. */
@@ -43147,7 +38163,6 @@
+/*
+ * Memory hotplug specific functions
+ */
-+
+void online_page(struct page *page)
+{
+ ClearPageReserved(page);
@@ -43159,36 +38174,23 @@
+
+#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;
++ struct zone *zone = pgdat->node_zones + ZONE_NORMAL;
+ unsigned long start_pfn = start >> PAGE_SHIFT;
+ unsigned long nr_pages = size >> PAGE_SHIFT;
+ int ret;
+
++ init_memory_mapping(start, (start + size -1));
++
+ 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__);
@@ -43202,7 +38204,17 @@
+}
+EXPORT_SYMBOL_GPL(remove_memory);
+
-+#else /* CONFIG_MEMORY_HOTPLUG */
++#if !defined(CONFIG_ACPI_NUMA) && defined(CONFIG_NUMA)
++int memory_add_physaddr_to_nid(u64 start)
++{
++ return 0;
++}
++EXPORT_SYMBOL_GPL(memory_add_physaddr_to_nid);
++#endif
++
++#endif /* CONFIG_MEMORY_HOTPLUG */
++
++#ifdef CONFIG_MEMORY_HOTPLUG_RESERVE
+/*
+ * Memory Hotadd without sparsemem. The mem_maps have been allocated in advance,
+ * just online the pages.
@@ -43228,7 +38240,7 @@
+ }
+ return err;
+}
-+#endif /* CONFIG_MEMORY_HOTPLUG */
++#endif
+
+static struct kcore_list kcore_mem, kcore_vmalloc, kcore_kernel, kcore_modules,
+ kcore_vsyscall;
@@ -43245,12 +38257,6 @@
+
+ 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);
+
@@ -43268,7 +38274,9 @@
+ init_page_count(&mem_map[pfn]);
+ totalram_pages++;
+ }
-+ reservedpages = end_pfn - totalram_pages - e820_hole_size(0, end_pfn);
++ reservedpages = end_pfn - totalram_pages -
++ absent_pages_in_range(0, end_pfn);
++
+
+ after_bootmem = 1;
+
@@ -43307,7 +38315,6 @@
+
+void free_init_pages(char *what, unsigned long begin, unsigned long end)
+{
-+#ifdef __DO_LATER__
+ unsigned long addr;
+
+ if (begin >= end)
@@ -43322,14 +38329,13 @@
+ free_page(addr);
+ totalram_pages++;
+ }
-+#endif
+}
+
+void free_initmem(void)
+{
-+#ifdef __DO_LATER__
+ memset(__initdata_begin, POISON_FREE_INITDATA,
+ __initdata_end - __initdata_begin);
++#ifdef __DO_LATER__
+ free_init_pages("unused kernel memory",
+ (unsigned long)(&__init_begin),
+ (unsigned long)(&__init_end));
@@ -43347,6 +38353,7 @@
+
+ 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
@@ -43366,15 +38373,30 @@
+
+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);
++#endif
++ unsigned long pfn = phys >> PAGE_SHIFT;
++ if (pfn >= end_pfn) {
++ /* This can happen with kdump kernels when accessing firmware
++ tables. */
++ if (pfn < end_pfn_map)
++ return;
++ printk(KERN_ERR "reserve_bootmem: illegal reserve %lx %u\n",
++ phys, len);
++ return;
++ }
++
++ /* Should check here against the e820 map to avoid double free */
++#ifdef CONFIG_NUMA
+ reserve_bootmem_node(NODE_DATA(nid), phys, len);
+#else
+ reserve_bootmem(phys, len);
+#endif
-+ if (phys+len <= MAX_DMA_PFN*PAGE_SIZE)
++ if (phys+len <= MAX_DMA_PFN*PAGE_SIZE) {
+ dma_reserve += len / PAGE_SIZE;
++ set_dma_reserve(dma_reserve);
++ }
+}
+
+int kern_addr_valid(unsigned long addr)
@@ -43433,14 +38455,15 @@
+__initcall(x8664_sysctl_init);
+#endif
+
-+/* A pseudo VMAs to allow ptrace access for the vsyscall page. This only
++/* A pseudo VMA 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
++ .vm_end = VSYSCALL_START + (VSYSCALL_MAPPED_PAGES << PAGE_SHIFT),
++ .vm_page_prot = PAGE_READONLY_EXEC,
++ .vm_flags = VM_READ | VM_EXEC
+};
+
+struct vm_area_struct *get_gate_vma(struct task_struct *tsk)
@@ -43468,9 +38491,9 @@
+{
+ return (addr >= VSYSCALL_START) && (addr < VSYSCALL_END);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/mm/Makefile linux-2.6.18-xen/arch/x86_64/mm/Makefile
---- linux-2.6.18.3/arch/x86_64/mm/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/x86_64/mm/Makefile 2006-11-19 14:26:34.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/mm/Makefile b/arch/x86_64/mm/Makefile
+--- a/arch/x86_64/mm/Makefile 2007-03-12 21:58:06.000000000 +0100
++++ b/arch/x86_64/mm/Makefile 2007-03-14 10:55:21.000000000 +0100
@@ -9,3 +9,13 @@
obj-$(CONFIG_ACPI_NUMA) += srat.o
@@ -43485,25 +38508,13 @@
+
+obj-y := $(call cherrypickxen, $(obj-y))
+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/mm/pageattr-xen.c linux-2.6.18-xen/arch/x86_64/mm/pageattr-xen.c
---- linux-2.6.18.3/arch/x86_64/mm/pageattr-xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/arch/x86_64/mm/pageattr-xen.c 2006-12-05 18:42:36.000000000 +0100
-@@ -0,0 +1,398 @@
-+/*
-+ * 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>
-+
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/mm/pageattr.c b/arch/x86_64/mm/pageattr.c
+--- a/arch/x86_64/mm/pageattr.c 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/mm/pageattr.c 2007-03-14 10:55:21.000000000 +0100
+@@ -13,6 +13,167 @@
+ #include <asm/tlbflush.h>
+ #include <asm/io.h>
+
+#ifdef CONFIG_XEN
+#include <asm/pgalloc.h>
+#include <asm/mmu_context.h>
@@ -43665,145 +38676,13 @@
+}
+#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.
-+ */
-+ 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 */
+ static inline pte_t *lookup_address(unsigned long address)
+ {
+ pgd_t *pgd = pgd_offset_k(address);
+@@ -154,7 +315,17 @@
+ 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
@@ -43811,85 +38690,16 @@
+ * replace it with a large page.
+ */
+#ifndef CONFIG_XEN
-+ BUG_ON(PageReserved(kpte_page));
+ BUG_ON(PageReserved(kpte_page));
+#else
+ if(!PageReserved(kpte_page))
+#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 -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/oprofile/Makefile linux-2.6.18-xen/arch/x86_64/oprofile/Makefile
---- linux-2.6.18.3/arch/x86_64/oprofile/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/x86_64/oprofile/Makefile 2006-11-19 14:26:34.000000000 +0100
+
+ if (page_private(kpte_page) == 0) {
+ save_page(kpte_page);
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/oprofile/Makefile b/arch/x86_64/oprofile/Makefile
+--- a/arch/x86_64/oprofile/Makefile 2007-03-12 21:58:06.000000000 +0100
++++ b/arch/x86_64/oprofile/Makefile 2007-03-14 10:55:21.000000000 +0100
@@ -11,9 +11,12 @@
oprofilefs.o oprofile_stats.o \
timer_int.o )
@@ -43904,10 +38714,10 @@
-
+endif
oprofile-y = $(DRIVER_OBJS) $(addprefix ../../i386/oprofile/, $(OPROFILE-y))
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/arch/x86_64/pci/Makefile linux-2.6.18-xen/arch/x86_64/pci/Makefile
---- linux-2.6.18.3/arch/x86_64/pci/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/arch/x86_64/pci/Makefile 2006-11-19 14:26:34.000000000 +0100
-@@ -15,11 +15,23 @@
+diff -x '.hg*' -x '.git*' -urN a/arch/x86_64/pci/Makefile b/arch/x86_64/pci/Makefile
+--- a/arch/x86_64/pci/Makefile 2007-03-16 18:49:21.000000000 +0100
++++ b/arch/x86_64/pci/Makefile 2007-03-14 10:55:21.000000000 +0100
+@@ -15,8 +15,13 @@
obj-$(CONFIG_NUMA) += k8-bus.o
@@ -43921,19 +38731,18 @@
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 -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/acpi/Kconfig linux-2.6.18-xen/drivers/acpi/Kconfig
---- linux-2.6.18.3/drivers/acpi/Kconfig 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/acpi/Kconfig 2006-11-19 14:26:34.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/acpi/Kconfig b/drivers/acpi/Kconfig
+--- a/drivers/acpi/Kconfig 2007-03-16 18:49:21.000000000 +0100
++++ b/drivers/acpi/Kconfig 2007-03-14 10:55:22.000000000 +0100
+@@ -11,7 +11,7 @@
+ bool "ACPI Support"
+ depends on IA64 || X86
+ depends on PCI
+- depends on PM
++ depends on PM || XEN
+ default y
+ ---help---
+ Advanced Configuration and Power Interface (ACPI) support for
@@ -45,7 +45,7 @@
config ACPI_SLEEP
@@ -43943,7 +38752,7 @@
depends on PM
default y
---help---
-@@ -305,6 +305,7 @@
+@@ -309,6 +309,7 @@
config X86_PM_TIMER
bool "Power Management Timer Support" if EMBEDDED
depends on X86
@@ -43951,9 +38760,9 @@
default y
help
The Power Management Timer is available on all ACPI-capable,
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/char/hangcheck-timer.c linux-2.6.18-xen/drivers/char/hangcheck-timer.c
---- linux-2.6.18.3/drivers/char/hangcheck-timer.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/char/hangcheck-timer.c 2006-11-19 14:26:36.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/char/hangcheck-timer.c b/drivers/char/hangcheck-timer.c
+--- a/drivers/char/hangcheck-timer.c 2007-03-16 18:49:21.000000000 +0100
++++ b/drivers/char/hangcheck-timer.c 2007-03-14 10:55:23.000000000 +0100
@@ -117,7 +117,7 @@
__setup("hcheck_dump_tasks", hangcheck_parse_dump_tasks);
#endif /* not MODULE */
@@ -43963,10 +38772,10 @@
# define HAVE_MONOTONIC
# define TIMER_FREQ 1000000000ULL
#elif defined(CONFIG_IA64)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/char/mem.c linux-2.6.18-xen/drivers/char/mem.c
---- linux-2.6.18.3/drivers/char/mem.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/char/mem.c 2006-11-19 14:26:36.000000000 +0100
-@@ -101,6 +101,7 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/char/mem.c b/drivers/char/mem.c
+--- a/drivers/char/mem.c 2007-03-16 18:49:21.000000000 +0100
++++ b/drivers/char/mem.c 2007-03-14 10:55:23.000000000 +0100
+@@ -102,6 +102,7 @@
}
#endif
@@ -43974,7 +38783,7 @@
/*
* This funcion reads the *physical* memory. The f_pos points directly to the
* memory location.
-@@ -223,6 +224,7 @@
+@@ -224,6 +225,7 @@
*ppos += written;
return written;
}
@@ -43982,7 +38791,7 @@
#ifndef __HAVE_PHYS_MEM_ACCESS_PROT
static pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
-@@ -776,6 +778,7 @@
+@@ -810,6 +812,7 @@
#define open_kmem open_mem
#define open_oldmem open_mem
@@ -43990,9 +38799,9 @@
static const struct file_operations mem_fops = {
.llseek = memory_lseek,
.read = read_mem,
-@@ -783,6 +786,9 @@
- .mmap = mmap_mem,
+@@ -818,6 +821,9 @@
.open = open_mem,
+ .get_unmapped_area = get_unmapped_area_mem,
};
+#else
+extern struct file_operations mem_fops;
@@ -44000,9 +38809,9 @@
static const struct file_operations kmem_fops = {
.llseek = memory_lseek,
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/char/tpm/Kconfig linux-2.6.18-xen/drivers/char/tpm/Kconfig
---- linux-2.6.18.3/drivers/char/tpm/Kconfig 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/char/tpm/Kconfig 2006-11-19 14:26:36.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/char/tpm/Kconfig b/drivers/char/tpm/Kconfig
+--- a/drivers/char/tpm/Kconfig 2007-03-12 21:58:07.000000000 +0100
++++ b/drivers/char/tpm/Kconfig 2007-03-14 10:55:24.000000000 +0100
@@ -31,7 +31,7 @@
config TCG_NSC
@@ -44027,18 +38836,18 @@
+ will be called tpm_xenu.
+endmenu
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/char/tpm/Makefile linux-2.6.18-xen/drivers/char/tpm/Makefile
---- linux-2.6.18.3/drivers/char/tpm/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/char/tpm/Makefile 2006-11-19 14:26:36.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/char/tpm/Makefile b/drivers/char/tpm/Makefile
+--- a/drivers/char/tpm/Makefile 2007-03-12 21:58:07.000000000 +0100
++++ b/drivers/char/tpm/Makefile 2007-03-14 10:55:24.000000000 +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 -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/char/tpm/tpm.c linux-2.6.18-xen/drivers/char/tpm/tpm.c
---- linux-2.6.18.3/drivers/char/tpm/tpm.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/char/tpm/tpm.c 2006-11-19 14:26:36.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/char/tpm/tpm.c b/drivers/char/tpm/tpm.c
+--- a/drivers/char/tpm/tpm.c 2007-03-16 18:49:21.000000000 +0100
++++ b/drivers/char/tpm/tpm.c 2007-03-14 10:55:24.000000000 +0100
@@ -30,7 +30,9 @@
enum tpm_const {
@@ -44159,10 +38968,10 @@
return ret_size;
}
EXPORT_SYMBOL_GPL(tpm_read);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/char/tpm/tpm.h linux-2.6.18-xen/drivers/char/tpm/tpm.h
---- linux-2.6.18.3/drivers/char/tpm/tpm.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/char/tpm/tpm.h 2006-12-05 18:42:36.000000000 +0100
-@@ -61,6 +61,7 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/char/tpm/tpm.h b/drivers/char/tpm/tpm.h
+--- a/drivers/char/tpm/tpm.h 2007-03-16 18:49:21.000000000 +0100
++++ b/drivers/char/tpm/tpm.h 2007-03-14 10:55:24.000000000 +0100
+@@ -62,6 +62,7 @@
const u8 req_complete_mask;
const u8 req_complete_val;
const u8 req_canceled;
@@ -44170,7 +38979,7 @@
void __iomem *iobase; /* ioremapped address */
unsigned long base; /* TPM base address */
-@@ -94,6 +95,7 @@
+@@ -95,6 +96,7 @@
/* Data passed to and from the tpm via the read/write calls */
u8 *data_buffer;
atomic_t data_pending;
@@ -44178,7 +38987,7 @@
struct semaphore buffer_mutex;
struct timer_list user_read_timer; /* user needs to claim result */
-@@ -121,6 +123,11 @@
+@@ -122,6 +124,11 @@
outb(value & 0xFF, base+1);
}
@@ -44190,9 +38999,9 @@
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 -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/char/tpm/tpm_vtpm.c linux-2.6.18-xen/drivers/char/tpm/tpm_vtpm.c
---- linux-2.6.18.3/drivers/char/tpm/tpm_vtpm.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/char/tpm/tpm_vtpm.c 2006-11-19 14:26:36.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/char/tpm/tpm_vtpm.c b/drivers/char/tpm/tpm_vtpm.c
+--- a/drivers/char/tpm/tpm_vtpm.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/char/tpm/tpm_vtpm.c 2007-03-14 10:55:24.000000000 +0100
@@ -0,0 +1,547 @@
+/*
+ * Copyright (C) 2006 IBM Corporation
@@ -44741,9 +39550,9 @@
+ tpm_remove_hardware(dev);
+ kfree(vtpms);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/char/tpm/tpm_vtpm.h linux-2.6.18-xen/drivers/char/tpm/tpm_vtpm.h
---- linux-2.6.18.3/drivers/char/tpm/tpm_vtpm.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/char/tpm/tpm_vtpm.h 2006-11-19 14:26:36.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/char/tpm/tpm_vtpm.h b/drivers/char/tpm/tpm_vtpm.h
+--- a/drivers/char/tpm/tpm_vtpm.h 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/char/tpm/tpm_vtpm.h 2007-03-14 10:55:24.000000000 +0100
@@ -0,0 +1,68 @@
+#ifndef TPM_VTPM_H
+#define TPM_VTPM_H
@@ -44813,9 +39622,9 @@
+}
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/char/tpm/tpm_xen.c linux-2.6.18-xen/drivers/char/tpm/tpm_xen.c
---- linux-2.6.18.3/drivers/char/tpm/tpm_xen.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/char/tpm/tpm_xen.c 2006-11-19 14:26:36.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/char/tpm/tpm_xen.c b/drivers/char/tpm/tpm_xen.c
+--- a/drivers/char/tpm/tpm_xen.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/char/tpm/tpm_xen.c 2007-03-14 10:55:24.000000000 +0100
@@ -0,0 +1,756 @@
+/*
+ * Copyright (c) 2005, IBM Corporation
@@ -45573,47 +40382,20 @@
+module_init(tpmif_init);
+
+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/char/tty_io.c linux-2.6.18-xen/drivers/char/tty_io.c
---- linux-2.6.18.3/drivers/char/tty_io.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/char/tty_io.c 2006-11-19 14:26:36.000000000 +0100
-@@ -130,6 +130,8 @@
- 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 */
- extern int pty_limit; /* Config limit on Unix98 ptys */
-@@ -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");
+diff -x '.hg*' -x '.git*' -urN a/drivers/char/tty_io.c b/drivers/char/tty_io.c
+--- a/drivers/char/tty_io.c 2007-03-16 18:49:21.000000000 +0100
++++ b/drivers/char/tty_io.c 2007-03-14 10:55:24.000000000 +0100
+@@ -3917,6 +3917,7 @@
+ device_create(tty_class, NULL, MKDEV(TTY_MAJOR, 0), "tty0");
vty_init();
+ out_vt:
#endif
return 0;
}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/firmware/Kconfig linux-2.6.18-xen/drivers/firmware/Kconfig
---- linux-2.6.18.3/drivers/firmware/Kconfig 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/firmware/Kconfig 2006-11-19 14:26:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/firmware/Kconfig b/drivers/firmware/Kconfig
+--- a/drivers/firmware/Kconfig 2007-03-16 18:49:22.000000000 +0100
++++ b/drivers/firmware/Kconfig 2007-03-14 10:55:24.000000000 +0100
@@ -7,7 +7,7 @@
config EDD
@@ -45623,10 +40405,10 @@
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 -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/ide/ide-lib.c linux-2.6.18-xen/drivers/ide/ide-lib.c
---- linux-2.6.18.3/drivers/ide/ide-lib.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/ide/ide-lib.c 2006-11-19 14:26:39.000000000 +0100
-@@ -408,10 +408,10 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/ide/ide-lib.c b/drivers/ide/ide-lib.c
+--- a/drivers/ide/ide-lib.c 2007-03-16 18:49:22.000000000 +0100
++++ b/drivers/ide/ide-lib.c 2007-03-14 10:55:24.000000000 +0100
+@@ -429,10 +429,10 @@
{
u64 addr = BLK_BOUNCE_HIGH; /* dma64_addr_t */
@@ -45641,20 +40423,20 @@
addr = HWIF(drive)->pci_dev->dma_mask;
}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/Makefile linux-2.6.18-xen/drivers/Makefile
---- linux-2.6.18.3/drivers/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/Makefile 2006-11-19 14:26:34.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/Makefile b/drivers/Makefile
+--- a/drivers/Makefile 2007-03-16 18:49:21.000000000 +0100
++++ b/drivers/Makefile 2007-03-14 10:55:22.000000000 +0100
@@ -31,6 +31,7 @@
obj-$(CONFIG_NUBUS) += nubus/
obj-$(CONFIG_ATM) += atm/
- obj-$(CONFIG_PPC_PMAC) += macintosh/
+ obj-y += macintosh/
+obj-$(CONFIG_XEN) += xen/
obj-$(CONFIG_IDE) += ide/
obj-$(CONFIG_FC4) += fc4/
obj-$(CONFIG_SCSI) += scsi/
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/oprofile/buffer_sync.c linux-2.6.18-xen/drivers/oprofile/buffer_sync.c
---- linux-2.6.18.3/drivers/oprofile/buffer_sync.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/oprofile/buffer_sync.c 2006-11-19 14:26:47.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/oprofile/buffer_sync.c b/drivers/oprofile/buffer_sync.c
+--- a/drivers/oprofile/buffer_sync.c 2007-03-16 18:49:24.000000000 +0100
++++ b/drivers/oprofile/buffer_sync.c 2007-03-14 10:55:31.000000000 +0100
@@ -6,6 +6,10 @@
*
* @author John Levon <levon at movementarian.org>
@@ -45775,9 +40557,9 @@
}
}
}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/oprofile/cpu_buffer.c linux-2.6.18-xen/drivers/oprofile/cpu_buffer.c
---- linux-2.6.18.3/drivers/oprofile/cpu_buffer.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/oprofile/cpu_buffer.c 2006-11-19 14:26:47.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/oprofile/cpu_buffer.c b/drivers/oprofile/cpu_buffer.c
+--- a/drivers/oprofile/cpu_buffer.c 2007-03-16 18:49:24.000000000 +0100
++++ b/drivers/oprofile/cpu_buffer.c 2007-03-14 10:55:31.000000000 +0100
@@ -6,6 +6,10 @@
*
* @author John Levon <levon at movementarian.org>
@@ -45887,9 +40669,9 @@
/*
* This serves to avoid cpu buffer overflow, and makes sure
* the task mortuary progresses
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/oprofile/cpu_buffer.h linux-2.6.18-xen/drivers/oprofile/cpu_buffer.h
---- linux-2.6.18.3/drivers/oprofile/cpu_buffer.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/oprofile/cpu_buffer.h 2006-11-19 14:26:47.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/oprofile/cpu_buffer.h b/drivers/oprofile/cpu_buffer.h
+--- a/drivers/oprofile/cpu_buffer.h 2007-03-16 18:49:24.000000000 +0100
++++ b/drivers/oprofile/cpu_buffer.h 2007-03-14 10:55:31.000000000 +0100
@@ -36,7 +36,7 @@
volatile unsigned long tail_pos;
unsigned long buffer_size;
@@ -45912,9 +40694,9 @@
+#define CPU_DOMAIN_SWITCH 4
#endif /* OPROFILE_CPU_BUFFER_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/oprofile/event_buffer.h linux-2.6.18-xen/drivers/oprofile/event_buffer.h
---- linux-2.6.18.3/drivers/oprofile/event_buffer.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/oprofile/event_buffer.h 2006-11-19 14:26:47.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/oprofile/event_buffer.h b/drivers/oprofile/event_buffer.h
+--- a/drivers/oprofile/event_buffer.h 2007-03-15 15:56:07.000000000 +0100
++++ b/drivers/oprofile/event_buffer.h 2007-03-14 10:55:31.000000000 +0100
@@ -29,15 +29,20 @@
#define CPU_SWITCH_CODE 2
#define COOKIE_SWITCH_CODE 3
@@ -45937,9 +40719,9 @@
/* add data to the event buffer */
void add_event_entry(unsigned long data);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/oprofile/oprof.c linux-2.6.18-xen/drivers/oprofile/oprof.c
---- linux-2.6.18.3/drivers/oprofile/oprof.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/oprofile/oprof.c 2006-11-19 14:26:47.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/oprofile/oprof.c b/drivers/oprofile/oprof.c
+--- a/drivers/oprofile/oprof.c 2007-03-12 21:58:09.000000000 +0100
++++ b/drivers/oprofile/oprof.c 2007-03-14 10:55:31.000000000 +0100
@@ -5,6 +5,10 @@
* @remark Read the file COPYING
*
@@ -45993,9 +40775,9 @@
int oprofile_setup(void)
{
int err;
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/oprofile/oprof.h linux-2.6.18-xen/drivers/oprofile/oprof.h
---- linux-2.6.18.3/drivers/oprofile/oprof.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/oprofile/oprof.h 2006-11-19 14:26:47.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/oprofile/oprof.h b/drivers/oprofile/oprof.h
+--- a/drivers/oprofile/oprof.h 2007-03-12 21:58:09.000000000 +0100
++++ b/drivers/oprofile/oprof.h 2007-03-14 10:55:31.000000000 +0100
@@ -35,5 +35,8 @@
void oprofile_timer_init(struct oprofile_operations * ops);
@@ -46005,9 +40787,9 @@
+int oprofile_set_passive(int passive_domains[], unsigned int pdomains);
#endif /* OPROF_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/oprofile/oprofile_files.c linux-2.6.18-xen/drivers/oprofile/oprofile_files.c
---- linux-2.6.18.3/drivers/oprofile/oprofile_files.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/oprofile/oprofile_files.c 2006-11-19 14:26:47.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/oprofile/oprofile_files.c b/drivers/oprofile/oprofile_files.c
+--- a/drivers/oprofile/oprofile_files.c 2007-03-15 15:56:07.000000000 +0100
++++ b/drivers/oprofile/oprofile_files.c 2007-03-14 10:55:31.000000000 +0100
@@ -5,15 +5,21 @@
* @remark Read the file COPYING
*
@@ -46241,9 +41023,9 @@
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 -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/pci/Kconfig linux-2.6.18-xen/drivers/pci/Kconfig
---- linux-2.6.18.3/drivers/pci/Kconfig 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/pci/Kconfig 2006-11-19 14:26:47.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/pci/Kconfig b/drivers/pci/Kconfig
+--- a/drivers/pci/Kconfig 2007-03-16 18:49:24.000000000 +0100
++++ b/drivers/pci/Kconfig 2007-03-14 10:55:32.000000000 +0100
@@ -5,6 +5,7 @@
bool "Message Signaled Interrupts (MSI and MSI-X)"
depends on PCI
@@ -46252,9 +41034,18 @@
help
This allows device drivers to enable MSI (Message Signaled
Interrupts). Message Signaled Interrupts enable a device to
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/serial/Kconfig linux-2.6.18-xen/drivers/serial/Kconfig
---- linux-2.6.18.3/drivers/serial/Kconfig 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/drivers/serial/Kconfig 2006-11-19 14:26:51.000000000 +0100
+@@ -55,7 +56,7 @@
+ config HT_IRQ
+ bool "Interrupts on hypertransport devices"
+ default y
+- depends on PCI && X86_LOCAL_APIC && X86_IO_APIC
++ depends on PCI && X86_LOCAL_APIC && X86_IO_APIC && !XEN
+ help
+ This allows native hypertransport devices to use interrupts.
+
+diff -x '.hg*' -x '.git*' -urN a/drivers/serial/Kconfig b/drivers/serial/Kconfig
+--- a/drivers/serial/Kconfig 2007-03-16 18:49:34.000000000 +0100
++++ b/drivers/serial/Kconfig 2007-03-14 10:55:34.000000000 +0100
@@ -11,6 +11,7 @@
config SERIAL_8250
tristate "8250/16550 and compatible serial support"
@@ -46263,9 +41054,9 @@
select SERIAL_CORE
---help---
This selects whether you want to include the driver for the standard
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/balloon/balloon.c linux-2.6.18-xen/drivers/xen/balloon/balloon.c
---- linux-2.6.18.3/drivers/xen/balloon/balloon.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/balloon/balloon.c 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/balloon/balloon.c b/drivers/xen/balloon/balloon.c
+--- a/drivers/xen/balloon/balloon.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/balloon/balloon.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,637 @@
+/******************************************************************************
+ * balloon.c
@@ -46363,8 +41154,8 @@
+static unsigned long balloon_low, balloon_high;
+
+/* Main work function, always executed in process context. */
-+static void balloon_process(void *unused);
-+static DECLARE_WORK(balloon_worker, balloon_process, NULL);
++static void balloon_process(struct work_struct *unusable);
++static DECLARE_WORK(balloon_worker, balloon_process);
+static struct timer_list balloon_timer;
+
+/* When ballooning out (allocating memory to return to Xen) we don't really
@@ -46599,7 +41390,7 @@
+ * 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 balloon_process(struct work_struct *unusable)
+{
+ int need_sleep = 0;
+ long credit;
@@ -46904,15 +41695,15 @@
+EXPORT_SYMBOL_GPL(balloon_release_driver_page);
+
+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/balloon/Makefile linux-2.6.18-xen/drivers/xen/balloon/Makefile
---- linux-2.6.18.3/drivers/xen/balloon/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/balloon/Makefile 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/balloon/Makefile b/drivers/xen/balloon/Makefile
+--- a/drivers/xen/balloon/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/balloon/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,2 @@
+
+obj-y += balloon.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blkback/blkback.c linux-2.6.18-xen/drivers/xen/blkback/blkback.c
---- linux-2.6.18.3/drivers/xen/blkback/blkback.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blkback/blkback.c 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blkback/blkback.c b/drivers/xen/blkback/blkback.c
+--- a/drivers/xen/blkback/blkback.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blkback/blkback.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,549 @@
+/******************************************************************************
+ * arch/xen/drivers/blkif/backend/main.c
@@ -47195,7 +41986,7 @@
+ wake_up(&blkif->wq);
+}
+
-+irqreturn_t blkif_be_int(int irq, void *dev_id, struct pt_regs *regs)
++irqreturn_t blkif_be_int(int irq, void *dev_id)
+{
+ blkif_notify_work(dev_id);
+ return IRQ_HANDLED;
@@ -47463,9 +42254,9 @@
+module_init(blkif_init);
+
+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blkback/common.h linux-2.6.18-xen/drivers/xen/blkback/common.h
---- linux-2.6.18.3/drivers/xen/blkback/common.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blkback/common.h 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blkback/common.h b/drivers/xen/blkback/common.h
+--- a/drivers/xen/blkback/common.h 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blkback/common.h 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,133 @@
+/*
+ * This program is free software; you can redistribute it and/or
@@ -47596,13 +42387,13 @@
+
+void blkif_xenbus_init(void);
+
-+irqreturn_t blkif_be_int(int irq, void *dev_id, struct pt_regs *regs);
++irqreturn_t blkif_be_int(int irq, void *dev_id);
+int blkif_schedule(void *arg);
+
+#endif /* __BLKIF__BACKEND__COMMON_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blkback/interface.c linux-2.6.18-xen/drivers/xen/blkback/interface.c
---- linux-2.6.18.3/drivers/xen/blkback/interface.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blkback/interface.c 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blkback/interface.c b/drivers/xen/blkback/interface.c
+--- a/drivers/xen/blkback/interface.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blkback/interface.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,171 @@
+/******************************************************************************
+ * arch/xen/drivers/blkif/backend/interface.c
@@ -47640,7 +42431,7 @@
+#include <xen/evtchn.h>
+#include <linux/kthread.h>
+
-+static kmem_cache_t *blkif_cachep;
++static struct kmem_cache *blkif_cachep;
+
+blkif_t *blkif_alloc(domid_t domid)
+{
@@ -47775,16 +42566,16 @@
+ blkif_cachep = kmem_cache_create("blkif_cache", sizeof(blkif_t),
+ 0, 0, NULL, NULL);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blkback/Makefile linux-2.6.18-xen/drivers/xen/blkback/Makefile
---- linux-2.6.18.3/drivers/xen/blkback/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blkback/Makefile 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blkback/Makefile b/drivers/xen/blkback/Makefile
+--- a/drivers/xen/blkback/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blkback/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,3 @@
+obj-$(CONFIG_XEN_BLKDEV_BACKEND) := blkbk.o
+
+blkbk-y := blkback.o xenbus.o interface.o vbd.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blkback/vbd.c linux-2.6.18-xen/drivers/xen/blkback/vbd.c
---- linux-2.6.18.3/drivers/xen/blkback/vbd.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blkback/vbd.c 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blkback/vbd.c b/drivers/xen/blkback/vbd.c
+--- a/drivers/xen/blkback/vbd.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blkback/vbd.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,119 @@
+/******************************************************************************
+ * blkback/vbd.c
@@ -47905,9 +42696,9 @@
+ out:
+ return rc;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blkback/xenbus.c linux-2.6.18-xen/drivers/xen/blkback/xenbus.c
---- linux-2.6.18.3/drivers/xen/blkback/xenbus.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blkback/xenbus.c 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blkback/xenbus.c b/drivers/xen/blkback/xenbus.c
+--- a/drivers/xen/blkback/xenbus.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blkback/xenbus.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,468 @@
+/* Xenbus code for blkif backend
+ Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
@@ -48377,10 +43168,10 @@
+{
+ xenbus_register_backend(&blkback);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blkfront/blkfront.c linux-2.6.18-xen/drivers/xen/blkfront/blkfront.c
---- linux-2.6.18.3/drivers/xen/blkfront/blkfront.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blkfront/blkfront.c 2006-12-05 18:42:37.000000000 +0100
-@@ -0,0 +1,846 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blkfront/blkfront.c b/drivers/xen/blkfront/blkfront.c
+--- a/drivers/xen/blkfront/blkfront.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blkfront/blkfront.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,847 @@
+/******************************************************************************
+ * blkfront.c
+ *
@@ -48447,8 +43238,8 @@
+
+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 irqreturn_t blkif_int(int irq, void *dev_id);
++static void blkif_restart_queue(struct work_struct *work);
+static void blkif_recover(struct blkfront_info *);
+static void blkif_completion(struct blk_shadow *);
+static void blkif_free(struct blkfront_info *, int);
@@ -48483,7 +43274,7 @@
+ info->xbdev = dev;
+ info->vdevice = vdevice;
+ info->connected = BLKIF_STATE_DISCONNECTED;
-+ INIT_WORK(&info->work, blkif_restart_queue, (void *)info);
++ INIT_WORK(&info->work, blkif_restart_queue);
+
+ for (i = 0; i < BLK_RING_SIZE; i++)
+ info->shadow[i].req.id = i+1;
@@ -48801,9 +43592,10 @@
+ }
+}
+
-+static void blkif_restart_queue(void *arg)
++static void blkif_restart_queue(struct work_struct *work)
+{
-+ struct blkfront_info *info = (struct blkfront_info *)arg;
++ struct blkfront_info *info = container_of(work, struct blkfront_info, work);
++
+ spin_lock_irq(&blkif_io_lock);
+ if (info->connected == BLKIF_STATE_CONNECTED)
+ kick_pending_request_queues(info);
@@ -48997,9 +43789,9 @@
+ if (RING_FULL(&info->ring))
+ goto wait;
+
-+ DPRINTK("do_blk_req %p: cmd %p, sec %lx, "
++ DPRINTK("do_blk_req %p: cmd %p, sec %llx, "
+ "(%u/%li) buffer:%p [%s]\n",
-+ req, req->cmd, req->sector, req->current_nr_sectors,
++ req, req->cmd, (u64)req->sector, req->current_nr_sectors,
+ req->nr_sectors, req->buffer,
+ rq_data_dir(req) ? "write" : "read");
+
@@ -49021,7 +43813,7 @@
+}
+
+
-+static irqreturn_t blkif_int(int irq, void *dev_id, struct pt_regs *ptregs)
++static irqreturn_t blkif_int(int irq, void *dev_id)
+{
+ struct request *req;
+ blkif_response_t *bret;
@@ -49227,9 +44019,9 @@
+module_exit(xlblk_exit);
+
+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blkfront/block.h linux-2.6.18-xen/drivers/xen/blkfront/block.h
---- linux-2.6.18.3/drivers/xen/blkfront/block.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blkfront/block.h 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blkfront/block.h b/drivers/xen/blkfront/block.h
+--- a/drivers/xen/blkfront/block.h 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blkfront/block.h 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,154 @@
+/******************************************************************************
+ * block.h
@@ -49385,9 +44177,9 @@
+void xlvbd_del(struct blkfront_info *info);
+
+#endif /* __XEN_DRIVERS_BLOCK_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blkfront/Kconfig linux-2.6.18-xen/drivers/xen/blkfront/Kconfig
---- linux-2.6.18.3/drivers/xen/blkfront/Kconfig 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blkfront/Kconfig 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blkfront/Kconfig b/drivers/xen/blkfront/Kconfig
+--- a/drivers/xen/blkfront/Kconfig 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blkfront/Kconfig 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,6 @@
+
+config XENBLOCK
@@ -49395,18 +44187,18 @@
+ depends on ARCH_XEN
+ help
+ Block device driver for Xen
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blkfront/Makefile linux-2.6.18-xen/drivers/xen/blkfront/Makefile
---- linux-2.6.18.3/drivers/xen/blkfront/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blkfront/Makefile 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blkfront/Makefile b/drivers/xen/blkfront/Makefile
+--- a/drivers/xen/blkfront/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blkfront/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,5 @@
+
+obj-$(CONFIG_XEN_BLKDEV_FRONTEND) := xenblk.o
+
+xenblk-objs := blkfront.o vbd.o
+
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blkfront/vbd.c linux-2.6.18-xen/drivers/xen/blkfront/vbd.c
---- linux-2.6.18.3/drivers/xen/blkfront/vbd.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blkfront/vbd.c 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blkfront/vbd.c b/drivers/xen/blkfront/vbd.c
+--- a/drivers/xen/blkfront/vbd.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blkfront/vbd.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,318 @@
+/******************************************************************************
+ * vbd.c
@@ -49726,10 +44518,10 @@
+ blk_cleanup_queue(info->rq);
+ info->rq = NULL;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blktap/blktapmain.c linux-2.6.18-xen/drivers/xen/blktap/blktapmain.c
---- linux-2.6.18.3/drivers/xen/blktap/blktapmain.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blktap/blktapmain.c 2006-12-05 18:42:37.000000000 +0100
-@@ -0,0 +1,1393 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blktap/blktapmain.c b/drivers/xen/blktap/blktapmain.c
+--- a/drivers/xen/blktap/blktapmain.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blktap/blktapmain.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,1389 @@
+/******************************************************************************
+ * drivers/xen/blktap/blktap.c
+ *
@@ -49780,6 +44572,7 @@
+#include <linux/major.h>
+#include <linux/gfp.h>
+#include <linux/poll.h>
++#include <linux/init.h>
+#include <asm/tlbflush.h>
+
+#define MAX_TAP_DEV 100 /*the maximum number of tapdisk ring devices */
@@ -49829,6 +44622,8 @@
+ ((_req) * BLKIF_MAX_SEGMENTS_PER_REQUEST * PAGE_SIZE) + \
+ ((_seg) * PAGE_SIZE))
+static int blkif_reqs = MAX_PENDING_REQS;
++module_param(blkif_reqs, int, 0);
++
+static int mmap_pages = MMAP_PAGES;
+
+#define RING_PAGES 1 /* BLKTAP - immediately before the mmap area, we
@@ -49866,13 +44661,6 @@
+static domid_translate_t translate_domid[MAX_TAP_DEV];
+static tap_blkif_t *tapfds[MAX_TAP_DEV];
+
-+static int __init set_blkif_reqs(char *str)
-+{
-+ get_option(&str, &blkif_reqs);
-+ return 1;
-+}
-+__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;
@@ -50746,7 +45534,7 @@
+ wake_up(&blkif->wq);
+}
+
-+irqreturn_t tap_blkif_be_int(int irq, void *dev_id, struct pt_regs *regs)
++irqreturn_t tap_blkif_be_int(int irq, void *dev_id)
+{
+ blkif_notify_work(dev_id);
+ return IRQ_HANDLED;
@@ -51123,9 +45911,9 @@
+module_init(blkif_init);
+
+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blktap/common.h linux-2.6.18-xen/drivers/xen/blktap/common.h
---- linux-2.6.18.3/drivers/xen/blktap/common.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blktap/common.h 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blktap/common.h b/drivers/xen/blktap/common.h
+--- a/drivers/xen/blktap/common.h 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blktap/common.h 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,120 @@
+/*
+ * This program is free software; you can redistribute it and/or
@@ -51240,16 +46028,16 @@
+
+void tap_blkif_xenbus_init(void);
+
-+irqreturn_t tap_blkif_be_int(int irq, void *dev_id, struct pt_regs *regs);
++irqreturn_t tap_blkif_be_int(int irq, void *dev_id);
+int tap_blkif_schedule(void *arg);
+
+int dom_to_devid(domid_t domid, int xenbus_id, blkif_t *blkif);
+void signal_tapdisk(int idx);
+
+#endif /* __BLKIF__BACKEND__COMMON_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blktap/interface.c linux-2.6.18-xen/drivers/xen/blktap/interface.c
---- linux-2.6.18.3/drivers/xen/blktap/interface.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blktap/interface.c 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blktap/interface.c b/drivers/xen/blktap/interface.c
+--- a/drivers/xen/blktap/interface.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blktap/interface.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,164 @@
+/******************************************************************************
+ * drivers/xen/blktap/interface.c
@@ -51287,7 +46075,7 @@
+#include "common.h"
+#include <xen/evtchn.h>
+
-+static kmem_cache_t *blkif_cachep;
++static struct kmem_cache *blkif_cachep;
+
+blkif_t *tap_alloc_blkif(domid_t domid)
+{
@@ -51415,9 +46203,9 @@
+ blkif_cachep = kmem_cache_create("blktapif_cache", sizeof(blkif_t),
+ 0, 0, NULL, NULL);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blktap/Makefile linux-2.6.18-xen/drivers/xen/blktap/Makefile
---- linux-2.6.18.3/drivers/xen/blktap/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blktap/Makefile 2006-12-05 18:42:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blktap/Makefile b/drivers/xen/blktap/Makefile
+--- a/drivers/xen/blktap/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blktap/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,6 @@
+LINUXINCLUDE += -I../xen/include/public/io
+
@@ -51425,9 +46213,9 @@
+
+blktap-y := xenbus.o interface.o blktapmain.o
+
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/blktap/xenbus.c linux-2.6.18-xen/drivers/xen/blktap/xenbus.c
---- linux-2.6.18.3/drivers/xen/blktap/xenbus.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/blktap/xenbus.c 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/blktap/xenbus.c b/drivers/xen/blktap/xenbus.c
+--- a/drivers/xen/blktap/xenbus.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/blktap/xenbus.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,366 @@
+/* drivers/xen/blktap/xenbus.c
+ *
@@ -51795,15 +46583,15 @@
+{
+ xenbus_register_backend(&blktap);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/char/Makefile linux-2.6.18-xen/drivers/xen/char/Makefile
---- linux-2.6.18.3/drivers/xen/char/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/char/Makefile 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/char/Makefile b/drivers/xen/char/Makefile
+--- a/drivers/xen/char/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/char/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,2 @@
+
+obj-y := mem.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/char/mem.c linux-2.6.18-xen/drivers/xen/char/mem.c
---- linux-2.6.18.3/drivers/xen/char/mem.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/char/mem.c 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/char/mem.c b/drivers/xen/char/mem.c
+--- a/drivers/xen/char/mem.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/char/mem.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,180 @@
+/*
+ * Originally from linux/drivers/char/mem.c
@@ -51985,10 +46773,10 @@
+ .mmap = mmap_mem,
+ .open = open_mem,
+};
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/console/console.c linux-2.6.18-xen/drivers/xen/console/console.c
---- linux-2.6.18.3/drivers/xen/console/console.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/console/console.c 2006-11-19 14:26:56.000000000 +0100
-@@ -0,0 +1,688 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/console/console.c b/drivers/xen/console/console.c
+--- a/drivers/xen/console/console.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/console/console.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,714 @@
+/******************************************************************************
+ * console.c
+ *
@@ -52048,6 +46836,7 @@
+#include <xen/interface/event_channel.h>
+#include <asm/hypervisor.h>
+#include <xen/evtchn.h>
++#include <xen/xenbus.h>
+#include <xen/xencons.h>
+
+/*
@@ -52055,17 +46844,23 @@
+ * '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'.
+ * [XC_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_DEFAULT, XC_TTY, XC_SERIAL } xc_mode = XC_DEFAULT;
++static enum {
++ XC_OFF, XC_DEFAULT, XC_TTY, XC_SERIAL, XC_XVC
++} xc_mode = XC_DEFAULT;
+static int xc_num = -1;
+
++/* /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
+
+static int __init xencons_setup(char *str)
@@ -52073,28 +46868,24 @@
+ char *q;
+ int n;
+
-+ if (!strncmp(str, "ttyS", 4))
++ if (!strncmp(str, "ttyS", 4)) {
+ xc_mode = XC_SERIAL;
-+ else if (!strncmp(str, "tty", 3))
++ str += 4;
++ } else if (!strncmp(str, "tty", 3)) {
+ xc_mode = XC_TTY;
-+ else if (!strncmp(str, "off", 3))
-+ xc_mode = XC_OFF;
-+
-+ switch (xc_mode) {
-+ case XC_SERIAL:
-+ n = simple_strtol(str+4, &q, 10);
-+ if (q > (str + 4))
-+ xc_num = n;
-+ break;
-+ case XC_TTY:
-+ n = simple_strtol(str+3, &q, 10);
-+ if (q > (str + 3))
-+ xc_num = n;
-+ break;
-+ default:
-+ break;
++ str += 3;
++ } else if (!strncmp(str, "xvc", 3)) {
++ xc_mode = XC_XVC;
++ str += 3;
++ } else if (!strncmp(str, "off", 3)) {
++ xc_mode = XC_OFF;
++ str += 3;
+ }
+
++ n = simple_strtol(str, &q, 10);
++ if (q != str)
++ xc_num = n;
++
+ return 1;
+}
+__setup("xencons=", xencons_setup);
@@ -52187,11 +46978,17 @@
+ if (!xen_start_info->console.domU.evtchn)
+ goto out;
+ if (xc_mode == XC_DEFAULT)
-+ xc_mode = XC_TTY;
++ xc_mode = XC_XVC;
+ kcons_info.write = kcons_write;
+ }
+
+ 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)
@@ -52296,16 +47093,16 @@
+/******************** User-space console driver (/dev/console) ************/
+
+#define DRV(_d) (_d)
-+#define DUMMY_TTY(_tty) ((xc_mode != XC_SERIAL) && \
++#define DUMMY_TTY(_tty) ((xc_mode == XC_TTY) && \
+ ((_tty)->index != (xc_num - 1)))
+
-+static struct termios *xencons_termios[MAX_NR_CONSOLES];
-+static struct termios *xencons_termios_locked[MAX_NR_CONSOLES];
++static struct ktermios *xencons_termios[MAX_NR_CONSOLES];
++static struct ktermios *xencons_termios_locked[MAX_NR_CONSOLES];
+static struct tty_struct *xencons_tty;
+static int xencons_priv_irq;
+static char x_char;
+
-+void xencons_rx(char *buf, unsigned len, struct pt_regs *regs)
++void xencons_rx(char *buf, unsigned len)
+{
+ int i;
+ unsigned long flags;
@@ -52316,7 +47113,7 @@
+
+ for (i = 0; i < len; i++) {
+#ifdef CONFIG_MAGIC_SYSRQ
-+ if (sysrq_enabled) {
++ if (sysrq_on()) {
+ if (buf[i] == '\x0f') { /* ^O */
+ sysrq_requested = jiffies;
+ continue; /* don't print the sysrq key */
@@ -52328,7 +47125,7 @@
+ spin_unlock_irqrestore(
+ &xencons_lock, flags);
+ handle_sysrq(
-+ buf[i], regs, xencons_tty);
++ buf[i], xencons_tty);
+ spin_lock_irqsave(
+ &xencons_lock, flags);
+ continue;
@@ -52393,14 +47190,13 @@
+}
+
+/* Privileged receive callback and transmit kicker. */
-+static irqreturn_t xencons_priv_interrupt(int irq, void *dev_id,
-+ struct pt_regs *regs)
++static irqreturn_t xencons_priv_interrupt(int irq, void *dev_id)
+{
+ static char rbuf[16];
+ int l;
+
+ while ((l = HYPERVISOR_console_io(CONSOLEIO_read, 16, rbuf)) > 0)
-+ xencons_rx(rbuf, l, regs);
++ xencons_rx(rbuf, l);
+
+ xencons_tx();
+
@@ -52619,8 +47415,8 @@
+ return rc;
+ }
+
-+ xencons_driver = alloc_tty_driver((xc_mode == XC_SERIAL) ?
-+ 1 : MAX_NR_CONSOLES);
++ xencons_driver = alloc_tty_driver((xc_mode == XC_TTY) ?
++ MAX_NR_CONSOLES : 1);
+ if (xencons_driver == NULL)
+ return -ENOMEM;
+
@@ -52635,14 +47431,23 @@
+ DRV(xencons_driver)->termios = xencons_termios;
+ DRV(xencons_driver)->termios_locked = xencons_termios_locked;
+
-+ if (xc_mode == XC_SERIAL) {
++ 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 = 0 + xc_num;
-+ } else {
++ 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;
+ }
+
+ tty_set_operations(xencons_driver, &xencons_ops);
@@ -52671,21 +47476,30 @@
+ printk("Xen virtual console successfully installed as %s%d\n",
+ DRV(xencons_driver)->name, xc_num);
+
++ /* Check about framebuffer messing up the console */
++ if (!is_initial_xendomain() &&
++ !xenbus_exists(XBT_NIL, "device", "vfb")) {
++ /* FIXME: this is ugly */
++ unregister_console(&kcons_info);
++ kcons_info.flags |= CON_CONSDEV;
++ register_console(&kcons_info);
++ }
++
+ return 0;
+}
+
+module_init(xencons_init);
+
+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/console/Makefile linux-2.6.18-xen/drivers/xen/console/Makefile
---- linux-2.6.18.3/drivers/xen/console/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/console/Makefile 2006-11-19 14:26:55.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/console/Makefile b/drivers/xen/console/Makefile
+--- a/drivers/xen/console/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/console/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,2 @@
+
+obj-y := console.o xencons_ring.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/console/xencons_ring.c linux-2.6.18-xen/drivers/xen/console/xencons_ring.c
---- linux-2.6.18.3/drivers/xen/console/xencons_ring.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/console/xencons_ring.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/console/xencons_ring.c b/drivers/xen/console/xencons_ring.c
+--- a/drivers/xen/console/xencons_ring.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/console/xencons_ring.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,143 @@
+/*
+ * This program is free software; you can redistribute it and/or
@@ -52772,7 +47586,7 @@
+ return sent;
+}
+
-+static irqreturn_t handle_input(int irq, void *unused, struct pt_regs *regs)
++static irqreturn_t handle_input(int irq, void *unused)
+{
+ struct xencons_interface *intf = xencons_interface();
+ XENCONS_RING_IDX cons, prod;
@@ -52783,7 +47597,7 @@
+ BUG_ON((prod - cons) > sizeof(intf->in));
+
+ while (cons != prod) {
-+ xencons_rx(intf->in+MASK_XENCONS_IDX(cons,intf->in), 1, regs);
++ xencons_rx(intf->in+MASK_XENCONS_IDX(cons,intf->in), 1);
+ cons++;
+ }
+
@@ -52830,9 +47644,9 @@
+{
+ (void)xencons_ring_init();
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/core/cpu_hotplug.c linux-2.6.18-xen/drivers/xen/core/cpu_hotplug.c
---- linux-2.6.18.3/drivers/xen/core/cpu_hotplug.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/core/cpu_hotplug.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/core/cpu_hotplug.c b/drivers/xen/core/cpu_hotplug.c
+--- a/drivers/xen/core/cpu_hotplug.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/core/cpu_hotplug.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,188 @@
+
+#include <linux/init.h>
@@ -53022,10 +47836,10 @@
+{
+ xenbus_allowed_cpumask = cpu_present_map;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/core/evtchn.c linux-2.6.18-xen/drivers/xen/core/evtchn.c
---- linux-2.6.18.3/drivers/xen/core/evtchn.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/core/evtchn.c 2006-11-19 14:26:56.000000000 +0100
-@@ -0,0 +1,868 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/core/evtchn.c b/drivers/xen/core/evtchn.c
+--- a/drivers/xen/core/evtchn.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/core/evtchn.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,840 @@
+/******************************************************************************
+ * evtchn.c
+ *
@@ -53401,7 +48215,7 @@
+
+int bind_evtchn_to_irqhandler(
+ unsigned int evtchn,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ irq_handler_t handler,
+ unsigned long irqflags,
+ const char *devname,
+ void *dev_id)
@@ -53423,7 +48237,7 @@
+int bind_virq_to_irqhandler(
+ unsigned int virq,
+ unsigned int cpu,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ irq_handler_t handler,
+ unsigned long irqflags,
+ const char *devname,
+ void *dev_id)
@@ -53445,7 +48259,7 @@
+int bind_ipi_to_irqhandler(
+ unsigned int ipi,
+ unsigned int cpu,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ irq_handler_t handler,
+ unsigned long irqflags,
+ const char *devname,
+ void *dev_id)
@@ -53500,13 +48314,12 @@
+ rebind_irq_to_cpu(irq, tcpu);
+}
+
-+static int retrigger(unsigned int irq)
++static int retrigger_vector(unsigned int irq)
+{
+ int evtchn = evtchn_from_irq(irq);
+ shared_info_t *s = HYPERVISOR_shared_info;
+ if (!VALID_EVTCHN(evtchn))
+ return 1;
-+ BUG_ON(!synch_test_bit(evtchn, &s->evtchn_mask[0]));
+ synch_set_bit(evtchn, &s->evtchn_pending[0]);
+ return 1;
+}
@@ -53515,7 +48328,7 @@
+ * Interface to generic handling in irq.c
+ */
+
-+static unsigned int startup_dynirq(unsigned int irq)
++static unsigned int startup_dynirq_vector(unsigned int irq)
+{
+ int evtchn = evtchn_from_irq(irq);
+
@@ -53524,15 +48337,7 @@
+ return 0;
+}
+
-+static void shutdown_dynirq(unsigned int irq)
-+{
-+ int evtchn = evtchn_from_irq(irq);
-+
-+ if (VALID_EVTCHN(evtchn))
-+ mask_evtchn(evtchn);
-+}
-+
-+static void enable_dynirq(unsigned int irq)
++static void unmask_dynirq_vector(unsigned int irq)
+{
+ int evtchn = evtchn_from_irq(irq);
+
@@ -53540,7 +48345,7 @@
+ unmask_evtchn(evtchn);
+}
+
-+static void disable_dynirq(unsigned int irq)
++static void mask_dynirq_vector(unsigned int irq)
+{
+ int evtchn = evtchn_from_irq(irq);
+
@@ -53548,7 +48353,7 @@
+ mask_evtchn(evtchn);
+}
+
-+static void ack_dynirq(unsigned int irq)
++static void ack_dynirq_vector(unsigned int irq)
+{
+ int evtchn = evtchn_from_irq(irq);
+
@@ -53560,7 +48365,7 @@
+ }
+}
+
-+static void end_dynirq(unsigned int irq)
++static void ack_dynirq_quirk_vector(unsigned int irq)
+{
+ int evtchn = evtchn_from_irq(irq);
+
@@ -53568,18 +48373,17 @@
+ unmask_evtchn(evtchn);
+}
+
-+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,
++static struct irq_chip dynirq_chip = {
++ .name = "Dynamic-irq",
++ .startup = startup_dynirq_vector,
++ .mask = mask_dynirq_vector,
++ .unmask = unmask_dynirq_vector,
++ .ack = ack_dynirq_vector,
++ .eoi = ack_dynirq_quirk_vector,
+#ifdef CONFIG_SMP
+ .set_affinity = set_affinity_irq,
+#endif
-+ .retrigger = retrigger,
++ .retrigger = retrigger_vector,
+};
+
+static inline void pirq_unmask_notify(int pirq)
@@ -53605,7 +48409,7 @@
+ */
+#define probing_irq(_irq) (irq_desc[(_irq)].action == NULL)
+
-+static unsigned int startup_pirq(unsigned int irq)
++static unsigned int startup_pirq_vector(unsigned int irq)
+{
+ struct evtchn_bind_pirq bind_pirq;
+ int evtchn = evtchn_from_irq(irq);
@@ -53637,26 +48441,7 @@
+ return 0;
+}
+
-+static void shutdown_pirq(unsigned int irq)
-+{
-+ struct evtchn_close close;
-+ int evtchn = evtchn_from_irq(irq);
-+
-+ if (!VALID_EVTCHN(evtchn))
-+ return;
-+
-+ mask_evtchn(evtchn);
-+
-+ 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] = IRQ_UNBOUND;
-+}
-+
-+static void enable_pirq(unsigned int irq)
++static void unmask_pirq_vector(unsigned int irq)
+{
+ int evtchn = evtchn_from_irq(irq);
+
@@ -53666,7 +48451,7 @@
+ }
+}
+
-+static void disable_pirq(unsigned int irq)
++static void mask_pirq_vector(unsigned int irq)
+{
+ int evtchn = evtchn_from_irq(irq);
+
@@ -53674,7 +48459,7 @@
+ mask_evtchn(evtchn);
+}
+
-+static void ack_pirq(unsigned int irq)
++static void ack_pirq_vector(unsigned int irq)
+{
+ int evtchn = evtchn_from_irq(irq);
+
@@ -53686,7 +48471,7 @@
+ }
+}
+
-+static void end_pirq(unsigned int irq)
++static void ack_pirq_quirk_vector(unsigned int irq)
+{
+ int evtchn = evtchn_from_irq(irq);
+
@@ -53696,18 +48481,17 @@
+ }
+}
+
-+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,
++static struct irq_chip pirq_chip = {
++ .name = "Phys-irq",
++ .startup = startup_pirq_vector,
++ .mask = mask_pirq_vector,
++ .unmask = unmask_pirq_vector,
++ .ack = ack_pirq_vector,
++ .eoi = ack_pirq_quirk_vector,
+#ifdef CONFIG_SMP
+ .set_affinity = set_affinity_irq,
+#endif
-+ .retrigger = retrigger,
++ .retrigger = retrigger_vector,
+};
+
+int irq_ignore_unhandled(unsigned int irq)
@@ -53875,7 +48659,8 @@
+ 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;
++ set_irq_chip_and_handler_name(dynirq_to_irq(i), &dynirq_chip,
++ handle_level_irq, "level");
+ }
+
+ /* Phys IRQ space is statically bound (1:1 mapping). Nail refcnts. */
@@ -53891,12 +48676,13 @@
+ 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;
++ set_irq_chip_and_handler_name(pirq_to_irq(i), &pirq_chip,
++ handle_level_irq, "level");
+ }
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/core/features.c linux-2.6.18-xen/drivers/xen/core/features.c
---- linux-2.6.18.3/drivers/xen/core/features.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/core/features.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/core/features.c b/drivers/xen/core/features.c
+--- a/drivers/xen/core/features.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/core/features.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,30 @@
+/******************************************************************************
+ * features.c
@@ -53928,9 +48714,9 @@
+ xen_features[i*32+j] = !!(fi.submap & 1<<j);
+ }
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/core/gnttab.c linux-2.6.18-xen/drivers/xen/core/gnttab.c
---- linux-2.6.18.3/drivers/xen/core/gnttab.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/core/gnttab.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/core/gnttab.c b/drivers/xen/core/gnttab.c
+--- a/drivers/xen/core/gnttab.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/core/gnttab.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,483 @@
+/******************************************************************************
+ * gnttab.c
@@ -54415,9 +49201,9 @@
+#ifdef CONFIG_XEN
+core_initcall(gnttab_init);
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/core/hypervisor_sysfs.c linux-2.6.18-xen/drivers/xen/core/hypervisor_sysfs.c
---- linux-2.6.18.3/drivers/xen/core/hypervisor_sysfs.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/core/hypervisor_sysfs.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/core/hypervisor_sysfs.c b/drivers/xen/core/hypervisor_sysfs.c
+--- a/drivers/xen/core/hypervisor_sysfs.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/core/hypervisor_sysfs.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,59 @@
+/*
+ * copyright (c) 2006 IBM Corporation
@@ -54478,9 +49264,9 @@
+
+device_initcall(hypervisor_subsys_init);
+EXPORT_SYMBOL_GPL(hypervisor_subsys);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/core/Makefile linux-2.6.18-xen/drivers/xen/core/Makefile
---- linux-2.6.18.3/drivers/xen/core/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/core/Makefile 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/core/Makefile b/drivers/xen/core/Makefile
+--- a/drivers/xen/core/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/core/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,13 @@
+#
+# Makefile for the linux kernel.
@@ -54495,10 +49281,10 @@
+obj-$(CONFIG_XEN_SKBUFF) += skbuff.o
+obj-$(CONFIG_XEN_REBOOT) += reboot.o
+obj-$(CONFIG_XEN_SMPBOOT) += smpboot.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/core/reboot.c linux-2.6.18-xen/drivers/xen/core/reboot.c
---- linux-2.6.18.3/drivers/xen/core/reboot.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/core/reboot.c 2006-11-19 14:26:56.000000000 +0100
-@@ -0,0 +1,384 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/core/reboot.c b/drivers/xen/core/reboot.c
+--- a/drivers/xen/core/reboot.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/core/reboot.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,385 @@
+#define __KERNEL_SYSCALLS__
+#include <linux/version.h>
+#include <linux/kernel.h>
@@ -54506,6 +49292,7 @@
+#include <linux/unistd.h>
+#include <linux/module.h>
+#include <linux/reboot.h>
++#include <linux/syscalls.h>
+#include <linux/sysrq.h>
+#include <linux/stringify.h>
+#include <asm/irq.h>
@@ -54579,8 +49366,8 @@
+
+/* Ignore multiple shutdown requests. */
+static int shutting_down = SHUTDOWN_INVALID;
-+static void __shutdown_handler(void *unused);
-+static DECLARE_WORK(shutdown_work, __shutdown_handler, NULL);
++static void __shutdown_handler(struct work_struct *work);
++static DECLARE_DELAYED_WORK(shutdown_work, __shutdown_handler);
+
+#if defined(__i386__) || defined(__x86_64__)
+
@@ -54723,7 +49510,7 @@
+
+ if ((shutting_down == SHUTDOWN_POWEROFF) ||
+ (shutting_down == SHUTDOWN_HALT)) {
-+ if (execve("/sbin/poweroff", poweroff_argv, envp) < 0) {
++ if (kernel_execve("/sbin/poweroff", poweroff_argv, envp) < 0) {
+ sys_reboot(LINUX_REBOOT_MAGIC1,
+ LINUX_REBOOT_MAGIC2,
+ LINUX_REBOOT_CMD_POWER_OFF,
@@ -54750,7 +49537,7 @@
+ return 0;
+}
+
-+static void __shutdown_handler(void *unused)
++static void __shutdown_handler(struct work_struct *unused)
+{
+ int err;
+
@@ -54810,7 +49597,7 @@
+ }
+
+ if (shutting_down != SHUTDOWN_INVALID)
-+ schedule_work(&shutdown_work);
++ schedule_delayed_work(&shutdown_work, 0);
+
+ kfree(str);
+}
@@ -54842,7 +49629,7 @@
+
+#ifdef CONFIG_MAGIC_SYSRQ
+ if (sysrq_key != '\0')
-+ handle_sysrq(sysrq_key, NULL, NULL);
++ handle_sysrq(sysrq_key, NULL);
+#endif
+}
+
@@ -54883,9 +49670,9 @@
+}
+
+subsys_initcall(setup_shutdown_event);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/core/skbuff.c linux-2.6.18-xen/drivers/xen/core/skbuff.c
---- linux-2.6.18.3/drivers/xen/core/skbuff.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/core/skbuff.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/core/skbuff.c b/drivers/xen/core/skbuff.c
+--- a/drivers/xen/core/skbuff.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/core/skbuff.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,144 @@
+
+#include <linux/module.h>
@@ -54903,7 +49690,7 @@
+#include <asm/hypervisor.h>
+
+/* Referenced in netback.c. */
-+/*static*/ kmem_cache_t *skbuff_cachep;
++/*static*/ struct kmem_cache *skbuff_cachep;
+EXPORT_SYMBOL(skbuff_cachep);
+
+/* Allow up to 64kB or page-sized packets (whichever is greater). */
@@ -54912,18 +49699,18 @@
+#else
+#define MAX_SKBUFF_ORDER 0
+#endif
-+static kmem_cache_t *skbuff_order_cachep[MAX_SKBUFF_ORDER + 1];
++static struct kmem_cache *skbuff_order_cachep[MAX_SKBUFF_ORDER + 1];
+
+static struct {
+ int size;
-+ kmem_cache_t *cachep;
++ struct kmem_cache *cachep;
+} skbuff_small[] = { { 512, NULL }, { 2048, NULL } };
+
+struct sk_buff *__alloc_skb(unsigned int length, gfp_t gfp_mask,
-+ int fclone)
++ int fclone, int node)
+{
+ int order, i;
-+ kmem_cache_t *cachep;
++ struct kmem_cache *cachep;
+
+ length = SKB_DATA_ALIGN(length) + sizeof(struct skb_shared_info);
+
@@ -54967,7 +49754,7 @@
+ return skb;
+}
+
-+static void skbuff_ctor(void *buf, kmem_cache_t *cachep, unsigned long unused)
++static void skbuff_ctor(void *buf, struct kmem_cache *cachep, unsigned long unused)
+{
+ int order = 0;
+
@@ -54982,7 +49769,7 @@
+ scrub_pages(buf, 1 << order);
+}
+
-+static void skbuff_dtor(void *buf, kmem_cache_t *cachep, unsigned long unused)
++static void skbuff_dtor(void *buf, struct kmem_cache *cachep, unsigned long unused)
+{
+ int order = 0;
+
@@ -55031,10 +49818,10 @@
+core_initcall(skbuff_init);
+
+EXPORT_SYMBOL(__dev_alloc_skb);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/core/smpboot.c linux-2.6.18-xen/drivers/xen/core/smpboot.c
---- linux-2.6.18.3/drivers/xen/core/smpboot.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/core/smpboot.c 2006-11-19 14:26:56.000000000 +0100
-@@ -0,0 +1,429 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/core/smpboot.c b/drivers/xen/core/smpboot.c
+--- a/drivers/xen/core/smpboot.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/core/smpboot.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,481 @@
+/*
+ * Xen SMP booting functions
+ *
@@ -55057,13 +49844,16 @@
+#include <asm/desc.h>
+#include <asm/arch_hooks.h>
+#include <asm/pgalloc.h>
++#if defined(__i386__)
++#include <asm/pda.h>
++#endif
+#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 *);
++extern irqreturn_t smp_reschedule_interrupt(int, void *);
++extern irqreturn_t smp_call_function_interrupt(int, void *);
+
+extern void local_setup_timer(unsigned int cpu);
+extern void local_teardown_timer(unsigned int cpu);
@@ -55112,8 +49902,6 @@
+#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
+
+void __init prefill_possible_map(void)
@@ -55182,9 +49970,24 @@
+}
+#endif
+
++#ifdef __i386__
++static inline void set_kernel_gs(void)
++{
++ /* Set %gs for this CPU's PDA. Memory clobber is to create a
++ barrier with respect to any PDA operations, so the compiler
++ doesn't move any before here. */
++ asm volatile ("mov %0, %%gs" : : "r" (__KERNEL_PDA) : "memory");
++}
++#endif
++
+void cpu_bringup(void)
+{
++#ifdef __i386__
++ set_kernel_gs();
++ secondary_cpu_init();
++#else
+ cpu_init();
++#endif
+ touch_softlockup_watchdog();
+ preempt_disable();
+ local_irq_enable();
@@ -55296,6 +50099,11 @@
+ }
+
+ for_each_possible_cpu (cpu) {
++#ifdef __i386__
++ struct i386_pda *pda;
++ struct desc_struct *gdt;
++#endif
++
+ if (cpu == 0)
+ continue;
+
@@ -55312,6 +50120,22 @@
+ }
+ gdt_descr->size = GDT_SIZE;
+ memcpy((void *)gdt_descr->address, cpu_gdt_table, GDT_SIZE);
++#ifdef __i386__
++ gdt = (struct desc_struct *)gdt_descr->address;
++ pda = kmalloc_node(sizeof(*pda), GFP_KERNEL, cpu_to_node(cpu));
++
++ if (unlikely(!pda)) {
++ printk(KERN_CRIT "CPU%d failed to allocate PDA\n",
++ cpu);
++ continue;
++ }
++ cpu_pda(cpu) = pda;
++ cpu_pda(cpu)->cpu_number = cpu;
++ pack_descriptor((u32 *)&gdt[GDT_ENTRY_PDA].a,
++ (u32 *)&gdt[GDT_ENTRY_PDA].b,
++ (unsigned long)pda, sizeof(*pda) - 1,
++ 0x80 | DESCTYPE_S | 0x2, 0); /* present read-write data segment */
++#endif
+ make_page_readonly(
+ (void *)gdt_descr->address,
+ XENFEAT_writable_descriptor_tables);
@@ -55326,8 +50150,8 @@
+ if (IS_ERR(idle))
+ panic("failed fork for CPU %d", cpu);
+
-+#ifdef __x86_64__
+ cpu_pda(cpu)->pcurrent = idle;
++#ifdef __x86_64__
+ cpu_pda(cpu)->cpunumber = cpu;
+ clear_ti_thread_flag(idle->thread_info, TIF_FORK);
+#endif
@@ -55346,17 +50170,17 @@
+
+ init_xenbus_allowed_cpumask();
+
-+#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:
+ */
++#ifdef CONFIG_X86_IO_APIC
+ if (!skip_ioapic_setup && nr_ioapics)
+ setup_IO_APIC();
+#endif
+}
+
-+void __devinit smp_prepare_boot_cpu(void)
++void __init smp_prepare_boot_cpu(void)
+{
+}
+
@@ -55430,7 +50254,7 @@
+
+#endif /* CONFIG_HOTPLUG_CPU */
+
-+int __devinit __cpu_up(unsigned int cpu)
++int __cpuinit __cpu_up(unsigned int cpu)
+{
+ int rc;
+
@@ -55458,15 +50282,30 @@
+{
+}
+
-+#ifndef CONFIG_X86_LOCAL_APIC
++#ifdef CONFIG_X86_MPPARSE
++/*
++ * If the BIOS enumerates physical processors before logical,
++ * maxcpus=N at enumeration-time can be used to disable HT.
++ */
++static int __init parse_maxcpus(char *arg)
++{
++ extern unsigned int maxcpus;
++
++ maxcpus = simple_strtoul(arg, NULL, 0);
++ return 0;
++}
++early_param("maxcpus", parse_maxcpus);
++#endif
++
++#if defined(CONFIG_XEN_UNPRIVILEGED_GUEST) && defined(CONFIG_X86_32)
+int setup_profiling_timer(unsigned int multiplier)
+{
+ return -EINVAL;
+}
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/core/xen_proc.c linux-2.6.18-xen/drivers/xen/core/xen_proc.c
---- linux-2.6.18.3/drivers/xen/core/xen_proc.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/core/xen_proc.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/core/xen_proc.c b/drivers/xen/core/xen_proc.c
+--- a/drivers/xen/core/xen_proc.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/core/xen_proc.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,18 @@
+
+#include <linux/proc_fs.h>
@@ -55486,9 +50325,9 @@
+{
+ remove_proc_entry(name, xen_base);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/core/xen_sysfs.c linux-2.6.18-xen/drivers/xen/core/xen_sysfs.c
---- linux-2.6.18.3/drivers/xen/core/xen_sysfs.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/core/xen_sysfs.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/core/xen_sysfs.c b/drivers/xen/core/xen_sysfs.c
+--- a/drivers/xen/core/xen_sysfs.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/core/xen_sysfs.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,378 @@
+/*
+ * copyright (c) 2006 IBM Corporation
@@ -55868,9 +50707,9 @@
+
+module_init(hyper_sysfs_init);
+module_exit(hyper_sysfs_exit);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/evtchn/evtchn.c linux-2.6.18-xen/drivers/xen/evtchn/evtchn.c
---- linux-2.6.18.3/drivers/xen/evtchn/evtchn.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/evtchn/evtchn.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/evtchn/evtchn.c b/drivers/xen/evtchn/evtchn.c
+--- a/drivers/xen/evtchn/evtchn.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/evtchn/evtchn.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,456 @@
+/******************************************************************************
+ * evtchn.c
@@ -56260,84 +51099,1085 @@
+
+ free_page((unsigned long)u->ring);
+
-+ for (i = 0; i < NR_EVENT_CHANNELS; i++) {
-+ int ret;
-+ if (port_user[i] != u)
-+ continue;
++ for (i = 0; i < NR_EVENT_CHANNELS; i++) {
++ int ret;
++ if (port_user[i] != u)
++ continue;
++
++ port_user[i] = NULL;
++ mask_evtchn(i);
++
++ close.port = i;
++ ret = HYPERVISOR_event_channel_op(EVTCHNOP_close, &close);
++ BUG_ON(ret);
++ }
++
++ spin_unlock_irq(&port_user_lock);
++
++ kfree(u);
++
++ return 0;
++}
++
++static struct file_operations evtchn_fops = {
++ .owner = THIS_MODULE,
++ .read = evtchn_read,
++ .write = evtchn_write,
++ .ioctl = evtchn_ioctl,
++ .poll = evtchn_poll,
++ .fasync = evtchn_fasync,
++ .open = evtchn_open,
++ .release = evtchn_release,
++};
++
++static struct miscdevice evtchn_miscdev = {
++ .minor = MISC_DYNAMIC_MINOR,
++ .name = "evtchn",
++ .fops = &evtchn_fops,
++};
++
++static int __init evtchn_init(void)
++{
++ int err;
++
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ spin_lock_init(&port_user_lock);
++ memset(port_user, 0, sizeof(port_user));
++
++ /* Create '/dev/misc/evtchn'. */
++ err = misc_register(&evtchn_miscdev);
++ if (err != 0) {
++ printk(KERN_ALERT "Could not register /dev/misc/evtchn\n");
++ return err;
++ }
++
++ printk("Event-channel device installed.\n");
++
++ return 0;
++}
++
++static void evtchn_cleanup(void)
++{
++ misc_deregister(&evtchn_miscdev);
++}
++
++module_init(evtchn_init);
++module_exit(evtchn_cleanup);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/evtchn/Makefile b/drivers/xen/evtchn/Makefile
+--- a/drivers/xen/evtchn/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/evtchn/Makefile 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,2 @@
++
++obj-y := evtchn.o
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/fbfront/Makefile b/drivers/xen/fbfront/Makefile
+--- a/drivers/xen/fbfront/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/fbfront/Makefile 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,2 @@
++obj-$(CONFIG_XEN_FRAMEBUFFER) := xenfb.o
++obj-$(CONFIG_XEN_KEYBOARD) += xenkbd.o
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/fbfront/xenfb.c b/drivers/xen/fbfront/xenfb.c
+--- a/drivers/xen/fbfront/xenfb.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/fbfront/xenfb.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,687 @@
++/*
++ * 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.
++ */
++
++/*
++ * TODO:
++ *
++ * Switch to grant tables when they become capable of dealing with the
++ * frame buffer.
++ */
++
++#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/freezer.h>
++#include <asm/hypervisor.h>
++#include <xen/evtchn.h>
++#include <xen/interface/io/fbif.h>
++#include <xen/xenbus.h>
++#include <linux/kthread.h>
++
++struct xenfb_mapping
++{
++ struct list_head link;
++ struct vm_area_struct *vma;
++ atomic_t map_refs;
++ int faults;
++ struct xenfb_info *info;
++};
++
++struct xenfb_info
++{
++ struct task_struct *kthread;
++ wait_queue_head_t wq;
++
++ 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 */
++
++ unsigned evtchn;
++ int irq;
++ struct xenfb_page *page;
++ unsigned long *mfns;
++ int update_wanted; /* XENFB_TYPE_UPDATE wanted */
++
++ struct xenbus_device *xbdev;
++};
++
++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 void xenfb_do_update(struct xenfb_info *info,
++ int x, int y, int w, int h)
++{
++ union xenfb_out_event event;
++ __u32 prod;
++
++ event.type = XENFB_TYPE_UPDATE;
++ event.update.x = x;
++ event.update.y = y;
++ event.update.width = w;
++ event.update.height = h;
++
++ 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;
++
++ notify_remote_via_evtchn(info->evtchn);
++}
++
++static int xenfb_queue_full(struct xenfb_info *info)
++{
++ __u32 cons, prod;
++
++ prod = info->page->out_prod;
++ cons = info->page->out_cons;
++ return prod - cons == XENFB_OUT_RING_LEN;
++}
++
++static void xenfb_update_screen(struct xenfb_info *info)
++{
++ unsigned long flags;
++ int y1, y2, x1, x2;
++ struct xenfb_mapping *map;
++
++ if (!info->update_wanted)
++ return;
++ if (xenfb_queue_full(info))
++ return;
++
++ mutex_lock(&info->mm_lock);
++
++ 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);
++
++ 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;
++ }
++
++ mutex_unlock(&info->mm_lock);
++
++ xenfb_do_update(info, x1, y1, x2 - x1, y2 - y1);
++}
++
++static int xenfb_thread(void *data)
++{
++ struct xenfb_info *info = data;
++
++ 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;
++}
++
++static int xenfb_setcolreg(unsigned regno, unsigned red, unsigned green,
++ unsigned blue, unsigned transp,
++ struct fb_info *info)
++{
++ u32 v;
++
++ if (regno > info->cmap.len)
++ return 1;
++
++ red >>= (16 - info->var.red.length);
++ green >>= (16 - info->var.green.length);
++ blue >>= (16 - info->var.blue.length);
++
++ v = (red << info->var.red.offset) |
++ (green << info->var.green.offset) |
++ (blue << info->var.blue.offset);
++
++ /* 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;
++}
++
++static void xenfb_timer(unsigned long data)
++{
++ struct xenfb_info *info = (struct xenfb_info *)data;
++ info->dirty = 1;
++ wake_up(&info->wq);
++}
++
++static void __xenfb_refresh(struct xenfb_info *info,
++ int x1, int y1, int w, int h)
++{
++ int y2, x2;
++
++ y2 = y1 + h;
++ x2 = x1 + w;
++
++ 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;
++
++ if (timer_pending(&info->refresh))
++ return;
++
++ mod_timer(&info->refresh, jiffies + HZ/xenfb_fps);
++}
++
++static void xenfb_refresh(struct xenfb_info *info,
++ int x1, int y1, int w, int h)
++{
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->dirty_lock, flags);
++ __xenfb_refresh(info, x1, y1, w, h);
++ spin_unlock_irqrestore(&info->dirty_lock, flags);
++}
++
++static void xenfb_fillrect(struct fb_info *p, const struct fb_fillrect *rect)
++{
++ struct xenfb_info *info = p->par;
++
++ cfb_fillrect(p, rect);
++ xenfb_refresh(info, rect->dx, rect->dy, rect->width, rect->height);
++}
++
++static void xenfb_imageblit(struct fb_info *p, const struct fb_image *image)
++{
++ struct xenfb_info *info = p->par;
++
++ cfb_imageblit(p, image);
++ xenfb_refresh(info, image->dx, image->dy, image->width, image->height);
++}
++
++static void xenfb_copyarea(struct fb_info *p, const struct fb_copyarea *area)
++{
++ struct xenfb_info *info = p->par;
++
++ cfb_copyarea(p, area);
++ xenfb_refresh(info, area->dx, area->dy, area->width, area->height);
++}
++
++static void xenfb_vm_open(struct vm_area_struct *vma)
++{
++ struct xenfb_mapping *map = vma->vm_private_data;
++ atomic_inc(&map->map_refs);
++}
++
++static void xenfb_vm_close(struct vm_area_struct *vma)
++{
++ struct xenfb_mapping *map = vma->vm_private_data;
++ struct xenfb_info *info = map->info;
++
++ mutex_lock(&info->mm_lock);
++ if (atomic_dec_and_test(&map->map_refs)) {
++ list_del(&map->link);
++ kfree(map);
++ }
++ mutex_unlock(&info->mm_lock);
++}
++
++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;
++
++ if (pgnr >= info->nr_pages)
++ return NOPAGE_SIGBUS;
++
++ mutex_lock(&info->mm_lock);
++ spin_lock_irqsave(&info->dirty_lock, flags);
++ page = info->pages[pgnr];
++ get_page(page);
++ map->faults++;
++
++ 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 (type)
++ *type = VM_FAULT_MINOR;
++
++ return page;
++}
++
++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)
++{
++ struct xenfb_info *info = fb_info->par;
++ struct xenfb_mapping *map;
++ int map_pages;
++
++ if (!(vma->vm_flags & VM_WRITE))
++ return -EINVAL;
++ if (!(vma->vm_flags & VM_SHARED))
++ return -EINVAL;
++ if (vma->vm_pgoff != 0)
++ return -EINVAL;
++
++ map_pages = (vma->vm_end - vma->vm_start + PAGE_SIZE-1) >> PAGE_SHIFT;
++ if (map_pages > info->nr_pages)
++ return -EINVAL;
++
++ map = kzalloc(sizeof(*map), GFP_KERNEL);
++ if (map == NULL)
++ return -ENOMEM;
++
++ map->vma = vma;
++ map->faults = 0;
++ map->info = info;
++ atomic_set(&map->map_refs, 1);
++
++ mutex_lock(&info->mm_lock);
++ list_add(&map->link, &info->mappings);
++ mutex_unlock(&info->mm_lock);
++
++ vma->vm_ops = &xenfb_vm_ops;
++ vma->vm_flags |= (VM_DONTEXPAND | VM_RESERVED);
++ vma->vm_private_data = map;
++
++ return 0;
++}
++
++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 irqreturn_t xenfb_event_handler(int rq, void *dev_id,
++ struct pt_regs *regs)
++{
++ /*
++ * 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;
++
++ if (page->in_cons != page->in_prod) {
++ info->page->in_cons = info->page->in_prod;
++ notify_remote_via_evtchn(info->evtchn);
++ }
++ return IRQ_HANDLED;
++}
++
++static unsigned long vmalloc_to_mfn(void *address)
++{
++ return pfn_to_mfn(vmalloc_to_pfn(address));
++}
++
++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 ret;
++
++ info = kzalloc(sizeof(*info), GFP_KERNEL);
++ if (info == NULL) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
++ return -ENOMEM;
++ }
++ 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;
++
++ /* set up shared page */
++ info->page = (void *)__get_free_page(GFP_KERNEL);
++ if (!info->page)
++ goto error_nomem;
++
++ xenfb_init_shared_page(info);
++
++ 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 = info->page->width;
++ fb_info->var.yres_virtual = fb_info->var.yres = info->page->height;
++ fb_info->var.bits_per_pixel = info->page->depth;
++
++ 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};
++
++ fb_info->var.activate = FB_ACTIVATE_NOW;
++ fb_info->var.height = -1;
++ fb_info->var.width = -1;
++ fb_info->var.vmode = FB_VMODE_NONINTERLACED;
++
++ 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;
++
++ 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;
++ }
++
++ 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;
++
++ /* 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;
++ }
++
++ 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;
++}
++
++static int xenfb_resume(struct xenbus_device *dev)
++{
++ struct xenfb_info *info = dev->dev.driver_data;
++
++ xenfb_disconnect_backend(info);
++ xenfb_init_shared_page(info);
++ return xenfb_connect_backend(dev, info);
++}
++
++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);
++ }
++ free_page((unsigned long)info->page);
++ vfree(info->mfns);
++ kfree(info->pages);
++ vfree(info->fb);
++ kfree(info);
++
++ return 0;
++}
++
++static void xenfb_init_shared_page(struct xenfb_info *info)
++{
++ int i;
++
++ for (i = 0; i < info->nr_pages; i++)
++ info->pages[i] = vmalloc_to_page(info->fb + i * PAGE_SIZE);
++
++ for (i = 0; i < info->nr_pages; i++)
++ info->mfns[i] = vmalloc_to_mfn(info->fb + i * PAGE_SIZE);
++
++ 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;
++}
++
++static int xenfb_connect_backend(struct xenbus_device *dev,
++ struct xenfb_info *info)
++{
++ int ret;
++ struct xenbus_transaction xbt;
++
++ ret = xenbus_alloc_evtchn(dev, &info->evtchn);
++ if (ret)
++ return ret;
++ ret = bind_evtchn_to_irqhandler(info->evtchn, xenfb_event_handler,
++ 0, "xenfb", info);
++ if (ret < 0) {
++ xenbus_free_evtchn(dev, info->evtchn);
++ xenbus_dev_fatal(dev, ret, "bind_evtchn_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",
++ info->evtchn);
++ 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;
++ }
++
++ xenbus_switch_state(dev, XenbusStateInitialised);
++ return 0;
++
++ error_xenbus:
++ xenbus_transaction_end(xbt, 1);
++ xenbus_dev_fatal(dev, ret, "writing xenstore");
++ return ret;
++}
++
++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)
++{
++ 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;
++
++ 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;
++
++ case XenbusStateClosing:
++ // FIXME is this safe in any dev->state?
++ xenbus_frontend_closed(dev);
++ break;
++ }
++}
++
++static struct xenbus_device_id xenfb_ids[] = {
++ { "vfb" },
++ { "" }
++};
++
++static struct xenbus_driver xenfb = {
++ .name = "vfb",
++ .owner = THIS_MODULE,
++ .ids = xenfb_ids,
++ .probe = xenfb_probe,
++ .remove = xenfb_remove,
++ .resume = xenfb_resume,
++ .otherend_changed = xenfb_backend_changed,
++};
++
++static int __init xenfb_init(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ /* Nothing to do if running in dom0. */
++ if (is_initial_xendomain())
++ return -ENODEV;
++
++ return xenbus_register_frontend(&xenfb);
++}
++
++static void __exit xenfb_cleanup(void)
++{
++ return xenbus_unregister_driver(&xenfb);
++}
++
++module_init(xenfb_init);
++module_exit(xenfb_cleanup);
++
++MODULE_LICENSE("GPL");
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/fbfront/xenkbd.c b/drivers/xen/fbfront/xenkbd.c
+--- a/drivers/xen/fbfront/xenkbd.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/fbfront/xenkbd.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,300 @@
++/*
++ * 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.
++ */
++
++/*
++ * TODO:
++ *
++ * Switch to grant tables together with xenfb.c.
++ */
++
++#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>
++
++struct xenkbd_info
++{
++ struct input_dev *dev;
++ struct xenkbd_page *page;
++ unsigned evtchn;
++ int irq;
++ struct xenbus_device *xbdev;
++};
++
++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.
++ */
++
++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->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;
++ event = &XENKBD_IN_RING_REF(page, cons);
++
++ switch (event->type) {
++ case XENKBD_TYPE_MOTION:
++ input_report_rel(info->dev, REL_X, event->motion.rel_x);
++ input_report_rel(info->dev, REL_Y, event->motion.rel_y);
++ break;
++ case XENKBD_TYPE_KEY:
++ input_report_key(info->dev, event->key.keycode, event->key.pressed);
++ break;
++ case XENKBD_TYPE_POS:
++ input_report_abs(info->dev, ABS_X, event->pos.abs_x);
++ input_report_abs(info->dev, ABS_Y, event->pos.abs_y);
++ break;
++ }
++ }
++ input_sync(info->dev);
++ mb(); /* ensure we got ring contents */
++ page->in_cons = cons;
++ notify_remote_via_evtchn(info->evtchn);
++
++ return IRQ_HANDLED;
++}
++
++int __devinit xenkbd_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ int ret, i;
++ struct xenkbd_info *info;
++ struct input_dev *input_dev;
++
++ 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;
++
++ 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;
++
++ input_dev = input_allocate_device();
++ if (!input_dev)
++ goto error_nomem;
++
++ input_dev->evbit[0] = BIT(EV_KEY) | BIT(EV_REL) | BIT(EV_ABS);
++ input_dev->keybit[LONG(BTN_MOUSE)]
++ = BIT(BTN_LEFT) | BIT(BTN_MIDDLE) | BIT(BTN_RIGHT);
++ /* TODO additional buttons */
++ input_dev->relbit[0] = BIT(REL_X) | BIT(REL_Y);
++
++ /* FIXME not sure this is quite right */
++ for (i = 0; i < 256; i++)
++ set_bit(i, input_dev->keybit);
++
++ input_dev->name = "Xen Virtual Keyboard/Mouse";
++
++ input_set_abs_params(input_dev, ABS_X, 0, XENFB_WIDTH, 0, 0);
++ input_set_abs_params(input_dev, ABS_Y, 0, XENFB_HEIGHT, 0, 0);
++
++ ret = input_register_device(input_dev);
++ if (ret) {
++ input_free_device(input_dev);
++ xenbus_dev_fatal(dev, ret, "input_register_device");
++ goto error;
++ }
++ info->dev = input_dev;
++
++ ret = xenkbd_connect_backend(dev, info);
++ if (ret < 0)
++ goto error;
++
++ return 0;
++
++ error_nomem:
++ ret = -ENOMEM;
++ xenbus_dev_fatal(dev, ret, "allocating device memory");
++ error:
++ xenkbd_remove(dev);
++ return ret;
++}
++
++static int xenkbd_resume(struct xenbus_device *dev)
++{
++ struct xenkbd_info *info = dev->dev.driver_data;
++
++ xenkbd_disconnect_backend(info);
++ return xenkbd_connect_backend(dev, info);
++}
++
++static int xenkbd_remove(struct xenbus_device *dev)
++{
++ struct xenkbd_info *info = dev->dev.driver_data;
++
++ xenkbd_disconnect_backend(info);
++ input_unregister_device(info->dev);
++ 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;
++
++ ret = xenbus_alloc_evtchn(dev, &info->evtchn);
++ if (ret)
++ return ret;
++ ret = bind_evtchn_to_irqhandler(info->evtchn, input_handler, 0,
++ "xenkbd", info);
++ if (ret < 0) {
++ xenbus_free_evtchn(dev, info->evtchn);
++ xenbus_dev_fatal(dev, ret, "bind_evtchn_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",
++ info->evtchn);
++ 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;
++ }
++
++ xenbus_switch_state(dev, XenbusStateInitialised);
++ return 0;
++
++ 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;
++}
+
-+ port_user[i] = NULL;
-+ mask_evtchn(i);
++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;
+
-+ close.port = i;
-+ ret = HYPERVISOR_event_channel_op(EVTCHNOP_close, &close);
-+ BUG_ON(ret);
-+ }
++ switch (backend_state) {
++ case XenbusStateInitialising:
++ case XenbusStateInitialised:
++ case XenbusStateUnknown:
++ case XenbusStateClosed:
++ break;
+
-+ spin_unlock_irq(&port_user_lock);
++ 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;
+
-+ kfree(u);
++ 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;
+
-+ return 0;
++ case XenbusStateClosing:
++ xenbus_frontend_closed(dev);
++ break;
++ }
+}
+
-+static struct file_operations evtchn_fops = {
-+ .owner = THIS_MODULE,
-+ .read = evtchn_read,
-+ .write = evtchn_write,
-+ .ioctl = evtchn_ioctl,
-+ .poll = evtchn_poll,
-+ .fasync = evtchn_fasync,
-+ .open = evtchn_open,
-+ .release = evtchn_release,
++static struct xenbus_device_id xenkbd_ids[] = {
++ { "vkbd" },
++ { "" }
+};
+
-+static struct miscdevice evtchn_miscdev = {
-+ .minor = MISC_DYNAMIC_MINOR,
-+ .name = "evtchn",
-+ .fops = &evtchn_fops,
++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,
+};
+
-+static int __init evtchn_init(void)
++static int __init xenkbd_init(void)
+{
-+ int err;
-+
+ if (!is_running_on_xen())
+ return -ENODEV;
+
-+ spin_lock_init(&port_user_lock);
-+ memset(port_user, 0, sizeof(port_user));
-+
-+ /* Create '/dev/misc/evtchn'. */
-+ err = misc_register(&evtchn_miscdev);
-+ if (err != 0) {
-+ printk(KERN_ALERT "Could not register /dev/misc/evtchn\n");
-+ return err;
-+ }
-+
-+ printk("Event-channel device installed.\n");
++ /* Nothing to do if running in dom0. */
++ if (is_initial_xendomain())
++ return -ENODEV;
+
-+ return 0;
++ return xenbus_register_frontend(&xenkbd);
+}
+
-+static void evtchn_cleanup(void)
++static void __exit xenkbd_cleanup(void)
+{
-+ misc_deregister(&evtchn_miscdev);
++ return xenbus_unregister_driver(&xenkbd);
+}
+
-+module_init(evtchn_init);
-+module_exit(evtchn_cleanup);
-+
-+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/evtchn/Makefile linux-2.6.18-xen/drivers/xen/evtchn/Makefile
---- linux-2.6.18.3/drivers/xen/evtchn/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/evtchn/Makefile 2006-11-19 14:26:56.000000000 +0100
-@@ -0,0 +1,2 @@
++module_init(xenkbd_init);
++module_exit(xenkbd_cleanup);
+
-+obj-y := evtchn.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/Kconfig linux-2.6.18-xen/drivers/xen/Kconfig
---- linux-2.6.18.3/drivers/xen/Kconfig 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/Kconfig 2006-11-19 14:26:55.000000000 +0100
-@@ -0,0 +1,260 @@
++MODULE_LICENSE("GPL");
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/Kconfig b/drivers/xen/Kconfig
+--- a/drivers/xen/Kconfig 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/Kconfig 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,283 @@
+#
+# This Kconfig describe xen options
+#
@@ -56512,6 +52352,29 @@
+ dedicated device-driver domain, or your master control domain
+ (domain 0), then you almost certainly want to say Y here.
+
++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.
++
++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.
++
+config XEN_SCRUB_PAGES
+ bool "Scrub memory before freeing it to Xen"
+ default y
@@ -56598,10 +52461,10 @@
+ depends on SMP
+
+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/Makefile linux-2.6.18-xen/drivers/xen/Makefile
---- linux-2.6.18.3/drivers/xen/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/Makefile 2006-11-19 14:26:55.000000000 +0100
-@@ -0,0 +1,17 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/Makefile b/drivers/xen/Makefile
+--- a/drivers/xen/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/Makefile 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,19 @@
+obj-y += core/
+obj-y += console/
+obj-y += evtchn/
@@ -56619,9 +52482,11 @@
+obj-$(CONFIG_XEN_NETDEV_FRONTEND) += netfront/
+obj-$(CONFIG_XEN_PCIDEV_BACKEND) += pciback/
+obj-$(CONFIG_XEN_PCIDEV_FRONTEND) += pcifront/
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/netback/common.h linux-2.6.18-xen/drivers/xen/netback/common.h
---- linux-2.6.18.3/drivers/xen/netback/common.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/netback/common.h 2006-12-05 18:42:37.000000000 +0100
++obj-$(CONFIG_XEN_FRAMEBUFFER) += fbfront/
++obj-$(CONFIG_XEN_KEYBOARD) += fbfront/
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/netback/common.h b/drivers/xen/netback/common.h
+--- a/drivers/xen/netback/common.h 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/netback/common.h 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,141 @@
+/******************************************************************************
+ * arch/xen/drivers/netif/backend/common.h
@@ -56749,7 +52614,7 @@
+
+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);
++irqreturn_t netif_be_int(int irq, void *dev_id);
+
+static inline int netbk_can_queue(struct net_device *dev)
+{
@@ -56764,9 +52629,9 @@
+}
+
+#endif /* __NETIF__BACKEND__COMMON_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/netback/interface.c linux-2.6.18-xen/drivers/xen/netback/interface.c
---- linux-2.6.18.3/drivers/xen/netback/interface.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/netback/interface.c 2006-12-05 18:42:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/netback/interface.c b/drivers/xen/netback/interface.c
+--- a/drivers/xen/netback/interface.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/netback/interface.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,352 @@
+/******************************************************************************
+ * arch/xen/drivers/netif/backend/interface.c
@@ -57120,10 +52985,10 @@
+ }
+ netif_free(netif);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/netback/loopback.c linux-2.6.18-xen/drivers/xen/netback/loopback.c
---- linux-2.6.18.3/drivers/xen/netback/loopback.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/netback/loopback.c 2006-11-19 14:26:56.000000000 +0100
-@@ -0,0 +1,320 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/netback/loopback.c b/drivers/xen/netback/loopback.c
+--- a/drivers/xen/netback/loopback.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/netback/loopback.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,322 @@
+/******************************************************************************
+ * netback/loopback.c
+ *
@@ -57181,6 +53046,8 @@
+#include <net/xfrm.h> /* secpath_reset() */
+#include <asm/hypervisor.h> /* is_initial_xendomain() */
+
++#include "../../../net/core/kmap_skb.h"
++
+static int nloopbacks = -1;
+module_param(nloopbacks, int, 0);
+MODULE_PARM_DESC(nloopbacks, "Number of netback-loopback devices to create");
@@ -57273,7 +53140,7 @@
+ np->stats.rx_bytes += skb->len;
+ np->stats.rx_packets++;
+
-+ if (skb->ip_summed == CHECKSUM_HW) {
++ if (skb->ip_summed == CHECKSUM_PARTIAL) {
+ /* Defer checksum calculation. */
+ skb->proto_csum_blank = 1;
+ /* Must be a local packet: assert its integrity. */
@@ -57444,18 +53311,18 @@
+module_exit(loopback_exit);
+
+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/netback/Makefile linux-2.6.18-xen/drivers/xen/netback/Makefile
---- linux-2.6.18.3/drivers/xen/netback/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/netback/Makefile 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/netback/Makefile b/drivers/xen/netback/Makefile
+--- a/drivers/xen/netback/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/netback/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,5 @@
+obj-$(CONFIG_XEN_NETDEV_BACKEND) := netbk.o
+obj-$(CONFIG_XEN_NETDEV_LOOPBACK) += netloop.o
+
+netbk-y := netback.o xenbus.o interface.o
+netloop-y := loopback.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/netback/netback.c linux-2.6.18-xen/drivers/xen/netback/netback.c
---- linux-2.6.18.3/drivers/xen/netback/netback.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/netback/netback.c 2006-12-05 18:42:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/netback/netback.c b/drivers/xen/netback/netback.c
+--- a/drivers/xen/netback/netback.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/netback/netback.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,1510 @@
+/******************************************************************************
+ * drivers/xen/netback/netback.c
@@ -57598,12 +53465,12 @@
+
+/*
+ * A gross way of confirming the origin of an skb data page. The slab
-+ * allocator abuses a field in the page struct to cache the kmem_cache_t ptr.
++ * allocator abuses a field in the page struct to cache the struct kmem_cache ptr.
+ */
+static inline int is_xen_skb(struct sk_buff *skb)
+{
-+ extern kmem_cache_t *skbuff_cachep;
-+ kmem_cache_t *cp = (kmem_cache_t *)virt_to_page(skb->head)->lru.next;
++ extern struct kmem_cache *skbuff_cachep;
++ struct kmem_cache *cp = (struct kmem_cache *)virt_to_page(skb->head)->lru.next;
+ return (cp == skbuff_cachep);
+}
+
@@ -57943,7 +53810,7 @@
+ copy_op = npo->copy + npo->copy_cons++;
+ if (copy_op->status != GNTST_okay) {
+ DPRINTK("Bad status %d from copy to DOM%d.\n",
-+ gop->status, domid);
++ copy_op->status, domid);
+ status = NETIF_RSP_ERROR;
+ }
+ } else {
@@ -58125,7 +53992,7 @@
+ id = meta[npo.meta_cons].id;
+ flags = nr_frags ? NETRXF_more_data : 0;
+
-+ if (skb->ip_summed == CHECKSUM_HW) /* local packet? */
++ if (skb->ip_summed == CHECKSUM_PARTIAL) /* local packet? */
+ flags |= NETRXF_csum_blank | NETRXF_data_validated;
+ else if (skb->proto_data_valid) /* remote but checksummed? */
+ flags |= NETRXF_data_validated;
@@ -58811,7 +54678,7 @@
+ netif_idx_release(page->index);
+}
+
-+irqreturn_t netif_be_int(int irq, void *dev_id, struct pt_regs *regs)
++irqreturn_t netif_be_int(int irq, void *dev_id)
+{
+ netif_t *netif = dev_id;
+
@@ -58878,7 +54745,7 @@
+}
+
+#ifdef NETBE_DEBUG_INTERRUPT
-+static irqreturn_t netif_be_dbg(int irq, void *dev_id, struct pt_regs *regs)
++static irqreturn_t netif_be_dbg(int irq, void *dev_id)
+{
+ struct list_head *ent;
+ netif_t *netif;
@@ -58967,9 +54834,9 @@
+module_init(netback_init);
+
+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/netback/xenbus.c linux-2.6.18-xen/drivers/xen/netback/xenbus.c
---- linux-2.6.18.3/drivers/xen/netback/xenbus.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/netback/xenbus.c 2006-12-05 18:42:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/netback/xenbus.c b/drivers/xen/netback/xenbus.c
+--- a/drivers/xen/netback/xenbus.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/netback/xenbus.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,438 @@
+/* Xenbus code for netif backend
+ Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
@@ -59409,9 +55276,9 @@
+{
+ xenbus_register_backend(&netback);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/netfront/Kconfig linux-2.6.18-xen/drivers/xen/netfront/Kconfig
---- linux-2.6.18.3/drivers/xen/netfront/Kconfig 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/netfront/Kconfig 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/netfront/Kconfig b/drivers/xen/netfront/Kconfig
+--- a/drivers/xen/netfront/Kconfig 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/netfront/Kconfig 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,6 @@
+
+config XENNET
@@ -59419,17 +55286,17 @@
+ depends on NETDEVICES && ARCH_XEN
+ help
+ Network driver for Xen
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/netfront/Makefile linux-2.6.18-xen/drivers/xen/netfront/Makefile
---- linux-2.6.18.3/drivers/xen/netfront/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/netfront/Makefile 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/netfront/Makefile b/drivers/xen/netfront/Makefile
+--- a/drivers/xen/netfront/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/netfront/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,4 @@
+
+obj-$(CONFIG_XEN_NETDEV_FRONTEND) := xennet.o
+
+xennet-objs := netfront.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/netfront/netfront.c linux-2.6.18-xen/drivers/xen/netfront/netfront.c
---- linux-2.6.18.3/drivers/xen/netfront/netfront.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/netfront/netfront.c 2006-12-05 18:42:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/netfront/netfront.c b/drivers/xen/netfront/netfront.c
+--- a/drivers/xen/netfront/netfront.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/netfront/netfront.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,2138 @@
+/******************************************************************************
+ * Virtual network driver for conversing with remote driver backends.
@@ -59550,7 +55417,7 @@
+{
+ return skb_is_gso(skb) &&
+ (!skb_gso_ok(skb, dev->features) ||
-+ unlikely(skb->ip_summed != CHECKSUM_HW));
++ unlikely(skb->ip_summed != CHECKSUM_PARTIAL));
+}
+#else
+#define netif_needs_gso(dev, skb) 0
@@ -59678,7 +55545,7 @@
+static void network_alloc_rx_buffers(struct net_device *);
+static int send_fake_arp(struct net_device *);
+
-+static irqreturn_t netif_int(int irq, void *dev_id, struct pt_regs *ptregs);
++static irqreturn_t netif_int(int irq, void *dev_id);
+
+#ifdef CONFIG_SYSFS
+static int xennet_sysfs_addif(struct net_device *netdev);
@@ -60358,7 +56225,7 @@
+ tx->flags = 0;
+ extra = NULL;
+
-+ if (skb->ip_summed == CHECKSUM_HW) /* local packet? */
++ if (skb->ip_summed == CHECKSUM_PARTIAL) /* local packet? */
+ tx->flags |= NETTXF_csum_blank | NETTXF_data_validated;
+#ifdef CONFIG_XEN
+ if (skb->proto_data_valid) /* remote but checksummed? */
@@ -60413,7 +56280,7 @@
+ return 0;
+}
+
-+static irqreturn_t netif_int(int irq, void *dev_id, struct pt_regs *ptregs)
++static irqreturn_t netif_int(int irq, void *dev_id)
+{
+ struct net_device *dev = dev_id;
+ struct netfront_info *np = netdev_priv(dev);
@@ -61569,9 +57436,9 @@
+module_exit(netif_exit);
+
+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/conf_space.c linux-2.6.18-xen/drivers/xen/pciback/conf_space.c
---- linux-2.6.18.3/drivers/xen/pciback/conf_space.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/conf_space.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/conf_space.c b/drivers/xen/pciback/conf_space.c
+--- a/drivers/xen/pciback/conf_space.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/conf_space.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,425 @@
+/*
+ * PCI Backend - Functions for creating a virtual configuration space for
@@ -61998,9 +57865,9 @@
+{
+ return pciback_config_capability_init();
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/conf_space_capability.c linux-2.6.18-xen/drivers/xen/pciback/conf_space_capability.c
---- linux-2.6.18.3/drivers/xen/pciback/conf_space_capability.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/conf_space_capability.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/conf_space_capability.c b/drivers/xen/pciback/conf_space_capability.c
+--- a/drivers/xen/pciback/conf_space_capability.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/conf_space_capability.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,71 @@
+/*
+ * PCI Backend - Handles the virtual fields found on the capability lists
@@ -62073,9 +57940,9 @@
+
+ return 0;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/conf_space_capability.h linux-2.6.18-xen/drivers/xen/pciback/conf_space_capability.h
---- linux-2.6.18.3/drivers/xen/pciback/conf_space_capability.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/conf_space_capability.h 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/conf_space_capability.h b/drivers/xen/pciback/conf_space_capability.h
+--- a/drivers/xen/pciback/conf_space_capability.h 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/conf_space_capability.h 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,23 @@
+/*
+ * PCI Backend - Data structures for special overlays for structures on
@@ -62100,9 +57967,9 @@
+};
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/conf_space_capability_pm.c linux-2.6.18-xen/drivers/xen/pciback/conf_space_capability_pm.c
---- linux-2.6.18.3/drivers/xen/pciback/conf_space_capability_pm.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/conf_space_capability_pm.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/conf_space_capability_pm.c b/drivers/xen/pciback/conf_space_capability_pm.c
+--- a/drivers/xen/pciback/conf_space_capability_pm.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/conf_space_capability_pm.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,113 @@
+/*
+ * PCI Backend - Configuration space overlay for power management
@@ -62217,9 +58084,9 @@
+ .capability = PCI_CAP_ID_PM,
+ .fields = caplist_pm,
+};
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/conf_space_capability_vpd.c linux-2.6.18-xen/drivers/xen/pciback/conf_space_capability_vpd.c
---- linux-2.6.18.3/drivers/xen/pciback/conf_space_capability_vpd.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/conf_space_capability_vpd.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/conf_space_capability_vpd.c b/drivers/xen/pciback/conf_space_capability_vpd.c
+--- a/drivers/xen/pciback/conf_space_capability_vpd.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/conf_space_capability_vpd.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,42 @@
+/*
+ * PCI Backend - Configuration space overlay for Vital Product Data
@@ -62263,9 +58130,9 @@
+ .capability = PCI_CAP_ID_VPD,
+ .fields = caplist_vpd,
+};
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/conf_space.h linux-2.6.18-xen/drivers/xen/pciback/conf_space.h
---- linux-2.6.18.3/drivers/xen/pciback/conf_space.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/conf_space.h 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/conf_space.h b/drivers/xen/pciback/conf_space.h
+--- a/drivers/xen/pciback/conf_space.h 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/conf_space.h 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,126 @@
+/*
+ * PCI Backend - Common data structures for overriding the configuration space
@@ -62393,10 +58260,10 @@
+int pciback_config_capability_add_fields(struct pci_dev *dev);
+
+#endif /* __XEN_PCIBACK_CONF_SPACE_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/conf_space_header.c linux-2.6.18-xen/drivers/xen/pciback/conf_space_header.c
---- linux-2.6.18.3/drivers/xen/pciback/conf_space_header.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/conf_space_header.c 2006-11-19 14:26:56.000000000 +0100
-@@ -0,0 +1,299 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/conf_space_header.c b/drivers/xen/pciback/conf_space_header.c
+--- a/drivers/xen/pciback/conf_space_header.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/conf_space_header.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,301 @@
+/*
+ * PCI Backend - Handles the virtual fields in the configuration space headers.
+ *
@@ -62419,12 +58286,13 @@
+
+static int command_write(struct pci_dev *dev, int offset, u16 value, void *data)
+{
-+ if (!dev->is_enabled && is_enable_cmd(value)) {
++ if (!atomic_read(&dev->enable_cnt) && is_enable_cmd(value)) {
+ if (unlikely(verbose_request))
+ printk(KERN_DEBUG "pciback: %s: enable\n",
+ pci_name(dev));
-+ pci_enable_device(dev);
-+ } else if (dev->is_enabled && !is_enable_cmd(value)) {
++ if (pci_enable_device(dev))
++ return -ENODEV;
++ } else if (atomic_read(&dev->enable_cnt) && !is_enable_cmd(value)) {
+ if (unlikely(verbose_request))
+ printk(KERN_DEBUG "pciback: %s: disable\n",
+ pci_name(dev));
@@ -62443,7 +58311,8 @@
+ printk(KERN_DEBUG
+ "pciback: %s: enable memory-write-invalidate\n",
+ pci_name(dev));
-+ pci_set_mwi(dev);
++ if (pci_set_mwi(dev))
++ return -EINVAL;
+ }
+
+ return pci_write_config_word(dev, offset, value);
@@ -62696,9 +58565,9 @@
+ out:
+ return err;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/conf_space_quirks.c linux-2.6.18-xen/drivers/xen/pciback/conf_space_quirks.c
---- linux-2.6.18.3/drivers/xen/pciback/conf_space_quirks.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/conf_space_quirks.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/conf_space_quirks.c b/drivers/xen/pciback/conf_space_quirks.c
+--- a/drivers/xen/pciback/conf_space_quirks.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/conf_space_quirks.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,128 @@
+/*
+ * PCI Backend - Handle special overlays for broken devices.
@@ -62828,9 +58697,9 @@
+ out:
+ return ret;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/conf_space_quirks.h linux-2.6.18-xen/drivers/xen/pciback/conf_space_quirks.h
---- linux-2.6.18.3/drivers/xen/pciback/conf_space_quirks.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/conf_space_quirks.h 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/conf_space_quirks.h b/drivers/xen/pciback/conf_space_quirks.h
+--- a/drivers/xen/pciback/conf_space_quirks.h 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/conf_space_quirks.h 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,35 @@
+/*
+ * PCI Backend - Data structures for special overlays for broken devices.
@@ -62867,9 +58736,9 @@
+int pciback_field_is_dup(struct pci_dev *dev, int reg);
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/Makefile linux-2.6.18-xen/drivers/xen/pciback/Makefile
---- linux-2.6.18.3/drivers/xen/pciback/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/Makefile 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/Makefile b/drivers/xen/pciback/Makefile
+--- a/drivers/xen/pciback/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,15 @@
+obj-$(CONFIG_XEN_PCIDEV_BACKEND) += pciback.o
+
@@ -62886,9 +58755,9 @@
+ifeq ($(CONFIG_XEN_PCIDEV_BE_DEBUG),y)
+EXTRA_CFLAGS += -DDEBUG
+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/passthrough.c linux-2.6.18-xen/drivers/xen/pciback/passthrough.c
---- linux-2.6.18.3/drivers/xen/pciback/passthrough.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/passthrough.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/passthrough.c b/drivers/xen/pciback/passthrough.c
+--- a/drivers/xen/pciback/passthrough.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/passthrough.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,157 @@
+/*
+ * PCI Backend - Provides restricted access to the real PCI bus topology
@@ -63047,9 +58916,9 @@
+ kfree(dev_data);
+ pdev->pci_dev_data = NULL;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/pciback.h linux-2.6.18-xen/drivers/xen/pciback/pciback.h
---- linux-2.6.18.3/drivers/xen/pciback/pciback.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/pciback.h 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/pciback.h b/drivers/xen/pciback/pciback.h
+--- a/drivers/xen/pciback/pciback.h 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/pciback.h 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,93 @@
+/*
+ * PCI Backend Common Data Structures & Function Declarations
@@ -63092,7 +58961,7 @@
+
+ unsigned long flags;
+
-+ struct work_struct op_work;
++ struct delayed_work op_work;
+};
+
+struct pciback_dev_data {
@@ -63136,18 +59005,18 @@
+void pciback_release_devices(struct pciback_device *pdev);
+
+/* 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);
++irqreturn_t pciback_handle_event(int irq, void *dev_id);
++void pciback_do_op(struct work_struct *work);
+
+int pciback_xenbus_register(void);
+void pciback_xenbus_unregister(void);
+
+extern int verbose_request;
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/pciback_ops.c linux-2.6.18-xen/drivers/xen/pciback/pciback_ops.c
---- linux-2.6.18.3/drivers/xen/pciback/pciback_ops.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/pciback_ops.c 2006-11-19 14:26:56.000000000 +0100
-@@ -0,0 +1,95 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/pciback_ops.c b/drivers/xen/pciback/pciback_ops.c
+--- a/drivers/xen/pciback/pciback_ops.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/pciback_ops.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,94 @@
+/*
+ * PCI Backend Operations - respond to PCI requests from Frontend
+ *
@@ -63175,7 +59044,6 @@
+
+ pci_write_config_word(dev, PCI_COMMAND, 0);
+
-+ dev->is_enabled = 0;
+ dev->is_busmaster = 0;
+ } else {
+ pci_read_config_word(dev, PCI_COMMAND, &cmd);
@@ -63194,16 +59062,16 @@
+ * 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);
++ schedule_delayed_work(&pdev->op_work, 0);
+}
+
+/* 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 pciback_do_op(struct work_struct *work)
+{
-+ struct pciback_device *pdev = data;
++ struct pciback_device *pdev = container_of(work, struct pciback_device, op_work.work);
+ struct pci_dev *dev;
+ struct xen_pci_op *op = &pdev->sh_info->op;
+
@@ -63235,7 +59103,7 @@
+ test_and_schedule_op(pdev);
+}
+
-+irqreturn_t pciback_handle_event(int irq, void *dev_id, struct pt_regs *regs)
++irqreturn_t pciback_handle_event(int irq, void *dev_id)
+{
+ struct pciback_device *pdev = dev_id;
+
@@ -63243,10 +59111,10 @@
+
+ return IRQ_HANDLED;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/pci_stub.c linux-2.6.18-xen/drivers/xen/pciback/pci_stub.c
---- linux-2.6.18.3/drivers/xen/pciback/pci_stub.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/pci_stub.c 2006-11-19 14:26:56.000000000 +0100
-@@ -0,0 +1,916 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/pci_stub.c b/drivers/xen/pciback/pci_stub.c
+--- a/drivers/xen/pciback/pci_stub.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/pci_stub.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,923 @@
+/*
+ * PCI Stub Driver - Grabs devices in backend to be exported later
+ *
@@ -64097,13 +59965,20 @@
+ err = pci_register_driver(&pciback_pci_driver);
+ if (err < 0)
+ goto out;
-+
-+ driver_create_file(&pciback_pci_driver.driver, &driver_attr_new_slot);
-+ driver_create_file(&pciback_pci_driver.driver,
-+ &driver_attr_remove_slot);
-+ driver_create_file(&pciback_pci_driver.driver, &driver_attr_slots);
-+ driver_create_file(&pciback_pci_driver.driver, &driver_attr_quirks);
-+ driver_create_file(&pciback_pci_driver.driver, &driver_attr_permissive);
++ err = driver_create_file(&pciback_pci_driver.driver, &driver_attr_new_slot);
++ if (err < 0)
++ goto out;
++ err = driver_create_file(&pciback_pci_driver.driver,
++ &driver_attr_remove_slot);
++ if (err < 0)
++ goto out;
++ err = driver_create_file(&pciback_pci_driver.driver, &driver_attr_slots);
++ if (err < 0)
++ goto out;
++ err = driver_create_file(&pciback_pci_driver.driver, &driver_attr_quirks);
++ if (err < 0)
++ goto out;
++ err = driver_create_file(&pciback_pci_driver.driver, &driver_attr_permissive);
+
+ out:
+ return err;
@@ -64163,9 +60038,9 @@
+module_exit(pciback_cleanup);
+
+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/slot.c linux-2.6.18-xen/drivers/xen/pciback/slot.c
---- linux-2.6.18.3/drivers/xen/pciback/slot.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/slot.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/slot.c b/drivers/xen/pciback/slot.c
+--- a/drivers/xen/pciback/slot.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/slot.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,151 @@
+/*
+ * PCI Backend - Provides a Virtual PCI bus (with real devices)
@@ -64318,9 +60193,9 @@
+ kfree(slot_dev);
+ pdev->pci_dev_data = NULL;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/vpci.c linux-2.6.18-xen/drivers/xen/pciback/vpci.c
---- linux-2.6.18.3/drivers/xen/pciback/vpci.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/vpci.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/vpci.c b/drivers/xen/pciback/vpci.c
+--- a/drivers/xen/pciback/vpci.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/vpci.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,204 @@
+/*
+ * PCI Backend - Provides a Virtual PCI bus (with real devices)
@@ -64526,9 +60401,9 @@
+ kfree(vpci_dev);
+ pdev->pci_dev_data = NULL;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pciback/xenbus.c linux-2.6.18-xen/drivers/xen/pciback/xenbus.c
---- linux-2.6.18.3/drivers/xen/pciback/xenbus.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pciback/xenbus.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pciback/xenbus.c b/drivers/xen/pciback/xenbus.c
+--- a/drivers/xen/pciback/xenbus.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pciback/xenbus.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,458 @@
+/*
+ * PCI Backend Xenbus Setup - handles setup with frontend and xend
@@ -64564,7 +60439,7 @@
+ pdev->evtchn_irq = INVALID_EVTCHN_IRQ;
+ pdev->be_watching = 0;
+
-+ INIT_WORK(&pdev->op_work, pciback_do_op, pdev);
++ INIT_DELAYED_WORK(&pdev->op_work, pciback_do_op);
+
+ if (pciback_init_devices(pdev)) {
+ kfree(pdev);
@@ -64988,9 +60863,9 @@
+{
+ xenbus_unregister_driver(&xenbus_pciback_driver);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pcifront/Makefile linux-2.6.18-xen/drivers/xen/pcifront/Makefile
---- linux-2.6.18.3/drivers/xen/pcifront/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pcifront/Makefile 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pcifront/Makefile b/drivers/xen/pcifront/Makefile
+--- a/drivers/xen/pcifront/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pcifront/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,7 @@
+obj-y += pcifront.o
+
@@ -64999,9 +60874,9 @@
+ifeq ($(CONFIG_XEN_PCIDEV_FE_DEBUG),y)
+EXTRA_CFLAGS += -DDEBUG
+endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pcifront/pci.c linux-2.6.18-xen/drivers/xen/pcifront/pci.c
---- linux-2.6.18.3/drivers/xen/pcifront/pci.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pcifront/pci.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pcifront/pci.c b/drivers/xen/pcifront/pci.c
+--- a/drivers/xen/pcifront/pci.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pcifront/pci.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,46 @@
+/*
+ * PCI Frontend Operations - ensure only one PCI frontend runs at a time
@@ -65049,9 +60924,9 @@
+
+ spin_unlock(&pcifront_dev_lock);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pcifront/pcifront.h linux-2.6.18-xen/drivers/xen/pcifront/pcifront.h
---- linux-2.6.18.3/drivers/xen/pcifront/pcifront.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pcifront/pcifront.h 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pcifront/pcifront.h b/drivers/xen/pcifront/pcifront.h
+--- a/drivers/xen/pcifront/pcifront.h 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pcifront/pcifront.h 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,40 @@
+/*
+ * PCI Frontend - Common data structures & function declarations
@@ -65093,9 +60968,9 @@
+void pcifront_free_roots(struct pcifront_device *pdev);
+
+#endif /* __XEN_PCIFRONT_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pcifront/pci_op.c linux-2.6.18-xen/drivers/xen/pcifront/pci_op.c
---- linux-2.6.18.3/drivers/xen/pcifront/pci_op.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pcifront/pci_op.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pcifront/pci_op.c b/drivers/xen/pcifront/pci_op.c
+--- a/drivers/xen/pcifront/pci_op.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pcifront/pci_op.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,273 @@
+/*
+ * PCI Frontend Operations - Communicates with frontend
@@ -65370,9 +61245,9 @@
+ kfree(bus_entry);
+ }
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/pcifront/xenbus.c linux-2.6.18-xen/drivers/xen/pcifront/xenbus.c
---- linux-2.6.18.3/drivers/xen/pcifront/xenbus.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/pcifront/xenbus.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/pcifront/xenbus.c b/drivers/xen/pcifront/xenbus.c
+--- a/drivers/xen/pcifront/xenbus.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/pcifront/xenbus.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,295 @@
+/*
+ * PCI Frontend Xenbus Setup - handles setup with backend (imports page/evtchn)
@@ -65669,15 +61544,15 @@
+
+/* Initialize after the Xen PCI Frontend Stub is initialized */
+subsys_initcall(pcifront_init);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/privcmd/Makefile linux-2.6.18-xen/drivers/xen/privcmd/Makefile
---- linux-2.6.18.3/drivers/xen/privcmd/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/privcmd/Makefile 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/privcmd/Makefile b/drivers/xen/privcmd/Makefile
+--- a/drivers/xen/privcmd/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/privcmd/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,2 @@
+
+obj-$(CONFIG_XEN_PRIVCMD) := privcmd.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/privcmd/privcmd.c linux-2.6.18-xen/drivers/xen/privcmd/privcmd.c
---- linux-2.6.18.3/drivers/xen/privcmd/privcmd.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/privcmd/privcmd.c 2006-12-05 18:42:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/privcmd/privcmd.c b/drivers/xen/privcmd/privcmd.c
+--- a/drivers/xen/privcmd/privcmd.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/privcmd/privcmd.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,294 @@
+/******************************************************************************
+ * privcmd.c
@@ -65973,9 +61848,9 @@
+}
+
+__initcall(privcmd_init);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/tpmback/common.h linux-2.6.18-xen/drivers/xen/tpmback/common.h
---- linux-2.6.18.3/drivers/xen/tpmback/common.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/tpmback/common.h 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/tpmback/common.h b/drivers/xen/tpmback/common.h
+--- a/drivers/xen/tpmback/common.h 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/tpmback/common.h 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,86 @@
+/******************************************************************************
+ * drivers/xen/tpmback/common.h
@@ -66063,9 +61938,9 @@
+}
+
+#endif /* __TPMIF__BACKEND__COMMON_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/tpmback/interface.c linux-2.6.18-xen/drivers/xen/tpmback/interface.c
---- linux-2.6.18.3/drivers/xen/tpmback/interface.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/tpmback/interface.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/tpmback/interface.c b/drivers/xen/tpmback/interface.c
+--- a/drivers/xen/tpmback/interface.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/tpmback/interface.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,182 @@
+ /*****************************************************************************
+ * drivers/xen/tpmback/interface.c
@@ -66084,7 +61959,7 @@
+#include <xen/balloon.h>
+#include <xen/gnttab.h>
+
-+static kmem_cache_t *tpmif_cachep;
++static struct kmem_cache *tpmif_cachep;
+int num_frontends = 0;
+
+LIST_HEAD(tpmif_list);
@@ -66249,17 +62124,17 @@
+{
+ kmem_cache_destroy(tpmif_cachep);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/tpmback/Makefile linux-2.6.18-xen/drivers/xen/tpmback/Makefile
---- linux-2.6.18.3/drivers/xen/tpmback/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/tpmback/Makefile 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/tpmback/Makefile b/drivers/xen/tpmback/Makefile
+--- a/drivers/xen/tpmback/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/tpmback/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,4 @@
+
+obj-$(CONFIG_XEN_TPMDEV_BACKEND) += tpmbk.o
+
+tpmbk-y += tpmback.o interface.o xenbus.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/tpmback/tpmback.c linux-2.6.18-xen/drivers/xen/tpmback/tpmback.c
---- linux-2.6.18.3/drivers/xen/tpmback/tpmback.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/tpmback/tpmback.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/tpmback/tpmback.c b/drivers/xen/tpmback/tpmback.c
+--- a/drivers/xen/tpmback/tpmback.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/tpmback/tpmback.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,944 @@
+/******************************************************************************
+ * drivers/xen/tpmback/tpmback.c
@@ -67205,9 +63080,9 @@
+}
+
+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/tpmback/xenbus.c linux-2.6.18-xen/drivers/xen/tpmback/xenbus.c
---- linux-2.6.18.3/drivers/xen/tpmback/xenbus.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/tpmback/xenbus.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/tpmback/xenbus.c b/drivers/xen/tpmback/xenbus.c
+--- a/drivers/xen/tpmback/xenbus.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/tpmback/xenbus.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,287 @@
+/* Xenbus code for tpmif backend
+ Copyright (C) 2005 IBM Corporation
@@ -67496,9 +63371,9 @@
+{
+ xenbus_unregister_driver(&tpmback);
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/util.c linux-2.6.18-xen/drivers/xen/util.c
---- linux-2.6.18.3/drivers/xen/util.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/util.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/util.c b/drivers/xen/util.c
+--- a/drivers/xen/util.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/util.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,70 @@
+
+#include <linux/mm.h>
@@ -67570,9 +63445,9 @@
+ preempt_enable();
+}
+EXPORT_SYMBOL_GPL(unlock_vm_area);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/xenbus/Makefile linux-2.6.18-xen/drivers/xen/xenbus/Makefile
---- linux-2.6.18.3/drivers/xen/xenbus/Makefile 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/xenbus/Makefile 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/xenbus/Makefile b/drivers/xen/xenbus/Makefile
+--- a/drivers/xen/xenbus/Makefile 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/xenbus/Makefile 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,12 @@
+obj-y += xenbus.o
+obj-$(CONFIG_XEN_BACKEND) += xenbus_be.o
@@ -67586,9 +63461,9 @@
+xenbus-objs += xenbus_xs.o
+xenbus-objs += xenbus_probe.o
+obj-$(CONFIG_XEN_XENBUS_DEV) += xenbus_dev.o
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/xenbus/xenbus_backend_client.c linux-2.6.18-xen/drivers/xen/xenbus/xenbus_backend_client.c
---- linux-2.6.18.3/drivers/xen/xenbus/xenbus_backend_client.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/xenbus/xenbus_backend_client.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/xenbus/xenbus_backend_client.c b/drivers/xen/xenbus/xenbus_backend_client.c
+--- a/drivers/xen/xenbus/xenbus_backend_client.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/xenbus/xenbus_backend_client.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,147 @@
+/******************************************************************************
+ * Backend-client-facing interface for the Xenbus driver. In other words, the
@@ -67737,10 +63612,10 @@
+EXPORT_SYMBOL_GPL(xenbus_dev_is_online);
+
+MODULE_LICENSE("Dual BSD/GPL");
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/xenbus/xenbus_client.c linux-2.6.18-xen/drivers/xen/xenbus/xenbus_client.c
---- linux-2.6.18.3/drivers/xen/xenbus/xenbus_client.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/xenbus/xenbus_client.c 2006-11-19 14:26:56.000000000 +0100
-@@ -0,0 +1,299 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/xenbus/xenbus_client.c b/drivers/xen/xenbus/xenbus_client.c
+--- a/drivers/xen/xenbus/xenbus_client.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/xenbus/xenbus_client.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,300 @@
+/******************************************************************************
+ * Client-facing interface for the Xenbus driver. In other words, the
+ * interface between the Xenbus and the device-specific code, be it the
@@ -67794,6 +63669,7 @@
+ };
+ return (state < ARRAY_SIZE(name)) ? name[state] : "INVALID";
+}
++EXPORT_SYMBOL_GPL(xenbus_strstate);
+
+int xenbus_watch_path(struct xenbus_device *dev, const char *path,
+ struct xenbus_watch *watch,
@@ -68040,9 +63916,9 @@
+ return result;
+}
+EXPORT_SYMBOL_GPL(xenbus_read_driver_state);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/xenbus/xenbus_comms.c linux-2.6.18-xen/drivers/xen/xenbus/xenbus_comms.c
---- linux-2.6.18.3/drivers/xen/xenbus/xenbus_comms.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/xenbus/xenbus_comms.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/xenbus/xenbus_comms.c b/drivers/xen/xenbus/xenbus_comms.c
+--- a/drivers/xen/xenbus/xenbus_comms.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/xenbus/xenbus_comms.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,203 @@
+/******************************************************************************
+ * xenbus_comms.c
@@ -68087,13 +63963,13 @@
+
+static int xenbus_irq;
+
-+extern void xenbus_probe(void *);
++extern void xenbus_probe(struct work_struct *);
+extern int xenstored_ready;
-+static DECLARE_WORK(probe_work, xenbus_probe, NULL);
++static DECLARE_WORK(probe_work, xenbus_probe);
+
+DECLARE_WAIT_QUEUE_HEAD(xb_waitq);
+
-+static irqreturn_t wake_waiting(int irq, void *unused, struct pt_regs *regs)
++static irqreturn_t wake_waiting(int irq, void *unused)
+{
+ if (unlikely(xenstored_ready == 0)) {
+ xenstored_ready = 1;
@@ -68247,9 +64123,9 @@
+
+ return 0;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/xenbus/xenbus_comms.h linux-2.6.18-xen/drivers/xen/xenbus/xenbus_comms.h
---- linux-2.6.18.3/drivers/xen/xenbus/xenbus_comms.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/xenbus/xenbus_comms.h 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/xenbus/xenbus_comms.h b/drivers/xen/xenbus/xenbus_comms.h
+--- a/drivers/xen/xenbus/xenbus_comms.h 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/xenbus/xenbus_comms.h 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,45 @@
+/*
+ * Private include for xenbus communications.
@@ -68296,9 +64172,9 @@
+extern int xen_store_evtchn;
+
+#endif /* _XENBUS_COMMS_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/xenbus/xenbus_dev.c linux-2.6.18-xen/drivers/xen/xenbus/xenbus_dev.c
---- linux-2.6.18.3/drivers/xen/xenbus/xenbus_dev.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/xenbus/xenbus_dev.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/xenbus/xenbus_dev.c b/drivers/xen/xenbus/xenbus_dev.c
+--- a/drivers/xen/xenbus/xenbus_dev.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/xenbus/xenbus_dev.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,356 @@
+/*
+ * xenbus_dev.c
@@ -68656,10 +64532,10 @@
+
+ return 0;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/xenbus/xenbus_probe.c linux-2.6.18-xen/drivers/xen/xenbus/xenbus_probe.c
---- linux-2.6.18.3/drivers/xen/xenbus/xenbus_probe.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/xenbus/xenbus_probe.c 2006-11-19 14:26:56.000000000 +0100
-@@ -0,0 +1,1189 @@
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/xenbus/xenbus_probe.c b/drivers/xen/xenbus/xenbus_probe.c
+--- a/drivers/xen/xenbus/xenbus_probe.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/xenbus/xenbus_probe.c 2007-03-14 10:55:36.000000000 +0100
+@@ -0,0 +1,1208 @@
+/******************************************************************************
+ * Talks to Xen Store to figure out what devices we have.
+ *
@@ -69282,8 +65158,13 @@
+ if (err)
+ goto fail;
+
-+ device_create_file(&xendev->dev, &dev_attr_nodename);
-+ device_create_file(&xendev->dev, &dev_attr_devtype);
++ err = device_create_file(&xendev->dev, &dev_attr_nodename);
++ if (err)
++ goto fail;
++
++ err = device_create_file(&xendev->dev, &dev_attr_devtype);
++ if (err)
++ goto fail;
+
+ return 0;
+fail:
@@ -69592,7 +65473,7 @@
+EXPORT_SYMBOL_GPL(unregister_xenstore_notifier);
+
+
-+void xenbus_probe(void *unused)
++void xenbus_probe(struct work_struct *unused)
+{
+ BUG_ON((xenstored_ready <= 0));
+
@@ -69660,8 +65541,12 @@
+ return -ENODEV;
+
+ /* Register ourselves with the kernel bus subsystem */
-+ bus_register(&xenbus_frontend.bus);
-+ bus_register(&xenbus_backend.bus);
++ err = bus_register(&xenbus_frontend.bus);
++ if (err)
++ goto err_frontend_bus;
++ err = bus_register(&xenbus_backend.bus);
++ if (err)
++ goto err_backend_bus;
+
+ /*
+ * Domain0 doesn't have a store_evtchn or store_mfn yet.
@@ -69671,8 +65556,10 @@
+
+ /* Allocate page. */
+ page = get_zeroed_page(GFP_KERNEL);
-+ if (!page)
-+ return -ENOMEM;
++ if (!page) {
++ err = -ENOMEM;
++ goto err_nomem;
++ }
+
+ xen_store_mfn = xen_start_info->store_mfn =
+ pfn_to_mfn(virt_to_phys((void *)page) >>
@@ -69731,8 +65618,12 @@
+ }
+
+ /* Register ourselves with the kernel device subsystem */
-+ device_register(&xenbus_frontend.dev);
-+ device_register(&xenbus_backend.dev);
++ err = device_register(&xenbus_frontend.dev);
++ if (err)
++ goto err;
++ err = device_register(&xenbus_backend.dev);
++ if (err)
++ goto err;
+
+ if (!is_initial_xendomain())
+ xenbus_probe(NULL);
@@ -69748,7 +65639,11 @@
+ * must exist because front/backend drivers will use them when they are
+ * registered.
+ */
-+
++ err_nomem:
++ bus_unregister(&xenbus_frontend.bus);
++ err_backend_bus:
++ bus_unregister(&xenbus_frontend.bus);
++ err_frontend_bus:
+ return err;
+}
+
@@ -69849,9 +65744,9 @@
+
+late_initcall(boot_wait_for_devices);
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/drivers/xen/xenbus/xenbus_xs.c linux-2.6.18-xen/drivers/xen/xenbus/xenbus_xs.c
---- linux-2.6.18.3/drivers/xen/xenbus/xenbus_xs.c 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/drivers/xen/xenbus/xenbus_xs.c 2006-11-19 14:26:56.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/drivers/xen/xenbus/xenbus_xs.c b/drivers/xen/xenbus/xenbus_xs.c
+--- a/drivers/xen/xenbus/xenbus_xs.c 1970-01-01 01:00:00.000000000 +0100
++++ b/drivers/xen/xenbus/xenbus_xs.c 2007-03-14 10:55:36.000000000 +0100
@@ -0,0 +1,853 @@
+/******************************************************************************
+ * xenbus_xs.c
@@ -70706,36 +66601,26 @@
+
+ return 0;
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/fs/Kconfig linux-2.6.18-xen/fs/Kconfig
---- linux-2.6.18.3/fs/Kconfig 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/fs/Kconfig 2006-11-19 14:26:56.000000000 +0100
-@@ -865,6 +865,7 @@
+diff -x '.hg*' -x '.git*' -urN a/fs/Kconfig b/fs/Kconfig
+--- a/fs/Kconfig 2007-03-16 18:49:35.000000000 +0100
++++ b/fs/Kconfig 2007-03-14 10:55:36.000000000 +0100
+@@ -1003,6 +1003,7 @@
config HUGETLBFS
bool "HugeTLB file system support"
- depends X86 || IA64 || PPC64 || SPARC64 || SUPERH || BROKEN
-+ depends !XEN
+ depends on 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 -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-generic/vmlinux.lds.h linux-2.6.18-xen/include/asm-generic/vmlinux.lds.h
---- linux-2.6.18.3/include/asm-generic/vmlinux.lds.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-generic/vmlinux.lds.h 2006-11-19 14:27:04.000000000 +0100
-@@ -194,3 +194,6 @@
- .stab.index 0 : { *(.stab.index) } \
- .stab.indexstr 0 : { *(.stab.indexstr) } \
- .comment 0 : { *(.comment) }
-+
-+#define NOTES \
-+ .notes : { *(.note.*) } :note
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/apic.h linux-2.6.18-xen/include/asm-i386/apic.h
---- linux-2.6.18.3/include/asm-i386/apic.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-i386/apic.h 2006-11-19 14:27:04.000000000 +0100
-@@ -119,10 +119,12 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/apic.h b/include/asm-i386/apic.h
+--- a/include/asm-i386/apic.h 2007-03-16 18:49:38.000000000 +0100
++++ b/include/asm-i386/apic.h 2007-03-14 10:55:42.000000000 +0100
+@@ -116,10 +116,12 @@
- extern int disable_timer_pin_1;
+ extern void enable_NMI_through_LVT0 (void * dummy);
+#ifndef CONFIG_XEN
- void smp_send_timer_broadcast_ipi(struct pt_regs *regs);
+ 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
@@ -70743,20 +66628,38 @@
extern int timer_over_8254;
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/fixmap.h linux-2.6.18-xen/include/asm-i386/fixmap.h
---- linux-2.6.18.3/include/asm-i386/fixmap.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-i386/fixmap.h 2006-12-05 18:42:37.000000000 +0100
-@@ -19,6 +19,7 @@
- * Leave one empty page between vmalloc'ed areas and
- * the start of the fixmap.
- */
-+
- #define __FIXADDR_TOP 0xfffff000
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/highmem.h b/include/asm-i386/highmem.h
+--- a/include/asm-i386/highmem.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-i386/highmem.h 2007-03-14 10:55:42.000000000 +0100
+@@ -68,6 +68,9 @@
+ void *kmap(struct page *page);
+ void kunmap(struct page *page);
+ void *kmap_atomic(struct page *page, enum km_type type);
++#ifdef CONFIG_XEN
++void *kmap_atomic_pte(struct page *page, enum km_type type);
++#endif
+ 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);
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/kmap_types.h b/include/asm-i386/kmap_types.h
+--- a/include/asm-i386/kmap_types.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-i386/kmap_types.h 2007-03-14 10:55:42.000000000 +0100
+@@ -22,7 +22,12 @@
+ D(10) KM_IRQ1,
+ D(11) KM_SOFTIRQ0,
+ D(12) KM_SOFTIRQ1,
++#ifdef CONFIG_XEN
++D(13) KM_SWIOTLB,
++D(14) KM_TYPE_NR
++#else
+ D(13) KM_TYPE_NR
++#endif
+ };
- #ifndef __ASSEMBLY__
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-default/mach_traps.h linux-2.6.18-xen/include/asm-i386/mach-default/mach_traps.h
---- linux-2.6.18.3/include/asm-i386/mach-default/mach_traps.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-i386/mach-default/mach_traps.h 2006-11-19 14:27:04.000000000 +0100
+ #undef D
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-default/mach_traps.h b/include/asm-i386/mach-default/mach_traps.h
+--- a/include/asm-i386/mach-default/mach_traps.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-i386/mach-default/mach_traps.h 2007-03-14 10:55:42.000000000 +0100
@@ -15,6 +15,18 @@
outb(reason, 0x61);
}
@@ -70776,9 +66679,9 @@
static inline unsigned char get_nmi_reason(void)
{
return inb(0x61);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/agp.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/agp.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/agp.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/agp.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/agp.h b/include/asm-i386/mach-xen/asm/agp.h
+--- a/include/asm-i386/mach-xen/asm/agp.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/agp.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,37 @@
+#ifndef AGP_H
+#define AGP_H 1
@@ -70817,29 +66720,26 @@
+ dma_free_coherent(NULL,PAGE_SIZE<<(order),(table),virt_to_bus(table))
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/desc.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/desc.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/desc.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/desc.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,164 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/desc.h b/include/asm-i386/mach-xen/asm/desc.h
+--- a/include/asm-i386/mach-xen/asm/desc.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/desc.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,222 @@
+#ifndef __ARCH_DESC_H
+#define __ARCH_DESC_H
+
+#include <asm/ldt.h>
+#include <asm/segment.h>
+
-+#define CPU_16BIT_STACK_SIZE 1024
-+
+#ifndef __ASSEMBLY__
+
+#include <linux/preempt.h>
+#include <linux/smp.h>
++#include <linux/percpu.h>
+
+#include <asm/mmu.h>
+
+extern struct desc_struct cpu_gdt_table[GDT_ENTRIES];
+
-+DECLARE_PER_CPU(unsigned char, cpu_16bit_stack[CPU_16BIT_STACK_SIZE]);
-+
+struct Xgt_desc_struct {
+ unsigned short size;
+ unsigned long address __attribute__((packed));
@@ -70855,52 +66755,109 @@
+ return (struct desc_struct *)per_cpu(cpu_gdt_descr, cpu).address;
+}
+
++extern struct desc_struct idt_table[];
++extern void set_intr_gate(unsigned int irq, void * addr);
++
++static inline void pack_descriptor(__u32 *a, __u32 *b,
++ unsigned long base, unsigned long limit, unsigned char type, unsigned char flags)
++{
++ *a = ((base & 0xffff) << 16) | (limit & 0xffff);
++ *b = (base & 0xff000000) | ((base & 0xff0000) >> 16) |
++ (limit & 0x000f0000) | ((type & 0xff) << 8) | ((flags & 0xf) << 20);
++}
++
++static inline void pack_gate(__u32 *a, __u32 *b,
++ unsigned long base, unsigned short seg, unsigned char type, unsigned char flags)
++{
++ *a = (seg << 16) | (base & 0xffff);
++ *b = (base & 0xffff0000) | ((type & 0xff) << 8) | (flags & 0xff);
++}
++
++#define DESCTYPE_LDT 0x82 /* present, system, DPL-0, LDT */
++#define DESCTYPE_TSS 0x89 /* present, system, DPL-0, 32-bit TSS */
++#define DESCTYPE_TASK 0x85 /* present, system, DPL-0, task gate */
++#define DESCTYPE_INT 0x8e /* present, system, DPL-0, interrupt gate */
++#define DESCTYPE_TRAP 0x8f /* present, system, DPL-0, trap gate */
++#define DESCTYPE_DPL3 0x60 /* DPL-3 */
++#define DESCTYPE_S 0x10 /* !system */
++
++#ifdef CONFIG_PARAVIRT
++#include <asm/paravirt.h>
++#else
+#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 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))
++#define load_tr(tr) __asm__ __volatile("ltr %0"::"m" (tr))
++#define load_ldt(ldt) __asm__ __volatile("lldt %0"::"m" (ldt))
+
+#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))
++#define store_tr(tr) __asm__ ("str %0":"=m" (tr))
++#define store_ldt(ldt) __asm__ ("sldt %0":"=m" (ldt))
+
-+/*
-+ * 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);
++#if TLS_SIZE != 24
++# error update this code.
++#endif
++
++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 _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))
++extern int write_ldt_entry(void *ldt, int entry, __u32 entry_a, __u32 entry_b);
++#define write_gdt_entry(dt, entry, a, b) write_dt_entry(dt, entry, a, b)
++#define write_idt_entry(dt, entry, a, b) write_dt_entry(dt, entry, a, b)
+
-+#ifndef CONFIG_X86_NO_TSS
-+static inline void __set_tss_desc(unsigned int cpu, unsigned int entry, void *addr)
++static inline void write_dt_entry(void *dt, int entry, __u32 entry_a, __u32 entry_b)
+{
-+ _set_tssldt_desc(&get_cpu_gdt_table(cpu)[entry], (int)addr,
-+ offsetof(struct tss_struct, __cacheline_filler) - 1, 0x89);
++ __u32 *lp = (__u32 *)((char *)dt + entry*8);
++ *lp = entry_a;
++ *(lp+1) = entry_b;
+}
+
-+#define set_tss_desc(cpu,addr) __set_tss_desc(cpu, GDT_ENTRY_TSS, addr)
-+#endif
++#define set_ldt native_set_ldt
++#endif /* CONFIG_PARAVIRT */
++
++static inline fastcall void native_set_ldt(const void *addr,
++ unsigned int entries)
++{
++ if (likely(entries == 0))
++ __asm__ __volatile__("lldt %w0"::"q" (0));
++ else {
++ unsigned cpu = smp_processor_id();
++ __u32 a, b;
++
++ pack_descriptor(&a, &b, (unsigned long)addr,
++ entries * sizeof(struct desc_struct) - 1,
++ DESCTYPE_LDT, 0);
++ write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_LDT, a, b);
++ __asm__ __volatile__("lldt %w0"::"q" (GDT_ENTRY_LDT*8));
++ }
++}
+
-+static inline void set_ldt_desc(unsigned int cpu, void *addr, unsigned int size)
++static inline void _set_gate(int gate, unsigned int type, void *addr, unsigned short seg)
++{
++ __u32 a, b;
++ pack_gate(&a, &b, (unsigned long)addr, seg, type, 0);
++ write_idt_entry(idt_table, gate, a, b);
++}
++
++#ifndef CONFIG_X86_NO_TSS
++static inline void __set_tss_desc(unsigned int cpu, unsigned int entry, const void *addr)
+{
-+ _set_tssldt_desc(&get_cpu_gdt_table(cpu)[GDT_ENTRY_LDT], (int)addr, ((size << 3)-1), 0x82);
++ __u32 a, b;
++ pack_descriptor(&a, &b, (unsigned long)addr,
++ offsetof(struct tss_struct, __cacheline_filler) - 1,
++ DESCTYPE_TSS, 0);
++ write_gdt_entry(get_cpu_gdt_table(cpu), entry, a, b);
+}
+
++#define set_tss_desc(cpu,addr) __set_tss_desc(cpu, GDT_ENTRY_TSS, addr)
++#endif
++
+#define LDT_entry_a(info) \
+ ((((info)->base_addr & 0x0000ffff) << 16) | ((info)->limit & 0x0ffff))
+
@@ -70926,19 +66883,6 @@
+ (info)->seg_not_present == 1 && \
+ (info)->useable == 0 )
+
-+extern int write_ldt_entry(void *ldt, int entry, __u32 entry_a, __u32 entry_b);
-+
-+#if TLS_SIZE != 24
-+# error update this code.
-+#endif
-+
-+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
-+}
-+
+static inline void clear_LDT(void)
+{
+ int cpu = get_cpu();
@@ -70955,22 +66899,16 @@
+/*
+ * load one particular LDT into the current CPU
+ */
-+static inline void load_LDT_nolock(mm_context_t *pc, int cpu)
++static inline void load_LDT_nolock(mm_context_t *pc)
+{
-+ void *segments = pc->ldt;
-+ int count = pc->size;
-+
-+ if (likely(!count))
-+ segments = NULL;
-+
-+ xen_set_ldt((unsigned long)segments, count);
++ xen_set_ldt((unsigned long)pc->ldt, pc->size);
+}
+
+static inline void load_LDT(mm_context_t *pc)
+{
-+ int cpu = get_cpu();
-+ load_LDT_nolock(pc, cpu);
-+ put_cpu();
++ preempt_disable();
++ load_LDT_nolock(pc);
++ preempt_enable();
+}
+
+static inline unsigned long get_desc_base(unsigned long *desc)
@@ -70982,25 +66920,45 @@
+ return base;
+}
+
++#else /* __ASSEMBLY__ */
++
++/*
++ * GET_DESC_BASE reads the descriptor base of the specified segment.
++ *
++ * Args:
++ * idx - descriptor index
++ * gdt - GDT pointer
++ * base - 32bit register to which the base will be written
++ * lo_w - lo word of the "base" register
++ * lo_b - lo byte of the "base" register
++ * hi_b - hi byte of the low word of the "base" register
++ *
++ * Example:
++ * GET_DESC_BASE(GDT_ENTRY_ESPFIX_SS, %ebx, %eax, %ax, %al, %ah)
++ * Will read the base address of GDT_ENTRY_ESPFIX_SS and put it into %eax.
++ */
++#define GET_DESC_BASE(idx, gdt, base, lo_w, lo_b, hi_b) \
++ movb idx*8+4(gdt), lo_b; \
++ movb idx*8+7(gdt), hi_b; \
++ shll $16, base; \
++ movw idx*8+2(gdt), lo_w;
++
+#endif /* !__ASSEMBLY__ */
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/dma-mapping.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/dma-mapping.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/dma-mapping.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/dma-mapping.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/dma-mapping.h b/include/asm-i386/mach-xen/asm/dma-mapping.h
+--- a/include/asm-i386/mach-xen/asm/dma-mapping.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/dma-mapping.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,151 @@
+#ifndef _ASM_I386_DMA_MAPPING_H
+#define _ASM_I386_DMA_MAPPING_H
+
-+/*
-+ * IOMMU interface. See Documentation/DMA-mapping.txt and DMA-API.txt for
-+ * documentation.
-+ */
-+
+#include <linux/mm.h>
++
+#include <asm/cache.h>
+#include <asm/io.h>
+#include <asm/scatterlist.h>
++#include <asm/bug.h>
+#include <asm/swiotlb.h>
+
+static inline int
@@ -71038,10 +66996,9 @@
+dma_unmap_single(struct device *dev, dma_addr_t dma_addr, size_t size,
+ enum dma_data_direction direction);
+
-+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
++dma_map_sg(struct device *dev, struct scatterlist *sg, int nents,
++ enum dma_data_direction direction);
+
+extern dma_addr_t
+dma_map_page(struct device *dev, struct page *page, unsigned long offset,
@@ -71052,6 +67009,10 @@
+ enum dma_data_direction direction);
+
+extern void
++dma_unmap_sg(struct device *dev, struct scatterlist *sg, int nhwentries,
++ enum dma_data_direction direction);
++
++extern void
+dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle, size_t size,
+ enum dma_data_direction direction);
+
@@ -71118,10 +67079,10 @@
+ return (1 << INTERNODE_CACHE_SHIFT);
+}
+
-+#define dma_is_consistent(d) (1)
++#define dma_is_consistent(d, h) (1)
+
+static inline void
-+dma_cache_sync(void *vaddr, size_t size,
++dma_cache_sync(struct device *dev, void *vaddr, size_t size,
+ enum dma_data_direction direction)
+{
+ flush_write_buffers();
@@ -71140,10 +67101,10 @@
+ dma_addr_t device_addr, size_t size);
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/fixmap.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/fixmap.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/fixmap.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/fixmap.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,156 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/fixmap.h b/include/asm-i386/mach-xen/asm/fixmap.h
+--- a/include/asm-i386/mach-xen/asm/fixmap.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/fixmap.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,163 @@
+/*
+ * fixmap.h: compile-time virtual memory allocation
+ *
@@ -71165,7 +67126,13 @@
+ * Leave one empty page between vmalloc'ed areas and
+ * the start of the fixmap.
+ */
++#ifndef CONFIG_COMPAT_VDSO
+extern unsigned long __FIXADDR_TOP;
++#else
++#define __FIXADDR_TOP 0xfffff000
++#define FIXADDR_USER_START __fix_to_virt(FIX_VDSO)
++#define FIXADDR_USER_END __fix_to_virt(FIX_VDSO - 1)
++#endif
+
+#ifndef __ASSEMBLY__
+#include <linux/kernel.h>
@@ -71245,6 +67212,7 @@
+extern void __set_fixmap(enum fixed_addresses idx,
+ maddr_t phys, pgprot_t flags);
+
++extern void reserve_top_address(unsigned long reserve);
+extern void set_fixaddr_top(unsigned long top);
+
+#define set_fixmap(idx, phys) \
@@ -71300,9 +67268,9 @@
+
+#endif /* !__ASSEMBLY__ */
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/floppy.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/floppy.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/floppy.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/floppy.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/floppy.h b/include/asm-i386/mach-xen/asm/floppy.h
+--- a/include/asm-i386/mach-xen/asm/floppy.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/floppy.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,147 @@
+/*
+ * Architecture specific parts of the Floppy driver
@@ -71349,14 +67317,14 @@
+static int virtual_dma_mode;
+static int doing_pdma;
+
-+static irqreturn_t floppy_hardint(int irq, void *dev_id, struct pt_regs * regs)
++static irqreturn_t floppy_hardint(int irq, void *dev_id)
+{
+ register unsigned char st;
+ register int lcount;
+ register char *lptr;
+
+ if (!doing_pdma)
-+ return floppy_interrupt(irq, dev_id, regs);
++ return floppy_interrupt(irq, dev_id);
+
+ st = 1;
+ for(lcount=virtual_dma_count, lptr=virtual_dma_addr;
@@ -71379,7 +67347,7 @@
+ virtual_dma_residue += virtual_dma_count;
+ virtual_dma_count=0;
+ doing_pdma = 0;
-+ floppy_interrupt(irq, dev_id, regs);
++ floppy_interrupt(irq, dev_id);
+ return IRQ_HANDLED;
+ }
+ return IRQ_HANDLED;
@@ -71451,169 +67419,9 @@
+#define EXTRA_FLOPPY_PARAMS
+
+#endif /* __ASM_XEN_I386_FLOPPY_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/highmem.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/highmem.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/highmem.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/highmem.h 2006-11-19 14:27:04.000000000 +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>
-+ */
-+
-+#ifndef _ASM_HIGHMEM_H
-+#define _ASM_HIGHMEM_H
-+
-+#ifdef __KERNEL__
-+
-+#include <linux/interrupt.h>
-+#include <linux/threads.h>
-+#include <asm/kmap_types.h>
-+#include <asm/tlbflush.h>
-+
-+/* declarations for highmem.c */
-+extern unsigned long highstart_pfn, highend_pfn;
-+
-+extern pte_t *kmap_pte;
-+extern pgprot_t kmap_prot;
-+extern pte_t *pkmap_page_table;
-+
-+/*
-+ * 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))
-+
-+extern void * FASTCALL(kmap_high(struct page *page));
-+extern void FASTCALL(kunmap_high(struct page *page));
-+
-+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)
-+
-+#endif /* __KERNEL__ */
-+
-+#endif /* _ASM_HIGHMEM_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/hw_irq.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/hw_irq.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/hw_irq.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/hw_irq.h 2006-12-05 18:42:37.000000000 +0100
-@@ -0,0 +1,72 @@
-+#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>
-+ */
-+
-+#include <linux/profile.h>
-+#include <asm/atomic.h>
-+#include <asm/irq.h>
-+#include <asm/sections.h>
-+
-+struct hw_interrupt_type;
-+
-+#define NMI_VECTOR 0x02
-+
-+/*
-+ * 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
-+ */
-+
-+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
-+
-+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);
-+
-+extern unsigned long io_apic_irqs;
-+
-+extern atomic_t irq_err_count;
-+extern atomic_t irq_mis_count;
-+
-+#define IO_APIC_IRQ(x) (((x) >= 16) || ((1<<(x)) & io_apic_irqs))
-+
-+#endif /* _ASM_HW_IRQ_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/hypercall.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/hypercall.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/hypercall.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/hypercall.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/hypercall.h b/include/asm-i386/mach-xen/asm/hypercall.h
+--- a/include/asm-i386/mach-xen/asm/hypercall.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/hypercall.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,389 @@
+/******************************************************************************
+ * hypercall.h
@@ -72004,9 +67812,9 @@
+
+
+#endif /* __HYPERCALL_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/hypervisor.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/hypervisor.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/hypervisor.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/hypervisor.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/hypervisor.h b/include/asm-i386/mach-xen/asm/hypervisor.h
+--- a/include/asm-i386/mach-xen/asm/hypervisor.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/hypervisor.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,234 @@
+/******************************************************************************
+ * hypervisor.h
@@ -72242,10 +68050,10 @@
+}
+
+#endif /* __HYPERVISOR_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/io.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/io.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/io.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/io.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,389 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/io.h b/include/asm-i386/mach-xen/asm/io.h
+--- a/include/asm-i386/mach-xen/asm/io.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/io.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,368 @@
+#ifndef _ASM_IO_H
+#define _ASM_IO_H
+
@@ -72401,7 +68209,11 @@
+ */
+#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))
++
++static inline void *isa_bus_to_virt(unsigned long address)
++{
++ return (void *)(__fix_to_virt(FIX_ISAMAP_BEGIN) + address);
++}
+
+/*
+ * However PCI ones are not necessarily 1:1 and therefore these interfaces
@@ -72485,33 +68297,6 @@
+
+#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)
-+{
-+ int retval = 0;
-+ do {
-+ if (readb(io_addr) != *signature)
-+ goto out;
-+ io_addr++;
-+ signature++;
-+ length--;
-+ } while (length);
-+ retval = 1;
-+out:
-+ return retval;
-+}
-+
+/*
+ * Cache management
+ *
@@ -72544,11 +68329,11 @@
+
+#endif /* __KERNEL__ */
+
-+#ifdef SLOW_IO_BY_JUMPING
-+#define __SLOW_DOWN_IO "jmp 1f; 1: jmp 1f; 1:"
++#if defined(CONFIG_PARAVIRT)
++#include <asm/paravirt.h>
+#else
++
+#define __SLOW_DOWN_IO "outb %%al,$0x80;"
-+#endif
+
+static inline void slow_down_io(void) {
+ __asm__ __volatile__(
@@ -72559,6 +68344,8 @@
+ : : );
+}
+
++#endif
++
+#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)
@@ -72635,10 +68422,10 @@
+#define ARCH_HAS_DEV_MEM
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/irqflags.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/irqflags.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/irqflags.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/irqflags.h 2006-12-05 18:42:37.000000000 +0100
-@@ -0,0 +1,80 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/irqflags.h b/include/asm-i386/mach-xen/asm/irqflags.h
+--- a/include/asm-i386/mach-xen/asm/irqflags.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/irqflags.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,96 @@
+/*
+ * include/asm-i386/irqflags.h
+ *
@@ -72651,12 +68438,13 @@
+#ifndef _ASM_IRQFLAGS_H
+#define _ASM_IRQFLAGS_H
+
++#ifdef CONFIG_PARAVIRT
++#include <asm/paravirt.h>
++#else
+#ifndef __ASSEMBLY__
+
-+#define raw_local_save_flags(flags) \
-+ do { (flags) = __raw_local_save_flags(); } while (0)
-+
+unsigned long __raw_local_save_flags(void);
++
+void raw_local_irq_restore(unsigned long flags);
+void raw_local_irq_disable(void);
+void raw_local_irq_enable(void);
@@ -72673,20 +68461,35 @@
+ */
+void halt(void);
+
-+static inline int raw_irqs_disabled_flags(unsigned long flags)
-+{
-+ return flags != 0;
-+}
-+
-+int raw_irqs_disabled(void);
-+
+/*
+ * For spinlocks, etc:
+ */
+unsigned long __raw_local_irq_save(void);
++
++#else
++#define DISABLE_INTERRUPTS(clobbers) GET_VCPU_INFO ; \
++ __DISABLE_INTERRUPTS
++#define ENABLE_INTERRUPTS(clobbers) GET_VCPU_INFO ; \
++ __ENABLE_INTERRUPTS
++#define ENABLE_INTERRUPTS_SYSEXIT NET_DONE_YET
++#define INTERRUPT_RETURN iret
++#define GET_CR0_INTO_EAX NOT_DONE_YET
++#endif /* __ASSEMBLY__ */
++#endif /* CONFIG_PARAVIRT */
++
++#ifndef __ASSEMBLY__
++#define raw_local_save_flags(flags) \
++ do { (flags) = __raw_local_save_flags(); } while (0)
++
+#define raw_local_irq_save(flags) \
+ do { (flags) = __raw_local_irq_save(); } while (0)
+
++static inline int raw_irqs_disabled_flags(unsigned long flags)
++{
++ return flags != 0;
++}
++
++int raw_irqs_disabled(void);
+#endif /* __ASSEMBLY__ */
+
+/*
@@ -72719,44 +68522,9 @@
+#endif
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/kmap_types.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/kmap_types.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/kmap_types.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/kmap_types.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,31 @@
-+#ifndef _ASM_KMAP_TYPES_H
-+#define _ASM_KMAP_TYPES_H
-+
-+
-+#ifdef CONFIG_DEBUG_HIGHMEM
-+# define D(n) __KM_FENCE_##n ,
-+#else
-+# define D(n)
-+#endif
-+
-+enum km_type {
-+D(0) KM_BOUNCE_READ,
-+D(1) KM_SKB_SUNRPC_DATA,
-+D(2) KM_SKB_DATA_SOFTIRQ,
-+D(3) KM_USER0,
-+D(4) KM_USER1,
-+D(5) KM_BIO_SRC_IRQ,
-+D(6) KM_BIO_DST_IRQ,
-+D(7) KM_PTE0,
-+D(8) KM_PTE1,
-+D(9) KM_IRQ0,
-+D(10) KM_IRQ1,
-+D(11) KM_SOFTIRQ0,
-+D(12) KM_SOFTIRQ1,
-+D(13) KM_SWIOTLB,
-+D(14) KM_TYPE_NR
-+};
-+
-+#undef D
-+
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/maddr.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/maddr.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/maddr.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/maddr.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/maddr.h b/include/asm-i386/mach-xen/asm/maddr.h
+--- a/include/asm-i386/mach-xen/asm/maddr.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/maddr.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,160 @@
+#ifndef _I386_MADDR_H
+#define _I386_MADDR_H
@@ -72918,10 +68686,10 @@
+#define __pte_ma(x) ((pte_t) { (x) } )
+
+#endif /* _I386_MADDR_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/mmu_context.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/mmu_context.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/mmu_context.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/mmu_context.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,108 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/mmu_context.h b/include/asm-i386/mach-xen/asm/mmu_context.h
+--- a/include/asm-i386/mach-xen/asm/mmu_context.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/mmu_context.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,107 @@
+#ifndef __I386_SCHED_H
+#define __I386_SCHED_H
+
@@ -72955,10 +68723,9 @@
+ * 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"
++ asm volatile ( "mov %%fs,%0"
++ : "=m" (current->thread.fs));
++ asm volatile ( "movl %0,%%fs"
+ : : "r" (0) );
+}
+
@@ -72994,7 +68761,7 @@
+ * load the LDT, if the LDT is different:
+ */
+ if (unlikely(prev->context.ldt != next->context.ldt)) {
-+ /* load_LDT_nolock(&next->context, cpu) */
++ /* load_LDT_nolock(&next->context) */
+ op->cmd = MMUEXT_SET_LDT;
+ op->arg1.linear_addr = (unsigned long)next->context.ldt;
+ op->arg2.nr_ents = next->context.size;
@@ -73013,14 +68780,14 @@
+ * tlb flush IPI delivery. We must reload %cr3.
+ */
+ load_cr3(next->pgd);
-+ load_LDT_nolock(&next->context, cpu);
++ load_LDT_nolock(&next->context);
+ }
+ }
+#endif
+}
+
-+#define deactivate_mm(tsk, mm) \
-+ asm("movl %0,%%fs ; movl %0,%%gs": :"r" (0))
++#define deactivate_mm(tsk, mm) \
++ asm("movl %0,%%fs": :"r" (0));
+
+static inline void activate_mm(struct mm_struct *prev, struct mm_struct *next)
+{
@@ -73030,43 +68797,10 @@
+}
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/mmu.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/mmu.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/mmu.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/mmu.h 2006-11-19 14:27:04.000000000 +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.
-+ */
-+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)
-+
-+/* 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
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/page.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/page.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/page.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/page.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,220 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/page.h b/include/asm-i386/mach-xen/asm/page.h
+--- a/include/asm-i386/mach-xen/asm/page.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/page.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,224 @@
+#ifndef _I386_PAGE_H
+#define _I386_PAGE_H
+
@@ -73176,6 +68910,7 @@
+ return (unsigned long long)x.pte_high << 32 | x.pte_low;
+}
+#define HPAGE_SHIFT 21
++#include <asm-generic/pgtable-nopud.h>
+#else
+typedef struct { unsigned long pte_low; } pte_t;
+typedef struct { unsigned long pgd; } pgd_t;
@@ -73197,6 +68932,7 @@
+ return ret;
+}
+#define HPAGE_SHIFT 22
++#include <asm-generic/pgtable-nopmd.h>
+#endif
+#define PTE_MASK PAGE_MASK
+
@@ -73245,12 +68981,9 @@
+
+#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)
+
+#ifdef CONFIG_XEN_COMPAT_030002
+#undef LOAD_OFFSET
@@ -73261,6 +68994,9 @@
+#define VMALLOC_RESERVE ((unsigned long)__VMALLOC_RESERVE)
+#define MAXMEM (__FIXADDR_TOP-__PAGE_OFFSET-__VMALLOC_RESERVE)
+#define __pa(x) ((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) __pa(RELOC_HIDE((unsigned long)(x),0))
+#define __va(x) ((void *)((unsigned long)(x)+PAGE_OFFSET))
+#define pfn_to_kaddr(pfn) __va((pfn) << PAGE_SHIFT)
+#ifdef CONFIG_FLATMEM
@@ -73283,197 +69019,15 @@
+#include <asm-generic/memory_model.h>
+#include <asm-generic/page.h>
+
++#ifndef CONFIG_COMPAT_VDSO
+#define __HAVE_ARCH_GATE_AREA 1
-+#endif /* __KERNEL__ */
-+
-+#endif /* _I386_PAGE_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/param.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/param.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/param.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/param.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,23 @@
-+#ifndef _ASMi386_PARAM_H
-+#define _ASMi386_PARAM_H
-+
-+#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
-+
-+#ifndef HZ
-+#define HZ 100
-+#endif
-+
-+#define EXEC_PAGESIZE 4096
-+
-+#ifndef NOGROUP
-+#define NOGROUP (-1)
-+#endif
-+
-+#define MAXHOSTNAMELEN 64 /* max length of hostname */
-+#define COMMAND_LINE_SIZE 256
-+
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/pci.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pci.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/pci.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pci.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,153 @@
-+#ifndef __i386_PCI_H
-+#define __i386_PCI_H
-+
-+
-+#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 */
-+
-+#ifdef CONFIG_PCI
-+extern unsigned int pcibios_assign_all_busses(void);
-+#else
-+#define pcibios_assign_all_busses() 0
-+#endif
-+#define pcibios_scan_all_fns(a, b) 0
-+
-+extern unsigned long pci_mem_start;
-+#define PCIBIOS_MIN_IO 0x1000
-+#define PCIBIOS_MIN_MEM (pci_mem_start)
-+
-+#define PCIBIOS_MIN_CARDBUS_IO 0x4000
-+
-+void pcibios_config_init(void);
-+struct pci_bus * pcibios_scan_root(int bus);
-+
-+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);
-+
-+/* Dynamic DMA mapping stuff.
-+ * i386 has everything mapped statically.
-+ */
-+
-+#include <linux/types.h>
-+#include <linux/slab.h>
-+#include <asm/scatterlist.h>
-+#include <linux/string.h>
-+#include <asm/io.h>
-+
-+struct pci_dev;
-+
-+#ifdef CONFIG_SWIOTLB
-+
-+
-+/* On Xen we use SWIOTLB instead of blk-specific bounce buffers. */
-+#define PCI_DMA_BUS_IS_PHYS (0)
-+
-+#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
-+
-+/* 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)
-+
-+/* 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);
-+}
-+
-+static inline struct page *
-+pci_dac_dma_to_page(struct pci_dev *pdev, dma64_addr_t dma_addr)
-+{
-+ return pfn_to_page(dma_addr >> PAGE_SHIFT);
-+}
-+
-+static inline unsigned long
-+pci_dac_dma_to_offset(struct pci_dev *pdev, dma64_addr_t dma_addr)
-+{
-+ return (dma_addr & ~PAGE_MASK);
-+}
-+
-+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 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();
-+}
-+
-+#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)
-+{
-+}
-+
-+#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
-+
+#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>
-+
-+/* On Xen we have to scan all functions since Xen hides bridges from
-+ * us. If a bridge is at fn=0 and that slot has a multifunction
-+ * device, we won't find the additional devices without scanning all
-+ * functions. */
-+#undef pcibios_scan_all_fns
-+#define pcibios_scan_all_fns(a, b) 1
-+
-+#endif /* __i386_PCI_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/pgalloc.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pgalloc.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/pgalloc.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pgalloc.h 2006-11-19 14:27:04.000000000 +0100
++#endif /* _I386_PAGE_H */
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/pgalloc.h b/include/asm-i386/mach-xen/asm/pgalloc.h
+--- a/include/asm-i386/mach-xen/asm/pgalloc.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/pgalloc.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,63 @@
+#ifndef _I386_PGALLOC_H
+#define _I386_PGALLOC_H
@@ -73538,40 +69092,13 @@
+#define check_pgt_cache() do { } while (0)
+
+#endif /* _I386_PGALLOC_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/pgtable-2level-defs.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pgtable-2level-defs.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/pgtable-2level-defs.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pgtable-2level-defs.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,21 @@
-+#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
-+#define PTRS_PER_PGD_NO_HV (HYPERVISOR_VIRT_START >> PGDIR_SHIFT)
-+
-+/*
-+ * 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 -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/pgtable-2level.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pgtable-2level.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/pgtable-2level.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pgtable-2level.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,87 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/pgtable-2level.h b/include/asm-i386/mach-xen/asm/pgtable-2level.h
+--- a/include/asm-i386/mach-xen/asm/pgtable-2level.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/pgtable-2level.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,77 @@
+#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.\n", __FILE__, __LINE__, (e).pte_low)
+#define pgd_ERROR(e) \
@@ -73582,6 +69109,7 @@
+ * within a page table are directly modified. Thus, the following
+ * hook is made available.
+ */
++#ifndef CONFIG_PARAVIRT
+#define set_pte(pteptr, pteval) (*(pteptr) = pteval)
+
+#define set_pte_at(_mm,addr,ptep,pteval) do { \
@@ -73589,30 +69117,21 @@
+ 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_pmd(pmdptr, pmdval) xen_l2_entry_update((pmdptr), (pmdval))
++#endif
+
+#define set_pte_atomic(pteptr, pteval) set_pte(pteptr,pteval)
-+
-+#define set_pmd(pmdptr, pmdval) xen_l2_entry_update((pmdptr), (pmdval))
++#define set_pte_present(mm,addr,ptep,pteval) set_pte_at(mm,addr,ptep,pteval)
+
+#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 ptep_get_and_clear(mm,addr,xp) __pte_ma(xchg(&(xp)->pte_low, 0))
-+#define pte_same(a, b) ((a).pte_low == (b).pte_low)
-+#define pte_mfn(_pte) ((_pte).pte_low >> PAGE_SHIFT)
-+#define pte_pfn(_pte) mfn_to_local_pfn(pte_mfn(_pte))
-+
-+#define pte_page(_pte) pfn_to_page(pte_pfn(_pte))
++#define raw_ptep_get_and_clear(mm,addr,xp) __pte_ma(xchg(&(xp)->pte_low, 0))
+
++#define pte_mfn(_pte) ((_pte).pte_low >> PAGE_SHIFT)
++#define pte_page(x) pfn_to_page(pte_pfn(x))
+#define pte_none(x) (!(x).pte_low)
++#define pte_pfn(x) mfn_to_local_pfn(pte_mfn(x))
+#define pfn_pte(pfn, prot) __pte(((pfn) << PAGE_SHIFT) | pgprot_val(prot))
+#define pfn_pmd(pfn, prot) __pmd(((pfn) << PAGE_SHIFT) | pgprot_val(prot))
+
@@ -73654,44 +69173,13 @@
+void vmalloc_sync_all(void);
+
+#endif /* _I386_PGTABLE_2LEVEL_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/pgtable-3level-defs.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pgtable-3level-defs.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/pgtable-3level-defs.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pgtable-3level-defs.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,25 @@
-+#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
-+#define PTRS_PER_PGD_NO_HV 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
-+
-+/*
-+ * entries per page directory level
-+ */
-+#define PTRS_PER_PTE 512
-+
-+#endif /* _I386_PGTABLE_3LEVEL_DEFS_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/pgtable-3level.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pgtable-3level.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/pgtable-3level.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pgtable-3level.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,185 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/pgtable-3level.h b/include/asm-i386/mach-xen/asm/pgtable-3level.h
+--- a/include/asm-i386/mach-xen/asm/pgtable-3level.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/pgtable-3level.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,194 @@
+#ifndef _I386_PGTABLE_3LEVEL_H
+#define _I386_PGTABLE_3LEVEL_H
+
-+#include <asm-generic/pgtable-nopud.h>
-+
+/*
+ * Intel Physical Address Extension (PAE) Mode - three-level page
+ * tables on PPro+ CPUs.
@@ -73733,6 +69221,7 @@
+ return pte_x(pte);
+}
+
++#ifndef CONFIG_PARAVIRT
+/* 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
@@ -73751,11 +69240,27 @@
+}
+# define set_pte_atomic(pteptr,pteval) \
+ set_64bit((unsigned long long *)(pteptr),pte_val_ma(pteval))
++
++/*
++ * Since this is only called on user PTEs, and the page fault handler
++ * must handle the already racy situation of simultaneous page faults,
++ * we are justified in merely clearing the PTE present bit, followed
++ * by a set. The ordering here is important.
++ */
++static inline void set_pte_present(struct mm_struct *mm, unsigned long addr, pte_t *ptep, pte_t pte)
++{
++ ptep->pte_low = 0;
++ smp_wmb();
++ ptep->pte_high = pte.pte_high;
++ smp_wmb();
++ ptep->pte_low = pte.pte_low;
++}
+#else
+/* no writable pagetables */
+# define set_pte(pteptr,pteval) \
+ xen_l1_entry_update((pteptr), (pteval))
+# define set_pte_atomic(pteptr,pteval) set_pte(pteptr,pteval)
++# define set_pte_pressent(mm,addr,ptep,pte) set_pte_at(mm,addr,ptep,pteval)
+#endif
+
+#define set_pte_at(_mm,addr,ptep,pteval) do { \
@@ -73764,39 +69269,12 @@
+ 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_pmd(pmdptr,pmdval) \
+ xen_l2_entry_update((pmdptr), (pmdval))
+#define set_pud(pudptr,pudval) \
+ xen_l3_entry_update((pudptr), (pudval))
+
+/*
-+ * 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) { }
-+
-+#define pud_page(pud) \
-+((struct page *) __va(pud_val(pud) & PAGE_MASK))
-+
-+#define pud_page_kernel(pud) \
-+((unsigned long) __va(pud_val(pud) & PAGE_MASK))
-+
-+
-+/* Find an entry in the second-level page table.. */
-+#define pmd_offset(pud, address) ((pmd_t *) pud_page(*(pud)) + \
-+ pmd_index(address))
-+
-+/*
+ * 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.
@@ -73810,7 +69288,7 @@
+
+#define pmd_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0)
+
-+static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
++static inline pte_t raw_ptep_get_and_clear(pte_t *ptep)
+{
+ pte_t res;
+
@@ -73821,7 +69299,28 @@
+
+ return res;
+}
++#endif
++
++/*
++ * 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) { }
++
++#define pud_page(pud) \
++((struct page *) __va(pud_val(pud) & PAGE_MASK))
+
++#define pud_page_vaddr(pud) \
++((unsigned long) __va(pud_val(pud) & PAGE_MASK))
++
++
++/* Find an entry in the second-level page table.. */
++#define pmd_offset(pud, address) ((pmd_t *) pud_page(*(pud)) + \
++ pmd_index(address))
++
++#define __HAVE_ARCH_PTE_SAME
+static inline int pte_same(pte_t a, pte_t b)
+{
+ return a.pte_low == b.pte_low && a.pte_high == b.pte_high;
@@ -73872,10 +69371,10 @@
+#define vmalloc_sync_all() ((void)0)
+
+#endif /* _I386_PGTABLE_3LEVEL_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/pgtable.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pgtable.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/pgtable.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/pgtable.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,508 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/pgtable.h b/include/asm-i386/mach-xen/asm/pgtable.h
+--- a/include/asm-i386/mach-xen/asm/pgtable.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/pgtable.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,561 @@
+#ifndef _I386_PGTABLE_H
+#define _I386_PGTABLE_H
+
@@ -73894,6 +69393,7 @@
+#include <asm/processor.h>
+#include <asm/fixmap.h>
+#include <linux/threads.h>
++#include <asm/paravirt.h>
+
+#ifndef _I386_BITOPS_H
+#include <asm/bitops.h>
@@ -73913,14 +69413,14 @@
+#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 struct kmem_cache *pgd_cache;
++extern struct kmem_cache *pmd_cache;
+extern spinlock_t pgd_lock;
+extern struct page *pgd_list;
+
-+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 pmd_ctor(void *, struct kmem_cache *, unsigned long);
++void pgd_ctor(void *, struct kmem_cache *, unsigned long);
++void pgd_dtor(void *, struct kmem_cache *, unsigned long);
+void pgtable_cache_init(void);
+void paging_init(void);
+
@@ -74127,20 +69627,88 @@
+# include <asm/pgtable-2level.h>
+#endif
+
-+static inline int ptep_test_and_clear_dirty(struct vm_area_struct *vma, unsigned long addr, pte_t *ptep)
-+{
-+ if (!pte_dirty(*ptep))
-+ return 0;
-+ return test_and_clear_bit(_PAGE_BIT_DIRTY, &ptep->pte_low);
-+}
++#ifndef CONFIG_PARAVIRT
++/*
++ * Rules for using pte_update - it must be called after any PTE update which
++ * has not been done using the set_pte / clear_pte interfaces. It is used by
++ * shadow mode hypervisors to resynchronize the shadow page tables. Kernel PTE
++ * updates should either be sets, clears, or set_pte_atomic for P->P
++ * transitions, which means this hook should only be called for user PTEs.
++ * This hook implies a P->P protection or access change has taken place, which
++ * requires a subsequent TLB flush. The notification can optionally be delayed
++ * until the TLB flush event by using the pte_update_defer form of the
++ * interface, but care must be taken to assure that the flush happens while
++ * still holding the same page table lock so that the shadow and primary pages
++ * do not become out of sync on SMP.
++ */
++#define pte_update(mm, addr, ptep) do { } while (0)
++#define pte_update_defer(mm, addr, ptep) do { } while (0)
++#endif
+
-+static inline int ptep_test_and_clear_young(struct vm_area_struct *vma, unsigned long addr, pte_t *ptep)
++/*
++ * 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) { \
++ if ( likely((__vma)->vm_mm == current->mm) ) { \
++ BUG_ON(HYPERVISOR_update_va_mapping((__address), (__entry), UVMF_INVLPG|UVMF_MULTI|(unsigned long)((__vma)->vm_mm->cpu_vm_mask.bits))); \
++ pte_update_defer((__vma)->vm_mm, (__address), (__ptep)); \
++ } else { \
++ xen_l1_entry_update((__ptep), (__entry)); \
++ pte_update_defer((__vma)->vm_mm, (__address), (__ptep)); \
++ flush_tlb_page((__vma), (__address)); \
++ } \
++ } \
++ } while (0)
++
++/*
++ * We don't actually have these, but we want to advertise them so that
++ * we can encompass the flush here.
++ */
++#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_DIRTY
++#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
++
++#define __HAVE_ARCH_PTEP_CLEAR_DIRTY_FLUSH
++#define ptep_clear_flush_dirty(vma, address, ptep) \
++({ \
++ int __dirty; \
++ __dirty = pte_dirty(*(ptep)); \
++ if (__dirty) { \
++ clear_bit(_PAGE_BIT_DIRTY, &(ptep)->pte_low); \
++ pte_update_defer((vma)->vm_mm, (address), (ptep)); \
++ flush_tlb_page(vma, address); \
++ } \
++ __dirty; \
++})
++
++#define __HAVE_ARCH_PTEP_CLEAR_YOUNG_FLUSH
++#define ptep_clear_flush_young(vma, address, ptep) \
++({ \
++ int __young; \
++ __young = pte_young(*(ptep)); \
++ if (__young) { \
++ clear_bit(_PAGE_BIT_ACCESSED, &(ptep)->pte_low); \
++ pte_update_defer((vma)->vm_mm, (address), (ptep)); \
++ flush_tlb_page(vma, address); \
++ } \
++ __young; \
++})
++
++#define __HAVE_ARCH_PTEP_GET_AND_CLEAR
++static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
+{
-+ if (!pte_young(*ptep))
-+ return 0;
-+ return test_and_clear_bit(_PAGE_BIT_ACCESSED, &ptep->pte_low);
++ pte_t pte = raw_ptep_get_and_clear(ptep);
++ pte_update(mm, addr, ptep);
++ return pte;
+}
+
++#define __HAVE_ARCH_PTEP_GET_AND_CLEAR_FULL
+static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm, unsigned long addr, pte_t *ptep, int full)
+{
+ pte_t pte;
@@ -74153,10 +69721,12 @@
+ return pte;
+}
+
++#define __HAVE_ARCH_PTEP_SET_WRPROTECT
+static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
+{
+ if (pte_write(*ptep))
+ clear_bit(_PAGE_BIT_RW, &ptep->pte_low);
++ pte_update(mm, addr, ptep);
+}
+
+/*
@@ -74246,11 +69816,11 @@
+#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))
++ ((pte_t *) pmd_page_vaddr(*(dir)) + pte_index(address))
+
+#define pmd_page(pmd) (pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT))
+
-+#define pmd_page_kernel(pmd) \
++#define pmd_page_vaddr(pmd) \
+ ((unsigned long) __va(pmd_val(pmd) & PAGE_MASK))
+
+/*
@@ -74273,8 +69843,6 @@
+ static inline int set_kernel_exec(unsigned long vaddr, int enable) { return 0;}
+#endif
+
-+extern void noexec_setup(const char *str);
-+
+#if defined(CONFIG_HIGHPTE)
+#define pte_offset_map(dir, address) \
+ ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)),KM_PTE0) + \
@@ -74292,29 +69860,19 @@
+#define pte_unmap_nested(pte) do { } while (0)
+#endif
+
++/* Clear a kernel PTE and flush it from the TLB */
++#define kpte_clear_flush(ptep, vaddr) \
++do { \
++ pte_clear(&init_mm, vaddr, ptep); \
++ __flush_tlb_one(vaddr); \
++} 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) { \
-+ if ( likely((__vma)->vm_mm == current->mm) ) { \
-+ BUG_ON(HYPERVISOR_update_va_mapping((__address), (__entry), UVMF_INVLPG|UVMF_MULTI|(unsigned long)((__vma)->vm_mm->cpu_vm_mask.bits))); \
-+ } else { \
-+ xen_l1_entry_update((__ptep), (__entry)); \
-+ flush_tlb_page((__vma), (__address)); \
-+ } \
-+ } \
-+ } while (0)
+
+#define __HAVE_ARCH_PTEP_ESTABLISH
+#define ptep_establish(__vma, __address, __ptep, __entry) \
@@ -74368,26 +69926,20 @@
+ unsigned long address,
+ unsigned long size);
+
-+#define io_remap_pfn_range(vma,from,pfn,size,prot) \
-+direct_remap_pfn_range(vma,from,pfn,size,prot,DOMID_IO)
++#define io_remap_pfn_range(vma,vaddr,pfn,size,prot) \
++ direct_remap_pfn_range(vma,vaddr,pfn,size,prot,DOMID_IO)
+
+#define MK_IOSPACE_PFN(space, pfn) (pfn)
+#define GET_IOSPACE(pfn) 0
+#define GET_PFN(pfn) (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_SET_WRPROTECT
-+#define __HAVE_ARCH_PTE_SAME
+#include <asm-generic/pgtable.h>
+
+#endif /* _I386_PGTABLE_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/processor.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/processor.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/processor.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/processor.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,741 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/processor.h b/include/asm-i386/mach-xen/asm/processor.h
+--- a/include/asm-i386/mach-xen/asm/processor.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/processor.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,759 @@
+/*
+ * include/asm-i386/processor.h
+ *
@@ -74410,6 +69962,7 @@
+#include <linux/threads.h>
+#include <asm/percpu.h>
+#include <linux/cpumask.h>
++#include <linux/init.h>
+#include <xen/interface/physdev.h>
+
+/* flag for disabling the tsc */
@@ -74463,6 +70016,7 @@
+#endif
+ unsigned char x86_max_cores; /* cpuid returned max cores value */
+ unsigned char apicid;
++ unsigned short x86_clflush_size;
+#ifdef CONFIG_SMP
+ unsigned char booted_cores; /* number of cores as seen by OS */
+ __u8 phys_proc_id; /* Physical processor id. */
@@ -74504,6 +70058,8 @@
+extern int cpu_llc_id[NR_CPUS];
+extern char ignore_fpu_irq;
+
++void __init cpu_detect(struct cpuinfo_x86 *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);
@@ -74536,75 +70092,16 @@
+#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));
-+}
-+
-+/* 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)
++static inline fastcall void native_cpuid(unsigned int *eax, unsigned int *ebx,
++ unsigned int *ecx, unsigned int *edx)
+{
++ /* ecx is often an input as well as an output. */
+ __asm__(XEN_CPUID
+ : "=a" (*eax),
+ "=b" (*ebx),
+ "=c" (*ecx),
+ "=d" (*edx)
-+ : "0" (op), "c" (count));
-+}
-+
-+/*
-+ * CPUID functions returning a single datum
-+ */
-+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)
-+{
-+ unsigned int eax, ebx;
-+
-+ __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;
-+
-+ __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;
++ : "0" (*eax), "2" (*ecx));
+}
+
+#define load_cr3(pgdir) write_cr3(__pa(pgdir))
@@ -74705,6 +70202,8 @@
+ : :"a" (eax), "c" (ecx));
+}
+
++extern void mwait_idle_with_hints(unsigned long eax, unsigned long ecx);
++
+/* from system description table in BIOS. Mostly for MCA use, but
+others may find it useful. */
+extern unsigned int machine_id;
@@ -74874,6 +70373,7 @@
+ .vm86_info = NULL, \
+ .sysenter_cs = __KERNEL_CS, \
+ .io_bitmap_ptr = NULL, \
++ .gs = __KERNEL_PDA, \
+}
+
+#ifndef CONFIG_X86_NO_TSS
@@ -74890,25 +70390,11 @@
+ .io_bitmap_base = INVALID_IO_BITMAP_OFFSET, \
+ .io_bitmap = { [ 0 ... IO_BITMAP_LONGS] = ~0 }, \
+}
-+
-+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) \
-+ HYPERVISOR_stack_switch(__KERNEL_DS, (thread)->esp0)
+#endif
+
+#define start_thread(regs, new_eip, new_esp) do { \
-+ __asm__("movl %0,%%fs ; movl %0,%%gs": :"r" (0)); \
++ __asm__("movl %0,%%fs": :"r" (0)); \
++ regs->xgs = 0; \
+ set_fs(USER_DS); \
+ regs->xds = __USER_DS; \
+ regs->xes = __USER_DS; \
@@ -74918,26 +70404,6 @@
+ regs->esp = new_esp; \
+} 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;
@@ -75029,6 +70495,106 @@
+
+#define cpu_relax() rep_nop()
+
++#ifdef CONFIG_PARAVIRT
++#include <asm/paravirt.h>
++#else
++#define paravirt_enabled() 0
++#define __cpuid native_cpuid
++
++#ifndef CONFIG_X86_NO_TSS
++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) \
++ HYPERVISOR_stack_switch(__KERNEL_DS, (thread)->esp0)
++#endif
++
++
++/*
++ * 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))
++
++#define set_iopl_mask native_set_iopl_mask
++#endif /* CONFIG_PARAVIRT */
++
++/*
++ * Set IOPL bits in EFLAGS from given mask
++ */
++static fastcall inline void native_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);
++}
++
++/*
++ * 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)
++{
++ *eax = op;
++ *ecx = 0;
++ __cpuid(eax, ebx, ecx, edx);
++}
++
++/* 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)
++{
++ *eax = op;
++ *ecx = count;
++ __cpuid(eax, ebx, ecx, edx);
++}
++
++/*
++ * CPUID functions returning a single datum
++ */
++static inline unsigned int cpuid_eax(unsigned int op)
++{
++ unsigned int eax, ebx, ecx, edx;
++
++ cpuid(op, &eax, &ebx, &ecx, &edx);
++ return eax;
++}
++static inline unsigned int cpuid_ebx(unsigned int op)
++{
++ unsigned int eax, ebx, ecx, edx;
++
++ cpuid(op, &eax, &ebx, &ecx, &edx);
++ return ebx;
++}
++static inline unsigned int cpuid_ecx(unsigned int op)
++{
++ unsigned int eax, ebx, ecx, edx;
++
++ cpuid(op, &eax, &ebx, &ecx, &edx);
++ return ecx;
++}
++static inline unsigned int cpuid_edx(unsigned int op)
++{
++ unsigned int eax, ebx, ecx, edx;
++
++ cpuid(op, &eax, &ebx, &ecx, &edx);
++ return edx;
++}
++
+/* generic versions from gas */
+#define GENERIC_NOP1 ".byte 0x90\n"
+#define GENERIC_NOP2 ".byte 0x89,0xf6\n"
@@ -75128,104 +70694,14 @@
+extern void enable_sep_cpu(void);
+extern int sysenter_setup(void);
+
-+#endif /* __ASM_I386_PROCESSOR_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/ptrace.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/ptrace.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/ptrace.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/ptrace.h 2006-11-19 14:27:04.000000000 +0100
-@@ -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
-+
-+/* this struct defines the way the registers are stored on the
-+ stack during a system call. */
-+
-+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;
-+};
-+
-+/* 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
-+
-+#define PTRACE_OLDSETOPTIONS 21
-+
-+#define PTRACE_GET_THREAD_AREA 25
-+#define PTRACE_SET_THREAD_AREA 26
-+
-+#define PTRACE_SYSEMU 31
-+#define PTRACE_SYSEMU_SINGLESTEP 32
-+
-+#ifdef __KERNEL__
-+
-+#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 (regs->xcs & 2) != 0;
-+}
-+static inline int user_mode_vm(struct pt_regs *regs)
-+{
-+ 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__ */
++extern int init_gdt(int cpu, struct task_struct *idle);
++extern void cpu_set_gdt(int);
++extern void secondary_cpu_init(void);
+
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/scatterlist.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/scatterlist.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/scatterlist.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/scatterlist.h 2006-11-19 14:27:04.000000000 +0100
++#endif /* __ASM_I386_PROCESSOR_H */
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/scatterlist.h b/include/asm-i386/mach-xen/asm/scatterlist.h
+--- a/include/asm-i386/mach-xen/asm/scatterlist.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/scatterlist.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,22 @@
+#ifndef _I386_SCATTERLIST_H
+#define _I386_SCATTERLIST_H
@@ -75249,216 +70725,10 @@
+#define ISA_DMA_THRESHOLD (0x00ffffff)
+
+#endif /* !(_I386_SCATTERLIST_H) */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/segment.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/segment.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/segment.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/segment.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,117 @@
-+#ifndef _ASM_SEGMENT_H
-+#define _ASM_SEGMENT_H
-+
-+/*
-+ * 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 TLS_SIZE (GDT_ENTRY_TLS_ENTRIES * 8)
-+
-+#define GDT_ENTRY_DEFAULT_USER_CS 14
-+#define __USER_CS (GDT_ENTRY_DEFAULT_USER_CS * 8 + 3)
-+
-+#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)
-+
-+#define GDT_ENTRY_DOUBLEFAULT_TSS 31
-+
-+/*
-+ * The GDT has 32 entries
-+ */
-+#define GDT_ENTRIES 32
-+
-+#define GDT_SIZE (GDT_ENTRIES * 8)
-+
-+/* Simple and small GDT entries for booting only */
-+
-+#define GDT_ENTRY_BOOT_CS 2
-+#define __BOOT_CS (GDT_ENTRY_BOOT_CS * 8)
-+
-+#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 */
-+
-+/*
-+ * 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
-+
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/setup.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/setup.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/setup.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/setup.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,81 @@
-+/*
-+ * Just a place holder. We don't want to have to test x86 before
-+ * we include stuff
-+ */
-+
-+#ifndef _i386_SETUP_H
-+#define _i386_SETUP_H
-+
-+#ifdef __KERNEL__
-+#include <linux/pfn.h>
-+
-+/*
-+ * Reserved space for vmalloc and iomap - defined in asm/page.h
-+ */
-+#define MAXMEM_PFN PFN_DOWN(MAXMEM)
-+#define MAX_NONPAE_PFN (1 << 20)
-+#endif
-+
-+#define PARAM_SIZE 4096
-+#define COMMAND_LINE_SIZE 256
-+
-+#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 */
-+
-+#ifndef __ASSEMBLY__
-+/*
-+ * This is set up by the setup-routine at boot-time
-+ */
-+extern unsigned char boot_params[PARAM_SIZE];
-+
-+#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))
-+
-+/*
-+ * Do NOT EVER look at the BIOS memory size location.
-+ * It does not work on many machines.
-+ */
-+#define LOWMEMSIZE() (0x9f000)
-+
-+struct e820entry;
-+
-+char * __init machine_specific_memory_setup(void);
-+
-+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);
-+
-+#endif /* __ASSEMBLY__ */
-+
-+#endif /* _i386_SETUP_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/smp.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/smp.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/smp.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/smp.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,103 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/smp.h b/include/asm-i386/mach-xen/asm/smp.h
+--- a/include/asm-i386/mach-xen/asm/smp.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/smp.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,115 @@
+#ifndef __ASM_SMP_H
+#define __ASM_SMP_H
+
@@ -75469,6 +70739,7 @@
+#include <linux/kernel.h>
+#include <linux/threads.h>
+#include <linux/cpumask.h>
++#include <asm/pda.h>
+#endif
+
+#ifdef CONFIG_X86_LOCAL_APIC
@@ -75517,7 +70788,7 @@
+ * 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)
++#define raw_smp_processor_id() (read_pda(cpu_number))
+
+extern cpumask_t cpu_possible_map;
+#define cpu_callin_map cpu_possible_map
@@ -75540,237 +70811,42 @@
+ return GET_APIC_ID(*(unsigned long *)(APIC_BASE+APIC_ID));
+}
+#endif
-+
-+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
+
++extern int safe_smp_processor_id(void);
+extern int __cpu_disable(void);
+extern void __cpu_die(unsigned int cpu);
++extern unsigned int num_processors;
+extern void prefill_possible_map(void);
++
+#endif /* !__ASSEMBLY__ */
+
+#else /* CONFIG_SMP */
+
++#define safe_smp_processor_id() 0
+#define cpu_physical_id(cpu) boot_cpu_physical_apicid
+
+#define NO_PROC_ID 0xFF /* No processor magic marker */
+
+#endif
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/spinlock.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/spinlock.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/spinlock.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/spinlock.h 2006-12-05 18:42:37.000000000 +0100
-@@ -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>
-+
-+/*
-+ * 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)
-+ */
-+
-+#define __raw_spin_is_locked(x) \
-+ (*(volatile signed char *)(&(x)->slock) <= 0)
-+
-+#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"
-+
-+/*
-+ * 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 __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");
-+}
-+
-+/*
-+ * 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 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;
-+}
-+
-+/*
-+ * __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)
-+ */
-+
-+#if !defined(CONFIG_X86_OOSTORE) && !defined(CONFIG_X86_PPRO_FENCE)
-+
-+#define __raw_spin_unlock_string \
-+ "movb $1,%0" \
-+ :"+m" (lock->slock) : : "memory"
-+
-+
-+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
-+{
-+ __asm__ __volatile__(
-+ __raw_spin_unlock_string
-+ );
-+}
+
-+#else
++#ifndef __ASSEMBLY__
+
-+#define __raw_spin_unlock_string \
-+ "xchgb %b0, %1" \
-+ :"=q" (oldval), "+m" (lock->slock) \
-+ :"0" (oldval) : "memory"
++extern u8 apicid_2_node[];
+
-+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
++#ifdef CONFIG_X86_LOCAL_APIC
++static __inline int logical_smp_processor_id(void)
+{
-+ char oldval = 1;
-+
-+ __asm__ __volatile__(
-+ __raw_spin_unlock_string
-+ );
++ /* 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 __raw_spin_unlock_wait(lock) \
-+ do { while (__raw_spin_is_locked(lock)) cpu_relax(); } while (0)
-+
-+/*
-+ * 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
-+ */
-+
-+/**
-+ * read_can_lock - would read_trylock() succeed?
-+ * @lock: the rwlock in question.
-+ */
-+#define __raw_read_can_lock(x) ((int)(x)->lock > 0)
-+
-+/**
-+ * 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");
-+}
-+
-+static inline void __raw_write_lock(raw_rwlock_t *rw)
-+{
-+ __build_write_lock(rw, "__write_lock_failed");
-+}
-+
-+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;
-+}
-+
-+static inline int __raw_write_trylock(raw_rwlock_t *lock)
-+{
-+ 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 __raw_read_unlock(raw_rwlock_t *rw)
-+{
-+ asm volatile(LOCK_PREFIX "incl %0" :"+m" (rw->lock) : : "memory");
-+}
-+
-+static inline void __raw_write_unlock(raw_rwlock_t *rw)
-+{
-+ asm volatile(LOCK_PREFIX "addl $" RW_LOCK_BIAS_STR ", %0"
-+ : "+m" (rw->lock) : : "memory");
-+}
-+
-+#endif /* __ASM_SPINLOCK_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/swiotlb.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/swiotlb.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/swiotlb.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/swiotlb.h 2006-11-19 14:27:04.000000000 +0100
++#endif
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/swiotlb.h b/include/asm-i386/mach-xen/asm/swiotlb.h
+--- a/include/asm-i386/mach-xen/asm/swiotlb.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/swiotlb.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,41 @@
+#ifndef _ASM_SWIOTLB_H
+#define _ASM_SWIOTLB_H 1
@@ -75813,9 +70889,9 @@
+#endif
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/synch_bitops.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/synch_bitops.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/synch_bitops.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/synch_bitops.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/synch_bitops.h b/include/asm-i386/mach-xen/asm/synch_bitops.h
+--- a/include/asm-i386/mach-xen/asm/synch_bitops.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/synch_bitops.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,141 @@
+#ifndef __XEN_SYNCH_BITOPS_H__
+#define __XEN_SYNCH_BITOPS_H__
@@ -75958,10 +71034,10 @@
+#define synch_cmpxchg_subword synch_cmpxchg
+
+#endif /* __XEN_SYNCH_BITOPS_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/system.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/system.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/system.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/system.h 2006-12-05 18:42:37.000000000 +0100
-@@ -0,0 +1,495 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/system.h b/include/asm-i386/mach-xen/asm/system.h
+--- a/include/asm-i386/mach-xen/asm/system.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/system.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,535 @@
+#ifndef __ASM_SYSTEM_H
+#define __ASM_SYSTEM_H
+
@@ -76060,6 +71136,9 @@
+#define savesegment(seg, value) \
+ asm volatile("mov %%" #seg ",%0":"=rm" (value))
+
++#ifdef CONFIG_PARAVIRT
++#include <asm/paravirt.h>
++#else
+#define read_cr0() ({ \
+ unsigned int __dummy; \
+ __asm__ __volatile__( \
@@ -76107,20 +71186,21 @@
+ : "=r" (__dummy): "0" (0)); \
+ __dummy; \
+})
-+
+#define write_cr4(x) \
+ __asm__ __volatile__("movl %0,%%cr4": :"r" (x))
-+/*
-+ * Clear and set 'TS' bit respectively
-+ */
++
++#define wbinvd() \
++ __asm__ __volatile__ ("wbinvd": : :"memory")
++
++/* Clear the 'TS' bit */
+#define clts() (HYPERVISOR_fpu_taskswitch(0))
++#endif/* CONFIG_PARAVIRT */
++
++/* Set the 'TS' bit */
+#define stts() (HYPERVISOR_fpu_taskswitch(1))
+
+#endif /* __KERNEL__ */
+
-+#define wbinvd() \
-+ __asm__ __volatile__ ("wbinvd": : :"memory")
-+
+static inline unsigned long get_limit(unsigned long segment)
+{
+ unsigned long __limit;
@@ -76238,6 +71318,9 @@
+#define cmpxchg(ptr,o,n)\
+ ((__typeof__(*(ptr)))__cmpxchg((ptr),(unsigned long)(o),\
+ (unsigned long)(n),sizeof(*(ptr))))
++#define sync_cmpxchg(ptr,o,n)\
++ ((__typeof__(*(ptr)))__sync_cmpxchg((ptr),(unsigned long)(o),\
++ (unsigned long)(n),sizeof(*(ptr))))
+#endif
+
+static inline unsigned long __cmpxchg(volatile void *ptr, unsigned long old,
@@ -76267,6 +71350,39 @@
+ return old;
+}
+
++/*
++ * Always use locked operations when touching memory shared with a
++ * hypervisor, since the system may be SMP even if the guest kernel
++ * isn't.
++ */
++static inline unsigned long __sync_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"(*__xg(ptr)), "0"(old)
++ : "memory");
++ return prev;
++ case 2:
++ __asm__ __volatile__("lock; cmpxchgw %w1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__xg(ptr)), "0"(old)
++ : "memory");
++ return prev;
++ case 4:
++ __asm__ __volatile__("lock; cmpxchgl %1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__xg(ptr)), "0"(old)
++ : "memory");
++ return prev;
++ }
++ return old;
++}
++
+#ifndef CONFIG_X86_CMPXCHG
+/*
+ * Building a kernel capable running on 80386. It may be necessary to
@@ -76457,10 +71573,10 @@
+void default_idle(void);
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/timer.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/timer.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/timer.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/timer.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,70 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/timer.h b/include/asm-i386/mach-xen/asm/timer.h
+--- a/include/asm-i386/mach-xen/asm/timer.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/timer.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,67 @@
+#ifndef _ASMi386_TIMER_H
+#define _ASMi386_TIMER_H
+#include <linux/init.h>
@@ -76501,11 +71617,8 @@
+extern struct timer_opts* __init select_timer(void);
+extern void clock_fallback(void);
+void setup_pit_timer(void);
-+
+/* Modifiers for buggy PIT handling */
-+
+extern int pit_latch_buggy;
-+
+extern struct timer_opts *cur_timer;
+extern int timer_ack;
+
@@ -76531,26 +71644,31 @@
+extern struct init_timer_opts timer_pmtmr_init;
+#endif
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/tlbflush.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/tlbflush.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/tlbflush.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/tlbflush.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,101 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/asm/tlbflush.h b/include/asm-i386/mach-xen/asm/tlbflush.h
+--- a/include/asm-i386/mach-xen/asm/tlbflush.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/asm/tlbflush.h 2007-03-14 10:55:43.000000000 +0100
+@@ -0,0 +1,106 @@
+#ifndef _I386_TLBFLUSH_H
+#define _I386_TLBFLUSH_H
+
+#include <linux/mm.h>
+#include <asm/processor.h>
+
-+#define __flush_tlb() xen_tlb_flush()
-+#define __flush_tlb_global() xen_tlb_flush()
-+#define __flush_tlb_all() xen_tlb_flush()
++#ifdef CONFIG_PARAVIRT
++#include <asm/paravirt.h>
++#else
++#define __flush_tlb() __native_flush_tlb()
++#define __flush_tlb_global() __native_flush_tlb_global()
++#define __flush_tlb_single(addr) __native_flush_tlb_single(addr)
++#endif
+
-+extern unsigned long pgkern_mask;
++#define __native_flush_tlb() xen_tlb_flush()
++#define __native_flush_tlb_global() xen_tlb_flush()
++#define __native_flush_tlb_single(addr) xen_invlpg(addr)
++#define __flush_tlb_all() xen_tlb_flush()
+
+#define cpu_has_invlpg (boot_cpu_data.x86 > 3)
+
-+#define __flush_tlb_single(addr) xen_invlpg(addr)
-+
+#define __flush_tlb_one(addr) __flush_tlb_single(addr)
+
+/*
@@ -76636,33 +71754,9 @@
+}
+
+#endif /* _I386_TLBFLUSH_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/asm/vga.h linux-2.6.18-xen/include/asm-i386/mach-xen/asm/vga.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/asm/vga.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/asm/vga.h 2006-11-19 14:27:04.000000000 +0100
-@@ -0,0 +1,20 @@
-+/*
-+ * Access to VGA videoram
-+ *
-+ * (c) 1998 Martin Mares <mj at ucw.cz>
-+ */
-+
-+#ifndef _LINUX_ASM_VGA_H_
-+#define _LINUX_ASM_VGA_H_
-+
-+/*
-+ * On the PC, we can just recalculate addresses and then
-+ * access the videoram directly without any black magic.
-+ */
-+
-+#define VGA_MAP_MEM(x,s) (unsigned long)isa_bus_to_virt(x)
-+
-+#define vga_readb(x) (*(x))
-+#define vga_writeb(x,y) (*(y) = (x))
-+
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/irq_vectors.h linux-2.6.18-xen/include/asm-i386/mach-xen/irq_vectors.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/irq_vectors.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/irq_vectors.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/irq_vectors.h b/include/asm-i386/mach-xen/irq_vectors.h
+--- a/include/asm-i386/mach-xen/irq_vectors.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/irq_vectors.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,125 @@
+/*
+ * This file should contain #defines for all of the interrupt vector
@@ -76789,9 +71883,9 @@
+#define irq_to_dynirq(_x) ((_x) - DYNIRQ_BASE)
+
+#endif /* _ASM_IRQ_VECTORS_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/mach_traps.h linux-2.6.18-xen/include/asm-i386/mach-xen/mach_traps.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/mach_traps.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/mach_traps.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/mach_traps.h b/include/asm-i386/mach-xen/mach_traps.h
+--- a/include/asm-i386/mach-xen/mach_traps.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/mach_traps.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,33 @@
+/*
+ * include/asm-xen/asm-i386/mach-xen/mach_traps.h
@@ -76826,18 +71920,104 @@
+static inline void reassert_nmi(void) {}
+
+#endif /* !_MACH_TRAPS_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/mach-xen/setup_arch.h linux-2.6.18-xen/include/asm-i386/mach-xen/setup_arch.h
---- linux-2.6.18.3/include/asm-i386/mach-xen/setup_arch.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-i386/mach-xen/setup_arch.h 2006-12-05 18:42:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mach-xen/setup_arch.h b/include/asm-i386/mach-xen/setup_arch.h
+--- a/include/asm-i386/mach-xen/setup_arch.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-i386/mach-xen/setup_arch.h 2007-03-14 10:55:43.000000000 +0100
@@ -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 -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/pgtable-2level-defs.h linux-2.6.18-xen/include/asm-i386/pgtable-2level-defs.h
---- linux-2.6.18.3/include/asm-i386/pgtable-2level-defs.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-i386/pgtable-2level-defs.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/mmu.h b/include/asm-i386/mmu.h
+--- a/include/asm-i386/mmu.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-i386/mmu.h 2007-03-14 10:55:43.000000000 +0100
+@@ -13,6 +13,19 @@
+ struct semaphore sem;
+ void *ldt;
+ void *vdso;
++#ifdef CONFIG_XEN
++ int has_foreign_mappings;
++#endif
+ } mm_context_t;
+
++#ifdef CONFIG_XEN
++/* mm/memory.c:exit_mmap hook */
++extern void _arch_exit_mmap(struct mm_struct *mm);
++#define arch_exit_mmap(_mm) _arch_exit_mmap(_mm)
++
++/* 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 /* CONFIG_XEN */
++
+ #endif
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/pci.h b/include/asm-i386/pci.h
+--- a/include/asm-i386/pci.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-i386/pci.h 2007-03-14 10:55:43.000000000 +0100
+@@ -42,6 +42,27 @@
+
+ struct pci_dev;
+
++#ifdef CONFIG_SWIOTLB
++
++
++/* On Xen we use SWIOTLB instead of blk-specific bounce buffers. */
++#define PCI_DMA_BUS_IS_PHYS (0)
++
++#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
++
+ /* The PCI address space does equal the physical memory
+ * address space. The networking and block device layers use
+ * this boolean for bounce buffer decisions.
+@@ -56,6 +77,8 @@
+ #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)
+
+@@ -110,10 +133,22 @@
+
+ #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>
+
++#ifdef CONFIG_XEN
++/* On Xen we have to scan all functions since Xen hides bridges from
++ * us. If a bridge is at fn=0 and that slot has a multifunction
++ * device, we won't find the additional devices without scanning all
++ * functions. */
++#undef pcibios_scan_all_fns
++#define pcibios_scan_all_fns(a, b) 1
++#endif /* CONFIG_XEN */
+ #endif /* __i386_PCI_H */
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/pgtable-2level-defs.h b/include/asm-i386/pgtable-2level-defs.h
+--- a/include/asm-i386/pgtable-2level-defs.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-i386/pgtable-2level-defs.h 2007-03-14 10:55:43.000000000 +0100
@@ -1,6 +1,8 @@
#ifndef _I386_PGTABLE_2LEVEL_DEFS_H
#define _I386_PGTABLE_2LEVEL_DEFS_H
@@ -76847,21 +72027,85 @@
/*
* traditional i386 two-level paging structure:
*/
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-i386/pgtable-3level-defs.h linux-2.6.18-xen/include/asm-i386/pgtable-3level-defs.h
---- linux-2.6.18.3/include/asm-i386/pgtable-3level-defs.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-i386/pgtable-3level-defs.h 2006-11-19 14:27:04.000000000 +0100
-@@ -1,6 +1,8 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/pgtable-3level-defs.h b/include/asm-i386/pgtable-3level-defs.h
+--- a/include/asm-i386/pgtable-3level-defs.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-i386/pgtable-3level-defs.h 2007-03-14 10:55:43.000000000 +0100
+@@ -1,6 +1,12 @@
#ifndef _I386_PGTABLE_3LEVEL_DEFS_H
#define _I386_PGTABLE_3LEVEL_DEFS_H
++#ifdef CONFIG_XEN
++#define HAVE_SHARED_KERNEL_PMD 0
++#else
+#define HAVE_SHARED_KERNEL_PMD 1
++#endif
+
/*
* PGDIR_SHIFT determines what a top-level page table entry can map
*/
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/agp.h linux-2.6.18-xen/include/asm-ia64/agp.h
---- linux-2.6.18.3/include/asm-ia64/agp.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/agp.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/segment.h b/include/asm-i386/segment.h
+--- a/include/asm-i386/segment.h 2007-03-16 18:49:38.000000000 +0100
++++ b/include/asm-i386/segment.h 2007-03-14 10:55:43.000000000 +0100
+@@ -87,7 +87,11 @@
+ #define GDT_SIZE (GDT_ENTRIES * 8)
+
+ /* Matches __KERNEL_CS and __USER_CS (they must be 2 entries apart) */
++#ifdef CONFIG_XEN
++#define SEGMENT_IS_FLAT_CODE(x) ((x) == __USER_CS || (x) == (__KERNEL_CS | get_kernel_rpl()))
++#else
+ #define SEGMENT_IS_FLAT_CODE(x) (((x) & 0xec) == GDT_ENTRY_KERNEL_CS * 8)
++#endif
+ /* Matches PNP_CS32 and PNP_CS16 (they must be consecutive) */
+ #define SEGMENT_IS_PNP_CODE(x) (((x) & 0xf4) == GDT_ENTRY_PNPBIOS_BASE * 8)
+
+@@ -132,6 +136,10 @@
+ #define SEGMENT_GDT 0x0
+
+ #ifndef CONFIG_PARAVIRT
++#ifdef CONFIG_XEN
++#define get_kernel_rpl() (xen_feature(XENFEAT_supervisor_mode_kernel)?0:1)
++#else
+ #define get_kernel_rpl() 0
++#endif /* CONFIG_XEN */
+ #endif
+ #endif
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/setup.h b/include/asm-i386/setup.h
+--- a/include/asm-i386/setup.h 2007-03-16 18:49:38.000000000 +0100
++++ b/include/asm-i386/setup.h 2007-03-14 10:55:43.000000000 +0100
+@@ -53,9 +53,15 @@
+ #define AUX_DEVICE_INFO (*(unsigned char *) (PARAM+0x1FF))
+ #define LOADER_TYPE (*(unsigned char *) (PARAM+0x210))
+ #define KERNEL_START (*(unsigned long *) (PARAM+0x214))
++#ifdef CONFIG_XEN
++#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))
++#else
+ #define INITRD_START (*(unsigned long *) (PARAM+0x218))
+ #define INITRD_SIZE (*(unsigned long *) (PARAM+0x21c))
+ #define EDID_INFO (*(struct edid_info *) (PARAM+0x140))
++#endif /* CONFIG_XEN */
+ #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))
+diff -x '.hg*' -x '.git*' -urN a/include/asm-i386/vga.h b/include/asm-i386/vga.h
+--- a/include/asm-i386/vga.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-i386/vga.h 2007-03-14 10:55:43.000000000 +0100
+@@ -12,7 +12,11 @@
+ * access the videoram directly without any black magic.
+ */
+
++#ifdef CONFIG_XEN
++#define VGA_MAP_MEM(x,s) (unsigned long)isa_bus_to_virt(x)
++#else
+ #define VGA_MAP_MEM(x,s) (unsigned long)phys_to_virt(x)
++#endif
+
+ #define vga_readb(x) (*(x))
+ #define vga_writeb(x,y) (*(y) = (x))
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/agp.h b/include/asm-ia64/agp.h
+--- a/include/asm-ia64/agp.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-ia64/agp.h 2007-03-14 10:55:43.000000000 +0100
@@ -19,13 +19,44 @@
#define flush_agp_cache() mb()
@@ -76907,9 +72151,9 @@
+#endif /* CONFIG_XEN */
#endif /* _ASM_IA64_AGP_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/dma-mapping.h linux-2.6.18-xen/include/asm-ia64/dma-mapping.h
---- linux-2.6.18.3/include/asm-ia64/dma-mapping.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/dma-mapping.h 2006-12-05 18:42:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/dma-mapping.h b/include/asm-ia64/dma-mapping.h
+--- a/include/asm-ia64/dma-mapping.h 2007-03-16 18:49:38.000000000 +0100
++++ b/include/asm-ia64/dma-mapping.h 2007-03-14 10:55:43.000000000 +0100
@@ -6,20 +6,67 @@
* David Mosberger-Tang <davidm at hpl.hp.com>
*/
@@ -77000,9 +72244,9 @@
static inline int
dma_set_mask (struct device *dev, u64 mask)
-@@ -61,4 +110,29 @@
+@@ -62,4 +111,29 @@
- #define dma_is_consistent(dma_handle) (1) /* all we do is coherent memory... */
+ #define dma_is_consistent(d, h) (1) /* all we do is coherent memory... */
+#ifdef CONFIG_XEN
+/* arch/i386/kernel/swiotlb.o requires */
@@ -77030,15 +72274,15 @@
+}
+
#endif /* _ASM_IA64_DMA_MAPPING_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/fixmap.h linux-2.6.18-xen/include/asm-ia64/fixmap.h
---- linux-2.6.18.3/include/asm-ia64/fixmap.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-ia64/fixmap.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/fixmap.h b/include/asm-ia64/fixmap.h
+--- a/include/asm-ia64/fixmap.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-ia64/fixmap.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,2 @@
+#define clear_fixmap(x) do {} while (0)
+#define set_fixmap(x,y) do {} while (0)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/gcc_intrin.h linux-2.6.18-xen/include/asm-ia64/gcc_intrin.h
---- linux-2.6.18.3/include/asm-ia64/gcc_intrin.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/gcc_intrin.h 2006-11-19 14:27:04.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/gcc_intrin.h b/include/asm-ia64/gcc_intrin.h
+--- a/include/asm-ia64/gcc_intrin.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-ia64/gcc_intrin.h 2007-03-14 10:55:43.000000000 +0100
@@ -26,7 +26,7 @@
register unsigned long ia64_r13 asm ("r13") __attribute_used__;
@@ -77228,9 +72472,9 @@
+#define __ia64_get_psr_i() (__ia64_getreg(_IA64_REG_PSR) & 0x4000UL)
+
#endif /* _ASM_IA64_GCC_INTRIN_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/hw_irq.h linux-2.6.18-xen/include/asm-ia64/hw_irq.h
---- linux-2.6.18.3/include/asm-ia64/hw_irq.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/hw_irq.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/hw_irq.h b/include/asm-ia64/hw_irq.h
+--- a/include/asm-ia64/hw_irq.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-ia64/hw_irq.h 2007-03-14 10:55:43.000000000 +0100
@@ -15,7 +15,11 @@
#include <asm/ptrace.h>
#include <asm/smp.h>
@@ -77256,9 +72500,9 @@
platform_send_ipi(smp_processor_id(), vector, IA64_IPI_DM_INT, 0);
}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/hypercall.h linux-2.6.18-xen/include/asm-ia64/hypercall.h
---- linux-2.6.18.3/include/asm-ia64/hypercall.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-ia64/hypercall.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/hypercall.h b/include/asm-ia64/hypercall.h
+--- a/include/asm-ia64/hypercall.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-ia64/hypercall.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,423 @@
+/******************************************************************************
+ * hypercall.h
@@ -77683,9 +72927,9 @@
+#define HYPERVISOR_update_va_mapping(va, new_val, flags) (0)
+
+#endif /* __HYPERCALL_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/hypervisor.h linux-2.6.18-xen/include/asm-ia64/hypervisor.h
---- linux-2.6.18.3/include/asm-ia64/hypervisor.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-ia64/hypervisor.h 2006-12-05 18:42:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/hypervisor.h b/include/asm-ia64/hypervisor.h
+--- a/include/asm-ia64/hypervisor.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-ia64/hypervisor.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,219 @@
+/******************************************************************************
+ * hypervisor.h
@@ -77906,9 +73150,9 @@
+#endif
+
+#endif /* __HYPERVISOR_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/intel_intrin.h linux-2.6.18-xen/include/asm-ia64/intel_intrin.h
---- linux-2.6.18.3/include/asm-ia64/intel_intrin.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/intel_intrin.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/intel_intrin.h b/include/asm-ia64/intel_intrin.h
+--- a/include/asm-ia64/intel_intrin.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-ia64/intel_intrin.h 2007-03-14 10:55:43.000000000 +0100
@@ -16,8 +16,10 @@
* intrinsic
*/
@@ -78032,9 +73276,9 @@
#define __builtin_trap() __break(0);
#endif /* _ASM_IA64_INTEL_INTRIN_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/io.h linux-2.6.18-xen/include/asm-ia64/io.h
---- linux-2.6.18.3/include/asm-ia64/io.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/io.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/io.h b/include/asm-ia64/io.h
+--- a/include/asm-ia64/io.h 2007-03-16 18:49:38.000000000 +0100
++++ b/include/asm-ia64/io.h 2007-03-14 10:55:43.000000000 +0100
@@ -66,9 +66,11 @@
#define PIO_RESERVED __IA64_UNCACHED_OFFSET
#define HAVE_ARCH_PIO_SIZE
@@ -78089,9 +73333,9 @@
# endif /* KERNEL */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/iosapic.h linux-2.6.18-xen/include/asm-ia64/iosapic.h
---- linux-2.6.18.3/include/asm-ia64/iosapic.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/iosapic.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/iosapic.h b/include/asm-ia64/iosapic.h
+--- a/include/asm-ia64/iosapic.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-ia64/iosapic.h 2007-03-14 10:55:43.000000000 +0100
@@ -53,6 +53,7 @@
#define NR_IOSAPICS 256
@@ -78108,9 +73352,9 @@
static inline void iosapic_eoi(char __iomem *iosapic, u32 vector)
{
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/irq.h linux-2.6.18-xen/include/asm-ia64/irq.h
---- linux-2.6.18.3/include/asm-ia64/irq.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/irq.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/irq.h b/include/asm-ia64/irq.h
+--- a/include/asm-ia64/irq.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-ia64/irq.h 2007-03-14 10:55:43.000000000 +0100
@@ -11,8 +11,39 @@
* 02/29/00 D.Mosberger moved most things into hw_irq.h
*/
@@ -78151,9 +73395,25 @@
static __inline__ int
irq_canonicalize (int irq)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/machvec_dig.h linux-2.6.18-xen/include/asm-ia64/machvec_dig.h
---- linux-2.6.18.3/include/asm-ia64/machvec_dig.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/machvec_dig.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/libata-portmap.h b/include/asm-ia64/libata-portmap.h
+--- a/include/asm-ia64/libata-portmap.h 2007-03-16 18:49:38.000000000 +0100
++++ b/include/asm-ia64/libata-portmap.h 1970-01-01 01:00:00.000000000 +0100
+@@ -1,12 +0,0 @@
+-#ifndef __ASM_IA64_LIBATA_PORTMAP_H
+-#define __ASM_IA64_LIBATA_PORTMAP_H
+-
+-#define ATA_PRIMARY_CMD 0x1F0
+-#define ATA_PRIMARY_CTL 0x3F6
+-#define ATA_PRIMARY_IRQ(dev) isa_irq_to_vector(14)
+-
+-#define ATA_SECONDARY_CMD 0x170
+-#define ATA_SECONDARY_CTL 0x376
+-#define ATA_SECONDARY_IRQ(dev) isa_irq_to_vector(15)
+-
+-#endif
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/machvec_dig.h b/include/asm-ia64/machvec_dig.h
+--- a/include/asm-ia64/machvec_dig.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-ia64/machvec_dig.h 2007-03-14 10:55:43.000000000 +0100
@@ -13,4 +13,19 @@
#define platform_name "dig"
#define platform_setup dig_setup
@@ -78174,9 +73434,9 @@
+#endif
+
#endif /* _ASM_IA64_MACHVEC_DIG_h */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/maddr.h linux-2.6.18-xen/include/asm-ia64/maddr.h
---- linux-2.6.18.3/include/asm-ia64/maddr.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-ia64/maddr.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/maddr.h b/include/asm-ia64/maddr.h
+--- a/include/asm-ia64/maddr.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-ia64/maddr.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,92 @@
+#ifndef _ASM_IA64_MADDR_H
+#define _ASM_IA64_MADDR_H
@@ -78270,29 +73530,29 @@
+typedef unsigned long maddr_t; // to compile netback, netfront
+
+#endif /* _ASM_IA64_MADDR_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/meminit.h linux-2.6.18-xen/include/asm-ia64/meminit.h
---- linux-2.6.18.3/include/asm-ia64/meminit.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/meminit.h 2006-11-19 14:27:05.000000000 +0100
-@@ -16,10 +16,15 @@
- * - command line string
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/meminit.h b/include/asm-ia64/meminit.h
+--- a/include/asm-ia64/meminit.h 2007-03-16 18:49:38.000000000 +0100
++++ b/include/asm-ia64/meminit.h 2007-03-14 10:55:43.000000000 +0100
+@@ -17,10 +17,15 @@
* - 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
*/
+#ifndef CONFIG_XEN
- #define IA64_MAX_RSVD_REGIONS 6
+ #define IA64_MAX_RSVD_REGIONS 7
+#else
-+#define IA64_MAX_RSVD_REGIONS 7
++#define IA64_MAX_RSVD_REGIONS 8
+#endif
struct rsvd_region {
unsigned long start; /* virtual address of beginning of element */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/page.h linux-2.6.18-xen/include/asm-ia64/page.h
---- linux-2.6.18.3/include/asm-ia64/page.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/page.h 2006-11-19 14:27:05.000000000 +0100
-@@ -126,7 +126,9 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/page.h b/include/asm-ia64/page.h
+--- a/include/asm-ia64/page.h 2007-03-16 18:49:38.000000000 +0100
++++ b/include/asm-ia64/page.h 2007-03-14 10:55:43.000000000 +0100
+@@ -125,7 +125,9 @@
# define pfn_valid(pfn) (((pfn) >= min_low_pfn) && ((pfn) < max_low_pfn) && ia64_pfn_valid(pfn))
#endif
@@ -78302,7 +73562,7 @@
#define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT)
#define pfn_to_kaddr(pfn) __va((pfn) << PAGE_SHIFT)
-@@ -227,5 +229,53 @@
+@@ -226,5 +228,53 @@
(((current->personality & READ_IMPLIES_EXEC) != 0) \
? VM_EXEC : 0))
@@ -78357,10 +73617,10 @@
+#endif /* __ASSEMBLY__ */
+#endif /* __KERNEL__ */
#endif /* _ASM_IA64_PAGE_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/pal.h linux-2.6.18-xen/include/asm-ia64/pal.h
---- linux-2.6.18.3/include/asm-ia64/pal.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/pal.h 2006-11-19 14:27:05.000000000 +0100
-@@ -82,6 +82,7 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/pal.h b/include/asm-ia64/pal.h
+--- a/include/asm-ia64/pal.h 2007-03-16 18:49:39.000000000 +0100
++++ b/include/asm-ia64/pal.h 2007-03-14 10:55:43.000000000 +0100
+@@ -92,6 +92,7 @@
#ifndef __ASSEMBLY__
#include <linux/types.h>
@@ -78368,9 +73628,9 @@
#include <asm/fpu.h>
/*
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/pgalloc.h linux-2.6.18-xen/include/asm-ia64/pgalloc.h
---- linux-2.6.18.3/include/asm-ia64/pgalloc.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/pgalloc.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/pgalloc.h b/include/asm-ia64/pgalloc.h
+--- a/include/asm-ia64/pgalloc.h 2007-03-16 18:49:39.000000000 +0100
++++ b/include/asm-ia64/pgalloc.h 2007-03-14 10:55:43.000000000 +0100
@@ -125,7 +125,11 @@
static inline void
pmd_populate(struct mm_struct *mm, pmd_t * pmd_entry, struct page *pte)
@@ -78383,9 +73643,9 @@
}
static inline void
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/privop.h linux-2.6.18-xen/include/asm-ia64/privop.h
---- linux-2.6.18.3/include/asm-ia64/privop.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-ia64/privop.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/privop.h b/include/asm-ia64/privop.h
+--- a/include/asm-ia64/privop.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-ia64/privop.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,59 @@
+#ifndef _ASM_IA64_PRIVOP_H
+#define _ASM_IA64_PRIVOP_H
@@ -78446,9 +73706,9 @@
+#endif /* !__ASSEMBLY */
+
+#endif /* _ASM_IA64_PRIVOP_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/processor.h linux-2.6.18-xen/include/asm-ia64/processor.h
---- linux-2.6.18.3/include/asm-ia64/processor.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/processor.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/processor.h b/include/asm-ia64/processor.h
+--- a/include/asm-ia64/processor.h 2007-03-16 18:49:39.000000000 +0100
++++ b/include/asm-ia64/processor.h 2007-03-14 10:55:43.000000000 +0100
@@ -18,6 +18,7 @@
#include <asm/kregs.h>
#include <asm/ptrace.h>
@@ -78456,10 +73716,10 @@
+#include <asm/privop.h>
#define IA64_NUM_DBG_REGS 8
- /*
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/synch_bitops.h linux-2.6.18-xen/include/asm-ia64/synch_bitops.h
---- linux-2.6.18.3/include/asm-ia64/synch_bitops.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-ia64/synch_bitops.h 2006-11-19 14:27:05.000000000 +0100
+
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/synch_bitops.h b/include/asm-ia64/synch_bitops.h
+--- a/include/asm-ia64/synch_bitops.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-ia64/synch_bitops.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,61 @@
+#ifndef __XEN_SYNCH_BITOPS_H__
+#define __XEN_SYNCH_BITOPS_H__
@@ -78522,9 +73782,9 @@
+#define synch_cmpxchg_subword synch_cmpxchg
+
+#endif /* __XEN_SYNCH_BITOPS_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/system.h linux-2.6.18-xen/include/asm-ia64/system.h
---- linux-2.6.18.3/include/asm-ia64/system.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-ia64/system.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/system.h b/include/asm-ia64/system.h
+--- a/include/asm-ia64/system.h 2007-03-12 21:58:14.000000000 +0100
++++ b/include/asm-ia64/system.h 2007-03-14 10:55:43.000000000 +0100
@@ -123,7 +123,7 @@
#define __local_irq_save(x) \
do { \
@@ -78543,9 +73803,9 @@
#define irqs_disabled() \
({ \
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-ia64/xen/privop.h linux-2.6.18-xen/include/asm-ia64/xen/privop.h
---- linux-2.6.18.3/include/asm-ia64/xen/privop.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-ia64/xen/privop.h 2006-11-19 14:27:05.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-ia64/xen/privop.h b/include/asm-ia64/xen/privop.h
+--- a/include/asm-ia64/xen/privop.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-ia64/xen/privop.h 2007-03-14 10:55:43.000000000 +0100
@@ -0,0 +1,304 @@
+#ifndef _ASM_IA64_XEN_PRIVOP_H
+#define _ASM_IA64_XEN_PRIVOP_H
@@ -78851,9 +74111,9 @@
+#define ia64_pal_call_static xen_pal_call_static
+
+#endif /* _ASM_IA64_XEN_PRIVOP_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-um/page.h linux-2.6.18-xen/include/asm-um/page.h
---- linux-2.6.18.3/include/asm-um/page.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-um/page.h 2006-11-19 14:27:09.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-um/page.h b/include/asm-um/page.h
+--- a/include/asm-um/page.h 2007-03-12 21:58:15.000000000 +0100
++++ b/include/asm-um/page.h 2007-03-14 10:55:49.000000000 +0100
@@ -114,7 +114,7 @@
extern struct page *arch_validate(struct page *page, gfp_t mask, int order);
#define HAVE_ARCH_VALIDATE
@@ -78863,12 +74123,12 @@
#define HAVE_ARCH_FREE_PAGE
#include <asm-generic/memory_model.h>
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/apic.h linux-2.6.18-xen/include/asm-x86_64/apic.h
---- linux-2.6.18.3/include/asm-x86_64/apic.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-x86_64/apic.h 2006-11-19 14:27:10.000000000 +0100
-@@ -98,11 +98,13 @@
- extern int disable_timer_pin_1;
-
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/apic.h b/include/asm-x86_64/apic.h
+--- a/include/asm-x86_64/apic.h 2007-03-16 18:49:42.000000000 +0100
++++ b/include/asm-x86_64/apic.h 2007-03-14 10:55:49.000000000 +0100
+@@ -95,11 +95,13 @@
+ #define K8_APIC_EXT_INT_MSG_EXT 0x7
+ #define K8_APIC_EXT_LVT_ENTRY_THRESHOLD 0
+#ifndef CONFIG_XEN
void smp_send_timer_broadcast_ipi(void);
@@ -78878,29 +74138,140 @@
#define ARCH_APICTIMER_STOPS_ON_C3 1
+#endif
- #endif /* CONFIG_X86_LOCAL_APIC */
+ extern unsigned boot_cpu_id;
+
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/bootsetup.h b/include/asm-x86_64/bootsetup.h
+--- a/include/asm-x86_64/bootsetup.h 2007-03-15 15:56:09.000000000 +0100
++++ b/include/asm-x86_64/bootsetup.h 2007-03-14 10:55:49.000000000 +0100
+@@ -24,9 +24,15 @@
+ #define AUX_DEVICE_INFO (*(unsigned char *) (PARAM+0x1FF))
+ #define LOADER_TYPE (*(unsigned char *) (PARAM+0x210))
+ #define KERNEL_START (*(unsigned int *) (PARAM+0x214))
++#ifdef CONFIG_XEN
++#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))
++#else
+ #define INITRD_START (*(unsigned int *) (PARAM+0x218))
+ #define INITRD_SIZE (*(unsigned int *) (PARAM+0x21c))
+ #define EDID_INFO (*(struct edid_info *) (PARAM+0x140))
++#endif
+ #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))
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/dma-mapping.h b/include/asm-x86_64/dma-mapping.h
+--- a/include/asm-x86_64/dma-mapping.h 2007-03-16 18:49:42.000000000 +0100
++++ b/include/asm-x86_64/dma-mapping.h 2007-03-14 10:55:49.000000000 +0100
+@@ -55,6 +55,7 @@
+ extern struct dma_mapping_ops* dma_ops;
+ extern int iommu_merge;
+
++#ifndef CONFIG_XEN
+ static inline int dma_mapping_error(dma_addr_t dma_addr)
+ {
+ if (dma_ops->mapping_error)
+@@ -195,6 +196,10 @@
+ }
+
+ extern struct device fallback_dev;
++#else
++#include <asm-i386/mach-xen/asm/dma-mapping.h>
++#endif /* CONFIG_XEN */
++
+ extern int panic_on_overflow;
+
+ #endif /* _X8664_DMA_MAPPING_H */
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/dmi.h b/include/asm-x86_64/dmi.h
+--- a/include/asm-x86_64/dmi.h 2007-03-12 21:58:16.000000000 +0100
++++ b/include/asm-x86_64/dmi.h 2007-03-14 10:55:49.000000000 +0100
+@@ -5,6 +5,10 @@
+
+ extern void *dmi_ioremap(unsigned long addr, unsigned long size);
+ extern void dmi_iounmap(void *addr, unsigned long size);
++#ifdef CONFIG_XEN
++extern void *bt_ioremap(unsigned long addr, unsigned long size);
++extern void bt_iounmap(void *addr, unsigned long size);
++#endif
+
+ #define DMI_MAX_DATA 2048
+
+@@ -21,7 +25,12 @@
+ return dmi_alloc_data + idx;
+ }
+
++#ifdef CONFIG_XEN
++#define dmi_ioremap bt_ioremap
++#define dmi_iounmap bt_iounmap
++#else
+ #define dmi_ioremap early_ioremap
+ #define dmi_iounmap early_iounmap
++#endif
+
+ #endif
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/dwarf2.h b/include/asm-x86_64/dwarf2.h
+--- a/include/asm-x86_64/dwarf2.h 2007-03-16 18:49:42.000000000 +0100
++++ b/include/asm-x86_64/dwarf2.h 2007-03-14 10:55:49.000000000 +0100
+@@ -13,7 +13,7 @@
+ away for older version.
+ */
+
+-#ifdef CONFIG_AS_CFI
++#ifdef CONFIG_UNWIND_INFO
+
+ #define CFI_STARTPROC .cfi_startproc
+ #define CFI_ENDPROC .cfi_endproc
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/e820.h b/include/asm-x86_64/e820.h
+--- a/include/asm-x86_64/e820.h 2007-03-16 18:49:42.000000000 +0100
++++ b/include/asm-x86_64/e820.h 2007-03-14 10:55:49.000000000 +0100
+@@ -41,13 +41,21 @@
+ extern void setup_memory_region(void);
+ extern void contig_e820_setup(void);
+ extern unsigned long e820_end_of_ram(void);
++#ifdef CONFIG_XEN
++extern void e820_reserve_resources(struct e820entry *e820, int nr_map);
++#else
+ extern void e820_reserve_resources(void);
++#endif
+ extern void e820_mark_nosave_regions(void);
+ 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);
+
++#ifdef CONFIG_XEN
++extern void e820_setup_gap(struct e820entry *e820, int nr_map);
++#else
+ extern void e820_setup_gap(void);
++#endif
+ extern void e820_register_active_regions(int nid,
+ unsigned long start_pfn, unsigned long end_pfn);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/ipi.h linux-2.6.18-xen/include/asm-x86_64/ipi.h
---- linux-2.6.18.3/include/asm-x86_64/ipi.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/asm-x86_64/ipi.h 2006-12-05 18:42:37.000000000 +0100
-@@ -49,6 +49,7 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/ipi.h b/include/asm-x86_64/ipi.h
+--- a/include/asm-x86_64/ipi.h 2007-03-12 21:58:16.000000000 +0100
++++ b/include/asm-x86_64/ipi.h 2007-03-14 10:55:49.000000000 +0100
+@@ -49,8 +49,12 @@
return SET_APIC_DEST_FIELD(mask);
}
-+#ifndef CONFIG_XEN_UNPRIVILEGED_GUEST
++
static inline void __send_IPI_shortcut(unsigned int shortcut, int vector, unsigned int dest)
{
++#ifdef CONFIG_XEN_UNPRIVILEGED_GUEST
++ BUG();
++#else
/*
-@@ -113,5 +114,6 @@
- }
- local_irq_restore(flags);
+ * Subtle. In the case of the 'never do double writes' workaround
+ * we have to lock out interrupts to be safe. As we don't care
+@@ -74,6 +78,7 @@
+ * Send the IPI. The write to APIC_ICR fires this off.
+ */
+ apic_write(APIC_ICR, cfg);
++#endif /* !CONFIG_XEN_UNPRIVILEGED_GUEST */
}
-+#endif /* CONFIG_XEN_UNPRIVILEGED_GUEST */
- #endif /* __ASM_IPI_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/arch_hooks.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/arch_hooks.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/arch_hooks.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/arch_hooks.h 2006-11-19 14:27:11.000000000 +0100
+
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/arch_hooks.h b/include/asm-x86_64/mach-xen/asm/arch_hooks.h
+--- a/include/asm-x86_64/mach-xen/asm/arch_hooks.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/arch_hooks.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,27 @@
+#ifndef _ASM_ARCH_HOOKS_H
+#define _ASM_ARCH_HOOKS_H
@@ -78918,7 +74289,7 @@
+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);
++extern irqreturn_t timer_interrupt(int irq, void *dev_id);
+
+/* these are the defined hooks */
+extern void intr_init_hook(void);
@@ -78929,56 +74300,10 @@
+extern void mca_nmi_hook(void);
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/bootsetup.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/bootsetup.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/bootsetup.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/bootsetup.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,42 @@
-+
-+#ifndef _X86_64_BOOTSETUP_H
-+#define _X86_64_BOOTSETUP_H 1
-+
-+#define BOOT_PARAM_SIZE 4096
-+extern char x86_boot_params[BOOT_PARAM_SIZE];
-+
-+/*
-+ * 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 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))
-+#define COMMAND_LINE saved_command_line
-+
-+#define RAMDISK_IMAGE_START_MASK 0x07FF
-+#define RAMDISK_PROMPT_FLAG 0x8000
-+#define RAMDISK_LOAD_FLAG 0x4000
-+
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/desc.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/desc.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/desc.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/desc.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,263 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/desc.h b/include/asm-x86_64/mach-xen/asm/desc.h
+--- a/include/asm-x86_64/mach-xen/asm/desc.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/desc.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,212 @@
+/* Written 2000 by Andi Kleen */
+#ifndef __ARCH_DESC_H
+#define __ARCH_DESC_H
@@ -78990,66 +74315,15 @@
+
+#include <linux/string.h>
+#include <linux/smp.h>
++#include <asm/desc_defs.h>
+
+#include <asm/segment.h>
+#include <asm/mmu.h>
+
-+// 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));
-+
-+struct n_desc_struct {
-+ unsigned int a,b;
-+};
-+
-+enum {
-+ GATE_INTERRUPT = 0xE,
-+ GATE_TRAP = 0xF,
-+ GATE_CALL = 0xC,
-+};
-+
-+// 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 PTR_LOW(x) ((unsigned long)(x) & 0xFFFF)
-+#define PTR_MIDDLE(x) (((unsigned long)(x) >> 16) & 0xFFFF)
-+#define PTR_HIGH(x) ((unsigned long)(x) >> 32)
-+
-+enum {
-+ DESC_TSS = 0x9,
-+ DESC_LDT = 0x2,
-+};
-+
-+// 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));
-+
-+struct desc_ptr {
-+ unsigned short size;
-+ unsigned long address;
-+} __attribute__((packed)) ;
++extern struct desc_struct cpu_gdt_table[GDT_ENTRIES];
+
+extern struct desc_ptr idt_descr, cpu_gdt_descr[NR_CPUS];
+
-+extern struct desc_struct cpu_gdt_table[GDT_ENTRIES];
-+
+#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))
+
@@ -79147,7 +74421,7 @@
+ * -1? seg base+limit should be pointing to the address of the
+ * last valid byte
+ */
-+ set_tssldt_descriptor(&cpu_gdt(cpu)[GDT_ENTRY_TSS],
++ set_tssldt_descriptor(&cpu_gdt(cpu)[GDT_ENTRY_TSS],
+ (unsigned long)addr, DESC_TSS,
+ IO_BITMAP_OFFSET + IO_BITMAP_BYTES + sizeof(unsigned long) - 1);
+}
@@ -79242,322 +74516,10 @@
+#endif /* !__ASSEMBLY__ */
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/dma-mapping.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/dma-mapping.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/dma-mapping.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/dma-mapping.h 2006-12-05 18:42:37.000000000 +0100
-@@ -0,0 +1,207 @@
-+#ifndef _X8664_DMA_MAPPING_H
-+#define _X8664_DMA_MAPPING_H 1
-+
-+/*
-+ * IOMMU interface. See Documentation/DMA-mapping.txt and DMA-API.txt for
-+ * documentation.
-+ */
-+
-+
-+#include <asm/scatterlist.h>
-+#include <asm/io.h>
-+#include <asm/swiotlb.h>
-+
-+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;
-+};
-+
-+extern dma_addr_t bad_dma_address;
-+extern struct dma_mapping_ops* dma_ops;
-+extern int iommu_merge;
-+
-+#if 0
-+static inline int valid_dma_direction(int dma_direction)
-+{
-+ return ((dma_direction == DMA_BIDIRECTIONAL) ||
-+ (dma_direction == DMA_TO_DEVICE) ||
-+ (dma_direction == DMA_FROM_DEVICE));
-+}
-+
-+static inline int dma_mapping_error(dma_addr_t dma_addr)
-+{
-+ if (dma_ops->mapping_error)
-+ return dma_ops->mapping_error(dma_addr);
-+
-+ return (dma_addr == bad_dma_address);
-+}
-+
-+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 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);
-+}
-+
-+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 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)
-+{
-+ 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();
-+}
-+
-+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();
-+}
-+
-+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);
-+ }
-+
-+ flush_write_buffers();
-+}
-+
-+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);
-+
-+ flush_write_buffers();
-+}
-+
-+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();
-+}
-+
-+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);
-+ }
-+
-+ flush_write_buffers();
-+}
-+
-+static inline int
-+dma_map_sg(struct device *hwdev, struct scatterlist *sg, int nents, int direction)
-+{
-+ 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);
-+}
-+
-+extern int dma_supported(struct device *hwdev, u64 mask);
-+
-+/* same for gart, swiotlb, and nommu */
-+static inline int dma_get_cache_alignment(void)
-+{
-+ return boot_cpu_data.x86_clflush_size;
-+}
-+
-+#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)
-+{
-+ flush_write_buffers();
-+}
-+
-+extern struct device fallback_dev;
-+#endif
-+extern int panic_on_overflow;
-+
-+#endif /* _X8664_DMA_MAPPING_H */
-+
-+#include <asm-i386/mach-xen/asm/dma-mapping.h>
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/dmi.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/dmi.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/dmi.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/dmi.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,29 @@
-+#ifndef _ASM_DMI_H
-+#define _ASM_DMI_H 1
-+
-+#include <asm/io.h>
-+
-+extern void *dmi_ioremap(unsigned long addr, unsigned long size);
-+extern void dmi_iounmap(void *addr, unsigned long size);
-+extern void *bt_ioremap(unsigned long addr, unsigned long size);
-+extern void bt_iounmap(void *addr, unsigned long size);
-+
-+#define DMI_MAX_DATA 2048
-+
-+extern int dmi_alloc_index;
-+extern char dmi_alloc_data[DMI_MAX_DATA];
-+
-+/* This is so early that there is no good way to allocate dynamic memory.
-+ Allocate data in an BSS array. */
-+static inline void *dmi_alloc(unsigned len)
-+{
-+ int idx = dmi_alloc_index;
-+ if ((dmi_alloc_index += len) > DMI_MAX_DATA)
-+ return NULL;
-+ return dmi_alloc_data + idx;
-+}
-+
-+#define dmi_ioremap bt_ioremap
-+#define dmi_iounmap bt_iounmap
-+
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/e820.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/e820.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/e820.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/e820.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,64 @@
-+/*
-+ * 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 */
-+
-+#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 HIGH_MEMORY (1024*1024)
-+
-+#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;
-+#endif/*!__ASSEMBLY__*/
-+
-+#endif/*__E820_HEADER*/
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/fixmap.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/fixmap.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/fixmap.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/fixmap.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,113 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/fixmap.h b/include/asm-x86_64/mach-xen/asm/fixmap.h
+--- a/include/asm-x86_64/mach-xen/asm/fixmap.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/fixmap.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,109 @@
+/*
+ * fixmap.h: compile-time virtual memory allocation
+ *
@@ -79599,13 +74561,9 @@
+ 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,
@@ -79671,10 +74629,10 @@
+}
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/floppy.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/floppy.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/floppy.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/floppy.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,206 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/floppy.h b/include/asm-x86_64/mach-xen/asm/floppy.h
+--- a/include/asm-x86_64/mach-xen/asm/floppy.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/floppy.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,207 @@
+/*
+ * Architecture specific parts of the Floppy driver
+ *
@@ -79691,6 +74649,7 @@
+
+#include <linux/vmalloc.h>
+
++
+/*
+ * The DMA channel used by the floppy controller cannot access data at
+ * addresses >= 16MB
@@ -79731,7 +74690,7 @@
+static int virtual_dma_mode;
+static int doing_pdma;
+
-+static irqreturn_t floppy_hardint(int irq, void *dev_id, struct pt_regs * regs)
++static irqreturn_t floppy_hardint(int irq, void *dev_id)
+{
+ register unsigned char st;
+
@@ -79743,7 +74702,7 @@
+ static int dma_wait=0;
+#endif
+ if (!doing_pdma)
-+ return floppy_interrupt(irq, dev_id, regs);
++ return floppy_interrupt(irq, dev_id);
+
+#ifdef TRACE_FLPY_INT
+ if(!calls)
@@ -79786,7 +74745,7 @@
+ dma_wait=0;
+#endif
+ doing_pdma = 0;
-+ floppy_interrupt(irq, dev_id, regs);
++ floppy_interrupt(irq, dev_id);
+ return IRQ_HANDLED;
+ }
+#ifdef TRACE_FLPY_INT
@@ -79881,9 +74840,9 @@
+#define EXTRA_FLOPPY_PARAMS
+
+#endif /* __ASM_XEN_X86_64_FLOPPY_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/hw_irq.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/hw_irq.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/hw_irq.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/hw_irq.h 2006-12-05 18:42:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/hw_irq.h b/include/asm-x86_64/mach-xen/asm/hw_irq.h
+--- a/include/asm-x86_64/mach-xen/asm/hw_irq.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/hw_irq.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,136 @@
+#ifndef _ASM_HW_IRQ_H
+#define _ASM_HW_IRQ_H
@@ -79906,8 +74865,7 @@
+#include <asm/irq.h>
+#include <linux/profile.h>
+#include <linux/smp.h>
-+
-+struct hw_interrupt_type;
++#include <linux/percpu.h>
+#endif
+
+#define NMI_VECTOR 0x02
@@ -79964,9 +74922,10 @@
+
+
+#ifndef __ASSEMBLY__
-+extern u8 irq_vector[NR_IRQ_VECTORS];
-+#define IO_APIC_VECTOR(irq) (irq_vector[irq])
-+#define AUTO_ASSIGN -1
++typedef int vector_irq_t[NR_VECTORS];
++DECLARE_PER_CPU(vector_irq_t, vector_irq);
++extern void __setup_vector_irq(int cpu);
++extern spinlock_t vector_lock;
+
+/*
+ * Various low-level irq details needed by irq.c, process.c,
@@ -80021,9 +74980,9 @@
+#endif
+
+#endif /* _ASM_HW_IRQ_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/hypercall.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/hypercall.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/hypercall.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/hypercall.h 2006-11-19 14:27:11.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/hypercall.h b/include/asm-x86_64/mach-xen/asm/hypercall.h
+--- a/include/asm-x86_64/mach-xen/asm/hypercall.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/hypercall.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,389 @@
+/******************************************************************************
+ * hypercall.h
@@ -80414,16 +75373,16 @@
+}
+
+#endif /* __HYPERCALL_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/hypervisor.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/hypervisor.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/hypervisor.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/hypervisor.h 2006-11-19 14:27:11.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/hypervisor.h b/include/asm-x86_64/mach-xen/asm/hypervisor.h
+--- a/include/asm-x86_64/mach-xen/asm/hypervisor.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/hypervisor.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,2 @@
+
+#include <asm-i386/mach-xen/asm/hypervisor.h>
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/io.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/io.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/io.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/io.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,327 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/io.h b/include/asm-x86_64/mach-xen/asm/io.h
+--- a/include/asm-x86_64/mach-xen/asm/io.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/io.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,300 @@
+#ifndef _ASM_IO_H
+#define _ASM_IO_H
+
@@ -80697,33 +75656,6 @@
+
+#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
-+ *
-+ * 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;
-+}
-+
+/* Nothing to do */
+
+#define dma_cache_inv(_start,_size) do { } while (0)
@@ -80751,9 +75683,9 @@
+#define ARCH_HAS_DEV_MEM
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/irqflags.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/irqflags.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/irqflags.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/irqflags.h 2006-12-05 18:42:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/irqflags.h b/include/asm-x86_64/mach-xen/asm/irqflags.h
+--- a/include/asm-x86_64/mach-xen/asm/irqflags.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/irqflags.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,65 @@
+/*
+ * include/asm-x86_64/irqflags.h
@@ -80820,10 +75752,10 @@
+#endif
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/irq.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/irq.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/irq.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/irq.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,38 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/irq.h b/include/asm-x86_64/mach-xen/asm/irq.h
+--- a/include/asm-x86_64/mach-xen/asm/irq.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/irq.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,34 @@
+#ifndef _ASM_IRQ_H
+#define _ASM_IRQ_H
+
@@ -80846,11 +75778,7 @@
+ return ((irq == 2) ? 9 : irq);
+}
+
-+#ifdef CONFIG_X86_LOCAL_APIC
+#define ARCH_HAS_NMI_WATCHDOG /* See include/linux/nmi.h */
-+#endif
-+
-+#define KDB_VECTOR 0xf9
+
+# define irq_ctx_init(cpu) do { } while (0)
+
@@ -80862,9 +75790,9 @@
+#define __ARCH_HAS_DO_SOFTIRQ 1
+
+#endif /* _ASM_IRQ_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/maddr.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/maddr.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/maddr.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/maddr.h 2006-11-19 14:27:11.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/maddr.h b/include/asm-x86_64/mach-xen/asm/maddr.h
+--- a/include/asm-x86_64/mach-xen/asm/maddr.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/maddr.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,139 @@
+#ifndef _X86_64_MADDR_H
+#define _X86_64_MADDR_H
@@ -81005,9 +75933,9 @@
+
+#endif /* _X86_64_MADDR_H */
+
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/mmu_context.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/mmu_context.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/mmu_context.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/mmu_context.h 2006-11-19 14:27:11.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/mmu_context.h b/include/asm-x86_64/mach-xen/asm/mmu_context.h
+--- a/include/asm-x86_64/mach-xen/asm/mmu_context.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/mmu_context.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,135 @@
+#ifndef __X86_64_MMU_CONTEXT_H
+#define __X86_64_MMU_CONTEXT_H
@@ -81144,52 +76072,10 @@
+}
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/mmu.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/mmu.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/mmu.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/mmu.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,38 @@
-+#ifndef __x86_64_MMU_H
-+#define __x86_64_MMU_H
-+
-+#include <linux/spinlock.h>
-+#include <asm/semaphore.h>
-+
-+/*
-+ * 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;
-+
-+/* mm/memory.c:exit_mmap hook */
-+extern void _arch_exit_mmap(struct mm_struct *mm);
-+#define arch_exit_mmap(_mm) _arch_exit_mmap(_mm)
-+
-+/* 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
-+
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/msr.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/msr.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/msr.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/msr.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,399 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/msr.h b/include/asm-x86_64/mach-xen/asm/msr.h
+--- a/include/asm-x86_64/mach-xen/asm/msr.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/msr.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,427 @@
+#ifndef X86_64_MSR_H
+#define X86_64_MSR_H 1
+
@@ -81258,14 +76144,25 @@
+#define rdtscl(low) \
+ __asm__ __volatile__ ("rdtsc" : "=a" (low) : : "edx")
+
++#define rdtscp(low,high,aux) \
++ asm volatile (".byte 0x0f,0x01,0xf9" : "=a" (low), "=d" (high), "=c" (aux))
++
+#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)
+
++#define rdtscpll(val, aux) do { \
++ unsigned long __a, __d; \
++ asm volatile (".byte 0x0f,0x01,0xf9" : "=a" (__a), "=d" (__d), "=c" (aux)); \
++ (val) = (__d << 32) | __a; \
++} while (0)
++
+#define write_tsc(val1,val2) wrmsr(0x10, val1, val2)
+
++#define write_rdtscp_aux(val) wrmsr(0xc0000103, val, 0)
++
+#define rdpmc(counter,low,high) \
+ __asm__ __volatile__("rdpmc" \
+ : "=a" (low), "=d" (high) \
@@ -81350,8 +76247,8 @@
+#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_FS_BASE 0xc0000100 /* 64bit FS base */
++#define MSR_GS_BASE 0xc0000101 /* 64bit GS base */
+#define MSR_KERNEL_GS_BASE 0xc0000102 /* SwapGS GS shadow (or USER_GS from kernel) */
+/* EFER bits: */
+#define _EFER_SCE 0 /* SYSCALL/SYSRET */
@@ -81370,6 +76267,7 @@
+
+#define MSR_IA32_PERFCTR0 0xc1
+#define MSR_IA32_PERFCTR1 0xc2
++#define MSR_FSB_FREQ 0xcd
+
+#define MSR_MTRRcap 0x0fe
+#define MSR_IA32_BBL_CR_CTL 0x119
@@ -81391,6 +76289,10 @@
+#define MSR_IA32_LASTINTFROMIP 0x1dd
+#define MSR_IA32_LASTINTTOIP 0x1de
+
++#define MSR_IA32_PEBS_ENABLE 0x3f1
++#define MSR_IA32_DS_AREA 0x600
++#define MSR_IA32_PERF_CAPABILITIES 0x345
++
+#define MSR_MTRRfix64K_00000 0x250
+#define MSR_MTRRfix16K_80000 0x258
+#define MSR_MTRRfix16K_A0000 0x259
@@ -81488,6 +76390,9 @@
+#define MSR_IA32_PERF_STATUS 0x198
+#define MSR_IA32_PERF_CTL 0x199
+
++#define MSR_IA32_MPERF 0xE7
++#define MSR_IA32_APERF 0xE8
++
+#define MSR_IA32_THERM_CONTROL 0x19a
+#define MSR_IA32_THERM_INTERRUPT 0x19b
+#define MSR_IA32_THERM_STATUS 0x19c
@@ -81558,138 +76463,50 @@
+#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
-+
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/nmi.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/nmi.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/nmi.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/nmi.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,93 @@
-+/*
-+ * linux/include/asm-i386/nmi.h
-+ */
-+#ifndef ASM_NMI_H
-+#define ASM_NMI_H
-+
-+#include <linux/pm.h>
-+#include <asm/io.h>
-+
-+#include <xen/interface/nmi.h>
-+
-+struct pt_regs;
-+
-+typedef int (*nmi_callback_t)(struct pt_regs * regs, int cpu);
-+
-+/**
-+ * 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);
-+
-+/**
-+ * unset_nmi_callback
-+ *
-+ * Remove the handler previously set.
-+ */
-+void unset_nmi_callback(void);
-+
-+#ifdef CONFIG_PM
-+
-+/** Replace the PM callback routine for NMI. */
-+struct pm_dev * set_nmi_pm_callback(pm_callback callback);
-+
-+/** Unset the PM callback routine back to the default. */
-+void unset_nmi_pm_callback(struct pm_dev * dev);
-+
-+#else
-+
-+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)
-+{
-+}
-+
-+#endif /* CONFIG_PM */
-+
-+extern void default_do_nmi(struct pt_regs *);
-+extern void die_nmi(char *str, struct pt_regs *regs);
-+
-+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;
-+}
++#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 int panic_on_timeout;
-+extern int unknown_nmi_panic;
++/* Intel Core-based CPU performance counters */
++#define MSR_CORE_PERF_FIXED_CTR0 0x309
++#define MSR_CORE_PERF_FIXED_CTR1 0x30a
++#define MSR_CORE_PERF_FIXED_CTR2 0x30b
++#define MSR_CORE_PERF_FIXED_CTR_CTRL 0x38d
++#define MSR_CORE_PERF_GLOBAL_STATUS 0x38e
++#define MSR_CORE_PERF_GLOBAL_CTRL 0x38f
++#define MSR_CORE_PERF_GLOBAL_OVF_CTRL 0x390
+
-+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);
-+
-+extern void nmi_watchdog_default(void);
-+extern int setup_nmi_watchdog(char *);
-+
-+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
-+
-+#endif /* ASM_NMI_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/page.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/page.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/page.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/page.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,208 @@
++#endif
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/page.h b/include/asm-x86_64/mach-xen/asm/page.h
+--- a/include/asm-x86_64/mach-xen/asm/page.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/page.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,213 @@
+#ifndef _X86_64_PAGE_H
+#define _X86_64_PAGE_H
+
@@ -81769,7 +76586,6 @@
+
+#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..
+ */
@@ -81837,7 +76653,7 @@
+#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 __PAGE_OFFSET 0xffff880000000000UL
+
+#else
+#define __PHYSICAL_START CONFIG_PHYSICAL_START
@@ -81863,6 +76679,12 @@
+#define KERNEL_TEXT_SIZE (40UL*1024*1024)
+#define KERNEL_TEXT_START 0xffffffff80000000UL
+
++#ifndef __ASSEMBLY__
++
++#include <asm/bug.h>
++
++#endif /* __ASSEMBLY__ */
++
+#define PAGE_OFFSET ((unsigned long)__PAGE_OFFSET)
+
+/* Note: __pa(&symbol_visible_to_c) should be always replaced with __pa_symbol.
@@ -81898,213 +76720,10 @@
+#endif /* __KERNEL__ */
+
+#endif /* _X86_64_PAGE_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/param.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/param.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/param.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/param.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,22 @@
-+#ifndef _ASMx86_64_PARAM_H
-+#define _ASMx86_64_PARAM_H
-+
-+#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
-+
-+#ifndef HZ
-+#define HZ 100
-+#endif
-+
-+#define EXEC_PAGESIZE 4096
-+
-+#ifndef NOGROUP
-+#define NOGROUP (-1)
-+#endif
-+
-+#define MAXHOSTNAMELEN 64 /* max length of hostname */
-+
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/pci.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/pci.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/pci.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/pci.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,173 @@
-+#ifndef __x8664_PCI_H
-+#define __x8664_PCI_H
-+
-+#include <asm/io.h>
-+
-+#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 */
-+
-+#ifdef CONFIG_PCI
-+extern unsigned int pcibios_assign_all_busses(void);
-+#else
-+#define pcibios_assign_all_busses() 0
-+#endif
-+#define pcibios_scan_all_fns(a, b) 0
-+
-+extern unsigned long pci_mem_start;
-+#define PCIBIOS_MIN_IO 0x1000
-+#define PCIBIOS_MIN_MEM (pci_mem_start)
-+
-+#define PCIBIOS_MIN_CARDBUS_IO 0x4000
-+
-+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);
-+
-+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);
-+
-+#include <linux/types.h>
-+#include <linux/slab.h>
-+#include <asm/scatterlist.h>
-+#include <linux/string.h>
-+#include <asm/page.h>
-+
-+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
-+ *
-+ * 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)
-+
-+#if defined(CONFIG_IOMMU) || defined(CONFIG_CALGARY_IOMMU)
-+
-+/*
-+ * 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)
-+
-+#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)
-+
-+#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)
-+
-+#endif
-+
-+#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);
-+}
-+
-+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));
-+}
-+
-+static inline unsigned long
-+pci_dac_dma_to_offset(struct pci_dev *pdev, dma64_addr_t dma_addr)
-+{
-+ return (dma_addr & ~PAGE_MASK);
-+}
-+
-+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 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();
-+}
-+
-+#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
-+
-+#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)
-+{
-+}
-+
-+#endif /* __KERNEL__ */
-+
-+/* generic pci stuff */
-+#ifdef CONFIG_PCI
-+#include <asm-generic/pci.h>
-+#endif
-+
-+/* On Xen we have to scan all functions since Xen hides bridges from
-+ * us. If a bridge is at fn=0 and that slot has a multifunction
-+ * device, we won't find the additional devices without scanning all
-+ * functions. */
-+#undef pcibios_scan_all_fns
-+#define pcibios_scan_all_fns(a, b) 1
-+
-+#endif /* __x8664_PCI_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/pgalloc.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/pgalloc.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/pgalloc.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/pgalloc.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,226 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/pgalloc.h b/include/asm-x86_64/mach-xen/asm/pgalloc.h
+--- a/include/asm-x86_64/mach-xen/asm/pgalloc.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/pgalloc.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,224 @@
+#ifndef _X86_64_PGALLOC_H
+#define _X86_64_PGALLOC_H
+
@@ -82127,18 +76746,6 @@
+ set_pmd(pmd, __pmd(_PAGE_TABLE | __pa(pte)));
+}
+
-+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 inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd)
+{
+ if (unlikely((mm)->context.pinned)) {
@@ -82171,6 +76778,18 @@
+ }
+}
+
++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 inline void pmd_free(pmd_t *pmd)
+{
+ pte_t *ptep = virt_to_ptep(pmd);
@@ -82184,19 +76803,17 @@
+ free_page((unsigned long)pmd);
+}
+
-+static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long addr)
++static inline pmd_t *pmd_alloc_one (struct mm_struct *mm, unsigned long addr)
+{
-+ pmd_t *pmd = (pmd_t *) get_zeroed_page(GFP_KERNEL|__GFP_REPEAT);
-+ return pmd;
++ return (pmd_t *)get_zeroed_page(GFP_KERNEL|__GFP_REPEAT);
+}
+
+static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long addr)
+{
-+ pud_t *pud = (pud_t *) get_zeroed_page(GFP_KERNEL|__GFP_REPEAT);
-+ return pud;
++ return (pud_t *)get_zeroed_page(GFP_KERNEL|__GFP_REPEAT);
+}
+
-+static inline void pud_free(pud_t *pud)
++static inline void pud_free (pud_t *pud)
+{
+ pte_t *ptep = virt_to_ptep(pud);
+
@@ -82331,10 +76948,10 @@
+#define __pud_free_tlb(tlb,x) pud_free((x))
+
+#endif /* _X86_64_PGALLOC_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/pgtable.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/pgtable.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/pgtable.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/pgtable.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,558 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/pgtable.h b/include/asm-x86_64/mach-xen/asm/pgtable.h
+--- a/include/asm-x86_64/mach-xen/asm/pgtable.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/pgtable.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,553 @@
+#ifndef _X86_64_PGTABLE_H
+#define _X86_64_PGTABLE_H
+
@@ -82381,12 +76998,9 @@
+
+#define swapper_pg_dir init_level4_pgt
+
-+extern int nonx_setup(char *str);
+extern void paging_init(void);
+extern void clear_kernel_mapping(unsigned long addr, unsigned long size);
+
-+extern unsigned long pgkern_mask;
-+
+/*
+ * ZERO_PAGE is a global shared page that is always zero: used
+ * for zero-mapped memory areas etc..
@@ -82433,22 +77047,27 @@
+#define set_pte_batched(pteptr, pteval) \
+ queue_l1_entry_update(pteptr, (pteval))
+
-+extern inline int pud_present(pud_t pud) { return !pud_none(pud); }
-+
+static inline void set_pte(pte_t *dst, pte_t val)
+{
+ *dst = val;
+}
++#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_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 void pud_clear (pud_t * pud)
++static inline void pud_clear (pud_t *pud)
+{
+ set_pud(pud, __pud(0));
+}
+
++#define set_pgd(pgdptr, pgdval) xen_l4_entry_update(pgdptr, (pgdval))
++
+#define __user_pgd(pgd) ((pgd) + PTRS_PER_PGD)
+
+static inline void pgd_clear (pgd_t * pgd)
@@ -82457,9 +77076,6 @@
+ set_pgd(__user_pgd(pgd), __pgd(0));
+}
+
-+#define pud_page(pud) \
-+ ((unsigned long) __va(pud_val(pud) & PHYSICAL_PAGE_MASK))
-+
+/*
+ * A note on implementation of this atomic 'get-and-clear' operation.
+ * This is actually very simple because Xen Linux can only run on a single
@@ -82607,41 +77223,32 @@
+
+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);
++ return pgd_val(pgd) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER);
+}
+
-+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);
-+}
++static inline unsigned long pud_bad(pud_t pud)
++{
++ return pud_val(pud) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER);
++}
+
-+#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)
++static inline unsigned long pmd_bad(pmd_t pmd)
++{
++ return pmd_val(pmd) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER);
++}
+
+#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_pfn(_pte) mfn_to_local_pfn(pte_mfn(_pte))
-+
++#define pages_to_mb(x) ((x) >> (20-PAGE_SHIFT)) /* FIXME: is this
++ right? */
+#define pte_page(x) pfn_to_page(pte_pfn(x))
++#define pte_pfn(x) mfn_to_local_pfn(pte_mfn(x))
++#define pte_mfn(_pte) (((_pte).pte & PTE_MASK) >> PAGE_SHIFT)
+
+static inline pte_t pfn_pte(unsigned long page_nr, pgprot_t pgprot)
+{
+ pte_t pte;
-+
+ (pte).pte = (pfn_to_mfn(page_nr) << PAGE_SHIFT);
+ (pte).pte |= pgprot_val(pgprot);
+ (pte).pte &= __supported_pte_mask;
@@ -82657,7 +77264,7 @@
+#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_exec(pte_t pte) { return !(__pte_val(pte) & _PAGE_NX); }
+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; }
@@ -82670,11 +77277,12 @@
+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_mkexec(pte_t pte) { __pte_val(pte) &= ~_PAGE_NX; 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; }
++static inline pte_t pte_clrhuge(pte_t pte) { __pte_val(pte) &= ~_PAGE_PSE; return pte; }
+
+struct vm_area_struct;
+
@@ -82722,7 +77330,8 @@
+ * Level 4 access.
+ * Never use these in the common code.
+ */
-+#define pgd_page(pgd) ((unsigned long) __va(pgd_val(pgd) & PTE_MASK))
++#define pgd_page_vaddr(pgd) ((unsigned long) __va((unsigned long)pgd_val(pgd) & PTE_MASK))
++#define pgd_page(pgd) (pfn_to_page(pgd_val(pgd) >> PAGE_SHIFT))
+#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))
@@ -82731,8 +77340,11 @@
+
+/* PUD - Level3 access */
+/* to find an entry in a page-table-directory. */
++#define pud_page_vaddr(pud) ((unsigned long) __va(pud_val(pud) & PHYSICAL_PAGE_MASK))
++#define pud_page(pud) (pfn_to_page(pud_val(pud) >> PAGE_SHIFT))
+#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_offset(pgd, address) ((pud_t *) pgd_page_vaddr(*(pgd)) + pud_index(address))
++extern inline int pud_present(pud_t pud) { return !pud_none(pud); }
+
+/* Find correct pud via the hidden fourth level page level: */
+
@@ -82744,18 +77356,17 @@
+}
+
+/* PMD - Level 2 access */
-+#define pmd_page_kernel(pmd) ((unsigned long) __va(pmd_val(pmd) & PTE_MASK))
++#define pmd_page_vaddr(pmd) ((unsigned long) __va(pmd_val(pmd) & PTE_MASK))
+#define pmd_page(pmd) (pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT))
+
+#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_offset(dir, address) ((pmd_t *) pud_page_vaddr(*(dir)) + \
++ pmd_index(address))
+#define pmd_none(x) (!pmd_val(x))
+/* 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_clear(xp) do { set_pmd(xp, __pmd(0)); } while (0)
-+#define pmd_bad(x) ((pmd_val(x) & (~PAGE_MASK & ~_PAGE_USER & ~_PAGE_PRESENT)) != (_KERNPG_TABLE & ~_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)
+
@@ -82773,7 +77384,8 @@
+static inline pte_t mk_pte_phys(unsigned long physpage, pgprot_t pgprot)
+{
+ pte_t pte;
-+ (pte).pte = physpage | pgprot_val(pgprot);
++ pte.pte = physpage | pgprot_val(pgprot);
++ pte.pte &= __supported_pte_mask;
+ return pte;
+}
+
@@ -82788,7 +77400,7 @@
+
+#define pte_index(address) \
+ (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1))
-+#define pte_offset_kernel(dir, address) ((pte_t *) pmd_page_kernel(*(dir)) + \
++#define pte_offset_kernel(dir, address) ((pte_t *) pmd_page_vaddr(*(dir)) + \
+ pte_index(address))
+
+/* x86-64 always has all page tables mapped. */
@@ -82893,10 +77505,10 @@
+#include <asm-generic/pgtable.h>
+
+#endif /* _X86_64_PGTABLE_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/processor.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/processor.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/processor.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/processor.h 2006-12-05 18:42:37.000000000 +0100
-@@ -0,0 +1,503 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/processor.h b/include/asm-x86_64/mach-xen/asm/processor.h
+--- a/include/asm-x86_64/mach-xen/asm/processor.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/processor.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,507 @@
+/*
+ * include/asm-x86_64/processor.h
+ *
@@ -83067,12 +77679,6 @@
+
+
+/*
-+ * Bus types
-+ */
-+#define MCA_bus 0
-+#define MCA_bus__is_a_macro
-+
-+/*
+ * User space process size. 47bits minus one guard page.
+ */
+#define TASK_SIZE64 (0x800000000000UL - 4096)
@@ -83384,6 +77990,16 @@
+ : :"a" (eax), "c" (ecx));
+}
+
++static inline void __sti_mwait(unsigned long eax, unsigned long ecx)
++{
++ /* "mwait %eax,%ecx;" */
++ asm volatile(
++ "sti; .byte 0x0f,0x01,0xc9;"
++ : :"a" (eax), "c" (ecx));
++}
++
++extern void mwait_idle_with_hints(unsigned long eax, unsigned long ecx);
++
+#define stack_current() \
+({ \
+ struct thread_info *ti; \
@@ -83400,166 +78016,28 @@
+#define HAVE_ARCH_PICK_MMAP_LAYOUT 1
+
+#endif /* __ASM_X86_64_PROCESSOR_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/ptrace.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/ptrace.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/ptrace.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/ptrace.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,125 @@
-+#ifndef _X86_64_PTRACE_H
-+#define _X86_64_PTRACE_H
-+
-+#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__ */
-+
-+/* top of stack page */
-+#define FRAME_SIZE 168
-+
-+#define PTRACE_OLDSETOPTIONS 21
-+
-+#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 */
-+};
-+
-+#endif
-+
-+/* 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
-+
-+/* 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 */
-+
-+#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
-+
-+void signal_fault(struct pt_regs *regs, void __user *frame, char *where);
-+
-+struct task_struct;
-+
-+extern unsigned long
-+convert_rip_to_linear(struct task_struct *child, struct pt_regs *regs);
-+
-+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 */
-+};
-+
-+#endif
-+
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/smp.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/smp.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/smp.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/smp.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,150 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/smp.h b/include/asm-x86_64/mach-xen/asm/smp.h
+--- a/include/asm-x86_64/mach-xen/asm/smp.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/smp.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,119 @@
+#ifndef __ASM_SMP_H
+#define __ASM_SMP_H
+
+/*
+ * 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
+
-+#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
+
+#include <asm/pda.h>
+
@@ -83576,7 +78054,6 @@
+
+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;
@@ -83604,28 +78081,22 @@
+
+#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
+
-+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 NO_PROC_ID 0xFF /* No processor magic marker */
+
+#endif
+
-+#ifndef ASSEMBLY
+/*
+ * Some lowlevel functions might want to know about
+ * the real APIC ID <-> CPU # mapping.
@@ -83634,12 +78105,6 @@
+extern u8 x86_cpu_to_log_apicid[NR_CPUS];
+extern u8 bios_cpu_apicid[];
+
-+#ifdef CONFIG_X86_LOCAL_APIC
-+static inline unsigned int cpu_mask_to_apicid(cpumask_t cpumask)
-+{
-+ return cpus_addr(cpumask)[0];
-+}
-+
+static inline int cpu_present_to_apicid(int mps_cpu)
+{
+ if (mps_cpu < NR_CPUS)
@@ -83647,13 +78112,9 @@
+ else
+ return BAD_APICID;
+}
-+#endif
-+
-+#endif /* !ASSEMBLY */
+
+#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>
@@ -83665,39 +78126,35 @@
+})
+#endif
+
-+#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
+
+#ifdef CONFIG_SMP
+#define cpu_physical_id(cpu) x86_cpu_to_apicid[cpu]
+#else
+#define cpu_physical_id(cpu) boot_cpu_id
++#endif /* !CONFIG_SMP */
+#endif
+
-+#endif
-+
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/synch_bitops.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/synch_bitops.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/synch_bitops.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/synch_bitops.h 2006-11-19 14:27:11.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/synch_bitops.h b/include/asm-x86_64/mach-xen/asm/synch_bitops.h
+--- a/include/asm-x86_64/mach-xen/asm/synch_bitops.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/synch_bitops.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,2 @@
+
+#include <asm-i386/mach-xen/asm/synch_bitops.h>
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/system.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/system.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/system.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/system.h 2006-12-05 18:42:37.000000000 +0100
-@@ -0,0 +1,264 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/system.h b/include/asm-x86_64/mach-xen/asm/system.h
+--- a/include/asm-x86_64/mach-xen/asm/system.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/system.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,266 @@
+#ifndef __ASM_SYSTEM_H
+#define __ASM_SYSTEM_H
+
+#include <linux/kernel.h>
+#include <asm/segment.h>
++#include <asm/alternative.h>
+#include <asm/synch_bitops.h>
+#include <asm/hypervisor.h>
+#include <xen/interface/arch-x86_64.h>
@@ -83717,12 +78174,13 @@
+#define __RESTORE(reg,offset) "movq (14-" #offset ")*8(%%rsp),%%" #reg "\n\t"
+
+/* 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"
++#define SAVE_CONTEXT "pushf ; pushq %%rbp ; movq %%rsi,%%rbp\n\t"
++#define RESTORE_CONTEXT "movq %%rbp,%%rsi ; popq %%rbp ; popf\t"
+
+#define __EXTRA_CLOBBER \
+ ,"rcx","rbx","rdx","r8","r9","r10","r11","r12","r13","r14","r15"
+
++/* Save restore flags to clear handle leaking NT */
+#define switch_to(prev,next,last) \
+ asm volatile(SAVE_CONTEXT \
+ "movq %%rsp,%P[threadrsp](%[prev])\n\t" /* save RSP */ \
@@ -83745,8 +78203,7 @@
+ [pda_pcurrent] "i" (offsetof(struct x8664_pda, pcurrent)) \
+ : "memory", "cc" __EXTRA_CLOBBER)
+
-+
-+extern void load_gs_index(unsigned);
++extern void load_gs_index(unsigned);
+
+/*
+ * Load a segment. Fall back on loading the zero
@@ -83785,11 +78242,12 @@
+ asm volatile("movq %0,%%cr0" :: "r" (val));
+}
+
-+#define read_cr3() ({ \
-+ unsigned long __dummy; \
-+ asm("movq %%cr3,%0" : "=r" (__dummy)); \
-+ machine_to_phys(__dummy); \
-+})
++static inline unsigned long read_cr3(void)
++{
++ unsigned long cr3;
++ asm("movq %%cr3,%0" : "=r" (cr3));
++ return machine_to_phys(cr3);
++}
+
+static inline unsigned long read_cr4(void)
+{
@@ -83957,9 +78415,9 @@
+extern void free_init_pages(char *what, unsigned long begin, unsigned long end);
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/timer.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/timer.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/timer.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/timer.h 2006-11-19 14:27:11.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/timer.h b/include/asm-x86_64/mach-xen/asm/timer.h
+--- a/include/asm-x86_64/mach-xen/asm/timer.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/timer.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,67 @@
+#ifndef _ASMi386_TIMER_H
+#define _ASMi386_TIMER_H
@@ -84028,28 +78486,45 @@
+extern struct init_timer_opts timer_pmtmr_init;
+#endif
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/tlbflush.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/tlbflush.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/tlbflush.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/tlbflush.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,103 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/tlbflush.h b/include/asm-x86_64/mach-xen/asm/tlbflush.h
+--- a/include/asm-x86_64/mach-xen/asm/tlbflush.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/tlbflush.h 2007-03-14 10:55:49.000000000 +0100
+@@ -0,0 +1,120 @@
+#ifndef _X8664_TLBFLUSH_H
+#define _X8664_TLBFLUSH_H
+
+#include <linux/mm.h>
+#include <asm/processor.h>
+
-+#define __flush_tlb() xen_tlb_flush()
++static inline unsigned long get_cr3(void)
++{
++ unsigned long cr3;
++ asm volatile("mov %%cr3,%0" : "=r" (cr3));
++ return machine_to_phys(cr3);
++}
+
-+/*
-+ * 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()
++static inline void set_cr3(unsigned long cr3)
++{
++ BUG();
++ /* What the hell is this supposed to do: JQ */
++ asm volatile("mov %0,%%cr3" :: "r" (cr3) : "memory");
++}
++
++#define __flush_tlb() xen_tlb_flush()
+
++static inline unsigned long get_cr4(void)
++{
++ unsigned long cr4;
++ asm volatile("mov %%cr4,%0" : "=r" (cr4));
++ return cr4;
++}
+
-+extern unsigned long pgkern_mask;
++static inline void set_cr4(unsigned long cr4)
++{
++ asm volatile("mov %0,%%cr4" :: "r" (cr4) : "memory");
++}
+
-+#define __flush_tlb_all() __flush_tlb_global()
++#define __flush_tlb_all() xen_tlb_flush()
+
+#define __flush_tlb_one(addr) xen_invlpg((unsigned long)addr)
+
@@ -84135,33 +78610,9 @@
+}
+
+#endif /* _X8664_TLBFLUSH_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/vga.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/vga.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/vga.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/vga.h 2006-11-19 14:27:11.000000000 +0100
-@@ -0,0 +1,20 @@
-+/*
-+ * Access to VGA videoram
-+ *
-+ * (c) 1998 Martin Mares <mj at ucw.cz>
-+ */
-+
-+#ifndef _LINUX_ASM_VGA_H_
-+#define _LINUX_ASM_VGA_H_
-+
-+/*
-+ * On the PC, we can just recalculate addresses and then
-+ * access the videoram directly without any black magic.
-+ */
-+
-+#define VGA_MAP_MEM(x,s) (unsigned long)isa_bus_to_virt(x)
-+
-+#define vga_readb(x) (*(x))
-+#define vga_writeb(x,y) (*(y) = (x))
-+
-+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/xor.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/xor.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/asm/xor.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/asm/xor.h 2006-11-19 14:27:11.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/asm/xor.h b/include/asm-x86_64/mach-xen/asm/xor.h
+--- a/include/asm-x86_64/mach-xen/asm/xor.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/asm/xor.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,328 @@
+/*
+ * x86-64 changes / gcc fixes from Andi Kleen.
@@ -84179,7 +78630,7 @@
+ preempt_disable(); \
+ if (!(current_thread_info()->status & TS_USEDFPU)) \
+ clts(); \
-+ __asm__ __volatile__ ( \
++ asm volatile ( \
+ "movups %%xmm0,(%1) ;\n\t" \
+ "movups %%xmm1,0x10(%1) ;\n\t" \
+ "movups %%xmm2,0x20(%1) ;\n\t" \
@@ -84491,9 +78942,9 @@
+ 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 -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/irq_vectors.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/irq_vectors.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/irq_vectors.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/irq_vectors.h 2006-11-19 14:27:11.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/irq_vectors.h b/include/asm-x86_64/mach-xen/irq_vectors.h
+--- a/include/asm-x86_64/mach-xen/irq_vectors.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/irq_vectors.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,123 @@
+/*
+ * This file should contain #defines for all of the interrupt vector
@@ -84618,9 +79069,9 @@
+#define irq_to_dynirq(_x) ((_x) - DYNIRQ_BASE)
+
+#endif /* _ASM_IRQ_VECTORS_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/mach_time.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/mach_time.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/mach_time.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/mach_time.h 2006-11-19 14:27:11.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/mach_time.h b/include/asm-x86_64/mach-xen/mach_time.h
+--- a/include/asm-x86_64/mach-xen/mach_time.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/mach_time.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,111 @@
+/*
+ * include/asm-i386/mach-default/mach_time.h
@@ -84733,9 +79184,9 @@
+}
+
+#endif /* !_MACH_TIME_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/mach_timer.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/mach_timer.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/mach_timer.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/mach_timer.h 2006-11-19 14:27:11.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/mach_timer.h b/include/asm-x86_64/mach-xen/mach_timer.h
+--- a/include/asm-x86_64/mach-xen/mach_timer.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/mach_timer.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,48 @@
+/*
+ * include/asm-i386/mach-default/mach_timer.h
@@ -84785,9 +79236,9 @@
+}
+
+#endif /* !_MACH_TIMER_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/setup_arch_post.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/setup_arch_post.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/setup_arch_post.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/setup_arch_post.h 2006-11-19 14:27:11.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/setup_arch_post.h b/include/asm-x86_64/mach-xen/setup_arch_post.h
+--- a/include/asm-x86_64/mach-xen/setup_arch_post.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/setup_arch_post.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,58 @@
+/**
+ * machine_specific_* - Hooks for machine specific setup.
@@ -84847,113 +79298,143 @@
+ }
+#endif
+}
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/asm-x86_64/mach-xen/setup_arch_pre.h linux-2.6.18-xen/include/asm-x86_64/mach-xen/setup_arch_pre.h
---- linux-2.6.18.3/include/asm-x86_64/mach-xen/setup_arch_pre.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/asm-x86_64/mach-xen/setup_arch_pre.h 2006-11-19 14:27:11.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mach-xen/setup_arch_pre.h b/include/asm-x86_64/mach-xen/setup_arch_pre.h
+--- a/include/asm-x86_64/mach-xen/setup_arch_pre.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/asm-x86_64/mach-xen/setup_arch_pre.h 2007-03-14 10:55:49.000000000 +0100
@@ -0,0 +1,5 @@
+/* Hook to call BIOS initialisation function */
+
+#define ARCH_SETUP machine_specific_arch_setup();
+
+static void __init machine_specific_arch_setup(void);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/linux/elfnote.h linux-2.6.18-xen/include/linux/elfnote.h
---- linux-2.6.18.3/include/linux/elfnote.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/linux/elfnote.h 2006-11-19 14:27:14.000000000 +0100
-@@ -0,0 +1,90 @@
-+#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)
-+ */
-+#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 /* !__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 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 -urN -x .hg -x .hgtags linux-2.6.18.3/include/linux/gfp.h linux-2.6.18-xen/include/linux/gfp.h
---- linux-2.6.18.3/include/linux/gfp.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/linux/gfp.h 2006-11-19 14:27:14.000000000 +0100
-@@ -99,7 +99,11 @@
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/mmu.h b/include/asm-x86_64/mmu.h
+--- a/include/asm-x86_64/mmu.h 2007-03-12 21:58:16.000000000 +0100
++++ b/include/asm-x86_64/mmu.h 2007-03-14 10:55:49.000000000 +0100
+@@ -15,6 +15,24 @@
+ 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;
++
++/* mm/memory.c:exit_mmap hook */
++extern void _arch_exit_mmap(struct mm_struct *mm);
++#define arch_exit_mmap(_mm) _arch_exit_mmap(_mm)
++
++/* 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
++
+ #endif
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/nmi.h b/include/asm-x86_64/nmi.h
+--- a/include/asm-x86_64/nmi.h 2007-03-16 18:49:42.000000000 +0100
++++ b/include/asm-x86_64/nmi.h 2007-03-14 10:55:49.000000000 +0100
+@@ -7,6 +7,10 @@
+ #include <linux/pm.h>
+ #include <asm/io.h>
+
++#ifdef CONFIG_XEN
++#include <xen/interface/nmi.h>
++#endif
++
+ /**
+ * do_nmi_callback
+ *
+@@ -39,7 +43,25 @@
+ extern void default_do_nmi(struct pt_regs *);
+ extern void die_nmi(char *str, struct pt_regs *regs, int do_panic);
+
++#ifdef CONFIG_XEN
++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;
++}
++#else
+ #define get_nmi_reason() inb(0x61)
++#endif
+
+ extern int panic_on_timeout;
+ extern int unknown_nmi_panic;
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/pci.h b/include/asm-x86_64/pci.h
+--- a/include/asm-x86_64/pci.h 2007-03-12 21:58:16.000000000 +0100
++++ b/include/asm-x86_64/pci.h 2007-03-14 10:55:49.000000000 +0100
+@@ -75,6 +75,23 @@
+ #define pci_unmap_len_set(PTR, LEN_NAME, VAL) \
+ (((PTR)->LEN_NAME) = (VAL))
+
++#elif defined(CONFIG_SWIOTLB)
++
++#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 */
+
+@@ -146,4 +163,13 @@
+ #include <asm-generic/pci.h>
+ #endif
+
++#ifdef CONFIG_XEN
++/* On Xen we have to scan all functions since Xen hides bridges from
++ * us. If a bridge is at fn=0 and that slot has a multifunction
++ * device, we won't find the additional devices without scanning all
++ * functions. */
++#undef pcibios_scan_all_fns
++#define pcibios_scan_all_fns(a, b) 1
++#endif
++
+ #endif /* __x8664_PCI_H */
+diff -x '.hg*' -x '.git*' -urN a/include/asm-x86_64/vga.h b/include/asm-x86_64/vga.h
+--- a/include/asm-x86_64/vga.h 2007-03-12 21:58:16.000000000 +0100
++++ b/include/asm-x86_64/vga.h 2007-03-14 10:55:49.000000000 +0100
+@@ -12,7 +12,11 @@
+ * access the videoram directly without any black magic.
+ */
+
++#ifdef CONFIG_XEN
++#define VGA_MAP_MEM(x,s) (unsigned long)isa_bus_to_virt(x)
++#else
+ #define VGA_MAP_MEM(x,s) (unsigned long)phys_to_virt(x)
++#endif
+
+ #define vga_readb(x) (*(x))
+ #define vga_writeb(x,y) (*(y) = (x))
+diff -x '.hg*' -x '.git*' -urN a/include/linux/gfp.h b/include/linux/gfp.h
+--- a/include/linux/gfp.h 2007-03-16 18:49:43.000000000 +0100
++++ b/include/linux/gfp.h 2007-03-14 10:55:50.000000000 +0100
+@@ -114,7 +114,11 @@
*/
#ifndef HAVE_ARCH_FREE_PAGE
@@ -84964,15 +79445,15 @@
+ */
+static inline int arch_free_page(struct page *page, int order) { return 0; }
#endif
-
- extern struct page *
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/linux/highmem.h linux-2.6.18-xen/include/linux/highmem.h
---- linux-2.6.18.3/include/linux/highmem.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/linux/highmem.h 2006-11-19 14:27:14.000000000 +0100
-@@ -24,10 +24,16 @@
-
+ #ifndef HAVE_ARCH_ALLOC_PAGE
+ static inline void arch_alloc_page(struct page *page, int order) { }
+diff -x '.hg*' -x '.git*' -urN a/include/linux/highmem.h b/include/linux/highmem.h
+--- a/include/linux/highmem.h 2007-03-16 18:49:43.000000000 +0100
++++ b/include/linux/highmem.h 2007-03-14 10:55:50.000000000 +0100
+@@ -26,10 +26,16 @@
/* declarations for linux/mm/highmem.c */
unsigned int nr_free_highpages(void);
+ extern unsigned long totalhigh_pages;
+#ifdef CONFIG_XEN
+void kmap_flush_unused(void);
+#endif
@@ -84984,12 +79465,12 @@
+static inline void kmap_flush_unused(void) { }
+#endif
- static inline void *kmap(struct page *page)
- {
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/linux/interrupt.h linux-2.6.18-xen/include/linux/interrupt.h
---- linux-2.6.18.3/include/linux/interrupt.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/linux/interrupt.h 2006-11-19 14:27:14.000000000 +0100
-@@ -166,6 +166,12 @@
+ #define totalhigh_pages 0
+
+diff -x '.hg*' -x '.git*' -urN a/include/linux/interrupt.h b/include/linux/interrupt.h
+--- a/include/linux/interrupt.h 2007-03-16 18:49:43.000000000 +0100
++++ b/include/linux/interrupt.h 2007-03-14 10:55:50.000000000 +0100
+@@ -184,6 +184,12 @@
#endif /* CONFIG_GENERIC_HARDIRQS */
@@ -85002,20 +79483,20 @@
#ifndef __ARCH_SET_SOFTIRQ_PENDING
#define set_softirq_pending(x) (local_softirq_pending() = (x))
#define or_softirq_pending(x) (local_softirq_pending() |= (x))
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/linux/mm.h linux-2.6.18-xen/include/linux/mm.h
---- linux-2.6.18.3/include/linux/mm.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/linux/mm.h 2006-11-19 14:27:14.000000000 +0100
-@@ -164,6 +164,9 @@
- #define VM_NONLINEAR 0x00800000 /* Is non-linear (remap_file_pages) */
+diff -x '.hg*' -x '.git*' -urN a/include/linux/mm.h b/include/linux/mm.h
+--- a/include/linux/mm.h 2007-03-16 18:49:43.000000000 +0100
++++ b/include/linux/mm.h 2007-03-14 10:55:50.000000000 +0100
+@@ -169,6 +169,9 @@
#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 */
+ #define VM_ALWAYSDUMP 0x04000000 /* Always include in core dumps */
+#ifdef CONFIG_XEN
-+#define VM_FOREIGN 0x04000000 /* Has pages belonging to another VM */
++#define VM_FOREIGN 0x08000000 /* 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
-@@ -1027,6 +1030,13 @@
+@@ -1129,6 +1132,13 @@
#define FOLL_GET 0x04 /* do get_page on page */
#define FOLL_ANON 0x08 /* give ZERO_PAGE if no pgtable */
@@ -85029,9 +79510,9 @@
#ifdef CONFIG_PROC_FS
void vm_stat_account(struct mm_struct *, unsigned long, struct file *, long);
#else
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/linux/oprofile.h linux-2.6.18-xen/include/linux/oprofile.h
---- linux-2.6.18.3/include/linux/oprofile.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/linux/oprofile.h 2006-11-19 14:27:14.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/linux/oprofile.h b/include/linux/oprofile.h
+--- a/include/linux/oprofile.h 2007-03-12 21:58:16.000000000 +0100
++++ b/include/linux/oprofile.h 2007-03-14 10:55:50.000000000 +0100
@@ -16,6 +16,10 @@
#include <linux/types.h>
#include <linux/spinlock.h>
@@ -85064,9 +79545,9 @@
/**
* Create a file of the given name as a child of the given root, with
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/linux/pfn.h linux-2.6.18-xen/include/linux/pfn.h
---- linux-2.6.18.3/include/linux/pfn.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/linux/pfn.h 2006-12-05 18:42:37.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/linux/pfn.h b/include/linux/pfn.h
+--- a/include/linux/pfn.h 2007-03-12 21:58:16.000000000 +0100
++++ b/include/linux/pfn.h 2007-03-14 10:55:50.000000000 +0100
@@ -4,6 +4,10 @@
#define PFN_ALIGN(x) (((unsigned long)(x) + (PAGE_SIZE - 1)) & PAGE_MASK)
#define PFN_UP(x) (((x) + PAGE_SIZE-1) >> PAGE_SHIFT)
@@ -85078,10 +79559,10 @@
+#endif
#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/linux/skbuff.h linux-2.6.18-xen/include/linux/skbuff.h
---- linux-2.6.18.3/include/linux/skbuff.h 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/include/linux/skbuff.h 2006-11-19 14:27:15.000000000 +0100
-@@ -203,6 +203,8 @@
+diff -x '.hg*' -x '.git*' -urN a/include/linux/skbuff.h b/include/linux/skbuff.h
+--- a/include/linux/skbuff.h 2007-03-16 18:49:44.000000000 +0100
++++ b/include/linux/skbuff.h 2007-03-14 10:55:51.000000000 +0100
+@@ -202,6 +202,8 @@
* @local_df: allow local fragmentation
* @cloned: Head may be cloned (check refcnt to be sure)
* @nohdr: Payload reference only, must not modify header
@@ -85090,7 +79571,7 @@
* @pkt_type: Packet class
* @fclone: skbuff clone status
* @ip_summed: Driver fed us an IP checksum
-@@ -282,7 +284,13 @@
+@@ -284,7 +286,13 @@
nfctinfo:3;
__u8 pkt_type:3,
fclone:2,
@@ -85104,9 +79585,9 @@
__be16 protocol;
void (*destructor)(struct sk_buff *skb);
-@@ -345,7 +353,8 @@
+@@ -347,7 +355,8 @@
- extern struct sk_buff *alloc_skb_from_cache(kmem_cache_t *cp,
+ extern struct sk_buff *alloc_skb_from_cache(struct kmem_cache *cp,
unsigned int size,
- gfp_t priority);
+ gfp_t priority,
@@ -85114,7 +79595,7 @@
extern void kfree_skbmem(struct sk_buff *skb);
extern struct sk_buff *skb_clone(struct sk_buff *skb,
gfp_t priority);
-@@ -1086,6 +1095,7 @@
+@@ -1088,6 +1097,7 @@
kfree_skb(skb);
}
@@ -85122,7 +79603,7 @@
/**
* __dev_alloc_skb - allocate an skbuff for receiving
* @length: length to allocate
-@@ -1106,6 +1116,9 @@
+@@ -1108,6 +1118,9 @@
skb_reserve(skb, NET_SKB_PAD);
return skb;
}
@@ -85132,9 +79613,9 @@
/**
* dev_alloc_skb - allocate an skbuff for receiving
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/balloon.h linux-2.6.18-xen/include/xen/balloon.h
---- linux-2.6.18.3/include/xen/balloon.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/balloon.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/balloon.h b/include/xen/balloon.h
+--- a/include/xen/balloon.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/balloon.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,57 @@
+/******************************************************************************
+ * balloon.h
@@ -85193,9 +79674,9 @@
+#define balloon_unlock(__flags) spin_unlock_irqrestore(&balloon_lock, __flags)
+
+#endif /* __ASM_BALLOON_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/cpu_hotplug.h linux-2.6.18-xen/include/xen/cpu_hotplug.h
---- linux-2.6.18.3/include/xen/cpu_hotplug.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/cpu_hotplug.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/cpu_hotplug.h b/include/xen/cpu_hotplug.h
+--- a/include/xen/cpu_hotplug.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/cpu_hotplug.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,43 @@
+#ifndef __XEN_CPU_HOTPLUG_H__
+#define __XEN_CPU_HOTPLUG_H__
@@ -85240,9 +79721,9 @@
+#endif /* !defined(CONFIG_HOTPLUG_CPU) */
+
+#endif /* __XEN_CPU_HOTPLUG_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/driver_util.h linux-2.6.18-xen/include/xen/driver_util.h
---- linux-2.6.18.3/include/xen/driver_util.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/driver_util.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/driver_util.h b/include/xen/driver_util.h
+--- a/include/xen/driver_util.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/driver_util.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,15 @@
+
+#ifndef __ASM_XEN_DRIVER_UTIL_H__
@@ -85259,9 +79740,9 @@
+extern void unlock_vm_area(struct vm_struct *area);
+
+#endif /* __ASM_XEN_DRIVER_UTIL_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/evtchn.h linux-2.6.18-xen/include/xen/evtchn.h
---- linux-2.6.18.3/include/xen/evtchn.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/evtchn.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/evtchn.h b/include/xen/evtchn.h
+--- a/include/xen/evtchn.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/evtchn.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,113 @@
+/******************************************************************************
+ * evtchn.h
@@ -85320,21 +79801,21 @@
+ */
+extern int bind_evtchn_to_irqhandler(
+ unsigned int evtchn,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ irq_handler_t handler,
+ unsigned long irqflags,
+ const char *devname,
+ void *dev_id);
+extern int bind_virq_to_irqhandler(
+ unsigned int virq,
+ unsigned int cpu,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ irq_handler_t handler,
+ unsigned long irqflags,
+ const char *devname,
+ void *dev_id);
+extern int bind_ipi_to_irqhandler(
+ unsigned int ipi,
+ unsigned int cpu,
-+ irqreturn_t (*handler)(int, void *, struct pt_regs *),
++ irq_handler_t handler,
+ unsigned long irqflags,
+ const char *devname,
+ void *dev_id);
@@ -85376,9 +79857,9 @@
+extern void notify_remote_via_irq(int irq);
+
+#endif /* __ASM_EVTCHN_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/features.h linux-2.6.18-xen/include/xen/features.h
---- linux-2.6.18.3/include/xen/features.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/features.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/features.h b/include/xen/features.h
+--- a/include/xen/features.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/features.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,20 @@
+/******************************************************************************
+ * features.h
@@ -85400,9 +79881,9 @@
+#define xen_feature(flag) (xen_features[flag])
+
+#endif /* __ASM_XEN_FEATURES_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/foreign_page.h linux-2.6.18-xen/include/xen/foreign_page.h
---- linux-2.6.18.3/include/xen/foreign_page.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/foreign_page.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/foreign_page.h b/include/xen/foreign_page.h
+--- a/include/xen/foreign_page.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/foreign_page.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,30 @@
+/******************************************************************************
+ * foreign_page.h
@@ -85434,9 +79915,9 @@
+ ( (void (*) (struct page *)) (page)->mapping )
+
+#endif /* __ASM_XEN_FOREIGN_PAGE_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/gnttab.h linux-2.6.18-xen/include/xen/gnttab.h
---- linux-2.6.18.3/include/xen/gnttab.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/gnttab.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/gnttab.h b/include/xen/gnttab.h
+--- a/include/xen/gnttab.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/gnttab.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,150 @@
+/******************************************************************************
+ * gnttab.h
@@ -85588,9 +80069,9 @@
+}
+
+#endif /* __ASM_GNTTAB_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/hvm.h linux-2.6.18-xen/include/xen/hvm.h
---- linux-2.6.18.3/include/xen/hvm.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/hvm.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/hvm.h b/include/xen/hvm.h
+--- a/include/xen/hvm.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/hvm.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,24 @@
+/* Simple wrappers around HVM functions */
+#ifndef XEN_HVM_H__
@@ -85616,9 +80097,9 @@
+}
+
+#endif /* XEN_HVM_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/hypervisor_sysfs.h linux-2.6.18-xen/include/xen/hypervisor_sysfs.h
---- linux-2.6.18.3/include/xen/hypervisor_sysfs.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/hypervisor_sysfs.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/hypervisor_sysfs.h b/include/xen/hypervisor_sysfs.h
+--- a/include/xen/hypervisor_sysfs.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/hypervisor_sysfs.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,32 @@
+/*
+ * copyright (c) 2006 IBM Corporation
@@ -85652,9 +80133,9 @@
+};
+
+#endif /* _HYP_SYSFS_H_ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/acm.h linux-2.6.18-xen/include/xen/interface/acm.h
---- linux-2.6.18.3/include/xen/interface/acm.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/acm.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/acm.h b/include/xen/interface/acm.h
+--- a/include/xen/interface/acm.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/acm.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,187 @@
+/*
+ * acm.h: Xen access control module interface defintions
@@ -85843,9 +80324,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/acm_ops.h linux-2.6.18-xen/include/xen/interface/acm_ops.h
---- linux-2.6.18.3/include/xen/interface/acm_ops.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/acm_ops.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/acm_ops.h b/include/xen/interface/acm_ops.h
+--- a/include/xen/interface/acm_ops.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/acm_ops.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,102 @@
+/*
+ * acm_ops.h: Xen access control module hypervisor commands
@@ -85949,9 +80430,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/arch-ia64.h linux-2.6.18-xen/include/xen/interface/arch-ia64.h
---- linux-2.6.18.3/include/xen/interface/arch-ia64.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/arch-ia64.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/arch-ia64.h b/include/xen/interface/arch-ia64.h
+--- a/include/xen/interface/arch-ia64.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/arch-ia64.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,448 @@
+/******************************************************************************
+ * arch-ia64/hypervisor-if.h
@@ -86401,9 +80882,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/arch-powerpc.h linux-2.6.18-xen/include/xen/interface/arch-powerpc.h
---- linux-2.6.18.3/include/xen/interface/arch-powerpc.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/arch-powerpc.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/arch-powerpc.h b/include/xen/interface/arch-powerpc.h
+--- a/include/xen/interface/arch-powerpc.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/arch-powerpc.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,115 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
@@ -86520,9 +81001,9 @@
+#endif
+
+#endif
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/arch-x86_32.h linux-2.6.18-xen/include/xen/interface/arch-x86_32.h
---- linux-2.6.18.3/include/xen/interface/arch-x86_32.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/arch-x86_32.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/arch-x86_32.h b/include/xen/interface/arch-x86_32.h
+--- a/include/xen/interface/arch-x86_32.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/arch-x86_32.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,255 @@
+/******************************************************************************
+ * arch-x86_32.h
@@ -86779,9 +81260,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/arch-x86_64.h linux-2.6.18-xen/include/xen/interface/arch-x86_64.h
---- linux-2.6.18.3/include/xen/interface/arch-x86_64.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/arch-x86_64.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/arch-x86_64.h b/include/xen/interface/arch-x86_64.h
+--- a/include/xen/interface/arch-x86_64.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/arch-x86_64.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,322 @@
+/******************************************************************************
+ * arch-x86_64.h
@@ -87105,9 +81586,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/callback.h linux-2.6.18-xen/include/xen/interface/callback.h
---- linux-2.6.18.3/include/xen/interface/callback.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/callback.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/callback.h b/include/xen/interface/callback.h
+--- a/include/xen/interface/callback.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/callback.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,74 @@
+/******************************************************************************
+ * callback.h
@@ -87183,9 +81664,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/dom0_ops.h linux-2.6.18-xen/include/xen/interface/dom0_ops.h
---- linux-2.6.18.3/include/xen/interface/dom0_ops.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/dom0_ops.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/dom0_ops.h b/include/xen/interface/dom0_ops.h
+--- a/include/xen/interface/dom0_ops.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/dom0_ops.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,102 @@
+/******************************************************************************
+ * dom0_ops.h
@@ -87289,9 +81770,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/domctl.h linux-2.6.18-xen/include/xen/interface/domctl.h
---- linux-2.6.18.3/include/xen/interface/domctl.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/domctl.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/domctl.h b/include/xen/interface/domctl.h
+--- a/include/xen/interface/domctl.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/domctl.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,392 @@
+/******************************************************************************
+ * domctl.h
@@ -87685,9 +82166,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/elfnote.h linux-2.6.18-xen/include/xen/interface/elfnote.h
---- linux-2.6.18.3/include/xen/interface/elfnote.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/elfnote.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/elfnote.h b/include/xen/interface/elfnote.h
+--- a/include/xen/interface/elfnote.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/elfnote.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,133 @@
+/******************************************************************************
+ * elfnote.h
@@ -87822,9 +82303,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/event_channel.h linux-2.6.18-xen/include/xen/interface/event_channel.h
---- linux-2.6.18.3/include/xen/interface/event_channel.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/event_channel.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/event_channel.h b/include/xen/interface/event_channel.h
+--- a/include/xen/interface/event_channel.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/event_channel.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,233 @@
+/******************************************************************************
+ * event_channel.h
@@ -88059,9 +82540,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/features.h linux-2.6.18-xen/include/xen/interface/features.h
---- linux-2.6.18.3/include/xen/interface/features.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/features.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/features.h b/include/xen/interface/features.h
+--- a/include/xen/interface/features.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/features.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,53 @@
+/******************************************************************************
+ * features.h
@@ -88116,9 +82597,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/grant_table.h linux-2.6.18-xen/include/xen/interface/grant_table.h
---- linux-2.6.18.3/include/xen/interface/grant_table.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/grant_table.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/grant_table.h b/include/xen/interface/grant_table.h
+--- a/include/xen/interface/grant_table.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/grant_table.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,362 @@
+/******************************************************************************
+ * grant_table.h
@@ -88482,9 +82963,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/hvm/e820.h linux-2.6.18-xen/include/xen/interface/hvm/e820.h
---- linux-2.6.18.3/include/xen/interface/hvm/e820.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/hvm/e820.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/hvm/e820.h b/include/xen/interface/hvm/e820.h
+--- a/include/xen/interface/hvm/e820.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/hvm/e820.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,32 @@
+#ifndef __XEN_PUBLIC_HVM_E820_H__
+#define __XEN_PUBLIC_HVM_E820_H__
@@ -88518,9 +82999,9 @@
+#define HVM_BELOW_4G_MMIO_LENGTH ((1ULL << 32) - HVM_BELOW_4G_MMIO_START)
+
+#endif /* __XEN_PUBLIC_HVM_E820_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/hvm/hvm_info_table.h linux-2.6.18-xen/include/xen/interface/hvm/hvm_info_table.h
---- linux-2.6.18.3/include/xen/interface/hvm/hvm_info_table.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/hvm/hvm_info_table.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/hvm/hvm_info_table.h b/include/xen/interface/hvm/hvm_info_table.h
+--- a/include/xen/interface/hvm/hvm_info_table.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/hvm/hvm_info_table.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,22 @@
+/******************************************************************************
+ * hvm/hvm_info_table.h
@@ -88544,9 +83025,9 @@
+};
+
+#endif /* __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/hvm/ioreq.h linux-2.6.18-xen/include/xen/interface/hvm/ioreq.h
---- linux-2.6.18.3/include/xen/interface/hvm/ioreq.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/hvm/ioreq.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/hvm/ioreq.h b/include/xen/interface/hvm/ioreq.h
+--- a/include/xen/interface/hvm/ioreq.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/hvm/ioreq.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,99 @@
+/*
+ * ioreq.h: I/O request definitions for device models
@@ -88647,9 +83128,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/hvm/params.h linux-2.6.18-xen/include/xen/interface/hvm/params.h
---- linux-2.6.18.3/include/xen/interface/hvm/params.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/hvm/params.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/hvm/params.h b/include/xen/interface/hvm/params.h
+--- a/include/xen/interface/hvm/params.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/hvm/params.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,24 @@
+#ifndef __XEN_PUBLIC_HVM_PARAMS_H__
+#define __XEN_PUBLIC_HVM_PARAMS_H__
@@ -88675,9 +83156,9 @@
+DEFINE_XEN_GUEST_HANDLE(xen_hvm_param_t);
+
+#endif /* __XEN_PUBLIC_HVM_PARAMS_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/hvm/vmx_assist.h linux-2.6.18-xen/include/xen/interface/hvm/vmx_assist.h
---- linux-2.6.18.3/include/xen/interface/hvm/vmx_assist.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/hvm/vmx_assist.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/hvm/vmx_assist.h b/include/xen/interface/hvm/vmx_assist.h
+--- a/include/xen/interface/hvm/vmx_assist.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/hvm/vmx_assist.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,98 @@
+/*
+ * vmx_assist.h: Context definitions for the VMXASSIST world switch.
@@ -88777,9 +83258,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/io/blkif.h linux-2.6.18-xen/include/xen/interface/io/blkif.h
---- linux-2.6.18.3/include/xen/interface/io/blkif.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/io/blkif.h 2006-11-19 14:27:15.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/io/blkif.h b/include/xen/interface/io/blkif.h
+--- a/include/xen/interface/io/blkif.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/io/blkif.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,87 @@
+/******************************************************************************
+ * blkif.h
@@ -88868,9 +83349,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/io/console.h linux-2.6.18-xen/include/xen/interface/io/console.h
---- linux-2.6.18.3/include/xen/interface/io/console.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/io/console.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/io/console.h b/include/xen/interface/io/console.h
+--- a/include/xen/interface/io/console.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/io/console.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,33 @@
+/******************************************************************************
+ * console.h
@@ -88905,9 +83386,241 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/io/netif.h linux-2.6.18-xen/include/xen/interface/io/netif.h
---- linux-2.6.18.3/include/xen/interface/io/netif.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/io/netif.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/io/fbif.h b/include/xen/interface/io/fbif.h
+--- a/include/xen/interface/io/fbif.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/io/fbif.h 2007-03-14 10:55:52.000000000 +0100
+@@ -0,0 +1,116 @@
++/*
++ * fbif.h -- Xen virtual frame buffer device
++ *
++ * Copyright (C) 2005 Anthony Liguori <aliguori at us.ibm.com>
++ * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster <armbru at redhat.com>
++ *
++ * 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.
++ */
++
++#ifndef __XEN_PUBLIC_IO_FBIF_H__
++#define __XEN_PUBLIC_IO_FBIF_H__
++
++#include <asm/types.h>
++
++/* Out events (frontend -> backend) */
++
++/*
++ * Out events may be sent only when requested by backend, and receipt
++ * of an unknown out event is an error.
++ */
++
++/* 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
++
++struct xenfb_update
++{
++ __u8 type; /* XENFB_TYPE_UPDATE */
++ __s32 x; /* source x */
++ __s32 y; /* source y */
++ __s32 width; /* rect width */
++ __s32 height; /* rect height */
++};
++
++#define XENFB_OUT_EVENT_SIZE 40
++
++union xenfb_out_event
++{
++ __u8 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
++
++union xenfb_in_event
++{
++ __u8 type;
++ char pad[XENFB_IN_EVENT_SIZE];
++};
++
++/* 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
++{
++ __u32 in_cons, in_prod;
++ __u32 out_cons, out_prod;
++
++ __s32 width; /* the width of the framebuffer (in pixels) */
++ __s32 height; /* the height of the framebuffer (in pixels) */
++ __u32 line_length; /* the length of a row of pixels (in bytes) */
++ __u32 mem_length; /* the length of the framebuffer (in bytes) */
++ __u8 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, that's 4 Megs. Two directory
++ * pages should be enough for a while.
++ */
++ unsigned long pd[2];
++};
++
++/*
++ * 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
++
++#endif
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/io/kbdif.h b/include/xen/interface/io/kbdif.h
+--- a/include/xen/interface/io/kbdif.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/io/kbdif.h 2007-03-14 10:55:52.000000000 +0100
+@@ -0,0 +1,108 @@
++/*
++ * kbdif.h -- Xen virtual keyboard/mouse
++ *
++ * Copyright (C) 2005 Anthony Liguori <aliguori at us.ibm.com>
++ * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster <armbru at redhat.com>
++ *
++ * 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.
++ */
++
++#ifndef __XEN_PUBLIC_IO_KBDIF_H__
++#define __XEN_PUBLIC_IO_KBDIF_H__
++
++#include <asm/types.h>
++
++/* In events (backend -> frontend) */
++
++/*
++ * 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
++{
++ __u8 type; /* XENKBD_TYPE_MOTION */
++ __s32 rel_x; /* relative X motion */
++ __s32 rel_y; /* relative Y motion */
++};
++
++struct xenkbd_key
++{
++ __u8 type; /* XENKBD_TYPE_KEY */
++ __u8 pressed; /* 1 if pressed; 0 otherwise */
++ __u32 keycode; /* KEY_* from linux/input.h */
++};
++
++struct xenkbd_position
++{
++ __u8 type; /* XENKBD_TYPE_POS */
++ __s32 abs_x; /* absolute X position (in FB pixels) */
++ __s32 abs_y; /* absolute Y position (in FB pixels) */
++};
++
++#define XENKBD_IN_EVENT_SIZE 40
++
++union xenkbd_in_event
++{
++ __u8 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
++{
++ __u8 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
++{
++ __u32 in_cons, in_prod;
++ __u32 out_cons, out_prod;
++};
++
++#endif
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/io/netif.h b/include/xen/interface/io/netif.h
+--- a/include/xen/interface/io/netif.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/io/netif.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,166 @@
+/******************************************************************************
+ * netif.h
@@ -89075,9 +83788,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/io/pciif.h linux-2.6.18-xen/include/xen/interface/io/pciif.h
---- linux-2.6.18.3/include/xen/interface/io/pciif.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/io/pciif.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/io/pciif.h b/include/xen/interface/io/pciif.h
+--- a/include/xen/interface/io/pciif.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/io/pciif.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,55 @@
+/*
+ * PCI Backend/Frontend Common Data Structures & Macros
@@ -89134,9 +83847,9 @@
+};
+
+#endif /* __XEN_PCI_COMMON_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/io/ring.h linux-2.6.18-xen/include/xen/interface/io/ring.h
---- linux-2.6.18.3/include/xen/interface/io/ring.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/io/ring.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/io/ring.h b/include/xen/interface/io/ring.h
+--- a/include/xen/interface/io/ring.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/io/ring.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,273 @@
+/******************************************************************************
+ * ring.h
@@ -89411,9 +84124,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/io/tpmif.h linux-2.6.18-xen/include/xen/interface/io/tpmif.h
---- linux-2.6.18.3/include/xen/interface/io/tpmif.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/io/tpmif.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/io/tpmif.h b/include/xen/interface/io/tpmif.h
+--- a/include/xen/interface/io/tpmif.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/io/tpmif.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,59 @@
+/******************************************************************************
+ * tpmif.h
@@ -89474,9 +84187,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/io/xenbus.h linux-2.6.18-xen/include/xen/interface/io/xenbus.h
---- linux-2.6.18.3/include/xen/interface/io/xenbus.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/io/xenbus.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/io/xenbus.h b/include/xen/interface/io/xenbus.h
+--- a/include/xen/interface/io/xenbus.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/io/xenbus.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,45 @@
+/*****************************************************************************
+ * xenbus.h
@@ -89523,9 +84236,9 @@
+typedef enum xenbus_state XenbusState;
+
+#endif /* _XEN_PUBLIC_IO_XENBUS_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/io/xs_wire.h linux-2.6.18-xen/include/xen/interface/io/xs_wire.h
---- linux-2.6.18.3/include/xen/interface/io/xs_wire.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/io/xs_wire.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/io/xs_wire.h b/include/xen/interface/io/xs_wire.h
+--- a/include/xen/interface/io/xs_wire.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/io/xs_wire.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,97 @@
+/*
+ * Details of the "wire" protocol between Xen Store Daemon and client
@@ -89624,9 +84337,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/memory.h linux-2.6.18-xen/include/xen/interface/memory.h
---- linux-2.6.18.3/include/xen/interface/memory.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/memory.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/memory.h b/include/xen/interface/memory.h
+--- a/include/xen/interface/memory.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/memory.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,243 @@
+/******************************************************************************
+ * memory.h
@@ -89871,9 +84584,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/nmi.h linux-2.6.18-xen/include/xen/interface/nmi.h
---- linux-2.6.18.3/include/xen/interface/nmi.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/nmi.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/nmi.h b/include/xen/interface/nmi.h
+--- a/include/xen/interface/nmi.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/nmi.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,60 @@
+/******************************************************************************
+ * nmi.h
@@ -89935,9 +84648,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/physdev.h linux-2.6.18-xen/include/xen/interface/physdev.h
---- linux-2.6.18.3/include/xen/interface/physdev.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/physdev.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/physdev.h b/include/xen/interface/physdev.h
+--- a/include/xen/interface/physdev.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/physdev.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,149 @@
+
+#ifndef __XEN_PUBLIC_PHYSDEV_H__
@@ -90088,9 +84801,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/platform.h linux-2.6.18-xen/include/xen/interface/platform.h
---- linux-2.6.18.3/include/xen/interface/platform.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/platform.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/platform.h b/include/xen/interface/platform.h
+--- a/include/xen/interface/platform.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/platform.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,125 @@
+/******************************************************************************
+ * platform.h
@@ -90217,9 +84930,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/sched.h linux-2.6.18-xen/include/xen/interface/sched.h
---- linux-2.6.18.3/include/xen/interface/sched.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/sched.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/sched.h b/include/xen/interface/sched.h
+--- a/include/xen/interface/sched.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/sched.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,103 @@
+/******************************************************************************
+ * sched.h
@@ -90324,9 +85037,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/sysctl.h linux-2.6.18-xen/include/xen/interface/sysctl.h
---- linux-2.6.18.3/include/xen/interface/sysctl.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/sysctl.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/sysctl.h b/include/xen/interface/sysctl.h
+--- a/include/xen/interface/sysctl.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/sysctl.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,151 @@
+/******************************************************************************
+ * sysctl.h
@@ -90479,9 +85192,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/trace.h linux-2.6.18-xen/include/xen/interface/trace.h
---- linux-2.6.18.3/include/xen/interface/trace.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/trace.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/trace.h b/include/xen/interface/trace.h
+--- a/include/xen/interface/trace.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/trace.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,87 @@
+/******************************************************************************
+ * include/public/trace.h
@@ -90570,9 +85283,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/vcpu.h linux-2.6.18-xen/include/xen/interface/vcpu.h
---- linux-2.6.18.3/include/xen/interface/vcpu.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/vcpu.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/vcpu.h b/include/xen/interface/vcpu.h
+--- a/include/xen/interface/vcpu.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/vcpu.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,121 @@
+/******************************************************************************
+ * vcpu.h
@@ -90695,9 +85408,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/version.h linux-2.6.18-xen/include/xen/interface/version.h
---- linux-2.6.18.3/include/xen/interface/version.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/version.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/version.h b/include/xen/interface/version.h
+--- a/include/xen/interface/version.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/version.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,73 @@
+/******************************************************************************
+ * version.h
@@ -90772,9 +85485,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/xencomm.h linux-2.6.18-xen/include/xen/interface/xencomm.h
---- linux-2.6.18.3/include/xen/interface/xencomm.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/xencomm.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/xencomm.h b/include/xen/interface/xencomm.h
+--- a/include/xen/interface/xencomm.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/xencomm.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,37 @@
+/*
+ * Copyright (C) 2006 Hollis Blanchard <hollisb at us.ibm.com>, IBM Corporation
@@ -90813,9 +85526,9 @@
+};
+
+#endif /* _XEN_XENCOMM_H_ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/xen-compat.h linux-2.6.18-xen/include/xen/interface/xen-compat.h
---- linux-2.6.18.3/include/xen/interface/xen-compat.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/xen-compat.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/xen-compat.h b/include/xen/interface/xen-compat.h
+--- a/include/xen/interface/xen-compat.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/xen-compat.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,26 @@
+/******************************************************************************
+ * xen-compat.h
@@ -90843,9 +85556,9 @@
+#endif
+
+#endif /* __XEN_PUBLIC_XEN_COMPAT_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/xen.h linux-2.6.18-xen/include/xen/interface/xen.h
---- linux-2.6.18.3/include/xen/interface/xen.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/xen.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/xen.h b/include/xen/interface/xen.h
+--- a/include/xen/interface/xen.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/xen.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,581 @@
+/******************************************************************************
+ * xen.h
@@ -91428,9 +86141,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/interface/xenoprof.h linux-2.6.18-xen/include/xen/interface/xenoprof.h
---- linux-2.6.18.3/include/xen/interface/xenoprof.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/interface/xenoprof.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/interface/xenoprof.h b/include/xen/interface/xenoprof.h
+--- a/include/xen/interface/xenoprof.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/interface/xenoprof.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,110 @@
+/******************************************************************************
+ * xenoprof.h
@@ -91542,9 +86255,9 @@
+ * indent-tabs-mode: nil
+ * End:
+ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/pcifront.h linux-2.6.18-xen/include/xen/pcifront.h
---- linux-2.6.18.3/include/xen/pcifront.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/pcifront.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/pcifront.h b/include/xen/pcifront.h
+--- a/include/xen/pcifront.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/pcifront.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,76 @@
+/*
+ * PCI Frontend - arch-dependendent declarations
@@ -91622,9 +86335,9 @@
+#endif /* __KERNEL__ */
+
+#endif /* __XEN_ASM_PCIFRONT_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/public/evtchn.h linux-2.6.18-xen/include/xen/public/evtchn.h
---- linux-2.6.18.3/include/xen/public/evtchn.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/public/evtchn.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/public/evtchn.h b/include/xen/public/evtchn.h
+--- a/include/xen/public/evtchn.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/public/evtchn.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,88 @@
+/******************************************************************************
+ * evtchn.h
@@ -91714,9 +86427,9 @@
+ _IOC(_IOC_NONE, 'E', 5, 0)
+
+#endif /* __LINUX_PUBLIC_EVTCHN_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/public/privcmd.h linux-2.6.18-xen/include/xen/public/privcmd.h
---- linux-2.6.18.3/include/xen/public/privcmd.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/public/privcmd.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/public/privcmd.h b/include/xen/public/privcmd.h
+--- a/include/xen/public/privcmd.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/public/privcmd.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,79 @@
+/******************************************************************************
+ * privcmd.h
@@ -91797,9 +86510,9 @@
+ _IOC(_IOC_NONE, 'P', 3, sizeof(privcmd_mmapbatch_t))
+
+#endif /* __LINUX_PUBLIC_PRIVCMD_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/xenbus.h linux-2.6.18-xen/include/xen/xenbus.h
---- linux-2.6.18.3/include/xen/xenbus.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/xenbus.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/xenbus.h b/include/xen/xenbus.h
+--- a/include/xen/xenbus.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/xenbus.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,306 @@
+/******************************************************************************
+ * xenbus.h
@@ -92107,9 +86820,9 @@
+int xenbus_frontend_closed(struct xenbus_device *dev);
+
+#endif /* _XEN_XENBUS_H */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/xencons.h linux-2.6.18-xen/include/xen/xencons.h
---- linux-2.6.18.3/include/xen/xencons.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/xencons.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/xencons.h b/include/xen/xencons.h
+--- a/include/xen/xencons.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/xencons.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,17 @@
+#ifndef __ASM_XENCONS_H__
+#define __ASM_XENCONS_H__
@@ -92121,16 +86834,16 @@
+void xencons_resume(void);
+
+/* Interrupt work hooks. Receive data, or kick data out. */
-+void xencons_rx(char *buf, unsigned len, struct pt_regs *regs);
++void xencons_rx(char *buf, unsigned len);
+void xencons_tx(void);
+
+int xencons_ring_init(void);
+int xencons_ring_send(const char *data, unsigned len);
+
+#endif /* __ASM_XENCONS_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/include/xen/xen_proc.h linux-2.6.18-xen/include/xen/xen_proc.h
---- linux-2.6.18.3/include/xen/xen_proc.h 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/include/xen/xen_proc.h 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/include/xen/xen_proc.h b/include/xen/xen_proc.h
+--- a/include/xen/xen_proc.h 1970-01-01 01:00:00.000000000 +0100
++++ b/include/xen/xen_proc.h 2007-03-14 10:55:52.000000000 +0100
@@ -0,0 +1,12 @@
+
+#ifndef __ASM_XEN_PROC_H__
@@ -92144,10 +86857,10 @@
+ const char *name);
+
+#endif /* __ASM_XEN_PROC_H__ */
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/kernel/fork.c linux-2.6.18-xen/kernel/fork.c
---- linux-2.6.18.3/kernel/fork.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/kernel/fork.c 2006-11-19 14:27:16.000000000 +0100
-@@ -276,6 +276,9 @@
+diff -x '.hg*' -x '.git*' -urN a/kernel/fork.c b/kernel/fork.c
+--- a/kernel/fork.c 2007-03-16 18:49:45.000000000 +0100
++++ b/kernel/fork.c 2007-03-14 10:55:52.000000000 +0100
+@@ -286,6 +286,9 @@
if (retval)
goto out;
}
@@ -92157,10 +86870,10 @@
retval = 0;
out:
up_write(&mm->mmap_sem);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/kernel/irq/manage.c linux-2.6.18-xen/kernel/irq/manage.c
---- linux-2.6.18.3/kernel/irq/manage.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/kernel/irq/manage.c 2006-12-05 18:42:37.000000000 +0100
-@@ -350,7 +350,6 @@
+diff -x '.hg*' -x '.git*' -urN a/kernel/irq/manage.c b/kernel/irq/manage.c
+--- a/kernel/irq/manage.c 2007-03-16 18:49:45.000000000 +0100
++++ b/kernel/irq/manage.c 2007-03-14 10:55:52.000000000 +0100
+@@ -358,7 +358,6 @@
struct irqaction **p;
unsigned long flags;
@@ -92168,9 +86881,9 @@
if (irq >= NR_IRQS)
return;
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/kernel/Kconfig.preempt linux-2.6.18-xen/kernel/Kconfig.preempt
---- linux-2.6.18.3/kernel/Kconfig.preempt 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/kernel/Kconfig.preempt 2006-11-19 14:27:16.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/kernel/Kconfig.preempt b/kernel/Kconfig.preempt
+--- a/kernel/Kconfig.preempt 2007-03-12 21:58:16.000000000 +0100
++++ b/kernel/Kconfig.preempt 2007-03-14 10:55:52.000000000 +0100
@@ -35,6 +35,7 @@
config PREEMPT
@@ -92179,35 +86892,23 @@
help
This option reduces the latency of the kernel by making
all kernel code (that is not executing in a critical section)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/lib/Kconfig.debug linux-2.6.18-xen/lib/Kconfig.debug
---- linux-2.6.18.3/lib/Kconfig.debug 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/lib/Kconfig.debug 2006-11-19 14:27:16.000000000 +0100
-@@ -325,7 +325,7 @@
-
- config UNWIND_INFO
- bool "Compile the kernel with frame unwind information"
-- depends on !IA64 && !PARISC
-+ depends on !IA64 && !PARISC && !X86_64_XEN
- depends on !MODULES || !(MIPS || PPC || SUPERH || V850)
- help
- If you say Y here the resulting kernel image will be slightly larger
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/lib/Makefile linux-2.6.18-xen/lib/Makefile
---- linux-2.6.18.3/lib/Makefile 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/lib/Makefile 2006-11-19 14:27:16.000000000 +0100
-@@ -52,6 +52,9 @@
- obj-$(CONFIG_AUDIT_GENERIC) += audit.o
+diff -x '.hg*' -x '.git*' -urN a/lib/Makefile b/lib/Makefile
+--- a/lib/Makefile 2007-03-16 18:49:45.000000000 +0100
++++ b/lib/Makefile 2007-03-14 10:55:52.000000000 +0100
+@@ -57,6 +57,9 @@
obj-$(CONFIG_SWIOTLB) += swiotlb.o
+ obj-$(CONFIG_FAULT_INJECTION) += fault-inject.o
+ifneq ($(CONFIG_XEN_IA64_DOM0_NON_VP),y)
+swiotlb-$(CONFIG_XEN) := ../arch/i386/kernel/swiotlb.o
+endif
- hostprogs-y := gen_crc32table
- clean-files := crc32table.h
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/mm/highmem.c linux-2.6.18-xen/mm/highmem.c
---- linux-2.6.18.3/mm/highmem.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/mm/highmem.c 2006-11-19 14:27:16.000000000 +0100
-@@ -142,6 +142,17 @@
+ lib-$(CONFIG_GENERIC_BUG) += bug.o
+
+diff -x '.hg*' -x '.git*' -urN a/mm/highmem.c b/mm/highmem.c
+--- a/mm/highmem.c 2007-03-16 18:49:45.000000000 +0100
++++ b/mm/highmem.c 2007-03-14 10:55:52.000000000 +0100
+@@ -148,6 +148,17 @@
return vaddr;
}
@@ -92225,18 +86926,10 @@
void fastcall *kmap_high(struct page *page)
{
unsigned long vaddr;
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/mm/Kconfig linux-2.6.18-xen/mm/Kconfig
---- linux-2.6.18.3/mm/Kconfig 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/mm/Kconfig 2006-11-19 14:27:16.000000000 +0100
-@@ -116,7 +116,6 @@
- config MEMORY_HOTPLUG
- bool "Allow for memory hot-add"
- depends on SPARSEMEM && HOTPLUG && !SOFTWARE_SUSPEND && ARCH_ENABLE_MEMORY_HOTPLUG
-- depends on (IA64 || X86 || PPC64)
-
- comment "Memory hotplug is currently incompatible with Software Suspend"
- depends on SPARSEMEM && HOTPLUG && SOFTWARE_SUSPEND
-@@ -127,11 +126,14 @@
+diff -x '.hg*' -x '.git*' -urN a/mm/Kconfig b/mm/Kconfig
+--- a/mm/Kconfig 2007-03-16 18:49:45.000000000 +0100
++++ b/mm/Kconfig 2007-03-14 10:55:52.000000000 +0100
+@@ -132,11 +132,14 @@
# 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.
@@ -92251,10 +86944,10 @@
default "4"
#
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/mm/memory.c linux-2.6.18-xen/mm/memory.c
---- linux-2.6.18.3/mm/memory.c 2006-12-06 09:06:09.000000000 +0100
-+++ linux-2.6.18-xen/mm/memory.c 2006-12-05 18:42:37.000000000 +0100
-@@ -403,7 +403,8 @@
+diff -x '.hg*' -x '.git*' -urN a/mm/memory.c b/mm/memory.c
+--- a/mm/memory.c 2007-03-16 18:49:45.000000000 +0100
++++ b/mm/memory.c 2007-03-14 10:55:53.000000000 +0100
+@@ -404,7 +404,8 @@
* and that the resulting page looks ok.
*/
if (unlikely(!pfn_valid(pfn))) {
@@ -92264,7 +86957,7 @@
return NULL;
}
-@@ -891,6 +892,7 @@
+@@ -896,6 +897,7 @@
tlb_finish_mmu(tlb, address, end);
return end;
}
@@ -92272,7 +86965,7 @@
/*
* Do a quick page-table lookup for a single page.
-@@ -1030,6 +1032,26 @@
+@@ -1035,6 +1037,26 @@
continue;
}
@@ -92299,7 +86992,7 @@
if (!vma || (vma->vm_flags & (VM_IO | VM_PFNMAP))
|| !(vm_flags & vma->vm_flags))
return i ? : -EFAULT;
-@@ -1369,6 +1391,102 @@
+@@ -1403,6 +1425,102 @@
}
EXPORT_SYMBOL(remap_pfn_range);
@@ -92402,10 +87095,10 @@
/*
* handle_pte_fault chooses page fault handler according to an entry
* which was read non-atomically. Before making any commitment, on
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/mm/mmap.c linux-2.6.18-xen/mm/mmap.c
---- linux-2.6.18.3/mm/mmap.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/mm/mmap.c 2006-11-19 14:27:16.000000000 +0100
-@@ -1963,6 +1963,10 @@
+diff -x '.hg*' -x '.git*' -urN a/mm/mmap.c b/mm/mmap.c
+--- a/mm/mmap.c 2007-03-16 18:49:45.000000000 +0100
++++ b/mm/mmap.c 2007-03-14 10:55:53.000000000 +0100
+@@ -1977,6 +1977,10 @@
unsigned long nr_accounted = 0;
unsigned long end;
@@ -92416,33 +87109,40 @@
lru_add_drain();
flush_cache_mm(mm);
tlb = tlb_gather_mmu(mm, 1);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/mm/page_alloc.c linux-2.6.18-xen/mm/page_alloc.c
---- linux-2.6.18.3/mm/page_alloc.c 2006-12-06 09:06:09.000000000 +0100
-+++ linux-2.6.18-xen/mm/page_alloc.c 2006-12-05 18:42:37.000000000 +0100
-@@ -443,7 +443,8 @@
+diff -x '.hg*' -x '.git*' -urN a/mm/page_alloc.c b/mm/page_alloc.c
+--- a/mm/page_alloc.c 2007-03-16 18:49:45.000000000 +0100
++++ b/mm/page_alloc.c 2007-03-14 10:55:53.000000000 +0100
+@@ -490,6 +490,8 @@
int i;
int reserved = 0;
-- arch_free_page(page, order);
+ if (arch_free_page(page, order))
+ return;
+ for (i = 0 ; i < (1 << order) ; ++i)
+ reserved += free_pages_check(page + i);
+ if (reserved)
+@@ -497,7 +499,6 @@
+
if (!PageHighMem(page))
- debug_check_no_locks_freed(page_address(page),
- PAGE_SIZE<<order);
-@@ -717,7 +718,8 @@
+ debug_check_no_locks_freed(page_address(page),PAGE_SIZE<<order);
+- arch_free_page(page, order);
+ kernel_map_pages(page, 1 << order, 0);
+
+ local_irq_save(flags);
+@@ -784,6 +785,9 @@
struct per_cpu_pages *pcp;
unsigned long flags;
-- arch_free_page(page, 0);
+ if (arch_free_page(page, 0))
+ return;
-
++
if (PageAnon(page))
page->mapping = NULL;
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/net/core/dev.c linux-2.6.18-xen/net/core/dev.c
---- linux-2.6.18.3/net/core/dev.c 2006-12-06 09:06:10.000000000 +0100
-+++ linux-2.6.18-xen/net/core/dev.c 2006-12-05 18:42:37.000000000 +0100
-@@ -118,6 +118,12 @@
+ if (free_pages_check(page))
+diff -x '.hg*' -x '.git*' -urN a/net/core/dev.c b/net/core/dev.c
+--- a/net/core/dev.c 2007-03-16 18:49:45.000000000 +0100
++++ b/net/core/dev.c 2007-03-14 10:55:53.000000000 +0100
+@@ -117,6 +117,12 @@
#include <linux/err.h>
#include <linux/ctype.h>
@@ -92455,7 +87155,7 @@
/*
* The list of packet types we will receive (as opposed to discard)
* and the routines to invoke.
-@@ -1391,6 +1397,43 @@
+@@ -1392,6 +1398,43 @@
} \
}
@@ -92484,7 +87184,7 @@
+ }
+ if ((skb->h.raw + skb->csum + 2) > skb->tail)
+ goto out;
-+ skb->ip_summed = CHECKSUM_HW;
++ skb->ip_summed = CHECKSUM_PARTIAL;
+ skb->proto_csum_blank = 0;
+ }
+ return 0;
@@ -92499,7 +87199,7 @@
/**
* dev_queue_xmit - transmit a buffer
* @skb: buffer to transmit
-@@ -1423,6 +1466,12 @@
+@@ -1424,6 +1467,12 @@
struct Qdisc *q;
int rc = -ENOMEM;
@@ -92512,7 +87212,7 @@
/* GSO will handle the following emulations directly. */
if (netif_needs_gso(dev, skb))
goto gso;
-@@ -1798,6 +1847,19 @@
+@@ -1799,6 +1848,19 @@
}
#endif
@@ -92521,7 +87221,7 @@
+ case CHECKSUM_UNNECESSARY:
+ skb->proto_data_valid = 1;
+ break;
-+ case CHECKSUM_HW:
++ case CHECKSUM_PARTIAL:
+ /* XXX Implement me. */
+ default:
+ skb->proto_data_valid = 0;
@@ -92532,7 +87232,7 @@
list_for_each_entry_rcu(ptype, &ptype_all, list) {
if (!ptype->dev || ptype->dev == skb->dev) {
if (pt_prev)
-@@ -3584,6 +3646,7 @@
+@@ -3563,6 +3625,7 @@
EXPORT_SYMBOL(net_enable_timestamp);
EXPORT_SYMBOL(net_disable_timestamp);
EXPORT_SYMBOL(dev_get_flags);
@@ -92540,18 +87240,18 @@
#if defined(CONFIG_BRIDGE) || defined(CONFIG_BRIDGE_MODULE)
EXPORT_SYMBOL(br_handle_frame_hook);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/net/core/skbuff.c linux-2.6.18-xen/net/core/skbuff.c
---- linux-2.6.18.3/net/core/skbuff.c 2006-12-06 09:06:10.000000000 +0100
-+++ linux-2.6.18-xen/net/core/skbuff.c 2006-12-05 18:42:37.000000000 +0100
-@@ -139,6 +139,7 @@
+diff -x '.hg*' -x '.git*' -urN a/net/core/skbuff.c b/net/core/skbuff.c
+--- a/net/core/skbuff.c 2007-03-16 18:49:45.000000000 +0100
++++ b/net/core/skbuff.c 2007-03-14 10:55:53.000000000 +0100
+@@ -141,6 +141,7 @@
* Buffers may only be allocated from interrupts using a @gfp_mask of
* %GFP_ATOMIC.
*/
+#ifndef CONFIG_HAVE_ARCH_ALLOC_SKB
struct sk_buff *__alloc_skb(unsigned int size, gfp_t gfp_mask,
- int fclone)
+ int fclone, int node)
{
-@@ -193,6 +194,7 @@
+@@ -196,6 +197,7 @@
skb = NULL;
goto out;
}
@@ -92559,15 +87259,15 @@
/**
* alloc_skb_from_cache - allocate a network buffer
-@@ -210,14 +212,17 @@
+@@ -213,14 +215,17 @@
*/
- struct sk_buff *alloc_skb_from_cache(kmem_cache_t *cp,
+ struct sk_buff *alloc_skb_from_cache(struct kmem_cache *cp,
unsigned int size,
- gfp_t gfp_mask)
+ gfp_t gfp_mask,
+ int fclone)
{
-+ kmem_cache_t *cache;
++ struct kmem_cache *cache;
struct sk_buff *skb;
u8 *data;
@@ -92580,7 +87280,7 @@
if (!skb)
goto out;
-@@ -241,10 +246,20 @@
+@@ -244,10 +249,20 @@
skb_shinfo(skb)->gso_segs = 0;
skb_shinfo(skb)->gso_type = 0;
skb_shinfo(skb)->frag_list = NULL;
@@ -92602,7 +87302,7 @@
skb = NULL;
goto out;
}
-@@ -464,6 +479,10 @@
+@@ -468,6 +483,10 @@
C(local_df);
n->cloned = 1;
n->nohdr = 0;
@@ -92613,43 +87313,48 @@
C(pkt_type);
C(ip_summed);
C(priority);
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/net/ipv4/netfilter/ip_nat_proto_tcp.c linux-2.6.18-xen/net/ipv4/netfilter/ip_nat_proto_tcp.c
---- linux-2.6.18.3/net/ipv4/netfilter/ip_nat_proto_tcp.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/net/ipv4/netfilter/ip_nat_proto_tcp.c 2006-11-19 14:27:18.000000000 +0100
-@@ -129,7 +129,12 @@
+diff -x '.hg*' -x '.git*' -urN a/net/ipv4/netfilter/ip_nat_proto_tcp.c b/net/ipv4/netfilter/ip_nat_proto_tcp.c
+--- a/net/ipv4/netfilter/ip_nat_proto_tcp.c 2007-03-16 18:49:45.000000000 +0100
++++ b/net/ipv4/netfilter/ip_nat_proto_tcp.c 2007-03-14 10:55:54.000000000 +0100
+@@ -129,8 +129,15 @@
if (hdrsize < sizeof(*hdr))
return 1;
-- hdr->check = ip_nat_cheat_check(~oldip, newip,
+#ifdef CONFIG_XEN
+ if ((*pskb)->proto_csum_blank)
-+ hdr->check = ip_nat_cheat_check(oldip, ~newip, hdr->check);
++ nf_csum_replace4(&hdr->check, oldip, newip);
+ else
+#endif
-+ hdr->check = ip_nat_cheat_check(~oldip, newip,
- ip_nat_cheat_check(oldport ^ 0xFFFF,
- newport,
- hdr->check));
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/net/ipv4/netfilter/ip_nat_proto_udp.c linux-2.6.18-xen/net/ipv4/netfilter/ip_nat_proto_udp.c
---- linux-2.6.18.3/net/ipv4/netfilter/ip_nat_proto_udp.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/net/ipv4/netfilter/ip_nat_proto_udp.c 2006-11-19 14:27:18.000000000 +0100
-@@ -114,7 +114,12 @@
- portptr = &hdr->dest;
++ {
+ nf_proto_csum_replace4(&hdr->check, *pskb, oldip, newip, 1);
+ nf_proto_csum_replace2(&hdr->check, *pskb, oldport, newport, 0);
++ }
+ return 1;
+ }
+
+diff -x '.hg*' -x '.git*' -urN a/net/ipv4/netfilter/ip_nat_proto_udp.c b/net/ipv4/netfilter/ip_nat_proto_udp.c
+--- a/net/ipv4/netfilter/ip_nat_proto_udp.c 2007-03-16 18:49:45.000000000 +0100
++++ b/net/ipv4/netfilter/ip_nat_proto_udp.c 2007-03-14 10:55:54.000000000 +0100
+@@ -115,8 +115,16 @@
}
- if (hdr->check) /* 0 is a special case meaning no checksum */
-- hdr->check = ip_nat_cheat_check(~oldip, newip,
+
+ if (hdr->check || (*pskb)->ip_summed == CHECKSUM_PARTIAL) {
+#ifdef CONFIG_XEN
+ if ((*pskb)->proto_csum_blank)
-+ hdr->check = ip_nat_cheat_check(oldip, ~newip, hdr->check);
++ nf_csum_replace4(&hdr->check, oldip, newip);
+ else
+#endif
-+ hdr->check = ip_nat_cheat_check(~oldip, newip,
- ip_nat_cheat_check(*portptr ^ 0xFFFF,
- newport,
- hdr->check));
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/net/ipv4/xfrm4_output.c linux-2.6.18-xen/net/ipv4/xfrm4_output.c
---- linux-2.6.18.3/net/ipv4/xfrm4_output.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/net/ipv4/xfrm4_output.c 2006-11-19 14:27:18.000000000 +0100
++ {
+ nf_proto_csum_replace4(&hdr->check, *pskb, oldip, newip, 1);
+ nf_proto_csum_replace2(&hdr->check, *pskb, *portptr, newport, 0);
++ }
++
+ if (!hdr->check)
+ hdr->check = CSUM_MANGLED_0;
+ }
+diff -x '.hg*' -x '.git*' -urN a/net/ipv4/xfrm4_output.c b/net/ipv4/xfrm4_output.c
+--- a/net/ipv4/xfrm4_output.c 2007-03-16 18:49:45.000000000 +0100
++++ b/net/ipv4/xfrm4_output.c 2007-03-14 10:55:54.000000000 +0100
@@ -18,6 +18,8 @@
#include <net/xfrm.h>
#include <net/icmp.h>
@@ -92667,21 +87372,22 @@
+ if (err)
+ goto error_nolock;
+
- if (skb->ip_summed == CHECKSUM_HW) {
- err = skb_checksum_help(skb, 0);
+ if (skb->ip_summed == CHECKSUM_PARTIAL) {
+ err = skb_checksum_help(skb);
if (err)
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/net/ipv6/addrconf.c linux-2.6.18-xen/net/ipv6/addrconf.c
---- linux-2.6.18.3/net/ipv6/addrconf.c 2006-09-20 05:42:06.000000000 +0200
-+++ linux-2.6.18-xen/net/ipv6/addrconf.c 2006-11-19 14:27:18.000000000 +0100
-@@ -2514,6 +2514,7 @@
+diff -x '.hg*' -x '.git*' -urN a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
+--- a/net/ipv6/addrconf.c 2007-03-16 18:49:45.000000000 +0100
++++ b/net/ipv6/addrconf.c 2007-03-14 10:55:54.000000000 +0100
+@@ -2553,6 +2553,8 @@
spin_lock_bh(&ifp->lock);
if (dev->flags&(IFF_NOARP|IFF_LOOPBACK) ||
++
+ !(dev->flags&IFF_MULTICAST) ||
- !(ifp->flags&IFA_F_TENTATIVE)) {
+ !(ifp->flags&IFA_F_TENTATIVE) ||
+ ifp->flags & IFA_F_NODAD) {
ifp->flags &= ~IFA_F_TENTATIVE;
- spin_unlock_bh(&ifp->lock);
-@@ -2598,6 +2599,7 @@
+@@ -2638,6 +2640,7 @@
if (ifp->idev->cnf.forwarding == 0 &&
ifp->idev->cnf.rtr_solicits > 0 &&
(dev->flags&IFF_LOOPBACK) == 0 &&
@@ -92689,9 +87395,9 @@
(ipv6_addr_type(&ifp->addr) & IPV6_ADDR_LINKLOCAL)) {
struct in6_addr all_routers;
-diff -urN -x .hg -x .hgtags linux-2.6.18.3/scripts/Makefile.xen linux-2.6.18-xen/scripts/Makefile.xen
---- linux-2.6.18.3/scripts/Makefile.xen 1970-01-01 01:00:00.000000000 +0100
-+++ linux-2.6.18-xen/scripts/Makefile.xen 2006-11-19 14:27:19.000000000 +0100
+diff -x '.hg*' -x '.git*' -urN a/scripts/Makefile.xen b/scripts/Makefile.xen
+--- a/scripts/Makefile.xen 1970-01-01 01:00:00.000000000 +0100
++++ b/scripts/Makefile.xen 2007-03-14 10:55:56.000000000 +0100
@@ -0,0 +1,14 @@
+
+# cherrypickxen($1 = allobj)
Added: dists/trunk/linux-2.6/debian/patches/features/all/xen/update.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/all/xen/update.patch Fri Mar 16 19:32:46 2007
@@ -0,0 +1,80 @@
+diff -r 563a16543a96 arch/i386/Kconfig
+--- a/arch/i386/Kconfig Mon Mar 12 14:07:24 2007 +0100
++++ b/arch/i386/Kconfig Fri Mar 16 20:29:31 2007 +0100
+@@ -200,7 +200,7 @@ config PARAVIRT
+ config PARAVIRT
+ bool "Paravirtualization support (EXPERIMENTAL)"
+ depends on EXPERIMENTAL
+- depends on !(X86_VISWS || X86_VOYAGER)
++ depends on !(X86_VISWS || X86_VOYAGER || X86_XEN)
+ help
+ Paravirtualization is a way of running multiple instances of
+ Linux on the same machine, under a hypervisor. This option
+diff -r 563a16543a96 arch/i386/kernel/microcode-xen.c
+--- a/arch/i386/kernel/microcode-xen.c Mon Mar 12 14:07:24 2007 +0100
++++ b/arch/i386/kernel/microcode-xen.c Fri Mar 16 20:15:51 2007 +0100
+@@ -52,30 +52,32 @@ MODULE_LICENSE("GPL");
+ /* no concurrent ->write()s are allowed on /dev/cpu/microcode */
+ static DEFINE_MUTEX(microcode_mutex);
+
+-static void __user *user_buffer; /* user area microcode data buffer */
+-static unsigned int user_buffer_size; /* it's size */
+-
+ static int microcode_open (struct inode *unused1, struct file *unused2)
+ {
+ return capable(CAP_SYS_RAWIO) ? 0 : -EPERM;
+ }
+
+
+-static int do_microcode_update (void)
++static int do_microcode_update (const void __user *ubuf, size_t len)
+ {
+ int err;
+- dom0_op_t op;
++ void *kbuf;
+
+- err = sys_mlock((unsigned long)user_buffer, user_buffer_size);
+- if (err != 0)
+- return err;
++ kbuf = vmalloc(len);
++ if (!kbuf)
++ return -ENOMEM;
+
+- op.cmd = DOM0_MICROCODE;
+- set_xen_guest_handle(op.u.microcode.data, user_buffer);
+- op.u.microcode.length = user_buffer_size;
+- err = HYPERVISOR_dom0_op(&op);
++ if (copy_from_user(kbuf, ubuf, len) == 0) {
++ dom0_op_t op;
+
+- (void)sys_munlock((unsigned long)user_buffer, user_buffer_size);
++ op.cmd = DOM0_MICROCODE;
++ set_xen_guest_handle(op.u.microcode.data, kbuf);
++ op.u.microcode.length = len;
++ err = HYPERVISOR_dom0_op(&op);
++ } else
++ err = -EFAULT;
++
++ vfree(kbuf);
+
+ return err;
+ }
+@@ -89,17 +91,9 @@ static ssize_t microcode_write (struct f
+ return -EINVAL;
+ }
+
+- if ((len >> PAGE_SHIFT) > num_physpages) {
+- printk(KERN_ERR "microcode: too much data (max %ld pages)\n", num_physpages);
+- return -EINVAL;
+- }
+-
+ mutex_lock(µcode_mutex);
+
+- user_buffer = (void __user *) buf;
+- user_buffer_size = (int) len;
+-
+- ret = do_microcode_update();
++ ret = do_microcode_update(buf, len);
+ if (!ret)
+ ret = (ssize_t)len;
+
Modified: dists/trunk/linux-2.6/debian/patches/series/1~experimental.1-extra
==============================================================================
--- dists/trunk/linux-2.6/debian/patches/series/1~experimental.1-extra (original)
+++ dists/trunk/linux-2.6/debian/patches/series/1~experimental.1-extra Fri Mar 16 19:32:46 2007
@@ -1,6 +1,7 @@
+ features/all/vserver/vs2.2.0-rc15.patch *_vserver *_xen-vserver
+ features/all/vserver/bindmount-dev.patch *_vserver *_xen-vserver
+ features/all/xen/vserver-clash.patch *_xen-vserver
-#+ features/all/xen/fedora-36252.patch *_xen *_xen-vserver
++ features/all/xen/fedora-2.6.20-48670.patch *_xen *_xen-vserver
++ features/all/xen/update.patch *_xen *_xen-vserver
+ features/all/xen/vserver-update.patch *_xen-vserver
+ bugfix/arm/nas100d-artop-temp-fix.patch arm
More information about the Kernel-svn-changes
mailing list